Thermodynamics & Chemical Equilibrium (P. Ellgen)

THERMODYNAMICS
AND CHEMICAL
EQUILIBRIUM
Paul Ellgen
Oklahoma School of Science Mathematics
Oklahoma School of Science Mathematics
Thermodynamics and Chemical Equilibrium
Paul Ellgen
This text is disseminated via the Open Education Resource (OER) LibreTexts Project (https://LibreTexts.org) and like the hundreds
of other texts available within this powerful platform, it is freely available for reading, printing and "consuming." Most, but not all,
pages in the library have licenses that may allow individuals to make changes, save, and print this book. Carefully
consult the applicable license(s) before pursuing such effects.
Instructors can adopt existing LibreTexts texts or Remix them to quickly build course-specific resources to meet the needs of their
students. Unlike traditional textbooks, LibreTexts’ web based origins allow powerful integration of advanced features and new
technologies to support learning.
The LibreTexts mission is to unite students, faculty and scholars in a cooperative effort to develop an easy-to-use online platform
for the construction, customization, and dissemination of OER content to reduce the burdens of unreasonable textbook costs to our
students and society. The LibreTexts project is a multi-institutional collaborative venture to develop the next generation of open-
access texts to improve postsecondary education at all levels of higher learning by developing an Open Access Resource
environment. The project currently consists of 14 independently operating and interconnected libraries that are constantly being
optimized by students, faculty, and outside experts to supplant conventional paper-based books. These free textbook alternatives are
organized within a central environment that is both vertically (from advance to basic level) and horizontally (across different fields)
integrated.
The LibreTexts libraries are Powered by NICE CXOne and are supported by the Department of Education Open Textbook Pilot
Project, the UC Davis Office of the Provost, the UC Davis Library, the California State University Affordable Learning Solutions
Program, and Merlot. This material is based upon work supported by the National Science Foundation under Grant No. 1246120,
1525057, and 1413739.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not
necessarily reflect the views of the National Science Foundation nor the US Department of Education.
Have questions or comments? For information about adoptions or adaptions contact [email protected]. More information on our
activities can be found via Facebook (https://facebook.com/Libretexts), Twitter (https://twitter.com/libretexts), or our blog
(http://Blog.Libretexts.org).
This text was compiled on 01/10/2024
TABLE OF CONTENTS
Licensing
About the Author
1: Introduction - Background and a Look Ahead

1.1: The Role of the Ideal Gas
1.2: Chemical Kinetics
1.3: Classical Thermodynamics
1.4: Statistical Thermodynamics
1.5: Heat Transfer in Practical Devices
1.6: The Concept of Equilibrium
1.7: Chemical Equilibrium and Predicting Chemical Change
1.8: Equilibrium and Classical Thermodynamics
1.9: A Few Ideas from the Philosophy of Science
1.10: A Few Ideas from Formal Logic
1.11: Problems
2: Gas Laws
2.1: Boyle's Law
2.2: Charles' Law
2.3: Avogadro's Hypothesis
2.4: Finding Avogadro's Number
2.5: The Kelvin Temperature Scale
2.6: Deriving the Ideal Gas Law from Boyle's and Charles' Laws
2.7: The Ideal Gas Constant and Boltzmann's Constant
2.8: Real Gases Versus Ideal Gases
2.9: Temperature and the Ideal Gas Thermometer
2.10: Deriving Boyle's Law from Newtonian Mechanics
2.11: The Barometric Formula
2.12: Van der Waals' Equation
2.13: Virial Equations
2.14: Gas Mixtures - Dalton's Law of Partial Pressures
2.15: Gas Mixtures - Amagat's Law of Partial Volums
2.16: Problems
3: Distributions, Probability, and Expected Values

3.1: The Distribution Function as a Summary of Experimental Results
3.2: Outcomes, Events, and Probability
3.3: Some Important Properties of Events
3.4: Applying the Laws of Probability
3.5: Bar Graphs and Histograms
3.6: Continuous Distribution Functions - the Envelope Function is the Derivative of the Area
3.7: A Heuristic View of the Probability Density Function
3.8: A Heuristic View of the Cumulative Distribution Function
3.9: Random Variables, Expected Values, and Population Sets
3.10: Statistics - the Mean and the Variance of a Distribution
3.11: The Variance of the Average- The Central Limit Theorem
1 https://chem.libretexts.org/@go/page/191336
3.12: The Normal Distribution
3.13: The Expected Value of a Function of Several Variables and the Central Limit Theorem
3.14: Where Does the N - 1 Come from?
3.15: Problems
4: The Distribution of Gas Velocities

4.1: Distribution Functions for Gas-velocity Components
4.2: Probability Density Functions for Velocity Components in Spherical Coordinates
4.3: Maxwell's Derivation of the Gas-velocity Probability-density Function
4.4: The Probability-density Function for Gas Velocities in One Dimension
4.5: Combining the One-dimensional Probability Density Functions
4.6: Boyle's Law from the Maxwell-Boltzmann Probability Density
4.7: Experimental Test of the Maxwell-Boltzmann Probability Density
4.8: Statistics for Molecular Speeds
4.9: Pressure Variations for Macroscopic Samples
4.10: Collisions between Gas Molecules Relative Velocity Coordinates
4.11: The Probability Density Function for the Relative Velocity
4.12: The Frequency of Collisions between Unlike Gas Molecules
4.13: The Rate of Collisions between Unlike Gas Molecules
4.14: Collisions between like Gas Molecules
4.15: The Geometry of A Collision between Spherical Molecules
4.16: The Energy of A Collision between Gas Molecules
4.17: Problems
5: Chemical Kinetics, Reaction Mechanisms, and Chemical Equilibrium

5.2: Reaction Rates and Rate Laws
5.3: Simultaneous Processes
5.4: The Effect of Temperature on Reaction Rates
5.5: Other Factors that Affect Reaction Rates
5.6: Mechanisms and Elementary Processes
5.7: Rate Laws for Elementary Processes
5.8: Experimental Determination of Rate Laws
5.9: First-order Rate Processes
5.10: Rate Laws by the Study of Initial Rates
5.11: Rate Laws from Experiments in a Continuous Stirred Tank Reactor
5.12: Predicting Rate Laws from Proposed Mechanisms
5.13: The Michaelis-Menten Mechanism for Enzyme-catalyzed Reactions
5.14: The Lindemann-Hinshelwood Mechanism for First-order Decay
5.15: Why Unimolecular Reactions are First Order
5.16: The Mechanism of the Base Hydrolysis of Co(NH₃)₅Xⁿ⁺
5.17: Chemical Equilibrium as the Equality of Rates for Opposing Reactions
5.18: The Principle of Microscopic Reversibility
5.19: Microscopic Reversibility and the Second Law
5.20: Problems
6: Equilibrium States and Reversible Processes

6.1: The Thermodynamic Perspective
6.2: Thermodynamic Systems and Variables
6.3: Equilibrium and Reversibility - Phase Equilibria
6.4: Distribution Equilibria
6.5: Equilibria in Chemical Reactions
6.6: Le Chatelier's Principle
6.7: The Number of Variables Required to Specify Some Familiar Systems
6.8: Gibbs' Phase Rule
6.9: Reversible vs. Irreversible Processes
6.10: Duhem's Theorem - Specifying Reversible Change in A Closed System
6.11: Reversible Motion of A Mass in A Constant Gravitational Field
6.12: Equilibria and Reversible Processes
6.13: The Laws of Thermodynamics
6.14: Thermodynamic Criteria for Change
6.15: State Functions in Systems Undergoing Spontaneous Change
6.16: Problems
7: State Functions and The First Law

7.1: Changes in a State Function are Independent of Path
7.2: The Total Differential
7.3: Line Integrals
7.4: Exact Differentials and State Functions
7.5: Determining Whether an Expression is an Exact Differential
7.6: The Chain Rule and the Divide-through Rule
7.7: Measuring Pressure-Volume Work
7.8: Measuring Work- Non-Pressure-Volume Work
7.9: Measuring Heat
7.10: The First Law of Thermodynamics
7.11: Other Statements of the First Law
7.12: Notation for Changes in Thermodynamic Quantities - E vs. ∆E
7.13: Heat Capacities for Gases- Cv, Cp
7.14: Heat Capacities of Solids- the Law of Dulong and Petit
7.15: Defining Enthalpy, H
7.16: Heat Transfer in Reversible Processes
7.17: Free Expansion of a Gas
7.18: Reversible vs. Irreversible Pressure-Volume Work
7.19: Isothermal Expansions of An Ideal Gas
7.20: Adiabatic Expansions of An Ideal Gas
7.21: Problems
8: Enthalpy and Thermochemical Cycles

8.1: Enthalpy
8.2: Using Thermochemical Cycles to Find Enthalpy Changes
8.3: How Enthalpy Depends on Pressure
8.4: Standard States and Enthalpies of Formation
8.5: The Ideal Gas Standard State
8.6: Standard Enthalpies of Reaction
8.7: Standard State Heat Capacities
8.8: How The Enthalpy Change for a Reaction Depends on Temperature
8.9: Calorimetry
8.10: Problems
9: The Second Law - Entropy and Spontaneous Change

9.1: The Second Law of Thermodynamics
9.2: The Carnot Cycle for an Ideal Gas and the Entropy Concept
9.3: The Carnot Cycle for Any Reversible System
9.4: The Entropy Change around Any Cycle for Any Reversible System
9.5: The Tiling Theorem and the Paths of Cyclic Process in Other Spaces
9.6: Entropy Changes for A Reversible Process
9.7: Entropy Changes for A Spontaneous Process in An Isolated System
9.8: The Entropy of the Universe
9.9: The Significance of The Machine-based Statement of The Second Law
9.10: A Slightly Philosophical Digression on Energy and Entropy
9.11: A Third Statement of the Second Law
9.12: Entropy and Predicting Change
9.13: Defining the Helmholtz and Gibbs Free Energies
9.14: The Fundamental Equation and Other Criteria for Reversible Change
9.15: Entropy and Spontaneous Change
9.16: Internal Entropy and the Second Law
9.17: Notation and Terminology- Conventions for Spontaneous Processes
9.18: The Heat Exchanged by A Spontaneous Process at Constant Entropy
9.19: The Energy Change for A Spontaneous Process at Constant S and V
9.20: The Enthalpy Change for A Spontaneous Process at Constant S and P
9.21: The Entropy Change for A Spontaneous Process at Constant E and V
9.22: The Entropy Change for A Spontaneous Process at Constant H and P
9.23: The Reversible Work is the Minimum Work at Constant Tˆ
9.24: The Free Energy Changes for A Spontaneous Process at Constant Tˆ
9.25: Summary- Thermodynamic Functions as Criteria for Change
9.26: Problems
10: Some Mathematical Consequences of the Fundamental Equation

10.1: Thermodynamic Relationships from dE, dH, dA and dG
10.2: dE = TdS - PdV and Internal consistency
10.3: Expressing Thermodynamic Functions with Other Independent Variables
10.4: Expressing Thermodynamic Functions with Independent Variables V and T
10.5: Expressing Thermodynamic Functions with Independent Variables P and T
10.6: The Transformation of Thermodynamic Variables in General
10.7: Reversibility and Thermodynamic Variables in Gerneral
10.8: Using the Pair (V, P) or the Pair (T, S) as Independent Variables
10.9: The Relationship Between Cv and Cp for Any Substance
10.10: The Dependence of Cv on Volume and of Cp on Pressure
10.11: The Gibbs-Helmholtz Equation
10.12: The Second Law and the Properties of Ideal Gases
10.13: The second-dependence of the Energy and Enthalpy of A Real Gas
10.14: The Joule-Thomson Effect
10.15: Problems
11: The Third Law, Absolute Entropy, and the Gibbs Free Energy of
Formation
11.1: Heat Capacity as a Function of Temperature
11.2: Enthalpy as a function of Temperature
11.3: The Third Law
11.4: Genesis of the Third Law - the Nernst Heat Theorem
11.5: Absolute Entropy
11.6: The Standard State for Third-law Entropies
11.7: The Fugacity of a Gas
11.8: A General Strategy for Expressing the Thermodynamic Properties of a Substance
11.9: The Standard Entropy and the Gibbs Free Energy of Formation
11.10: The Nature of Hypothetical States
11.11: The Fugacity and Gibbs Free Energy of A Substance in Any System
11.12: Evaluating Entropy Changes Using Thermochemical Cycles
11.13: Absolute Zero is Unattainable
11.14: Problems
12: Applications of the Thermodynamic Criteria for Change

12.1: Mechanical Processes
12.2: The Direction of Spontaneous Heat Transfer
12.3: Phase Changes - the Fusion of Ice
12.4: Measuring the Entropy Change for Any Reversible Process
12.5: Another Perspective on the Principle of Le Chatelier
12.6: Phase Equilibria - Temperature Dependence of the Boiling Point
12.7: Phase Equilibria - Temperature Dependence of the Melting Point
12.8: The Clapeyron Equation
12.9: The Clausius-Clapeyron Equation
12.10: Problems
13: Equilibria in Reactions of Ideal Gases

13.1: The Gibbs Free Energy of an Ideal Gas
13.2: The Gibbs Free Energy Change for A Reaction of Ideal Gases
13.3: The Thermodynamics of Mixing Ideal Gases
13.4: The Gibbs Free Energy Change for Reaction at Constant Partial Pressures
13.5: ∆rG is the rate at which the Gibbs Free Energy Changes with The Extent of Reaction
13.6: The Standard Gibbs Free Energy Change and Equilibrium in Ideal Gas Reactions
13.7: The Gibbs Free Energy of Formation and Equilibrium in Ideal Gas Reactions
13.8: Equilibrium When A Component is Also Present as A Condensed Phase
13.9: Equilibrium When An Ideal Gas Component is Also Present as A solute
13.10: Problems
14: Chemical Potential - Extending the Scope of the Fundamental

Equation
14.1: Dependence of the Internal Energy on the Composition of the System
14.2: Dependence of Other Thermodynamic Functions on the Composition
14.3: Partial Molar Quantities
14.4: Chemical Potentials and Stoichiometry
14.5: ∑µjdnj = 0 and Primitive Vs. Gibbsian Equilibrium
14.6: The Change Criteria in A System Composed of Subsystems
14.7: At Constant P and T, ∆rµ is the Change in Gibbs Free Energy
14.8: Gibbs-Duhem Equation
14.9: The Dependence of Chemical Potential on Other Variables
14.10: Chemical Activity
14.11: Back to the Fugacity- the Fugacity of A Component of A Gas Mixture
14.12: Relating Fugacity and Chemical Activity
14.13: Relating the Differentials of Chemical Potential and Activity
14.14: Dependence of Activity on Temperature- Relative Partial Molar Enthalpies
14.15: Problems
15: Chemical Potential, Fugacity, Activity, and Equilibrium
15.1: The Chemical Potential and Fugacity of a Gas
15.2: The Chemical Potential and Activity of a Gas
15.3: The Pressure-dependence of the Fugacity and Activity of a Condensed Phase
15.4: Standard States for the Fugacity and Activity of a Pure Solid
15.5: The Chemical Potential, Fugacity, and Activity of a Pure Solid
15.6: Chemical Potential, Fugacity, and Equilibrium
15.7: Chemical Potential, Activity, and Equilibrium
15.8: The Rate of Gibbs Free Energy Change with Extent of Reaction
15.9: Problems
16: The Chemical Activity of the Components of a Solution

16.1: Solutions Whose Components are in Equilibrium with Their Own Gases
16.2: Raoult's Law and Ideal Solutions
16.3: Expressing the Activity Coefficient as a Deviation from Raoult's Law
16.4: Henry's Law and the Fugacity and Activity of A Solution Component
16.5: Expressing the Activity Coefficient as A Deviation from Henry's Law
16.6: Henry's Law and the Hypothetical One-molal Standard State
16.7: Finding the Activity of a Solute from the Activity of the Solvent
16.8: When the Solute Obeys Henry's Law, the Solvent Obeys Raoult's Law
16.9: Properties of Ideal Solutions
16.10: Colligative Properties - Boiling-point Elevation
16.11: Colligative Properties - Freezing-point Depression
16.12: Colligative Properties - Osmotic Pressure
16.13: Colligative Properties - Solubility of a Solute in an Ideal Solution
16.14: Colligative Properties - Solubility of a Gas
16.15: Solvent Activity Coefficients from Freezing-point Depression Measurements
16.16: Electrolytic Solutions
16.17: Activities of Electrolytes - The Mean Activity Coefficient
16.18: Activities of Electrolytes - The Debye-Hückel Theory
16.19: Finding Solute Activity Using the Hypothetical One-molal Standard State
16.20: Problems
17: Electrochemistry
17.1: Oxidation-reduction Reactions
17.2: Electrochemical Cells
17.3: Defining Oxidation States
17.4: Balancing Oxidation-reduction Reactions
17.5: Electrical Potential
17.6: Electrochemical Cells as Circuit Elements
17.7: The Direction of Electron Flow and its Implications
17.8: Electrolysis and the Faraday
17.9: Electrochemistry and Conductivity
17.10: The Standard Hydrogen Electrode (S.H.E)
17.11: Half-reactions and Half-cells
17.12: Standard Electrode Potentials
17.13: Predicting the Direction of Spontaneous Change
17.14: Cell Potentials and the Gibbs Free Energy
17.15: The Nernst Equation
17.16: The Nernst Equation for Half-cells
17.17: Combining two Half-cell Equations to Obtain a new Half-cell Equation
17.18: The Nernst Equation and the Criterion for Equilibrium
17.19: Problems
18: Quantum Mechanics and Molecular Energy Levels

18.1: Energy Distributions and Energy Levels
18.2: Quantized Energy - De Broglie's Hypothesis and the Schroedinger Equation
18.3: The Schrödinger Equation for A Particle in A Box
18.4: The Schrödinger Equation for a Molecule
18.5: Solutions to Schroedinger Equations for Harmonic Oscillators and Rigid Rotors
18.6: Wave Functions, Quantum States, Energy Levels, and Degeneracies
18.7: Particle Spins and Statistics- Bose-Einstein and Fermi-Dirac Statistics
19: The Distribution of Outcomes for Multiple Trials

19.1: Distribution of Results for Multiple Trials with Two Possible Outcomes
19.2: Distribution of Results for Multiple Trials with Three Possible Outcomes
19.3: Distribution of Results for Multiple Trials with Many Possible Outcomes
19.4: Stirling's Approximation
19.5: Problems
20: Boltzmann Statistics

20.1: The Independent-Molecule Approximation
20.2: The Probability of An Energy Level at Constant N, V, and T
20.3: The Population Sets of a System at Equilibrium at Constant N, V, and T
20.4: How can Infinitely Many Probabilities Sum to Unity?
20.5: The Total Probability Sum at Constant N, V, and T
20.6: The Most Probable Population Set at Constant N, V, and T
20.7: The Microstates of a Given Population Set
20.8: The Probabilities of Microstates that Have the Same Energy
20.9: The Probabilities of the Population Sets of an Isolated System
20.10: Entropy and Equilibrium in an Isolated System
20.11: Thermodynamic Probability and Equilibrium in an Isomerization Reaction
20.12: The Degeneracy of an Isolated System and Its Entropy
20.13: The Degeneracy of an Isolated System and its Entropy
20.14: Effective Equivalence of the Isothermal and Constant-energy Conditions
20.15: Problems
21: The Boltzmann Distribution Function

21.1: Finding the Boltzmann Equation
21.2: Lagrange's Method of Undetermined Multipliers
21.3: Deriving the Boltzmann Equation I
21.4: Deriving the Boltzmann Equation II
21.5: Partition Functions and Equilibrium - Isomeric Molecules
21.6: Finding ß and the Thermodynamic Functions for Distinguishable Molecules
21.7: The Microscopic Model for Reversible Change
21.8: The Third Law of Thermodynamics
21.9: The Partition Function for a System of N Molecules
21.10: Problems
22: Some Basic Applications of Statistical Thermodynamics
22.1: Interpreting the Partition Function
22.2: Conditions under which Integrals Approximate Partition Functions
22.3: Probability Density Functions from the Energies of Classical-mechanical Models
22.4: Partition Functions and Average Energies at High Temperatures
22.5: Energy Levels for a Three-dimensional Harmonic Oscillator
22.6: Energy and Heat Capacity of the "Einstein Crystal"
22.7: Applications of Other Entropy Relationships
22.8: Problems
23: The Ensemble Treatment

23.1: Ensembles of N-molecule Systems
23.2: The Ensemble Entropy and the Value of ß
23.3: The Thermodynamic Functions of the N-molecule System
24: Indistinguishable Molecules - Statistical Thermodynamics of Ideal

Gases
24.1: The Partition Function for N Distinguishable, Non-interacting Molecules
24.2: The Partition Function for N Indistinguishable, Non-interacting Molecules
24.3: Occupancy Probabilities for Translational Energy Levels
24.4: The Separable-modes molecular Model
24.5: The Partition Function for A Gas of Indistinguishable, Non-interacting, Separable-modes Molecules
24.6: The Translational Partition Function of An Ideal Gas
24.7: The Electronic Partition Function of an Ideal Gas
24.8: The Vibrational Partition Function of A Diatomic Ideal Gas
24.9: The Rotational Partition Function of A Diatomic Ideal Gas
24.10: The Gibbs Free Energy for One Mole of An Ideal Gas
24.11: The Standard Gibbs Free Energy for H₂(g), I₂(g), and HI(g)
24.12: The Gibbs Free Energy Change for Forming HI(g) from H₂(g) and I₂(g)
24.13: The Reference State for Molecular Partition Functions
24.14: Problems
25: Bose-Einstein and Fermi-Dirac Statistics

25.1: Quantum Statistics
25.2: Fermi-Dirac Statistics and the Fermi-Dirac Distribution Function
25.3: Bose-Einstein Statistics and the Bose-Einstein Distribution Function
26: Appendices
26.1: Appendix A. Standard Atomic Weights 1999†
26.2: Appendix B.
26.3: Appendix B. Fundamental Constants†
26.4: Appendix C.
26.5: Units and Conversion Factors
26.6: Appendix D.
26.7: Appendix D. Some Important Definite Integrals
Index
Index
Detailed Licensing
Licensing
A detailed breakdown of this resource's licensing can be found in Back Matter/Detailed Licensing.
About the Author
Following undergraduate work at Carleton College, the author earned a Ph.D. in physical chemistry at Northwestern University.
He taught for several years at the University of California at Riverside, and then spent many years in industrial research and
industrial research management. He has authored 16 technical publications and is the inventor on 15 U. S. Patents. Following his
retirement, he presented a college-level physical chemistry course for several years at The Oklahoma School of Science and
Mathematics. This book arose from his perception that many of the obstacles to understanding the subject can be reduced by
increased attention to the background material that is included in the initial presentation. Accordingly, this text sacrifices breadth
in order to emphasize basic concepts and clarify them by treating them from alternative perspectives.
CHAPTER OVERVIEW
1: Introduction - Background and a Look Ahead

Physical chemistry encompasses a wide variety of ideas that are intimately linked. For the most part, we cannot understand one
without having some understanding of many others. We overcome this problem by looking at the same idea from a series of
slightly different and increasingly sophisticated perspectives. This book focuses on the theories of physical chemistry that describe
and make predictions about chemical equilibrium. We omit many topics that are usually understood to be included in the subject of
physical chemistry. In particular, we treat quantum mechanics only briefly and spectroscopy not at all.
The goals of chemistry are to predict molecular structures and chemical reactivity. For a given empirical formula, we want to be
able to predict all of the stable, three-dimensional atomic arrangements that can exist. We also want to be able to predict all of the
reactions in which each such molecule can participate. And, while we are drawing up our wish list, we want to be able to predict
how fast each reaction goes at any set of reaction conditions that may happen to be of interest.
Many different kinds of theories enable us to make useful predictions about chemical reactivity. Sometimes we are able to make
predictions based on detailed quantum mechanical calculations. When thermodynamic data are available, we can make precise
predictions about the extent to which a particular reaction can occur. Lacking such data about a compound, but given its structure,
we can usually make some worthwhile predictions based on generalizations (models) that reflect our accumulated experience with
particular classes of compounds and their known reactions. Usually these predictions are qualitative, and they fail to include many
noteworthy features that emerge when experimental studies are made.
Physical chemistry is the general theory of the properties of chemical substances, with chemical reactivity being a pivotally
important property. This book focuses on the core ideas in the subjects of chemical kinetics, chemical thermodynamics, and
statistical thermodynamics. These ideas apply to the characterization, correlation, and prediction of the extent to which any
chemical reaction can occur. Because predicting how a system will react is substantially equivalent to predicting its equilibrium
position, we direct our efforts to understanding how each of these subjects contributes to our understanding of the equilibrium
processes that are important in chemistry. In this chapter, we review the general characteristics of these subjects.
The study of chemical kinetics gives us one way to think about chemical equilibrium that is simple and direct. Classical chemical
thermodynamics gives us a way to predict chemical equilibria for a wide range of reactions from experimental observations made
under a much smaller range of conditions. That is, once we have measured the thermodynamic functions that characterize a
compound, we can use these values to predict the behavior of the compound in a wide variety of reactions. Statistical mechanics
gives us a conceptual basis for understanding why the laws of chemical thermodynamics take the form that they do. It also provides
a way to obtain accurate values for the thermodynamic properties of many compounds.
1.11: Problems
This page titled 1: Introduction - Background and a Look Ahead is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available
upon request.
1
The concept of ideal gas behavior plays a pivotal role in the development of science and particularly in the development of
thermodynamics. As we shall emphasize, intermolecular forces do not influence the behavior of an ideal gas. Ideal gas molecules
are neither attracted to one another nor repelled by one another. For this reason, the properties of an ideal gas are particularly
simple. Because ideal gas behavior is so important, we begin by studying ideal gases from both an experimental and a theoretical
perspective.
In Chapter 2, we review the experimental observations that we can make on gases and the idealizations that we introduce to
extrapolate the behavior of ideal gases from the observations we make on real gases. We also develop Boyle’s law from a very
simple model for the interactions between point-mass gas molecules and the walls of their container.
In Chapter 4, we develop a detailed model for the behavior of an ideal gas. The physical model is the one we use in Chapter 2, but
the mathematical treatment is much more sophisticated. For this treatment we need to develop a number of ideas about probability,
distribution functions, and statistics. Chapter 3 introduces these topics, all of which again play important roles when we turn to the
development of statistical thermodynamics in Chapter 19.
This page titled 1.1: The Role of the Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.
1.1.1 https://chem.libretexts.org/@go/page/151804
Chemical kinetics is the study of how fast chemical reactions occur. In Chapter 5, we see that there is a unique way to specify what
we mean by “how fast.” We call this specification the reaction rate. Chemical kinetics is the study of the factors that determine the
rate of a particular reaction. There are many such factors, among them:
temperature
pressure
concentrations of the reactants and products
nature and concentrations of “spectator species” like a solvent or dissolved salts
isotopic substitution
presence or absence of a catalyst.
We will look briefly at all of these, but the thrust of our development will be to understand how the rate of a reaction depends on
the concentrations of the reaction’s reactants and products.
Many reactions that we observe actually occur as a sequence of more simple reactions. Such a sequence of simple reaction steps is
called a reaction mechanism. Our principal goal is to understand the relationships among concentrations, reaction rates, reaction
mechanisms, and the conditions that must be satisfied when a particular reaction reaches equilibrium. We will find that two related
ideas characterize equilibrium from a reaction-rate perspective. One is that concentrations no longer change with time. The other is
a fundamental postulate, called the principle of microscopic reversibility, about the relative rates of individual steps in an overall
chemical reaction mechanism when the reacting system is at equilibrium.
This page titled 1.2: Chemical Kinetics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.
One goal of chemical thermodynamics is to predict whether a particular chemical reaction can occur. We say can, not will, because
chemical thermodynamics is unable to make predictions about reaction rates. If we learn from our study of chemical
thermodynamics that a particular reaction can occur, we still do not know whether it will occur in a millisecond—or so slowly that
no change is detectable. The science of thermodynamics builds on the idea that a particular chemical system can be characterized
by the values of certain thermodynamic functions. These state functions include such familiar quantities as pressure, temperature,
volume, concentrations, and energy, as well as some that are not so well known, notably enthalpy, entropy, Gibbs free energy,
Helmholtz free energy, chemical potential, fugacity, and chemical activity. We can think of a state function as a quasi-
mathematical function whose argument is a physical system. That is, a state function maps a real system onto a real number. When
we insert a thermometer into a mixture, the measurement that we make maps the state of the mixture onto the real number that we
call temperature.
The word “thermodynamics” joins roots that convey the ideas of heat and motion. In general, motion involves kinetic energy and
mechanical work. The interconversion of heat and mechanical work is the core concern of the science of thermodynamics. We are
familiar with the idea that kinetic energy can be converted into work; given a suitable arrangement of ropes and pulleys, a falling
object can be used to lift another object. Kinetic energy can also be converted—or, as we often say, degraded—into heat by the
effects of friction. We view such processes as the conversion of the kinetic energy of a large object into increased kinetic energy of
the atoms and molecules that comprise the warmed objects. We can say that easily visible mechanical motions are converted into
invisible mechanical motions. The idea that heating an object increases the kinetic energy of its component atoms is called the
kinetic theory of heat.
It is often convenient to use the term microscopic process to refer to an event that occurs at the atomic or molecular level. We call a
process that occurs on a larger scale a macroscopic process, although the usual connotation is that a macroscopic process is
observable in a quantity of bulk matter. When friction causes the degradation of macroscopic motion to heat, we can say that
macroscopic motion is converted to microscopic motion.
While this terminology is convenient, it is not very precise. Changes visible under an optical microscope are macroscopic
processes. Of course, all macroscopic changes are ultimately attributable to an accumulation of molecular-level processes. The
Brownian motion of a colloidal particle suspended in a liquid medium is noteworthy because this relationship is visible. Viewed
with an optical microscope, a suspended, macroscopic, colloidal particle is seen to undergo a rapid and random jiggling motion.
Each jiggle is the accumulated effect of otherwise invisible collisions between the particle and the molecules of the liquid. Each
collision imparts momentum to the particle. Over long times, the effects average out; momentum transfer is approximately equal in
every direction. During the short time of a given jiggle, there is an imbalance of collisions such that more momentum is transferred
to the particle in the direction of the jiggle than in any other.
We are also familiar with the idea that heat can be converted into mechanical motion. In an earlier era, steam engines were the
dominant means by which heat was converted to work. Steam turbines remain important in large stationary facilities like power
plants. For applications we encounter in daily life, the steam engine has been replaced by the internal-combustion engine. When we
want to create mechanical motion (do work) with a heat engine, it is important to know how much heat we need in order to produce
a given quantity of work. Sadi Carnot was the first to analyze this problem theoretically. In doing so, he discovered the idea that we
call the second law of thermodynamics.
The interconversion of heat and work involves an important asymmetry. We readily appreciate that the conversion of kinetic energy
to heat can be complete, because we have seen countless examples of objects coming to a complete standstill as the result of
frictional forces. However, ordinary experience leaves us less prepared to deal with the question of whether heat can be completely
converted into work. Possibly, we remember hearing that it cannot be done and that the reason has something to do with the second
law of thermodynamics. If we have heard more of the story, we may remember that it is slightly more complicated. Under idealized
circumstances, heat can be converted into work completely. If we confine an ideal gas in a frictionless piston and arrange to add
heat to the gas while increasing the volume of the piston in a coordinated way, such that the temperature of the gas remains
constant, the expanding piston will do work on some external entity, and the amount of this work will be just equal to the thermal
energy added to the gas. We call this process a reversible isothermal expansion. This process does not involve a cycle; the volume
of the gas at the end of the process is greater than its volume at the start.
What Carnot realized is that an engine must operate in a cyclic fashion, and that no device—not even an idealized frictionless
device—operating around a cycle can convert heat to work with 100% efficiency. Carnot analyzed the process of converting heat
into work in terms of an ideal engine that accepts thermal energy (heat) at a high temperature, uses some of this thermal energy to
do work on its surroundings, and rejects the rest of its thermal-energy intake to the surroundings in the form of thermal energy at a
lower temperature. Carnot’s analysis preceded the development of our current ideas about the nature of thermal energy. He
expressed his ideas using a now-abandoned theory of heat. In this theory, heat is considered to be a fluid-like quantity—called
caloric. Transfers of heat comprise the flow of caloric from one object to another. Carnot’s ideas originated as an analogy between
the flow of caloric through a steam engine and the flow of water through a water wheel. In this view, the temperature of the steam,
entering and leaving the engine, is analogous to the altitude of the water entering and leaving the wheel . 1
Such considerations are obviously relevant if we are interested in building engines, but we are interested in chemical reactivity.
How does chemical change relate to engines and the conversion of heat into work? Well, rather directly, actually; after all, a
chemical reaction usually liberates or absorbs heat. If we can relate mechanical work to heat, and we can relate the amount of heat
liberated to the extent of a chemical reaction, then we can imagine allowing the reaction to go to equilibrium in a machine that
converts heat to work. We can expect that the amount of work produced will have some relationship to the extent of the reaction.
The nature of this relationship is obscure at this point, but we can reasonably expect that one exists.
This page titled 1.3: Classical Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.
Statistical thermodynamics is a theory that uses molecular properties to predict the behavior of macroscopic quantities of
compounds. While the origins of statistical thermodynamics predate the development of quantum mechanics, the modern
development of statistical thermodynamics assumes that the quantized energy levels associated with a particular system are known.
From these energy-level data, a temperature-dependent quantity called the partition function can be calculated. From the partition
function, all of the thermodynamic properties of the system can be calculated. We begin our development of statistical
thermodynamics by using the energy levels of an individual molecule to find its molecular partition function and the
thermodynamic properties of a system that contains N non-interacting molecules of that substance. Later, we see that the partition
function of a system containing molecules that do interact with one another can be found by very similar arguments.
Statistical thermodynamics has also been applied to the general problem of predicting reaction rates. This application is called
transition state theory or the theory of absolute reaction rates. In principle, we should be able to predict the rate of any reaction.
To do so, we need only to solve the quantum mechanical equations that give the energy levels associated with the reactants and the
energy levels associated with a transitory chemical structure called the transition state for the reaction. From the energy levels we
calculate partition functions; from partition functions we calculate thermodynamic functions; and from these thermodynamic
functions we obtain the reaction rate. There is a big difference between “in principle” and “in practice.” While increases in
computer speed make it increasingly feasible to do quantum mechanical calculations to useful degrees of accuracy, the results of
such calculations remain too inaccurate to give generally reliable reaction rate predictions. The theory of absolute reaction rates is
an important application of statistical thermodynamics. However, it is not included in this book.
Quantum mechanical calculations are not the only way to obtain the energy-level information that is needed to evaluate partition
functions. Particularly for small molecules, these energy levels can be deduced from spectroscopic data. In these cases, the theory
of statistical thermodynamics enables us to calculate thermodynamic properties from spectroscopic measurements. Excellent
agreement is obtained between the values of thermodynamic functions obtained from classical thermodynamic (thermochemical)
measurements and those obtained from statistical-thermodynamic calculations based on energy levels derived from spectroscopic
measurements. In Chapter 24, we consider a particular example to illustrate this point.
This page titled 1.4: Statistical Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
The amount of heat transferred to or from a system undergoing change is an important thermodynamic variable. In practical
devices, the rate at which heat can be transferred to or from a system plays a very important role also. Consider again the work
produced by heating a gas that is confined in a cylinder that is closed by a piston. Clearly, the rate at which heat can be transferred
from the outside to the gas determines the rate at which the piston moves outward and thus the rate at which work is done on the
environment.
Does it matter whether the heat-transfer process is fast or slow? If the heat cost nothing, would we care if our engine produced
work only very slowly? After all, if we want more work and the heat is free, we need only build more engines; eventually we will
have enough of them to produce any required amount of work. Of course, heat is not free; more significantly for our present
considerations, the engines are not free either. Engineers and accountants call the cost of heat an operating costcost:operating.
There are many other operating costs, like labor, supplies, insurance, and taxes. The cost of the engine is called a capital
costcost:capital. To find the total cost of a unit of work, we need to add up the various operating costs and a part of the cost of the
engine.
cost of a unit of work = fuel cost + other operating costs + capital cost
The difference between an operating cost and a capital cost is that an operating cost is incurred at (about) the same time that the
product, in this case a unit of work, is created. In contrast, a capital cost is incurred well before the product is created. The purchase
of a machine is a typical capital expense. The cost of the machine is incurred long before the machine makes its last product. This
occurs because the machine must be paid for when it is acquired, but it continues to function over a useful lifetime that is typically
many years. For example, if an engine that costs $1,000,000 can produce a maximum of 1,000,000 units of work before it wears
out, the minimum contribution that the cost of the engine makes to the cost of the work it produces is $1 per unit. The life of the
engine also enters into the estimation of capital cost. If some of the work done by the engine will be produced ten years in the
future, we will be foregoing the interest that we could otherwise have earned on the money that we invested in the engine while we
wait around to get the future work. Operating costs are well defined because they are incurred here and now. Capital costs are more
problematic, because they depend upon assumptions about things like the life of the machine and the variation in interest rates
during that life.
Suppose that we are developing a new engine. All else being equal, we can decrease the capital-cost component of the work our
engine produces by decreasing the time it needs to produce a unit of work. The savings occurs because we can get the same amount
of work from a smaller and hence less-costly engine. Since each unit of work requires that the same amount of heat be moved, we
can make the engine smaller only if we can move heat around more quickly. In internal combustion engines, we get heat into the
engine:internal combustion with a combustion reaction (an explosion) and take most of it out again by venting the combustion
products (the exhaust gas). So internal combustion engines have the great advantage that both of these steps can be fast. Steam
engines are successful because we can get heat into the engine quickly by allowing steam to flow from a boiler into the engine. We
can remove heat from the steam engine quickly by venting the spent steam, which is feasible because the working fluid is water.
The Stirling engine is a type of external combustion engine that works by alternately heating (expanding) and cooling
(compressing) an enclosed working fluid. Stirling engines have theoretical advantages, but they are not economically competitive,
essentially because heat transfer to and from the working fluid cannot be made fast enough.
Why does anyone care about capital cost? Well, we can be sure that the owner of an engine will be keenly interested in minimizing
the dollars that come out of his pocket. But capital cost is also a measure of the consumption of resources—resources that may
have more valuable alternative uses. So if any segment of an economy uses resources inefficiently, other segments of that economy
must give up other goals that could have been achieved using the wasted resources. Economic activity benefits many people
besides the owners of capital. If capital is used inefficiently, society as a whole is poorer as a result.
Heat transfer has a profound effect also on the design of the machines that manufacture chemicals. This occurs most conspicuously
in processes that involve very exothermic reactions. If heat cannot be removed from the reacting material fast enough, the
temperature of the material rises. The higher temperature may cause side reactions that decrease the yield of the product. If the
temperature rises enough, there may be an explosion. For such reactions, the equipment needed to achieve rapid heat transfer, and
to manage the rate of heat production and dissipation, may account for a large fraction of the cost of the whole plant. In some cases,
chemical reactions used for the production of chemicals produce enough heat that it is practical to use this “waste heat” for the
production of electricity.
This page titled 1.5: Heat Transfer in Practical Devices is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon
request.
We are familiar with the idea that a system undergoing change eventually reaches equilibrium. We say that a system is at
equilibrium when no further change is possible. When we talk about change, we always have in mind some particular property. We
measure the change in the system by the amount of change in this property. When the property stops changing, we infer that the
system has stopped changing, and we say that the system has come to equilibrium. Of course, we may interest ourselves in a
system in which many properties undergo change. In such cases, we recognize that the system as a whole cannot be at equilibrium
until all of these properties stop changing.
On the other hand, we also recognize that the absence of observable change is not enough to establish that a system is at
equilibrium with respect to all of the possible changes that it could undergo. We know that hydrogen burns readily in oxygen to
form water, but a mixture of hydrogen and oxygen undergoes no change under ordinary conditions. This unchanging mixture is
plainly not at equilibrium with respect to the combustion reaction. Only when a catalyst or an ignition source is introduced does
reaction begin.
It is also possible, indeed probable, that a system can be at equilibrium with respect to one process and not be at equilibrium with
respect to other processes, which, while possible, simply do not occur under the conditions at hand. For example, if an aqueous
solution of the oxygen-carrying protein hemoglobin is added to the hydrogen—air system, the protein will add or lose coordinated
oxygen molecules until the equilibrium composition is reached. If our investigation is focused on the protein–oxygenation reaction,
we do not hesitate to say that the system is at equilibrium. The non-occurrence of the oxygen—hydrogen reaction is not relevant to
the phenomenon we are studying.
It is even possible to reach a non-equilibrium state in which the concentrations of the reactants and products are constant. Such a
system is said to have reached a steady state. In order for this to occur, the reaction must occur in an open system; that is, one in
which materials are being added or removed; there must be continuous addition of reactants, and continuous removal of products.
In Chapter 5, we discuss a simple system in which this can be achieved. A closed system is one that can neither gain nor lose
material. An isolated system is a closed system that can neither gain nor lose energy; in consequence, its volume is fixed. In an
isolated system, change ceases when equilibrium is reached, and conversely.
We will consider several commonly encountered kinds of change, including mechanical motions, heat transfers, phase changes,
partitioning of a solute between two phases, and chemical reactions. Here we review briefly what occurs in each of these kinds of
change. In Chapter 6, we review the characteristics that each of these kinds of change exhibits at equilibrium.
A system in mechanical equilibrium is stationary because the net force acting on any macroscopic portion of the system is zero.
Another way of describing such a situation is to say that the system does not move because of the presence of constraints that
prevent movement.
Two macroscopic objects are in thermal equilibrium if they are at the same temperature. We take this to be equivalent to saying
that, if the two objects are in contact with one another, no heat flows between them. Moreover, if object A is in thermal equilibrium
with each of two other objects, B and C, then we invariably find that objects B and C are in thermal equilibrium with one another.
This observation is sometimes called the zeroth law of thermodynamics. It justifies the concept of temperature and the use of a
standard system—a thermometer—to measure temperature.
For an isolated system to be in phase equilibrium, it must contain macroscopic quantities of two or more phases, and the amount of
each phase present must be unchanging. For example, at 273.15 K and 1 bar, and in the presence of one atmosphere of air, liquid
water and ice are in equilibrium; the amounts of water and ice remain unchanged so long as the system remains isolated. Similarly,
a saturated aqueous solution of copper sulfate is in equilibrium with solid copper sulfate; if the system is isolated, the amounts of
solid and dissolved copper sulfate remain constant.
If a system is in phase equilibrium, we can remove a portion of any phase without causing any change in the other phases. At
equilibrium, the concentrations of species present in the various phases are independent of the absolute amount of each phase
present. It is only necessary that some amount of each phase be present. To describe this property, we say that the condition for
equilibrium for is the same irrespective of the amounts of the phases present in the particular system. For example, if one of the
species is present in both a gas phase and a condensed phase, we can specify the equilibrium state by specifying the pressure and
temperature of the system. However, we can change the relative amounts of the phases present in this equilibrium state by changing
the volume of the system. (If its volume can change, the system is not isolated.)
Partitioning of a solute between two immiscible condensed phases is important in many chemical systems. If we add water and
chloroform to the same vessel, two immiscible liquid phases are formed. Elemental iodine is very sparingly soluble in water and
substantially more soluble in chloroform. If we add a small amount of iodine to the water–chloroform system, some of the iodine
dissolves in the water and the remainder dissolves in the chloroform layer. We say that the iodine is distributed between the two
phases. When the iodine concentrations become constant, we say that the system has reached distribution equilibrium.
In a chemical reaction, one or more chemical substances (reactants) undergo a change to produce one or more new chemical
substances (products). We are accustomed to representing chemical substances by symbols and representing their reactions by
chemical equations. Thus, for the hydrolysis of ethyl acetate, we write
C H3 C O2 C H2 C H3 + H2 O ⇌ C H3 C O2 H + C H3 C H2 OH
A chemical equation like this expresses a stoichiometric relationship between reactants and products. Often we invoke it as a
symbol for various distinctly different physical situations. For example:
1. We may view the equation as a symbolic representation of a single solution that contains the four compounds ethyl acetate,
water, acetic acid, and ethanol—and possibly other substances.
2. We may view the equation as a symbolic representation of a relationship between two systems whose proportions are arbitrary.
The first system comprises ethyl acetate and water. The second system comprises acetic acid and ethanol. The equation
represents the idea that the first system can be converted into the second.
3. We may view the symbols on each side of the equation as representing mixtures of the indicated chemical substances in the
specified stoichiometric proportions.
4. We may view the equation as representing the specified stoichiometric proportions of pure, unmixed chemical substances.
When we are discussing changes in “standard” thermodynamic properties that accompany a chemical reaction, this is the
interpretation that we have in mind.
When we discuss a chemical equation, the intended interpretation is normally evident from the context. Indeed, we often skip back
and forth among these interpretations in the course of a single discussion. Nevertheless, it is important to avoid confusing them.
By doing experiments, we can discover that there is an equation that uniquely defines the position of a chemical reaction at
equilibrium, an equation that we usually think of as the definition of the equilibrium constant. If our measurements are not too
accurate, or we confine our study to a limited range of concentrations, or the system is particularly well behaved, we can express
the equilibrium constant as a function of concentrations . For the hydrolysis of ethyl acetate, we find
2
[C H3 C O2 H ] [H OC H2 C H3 ]
K =
[C H3 C O2 C H2 C H3 ] [ H2 O]
In general, for the reaction
a A+b B ⇌ c C +d D
we find
c d
[C ] [D]
K =
a b
[A] [B]
That is, at equilibrium the indicated function of reactant concentrations always computes to approximately the same numerical
value.
When our concentration measurements are more accurate, we find that we must introduce new quantities that we call chemical
activities. We can think of an activity as a corrected concentration. The correction compensates for the effects of intermolecular
attraction and repulsion. Denoting the activity of substance X as a
~
, we find that
x
~c ~d
a a
C D
Ka =
~a ~b
a a
A B
gives a fully satisfactory characterization of the equilibrium states that are possible for systems in which this reaction occurs. K ,
a
the equilibrium constant computed as a function of reactant activities, always has exactly the same numerical value.
We can develop the equilibrium constant expression from three distinctly different theoretical treatments. We develop it first from
some basic ideas about the rates of chemical reactions. Then we obtain same result from both the macroscopic-behavior
considerations of classical thermodynamics and the molecular-property considerations of statistical thermodynamics.
Our most basic concept of equilibrium is based on the observation that change in an isolated system eventually ceases; once change
ceases, it never resumes. In this book, we call the idea of a static state of an isolated system the primitive equilibrium. We also
observe that change eventually ceases in a closed system that is not isolated but whose temperature, pressure, and volume are kept
constant. Conversely, if a system is at equilibrium, its temperature, pressure, and volume are necessarily constant; all interactions
between such a system and its surroundings can be severed without changing any of the properties of the system. We can view any
particular equilibrium state as a primitive equilibrium state.
A system whose temperature, pressure, or volume is established by interactions between the system and its surroundings is
inherently more variable than an isolated system. For a given isolated system, only one equilibrium state exists; for a system that
interacts with its surroundings, many different equilibrium states may be possible. In chemical thermodynamics, our goal is to
develop mathematical models that specify the equilibrium states available to a system; we seek models in which the independent
variables include pressure, temperature, volume, and other conditions that can be imposed on the system by its surroundings. In this
conception, an equilibrium system is characterized by a set of points in a variable space. We can think of this set of points as a
surface or a manifold in the variable space; every point in the set is a different primitive-equilibrium state of the system. By
imposing particular changes on some variables, a particular equilibrium system can be made to pass continuously through a series
of primitive-equilibrium states.
For reasons that become apparent as we proceed, we use the name Gibbsian equilibrium to denote this more general conception.
When we talk about equilibrium in thermodynamics, we usually mean Gibbsian equilibrium. In Chapter 6, we see that the idea of
(Gibbsian) equilibrium is closely related to the idea of a reversible process. We also introduce Gibbs’ phase rule, which amounts to
a more precise definition, from the perspective of classical thermodynamics, of what we mean by (Gibbsian) equilibrium in
chemical systems.
This page titled 1.6: The Concept of Equilibrium is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
When we talk about predicting chemical reactions, we imagine taking quantities of various pure compounds and mixing them
under some set of conditions. We suppose that they react until they reach a position of equilibrium in which one or more new
compounds are present. We want to predict what these new compounds are and how much of each will be produced.
For any given set of reactants, we can accomplish this predictive program in two steps. First we find all of the sets of products that
can be obtained from the given reactants. Each such set represents a possible reaction. We suppose that, for each set of possible
products, we are able to predict the equilibrium composition. Predicting which reaction will occur is equivalent to finding the
reaction whose position of equilibrium lies farthest in the direction of its products. From this perspective, being able to predict the
position of equilibrium for the reactants and any stoichiometrically consistent set of products is the same thing as being able to
predict what reaction will occur. (If there is no single reaction whose position of equilibrium is much further in the direction of its
products than that of any other reaction, multiple reactions can occur simultaneously.)
This two-step procedure corresponds to the sense in which chemical thermodynamics enables us to predict reaction products. We
measure values for certain characteristic thermodynamic functions for all relevant compounds. Given the values of these functions
for all of the compounds involved in a hypothesized reaction, we calculate the position of equilibrium. That we must begin by
guessing the products makes this approach cumbersome and uncertain. We can never be positive that the true products are among
the possibilities that we consider. Nevertheless, as a practical matter for most combinations of reactants, the number of plausible
product sets is reasonably small.
This page titled 1.7: Chemical Equilibrium and Predicting Chemical Change is shared under a CC BY-SA 4.0 license and was authored, remixed,
and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is
available upon request.
We develop classical thermodynamics by reasoning about reversible processes—processes in which a system passes through a
series of equilibrium states. Any such process corresponds to a path on one or more of the Gibbsian manifolds that are available to
the system. The resulting theory consists of equations that relate the changes in the values of the system’s state functions as the
system undergoes a reversible change. For this reason, the body of theory that we are calling classical thermodynamics is often
called equilibrium thermodynamics or reversible thermodynamics.
As we discuss further in Chapter 6, any change that we can actually observe in a real system must be the result of a spontaneous
process. In a reversible process, both the initial and the final states are equilibrium states. In a spontaneous process, the initial state
of the system is not an equilibrium state. A spontaneous process begins with the system in a non-equilibrium state and proceeds
until an equilibrium state is reached.
The domain of classical thermodynamics—reversible processes—is distinct from the domain of real observations, because real
observations can be made only for spontaneous processes. We bridge this gap by careful selection of real-world systems to serve as
models for the reversible systems that inhabit our theory. That is, we find that we can make measurements on non-equilibrium
systems and irreversible processes from which we can estimate the properties of equilibrium systems and reversible processes.
Saying almost the same thing from another perspective, we find that the classical thermodynamic equations that apply to
equilibrium states can also be approximately valid for non-equilibrium states. For many non-equilibrium states, notably those
whose individual phases are homogenous, the approximations can be very good. For other non-equilibrium states, notably those
whose individual phases are markedly inhomogeneous, these approximations may be very poor.
This page titled 1.8: Equilibrium and Classical Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The goal of science is to create theories that accurately describe physical reality. In this book, we explore some of the most useful
scientific theories that exist. They have been tested extensively. We know that there are limits to their applicability. We expect that
further thought and experimentation will expand their scope. We expect that some elements of these theories will need to be
modified in ways that we cannot anticipate, but we do not expect that the core concepts will be invalidated.
We love theories because they rationalize our environment. This is true not only of scientific theories but also of the panoply of
conceptual frameworks that we use to organize our views about—and responses to— all of life’s issues. We are addicted to
theories. We find nothing more disconcerting than information that we cannot put into a coherent context. Indeed, the term
cognitive dissonance has entered the language to describe the feeling of disorientation that we experience when “things just don’t
add up.”
Logically, confronted with a fact that contradicts one of our theories, we are compelled to give up the theory. We are not always
logical. A fact that contradicts a pet theory is unlikely to be accepted at face value. We challenge it, as indeed we should. We
scrutinize the offending fact and try to convince ourselves that it is no fact at all, merely a spurious artifact. Often, of course, this
proves to be the case. Sometimes we conclude that the offending fact is spurious when it is not. We get stuffy about our theories.
When we find one that suits us, we resist giving it up. It has been observed that a revolutionary scientific theory often achieves
universal acceptance only after all those who grew up with the predecessor theory have died.
Science is ultimately a social enterprise. To develop and test theories about physical reality, participants in this enterprise must be
in general agreement about the criteria that are to be applied. These criteria are frequently called “the philosophy of science.” To
summarize the philosophy of science, we begin by observing that the goal of science is to explain the world that we experience
through our sensory perceptions. It is easy to generate putative explanations that have little or no real value. Unfortunate
experiences with past explanations have led to a broad consensus that scientific theories must have the following properties:
operational definitions,
logical structure,
predictive capability and testability,
internal consistency, and
consistency with any and all experimental observations.
The theory must be about the properties of some set of things. By operational definitions, we mean that the subjects of the theory
must measurable, and the theory must specify a set of operations for making each of these measurements. By logical structure, we
mean that a satisfactory theory must include well-defined rules to specify how the subjects of the theory relate to one another. By
predictive capability we mean that a satisfactory theory must be capable of predicting the results of experiments that have not been
performed. By internal consistency we mean that a satisfactory theory must not allow us to logically derive contradictory
conclusions, which also means that it must not predict more than one outcome for any particular experiment. Because a satisfactory
theory makes predictions, it is also testable. It is possible to check whether the predictions correspond to reality. We require that the
theory’s predictions be consistent with the results that we observe when we do the experiment.
The first four of these requirements really detail the characteristics that a theory must have in order to be considered a proper
subject for scientific investigation. Only the last requirement speaks to the all-important issue of whether the theory accurately
mirrors physical reality. We can never prove that any theory is true. What we can prove is that a theory fails to meet one of our
criteria. Science progresses when we discover a fatal flaw in a currently accepted theory.
Let us think further about what we mean when we say that a theory must be a logical structure. Consider a simple classical
syllogism.
Major premise: All dogs are cats.
Minor premise: All cats are white.
Conclusion: All dogs are white.
As a logical structure, this seems to be satisfactory. We can represent the whole of its content in a simple diagram (See Figure 1), so
if we want to view this syllogism as a theory about nature, its internal consistency is more or less self-evident. Moreover, viewed as
a theory, it makes a prediction: All dogs are white. If we have operational definitions for “dog”, “cat”, and “white” that conform to
customary usage, we can say that this syllogism meets our criteria for a proper subject for scientific investigation. Of course, as a
mirror of reality, it fails.
Figure 1. Venn diagram of a classical syllogism

To see the issue of logical structure from another perspective, let us consider the theory of evolution. Some people summarize the
theory of evolution as teaching that the fittest individuals survive and defining survivors as those individuals who are most fit. They
then point out that these are circular statements and proceed from this observation to the conclusion that the theory of evolution is
devoid of real content. So it can be dismissed. Now, if we are not closed-minded about evolution, this analysis looks like a case of
throwing out the baby with the bath water. Even so, we are likely to be troubled, because the circularity is undeniable. Does this
circularity mean that the theory of evolution is bad science?
A tautology is a statement that must be true. Our analysis attempts to recast the entire content of the theory of evolution as one
tautologous statement. If the whole of a theory is a single statement, and that statement must be true, then the theory cannot be
tested. Our rules require that we reject it. However, our tautologous summary fails to capture the whole of the theory of evolution.
If a theory that contains tautologous statements also makes predictions that are not tautologous, then it can be tested. In the present
example, we can predict from the theory of evolution that selection of a particular trait through any process will cause increased
expression of that trait in succeeding generations. Evolution is based on natural selection, but it postulates a mechanism that must
be valid for any consistent selection process. It predicts that a farmer who selects for cows that produce more milk will eventually
get cows that produce more milk. Thus, attempts to apply selective breeding are tests of the central element of the theory of
evolution, and the success of selective breeding in every aspect of agriculture verifies a prediction of the theory.
There is no reason to object to a theory that has tautological elements so long as the content of the theory has real substance. What
we require is that a theory’s predictions be non-trivial. We object when substantially all of a theory’s purported predictions are
merely restatements of its premises, so that the whole of the theory is an exercise in verbiage.
We require scientific theories to be internally consistent. (Normally, we do not expect to be able to prove internal consistency. What
we really mean is that we will discard any theory that we can show to be internally inconsistent.) The presence of tautologous
statements cannot make a theory internally inconsistent. Indeed, we can expect any internally consistent theory to have tautological
elements. After all, if we try to define all of the subjects of a theory, at least some of our definitions will inevitably be circular.
Another way to describe the logical structure of a physical theory is to say that a theory is a model for some observable part of the
world. We want the model to include things and rules. The rules should specify how the things of the model change. When we talk
about comparing predictions of the theory to the results of experiment, we mean that the changes that occur in the model when we
apply the rules of the theory should parallel the changes that occur in the real world when we do the corresponding experiment. The
idea is analogous to the mathematical concept of a homomorphism. A dictionary definition of a homomorphism is “a mapping of
3
a mathematical group, ring, or vector space onto another in such a way that the result obtained by applying an operation to elements
of the domain is mapped onto the result of applying the same or a corresponding operation to their images in the range.” Stated
more picturesquely, the idea of a homomorphism is that two mathematical structures have the same form: If one is “laid on top of
the other”, there is a perfect correspondence between them. (See Figure 2.)
Figure 2. A homomorphism.
The idea of a one-to-one correspondence between two structures can be extended to the case where one structure is a logical
structure and the other is a physical structure. Consider the logical structure, usually called a truth table, associated with the truth of
the proposition “sentence A is true and sentence B is true.” The proposition is false unless both A and B are true. Now consider the
performance of an electrical circuit that consists of a battery, switches A and B, and a light bulb, all connected in series. The light is
off unless both A and B are on. There is a perfect correspondence between the elements of the logical structure and the
performance of the circuit. (See Figure 3.) In fact, this is a special case of a much more extensive parallelism. For any sentence in
propositional logic, there is a corresponding circuit—involving batteries, switches, and a light bulb—such that the bulb is on if the
corresponding sentence is true and off if the sentence is false. Moreover, there is a mathematical structure, called Boolean algebra,
which is homomorphic to propositional logic and exhibits the same parallelism to circuits. Digital computers carry this parallelism
to structures of great complexity.
If we stretch our definition of a homomorphism to include comparisons of logical constructs with physical things, we can view a
scientific theory as a logical construct that we are trying to make homomorphic with physical reality. We want our theory to map
onto reality in such a way that changes in the logical construct map onto changes in physical reality. Of course, what we mean by
this is the same thing we meant previously when we said that the predictions of our theory should agree with the results of
experiment.
Figure 3. Mapping a physical system onto a logical construct.

It is implicit in this view that any scientific theory describes an abstraction from reality. If, for example, we say that a physical
system is at a particular temperature, this statement represents an approximation in at least two respects. First, we believe that we
can measure temperature in continually more refined ways. If we know the temperature to six significant figures, this represents a
very good approximation, but we believe that the “true” temperature can only be expressed using indeterminably many significant
figures. To say that the temperature of the system has some value expressed to six significant figures is an approximation, albeit a
very good one. Second, it is unlikely that all parts of any real system are actually at the same temperature. When we say that a
physical system has a particular temperature, we mean that any differences in temperature between different parts of the system are
too small to affect the behavior in which we are currently interested. In our discussions and deliberations about the physical system,
we replace its actual properties by our best measurement. In doing so, we abstract something that we believe to be an essential
feature from the reality we actually observe.
Implicit also is the idea that our theories about reality are subject to inevitable limitations. In the example above, any theory that,
however accurately, predicts the single temperature that we use to describe the real system is inadequate to describe whatever small
variations there may be from point to point within it. If we expand our theory to encompass variations over, say, distances of
millimeters, then the expanded theory will be inadequate to describe variations over some smaller distance. Any “perfect” theory
must exactly describe the motions of all of the system’s constituent particles. This is impossible not only because it conflicts with
the basis premises of quantum mechanics, but also because it requires the theory to contain information at a level of detail equal to
that in the physical system itself.
One way to express all of the same information at exactly the same level of detail is to have an exact replica of the system. This
idea has been expressed by saying that an absolutely accurate map of Oklahoma City would have to be as large as Oklahoma City
itself. For use as a map, such a thing would be useless. Similarly, a theory that predicts temperature is useful only if it predicts the
temperature we measure (measurement being a part of the process by which we effect abstraction) experimentally. We can make
use of multiple theories of the same phenomenon, if each of them has advantages and limitations that we recognize and respect. We
see however, that in the end, there can be no single, all-encompassing theory. Any theory must model an approximation to reality.
In the final analysis, reality is … reality.
This page titled 1.9: A Few Ideas from the Philosophy of Science is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Formal logic deals with relationships among propositions, where a proposition is any statement of (alleged) fact. Any proposition
can be expressed as an ordinary English sentence, although it may be more convenient to use mathematical symbols or some other
notation. The following are all propositions:
Albert Einstein is deceased.
Tulsa is in Oklahoma.
Two plus two equals four.
2 +2 = 4
2
∫ x dx = x + 1
A proposition need not be true. The last of these examples is a false proposition. We represent an arbitrary proposition by any
convenient symbol, usually a letter of the alphabet. Thus, we could stipulate that “p” represents any of the propositions above.
Once we have associated a symbol with a particular proposition, the symbol itself is taken to represent an assertion that the
proposition is true. It is an axiom of ordinary logic that any proposition must be either true or false. If we associate the symbol “p”
with a particular proposition, we write “∼ p ” to represent the statement: “The proposition represented by the symbol ‘p’ is false.”
∼ p is called the negation of p. We can use the negation of p , ∼ p , to state the axiom that a proposition must be either true or false.
To do so, we write: Either p or ∼ p is true. We can write this as the proposition “p or ∼ p ”. The negation of the negation of p is an
assertion that p is true; that is, ∼ ∼ p = p .
Logic is concerned with relationships among propositions. One important relationship is that of implication. If a proposition, q,
follows logically from another proposition, p, we say that q is implied by p. Equivalently, we say that proposition p implies
proposition q. The double-shafted arrow, ⇒, is used to symbolize this relationship. We write “p ⇒ q ” to mean, “That proposition p
is true implies that proposition q is true.” We usually read this more tersely, saying, “p implies q.” Of course, “p ⇒ q ” is itself a
proposition; it asserts the truth of a particular logical relationship between propositions p and q.
For example, let p be the proposition, “Figure A is a square.” Let q be the proposition, “Figure A is a rectangle.” Then, writing out
the proposition, p ⇒ q , we have: Figure A is a square implies figure A is a rectangle. This is, of course, a valid implication; for this
example, the proposition p ⇒ q is true. For reasons that will become clear shortly, p ⇒ q is called the conditional of p and q.
Proposition p is often called a sufficient condition, while proposition q is called a necessary condition. That is, the truth of p is
sufficient to establish the truth of q.
sufficient condition ⇒ necessary condition
Now, if proposition p ⇒ q is true, and proposition q is also true, can we infer that proposition p is true? We most certainly cannot!
In the example we just considered, the fact that figure A is a rectangle does not prove that figure A is a square. We call q ⇒ p the
converse of p ⇒ q . The conditional of p and q can be true while the converse is false. Of course, it can happen that both p ⇒ q and
q ⇒ p are true. We often write “p ⇔ q ” to express this relationship of mutual implication. We say that, “p implies q and
conversely.”
What if p ⇒ q , and q is false? That is, ∼ q is true. In this case, p must be false! If ∼ q is true, it must also be that ∼ p is true. Using
our notation, we can express this fact as
(p ⇒ q and ∼ q) ⇒∼ p
Equivalently, we can write
(p ⇒ q)⇔(∼ q ⇒∼ p)
That is, p ⇒ q and ∼ q ⇒∼ p are equivalent propositions; if one is true, the other must be true. ∼ q ⇒∼ p is called the
contrapositive of p ⇒ q . The equivalence of the conditional and its contrapositive is a theorem that can be proved rigorously in an
axiomatic formulation of logic. In our later reasoning about thermodynamic principles, we use the equivalence of the conditional
and the contrapositive of p and q.
The equivalence of the conditional, p ⇒ q , and the contrapositive, ∼ q ⇒∼ p , is the reason that q is called a necessary condition.
If p ⇒ q , it is necessary that q be true for p to be true. (If figure A is to be a square, it must be a rectangle.)
It is also intimately related to proof by contradiction. Suppose that we know p to be true. If, by assuming that q is false (∼ q is
true), we can validly demonstrate that p must also be false (∼ q⇒ ∼ p , so that ∼ p is true), we have the contradiction that p is both
true and false (p and ∼ p ). Since p cannot be both true and false, it must be false that q is false (∼∼ q = q ). Otherwise stated, the
equivalence of the conditional and the contrapositive leads not only to (p and ∼ p ) but also to (q and ∼ q ).
∼ q⇒ ∼ p
implies
p⇒q.
In summary, since we know p to be true, our assumption that q is false, together with the valid implication ∼ q⇒ ∼ p , leads to the
conclusion that q is true, which contradicts our original assumption, so that the assumption is false, and q is true.
This page titled 1.10: A Few Ideas from Formal Logic is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
1.11: Problems
1. Philosophers argue about the feasibility of a private language, a language that is known by only one individual. The case against
the possibility of a private language is based on the assumption that language comes into existence as a tool for communication in a
society comprising two or more individuals. Solipsism is a philosophical conception in which your sensory perceptions are
internally generated; they are not the result of your interactions with the world. Solipsism assumes that the world you think you
perceive does not in fact exist. Since a solipsistic individual is the only being that exists, any language he uses is necessarily a
private language. Evidently, the existence of a solipsistic individual who uses some language to think and the impossibility of a
private language are mutually exclusive: If a private language is impossible, a solipsistic individual cannot use any language. Since
you are reading this, you are using a language, and therefore you cannot be a solipsistic individual. Does this argument convince
you that you are not a solipsistic individual; that is, does this argument convince you of the existence of a physical reality that is
external to yourself? Why or why not? Can a solipsistic individual engage in scientific inquiry?
2. Many people find the theory of evolution deeply repugnant. Some argue that the theory of evolution has not been proved and that
creationism is an alternative scientific theory, where creationism is the Biblical description of God’s creation of the world in six
days.
(a) Is it valid to say, “The theory of evolution is unproved.”?
(b) Comment very briefly on whether or not the theory of evolution meets each of our criteria for a scientific theory.
(c) Comment very briefly on whether or not creationism meets each of our criteria for a scientific theory.
3. Use an ordinary English sentence to state the meaning of propositions (a) and (b):
(a) ∼ (p and q) ⇒ (∼ p or ∼ q)
(b) ∼ (p or q) ⇒ (∼ p and ∼ q)
Are propositions (a) and (b) true or false?

Using propositions (a) and (b), prove propositions (c) and (d):
(c) [(p and q) ⇒r ] ⇒ [∼ r ⇒ (∼ p or ∼ q) ]
(d) [(p or q) ⇒r ] ⇒ [∼ r ⇒ (∼ p and ∼ q) ]
Notes
1
R. Clausius, The Mechanical Theory of Heat, translated by Walter R. Browne, Macmillan and Co., London, 1879, p. 76.
2
We use square brackets around the symbol for a chemical substance to denote the concentration of that substance in molarity
(moles per liter of solution) units.
3
Webster’s Ninth New Collegiate Dictionary, Merriam-Webster, Inc., Springfield, Massachusetts, 1988.
This page titled 1.11: Problems is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.
CHAPTER OVERVIEW
2: Gas Laws
Early experimenters discovered that the pressure, volume, and temperature of a gas are related by simple equations. The classical
gas laws include Boyle’s law, Charles’ law, Avogadro’s hypothesis, Dalton’s law of partial pressures, and Amagat’s law of partial
volumes. These laws were inferred from experiments done at relatively low pressures and at temperatures well above those at
which the gases could be liquefied. We begin our discussion of gas laws by reviewing the experimental results that are obtained
under such conditions. As we extend our experiments to conditions in which gas densities are greater, we find that the accuracy of
the classical gas laws decreases.
2.1: Boyle's Law
2.2: Charles' Law
2.16: Problems
1
We use the over-bar to indicate that the quantity is per mole of substance. Thus, we write N to indicate the number of particles
¯
¯¯¯
¯
¯¯¯¯
¯
per mole. We write M to represent the gram molar mass. In Chapter 14, we introduce the use of the over-bar to denote a partial
molar quantity; this is consistent with the usage introduced here, but carries the further qualification that temperature and pressure
are constant at specified values. We also use the over-bar to indicate the arithmetic average; such instances will be clear from the
context.
2
The unit of temperature is named the kelvin, which is abbreviated as K.
3
A redefinition of the size of the unit of temperature, the kelvin, is under consideration. The practical effect will be
inconsequential for any but the most exacting of measurements.
4
For a thorough discussion of the development of the concept of temperature, the evolution of our means to measure it, and the
philosophical considerations involved, see Hasok Chang, Inventing Temperature, Oxford University Press, 2004.
5
See T. L. Hill, An Introduction to Statistical Thermodynamics, Addison-Wesley Publishing Company, 1960, p 286.
6
See S. M. Blinder, Advanced Physical Chemistry, The Macmillan Company, Collier-Macmillan Canada, Ltd., Toronto, 1969, pp
185-18926
This page titled 2: Gas Laws is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
1
2.1: Boyle's Law
Robert Boyle discovered Boyle’s law in 1662. Boyle’s discovery was that the pressure, P, and volume, V, of a gas are inversely
proportional to one another if the temperature, T, is held constant. We can imagine rediscovering Boyle’s law by trapping a sample
of gas in a tube and then measuring its volume as we change the pressure. We would observe behavior like that in Figure 1. We can
represent this behavior mathematically as
∗
P V = α (n, T )
where we recognize that the “constant”, α , is actually a function of the temperature and of the number of moles, n , of gas in the
∗
sample. That is, the product of pressure and volume is constant for a fixed quantity of gas at a fixed temperature.
A little thought convinces us that we can be more specific about the dependence on the quantity of gas. Suppose that we have a
volume of gas at a fixed pressure and temperature, and imagine that we introduce a very thin barrier that divides the volume into
exactly equal halves, without changing anything else. In this case, the pressure and temperature of the gas in each of the new
containers will be the same as they were originally. But the volume is half as great, and the number of moles in each of the half-size
containers must also be half of the original number. That is, the pressure–volume product must be directly proportional to the
number of moles of gas in the sample:
P V = nα(T )
where α(T ) is now a function only of temperature. When we repeat this experiment using different gaseous substances, we
discover a further remarkable fact: Not only do they all obey Boyle’s law, but also the value of α(T ) is the same for any gas.
Figure 1. Gas volume versus pressure.
This page titled 2.1: Boyle's Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
2.2: Charles' Law
Quantitative experiments establishing the law were first published in 1802 by Gay-Lussac, who credited Jacques Charles with
having discovered the law earlier. Charles’ law relates the volume and temperature of a gas when measurements are made at
constant pressure. We can imagine rediscovering Charles’ law by trapping a sample of gas in a tube and measuring its volume as
we change the temperature, while keeping the pressure constant. This presumes that we have a way to measure temperature,
perhaps by defining it in terms of the volume of a fixed quantity of some other fluid—like liquid mercury. At a fixed pressure, P , 1
we observe a linear relationship between the volume of a sample of gas and its temperature, like that in Figure 2. If we repeat this
experiment with the same gas sample at a higher pressure, P , we observe a second linear relationship between the volume and the
2
temperature of the gas. If we extend these lines to their intersection with the temperature axis at zero volume, we make a further
important discovery: Both lines intersect the temperature axis at the same point.
Figure 2: Gas volume versus temperature.

We can represent this behavior mathematically as
∗ ∗ ∗
V =β (n, P ) T +γ (n, P )
where we recognize that both the slope and the V-axis intercept of the graph depend on the pressure of the gas and on the number
of moles of gas in the sample. A little reflection shows that here too the slope and intercept must be directly proportional to the
number of moles of gas, so that we can rewrite our equation as
∗
V = nβ (P ) T + nγ(P )
When we repeat these experiments with different gaseous substances, we discover an additional important fact: β(P ) and γ(P ) are
the same for any gas. This means that the temperature at which the volume extrapolates to zero is the same for any gas and is
independent of the constant pressure we maintain as we vary the temperature (Figure 2).
This page titled 2.2: Charles' Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
Avogadro’s hypothesis is another classical gas law. It can be stated: At the same temperature and pressure, equal volumes of
different gases contain the same number of molecules.
When the mass, in grams, of an ideal gas sample is equal to the gram molar mass (traditionally called the molecular weight) of the
gas, the number of molecules in the sample is equal to Avogadro’s number, N . Avogadro’s number is the number of molecules in
¯
¯¯¯
¯1
a mole. In the modern definition, one mole is the number of atoms of C in exactly 12 g of C . That is, the number of atoms of
12 12
C
12
in exactly 12 g of C is Avogadro’s number. The currently accepted value is 6.02214199×10 molecules per mole. We can
12 23
find the gram atomic mass of any other element by finding the mass of that element that combines with exactly 12 g of C in a 12
compound whose molecular formula is known.

The validity of Avogadro’s hypothesis follows immediately either from the fact that the Boyle’s law constant, α(T ), is the same for
any gas or from the fact that the Charles’ law constants, β(P ) and γ (P ), are the same for any gas. However, this entails a
significant circularity; these experiments can show that α(T ), β(P ), and γ (P ) are the same for any gas only if we know how to
find the number of moles of each gas that we use. To do so, we must know the molar mass of each gas. Avogadro’s hypothesis is
crucially important in the history of chemistry: Avogadro’s hypothesis made it possible to determine relative molar masses. This
made it possible to determine molecular formulas for gaseous substances and to create the atomic mass scale.
This page titled 2.3: Avogadro's Hypothesis is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
This use of Avogadro’s number raises the question of how we know its value. There are numerous ways to measure Avogadro’s
number. One such method is to divide the charge of one mole of electrons by the charge of a single electron. We can obtain the
charge of a mole of electrons from electrolysis experiments. The charge of one electron can be determined in a famous experiment
devised by Robert Millikan, the “Millikan oil-drop experiment”. The charge on a mole of electrons is called the faraday.
Experimentally, it has the value 96, 485 C mol (coulombs per mole). As determined by Millikan’s experiment, the charge on one
−1
electron is 1.6022 × 10 −19

C . Then
96, 485 C 1 electron electrons

23
( )( ) = 6.022 × 10
−19
mole electrons 1.6022 × 10 C mole electrons
This page titled 2.4: Finding Avogadro's Number is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
Thus far, we have assumed nothing about the value of the temperature corresponding to any particular volume of our standard
fluid. We could define one unit of temperature to be any particular change in the volume of our standard fluid. Historically,
Fahrenheit defined one unit (degree) of temperature to be one one-hundredth of the increase in volume of a fixed quantity of
standard fluid as he warmed it from the lowest temperature he could achieve, which he elected to call 0 degrees, to the temperature
of his body, which he elected to call 100 degrees. Fahrenheit’s zero of temperature was achieved by mixing salt with ice and water.
This is not a very reproducible condition, so the temperature of melting ice (with no salt present), soon became the calibration
standard. Fahrenheit’s experiments put the melting point of ice at 32 F. The normal temperature for a healthy person is now taken to
be 98.6 F; possibly Fahrenheit had a slight fever when he was doing his calibration experiments. In any case, human temperatures
vary enough so that Fahrenheit’s 100-degree point was not very practical either. The boiling point of water, which Fahrenheit’s
experiments put at 212 F, became the calibration standard. Later, the centigrade scale was developed with fixed points at 0 degrees
and 100 degrees at the melting point of ice and the boiling point of water, respectively. The centigrade scale is now called the
Celsius scale after Anders Celsius, Anders, a Swedish astronomer. In 1742, Celsius proposed a scale on which the temperature
interval between the boiling point and the freezing point of water was divided into 100 degrees; however, a more positive number
corresponded to a colder condition.
Further reflection convinces us that the Charles’ law equation can be simplified by defining a new temperature scale. When we
extend the straight line in any of our volume-versus-temperature plots, it always intersects the zero-volume horizontal line at the
same temperature. Since we cannot associate any meaning with a negative volume, we infer that the temperature at zero volume
represents a natural minimum point for our temperature scale. Let the value of T at this intersection be T . Substituting into our
∗
0
∗
volume-temperature relationship, we have

∗
0 = nβ (P ) T + nγ(P )
0
or
∗
γ (P ) = −β(P )T
0
So that
∗ ∗
V = nβ (P ) T − nβ(P )T (2.5.1)
0
∗ ∗
= nβ (P ) [ T −T ] (2.5.2)
0
= nβ (P ) T (2.5.3)
where we have created a new temperature scale. Temperature values on our new temperature scale, T, are related to temperature
values on the old temperature scale, T , by the equation
∗
∗ ∗
T =T −T
0
When the size of one unit of temperature is defined using the Celsius scale (i.e., T is the temperature in degrees Celsius), this is
∗
the origin of the Kelvin temperature scale . Then, on the Kelvin temperature scale, T is -273.15 degrees. (That is, T = 0 when
2 ∗
0
T
0
∗
= 273.15; 0 K is −273.15 degrees Celsius.) The temperature at which the volume extrapolates to zero is called the absolute
zero of temperature. When the size of one unit of temperature is defined using the Fahrenheit scale and the zero of temperature is
set at absolute zero, the resulting temperature scale is called the Rankine scale, after William Rankine, a Scottish engineer who
proposed it in 1859.
This page titled 2.5: The Kelvin Temperature Scale is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
We can solve Boyle’s law and Charles’ law for the volume. Equating the two, we have
nα(T )
= nβ (P ) T
P
The number of moles, n , cancels. Rearranging gives

α(T )
= P β(P )
T
In this equation, the left side is a function only of temperature, the right side only of pressure. Since pressure and temperature are
independent of one another, this can be true only if each side is in fact constant. If we let this constant be R , we have
α (T ) = RT
and
β (P ) = R/P
Since the values of α(T ) and β(P ) are independent of the gas being studied, the value of R is also the same for any gas. R is
called the gas constant, the ideal gas constant, or the universal gas constant. Substituting the appropriate relationship into either
Boyle’s law or Charles’ law gives the ideal gas equation
P V = nRT
The product of pressure and volume has the units of work or energy, so the gas constant has units of energy per mole per degree.
(Remember that we simplified the form of Charles’s law by defining the Kelvin temperature scale; temperature in the ideal gas
equation is in degrees Kelvin.)
This page titled 2.6: Deriving the Ideal Gas Law from Boyle's and Charles' Laws is shared under a CC BY-SA 4.0 license and was authored,
remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit
history is available upon request.
Having developed the ideal gas equation and analyzed experimental results for a variety of gases, we will have found the value of
R. It is useful to have R expressed using a number of different energy units. Frequently useful values are
3 −1 −1
R = 8.314 Pa m K mol
−1 −1
= 8.314 J K mol
−1 −1
= 0.08314 L bar K mol
−1 −1
= 1.987 cal K mol
−1 −1
= 0.08205 L atm K mol
We also need the gas constant expressed per molecule rather than per mole. Since there is Avogadro’s number of molecules per
¯
¯¯¯
¯
mole, we can divide any of the values above by N to get R on a per-molecule basis. Traditionally, however, this constant is given a
different name; it is Boltzmann’s constant, usually given the symbol k .
¯
¯¯¯
¯ −23 −1 −1
k = R/ N = 1.381 × 10 J K molecule
This means that we can also write the ideal gas equation as P V = nRT = nN kT
¯
¯¯¯
¯
. Because the number of molecules in the
¯
¯¯¯
¯
sample, N , is N = nN , we have
P V = N kT .
This page titled 2.7: The Ideal Gas Constant and Boltzmann's Constant is shared under a CC BY-SA 4.0 license and was authored, remixed,
Now, we need to expand on the qualifications with which we begin this chapter. We imagine that the results of a large number of
experiments are available for our analysis. Our characterization of these results has been that all gases obey the same equations—
Boyle’s law, Charles’ law, and the ideal gas equation—and do so exactly. This is an oversimplification. In fact they are always
approximations. They are approximately true for all gases under all “reasonable” conditions, but they are not exactly true for any
real gas under any condition. It is useful to introduce the idea of hypothetical gases that obey the classical gas equations exactly. In
the previous section, we call the combination of Boyle’s law and Charles’ law the ideal gas equation. We call the hypothetical
substances that obey this equation ideal gases. Sometimes we refer to the classical gas laws collectively as the ideal gas laws.
At very high gas densities, the classical gas laws can be very poor approximations. As we have noted, they are better
approximations the lower the density of the gas. In fact, experiments show that the pressure—volume—temperature behavior of
any real gasreal gas becomes arbitrarily close to that predicted by the ideal gas equation in the limit as the pressure goes to zero.
This is an important observation that we use extensively.
At any given pressure and temperature, the ideal gas laws are better approximations for a compound that has a lower boiling point
than they are for a compound with a higher boiling point. Another way of saying this is that they are better approximations for
molecules that are weakly attracted to one another than they are for molecules that are strongly attracted to one another.
Forces between molecules cause them to both attract and repel one another. The net effect depends on the distance between them. If
we assume that there are no intermolecular forcesintermolecular forces acting between gas molecules, we can develop exact
theories for the behavior of macroscopic amounts of the gas. In particular, we can show that such substances obey the ideal gas
equation. (We shall see that a complete absence of repulsive forces implies that the molecules behave as point masses.) Evidently,
the difference between the behavior of a real gas and the behavior it would exhibit if it were an ideal gas is just a measure of the
effects of intermolecular forces.
The ideal gas equation is not the only equation that gives a useful representation for the interrelation of gas pressure–volume–
temperature data. There are many such equations of state. They are all approximations, but each can be a particularly useful
approximation in particular circumstances. We discuss van der Waal’s equation equation and the virial equations later in this
chapter. Nevertheless, we use the ideal gas equation extensively.
We will see that much of chemical thermodynamics is based on the behavior of ideal gases. Since there are no ideal gases, this may
seem odd, at best. If there are no ideal gases, why do we waste time talking about them? After all, we don’t want to slog through
tedious, long-winded, pointless digressions. We want to understand how real stuff behaves! Unfortunately, this is more difficult.
The charm of ideal gases is that we can understand their behavior; the ideal gas equation expresses this understanding in a
mathematical model. Real gases are another story. We can reasonably say that we can best understand the behavior of a real gas by
understanding how and why it is different from the behavior of a (hypothetical) ideal gas that has the same molecular structure.
This page titled 2.8: Real Gases Versus Ideal Gases is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
In Section 2.2 we suppose that we have a thermometer that we can use to measure the temperature of a gas. We suppose that this
thermometer uses a liquid, and we define an increase in temperature by the increase in the volume of this liquid. Our statement of
Charles’ law asserts that the volume of a gas is a linear function of the volume of the liquid in our thermometer, and that the same
linear function is observed for any gas. As we note in Section 2.8, there is a problem with this statement. Careful experiments with
such thermometers produce results that deviate from Charles’ law. With sufficiently accurate volume measurements, this occurs to
some extent for any choice of the liquid in the thermometer. If we make sufficiently accurate measurements, the volume of a gas is
not exactly proportional to the volume of any liquid (or solid) that we might choose as the working substance in our thermometer.
That is, if we base our temperature scale on a liquid or solid substance, we observe deviations from Charles’ law. There is a further
difficulty with using a liquid as the standard fluid on which to base our temperature measurements: temperatures outside the liquid
range of the chosen substance have to be measured in some other way.
Evidently, we can choose to use a gas as the working fluid in our thermometer. That is, our gas-volume measuring device is itself a
thermometer. This fact proves to be very useful because of a further experimental observation. To a very good approximation, we
find: If we keep the pressures in the thermometer and in some other gaseous system constant at low enough values, both gases
behave as ideal gases, and we find that the volumes of the two gases are proportional to each other over any range of temperature.
Moreover, this proportionality is observed for any choice of either gas. This means that we can define temperature in terms of the
expansion of any constant-pressure gas that behaves ideally. In principle, we can measure the same temperature using any gas, so
long as the constant operating pressure is low enough. When we do so, our device is called the ideal gas thermometer. In so far as
any gas behaves as an ideal gas at a sufficiently low pressure, any real gas can be used in an ideal gas thermometer and to measure
any temperature accurately. Of course, practical problems emerge when we attempt to make such measurements at very high and
very low temperatures.
The (very nearly) direct proportionality of two low-pressure real gas volumes contrasts with what we observe for liquids and solids.
In general, the volume of a given liquid (or solid) substance is not exactly proportional to the volume of a second liquid (or solid)
substance over a wide range of temperatures.
In practice, the ideal-gas thermometer is not as convenient to use as other thermometers—like the mercury-in-glass thermometer.
However, the ideal-gas thermometer is used to calibrate other thermometers. Of course, we have to calibrate the ideal-gas
thermometer itself before we can use it.
We do this by assigning a temperature of 273.16 K to the triple point of water. (It turns out that the melting point of ice isn’t
sufficiently reproducible for the most precise work. Recall that the triple point is the temperature and pressure at which all three
phases of water are at equilibrium with one another, with no air or other substances present. The triple-point pressure is 611 Pa or
6.03×10
−3
atm. See Section 6.3.) From both theoretical considerations and experimental observations, we are confident that no
system can attain a temperature below absolute zero. Thus, the size of the kelvin (one degree on the Kelvin scale) is fixed by the
3
difference in temperature between a system at the triple point of water and one at absolute zero. If our ideal gas thermometer has
volume V at thermal equilibrium with some other constant-temperature system, the proportionality of V and T means that
T 273.16
=
V V273.16
With the triple point fixed at 273.16 K, experiments find the freezing point of air-saturated water to be 273.15 K when the system
pressure is 1 atmosphere. (So the melting point of ice is 273.15 K, and the triple-point is 0.10 C. We will find two reasons for the
fact that the melting point is lower than the triple point: In Section 6.3 we find that the melting point of ice decreases as the
pressure increases. In Section 16.10 we find that solutes usually decrease the temperature at which the liquid and solid states of a
substance are in phase equilibrium.)
If we could use an ideal gas in our ideal-gas thermometer, we could be confident that we had a rigorous operational definition of
temperature. However, we note in Section 2.8 that any real gas will exhibit departures from ideal gas behavior if we make
sufficiently accurate measurements. For extremely accurate work, we need a way to correct the temperature value that we associate
with a given real-gas volume. The issue here is the value of the partial derivative
∂V
( )
∂T
P
For one mole of an ideal gas,
∂V R V
( ) = =
∂T P T
P
is a constant. For a real gas, it is a function of temperature. Let us assume that we know this function. Let the molar volume of the
real gas at the triple point of water be V and its volume at thermal equilibrium with a system whose true temperature is V be
273.16
V . We have
T
T VT
∂V
∫ ( ) dT = ∫ dV = VT − V273.16
273.16 ∂T P V273.16
When we know the integrand on the left as a function of temperature, we can do the integration and find the temperature
corresponding to any measured volume, V . T
When the working fluid in our thermometer is a real gas we make measurements to find (∂V /∂T ) as a function of temperature.
P
Here we encounter a circularity: To find (∂V /∂T ) from pressure-volume-temperature data we must have a way to measure
P
temperature; however, this is the very thing that we are trying to find.
In principle, we can surmount this difficulty by iteratively correcting the temperature that we associate with a given real-gas
volume. As a first approximation, we use the temperatures that we measure with an uncorrected real-gas thermometer. These
temperatures are a first approximation to the ideal-gas temperature scale. Using this scale, we make non-pressure-volume-
temperature measurements that establish (∂V /∂T ) as a function of temperature for the real gas. [This function is
P
∂V V +μ CP
JT
( ) =
∂T T
P
where C is the constant-pressure heat capacity and μ is the Joule-Thomson coefficient. Both are functions of temperature.
P JT
We introduce C in Section 7.9. We discuss the Joule-Thomson coefficient further in Section 2.10 below, and in detail in Section
P
10.14. Typically V ≫ C , and the value of (∂V /∂T ) is well approximated by V /T = R/P . With (∂V /∂T ) established using
P P P
this scale, integration yields a second-approximation to the ideal-gas temperatures. We could repeat this process until successive
temperature scales converge at the number of significant figures that our experimental accuracy can support.
In practice, there are several kinds of ideal-gas thermometers, and numerous corrections are required for very accurate
measurements. There are also numerous other ways to measure temperature, each of which has its own complications. Our
development has considered some of the ideas that have given rise to the concept that temperature is fundamental property of
4
nature that can be measured using a thermodynamic-temperature scale on which values begin at zero and increase to arbitrarily
high values. This thermodynamic temperature scale is a creature of theory, whose real-world counterpart would be the scale
established by an ideal-gas thermometer whose gas actually obeyed P V = nRT at all conditions. We have seen that such an ideal-
gas thermometer is itself a creature of theory.
The current real-world standard temperature scale is the International Temperature Scale of 1990 (ITS-90). This defines
temperature over a wide range in terms of the pressure-volume relationships of helium isotopes and the triple points of several
selected elements. The triple points fix the temperature at each of several conditions up to 1357.77 K (the freezing point of copper).
Needless to say, the temperatures assigned at the fixed points are the results of painstaking experiments designed to give the closest
possible match to the thermodynamic scale. A variety of measuring devices—thermometers—can be used to interpolate
temperature values between different pairs of fixed points.
This page titled 2.9: Temperature and the Ideal Gas Thermometer is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
We can derive Boyle’s law from Newtonian mechanics. This derivation assumes that gas molecules behave like point masses that
do not interact with one another. The pressure of the gas results from collisions of the gas molecules with the walls of the container.
The contribution of one collision to the force on the wall is equal to the change in the molecule’s momentum divided by the time
between collisions. The magnitude of this force depends on the molecule’s speed and the angle at which it strikes the wall. Each
such collision makes a contribution to the pressure that is equal to the force divided by the area of the wall. To find the pressure
from this model, it is necessary to average over all possible molecular speeds and all possible collision angles. In Chapter 4, we
derive Boyle’s law in this way.
We can do a simplified derivation by making a number of assumptions. We assume that all of the molecules in a sample of gas have
the same speed. Let us call it u. As sketched in Figure 3, we assume that the container is a cubic box whose edge length is d . If we
consider all of the collisions between molecules and walls, it is clear that each wall will experience 1/6 of the collisions; or, each
pair of opposing walls will experience 1/3 of the collisions. Instead of averaging over all of the possible angles at which a
molecule could strike a wall and all of the possible times between collisions, we assume that the molecules travel at constant speed
back and forth between opposite faces of the box. Since they are point masses, they never collide with one another. If we suppose
that 1/3 of the molecules go back and forth between each pair of opposite walls, we can expect to accomplish the same kind of
averaging in setting up our artificial model that we achieve by averaging over the real distribution of angles and speeds. In fact, this
turns out to be the case; the derivation below gets the same result as the rigorous treatment we develop in Chapter 4.
Figure 3. Simplified model for velocities of gas molecules in a cubic box.

Since each molecule goes back and forth between opposite walls, it collides with each wall once during each round trip. At each
collision, the molecule’s speed remains constant, but its direction changes by 180 ; that is, the molecule’s velocity changes from ⇀
o
u
to − u . Letting Δt be the time required for a round trip, the distance traversed in a round trip is
⇀
⇀
2d = ∣
∣ u ∣
∣ Δt
= uΔt
The magnitude of the momentum change for a molecule in one collision is

∣ ⇀ ⇀ ⇀
∣Δ(m u )∣
∣ =∣
∣m u f inal − m u initial ∣
∣
⇀ ⇀
=∣
∣m u f inal − (−m u f inal )∣
∣
= 2mu
The magnitude of the force on the wall from one collision is
∣ ⇀ 2
∣Δ (m u )∣
∣ 2mu mu
F = = =
Δt (2d/u) d
and the pressure contribution from one collision on the wall, of area d , is 2
2 2 2
F mu mu mu
P = = = =
2 3
A d∙d d V
so that we have
2
P V = mu
from the collision of one molecule with one wall.

If the number of molecules in the box is N , N /3 of them make collisions with this wall, so that the total pressure on one wall
attributable to all N molecules in the box is
2
mu N
P =
V 3
or
2
N mu
PV =
3
Since the ideal gas equation can be written as P V = N kT we see that N mu 2

/3 = N kT so that mu 2
= 3kT and
−−−−
3kT
u =√
m
Thus we have found a relationship between the molecular speed and the temperature of the gas. (The actual speed of a molecule, v ,
can have any value between zero and—for present purposes—infinity. When we average the values of v for many molecules, we 2
¯
¯¯¯
¯ ¯
¯¯¯
¯
find the average value of the squared speeds, v
2
. In Chapter 4, we find that u
2
=v
2
. That is, the average speed we use in our
−
−
¯
¯¯¯
¯
derivation turns out to be a quantity called the root-mean-square speed, v rms =u =
√ 2
v .) This result also gives us the (average)
kinetic energy of a single gas molecule:
2
mu 3kT
KE = =
2 2
From this derivation, we have a simple mechanical model that explains Boyle’s law as the logical consequence of point-mass
molecules colliding with the walls of their container. By combining this result with the ideal gas equation, we find that the average
speed of ideal gas molecules depends only on the temperature. From this we have the very important result that the translational
kinetic energy of an ideal gas depends only on temperature.
Since our non-interacting point-mass molecules have no potential energy arising from their interactions with one another, their
translational kinetic energy is the whole of their energy. (Because two such molecules neither attract nor repel one another, no work
is required to change the distance between them. The work associated with changing the volume of a confined sample of an ideal
gas arises because of the pressure the molecules exert on the walls of the container; the pressure arises because of the molecules’
kinetic energy.) The energy of one mole of monatomic ideal gas molecules is
KE = (3/2) RT
When we expand our concept of ideal gases to include molecules that have rotational or vibrational energy, but which neither
attract nor repel one another, it remains true that the energy of a macroscopic sample depends only on temperature. However, the
molar energy of such a gas is greater than (3/2) RT , because of the energy associated with these additional motions.
We make extensive use of the conclusion that the energy of an ideal gas depends only on temperature. As it turns out, this
conclusion follows rigorously from the second law of thermodynamics. In Chapter 10, we show that
∂E ∂E
( ) =( ) =0
∂V ∂P
T T
for a substance that obeys the ideal gas equation; at constant temperature, the energy of an ideal gas is independent of the volume
and independent of the pressure. So long as pressure, volume, and temperature are the only variables needed to specify its state, the
laws of thermodynamics imply that the energy of an ideal gas depends only on temperature.
While the energy of an ideal gas is independent of pressure, the energy of a real gas is a function of pressure at a given temperature.
At ordinary pressures and temperatures, this dependence is weak and can often be neglected. The first experimental investigation of
this issue was made by James Prescott Joule, for whom the SI unit of energy is named. Beginning in 1838, Joule did a long series
of careful measurements of the mechanical equivalent of heat. These measurements formed the original experimental basis for the
kinetic theory of heat. Among Joule’s early experiments was an attempt to measure the heat absorbed by a gas as it expanded into
an evacuated container, a process known as a free expansion. No absorption of heat was observed, which implied that the energy of
the gas was unaffected by the volume change. However, it is difficult to do this experiment with meaningful accuracy.
Subsequently, Joule collaborated with William Thomson (Lord Kelvin) on a somewhat different experimental approach to
essentially the same question. The Joule-Thomson experiment provides a much more sensitive measure of the effects of
intermolecular forces of attraction and repulsion on the energy of a gas during its expansion. Since our definition of an ideal gas
includes the stipulation that there are no intermolecular forces, the Joule-Thomson experiment is consistent with the conclusion that
the energy of an ideal gas depends only on temperature. However, since intermolecular forces are not zero for any real gas, our
analysis reaches this conclusion in a somewhat indirect way. The complication arises because the Joule-Thomson results are not
entirely consistent with the idea that all properties of a real gas approach those of an ideal gas at a sufficiently low pressure. (The
best of models can have limitations.) We discuss the Joule-Thomson experiment in Section 10.14.
This page titled 2.10: Deriving Boyle's Law from Newtonian Mechanics is shared under a CC BY-SA 4.0 license and was authored, remixed,
We can measure the pressure of the atmosphere at any location by using a barometer. A mercury barometer is a sealed tube that
contains a vertical column of liquid mercury. The space in the tube above the liquid mercury is occupied by mercury vapor. Since
the vapor pressure of liquid mercury at ordinary temperatures is very low, the pressure at the top of the mercury column is very low
and can usually be ignored. The pressure at the bottom of the column of mercury is equal to the pressure of a column of air
extending from the elevation of the barometer all the way to the top of the earth’s atmosphere. As we take the barometer to higher
altitudes, we find that the height of the mercury column decreases, because less and less of the atmosphere is above the barometer.
Figure 4. Atmospheric pressure versus altitude.

If we assume that the atmosphere is composed of an ideal gas and that its temperature is constant, we can derive an equation for
atmospheric pressure as a function of altitude. Imagine a cylindrical column of air extending from the earth’s surface to the top of
the atmosphere (Figure 4). The force exerted by this column at its base is the weight of the air in the column; the pressure is this
weight divided by the cross-sectional area of the column. Let the cross-sectional area of the column be A .
Consider a short section of this column. Let the bottom of this section be a distance h from the earth’s surface, while its top is a
distance h + Δh from the earth’s surface. The volume of this cylindrical section is then V = AΔh . Let the mass of the gas in this
S
section be M . The pressure at h + Δh is less than the pressure at h by the weight of this gas divided by the cross-sectional area.
S
The weight of the gas is M g . The pressure difference is ΔP = −M g/A . We have

S S
P (h + Δh) − P (h) ΔP −MS g −MS g

= = =
Δh Δh AΔh VS
Since we are assuming that the sample of gas in the cylindrical section behaves ideally, we have V S = nS RT /P . Substituting for
VS and taking the limit as Δh → 0 , we find
¯¯¯¯
¯
dP −M S g −n S M g −mg
=( )P =( )P =( )P
dh nS RT nS RT kT
where we introduce nS as the number of moles of gas in the sample, ¯¯¯¯

M
¯
as the molar mass of this gas, and m as the mass of an
¯¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
individual atmosphere molecule. The last equality on the right makes use of the identities M = mN and R = Nk . Separating
variables and integrating between limits P (0) = P and P (h) = P , we find
0
P h
dP −mg
∫ =( )∫ dh
P0 P kT 0
so that
P −mgh
ln( ) =
P0 kT
and
−mgh
P = P0 exp ( )
kT
Either of the latter relationships is frequently called the barometric formula.

If we let η be the number of molecules per unit volume, η = N /V , we can write P = N kT /V = ηkT and P 0 = η0 kT so that the
barometric formula can be expressed in terms of these number densities as
−mgh
η = η0 exp ( )
kT
This page titled 2.11: The Barometric Formula is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
An equation due to van der Waals extends the ideal gas equation in a straightforward way. Van der Waals’ equation is
2
an
(P + ) (V − nb) = nRT
2
V
It fits pressure-volume-temperature data for a real gas better than the ideal gas equation does. The improved fit is obtained by
introducing two parameters (designated “a ” and “b ”) that must be determined experimentally for each gas. Van der Waals’ equation
is particularly useful in our effort to understand the behavior of real gases, because it embodies a simple physical picture for the
difference between a real gas and an ideal gas.
In deriving Boyle’s law from Newton’s laws, we assume that the gas molecules do not interact with one another. Simple arguments
show that this can be only approximately true. Real gas molecules must interact with one another. At short distances they repel one
another. At somewhat longer distances, they attract one another. The ideal gas equation can also be derived from the basic
assumptions that we make in §10 by an application of the theory of statistical thermodynamics. By making different assumptions
about molecular properties, we can apply statistical thermodynamics to derive van der Waals’ equation. The required assumptions
5
are that the molecules occupy a finite volume and that they attract one another with a force that varies as the inverse of a power of
the distance between them. (The attractive force is usually assumed to be proportional to r .) −6
To recognize that real gas molecules both attract and repel one another, we need only remember that any gas can be liquefied by
reducing its temperature and increasing the pressure applied to it. If we cool the liquid further, it freezes to a solid. Now, two
distinguishing features of a solid are that it retains its shape and that it is almost incompressible. We attribute the incompressibility
of a solid to repulsive forces between its constituent molecules; they have come so close to one another that repulsive forces
between them have become important. To compress the solid, the molecules must be pushed still closer together, which requires
inordinate force. On the other hand, if we throw an ice cube across the room, all of its constituent water molecules fly across the
room together. Evidently, the water molecules in the solid are attracted to one another, otherwise they would all go their separate
ways—throwing the ice cube would be like throwing a handful of dry sand. But water molecules are the same molecules whatever
the temperature or pressure, so if there are forces of attraction and repulsion between them in the solid, these forces must be present
in the liquid and gas phases also.
In the gas phase, molecules are far apart; in the liquid or the solid phase, they are packed together. At its boiling point, the volume
of a liquid is much less than the volume of the gas from which it is condensed. At the freezing point, the volume of a solid is only
slightly different from the volume of the liquid from which it is frozen, and it is certainly greater than zero. These commonplace
observations are readily explained by supposing that any molecule has a characteristic volume. We can understand this, in turn, to
be a consequence of the nature of the intermolecular forces; evidently, these forces become stronger as the distance between a pair
of molecules decreases. Since a liquid or a solid occupies a definite volume, the repulsive force must increase more rapidly than the
attractive force when the intermolecular distance is small. Often it is useful to talk about the molar volume of a condensed phase.
By molar volume, we mean the volume of one mole of a pure substance. The molar volume of a condensed phase is determined by
the intermolecular distance at which there is a balance between intermolecular forces of attraction and repulsion.
Evidently molecules are very close to one another in condensed phases. If we suppose that the empty spaces between molecules are
negligible, the volume of a condensed phase is approximately equal to the number of molecules in the sample multiplied by the
volume of a single molecule. Then the molar volume is Avogadro’s number times the volume occupied by one molecule. If we
know the density, D, and the molar mass, M , we can find the molar volume, V , as
¯¯¯¯
¯ ¯
¯¯¯
¯¯¯¯
¯
¯
¯¯¯ M
V =
D
The volume occupied by a molecule, V molecule , becomes

¯
¯¯¯
V
Vmolecule =
¯
¯¯¯
¯
N
The pressure and volume appearing in van der Waals’ equation are the pressure and volume of the real gas. We can relate the terms
in van der Waals’ equation to the ideal gas equation: It is useful to think of the terms (P + an /V ) and (V − nb) as the pressure
2 2
and volume of a hypothetical ideal gas. That is
2
an
Pideal gas Videal gas = ( Preal gas + ) (Vreal gas − nb)
2
V
real gas
= nRT
Then we have
Vreal gas = Videal gas + nb
We derive the ideal gas equation from a model in which the molecules are non-interacting point masses. So the volume of an ideal
gas is the volume occupied by a gas whose individual molecules have zero volume. If the individual molecules of a real gas
effectively occupy a volume b/N , then n moles of them effectively occupy a volume
¯
¯¯¯
¯
¯
¯¯¯
¯ ¯
¯¯¯
¯
(b/ N ) (nN ) = nb.
Van der Waals’ equation says that the volume of a real gas is the volume that would be occupied by non-interacting point masses,
V , plus the effective volume of the gas molecules themselves. (When data for real gas molecules are fit to the van der
ideal gas
Waals’ equation, the value of b is usually somewhat greater than the volume estimated from the liquid density and molecular
weight. See problem 24.)
Similarly, we have
2
an
Preal gas = Pideal gas −
2
V
real gas
We can understand this as a logical consequence of attractive interactions between the molecules of the real gas. With a > 0 , it
says that the pressure of the real gas is less than the pressure of the hypothetical ideal gas, by an amount that is proportional to
2
(n/V ) . The proportionality constant is a . Since n/V is the molar density (moles per unit volume) of the gas molecules, it is a
measure of concentration. The number of collisions between molecules of the same kind is proportional to the square of their
concentration. (We consider this point in more detail in Chapters 4 and 5.) So (n/V ) is a measure of the frequency with which the
2
real gas molecules come into close contact with one another. If they attract one another when they come close to one another, the
effect of this attraction should be proportional to (n/V ) . So van der Waals’ equation is consistent with the idea that the pressure
2
of a real gas is different from the pressure of the hypothetical ideal gas by an amount that is proportional to the frequency and
strength of attractive interactions.
But why should attractive interactions have this effect; why should the pressure of the real gas be less than that of the hypothetical
ideal gas? Perhaps the best way to develop a qualitative picture is to recognize that attractive intermolecular forces tend to cause
the gas molecules to clump up. After all, it is these attractive forcesattractive force that cause the molecules to aggregate to a liquid
at low temperatures. Above the boiling point, the ability of gas molecules to go their separate ways limits the effects of this
tendency; however, even in the gas, the attractive forces must act in a way that tends to reduce the volume occupied by the
molecules. Since the volume occupied by the gas is dictated by the size of the container—not by the properties of the gas itself—
this clumping-up tendency finds expression as a decrease in pressure.
Figure 5. Potential energy versus distance for "hard sphere molecules."
It is frequently useful to describe the interaction between particles or chemical moieties in terms of a potential energy versus
distance diagram. The van der Waals’ equation corresponds to the case that the repulsive interaction between molecules is non-
existent until the molecules come into contact. Once they come into contact, the energy required to move them still closer together
becomes arbitrarily large. Often this is described by saying that they behave like “hard spheres”. The attractive force between two
molecules decreases as the distance between them increases. When they are very far apart the attractive interaction is very small.
We say that the energy of interaction is zero when the molecules are infinitely far apart. If we initially have two widely separated,
stationary, mutually attracting molecules, they will spontaneously move toward one another, gaining kinetic energy as they go.
Their potential energy decreases as they approach one another, reaching its smallest value when the molecules come into contact.
Thus, van der Waals’ equation implies the potential energy versus distance diagram sketched in Figure 5.
This page titled 2.12: Van der Waals' Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
It is often useful to fit accurate pressure-volume-temperature data to polynomial equations. The experimental data can be used to
compute a quantity called the compressibility factor, Z , which is defined as the pressure–volume product for the real gas divided
by the pressure–volume product for an ideal gas at the same temperature.
We have
(P V ) = nRT
ideal gas
¯
¯¯¯
Letting P and V represent the pressure and volume of the real gas, and introducing the molar volume, V = V /n , we have
¯
¯¯¯
(P V )real gas PV PV
Z = = =
(P V ) nRT RT
ideal gas
Since Z = 1 if the real gas behaves exactly like an ideal gas, experimental values of Z will tend toward unity under conditions in
which the density of the real gas becomes low and its behavior approaches that of an ideal gas. At a given temperature, we can
−1
¯
¯¯¯
conveniently ensure that this condition is met by fitting the Z values to a polynomial in P or a polynomial in V . The coefficients
are functions of temperature. If the data are fit to a polynomial in the pressure, the equation is
∗ ∗ 2 ∗ 3
Z = 1 +B (T ) P + C (T ) P +D (T ) P +…
−1
¯
¯¯¯
For a polynomial in V , the equation is
B (T ) C (T ) D (T )
Z =1+ + + +…
¯
¯¯¯ 2 3
¯
¯¯¯ ¯
¯¯¯
V V V
These empirical equations are called virial equations. As indicated, the parameters are functions of temperature. The values of
B (T ) , C (T ), D (T ), …, and B (T ) , C (T ), D (T ),…, must be determined for each real gas at every temperature. (Note also
∗ ∗ ∗
that B (T ) ≠ B (T ) , C (T ) ≠ C (T ) , D (T ) ≠ D (T ) , etc. However, it is true that B = B/RT .) Values for these parameters
∗ ∗ ∗ ∗
are tabulated in various compilations of physical data. In these tabulations, B (T ) and C (T ) are called the second virial coefficient
and third virial coefficient, respectively.
This page titled 2.13: Virial Equations is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
Thus far, our discussion of the properties of a gas has implicitly assumed that the gas is pure. We turn our attention now to mixtures
of gases—gas samples that contain molecules of more than one compound. Mixtures of gases are common, and it is important to
understand their behavior in terms of the properties of the individual gases that make it up. The ideal-gas laws we have for mixtures
are approximations. Fortunately, these approximations are often very good. When we think about it, this is not surprising. After all,
the distinguishing feature of a gas is that its molecules do not interact with one another very much. Even if the gas is composed of
molecules of different kinds, the unimportance of molecule—molecule interactions means that the properties of one kind of
molecules should be nearly independent of the properties of the other kinds.
Consider a sample of gas that contains a fixed number of moles of each of two or more compounds. This sample has a pressure, a
volume, a temperature, and a specified composition. Evidently, the challenge here is to describe the pressure, volume, and
temperature of the mixture in terms of measurable properties of the component compounds.
There is no ambiguity about what we mean by the pressure, volume, and temperature of the mixture; we can measure these
properties without difficulty. Given the nature of temperature, it is both reasonable and unambiguous to say that the temperature of
the sample and the temperature of its components are the same. However, we cannot measure the pressure or volume of an
individual component in the mixture. If we hope to describe the properties of the mixture in terms of properties of the components,
we must first define some related quantities that we can measure. The concepts of a component partial pressure and a component
partial volume meet this need.
We define the partial pressure of a component of a gas mixture as the pressure exerted by the same number of moles of the pure
component when present in the volume occupied by the mixture, V , at the temperature of the mixture. In a mixture of n
mixture A
moles of component A , n moles of component B , etc., it is customary to designate the partial pressure of component A as P . It
B A
is important to appreciate that the partial pressure of a real gas can only be determined by experiment.
We define the partial volume of a component of a gas mixture as the volume occupied by the same number of moles of the pure
component when the pressure is the same as the pressure of the mixture, P , at the temperature of the mixture. In a mixture of
mixture
components A , B , etc., it is customary to designate the partial volume of component A as V . The partial volume of a real gas can
A
only be determined by experiment.

Dalton’s law of partial pressures asserts that the pressure of a mixture is equal to the sum of the partial pressures of its
components. That is, for a mixture of components A, B, C, etc., the pressure of the mixture is
Pmixture = PA + PB + PC + … (2.14.1)
Under conditions in which the ideal gas law is a good approximation to the behavior of the individual components, Dalton’s law is
usually a good approximation to the behavior of real gas mixtures. For mixtures of ideal gases, it is exact. To see this, we recognize
that, for an ideal gas, the definition of partial pressure becomes
nA RT
PA =
Vmixture
The ideal-gas mixture contains n mixture = nA + nB + nC + … moles , so that

nmixture RT
Pmixture =
Vmixture
(nA + nB + nC + …) RT
=
Vmixture
nA RT nB RT nC RT
= + + +…
Vmixture Vmixture Vmixture
= PA + PB + PC + …
Applied to the mixture, the ideal-gas equation yields Dalton’s law (Equation 2.14.1). When xA is the mole fraction of A in a
mixture of ideal gases,
PA = xA Pmixture .
This page titled 2.14: Gas Mixtures - Dalton's Law of Partial Pressures is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Amagat’s law of partial volumes asserts that the volume of a mixture is equal to the sum of the partial volumes of its
components. For a mixture of components A , B , C , etc., Amagat’s law gives the volume as
Vmixture = VA + VB + VC + …
For real gases, Amagat’s law is usually an even better approximation than Dalton’s law . Again, for mixtures of ideal gases, it is
6
exact. For an ideal gas, the partial volume is

nA RT
VA =
Pmixture
Since n mixture = nA + nB + nC + … , we have, for a mixture of ideal gases,

nmixture RT
Vmixture =
Pmixture
(nA + nB + nC + …) RT
=
Pmixture
= VA + VB + VC + …
Applied to the mixture, the ideal-gas equation yields Amagat’s law. Also, we have V A = xA Vmixture .
This page titled 2.15: Gas Mixtures - Amagat's Law of Partial Volums is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
2.16: Problems
1. If A is an ideal gas in a mixture of ideal gases, prove that its partial pressure, P , is given by P = x P
A A A. mixture
2. If A is an ideal gas in a mixture of ideal gases, prove that its partial volume, V , is given by V = x V
A A .
A mixture
3. A sample of hydrogen chloride gas, H C l, occupies 0.932 L at a pressure of 1.44 bar and a temperature of 50 C. The sample is
dissolved in 1 L of water. What is the resulting hydronium ion, H O , concentration?
3
+
4. Ammonia gas, N H , also dissolves quantitatively in water. If it is measured at 0.720 bar and 50 C, what volume of N H gas is
3 3
required to neutralize the solution prepared in problem 3? For present purposes, assume that the neutralization reaction occurs
quantitatively.
5. Two pressure vessels are separated by a closed valve. One contains 10.0 moles of helium, H e, at 5.00 bar. The other contains
5.00 moles of neon, N e , at 20.0 bar. Both vessels are at the same temperature. The valve is opened and the gases are allowed to
mix. The temperature remains constant. What is the final pressure?
6. What is the average velocity of a molecule of nitrogen, N , at 300 K? Of a molecule of hydrogen, H , at the same temperature?
2 2
7. The Homestake gold mine near Lead, South Dakota, is excavated to 8000 feet below the surface. Lead is nearly a mile high; the
bottom of the Homestake is about 900 m below sea level. Nearby Custer Peak is about 2100 m above sea level. What is the
ratio of the barometric pressure on top of Custer Peak to the barometric pressure at the bottom of the Homestake? Assume that
the entire atmosphere is at 300 K and that it behaves as a single ideal gas whose molar mass is 29.
8. On the sidewalk in front of a tall building, the barometric pressure is 740 torr and the temperature is 25 C. On the roof of this
building, the barometric pressure is 732 torr. Assuming that the entire atmosphere behaves as an ideal gas of molecular weight
29 at 25 C, estimate the height of the building. Comment on the likely accuracy of this estimate.
9. At 1 bar, the boiling point of water is 372.78 K. At this temperature and pressure, the density of liquid water is 958.66 kg m −− 3
and that of gaseous water is 0.59021 kg m −−3

. What are the molar volumes, in m mol , of liquid and gaseous water at this
3 −1
temperature and pressure? In L mol ? −1
10. Refer to your results in Problem 9. Assuming that a water molecule excludes other water molecules from a cubic region
centered on itself, estimate the average distance between nearest-neighbor water molecules in the liquid and in the gas.
11. Calculate the molar volume of gaseous water at 1 bar and 372.78 K from the ideal gas equation. What is the error, expressed as
a percentage of the value you calculated in Problem 9?
12. At 372.78 K, the virial coefficient B* for water is −1.487 × 10 Pa . Calculate the molar volume of gaseous water at 1 bar
−7 −1
¯
¯¯¯
and 372.78 K from the virial equation: Z = P V /RT = 1 + B P . What is the error, expressed as a percentage of the value
∗
you calculated in Problem 9?

13. Calculate the molar volume of gaseous water at 1 bar and 372.78 K from van der Waals’ equation. The van der Waals’
parameters for water are a = 5.537 bar L mol and b = 0.0305 L mol . What is the error, expressed as a percentage of
2 −1 −1
the value you calculated in Problem 9?

14. Comment on the results in Problems 11 – 13. At this temperature, would you expect the accuracy to increase or decrease at
lower pressures?
15. The critical temperature for water is 647.1 K. At 10 bar and 700 K, the density of supercritical water is 651.37 kg m . Note
3 3
that this is about 68% of the value for liquid water at the boiling point at 1 bar. What is the molar volume, in m mol , of 3 −1
water at this temperature and pressure? In L mol ? −1
16. Refer to your results in Problem 15. Assuming that a water molecule excludes other water molecules from a cubic region
centered on itself, estimate the average distance between nearest-neighbor water molecules in supercritical water at 10 bar and 3
700 K.
17. Calculate the molar volume of supercritical water at 10 bar and 700 K from the ideal gas equation. What is the error, expressed
3
as a percentage of the value you calculated in Problem 15?

18. At 700 K, the virial coefficient B* for water is −1.1512 × 10 Pa Calculate the molar volume of supercritical water at 10
−8 −1 3
bar and 700 K from the virial equation. (See Problem 12.) What is the error, expressed as a percentage of the value you
calculated in Problem 15?
19. Calculate the molar volume of supercritical water at 10 bar and 700 K from van der Waals’ equation. (See Problem 13.) What
3
is the error, expressed as a percentage of the value you calculated in Problem 15?
20. Comment on the results in Problems 16 – 19.
21. Comment on the results in Problems 10 – 13 versus the results in Problems 16 – 19.
22. A 1.000 L combustion bomb is filled with natural gas at 2.00 bar and 300 K. Pure oxygen is then pressured into the bomb until
the pressure reaches 7.00 bar, at 300 K. Combustion is initiated. When reaction is complete, the bomb is thermostatted at 500 K,
and the pressure is measured to be 12.08 bar. Thereafter, the bomb is cooled to 260 K, so that all of the water freezes. The
pressure is then found to be 2.812 bar. The natural gas is a mixture of helium, methane, and ethane. How many moles of each
gas are in the original sample?
23. An unknown liquid compound boils at 124 C. A classical method is used to find the approximate molecular weight of this
compound. This method uses a glass bulb whose only opening is a long thin capillary tube, so that a gas sample inside the bulb
can mix with the air outside only slowly. Filled with water, the bulb weighs 102.7535 grams. Empty, it weighs 50.0230 grams.
A quantity of the unknown liquid is put into the bulb, and the body of the bulb is immersed in an oil bath at 150 C. The end of
the capillary tube extends out of the oil bath. The liquid vaporizes filling the bulb with its gas. The total amount of vapor
generated is large compared to the volume of the bulb, so the escaping vapor effectively sweeps all of the air out of the bulb,
leaving the bulb filled with just the vapor of the unknown compound. When the last drop of liquid has just vaporized, the bulb
is filled with the vapor of the unknown substance at the ambient atmosphere pressure, which is 0.980 bar, and a temperature of
150 C. The bulb is then removed from the oil bath and allowed to cool quickly so that the vapor condenses to a liquid film on
the inside of the bulb. The oil is cleaned from the outside of the bulb, and the bulb is reweighed. The bulb and the liquid inside
weigh 50.1879 grams. What is the approximate molecular weight of the liquid?
24. From the data below, calculate the molar volume, in liters, of each substance. For each substance, divide van der Waals’ b by
the molar volume you calculate. Comment.
–1 –1 –1
Compound Mol Mass, g mol Density, g mL Van der Waals b, L mol
Acetic acid 60.05 1.0491 0.10680
Acetone 58.08 0.7908 0.09940
Acetonitrile 41.05 0.7856 0.11680
Ammonia 17.03 0.7710 0.03707
Aniline 93.13 1.0216 0.13690
Benzene 78.11 0.8787 0.11540
Benzonitrile 103.12 1.0102 0.17240
iso-Butylbenzene 134.21 0.8621 0.21440
Chlorine 70.91 3.2140 0.05622
Durene 134.21 0.8380 0.24240
Ethane 30.07 0.5720 0.06380
Hydrogen chloride 36.46 1.1870 0.04081
Mercury 200.59 13.5939 0.01696
Methane 16.04 0.4150 0.04278
Nitrogen dioxide 46.01 1.4494 0.04424
Silicon tetrafluoride 104.08 1.6600 0.05571
Water 18.02 1.0000 0.03049
Notes
1 ¯
¯¯¯
¯
We use the over-bar to indicate that the quantity is per mole of substance. Thus, we write N to indicate the number of particles per
¯¯¯¯
¯
mole. We write M to represent the gram molar mass. In Chapter 14, we introduce the use of the over-bar to denote a partial molar
quantity; this is consistent with the usage introduced here, but carries the further qualification that temperature and pressure are
constant at specified values. We also use the over-bar to indicate the arithmetic average; such instances will be clear from the
context.
2The unit of temperature is named the kelvin, which is abbreviated as K.
3
A redefinition of the size of the unit of temperature, the kelvin, is under consideration. The practical effect will be inconsequential
for any but the most exacting of measurements.
4
For a thorough discussion of the development of the concept of temperature, the evolution of our means to measure it, and the
philosophical considerations involved, see Hasok Chang, Inventing Temperature, Oxford University Press, 2004.
5
See T. L. Hill, An Introduction to Statistical Thermodynamics, Addison-Wesley Publishing Company, 1960, p 286.
6
See S. M. Blinder, Advanced Physical Chemistry, The Macmillan Company, Collier-Macmillan Canada, Ltd., Toronto, 1969, pp
185-189
CHAPTER OVERVIEW
3: Distributions, Probability, and Expected Values

3.6: Continuous Distribution Functions - the Envelope Function is the Derivative of the Area
3.13: The Expected Value of a Function of Several Variables and the Central Limit Theorem
3.15: Problems
This page titled 3: Distributions, Probability, and Expected Values is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
1
In Section 2.10, we derive Boyle’s law from Newton’s laws using the assumption that all gas molecules move at the same speed at
a given temperature. This is a poor assumption. Individual gas molecules actually have a wide range of velocities. In Chapter 4, we
derive the Maxwell–Boltzmann distribution law for the distribution of molecular velocities. This law gives the fraction of gas
molecules having velocities in any range of velocities. Before developing the Maxwell–Boltzmann distribution law, we need to
develop some ideas about distribution functions. Most of these ideas are mathematical. We discuss them in a non-rigorous way,
focusing on understanding what they mean rather than on proving them.
The overriding idea is that we have a real-world source of data. We call this source of data the distribution. We can collect data
from this source to whatever extent we please. The datum that we collect is called the distribution’s random variable. We call each
possible value of the random variable an outcome. The process of gathering a set of particular values of the random variable from a
distribution is often called sampling or drawing a sample. The set of values that is collected is called the sample. The set of values
that comprise the sample is often called “the data.” In scientific applications, the random variable is usually a number that results
from making a measurement on a physical system. Calling this process “drawing a sample” can be inappropriate. Often we call the
process of getting a value for the random variable “doing an experiment”, “doing a test”, or “making a trial”.
As we collect increasing amounts of data, the accumulation quickly becomes unwieldy unless we can reduce it to a mathematical
model. We call the mathematical model we develop a distribution function, because it is a function that expresses what we are able
to learn about the data source—the distribution. A distribution function is an equation that summarizes the results of many
measurements; it is a mathematical model for a real-world source of data. Specifically, it models the frequency of an event with
which we obtain a particular outcome. We usually believe that we can make our mathematical model behave as much like the real-
world data source as we want if we use enough experimental data in developing it.
Often we talk about statistics. By a statistic, we mean any mathematical entity that we can calculate from data. Broadly speaking a
distribution function is a statistic, because it is obtained by fitting a mathematical function to data that we collect. Two other
statistics are often used to characterize experimental data: the mean and the variance. The mean and variance are defined for any
distribution. We want to see how to estimate the mean and variance from a set of experimental data collected from a particular
distribution.
We distinguish between discrete and continuous distributions. A discrete distribution is a real-world source of data that can
produce only particular data values. A coin toss is a good example. It can produce only two outcomes—heads or tails. A
continuous distribution is a real-world source of data that can produce data values in a continuous range. The speed of an
automobile is a good example. An automobile can have any speed within a rather wide range of speeds. For this distribution, the
random variable is automobile speed. Of course we can generate a discrete distribution by aggregating the results of sampling a
continuous distribution; if we lump all automobile speeds between 20 mph and 30 mph together, we lose the detailed information
about the speed of each automobile and retain only the total number of automobiles with speeds in this interval.
This page titled 3.1: The Distribution Function as a Summary of Experimental Results is shared under a CC BY-SA 4.0 license and was authored,
We also need to introduce the idea that a function that successfully models the results of past experiments can be used to predict
some of the characteristics of future results.
We reason as follows: We have results from drawing many samples of a random variable from some distribution. We suppose that a
mathematical representation has been found that adequately summarizes the results of these experiences. If the underlying
distribution—the physical system in scientific applications—remains the same, we expect that a long series of future results would
give rise to essentially the same mathematical representation. If 25% of many previous results have had a particular characteristic,
we expect that 25% of a large number of future trials will have the same characteristic. We also say that there is one chance in four
that the next individual result will have this characteristic; when we say this, we mean that 25% of a large number of future trials
will have this characteristic, and the next trial has as good a chance as any other to be among those that do. The probability that an
outcome will occur in the future is equal to the frequency with which that outcome has occurred in the past.
Given a distribution, the possible outcomes must be mutually exclusive; in any given trial, the random variable can have only one
of its possible values. Consequently, a discrete distribution is completely described when the probability of each of its outcomes is
specified. Many distributions are comprised of a finite set of N mutually exclusive possible outcomes. If each of these outcomes is
equally likely, the probability that we will observe any particular outcome in the next trial is 1/N .
We often find it convenient to group the set of possible outcomes into subsets in such a way that each outcome is in one and only
one of the subsets. We say that such assignments of outcomes to subsets are exhaustive, because every possible outcome is
assigned to some subset; we say that such assignments are mutually exclusive, because no outcome belongs to more than one
subset. We call each such subset an event. When we partition the possible outcomes into exhaustive and mutually exclusive events,
we can say the same things about the probabilities of events that we can say about the probabilities of outcomes. In our discussions,
the term “events” will always refer to an exhaustive and mutually exclusive partitioning of the possible outcomes. Distinguishing
between outcomes and events just gives us some language conventions that enable us to create alternative groupings of the same set
of real world observations.
Suppose that we define a particular event to be a subset of outcomes that we denote as U. If in a large number of trials, the fraction
of outcomes that belong to this subset is F, we say that the probability is F that the outcome of the next trial will belong to this
event. To express this in more mathematical notation, we write P (U ) = F . When we do so, we mean that the fraction of a large
number of future trials that belong to this subset will be F, and the next trial has as good a chance as any other to be among those
that do. In a sample comprising M observations, the best forecast we can make of the number of occurrences of U is M × P (U ) ,
and we call this the expected number of occurrences of U in a sample of size M.
The idea of grouping real world observations into either outcomes or events is easy to remember if we keep in mind the example of
tossing a die. The die has six faces, which are labeled with 1, 2, 3, 4, 5, or 6 dots. The dots distinguish one face from another. On
any given toss, one face of the die must land on top. Therefore, there are six possible outcomes. Since each face has as good a
chance as any other of landing on top, the six possible outcomes are equally probable. The probability of any given outcome is 1/6.
If we ask about the probability that the next toss will result in one of the even-numbered faces landing on top, we are asking about
the probability of an event—the event that the next toss will have the characteristic that an even-numbered face lands on top. Let us
call this event X. That is, event X occurs if the outcome is a 2, a 4, or a 6. These are three of the six equally likely outcomes.
Evidently, the probability of this event is 3/6 = 1/2.
Having defined event X as the probability of an even-number outcome, we still have several alternative ways to assign the odd-
number outcomes to events. One assignment would be to say that all of the odd-number outcomes belong to a second event—the
event that the outcome is odd. The events “even outcome” and “odd outcome” are exhaustive and mutually exclusive. We could
create another set of events by assigning the outcomes 1 and 3 to event Y , and the outcome 5 to event Z . Events X, Y , and Z are
also exhaustive and mutually exclusive.
We have a great deal of latitude in the way we assign the possible outcomes to events. If it suits our purposes, we can create many
different exhaustive and mutually exclusive partitionings of the outcomes of a given distribution. We require that each partitioning
of outcomes into events be exhaustive and mutually exclusive, because we want to apply the laws of probability to events.
This page titled 3.2: Outcomes, Events, and Probability is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
If we know the probabilities of the possible outcomes of a trial, we can calculate the probabilities for combinations of outcomes.
These calculations are based on two rules, which we call the laws of probability. If we partition the outcomes into exhaustive and
mutually exclusive events, the laws of probability also apply to events. Since, as we define them, “events” is a more general term
than “outcomes,” we call them the law of the probability of alternative events and the law of the probability of compound events.
These laws are valid so long as three conditions are satisfied. We have already discussed the first two of these conditions, which are
that the outcomes possible in any individual trial must be exhaustive and mutually exclusive. The third condition is that, if we make
more than one trial, the outcomes must be independent; that is, the outcome of one trial must not be influenced by the outcomes of
the others.
We can view the laws of probability as rules for inferring information about combinations of events. The law of the probability of
alternative events applies to events that belong to the same distribution. The law of the probability of compound events applies to
events that can come from one or more distributions. An important special case occurs when the compound events are N
successive samplings of a given distribution that we identify as the parent distribution. If the random variable is a number, and we
average the numbers that we obtain from N successive samplings of the parent distribution, these “averages-of-N ” themselves
constitute a distribution. If we know certain properties of the parent distribution, we can calculate corresponding properties of the
“distribution of averages-of-N values obtained by sampling the parent distribution.” These calculations are specified by the central
limit theorem, which we discuss in Section 3.11.
In general, when we combine events from two distributions, we can view the result as an event that belongs to a third distribution.
At first encounter, the idea of combining events and distributions may seem esoteric. A few examples serve to show that what we
have in mind is very simple.
Since an event is a set of outcomes, an event occurs whenever any of the outcomes in the set occurs. Partitioning the outcomes of
tossing a die into “even outcomes” and “odd outcomes” illustrates this idea. The event “even outcome” occurs whenever the
outcome of a trial is 2, 4, or 6. The probability of an event can be calculated from the probabilities of the underlying outcomes. We
call the rule for this calculation the law of the probabilities of alternative events. (We create the opportunity for confusion here
because we are illustrating the idea of alternative events by using an example in which we call the alternatives “alternative
outcomes” rather than “alternative events.” We need to remember that “event” is a more general term than “outcome.” One possible
partitioning is that which assigns every outcome to its own event.) We discuss the probabilities of alternative events further below.
To illustrate the idea of compound events, let us consider a first distribution that comprises “tossing a coin” and a second
distribution that comprises “drawing a card from a poker deck.” The first distribution has two possible outcomes; the second
distribution has 52 possible outcomes. If we combine these distributions, we create a third distribution that comprises “tossing a
coin and drawing a card from a poker deck.” The third distribution has 104 possible outcomes. If we know the probabilities of the
outcomes of the first distribution and the probabilities of the outcomes of the second distribution, and these probabilities are
independent of one another, we can calculate the probability of any outcome that belongs to the third distribution. We call the rule
for this calculation the law of the probability of compound events. We discuss it further below.
A similar situation occurs when we consider the outcomes of tossing two coins. We assume that we can tell the two coins apart.
Call them coin 1 and coin 2. We designate heads and tails for coins 1 and 2 as H , T , H , and T , respectively. There are four
1 1 2 2
possible outcomes in the distribution we call “tossing two coins:” H H , H T , T H , and T T . (If we could not tell the coins
1 2 1 2 1 2 1 2
apart, H T would be the same thing as T H ; there would be only three possible outcomes.) We can view the distribution
1 2 1 2
“tossing two coins” as being a combination of the two distributions that we can call “tossing coin 1” and “tossing coin 2 .” We can
also view the distribution “tossing two coins” as a combination of two distributions that we call “tossing a coin a first time” and
“tossing a coin a second time.” We view the distribution “tossing two coins” as being equivalent to the distribution “tossing one
coin twice.” This is an example of repeated trials, which is a frequently encountered type of distribution. In general, we call such a
distribution a “distribution of events from a trial repeated N times,” and we view this distribution as being completely equivalent
to N simultaneous trials of the same kind. Chapter 19 considers the distribution of outcomes when a trial is repeated many times.
Understanding the properties of such distributions is the single most essential element in understanding the theory of statistical
thermodynamics. The central limit theorem relates properties of the repeated-trials distribution to properties of the parent
distribution.
The Probability of Alternative Events
If we know the probability of each of two mutually exclusive events that belong to an exhaustive set, the probability that one or the
other of them will occur in a single trial is equal to the sum of the individual probabilities. Let us call the independent events A and
B, and represent their probabilities as P (A) and P (B), respectively. The probability that one of these events occurs is the same
thing as the probability that either A occurs or B occurs. We can represent this probability as P (A or B) . The probability of this
combination of events is the sum: P (A) + P (B) . That is,
P (A or B) = P (A) + P (B)
Above we define Y as the event that a single toss of a die comes up either 1 or 3. Because each of these outcomes is one of six,
mutually-exclusive, equally-likely outcomes, the probability of either of them is 1/6: P (tossing a 1) = P (tossing a 3) = 1/6.
From the law of the probability of alternative events, we have
P (event Y ) = (tossing a 1 or tossing a 3)
= P (tossing a 1) orP (tossing a 3)
= 1/6 + 1/6
= 2/6
We define X as the event that a single toss of a die comes up even. From the law of the probability of alternative events, we have
P (event X) = P (tossing 2 or 4 or 6)
= P (tossing a 2) + P (tossing a 4) + P (tossing a 6)
= 3/6
We define Z as the event that a single toss comes up 5.
P (event Z) = P (tossing a 5) = 1/6
If there are ω independent events (denoted E , E , … , E , … , E ), the law of the probability of alternative events asserts that the
1 2 i ω
probability that one of these events will occur in a single trial is

P (E1 or E2 or … Ei … or Eω ) = P (E1 ) + P (E2 ) + ⋯ + P (Ei ) + ⋯ + P (Eω )
= ∑ P (Ei )
i=1
If these ω independent events encompass all of the possible outcomes, the sum of their individual probabilities must be unity.
Figure 1. A simple case that illustrates the laws of probability.
The Probability of Compound Events

Let us now suppose that we make two trials in circumstances where event A is possible in the first trial and event B is possible in
the second trial. We represent the probabilities of these events by P (A) and P (B) and stipulate that they are independent of one
another; that is, the probability that B occurs in the second trial is independent of the outcome of the first trial. Then, the
probability that A occurs in the first trial and B occurs in the second trial, P (A and B) , is equal to the product of the individual
probabilities.
P (A and B) = P (A) × P (B)
To illustrate this using outcomes from die-tossing, let us suppose that event A is tossing a 1 and event B is tossing a 3. Then,
P (A) = 1/6 and P (B) = 1/6 . The probability of tossing a 1 in a first trial and tossing a 3 in a second trial is then
P (tossing a 1 first and tossing a 3 second) = P (tossing a 1) × P (tossing a 3)
= 1/6 × 1/6
= 1/36
If we want the probability of getting one 1 and one 3 in two tosses, we must add to this the probability of tossing a 3 first and a 1
second.
If there are ω independent events (denoted E , E , … , E , … , E ), the law of the probability of compound events asserts that the
1 2 i ω
probability that E will occur in a first trial, and E will occur in a second trial, etc., is
1 2
P (E1 and E2 and … Ei … and Eω ) = P (E1 ) × P (E2 ) × ⋯ × P (Ei ) × ⋯ × P (Eω )
= ∏ P (Ei )
i=1
This page titled 3.3: Some Important Properties of Events is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
The laws of probability apply to events that are independent. If the result of one trial depends on the result of another trial, we may
still be able to use the laws of probability. However, to do so, we must know the nature of the interdependence.
If the activity associated with event C precedes the activity associated with event D, the probability of D may depend on whether C
occurs. Suppose that the first activity is tossing a coin and that the second activity is drawing a card from a deck; however, the deck
we use depends on whether the coin comes up heads or tails. If the coin is heads, we draw a card from an ordinary deck; if the coin
is tails, we draw a coin from a deck with the face cards removed. Now we ask about the probability of drawing an ace. If the coin is
heads, the probability of drawing an ace is 4/52 = 1/13. If the coin is tails, the probability of drawing an ace is 4/40 = 1/10. The
combination coin is heads and card is ace has probability: (1/2) (1/13) = 1/26. The combination coin is tails and card is ace has
probability (1/2) (1/10) = 1/20. In this
case, the probability of drawing an ace depends on the modification we make to the deck based on the outcome of the coin toss.
Applying the laws of probability is straightforward. An example that illustrates the application of these laws in a transparent way is
provided by villages First, Second, Third, and Fourth, which are separated by rivers. (See Figure 1.) Bridges 1, 2, and 3 span the
river between First and Second. Bridges a and b span the river between Second and Third. Bridges A , B , C , and D span the river
between Third and Fourth. A traveler from First to Fourth who is free to take any route he pleases has a choice from among
3 × 2 × 4 = 24 possible combinations. Let us consider the probabilities associated with various events:
There are 24 possible routes. If a traveler chooses his route at random, the probability that he will take any particular route is
1/24. This illustrates our assumption that each event in a set of N exhaustive and mutually exclusive events occurs with
probability 1/N .
If he chooses a route at random, the probability that he goes from First to Second by either bridge 1 or bridge 2 is
P (1) + P (2) = 1/3 + 1/3 = 2/3 . This illustrates the calculation of the probability of alternative events.
The probability of the particular route 2 → a → C is P (2) × P (a) × P (C ) = (1/3) (1/2) (1/4) = 1/24 , and we calculate
the same probability for any other route from First to Fourth. This illustrates the calculation of the probability of a compound
event.
If he crosses bridge 1, the probability that his route will be 2 → a → C is zero, of course. The probability of an event that has
already occurred is 1, and the probability of any alternative is zero. If he crosses bridge 1, P (1) = 1 , and P (2) = P (3) = 0 .
Given that a traveler has used bridge 1, the probability of the route 1 → a → C becomes the probability of path a → C , which
is P (a) × P (C ) = (1/2) (1/4) = 1/8 . Since P (1) = 1 , the probability of the compound event 1 → a → C is the
probability of the compound event a → C .
The outcomes of rolling dice, rolling provide more illustrations. If we roll two dice, we can classify the possible outcomes
according to the sums of the outcomes for the individual dice. There are thirty-six possible outcomes. They are displayed in Table
1.
Table 1: Outcomes from tossing two dice
Outcome for first die
1 2 3 4 5 6
1 2 3 4 5 6 7
2 3 4 5 6 7 8
Outcome for
3 4 5 6 7 8 9
second die
4 5 6 7 8 9 10
5 6 7 8 9 10 11
6 7 8 9 10 11 12
Let us consider the probabilities associated with various dice-throwing events:

The probability of any given outcome, say the first die shows 2 and the second die shows 3, is 1/36.
Since the probability that the first die shows 3 while the second die shows 2 is also 1/36, the probability that one die shows 2
and the other shows 3 is
P (3) × P (2) + P (2) × P (3) = (1/36) + (1/36) = 1/18.
Four different outcomes correspond to the event that the score is 5. Therefore, the probability of rolling 5 is
P (1) × P (4) + P (2) × P (3) + P (3) × P (2) + P (4) × P (1) = 1/9
The probability of rolling a score of three or less is the probability of rolling 2, plus the probability of rolling 3 which is
(1/36) + (2/36) = 3/36 = 1/12
Suppose we roll the dice one at a time and that the first die shows 2. The probability of rolling 7 when the second die is thrown
is now 1/6, because only rolling a 5 can make the score 7, and there is a probability of 1/6 that a 5 will come up when the
second die is thrown.
Suppose the first die is red and the second die is green. The probability that the red die comes up 2 and the green die comes up
3 is (1/6) (1/6) = 1/36.
Above we looked at the number of outcomes associated with a score of 3 to find that the probability of this event is 1/18. We can
use another argument to get this result. The probability that two dice roll a score of three is equal to the probability that the first die
shows 1 or 2 times the probability that the second die shows whatever score is necessary to make the total equal to three. This is:
P (f irst die shows 1 or 2) × (1/6) = [(1/6) + (1/6)] × 1/6
= 2/36
= 1/18
Application of the laws of probability is frequently made easier by recognizing a simple restatement of the requirement that events
be mutually exclusive. In a given trial, either an event occurs or it does not. Let the probability that an event A occurs be P (A). Let
the probability that event A does not occur be P (∼ A) . Since in any given trial, the outcome must belong either to event A or to
event ∼ A , we have
P (A) + P (∼ A) = 1
For example, if the probability of success in a single trial is 2/3 , the probability of failure is . If we consider the outcomes of
1/3
two successive trials, we can group them into four events.

Event SS: First trial is a success; second trial is a success.
Event SF: First trial is a success; second trial is a failure.
Event FS: First trial is a failure; second trial is a success.
Event FF: First trial is a failure; second trial is a failure.
Using the laws of probability, we have
1 = P (Event SS) + P (Event SF ) + P (Event F S) + P (Event F F )
= P1 (S) × P2 (S) + P1 (S) × P2 (F ) + P1 (F ) × P2 (S) + P1 (F ) × P2 (F )
where P 1 (X) and P

2 (X) are the probability of event X in the first and second trials, respectively.
This situation can be mapped onto a simple diagram. We represent the possible outcomes of the first trial by line segments on one
side of a unit square P (S) + P (F ) = 1 . We represent the outcomes of the second trial by line segments along an adjoining side
1 1
of the unit square. The four possible events are now represented by the areas of four mutually exclusive and exhaustive portions of
the unit square as shown in Figure 2.
Figure 2. Success and failure in successive trials.
This page titled 3.4: Applying the Laws of Probability is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
Since a discrete distribution is completely specified by the probabilities of each of its events, we can represent it by a bar graph.
The probability of each event is represented by the height of one bar. We can generalize this graphical representation to represent
continuous distributions. To see what we have in mind, let us consider a particular example.
Let us suppose that we have a radar gun and that we decide to interest ourselves in the typical speeds of cars on a highway just
outside of town. As we think about this project, we recognize that speeds might vary with the time of day and the day of the week.
Random variations in many other factors might also be important; these include weather conditions and accidents in the vicinity. To
eliminate as many atypical factors as possible, we might decide that typical speeds are those of cars going north between 1:00 pm
and 4:00 pm on weekdays when the road surface is dry and there are no disabled vehicles in view. If we have a lot of time and the
road is busy, we could collect a lot of data. Let us suppose that we record the speeds of 10, 000 cars. Each datum would be the
speed of a car on the road at a time when the selected conditions are satisfied.
To use this data, we want to summarize it in a form that is easy to visualize. One way to do this is to aggregate the data to give the
number of cars in each 20 mph range; the results might look something like the data in Table 2. Figure 3 is a five-channel bar graph
that displays the number of cars in each 20 mph range. A great deal of information is lost in the aggregating process. In particular,
nothing on the graph represents the number of automobiles in narrower speed intervals.
Table 2. Vehicle speed data.
Height for bar area to equal
Speed(mph) Number of cars Fraction of cars
fraction
–10
200 0.020 0.20/20 = 0.0010
10
800 0.08 0.08/20 = 0.0040
30
2500 0.25 0.25/20 = 0.0125
50
5500 0.55 0.55/20 = 0.0275
70
1000 0.10 0.10/20 = 0.0050
90
Now, suppose that we repeat this task, but that we do not have enough time to collect data on as many as 10, 000 more cars. We
will be curious about the extent to which our two samples agree with one another. Since the total number of vehicles will be
different, the appropriate way to go about this is obviously to compare the fraction of cars in each speed range. In fact, using
fractions enables us to compare any number of such studies. To the extent that these studies measure the same thing—typical
speeds under the specified conditions—the fraction of automobiles in any particular speed interval should be approximately
constant. Dividing the number of automobiles in each speed interval by the total number of automobiles gives a representation that
focuses attention on the proportion of automobiles with various speeds. The shape of the bar graph remains the same; all that
changes is the scale we use to label the ordinate. (See Figure 4.)
Figure 3. Number of cars versus speed.
Insofar as any repetition of this experiment gives nearly the same results, this is a useful change. However, the fundamental
limitations of the graph remain. For example, if we want to use the graph to estimate how speeds are distributed in any other set of
intervals, we have to read values off the ordinate and manipulate them in ways that may not be very satisfactory. To estimate the
fraction with speeds between 20 mph and 40 mph, we might assign half of the automobiles in the 10 − 30 mph interval and half of
those in the 30 − 50 mph interval to the new interval. This enables us to estimate that the fraction in the 20 − 40 mph interval is
0.165. This estimate is much less reliable than one that could be made by going back to the raw data for all 10, 000 automobiles.
Figure 4. Fraction of cars versus speed.

The data can also be represented as a histogram. In a histogram, the information is represented by the area rather than the height of
the bar. In the present case, the only visible change to the graph is another change in the numerical values on the ordinate. In Figure
5, the area of a bar represents the fraction of automobiles with speeds in the given interval. As the speed interval is made smaller,
any of these bar graphs looks increasingly like a continuous curve. (See Figure 6.) The histogram has the advantage that, as the
curve becomes continuous, the interpretation remains constant: the area under the curve between any two speeds always represents
the fraction of automobiles with speeds in this interval. It turns out that we are adept at visually estimating the relative areas of
different parts of a histogram. That is, from a quick glance at a histogram, we are able to obtain a good semi-quantitative
appreciation of the significance of the underlying data.
Figure 5. Speed data presented as a histogram.
If the histogram captures our experience, and we expect future events to have the same characteristics, the histogram becomes an
expression of probability. All that is necessary is that we construct the histogram so that the total area under the graph is unity. If
we let f (u) be the area under the graph from u = −∞ to u = u , then f (u) represents the probability that the speed of a randomly
selected automobile will lie between −∞ and u. For any a and b, the probability that u lies in the interval a < b > is f (b) − f (a) .
The function f (u) is called the cumulative probability distribution function, because its value for any u is the fraction of
automobiles that have a speed less than u. f (a) is the frequency with which we observe values of the random variable, u, that are
less than a . Equivalently, we can say that f (u) is the probability that any randomly selected automobile will have a speed less than
u. If we let the width of every interval go to zero, the bar graph representation of the histogram becomes a curve, and the histogram
becomes a continuous function of the random variable, u. (See Figure 7.) Note that the curve—the enclosing envelope—is not
f (u). f (u) is the area under the enclosing envelope curve.
Figure 6. Histogram with narrower speed intervals.
Figure 7. The histogram can be a continuous function.
This page titled 3.5: Bar Graphs and Histograms is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
3.6: Continuous Distribution Functions - the Envelope Function is the Derivative of
the Area
When we can represent the envelope curve as a continuous function, the envelope curve is the derivative of the cumulative
probability distribution function: The cumulative distribution function is f (u); the envelope function is df (u)/du. The envelope
function is a probability density, and we will refer to the envelope function, df (u)/du, as the probability density function. The
probability density function is the derivative, with respect to the random variable, of the cumulative distribution function. This is an
immediate consequence of the fundamental theorem of calculus.
If H (u) is the anti-derivative of a function h (u), we have dH (u)/du = h (u) , and the fundamental theorem of calculus asserts
that the area under h (u), from u = a to u = b is
b b
dH (u)
∫ h(u)du = ∫ ( ) du
a a
du
= H (b) − H (a)
In the present instance, H (u) = f (u), so that

b
df (u)
∫ ( ) du = f (b) − f (a)
a
du
and
df (u)
h(u) =
du
The envelope function, h(u), and df (u) are the same function.
This point is also apparent if we consider the incremental change in the area, dA , under a histogram as the variable increases from
u to u + du . If we let the envelope function be h(u), we have
dA = h(u)du
or
dA
h(u) =
du
That is, the envelope function is the derivative of the area with respect to the random variable, u. The area is f (u), so the envelope
function is h(u) = df (u)/du.
Calling the envelope curve the probability density function emphasizes that it is analogous to a function that expresses the density
of matter. That is, for an incremental change in u, the incremental change in probability is
df
Δ(probability) = Δu
du
analogous to the incremental change in mass accompanying an incremental change in volume

Δ(mass) = density × Δ(volume)
where
d (mass)
density = .
d (volume)
In this analogy, we suppose that mass is distributed in space with a density that varies from point to point in the space. The mass
enclosed in any particular volume is given by the integral of the density function over the volume enclosed; that is,
mass = ∫ (density) dV .
V
Conversely, the density at any given point is the limit, as the enclosing volume shrinks to zero, of the enclosed mass divided by the
magnitude of the enclosing volume.
Similarly, for any value of the random variable, the probability density is the limit, as an interval spanning the value of the
random variable shrinks to zero, of the probability that the random variable is in the interval, divided by the magnitude of
the interval.
This page titled 3.6: Continuous Distribution Functions - the Envelope Function is the Derivative of the Area is shared under a CC BY-SA 4.0
license and was authored, remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts
platform; a detailed edit history is available upon request.
Suppose that we have a probability density function like that sketched in Figure 8 and that the area under the curve in the interval
a < u < b is 0.25. If we draw a large number of samples from the distribution, our definitions of probability and the probability
density function mean that about 25% of the values we draw will lie in the interval a < u < b . We expect the percentage to become
closer and closer to 25% as the total number of samples drawn becomes very large. The same would be true of any other interval,
c < u < d , where the area under the curve in the interval c < u < d is 0.25.
If we draw exactly four samples from this distribution, the values can be anywhere in the domain of u. However, if we ask what
arrangement of four values best approximates the result of drawing a large number of samples, it is clear that this arrangement must
have a value in each of the four, mutually-exclusive, 25% probability zones. We can extend this conclusion to any number of
representative points. If we ask what arrangement of N points would best represent the arrangement of a large number of points
drawn from the distribution, the answer is clearly that one of the N representative points should lie within each of N , mutually-
exclusive, equal-area segments that span the domain of u.)
Figure 8. A sample of four that approximates its distribution.

We can turn this idea around. In the absence of information to the contrary, the best assumption we can make about a set of N
values of a random variable is that each represents an equally probable outcome. If our entire store of information about a
distribution consists of four data points drawn from the distribution, the best description that we can give of the probability density
function is that one-fourth of the area under the curve lies above a segment of the domain that is associated with each point. If we
th
have N points, the best estimate we can make of the distribution from which the N points are drawn is that (1/N ) of the area
lies above each of them.
This view tells us to associate a probability of 1/N with an interval around each data point, but it does not tell us where to begin or
end the interval. If we could decide where the interval about each data point began and ended, we could estimate the shape of the
probability density function. For a small number of points, we could not expect this estimate to be very accurate, but it would be
the best possible estimate based on the given data.
Figure 9. The sample of four that best approximates its distribution.
Now, instead of trying to find the best interval to associate with each data point, let us think about the intervals into which the data
points divide the domain. This small change of perspective leads us to a logical way to divide the domain of u into specific
intervals of equal probability. If we put N points on any line, these points divide the line into N + 1 segments. There is a segment
to the left of every point; there are N such segments. There is one final segment to the right of the right-most point, and so there
are N + 1 segments in all.
In the absence of information to the contrary, the best assumption we can make is that N data points divide their domain into
N + 1 segments, each of which is associated with equal probability. The fraction of the area above each of these segments is
1/(N + 1) ; also, the probability associated with each segment is 1/(N + 1) . If, as in the example above, there are four data
points, the best assumption we can make about the probability density function is that 20% of its area lies between the left
boundary and the left-most data point, and 20% lies between the right-most data point and the right boundary. The three intervals
between the four data points each represent an additional 20% of the area. Figure 9 indicates the N data points that best
approximate the distribution sketched in Figure 8.
The sketches in Figure 10 describe the probability density functions implied by the indicated sets of data points.
Figure 10. Approximate probability density functions.
This page titled 3.7: A Heuristic View of the Probability Density Function is shared under a CC BY-SA 4.0 license and was authored, remixed,
We can use these ideas to create a plot that approximates the cumulative probability distribution function given any set of N
measurements of a random variable u. To do so, we put the u values found in our N measurements in order from smallest to
i
largest. We label the ordered values u , u ,…,u ,…,u , where u is the smallest. By the argument that we develop in the previous
1 2 i N 1
section, the probability of observing a value less than u is about 1/(N + 1) . If we were to make a large number of additional
1
measurements, a fraction of about 1/(N + 1) of this large number of additional measurements would be less than u . This fraction 1
is just f (u ), so we reason that f (u ) ≈ 1/(N + 1) . The probability of observing a value between u and u is also about
1 1 1 2
1/(N + 1) ; so the probability of observing a value less than u is about 2/(N + 1) , and we expect f (u ) ≈ 2/(N + 1) . In
2 2
general, the probability of observing a value between u and u is also about 1/(N + 1) , and the probability of observing a
i−1 i
value less than u is about i/(N + 1) . In other words, we expect the cumulative probability distribution function for u to be such
i i
that the i smallest observation corresponds to f (u ) ≈ i/(N + 1) . The quantity i/(N + 1) is often called the rank probability
th
i
of the i data point.

th
Figure 11. An approximate cumulative probability distribution function.

Figure 11 is a sketch of the sigmoid shape that we usually expect to find when we plot i/(N + 1) versus the i value of u. This th
plot approximates the cumulative probability distribution function, f (u). We expect the sigmoid shape because we expect the
observed values of u to bunch up around their average value. (If, within some domain of u values, all possible values of u were
equally likely, we would expect the difference between successive observed values of u to be roughly constant, which would make
the plot look approximately linear.) At any value of u, the slope of the curve is just the probability-density function, df (u)/du.
These ideas mean that we can test whether the experimental data are described by any particular mathematical model, say F (u).
To do so, we use the mathematical model to predict each of the N rank probability values: 1/(N + 1) , 2/(N + 1) ,…,
i/(N + 1) ,…, N /(N + 1) . That is to say, we calculate F (u ), F (u ),…, F (u ), …, F (u ); if F (u) describes the data well,
1 2 i N
we will find, for all i, F (u ) ≈ i/(N + 1) . Graphically, we can test the validity of the relationship by plotting i/(N + 1) versus
i
F (u ). If F (u) describes the data well, this plot will be approximately linear, with a slope of one.
i
In Section 3.12, we introduce the normal distribution, which is a mathematical model that describes a great many sources of
experimental observations. The normal distribution is a distribution function that involves two parameters, the mean, μ , and the
standard deviation, σ. The ideas we have discussed can be used to develop a particular graph paper—usually called normal
probability paper. If the data are normally distributed, plotting them on this paper produces an approximately straight line.
We can do essentially the same test without benefit of special graph paper, by calculating the average, ū ≈ μ , and the estimated
¯
¯
standard deviation, s ≈ σ , from the experimental data. (Calculating ū and s is discussed below.) Using ū and s as estimates of μ
¯
¯ ¯
¯
and σ, we can find the model-predicted probability of observing a value of the random variable that is less than u . This value is i
f (u ) for a normal distribution whose mean is ū and whose standard deviation is s . We can find f (u ) by using standard tables
i
¯
¯
i
(usually called the normal curve of error in mathematical compilations), by numerically integrating the normal distribution’s
probability density function, or by using a function embedded in a spreadsheet program, like Excel . If the data are described by
®
the normal distribution function, this value must be approximately equal to the rank probability; that is, we expect
f (u ) ≈ i/(N + 1) . A plot of i/(N + 1) versus f (u ) will be approximately linear with a slope of about one.
i i
This page titled 3.8: A Heuristic View of the Cumulative Distribution Function is shared under a CC BY-SA 4.0 license and was authored,
When we sample a particular distribution, the value that we obtain depends on chance and on the nature of the distribution
described by the function f (u). The probability that any given trial will produce u in the interval a < u < b is equal to
f (b) − f (a) . We often find situations in which a second function of u, call it g (u) , is also of interest. If we sample the distribution
and obtain a value of the random variable, u , then the value of g associated with that trial is g (u ) . The question arises: Given
k k
g(u) and the distribution function f (u), what should we expect the value of g (u ) to be? That is, if we get a value of u from the
k
distribution and then find g (u) , what value should we expect to find for g (u) ? While this seems like a reasonable question, it is
obvious that we can give a meaningful answer only when we can define more precisely just what we mean by “expect.”
To understand our definition of the expected value (sometimes called the expectation value) of g (u) , let us consider a game of
chance. Suppose that we have a needle that rotates freely on a central axis. When spun, the needle describes a circular path, and its
point eventually comes to rest at some point on this path. The location at which the needle stops is completely random. Imagine
that we divide the circular path into six equal segments, which we number from one to six. When we spin the needle, it is equally
likely to stop over any of these segments. Now, let us suppose that we conduct a lottery by selling six tickets, also numbered from
one to six. We decide the winner of the lottery by spinning the needle. The holder of the ticket whose number matches the number
on which the needle stops receives a payoff of $6000. After the spin, one ticket is worth $6000, and the other five are valueless. We
ask: Before the spin, what is any one of the lottery tickets worth?
In this context, it is reasonable to define the expected value of a ticket as the amount that we should be willing to pay to buy a
ticket. If we buy them all, we receive $6000 when the winning ticket is selected. If we pay $1000 per ticket to buy them all, we get
our money back. If we buy all the tickets, the expected value of each ticket is $1000. What if we buy only one ticket? Is it
reasonable to continue to say that its expected value is $1000? We argue that it is. One argument is that the expected value of a
ticket should not depend on who owns the ticket; so, it should not depend on whether we buy one, two, or all of them. A more
general argument supposes that repeated lotteries are held under the same rules. If we spend $1000 to buy one ticket in each of a
very large number of such lotteries, we expect that we will eventually “break even.” Since the needle comes to rest at each number
with equal probability, we reason that
Expected value of a ticket
= $6000 (f raction of times our ticket would be selected)
= $6000 (1/6)
= $1000
Since we assume that the fraction of times our ticket would be selected in a long series of identical lotteries is the same thing as the
probability that our ticket will be selected in any given drawing, we can also express the expected value as
Expected value of a ticket
= $6000 (probability that our ticket will be be selected)
= $6000 (1/6)
= $1000
Clearly, the ticket is superfluous. The game depends on obtaining a value of a random variable from a distribution. The distribution
is a spin of the needle. The random variable is the location at which the needle comes to rest. We can conduct essentially the same
game by allowing any number of participants to bet that the needle will come to rest on any of the six equally probable segments of
the circle. If an individual repeatedly bets on the same segment in many repetitions of this game, the total of his winnings
eventually matches the total amount that he has wagered. (More precisely, the total of his winnings divided by the total amount he
has wagered becomes arbitrarily close to one.)
Suppose now that we change the rules. Under the new rules, we designate segment 1 of the circle as the payoff segment.
Participants pay a fixed sum to be eligible for the payoff for a particular game. Each game is decided by a spin of the needle. If the
needle lands in segment 1, everyone who paid to participate in that game receives $6000. Evidently, the new rules have no effect on
the value of participation. Over the long haul, a participant in a large number of games wins $6000 in one-sixth of these games. We
take this to be equivalent to saying that he has a probability of one-sixth of winning $6000 in a given game in which he
participates. His expected payoff is
Expected value of game
= $6000 (probability of winning $6000)
= $6000 (1/6)
= $1000
Let us change the game again. We sub-divide segment 2 into equal-size segments 2A and 2B. The probability that the needle lands
in 2A or 2B is 1/12. In this new game, the payoff is $6000 when the needle lands in either segment 1 or segment 2A. We can use
any of the arguments that we have made previously to see that the expected payoff game is now $6000 (1/4) = $1500. However,
the analysis that is most readily generalized recognizes that the payoff from this game is just the sum of the payout from the
previous game plus the payout from a game in which the sole payout is $6000 whenever the needle lands in segment 2A. For the
new game, we have
Expected value of a game
= $6000 × P (segment 1) + $6000 × P (segment 2A)
= $6000 (1/6) + $6000 (1/12)
= $1500
We can devise any number of new games by dividing the needle’s circular path into Ω non-overlapping segments. Each segment is
a possible outcome. We number the possible outcomes 1, 2, …, i, …, Ω, label these outcomes u , u ,…, u ,…, u , and denote 1 2 i Ω
their probabilities as P (u ), P (u ),...,P (u ),…, P (u ). We say that the probability of outcome u , P (u ), is the expected
1 2 i Ω i i
frequency of outcome u . We denote the respective payoffs as g (u ) , g (u ) ,...,g (u ) ,…, g (u ) . Straightforward generalization of
i 1 2 i Ω
our last analysis shows that the expected value for participation in any game of this type is
Ω
∑ g (ui ) × P (ui )
i=1
Moreover, the spinner is representative of any distribution, so it is reasonable to generalize further. We can say that the expected
value of the outcome of a single trial is always the probability-weighted sum, over all possible outcomes, of the value of each
outcome. A common notation uses angular brackets to denote the expected value for a function of the random variable; the
expected value of g (u) is ⟨g (u)⟩. For a discrete distribution with Ω exhaustive mutually-exclusive outcomes u , probabilities i
P (u ), and outcome values (payoffs) g (u ) , we define the expected value expected value of g (u) to be
i i
⟨g (u)⟩ = ∑ g (ui ) × P (ui )
i=1
Now, let us examine the expected value of g (u) from a slightly different perspective. Let the number of times that each of the
various outcomes is observed in a particular sample of N observations be N , N , … , N , … , N . We have 1 2 3 Ω
N = N + N +⋯ +N +⋯ +N
1 2 i . The set {N , N , … , N , … , N } specifies the way that the possible outcomes are
Ω 1 2 i Ω
populated in this particular series of N observations. We call {N , N , … , N , … , N } a population set. If we make a second
1 2 i Ω
series of N observations, we obtain a second population set. We infer that the best forecast we can make for the number of
occurrences of outcome u in any future series of N observations is N × P (u ) . We call N × P (u ) the expected number of
i i i
observations of outcome u in a sample of size N .

i
In a particular series of N trials, the number of occurrences of outcome ui , and hence of g (ui ) , is Ni . For the set of outcomes
{ N1 , N2 , … , N3 , … , NΩ } , the average value of g (u) is
Ω
¯
¯¯¯¯¯¯¯¯
¯ 1
g (u) = ∑ g (ui ) × Ni
N
i=1
¯
¯¯¯¯¯¯¯¯
¯ ¯
¯¯¯¯¯¯¯¯
¯
Collecting a second sample of N observations produces a second estimate of g (u) . If N is small, successive estimates of g (u)
may differ significantly from one another. If we make a series of N observations multiple times, we obtain multiple population
sets. In general, the population set from one series of N observations is different from the population set for a second series of N
observations. If N ≫ Ω , collecting such samples of N a sufficiently large number of times must produce some population sets
more than once, and among those that are observed more than once, one must occur more often than any other. We call it the most
probable population set. Let the elements of the most probable population set be {N , N , … , N , … , N }. We infer that the 1 2 i Ω
most probable population set is the best forecast we can make about the outcomes of any future sample of N from this distribution.
Moreover, we infer that the best estimate we can make of N is that it equals the expected number of observations of outcome u ;
i i
that is,
Ni ≈ N × P (ui )
Now, N and N must be natural numbers, while N × P (u ) need only be real. In particular, we can have 0, but\(N must be 0 or
i i i i
1 (or some higher integer). This is a situation of practical importance, because circumstances may limit the sample size to a
number, N , that is much less than the number of possible outcomes, Ω. (We encounter this situation in our discussion of statistical
thermodynamics in Chapter 21. We find that the number of molecules in a system can be much smaller than the number of
outcomes—observable energy levels—available to any given molecule.)
If many more than N outcomes have about the same probability, repeated collection of samples of N observations can produce a
series of population sets (each population set different from all of the others) in each of which every element is either zero or one.
When this occurs, it may be that no single population set is significantly more probable than any of many others. Nevertheless,
every outcome occurs with a well-defined probability. We infer that the set
{N × P (u ) , N × P (u ) , … , N × P (u ) , … , N × P (u )}
1 2 is always an adequate proxy for calculating the expected value
i Ω
for the most probable population set.

To illustrate this kind of distribution, suppose that there are 3000 possible outcomes, of which the first and last thousand have
probabilities that are so low that they can be taken as zero, while the middle 1000 outcomes have approximately equal
probabilities. Then P (u ) ≈ 0 for $1<$>1000 and 2001$<$>3000, while P (u ) ≈ 10 for 1001 < 2000>. We are illustrating
i i
−3
the situation in which the number of outcomes we can observe, N , is much less than the number of outcomes that have appreciable
probability, which is 1000. So let us take the number of trials to be N = 4 . If the value of g (u) for each of the 1000 middle
outcomes is the same, say g (u ) = 100 for $1001<2000$>, then our calculation of the expected value of g (u) will be
i
3000 2000
1 1 400
⟨g (u)⟩ = ∑ g (ui ) × N × P (ui ) = ∑ 100 × Ni = = 100
4 4 4
i=1 i=1001
regardless of which population set results from the four trials. That is, because all of the populations sets that have a significant
chance to be observed have N = 1 and g (u ) = 100 for exactly four values of i in the range $1001<2011$>, all of the
i i
population sets that have a significant chance to be observed give rise to the same expected value.
¯
¯¯¯¯¯¯¯¯
¯
Let us compute the arithmetic average, g (u) , using the most probable population set for a sample of N trials. In this case, the
number of observations of the outcome u is N = N × P (u ) . i i i
¯
¯¯¯¯¯¯¯¯
¯ 1 Ω 1 Ω Ω
g (u) = ∑ g (ui ) × Ni = ∑ g (ui ) × N × P (ui ) = ∑ g (ui ) × P (ui ) = ⟨g (u)⟩
N i=1 N i=1 i=1
¯
¯¯¯¯¯¯¯¯
¯
For a discrete distribution, ⟨g (u)⟩ is the value of g (u) that we calculate from the most probable population set,
{ N , N , … , N , … , N }, or its proxy {N × P (u ) , N × P (u ) , … , N × P (u ) , … , N × P (u )} .
1 2 i Ω 1 2 i Ω
We can extend the definition of the expected value, ⟨g (u)⟩, to cases in which the cumulative probability distribution function,
f (u), and the outcome-value function, g (u) , are continuous in the domain of the random variable, $u_{min}<u_{max}$>. To do
so, we divide this domain into a finite number, Ω, of intervals, Δu . We let u be the lower limit of u in the interval Δu . Then the
i i i
probability that a given trial yields a value of the random variable in the interval Δu is P (Δu ) = f (u + Δu ) − f (u ) , and i i i i i
we can approximate the expected value of g (u) for the continuous distribution by the finite sum
Ω Ω Ω
f (ui + Δui ) − f (ui )
⟨g (u)⟩ = ∑ g (ui ) × P (Δui ) = ∑ g (ui ) × [f (ui + Δui ) − f (ui )] = ∑ g (ui ) × [ ]
Δui
i=1 i=1 i=1
× Δui
In the limit as Ω becomes arbitrarily large and all of the intervals Δui become arbitrarily small, the expected value of g (u) for a
continuous distribution becomes
∞
df (u)
⟨g (u)⟩ =∫ g (u) [ ] du
−∞ du
This integral is the value of ⟨g (u)⟩, where df (u)/du is the probability density function for the distribution. If c is a constant, we
have
⟨g (cu)⟩ = c ⟨g (u)⟩
If h (u) is a second function of the random variable, we have
⟨g (u) + h (u)⟩ = ⟨g (u)⟩ + ⟨h (u)⟩
This page titled 3.9: Random Variables, Expected Values, and Population Sets is shared under a CC BY-SA 4.0 license and was authored,
There are two important statistics associated with any probability distribution, the mean of a distribution and the variance of a
distribution. The mean is defined as the expected value of the random variable itself. The Greek letter μ is usually used to represent
the mean. If f (u) is the cumulative probability distribution, the mean is the expected value for g (u) = u . From our definition of
expected value, the mean is
∞
df
μ =∫ u( ) du
−∞ du
The variance is defined as the expected value of (u − μ) . The variance measures how dispersed the data are. If the variance is
2
large, the data are—on average—farther from the mean than they are if the variance is small. The standard deviation is the square
root of the variance. The Greek letter σ is usually used to denote the standard deviation. Then, σ denotes the variance, and 2
∞
2
df
2
σ =∫ (u − μ) ( ) du
−∞
du
If we have a small number of points from a distribution, we can estimate μ and σ by approximating these integrals as sums over the
domain of the random variable. To do this, we need to estimate the probability associated with each interval for which we have a
sample point. By the argument we make in Section 3.7, the best estimate of this probability is simply 1/N , where N is the number
of sample points. We have therefore
∞ N
df 1
¯¯
μ =∫ u( ) du ≈ ∑ ui ( ) = ū
−∞ du N
1
That is, the best estimate we can make of the mean from N data points is u, where u is the ordinary arithmetic average. Similarly,
¯¯
¯ ¯¯
¯
the best estimate we can make of the variance is

∞ N
2 2
df 2
1
σ =∫ (u − μ) ( ) du ≈ ∑(ui − μ) ( )
−∞
du N
i=1
Now a complication arises in that we usually do not know the value of μ . The best we can do is to estimate its value as μ ≈ ū . It ¯
¯
turns out that using this approximation in the equation we deduce for the variance gives an estimate of the variance that is too
small. A more detailed argument (see Section 3.14) shows that, if we use ū to approximate the mean, the best estimate of σ ,
¯
¯ 2
usually denoted s , is
2
N
2 2 2 1
¯¯
estimated σ =s = ∑ (ui − ū) ( )
N −1
i=1
Figure 3.10.1 : Variance is analogous to a moment of inertia.

Dividing by N − 1 , rather than N , compensates exactly for the error introduced by using ū rather than μ . ¯
¯
The mean is analogous to a center of mass. The variance is analogous to a moment of inertia. For this reason, the variance is also
called the second moment about the mean. To show these analogies, let us imagine that we draw the probability density function
on a uniformly thick steel plate and then cut along the curve and the u-axis (Figure 3.10.1). Let M be the mass of the cutout piece
of plate; M is the mass below the probability density curve. Let dA and dm be the increments of area and mass in the thin slice of
the cutout that lies above a small increment, du, of u. Let ρ be the density of the plate, expressed as mass per unit area. Since the
plate is uniform, ρ is constant. We have dA = (df /du) du and dm = ρdA so that
df
dm = ρ ( ) du
du
The mean of the distribution corresponds to a vertical line on this cutout at u = μ . If the cutout is supported on a knife-edge along
the line u = μ , gravity induces no torque; the cutout is balanced. Since the torque is zero, we have
M ∞
df
0 =∫ (u − μ) dm = ∫ (u − μ) ρ ( ) du
m=0 −∞
du
Since μ is a constant property of the cut-out, it follows that

∞
df
μ =∫ u( )du
−∞
du
The cutout’s moment of inertia about the line u = μ is

M
2
I =∫ (u − μ) dm
m=0
∞
2 df
=∫ (u − μ) ρ( ) du
−∞
du
2
= ρσ
The moment of inertia about the line u −μ is simply the mass per unit area, ρ, times the variance of the distribution. If we let
ρ = 1 , we have I = σ .
2
We define the mean of f (u) as the expected value of u. It is the value of u we should “expect” to get the next time we sample the
distribution. Alternatively, we can say that the mean is the best prediction we can make about the value of a future sample from the
distribution. If we know μ , the best prediction we can make is u = μ . If we have only the estimated mean, ū, then ū is the
predicted
¯
¯ ¯
¯
best prediction we can make. Choosing u = ū makes the difference, |u − u

predicted
¯
¯
| , as small as possible.
predicted
These ideas relate to another interpretation of the mean. We saw that the variance is the second moment about the mean. The first
moment about the mean is
∞
df
st
1 moment = ∫ (u − μ) ( ) du
−∞ du
∞ ∞
df df
=∫ u( ) du − μ ∫ ( ) du
−∞ du −∞ du
= μ−μ
=0
Since the last two integrals are μ and 1, respectively, the first moment about the mean is zero. We could have defined the mean as
the value, μ , for which the first moment of u about μ is zero.
The first moment about the mean is zero. The second moment about the mean is the variance. We can define third, fourth, and
higher moments about the mean. Some of these higher moments have useful applications.
This page titled 3.10: Statistics - the Mean and the Variance of a Distribution is shared under a CC BY-SA 4.0 license and was authored, remixed,
The central limit theorem establishes very important relationships between the statistics for two distributions that are related in a
particular way. It enables us to understand some important features of physical systems.
The central limit theorem concerns the distribution of averages. If we have some original distribution and sample it three times, we
can calculate the average of these three data points. Call this average A . We could repeat this activity and obtain a second
3,1
average of three values, A . We can do this repeatedly, generating averages A ,…, A . Several things will be true about these
3,2 3,3 3,n
averages:
The set of all of the possible averages-of-three, {A }, is itself a distribution. This averages-of-three distribution is different
3,i
from the original distribution. Each average-of-three is a value of the random variable associated with the averages-of-three
distribution.
Each of the A is an estimate of the mean of the original distribution.
3,i
The distribution of the A will be less spread out than the original distribution.
3,i
There is nothing unique about averaging three values. We could sample the original distribution seven times and compute the
average of these seven values, calling the result A . Repeating, we could generate averages A ,…, A . All of the things we
7,1 7,2 7,m
say about the averages-of-three are also true of these averages-of-seven. However, we can now say something more: The
distribution of the A will be less spread out than the distribution of the A . The corresponding probability density functions are
7,i 3,i
sketched in Figure 13.
Figure 13. The Variance of an Average of N is proportional to 1/N.

The central limit theorem relates the mean and variance of the distribution of averages to the mean and variance of the original
distribution:
If random samples of N values are taken from a distribution, whose mean is µ and whose variance is σ , averages of these N
2
values, A , are approximately normally distributed with a mean of µ and a variance of σ /N . The approximation to the
N
2
normal distribution becomes better as N becomes larger.

It turns out that the number, N , of trials that is needed to get a good estimate of the variance is substantially larger than the number
required to get a good estimate of the mean.
This page titled 3.11: The Variance of the Average- The Central Limit Theorem is shared under a CC BY-SA 4.0 license and was authored,
The normal distribution is very important. The central limit theorem says that if we average enough values from any distribution,
the distribution of the averages we calculate will be the normal distribution. The probability density function for the normal
distribution is
2
df 1 − (u − μ)
= −− exp [ ]
du σ √2π 2σ 2
The integral of the normal distribution from u = −∞ to u = ∞ is unity. However, the definite integral between arbitrary limits
cannot be obtained as an analytical function. This turns out to be true for some other important distributions also; this is one reason
for working with probability density functions rather than the corresponding cumulative probability functions. Of course, the
definite integral can be calculated to any desired accuracy by numerical methods, and readily available tables give values for
definite integrals from u = −∞ to u = u . (We mention normal curve of error tables in Section 3.8, where we introduce a method
for testing whether a given set of data conforms to the normal distribution equation.)
This page titled 3.12: The Normal Distribution is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
3.13: The Expected Value of a Function of Several Variables and the Central Limit
Theorem
We can extend the idea of an expected value to a function of multiple random variables. Let U and V be distributions whose
random variables are u and v , respectively. Let the probability density functions for these distributions be df (u)/du and u
df (v)/dv . In general, these probability density functions are different functions; that is, U and V are different distributions. Let
v
g (u, v) be some function of these random variables. The probability that an observation made on U produces a value of u in the
range u < u < u + du is

∗ ∗
∗
dfu (u )
∗ ∗
P (u <u <u + du) = du
du
and the probability that an observation made on V produces a value of v in the range v ∗
<v<v
∗
+ dv is
∗
dfv (v )
∗ ∗
P (v <v<v + dv) = dv
dv
The probability that making one observation on each of these distributions produces a value of u that lies in the range
u < u < u + du and a value of v that lies in the range v < v < v + dv is
∗ ∗ ∗ ∗
∗ ∗
dfu (u ) dfv (v )
du dv
du dv
In a straightforward generalization, we define the expected value of g (u, v), ⟨g (u, v) ⟩ , as

∞ ∞
dfu (u) dfv (v)
⟨g (u, v) ⟩ = ∫ ∫ g (u, v) dudv
v=−∞ u=−∞
du dv
If g (u, v) is a sum of functions of independent variables, g (u, v) = h (u) + k (v) , we have

∞ ∞ ∞ ∞
dfu (u) dfv (v) dfu (u) dfv (v)
⟨g (u, v)⟩ = ∫ ∫ [h (u) + k (v)] dudv = ∫ h (u) du + ∫ k (v) dv
−∞ −∞ du dv −∞ du −∞ dv
= ⟨h (u) ⟩ + ⟨k (v) ⟩
If g (u, v) is a product of independent functions, g (u, v) = h (u) k (v) , we have

∞ ∞ ∞ ∞
dfu (u) dfv (v) dfu (u) dfv (v)
⟨g (u, v)⟩ = ∫ ∫ h (u) k (v) dudv =∫ h (u) du × ∫ k (v) d
−∞ −∞ du dv −∞ du −∞ dv
= ⟨h (u)⟩ ⟨k (v)⟩
We can extend these conclusions to functions of the random variables of any number of distributions. If u is the random variable i
of distribution U whose probability density function is df (u )/du , the expected value of

i i i i
g (u1 , … , ui , … , uN ) = h1 (u1 ) + ⋯ + hi (ui ) + ⋯ + hN (uN )
becomes
N
⟨g (u1 , … , ui , … , uN )⟩ = ∑ ⟨hi (ui )⟩
i=1
and the expected value of
g (u1 , … , ui , … , uN ) = h1 (u1 ) … hi (ui ) … hN (uN )
becomes
N
⟨g (u1 , … , ui , … , uN )⟩ = ∏ ⟨hi (ui )⟩
i=1
We are particularly interested in expected values for repeated trials made on the same distribution. We consider distributions for
which the outcome of one trial is independent of the outcome of any other trial. The probability density function is the same for
every trial, so we have f (u) = f (u ) = ⋯ = f (u ) = ⋯ = f (u ) . Let the values obtained for the random variable in a
1 1 i i N N
series of trials on the same distribution be u ,…, u ,…, u . For each trial, we have
1 i N
∞
dfi (ui )
⟨hi (ui )⟩ = ∫ hi (ui ) dui
−∞
dui
If we consider the special case of repeated trials in which the functions hi (ui ) are all the same function, so that
h (u) = h (u ) = ⋯ = h (u ) = ⋯ = h
1 1 (u ) , the expected value of
i i N N
g (u1 , … , ui , … , uN )
= h1 (u1 ) + ⋯ + hi (ui ) + ⋯ + hN (uN )
becomes
N
⟨g (u1 , … , ui , … , uN )⟩ = ∑ ⟨hi (ui )⟩ = N ⟨h (u)⟩
i=1
and the expected value of
g (u1 , … , ui , … , uN ) = h1 (u1 ) … hi (ui ) … hN (uN )
becomes
N
N
⟨g (u1 , … , ui , … , uN )⟩ = ∏ ⟨hi (ui )⟩ = ⟨h (u)⟩
i=1
Now let us consider N independent trials on the same distribution and let h i (ui ) = h (ui ) = ui . Then, the expected value of
g (u1 , … , ui , … , uN ) = h1 (u1 ) + ⋯ + hi (ui ) + ⋯ + hN (uN )
becomes
N
⟨u1 + ⋯ + ui + ⋯ + uN ⟩ = ∑ ⟨ui ⟩ = N ⟨u⟩ = N μ
i=1
By definition, the average of N repeated trials is

¯¯
¯
uN = (u1 + ⋯ + ui + ⋯ + uN )/N , so that the expected value of the mean of a distribution of an average-of- N repeated trials
is
⟨u1 + ⋯ + ui + ⋯ + uN ⟩
¯¯
⟨ūN⟩ = =μ
N
This proves one element of the central limit theorem: The mean of a distribution of averages-of- N values of a random variable
drawn from a parent distribution is equal to the mean of the parent distribution.
The variance of these averages-of- N is
2 2 2
N N N
2 1 1 Nμ 1
2 ¯¯
¯
σ = ⟨ (uN − μ) ⟩ = ⟨ [( ∑ ui ) − μ] ⟩ = ⟨ [( ∑ ui ) − ] ⟩ = ⟨ [( ∑ ui ) − N μ] ⟩
N 2
N N N N
i=1 i=1 i=1
2
N N N −1 N
1 1 2
2 1
= ⟨ [( ∑ (ui − μ))] ⟩ = ⟨ ∑ (ui − μ) ⟩+ ⟨ ∑ (ui − μ) ∑ (uj − μ)⟩ =
2 2 2 2
N N N N
i=1 i=1 i=1 j=i+1
N N −1 N
2
2
∑ ⟨ (ui − μ) ⟩+ ⟨ ∑ (ui − μ)⟩ ⟨ ∑ (uj − μ)⟩
2
N
i=1 i=1 j=i+1
Where the last term is zero, because
N −1 N −1
⟨ ∑ (ui − μ) ⟩ = ∑ ⟨ (ui − μ) ⟩
i=1 i=1
and
⟨ (ui − μ) ⟩ =0
By definition, σ 2
= ⟨ (ui − μ)
2
⟩ , so that we have
2 2
Nσ σ
2
σ = =
N 2
N N
This proves a second element of the central limit theorem: The variance of an average of N values of a random variable drawn
from a parent distribution is equal to the variance of the parent distribution divided by N .
This page titled 3.13: The Expected Value of a Function of Several Variables and the Central Limit Theorem is shared under a CC BY-SA 4.0
If we know μ and we have a set of N data points, the best estimate we can make of the variance is
uma x N
2
df 2
1
2
σ =∫ (u − μ) ( ) du ≈ ∑ (ui − μ) ( )
umin
du N
i=1
We have said that if we must use u to approximate the mean, the best estimate of σ , usually denoted s , is
¯¯
¯ 2 2
2 2 2 1
¯¯
¯
estimated σ =s = ∑ (ui − u) ( )
N −1
i=1
The use of N − 1 , rather than N , in the denominator is distinctly non-intuitive; so much so that this equation often causes great
irritation. Let us see how this equation comes about.
Suppose that we have a distribution whose mean is μ and variance is σ . Suppose that we draw N values of the random variable,
2
u, from the distribution. We want to think about the expected value of (u − μ) . Let us write (u − μ) as
2
¯¯
¯ ¯¯
¯
(u − μ) = (u − u) + (u − μ) .
Squaring this gives

2 2 2
¯¯
¯ ¯¯
¯ ¯¯
¯ ¯¯
¯
(u − μ) = (u − u) + (u − μ) + 2 (u − u) (u − μ) .
From our definition of expected value, we can write:

2
Expected value of (u − μ) =
2
¯¯
= expected value of (u − ū)
2
¯¯
+ expected value of (ū − μ)
¯¯ ¯¯
+ expected value of 2 (u − ū) (ū − μ)
From our discussion above, we can recognize each of these expected values:
The expected value of (u − μ) is the variance of the original distribution, which is σ . Since this is a definition, it is exact.
2 2
2
The best possible estimate of the expected value of (u − u) is ¯¯
¯
N
2
1
¯¯
∑ (ui − ū) ( )
N
i=1
2
The expected value of (u − μ) is the expected value of the variance of averages of N random variables drawn from the
¯¯
¯
original distribution. That is, the expected value of (ū − μ) is what we would get if we repeatedly drew N values from the
¯
¯
2
original distribution, computed the average of each set of N values, and then found the variance of this new distribution of
2
average values. By the central limit theorem, this variance is σ /N . Thus, the expected value of (ū − μ) is exactly σ /N .
2 ¯
¯ 2
Since (ū − μ) is constant, the expected value of 2 (u − ū) (ū − μ) is

¯
¯ ¯
¯ ¯
¯
N
1
¯¯ ¯¯
2 (ū − μ) [ ∑ (ui − ū)]
N
i=1
which is equal to zero, because

N N
¯¯ ¯¯
∑ (ui − ū) = ( ∑ ui ) − N ū =0
i=1 i=1
by the definition of u.
¯¯
¯
Substituting, our expression for the expected value of (u − μ) becomes: 2
N 2
2 2 1 σ
¯¯
¯
σ ≈ ∑ (ui − u) ( )+
N N
i=1
so that
N 2
¯¯
1 N −1 (ui − ū)
2 2
σ (1 − ) =σ ( ) ≈∑
N N N
i=1
and
N 2
¯¯
¯
(ui − u)
2
σ ≈∑
N −1
i=1
That is, as originally stated, when we must use ū rather than the true mean, μ , in the sum of squared differences, the best possible
¯
¯
estimate of σ , usually denoted s , is obtained by dividing by N − 1 , rather than by N .

2 2
This page titled 3.14: Where Does the N - 1 Come from? is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
3.15: Problems
Problems
1. At each toss of a die, the die lands with one face on top. This face is distinguished from the other five faces by the number of
dots that appear on it. Tossing a die produces data. What is the distribution? What is the random variable of this distribution? What
outcomes are possible for this distribution? How would we collect a sample of ten values of the random variable of this
distribution?
2. Suppose that we toss a die three times and average the results observed. How would you describe the distribution from which
this average is derived? What is the random variable of this distribution? What outcomes are possible for this distribution? What
would we do to collect a sample of ten values of the random variable of this distribution?
3. Suppose that we toss three dice simultaneously and average the results observed. How would you describe the distribution from
which this average is derived? What is the random variable of this distribution? What outcomes are possible for this distribution?
What would we do to collect a sample of ten values of the random variable of this distribution? Suppose that some third party
collects a set, call it A, of ten values from this distribution and a second set, call it B, of values from the distribution in problem 2.
If we are given the data in each set but are not told which label goes with which set of data, can we analyze the data to determine
which set is A and which is B?
4. The manufacturing process for an electronic component produces 3 bad components in every 1000 components produced. The
bad components appear randomly. What is the probability that
(a) a randomly selected component is bad?
(b) a randomly selected component is good?
(c) 2 bad components are produced in succession?
(d) 100 good components are produced in succession?
5. A product incorporates two of the components in the previous problem. What is the probability that
(a) both components are good?
(b) both components are bad?
(c) one component is good and one component is bad?
(d) at least one component is good?
6. A card is selected at random from a well-shuffled deck. A second card is then selected at random from among the remaining 51
cards. What is the probability that
(a) the first card is a heart?
(b) the second card is a heart?
(c) neither card is a heart?
(d) both cards are hearts?
(e) at least one card is a heart?
7. A graduating class has 70 men and 77 women. How many combinations of homecoming king and queen are possible?
8. After the queen is selected from the graduating class of problem 7, one woman is selected to “first attendant” to the homecoming
queen. Thereafter, another woman is selected to be “second attendant.” After the queen is selected, how many ways can two
attendants be selected?
9. A red die and a green die are rolled. What is the probability that
(a) both come up 3?
(b) both come up the same?
(c) they come up different?
(d) the red die comes up less than the green die?
(e) the red die comes up exactly two less than the green die?
(f) together they show 5?
10. A television game show offers a contestant a new car as the prize for correctly guessing which of three doors the car is behind.
After the contestant selects a door, the game-show host opens an incorrect door. The host then gives the contestant the option of
switching from the door he originally chose to the other door that remains unopened. Should the contestant change his selection?
[Hint: Consider the final set of outcomes to result from a sequence of three choices. First, the game-show producer selects a door
and places the car behind this door. Diagram the possibilities. What is the probability of each? Second, the contestant selects a door.
There are now nine possible outcomes. Diagram them. What is the probability of each? Third, the host opens a door. There are now
twelve possible outcomes. Diagram them. What is the probability of each? Note that these twelve possibilities are not all equally
probable.]
11. For a particular distribution, possible values of the random variable, x, range from zero to one. The probability density function
for this distribution is df /dx = 1 .
(a) Show that the probability of finding x in the range 0 ≤ x ≤ 1 is one.
(b) What is the mean of this distribution?
(c) What is the variance of this distribution? The standard deviation?
(d) A quantity, g , is a function of x: g (x) = x . What is the expected value of g ?
2
12. For a particular distribution, possible values of the random variable, x , range from one to three. The probability density
function for this distribution is df /dx = cx, where c is a constant.
(a) What is the value of the constant, c?
(d) If g (x) = x , what is the expected value of g ?
2
13. For a particular distribution, possible values of the random variable, x, range from two to four. The probability density function
for this distribution is df /dx = cx , where c is a constant.
3
(a) What is the value of the constant, c?

2
14. For a particular distribution, possible values of the random variable, x, range from zero to four. For 0 ≤ x ≤≤ 1 , the
probability density function is df /dx = x/2. For 1, theprobabilitydensityf unctionis\(df /dx = (4 − x)/6.
(a) Show that the area under this probability distribution function is one.
2
15. The following values, xi , of the random variable, x, are drawn from a distribution: , ,
9.63 9.00 11.87 , ,
10.13 10.83, 9.50,
10.40, 9.83, and 10.09.
(a) Arrange these values in increasing order and calculate the “rank probability,” i/(N + 1) , associated with each of the xi
values.
(b) Plot the rank probability (on the ordinate) versus the random-variable value (on the abscissa). Sketch a smooth curve
through the points on this plot.
(c) What function is approximated by the curve sketched in part b?
(d) Plot the data points along a horizontal axis. Then create a bar graph (histogram) by erecting bars of equal area between
each pair of data points.
(e) What function is approximated by the tops of the bars erected in part d?
16. For a particular distribution, possible values of the random variable range from zero to four. The following values of the
random variable are drawn from this distribution: 0.1, 1.0, 1.1, 1.5, 2.1. Sketch an approximate probability density function for this
distribution.
17. The possible values for the random variable of a particular distribution lie in the range 0 ≤ x ≤ 10 . In six trials, the following
values are obtained: 1.0, 1.9, 2.3, 2.7, 3.0, 3.8.
(a) Sketch an approximate probability density function for this distribution.
(b) What is the best estimate we can make of the mean of this distribution?
(c) What is the best estimate we can make of the variance of this distribution?
(d) What is the best estimate we can make of the variance of averages-of-six drawn from this distribution?
(e) What is the best estimate we can make of the variance of averages-of-sixteen drawn from this distribution?
18. A computer program generates numbers from a normal distribution with a mean of zero and a standard deviation of 10. Also,
for any integer N , the program will generate and average N values from this distribution. It will repeat this operation until it has
produced 100 such averages. It will then compute the estimated standard deviation of these 100 average values. The table below
gives various values of N and the estimated standard deviation, s , that was found for 100 averages of that N . Plot these data in a
way that tests the validity of the central limit theorem.
N s
4 5.182
9 2.794
16 2.206
25 2.152
36 1.689
49 1.092
64 1.001
81 1.004
100 1.074
144 0.601
196 0.546
256 0.690
324 0.545
19. If f (u) is the cumulative probability distribution function for a distribution, what is the expected value of f (u) ? What
interpretation can you place on this result?
20. Five replications of a volumetric analysis yield concentration estimates of 0.3000, 0.3008, 0.3012, 0.3014, and 0.3020 mol
L
−1
. Calculate the rank probability of each of these results. Sketch, over the concentration range $0.3000<0.3020$> mol L , an −1
approximation of the cumulative probability distribution function for the distribution that yielded these data.
21. The Louisville Mudhens play on a square baseball field that measures 100 meters on a side. Casey’s hits always fall on the
field. (He never hits a foul ball or hits one out of the park.) The probability density function for the distance that a Casey hit goes
parallel to the first-base line is df (x)/dx = (2 × 10 ) x . (That is, we take the first-base line as our x-axis; the third-base line as
x
−4
our y -axis; and home plate is at the origin. df (x)/dx is independent of the distance that the hit goes parallel to the third-base line,
x
our y -axis.) The probability density function for the distance that a Casey hit goes parallel to the third-base line is
dfy (y)/dy = (3 × 10
−6
)y
2
. (df y (y)/dy is independent of the distance that the hit goes parallel to the first-base line, our x-
axis.)
(a) What is the probability that a Casey hit lands at a point (x, y) such that $x^*<x^*+dx$> and $y^*<y^*+dy$> ?
(b) What is the two-dimensionally probability density function that describes Casey’s hits, expressed in this Cartesian
coordinate system?
(c) Recall that polar coordinates transform to Cartesian coordinates according to x = rcosθ and y = rsinθ . What is the
probability density function for Casey’s hits expressed using polar coordinates?
(d) Recall that the differential element of area in polar coordinates is rdrdθ . Find the probability that a Casey hit lands within
the pie-shaped area bounded by $0<50$> m and 0 < θ < π/4 .
22. In Chapter 2, we derived the Barometric Formula, η (h) = η (0) exp (−mgh/kT ) for molecules of mass m in an isothermal
atmosphere at a height h above the surface of the earth. η (h) is the number of molecules per unit volume at height h ; η (0) is the
number of molecules per unit volume at the earth’s surface, where h = 0 . Consider a vertical cylinder of unit cross-sectional area,
extending from the earth’s surface to an infinite height. Let f (h) be the fraction of the molecules in this cylinder that is at a height
less than h . Prove that the probability density function is df /dh = (mg/kT ) exp (−mgh/kT ) .
23. A particular distribution has six outcomes. These outcomes and their probabilities are a [GrindEQ__0_1_]; b
[GrindEQ__0_2_]; c [GrindEQ__0_3_]; d [GrindEQ__0_2_]; e [GrindEQ__0_1_]; and f [GrindEQ__0_1_].

(a) Partitioning I assigns these outcomes to a set of three events: Event A = a or b or c ; Event B = d ; and Event C = e or
f . What are the probabilities of Events A , B , and C ?
(b) Partitioning II assigns the outcomes to two events: Event D = a or b or c ; and Event E = d or e or f . What are the
probabilities of Events D and E ? Express the probabilities of Events D and E in terms of the probabilities of Events A , B ,
and C .
(c) Partitioning III assigns the outcomes to three events: Event F = a or b ; Event G = c or d ; and Event H = e or f .
What are the probabilities of Events F , G, and H ? Can the probabilities of Events F , G, and H be expressed in terms of the
probabilities of Events A , B , and C ?
24. Consider a partitioning of outcomes into events that is not exhaustive; that is, not every outcome is assigned to an event. What
problem arises when we want to describe the probabilities of these events?
25. Consider a partitioning of outcomes into events that is not mutually exclusive; that is, one (or more) outcome is assigned to two
(or more) events. What problem arises when we want to describe the probabilities of these events?
26. For integer values of p (p ≠ 1) , we find
p+1 p+1
x x
p
∫ x ln(x) dx = ( ) ln(x) −
2
p +1 (p + 1)
(a) Sketch the function, h (x) = df (x)/dx = −4xln(x) , over the interval 0 ≤ x ≤ 1 .
(b) Show that we can consider h (x) = df (x)/dx = −4xln(x) to be a probability density function over this interval; that
is, show f (1) − f (0) = 1 . Let us name the corresponding distribution “Sam.”
(c) What is the mean, μ , of Sam?
(d) What is the variance, σ , of Sam?
2
(e) What is the standard deviation, σ, of Sam?

(f) What is the variance of averages-of-four samples taken from Sam?
(g) The following four values are obtained in random sampling of an unknown distribution: 0.050; 0.010; 0.020; and 0.040.
Estimate the mean, μ , variance (σ or s ), and the standard deviation (σ or s) for this unknown distribution.
2 2
(h) What is the probability that a single sample drawn from Sam will lie in the interval 0 ≤ x ≤ 0.10 ? Note: The upper limit
of this interval is 0.10, not 1.0 as in part (a).
(i) Is it likely that the unknown distribution sampled in part g is in fact the distribution we named Sam? Why, or why not?
∞ N
27. We define the mean, μ , as the expected value of the random variable: μ = ∫ u (df /du)du . Define
−∞
¯¯
¯
u = ∑i=1 (ui /N ) ,
where the u are N independent values of the random variable. Show that the expected value of u is μ .
i
¯¯
¯
28. A box contains a large number of plastic balls. An integer, W , in the range 1 ≤ W ≤ 20 is printed on each ball. There are
many balls printed with each integer. The integer specifies the mass of the ball in grams. Six random samples of three balls each are
drawn from the box. The balls are replaced and the box is shaken between drawings. The numbers on the balls in drawings I
through VI are:
I: 3, 4, 9
II: 1, 6, 17
III: 2, 5, 8
IV: 2, 6, 7
V: 3, 5, 6
VI: 2, 3, 10
(a) What are the population sets represented by the samples I through VI?
(b) Sketch the probability density function as estimated from sample I.
(c) Sketch the probability density function as estimated from sample II
(d) Using the data from samples I through VI, estimate the probability of drawing a ball of each mass in a single trial.
(e) Sketch the probability density function as estimated from the probability values in part (d).
(f) From the data in sample I, estimate the average mass of a ball in the box.
(g) From the data in sample II, estimate the average mass of a ball in the box.
(h) From the probability values calculated in part (d), estimate the average mass of a ball in the box.
CHAPTER OVERVIEW
4: The Distribution of Gas Velocities

4.17: Problems
This page titled 4: The Distribution of Gas Velocities is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
1
In Chapter 2, we assume that all of the molecules in a gas move with the same speed and use a simplified argument to conclude that
this speed depends only on temperature. We now recognize that the individual molecules in a gas sample have a wide range of
speeds; the velocities of gas molecules must be described by a distribution function. It is true, however, that the average speed
depends only on temperature.
James Clerk Maxwell was the first to derive the distribution function for gas velocities. He did it about 1860. We follow Maxwell’s
argument. For a molecule moving in three dimensions, there are three velocity components. Maxwell’s argument uses only one
assumption: the speed of a gas molecule is independent of the direction in which it is moving. Equivalently, we can say that the
components of the velocity of a gas molecule are independent of one another; knowing the value of one component of a molecule’s
velocity does not enable us to infer anything about the values of the other two components. When we use Cartesian coordinates,
Maxwell’s assumptionMaxwell’s assumption means also that the same mathematical model must describe the distribution of each
of the velocity components.
Since the velocity of a gas molecule has three components, we must treat the velocity distribution as a function of three random
variables. To understand how this can be done, let us consider how we might find probability distribution functions for velocity
components. We need to consider both spherical and Cartesian coordinate systems.
Let us suppose that we are able to measure the Cartesian-coordinate components v , v , and v of the velocities of a large number
x y z
of randomly selected gas molecules in a particular constant-temperature sample. Then we can transform each set of Cartesian
components to spherical-coordinate velocity covelocity componentsmponents v , θ , and φ . We imagine accumulating the results of
these measurements in a table like Table 1. As a practical matter, of course, we cannot make the measurements to complete such a
table. However, there is no doubt that, at every instant, every gas molecule can be characterized by a set of such velocity
components; the values exist, even if we cannot measure them. We imagine that we have such data only as a way to clarify the
properties of the distribution functions that we need.
Table 1. Molecular Velocity Components
Molecule Number vx vx vx v θ φ
1 vx (1) vy (1) vz (1) v (1) θ (1) φ(1)
2 vx (2) vy (2) vz (2) v (2) θ (2) φ(2)
3 vx (3) vy (3) vz (3) v (3) θ (3) φ(3)
4 vx (4) vy (4) vz (4) v (4) θ (4) φ(4)
… … … … … … …
N vx (N ) vx (N ) vz (N ) v (N ) θ (N ) φ(N )
These data have several important features. The scalar velocity, v , ranges from 0 to +∞ ; v , v , and v range from −∞ to +∞ .
x y z
In §2, we see that θ varies from 0 to π; and φ ranges from 0 to 2π. Each column represents data sampled from the distribution of
the corresponding random variable. In Chapter 3, we find that we can use such data to find mathematical models for such
distributions. Here, we can find mathematical models for the cumulative distribution functions f (v ), f (v ), and f (v ) . We x x y y z z
can approximate the graph of f (v ) by plotting the rank probability of v versus v . We expect this plot to be sigmoid; at any v ,
x x x x x
the slope of this plot is the probability-density function, df (v )/dv . The probability density function for v depends only on v ,
x x x x x
because the value measured for v is independent of the values measured for v and v . However, by Maxwell’s assumption, the
x y z
functions describing the distribution of v and v are the same as those describing the distribution of v . While redundant, it is
y z x
convenient to introduce additional symbols to represent these probability density functions. We define ρ (v ) = df (v )/dv , x x x x x
ρ (v ) = df (v )/dv , and ρ (v ) = df (v )/dv .

y y y y y z z z z z
When we find these one-dimensional distribution functions by modeling the experimental data in this way, each v datum that we x
use in our analysis comes from an observation on a molecule and is associated with particular v and v values. These values of v y z y
and v can be anything from −∞ to +∞ . This is a significant point. The functions f (v ) and df (v )/dv are independent of
z x x x x x
v and v . We can also say that df (v )/dv

y z x describes the distribution of v when v and v are averaged over all the values it is
x x x y z
possible for them to have.
To clarify this, let us consider another cumulative probability distribution function, f (v , v , v ), which is just the fraction of all xyz x y z
molecules whose respective Cartesian velocity components are less than v , v , v . Since f (v ), f (v ), and f (v ) are the x y z x x y y z z
fractions whose components are less than v , v , and v , respectively, their product is equal to f (v , v , v ) We have x y z xyz x y z
(v , v , v ) = f (v ) f (v ) f (v ) . For the velocity of a randomly selected molecule, (v , v , v ), to be included in the

∗ ∗ ∗
fxyz x y z x x y y z z x y z
fraction represented by f (v , v , v ), the velocity must be in the particular range +∞<v <v , +∞<v <v , and
xyz x y z
∗
x x
∗
y y
+∞<v <v .
∗
z z
However, for a velocity v to be included in f (v ), we must have v < v , v < ∞ , and v < ∞ ; that is, the components v and
∗
x x x
∗
x x
∗
y
∗
z
∗
y
v can have any values. Since the probability that v , v , and v satisfy v < v , v < v , and v < v is
∗ ∗ ∗ ∗
z x y z x x y y z z
∗ ∗ ∗
P (vx < vx , vy < vy , vz < vz ) = fxyz (vx , vy , vz )
= fx (vx )fy (vy )fz (vz )
the probability that v is included in f ∗

x x (vx ) becomes
∗ ∗ ∗
P (vx < vx , vy < ∞, vz < ∞) = fxyz (vx , ∞, ∞)
= fx (vx ) fy (∞) fz (∞)
= fx (vx )
For our purposes, we need to be able to express the probability that the velocity lies within any range of velocities. Let us use ʋ to
designate a particular “volume” region in velocity space and use P (ʋ ) to designate the probability that the velocity of a randomly
selected molecule is in this region. When we let ʋ be the region in velocity space in which x-components lie between v x
and v + dv , y -components lie between v , and v + dv , and z -components lie between v and v + dv ,
x x y y y z z z dP ( ʋ ) denotes the
probability that the velocity of a randomly chosen molecule, (v , v , v ) , satisfies the conditions v ∗
x
∗
y
∗
z x
∗
< vx < vx + dvx ,
v < v < v + dv
y
∗
y , and v < v < v + dv .
y y z
∗
z z z
dP ( ʋ ) is an increment of probability. The dependence of dP ( ʋ ) on vx , vy , vz , dvx , dvy , and dvz can be made explicit by
introducing a new function, ρ (v , v , v ), defined by x y z
dP ( ʋ ) = ρ (vx , vy , vz ) dvx dvy dvz
Since dv dv dv is the volume available in velocity space for velocities whose x-components are between v and v + dv ,
x y z x x x
whose y -components are between v , and v + dv , and whose z -components are between v and v + dz , we see that
y y y z z
ρ (v , v , v ) is a probability density function in three dimensions. The value of ρ (v , v , v ) is the probability, per unit volume in
x y z x y z
velocity space, that a molecule has the velocity (v , v , v ) . For any velocity, (v , v , v ) , there is a value of ρ (v , v , v ); this
x y z x y z x y z
value is just a number. If we want the probability of finding a velocity within some small volume of velocity space around
(v , v , v ) , we can find it by multiplying ρ (v , v , v ) by this volume.
x y z x y z
From the one-dimensional probability-density functions, the probability that the x-component of a molecular velocity lies between
vx and v + dv , is just (df (v )/dv ) dv , whatever the values of v and v . The probability that the y -component lies between
x x x x x x y z
v and v + dv , is just (df (v )/dv ) dv , whatever the values of v

y y y and v . The probability that the z -component lies between
y y y x z
y
v and v + dv , is just (df (v )/dv ) dv , whatever the values of v

z z z z and v . When we interpret Maxwell’s assumption to mean
z z z x y
that these are independent probabilities, the probability that all three conditions are realized simultaneously is
dfx (vx ) dfy (vy ) dfz (vz )

dP (ʋ) = ( )( )( ) dvx dvy dvz = ρ (vx , vy , vz ) dvx dvy dvz
dvx dvy dvz
Evidently, the product of these three one-dimensional probability densities is the three-dimensional probability density. We have
ρ (vx , vy , vz ) = ( )( )( ) = ρx (vx ) ρy (vy ) ρz (vz )
dvx dvy dvz
From Maxwell’s assumption, we have derived the conclusion that ρ (v , v , v ) can be expressed as a product of the one- x y z
dimensional probability densities (df (v )/dv ) dv , (df (v )/dv ) dv , and (df (v )/dv ) dv . Since these are probability
x x x y y y z z z
densities, we have
∞ ∞ ∞
∫ ( )dvx = ∫ ( )dvy = ∫ ( )dvz = 1
−∞
dvx −∞
dvy −∞
dvz
and
∞
∫∫∫ ρ (vx , vy , vz ) dvx dvy dvz = 1

−∞
Moreover, because the Cartesian coordinates differ from one another only in orientation, (df (vx )/dvx ) dvx , (df (vy )/dvy ) dvy ,
and (df (v )/dv ) dv must all be the same function.
z z z
Figure 1. Transformation from Cartesian to spherical coordinates.

To summarize the development above, we define ρ (v , v , v ) independently of df (v )/dv , df (v
x y z x x x y y )/dvy , and dfz (vz )/dvz .
Then, from Maxwell’s assumption that the three one-dimensional probabilities are independent, we find
ρ (vx , vy , vz ) = ( )( )( )
dvx dvy dvz
= ρx (vx ) ρy (vy ) ρz (vz )
Alternatively, we could take Maxwell’s assumption to be that the three-dimensional probability density function is expressible as a
product of three one-dimensional probability densities:
ρ (vx , vy , vz ) = ρx (vx ) ρy (vy ) ρz (vz )
In this case, the relationships of ρ (v ), ρ (v ), and ρ (v ), to the one-dimensional cumulative probabilities (f (v ), etc.) must
x x y y z z x x
be deduced from the properties of ρ (v , v , v ). As emphasized above, our deduction of f (v ) from experimental data uses v
x y z x x x
values that are associated with all possible values of v and v . That is, what we determine in our (hypothetical) experiment is
y z
vx ∞
fx (vx ) = ∫ ∫∫ ρ (vx , vy , vz ) dvx dvy dvz

vx =−∞ vy,z =−∞
vx ∞ ∞
=∫ ρx (vx ) dvx ∫ ρy (vy ) dvy ∫ ρz (vz ) dvz

−∞ −∞ −∞
vx
=∫ ρx (vx ) dvx
−∞
from which it follows that

dfx (vx )
=ρ (vx )
x
dvx
Figure 2. The differential volume element in spherical coordinates.
This page titled 4.1: Distribution Functions for Gas-velocity Components is shared under a CC BY-SA 4.0 license and was authored, remixed,
We introduce the idea of a three-dimensional probability-density function by showing how to find it from data referred to a
Cartesian coordinates system. The probability density associated with a particular molecular velocity is just a number—a number
that depends only on the velocity. Given a velocity, the probability density associated with that velocity must be independent of our
choice of coordinate system. We can express the three-dimensional probability density using any coordinate system. We turn now
to expressing velocities and probability density functions using spherical coordinates.
Just as we did for the Cartesian velocity components, we deduce the cumulative probability functions f (v) , f (θ) , and f (φ) for v θ φ
the spherical-coordinate components. Our deduction of f (v) from the experimental data uses v -values that are associated with all
v
possible values of θ and φ . Corresponding statements apply to our deductions of f (θ) , and f (φ). We also obtain their θ φ
derivatives, the probability-density functions df (v)/dv , df (θ)/dθ , and df (φ)/dφ. From the properties of probability-density
v θ φ
functions, we have
∞ π 2π
dfv (v) dfθ (θ) df (φ)
φ
∫ ( )dv = ∫ ( )dθ = ∫ ( )dφ = 1
0
dv 0
dθ 0
dφ
Let ʋ ′ be the arbitrarily small increment of volume in velocity space in which the v -, θ -, and φ -components of velocity lie between
v and v + dv , θ and θ + dθ , and φ and φ + dφ . Then the probability that the velocity of a randomly selected molecule lies within
ʋ ′ is
dfv (v) dfθ (θ) dfφ (φ)

dP ( ʋ ′) = ( )( )( ) dvdθdφ
dv dθ dφ
Note that the product

( )( )( )
dv dθ dφ
is not a three-dimensional probability density function. This is most immediately appreciated by recognizing that dvdθdφ is not an
incremental “volume” in velocity space. That is, ʋ ′ ≠ dvdθdφ
We let ρ (v, θ, φ) be the probability-density function for the velocity vector in spherical coordinates. When v , θ , and φ specify the
velocity, ρ (v, θ, φ) is the probability per unit volume at that velocity. We want to use ρ (v, θ, φ) to express the probability that an
arbitrarily selected molecule has a velocity vector whose magnitude lies between v and v + dv , while its θ -component lies between
θ and θ + dθ , and its φ -component lies between φ and φ + dφ . This is just ρ (v, θ, φ) times the velocity-space “volume”
included by these ranges of v , θ , and φ .

When we change from Cartesian coordinates, ⇀ v = (v , v , v ) , to spherical coordinates, v = (v, θ, φ) , the transformation is
x y z
⇀
v = vsinθ cosφ , v = vsinθ sinφ , v = vcosθ . (See Figure 1.) As sketched in Figure 2, an incremental increase in each of the
x y z
coordinates of the point specified by the vector (v, θ, φ) advances the vector to the point (v + dv, θ + dθ, φ + dφ) . When dv , dθ ,
and dφ are arbitrarily small, these two points specify the diagonally opposite corners of a rectangular parallelepiped, whose edges
have the lengths dv , vdθ , and vsinθ dφ . The volume of this parallelepiped is v sinθ dvdθdφ . Hence, the differential volume
2
elementdifferential volume element in Cartesian coordinates, dv dv dv , becomes v sinθ dvdθdφ in spherical coordinates.
x y z
2
vx , vy , vz
Mathematically, this conversion is obtained using the absolute value of the Jacobian, J ( v,θ,φ
) , of the transformation. That is,
∣ vx , vy , vz ∣
dvx dvy dvz = ∣J ( ) ∣ dvdθdφ
∣ v, θ, φ ∣
where the Jacobian is a determinate of partial derivatives

∣ ∂ vx /∂v ∂ vx /∂θ ∂ vx /∂φ ∣
vx , vy , vz ∣ ∣
J( ) = ∣ ∂ vy /∂v ∂ vy /∂θ ∂ vy /∂φ ∣
v, θ, φ ∣ ∣
∣ ∂ vz /∂v ∂ vz /∂θ ∂ vz /∂φ ∣
2
= v sinθ
Since the differential unit of volume in spherical coordinates is 2
v sinθ dvdθdφ , the probability that the velocity components lie
within the indicated ranges is
dP ( ʋ ′) = ρ (v, θ, φ) v2 sinθ dvdθdφ
We can develop the next step in Maxwell’s argument by taking his assumption to mean that the three-dimensional probability
density function is expressible as a product of three one-dimensional functions. That is, we take Maxwell’s assumption to assert the
existence of independent functions ρ (v) , ρ (θ) , and ρ (φ) such that ρ (v, θ, φ) = ρ (v) ρ (θ) ρ (φ) . The probability that the
v θ φ v θ φ
v -, θ -, and φ -components of velocity lie between v and v + dv , θ and θ + dθ , and φ and φ + dφ becomes

dP (ʋ ′) =( )( )( ) dvdθdφ
dv dθ dφ
2
= ρ (v, θ, φ) v sinθdvdθdφ
2
= ρv (v) ρθ (θ) ρφ (φ) v sinθdvdθdφ
Since v , θ , and φ are independent, it follows that

dfv (v)
2
=v ρ (v)
v
dv
dfθ (θ)
= ρθ (θ) sinθ
dθ
dfφ (φ)
= ρφ (φ)
dφ
Moreover, the assumption that velocity is independent of direction means that ρ (θ) must actually be independent of θ ; that is,
θ
ρ (θ) must be a constant. We let this constant be α ; so ρ (θ) = α . By the same argument, we set ρ (φ) = α . Each of these
θ θ φ
θ θ φ
probability-density functions must be normalized. This means that

∞
2
1 =∫ v ρv (v)dv
0
1 =∫ αθ sinθ dθ = 2 αθ
0
2π
1 =∫ αφ dφ = 2π αφ
0
from which we see that ρ (θ) = α = 1/2 and ρ (φ) = α = 1/2π . It is important to recognize that, while ρ (v ), ρ (v ), and
θ θ φ φ x x y y
ρ (v ) are probability density functions, ρ (θ) and ρ (v) are not. (However, ρ (φ) is a probability density function.) We can see
z
z θ v φ
this by noting that, if ρ (θ) were a probability density, its integral over all possible values of θ (0 < θ < π) would be one. Instead,
θ
we find
π π
∫ ρθ (θ) dθ = ∫ dθ/2 = π/2

0 0
Similarly, when we find ρ v

(v) , we can show explicitly that
∞
∫ ρv (v) dv ≠ 1
0
Our notation now allows us to express the probability that an arbitrarily selected molecule has a velocity vector whose magnitude
lies between v and v + dv , while its θ -component lies between θ and θ + dθ , and its φ -component lies between φ and φ + dφ
using three equivalent representations of the probability density function:
1
2 2 2
dP ( ʋ ′) = ρ (v, θ, φ) v sinθdvdθdφ − ρv (v) ρθ (θ) ρφ (φ) v sinθdvdθdφ = ( ) ρv (v) v sinθ dvdθdφ
4π
The three-dimensional probability-density function in spherical coordinates is
ρv (v)
ρ (v, θ, φ) = ρv (v) ρθ (θ) ρφ (φ) =
4π
This shows explicitly that ρ (v, θ, φ) is independent of θ and φ ; if the speed is independent of direction, the probability density
function that describes velocity must be independent of the coordinates, θ and φ , that specify its direction.
This page titled 4.2: Probability Density Functions for Velocity Components in Spherical Coordinates is shared under a CC BY-SA 4.0 license and
was authored, remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a
detailed edit history is available upon request.
To this point, we have been developing our ability to characterize the gas-velocity distribution functions. We now want to use
Maxwell’s argument to find them. We have already introduced the first step, which is the recognition that three-dimensional
probability-density functions can be expressed as products of independent one-dimensional functions, and that ρ (θ) , and ρ (φ) θ φ
are the constants 1/2 and 1/2π. Now, because the probability density associated with any given velocity is just a number that is
independent of the coordinate system, we can equate the three-dimensional probability-density functions for Cartesian and
spherical coordinates: ρ (v , v , v ) = ρ (v, θ, φ) so that
x y z
ρ (v)
v
ρ (vx ) ρ (vy ) ρ (vz ) =
x y z
4π
We take the partial derivative of this last equation with respect to v . The probability densities ρ x y
(vy ) and ρ
z
(vz ) are independent
of v . However, v is a function of v , because v = v + v + v . We find
x x
2 2
x
2
y
2
z
dρx (vx ) 1 ∂ρv (v) 1 dρv (v) ∂v

ρy (vy ) ρz (vz ) = ( ) = ( )( )
dvx 4π ∂vx 4π dv ∂vx
vy vv vy vz
Since v 2 2 2 2
= vx + vy + vz , 2v(∂v/∂ v x )v v
y z
= 2 vx and
∂v vx
( ) =
∂vx vy vz
v
Making this substitution and dividing by the original equation gives
dρ (vx ) ρy (vy ) ρz (vz ) vx 1 dρ (v)

x v
=
dvx ρx (vx ) ρy (vy ) ρz (vz ) v ρv (v) dv
Cancellation and rearrangement of the result leads to an equation in which the independent variables v and v are separated. This x
means that each term must be equal to a constant, which we take to be −λ . We find
1 dρx (vx ) 1 dρv (v)
( ) =( ) = −λ
vx ρx (vx ) dvx vρv (v) dv
so that
dρx (vx )
= −λ vx dvx
ρx (vx )
and
dρv (v)
= −λvdv
ρv (v)
From the first of these equations, we obtain the probability density function for the distributions of one-dimensional velocities. (See
Section 4.4.) The three-dimensional probability density function can be deduced from the one-dimensional function. (See Section
4.5.)
From the second equation, we obtain the three-dimensional probability-density function directly. Integrating from v=0 , where
ρ (0) has a fixed value, to an arbitrary scalar velocity, v , where the scalar-velocity function is ρ (v) , we have
v v
ρv (v) v
dρv (v)
∫ = −λ ∫ vdv
ρ (0) ρv (v) 0
v
or
2
−λv
ρ (v) = ρ (0) exp ( )
v v
2
The probability-density function for the scalar velocity becomes
2
dfv (v) −λv
2 2
= v ρv (v) = ρv (0) v exp ( )
dv 2
This is the result we want, except that it contains the unknown parameters ρv (0) and λ . The value of ρ v (0) must be such as to
make the integral over all velocities equal to unity. We require
∞
dfv (v)
1 =∫ ( ) dv
0 dv
∞ 2
−λv
2
= ρv (0) ∫ v exp ( ) dv
0
2
3/2
ρv (0) 2π
= ( )
4π λ
so that
3/2
λ
ρv (0) = 4π ( )
2π
∞ −−−−
where we use the definite integral ∫ x exp (−ax ) dx = (1/4) √π/a . (See Appendix D.) The scalar-velocity function in the
0
2 2 3
three-dimensional probability-density function becomes

3/2 2
λ −λv
ρv (v) = 4π ( ) exp ( )
2π 2
The probability-density function for the scalar velocity becomes

dfv (v)
2
= v ρv (v)
dv
3/2 2
λ −λv
2
= 4π ( ) v exp ( )
2π 2
The three-dimensional probability density in spherical coordinates becomes

ρ (v, θ, φ) = ρv (v) ρθ (θ) ρφ (φ)
3/2 2
λ −λv
=( ) exp ( )
2π 2
The probability that an arbitrarily selected molecule has a velocity vector whose magnitude lies between v and v + dv , while its θ -
component lies between θ and θ + dθ , and its φ -component lies between φ and φ + dφ becomes

dP ( ʋ ′) =( )( )( ) dvdθdφ
dv dθ dφ
2
1
2
=( ) ρv (v) v sinθdvdθdφ
4π
3/2 2
λ 2
−λv
=( ) v exp ( ) sinθdvdθdφ
2π 2
In Section 4.6, we again derive Boyle’s law and use the ideal gas equation to show that λ = m/kT .
This page titled 4.3: Maxwell's Derivation of the Gas-velocity Probability-density Function is shared under a CC BY-SA 4.0 license and was
authored, remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a
In Section 4.3, we find a differential equation in the function ρ (v ). Unlike the velocity, which takes values from zero to infinity,
x x
the x-component, v , takes values from minus infinity to plus infinity. The probability density at an infinite velocity, in either
x
direction, is necessarily zero. Therefore, we cannot evaluate the integral of dρ (v ) /ρ (v ) from v = −∞ to an arbitraryx x x x x
velocity, v . However, we know from Maxwell’s assumption that the probability density for v must be independent of whether the
x x
molecule is traveling in the direction of the positive x-axis or the negative x-axis. That is, ρ (v ) must be an even function; the x x
probability density function must be symmetric around v = 0 ; ρ (v ) = ρ (−v ) . Hence, we can express ρ (v ) relative to its
x x x x x x x
fixed value,ρ (0), at v = 0 . We integrate dρ (v ) /ρ (v ) from ρ (0) to ρ (v ) as v goes from zero to an arbitrary velocity,
x x x x x x x x x x
v , to find
x
ρ ( vx ) vx
x
dρx (vx )
∫ = −λ ∫ vx dvx
ρx (0)
ρx (vx ) 0
or
2
dfx (vx ) −λvx
ρx (vx ) = = ρx (0) exp ( )
dvx 2
The value of ρ (0) must be such as to make the integral of ρ

x x (vx ) over all possible values of v , $-\infty < v_x <\infty $, equal to
x
unity. That is, we must have

∞
1 =∫ ρx (vx ) dvx
−∞
∞
dfx (vx )
=∫ dvx
−∞ dvx
∞ 2
−λvx
= ρx (0) ∫ exp ( ) dvx
−∞
2
−−
−
2π
= ρx (0) √
λ
∞ −−− 1/2
where we use the definite integral ∫ −∞
2
exp (−ax ) dx = √π/a . (See Appendix D.) It follows that ρ x (0) = (λ/2π) . The one-
dimensional probability-density function becomes
dfx (vx )
ρx (vx ) =
dvx
1/2 2
λ −λvx
=( ) exp ( )
2π 2
Note that this is the normal distribution with μ = 0 and σ = λ . So λ is the variance of the normal one-dimensional
2 −1 −1
probability-density function. As noted above, in Section 4.6 we find that λ = m/kT .
This page titled 4.4: The Probability-density Function for Gas Velocities in One Dimension is shared under a CC BY-SA 4.0 license and was
In Section 4.4, we derive the probability density function for one Cartesian component of the velocity of a gas molecule. The
probability density functions for the other two Cartesian components are the same function. For ⇀ v = (v , v , v ) , we have x y z
2 2
v = v +v +v
x
2
y , and 2
z
1/2
2
dfx (vx ) λ −λvx
=( ) exp ( )
dvx 2π 2
1/2 2
dfy (vy ) λ −λvy
=( ) exp ( )
dvy 2π 2
1/2 2
dfz (vz ) λ −λvz
=( ) exp ( )
dvz 2π 2
We now want to derive the three-dimensional probability density function from these relationships. Given these probability density
functions for the Cartesian components of ⇀
v , we can find the probability density function in spherical coordinates
dfx ( vx ) dfy ( vy ) dfz ( vz )

( )( )( )
dvx dvy dvz
2 2 2
3/2 −λvx −λvy −λvz
λ
=( ) exp ( ) exp ( ) exp ( )
2π 2 2 2
2
λ 3/2 −λv
=( ) exp ( )
2π 2
= ρ (v, θ, φ)
Since the differential volume element in spherical coordinates is v sinθ dvdθdφ , the probability that a molecule has a a velocity
2
vector whose magnitude lies between v and v + dv , while its θ -component lies between θ and θ + dθ , and its φ -component lies
between φ and φ + dφ becomes
dfv (v) dfθ (θ) df (φ)
φ
( )( )( ) dvdθdφ
dv dθ dφ
2
3/2 2
λ 2 −λv
=( ) v exp ( ) sinθdvdθdφ
2π 2
(We found the same result in Section 4.3, of course.) We can find the probability-density function for the scalar velocity by
eliminating the dependence on the angular components. To do this, we need only sum up, at a given value of v , the contributions
from all possible values of θ and φ , recalling that 0 ≤ θ < π and 0 ≤ φ < 2π . This sum is just
π 2π
∫ ( ) dθ ∫ ( ) dφ =
dv θ=0
dθ φ=0
dφ
3/2 2 π 2π
λ −λv
2
=( ) v exp ( )∫ sinθdθ ∫ dφ
2π 2 θ=0 φ=0
π dfθ (θ) 2π dfφ (φ) π 2π

Since ∫
θ=0
(
dθ
)dθ = ∫
φ=0
(
dφ
) dφ = 1 , ∫
0
sinθdθ = 2 , and ∫
0
dφ = 2π , we again obtain the Maxwell-Boltzmann
probability-density function for the scalar velocity:
3/2 2
dfv (v) λ −λv
2
= 4π ( ) v exp ( )
dv 2π 2
Unlike the distribution function for the Cartesian components of velocity, the Maxwell-Boltzmann distribution for scalar velocities
is not a normal distribution. Possible speeds lie in the interval 0 ≤ v < ∞ . Because of the v term, the Maxwell-Boltzmann 2
equation is asymmetric; it has a pronounced tail at high velocities.
This page titled 4.5: Combining the One-dimensional Probability Density Functions is shared under a CC BY-SA 4.0 license and was authored,
In Chapter 2, we derive Boyle’s lawBoyle’s law using simplifying assumptions. We are now able to do this derivation much more
rigorously. We consider the collisions of gas molecules with a small portion of the wall of their container. We suppose that the wall
is smooth, so that we can select a small and compact segment of it that is arbitrarily close to being planar. We denote both the
segment of the wall and its area as A . A can have any shape so long as it is a smooth, flat surface enclosed by a smooth curve.
Let the volume of the container be V and the number of gas molecules in the container be N . We imagine that we follow the
trajectory of one particular molecule as it moves to hit the wall somewhere within A . We begin our observations at time t = 0 and
suppose that the collision occurs at time t .
Figure 3. Trajectory of a molecule colliding with a wall of its container.

As sketched in Figure 3, we erect a Cartesian coordinate system with its origin at the location in space of the molecule at time
t = 0 . We orient the axes of this coordinate system so that the xy-plane is parallel to the plane of A , and the z-axis is pointed
toward the wall. Then the unit vector along the z -axis and a vector perpendicular to A are parallel to one another. It is convenient
to express the velocity of the selected molecule in spherical coordinates. We suppose that, referred to the Cartesian coordinate
⇀
system we have erected, the velocity vector of the selected molecule is (v , θ , φ ). The vector v t, drawn from the origin of our
∗ ∗ ∗ ∗
Cartesian system to the point of impact on the wall, follows the trajectory of the molecule from time zero to time t . The z -
component of the molecular velocity vector is normal to the plane of A at the point of impact; the magnitude of the z -component
v cosθ . The perpendicular distance from the plane of A to the xy-plane of the Cartesian system is v tcosθ .
∗ ∗ ∗ ∗
We assume that the collision is perfectly elastic. Before collision, the velocity component perpendicular to the wall is
∗
v = v cosθ
z . Afterward, it is v = −v cosθ . Only this change in the v component contributes to the force on the wall within
∗
z
∗ ∗
z
A . (The v x and v components are not changed by the collision.) During the collision, the molecule’s momentum change is
y
−2mv cosθ . During our period of observation, the average force on the molecule is thus (−2mv cosθ )/t . The force that the
∗ ∗ ∗ ∗
molecule exerts on the wall is (2mv cosθ )/t, and hence the contribution that this particular collision—by one molecule traveling
∗ ∗
at velocity v — makes to the pressure on the wall is

∗
∗ ∗
2m v cosθ
∗
P1 (v ) =
At
We want to find the pressure on segment A of the wall that results from all possible impacts. To do so, we recognize that any other
molecule whose velocity components are v , θ , and φ , and whose location at time t = 0 enables it to reach A within time t ,
∗ ∗ ∗
makes the same contribution to the pressure as the selected molecule does. Let us begin by assuming that the velocities of all N of
the molecules in the volume, V , are the same as that of the selected molecule. In this case, we can find the number of the molecules
in the container that can reach A within time t by considering a tubular segment of the interior of the container. The long axis of
this tube is parallel to the velocity vector of the selected molecule. The sides of this tube cut the container wall along the perimeter
of A . This tube also cuts the xy-plane (the z = 0 plane) of our coordinate system in such a way as to make an exact replica of A in
this plane. Call this replica A . o
The area of A is A ; the plane of A is parallel to the plane of A ; and the perpendicular distance between the plane of A and the
o o
plane of A is v tcosθ . The volume of this tube is therefore Av tcosθ . Since there are N /V molecules per unit volume, the
o ∗ ∗ ∗ ∗
total number of molecules in the tube is (AN v tcosθ )/V . When we assume that every molecule has velocity components v , θ ,
∗ ∗ ∗ ∗
and φ , all of the molecules in the tube reach A within time t , because each of them travels parallel to the selected molecule, and
∗
each of them is initially at least as close to A as is the selected molecule. Therefore, each molecule in the tube contributes
P (v ) = 2m v cosθ /At to the pressure at A . The total pressure is the pressure per molecule multiplied by the number of
∗ ∗ ∗
1
molecules:
∗ ∗ ∗ ∗ 2
∗ ∗
2m v cosθ AN v tcosθ 2mN (v cosθ )
( )( ) =
At V V
However, the molecular velocities are not all the same, and the pressure contribution 2mN (v cosθ ) /V is made only by that ∗ ∗ 2
fraction of the molecules whose velocity components lie in the intervals θ < θ < θ + dθ and φ < φ < φ + dφ . This fraction ∗ ∗ ∗ ∗
is
3/2 ∗ 2
2
λ 2
−λ(v )
∗ ∗ ∗ ∗ ∗ ∗ ∗
ρ (v , θ , φ ) (v ) sinθ dvdθdφ = ( ) (v ) exp ( ) sinθ dvdθdφ
2π 2
so that the pressure contribution from molecules whose velocity components lie in these ranges is
∗ ∗ 2 3/2 ∗ 2
2mN (v cosθ ) λ 2
−λ(v )
∗ ∗
dP = ×( ) (v ) exp ( ) sinθ dvdθdφ
V 2π 2
The total pressure at A is just the sum of the contributions from molecules with all possible combinations of velocities v , θ , and ∗ ∗
φ . To find this sum, we integrate over all possible velocity vectors. The allowed values of v are 0 ≤ v < ∞ . There are no
∗
constraints on the values of φ ; we have 0 ≤ φ < 2π . However, since all of the impacting molecules must have a velocity
component in the positive z-direction, the possible values of θ lie in the interval 0 ≤ θ < π/2 . We designate the velocity of the
original molecule as (v , θ , φ ) and retain this notation to be as specific as possible in describing the tube bounded by A and A .
∗ ∗ ∗ o
However, the velocity components of an arbitrary molecule can have any of the allowed values. To integrate (See Appendix D)
over the allowed values, we drop the superscripts. The pressurepressure:on wall at A becomes
3/2
2mN λ
P = ( ) ×
V 2π
∞ 2 π/2 2π
4
−λv 2
∫ v exp ( ) dv ∫ cos θ sinθ dθ ∫ dφ
0
2 0 0
3/2 2 1/2
2mN λ 3 2 2π 1
= ( ) [ ( ) ( ) ][ ] [2π] = mN /V λ
V 2π 8 λ λ 3
and the pressure–volume product becomes

mN
PV =
λ
Since m, N , and λ are constants, this is Boyle’s law. Equating this pressure–volume product to that given by the ideal gas
equation, we have mN /λ = N kT so that
m
λ =
kT
Finally, the Maxwell-Boltzmann equation becomes

3/2 2
dfv (v) m −mv
2
= 4π ( ) v exp ( )
dv 2πkT 2kT
and the probability density becomes
3/2 2
m 2
−mv
ρ (v, θ, φ) = ( ) v exp ( )
2πkT 2kT
This derivation can be recast as a computation of the expected value of the pressurepressure:expected value. To do so, we rephrase
our description of the system: A molecule whose velocity components are (v , θ , φ ) creates a pressure 2mv cosθ /At on the
∗ ∗ ∗ ∗ ∗
area A with a probability of Av tcosθ /V . (The latter term is the probability that a molecule, whose velocity is (v , θ , φ ), is, at
∗ ∗ ∗ ∗ ∗
time t = 0 , in a location from which it can reach A within time t. If the molecule is to hit the wall within time t , at time t = 0 the
molecule must be within the tubular segment of volume is Av tcosθ . The probability that the molecule is within this tubular
∗ ∗
segment is equal to the fraction of the total volume that this segment occupies.) Therefore, the product
∗ ∗ ∗ ∗
2m v cosθ Av tcosθ 2m ∗ 2
∗
( )( ) = (v cosθ )
At V V
is the pressure contribution of a molecule with velocity (v , θ , φ ), when θ is in the interval 0 ≤ θ < π/2 . The total pressure
∗ ∗ ∗ ∗ ∗
per molecule is the expected value of this pressure contribution; the expected value is the integral, over the entire volume of
velocity space, of the pressure contribution times the probability density function for velocities.
It is useful to view the Maxwell-Boltzmann equation as the product of a term
2
exp (−m v /2kT )
—called the Boltzmann factorBoltzmann factor—and a pre-exponential term that is proportional to the number of ways that a
molecule can have a given velocity, v . If there were no constraints on a molecule’s speed, we would expect that the number of
molecules with speeds between v and v + dv would increase as v increases, because the probability that a molecule has a speed
between v and v + dv is proportional to the volume in velocity space of a spherical shell of thickness dv . The volume of a
spherical shell of thickness dv is 4π v dv, which increases as the square of v . However, the number of molecules with large values
2
of v is constrained by the conservation of energy. Since the total energy of a collection of molecules is limited, only a small
proportion of the molecules can have very large velocities. The Boltzmann factor introduces this constraint. A molecule whose
mass is m and whose scalar velocity is v has kinetic energy ϵ = mv /2 . The Boltzmann factor is often written as exp (−ϵ/kT ) .
2
This page titled 4.6: Boyle's Law from the Maxwell-Boltzmann Probability Density is shared under a CC BY-SA 4.0 license and was authored,
There are numerous applications of the Maxwell-Boltzmann equation. These include predictions of collision frequencies, mean-
free paths, effusion and diffusion rates, the thermal conductivity of gases, and gas viscosities. These applications are important, but
none of them is a direct test of the validity of the Maxwell-Boltzmann equation.
The validity of the equation has been demonstrated directly in experiments in which a gas of metal atoms is produced in an oven at
a very high temperature. As sketched in Figure 4, the gas is allowed to escape into a vacuum chamber through a very small hole in
the side of the oven. The escaping atoms impinge on one or more metal plates. Narrow slits cut in these plates stop any metal atoms
whose flight paths do not pass though the slits. This produces a beambeam:of metal atoms of metal atoms whose velocity
distribution is the same as that of the metal-atom gas inside the oven. The rate at which metal atoms arrive at a detector is
measured. Various methods are used to translate the atom-arrival rate into a measurement of their speed.
Figure 4. Producing a beam of metal atoms.

One device uses a solid cylindrical drum, which rotates on its cylindrical axis. As sketched in Figure 5, a spiral groove is cut into
the cylindrical face of this drum. This groove is cut with a constant pitch. When the drum rotates at a constant rate, an atom
traveling at a constant
velocity parallel to the cylindrical axis can traverse the length of the drum while remaining within the groove. That is, for a given
rotation rate, there is one critical velocity at which an atom can travel in a straight line while remaining in the middle of the groove
all the way from one end of the drum to the other. If the atom moves significantly faster or slower than this critical velocity, it
collides with—and sticks to—one side or the other of the groove.
Figure 5. Device to select metal atoms having a specified velocity.

Since the groove has a finite width, atoms whose velocities lie in a narrow range about the critical velocity can traverse the groove
without hitting one of the sides.
Let us assume that the groove is cut so that the spiral travels half way around the cylinder. That is, if we project the spiral onto one
of the circular faces of the drum, the projection traverses an angle of 180 on the face. In order to remain in the middle of this
o
groove all the way from one end of the drum to the other, the atom must travel the length of the cylindrical drum in exactly the
same time that it takes the drum to make a half-rotation. Let the critical velocity be v . Then the time required for the atom to
critical
traverse the length, d, of the drum is d/v . If the drum rotates at u cycles/sec, the time required for the drum to make one-half
critial
rotation is 1/2u. Thus, the atom will remain in the middle of the groove all the way through the drum if
vcritial = 2ud
By varying the rotation rate, we can vary the critical velocity.

Because the groove has a finite width, atoms whose velocities are in a range $v_{min}<v_{max}$ can successfully traverse the
groove. Whether or not a particular atom can do so depends on its velocity, where it enters the groove, and the width of the groove.
Let the width of the groove be w and the radius of the drum be r, where the drum is constructed with r ≫ w . A slower atom that
enters the groove at the earliest possible time—when the leading edge of the groove first encounters the beam of atoms—can
traverse the length of the groove in a longer time, t . A point on the circumference of the drum travels with speed 2πru. The
max
slowest atom traverses the length of the drum while a point on the circumference of the drum travels a distance πr + w . (To
intercept the slowest atom, the trailing edge of the groove must travel a distance equal to half the circumference of the drum, πr,
plus the width of the groove, w.) The time required for this rotation is the maximum time a particle can take to traverse the length,
so
tmax = (πr + w) / (2πru)
and
vmin = d/ tmax = 2πrud/ (πr + w)
A fast atom that enters the groove at the last possible moment—when the trailing edge of the grove just leaves the beam of atoms—
can still traverse the groove if it does so in the time, t that it takes the trailing edge of the groove to travel a distance πr − w .
min
So,
tmin = (πr − w) / (2πru)
and
vmax = d/ tmin = 2πrud/ (πr − w)
At a given rotation rate, the drum will pass atoms whose speeds are in the range
πr πr 2πrw 2w
Δv = vmax − vmin = 2ud ( − ) = 2ud ( ) ≈ vcritical ( )
2
πr − w πr + w (πr) −w
2 πr
So that
Δv 2w
≈
vcritial πr
The fraction of the incident atoms that successfully traverse the groove is equal to the fraction that have velocities in the interval
Δv centered on the critical velocity, v = 2ud .
critical
This page titled 4.7: Experimental Test of the Maxwell-Boltzmann Probability Density is shared under a CC BY-SA 4.0 license and was authored,
Expected values for several quantities can be calculated from the Maxwell-Boltzmann probability density function. The required
definite integrals are tabulated in Appendix D.
The most probable speed, v mp , is the speed at which the Maxwell-Boltzmann equation takes on its maximum value. At this speed,
we have
3/2 2
d df (v) d m −mv
2
0 = ( ) = [4π ( ) v exp( )]
dv dv dv 2πkT 2kT
3/2 2 3
m −mv mv
= [4π ( ) exp( )] [2v − ]
2πkT 2kT kT
from which
−−−− −−−
2kT kT
vmp = √ ≈ 1.414 √
m m
The average speed, v or ⟨v⟩ , is the expected value of the scalar velocity (g (v) = v ). We find
¯
¯¯
∞ 2
−−−− −−−
3/2
m 3
−mv 8kT kT
¯
¯¯
v = ⟨v⟩ = ∫ 4π ( ) v exp ( )dv = √ ≈ 1.596 √
0
2πkT 2kT πm m
¯
¯¯¯
¯
The mean-square speed, v or ⟨v
2 2
⟩ , is the expected value of the velocity squared (g (v) = v ): 2
∞ 2
¯
¯¯¯
¯ m 3/2 −mv 3kT
2 2 4
v = ⟨v ⟩ = ∫ 4π ( ) v exp ( )dv =
0 2πkT 2kT m
and the root mean-square speed, v rms , is

−−−− −−−
−−−− 3kT kT
2
vrms = √⟨v ⟩ = √ ≈ 1.732 √
m m
Example 4.8.1
Figure 6. The Maxwell-Boltzmann distribution function for N at 300K. 2
Figure 6 shows the velocity distribution 300 K for nitrogen molecules at 300 K.
Solution
2
Finally, let us find the variance of the velocity; that is, the expected value of (v − ⟨v⟩) :
2
variance (v) = σv
∞
df (v)
2
=∫ (v − ⟨v⟩) ( ) dv
0 dv
∞ ∞ ∞
df df df
2 2
=∫ v ( ) dv − 2⟨v⟩ ∫ v( ) dv + ⟨v⟩ ∫ ( )
0
dv 0
dv 0
dv
2 2
= ⟨v ⟩ − 2⟨v⟩⟨v⟩ + ⟨v⟩
2 2
= ⟨v ⟩ − ⟨v⟩
For N at 300 K, we calculate:

2
−1
vmp = 422 m s
¯¯ −1
⟨v⟩ = v̄ = 476 m s
−1
vrms = 517 m s
2 −3 −1
Variance (v) = σv = 40.23 × 10 m s
−1
σv = 201 m s
This page titled 4.8: Statistics for Molecular Speeds is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
At 300 K, the standard deviation of N speeds is about 40% of the average speed. Clearly the relative variation among molecular
2
speeds in a sample of ordinary gas is very large. Why do we not observe macroscopic effects from this variation? In particular, if
we measure the pressure at a small area of the container wall, why do we not observe pressure variations that reflect the wide
variety of speeds with which molecules strike the wall?
Qualitatively, the answer is obvious. A single molecule whose scalar velocity is v contributes P (v) = mv /3V to the pressure on
1
2
the walls of its container. (See problem 20.) When we measure pressure, we measure an average squared velocity. Even if we
measure the pressure over a very small area and a very short time, the
number of molecules striking the wall during the time of the measurement is very large. Consequently, the average speed of the
molecules hitting the wall during any one such measurement is very close to the average speed in any other such measurement.
We are now able to treat this question quantitatively. For N gas at 300 K and 1 bar, roughly 3×10 molecules collide with a
2
15
square millimeter of wall every microsecond. (See problem 12.) The standard deviation of the velocity of an N molecule is 2
201 m s . Using the central limit theorem, the standard deviation of the average of 3 × 10 molecular speeds is
−1 15
−1
201 m s −6 −1
− −−−−−− ≈ 4 × 10 ms
√ 3 × 1015
The distribution of the average of 3 × 10 15

molecular speeds is very narrow indeed.
Similarly, when molecular velocities follow the Maxwell-Boltzmann distribution function, we can show that the expected value of
the pressure for a single-molecule collision is ⟨P (v)⟩ = kT /V . (See problem 21.) The variance of the distribution of these
1
individual pressure measurements is σ 2

= 2 k T /3 V
P1 (v)
2
, so that the magnitude of the standard deviation is comparable to that of
2 2
the average:
−−−
σP (v) / ⟨P1 (v)⟩ = √2/3
1
For the distribution of averages of 3 × 10 15

pressure contributions, we find
Pavg = ⟨P1 (v)⟩
−−−
= √3/2σP
1 (v)
σP
1 (v)
σavg =
− −−−−−−
√ 3 × 1015
and
σavg
−8
≈ 1.5 × 10
Pavg
This page titled 4.9: Pressure Variations for Macroscopic Samples is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The pressure of a gas depends on the frequency with which molecules collide with the wall of their container. The rate at which gas
molecules escape through a very small opening in their container is called the effusion rate. The effusion rate rate also depends on
the frequency of collisions with the wall. (See problem 4.12.) Other gas properties depend not on the rate of collision with the wall,
but on the rate with which gas molecules collide with one another. We turn now to some of these properties. For these
considerations, we need to describe the motion of one molecule relative to another. We need the probability density function for the
relative velocity of two particles.
To describe the relative velocity of two particles, we introduce relative velocity coordinates. Let us begin by considering a
Cartesian coordinate frame, with x-, y -, and z-axes, whose origin is at a point O; we will use Oxyz to designate this set of axes.
⇀ ⇀ ⇀
We specify the location of particle 1 by the vector ⇀ r = (x , y , z ) = x
1 1 1 1 1 i + y1 j + z1 k and that of particle 2 by
= (x , y , z ) . We let the location of the center of mass of this two-particle
⇀
r 2 2 2 2
system be specified by ⇀
r 0 = (x0 , y0 , z0 ) . The vector from particle 1 to particle 2, ⇀
r 12 = (x12 , y12 , z12 ) , is the vector difference
⇀ ⇀ ⇀
r 12 = r 2 − r 1 = (x2 − x1 , y2 − y1 , z2 − z1 )
When the particles are moving, these vectors and their components are functions of time. Using the notation ẋ = dx /dt , we can 1 1
specify the velocity of particle 1, for example, as ⇀

v =d r
⇀
/dt = (ẋ , ẏ , ż ) . Our goal is to find the relative velocity vector,
1 1 1 1 1
/dt . We call the components of v the relative velocity coordinates.

⇀ ⇀ ⇀
v12 =d r 12 12
Figure 7. The center of mass frame

The essential idea underlying relative velocity coordinates is that the vectors ⇀
r and ⇀r contain the same information as the
0 12
vectors r and r . This is equivalent to saying that we can transform the locations as specified by (x , y , z ) and (x , y , z ) to
⇀
1
⇀
2 1 1 1 2 2 2
the same locations as specified by (x , y , z ) and (x , y , z ), and vice versa. To accomplish this, we write the equation
0 0 0 12 12 12
defining the x-component of the center of mass, x : 0
m1 (x1 − x0 ) + m2 (x2 − x0 ) = 0
which we rearrange to
x1 x2 1 1
+ =( + ) x0
m2 m1 m1 m2
Corresponding relationships can be written for the y - and z -components. It proves to be useful to introduce the reduced mass, μ ,
defined by
1 1 1
= +
μ m1 m2
Using the reduced mass, we can express the coordinates of the center of mass in terms of the coordinates of the individual particles.
That is,
μ μ
x0 = ( ) x1 + ( ) x2
m2 m1
μ μ
y0 = ( ) y1 + ( ) y2
m2 m1
μ μ
z0 = ( ) z1 + ( ) z2
m2 m1
Since, by definition, we also have
x12 = x2 − x1
y12 = y2 − y1
z12 = z2 − z1
we have developed the transformation from (x0 , y0 , z0 ) and (x12 , y12 , z12 ) to (x1 , y1 , z1 ) and (x2 , y2 , z2 ) . The inverse
transformation is readily found to be
x1 = x0 − (μ/ m1 ) x12
y1 = y0 − (μ/ m1 ) y12
z1 = z0 − (μ/ m1 ) z12
x2 = x0 + (μ/ m2 ) x12
y2 = y0 + (μ/ m2 ) y12
z2 = z0 + (μ/ m2 ) z12
Now we can create two new Cartesian coordinate frames. Which of these is more useful depends on the objective of the particular
analysis we have at hand. We call the first one the center of mass frame, O x y z . It is sketched in Figure 7. The x -, y -, and z -
′ ′ ′
′ ′ ′
O
axes of O x y z are parallel to the corresponding axes of Oxyz, but their origin, O , is always at the point occupied by the center
′ ′ ′
O O
of mass of the two-particle system. In this reference frame, the coordinates of particles 1 and 2 are their displacements from the
center of mass:
′
x = x1 − x0 = − (μ/ m1 ) x12
1
y = y1 − y0 = − (μ/ m1 ) y12
1
z = z1 − z0 = − (μ/ m1 ) z12
1
x = x2 − x0 = (μ/ m2 ) x12
2
y = y2 − y0 = (μ/ m2 ) y12
2
z = z2 − z0 = (μ/ m2 ) z12
2
The center of mass frame is particularly useful for analyzing interactions between colliding particles.
Figure 8. The particle-one centered frame.
For our purposes, a third Cartesian coordinate frame, which we will denote the O x y z frame, is more useful. It is sketched in
′′ ′′ ′′
Figure 8. The x -, y -, and z -axes of O x y z are parallel to the corresponding axes of Oxyz, but their origin, O , is always at
′′ ′′ ′′ ′′ ′′ ′′
1 1
the point occupied by particle 1. In this reference frame, the coordinates of particles 1 and 2 are
′′
x =0
1
′′
y =0
1
′′
z =0
1
′′
x = x2 − x1 = x12
2
′′
y = y2 − y1 = y12
2
′′
z = z2 − z1 = z12
2
and the coordinates of the center of mass are

′′
x = x0 − x1 = μx12 / m1
0
′′
y = y0 − y1 = μy12 / m1
0
′′
z = z0 − z1 = μz12 / m1
0
The O x y z frame is sometimes called the center of mass frame also. To avoid confusion, we call O x y z the particle-one
′′ ′′ ′′ ′′ ′′ ′′
1 1
centered frame. In the particle-one centered frame, particle 1 is stationary at the origin. With its tail at the origin, the vector
) specifies the position of particle 2 .
⇀
r12 = (x ,y ,z
12 12 12
We are interested in the relative velocity of particles 1 and 2. The velocity components for particles 1 and 2, and for their relative
velocity, are obtained by finding the time-derivatives of the corresponding displacement components. Since the transformations of
the displacement coordinates are linear, the velocity components transform from one reference frame to another in exactly the same
way that the displacement components do. We have
⇀ ⇀
v 0 = d r 0 /dt = (ẋ0 , ẏ 0 , ż 0 )
and
⇀ ⇀
v 12 = d r 12 /dt = (ẋ12 , ẏ 12 , ż 12 )
The vector ⇀ v12 specifies the velocity of particle 2, relative to a stationary particle 1. Just as ⇀ r and ⇀
0r contain the same
12
information as the vectors ⇀ r 1and ⇀r , the vectors v

2
⇀
and ⇀
0 v contain the same information as ⇀
12 v and ⇀v . Since a parallel
1 2
displacement leaves a vector unchanged, each of these vectors is the same in any of the three reference frames. In §11, we find the
probability density function for the magnitude of the scalar relative velocity, v = ∣∣⇀
12v ∣ . Since the probability is independent of
∣
12
direction, the probability that two molecules have relative velocity ⇀ v is the same as that they have relative velocity −⇀
12 v . (In 12
spherical coordinates, if ⇀ v 12= (v , θ, φ) , then − v

12
⇀
= (v
12 , θ + π, φ + π) .) The probability and magnitude of the relative
12
velocity are independent of which particle—if either—we choose to view as being stationary; they are independent of whether the
particles are approaching or receding from one another.
This page titled 4.10: Collisions between Gas Molecules Relative Velocity Coordinates is shared under a CC BY-SA 4.0 license and was authored,
From our development of the Maxwell-Boltzmann probability density functions, we can express the probability that the velocity
components of particle 1 lie in the intervals v to v + dv ; v to v + dv ; v to v + dv ; while those of particle 2
1x 1x 1x 1y 1y 1y 1z 1z 1z
simultaneously lie in the intervals v to v + dv ; v to v + dv ; v to v + dv as

2x 2x 2x 2y 2y 2y 2z 2z 2z
df (v1x ) df (v1y ) df (v1z ) df (v2x ) df (v2y ) df (v2z )

( )( )( )( )( )( )
dv1x dv1y dv1z dv2x dv2y dv2z
×dv1x dv1y dv1z dv2x dv2y dv2z
df (v1 ) df (v2 )
=( )( ) dv1 dv2
dv1 dv2
We want to express this probability using the relative velocity coordinates. Since the velocity of the center of mass and the relative
velocity are independent, we might expect that the Jacobian of this transformation is just the product of the two individual
Jacobians. This turns out to be the case. The Jacobian of the transformation
(ẋ1 , ẏ 1 , ż 1 , ẋ2 , ẏ 2 , ż 2 ) → (ẋ0 , ,˙

y0 , ż 0 , ẋ12 , ẏ 12 , ż 12 )
is a six-by-six determinate. It is messy, but straightforward, to show that it is equal to the product of two three-by-three
determinants and that the absolute value of this product is one. Therefore, we have
dv1x dv1y dv1z dv2x dv2y dv2z =dẋ1 dẏ 1 d ż 1 dẋ2 dẏ 2 d ż 2
= dẋ0 dẏ 0 d ż 0 dẋ12 dẏ 12 d ż 12
We transform the probability density by substituting into the one-dimensional probability density functions. That is,
2 2 2 2 2 2
3/2
⎛ −m1 (v +v +v ) ⎞ 3/2
⎛ −m2 (v +v +v ) ⎞
df (v1 ) df (v2 ) m1 1x 1y 1z m2 2x 2y 2z
( )( ) =( ) exp ⎜ ⎟×( ) exp ⎜ ⎟

dv1 dv2 2πkT 2kT 2πkT 2kT
⎝ ⎠ ⎝ ⎠
2 2 2 2 2 2
⎛ −m1 (v +v +v ) − m2 ( v +v +v ) ⎞
m1 m2 3/2 1x 1y 1z 2x 2y 2z
=( ) × exp ⎜ ⎟
2 2 2
4π k T 2kT
⎝ ⎠
m1 m2 2 2 2 2 2
2
3/2 − (ẋ0 + ẏ 0 + z ) − μ (ẋ12 + ẏ 12 + ż 12 )
m1 m2 μ 0
=( ) × exp ( )
2 2 2
4π k T 2kT
where the last expression specifies the probability density as a function of the relative velocity coordinates.
Next, we make a further transformation of variables. We convert the velocity of the center of mass, (ẋ , ẏ , ż ), and the relative 0 0 0
velocity, (ẋ , ẏ , ż ) , from Cartesian coordinates to spherical coordinates, referred to the Oxyz axis system. (The motion of the
12 12 12
center of mass is most readily visualized in the original frame Oxyz. The relative motion, ⇀ v , is most readily visualized in the 12
′′ ′′
Particle-One Centered Frame, O x y z . In O x y z , the motion of particle 2 is specified by ẋ = ẋ , ẏ = ẏ , and

′′ ′′ ′′ ′′ ′′ ′′
1 1 2 12 2 12
′′ ′′ ′′ ′′
ż
2
= ż 12. The motion of the center of mass is specified by ẋ = μẋ / m , ẏ = μẏ / m , and ż = μż
0
/ m . Since it is the
12 1 0 12 1 0 12 1
relative motion that is actually of interest, it might seem that we should refer the spherical coordinates to the O x y z frame. This
′′ ′′ ′′
is an unnecessary distinction because all three coordinate frames are parallel to one another, and ⇀ r and ⇀r are the same vectors 0 12
in all three frames.) Letting

2 2 2 2
v = ẋ0 + ẏ 0 + z
0 0
2 2 2 2
v = ẋ12 + ẏ 12 + z
12 12
the Cartesian velocity components are expressed in spherical coordinates by
ẋ0 = v0 sin θ0 cosφ0
ẏ 0 = v0 sin θ0 sin φ0
ż 0 = v0 cosθ0
ẋ12 = v12 sin θ12 cosφ12
ẏ 12 = v12 sin θ12 sin φ12
ż 12 = v12 cosθ12
The angles θ , θ , φ , and φ are defined in the usual manner relative to the Oxyz axis system. The Jacobian of this
0 12 0 12
transformation is a six-by-six determinate; which can again be converted to the product of two three-by-three determinates. We find
dẋ0 dẏ 0 dż 0 dẋ12 dẏ 12 dż 12 =
2 2
= v sin θ0 dv0 dθ0 dφ0 v sin θ12 dv12 dθ12 dφ12
0 12
The probability that the components of the velocity of the center of mass lie in the intervals v to v + dv ; θ to θ + dθ ; φ to 0 0 0 0 0 0 0
φ + dφ ; while the components of the relative velocity lie in the intervals v

0 0
to v + dv ; θ to θ + dθ ; φ to 12 12 12 12 12 12 12
φ12
+ dφ 12
; becomes
2 2
m1 m2 3/2 −m1 m2 v −μv
0 12
( ) exp ( ) exp ( )×
4π 2 k2 T 2
2μkT 2kT
2 2
v sin θ0 dv0 dθ0 dφ0 v sin θ12 dv12 dθ12 dφ12
0 12
We are interested in the probability increment for the relative velocityrelative velocity:probability density function irrespective of
the velocity of the center of mass. To sum the contributions for all possible motions of the center of mass, we integrate this
expression over the possible ranges of v , θ , and φ . We have
0 0 0
df (v12 ) df (θ12 ) df (φ12 )

( )( )( ) dv12 dθ12 dφ =
12
dv12 dθ12 dφ
12
∞ 2 π 2π
m1 m2 3/2 −m1 m2 v
2 0
=( ) ∫ v exp ( )dv0 × ∫ sin θ0 dθ0 ∫ dφ0
0
4π 2 k2 T 2
0
2μkT 0 0
2
−μv
2 12
× [v exp ( ) sin θ12 dv12 dθ12 dφ ]
12 12
2kT
2
μ 3/2 −μv
2 12
=( ) v exp ( ) sin θ12 dv12 dθ12 dφ12
12
2πkT 2kT
This is the same as the probability increment for a single-particle velocity—albeit with μ replacing m; v replacing v ; θ 12 12
replacing θ ; and φ replacing φ . As in the single-particle case, we can obtain the probability increment for the scalar component
12
of the relative velocity by integrating over all possible values of θ and φ . We find 12 12
2
df (v12 ) μ 3/2 −μv
2 12
= 4π ( ) v exp ( ) dv12
12
dv12 2πkT 2kT
In §8, we find the most probable velocity, the mean velocity, and the root-mean-square velocity for a gas whose particles have mass
m . By identical arguments, we obtain the most probable relative velocity, the mean relative velocity, and the root-mean-square
relative velocity. To do so, we can simply substitute μ for m in the earlier results. In particular, the mean relative velocity is
1/2 1/21
8kT kT
¯
¯¯
v12 = ⟨v12 ⟩ = ( ) ≈ 1.596 ( )
πμ πμ
If particles 1 and 2 have the same mass, m, the reduced mass becomes μ = m/2 . In this case, we have
1/2
2 (8kT ) –
⟨v12 ⟩ = ( ) = √2 ⟨v⟩
πm
We can arrive at this same conclusion by considering the relative motion of two particles that represents the average case. As
illustrated in Figure 9, this occurs when the two particles have the same speed, ⟨v⟩ , but are moving at 90-degree angles to one
another. In this situation, the length of the resultant vector—the relative speed— is just
–
¯¯
| v̄12 | = ⟨v12 ⟩ = √2 ⟨v⟩ .
This page titled 4.11: The Probability Density Function for the Relative Velocity is shared under a CC BY-SA 4.0 license and was authored,
Thus far in our theoretical development of the properties of gases, we have assumed that ideal gas molecules are point masses.
While they can collide with the walls of their container, point masses cannot collide with one another. As we saw in our discussion
of van der Waals equation, the deviation of real gases from ideal gas behavior is one indication that an individual gas molecule
occupies a finite volume.
To develop a model for molecular collisions, we need to know the size and shape of the colliding molecules. For a general model,
we want to use the simplest possible size and shape. Accordingly, we consider a model in which gas molecules are spheres with
well-defined radii. We let the radii of molecules 1 and 2 be σ and σ , respectively. See Figure 10. When such molecules collide,
1 2
their surfaces must come into contact, and the distance between their centers must be σ = σ + σ . We call σ the collision
12 1 2 12
radius.
Figure 10. The molecular collision radius.

Let us consider a molecule of type 1 in a container with a large number of molecules of type 2. We suppose that there are N 2
molecules of type 2 per unit volume. Every molecule of type 2 has some velocity, v , relative to the molecule of type 1. From our
12
development above, we know both the probability density function for v and the expected value ⟨v ⟩. Both molecule 1 and all of
12 12
the molecules of type 2 are moving with continuously varying speeds. However, it is reasonable to suppose that—on average—the
encounters between molecule 1 and molecules of type 2 are the same as they would be if all of the type 2 molecules were fixed at
random locations in the volume, and molecule 1 moved among them with a speed equal to the average relative velocity, ⟨v ⟩. 12
Under this assumption, a molecule 1 travels a distance equal to ⟨v ⟩ in unit time. As it does so, it collides with any type 2
12
molecule whose center is within a distance σ of its own center. For the moment, let us suppose that the trajectory of molecule 1 is
12
unaffected by the collisions it experiences. Then, in unit time, molecule 1 sweeps out a cylinder whose length is ⟨v ⟩ and whose 12
cross-sectional area is πσ . The volume of this cylinder is π σ ⟨v ⟩. (See Figure 11.)

2
12
2
12 12
Figure 11. The collision volume of a gas molecule in unit time.

Since there are N molecules of type 2 per unit volume, the number of type 2 molecules in the cylinder is N π σ ⟨v ⟩. Each of
2 2
2
12 12
these molecules is a molecule of type 2 that experiences a collision with molecule 1 in unit time. Letting ν̃ be the frequency
12
(number of collision per unit time) with which molecule 1 collides with molecules of type 2, we have
1/2
8kT
2 2
ν̃ 12 = N2 π σ ⟨v12 ⟩ = N2 π σ ( )
12 12
πμ
1/2
8πkT
2
= N2 σ ( )
12
μ
Two additional parameters that are useful for characterizing molecular collisions are τ , the mean time between collisions, and
12
λ12, the mean distance that molecule 1 travels between collisions with successive molecules of type 2. λ is called the mean free
12
path. The mean time between collisions is simply the reciprocal of the collision frequency,
τ12 = 1/ ν̃ 12
and the mean free path for molecule 1 is the distance that molecule 1 actually travels in this time, which is ⟨v ⟩ , not ⟨v 1 12 ⟩ , so that
1/2
(μ/m)
λ12 = ⟨v1 ⟩ τ12 =
2
N2 π σ
12
Now, we need to reevaluate the assumption that the trajectory of a molecule of a molecule 1 is unaffected by its collisions with
molecules of type 2. Clearly, this is not the case. The path of molecule 1 changes abruptly at each collision. The actual cylinder that
molecule 1 sweeps out will have numerous kinks, as indicated in Figure 12. The kinked cylinder can be produced from a straight
one by making a series oblique cuts (one for each kink) across the straight cylinder and then rotating the ends of each cut into
convergence. If we think of the cylinder as a solid rod, its volume is unchanged by these cuttings and rotations. The volume of the
kinked cylinder is the same as that of the straight cylinder. Thus, our conclusions about the collision frequency, the mean time
between collisions, and the mean free path are not affected by the fact that the trajectory of molecule 1 changes at each collision.
Figure 12. The collision volume is unaffected by collisions.
This page titled 4.12: The Frequency of Collisions between Unlike Gas Molecules is shared under a CC BY-SA 4.0 license and was authored,
We define the collision frequencycollision frequency, ν̃ , as the number of collision per unit time between a single molecule of
12
type 1 and any of the molecules of type 2 present in the same container. We find ν̃ = N π σ ⟨v ⟩ . If there are N molecules of
12 2
2
12 12 1
type 1 present in a unit volume of the gas, the total number of collisions between type 1 molecules and type 2 molecules is N 1
times greater. For clarity, let us refer to the total number of such collisions, per unit volume and per unit time, as the collision rate,
ρ12
. We have
1/2
8πkT
2 2
ρ12 = N1 ν̃ 12 = N1 N2 π σ ⟨v12 ⟩ = N1 N2 σ ( )
12 12
μ
This page titled 4.13: The Rate of Collisions between Unlike Gas Molecules is shared under a CC BY-SA 4.0 license and was authored, remixed,
When we consider collisions between different gas molecules of the same substance, we can denote the relative velocity and the
expected value of the relative velocity as v and ⟨v ⟩, respectively. By the argument we make above, we can find the number of
11 11
collisions between any one of these molecules and all of the others. Letting this collision frequency be ν̃ , we find 11
2
ν̃ 11 = N1 π σ ⟨v11 ⟩ ,
11
where σ 11 = 2 σ1 . Since we have

–
⟨v11 ⟩ = √2 ⟨v1 ⟩ ,
while
−−−−−−−−
⟨v1 ⟩ = √8kT /πm1,
−−−−−−−
we have ⟨v 11 ⟩ = 4 √kT /πm1 . The frequency of collisions between molecules of the same substance becomes
1/2
πkT
2 2
ν̃ 11 = N1 π σ ⟨v11 ⟩ = 4 N1 σ ( )
11 11
m1
The mean time between collisions, τ , is 11
τ11 = 1/ ν̃ 11
and the mean free path, λ , 11
1
2
λ11 = ⟨v1 ⟩ τ11 = – πσ
11
√2N 1
When we consider the rate of collisions between all of the molecules of type 1 in a container, ρ , there is a minor complication. If
11
we multiply the collision frequency per molecule, ν̃ , by the number of molecules available to undergo such collisions, N , we
11 1
count each collision twice, because each such collision involves two type 1 molecules. To find the collision rate among like
molecules, we must divide this product by 2. That is,
1/2
N1 ν̃ 11 πkT
2 2
ρ = = 2N σ ( )
11 1 11
2 m1
This page titled 4.14: Collisions between like Gas Molecules is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon
request.
Thus far we have not concerned ourselves with the relative orientation of a pair of colliding molecules. We want to develop a more
detailed model for the collision process itself, and the first step is to specify what we mean by relative orientation.
1
As before, we consider a molecule of type 1 moving with the relative velocity v through a gas of stationary type 2 molecules. In
12
unit time, molecule 1 travels a distance v and collides with many molecules of type 2. We can characterize each such collision by
12
the angle, θ , between the velocity vector and the line of centers of the colliding pair. For glancing collisions, we have θ = π/2 . For
head-on collisions, we have θ = 0 . All else being equal, the collision will be more violent the smaller the angle θ . Evidently, we
can describe the average effect of collisions more completely if we can specify the frequency of collisions as a function of θ . More
precisely, we want to find the frequency of collisions in which this angle lies between θ and θ + dθ .
When a collision occurs, the distance between the molecular centers is σ . We can say that the center of molecule 2 is at a
12
particular point on the surface of a sphere, of radius σ , circumscribed about molecule 1. As sketched in Figure 13, we can rotate
12
the line of centers around the velocity vector, while keeping the angle between them constant at θ . As we do so, the line of centers
traces out a circle on the surface of the sphere; collisions that put the center of molecule 2 at any two points on this circle are
completely equivalent. Letting the radius of this circle be r, we see that r = σ sinθ . Evidently, for spherical molecules, specifying
12
θ specifies the relative orientation at the time of collision.
If we now allow θ to vary by dθ , the locus of equivalent points on the circumscribed sphere expands to a band. Measured along the
surface of the sphere, the width of this band is σ dθ . As molecule 1 moves through the gas of stationary type 2 molecules, this
12
band sweeps out a cylindrical shell. Molecule 1 collides, at an angle between θ and θ + dθ , with every type 2 molecule in this
cylindrical shell. Conversely, every type 2 molecule in this cylindrical shell collides with molecule 1 at an angle between θ and
θ + dθ . (Molecule 1 also collides with many other type 2 molecules, but those collisions are at other angles; they have different
orientations.) In unit time, the length of the cylindrical shell is v . The volume of the cylindrical shell is its length times its cross-
12
sectional area.
Figure 13. The geometry of collisions between spheres.

The cross-section of the cylindrical shell is a circular annulus. Viewing the annulus as a rectangular strip whose length is the
circumference of the shell and whose width is the radial thickness of the annulus, the area of the annulus is the circumference times
the radial thickness. Since the radius of the shell is r = σ sinθ , its circumference is 2π σ sinθ. The radial thickness of the
12 12
annulus is just the change in the distance, r = σ sinθ , between the velocity vector and the wall of the cylinder when θ changes by
12
a small amount dθ . This is

dr
dr = ( ) dθ = σ12 cosθdθ
dθ
Therefore, the area of the annulus is
2
2π σ sinθcosθdθ
12
and the volume of the cylindrical shell swept out by a type 1 molecule (traveling at exactly the speed v ) in unit time is 12
2
2π σ v12 sinθcosθdθ
12
We again let N be the number of molecules of type 2 per unit volume. The number of collisions, per unit time, between a
2
molecule of type 1, traveling at exactly v , and molecules of type 2, in which the collision angle lies between θ and θ + dθ is
12
2
2π N2 σ v12 sinθcosθdθ
12
We need to find the number of such collisions in which the relative velocity lies between v and v + dv . The probability of 12 12 12
finding v in this interval is (df (v ) /dv ) dv . Let dν̃ (ν , θ) be the number of collisions made in unit time, by a type 1
12 12 12 12 12 12
molecule, with molecules of type 2, in which the collision angle is between θ and θ + dθ , and the scalar relative velocity is
between v and v + dv . This is just the number of collisions when the relative velocity is v multiplied by the probability that
12 12 12 12
the relative velocity is between v and v + dv . We have the result we need:

12 12 12
df (v12 )
2
dν̃ 12 (v12 , θ) = 2π N2 σ v12 ( ) sinθ cosθ dθdv12
12
dv12
2
3/2 −μv
2
μ 12
2 3
= 8 π N2 σ ( ) v exp ( ) × sinθ cosθ dθdv12
12 12
2πkT 2kT
Recognizing that possible values of θ lie in the range 0 ≤ θ < π/2 and that possible values of v lie in the range 12 0 ≤ v12 < ∞ ,
we can find the frequency of all possible collisions, ν̃ , by summing over all possible values of θ and v . That is,
12 12
∞ 2 π/2
μ 3/2 −μv
2 2 3 12
ν̃ 12 = 8 π N2 σ ( ) ∫ v exp ( ) dv12 × ∫ sinθ cosθ dθ
12 12
2πkT 0
2kT 0
2
μ 3/2 kT 1
2 2
= 8 π N2 σ ( ) [2 ( ) ][ ]
12
2πkT μ 2
1/2
8πkT
2
= N2 σ ( )
12
μ
In Section 4.12, we obtained this result by a slightly different argument, in which we did not explicitly consider the collision angle,
θ.
This page titled 4.15: The Geometry of A Collision between Spherical Molecules is shared under a CC BY-SA 4.0 license and was authored,
It is useful to extend our model of molecular collisions to suppose that one or both of the molecules can undergo chemical change
as a result of the collision. In doing so, we are introducing some ideas that we develop further in Chapter 5.
When we ask about the factors that determine whether such a reaction can occur, there can be several possibilities. We want to
focus on one such factor—the violence of the collision. We expect that a collision is more likely to result in a reaction the harder
the two molecules hit one another. When we try to formulate our basis for this expectation, we see that the underlying idea is that a
collision deforms the colliding molecules. The more violent the collision, the greater the deformation, and the greater the likelihood
of reaction becomes.
To proceed, we need to be more precise about what we mean by the violence of the collision. Evidently, what we have in mind has
two components: the relative velocity and the collision angle. If the collision is a glancing one, θ = π/2 , we expect the effect on
the molecules to be minimal, even if the relative velocity is high. On the other hand, a direct collision, θ ≈ 0 , might lead to reaction
even if the relative velocity is comparatively low. With these ideas in mind, we see that a reasonable model is to suppose that forces
acting along the line of centers can lead to reaction, whereas forces acting perpendicular to the line of centers cannot. If the
colliding molecules have complex shapes, this may be a poor assumption.
We also need a way to specify how much deformation occurs in a collision. If we want to specify the deformation by describing
specific changes in the molecular structures, this is a complex problem. For a general model, however, we can avoid this level of
detail. To do so, we recognize that any deformation can proceed only until the work done in deforming the molecules equals the
energy that can be expended to do this work. As the molecules are deformed, their potential energies change. The maximum change
in this potential energy is just the amount of kinetic energy that the colliding molecules can use to effect this deformation. We can
identify this amount of kinetic energy with the component of the molecules’ kinetic energy that is associated with their relative
motion along the line of centers.
If we now associate a threshold level of deformation with the occurrence of a chemical change, the kinetic energy required to effect
this deformation determines whether the change can occur. If the available kinetic energy is less than that required to achieve the
threshold level of deformation, reaction cannot occur. If the available kinetic energy exceeds this minimum, reaction takes place.
We call the minimum kinetic energy the activation energy and usually represent it by the symbol ϵ . (In discussing reaction rates, a
we usually express the activation energy per mole and represent it as E , where a
¯
¯¯¯
¯
Ea = N ϵa . )
We can apply these ideas to our model for collision between spherical molecules. In Section 4.10, we develop relative velocity
coordinates. It follows that we can partition kinetic energy of the two-particle system into a component that depends on the velocity
of the center of mass and a component that depends on the relative velocity. That is, we have
2 2
m1 v m2 v
1 2
KE = +
2 2
m1 m2
2 2 2 2 2 2
= (ẋ + ẏ + z ) + μ (ẋ + ẏ + ż )
0 0 0 12 12 12
2μ
2 2
m1 m2 v μv
0 12
= +
2μ 2
Only the component that depends on the relative velocity can contribute to the deformation of the colliding molecules. The relative
velocity can be resolved into components parallel and perpendicular to the line of centers. The parallel component is the projection
of the velocity vector onto the line of centers. This is v cosθ, and the perpendicular component is v sinθ. We see that the kinetic
12 12
energy associated with the relative motion of particles 1 and 2 has a component
2 2
μv cos θ
12
parallel to the line of centers and a component

2 2
μv sin θ
12
perpendicular to it.
The idea that the kinetic energy parallel to the line of centers must exceed εa for reaction to occur can now be expressed as the
requirement that
2 2
μv cos θ
12
ϵa <
2
When we consider all possible collisions between molecules 1 and 2, the collision angle varies from 0 to π/2. However, only those
collisions for which v satisfies the inequality above will have sufficient kinetic energy along the line of centers for reaction to
12
occur. The smallest value of v that can satisfy this inequality occurs when θ = 0 . This minimum relative velocity is
12
minimun 1/2
v = (2 ϵa /μ)
12
For relative velocities in excess of this minimum, collisions are effective only when
1/2
2
cosθ > (2 ϵa /μv )
12
so that
−1 1/2
2
θ < cos (2 ϵa /μv )
12
Let us designate the frequency of collisions satisfying these constraints as ν̃ 12 (ϵa ) . Recalling that
2
μ 3/2 −μv
2 2 3 12
dν̃ 12 (ν12 , θ) = 8 π N2 σ ( ) v exp ( ) × sinθ cosθ dθdv12
12 12
2πkT 2kT
we see that
1/2
−1 2
∞ cos (2 ϵa /μ v ) 2
μ 3/2 12 −μv
2 2 3 12
ν̃ 12 (εa ) = 8 π N2 σ ( ) ×∫ ∫ v exp ( ) × sinθ cosθ dθdv12
12 12
2πkT v12 =(2 ϵa /μ)
1/2
θ=0 2kT
The integral involving θ is

1/2
1/2 −1 2
cos
−1 2
(2 ϵa /μ v ) 2 cos (2 ϵa /μ v )
12
12
sin θ 1 2ϵa
∫ sinθ cosθ dθ = [ ] = [1 − ]
2
θ=0
2 2 μv
0 12
1/2
where, to evaluate the integral at its upper limit, we note that the angle θ = cos
−1
(2 ϵa /μv
2
12
) lies in a triangle whose sides
have lengths as indicated in Figure 14.
Figure 14. Maximum angle for an effective collision.

The collision frequency becomes
∞ 2
μ 3 2ϵa −μv
2 2 3 12
ν̃ 12 (ϵa ) = 4 π N2 σ ( ) /2 × ∫ [1 − ]v exp ( ) dv12
12 2 12
2πkT v12 =(2 ϵa /μ)
1/2
μv 2kT
12
1/2
This integral can be evaluated by making the substitution v 12 = (2ϵ/μ) . The lower limit of integration becomes ϵ ; we have a
∞ 2
2ϵa −μv
3 12
∫ [1 − ]v exp ( ) dv12
2 12
v12 =(2 ϵa /μ)
1/2
μv 2kT
12
∞
2 −ϵ
= ∫ (ϵ − ϵa ) exp ( ) dϵ
2
μ ϵa
kT
2
kT −ϵa
= 2( ) exp ( )
μ kT
Then
2
μ 3/2 kT −ϵa
2 2
ν̃ 12 (εa ) = 4 π N2 σ ( ) × 2( ) exp ( )
12
2πkT μ kT
1/2
8πkT −ϵa
2
= N2 σ ( ) exp ( )
12
μ kT
Note that when ϵ = 0 , this reduces to the same expression for ν̃ that we have obtained twice previously. The frequency of
a 12
collisions having kinetic energy along the line of centers in excess of ϵ depends exponentially on −ϵ /kT . All else being equal,
a a
this frequency increases as the temperature increases; it decreases as the activation energy increases.
This page titled 4.16: The Energy of A Collision between Gas Molecules is shared under a CC BY-SA 4.0 license and was authored, remixed,
4.17: Problems
1. For an oxygen molecule at 25 C, calculate (a) the most probable velocity, (b) the average velocity, (c) the root-mean-square
velocity.
2. For a gas of oxygen molecules at 25 C and 1.00 bar, calculate (a) the collision frequency, (b) the mean time between collisions,
(c) the mean free path. The diameter of an oxygen molecule, as estimated from gas-viscosity measurements, is 3.55 x 10 m. −10
3. For oxygen molecules at 25 C, calculate (a) the fraction with speeds between 150 and 151 m s −1
, (b) the fraction with speeds
between 400 and 401 m s , (c) the fraction with speeds between 550 and 551 m s .
−1 −1
4. For a hydrogen molecule at 100 C, calculate (a) the most probable velocity, (b) the average velocity, (c) the root-mean-square
velocity.
5. For a gas of hydrogen molecules at 100 C and 1.00 bar, calculate (a) the collision frequency, (b) the mean time between
collisions, (c) the mean free path. The diameter of a hydrogen molecule, as estimated from gas-viscosity measurements, is 2.71 x
10−10
m.
6. For a uranium hexafluoride (UF ) molecule at 100 C, calculate (a) the most probable velocity, (b) the average velocity, (c) the
6
root-mean-square velocity.
7. For a gas of uranium hexafluoride molecules at 100 C and 1.00 bar, calculate (a) the collision frequency, (b) the mean time
between collisions, (c) the mean free path. Assume that the diameter of a uranium hexafluoride molecule is 7.0 × 10 m.
−10
8. What is the average kinetic energy of hydrogen molecules at 100 C? What is the average kinetic energy of uranium hexafluoride
(U F ) molecules at 100 C?
6
9. Assuming the temperature in interstellar space is 2.73 K, calculate, for a hydrogen atom, (a) the most probable velocity, (b) the
average velocity, (c) the root-mean-square velocity.
10. Assuming that interstellar space is occupied entirely by hydrogen atoms at a particle density of 10 molecules m , calculate 2 −3
(a) the collision frequency, (b) the mean number of years between collisions, (c) the mean free path. Assume that the diameter of a
hydrogen atom is 2.40 × 10 m.
−10
11. Ignoring any effects attributable to its charge and assuming that the temperature is 2.73 K, calculate, for an electron in
interstellar space, (a) the most probable velocity, (b) the average velocity, (c) the root-mean-square velocity.
12. If a wall of a gas-filled container contains a hole, gas molecules escape through the hole. If all of the molecules that hit the hole
escape, but the hole is so small that the number escaping has no effect on the velocity distribution of the remaining gas molecules,
we call the escaping process effusion. That is, we call the process effusion only if it satisfies three rather stringent criteria. First, the
hole must be large enough (and the wall must be thin enough) so that most molecules passing through the hole do not hit the sides
of the hole. Second, a molecule that passes through the hole must not collide with anything on the other side that can send it back
through the hole into the original container. Third, the hole must be small enough so that the escaping molecules do not create a
pressure gradient; the rate at which gas molecules hit the hole and escape must be determined entirely by the equilibrium
distribution of gas velocities and, of course, the area of the hole. Show that the number of molecules effusing through a hole of area
A in time t is
1/2
N kT
At ( )( )
V 2πm
where (N /V ) is the number density of molecules in the container, and m is their molecular mass.
13. A vessel contains hydrogen and oxygen at 350 K and partial pressures of 0.50 bar and 1.50 bar, respectively. These gases
effuse into a vacuum. What is the ratio of hydrogen to oxygen in the escaping gas?
14. How could we use effusion to estimate the molecular weight of an unknown substance?
15. An equimolar mixture of 235
U F6 and 238
U F6 is subjected to effusion. What is the ratio of 235
U to 238
U in the escaping gas?
16. Calculate the number of nitrogen molecules that collide with 10
−6
m
2
of wall in −6
10 s , if the pressure is 1.00 bar and the
temperature is 300 K.
17. Air is approximately 20% oxygen and 80% nitrogen by volume. Assume that oxygen and nitrogen molecules both have a
radius of 1.8 × 10 m. For air at 1.0 bar and 298 K, calculate:
−8
(a) The number of collisions that one oxygen molecule makes with nitrogen molecules every second.
(b) The number of collisions that occur between oxygen and nitrogen molecules in one cubic meter of air every second.
(c) The number of collisions that one oxygen molecule makes with other oxygen molecules every second.
(d) The number of collisions that occur between oxygen molecules in one cubic meter of air every second.
(e) The number of collisions that occur between oxygen and nitrogen molecules in one cubic meter each second in which the
kinetic energy along the line of centers exceeds 100 kJ mol or 1.66 × 10 J per collision.
−1 −19
(f) The number of oxygen-nitrogen collisions that occur in which the kinetic energy along the line of centers exceeds 50 kJ mol . −1
∞
18. Show that ∫ 0
ρ
v
(v) dv ≠ 1.
19. For what volume element, ʋ, is
P (ʋ ) = fxyz (vx , vy , vz )?
20. Using the model we develop in Section 2.10:

(a) Show that the pressure, P 1 (v) , attributable to a single molecule of mass m and velocity v in a container of volume V is
2
mv
P1 (v) =
3V
(b) In Section 4.6, we find that this pressure is

2 2
2m v cos θ
δP1 (v) =
V
for a molecule whose velocity vector lies between θ and θ + dθ and between φ and φ + dφ . This angular region comprises a solid
angle whose magnitude is dΩ = sinθdθdφ . Since the solid angle surrounding a given point is 4π, the probability that a randomly
oriented velocity vector lies between θ and θ + dθ and between φ and φ + dφ is
dΩ sinθdθdφ
=
4π 4π
Therefore, given that the scalar component of a molecule’s velocity is v , its contribution to the pressure at A is
2
mv
2
dP1 (v) = ( ) cos θ sinθ dθdφ
2πV
To find the pressure contribution made by this molecule irrespective of the values of θ and φ , we must integrate dP (v) over all 1
values of θ and φ that allow the molecule to impact the wall at A . Recalling that these ranges are 0 ≤ θ < π/2 and 0 ≤ φ < 2π ,
show that
2
mv
P1 (v) =
3V
21. Taking P 1
2
(v) = m v /3V as the contribution made to the pressure by one molecule whose velocity is v :
(a) Show that the expected value for the contribution made to the pressure by one molecule when the Maxwell–Boltzmann
distribution function describes the distribution of molecular velocities is
kT
⟨P1 (v)⟩ =
V
(b) Show that the variance of the contribution made to the pressure by one molecule is
2 2
2k T
2
σ =
P1 (v) 2
3V
What is the standard deviation, σ P1 (v)

?
(c) What is the value of the ratio
σP
1 (v)
⟨P1 (v)⟩
(d) Taking 3 × 10 as the number of collisions of N molecules at 1 bar and 300 K with one one square millimeter per
15
2
microsecond, what pressure, P , would we find if we could measure the individual contribution made by each collision and
avg
compute their average? What would be the variance, σ , of this average? The standard deviation, σ ? The ratio σ /P ?
2
avg avg avg avg
22. Let ϵ = mv 2
/2 be the translational kinetic energy of a gas molecule whose mass is m . Show that the probability density
function for ϵ is
3/2
df 1 −ϵ
1/2
= 2π ( ) ϵ exp ( )
dϵ πkT kT
¯
¯¯¯
¯
Letting the translational kinetic energy per mole be E = N ϵ , show that
3/2
df 1 −E
1/2
= 2π ( ) E exp ( )
dE πRT RT
Notes
1
Our collision model and quantitative treatment of the role of activation energy in chemical reaction rates follow those given by
Arthur A. Frost and Ralph G. Pearson, Kinetics and Mechanism, 2 Ed., John Wiley and Sons, New York, 1961, pp 65-68. See
nd
also R. H. Fowler, Statistical Mechanics, Cambridge University Press, New York, 1936.
CHAPTER OVERVIEW
5: Chemical Kinetics, Reaction Mechanisms, and Chemical Equilibrium

Chemical kinetics is the study of how fast chemical reactions occur and of the factors that affect these rates. The study of reaction
rates is closely related to the study of reaction mechanisms, where a reaction mechanism is a theory that explains how a reaction
occurs.
5.20: Problems
This page titled 5: Chemical Kinetics, Reaction Mechanisms, and Chemical Equilibrium is shared under a CC BY-SA 4.0 license and was
1
A reaction mechanism describes the sequence of bond-making, bond-breaking, and intramolecular-rearrangement steps that results
in the overall chemical change. These individual steps are called elementary reactions or elementary processes. In an elementary
reaction, no intermediate species is formed; that is, none of the arrangements of atoms that occur during the elementary reaction has
a lifetime greater than the duration of a molecular vibration, which is typically from 10 to 10
−12
seconds. We expand on this
−14
point in Section 5.6.

We can distinguish two levels of detail in a chemical reaction mechanism: The first is the series of elementary processes that occurs
for a given net reaction. This is called the stoichiometric mechanism. Frequently it is also possible to infer the relative positions of
all of the atoms during the course of a reaction. This sort of model is called an intimate mechanism or detailed mechanism.
A rate law is an equation that describes how the observed reaction rate depends on the concentrations of the species involved in the
reaction. This concentration dependence can be determined experimentally. We will see that any series of elementary reactions
predicts the dependence of reaction rates on concentrations, so one of the first tests of a proposed mechanism is that it be consistent
with the rate law that is observed experimentally. (If the overall reaction proceeds in more than one step and the concentration of an
intermediate species becomes significant, we may need more than one equation to adequately describe the rates of all of the
reactions that occur.)
The rate law plays a central role in our study of reaction rates and mechanisms. We infer the rate law from experimental
measurements. We must be able to prove that the experimental rate law is consistent with any mechanism that we propose. The rate
law that we deduce from experimental rate data constitutes an experimental fact. Our hypothesized mechanism is a theory. We can
entertain the idea that the theory may be valid only so long as its predictions about the rate law are consistent with the experimental
result. We can predict rate laws for elementary processes by rather simple arguments. For a mechanism involving a series of
elementary processes, we can often predict rate laws by making simplifying assumptions. When simplifying assumptions are
inadequate, we can use numerical integration to test agreement between the proposed mechanism and experimental observations of
the dependence of the reaction rate on the concentrations of the species involved in the reaction. We will see that a given
experimental rate law may be consistent with any of several mechanisms. In such cases, we must develop additional information in
order to discriminate among the several mechanisms.
This page titled 5.1: Chemical Kinetics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
Chemical reactions occur under a wide variety of circumstances: Many chemicals are manufactured by passing a homogeneous
mixture of gaseous reactants through a bed of solid catalyst pellets. Corrosion of metals is a reaction between the solid metal and
oxygen from the air, often catalyzed by other common chemical species like water or chloride ion. An enormous number of
biological reactions occur within the cells of living organisms. In the laboratory, we typically initiate a reaction by mixing the
reactants, often with a solvent, in a temperature-controlled vessel.
Chemical reactions can be carried out in batch reactors and in a wide variety of flow reactors. A batch reactor is simply a
container, in which we initiate the reaction by mixing the reactants with one another (and any additional ingredients), and in which
the reaction occurs and the products remain—until we get around to removing them. A reaction carried out under these conditions
is called a batch reaction. If any of the reactants are gases, a batch reactor must be sealed to prevent their escape. Otherwise, we
may leave the reactor open to the atmosphere as the reaction occurs.
Flow reactor have been designed to achieve a variety of objectives. Nevertheless, they have a number of characteristics in common.
A flow reactor is a container, into which reactants and other ingredients are injected. The products are recovered by withdrawing
portions of the reaction mixture from one or more locations within the reactor. The rates at which materials are injected or
withdrawn are usually constant. In the simplest case, the reactants are mixed at one end of a long tube. The reacting mixture flows
through the tube. If the tube is long enough, the mixture emerging from the other end contains the equilibrium concentrations of
reactants and products. In such a tubular reactor, it is usually a good approximation to assume that the material injected during one
short time interval does not mix with the material injected during the next time interval as they pass through the tube. We view the
contents of the reactor as a series of fluid “plugs” that traverse the reactor independently of one another and call this behavior plug
flow.
In Section 5.11, we discuss another simple flow stirred tank, called a continuous stirred-tank reactor (CSTR) or a capacity-flow
reactor. A CSTR consists of a single constant-volume vessel into which reactants are continuously injected and from which
reaction mixture is continuously withdrawn. The contents of this container are constantly stirred. In our discussion, we assume that
the reactor is completely filled with a homogeneous liquid solution. We express the rate of reaction within the CSTR in moles per
liter of reactor volume per second, mol L s .−1 −1
When we talk about the rate of a particular reaction, we intend to specify the amount of chemical change that occurs in unit time
because of that reaction. It is usually advantageous to specify the amount of chemical change in units of moles. We can specify the
amount of chemical change by specifying the number of moles of a reactant that are consumed, or the number of moles of a
product that are produced, per second, by that reaction. If we do so, the amount of chemical change depends on the stoichiometric
coefficient of the reactant or product that we choose. Moreover, the rate is proportional to the size of the system. Since the
properties of reaction rates that are of interest to us are usually independent of the size of the system, we find it convenient to
express reaction rates as moles per second per unit system size, so that the most convenient units are usually concentration per
second.
For reactors containing heterogeneous catalysts, we typically express the reaction rate in moles per unit volume of catalyst bed per
second. For corrosion of a metal surface, we often express the rate in moles per unit area per second. For biological reactions, we
might express the reaction rate in moles per gram of biological tissue per second. For reactions in liquid solutions, we typically
express the rate in moles per liter of reaction mixture per second, mol L s or M s .
−1 −1 −1
–––
Evidently, we need to express the rate of a reaction in a way that accounts for the stoichiometry of the reaction and is independent
of the size of the system. Moreover, we must distinguish the effect of the reaction on the number of moles of a reagent present from
the effects of other processes, because competing reactions and mechanical processes can affect the amount of a substance that is
present.
To develop the basic idea underlying our definition of reaction rate, let us consider a chemical substance, A , that undergoes a single
reaction in a closed system whose volume is V . For a gas-phase reaction, this volume can vary with time, so that V = V (t) . (The
volume of any open system can vary with time.) Since the system is closed, the reaction is the is the only process that can change
the amount of A that is present. Let Δn be the increase in the number of moles of A in a short interval, Δt, that includes time t .
A
Let the average rate at which the number of moles of A increases in this interval be r̄ (A) = Δn /Δt . The corresponding
¯
A
instantaneous rate is
ΔnA
r (A) = lim ( )
Δt→0 Δt
dnA
=
dt
To express this information per unit volume, we can define the instantaneous rate of reaction of A , at time t , as
1 ΔnA
R (A) = lim ( )
V (t) Δt→0 Δt
1 dnA
=
V (t) dt
Experimental studies of reaction rate are typically done in constant-volume closed systems under conditions in which only one
reaction occurs. Most of the discussion in this chapter is directed toward reactions in which these conditions are satisfied, and the
system comprises a single homogeneous phase. In the typical case, we mix reactants with a liquid solvent in a reactor and immerse
the reactor in a constant-temperature bath. Under these conditions, the rate at which a particular substance reacts is equal to the rate
at which its concentration changes. Writing [A] to designate the molarity, mol L , of A , we have −1
1 dnA
R (A) =
V dt
d (nA /V )
=
dt
d [A]
=
dt
If we express a reaction rate as a rate of concentration change, it is essential that both conditions be satisfied. If both n and V vary A
with time, we have

d [A] d (nA /V )
=
dt dt
1 dnA nA dV
= −
2
V dt V dt
The instantaneous rate at which substance A undergoes a particular reaction is equal to V (dn /dt) only if the reaction is the
−1
A
sole process that changes n ; the contribution to d [A]/dt made by (n V ) dV /dt vanishes only if the volume is constant.
A A
−2
If a single reaction is the only process that occurs in a particular system, the rate at which the number of moles of any reactant or
product changes is a measure of the rate of the reaction. However, these rates depend on the stoichiometric coefficients and the size
of the system. For a reaction of specified stoichiometry, we can use the extent of reaction, ξ to define a unique reaction rate, R .
The amounts of reactants and products present at any time are fixed by the initial conditions and the stoichiometry of the reaction.
Let us write n to denote the number of moles of reagent A present at an arbitrary time and n to denote the number of moles of
A
o
A present at the time (t = 0 ) that the reaction is initiated. We define the extent of reaction as the change in the number of moles of
a product divided by the product’s stoichiometric coefficient or as the change in the number of moles of a reactant divided by the
negative of the reactant’s stoichiometric coefficient. For the stoichiometry
aA + bB + ⋯ → cC + dD + …
we have
o o
nA − n nB − n
A B
ξ = = =…
−a −b
o o
nC − n nD − n
C D
= = =…
c d
If Ais the limiting reagentlimiting reagent, ξ varies from zero, when the reaction is initiated with nA = n
o
A
, to n
o
A
/a , when
nA = 0 . At any time, t , we have
dξ 1 dnA 1 dnB
=− =− =…
dt a dt b dt
1 dnC 1 dnD
= = =…
c dt d dt
and we can define a unique reaction rate as

1 dξ
R =
V (t) dt
1 dnA
=−
aV (t) dt
1 dnB
=− =…
bV (t) dt
1 dnC
=
cV (t) dt
1 dnD
= =…
dV (t) dt
The relationship between the instantaneous rate at which reactant A undergoes this reaction, R (A) , and the reaction rate, R , is
1 dnA a dξ
R (A) = =− = −aR
V (t) dt V (t) dt
If the volume is constant, we have
ξ [A] − [A] [B] − [B]

0 0
= = =…
V −a −b
[C ] − [C ] [D] − [D]
0 0
= = =…
c d
and the reaction rate is

1 dξ 1 d [A] 1 d [B]
R = =− =− =…
V dt a dt b dt
1 d [C ] 1 d [D]
= = =…
c dt d dt
The name “extent of reaction” is sometimes given to the fraction of the stoichiometrically possible reaction that has occurred. To
distinguish this meaning, we call it the fractional conversion, χ. When A is the stoichiometrically limiting reactant, the fractional
conversion is
o
nA − n
A
χ =−
o
n
A
aξ
The extent of reaction, ξ , and the fractional conversion, χ, are related asχ = n
o
A
We have 0 ≤ ξ ≤ n o
A
/a and 0 ≤ χ ≤ 1 .
The rate of a reaction usually depends on the concentrations of some or all of the substances involved. The dependence of reaction
rate on concentrations is the rate law. It must be determined by experiment. For reaction
aA + bB + ⋯ → cC + dD + …
the observed rate law is often of the form

1 dξ
R =
V dt
m n p q
= k[A] [B] … [C ] [D]
where m, n , …, p, q, … are small positive or negative integers or (less often) simple fractions.
We use a conventional terminology to characterize rate laws like this one. We talk about the order in a chemical species and the
order of the reaction. To characterize the rate law above, we say that the reaction is “m order in compound A ,” “n order in
th th
compound B ,” “p th
order in compound C ,” and “q order in compound D.” We also say that the reaction is “
th
(m + n + ⋯ + p + q + …) order overall.” Here k is an experimentally determined parameter that we call the rate constant or
th
rate coefficient.
It frequently happens that we are interested in an overall chemical change whose stoichiometric mechanism involves two or more
elementary reactions. In this case, an exact rate model includes a differential equation for each elementary reaction. Nevertheless, it
is often possible to approximate the rate of the overall chemical change by a single differential equation, which may be a relatively
complex function of the concentrations of the species present in the reaction mixture. For the reaction above, the experimental
observation might be
1 dξ
R =
V dt
k1 [A] [B]
=
k2 [C ] + k3 [D]
In such cases, we continue to call the differential equation the rate law. The concept of an overall order is no longer defined. The
constants (k , k , and k ) may or may not be rate constants for elementary reactions that are part of the overall process.
1 2 3
Nevertheless, it is common to call any empirical constant that appears in a rate law a rate constant. In a complex rate law, the
constants can often be presented in more than one way. In the example above, we can divide numerator and denominator by, say,
k , to obtain a representation in which the constant coefficients have different values, one of which is unity.
3
Most of the rest of this chapter is devoted to understanding the relationship between an observed overall reaction rate and the rates
of the elementary processes that contribute to it. Our principal objective is to understand chemical equilibrium rates at in terms of
competing forward and reverse reactions. At equilibrium, chemical reactions may be occurring rapidly; however, no concentration
changes can be observed because each reagent is produced by one set of reactions at the same rate as it is consumed by another set.
For the most part, we focus on reactions that occur in closed constant-volume systems.
This page titled 5.2: Reaction Rates and Rate Laws is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
The number of moles of a substance in a system can change with time because several processes occur simultaneously. Not only
can a given substance participate in more than one reaction, but also the amount of it that is present can be affected by processes
that are not chemical reactions. A variety of transport process can operate to increase or decrease the amount of the substance that
is present in the reaction mixture: A pure solid reactant could dissolve in a reacting solution, or a product could precipitate from it,
as the reaction proceeds. A reacting species could diffuse into a reactor across a semi-permeable membrane. Controlled amounts of
a reacting species could be added, either continuously or at specified intervals.
Each of the simultaneous processes contributes to the change in the number of moles of A present. At every instant, each of these
contributions can be characterized by a rate. Over a short time interval, Δt, let Δ n be the contribution that the i process makes
i A
th
to the change in the amount of A in volume V . If, even though the i process may not be a reaction, we use R (A) to represent its
th
i
rate, its contribution to the rate at which the amount of A changes is

Δi nA 1 di nA
Ri (A) = lim ( ) =
Δt→0 V (t) Δt V (t) dt
If there are numerous such processes, whose rates are R 1 (A) ,R 2 (A) , ..., R i (A) ,…, R ω (A) , the observed overall rate is
ω ω
1 di nA
R (A) = ∑ Ri (A) = ∑
V (t) dt
i=1 i=1
If the volume is constant,

1 di nA di [A]
Ri (A) = =
V dt dt
and
ω ω
di [A] d [A]
R (A) = ∑ Ri (A) = ∑ =
dt dt
i=1 i=1
To illustrate these ideas, let us consider the base hydrolyses of methyl and ethyl iodide. No intermediates are observed in these
reactions. If we carry out the base hydrolysishydrolysis:ethyl iodide of methyl iodide,
− −
CH I + OH ⟶ CH OH + I
3 3
in a closed constant-volume system, we can express the reaction rate in several equivalent ways:
− −
1 dξ (C H3 I ) d [C H3 OH ] d [I ] d [C H3 I ] d [OH ]
R (C H3 I ) = = = =− =−
V dt dt dt dt dt
If a mixture of methyl and ethyl iodide is reacted with aqueous base, both hydrolysis reactions consume hydroxide ion and produce
iodide ion. The rates of these individual processes can be expressed as
1 dξ (C H3 I ) d [C H3 OH ] d [C H3 I ]
R (C H3 I ) = = =−
V dt dt dt
and
1 dξ (C H3 C H2 I ) d [C H3 C H2 OH ]
R (C H3 C H2 I ) = =
V dt dt
d [C H3 C H2 I ]
=−
dt
but the rates at which the concentrations of hydroxide ion and iodide ion change depend on the rates of both reactions. We have
− −
d [I ] d [OH ]
=−
dt dt
= R (C H3 I ) + R (C H3 C H2 I )
1 dξ (C H3 I ) 1 dξ (C H3 C H2 I )
=− −
V dt V dt
In principle, either of the reaction rates can be measured by finding the change, over a short time interval, in the number of moles
of a particular substance present.
Simultaneous processes occur when a reaction does not go to completion. The hydrolysis of ethyl acetate,
CH CO CH CH + H O ⇌ CH CO H + CH CH OH
3 2 2 3 2 3 2 3 2
can reach equilibrium before the limiting reactant is completely consumed. The reaction rate, defined as R = V (dξ/dt) , falls to −1
zero. However, ethyl acetate molecules continue to undergo hydrolysis; the extent of reaction becomes constant because ethyl
acetate molecules are produced from acetic acid and ethanol at the same rate as they are consumed by hydrolysis. Evidently, the
rate of the forward reaction does not fall to zero even though the net reaction rate does.
Let R represent the number of moles of ethyl acetate undergoing hydrolysis per unit time per unit volume. Let R represent the
f r
number of moles of ethyl acetate being produced per unit time per unit volume. The net rate of consumption of ethyl acetate is
R = R − R . At equilibrium, R = 0 , and R = R > 0 . In such cases, it can be ambiguous to refer to “the reaction” or “the rate
f r f r
of reaction.” The rates of the forward and of the net reaction are distinctly different things.
So long as no intermediate species accumulate to significant concentrations in the reaction mixture, we can find the forward and
reverse rates for a reaction like this, at any particular equilibrium composition, in a straightforward way. When we initiate reaction
with no acetic acid or ethanol present, the rate of the reverse reaction must be zero. We can find the rate law for the forward
reaction by studying the rate of the hydrolysis reaction when the product concentrations are low. Under these conditions, R = 0 r
and R ≈ R . From the rate law that we find and the equilibrium concentrations, we can calculate the rate of the forward reaction at
f
equilibrium. Likewise, when the ethyl acetate concentration is low, the rate of the hydrolysis reaction is negligible in comparison to
that of the esterification reaction. We have R ≈ 0 and R = −R , and we can find the rate law for the esterification reaction by
f r
studying the rate of the esterification reaction when the concentration of ethyl acetate is negligible. From this rate law, we can
calculate the rate of the reverse reaction at the equilibrium concentrations.
This page titled 5.3: Simultaneous Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
In practice, rate constants vary in response to changes in several factors. Indeed, they are usually the same in two experiments only
if we keep everything but the reagent concentrations the same. Another way of saying this is that the rate law captures the
dependence of reaction rate on concentrations, while the dependence of reaction rate on any other variable appears as a dependence
of rate constants on that variable.
Temperature usually has a big effect. The experimentally observed dependence of rate constants on temperature can be expressed in
a compact fashion. Over small temperature ranges it can usually be expressed adequately by the Arrhenius equation:
−Ea
k = A exp ( )
RT
where E and A are called the Arrhenius activation energy and the frequency factor (or pre-exponential factor), respectively.
a
The Arrhenius equation is an empirical relationship. As we see below for our collision-theory model, theoretical treatments predict
that the pre-exponential term, A, is weakly temperature dependent. When we investigate reaction rates experimentally, the
temperature dependence of A is usually obscured by the uncertainties in the measured rate constants. It is often said, as a rough rule
of thumb, that the rate of a chemical reaction doubles if the temperature increases by 10 K. However, this rule can fail
spectacularly. A reaction can even proceed more slowly at a higher temperature, and there are multi-step reactions for which this is
observed.
This page titled 5.4: The Effect of Temperature on Reaction Rates is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
A reaction that occurs in one solvent usually occurs also in a number of similar solvents. For example, a reaction that occurs in
water will often occur with a low molecular weight alcohol—or an alcohol-water mixture —as the solvent. Typically, the same rate
law is observed in a series of solvents, but the rate constants are solvent-dependent.
Other chemical species that are present in the reaction medium (but which are neither products nor reactants) can also affect
observed reaction rates. Any such species meets the usual definition of a catalyst. However, common practice restricts use of the
word “catalyst” to a chemical species that substantially increases the rate of the reaction. A chemical species that decreases the rate
of the reaction is usually called an inhibitor. If we think that the rate effect of the non-reacting species results from a non-specific
or a greater-than-bonding-distance interaction with one or more reacting species, we call the phenomenon a medium effect. A
solvent effect is a common kind of medium effect; altering the solvent affects the reaction rate even though the solvent does not
form a chemical bond to any of the reactants or products. Dissolved salts can affect reaction rates in a similar way. Such effects
often occur when the degree of charge separation along the path of an elementary reaction is significantly different from that in the
reactants.
Isotopic substitution in a reactant can affect the reaction rate. (Replacement of a hydrogen atom with a deuterium atom is the most
common case.) The effect of an isotopic substitution on a reaction rate is called a kinetic isotope effect. Kinetic isotope effects can
provide valuable information about the reaction mechanism. A kinetic isotope effect is expected if the energy needed make or
break a chemical bond to the isotopically substituted atom is a significant component of the activation energy for the reaction.
Kinetic isotope effects are usually small in comparison to other factors that affect reaction rates. A ten-fold change in the reaction
rate is a big kinetic isotope effect. Effects much smaller than this are often useful; indeed, the absence of a kinetic isotope effect can
help distinguish among alternative mechanisms.
In studies of reaction rates that are focused on finding the reaction mechanism, many characteristics of the reaction that are not
strictly rate-related can be important. These include the stereochemistry of the product; the Walden inversion that accompanies S
N 2 reactions at tetrahedral carbon centers is a notable example. Isotopic substitution that occurs incidental to a reaction can help
establish that an intermediate is formed. The effects of competing reactions are often significant. The study of competing reactions
is frequently helpful when the reaction involves a short-lived and otherwise undetectable intermediate. The use of isotopic
substitution and competing reactions is illustrated in Section 5.16, in which we review the base hydrolysis of cobalt (III)cobalt
pentaammine complexes, C o(N H ) X . 3 5
n+
This page titled 5.5: Other Factors that Affect Reaction Rates is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
To see what we mean by an elementary process, let us consider some possible mechanisms for the base hydrolysis of methyl
iodide:
− −
CH I + OH → CH OH + I .
3 3
In this reaction, a carbon–iodide bond is broken and a carbon–oxygen bond is formed. While any number of reaction sequences
sum to this overall equation, we can write down three that are reasonably simple and plausible. The C−I could be broken first and
the C−OH bond formed thereafter. Alternatively, the C−OH bond could be formed first and the C−I bond broken thereafter. In
the first case, we have an intermediate species, CH , of reduced coordination number, and in the second we have an intermediate,
+
3
HO−CH −I , of increased coordination number. Finally, we can suppose that the bond-forming and bond-breaking steps occur
−
simultaneously, so that no intermediate species is formed at all.

Heterolytic bond-breaking precedes bond-making
+ −
CH I → CH3 + I
3
(mechanism a)
+ −
CH3 + OH → CH OH
3
Bond-making precedes bond-breaking

− −
CH I + OH → HO⋅⋅⋅ CH ⋅⋅⋅ I (mechanism b)
3 3
− −
HO⋅⋅⋅ CH ⋅⋅⋅ I ⟶ CH OH + I
3 3
Bond-breaking and bond-making are simultaneous

− − ‡ −
CH I + OH ⟶ [HO⋅⋅⋅ CH ⋅⋅⋅ I ] ⟶ CH OH + I (mechanism c)
3 3 3
The distinction between mechanism (b) and mechanism (c) is that an intermediate is formed in the former but not in the latter.
Nevertheless, mechanism (c) clearly involves an intermediate structure in which both the incoming and the leaving group are
bonded to the central carbon atom. The distinction between mechanisms (b) and (c) depends on the nature of the intermediate
structure. In mechanism (b), we suppose that the intermediate is a bona fide chemical entity; once a molecule of it is formed, that
molecule has a finite lifetime. In (c), we suppose that the intermediate structure is transitory; it does not correspond to a molecule
with an independent existence.
For this distinction to be meaningful, we must have a criterion that establishes the shortest lifetime we are willing to associate with
“real molecules.” It might seem that any minimum lifetime we pick must be wholly arbitrary. Fortunately this is not the case; there
is a natural definition for a minimum molecular lifetime. The definition arises from the fact that molecules undergo vibrational
motions. If a collection of atoms retains a particular relative orientation for such a short time that it never undergoes a motion that
we would recognize as a vibration, it lacks an essential characteristic of a normal molecule. This means that the period of a high-
frequency molecular vibration (roughly 10 −14
s) is the shortest time that a collection of atoms can remain together and still have all
of the characteristics of a molecule. If a structure persists for more than a few vibrations, it is reasonable to call it a molecule, albeit
a possibly very unstable one.
− ‡
In mechanism (c) the structure designated [HO⋅⋅⋅CH ⋅⋅⋅I ] depicts a transitory arrangement of the constituent atoms. The atomic
3
arrangement does not persist long enough for the HO−CH bond or the CH −I bond to undergo vibrational motion. A structure
3 3
with these characteristics is called an activated complex or a transition state for the reaction, and a superscript double dagger, ‡ , is
conventionally used to signal that a structure has this character. The distinction between a bona fide intermediate and a transition
state is clear enough in principle, but it can be very difficult to establish experimentally.
These considerations justify our earlier definition: An elementary reaction is one in which there are no intermediates. Any atomic
arrangement that occurs during an elementary reaction does not persist long enough to vibrate before the arrangement goes on to
become products or reverts to reactants.
An elementary reaction is one in which there are no intermediates.

We can distinguish a small number of possible kinds of elementary reaction: termolecular elementary reactions, bimolecular
elementary reactions, and unimolecular reactions. A single molecule can spontaneously rearrange to a new structure or break into
smaller pieces. Two molecules can react to form one or more products. Three molecules can react to produce products. Or we can
imagine that some larger number of molecules reacts. We refer to these possibilities as unimolecular, bimolecular, termolecular,
and higher-molecularity processes.
The stoichiometry of many reactions is so complicated as to preclude the possibility that they could occur as a single elementary
process. For example, the reaction
2 + − + 3 + 3 +
3 Fe + HCrO +7 H ⟶ 3 Fe + Cr +4 H O
4 2
can not plausibly occur in a single collision of three ferrous ions, one chromate ion, and seven hydronium ions. It is just too
unlikely that all of these species could find themselves in the same place, at the same time, in the proper orientation, and with
sufficient energy to react. In such cases, the stoichiometric mechanism must be a series of elementary steps. For this reaction, a
skeletal representation of one plausible series is
Fe(II) + Cr(VI) → Fe(III) + Cr(V)
Fe(II) + Cr(V) → Fe(III) + Cr(IV)
Fe(II) + Cr(IV) → Fe(III) + Cr(III)
This page titled 5.6: Mechanisms and Elementary Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
Bimolecular Elementary Processes
If we think about an elementary bimolecular reaction rate law between molecules A and B , we recognize that the reaction can
occur only when the molecules come into contact. They must collide before they can react. So the probability that they react must
be proportional to the probability that they collide, and the number of molecules of product formed per unit time must be
proportional to the number of A − B collisions that occur in unit time. In our development of the collision theory for bimolecular
reactions in the gas phase, (§4-12 to §4-16), we find that the number of such collisions is proportional to the concentration of each
reactant. It is clear that this conclusion must apply to any bimolecular reaction.
If we have a vessel containing some concentration of A molecules and some concentration of B molecules, the collection
experiences some number of A − B collisions per unit time. If we double the concentration of B molecules, each A molecule is
twice as likely as before to encounter a B molecule. Indeed, for any increase in the concentration of B molecules, the number of
collisions of an A molecule with B molecules increases in the same proportion. The number of A − B collisions must be
proportional to the concentration of B molecules. Likewise, increasing the concentration of A molecules must increase the number
of A − B collisions proportionately; the number of A − B collisions must also be proportional to the concentration of A
molecules. We conclude that the rate for any bimolecular reaction between molecular substances A and B is described by the
equations
1 dnA 1 dnB
R = = = −k2 [A] [B]
V dt V dt
This is a second-order rate law, and the proportionality constant, k , is called a second-order rate constant.
2
In Section 5.4-5.16, we derive an equation for the frequency with which a type 1 molecule collides with type 2 molecules in the gas
phase when the concentration of type 2 molecules is N and the kinetic energy along the line of centers exceeds a threshold value,
2
ϵ per molecule, or E per mole. The rate at which such collisions occur is
a a
1/2
8πkT −ϵa
2
ρ12 (ϵa ) = N1 N2 σ ( ) exp ( )
12
μ kT
which is just exp (−ϵ /kT ) times the rate at which collisions of any energy occur between molecules of type 1 and molecules of
a
type 2. If reaction occurs at every collision between a molecule 1 and a molecule 2 in which the kinetic energy along the line of
centers exceeds ϵ , the collision rate, ρ (ϵ ) , equals the reaction rate. We have R = ρ (ϵ ) .
a 12 a 12 a
If the temperature-dependence of the rate constant is given by the Arrhenius equation, the rate of the bimolecular reaction between
species 1 and species 2 is
−Ea
R = k2 N1 N2 = [A exp ( )] N1 N2
RT
where A is independent of temperature and E = N ϵ . The collision-theory model for the bimolecular reaction is almost the same;
a
¯
¯¯¯
¯
a
the difference being a factor of T 1/2

in the pre-exponential factor, σ (8πkT /μ) . The effect of the T
2
12
1/2
term is usually small1/2
in comparison to the effect of temperature in the exponential term. Thus, the temperature dependence predicted by collision theory,
which is a highly simplified theoretical model, and that predicted by the Arrhenius equation, which is an empirical generalization
usually used to describe data taken over a limited temperature range, are in substantial agreement.
Experimentally determined values of the pre-exponential factor for gas-phase bimolecular reactions can approach the value
calculated from collision theory. However, particularly for reactions between polyatomic molecules, the experimental value is often
much smaller than the calculated collision frequency. We rationalize this observation by recognizing that our colliding-spheres
model provides no role for the effect of molecular structures. When the colliding molecules are not spherical, the collision angle is
an incomplete description of their relative orientation. If the relative orientation of two colliding molecules is unfavorable to
reaction, it is entirely plausible that they can fail to react no matter how energetic their collision. To recognize this effect, we
suppose that the reaction rate is proportional to a steric factor, γ, where γ represents the probability that a colliding pair of
molecules have the relative orientation that is necessary for reaction to occur. Of course, γ must be less than one. Taking this
amplification of the collision model into account, the relationship between the reaction rate and the collision frequency becomes
dξ
= γ ρ12 (εa )
dt
When we consider reactions in solution, we recognize that there are usually many more solvent molecules than reactant molecules.
As a result, collisions of a reactant molecule with solvent molecules are much more frequent than collisions of a molecule of one
reactant with a molecule of another reactant. The high frequency of collisions with solvent molecules means that the net distance
moved by a reactant molecule in unit time is much less in solution than in a gas. This decreases the probability that two reactant
molecules will meet. On the other hand, once two reactant molecules near one another, the solvent molecules tend to keep them
together, and they are likely to collide with one another many times before they finally drift apart. (This is known as the solvent-
cage effect.) We can expect these effects to roughly offset one another.
Termolecular Elementary Processes

A termolecular elementary process is a reaction in which three reactant molecules collide. For this to happen, an A molecule and a
B molecule must be very close to one another at exactly the time that a C molecule encounters the pair of them. If the reactants are
not very concentrated, the probability that a given A molecule is very close to a B molecule during any short time interval is small.
The probability that this A molecule will be hit by a C molecule during the same time interval is also very small. The probability
that all three species will collide at the same time is the product of two small probabilities; under any given set of conditions, the
number of collisions involving three molecules is smaller than the number of collisions between two molecules. The probability of
a termolecular collision and hence the rate of a termolecular elementary process is proportional to the concentrations of all three
reacting species
1 dnA 1 dnB 1 dnC
R = = = = −k2 [A] [B] [C ]
V dt V dt V dt
However, the low probability of a termolecular collision means that we can expect the termolecular rate constant, k , to be very 3
small. If termolecular mechanisms are rare, higher-molecularity mechanisms must be exceedingly rare, if, indeed, any occur at all.
For most chemical reactions, the mechanism is a series of unimolecular and bimolecular elementary reactions.
Unimolecular Elementary Processes

A unimolecular elementary process is one in which a molecule spontaneously undergoes a chemical change. If we suppose that
there is a constant probability that any given A molecule undergoes reaction in unit time, then the total number reacting in unit time
¯¯¯¯¯¯
¯
is proportional to the number of A molecules present. Let the average number of moles reacting in unit time be Δn , the number A
of molecules in the system be n , and the proportionality constant be k . (We choose a unit of time that is small enough to insure
A
¯¯¯¯¯¯
¯ ¯¯¯¯¯¯
¯ ¯¯¯¯¯¯
¯
that Δn ≪ n .) If the probability of reaction is constant, we have Δn = −kn . Since Δn is the number of moles that react
A A A A A
¯¯¯¯¯¯
¯
in unit time, the number of moles that react in time Δt is Δn = Δn Δt , so that
A A
¯¯¯¯¯¯
¯ ΔnA
ΔnA = = −knA
Δt
Dividing by the volume of the system, we have

1 ΔnA nA
= −k
V Δt V
In the limit that Δt → 0 , the term on the left becomes the reaction rate, R = V −1
(dnA /dt) , and since n A /V = [A] , we have
1 dnA
R = = −k [A]
V dt
Thus, a constant reaction probability implies that a unimolecular reaction has a first-order rate law. If the volume is constant, we
have
d [A]
= −k [A]
dt
The idea that a unimolecular reaction corresponds to a constant reaction probability can be rationalized by introducing a simple
model of the reaction process. This model assumes that reactant molecules have a distribution of energies, that only molecules
whose energies exceed some minimum can react, and that this excess energy must be in some specific internal motion before the
reaction can occur. Molecules exchange energy by colliding with one another. When a molecule acquires excess energy as the
result of a collision, redistribution of this energy among the motions available to the molecule is not instantaneous. A characteristic
length of time is required for excess energy to reach the specific internal motion that leads to reaction. Any given molecule can
retain excess energy only for the short time between two collisions. The molecule gains excess energy in one collision and loses it
in a subsequent one. Reaction can occur only if the excess energy reaches the specific internal motion before the molecule
undergoes a deactivating collision. (We return to these ideas in §14 and §15.)
In summary, only two kinds of elementary processes are needed to develop a mechanism for nearly any chemical change. These
elementary processes and their rate laws are:
Unimolecular A →
d [A]
= −k [A]
dt
Bimolecular A + B →
d [A]
= −k [A] [B]
dt
Finally, we should note that we develop these rate laws for elementary processes under the assumption that the rate at which
molecules collide is proportional to the concentrations of the colliding species. In doing so, we implicitly assume that
intermolecular forces of attraction or repulsion have no effect on this rate. When our goal is to predict rate laws from reaction
mechanisms, this assumption is almost always an adequate approximation. However, when we study chemical equilibria, we often
find that we must allow for the effects of intermolecular forces in order to obtain an adequate description. In chemical
thermodynamics, we provide for the effects of such forces by introducing the idea of a chemical activity. The underlying idea is
that the chemical activity of a compound is the effective concentration of the compound—we can view it as the concentration
“corrected” for the effects of intermolecular forces.
This page titled 5.7: Rate Laws for Elementary Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
The determination of a rate law is a matter of finding an empirical equation that adequately describes reaction-rate data. We can
distinguish two general approaches to this task. One approach is to measure reaction rate directly. That is, for A + B → C , we
measure the reaction rate in experiments where the concentrations, [A], [B], and [C ], of reactants and products are known. The
other is to measure a concentration at frequent time intervals as a batch reaction goes nearly to completion. We then seek a
differential equation that is consistent with this concentration-versus-time data.
If the reaction is the only process that affects Δn , direct measurement of the reaction rate can be effected by measuring Δn
C C
over a short time interval, Δt, in which the concentrations, [A], [B], and [C ], do not change appreciably. This is often difficult to
implement experimentally, primarily because it is difficult to measure small values of Δn with the necessary accuracy at known
C
values of [A], [B], and [C ].
Method of Initial Rates

The method of initial rates is an experimentally simple method in which the reaction rate is measured directly. Initial-rate
measurements are extensively used in the study of enzyme-catalyzed reactions. Direct measurement of reaction rate can also be
accomplished using a flow reactor. We discuss the method of initial rates, a particular kind of flow reactor known as a CSTR, and
enzyme catalysis in Sections 5.10, 5.11, and 5.13, respectively.
The most common reaction-rate experiment is a batch reaction in which we mix the reactants as rapidly as possible and then
monitor the concentration vs. time of one (or more) of the reactants or products as the reaction proceeds. We do the mixing so that
the initially mixed reactants are at a known temperature, which can be maintained constant for the remainder of the experiment.
The data from such an experiment are a set of concentrations and the times at which they are measured. To find the rate law
corresponding to these concentration-versus-time data, we employ a trial-and-error procedure. We guess what the rate law is likely
to be. We then obtain a general solution for this differential equation. This solution predicts the dependence of concentrations
versus time as a function of one or more rate constants. If we can obtain a satisfactory fit of experimental concentration-versus-time
data to the concentration-versus-time equation predicted by the rate law, we conclude that the rate law is a satisfactory
representation of the experimental data.
For a reaction A → C in a closed constant-volume system, we would want to test a first-order rate law rate law, which we can
express in several alternative ways:
1 dξ d [A] d [C ]
=− = = k [A]
V dt dt dt
Using the changing concentration of A to express the rate, separating variables, and integrating between the initial concentration
[A] = [A] at t = 0 and concentration [A] at time t gives
0
[A] t
d [A]
∫ = −k ∫ dt
[A] [A] 0
0
so that
[A]
ln = −kt
[A]
0
or
[A] = [A] exp (−kT )

0
Frequently it is convenient to introduce the extent of reaction or the concentration of a product as a parameter. In the present
instance, if the initial concentration of C is zero, [C ] = ξ/V = x . Then at any time, t, we have [A] = [A] − x , and the first-order
0
rate equation can be written as

dx
= k ([A] − x)
0
dt
which we rearrange and integrate between the limits x (0) = 0 and x (t) = x as
x t
−dx
∫ = −k ∫ dt
0 [A] −x 0
0
To give
[A] −x
0
ln( ) = −kt
[A]
0
It is easy to test whether concentration versus time data conform to the first-order decay model. If they do, a plot of ln([A] 0
− x) or
ln[A], versus time, t , is a straight line.
For a reaction 2A → C , we would want to test a rate law rate of the form
1 dξ 1 d [A] d [C ] 2
=− = = k[A]
V dt 2 dt dt
If the initial concentration of C is zero, [C ] = ξ/V =x , and [A] = [A] 0

− 2x at any time t . The rate law can be written as
dx 2
= k ([A]0 − 2x)
dt
and rearranged and integrated as

x t
−dx
∫ = −k ∫ dt
2
0 ([A] − 2x) 0
0
to give
1 1
− = 2kt
[A] − 2x [A]
0 0
or
1 1
− = 2kt
[A] [A]0
−1
If concentration-versus-time data conform to this second-order rate law, a plot of [A] versus time is a straight line.
For a reaction A + B → C , we would want to test a rate law of the form
1 dξ d [A] d [B] d [C ]
=− =− = = k [A] [B]
V dt dt dt dt
If the initial concentration of C is again zero, [C ] = ξ/V =x , [A] = [A] 0

−x and [B] = [B] 0
−x at any time t . The rate law can
be written as
dx
= k ([A] − x) ([B] − x)
0 0
dt
If [A]
0
≠ [B]
0
, this can be integrated (by partial fractions) to give
1 [A] ([B] − x)
0 0
ln = kt
[B] − [A] [B] ([A] − x)
0 0 0 0
If experimental data conform to this equation, a plot of

([B] − x)
0
ln
([A] − x)
0
versus time is linear. In practice, this often has disadvantages, and experiments to study reactions like this typically exploit the
technique of flooding.
Flooding is a widely used experimental technique that enables us to simplify a complex rate law in a way that makes it more
convenient to test experimentally. In the case we are considering, we can often arrange to carry out the reaction with the initial
concentration of B much greater than the initial concentration of A . Then the change that occurs in the concentration of B during
the reaction has much less effect on the reaction rate than the change that occurs in the concentration of A ; in the rate equation, it
becomes a good approximation to let [B] = [B] at all times. (For a fuller consideration of this point, see problem 5.23.) The
0
second-order rate equation simplifies to

d [A] d [C ]
=− = −kobs [A]
dt dt
where
kobs = k[B]
0
Since the simplified rate equation is approximately first order, the observed rate constant, k obs , is the slope of a plot of ln[A] versus
t. kobsis called a pseudo-first-order rate constant.
Of course, one such experiment tests only whether the true rate law is first order in [A]. It tells nothing about the dependence on
[B]. If we do several such experiments at different initial concentrations of B , the resulting set of k values must be directly
obs
proportional to the corresponding [B] values. This can be tested graphically by plotting k versus [B] . If the rate law is first
0 obs 0
order in [B], the resulting plot is linear with an intercept of zero. The slope of this plot is the second-order rate constant, k .
Flooding works by simplifying the rate law that is observed in a given experiment. Similar simplification can be achieved by
designing the experiment so that the initial concentrations of two or more reactants are proportional to their stoichiometric
coefficients. For the reaction A + B → C and the expected rate law
d [C ]
= k [A] [B]
dt
we would initiate the experiment with equal concentration of reactants A and B . Letting [A] = [B] = α and [C ] = ξ/V
0 0
=x ,
the concentrations of A and B at longer times become [A] = [B] = α − x . The rate law becomes effectively second order.
dx 2
= k (α − x)
dt
This page titled 5.8: Experimental Determination of Rate Laws is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
First-order rate processes are ubiquitous in nature—and commerce. In chemistry we are usually interested in first-order decay
processes; in other subjects, first-order growth is common. We can develop our appreciation for the dynamics—and mathematics—
of first-order processes by considering the closely related subject of compound interest.
Compound Interest
When a bank says that it pays 5% annual interest, compounded annually, on a deposit, it means that for every $1.00 we
deposit at the beginning of a year, the bank will add 5% or $0.05 to our account at the end of the year, making our deposit
worth $1.05. If we let the value of our deposit at the end of year n be P (n), and the interest rate (expressed as a fraction) be r,
with r > 0 , we can write
P (1) = P (0) + ΔP = P (0) + rP (0) = (1 + r) P (0)
where we represent the first year’s interest by ΔP = rP (0) . If we leave all of the money in the account for an additional year,
we will have
2
P (2) = (1 + r) P (1) = (1 + r) P (0)
and after t years we will have

t
P (t) = (1 + r) P (0)
Sometimes a bank will say that it pays 5% annual interest, compounded monthly. Then the bank means that it will compute a
new balance every month, based on r = 0.05 year = (0.05/12) month . After one month
−1 −1
0.05
P (1 month) = (1 + ) P (0)
12
and after n months

n
0.05
P (n months) = (1 + ) P (0)
12
If we want the value of the account after t years, we have, since n = 12t ,
12t
0.05
P (t) = (1 + ) P (0)
12
If the bank were to say that it pays interest at the rate r, compounded daily, the balance at the end of t years would be
r 365t
P (t) = (1 + ) P (0)
365
For any number of compoundings, m, at rate r, during a year, the balance at the end of t years would be
mt
r
P (t) = (1 + ) P (0)
m
Sometimes banks speak of continuous compounding, which means that they compute the value of the account at time t as the
limit of this equation as m becomes arbitrarily large. That is, for continuous compounding, we have
mt
r
P (t) = lim [ (1 + ) ] P (0)
m→∞ m
Fortunately, we can think about the continuous compounding of interestinterest:continuous compounding in another way. What
we mean is that the change in the value of the account, ΔP , over a short time interval, Δt, is given by
ΔP = rP Δt
where P is the (initial) value of the account for the interval Δt, and r is the fractional change in P during one unit of time. So
we can write
ΔP dP
lim ( ) = = rP
Δt→0 Δt dt
Separating variables to obtain dP /P = rdt and integrating between the limits P = P (0) at t = 0 and P = P (t) at t = t , we
obtain
P (t)
ln = rt
P (0)
or
P (t) = P (0) exp (rt)
Comparing the two equations we have derived for continuous compounding, we see that
mt
r
exp (rt) = lim (1 + )
m→∞ m
Continuous compounding of interest is an example of first-order or exponential growth. Other examples are found in nature; the
growth of bacteria normally follows such an equation. Reflection suggests that such behavior should not be considered remarkable.
It requires only that the increase per unit time in some quantity, P , be proportional to the amount of P that is already present:
ΔP = rP Δt . Since P measures the number of items (dollars, molecules, bacteria) present, this is equivalent to our observation in
Section 5.7 that a first-order process corresponds to a constant probability that a given individual item will disappear (first-order
decay) or reproduce (first-order growth) in unit time. For a first-order decay we have, keeping r > 0 ,
ΔP = −rP Δt
In the limit as Δt → 0 ,
dP
= −rP
dt
which has solution
P (t) = P (0) exp (−rt)
First-order growth and first-order decay both depend exponentially on rt . The difference is in the sign of the exponential term. For
exponential growth, P (t) becomes arbitrarily large as t → ∞ ; for exponential decay, P (t) goes to zero. If the concentration of a
chemical species A decreases according to a first-order rate law, we have
[A]
ln = −kt
[A]0
The units of the rate constant, k , are s . The half-life of a chemical reaction is the time required for one-half of the
−1
stoichiometrically possible change to occur. For a first-order decay, the half-life, t , is the time required for the concentration of
1/2
the reacting species to decrease to one-half of its value at time zero; that is, when the time is t , the concentration is 1/2
[A] = [A] /2 . Substituting into the integrated rate law, we find that the half-life of a first-order decay is independent of
0
concentration; the half-life is

ln 2
t1/2 =
k
This page titled 5.9: First-order Rate Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
In concept, the most straightforward way to measure reaction rate directly is to measure the change in the concentration of one
reagent in a short time interval immediately following initiation of the reaction. The initial concentrations are known from the way
the reaction mixture is prepared. If necessary, the initial mixture can be prepared so that known concentrations of products or
reaction intermediates are present. The initial reaction rate is approximated as the measured concentration change divided by the
elapsed time. The accuracy of initial-rate measurements is often poor. This can result from concentration variations associated with
initiation of the reaction; the actual mixing process is not instantaneous and significant reaction can occur before the mixture
becomes truly homogeneous. Measuring small changes in concentration with sufficient accuracy can also be difficult.
Enzymes are naturally occurring catalysts for biochemical reactions. In the study of enzyme-catalyzed reactions, it is usually
possible to select the enzyme concentration and other reaction conditions so that the initial rate can be measured with adequate
accuracy. For such studies, initial-rate measurements are used extensively. For other types of reactions, the method of initial rates is
usually less effective than alternative methods.
To illustrate the application of the method, suppose we have a reaction
A + B + C → P roducts
and that we are able to measure small changes in [A] with good accuracy. We seek a rate law of the form
d[A]
− = f ([A], [B], [C ])
dt
For any given experiment we approximate d[A]/dt by Δ[A]/Δt, and approximate the average concentrations of the reagents over
the interval Δt by their initial values: [A] = [A] , [B] = [B] , and [C ] = [C ] . By carrying out an number of such experiments
0 0 0
with suitably chosen initial concentrations, we can determine the functional form of the rate law and evaluate the rate constants that
appear in it.
Table 5.10.1 : Hypothetical reaction rate data
Δ[A]/Δt ,–
M
––
,
Δt s
−1
[A]
0
[B]
0
[C]
0
−2 × 10
−7
1000 0.010 0.010 0.010
−4 × 10
−7
500 0.010 0.010 0.020
−2 × 10
−7
1000 0.010 0.020 0.010
−8 × 10
−7
2500 0.020 0.010 0.010
Table 5.10.1 presents data for a hypothetical reaction that serve to illustrate the basic concept. We suppose that initial rates have
been determined for four different combinations of initial concentrations. Comparison of the first and second experiments indicates
that doubling [C ] doubles the reaction rate, indicating that the rate depends on [C ] to the first power. Comparison of the first and
third experiments indicates that doubling [B] leaves the reaction rate unchanged, implying that the rate is independent of [B].
Comparison of the first and fourth experiments indicates that doubling [A] increases the reaction rate by a factor of four, implying
that the rate is proportional to the second power of [A]. We infer that the rate law is
d[A] 2 0 1
− = k[A] [B] [C ]
dt
Given the form of the rate law, an estimate of the value of the rate constant, k , can be obtained from the data for each experiment.
For this illustration, we calculate k = 0.20 ––
M s from each of the experiments.
−1
This page titled 5.10: Rate Laws by the Study of Initial Rates is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
A continuous stirred tank reactor (CSTR)—or capacity-flow reactor—is a superior method of collecting kinetic data when the rate
law is complex. Unfortunately, a CSTR tends to be expensive to construct and complex to operate. Figure 5.11.1 gives a schematic
representation of the essential features of a CSTR. Fresh reagents are fed to a reactor vessel of volume V at a constant rate. A
portion of the reactor contents is continuously removed at the same volumetric flow rate. Because the addition and removal of
material occur at the same rate, the reactor is always filled with a fixed volume of reaction mixture. The reaction vessel and its
contents are maintained at a constant temperature. The vessel contains a stirrer, which operates continuously and at a high enough
speed to keep the contents of the vessel homogeneous (free of concentration and temperature gradients) at all times.
Figure 5.11.1 : A continuous stirred-tank reactor.

The essential idea involved in the operation of a CSTR is that, after the passage of sufficient time, the concentrations of the various
species present in the reactor become constant. We say that the reactor contains steady-state concentrations of the reactants and
products. When the reactor reaches this steady state, processes that increase reagent concentrations are occurring at the same rate as
processes that decrease them.
Let the reaction be of the form:
A + B → Products.
The concentrations of the reagents in the feed solution are known. Let the concentration of A in the fresh feed solution be [A] . Let
0
the rate at which fresh reagent-containing solution is fed to the reactor be u L s . Homogeneous reaction mixture is withdrawn
−1 −1
from the vessel at the same flow rate. The amount of A in the reactor is increased by the flow of fresh reactant solution into the
reactor. It is decreased both by reaction and by the flow of solution out of the reactor. The steady-state reaction rate, R = R (A) , is
the number of moles of reactant A consumed by the reaction per unit time per unit volume of reaction vessel after all of the reagent
concentrations have become constant. Since A is a reactant, this rate is
1 dr nA
R =−
V dt
where d n /dt is the contribution that the reaction makes to the rate at which the number of moles of A in the reactor changes.
r A
Since all of the reaction occurs within the vessel, and the vessel is entirely filled with the solution, R is also the number of moles of
A consumed by reaction per unit time per unit volume of solution.
At steady state, the number of moles of A in the reactor is determined by:

1. the number of moles of A entering the reactor per unit time,
2. the number of moles of A being consumed by reaction per unit time, and
3. the number of moles of A leaving the reactor in the effluent stream per unit time.
In unit time, the number of moles entering with the feed is given by u[A] ; the number leaving with the effluent is given by u [A].
0
In unit time, the contribution that the reaction makes to the change in the number of moles of A present is −RV . When the steady
state is reached, the number of moles entering, plus the change due to reaction, must equal the number of moles leaving:
moles flowing in + change in moles due to reaction − moles flowing out = 0
or, in unit time,
u [A] − RV − u [A] = 0
0
Solving for R , we have

u
R =− ([A] − [A]0 )
V
(We define reaction rate so that R > 0 . If A is produced by the reaction, the mass-balance equation is u [A] 0
+ RV − u [A] = 0 .)
As with the method of initial rates, the rate law is determined by measuring reaction rates in a series of experiments in which the
steady-state concentrations of the various reactants and products vary. For each experiment it is necessary to determine both the
reaction rate and the steady-state concentration of each reagent that might be involved in the rate law. Using the equation above, the
rate is calculated from the difference between a reagent concentration in the feed solution and its steady-state concentration in the
reactor. The concentration of each reagent in the effluent is the same as its concentration in the reactor, so the necessary
concentration information can be obtained by chemical analysis of the effluent solution. The chemical analysis must be done in
such a way that no significant reaction occurs between the time the material leaves the reaction vessel and the time the analysis is
completed.
This page titled 5.11: Rate Laws from Experiments in a Continuous Stirred Tank Reactor is shared under a CC BY-SA 4.0 license and was
Because a proposed mechanism can only be valid if it is consistent with the rate law found experimentally, the rate law plays a
central role in the investigation of chemical reaction mechanisms. The discussion above introduces the problems and methods
associated with collecting rate data and with finding an empirical rate law that fits experimental concentration-versus-time data. We
turn now to finding the rate laws that are consistent with a particular proposed mechanism. For simplicity, we consider reactions in
closed constant-volume systems.
In principle, numerical integration can be used to predict the concentration at any time of each of the species in any proposed
reaction mechanism. This prediction can be compared to experimental observations to see whether they are consistent with the
proposed mechanism. To do the numerical integration, it is necessary to know the initial concentrations of all of the chemical
species and to know, or assume, values of all of the rate constants. The initial concentrations are known from the procedure used to
initiate the reaction. However, the rate constants must be determined by some iterative procedure in which initial estimates of the
rate constants are used to predict concentration-versus-time data that can be compared to the experimental results to produce
refined estimates.
In practice, we tailor our choice of reaction conditions so that we can use various approximations to test whether a proposed
mechanism can explain the data. We now consider the most generally useful of these approximations.
In this discussion, we assume that the overall reaction goes to completion; that is, at equilibrium the concentration of the reactant
whose concentration is limiting has become essentially zero. If the overall reaction involves more than one elementary step, then an
intermediate compound is involved. A valid mechanism must include this intermediate, and more than one differential equation
may be needed to characterize the time rate of change of all of the species involved in the reaction. We focus on conditions and
approximations under which the rate of appearance of the final products in a multi-step reaction mechanism can be described by a
single differential equation, the rate law.
We examine the application of these approximations to a particular reaction mechanism. When we understand the application of
these approximations to this mechanism, the ways in which they can be used in other situations are clear.
Consider the following sequence of elementary steps
k1 k3
A+B ⇌ C −
→D
k2
whose kinetics are described by the following simultaneous differential equations:

d[A] d[B]
= = −k 1 [A][B] + k2 [C ]
dt dt
d [C ]
= k1 [A][B] − k2 [C ] − k3 [C ]
dt
d [D]
= k3 [C ]
dt
The general analytical solution for this system of coupled differential equations can be obtained, but it is rather complex, because
[C ] increases early in the reaction, passes through a maximum, and then decreases at long times. In principle, experimental data
could be fit to these equations. The numerical approach requires that we select values for k , k , k , [A] , [B] , [C ] , and [D] ,
1 2 3 0 0 0 0
and then numerically integrate to get [A], [B], [C ], and [D] as functions of time. In principle, we could refine our estimates of k , 1
k , and k by comparing the calculated values of one or more concentrations to the experimental ones. In practice, the approximate
2 3
treatments we consider next are more expedient.

When we begin a kinetic study, we normally have a working hypothesis about the reaction mechanism, and we design our
experiments to simplify the differential equations that apply to it. For the present example, we will assume that we always arrange
the experiment so that [C ] = 0 and [D] = 0 . In consequence, at all times:
0 0
[A] = [A] + [C ] + [D] .

0
Also, we restrict our considerations to experiments in which [B] ≫ [A] . This exemplifies the use of flooding. The practical
0 0
effect is that the concentration of B remains effectively constant at its initial value throughout the entire reaction, which simplifies
the differential equations significantly. In the present instance, setting [B]
0
≫ [A]
0
means that the rate-law term k1 [A][B] can be
replaced, to a good approximation, by k [A], where k = k [B] .
obs obs 1 0
Once we have decided upon the reaction conditions we are going to use, whether the resulting concentration-versus-time data can
be described by a single differential equation depends on the relative magnitudes of the rate constants in the several steps of the
overall reaction. Particular combinations of relationships that lead to simplifications are often referred to by particular names; we
talk about a combination that has a rate-determining step, or one that involves a prior equilibrium, or one in which a steady-state
approximation is applicable. To see what we mean by these terms, let us consider some particular relationships that can exist
among the rate constants in the mechanism above.
Case I
Suppose that k [A][B] ≫ k [C ] and k ≫ k . We often describe this situation by saying, rather imprecisely, that the reaction to
1 2 3 2
convert C to D is very fast and that the reaction to convert C back to A and B is very slow—compared to the reaction that forms
C from A and B . When C is produced in these circumstances, it is converted to D so rapidly that we never observe a significant
concentration of C in the reaction mixture. The formation of a molecule of C is tantamount to the formation of a molecule of D,
and the reaction produces D at essentially the same rate that it consumes A or B . We say that the first step, A + B → C , is the
rate-determining step in the reaction. We have
d[A] d[B] d [D]
− = − ≈
dt dt dt
The assumption that k1 [A][B] ≫ k2 [C ] means that we can neglect the smaller term in the equation for d[A]/dt , giving the
approximation
d[A] d[B] d [D]
= = − = − k1 [A][B]
dt dt dt
Letting [D] = x and recognizing that our assumptions make [C ] ≈ 0 , the mass-balance condition, [A] = [A] + [C ] + [D] , 0
becomes [A] = [A] − x . Choosing [B] ≫ [A] means that k [B] ≈ k [B] = k
0 0 0
. The rate equation becomes first-order:
1 1 0 I,obs
dx
= kI,obs ( [A]0 − x)
dt
Since kI,obs is not strictly constant, it is a pseudo-first-order rate constant. The disappearance of A is said to follow a pseudo-first-
order rate equation.
The concept of a rate-determining step step is an approximation. In general, the consequence we have in mind when we invoke
this approximation is that no intermediate species can accumulate to a significant concentration if it is produced by the rate-
determining step or by a step that occurs after the rate-determining step. We do not intend to exclude the accumulation of a species
that is at equilibrium with another product. Thus, in the mechanism
k
A → B−
↽⇀
− C
we suppose that the conversion of A to B is rate-determining and that the interconversion of B and C is so rapid that their
concentrations always satisfy the equilibrium relationship
[C ]
K = .
[B]
For the purpose at hand, we do not consider B to be an intermediate; B is a product that happens to be at equilibrium with the co-
product, C .
Case II
Suppose that k [A][B] ≫ k [C ]. In this case A + B → C is fast compared to the rate at which C is converted to D, and we say
1 3
that C → D is the rate-determining step. We can now distinguish three sub-cases depending upon the way [C ] behaves during the
course of the reaction.
Case IIa: Suppose that k [A][B] ≫ k [C ] and k ≫ k . Then A + B → C is rapid and essentially quantitative. That is, within a
1 3 3 2
short time of initiating the reaction, all of the stoichiometrically limiting reactant is converted to C . Letting [D] = x and
recognizing that our assumptions make [A] ≈ 0 , the mass-balance condition,
[A] = [A] + [C ] + [D]

0
becomes
[C ] = [A] − x.
0
After a short time, the rate at which D is formed becomes

d [D]
= k3 [C ]
dt
or
dx
= k3 ( [A] − x)
0
dt
The disappearance of C and the formation of D follow a first-order rate law.

Case IIb: If the forward and reverse reactions in the first elementary process are rapid, then this process may be effectively at
equilibrium during the entire time that D is being formed. (This is the case that k [A][B] ≫ k [C ] and k ≫ k .) Then, 1 3 2 3
throughout the course of the reaction, we have
Keq = [C ]/[A][B]
Letting [D] = x and making the further assumption that [A] ≫ [C ] ≈ 0 throughout the reaction, the mass-balance condition,
[A] = [A] + [C ] + [D] , becomes [A] = [A] − x . Substituting into the equilibrium-constant expression, we find
0 0
[C ] = Keq [B] ([A] − x)

0 0
Substituting into d [D]/dt = k 3 [C ] we have

dx
= k3 Keq [B] ([A] − x) = kIIa,obs ([A] − x)
0 0 0
dt
where k = k K [B] . The disappearance of A and the formation of D follow a pseudo-first-order rate equation. The
IIa,obs 3 eq 0
pseudo-first-order rate constant is a composite quantity that is directly proportional to [B] . 0
Case IIc: If we suppose that the first step is effectively at equilibrium during the entire time that D is being produced (as in case
IIb) but that [C ] is not negligibly small compared to [A], we again have K = [C ]/[A][B]. With [D] = x , the mass-balance eq
condition becomes [A] = [A] − [C ] − x . Eliminating [A] between the mass-balance and equilibrium-constant equations gives
0
Keq [B] ([A] − x)

0 0
[C ] =
1 + Keq [B]
0
so that d [D]/dt = k 3 [C ] becomes
dx k3 Keq [B]
0
=( ) ([A] − x) = kIIc,obs ([A] − x)
0 0
dt 1 + Keq [B]
0
The formation of D follows a pseudo-first-order rate equation. (The disappearance of A is also pseudo-first-order, but the pseudo-
first-order rate constant is different.) As in Case IIb, the pseudo-first-order rate constant, k , is a composite quantity, but now IIc,obs
its dependence on [B] is more complex. The result for Case IIc reduces to that for Case IIb if K [B] ≪ 1 .
0 eq 0
Case III
In the cases above, we have assumed that one or more reactions are intrinsically much slower than others are. The differential
equations for this mechanism can also become much simpler if all three reactions proceed at similar rates, but do so in such a way
that the concentration of the intermediate is always very small, [C ] ≈ 0 . If the concentration of C is always very small, then we
expect the graph of [C ] versus time to have a slope, d [C ]/dt, that is approximately zero. In this case, we have
d [C ]
= k1 [A][B] − k2 [C ] − k3 [C ] ≈ 0
dt
so that
k1 [A][B]
[C ] =
k2 + k3
With [D] = x , d [D]/dt = k 3 [C ] becomes
dx k1 k3 [B]
0
=( ) ([A] − x) = kIII,obs ([A] − x)
0 0
dt k2 + k3
As in the previous cases, the disappearance of A and the formation of D follow a pseudo-first-order rate equation. The pseudo-
first-order rate constant is again a composite quantity, which depends on [B] and the values of all of the rate constants.
0
Case III illustrates the steady-state approximation, in which we assume that the concentration of an intermediate species is much
smaller than the concentrations of other species that affect the reaction rate. Under these circumstances, we can be confident that
the time-derivative of the intermediate’s concentration is negligible compared to the reaction rate, so that it is a good approximation
to set it equal to zero. The idea is simply that, if the concentration is always small, its time-derivative must also be small. If the
graph of the intermediate’s concentration versus time is always much lower than that of other participating species, then its slope
will be much less.
Equating the time derivative of the steady-state intermediate’s concentration to zero produces an algebraic expression that involves
the intermediate’s concentration. Solving this expression for the concentration of the steady-state intermediate makes it possible to
greatly simplify the set of simultaneous differential equations that is predicted by the mechanism. When there are multiple
intermediates to which the approximation is applicable, remarkable simplifications can result. This often happens when the
mechanism involves free-radical intermediates.
The name “steady-state approximation” is traditional. When we use it, we do so on the understanding that the “state” which is
approximately “steady” is the concentration of the intermediate, not the state of the system. Since a net reaction is occurring, the
state of the system is distinctly not constant.
This page titled 5.12: Predicting Rate Laws from Proposed Mechanisms is shared under a CC BY-SA 4.0 license and was authored, remixed,
An enzyme is a molecule produced by a living organism. Its enzymes are essential to the life processes of the organism. Often
enzymes are large molecules that are catalytically active only when folded into a particular conformation. Molecules whose
reactions are catalyzed by an enzyme are customarily referred to as substrates for that enzyme. Two aspects of enzymatic catalysis
are remarkable. First, it is often found that an enzyme can discriminate between two very similar substrates, greatly enhancing the
reaction rate of one while having a much smaller effect on the rate of the other. Second, the rate enhancements achieved by enzyme
catalysis are often very large; rate increases by a factor of 10 are observed.
6
Simple mechanistic and kinetic models are sufficient to explain these essential features of enzyme catalysis. We consider the
simplest case. Both the mechanisms and the rate laws for enzymatic reactions can become much more complex than the model we
develop. (The literature of enzyme catalysis uses a specialized vocabulary. Problem 32 introduces some of this terminology and
some of the mechanistic complications that can be observed.)
The catalytic specificity of enzymes is explained by the idea that the enzyme and the substrate have complex three-dimensional
structures. These structures complement one another in the sense that enzyme and substrate can fit together tightly, bringing the
catalytically active parts of the enzyme’s structure into close proximity with those substrate chemical bonds that are changed in the
reaction. This is often called the lock and key model for enzyme specificity, invoking the idea that the detailed features of the
enzyme’s structure are shaped to fit into the structure of its substrate, just as a key is machined to match the arrangement of
tumblers in the lock that it opens. Figure 2 illustrates this idea.
Figure 2. The lock and key model.

The rates of enzyme-catalyzed reactions can exhibit complex dependence on the relative concentrations of enzyme and substrate.
Most of these features are explained by the Michaelis-Menten mechanism, which postulates a rapid equilibration of enzyme and
substrate with their enzyme-substrate complex. Transformation of the substrate occurs within this complex. The reaction products
do not complex strongly with the enzyme. After the substrate has been transformed, the products diffuse away. The enzyme can
then complex with another substrate molecule and catalyze its reaction. Representing the enzyme, substrate, enzyme–substrate
complex, and products as E , S , ES , and P , respectively, the simplest-case Michaelis-Menten mechanism is
Complexation equilibrium:
ES ⇌ E + S
Substrate transformation:
ES → E + P
where K is the equilibrium constant for the dissociation of the enzyme–substrate complex, and k is the rate constant for the rate-
determining transformation of the enzyme–substrate complex into products.
Since the first-order decay of the enzyme–substrate complex is the product-forming step, the reaction rate is expected to be
d [P ]
= k [ES]
dt
Letting the product concentration be [P ] = x and the initial concentrations of enzyme and substrate be E0 and S0 , respectively,
material balance requires
E0 = [E] + [ES] and S 0 = [S] + [ES] + x
Most experiments are done with the substrate present at a much greater concentration than the enzyme, S ≫ E . In this case, 0 0
[ES] is negligible in the S equation, and the concentration of free substrate becomes [S] = S − x . The enzyme-substrate
0 0
equilibrium imposes the relationship

[E] [S] [E] (S0 − x)
K = =
[ES] [ES]
Using this relationship to eliminate [E] from the expression for E gives 0
K [ES] K
E0 = + [ES] = ( + 1) [ES]
S0 − x S0 − x
The rate law for product formation becomes

dx kE0 (S0 − x)
= k [ES] =
dt K + (S0 − x)
If K ≪ S 0 −x , the dependence on substrate cancels, and the rate law becomes pseudo-zero-order
dx
= kE0 = kobs
dt
where the observed zero-order rate constant depends on E . If K ≫ S 0 0 −x , the rate law becomes pseudo-first-order
dx kE0
= (S0 − x) = kobs (S0 − x)
dt K
with pseudo-first-order rate constant k = kE

obs 0 /K . If neither of these simplifications is applicable, fitting experimental data to
the integrated rate law becomes inconvenient.
In practice, the integrated form of the rate law is seldom used. As noted earlier, enzyme catalysis is usually studied using the
method of initial rates. That is, the rate is approximated as dx/dt ≈ Δx/Δt by measuring the amount of product formed over a
short time interval, Δt, in which x ≪ S . The rate and equilibrium constants can then be found by fitting the rate measured at
0
various initial concentrations, E and S , to the equation

0 0
Δx kE0 S0
≈
Δt K + S0
Since the concentration of the enzyme–substrate complex is normally small compared to the concentration of the substrate, the
Michaelis-Menten mechanism can also be analyzed by applying the steady-state approximation to [ES]. If the product-forming
step occurs so rapidly that the concentration of the enzyme–substrate complex is not maintained at the equilibrium value, the
steady-state treatment becomes necessary.
This page titled 5.13: The Michaelis-Menten Mechanism for Enzyme-catalyzed Reactions is shared under a CC BY-SA 4.0 license and was
First-order kinetics for a unimolecular reaction corresponds to a constant probability that a given molecule reacts in unit time. In
Section 5.7, we outline a simple mechanism that rationalizes this fact. This mechanism assumes that the probability of reaction is
zero unless the molecule has some minimum energy. For molecules whose energy exceeds the threshold value, we assume that the
probability of reaction is constant. However, when collisions are frequent, a molecule can have excess energy only for brief
intervals.
The Lindemann-Hinshelwood mechanism for gas-phase unimolecular reactions provides a mathematical model of these ideas.
Since molecules exchange energy via collisions, any given molecule acquires excess energy by collisions with other molecules, and
loses it within a short time through other collisions. If it retains its excess energy long enough, it will react. If collisions are very
infrequent, every molecule that acquires excess energy reacts before it undergoes a deactivating collision. In this case the reaction
rate is proportional to the rate at which molecules acquire excess energy, which is proportional to the number of collisions. In a
collection of A molecules, the total number of A−A collisions is proportional to [A] not [A] and so the reaction rate depends on
2
2
[A] not [A].
We represent molecules with excess energy as A , and assume that all A molecules undergo reaction with a constant probability.
∗ ∗
A molecules are formed in collisions between A molecules, and they are deactivated by subsequent collisions with A molecules.
∗
k1
∗
A+ A ⇌ A +A (step 1)
k2
k3
∗
A −
→P (step 2)
where P is the product(s) of the reaction. The rate at which the number of moles of A molecules changes is ∗
1 dnA∗ 2 ∗ ∗
= k1 [A] − k2 [ A ][A] − k3 [ A ]
V dt
and since we suppose that [A ∗

] is always very small, the steady-state approximation applies, so that dn A
∗ /dt ≈ 0 , and
2
k1 [A]
∗
[A ] =
k2 [A] + k3
The reaction rate is given by

1 dξ ∗
= k3 [ A ]
V dt
2
k1 k3 [A]
= (5.14.1)
k2 [A] + k3
When k [A] ≫ k , the rate of deactivating collisions between A and A is greater than the rate at which
2 3
∗
A
∗
molecules go on to
become products, the rate law (Equation 5.14.1) for consumption of [A] becomes first order:
2 2
k1 k3 [A] k1 k3 [A]
lim =
k2 [A]≫k3 k2 [A] + k3 k2 [A]
k1 k3
= [A]
k2
= khigh pressure [A]
where the first order rate constant k high pressure is a function of k , k and k . This is termed the rate law’s high-pressure limit.
1 2 3
The low-pressure limit occurs when k 2 [A] ≪ k3 . The rate law (Equation 5.14.1) becomes second order in [A]:
2
2
k1 k3 [A] k1 k3 [A]
lim =
k2 [A]≪k3 k2 [A] + k3 k3
2
= klow pressure[A]
where the second order rate constant k low pressure is a function of only k . The rate of product formation becomes equal to the rate at
1
which A molecules are formed.

∗
This page titled 5.14: The Lindemann-Hinshelwood Mechanism for First-order Decay is shared under a CC BY-SA 4.0 license and was authored,
In the discussion above, we assume that a molecule with energy in excess of the minimum activation energy undergoes reaction
with some fixed probability, represented by the rate constant k . A complete answer to the question of why unimolecular processes
3
are characteristically first-order (in the high pressure limit) requires that we rationalize this assumption. Another way of phrasing
this question is to ask why the activated molecule does not react immediately: Why isn’t k ≈ ∞ ? 3
The total energy of a molecule is distributed among numerous degrees of freedom. The molecule has translational kinetic energy,
rotational kinetic energy, and vibrational energy. When it acquires excess energy through a collision with another molecule, the
additional energy could go directly into any of these modes. However, before the molecule can react, enough energy must find its
way into some rather particular mode. If, for example, the reaction involves the breaking of a chemical bond, and the collision puts
excess kinetic energy into the molecule’s translational modes, the reaction can occur only after some part of the excess translational
energy has been converted to excess vibrational energy in the breaking bond. This intramolecular transfer of energy among the
molecule’s various internal modes is time-dependent.
From this perspective, the probability that an excited molecule will react in unit time is the probability that the necessary energy
will reach the critical locus in unit time. The reshuffling of energy among the molecule’s internal modes is a stochastic process, and
the probability that the reshuffling will put the necessary energy where it is needed is a constant characteristic of the molecule.
This page titled 5.15: Why Unimolecular Reactions are First Order is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
This chapter focuses on the relationship between rate laws and reaction mechanisms. We have noted that the rate law is rarely
sufficient to establish the mechanism of a particular reaction. The base hydrolysis of cobalt pentammine complexes is a reaction for
which numerous lines of evidence converge to establish the mechanism. To illustrate the range of data that can be useful in the
determination of reaction mechanisms, we summarize this evidence here.
Cobalt(III) complexes usually undergo substitution reactions at readily measurable rates. Cobalt pentammine complexes,
, have been studied extensively.
+
n
Co(NH ) X
3 5
Mechanism 1
In acidic aqueous solutions, the reaction
+ − + −
n p m q
Co (NH ) X +Y ⟶ Co (NH ) Y +X
3 5 3 5
− −
(X ,
q
Y
p
= Cl
−
, Br , NO , SCN , CH CO , etc.) usually proceeds exclusively through the aquo complex,
− −
2
−
3
−
Co(NH ) OH
3 5 2
3 +
. The first step in the reaction is the breaking of a Co−X bond and the formation of a Co−OH bond (step 1). 2
Subsequently, a Y p−
moiety can replace the aquo group (step 2). For example:
2 + 3 + −
Co (NH ) Cl + H O ⇌ Co (NH ) (OH ) + Cl (step 1)
3 5 2 3 5 2
3 + − 2 +
Co (NH ) (OH ) + Br ⇌ Co (NH ) Br +H O (step 2)
3 5 2 3 5 2
In aqueous solution, water is always present at a much higher concentration than the various possible entering groups Y , so it is p−
reasonable that it should be favored in the competition to form the new bond to Co(III). Nevertheless, we expect the strength of
the Co−Y bond to be an indicator of the nucleophilicity of Y in these substitution reactions. The fact that the aquo complex is
p−
the predominant reaction product strongly suggests that the energetics of the reaction are dominated by the breaking of the Co−X
bond; formation of the new bond to the incoming ligand apparently has little effect. Whether the old Co−X bond has been
completely broken (so that Co(NH ) is a true intermediate) before the new Co−OH bond has begun to form remains an issue
3
3 +
5 2
on which it is possible to disagree.
Mechanism 2
There is a conspicuous exception to the description given above. When the entering group, Y p−
, is the hydroxide ion, the reaction
is
+ −
n − 2 + q
Co (NH ) X + OH → Co (NH ) OH +X
3 5 3 5
This is called the base-hydrolysis reaction. It is faster than the formation of the aquo complex in acidic solutions, and the rate is
law found to be
2 +
d [Co (NH ) OH ] +
3 5 n −
= k [Co (NH ) X ] [ OH ]
3 5
dt
This rate law is consistent with S 2 nucleophilic attack by the hydroxide ion at the cobalt center, so that Co−OH bond formation
N
occurs simultaneously with breaking of the Co−X bond. However, this interpretation means that the hydroxide ion is a uniquely
effective nucleophile toward cobalt(III). Nucleophilic displacements have been investigated on many other electrophiles. In
general, hydroxide is not a particularly effective nucleophile toward other electrophilic centers. So, assignment of an S 2 N
mechanism to this reaction is reasonable only if we can explain why hydroxide is uniquely reactive in this case and not in others.
Mechanism 3
An alternative mechanism, usually labeled the S 1CB (Substitution, Nucleophilic, first-order in the Conjugate Base mechanism)
N
mechanism, is also consistent with the second-order rate law. In this mechanism, hydroxide removes a proton from one of the
ammine ligands, to give a six-coordinate intermediate, containing an amido (NH ) ligand (step 1). This intermediate loses the −
2
leaving group X in the rate determining step to form a five-coordinate intermediate, Co(NH ) NH
q−
(step 2). This 3 4
2 +
intermediate picks up a water molecule to give the aquo complex (step 3). In a series of proton transfers to (step 4) and from (step
5) the aqueous solvent, the aquo complex rearranges to the final product. With Cl as the leaving group, the S 1CB mechanism is
−
N
2 + − +
Co (NH ) Cl + OH ⇌ Co (NH ) (NH )Cl +H O (step 1)
3 5 3 4 2 2
+ 2 + −
Co (NH ) (NH )Cl ⇌ Co (NH ) (NH ) + Cl (step 2)
3 4 2 3 4 2
2 + 2 +
Co (NH ) (NH ) + H O ⇌ Co(NH ) (NH )(OH ) (step 3)
3 4 2 2 3 4 2 2
2 + − +
Co (NH ) (NH )(OH ) + OH ⇌ Co (NH ) (NH )OH +H O (step 4)
3 4 2 2 3 4 2 2
+ 2 + −
Co (NH ) (NH )OH + H O ⇌ Co (NH ) OH + OH (step 5)
3 4 2 2 3 5
The evidence in favor of the S 1CB mechanism is persuasive. It requires that the ammine protons be acidic, so that they can
N
undergo the acid–base reaction in the first step. That this reaction occurs is demonstrated by proton-exchange experiments. In basic
D O , the ammine protons undergo H−D exchange according to
2
2 + 2 +
Co (NH ) Cl + D O ⇌ Co (ND ) Cl + HDO
3 5 2 3 5
The ammine protons are also necessary; base hydrolysis does not occur for similar compounds, like
′
Co(2 , 2 −bipyridine) (O CCH )
2 2 3
+
2
, in which there are no protons on the nitrogen atoms that are bound to cobalt (i.e., there are
no H−N−Co moieties).
The evidence that Co(NH ) (NH ) is an intermediate is also persuasive. When the base hydrolysis reaction is carried out in the
3 4 2
2 +
presence of other possible entering groups, Y , the rate at which Co(NH ) X is consumed is unchanged, but the product is a
p−
3 5
n
mixture of Co(NH ) OH and Co(NH ) Y . If this experiment is done with a variety of leaving groups, X , the proportions
3 5
2 +
3 5
n q−
of Co(NH ) OH and Co(NH ) Y are constant—independent of which leaving group the reactant molecule contains. These
3 5
2 +
3 5
n
observations are consistent with the hypothesis that all reactants, Co(NH ) X , give the same intermediate, 3 5
n
2 +
Co(NH ) (NH )
3 4
. The product distribution is always the same, because it is always the same species undergoing the product-
2
forming reaction.
This page titled 5.16: The Mechanism of the Base Hydrolysis of Co(NH₃)₅Xⁿ⁺ is shared under a CC BY-SA 4.0 license and was authored,
Suppose that the bimolecular reaction
kf
A+B ⇌ C +D
kr
occurs as an elementary process. From our conclusions about the concentration dependencies of elementary reactions, the rate of
the net reaction is
1 dξ
R =
V dt
1 dnA
=−
V dt
= kf [A][B] − kr [C ][D]
at any time. In particular, this rate equation must remain true at arbitrarily long times—times at which the reaction has reached
equilibrium and at which dn /dt = 0 . Therefore, at equilibrium we have
A
kf
K = (5.17.1)
kr
[C ]eq [D]eq
= (5.17.2)
[A]eq [B]eq
where the concentration-term subscripts serve to emphasize that the concentration values correspond to the reaction being in a state
of equilibrium. We see that the ratio of rate constants, k /k , characterizes the equilibrium state. This constant is so useful we give
f r
it a separate name and symbol, the equilibrium constant, K .

Now, let us consider the possibility that the reaction is not an elementary process, but instead proceeds by a two-step mechanism
involving an intermediate, E :
A+ B −
↽⇀
− [ k ][ k ]E
1 2
E −
↽⇀
− [ k ][ k ]C + D
3 4
The sum of these elementary processes yields the same overall reaction as before. This mechanism implies the following
differential equations:
1 dnA
= −k1 [A][B] + k2 [E]
V dt
1 dnD
= k3 [E] − k4 [C ][D]
V dt
At equilibrium, both n and n must be constant, so both differential equations must be equal to zero. Hence, at equilibrium,
A D
k1 [E]eq
=
k2 [A]eq [B]eq
k3 [C ]eq [D]eq
=
k4 [E]eq
Multiplying these, we have

k1 k3
K = (5.17.3)
k2 k4
[C ]eq [D]eq
= (5.17.4)
[A]eq [B]eq
The concentration dependence of the equilibrium constant for the two-step mechanism (Equation 5.17.4) is the same as for case
that the reaction is an elementary process (Equation 5.17.2). As far as the description of the equilibrium system is concerned, the
only difference is that the equilibrium constant is interpreted as a function of different rate constants (Equation 5.17.1 vs. Equation
5.17.3).
For the general reaction
aA + bB + … −
↽⇀
− fF + gG + …
we see that any sequence of elementary reactions will give rise to the same concentration expression for the equilibrium system.
Whatever the mechanism, reactant A must appear a times more often on the left side of elementary reactions than it does on the
right. Product F must appear f times more often on the right side of elementary reactions than it does on the left. Any
intermediates must appear an equal number of times on the left and on the right in the various elementary reactions. As a result,
when we form the ratio of forward to reverse rate constants for each of the elementary reactions and multiply them, the
concentration of reactant A must appear in the product to the −a power, the concentration of product F must appear to the +f
power, and the concentrations of the intermediates must all cancel out. We conclude that the condition for equilibrium in the
general case is
f g
[F ] [G] …
K =
a b
[A] [B] …
where we drop the “eq ” subscripts, trusting ourselves to remember that the equation is valid only when the concentration terms
apply to the equilibrated system.
Pure Phases
When the reaction involves a pure phase as a reactant or product, we observe experimentally that the amount of the pure phase
present in the reaction mixture does NOT affect the position of equilibrium. The composition of the reaction solution is the same so
long as the solution is in contact with a finite amount of the pure phase. This means that we can omit the concentration of the
substance that makes up the pure phase when we write the equilibrium-constant expression. In writing the equilibrium constant
expression, we can take the concentration the substance to be an arbitrary constant. Unity is usually the most convenient choice for
this constant.
To rationalize this experimental observation within our kinetic model for equilibrium, we postulate that the rate at which molecules
leave the pure phase is proportional to the area, S , of the phase that is in contact with the reaction solution; that is,
Rleaving = k1 S.
We postulate that the rate at which molecules return to the pure phase from the reaction solution is proportional to both the area and
the concentration of the substance in the reaction solution. If the pure phase consists of substance A , we have
Rreturning = k2 S[A].
At equilibrium, we have R leaving = Rreturning , so that k

1S = k2 S[A] , and [A] = k
1 / k2 = constant
This page titled 5.17: Chemical Equilibrium as the Equality of Rates for Opposing Reactions is shared under a CC BY-SA 4.0 license and was
The equilibrium constant expression is an important and fundamental relationship that relates the concentrations of reactants and
products at equilibrium. We deduce it above from a simple model for the concentration dependence of elementary-reaction rates. In
doing so, we use the criterion that the time rate of change of any concentration must be zero at equilibrium. Clearly, this is a
necessary condition; if any concentration is changing with time, the reaction is not at equilibrium. However, our deduction uses
another assumption that we have not yet emphasized. We assume that the forward and reverse rates of each elementary step are
equal when the overall reaction is at equilibrium. This is a special case of the principle of microscopic reversibility
Definition: Principle of Microscopic Reversibility

Any molecular process and its reverse occur with equal rates at equilibrium
The principle of microscopic reversibility applies to any molecular process; it is inferred from the fact that such processes can be
described by their equations of motion if the initial state of the constituent particles can be specified. The equations of motion can
be either classical mechanical or quantum mechanical. We consider the implications of the principle for molecular processes that
constitute elementary reactions. However, the principle also applies to equilibria in other molecular processes, notably the
absorption and emission of radiation.
When we apply it to elementary reactions, we see that the principle of microscopic reversibility provides a necessary and sufficient
condition for equilibrium from a reaction-mechanism perspective. The principle also imposes several significant conditions on the
sequences of elementary processes that constitute a mechanism and on their relative rates.
In the previous section, we see that microscopic reversibility provides a sufficient basis for deducing the relationship relating
reactant and product concentrations at equilibrium—the equilibrium constant expression—from our rate equations for elementary
reactions. We now want to see that the principle of microscopic reversibility is indeed necessary. That is, setting dn /dt = 0 for
x
all species, X, involved in the reaction is not in itself sufficient to assure that the system is at equilibrium.
Figure 3. Cyclic equilibrium

We consider the triangular network of elementary reactions 3,4
shown in Figure 3. This network gives rise to the following reaction-
rate equations:
−1
dnA
V = −k1f [A] + k1r [B] + k3f [C ] − k3r [A]
dt
−1
dnB
V = +k1f [A] − k1r [B] + k2r [C ] − k2f [B]
dt
−1
dnC
V = +k2f [B] − k2r [C ] + k3r [A] − k3f [C ]
dt
At equilibrium each of these equations must equal zero. Since we have three equations in three unknowns, it might at first appear
that we can solve for the three unknowns [A], [B], and [C ]. We can see, however, either from the equations themselves or by
considering the physical situation that they represent, that only two of these equations are independent. That is, we have
dnA dnB dnC
+ + =0
dt dt dt
While we cannot solve for [A], [B], and [C ] independently, we can solve for their ratios, which are
[B] k1f k2r + k2r k3r + k1f k3f
=
[A] k2f k3f + k1f k3f + k1r k2r
[C ] k1f k2f + k1r k3r + k2f k3r

=
[B] k1f k2r + k2r k3r + k1f k3f
[A] k2f k3f + k1r k3f + k1r k2r

=
[C ] k1f k2f + k1r k3r + k2f k3r
Since we deduce these equations from the condition that all the time derivatives are zero, it might seem that they should represent
the criteria for the system of reactions to be at equilibrium. Purely as a name for easy reference, let us call these equations the cyclic
equilibrium set.
When we consider the reactions one at a time, we deduce the following equilibrium relationships:
[B] k1f [C ] k2f [A] k3f
= = =
[A] k1r [B] k2r [C ] k3r
For easy reference, let us call these equations the one-at-a-time set.
Now, it cannot be true that both sets of relationships specify a sufficient condition for the system to be at equilibrium. To see this,
let us first suppose that the principle of microscopic reversibility is a sufficient condition for equilibrium. Then the one-at-a-time
set of equations must be sufficient to uniquely specify the position of equilibrium. It is easy to show that a set of rate constants that
satisfies the one-at-a-time set also satisfies the cyclic set. Therefore, if microscopic reversibility is a sufficient condition for
equilibrium, the cyclic network rate equations are necessarily equal to zero at equilibrium. In short, if we assume that microscopic
reversibility is a sufficient condition for equilibrium, we encounter no inconsistencies, because the cyclic set of equations is
satisfied by the same equilibrium-concentration ratios.
On the other hand, if we suppose that setting dn /dt = 0 for all species, X, is a sufficient condition for equilibrium, then the
x
cyclic set of equations must be sufficient to uniquely specify the position of equilibrium. Let us consider a particular set of rate
constants:
k1f = k2f = k3f = 1 and
k1r = k2r = k3r = 2 .
This set of rate constants satisfies the cyclic set of equations and requires that each of the equilibrium-concentration ratios be equal
to 1. In this case, the one-at-a-time set of equations implied by microscopic reversibility cannot be satisfied. (We have [B]/[A] = 1
and k /k = 1/2 . Therefore, [B]/[A] ≠ k /k .) That is, if we assume that setting dn /dt = 0 , for all species, X, is a
1f 1r 1f 1r x
sufficient condition for equilibrium, we must conclude that the principle of microscopic reversibility is false. Using the
contrapositive: If the principle of microscopic reversibility is true, it is false that setting dn /dt = 0 for all species, X, is a
x
sufficient condition for equilibrium.
Figure 4. Potential energy versus reaction coordinate.
Setting the derivatives for the reaction network equal to zero is not sufficient to assure that the system is at equilibrium. It is merely
necessary. To assure that the network is at equilibrium, we must apply the principle of microscopic reversibility and require that
each elementary process in the network be at equilibrium.
The principle of microscopic reversibility requires that any elementary process occur via the same sequence of transitory
molecular structures in both the forward and reverse directions. Consequently, if a sequence of elementary steps is a mechanism for
a forward reaction, the same sequence of steps—traversed backwards—must be a mechanism for the reverse reaction. The
principle does not exclude the possibility that a given reaction can occur simultaneously by two different mechanisms. However, it
does mean that a given reaction cannot have one mechanism in the forward direction and a second, different mechanism in the
reverse direction.
In describing reaction mechanisms, we assume that the energy of the reacting molecules depends on their progress along the path
that they follow during the course of the reaction. We call this path the reaction coordinate. We suppose that we can plot the energy
of the system as a function of the system’s position on the path, or displacement along the reaction coordinate. In the context of
such a graph, the principle of microscopic reversibility is essentially the observation that the path is the same irrespective of the
direction in which it is traversed. Two such paths are sketched in Figure 4. In this sketch, E and E are the activation energies
1f 2f
for the two forward reactions; E and E are the activation energies for the reverse reactions.
1r 2r
This page titled 5.18: The Principle of Microscopic Reversibility is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The principle of microscopic reversibility requires that parallel mechanisms give rise to the same expression for the concentration-
dependence of the equilibrium constant. That is, the function that characterizes the equilibrium composition must be the same for
each mechanism. If, for the reaction aA + bB ⇌ cC + dD , the equilibrium composition for mechanism 1 is [A] , [B] , [C ] ,
1 1 1
[D] , and that for mechanism 2 is [A] , [B] , [C ] , [D] , microscopic reversibility asserts that
1 2 2 2 2
c d
[C ] [D]
1 1
K1 =
a b
[A] [B]
1 1
and
c d
[C ] [D]
2 2
K2 =
a b
[A]2 [B]2
In and of itself, microscopic reversibility makes no assertion about the value of [A] compared to that of [A] . While microscopic
1 2
reversibility asserts that the same function characterizes the concentration relationships for parallel mechanisms, it does not assert
that the numerical value of this function is necessarily the same for each of the mechanisms.
However, that these numerical values must be equal follows directly when we introduce another of our most basic observations. No
matter how many mechanisms may be available to a reaction in a particular system, the concentration of any reagent can have only
one value in an equilibrium state. At equilibrium, [A] = [A] , etc.; therefore, the numerical values of the equilibrium constants
1 2
must be the same: K = K .

1 2
The uniqueness of the equilibrium composition is a fundamental feature of our ideas about what chemical equilibrium means.
Nevertheless, it is of interest to show that we can arrive at this conclusion from a different perspective: We can use an idealized
machine to show that the second law of thermodynamics requires that parallel mechanisms must produce the same the equilibrium
composition. Our argument is a proof by contradiction.
Let us suppose that A , B , and C are gases. Suppose that the reaction A → B + C occurs in the absence of a catalyst, but that
reaction occurs in the opposite direction, C + B → A , when a catalyst is present. More precisely, we assume that the position of
equilibrium A ⇌ B + C lies to the right in the absence of the catalyst and to the left in its presence, while all other reaction
conditions are maintained constant. These assumptions mean that the equilibrium composition for the catalyzed mechanism is
different from that of the mechanism that does not involve the catalyst.
We can show that these assumptions imply that the second law of thermodynamics is false. If we accept the validity of the second
law, this violation means that the assumptions cannot in fact describe any real system. (We are getting a bit ahead of ourselves here,
inasmuch as our detailed consideration of the laws of thermodynamics begins in Chapter 6.)
Given our assumptions, we can build a machine consisting of a large cylinder, closed by a frictionless piston. The cylinder contains
a mixture of A , B , and C , and a quantity of the catalyst. We provide a container for the catalyst, and construct the device so that
the catalyst container can be opened and closed from outside the cylinder. Finally, we immerse the entire cylinder in a fluid, which
we maintain at a constant temperature.
When the catalyst container is sealed, so that the gaseous contents of the cylinder are not in contact with the catalyst, reaction
occurs according to A → B + C , and the piston moves outward, doing work on the surroundings. When the catalyst container is
open, reaction occurs according to C + B → A , and the piston moves inward. Figure 5.19.1 shows these changes schematically.
At the end of a cycle, the machine is in exactly the same state as it was in the beginning, and the temperature of the reaction
mixture is the same at the end of a cycle as it was at the beginning. By connecting the piston to a load, we can do net work on the
load as the machine goes through a cycle. For example, if we connect the piston to a mechanical device that converts the
reciprocating motion of the piston into rotary motion, we can wind a rope around an axle and thereby lift an attached weight.
Figure 5.19.1 : A machine that violates the second law.
We can operate this machine as an engine by alternately opening and closing the catalyst container. We can make the cylinder as
large as we want, so the energy we expend in opening and closing the catalyst container can be made arbitrarily small compared to
the amount of work we get out of the machine in a given cycle. All of this occurs with the machine maintained at a constant
temperature. If energy is conserved, the machine must absorb heat from the bath during the cycle; otherwise, the machine would be
doing work with no offsetting consumption of energy. This would be a violation of the first law of thermodynamics. (See Sections
5.7-5.11.)
From experience, we know that this machine cannot function in the manner we have described. This experience is embodied in the
second law of thermodynamics we know that it is impossible to construct a machine that operates in a cycle, exchanges heat
with its surroundings at only one temperature, and produces work in the surroundings (Section 9-1). Our argument assumes
that two reaction mechanisms are available in a particular physical system, that they consume the same reactants, that they produce
the same products, and that the equilibrium compositions are different. These assumptions imply that the second law is false. Since
we are confident that it is possible for some system to satisfy the first three of these assumptions, the second law requires that the
last one be false: the equilibrium compositions must be the same.
We see that there is a complementary relationship between microscopic reversibility and this statement of the second law.
Microscopic reversibility asserts that a unique function of concentrations characterizes the equilibrium state for any reaction
mechanism, but does not require that every mechanism reach the same state at equilibrium. This statement of the second law
implies that a reaction’s equilibrium composition unique, but it does not specify a law relating the equilibrium concentrations of the
reacting species. (In Chapters 9 and 13, we see that, by augmenting this statement of the second law with some additional ideas, we
are led to a more rigorous statement, from which we are eventually able to infer the same functional form for the equilibrium
constant.)
Microscopic reversibility asserts that a unique function of concentrations characterizes

the equilibrium state for any reaction mechanism, but does not require that every
mechanism reach the same state at equilibrium.
G. N. Lewis gave an early statement of the principle of microscopic reversibility. He called it “the law of entire equilibrium,” and
observed that it is “a law which in its general form is not deducible from thermodynamics, but proves to be compatible with the
laws of thermodynamics in all cases where a comparison is possible.”
It is worth noting that we have not shown that the existence of a unique equilibrium state implies either microscopic reversibility or
the second law. Also, even though the principle of microscopic reversibility is inferred from the laws of mechanics, our
development of the equilibrium constant relationship—which we do view as a law of thermodynamics—depends on our equations
for the rates of elementary reactions. Our rate equations are not logical consequences of the laws of motion. Rather, they follow
from assumptions we make about the average behavior of systems that contain many molecules. Consequently, we should not
suppose that we have deduced a thermodynamic result (the condition for chemical equilibrium) solely from the laws of mechanics.
In Section 12-2, we give brief additional consideration to the relationship between the theories of mechanics and thermodynamics.
Beginning in Chapter 20, we develop thermodynamic equations by applying statistical models to the distribution of molecular
energy levels.
This page titled 5.19: Microscopic Reversibility and the Second Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
5.20: Problems
1. A dimeric molecule, A , dissociates in aqueous solution according to A → 2A . One millimole of
2 2 A2 is dissolved rapidly in
one liter of pure water. After 100 seconds, the concentration of A is 8.0 × 10 ––
M. 2
−4
(a) What is the concentration of A ?

(b) What is the average rate at which A has been formed, in moles per liter per second?
(c) What is the average rate at which A has reacted?
2
(d) What is the average reaction rate?

2. The initial concentration of A in a solution is 10 ––
M . The initial concentrations of B and C are both 10
−2
M . The volume of
––
−3
the solution is 2 L. Reaction occurs according to the stoichiometry: A + 2B → 3C . After 50 seconds, the concentration of C is
M.
−3
1.3 × 10
––
(a) What is the concentration of A ?

(b) What is the concentration of B ?
(c) What is the change in the extent of this reaction during this 50 s period?
(d) What is the average reaction rate?
3. When A dissociates according to A
2 2 → 2A , the observed rate law is
1 dnA k 3/2
=( ) [ A2 ]
V dt 2
(a) What is the order of the reaction in [A ] ? 2
(b) What is the order of the reaction overall?

4. For the reaction A + 2B → 3C , the observed rate law is
2
1 dξ k [A] [B]
=
V dt [C ]
(a) What is the order of the reaction in [A]?

(b) In[B]?
(c) In [C ]?
(d) What is the order of the reaction overall?
5. You deposit $1000 in a bank that pays interest at a 5% annual rate. How much will your account be worth after one year if the
bank compounds interest annually? How much if it compounds interest monthly? Daily? Continuously?
mt
6. We deduced that exp (rt) = lim m→∞ (1 +
r
m
) .
mt
Take r = 0.2 and t = 10 . Calculate exp (rt) . Calculate (1 +
r
m
) for 3
m = 1, 10, 100, 10 , 10 .
4
Do the same for r = −0.2
and t = 10 .
7. Suppose that you invest $10, 000 in the stock market and that your nest egg grows at the rate of 11% per year. (This number, or
something close to it, is often cited as the historical long-term average performance of equities traded on the New York Stock
Exchange.) Assuming continuous compounding, what will be the value of your nest egg at the end of 30 years?
8. Suppose instead that you “invest” your $10, 000 in an automobile. The value of the automobile will most likely decrease with
time, by, say, roughly 20% per year. Assuming continuous decay at this rate, what will be the value of the automobile at the end of
5 years? 30 years?
9. At particular reaction conditions, a compound C decays in a first-order reaction with rate constant 10 s . If the initial −3 −1
concentration of C , is [C ] = 10 mol L , how much does [C ] change in the first second? What is [C ] after 100 seconds? 1000
0
−2 −1
seconds? 2000 seconds? 4000 seconds?
10. In problem 9, how long does it take for one-half of the original concentration of C to disappear? How does the half-life depend
on the initial concentration of C?
11. C is produced continuously in the upper atmosphere. It decays with a half-life of 5715 years. Since this has been going on for
14
a long time, the concentration of C in the atmosphere has reached a steady-state value. Living things continuously exchange
14
carbon with the atmosphere, so the concentration of C (i.e., the fraction of the C that is C ) in the biosphere is the same as it is
14 14
in the atmosphere. When an organism dies, it ceases to exchange carbon with the biosphere, and the concentration of C in its 14
remains begins to decrease. Charcoal found at an ancient campsite during an archeological dig has a C content that is 22% of the
14
atmospheric value.
(a) What is the rate constant, in y −1
, for C 14
decay?
(b) How old is the charcoal?
(c) Reverend Smith tells his parishioners that God created the universe about 4000 B.C. If Smith accepts that the atmospheric
concentration and the decay rate of C have been constant since the time of creation, what would he conclude about the C
14 14
content of the charcoal when God created it?

12. Mordred has introduced an exotic fungus that is growing on the surface of King Arthur’s favorite pond at Camelot. Merlin has
calculated that the area covered by the fungus increases by 10% per day. That is,
d(area)/dt = (0.10/day) × area
Fortunately, a trained Knight of the Round Table can clear 100 m of fungus per day. Arthur has six trained knights who would
2
cheerfully perform this remediation work, but all six are committed to out-of-town dragonslaying activities for the next 10 days.
Merlin says that the fungus covers 2874 m at 8:00 a.m. this morning. (The total area of the pond is about 11, 200 m .) Can Arthur
2 2
wait for the dragon slayers to return, or does he need to develop an alternative effective management action plan?
13. What happens to the balance in a bank account, P (t), if P (0) < 0 ? What do bankers call this sort of account?
14. Suppose that you have an account whose initial balance is −$1000. The bank will continuously compound interest on this
account at the annual rate of 11%. You make continuous payments to this account at a rate of q dollars/year. What must q be if you
want to increase the value of the account to exactly zero at the end of 10 years?
For problems 15 – 18, prove that your conclusion is correct by making an appropriate plot. In each case, the reaction occurs at
constant volume.
15. The reaction A + B → C is studied with a large excess of B. ([A] = 10 –M 0
––
. [B] = 10 –M
−2
0
––
. [C ] = 0.0–M
−1
––
). 0
Concentration versus time data are given in the table below. What is the order of the reaction in the concentration of A, and what is
the rate constant?
Time, s [A], M
–––
−3
100 9.1 × 10
−3
300 7.4 × 10
−3
500 6.1 × 10
−3
800 4.6 × 10
−3
1000 3.6 × 10
−3
1500 2.3 × 10
−3
2000 1.3 × 10
−4
2500 8.3 × 10
16. The following data are collected for a reaction in which A dimerizes: 2A → A . What is the order of the reaction in
2 [A] , and
what is the rate constant?
Time, hr [A], M
––
−2
0.0 1.0 × 10
−3
0.28 9.1 × 10
−3
0.56 8.3 × 10
−3
1.39 6.7 × 10
−3
2.78 5.0 × 10
−3
5.56 3.3 × 10
−3
11.10 2.0 × 10
−3
16.70 1.4 × 10
17. In the reaction A + B → C , the rate at which B is consumed is first-order in [B]. In a series of experiments whose results are
tabulated below, the observed first-order rate constant, k is measured for the disappearance of B in the presence of large
obs
excesses of A . What is the order of the reaction in [A]? The rate law? The rate constant? ([B] = 10 ––
0 M in all experiments.)
−4
−1
[A]0 , M kobs , s
––
−1 −5
2.0 × 10 2.6 × 10
−1 −5
1.1 × 10 1.4 × 10
−2 −6
6.3 × 10 8.2 × 10
−2 −6
2.5 × 10 3.3 × 10
−3 −6
9.1 × 10 1.2 × 10
18. In the reaction A + B → C , the rate at which B is consumed is first-order in [B]. The table below presents first-order rate
constants for the disappearance of B in the presence of large excesses of A . Plot these data to test the hypothesis that
k1 [A]0
kobs =
1 + k2 [A]0
What are the values of k and k ?

1 2
−1
[A]0 , M kobs , s
––
−1 −5
5.0 × 10 8.3 × 10
−1 −5
2.0 × 10 6.7 × 10
−1 −5
1.0 × 10 5.0 × 10
−2 −5
5.0 × 10 3.3 × 10
−2 −5
1.4 × 10 1.2 × 10
−3 −6
7.6 × 10 7.1 × 10
−3 −6
3.0 × 10 3.0 × 10
19. For the reaction A + 2B → C + D , the rate law is

d[C ]
2
= k[A][B]
dt
The volume is constant. Suggest a mechanism for this reaction that does not include a termolecular elementary process. Show that
this mechanism is consistent with the rate law.
20. For the reaction X + 2Y → W +Z , the rate law is
d[W ]dt = k[X]
The volume is constant. Suggest a mechanism for this reaction. Show that this mechanism is consistent with the rate law.
21. For the reaction A
2 + 2B → 2C , the rate law is
d[C ]
1/2
= k[ A2 ] [B]
dt
22. For the reaction AB + C → A+D , the rate law is
\[d[D] dt = \frac{k_u [AB][C]}{k_v [A] +k_w[C]}\)
23. When we use the flooding technique to study a reaction rate, we often say that the concentrations of species present in great
stoichiometric excess are essentially constant. This is a convenient but rather imprecise way to describe a useful approximation.
Consider the reaction A + B → C . Over any time interval, Δt, we have Δ[B] = Δ[A] . In absolute terms, [B] is no more constant
than [A]. Suppose that the reaction rate is described by R = k[A][B] and that [B] = 100[A] . Define the extent of reaction by
0 0
ξ = [A] − [A] = [B] − [B] . Find

0 0
∂R / ∂[B]
∂R / ∂[A]
and evaluate this relative concentration dependence at 0% conversion, ξ = 0 (where [A] = [A] ), and at 90% conversion, 0
ξ = 0.9[A] (where [A] = 0.9[A] ). Give a more precise statement of what we mean when we say that “the concentration of B is
0 0
essentially constant” in such circumstances.

24. For the reaction aA + bB ⇌ cC + dD , we define the extent of reaction ξ = −(n − n )/a . When the reaction reaches
A
o
equilibrium (at t = ∞ ), the extent of reaction becomes ξ = −(n − n ) / a . If A is the limiting reagent and the reaction goes to
∞
A
o
A
completion, the theoretical extent of reaction is ξ = n /a . Why? It is often useful to describe the amount of reaction that
theoretical
0
A
has occurred as a dimensionless fraction. If the reaction does not go to completion, n > 0 . Use ξ and ξ to express the “extent of
∞
0 ∞
equilibration,” f , as a dimensionless fraction. Use ξ and ξ

equilibrium to express the “conversion,” f
theoretical , as a conversion
dimensionless fraction. How would you define the “equilibrium conversion”?

25. We often exercise a degree of poetic license in talking about “fast” and “slow” steps in reaction mechanisms. In Section 5.12,
Case I, for example, we say that the step that consumes A to produce intermediate C is “slow” but the step that consumes C to
produce D is “fast.” We then write −d[A] / dt ≈ d[D] / dt . Discuss.
26. Find the rate law for the simplest-case Michaelis-Menten mechanism by applying the steady-state approximation to the
concentration of the enzyme– substrate complex. Under what conditions do this treatment and the result developed in the text
converge to the same rate law?
27. What is the half-life of a constant-volume secondorder reaction, 2A → C , for which
d[A] 2 dξ
2
=− = −2k[A]
dt V dt
28. For the reaction between oxygen and nitric oxide, 2N O + O 2 → 2N O2 , the observed rate law, at constant volume, is
d[N O2 ]
2
= k[N O] [ O2 ]
dt
Show that this rate law is consistent with either of the following mechanisms:
N O ⇌ N2 O2 (fast equilibrium)
(i)
N2 O2 + O2 → 2N O2 (rate-determining step)
O + O2 ⇌ N O3 (fast equilibrium)
(ii)
N O3 + N O → 2N O2 (rate-determining step)
29. For the reaction between gaseous chlorine and nitric oxide, 2N O + C l 2 → 2N OC l , doubling the nitric oxide concentration
quadruples the rate, and doubling the chlorine concentration doubles the rate.
(a) Deduce the rate law for this reaction.
(b) Keeping mind the mechanisms in problem 28, write down two possible mechanisms that are consistent with the rate law you
deduced in part (a). Show that each of these mechanisms is consistent with the rate law in part (a).
30. Nitric oxide reacts with hydrogen according to the equation, 2N O + 2H → N + 2H O . At constant volume, the following
2 2 2
kinetic data have been obtained for this reaction at 1099 K. [ 1 mm = 1 torr = (1⁄760) atm.] C. N. Hinshelwood and T. Green, J.
Chem. Soc., 730 (1926)]
Initial reaction
0 0
P (H2 ), mm P (N O), mm
−1
rate, mm s
289 400 0.162
205 400 0.110
147 400 0.079
400 359 0.150
400 300 0.103
400 152 0.025
(a) What is the rate law for this reaction?

(b) Suggest two mechanisms for this reaction that are consistent with the rate law you deduce in part (a).
31. Review the reactions, rate laws, and mechanisms that you considered in problems 28, 29, and 30.
(a) Does comparing these three reactions and their rate laws provide any basis for preferring one set of mechanisms to the other?
(b) Which set of mechanisms do you prefer; that is, which mechanism in each of problems 29, 30, and 31 seems more likely to
you? Why?
32. The rate of an enzyme-catalyzed
reaction, commonly called the velocity, v , is measured directly as
v = d[P ] / dt ≈ Δ[P ] / Δt = − Δ [S] / Δt . For small S , a plot of v versus. the initial substrate concentration, S , increases
0 0
with increasing S . For large values of S , v reaches a constant value, v ax . The substrate concentration at which the reaction rate
0 0 m
is equal to v / 2 is defined to be the Michaelis constant, K

max . It is customary to express the equilibrium constant as the
M
dissociation constant for the enzyme–substrate complex. Let the total enzyme concentration be E . For the mechanism 0
KS
ES ⇌ E + S KS = [E][S]/[ES]
ES → E + P (rate-determining step)
(a) Show that the velocity is given by

KS
v = kE0 (1 + )
S0
(b) What is vmax ?

(c) What is the Michaelis constant, K ? M
(d) Does a larger value of K M correspond to stronger or weaker complexation of the substrate by the enzyme?
(e) Sketch the curve of v versus. S for the reaction rate described in (a). On this sketch, identify v
0 max ,v
max /2 , and K .
M
(f) If a second substrate, I , can form a complex with the enzyme, the reaction rate for substrate S decreases in the presence of I .
Such substrates, I , are called inhibitors. Many kinds of inhibition are observed. One common distinction is between inhibitors that
are competitive and inhibitors that are not competitive. Competitive inhibition can be explained in terms of a mechanism in which
the enzyme equilibrates with both substrates.
KS
ES ⇌ E + S KS = [E][S] / [ES]
KI
EI ⇌ E + I KI = [E][I ] / [EI ]
ES → E + P (rate-determining step)
Show that the velocity is given by

KS I0
v = kE0 / [1 + (1 + )]
S0 KI
(g) A series of experiments is done in which S is varied, while I is maintained constant. The results are described by the equation
0 0
in (f). What is v maxin this series of experiments?
(h) For the series of experiments done in (g), what is the Michaelis constant, K ?M
33. Consider a bimolecular reaction between molecules of substances A and B . If there are no forces of attraction or repulsion
between A molecules and B molecules, we expect their collision rate to be k[A][B], where k is a constant whose value is
independent of the values of [A] and [B]. Now suppose that molecules of A and B experience a strong attractive force whenever
their intermolecular separation becomes comparable to, say, twice the diameter of an A molecule. Will the value of k be different
when there is a strong force of attraction than when there is no such force?
Notes
1
See Fred Basolo and Ralph G. Pearson, Mechanisms of Inorganic Reactions, 2
nd
Ed., John Wiley & Sons, Inc., New York,
1967, pp 177-193.
2
R.C. Tolman, The Principles of Statistical Thermodynamics, Dover Publications, 1979, (published originally in 1938 by Oxford
University Press), p 163.
3
R. L. Burwell and R. G. Pearson, J. Phys. Chem., 79, 300, (1966).
4
George M. Fleck, Chemical Reaction Mechanisms, Holt, Rinehard, and Winston, Inc., New York, NY, 1971, pp 104-112.
5
G. N. Lewis, Proc. Nat. Acad. Sci. U. S., 11, 179 (1925).
CHAPTER OVERVIEW
6: Equilibrium States and Reversible Processes

6.16: Problems
Thumbnail: Illustration of a system exhibiting an irreversible process as rapid and slow particles mix together. (CC BY 2.5
Generic; Htkym and Dhollm via Wikipedia)
This page titled 6: Equilibrium States and Reversible Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
1
Classical thermodynamics does not consider the atomic and molecular characteristics of matter. In developing it, we focus
exclusively on the measurable properties of macroscopic quantities of matter. In particular, we study the relationship between the
thermodynamic functions that characterize a system and the increments of heat and work that the system receives as it undergoes
some change of state. In doing so, we adopt some particular perspectives. The first is to imagine that we can segregate the
macroscopic sample that we want to study from the rest of the universe. As sketched in Figure 1, we suppose that we can divide the
universe into two mutually exclusive pieces: the system that we are studying and the surroundings, which we take to encompass
everything else.
Figure 1. The thermodynamic universe.

We imagine the system to be enclosed by a boundary, which may or may not correspond to a material barrier surrounding the
collection of matter that we designate as the system. (For our purposes, a system will always contain a macroscopic quantity of
matter. However, this is not necessary; thermodynamic principles can be applied to a volume that is occupied only by radiant
energy.) Everything inside the boundary is part of the system. Everything outside the boundary is part of the
surroundingssurroundings. Every increment of energy that the system receives, as either heat or work, is passed to it from the
surroundings, and conversely.
An open system can exchange both matter and energy with its surroundings. A closed system can exchange energy but not matter
with its surroundings. An isolated system can exchange neither matter nor energy.
Together, system and surroundings comprise the universe, thermodynamic.
If we are too literal-minded, this reference to “the universe” can start us off on unnecessary ruminations about cosmological
implications. All we really have in mind is an energy-accounting scheme, much like the accountants’ system of double-entry
bookkeeping, in which every debit to one account is a credit to another. When we talk about “the universe,” we are really just
calling attention to the fact that our scheme involves only two accounts. One is labeled “system,” and the other is labeled
“surroundings.” Since we do our bookkeeping one system at a time, the combination of system and surroundings encompasses the
universe of things affected by the change.
Figure 2. Transferring heat and work in a thermodynamic universe.
Figure 2 schematically depicts a closed system that can exchange heat and work with its surroundings. The surroundings comprise
a heat reservoir and a device that can convert potential energy in the surroundings into work exchanged with the system. The heat
reservoir can exchange heat but not work with the system. In this sketch, the heat reservoir is at a constant temperature,
ˆ
T (= T surroundings) . It might comprise, for example, a large quantity of ice and water in phase equilibrium. The work-generating
device cannot exchange heat, but it can exchange work with the system. A partially extended spring represents the potential energy
available in the surroundings. The system can do work on the device and increase the potential energy of the spring. Alternatively,
the surroundings can transfer energy to the system at the expense of the potential energy of the spring. Since nothing else in the rest
of the universe is affected by these exchanges, our sketch encompasses the entire universe insofar as these changes are concerned.
A system that cannot interact with anything external to itself is isolated. The combination of system and surroundings depicted in
Figure 2 is itself an isolated system.
When we deal with the entropy change that accompanies some change in the state of the system, the properties of the surroundings
become important. We develop the reason for this in Chapter 9. It is useful to introduce notation to distinguish properties of the
surroundings from properties of the system. In Figure 2, we indicate the temperatures of system and surroundings by T and T^,
respectively. We adopt this general rule:
When a thermodynamic quantity appears with a superscripted caret, the quantity is that of the surroundings. If there is no
superscripted caret, the quantity is that of the system.
Thus, T^, E
^
, and S^ are the temperature, the energy, and the entropy of the surroundings, respectively, whereas T , E , and S are the
corresponding quantities for the system.
We develop thermodynamics by reasoning about closed chemical systems that consist of one or more homogeneous phases. A
phase can be a solid, a liquid, or a gas. A phase can consist of a single chemical substance, or it can be a homogeneous solution
containing two or more chemical substances. When we say that a phase is homogeneous, we mean that the pressure, temperature,
and composition of the phase are the same in every part of the phase. Since gases are always miscible, a system cannot contain two
gas phases that are in contact with one another. However, multiple solid and immiscible-liquid phases can coexist.
This page titled 6.1: The Thermodynamic Perspective is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
We characterize the system by specifying the values of enough variables so that the system can be exactly replicated. By “exactly
replicated” we mean, of course, that we are not able to distinguish the system from its replicate by any experimental measurement.
Any variable that can be used to characterize the system in this way is called a variable of state, a state variable, or a state
function.
We can say the same thing in slightly different words by saying that the state of a system is completely specified when the values of
all of its state variables are specified. If we initially have a system in some equilibrium state and change one or more of the
variables that characterize it, the system will eventually reach a new equilibrium state, in which some state variables will have
values different from those that characterized the original state. If we want to return the system to its original state, we must arrange
matters so that the value of every state variable is the same as it was originally.
The variables that are associated with a chemical system include pressure, volume, temperature, and the number of moles of each
substance present. All of these variables can be measured directly; that is, every equilibrium state of a system is associated with a
specific value of each of these variables, and this value can be determined without reference to any other state of the system.
Energy and entropy are also variables that are associated with a thermodynamic system. We can only measure changes in energy
and entropy; that is, we can only measure energy and entropy for a process in which a system passes from one state to another.
Other important thermodynamic variables are defined as functions of pressure, volume, temperature, energy and entropy. These
include enthalpy, the Gibbs free energy, the Helmholtz free energy, chemical activity, and the chemical potential. Our goals in
developing the subject of chemical thermodynamics are to define each of these state functions, learn how to measure each of them,
and provide a theory that relates the change that occurs in any one of them to the changes that occur in the others when a chemical
system changes from one state to another.
Any interaction through which a chemical system can exchange work with its surroundings can affect its behavior. Work-producing
forces can involve many phenomena, including gravitational field, electric field, and magnetic fields; surface properties; and sound
(pressure) waves. In Chapter 17, we discuss the work done when an electric current passes through an electrochemical cell.
Otherwise, this book focuses on pressure–volume workpressure–volume work and gives only passing attention to the job of
incorporating other forms of work into the general theory. We include pressure–volume work because it occurs whenever the
volume of a system changes. A thermodynamic theory that did not include volume as a variable would be of limited utility.
Thermodynamic variables can be sorted into two classes in another way. Consider the pressure, temperature, and volume of an
equilibrium system. We can imagine inserting a barrier that divides this original system into two subsystems—without changing
anything else. Each of the subsystems then has the temperature and pressure of the original system; however, the volume of each
subsystem is different from the volume of the original system. We say that temperature and pressure are intensive variables, by
which we mean that the temperature or pressure of an equilibrium system is independent of the size of the system and the same at
any location within the system. Intensive variables stand in contrast to extensive variables. The magnitude of an extensive variable
is directly proportional to the size of the system. Thus, volume is an extensive variable. Energy is an extensive variable. We shall
see that entropy, enthalpy, the Helmholtz free energy, and the Gibbs free energy are extensive variables also.
For any extensive variable, we can create a companion intensive variable by dividing by the size of the system. For example, we
can convert the mass of a homogeneous system into a companion variable, the density, by dividing by the system’s volume. We
will discover that it is useful to define certain partial molar quantities, which have units like energy per mole. Partial molar
quantities are intensive variables. We will find a partial molar quantity that is particularly important in describing chemical
equilibrium. It is called the chemical potential, and since it is a partial molar quantity, the chemical potentialchemical potential is
an intensive thermodynamic variable.
We think of a system as a specific collection of matter containing specified phases. Our goal is to develop mathematical models
(equations) that relate a system’s state functions to one another. A system can be at equilibrium under a great many different
circumstances. We say that the system can have many equilibrium positions. A complete description of all of these equilibrium
positions requires models that can specify how much of each of the substances that make up the system is present in each phase.
However, if a system is at equilibrium, a half-size copy of it is also at equilibrium; whether a system is at equilibrium can be
specified without specifying the sizes of the phases that make it up. This means that we can characterize the equilibrium states of
any system that contains specified substances and phases by specifying the values of the system’s intensive variables. In general,
not all of these intensive variables will be independent. The number of intensive variables that are independent is called the number
of degrees of freedom available to the system. This is also the number of intensive variables that can change independently while a
given system remains at equilibrium.
To completely define a particular system we must specify the size and composition of each phase. To do so, we must specify the
values of some number of extensive variables. These extensive variables can change while all of the intensive variables remain
constant and the system remains at equilibrium. In the next section, we review the phase equilibria of water. A system comprised of
liquid and gaseous water in phase equilibrium illustrates these points. Specifying either the pressure or temperature specifies the
equilibrium state to within the sizes of the two phases. For a complete description, we must specify the number of moles of water in
each phase. By adding or removing heat, while maintaining the original pressure and temperature, we can change the distribution of
the water between the two phases.
In 1875, J. Willard Gibbs developed an equation, called Gibbs’ phase rule, from which we can calculate the number of degrees of
freedom available to any particular system. We introduce Gibbs’ phase rule in Section 6.8. The perspective and analysis that
underlie Gibbs’ phase rule have a significance that transcends use of the rule to find the number of degrees of freedom available to
a system. In essence, the conditions assumed in deriving Gibbs’ phase rule define what we mean by equilibrium in chemical
systems. From experience, we are usually confident that we know when a system is at equilibrium and when it is not. One of our
goals is to relate thermodynamic functions to our experience-based ideas about what equilibrium is and is not. To do so, we need to
introduce the idea of a reversible process, in which the system undergoes a reversible change.
We will see that the states that are accessible to a system that is at equilibrium in terms of Gibbs’ phase rule are identically the
states that the system can be in while undergoing a reversible change. A principal goal of the remainder of this chapter is to clarify
this equivalence between the range of states accessible to the system at equilibrium and the possible paths along which the system
can undergo reversible change.
The thermodynamic theory that we develop predicts quantitatively how a system’s equilibrium position changes in response to a
change that we impose on one or more of its state functions. The principle of Le Chatelier makes qualitative predictions about
such changes. We introduce the principle of Le Chatelier and its applications later in this chapter. In Chapter 12, we revisit this
principle to understand it as a restatement, in qualitative terms, of the thermodynamic criteria for equilibrium.
This page titled 6.2: Thermodynamic Systems and Variables is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
To review the general characteristics of phase equilibria, let us consider a familiar system. Suppose that we have a transparent but
very strong cylinder, sealed with a frictionless piston, within which we have trapped a quantity of pure liquid water at some high
pressure. We can fix the pressure of the liquid water at any value we choose by applying an appropriate force to the piston. Suppose
that we hold the temperature constant and force the volume to increase by withdrawing the piston in very small increments.
Because pure water is not compressed easily, we find initially that the pressure of the water decreases and does so in very large
increments.
However, after some small increase in the volume, we find that imposing a further volume increase changes the system’s behavior
abruptly. The system undergoes a profound change. What was formerly pure liquid becomes a mixture of liquid and gas. As we
impose still further volume increases, the pressure of the system remains constant, additional liquid passes from the liquid to the
gas phase, and we find that we must supply substantial amounts of heat in order to keep the temperature of the system constant. If
we continue to force volume increases in this manner, vaporization continues until all of the liquid evaporates.
If we impose a decrease in the volume of the two-phase system, we see the process reverse. The pressure of the system remains
constant, some of the gas condenses to liquid, and the system gives up heat to the surroundings. For any given temperature, these
conversions are precisely balanced at some particular pressure, and these conditions characterize a state of liquid–vapor
equilibrium. At any given pressure, the equilibrium temperature is called the boiling point of the liquid. The equilibrium pressure
and temperature completely specify the state of the system, except for the exact amounts of liquid and gaseous water present.
If we begin with this system in a state of liquid–vapor equilibrium, we can increase the amount of vapor by imposing a small
volume increase. Conversely, we can decrease the amount of vapor by imposing a very small volume decrease. At the equilibrium
temperature and pressure, changing the imposed volume by an arbitrarily small amount (from V to V ± dV ) is sufficient to reverse
the direction of the change that occurs in the system. We call any process whose direction can be reversed by an arbitrarily small
change in a thermodynamic state function a reversible process. Evidently, there is a close connection between reversible
processes and equilibrium states. If a process is to occur reversibly, the system must pass continuously from one equilibrium
state to another.
In this description, the reversible, constant-temperature vaporization of water is driven by arbitrarily small volume changes. The
system responds to these imposed volume changes so as to maintain a constant equilibrium vapor pressure at the specified
temperature. We say that the reversible process “takes place at constant pressure and temperature.” We can also describe this
process as being driven by arbitrarily small changes in the applied pressure: If the applied pressure exceeds the equilibrium vapor
pressure by an arbitrarily small increment, dP > 0 , condensation occurs; if the applied pressure is less than the equilibrium vapor
pressure by an arbitrarily small increment, dP < 0 , vaporization occurs. To describe this tersely, we introduce a figure of speech
and say that the reversible process occurs “while the system pressure and the applied pressure are equal.” Literally, of course, there
can be no change when these pressures are equal.
To cause water to vaporize at a constant temperature and pressure, we must add heat energy to the system. This heat is called the
latent heat of vaporization or the enthalpy of vaporization, and it must be supplied from some entity in the surroundings. When
water vapor condenses, this latent heat must be removed from the system and taken up by the surroundings. (The enthalpy change
for vaporizing one mole of a substance is usually denoted Δ H. It varies with temperature and pressure. Tables usually give
vap
experimental values of the equilibrium boiling temperature at a pressure of 1 bar or 1 atm; then they give the enthalpy of
vaporization at this temperature and pressure. We discuss the enthalpy function in Chapter 8.)
Four conditions are sufficient to exactly specify either the initial or the final state: the number of moles of liquid, the number of
moles of gas, the pressure, and the temperature. The change is a conversion of some liquid to gas, or vice versa. We can represent
this change as a transition from an initial state to a final state where n o
and n are the initial numbers of moles of liquid and
liquid
o
gas
gas, respectively, and δn is the incremental number of moles vaporized:

o o o o
(P , T , n , ngas ) → (P , T , n − δn, ngas + δn)
liquid liquid
The initial pressure and temperature are the same as the final pressure and temperature. Effecting this change requires that a
quantity of heat, (Δ H ) δn, be added to the system, without changing the temperature of the system.
vap
This introduces another requirement that a reversible process must satisfy. If the reversibly vaporizing water is to take up an
arbitrarily small amount of heat, the system must be in contact with surroundings that are hotter than the system. The temperature
difference between the system and its surroundings must be arbitrarily small, because we can describe exactly the same process as
being driven by contacting the system, at temperature T , with surroundings at temperature T^ + δT^ . If we keep the applied
pressure constant at the temperature-T equilibrium vapor pressure, the system volume increases. We can reverse the direction of
change by changing the temperature of the surroundings from T^ + δT^ to T^ − δT^ . If the process is to satisfy our criterion for
reversibility, the difference between these two temperatures must be arbitrarily small. To describe this requirement tersely, we again
introduce a figure of speech and say that the reversible process occurs “while the system temperature and the surroundings
temperature are equal.”
If we repeat the water-in-cylinder experiment with the temperature held constant at a slightly different value, we get similar results.
There is again a pressure at which the process of converting liquid to vapor is at equilibrium. At this temperature and pressure, both
liquid and gaseous water can be present in the system, and, so long as no heat is added or removed from the system, the amount of
each remains constant. When we hold the pressure of the system constant at the equilibrium value and supply a quantity of heat to
the system, a quantity of liquid is again converted to gaseous water. (The quantity of heat required to convert one mole of liquid to
gaseous water, phase equilibria is slightly different from the quantity required in the previous experiment. This is what we mean
when we say that the enthalpy of vaporization varies with temperature.)
This experiment can be repeated for many temperatures. So long as the temperature is in the range \(\mathrm{273.16}, we find a
pressure at which liquid and gaseous water are in equilibrium. If we plot the results, they lie on a smooth curve, which is sketched
in Figure 3. This curve represents the combinations of pressure and temperature at which liquid water and gaseous water are in
equilibrium.
Figure 3. Water liquid–vapor equilibrium.

Below 273.16 K, an equilibrium system containing only liquid and gaseous water cannot exist. At high pressures, a two-phase
equilibrium system contains solid and liquid; at sufficiently low pressures, it contains solid and gas. Above 647.1 K, the distinction
between liquid and gaseous water vanishes. The water exists as a single dense phase. This is the critical temperature. Above the
critical temperature, there is a single fluid phase at any pressure.
If we keep the pressure constant and remove heat from a quantity of liquid water, the temperature decreases until we eventually
reach a temperature at which the water begins to freeze to ice. At this point, water and ice are in equilibrium. Further removal of
heat does not decrease the temperature of the water–ice system; rather, the temperature remains constant and additional water
freezes into ice. Only when all of the liquid has frozen does further removal of heat cause a further decrease in the temperature of
the system. When we repeat this experiment at a series of temperatures, we find a continuous line of pressure–temperature points
that are liquid–ice equilibrium points.
Figure 4. The phase diagram for water. Note that the ordinate values for the solid – liquid and liquid – gas equilibria are severely
compressed. 1
As sketched in Figure 4, the liquid–ice equilibrium line intersects the liquid–vapor equilibrium line. At this intersection, liquid
water, ice, and water vapor are all in equilibrium with one another. There is only one such point. It is called the triple point of
water. The ice point or melting point of water is the temperature at which solid and liquid water are in equilibrium at one
atmosphere in the presence of air. The water contains dissolved air. The triple point occurs in a pure-water system; it is the
temperature and pressure at which gaseous, liquid, and solid water are in equilibrium. By definition, the triple point temperature is
273.16 K. Experimentally, the pressure at the triple point is 611 Pa. Experimentally, the melting point is 273.15 K.
To freeze a liquid, we must remove heat. To fuse (melt) the same amount of the solid at the same temperature and pressure, we
must add the same amount of heat. This heat is called the latent heat of fusion or the enthalpy of fusion. The enthalpy of fusion for
one mole of a substance is usually denoted Δ H. It varies slightly with temperature and pressure. Tables usually give
f us
experimental values of the equilibrium melting temperature at a pressure of 1 bar or 1 atm; then they give the enthalpy of fusion at
this temperature and pressure.
At low pressures and temperatures, ice is in equilibrium with gaseous water. A continuous line of pressure–temperature points
represents the conditions under which the system contains only ice and water vapor. As the temperature increases, the ice–vapor
equilibrium line ends at the triple point. The conversion of a solid directly into its vapor is called sublimation. To sublime a solid to
its vapor requires the addition of heat. This heat is called the latent heat of sublimation or the enthalpy of sublimation. The
enthalpy of sublimation for one mole of a substance is usually denoted Δ H . It varies slightly with temperature and pressure.
sub
This page titled 6.3: Equilibrium and Reversibility - Phase Equilibria is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
A system can contain more than one phase, and more than one chemical substance can be present in each phase. If one of the
substances is present in two phases, we say that the substance is distributed between the two phases. We can describe the
equilibrium distribution quantitatively by specifying the concentration of the substance in each phase. At constant temperature, we
find experimentally that the ratio of these concentrations is approximately constant. Letting A be the substance that is distributed,
we find for the distribution equilibrium
A (phase 1) ⇌ A (phase 2)
the equilibrium constant

[A]
phase 2
K =
[A]
phase 1
where K varies with temperature and pressure.

For example, iodineiodine is slightly soluble in water and much more soluble in chloroform. Since water and chloroform are
essentially immiscible, a system containing water, chloroform, and iodine will contain two liquid phases. If there is not enough
iodine present to make a saturated solution with both liquids, the system will reach equilibrium with all of the iodine dissolved in
the two immiscible solvents. Experimentally, the equilibrium concentration ratio
[ I2 ]
water
K =
[ I2 ]
chlorof orm
is approximately constant, whatever amounts of the three substances are mixed.

We begin our development of physical chemistry by reasoning about the effects of concentrations on the properties of chemical
systems. In Chapter 5, we find that rate laws expressed using concentration variables are adequate for the analysis of reaction
mechanisms. Consideration of these rate laws leads us to the equilibrium constant for a chemical reaction expressed as a function
of concentrations. Eventually, however, we discover that an adequately accurate theory of chemical equilibrium must be expressed
using new quantities, which we call chemical activities . We can think of a chemical activity as an “effective concentration” or a
2
“corrected concentration,” where the correction is for the effects of intermolecular interactions. When we allow for the effects of
intermolecular interactions, we find that we must replace the concentration terms by chemical activities. For the distribution
equilibrium constant, we have
~
aA,phase 2
K =
~
aA,phase 1
where a
~
A,phase 1 denotes the chemical activity of species A , in phase 1, at equilibrium.
This page titled 6.4: Distribution Equilibria is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
Equilibria involving chemical reactions share important characteristics with phase and distribution equilibria. In Chapter 5, we
develop the equilibrium constant expression from ideas about reaction rates. For the present comparison, let us consider the
equilibrium between the gases nitrogen dioxide, N O , and dinitrogen tetroxide, N O :
2 2 4
N O (g) −
↽⇀
− 2 NO (g)
2 4 2
Suppose that we trap a quantity of pure N O in a cylinder closed with a piston. If we fix the temperature and volume of this
2 4
system, the dissociation reaction occurs until equilibrium is achieved at some system pressure. For present purposes, let us assume
that both N O and N O behave as ideal gases. The equilibrium system pressure will be equal to the sum of the partial pressures:
2 4 2
P =P +P
N2 O4 . If we now do a series of experiments, in which we hold the volume constant while allowing the temperature
N O2
to change, we find a continuous series of pressure–temperature combinations at which the system is at equilibrium. This curve is
sketched in Figure 5. It is much like the curve describing the dependence of the water–water-vapor equilibrium on pressure and
temperature.
Figure 5. System pressure versus temperature for N 2 O4 dissociation.

If we hold the temperature constant and allow the volume to vary, we can change the force on the piston to keep the total pressure
constant at a new value, P . The position of the chemical equilibrium will change. At the new equilibrium position, the new
∗
total
N O
2 4and N O partial pressures will satisfy the total pressure relationship. When we repeat this experiment, we find that,
2
whatever the total pressure, the equilibrium partial pressures are related to one another as sketched in Figure 6.
Figure 6. Equilibrium compositions for N 2 O4 dissociation at a fixed temperature.

We find that the experimental data fit the equation
2
P
N O2
KP = ,
PN2 O4
where K is the equilibrium constant for the reaction. (Pressure is a measure of gas concentration. Later, we see that the
P
equilibrium constant can be expressed more rigorously as a ratio of fugacities—or activities.)
This page titled 6.5: Equilibria in Chemical Reactions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
In Chapter 1, we describe a very general goal: given that we create a system in some arbitrary initial state (by some “change of
conditions” or “removal of some constraint”), we want to predict how the system will respond as it changes on its own
(“spontaneously”) to some new equilibrium state. Under these circumstances, we need a lot of information about the system before
we can make any useful prediction about the spontaneous change. In this case, we have said nothing about the condition of the
system before we effect the change of conditions that creates the arbitrary “initial state”.
Our ability to make useful predictions is much greater if the system and the change of conditions have a particular character. If we
start with a system that is at equilibrium, and we impose a change in conditions on it, the “initial” state of the system after the
imposed change of conditions will generally not be an equilibrium state. Experience shows that the system will undergo some
spontaneous change to arrive at a new equilibrium state. In these particular circumstances, Le Chatelier’s principle enables us to
predict the spontaneous change that occurs.
Definition: Le Chatelier’s principle

If a change is imposed on the state of a system at equilibrium, the system adjusts to reach a new equilibrium state. In doing so,
the system undergoes a spontaneous change that opposes the imposed change.
Le Chatelier’s principle is useful, and it is worthwhile to learn to apply it. The principle places no limitations on the nature of the
imposed change or on the number of thermodynamic variables that might change as the system responds. However, since our
reasoning based on the principle is qualitative, it is frequently useful to suppose that the imposed change is made in just one
variable and that the opposing change involves just one other variable. That is, we ask how changing one of the variables that
characterizes the equilibrated system changes a second such variable, “all else being equal.” Successful use of the principle often
requires careful thinking about the variable on which change is imposed and the one whose value changes in response. Let us
consider some applications.
Vapor–liquid equilibrium
Vapor–liquid equilibrium. Suppose that we have a sealed vial that contains only the liquid and vapor phases of a pure compound.
We suppose that the vial and its contents are at a single temperature and that the liquid and the vapor are in equilibrium with one
another at this temperature. What will happen if we now thermostat the vial at some new and greater temperature?
We see that the imposed change is an increase in temperature or, equivalently, an addition of heat to the system. The system cannot
respond by decreasing its temperature, because the temperature change is the imposed change. Similarly, it cannot respond by
changing its volume, because the system volume is fixed. Evidently, the observable consequence of increasing temperature—
adding heat— must be a change in the pressure of the system. The principle asserts that the system will respond so as to consume
heat. Converting liquid to vapor consumes the latent heat of vaporization, so the system can oppose the imposed addition of heat by
converting liquid to vapor. This increases the pressure of the vapor. We can conclude from Le Chatelier’s principle that increasing
the temperature of a system at liquid-vapor equilibrium increases the equilibrium vapor pressure.
Now suppose that we have the liquid and vapor phases of the same pure compound in a thermally isolated cylinder that is closed by
a piston. We ask what will happen if we decrease the volume. That is, the imposed change is a step decrease in volume,
accompanied by an increase in pressure. The new volume is fixed, but the pressure is free to adjust to a new value at the new
equilibrium position. The principle asserts that the system will respond so as to decrease its pressure. Decreasing the system
pressure is accomplished by condensing vapor to liquid, which is accompanied by the release of the latent heat of vaporization.
Since we suppose that the system is thermally isolated during this process, the heat released must result in an increase in the
temperature of the system. While the pressure can decrease from the initial non-equilibrium value, it cannot decrease to its original-
equilibrium value; evidently, the new equilibrium pressure must be greater than the original pressure.
We again conclude that an increase in the equilibrium vapor pressure requires an increase in the temperature of the system. (If the
volume decrease were imposed with the system immersed in a constant temperature bath, the heat evolved would be transferred
from the system to the bath. The system would return to its original pressure and original temperature, albeit with fewer moles of
the substance present in the gas phase.)
Ice–water equilibrium
Suppose that we have a closed system consisting of ice in equilibrium with liquid water at some temperature and pressure. What
will happen if we impose an increase in the temperature this system? We suppose that the system occupies a container of fixed
volume. Initially it is at equilibrium with a constant-temperature bath. We impose the change by moving the container to a new bath
whose temperature is higher—but not high enough to melt all of the ice. The imposed change is a temperature increase or,
equivalently, an addition of heat. The principle asserts that the system will respond by consuming heat, which it can do by
converting ice to liquid. Since liquid water occupies less volume than the same mass of ice, the system pressure will decrease. We
conclude that the pressure at which ice and water are at equilibrium decreases when the temperature increases. That is, the melting
point increases as the pressure decreases.
Again, we can imagine that the equilibrium mixture of ice and water is contained in a thermally isolated cylinder that is closed by a
piston and ask how the system must respond if we impose a step decrease in its volume. We impose the volume decrease by
applying additional force to the piston. The imposed step change in the volume is accompanied by an increase in the system
pressure; the new volume is fixed, but the system pressure can adjust. The principle asserts that the system will respond by
decreasing its pressure. The system pressure will decrease if some of the ice melts. Melting ice consumes heat. Since we are now
assuming that the system is thermally isolated, this heat cannot come from outside the system, which means that the temperature of
the system must decrease. While the pressure can decrease from its initial non-equilibrium value, it cannot decrease to the value
that it had in the original equilibrium position. We again conclude that increasing the pressure results in a decrease in temperature;
that is, the melting point of ice increases as the pressure decreases.
Chemical reaction between gases

Chemical reaction between gases. Finally, suppose that we have a chemical equilibrium involving gaseous reagents. To be specific,
let us again consider the reaction
N2 O4 (g) ⇌ 2 N O2 (g)
We suppose that this system is initially at equilibrium at some temperature and that we seek to increase the pressure while
maintaining the temperature constant. (We can imagine that the system is contained in a cylinder that is closed by a piston. The
cylinder is immersed in a constant-temperature bath. We increase the pressure by applying additional force to the piston. As in the
examples above, we view this as a step change in volume that is accompanied by an increase of the pressure to a transitory non-
equilibrium value.) The principle asserts that the system will respond by undergoing a change that opposes this pressure increase.
The system can reduce its pressure by decreasing the number of moles of gas present, and it can do this by converting N O 2
molecules to N O molecules. We conclude that there will be less N O present at equilibrium at the higher pressure.
2 4 2
When we first encounter it, Le Chatelier’s principle seems to embody a remarkable insight. As, indeed, it does. However, as we
think about it, we come to see it as a logical necessity. Suppose that the response of an equilibrium system to an imposed change
were to augment the change rather than oppose it. Then an imposed change would reinforce itself. The slightest perturbation of any
equilibrium system would cause the system to “run away” from that original position. Since no real system can be maintained at a
perfectly constant set of conditions, any real system could undergo spontaneous change. Equilibrium would be unattainable. If we
assume that a system must behave oppositely to the way that is predicted by Le Chatelier’s principle, we arrive at a prediction that
contradicts our experience.
Le Chatelier’s principle is inherently qualitative. We will discuss it further after we develop the thermodynamic criteria for
equilibrium. We will find that the thermodynamic criteria for equilibrium tell us quantitatively how two (or more) thermodynamic
variables must change in concert if a system is to remain at equilibrium while also undergoing some change of condition.
This page titled 6.6: Le Chatelier's Principle is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
If we are to model a physical system mathematically, we must abstract measurable properties from it—properties that we can treat
as variables in our model. In Section 6.2 we found that the size of the system does not matter when we consider the variables that
specify an equilibrium state. A half-size version of an equilibrium system has the same equilibrium properties. We can say that only
intensive properties are relevant to the question of whether a system is at equilibrium.
The idea that we can subdivide a system without changing its equilibrium properties is subject to an important qualification. We
intend that both subsystems be qualitatively equivalent to the original. For example, if we divide a system at vapor–liquid
equilibrium into subsystems, each subsystem must contain some liquid and some vapor. If we subdivide it into one subsystem that
is all liquid and another that is all vapor, the subsystems are not qualitatively equivalent to the original.
We can be more precise about the criterion we have in mind: An equilibrium system consists of one or more homogenous phases.
Two systems can be in the same equilibrium condition only if all of the phases present in one are also present in the other. If a
process changes the number of phases present in a system, we consider that the system changes from one kind of equilibrium
system to a second one. We can describe one kind of equilibrium system by specifying a sufficient number of intensive variables.
This description will be complete to within a specification of the exact amount of each phase present.
If we apply these ideas to a macroscopic sample of a pure gas, we know that we need four variables to completely describe the state
of the gas: the number of moles of the gas, its pressure, its volume, and its temperature. This assumes that we are not interested in
the motion of the container that contains the gas. It assumes also that no other extrinsic factors—like gravitational, electric, or
magnetic fields— affect the behavior that we propose to model
When we do experiments in which the amount, pressure, volume, and temperature of a pure gas vary, we find that we can develop
an equation that relates the values that we measure. We call this an equation of state, because it is a mathematical model that
describes the state of the system. In Chapter 2, we reviewed the ideal gas equation, van der Waals equation, and the virial equation;
however, we can devise many others. Whatever equation of state we develop, we know that it must have a particular property: At
constant pressure and temperature, the volume must be directly proportional to the number of moles. This means that any equation
of state can be rewritten as a function of concentration. For the case of an ideal gas, we have P = (n/V ) RT , where n/V , the
number of moles per unit volume, is the gas concentration. We see that any equation of state can be expressed as a function of three
intensive variables: pressure, temperature, and concentration.
The existence of an equation of state means that only two of the three intensive variables that describe the gas sample are
independent of one another. At equilibrium, a sample of pure gas has two degrees of freedom. Viewed as a statement about the
mathematical model, this is true because knowledge of the equation of state and any two of the intensive variables enables us to
calculate the third variable. Viewed as a statement about our experimental observations, this is true because, so long as the changes
are consistent with the system remaining a gas, we can change any two of these variables independently. That only two are
independent is shown experimentally by the observation that we can start with a fixed quantity of gas at any pressure, temperature,
and concentration and find, after taking the system through any sequence of changes whatsoever, that returning to the original
pressure and temperature also restores the original concentration.
In the experiment or in the mathematical model, fixing two of the three intensive variables is sufficient to fix the equilibrium
properties of the system. Fixing the equilibrium properties means, of course, that the state of the system is fixed to within an
arbitrary factor, which can be specified either as the number of moles present or as the system volume.
Similar results are obtained when we study the pressure–volume–temperature behavior of pure substances in condensed phases. At
equilibrium, a pure liquid or a pure solid has two degrees of freedom.
If we consider a homogeneous mixture of two non-reacting gases, we discover that three variables are necessary to fix the
equilibrium properties of the system. We must know the pressure and temperature of the system and the concentration of each gas.
Because the mixture must obey an equation of state, determination of any three of these variables is sufficient to fix the value of the
fourth. Note that we can conclude that three intensive variables are sufficient to determine the equilibrium properties of the system
even if we do not have a mathematical model for the equation of state.
If we experiment with a system in which the liquid and vapor of a pure substance are in phase equilibrium with one another, we
find that there is only one independent intensive variable. (Figure 4 illustrates this for water.) To maintain phase equilibrium, the
system pressure must be the equilibrium vapor pressure of the substance at the system temperature. If we keep the pressure and
temperature constant at equilibrium values, we can increase or decrease the concentration (moles per unit system volume) by
removing or adding heat. In this process, we change one variable, concentration, while maintaining phase equilibrium.
If we keep the pressure constant and impose a temperature increase, vaporization continues (the concentration decreases) until the
liquid phase is completely consumed. In this process, two variables change, and phase equilibrium cannot be maintained. To reach
a new equilibrium state in which both liquid and gas are present at a higher temperature, we must increase the pressure to the new
equilibrium vapor pressure; the magnitude of the temperature increase completely determines the required pressure increase. Two
intensive variables change, but the changes are not independent.
If we have pure gas, there are two independent intensive variables. If we have pure liquid, there are two independent intensive
variables. However, if we have liquid and gas in equilibrium with one another, there is only one independent intensive variable. In
the liquid region of the water phase diagram, we can vary pressure and temperature and the system remains liquid water. Along the
liquid-gas equilibrium line, we can vary the temperature and remain at equilibrium only if we simultaneously vary the pressure so
as to remain on the liquid-gas equilibrium line.
Similar statements apply if we contrast varying pressure and temperature for the pure solid to varying the pressure and temperature
along the solid-liquid or the solid-gas equilibrium line. At the triple point, nothing is variable. For a fixed quantity of water, the
requirement that the system be at equilibrium at the triple point fixes the system pressure, temperature, and concentration.
Evidently, maintaining a phase equilibrium in a system imposes a constraint that reduces the number of intensive variables that we
can control independently.
The equilibrium between water and ice is completely unaffected by the state of subdivision of the ice. The ice can be present in a
single lump or as a large number of very small pieces; from experience, we know that the equilibrium behavior of the system is the
same so long as some ice and some water are both present. A system contains as many phases as there are kinds of macroscopic,
homogeneous, bounded portions that are either solid, liquid, or gas.
If we add a lump of pure aluminum to our ice-water system, the new system contains three phases: water, ice, and aluminum. The
equilibrium properties of the new system are the same if the aluminum is added as a ground-up powder. The powder contains many
macroscopic, homogeneous, bounded portions that are aluminum, but each of these portions has the same composition; there is
only one kind of aluminum particle. (Molecules on the surface of a substance can behave differently from those in the bulk. When a
substance is very finely divided, the fraction of the molecules that is on the surface can become large enough to have a significant
effect on the behavior of the system. In this book, we do not consider systems whose behavior is surface-area dependent.)
This page titled 6.7: The Number of Variables Required to Specify Some Familiar Systems is shared under a CC BY-SA 4.0 license and was
Gibbs found an important relationship among the number of chemical constituents, the number of phases present, and the number
of intensive variables that must be specified in order to characterize an equilibrium system. This number is called the number of
degrees of freedom available to the system and is given the symbol F . By specifying F intensive variables, we can specify the
state of the system—except for the amount of each phase. The number of chemical constituents is called the number of
components and is given the symbol C . The number of components is the smallest number of pure chemical compounds that we
can use to prepare the equilibrium system so that it contains an arbitrary amount of each phase. The number of phases is given the
symbol P . The relationship that Gibbs found between C , P , and F is called Gibbs’ phase rule or just the phase rule. The phase
rule applies to equilibrium systems in which any component can move freely between any two phases in which that component is
present.
We suppose that the state of the system is a continuous function of its state functions. If F , intensive, independent variables, X , 1
X , …, X , are sufficient to specify the state of an equilibrium system, then X + dX , X + dX , …, X

2 F 1 1 2 2 + dX
F Fspecify an
incrementally different equilibrium state of the same system. This means that the number of degrees of freedom is also the number
of intensive variables that can be varied independently while the system changes reversibly—subject to the condition that there is
no change in either the number or kinds of phases present. Moreover, if we keep the system’s intensive variables constant, we can
change the size of any phase without changing the nature of the system. This means that Gibb’s phase rule applies to any
equilibrium system, whether it is open or closed.
A system containing only liquid water contains one component and one phase. By adjusting the temperature and pressure of this
system, we can arrive at a state in which both liquid and solid are present. For present purposes, we think of this as a second
system. Since the second system can be prepared using only liquid water (or, for that matter, only ice) it too contains only one
component. However, since it contains both liquid and solid phases, the second system contains two phases. We see that the
number of components required to prepare a system in such a way that it contains an arbitrary amount of each phase is not affected
by phase equilibria. However, the number of componentsnumber of components is affected by chemical equilibria and by any other
stoichiometric constraints that we impose on the system. The number of components is equal to the number of chemical substances
present in the system, less the number of stoichiometric relationship among these substances.
Let us consider an aqueous system containing dissolved acetic acid, ethanol, and ethyl acetate. For this system to be at equilibrium,
the esterification reaction
CH CO H + CH CH OH −
↽⇀
− CH CO CH CH +H O
3 2 3 2 3 2 2 3 2
must be at equilibrium. In general we can prepare a system like this by mixing any three substances chosen from the set: acetic
acid, ethanol, ethyl acetate, and water. Hence, there are three components. The esterification reaction, or its reverse, then produces
an equilibrium concentration of the fourth substance. However, there is a special case with only two components. Suppose that we
require that the equilibrium concentrations of ethanol and acetic acid be exactly equal. In this case, we can prepare the system by
mixing ethyl acetate and water. Then the stoichiometry of the reaction assures that the concentration condition will be met; indeed,
this is the only way that the equal-concentration condition can be met exactly.
In this example, there are four chemical substances. The esterification reaction places one stoichiometric constraint on the amounts
of these substances that can be present at equilibrium, which means that we can change only three concentrations independently.
The existence of this constraint reduces the number of components from four to three. An additional stipulation that the product
concentrations be equal is a second stoichiometric constraint that reduces the number of independent components to two.
If we have a one-phase system at equilibrium, we see that the pressure, the temperature, and the C component-concentrations
constitute a set of variables that must be related by an equation of state. If we specify all but one of these variables, the remaining
variable is determined, and can be calculated from the equation of state. There are C + 2 variables, but the existence of the
equation of state means that only C + 1 of them can be changed independently. Evidently, the number of degrees of freedom for a
one-phase system is F = C + 1 .
To find the number of degrees of freedom when P such phases are in equilibrium with one another requires a similar but more
extensive analysis. We first consider the number of intensive variables that are required to describe completely a system that
contains C components and P phases, if the phases are not at equilibrium with one another. (Remember that the description we
seek is complete except for a specification of the absolute amount of each phase present. For the characterization of equilibrium
that we seek, these amounts are arbitrary.) In this case, each phase is a subsystem in its own right. Each phase can have a pressure,
a temperature, and a
concentration for each component. Each of these properties can have a value that is independent of its value in any other phase.
There are C + 2 variables for each phase or P (C + 2) variables for all P phases. Table 1 displays these variables.
1 2 … P
Pressure P1 P2 … PP
Temperature T1 T2 … TP
Component 1 n1,1 / V1 n1,2 / V2 … n1,P / VP
Component 2 n2,1 / V1 n2,2 / V2 … n2,P / VP
… … … … …
Component C nC ,1 / V1 nC ,2 / V2 … nC ,P / VP
If the system is at equilibrium, there are numerous relationships among these P (C + 2) variables. We want to know how many
independent relationships there are among them. Each such relationship decreases by one the number of independent intensive
variables that are needed to specify the state of the system when all of the phases are at equilibrium. Let us count these
relationships.
The pressure must be the same in each phase. That is, P = P , P = P , …, P = P , P = P , …, P = P , etc. Since
1 2 1 3 1 P 2 3 2 P
P =P
1 2and P = P implies that P = P , etc., there are only P − 1 independent equations that relate these pressures to one
1 3 2 3
another.
The temperature must be the same in each phase. As for the pressure, there are P − 1 independent relationships among the
temperature values.
The concentration of species A in phase 1 must be in equilibrium with the concentration of species A in phase 2, and so forth.
We can write an equation for phase equilibrium involving the concentration of A in any two phases; for example,
(nA,2 / VA,2 )
K =
(nA,1 / VA,2 )
(In Chapter 14, we will find that this requirement can be stated more rigorously using a thermodynamic function that we call the
chemical potential. At equilibrium, the chemical potential of species A must be the same in each phase.) For the P phases, there
are again P − 1 independent relationships among the component-A concentration values. This is true for each of the C
components, so the total number of independent relationships among the concentrations is C (P − 1) .
While every component need not be present in each phase, there must be a finite amount of each phase present. Each phase must
have a non-zero volume. To express this requirement using intensive variables, we can say that the sum of the concentrations in
each phase must be greater than zero. For phase 1, we must have
(nA,1 / V1 ) + (nB,1 / V1 ) + ⋯ + (nZ,1 / V1 ) > 0
and so on for each of the P phases. There are P such relationships that are independent of one another.
If we subtract, from the total number of relevant relationships, the number of independent relationships that must be satisfied at
equilibrium, we find Gibbs’ phase rule: There are
F = P (C + 2) − (P − 1) − (P − 1) − C (P − 1) − P
= 2 +C −P
independent relationships or degrees of freedom needed to describe the equilibrium system containing C components and P
phases.
A component may not be present in some particular phase. If this is the case, the total number of relationships is one less than the
number that we used above to derive the phase rule. The number of equilibrium constraints is also one less than the number we
used. Consequently, the absence of a component from any particular phase has no effect on the number of degrees of freedom
available to the system at equilibrium.
This page titled 6.8: Gibbs' Phase Rule is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
When we think about a physical system that is undergoing a reversible change, we imagine that the system passes through a series
of states. In each of these states, every thermodynamic variable has a well-defined value in every phase of the system. We suppose
that successive states of the changing system are arbitrarily close to one another in the sense that the successive values of every
thermodynamic are arbitrarily close to one another. These suppositions are equivalent to assuming that the state of the system and
the value of every thermodynamic variable are continuous functions of time. Then every thermodynamic variable is either constant
or a continuous function of other thermodynamic variables. When we talk about a reversible process, we have in mind a physical
system that behaves in this way and in which an arbitrarily small change in one of the thermodynamic variables can reverse the
direction in which other thermodynamic variables change.
A process that is not reversible is said to be irreversible. We distinguish between two kinds of irreversible processes. A process that
cannot occur under a given set of conditions is said to be an impossible process. A process that can occur, but does not do so
reversibly, is called a possible process or a spontaneous process.
Another essential characteristic of a reversible process is that changes in the system are driven by conditions that are imposed on
the system by the surroundings. In our discussion of the phase equilibria of water, we note that the surroundings can transfer heat to
the system only when the temperature of the surroundings is greater than that of the system. However, if the process is to be
reversible, this temperature difference must be arbitrarily small, so that heat can be made to flow from the system to the
surroundings by an arbitrarily small decrease in the temperature of the surroundings.
Similar considerations apply when the process involves the exchange of work between system and surroundings. We focus on
changes in which the work exchanged between system and surroundings is pressure–volume work. A process can occur reversibly
only if the pressure of the system and the pressure applied to the system by the surroundings differ by an arbitrarily small amount.
To abbreviate these statements, we customarily introduce a figure of speech and say that, for a reversible process,
Tsystem =T surroundings (or T = T^ ) and that P =P
system (or P = P
surroundings ).
applied
Since a reversible process involves a complementary exchange of energy increments between system and surroundings, it is
evident that an isolated system cannot undergo a reversible change. Any change that occurs in an isolated system must be
spontaneous. By the contrapositive, an isolated system that cannot undergo change must be at equilibrium.
While T = T^ and P = P appliedare necessary conditions for a reversible process, they are not sufficient. A spontaneous process
can occur under conditions in which the system temperature is arbitrarily close to the temperature of the surroundings and the
system pressure is arbitrarily close to the applied pressure. Consider a mixture of hydrogen and oxygen in a cylinder closed by a
frictionless piston. We suppose that the surroundings are maintained at a constant temperature and that the surroundings apply a
constant pressure to the piston. We suppose that the system contains a small quantity of a poorly effective catalyst. By controlling
the activity of the catalyst, we can arrange for the formation of water to occur at an arbitrarily slow rate—a rate so slow that the
temperature and pressure gradients that occur in the neighborhood of the catalyst are arbitrarily small. Nevertheless, the reaction is
a spontaneous process, not a reversible one. If the process were reversible, an arbitrarily small increase in the applied pressure
would be sufficient to reverse the direction of reaction, causing water to decompose to hydrogen and oxygen.
This page titled 6.9: Reversible vs. Irreversible Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
We view a chemical system as a collection of substances that occupies some volume. Let us consider a closed system whose
volume is variable, and in which no work other than pressure–volume work is possible. If this system is undergoing a reversible
change, it is at equilibrium, and it is in contact with its surroundings. Because the system is at equilibrium, all points inside the
system have the same pressure and the same temperature. Since the change is reversible, the interior pressure is arbitrarily close to
the pressure applied to the system by the surroundings. If the reversibly changing system can exchange heat with its surroundings,
the temperature of the surroundings is arbitrarily close to the temperature of the system. (If a process takes place in a system that
cannot exchange heat with its surroundings, we say that the process is adiabatic.)
We can measure the pressure, temperature, and volume of such a system without knowing anything about its composition. For a
system composed of a known amount of a single phase of a pure substance, we know from experience that any cyclic change in
pressure or temperature restores the initial volume. That is, for a pure phase, there is an equation of state that we can rearrange as
V = V (P , T ) , meaning that specifying P and T is sufficient to specify V uniquely.
For other reversible systems, the function V = V (P , T ) may not exist. For example, consider a system that consists of a known
amount of water at liquid–vapor equilibrium and whose pressure and temperature are known. For this system, the volume can have
any value between that of the pure liquid and that of the pure gas. Specifying the pressure and temperature of this system is not
sufficient to specify its state. However, if we specify the temperature of this system, the pressure is fixed by the equilibrium
condition; and if we specify the volume of the system, we can find how much water is in each phase from the known molar
volumes of the pure substances at the system pressure and temperature. For the water–water-vapor equilibrium system, we can
write P = P (V , T ) .
In each of these cases, we can view one of the variables as a function of the other two and represent it as a surface in a three
dimensional space. The two independent variables define a plane. Projecting the system’s location in this independent-variable
plane onto the surface establishes the value of the dependent variable. The two independent-variable values determine the point on
the surface that specifies the state of the system. In the liquid–vapor equilibrium system, the pressure is a surface above the
volume–temperature plane.
A complete description of the state of the system must also include the number of moles of liquid and the number of mole of vapor
present. Each of these quantities can also be described as a surface in a three dimensional space in which the other two dimensions
are volume and temperature. Duhem’s theorem asserts that these observations are special cases of a more general truth:
Duhem’s theorem
For a closed, reversible system in which only pressure–volume work is possible, specifying how some pair of state functions
changes is sufficient to specify how the state of the system changes.
Duhem’s theorem asserts that two variables are sufficient to specify the state of the system in the following sense: Given the values
of the system’s thermodynamic variables in some initial state, say {X , Y , Z , W ,…}, specifying the change in some pair of
1 1 1 1
variables, say ΔX and ΔY , is sufficient to determine the change in the remaining variables, ΔZ , ΔW ,… so that the system’s
thermodynamic variables in the final state are {X , Y , Z , W ,…}, where W = W + ΔW , etc. The theorem does not specify
2 2 2 2 2 1
which pair of variables is sufficient. In fact, from the discussion above of the variables that can be used to specify the state of a
system containing only water, it is evident that a particular pair may not remain sufficient if there is a change in the number of
phases present.
In Chapter 10, we see that Duhem’s theorem follows from the first and second laws of thermodynamics, and we consider the
particular pairs of variables that can be used. For now, let us consider a proof of Duhem’s theorem for a system in which the
pressure, temperature, volume, and composition can vary. We consider systems in which only pressure–volume work is possible.
Let the number of chemical species present be C and the number of phases be P . (C , the number of component in the phase rule,
′
and C differ by the number of stoichiometric constraints that apply to the system: C is C less the number of stoichiometric
′ ′
constraints.) We want to know how many variables can be changed independently while the system remains at equilibrium.
This is similar to the question we answered when we developed Gibbs’ phase rule. However, there are important differences. The
phase rule is independent of the size of the system; it specifies the number of intensive variables required to prescribe an
equilibrium state in which specified phases are present. The size of the system is not fixed; we can add or remove matter to change
the size of any phase without changing the number of degrees of freedom. In the present problem, the system cannot exchange
matter with its surroundings. Moreover, the number of phases present can change. We require only that any change be reversible,
and a reversible process can change the number of phases. (For example, reversible vaporization can convert a two-phase system to
a gaseous, one-phase system.)
We want to impose a change on an initial state of a closed system. This initial state is an equilibrium state, and we want to impose a
change that produces a new Gibbsian equilibrium state of the same system. This means that the change we impose can neither
eliminate an existing chemical species nor introduce a new one. A given phase can appear or disappear, but a given chemical
species cannot.
We can find the number of independent variables for this system by an argument similar to the one we used to find the phase rule.
To completely specify this system, we must specify the pressure, temperature, and volume of each phase. We must also specify the
number of moles of each of C chemical species in each phase. This means that P (C + 3) variables must be specified. Every
′ ′
relationship that exists among these variables decreases by one the number that are independent. The following relationships exist:
1. The pressure is the same in each phase. There are P − 1 pressure constraints.
2. The temperature is the same in each phase. There are P − 1 temperature constraints.
3. The volume of each phase is determined by the pressure, the temperature, and the number of moles of each species present in
¯
¯¯¯
that phase. (In Chapter 14, we find that the volume of a phase, V , is given rigorously by the equation V = ∑ n V , where
k k k
¯
¯¯¯ ¯
¯¯¯
n and V are the number of moles and the partial molar volume of the k species in that phase. The V depend only on
th
k k k
pressure, temperature, and composition.) For P phases, there are P constraints, one for the volume of each phase.
4. To completely specify the system, the concentration of each species must be specified in each phase. This condition creates
C P constraints. (We can also reach this conclusion by a slightly different argument. To specify the concentrations of C
′ ′
species in some one phase requires C constraints. A distribution equilibrium relates the concentrations of each species in every
′
pair of phases. There are P − 1 independent pairs of phases. For C chemical species, there are C (P − 1) such constraints.
′ ′
This is equivalent to the requirement in our phase rule analysis that there are C (P − 1) equilibrium relationships among C
′ ′ ′
components in P phases. In the present problem, the total number of concentration constraints is C + C (P − 1) = C P.
Subtracting the number of constraints from the number of variables, we find that there are
′ ′
P (C + 3) − (P − 1) − (P − 1) − P − C P = 2
independent variables for a reversible process in a closed system, if all work is pressure–volume work. The number of independent
variables is constant; it is independent of the species that are present and the number of phases.
It is important to appreciate that there is no conflict between Duhem’s theorem and the phase-rule conclusion that F degrees of
freedom are required to specify an equilibrium state of a system containing specified phases. When we say that specifying some
pair of variables is sufficient to specify the state of a particular closed system undergoing reversible change, we are describing a
system that is continuously at equilibrium as it goes from a first equilibrium state to a second one. Because it is closed and
continuously in an equilibrium state, the range of variation available to the system is circumscribed in such a way that specifying
two variables is sufficient to specify its state. On the other hand, when we say that F degrees of freedom are required to specify an
equilibrium state of a system containing specified phases, we mean that we must know the values of F intensive variables in order
to establish that the state of the system is an equilibrium state.
To illustrate the compatibility of these ideas and the distinction between them, let us consider a closed system that contains
nitrogen, hydrogen, and ammonia gasammonia gases. In the presence of a catalyst, the reaction
N +3 H −
↽⇀
− 2 NH
2 2 3
occurs. For simplicity, let us assume that these gases behave ideally. (If the gases do not behave ideally, the argument remains the
same, but more complex equations are required to express the equilibrium constant and the system pressure as functions of the
molar composition.) This system has two components and three degrees of freedom. When we say that the system is closed, we
mean that the total number of moles of the elements nitrogen and hydrogen are known and constant. Let these be n and n , N H
respectively. Letting the moles of ammonia present be n = x , the number of moles of dihydrogen and dinitrogen are
N H3
nH2 = (n H − 3x)/2 and n N2 = (n N − x)/2 , respectively.
If we know that this system is at equilibrium, we know that the equilibrium constant relationship is satisfied. We have
2 2 −2 −2
P n 2
N H3 N H3 RT 16x RT
KP = = ( ) = ( )
3 3 3
P PN n nN V (nH − 3x) (nN − x) V
H2 2 H2 2
where V is the volume of the system. The ideal-gas equilibrium constant is a function only of temperature. We assume that we
know this function; therefore, if we know the temperature, we know the value of the equilibrium constant. The pressure of the
system can also be expressed as a function of x and V . We have
(nH + nN ) RT
P = PH + PN + PN H =[ − x] ( )
2 2 3
2 V
If we know the system pressure and we know that the system is at equilibrium, we can solve the equations for K and P
simultaneously to find the unknowns x and V . From these, we can calculate the molar composition of the system and the partial
pressure of each of the gases. (We discuss ideal-gas equilibrium calculations in detail in Chapter 13.) Thus, if we know that the
system is at equilibrium, knowledge of the pressure and temperature is sufficient to determine its composition and all of its other
properties.
If we do not know that this system is at equilibrium, but instead want to collect sufficient experimental data to prove that it is, the
phase rule asserts that we must find the values of some set of three intensive variables. Two are not sufficient. From the perspective
provided by the equations developed above, we can no longer use the equilibrium constant relationship to find x and V . Instead,
our problem is to find the composition of the system by other means, so that we can test for equilibrium by comparing the value of
the quantity
2 −2
P 2
N H3 16x RT
= ( )
3 3
P PN (nH − 3x) (nN − x) V
H2 2
to the value of the equilibrium constant. We could accomplish this goal by measuring the values of several different combinations
of three intensive variables. A convenient combination is pressure, temperature, and ammonia concentration, x/V . When we
rearrange the equation for the system pressure to
(nH + nN ) x
P =[ −( )] RT
2V V
it is easy to see that knowing P, T, and x/V enables us to find the volume of the system. Given the volume, we can find the molar
composition of the system and the partial pressure of each of the gases. With these quantities in hand, we can determine whether
the equilibrium condition is satisfied.
This page titled 6.10: Duhem's Theorem - Specifying Reversible Change in A Closed System is shared under a CC BY-SA 4.0 license and was
Let us explore our ideas about reversibility further by considering the familiar case of a bowling ball ball that can move vertically
in the effectively constant gravitational field near the surface of the earth.
We begin by observing that we develop our description by abstracting from reality. We consider idealized models because we want
to develop theories that capture the most important features of real systems. We ignore less important features. In the present
example, we know that the behavior of the bowling ball will be slightly influenced by its frictional interaction with the surrounding
atmosphere. (We attribute these interactions to a property of air that we call viscosity.) We assume that this effect can be ignored.
This causes no difficulty so long as our experiments are too insensitive to observe the effects of this atmospheric drag. If necessary,
of course, we could do our experiments inside a vacuum chamber, so that the system we study experimentally better meets the
assumptions we make in our analysis. Alternatively, we could expand our theory to include the effects of atmospheric drag.
To raise an initially stationary bowling ball to a greater height requires that we apply a vertical upward force that exceeds the
downward gravitational force on the ball. Let height increase in the upward direction, and let h (t) and v (t) be the height and
(vertical) velocity of the ball at time t . Let the mass of the ball be m, and let the ball be at rest at time zero. Representing the initial
velocity and height as v and h , we have v = v (0) = 0 and h = h (0) = 0 . Letting the gravitational acceleration be g, the
0 0 0 0
gravitational force on the ball is f = −mg . To raise the ball, we must apply a vertical force, f
gravitation > 0 , that makes the applied
net force on the ball greater than zero. That is, we require
fnet = fapplied − mg > 0
so that
2
d h
m =f
2
dt net
If f
applied is constant, f
net is constant; we find for the height and velocity of the ball at any later time t ,
fnet
v (t) = ( )t
m
and
2
fnet t
h (t) = ( )
m 2
Let us consider the state of the system when the ball reaches a particular height, hS . Let the corresponding time, velocity, kinetic
energy, and potential energy at h , be t , v , τ , and ʋ , respectively. Since
S S S S S
fnet
vS = ( ) tS
m
and
2
fnet t
S
hS = ( )
m 2
we have
2 2
mv m fnet
S 2
τ (hS ) = = ( ) t = fnet hS
S
2 2 m
The energy we must supply to move the ball from height zero to h is equal to the work done by the surroundings on the ball. The
S
increase in the energy of the ball is −w

^ . At h this input energy is present as the kinetic and potential energy of the ball. We have
S
hS hS
−w
^ =∫ fapplied dh = ∫ (fnet + mg) dh = fnet hS + mghS = τ (hS ) + ʋ (hS )
h=o h=0
where the kinetic and potential energies are τ (h S) = fnet hS and ʋ (h S) = mghS , respectively.
The ball rises only if the net upward force is positive: f = f net − mg > 0 . Then the ball arrives at h
applied with a non-zero
S
velocity and kinetic energy. If we make f smaller and smaller, it takes the ball longer and longer to reach h ; when it arrives, its
net S
velocity and kinetic energy are smaller and smaller. However, no matter how long it takes the ball to reach h , when it arrives, its
S
potential energy is ʋ (h ) = mgh ,

S S
Now, let us consider the energy change in a process in which the ball begins at rest at height zero and ends at rest at h . At the end,
S
we have τ (h ) = 0 . To effect this change in a real system, we must apply a net upward force to the ball to get it moving; later we
S
must apply a net downward force to slow the ball in such a way that its velocity becomes zero at exactly the time that it reaches h . S
There are infinitely many ways we could apply forces to meet these conditions. The net change in the ball’s energy is the same for
all of them.
We find it useful to use a hypothetical process to calculate this energy change. In this hypothetical process, the upward force is
always just sufficient to oppose the gravitational force on the ball. That is, f = 0 so that f net = mg , and from the
applied
development above v = 0 and τ (h ) = 0 . Of course, t = ∞ . This is a hypothetical process, because the ball would not
S S ∞
actually move under these conditions. We see that the hypothetical process is the limiting case in a series of real processes in which
we make f > 0 smaller and smaller. In all of these processes, the potential energy change is
net
hS
ʋ (hS ) = ∫ mgdh =mghS

h=0
If the ball is stationary and fapplied= mg , the ball remains at rest, whatever its height. If we make f > mg , the ball rises. If
applied
we make f applied < mg >, the ball falls. If f ≈ 0 and the ball is moving only slowly in either direction, a very small change in
net
fnet =f applied − mg can be enough to reverse the direction of motion. These are the characteristics of a reversible process: an
arbitrarily small change in the applied force changes the direction of motion.
The advantage of working with the hypothetical reversible process is that the integral of the applied force over the distance through
which it acts is the change in the potential energy of the system. While we cannot actually carry out a reversible process, we can
compute the work that must be done if we know the limiting force that is required in order to effect the change. This is true because
the velocity and kinetic energy of the ball are zero throughout the process. When the process is reversible, the change in the
potential energy of the ball is equal to the work done on the ball; we have
^ (hS ) = w (hS ) =
−w ʋ (hS )
Gravitational potential energy is an important factor in some problems of interest in chemistry. Other forms of potential energy are
important much more often. Typically, our principal interest is in the potential energy change associated with a change in the
chemical composition of a system. We are seldom interested in the kinetic energy associated with the motion of a macroscopic
system as a whole. We can include effects that arise from gravitational forces or from the motion of the whole system in our
thermodynamic models, but we seldom find a need to do so. For systems in which the motion of the whole system is important, the
laws of mechanics are usually sufficient; we find out what we want to know about such systems by solving their equations of
motion.
When we discuss the first law of thermodynamics, we write E = q + w (or dE = dq + dw ) for the energy change that
accompanies some physical change in a system. Since chemical applications rarely require that we consider the location of the
system or the speed with which it may be moving, “w” usually encompasses only work that changes the energy of the system itself.
Then, E designates the energy of the macroscopic system itself. As noted earlier, we often recognize this by calling the energy of
the system its internal energy. Some writers use the symbol U to represent the internal energy, intending thereby to make it
explicit that the energy under discussion is independent of the system’s location and motion.
This page titled 6.11: Reversible Motion of A Mass in A Constant Gravitational Field is shared under a CC BY-SA 4.0 license and was authored,
The distinction between a system at equilibrium and a system undergoing reversible change is razor-thin. What we have in mind
goes to the way we choose to define the system and centers on the origin of the forces that affect its energy. For a system at
equilibrium, the forces are fixed. For a system undergoing reversible change, some of the forces originate in the surroundings, and
those that do are potentially variable.
To raise a bowling ball reversibly, we apply an upward force, +mg, exactly equal and opposite to the downward force, −mg, due
to gravity. At any point in this reversible motion, the ball is stationary, which is the reason we say that a reversible process is a
hypothetical change. If we were to change the system slightly, by adding a shelf to support the ball at exactly the same height, the
forces on the ball would be the same; however, the forces would be fixed and we would say that the ball is at equilibrium.
We can further illustrate this distinction by returning to the water–water-vapor system. If an unchanging water–water-vapor mixture
is enclosed in a container whose dimensions are fixed (like a sealed glass bulb) we say that the system is at equilibrium. If a piston
encloses the same collection of matter, and the surroundings apply a force on the piston that balances the pressure exerted by the
mixture, we can say that the system is changing reversibly.
In Section 1.6, we used the term “primitive equilibrium” to refer to an equilibrium state in which all of the state functions are fixed.
A system that can undergo reversible change without changing the number or kinds of phases present can be in an infinite number
of such states. Since the set of such primitive equilibrium states encompasses the accessible equilibrium conditions in the sense of
Gibb’s phase rule, we can call this set a Gibbsian equilibrium manifold.
This page titled 6.12: Equilibria and Reversible Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
We usually consider that the first, second, and third laws of thermodynamics are basic postulates. One of our primary objectives is
to understand the ideas that are embodied in these laws. We introduce these ideas here, using statements of the laws of
thermodynamics that are immediately applicable to chemical systems. In the next three chapters, we develop some of the most
important consequences of these ideas. In the course of doing so, we examine other ways that these laws have been stated.
The first law deals with the definition and properties of energy. The second and third laws deal with the definition and properties of
entropy. The laws of thermodynamics assert that energy and entropy are state functions. In the next chapter, we discuss the
mathematical properties of state functions. Energy and entropy changes are defined in terms of the heat and work exchanged
between a system and its surroundings. We adopt the convention that heat and work are positive if they increase the energy of the
system. In a process in which a closed system accepts increments of heat, dq, and work, dw, from its surroundings, we define the
changes in the energy, dE , and the entropy, dS , of the system in terms of dq, dw, and the temperature.
The meaning of the first law is intimately related to a crucial distinction between the character of energy on the one hand and that
of the variables heat and work on the other. When we say that energy is a state function, we mean that the energy is a property of
the system. In contrast, heat and work are not properties of the system; rather they describe a process in which the system changes.
When we say that the heat exchanged in a process is q, we mean that q units of thermal energy are transferred from the
surroundings to the system. If q > 0 , the energy of the system increases by this amount, and the energy of the surroundings
decreases by the same amount. q has meaning only as a description of one aspect of the process.
When the process is finished, the system has an energy, but q exists only as an accounting record. Like the amount on a cancelled
check that records how much we paid for something, q is just a datum about a past event. Likewise, w is the record of the amount
of non-thermal energy that is transferred. Because we can effect the same change in the energy of a system in many different ways,
we have to measure q and w for a particular process as the process is taking place. We cannot find them by making measurements
on the system after the process has gone to completion.
In Section 6.1, we introduce a superscripted caret to denote a property (state function) of the surroundings. Thus, E is the energy of
the system; E ^
is the energy of the surroundings; dE is an incremental change in the energy of the system; and dE ^
is an
incremental change in the energy of the surroundings. If we are careful to remember that heat and work are not state functions, it is
useful to extend this notation to increments of heat and work. If q units of energy are transmitted to the system as heat, we let q^ be
the thermal energy transferred to the surroundings in the same process. Then q^ = −q , and q^ + q = 0 . Likewise, we let w be the
work done on the system and w ^ be the work done on the surroundings in the same process, so that w ^ = −w , and w ^ +w = 0 .
Unlike E and E , which are properties of different systems, q and q^ (or w and w
^ ^ ) are merely alternative expressions of the same
thing—the quantity of energy transferred as heat (or work).

We define the incremental change in the energy of a closed system as dE = dq + dw . The accompanying change in the energy of
the surroundings is dE^
= dq ^ , so that dE + dE
^ + dw ^
= 0 . Whereas q
^ + q = 0 (or d q
^ + dq = 0 ) is a tautology, because it merely
defines q^ as −q, the first law asserts that dE

universe= dE + dE is a fundamental property of nature. Any increase in the energy
^
of the system is accompanied by a decrease in the energy of the surroundings, and conversely. Energy is conserved; heat is not;
work is not.
The first law of thermodynamics

In a process in which a closed system accepts increments of heat, dq, and work, dw, from its surroundings, the change in the
energy of the system, dE , is dE = dq + dw . Energy is a state function. For any process, dE = 0.
universe
For a reversible process in which a system passes from state A to state B, the amount by which the energy of the system changes is
the line integral of dE along the path followed. Denoting an incremental energy change along this path as d E , we haveAB
B
Δ AB E =∫ d AB E . (We review line integrals in the next chapter.) The energy change for the surroundings is the line integral of
A
dE along the path followed by the surroundings during the same process:
^
^ ^
ΔAB E = ∫ dAB E .
A
For any process in which energy is exchanged with the surroundings, the change in the system’s energy is ΔE = q + w , where q
and w are the amounts of thermal and non-thermal energy delivered to the system. We can compute ΔE from q and w whether the
process is reversible or irreversible.
In contrast, the definition of entropy change applies only to reversible processes. In a process in which a system reversibly accepts
an increment of heat, dq , from its surroundings, the entropy change is defined by dS = dq /T . (We introduce the superscript,
rev rev
“rev”, to distinguish heat and work exchanged in reversible processes from heat and work exchanged in irreversible, “irrev”, or
spontaneous, “spon”, processes.) When a system passes reversibly from state A to state B, the entropy change for the system is the
line integral of d S = q /T along the path followed:
AB
rev
B
rev
ΔAB S = ∫ dAB q /T .
A
The entropy change for the surroundings is defined by the same relationship,
rev
^ ^
dAB S = dAB q /T .
Every system has an entropy. The entropies of the system and of its surroundings can change whenever a system undergoes a
change. If the change is reversible, ΔS = −ΔS^ .
The second law of thermodynamics

In a reversible process in which a closed system accepts an increment of heat, dq , from its surroundings, the change in the
rev
entropy of the system, dS , is dS = dq /T . Entropy is a state function. For any reversible process, dS
rev
= 0 , and universe
conversely. For any spontaneous process, dS >0 , and conversely.

universe
We define the entropy change of the universe by dS = dS + dS ; it follows that Δ

universe
^
S =Δ S +Δ
AB S for any
^
universe AB AB
process in which a system passes from a state A to a state B, whether the process is reversible or not. Since dS = 0 for universe
every part of a reversible process, we have ΔS = 0 for any reversible process. Likewise, since dS
universe > 0 for every universe
part of a spontaneous process, we have ΔS > 0 for any spontaneous process.

universe
The third law deals with the properties of entropy at temperatures in the neighborhood of absolute zero. It is possible to view the
third law as a statement about the properties of the temperature function. It is also possible to view it as a statement about the
properties of heat capacities. A statement in which the third law attributes particular properties to the entropy of pure substances is
directly applicable to chemical systems. This statement is that of Lewis and Randall : 3
third law of thermodynamics

If the entropy of each element in some crystalline state be taken as zero at the absolute zero of temperature, every substance
has a positive finite entropy; but at the absolute zero of temperature the entropy may become zero, and does so become in the
case of perfect crystalline substances.
The Lewis and Randall statement focuses on the role that the third law plays in our efforts to express the thermodynamic properties
of pure substances in useful ways. To do so, it incorporates a matter of definition when it stipulates that “the entropy of each
element be taken as zero at the absolute zero of temperature.” The third law enables us to find thermodynamic properties (“absolute
entropies” and Gibbs free energies of formation) from which we can make useful predictions about the equilibrium positions of
reactions. The third law can be inferred from experimental observations on macroscopic systems. It also arises in a natural way
when we develop the theory of statistical thermodynamics. In both developments, the choice of zero for the entropy of “each
element in some crystalline state” at absolute zero is—while arbitrary—logical, natural, and compellingly convenient.
This page titled 6.13: The Laws of Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
When the state of an isolated system can change, we say that the system is capable of spontaneous change. When an isolated
system is incapable of spontaneous change, we say that it is at equilibrium. Ultimately, this statement defines what we mean by
(primitive) equilibrium. From our statement of the second law of thermodynamics, we have criteria for spontaneous change and for
equilibrium in any macroscopic system:
An isolated system can undergo any change that results in an increase in the entropy of the system. The converse is also true; an
isolated system whose entropy can increase can undergo change. Any such change is said to be spontaneous. If an isolated
system cannot change in such a way that its entropy increases, the system cannot change at all and is said to be at equilibrium.
A system that is not isolated can undergo any change that results in an increase in the entropy of the universe, and conversely.
Such changes are also said to be spontaneous.
If a system that is not isolated undergoes a change, but the entropy of the universe remains constant, the change is not
spontaneous. The entropy changes for the system and the surroundings are equal in magnitude and opposite in sign. The change
is said to be reversible.
Although the first statement applies to isolated systems and the second applies to systems that are not isolated, we usually consider
that both are statements of the same criterion, because the second statement follows from the first when we view the universe as an
isolated system. We can restate these criteria for spontaneous change and equilibrium using the compact notation that we introduce
in Section 6.13.
From our definitions, any change that occurs in an isolated system must be spontaneous. From our statement of the second law, the
entropy of the universe must increase in any such process. To indicate this, we write ΔS > 0 . The surroundings must be
universe
unaffected by any change in an isolated system; hence, none of the surroundings’ state functions can change. Thus, ΔS^ = 0 , and
since ΔS + ΔS^ = ΔS > 0 , we have ΔS > 0 .
universe
For a spontaneous change in a system that is not isolated, ΔS can be greater or less than zero. However, ΔS and ΔS^ must satisfy
^
ΔS + ΔS = ΔSuniverse > 0.
In a system that is not isolated, reversible change may be possible. A system that undergoes a reversible change is at—or is
arbitrarily close to—one of its equilibrium states during every part of the process. For a reversible change, it is always true that
^
ΔS = −ΔS , so that
^
ΔS + ΔS = ΔSuniverse = 0.
Our criteria for change are admirably terse, but to appreciate them we need to understand precisely what is meant by “entropy”. To
use the criteria to make predictions about a particular system, we need to find the entropy changes that occur when the system
changes. To use these ideas to understand chemistry, we need to relate these statements about macroscopic systems to the properties
of the molecules that comprise the system.
Since an isolated system does not interact with its surroundings in any way, no change in an isolated system can cause a change in
its surroundings. If an isolated system is at equilibrium, no change is possible, and hence there is no system change for which the
entropy of the universe can increase. Evidently, the entropy of the universe is at a maximum when the system is at equilibrium.
Typically, we are interested in what happens when the interaction between the system and surroundings serves to impose conditions
on the final state of a system. A common example of such conditions is that the surroundings maintain the system at a constant
pressure, while providing a constant-temperature heat reservoir, with which the system can exchange heat. In such cases, the
system is not isolated. It turns out that we can use the entropy criterion to develop supplemental criteria based on other
thermodynamic functions. These supplemental criteria provide the most straightforward means to discuss equilibria and
spontaneous change in systems that are not isolated. Which thermodynamic function is most convenient depends upon the
conditions that we impose on the system.
This page titled 6.14: Thermodynamic Criteria for Change is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
In this chapter, we introduce ideas that underlie classical thermodynamics. Because the development of classical thermodynamics
relies on the properties of reversible processes, we have devoted considerable attention to specifying what we mean by a reversible
process and to the relationship between reversible processes and the equilibrium states available to a system. In Section 7.17, we
develop this theory. The development assumes that we can measure the heat and work exchanged between a system and its
surroundings. It assumes that we can measure the state functions volume, pressure, temperature, and the amounts (moles) of
chemical substances in a system. We define other state functions, whose values can be computed from these measurable quantities.
From observations that we make on systems that are undergoing reversible change, we develop numerous relationships among
these state functions.
No real system can change in exactly the manner we have in mind when we talk about a reversible process. Strictly speaking, any
process that actually occurs must be spontaneous. The idea of reversible change is clearly an abstraction from reality. Nevertheless,
we can determine—to a good approximation—the way in which one variable depends on another in a reversible process. We
accomplish this by making measurements on a real system whose behavior approximates the ideal of reversibility as closely as
possible. We express the—approximate—result of any such measurement as a number. Normally, we view the approximate
character of the number to be a consequence of experimental error. When we say that we make measurements on a system that is
undergoing a reversible change, we mean that we are making the measurements on a process that satisfies our definition of
reversibility closely enough for the purpose at hand.
There are two reasons for the fact that reversible processes play an essential role in the development of the equations of
thermodynamics. The first is that we can measure the entropy change for a process only if the process is reversible. The second and
subtler reason is that an intensive variable may not have a unique value in a system that is undergoing a spontaneous change. If the
temperature, the pressure, or the concentration of a component varies from point to point within the system, then that state function
does not have a unique value, and we cannot use it to model the change. This occurs, for example, when gasoline explodes in a
cylinder of a piston engine. The system consists of the contents of the cylinder. At any given instant, the pressure, temperature, and
component concentrations vary from place to place within the cylinder. In general, no single value of any of these intensive
variables is an adequate approximation for use in the thermodynamic equations that characterize the system as a whole.
To explore this idea further, let us think about measuring changes in extensive state functions during a spontaneous process. Since
we are free to define the system as we please, we can choose a definition that makes the volume readily measurable. In the piston-
engine example, there is no ambiguity about the volume of the system at any instant. While point-to-point variability means that the
concentrations of the chemical components are not defined for the system as a whole, we are confident that there is some specific
number of moles of each component present in the system at every instant. We can reach this conclusion by imagining that we can
instantaneously freeze the composition by stopping all reactions. We can then find the number of moles of each component at our
leisure.
If the system is not too inhomogeneous, we can devise an alternative procedure for making—in concept—such composition
measurements. We imagine dividing the system into a large number of macroscopic subsystems. Each of these subsystems has a
well-defined volume. We suppose also that each of them has well-defined thermodynamic functions at any given instant; that is, we
assume that the pressure, temperature, and concentrations are approximately homogeneous within each of these subsystems. If this
condition is satisfied, we can sum up the number of moles of a component in each of the sub-volumes to obtain the number of
moles of that component in the whole system.
We can make a similar argument for any extensive thermodynamic function, so it applies to the energy, entropy, enthalpy, and the
Helmholtz and Gibbs free energies. As long as the point-to-point variability within the system is small enough so that a division of
the system into macroscopic subsystems produces subsystems that are approximately homogeneous, we can find the value for any
extensive thermodynamic function in each individual sub-system and for the system as a whole. The measurement we propose has
the character of a gedanken experiment. We can describe a procedure for making the measurement, whether we can actually
perform the procedure or not.
This argument does not work for intensive thermodynamic functions. It is true that we could produce a weighted-average value for
the temperature by multiplying the temperature of each subsystem by the subsystem volume, adding up the products, and dividing
the sum by the volume of the whole system; however, the result would not be an intensive property of the whole system. For one
thing, we could produce a different average temperature for every extensive variable by using it rather than the volume as the
weighting factor in the average-temperature computation. Moreover, no such weighted-average temperature can reflect the fact that
different temperatures in different subsystems result in grossly different reaction rates. No single temperature represents the state of
the whole system, and we can make the same statement about any other intensive thermodynamic function. For a non-
homogeneous system that can be subdivided into approximately homogeneous macroscopic subsystems, we can measure, in
principle, the values of the system’s extensive state functions; however, its intensive state functions are essentially undefined.
On the other hand, we may be able to assume that an effectively homogeneous subsystem of macroscopic proportions does have
well-defined extensive and intensive state functions, even if it is not in an equilibrium state. While a spontaneously changing
system need not be homogenous, we commonly encounter systems that are homogeneous to within some arbitrarily small
deviation. Consider a closed and well-stirred system in which some chemical reaction is occurring slowly. We immerse this system
in a constant-temperature bath, and arrange for the applied pressure to be constant. From experience, we know that the temperature
and pressure within such a system will be essentially constant, equal to the bath temperature and the applied pressure, respectively,
and homogeneous throughout the system. In such a system, the temperature and the pressure of the system are at equilibrium with
those imposed by the surroundings. The chemical process is not at equilibrium, but the component concentrations are
homogeneous.
An important question now arises: Are all of the equations of equilibrium thermodynamics applicable to a system in which some
processes occur spontaneously? That they are not is evident from the fact that we can calculate an entropy change from its defining
equation, , only if the behavior of the system is reversible.
Nevertheless, we will find that the relationships among state functions that we derive for reversible processes can be augmented to
describe spontaneous processes that occur in homogeneous systems. The necessary augmentation consists of the addition of terms
that express the effects of changing composition. (In Section 9.14, we develop the fundamental equation,
which applies to any reversible process in a closed system. In Section 14.1, we infer that the fundamental equation becomes
for a spontaneous process in which and are the chemical potentialchemical potential and the change in the number of moles
of component , respectively.)
The distinction between reversiblereversible process and spontaneous processspontaneous processes plays a central role in our
theory. We find a group of relationships that express this distinction, and we call these relationships criteria for change. (In Section
9.19, we find that if and only if the process is reversible, while if and only if the process is
spontaneous. We find a close connection between the criteria for change and the composition-dependent terms that are needed to
model the thermodynamic functions during spontaneous processes. We find that if and only if the process is
reversible, while if and only if the process is spontaneous.)
In thinking about spontaneous processes, we should also keep in mind that the validity of our general relationships among state
functions does not depend on our ability to measure the state functions of any particular state of a system. For example, we can find
relationships among the molar volume and other thermodynamic properties of liquid water. Liquid water does not exist at 200 C
and 1 bar, so we cannot undertake to measure its thermodynamic properties. However, by using our relationships among state
functions and properties that we measure for liquid water where it does exist, we can estimate the thermodynamic properties of
liquid water at 200 C and 1 bar. The results are two steps removed from reality; they are the estimated properties of a hypothetical
substancehypothetical substance. Nevertheless, they have predictive value; for example, we can use them to predict that liquid
water at 200 C and 1 bar will spontaneously vaporize to form gaseous water at 1 bar. The equations of thermodynamics are
creatures of theory. We should not expect every circumstance that is described by the theory to exist in reality. What we require is
that the theory accurately describe every circumstance that actually occurs.
To develop the equations of classical thermodynamics, we consider reversible processes. We then find general criteria for change
that apply to any sort of change in any system. Later, we devise criteria based on the changes that occur in the composition of the
system. In this book, we consider such composition-based criteria only for homogeneous systems. An extensive theory has been
developed to model spontaneous processes in systems that are not necessarily homogeneous. This theory is often called irreversible
thermodynamics or non-equilibrium thermodynamics. Development of this theory has led to a wide variety of useful insights
about various molecular processes. However, much of what we are calling classical thermodynamics also describes irreversible
processes. Even as we develop our theory of reversible thermodynamics, we use arguments that apply the equations we infer from
reversible processes to describe closely related systems that are not at equilibrium.
This page titled 6.15: State Functions in Systems Undergoing Spontaneous Change is shared under a CC BY-SA 4.0 license and was authored,
6.16: Problems
Use Le Chatelier’s principle to answer questions 1-8.
1. One gram of iodine just dissolves in 2950 mL of water at ambient temperatures. One gram of iodine is added to 1000 mL of
pure water and the resulting system is allowed to come to equilibrium. When equilibrium is reached, will all of the iodine have
dissolved? What will happen if a small amount of water is added to the equilibrated system?
2. A saturated solution of barium sulfate is in contact with excess solid barium sulfate.
2+ 2−
BaS O4 (s) ⇌ Ba + SO
4
A small amount of a concentrated solution of BaC l is added. How does the system respond?
2
3. In the presence of a catalyst, oxygen reacts with sulfur dioxide to produce sulfur trioxide.
SO2 (g) + ½O2 (g) ⇌ SO3 (g)
A particular system consists of an equilibrated mixture of these three gases. A small amount of oxygen is added. How does the
system respond?
4. A system containing these three gases is at equilibrium.
SO2 (g) + ½O2 (g) ⇌ SO3 (g)
We suddenly decrease the volume of this system. How does the system respond?
5. In the presence of a catalyst, oxygen reacts with nitrogen to produce nitric oxide.
N2 (g) + O2 (g) ⇌ 2N O (g)
A particular system consists of an equilibrated mixture of these three gases. While keeping the temperature constant, we suddenly
increase the volume of this system. How does the system respond?
6. Nitric oxide formation
N2 (g) + O2 (g) ⇌ 2N O (g)
is endothermic. (At constant temperature, the system absorbs heat as reaction occurs from left to right.) How does the position of
equilibrium change when we increase the temperature of this system?
7. Pure water dissociates to a slight extent, producing hydronium, H3 O
+
, and hydroxide, OH
−
, ions. This reaction is called the
autoprotolysis of water.
+ −
2 H2 O ⇌ H 3 O + OH
Is the autoprotolysisautoprotolysis reaction endothermic or exothermic? (What happens to the temperature when we mix an acid
with a base?) How does the autoprotolysis equilibrium change when we increase the temperature of pure water?
8. At the melting point, most substances are more dense in their solid state than they are in their liquid state. Such a substance is at
its melting point at a particular pressure. Suppose that we now increase the pressure on this system. Does the melting point of the
substance increase or decrease?
For each of the systems 9 – 20, specify
(a) what phases are present,
(b) the number of phases, P ,
(c) the substances that are present,
(d) the number of components*, C , and
(e) the number of degrees of freedom, F .
Assume that the temperature and pressure of each system is constant and that all relevant chemical reactions are at equilibrium.
9. Pure helium gas, H e, sealed in a glass bulb.
10. A mixture of helium gas, H e, and neon gas, N e , sealed in a glass bulb.
11. A mixture of N 2 O4 gas and N O gas sealed in a glass bulb. These compounds react according to the equation
2
N2 O4 (g) ⇌ 2N O2 (g)
12. A mixture of N 2 O4 gas, N O gas, and H e gas sealed in a glass bulb.

2
13. A mixture of P C l gas, P C l gas, and C l gas, sealed in a glass bulb. The proportions of P C l and C l are arbitrary. These
5 3 2 3 2
compounds react according to the equation
P C l5 (g) ⇌ P C l3 (g) + C l2 (g)
14. A mixture of P C l gas, P C l gas, C l gas, and H e gas sealed in a glass bulb. The proportions of P C l and C l are arbitrary.
5 3 2 3 2
15. A mixture of P C l gas, P C l gas, and C l gas sealed in a glass bulb. In this particular system, the number of moles of P C l
5 3 2 3
is the same as the number of moles of C l . 2
16. A saturated aqueous solution of iodine, I . The solution is in contact with a quantity of solid I .
2 2
17. An aqueous solution that contains potassium ion, K , iodide ion, I , triiodide ion,
+ −
I
3
−
, and dissolved I2 . The solution is in
contact with a quantity of solid I . Recall that triiodide ion is formed by the reaction
2
− −
I + I2 ⇌ I
3
18. An aqueous solution that contains K , and I . This solution is in contact with a quantity of solid silver iodide, AgI . Recall
+ −
that AgI is quite insoluble. Neutral molecules of AgI do not exist as such in aqueous solution. The solid substance equilibrates
with its dissolved ions according to the reaction
+ −
AgI (s) ⇌ Ag +I
19. An aqueous solution that contains K

+
, I , and chloride ions,
−
Cl
−
. This solution is in contact with a mixture of (pure) solid
AgI and (pure) solid AgC l .
20. An aqueous solution that contains K

+
, I , C l , and nitrate ions,
− −
NO
−
3
. This solution is in contact with a mixture of (pure)
solid AgI and (pure) solid AgC l.
21. A large vat contains oil and water. The oil floats as a layer on top of the water. Orville has another tank with a reserve supply of
oil. He also has pipes and pumps that enable him to pump oil between his tank and the vat. Wilbur has a third tank with a reserve
supply of water. Wilbur has pipes and pumps that enable him to pump water between his tank and the vat. Their pumps are
calibrated to show the volume of oil or water added to or removed from the vat. Normally, Orville and Wilbur work as a team to
keep the total mass of liquid in the vat constant. The oil and water have densities of 0.80 kg L and 1.00 kg L , respectively. −1 −1
Let M be the total mass of the liquids in the vat.

vat
(a) If Orville pumps a small volume of oil, dV , into or out of the vat, while Wilbur does nothing, what is the change in the mass
oil
of liquid in the vat? (i.e., dM =? )

vat
(b) If Wilbur pumps a small volume of water, dV water , into or out of the vat, while Orville does nothing, what is the change in the
mass of liquid in the vat? (i.e., dM =? ) vat
(c) Suppose that Orville and Wilbur make adjustments, dV and dV at the same time, but contrary to their customary
oil water
practice, they do not coordinate their adjustments with one another. What would be the change in the mass of liquid in the vat? (i.e.,
dM vat=? )
(d) If Orville and Wilbur make adjustments, dV and dV , at the same time, in such a way as to keep the mass of liquid in the
oil water
vat constant, what value of dM results from this combination of adjustments? (i.e., dM =? )
vat vat
(e) From your answers to (c) and (d), what relationship between dVoil and dVwater must Orville and Wilbur maintain in order to
keep dM constant?
vat
(f) One day, the boss, Mr. Le Chatelier, instructs Orville to add 1.00 L of oil to the vat. Qualitatively, what change does Mr. Le
Chatelier impose on the mass of the vat’s contents? (That is, what is the direction of the imposed change?)
(g) Quantitatively, what is the change in mass that Mr. Le Chatelier imposes? (That is, what is the equation for the change in the
mass in the vat in kg?
(h) Qualitatively, how must Wilbur respond?
(i) Quantitatively, how must Wilbur respond? (That is, what is the equation for the change in the mass in the vat in kg?)
22. Which of the following processes can be carried out reversibly?
(a) Melting an ice cube.
(b) Melting an ice cube at 273.15 K and 1.0 bar.
(c) Melting an ice cube at 275.00 K.
(d) Melting an ice cube at 272.00 K and 1.0 bar.
(e) Frying an egg.
(f) Riding a roller coaster.
(g) Riding a roller coaster and completing the ride in 10 minutes.
(h) Separating pure water from a salt solution at 1 bar and 280.0 K.
(i) Dissolving NaCl in an aqueous solution that is saturated with NaCl.
(j) Compressing a gas.
(k) Squeezing juice from a lemon.
(l) Growing a bacterial culture.
(m) Bending (flexing) a piece of paper.
(n) Folding (creasing) a piece of paper.
Notes
1
The ordinate (pressure) values for the solid–liquid and liquid–gas equilibrium lines are severely compressed. The ranges of
pressure values are so different that the three equilibrium lines cannot otherwise be usefully exhibited on the same graph.
2
We also use closely related quantities that we call fugacities. We think of a fugacity as a “corrected pressure.” For present
purposes, we can consider a fugacity to be a particular type of activity.
3
G. N. Lewis and M. Randall, Thermodynamics and the Free Energy of Chemical Substances, 1 Ed., McGraw-Hill, Inc., New
st
York, 1923, p. 448.

4
Ilya Prigogine, Introduction to the Thermodynamics of Irreversible Processes, Second Edition, Interscience Publishers, 1961.
5
S. R. de Groor and P. Mazur, Non-Equilibrium Thermodynamics, Dover Publications, New York, 1984. (Published originally by
North Holland Publishing Company, Amsterdam, 1962.)
CHAPTER OVERVIEW
7: State Functions and The First Law

7.3: Line Integrals
7.9: Measuring Heat
7.21: Problems
This page titled 7: State Functions and The First Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
1
We can specify an equilibrium state of a physical system by giving the values of a sufficient number of the system’s measurable
properties. We call any measurable property that can be used in this way a state function or a state variable. If a system undergoes a
series of changes that return it to its original state, any state function must have the same value at the end as it had at the beginning.
The relationship between our definition of a physical state and our definition of a state function is tautological. A system can return
to its initial state only if every state variable returns to its original value.
Figure 1. Paths between states I and II.

It is evident that the change in a state function when the system goes from an initial state, I , to some other state, I I , must always be
the same. Consider the state functions X, Y , Z , and W . Suppose that functions Y , Z , and W are sufficient to specify the state of a
particular system. Let their values in state I be X , Y , Z , and W . We can express their interdependence by saying that X is a
I I I I I
function of the other state functions Y , Z , and W : X = f (Y , Z , W ) . In state I I , this relationship becomes
I I I I I I I
X II = f (Y II, Z ,W
II ) . The difference
II
XII − XI = f (YII , ZII , WII ) − f (YI , ZI , WI )
depends only on the states I and I I . In particular, X − X is independent of the values of Y , Z , and W in any intermediate
II I
states that the system passes through as it undergoes the change from state I to state I I . We say that the change in the value of a
state function depends only on the initial and final states of the system; the change in the value of a state function does not depend
on the path along which the change is effected.
We can also develop this conclusion by a more explicit argument about the path. Suppose that the system goes from state I to state
I I by path A and then returns to state I by path A
out , as sketched in Figure 1. Let X be some state function. If the change in X
back
as the system traverses path A outis X − X = ΔX (A ) , and the change in X as the system traverses A
II I out is back
X −X
I II = ΔX (A ) , we must have ΔX (A
back ) + ΔX (A
out ) = 0 , so that back
ΔX (Aout ) = −ΔX (Aback )
For some second path comprising B out followed by B back , the same must be true:
ΔX (Bout ) + ΔX (Bback ) = 0
and
ΔX (Bout ) = −ΔX (Bback )
The same is true for any other path. In particular, in must be true for the path Aout followed by Bback , so that
ΔX (A ) + ΔX (B
out ) = 0 , and hence
back
ΔX (Aout ) = −ΔX (Bback )
But this means that
ΔX (Aout ) = ΔX (Bout )
Since the paths A and B are arbitrary, the change in X in going from state I to state I I must have the same value for any path.
Notes
1
Since the temperature of the water increases and the process is to be reversible, we must keep the temperature of the thermal
reservoir just dT greater than that of the water throughout the process. We can accomplish this by using a quantity of ideal gas as
the heat reservoir. By reversibly compressing the ideal gas, we can reversibly deliver the required heat while maintaining the
required temperature. We consider this operation further in §12-5.
This page titled 7.1: Changes in a State Function are Independent of Path is shared under a CC BY-SA 4.0 license and was authored, remixed,
If is a continuous function of the variables x and y , we can think of f (x, y) as a surface in a three-dimensional space.
f (x, y)
f (x, y) is the height of the surface above the xy-plane at the point (x, y) in the plane. If we consider points (x , y ) and (x , y ) 1 1 2 2
in the xy-plane, the vertical separation between the corresponding points on the surface, f (x , y ) and f (x , y ), is 1 1 2 2
Δf = f (x2 , y2 ) − f (x1 , y1 )
We can add f (x 1, y2 ) − f (x1 , y2 ) to Δf without changing its value. Then
Δf = [f (x2 , y2 ) − f (x1 , y2 )] + [f (x1 , y2 ) − f (x1 , y1 )]
If we consider a small change, such that x 2 = x1 + Δx and y 2 = y1 + Δy , we have

[f (x1 + Δx, y1 + Δy1 ) − f (x1 , y1 + Δy1 )] Δx [f (x1 , y1 + Δy1 ) − f (x1 , y1 )] Δy
Δf = +
Δx Δy
Letting df = lim Δx→0 Δf , we have

Δy→0
[f (x1 + Δx, y1 + Δy1 ) − f (x1 , y1 + Δy1 )] Δx [f (x1 , y1 + Δy1 ) − f (x1 , y1 )] Δy

df = lim { } + lim { }
Δx→0 Δx Δy→0 Δy
Δy→0
∂f (x1 , y1 + Δy) ∂f (x1 , y1 ) ∂f (x1 , y1 ) ∂f (x1 , y1 )

= lim { ( ) dx} + ( ) dy = ( ) dx + ( ) dy
Δy→0 ∂x y
∂y x
∂x y
∂y x
We call df the total differential of the function f (x, y):

∂f ∂f
df = ( ) dx + ( ) dy
∂x ∂y
y x
where df is the amount by which f (x, y) changes when x changes by an arbitrarily small increment, dx , and y changes by an
arbitrarily small increment, dy . We use the notation
∂f
fx (x, y) = ( )
∂x y
and
∂f
fy (x, y) = ( )
∂y
x
to represent the partial derivatives more compactly. In this notation, df = f (x, y) dx + f (x, y) dy . We indicate the partial
x y
derivative with respect to x with y held constant at the particular value y = y by writing f (x, y ). 0 x 0
We can also write the total differential of f (x, y) as
df = M (x, y) dx + N (x, y) dy (7.2.1)
in which case M (x, y) and N (x, y) are merely new names for (∂f /∂x) and (∂f /∂y) , respectively. To express the fact that
y x
there exists a function, f (x, y), such that M (x, y) = (∂f /∂x) and N (x, y) = (∂f /∂y) , we say that df is an exact differential.
y x
Inexact Differentials
It is important to recognize that a differential expression in Equation 7.2.1, may not be exact. In our efforts to model physical
systems, we encounter differential expressions that have this form, but for which there is no function, f (x, y), such that
M (x, y) = (∂f /∂x) and N (x, y) = (∂f /∂y) . We call a differential expression, df (x, y), for which there is no corresponding
y x
function, f (x, y), an inexact differential. Heat and work are important examples. We will develop differential expressions that
describe the amount of heat, dq, and work, dw, exchanged between a system and its surroundings. We will find that these
differential expressions are not necessarily exact. (We develop examples in Section 7.17 to Section 7.20.) It follows that heat and
work are not state functions.
This page titled 7.2: The Total Differential is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
7.3: Line Integrals
The significance of the distinction between exact and inexact differential expressions comes into focus when we use the
differential, df , to find how the quantity, f , changes when the system passes from the state defined by (x , y ) to the state defined 1 1
by (x , y ). We suppose that the system undergoes this change along some continuous path in the xy-plane. We can specify such a
2 2
path as a function, c = g (x, y) , where c is a constant, or as y = h (x) . Whether the differential is exact or inexact, we can sum up
increments of change, Δf , along short segments of the path to find the change in f between (x , y ) and (x , y ). Let (x , y ) and 1 1 2 2 i i
(x + Δx, y + Δy) be two neighboring points on the curve c = g (x, y) . As the system traverses c = g (x, y) between these
i i
points, the change in f is
Δf ≈ M (xi , yi ) Δx + N (xi , yi ) Δy
If we sum up such increments of Δf , along the curve c = g (x, y) , from (x , y ) to (x , y ), the sum approximates the change in f
1 1 2 2
along this path. In the limit that all of the incremental Δx and Δy become arbitrarily small, the approximation becomes exact. The
limit of this sum is called the line integral of df along the path c = g (x, y) , between (x , y ) and (x , y ). 1 1 2 2
Whether df is exact or inexact, the line integral of df is defined along any continuous path in the xy-plane. If the path is
c = g (x, y) and it connects the points (x , y ) and (x
1 1 2, y2 )in the xy-plane, we designate the value of the line integral as
c=g( x2 , y2 )
Δf = ∫ df = ∫ df
g c=g( x1 , y1 )
(any differential expression)

However, if df is exact, we know that Δf = f (x2 , y2 ) − f (x1 , y1 ) . In this case, the line integral of df along curve c = g (x, y)
between these points has the value

c=g( x2 , y2 )
Δf = f (x2 , y2 ) − f (x1 , y1 ) = ∫ df
c=g( x1 , y1 )
(for exact differential df )

Because the value of the line integral depends only on the values of f (x, y) at the end points of the integration path, the line
integral of the total differential, df , is independent of the path, c = g (x, y) . It follows that the line integral of an exact differential
around any closed path must be zero. A circle in the middle of the integral sign is often used to indicate that the line integral is
being taken around a closed path. In this notation, writing ∮ df = 0 indicates that df is exact and f is a state function.
In concept, the evaluation of line integrals is straightforward. Since the path of integration is a line, the integrand involves only one
dimension. A line integral can always be expressed using a single variable of integration. Three approaches to the evaluation of line
integrals are noteworthy.
If we are free to choose an arbitrary path, we can choose the two-segment path (x1 , y1 ) → (x2 , y1 ) → (x2 , y2 ) . Along the first
segment, y is constant at y , so we can evaluate the change in f as
1
x2
ΔfI = ∫ M (x, y1 ) dx
x1
Along the second segment, x is constant at x , so we can evaluate the change in f as

2
y2
ΔfII = ∫ N (x2 , y) dy
y
1
Then Δf = ΔfI + ΔfII .

If the path, c = g (x, y) , is readily solved for y as a function of x, say y = h (x) , substitution converts the differential expression
into a function of only x:
dh
df = M (x, h (x)) dx + N (x, h (x)) ( ) dx
dx
Integration of this expression from x to x gives Δf .

1 2
The path, c = g (x, y) , can always be expressed as a parametric function of a dummy variable, t. That is, we can always find
functions x = x (t) and y = y (t) such that c = g (x (t) , y (t)) = g (t) , x = x (t ) , y = y (t ) , x = x (t ) , and y = y (t ) .
1 1 1 1 2 2 2 2
Then substitution converts the differential expression into a function of t :
df = M (x (t) , y (t)) dt + N (x (t) , y (t)) dt
Integration of this expression from t to t gives Δf .

1 2
While the line integral of an exact differential between two points is independent of the path of integration, this not the case for an
inexact differential. For an inexact differential, the integral between two points depends on the path of integration. To illustrate
these ideas, let us consider some examples. These examples illustrate methods for finding the integral of a differential along a
particular path. They illustrate also the path-independence of the integral of an exact differential and the path-dependence of the
integral of an inexact differential.
Example 7.3.1: An exact Differential
We begin by considering the function

2
f (x, y) = xy
for which df = y dx + 2xy dy . Since f (x, y) exists, df must be exact. Let us integrate df between the points
2
(1, 1) and
(2, 2) along four different paths, sketched in Figure 2, that we denote as paths a, b, c, and d.
Figure 2. Paths a, b, c, and d.

Path a has two linear segments. The first segment is the portion of the line y = 1 from x = 1 to x = 2 . Along this
segment, dy = 0 . The second segment is portion of the line x = 2 from y = 1 to y = 2 . Along the second segment,
dx = 0 .
Path b has two linear segments also. The first segment is the portion of the line x = 1 from y = 1 to y = 2 . Along the first
segment, dx = 0 . The second segment is portion of the line y = 2 from x = 1 to x = 2 . Along the second segment,
dy = 0 .
Path c is the line y = x , from x = 1 to x = 2 , and for which dy = dx .

Path d is the line y = x − 2x + 2 , which we can express in parametric form as y = t + 1 and x = t + 1 . At (1, 1),
2 2
t = 0 . At (2, 2), t = 1 . Also, dx = dt and dy = 2t dt .
The integrals along these paths are

Path a:
x=2 y=2
2
∫ df = ∫ 1 dx + ∫ (2) (2) y dy
a x=1 y=1
2 2
2
= x| + 2y ∣∣
1 1
=7
Path b:
x=2 y=2
2
∫ df = ∫ 2 dx + ∫ (2) (1) y dy
b x=1 y=1
2 2 2
= 4x| +y ∣
∣
1 1
=7
Path c:
x=2
2
∫ df = ∫ 3x dx
c x=1
2
3
=x ∣∣
1
=7
Path d:
t=1
2 2 2
∫ df =∫ {(t + 1) + 2(t + 1)(t + 1)(2t)} dt
d t=0
t=1
4 3 2
=∫ {5 t + 4t + 6t + 4t + 1} dt
t=0
1
5 4 3 2
= t +t + 2t + 2t + t∣
∣ =7
0
The integrals along all four paths are the same. The value is 7, which, as required, is the difference f (2, 2) − f (1, 1) = 7 .
Example 7.3.2: An inexact Differential
Now, let us consider the differential expression
dh = y dx + 2xy dy.
This expression has the form of a total differential, but we will see that there is no function, h (x, y), for which this expression
is the total differential. That is, dh is an inexact differential. If we integrate dh over the same four paths, we find
Path a:
x=2 y=2
∫ dh = ∫ (1) dx + ∫ (2)(2)y dy
a x=1 y=1
2 2
2
= x] + 2y ∣
∣
1 1
=7
Path b:
x=2 y=2
∫ dh = ∫ (2) dx + ∫ (2) (1) y dy

b x=1 y=1
2 2
2
= 2x| +y ∣
∣
1 1
=5
Path c:
x=2
2
∫ dh = ∫ (x + 2 x ) dx
c x=1
2
2 3
x 2x
= [ + ]
2 3
1
1
=6
6
Path d:
t=1
2 2
∫ dh =∫ {(t + 1) + 2(t + 1)(t + 1)(2t)} dt
d t=0
t=1
4 3 2
=∫ {4 t + 4t + 5t + 4t + 1} dt
t=0
5 4 3 2 1
= [4 t /5 + t + 5 t /3 + 2 t + t]
0
7
=6
15
For dh (x, y), the value of the integral depends on the path of integration, confirming that dh (x, y) is an inexact differential:
Since the value of the integral depends on path, there can be no h (x, y) for which
( x2 , y )
2
Δh = h (x2 , y2 ) − h (x1 , y1 ) = ∫ dh
( x1 , y1 )
That is, h (x 2, y2 ) − h (x1 , y1 ) cannot have four different values.
This page titled 7.3: Line Integrals is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
Now, let us consider the general case of a continuous function f (x, y), for which the exact differential is
df = fx (x, y) dx + fy (x, y) dy.
We want to integrate the exact differential over very short paths like paths a and b in Section 7.3. Let us evaluate the integral
between (x , y ) and (x + Δx, y + Δy) over the paths a* and b* sketched in Figure 3.
0 0 0 0
Figure 3. Alternative paths from (x 0, y0 ) to (x 0 + Δx, y0 + Δy )

Path a* has two linear segments. The first segment is the portion of the line y = y as x goes from x to x + Δx . Along the 0 0 0
first segment Δy = 0 . The second segment is the portion of the line x = x + Δx as y goes from y to y + Δy . Along the
0 0 0
second segment, Δx = 0 .
Path b* has two linear segments also. The first segment is the portion of the line x = x as y goes from y to y + Δy . Along 0 0 0
the first segment, Δx = 0 . The second segment is the portion of the line y = y + Δy as x goes from x to x + Δx . Along
0 0 0
the second segment, Δy = 0 .

Along path a*, we have
Δa∗ f = fx (x0 , y0 ) Δx + fy (x0 + Δx, y0 ) Δy
Along path b*,
Δb∗ f = fx (x0 , y0 + Δy) Δx + fy (x0 , y0 ) Δy
In the limit as Δx and Δy become arbitrarily small, we must have Δ a

∗ f = Δb∗ f , so that
fx (x0 , y0 ) Δx + fy (x0 + Δx, y0 ) Δy = fx (x0 , y0 + Δy) Δx + fy (x0 , y0 ) Δy
Rearranging this equation so that terms in f are on one side and terms in f are on the other side, dividing both sides by ΔxΔy,
x y
and taking the limit as Δx → 0 and Δy → 0 , we have

fy (x0 + Δx, y0 ) − fy (x0 , y0 ) fx (x0 , y0 + Δy) − fx (x0 , y0 )
lim [ ] = lim [ ]
Δx→0 Δx Δy→0 Δy
These limits are the partial derivative of f y (x0 , y0 ) with respect to x and of f x (x0 , y0 ) with respect to y . That is
2
∂ ∂ ∂f (x0 , y0 ) ∂ f (x0 , y0 )
[ f (x0 , y0 )] = [ ( )] =
∂x y ∂x ∂y ∂y∂x
y y
and
2
∂ ∂ ∂f (x0 , y0 ) ∂ f (x0 , y0 )
[ f (x0 , y0 )] = [ ( )] =
∂y x
∂y ∂x ∂x∂y
x x
This shows that, if f (x, y) is a continuous function of x and y whose partial derivatives exist, then
2 2
∂ f (x0 , y0 ) ∂ f (x0 , y0 )
=
∂y∂x ∂x∂y
The mixed second partial derivative of f (x, y) is independent of the order of differentiation. We also write these second partial
derivatives as f (x , y ) and f (x , y ).
xy 0 0 yx 0 0
To summarize these points, if f (x, y) is a continuous function of x and y , all of the following are true:
1. f (x, y) represents a surface in a three-dimensional space.
2. f (x, y) is a state function.
3. The total differential is
df = (∂f /∂x) dx + (∂f /∂y) dy.

y x
4. The total differential is exact.

5. The line integral of df between two points is independent of the path of integration.
6. The line integral of df around any closed path is zero: ∮ df = 0 .
7. The mixed second-partial derivatives are equal; that is,
2 2
∂ f ∂ f
=
∂y∂x ∂x∂y
This page titled 7.4: Exact Differentials and State Functions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
Since exact differentials have these important characteristics, it is valuable to know whether a given differential expression is exact
or not. That is, given a differential expression of the form
df = M (x, y) dx + N (x, y) dy, (7.5.1)
we would like to be able to determine whether df is exact or inexact. It turns out that there is a simple test for exactness:
test for exactness

The differential in the form of Equation 7.5.1 is exact if and only if
∂M ∂N
= . (7.5.2)
∂y ∂x
That is, this condition is necessary and sufficient for the existence of a function, f (x, y) , for which M (x, y) = fx (x, y) and
N (x, y) = f (x, y).
y
In §4 we demonstrate that the condition is necessary. Now we want to show that it is sufficient. That is, we want to demonstrate: If
Equation 7.5.2 hold, then there exists a f (x, y), such that M (x, y) = f (x, y) and N (x, y) = f (x, y). To do this, we show how
x y
to find a function, f (x, y), that satisfies the given differential relationship. If we integrate M (x, y) with respect to x, we have
f (x, y) = ∫ M (x, y) dx + h (y)
where h (y) is a function only of y ; it is the arbitrary constant in the integration with respect to x, which we carry out with y held
constant.
To complete the proof, we must find a function h (y) such that this f (x, y) satisfies the conditions:
M (x, y) = fx (x, y) ⇔ (7.5.3)
∂
= [∫ M (x, y) dx + h (y)] (7.5.4)
∂x
N (x, y) = fy (x, y) ⇔ (7.5.5)
∂
= [∫ M (x, y) dx + h (y)] (7.5.6)
∂y
The validity of condition in Equation 7.5.3 follows immediately from the facts that the order of differentiation and integration can
be interchanged for a continuous function and that h (y) is a function only of y , so that ∂h/∂x = 0 .
To find h (y) such that condition in Equation 7.5.5 is satisfied, we observe that
∂ ∂M (x, y) dh (y)
[∫ M (x, y) dx + h (y)] = ∫ ( )dx +
∂y ∂y dy
But since
∂M (x, y) ∂N (x, y)
=
∂y ∂x
this becomes
∂ ∂N (x, y) dh (y) dh (y)
[∫ M (x, y) dx + h (y)] = ∫ ( )dx + y = N (x, y) +
∂y ∂x d dy
Hence, condition in Equation 7.5.5 is satisfied if and only if dh (y)/dy = 0 , so that h (y) is simply an arbitrary constant.
This page titled 7.5: Determining Whether an Expression is an Exact Differential is shared under a CC BY-SA 4.0 license and was authored,
If we have f (x, y) while x and y are functions of another variable, u, the chain rule states that
df ∂f dx ∂f dy
=( ) +( )
du ∂x du ∂y du
y x
If x and y are functions of variables u and v ; that is, x = x (u, v) and y = y (u, v) , the chain rule for partial derivatives is
∂f ∂f ∂x ∂f ∂y
( ) =( ) ( ) +( ) ( )
∂u ∂x ∂u ∂y ∂u
v y v x v
A useful mnemonic recognizes that these equations can be generated from the total differential by “dividing through” by du. We
must specify that the “new” partial derivatives are taken with v held constant. This is sometimes called the divide-through rule.
The divide-through rule is a reliable expedient for generating new relationships among partial derivatives. As a further example,
dividing by dx and specifying that any other variable is to be held constant produces a valid equation. Letting w be the variable
held constant, we obtain
∂f ∂f ∂x ∂f ∂y
( ) =( ) ( ) +( ) ( )
∂x w
∂x y
∂x w
∂y x
∂x w
∂f ∂f ∂y
=( ) +( ) ( )
∂x ∂y ∂x
y x w
where we recognize that (∂x/∂x) w

=1 . The result is just the chain rule for(∂f /∂x) when f w
= f (x, y) and ; that
y = y (x, w)
is, when f = f (x, y (x, w)).

If we require that f (x, y) remain constant while x and y vary, we can use the divide-though rule to obtain another useful
relationship from the total differential. If f (x, y) is constant, df (x, y) = 0 . This can only be true if there is a relationship between
x and y . To find this relationship we use the divide-through rule to find (∂f /∂y) when f = f (x (y) , y). Dividing
f
∂f ∂f
df = ( ) dx + ( ) dy
∂x ∂y
y x
by dy , and stipulating that f is constant, we find

∂f ∂f ∂x ∂f ∂y
( ) =( ) ( ) +( ) ( )
∂y ∂x ∂y ∂y ∂y
f y f x f
Since (∂f /∂y) f

=0 and (∂y/∂y) f
=1 , we have
∂f ∂f ∂x
( ) = −( ) ( )
∂y x
∂x y
∂y f
In Chapter 10, we find that the divide-through rule is a convenient way to generate thermodynamic relationships.
This page titled 7.6: The Chain Rule and the Divide-through Rule is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
By definition, the energy of a system can be exploited to produce a mechanical change in the surroundings. The energy of the
surroundings increases; the energy of the system decreases. Raising a weight against the earth’s gravitational force is the classical
example of a mechanical change in the surroundings. When we say that work is done on a system, we mean that the energy of the
system increases because of some non-thermal interaction between the system and its surroundings. The amount of work done on a
system is determined by the non-thermal energy change in its surroundings. We define work as the scalar product of a vector
⇀
representing an applied force, F , and a second vector, ⇀
applied r , representing the displacement of the object to which the force is
applied. The definition is independent of whether the process is reversible or not. If the force is a function of the displacement, we
have
⇀
⇀ ⇀
dw = F ( r ) d r
applied
Pressure–volume work is done whenever a force in the surroundings applies pressure on the system while the volume of the
system changes. Because chemical changes typically do involve volume changes, pressure–volume work often plays a significant
role. Perhaps the most typical chemical experiment is one in which we carry out a chemical reaction at the constant pressure
imposed by the earth’s atmosphere. When the volume of such a system increases, the system pushes aside the surrounding
atmosphere and thereby does work on the surroundings.
When a pressure, P , is applied to a surface of area A , the force normal to the area is F = P
applied A . For a displacement, dx,
applied
normal to the area, the work is F dx = dw = P A dx . We can find the general relationship between work and the change in
applied
the volume of a system by supposing that the system is confined within a cylinder closed by a piston. (See Figure 4.) The
surroundings apply pressure to the system by applying force to the piston. We suppose that the motion of the piston is frictionless.
Figure 4. Pressure–volume work.

The system occupies the volume enclosed by the piston. If the cross-sectional area of the cylinder is A , and the system occupies a
length x, the magnitude of the system’s volume is V = Ax . If an applied pressure moves the piston a distance dx, the volume of
the system changes by dV = A dx . The magnitude of the work done in this process is therefore
system
|dwsystem | = | Papplied A dx|
= | Papplied dVsystem |
work is positive if it is done on the system

We are using the convention that work is positive if it is done on the system. This means that a compression of the system, for
which dx < 0 and dV < 0 , does a positive quantity of work on the system. Therefore, the work done on the system is
system
dw = −P
system dV
applied or, using our convention that unlabeled variables always characterize the system,
system
dw = −Papplied dV
This page titled 7.7: Measuring Pressure-Volume Work is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
For chemical systems, pressure–volume work is usually important. Many other kinds of work are possible. From our vector
definition of work, any force that originates in the surroundings can do work on a system. The force drives a displacement in space
of the system or some part of the system. Stretching a strip of rubber is a one-dimensional analog of pressure–volume work.
Changing the surface area of a liquid is a two-dimensional analog of pressure–volume work. When only internal forces act, a liquid
system minimizes its surface area. We can model this property by attributing a surface-area minimizing force, which we call the
surface tension, to the surface of the liquid. We can think of the layer of molecules at the surface as a film that separates the bulk
liquid from its surroundings. To increase the area of a liquid system requires an expenditure of work by the surroundings against
the surface tension of the film. Gravitational, electrical, and magnetic forces can all do work on particular systems.
In this book, we give little attention to the details of the various kinds of non-pressure–volume work that can be important. (There
are two exceptions: Electrical work is important in electrochemistry, which we discuss in Chapter 17. We discuss gravitational
work in examples that illustrate reversible processes and some aspects of the criteria for change.) Nevertheless, no development of
the basic concepts can be complete without including the effects of non-pressure–volume work. For this reason, we include non-
pressure–volume work in our discussions frequently. For the most part, however, we do so in a generalized or abstract way. To do
so, we must identify some essential features of any process that does work on a system.
Whenever a particular kind of work is done on a system, some change occurs in a thermodynamic variable that is characteristic of
that kind of work. For pressure–volume work this is the volume change. For stretching a strip of rubber, it is the change in length.
For gravitational work, it is the displacement of a mass in a gravitational field. For changing the shape of a liquid, it is the change
in surface area. For electrical work, it is the displacement of a charge in an electrical field. For magnetic work, it is the
displacement of a magnetic moment in a magnetic field. For an arbitrary form of non-pressure–volume work, let us use θ to
represent this variable. We can think of θ as a generalized displacement. When there is an incremental change, dθ , in this variable,
there is a corresponding change, dE , in the energy of the system.
For a displacement, dθ , let the increase in the energy of the system be dw . The energy increase also depends on the magnitude of
θ
the force that must be applied to the system, parallel to the displacement dθ . Let this force be f . Then, for this arbitrary abstract
θ
process, we have dw = f dθ , or f = dw /dθ . Since dw is the contribution to the incremental change in the energy of the
θ θ θ θ θ
system associated with the displacement dθ , we can also write this as

dwθ ∂E
fθ = =
dθ ∂θ
We can generalize this perspective. θ need not be a vector, and ∂E/∂θ need not be a mechanical force. So long as dθ determines
the energy change, dE , we have
∂E
d wθ = ( ) dθ
∂θ
We call ∂E/∂θ a potential. If we let

∂E
Φθ = ( )
∂θ
the energy increment becomes dw = Φ dθ . If multiple forms of work are possible, we can distinguish them by their characteristic
θ θ
variables, which we label θ , θ ,…, θ ,…, θ . For each of these characteristic variables, there is a corresponding potential, Φ ,
1 2 k ω 1
Φ ,…, Φ ,…, Φ . The total energy increment, which we also call the non-pressure–volume work, dw
2 k ω , becomes
N PV
dwN P V = ∑ Φk dθk
k=1
For pressure–volume work, dw = −P dV . The characteristic variable is volume, dθ = dV , and the potential is the negative of
PV
the pressure, Φ = −P . For gravitational work, the characteristic variable is elevation, dθ = dh ; for a given system, the potential
V
depends on the gravitational acceleration, g , and the mass of the system: Φ = mg . h
When a process changes the composition of a system, it is often important to relate the work done on the system to the composition
change. Formally, we express the incremental work resulting from the k -th generalized displacement as
∂E ∂θk
d wk = ( )( ) dn
∂θk ∂n
where dw and dn are the incremental changes in the work done on the system and the number of moles of the substance in the
k
system. To see how this works out in practice, let us consider the particular case of electrical work. The electrodes of an
electrochemical cell can be at different electric potentials. We usually designate the potential difference between the electrodes as
Ecell. (We can also write Φ = E cell when we want to keep our notation uniform. The unit of electrical potential is the volt, V. One
cell
volt is one joule per coulomb, 1 V = 1 J C .) We are usually interested in cases in which we can assume that E is constant.
−1
cell
Whenever a current flows in an electrochemical cell, electrons flow through an external circuit from one electrode to the other. By
our definition of electrical potential, the energy change that occurs when a charge dq passes through a potential difference, E , is cell
dE = Ecell dq = dwelect
We have E cell = (∂E/∂q) . Evidently, charge is the characteristic variable for electrical work; we have θ elect =q , and
∂E ∂E
( ) =( ) = Ecell
∂θelect ∂q
¯
¯¯¯
¯ ¯
¯¯¯
¯
Letting the magnitude of the electron charge be e , dN electrons carry charge dq = −e dN . Then, dq = (−eN ) (dN / N ) . The
magnitude of the charge carried by one mole of electrons is the faraday, F . That is, 1 F ¯
¯¯¯
¯
= ∣eN ∣ = 96, 485 C mol
∣ ∣
−1
. (See §17-8.)
Letting dn be the number of moles of electrons, we have dn = dN /N and ¯
¯¯¯
¯
dq = −F dn , so that
∂θelect ∂q
( ) =( ) = −F
∂n ∂n
The work done when dn moles of electrons pass through the potential difference E cell becomes
∂E ∂θcell
dwelect = ( )( ) dn
∂θcell ∂n
∂E ∂q
=( )( ) dn = −F Ecell dn
∂q ∂n
We find the work done when ions pass through a potential difference E by essentially the same argument. If ions of species j carry
charge z e, then dN ions carry charge dq = z e dN = z F dn , and the electrical work is (dw
j j j j j j ) = z F Edn . If ω different
elect j j j
species pass through the potential difference, the total electrical work becomes
ω
dwelect = ∑ zj F Ednj
j=1
This page titled 7.8: Measuring Work- Non-Pressure-Volume Work is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
7.9: Measuring Heat
As the idea of heat as a form of transferring energy was first being developed, a unit amount of heat was taken to be the amount
that was needed to increase the temperature of a reference material by one degree. Water was the reference material of choice, and
the calorie was defined as the quantity of heat that raised the temperature of one gram of water one degree kelvin. The amount of
heat exchanged by a known amount of water could then be calculated from the amount by which the temperature of the water
changed. If, for example, introducing 63.55 g (1 mole) of copper metal, initially at 274.0 K, into 100 g of water, initially at 373.0
K, resulted in thermal equilibrium at 288.5 K, the water surrendered
−1 −1
100 g × 1 cal g K × 84.5 K = 8450 cal
This amount of heat was taken up by the copper, so that 0.092 cal was required to increase the temperature of one gram of copper
by one degree K. Given this information, the amount of heat gained or lost by a known mass of copper in any subsequent
experiment can be calculated from the change in its temperature.
Joule developed the idea that mechanical work can be converted entirely into heat. The quantity of heat that could be produced
from one unit of mechanical work was called the mechanical equivalent of heat. Today we define the unit of heat in mechanical
units. That is, we define the unit of energy, the joule (J), in terms of the mechanical units mass (kg), distance (m), and time (s).
One joule is one newton-meter or one kg m s . One calorie is now defined as 4. 184 J , exactly. This definition assumes that
2 −2
heat and work are both forms of energy. This assumption is an intrinsic element of the first law of thermodynamics. This aspect of
the first law is, of course, just a restatement of Joule’s original idea.
When we want to measure the heat added to a system, measuring the temperature increase that occurs is often the most convenient
method. If we know the temperature increase in the system, and we know the temperature increase that accompanies the addition of
one unit of heat, we can calculate the heat input to the system. Evidently, it is useful to know how much the temperature increases
when one unit of heat is added to various substances. Let us consider a general procedure for accumulating such information.
Figure 5. Heat capacity is the slope of q versus T .

First, we need to choose some standard amount of the substance in question. After all, if we double the amount, it takes twice as
much heat to effect the same temperature change. One mole is a natural choice for this standard amount. If we add small
increments of heat to one mole of a pure substance, we can measure the temperature after each addition and plot heat versus
temperature. Figure 5 shows such a plot. (In experiments like this, it is often convenient to introduce the heat by passing a known
electrical current, I , through a known resistance, R , immersed in the substance. The rate at which heat is produced is I R. Except
2
for the usually negligible amount that goes into warming the resistor, all of it is transferred to the substance.) At any particular
temperature, the slope of the graph is the increment of heat input divided by the incremental temperature increase. This slope is so
useful, it is given a name; it is the molar heat capacity of the substance, C . Since this slope is also the derivative of the q-versus-T
curve, we have
dq
C =
dT
The temperature increase accompanying a given heat input varies with the particular conditions under which the experiment is
done. In particular, the temperature increase will be less if some of the added heat is converted to work, as is the case if the volume
of the system increases. If the volume increases, the system does work on the surroundings. For a given q, ΔT will be less when
the system is allowed to expand, which means that q/ΔT will be greater. Heat capacity measurements are most conveniently done
with the system at a constant pressure. However, the heat capacity at constant volume plays an important role in our theoretical
development. The heat capacity is denoted C when the pressure is constant and C when the volume is constant. We have the
P V
important definitions
∂q
CP = ( )
∂T
P
and
∂q
CV = ( )
∂T
V
Since no pressure–volume work can be done when the volume is constant, less heat is required to effect a given temperature
change, and we have C > C , as a general result. (In §14, we consider this point further.) If the system contains a gas, the effect
P V
of the volume increase can be substantial. For a monatomic ideal gas, the temperature increase at constant pressure is only 60% of
the temperature increase at constant volume for the same input of heat.
This page titled 7.9: Measuring Heat is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
A state function must return to its original value if a system is taken through a series of changes and finally returned to its original
state. We say that the change in a state function must be zero if the system is taken through a cyclic process, or somewhat more
picturesquely, if the system traverses a cyclic path. While we can measure the heat and work that a system exchanges with its
surroundings, neither the heat nor the work is necessarily zero when the system traverses a cycle. Heat and work are not state
functions. Nevertheless, adding heat to a system increases its energy. Likewise, doing work on a system increases its energy. If the
system surrenders heat to the surroundings or does work on the surroundings, the energy of the system is decreased. In any change
that a closed system undergoes, the total energy change is E = q + w , where q and w can be either positive or negative. For very
small changes, we write
dE = dq + dw.
Anything we do to increase the energy of a closed system can be classified as either adding heat to the system or doing work on the
system.
Heat, work, and energy are all extensive variables. They are additive. If a system acquires an increment of heat q from one source
1
and an increment q from another source, the total heat acquired by the system is q + q . If work, w , of one kind, and work, w ,
2 1 2 1 2
of a second kind are done on the system, the total work is w + w .

1 2
In keeping with the thermodynamic perspective that we can partition the universe into system and surroundings, we assume that
any energy lost by the system is taken up by the surroundings, and vice versa. By definition, we have q = −q^ , w = −w ^ , and
E = −E ; for any process, E = E + E . This is the principle of conservation of energy, which is usually stated:
^ ^
universe
For any change in any system, the energy of the universe remains constant.
So, conservation of energy is built into our energy accounting scheme. It is a consequence of the thermodynamic perspective and
our rules for keeping track of exchanges of heat and work between system and surroundings. Conservation of energy is an
“accounting convention,” but it is not arbitrary. That is, we are not free to choose another convention for this “energy accounting.”
Ample experimental evidence supports our assumption that energy conservation is a fundamental property of nature.
In summary, we postulate that for any change whatsoever that a closed system may undergo, we can identify energy inputs either as
heat or as one or more forms of work such that
E = q +w
and, if a system undergoes a series of changes that ultimately return it to its original state, the energy change for the entire series of
changes will be zero. There are two components to this postulate. The first component is an operational definition of energy. An
important aspect of this definition is that the principle of conservation of energy is embedded in it. The second is an assertion that
energy is a state function. Our operational definition of energy is open-ended. An essential element of the postulate is that we can
always identify work inputs that make the energy, whose changes we compute as q + w , a state function. The facts that energy is
conserved and that energy is a state function are related properties of a single aspect (energy) of nature. The relationship between
these facts is a characteristic property of physical reality; it is not a matter of logic in the sense that one fact implies the other.
All of these ideas are essential components of the concept of energy. We roll them all together and assert them as a postulate that
we call the first law of thermodynamics. We introduce the first law in Chapter 6. We repeat it here.
The first law of thermodynamics

In a process in which a closed system accepts increments of heat, dq, and work, dw, from its surroundings, the change in the
energy of the system, dE , is dE = dq + dw . Energy is a state function. For any process, dE =0.
universe
This statement of the first law does not deal explicitly with the mechanical energy of the system as a whole or with the energy
effects of a transport of matter across the boundary of an open system. Because dw can include work that changes the position or
motion of a system relative to an external reference frame, increments of mechanical energy can be included in dE . In chemical
applications, we seldom need to consider the mechanical energy of the system as a whole; we can assume that the system has no
kinetic or potential energy associated with the movement or location of its mass. When this is case, the total incremental energy
change, dE , is the same thing as the incremental change in the internal energy of the system. When we need to distinguish the
internal energy of a system from its total energy, we write U for the internal energy and E for the total energy. Letting incremental
changes in the kinetic and potential energy of the whole system be dτ and dʋ , respectively, we have
dU = dq + dw
and
dE = dU + dτ + d ʋ.
For processes of interest in chemical systems, we normally have dτ = d ʋ =0 . Then the total energy and the internal energy are
the same thing: dE = dU .
In Section 7.8, we introduce characteristic variables, θ , to represent changes in the system that result from various forms of non-
k
pressure–volume work done on the system. We let Φ = (∂E/∂ θ ) , so that we can represent the incremental energy change that
k k
results from the non-pressure–volume work of all kinds as

ω
k=1
When both pressure–volume and non-pressure–volume work occur, we have

dw = dwP V + dwN P V
= −P dV + ∑ Φk dθk
k=1
When a non-thermal process changes the energy of a closed, constant-volume system, we have dE = dw N PV .
We state the first law for a closed system. Extending the first law to open systems is straightforward. The energy of a system
depends on the substances that are present, their amounts, and their states. At any specified conditions, a given amount of a
particular substance makes a fixed contribution to the energy of the system. If we transfer matter across the boundary of a system,
we change the energy of the system. We can always alter the original system to include the matter that is to be transferred. The
altered system is closed; and so, by the first law, its energy is the same after the transfer as it was before. In Section 14-2 we
develop an explicit mathematical function to model the contribution made to the energy of a system by a specified quantity of
matter in a specified state. If matter crosses the boundary of a system, the energy models for the separate collections of substances
pre-transfer must equate to that for the new system post-transfer.
Finally, we make a further simple but important observation: We imagine that we can always identify an energy increment that
crosses a system boundary as work, dw, or heat, dq. However, the essence of the first law is that these increments lose their
identities—so to speak—in the system. The effect of a work input, dw, doesn’t necessarily appear as an increase in the mechanical
energy of the system; a heat input, dq, doesn’t necessarily appear as in increase in the thermal energy of the system.
To illustrate this point, let us consider a reversible process and an irreversible process, each of which increases the temperature of
one gram of water by one degree K. The initial and final states of the system are the same for both processes. In the reversible
process, we bring the water, whose temperature is T , into contact with a thermal reservoir at an incrementally higher temperature
T + dT and allow 4.184 J of heat to transfer to the system by convection . No work is done. We have
1
^ = 4.184 J
q = −q ,
^ =0
w = −w , and
E = q = 4.184 J .
In this case, there is no inter-conversion of heat and work.

In the irreversible process, we stir one gram of water that is thermally isolated. The stirring generates heat in the system. We supply
the energy to drive the stirrer from the surroundings, perhaps by allowing a spring to uncoil. When 4.184 J of work from the
surroundings has been frictionally dissipated in the system, the state of the one-gram system is the same as it was at the end of the
reversible process. In this irreversible process, all of the energy traverses the system boundary as work (non-thermal energy). No
heat traverses the system boundary. We have
w = −w
^ = 4.184 J ,
q = −q
^ =0 , and
ΔE = w = 4.184 J .
Uncoiling the spring generates bulk motion within the water. Within a short time, the energy of this bulk motion is completely
dissipated as molecular-level kinetic energy, or heat.
Beginning in §18, we consider the reversible isothermal expansion of an ideal gas. This simple process provides a further
illustration of the inter-conversion of heat and work within a system. For this process, we find that thermal energy, q > 0 , crosses
the system boundary and is converted entirely into work that appears in the surroundings, w ^ = −w > 0 . The energy of the system
is unchanged. The energy of an ideal gas depends only on temperature. In an isothermal expansion, the temperature is constant, so
the transport of heat across the system boundary has no effect on the energy of the system.
This page titled 7.10: The First Law of Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
The first law has been stated in many ways. Some are intended to be humorous or evocative rather than precise statements; for
example, “You can’t get something (useful work in some system) for nothing (no decrease in the energy of some other system).”
Others are potentially ambiguous, because we construct them to be as terse as possible. To make them terse, we omit ideas that we
deem to be implicit.
A compact and often used statement is, “E = q + w , and E is a state function.” In this statement, the fact that energy is conserved
is taken to be implicit in the operational definition, E = q + w . We can give an equally valid statement by saying, “Energy is
conserved (E = E + E = 0 ) in all processes.” In making this statement, we assume that the definition of energy (
universe
^
E = q + w ) is understood and that the state-function postulate is implicit in this definition.
To see that the postulate that energy is conserved and the postulate that energy is a state function are logically independent, let us
consider a system that undergoes a particular cyclic process, which we call “Cycle A.” In Cycle A, the final state of the system is
the same as its initial state; the postulate that energy is a state function is then equivalent to the statement that E = 0 . The
C ycle A
postulate that energy is conserved is equivalent to the statement that E C ycle A +E

^
= 0 . Now, what can we say about
C ycle A
E
^
? Obviously, if we combine the information from the two postulates, it follows that E
C ycle A
^
= 0 . The essential point,
C ycle A
however, is that E^ = 0 is not required by either postulate alone.

C ycle A
^
E C ycle A = 0is not required by the postulate that energy is a state function, because the surroundings do not necessarily
traverse a cycle whenever the system does.
= 0 is not required by conservation of energy, which merely requires E , and absent the
^ ^
E C ycle A = −E
C ycle A C ycle A
requirement that E be a state function, E could be anything.

C ycle A
In Chapter 9, we explore a statement of the second law that denies the possibility of constructing a “perpetual motion machine of
the second kind.” Such a perpetual motion machine converts heat from a constant-temperature reservoir into work. This statement
is: “It is impossible to construct a machine that operates in a cycle, exchanges heat with its surroundings at only one temperature,
and produces work in the surroundings.”
A parallel statement is sometimes taken as a statement of the first law. This statement denies the possibility of constructing a
“perpetual motion machine of the first kind.” This statement is, “It is impossible to construct a machine that operates in a cycle and
converts heat into a greater amount of work.” The shared perspective and phrasing of these statements is esthetically pleasing. Let
us consider the relationship between this statement of the first law and the statement given in §10. (For brevity, let us denote this
impossibility statement as the “machine-based” statement of the first law and refer to it as proposition “MFL.” We refer to the
statement of the first law given in §10 as proposition “FL.”)
In the machine-based statement (MFL), we mean by “a machine” a system that accepts heat from its surroundings and produces a
greater amount of work, which appears in the surroundings. If such a machine exists, the machine-based statement of the first law
is false, and proposition ∼MFL is true. For one cycle of this first-law violating machine, we have w ^ > q > 0 . Since q = −q ^ , we
have w ^ > −q ^ > 0 . It follows that E = q

^
^+w ^ > 0 . Our statement of the principle of conservation of energy (E + E = 0 ) then
^
requires that, for one cycle of this perpetual motion machine, E < 0 . Our statement of the first law, FL, requires that, since energy
is a state function, E = 0 . Since this is a contradiction, the existence of a perpetual motion machine of the first kind (proposition
∼MFL) implies that the first law (energy is a state function, proposition FL) is false (∼MFL ⇒ ∼FL).
From this result, we can validly conclude: If the first law is true, the existence of a perpetual motion machine of the first kind is
impossible:
(∼MFL ⇒ ∼FL) ⇒(∼∼FL⇒∼∼MFL)⇒(FL⇒MFL)
We cannot conclude that the impossibility of perpetual motion of the first kind implies that energy is a state function
(∼MFL ⇒ ∼FL) does not imply (MFL⇒FL)
That is, the impossibility of perpetual motion of the first kind, as we have interpreted it, is not shown (by this argument) to be
equivalent to the first law, as we have stated it. (It remains possible, of course, that this equivalence could be proved by some other
argument.)
Evidently, when we take the impossibility of constructing a perpetual motion machine of the first kind as a statement of the first
law, we have a different interpretation in mind. The difference is this: When we specify a machine that operates in a cycle, we
intend that everything about the machine shall be the same at the end of the cycle as at the beginning—including its energy. That is,
we intend the statement to be understood as requiring that, for one cycle of the perpetual motion machine E = 0 . Equivalently, we
intend the statement to be understood to include the stipulation that energy is a state function.
Now, for one cycle of the perpetual motion machine, we have E = 0 and E ^
> 0 . Given the basic idea that energy is additive, so
that E universe= E + E , we have that E

^
> 0 . The impossibility statement asserts that this is false; equivalently, the
universe
impossibility statement asserts that energy cannot be created. This conclusion is a weak form of the principle of conservation of
energy; it says less than we want the first law to say. We postulate that energy can be neither created nor destroyed. That is,
E universe> 0 and E < 0 are both false. When we consider the impossibility statement to assert the principle of energy
universe
conservation, we implicitly stipulate that the machine can also be run in reverse. (See problem 10.)
We intend the first law to assert the existence of energy and to summarize its properties. However we express the first law, we
recognize that the concept of energy encompasses several closely interrelated ideas.
This page titled 7.11: Other Statements of the First Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
From the outset of our study of energy, we recognize that we are always dealing with energy changes. Even when we write
E = E (P , T , h) to indicate that energy is a function of P , T , and h , we recognize that E represents the energy difference
between the state of the system characterized by P , T , and h and the state of the system when the independent variables
correspond to a reference state in which, by definition, E = 0 . As we observe in §6-2, we can sort thermodynamic variables into
two classes. Some, like P , V , and T , can be measured only for a system. Others, like q, w, E , S , H , G, and A , can be measured
only for a process. To say that the volume of a system is one cubic meter has absolute significance. To say that the energy of a
system is one joule means nothing unless we know the reference state.
When we intend to specify that the reference state for energy is the particular state specified by P = P , T = T , and h= h , we
0 0 0
write “E (P , T , h ) = 0 .” Otherwise, when we write “E = E (P , T , h) ,” we could equally well write “ΔE = E (P , T , h) .” We

0 0 0
intend either of these formulations to mean the same thing as “E = E (P , T , h) − E (P , T , h ) and E (P , T , h ) = 0 .”

0 0 0 0 0 0
Whether we write E or ΔE, the quantity represented is the difference in energy between some initial and some final state. When
we focus on very small changes, we can write dE or d (ΔE) . If our perspective is that we are describing a process, we may prefer
to write “E ”; if our perspective is that we are describing a change in the system, we may prefer to write “ΔE.” In practice, our
choice depends primarily on what we have grown accustomed to in the context at hand. In the discussion above, we write
E = q + w . We could equally well write ΔE = q + w . The meaning is the same. We can make similar statements about most
thermodynamic functions. Often there is no particular reason to prefer X over ΔX, or vice versa.
However, there are circumstances in which the delta notation serves particular purposes. If a system undergoes a change in which
some thermodynamic variables remain constant, the delta notation provides a convenient way to indicate that a particular variable
is not constant. For example, if the volume of a system changes while the applied pressure remains constant, we write
w = −P ΔV .
applied
Similarly, we often want to describe processes in which some state functions are different in the final state than they are in the
initial state, while other state functions are the same in both states, but not necessarily constant throughout the process. In the next
few chapters, we develop properties of the state functions entropy, S , enthalpy, H , and Gibbs free energy, G. We define the Gibbs
free energy by the relationship G = H − T S . To specify the relationship among the changes in these state functions when the final
temperature is the same as the initial temperature, we write ΔG = ΔH − T ΔS . Here too, we often say that this relationship
relates the changes in ΔG, ΔH , and ΔS when “the temperature is constant.” This is another useful, but potentially misleading,
figure of speech. It is important to remember that the equation is valid for any path between the same two states, even if the
temperature varies wildly along that path, so long as the initial and final states are at the same temperature.
Finally, we find it convenient to use subscripted versions of the delta notation to specify particular kinds of processes. For a process
in which one mole of a pure substance vaporizes to its gas at a particular temperature, we write Δ H and Δ G to denote the
vap vap
changes in enthalpy and Gibbs free energy, respectively. (We can write Δ E to denote the change in the energy; however, Δ E
vap vap
is not a quantity that we find useful very often.) Similarly for the fusion and sublimation of one mole of a pure substance at a
particular temperature, we write Δ G, Δ G, Δ H, and Δ H . We also find it convenient to write Δ H and Δ G to denote
f us sub f us sub r r
the changes in these quantities when a chemical reaction occurs. When we do so, it is essential that we specify the corresponding
stoichiometric equation.
This page titled 7.12: Notation for Changes in Thermodynamic Quantities - E vs. ∆E is shared under a CC BY-SA 4.0 license and was authored,
If we heat or do work on any gas—real or ideal—the energy change is E = q + w . When we investigate the energy change that
accompanies a temperature change, we can obtain reproducible results by holding either the pressure or the volume constant. With
volume held constant, we measure C . With pressure held constant, the energy change we measure depends on both C and the
V P
relationship among the pressure, volume, and temperature of the gas. If we know an equation of state for the gas and the values of
both C and C , we can find the energy change between any two states of the gas, because the same change of state can be
V P
achieved in two steps, one at constant pressure and one at constant volume.
To see this, we recognize that the state of any pure gas is completely specified by specifying its pressure, temperature, and volume.
Any change of state necessarily involves changing at least two of these state functions. Any change of state that changes all three of
them can be achieved in an alternate way that involves two changes, each of which occurs with one variable held constant. For
example, the change
(P1 , V1 , T1 ) → (P2 , V2 , T2 )
can be achieved by the constant-pressure sequence
(P1 , V1 , T1 ) → (P1 , V2 , Ti )
followed by the constant-volume sequence
(P1 , V2 , Ti ) → (P2 , V2 , T2 )
where T is some intermediate temperature. Note that this sequence has to be possible: with P held constant, specifying a change in
i
T is sufficient to determine the change in V ; with V held constant, specifying a change in T is sufficient to determine the change
in P .
Let us consider how the energy of one mole of any pure substance changes with temperature at constant volume. The rate of
change of E with T is
∂E ∂q ∂w ∂w
( ) =( ) +( ) = CV + ( )
∂T ∂T ∂T ∂T
V V V V
where we use the definition of C . For any system, and hence for any substance, the pressure–volume work is zero for any process
V
in which the volume remains constant throughout; therefore, we have (∂w/∂T ) = 0 and V
∂E
( ) = CV
∂T
V
(one mole of any substance, only PV work possible)

When we develop the properties of ideal gases by treating them as point mass molecules, we find that their average translational
kinetic energy is 3RT /2 per mole or 3kT /2 per molecule, which clearly depends only on temperature. Translational kinetic energy
is the only form of energy available to a point-mass molecule, so these relationships describe all of the energy of any point-mass
molecule. In particular, they describe all of the energy of a monatomic ideal gas. Since the energy of a monatomic ideal gas is
independent of pressure and volume, the temperature derivative must be independent of pressure and volume. The ordinary
derivative and the partial derivatives at constant pressure and constant volume all describe the same thing, which, we have just
seen, is C . V
dE ∂E ∂E 3
=( ) =( ) = CV = R
dT ∂T ∂T 2
P V
(one mole of a monatomic ideal gas)

It is useful to extend the idea of an ideal gas to molecules that are not monatomic. When we do so, we have in mind molecules that
do not interact significantly with one another. Another way of saying this is that the energy of the collection of molecules is not
affected by any interactions among the molecules; we can get the energy of the collection by adding up the energies that the
individual molecules would have if they were isolated from one another. In our development of statistical thermodynamics, we find
that the energy of a collection of non-interacting molecules depends only on the molecules’ energy levels and the temperature. The
molecules’ energy levels are fixed. This means that if we extend our idea of ideal gases to include non-interacting polyatomic
compounds, the energies of such gases still depend only on temperature. For any ideal gas, we have
dE ∂E ∂E
=( ) =( ) = CV
dT ∂T ∂T
P V
(one mole of any ideal gas)

However, for polyatomic molecules it will no longer be true that C = 3R/2 . Let us see why. Recall that we construct our
V
absolute temperature scale by extrapolating the Charles’ law graph of volume versus temperature to zero volume. (Figure 2-2.) By
experiment, we find that this graph is the same for one mole of a polyatomic ideal gas as it is for one mole of a monatomic ideal
gas. Evidently, our definition of temperature depends only on the translational energy of ideal gas molecules and vice-versa. At a
fixed temperature, the average translational kinetic energy is the same for any ideal gas; it is independent of the mass of the
molecule and of the kinds of atoms in it. To increase the temperature by one degree requires that the translational kinetic energy
increase by 3R/2, and vice versa.
Consider what happens when we add energy to a polyatomic ideal gas. Polyatomic gas molecules have energy in rotational and
vibrational modes of motion. When we add energy to such molecules, some of the added energy goes into these rotational and
vibrational modes. To achieve the same increase in translational kinetic energy, the total amount of energy added must be greater.
We find that we need a larger ΔE to achieve the same ΔT , which means that the heat capacity (either C or C ) of the V P
polyatomic ideal gas is greater than that of a monatomic ideal gas.

Now let us consider the rate of change of E with T at constant pressure. For one mole of any substance, we have
∂E ∂q ∂w ∂w
( ) =( ) +( ) = CP + ( )
∂T ∂T ∂T ∂T
P P P P
This equation is as far as we can go, unless we can focus on a particular situation for which we know how work varies with
temperature at constant pressure.
For one mole of an ideal gas, we have this information. From P V = RT at constant P , we have P dV = RdT . If reversible work
is done on the ideal gas,
w = ∫ −Papplied dV = ∫ −P dV and
∂w ∂ ∂
( ) = [ ∫ −P dV ] = [ ∫ −RdT ] = −R
∂T ∂T ∂T
P P P
(any ideal gas)

That is, when enough heat is added to increase the temperature of one mole of ideal gas by one degree kelvin at constant pressure,
−R units of work are done on the gas. This is the energy change that occurs because of the increase in volume that accompanies
the one-degree temperature increase. Since, for any ideal gas,

∂E ∂q ∂w
CV = ( ) =( ) +( ) = CP − R
∂T ∂T ∂T
P P P
we have
CP = CV + R

For a monatomic ideal gas,
3 5
CP = CV + R = R+R = R
2 2

The heat capacity functions have a pivotal role in thermodynamics. We consider many of their properties further in the next section
and in later chapters (particularly §10-9 and §10-10.) Because we want to use these properties before we get around to justifying
them all, let us summarize them now:
1. For monatomic ideal gases, C and C are independent of temperature.
V P
2. For polyatomic gases, real or ideal, C and C are functions of temperature.

V P
3. C is always greater than C , but as the temperature decreases, their values converge, and both vanish at absolute zero.
P V
4. At ordinary temperatures, C and C increase only slowly as temperature increases. For many purposes they can be taken to be
V P
constant over rather wide temperature ranges.

5. For real substances, C is a weak function of volume, and C is a weak function of pressure. These dependencies are so small
V P
that they can be neglected for many purposes.

6. For ideal gases, C is independent of volume, and C is independent of pressure.
V P
This page titled 7.13: Heat Capacities for Gases- Cv, Cp is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
It is easy to maintain a constant pressure on a solid while varying its temperature. To keep its volume rigorously constant over a
range of temperatures is difficult. Because the direct measurement of C is straightforward, most heat-capacity experiments on
P
solids measure C . In Section 10.9, we derive a general relationship between C , C , and other measurable properties of a
P P V
substance. This relationship makes it possible to evaluate C indirectly. For a solid, this relationship shows that C and C are
V P V
usually about the same.

Heat capacities of solids have been investigated over wide temperature ranges. For most solids, C is approximately constant at
P
room temperature and above. For any of the heavier elements, this constant has about the same value. This observation was first
made in 1819. It is called the law of Dulong and Petit, in honor of the discoverers. It played an important role in the establishment
of correct atomic weights for the elements. The value of the constant found by Dulong and Petit is about 3R. Remarkably, the law
can be extended to polyatomic molecules containing only the heavier elements. Often the solid-state heat capacity of such
molecules is about 3R per mole of atoms in the molecule. Correlations that are more detailed have been developed. These relate
the heat capacity of a mole of a molecular solid to its molecular formula. In such correlations, the heat capacity per mole increases
by a fixed increment for each atom of, say, carbon in the molecule; by a different fixed increment for each atom of nitrogen in the
molecule; etc. For the lighter elements, the increments are less than 3R. For the heavier elements, the increment is approximately
3R, as observed by Dulong and Petit.
Figure 6. Heat capacity of solid rhombohedral mercury. Data from D. R. Linde, Editor, The Handbook of Chemistry and Physics,
CRC Press, 79th edition (1998-1999), p 6-134
As the temperature of any solid decreases, its heat capacity eventually begins to decrease. At temperatures near absolute zero, the
heat capacity approaches zero. The graph in Figure 6 shows the shape of the heat capacity versus temperature curve for solid
mercury. The shape of this curve can be predicted from a very simple model for the energy modes available to the atoms in a solid.
Albert Einstein developed this model in 1907. Einstein’s model for the heat capacity of a solid was an important milestone in the
development of quantum mechanics. Since then, the basic ideas have been extended and refined to create more detailed theories
that achieve good quantitative agreement with the experimental results for particular substances. We discuss Einstein’s treatment in
Section 22.6.
This page titled 7.14: Heat Capacities of Solids- the Law of Dulong and Petit is shared under a CC BY-SA 4.0 license and was authored, remixed,
Any mathematical expression that involves only state functions must itself be a state function. We could define Γ = E P V T , 2 2
and Γ would be a state function. However, it is not a useful state function. We can define several state functions that have the units
of energy and that turn out to be particularly useful. One of them is named enthalpy and is customarily represented by the symbol
H . We define enthalpy:
H = E +PV .
One reason that enthalpy is a useful state function emerges if we examine the change in H when the system pressure is equal to the
applied pressure, and both are constant. (When these conditions are satisfied, we usually denote the heat accepted by the system as
“q .”) If all of the work is pressure–volume work, we have
P
ΔH = ΔE + P ΔV
= qP + w + P ΔV
= qP − Papplied ∫ dV + P ΔV
= qP − (Papplied − P ) ΔV
= qP
If these conditions are satisfied, the enthalpy change is the same thing as the heat added to the system. When we want to express
the requirement that the system and applied pressures are equal and constant, we often just say that the process “occurs at a
constant pressure.” This is another convenient figure of speech. It also reflects our expectation that the system pressure and the
applied pressure will equilibrate rapidly in most circumstances.
For an ideal gas, the molar energy depends only on temperature. Since P V = RT for an ideal gas, P V depends only on
temperature. Hence, the molar enthalpy of an ideal gas also depends only on temperature. For an ideal gas, we have the parallel
relationships:
∂E ∂E ∂q
( ) =( ) =( ) = CV
∂T ∂T ∂T
V P V
and
∂H ∂H ∂q
( ) =( ) =( ) = Cp
∂T ∂T ∂T
V P P
Earlier we asserted that, while energy is a state function, heat and work are not. Hess’s law, as originally formulated in 1840, says
that the heat changes for a series of chemical reactions can be summed to get the heat change for the overall process described by
the sum of the chemical reactions. This amounts to saying that heat is a state function. As it stands, this is a contradiction. The
resolution is, of course, that Hess’s law was formulated for a series of chemical reactions that occur at the same constant pressure.
Then the heat involved in each step is the enthalpy change for that step, and since enthalpy is a state function, there is no
contradiction. Modern statements of Hess’s law frequently forego historical accuracy in favor of scientific accuracy to assert that
the enthalpy change for a series of reactions can be summed to get the enthalpy change for the overall process. Thus revised, Hess’s
law ceases to be a seminal but imperfect conjecture and becomes merely a special case of the principle that enthalpy is a state
function.
This page titled 7.15: Defining Enthalpy, H is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
If a system is in thermal contact with its surroundings, a reversible change can involve the exchange of heat between the system
and the surroundings. In Chapter 6, we make a number of important observations about the nature of any heat transfer that occurs
during a reversible process. Let us review these ideas.
A system can undergo a change in which it accepts (or liberates) heat while its temperature remains constant. If we boil a liquid at
constant pressure, a thermometer immersed in the liquid continues to show the same temperature even though we add more and
more heat energy. The added heat is used within the system to convert the liquid to its vapor. If the liquid is stirred well, any
localized temperature excursions away from the equilibrium temperature are small; it is a good approximation to say that the
temperature of the system is homogenous throughout the system and that it has a constant value.
Nevertheless, for a finite boiling rate we recognize that the idea of an isothermal process is indeed an approximation. For heat
transfer to occur from the surroundings to the system, the surroundings must be at a higher temperature than the system. The
portion of the system in immediate contact with the wall of the vessel must be at a higher temperature than the portion in the
interior of the vessel.
When we think about a constant-temperature system undergoing a reversible change while in thermal contact with its surroundings,
we imagine that heat can be transferred in either direction with equal facility. If the system is taking up heat as the process
proceeds, we imagine that we can reverse the direction of the change simply by changing the direction of heat transfer. Heat will
flow from the system to the surroundings, and the process will run backwards. We can reverse the direction of heat flow by
changing the temperature of the surroundings. Initially the surroundings must be hotter than the system. To reverse the direction of
heat flow, we must make the temperature of the surroundings less than that of the system. Since a reversible process is one whose
direction can be reversed by an arbitrarily small change in some state function, the original temperatures must be arbitrarily close to
one another.
For a system that exchanges heat with its surroundings, a process can be reversible only if the temperatures of the system and the
surroundings are arbitrarily close to one another. In a reversible process, net heat transfer occurs between two entities—the
system and its surroundings—that are arbitrarily close to thermal equilibrium. Such a process is an idealization. As we have noted
several times, a reversible process is a creature of theory that is merely approximated in real systems. A reversible process does not
have to be a constant-temperature process. If the temperatures of system and surroundings change simultaneously, they can remain
arbitrarily close to one another throughout the process. Nor must a system undergoing reversible change be in thermal contact with
its surroundings. A system can undergo a reversible change adiabatically.
Finally, we have noted that the term “isothermal process” is often intended to mean a constant-temperature thermally-reversible
process. However, the same words are frequently intended to indicate only that the final temperature of the system is the same as
the initial temperature. This is the case whenever the “isothermal process” is a spontaneous process. The intended meaning is
usually clear from the context.
This page titled 7.16: Heat Transfer in Reversible Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
To develop the theory of thermodynamics, we must be able to model the thermodynamic properties of gases as functions of
pressure, temperature, and volume. To do so, we consider processes in which the volume of a gas changes. For the expansion (or
compression) of a gas to be a reproducible process, the exchange of heat between the system and its surroundings must be
controlled. There are two straightforward ways to do this. We can immerse the system in a constant temperature bath whose
temperature is the same as that of the system; in this case, ΔT = 0 , and we can say that the process is isothermal.
Alternatively, we can isolate the system so that it cannot exchange heat with the surroundings; in this case q = 0 , and the process is
said to be adiabatic. In §7 we find that the work done on a system when its volume changes by dV under the influence of an
applied pressure, P , is
applied
dw = −Papplied dV .
Any expansion of a system in which the applied pressure is less than the system pressure can be called a free expansion. In Section
10.14 we consider the adiabatic expansion of a real gas against a constant applied pressure—a process known as a Joule-Thomson
expansion. We find that we must introduce a new parameter—the Joule-Thomson coefficient—in order to describe the behavior of
a real gas in a free expansion. The Joule-Thomson coefficient varies with pressure and temperature.
Literally, an isothermal process is one in which the temperature of the system remains the same throughout the process. However,
we often use the term to mean merely that the process occurs while the system is in thermal contact with constant-temperature
surroundings. The free expansion of a gas is an irreversible process; in principle, the temperature of a gas undergoing a free
expansion is not a meaningful quantity. When we talk about an isothermal free expansion of a gas, we mean that the final
temperature is the same as the initial temperature.
Here we consider the behavior of ideal gases, and we begin by considering the limiting case of a free expansion in which the
applied pressure is zero. Physically, this corresponds to the expansion of a system into a (very large) evacuated container. Under
this condition, dw = 0 , and the energy change is dE = dq . For one mole of any substance, C = (∂E/∂T ) . If only pressure–
V V
volume work is possible and the applied pressure is zero, we have dE = dq = C dT , and V
T2
ΔE = q = ∫ CV dT
T1
where T and T are the temperatures of the substance before and after the expansion, respectively.
1 2
At ordinary temperatures, C changes only slowly as the temperature changes. Over a short temperature range, it is usually a good
V
approximation to assume that C is constant. We have

V
ΔE = q = CV (T2 − T1 )
(one mole of any gas or other substance)

For a monatomic ideal gas, the energy change is exactly
3
ΔE = q = R (T2 − T1 )
2

The enthalpy change for any process is ΔH = ΔE + Δ (P V ) . If the system is one mole of an ideal gas, we have, because
Δ (P V ) = RΔT = R (T − T ) , 2 1
ΔH = CV (T2 − T1 ) + R (T2 − T1 ) = (CV + R) (T2 − T1 ) = CP (T2 − T1 )

For an isothermal free expansion against an applied pressure of zero, we have ΔT = 0 , and so neither the energy nor the enthalpy
of the gas changes. Since also dw = 0 , there can be no exchange of heat with the surroundings. We have
w = q = ΔT = ΔE = ΔH = 0
(free expansion, ideal gas)
For an adiabatic free expansion, we have dq = 0 and dw = 0 , and it follows again that w = q = ΔT = ΔE = ΔH = 0 . We see
that the isothermal and adiabatic expansions of an ideal gas into a vacuum are equivalent processes. If the expansion is opposed by
a non-zero applied pressure, the two processes cease to be equivalent.
This page titled 7.17: Free Expansion of a Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
In Section 7.16 we consider heat transfer in reversible processes. Similar considerations apply to the exchange of work between a
system and its surroundings. When we use a piston to compress a gas in a cylinder, we must apply sufficient inward force on the
piston to overcome the outward force applied by the gas. In any real system, it is necessary also to overcome the force of friction in
order to slide the piston into the cylinder. We ignore friction, imagining that we can make its effects arbitrarily small.
The gas can be compressed only if the applied pressure exceeds the gas pressure. If the applied pressure equals the gas pressure, the
piston remains stationary. If the applied pressure is greater than the gas pressure by any ever-so-small amount, the gas will be
compressed. Conversely, if the applied pressure is infinitesimally less than the gas pressure, the gas will expand. The work done
under such conditions is reversible work; an arbitrarily small change in the relative pressures can reverse the direction in which the
piston moves. We summarize these conditions by saying that reversible pressure–volume work can occur only if the system and its
surroundings are at mechanical equilibrium.
Now, let us think about calculating the reversible work for isothermally compressing a gas by sliding a piston into a cylinder. In any
real experiment, we must have P >P
applied , and any real experiment is necessarily irreversible. In a reversible experiment, we
gas
have P =P
applied = P , and the reversible work, w
gas , is rev
V2
rev
w =∫ −Papplied dV
V1
V2
=∫ −Pgas dV
V1
For one mole of an ideal gas, we have P = RT /V . Since the temperature is constant, the reversible isothermal work becomes
V2
RT V2
rev
w =∫ − dV = −RT ln
V1
V V1
where V and V are the initial and final volumes of the gas, respectively. This has a straightforward graphical interpretation. For
1 2
an ideal gas at constant temperature, −P is inversely proportional to V . As sketched in Figure 7, the reversible work
gas gas
corresponds to the area between this curve and the abscissa and between the initial, V , and the final, V , gas volumes.
1 2
Figure 7. Reversible versus irreversible expansion of an ideal gas.

In contrast, an irreversible expansion corresponds to movement of the piston when P > P , or equivalently,
gas applied
−P gas < −P . Therefore, the work done on the gas is less in the reversible case than it is in the irreversible case. (Both work
applied
terms are less than zero. The absolute value of the reversible work is greater than the absolute value of the irreversible work.) From
our definitions of reversible and irreversible pressure–volume work, we have dw < dw and w < w
rev irrev
, so long as the
rev irrev
initial and final states are the same in the irreversible process as they are in the reversible constant-temperature process. The shaded
area in Figure 7 represents the work done on the gas when the applied pressure is instantaneously decreased to the final pressure,
P , attained by the gas in the reversible process.
a
For the reversible process, the pressure–volume curve accurately depicts the state of the gas as the volume increase takes place. The
temperature of the gas is constant along this curve. While we can trace a similar line of pressure–volume points for the irreversible
expansion, this line does not define a set of intermediate states that the system occupies during the irreversible expansion. The state
of the gas is well defined only in the equilibrium state that precedes the irreversible pressure drop and in the equilibrium state that
the system ultimately attains. It is convenient to describe these two processes as a reversible process and a spontaneous process that
“take the system from the same initial state to the same final state.” However, this language obscures a significant point. In the
initial state for the reversible process, we have
Pgas = RT / V1 = Papplied
In the initial state for the spontaneous process, we have

RT
Pgas =
V1
and
RT
Papplied =
V2
What we mean, of course, is that the values of all of the state functions for the hypothetical initial state of the spontaneous process
are the same as those for the equilibrium initial state of the reversible process.
So long as we can say that the process takes the system from the same initial state to the same final state, a similar argument can be
made for reversible and irreversible work of any kind. Whatever the force, the isothermal reversible work done on the system is
always less than the irreversible work for taking the system between the same two states. This is an important result. In Chapter 9,
we find that it is a logical consequence of the second law of thermodynamics.
Finally, let us consider a reversible process in which a system completes a pressure–volume cycle. The system traverses a closed
path in the pressure–volume plane. Such a path is depicted in Figure 8. We let the smallest and largest volumes reached during the
cycle be V and V , respectively. The closed path is composed of a high-pressure segment and a low-pressure segment that meet at
ℓ h
Vℓ and V . On each of these segments, the pressure is a function of volume. We let pressures on the high- and low-pressure
h
segments be P (V ) and P (V ), respectively. (In the interval $V_{\ell }<v_h$>, we have P (V ) > P (V ) > 0 . At the limiting
h ℓ h ℓ
volumes, we have P (V ) = P (V ) and P (V ) = P (V ) .) The system temperature varies continuously around the closed path.
h h ℓ h h ℓ ℓ ℓ
The work done on the system as it traverses the high-pressure segment from V to V is represented in Figure 8 by area A .
ℓ h h
Figure 8. Reversible pressure-volume work in a cycle.

We have
Vh
wh (Vℓ →Vh ) = ∫ +P h (V ) dV = Ah < 0

Vℓ
The work done on the system as it traverses the low-pressure segment from V to V is represented by area A . We have
ℓ h ℓ
Vh
wℓ (Vℓ → Vh ) = ∫ −P ℓ (V ) dV = Aℓ < 0
Vℓ
When the low-pressure segment is traversed in the opposite direction, we have w (V ℓ h → Vℓ ) = −Aℓ > 0 . When the system
traverses the cycle in the counterclockwise direction, the net work done on the system is
wnet = wh (Vℓ → Vh ) + wℓ (Vh → Vℓ )
Vh Vℓ
=∫ −Ph (V ) dV + ∫ −P ℓ (V ) dV
Vℓ Vh
=∮ −P (V ) dV
= Ah − Aℓ < 0
Thus the net work done on the system is represented on the graph by the area Ah − Aℓ , which is just the (negative) area in the
pressure–volume plane that is bounded by the closed path.
This page titled 7.18: Reversible vs. Irreversible Pressure-Volume Work is shared under a CC BY-SA 4.0 license and was authored, remixed,
For an isothermal reversible expansion of an ideal gas, we have by definition that ΔT = 0 . Since the energy of an ideal gas
depends only on the temperature, a constant temperature implies constant energy, so that ΔE = 0 = q + w . Using the rev rev
equation we find for w in the previous section, we have

rev
V2
rev rev
−q =w = −RT ln
V1
(ideal gas, isothermal reversible expansion)

where V1 andare the initial and final volumes, respectively. Since enthalpy is defined as
V2 H = E +PV , we have
ΔH = ΔE + Δ (P V ) = ΔE + Δ (RT ) = 0 .
For the spontaneous isothermal expansion of an ideal gas from V to V against a constant applied pressure, we again have
1 2
ΔT = ΔE = ΔH = 0 . These are state functions, and the amounts by which they change in this spontaneous process must be the
same as those for the reversible process between the same two states. The heat and work exchanged in the spontaneous process are
different, demonstrating that heat and work are not state functions. We have
V2
Papplied
−q = w = ∫ −Papplied dV = −Papplied (V2 − V1 ) = RT ( − 1)
V1 P1
(one mole ideal gas, isothermal free expansion, P applied >0 )
This page titled 7.19: Isothermal Expansions of An Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
Consider an ideal gas that undergoes a reversible adiabatic expansion from an initial state, specified by known values V and T , to 1 1
a new state in which the value of the volume, V , is known but the value of the temperature, T , is not known. For an adiabatic
2 2
reversible process, q = 0 , and w = ΔE . Since (∂E/∂T ) = C , we have dE = C dT , so that

V V V
T2
w = ΔE = ∫ CV dT
T1
For any gas, we can assume that C is approximately constant over a small temperature range. Taking
V CV to be constant in the
interval $T_1<t_2$>, we have w = ΔE = C (T − T ) . We obtain the enthalpy change from
V 2 1
ΔH = ΔE + Δ (P V ) = ΔE + Δ (RT ) = CV (T2 − T1 ) + R (T2 − T1 ) = CP (T2 − T1 )
where we use our ideal-gas result from Section 7.16, C P = CV + R .

While these relationships yield the values of the various thermodynamic quantities in terms of the temperature difference, T − T , 2 1
we have yet to find the final temperature, T . To find T , we return to the first law: dE = dq + dw . Substituting for dE , dq, and
2 2
dw , and making use of the ideal gas equation, we have
RT
CV dT = −P dV = − dV
V
from which, by separation of variables, we have

T2 V2
∫
T1
CV
dT
T
= −R ∫
V1
dV
V
(one mole ideal gas, reversible adiabatic expansion)
If we know C as a function of temperature, we can integrate to find a relationship among T , T , V , and V . Given any three of
V 1 2 1 2
these quantities, we can use this relationship to find the fourth. If C is independent of temperature, as it is for a monatomic ideal
V
gas, we have
R/CV
T2 R V2 R V1 V1
ln =− ln = ln = ln ( )
T1 CV V1 CV V2 V2
so that
R/CV
T2 V1
=( )
T1 V2
(monatomic ideal gas, reversible adiabatic expansion)

For the spontaneous adiabatic expansion of an ideal gas against a constant applied pressure, we have dq = 0 , so that dE = dw ,
and C dT = −P
V dV . Given the initial conditions, we can find the final temperature from
applied
T2 V2
RT2 RT1 Papplied T1
∫ CV dT = ∫ −Papplied dV = −Papplied ( − ) = R( − T2 )
T1 V1 Papplied P1 P1
(spontaneous adiabatic process)

The changes in the remaining state functions can then be calculated from the relationships above. In this spontaneous adiabatic
process, all of the other thermodynamic quantities are different from those of a reversible adiabatic process that reaches the same
final volume.
This page titled 7.20: Adiabatic Expansions of An Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
7.21: Problems
1. Which of the following differential expressions are exact?
a. df = ydx + xdy
b. df 2
= 2x y dx + 2 x ydy
2
c. df = 2xydx + 2 x ydy
2
d. df = [(1 − xy) e
−xy 2
] dx − [ x e
−xy
] dy
e. df = (cosx cosy ) dx − (sinxsiny ) dy
f. df = (cosx cosy ) dx − (siny ) dy
2. Show that df =e
−y
dx − x e
−y
dy is exact. Find f (x, y) by integrating the dx term. Find f (x, y) by integrating the dy term.
3. A marble of mass m is free to move on a surface whose height above the x, y-plane is h = ax 2
+ by
2
.
a. What is the gravitational potential energy of the marble expressed as a function of x and y , E (x, y)?
b. The force experienced by the marble due to gravity is the vector function
⇀ ∂E ⇀ ∂E ⇀
f (x, y) = −∇E (x, y) = −( ) i −( ) j
∂x ∂y
y x
⇀
What is f (x, y) on this surface?
c. What is the differential of E ? Is dE exact or inexact?
⇀ ⇀ ⇀ ⇀
d. The vector description of a general path, {(x, y)}, is the position vector, ⇀
r = x i + y j , and so d r = dx i
⇀
+ dy j . If we
push the marble up the surface from point (0, 0, E (0, 0)) to point (2, 2, E (2, 2)) along the path y = x , express ⇀
d r as a vector
function of dx.
e. If we push the marble along the path in part d with a force just large enough to overcome the force of gravity, what is the
increment of work, dw, associated with an increment of motion, d ⇀
r ?
f. How much work must we do if we are to move the marble from (0, 0, E (0, 0)) to point (2, 2, E (2, 2)) along the path in part d,
using the force in part e? What is the relationship between this amount of work and the change in the energy of the marble during
this process?
g. Suppose that we push the marble up the surface from point (0, 0, E (0, 0)) to point (2, 2, E (2, 2)) along the path 2
y = x /2 .
What is the vector description of this path?
h. How much work must we do if we are to move the marble from point (0, 0, E (0, 0)) to point (2, 2, E (2, 2)) along the path in
part g using the force in part b? Compare this result to your result in part f. Explain.
4. Consider the plane, f (x, y) = 1 − 2x − 3y . What is df for this surface? Evaluate Δf = f (1, 1) − f (−1, −1) by integrating
df along each of the following paths:
a. y = x
b. y = x 3
c. y = 1 + x − x 2
d. y = sin (πx/2)
5. A 2.00 mole sample of a monatomic ideal gas is expanded reversibly and isothermally at 350 K from 5.82 L to 58.20 L. How
much work is done on the gas? What are q, ΔE, and ΔH for the gas in this process?
6. A 2.00 mole sample of a monatomic ideal gas is expanded irreversibly from 5.82 L to 58.20 L at a constant applied pressure
equal to the final pressure of the gas. The initial and final temperatures are 350 K. How much work is done on the gas? What are q,
ΔE, and ΔH for the gas in this process? Compare w , q , ΔE, and ΔH for this process to the corresponding quantities for the
process in problem 5. Compare the initial and final states of the gas to the corresponding states in problem 5.
7. A 2.00 mole sample of a monatomic ideal gas is expanded reversibly and adiabatically from 5.82 L to 58.20 L. The initial
temperature is 350 K. What is the final temperature? What are the initial and final pressures? How much work is done on the gas?
What are q, ΔE, and ΔH for the gas in this process?
8. The equation of state for a “hard-sphere gas” is P (V − nb) = nRT , where n is the number of moles and b is the molar volume
of the hard spheres. How much work is done on this gas when n moles of it expand reversibly and isothermally from V to V ? 1 2
9. Strictly speaking, can the spontaneous expansion of a real gas be isothermal? Can it be free? Can it be adiabatic? Can the
reversible expansion of a gas be isothermal? Can it be free? Can it be adiabatic?
10. Consider a machine that operates in a cycle and converts heat into a greater amount of work. What would happen to the energy
of the universe if this machine could be operated in reverse?
11. Show that the product of pressure and volume has the units of energy.
12. Give a counter-example to prove that each of the following propositions is false:
a. If X is a state function, X is conserved.
b. If X is an extensive quantity that satisfies X + X

^
= 0 , X is a state function.
Notes
1
Since the temperature of the water increases and the process is to be reversible, we must keep the temperature of the thermal
reservoir just dT greater than that of the water throughout the process. We can accomplish this by using a quantity of ideal gas as
the heat reservoir. By reversibly compressing the ideal gas, we can reversibly deliver the required heat while maintaining the
required temperature. We consider this operation further in Section 12.5.
CHAPTER OVERVIEW
8: Enthalpy and Thermochemical Cycles

8.1: Enthalpy
8.9: Calorimetry
8.10: Problems
This page titled 8: Enthalpy and Thermochemical Cycles is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
1
8.1: Enthalpy
In Chapter 7, we introduce the enthalpy function, which we define as
H = E + PV
When the only form of work possible is pressure–volume work and a system change occurs at constant pressure, the enthalpy
change is synonymous with the heat added to the system.
H = qP (only PV work)
Since we define C = (∂q/∂T ) , it follows that
P P
(∂H /∂T )
P
= CP . Recalling our earlier discovery that (∂E/∂T )
V
= CV , we
have the important parallel relationships:
∂q ∂H
CP = ( ) =( )
∂T ∂T
P P
and
∂q ∂E
CV = ( ) =( )
∂T ∂T
V V
We can find the enthalpy change for heating a substance at constant pressure by integrating its constant-pressure heat capacity, C , P
over the change in temperature. That is, C = (∂H /∂T ) implies that
P P
T2
ΔH = ∫ CP dT
T1
Similarly, we have
T2
ΔE = ∫ CV dT
T1
for a process in which a substance is heated at constant volume.

One reason that the enthalpy function is useful in chemistry is that many processes are carried out at conditions (constant pressure,
only P V work) where the enthalpy change is synonymous with the heat exchanged. The heat exchanged in a process is frequently
an important consideration. If we want to carry out an endothermic process, we must provide means to add sufficient
heat. If we want to carry out an exothermic process, we may have to make special arrangements to safely transfer the heat evolved
from the system to its surroundings.
One of our principal objectives is to predict whether a given process can occur spontaneously. We will see that the heat evolved in a
process is not a generally valid predictor of whether or not the process can occur spontaneously; however, it is true that a very
exothermic process is usually one that can occur spontaneously. (We will see that ΔH < 0 is a rigorous criterion for whether the
process can occur spontaneously if and only if the process is one for which both the entropy and the pressure remain constant.)
This page titled 8.1: Enthalpy is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
Page ID
151705
Figure 1. A cyclic path on the phase diagram for water.

Because enthalpy is a state function, the enthalpy change in going between any two states of a system is independent of the path.
For a series of changes that restore a system to its original state, the sum of all the enthalpy changes must be zero. This fact enables
us to find the enthalpy changes for many processes for which it is difficult to measure heat and work directly. It is easiest to see
what is involved by considering a specific example. Figure 1 shows a cyclic path, A→A*→B→…→A, superimposed on a not-to-
scale presentation of the phase diagram for water. Let us look at the sublimation of ice at the melting point of pure water. The
sublimation of ice is the conversion of pure ice to pure water vapor. (The melting point of pure water is the temperature at which
pure ice is at equilibrium with pure liquid water at a pressure of one atmosphere; it is represented by points A and A* on the
diagram. We want to find the enthalpy of sublimation at the temperature and pressure represented by points D and D*.)
Points A, A*, D, and D* are all at the same temperature; this temperature is about 273.153 K or 0.003 C. (This temperature is very
slightly greater than 273.15 K or 0 C—which is the temperature at which ice and water are at equilibrium in the presence of air at a
total pressure of one atmosphere.) We want to calculate the enthalpy change for the equilibrium conversion of one mole of ice to
gaseous water at the pressure where the solid–gas equilibrium line intersects the line T = 273.153 K ≈ 0 C .
On the diagram, this sublimation pressure is represented as P and the sublimation process is represented as the transition from
sub
D* to D. P is less than the triple-point pressure of 611 Pa or 6.03 × 10 atm. However, the difference is less than
sub
−3
atm or 1.4 Pa. In equation form, the successive states traversed in this cycle are:
−5
1.4 × 10
A (ice at 0 C and 1 atm) →

A* (water at 0 C and 1 atm) →
B (water at 100 C and 1 atm) →
B* (water vapor at 100 C and 1 atm) →
C (water vapor at 100 C and P sub )→
D (water vapor at 0 C and P sub )→
D*(ice at 0 C and P sub )→
A (ice at 0 C and 1 atm)
We select these steps because it is experimentally straightforward to find the enthalpy change for all of them except the sublimation
step (D*→D). All of these steps can be carried out reversibly. This strategy is useful in general. We make extensive use of
reversible cycles to find thermodynamic information for chemical systems. The enthalpy changes for these steps are
H2 O (s, 0 C, 1 atm) → H 2O (liq, 0 C, 1 atm)
∗
ΔH (A→A ) = Δf us H
H2 O (liq, 0 C, 1 atm) → H 2O (liq, 100 C, 1 atm)
372.15 K
∗
ΔH (A →B) = ∫ CP (H2 O, liq) dT
273.15 K
H2 O (liq, 100 C, 1 atm) → H 2O (g, 100 C, 1 atm)

∗
ΔH (B→B ) = Δvap H
H2 O (g, 100 C, 1 atm) → H 2O (g, 100 C, P sub )

P =Psub
∂H (H2 O, g)
∗
ΔH (B →C) = ∫ ( ) dP ≈ 0
P =1
∂P
T
H2 O (g, 100 C, P sub ) →H 2O (g, 0 C, P sub )

272.15 K
ΔH (C→D) = ∫ CP (H2 O, g) dT
373.15 K
H2 O (g, 0 C, P sub )→H 2O (s, 0 C, Psub )

∗
ΔH (D→D ) = −Δ sub H
H2 O (s, 0 C, P
sub ) →H2O (s, 0 C, 1 atm)
P =1
∂H (H2 O, s)
∗
ΔH (D →A) = ∫ ( ) dP ≈ 0
P =Psub
∂P
T
Summing the enthalpy changes around the cycle gives

372.15 K
0 = Δf us H + ∫ CP (H2 O, liq) dT + Δvap H

273.15 K
272.15 K
∗
+ΔH (B →C) + ∫ CP (H2 O, g) dT −Δ sub H
373.15 K
∗
+ΔH (D →A)
Using results that we find in the next section, H (B ∗

→C) ≈ 0 and ΔH (D ∗
→A) ≈ 0 , we have
372.15 K
0 = Δf us H + ∫ CP (H2 O, liq) dT + Δvap H

273.15 K
272.15 K
+∫ CP (H2 O, g) dT −Δ sub H
373.15 K
The enthalpy of fusion, the enthalpy of vaporization, and the heat capacities are measurable in straightforward experiments. Their
values are given in standard compilations, so we are now able to evaluate Δ H , a quantity that is not susceptible to direct
sub
measurement, from other thermodynamic quantities that are. (See Problem 8.)
This page titled 8.2: Using Thermochemical Cycles to Find Enthalpy Changes is shared under a CC BY-SA 4.0 license and was authored,
Let us look briefly at the approximations ΔH (B →C) ≈ 0 and ΔH (D →A) ≈ 0 that we used in Section 8.2. In these steps, the
∗ ∗
pressure changes while the temperature remains constant. In Chapter 10, we find a general relationship for the pressure-dependence
of a system’s enthalpy:
∂H ∂V
( ) = −T ( ) +V
∂P ∂T
T P
This evaluates to zero for an ideal gas and to a negligible quantity for many other systems.
For liquids and solids, information on the variation of volume with temperature is collected in tables as the coefficient of thermal
expansion, α , where
1 ∂V
α = ( )
V ∂T
P
Consequently, the dependence of enthalpy on pressure is given by

∂H
( ) = V (1 − αT )
∂P
T
For ice, α ≈ 50 × 10 K −6 −1
and the molar volume near 0 C is 19.65 cm mol . The enthalpy change for compressing one mole
3 −1
of ice from the sublimation pressure to 1 atm is ΔH (D →A) = 2 J mol .

∗ −1
To find the enthalpy change for expanding one mole of water vapor at 100 C from 1 atm to the sublimation pressure, we use the
virial equation and tabulated coefficients for water vapor to calculate (∂H /∂P ) . We find ΔH (B →C) = 220 J mol . (See
398 K
∗ −1
problem 9.)
This page titled 8.3: How Enthalpy Depends on Pressure is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
A useful convention makes it possible to tabulate enthalpy data for individual compounds in such a way that the enthalpy change
for any chemical reaction can be calculated from the tabulated information for the reaction’s reactants and products. The
convention comprises the following rules:
I. At any particular temperature, we define the standard state of any liquid or solid substance to be the most stable form of
that substance at a pressure of one bar. For example, for water at −10 C, the standard state is ice at a pressure of one bar; at
+10 C, it is liquid water at a pressure of one bar.
II. At any particular temperature, we define the standard state of a gas to be the ideal gas standard state at that temperature. By the
ideal gas standard state, we mean a finite low pressure at which the real gas behaves as an ideal gas. We know that it is possible to
find such a pressure, because any gas behaves as an ideal gas at a sufficiently low pressure. Since the enthalpy of an ideal gas is
independent of pressure, we can also think of a substance in its ideal gas standard state as a hypothetical substance whose
pressure is one bar but whose molar enthalpy is that of the real gas at an arbitrarily low pressure.
III. For any substance at any particular temperature, we define the standard enthalpy of formation as the enthalpy change for a
reaction in which the product is one mole of the substance and the reactants are the compound’s constituent elements in
their standard states.
For water at –10 C, this reaction is
1
H2 (g, −10 C, 1 bar) + O2 (g, −10 C, 1 bar) →H2 O (s, −10 C, 1 bar)
2
For water at +10 C, it is

1
H2 (g, +10 C, 1 bar) + O2 (g, +10 C, 1 bar) →H2 O (liq, +10 C, 1 bar)
2
For water at +110 C, it is

1
H2 (g, +110 C, 1 bar) + O2 (g, +110 C, 1 bar) →H2 O (g, +110 C, 1 bar)
2
IV. The standard enthalpy of formation is given the symbol Δ H , where the superscript degree sign indicates that the
f
o
reactants and products are all in their standard states. The subscript, f , indicates that the enthalpy change is for the
formation of the indicated compound from its elements. Frequently, the compound and other conditions are specified in
parentheses following the symbol. The solid, liquid, and gas states are usually indicated by the letters “s”, “ℓ ” (or “liq”), and “g”,
respectively. The letter “c” is sometimes used to indicate that the substance is in a crystalline state. In this context, specification of
the gas state normally means the ideal gas standard state.
Thermochemical-data tables that include standard enthalpies of formation can be found in a number of publications or on the
internet. For some substances, values are available at a number of temperatures. For substances for which less data is available,
these tables usually give the value of the standard enthalpy of formation at 298.15 K. (In this context, 298.15 K is frequently
abbreviated to 298 K.)
V. For any element at any particular temperature, we define the standard enthalpy of formation to be zero. When we define
standard enthalpies of formation, we choose the elements in their standard states as a common reference state for the enthalpies of
all substances at a given temperature. While we could choose any arbitrary value for the enthalpy of an element in its standard
state, choosing it to be zero is particularly convenient.
This page titled 8.4: Standard States and Enthalpies of Formation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The ideal gas standard state is a useful invention, which has additional advantages that emerge as our development proceeds. For
permanent gases—gases whose behavior is approximately ideal anyway—there is a negligible difference between the enthalpy in
the ideal gas state and the enthalpy at 1 bar.
For volatile substances that are normally liquid or solid at 1 bar, the ideal gas standard state becomes a second standard state. For
such substances, data tables frequently give the standard enthalpy of formation for both the condensed phase (designated
Δ H (liq) or Δ H (s) ) and the ideal gas standard state (designated Δ H (g)). For example, the CODATA values for the
o o o 1
f f f
standard enthalpies of formation for liquid and ideal-gas methanol are −239.2 and −201.0 kJ mol , respectively, at 298.15 K.
−1
The difference between these values is the enthalpy change in vaporizing one mole of liquid methanol to its ideal gas standard state
at 298.15 K:
C H3 OH (liq, 298.15 K, 1 bar) →C H3 OH (ideal gas, 298.15 K, ∼ 0 bar)
Since this is the difference between the enthalpy of methanol in its standard state as an ideal gas and methanol in its standard state
as a liquid, we can call this difference the standard enthalpy of vaporization for methanol:
o o o −1
Δvap H = Δf H (g, 298.15 K, ∼ 0 bar) − Δf H (g, 298.15 K, 1 bar) = 37.40 kJ mol
This is not a reversible process, because liquid methanol at 1 bar is not at equilibrium with its vapor at an arbitrarily low pressure at
298.15 K.
Note that Δ H is not the same as the ordinary enthalpy of vaporization, Δ H. The ordinary enthalpy of vaporization is the
vap
o
vap
enthalpy change for the reversible vaporization of liquid methanol to real methanol vapor at a pressure of 1 atm and the normal
boiling temperature. We write it without the superscript degree sign because methanol vapor is not produced in its standard state.
For methanol, the normal boiling point and enthalpy of vaporization are 337.8 K and 35.21 kJ mol , respectively.
2 −1
We can devise a cycle that relates these two vaporization processes to one another: Summing the steps below yields the process for
vaporizing liquid methanol in its standard state to methanol vapor in its standard state.
1. C H 3 OH (liq, 298.15 K, 1 bar) → C H OH (liq, 337.8 K, 1 bar) Δ
3 (1) H
2. C H 3 OH (liq, 337.8 K, 1 bar) → C H OH (liq, 337.8 K, 1 atm) Δ

3 (2)
H
3. C H 3 OH (liq, 337.8 K, 1 atm) → C H OH (g, 337.8 K, 1 atm)

3
Δ(3) H = Δvap H
4. C H 3 OH (g, 337.8 K, 1 atm) → C H OH (g, 337.8 K, ∼0 bar) Δ H

3 (4)
5. C H 3 OH (g, 337.8 K, ∼0 bar) → C H OH (g, 298.15 K, ∼0 bar) Δ

3 H (5)
Thus, we have
o
Δvap H = Δ(1) H + Δ(2) H + Δvap H + Δ(4) H + Δ(5) H
Δ(1) H and Δ H can be evaluated by integrating the heat capacities for the liquid and gas, respectively. Δ H and Δ H can be
(5) (2) (4)
evaluated by integrating (∂H /∂P ) for the liquid and gas, respectively. Δ H is negligible. (For the evaluation of these
T (2)
quantities, see problem 10.)
This page titled 8.5: The Ideal Gas Standard State is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
The benefit of these conventions is that, at any particular temperature, the standard enthalpy change for a reaction
aA + bB + … → cC + dD + …
which we designate as Δ rH
o
, is given by
o o o o o
Δr H = c Δf H (C ) + dΔf H (D) + … − aΔf H (A) − b Δf H (B) − …
 
product enthalpies reactant enthalpies
If we have the enthalpies of formation, we can compute the enthalpy change for the reaction. We can demonstrate this by writing
out the chemical equations corresponding to the formation of A, B, C, and D from their elements. When we multiply these
chemical equations by the appropriately signed stoichiometric coefficient and add them, we obtain the chemical equation for the
indicated reaction of A and B to give C and D. (See below.) Because enthalpy is a state function, the enthalpy change that we
calculate this way will be valid for any process that converts the specified reactants into the specified products.
The oxidation of methane to methanol is a reaction that illustrates the value of this approach. The normal products in the oxidation
of methane are, of course, carbon dioxide and water. If the reaction is done with an excess of methane, a portion of the carbon-
containing product will be carbon monoxide rather than carbon dioxide. In any circumstance, methanol is, at best, a trace product.
Nevertheless, it would be very desirable to devise a catalyst that quantitatively—or nearly quantitatively—converted methane to
methanol according to the equation
1
CH + O ⟶ CH OH
4 2 2 3
(This is frequently called a selective oxidation, to distinguish it from the non-selective oxidation that produces carbon dioxide and
water.)
If the catalyst were not inordinately expensive or short-lived, and the operating pressure were sufficiently low, this would be an
economical method for the manufacture of methanol. (Methanol is currently manufactured from methane. However, the process
involves two steps and requires a substantial capital investment.) If the cost of manufacturing methanol could be decreased
sufficiently, it would become economically feasible to convert natural gas, which cannot be transported economically unless it is
feasible to build a pipeline for the purpose, into liquid methanol, which is readily transported by ship. (At present, the economic
feasibility of marine transport of liquefied natural gas, LNG, is marginal, but it appears to be improving.) This technology would
make it possible to utilize the fuel value of known natural gas resources that are presently useless because they are located too far
from population centers.
When we contemplate trying to develop a catalyst and a manufacturing plant to carry out this reaction, we soon discover reasons
for wanting to know the enthalpy change. One is that the oxidative manufacture of methanol will be exothermic, so burning the
methanol produced will yield less heat than would be produced by burning the methane from which it was produced. We want to
know how much heat energy is lost in this way.
Another reason is that a manufacturing plant will have to control the temperature of the oxidation reaction in order to maintain
optimal performance. (If the temperature is too low, the reaction rate will be too slow. If the temperature is too high, the catalyst
may be deactivated in a short time, and the production of carbon oxides will probably be excessive.) A chemical engineer designing
a plant will need to know how much heat is produced so that he can provide adequate cooling equipment.
Because we do not know how to carry out this reaction, we cannot measure its enthalpy change directly. However, if we have the
enthalpies of formation for methane and methanol, we can compute this enthalpy change:
1
C(s) + 2 H (g) + O (g) ⟶ CH OH(g))
2 2 2 3
o
ΔH = Δf H (C H3 OH , g)
C H4 (g) → C (s) + 2 H2 (g)
o
ΔH = −Δ f H (C H4 , g)
1/2 O2 (g) → 1/2 O2 (g)
o
ΔH = −1/2 Δf H (O2 , g) = 0
Summing the reactions gives
1
CH (g) + O (g) → CH OH(g)
4 2 2 3
o
ΔH = Δr H
and summing the enthalpy changes gives

o o o o
Δr H = Δf H (C H3 OH , g) − Δf H (C H4 , g) − 1/2 Δf H (O2 , g)
Figure 2. A thermochemical cycle to find Δ rH

o
.
The diagram in Figure 2 shows how these conventions, and the fact that enthalpy is a state function, work together to produce, for
the reaction aA + bB + ⋯ → cC + dD + … , the result that the standard reaction enthalpy is given by
o o o o o
Δr H = c Δf H (C ) + dΔf H (D) + ⋯ − aΔf H (A) − b Δf H (B) − …
This cycle highlights another aspect of the conventions that we have developed. Note that Δ H is the difference between the r
o
enthalpies of formation of the separated products and the enthalpies of formation of the separated reactants. We often talk about
Δ H
r
o
as if it were the enthalpy change that would occur if we mixed a moles of A with b moles of B and the reaction proceeded
quantitatively to yield a mixture containing c moles of C and d moles of D. This is usually a good approximation. However, to
relate rigorously the standard enthalpy of reaction to the enthalpy change that would occur in a real system in which this reaction
took place, it is necessary to recognize that there can be enthalpy changes associated with the pressure–volume changes and with
the processes of mixing the reactants and separating the products.
Let us suppose that the reactants and products are gases in their hypothetical ideal-gas states at 1 bar, and that we carry out the
reaction by mixing the reactants in a sealed pressure vessel. We suppose that the reaction is then initiated and that the products are
formed rapidly, reaching some new pressure and an elevated temperature. (To be specific, we could imagine the reaction be the
combustion of methane. We would mix known amounts of methane and oxygen in a pressure vessel and initiate the reaction using
an electrical spark.) We allow the temperature to return to the original temperature of the reactants; there is an accompanying
pressure change.
Experimentally, we measure the heat evolved as the mixed reactants are converted to the mixed products, at the original
temperature. To complete the process corresponding to the standard enthalpy change, however, we must also separate the products
and bring them to a pressure of 1 bar. That is, the standard enthalpy of reaction and the enthalpy change we would measure are
related by the following sequence of changes, where the middle equation corresponds to the process whose enthalpy change we
actually measure.
(aA + bB) → (aA + bB)

separate reactants at P =1 bar homogenous mixture at P
ΔHcompression
(aA + bB) separate reactants at P

→ (aA + bB) homogeneous mixture at P
ΔHmixing
(aA + bB) → (cC + dD) ∗

homogeneous mixture at P homogeneous mixture at P
ΔHmeasured
(cC + dD) ∗ → (cC + dD) ∗

homogeneous mixture at P separate products at P
ΔHseparation
(cC + dD) ∗ → (cC + dD)

separate products at P separate products at P =1 bar
ΔHexpansion
Summing the reaction equations gives
(aA + bB) separate reactants at P =1 bar

→ (cC + dD) separate products at P =1 bar
o
Δr H
and summing the enthalpy changes for the series of steps gives the standard enthalpy change for the reaction:
o
Δr H = ΔHcompression + ΔHmixing + ΔHmeasured + ΔHseparation + ΔHexpansion
It turns out that the enthalpy changes for the compression, mixing, separation, and expansion processes are usually small compared
to Δ H . This is the principal justification for our frequent failure to consider them explicitly. For ideal gases, these enthalpy
r
o
changes are identically zero. (In Chapter 13, we see that the entropy changes for the mixing and separation processes are
important.)
When we call Δ H the standard enthalpy change “for the reaction,” we are indulging in a degree of poetic license. Since Δ H
r
o
r
o
is a computed difference between the enthalpies of the pure products and those of the pure reactants, the corresponding “reaction”
is a purely formal change, which is a distinctly different thing from the real-world process that actually occurs.
This page titled 8.6: Standard Enthalpies of Reaction is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
We have observed that C depends on volume and temperature, while C depends on pressure and temperature. Compilations of
V P
heat capacity data usually give values for C , rather than C . When the temperature-dependence of C is known, such
P V P
compilations usually express it as an empirical polynomial function of temperature. In Chapter 10, we find an explicit function for
the dependence of C on pressure:
P
2
∂CP ∂ V
( ) = −T ( )
2
∂P ∂T
T
P
If we have an equation of state for a substance, we can find this pressure dependence immediately. It is usually negligible. For ideal
gases, it is zero, and C is independent of pressure.
P
Compilations often give data for the standard state heat capacity, C , at a specified temperature. For condensed phases, this is the
o
P
heat capacity for the substance at one bar. For gases, this is the heat capacity of the substance in its ideal gas standard state.
300 K 400 K a (J) b (J K

−1
)
C_s 0 0 -1.482 0.03364
H2 (g) 0 0 27.853 0.00332
O2 (g) 0 0 27.221 0.00722
CH 4 (g) -74.656 -77.703 21.167 0.04866
CH 3 OH (g) -201.068 -204.622 21.737 0.07494
This page titled 8.7: Standard State Heat Capacities is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
In Section 8.6, we see how to use tabulated enthalpies of formation to calculate the enthalpy change for a particular chemical
reaction. Such tables typically give enthalpies of formation at a number of different temperatures, so that the enthalpy change for a
given reaction can also be calculated at these different temperatures; it is just a matter of repeating the same calculation at each
temperature.
We often need to find the enthalpy change associated with increasing the temperature of a substance at constant pressure. As we
observe in §1, this enthalpy change is readily calculated by integrating the heat capacity over the temperature change. We may want
to know, for example, the enthalpy change for increasing the temperature of one mole of methane from 300 K to 400 K, with the
pressure held constant at one bar. In Table 1, we find
o −1
Δf H (C H4 , g, 300 K) = −74.656 kJ mol
o −1
Δf H (C H4 , g, 400 K) = −77.703 kJ mol
We might be tempted to think that the difference represents the enthalpy change associated with heating the methane. This is not
so! The reason becomes immediately apparent if we consider a cycle in which we go from the elements to a compound at two
different temperatures. For methane, this cycle is shown in Figure 3.
Figure 3. A thermochemical cycle relating Δ f H

o
(C H4 ) at two temperatures.
The difference between the standard enthalpies of formation of methane at 300 K and 400 K reflects the enthalpy change for
increasing the temperatures of all of the reactants and products from 300 K to 400 K. That is,
o o
Δf H (C H4 , g, 400 K) − Δf H (C H4 , g, 300 K)
400 400 400
=∫ CP (C H4 , g) dT − ∫ CP (C , s) dT − 2 ∫ CP (H2 , g) dT
300 300 300
Over the temperature range from 300 K to 400 K, the heat capacities of carbon, hydrogen, and methane are approximated by
CP = a + bT , with values of a and b given in Table 1. From this information, we calculate the enthalpy change for increasing the
temperature of one mole of each substance from 300 K to 400 K at 1 bar: ΔH (C ) = 1, 029 J mol , −1
ΔH (H ) = 2, 902 J mol
2 , and ΔH (C H ) = 3, 819 J mol . Thus, from the cycle, we calculate:
−1
4
−1
o −1 −1
Δf H (C H4 , g, 400 K) = −74, 656 + 3, 819 − 1, 029 − 2 (2, 902) J mol = − 77, 670 J mol
The tabulated value is −77, 703 J mol . The two values differ by
−1
33 J mol
−1
, or about 0.04%. This difference arises from the
limitations of the two-parameter heat-capacity equations.
As another example of a thermochemical cycle, let us consider the selective oxidation of methane to methanol at 300 K and 400 K.
From the enthalpies of formation in Table 1, we calculate the enthalpies for the reaction to be
o
Δ H (300 K) = −126.412 kJ mol
r and Δ H (400 K) = −126.919 kJ mol . As in the previous example, we use the
−1
r
o −1
tabulated heat-capacity parameters to calculate the enthalpy change for increasing the temperature of one mole of each of these
gases from 300 K to 400 K at 1 bar. We find: ΔH (C H OH ) = 4, 797 J mol , ΔH (C H ) = 3, 819 J mol , and
3
−1
4
−1
ΔH (O ) = 2, 975 J mol
2 .−1
Figure 4. A thermochemical cycle relating ΔrH
o
at two temperatures.
The cycle is shown in Figure 4. Inspecting this cycle, we see that we can calculate the enthalpy change for warming one mole of
methanol from 300 K to 400 K by summing the enthalpy changes around the bottom, left side, and top of the cycle; that is,
1 −1 −1
ΔH (C H3 OH ) = 126, 412 + 3, 819 + ( ) 2, 975 − 126, 919 J mol = 4, 800 J mol
2
This is 3 J or about 0.06 % larger than the value obtained (4, 797 J) by integrating the heat capacity for methanol.
This page titled 8.8: How The Enthalpy Change for a Reaction Depends on Temperature is shared under a CC BY-SA 4.0 license and was
8.9: Calorimetry
Calorimetry is the experimental science of measuring the heat changes that accompany chemical or physical changes. The accurate
measurement of small amounts of heat is experimentally challenging. Nevertheless, calorimetry is an area in which great
experimental sophistication has been achieved and remarkably accurate measurements can be made. Numerous devices have been
developed to measure heat changes.
Some of these devices measure a (usually small) temperature change. Such devices are calibrated by measuring how much their
temperature increases when a known amount of heat is introduced. This is usually accomplished by passing a known electric
current through a known resistance for a known time. Other calorimeters measure the amount of some substance that undergoes a
phase change. The ice calorimeter is an important example of the latter method. In an ice calorimeter, the heat of the process is
transferred to a mixture of ice and water. The amount of ice that melts is a direct measure of the amount of heat released by the
process. The amount of ice melted can be determined either by direct measurement of the increase in the amount of water present
or by measuring the change in the volume of the ice–water mixture. (Since ice occupies a greater volume than the same mass of
water, melting is accompanied by a decrease in the total volume occupied by the mixture of ice and water.)
The processes that can be investigated accurately using calorimetry are limited by two important considerations. One is that the
process must go to completion within a relatively short time. No matter how carefully it is constructed, any calorimeter will
exchange thermal energy with its environment at some rate. If this rate is not negligibly small compared to the rate at which the
process evolves heat, the accuracy of the measurement is degraded. The second limitation is that the process must involve complete
conversion of the system from a known initial state to a known final state. When the processes of interest are chemical reactions,
these considerations mean that the reactions must be quantitative and fast.
Combustion reactions and catalytic hydrogenation reactions usually satisfy these requirements, and they are the most commonly
investigated. However, even in these cases, there can be complications. For a compound containing only carbon, hydrogen, and
oxygen, combustion using excess oxygen produces only carbon dioxide and water. For compounds containing heteroatoms like
nitrogen, sulfur, or phosphorus, there may be more than one heteroatom-containing product. For example, combustion of an
organosulfur compound might produce both sulfur dioxide and sulfur trioxide. To utilize the thermochemical data obtained in such
an experiments, a chemical analysis must be done to determine the amount of each oxide present.
This page titled 8.9: Calorimetry is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
8.10: Problems
1. One mole of an ideal gas reversibly traverses Cycle I above. Step a is isothermal. Step b is isochoric (constant volume). Step c is
isobaric (constant pressure). Assume C and C are constant. Find q, w, ΔE, and ΔH for each step and for the cycle. Prove
V P
CP =C V +R .
2. One mole of an ideal gas reversibly traverses Cycle II below. Step a is the same isothermal process as in problem 1. Step d is
adiabatic. Step e is isobaric. Assume C and C are constant. Find q, w, ΔE, and ΔH for each step and for the cycle.
V P
3. One mole of an ideal gas reversibly traverses Cycle III below. Step a is the same isothermal process as in
problem 1. Step f is adiabatic. Step g is isochoric. Assume C and C are constant. Find q, w, ΔE, and ΔH for each step and for
V P
the cycle.
4. One mole of an ideal gas reversibly traverses Cycle IV. Step h is isobaric. Step f is the same adiabatic process as in problem 3.
Step i is isochoric. Assume C and C are constant. Find q, w, ΔE, and ΔH for each step and for the cycle.
V P
5. Prove that the work done on the system is positive when the system traverses Cycle I. Note that Cycle I traverses the region of
the P V plane that it encloses in a counter-clockwise direction. Hint: Note that T < t . Show that V /V = T /T .
2 1 2 1 2 1
6. Cycles III and IV share a common adiabatic step. Express the work done in each of these cycles in terms of V1 , V2 , and T1 .
Prove that the work done in Cycle IV is greater than the work done in Cycle III.
7. Cycles I, II, and III share a common first step, a. Express V , T , and T in terms of V , V , and T . For V
3 3 4 1 2 1 1 = 10 L , V2 = 2 L ,
and T = 400 K , show that the work done decreases in the order Cycle I > Cycle III > Cycle II.
1
8. For water, the enthalpies of fusion and vaporization are 6.009 and 40.657 kJ mol , respectively. The heat capacity of liquid
−1
water varies only weakly with temperature and can be taken as 75.49 J mol K . The heat capacity of water vapor varies with
−1 −1
temperature:
−2
CP (H2 O, g) = 30.51 + (1.03 × 10 )T
where T is in degrees K and the heat capacity is in J mol −1

K
−1
. Estimate the enthalpy of sublimation of water.
9. If we truncate the virial equation (Z = 1 + B (T ) P + …) and make use of B (T ) = RT B
∗ ∗
(T ) , where B (T ) is the “second
virial coefficient” most often given in data tables, the molar volume is
¯
¯¯¯
RT
V = + B (T )
P
Show that
∂H dB
( ) = B (T ) − T ( )
∂P dT
T
The Handbook of Chemistry and Physics (CRC Press, 79 th

Ed., 1999, p. 6–25) gives the temperature dependence of B for water
vapor as
2 3 4
B = −1158 − 5157t − 10301 t − 10597 t − 4415 t
where t = (298.15/T ) − 1 , T is in degrees kelvin, and the units of B are cm mol . Estimate the enthalpy change when one
−3 −1
mole of water vapor at 1 atm and 100 C is expanded to the equilibrium sublimation pressure, which for this purpose we can
approximate as the triple-point pressure, 610 Pa. How does this value compare to the result of problem 8?
10. The heat capacities of methanol liquid and gas are 81.1 and 44.1 J mol −1
K
−1
, respectively. The second virial coefficient for
methanol vapor is
B = −1752 − 4694t
where t = (298.15/T ) − 1 , T is in degrees kelvin, and the units of B are cm mol . Referring to the discussion of methanol
−3 −1
vaporization in §5, calculate Δ H , Δ H , Δ H , Δ

(1) (4) (5) H . Compare this value of Δ
(vap)
o
H to the value given in the text.
(vap)
o
[Data from the Handbook of Chemistry and Physics, CRC Press, 79 Ed., 1999, p. 5-27 and p. 6-31.]
th
o −1
Molecular formula Name Δf H (kJ mol )
H2 O (liq) Water −285.8
CO (g) Carbon monoxide −110.5
C O2 (g) Carbon dioxide −393.5
C H4 (g) Methane −74.6
C2 H4 (g) Ethylene 52.4
C2 H6 (g) Ethane −84.0
C H3 C H2 OH (liq) Ethanol −277.6
C H3 CHO (liq) Acetaldehyde −192.2
C H3 C O2 H (liq) Acetic acid −484.3
C H3 C H2 CHO (liq) Propanal −215.6
C6 H6 (liq) Benzene 49.1
C6 H5 C O2 H (s) Benzoic acid −385.2
11. Using data from the table above, find the enthalpy change for each of the following reactions at 298 K.
(a) $C_2H_6\left(\mathrm{g}\right)+
ParseError: invalid DekiScript (click for details)
\ O_2\left(\mathrm{g}\right)\to CH_3CH_2OH\left(\mathrm{liq}\right)$
(b) $C_2H_4\left(\mathrm{g}\right)+
\ O_2\left(\mathrm{g}\right)\to CH_3CHO\left(\mathrm{liq}\right)$
(c) $C_2H_6\left(\mathrm{g}\right)+
\ O_2\left(\mathrm{g}\right)\to CH_3CHO\left(\mathrm{liq}\right)+H_2O\left(\mathrm{liq}\right)$
(d) C 6 H6 (liq) + C O2 (g) → C6 H5 C O2 H (s)
(e) $CH_3CHO\left(\mathrm{liq}\right)+
\ O_2\left(\mathrm{g}\right)\to CH_3CO_2H\left(\mathrm{liq}\right)$
(f) C H 4 (g) + H2 O (liq) → C O (g) + 3 H2 (g)
(g) $CH_4\left(\mathrm{g}\right)+H_2O\left(\mathrm{liq}\right)+
\ O_2\left(\mathrm{g}\right)\to CO_2\left(\mathrm{g}\right)+3\ H_2\left(\mathrm{g}\right)$

(h) C 2 H4 (g) + C O (g) + H2 (g) → C H3 C H2 C H O (liq)
Notes
1
Data compiled by The Committee on Data for Science and Technology (CODATA) and reprinted in D. R. Linde, Editor, The
Handbook of Chemistry and Physics, 79 Edition (1998-1999), CRC Press, Section 5.
th
2
D. R. Linde, op. cit., p. 6-104.
CHAPTER OVERVIEW
9: The Second Law - Entropy and Spontaneous Change

9.26: Problems
This page titled 9: The Second Law - Entropy and Spontaneous Change is shared under a CC BY-SA 4.0 license and was authored, remixed,
1
The first law of thermodynamics is concerned with energy and its properties. As we saw in Chapter 7, the first law arose from the
observation that the dissipation of mechanical work through friction creates heat. In a synthesis that was partly definition and partly
a generalization from experience, it was proposed that mechanical energy and heat are manifestations of a common quantity,
energy. Later, by further definition and generalization, the concept was expanded to include other forms of energy. The energy
concept evolved into the prescript that there exists a quantity (state function) that is conserved through any manner of change
whatsoever.
The element of definition arises from the fact that we recognize new forms of energy whenever necessary in order to ensure that the
conservation condition is satisfied. The element of experience arises from the fact that this prescript has resulted in a body of theory
and a body of experimental results that are mutually compatible. When we define and measure energy “correctly” we do indeed
find that energy is a state function and that it is conserved.
The theory of relativity introduced a significant expansion of the energy concept. For chemical processes, we can view mass and
energy conservation as independent postulates. For processes in which fundamental particles undergo changes and for systems
moving at velocities near that of light, we cannot. Relativity asserts that the energy of a particle is given by Einstein’s equation,
2 2 2 2 4
E =p c +m c .
0
In this equation, E is the particle energy, p is its momentum, m is its rest mass, and c is the speed of light. In transformations of
0
fundamental particles in which the sum of the rest masses of the product particles is less than that of the reactant particles,
conservation of energy requires that the sum of the momenta of the product particles exceed that of the reactant particles. The
momentum increase means that the product particles have high velocities, corresponding to a high temperature for the product
system. The most famous expression of this result is that E = m c , meaning that we can associate this quantity of energy with the
0
2
mass, m , of a stationary particle, for which p = 0 .

0
The situation with respect to the second law is similar. From experience with devices that convert heat into work, the idea evolved
that such devices must have particular properties. Consideration of these properties led to the discovery of a new state function,
which we call entropy, and to which we customarily assign the symbol “S ”. We introduce the laws of thermodynamics in §6-13.
We repeat our statement of the second law here:
The Second Law of Thermodynamics

In a reversible process in which a closed system accepts an increment of heat, dq , from its surroundings, the change in the
rev
entropy of the system, dS, is dS = dq /T . Entropy is a state function. For any reversible process, dS
rev
= 0 , and
universe
conversely. For any spontaneous process, dS > 0 , and conversely.

universe
If a spontaneous process takes a system from state A to state B, state B may or may not be an equilibrium state. State A cannot be
an equilibrium state. Since we cannot use the defining equation to find the entropy change for a spontaneous process, we must use
some other method if we are to estimate the value of the entropy change. This means that we must have either an empirical
mathematical model from which we can estimate the entropy of a non-equilibrium state or an equilibrium system that is a good
model for the initial state of the spontaneous process.
We can usually find an equilibrium system that is a good model for the initial state of a spontaneous process. Typically, some
alteration of an equilibrium system makes the spontaneous change possible. The change-enabled state is the initial state for a
spontaneous process, but its thermodynamic state functions are essentially identical to those of the pre-alteration equilibrium state.
For example, suppose that a solution contains the reactants and products for some reaction that occurs only in the presence of a
catalyst. In this case, the solution can be effectively at equilibrium even when the composition does not correspond to an
equilibrium position of the reaction. (In an effort to be more precise, we can term this a quasi-equilibrium state, by which we mean
that the system is unchanging even though a spontaneous change is possible.) If we introduce a very small quantity of catalyst, and
consider the state of the system before any reaction occurs, all of the state functions that characterize the system must be essentially
unchanged. Nevertheless, as soon as the catalyst is introduced, the system can no longer be considered to be in an equilibrium state.
The spontaneous reaction proceeds until it reaches equilibrium. We can find the entropy change for the spontaneous process by
finding the entropy change for a reversible process that takes the initial, pre-catalyst, quasi-equilibrium state to the final, post-
catalyst, equilibrium state.
Our statement of the second law establishes the properties of entropy by postulate. While this approach is rigorously logical, it does
not help us understand the ideas involved. Like the first law, the second law can be stated several ways. To develop our
understanding of entropy and its properties, it is useful to again consider a more traditional statement of the second law:
A Traditional statement of the second law

It is impossible to construct a machine that operates in a cycle, exchanges heat with its surroundings at only one temperature,
and produces work in the surroundings.
When we introduce the qualification that the machine “exchanges heat with its surroundings at only one temperature,” we mean
that the temperature of the surroundings has a particular value whenever the machine and surroundings exchange heat. The
statement does not place any conditions on the temperature of the machine at any time.
In this chapter, we have frequent occasion to refer to each of these statements. To avoid confusing them, we will refer to our
statement of the second law as the entropy-based statement. We will refer to the statement above as the machine-based statement
of the second law.
By “a machine”, we mean a heat engine—a device that accepts heat and produces mechanical work. This statement asserts that a
“perpetual motion machine of the second kind” cannot exist. Such a machine accepts heat energy and converts all of it into work,
while itself returning to the same state at the end of each cycle. (In §7-11, we note that a “perpetual motion machine of the first
kind” is one whose operation violates the principle of conservation of energy.) Normally, we view this statement as a postulate. We
consider that we infer it from experience. Unlike our statements about entropy, which are entirely abstract, this statement makes an
assertion about real machines of the sort that we encounter in daily life. We can understand the assertion that it makes in concrete
terms: A machine that could convert heat from a constant-temperature source into work could extract heat from ice water,
producing ice cubes in the water and an equivalent amount of work elsewhere in the surroundings. This machine would not
exchange heat with any other heat reservoir. Our machine-based statement of the second law postulates that no such machine can
exist.
Our entropy-based statement of the second law arose from thinking about the properties of machines that do convert heat into
work. We trace this thinking to see how our entropy-based statement of the second law was developed. Understanding this
development gives us a better appreciation for the meaning of entropy. We find that we must supplement the machine-based
statement of the second law with additional assumptions in order to arrive at all of the properties of the entropy function that are
asserted in the entropy-based statement.
However, before we undertake to develop the entropy-based statement of the second law from the machine-based statement, let us
develop the converse; that is, let us show that the machine-based statement is a logical consequence of the entropy-based statement.
To do so, we assume that a perpetual motion machine of the second kind is possible. To help keep our argument clear, let
proposition MSL be the machine-based statement. We are assuming that proposition MSL is false, so that proposition ∼ MSL is
true. We let SL be the entropy-based statement of the second law.
The sketch in Figure 1 describes the interaction of this perpetual motion machine, PPM, with its surroundings. From our entropy-
based statement of the second law, we can assert some important facts about the entropy changes that accompany operation of the
machine. Since entropy is a state function, ΔS = 0 for one cycle of the machine. If the machine works (that is, ∼ MSL is true),
then the entropy-based statement requires that ΔS universe= ΔS + ΔS ≥ 0 . Since ΔS = 0 , it follows that ΔS ≥ 0 . We can
^ ^
make this more explicit by writing: (SL and ∼ MSL) ⇒ ΔS^ ≥ 0 .

The machine-based statement of the second law also enables us to determine the entropy change in the surroundings from our
second-law definition of entropy. In one cycle, this machine (system) delivers net work, w^ > 0 , to the surroundings; it accepts a net
quantity of heat, q > 0 , from the surroundings, which are at temperature, T . Simultaneously, the surroundings surrender a quantity
^
of heat, q^, where q^ = −q , and q^ < 0 . The change that occurs in one cycle of the machine need not be reversible. However,
whether the change is reversible or not, the entire thermal change in the surroundings consists in the exchange of an amount of
heat, q^ < 0 , by a constant temperature reservoir at T^. We can effect identically the same change in the surroundings using some
other process to reversibly extract this amount of heat. The entropy change in the surroundings in this reversible process will be
^/ T
q
^
, and this will be the same as the entropy change for the surroundings in one cycle of the machine. (We consider this
conclusion further in §15.) It follows that ΔS^ = q^/T^ , and since q^ < 0 , while T^ > 0 , we have ΔS^ < 0 . We can write this
conclusion more explicitly: (SL and ∼ MSL) ⇒ ΔS^ < 0 .
Figure 1. A perpetual motion machine (PPM) that violates the second law.
By assuming a perpetual motion machine of the second kind is possible—that is, by assuming ∼ MSL is true—we derive the
contradiction that both ΔS^ ≥ 0 and ΔS^ < 0 . Therefore, proposition ∼ MSL must be false. Proposition MSL must be true. The
entropy-based second law of thermodynamics implies that a perpetual motion machine of the second kind is not possible. That is,
the entropy-based statement of the second law implies the machine-based statement. (We prove that ∼ (SL and ∼ MSL) ; it
follows that SL ⇒MSL . For a more detailed argument, see problem 2.)
This page titled 9.1: The Second Law of Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
Historically, the steam engine was the first machine for converting heat into work that could be exploited on a large scale. The
steam engine played a major role in the industrial revolution and thus in the development of today’s technology-intensive economy.
It was important also in the development of the basic concepts of thermodynamics. A steam engine produces work when hot steam
under pressure is introduced into a cylinder, driving a piston outward. A shaft connects the piston to a flywheel. When the
connecting shaft reaches its greatest extension, the spent steam is vented to the atmosphere. Thereafter the flywheel drives the
piston inward.
The economic viability of the steam engine derives, in part, from the fact that the spent steam can be vented to the atmosphere at
the end of each cycle. However, this is not a necessary feature of heat engines. We can devise engines that alternately heat and cool
a captive working fluid to convert heat energy into mechanical work. Stirling engines are practical devices of this type. A Carnot
engine is a conceptual engine that exploits the response of a closed system to temperature changes. A Carnot engine extracts heat
from one reservoir at a fixed high temperature and discharges a lesser amount of heat into a second reservoir at a fixed lower
temperature. An amount of energy equal to the difference between these increments of heat energy appears in the surroundings as
work.
For one cycle of the Carnot engine, let the heat transferred to the system from the hot and cold reservoirs be q and q respectively.
h ℓ
We have q > 0 and q < 0 . Let the net work done on the system be w
h ℓ and the net work that appears in the surroundings be
net
w
^net . We have
^ net > 0
w ,w
^ net = −wnet , and w net <0 . For one cycle of the engine, ΔE = 0 , and since
^ net ,
ΔE = qh + qℓ + wnet = qh + qℓ − w
it follows that w
^ = q + q . The energy input to the Carnot engine is q , and the useful work that appears in the surroundings is
net h ℓ h
w^
net . (The heat accepted by the low-temperature reservoir, q^ = −q > 0 , is a waste product, in the sense that it represents energy
ℓ ℓ
that cannot be converted to mechanical work using this cycle. All feasible heat engines share this feature of the Carnot engine. In
contrast, a perpetual motion machine of the second kind converts its entire heat intake to work; no portion of its heat intake goes
unused.) The efficiency, ϵ, with which the Carnot engine converts the input energy, q , to useful output energy, w
h
^ , is therefore,
net
w
^ net qh + qℓ qℓ
ϵ= = =1+
qh qh qh
We can generalize our consideration of heat engines to include any series of changes in which a closed system exchanges heat with
its surroundings at more than one temperature, delivers a positive quantity of work to the surroundings, and returns to its original
state. We use the Carnot cycle and the machine-based statement of the second law to analyze systems that deliver pressure–volume
work to the surroundings. We consider both reversible and irreversible systems. We begin by considering reversible Carnot cycles.
If any system reversibly traverses any closed path on a pressure–volume diagram, the area enclosed by the path represents the
pressure–volume work exchanged between the system and its surroundings. If the area is not zero, the system temperature changes
during the cycle. If the cycle is reversible, all of the heat transfers that occur must occur reversibly. We can apply our reasoning
about reversible cycles to any closed system containing any collection of chemical substances, so long as any phase changes
or chemical reactions that occur do so reversibly. This means that all phase and chemical changes that occur in the system must
adjust rapidly to the new equilibrium positions that are imposed on them as a system traverses a Carnot cycle reversibly.
Figure 2. An ideal gas Carnot cycle. Note that the pressure axis is compressed: In P is plotted vs. V.
In Figure 2, we describe the operation of a reversible Carnot engine in which the working fluid is an ideal gas. We designate the
system’s initial pressure, volume, and temperature by P , V , and T . From this initial state, we cause the ideal gas to undergo a
1 1 h
reversible isothermal expansion in which it absorbs a quantity of heat, q , from a high-temperature heat reservoir at T^ . We
h h
designate the pressure, volume, and temperature at the end of this isothermal expansion as P , V , and T . In a second step, we
2 2 h
reversibly and adiabatically expand the ideal gas until its temperature falls to that of the second, low-temperature, heat reservoir.
We designate the pressure, volume, and temperature at the end of this adiabatic expansion as P , V , and T . We begin the return
3 3 ℓ
portion of the cycle by reversibly and isothermally compressing the ideal gas at the temperature of the cold reservoir. We continue
this reversible isothermal compression until the ideal gas reaches the pressure and volume from which an adiabatic compression
will just return it to the initial state. We designate the pressure, volume, and temperature at the end of this isothermal compression
by P , V , and T . During this step, the ideal gas gives up a quantity of heat, q < 0 , to the low-temperature reservoir. Finally, we
4 4 ℓ ℓ
reversibly and adiabatically compress the ideal gas to its original pressure, volume, and temperature.
For the high-temperature isothermal step, we have
V2
−qh = wh = −RTh ln( )
V1
and for the low-temperature isothermal step, we have

V4
−qℓ = wℓ = −RTℓ ln( )
V3
For the adiabatic expansion and compression, we have
qexp = qcomp = 0
The corresponding energy and work terms are

Tℓ
Δexp E = wexp = ∫ CV dT
Th
for the adiabatic expansion and

Th
Δcomp E = wcomp = ∫ CV dT
Tℓ
for the adiabatic compression. The heat-capacity integrals are the same except for the direction of integration; they sum to zero, and
we have w + w exp =0 .
comp The net work done on the system is the sum of the work for these four steps,
wnet = w +w
h exp+w +w ℓ = w +w
comp h . The heat input occurs at the high-temperature reservoir, so that q > 0 . The heat
ℓ h
discharge occurs at the low-temperature reservoir, so that q < 0 . ℓ
For one cycle of the reversible, ideal-gas Carnot engine,
qℓ RTℓ ln(V4 / V3 )
ϵ=1+ =1+
qh V2
RTh ln( )
V1
Because the two adiabatic steps involve the same limiting temperatures, the energy of an ideal gas depends only on temperature,
and dE = dw for both steps, we see from Section 9.7-9.20 that
Tℓ V3
CV R V3
∫ dT = − ∫ dV = −Rln( )
Th
T V2
V V2
and
Th V1
CV R V1
∫ dT = − ∫ dV = −Rln( )
Tℓ
T V4
V V4
The integrals over T are the same except for the direction of integration. They sum to zero, so that
−Rln(V3 / V2 ) − Rln(V1 / V4 ) = 0 and
V2 V3
=
V1 V4
Using this result, the second equation for the reversible Carnot engine efficiency becomes
Tℓ
ϵ=1−
Th
Equating our expressions for the efficiency of the reversible Carnot engine, we find
qℓ Tℓ
ϵ=1+ =1−
qh Th
from which we have

qh qℓ
+ =0
Th Tℓ
Since there is no heat transfer in the adiabatic steps, q exp = qcomp = 0, and we can write this sum as
qi
∑ =0
Ti
cycle
If we divide the path around the cycle into a large number of very short segments, the limit of this sum as the q become very small i
is
rev
dq
∮ =0
T
where the superscript “rev ” serves as a reminder that the cycle must be traversed reversibly. Now, we can define a new function, S ,
by the differential expression
rev
dq
dS =
T
In this expression, dS is the incremental change in S that occurs when the system reversibly absorbs a small of increment of heat,
dq , at a particular temperature, T . For an ideal gas traversing a Carnot cycle, we have shown that
rev
rev
dq
ΔS = ∮ dS = ∮ =0
T
S is, of course, the entropy function described in our entropy-based statement of the second law.
We now want to see what the machine-based statement of the second law enables us to deduce about the properties of S . Since the
change in S is zero when an ideal gas goes around a complete Carnot cycle, we can conjecture that S is a state function. Of course,
the fact that ΔS = 0 around one particular cycle does not prove that S is a state function. If S is a state function, it must be true
that ΔS = 0 around any cycle whatsoever. We now prove this for any reversible cycle.
The proof has two steps. In the first, we show that ∮ dq /T = 0 for a machine that uses any reversible system operating between
rev
two constant-temperature heat reservoirs to convert heat to work. In the second step, we show that ∮ dq /T = 0 for any system
rev
that reversibly traverses any closed path.
This page titled 9.2: The Carnot Cycle for an Ideal Gas and the Entropy Concept is shared under a CC BY-SA 4.0 license and was authored,
To show that ∮ dq /T = 0 for any reversible system taken around a Carnot cycle, we first observe that the Carnot cycle can be
rev
traversed in the opposite direction. In this case, work is delivered to the engine and a quantity of heat is transferred from the low-
temperature reservoir to the high-temperature reservoir. Operated in reverse, the Carnot engine is a refrigerator. Suppose that we
have two identical ideal-gas Carnot machines, one of which we operate as an engine while we operate the other as a refrigerator. If
we configure them so that the work output of the engine drives the refrigerator, the effects of operating them together cancel
completely. The refrigerator exactly consumes the work output of the engine. The heat transfers to and from the heat reservoirs
offset exactly.
Now, let us consider an ideal-gas Carnot engine and any other reversible engine that extracts heat from a high-temperature reservoir
and rejects a portion of it to a low-temperature reservoir. Let us call these engines A and B. We suppose that one is operated to
produce work in its surroundings (w < 0 ); the other is operated to consume this work and transfer net heat energy from the low-
temperature to the high-temperature reservoir. Let the net work done in one cycle on machines A and B be w and w ,
netA netB
respectively. We can choose to make these engines any size that we please. Let us size them so that one complete cycle of either
engine exchanges the same quantity of heat with the high-temperature reservoir. That is, if the high-temperature reservoir delivers
heat q to engine A, then it delivers heat q = q
hA hB hA to engine B. Figure 3 diagrams these engines. With one operating as an
engine and the other operating as a refrigerator, we have q + q = 0 . When both engine and refrigerator have completed a
hA hB
cycle, the high temperature reservoir has returned to its original state.
Figure 3. Matched heat engine and refrigerator and a system that combines them.
We can create a combined device that consists of A running as an engine, B running as a refrigerator, and the high-temperature
reservoir. Figure 3 also diagrams this combination. When it executes one complete cycle, the initial condition of the combined
device is restored. Therefore, since E is a state function, we have
ΔE = wnetA + qhA + qℓA + wnetB + qhB + qℓB (9.3.1)
= wnetA + wnetB + qℓA + qℓB (9.3.2)
= 0. (9.3.3)
where we use the constraint q + q = 0 . Let us consider the possibility that

hA hB wnetA + wnetB < 0 ; that is, the combined device
does net work on the surroundings. Then, ΔE = 0 implies that q + q > 0 . ℓA ℓB
In this cyclic process, the combined device takes up a positive quantity of heat from a constant-temperature reservoir and delivers a
positive quantity of work to the surroundings. There is no other change in either the system or the surroundings. This violates the
machine-based statement of the second law. Evidently, it is not possible for the combined device to operate in the manner we have
hypothesized. We conclude that any such machine must always operate such that w +w ≥ 0 ; that is, the net work done on
netA netB
the combined machine during any complete cycle must be either zero or some positive quantity.
In concluding that w +w
netA ≥ 0 , we specify that the combined machine has A running as a heat engine and B running as a
netB
refrigerator. Now, suppose that we reverse their roles, and let w and w
∗
netA
represent the net work for the reversed combination.
∗
netB
Applying the same argument as previously, we conclude that w +w

∗
netA
≥ 0 . But, since the direction of operation is reversed
∗
netB
for both machines, we must also have w ∗
netA
= −wnetA and w
∗
netB
= −wnetB . Hence we have −wnetA − wnetB ≥ 0 or
w netA
+w ≤ 0 . We conclude, therefore, that
netB
wnetA + wnetB = 0
for any two, matched, reversible engines operating around a Carnot cycle.
This conclusion can be restated as a condition on the efficiencies of the two machines. The individual efficiencies are
ϵA = −w /q
netA and ϵ = −w
hA B /q .
netB hB
(The efficiency equation is unaffected by the direction of operation, because changing the direction changes the sign of every
energy term in the cycle. Changing the direction of operation is equivalent to multiplying both the numerator and denominator by
minus one.) Then, from w +w
netA = 0 , it follows that
netB
ϵA qhA + ϵB qhB = 0
Since we sized A and B so that q hA + qhB = 0 , we have
ϵA qhA − ϵB qhA = 0
so that
ϵA = ϵB
for any reversible Carnot engines A and B operating between the same two heat reservoirs.
For the ideal gas engine, we found ϵ = 1 − T ℓ / Th . For any reversible Carnot engine, we have ΔE = 0 = w net + qh + qℓ , so that
−w = q + q , and
net h ℓ
−wnet qℓ
ϵ= =1+
qh qh
This means that the efficiency relationship

Tℓ qℓ
ϵ=1− =1+
Th qh
applies to any reversible Carnot engine. It follows that the integral of dq

rev
/T around a Carnot cycle is zero for any reversible
system.
The validity of these conclusions is independent of type of work that the engine produces; if engine A is an ideal-gas engine,
engine B can be comprised of any system and can produce any kind of work. In obtaining this result from the machine-based
statement of the second law, we make the additional assumption that pressure–volume work can be converted entirely to any other
form of work, and vice versa. That is, we assume that the work produced by engine A can reversibly drive engine B as a
refrigerator, whether engines A and B produce the same or different kinds of work.
This page titled 9.3: The Carnot Cycle for Any Reversible System is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Any system reversibly traversing any closed curve on a pressure–volume diagram exchanges work with its surroundings, and the
area enclosed by the curve represents the amount of this work. In the previous section, we found ∮ dq /T = 0 for any system rev
that traverses a Carnot cycle reversibly. We now show that this is true for any system that traverses any closed path reversibly. This
establishes that ΔS is zero for any system traversing any closed path reversibly and proves that S , defined by dS = dq /T , is a rev
state function.
To do so, we introduce an experience-based theorem: The pressure–volume diagram for any reversible system can be tiled by
intersecting lines that represent isothermal and adiabatic paths. These lines can be packed as densely as we please, so that the tiling
of the pressure–volume diagram can be made as closely spaced as we please. The perimeter of any one of the resulting tiles
corresponds to a path around a Carnot cycle. Given any arbitrary closed curve on the pressure–volume diagram, we can select a set
of tiles that just encloses it. See Figure 4. The perimeter of this set of tiles approximates the path of the arbitrary curve. Since the
tiling can be made as fine as we please, the perimeter of the set of tiles can be made to approximate the path of the arbitrary curve
as closely as we please.
Figure 4. Tiling the PV-plane with isotherms and adiabats.

Suppose that we traverse the perimeter of each of the individual tiles in a clockwise direction, adding up q /T as we go. rev
Segments of these perimeters fall into two groups. One group consists of segments that are on the perimeter of the enclosing set of
tiles. The other group consists of segments that are common to two tiles. When we traverse both of these tiles in a clockwise
direction, the shared segment is traversed once in one direction and once in the other. When we add up q /T for these two rev
traverses of the same segment, we find that the sum is zero, because we have q /T in one direction and −q /T in the other.
rev rev
This means that the sum of q /T around all of the tiles will just be equal to the sum of q /T around those segments that lie on
rev rev
the perimeter of the enclosing set. That is, we have
⎧
⎪ ⎫
⎪
⎪
⎪ ⎪
⎪
rev rev ⎪ rev ⎪
q q q
∑ + ∑ = ∑ ⎨ ∑ ⎬
T T ⎪ T ⎪
cycle interior all ⎪
⎪ tile ⎪
⎪
⎩
⎪ ⎭
⎪
perimeter seqments tiles perimeter
where
rev
q
∑ =0
T
interior
seqments
because each interior segment is traversed twice, and the two contributions cancel exactly.
This set of tiles has another important property. Since each individual tile represents a reversible Carnot cycle, we know that
rev
q
∑ =0
T
tile
perimeter
around each individual tile. Since the sum around each tile is zero, the sum of all these sums is zero. It follows that the sum of
/T around the perimeter of the enclosing set is zero:
rev
q
rev
q
∑ =0
T
cycle
perimeter
By tiling the pressure–volume plane as densely as necessary, we can make the perimeter of the enclosing set as close as we like to
any closed curve. The heat increments become arbitrarily small, and
⎡ ⎤
rev rev
⎢ q ⎥ dq
lim ⎢ ∑ ⎥ =∮ =0
rev ⎢ ⎥
q
rev
→dq ⎢ T ⎥ T
cycle
⎣ ⎦
perimeter
For any reversible engine producing pressure–volume work, we have ∮ dS = 0 around any cycle.
We can extend this analysis to reach the same conclusion for a reversible engine that produces any form of work. To see this, let us
consider the tiling theorem more carefully. When we say that the adiabats and isotherms tile the pressure–volume plane, we mean
that each point in the pressure–volume plane is intersected by one and only one adiabat and by one and only one isotherm. When
only pressure–volume work is possible, every point in the pressure–volume plane represents a unique state of the system.
Therefore, the tiling theorem asserts that every state of the variable-pressure system can be reached along one and only one adiabat
and one and only one isotherm.
From experience, we infer that this statement remains true for any form of work. That is, every state of any reversible system can
be reached by one and only one isotherm and by one and only one adiabat when any form of work is done. If more than one form
of work is possible, there is an adiabat for each form of work. If changing θ and changing θ change the energy of the system, the
1 2
effects on the energy of the system are not necessarily the same. In general, Φ is not the same as Φ , where
1 2
∂E
Φi = ( )
∂θi V ,θm≠i
From §3, we know that a reversible Carnot engine doing any form of work can be matched with a reversible ideal-gas Carnot
engine in such a way that the engines complete the successive isothermal and adiabatic steps in parallel. At each step, each engine
experiences the same heat, work, energy, and entropy changes as the other. Just as we can plot the reversible ideal-gas Carnot cycle
as a closed path in pressure–volume space, we can plot a Carnot cycle producing any other form of work as a closed path with
successive isothermal and adiabatic steps in Φ − − θ space. Just as any closed path in pressure–volume space can be tiled (or
i i
built up from) arbitrarily small reversible Carnot cycles, so any closed path in Φ − − θ space can be tiled by such cycles.
i i
Therefore, the argument we use to show that ∮ dS = 0 for any closed reversible cycle in pressure–volume space applies equally
well to a closed reversible cycle in which heat is used to produce any other form of work.
This page titled 9.4: The Entropy Change around Any Cycle for Any Reversible System is shared under a CC BY-SA 4.0 license and was
We view the tiling theorem as a generalization from experience, just as the machine-based statement of the second law is such a
generalization. Let us consider the kinds of familiar observations from which we infer that every equilibrium state of any system is
intersected by one and only one adiabat and by one and only one isotherm.
When only pressure–volume work is possible, each pressure–volume point specifies a unique equilibrium state of the system. Since
temperature is a state function, the temperature of this state has one and only one value. When another form of work is possible,
every Φ −θ point specifies a unique state for which the temperature has one and only one value. From experience, we know that
i i
we can produce a new state of the system, at the same temperature, by exchanging heat and work with it in a concerted fashion. We
can make this change of state arbitrarily small, so that successive equilibrium states with the same temperature are arbitrarily close
to one another. This succession of arbitrarily close equilibrium states is an isotherm. Therefore, at least one isotherm intersects any
equilibrium state. There cannot be two such isotherms. If there were two isotherms, the system would have two temperatures,
violating the principle that temperature is a state function.
In an adiabatic process, the system exchanges energy as work but not as heat. From experience, we know that we can effect such a
change with any reversible system. The result is a new equilibrium state. When we make the increment of work arbitrarily small,
the new equilibrium state is arbitrarily close to the original state. Successive exchanges of arbitrarily small work increments
produce successive equilibrium states that are arbitrarily close to one another. This succession of arbitrarily close equilibrium states
is an adiabat.
If the same state of a system could be reached by two reversible adiabats involving the same form of work, the effect of doing a
given amount of this work on an equilibrium system would not be unique. From the same initial state, two reversible adiabatic
experiments could do the same amount of the same kind of work and reach different final states of the system. For example, in two
different experiments, we could raise a weight reversibly from the same initial elevation, do the same amount of work in each
experiment, and find that the final elevation of the weight is different. Any such outcome conflicts with the observations that
underlie our ideas about reversible processes.
More specifically, the existence of two adiabats through a given point, in any Φ −θ space, violates the machine-based statement of
i i
the second law. Two such adiabats would necessarily intersect a common isotherm. A path along one adiabat, the isotherm, and the
second adiabat would be a cycle that restored the system to its original state. This path would enclose a finite area. Traversed in the
appropriate direction, the cycle would produce work in the surroundings. By the first law, the system would then accept heat as it
traverses the isotherm. The system would exchange heat with surroundings at a single temperature and produce positive work in the
surroundings, thus violating the machine-based statement.
If an adiabatic process that connects two states A and B is reversible, we see that the system follows the same path, in opposite
directions, when it does work going from A to B as it does when work is done on it as it goes from B to A.
From another perspective, we can say that the tiling theorem is a consequence of our assumptions about reversible processes. Our
conception of a reversible process is that the energy, pressure, temperature, and volume are continuous functions of state, with
continuous derivatives. That there is one and only one isotherm for every state is equivalent to the assumption that temperature is a
continuous (single-valued) function of the state of the system. That there is one and only one adiabat for every state is equivalent to
the assumption that (∂E/∂V ) , or generally, (∂E/∂ θ )
T ,θ1 i , is a continuous, single-valued function of the state of the
T ,V ,θm≠i
system.
With these ideas in mind, let us now observe that any reversible cycle can be described by a closed path in a space whose
coordinates are T and q /T (entropy). In Figure 5, we sketch this space with q /T on the abscissa; then an isotherm is a
rev rev
horizontal line, and line of constant entropy (an isentrope) is vertical. A reversible Carnot cycle is a closed rectangle, and the area
of this rectangle corresponds to the reversible work done by the system on its surroundings in one cycle. Any equilibrium state of
the system corresponds to a particular point in this space. Any closed path can be tiled arbitrarily densely by isotherms and
isentropes. Any reversible cycle involving any form of work is represented by a closed path in this space. Figure 5 is an alternative
illustration of the argument that we make in Section 9.4. The path in this space is independent of the kind of work done, reinforcing
the conclusion that ∮ dq /T = 0 for a reversible Carnot cycle producing any form of work. The fact that a cyclic process
rev
corresponds to a closed path in this space is equivalent to the fact that entropy is a state function.
Figure 5. A reversible cycle described using coordinates T and q rev
/T .
To appreciate this aspect of the path of a cyclic process in T − q /T space, let us describe the path of the same process in a space
rev
whose coordinates are T and q . With q on the abscissa, isotherms are again horizontal lines and adiabats are vertical lines. In
rev rev
this space, a reversible Carnot cycle does not begin and end at the same point. The path is not closed. Similarly, the representation
of an arbitrary reversible cycle is not a closed figure. See Figure 6. The difference between the representations of a reversible cyclic
process in these two spaces illustrates graphically the fact that entropy is a state function while heat is not.
Figure 6. A reversible cycle described using T and q rev
This page titled 9.5: The Tiling Theorem and the Paths of Cyclic Process in Other Spaces is shared under a CC BY-SA 4.0 license and was
Let us consider a closed system that undergoes a reversible change while in contact with its surroundings. Since the change is
reversible, the portion of the surroundings that exchanges heat with the system is at the same temperature as the system: T = T^ .
From q = −q^
rev
and the definition, dS = dq /T , the entropy changes are
rev rev
rev
ΔS = q /T
and
^ rev rev
ΔS = q
^ /T = −q /T = −ΔS
Evidently, for any reversible process, we have

^
ΔSuniverse = ΔS + ΔS = 0
Note that these ideas are not sufficient to prove that the converse is true. From only these ideas, we cannot prove that
ΔS = 0 for a process means that the process is reversible; it remains possible that there could be a spontaneous process for
universe
which ΔS = 0 . However, our entropy-based statement of the second law does assert that the converse is true, that
universe
ΔS = 0 is necessary and sufficient for a process to be reversible.

universe
In the next section, we use the machine-based statement of the second law to show that ΔS ≥ 0 for any spontaneous process in an
isolated system. We introduce heuristic arguments to infer that ΔS = 0 is not possible for a spontaneous process in an isolated
system. From this, we show that ΔS > 0 for any spontaneous process and hence that ΔS
universe = 0 is not possible for any
universe
spontaneous process. We conclude that ΔS = 0 is sufficient to establish that the corresponding process is reversible.
universe
This page titled 9.6: Entropy Changes for A Reversible Process is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
In Section 9.6, we consider the entropy changes in a system and its surroundings when the process is reversible. We consider now
the diametrically opposite situation in which an isolated system undergoes a spontaneous change. From the entropy-based
statement of the second law, we know how the entropy of this system and its surroundings change. Since the system is isolated, no
change occurs in the surroundings. Thus, ΔS^ = 0 ; and since ΔS + ΔS^ > 0 , we have ΔS > 0 .
Let us attempt to develop these conclusions from the machine-based statement of the second law. Since the process occurs
irreversibly, we cannot use the heat of the process to find the entropy change for the system. We can calculate the entropy change
for a process from the defining equation only if the process is reversible. However, entropy is a state function; using the figure of
speech that we introduce in Section 7.21, we can find the entropy change for the spontaneous process by evaluating ΔS along a
second and reversible path that connects the same initial and final states.
Figure 7. A cycle that includes a spontaneous process in an isolated system.

In Figure 7, these paths are diagrammed in temperature–entropy space. The transition from state A to state B occurs irreversibly,
and therefore it does not necessarily correspond to a path that we can specify on this diagram. The dashed line drawn for this
transition is supposed to remind us of this fact. We can readily devise a reversible path from B back to A. First, we reversibly and
adiabatically return the temperature of the system to its original value T^. In this step, the system does work on the surroundings, or
vice versa. The system reaches point C on the diagram. Then we reversibly and isothermally add or remove heat from the system to
return to the original state at point A. For the transfer of heat to be reversible, we must have T = T^ for this step. Hence, the final
(and original) temperature of the system at point A is equal to the temperature of the surroundings. The reversible path
B → C → A must exist, because the tiling theorem asserts that adiabats (vertical lines) and isotherms (horizontal lines) tile the
T − −S -plane arbitrarily densely.
Taken literally, this description of state A is inconsistent. We suppose that the initial state A is capable of spontaneous change;
therefore, it cannot be an equilibrium state. We suppose that the final state A is reached by a reversible process; therefore, it must
be an equilibrium state. We bridge this contradiction by refining our definition of the initial state. The final state A is an
equilibrium state with well-defined state functions. What we have in mind is that these final equilibrium-state values also
characterize the initial non-equilibrium state. Evidently, the initial state A that we have in mind is a hypothetical state. This
hypothetical state approximates the state of a real system that undergoes spontaneous change. By invoking this hypothetical initial
state, we eliminate the contradiction between our descriptions of initial state A and final state A. Given a real system that
undergoes spontaneous change, we must find approximate values for the real system’s state functions by finding an equilibrium—
or quasi-equilibrium—system that adequately models the initial state of the spontaneously changing system.
In the development below, we place no constraints on the nature of the system or the spontaneous process. We assume that the state
functions of any hypothetical initial state A can be adequately approximated by some equilibrium-state model. However, before we
consider the general argument, let us show how these conditions can be met for another specific system. Consider a vessel whose
interior is divided by a partition. The real gas of a pure substance occupies the space on one side of the partition. The space on the
other side of the partition is evacuated. We suppose that this vessel is isolated. The real gas is at equilibrium. We can measure its
state functions, including its pressure, volume, and temperature. Now suppose that we puncture the partition. As soon as we do so,
the gas expands spontaneously to fill the entire vessel, reaching a new equilibrium position, at a new pressure, volume, and
temperature. The gas undergoes a free expansion, as defined in Section 7.17.
At the instant the partition is punctured, the system becomes able to undergo spontaneous change. In this hypothetical initial state,
before any significant quantity of gas passes through the opening, neither the actual condition of the gas nor the values of its state
functions have changed. After the expansion to the new equilibrium state, the original state can be restored by reversible processes
of adiabatic compression and isothermal volume adjustment. (Problems 13 and 14 in Chapter 10 deal with the energy and entropy
changes for ideal and real gases around a cycle in which spontaneous expansion in an isolated system is followed by reversible
restoration of the initial state.)
Returning to the general cycle depicted in Figure 7, we see that there are some important conditions on the heat and work terms in
the individual steps. Since the system is isolated while it undergoes the transition from A to B, it exchanges no heat or work with
spon spon
the surroundings in this step: q AB
=w = 0 . For the reversible adiabatic transition from B to C, d
AB
q = 0 in every BC
rev
incremental part of the path. The transition from C to A occurs reversibly and isothermally; letting the heat of this step be q , the rev
CA
entropy changes for these reversible steps are, from the defining equation,
TC rev
dBC q
ΔBC S = ∫ =0
TB
T
and
rev
q
CA
ΔC A S =
^
T
The energy and entropy changes around this cycle must be zero, whether the individual steps occur reversibly or irreversibly. We
have
spon spon rev rev rev rev rev rev rev
ΔE = q +w +q +w +q +w =q +w +w =0
AB AB BC BC CA CA CA BC CA
and
rev ^
ΔS = ΔAB S + ΔBC S + ΔC A S = ΔAB S + q /T = 0
CA
rev rev
We want to analyze this cycle using the machine-based statement of the second law. We have w = −w ^ ,w rev
BC BC
rev
CA
^
= −w CA
, and
q
rev
CA
= −q^
CA
. Let us assume that the system does net work on the surroundings as this cycle is traversed so that w
rev
^
rev
BC
^
+w
rev
CA
>0 .
Then,
rev rev rev rev
^
− (w ^
+w ) =w +w <0
BC CA BC CA
and it follows that q > 0 . The system exchanges heat with the surroundings in only one step of this process. In this step, the
rev
CA
system extracts a quantity of heat from a reservoir in the surroundings. The temperature of this reservoir remains constant at T^
throughout the process. The heat extracted by the system is converted entirely into work. This result contradicts the machine-based
statement of the second law. Hence, w + w < 0 is false; it follows that
rev
BC
rev
CA
rev rev
w +w ≥0
BC CA
and that
rev
q ≤0
CA
For the entropy change in the spontaneous process in the isolated system, we have
rev ^
ΔAB S = −q /T ≥ 0
CA
Now, we introduce the premise that q

CA
rev
≠0 . If this is true, the entropy change in the spontaneous process in the isolated system
becomes
ΔAB S > 0
(The converse is also true; that is, Δ S > 0 implies that q ≠ 0 .) The premise that q ≠ 0 is independent of the machine-
AB
rev
CA
rev
CA
based statement of the second law, which requires only that q ≤ 0 , as we just demonstrated. It is also independent of the first
rev
CA
law, which requires only that q = −w − w

rev
CA
rev
BC
. If q ≠ 0 , we can conclude that, for a spontaneous process in an isolated
rev
CA
rev
CA
system, we must have w + w > 0 and q < 0 . These conditions correspond to doing work on the system and finding that
BC CA
rev
CA
heat is liberated by the system. There is no objection to this; it is possible to convert mechanical energy into heat quantitatively.
The conclusions that q < 0 and Δ S > 0 have important consequences; we consider them below. First, however, we consider
rev
CA AB
a line of thought that leads us to infer that q ≠ 0 and hence that q < 0 must be true.
rev
CA
rev
CA
Because q AB = 0 and w = 0 , we have E

AB = E . The system can be taken from state A to state B by the reversible process A
A B
→ C → B . Above we see that if q = 0 , we have S = S . In §6-10, we introduce Duhem’s theorem, which asserts that two
rev
CA
A B
thermodynamic variables are sufficient to specify the state of a closed reversible system in which only pressure–volume work is
possible. We gave a proof of Duhem’s theorem when the two variables are chosen from among the pressure, temperature, and
composition variables that describe the system. We avoided specifying whether other pairs of variables can be used. If we assume
now that specifying the variables energy and entropy is always sufficient to specify the state of such a system, it follows that states
A and B must in fact be the same state. (In §14, and in greater detail in Chapter 10, we see that the first law and our entropy-based
statement of the second law do indeed imply that specifying the energy and entropy specifies the state of a closed reversible system
in which only pressure–volume work is possible.)
If state A and state B are the same state; that is, if the state functions of state A are the same as those of state B, it is meaningless to
say that there is a spontaneous process that converts state A to state B. Therefore, if A can be converted to B in a spontaneous
process in an isolated system, it must be that q ≠ 0 . That is,
rev
CA
rev
[(q = 0) ⇒∼ (A can go to B spontaneously)]
CA
rev
⇒ [(A can go to B spontaneously) ⇒ (q ≠ 0)]
CA
From the machine-based statement of the second law, we find Δ S = −q /T^ ≥ 0 . When we supplement this conclusion with
AB
rev
CA
our Duhem’s theorem-based inference that q ≠ 0 , we can conclude that ΔS > 0 for any spontaneous process in any isolated
rev
CA
system. Because the system is isolated, we have q^ = 0 , and ΔS^ = 0 . For any spontaneous process in any isolated system we
have
^
ΔSuniverse = ΔS + ΔS > 0.
We can also conclude that the converse is true; that is, if Δ S = S − S > 0 for a process in which an isolated system goes
AB B A
from state A to state B, the process must be spontaneous. Since any process that occurs in an isolated system must be a spontaneous
process, it is only necessary to show that Δ S > 0 implies that state B is different from state A. This is trivial. Because entropy
AB
is a state function, S − S > 0 requires that state B be different from state A.

B A
None of our arguments depends on the magnitude of the change that occurs. Evidently, the same inequality must describe every
incremental portion of any spontaneous process; otherwise, we could define an incremental spontaneous change for which the
machine-based statement of the second law would be violated. For every incremental part of any spontaneous change in any
isolated system we have dS > 0 and
^
dSuniverse = dS + dS > 0.
These are pivotally important results; we explore their ramifications below. Before doing so, however, let us again consider a
system in which only pressure–volume work is possible. There is an alternative way to express the idea that such a system is
isolated. Since an isolated system cannot interact with its surroundings in any way, it cannot exchange energy with its surroundings.
Its energy must be constant. Since it cannot exchange pressure–volume work, its volume must be constant. Hence, isolation implies
constant E and V . If only pressure–volume work is possible, the converse must be true; that is, if only pressure–volume work is
possible, constant energy and volume imply that there are no interactions between the system and its surroundings. Therefore,
constant E and V imply that the system is isolated, and it must be true that ΔS^ = 0 . In this case, a spontaneous process in which
E and V are constant must be accompanied by an increase in the entropy of the system. (If V is constant and only pressure–
volume work is possible, the process involves no work.) We have a criterion for spontaneous change:
(ΔS) >0
EV
(spontaneous process, only pressure–volume work)
where the subscripts indicate that the energy and volume of the system are constant. (In Section 9.21, we arrive at this conclusion
by a different argument.)
This page titled 9.7: Entropy Changes for A Spontaneous Process in An Isolated System is shared under a CC BY-SA 4.0 license and was
In Section 9.7, we conclude that the entropy change is positive for any spontaneous change in an isolated system. Since we can
consider the universe to be an isolated system, it follows that ΔS > 0 for any spontaneous process.
universe
To reach this conclusion by a more detailed argument, let us consider an arbitrary system that is in contact with its surroundings.
We can subdivide these surroundings into subsystems. As diagrammed in Figure 8, we define a surroundings subsystem
(Surroundings 1) that interacts with the system and a more remote surroundings subsystem (Surroundings 2) that does not. That is,
we assume that we can define Surroundings 2 so that it is unaffected by the process. Then we define an augmented system
consisting of the original system plus Surroundings 1. The augmented system is isolated from the remote portion of the
surroundings, so that the entropy change for the augmented system is positive by the argument in the previous section. Denoting
entropy changes for the system, Surroundings 1, Surroundings 2, and the augmented system by ΔS , ΔS^ , ΔS^ , and ΔS 1 2 , augmented
respectively, we have ΔS = ΔS + ΔS > 0 , and

augmented
^
1
^ ^ ^
ΔS = ΔS 1 + ΔS 2 > 0 . Since the remote portion of the surroundings is unaffected by the change, we have ^
ΔS 2 = 0 . For any
spontaneous change, whether the system is isolated or not, we have
^ ^ ^
ΔSuniverse = ΔS + ΔS 1 + ΔS 2 = ΔSaugmented + ΔS 2 = ΔSaugmented > 0
(any spontaneous change)
Figure 8. Expanding a system to create a new, augmented, isolated system.

This statement is an essential part of the entropy-based statement of the second law. We have now developed it from the machine-
based statement of the second law by convincing, but not entirely rigorous arguments. In Section 9.6 we find that
= ΔS + ΔS = 0 for any reversible process. Thus, for any possible process, we have
^
ΔS universe
^
ΔSuniverse = ΔS + ΔS ≥ 0
The equality applies when the process is reversible; the inequality applies when it is spontaneous.
Because entropy is a state function, ΔS and ΔS^ change sign when the direction of a process is reversed. We say that a process for
which ΔS + ΔS^ < 0 is an impossible process. Our definitions mean that these classifications—reversible, spontaneous, and
impossible—are exhaustive and mutually exclusive. We conclude that ΔS = ΔS + ΔS = 0 is necessary and sufficient for
^
universe
a process to be reversible; ΔS = ΔS + ΔS > 0 is necessary and sufficient for a process to be spontaneous. (See problem
universe
^
19.)
This page titled 9.8: The Entropy of the Universe is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
Our entropy-based statement of the second law asserts the definition and basic properties of entropy that we need in order to make
predictions about natural processes. The ultimate justification for these assertions is that the predictions they make agree with
experimental observations. We have devoted considerable attention to arguments that develop the definition and properties of
entropy from the machine-based statement of the second law. These arguments parallel those that were made historically as these
concepts were developed. Understanding these arguments greatly enhances our appreciation for the relationship between the
properties of the entropy function and the changes that can occur in various physical systems.
While these arguments demonstrate that our machine-based statement implies the entropy-based statement, we introduce additional
postulates in order to make them. These include: the premise that the pressure, temperature, volume, and energy of a reversible
system are continuous functions of one another; Duhem’s theorem; the tiling theorem; and the presumption that the conclusions we
develop for pressure–volume work are valid for any form of work. We can sum up this situation by saying that our machine-based
statement serves a valuable heuristic purpose. The entropy-based statement of the second law is a postulate that we infer by
reasoning about the consequences of the machine-based statement. When we want to apply the second law to physical systems, the
entropy-based statement and other statements that we introduce below are much more useful.
Finally, we note that our machine-based statement of the second law is not the only statement of this type. Other similar statements
have been given. The logical relationships among them are interesting, and they can be used to develop the entropy-based
statement of the second law by arguments similar to those we make in Section 9.2 to Section 9.8.
This page titled 9.9: The Significance of The Machine-based Statement of The Second Law is shared under a CC BY-SA 4.0 license and was
The content of the first law of thermodynamics is that there is a state function, which we call energy, which has the property that
ΔE = 0 for any process that can occur. The content of the second law is that there is a state function, which we call
universe
entropy, which has the property that ΔS > 0 for any spontaneous process.
universe
These two state functions exhaust the range of independent possibilities: Suppose that we aspire to find a new and independent
state function, call it B , which further characterizes the possibilities open to the universe. What other condition could B impose on
the universe—or vice versa? The only available candidate might appear to be ΔB < 0 . However, this does not represent an
universe
independent condition, since its role is already filled by the quantity −ΔS .
universe
Of course, we can imagine a state function, B , which is not simply a function of S , but for which
ΔBuniverse > 0 , ΔB = 0 , or ΔB
universe < 0 , according as the process is spontaneous, reversible, or impossible,
universe
respectively. For any given change, ΔB would not be the same as ΔS ; however, ΔB and ΔS would make exactly the same
predictions. If ΔB were more easily evaluated than ΔS
universe , we would prefer to use B rather than S . Nevertheless, if
universe
there were such a function B , its role in our description of nature would duplicate the role played by S .
This page titled 9.10: A Slightly Philosophical Digression on Energy and Entropy is shared under a CC BY-SA 4.0 license and was authored,
Let us consider another frequently cited alternative statement of the second law, which, for easy reference, we call the
temperature-based statement of the second law:
The temperature-based statement of the second law
The spontaneous transfer of heat from a colder body to a warmer body is impossible.
In the discussion below, we refer to this statement as proposition TSL. By “body”, we simply mean any system or object. By the
“spontaneous transfer of heat,” we mean that the transfer of heat energy can be initiated by bringing the two bodies into contact
with one another or by enabling the transmission of radiant energy between them. The surroundings do no work and exchange no
heat with either reservoir; there is no change of any sort in the surroundings.
We can show that the entropy-based statement and the temperature-based statement of the second law are equivalent: Given the
definition of entropy, one implies the other.
Let us begin by showing that the entropy-based statement implies the temperature-based statement of the second law. That is, we
prove SL ⇒ TSL . To do so, we prove ∼ TSL ⇒∼ SL . That is, we assume that spontaneous transfer of heat from a colder to a
warmer body is possible and show that this leads to a contradiction of the entropy-based statement of the second law. Let the
quantity of heat received by the warmer body be dq > 0 , and let the temperatures of the warmer and colder bodies be
warmer
Twarmer and T , respectively. We have T

colder −T
warmer > 0. The colder body receives heat
colder
dqcolder = −dqwarmer < 0 . We make the heat increment so small that there is no significant change in the temperature of either
body. No other changes occur. The two bodies are the only portions of the universe that are affected. Let the entropy changes for
the warmer and colder bodies be dS and dS
colder , respectively.
warmer
To find dS colderand dS we must find a reversible path to effect the same changes. This is straightforward. We can effect
warmer
identically the same change in the warmer body by transferring heat, q > 0 , to it through contact with some third body,
warmer
whose temperature is infinitesimally greater than T . This process is reversible, and the entropy change is
warmer
dS warmer = dq warmer/T .
warmer Similarly, the entropy change for the colder body is
dS colder= dq /T
colder = −dq
colder /T
warmer . It follows that
colder
dSuniverse = dSwarmer + dScolder
dqwarmer dqwarmer
= −
Twarmer Tcolder
Twarmer − Tcolder
= −dqwarmer ( )
Twarmer Tcolder
<0
However, if dS < 0 for a spontaneous process, the second law (SL) must be false. We have shown that a violation of the
universe
temperature-based statement implies a violation of the entropy-based statement of the second law: ∼ TSL ⇒∼ SL , so that
SL ⇒ TSL .
It is equally easy to show that the temperature-based statement implies the entropy-based statement of the second law. To do so, we
assume that the entropy-based statement is false and show that this implies that the temperature-based statement must be false. By
the arguments above, the entropy change that the universe experiences during the exchange of the heat increment is
Twarmer − Tcolder
dSuniverse = −dqwarmer ( )
Twarmer Tcolder
If the entropy-based statement of the second law is false, then dS < 0 . It follows that dq
universe > 0 ; that is, the spontaneous
warmer
process transfers heat from the colder to the warmer body. This contradicts the temperature-based statement. That is,
∼ SL ⇒∼ TSL , so that TSL ⇒ SL .
This page titled 9.11: A Third Statement of the Second Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
The entropy-based criteria that we develop in Section 9.2 through Section 9.8 are of central importance. If we are able to evaluate
the change in the entropy of the universe for a prospective process and find that it is greater than zero, we can conclude that the
process can occur spontaneously. The reverse of a spontaneous process cannot occur; it is an impossible process and the change in
the entropy of the universe for such a process must be less than zero. Since an equilibrium process is a reversible process, the
entropy of the universe must remain unchanged when a system goes from an initial state to a final state along a path whose every
point is an equilibrium state. Using another figure of speech, we often say that a change that occurs along a reversible path is a
change that “occurs at equilibrium.”
These conclusions are what make the entropy function useful: If we can calculate ΔS for a prospective process, we know
universe
whether the system is at equilibrium with respect to that process; whether the process is possible; or whether the process cannot
occur. If we find ΔS > 0 for a process, we can conclude that the process is possible; however, we cannot conclude that the
universe
process will occur. Indeed, many processes can occur spontaneously but do not do so. For example, hydrocarbons can react
spontaneously with oxygen; most do so only at elevated temperatures or in the presence of a catalyst.
The criteria ^
ΔSuniverse = ΔS + ΔS ≥ 0 are completely general. They apply to any process occurring under any conditions. To
apply them we must determine both ΔS and ΔS^. By definition, the system comprises the part of the universe that is of interest to
us; the need to determine ΔS^ would appear to be a nuisance. This proves not to be the case. So long as the surroundings have a
well-defined temperature, we can develop additional criteria for equilibrium and spontaneous change in which ΔS^ does not occur
explicitly. In §14, we develop criteria that apply to reversible processes. In §15, we find a general relationship for ΔS^ that enables
us to develop criteria for spontaneous processes.
To develop the criteria for spontaneous change, we must define what we mean by spontaneous change more precisely. To define a
spontaneous process in an isolated system as one that can take place on its own is reasonably unambiguous. However, when a
system is in contact with its surroundings, the properties of the surroundings affect the change that occurs in the system. To specify
a particular spontaneous process we must specify some properties of the surroundings or—more precisely—properties of the
system that the surroundings act to establish. The ideas that we develop in §15 lead to criteria for changes that occur while one or
more thermodynamic functions remain constant. These criteria supplement the second-law criteria ΔS + ΔS^ ≥ 0 . In using these
criteria, we can say that the change occurs subject to one or more constraints.
Some of these criteria depend on the magnitudes of ΔE and ΔH in the prospective process. We also find criteria that are
expressed using new state functions that we call the Helmholtz and Gibbs free energies. In the next section, we introduce these
functions.
This page titled 9.12: Entropy and Predicting Change is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
The first and second laws of thermodynamics define energy and entropy. Energy and entropy are fundamental state functions that
we use to define other state functions. In Chapter 8, we use the energy function to define enthalpy. We use the energy and entropy
functions to define two more state functions that also prove to have useful properties. These are the Helmholtz and Gibbs free
energies. The Helmholtz free energy is usually given the symbol A , and the Gibbs free energy is usually given the symbol G. We
define them by
A = E −TS (Helmholtz free energy)
and
G = H −TS (Gibbs free energy)
Note that P V , T S , H , A , and G all have the units of energy, E .

The sense of the name “free energy” is that a constant-temperature process in which a system experiences an entropy increase (
ΔS > 0 ) is one in which the system’s ability to do work in the surroundings is increased by an energy increment T ΔS . Then,
adding T ΔS to the internal energy lost by the system yields the amount of energy that the process actually has available (energy
that is “free”) to do work in the surroundings. When we consider how ΔA and ΔG depend on the conditions under which system
changes, we find that this idea leads to useful results.
The rest of this chapter develops important equations for ΔE ΔH, , ΔS , ΔA, and ΔG that result when we require that a system
change occur under particular sets of conditions.
This page titled 9.13: Defining the Helmholtz and Gibbs Free Energies is shared under a CC BY-SA 4.0 license and was authored, remixed,
To begin exploring the possibilities for stating the criteria for change using only the properties of the system, let us consider how
some thermodynamic functions change when a process is reversible. We consider a closed system and focus on making incremental
changes in the state of the system. For a reversible process, we have dq = T dS . The reversible pressure–volume work is
rev
= −P dV . If non-pressure–volume work is also possible, the reversible work becomes dw , where

rev rev rev
dw PV
= −P dV + dw N PV
dw
rev
N PV
is the increment of reversible, non-pressure–volume work. The energy change is
rev rev rev
dE = dq + dw = T dS − P dV + dw
N PV
(any reversible process)

This equation is of central importance. It is sometimes called the combined first and second laws of thermodynamics or the
fundamental equation. It applies to any closed system that is undergoing reversible change. It specifies a relationship among the
changes in energy, entropy, and volume that must occur if the system is to remain at equilibrium while an increment of non-
pressure–volume work, dw rev
N PV
, is done on it. The burden of our entire development is that any reversible process must satisfy this
equation. Conversely, any process that satisfies this equation must be reversible.
For a reversible process at constant entropy, we have dS = 0 , so that (dE) = −P dV + dw S
rev
N PV
. Since −P dV is the reversible
pressure–volume work, dw , and the sum dw = −P dV + dw
rev
PV
rev
net
is the net work, we have
rev
N PV
rev
(dE) = dw
S net
(reversible process, constant S)

where the subscript “S ” specifies that the entropy is constant. For a reversible process in which all of the work is pressure–volume
work, we have dw rev
N PV
= 0 , and the fundamental equation becomes
dE = T dS − P dV
(reversible process, only pressure–volume work)

For a reversible process in which only pressure–volume work is possible, this equation gives the amount, dE , by which the energy
must change when the entropy changes by dS and the volume changes by dV .
Now, let us apply the fundamental equation to an arbitrary process that occurs reversibly and at constant entropy and constant
volume. Under these conditions, dS = 0 and dV = 0 . Therefore, at constant entropy and volume, a necessary and sufficient
condition for the process to be reversible—and hence to be continuously in an equilibrium state as the process takes place—is that
rev
(dE)SV = dw
N PV
(reversible process)
and if only pressure–volume work is possible,
(dE) =0
SV

where the subscripts indicate that entropy and volume are constant.
If we consider an arbitrary reversible process that occurs at constant energy and volume, we have dE = 0 and dV = 0 , and the
fundamental equation reduces to
rev
dw
N PV
(dS) =−
EV
T
(dS) =0
EV
In this case, as noted in §7, the system is isolated. In §1-6, we note that an isolated system in an equilibrium state can undergo no
further change. Thus, the condition (dS) = 0 defines a unique or primitive equilibrium state.
EV
If a closed system behaves reversibly, any composition changes that occur in the system must be reversible. For chemical
applications, composition changes are of paramount importance. We return to these considerations in Chapter 14, where we relate
the properties of chemical substances—their chemical potentials—to the behavior of systems undergoing both reversible and
spontaneous composition changes.
If a closed system behaves reversibly and only pressure–volume work is possible, we see from the fundamental equation that
specifying the changes in any two of the three variables, E , S , and V , is sufficient to specify the change in the system. In
particular, if energy and entropy are constant, dE = dS = 0 , the volume is also constant, dV = 0 , and the system is isolated.
Thus, the state of an equilibrium system whose energy and entropy are fixed is unique; dE = dS = 0 specifies a primitive
equilibrium state. We see that the internal consistency of our model passes a significant test: From the entropy-based statement of
the second law, we deduce the same proposition that we introduce in §7 as a heuristic conjecture. In Chapter 10, we expand on this
idea.
Starting from the fundamental equation, we can find similar sets of relationships for enthalpy, the Helmholtz free energy, and the
Gibbs free energy. We define H = E + P V . For an incremental change in a system we, have
dH = dE + P dV + V dP
Using the fundamental equation to substitute for dE, this becomes

rev rev
dH = T dS − P dV + dw + P dV + V dP = T dS + V dP + dw
N PV N PV
For a reversible process in which all of the work is pressure–volume work, we have
dH = T dS + V dP

For a reversible process in which only pressure–volume work is possible, this equation gives the amount, dH , by which the
enthalpy must change when the entropy changes by dS and the pressure changes by dP . If a reversible process occurs at constant
entropy and pressure, then dS = 0 and dP = 0 . At constant entropy and pressure, the process is reversible if and only if
rev
(dH ) = dw
SP N PV
If only pressure–volume work is possible,
(dH ) =0
SP

where the subscripts indicate that entropy and pressure are constant.
If we consider an arbitrary reversible process that occurs at constant enthalpy and pressure, we have dH = 0 and dP = 0 , and
the total differential for dH reduces to
rev
dw
(dS)
HP
=−
NPV
T
(dS) =0
HP

From A = E − T S , we have dA = dE − T dS − SdT . Using the fundamental equation to substitute for dE , we have
rev rev
dA = T dS − P dV + dw − T dS − SdT = −P dV − SdT + dw
N PV N PV
For a reversible process in which all of the work is pressure–volume work,

dA = −SdT − P dV
For a reversible process in which only pressure–volume work is possible, this equation gives the amount, dA , by which the
Helmholtz free energy must change when the temperature changes by dT and the volume changes by dV . For a reversible
isothermal process, we have dT = 0 , and from dA = −P dV − SdT + dw we have rev
N PV
rev rev rev rev

(dA)T = −P dV + dw = dw + dw = dwnet
N PV PV N PV
(reversible isothermal process)

where we recognize that the reversible pressure–volume work is dw = −P dV , and the work of all kinds is
rev
PV
dw
rev
net
= dw
rev
PV
+ dw
rev
N PV
. We see that (dA) is the total of all the work done on the system in a reversible process at constant
T
temperature. This is the reason that “A ” is used as the symbol for the Helmholtz free energy: “A ” is the initial letter in “Arbeit,” a
German noun whose meaning is equivalent to that of the English noun “work.”
If a reversible process occurs at constant temperature and volume, we have dT = 0 and dV = 0 . At constant temperature and
volume, a process is reversible if and only if
rev
(dA)T V = dw
N PV
(dA) =0
TV

where the subscripts indicate that volume and temperature are constant. (Of course, these conditions exclude all work, because
constant volume implies that there is no pressure–volume work.)
From
G = H −TS = E +PV −TS
and the fundamental equation, we have

rev
dG = dE + P dV + V dP − SdT − T dS = T dS − P dV + dw − T dS + P dV + V dP
N PV
rev
− SdT = −SdT + V dP + dw
N PV
For a reversible process in which all of the work is pressure–volume work,
dG = −SdT + V dP

For a reversible process in which only pressure–volume work is possible, this equation gives the amount, dG, by which the Gibbs
free energy must change when the temperature changes by dT and the pressure changes by dP . For a reversible process that occurs
at constant temperature and pressure, dT = 0 and dP = 0 . At constant temperature and pressure, the process will be reversible if
and only if
rev
(dG) = dw
TP N PV
(any reversible process)

(dG) =0
TP

where the subscripts indicate that temperature and pressure are constant.
In this section, we develop several criteria for reversible change, stating these criteria as differential expressions. Since each of
these expressions applies to every incremental part of a reversible change that falls within its scope, corresponding expressions
apply to finite changes. For example, we find (dE) = dwS
for every incremental part of a reversible process in which the
rev
net
entropy has a constant value. Since we can find the energy change for a finite amount of the process by summing up the energy
changes in every incremental portion, it follows that
rev
(ΔE) =w
S net
Each of the other differential-expression criteria for reversible change also gives rise to a corresponding criterion for a finite
reversible change. These criteria are summarized in §25.
In developing the criteria in this section, we stipulate that various combinations of the thermodynamic functions that characterize
the system are constant. We develop these criteria for systems undergoing reversible change; consequently, the requirements
imposed by reversibility must be satisfied also. In particular, the system must be composed of homogeneous phases and its
temperature must be the same as that of the surroundings. The pressure of the system must be equal to the pressure applied to it by
the surroundings. When we specify that a reversible process occurs at constant temperature, we imagemean that
T =T ^ = constant . When we specify that a reversible process occur at constant pressure, we mean that
P =P = constant .
applied
This page titled 9.14: The Fundamental Equation and Other Criteria for Reversible Change is shared under a CC BY-SA 4.0 license and was
In a reversible process, the changes that occur in the system are imposed by the surroundings; reversible change occurs only
because the system responds to changes in the conditions imposed on it by its surroundings. A reversible process is driven by the
surroundings. In contrast, a spontaneous process is driven by the system. Nevertheless, when a spontaneous process occurs under
some specific set of imposed conditions (specific values of the temperature and pressure, for example) the system’s equilibrium
state depends on these conditions. To specify a particular spontaneous change, we must specify enough constraints to fix the final
state of the system.
To see these points from a slightly different perspective, let us consider a closed reversible system in which only pressure–volume
work is possible. Duhem’s theorem asserts that a change in the state of this system can be specified by specifying the changes in
some pair of state functions, say X and Y . If the imposed values of X and Y are constant at their eventual equilibrium values, but
the system is changing, the system cannot be on a Gibbsian equilibrium manifold. We say that the system is undergoing a
spontaneous change at constant X and Y .
This description is a figure of speech in that the system’s X and Y values do not necessarily attain the imposed values and become
constant until equilibrium is reached. An example is in order: A system whose original pressure and temperature are P and T can
i i
undergo a spontaneous change while the surroundings impose a constant pressure, P = P , and the system is immersed in
applied f
constant temperature bath at T = T . The pressure and temperature of the system may be indeterminate as the process occurs, but
f
the equilibrium pressure and temperature must be P and T .

f f
If the surroundings operate to impose particular values of X and Y on the system, then the position at which the system
eventually reaches equilibrium is determined by these values. The same equilibrium state is reached for any choice of
surroundings that imposes the same values of X and Y on the system at the time that the system reaches equilibrium. For every
additional form of non-pressure–volume work that affects the system, we must specify the value of one additional variable in order
to specify a unique equilibrium state.
The entropy changes that occur in the system and its surroundings during a spontaneous process have predictive value. However,
our definitions do not enable us to find the entropy change for a spontaneous process, and the temperature of the system may not
have a meaningful value. On the other hand, we can always carry out the process so that the temperature of the surroundings is
known at every point in the process. Indeed, if the system is in thermal contact with its surroundings as the process occurs, we
cannot specify the conditions under which the process occurs without specifying the temperature of the surroundings along this
path.
Figure 9 describes a spontaneous process whose path can be specified by the values of thermodynamic variable Y and the
temperature of the surroundings, T^, as a function of time, t . Let us denote the curve that describes this path as C . We can divide
this path into short intervals. Let C denote a short segment of this path along which the temperature of the surroundings is
k
approximately constant. For our present purposes, the temperature of the system, T , is irrelevant; since the process is spontaneous,
the temperature of the system may have no meaningful value within the interval C . As the system traverses segment C , it accepts
k k
a quantity of heat, q , from the surroundings, which are at temperature T^ . The heat exchanged by the surroundings within C is
k k k
^ = −q . Below, we show that it is always possible to carry out the process in such a way that the change in the surroundings
q k
k
occurs reversibly. Then

^
q qk
^ k
ΔS k = =−
^ ^
T k T k
and since ΔS k
^
+ ΔS k > 0 , it follows that
qk
ΔSk >
^
T k
This is the Clausius inequality. It plays a central role in the thermodynamics of spontaneous processes. When we make the
intervals C arbitrarily short, we have
k
dqk
dSk >
^
T k
To demonstrate that we can measure the entropy change in the surroundings during a spontaneous process, let us use a conceptual
device to transfer the heat, q , that must be exchanged from the surroundings at temperature, T^ to the system. As sketched in
k k
Figure 10, we imagine a very small, reversible, ideal-gas Carnot engine, whose high-temperature reservoir is also very small. We
suppose that the Carnot engine delivers a very small heat increment δq to the high temperature reservoir in every cycle. While the
system is within C , we maintain the Carnot engine’s high temperature reservoir at T^ , and allow heat q to pass from the high
k k k
temperature reservoir to the system. The high temperature reservoir is the only part of the surroundings that is in thermal contact
with the system; q is the only heat exchanged by the system while it is within C .
k k
Figure 10. Using a reversible Carnot engine to exchange heat with a spontaneously changing system.
To maintain the high temperature reservoir at T^ we operate the Carnot engine for a large integral number of cycles, n , such that
k
q ≈ n × δq , and do so at a rate that just matches the rate at which heat passes from the high-temperature reservoir to the system.
k
When the system passes from path-segment C to path-segment C , we alter the steps in the reversible Carnot cycle to maintain
k k+1
the high-temperature reservoir at the new surroundings temperature, T^ . The low-temperature heat reservoir for this Carnot
k+1
engine is always at the constant temperature T^ . Let the heat delivered from the high-temperature reservoir to the Carnot engine
ℓ
within C be (q^ ) . We have q = −(q^ ) . Let the heat delivered from the low-temperature reservoir to the Carnot engine within
k k h k k h
Ck be (q^ ) . Let the heat delivered to the low-temperature reservoir within C be q . We have q = −(q^ ) . Since the Carnot
k ℓ k ℓ ℓ k ℓ
engine is reversible, we have

(q
^k ) (q
^k )
h ℓ
+ =0
^
T ^
T
k ℓ
and
qk qℓ
− − =0
^ ^
T k T ℓ
so that
qℓ qk
=−
^ ^
T ℓ T k
While the system is within C , it receives an increment of heat q from the high temperature reservoir. Simultaneously, three
k k
components in the surroundings also exchange heat. Let the entropy changes in the high-temperature reservoir, the Carnot engine,
and the low-temperature reservoir be ^
(ΔS H T ) , ^
(ΔS C E ) , and ^
(ΔS LT ) , respectively. The high temperature reservoir
k k k
receives heat qk from the Carnot engine and delivers the same quantity of heat to the system. The net heat accepted by the high
temperature reservoir is zero. No change occurs in the high-temperature reservoir. We have (ΔS^ HT ) =0 . The reversible Carnot
k
engine completes an integral number of cycles, so that (ΔS^ CE ) =0 . The low temperature reservoir accepts heat − (q^ k
)
ℓ
= qℓ ,
k
at the fixed temperature T^ , during the reversible operation of the Carnot engine, so that
ℓ
qℓ qk
^
(ΔS LT ) = =−
k ^
T ^
T
ℓ k
The entropy change in the surroundings as the system passes through C is k
qℓ qk
^ ^ ^ ^
ΔS k = (ΔS H T ) + (ΔS C E ) + (ΔS LT ) = =−
k k k ^ ^
T ℓ T k
so that, as we observed above,

qk
^
ΔSk > −ΔS k =
^
T k
Since C can be any part of path C, and C can be made arbitrarily short, we have for every increment of any spontaneous process
k k
occurring in a closed system that can exchange heat with its surroundings, dS^ = −dq/T^ , and
dq
dS >
^
T
If the temperature of the surroundings is constant between any two points A and B on curve C, we can integrate over this interval to
obtain Δ ˆ
AB S=
ˆ
−qAB / T and
qAB
ΔAB S >
ˆ
T
For an adiabatic process, q = 0 . For any arbitrarily small increment of an adiabatic process, dq = 0 . It follows that ΔS > 0 and
dS > 0 for any spontaneous adiabatic process.
This page titled 9.15: Entropy and Spontaneous Change is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
For every incremental part of any process, we have dS + dS^ ≥ 0 . Let us define a new quantity, the external entropy change, as
d S = −dS . The change criteria become dS − d S ≥ 0 . Now, let us define the internal entropy change as d S = dS − d S .
^ 1
e e i e
The entropy change for a system is the sum of its internal and external entropy changes, dS = d S + d S . We use d S and d S to
i e i e
represent incremental changes. To represent macroscopic changes, we use Δ S and Δ S . Since two processes can effect different
i e
changes in the surroundings while the change that occurs in the system is the same, ΔS^ and Δ S are not completely determined by
e
the change in the state of the system. Neither the internal nor the external entropy change depends solely on the change in the state
of the system. Nevertheless, we see that d S ≥ 0 or Δ S ≥ 0 is an alternative expression of the thermodynamic criteria.
i i
The external entropy change is that part of the entropy change that results from the interaction between the system and its
surroundings. The internal entropy is that part of the entropy change that results from processes occurring entirely within the
system. (We also use the term “internal energy.” The fact that the word “internal” appears in both of these terms does not reflect
any underlying relationship of material significance.) The criterion d S > 0 makes it explicit that a process is spontaneous if and
i
only if the events occurring within the system act to increase the entropy of the system. In one common figure of speech, we say
“entropy is produced” in the system in a spontaneous process. (It is, of course, possible for a spontaneous process to have d S > 0 i
while d S < 0 , and dS < 0 .)

e
In Section 14.1, we introduce a quantity,

ω
∑ μj dnj
j=1
that we can think of as a change in the chemical potential energy of a system. The internal entropy change is closely related to this
quantity: We find
ω
1
di S = − ∑ μj dnj
T
j=1
As required by the properties of d S, we find that ∑ μ dn ≤ 0 is an expression of the thermodynamic criteria for change.
i
ω
j=1 j j
Internal entropy is a useful concept that is applied to particular advantage in the analysis of many different kinds of spontaneous
processes in non-homogeneous systems.
This page titled 9.16: Internal Entropy and the Second Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
We now want to consider criteria for a spontaneous process in which a closed system passes from state A to state B. State B can be
an equilibrium state, but state A is not. We can denote the energy change for this process as Δ E , and we can find it by
AB
measuring the heat and work exchanged with the surroundings as the process takes place, Δ E = q + w , or for a process in
AB
which the increments of heat and work are arbitrarily small, d E = dq + dw . Likewise, we can denote the entropy change for
AB
the spontaneous process as Δ S or d S , but we cannot find the entropy change by measuring q
AB AB or dqspon
. If we cannot find
spon
the entropy change, we cannot find the Helmholtz or Gibbs free energy changes from their defining relationships, A = E − T S
and G = H − T S . Moreover, intensive variables—pressure, temperature, and concentrations—may not have well-defined values
in a spontaneously changing system.
When we say that a reversible process occurs with some thermodynamic variable held constant, we mean what we say: The
thermodynamic variable has the same value at every point along the path of reversible change. In the remainder of this chapter, we
develop criteria for spontaneous change. These criteria are statements about the values of ΔE, ΔH , ΔA, and ΔG for a system that
can undergo spontaneous change under particular conditions. In stating some of these criteria, we specify the conditions by saying
that the pressure or the temperature is constant. As we develop these criteria, we will see that these stipulations have specific
meanings. When we say that a process occurs “at constant volume” (isochorically), we mean that the volume of the system remains
the same throughout the process. When we say that a spontaneous process occurs “at constant pressure” (isobarically or
isopiestically), we mean that the pressure applied to the system by the surroundings is constant throughout the spontaneous process
and that the system pressure is equal to the applied pressure, P = P , at all times. When we say that a spontaneous process
applied
occurs “at constant temperature”, we may mean only that

1. the system is continuously in thermal contact with its surroundings
2. the temperature of the surroundings is constant
3. in the initial and final states, the system temperature is equal to the surroundings temperature.
This page titled 9.17: Notation and Terminology- Conventions for Spontaneous Processes is shared under a CC BY-SA 4.0 license and was
To continue our effort to find change criteria that use only properties of the system, let us consider a spontaneous process, during
which the system is in contact with its surroundings and the entropy of the system is constant. For every incremental part of this
process, we have dS = 0 and dS + dS^ > 0 . Hence, dS^ > 0 . It follows that ΔS = 0 , ΔS + ΔS^ > 0 , and ΔS^ > 0 . (Earlier, we
found that the entropy changes for a spontaneous process in an isolated system are ΔS > 0 and ΔS^ = 0 . The present system is
not isolated.) Since the change that occurs in the system is irreversible, dS = 0 does not mean that dq = 0 . The requirement that
dS = 0 places no constraints on the temperature of the system or of the surroundings at any time before, during, or after the
process occurs.
In Section 9.15, we find dS > dq spon ^
/T for any spontaneous process in a closed system. If the entropy of the system is constant,
we have
spon
dq <0
(spontaneous process, constant entropy)

for every incremental part of the process. For any finite change, it follows that the overall heat must satisfy the same inequality:
spon
q <0
(spontaneous process, constant entropy)

For a spontaneous process that occurs with the system in contact with its surroundings, but in which the entropy of the system is
constant, the system must give up heat to the surroundings. dq < 0 and q < 0 are criteria for spontaneous change at constant
system entropy.
In Section 9.14, we develop criteria for reversible processes. The criteria relate changes in the system’s state functions to the
reversible non-pressure–volume work that is done on the system during the process. Now we can develop parallel criteria for
spontaneous processes.
This page titled 9.18: The Heat Exchanged by A Spontaneous Process at Constant Entropy is shared under a CC BY-SA 4.0 license and was
From the fundamental equation, dE − T dS + P dV = dw for a reversible process. We find that the criterion for reversible
rev
N PV
change at constant entropy is (dE) = dw . For a reversible process at constant entropy and volume, we find
S
rev
net
rev
(dE) = dw
SV N PV
To consider the energy change for a spontaneous process, we begin with dE = dq + dw , which is independent of whether the
change is spontaneous or reversible. For a spontaneous process in which both pressure–volume, dw , and non-pressure–volume
spon
PV
work, dw spon
N PV
, are possible, we have dE = dq spon
+ dw
spon
PV
+ dw , which we can rearrange to
spon
N PV
spon spon spon

dE − dw − dw = dq
PV N PV
For a spontaneous, constant-entropy change that occurs while the system is in contact with its surroundings, we have dq
spon
<0 .
spon spon spon spon spon
Hence, we have (dE) − dwS PV
− dw < 0 . Lettting dw
N PV
= dw
net
+ dw
PV
, we can express this as
N PV
spon
(dE) < dwnet
S
(spontaneous process, constant S)

If we introduce the further condition that the spontaneous process occurs while the volume of the system remains constant, we have
= 0 . Making this substitution and repeating our earlier result for a reversible process, we have the parallel relationships
spon
dw
PV
spon
(dE) < dw
SV N PV
(spontaneous process, constant S and V)

rev
(dE) = dw
SV N PV
(reversible process, constant S and V)

spon
If we introduce the still further requirement that only pressure–volume work is possible, we have dw
N PV
=0 . The parallel
relationships become
(dE) <0
SV
(spontaneous process, constant S and V, only PV work)
(dE)SV = 0
(reversible process, constant S and V, only PV work)

These equations state the criteria for change under conditions in which the entropy and volume of the system remain constant. If
the process is reversible, the energy change must be equal to the non-pressure–volume work. If the process is spontaneous, the
energy change must be less than the non-pressure volume work. If only pressure–volume work is possible, the energy of the system
must decrease in a spontaneous process and remain constant in a reversible process. Each of these differential-expression criteria
applies to every incremental part of a change that falls within its scope. In consequence, corresponding criteria apply to finite
spontaneous changes. These criteria are listed in the summary in Section 9.25.
Now the question arises: What sort of system can undergo a change at constant entropy? If the process is reversible and involves no
heat, the entropy change will be zero. If we have a system consisting of a collection of solid objects at rest, we can rearrange the
objects without transferring heat between the objects and their surroundings. For such a process, the change in the energy of the
system is equal to the net work done on the system. Evidently, reversible changes in mechanical systems occur at constant entropy
and satisfy the criterion
rev
(dE) = dw
S net
For a change that occurs reversibly and in which the entropy of the system is constant, the energy change is equal to the net work
(of all kinds) done on the system. A spontaneous change in a mechanical system dissipates mechanical energy as heat by friction. If
this heat appears in the surroundings and the thermal state of the system remains unchanged, such a spontaneous processes satisfies
the criterion
spon
(dE) < dw
S net
We have arrived at the criterion for change that we are accustomed to using when we deal with a change in the potential energy of a
constant-temperature mechanical system: A spontaneous change can occur in such a system if and only if the change in the
system’s energy is less than the net work done on it. The excess work is degraded to heat that appears in the surroundings. This
convergence notwithstanding, the principles of mechanics and those of thermodynamics, while consistent with one another, are
substantially independent. We address this issue briefly in Section 12.2.
In the next section, we develop spontaneous-change criteria based on the enthalpy change for a constant-entropy process. In
subsequent sections, we consider other constraints and find other criteria. We find that the Helmholtz and Gibbs free energy
functions are useful because they provide criteria for spontaneous change when the process is constrained to occur isothermally.
This page titled 9.19: The Energy Change for A Spontaneous Process at Constant S and V is shared under a CC BY-SA 4.0 license and was
From H = E + P V , we have dH = dE + d (P V ) . For a spontaneous process in which both pressure–volume and non-pressure–
spon spon
volume work are possible, we can write this as dH = dq + dw + dw
spon
+ d (P V ) , which we can rearrange to
PV N PV
spon spon
dH − dw
PV
− dw
N PV
− d (P V ) = dq . For a spontaneous constant-entropy change that occurs while the system is in
spon
contact with its surroundings, we have dq < 0 , so that

spon
spon spon
(dH ) − dw − dw − d (P V ) < 0.
S PV N PV
Now, let us introduce the additional constraint that the system is subjected to a constant applied pressure, P , throughout the
applied
process. Thus P is a well-defined property that can be measured at any stage of the process. The incremental pressure–
applied
volume work done by the surroundings on the system is dw = −P

spon
PV
dV . In principle, the system can undergo spontaneous
applied
change so rapidly that there can be a transitory difference between the system pressure and the applied pressure. In practice,
pressure adjustments occur very rapidly. Except in extreme cases, we find that P = P is a good approximation at all times.
applied
Then the change in the pressure volume product is d (P V ) = P dV . Making these substitutions, the enthalpy inequality
applied
becomes
spon
(dH ) < dw
SP N PV
(spontaneous process, constant S and P applied )

From our earlier discussion of reversible processes, we have the parallel relationship
rev
(dH )SP = dwN P V
(any reversible process, constant S and P applied )

If we introduce the still further requirement that only pressure–volume work is possible, we have dwN P V = 0 . The parallel
relationships become
(dH ) <0
SP
(spontaneous process, constant S and P, only PV work)
(dH ) =0
SP
(reversible process, constant S and P, only PV work)

These equations state the criteria for change under conditions in which the entropy and pressure of the system remain constant. If
the process is reversible, the enthalpy change must be equal to the non-pressure–volume work. If the process is spontaneous, the
enthalpy change must be less than the non-pressure–volume work. If only pressure–volume work is possible, the enthalpy of the
system must decrease in a spontaneous process and remain constant in a reversible process. Since each of these differential criteria
applies to every incremental part of a reversible change that falls within its scope, corresponding criteria apply to finite spontaneous
changes. These criteria are listed in the summary in Section 9.25.
This page titled 9.20: The Enthalpy Change for A Spontaneous Process at Constant S and P is shared under a CC BY-SA 4.0 license and was
For any spontaneous process, we have dE = dq spon
+dwspon
, which we can rearrange to dq spon
= dE − dw
spon
. Substituting our
result from Section 9.15, we have
^dS > dE − dwspon
T
(spontaneous process)
If the energy of the system is constant throughout the process, we have dE = 0 and
^(dS) spon
T > −dw
E
(spontaneous process, constant energy)

spon spon
The spontaneous work is the sum of the pressure–volume work and the non-pressure–volume work, dw = dw
spon
+ dw . PV N PV
If we introduce the further condition that the spontaneous process occurs while the volume of the system remains constant, we have
spon
dw = 0 . Making this substitution and repeating our earlier result for a reversible process, we have the parallel relationships
PV
spon
−dw
N PV
(dS) >
EV
^
T
(spontaneous process, constant E and V )

spon
−dw
N PV
(dS) =
EV
^
T
(reversible process, constant E and V )

(For a reversible process, T = T^ .) If the spontaneous process occurs while ^
T is constant, summing the incremental contributions
to a finite change of state produces the parallel relationships
spon
−w
N PV
(ΔS) >
EV
^
T
(spontaneous process, constant E , V , and T^)

spon
−w
N PV
(ΔS) =
EV
^
T
(reversible process, constant E , V , and T^)

Constant T^ corresponds to the common situation in chemical experimentation in which we place a reaction vessel in a constant-
temperature bath. If we introduce the further condition that only pressure–volume work is possible, we have dw = 0 . The
spon
N PV
parallel relationships become
(dS) >0
EV
(spontaneous process, constant E and V , only P V work)
(dS) =0
EV
(reversible process, constant E and V , only P V work)

If the energy and volume are constant for a system in which only pressure–volume work is possible, the system is isolated. The
conditions we have just derived are entirely equivalent to our earlier conclusions that dS = 0 and dS > 0 for an isolated system
that is at equilibrium or undergoing a spontaneous change, respectively. Summing the incremental contributions to a finite change
of state produces the parallel relationships
(ΔS) EV > 0
(spontaneous process, only P V work)
(ΔS) =0
EV
(reversible process, only P V work)

The validity of these expressions is independent of any variation in either T or T^.
This page titled 9.21: The Entropy Change for A Spontaneous Process at Constant E and V is shared under a CC BY-SA 4.0 license and was
For any spontaneous process, we have
spon spon
dH = dE + P dV + V dP = dq − Papplied dV + dw + P dV + V dP
N PV
spon
If the pressure is constant (P = Papplied = constant ), this becomes dq spon
= dH − dw
N PV
. Substituting our result from Section
9.15, we have
spon
^(dS)
T > dH − dw
P N PV
(spontaneous process, constant P )

If the enthalpy of the system is also constant throughout the process, we have
spon
^
T (dS) H P > −dw
N PV
(spontaneous process, constant H and P )

Dividing by T^ and repeating our earlier result for a reversible process, we have the parallel relationships
spon
−dw
N PV
(dS) >
HP
^
T
(spontaneous process, constant H and P )

rev
−dw
N PV
(dS) =
HP
^
T
(reversible process, constant H and P )

If it is also true that the temperature of the surroundings is constant, summing the incremental contributions to a finite change of
state produces the parallel relationships
spon
−w
N PV
(ΔS) >
HP
^
T
(spontaneous process, constant H , P , and T^)

rev
−w
N PV
(ΔS) >
HP
^
T
(reversible process, constant H , P , and T^ = T )

spon
If only pressure–volume work is possible, we have dw N PV
=0 , and
(dS) >0
HP
(spontaneous process, constant H , P , only P V work)
(dS) =0
HP
(reversible process, constant H and P , only P V work)

and for a finite change of state,
(ΔS) >0
HP
(spontaneous process, only P V work)
(ΔS) =0
HP
(reversible process, only P V work)

In this and earlier sections, we develop criteria for spontaneous change that are based on dE and dH . We are now able to develop
similar criteria for a spontaneous change in a system that is in thermal contact with constant-temperature surroundings. These
criteria are based on dA and dG. However, before doing so, we develop a general relationship between the isothermal work in a
spontaneous process and the isothermal work in a reversible process, when these processes take a system from a common initial
state to a common final state.
This page titled 9.22: The Entropy Change for A Spontaneous Process at Constant H and P is shared under a CC BY-SA 4.0 license and was
The Clausius inequality leads to an important constraint on the work that can be done on a system during a spontaneous process in
which the temperature of the surroundings is constant. As we discuss in Section 9.7, the initial state of the spontaneous process
cannot be a true equilibrium state. In our present considerations, we assume that the initial values of all the state functions of the
spontaneously changing system are the same as those of a true equilibrium system. Likewise, we assume that the final state of the
spontaneously changing system is either a true equilibrium state or a state whose thermodynamic functions have the same values as
those of a true equilibrium system.
From the first law applied to any spontaneous process in a closed system, we have ΔE = ΔE rev
and spon
q
rev
+w
rev
=q
spon
+w
spon
. Since the temperature of the system and its surroundings are equal and constant for the reversible
process, we have q = T ΔS = T^ΔS . So long as the temperature of the surroundings is constant, we have q
rev
< T ΔS for the
^ spon
spontaneous process. It follows that

^ΔS + wrev − wspon = q spon < T
T ^ΔS
so that
rev spon
w <w
(T^ constant)
A given isothermal process does the minimum possible amount of work on the system when it is carried out reversibly. (In Section
7.20, we find this result for the special case in which the only work is the exchange of pressure–volume work between an ideal gas
and its surroundings.) Equivalently, a given isothermal process produces the maximum amount of work in the surroundings when it
is carried out reversibly: Since w = −w
rev
^
rev
and w spon
= −w ^ , we have −w
spon
^
rev
< −w ^ or
spon
rev spon
w
^ >w
^
This page titled 9.23: The Reversible Work is the Minimum Work at Constant Tˆ is shared under a CC BY-SA 4.0 license and was authored,
Now let us consider the change in the Helmholtz free energy when a system undergoes a spontaneous change while in thermal
contact with surroundings whose temperature remains constant at T^. We begin by considering an arbitrarily small increment of
change in a process in which the temperature of the system remains constant at T = T^ . The change in the Helmholtz free energy
for this process is (dA) = dE − T dS . Substituting dE = dq
T
+ dw gives spon spon
spon spon
(dA) = dq + dw − T dS
T
(spontaneous process, constant T )

Rearranging, we have (dA) T
− dw
spon
+ T dS = dq
spon
. Using the inequality dq spon ^
< T dS , we have
spon ^
(dA) − dw + T dS < T dS
T
When we stipulate that T ^ = constant

=T , this becomes
spon
(dA) < dw
T

where dw spon
is all of the work of any kind done on the system during a small increment of the spontaneous process. If we
spon spon
introduce the still further requirement that the volume is constant, we have dw = 0 and dw = dw . Then
PV
spon
N PV
spon
(dA) < dwnpv
TV
(spontaneous process, constant T and V )

(dA) <0
TV
(spontaneous process, constant T and V , only P V work)

From our earlier discussion of reversible processes, we have the parallel relationships
rev
(dA) = dw
T net
(reversible isothermal process)

rev
(dA) = dw
TV N PV
(reversible process at constant T and V )
(dA) =0
TV
(reversible process at constant T and V , only P V work)

Similarly, under these conditions, the change in the Gibbs free energy for a spontaneous isothermal process is
(dG) = dH − T dS
T
= dE + d (P V ) − T dS
spon spon spon
= dq + dw + dw + d (P V ) − T dS
PV N PV
Rearranging, we have
spon spon spon
(dG) − dw − dw − d (P V ) + T dS = dq ^dS
<T
T PV N PV
and since T ^
= T = constant ,
spon spon
(dG) < dwpv + dwnpv + d (pv)
T
As we did when considering the enthalpy change for a spontaneous process, we introduce the additional constraints that the system
is subjected to a constant applied pressure, P , and that P = P
applied throughout the process. The irreversible pressure–
applied
volume work done by the surroundings on the system becomes dw = −P dV , and the change in the pressure volume
spon
PV applied
product becomes d (P V ) = P dV . The Gibbs free energy inequality becomes

applied
spon
(dG) < dwnpv
TP
(spontaneous process, constant P applied and T )

If only pressure–volume work is possible, this becomes
(dG) <0
TP
(spontaneous process, constant P applied and T , only P V work)

From our earlier discussion of reversible processes, we have the parallel relationships
rev
(dG) = dw
TP N PV
(reversible process, constant P and T )
(dG) =0
TP
(reversible process, constant P and T ,

only P V work)
Since each of these differential-expression criteria applies to every incremental part of a reversible change that falls within its
scope, we have the following criteria for finite spontaneous changes when the temperature of the system is constant:
spon
(ΔA) <w
T

spon
(ΔA) < wnpv
TV
(spontaneous process, constant T and V )
(ΔA) T V < 0
(spontaneous process, constant T and V , only P V work)

spon
(ΔG) T P < wnpv
(spontaneous process, constant P applied and T )
(ΔG) =0
TP
(spontaneous process, constant P applied and T , only P V work)

While the development we have just made assumes that the system temperature is strictly constant, the validity of these finite-
change inequalities is not restricted to the condition of strictly constant system temperature. We can derive these finite-change
inequalities by essentially the same argument from less restrictive conditions.
Let us consider a spontaneous process in which a system goes from state B to state C while in contact with surroundings whose
temperature remains constant at T^. We suppose that in both state B and state C the system temperature is equal to the surroundings
temperature; that is, T = T = T^ = constant . However, at any intermediate point in the process, the system can have any
B C
temperature whatsoever. In states B and C, the Helmholtz free energies are A = E − T^S and A = E − T^S . The change
B B B C C C
in the Helmholtz free energy is (A − A ) = (E −E ) − T^ (S − S ) or (ΔA) = ΔE − T^ΔS = q

C B C B C B +w
T
^
^
− T ΔS . spon spon
Rearranging, and using q spon ^

< T ΔS , we have
(ΔA)
T
^ −w
spon ^
+ T ΔS = q
spon ^
< T ΔS , so that
spon
(ΔA) ^ <w
T
(spontaneous process, constant T^)
If we require further that the system volume remain constant, there is no pressure–volume work, and we have
spon
(ΔA) ^ < wnpv
VT
(spontaneous process, constant T^ and V )

If only pressure–volume work is possible, w spon
N PV
=0 , and
(ΔA) ^ <0
VT
(spontaneous process, constant T^ and V , only P V work)

Under the same temperature assumptions, and assuming that P = P = P B C applied = constant , the Gibbs free energies are
G B=E +P
B V
applied
^
−T S B B and G =E +P c C applied VC
^
− T SC . So that
(G C− G ) = (E −E ) + P
B C B (V applied
^
− V ) − T (SC −S ) Bor C B
^
(ΔG) ^ = ΔE + Papplied ΔV − T ΔS
PT
spon spon spon

^
=q +w +w + Papplied ΔV − T ΔS
PV N PV
spon
The pressure–volume work is w PV
= −Papplied ΔV . Cancelling and rearranging, we have
spon spon
^ ^
(ΔG) ^ −w + T ΔS = q < T ΔS
PT N PV
and
spon
(ΔG) ^ < wnpv
PT
(spontaneous process, constant T^ and P )

(ΔG) ^ <0
PT
(spontaneous process, constant T^ and P , only P V work)

We find (ΔG)
PT
^
spon
< wnpv > for any spontaneous process that occurs at constant pressure, while the system is in contact with
surroundings at the constant temperature T^, and in which the initial and final system temperatures are equal to T^. These are the
most common conditions for carrying out a chemical reaction. Consider the situation after we mix non-volatile reactants in an open
vessel in a constant-temperature bath. We suppose that the initial temperature of the mixture is the same as that of the bath. The
atmosphere applies a constant pressure to the system. The reaction is an irreversible process. It proceeds spontaneously until its
equilibrium position is reached. Until equilibrium is reached, the reaction cannot be reversed by an arbitrarily small change in the
spon
applied pressure or the temperature of the surroundings. (ΔG) <w and (ΔG)
PT
^ < 0 are criteria for spontaneous change
npv
PT
^
that apply to this situation whatever the temperature of the system might be during any intermediate part of the process.
This page titled 9.24: The Free Energy Changes for A Spontaneous Process at Constant Tˆ is shared under a CC BY-SA 4.0 license and was
For a spontaneous process, we conclude that the entropy change of the system must satisfy the inequality ΔS + ΔS^ > 0. For any
process that occurs reversibly, we conclude that ΔS + ΔS^ = 0 . For every incremental part of a reversible process that occurs in a
closed system, we have the following relationships:
rev
dE = T dS − P dV + dw
N PV
rev
dH = T dS + V dP + dw
N PV
rev
dA = −SdT − P dV + dw
N PV
rev
dG = −SdT + V dP + dw
N PV
At constant entropy, the energy relationship becomes:

rev
(dE) = dw
S net
rev
(ΔE) =w
S net
At constant temperature, the Helmholtz free energy relationship becomes:

rev
(dA) = dw
T net
rev
(ΔA) =w
T net
For reversible processes in which all work is pressure–volume work:
dE = T dS − P dV
dH = T dS + V dP
dA = −SdT − P dV
dG = −SdT + V dP
From these general equations, we find the following relationships for reversible processes when various pairs of variables are held
constant:
rev rev
(dS) EV = −dw /T (ΔS) EV = −w /T
N PV N PV
rev rev
(dS) = −dw /T (ΔS) = −w /T
HP N PV HP N PV
rev rev
(dE) = dw (ΔE) =w
SV N PV SV N PV
rev rev
(dH ) = dw (ΔH ) =w
SP N PV SP N PV
rev rev
(dA) = dw (ΔA) =w
TV N PV TV N PV
rev rev
(dG) = dw (ΔG) =w
TP N PV TP N PV
If the only work is pressure–volume work, then dw rev

N PV
=0 ,w rev
N PV
=0 , and these relationships become:
(dS) = 0 (ΔS) =0
EV EV
(dS) = 0 (ΔS) =0
HP HP
(dE) = 0 (ΔE) =0
SV SV
(dH ) = 0 (ΔH ) =0
SP SP
(dA) = 0 (ΔA) =0
TV TV
(dG) = 0 (ΔG) =0
TP TP
For every incremental part of an irreversible process that occurs in a closed system at constant entropy:
spon
dq <0
and
spon
(dE) < dw
S net
and
spon
q <0
and
spon
(ΔE) < wnet
S
For an irreversible process at constant temperature:

spon ^dS
dq <T
and
spon
(dA) ^ < dw
T net
and
spon ^
q < T ΔS
and
spon
(ΔA) ^ < wnet
T
When an irreversible process occurs with various pairs of variables held constant, we find:
spon ^ spon ^
(dS) > −dw / T (ΔS) = −w /T
EV N PV EV N PV
spon ^ spon ^
(dS) > −dw / T (ΔS) > −w /T
HP N PV HP N PV
(dE)
SV
(dH )
SP
(dA) ^
T V
(dG) ^
T P
For irreversible processes in which the only work is pressure–volume work, these inequalities become:
(dS) > 0 (ΔS) >0

EV EV
(dS) > 0 (ΔS) >0

HP HP
(dE) < 0 (ΔE) <0

SV SV
(dH )SP < 0 (ΔH ) SP < 0
(dA) ^ < 0 (ΔA) ^ <0

T V T V
(dG) ^ < 0 (ΔG) ^ <0

T P T P
This page titled 9.25: Summary- Thermodynamic Functions as Criteria for Change is shared under a CC BY-SA 4.0 license and was authored,
9.26: Problems
Problems
1. Does a perpetual motion machine of the second kind violate the principle of conservation of energy?
2. What is the contrapositive of ^

(SL and ∼ MSL) ⇒ (ΔS < 0) ? It is a theorem of logic that
∼ (B and C) ⇒ (∼ Band/or ∼ C) . Interpret this theorem. Given that SL is true and that ∼ (SL and ∼ MSL) is true, prove
that ∼ MSL is true.
3. Max Planck introduced the following statement of the second law:
“It is impossible to construct an engine which will work in a complete cycle, and produce no effect except the raising of a weight
and the cooling of a heat-reservoir.”
(M. Planck, Treatise on Thermodynamics, 3rd Edition, translated from the seventh German Edition, Dover Publications, Inc., p
89.) Since we take “raising a weight” to be equivalent to “produces work in the surroundings,” the Planck statement differs from
our machine-based statement only in that it allows the temperature of the heat source to decrease as the production of work
proceeds. We can now ask whether this difference has any material consequences. In particular, can we prove that the Planck
statement implies our machine-based statement, or vice versa? (Suggestion: Suppose that we have identical Planck-type machines,
each with its own heat reservoir. We dissipate by friction the work produced by one machine in the heat reservoir of the other.)
4. Our statements of the first and second laws have a common format: Assertion that a state function exists; operational definition
by which the state function can be measured; statement of a property exhibited by this state function. Express the zero-th law of
thermodynamics (Chapter 1) in this format.
5. A 0.400 mol sample of N is compressed from 5.00 L to 2.00 L, while the temperature is maintained constant at 350 K. Assume
2
that N is an ideal gas. Calculate the change in the Helmholtz free energy, ΔA.
2
6. Show that ΔG = ΔA when an ideal gas undergoes a change at constant temperature.

7. Calculate ΔE, ΔH , and ΔG for the process in problem 5.
8. A sample of 0.200 mol of an ideal gas, initially at 5.00 bar, expands reversibly and isothermally from 1.00 L to 10.00 L.
Calculate ΔE, ΔH , and ΔG for this process.
9. A 100.0 g sample of carbon tetrachloride is compressed from 1.00 bar to 10.00 bar at a constant temperature of 20 C. At 20 C,
carbon tetrachloride is a liquid whose density is 1.5940 g mL . Assume that the density does not change significantly with
−1
pressure. What is ΔG for this process?

10. Calculate the Helmholtz free energy change (ΔA) in problem 9.
11. If C is constant, show that the initial and final temperatures and volumes for an adiabatic ideal-gas expansion are related by
V
the equation
R/CV
Tf Vi
( ) =( )
Ti Vf
12. At 25 C, the initial volume of a monatomic ideal gas is 5 L at 10 bar. This gas expands to 20 L against a constant applied
pressure of 1 bar.
(a) Is this process impossible, spontaneous, or reversible?
(b) What is the final temperature?
(c) Find q, w, ΔE, and ΔH for this process.
13. The same change of state experience by the monatomic ideal gas in problem 12 can be effected in two steps. Let step A be the
reversible cooling of the gas to its final temperature while the pressure is maintained constant at 10 bar. Let step B be the reversible
isothermal expansion of the resulting gas to a pressure of 1 bar.
(a) Find q, w, ΔE, and ΔH for step A.
(b) Find q, w, ΔE, and ΔH for step B.
(c) From your results in (a) and (b), find q , w, ΔE, and ΔH for the overall process of step A followed by step B.
(d) Compare the values of q, w , ΔE , and ΔH that you find in (c) to the values for the same overall process that you found in
problem 12.
(e) Find ΔS and ΔS^ for step A.

(f) Find ΔS and ΔS^ for step B.
(g) Find ΔS , ΔS^, and ΔS universe for the overall process.

14. Assume that the process in problem 12 occurs while the gas is in thermal contact with its surroundings and that the temperature
of the surroundings is always equal to the final temperature of the gas. Find ΔS^ and ΔS for this process.
universe
15. At 25 C, the initial volume of a monatomic ideal gas is 5 L at 10 bar. The gas expands to 20 L while in thermal contact with
surroundings at 125 C. During the expansion, the applied pressure is constant and equal to the equilibrium pressure at the final
volume and temperature.
(b) Find q, w, ΔE, ΔH , and ΔS^ for this process.

(c) Find ΔS and ΔS for this process. To find
universe ΔS , it is necessary to find a reversible alternative path that effects the same
change in the system’s state functions.
16. At 60 C, the density of water is 0.98320 g cm , the vapor pressure is 19,932 Pa, and the enthalpy of vaporization is
−3
42, 482 J mol . Assume that gaseous water behaves as an ideal gas. A vessel containing liquid and gaseous water is placed in a
−1
constant 60 C bath, and the applied pressure is maintained at 19,932 Pa while 100 g of water vaporizes.
(b) Find q, w, ΔE, ΔH , ΔS , ΔA, and ΔG for this process.
(c) Is (ΔS)
EV
=0 a criterion for equilibrium that applies to this system? Why or why not? (ΔH )
SP
=0 ? (ΔA)
VT
=0 ?
(ΔG)
PT
=0 ?
17. This problem compares the efficiency and ∑ q/T for one mole of a monatomic ideal gas taken around a reversible Carnot
cycle to the same quantities for the same gas taken around an irreversible cycle using the same two heat reservoirs.
(i) Let the successive step of the reversible Carnot cycle be a, b, c, and d. Isothermal step a begins with the gas occupying 5.00 L at
600 K and ends with the gas occupying 20.00 L. Adiabatic expansion step b ends with the gas at 300 K. After the isothermal
compression step c, the gas is adiabatically compressed in step d to the original state. Find P , V , and T for the gas at the end of
each step of this reversible cycle. Find ∑ q/T^, ΔS , and ΔS^ for the cycle a, b, c, d. What is the efficiency of this cycle?
(ii) Suppose that following step b, the ideal gas is warmed at constant volume to 400 K by exchanging heat with the 600 K
reservoir. Call this step e. Following step e, the gas is cooled at constant pressure to 300 K by contact with the 300 K reservoir. Call
this step f. Following step f, the gas is isothermally and reversibly compressed at 300 K to the same P , V , and T as the gas reaches
at the end of step c. Call this step g. Find P , V , and T for the gas at the ends of steps e, f, and g. Although steps e and f are not
reversible, the same changes can be effected reversibly by keeping T = T^ as the gas is warmed at constant volume (step e) and
400 K CV
cooled at constant pressure (step f). (We discuss this point in further in Section 12.4.) Consequently, Δe S = ∫
300 K T
dT and
400 K CP
Δf S = ∫
300 K T
dT . Find ^
∑ q/T ΔS , , and ^
ΔS , and ΔS universe for the cycle a, b, e, f, g, d. What is the efficiency of this cycle?
(iii) Compare the value of ∑ q/T^ that you obtained in part (ii) to value of ∑ q/T^ that you obtained in part (i).
(iv) Clausius’ theorem states that ∑ q/T^ = 0 for a cycle traversed reversibly, and ∑ q/T^ < 0 for a cycle traversed spontaneously.
Comment.
18. For a spontaneous cycle traversed while the temperature changes continuously, Clausius’ theorem asserts that ∮ dq/T^ < 0 .
Show that this inequality follows from the result, dS > dq/T^ , that we obtained in Section 9.15 for any spontaneous process in a
closed system.
19. In Sections 9.6 through 9.8, we conclude that ΔS + ΔS^ = 0 is necessary for a reversible process, ΔS + ΔS^ > 0 is necessary
for a spontaneous process, and ΔS + ΔS^ < 0 is necessary for an impossible process. That is:
^
(Process is reversible) ⇒ (ΔS + ΔS = 0)
(Process is spontaneous) ^
⇒ (ΔS + ΔS > 0) , and (Process is impossible) ^
⇒ (ΔS + ΔS < 0) .
Since we have defined the categories reversible, spontaneous, and impossible so that they are exhaustive and mutually exclusive,
the following proposition is true:
∼ (Process is spontaneous) and ∼ (Process is impossible)
⇒ (Process is reversible)
(a) Prove that ΔS + ΔS^ = 0 is sufficient for the process to be reversible; that is, prove:
^
(ΔS + ΔS = 0) ⇒ (Process is reversible)
(b) Prove that ΔS + ΔS^ > 0 is sufficient for the process to be spontaneous; that is, prove:
^
(ΔS + ΔS > 0 ) ⇒ (Process is spontaneous)
(c) Prove that ΔS + ΔS^ < 0 is sufficient for the process to be impossible; that is, prove:
^
(ΔS + ΔS < 0) ⇒ (Process is impossible)
20. Label the successive steps in a reversible Carnot cycle A, B, C, and D, where A is the point at which the pressure is greatest.
(a) Sketch the path ABCD in P –V space.
(b) Sketch the path ABCD in T –dq rev
/T space.
(c) Sketch the path ABCE in T – q rev
space.
(d) Sketch the path BCDA in T – q rev
space.
(e) Sketch the path CDAB in T – q rev
space.
(d) Sketch the path DABC in T – q rev
space.
21. Assume that the earth’s atmosphere is pure nitrogen and that it behaves as an ideal gas. Assume that the molar energy of this
¯¯
¯ ¯
¯¯¯ ¯
¯¯¯
nitrogen is constant and that its molar entropy changes are adequately modeled by dS = (CV /T ) dT + (R/ V ) dV . For this
atmosphere, show that
¯¯¯¯
¯
∂T −M g
( ) =
∂h CV
E
¯¯¯¯
¯
where h is the height above the earth’s surface, M is the molar mass of dinitrogen (0.0280 kg mol ), g is the acceleration due to −1
gravity (9.80 m s ), and C is the constant-volume heat capacity (20.8 J K mol ). [Suggestion: Write the total differential
−1
V
−1 −1
¯¯
¯ ¯¯
¯
for E = E (S , V
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
, h) . What are (∂ E /∂ S )
¯
¯¯¯
¯
¯¯¯
, (∂ E /∂ V
¯
¯¯¯ ¯
¯¯¯
)
¯¯
¯¯
, and (∂ E /∂h)¯
¯¯¯
¯¯
¯¯ ¯
¯¯¯
?]
V ,h S ,h S ,V
If the temperature at sea level is 300 K, what is the temperature on the top of a 3000 m mountain?
22. Assume that the earth’s atmosphere is pure nitrogen and that it behaves as an ideal gas. Assume that the molar enthalpy of this
¯¯
¯
nitrogen is constant and that its molar entropy changes are adequately modeled by dS = (C /T ) dT − (R/P ) dP . For this P
atmosphere, show that

¯¯¯¯
¯
∂T −M g
( ) =
∂h CP
S
¯¯¯¯
¯
where h is the height above the earth’s surface, M is the molar mass of dinitrogen (0.0280 kg mol ), g is the acceleration due to −1
gravity (9.80 m s ), and C is the constant-pressure heat capacity (29.1 J K mol ). [Suggestion: Write the total differential
−1
P
−1 −1
¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯¯
¯ ¯
¯¯¯
¯ ¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
for H = H (S , P , h) . What are (∂ H /∂ S ) , (∂ H /∂P ) ¯
¯¯¯
, and (∂ H /∂h) ¯¯
¯¯ ¯
¯¯¯
?]
P ,h V ,h S ,V
Use this approximation to calculate the temperature on the top of a 3000 m mountain when the temperature at sea level is 300 K.
23. Hikers often say that, as a rule-of-thumb, the temperature on a mountain decreases by 1 C for every 100 m increase in
elevation. Is this rule in accord with the relationships developed in problems 21 and 22? In these problems, we assume that the
temperature of an ideal-gas atmosphere varies with altitude but that the molar energy or enthalpy does not. Does this assumption
contradict the principle that the energy and enthalpy of an ideal gas depend only on temperature?
24. Derive the barometric formula (Section 2.10) from the assumptions that the earth’s atmosphere is an ideal gas whose molar
¯
¯¯¯¯
¯
mass is M and whose temperature and Gibbs free energy are independent of altitude.
25. Run in reverse, a Carnot engine consumes work (w > 0) and transfers heat (q > 0) from a low-temperature reservoir to a high
ℓ
temperature reservoir (q < 0) . The work done by the machine is converted to heat that is discharged to the high-temperature
h
reservoir. In one cycle of the machine, ΔE = q + q = 0 . For a refrigerator—or for a heat pump operating in air-conditioning
ℓ h
mode—we are interested in the quantity of heat removed (q ) per unit of energy expended (w). We define the coefficient of
ℓ
performance as C OP (cooling) = q /w . This is at a maximum for the reversible Carnot engine. Show that the theoretical
ℓ
maximum is
1 −ϵ Tℓ
C OP (cooling) = =
ϵ Th − Tℓ
where ϵ is the reversible Carnot-engine efficiency,

Tℓ
ϵ=1−
Th
26. For a heat pump operating in heating mode—as a “furnace”—we are interested in the quantity of heat delivered to the space
being heated (−q ) per unit of energy expended (w). We define the coefficient of performance as C OP (heating) = −q /w .
h h
Show that the theoretical maximum is

1 Th
C OP (heating) = =
ϵ Th − Tℓ
27. For T ℓ = 300 K and T h = 500 K , calculate the theoretical maxima for C OP (cooling) and C OP (heating) .
28. Find the theoretical maximum C OP (cooling) for a refrigerator at 40 F in a room at 72 F.
29. Find the theoretical maximum C OP (cooling) for a heat pump that keeps a room at 72 F when the outside temperature is
100 F.
30. Find the theoretical maximum C OP (heating) for a heat pump that keeps a room at 72 F when the outside temperature is
32 F .
Notes
1
For an introduction to the concept of internal entropy and its applications, see Ilya Prigogine, Introduction to the
Thermodynamics of Irreversible Processes, Interscience Publishers, 1961.
CHAPTER OVERVIEW
10: Some Mathematical Consequences of the Fundamental Equation

10.15: Problems
This page titled 10: Some Mathematical Consequences of the Fundamental Equation is shared under a CC BY-SA 4.0 license and was authored,
1
In Chapter 9, we substitute dq rev
= T dS , from the second law, into
dE = dq + P dV
from the first law, to obtain, for any closed system undergoing a reversible change in which the only work is pressure–volume
work, the fundamental equation, dE = T dS + P dV . In view of the mathematical properties of state functions that we develop in
Chapter 7, this result means that we can express the energy of the system as a function of entropy and volume, E = E (S, V ) . With
this choice of independent variables, the total differential of E is
(dE = (∂E/∂S) dS + (∂E/∂V ) dV .

V S
Equating these expressions for dE , we find

∂E ∂E
[( ) + T ] dS + [( ) + P ] dV = 0
∂S V
∂V S
for any such system. Since S and V are independent variables, this equation can be true for any arbitrary state of the system only if
the coefficients of dS and dV are each identically equal to zero. It follows that
∂E
( ) =T
∂S V
and
∂E
( ) = −P
∂V S
Moreover, because dE is an exact differential, we have

∂ ∂E ∂ ∂E
( ) = ( )
∂V ∂S ∂S ∂V
V S
so that
∂T ∂P
( ) = −( )
∂V ∂S
S V
Using the result dH = T dS + V dP , parallel arguments show that enthalpy can be expressed as a function of entropy and
pressure, H = H (S, P ) , so that
∂H
( ) =T
∂S
P
and
∂H
( ) =V
∂P
S
and
∂T ∂V
( ) =( )
∂P ∂S
S P
Since dA= −SdT + P dV , the Helmholtz free energy must be a function of temperature and volume, A = A (T , V ) , and we have
∂A
( ) = −S
∂T
V
and
∂A
( ) = −P
∂V
T
and
∂S ∂P
( ) =( )
∂V ∂T
T V
Likewise, dG= −SdT + V dP implies that the Gibbs free energy is a function of temperature and pressure, G = G (P , T ) , so that
∂G
( ) = −S
∂T
P
and
∂G
( ) =V
∂P
T
and
∂V ∂S
( ) = −( )
∂T ∂P
P T
This page titled 10.1: Thermodynamic Relationships from dE, dH, dA and dG is shared under a CC BY-SA 4.0 license and was authored,
In Chapter 1, we observe that the business of science is the creation of models that are internally consistent and that accurately
describe reality. Logical deduction from tentative hypotheses is a valuable tool in our effort to create new models. Such logical
arguments often take the form of “Gedanken (thought) experiments,” as exemplified by our various arguments about the properties
of hypothetical, friction-free, piston engines. Nevertheless, the route by which we arrive at a theory is irrelevant; what counts are
the theory’s internal consistency and predictive capability. Let us pause, therefore, to note that we have arrived at mathematical
expressions of ideas that we initially introduced as principles inferred from experience.
In Chapter 6, we prove Duhem’s theorem when the variables are chosen from the set pressure, temperature, volume, and
component concentrations. However, the theorem is more general. It asserts that two variables are sufficient to specify changes in
the state of a closed, reversible system, in which only pressure–volume work is possible. Our derivations have now led us to the
conclusion that the energy of such a system can be expressed as a function of entropy and volume. Given the entropy, the volume,
and the function E = E (S, V ) , the relationships developed above mean that we know S , V , E , P , and T for the system. Given
these, we can calculate H , A , and G. That is, specifying the changes in the two variables S and V is sufficient to specify the
change in the state of the system. Moreover, we can rearrange the fundamental equation to
T 1
dV = ( ) dS + ( ) dE
P P
so that the volume can be expressed as a function of entropy and energy. Given S , E , and the function V = V (S, E) , we can find
P and T . Specifying the changes in S and E is sufficient to specify the change in the system. Finally, we can rearrange the
fundamental equation to
1 P
dS = ( ) dE + ( ) dV
T T
so that S = S (V , E) and specifying changes in E and V is sufficient to specify the change in the system.
Now, let us return to our discussion in Section 9.7 of the entropy change for an isolated system undergoing a spontaneous change.
That discussion explores the use of the machine-based statement of the second law to establish that the entropy of an isolated
system must increase during any spontaneous process. To infer that the system’s entropy must increase in such a process, we
consider the special case in which only pressure–volume work is possible and argue that a change in which ΔE = ΔS= 0 is no
change at all. That is, we assume that specifying the change in E and S is sufficient to specify the change in the state of such a
system. It is, therefore, a significant check on the internal consistency of our thermodynamic model to see that dE = T dS + P dV
implies that E and S are indeed a sufficient pair.
Finally, let us consider the relationship of a spontaneous process in a closed system to the surface that describes reversible
processes in the same system. The energy of the system undergoing reversible change is expressed as E = E (S, V ) . An energy
surface in V − S − E -space is sketched in Figure 1. At any point on this surface, the system is at equilibrium. The point
(V , S , E ) is such a point. The tangent to this surface at (V , S , E ) and in the plane V = V
0 0 0 0 0 0 0 is the partial derivative
T = (∂E/∂S)
V
. The tangent to the surface at (V , S , E ) and in the plane S = S is the partial derivative P = −(∂E/∂V ) .
0 0 0 0 S
Figure 1. An energy surface in volume-entropy-energy-space
No point in V − S − E -space that is off the E = E (S, V ) surface can describe an equilibrium state of the closed system. As a
practical matter, some of these points may represent states that the system can attain. If so, they are transient states of a
spontaneously changing system. Let us suppose, for example, that we are able to
maintain S = S and V = V while a slow chemical reaction occurs in the system. At every instant, such a state must have a non-
0 0
equilibrium composition. It must have an energy, and this energy must exceed E ; since we must have (ΔE) < 0 for a
0 SV
spontaneous process, a point (V , S , E ) can represent a state of the system during a spontaneous change only if E >E . If
0 0 1 1 0
E <E , the point (V , S , E ) cannot represent a state of the system that can spontaneously go to equilibrium at (V , S , E ).
2 0 0 0 2 0 0 0
Similarly, if S <S and the system is isolated with E = E and V = V , the point (V , S , E ) represents a state of the system
1 0 0 0 0 1 0
that can go to equilibrium at (V , S , E ) spontaneously. This process would satisfy the entropy criterion, (ΔS) > 0 .
0 0 0 EV
This page titled 10.2: dE = TdS - PdV and Internal consistency is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
We have found simple differential expressions for E , S , H , A, and G that apply to closed, reversible systems in which only
pressure–volume work is possible. From dE = T dS + P dV , we infer that E = E (S, V ) . From dH = T dS + V dP , we infer
that H = H (S, P ) . An argument parallel to that above leads us to the conclusion that specifying the changes in S and P is
sufficient to specify the change in the state of the system. Similarly, from dA= −SdT + P dV and dG= −SdT + V dP , we see
that it is sufficient to specify the changes in either V and T or P and T . These total differentials show that that specifying the
change in two state functions is sufficient to specify the change that occurs in the state of a closed system, when the change is
reversible and all of the work is pressure–volume work. We have now found seven pairs of state functions that are sufficient; they
are {S, V }, {S, E}, {E, V }, and the four pairs in which we choose one variable from the set {S, T } and one from the set
{P , V }. However, each of the equations we have obtained so far uses a different pair of independent variables.
Evidently, we should be able to express any thermodynamic function using various pairs of state functions. We can do this by
transforming the equations that we have already derived. We are particularly interested in P , V , and T as independent variables,
because these quantities are readily measured for most systems. In the sections below, we find exact differentials for dS , dE , dH ,
dA , and dG with V and T and with P and T as the independent variables.
While specifying the change in some pair of variables is always sufficient to specify the change in the state of a closed reversible
system, we should note that it is not always necessary. If the system has only one degree of freedom, specifying some single
variable is sufficient. For example, so long as both phases remain present, the change in the state of a pure substance at liquid–
vapor equilibrium can be specified by specifying the change in the temperature, the pressure, the volume, or the number of moles
of either phase. We discuss this further in Section 10.7.
At present, we are developing relationships among state functions that are valid for any closed reversible system in which all work
is pressure–volume work. The next several chapters explore the implications of these results. If the composition of the closed
reversible system changes during these processes, this composition change does not affect the relationships we develop here. Of
course, any composition change that occurs during a reversible process must be reversible; if the components of the system can
react, this reaction must be at equilibrium throughout the process. In Chapter 14, we extend the relationships that we develop here
to explicitly include molar compositions as independent variables. This enables us to express our theory for equilibrium using
composition variables.
In Section 6.10, we assume—infer from experience—that specifying the changes in P and T is sufficient to specify a change in the
state of a closed equilibrium system whose phase composition is fixed and in which only pressure–volume work is possible. We
use this assumption to give a partial proof of Duhem’s theorem. In Section 10.5, we see that this assumption is also a consequence
of the theory we have developed. This is another check on the internal consistency of the theory.
Finally, it is time to consider a question we have thus far avoided: Is any arbitrary pair of state functions a sufficient set? The
answer is no. In Section 10.8, we find that neither {P , V } nor {S, T } is a sufficient pair in all cases.
This page titled 10.3: Expressing Thermodynamic Functions with Other Independent Variables is shared under a CC BY-SA 4.0 license and was
If we choose V and T as the independent variables, we can express the differential of E as a function of V and T . We also have
the differential relationship dE = T dS − P dV . These expressions for dE must be equal:
∂E ∂E
dE = ( ) dV + ( ) dT = T dS + P dV
∂V ∂T
T V
Rearranging, we find a total differential for dS with V and T as the independent variables:
1 ∂E 1 ∂E
dS = ( ) dT + [( ) + P ] dV
T ∂T T ∂V
V T
From the coefficient of dT , we have

∂S 1 ∂E CV
( ) = ( ) =
∂T T ∂T T
V V
where we use the definition (∂E/∂T ) = C . (When we write “C ,” we usually think of it as a property of a pure substance. The
V V V
relationship above is valid for any reversible system. When we are describing a system that is not a pure substance, C is just an V
abbreviation for (∂E/∂T ) .) From the coefficient of dV , we have

V
∂S 1 ∂E ∂P
( ) = [( ) +P] = ( )
∂V T ∂V ∂T
T T V
where we use the relationship (∂S/∂V ) T

= (∂P /∂T )
V
that we find in §1. Substituting into the expression for dS , we find
CV ∂P
dS = dT + ( ) dV
T ∂T
V
Now, from dE = T dS − P dV , we have

∂P
dE = CV dT + [T ( ) + P ] dV
∂T
V
From H = E + P V , we have
dH = dE + d (P V )
∂ (P V ) ∂ (P V )
= dE + ( ) dT + ( ) dV
∂T ∂V
V T
∂P ∂P
= dE + V ( ) dT + [P + V ( ) ] dV
∂T ∂V
V T
∂P ∂P ∂P
= [CV + V ( ) ] dT + [T ( ) +V ( ) ] dV
∂T ∂T ∂V
V V T
Of course, we already have
dA= −SdT + P dV
From G = H − T S , by an argument that parallels the above derivation of dH , we obtain

∂P ∂P
dG = [V ( ) + S] dT + V ( ) dV
∂T ∂V
V T
Finally, we can write P = P (T , V ) to find

∂P ∂P
dP = ( ) dT + ( ) dV
∂T ∂V
V T
P , T , V , C , (∂P /∂T ) , and (∂P /∂V ) are all experimentally accessible for any reversible system. If we have this information
V V T
for a system that undergoes a change from a state specified by T and V to a second state specified by T and V , we can use these
1 1 2 2
relationships to calculate ΔE, ΔS , and ΔH . To do so, we calculate the appropriate line integral along a reversible path. One such
path is an isothermal reversible change, at T , from V to V , followed by a constant-volume change, at V , from T to T . In
1 1 2 2 1 2
principle, the same procedure can used to calculate ΔA and ΔG. However, because S appears in the differentials dA and dG, this
requires that we first find S as a function of V and T .
If the system is a pure substance for which we have an equation of state, we can find (∂P /∂T ) , and (∂P /∂V ) by V T
straightforward differentiation. When the substance is a gas, an equation of state may be available in the literature. When the
substance is a liquid or a solid, these partial derivatives can still be related to experimentally accessible quantities. The
compressibility of a substance is the change in its volume that results from a change in the applied pressure, at a constant
temperature. The thermal expansion of a substance is the change in its volume that results from a change in its temperature, at a
constant applied pressure. It is convenient to convert measurements of these properties into intensive functions of the state of the
substance by expressing the volume change as a fraction of the original volume. That is, we define the coefficient of thermal
expansion:
1 ∂V
α = ( )
V ∂T P
and the coefficient of isothermal compressibility:

1 ∂V
β= − ( )
V ∂P
T
Coefficients of thermal expansion and isothermal compressibility are available in compilations of thermodynamic data for many
liquids and solids. In general, both coefficients are weak functions of temperature. We have
∂P 1
( ) =−
∂V βV
T
and
∂P ∂V ∂V α
( ) = −( ) /( ) =
∂T ∂T ∂P β
V P T
Using these coefficients, we can estimate a pressure change, for example, as a line integral of
α 1
dP = ( ) dT − ( ) dV
β βV
This page titled 10.4: Expressing Thermodynamic Functions with Independent Variables V and T is shared under a CC BY-SA 4.0 license and
We can follow a parallel development to express these thermodynamic functions with P and T as the independent variables. We
have the differential relationship dH = T dS + V dP . We expand dH with P and T as the independent variables. Equating these,
we obtain
∂H ∂H
dH = ( ) dP + ( ) dT (10.5.1)
∂P ∂T
T P
= T dS + V dP (10.5.2)
so that we have
1 ∂H 1 ∂H
dS = ( ) dT + [( ) − V ] dP
T ∂T T ∂P
P T
From the coefficient of dT and the definition (∂H /∂T ) P

= CP , we have
∂S 1 ∂H CP
( ) = ( ) =
∂T T ∂T T
P P
(When we are describing a reversible system that is not a pure substance, C is just an abbreviation for P (∂H /∂T )
P
.) From the
coefficient of dP and the relationship (∂S/∂P ) = −(∂V /∂T ) that we find in Section 10.1, we have
T P
∂S 1 ∂H ∂V
( ) = [( ) − V ] = −( )
∂P T ∂P ∂T
T T P
Substituting into the expression for dS , we find

CP ∂V
dS = dT − ( ) dP
T ∂T
P
Using the same approach as in the previous section, we can now obtain
∂V ∂V ∂V
dE = [CP − P ( ) ] dT − [P ( ) +T( ) ] dP (10.5.3)
∂T ∂P ∂T
P T P
∂V
dH = CP dT + [V − T ( ) ] dP (10.5.4)
∂T
P
∂V ∂V
dA = − [S + P ( ) ] dT − P ( ) dP (10.5.5)
∂T ∂P
P T
and, we already have
dG = V dP − SdT
Finally, we can write V = V (P , T ) to find

∂V ∂V
dV = ( ) dP + ( ) dT
∂P ∂T
T P
so that we have total differentials for all of the principal thermodynamic functions when they are expressed as functions of P and
T . If an equation of state is not known but the coefficients of thermal expansion and isothermal compressibility are available, we
have (∂V /∂T ) = αV and (∂V /∂P ) = −βV . Then we can estimate a volume change, for example, as a line integral of
P T
dV = αV dT − βV dP
This page titled 10.5: Expressing Thermodynamic Functions with Independent Variables P and T is shared under a CC BY-SA 4.0 license and was
Let us suppose that M , Q, R , X, and Y are state functions and that we know the total differentials
∂M ∂M
dM = ( ) dX + ( ) dY
∂X ∂Y
Y X
∂Q ∂Q
dQ = ( ) dX + ( ) dY
∂X Y
∂Y X
∂R ∂R
dR = ( ) dX + ( ) dY
∂X ∂Y
Y X
To find the total differential of M (Q, R),

∂M ∂M
dM = ( ) dQ + ( ) dR
∂Q ∂R
R Q
we solve the total differentials of Q (X, Y ) and R (X, Y ) to find dX and dY in terms of dQ and dR . Since dQ and dR are
simultaneous equations in the variables dX and dY , we can apply Cramer’s rule to obtain
∣ dQ (∂Q/∂Y ) ∣ ∂R ∂Q
X
∣ ∣ ( ) dQ − ( ) dR
∣ dR (∂R/∂Y )X ∣ ∂Y ∂Y
X X
dX = =
Q, R Q, R
J( ) J( )
X, Y X, Y
and
∣ (∂Q/∂X) dQ ∣ ∂R ∂Q
Y
∣ ∣ −( ) dQ + ( ) dR
∣ (∂R/∂X)Y dR ∣ ∂X
Y
∂X
Y
dY = =
Q, R Q, R
J( ) J( )
X, Y X, Y
where J ((Q, R)/(X, Y )) is the Jacobian of the transformation of variables X and Y to variables Q and R :
Q, R ∣ (∂Q/∂X) (∂Q/∂Y ) ∣
Y X
J( ) =∣ ∣
X, Y ∣ (∂R/∂X)Y (∂R/∂Y )X ∣
∂Q ∂R ∂Q ∂R
=( ) ( ) +( ) ( )
∂X ∂Y ∂Y ∂X
Y X X Y
To find
∂M ∂M
dM = ( ) dQ + ( ) dR
∂Q ∂R
R Q
We substitute these results for dX and dY into the total differential of M = M (X, Y ) :
∂M ∂M
dM = ( ) dX + ( ) dY
∂X ∂Y
Y X
∂M ∂R ∂Q ∂M ∂R ∂Q
( ) [( ) dQ − ( ) dR] ( ) [− ( ) dQ + ( ) dR]
∂X ∂Y ∂Y ∂Y ∂X ∂X
Y X X X Y Y
= +
Q, R Q, R
J( ) J( )
X, Y X, Y
∂M ∂R ∂M ∂R ∂M ∂Q ∂M ∂Q
⎡ ( ) ( ) −( ) ( ) ⎤ ⎡ −( ) ( ) +( ) ( ) ⎤
⎢ ∂X ∂Y ∂Y ∂X ⎥ ⎢ ∂X ∂Y ∂Y ∂X ⎥
Y X X Y Y X X Y
=⎢ ⎥ dQ + ⎢ ⎥ dR
⎢ ⎥ ⎢ ⎥
⎢ Q, R ⎥ ⎢ Q, R ⎥
J( ) J( )
⎣ ⎦ ⎣ ⎦
X, Y X, Y
where the coefficients of dQ and dR are (∂M /∂Q) and (∂M /∂R) , respectively. In §5, we find the other total differentials in
R Q
terms of dP and dT . If we set X = T and Y = P , we can use these relationships to find the total differential for any state function
expressed in terms of any two other state functions.
To illustrate this point, let us use these relationships to find the total differential of S expressed as a function of P and V ,
S = S (P , V ) . In this case, we are transforming from the variables (P , T ) to the variables (P , V ). This is a one-variable
transformation. To effect it, we make the additional substitutions M = S , Q = V , and R = P . Since we have Y = R = P , the
transformation equations simplify substantially. We have
∂R ∂P
( ) =( ) =1
∂Y X
∂P X
∂R ∂P
( ) =( ) =0
∂X ∂T
Y P
∂Q ∂V
( ) =( )
∂Y ∂P
X T
∂Q ∂V
( ) =( )
∂X ∂T
Y P
The Jacobian becomes

Q, R ∂V
J( ) =( )
X, Y ∂T
P
and the partial derivatives of S become

∂S ∂S ∂V CP ∂T
( ) =( ) /( ) = ( )
∂V ∂T ∂T T ∂V
P P P P
and
∂S ∂V ∂S ∂V
−( ) ( ) +( ) ( )
∂S ∂T ∂P ∂P ∂T
P T T P
( ) = (10.6.1)
∂P ∂V
V
( )
∂T
P
CP ∂T ∂S
= ( ) +( ) (10.6.2)
T ∂P ∂P
V T
This page titled 10.6: The Transformation of Thermodynamic Variables in General is shared under a CC BY-SA 4.0 license and was authored,
We have found total differentials for the principal thermodynamic functions with V and T and with P and T as the independent
variables. In §6, we see how to find such differentials, dM = M dX + M dY , for any pair of independent variables, X and Y .
x Y
These equations express our physical theory as a mathematical structure. Because E , T , S , P , V , H , A , and G are state functions,
the mathematical properties of state functions enable us to obtain the relationships M = (∂M /∂X) , M = (∂M /∂Y ) , and X Y Y X
(∂ M /∂X) = (∂ M /∂Y ) that we find in §1. These equations apply to reversible processes in closed systems when only
Y X
pressure–volume work is possible. Using these equations to describe reversible processes involves a number of important ideas. Let
us consider four cases: [GrindEQ__1_] None of M , X, and Y are constant; [GrindEQ__2_] M is constant, but X and Y are not;
[GrindEQ__3_] X is constant, but M and Y are not; [GrindEQ__4_] X and Y are constant, but M is not.
None of M , , and
is constant.– The properties of state functions and the existence of the exact differential ,
X Y
dM = MX dX + MY dY imply that M is a function of X and Y , M = M (X, Y ) . If M and M are single-valued and X Y
continuous along some path in the XY -plane, we can evaluate the line integral of dM along this path. Given M at a first point,
M (X , Y ) , we can find M at a second point, M (X , Y ) , by evaluating this line integral along some path in the XY -plane
1 1 2 2
between the points (X , Y ) and (X , Y ) . Given the values of X and Y , the value of M is uniquely determined; M (X, Y ) is a
1 1 2 2
two-dimensional surface in the three-dimensional space whose dimensions are M , X, and Y .

The M (X, Y ) surface is just the set of points that are accessible to the system in reversible processes in which X and Y change.
Conversely, the only values of M that are accessible to the system in reversible processes in which X and Y change correspond to
those points that lie on the surface. In §5, we find the total differentials for dE , dS , dH , dA , and dG using P and T as the
independent variables. Evidently, for a reversible process in a closed system, there is a surface representing each of E , S , H , A ,
and G over the P − −T -plane. Since the system is also characterized by an equation of state that relates the values of P , V, and T ,
there is also a surface representing V over the P − −T -plane.
In general, a given system can also undergo spontaneous changes. Suppose that a system is originally at equilibrium at temperature
T and pressure P . If we contact this system with surroundings at some arbitrary T
1 1
^ and we arrange for the pressure applied to the
system to have some arbitrary value, P , the system will respond, eventually reaching equilibrium with the system
applied
temperature equal to T and the system pressure equal to P

^
. (The change-enabled state in which P
applied =P and system 1
Tsystem = T , while the applied pressure is P and the surroundings temperature is T^, is a hypothetical state. It is not an
applied
1
equilibrium state, because P ≠P and T ≠T . The change-enabled state can undergo spontaneous change; however,
^
system applied system
its thermodynamic functions have the same values as they have in the original equilibrium state, in which P =P and system 1
Tsystem = T .) Since this change is spontaneous, it may not be possible to trace the path of the system in the P − −T -plane as the
1
change occurs. If we can trace the path in the P − −T -plane, the energy of the system can be described as a line in
E − −P − −T -space, but this line will not lie on the reversible-process surface specified by the function E = E (P , T ) .
Nevertheless, we can select paths in the P − −T -plane that connect the initial point (P1 , T1 ) to the final point ^
(Papplied , T ) .
There are reversible processes that correspond to these paths. By evaluating the line integral for any state function along any of
these paths, we can find the change that occurs in the state function during the spontaneous process.
Cases arise in which M is not single-valued or continuous along some or all of the paths that connect points (X , Y ) and 1 1
(X , Y ) . Then M
2 2 or M may not exist for some points (X, Y ). In this case, it may not be possible to evaluate the line integral
X Y
to find the change M (X , Y ) + M (X , Y ) . This can occur when there is a phase change. If (P , T ) specifies a state of
2 2 1 1 vp bp
liquid–vapor equilibrium, the enthalpy of the system is not single-valued. Below, we consider the thermodynamic surfaces of water
when a phase change occurs. In §8, we see that M may not be single-valued when V and P or when T and S are the independent
variables.
M is constant, X and Y are not.– If M is constant, we have dM = 0 . If X and Y are not constant, if M and M are defined, X Y
and if M ≠0, we can apply the divide-through rule to obtain

Y
∂Y MX
( ) =−
∂X MY
M
Such relationships are useful. In Chapter 12, we discuss the Clapeyron and Clausius-Clapeyron equations, which we obtain from
dG= −SdT + V dP using this argument.
X is constant, M and Y are not.– If X is constant, we have dX= 0 . Instead of , dM = M dX + M dY , we have , X Y
dM = MX dX , and we must ask whether M is indeed expressible as a function of Y only. If M can be expressed as a function of
Y only, so that M is single-valued and continuous, we can integrate to find
Y
Y2
M2 + M1 = ∫ MY dY
Y1
X and Y are constant, M is not constant.– An interesting and important case arises when X and Y are constant, but M is not
constant. When X and Y are constant, dX = dY = 0 , and from dM = M dX + M dY , it follows that dM = 0 . Nevertheless,
X Y
we can readily identify processes in which some state function, M , changes while two (or more) others, X and Y , remain constant.
In this case, it is clear that X and Y are not sufficient to model the change in M . Recall from our discussion of Duhem’s Theorem
that, while two independent variables are sufficient to describe a reversible process in which pressure-volume work is the only
work, which pair of variables is adequate depends on the system.
Below we discuss the reversible vaporization of water at constant P and T . For this process we have dG = 0 . However, we know
that dS > 0 . For independent variables P and T , our differential expressions for dG and dS are
dG = V dP − SdT
and
1 ∂H ∂V
dS = ( ) dT − ( ) dP
T ∂T ∂T
P T
Setting dP = dT = 0 in these equations correctly gives dG = 0 ; however, dS = 0 is false. Evidently, variables P and T are not
sufficient to model the entropy change in this process.
However, at constant P and T , variables V and T are adequate. We have
1 ∂E ∂P
dS = ( ) dT + ( ) dV
T ∂T ∂T
V V
In Section 12.10 we develop the Clausius-Clapeyron equation for this vaporization process; we find
¯
¯¯¯
¯
∂P P Δvap H
( ) =
2
∂T RT
V
¯
¯¯¯
¯
(Δ H is the enthalpy of vaporization per mole.) Since the volume of the system is essentially the volume of the gas phase, we
vap
have, assuming the vapor behaves as an ideal gas, V = n RT /P , and

g
RT
dV = dng
P
The entropy equation then becomes

∂P
dS = ( ) dV
∂T
V
¯
¯¯¯
¯
P Δvap H RT
= ⋅ dng
2
RT P
¯
¯¯¯
¯
Δvap H
= dng
T
so that the entropy change for this reversible process is directly proportional to the number of moles of vapor produced.
We see that we must introduce an extensive variable to model the entropy change in the vaporization process. The system volume
serves this purpose, although we wind up expressing this volume in terms of the number of moles of vapor in the system.
From another perspective, we can write the entropy as a function of P , T , and n : S = S (P , T , n ) . Then g g
1 ∂H ∂V ∂S
dS = ( ) dT − ( ) dP + ( ) dng
T ∂T ∂T ∂ng
P T PT
which becomes
∂S
(dS) =( ) dng
PT
∂ng
PT
with
¯
¯¯¯
¯
∂S Δvap H
( ) =
∂ng T
PT
In Chapter 14, we extend all of our thermodynamic models to include variables that specify the system composition—the number
of moles of the substances present in the system.
Thermodynamic surfaces in the reversible vaporization of water

To illustrate the fact that dT = dP = 0 has different implications for dS than it does for dG, let us consider the reversible
vaporization of one mole of water at constant P and T . ΔG for this process is zero, but ΔS and ΔH are not. We can describe a
system comprised of one mole of liquid water using pressure and temperature as the independent variables. There is a wide range
of pressure and temperature values that is consistent with the system remaining entirely liquid. Every combination of pressure and
temperature at which the system remains entirely liquid can be reached by a reversible process from any other such combination of
pressure and temperature. For every combination of pressure and temperature within this range, there is one and only one value for
every other thermodynamic function. Choosing the enthalpy function to be specific, we can say that the set of enthalpy–pressure–
temperature-points for which the system remains entirely liquid is a thermodynamic surface on which reversible change is possible.
We can say all of these same things about a system that consists of one mole of gaseous water. Of course, the enthalpy surface for
gaseous water is a different surface from the enthalpy surface for liquid water. At any given temperature, there is a pressure at
which liquid and gaseous water are at equilibrium. Above this pressure, the system is entirely liquid; below it, the system is entirely
gas. The enthalpy surface of the liquid lies over a different part of the pressure–temperature-plane than does that of the gas. (If the
liquid can be superheated or the gas can be supercooled, a given pressure and temperature may be represented by a point on the
enthalpy surfaces for both the pure liquid and the pure gas.) The enthalpy surface for the gaseous system lies at higher energies
than that for the liquid system; the two enthalpy surfaces do not intersect.
To reversibly transform pure liquid water to pure gaseous water, we must move on the enthalpy surface of the liquid to a pressure
and temperature at which the liquid and the gas are at equilibrium. At this pressure and temperature, we can reversibly increase the
volume of the system, causing the reversible vaporization of liquid water, and we can continue this process until all of the liquid
has been vaporized. When all of the liquid has been vaporized, the system is on the enthalpy surface of the gas. Thereafter, we can
change the pressure and temperature of the system to reversibly change the state of the pure gas. While we can describe this
process in terms of the successive changes that we impose on the state functions of a system that consists of one mole of water, we
are considering three different systems when we describe the overall process from the perspective afforded by Gibbs’ phase rule.
1. The first system is one mole of pure liquid. This system has one phase. There are two degrees of freedom, which we take to be
pressure and temperature.
2. The second system is one mole of water, of which x mole is liquid and 1−x mole is gas, at equilibrium, at a fixed pressure
ℓ ℓ
and temperature. In the vaporization process, the pressure and temperature are constant while the volume of the system
increases (x decreases) reversibly. The one mole of water is described by this system from the time the first bubble of gas
ℓ
appears to the time the last drop of liquid vaporizes. There are two phases and one degree of freedom. When we reversibly
vaporize water at a fixed pressure and temperature, one variable must describe the composition: We can take it to be the volume
of the system or the liquid mole fraction, x . (Of course, we can reversibly vaporize water in a process in which the pressure,
ℓ
temperature, and composition all change; however, because there is only one degree of freedom, specifying a temperature
change uniquely determines the pressure change, and conversely. In H − −P − −T -space, a reversibly vaporizing system
traces a path on a vertical plane between the enthalpy surfaces of the liquid and the gas. If pressure and temperature are
constant, this path is a vertical line. If reversibility is achieved through synchronous variation of pressure and temperature, the
path is not vertical, but it remains in a vertical plane.)
3. The third system describes the mole of water after all of the water has been converted to the gas. This system has one phase and
two degrees of freedom, which we again take to be pressure and temperature.
We can say that this description of the reversible conversion of liquid water to gaseous water involves three Gibbsian H − P − T -
manifolds. Two of these are the enthalpy surfaces for the gas and the liquid. The third is a line of enthalpies at constant pressure
and temperature; successive points on this line represent different mole fractions of liquid water.
We can track the reversible conversion of one mole of pure liquid water to pure gas on other thermodynamic surfaces. For example,
if we consider enthalpy as a function of volume and temperature, the entire process can be traced on a single H (V , T ) surface; that
is, every volume–temperature point, (V , T ), specifies a unique state of the system, and conversely.
When we use pressure and temperature as the independent variables, the Gibbs free energy provides the criterion for reversibility.
Unlike the corresponding enthalpy surfaces, which never meet, the Gibbs free energy surfaces for the pure liquid and the pure gas
intersect along a line of pressure and temperature values. At an equilibrium pressure and temperature, the Gibbs free energy change
for the reversible vaporization of water is zero, which means that the Gibbs free energy for a mole of liquid water is the same as the
Gibbs free energy for a mole of gaseous water at that pressure and temperature.
When we trace the reversible conversion of a mole of liquid water to a mole of gaseous water on the Gibbs free energy surfaces,
the point representing the state of the mole of water moves on the Gibbs free energy surface of the liquid from the initial pressure
and temperature to the pressure–temperature equilibrium line. The pressure–temperature equilibrium line is formed by the
intersection of the Gibbs free energy surface of the liquid with the Gibbs free energy surface of the gas. (The projection of this line
of intersection onto the P − −T -plane is a line of points in the P − −T -plane that satisfies the differential relationship
0 = −Δ vapSdT + Δ V dP . The point P = 1 atm , T = 373.15 K, lies on this line.)
vap
The conversion of liquid to gas can occur while the mole of water remains at the same point in pressure–temperature–Gibbs free
energy space, and the mole fractions, x and 1−x , vary continuously over the range 0 <x < 1 . During this reversible
ℓ ℓ ℓ
vaporization process, dG = dT = dP = 0 , while Δ V > 0 , Δ H > 0, and Δ S> 0 . (We find Δ H and Δ S by
vap vap vap vap vap
measuring the heat required to vaporize a mole of water at P and T . Then q = Δ H , and q /T = Δ S . Since the process is
P vap P vap
reversible, we have Δ S= −Δ S^ , and Δ S + Δ S^= 0 .) When the conversion of liquid to gas is complete, reversible
vap vap vap vap
changes to the one mole of gaseous water correspond to motion of a point on the Gibbs free energy surface of the gas.
This page titled 10.7: Reversibility and Thermodynamic Variables in Gerneral is shared under a CC BY-SA 4.0 license and was authored,
When only pressure–volume work is possible, various pairs of state functions can specify the state of any closed equilibrium
system. A given pair may be sufficient to specify the state on a given Gibbsian manifold but not to specify the state of the same
system on a different Gibbsian manifold. (In Section 10.7, we see that the set {T , P } is sufficient to specify the state of a closed
system containing only liquid water or only gaseous water. However, specifying T and P is not sufficient to establish the amount of
water vapor in a closed system in which the liquid is vaporizing reversibly.)
It can also happen that a given pair of state functions specifies the state of a closed system over part but not all of a given Gibbsian
manifold. Specifying the values of such a pair is not sufficient to describe the entire manifold. In particular, we can show that
neither the set {P , V } nor the set {T , S} is always a sufficient pair in this sense.
When we say that specifying P and V is sufficient to specify the state of a system on a particular Gibbsian manifold, we mean that
any state function, M , must be uniquely specified when P and V are specified; a single-valued function, M (P , V ), must exist.
Conversely, if for any choice of M in any system, M (P , V ) is not single-valued, P and V are not always a sufficient set. In §6,
we see how to find the total differential,
∂M ∂M
dM = ( ) dP + ( ) dV
∂P V
∂V P
It might seem that this is sufficient to ensure that specifying P and V always enables us to find M (P , V ) relative to its value in an
initial reference state M (P , V ) . To do so, we need only evaluate dM as a line integral along some reversible path that leads from
1 1
(P , V ) to (P , V ). However, we can evaluate this line integral only if both of the partial derivatives can be integrated. If one of
1 1 2 2
the partial derivatives is undefined along any path that connects (P , V ) to (P , V ), we cannot find M (P , V ) by this method.
1 1 2 2
Figure 2. Molar volume of water versus temperature.

Let us consider a closed reversible system that consists of one mole of liquid water. At ordinary pressures, the density of liquid
water is not a monotonic function of temperature. At one atmosphere, the density of liquid water reaches a maximum at 4 C.
Therefore, at a pressure of one atmosphere, the molar volume of water is a minimum at 4 C, as indicated in Figure 2. This means
¯
¯¯¯
that, at one atmosphere and a range of volumes, liquid water can be at either of two temperatures for specified values of P and V .
¯
¯¯¯
Therefore, specifying P and V does not specify T ; temperature is not a single-valued function of pressure and volume; we cannot
¯
¯¯¯
uniquely express the temperature as the required function T = T (P , V ) . Moreover, because the density has a maximum, we
have
¯
¯¯¯
∂V
( ) =0
∂T
P
at this maximum, and it follows that
∂T
( )
¯
¯¯¯
∂V P
is not defined at this temperature and pressure. In Section 10.6, we find

¯¯
¯
∂S CP ∂T
( ) = ( )
¯
¯¯¯ ¯
¯¯¯
T
∂V ∂V P
P
so that (∂ S /∂ V
¯¯
¯ ¯
¯¯¯
) is also undefined. Hence, we cannot evaluate ΔS by evaluating the line integral of dS (P , V
¯¯
¯ ¯¯
¯ ¯
¯¯¯
) along any path
P
that includes a point of maximum density. These examples show that pressure and volume are not sufficient to describe the entire
Gibbsian manifold for liquid water.
Temperature and entropy are likewise not sufficient. Since we have
¯
¯¯¯
¯¯
¯ CP ∂V
dS = dT + ( ) dP
T ∂T
P
the total differential for pressure as a function of entropy and temperature is

CP ¯¯
¯ ∂T
dP = [ dT + dS ] ( )
¯
¯¯¯
T
∂V P
so that dP is not defined at pressures and temperatures of maximum density. Consequently, we cannot express the pressure as
¯¯
¯
P = P (S , T ) over the entire liquid region.
This page titled 10.8: Using the Pair (V, P) or the Pair (T, S) as Independent Variables is shared under a CC BY-SA 4.0 license and was authored,
In Chapter 7, we derive the relationship between C and C for an ideal gas. It is useful to have a relationship between these
P V
quantities that is valid for any substance. We can derive this relationship from the equations for dS that we develop in Sections
10.4 and 10.5. If we apply the divide-through rule to dS expressed as a function of dT and dV, at constant pressure, we have
∂S CV ∂P ∂V
( ) = +( ) ( )
∂T T ∂T ∂T
P V P
From dS expressed as a function of T and P ,

CP ∂V
dS = dT + ( ) dP
T ∂T
P
we have
∂S CP
( ) =
∂T T
P
so that
CP CV ∂P ∂V
= +( ) ( )
T T ∂T ∂T
V P
and
∂P ∂V
CP + CV = T ( ) ( ) (10.9.1)
∂T ∂T
V P
For an ideal gas, the right side of Equation 10.9.1 reduces to R , in agreement with our previous result. Note also that, for any
substance, C and C become equal when the temperature goes to zero.
P V
The partial derivatives on the right hand side can be related to the coefficients of thermal expansion, α , and isothermal
compressibility, β. Using
α ∂P
=( )
β ∂T V
we can write the relationship between C and C as

P V
2
V Tα
CP + CV =
β
This page titled 10.9: The Relationship Between Cv and Cp for Any Substance is shared under a CC BY-SA 4.0 license and was authored,
The heat capacities of a substance increase with temperature. The rate of increase decreases as the temperature increases. To
achieve adequate accuracy in calculations, we often need to know how heat capacities depend on temperature. In contrast, the
dependence of heat capacities on pressure and volume is usually negligible; that is, the dependence of C on V and the V
dependence of C on P can usually be ignored. Nevertheless, we need to know how to find them.
P
An exact equation for the dependence of C on V follows readily from dS expressed as a function of dT and dV
V
CV ∂P
dS = dT + ( ) dV
T ∂T
V
Since the mixed second-partial derivatives must be equal, we have

∂ CV ∂ ∂P
[ ( )] =[ ( ) ]
∂V T ∂T ∂T
T V V
and thus
2
∂CV ∂ P
( ) = T( )
2
∂V T
∂T V
Similarly, the dependence of C on P follows from dS expressed as a function of dT and dP ,

P
CP ∂V
dS = dT + ( ) dP
T ∂T P
Equating the mixed second-partial derivatives, we have

∂ CP ∂ ∂V
[ ( )] = [+ ( ) ]
∂P T ∂T ∂T
T P P
and thus
2
∂CP ∂ V
( ) = −T ( )
2
∂P ∂T
T P
For an ideal gas, it follows that C is independent of V , and C is independent of P .

V P
When we use the coefficient of thermal expansion to describe the variation of volume with temperature, we have
∂V
( ) = αV
∂T
P
When it is adequate to approximate α as a constant, another partial differentiation with respect to temperature gives
∂CP ∂ (αV )
2
( ) = −T ( ) = −α T V
∂P ∂T
T P
Since α is normally small, this result predicts weak dependence of CP on P . If α and β are both adequately approximated as
constants, we have from
∂P α
( ) =
∂T β
V
that
∂CV ∂ (α/β)
( ) = T( ) =0
∂V ∂T
T V
This page titled 10.10: The Dependence of Cv on Volume and of Cp on Pressure is shared under a CC BY-SA 4.0 license and was authored,
When temperature and pressure are the independent variables, the Gibbs free energy is the change criterion that takes the most simple
form:
dG = −SdT + V dP .
In chemical applications, temperature and pressure are often the most convenient choice of independent variables, making the Gibbs free
energy a particularly useful function. Constant Gibbs free energy is the criterion for equilibrium at constant pressure and temperature.
The Gibbs free energy of the system does not change when the system goes from one equilibrium state to another at the same
temperature and pressure, (dG) = 0 . An equilibrium system of ice and water is one example; we can melt a portion of the ice,
PT
changing the composition of the system, while maintaining equilibrium at constant pressure and temperature. Similarly, we may be able
to change the equilibrium composition of an equilibrium system that consists of reacting gases by changing the volume of the system
while maintaining constant pressure and temperature.
Consider a system that undergoes some arbitrary change from a state A, in which its Gibbs free energy is G , to a second state B, in
A
which its Gibbs free energy is G . In general, G ≠G ; in the most general case, the pressures and temperatures of states A and B are
B A B
different. (For example, state A might be a mole of ice at −10 C and 0.5 bar, while state B is a mole of water at +10 C and 2.0 bar.
Either of these states can be converted to the other; however, they are not at equilibrium with one another, and their Gibbs free energies
are not equal.) Representing the pressures and temperatures as P , T , P , T , we can express the Gibbs free energies of these two state
1 1 2 2
as G = G (P , T ) and G = G (P , T ) , respectively. The difference is the change in the Gibbs free energy when the system
A A 1 1 B B 2 2
passes from state A to state B:
ΔAB G = GB (P2 , T2 ) − GA (P1 , T1 )
Often, we are interested in Gibbs free energy differences between states that are at the same pressure and temperature, say P1 and T1 .
Then the Gibbs free energy difference is
ΔAB G (P1 , T2 ) = GB (P1 , T2 ) − GA (P1 , T1 )
(For example, state A might be a mole of ice at −10 C and 0.5 bar, while state B is a mole of water, also at −10 C and 0.5 bar. This
would be a super-cooled state of liquid water. These states are not at equilibrium with one another, and their Gibbs free energies are not
equal. The difference between the Gibbs free energies of these states is the change in the Gibbs free energy when ice goes to super-
cooled water at −10 C and 0.5 bar.)
Similar considerations apply to expressing differences between the enthalpies and the entropies of two states that are available to a
system. The Gibbs free energy is defined by
G = H − T S.
When we are interested in a process that converts some state A to a second state B at constant pressure and temperature, we usually write
ΔG = ΔH − T ΔS
which relies on the context for the information about the pressure and temperature and the initial and final states. To explicitly denote that
the change is one that occurs at a constant temperature, T , we can write
0
ΔG (T0 ) = ΔH (T0 ) − T0 ΔS (T0 ) .
Frequently we are interested in the way that ΔG, ΔH , and ΔS vary with temperature at constant pressure. If we know how G, H , and S
vary with temperature for each of the two states of interest, we can find the temperature dependence of ΔG, ΔH , and ΔS . The Gibbs-
Helmholtz equation is a frequently useful expression of the temperature dependence of G or ΔG. Since it is a mathematical consequence
of our thermodynamic relationships, we derive it here.
At constant pressure, the temperature derivative of the Gibbs free energy is +S ; that is,
∂G
( ) =−S
∂T
P
Using this result and the definition, G = H − T S , we obtain the temperature dependence of G as
∂G G−H
( ) =
∂T T
P
However, the Gibbs-Helmholtz equation can be expressed most compactly as the temperature derivative of G/T . As a matter of calculus,
we have
∂ (G/T ) G 1 ∂G
( ) =− + ( )
2
∂T T T ∂T
P P
Using the relationships above, this becomes

∂ (G/T ) H +TS S H
( ) =− + =−
2 2
∂T T T T
P
Since Δ AB G = GB + GA , we have
(∂ ΔAB G/∂T ) = − (SB + SA ) = −ΔAB S

P
and
∂ (ΔAB G/T ) ∂ (GB /T ) ∂ (GA /T )
( ) =( ) +( )
∂T ∂T ∂T
P P P
HB HA
= −( + ) (The Gibbs-Helmholtz equation)
2 2
T T
ΔAB H
=−
2
T
If we know the temperature dependence of ΔS or ΔH , we can find the temperature dependence of ΔG by integrating the relationships
above. That is, given ΔG at T , we can find ΔG at T . Thus, from (∂G/∂T ) = −S , we have
1 2 P
ΔG( T2 ) T2
∂ΔG
∫ ( ) dT = ΔG (T2 ) − ΔG (T1 ) = − ∫ ΔSdT
ΔG( T1 ) ∂T P T1
and from (∂ (ΔG/T ) /∂T ) P

= −ΔH / T
2
, we have
ΔG( T2 )/ T2 T2
∂ (ΔG/T ) ΔG (T2 ) ΔG (T1 ) ΔH
∫ ( ) dT = − = −∫ dT
2
ΔG( T1 )/ T1 ∂T T2 T1 T1 T
P
For small temperature differences, ΔH is often approximately constant. Then, we can evaluate the change in ΔG from
ΔG (T2 ) ΔG (T1 ) 1 1
− = ΔH ( + )
T2 T1 T2 T1
Another common application arises when we know ΔG at several temperatures. A plot of ΔG/T versus 1/T is then approximately
linear with a slope that approximates the average value of ΔH in the temperature interval.
This page titled 10.11: The Gibbs-Helmholtz Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
We make extensive use of the principle that the energy of an ideal gas depends only on temperature when only pressure–volume
work is possible. In Chapter 2, we consider the Joule experiment, which provides weak evidence that this principle is correct. In the
Joule experiment, no temperature change is observed during the adiabatic free expansion of a gas whose behavior is approximately
ideal at the initial temperature and pressure. While this observation supports the principle, the accuracy attainable in the Joule
experiment is poor. Otherwise, the most compelling evidence for this principle that we have developed is the theoretical
relationship between the pressure–volume product of an ideal gas model and the mean-square velocity of its molecules. We derive
this relationship from the Maxwell-Boltzmann distribution law for gas velocities and use the ideal gas equation to find that the
mean squared velocity depends only on temperature.
To appreciate the importance of this principle, let us review some of the important steps in our development of the second law. In
Chapter 7, we observe that, since its energy depends only on temperature, C for an ideal gas must also depend only on
V
temperature; this follows immediately from the definition, C = (∂E/∂T ) . In Chapter 9, we use the conclusion that C depends
V V V
only on temperature in our development of the relationships among the heat, work, volume, and temperature changes for an ideal
gas traversing a Carnot cycle. In considering these relationships, we observe that the values of the terms q /T for the steps in this
rev
cycle sum to zero, as required for a state function. This leads us to define entropy by the differential expression dS = dq /T and rev
to infer that the entropy so defined is a state function. Reasoning from the machine-based statement of the second law, we conclude
that this inference is correct.
That the energy of an ideal gas depends only on temperature is therefore of central importance to the internal consistency of the
thermodynamic theory we have developed. It is easy to demonstrate this internal consistency. From the ideal gas equation and the
relationships developed earlier in this chapter, we can show that the quantities (∂E/∂V ) , (∂E/∂P ) , (∂H /∂V ) , (∂H /∂P ) ,
T T T T
(∂ C /∂V ) , and (∂ C /∂P )

V P are all identically zero.
T T
The fact that our theory passes this test of internal consistency is independent of the properties of real gases. However, since we
want to make predictions about the behavior of real gases, we need to be able to measure these quantities for real gases. Moreover,
because we want to understand the properties of real gases in terms of their molecular characteristics, we want to be able to
interpret these quantities for real gases using real-gas models that explain the differences between real gas molecules and ideal gas
molecules. The van der Waals equation of state provides a simple model for the effects of attractive and repulsive molecular
interactions. In the next section, we first consider simple qualitative arguments about the effects of intermolecular interactions on
the energy of a real gas. We then investigate these effects for a van der Waals gas. We see that the van der Waals model and our
qualitative arguments are consistent.
This page titled 10.12: The Second Law and the Properties of Ideal Gases is shared under a CC BY-SA 4.0 license and was authored, remixed,
Let us consider the effects that intermolecular forces of attraction and repulsion must have in the adiabatic free expansion
(P = 0) of a real gas. In such an expansion, no energy can be exchanged between the gas and its surroundings.
applied
Suppose that the molecules of the gas are attracted to one another. Then energy must be expended to separate the molecules as the
expansion takes place. (To achieve the expansion, work must be done against the intermolecular attractive forces.) Since the system
cannot obtain this energy from its surroundings, it must be obtained by decreasing the translational kinetic energy (and the
rotational and vibrational energy) of the gas molecules themselves. This means that the temperature of the gas must decrease during
the expansion.
Conversely, if the molecules repel one another, energy is released as the expansion takes place, and the temperature of the gas
increases during the expansion. The temperature can remain unchanged after the adiabatic free expansion only if the effects of the
intermolecular forces of attraction and repulsion offset one another exactly.
We can express these conclusions more precisely by saying that we expect (∂T /∂P ) > 0 if forces of attraction dominate the
E
intermolecular interactions. We expect (∂T /∂P ) < 0 if forces of repulsion dominate. Now, as a matter of mathematics, we have
E
∂E ∂E ∂T
( ) = −( ) ( )
∂P ∂T ∂P
T P E
As a matter of experience, increasing the temperature of any gas at constant pressure always increases the energy of the gas; that is,
we observe (∂E/∂T ) > 0 . It follows that we can expect
P
∂E
( ) <0
∂P
T
(attraction dominates)
when intermolecular forces of attraction dominate and
∂E
( ) >0
∂P
T
(repulsion dominates)
when forces of repulsion dominate.
In the Joule experiment, a gas is allowed to expand into an initially evacuated container. The Joule experiment is a direct test of
these ideas; however, as we have noted, it is difficult to carry out accurately. Fortunately, a simple modification of the Joule
experiment produces an experiment that is much more sensitive. Instead of allowing the gas to expand freely into a fixed volume,
we allow it to expand adiabatically against a constant applied pressure. This is the Joule-Thomson experiment. In the next section,
we show that the enthalpy of the gas does not change in such a process. We measure the temperature change as the gas expands
adiabatically from an initial, constant, higher pressure to a final, constant, lower pressure. Since this process occurs at constant
enthalpy, the Joule-Thomson experiment measures
∂T
( )
∂P
H
from which we can obtain

∂H
( )
∂P
T
To interpret the Joule-Thomson experiment in terms of intermolecular forces, we need to show that
∂H
( ) <0
∂P
T
at pressures and temperatures where intermolecular forces of attraction dominate and
∂H
( ) >0
∂P
T
where forces of repulsion dominate. To do this using an explicit mathematical model for a real gas, let us find
∂E
( )
∂P
T
and
∂H
( )
∂P
T
2
¯
¯¯¯ ¯
¯¯¯
for a van der Waals gas. Writing van der Waals equation in terms of the molar volume, (P + a/ V ) (V − b) = RT , and
introducing
¯
¯¯¯ a 2ab
γ (P , V ) = P − +
2 3
¯
¯¯¯ ¯
¯¯¯
V V
so that we can express the results more compactly, we find

¯
¯¯¯
γ (P , V )
∂P
( ) =−
¯
¯¯¯ ¯
¯¯¯
∂V T V +b
so that
¯
¯¯¯ ¯
¯¯¯
∂V V +b
( ) =−
∂P ¯
¯¯¯
T γ (P , V )
and
¯
¯¯¯
∂V R
( ) =
∂T ¯
¯¯¯
P γ (P , V )
Substituting into results we develop in Section 10.5, we have

¯
¯¯¯
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ a (V − b)
∂E ∂V ∂V
( ) = −P ( ) −T( ) =−
2
∂P ∂P ∂T ¯
¯¯¯ ¯
¯¯¯
T T P V γ (P , V )
and
¯
¯¯¯
¯ ¯
¯¯¯
∂H ¯
¯¯¯
∂V ¯
¯¯¯
RT
( ) = V +T( ) =V +
∂P ∂T ¯
¯¯¯
T P γ (P , V )
2
We introduce van der Waals equation in Section 2.12. By the argument we make there, the (P + a/ V
¯
¯¯¯
) term models the effects
of attractive intermolecular interactions when a > 0 . By a parallel argument, we can see that it models the effects of repulsive
interactions when a < 0 . Parameter b models the effects of intermolecular repulsive interactions that come into play when the
molecules come into contact with one another. For present purposes, we can consider molecules for which b = 0 ; this simplifies
our equations without affecting the description they give of the phenomena that are of current interest.
This gives us a model in which the effects of intermolecular interactions are described by the values of a single parameter that has a
straightforward physical interpretation. Thus, we can write
2
¯
¯¯¯ ¯
¯¯¯
(P + a/ V )V = RT
to describe a gas of point-mass molecules that experience intermolecular forces. When a > 0 , these forces are attractive; when
a < 0 they are repulsive. (For any given real gas, our equation can only be an approximation that is valid over a limited range of
conditions. In some ranges, a > 0 ; in others, a < 0 .) With b = 0 we must have

¯
¯¯¯ a
γ (P , V ) = P − >0
2
¯
¯¯¯
V
¯
¯¯¯ ¯
¯¯¯
(If γ (P , V ) ≤0 , we have (∂P /∂ V ) ≥0 . As a matter of experience, the pressure of a gas always decreases with increasing
T
2
¯
¯¯¯
volume at constant temperature. It follows that van der Waals equation with b =0 and (P − a/ V ) ≤0 cannot describe any
gas.) With b = 0 we have
¯
¯¯¯
∂E −a
( ) =
2
∂P ¯
¯¯¯ ¯
¯¯¯
T V (P − a/ V )
and
¯
¯¯¯
¯
∂H ⎡ 1 1 ⎤
( ) = RT −
2 2
∂P ⎣ ¯
¯¯¯ ¯
¯¯¯ ⎦
T P + a/V P − a/V
For a gas at conditions in which forces of attraction dominate, we have a > 0 , so that
¯
¯¯¯ ¯
¯¯¯
¯
(∂ E /∂P ) <0 and (∂ H /∂P ) <0
T T
Conversely, at conditions in which forces of repulsion dominate, we have a < 0 , and
¯
¯¯¯ ¯
¯¯¯
¯
(∂ E /∂P ) >0 and (∂ H /∂P ) >0
T T
This page titled 10.13: The second-dependence of the Energy and Enthalpy of A Real Gas is shared under a CC BY-SA 4.0 license and was
In practice, the Joule-Thomson experiment is done by allowing gas from a pressure vessel to pass through an insulated tube. The
tube contains a throttling valve or a porous plug through which gas flows slowly enough so that the gas upstream from the plug is
at a uniform pressure P , and the gas downstream is at a uniform pressure P . In general, the temperature of the downstream gas is
1 2
different from that of the upstream gas. Depending on the initial temperature and pressure, the pressure drop, and the gas, the
temperature of the gas can either decrease or increase as it passes through the plug. (We see below that it must be constant if the gas
is ideal.)
The temperature change is called the Joule-Thomson effect. The enthalpy of the gas remains constant. If the measured temperature
and pressure changes are ΔT and ΔP , their ratio is called the Joule-Thomson coefficient, μ . We define JT
∂T ΔT
μJT = ( ) ≈
∂P ΔP
H
Figure 3. The idealized Joule-Thomson experiment

To see that the enthalpy of the gas is the same on both sides of the plug, we consider an idealized version of the experiment, in
which the flow of gas through the plug is controlled by the coordinated movement of two pistons. (See Figure 3.) We suppose that
the gas is pushed through the plug in such a way that the upstream pressure remains constant at P and the downstream pressure 1
remains constant at P . Let us consider the changes that result when one mole of gas passes through the plug under these
2
¯
¯¯¯
conditions. Initially, there are n + 1 moles of gas on the upstream side at a pressure P , occupying a volume (n + 1) V , at a
1 1 1 1
temperature T , and having an energy per mole of E . On the downstream side, there are n moles of gas at a pressure P ,
1
¯
¯¯¯
1 2 2
occupying a volume n V , but having a temperature T and an energy per mole of E .

2
¯
¯¯¯
2 2
¯
¯¯¯
2
When the process is complete, there are n moles of gas on the upstream side, still at a pressure P and temperature T , but
1 1 1
¯
¯¯¯ ¯
¯¯¯
occupying a volume n V . On the downstream side, there are n + 1 moles of gas at pressure P , occupying volume (n + 1) V
1 1 2 2 2 2
at a temperature T and with an energy per mole of E . On the upstream side, ΔE = −E

2
¯
¯¯¯
2 and 1
¯
¯¯¯
1
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
w1 = −P1 [n1 V 1 − (n1 + 1) V 1] = P1 V 1
On the downstream side, ΔE 2 = E2

¯
¯¯¯
, and
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
w2 = −P2 [(n2 + 1) V 1 − n2 V 1] = −P2 V 2
Since the process is adiabatic, any heat taken up by the upstream gas must be surrendered by the downstream gas, so that
q + q = 0 . For the process of moving the mole of gas across the plug,
1 2
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
ΔE = ΔE1 + ΔE2 = −E 1 + E 2 = q1 + q2 + w1 + w2 = P1 V 1 − P2 V 2
from which
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
E 1 + P1 V 1 = E 2 + P2 V 2
or
¯
¯¯¯
¯ ¯
¯¯¯
¯
H1 = H2
so that we have ΔH = 0 for the expansion.
In practice, it is convenient to measure downstream pressures and temperatures, P and T , in a series of experiments in which the 2 2
upstream pressure and temperature, P and T , are constant. The enthalpy of the gas is the same at each of these pressure-
1 1
temperature points. A graph of these points is an isenthalpic (constant enthalpy) curve. At any given pressure and temperature, the
Joule-Thomson coefficient, μ , is the slope of this curve.
JT
We can also express μ

JT
as a function of the heat capacity, CP , and the coefficient of thermal expansion, α , where
¯
¯¯¯
¯
α =V
−1
(∂V /∂T )P . We begin by expressing dH as a function of temperature and pressure:
¯
¯¯¯
¯ ¯
¯¯¯
¯
¯
¯¯¯
¯ ∂H ∂H
dH = ( ) dT + ( ) dP
∂T ∂P
P T
¯
¯¯¯
¯
If we divide through by dP and hold H constant, we obtain
¯
¯¯¯
¯ ¯
¯¯¯
¯
∂H ∂T ∂H
0 =( ) ( ) +( )
∂T ∂P ¯
¯¯¯
¯ ∂P
H
P T
so that
¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
∂T ∂H ∂H 1 ∂H
μ =( ) = −( ) /( ) =− ( )
JT
∂P ¯
¯¯¯
¯ ∂P ∂T CP ∂P
H
T P T
¯
¯¯¯
¯
If we substitute the coefficient of thermal expansion into the expression for (∂ H /∂P ) that we develop in Section 10.5, we have
T
¯
¯¯¯
¯ ¯
¯¯¯
∂H ¯
¯¯¯ ∂V ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
( ) = V −T( ) = V − α V T = V (1 − αT )
∂P ∂T
T P
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
¯
For an ideal gas, (∂ V /∂T ) = V /T , so that both (∂ H /∂P ) and μJT are zero. For real gases, we substitute into the
P T
expression for μ JT
to find
¯
¯¯¯
¯ ¯
¯¯¯
∂T 1 ∂H V
μJT = ( ) =− ( ) =− (1 − αT )
∂P ¯
¯¯¯
¯ CP ∂P CP
H
T
¯
¯¯¯
Given V and any two of μ JT
, C , or α , we can find the third from this relationship.
P
Making the same substitutions using the partial derivatives we found above for a van der Waals gas, we find
⎛ ⎞
1 ¯
¯¯¯ RT
μJT = − ⎜V − ⎟
CP ¯
¯¯¯
⎝ γ (P , V ) ⎠
Given that the van der Waals equation oversimplifies the effects of intermolecular forces, we can anticipate that calculation of the
Joule-Thomson coefficient from the van der Waals parameters is likely to be qualitatively correct, but in poor quantitative
agreement with experimental results. Figure 4 compares calculated and experimental curves for the Joule-Thomson coefficient of
nitrogen gas at 0 C from 1 to 200 bar. (Calculated values take a = 0.137 Pa m mol and b = 3.81 × 10 m mol . The 6 −−2 −5 3 −−1
experimental data are from reference 1.)
Figure 4. The Joule-Thomson coefficient for N at 0 C
2
We anticipate that the Joule-Thomson coefficient becomes zero at pressures and temperatures where the effects of intermolecular
attractions and repulsions exactly offset one another. For interactions between molecules, attractive forces have the dominant effect
at long distances, while repulsive forces dominate at short distances. The lower the pressure, the greater the average distance
between gas molecules. Therefore, at any given temperature and a sufficiently low pressure, the effects of intermolecular attractive
forces are more important than those of intermolecular repulsive forces. At low pressures, the Joule-Thomson coefficient should be
positive. As the pressure increases, the effects of both attractive and repulsive forces must both increase, but at a sufficiently high
pressure, the average intermolecular distance becomes so small that the effects of intermolecular repulsive forces become
dominant. Therefore, we anticipate that the Joule-Thomson coefficient decreases as the pressure increases, eventually becoming
negative.
Experiments confirm these expectations. A temperature and pressure at which the Joule-Thomson coefficient becomes zero is
called a Joule-Thomson inversion point. The experimentally determined curve for nitrogen gas is graphed in Figure 5. The van
1
der Waals model also exhibits this effect. The inversion curve can be found from the expression for μ developed above for a van
JT
der Waals gas. The inversion curve for nitrogen that is found in this way is also graphed in Figure 5. Qualitatively, the agreement is
a satisfying confirmation of the basic interpretation that we have given for the role of intermolecular forces. Quantitatively, the
agreement is poor, as we expect given the overly simple character of the van der Waals model.
Figure 5. The Joule-Thomson inversion temperature for N (see reference 1)

2
The Joule-Thomson coefficient for an ideal gas is zero, and we normally expect the properties of real gases to approach those of an
ideal gas as the pressure falls to zero. However, both experiment and the van der Waals model indicate that the Joule-Thomson
coefficient converges to a finite value as the pressure decreases to zero at a fixed temperature. A statistical thermodynamic model 2
also predicts this outcome. This model calculates the coefficients in the virial equation of state. In it, the second virial coefficient
reflects the net effect of attractive and repulsive forces between a pair of molecules, and it is the second virial coefficient and its
¯
¯¯¯
¯
temperature derivative determine that the value of (∂ H /∂P ) . (Higher-order virial coefficients reflect interactions among larger
T
numbers of molecules.)
This page titled 10.14: The Joule-Thomson Effect is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
10.15: Problems
1. Show that
∂ (A/T ) E
( ) =−
2
∂T T
V
2. At 60 C, the vapor pressure of water is 19,932 Pa, and the enthalpy of vaporization is 42.482 kJ mol −1
.
(a) Is the vaporization of water at these conditions impossible, spontaneous, or reversible? What is ΔG for this process?
(b) Estimate ΔG for the vaporization of liquid water at 19,932 Pa and 70 C. Is this process impossible, spontaneous, or reversible?
(c) Estimate ΔG for the vaporization of liquid water at 19,932 Pa and 50 C. Is this process impossible, spontaneous, or reversible?
3. At 298.15 K and 1 bar, the Gibbs free energy of one mole of N O is 4.729 kJ less than the Gibbs free energy of two moles of
2 4
N O . The enthalpy of one mole of N O is 57.111 kJ less than the enthalpy of two moles of N O . We customarily express these
2 2 4 2
facts by saying that the Gibbs free energy and the enthalpy changes for the reaction 2 N O → N O are 2 2 4
Δ G (298.15 K) = −4.729 kJ and Δ H (298.15 K) = −57.111 kJ. Assume that the enthalpy change for this process is
o o
r r
independent of temperature. Estimate the Gibbs free energy change for this reaction at 500 K and 1 bar, Δ G (500 K) . r
o
4. Over the temperature range \(300\ \mathrm{K}, the Gibbs free energy change for the formation of ammonia from the elements,
H → N H , is well approximated by
1 3
N + 2 2 3
2 2
o 2 3
Δf G (N H3 ) = a + b (T − 600) + c (T − 600) + d (T − 600)
where a = 15.824 kJ, b = 0.1120 kJ K , c = 1.316 × 10 −−1 −5

kJ K
−−2
, and d = −1.324 × 10
−8
kJ K
−−3
. Estimate the
enthalpy change for this process, Δ H (N H ) , at 600 K. f
o
3
5. Consider the total differentials for S = S (P , T ) , E = E (P , T ) , H = H (P , T ) , A = A (P , T ) , and G = G (P , T ) . Can we

ever encounter an undefined integrand when we evaluate the line integral of one of these total differentials between any two points
(P T ) and (P T )? (In the next chapter, we find that, because of the third law of thermodynamics, no real system can ever reach
1 1 2 2
the absolute zero of temperature.)

6. Consider the total differentials for S = S (V , T ) , E = E (V , T ) , H = H (V , T ) , A = A (V , T ) , and G = G (V , T ) . Can we
ever encounter an undefined integrand when we evaluate the line integral of one of these total differentials between any two points
(V T ) and (V T )?
1 1 2 2
7. The normal boiling point of methanol is 337.8 K at 1 atm. The enthalpy of vaporization at the normal boiling point is
Δ H = 35.21 kJ mol
vap . Is the process impossible, spontaneous, or reversible? Find q, w, Δ E , Δ S , Δ A, Δ G for
−−1
vap vap vap vap
the vaporization of one mole of methanol at the normal boiling point. Assume that methanol vapor behaves as an ideal gas.
8. For S = S (P , V ) , we obtain
∂S CP ∂T
( ) = ( )
∂V T ∂V
P P
For S = S (P , T ) , we obtain
∂S CP
( ) =
∂T T
P
For temperatures near 4 C and at a pressure of 1 atm, the molar volume of water is given by
¯
¯¯¯ ¯
¯¯¯ 2
V =V 4 + a (T − 277.15)
where V ¯
¯¯¯
4 = 1.801575 × 10
−6
m
3
mol
−−1
and a = 1.45 × 10 −11
m
3
K
−1
. The heat capacity of liquid water is 75.49 J mol −1
.
(a) Using (∂S/∂T ) , calculate the entropy change when one mole of water is warmed from 2 C to 6 C while the pressure is
P
constant at 1 atm.
(b) Repeat the calculation in (a), for warming the water from 4 C to 6 C.
(c) Can we calculate the entropy change when one mole of water is warmed from 2 C to 6 C using (∂S/∂V ) ? Why, or why not? P
The required integral can be transformed to

A du 2A 1/2
∫ =( ) ln(1 + β u )+C
1/2 B
u + Bu
where C is an arbitrary constant.

(d) Using (∂S/∂V ) , calculate the entropy change when one mole of water is warmed from 4 C to 6 C. Compare this result to the
P
value obtained in (b).

9. For an ideal gas, show that (∂E/∂V ) , (∂E/∂P ) , (∂H /∂V ) , (∂H /∂P ) , (∂ C
T T T T V /∂V )
T
, and (∂ C P /∂P )
T
are all zero.

¯
¯¯¯
10. Find (∂E/∂P )
T
for a gas that obeys the virial equation of state P [V − B (T )] = RT , in which B(T ) is a function of
temperature.
11. Derive the following relationships for an ideal gas:
(a) dE = C V dT
(b) dS = (C V /T ) dT + (R/V ) dV
(c) dS = (C P /T ) dT − (R/P ) dP
12. Derive the following relationships for a gas that obeys the virial equation, ¯
¯¯¯
P [V − B (T )] = RT , where B (T ) is a function
of temperature:
(a)
⎡ ⎤
2
¯
¯¯¯ RT RT dB RT dB ¯
¯¯¯
dE = CV dT − ⎢ + ( )− ( ) ⎥ dV
⎢ ¯
¯¯¯ ¯
¯¯¯ 2 ⎥
dT ¯
¯¯¯ dT
V −B V −B
⎣ (V − B) ⎦
(b)
¯
¯¯¯ dB dB
dE = [CP − R − P ( )] dT − T ( ) dP
dT dT
(c)
⎡ ⎤
¯¯
¯
CV R RT dB ¯
¯¯¯
dS = dT + ⎢
⎢ + ( )⎥
⎥ dV
¯
¯¯¯ 2
T ¯
¯¯¯ dT
V −B
⎣ (V − B) ⎦
(d)
¯¯
¯ CP R dB
dS = dT − [ +( )] dP
T P dT
13. One mole of a monatomic ideal gas (C = 3R/2 ), originally at 10 bar and 300 K (state A), undergoes an adiabatic free
V
expansion against a constant applied pressure of 1 bar to reach state B. Thereafter the gas is warmed reversibly at constant volume
back to 300 K, reaching state C. Finally, the warmed gas is compressed reversibly and isothermally to the original pressure. What is
are the temperature and volume in state B, following the original adiabatic free expansion? Find q, w, ΔE, ΔH , and ΔS for each
of the steps and for the cycle A→B→C→A.
14. As in problem 13, one mole of a monatomic ideal gas (C = 3R/2 ), originally at 10 bar and 300 K (state A), undergoes an
V
adiabatic free expansion against a constant applied pressure of 1 bar to reach state B. The gas is then returned to its original state in
a different two-step process. From state A a reversible constant-pressure warming takes the gas to state D at the original
temperature of 300 K. The gas is then returned to state A by an isothermal compression to the original volume. What are the
temperature and volume after the constant-pressure warming step? Find q, w, ΔE, ΔH , and ΔS for each of the steps and for the
cycle A→B→D→A.
15. As in problem 13, one mole of a monatomic ideal gas (C = 3R/2 ), originally at 10 bar and 300 K (state A), undergoes an
V
adiabatic free expansion against a constant applied pressure of 1 bar to reach state B. Now consider a reversible adiabatic
expansion from the same initial state, A, that reaches the same temperature as the gas in state B. Call this state F. Find q, w, ΔE,
ΔH , and ΔS for the step A→F. Find q , w , ΔE, ΔH , and ΔS for reversible isothermal expansion from state F to state B. What
are q, w, ΔE, ΔH , ΔS , and ΔS^ for the cycle A→F→B→A. Does this cycle violate the machine-based statement of the second
law?
16. One mole of carbon dioxide, originally at 10 bar and 300 K, is taken around the cycle in problem 13. Find the energy and
entropy changes for the steps in this cycle using the ideal gas equation and the temperature-dependent heat capacity. The constant-
volume heat capacity is C = 14.7 + 0.046 × T . Find q, w, ΔE, ΔH , and ΔS for each of the steps and for the cycle when C O
V 2
is taken around the cycle A→B→C→A.

17. Ten moles of a monatomic ideal gas, initially occupying a volume of 30 L at 25 C, is expanded against a constant applied
pressure of 2 bar. The final temperature is 25 C.
(a) What is the initial pressure? The final volume?
(b) Is this process impossible, spontaneous, or reversible?
(c) Find q, w, ΔE, ΔH , ΔA, ΔS , and ΔG for this process.
18. One mole of C O , originally at 1.00 bar and 300 K, expands adiabatically against a constant applied pressure of 0.200 bar.
2
Assume that C O behaves as an ideal gas with constant heat capacity, C = 28.5 J mol
2 K . V
−−1 −−1
(a) For the spontaneous expansion, we have dE = C V dT − Papplied dV . Find the final temperature and volume for this
spontaneous expansion. What is ΔE for this process?
(b) Find the volume and pressure after the gas is compressed adiabatically and reversibly to the original temperature of 300 K.
What are ΔS and ΔE for this step?
(c) Find ΔE when the gas in the final state of part (b) is compressed isothermally to the original volume. What is ΔS for this step?
(d) What are ΔE and ΔS for the cycle comprised of the spontaneous expansion of part (a), the adiabatic compression of part (b),
and the isothermal compression of part (c)?
(e) What are ΔS , ΔS^, and ΔS universe for the spontaneous expansion?
19. Consider the energy surface depicted in Figure 1. As sketched, E increases monotonically as S increases. E decreases
monotonically as V increases. Could the energy surface decrease as S increases or increase as V increases?
20. At 298.15 K, the vapor pressure of water is 3.169 × 10 −3
Pa . Some thermodynamic properties for liquid and gaseous water at
this temperature and pressure are given in the table below.
liquid gas
¯
¯¯¯
G, kJ mol
−−1
– 237.1 – 237.1
¯¯
¯
S , J mol
−−1
K
−−1
70.0 217.5
¯
¯¯¯
E , kJ mol
−−1
– 285.5 – 245.1
CP , J mol
−−1
K
−−1
75.3 33.6
CV , J mol
−−1
K
−−1
67.0 25.3
¯
¯¯¯ ¯¯
¯ ¯
¯¯¯
1. Find Δ vap G ,Δ vap S ,Δ vap E for water at this temperature and pressure. Is this process reversible, spontaneous, or impossible?
¯
¯¯¯ ¯
¯¯¯
(b) Sketch G (ℓ) and G (g) vs. T for \(288.15. What path is followed when one mole of water at 288.15 K and 3.169 × 10
3
Pa
goes reversibly to 308.15 K at the same pressure?

(c) On the graph of part (b), indicate the transition in which superheated liquid water at 300 K and 3.169 × 10 3
Pa goes to gaseous
¯
¯¯¯
water at 300 K and the same pressure. Is this process spontaneous, reversible, or impossible? Is ΔG for this process positive, zero,
or negative?
¯
¯¯¯ ¯
¯¯¯
(d) Sketch E (ℓ) and E (g) vs. T for \(288.15. What path is followed when one mole of water at 288.15 K and 3.169 × 10
3
Pa
goes reversibly to 308.15 K at the same pressure?

(e) On the graph of part (b), indicate the transition in which superheated liquid water at 300 K and 3.169 × 10 3
Pa goes to gaseous
water at 300 K and the same pressure. Is ΔE for this process positive, zero, or negative?
¯
¯¯¯
21. At 273.15 K and 1 bar, the enthalpy of fusion of ice is 6010 J mol −−1
. Estimate the Gibbs free energy change for the fusion of
ice at 283.15 K and 1 bar.
Notes
1
J. R. Roebuck and H. Osterberg, The Joule-Thomson Effect in Nitrogen, Phys. Rev., Vol. 48, pp 450-457 (1935).
2
See T. L. Hill, An Introduction to Statistical Thermodynamics, Addison-Wesley Publishing Co., Reeding, MA, 1960, pp 266-
268.
CHAPTER OVERVIEW
11: The Third Law, Absolute Entropy, and the Gibbs Free Energy of Formation
11.3: The Third Law
11.8: A General Strategy for Expressing the Thermodynamic Properties of a Substance
11.14: Problems
This page titled 11: The Third Law, Absolute Entropy, and the Gibbs Free Energy of Formation is shared under a CC BY-SA 4.0 license and was
1
It is relatively easy to measure heat capacities as a function of temperature. If we measure the constant-pressure heat capacity of a
pure substance over a wide temperature range, we typically observe a curve like that in Figure 1. The heat capacity is a smooth,
continuous function of temperature except for a small number of discontinuities. These occur at temperatures where the substance
undergoes phase changes. These can be changes from one solid phase to another, melting to convert a solid phase to the liquid, or
vaporization to convert the liquid to the gas. The details of the curve are pressure dependent; for example, at a low pressure, we
might observe sublimation of the material from a solid phase directly into its gas phase.
Figure 1. Heat capacity of mercury versus temperature

Another general feature of these curves is that the heat capacity of the solid substance decreases to zero as the absolute temperature
decreases to zero; the curve meets the abscissa at the zero of temperature and does so asymptotically. That this is true for all
substances seems like an odd sort of coincidence. Why should all solid substances exhibit essentially the same heat capacity (zero)
at one temperature (absolute zero)?
As it turns out, this result has a straightforward molecular interpretation in the theory of statistical thermodynamics. In Section
22.6, we consider a theory of low-temperature heat capacity developed by Einstein. Einstein’s theory explains all of the qualitative
features that are observed when we measure heat capacities at low temperatures, but its predictions are not quantitatively exact.
Debye extended the Einstein model and developed a theory that gives generally excellent quantitative predictions. The Debye
theory predicts that, at temperatures near absolute zero, the heat capacity varies as the cube of temperature: C = AT , where A
P
3
is a constant. If we have heat capacity data down to a temperature near absolute zero, we can estimate the value of A from the
value of C at the lowest available temperature.
P
Anticipating results that we develop in Chapter 22, we can characterize the statistical interpretation as follows: When a system of
molecules gives up heat to its surroundings, some of the molecules move from higher energy levels to lower ones. Statistical
thermodynamics posits that the fraction of the molecules that are in the lowest energy level approaches one as the temperature goes
to zero. If nearly all of the molecules are already in the lowest energy level, decreasing the temperature still further has a negligible
effect on the energy and enthalpy of the system.
Given such heat capacity data, we can find the enthalpy or entropy change that occurs as we change the temperature of a quantity
of the substance from some reference temperature to any other value. When we use pressure and temperature as the independent
variables, we have
∂V
dH = CP dT + [V − T ( ) ] dP
∂T
P
and
CP ∂V
dS = dT − ( ) dP
T ∂T
P
At constant pressure, we have
(dH ) = CP dT
P
so that
T
H (T ) − H (Tref ) = ∫ CP dT
Tref
and
CP
(dS) = dT
P
T
so that
T
CP
S (T ) − S (Tref ) = ∫ dT
Tref T
If phase transitions occur as the temperature goes from the reference temperature to the temperature of interest, these integrations
must be carried out in steps. Also, we must include the enthalpy and entropy changes that occur during these phase changes.
This page titled 11.1: Heat Capacity as a Function of Temperature is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The fact that C goes to zero asymptotically as the temperature goes to zero has no practical ramifications for the measurement or
P
use of enthalpy. We can only measure changes in energy and enthalpy; no particular state of any system is a uniquely useful
reference state for the enthalpy function. Experimental convenience is the only consideration that makes one reference state a better
choice than another. In Chapter 8, we define standard enthalpies of formation for elements and compounds. For this purpose, we
choose to define the standard enthalpy of formation of each element to be zero at every temperature. For standard enthalpies of
formation, the reference state is different at every temperature.
Compilations of thermodynamic data often choose 298.15 K and one bar as the zero of enthalpy for pure substances. (In citing data
from such compilations, 298.15 K is frequently abbreviated to “298 K.”) Because it is near the ambient temperature of most
laboratories, much thermochemical data has been collected at or near this temperature. Choosing a reference temperature near
ambient laboratory temperatures helps minimize the errors introduced when we extrapolate experimental thermodynamic data to
the reference temperature. Expressed relative to a reference temperature, the substance’s enthalpy at any other temperature is the
change in enthalpy that occurs when the substance is taken from the reference temperature to that temperature. Such enthalpy
changes are often called absolute enthalpies. Enthalpy-data tables frequently include values for “H − H ” at a series of
o
T
o
298K
temperatures. These data should not be confused with enthalpies of formation.
This page titled 11.2: Enthalpy as a function of Temperature is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
11.3: The Third Law
For entropy on the other hand, the fact that the heat capacity goes to zero as the temperature decreases has important consequences.
Consider the change in the entropy of a pure substance whose heat capacity approaches some finite limiting value as its
temperature decreases to absolute zero. For such a substance, C /T becomes arbitrarily large as the temperature decreases, and the
P
entropy integral
0
CP
∫ dT
T
T
approaches minus infinity as the temperature approaches zero. For real substances, this does not occur. In the neighborhood of
absolute zero, heat capacities decrease more rapidly than temperature. The entropy change approaches zero as the temperature
approaches zero.
The idea that the entropy change for a pure substance goes to zero as the temperature goes to zero finds expression as the third law
of thermodynamics. In 1923, Lewis and Randall gave a statement of the third law that is particularly convenient in chemical
1
applications:
If the entropy of each element in some crystalline state be taken as zero at the absolute
zero of temperature, every substance has a positive finite entropy; but at the absolute zero
of temperature the entropy may become zero, and does so become in the case of perfect
crystalline substances.
Implicitly, the Lewis and Randall statement defines the entropy of any substance, at any temperature, T , to be the difference
between the entropy of the constituent elements, at absolute zero, and the entropy of the substance at temperature T . Equivalently,
we can say that it is the entropy change when the substance is formed at temperature T from its constituent elements at absolute
zero. Arbitrarily, but very conveniently, the statement sets the entropy of an element to zero at absolute zero.
The distinction between perfect crystalline substances and less-than-perfect crystalline substances lies in the regularity of the
arrangement of the molecules within the crystal lattice. In any lattice, each molecule of the substance is localized at a specific site
in the lattice. In a perfect crystal, all of the molecules are in oriented the same way with respect to the lattice. Some substances
form crystals in which the molecules are not all oriented the same way. This can happen when the molecule can fit into a lattice site
of the same shape in more than one way. For example, in solid carbon monoxide, the individual molecules occupy well-defined
lattice sites. If the carbon monoxide crystal were perfect, all of the molecules would point in the same direction, as diagrammed in
Figure 2. Instead, they point randomly in either of two possible directions.
Figure 2. A two-dimensional representation of a perfect CO crystal.
This page titled 11.3: The Third Law is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
The third law arises in a natural way in the development of statistical thermodynamics. It is probably fair to say that the classical
thermodynamic treatment of the third law was shaped to a significant degree by the statistical thermodynamic treatment that
developed about the same time. Nevertheless, we can view the third law as an inference from thermochemical observations.
Walther Nernst was the first to recognize the principle that underlies the third law. From published experimental results, Nernst
inferred a postulate known as the Nernst heat theorem. The experimental results that inspired Nernst were measurements of
enthalpy and Gibbs free energy differences, Δ H and Δ G, for particular reactions at a series of temperatures. (We define Δ H
r r r
o
in Section 8.6. We define Δ H the same way, except that the reactants and products are not all in their standard states. Likewise,
r
Δ G and Δ S are differences between Gibbs free energies and entropies of reactants and products. We give a more precise
r r
definition for Δ G in Section 13.2.) As the temperature decreased to a low value, the values of Δ H and Δ G converged. Since
r
o
r r
T Δ S = Δ H − Δ G , this observation was consistent with the fact that the temperature was going to zero. However, Nernst
r r r
concluded that the temperature factor in T Δ S was not, by itself, adequate to explain the observed dependence of Δ H − Δ G on
r r r
temperature. He inferred that the entropy change for these reactions decreased to zero as the temperature decreased to absolute zero
and postulated that this observation would prove to be generally valid. The Nernst heat theorem asserts that the entropy change for
any reaction of pure crystalline substances goes to zero as the temperature goes to zero.
Subsequently, Max Planck suggested that the entropy of reaction goes to zero because of a still more basic phenomenon: the
entropy of every crystalline substance goes to zero as the temperature goes to zero. Further investigation then showed that Planck’s
formulation fails for substances like carbon monoxide, in which the crystalline solid does not become perfectly ordered at the
temperature goes to zero. The Lewis and Randall statement refines the Planck formulation by recognizing that non-zero entropies
will be observed at absolute zero for solids that are not crystalline and for crystalline solids that are not perfectly ordered. The
Lewis and Randall statement also makes a choice (implicit also in the Planck formulation) of the zero point for the entropies of
chemical substances—namely, “some crystalline state” of each element at absolute zero. This choice ensures that, at any
temperature greater than zero, the entropy of every substance will be greater than zero.
This page titled 11.4: Genesis of the Third Law - the Nernst Heat Theorem is shared under a CC BY-SA 4.0 license and was authored, remixed,
By the Lewis and Randall statement of the third law, the entropy of a substance that forms a perfect crystal is identically equal to
zero at absolute zero. Much as the ideal gas temperature scale has a natural zero at the temperature at which the volume
extrapolates to zero, a perfect crystalline substance has a natural zero of entropy at this same temperature. We can choose a non-
zero value for the absolute zero of temperature. The Centigrade scale is based on such a choice. However, for thermodynamic
purposes, any such choice is much less convenient. Similarly, we could choose arbitrary values for the entropies of the elements at
the absolute zero of temperature. The entropy of a perfect crystalline substance at absolute zero would then be the sum of the
entropies of its constituent elements. (See problem 5.) However, choosing non-zero values proves to be much less convenient.
Given the entropy of a substance at absolute zero, its entropy at any higher temperature can be calculated from the entropy changes
that occur as the substance is warmed to the new temperature. At the very lowest temperatures, this entropy change is calculated by
integrating C /T , using Debye’s theoretical relationship, C = AT ; A is obtained from the value of C at the lowest
P P
3
P
temperature for which an experimental value of C is available. In temperature ranges where experimental heat capacity data are
P
available, the entropy change is obtained by integration using these data.

Phase changes are isothermal and reversible. Where the substance undergoes phase changes, the contribution that the phase change
makes to the entropy of the substance is equal to the enthalpy change for the phase change divided by the temperature at which it
occurs.
At any given temperature, the entropy value that is obtained in this way is called the substance’s absolute entropy or its third-law
entropy. When the entropy value is calculated for one mole of the substance in its standard state, the resulting absolute entropy is
called the standard entropy. The standard entropy is usually given the symbol S . It is usually included in compilations of
o
thermodynamic data for chemical substances.

We write S (T ) to indicate the absolute entropy of substance A in its standard state at temperature T . S (T ) is the entropy of the
o
A
o
substance in its standard state at absolute zero plus the entropy increase that occurs as the substance changes reversibly to its
standard state at T . So long as substance A forms a perfect crystal at absolute zero, S (T ) is the difference between its molar
o
A
entropy at T and its molar entropy at absolute zero—as calculated from heat capacity and phase-change enthalpy data.
If substance A does not form a perfect crystal at absolute zero, the true value of S (T ) exceeds the calculated value. The excess is
o
the molar entropy of the imperfect crystal at absolute zero. We observe the discrepancy when measured values, at T , of entropies of
reactions that involve A fail to agree with those calculated using the incorrect value of S (T ) . o
A
In Section 11.2 we note that many tables of thermochemical properties present “absolute enthalpy” data for chemical substances.
An absolute enthalpy is the difference between the enthalpies of a substance at two different temperatures, but the reference
temperature is not absolute zero. In Section 6.4 and Section 6.5, we introduce enthalpy standard states and the standard enthalpy of
formation of substance A at temperature T , which we designate as Δ H (T ). We define the standard enthalpy of formation of
f
o
any element at any temperature to be zero. In Section 8.6, we find that the enthalpy difference between reactants in their standard
states and products in their standard states is readily calculated from the standard enthalpies of formation of the participating
substances. As illustrated in Figure 8-2, this calculation is successful because it utilizes an isothermal cycle which begins and ends
in a common set of elements, all of which are at the same temperature.
We can also define the standard entropy of formation of any substance to be the difference between its standard entropy, S (T ) , o
A
and those of its pure constituent elements in their standard states at the same temperature. This definition is embedded in the Lewis
and Randall statement of the third law. For example, the standard entropy of formation of water at 400 K is the difference
o o
Δf S (H2 O, 400 K) = ΔS (H2 O, 400K)
1
o o
−ΔS (H2 , 400 K) − ΔS (O2 , 400K)
2
Because of this definition, the standard entropy of formation of an element in its standard state is zero. We can calculate the
standard entropy change for any reaction, Δ S (T ), either as the difference between the standard entropies of formation (the
r
o
Δ S (T ) values) of the reactants and products or as the difference between their standard entropies (the S (T ) values). Either
o o
f
A
calculation is successful because it begins and ends with a common set of elements, all of which are at the same temperature. When
we compute Δ S (T ) using values of S (T ) for the reactants and products the reference temperature for the elements is absolute
r
o
A
o
zero. When we compute Δ S (T ) using values of Δ S (T ) for the reactants and products, the reference temperature is T .
r
o
f
o
Given Δf H
o
(T ) and , the standard Gibbs free energy of formation is immediately obtained from
Δf S
o
(T )
Δf G
o
(T ) = Δf H
o
. For any element at any temperature, we have Δ H = 0 and Δ S = 0 ; it follows that
(T ) − T Δf S
o
(T ) f
o
f
o
the standard Gibbs free energy of formation of an element in its standard state is zero. Tables of thermodynamic data usually give
values forΔ H , Δ G , and S . (A set of standard entropies contains the same information as the corresponding set of entropies
f
o
f
o o
of formation. Entropies of formation are seldom tabulated. If Δ S is needed, it can be calculated either from Δ H and Δ G or
f
o
f
o
f
o
from the absolute entropies of the substance and the elements from which it is formed.)
This page titled 11.5: Absolute Entropy is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
The standard state for entropies is essentially the same as the standard state for enthalpies. For liquids and solids, the standard state
for entropies is identical to that for enthalpies: At any given temperature, the standard state is the most stable form of the substance
at that temperature and a pressure of 1 bar.
For gases, the ideal gas standard state for entropy tabulations is the hypothetical ideal gas state at a pressure of 1 bar. A substance in
a hypothetical ideal gas state is a creature of theory, and we obtain its thermodynamic properties from calculations that use the
experimentally determined properties of the real gas. The idea behind these calculations is that we can—by calculation—remove
the effects of intermolecular interactions from the measured properties. When we want the properties of the real gas, we put these
effects back. From this perspective, we can say that a substance in its hypothetical ideal gas standard state has the entropy that the
real gas would have if it behaved as an ideal gas at pressures of 1 bar and below.
In principle, this differs from the standard state chosen for enthalpy, because the enthalpy standard state is defined to be an arbitrary
low pressure at which the substance behaves as an ideal gas. However, because the enthalpy of an ideal gas is independent of
pressure, we can consider the standard state for either enthalpy or entropy to be the hypothetical ideal gas at a pressure of 1 bar.
(Below, we write “H I G ” to designate this state.) Since the Gibbs free energy is defined by G = H − T S , we can also describe
o
the standard state for the Gibbs free energy of gases as the hypothetical ideal gas at 1 bar.
To see precisely what we mean by the hypothetical ideal gas standard state, let us consider the conversion of one mole of a real gas,
initially at pressure P and temperature T , to its hypothetical ideal gas state at 1 bar and the same temperature. We accomplish this
in a three-step process. In the first step, we reversibly and isothermally expand the real gas to some arbitrary low pressure at which
the real gas behaves ideally. The second step is purely conceptual: We suppose that the real gas changes into a hypothetical gas that
behaves ideally. The third step is the reversible isothermal compression of the hypothetical ideal gas to a pressure of 1 bar. Letting
A denote the gas, these steps are
1. A (real gas, P ) → A (real gas, P

∗
≈ 0)
2. A (real gas, P
∗
≈ 0) → A (ideal gas, P
∗
≈ 0)
3. A (ideal gas, P
∗ o
≈ 0) → A (H I G , P
o
= 1 bar)
We use the symbol “P ” to designate the standard-state pressure. P is a constant, whose value is 1 bar.
o o
¯¯
¯ ¯
¯¯¯
Let the enthalpy, entropy, and Gibbs free energy changes for the i step be Δ H , Δ S , and Δ G. (We use the overbar to
th
i
¯
¯¯¯
¯
i i
emphasize that the system consists of one mole of a pure substance. Since the superscript zero implies that H (H I G ) , for A
o o
example, is a property of one mole of A in its standard state, we omit the overbar from standard state properties.) From our
expressions for dH and dS as functions of pressure and temperature, we have, at constant temperature,
¯
¯¯¯
¯
¯¯¯
¯ ¯
¯¯¯
∂V
(dH ) = [V − T ( ) ] dP
T ∂T
P
and
¯
¯¯¯
¯¯
¯ ∂V
(dS ) = −( ) dP
T ∂T
P
For an ideal gas,

¯
¯¯¯
¯
(dH ) =0
T
and
¯¯
¯
(dS ) = −R/P
T
For step [GrindEQ__1_], the enthalpy change is

∗
P ¯
¯¯¯
¯
¯¯¯
¯ ¯
¯¯¯
∂V
Δ1 H = ∫ [V − T ( ) ] dP
P ∂T
P
the entropy change is
∗
P ¯
¯¯¯
¯¯
¯ ∂V
Δ1 S = ∫ −( ) dP
P
∂T
P
We must evaluate the partial derivative using an equation of state that describes the real gas.
Step [GrindEQ__2_] is merely a change in our perspective; nothing actually happens to the gas. There is no enthalpy or entropy
¯¯
¯ ¯
¯¯¯
¯
change: Δ 2S =0 and Δ 2H =0 .
For step [GrindEQ__3_], the enthalpy change is zero because the gas is ideal, Δ ¯
¯¯
3H
¯
¯
=0 . We evaluate the entropy change using the
ideal gas equation; then
o
P
¯¯
¯ R
Δ3 S = ∫ − dP
P
∗ P
¯¯
¯ ¯
¯¯¯
Let H (P ), S (P ), and G (P ) be the molar enthalpy, entropy, and Gibbs free energy of real gas A at pressure P . Let us
¯
¯¯¯
¯
A A A
express the enthalpy, entropy, and Gibbs free energy of the standard state substance relative to the corresponding properties of the
constituent elements in their standard states at temperature T . Then the molar enthalpy, entropy, and Gibbs free energy of A in its
hypothetical ideal gas standard state are Δ H (H I G ) , Δ S (H I G ) , and Δ G (H I G ) . Since these are state functions, we
f
o
A
o
f
A
o o
f
o
A
o
have
∗
P ¯
¯¯¯
o ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯ ∂V
o
Δf H (H I G ) = H A (P ) + Δ1 H + Δ2 H + Δ3 H = H A (P ) + ∫ [V − T ( ) ] dP
A
P
∂T
P
and
∗ o
P ¯
¯¯¯ P
¯¯
¯ ¯¯
¯ ¯¯
¯ ¯¯
¯ ¯¯
¯ ∂V R
o o
Δf S (H I G ) = S A (P ) + Δ1 S + Δ2 S + Δ3 S = S A (P ) − ∫ ( ) dP − ∫ dP
A
P ∂T P
∗
P
P
These equations relate the enthalpy and entropy of the hypothetical ideal gas standard state to the enthalpy and entropy of the real
gas at pressure P , at the same temperature.
To evaluate the enthalpy change, we can use the virial equation for the volume of a real gas, set P = 0 , and evaluate the resulting ∗
integral. (See problem 1a.) For the entropy, there is a complication. If we set P = 0 , neither integral is finite. We can overcome∗
this difficulty by choosing a small positive value for P . Then both integrals are finite, and the value of their sum remains finite in
∗
the limit as P → 0 . This occurs because the molar volume of any gas approaches the molar volume of an ideal gas at a
∗
sufficiently low pressure. (See problem 1.)
This page titled 11.6: The Standard State for Third-law Entropies is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
As it turns out, the pressure-dependence of the Gibbs free energy is useful more often than that of the enthalpy or entropy. From the
defining relationship, G = H − T S , and the results above, we find
o o o o o o
Δf G (H I G ) = Δf H (H I G ) − T Δf S (H I G )
A A A
∗ ∗ o
P ¯
¯¯¯ P ¯
¯¯¯ P
¯
¯¯¯
¯ ¯
¯¯¯ ∂V ¯¯
¯ ∂V RT
= H A (P ) + ∫ [V − T ( ) ] dP − T S A (P ) + ∫ T( ) dP + ∫ dP
P
∂T P
∂T P
∗ P
P P
∗
P P
¯
¯¯¯ ¯
¯¯¯ RT
= GA (P ) − ∫ V dP − ∫ dP
P
∗
P
o P
where we have reversed the direction of integration in the remaining integrals. Again, we cannot evaluate the integrals with
= 0 . To overcome this difficulty, we introduce a clever expedient: we add and subtract the same integral, obtaining
∗
P
∗
P P
o ¯
¯¯¯ ¯
¯¯¯ RT
o
Δf G (H I G ) = GA (P ) − ∫ V dP − ∫ dP
A
P
∗
P
o P
P P
RT RT
−∫ dP + ∫ dP
P
∗ P P
∗ P
P P
¯
¯¯¯ ¯
¯¯¯
RT RT
= GA (P ) − ∫ [V − ] dP − ∫ dP
P
∗ P P
o P
This result is pivotal. We will derive it again later by a slightly different argument. For now, we evaluate the last integral and
rearrange the equation to the form in which we write it most often:
P ¯
¯¯¯
¯
¯¯¯
P V 1
o o
GA (P ) = Δf G (H I G ) + RT ln( ) + RT ∫ [ − ] dP
A 0
P 0 RT P
In doing so, we use the fact that

¯
¯¯¯
V 1
−
RT P
vanishes for any gas when the pressure is sufficiently low, which means that the integral from P to P remains finite when we let ∗
→ 0.
∗
P
¯
¯¯¯
Recapitulating: The last equation relates the Gibbs free energy of one mole of a real gas at pressure P and temperature T , G (P ), A
to the Gibbs free energy of formation of the gas in its standard state, Δ G (H I G ) , where the standard state is the hypothetical
f
o
A
o
ideal gas at P = P = 1 bar and the same temperature. (We write “RT ln(P /P ) ” rather than “RT ln P ” to emphasize that the
o o
argument of the logarithm is a dimensionless quantity.)

If the gas is ideal, the integrand is zero at any pressure. For an ideal gas, the Gibbs free energy, at pressure P and temperature T , is
related to the Gibbs free energy in the standard state by
¯
¯¯¯ P
o
GA (P ) = G (1 bar) + RT ln( )
A o
P
(ideal gas)
This observation enables us to give a physical interpretation to the integral
P P ¯
¯¯¯
¯
¯¯¯
RT V 1
∫ [V − ] dP = RT ∫ [ − ] dP
0
P 0
RT P
image
Evidently, this integral is the difference between the Gibbs free energy of one mole of the real gas and the Gibbs free energy that it
would have if it behaved as an ideal gas. When we add it to the Gibbs free energy of the hypothetical ideal gas, we get the Gibbs
free energy of the real gas.
For an ideal gas, the Gibbs free energy is a simple function of its pressure. It turns out to be useful to view the integral as a
contribution to a “corrected pressure.” The “correction” is an adjustment to the pressure that, in our calculations, makes the real gas
behave as an ideal gas. The idea is that we can express the Gibbs free energy as a function of this corrected pressure, which we call
the fugacity, and to which we give the symbol “f ”. Fugacity is therefore a function of pressure. Fugacity also has the units of
pressure, which we always take to be bars. We define the fugacity of A at pressure P , f (P ), by A
¯
¯¯¯
fA (P )
o o
GA (P ) = Δf G (H I G ) + RT ln( )
A o
fA (H I G )
(real gas)
The fugacity, f (H I G ) , and the standard Gibbs free energy of formation, Δ G (H I G ) , describe the same state. We define
A
o
f
o
A
o
the fugacity in this standard state to be one bar. That is, the fugacity of a substance in its hypothetical ideal-gas standard state,
(H I G ) , is a constant whose value is 1 bar. (If we want to express fugacity in units of, say, pascals, then
o
fA
(H I G ) = 10 Pa .)
o o 5
P =f A
We can calculate the fugacity of a real gas, at pressure P and temperature T , from
P ¯
¯¯¯
fA (P ) P V 1
RT ln( ) = RT ln( ) + RT ∫ [ − ] dP
o
fA (H I G ) P0 0
RT P
We find it useful to introduce a function of pressure that we call the fugacity coefficient, γ A
(P ) . We define it as
fA (P )
γ (P ) =
A
P
Since fA
o
(H I G ) = P
o
= 1 bar , we have
P ¯
¯¯¯
V 1
ln γA (P ) =∫ [ − ] dP
0
RT P
Figure 3. Graphical description of ln γ

Figure 3 exhibits these quantities graphically. The shaded area is ln γ A
(P ) .
This page titled 11.7: The Fugacity of a Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
11.8: A General Strategy for Expressing the Thermodynamic Properties of a
Substance
¯
¯¯¯
¯
The equations we develop in Sections 11.7 and 11.8 express the differences Δf H
A
o o
(H I G ) − H A (P ) ,
¯¯
¯ ¯
¯¯¯
Δf S
o
A
o
, and Δ G (H I G ) − G (P ) between thermodynamic functions for one mole of a gas at two
(H I G ) − S A (P ) f
o
A
o
A
different pressures and the same temperature. They follow from the properties of gases and the relationships that result when we
express the enthalpy, entropy, and Gibbs free energy as functions of temperature and pressure. Because Δ H (H I G ) , f
o
A
o
(H I G ) , and Δ G (H I G ) are measured relative to the constituent elements of substance A at the same temperature,
o o o o
Δ S
f f
A A
¯
¯¯¯
¯ ¯¯
¯ ¯
¯¯¯
,
H A (P ) S A (P ) , and G (P ) are the differences in these properties between the real substance at pressure P and temperature
A T
and the real constituent elements in their standard states at the same temperature.
We have just found a way to express the thermodynamic functions of a pure real gas at any pressure and temperature. This
development shows us the way toward a broader goal. Ultimately, we want to be able to express—and to find the values of—the
thermodynamic functions of any substance in any system at any temperature and pressure.
Our goal is to create a scheme in which the enthalpy, the entropy, or the Gibbs free energy
of any substance in any arbitrary state is equal to the change in that thermodynamic
property when the substance is produced, in that state, from its pure, separate, constituent
elements, in their standard states at the same temperature.
The scheme we create uses two steps to convert the constituent elements into the substance in the arbitrary state. The elements are
first converted into the pure substance in its standard state at the same temperature. The substance is then taken from its standard
state to the state it occupies in the arbitrary system, at the same temperature. While straightforward in principle, finding changes in
thermodynamic properties for this last step is often difficult in practice.
¯
¯¯¯
The value of the Gibbs free energy of substance A in any arbitrary system, G , becomes equal to the sum of its Gibbs free energy
A
in the standard state, Δ G , and the Gibbs free energy change when the substance passes from its standard state into its state in the
f
o
A
arbitrary system, ΔG . That is,

¯
¯¯¯
A
¯
¯¯¯ o ¯
¯¯¯
GA = Δf G + ΔGA
A
¯
¯¯¯
and G is the same thing as the Gibbs free energy change when the substance is formed, as it is found in the arbitrary system, from
A
its constituent elements. The same relationships apply to the enthalpies and entropies of these states:
¯
¯¯¯
¯ o ¯
¯¯¯
¯
H A = Δf H + ΔH A
A
and
¯¯
¯ o ¯¯
¯
S A = Δf S + ΔS A
A
¯¯
¯ ¯
¯¯¯
Measuring the differences ΔH , ΔS , and ¯
¯¯¯
¯
A A
ΔGA become important objectives—and major challenges—in the study of the
thermodynamics of chemical systems.
In Section 11.6, we find the changes that occur in the enthalpy, entropy, and Gibbs free energy when one mole of a pure gas is
taken from its real gas state at any pressure to its hypothetical ideal-gas standard state, at the same temperature. In Chapter 13, we
extend this development to express the thermodynamic properties of any mixture of ideal gases in terms of the properties of the
individual pure gases. As a result, we can find the equilibrium position for any reaction of ideal gases from the thermodynamic
¯
¯¯¯
properties of the individual pure gases. This application is successful because we can find both Δ G and ΔG for an ideal gas at f
o
A
A
¯
¯¯¯
any pressure. Beginning in Chapter 14, we extend this success by finding ways to measure ΔG for the process of taking A from A
its standard state as a pure substance to any arbitrary state in which A may be only one component of a solution or mixture.
This page titled 11.8: A General Strategy for Expressing the Thermodynamic Properties of a Substance is shared under a CC BY-SA 4.0 license
and was authored, remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform;
a detailed edit history is available upon request.
Given Δ G for all of the species involved in a reaction, we can calculate the difference between the Gibbs free energies of
f
o
formation of the pure separate products and those of the pure separate reactants. We call this difference the standard Gibbs free
energy change for the reaction, Δ G . A standard Gibbs free energy of formation is the standard Gibbs free energy change for the
r
o
reaction that forms a substance from its elements. Likewise, from the absolute entropies, S , of the reactants and products, we can
o
calculate the standard entropy change for the reaction, Δ S . In doing so, we utilize the thermochemical cycle that we introduced
r
o
to calculate Δ H from the values of Δ H for the reacting species. For

r
o
f
o
a A+b B → c C +d D
we have
o o o o o
Δr G = c Δf G (C ) + d Δf G (D) − a Δf G (A) − b Δf G (B)
and
o o o o o
Δr S =c S (C ) + d S (D) − a S (A) − b S (B)
We use Δ G (H I G ) to denote the Gibbs free energy of one mole of a gas in its hypothetical ideal-gas standard state. Because
f
o
A
o
the fugacity of the ideal gas standard state is 1 bar, f (H I G ) = P , the Gibbs free energy of a gas at unit fugacity becomes the
A
o o
Gibbs free energy change for the formation of the substance in its hypothetical ideal gas standard state. For an ideal gas, unit
fugacity occurs at a pressure of one bar. For real gases, the standard state of unit fugacity occurs at a real-gas pressure that is, in
general, different from one bar.
¯
¯¯¯
The Gibbs free energy of the gas at any other pressure, G (P ), becomes identical to the difference between the Gibbs free energy
A
of the gas in that state and the Gibbs free energy of its constituent elements in their standard states at the same temperature. This
convention makes the Gibbs free energy of the elements the “zero point” for the Gibbs free energy of the gas. As indicated in
Figure 4, the Gibbs free energy of the gas at any pressure, P, becomes
¯
¯¯¯
fA (P )
o o
GA (P ) = Δf G (H I G ) + RT ln( )
A o
fA (H I G )
fA (P )
o o
= Δf G (H I G ) + RT ln( )
A o
P
This page titled 11.9: The Standard Entropy and the Gibbs Free Energy of Formation is shared under a CC BY-SA 4.0 license and was authored,
It is worthwhile to call attention to some important aspects of this development. The hypothetical ideal gas standard state is a
wholly theoretical construct. We create this “substance” only because it is convenient to have a name for the “unreal” state of
substance A, whose Gibbs free energy we have denoted as G (H I G ) = Δ G (H I G ) . We have developed procedures for
o
A
o
f
o
A
o
calculating Δ G (H I G ) from the properties of the corresponding real gas. Given the properties of real gas A, these procedures
f
o
A
o
determine Δ G (H I G ) uniquely. Δ G (H I G ) is a useful quantity; the calculation of Δ G (H I G ) is “real” even though

f
o
A
o
f
o
A
o
f
o
A
o
the substance it putatively describes is not.

Other hypothetical states are frequently useful. Problem 8 in this chapter considers a hypothetical liquid state of methanol at 500 K
and 1 atm—conditions at which the real substance is a gas. Alternative approximations enable us to calculate the Gibbs free energy
of this hypothetical state in different ways. The results have predictive value. Not surprisingly, however, the alternative
approximations produce Gibbs free energy values whose quantitative agreement is poor. Were it useful to do so, we could select
one particular approximation and define the Gibbs free energy of the hypothetical superheated liquid methanol to be the value
produced using that approximation. This would not make the superheated liquid methanol any more real, but it would uniquely
define the Gibbs free energy of the hypothetical substance.
Later in our development, we create other hypothetical reference states. As for the hypothetical ideal gas standard state, we specify
unique ways to calculate the properties of these hypothetical states from measurements that we can make on real systems.
This page titled 11.10: The Nature of Hypothetical States is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
We can find the Gibbs free energy of formation for substances whose standard states are condensed phases. As indicated in Section
11.9, we adopt the same rule for any substance; we set G = Δ G (A) for any substance, whether its standard state is a gas,
o
A f
o
liquid, or solid. The Gibbs free energy of the elements becomes the “zero point” for the Gibbs free energy of any substance.
In Chapters 14 and 15, we see that we can also define the fugacity of any substance in any system; that is, we can define the
fugacity for a pure liquid, a pure solid, or for one component in any mixture. When we do so, the Gibbs free energy of one mole of
¯
¯¯¯
the substance in the system, G A (system, P ) , is given by the same relationship we developed for the molar Gibbs free energy of a
pure gas. We find
¯
¯¯¯
fA (system, P )
o o
GA (system, P ) = Δf G (H I G ) + RT ln( )
A o
fA (H I G )
To obtain this result and to see how to find the fugacity of A in any system, fA (system, P ) , we must introduce a number of
additional ideas. For now, let us note some of the consequences.
Figure 4. Fugacity and Gibbs free energy

The essential consequence is that the difference between the Gibbs free energy of one mole of a substance in two different systems,
say system X and system Y, can be expressed using the ratio of the corresponding fugacities. That is,
¯
¯¯¯ ¯
¯¯¯
fA (system, PX )
GA (system X, PX ) − GA (system Y, PY ) = RT ln( )
fA (system Y, PY )
where P and P are the pressures of systems X and Y, respectively, and both systems are at the same temperature.
X Y
For liquids and solids, the standard state is the pure substance in its most stable form at one bar and the temperature of interest. The
fugacity in the standard state must be determined experimentally. If the liquid or solid has negligible vapor pressure, this may not
be possible. Since we intend “any system” to include all manner of mixtures and solutions, it can be very difficult to find the Gibbs
free energy change for taking the substance from its standard state to the arbitrary system in which its fugacity is f (system, P ).
A
In Chapter 14, we introduce the chemical activity of the substance to cope with such cases.
When we define the chemical activity of a substance in a particular system, we also introduce a new standard state. The primary
criterion for our choice of this activity standard state is that we be able to measure how much the Gibbs free energy of the
substance differs between the activity standard state and other states of the system. A principal object of the next seven chapters is
to introduce ideas for measuring the difference between the Gibbs free energy of a substance in two states of a given system. Even
so, our treatment of the issues involved in this step is quite incomplete.
This page titled 11.11: The Fugacity and Gibbs Free Energy of A Substance in Any System is shared under a CC BY-SA 4.0 license and was
As for the standard enthalpy of reaction, we can obtain the standard entropy of reaction at a new temperature by evaluating entropy
changes around a suitable thermochemical cycle. To do so, we need the standard entropy change at one temperature. We also need
heat capacity data for all of the reactants and products. For the reaction
a A+b B → c C +d D
we can evaluate the entropy change at a second temperature by summing the individual contributions to the change in entropy
around the cycle in Figure 5. For this cycle, we have
T2 T2 T2 T2
CP (C ) CP (D) CP (A) CP (B)
o o
Δr S (T2 ) = Δr S (T1 ) + c ∫ dT + d ∫ dT − a ∫ dT − b ∫ dT
T1
T T1
T T1
T T1
T
Figure 5. Cycle relating the entropy changes for a reaction at two temperatures.
This page titled 11.12: Evaluating Entropy Changes Using Thermochemical Cycles is shared under a CC BY-SA 4.0 license and was authored,
The third law postulates that the entropy of a substance is always finite and that it approaches a constant as the temperature
approaches zero. The value of this constant is independent of the values of any other state functions that characterize the substance.
For any given substance, we are free to assign an arbitrarily selected value to the zero-temperature limiting value. However, we
cannot assign arbitrary zero-temperature entropies to all substances. The set of assignments we make must be consistent with the
experimentally observed zero-temperature limiting values of the entropy changes of reactions among different substances. For
perfectly crystalline substances, these reaction entropies are all zero. We can satisfy this condition by assigning an arbitrary value to
the zero-temperature molar entropy of each element and stipulating that the zero-temperature entropy of any compound is the sum
of the zero-temperature entropies of its constituent elements. This calculation is greatly simplified if we let the zero-temperature
entropy of every element be zero. This is the essential content of the third law.
The Lewis and Randall statement incorporates this selection of the zero-entropy reference state for entropies, specifying it as “a
crystalline state” of each element at zero degrees. As a result, the entropy of any substance at zero degrees is greater than or equal
to zero. That is, the Lewis and Randall statement includes a convention that fixes the zero-temperature limiting value of the entropy
of any substance. In this respect, the Lewis and Randall statement makes an essentially arbitrary choice that is not an intrinsic
property of nature. We see, however, that it is an overwhelmingly convenient choice.
We have discussed alternative statements of the first and second laws. A number of alternative statements of the third law are also
possible. We consider the following:
It is impossible to achieve a temperature of absolute zero.

This statement is more general than the Lewis and Randall statement. If we consider the application of this statement to the
temperatures attainable in processes involving a single substance, we can show that it implies, and is implied by, the Lewis and
Randall statement.
The properties of the heat capacity, C , play a central role in these arguments. We have seen that C is a function of temperature.
P P
While it is not useful to do so, we can apply the defining relationship for C to a substance undergoing a phase transition and find
P
CP = ∞ . If we think about a substance whose heat capacity is less than zero, we encounter a contradiction of our basic ideas about
heat and temperature: If q > 0 and q/ΔT < 0 , we must have ΔT < 0 ; that is, heating the substance causes its temperature to
decrease. In short, the theory we have developed embeds premises that require C > 0 for any system on which we can make P
measurements.
Let us characterize a pure-substance system by its pressure and temperature and consider reversible constant-pressure processes in
which only pressure–volume work is possible. Then (∂S/∂T ) = C /T and dS = C dT /T . We now want to show: the Lewis
P P P
and Randall stipulation that the entropy is always finite requires that the heat capacity go to zero when the temperature goes to
zero. (Since we are going to show that the third law prohibits measurements at absolute zero, this conclusion is consistent with our
conclusion in the previous paragraph.) That the heat capacity goes to zero when the temperature goes to zero is evident from
S = C dT /T . If C
P does not go to zero when the temperature goes to zero, dS becomes arbitrarily large as the temperature goes
P
to zero, which contradicts the Lewis and Randall statement.

To develop this result more explicitly, we let the heat capacities at temperatures T and zero be C (T ) and C (0), respectively. P P
Since C (T ) > 0 for any T > 0 , we have S (T ) − S (T ) > 0 for any T > T > 0 . Since the entropy is always finite,
P
∗ ∗
∞ > S (T ) − S (T ) > 0 , so that

∗
∗
∞ > lim [S (T ) − S (T )] > 0
∗
T →0
and
T
CP
∞ > lim ∫ dT > 0
∗
T →0
T
∗ T
For temperatures in the neighborhood of zero, we can expand the heat capacity, to arbitrary accuracy, as a Taylor series polynomial
in T :
2
∂ CP (0) 1 ∂ CP (0)
2
CP (T ) = CP (0) + ( ) T + ( ) T +…
2
∂T 2 ∂T
P
P
The inequalities become
2
T ∂ CP (0) 1 ∂ CP (0)
∗ ∗ 2
∞ > lim { CP (0) ln +( ) (T − T )+ ( ) (T − T ) + …} > 0
∗ ∗ 2
T →0 T ∂T 4 ∂T
P
P
The condition on the left requires C P (0) = 0 .

We could view the third law as a statement about the heat capacities of pure substances. We infer not only that CP > 0 for all
T > 0 , but also that
CP
lim ( ) =0
T →0 T
More generally, we can infer corresponding assertions for closed reversible systems that are not pure substances: (∂H /∂T ) > 0 P
for all T > 0 , and lim T

T →0
−1
(∂H /∂T ) = 0 . (The zero-temperature entropies of such systems are not zero, however.) In the
P
discussion below, we describe the system as a pure substance. We can make essentially the same arguments for any system; we
need only replace C by (∂H /∂T ) . The Lewis and Randall statement asserts that the entropy goes to a constant at absolute zero,
P P
irrespective of the values of any other thermodynamic functions. It follows that the entropy at zero degrees is independent of the
value of the pressure. For any two pressures, P and P , we have S (P , 0) − S (P , 0) = 0 . Letting P = P and P = P + ΔP
1 2 2 1 1 2
and, we have
S (P + ΔP , 0) − S (P , 0)
=0
ΔP
for any ΔP . Hence, we have

∂S
( ) =0
∂P
T =0
In Chapter 10, we find (∂S/∂P )T = −(∂V /∂T )P , so both the entropy and the volume approach their zero-temperature values
asymptotically.
When we say that absolute zero is unattainable, we mean that no system can undergo any change in which its final temperature is
zero. To see why absolute zero must be unattainable, let us consider processes that can decrease the temperature of a system. In
general, we have heat reservoirs available at various temperatures. We can select the available reservoir whose temperature is
lowest, and bring the system to this temperature by simple thermal contact. This much is trivial; clearly, the challenge is to decrease
the temperature further. To do so, we must effect some other change. Whatever this change may be, it cannot be aided by an
exchange of heat with the surroundings. Once we have brought the system to the temperature of the coldest available portion of the
surroundings, any further exchange of heat with the surroundings can only be counter-productive. We conclude that any process
suited to our purpose must be adiabatic. Since an adiabatic process exchanges no heat with the surroundings, ΔS^ = 0 .
The process must also be a possible process, so that ΔS + ΔS^ ≥ 0 , and since it is adiabatic, ΔS ≥ 0 . Let us consider a reversible
process and an irreversible process in which the same system goes from the state specified by P and T to a second state in
2
1 1
which the pressure is P . The final temperatures and the entropy changes of these processes are different. For the reversible
2
process, ΔS = 0 ; we designate the final temperature as T . For the irreversible process, ΔS > 0 ; we designate the final
2
temperature as T . As it turns out, the temperature change is less for the irreversible process than for the reversible process; that is,
2
∗
T −T < t −t
2 1
∗
2 1 >. Equivalently, the reversible process reaches a lower temperature: T < t >. From 2
∗
2
CP ∂V
dS = dT − ( ) dP
T ∂T
P
we can calculate the entropy changes for these processes. For the reversible process, we calculate
rev
ΔS = S (P2 , T2 ) − S (P1 , T1 )
To do so, we first calculate
(ΔS) = S (P2 , T1 ) − S (P1 , T1 )

T
for the isothermal reversible transformation from state P , T to the state specified by P and T . For this step, dT is zero, and so
1 1 2 1
P2
∂V
(ΔS) T = ∫ ( ) dP
P1
∂T
P
We then calculate
(ΔS) P = S (P2 , T2 ) − S (P2 , T1 )
for the isobaric reversible transformation from state P , T to state P , T . For this transformation, dP is zero, and
2 1 2 2
T2
CP
(ΔS) = −∫ dT
P
T1
T
Then,
T2 P2
rev
CP ∂V
ΔS = S (P2 , T2 ) − S (P1 , T1 ) = ∫ dT − ∫ ( ) dP = 0
T1
T P1
∂T
P
Because ΔS = 0 , the reversible process is unique; that is, given

rev
P1 , T1 , and P2 , the final temperature of the system is
determined. We find T from 2
T2 P2
CP ∂V
∫ dT = ∫ ( ) dP
T1
T P1 ∂T P
To understand the entropy change for the irreversible process, we note first that there are an infinite number of such processes.
There is nothing unique about the final temperature. Given P , T , and P , the final temperature, T , can have any value
1 1 2
∗
2
consistent with the properties of the substance. To specify a particular irreversible process, we must specify all four of the
quantities P , T , P , and T . Having done so, however, we can calculate the entropy change for the irreversible process,
1 1 2 2
∗
irrev ∗
ΔS = S (P2 , T ) − S (P1 , T1 ) > 0
2
by computing the entropy changes as we reversibly carry the system along the isothermal two-step path from P , T to P , T and 1 1 2 1
then along the isobaric path from P , T to P , T . The calculation of ΔS

2 1 2 2
∗
for this reversible path from P , T to P , T irrev
1 1 2
∗
2
employs the same logic as the calculation, in the previous paragraph, of ΔS for the reversible path from P , T to P , T . The 1 1 2 2
difference is that T replaces T as the upper limit in the temperature integral. The pressure integral is the same. We have
∗
2 2
∗
T P2
irrev
2
CP ∂V
∗
ΔS = S (P2 , T ) − S (P1 , T1 ) = ∫ dT − ∫ ( ) dP > 0
2
T1
T P1
∂T
P
From ΔS irrev
> ΔS
rev
, we have
∗
T T2
2
CP CP
∫ dT > ∫ dT
T1
T T1
T
Since the integrands are the same and positive, it follows that T 2
∗
> T2 , as asserted above.
Figure 6. Temperature versus entropy for spontaneous and reversible processes.
Figure 6 shows the relationships among the various quantities discussed in this argument. In the first instance, Figure 6 shows a
plot of two of the system’s isobars in temperature—entropy space. That is, the line labeled P = P depicts the set of temperature— 1
entropy points at which the equilibrated system has pressure P ; the line labeled P = P , depicts the equilibrium positions at
1 2
pressure P . Other lines in this sketch represent paths along which the system can undergo reversible changes at constant entropy
2
or constant temperature. The dotted line represents the irreversible process in which the system goes from the state specified by P , 1
T to the state specified by P , T . This line is dotted to represent the fact that the system’s temperature may not be well defined
∗
1 2 2
during the irreversible process.

Effective cooling can be achieved using pressure changes if the system is a gas. However, for liquids and solids, (∂V /∂T ) is P
small; consequently, the temperature change for a reversible pressure change is also small. At temperatures near absolute zero,
nearly all substances are solid; to achieve effective cooling we must change a thermodynamic variable for which a solid’s
temperature coefficient is as large as possible. To consider the general problem of decreasing the temperature of a system by
varying something other than pressure, we must consider a system in which some form of non-pressure–volume work is possible.
Such a system is subject to an additional force, and its energy changes as this force changes.
Adiabatic Demagnetization
The practical method by which extremely low temperatures are achieved is called adiabatic demagnetization. This method exploits
the properties of paramagnetic solids. In such solids, unpaired electrons localized on individual atoms give rise to a magnetic
moment. Quantum mechanics leads to important conclusions about the interaction between such magnetic moments and an applied
magnetic field: In an applied magnetic field, the magnetic moment of an individual atom is quantized. In the simplest case, it can
be aligned in only two directions; it must be either parallel or anti-parallel to the applied magnetic field. When an atom’s magnetic
moment is parallel to the magnetic field the energy of the system is less than when the alignment is anti-parallel. The applied
magnetic field exerts a force on the magnetic moments associated with individual atoms. The energy of the system depends on the
magnitude of the applied magnetic field.
Rather than focus on the particular case of adiabatic demagnetization, let us consider the energy and entropy changes associated
with changes in a generalized potential, Φ , and its generalized displacement, θ . (For adiabatic demagnetization, θ would be the
θ
applied magnetic field.) Three variables are required to describe reversible changes in this system. We can express the energy and
entropy as functions of temperature, pressure, and θ :
E = E (T , P , θ) and S = S (T , P , θ) . The total differential of the entropy includes a term that specifies the dependence of
entropy on θ . We have
∂S ∂S ∂S C (T , P , θ) ∂V ∂S
dS = ( ) dT + ( ) dP + ( ) dθ = dT − ( ) dP + ( ) dθ
∂T ∂P ∂θ T ∂T ∂θ
P ,θ T ,θ T ,P P ,θ T ,P
where we write C (T , P , θ) to emphasize that our present purposes now require that we measure the heat capacity at constant
pressure and constant θ .
For constant pressure, P, and constant displacement, θ , the entropy depends on temperature as
T T
∂S C (T , P , θ)
S (T , P , θ) = S (0, P , θ) + ∫ ( ) dT = S (0, P , θ) + ∫ dT
0
∂T 0
T
P ,θ
The postulate that the entropy be finite at any temperature implies that the pressure- and θ -dependent heat capacity becomes zero at
absolute zero. That is, at absolute zero, the heat capacity vanishes whatever the values of P and θ . The argument is exactly the same
as before. Earlier, we wrote C (0) = 0 ; for the present generalized case, we write C (0, P , θ) = 0 .
P
Similarly, from the postulate that the entropy goes to a constant at absolute zero for all values of the other thermodynamics
variables, it follows that, for any two pressures P and P , and for any two values of the generalized displacement, θ and θ ,
1 2 1 2
S (0, P1 , θ1 ) = S (0, P2 , θ1 ) = S (0, P1 , θ2 ) = S (0, P2 , θ2 ) = 0
and hence that

∂S ∂S (0, P , θ)
( ) =( ) =0
∂P ∂P
T =0,θ T ,θ
and
∂S ∂S (0, P , θ)
( ) =( ) =0
∂θ T =0,P
∂θ T ,P
We want to consider a process in which a system goes from the lowest temperature available in the surroundings to a still lower
temperature. To minimize the final temperature, this process must be carried out adiabatically. It must also be a possible process, so
that dS ≥ 0 . For simplicity, let us now assume that we carry out this process at a constant pressure, P , and that the system goes
from the state specified by P , T , θ to the state specified by P , T , θ where T > T . The entropies of these two states are
1 1 2 2 1 2
T1
C (T , P , θ1 )
S (T1 , P , θ1 ) = S (0, P , θ1 ) + ∫ dT
0 T
and
T2
C (T , P , θ2 )
S (T2 , P , θ2 ) = S (0, P , θ2 ) + ∫ dT
0
T
The entropy change for this process is
S (T2 , P , θ2 ) − S (T1 , P , θ1 ) = S (0, P , θ2 ) − S (0, P , θ1 )
T2 T1
C (T , P , θ2 ) C (T , P , θ1 )
+∫ dT − ∫ dT ≥ 0
0
T 0
T
Now, let us suppose that the final temperature is zero; that is, T 2 =0 , so that
T2
C (T , P , θ2 )
∫ dT = 0
0 T
It follows that
T1
C (T , P , θ1 )
S (0, P , θ2 ) − S (0, P , θ1 ) ≥ ∫ dT > 0
0
T
where the inequality on the right follows from the fact that that C (T , P , θ 1) >0 . Then, it follows that
S (0, P , θ2 ) − S (0, P , θ1 ) > 0
which contradicts the Lewis and Randall statement of the third law. The assumption that the system can reach absolute zero leads to
a contradiction of the Lewis and Randall statement of the third law. Therefore, if the Lewis and Randall statement is true, absolute
zero is unattainable.
The converse applies also; that is, from the proposition that absolute zero is unattainable, we can show that the Lewis and Randall
statement is true. To do so, we rearrange the above equation for ΔS ,
T2
C (T , P , θ2 )
∫ dT ≥
0 T
T1
C (T , P , θ1 )
∫ dT − S (0, P , θ2 ) + S (0, P , θ1 )
0
T
If we now assume that the Lewis and Randall statement is false, the expression on the right can be less than or equal to zero. The
integral on the left can then be zero, in which case the system can reach absolute zero. If the Lewis and Randall statement is false, it
is true that the system can reach absolute zero. Therefore: If the system cannot reach absolute zero, the Lewis and Randall
statement is true.
Figures 7 and 8 depict these ideas using contour plots in temperature–entropy space. Each figure shows two contour lines. One of
these contour lines is a set of temperature and entropy values along which the pressure is constant at P and θ is constant at θ . The 1
other contour line is a set of temperature and entropy values along which the pressure is constant at P and θ is constant at θ . The 2
slope of a contour line is

∂T T
( ) =
∂S C (T , P , θ)
P ,θ
Because the heat capacity is always positive, this slope is always positive.
Figure 7. Reversible change in a system that satisfies the Lewis and Randall statement.
In Figure 7, the Lewis and Randall statement is satisfied. When the temperature goes to zero, the contour lines meet at the same
value of the entropy; these contours satisfy the relationship
S (0, P , θ1 ) = S (0, P , θ2 )
An adiabatic (vertical) path from the contour for P and θ meets the contour for P and θ at a positive temperature, T > 0 . Since
1 2 2
this is evidently true for any P and any θ , the final state for any adiabatic process will have T > 0 . Because the Lewis and
2 2
Randall statement is satisfied, the system cannot reach absolute zero, and vice versa.
Figure 8. Reversible change in a system that does not satisfy the Lewis and Randall statement.
In Figure 8, the Lewis and Randall statement is violated, because we have S (0, P , θ ). In this case, an adiabatic process initiated
1
from a low enough initial temperature, T , will reach absolute zero without intersecting the contour for constant P and θ . Because
1 2
the Lewis and Randal statement is violated, the system can reach absolute zero, and vice versa.
This page titled 11.13: Absolute Zero is Unattainable is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
11.14: Problems
¯
¯¯¯
¯ ¯¯
¯
1. The relationships between H
o
, S
o
, and G
o
for the standard state of a gas and the molar enthalpy, H (P ) , entropy, S (P ) , and
¯
¯¯¯
Gibbs free energy, G (P ), of the real gas at pressure P and temperature T involve several integrals. Given the virial equation for a
real gas,
¯
¯¯¯
PV ∗ ∗ 2 ∗ 3
Z = = 1 +B (T ) P + C (T ) P +D (T ) P +…
RT
evaluate the following:

(a)
P ¯
¯¯¯
¯
¯¯¯
∂V
∫ [V − T ( ) ] dP
0 ∂T
P
(b)
P ¯
¯¯¯
V 1
∫ [ − ] dP
0
RT P
(c)
P ¯
¯¯¯ 1 bar
∂V R
∫ ( ) dP − ∫ dP
P
∗
∂T P
∗ P
P
2. Why does (∂S/∂P ) = − (∂V /∂T )

T =0
imply that both the entropy and the volume approach their zero-temperature
P , T =0
values asymptotically? Is this consistent with defining absolute zero to be the temperature at which the volume of an ideal gas
extrapolates to zero—at constant pressure?
3. Prove that
∗
P ¯
¯¯¯ 1 bar
¯¯
¯ ∂V R
0
S A (P ) − S =∫ ( ) dP + ∫ dP
A
P
∂T P
∗ P
P
remains finite for any gas in the limit as P ∗

→ 0 . Hint: Express the integral from P to P as the sum of integrals from P to 1 bar ∗ ∗
and from 1 bar to P .

4. Let A , B , C , and D be elements, whose absolute entropies at 1 bar and temperature T are S (A, T ) , S (B, T ) , S (C , T ), o o o
and S (D, T ), respectively. Let AB, C D, AC , and BD be binary compounds of these elements, and represent their absolute
o
entropies at these conditions by S (AB, T ) , S (C D, T ), S (AC , T ), and S (BD, T ) .

o o o o
(a) What is the entropy change, Δ rS

o
(T ) , for the reaction AB + C D → AC + BD ?
(b) What are Δ f S
o
(AB, T ) ,Δ f S
o
(C D, T ) Δf S , o
(AC , T ) , and Δ f S
o
(BD, T ) ?
(c) Show that
o o o o o
Δr S (T ) = Δf S (AC , T ) + Δf S (BD, T ) − Δf S (AB, T ) − Δf S (C D, T )
5. Let A and B be elements; let A B be a binary compound of these elements. At temperature T and 1 bar, let the entropy of
a b
these substances be S (A, T ) , S (B, T ) , and S (A B , T ), respectively. At absolute zero and 1 bar, let these entropies be
o o o
a b
S (A, 0) , S (B, 0) , and S (A B , 0). Represent the change in entropy when these substances are warmed from 0 K to T , at a
o o o
a b
constant pressure of 1 bar, as Δ S (A) , Δ S (B) , and Δ

0→T S (A B ). These quantities are related by the following
o
0→T
o
0→T
o
a b
equations:
o o o
S (A, T ) = S (A, 0) + Δ0→T S (A)
o o o
S (B, T ) = S (B, 0) + Δ0→T S (B)
o o o
S (Aa Bb , T ) = S (Aa Bb , 0) + Δ0→T S (Aa Bb )
By the Nernst Heat Theorem, the entropy change for formation of A a Bb at absolute zero must be zero:
o o o o
Δf S (Aa Bb , 0) = S (Aa Bb , 0) − a S (A, 0) − b S (B, 0)
When, following Planck and Lewis and Randall, we choose to let the entropies of the elements be zero at absolute zero, the entropy
of A B and its entropy of formation also become zero at absolute zero:
a b
o o
Δf S (Aa Bb , 0) = S (Aa Bb , 0) = 0
Suppose that we decide to create an alternative set of absolute entropies by assigning non-zero values to the entropies of the
elements at absolute zero. Let us distinguish entropy values in this new scheme with a tilde. Then, the non-zero values that we
~ ~ o o
assign to the elements at absolute zero are S (A, 0) ≠ 0 and S (B, 0) ≠ 0 . By the Nernst Heat Theorem, we have
~o ~o ~o ~o
Δf S (Aa Bb , 0) = S (Aa Bb , 0) − a S (A, 0) − b S (B, 0)
Evidently, we have
~o o
Δf S (Aa Bb , 0) = Δf S (Aa Bb , 0) = 0
~o ~o ~o
and, since the values of S (A, 0) and S (B, 0) are arbitrary, we can choose them so that S (Aa Bb , 0) is non-zero also
~o ~o ~o
S (Aa Bb , 0) = aS (A, 0) + b S (B, 0) ≠0
~o ~o ~o
(a) What are S (A, T ) ,S (B, T ) , and S (Aa Bb , T ) ?
~o
(b) What is Δ f S (Aa Bb , T ) ?
~o
(c) Show that o
Δf S (Aa Bb , T ) = Δf S (Aa Bb , T ) .
(d) Consider a reaction in which A B is a reactant or a product. How will the alternative choice of values for the entropies of A
a b
and B at absolute zero affect the values we calculate for Δ S , Δ H , Δ G ? r

o
r
o
r
o
~o ~o
(e) Can you think of any circumstance in which there would be an advantage to choosing S (A, 0) ≠ 0 and S (B, 0) ≠ 0 ?
¯
¯¯¯
6. Find ln γ for a gas that obeys the equation of state P (V − b) = RT . For C O2 at 300 K, the value of b (the second virial
coefficient) is −1.26 × 10 −4
m
3
mol
−1
.
Calculate the fugacity coefficient and the fugacity of C O at 300 K and pressures of 1, 10, and 100 bar.
2
7. Consider the following sequence of steps that convert a van der Waals gas, vdwg, at an arbitrary pressure to the corresponding
hypothetical ideal gas in its standard state.
ΔI G
(I) A (vdwg, P , T ) A (vdwg, P
∗
≈ 0, T )
→
ΔII G ΔIII G
(II) A (vdwg, P ∗
≈ 0, T ) A (ideal gas, P
∗
≈ 0, T ) (III) A (ideal gas, P ∗
≈ 0, T )
o
A (H I G , T )
→ →
Show that the fugacity of the van der Waals gas is given by
b 2a RT
ln fvdw = − + ln
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
V −b RT V V −b
Hint: Find ΔG = Δ I G + ΔII G + ΔIII G . To calculate Δ I G , use integration by parts:

∗
∗ ¯
¯¯¯
P ∗ V
P
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
ΔI G = ∫ V vdw dP = [P V ] −∫ P dV
¯
¯¯¯
P
P V
∗
¯
¯¯¯
When P ∗
→ 0 ,V → ∞ .
8. The normal boiling point of methanol is 337.8 K at 1 atm. The enthalpy of vaporization at the normal boiling point is
Δ H = 35.21 kJ mol
vap . −1
(a) What are Δ vap G and Δ vap S for methanol at its normal boiling point?
(b) At 1 bar, the heat capacity of gaseous methanol depends on temperature as
−1 −1
CP (C H3 OH , g, 1 bar) = 21.737 + 0.07494 T [J K mol ]
Assume that the heat capacity at 1 atm is the same as it is at 1 bar. Calculate the enthalpy change and the entropy change when one
mole of gaseous methanol is heated from the normal boiling point to 500 K at 1 atm.
(c) At 1 bar, the absolute entropy of gaseous methanol depends on temperature as
o −5 2 −1 −1
S (C H3 OH , g, 1 bar) = 192.8 + 0.1738 T − (5.367 × 10 )T [J K mol ]
Assume that the absolute entropy at 1 atm is the same as it is at 1 bar. Calculate the Gibbs Free Energy change, ΔG , when one
mole of gaseous methanol is heated from the normal boiling point to 500 K at a constant pressure of 1 atm.
(d) The heat capacity of liquid methanol,
CP (C H3 OH , liq, 298.15 K, 1 bar)
is 1.1 J K mol . Assume that this heat capacity remains constant for superheated liquid methanol. Calculate the enthalpy and
−1 −1
entropy changes when liquid methanol is heated from the normal boiling point to 500 K.
(e) The molar entropy of liquid methanol at 1 bar and 298.15 K is 126.8 J K mol . We can estimate the molar entropy of
−1 −1
superheated liquid methanol by using the heat capacity at 298.15 K to estimate the entropy at higher temperatures:
T
o o
CP
S (T ) ≈ S (298.15 K) + ∫ dT
298.15 T
Find S o
(T ) and use this equation to calculate ΔG for heating liquid methanol from the normal boiling point to 500 K.
(f) Devise a cycle that enables you to use the results you obtain in parts (a)–(e) to calculate Δ vap H ,
(500 K) Δvap S (500 K) , and
Δ vapG (500 K) when one mole of methanol vaporizes at 1 atm and 500 K.
(g) Use the values you obtain for Δ vap H (500 K) and Δ vap S (500 K) in part (f) to calculate Δ vap G (500 K) .
(h) Use the Gibbs-Helmholtz equation to calculate the Gibbs free energy change, Δ G (500 K), when one mole of methanol vap
vaporizes at 1 atm and 500 K. Compare this result to those you obtained in parts (f) and (g). Is this process impossible,
spontaneous, or reversible?
(i) How much heat is taken up by the system when one mole of methanol is vaporized reversibly at its normal boiling point and the
resulting vapor is heated reversibly at 1 atm to 500 K? How much work is done on the system in this process? What are ΔE and
ΔS for this process?
(j) Suppose that the change of state in part (i) is effected irreversibly by contacting the liquid methanol with surroundings at 500 K,
while maintaining the applied pressure constant at 1 atm. How much work is done on the system in this spontaneous process? How
much heat is taken up by the system? What are ΔS^ and ΔS for this spontaneous process?
universe
9. To maximize the temperature change for a given change in system pressure, the value of | (∂T /∂P ) |
S
should be as large as
possible. If only pressure–volume work is possible, we have
CP ∂V
dS = dT − ( ) dP
T ∂T
P
Show that
∂T T ∂V
( ) = ( )
∂P CP ∂T
S P
What happens to T /C as the temperature approaches absolute zero? For condensed phases, we find (∂V /∂T )
P P
≪ V . Consider
the behavior of the molecules in a lattice as the temperature approaches absolute zero. Is it reasonable to expect
∂V
limT →0 ( ) =0
∂T
P
Why?
Notes
1
Lewis, G.N., Randall, M., K. S. Pitzer, and L. Brewer, Thermodynamics, 2 nd
Edition, McGraw-Hill, New York, 1981, p 130.
2
More precisely, we consider a reversible process and a spontaneous process whose initial state is a “change-enabled”
modification of the reversible-process initial state. The state functions are the same in both initial states.
CHAPTER OVERVIEW
12: Applications of the Thermodynamic Criteria for Change

The equations we derive in Chapters 9 and 10 are the core of chemical thermodynamics. However, we have yet to deal with the
effects of changing the concentrations of the substances present in the system. To apply our theory to chemical changes, we must
extend our theory so that it can model these effects. In this chapter, we consider some basic applications that do not involve
chemical reactions and in which both intermolecular interactions and the effects of mixing can be ignored. In Chapters 13-16, we
develop the application of thermodynamic concepts to processes in which a chemical reaction occurs. We do so in two steps. In
Chapter 13, we consider an approximation in which the properties of a multi-component system are determined by the effects of
mixing pure substances whose molecules neither attract nor repel one another. In Chapter 14, we begin to consider the general case
in which intermolecular interactions can be important.
12.10: Problems
This page titled 12: Applications of the Thermodynamic Criteria for Change is shared under a CC BY-SA 4.0 license and was authored, remixed,
1
When we talk about a purely mechanical process, we have in mind a system in which one or more unchanging objects can move
relative to some reference frame. Their movements are described completely by Newton’s laws of motion. The objects are
characterized by their masses, locations, velocities, and accelerations. They may be subject to the effects of force fields, whose
magnitudes can vary with location and time. We stipulate that the volume, pressure, temperature, entropy, composition, and the
internal energy, U , of an individual object remain constant. Since entropy and volume are extensive state functions, we can obtain
the entropy and volume of the objects in aggregate by summing up those for the individual objects. Moreover, the total entropy and
the total volume are constants.
The energy of a purely mechanical system is the sum of its internal energy, U , its kinetic energy, τ , and its mechanical potential
energy, υ ; that is,
E = U + τ + υ.
U is the sum of the internal energies of the unchanging constituent objects. Since it is constant, the internal energy can be given an
arbitrary value that we usually choose to be zero. When we do so, the energy of the system becomes the sum of its kinetic and
potential energies. Noting explicitly that the entropy and volume are constant, we have
(ΔE) = τ + υ.
SV
The essential distinction between a purely mechanical system and a thermodynamic system is that our models for mechanical
systems focus on the motions of unchanging objects; our models for thermodynamic systems focus on the internal changes of
stationary objects.
An important aspect of this distinction is that our definitions of equilibrium, reversibility, and spontaneous change in mechanical
systems are not wholly congruent with the definitions we use in developing the principles of thermodynamics. Thus, an equilibrium
state of a mechanical system is one in which the objects comprising the system are stationary with respect to some reference frame.
For a mechanical system at equilibrium, the kinetic energy is constant and can usually be taken to be zero, τ = 0 . The motion of a
system generally has no bearing on whether the substances that comprise the system are at equilibrium in a thermodynamic sense.
We normally consider that a sufficient condition for a mechanical system to be reversible is that—following some excursion to
other states—the initial conditions of both the system and the surroundings can be restored exactly . In our thermodynamic view of
1
reversibility, this condition is necessary but not sufficient: We say that a thermodynamic system is reversible only if the direction in
which it is changing can be reversed at any time by an arbitrarily small change in its interaction with its surroundings. The initial
conditions can be restored after any excursion.
Our treatments of the frictionless harmonic oscillator illustrate this imperfect congruence. Viewed as a mechanical system, a
frictionless harmonic oscillator is a reversible system. If we adopt the view that it is continuously undergoing spontaneous change,
our thermodynamic principles mean that its entropy is continuously increasing. However, since its state and that of its surroundings
reproduce themselves exactly after every period of oscillation, our thermodynamic principles mean also that there is zero net
change in the entropy over any complete oscillation. Clearly, this is a contradiction. If we attempt to salvage the situation by
supposing that the entropy of an isolated freely moving harmonic oscillator is constant, our thermodynamic principles require us to
say that it is at equilibrium. This contradicts the view that a mechanical system is at equilibrium only if its kinetic energy is zero.
Since neither formulation is satisfactory, we recognize that we cannot expect to describe every mechanical system in purely
thermodynamic terms.
Nor can we expect to describe every thermodynamic system in mechanical terms. This becomes obvious when we observe that the
second law of thermodynamics is essential to our description of thermodynamic systems, but it is not among the principles of
mechanics. Beginning in Chapter 20, we find that we can model the thermodynamic properties of a system that is itself a collection
of a large number of subsystems by focusing on the average values of the properties of the subsystems. The laws of motion model
the movements of the individual particles of a system. The laws of thermodynamics model the average properties of the particles in
a system that contains a very large number of particles. While we cannot usefully describe an individual harmonic oscillator as a
thermodynamic system, we see in §22-6 that the thermodynamic properties of a system composed of many identical harmonic
oscillators can be modeled very successfully.
In short, mechanics and thermodynamics model different kinds of systems from fundamentally different perspectives. Nevertheless,
when we limit our consideration of mechanical systems to the prediction of spontaneous change from one equilibrium state to
another, we can recognize that the criteria we are accustomed to apply to mechanical systems are analogous to our thermodynamic
criteria as they apply to such processes.
From our thermodynamic perspective, a purely mechanical process involves no change in the entropy or volume of the system, and
the criterion for irreversible change is (ΔE) < w >. From our mechanical perspective, an irreversible transition between
SV npv
equilibrium states is one in which various objects interact with one another or with various force fields. We consider that the
process can occur if the change in the potential energy of the system is less than the work done on it; that is Δυ < w >. In the
npv
mechanical system, some of the work done on the system is dissipated by frictional forces as heat that appears in the surroundings.
(If the process involves no exchange of work with the surroundings, the criterion becomes Δυ < 0 .) If the mechanical system
begins and ends at rest, we have (ΔE) < Δυ , so that again (ΔE) < w .
SV SV npv
If no work is exchanged with the surroundings, potential energy minimization, Δυ < 0 , is a sufficient condition for spontaneous
change to be possible in mechanical systems under the circumscribed conditions we have outlined. In Section 14-2, we find that
minimization of the chemical potential energy,
ω
∑ μ dnj < 0
j
j=1
is a necessary and sufficient condition for spontaneous change to be possible in a thermodynamic system. These conditions are
parallel, but they are not equivalent to one another.
This page titled 12.1: Mechanical Processes is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
The idea that thermal energy can be transferred from a warmer body to a colder one, but not in the opposite direction, is a
fundamental assumption in our development of the thermodynamic criteria for change. Therefore, if our theory is to be internally
consistent, we must be able to deduce this principle from the criteria we have developed. Let us consider one way in which this can
be done: We consider an isolated system comprised of two subsystems, A and B , that are in thermal contact with one another. We
suppose that the temperatures are T and T and that T ≠ T . If the energy of subsystem A increases, heat transfers from
A B A B
subsystem B to subsystem A . In this case, we know that q > 0 and T > T .

A B A
When we seek to analyze this process using our thermodynamic theory, we encounter a problem that arises for any spontaneous
process: Since the process is not reversible, we must introduce approximating assumptions. For the present analysis, we want to
estimate the entropy change that occurs in each subsystem. To do so, we suppose that an increment of heat, dq, can pass from one
subsystem to the other without significantly changing the temperature of either one. It is evident that we could—by some other
process—effect this change in either subsystem as nearly reversibly as we wish. (In §5, we consider such a process.) Even though
the present process is not reversible, we have good reason to assume that the entropy changes in the subsystems are well
approximated as dq /T and dq /T = −dq /T . Since the system is isolated, the process can be spontaneous only if its
A A B B A B
entropy change is positive; that is, the relevant thermodynamic criterion is dq /T − dq /T > 0 . With dq > 0 , we find
A A A B A
1/ TA − 1/ T B> 0 , or T >TB . When heat is spontaneously transferred from B to A , our thermodynamic criterion also requires
A
that subsystem B be warmer than subsystem A .
This page titled 12.2: The Direction of Spontaneous Heat Transfer is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Let us consider processes in which transfer of heat from the surroundings melts one mole of ice. We suppose that the ice is initially
at 0 ºC and one bar. At these conditions, the enthalpy change for melting a mole of ice is 6010 J. If the ice melts reversibly at these
conditions, the temperature of the surroundings is also 0 ºC. As it melts, the ice takes up 6010 J of heat, which is given up by the
surroundings. For this process, we have q = Δ H = 6010 J . The temperature is constant, and the entropy change for the
rev
P f us
system is
rev
q Δf us H
P
ΔS = =
T T
Since q^ = −q rev
P
, we have
rev
−q −Δf us H
^ P
ΔS = =
^ ^
T T
so that
−1
ΔS = 6010 J/273.15 K = 22.00 J K
and
^ −1
ΔS = −6010 J/273.15 K = −22.00 J K
As required for a reversible process, we have ΔS + ΔS^ = 0 . The Gibbs free energy change is
rev rev
(ΔG) = (ΔH ) − T (ΔS) =q − T (q /T ) = 0
PT PT PT P P
which is also as required for a reversible process.

Now let us consider a spontaneous process, in which the ice melts while in thermal contact with surroundings at 10 ºC. To reach
equilibrium, the system must reach the temperature of the surroundings, which we assume to be constant. In this process, the ice
melts and the melt water warms to 10 ºC. To find the entropy change, we must find a reversible process that effects the same
change. A two-step process effects this conveniently. The first step is the one we have just considered: Surroundings at 0 ºC
transfer 6010 J of heat to the system, reversibly melting the ice to water at 0 ºC. We have q = 6010 J and 1
−1
ΔS1 = 22.00 J K
In the second reversible step, surroundings that are always at the same temperature as the system transfer heat to the system as the
temperature increases from 273.15 K to 283.15 K. The heat capacity of liquid water is 75.3 J mol . For this step, −1
283.15
−1
q2 = ∫ CP dT = (75.3 J K ) (10 K) = 753 J
273.15
and
283.15
CP −1
283.15 −1
ΔS2 = ∫ dT = (75.3 J K ) ln = 2.71 J K
273.15
T 273.15
For these reversible changes in the system, we have ΔS = ΔS + ΔS = 24.71 J K . This is also the value of ΔS for the
1 2
−1
spontaneous process. The heat taken up by the system in the two-step reversible process is q = q + q = 6763 J . This heat is 1 2
surrendered by the surroundings, and we could effect identically the same change in the surroundings by exchanging this quantity
of heat reversibly. For the spontaneous process, therefore, we have q^ = −6763 J and
^ −1
ΔS = −6763 J/283.15 K = −23.88 J K
For the universe, we have

^ −1
ΔSuniverse = ΔS + ΔS = +0.83 J K
which is greater than zero, as required for a spontaneous process.
Because this reversible two-step process does not occur at a constant temperature, its Gibbs free energy change is not zero.
However, we can use the Gibbs-Helmholtz equation to estimate the Gibbs free energy change for the related process in which ice at
10 ºC and 1 bar (a hypothetical substance) melts to form liquid water at the same temperature and pressure. For this process, we
estimate
−1 −1
ΔG = −220 J mol K
(See problem 10-21.) Since we have ΔG < 0 for the process, our change criterion asserts that, in agreement with our experience,
superheated ice melts spontaneously at 10 C.
This page titled 12.3: Phase Changes - the Fusion of Ice is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
We define entropy in terms of its differential as
rev
dS = dq /T .
To measure an entropy change using this definition requires that the process be reversible, which means that the system and
surroundings must be at the same temperature as the heat transfer occurs. We understand reversible heat transfer to be the limiting
case in which the temperature difference between system and surroundings becomes arbitrarily small. Here we want to consider the
conceptual problems associated with keeping the temperature of the surroundings arbitrarily close to the temperature of the system
while the system undergoes an arbitrary reversible change, which may include a temperature change.
We can illuminate one necessary aspect by posing a trivial dilemma: Since system and surroundings jointly comprise the universe,
the requirement that system and surroundings be at the same temperature might seem to require that the entire universe be at a
single temperature. Plainly, this condition is not met; the temperature of the universe varies from place to place. In fact, the
requirement is only that the system and that portion of the surroundings with which the system exchanges heat be at the same
temperature. We can satisfy this requirement by permitting the exchange of heat between the system and a portion of the
surroundings, under conditions in which the combination of the two is thermally isolated from the rest of the universe.
The non-trivial aspect of this problem arises from the requirement that the temperature of the surroundings remain arbitrarily close
to the temperature of the system, while both temperatures change and heat is exchanged between the system and the surroundings.
A clumsy solution to this problem is to suppose that we exchange one set of surroundings (at temperature T ) to a new set (at
temperature T + ΔT ) whenever the temperature of the system changes by ΔT . A more elegant solution is to use a machine that
can measure the entropy change associated with an arbitrary reversible change in any closed system. This is a conceptual device,
not a practical machine. We can use it in gedanken experiments to make arbitrarily small changes in the temperature and pressure
of the system along any reversible path. At every step along this path, the entropy change is
1 ∂H ∂V
dS = ( ) dT − ( ) dP
T ∂T ∂T
P P
Figure 1. An entropy-measuring machine.

The entropy-measuring machine is sketched in Figure 1. In this device, the portion of the surroundings with which the system can
exchange heat is a quantity of ideal gas, which functions as a heat reservoir. This heat reservoir is in thermal contact with the
system. The combination of system and ideal-gas heat reservoir is thermally isolated from the rest of the universe. We consider the
case in which only pressure–volume work can be done on either the system or the ideal-gas heat reservoir. In this device, all
changes are driven by changes in the pressures applied to the surroundings (the ideal-gas heat reservoir) and the system. The
pressure applied to the system and the pressure applied to the ideal-gas heat reservoir can be varied independently. We suppose that
the system is initially at equilibrium and that changes in the applied pressures are effected in such a manner that all changes in the
system and in the ideal-gas heat reservoir occur reversibly. For any change effected in the entropy-measuring machine, the heat and
entropy changes in the heat reservoir are q^ and ΔS^.
Isothermal Process
Let us consider first a process in which a quantity of heat must be transferred from the surroundings to the system while both are at
the constant temperature T . To be specific, let this be a process in which a mole of pure liquid vaporizes at constant pressure,
P
taking up a quantity of heat equal to the molar enthalpy of vaporization. We can supply this heat by reversibly and isothermally
compressing the ideal-gas heat reservoir. To keep the temperature of the ideal-gas heat reservoir constant, we reversibly withdraw
the piston that controls the pressure of the system, causing the vaporization of liquid in the system and absorption by the system of
the heat given up by the ideal-gas heat reservoir. Overall, we apply forces to the two pistons to achieve reversible isothermal
compression of the ideal gas in the heat reservoir and reversible isothermal vaporization of a quantity of liquid in the system.
Because q = −q^ , the magnitude of the entropy change for the surroundings is equal to that for the ideal-gas heat reservoir. The
entire process is reversible, the entropy change for the system and the entropy change for the surroundings sum to zero:
ΔS + ΔS = 0 .
^
We can calculate the entropy change for the ideal-gas heat reservoir. Overall, the n
^ moles of ideal gas in the heat reservoir go from
the state specified by P^ and T 1 to the state specified by P^ and T . From the general relationship,
P 2 P
(∂H /∂T ) dT − (∂V /∂T ) dP , with dT = 0 , we have dS = − (n ^ R/P ) dP and

−1 ^
dS = T
P P
^ ^
P2 V2
^
ΔS = −n
^ R ln =n
^ R ln
^ ^
P1 V1
The entropy change for the system is ΔS = −ΔS^ . So long as we carry out the process isothermally and reversibly, we can
determine the entropy change for the system simply by measuring the initial and final pressures (or volumes) of the ideal-gas heat
reservoir.
Any Reversible Process

If the system undergoes a reversible change in which the temperature of the system is not constant, we can operate the entropy-
measuring machine in essentially the same manner as before. The only difference is that we must adjust the pressure applied to the
system so that the system temperature changes in the manner required to keep the process reversible—that is, to maintain the
system at equilibrium. Then the change to the system and the change to the ideal-gas heat reservoir both take place reversibly. Even
though the temperatures change, appropriate control of the applied pressures assures that the system is always in an equilibrium
state and that the temperature of the system is always arbitrarily close to the temperature of the ideal-gas heat reservoir.
We can calculate the entropy change for the ideal-gas heat reservoir. Overall, the n^ moles of ideal gas in the heat reservoir go from
the state specified by P and T to the state specified by P and T . We can evaluate the entropy change for taking the ideal gas
^
1
^
1
^ ^
2 2
from state 1 to state 2 by a two-step path. We first compress the gas isothermally at T^ from P^ to P^ . We then warm the gas at
1 1 2
constant pressure P^ from T^ to T^ . For the first step, dT^ = 0 , and, as before, we find
2 1 2
^ ^
P2 V2
^ ^ Rln ^ Rln
ΔS isothermal = −n =n
^ ^
P1 V1
For the second step, dP^ = 0 , and

^
T 2
^ CP ^
n T 2
^ ^ CP ln
ΔS isobaric = ∫ dT = n
^ T ^
T 1 T 1
The entropy change for the ideal-gas heat reservoir is thus

^ ^
P2 T 2
^ ^ ^
ΔS = ΔS isothermal + ΔS isobaric = −n
^ R ln +n
^ CP ln
^ ^
P1 T 1
and we have ΔS = −ΔS^ .

The essential point of the entropy-measuring machine is that we can determine the entropy change for any process without
knowing anything about the process except how to control the system pressure and temperature so that the process occurs
reversibly. Of course, this one reason that the entropy-measuring machine is not a practical device. To control the machine in the
required manner, we must know how the thermodynamic properties of the system are related to one another on the Gibbsian
manifold that defines the system’s equilibrium states. If we know this, then we know (∂H /∂T ) and (∂V /∂T ) for the systemP P
along any reversible path, and we can calculate the entropy change for the system in the same way that we calculate the entropy
change for the ideal-gas heat reservoir. If we have the information needed to perform the measurement, we can calculate the
entropy change without using the machine.
This page titled 12.4: Measuring the Entropy Change for Any Reversible Process is shared under a CC BY-SA 4.0 license and was authored,
When we apply Le Chatelier’s principle, we imagine an equilibrium system on which we impose some step-wise change.
Immediately following the imposition of this change, we isolate the system from all further interactions with its surroundings. This
isolated system is a hypothetical construct, which can be only approximated in any real experiment. It has peculiar features: While
the changed and isolated system has the properties of the original system, it is also free to undergo a further change that the original
system could not. The hypothetical isolated system is no longer at equilibrium; it can undergo a spontaneous process of further
change until it reaches a new position of equilibrium. The principle asserts that this further change opposes the imposed change.
The principle is inherently qualitative. This contributes to its utility in that we do not have to have quantitative data in order to use
it. However, a qualitative prediction is less useful than a quantitative one. Let us now attempt to apply our second-law based
quantitative models to the sequence of changes envisioned by Le Chatelier’s principle. We begin by restating the principle in more
mathematical language. We then illustrate these ideas for the specific case of vapor–liquid equilibrium with temperature and
pressure as the independent variables.
Let us suppose that W , X, Y , and Z are a set of thermodynamic variables that is adequate to specify the state of the system. In any
equilibrium state, the entropy of the system is then a function of these variables; we have S = S (W , X, Y , Z) . For present
purposes, we assume that we know the function S (W , X, Y , Z). Given small changes, dW , dX, dY , and dZ , in the independent
variables, we can find the change in dS for a reversible transition from (W , X, Y , Z) to (W + dW , X + dX, Y + dY , Z + dZ) :
∂S ∂S ∂S ∂S
dS = ( ) dW + ( ) dX + ( ) dY + ( ) dZ
∂W XY Z
∂X WYZ
∂Y W XZ
∂Z W XY
When we impose the change creating the hypothetical isolated system, we imagine that some characteristic of the system changes
instantaneously, and that it does so without changing the other properties of the system. Since we suppose that nothing about the
system changes in the perturbation and isolation step, the entropy of the perturbed, isolated, hypothetical system remains the same
as that of the original equilibrium system.
Figure 2. Spontaneous and reversible processes connecting the same equilibrium states.
Figure 2 shows the entropies for three states in the cycle that comprises the Le Chatelier model for change. The entropy of the
original equilibrium system is S (W , X, Y , Z) and that of the final equilibrium system is
S (W + dW , X + dX, Y + dY , Z + dZ) . The same final equilibrium state is reached by both the irreversible transition from the
change-enabled hypothetical state and by a reversible transition from the initial equilibrium state. Since entropy is a state function,
its change around this cycle must be zero. Hence, the incremental changes dW , dX, dY , and dZ that occur in the thermodynamic
variables must satisfy the inequality
∂S ∂S ∂S ∂S
( ) dW + ( ) dX + ( ) dY + ( ) dZ > 0
∂W ∂X ∂Y ∂Z
XY Z WYZ W XZ W XY
We can view the application of this inequality to the hypothetical, change-enabled, isolated system as a mathematical expression of
Le Chatelier’s principle. To see this more clearly, let us suppose that we are able to keep W and Z constant. We suppose that the
imposed change requires that the final value of X be X + dX . For the system to remain at equilibrium, the remaining variable, Y ,
must change by an amount, dY , that satisfies this inequality. That is, to reach the new equilibrium state, the change in Y must
satisfy
∂S ∂S
( ) dX + ( ) dY > 0
∂X ∂Y
WYZ W XZ
In this model, variables X and Y drive the entropy change as the hypothetical system moves toward its new equilibrium position.
The imposed change in X changes the entropy of the system by
∂S
dSimposed = ( ) dX
∂X
WYZ
Since the effect of the imposed change is to drive the system away from its original equilibrium position, we have dS imposed <0 .
The system’s response changes the entropy of the system by
∂S
dSresponse = ( ) dY
∂Y
W XZ
We have dS response> −dS > 0 , so that we can reasonably describe the response,
imposed dY , that makes dSresponse > 0 , as a
change that opposes the imposed change, dX, that makes dS < 0.
imposed
Applying Le Chatelier’s principle is something of an art. Central to this art is an ability to devise a hypothetical, change-enabled,
isolated, non-equilibrium state that is a good model for the initial state of the spontaneous process. In Section 6.6, we use
qualitative arguments to apply Le Chatelier’s principle to vapor–liquid equilibrium. To relate these qualitative arguments to the
mathematical model we have developed, let us consider the gedanken experiment depicted in Figure 3. We suppose that the initial
equilibrium system contains the liquid and vapor of a pure substance at pressure, P , and temperature, T . We imagine that we can
create the hypothetical isolated system by imposing a step change to the applied pressure without changing the pressure of the
system.
Figure 3. Spontaneous and reversible processes connecting the same liquid-vapor equilibrium states.
To model the pressure perturbation, let us view the liquid–vapor mixture as a subsystem, which is enclosed in a vertical cylinder
that is sealed by a frictionless piston. A mass, M , rests on top of the piston. For present purposes, we consider this mass to be a
portion of a second subsystem. The gravitational force on this mass creates the pressure applied to the liquid–vapor mixture. Since
this is an equilibrium state of the substance, this applied pressure is equal to the pressure, P , of the liquid–vapor subsystem. A
small mass, dM , is also a part of the second subsystem. In this original equilibrium state of the system, this smaller mass is
supported in some manner, so that it does not contribute to the applied pressure. We assume that the piston is a perfect thermal
insulator, so that no heat can be exchanged between the two subsystems.
As sketched in Figure 3, we create the hypothetical change-enabled system by moving the smaller mass so that it too rests on top of
the piston. Immediately thereafter, we completely isolate the system from the rest of the universe. We suppose that the applied
pressure instantaneously increases to P + dP . However, since the liquid–vapor subsystem is unchanged, we suppose that the
pressure, entropy, and all other thermodynamic properties of the liquid–vapor subsystem remain unchanged in this hypothetical
state. The system is not at equilibrium in this hypothetical state, because the applied pressure is not equal to that of the liquid–vapor
subsystem. Spontaneous change to a new equilibrium state can occur. Because the system is isolated, ΔS^ = 0 . Therefore, we have
ΔS > 0 . The final equilibrium temperature is T + dT .
With pressure and temperature as the independent variables, this model for Le Chatelier’s principle gives rise to the following
mathematical requirement:
−1
dS = T (∂H /∂T ) dT − (∂V /∂T ) dP > 0.
P P
We know that T , (∂H /∂T ) , and (∂V /∂T ) are positive. Therefore we can rearrange the inequality to find
P P
∂V ∂H
dT > T [( ) /( ) ] dP
∂T ∂T
P P
If we have dP >0 , it follows that dT >0 ; that is, the liquid–vapor equilibrium temperature increases with pressure.
This page titled 12.5: Another Perspective on the Principle of Le Chatelier is shared under a CC BY-SA 4.0 license and was authored, remixed,
In Sections 12.4 and 12.6, we explore two approaches to using the entropy-based criterion for spontaneous change. In discussing
the melting of ice at +10 C, we calculate the entropy changes for both the system and the surroundings to show that
ΔS universe > 0 , as the second law requires for a spontaneous process. In discussing the pressure-dependence of a liquid’s boiling
point in Section 12.6, we relate the second law criterion for spontaneous change to Le Chatelier’s principle. We turn now to
specifying the pressures and temperatures at which two phases of a pure substance are in equilibrium. When we choose pressure
and temperature as the independent variables, the Gibbs free energy criteria specify the equilibrium state and the direction of
spontaneous change.
For a reversible process in which all the work is pressure–volume work and in which the pressure and temperature change by dP
and dT , the change in the Gibbs free energy is dG = V dP − SdT . Let us apply this relationship to the liquid–vapor equilibrium
problem that we discuss in Section 12.6. To do so, we view the process from a slightly different perspective. We suppose that we
have two systems. These are identical to the initial and final states of the system in our discussion above. One of these systems is at
liquid–vapor equilibrium at a particular pressure, P , and temperature, T . The other is at liquid–vapor equilibrium at P + dP and
T + dT . We consider the change in the Gibbs free energy of a mole of the substance as it reversibly traverses the cycle sketched in
Figure 4.
Figure 4. Molar Gibbs free energy differences between two liquid-vapor equilibrium states.
The pressure and temperature are constant in each of the two equilibrium states. In either of these equilibrium states, the Gibbs free
energy does not change when a mole of liquid is converted to its gas, Δ G (P , T ) = 0 and Δ G (P + dP , T + dP ) = 0 .
vap vap
When the pressure and temperature of one mole of liquid change from P and T to P + dP and T + dT , the Gibbs free energy
¯
¯¯¯ ¯¯
¯
change is dG (ℓ) = V dP − S dT . For a mole of gas, this change in the pressure and temperature change the Gibbs free energy
¯
¯¯¯
ℓ ℓ
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯¯
¯
by dG (g) = V dP − S dT . (V , V , S , and S are evaluated at P and T . However, since dP and dT are small, these
g g ℓ g ℓ g
quantities are essentially constant over the pressure and temperature ranges involved.) For the individual steps in this cycle, we
have
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯
(ℓ, P , T ) → (ℓ, P + dP , T + dT ) dG (ℓ) = V ℓ dP − S ℓ dT
(ℓ, P + dP , T + dT ) → (g, P + dP , T + dT ) Δvap G (P + dP , T + dT ) = 0
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯
(g, P + dP , T + dT ) → (g, P , T ) − dG (g) = − (V g dP − S g dT )
(g, P , T ) → (ℓ, P , T ) − Δvap G (P , T ) = 0
Since the Gibbs free energy is a state function, the sum of these terms is zero. We have
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
Δvap G (P + dP , T + dT ) − Δvap G (P , T ) + dG (ℓ) − dG (g) = dG (ℓ) − dG (g) = 0
¯
¯¯¯ ¯
¯¯¯
so that dG (ℓ) = dG (g) . That is, the Gibbs free energy of the liquid changes by the same amount as the Gibbs free energy of the
gas when a mole of either is taken from one liquid–vapor equilibrium state to another. Substituting, we find a condition that the
pressure and temperature changes must satisfy when the system goes from the liquid–vapor equilibrium state at (P , T ) to the
liquid–vapor equilibrium state at (P + dP , T + dT ) :
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯¯
¯
(V g −V ℓ) dP − (S g − S ℓ ) dT = 0
¯¯
¯ ¯¯
¯ ¯¯
¯ ¯¯
¯
We let Δvap V
¯
¯¯¯
=V
¯
¯¯¯
g
¯
¯¯
−V
¯
ℓ and Δvap S = S g − S ℓ , where Δvap V
¯
¯¯¯
and Δvap S are the volume and entropy changes that
¯
¯¯¯ ¯¯
¯
accompany the vaporization of one mole of the liquid at P and T . Δvap V and Δvap S are essentially constant over the small
¯
¯¯¯ ¯¯
¯
pressure and temperature ranges involved. Substituting, we have Δ vap V dP − Δvap S dT = 0 , which we can rearrange to give
¯¯
¯
dP Δvap S
=
dT ¯
¯¯¯
Δvap V
¯
¯¯¯
¯
As one mole of liquid vaporizes reversibly at P and T , the system accepts heat q rev
P
= Δvap H . Hence, the entropy of vaporization
¯¯
¯ ¯
¯¯¯
¯
at P and T is Δ vap S = Δvap H /T , and the relationship between dP and dT becomes
¯
¯¯¯
¯
dP Δvap H
=
dT ¯
¯¯¯
T Δvap V
Below we see that such a relationship holds for any equilibrium between two pure phases. The general relationship is called the
Clapeyron equation.
This analysis is successful because the constituents are pure phases; the properties of the liquid are independent of how much vapor
is present and vice versa. When we analyze the equilibrium between a liquid solution and a gas of the solution’s components, the
problem is more complex, because the properties of the phases depend on their compositions.
This page titled 12.6: Phase Equilibria - Temperature Dependence of the Boiling Point is shared under a CC BY-SA 4.0 license and was authored,
We can also represent reversible changes by paths on contour maps. In Figure 5, a Gibbs free energy surface is represented as a
contour map.
Figure 5. The path of a changing system depicted on a map of constant Gibbs free energy contours.
For small changes in T and P , we can evaluate
dG = G (P0 + dP , T0 + dT ) − G (P0 , T0 )
from
dG = −S (P0 , T0 ) dT + V (P0 , T0 ) dP
For larger changes, we can integrate along the paths P = P0 and T = T0 + ΔT to find
T0 +ΔT P0 +ΔP
ΔG = ∫ −S (P0 , T ) dT + ∫ V (P , T0 + ΔT ) dP
T0 P0
The calculation of ΔS in Section 12.5 could be similarly represented as a path in the temperature–pressure plane that connects two
constant-entropy contours.
Analysis of solid–liquid equilibrium parallels that of liquid–vapor equilibrium. Let us again consider the equilibrium between ice
and water. Given that ice and water are at equilibrium at a particular temperature and pressure, and supposing that we increase the
pressure from this equilibrium value, how must the temperature change in order that the system remain at equilibrium? In Section
6.6, we use Le Chatelier’s principle to answer this question qualitatively. Now, we find a quantitative answer by an argument that
closely parallels that in Section 12.7.
Figure 6. Gibbs free energy changes in a cycle connecting two solid-liquid equilibrium states.
Figure 6 depicts the line of pressures and temperatures along which ice and water are in equilibrium. We can view this as a contour
¯
¯¯¯
map. In this case, the contours are sets of pressures and temperatures for which Δ G is constant. Only the contour for f us
¯
¯¯¯
Δf us G = 0 is shown. The figure also depicts paths along which ice and water can individually be taken from their equilibrium
state at P and T to their equilibrium state at P + dP and T + dT . The Gibbs free energy change for the ice must equal that for
¯
¯¯¯ ¯¯
¯
water. Letting G , S , and V be the Gibbs free energy, the entropy, and the volume of one mole of water at temperature T and
ℓ ℓ
¯
¯¯¯
ℓ
pressure P , the equation

¯
¯¯¯ ¯
¯¯¯ ¯¯
¯
dGℓ = V ℓ dP − S ℓ dT
specifies the change in the Gibbs free energy of one mole of water when the pressure changes P to P + dP and the temperature
changes from T to T + dT . Similarly, using the subscript “s” to denote ice, we have
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯
dGs = V s dP − S s dT
Since these Gibbs free energy changes connect states of ice–water equilibrium, they must be equal, and we have
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯¯
¯ ¯
¯¯¯ ¯¯
¯
dGℓ − dGs = (V ℓ −V s) dP − (S ℓ − S s ) dT = Δf us V dP − Δf us S dT = 0
¯¯
¯ ¯
¯¯¯
where we introduce Δ S and Δ V to represent the entropy and volume changes that occur when one mole of ice melts
f us f us
reversibly at P and T . Rearranging gives

¯¯
¯
dP Δf us S
=
¯
¯¯¯
dT
Δf us V
Since
¯¯
¯ ¯
¯¯¯
¯
Δf us S = Δf us H /T ,
the Clapeyron equation becomes

¯
¯¯¯
¯
dP Δf us H
=
¯
¯¯¯
dT
T Δf us V
At a pressure of one bar and a temperature of 273.15 K, the enthalpy of fusion is 6010 J mol K . The enthalpy value changes −1 −1
only slowly as the equilibrium temperature changes. The volumes of one mole of ice and one mole of water are 19.651 and 18.019
cm , respectively. At 273.15 K, we obtain
3
dP
−1
= −143.7 bar K
dT
If the pressure increases to 1000 bar, the change in the melting point is about –6.96 K, so that T mp (1000 bar) = 266.2 K .
Again, this analysis is successful because the constituents are pure phases; the properties of the ice are independent of how much
water is present and vice versa. When we analyze the equilibrium between ice and salt water, the properties of the salt water
depend on the kind of salt present and on its concentration.
This page titled 12.7: Phase Equilibria - Temperature Dependence of the Melting Point is shared under a CC BY-SA 4.0 license and was authored,
The analysis in the two previous sections can be repeated for any phase change of a pure substance. Let α and β denote the two
phases that are at equilibrium.
α ⇌ β
¯
¯¯¯ ¯¯
¯ ¯
¯¯¯
Let G , S , and V
α α α represent the Gibbs free energy, the entropy, and the volume of one mole of pure phase α at pressure P and
¯
¯¯¯ ¯¯
¯ ¯
¯¯¯
temperature T . Let G , S , and V
β β β represent the corresponding properties of one mole of pure phase β. The equations
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯
dG (α) = V α dP − S α dT
and
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯
dG (β) = V β dP − S β dT
describe the changes in the Gibbs free energy of a mole of α and a mole of β when they go from one α − β -equilibrium state at P
and T to a second α − β -equilibrium state at P + dP and T + dT . Since these Gibbs free energy changes must be equal, we have
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯¯
¯
dG (β) − dG (α) = (V β
−V α) dP − (S β − S α ) dT
¯
¯¯¯ ¯¯
¯
= ΔV dP − ΔS dT
=0
and
¯¯
¯
dP ΔS
=
¯
¯¯¯
dT
ΔV
¯¯
¯
where ΔS and ΔV are the entropy and volume changes that occur when one mole of the substance goes from phase α to phase β.
¯
¯¯¯
¯¯
¯ ¯
¯¯¯
¯
Since ΔS = ΔH /T , the condition for equilibrium between phases α and β becomes
¯
¯¯¯
¯
dP ΔH
= (12.8.1)
dT ¯
¯¯¯
T ΔV
Equation 12.8.1 is known as the Clapeyron equation.
This page titled 12.8: The Clapeyron Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
To use the Clapeyron equation we must know the enthalpy and volume differences at one equilibrium temperature and pressure. In
general, these properties are readily measured. If we fix the pressure, we can measure the corresponding equilibrium temperature.
We can obtain the enthalpy change at this pressure by measuring the heat required to convert a mole of the substance from one
phase to the other. We can obtain the volume change from the molar volumes, which we can obtain by measuring the density of
each phase. The enthalpy of the phase change varies only weakly as the equilibrium pressure and temperature vary. Similarly, for
condensed phases, the densities are weak functions of temperature. This means that, for transitions between condensed phases,
¯
¯¯¯
¯ ¯
¯¯¯
ΔH /ΔV is approximately constant over a modest temperature range.
For a sublimation or vaporization process, the product is a gas. Then the molar volume of the product is a sensitive function of both
pressure and temperature. However, the molar volume of the product phase is much greater than the molar volume of the initial
solid or liquid phase. To a good approximation, the volume change for the process equals the volume of the gas produced. If we
have an equation of state for the gas, the volume calculated from the equation of state is a good approximation to ΔV for the phase
change. The ideal gas equation is usually adequate for this purpose. Then, ΔV ≈ RT /P , and ¯
¯¯¯
¯
¯¯¯
¯
dP P ΔH
= (Clausius-Clapeyron equation)
2
dT RT
This equation for the pressure–temperature relationship for a phase equilibrium involving a gas is called the Clausius-Clapeyron
equation. Dividing both sides by the pressure, we can put the Clausius-Clapeyron equation into an alternative and often-useful
form:
¯
¯¯¯
¯
dln P ΔH
=
2
dT RT
If we can assume that ΔH is independent of pressure, we can separate variables and integrate to obtain the Clausius-Clapeyron
¯
¯¯¯
¯
equation in integrated form. If we can assume further that ΔH is constant, the integration yields
¯
¯¯¯
¯
P ¯
¯¯¯
¯ T
dP ΔH dT
∫ = ∫
2
P0
P R T0
T
and
¯
¯¯¯
¯
P ΔH 1 1
ln =− ( − )
P0 R T T0
where P and T are the initial equilibrium position.

0 0
This page titled 12.9: The Clausius-Clapeyron Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
12.10: Problems
1. For any change in a reversible system, we have dG = −SdT + V dP . Consider two systems, α and β, where system α can be
converted to system β. (Below, we will let α and β be the solid and liquid phases of the same pure substance, but this is not a
necessary restriction.) For incremental changes in temperature and pressure, represented by dT and dP , we have
dGα = Vα dP − Sα dT
and dG β
= Vβ dP − Sβ dT
We can subtract to find
d (Gβ − Gα ) = (Vβ − Vα ) dP − (Sβ − Sα ) dT
or
d (Δα→β G) = (Δα→β V ) dP − (Δα→β S) dT
which we usually write as
d (ΔG) = ΔV dP − ΔSdT
Here Δ X (or ΔX) is the change in the state function X that occurs when system α is converted to system β. For many inter-
α→β
convertible systems, it is a good approximation to say that Δ S (or ΔS ) and Δ α→βV (or ΔV ) are constant for modest changes
α→β
in temperature or pressure. Then, representing the pressure and temperature in the initial and final states as (P , T ) and (P , T ), 1 1 2 2
respectively, the change in Δ G (or ΔG) can be obtained by integration:

α→β
P2 , T2 T2 P2
∫ d (ΔG) = −ΔS ∫ dT + ΔV ∫ dP
P1 , T1 T1 P1
or
Δ (ΔG) = ΔG (P2 , T2 ) − ΔG (P1 , T1 ) = −ΔS (T2 − T1 ) + ΔV (P2 − P1 )
Note that α and β need not be in equilibrium with one another at either the condition specified by (P1 , T1 ) or that specified by
(P , T ).
2 2
However, in the important special case that α and β are in equilibrium at (P1 , T1 ) , we have
Δα→β G (P1 , T1 ) = ΔG (P1 , T1 ) = 0 . Then
Δ (ΔG) = ΔG (P2 , T2 ) = −ΔS (T2 − T1 ) + ΔV (P2 − P1 )
Consider the application of these observations to the case where α and β are solid and liquid aluminum metal, respectively. At one
¯
¯¯¯
¯
bar, aluminum melts at 933.47 K. At the melting point, the enthalpy of fusion is Δ H = +10.71 kJ mol . The atomic weight
f us
−1
of aluminum is 26.9815 g mol . The density of liquid aluminum at the melting point is 2.375 g cm ; the molar volume of the
−1 −1
liquid is therefore 1.1361 × 10 m mol . At 20 C, the density of solid aluminum is 2.70 g cm , and the molar volume is
−5 3 −1 −1
0.9993 × 10
−5
m mol . Assuming the molar volume of the solid to be independent of temperature, the change in the molar
3 −1
¯
¯¯¯
volume that occurs when aluminum melts is Δ V = 1.368 × 10 m mol .
f us
−6 3 −1
(a) What is Δ f us V
¯
¯¯¯
at the melting point at one bar (point A on the diagram)?
(b) Is the conversion of solid aluminum to liquid aluminum a reversible, spontaneous, or impossible process at (933.47 K, 1 bar)—
that is, at point A?
¯¯
¯
(c) What is Δ f us S at the melting point at one bar—that is, at point A?
¯
¯¯¯
(d) What is Δ (Δ f us G) when the pressure and temperature go from (933.47 K, 1 bar) to (934.47 K, 1 bar)? (That is from point A
¯
¯¯¯
to point B on the diagram.) What is Δ f us G at (934.47 K, 1 bar)?
(e) Is the conversion of solid aluminum to liquid aluminum a reversible, spontaneous, or impossible process at (934.47 K, 1 bar)—
that is, at point B?
¯
¯¯¯
(f) What is Δ (Δ f us G) when the pressure and temperature go from (933.47 K, 1 bar) to (932.47 K, 1 bar)? (That is from point A
¯
¯¯¯
to point C on the diagram.) What is Δ f us G at (932.47 K, 1 bar)?
(g) Is the conversion of solid aluminum to liquid aluminum a reversible, spontaneous, or impossible process at (932.47 K, 1 bar)—
that is, at point C?
(h) What is Δ (Δ f us
¯
¯¯
G)
¯
when the pressure and temperature go from (933.47 K, 1 bar) to (933.47 K, 101 bar)? (That is from point
¯
¯¯¯
A to point D on the diagram.) What is Δ f us G at (933.47 K, 101 bar)?
(i) Is the conversion of solid aluminum to liquid aluminum a reversible, spontaneous, or impossible process at (933.47 K, 101 bar)
—that is, at point D?
(j) If we maintain the pressure constant at 101 bar, how much would we have to change the temperature to just offset the change in
¯
¯¯¯
Δ (Δf us G) that occurred in part h? Note that this change will reach the conditions represented by point E on the diagram
(k) Is the conversion of solid aluminum to liquid aluminum a reversible, spontaneous, or impossible process at point E?
¯
¯¯¯
(l) What is Δ (Δf us G) in going from point D (933.47 K, 101 bar) to point B (934.47 K, 1 bar)? Is this value equal to the
difference between the Gibbs free energy of a mole of liquid aluminum at point B and a mole of solid aluminum at point D?
2. In the temperature interval 0.01 > t > −10 C , the vapor pressure of water (in Pa) above pure ice is aproximated by
ln P = a + bt + ct , where a = 6.41532, b = 8.229 × 10 , c = −3.2 × 10 , t = T − 273.15 , and T is the temperature in
2 −2 −4
degrees kelvin. Estimate the enthalpy of sublimation of ice at 273.15 K.

3. In the temperature interval , the vapor pressure of water (in Pa) is approximated by
373.15 > T > 273.15 K
ln P = a + bt + c t
2
+ dt
3
+ et
4
,
where a = 9.42095, b = 4.960 × 10 , c = −1.7536 × 10 , d = 6.02 × 10 , −2 −4 −7
e = −2.0 × 10 , t = T − 323.15 , and T is the temperature in degrees kelvin. Estimate the enthalpy of vaporization of water at
−9
323.15 K and 373.15 K.

4. The normal (1 atm) boiling point of acetone is 56.05 C. The enthalpy of vaporization at the normal boiling point is
29.10 kJ mol . What is the entropy of vaporization of acetone at the normal boiling point? Estimate the vapor pressure of
−1
acetone at 25.0 C.
5. Two allotropic forms of tin, gray tin and white tin, are at equilibrium at 13.2 C and 1 atm. The density of gray tin is
5.769 g cm ; the density of white tin is 7.265 g cm . Assume that the densities are independent of temperature. At 298.15 K
−1 −1
and 1 bar, S and C for gray tin are 44.1 J mol K and 25.8 J mol K , respectively. For white tin, S and C are
o
P
−1 −1 −1 −1 o
P
¯¯
¯ ¯
¯¯¯
51.2 J mol K
−1
and 27.0 J mol K , respectively. Estimate ΔS and ΔV for the conversion of gray tin to white tin at 13.2
−1 −1 −1
C. At what temperature are gray tin and white tin at equilibrium at a pressure of 100 atm?
6. At 1000K, the standard Gibbs free energies of formation of graphite and diamond are 0.000 and +5.905 kJ mol , respectively. −1
At 298.15 K and 1.000 bar, the molar volumes of graphite and diamond are 5.46 × 10 m mol and 3.42 × 10 m mol , −6 3 −1 −6 3 −1
respectively. Let P be the pressure at which graphite and diamond are at equilibrium at 1000 K.
#
¯
¯¯¯
(a) What is the value of ΔG for
C (graphite, 1 bar, 1000K) → C (diamond, 1 bar, 1000K)?
¯
¯¯¯
(b) Express ΔG for
#
C (graphite, 1 bar, 1000K) → C (graphite, P , 1000K)
as a function of P .#
¯
¯¯¯
(c) Express ΔG for
#
C (diamond, 1 bar, 1000K) → C (diamond, P , 1000K)
as a function of P .#
¯
¯¯¯
(d) What is the value of ΔG for
# #
C (graphite, P , 1000K) → C (diamond, P , 1000K)?
(e) Assume that the molar volumes are independent of pressure and temperature. Estimate the value of the equilibrium pressure,
P
#
.
Notes
See Robert Bruce Lindsay and Henry Margenau, Foundations of Physics, Dover Publications, Inc., New York, 1963, p 195.
CHAPTER OVERVIEW
13: Equilibria in Reactions of Ideal Gases

13.5: ∆rG is the rate at which the Gibbs Free Energy Changes with The Extent of Reaction
13.6: The Standard Gibbs Free Energy Change and Equilibrium in Ideal Gas Reactions
13.10: Problems
This page titled 13: Equilibria in Reactions of Ideal Gases is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
1
In Chapter 11, we find a general equation for the molar Gibbs free energy of a pure gas. We adopt the Gibbs free energy of
formation of the hypothetical ideal gas, in its standard state at 1 bar, P , as the reference state for the Gibbs free energy of the gas
o
at other pressures and the same temperature. Then, the molar Gibbs free energy of pure gas A , at pressure P , is
P ¯
¯¯¯
¯
¯¯¯ o o
P V 1
GA (P ) = Δf G (A, H I G ) + RT ln( ) + RT ∫ ( − ) dP
o
P 0 RT P
(any pure gas)

¯
¯¯¯
GA (P ) is the difference between the Gibbs free energy of the gas at pressure P and that of its constituent elements at 1 bar and the
same temperature. If gas A is an ideal gas, the integral is zero, and the standard-state Gibbs free energy of formation is that of an
“actual” ideal gas, not a “hypothetical state” of a real gas. To recognize this distinction, let us write Δ G (A, P ) , rather than f
o o
Δ G (A, H I G ) , when the gas behaves ideally. In a mixture of ideal gases, the partial pressure of gas A is given by P =x P ,
o o
f A A
where x is the mole fraction of A and P is the pressure of the mixture. In §3, we find that the Gibbs free energy of one mole of
A
pure ideal gas A at pressure P has the same Gibbs free energy as one mole of gas A in a gaseous mixture in which the partial
A
pressure of A is P = x P . Recognizing these properties of an ideal gas, we can express the molar Gibbs free energy of an ideal
A A
gas—pure or in a mixture—as
¯
¯¯¯ o o
PA
GA (PA ) = Δf G (A, P ) + RT ln( )
o
P
(ideal gas)
¯
¯¯¯ ¯
¯¯¯
Note that we can obtain this result for pure gas A directly from (∂ G/∂P ) =V = RT /P by evaluating the definite integrals
T
¯
¯¯¯
GA ( PA ) PA
¯
¯¯¯ RT
∫ dG = ∫ dP
Δf G (A, P
o o
) P
o P
Including the constant, P , in these relationships is a useful reminder that RT ln(P /P ) represents a Gibbs free energy
o
A
o
difference. Including P makes the argument of the natural-log function dimensionless; if we express P in bars, including
o
= 1 bar leaves the numerical value of the argument unchanged. If we express P in other units, P becomes the conversion
o o
P
factor for converting those units to bars; if we express P in atmospheres, we have P = 1 bar = 0.986923 atm. o
However, including the “P ” is frequently a typographical nuisance. Therefore, let us introduce another bit of notation; we use a
o
lower-case “p” to denote the ratio “P /P ”. That is, p is a dimensionless quantity whose numerical value is that of the partial
0
A
pressure of A , expressed in bars. The molar Gibbs free energy becomes

¯
¯¯¯ o o
GA (PA ) = Δf G (A, P ) + RT ln pA
(ideal gas)
This page titled 13.1: The Gibbs Free Energy of an Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
Let us consider a system that consists of a mixture of ideal gases A , B , C , and D , at a particular fixed temperature. We suppose
that reaction occurs according to
We want to think about what happens when a moles of A (at pressure P ) and b moles of B (at pressure P ) react to form c moles
A B
of C (at pressure P ) and d moles of D (at pressure P ) under conditions in which the partial pressures in the mixture remain
C D
constant at P , P , P , and P . At first encounter, these conditions may appear to be impossible; if A reacts, its partial pressure
A B C D
must change. However, on reflection, we recognize that by making the system very large, a moles of A and b moles of B can
disappear without changing P or P very much. In fact, we can make the change in P and P as small as we like just by
A B A B
making the original system large enough. The same considerations apply to P and P . We let Δ G be the Gibbs free energy C D r
change under these conditions. We want to know how the Gibbs free energy of this system, G, and the Gibbs free energy change,
Δ G, depend on the partial pressures of the gases involved in the reaction.
r
We have observed repeatedly that the temperature and pressure of a system undergoing spontaneous change may not be well
defined, and the concentration of a component may vary from point to point within a given phase. If any of these inhomogeneities
are substantial, the reaction conditions in the previous paragraph are not met. On the other hand, if any pressure and temperature
variations are too small to have observable effects, and there are no point-to-point concentration variations, it is entirely reasonable
to suppose that the Gibbs free energies of the system, and of its individual components, are described by the thermodynamic
models we have developed for reversible systems.
To see that there can be no objection in principle to measuring Gibbs free energies in a non-equilibrium system, we need only find
a hypothetical equilibrium system whose state functions must have the same values. Thus, if the reaction
a A+b B ⇌ c C +d D does not occur, mixtures of gases A , B , C , and D in any proportions can be at equilibrium, and the
thermodynamic properties of any such mixture are well defined. In concept, we can produce a hypothetical equilibrium state
equivalent to any intermediate state in the spontaneous process if we suppose that the reaction occurs only in the presence of a solid
catalyst. By introducing and then removing the catalyst from the reaction mixture, we can produce a quasi-equilibrium state whose
composition is identical to that of the spontaneous reaction at any particular extent of reaction. In this quasi-equilibrium state, the
pressure is equal to the applied pressure, and the temperature is equal to that of the surroundings. If the spontaneous uncatalyzed
reaction is slow compared to the rate at which the pressure and temperature of the system can equilibrate with the applied pressure
and the surroundings temperature, the state functions of a spontaneously reacting system and a static, quasi-equilibrium system
with the same composition must be essentially identical.
It turns out that we can find the dependence of Δ G on concentrations by considering a fundamentally different system—one that
r
is composed of exactly the same amount of each of these gases, but in which the gases are not mixed. Each gas occupies its own
container. For the partial pressures P , P , P , and P , the Gibbs free energies per mole are
A B C D
¯
¯¯¯ o o
¯
¯¯¯ o o
GB (PB ) = Δf G (B, P ) + RT ln pB
¯
¯¯¯ o o
GA (PC ) = Δf G (C , P ) + RT ln pC
¯
¯¯¯ o o
GD (PD ) = Δf G (D, P ) + RT ln pD
The Gibbs free energy of this system, which is just a composite of the separated gases, is
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
G = nA GA (PA ) + nB GB (PB ) + nC GC (PC ) + nD GD (PD )
If we subtract the Gibbs free energies of a moles of reactant A and b moles of reactant B from the Gibbs free energies of c moles
of product C and d moles of product C , we find
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
c GC (PC ) + d G D (PD ) − a G A (PA ) + b GB (PB )
pC pD
o o o o o o o o
= c Δf G (C , P ) + d Δf G (D, P ) − a Δf G (A, P ) − b Δf G (B, P ) + RT ln
pA pB
To represent these free energy differences when the gases are in separate containers, we introduce the abbreviations
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
Δsep G = c GC (PC ) + d G D (PD ) − a G A (PA ) − b GB (PB )
and
o o o o o o o o o
Δr G = c Δf G (C , P ) + d Δf G (D, P ) − a Δf G (A, P ) − b Δf G (B, P )
so that the difference between the Gibbs free energies of the separated reactants and products can be written more compactly as
o
pC pD
Δsep G = Δr G + RT ln
pA pB
Δsep G is the difference in the Gibbs free energies when the pressures of the separated gases are fixed at P , P , P , and P . A B C D
Note that, if any of the pressures changes, Δ G changes. When we introduce Δ G in Section 11.10, we emphasize that this
sep r
o
quantity is the difference between the standard Gibbs free energies of the separated products
c moles of C at 1 bar and d moles of D at 1 bar.
and the separated reactants
a moles of A at 1 bar and b moles of B at 1 bar.
We call Δ G the standard Gibbs free energy change for the reaction. At a given temperature, Δ G is a constant. Our choice of
r
o
r
o
standard state for the Gibbs free energy of a compound means that we can calculate the standard free energy change for a reaction
from the standard free energies of formation of the products and reactants:
o o o o o o o o o
Δr G = c Δf G (C , P ) + d Δf G (D, P ) − a Δf G (A, P ) − b Δf G (B, P )
Let us recapitulate: In the first system, we are interested in the Gibbs free energy change, Δ G, for a process in which a moles ofr
A and b moles of B are converted to c moles of C and d moles of D in a (large) mixed system where the partial pressures are
constant at the values P , P , P , and P . In the second system, there is actually no process at all. The Gibbs free energy change,
A B C D
Δ sepG, is merely a computed difference between the Gibbs free energies of the specified quantities of product and reactant gases,
where each gas is in its own container at its specified pressure. When the gases are ideal, the Gibbs free energy differences
(changes) for these two systems turn out to be the same. That is, Δ G = Δ G. sep r
This relationship is valid for systems in which the properties of one substance are not affected by the concentrations of other
substances present. It is not true in general. On the other hand, Δ G is always the computed difference between the standard
r
o
Gibbs free energies of the pure separated products and reactants. For ideal gases,
pC pD
o
Δr G = Δr G + RT ln
pA pB
(reaction of ideal gases)

We have asserted that we can equate Δ G and Δ G for systems composed of ideal gases. Now we need to show that this is true.
sep r
This is easy if we first understand a related problem—the thermodynamics of mixing ideal gases.
This page titled 13.2: The Gibbs Free Energy Change for A Reaction of Ideal Gases is shared under a CC BY-SA 4.0 license and was authored,
When we talk about the thermodynamics of mixing, we have a very particular process in mind. By convention, the process of
mixing two gases, call them A and B , is the process in which the two gases initially occupy separate containers, but are both at a
common pressure and temperature. (We denote the common initial pressure by “P ”. P is not to be confused with the constant
0 0
P .) The final state after the mixing process is one in which there is a homogeneous mixture of A and B at the same temperature
o
as characterized the initial state. The final volume is the sum of the initial volumes. If the gases are ideal, the final pressure is the
same as the initial pressure, and the partial pressures are P = n RT /(V + V ) and P = n RT /(V + V ) .
A A A B B B A B
Figure 1. Changes in thermodynamic functions during the mixing of ideal gases.

The mixing process is represented by the change on the right side of Figure 1. The gases are always in thermal contact with
constant-temperature surroundings. We imagine that we bring the initially separate containers together and then remove the
overlapping walls. (Or we can imagine connecting the two containers by a tube—whose volume is negligibly small—that allows
the molecules to move from one container to the other.) Molecular diffusion eventually causes the concentration of either gas to be
the same in any macroscopic portion of the combined volumes. This diffusive mixing begins as soon as we provide a path for the
molecules to move between their containers. Isothermal mixing is a spontaneous process. The reverse process does not occur.
Isothermal mixing is irreversible.
Since the temperature is constant and the gases are ideal, the energy of the A molecules is constant; likewise, the energy of the B
molecules constant. It follows that the energy of a system containing molecules of A and B is independent of their concentrations
and that the energy of the mixture is the sum of the energies of the separated components. That is, we have Δ E = 0 . Since the
mix
volume and the pressure of the two-gas system are constant, it follows that Δ (P V ) = 0 and Δ
mix H = 0 . Constant volume
mix
also means that w mix = 0 . It follows that qmix= 0 , and that Δ S = 0 . Then, because the mixing process is spontaneous, we
^
mix
have Δ S > 0 . Finally, since Δ H = 0 , we have Δ G = −T Δ S < 0 . However, in order to calculate the values of
mix mix mix mix
Δ mix S = 0 and Δ mix G < 0 , we must find a reversible path for the state change that occurs during mixing.
The two-step path represented by the changes along the left side and the bottom of Figure 1 is such a reversible path. The first step
is simply the reversible isothermal expansion of the separate gases from their initial volumes (V and V , respectively) to the
A B
common final volume, V + V . At the end of this step, the two gases are still in separate containers. Reversible isothermal
A B
expansion of an ideal gas is a familiar process; for this step, we know:
ΔEA = ΔHA = 0
ΔEB = ΔHB = 0
ΔSA = −nA R ln(PA / P0 ) > 0
ΔSB = −nB R ln(PB / P0 ) > 0
ΔGA = −T ΔSA < 0
ΔGB = −T ΔSB < 0
We call the change along the bottom of the diagram the merging process. The merging process is the isothermal, reversible
blending of the separate gas samples, each initially occupying a volume V + V , in such a way that the final state has all of the
A B
molecules of the two gases in the same container, whose volume is also V + V . While the final state of the merging process is
A B
identical to the final state of the mixing process, the initial state is distinctly different.
For ideal gases, it turns out that all of the thermodynamic functions are unchanged during the merging process. Consequently, the
thermodynamic functions for mixing are just the sums of the thermodynamic functions for the reversible expansions of A and B
separately. Equating thermodynamic functions for the two paths to the mixed state, we have
Δmix E = ΔEA + ΔEB + Δmerge E = 0
Δmix H = ΔHA + ΔHB + Δmerge H = 0
Δmix S = ΔSA + ΔSB + Δmerge S = −nA R ln(PA / P0 ) − nB RT ln(PB / P0 ) > 0
Δmix G = −T Δmix S < 0
The pressure ratios equal the mole fractions of the compounds in the mixture. Therefore, the entropy of mixing is also given by
Δmix S = −nA Rln xA − nB Rln xB >0
If we calculate the entropy of mixing per mole of A –B -mixture, Δ mix

¯¯
S
¯
, we find
¯¯
¯
Δmix S nA nB
Δmix S = = −( ) R ln xA − ( ) R ln xB = −xA R ln xA − xB R ln xB
nA + nB nA + nB nA + nB
It remains to prove our assertion that the thermodynamic functions are unchanged during the merging process. That the energy is
unchanged follows from the fact that the energy of an ideal gas depends only on temperature; it is therefore independent of pressure
and of the presence of any other substance. We can give a qualitative argument for the idea that other thermodynamic quantities are
also unchanged. In this argument, the enthalpy, entropy, and free energy functions are unchanged because ideal gas molecules do
not interact with one another. If A molecules do not interact with B molecules, it follows that the properties of the A molecules are
independent of whether the B molecules share the same container or are present in a separate container of identical volume. At the
same temperature, the A molecules generate a pressure P , and the B molecules independently generate a pressure P . Since
A B
these pressures are generated independently, we conclude that the total pressure is the sum of the two partial pressures—which is,
of course, just Dalton’s law of partial pressures. The same argument applies to the enthalpy, entropy, and free energy functions, so
these should also be unchanged during the merging process.
Figure 2. A device to merge gases.

We can also create a device in which we can—in concept—carry out the reversible isothermal merging process and calculate
thermodynamic-function changes. (We rely on the argument above to establish Dalton’s law of partial pressures and the conclusion
that Δmerge E = 0 . We use the device to show from these results that there is no change in the other thermodynamic functions.)
The device is sketched in Figure 2, at an intermediate stage of the process. It consists of three cylinders, each closed by a
frictionless piston. The first contains unmerged gas A , the second contains unmerged gas B , and the third contains the mixture that
results from merging them. The head of the A cylinder is shared with part of the head of the merging cylinder. Likewise, the head
of the B cylinder is shared with another part of the head of the merging cylinder. We suppose that the A and B cylinder heads are
comprised in part of molecule-selective gas-permeable membranes. The membrane in the head of the A cylinder allows the
diffusion of A molecules in either direction, but does not permit B molecules to pass. The membrane in the head of the B cylinder
allows the diffusion of B molecules in either direction, but does not permit A molecules to pass.
In the merging cylinder, we reversibly accumulate A molecules from the A cylinder and B molecules from the B cylinder. We
accomplish this by controlling the pressures in the three cylinders. The pressure in the A cylinder is always infinitesimally greater
than P . The pressure in the B cylinder is always infinitesimally greater than P . The partial pressure of A in the merging
A B
cylinder is infinitesimally less than P ; the partial pressure of B in the merging cylinder is infinitesimally less than P ; and the
A B
total pressure in the merging cylinder is infinitesimally less than P + P . (There is always a difference in the total pressure across
A B
each membrane.) The system consists of the contents of the cylinders. Work is done on the system by the forces acting on the A
and B cylinders. Work is done by the system on the surroundings as the merging cylinder fills with the mixture.
As far as the A cylinder is concerned, the process is reversible because the pressure due to the A molecules is just infinitesimally
greater in the A cylinder than in the merging cylinder. Therefore, a very small decrease in the pressure of the A cylinder would
cause the net flow of A molecules to change direction. The same is true for the B cylinder.
Figure 3. The device before and after merging the gases.

The initial and final states of the apparatus when we use it to merge n moles of A at P with n moles of B at P are shown in
A A B A
Figure 3. Let Vmix =V A +V B be the sum of the initial volumes of cylinders A and B . This is also the final volume of the merging
cylinder. The reversible work done in the A cylinder is
0
rev
w = −∫ PA dV = −PA (0 − VA VB ) = PA (VA + VB ) = nA RT
A
VA +VB
Similarly, the work in the B cylinder is

rev
w = PB (VA + VB ) = nB RT
B
In the merging cylinder it is

rev
w = − (PA + PB ) (VA + VB ) = − (nB + nB ) RT
A+B
For the merging process the net work is

rev rev rev rev
wmerge = w +w +w =0
A B A+B
and the net change in the pressure-volume product is
Δmerge (P V ) = (PA + PB ) (VA + VB )
−PA (VA + VB ) − PB (VA + VB )
=0
Since constant temperature ensures that Δ E = 0 , it follows from Δ

merge (P V ) = 0
mergethat Δ merge H =0 , and from
= 0 that q = 0 . Since the merging process is reversible, we have also that Δ S = 0 and Δ .
rev rev
w merge merge merge merge G=0
These arguments can be extended to merging any number of ideal gases. In the initial state for this merging process each gas is at
the same temperature, but occupies a separate container; all of these containers have the same volume. Each gas can be at a
different pressure. In the final state, all of the gases occupy a common container, whose volume is the same as the common volume
of their initial containers. The temperature of the mixture in the final state is equal to the common initial temperature of the
separate gases. In the final state, the partial pressure of each gas is equal to its pressure in the initial state. For any number of gases,
we have
Δmerge E = Δmerge H = Δmerge S = Δmerge G = 0
Likewise, these arguments can be extended to the mixing of multiple ideal gases, all at the same original pressure and temperature,
into a final volume that is equal to the sum of the initial volumes—and is at the original temperature. If there are ω such gases,
Δmix E = Δmix H = 0
Δmix S = ∑ −ni R ln xi > 0
i=1
Δmix G = −T Δmix S < 0
This page titled 13.3: The Thermodynamics of Mixing Ideal Gases is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Now we can compare the difference between the Gibbs free energies of reactants and products in the reaction
when all of the gases are present in the same system to the same difference when each gas is in its own container.
In the first case, the gases are present in a mixture, and their partial pressures remain constant at P , P , P , and P respectively.
A B C D
We called the Gibbs free energy change under these conditions Δ G. These conditions can be satisfied if we suppose that the
r
system is very large. That is, Δ G is the limiting Gibbs free energy change for the conversion of an initial mixture into a final
r
mixture. The initial mixture contains (n + a) moles of A , (n + b) moles of B , n moles of C , and n moles of D. The final
A B C D
mixture contains n moles of A , n moles of B , (n + c) moles of C , and (n + d) moles of D. Δ G is the Gibbs free energy
A B C D r
change for this conversion in the limit as n , n , n , and n become arbitrarily large. In this limit, we have n ≫ a , n ≫ b ,
A B C D A B
nC ≫ c , , and n D≫ d . Since the partial pressures remain (essentially) constant, the n must also satisfy
i
PA = nA /(nA + nB + nC + nD )
PB = nB /(nA + nB + nC + nD )
PC = nC /(nA + nB + nC + nD )
PD = nD /(nA + nB + nC + nD )
When the ideal gases are separated from one another, the Gibbs free energy difference is the Gibbs free energy of c moles of gas C
(at pressure P ) plus the Gibbs free energy of d moles of gas D (at pressure P ) minus the Gibbs free energy of a moles of gas A
C D
(at pressure P ) and minus the Gibbs free energy of b moles of gas B (at pressure P ). In §2, we call the Gibbs free energy change
A B
under these conditions Δ G. sep
In Section 13.2, we assert that Δsep G = Δr G . Now we can see why this is so. The cycle shown in Figure 4 relates these two
Gibbs free energy differences.
Figure 4. Cycle demonstrating that Δ sep G = Δr G .

We have
Δsep G (PA , PB , PC , PD ) + Δmerge Gf inal
= Δmerge Ginitial + Δr G (PA , PB , PC , PD )
and since
Δmerge Gf inal = Δmerge Ginitial
we have
Δsep G (PA , PB , PC , PD ) = Δr G (PA , PB , PC , PD )
¯
¯¯¯ ¯
¯¯¯
Letting G (P ) be the Gibbs free energy of one mole of pure ideal gas
A A A at pressure PA ,GB (PB ) be the Gibbs free energy of
one mole of pure ideal gas B at pressure P , etc., we have
B
¯
¯¯¯ ¯
¯¯¯
Δsep G (PA , PB , PC , PD ) = nA GA (PA ) + nB GB (PB )
¯
¯¯¯ ¯
¯¯¯
+ (nC + c) GC (PC ) + (nD + d) GD (PD )
¯
¯¯¯ ¯
¯¯¯
− (nA + a) GA (PA ) − (nB + b) GB (PB )
¯
¯¯¯ ¯
¯¯¯
− nC GC (PC ) − nD GD (PD )
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
= c GC (PC ) + dGD (PD ) − a GA (PA ) − b GB (PB )
Since Δ sep G = Δr G , we have

¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
Δr G = c GC (PC ) + d GD (PD ) − a GA (PA ) − b GB (PB )
This page titled 13.4: The Gibbs Free Energy Change for Reaction at Constant Partial Pressures is shared under a CC BY-SA 4.0 license and was
13.5: ∆rG is the rate at which the Gibbs Free Energy Changes with The Extent of
Reaction
For the reaction a A + b B ⇌ c C + d D , let us call the consumption of a moles of A one “unit of reaction.” Δ G corresponds r
to the actual Gibbs free energy change for one unit of reaction only in the limiting case where the reaction occurs in an arbitrarily
large system. For a closed system of specified initial composition, n , n , n , and n , whose composition at any time is specified
o
A
o
B
o
C
o
D
by n , n , n , and n , the extent of reaction, ξ , is

A B C D
o o o o
nA − n nB − n nC − n nD − n
A B C D
ξ = −( ) = −( ) = =
a b c d
At constant pressure and temperature, every possible state of this system is represented by a point on a plot of Δ rG versus ξ . Every
such state is also represented by a point on a plot of G versus ξ . system
From the general result that (dG ) = 0 if and only if the system is at equilibrium, it follows that Δ G (ξ ) = 0 if and only
system
PT
r eq
if ξ specifies the equilibrium state. (We can arrive at the same conclusion by considering the heat exchanged for one unit of
eq
reaction in an infinitely large system at equilibrium. This process is reversible, and it occurs at constant pressure and temperature,
so we have Δ H = q , Δ S = q /T , and Δ G = q + T (q /T ) = 0 .)
r
rev
P r P
rev
r
rev
P P
rev
Below, we show that

∂Gsystem
Δr G (ξ) = ( )
∂ξ
PT
for any value of ξ . (In Section 15.9, we use essentially the same argument to show that this conclusion is valid for any reaction
among any substances.) Given this result, we see that the equilibrium composition corresponds to the extent of reaction, ξ , for eq
which the Gibbs free energy change for one unit of the reaction is zero
Δr G (ξ )= 0
eq
and
∂Gsystem
( ) =0
∂ξ
PT
So that the Gibbs free energy of the system is a minimum.

In the next section, we show that the condition Δ G (ξ ) = 0 makes it easy to calculate the equilibrium extent of reaction, ξ .
r eq eq
Given the stoichiometry and initial composition, the equation for Δ G (ξ ) specifies the equilibrium composition and the partial
r eq
pressures P , P , P , and P . This is the usual application of these results. Setting Δ G = 0 enables us to answer the question:
A B C D r
If we initiate reaction at a given composition, what will be the equilibrium composition of the system? Usually this is what we want
to know. The amount by which the Gibbs free energy changes as the reaction goes to equilibrium is seldom of interest.
To show that Δ G = (∂G/∂ξ)r for any reaction, it is helpful to introduce modified stoichiometric coefficients, ν , defined such
PT j
that ν > 0 if the j-th species is a product and ν < 0 if the j-th species is a reactant. That is, for the reaction
j j
a A + b B ⇌ c C + d D , we define ν = −a , ν = −b , ν = c , and ν
A = d . Associating successive integers with the
B C D
reactants and products, we represent the j-th chemical species as X and an arbitrary reaction as j
| ν1 | X1 + | ν2 | X2 + ⋯ + | νi | Xi ⇌ νj Xj + ⋯ + νω Xω
Let the initial number of moles of ideal gas Xj be n ; then n

o
j j =n
o
j
+ νj ξ . (For species that are present but do not participate in
the reaction, we have ν = 0 .) j
We have shown that the Gibbs free energy of a mixture of ideal gases is equal to the sum of the Gibbs free energies of the
components. In calculating Δ G, we assume that this is as true for a mixture undergoing a spontaneous reaction as it is for a
r
mixture at equilibrium. In doing so, we assume that the reacting system is homogeneous and that its temperature and pressure are
well defined. In short, we assume that the Gibbs free energy of the system is the same continuous function of temperature, pressure,
and composition, G = G (T , P , n , n , …) , whether the system is at equilibrium or undergoing a spontaneous reaction. For the
1 2
equilibrium system, we have (∂G/∂T ) P ,nj
= −S and (∂G/∂P ) T ,nj
=V . When we assume that these functions are the same for
a spontaneously changing system as they are for a reversible system, it follows that
∂G ∂G ∂G ∂G
dG = ( ) dT + ( ) dP + ∑ ( ) dnj = −SdT + V dP + ∑ ( ) dnj
∂T ∂P ∂nj ∂nj
P ,nj T ,nj j P ,T ,ni≠j j P ,T ,ni≠j
whether the system is at equilibrium or undergoing spontaneous change. At constant temperature and pressure, when pressure–
volume work is the only work, the thermodynamic criteria for change, dG ≤ 0 become TP
∂G
∑( ) dnj ≤ 0
∂nj
j P ,T ,ni≠j
When a reaction occurs in the system, the composition is a continuous function of the extent of reaction. We have
G = G (T , P , n + ν ξ, n + ν ξ, …) . At constant temperature and pressure, the dependence of the Gibbs free energy on the
o o
1 1 2 2
extent of reaction is
o
∂G ∂G ∂ (n + νj ξ)
J
( ) = ∑( ) ( )
o
∂ξ ∂ (n + νj ξ) ∂ξ
P ,T ,nm j J P ,T ,nm≠j P ,T ,nm≠j
Since
∂G ∂G ¯
¯¯¯
( ) =( ) = Gj
o
∂ (n + νj ξ) ∂nj
J P ,T ,nm≠j
P ,T ,nm≠j
and
o
∂ (n + νj ξ)
J
( ) = νj
∂ξ
P ,T ,nm≠j
it follows that
∂G ¯
¯¯¯
( ) = ∑ νj Gj = Δr G
∂ξ
P ,T ,nm j
Moreover, we have
∂G
(dG) =( ) dξ
PT
∂ξ
P ,T ,nm
The criteria for change, (dG) PT

≤0 , become
∂G
( ) dξ ≤ 0
∂ξ
P ,T ,nm
From our definition of ξ , we have dξ > 0 for a process that proceeds spontaneously from left to right, so the criteria become
∂G
( ) ≤0
∂ξ P ,T ,nm
or, equivalently,
¯
¯¯¯
∑ νj Gj = Δr G ≤ 0
This page titled 13.5: ∆rG is the rate at which the Gibbs Free Energy Changes with The Extent of Reaction is shared under a CC BY-SA 4.0
13.6: The Standard Gibbs Free Energy Change and Equilibrium in Ideal Gas
Reactions
Figure 5. Cycle demonstrating the relationship between Δ rG and ΔrG

o
.
The relationship between Δr G andis evident from the cycle in Figure 5. Since we have shown that
Δr G
o
Δ sepG (P , P , P , P ) = Δ G (P , P , P , P ) , we can consider the bottom equation in this cycle to represent the reaction
A B C D r A B C D
occurring in a mixture while calculating its free energy change as the free energy difference between pure products and pure
reactants. Since Δ G = 0,
cycle
o c d a B
Δcycle G = Δr G − Δr G + RT (ln p + ln p − ln p − ln p ) =0
C D A B
which can be rearranged to the result obtained in §2:

c d
p p
o C D
a b
p p
A B
Δr G is the Gibbs free energy change for one unit of the reaction occurring in a system whose composition is specified by P , A PB ,
PC , and P . In this spontaneously reacting system, the molar Gibbs free energy of ideal gas A is
D
¯
¯¯¯ o o
If the system is at equilibrium, P , P , P , and P are equilibrium pressures; these values characterize an equilibrium state. Then
A B C D
Δ G is the free energy change for a reaction occurring at equilibrium at constant pressure and temperature, and Δ G is zero. The
r r
equation
c d
p p
o C D
0 = Δr G + RT ln
a b
p p
A B
is exact. We have, when the partial pressures are those for a system at equilibrium,
c d
p p
o C D
Δr G = −RT ln
a b
p p
A B
Since Δ rG
o
is a constant, it follows that
c d
p p
C D
a b
p p
A B
is a constant. It is, of course, the equilibrium constant. We have

c d
p p
C D
KP =
a b
p p
A B
and
o
Δr G = −RT ln KP
or, solving for K P
o
Δr G
KP = exp (+ )
RT
Note that the value of the equilibrium constant is calculated from the Gibbs free energy change at standard conditions, not the
Gibbs free energy change at equilibrium, which is zero.
This page titled 13.6: The Standard Gibbs Free Energy Change and Equilibrium in Ideal Gas Reactions is shared under a CC BY-SA 4.0 license
In Section 13.6, we develop a relationship between the standard Gibbs free energy change for a reaction and the equilibrium
constant for that reaction. The standard Gibbs free energy change for a reaction is a specific quantity of energy, which depends only
on the temperature. In Chapter 11, we develop the Gibbs free energy of formation for a substance in its standard state. We assert
that it is convenient to set the Gibbs free energy of every substance in its standard state equal to its Gibbs free energy of formation.
Now we can see the reason: When we express the molar Gibbs free energy of an ideal gas as
¯
¯¯¯ PA
o o
GA (PA ) = Δf G (A, P ) + RT ln
o
P
¯
¯¯¯
GA (PA ) is the Gibbs free energy change for producing one mole of ideal gas A , at pressure P , from the elements at the same A
temperature. Because the Gibbs free energies of the reactants and products are measured from a common starting point, we can use
them to calculate the Gibbs free energy change for their reaction.
To demonstrate the role of our choice of the elements as the reference state for the Gibbs free energies of chemical substances, we
need only expand the Gibbs free energy cycle in Figure 4. In Figure 6, we introduce the Gibbs free energy change for producing the
separate components of the very large system.
Figure 6. Cycle demonstrating the relationship between Δsep G and Δr G for the formation of reactants and products from their
constituent elements.
Computing the Gibbs free energy change in a clockwise direction around the cycle in which the elements are converted first to
isolated reactants, then to components of a large equilibrium system, then to separated products, and finally back to the elements,
we find
0 =Δf G (initial) + Δmerge Ginitial
+Δr G (PA , PB , PC , PD ) + Δmerge Gf inal + Δf G (f inal)
Since the partial pressures PA , P , P , and P characterize the system at equilibrium, Δ G (P , P , P , P

B C D r A B C D) =0 . Also,
Δmerge Ginitial = Δmerge Gf inal = 0, so the equation for the Gibbs free energy change around the cycle simplifies to
0 = Δf G (initial) − Δf G (f inal)
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
= aGA (pA ) + b GB (pB ) − c GC (pC ) − dGD (pD )
o o o o o o o o a b c
= a Δf G (A, P ) + b Δf G (B, P ) − c Δf G (C , P ) − dΔf G (D, P ) + RT ln p + RT ln p − RT ln p
A B C
d
− RT ln p
D
c d
p p
o C D
= −Δr G − RT ln
a b
p p
A B
So that we have
c d
p p
o C D
Δr G = −RT ln
a b
p p
A B
To illustrate the application of these ideas, let us consider equilibrium in the oxidation of nitric oxide to nitrogen dioxide
2 N O + O2 ⇌ 2 N O2
At 800 K, the standard Gibbs free energy of formation of nitric oxide is +81.298 kJ mol , and that of nitrogen dioxide is
−1
+83.893 kJ mol . Hence, the standard Gibbs free energy change for the oxidation is +5, 190 J mol , and the equilibrium
−1 −1
constant is
o
Δr G
KP = exp (− ) = 0.458
RT
Suppose that one mole of N O and one mole of O are mixed, that the mixture is thermostated at 800 K, and that the applied
2 2
pressure remains constant at 10 bar while the reaction goes to equilibrium.

Let the number of moles of N O present at equilibrium be n . Then the moles of N O and O present at equilibrium are
NO 2 2
nN O2= 1−n NOand n = 1 + n /2 . The partial pressures are P

O2 NO =n RT /V , P =n NO RT /V ,
NO and N O2 N O2
PO2 =n RT /V , so that
O2
2
(1−nN O )
KP =
2
n (1 + nN O /2) (RT /V )
NO
The value of RT /V depends on the total pressure; we have
P = PN O + PN O2 + PO2 = (2 + nN O /2) (RT /V ) = 10 bar
so that
RT 20
=
V 4 + nN O
and
2
(1−nN O ) (4 + nN O )
KP = = 0.458
2
10 n (2 + nN O )
NO
Solving, we find n NO = 0.388 .
This page titled 13.7: The Gibbs Free Energy of Formation and Equilibrium in Ideal Gas Reactions is shared under a CC BY-SA 4.0 license and
Suppose that the very large equilibrium system with ideal gas components at pressures P , P , P , and P , also contains a
A B C D
quantity of liquid A . (For the present, we assume that this is pure liquid A ; components B , C , and D are insoluble in liquid A .) If
this augmented system is at equilibrium, we know that liquid A is in phase equilibrium with ideal gas A at pressure P . That is, A
PA is the vapor pressure of liquid A at the fixed temperature that we are considering, and the Gibbs free energy change for
converting liquid A to its ideal gas at P is zero. As long as the liquid is in phase equilibrium with its ideal gas, the relationship
A
between the ideal gas partial pressures and the standard Gibbs free energy change for the reaction is not affected by the presence of
the liquid. These same considerations apply when the very large equilibrium system contains both ideal gas A and solid A , so long
as the equilibrium sublimation pressure is equal to the partial pressure of ideal gas A, P . A
Now, let us suppose that substance A is a non-volatile liquid or solid. In this case, we may not be able to measure the standard
Gibbs free energy of formation of ideal gas A . From a practical standpoint, this is an important consideration; if we cannot find the
standard Gibbs free energy of formation of the ideal gas, we cannot use it to calculate equilibrium constants. From a theoretical
standpoint, it is less crucial; we can reasonably imagine that any substance has a finite vapor pressure at any temperature, even if
the value is much too small to measure experimentally. We can reason about the relationship of the equilibrium vapor pressure to
other quantities whether we can measure it or not.
If substance A is a non-volatile liquid (or solid) at the temperature of interest, it is useful to modify the cycle introduced in the
preceding section. We can find the standard Gibbs free energy of formation for the liquid from thermal measurements, and we can
use liquid A rather than ideal gas A as the standard state. If A is non-volatile, it is present in the equilibrium system only as the
liquid. For present purposes, we again assume that the other components are insoluble in liquid A ; then, the only difference
between A in its standard state and A in the equilibrium system is that the pressure on A in its standard state is one bar and the
pressure on liquid A in the equilibrium system is P = P + P + P . Therefore, when a moles of liquid A go from their
B C D
standard state to the equilibrium system, the Gibbs free energy change is
P
¯
¯¯¯
Δpress GA = a ∫ V A
(ℓ) dP
o
P
¯
¯¯¯
where V (ℓ) is the molar volume of pure liquid A . When a moles of liquid A are produced at the equilibrium pressure from the
A
elements, the Gibbs free energy change is

P
¯
¯¯¯ o ¯
¯¯¯
aGA (ℓ, P ) = a Δf G (A, ℓ) + a ∫ V (ℓ) dP
A
o
P
In Section 15.3, we give further attention to the value of this integral; for now, let us simply note that it is negligible in essentially
all circumstances. These considerations mean that we can modify the left side of the cycle in Figure 5 as indicated in Figure 7.
Figure 7. The Gibbs free energy of liquid A in an equilibrium system with ideal gases B , C , and D .
When we sum around the modified cycle in the same manner as before, we find
o o o o o
0 = a Δf G (A, ℓ) + b Δf G (B, P ) − c Δf G (C , P )
P
o o ¯
¯¯¯ b
−d Δf G (D, P )+a∫ V A
(ℓ) dP + RT ln p
B
o
P
c d
−RT ln p − RT ln p
C D
where we have supplemented our notation to emphasize that A is a non-volatile liquid while B , C , and D are ideal gases. We
continue to use Δ G to represent the difference between the standard Gibbs free energies of the products and those of the
r
o
reactants. In the present circumstances, we have

o o o o o o o o
Δr G = c Δf G (C , P ) + d Δf G (D, P ) − a Δf G (A, ℓ) − b Δf G (B, P )
Taking the value of the integral as zero, our result simplifies to

c d
p p
o C D
Δr G = −RT ln
b
p
B
We have again arrived at the conclusion that the “concentration” of a pure solid or liquid can be set equal to unity in the
equilibrium constant expression for a reaction in which it participates. When we do so, we must use the Gibbs free energy of
formation of the condensed phase in the calculation of Δ G . r
o
This page titled 13.8: Equilibrium When A Component is Also Present as A Condensed Phase is shared under a CC BY-SA 4.0 license and was
Finally, let us consider a very large equilibrium system that contains ideal gas components A , B , C , and D, at pressures P , P , A B
P , and P . We suppose that this system also contains a very large quantity of an inert solvent. This solvent is a liquid in which
C D
components A , B , C , and D are soluble. Let the concentrations of components A , B , C , and D in this solvent be [A], [B], [C ],
and [D]. Since this solvent-containing system is at equilibrium, we know that A dissolved in the solvent at concentration [A] is in
equilibrium with ideal gas A at pressure P . Since we assume that both the gas and the solution phases are very large, the transport
A
of a moles of A from one to the other does not significantly change any concentration in either phase. The Gibbs free energy
change is zero for this phase-transfer process. The Gibbs free energy change for the ideal-gas reaction and its equilibrium position
are unaffected by the presence of the solution.
In Chapter 6, we observe that a distribution equilibrium is characterized by an equilibrium constant; the ratio of the concentration
of a given chemical species in one phase to its concentration in a second phase is (approximately) a constant. In the present
instance, the partial pressure of component A is a measure of its gas-phase concentration and [A] is a measure of its solution-phase
concentration. Letting
o
PA / P pA
κA = ( ) =
[A] [A]
be the distribution constant for component A , we have p = κ [A] , and parallel relationships for components B ,
A A C , and D .
Substituting the distribution equilibrium relationships into the equilibrium constant equation, we have
c d c d c d
p p κ κ [C ] [D]
o C D C D
Δr G = −RT ln = −RT ln − RT ln
a b a b a b
p p κ κ [A] [B]
A B A B
Evidently, we can characterize the position of equilibrium in this system using either the pressure-based constant,
c d
p p
C D
KP =
a b
p p
A B
or the concentration based constant,

c d
[C ] [D]
KC =
a b
[A] [B]
The relationship between the pressure-based and the concentration-based constants is

c d
κ κ
C D
KC = ( ) KP
a b
κ κ
A B
In our present discussion, the concentrations can be expressed in any convenient units. The numerical value of the concentration-
based constant depends on the units of concentration and the values of the distribution-equilibrium constants as well as the standard
Gibbs free energy change for the reaction of the ideal gases.
When all of the reacting species are non-volatile, all of the reacting substances are present in the solution. The partial pressures of
these species in any gas phase above the solution is immeasurably small, and the Gibbs free energies of formation of the ideal gases
are not accessible by thermal measurements. Since Δ G for the ideal-gas reaction is not available, it cannot be used to find K .
r
o
P
Nevertheless, our thermodynamic model presumes that these parameters have finite—albeit immeasurable—values. The
concentration-based constant
c d
[C ] [D]
KC =
a b
[A] [B]
characterizes the position of equilibrium in solution even when the data to characterize the gas-phase equilibrium process are
immeasurable.
We have again arrived at the same function of concentrations to characterize the position of equilibrium. It is the same whether the
equilibrating species are present in the gas phase or in an inert-liquid solvent. To obtain this result, we have used our general
thermodynamic model for equilibrium, but we have made special assumptions about the properties of the reacting species. We have
assumed that the gases behave ideally and that the distribution-equilibrium constants can be expressed using the species’
concentrations. In Chapter 15, we return to this subject and develop a rigorous model for chemical equilibrium that does not require
these special assumptions.
This page titled 13.9: Equilibrium When An Ideal Gas Component is Also Present as A solute is shared under a CC BY-SA 4.0 license and was
13.10: Problems
1. Ethylene is the most important feedstock for the organic chemicals industry. The ethylene-production process with the lowest
processing cost is the “thermal cracking” of ethane to produce hydrogen and ethylene. The table below gives Δ G for ethane and f
o
ethylene at 1000 K, 1100 K, and 1200 K.
T, K C6 H6 C6 H4
1000 110,750 119,067
1100 132,385 127,198
1200 154,096 135,402
(a) Calculate Δ rG
o
and K at each temperature.
P
(b) At each temperature, calculate the extent of reaction when one mole of pure ethane reacts to reach equilibrium, while the
pressure of the system is maintained constant at one bar.
(c) At each temperature, calculate the extent of reaction when one mole of pure ethane reacts to reach equilibrium, while the
pressure of the system is maintained constant at 0.100 bar.
(d) To minimize side reactions, it is desirable to operate a cracking reactor at the lowest possible temperature. The feed to a
cracking reactor contains many moles of water (steam) for every mole of ethane. Steam is simply an inexpensive inert gas in this
system. Why is steam fed to cracking reactors?
2. A system initially contains one mole of ethane. The cracking reaction occurs while the system is maintained at a constant
pressure of 0.1 bar and a constant temperature of 1000 K.
(a) Write the equations for the molar Gibbs free energies of ethane, ethylene, and hydrogen as a function of the extent of reaction,
ξ.
(b) Write the equation for Δ G as a function of the extent of reaction. For a constant system pressure of 0.1 bar, calculate Δ
r
o
rG
o
for ξ = 0, 0.4, 0.7, 0.8, 0.9, and 1.0. Roughly, what extent of reaction corresponds to Δ G = 0 ? r
o
(c) Write the equation for the Gibbs free energy of the system, G , as a function of the extent of reaction, ξ . For a constant
system
system pressure of 0.1 bar, calculate G for ξ = 0, 0.4, 0.7, 0.8, 0.9, and 1.0. Roughly, at what extent of reaction is G
system a system
minimum?
(d) From the equation for G system in part (c), find (∂ G system /∂ξ)
T ,P ,nj
. (When the n are constant, the composition of the system
j
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
is fixed: G C6 H6 ,GC6 H4 , and G H2 are constants.)
3. At –56.6 C, the vapor pressure of solid C O is 5.18 bar. (This is the triple point.) At –78.5 C, the vapor pressure of solid C O is
2 2
1.01 bar.
(a) A cold bath is prepared by mixing solid C O2 with methanol in a Dewar flask that is open to the atmosphere. What is the
temperature of this cold bath?
(b) Use the Clausius-Clapeyron equation to estimate the enthalpy of sublimation of C O . 2
(c) Assume that we know the Gibbs free energy of formation for both solid and ideal-gas C O at a particular temperature, T. That
2
is, we know Δ G (C O , g, T ) and Δ G (C O , s, T ). What is the Gibbs free energy of gaseous C O as a function of pressure
f
o
2 f
o
2 2
at constant temperature, T?
(d) What is the Gibbs free energy of solid C O at a constant temperature, T?
2
(e) Assume that the variation of the Gibbs free energy of the solid with pressure is negligible. Using your answers to (c) and (d),
write the Gibbs free energy change, ΔG, for the process C O (solid, P , T ) → C O (gas, P , T ). If (P , T ) is a point of solid—
2 2
gas equilibrium, what is the value of ΔG? How is the equilibrium constant for this process related to P ? C O2
(f) Use the Gibbs-Helmholtz equation to estimate the enthalpy of sublimation of C O . 2
(g) Compare the equation you used in part (b) to the equation that you used in part (f).
4. Ethanol is manufactured by the addition of water to ethylene over a solid-acid catalyst. The process is known as “the hydration
of ethylene.” One mole of ethylene, 10 moles of water, and a quantity of solid catalyst are charged to a reactor. When equilibrium is
reached at 400 K, the system pressure is 10.00 bar and the conversion to ethanol is 64.17%. Assume that ethylene, water, and
ethanol behave as ideal gases. What is Δ G for the hydration of ethylene at 400 K?
r
o
5. At 298.15 K, the Gibbs free energies for formation of N O2 and N 2 O4 , in their hypothetical ideal gas standard states, are 51.3
and 99.8 kJ mol , respectively.
−−1
(a) At this temperature, what is the value of the equilibrium constant for the reaction N 2 O4 ⇌ 2 N O2 ?
(b) A 10 L sample of gas initially contains one mole of N O . The balance of the sample is N . The pressure and temperature of
2 4 2
the system are maintained constant at 3 bar and 298.15 K. How many moles of N O are present at equilibrium? What fraction of 2
the N O is converted to N O ?
2 4 2
(c) A 10 L sample of gas initially contains only one mole of N O at 298.15 K. The volume and temperature of the system are
2 4
maintained constant. What fraction of the N O is converted to N O ?

2 4 2
(d) A 10 L sample of gas initially contains only 10 mole of N O at 298.15 K. The pressure and temperature of the system are
−1
2 4

2 4 2
(e) A 10 L sample of gas initially contains only 10 mole of N O at 298.15 K. The volume and temperature of the system are
−1
2 4

2 4 2
6. At 800 K, the Gibbs free energies for formation of methanol and formaldehyde, in their hypothetical ideal gas standard states,
are –88.063 and –87.893 kJ mol , respectively.
−1
(a) At 800 K, what is the value of the equilibrium constant for the reaction C H 3 OH ⇌ C H2 O + H2 ?
(b) Initially, one mole of pure methanol occupies 1 m at 800 K. The volume and temperature of this system are maintained
3
constant. Assume that the presence of a catalyst allows this reaction to occur selectively. How many moles of formaldehyde are
present at equilibrium? What fraction of the methanol is converted to formaldehyde?
(c) The initial system in (b) is allowed to equilibrate at constant pressure and temperature. What fraction of the methanol is
converted to formaldehyde?
(d) Initially, one mole of hydrogen and one mole of methanol occupy 1 m at 800 K. The volume and temperature of this system
3
are maintained constant. What fraction of the methanol is converted to formaldehyde at equilibrium?
7. At 298.15 K, the Gibbs free energies of formation of Br and BrC l, in their hypothetical ideal gas standard states, are +3.1 and
2
–1.0 kJ mol , respectively. Why is Δ G (Br , H I G , 298.15 K) not zero? A 1 m vessel initially contains 0.100 mole of Br
−1
f
o
2
o 3
2
and 0.200 mole of C l at 298.15 K. The volume and temperature are maintained constant. What are the partial pressures of Br ,
2 2
C l , and BrC l at equilibrium?

2
8. Devise a non-isothermal process for the reversible mixing of two gases. (Suppose that one of the substances can be condensed to
a liquid, whose vapor pressure is negligible, at a temperature at which the other substance remains a gas.)
9. The following thermodynamic data are available:
o −1
Δf G (SO2 , g, 298.15 K) = −300.1 kJ mol
o −1
Δf G (SO3 , g, 298.15 K) = −371.1 kJ mol
o −1
Δf H (SO2 , g, 298.15 K) = −296.8 kJ mol
o −1
Δf H (SO3 , g, 298.15 K) = −395.7 kJ mol
(a) What are Δ rG

o
and ΔrH
o
for the reaction 2 SO 2 + O2 ⇌ 2 SO3 at 298.15 K?
(b) Estimate the temperature, T , at which the standard Gibbs free energy change for this reaction is zero.
0
(c) A mixture is prepared to contain 1.00 mole of SO and 1.00 mole of 2 O2 . The mixture is allowed to equilibrate at T0 and a
pressure of 10.0 bar. What fraction of the SO is converted to SO ?
2 3
(d) A mixture is prepared to contain 1.00 mole of SO and 10.0 mole of 2 O2 . The mixture is allowed to equilibrate at T0 and a
pressure of 10.0 bar. What fraction of the SO is converted to SO ?
2 3
10. At 400 K, the Gibbs free energies of formation for carbon monoxide and methanol are −146.341 kJ mol and −1
−148.509 kJ mol , respectively. What is Δ G for the reaction C O + 2 H ⇌ C H OH ? A mixture of carbon monoxide and
−1
r
o
2 3
hydrogen in the stoichiometric proportions is prepared. Assume that the presence of a catalyst allows this reaction to occur
selectively. When this mixture reaches equilibrium at 400 K and 1.00 bar, what fraction of the carbon monoxide is converted to
methanol?
11. At 500 K, the Gibbs free energies of formation for carbon monoxide and methanol are −155.412 kJ mol and −1
−134.109 kJ mol
−1
, respectively. What is Δ G for the reaction C O + 2 H ⇌ C H OH ? Assume that the presence of a
r
o
2 3
catalyst allows this reaction to occur selectively. What ratio of hydrogen to carbon monoxide must be charged to a reactor in order
for 50% of the carbon monoxide to be converted to methanol when the system reaches equilibrium at a pressure of 100.0 bar?
12. The reaction of carbon monoxide with water to produce carbon dioxide and hydrogen, C O + H O ⇌ C O + H , is known as
2 2 2
“the water-gas shift reaction.” It is used in the commercial manufacture of hydrogen. The Gibbs free energy change for this
reaction becomes less favorable as the temperature increases. At 1100 K, the standard Gibbs free energy change for this reaction is
o
Δ G = +0.152 kJ mol
r . Equal numbers of moles of carbon monoxide and water are charged to a reactor. What fraction of the
−1
water is converted to hydrogen when the system reaches equilibrium at 1100 K and 10.0 bar?
CHAPTER OVERVIEW
14: Chemical Potential - Extending the Scope of the Fundamental Equation

14.15: Problems
This page titled 14: Chemical Potential - Extending the Scope of the Fundamental Equation is shared under a CC BY-SA 4.0 license and was
1
We develop the thermodynamic criteria for change in a closed system without specifying the composition of the system. It is clear,
therefore, that the validity of the results is not restricted in any way by the composition of the system. For the most part, our
development proceeds as if the system is a single substance. However, the validity of the thermodynamic criteria for change is
independent of whether the system comprises a single substance in a single phase or multiple substances in multiple phases. The
criteria for reversible change are independent of whether the change involves interconversions among the substances comprising
the system, so long as these interconversions occur reversibly.
While we develop these criteria without specifying the composition of the system undergoing change, we use them to make
predictions about processes in which system compositions do change. In Chapter 12, we apply the Gibbs free energy criterion for
reversible change to equilibria between two phases of a pure substance. This application is successful because the thermodynamic
properties of one pure phase are independent of how much of any other pure phase is present.
In Chapter 13, we find a relationship between the standard Gibbs free energy change and the equilibrium constant for a reaction of
ideal gases. We also use the equation for the Gibbs free energy of an ideal gas to find the value of the Gibbs free energy change for
a spontaneous reaction of ideal gases at a constant temperature. This is a noteworthy result. Because we can find this value, we can
predict the spontaneous transformation of one non-equilibrium state to a second non-equilibrium state. These applications are
successful because ideal gas molecules do not interact with one another; the laws describing the behavior of an ideal gas do not
depend on the properties of other substances that may be present.
In general, the behavior of a system depends on the molecular characteristics of all of the components and on their concentrations.
For example, the pressure at which pure ice is in equilibrium with a salt solution depends not only on the temperature but also on
the choice of salt and its concentration. When it is expressed as a function of reagent concentrations, the equilibrium constant for a
reaction of real gases varies to some extent with the composition and with the total pressure of the system. Such observations are
consistent with the idea that the thermodynamic properties of a system depend on the attractive and repulsive forces among its
molecules. Such forces act between any pair of real molecules, whether they are molecules of the same substance or molecules of
different substances.
Our treatment of ideal-gas reactions shows that, if we can model the effects of compositional changes on the values of a system’s
thermodynamic functions, we can use our thermodynamic theory to predict spontaneous changes in chemical composition. To
extend to real substances the treatment that we develop for ideal gases in Chapter 13, we must find general relationships between a
system’s thermodynamic properties and its chemical composition. These relationships must reflect the dependence of the
thermodynamic properties of one substance on the molecular characteristics of the other substances present. To find them, we must
introduce additional inferences and assumptions. Since we want to describe spontaneous processes, these ideas must apply to non-
equilibrium systems.
To introduce these ideas, let us consider an open system that can undergo a spontaneous change. We begin by considering its
energy. We want to model any change in the energy, dE , of this system as a function of a sufficient set of independent variables.
Thus far, we have developed equalities that relate thermodynamic functions only for closed systems undergoing reversible change.
Consequently, when we now seek an equation for dE that is valid for an open, spontaneously changing system, our task is one of
scientific inference, not mathematical deduction from the reversible-process equations. Nevertheless, the reversible-process
equations are our primary resource.
For a closed, reversible system, we interpret the fundamental equation to mean that dS determines the effect of thermal processes
and that dV determines the effect of pressure–volume work on the energy of the system. We infer that entropy and volume changes
will play these same roles in the description of open, spontaneously changing systems. If λ different kinds of non-pressure–volume
work are possible, the energy of the system is also a function of λ generalized extensive thermodynamic variables, θ , θ ,…, θ ,…,
1 2 k
θ . If the system contains ω substances, adding or removing some amount of any of these substances changes the energy of the
λ
system. Evidently, the energy of the system is a function of the number of moles of each substance present. Let n , n ,…, n ,
1 2 ω
represent the number of moles of each component in the system. We can recognize these dependencies formally by writing
E = E (S, V , θ1 , θ2 , … , θλ , n1 , n2 , … , nω )
This formalism applies to both open and closed systems. If the system is open, all of these variables are independent; any one of
them can be changed independently of the others. If the system is closed, changes in the n are not independent, because the mass
j
of the system is constant. (Letting M be the molar mass of the j-th substance, we have 0 = ∑
¯¯¯¯
¯
j
ω ¯¯¯¯
1
¯
M j dnj .) If the energy is a
continuous and differentiable function of each of these variables, the total differential becomes
∂E ∂E
dE = ( ) dS + ( ) dV
∂S ∂V
V , θm , nj S, θm , nj
λ
∂E
+∑( ) dθk
∂θk
k=1 S,V , θm≠k , nj
ω
∂E
+∑( ) dnj
∂nj
j=1 S,V , θm , np≠j
(Inclusion of “θm≠k ” and “n ” in the subscripted variable lists indicates that the partial taken with respect to any θ is taken with
p≠j k
the other θ -values held constant, and the partial taken with respect to any of the n is taken with the other n -values held constant.)
j
We hypothesize that this total differential describes energy changes in any system whose energy, entropy, volume, work variables,
θ , and composition variables, n , vary in a continuous manner. In particular, we hypothesize that it describes the energy change
k j
that results from any change in composition, specified by the dn , whether the system is open or closed. All of these partial
j
derivatives are intensive variables. Each of them specifies how much the energy of the system changes when its associated
extensive variable undergoes an incremental change. We find it convenient to refer to the partial derivatives as potentials. In this
book, we restrict our attention to systems in which a given potential has the same value at every location within the system.
Since
∂E
( )
∂nj
S,V , θm , np≠j
is the energy change that occurs when the amount of the j-th chemical substance changes by dn , we call this partial derivative a j
chemical potential. We introduce the new symbol μ to denote the chemical potential of the j-th substance; that is,
j
∂E
μj = ( )
∂nj
S,V , θm , np≠j
The value of μ depends exclusively on the properties of the system. When the amount of the j-th substance changes by dn , the
j j
contribution to the energy change is dE = μ dn . When the system is open, the amount of the j-th substance can change either
j
j
because of a process that occurs entirely within the system or because some of the substance moves across the system’s boundary.
Reversible processes in closed systems

In Section 9.14, we find, for a reversible process in a closed system, that dE = T dS − P dV + dwN P V , where
λ
k=1
In our hypothesized total differential, the contribution that composition changes make to dE is
ω ω
∂E
∑( ) dnj = ∑ μj dnj
∂nj
j=1 S,V , θm , np≠j j=1
From the empirical fact that the energy change for a reversible process in a closed system can be specified by specifying dS , dV ,
and the dθ , it follows that
k
∑ μj dnj = 0
j=1
for any such process. Implicitly using this fact in Section 10.1, we find that (∂E/∂S) = T and (∂E/∂V ) = −P for reversible
V S
processes in which all of the work is pressure–volume work. We can extend the argument to a reversible process involving any
form of work in a closed system. It follows that
∂E
( ) =T
∂S
V , θm , nj
∂E
( ) = −P
∂V
S, θm , nj
∂E
( ) = Φk
∂θk
S,V , θm≠k , nj
and
λ
∂E
∑( ) dθk = dwN P V
∂θk
Substituting into
∂E ∂E
dE = ( ) dS + ( ) dV
∂S ∂V
V , θm , nj S, θm , nj
λ
∂E
+∑( ) dθk
∂θk
ω
∂E
+∑( ) dnj
∂nj
j=1 S,V , θm , np≠j
we have
ω
dE = T dS − P dV + dwN P V + ∑ μj dnj
j=1
with
ω
∑ μ dnj = 0
j
j=1
for any reversible process in a closed system.
Reversible processes in open systems

Let us suppose that an open system is undergoing a reversible change and that we instantaneously stop any further exchange of
matter between the system and its surroundings. Before this action, the system is open; thereafter, it is closed. The act of closure
does not change the system’s state functions. If, after closure, the system continues to change reversibly,
dE = T dS − P dV + dw N PV must be valid. Closure of the system introduces a constraint that reduces the number of
independent variables. However, S , V , and the θ remain independent after closure, and the act of closure cannot alter the laws that
k
describe the interaction of the system with its surroundings during a reversible process. The relationships (∂E/∂S) = T , V
(∂E/∂V ) = −P , and ∑ are valid for the closed system. We infer that they are also valid
λ
(∂E/∂ θ ) k dθ = dw k N PV
S k=1 S,V , θm≠k , nj
for the open system, so that the total differential for a system undergoing reversible change is
ω
j=1
whether the system is open or closed. This comparison implies that

ω
∑ μj dnj = 0
j=1
for any reversible process, whether the system is open or closed. (As noted below, however, our model implicitly assumes that
μ = μ̂ ; that is, μ has the same value whether the change dn is the result of internal transformation, movement of the j-th
j
j j j
substance across the system boundary, or a combination of the two.)
Spontaneous processes
We hypothesize that the same mathematical model describes energy changes for both spontaneous and reversible processes in both
open and closed systems, so long as the process occurs in such a way that all of the system’s state functions behave as continuous
variables. For reversible systems, we conclude that this model is
ω
j=1
By hypothesis, the model also describes spontaneous processes, so long as the state functions behave as continuous variables. We
base the development of our thermodynamic theory for spontaneous reactions on this hypothesis. It is, of course, the utility of the
resulting theory, not the logical force of the reasoning by which we reach it, that justifies our acceptance of the hypothesis.
In Section 9.19, we find that (dE) < dw
SV
for a spontaneous process, and we reach this conclusion by an argument that is
N PV
independent of the composition of the system. For a spontaneous process at constant entropy and volume, our model becomes
ω
(dE) = dwN P V + ∑ μ dnj

SV j
j=1
and since (dE) SV

< dwN P V , it follows that
ω
∑ μ dnj < 0
j
j=1
for any spontaneous process.

All of the variables that appear in our model for dE are properties of the system. The surroundings can affect the behavior of the
system by imposing particular values on one or more of the system’s thermodynamic functions. (For example, immersing a reacting
system in a constant-temperature bath ensures that the system reaches equilibrium only when T = T^ .) When the surroundings
impose particular values on the system’s thermodynamic functions, the system’s state functions must approach the imposed values
as the system approaches equilibrium.
Internal entropy and chemical potential

Substituting the internal and external entropies that we introduce in Section 9.16, our model for the incremental energy change
during a spontaneous composition change becomes
ω
dE = T di S + dq + dwP V + dwN P V + ∑ μj dnj
j=1
Our first-law equation is dE = dq + dw PV + d wN P V . It follows that

ω
T di S + ∑ μj dnj = 0
j=1
and we find
ω
1
T
j=1
The change criteria
di S ≥ 0
and
ω
∑ μj dnj ≤ 0
j=1
are equivalent.
This page titled 14.1: Dependence of the Internal Energy on the Composition of the System is shared under a CC BY-SA 4.0 license and was
Using the appropriate sets of independent variables, we can obtain similar expressions for dH , dA , dG, and dS under the
conditions that we assume in Section 14.1. Since entropy and pressure are the natural variables for enthalpy in a closed system, we
infer that a change in the enthalpy of any system can be expressed as a function of dS , dP , changes in the non-pressure–volume
work variables, dθ , and changes in the composition, dn . That is, from
k j
H = H (S, P , θ1 , θ2 , … , θλ , n1 , n2 , … , nω )
we have
∂H ∂H
dH = ( ) dS + ( ) dP
∂S ∂P
P , θm , nj S, θm , nj
λ
∂H
+∑( ) dθk
∂θk
k=1 S,P , θm≠k , nj
ω
∂H
+∑( ) dnj
∂nj
j=1 S,P , θm , np≠j
From H = E + P V , we have
dH = dE + P dV + V dP
for any system. Substituting our result from Section 14.1.

ω
dE = T dS − P dV + dwN P V + ∑ μ dnj
j
j=1
where
∂E
μj = ( )
∂nj
S,V , θm , np≠j
we have
ω
dH = T dS + V dP + dwN P V + ∑ μ dnj
j
j=1
Since we assume that both of these equations for dH describe open systems, all of the variables are independent, and the
corresponding coefficients must be equal. For systems that satisfy our assumptions, we have
∂H
T =( )
∂S
P , θm , nj
∂H
V =( )
∂P S, θm , nj
λ
∂H
dwN P V = ∑ ( ) dθk
∂θk
and
∂H ∂E
μj = ( ) =( )
∂nj ∂nj
S,P , θm , np≠j S,V , θm , np≠j
By parallel arguments, we find

ω
dA = −SdT − P dV + dwN P V + ∑ μ dnj

j
j=1
with
∂A
μj = ( )
∂nj
T ,V , θm , np≠j
and
ω
dG = −SdT + V dP + dwN P V + ∑ μj dnj
j=1
with
∂G
μj = ( )
∂nj
T ,P , θm , np≠j
Rearranging the result for dE , we find

ω
1 P dwN P V 1
dS = dE + dV − − ∑ μj dnj
T T T T
j=1
so that
∂S
μ = −T ( )
j
∂nj
E,V , θm , np≠j
and from the result for dH , we have

ω
1 V dwN P V 1
dS = dH − dP − − ∑ μj dnj
T T T T
j=1
so that
∂S
μ = −T ( )
j
∂nj
H ,P , θm , np≠j
Evidently, for any system whose condition corresponds to our assumptions, all of the relationships that we develop in Section 10.1
remain valid. The non-pressure–volume work can be expressed in terms of partial derivatives in several equivalent ways:
λ
∂E
dwN P V = ∑( ) dθk
∂θk
λ
∂H
= ∑( ) dθk
∂θk
k=1 S,P , θm≠k , nj
λ
∂A
= ∑( ) dθk
∂θk
k=1 V ,T , θm≠k , nj
λ
∂G
= ∑( ) dθk
∂θk
k=1 P ,T , θm≠k , nj
λ
∂S
= −T ∑ ( ) dθk
∂θk
k=1 E,V , θm≠k , nj
λ
∂S
= −T ∑ ( ) dθk
∂θk
k=1 H ,P , θm≠k , nj
Most importantly, the chemical potential and the criteria for change can be expressed in several alternative ways. We have
∂E ∂H
μj = ( ) =( )
∂nj ∂nj
S,V , θm , np≠j S,P , θm , np≠j
∂A ∂G
=( ) =( )
∂nj ∂nj
V ,T , θm , np≠j P ,T , θm , np≠j
∂S ∂S
= −T ( ) = −T ( )
∂nj ∂nj
E,V , θm , np≠j H ,P , θm , np≠j
For any system, open or closed, that satisfies the assumptions we introduce in §1, the criteria for change are
ω
∑ μ dnj ≤ 0
j
j=1
where the chemical potential is equivalently expressed as any of the partial derivatives above.
Let us review the scope and significance of these results. We develop the fundamental equation, dE = T dS − P dV + dw , by N PV
reasoning about the behavior of closed, reversible systems. For any process in a closed system, we develop the criteria for change
(ΔE) ≤ wN P V
SV
(ΔH ) ≤ wN P V
SP
(ΔA) ≤ wN P V
VT
(ΔG) ≤ wN P V
PT
^
(ΔS) ≥ −wN P V / T
EV
(ΔS) ^
≥ −wN P V / T
HP
and the corresponding relationships among differentials for incremental changes. Now we are extending these conclusions to
produce equations for the changes in the various thermodynamic functions when a system undergoes a spontaneous composition
change. To do so, we introduce the idea that composition variables must be included in a complete model for a thermodynamic
function. When we write
E = E (S, V , θ1 , θ2 , … , θλ , n1 , n2 , … , nω )
we assert that every set of values, {S, V , θ , θ , … ,θ , n , n , … , n }, corresponds to a state of the system in which the system
1 2 λ 1 2 ω
has a definite energy. (Of course, we are not asserting that every such set of values can actually be attained by the real system.
Many such sets correspond to hypothetical states. The particular sets of values that do correspond to realizable states of the system
lie on a manifold. Given one such set of values—one point on the manifold—our differential expressions specify all of the other
states that lie on the same manifold.)
When we assume that the various partial derivatives, (∂E/∂S) , (∂E/∂V ) , (∂E/∂ θ )
V , θk , nj
, and S, θk , nj k S,V θm≠k , nj
(∂E/∂ n ) j , exist, we are assuming that E is a smooth, continuous function of each of these variables. Since some of these
S,V , θk , np≠j
partial derivatives are synonymous with intensive variables, we are assuming that these intensive variables are well defined.
Moreover, since these intensive variables characterize the system as a whole, we are assuming that each intensive variable has the
same value in every part of the system . When we assume that μ = (∂E/∂ n )
1
j
has a well-defined value, we are
j
S,V , θm , np≠j
assuming that E is a continuous function of n ; we are assuming that μ exists for any arbitrary state of the system and not just for
j j
states undergoing reversible change.

ω
When we assume that an arbitrary change is described by the total differential dE = T dS − P dV + dw +∑ μ dn , we N PV j=1 j j
are going beyond our conclusion that this total differential describes paths of reversible change. We are asserting that it describes
any process in which a change in composition is the sole source of irreversibility. As a practical matter, we expect it to describe any
process that occurs in a system whose potential functions are well defined. By well defined, we mean, of course, that that they can
be measured and that the measurements are reproducible and consistent. We expect these conclusions to apply to multiple-phase,
open systems, so long as each potential has the same value in every phase. In thus assuming that we can expand the scope of the
fundamental equation, we are not modifying the change criteria that we develop in Chapter 9. Our conclusions that
(dE)
SV
= dw for a reversible process and (dE) < dw
N PV SV
for a spontaneous process in a closed system are not affected.
N PV
Again, while our arguments for them are compelling, these results are not rigorously logical consequences of our earlier
conclusions about reversible processes. As for any scientific principle, their validity depends on their predictive capability, not their
provenance.
This page titled 14.2: Dependence of Other Thermodynamic Functions on the Composition is shared under a CC BY-SA 4.0 license and was
Because they are easy to control in typical laboratory experiments, pressure, temperature, and the number of moles of each
component are the independent variables that we find useful most often. Partial derivatives of thermodynamic quantities, taken
with respect to the number of moles of a component, at constant pressure, temperature, and θ , are given a special designation; k
they are called partial molar quantities. That is,

∂E
( )
∂nA
P ,T , θm , np≠A
is the partial molar energy of component A ,

∂G
( )
∂nA
P ,T , θm , np≠A
is the partial molar Gibbs free energy, etc. All partial molar quantities are intensive variables.
Because partial molar quantities are particularly useful, it is helpful to have a distinctive symbol to represent them. We use a
horizontal bar over a thermodynamic variable to represent a partial molar quantity. (We have been using the horizontal over-bar to
mean simply per mole. When we use it to designate a partial molar quantity, it means per mole of a specific component.) Thus, we
write
∂E ¯
¯¯¯
( ) = EA
∂nA
P ,T , θm , np≠A
∂V ¯
¯¯¯
( ) =V A
∂nA
P ,T , θm , np≠A
∂G ¯
¯¯¯
( ) = GA
∂nA
P ,T , θm , np≠A
etc.
In Sectiona 14.1 and 14.2, we introduce the chemical potential for substance A , μ , and find that the chemical potential of
A
substance A is equivalently expressed by several partial derivatives. In particular, we have

∂G ¯
¯¯¯
μA = ( ) = GA
∂nA
P ,T , θm , np≠A
that is, the chemical potential is also the partial molar Gibbs free energy.
It is important to recognize that the other partial derivatives that we can use to calculate the chemical potential are not partial molar
quantities. Thus,
∂E ∂E
μA = ( ) ≠( )
∂nA ∂nA
S,V , θm , np≠A P ,T , θm , np≠A
¯
¯¯¯ ¯
¯¯¯
¯ ¯
¯¯¯ ¯¯
¯
That is, μA
≠ EA . Similarly, μ
A
≠ HA ,μ
A
≠ AA , and μ A
≠ −T S A .
We can think of a thermodynamic variable as a manifold—a “surface” in a multidimensional space. If there are two independent
variables, the dependent thermodynamic variable is a surface in a three-dimensional space. Then we can visualize the partial
derivative of the dependent thermodynamic variable with respect to an independent variable as the slope of a line tangent to the
surface. This tangent lies in a plane in which the other independent variable is constant. If the independent variables are pressure,
temperature, θ -values, and compositions, the slope of the tangent line at (P , T , θ , θ , … ,θ , n , n , … , n ) is the value of a
k 1 2 λ 1 2 ω
partial molar quantity at that point.

A more concrete way to think of a partial molar quantity for component A is to view it as the change in that quantity when we add
one mole of A to a very large system having the specified pressure, temperature, θ -values, and composition. When we add one
k
mole of A to this system, the relative change in any of the system’s properties is very small; for example, the ratio of the final
volume to the initial volume is essentially unity. Nevertheless, the volume of the system changes by a finite amount. This amount
approximates the partial molar volume of substance A . This approximation becomes better as the size of the system becomes
larger. We expect the change in the volume of the system to be approximately equal to the volume of one mole of pure A , but we
know that in general it will be somewhat different because of the effects of attractive and repulsive forces between the additional A
molecules and the molecules comprising the original system.
Partial molar quantities can be expressed as functions of other thermodynamic variables. Because pressure and temperature are
conveniently controlled variables, functions involving partial molar quantities are particularly useful for describing chemical
change in systems that conform to the assumptions that we introduce in §1. Because the chemical potential is the same thing as the
partial molar Gibbs free energy, it plays a prominent role in these equations.
To use these equations to describe a real system, we must develop empirical models that relate the partial molar quantities to the
composition of the system. In general, these empirical models are non-linear functions of the system composition. However, simple
approximations are sometimes adequate. The simplest approximation is a case we have already considered. If we can ignore the
attractive and repulsive interactions among the molecules comprising the system, the effect of increasing n by a small amount,
A
¯
¯¯¯
dnA, is simply the effect of adding dn moles of pure component A to the system. If we let E be the energy per mole of pure
A A
component A , the contribution to the energy of the system, at constant temperature and pressure, is
∂E ¯
¯¯¯
( ) dnA = E A dnA
∂nA
P ,T , θm , np≠A
In Chapter 12, we apply the thermodynamic criteria for change to the equilibria between phases of a pure substance. To do so, we
use the Gibbs free energies of the pure phases. In Chapter 13, we apply these criteria to chemical reactions of ideal gases, using the
Gibbs free energies of the pure gases. In these cases, the properties of a phase of a pure substance are independent of the amounts
of any other substances that are present. That is, we use the approximation
∂G ¯
¯¯¯
( ) dnA = GA dnA
∂nA
P ,T , θm , np≠A
albeit without using the over-bar or the bullet superscript to indicate that we are using the partial molar Gibbs free energy of the
pure substance. In Section 14.1, we develop the principle that ∑ μ dn ≤ 0 are general criteria for change that are applicable
ω
j=1 j j
not only to closed systems but also to open systems composed of homogeneous phases.
Thus far in this chapter, we have written each partial derivative with a complete list of the variables that are held constant. This is
typographically awkward. Clarity seldom requires that we include the work-related variables, θ , and composition variables, n , in
k j
this list. From here on, we usually omit them.
This page titled 14.3: Partial Molar Quantities is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
ω
Let us now apply our chemical-potential equilibrium criterion, ∑ μ dn = 0 , to a simple, closed system in which phases α and
j=1 j j
β of a single component are at equilibrium. The total number of moles, n , is fixed; we have n = n + n . Reversible change is α β
possible while the system remains at equilibrium. However, there is a stoichiometric constraint; growth in one phase is exactly
matched by shrinkage in the other; dn = dn + dn = 0 , so that dn = −dn . The equilibrium criterion becomes
α β α β
(μ − μ ) dn = 0 . If reversible change occurs, we have dn ≠ 0 , and it follows that μ

α β α α =μ . α β
We infer that this result must always be valid. That is, if a substance is present in two phases of an equilibrium system, the
chemical potential of the substance is the same in each phase; no matter what other processes involving the substance may occur.
This is so because all processes that occur in the system must be at equilibrium in order for the system as a whole to be at
equilibrium. In our discussion of chemical kinetics, we saw this conclusion expressed as the principle of microscopic reversibility.
In our development of Gibbs phase rule, we asserted that a distribution-equilibrium constant relates the concentrations of a species
in any two phases. Our present conclusion that the chemical potential of a species must be the same in each phase is a more
rigorous statement of the same principle.
For a process in which dn moles of a substance transfer spontaneously from phase β to phase α , we have dn = dn α = −dnβ . The
ω
criterion for spontaneous change, ∑ μ dn < 0 , becomes
j=1 j j
μα dnα + μβ dnβ = μα dnα − μβ dnα < 0
and it follows that μ < μ . We infer that this result, too, must always be valid. That is, if a substance spontaneously transfers
α β
between any two phases, the chemical potential of the transferring substance must be greater in the donor phase than it is in the
acceptor phase.
While ∑ μ dn ≤ 0 summarizes our criteria for change, it does not specify the process that occurs in the system. In order to
ω
j=1 j j
describe a particular chemical reaction, we must incorporate the constraints that the reaction stoichiometry imposes on the dn . For j
a description of the equilibrium system that does not require the size of the system to be specified, the position of equilibrium must
be given as a function of intensive variables. While the μ are intensive variables, the dn are not. When we include information
j j
about the stoichiometry of the equilibrium process, we obtain a criterion for equilibrium that is specific to that process.
Consider the reaction a A + b B ⇌ c C + d D in a closed system. Let dξ be the incremental extent of reaction. Using the
modified stoichiometric coefficients, ν , that we introduce in Section 13.5, the incremental extent of reaction becomes
j
dnA dnB dnC dnD

dξ = = = =
νA νB νC νD
and we can use the stoichiometric coefficients to express the incremental composition changes, dnj , as dn j = νj dξ . The criteria
for change become
ω
∑ μj νj dξ ≤ 0
j=1
For the reaction to occur left to right, we must have dξ > 0 , so that
ω
∑ μ νj ≤ 0
j
j=1
In this form, the criteria involve only intensive variables, and the stoichiometry of the equilibrium process is uniquely specified. In
most circumstances, there are additional relationships among the μ for which we may have little or no specific information.
j
To see the nature of these additional relationships, let us consider a chemical reaction involving three species, A + B ⇌ C , at
constant pressure and temperature. No other substances are present. There are two components; so there are three degrees of
freedom. Pressure and temperature account for two of the degrees of freedom. The chemical potential of any one of the
equilibrating species, say A , accounts for the third. Evidently, fixing pressure, temperature, and μ is sufficient to fix the values of
A
the remaining chemical potentials, μ and μ . μ must contain information about the other chemical potentials, μ and μ .
B C A B C
If the same reaction occurs in the presence of a solvent, there are three components and four degrees of freedom. In this case, two
of the chemical potentials can vary independently. Fixing pressure, temperature, and, say μ and μ , fixes μ and the chemical A B C
potential of the solvent. We can expect the chemical potential of the solvent to be nearly constant, and we often omit it when we
write dG for this system. This omission notwithstanding, the chemical potentials of the equilibrating species include information
about the chemical potential of the solvent. In general, the chemical potential of any component of an equilibrium system is a
function of the chemical potentials of all of the other components.
This page titled 14.4: Chemical Potentials and Stoichiometry is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
ω
We conclude that ∑ μ dn = 0 is a criterion for reversible change in any system. When the change involves equilibria among
j=1 j j
two or more phases or substances, it alters the number of moles of the components present. An extent of reaction,
ξ = (nj − n )/ νj
o
j
, characterizes the displacement of every such equilibrium. The magnitude of each incremental equilibrium
displacement is specified by composition changes, dn = ν dξ , and conversely. The criterion for reversible change becomes
j j
μ ν dξ = 0 . When this criterion is satisfied because ∑ μ ν = 0 , dξ is arbitrary, and the system can reversibly traverse a
ω ω
∑ j j
j=1 j j=1 j
range of equilibrium states. In other words, ∑ μ ν = 0 defines a Gibbsian equilibrium manifold.

ω
j=1 j j
We can also have a reversible process for which dξ = 0. If the process is reversible, the state of the system corresponds to a point
on the Gibbsian manifold, but dξ = 0 stipulates that the system cannot change: it must remain at the specified point on the
manifold. This corresponds to what we are calling a primitive equilibrium state. The system is constrained to remain in this state by
the nature of its interactions with its surroundings: the system may be isolated, or the surroundings may act to maintain the system
in a fixed state.
This page titled 14.5: ∑µjdnj = 0 and Primitive Vs. Gibbsian Equilibrium is shared under a CC BY-SA 4.0 license and was authored, remixed,
Let us now consider a closed, constant temperature, constant pressure system that is composed of open subsystems. Chemical
substances can pass from one subsystem to another, but they cannot enter or leave the system. We assume that our model for dG
applies in every subsystem. Each subsystem is at the same temperature and pressure. For the r-th subsystem,
Gr = Gr (P , T , θr,1 θr,2 , ⋯ , θr,λ , nr,1 , nr,2 , ⋯ , nr,ω )
For a physical system in which all of these assumptions correspond closely to physical reality, we have, for the r-th subsystem,
ω
dGr = −Sr dT + Vr dP + (dwN P V ) + ∑ μj dnr,j

r
j=1
For the closed system, we have

ω
∑ dGr = −dT ∑ Sr + dP ∑ Vr + ∑ (dwN P V ) + ∑ ∑ μj dnr,j

r
r r r r r j=1
Since Gibbs free energy, entropy, volume, and work are extensive variables, we have, for the closed system, dG = ∑
r
dGr ,
dS = ∑ S , V = ∑ V , and dw
r r = ∑ (dw N PV ) . Therefore,N PV
r r r r
dG = −SdT + V dP + dwN P V + ∑ ∑ μj dnr,j
r j=1
For any process that occurs in this closed system at constant pressure and temperature, we have (dG) PT
≤ d wN P V , and
ω
∑ ∑ μj dnr,j ≤ 0
r j=1
expresses the criteria for change in the closed system as a sum of conditions on the open subsystems.
Now let us consider the possibility that, for the ρ-th open subsystem, we have
ω
∑ μj dnρ,j > 0
j=1
If this were true, the sum over all of the subsystems could still be less than or equal to zero. In this case, the energy increase
occurring in the ρ-th subsystem would have to be offset by energy decreases occurring in the other subsystems. This is at odds with
the way that physical systems are observed to behave. To see this, let us suppose that the process is a chemical reaction. Then the
composition changes are related to the extent of reaction as dn = ν dξ . For the open subsystem, we have
ρ,j j ρ
∑ μj νj dξρ > 0
j=1
Now, we can alter the boundary of this subsystem to make it impermeable to matter, while keeping its state functions unchanged.
This change converts the open subsystem to a closed system, for which we know that
ω
∑ μj νj dξρ < 0
j=1
If the criterion for spontaneous change switches from ∑ μ ν dξ > 0 to ∑ μ ν dξ < 0 the sign of dξ must change. The
ω
j=1 j j ρ
ω
j=1 j j ρ ρ
ω
supposition that ∑ μ ν dξ > 0 is possible in an open subsystem implies that the direction of a spontaneous change in a closed
j=1 j j ρ
system can be opposite the direction of a spontaneous change in an otherwise identical open system. No such thing is ever
observed. We conclude that the criterion for spontaneous change,
ω
∑ μj dnj < 0
j=1
must be satisfied in every part of any system in which the various potentials are the same throughout. Since
ω
1
T
j=1
it follows that d iS >0 must also be satisfied in every part of the system.
This page titled 14.6: The Change Criteria in A System Composed of Subsystems is shared under a CC BY-SA 4.0 license and was authored,
At constant temperature and pressure, the chemical potential of component A is the contribution that one mole of A makes to the
Gibbs free energy of the system. The Gibbs free energy of the system is the sum of the contributions made by all of its components.
The chemical potential change that occurs during a reaction,
ω
Δr μ = ∑ μj νj
j=1
is the same thing as the Gibbs free energy change. To establish these points, we introduce Euler’s theorem on homogenous
functions. For simplicity, we consider systems in which only pressure–volume work occurs.
A function f (x, y, z) is said to be homogeneous of order n if
n
λ f (x, y, z) = f (λx, λy, λz)
Then,
d ∂f d (λx) ∂f d (λy) ∂f d (λz)
n n−1
[ λ f (x, y, z)] = nλ f (x, y, z) = ( ) +( ) +( )
dλ ∂ (λx) dλ ∂ (λy) dλ ∂ (λz) dλ
yz xz xy
∂f ∂f ∂f
= x( ) + y( ) + z( )
∂ (λx) ∂ (λy) ∂ (λz)
yz xz xy
Since this must be true for any λ , it must be true for λ = 1 . Making this substitution, we have Euler’s theorem for order n :
∂f ∂f ∂f
nf (x, y, z) = x ( ) + y( ) + z( )
∂x ∂y ∂z
yz xz xy
An extensive state function is homogeneous of order one in its extensive variables. In particular, we have
λG (P , T , nA , nB , nC , nD ) = G (P , T , λnA , λnB , λnC , λnD )
for any λ . (For, say, λ = 1/2 , this says only that, if we divide a homogeneous equilibrium system into two equal portions, all else
remaining constant, the Gibbs free energy of each portion will be half of that of the original system. The pressure and temperature
are independent of λ .) Taking the derivative with respect to λ , we find
d ∂G ∂G
[λG (P , T , nA , nB , nC , nD )] = G (P , T , nA , nB , nC , nD ) = nA ( ) + nB ( )
dλ ∂ (λ nA ) ∂ (λ nB )
PT PT
∂G ∂G
+ nC ( ) + nD ( )
∂ (λ nc ) ∂ (λ nD )
PT PT
Since this must be true for any λ , it must be true for λ = 1 . We find
G (P , T , nA , nB , nC , nD ) = μA nA + μB nB + μC nC + μD nD
The same equation follows from applying Euler’s theorem to other state functions. For example, viewing the internal energy as a
function of entropy, volume, and composition, we have
λE (S, V , nA , nB , nC , nD ) = E (λS, λV , λnA , λnB , λnC , λnD )
so that
d ∂E ∂E ∂E
[λE (S, V , nA , nB , nC , nD )] = E (S, V , nA , nB , nC , nD ) = S ( ) +V ( ) + nA ( )
dλ ∂ (λS) ∂ (λV ) ∂ (λ nA )
V S SV
∂E ∂E ∂E
+ nB ( ) + nC ( ) + nD ( )
∂ (λ nB ) ∂ (λ nc ) ∂ (λ nD )
SV SV SV
Setting λ = 1 , we have
E (S, V , nA , nB , nC , nD )
= T S − P V + μA nA + μB nB + μC nC + μD nD
which we can rearrange to
G = E − T S + P V = μA nA + μB nB + μC nC + μD nD
This equation describes any system whose thermodynamic functions are continuous functions of one another. Evidently, we can
model the thermodynamic functions of any such system by modeling the chemical potentials of its components. In the remainder of
this chapter, we develop this idea.
By definition, Δ G is the Gibbs free energy change for converting a moles of A and b moles of B to c moles of C and d moles of
r
D at constant pressure and temperature and while the composition remains constant at n , n , n , and n . That is,
A B C D
Δr G = G (nA , nB , nC + c, nD + d) − G (nA + a, nB + b, nC , nD )
where n A ≫ a ,n B ≫ b ,n
C ≫ c , and n
D ≫ d . Using the relationship we have just found between G and the n , we have i
Δr G = [ μA nA + μB nB + μC (nC + c) + μD (nD + d)]
− [ μA (nA + a) + μB (nB + b) + μC nC + μD nD ]
= c μC + dμD − aμA − dμd
= Δr μ
Since any other extensive thermodynamic function is also homogeneous of order one in the composition variables, a similar
relationship will exist between the change in the function itself and its partial molar derivatives. For example, expressing the
system volume as a function of pressure, temperature, and composition, we have
λV (P , T , nA , nB , nC , nD ) = V (P , T , λnA , λnB , λnC , λnD )
and a completely parallel derivation shows that the volume of the system is related to the partial molar volumes of the components
by
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
V (P , T , nA , nB , nC , nD ) = V A nA +V B nB +V C nC + V D nD
The volume change for the reaction is given by

¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
Δr V = c V C + dV D − aV A − bV B
This page titled 14.7: At Constant P and T, ∆rµ is the Change in Gibbs Free Energy is shared under a CC BY-SA 4.0 license and was authored,
An important relationship among the differentials of the chemical potentials for a system follows from the relationships we have
just developed. From the fact that the Gibbs free energy, G (P , T , n , n , n , n ), is homogeneous of order one in the
A B C D
composition variables, we find that the Gibbs free energy of the system is related to its partial molar derivatives by
G (P , T , nA , nB , nC , nD ) = μA nA + μB nB + μC nC + μD nD
The differential of the left hand side is
dG = −SdT + V dP + μ dnA + μ dnB + μ dnC + μ dnD

A B C D
and the differential of the right hand side is
dG = μ dnA + nA dμ +μ dnB + nB dμ +μ dnC + nC dμ +μ dnD + nD dμ

A A B B C C D D
Since these differential expressions must be equal, we have
−SdT + V dP = nA dμ + nB dμ + nC dμ + nD dμ
A B C D
for any change in this system.

While we have considered the particular case of a system containing the species A , B , C , and D, it is clear that the same
arguments apply to any system. For a system that contains ω species, we can write the result in general form as
ω
−SdT + V dP = ∑ nj dμj
j=1
This relationship is known as the chemical-potential Gibbs-Duhem equation. It is a constraint on the dμ that must be satisfied j
when any change occurs in a system whose thermodynamic functions are continuous functions of its composition variables. If
pressure and temperature are constant and this equation is satisfied, the system is at equilibrium—it is on a Gibbsian equilibrium
manifold—and the chemical-potential Gibbs-Duhem equation becomes
ω
0 = ∑ nj dμj
j=1
In the next two sections, we develop a particularly useful expression for dμ . We can obtain similar relationships for other partial
j
molar quantities. (See problems 14.2 and 14.3.) These relationships are also called Gibbs-Duhem equations. Because the derivation
requires only that the thermodynamic function be homogeneous of order one, the same relationships exist among the differentials
of the partial molar derivatives of any extensive thermodynamic function. For partial molar volumes at constant pressure and
temperature, we find
ω
¯
¯¯¯
0 = ∑ nj dV j
j=1
This page titled 14.8: Gibbs-Duhem Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
The chemical potential of a substance in a particular system is a function of all of the variables that affect the Gibbs free energy of
the system. For component A , we can express this by writing
μA = μA (P , T , n1 , n2 , … , nA , … , nω )
for which the total differential is

ω
∂μ ∂μ ∂μ
A A A
dμA = ( ) dT + ( ) dP + ∑ ( ) dnj
∂T ∂P ∂nj
P T j=1 PT
Recalling the definition of the chemical potential and the fact that the mixed second-partial derivatives of a state function are equal,
we have
∂μA ∂ ∂G ∂ ∂G ∂S ¯¯
¯
( ) =( ) ( ) =( ) ( ) = −( ) = −S A
∂T P
∂T P
∂nA TP
∂nA TP
∂T P
∂nA TP
Similarly,
∂μ ∂ ∂G ∂ ∂G ∂V ¯
¯¯¯
A
( ) =( ) ( ) =( ) ( ) =( ) =V A
∂P ∂P ∂nA ∂nA ∂P ∂nA
T T TP TP T TP
Thus, the total differential of the chemical potential for species A can be written as
ω
¯¯
¯ ¯
¯¯¯
∂μA
dμ = −S A dT + V A
dP + ∑ ( ) dnj
A
∂nj
j=1 PT
To illustrate the utility of this result, we can use it to derive the Clapeyron equation for equilibrium between two phases of a pure
substance. In Chapter 12, we derived the Clayeyron equation using a thermochemical cycle. We can now use the total differential
of the chemical potential to present essentially the same derivation using a simpler argument. Letting the two phases be α and β,
the total differentials for a system that contains both phases becomes
¯¯
¯ ¯
¯¯¯
∂μα ∂μα
dμα = −S α dT + V α dP +( ) dnα + ( ) dnβ
∂nα ∂nβ
PT PT
and
∂μ ∂μ
¯¯
¯ ¯
¯¯¯ β β
dμβ = −S β dT + V β dP +( ) dnα + ( ) dnβ
∂nα ∂nβ
PT PT
Since equilibrium between phases α and β means that μ = μ , we have also that dμ = dμ for any process in which the phase
α β α β
equilibrium is maintained. Moreover, α and β are pure phases, so that μ and μ are independent of n and n . Then α β α β
∂μα ∂μβ ∂μα ∂μβ

( ) =( ) =( ) =( ) =0
∂nα ∂nα ∂nβ ∂nβ
PT PT PT PT
Hence,
¯¯
¯ ¯
¯¯¯ ¯¯
¯ ¯
¯¯¯
−S α dT + V α dP = −S β dT + V β dP
and the rest of the derivation follows as before.
This page titled 14.9: The Dependence of Chemical Potential on Other Variables is shared under a CC BY-SA 4.0 license and was authored,
In Section 5.17, we use mathematical models for the rates of chemical reactions to find the equilibrium constant equation, which
specifies the equilibrium position of a reacting system. The variables in the reaction-rate models are the concentrations of reactants
and products. Consequently, these concentrations are also the variables in the equilibrium constant equation. We develop our
reaction-rate models from simple arguments about the dependence of collision frequencies on reagent concentrations. These
arguments ignore the possibility that the effects of intermolecular forces on collision frequencies can change as the composition of
the system changes. For a great many purposes, the concentration-based models describe experimental results with sufficient
accuracy. However, as we note in our introductory discussion in Section 1.6, the concentration-based equilibrium constant equation
proves to be only approximately constant: To obtain rigorously accurate results, we must introduce new quantities, having the
character of “corrected concentrations”, that we call activities.
We are now in a position to define the chemical activity of a substance more precisely. We introduce the activity concept in the
particular context of chemical equilibria, but its scope is broader. The activity of a component is a wholly new thermodynamic
quantity. We view it as the property that determines the chemical potential, μ , of a substance in a particular system at a given
temperature. As such, the activity directly links the properties of the substance to the behavior of the system. Introducing the
activity makes our equations for μ and dμ more compact. It also provides an avenue to relate qualitative ideas about the properties
of the substance—ideas that we may not be able to express in mathematical models—to the effects that the substance has on the
properties of the system.
The heuristic idea that an activity is a corrected concentration arises in the first instance because we use the pressure-dependence of
the Gibbs free energy of an ideal gas as the model for our definition of activity. In Section 15.7, we find that the equilibrium
constant for a chemical reaction can be specified using the activities of the reactants and products and that the form of the
equilibrium constant is the same for activities as it is for concentrations. In Chapter 16 we see that assuming ideal behavior for
solutions enables us to equate the activity and the mole fraction of a component in such solutions.
Our consideration of the chemical reactions of gases in Chapter 13 introduces the basic ideas that we use: The Gibbs free energy of
reaction is the same thing as the change in the chemical potential of the system; a reaction reaches equilibrium when the Gibbs free
energy of the system cannot decrease further; the Gibbs free energy of reaction is a function of the chemical potentials of the
reactants and products; and the chemical potential of a gas is a function of its partial pressure. In ideal gas mixtures, the chemical
potential of a component depends on the partial pressure of that component, but not on the partial pressures of other species that
may be present. The model we develop for ideal gases becomes unsatisfactory for real gases under the kinds of conditions that
cause the experimentally observed properties of gases to depart from ideal gas behavior—conditions in which intermolecular forces
cannot be neglected.
In Section 14.9, we observe that the chemical potential of substance A can be expressed as a function of temperature, pressure, and
its molar composition:
μA (P , T , n1 , n2 , … , nA , … , nω )
The total differential becomes

ω
¯¯
¯ ¯
¯¯¯
∂μA
dμA = −S A dT + V A dP + ∑ ( ) dnj
∂nj
j=1 PT
We can write this as

¯¯
¯ ¯
¯¯¯
dμA = −S A dT + V A
dP + (dμA )
PT
where
ω
∂μA
(dμA ) = ∑( ) dnj
PT
∂nj
j=1 PT
In general, the change in the chemical potential of A depends on partial derivatives with respect to the amount of every substance
present. The dependence on the amount of substance j is given by
∂μA
( )
∂nj
P ,T ,nm≠j
and the dependence on the amount of A is given by

∂μA
( )
∂nA
P ,T ,nm≠A
In a multi-component system whose components do not interact, (∂ μ /∂ n ) is not zero, but all of the other partial
A A P ,T ,nm≠A
derivatives (∂ μ /∂ n )
A j must, by definition, vanish. This is exemplified in an ideal-gas system, in which the chemical
P ,T ,nm≠j
potential of component A is given by

o o o
μA (PA ) = Δf G (A, P ) + RT ln(PA / P )
and the partial pressure of A depends only on the mole fraction of A in the system, P = x P . For an ideal gas at constant
A A system
temperature, changes in the chemical potential do not depend on the amounts of other substances that may be present:
(dμ ) = RT (dln PA )
A T T
In the special case that the system pressure is constant, this becomes:
(dμ ) = RT (dln xA )
A T T
foreshadowing our eventual recognition that mole fraction is the most natural concentration unit for theoretical modeling of the
chemical potential.
In real systems, the chemical potential of a component depends on its own concentration, but it also depends—more weakly—on
the concentrations of the other species present. To treat real systems adequately, we need a general method to express the chemical
potential of a component as a function of the component’s concentration, in a way that fully accounts for the effects of the other
species. Our grand strategy is to develop the activity of a component as an abstract, dimensionless quantity. We do this by choosing
a simple function to define the activity of a component in terms of its chemical potential. This makes it possible to express the
thermodynamic properties of any system as relatively simple functions of the activities of its components. To apply these results to
specific systems, we must then find empirical equations that express the component activities as functions of component
concentrations. So, from one perspective, the activity is merely a convenient intermediary in our overall effort to express the
chemical potential as a function of composition.
To begin expressing this strategy in mathematical notation, let us represent the activity and concentration of a component, A , as a
~
A
and c , respectively. For simplicity, we consider systems in which pressure–volume work is the only work. The chemical potential
A
is a function of pressure, temperature, and composition; that is μ = μ (P , T , c , c , c , …) . We suppose that we have a large
A A A B C
volume of such data available; that is, we have measured μ at many conditions represented by widely varying values of the
A
variables P , T , c , c , c , …. Our strategy is to find two empirical functions. The first empirical function expresses the chemical
A B C
potential of A as a function of its activity and the temperature of the system, μ = μ (T , a ~

) . The second expresses the activity
A A A
of A as a function of the concentrations of the species present and of the pressure and temperature of the system,
(P , T , c , c , c , …). The mathematical composition of these two functions expresses the chemical potential as a
~ ~
aA =a A A B C
function of P , T , c , c , c , …; that is, the mathematical composition is the function μ = μ (P , T , c , c , c , …) .

A B C A A A B C
If we focus only on finding a function, μ (P , T , c , c , c , …), that fits the data, we can let the dependence of chemical
A A B C
potential on activity be anything we please. All we have to do is fix up the activity function, a
~
=a
~
(P , T , c , c , c , …), so
A A A B C
that the mathematical composition of the two accurately reproduces the experimental data. This is just another way of saying that,
if all we want is an empirical correlation of the form μ = μ (P , T , c , c , c , …) , we do not need to introduce the activity at
A A A B C
all.
However, our goal is more exacting. We want to construct the activity function so that it behaves as much like a concentration as
possible. In particular, we want the activity of a component to become equal to its concentration in cases where intermolecular
interactions vanish—or become independent of the component’s concentration for some other reason. To achieve these objectives,
we must choose the form of the function μ = μ (T , a
A A
~
) carefully. Here the behavior of ideal gases provides a valuable template.
A
Our observation that the chemical potential is related to the partial pressure by the differential expression
(dμA )
T
= RT (dln P A ) suggests that it might be possible to develop the activity function using a relationship of the same form.
T
In fact, this strategy proves to be successful.

We relate the chemical potential of a component to its activity by the differential expression
~
(dμ ) = RT (dln aA )
A T T
The activity becomes a single function that captures the dependence of (dμ A
)
T
on the number of moles of each of the components
of the system:
ω
∂μA
~
RT (dln aA ) = ∑( ) dnj
T
∂nj
j=1 PT
If we hold the temperature constant, we can integrate between some arbitrary base state (in which the activity and chemical
# #
potential are a
~
A
and μ ) and a second state (in which these quantities are a
A
~
and μ ). That is, at any given temperature, we can
A A
select a particular state that we define to be the base state of A for the purpose of creating an activity scale. We denote the activity
# #
in the base state as a ~
A
. We let the chemical potential of A be μ when A is in its base state at this temperature. At a fixed
A
#
temperature, μ is fixed. However, even if we keep all the other variables the same, the properties of the base state change when
A
the temperature changes. This means that the chemical potential in the base state is a function of temperature—and only of
temperature.
At constant temperature, the relationship between chemical potential and activity in any other state becomes
μ aA
A
~
∫ dμA = RT ∫ d ln aA
# #
μ a
A A
or
~
# aA
μA = μ + RT ln( )
A
~#
a
A
where it is implicit in the last equation that the activity is a function of pressure, temperature, and the chemical composition of the
system. The temperature and chemical potential are state functions, so this expression requires that the activity be a state function
also.
It is convenient to let the activity be unity in the base state. When we do this, we give the base state another name; we call it the
activity standard state. When we stipulate that the activity in the base state is unity, we use a superscript zero to designate the
activity and chemical potential. That is, in the standard state, a
~
= 1 , and the value of the chemical potential of A is μ̃
o
A
when A is o
A
in this state. (In many cases, it turns out to be convenient to choose a standard state that does not correspond to any physical
condition that can actually be achieved. In such cases, the standard state is a hypothetical system, whose properties we establish by
mathematical extrapolation of measurements made on a real system in non-standard states.) With this convention, the chemical
potential and activity are related by the equation
o ~
μA = μ̃A + RT ln aA
at the specified temperature. We adopt this equation as our formal definition of the activity of component A . The activity and the
chemical potential depend on the same variables and contain equivalent information.
Note that we have done nothing to restrict the state we choose to designate as the activity standard state. This creates opportunity
for confusion, because it allows us to choose an activity standard state for A in a particular system that is different from the
standard state for pure A that we define for the tabulation of thermodynamic data for pure substances. This means that we choose
to let the meaning of the words “standard state” be context dependent. The practical effect of this ambiguity is that, whenever we
are dealing with a chemical activity, we must be careful to understand what state of the substance is being designated as the
standard state—and thereby being assigned chemical potential μ̃ and unit activity. Because the standard state is a fixed
o
composition, μ̃ is a function of temperature, but not of pressure or composition. At a given temperature, changes in the system
o
pressure or the concentration of A affect the chemical potential, μ , only by their effect on the activity, a
A
~
. A
To complete our program, we define a function that we call the activity coefficient, γ A
(P , T , cA , cB , cC , …) , of component A , in
a system characterized by variables P , T , c , c , c , …, by the equation
A B C
~
aA (P , T , cA , cB , cC , …) = cA γA (P , T , cA , cB , cC , …)
This definition places only one constraint on the form of γ ; the function γ (P , T , c , c , c , …) can be anything that
A A A B C
adequately accounts for the experimental data, so long as γ → 1 when the effects of intermolecular interactions are negligible.
A
Introducing the activity coefficient does not simplify the job of finding a suitable, empirical, activity function; it just imposes a
condition on its form. In Chapters 15 and 16, we find that standard states are best described using mole fraction or molality as the
concentration unit. (Note that mole fraction is dimensionless and molality is proportional to mole fraction for dilute solutes.)
In Section 16.6 we consider the use of molality when A is a solute in a solution whose other components are in fixed proportions to
one another. At very low concentrations of A , the environment around every A molecule is essentially the same. The chemical
potential of A is observed to be a linear function of ln –– m , because a small increase in the concentration of A does not
A
significantly change the intermolecular interactions experienced by A molecules. As the concentration of A increases further,
intermolecular interactions among A molecules become increasingly important, and the chemical potential ceases to be a simple
linear function of ln ––
m A
. The activity coefficient is no longer equal to one.
In the remainder of this chapter, we focus on some of the general properties of the activity function. In the next two chapters, we
develop a few basic applications of these ideas. First, however, we digress to relate the ideas that we have just developed about
activity to the ideas that we developed in Chapter 11 about fugacity.
This page titled 14.10: Chemical Activity is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
In Chapter 11, we introduce the fugacity as an alternative measure of the difference between the Gibbs free energy of one mole of a
pure gas in its hypothetical ideal gas standard state and its Gibbs free energy in any other state at the same temperature. This
definition makes the fugacity of a gas an intensive function of pressure and temperature. At a fixed temperature, the state of one
mole of a pure gas is specified by its pressure, and the fugacity is a function of pressure only. Fugacity has the units of pressure.
Giving effect to our decision to let the fugacity of the gas be unity when the gas is in its hypothetical ideal gas standard state
(H I G ) and using the Gibbs free energy of formation for the gas in this state as the standard Gibbs free energy for fugacity, we
o
define the fugacity of a pure gas, A , by the equation
¯
¯¯¯
fA (P )
o o
GA (P ) = Δf G (A, H I G ) + RT ln[ ]
o
fA (H I G )
¯
¯¯¯
For substance A in any system, the chemical potential is the partial molar free energy; that is, μ (P , T ) = G (P , T ) . Since the
A A
Gibbs free energy of formation is defined for one mole of pure substance at a specified pressure and temperature, it is a partial
molar quantity. When we elect to use the hypothetical ideal gas at a pressure of one bar as the standard state for the Gibbs free
energy of formation of the gas, we also establish the Gibbs free energy of formation of the hypothetical ideal gas in its standard
state as the standard-state chemical potential; that is, μ = Δ G (A, H I G ) . Hence, we can also express the fugacity of a gas by
o
A
f
o o
the equation
fA (P )
o o
μA (P ) = Δf G (A, H I G ) + RT ln[ ]
o
fA (H I G )
or
fA (P )
o
μA (P ) = μ + RT ln[ ]
A o
fA (H I G )
(pure real gas)

For a mixture of real gases, we can extend the definition of fugacity in a natural way. We want the fugacity of a component gas to
measure the difference between its chemical potential in the mixture, μ , and its chemical potential in its standard state, μ , where o
its standard state is the pure hypothetical ideal gas at one bar pressure. If gas A is a component of a constant-temperature mixed-
¯
¯¯¯ ¯
¯¯¯
gas system, we have dμ = V dP , where V is the partial molar volume of A in the system, and P is the pressure of the
A A A
system. Let us find μ in a binary mixture that contains one mole of A and n moles of a second component, B . Let the partial
A B
molar volume of B be V . The system volume is V = V + n V . The mole fractions of A and B are x = 1/(n + 1) and
¯
¯¯¯
B
¯
¯¯¯
A B
¯
¯¯¯
B A B
xB = n /(n
B B + 1) .
Figure 1. Finding the Gibbs free energy of a real gas in a mixture.
To find the change in μ , we need a reversible process that takes one mole of A in its standard state to a mixture of one mole A
A
with n moles of B , in which the pressure of the mixture is P . The four-step process described in Figure 1 answers our
B
requirements: One mole of A and n moles of B are separately expanded from their hypothetical ideal gas standard states, at P ,
B
o
to the arbitrary low pressures x P and x P , respectively. For this expansion, the change in the Gibbs free energy of one mole
A
∗
B
∗
of A , which remains in its hypothetical ideal gas state, is

∗ ∗
xA P xA P ∗
¯
¯¯¯ ¯
¯¯¯
RT xA P
Δexps GA = ∫ V dP = ∫ dP = RT ln( )
A o
P
o
P
o P P
Next, these low-pressure ideal gases are merged to form a mixture of one mole of A with n moles of B at the total pressure P . B
∗
¯
¯¯¯
For this merging process, Δ mergeG = 0 . Then, we suppose that the ideal gases become real gases in a mixture whose pressure is
A
¯
¯¯¯
P . Since this is merely a conceptual change, we have Δ
∗
concept GA =0 . Finally, we compress the mixture of real gases from P to ∗
¯
¯¯¯ ¯
¯¯¯
an arbitrary pressure, P . Since the volume of the mixture is V =V A +n V B , the Gibbs free energy change for this compression
B
of the mixture is
P P P
¯
¯¯¯ ¯
¯¯¯
Δcomp Gmixture = ∫ V dP = ∫ V A dP + nB ∫ V B dP
∗ ∗ ∗
P P P
We see that the Gibbs free energy change for the real-gas system is the sum of the Gibbs free energy changes for the components;
we have
P
¯
¯¯¯ ¯
¯¯¯
Δcomp GA = ∫ V A dP
∗
P
For this process, we have

fA (P )
o
μA − μ = RT ln[ ]
A o
fA (H I G )
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
= Δexps GA + Δmerge GA + Δconcept GA + Δcomp GA
∗ P P P
xA P ¯
¯¯¯
RT RT
= RT ln[ ]+∫ V A
dP − ∫ dP + ∫ dP
o
P P
∗
P
∗ P P
∗ P
where we have added and subtracted the quantity
P
RT
∫ dP
P
∗ P
Dividing by RT and evaluating the last integral, we find

o
μA − μ fA (P )
A
= ln[ ]
o
RT fA (H I G )
P ¯
¯¯¯
∗ o
V A 1 ∗
= ln xA + ln P − ln P +∫ ( − ) dP + ln P − ln P
P
∗ RT P
P is a finite pressure arbitrarily near zero. At very low pressures, real gas A behaves as an ideal gas; hence, at very low pressures,
∗
the partial molar volume of the real gas is well approximated by the partial molar volume of pure gas A . That is, we have
¯
¯¯¯¯
¯
¯
¯¯
V
¯
A
≈V
A
, and
¯
¯¯¯¯
¯
∗ ∗
P ¯
¯¯¯ P ⎛ V ⎞
V A 1 A 1
∫ ( − ) dP ≈∫ ⎜ − ⎟ dP
0
RT P 0
RT P
⎝ ⎠
≈0
where the approximation becomes exact in the limit as P

∗
→ 0 . Simplifying the natural logarithm terms and expanding the
integral, we obtain
o
μA − μ fA (P )
A
= ln[ ]
o
RT fA (H I G )
∗
P ¯
¯¯¯ P ¯
¯¯¯
xA P V A 1 V A 1
= ln +∫ ( − ) dP − ∫ ( − ) dP
o
P 0
RT P 0
RT P
P ¯
¯¯¯
xA P V A 1
= ln +∫ ( − ) dP
o
P 0
RT P
Defining the fugacity coefficient for A in this mixture, γ , by A
fA (P )
γA =
xA P
and recalling that f A

o
(H I G ) = P
o
, we use this result to find
P ¯
¯¯¯
fA (P ) V A 1
ln γA = ln( ) = ∫ ( − )dP
xA P 0 RT P
This differs from the corresponding relationship for the fugacity of a pure gas only in that the partial molar volume is that of gas A
in a mixture with other gases. This is a trifling difference in principle, but a major difference in practice. To find the fugacity of
pure A , we use the partial molar volume of the pure gas, which is readily calculated from any empirical pure-gas equation of state.
However, to experimentally obtain the partial molar volume of gas A in a gas mixture, we must collect pressure–volume–
temperature data as a function of the composition of the system. If we contemplate creating a catalog of such data for the mixtures
of even a modest number of compounds, we see that an enormous amount of data must be collected. Just the number of systems
involving only binary mixtures is large. For N compounds, there are N (N − 1)/2 binary mixtures—each of which would have to
be studied at many compositions in order to develop good values for the partial molar volumes.
Fortunately, practical experience shows that a simple approximation often gives satisfactory results. In this approximation, we
assume that the partial molar volume of gas A —present at mole fraction x in a system whose pressure is P — is equal to the A
partial molar volume of the pure gas at the same pressure. That is, for a binary mixture of gases A and B , we assume
¯
¯¯¯¯
¯
¯
¯¯¯
V A (P , xA , xB ) = V (P )
A
In this approximation, we have
¯
¯¯¯¯
¯
o P ⎛ V ⎞
μ −μ xA P 1
A A A
= ln +∫ ⎜ − ⎟dP
o
RT P 0
RT P
⎝ ⎠
and
¯
¯¯¯¯
¯
P ⎛ ⎞
fA (P ) V 1
A
ln γ = ln( ) = ∫ ⎜ − ⎟dP
A
xA P 0 RT P
⎝ ⎠
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
We make the same assumption for gas B . From Euler’s theorem on homogeneous functions, we have V = xA V A + xB V B .
Therefore, in this approximation, we have
¯
¯¯¯¯
¯ ¯
¯¯¯¯
¯
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
V = xA V A + xB V B ≈ xA V + xB V
A B
The last sum is the Amagat’s law representation of the molar volume of the gas mixture. We see that our approximation is
equivalent to assuming that the system obeys Amagat’s law. Physically, this assumes that the gas mixture is an ideal (gaseous)
solution. We discuss ideal solutions in Chapter 16. In an ideal solution, the intermolecular interactions between an A molecule and
a B molecule are assumed to have the same effect as the interactions between two A molecules or between two B molecules. This
differs from the ideal-gas assumption; there is no effect from the interactions between any two ideal-gas molecules.
This page titled 14.11: Back to the Fugacity- the Fugacity of A Component of A Gas Mixture is shared under a CC BY-SA 4.0 license and was
While we develop the fugacity concept by thinking about the Gibbs free energy of a pure real gas, our definition means that the
fugacity of a substance in any system depends only on the difference between the chemical potential in that system and its chemical
potential in the standard state. There is no reason to confine the definition of fugacity to gaseous systems. We generalize it: Our
defining relationship specifies the fugacity of any substance in any system as a function of the difference between its chemical
potential in that system and its chemical potential in its hypothetical ideal gas standard state.
Since the Gibbs free energy of an arbitrary system depends on the pressure, temperature, and composition of the system, the
fugacity of any component also depends on these variables. However, there is an additional constraint on the fugacity. At a
sufficiently low system pressure, the gases in any system behave ideally, and they obey Boyle’s law of partial pressures; the
integral in the fugacity-coefficient equation becomes zero, and the fugacity coefficient becomes unity. In the limit as P → 0 ,
fA → x P , where x
A is the mole fraction of A in the gas phase. At a sufficiently low pressure, any gas mixture behaves ideally,
A
and the fugacity of a constituent species becomes equal to its gas-phase partial pressure. (For a mixture of ideal gases, the fugacity
of a component is always equal to its partial pressure.)
In Section 14.10, we define the activity of component A in an arbitrary system by the relationship
o ~
μ = μ̃ + RT ln[ aA (P , T , cA , cB , cC , …)]
A A
where μ̃ is the chemical potential of A in an activity standard state in which we stipulate that the activity of A is unity. Since the
o
A
defining equations for activity and for fugacity are formally identical, the distinction between activity and fugacity lies in our
choices of standard states and in the facts that activity is dimensionless while fugacity has the units of pressure. If we use the
hypothetical ideal-gas standard state for activity and measure the concentration in bars, the practical distinction between activity
and fugacity vanishes. We can view the fugacity as a specialization of the activity concept.
In summary: The fugacity function proves to be a useful way to express the difference between the chemical potential of a
substance in two different states. Fugacity is measured in bars with the hypothetical ideal gas standard state as the reference state.
We add chemical activity to our list of useful thermodynamic properties because it extends the advantages of the fugacity
representation to non-volatile components of systems that contain condensed phases. The standard state for activity can be any
particular state of any convenient system that contains the substance. We define the activity of the substance in this reference
system to be unity. (As with the hypothetical ideal gas standard state, we often find it useful to define a hypothetical system as the
standard state for the activity.) The chemical potential of the substance in the standard-state system is, by definition, the standard
chemical potential, μ̃ , for this particular activity scale.
o
A
In the remainder of this chapter and in Chapters 15-17, we consider additional properties and applications of fugacity and chemical
activity. Before doing so however, we digress to observe that we can choose an alternate definition of activity: chemical activity
can be defined as a ratio of fugacities. Let us consider three systems that contain substance A .
The first is pure gas A in its hypothetical ideal gas standard state at temperature T , A (H I G , T ) . In this state, the fugacity of A is
o
unity, f (H I G , T ) = 1 bar . The chemical potential of A is the standard-state chemical potential, which we equate to the Gibbs
A
o
free energy of formation:

¯
¯¯¯ o o o o
GA (H I G , T ) = Δf G (A, H I G , T ) = μ
A
The second is a system that we define as the standard state for the activity of A at temperature T . Apart from convenience, there is
no reason to prefer any particular system for this role. We denote this system as A (activity standard state, T ) or, for short,
A (ss, T ) . By definition, the activity of A in this state is unity, a (ss) = 1 , and the chemical potential of A in this system is the
~ o
A
¯
¯¯¯
standard chemical potential for this particular activity scale: GA (ss, T ) = μ̃A
o
. We denote the fugacity of A in this state as
fA (ss) . By our definition of fugacity, we have
fA (ss)
o o
μ̃ =μ + RT ln[ ]
A A o
fA (H I G )
The third is an arbitrary system that we denote as A (P , T , c , c , c , …). We denote the chemical potential, fugacity, and activity
A B C
¯
¯¯¯
of A in this system as G (P , T , c , c , c , …) = μ (P , T , c , c , c , …), f (P , T , c , c , c , …), and
A A B C A A B C A A B C
(P , T , c , c , c , …), respectively. By our definition of fugacity, we have

~
aA A B C
fA (P , T , cA , cB , cC , …)
o
μ (P , T , cA , cB , cC , …) = μ + RT ln[ ]
A A o
fA (H I G )
and by our definition of activity,

o ~
μA (P , T , cA , cB , cC , …) = μ̃ + RT ln[ aA (P , T , cA , cB , cC , …)]
A
Figure 2. Relating the chemical activity of a substance to its activity.

Figure 2 summarizes the relationships among the Gibbs free energies of these three states of substance A . From the cycle in Figure
2, we have Δ G + Δ G = Δ G , so that
1 3 2
fA (ss, T ) fA (P , T , cA , cB , … . )
RT ln[ ] + RT ln [ ]
o
fA (H I G ) fA (ss, T )
fA (P , T , cA , cB , cC , …)
= RT ln[ ]
o
fA (H I G )
~
= RT ln[ aA (P , T , cA , cB , … . )]
and
fA (P , T , cA , cB , cC , …)
~
aA (P , T , cA , cB , cC , …) =
fA (ss, T )
That is, the activity of a substance in a particular system is always equal to its fugacity in that system divided by its fugacity in the
standard state for activity. From this relationship, it is evident that activity is always a dimensionless function of concentrations.
This page titled 14.12: Relating Fugacity and Chemical Activity is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Let us write (dμ ) to represent the differential of μ at constant pressure and temperature. From the general expression for dμ
A PT A A
and the definition of activity, we can write the total differential of the chemical potential of substance A in a particular system in
several equivalent ways
∂μA ∂μA
dμA = ( ) dP + ( ) dT + (dμA )P T
∂P ∂T
T P
ω
∂μA ∂μA ∂μA
=( ) dP + ( ) dT + ∑ ( ) dnj
∂P ∂T ∂nj
T P j=1 PT
¯
¯¯¯ ¯¯
¯ ~
=V A dP − S A dT + RT (d ln aA )
PT
~
∂ ln aA ∂μ
A ~
= RT ( ) dP + ( ) dT + RT (d ln aA )
PT
∂P ∂T
T P
In short, we have developed several alternative notations for the same physical quantities. From the dependence of chemical
potential on pressure, and because μ̃ is not a function of pressure, we have a very useful relationship:
o
A
~
∂μ ∂ln aA ¯
¯¯¯
A
( ) = RT ( ) =V A
∂P ∂P
T T
From the definition of activity and the dependence of chemical potential on temperature, we have:
o ~
∂μ ∂μ̃ ∂ln aA ¯¯
¯
A A ~
( ) =( ) + Rln aA + RT ( ) = −S A
∂T ∂T ∂T
P P P
From the dependence of chemical potential on the composition of the system, we have
ω
∂μA
~
∑( ) dnj = (dμ ) = RT (dln aA )
A PT PT
∂nj
j=1 PT
This last equation shows explicitly that the activity of component A depends on all of the species present. The effects of
interactions between A molecules and B molecules are represented in this sum by the term (∂ μ /∂ n ) . When the effects of A B PT
intermolecular interactions on the chemical potential are independent of the component concentrations, (∂ μ /∂ n ) = 0 , and A B PT
the only surviving term is (∂ μ /∂ n ) . If the interactions between A molecules and the rest of the system are constant over a
A A PT
range of concentrations of A , γ is constant over this range.

A
This page titled 14.13: Relating the Differentials of Chemical Potential and Activity is shared under a CC BY-SA 4.0 license and was authored,
Having found the activity of a component at one temperature, we want to be able to find it at a second temperature. The equation
developed in Section 14.13 does not provide a practical way to find the temperature dependence of a ~
or ln a
~
. We can obtain a A A
useful equation by rearranging the defining equation, taking the partial derivative of ln a with respect to temperature, and making
~
A
use of the Gibbs-Helmholtz equation:

~ o
∂ ln aA ∂ μ μ̃
A A
( ) = [ ( − )]
∂T ∂T RT RT
P P
o
1 ∂ (μ /T ) ∂ (μ̃ /T )
A A
= [( ) −( ) ]
R ∂T ∂T
P P
¯
¯¯¯
¯ ~o
HA HA
=− +
2 2
RT RT
¯
¯¯¯
¯ ~o
HA is the partial molar enthalpy of component A as it is present in the system. H is the partial molar enthalpy of component A in
A
~ o
its activity standard state. Since this standard state need not correspond to any real system, H can be the partial molar enthalpy of A
the substance in a hypothetical state.

¯
¯¯¯
¯ ~o
In general, it is not possible to find − (H A − HA ) by experimental measurement of the heat exchanged in a single-step process in
which A passes from its state in the system of interest to its activity standard state. Instead, we devise a multi-step cycle in which
we can determine the enthalpy change for each step. This cycle includes yet another state of substance A , which we call the
ref
¯
¯¯¯
¯
reference state and whose molar enthalpy we designate as H . We devise this cycle to find two enthalpy changes. One is the
A
enthalpy change that occurs when one mole of A passes from the system of interest to the reference state; this enthalpy change is
ref
represented by the difference − (H ¯

¯¯¯
¯
A
−H
¯
¯¯¯
¯
A
) . The other is the enthalpy change that occurs when one mole of A passes from the
~o ref
¯
¯¯¯
¯
activity standard state to the reference state. This enthalpy change is represented by the difference − (H A
−HA ) . The enthalpy
change we seek is the difference between these two differences:
¯
¯¯¯
¯ ~o ¯
¯¯¯
¯ ¯
¯¯¯
¯
ref ~o ¯
¯¯¯
¯
ref
− (H A − H A ) = − (H A − H A ) + (H A − H A )
Because we explicitly choose the reference state so that these differences are experimentally measurable, it is useful to introduce
¯
¯¯¯
still more terminology. We define the relative partial molar enthalpy of substance A , L , as the difference between the partial A
ref
molar enthalpy of A in the state of interest, ¯

¯¯
HA
¯
¯
, and the partial molar enthalpy of A in the reference state, ¯
¯
H
¯¯
¯
A
; that is,
ref ~o ~o ref ~o
¯
¯¯¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
LA = H A − H A and LA = H A − H A where is the relative partial molar enthalpy of substance A in the activity standard
LA
¯
¯¯¯ ~ o ~ ¯
¯¯¯ o
state. Clearly, the values of L and L depend on the choice of reference state. We choose the reference state so that L and L
A A A A
can be measured directly. The quantity that we must evaluate experimentally in order to find (∂ln a
~
/∂T ) becomes A P
¯
¯¯¯
¯ ~o ¯
¯¯¯ ~o
− (H A − H A ) = − (LA − LA )
To see how these ideas and definitions can be given practical effect, let us consider a binary solution that comprises a relatively low
concentration of solute, A , in a solvent, B , at a fixed pressure, normally one bar. We suppose that pure A is a solid and pure B is a
liquid in the temperature range of interest. We need to choose an activity standard state and an enthalpy reference state for each
substance. The most generally useful choices use the concept of an infinitely dilute solution. In an infinitely dilute solution, A
molecules are dispersed so completely that they can interact only with B molecules. Consequently, the energy of the A molecules
cannot change if additional pure solvent (initially at the same temperature and pressure) is added. Operationally then, we can
recognize an infinitely dilute solution by mixing it with additional pure solvent; if no heat must be exchanged with the
surroundings in order to keep the temperature constant, the original solution is infinitely dilute.
For the solvent, the concept of an infinitely dilute solution gives rise to the following choices, which are shown schematically in
Figure 3:
Figure 3. Cycles relating the molar enthalpy of a solvent to its molar enthalpies in the pure liquid, in an activity standard state, and
in an enthalpy standard state.
For the activity standard state of the solvent, B , we choose pure liquid B . Then, the activity of pure liquid B is unity at any
~ o
temperature. Letting H be the partial molar enthalpy of B in the activity standard state and H be the molar enthalpy of pure
B B
~o
liquid B , we have H B
=H
B
.
For the enthalpy reference state of the solvent, we choose B in an infinitely dilute solution. We represent the partial molar
ref
¯
¯¯¯
¯
enthalpy of B in this infinitely dilute solution by H B
.
For the solute, the infinitely dilute solution involves the following choices, which are shown schematically in Figure 4:
For the activity standard state of the solute, we choose the hypothetical solution in which the concentration of A is one molal,
and the activity of A is unity, but all of the effects of intermolecular interactions are the same as they are in an infinitely dilute
~ o
solution. We represent the partial molar enthalpy of A in the activity standard state by H . A
For the enthalpy reference state of the solute, we choose the infinitely dilute solution and designate the partial molar enthalpy of
ref
¯
¯¯¯
¯
A in this reference state by H . Since all of the intermolecular interactions are the same in the enthalpy reference state as they
A
are in the activity standard state, there can be no energy change when one mole of A goes from one of these states to the other. It
~o ref
¯
¯¯¯
¯
follows that H A
= HA .
It follows from these choices that the relative partial molar enthalpies of A and B in their activity standard states are
~o ~o ¯
¯¯¯
¯
ref
LA = H A − H A =0 and
~o ~o ¯
¯¯¯
¯
ref ∂ Δmix H
L =H −H = lim (− )
B B B
nA →0 ∂nB
P ,T ,nA
The relevant process for which we can measure an enthalpy change is the isothermal mixing of n moles of pure solid A with n A B
moles of pure liquid B to form a solution:

Δmix H
nA A (pure solid) + n B (pure liquid) A/B (solution, nA , nB )
B
→
Letting the enthalpy of mixing be Δ mix H and the enthalpy of the resulting solution be H , we have
H = Δmix H + nA H + nB H
A B
Then
¯
¯¯¯
¯
∂H ∂ Δmix H
HA = ( ) =( ) +H
A
∂nA ∂nA
PT PT
and
¯
¯¯¯
¯ ∂H ∂ Δmix H
HB = ( ) =( ) +H
B
∂nB ∂nB
PT PT
where, of course, H , Δ mix H , H , and H are all functions of n and n . The partial molar enthalpies in the reference states are
¯
¯¯¯
¯
A
¯
¯¯¯
¯
B A B
¯
¯¯¯
¯ ¯
¯¯¯
¯
the limiting values of H and H as n
A B A → 0 . That is,
¯
¯¯¯
¯
ref ∂ Δmix H
H = lim ( ) +H
A A
nA →0 ∂nA
P ,T ,nB
and
¯
¯¯¯
¯
ref ∂ Δmix H
HB = lim ( ) +H
B
nB →0 ∂nB
P ,T ,nA
When we base the enthalpy reference states on the infinitely dilute solution, we find for the solute
~o ref ~o ref
¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
− (H A − H A ) = − (H A − H A ) + (H A − H A )
¯
¯¯¯ ~o ¯
¯¯¯
∂ Δmix H ∂ Δmix H
= − (LA − LA ) = −LA = −( ) + lim ( )
∂nA nA →0 ∂nA
P ,T ,nB P ,T ,nB
and for the solvent
¯
¯¯¯
¯ ~o ¯
¯¯¯
¯ ¯
¯¯¯
¯
ref ~o ¯
¯¯¯
¯
ref
¯
¯¯¯ ~o ∂ Δmix H
− (H B − H B ) = − (H B − H B ) + (H B − H B ) = − (LB − LB ) = −( )
∂nB
P ,T ,nA
The temperature dependence of the activities becomes

~
∂ln aA 1 ∂ Δmix H ∂ Δmix H
( ) = [−( ) + lim ( ) ]
2
∂T P
RT ∂nA P ,T ,nB
nA →0 ∂nA P ,T ,nB
and
~
∂ln aB 1 ∂ Δmix H
( ) =− ( )
2
∂T RT ∂nB
P P ,T ,nA
To a good first approximation, we can measure Δmix H as a function of composition at a single temperature, determine ¯
¯¯
LA
¯
and
¯
¯¯¯ ~o
LB − LB at that temperature, and assume that these values are independent of temperature. For a more exact treatment, we can
~ ¯
¯¯¯ ¯
¯¯¯ o
measure Δ H as a function of composition at several temperatures and find L and L − L as functions of temperature. It
mix A B B
proves to be useful to define the relative partial molar heat capacity of A , to which we give the symbol, J , as the temperature ¯¯
¯
A
derivative of L :
¯
¯¯¯
A
¯
¯¯¯
¯¯
¯
∂LA
JA = ( )
∂T
P
To illustrate the use of these ideas, let us suppose that we measure the enthalpy of mixing of solute A in 1 kg water (solvent B ). We
make this measurement for several quantities of A at each of several temperatures between 273.15 K and 293.15 K. For each
experiment in this series, n is 55.51 mole and n is equal to the molality of A , m , in the solution. We fit the experimental data to
B A
––
empirical equations. Let us suppose that the enthalpy of mixing data at any given temperature are described adequately by the
equation Δ H = α m + α m and that α and α depend linearly on T according to
mix 1 2
2
1 2
–– ––
α1 = β11 + β12 (T − 273.15)
and
α2 = β21 + β22 (T − 273.15)
Then
∂ Δmix H ∂ Δmix H
( ) =( ) = α1 + 2 α2 m
∂nA ∂m ––
P ,T ,nB P ,T ,nB
––
and
∂ Δmix H
lim ( ) = α1
nA →0 ∂nA
P ,T ,nB
so that
¯
¯¯¯ ∂ Δmix H ∂ Δmix H
LA = ( ) − lim ( ) = 2 α2 m
∂nA nA →0 ∂nA ––
P ,T ,nB P ,T ,nB
and
~ ¯
¯¯¯
∂ ln aA LA 2α2 m
( ) =− =− ––
2 2
∂T P
RT RT
and
¯
¯¯¯
¯¯
¯ ∂LA ∂α2
JA = ( ) = 2 m( ) = 2 mβ22
∂T –– ∂T ––
P
P
In the experiments of this illustrative example, n is constant. This might make it seem that we would have to do an additional set
B
of experiments, a set in which n is varied, in order to find L , (∂ln a

B
~
/∂T ) , and J
¯
¯¯¯
B . However, this is not the case. Since L
B P
¯¯
¯
B
¯
¯¯¯
A
and L are partial molar quantities, we have

¯
¯¯¯
B
¯
¯¯¯ ¯
¯¯¯
nA dLA + nB dLB = 0
so that
¯
¯¯¯ nA ¯
¯¯¯
dLB = − ( ) dLA
nB
¯
¯¯¯
and we can find L from
B
m 2
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
–– 2mα2 m α2
LB (m) − LB (0) = LB (m) = ∫ −( –– ) d m = − ––
–– –– 55.51 –– 55.51
0
and
¯
¯¯¯ 2 2
¯¯
¯ ∂LB m ∂α2 m β22
–– ––
JB = ( ) =− ( ) =−
∂T 55.51 ∂T 55.51
P ,nA
P
This page titled 14.14: Dependence of Activity on Temperature- Relative Partial Molar Enthalpies is shared under a CC BY-SA 4.0 license and
14.15: Problems
Problems
1. When we express the energy of a system as a function of entropy, volume, and composition, we have
E = E (S, V , n , n , … , n ) . Since S and V are extensive variables, we have λE = E (λS, λV , λn , λ n , … , λ n ) . Find
1 2 ω 1 2 ω
(∂ (λE)/∂λ) . From this result, show that

SV
G = ∑ μ nj
j
j=1
2. When we express the energy of a system as a function of pressure, temperature, and composition, we have
E = E (P , T , n , n , … , n ) . Because P and T are independent of λ , λE = E (P , T , λn , λ n , … , λ n ) . Show that
1 2 ω 1 2 ω
ω
¯
¯¯¯
E = ∑ E j nj
j=1
3. From E= E (P , T , n 1, n2 , … , nω ) and the result in problem 2, show that

ω
∂H ∂V ∂V ∂V ¯
¯¯¯
[( ) +P( ) ] dT + [P ( ) +T( ) ] dP = ∑ nj dE j
∂T ∂T ∂P ∂T
P P T P j=1
Note that at constant pressure and temperature,

ω
¯
¯¯¯
∑ nj dE j = 0
j=1
4. If pressure and temperature are constant, E = E (n1 , n2 , … , nω ) and λE = E (λn1 , λ n2 , … , λ nω ) . Show that
ω ¯
¯¯¯
∑ n dE = 0 follows from these relationships.
j j
j=1
5. A solution contains n moles of component 1, n moles of component 2,

1 2 n3 moles of component 3, etc. Let
n = n + n + n +. . . The mole fraction of component j is x = n /n . Show that
1 2 3 j j
∂xj n − nj
( ) =
2
∂nj n
and, for j ≠ k ,
∂xj −nj
( ) =
2
∂nk n
What are
∂x1
( )
∂n1
and
∂x2
( )
∂n2
if the solution has only two components?

6. For any extensive state function, Y (P , T , n1 , n2 , … , nω ) , the arguments developed in this chapter lead, at constant P and T ,
to the equations
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
Y = n1 Y 1 + n2 Y 2 + ⋯ + nω Y ω
and
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
0 = n1 dY 1 + n2 dY 2 + ⋯ + nω dY ω
Where Y is the partial molar quantity (∂Y /∂ n
¯
¯¯¯
j j)
P ,T ,nm≠j
.
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
(a) Prove that 0 = x 1 dY 1 + x2 dY 2 + ⋯ + xω dY ω
(b) Prove that

¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
∂Y 1 ∂Y 2 ∂Y ω
0 = n1 ( ) + n2 ( ) + ⋯ + nω ( )
∂n1 ∂n2 ∂nω
(c) Prove that

¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
∂Y 1 ∂Y 2 ∂Y ω
0 = x1 ( ) + x2 ( ) + ⋯ + xω ( )
∂x1 ∂x2 ∂xω
7. The enthalpy of mixing is measured in a series of experiments in which solid solute, A , dissolves to form an aqueous solution.
These enthalpy data are represented well by empirical equations Δ H = α –– m +α m , α = β
––
mix +β (T − 273.15) and
1 2
2
1 11 12
α2 = β21 + β22 (T − 273.15) with

−1
β11 = 10.0 kJ molal
−2 −1
β12 = −0.14 kJ molal K
−1
β21 = −3.00 kJ molal
−2 −1
β22 = −0.040 kJ molal K
¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯¯
¯ ¯
¯¯¯ ¯
¯¯¯ ¯¯
¯ ¯¯
¯
Find L , L
A , J , and J
H2 O A H2 O as functions of ––
m and T . Find L , L
A A H2 O , J , and J
A H2 O for a one molal solution at 209 K. What
is the value of
~
aA (1 molal, 290 K)
ln
~
aA (1 molal, 273.15 K)
Notes
1
We can make other assumptions. It is possible to describe an inhomogeneous system as a collection of many macroscopic,
approximately homogeneous regions.
CHAPTER OVERVIEW
15: Chemical Potential, Fugacity, Activity, and Equilibrium

In Chapters 11-14, we define chemical potential, fugacity, and activity. We find numerous relationships among these quantities. In
Sections 15.1 and 15.2, we summarize the principal relationships between chemical potential and fugacity and between chemical
potential and activity. Thereafter, we introduce some basic ideas about the chemical potentials, fugacities, and activities of liquids,
solids, solvents, and solutes. We use these ideas to relate standard Gibbs free energy changes to fugacities and activities in systems
at equilibrium.
15.9: Problems
This page titled 15: Chemical Potential, Fugacity, Activity, and Equilibrium is shared under a CC BY-SA 4.0 license and was authored, remixed,
1
The third law and the fugacity of a pure real gas.
In Chapter 11, we introduce the fugacity as a measure of the difference between the molar Gibbs free energy of a real gas,
¯
¯¯¯
G (P , T ) at pressure P , and that of the pure gas in its hypothetical ideal-gas standard state at the same temperature. We choose the
standard Gibbs free energy of formation, Δ G (H I G , T ) , to be the Gibbs free energy of the real gas in its hypothetical ideal-gas
f
o o
standard state. Letting the gas be A , we find
f (P )
¯
¯¯¯ o o A
G (P , T ) = Δf G (H I G , T ) + RT ln[ ]
o
fA (H I G )
(real gas)
where the fugacity depends on pressure according to
P ¯
¯¯¯
f (P ) P ⎡V 1 ⎤
A A
ln[ ] = ln[ ] +∫ − dP
o o
fA (H I G ) P 0 ⎣ RT P ⎦
(real gas)
¯
¯¯¯
and V is the molar volume of the pure real gas. (In Chapter 14, we introduce a solid-bullet superscript to indicate that a particular
A
property is that of a pure substance.) Given Δ G (H I G , T ) and an equation of state for the real gas, we can calculate the
f
o o
fugacity and molar Gibbs free energy of the real gas at any pressure.
The fugacity of a pure ideal gas

For a pure ideal gas, we have
¯
¯¯¯
V A
1
− =0
RT P
(ideal gas)
The fugacity becomes equal to the ideal-gas pressure
f (P ) = P
A
(ideal gas)
and the Gibbs free energy relationship becomes
¯
¯¯¯ P
o o
GA (P , T ) = Δf G (A, P , T ) + RT ln[ ]
o
P
(ideal gas)
For pure gases, the system pressure that appears in these equations, P , is the same thing as the pressure of the gas.
The fugacity of an ideal gas in a mixture

In Chapter 13, we find that the molar Gibbs free energy of a component of an ideal gas mixture is unaffected by the presence of the
other gases. For an ideal gas, A , present at mole fraction x , in a system whose pressure is P , the partial pressure is P = x P .
A A A
Since the partial pressure is the pressure that the system would exhibit if only ideal gas A were present, the molar Gibbs free
energy of an ideal gas in a mixture is
¯
¯¯¯ xA P
o o
GA (xA , P , T ) = Δf G (A, P , T ) + RT ln[ ]
o
P
(ideal gas)
The chemical potential and fugacity of real gases
In Chapter 14, we introduce the chemical potential as the partial molar Gibbs free energy. The defining relationship is
¯
¯¯¯ ∂G
μA = GA = ( )
∂nA
P ,T ,ni≠A
(any substance in any system)

When the system is a pure substance, the chemical potential is identical to the Gibbs free energy per mole of the pure substance at
the same temperature and pressure. For the chemical potential of A in a system comprised of pure A , we can write
¯
¯¯¯ G dG
μ = GA = =
A
nA dnA
(any system comprised of pure A)

From Euler’s theorem, we find that the Gibbs free energy of any system is the composition-weighted sum of the chemical
potentials of the substances present:
ω
G = ∑ ni μi
i=1
For a pure real gas, the partial molar Gibbs free energy and the molar Gibbs free energy are the same thing; we also write
f (P )
o o A
μ (P , T ) = Δf G (A, H I G , T ) + RT ln[ ]
A o
fA (H I G )
(pure real gas A)

and introducing μ o
A
(T ) = Δf G
o o
(A, H I G , T ) , we write
f (P )
o A
μ (P , T ) = μ (T ) + RT ln[ ]
A A o
fA (H I G )
(pure real gas A)

Since μ , Δ G (A) , and f (H I G ) are defined to be properties of one mole of pure A , it is not necessary to include either the
o
A
f
o
A
o
solid-bullet superscript or the solid over-bar in these symbols.

In Section 14.11, we find that the partial molar Gibbs free energy of a component of a real-gas mixture is
fA (P )
o
μA (P , T ) = μ (T ) + RT ln[ ]
A o
fA (H I G )
(real gas A in a mixture)

where the fugacity of A , present at mole fraction x in a system whose pressure is P , is given by
A
P ¯
¯¯¯
fA (P ) xA P V A 1
RT ln[ ] = ln[ ] +∫ [ − ]dP
o o
fA (H I G ) P 0
RT P
(real gas A in a mixture)

where f A
o
(H I G ) = P
o
= 1 bar . The partial molar volume is a function of the system’s pressure, temperature, and composition;
that is,
¯
¯¯¯ ¯
¯¯¯
V A (P ) = V A (P , T , xA , xB , … , xω )
and the fugacity depends on the same variables,
fA (P ) = fA (P , T , xA , xB , … , xω )
If the system is a mixture of ideal gases,
V = (nA + nB + ⋯ + nω ) RT /P
and
¯
¯¯¯
V A = (∂V /∂ nA ) = RT /P
P T nm≠A
The integrand becomes zero, and the fugacity relationship reduces to the ideal-gas fugacity equation introduced in Chapter 13 and
repeated above.
The fugacity of a gas in any system is a measure of the difference between its chemical potential in that system and its chemical
potential in its hypothetical ideal-gas standard state at the same temperature. The chemical potential of A in a particular system,
μ , is the change in the Gibbs free energy when the amounts of the elements that form one mole of A pass from their standard
A
states as elements into the (very large) system as one mole of substance A .
This page titled 15.1: The Chemical Potential and Fugacity of a Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
To make predictions about processes involving substance A , we need information about the chemical potential of A . Introducing
the fugacity does not introduce new information; the fugacity is merely a convenient way to relate the chemical potential to the
composition of the system. The fugacity relationship is valid whether we can actually measure μ and μ (H I G ) or not. To use
A
o
A
o
the relationship for practical calculations, we must know both, of course.

We introduce the activity function to cope with situations in which we cannot measure the fugacity. For volatile liquids—or solids
—we can obtain the Gibbs free energy of formation for both the condensed phase and for the hypothetical ideal-gas standard state.
In Section 15.4, we consider the relationship between the two.
The chemical activity of substance A measures the change in the chemical potential when one mole of A in some arbitrarily chosen
standard state passes into a very large system of specified composition. We introduce
~
o
aA
μA = μ̃A + RT ln[ ]
~
aA (ss)
where, as always,
∂G
μ =( )
A
∂nA
P ,T ,ni≠A
We let the activity of A in the arbitrarily chosen standard state, designated “ss ”, be unity, so that a
~
(ss) = 1 and the chemical
A
potential of A in this standard state is μ̃ . The activity, a , is a function of the pressure, temperature, and composition of the
o
A
~
A
system.
While we are free to choose any standard state we please for the activity of a gas, the hypothetical ideal-gas standard state is the
most practical. In this case, the activity of a gas is given by
fA (P )
~
aA (P ) =
o
fA (H I G )
and
o o o
μ̃ =μ (H I G )
A A
Then the only difference between fugacity and activity is that fugacity has the units of pressure, whereas activity is dimensionless.
For any gas in any state, we have
o o
μ −μ (H I G ) fA (P )
A A ~
= ln[ ] = ln[ aA (P )]
o
RT fA (H I G )
For an ideal-gas mixture whose pressure is P and in which the mole fraction of A is x , we have A
fA (xA , P ) xA P
~
aA (xA , P ) = =
o o
fA (H I G ) P
(ideal gas mixture)
This page titled 15.2: The Chemical Potential and Activity of a Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
So far, we have investigated fugacity and activity only for gases. Let us now consider a system that consists entirely of substance A
present as either a pure liquid or a pure solid. We assume that the temperature is fixed and that the pressure on this condensed phase
can be varied. For our present discussion, it does not matter whether the condensed phase is a liquid or a solid. For specificity, let
us assume that it is a liquid. We can imagine that the pressure changes are effected with the pure substance contained in a cylinder
that is sealed by a frictionless piston. We ask how the fugacity and activity vary when the system pressure changes. Let the molar
¯
¯¯¯
volume of the pure condensed phase be V A
. Since the coefficient of isothermal compression is small for condensed phases, it is
¯
¯¯¯
often adequate to assume that V A
is a constant. We do so here.
We have developed several designations for the molar Gibbs free energy of the condensed phase. For pure liquid A at pressure P ,
we can write
¯
¯¯¯ ¯
¯¯¯
G (A, ℓ, P ) = GA (ℓ, P ) = μ (ℓ, P )
A
(pure liquid)
For the pure liquid at a pressure of one bar, it is convenient to let the molar Gibbs free energy of the liquid be equal to the molar
Gibbs free energy of formation. We let
¯
¯¯¯ o ¯
¯¯¯ o o o o
G (A, ℓ, P ) = GA (ℓ, P ) =μ (ℓ, P ) = Δf G (A, ℓ, P )
A
(pure liquid, P = 1 bar )

We also have several designations for the pressure dependence of the Gibbs free energy of this liquid. Since the system consists
entirely of the pure liquid, we have
¯
¯¯¯ ¯
¯¯¯
(∂ G (A, ℓ, P ) /∂P ) = (∂ GA (ℓ, P ) /∂P )
T T
= (∂ μ (ℓ, P ) /∂P )
A
T
= RT (∂ {ln[f (ℓ, P )]} /∂P )

A
T
~
= RT (∂ {ln[a (ℓ, P )]} /∂P )
A
T
¯
¯¯¯
=V A
(ℓ, P ) dP
We are free to choose any state of any system that contains A as the standard state for the activity of substance A . Often, it is
convenient to let this standard state be the pure liquid (or the pure solid) at a pressure of one bar. The activity of A is unity in the
¯
¯¯¯ ¯
¯¯¯
standard state. Taking the partial molar volume to be constant, V (ℓ, P ) = V , and integrating between one bar and an arbitrary
A A
pressure, P , we can express the pressure dependence of the Gibbs free energy of the pure liquid in several equivalent ways:
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯ o
GA (ℓ, P ) = G (A, ℓ, P ) − G (A, ℓ, P )
¯
¯¯¯ o
= G (A, ℓ, P ) − Δf G (A, ℓ)
o
=μ (ℓ, P ) − μ (ℓ, P )
A A
f (ℓ, P )
A
= RT ln[ ]
o
f (ℓ, P )
A
~
= RT ln[a (ℓ, P )]
A
¯
¯¯¯ o
=V A
(P − P )
From the last equations, we see that the activity and fugacity of the pure liquid vary with the system pressure as
¯
¯¯¯ o
f (ℓ, P ) ⎡ V (P − P ) ⎤
~ A A
a (ℓ, P ) = = exp
A
o ⎣ RT ⎦
f (ℓ, P )
A
¯
¯¯¯
At ordinary temperatures and pressures, V (P − P ) ≪ RT . In consequence, the system pressure must become much greater
A
o
than one bar before the exponential term becomes significantly different from one. Thus, the activity of a condensed phase is
approximately unity until the system reaches high pressures. At pressures near one bar, the fugacity of a condensed phase is only a
weak function of pressure.
This argument provides rigorous justification for treating the activities (or concentrations) of pure solids and liquids as constants
when we use equilibrium constant data to calculate the compositions of systems that are at equilibrium. (We address this issue
previously in Section 5.17 and Section 13.8.)
This page titled 15.3: The Pressure-dependence of the Fugacity and Activity of a Condensed Phase is shared under a CC BY-SA 4.0 license and
If substance A is a liquid at one bar and the temperature of interest, pure liquid A is the standard state for the calculation of the
enthalpy and Gibbs free energy of formation. From thermal measurements, we can find the standard Gibbs free energy of formation
of this liquid, Δ G (A, ℓ) = μ (ℓ, P ) . If we can measure the vapor pressure of the substance and find an equation of state that
f
o
A
o
describes the behavior of the real vapor, we can also find its fugacity and the standard Gibbs free energy of formation of its
hypothetical ideal gas, Δ G (A, H I G ) = μ (H I G ) . From the principle that the chemical potential of substance A is the
f
o o o
A
o
same in any two phases that are at equilibrium, it follows that the fugacity is the same in each phase.
If we choose the hypothetical ideal-gas standard state as the standard state for the activity of A , then the activity and fugacity are
the same thing, and the standard state chemical potential is the same thing as the Gibbs free energy of formation of the hypothetical
ideal gas.
o o o o
μ = μ̃ = Δf G (A, H I G )
A A
(activity standard state is the hypothetical ideal gas)

Alternatively, we can choose the pure liquid as the standard state for the activity of A . In this case, there are two further options:
We can choose the pure liquid either at one bar pressure, P , or at its equilibrium vapor pressure, P . If we choose the pure liquid
o
vp
at P , we have
o
o o o o
μ̃ =μ (ℓ, P ) = Δf G (A, ℓ, P )
A A
(activity standard state is the pure liquid at P ) o
In Chapter 16, we see that the pure liquid at Pvp proves to be the most generally useful choice for A in a solution. For the pure
liquid at P , we have
vp
o
μ̃ =μ (ℓ, Pvp )
A A
(activity standard state is the pure liquid at P ) vp
Evidently, it is useful to be able to relate the quantities Δf G

o
(A, H I G )
o
, μ
A
(ℓ, P
o
) = Δf G
o
(A, ℓ, P
o
) , and μ
A
(ℓ, Pvp ) to
one another.
The difference between the Gibbs free energy of formation of the ideal gas in its hypothetical ideal-gas state and the Gibbs free
energy of formation of the liquid in its standard state is a quantity that we can call the standard Gibbs free energy of vaporization,
G (A, P ) , because both the initial and final states are at a pressure of one bar. At an arbitrary temperature, a liquid at one
o o
Δ vap
bar is not at equilibrium with its own ideal gas at one bar, and the standard Gibbs free energy of vaporization is not zero. We have
o o o o o o
Δvap G (A, P ) = Δf G (A, H I G ) − Δf G (A, ℓ, P )
o
fA (H I G )
o o
=μ −μ (ℓ, P ) = RT ln[ ]
A A
o
f (ℓ, P )
A
Figure 1. Cycle that takes a pure liquid to its hypothetical ideal gas state.
Figure 1 describes a reversible process that takes one mole of A from the pure liquid state at one bar to its hypothetical ideal-gas
standard state. We first reversibly decrease the pressure applied to pure liquid A until we reach its equilibrium vapor pressure, P , vp
at the temperature of interest. The molar Gibbs free energy change for this process is
Pvp
o o o ¯
¯¯¯
Δpress G (A, ℓ) = μ (ℓ, Pvp ) − Δf G (A, ℓ, P ) =μ (ℓ, Pvp ) − μ (ℓ, P ) =∫ V A
(ℓ) dP
A A A
o
P
To reach the hypothetical ideal-gas standard state, we then reversibly and isothermally evaporate one mole of the liquid to its real
gas. The Gibbs free energy change for this reversible process is zero, Δvap G (A, Pvp ) = 0 . Finally, we isothermally and
reversibly expand the real gas to an arbitrarily low pressure, P , conceptually convert the real gas to an ideal gas, and compress
∗
this ideal gas from P to one bar. The Gibbs free energy change for these latter steps is
∗
⎡ o ⎤
fA (H I G )
Δpress G (A, g) = RT ln⎢ ⎥
⎣f (RG, Pvp ) ⎦
A
Pvp ¯
¯¯¯
Pvp ⎡V (g) 1 ⎤
A
= −RT ln[ ] − RT ∫ − dP
o
P 0 ⎣ RT P ⎦
so that we can also express the standard Gibbs free energy of vaporization as
o o
Δvap G (A, P ) = Δpress G (A, ℓ) + Δpress G (A, g)
P P ¯
¯¯¯
vp
Pvp vp ⎡V (g) 1 ⎤
¯
¯¯¯ A
=∫ V A
(ℓ) dP − RT ln[ ] − RT ∫ − dP
o
P
o P 0 ⎣ RT P ⎦
Equating expressions for Δ vap G

o
(A, P
o
) , we find
o
fA (H I G )
o o o o
Δf G (A, H I G ) − Δf G (A, ℓ, P ) = RT ln[ ]
o
f (ℓ, P )
A
Pvp Pvp ¯
¯¯¯
Pvp ⎡V (g) 1 ⎤
¯
¯¯¯ A
=∫ V (ℓ) dP − RT ln[ ] − RT ∫ − dP
A o
P
o P 0 ⎣ RT P ⎦
Thus, we can find the standard Gibbs free energy of formation of the hypothetical ideal gas from the standard Gibbs free energy of
formation of the liquid, the equilibrium vapor pressure of the pure substance, and the equation of state of the pure gas:
o o o o o o o
Δf G (A, H I G ) − Δf G (A, ℓ, P ) = Δf G (A, H I G ) − μ (ℓ, P )
A
Pvp Pvp ¯
¯¯¯
Pvp ⎡V (g) 1 ⎤
¯
¯¯¯ A
=∫ V (ℓ) dP − RT ln[ ] − RT ∫ − dP
A o
P
o P 0 ⎣ RT P ⎦
If the vapor of the substance behaves as an ideal gas, the last integral vanishes. If we also neglect the integral of the molar volume
of the liquid, we have
Pvp Pvp
o o o o o
Δf G (A, H I G ) = Δf G (A, ℓ, P ) − RT ln[ ] =μ (ℓ, P ) − RT ln[ ]
o A o
P P
(ideal gas)
and fA
(ℓ, P
o
) = Pvp .
In this development, we suppose that we know Δ G (A, ℓ, P ) from thermal measurements. We can calculate the Gibbs free
f
o o
energy difference between the liquid and hypothetical ideal-gas standard states if we have an equation of state for the vapor. The
chemical potential in the hypothetical ideal-gas standard state is
o o o o
μ (H I G ) = Δf G (A, H I G )
A
and the chemical potential of the pure liquid

o o o
μ (ℓ, P ) = Δf G (A, ℓ, P )
A
is expressed as a function of the fugacity of the pure liquid:

o
f (ℓ, P )
o o o A
μ (ℓ, P ) =μ (H I G ) + RT ln[ ]
A A o
fA (H I G )
The activity formalism provides an alternative way to express the same information. When we choose the pure liquid at one bar as
the activity standard state; we set a
~
(ℓ, P ) = 1 . For this activity scale, the standard chemical potential becomes the Gibbs free
A
o
energy of formation of the pure liquid, μ̃ (ℓ) = Δ G (A, ℓ, P ) . Since the chemical potential of the hypothetical ideal-gas
o
A f
o o
standard state is μ (H I G ) = Δ G (A, H I G ) , the activity relationship becomes

o
A
o
f
o o
o o o o o o o o o o ~ o
Δf G (A, H I G ) − Δf G (A, ℓ, P ) =μ (H I G ) − μ̃ (ℓ) = − (μ (ℓ, P )−μ (H I G )) = RT ln[ aA (H I G )]
A A A A
Comparison with the previous equation shows that

o
fA (H I G )
~ o
ln[ aA (H I G )] = ln[ ]
o
f (ℓ, P )
A
If the pure liquid is at equilibrium with a gas mixture at P in which the mole fraction of A is x , the fugacity of the pure liquid is
o eq
equal to the fugacity of the gas in the mixture; that is,

o eq o
f (ℓ, P ) = fA (RG, x ,P )
A A
so that
o
fA (H I G )
~ o
ln[ aA (H I G )] = ln[ ]
eq
o
fA (RG, x ,P )
A
and
o
fA (H I G )
~ o
aA (H I G ) =
eq
o
fA (RG, x ,P )
A
Finally, let us take the activity standard state to be pure liquid A at 1 bar and find the activity of A in an arbitrary real-gas mixture
whose pressure is P and in which the mole fraction of A is x . Let us represent this state as A (RG, x , P ) . The activity of real
A A
gas A in this state is

fA (RG, xA , P ) fA (RG, xA , P )
~
aA (RG, xA , P ) = =
eq o
f (ℓ, P
o
) fA (RG, x ,P )
A A
and the chemical potential is

o ~
μA (RG, xA , P ) = μ̃A (ℓ) + RT ln[ aA (RG, xA , P )]
If the real gas that is present at mole fraction x A in a system whose pressure is P can be treated approximately as an ideal gas,
these gas-fugacity terms can be approximated as
fA (RG, xA , P ) ≈ fA (I G, xA , P ) = xA P
and
eq o eq o
fA (RG, x ,P ) ≈ fA (I G, x ,P ) = fA (I G, Pvp ) = Pvp
A A
The activity becomes
~ ~
aA (RG, xA , P ) ≈ aA (I G, xA , P ) ≈ xA P / Pvp
and the chemical potential becomes
o
xA P
μA (RG, xA , P ) ≈ μA (I G, xA , P ) = μ̃A (ℓ) + RT ln( )
Pvp
This page titled 15.4: Standard States for the Fugacity and Activity of a Pure Solid is shared under a CC BY-SA 4.0 license and was authored,
The relationship between the standard Gibbs free energy of formation of a substance whose standard state is a solid and the Gibbs
free energy of the substance in its hypothetical ideal-gas standard state is essentially the same as described in the previous section
for a liquid. In each case, to find the fugacity of the condensed phase in its standard state, it is necessary to find a reversible path
that takes the condensed-phase substance to its hypothetical ideal-gas standard state. If the solid substance has a significant vapor
(sublimation) pressure, the paths described for a liquid in the previous section are also available for the solid. Otherwise, it may be
possible to determine the Gibbs free energy change along some more complicated path.
Of course, whether the Gibbs free energy of formation for the hypothetical ideal-gas standard state can be evaluated or not, the
fugacity and activity relationships remain valid. For substance A in another state—in which A need not be pure and the pressure is
generally not one bar—we have
fA (P , T , xA , xB , … . )
o
μA = μ + RT ln[ ]
A o
fA (H I G )
where μ = Δ G (A, H I G , T ) , and the fugacity of substance A , f (P , T , x , x , … . ), is simply an alternative expression

o
A
f
o o
A A B
of the difference between the chemical potential of the substance, as it occurs in the system, and its chemical potential in the
hypothetical ideal-gas standard state. We write f (P , T , x , x , … . ) to indicate that the state of the system is specified by its
A A B
pressure, temperature, and composition.

When the fugacity is difficult to measure, the activity function becomes essential. Choosing the standard state for the activity to be
pure solid A in the same standard state that we use for the Gibbs free energy of formation, we have
o o
μ̃A (s) = Δf G (A, s)
and
o
f (s, P )
o o A
μ̃ (s) = μ + RT ln[ ]
A A o
fA (H I G )
Then, for substance A in an arbitrary state at the temperature of interest:

fA (P , T , xA , xB , … . )
o
μA = μ + RT ln[ ]
A o
fA (H I G )
o
f (s, P ) fA (P , T , xA , xB , … . )
o A
= μ̃A (s) − RT ln[ ] + RT ln[ ]
o o
fA (H I G ) fA (H I G )
o
fA (P , T , xA , xB , … . )
= μ̃ (s) + RT ln[ ]
A
o
f (s, P )
A
o ~
= μ̃ (s) + RT ln aA
A
From one perspective, the activity function is simply a mathematical device that expresses the chemical potential relative to an
arbitrarily chosen reference state. If we can measure this difference experimentally, we can find a ~
whether we can measure the A
fugacity of A or not. For clarity, we designate the chemical potential in the reference state as μ̃ (s) . When we let μ̃ (s) be the o
A
o
A
Gibbs free energy of formation of the substance in the arbitrarily chosen reference state and let the activity in this reference state be
unity, μ is the chemical potential difference between the substance in the state of interest and the chemical potential of its
A
constituent elements in their standard states at the same temperature.
This page titled 15.5: The Chemical Potential, Fugacity, and Activity of a Pure Solid is shared under a CC BY-SA 4.0 license and was authored,
In Chapter 13, we develop the relationship between the standard Gibbs free energy change for a reaction and the equilibrium
constant for that reaction, under the assumption that all of the substances involved in the reaction behave ideally. In the gas phase,
they behave as ideal gases; when dissolved in a solution, their concentrations are proportional to their mole fractions in a gas phase
at equilibrium with the solution.
We can now repeat this development using the fugacities instead of the pressures of the reacting species. Let the reaction be
aA + bB → cC + dD.
We introduce the reactions that create the reactants and the products in their hypothetical ideal-gas standard states from their
elements in their standard states. The Gibbs free energy change for creating the reactants in their hypothetical ideal-gas standard
states from the elements in their standard states is
o o o o
aΔf G (A, H I G ) + b Δf G (B, H I G )
The Gibbs free energy change for creating the products in their standard states, from the same set of elements, is
o o o o
c Δf G (C , H I G ) + dΔf G (D, H I G )
Next, we introduce a set of processes, each of which adds a further quantity of a reactant or product to a very large system at
equilibrium. That is, we transfer an additional a moles of pure A from its hypothetical ideal-gas standard state into a very large
system in which its fugacity is f (P , T , x , x , … . ). The Gibbs free energy change for this process is
A A B
fA (P , T , xA , xB , … . )
a RT ln[ ]
o
fA (H I G )
Corresponding processes add b moles of pure B at fugacity f (P , T , x , x , … . ), etc. These processes are diagrammed in
B A B
Figure 2. Since the fugacity of a substance in any state is a rigorous measure of the difference between its chemical potential in that
state and its chemical potential in its hypothetical ideal-gas standard state, these Gibbs free energy changes are exact. Since the
very large system is at equilibrium, there is no Gibbs free energy change when a moles of A and b moles of B react according to
aA + bB → cC + dD .
Figure 2. Relating fugacities in an equilibrium system to the Gibbs free energies of formation of the components in their
hypothetical ideal gas states.
Let us compute the Gibbs free energy change in a clockwise direction around the reversible cycle in Figure 2. The elements in their
standard states are converted first to isolated products, then to components of the large equilibrium system, then to separated
reactants, and finally back to the elements in their standard states. We have
o o o o o o o o
0 = c Δf G (C , H I G ) + d Δf G (D, H I G ) − a Δf G (A, H I G ) − b Δf G (B, H I G )
fC (P , T , xA , xB , … . ) fD (P , T , xA , xB , … . ) fA (P , T , xA , xB , … . )
+ c RT ln[ ] + d RT ln[ ] − a RT ln[ ]
o o o
fC (H I G ) fD (H I G ) fA (H I G )
fB (P , T , xA , xB , … . )
− b RT ln[ ]
o
fB (H I G )
To express the fugacity ratios more compactly, we let

fA (P , T , xA , xB , … . )
fA =
o
fA (H I G )
etc. Letting
o o o o o o o o o
Δr G = c Δf G (C , H I G ) + d Δf G (D, H I G ) − a Δf G (A, H I G ) − b Δf G (B, H I G )
the Gibbs free energy around this reversible cycle simplifies to

c d
f f
o C D
Δr G = −RT ln
a b
f f
A B
We can express the criterion for equilibrium as

c d
f f
C D
Kf =
a b
f f
A B
where
o
−Δr G
Kf = exp ( )
RT
We introduce the subscript, “f ”, to indicate that the equilibrium constant, K , is a function of the fugacities of the reacting
f
substances. These relationships parallel those that we found for equilibrium among ideal gases, with real-gas fugacities replacing
ideal-gas pressures.
As we did when we considered ideal-gas equilibria, let us suppose that the very large equilibrium system contains a liquid phase in
which the reactants and products are soluble. The reaction can also occur in this liquid phase, and this liquid-phase reaction must be
at equilibrium. Since the system is at equilibrium, each chemical species must have the same chemical potential in the solution as it
does in the gas phase. Hence, the fugacities and the fugacity-based equilibrium constant are the same in both phases.
At this point, we have obtained—in principle—a complete solution to the problem of predicting the equilibrium position for any
reaction. If we can find the Gibbs free energy of formation of each substance in its hypothetical ideal-gas standard state, and we can
find its fugacity as a function of the composition and pressure of the system in which the reaction occurs, we can find the
equilibrium constant and the equilibrium composition of the system.
In practice, a great many substances are non-volatile. The Gibbs free energy of formation of their hypothetical ideal-gas standard
states and their fugacities cannot be measured. For such substances, we have recourse to other standard states and use activities to
express the equilibrium constant.
This page titled 15.6: Chemical Potential, Fugacity, and Equilibrium is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
When we choose a standard state for the activity of a substance, we want the chemical potential that we calculate from the
measured activity of a substance in a particular system to be the Gibbs free energy difference for the formation of the substance in
that system from its constituent elements in their standard states.
Thus we must be able to measure the Gibbs free change for the process that produces the substance in its activity standard state
~ o
from its constituent elements; this is the quantity that we designate as Δ G (A, ss) = μ̃ . We must also be able to measure the
f
o
A
difference between the chemical potential of the substance in this standard state and its chemical potential in the system of interest;
this is the quantity that we designate as RT ln a
~
A. The sum
o ~
μA = μ̃A + RT ln aA
is then the desired Gibbs free energy change for formation of the substance in the system of interest.
When it is convenient to choose the activity standard state to be the standard state of the pure liquid or the pure solid, we have
~o o o
Δf G (A, ss) = μ̃A = Δf G (A, ℓ)
or
~o o o
Δf G (A, ss) = μ̃A = Δf G (A, s)
For other choices, we may be unable to measure the difference between the chemical potential of the substance in the activity
~ o
standard state and that of its constituent elements in their standard states; that is, the value of Δ G (A, ss) = μ̃ may be
f
o
A
unknown. Nevertheless, we can describe the Gibbs free energy changes in a cycle that goes from the elements in their standard
states to the chemical species in the equilibrium system and back.
Figure 3. Relating activities in an equilibrium system to the Gibbs free energies of formation of the components in their activity
standard states.
For the reaction aA + bB → cC + dD , this cycle is shown in Figure 3. Summing the Gibbs free energy changes in a clockwise
direction around this cycle, we have
~o ~o ~o ~o
0 = c Δf G (C , ss) + d Δf G (D, ss) − a Δf G (A, ss) − b Δf G (B, ss)
~c ~d ~a ~b
+c RT ln aC + d RT ln aD − a RT ln aA − b RT ln aB
~o
Writing Δ G to emphasize that the standard Gibbs free energy change for the reaction is now a difference between the chemical
r
potentials of reactants and products in their activity standard states, we have

~o ~o ~o ~o ~o o
Δr G = c Δf G (C , ss) + d Δf G (D, ss) − a Δf G (A, ss) − b Δf G (B, ss) = Δr μ̃
and the equation for the Gibbs free energy change around the cycle becomes
~c ~d
~o aC aD
o
Δr G = Δr μ̃ = −RT ln
~a ~b
aA aB
Since the chemical reaction is at equilibrium, the activities term is constant. We have
~c ~d
aC aD
Ka =
~a ~b
aA aB
and
~o o
−Δr G −Δr μ̃
Ka = exp ( ) = exp ( )
RT RT
We append the subscript, “a ”, to indicate that the equilibrium constant, K , is expressed in terms of the activities of the reacting
a
substances. This cycle demonstrates the underlying logic for our calculation of K from our definition of activity, a
μ
A
= μ̃
o
A
~
+ RT ln a , our definition of Δ μ, and the fact that
A r
Δr μ = ∑ μj νj = 0
j=1
is a criterion for equilibrium.

Introducing activity coefficients, we have
~c ~d c d c d
a a c c γ γ
C D C D C D
Ka = =( )( )
~a ~b a b a b
a a c c γ γ
A B A B A B
When we know K and can estimate the activity coefficients, we can predict the thermodynamic properties of a system whose
a
component concentrations are known. In general, to determine K and the activity coefficients is more difficult than to determine
a
the equilibrium concentrations. Over narrow ranges of concentrations it is often adequate to assume that the activity-coefficient
function,
c d
γ γ
C D
Kγ =
a b
γ γ
A B
is approximately constant. When this is the case, the concentration function,

c d
c c
C D
Kc =
a b
c c
A B
is approximately constant. Then, a direct measurement of K in one equilibrium system makes it possible to predict the position of
c
equilibrium in other similar systems.

The relationship between the equilibrium constant and the standard Gibbs free energy change for a reaction is extremely useful. If
we can calculate the standard Gibbs free energy change from tabulated values, we can find the equilibrium constant and predict the
position of equilibrium for a particular system. Conversely, if we can measure the equilibrium constant, we can find the standard
~ o
Gibbs free energy change, Δ G . r
This brings us back to a central challenge in our development. We now have rigorous relationships between the Gibbs free energy
and the fugacity or activity of a substance. To use these relationships, we must be able to relate the fugacity or activity to the
concentration of a substance in any particular system that we want to study. For many common systems, this is difficult. In Chapter
16, we discuss some basic approaches to accomplishing it.
This page titled 15.7: Chemical Potential, Activity, and Equilibrium is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
In Section 13.5, we demonstrate that the Gibbs free energy change for a reaction among ideal gases is the same thing as the rate at
which the Gibbs free energy of the system changes with the extent of reaction. That is, for an ideal-gas reaction at constant
temperature and pressure, we find Δ G = (∂G/∂ξ) . We can now show that this conclusion is valid for any reaction.
r TP
With the introduction of the activity function, we have developed a very general expression for the Gibbs free energy of any
substance in any system. For substance A at a fixed temperature, we have
¯
¯¯¯ o ~
GA = μ = μ̃ + RT ln aA
A A
For a reaction that we describe with generalized substances and stoichiometric coefficients as
| ν1 | X1 + | ν2 | X2 + ⋯ + | νi | Xi → | νj | Xj + | νk | Xk + ⋯ + | νω | Xω
we can write the Gibbs free energy change in several equivalent ways:
ω
¯
¯¯¯
Δr G = ∑ νj Gj
j=1
= Δr μ
= ∑ νj μj ≤ 0
j=1
The Gibbs free energy of the system is a function of temperature, pressure, and composition,
G = G (P , T , n1 , n2 , … , nω )
To introduce the dependence of the Gibbs free energy of the system on the extent of reaction, we use the stoichiometric
relationships n = n + ν ξ . (n is the number of moles of the j reacting species; n is the number of moles of the j reacting
j
o
j j j
th o
j
th
species when ξ =0. If the k substance does not participate in the reaction, ν = 0 .) Then,
th
k
o o o
G = G (P , T , n + ν1 ξ, n + ν2 ξ, … , nω + νω ξ)
1 2
At constant temperature, pressure, and composition, the dependence of the Gibbs free energy on the extent of reaction is
o
ω ⎛ ⎞ ⎛ ∂ (n + νj ξ) ⎞
∂G ∂G j
( ) = ∑⎜ ⎟ ⎜ ⎟
∂ξ o ∂ξ
P T nm j=1 ⎝ ∂ (n + νj ξ) ⎠ ⎝ ⎠
j
P T nm≠j P T nm
= ∑ νj μj
j=1
It follows that
∂G
( ) = Δr G ≤ 0
∂ξ
P T nm
expresses the thermodynamic criteria for change when the process is a chemical reaction.
If a reacting system is not at equilibrium, the extent of reaction is time-dependent. We see that the Gibbs free energy of a reacting
system depends on time according to
dG ∂G dξ dξ
=( ) ( ) = Δr μ ( )
dt ∂ξ dt dt
P T nm
This page titled 15.8: The Rate of Gibbs Free Energy Change with Extent of Reaction is shared under a CC BY-SA 4.0 license and was authored,
15.9: Problems
Problems
Use data from the table below to find the thermodynamic properties requested in problems 1 to 7.
Properties CH3 OH CH3 CH2 OH
Density, g cm
−3
at 20 C 0.7914 0.7893
Mol mass, g mol

−1
32.04 46.07
bp, C 64.6 78.2
Δf G
o o
(300K, HI G )kJ mol
−1
–159.436 –162.934
Vapor pressure at320 K, bar 0.5063 0.2764
Virial coefficient, B , m
3
mol
−1
−1.421 × 10
−3
−2.710 × 10
−3
1. Find the chemical potentials of the pure gases, assuming that they are ideal, taking the hypothetical ideal gas standard state as the
standard state for activity (f = a
~
= P ).
2. Find the chemical potentials of the mixed gases, assuming that they are ideal, taking the hypothetical ideal gas standard state as
the standard state for activity (f = a
~
= x P) . A A
3. Find the chemical potentials of the mixed gases, assuming them to obey the Virial equation,
P V /RT = 1 + (BP /RT ),
assuming that the partial molar volumes in the mixture are equal to the partial molar volumes of the pure gases at the same
pressure, and taking the hypothetical ideal gas standard state as the standard state for activity.
4. Find the standard chemical potentials of the pure liquids at 320 K, assuming that the gases behave ideally.
5. Find the standard chemical potentials of the pure liquids at 320 K, assuming that the gases obey the Virial equation.
6. Find the chemical potential of the pure liquids as a function of pressure, assuming that the partial molar volumes of the pure
liquids are constant and that the gases obey the Virial equation.
7. Find the activity and chemical potential of the pure liquids at 101 bar, taking the pure liquids at 1 bar as the standard state for
activity.
8. A system is created by mixing one mole of gas A with one mole of gas C . Reaction occurs according to the stoichiometry
A+C ⇌ D . Assume that the behaviors of these gases in their equilibrium mixture are adequately approximated by the Virial
¯
¯¯¯
equations P V A /RT = 1 + (BA P /RT ) , etc.
(a) Show that the fugacity of gas A is given by
xA P BA P
ln fA = ln( ) +
o
P RT
(b) Write an equation for μ A

=μ
A
(xA , P ) at constant temperature, T .
(c) Write an equation for Δ rμ .
(d) Assume that
Δr μ
o
=μ
o
D
−μ
o
A
−μ
o
C
. If all three gases behave ideally (B = B = B
= 1000 J mol
−1
A C D =0 ), what is KP for this reaction
at 300 K? At equilibrium at 300 K and 1 bar, what are the mole fractions of A , C , and D?
(e) Under the assumptions in part (d), what are the equilibrium mole fractions of A , C , and D at 300 K and 10 bar?
(f) Suppose that, contrary to the assumptions in (d) and (e), the Virial coefficients are not zero and that
B A =B =BC = −10 m mol
D . At equilibrium at 300 K and 1 bar, what are the mole fractions of A , C , and D?
−3 3 −1
(g) Under the assumptions in part (f), what are the equilibrium mole fractions of A , C , and D at 300 K and 10 bar?
9. Suppose that the reaction A + C ⇌ D occurs in an inert solvent and that it is convenient to express concentrations as
molarities. A frequently convenient choice of activity standard state for a solute is a hypothetical one-molar solution in which the
chemical potential of the solute is equal to the chemical potential of the solute in a very (“infinitely”) dilute solution in the same
solvent. Then a
~
A = [A] γ , and γ
A
→ 1 as [A] → 0 . The thermodynamic equilibrium constant becomes
A
~ d
aD [D] γ
D
Ka = = = Kc Kγ
~ ~ a c
aA aC [A] [C ] γ γ
A C
where we introduce
[D]
Kc =
[A] [C ]
and
d
γ
D
Kγ =
a c
γ γ
A C
In a very dilute solution, K → 1 and K = K . Therefore, we can estimate K by finding the limiting value of K as all of the
γ c a a c
concentrations become very small. From values of K at higher concentrations, we can develop an empirical equation for K . The
c γ
form of this equation can be anything that can adequately represent the experimental data. Note, however, that finding an empirical
model for K does not solve the problem of finding empirical models for γ , γ , and γ individually.
γ A C D
(a) Given that the hypothetical one-molar solution is chosen to be the activity standard state for all three species, what is the
physical significance of Δ μ̃ ? r
o
(b) A simple function that has the properties required of γ is A

γ
A
=α
[A]
, where α is a constant. Represent , γ and , C
γ
D
by
similar functions and show that this leads to ln K = β [A] + β
γ A C
[C ] + βD [D] , where β , β , and β are constants.
A C D
(c) A series of solutions is prepared. The equilibrium concentrations of A , B , and D in these solutions are given below. Calculate
K for each solution. Estimate K and the parameters β , β , and β
c a in the equation of part (b).
A C D
[A] [B] [C]
1.96 x 10 −3
1.96 x 10 −3
3.84 x 10 −5
7.85 x 10 −3
1.85 x 10 −3
1.45 x 10 −4
3.94 x 10 −2
1.44 x 10 −3
5.57 x 10 −4
1.99 x 10 −1
7.15 x 10 −4
1.29 x 10 −3
9.98 x 10 −1
2.84 x 10 −4
1.72 x 10 −3
1.85 x 10 −3
7.85 x 10 −3
1.45 x 10 −4
1.44 x 10 −3
3.94 x 10 −2
5.60 x 10 −4
6.97 x 10 −4
1.99 x 10 −1
1.30 x 10 −3
2.38 x 10 −4
9.98 x 10 −1
1.76 x 10 −3
9.32 x 10 −3
7.32 x 10 −3
6.78 x 10 −4
3.24 x 10 −2
3.04 x 10 −2
9.57 x 10 −3
1.06 x 10 −1
1.04 x 10 −1
9.59 x 10 −2
(d) Using the values you find in part (c), estimate the equilibrium concentrations of A , B , and D when a solution is prepared by
mixing one mole of A with one mole of C and sufficient solvent to make 1 L of solution at equilibrium.
10. An ester, RC O 2R
′
, undergoes hydrolysis in an ether solvent:
′ ′
RCO R + H O −
↽⇀
− RCO H + HOR
2 2 2
We can express the activity of any of these species as the product of a concentration (in any convenient units) and an activity
coefficient. When all of the reactants and products are present at low concentrations, the activity coefficients are approximately
unity. The standard state for each species becomes a hypothetical solution of unit concentration in which the chemical potential (per
mole) of that species is the same as its chemical potential in an arbitrarily (infinitely) dilute solution. A solution is prepared by
mixing 2 × 10 mole of the ester and 10 mole water in sufficient ether to make 1 L of solution. When equilibrium is reached,
−3 −1
the acid and alcohol concentrations are 9.66 × 10 molar.

−3
(a) What is the equilibrium constant for this reaction?

(b) A solution is prepared by mixing 2 × 10 mole of the acid, 3×10 mole of the alcohol, and 10
−2 −2 −1
mole of water in sufficient
ether to make 1 L of solution. When equilibrium is reached, what is the concentration of the ester?
(c) With this choice of the standard states, what physical process does the standard chemical-potential change, Δμ̃ , describe?
o
CHAPTER OVERVIEW
16: The Chemical Activity of the Components of a Solution

We are frequently interested in equilibrium processes that occur in a solution at a constant temperature. If we are able to find the
activities of the species making up the solution, we can describe the thermodynamics of such processes. Many experimental
methods have been developed for the measurement of the activities of species in solution. In general, the accurate measurement of
chemical activities is experimentally exacting. In this chapter, we consider some of the basic concepts involved. We focus primarily
on molecular solvents and solutes; that is, neutral molecules that exist as such in solution. We introduce a simplified model, called
the ideal solution model, which is often a useful approximation, particularly for dilute solutions. In Sections 16.16-16.18, we touch
on the special issues that arise when we consider the activities of dissolved ions.
16.20: Problems
Thumbnail: The Effect of Solution Formation on Entropy. (CC BY-SA-NC; anonymous by request)
This page titled 16: The Chemical Activity of the Components of a Solution is shared under a CC BY-SA 4.0 license and was authored, remixed,
1
One way to find activities is to find the composition and pressure of the gas phase that is in equilibrium with the solution. If the
gases are not ideal, we also need experimental data on the partial molar volumes of the components in the gas phase. Collecting
such data is feasible for solutions of volatile molecular liquids. For solutions of electrolytes or other non-volatile components, other
methods are required.
Figure 1. Equilibrium mole fractions in gas and liquid phases.

The curves sketched in Figure 1 describe a system containing components A and B . The mole fractions in the solution and the
mole fractions in the gas are related in a non-linear way. Let the mole fractions in the gas be x and x ; let those in the solution be
A B
yA and y . We have x + x = 1 and y + y = 1 . At equilibrium, both phases are at the same pressure, P .
B A B A B
We imagine obtaining the data we need about this system by preparing many mixtures of A and B . Beginning with an entirely
liquid system at some applied pressure, we slowly decrease the applied pressure until the applied pressure becomes equal to the
equilibrium pressure, P , and the liquid begins to vaporize. Figure 2 shows this system schematically.
Figure 2. Liquid-gas equilibrium for a solution.

We determine the compositions of the gas and liquid phases by chemical analysis; for each system, we determine P , x , x , y , A B A
and y . From these data we can develop empirical equations that express P , x , and x as functions of y ; that is, we have
B A B A
P = P (y ) , x
A A =x A(y ) , and x
A =x
B (y ) . Finally, we can find the value of the products x P and x P . Figure 3 illustrates
B A A B
a possible function P = P (y ) and the products x P , and x P , when the gas-phase mole fractions depend on y as shown in
A A B A
Figure 1.
Figure 3. Equilibrium pressures over a solution.
With the hypothetical ideal gas standard state as the standard state for A in the gas phase, we see in Section 14.11 that the chemical
potential of A in the gas phase is
P ¯
¯¯¯
xA P V A (g) 1
o o
μA (g, P , xA , xB ) = Δf G (A, H I G ) + RT ln[ ] + RT ∫ ( − ) dP
o
P 0
RT P
(any gas; activity standard state is HIG ) o
¯
¯¯¯
where V (g) is the partial molar volume and x is the mole fraction of A in the gaseous mixture. The fugacity and activity of A
A A
in the gas phase are given by

P ¯
¯¯¯
fA (g, P , xA , xB ) xA P V A (g) 1
~
ln[ aA , P , xA , xB ] = ln[ ] = ln[ ]+∫ ( − ) dP
o o
fA (H I G ) P 0
RT P
and the standard state chemical potential is

o o o
μ = Δf G (A, H I G )
A
We want to express the chemical potential of A in the liquid solution using the properties of the solution. To do so, we introduce
the chemical activity of component A . We write a ~
(P , y , y ) to represent the activity of A in a solution at pressure P and in
A A B
which the composition is specified by the mole fractions y and y . If it suits our purposes, we are free to choose a standard state
A B
for the activity of A in the liquid solution that is different from the standard state we choose for A in the gas phase. For reasons that
become apparent below, it is often useful to choose the standard state for the activity of A in the liquid solution to be pure liquid A
at its equilibrium vapor pressure, P . We represent the chemical potential of A in this standard state by μ̃
A
o
A
(ℓ, P
A
) . Note that this
state is not identical to the standard state for the pure liquid, for which the pressure is one bar and the chemical potential is
Δ G (A, ℓ) . The chemical potential and the activity of A in the solution are related by
o
f
o ~
μA (P , yA , yB ) = μ̃A (ℓ, P ) + RT ln[ aA (P , yA , yB )]
A
(any solution; activity standard state for A in solution is pure liquid A at its equilibrium vapor pressure)
Since the system is at equilibrium, we have
μ (g, P , xA , xB ) = μ (P , yA , yB )
A A
Equating our equations for these quantities, we find

o o o
Δf G (A, H I G ) − μ̃ (ℓ, P ) P ¯
¯¯¯
A A xA P V (g) 1
~ A
ln[ aA (P , yA , yB )] = + ln[ ]+∫ ( − ) dP
o
RT P 0
RT P
This equation gives the activity of A in a liquid solution whose state is specified by the liquid-phase mole fraction y . In particular, A
it must give the activity of A in the “solution” for which y = 1 and y = 0 ; of course, this “solution” is pure liquid A . At
A B
equilibrium with pure liquid A , the gas phase contains pure gaseous A ; therefore, we have xA = 1 and P =P
A
. The gas-phase
partial molar volume is that for the pure gas, V ¯
¯¯¯
A
(g) . Moreover, this “solution” is the standard state for the activity of component
A , for which the activity of A is unity; that is,
~ ~
aA (P , yA , yB ) = aA (P
A
, 1, 0) = 1 . Making these substitutions into our equation for ~
ln[ aA (P , yA , yB )] and rearranging, we
find
o o o
Δf G (A, H I G ) − μ̃A (ℓ, P ) P
⎛V
¯
¯¯¯
A P
A
A
A
(g) 1 ⎞
= −ln[ ] −∫ − dP
o
RT P 0 ⎝ RT P ⎠
Substituting this result into our general equation for ~

ln[ aA (P , yA , yB )] we find a completely general function for the activity of
component A .
cdot
⋅
PA ¯
¯¯¯ P ¯
¯¯¯
xA P V (g) 1 V A (g) 1
~ A
ln[ aA (P , yA , yB )] = ln[ ]−∫ ( − ) dP + ∫ ( − ) dP
⋅
P 0
RT P )
RT P
A
(any solution; xA is the mole fraction in the gas; the standard state for A in solution is the pure liquid at its equilibrium vapor
pressure)
In such a system, the roles of solute and solvent are interchangeable. Interchanging the labels “A ” and “B” gives an equation for
the activity of component B .
As circumstances warrant, several approximations can be applied to this result. When the partial molar volume of A in the gas,
(g) , is not available, some approximation is required. Perhaps the least drastic approximation is that introduced in Section
¯
¯¯¯
V A
14.11. We equate the unknown partial molar volume to the partial molar volume of the pure real gas at the same system pressure.
Setting V
¯
¯¯¯
A
(g) = V
¯
¯¯¯
A
(g) , we have
P ¯
¯¯¯
xA P ⎛V (g) 1 ⎞
~ A
ln[ aA (P , yA , yB )] = ln[ ] +∫ − dP
P ⎝ RT P ⎠
P A
A
Other approximations lead to greater simplifications. In the following sections, we discuss several. All of them assume that the
components behave ideally in the gas phase. In this case, the integrals in our general equation for ln[a~
( P , y , y )] vanish. A A B
Then,
xA P
~
ln[ aA (P , yA , yB )] = ln[ ]
P
A
(solution; x is the mole fraction in the ideal gas that is at equilibrium with the solution)
A
Our general result gives the activity of component A in solution using the mole fraction of its own vapor in the equilibrium system.
When we have an empirical function, x = x (y ) , that relates the mole fraction in the gas to that in the solution, we can make
A A A
this substitution and express the activity of A in the solution using its concentration in the solution. In the next several sections, we
develop some basic methods for finding x = x (y ) . A A A
This page titled 16.1: Solutions Whose Components are in Equilibrium with Their Own Gases is shared under a CC BY-SA 4.0 license and was
An ideal solution is a homogeneous liquid solution that is at equilibrium with an ideal-gas solution in which the vapor pressure of
each component satisfies Raoult’s law . Since the gas is ideal, the partial pressure of A is P = x P . Raoult’s law asserts a
1
A A
relationship among the gas- and solution-phase mole fractions of A , the vapor pressure of the pure liquid, and the pressure of the
system:
PA = xA P = yA P
A
(Raoult’s law)
For a binary mixture of A and B that satisfies Raoult’s law, we have also that P B = xB P = yB P , and the total pressure becomes
B
P =P A +P B=y P A
A
+y PB . The lines sketched in Figure 4 show how
B
PA , PB , and P vary with the solution-phase
composition when the solution is ideal.
Figure 4. Equilibrium pressures over an ideal solution.

When the standard state for A in solution is taken to be pure liquid A at its equilibrium vapor pressure, substitution of Raoult’s law
into the results in Section 16.1 gives the activity of component A in an ideal solution as
xA P yA P
~ A
ln[ aA (P , yA , yB )] = ln[ ] = ln[ ] = ln yA
P P
A A
and
~
aA (P , yA , yB ) = yA
(ideal solution, Raoult’s law)

In general, the activity and chemical potential of a component depend on pressure. If the solution is ideal, we see that the system
pressure is fixed by P = y P + y P , and the pure-component vapor pressures depend only on temperature. Since for the
A
A
B B
binary solution, y = 1 − y , we can write the chemical potential of component A as

B A
o
μ (P , yA , yB ) = μ (yA ) = μ̃ (ℓ, P ) + RT ln yA
A A A A
(ideal solution)
We can also use relationships we develop earlier to find another representation for μ̃ o
A
(ℓ, P
A
) . The chemical potential of A in the
liquid phase is the same as in the gas. Using the chemical potential for A in the gas phase that we find in Section 16.1, we have
xA P P
o o o o A
μA (P , yA , yB ) = μA (g, P , xA , xB ) = Δf G (A, H I G ) + RT ln[ ] = Δf G (A, H I G ) + RT ln[ ]
o o
P P
+ RT ln yA
and hence,
P
o o o A
μ̃A (ℓ, P ) = Δf G (A, H I G ) + RT ln[ ]
A o
P
In Section 15.4, we find, for an ideal gas,

P
P A
o o A o ¯
¯¯¯
Δf G (A, H I G ) + RT ln[ ] = Δf G (A, ℓ) + ∫ V (ℓ)dP
o A
P P
o
so that the chemical potential of the pure liquid at its vapor pressure is also given by
P
A
o o ¯
¯¯¯
μ̃A (ℓ, P ) = Δf G (A, ℓ) + ∫ V A
(ℓ)dP
A
o
P
The integral is the difference between the Gibbs free energy of the pure liquid at its vapor pressure and that of the pure liquid at
¯
¯¯¯
P
o
= 1 bar . Note that we can obtain the same result much more simply by integrating (dG A
) =V A
dP between the same two
T
states. In Section 15.3, we see that the value of the integral is usually negligible. To a good approximation, we have
o o
μ̃ (ℓ, P ) ≈ Δf G (A, ℓ)
A A
(ideal solution)
This page titled 16.2: Raoult's Law and Ideal Solutions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
If the components behave ideally in the gas phase and if pure liquid A at its equilibrium vapor pressure, P
A
, is the standard state
for the activity of A in solution, we find in Section 16.2 that the activity of component A is a ~
(P , y , y ) = x P / P . A A B A
A
Experimentally, we determine a relationship between the mole fractions of A in the gas and liquid phases. By expressing this
relationship as the function x = x (y ) , we can express the activity as a function of y . If Raoult’s law is obeyed, we have seen
A A A A
that this function is x (y ) = y P /P , and the activity is a

A A A
A
~
(P , y , y ) = y . A A B A
If Raoult’s law is not obeyed, we must find an alternative function that adequately describes the experimentally observed
relationship between x and y . As we note in Section 14.10, we want to construct this function so that it approaches y whenever
A A A
the behavior of the solution approaches the behavior of an ideal solution. We can accomplish this by defining the activity
coefficient for component A , γ = γ (P , y , y ) , by the equation
A A A B
~
aA (P , yA , yB ) = yA γA (P , yA , yB )
(Raoult’s law activity)

where the argument lists serve to emphasize that the activity and the activity coefficient are functions of the same thermodynamic
variables.
Dropping the argument lists and equating the two activity relationships, we have
xA P
= yA γ
A
P
A
so that the activity coefficient is

xA P
γA =
yA P
A
(Raoult’s Law activity coefficient)

Since we are using pure liquid A at its equilibrium vapor pressure as the standard state for component A , the chemical potential
can be expressed as
o
μ (P , yA , yB ) = μ̃ (ℓ, P ) + RT ln yA γ
A A A A
Introduction of the activity coefficient adds nothing to our store of information about the system. It merely provides a convenient
way to recast the available information, so that the solute mole fraction, y , becomes the independent variable in the chemical-
A
potential equation. (For a one-phase two-component system, γ is completely determined by the temperature, system pressure, and
A
y . Then y
A A is the concentration variable in the chemical-potential equation. If there are more than two components, additional
concentration variables are required to specify the composition of the system and the values of a
~
and γ .) A A
In summary, since the gas is ideal, the partial pressure of A above the solution is P A = xA P , whether the solution is ideal or not. If
the solution is ideal, we have yA P
A
= xA P , so that yA P
A
/P
o
= xA P / P
o
. If the solution is not ideal, we introduce the activity
coefficient, γ , as the “fudge factor” that makes
A
γA yA P
A
= xA P true. That is, the activity coefficient is just the actual value of
the partial pressure of ideal gas A , x P , divided by the value it would have if the solution were ideal, y P . The activity
A A A
coefficient corrects for the departure of the real solution from the behavior that Raoult’s law predicts for the ideal solution. When
we define the activity coefficient so that a
~
A = yA γA , we have a
~
A PA = xA P , so that
xA P
~ o
aA P /P =
A o
P
thus preserving the form of the ideal solution result—with ~

aA replacing y . For an ideal solution,
A γA = 1 , and our result for the
real solution activity reduces to the ideal solution activity.
This page titled 16.3: Expressing the Activity Coefficient as a Deviation from Raoult's Law is shared under a CC BY-SA 4.0 license and was
In describing the activities of solution components, we have taken the standard state of component A to be the pure liquid at its
equilibrium vapor pressure, P , at the temperature of the solution. We can also express the activity using Henry’s law. Henry’s law
A
states that the partial pressure of a component above its solution is directly proportional to the concentration of the component. We
can choose any convenient unit to express the solute concentration. The value of the proportionality constant depends on this
choice. Using mole fraction as the unit of concentration, Henry’s law is
PA = xA P = κA yA
(Henry’s law)
where the proportionality constant, κ , is called the Henry’s-law constant. (When we write P = x P , we implicitly assume that
A A A
the gas-phase components of the equilibrium system behave as ideal gases.) Henry’s law is more general than Raoult’s law. Indeed,
Raoult’s law is a special case of Henry’s law; if the solute obeys Raoult’s law, the Henry’s-law constant is P .
A
The value of the Henry’s-law constant depends on the components and the temperature. Experimentally, the value of the constant is
determined by finding the slope of a plot of P versus y in the limit as y → 0 . The sketch in Figure 5 illustrates the relationship
A A A
between the P curve and the Henry’s-law tangent to it at y = 0 . Note that the slope of the tangent line at y = 0 is equal to its
A A A
intercept at y = 1 .
A
Figure 5. Henry's law for a solute.

In practice, a sufficiently low concentration of any non-electrolyte component, in any liquid solution, obeys Henry’s law. For this
reason, we refer to the component that obeys Henry’s law as the solute. We designate the higher concentration component as the
solvent. The universal validity of Henry’s law as a low-concentration approximation has a simple physical interpretation. The
solute vapor pressure depends upon the net effects of solute–solute, solute–solvent, and solvent–solvent intermolecular forces.
If all of these intermolecular forces are the same, the intermolecular interactions that determine the gas-phase composition are the
same for solvent molecules as they are for solute molecules; the vapor pressures of the solvent and the solute are the same; the gas
above the solution has the same composition as the solution; the partial pressure of the solute is proportional to the solute
concentration, y ; the proportionality constant is the pure-solvent vapor pressure, P ; and the solution obeys Raoult’s law.
A
A
However, if the intermolecular forces are not all the same, their net effect changes as the solute concentration changes. As the
solute concentration increases, the effects of solute–solute interactions become increasingly important. If these are different from
the effects of solute–solvent and solvent–solvent interactions, the solute partial pressure is not proportional to the solute
concentration.
Conversely, at some sufficiently low concentration, solute molecules are so far apart that the effects of solute–solute interactions
become negligible. Only solute–solvent and solvent–solvent interactions affect the solute vapor pressure. Because these remain
constant as the solute concentration decreases further, the solute partial pressure is proportional to the solute concentration in this
low-concentration regime. However, if the effects of solute–solvent interactions are different from those of solvent–solvent
interactions, the pure-solvent vapor pressure, P A
, is not the proportionality constant.
We can assume the existence of a hypothetical ideal gas state for any substance, even a substance that has no measureable vapor
pressure at any attainable temperature. We can assume that the substance has a well-defined Gibbs free energy of formation in this
state even though there may be no possibility of measuring its value. Likewise, we can assume that any solute exerts some partial
pressure over its solution. We consider that this partial pressure has some finite value, even if it is much too small to measure. It
follows that the fugacity of the solute has a finite value. Henry’s law implies that the fugacity is proportional to the solute
concentration, at least in the limit of arbitrarily low concentration. Expressing the concentration of solute A as its mole fraction, we
have
o o o
fA (P , yA , yB ) = PA / P = xA P / P = κA yA / P
and the chemical potential of the solute is

o o
μ (P , yA , yB ) = Δf G (A, H I G ) + RT ln fA (P , yA , yB )
A
If the solute behaves ideally in the gas phase, Henry’s law leads to simple expressions for its chemical potential and activity. The
development of these expressions is very similar to the corresponding development from Raoult’s law. The essential differences
arise from the introduction of a different standard state for the activity of the solute. We begin with our basic equation for the
chemical potential of A in the gas phase:
P ¯
¯¯¯
o
xA P V A (g) 1
o
μ (g, P , xA , xB ) = Δf G (A, H I G ) + RT ln[ ] +∫ ( − )dP
A o
P 0
RT P
The integral term vanishes because we assume ideal gas behavior. We write a ~
(solute, P , y , y ) to represent the activity of A in
A A B
a solution at pressure P and in which the composition is specified by the mole fractions y and y . A B
We choose a hypothetical liquid to be the standard state for the activity of the solute. This hypothetical liquid is pure liquid A at the
vapor pressure it would exhibit if it followed Henry’s law over the entire range of possible system compositions. This pressure is
equal to its Henry’s law constant κ . (See Figure 5.) Let us denote the chemical potential of this standard state by μ̃ (H yp ℓ, κ ).
A
o
A A
The chemical potential of solute A and the activity of A in the solution are related by
o ~
μA (P , yA , yB ) = μ̃A (H yp ℓ, κA ) + RT ln aA (P , yA , yB )
and since μ A
(g, P , xA , xB ) = μA (P , yA , yB ) , and the gas is ideal, we have
o o o
Δf G (A, H I G ) − μ̃ (H yp ℓ, κA ) xA P
~ A
ln[ aA (P , yA , yB )] = + ln[ ]
o
RT P
Except for the solute standard state, this is same as the equation that we develop in Section 16.1.
This equation must give the activity of solute A in its standard state, which is pure hypothetical liquid at a pressure equal to the
Henry’s law constant: P = P = κ . In this state, x = y = 1 and x = y = 0 . In its standard state, the activity of solute A
A A A A B B
is unity; we have a
~
A A
~
(P , y , y ) = a
B (κ , 1, 0) = 1 . Making these substitutions and rearranging, we find
A A
o o o
Δf G (A, H I G ) − μ̃ (H yp ℓ, κA ) κA
A
= −ln[ ]
o
RT P
Substituting this result into our general equation for ln[a

~
A (solute, P , yA , yB )] , we find that the activity of solute A is
xA P
~
aA (P , yA , yB ) =
κA
(any solution, ideal gas)

The chemical potential of the standard-state hypothetical pure liquid whose vapor pressure at T is κ is A
o o o
κA
μ̃A (H yp ℓ, κA ) = Δf G (A, H I G ) + RT ln[ ]
o
P
This is a general result for the activity of solute A when the standard state is the hypothetical pure liquid whose pressure is κ , and A
A behaves ideally in the gas phase. If Henry’s law is obeyed, we have x P = y κ . Substituting, we find A A A
~
aA (P , yA , yB ) = yA
(Henry’s law is obeyed)

The Henry’s law development and the Raoult’s law development give the same value for the chemical activity. However, the
standard states are different.
If the solute obeys Raoult’s law, the standard state we choose for the solute is the pure solute at its equilibrium vapor pressure; in
this state, the pure solute is in equilibrium with its own gas. Real substances can satisfy this condition. If the solute obeys Henry’s
law, the standard state we choose for the solute is a hypothetical pure-liquid solute at a pressure equal to κ . In this state, we
A
assume that the hypothetical liquid is in equilibrium with its own gas, also at pressure κ . The hypothetical standard-state liquid is
A
a substance in which the interactions among A molecules have the same effects as the interactions, in a very dilute solution,
between A molecules and the B molecules that comprise the solvent. If solutions of A and B are described poorly by Raoult’s law,
the vapor pressure of pure liquid A , P , is likely to be very different from the vapor pressure, κ , of the hypothetical standard-
A A
state liquid that we define using Henry’s law.

Our development produces a model for the chemical potential of the solute in which the activity is equal to the solute mole fraction.
At the same mole fraction, every solute has the same activity. The chemical potentials of different solutes vary because the
chemical potential in its standard state is different for every solute. We conclude that we can let a
~
A=y for any sufficiently dilute
A
solute, even when it is not feasible to measure the chemical potential of the solute in its standard state experimentally.
This page titled 16.4: Henry's Law and the Fugacity and Activity of A Solution Component is shared under a CC BY-SA 4.0 license and was
Even if Henry’s law is valid only for solute concentrations very close to zero, we can use it to express the activity of the real system
as a function of solute concentration. Let us suppose that we have data on the mole fraction of A , x , in a gas whose pressure is P
A
and which is at equilibrium with a solution in which its mole fraction is y . In the preceding section, we find that we can choose
A
the solute’s standard state so that its activity in any state is a~

= x P / κ . Introducing the activity coefficient, defined by
A A A
= y γ , we have x P / κ = y γ . The activity coefficient is

~
aA A A A A A A
xA P
γ =
A
yA κA
(Henry’s law activity coefficient)

and the chemical potential is
o
μ = μ̃ (H yp ℓ, κA ) + RT ln yA γ
A A A
Just as when we define the activity coefficient using the deviation from Raoult’s law, this development provides a way to recast the
available information in a way that makes the solute mole fraction, y , the independent variable in the chemical-potential equation.
A
In Section 16.4, we note that Raoult’s law is the special case of Henry’s law in which P = κ . If we make this substitution into
A A
the Henry’s-law based activity coefficient, we recover the Raoult’s-law based activity coefficient.
This page titled 16.5: Expressing the Activity Coefficient as A Deviation from Henry's Law is shared under a CC BY-SA 4.0 license and was
When the solution is dilute, it is often convenient to use the molality of the solute rather than its mole fraction. We define the
¯¯¯¯
¯
molality as the number of moles solute, A , per kilogram of solvent, B . We use ––m to represent the molality of A and M to
A B
represent the gram-molar mass of B . A solution that contains n and n moles of A and B , respectively, contains a mass of
A B
solvent given in kilograms by

¯¯¯¯
¯
nB M B
mB =
1000
Then the molality of A is

nA 1000nA
m = =
––A mB ¯¯¯¯
¯
M B nB
so that
¯¯¯¯
¯
m nB M B
––A
nA =
1000
The mole fraction of A is given by

¯¯¯¯
¯
nA m MB
––A
yA = =
¯¯¯¯
¯
nA + nB
m M B + 1000
––A
In dilute solution, where n A

≪ nB and m ¯¯¯¯
¯
M B ≪ 1000 , the mole fraction and molality of the solute are related by
––A
¯¯¯¯
¯
MB
yA ≈ ( )m
1000 ––A
(dilute solution)
Using this approximation, assuming that solute A obeys Henry’s law and that gas A behaves ideally, we have
o
xA P
o
μA (P , m ) = μA (P , yA , yB ) = μA (g, P , xA , xB ) = Δf G (A, H I G ) + RT ln[ ]
––A P
o
κA yA
o o
= Δf G (A, H I G ) + RT ln[ ]
o
P
¯¯¯¯
¯
o
κA M B
o
= Δf G (A, H I G ) + RT ln[ ] + RT ln m
o
1000P ––A
(solute A obeys Henry’s law)

When Henry’s law is not obeyed over the composition range of interest, it is often convenient to choose the standard state of the
solute to be a one-molal solution of a hypothetical substance that obeys Henry’s law with the Henry’s law constant κ . Then the A
activity of this solution is unity, and its chemical potential is the chemical potential of A in this hypothetical standard state. Letting
this standard-state chemical potential be μ̃ (H yp 1 ––
o
A
m solute, P ) , we have
¯¯¯¯
¯
o o o
κA M B
μ̃A (H yp 1 m solute, P ) = Δf G (A, H I G ) + RT ln[ ]
–– o
1000P
The chemical potential of a substance that satisfies Henry’s law is

o
μA (H L, P , m ) = μ̃ (H yp 1 m solute, P ) + RT ln m
––A A –– ––A
If A behaves as an ideal gas and the solution is dilute (m M ≪ 1000), but m is above the range in which Henry’s law is
A
¯¯¯¯
¯
B A
–– ––
obeyed, we introduce the Henry’s law activity coefficient, γ = a ~
/ m , to measure the departure of the real solution behavior
A A A
––
from that predicted by Henry’s law. Then the chemical potential of A in any solution is
o ~ o
μA (P , m ) = μ̃ (H yp 1 m solute, P ) + RT ln aA = μ̃ (H yp 1 m solute, P ) + RT ln m γA
––A A –– A –– ––A
o
= μ̃ (H yp 1 m solute, P ) + RT ln γ + RT ln m
A A
–– ––A
and the logarithm of the activity coefficient measures the difference between the chemical potential of the real solute and that of a
solute that obeys Henry’s law over an extended concentration range:
μA (P , mA ) − μA (H L, P , mA )
–– ––
ln γA =
RT
In Section 16.19, we consider the determination of Henry’s law-based activity coefficients further.
This page titled 16.6: Henry's Law and the Hypothetical One-molal Standard State is shared under a CC BY-SA 4.0 license and was authored,
We have seen that the activity of any component of an equilibrium system contains information about the activities of all the other
components. From
¯¯
¯ ¯
¯¯¯ ~
dμ = −S dT + V P + RT dln a
and the Gibbs-Duhem equation, we can find a general relationship among the activities. Substituting dμ and dμ into the Gibbs- A B
Duhem equation, we have, for a two-component solution,

−SdT + V dP = nA dμ + nB dμ
A B
¯¯
¯ ¯
¯¯¯ ~ ¯¯
¯ ¯
¯¯¯ ~
= nA (−S A dT + V AP + RT dln aA ) + nB (−S B dT + V B P + RT dln aB )
¯¯
¯ ¯¯
¯ ¯
¯¯¯ ¯
¯¯¯ ~ ~
= − (nA S A + nB S B ) dT + (nA V A + nB V B) dP + nA RT dln aA + nB RT dln aB
¯¯
¯ ¯¯
¯
Since S = n A
S A + nB S B and V = nA V
¯
¯¯¯
A
+ nB V
¯
¯¯¯
B , this simplifies to
~ ~
0 = nA dln aA + nB dln aB
or, dividing by n A + nB ,
~ ~
0 = yA dln aA + yB dln aB
For simplicity, let us consider a system in which a non-volatile solute, A , is dissolved in a volatile solvent, B . Measuring the
pressure of the system and applying the equations that we developed in Section 16.1 for volatile component A to the volatile
solvent, B , in the present system, we can determine the activity of the solvent, B . Let us use mole fractions to measure
concentrations and take pure liquid B at its equilibrium vapor pressure as the activity standard state for both liquid- and gas-phase
B . When B is in its standard state, we have x = 0 , x = 1 , and V (g) = V (g) . Then, since the solute is non-volatile, we can
A B
¯
¯¯¯
B
¯
¯¯¯
determine the activity of the solvent, B , from the pressure of the system. We have
P ¯
¯¯¯
P ⎛V (g) 1 ⎞
~ B
ln[ aB (P , yA , yB )] = RT ln[ ] −∫ − dP
P ⎝ RT P ⎠
P B
B
Assuming that the integral makes a negligible contribution to the activity, we have
P
~ ~
aB (P , yA , yB ) = aB = yB γB (P , yA , yB ) =
P
B
(solvent)
so that
P
γB (P , yA , yB ) = γB =
yB P
B
(solvent)
Since the gas-phase concentration of A is immeasurably small, we must determine its activity indirectly. Let the standard state for
solute activity be the hypothetical pure liquid, y = 1 , whose equilibrium vapor pressure is equal to the Henry’s law constant of
A
solute A . (We can determine the solute’s activity without measuring its Henry’s law constant.) We have
o ~
μA (P , yA , yB ) = μ̃A (H yp ℓ, ҝA ) + RT ln[a A (P , yA , yB )]
where
~ ~
aA (P , yA , yB ) = aA = yA γA
(solute)
Since we are able to measure the activity of the solvent, we can determine the activity of the solute from the relationship
A
~
0 = y dln a
~
+ y dln a
A B . Rearranging, we have
B
yB 1 −y
~ ~ A ~
dln aA =− dln aB = −( ) dln aB
yA yA
For two solutions in which the mole fractions of A are y and y , and in which the activities of A and B are a
A
#
A
~
,a
~
,a
~
, and A B
#
A
~#
a
B
,
we have
~ yA
aA 1 − yA
~
ln = −∫ ( ) d ln aB
~# # yA
a y
A
A
#
Graphically, the integral is the area under a plot of −(1 − y A )/ yA versus ln a
~
, from y to y .
B A A
Typically, we are interested in solutions for which y ≪ y . In the limit as the solution becomes very dilute, the activity, mole
A B
fraction, and activity coefficient of the solvent, B , all approach unity: a

~
→ 1, y → 1 , and γ → 1 . The activity of the solute,
B B B
A , approaches the mole fraction of A . As a matter of experience, the approach is asymptotic: as the mole fraction approaches zero,
→ 0 , the solute activity coefficient approaches unity, γ → 1 , and does so asymptotically, so that a → y . For dilute
~
yA A A A
solutions, ln a → −∞ and ln γ → 0 asymptotically. In consequence,

~
A A
lim (dln γA ) = 0
yA →0
Because the activity coefficient approaches a finite limit while the activity does not, we can express the solute’s activity most
simply by finding the solute’s activity coefficient. Since a
~
=y γ and a~
A=y γ = (1 − y ) γ
A A , we have B B B A B
~ ~
0 = yA dln aA + yB dln aB
= yA dln yA γA + yB dln yB γB = yA dln yA + yA dln γA + yB dln yB + yB dln γB
= (dyA + dyB ) + yA dln γA + yB dln γB = yA dln γA + yB dln γB
(Since y A + yB = 1 , we have dyA

+ dyB = 0 .) We can rearrange this to
1 −y
A
dln γA = −( ) dln γB
yA
As the solute concentration approaches zero, (1 − yA )/ yA becomes arbitrarily large. However, since limy
A
→0 (dln γA ) = 0 , it
follows that
lim dln γB =0
yA →0
We see that the solvent activity coefficient also approaches unity asymptotically as the solute concentration goes to zero. The solute
activity coefficient at any y > 0 is then given by
A
yA yA
1 − yA
∫ d ln[ γA (yA )] ln[ γA (yA )] ∫ −( ) d ln[ γB (yA )]
0 0
yA
Figure 6. Graphical representation of ln γ for the solute when γ of the solvent is known.
As sketched in Figure 6, the latter integral is the area under a graph of − (1 − y )/y versus ln[γ (y )] , between y = 0 and
A A B A A
y . Since γ
A (y ) → 1 as y
A A → 0 , this integral must remain finite even though − (1 − y )/ y
A → −∞ as y → 0 . This can A A A
occur, because lim dln γ

yA →0 = 0 , as we observe above. Nevertheless, the fact that the integrand is unbounded can limit the
B
accuracy of the necessary integration. For accurate measurement of the solute activity coefficient, it is important to obtain solvent-
activity data at the lowest possible solute concentration.
The most desirable situation is to collect solvent-activity data down to solute concentrations at which the solvent activity
coefficient, γB , becomes unity. If γB (y
A
#
) =1 when the solute concentration is y
#
A
, ln[ γA (yA )] can be evaluated with y
A
#
,
rather than zero, as the lower limit of integration. In some cases, ln[ γ (y )] may be known from some other measurement at a
A
△
A
particular concentration, y ; if so, we can find ln[γ (y )] by carrying out the numerical integration between the limits y and
△
A
A A
△
y .
A
If the measurement of γ cannot be extended to values of y at which γ (y ) = 1 , we must find an empirical function, call it
B A B A
f (y ), that fits the experimental values of ln[ γ

A (y )] , for the smallest values of y . (That is, the empirical function is
B A A
f (y ) = ln [ γ
A (y )] .) The differential of f (y ) is then a mathematical model for dln[ γ
B A A (y )] over the region of low solute B A
concentrations. Letting y be the smallest solute concentration for which the solvent activity can be determined, we can integrate,
#
#
using the function for dln[γ B (yA )] that we derive from this model, to estimate ln[γ B (y
A
)] . Uncertainty about the accuracy of
the mathematical model becomes a significant source of uncertainty in the calculated values of γ . A
Of course, if we can find an analytical function that provides a good mathematical model for all of the solvent-activity data, the
differential of this function can be used in the integral to evaluate ln[γ (y )] over the entire range of the experimental data. If
A A
necessary, the evaluation of this integral can be accomplished using numerical methods.
It is essential that any empirical function, f (y ) = ln [γ (y )] , have the correct mathematical properties over the concentration
A B A
range to which it is applied. If it is to be used to extend the integration to y = 0 , f (y ) must satisfy f (0) = 0 and df (0) = 0 .
A A
This is a significant condition. For example, consider the approximation

αB
f (yA ) = ln[ γB (yA )] = cB y
A
This model gives

dln[ γA (yA )] αB −2
= − (1 − yA ) c αB y
B A
dyA
and
dln[ γA (yA )]
lim =0
yA →0 dyA
requires α B >2 .
This page titled 16.7: Finding the Activity of a Solute from the Activity of the Solvent is shared under a CC BY-SA 4.0 license and was authored,
In Section 16.4, we conclude that any sufficiently dilute solute obeys Henry’s law, and define a hypothetical, pure-liquid standard
state that makes the solute activity equal to its mole fraction, a
~
(P , y , y ) = y . In Section 16.7, we find that the mole fractions
A A B A
and activities of the components of any binary solution are related by

~ ~
yA d ln aA + yB d ln aB = 0.
For a solute that obeys Henry’s law, we have

yA
~
dln aB = −( ) dln yA
yB
yA d ln yA
= −( ) (dln yB ) ( )
yB d ln yB
yA dyA / yA
= −( ) (d ln yB ) ( )
yB dyB / yB
yA −dyB / yA
= −( ) (d ln yB ) ( )
yB dyB / yB
= d ln yB
This result follows for any choice of standard state for the activity of solvent B . It is satisfied by a
~
= ky , where k is a constant. B B
It is valid even if A is completely nonvolatile. When gas-phase B behaves as an ideal gas, and we choose the ideal gas at P as the o
standard state for both gas- and solution-phase B, we have

~ o o
aB (gas) = fB = PB / P = xB P / P
Since the standard states are the same, the fugacity and activity of B in solution are the same as they are in the gas phase above it.
We have a~
(solution) = ky
B = x P /P
B B . To find k , we consider the system comprised of pure B , for which y = x = 1 and
o
B B
P = P . Substituting, we find k = P / P . With this value for k ,

o
B B
~ o o
aA (solution) = P yB / P = xB P / P
B
so that
yB P = xB P .
B
This is Raoult’s law.
Thus when the solute obeys Henry’s law and the solvent behaves as an ideal gas in the gas
phase above its solution, the solvent obeys Raoult’s law.
Evidently the converse is also true. If the solvent obeys Raoult’s law, y B PB = xB P . With pure ideal gas B as the standard state for
B in both the gas phase and the solution phase, we have
~ ~ o o o
aB (solution) = aB (gas) = fB = PB / P = xB P / P = yB (P /P )
B
so that
~
dln aB (solution) = dln yB
From y A dln
~ ~
aA + yB dln aB =0 and dln a
~
B = dln yB , we have
yB yB dyB dyA
~
dln aA = −( ) dln yB = −( )( ) = = dln yA
yA yA yB yA
so that ~
aA (solution) = kyA , where k is a constant. When we choose the standard state such that a ~
(ss, solution) = 1 when A
yA = 1 , we find k = 1 and ~
aA (solution) = yA . The activity of the solute is related to its fugacity and the fugacity of its standard
state by
fA (solution)
~
aA (solution) = yA =
fA (ss, solution)
When y = 1 , the fugacity is that of the standard state, which is a system of the hypothetical pure liquid in equilibrium with its
A
own ideal gas. Letting the pressure of this ideal gas be ĸ , we have f (ss, solution) = ĸ , so that f (solution) = ĸ y , which
A A A A A A
is equal to the fugacity of the gas with which it is at equilibrium. The fugacity of the ideal gas is x P , so that
A
xA P = ĸ A yA .
This is Henry’s law. Thus, if solvent B obeys Raoult’s law, solute A obeys Henry’s law.
This page titled 16.8: When the Solute Obeys Henry's Law, the Solvent Obeys Raoult's Law is shared under a CC BY-SA 4.0 license and was
We have found the chemical potential of any component in an ideal solution. Now let us find some other thermodynamic properties
of an ideal solution. The value of an extensive thermodynamic property of the solution will be the sum of the values of that
property for the separate pure components plus the change that occurs when these components are mixed. (The initial state of the
system comprises the pure, separate components at a particular temperature and pressure. The mixed state is a homogeneous liquid
solution at the same temperature and pressure.) If the solution contains n moles of component A and n moles of component B ,
A B
the Gibbs free energy is
Gsolution (nA , nB ) = nA μA + nB μB
Dividing through by n A
+ nB to find the Gibbs free energy of the mixture per mole of solution, we have
¯
¯¯¯
Gsolution (yA , yB ) = yA μA + yB μB
o o
= yA Δf G (A) + yA RT ln yA + yB Δf G (B) + yB RT ln yB
To make this mixture, we need y moles of A and y moles of B . The Gibbs free energy of these amounts of unmixed pure A and
A B
B , each in its standard state, is
¯
¯¯¯ o o
Ginitial = yA Δf G (A) + yB Δf G (B)
For the process of mixing pure A and pure B , each originally in its standard state, to form one mole of an ideal solution, the Gibbs
free energy change is
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
Δmix G = Gsolution − Ginitial = yA RT ln yA + yB RT ln yB
In Section 13.3, we found this same relationship for mixing ideal gases:
¯
¯¯¯
Δmix G (gas) = xA RT ln xA + xB RT ln xB
From (∂ Δ mix G/∂T )

P
= −Δmix S , we find
¯¯
¯
Δmix S = −yA Rln yA − yB Rln yB
and the entropy of the liquid solution is

¯¯
¯
o o
S solution = yA S + yB S − yA Rln yA − yB Rln yB
A B
From (∂ Δ mix G/∂P )

T
= Δmix V , we find
¯
¯¯¯
Δmix V =0
and from Δ mix H = Δmix G + T Δmix S , we find

¯
¯¯¯
¯
Δmix H = 0
¯¯
¯
Thus, Δ mix S ,Δ
mix V
¯
¯¯¯
, and Δ ¯
¯
mix H
¯¯
¯
for forming an ideal solution are identical also to the relationships we found for mixing ideal
gases.
¯
¯¯¯
These results have an important physical interpretation. That Δ V = 0 implies that the molecules of A and the molecules of B
mix
occupy the same volume in the mixture as they do in the pure state. From Δ V = 0 and Δ H = 0 , it follows that mix
¯
¯¯¯
mix
¯
¯¯¯
¯
E = 0 at constant pressure. In turn, this implies that the forces between an A molecule and a B molecule are the same as the
¯
¯¯¯
Δ mix
forces between two A molecules or between two B molecules. If the force of attraction between an A molecule and a B molecule
were stronger than that between two A molecules or between two B molecules, molecules in the mixture would be—on average—
¯
¯¯¯
closer together in the mixture than in the separate components; we would find Δ V < 0 . Moreover, the potential energy of the mix
mixed state would be lower than that of the separate components; the mixing process would evolve heat at constant temperature;
¯
¯¯¯
we would find Δ E < 0 . mix
Conversely, if the repulsive force between an A molecule and a B molecule were stronger than the repulsive forces between two A
molecules or between two B molecules, the average separation would be greater in the mixture; we would find Δ V > 0 . The mix
¯
¯¯¯
potential energy of the mixed state would be greater than that of the separate components; the mixing process would consume heat
at constant temperature; we would find Δ E > 0 .
mix
¯
¯¯¯
In an ideal gas, molecules do not interact at all. In an ideal solution, the molecules must interact, because only their mutual
attraction can keep them in the liquid state. The ideal solution behaves ideally not because the intermolecular interactions are zero
but rather because the intermolecular interactions are the same for all of the kinds of molecules present in the mixture. This
interpretation implies that the vapor pressures of the pure components of an ideal solution should be equal. Even for solutions that
follow Raoult’s law quite closely, this expected equality is often imperfectly realized. Not surprisingly, ideal-solution behavior is
best exhibited when the components are isotopically-substituted versions of the same compound.
In an ideal solution, the activities of the components are equal to their mole fractions. The activity of the solvent depends only on
the solvent mole fraction. The properties of the solvent in an ideal solution are independent of the specific substance that comprises
the solute; they depend only on the concentration of solute particles present. Systems in which this is a useful approximation are
sufficiently common that their properties are given a special name: A colligative property of a solution is a property that depends
only on the concentration of solute particles and not on the specific chemical properties of the solute. We expect this approximation
to become better as the solute concentration approaches zero. When a solute obeys Raoult’s law or Henry’s law, its effects on the
thermodynamic properties of the solvent depend only on the concentration of the solute. Consequently, Raoult’s law and Henry’s
law prove to be useful when we seek to model colligative properties.
In Sections 16.10 through 16.14, we evaluate five colligative properties: boiling-point elevation, freezing-point depression, osmotic
pressure, solid-solute solubility and gas-solute solubility. We derive the first three of these properties from the perspective that they
enable us to determine the molar mass of solutes. However, boiling-point elevation, freezing-point depression, and osmotic
pressure are important methods for the measurement of activity coefficients in non-ideal solutions. To illustrate the measurement of
activity coefficients, we develop a more detailed analysis of freezing-point depression in Section 16.15.
This page titled 16.9: Properties of Ideal Solutions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
The system we envision when we talk about boiling-point elevation is described schematically in Figure 7. We consider a solution
of two components, A and B . The mole fractions of A and B , y and y , specify the composition of the solution. We suppose that
A B
one of the components is present at a low concentration. We call this component the solute, and designate it as compound A . Under
these assumptions, we have y ≈ 0 and y = 1 − y ≈ 1 . We assume further that A is nonvolatile, by which we mean that the
A B A
vapor pressure of pure A , P , is very small. Then the second component, B , comprises most of the material of the system. We call
A
component B the solvent. We suppose that the A –B solution is in equilibrium with a gas phase. In principle, molecules of both
components are present in this gas. Since we assume that essentially no component A is present in the gas phase, we have x = 0 A
and x = 1 . We assume also that gas-phase B behaves as an ideal gas and solute A obeys Henry’s law.
B
Figure 7. Schematic description of boiling-point elevation.

When we measure the boiling point of a liquid system, we find the temperature at which the vapor pressure of the system becomes
equal to a specified value. For the normal boiling point, this pressure is 1 atmosphere, or 1.01325 bars. At the boiling point, liquid-
phase solvent is in equilibrium with gas-phase solvent, so that the chemical potential of liquid-phase solvent is equal to the
chemical potential of gas-phase solvent. That is, we have
μB,solution = μB,gas.
We want to describe the change in the equilibrium position that occurs when there is an incremental change in the solute
concentration, dy , while the pressure of the system remains constant. If the system is to remain at equilibrium,
A
μ =μ
B,solution must remain true. It follows that the chemical potentials of the two phases must change in tandem. Continued
B,gas
equilibrium implies that dμ = dμ

B,solution when the solute concentration changes by dy .
B,gas A
We can analyze the boiling-point elevation phenomenon for any fixed pressure at which pure liquid B can be at equilibrium with
pure gas B . Let us designate the fixed pressure as P . We designate the boiling-point temperature of pure solvent B , at P , as
# #
TB ; thus, P = P (T ) , where P (T ) designates the equilibrium vapor pressure of pure solvent B at temperature T . Our
#
B B B B B
goal is to find the temperature at which a binary solution is in equilibrium with pure gas B at the fixed pressure P . We let T be #
bp
the boiling temperature of the solution at P . The composition of the solution is specified by the solute concentration,
#
yA = 1 −y B.
Since we assume that the solute obeys Henry’s law, we choose the standard state for solute A to be the pure hypothetical liquid A
whose vapor pressure is ĸ at T . We suppose that ĸ is exceedingly small. From Section 16.4, we then have a
A A
~
= y , so A,solution A
that
~
d ln aA,solution = dln yA
at any temperature. From Section 16.8, we have

~
d ln aB,solution = dln yB
Pure liquid-phase solvent is at equilibrium with gas-phase solvent at P = P (T ) and T . We imagine that we create a solution
#
B B B
by adding a small amount of solute A , making the concentrations of solute and solvent y and y = 1 − y , respectively. We
A B A
maintain the system pressure constant at P , while changing the temperature to maintain equilibrium between gas-phase and
#
solution-phase solvent B . The new temperature is T . bp
The pressure of gas-phase B is constant at P . The temperature goes from T to T . We choose the activity standard state to be
#
B bp
pure gas B at P and T. This means that the activity of the pure gas is unity at every temperature, so that dln a
# ~
=0 . B,gas
It is worthwhile to note that we can arrive at this conclusion from a different perspective: From Section 14.14, the incremental
change in the activity is
¯
¯¯¯
¯ ~o
HB H
~ B
d ln aB,gas = (− + ) dT
2 2
RT RT
¯
¯¯¯
¯ ~o
where H is the partial molar enthalpy of gas-phase B at T , and H is the partial molar enthalpy of B in its activity standard state
B B
~ ¯
¯¯¯
¯ o
at T . Since we assume that the gas phase is essentially pure B , we have H = H and, again, d ln a~
=0 . B B B,gas
From Section 14.3, we have the general result that

¯
¯¯¯ ¯¯
¯ ~
dμB = V B dP − S B dT + RT (dln aB )
The system pressure and temperature are P = P #

and T = Tbp . For both the gas phase and the solution phase, we have dP = 0
and dT = dT . Since dln a

bp
~
= 0 , we have
B,gas
¯¯
¯
dμB,gas = −S B,gasdTbp
Since d ln a
~
B,solution = dln yB , we have
¯¯
¯
dμB,solution = −S B,solutiondTbp + RTbp (dln yB )
The chemical potential of the pure, constant-pressure, gas-phase solvent depends only on temperature. The chemical potential of
the constant-pressure, solution-phase solvent depends on temperature and solute concentration. Equilibrium is maintained if
dμB,solution = dμB,gas
Substituting, we have
¯¯
¯ ¯¯
¯
−S B,solutiondTbp + RTbp (dln yB ) = −S B,gasdTbp
Since y B = 1 − yA ≈ 1 ,
dln yB = dln(1 − yA ) = −dyA /(1 − yA ) ≈ −dyA
The relationship dμ B,solution = dμB,gas becomes

¯¯
¯ ¯¯
¯
−S B,solutiondTbp − RTbp dyA = −S B,gasdTbp
or
¯¯
¯
S B,gas ¯¯
¯
dyA = ( S B,solutionRTbp ) dTbp
−
We consider systems in which the boiling point of the solution, T , is little different from the boiling point of the pure solvent, T .
bp B
Then, T ≈ T , and T /T ≈ 1 . We let ΔT = T − T , where |ΔT | ≪ T . Since the solution is almost pure B , the partial
B bp bp B bp B B
molar entropy of B in the solution is approximately that of pure B. Consequently, this partial molar entropy difference is, to a good
approximation, just the entropy of vaporization of the solvent, at equilibrium, at the boiling point for the specified system pressure,
P
#
=P
B
(TB ) . Then, since the vaporization of pure B at P #
and T is a reversible process,
B
¯¯
¯ ¯¯
¯ ¯¯
¯ ¯¯
¯
(S B,gas − S B,solution) ≈ ( S B,gas − S B,liquid) = Δvap SB = Δvap HB / TB
#
P , Tbp #
P , TB
so that
Δvap HB
dyA = ( ) dTbp
RTbp TB
In the solution, the solute mole fraction is y ; in the pure solvent, it is zero. At
A P
#
and TB , Δvap HB is a constant. Integrating,
between the limits (0, T ) and (y , T ), we have
B A bp
yA Tbp
Δvap HB dTbp
∫ dyA = ∫
0
RTB TB
Tbp
and
Δvap HB Tbp
yA = ln
RTB TB
Introducing the approximation ln x ≈ x − 1 , which is valid for x ≈ 1 , we have

Δvap HB Tbp Δvap HB
yA = ( − 1) = ΔT
2
RTB TB RT
B
Solving for ΔT ,
2
RT
B
ΔT = ( ) yA
Δvap HB
Since the enthalpy of vaporization and the mole fraction are both greater than zero, ΔT = T − T > 0 ; that is, the addition of a bp B
non-volatile solute increases the boiling point of a liquid system. By measuring ΔT , we can find y ; if we know the molar mass of A
the solvent, we can calculate the number of moles of solute in the solution. If we know the mass of the solute used to prepare the
solution, we can calculate the molar mass of the solute.
Frequently it is useful to express the solute concentration as a molality rather than a mole fraction. Using the dilute-solution
¯¯¯¯
¯
relationship between mole fraction and molality from Section 16.6, y A = M B mA /1000
––
, the boiling-point elevation becomes:
2 ¯¯¯¯
¯
RT MB
B
ΔT = ( )( ) mA
Δvap HB 1000 ––
Our theory predicts that the boiling-point elevation observed for a given solvent is proportional to the solute concentration and
independent of the molecular characteristics of the solute. Experiments validate this prediction; however, its accuracy decreases as
the solute concentration increases. Letting
2
RT
B
ĸB =
Δvap HB
and
2 ¯¯¯¯
¯
RT MB
∗ B
ĸB =
1000 Δvap HB
∗ ∗
we have ΔT = ĸ y and ΔT = ĸ m . We call ĸ or ĸ the boiling-point (or boiling-temperature) elevation constant for
B A B A B B
–– ∗
solvent B . For practical determination of molecular weights, we usually find ĸ or ĸ by measuring the increase in the boiling
B B
point of a solution of known composition.
This page titled 16.10: Colligative Properties - Boiling-point Elevation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The boiling point of a pure solvent, at a specified pressure, is the temperature at which the chemical potential of the pure solvent
gas is equal to the chemical potential of the pure solvent liquid. The boiling point of a solution that contains a nonvolatile solute is
the temperature at which the chemical potential of the pure solvent gas is equal to the chemical potential of the solvent in the
solution. In the preceding section, we found that the boiling point of the solution is greater than the boiling point of the pure
solvent. The temperature difference is the boiling-point elevation.
Similarly, the freezing point of a pure solvent, at a specified pressure, is the temperature at which the chemical potential of the
pure-solid solvent is equal to the chemical potential of the pure-liquid solvent. The freezing point of a solution is the temperature at
which the chemical potential of the pure-solid solvent is equal to the chemical potential of the solution-phase solvent. We find that
the freezing point of the solution is less than the freezing point of the pure solvent. The temperature difference is the freezing-point
depression.
In the boiling-point elevation case, we assume that the pure solvent gas contains no solute. In the freezing-point depression case,
we assume that the pure solvent solid contains no solute. We find the relationship between composition and freezing-point
depression by an argument very similar to that for boiling-point elevation. The equilibrium state in the freezing-point depression
experiment is described schematically in Figure 8.
Figure 8. Schematic description of freezing-point depression.

Since the equilibrium temperature decreases as the solute concentration increases, we can realize the equilibrium state
experimentally by slowly cooling a solution of the specified composition. We determine the temperature at which the first, very
small, crystal of solid solvent forms. Because the pure solvent freezes at a higher temperature than any solution, the first crystal
formed is nearly pure solid solvent. Since this first crystal is very small, its formation does not change the composition of the
solution significantly. Hence, the solution is in equilibrium with pure solid solvent at this temperature; we call this temperature the
freezing point of the solution.
In practice, it is common to determine the melting point of a solid mixture rather than the freezing point of the liquid solution. The
temperature of the solid mixture is increased slowly. As the mixture melts to form a homogeneous solution, the solute–solvent ratio
in the melt approaches the ratio in which the mixture was prepared. When the last bit of solid melts, the composition of the solution
is known from the manner of preparation. This last bit of solid melts at the highest temperature of any part of the mixture. It
contains therefore the smallest proportion of solute. If this last bit of solid is in fact pure solvent, the temperature at which the last
solid melts is the freezing point of the liquid solution. In the limit that the freezing-point and melting-point experiments are carried
out reversibly, the state of the freezing-point system just after the first bit of solid freezes is the same as the state of the melting-
point system just before the last bit of solid melts.
We again specify the composition of the solution by the mole fractions of A and B . We let the solute be compound A , and assume
that its concentration is low. We let the solute concentration be y , where y = 1 − y , y ≈ 0 and y = 1 − y ≈ 1 . The A –B
A A B A B A
solution is in equilibrium with pure solid B . We want to find the temperature at which these phases are in equilibrium. At this
temperature, μ =μ
B,solution , and hence dμ
B,solid = dμ for any change that takes the system to a new equilibrium
B,solution B,solid
state.
We can analyze the freezing-point depression phenomenon for any fixed pressure at which pure-liquid B can be in equilibrium
with pure-solid B . Let us designate the fixed pressure as P and the freezing-point temperature of pure-liquid B, at P , as T .
# #
F
Our goal is to find the temperature at which a binary solution is in equilibrium with pure-solid B at the fixed pressure P . We let #
Tfp be the freezing-point temperature of the solution at P . We base our analysis on the assumption that A obeys Henry’s law.
#
We let the pure-solid solvent be the standard state for the solid solvent (see Section 15.5). Then, at every temperature,
μ = μ̃
B,solid
o
, and
B,solid
o ~
μB,solid − μ̃B,solid = RT ln aB,solid = 0
At every temperature, ~
aB,solid = 1 so that ~
(ln aB,solid )
PT
=0 . The system pressure is constant at P , so dP #
=0 . In the general
expression
¯
¯¯¯ ¯¯
¯
~
dμB,solid = V B,soliddP − S B,soliddT + RT (dln aB,solid )
PT
only the term in dT is non-zero. Recognizing that dT is the change in the freezing-point temperature at P
#
, for some change in
the chemical potential of pure solid B , we have
¯¯
¯
dμB,solid = −S B,solid dTf p
We let the pure-liquid solvent at its equilibrium vapor pressure be the standard state for the solution-phase solvent (see Section
16.2). We designate this equilibrium vapor pressure as P (T ) . Now, since we ultimately find T < t >, the pure liquid freezes
B
fp fp f
spontaneously at T . The standard state for the liquid solvent is therefore a hypothetical state; it is a pure, super-cooled liquid. The
fp
properties of this hypothetical liquid can be estimated from our theory; however, except possibly in unusual circumstances, they
cannot be measured directly. Since we assume that the solute obeys Henry’s law, we have from Section 16.8 that
~
dln a = dln y
B,solution . Thus, while the activity of the pure-solid solvent is constant, the activity of the solvent in the solution
B
varies with the solute concentration. We have

¯¯
¯
dμB,solution = −S B,solution dTf p + RTf p (dln yB )
PT
Using dln y B ≈ −dyA , the relationship dμ B,solution = dμB,solid becomes

¯¯
¯ ¯¯
¯
−S B,solution dTf p − RTf p dyA = −S B,solid dTf p
or
¯¯
¯ ¯¯
¯
⎛ S ⎞
B,solution − S B,solid
dyA = − ⎜ Tf p ⎟ dTf p
R
⎝ ⎠
We consider systems in which the freezing point of the solution, T , is little different from the freezing point of the pure solvent,
fp
T . Then, T
F F≈T , and T /T ≈ 1 . We let |ΔT | = T − T , where |ΔT | ≪ T . Since the solution is almost pure B , the
fp fp F F fp F
partial molar entropy of B in the solution is approximately that of pure liquid B . Consequently, the partial molar entropy difference
is, to a good approximation, just the entropy of fusion of the pure solvent, at equilibrium, at the freezing point for the specified
system pressure. That is,
¯¯
¯ ¯¯
¯ ¯¯
¯ ¯¯
¯
( S B,solution − S B,solid) ≈ ( S B,liquid − S B,solid) = Δf us SB = Δf us HB / TF
# #
P , Tfp P , TF
so that
Δf us HB
dyA = − ( ) dTf p
RTf p TF
At P and T , Δ H is a constant. In the solution, the solute mole fraction is y ; in the pure solid solvent, it is zero. Integrating
#
F f us B A
between the limits (0, T ) and (y , T ), we have

F A fp
yA Tfp
Δf us HB dTf p
∫ dyA = − ∫
0
RTF TF
Tf p
and
Δf us HB Tf p
yA = − ln
RTF TF
Introducing ln x ≈ x − 1 , we have
Δf us HB Tf p Δf us HB
yA = − ( − 1) = ΔT
2
RTF TF RT
F
Solving for ΔT ,
2
RT
F
ΔT = ( ) yA
Δf us HB
The fusion process is endothermic, and Δ H > 0 . Therefore, we find ΔT = T − T > 0 ; that is, the addition of a solute
f us B F fp
decreases the freezing point of a liquid. The depression of the freezing point is proportional to the solute concentration.
Since measurement of ΔT enables us to find y , freezing-point depression—like boiling point elevation—enables us to determine
A
the molar mass of a solute. In our discussion of boiling-point elevation, we noted that it is often convenient to express the
concentration of a dilute solute in units of molality rather than mole fraction. This applies also to freezing-point depression.
Likewise, for practical applications, we usually find the freezing-point depression constant by measuring the depression of the
freezing point of a solution of known composition.
This page titled 16.11: Colligative Properties - Freezing-point Depression is shared under a CC BY-SA 4.0 license and was authored, remixed,
The phenomena of boiling-point elevation and freezing-point depression involve relationships between composition and
equilibrium temperature—at constant system pressure. We turn now to a phenomenon, osmotic pressure, which involves a
relationship between composition and equilibrium pressure—at constant system temperature.
To analyze boiling-point elevation, we equate the chemical potential of the solvent in two subsystems, a solution and the gas phase
above it. To analyze freezing-point depression, we equate the chemical potential of the solvent in solution and solid subsystems.
Similarly, to analyze osmotic pressure, we equate the chemical potential of the pure solvent—at one pressure—to the chemical
potential of the solvent in a solution—at a second pressure. We find that equilibrium can be obtained only when the pressure in the
solution subsystem exceeds the pressure in the solvent subsystem. The difference between these two pressures is the osmotic
pressure.
In the boiling-point elevation and freezing-point depression phenomena, the subsystems are separated by a phase boundary. In the
osmotic pressure phenomenon, a pure solvent phase is separated from a solution phase by a semi-permeable membrane. A semi-
permeable membrane allows free passage to solvent molecules; however, solute molecules cannot pass through it. In practice, the
semi-permeable membrane is a material that is penetrated by pores, or channels, whose cross-sectional dimensions are nearly as
small as typical solvent molecules. Solvent molecules can diffuse through these pores and pass from one side of the membrane to
the other. With such a membrane, we can satisfy the osmotic pressure conditions by choosing a solute whose molecules are larger
than the pore diameters, because large molecules will be unable to pass through the pores. In practice, the solute in osmotic
pressure experiments is typically a polymer or a biologically derived molecule of high molecular weight. Osmotic pressure
measurements have been an important source of data on the molar masses of such substances.
Figure 9. Schematic description of osmotic pressure.

The osmotic pressure experiment is described schematically in Figure 9. The semi-permeable membrane must be sufficiently
robust to support the pressure drop between the two subsystems. At constant pressure, mixing of the two subsystems is a
spontaneous process. Were we to remove the membrane and the pressure drop that it supports, the subsystems would mix to form a
single, more dilute solution. We see therefore that there is a tendency for net migration of solvent molecules from the solvent side
of the membrane to the solution side. We can oppose this tendency by applying additional pressure on the solution side. Evidently,
for any given solution composition, there will be an applied pressure at which the subsystems are in equilibrium with one another.
We let the pure-liquid solvent at its equilibrium vapor pressure be the standard state for both the pure-liquid and the solution-phase
solvent (see Section 16.2). For the two subsystems to be in equilibrium, we must have μ =μ
B,soluton
. For any change that
B,solvent
takes one equilibrium state to another, we have dμ = dμ

B,soluton
. Since the pure solvent subsystem contains only B , we
B,solvent
have a~
B,solvent= constant so that dln a
~
= 0 . Since the temperature is constant, we have dT = 0 . For the solvent
B,solvent
subsystem, the general expression for dμ reduces to

B,solvent
¯
¯¯¯
dμB,solvent = V B,solvent
dPsolvent
For the solution subsystem, dT = 0 . Assuming the solvent in the solution obeys Raoult’s law, we have ~
aB,solution = yB . The
general equation for dμ B,soluton
reduces to
¯
¯¯¯
dμ =V B,solutiondPsolution + RT (dln yB )
B,soluton
Using dln y B ≈ −dyA , the relationship dμ B,solution

= dμB,solvent becomes
¯
¯¯¯ ¯
¯¯¯
V B,solutiondPsolution − RT dyA = V B,solvent
dPsolvent
The molar volume of a liquid is nearly independent of the system pressure. Because the solution is nearly pure solvent, the molar
¯
¯¯¯ ¯
¯¯¯ ¯
¯¯¯
volume of B in the solution is approximately equal to the molar volume of pure solvent B . Letting V B,solution =V B,solvent
=V B
,
this becomes
¯
¯¯¯ ¯
¯¯¯
⎛V ⎞ ⎛V ⎞
B B
dyA = (dPsolution − dPsolvent ) = d (Psolution − Psolvent )
⎝ RT ⎠ ⎝ RT ⎠
This pressure difference is the osmotic pressure; it is often represented by the Greek alphabet capital pi: Π = P −P . solution solvent
The osmotic pressure of the pure solvent must be zero; that is, Π = 0 when y = 0 . Integrating between the limits (0, 0) and
A
(y , Π) , we have
A
yA ¯
¯¯¯ Π
V B
∫ dyA = ∫ dΠ
0 RT 0
and
¯
¯¯¯
V Π
B
yA =
RT
or
¯
¯¯¯
ΠV B
= yA RT
From this equation, we see that the osmotic pressure must be positive; that is, at equilibrium, the pressure on the solution must be
greater than the pressure on the solvent: Π = P −P >0 .
solution solvent
The osmotic pressure equation can be put into an easily remembered form. For n A
≪ nB ,y A
= nA /(nA + nB ) ≈ nA / nB . With
this substitution, ¯
¯¯
Π ( nB V
¯
B
) = nA RT , but since ¯
¯¯
V
¯
B
is the molar volume of pure B , V = nB V
¯
¯¯¯
B
is just the volume of the
solvent and essentially the same as the volume of the solution. The osmotic pressure equation has the same form as the ideal gas
equation:
¯
¯¯¯
Π ( nB V B
) = Π V = nA RT
This page titled 16.12: Colligative Properties - Osmotic Pressure is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Although the result has few practical applications, we can also use these ideas to calculate the solubility of a solid solute in an ideal
solution. The arguments are similar to those we used to estimate the freezing-point depression of a solution. The freezing point of a
solution is the temperature at which the solution is in equilibrium with its pure-solid solvent. The solubility of a solute is the mole
fraction of the solute in a solution that is at equilibrium with pure-solid solute. In this analysis, we assume that the solid phase is
pure solute. Our analysis does not apply to a solid solution in equilibrium with a liquid solution. The properties of the solvent have
no role in our description of the solid–ideal-solution equilibrium state. Consequently, our analysis produces a model in which the
solubility of an ideal solute depends only on the properties of the solute; for a given solute, the ideal-solution solubility is the same
in every solvent.
We specify the composition of the solution by the mole fractions of A and B , again letting the solute be compound A . When we
consider freezing-point depression, an A –B solution of specified composition is in equilibrium with pure solid B , and we want to
find the equilibrium temperature. When we consider solute solubility, the A –B solution is in equilibrium with pure solid A at a
specified pressure, P , and temperature, T ; we want to find the equilibrium composition. Since pure solid A is present, the
#
temperature must be less than the melting point of pure A . We let the melting point of the pure solute be T , at the specified FA
system pressure.
The activity of pure solid A and the system pressure are both constant; we have dln a
~
A,solid
=0 , dP =0 , and
¯¯
¯
dμA,solid = −S A,soliddT
In the saturated ideal solution in equilibrium with this solid, we have a

~
A,solution = yA , dP =0 , and
¯¯
¯
dμ = −S A,solutiondT + RT (dln yA )
A,solution
The relationship dμ A,solution

= dμA,solid becomes
¯¯
¯ ¯¯
¯
−S A,solutiondT + RT (dln yA ) = −S dT
A,solid
and
¯¯
¯ ¯¯
¯
⎛ S −S ⎞
A,solution A,solid
dln yA =⎜ ⎟ dT
RT
⎝ ⎠
¯¯
¯ ¯¯
¯
Now, S −S
A,solution is the entropy change for the reversible (equilibrium) process in which one mole of pure solid A
A,solid
dissolves in a very large volume of a saturated solution; the mole fraction of A in this solution is constant at y . During this A
process, the pressure and temperature are constant at P and T . Letting the heat absorbed by the system during this process be
#
q
rev
P
, we have
#
¯¯
¯ ¯¯
¯ rev
( S A,solution − S ) =q /T
A,solid #
P
#
P ,T
The heat absorbed is also expressible as the difference between the partial molar enthalpy of A in the solution and that of the pure
solid; that is,
rev ¯
¯¯¯
¯ ¯
¯¯¯
¯
q #
= ( H A,solution − H A,solid)
P
#
P ,T
One of the properties of an ideal solution is that the enthalpy of mixing is zero. Thus, the partial molar enthalpy of A in an ideal
solution is independent of y , so that the partial molar enthalpy of A in an ideal solution is the same as the partial molar enthalpy
A
¯
¯¯¯
¯ ¯
¯¯¯
¯
of pure liquid A ; that is, H A,solution = H A,liquid , and
rev ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
q #
= ( H A,solution − H A,solid) = ( H A,liquid − H A,solid) = (Δf us HA ) # ≈ (Δf us HA ) #
P P ,T P , TFA
# #
P ,T P ,T
Then,
(Δf us HA ) #
¯¯
¯ ¯¯
¯ P , TFA
( S A,solution − S A,solid) ≈
# T
P ,T
Dropping the subscript information and replacing the approximate equality, we have
Δf us HA
dln yA =( ) dT
2
RT
At P and T , Δ H is a property of pure A and is independent of the solution composition. When the pure solid solute melts
#
FA f us A
at T , the solute mole fraction is unity in the liquid phase with which it is in equilibrium: At T , y = 1 . At temperature T , y
FA FA A A
is the solute mole fraction in the liquid-phase solution that is at equilibrium with the pure-solid solute. Integrating between the
limits (1, T ) and (y , T ), we have
FA A
yA T
Δf us HA dT
∫ dln yA = ∫
2
1
R TFA T
and
−Δf us HA 1 1
ln yA = ( − )
R T TF A
For a given solute, Δ H and T are fixed and are independent of the characteristics of the solvent. The mole fraction of A in
f us A FA
the saturated solution depends only on temperature. Since Δ H > 0 and T < t >, we find ln y < 0 . Therefore, we find that
f us A fa A
yA < 1 , as it must be. However, y increases, with T, implying that the solubility of a solid increases as the temperature increases,
A
as we usually observe.
This page titled 16.13: Colligative Properties - Solubility of a Solute in an Ideal Solution is shared under a CC BY-SA 4.0 license and was
A similar analysis yields an equation for the solubility of a gas in a liquid solvent as a function of temperature at a fixed pressure,
P
#
. We refer to the gas component as A and the liquid component as B . We assume that solvent B is nonvolatile, so that the gas
phase with which it is in equilibrium is essentially pure gaseous solute A . We again find that the properties of the solvent have no
role in our model, and the solubility of gas A is the same in every solvent.
We assume that low concentrations of the solute obey Henry’s law and choose the solution-phase standard state for solute A to be
the pure hypothetical liquid A whose vapor pressure is ĸ at T . From Section 16.4, we then have a
A
~
= y , so that A,solution A
~
dln a = dln y
A,solution at any temperature. Substituting into the general equation
A
¯
¯¯¯ ¯¯
¯ ~
dμA = V A dP − S A dT + RT (dln aA )
we have
¯¯
¯
dμA,solution = −S A,solution dT + RT (dln yA )
The pressure of gas-phase A is constant at P , and dP = 0 . We choose the gas-phase activity standard state to be pure gas A at
#
P
#
and T . Since this makes the activity of the pure gas unity at any temperature, we have dln a
~
= 0 . Substituting, we have B,gas
¯¯
¯
dμA,gas = −S A,gas dT
Any constant-pressure process that maintains equilibrium between gas-phase A and solution-phase A must involve the same
change in the chemical potential of A in
each phase, so that dμ A,gas
= dμA,solution , and
¯¯
¯ ¯¯
¯
−S A,gas dT = −S A,solution dT + RT (dln yA )
so that
¯¯
¯ ¯¯
¯
(S A,gas − S A,solution)
dln yA =− dT
RT
¯¯
¯ ¯¯
¯
The difference S −SA,gas is the entropy change for the equilibrium—and hence reversible—process in which one mole of
A,solution
substance A originally in solution vaporizes into a gas phase consisting of essentially pure gas A while the system is at the constant
pressure P . Let us designate the enthalpy change for this reversible process at P and T as Δ H
#
. Then, we have
#
vap
¯
¯¯¯
¯
A,solution
¯
¯¯¯
¯
Δvap H A,solution
¯¯
¯ ¯¯
¯
S A,gas − S A,solution =
T
so that
¯
¯¯¯
¯
Δvap H A,solution
dln yA =− dT
2
RT
Since enthalpy changes are generally relatively insensitive to temperature, we expect that, at least over small ranges of y and T , A
¯
¯¯¯
¯
Δ vapH is approximately constant. Since the vaporization process takes A from a state in which it has some of the
A,solution
characteristics of a liquid into a gaseous state, we can be confident that Δ H > 0 . This conclusion implies that
vap
¯
¯¯¯
¯
A,solution
dln yA
<0
dT
Thus, our thermodynamic model leads us to the conclusion that the solubility of gas A decreases as the temperature increases. That
the solubilities of gases generally decrease with increasing temperature is a well-known experimental observation. It stands in
contrast to the observation that the solubilities of liquid or solid—at P and T —substances generally increase with increasing
#
temperature. Our analysis of gas solubility provides a satisfying theoretical interpretation for an experimental observation which
otherwise appears to be counterintuitive.
The meaning of ¯
¯¯¯
¯
Δvap H A,solution is unambiguous. Our analysis enables us to measure it by experimentally measuring yA as a
¯
¯¯¯
¯
function of T . We can estimate Δ H vap from another perspective: When we consider the “solution” in which y = 1 , the
A,solution A
vaporization process is the vaporization of liquid A into a gas phase of pure A at P and T . Since we assume that A is stable as a
#
gas at T , the boiling point of pure liquid A must be less than T at P and the vaporization A must be a spontaneous process at
#
P
#
and T . The enthalpy of vaporization datum which is most accessible for liquid A is that for the reversible vaporization at one
atmosphere and the normal boiling point, T , which we designate as Δ H . If we stipulate that P is one atmosphere; assume
B vap A
o #
that our solubility equation remains valid as y increases from y ≈ 0 to y = 1 ; and assume that the enthalpy of vaporization is
A A A
approximately constant between the boiling point of A and T , we have Δ H =Δ H . Then,

vap
¯
¯¯¯
¯
A,solution vap
o
A
o
Δvap H
A
dln yA =− dT
2
RT
and
yA T o
Δvap H
A
∫ dln yA =∫ − dT
2
1 TB RT
so that
o
Δvap H 1 1
A
ln yA = ( − )
R T TB
Check
Viewed critically, the accuracy of the approximation
¯
¯¯¯
¯ o
Δvap H A,solution ≈ Δvap H
A
is dubious. The assumptions we make to reach it are essentially equivalent to assuming that the cohesive forces in solution are
about the same between A molecules as they are between A molecules and B molecules. We expect this approximation to be
more accurate the more closely the solution exhibits ideal behavior. However, if solvent B is to satisfy our assumption that the
solvent is nonvolatile, the cohesive interactions between B molecules must be greater than those between A molecules, and
this not consistent with ideal-solution behavior.
This page titled 16.14: Colligative Properties - Solubility of a Gas is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
The analysis of freezing-point depression that we present in Section 16.11 introduces a number of simplifying assumptions. We now undertake a
more rigorous analysis of this phenomenon. This analysis is of practical importance. Measuring the freezing-point depression of a solution is one
way that we can determine the activity and the activity coefficient of the solvent component. As we see in Section 16.7, if we have activity
coefficients for the solvent over a range of solute concentrations, we can use the Gibbs-Duhem equation to find activity coefficients for the solute.
Freezing-point depression measurements have been used extensively to determine the activity coefficients of aqueous solutes by measuring the
activity of water in their solutions.
As in our earlier discussion of freezing-point depression, the equilibrium system is a solution of solute A in solvent B , which is in phase
equilibrium with pure solid solvent B . Our present objective is to determine the activity of the solvent in its solutions at the melting point of the
pure solvent. Having obtained this information, we can use the Gibbs-Duhem relationship to find the activity of the solute, as a function of solute
concentration, at the melting point of the pure solvent. Once we have the solute activity at the melting point of the pure solvent, we can use the
methods developed in Section 14.14 to find the solute activity in a solution at any higher temperature.
In Section 14.14, we find the temperature dependence of the natural logarithm of the chemical activity of a component of a solution. For a
particular choice of activity standard states and enthalpy reference states, we develop a method to obtain the experimental data that we need to
apply this equation. For brevity, let us refer to these choices as the infinite dilution standard states. In order to determine the activity of a solvent in
its solutions at the melting point of the pure solvent, it is useful to define an additional standard state for the solvent. At temperatures below the
normal melting point, which we again designate as T , we let the activity standard state of the solvent be pure solid B . Above the melting point, we
F
use the infinite dilution standard state that we define in Section 14.14; that is, we let the activity standard state of the solvent be pure liquid solvent
B.
At and below the melting point, T , the activity standard state for the solvent, B , is pure solid B . At and above the melting point, the activity
F
standard state for the solvent is pure liquid B . At the melting point, pure solid solvent is in equilibrium with pure liquid solvent, which is also the
solvent in an infinitely dilute solution. At T , the activity standard state chemical potentials of the pure solid solvent, the pure liquid solvent, and
F
the solvent in an infinitely dilute solution are all the same. It follows that the value that we obtain for the activity of the solvent at T , for any
F
particular solution, will be the same whether we determine it from measurements below T using the pure solid standard state or from
F
measurements above T using the infinitely dilute solution standard state.

F
Now let us consider the chemical potential of liquid solvent B in a solution whose composition is specified by the molality of solute A , ––
m , when
A
the activity standard state is pure solid B . We want to find this chemical potential at temperatures in the range $T_{fp}<t_f$>, where T is the fp
freezing point of the solution whose composition is specified by ––

m . In this temperature range, we have
A
μB (solution, m , T )
––A
o ~
= μ̃B (pure solid, T ) + RT ln aB (solution, m , T )
––A
Using the Gibbs-Helmholtz equation, we obtain

~
∂ln aB (solution, mA , T )
––
( )
∂T
P , mA
––
¯
¯¯¯
¯ ~o
−H B (solution, m , T ) H B (T )
––A
= +
2 2
RT RT
¯
¯¯¯
¯ ¯
¯¯¯
¯
− [H B (solution, mA , T ) − H B (solid, T )]
––
=
2
RT
~o ¯
¯¯¯
¯
where we have H (T ) = H (solid, T ) , because pure solid B is the activity standard state. Using the ideas developed in Section 14.14, we can
B B
use the thermochemical cycle shown in Figure 10 to evaluate

¯
¯¯¯
¯ ¯
¯¯¯
¯
H B (solution, mA , T ) − H B (solid, T )
––
Figure 10. Enthalpy cycle for the temperature dependence of ln a
~
. B
¯
¯¯¯
¯ ¯
¯¯¯
In this cycle, Δ H (T ) is the molar enthalpy of fusion of pure B at the melting point, T . L (solution, ––
f us B F m F B A
,T) is the relative partial molar
enthalpy of B at T in a solution whose composition is specified by ––
F m . The only new quantity in this cycle is
A
T ¯
¯¯¯
¯
∂ H B (solution, m , T )
––A
∫ ( ) dT
Tf
∂T
P,m
––A
We can use the relative partial molar enthalpy of the solution to find it. By definition,
¯
¯¯¯
LB (solution, m , T )
––A
∂H (solution, m , T ) ¯
¯¯¯
¯
ref
––A
=( ) −HB (T )
∂nB
P ,T ,nA
or, dropping the parenthetical information,

ref
¯
¯¯¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
LB = H B − H B (T )
so that
ref
¯
¯¯¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
∂LB ∂H B ⎛ ∂H ⎞
B
( ) =( ) −
∂T ∂T ⎝ ∂T ⎠
P P P
In Section 14.14, we introduce the relative partial molar heat capacity,

¯
¯¯¯
¯¯
¯ ∂LB
J B (T ) = ( )
∂T
P
Since the infinitely dilute solution is the enthalpy reference state for B in solution, we expect the molar enthalpy of pure liquid B to be a good
ref
¯
¯¯¯
¯
approximation to the partial molar enthalpy of liquid B in the enthalpy reference state. Then, (∂ H B
/∂T ) is just the molar heat capacity of pure
P
liquid B , CP (B, liquid, T ) . (See problem 16-11.) We find

¯
¯¯¯
¯
∂H B ¯¯
¯
( ) = J B (T ) + CP (B, liquid, T )
∂T
P
Using this result, the enthalpy changes around the cycle in Figure 10 yield
¯
¯¯¯
¯ ¯
¯¯¯
¯
H B (solution, mA , T ) − H B (solid, T )
––
TF T
¯
¯¯¯
¯ o ¯
¯¯¯ ¯¯
¯
=∫ CP (B, solid, T )dT + Δf us H B (TF ) − L (TF ) + LB (m , TF ) + ∫ [J B (T ) + CP (B, liquid, T )]dT
B ––A
T TF
TF
¯
¯¯¯
¯ ¯
¯¯¯ o ¯¯
¯
= Δf us H B (TF ) + LB (mA , TF ) − L (TF ) − ∫ [CP (B, liquid, T ) − CP (B, solid, T ) + J B ]dT
–– B
T
¯¯
¯
Since we know how to determine J and the heat capacities as functions of temperature, we can evaluate this integral to obtain a function of
B
¯¯
¯
temperature. For present purposes, let us assume that J and the heat capacities are essentially constant and introduce the abbreviation
B
ΔCP = CP (B, liquid, T ) − CP (B, solid, T ) , so that
TF
¯¯
¯ ¯¯
¯
∫ [CP (B, liquid, T ) − CP (B, solid, T ) + J B ]dT = (ΔCP + J B ) (TF − T )
T
The temperature derivative of ln a

~
B (solution, m , T )
––A
becomes
~
∂ln aB (solution, mA , T )
––
( )
∂T
P , mA
––
¯¯
¯
¯
¯¯¯
¯ ¯
¯¯¯ o (ΔCP + J B ) (TF − T )
−Δf us H B (TF ) − LB (m , TF ) + L (TF )
––A B
= +
2 2
RT RT
Tf pis the freezing point of the solution whose composition is specified by –– m . At T the solvent in this solution is at equilibrium with pure solid
A fp
solvent. Hence, the chemical potential of the solution solvent is equal to that of the pure-solid solvent. Then, because the pure solid is the activity
standard state for both solution solvent and pure-solid solvent at T , the activity of the solution solvent is equal to that of the pure-solid solvent.
fp
Because the pure solid is the activity standard state, the solvent activity is unity at T . This means that we can integrate the temperature derivative
fp
from T to T to obtain
fp F
TF
~
∫ d ln aB (solution, m , T )
––A
Tfp
~
= ln aB (solution, mA , TF )
––
¯
¯¯¯
¯ ¯
¯¯¯ o ¯¯
¯
Δfus H B (TF ) + LB (mA , TF ) + L (TF ) 1 1 ΔCP + J B TF TF
–– B
=( )( − ) −( ) (1 − + ln( ))
R TF Tf p R Tf p Tf p
Thus, from the measured freezing point of a solution whose composition is specified by m
––A
, we can calculate the activity of the solvent in that
solution at T . F
Several features of this result warrant mention. It is important to remember that we obtained it by assuming that ΔC + J (T ) is a constant. This P
¯¯
¯
B
is usually a good assumption. It is customary to express experimental results as values of the freezing-point depression, ΔT = T − T . The F fp
activity equation becomes

~
ln aB (solution, mA , TF )
––
¯
¯¯¯
¯ ¯
¯¯¯ o
Δf us H B (TF ) + LB (mA , TF ) + L (TF )
–– B
= −( ) ΔT
RTF Tf p
¯¯
¯
ΔCP + J B ΔT ΔT
+( )( − ln(1 + ) )
R Tf p Tf p
¯
¯¯¯ ¯¯
¯
The terms involving L , L , ΔC , and J are often negligible, particularly when the solute concentration is low. When T /T ≈ 1 , that is,
B
o
B P B F fp
¯¯
¯
when the freezing-point depression is small, the coefficient of ΔC + J is approximately zero. When these approximations apply, the activity
P B
equation is approximated by
¯
¯¯¯
¯
Δf us H B (TF )
~
ln aB (solution, mA , TF ) = − ( ) ΔT
–– 2
RT
F
This page titled 16.15: Solvent Activity Coefficients from Freezing-point Depression Measurements is shared under a CC BY-SA 4.0 license and was authored,
remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available
upon request.
Thus far in our discussion of solute activities, we have assumed that the solute is a molecular species whose chemical structure is
unchanged when the pure substance dissolves. This is not the case when salts dissolve in water and other polar solvents. A pure
solid salt exists as a lattice of charged ions, rather than electrically neutral molecular moieties, and its solutions contain solvated
ions. Since salt solutions conduct electricity, we often call them electrolytic solutions. Solutions of salts in water are extremely
important from both practical and theoretical standpoints. Accordingly, we focus our discussion on aqueous solutions; however, the
ideas that we develop apply to salt solutions in any solvent that supports the formation of solvated ions.
We can apply the concepts that we develop in this chapter to measure the activities of aqueous salt solutions. When we do so, we
find new features. These features arise from the formation of aquated ionic species and from electrical interactions among these
species. In this chapter, we consider only the most basic issues that arise when we investigate the activities of dissolved salts. We
consider only strong electrolytes; that is, salts that are completely dissociated in solution. In this section, we briefly review the
qualitative features of such solutions.
Departure from Henry’s law behavior begins at markedly lower concentrations when the solute is a salt than when it is a neutral
molecular species. This general observation is easily explained: Departures from Henry’s law are caused by interactions among
solution species. For neutral molecules separated by a distance r, the variation of the interaction energy with distance is
approximately proportional to r . This means that only the very closest molecules interact strongly with one another. For ions,
−6
Coulomb’s law forces give rise to interaction energies that vary as r . Compared to neutral molecules, ions interact with one
−1
another at much greater distances, so that departures from Henry’s law occur at much lower concentrations.
Our qualitative picture of an aqueous salt solution is that the cations and anions that comprise the solid salt are separated from one
another in the solution. Both the cations and the anions are surrounded by layers of loosely bound water molecules. The binding
results from the electrical interaction between the ions and the water-molecule dipole. The negative (oxygen) end of the water
dipole is preferentially oriented toward cations and the positive (hydrogen) end is preferentially oriented toward anions.
In aqueous solution, simple metallic cations are coordinated to a first layer of water molecules that occupy well-defined positions
around the cation. In this layer, the bonding can have a covalent component. Such combinations of metal and coordinated water
molecules are called aquo complexes. For most purposes, we can consider that the aquo complex is the cationic species in solution.
Beyond the layer of coordinated water molecules, a second layer of water molecules is less tightly bound. The positions occupied
by these molecules are more variable. At still greater distances, water molecules interact progressively more weakly with the
central cation. In general, when we consider the water molecules that surround a given anion, we find that even the closest solvent
molecules do not occupy well-defined positions.
In any macroscopic quantity of solution, each ion has a specific average concentration. On a microscopic level, the Coulomb’s law
forces between dissolved ions operate to make the relative locations of cations and anions less random. It is useful to think about a
spherical volume that surrounds a given ion. We suppose that the diameter of this sphere is several tens of nanometers. Within such
a sphere centered on a particular cation, the concentration of anions will be greater than the average concentration of anions; the
concentration of cations will be less than the average concentration of cations. Likewise, within a microscopic sphere centered on a
given anion, the concentration of cations will be greater than the average concentration of cations; the concentration of anions will
be below average.
As the concentration of a dissolved salt increases, distinguishable species can be formed in which a cation and an anion are nearest
neighbors. We call such species ion pairs. At sufficiently high salt concentrations, a significant fraction of the ions can be found in
such ion-pair complexes. Compared to other kinds of chemical bonds, ion-pair bonds are weak. The ion-pair bond is labile; the
lifetime of a given ion pair is short. At still higher salt concentrations, the formation of significant concentrations of higher
aggregates becomes possible. Characterizing the species present in an electrolytic solution becomes progressively more difficult as
the salt concentration increases.
This page titled 16.16: Electrolytic Solutions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
We can find the activity of a salt in its aqueous salt solutions. For example, we can measure the freezing point depression for
aqueous solutions of sodium chloride, find the activity of water in these solutions as a function of the sodium chloride
concentration, and use the Gibbs-Duhem equation to find the activity of the dissolved sodium chloride as a function of its
concentration. When we do so, we find some marked differences from our observations on molecular solutes.
For molecular solutes, the activity approaches the solute concentration as the concentration approaches zero; that is, the activity
coefficient for a molecular solute approaches unity as the concentration approaches zero. For sodium chloride, and other 1:1
electrolytes, we find that the activity we measure in this way approaches the square of the solute concentration as the concentration
approaches zero. For other salts, the measured activity approaches other powers of the solute concentration as the concentration
approaches zero.
The dissociation of the solid salt into solvated ions explains these observations. Let us consider a solution made by dissolving n
moles of a salt, A B , in n
p q moles of solvent. (Let A be the cation and B the anion.) For present purposes, the cation and
solvent
anion charges are not important. We use p and q to designate the composition of the salt. Typically, we are interested in dilute
solutions, and it is convenient to use the hypothetical one-molal solution as the standard state for the activity of a solute species. We
can represent the Gibbs free energy of this solution as
G = nsolvent μ + nμ
solvent Ap Bq
where μ solvent
and μ Ap Bq
are the partial molar Gibbs free energies of the solvent and the solute in the solution. We can also write
o ~
μAp Bq = μ̃A Bq
+ RT ln aAp Bq
p
where μ̃ o
Ap Bq
is the partial molar Gibbs free energy when a
~
Ap Bq = mA B γA B = 1
–– p q p q
in the activity standard state of the salt.
We assume that A B is a strong electrolyte; its solution contains np moles of the cation,
p q A , and nq moles of the anion, B . In
principle, we can also represent the Gibbs free energy of the solution as
G = nsolvent μsolvent + np μA + nq μB
and the individual-ion chemical potentials as μ = μ̃ + RT ln a ~

and μ = μ̃ + RT ln a
A
o
A
~
, where μ̃ and μ̃ are the partial
A B
o
B B
o
A
o
B
molar Gibbs free energies of the ions A and B in their hypothetical one-molal activity standard states. Equating the two equations
for the Gibbs free energy of the solution, we have
μA Bq
= p μA + q μB
p
and
o ~ o ~p o ~q
μ̃A + RT ln aAp Bq = pμ̃A + RT ln a + qμ̃B + RT ln a
p Bq A B
While it is often experimentally challenging to do so, we can measure μ̃ and a

~
. In principle, the meanings of the o
Ap Bq Ap Bq
individual-ion activities, a
~
and a
~
, and their standard-state chemical potentials, μ̃ and μ̃ , are unambiguous; however, since we
A B
o
A
o
B
cannot prepare a solution that contains cation A and no anion, we cannot make measurements of a ~
or μ̃ that are independent of A
o
A
the properties of B , or some other anion. Consequently, we must adopt some conventions to relate these properties of the ions,
which we cannot measure, to those of the salt solution, which we can.
The universally adopted convention for the standard chemical potentials is to equate the sum of the standard chemical potentials of
the constituent ions to that of the salt. We can think of this as assigning an equal share of the standard-state chemical potential of
the salt to each of its ions; that is, we let
o
μ̃
Ap Bq
o o
μ̃A = μ̃B =
p +q
Then,
o o o
μ̃ = p μ̃ + q μ̃
Ap Bq A B
and the activities of the individual ions are related to that of the salt by
~ ~p ~ q
aA Bq =a a
p A B
We can develop the convention for the activities of the individual ions by representing each activity as the product of a
concentration and an activity coefficient. That is, we represent the activity of each individual ion in the same way that we represent
the activity of a molecular solute. In effect, this turns the problem of developing a convention for the activities of the individual
ions into the problem of developing a convention for their activity coefficients. Using the hypothetical one-molal standard state for
each ion, we write a~
A= m γ
–– A
and a
A
~
= m γ , where m , γ , m , and γ
B
–– B B
–– –– A A
are the molalities and activity coefficients for ions
B B
A and B , respectively. Let the molality of the salt, A B , be m . Then m

p q = p m and m = q m , and
A B
–– –– –– –– ––
~ ~p ~ q p q p q p+q p q
aA =a a = (p mγ ) (q mγ ) = (p q ) m γ γ
p Bq A B –– A –– B –– A B
Now we introduce the geometric mean of the activity coefficients γA and γB ; that is, we define the geometric mean activity
coefficient, γ , by
±
p q 1/(p+q)
γ± = (γ γ )
A B
The activity of the dissolved salt is then given by

~ p q p+q p+q
aAp Bq = (p q ) m γ±
––
The mean activity coefficient, γ , can be determined experimentally as a function of m

±
, but the individual activity Ap Bq
––
coefficients, γ and γ , cannot. It is common to present the results of activity measurements on electrolytic solutions as a table or
A B
a graph that shows the mean activity coefficient as a function of the salt molality.
While we cannot determine the activity or activity coefficient for an individual ion experimentally, no principle prohibits a
theoretical model that estimates individual ion activities. Debye and Hückel developed such a theory. The Debye-Hückel theory
gives reasonably accurate predictions for the activity coefficients of ions for solutions in which the total ion concentration is about
0.01 molal or less. We summarize the results of the Debye-Hückel theory in Section 16.18.
This page titled 16.17: Activities of Electrolytes - The Mean Activity Coefficient is shared under a CC BY-SA 4.0 license and was authored,
In earlier sections, we introduce some basic methods for the experimental measurement of activities and activity coefficients. The
Debye-Hückel theory leads to an equation for the activity coefficient of an ion in solution. The theory gives accurate values for the
activity of an ion in very dilute solutions. As salt concentrations become greater, the accuracy of the Debye-Hückel model
decreases. As a rough rule of thumb, the theory gives useful values for the activity coefficients of dissolved ions in solutions whose
total salt concentrations are less than about 0.01 molal. The theory is based on an electrostatic model. We describe this model and
2
present the final result. We do not, however, present the argument by which the result is obtained.
We begin by reviewing some necessary ideas from electrostatics. When point charges q1 and q2 are embedded in a continuous
medium, the Coulomb’s law force exerted on q by q is
1 2
⇀ ^21
q1 q2 r
F 21 =
2
4π ε0 Dr
12
where ε is a constant called the permittivity of free space, and D is a constant called the dielectric coefficient of the continuous
0
medium. r^ is a unit vector in the direction from the location of q to the location of q . When q and q have the same sign, the
21 2 1 1 2
⇀ ⇀
force is positive and acts to increase the separation between the charges. The force exerted on q by q is 2 1 F 12 = − F 21 ; the net
force on the system of charges is
⇀ ⇀ ⇀
F net = F 12 + F 21 = 0.
When the force is expressed in newtons, the point charges are expressed in coulombs, and distance is expressed in meters,
ε = 8.854 × 10
0
−12
C N
2 −1
m . The dielectric coefficient is a dimensionless quantity whose value in a vacuum is unity. In liquid
−2
water at 25 ºC, D = 78.4 We are interested in the interactions between ions whose charges are multiples of the fundamental unit of
charge, e . We designate the charge on a proton and an electron as e and −e , respectively, where e = 1.602 × 10 C . We express
−19
the charge on a cation, say A , as z e , and that on an anion, say B , as z e , where z = +m > 0 and z = −n < 0 .
m+
A
n−
B A B
The Debye-Hückel theory models the environment around a particular central ion—the ion whose activity coefficient we calculate.
We assume that the interactions between the central ion and all other ions result exclusively from Coulomb’s law forces. We
assume that the central ion is a hard sphere whose charge, q , is located at the center of the sphere. We let the radius of this sphere
C
be a . Focusing on the central ion makes it possible to simplify the mathematics by fixing the origin of the coordinate system at
C
the center of the central ion; as the central ion moves through the solution, the coordinate system moves with it. The theory
develops a relationship between the activity coefficient of the central ion and the electrical work that is done when the central ion is
brought into the solution from an infinite distance—where its potential energy is taken to be zero.
The theory models the interactions of the central ion with the other ions in the solution by supposing that, for every type of ion, k ,
in the solution, there is a spherically symmetric function, ρ (r) , which specifies the concentration of k -type ions at the location
k
specified by r, for r ≥ a . That is, we replace our model of mobile point-charge ions with a model in which charge is distributed
C
continuously. The physical picture corresponding to this assumption is that the central ion remains discrete while all of the other
ions are “ground up” into tiny charged bits that are spread smoothly—but not uniformly—throughout the solution that surrounds
the central ion. The introduction of ρ (r) changes our model from one involving point-charge neighbor ions—whose effects would
k
have to be obtained by summing an impracticably large number of terms and whose locations are not well defined anyway—to one
involving a mathematically continuous function. From this perspective, we adopt, for the sake of a quantitative mathematical
treatment, a physical model that violates the atomic description of everything except the central ion.
It is useful to have a name for the collection of charged species around the central ion; we call it the ionic atmosphere. The ionic
atmosphere occupies a microscopic region around the central ion in which ionic concentrations depart from their macroscopic-
solution values. The magnitudes of these departures depend on the sign and magnitude of the charge on the central ion.
The essence of the Debye-Hückel model is that the charge of the central ion gives rise to the ionic atmosphere. To appreciate why
this is so, we can imagine introducing an uncharged moiety, otherwise identical to the central ion, into the solution. In such a
process, no ionic atmosphere would form. As far as long-range Coulombic forces are concerned, no work would be done.
When we imagine introducing the charged central ion into the solution in this way, Coulombic forces lead to the creation of the
ionic atmosphere. Since formation of the ionic atmosphere entails the separation of charge, albeit on a microscopic scale, this
process involves electrical work. Alternatively, we can say that electrical work is done when a charged ion is introduced into a salt
solution and that this work is expended on the creation of the ionic atmosphere.
In the Debye-Hückel model, this electrical work is the energy change associated with the process of solvating the ion. Since the
reversible, non-pressure–volume work done in a constant-temperature, constant-pressure process is also the Gibbs free energy
change for that process, the work of forming the ionic atmosphere is the same thing as the Gibbs free energy change for introducing
the ion into the solution.
The Debye-Hückel theory makes these ideas quantitative by finding the work done in creating the ionic atmosphere. To do this, it
proves to be useful to define a quantity that we call the ionic strength of the solution. By definition, the ionic strength is
n
2
I = ∑ z m /2
k ––k
k=1
where the sum is over all of the ions present in the solution. The factor of 1/2 is essentially arbitrary. We introduce it in order to
make the ionic strength of a 1:1 electrolyte equal to its molality. (z is dimensionless.)
k
For the hypothetical one-molal standard state that we consider in §6, the activity coefficient for solute C , γ , is related to the C
chemical potential of the real substance, μ (P , ––

Cm ), and that of a hypothetical ideal solute C at the same concentration,
C
μC (Hyp solute, P , m ) , by
C
––
μC (P , m ) − μC (Hyp solute, P , m )
ln γC = ––C ––C
RT
The Debye-Hückel model equates this chemical-potential difference to the electrical work that accompanies the introduction of the
central ion into a solution whose ionic strength is I. The final result is
2 2 ¯
¯¯¯
¯
−z e κN
C
ln γC =
8π ε0 D (1 + κ aC )
(While it is not obvious from our discussion, the parameter,

1/2
2¯
¯¯¯
¯
2 e N dw I
κ =( )
ε0 DkT
characterizes the ionic atmosphere around the central ion. The quantity dw is the density of the pure solvent, which is usually
water.)
For sufficiently dilute solutions, 1 + κa ≈ 1 . (See problem 14.) Introducing this approximation, substituting for κ , and dividing
C
by 2.303 to convert to base-ten logarithms, we obtain the Debye-Hückel limiting law in the form in which it is usually presented:
2 1/2
log10 γC = −Aγ z I
C
where
2
1/2 ¯
¯¯¯
¯ 3/2
2
(2 dw ) N e
Aγ = ( )
2.303 (8π) ε0 DRT
For aqueous solutions at 25 C, A γ = 0.510 .

The Debye-Hückel model finds the activity of an individual ion. In §18, we note that the activity of an individual ion cannot be
determined experimentally. We introduce the mean activity coefficient, γ , for a strong electrolyte to as a way to express the
±
departure of a salt solution from ideal-solution behavior. Adopting the hypothetical one-molal ideal-solution state as the standard
state for the salt, A B , we develop conventions that express the Gibbs free energy of a real salt solution and find
p q
p q 1/(p+q)
γ± = (γ γ ) .
A B
Using the Debye-Hückel limiting law values for the individual-ion activity coefficients, we find
2 1/2 2 1/2 2 2
p log10 γA + q log10 γB −p Aγ z I − q Aγ z I pz + qz
A B A B 1/2 1/2
log10 γ± = = = −( ) Aγ I = −Aγ zA zB I
p +q p +q p +q
where we use the identity

2 2
pz + qz
A B
= −zA zB
p +q
(See problem 16.12.)
This page titled 16.18: Activities of Electrolytes - The Debye-Hückel Theory is shared under a CC BY-SA 4.0 license and was authored, remixed,
In this chapter, we introduce several ways to measure the activities and chemical potentials of solutes. In Sections 16.1–16.6 we
consider the determination of the activities and chemical potentials of solutes with measurable vapor pressures. To do so, we use
the ideal behavior expressed by Raoult’s Law and Henry’s Law. In Section 16.15 we discuss the determination of solvent activity
coefficients from measurements of the decrease in the freezing point of the solvent. In Section 16.7 we discuss the mathematical
analysis by which we can obtain solute activity coefficients from measured solvent activity coefficients. Electrical potential
measurements on electrochemical cells are an important source of thermodynamic data. In Chapter 17, we consider the use of
electrochemical cells to measure the Gibbs free energy difference between two systems that contain the same substances but at
different concentrations.
We define the activity of substance A in a particular system such that
¯
¯¯¯ o ~
GA = μA = μ̃ + RT ln aA .
A
In the activity standard state the chemical potential is μ̃ and the activity is unity, a
o
A
~
= 1 . It is often convenient to choose the
A
standard state of the solute to be the hypothetical one-molal solution, particularly for relatively dilute solutions. In the hypothetical
one-molal standard state, the solute molality is unity and the environment of a solute molecule is the same as its environment at
infinite dilution. The solute activity is a function of its molality, a
~
(m ) . We let the molality of the actual solution of unit activity
A A
––
be m . That is, we let a (m ) = 1 ; consequently, we have μ (m ) = μ̃ even though the actual solution whose molality is
o
A
~
A
o
A A
o
A
o
A
–– –– ––
m
––
o
A
is not the standard state. To relate the solute activity and chemical potential in the actual solution to the solute molality, we
must find the activity coefficient, γ , as a function of the solute molality,
A
γA = γA (m )
––A
Then
~
aA (mA ) = mA γA (mA )
–– –– ––
and
~ o o o
aA (m ) = m γA (m ) = 1
––A ––A ––A
To introduce some basic approaches to the determination of activity coefficients, let us assume for the moment that we can measure
the actual chemical potential, μ , in a series of solutions where m varies. We have
A
––A
o ~ o
μA = μ̃ + RT ln aA = μ̃ + RT ln m + RT ln γA
A A ––A
We know ––m
A
from the preparation of the system—or by analysis. If we also μ̃ , we can calculate γ (–– m
o
A A A
) from our experimental
values of μ . If we don’t know μ̃ , we need to find it before we can proceed. To find it, we recall that
A
o
lim RT ln γA = 0
m →0
––A
Then
o o
lim (μA − RT ln mA ) = lim (μ̃A + RT ln γA ) = μ̃A
m →0 –– m →0
––A ––A
and a plot of (μ A − RT ln mA )
––
versus ––
m will intersect the line ––
A
m A
=0 at μ̃ .
o
A
Now, in fact, we can measure only Gibbs free energy differences. In the best of circumstances what we can measure is the
difference between the chemical potential of A at two different concentrations. If we choose a reference molality, m , the ref
A
––
chemical potential difference Δμ (–– m ) =μ A (m ) − μ
A
––
(m
––
) is a measurable quantity. A series of such results can be
A A A
ref
A
displayed as a plot of Δμ (–– m ) versus m —or any other function of m

A A –– A ––
that proves to suit our purposes. The reverence
A
molality, ––
m , can be chosen for experimental convenience.
ref
If our theoretical structure is valid, the results are represented by the equations
~
aA (m )
A ~
ΔμA (m ) = μA (m ) − μA (m ) = RT ln ––ref
= RT ln m + RT ln γA (m ) − RT ln aA (m
ref
)
––A ––A ––A ~ ref ––A ––A ––A
aA (m )
––A
When ––
m
A
= m
––A
o
, we have
~ o
aA (m )
o o ref 0 ref ––A ~ ref
ΔμA (m ) = μA (m ) − μA (m ) = μ̃ − μA (m ) = RT ln = −RT ln aA (m )
A ~
––A ––A ––A ––A ref ––A
aA (m )
––A
and
o
ΔμA (mA ) − ΔμA (m ) = RT ln mA + RT ln γA (mA )
–– ––A –– ––
so that
o
RT ln γA (mA ) = −ΔμA (m ) + ΔμA (mA ) − RT ln mA
–– ––A –– ––
Since lim m →0 γA (m ) = 1 , we have

––A ––A
o
0 = lim RT ln γA (m ) = −ΔμA (m ) + lim [ΔμA (m ) − RT ln m ]
m →0 ––A ––A m →0 ––A ––A
––A ––A
Letting
β (m ) = ΔμA (m ) − RT ln m
––A ––A ––A
and
o
β = lim [ΔμA (m ) − RT ln m ]
m →0 ––A ––A
––A
we have
o o
ΔμA (m ) = β
––A
Then
o o
RT ln γA (m ) = −ΔμA (m ) + ΔμA (m ) − RT ln m = −β + ΔμA (m ) − RT ln m
––A ––A ––A ––A ––A ––A
so that we know both the activity coefficient, γ = γ (––

m ) , and the activity, a
A
~
(m ) = m γ
–– ––
A (m ) , of A as a function of its
–– A A A A A A
molality. Consequently, we know the value of Δμ (–– m ) − Δμ (m ) as a function of molality. Since this difference vanishes
––
A A A
o
A
when m = m , we can find m from our experimental data. Finally, the activity equation becomes
o 0
––A ––A ––A
~ o
RT ln aA (m ) = ΔμA (m ) − β
––A ––A
This procedure yields the activity of A as a function of the solute molality. We obtain this function from measurements of
) . These measurements do not yield a value for μ (m ) ; what we obtain from our analysis is
ref
Δμ (m ) = μ
A
(m ) − μ (m A A A
A A A A
–– –– –– ––
an alternative expression,
~
aA (m )
RT ln ––A
~ ref
aA (m )
––A
ref
for the chemical potential difference, μ A (mA ) − μA ( m
–– ––A
) between two states of the same substance. μ A (mA )
––
is the difference
between the chemical potential of solute A at ––
m and the chemical potential of its constituent elements in their standard states at
A
the same temperature. To find this difference is a separate experimental undertaking. If, however, we can find μ (–– m ) for some A
∗
A
m , our activity equation yields μ̃ as

∗ o
A
––A
0 ∗ ~ ∗
μ̃ = μA (m ) − RT ln aA (m )
A
––A ––A
This analysis of the Δμ (––

m ) data assumes that we can find β
A A
o
. To find an accurate
= limm
––A
→0 [ΔμA (mA ) − RT ln mA ]
–– ––
value for β , it is important to collect data for
o
at the lowest possible values for
ΔμA (mA )
––
m $ m . I nevitably, however, theexperimentalerrorin\(Δμ
A A (m ) increases as m decreases. Our theory requires that
A A A
–– –– ––
β (m ) = β + f (m ) , where lim f (m ) = 0 , so that the graph of β (m ) versus f (m ) has an intercept at β .
o o
A A m →0 A A A
–– –– –– ––A –– ––
Accurate extrapolation of the data to the intercept at ––m = 0 is greatly facilitated if we can choose f (m ) so that the graph is
A –– A
linear. In practice, the increased experimental error in β (––

m ) at the lowest values of m
––
causes the uncertainty in the extrapolated
A A
value of β for a given choice of f (––

o
m ) to be similar to the range of β
A
values estimated using different functions. For some p in
o
the range 0.5 < p < 2 , letting f (––
m
A
) = m
––A
P
often provides a fit that is as satisfactorily linear as the experimental uncertainty can
justify.
This procedure yields the activity of A as a function of the solute molality. We obtain this function from measurements of
) . These measurements do not yield a value for μ (m ) ; what we obtain from our analysis is
ref
Δμ (m ) = μ
A A A (m ) − μ
A
)mA A A
–– –– –– A ––
an alternative expression,
~
aA (mA )
––
RT ln
~ ref
aA (m )
––A
for the chemical potential difference, μ (––

m )−μA )m
–– A
) between two states of the same substance. μ
A
ref
A
(m ) is the difference
––
A A
between the chemical potential of solute A at m and the chemical potential of its constituent elements in their standard states at
A
––
the same temperature. To find this difference is a separate experimental undertaking. If, however, we can find μ (m ) for some A
∗
A
––
m , our activity equation yields μ as
∗ ~ o
––A A
~o ∗ ~ ∗
μ = μA (m ) − RT ln aA (m )
A ––A ––A
Finally, let us contrast this analysis to the analysis of chemical equilibrium that we discuss briefly in Chapter 15. In the present
analysis, we use an extrapolation to infinite dilution to derive activity values from the difference between the chemical potentials of
the same substance at different concentrations. In the chemical equilibrium analysis for aA + bB ⇌ cD + dD , we have
~c ~d c d c d
aC aD m m γ γ̃ D
o o ––C ––D C
Δr μ = Δr μ̃ + RT ln = Δr μ̃ + RT ln + RT ln
~a ~b a b a b
aA aB m m γ γ
––A ––B A B
When the system is at equilibrium, we have Δ rμ =0 . Since lim m →0 γi = 1 , we have, in the limit that all of the concentrations
––i
go to zero in an equilibrium system,

c d c d
γ γ m m
C D o ––C ––D
0 = lim RT ln = Δr μ̃ + lim RT ln
m a b m a b
––i→ 0 γ γ ––i→ 0 m m
B
A ––A ––B
Letting
c d
m m
––C ––D
Kc =
a b
m m
––A ––B
We have
o
Δr μ̃ = −RT lim Kc
m →0
––i
Since Δ μ = 0 whenever the system is at equilibrium, measurement of

r Kc for any equilibrium state of the reaction yields the
corresponding ratio of activity coefficients:
c d
γ γ
C D o
RT ln = −Δr μ̃ − RT ln Kc
a b
γ γ
A B
This page titled 16.19: Finding Solute Activity Using the Hypothetical One-molal Standard State is shared under a CC BY-SA 4.0 license and was
16.20: Problems
1. At 100 C, the enthalpy of vaporization of water is 40.657 kJ mol . Calculate the boiling-point elevation constant for water
−1
when the solute concentration is expressed in molality units.

2. At 0 C, the enthalpy of fusion of water is 6.009 kJ mol . Calculate the freezing-point depression constant for water when the
−1
solute concentration is expressed in molality units.

3. A solution is prepared by dissolving 20.0 g of ethylene glycol (1,2-ethanediol) in 1 kg of water. Estimate the boiling point and
the freezing point of this solution.
4. A biopolymer has a molecular weight of 250,000 dalton. At 300 K, estimate the osmotic pressure of a solution that contains 1 g
of this substance in 10 mL of water.
5. Cyclohexanol melts at 25.46 C; the enthalpy of fusion is 1.76 kJ mol . Estimate the freezing-point depression constant when
−1
the solute concentration is expressed as a mole fraction and when it is expressed in molality units. A solution is prepared by mixing
1 g of ethylene glycol with 50 g of liquid cyclohexanol. How much is the freezing point of this solution depressed relative to the
freezing point of pure cyclohexanol?
6. Freezing-point depression data for numerous solutes in aqueous solution are reproduced below. Calculate the freezing-point
3
depression, −ΔT / m, for each of these solutes. Compare these values to the freezing-point depression constant that you
fp
––
calculated in problem 2. Explain any differences.
Solute molality −ΔTf p , K
Acetone 0.087 0.16
ethanol 0.109 0.20
ethylene glycol 0.081 0.15
ammonia 0.295 0.55
glycerol 0.110 0.18
lithium chloride 0.119 0.42
nitric acid 0.080 0.28
potassium bromide 0.042 0.15
barium chloride 0.024 0.12
7. In a binary solution of solute A in solvent B , the mole fractions in the pure solvent are y = 0 and y = 1 . We let the pure
A B
solvent be the solvent standard state; when y = 0 , y = a

A
~
B= 1 , and ln a
B
~
= 0 . What happens to the value of ln a
B
~
as B
→ 1 ? Sketch the graph of (1 − y )/ y versus ln a . For $0<y^*_a><1$>, shade the area on this graph that represents the
~
yA A A B
integral
∗
y
A
1 − yA
∫ ( ) dln γ
A
y yA
A
Is this area greater or less than zero?

8. In a binary solution of solute A in solvent B , the activity coefficient of the solvent can be modeled by the equation
p
ln γ B
(y ) = c y
A , where the constants c and p are found by using least squares to fit experimental data to the equation. Find an
A
equation for ln γ (y ) . For c = 8.4 and p = 2.12, plot ln γ (y ) and ln γ (y ) versus y .

A A B A A A A
9. A series of solutions contains a non-volatile solute, A , dissolved in a solvent, B . At a fixed temperature, the vapor pressure of
solvent B is measured for these solutions and for pure B (y = 0 ). At low solute concentrations, the vapor pressure varies with the
A
solute mole fraction according to P =P (1 − yA ) exp (−α y

β
A
) .
(a) If the pure solvent at one bar is taken as the standard state for liquid A , and gaseous B behaves as an ideal gas, now does the
activity of solvent B vary with y ? A
(b) How does ln γ B
vary with y ? A
(c) Find ln γ A
(yA ) .
10. In Section 14.14, we find for liquid solvent B ,
¯
¯¯¯
¯
¯
¯¯¯
o
¯
¯¯¯
¯
ref ∂Δmix H
o
LB = H −HB = lim (− )
B
T →0 ∂nB
P ,T ,nA
¯
¯¯¯
¯
Since H is the molar enthalpy of pure liquid B , we have
B
¯
¯¯¯
¯
⎛ ∂H ⎞
B
= CP (B, liquid, T )
⎝ ∂T ⎠
P
In Section 16.15, we set

ref
¯
¯¯¯
¯
⎛ ∂H ⎞
B
= CP (B, liquid, T )
⎝ ∂T ⎠
P
Show that this is equivalent to the condition

o
¯
¯¯¯
∂LB
( ) ≪ CP (B, liquid, T )
∂T
P
11. If pz A
= −q zB , prove that
2 2
pz + qz
A B
= −zA zB
p +q
12. At temperatures of 5 C, 25 C, and 45 C, evaluate Debye-Hückel parameter κ for aqueous sodium chloride solutions at
concentrations of 10 m , 10 m , and 10 m .
−3 −2 −1
–– –– ––
13. Introducing the approximation 1 + κa ≈ 1 produces the Debye-Hückel limiting law, which is strictly applicable only in the
c
limiting case of an infinitely dilute solution. Introducing the approximation avoids the problem of choosing an appropriate value for
a . If a = 0.2 nm , calculate 1 + κa for aqueous solutions in which the ionic strength, I , is 10 ––
m , 10 m , and 10 m.
−3 −2 −1
c c c
–– ––
What does the result suggest about the ionic-strength range over which the limiting law is a good approximation?
14. The solubility product for barium sulfate, K sp
~ ~
= aBa2+ a
SO4
2− , is 1.08 × 10 −10
. Estimate the solubility of barium sulfate in pure
water and in 10 m potassium perchlorate.
−2
––
15. The enthalpy of vaporization of n-butane at its normal boiling point, 272.65 K, is 22.44 kJ mol . In the temperature range \
3 −1
(273.15, the solubility of n-butane in water is given by

3
100 B T
ln yA = A+ + C ln( )
T 100
where A = −102.029 , B = 146.040 K , and C = 38.7599. From the result we develop in Section 16.14, calculate
−1
Δ vapH
¯
¯¯¯
¯
for n-butane at is normal boiling point. (Note that the normal boiling temperature is slightly below the temperature
A,solution
range to which the equation for ln y is valid.) Comment.

A
16. The enthalpy of vaporization of molecular oxygen at its normal boiling point, 90.02 K, is 6.82 kJ mol . In the temperature
3 −1
range $273.15<$>348.15, the solubility of oxygen in water is given by

3
100 B T
ln yA = A+ + C ln( )
T 100
where A = −66.7354 , B = 87.4755 K , and C = 24.4526. From the result we develop in Section 16.14, calculate
−1
¯
¯¯¯
¯
Δvap H A,solution for oxygen at 273.25 K and at its normal boiling point, 90.02 K. Comment.
Notes
1
Raoult’s law and ideal solutions can be defined using fugacities in place of partial pressures. The result is more general but—for
those whose intuition has not yet embraced fugacity—less transparent.
2
For a discussion of the concentration range in which the Debeye-Huckel model is valid and of various supplemental models that
allow for the effects of forces that are specific to the chemical characteristics of the interacting ions, see Lewis and Randall, Pitzer
and Brewer, Thermodynamics, 2 Edition, McGraw Hill Book Company, New York, 1961, Chapter 23.
nd
3
Data from CRC Handbook of Chemistry and Physics, 79 th
Edition, David R. Lide, Ed., CRC Press, 1998-1999.
CHAPTER OVERVIEW
17: Electrochemistry
17.19: Problems
Thumbnail: Schematic of Zn-Cu galvanic cell. (CC BY-SA 3.0; Ohiostandard).
This page titled 17: Electrochemistry is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
1
We find it useful to classify reactions according to the type of change that the reagents undergo. Many classification schemes exist,
often overlapping one another. The three most commonly used classifications are acid–base reactions, substitution reactions, and
oxidation–reduction reactions.
Acids and bases can be defined in several ways, the most common being the Brønsted-Lowry definition, in which acids are proton
donors and bases are proton acceptors. The Brønsted-Lowry definition is particularly useful for reactions that occur in aqueous
solutions. A prototypical example is the reaction of acetic acid with hydroxide ion to produce the acetate ion and water.
− −
C H3 C O2 H + OH → C H3 C O + H2 O
2
Here acetic acid is the proton donor, hydroxide ion is the proton acceptor. The products are also an acid and a base, since water is a
proton donor and acetate ion is a proton acceptor.
When we talk about substitution reactions, we focus on a particular substituent in a chemical compound. The original compound is
often called the substrate. In a substitution reaction, the original substituent is replaced by a different chemical moiety. A
prototypical example is the displacement of one substituent on a tetrahedral carbon atom by a more nucleophilic group, as in the
reaction of methoxide ion with methyl iodide to give dimethyl ether and iodide ion.
− −
C H3 I + C H3 O → C H3 OC H3 + I
We could view a Brønsted-Lowry acid-base reaction as a substitution reaction in which one group (the acetate ion in the example
above) originally bonded to a proton is replaced by another (hydroxide ion). Whether we use one classification scheme or another
to describe a particular reaction depends on which is better suited to our immediate purpose.
In acid-base reactions and substitution reactions, we focus on the transfer of a chemical moiety from one chemical environment to
another. In a large and important class of reactions we find it useful to focus on the transfer of one or more electrons from one
chemical moiety to another. For example, copper metal readily reduces aqueous silver ion. If we place a piece of clean copper wire
in an aqueous silver nitrate solution, reaction occurs according to the equation
+ 0 0 2+
2Ag + Cu → 2Ag + Cu
We have no trouble viewing this reaction as the transfer of two electrons from the copper atom to the silver ions. In consequence, a
cupric ion, formed at the copper surface, is released into the solution. Two atoms of metallic silver are deposited at the copper
surface. Reactions in which electrons are transferred from one chemical moiety to another are called oxidation–reduction
reactions, or redox reactions, for short.
We define oxidation as the loss of electrons by a chemical moiety. Reduction is the gain of electrons by a chemical moiety. Since a
moiety can give up electrons only if they have some place to go, oxidation and reduction are companion processes. Whenever one
moiety is oxidized, another is reduced. In the reduction of silver ion by copper metal, it is easy to see that silver ion is gaining
electrons and copper metal is losing them. In other reactions, it is not always so easy to see which moieties are gaining and losing
electrons, or even that electron transfer is actually involved. As an adjunct to our ideas about oxidation and reduction, we develop a
scheme for formally assigning electrons to the atoms in a molecule or ion. This is called the oxidation state formalism and
comprises a series of rules for assigning a number, which we call the oxidation state (or oxidation number), to every atom in the
molecule. When we adopt this scheme, the redox character of a reaction is determined by which atoms increase their oxidation state
and which decrease their oxidation state as a consequence of the reaction. Those whose oxidation state increases lose electrons and
are oxidized, while those whose oxidation state decreases gain electrons and are reduced.
A process of electron loss is called an oxidation because reactions with elemental oxygen are viewed as prototypical examples of
such processes. Since many observations are correlated by supposing that oxygen atoms in compounds are characteristically more
electron-rich than the atoms in elemental oxygen, it is useful to regard a reaction of a substance with oxygen as a reaction in which
the atoms of the substance surrender electrons to oxygen atoms. It is then a straightforward generalization to say that a substance is
oxidized whenever it loses electrons, whether oxygen atoms or some other chemical moiety takes up those electrons. So, for
example, the reaction of sodium metal with oxygen in a dry environment produces sodium oxide, N a O, in which the sodium is
2
usefully viewed as carrying a positive charge. (The oxidation state of sodium is 1+; the oxidation state of oxygen is 2–.)
The conversion of a metal oxide to the corresponding metal is described as reducing the oxide. Since converting a metal oxide to
the metal reverses the change that occurs when we oxidize it, generalization of this idea leads us to apply the term reduction to any
process in which a chemical moiety gains electrons. It is a fortunate coincidence that a reduction process is one in which the
oxidation number of an atom becomes smaller (more negative) and is therefore reduced, in the sense of being decreased.
Another feature of oxidation–reduction reactions, and one that relates to the utility of viewing these reactions in terms of electron
gain and loss, emerges when we observe the reaction of aqueous silver ions with copper metal closely. As the reaction proceeds, the
aqueous solution becomes blue as cupric ions accumulate. Long needle-like crystals of silver metal grow out from the copper
surface. The simplest mechanism that we can imagine for the growth of well-formed silver crystals is that silver ions from the
solution plate out on the surface of the growing silver crystal, accepting an electron from the metallic crystal as they do so. The
silver metal acquires this electron from the copper metal, with which it is in contact, but at a large (on an atomic scale) distance
from the site at which the new atom of silver is deposited. Evidently the processes of electron loss and gain that characterize an
overall reaction can occur at different locations, if there is a suitable process for moving the electron from one location to the other.
This page titled 17.1: Oxidation-reduction Reactions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
We can extend the idea of carrying out the electron loss and electron gain steps in different physical locations. Suppose that the
only aqueous species in contact with the silver metal are silver ions and nitrate ions; the silver metal is also in contact with a length
of copper wire, whose other end dips into a separate reservoir containing an aqueous solution of sodium nitrate. This arrangement
is sketched in Figure 1. When we create this arrangement, nothing happens. We do not see any visible change in the silver metal,
and the water contacting the copper wire never turns blue. On the one hand, this result does not surprise us. We are accustomed to
the idea that reactants must be able to contact one another in order for reaction to occur.
Figure 1. Electron transport between compartments is possible, but ion transport is not.
On the other hand, the original experiment really does show that silver ions can accept electrons in one location while copper atoms
give them up in another, so long as we provide a metal bridge on which the electrons can move between the two locations. Why
should this not continue to happen in the new experimental arrangement? In fact, it does. It is just that the reaction occurs to only a
very small extent before stopping altogether. The reason is easy to appreciate. After a very small number of silver ions are reduced,
the silver nitrate solution contains more nitrate ions than silver ions; the solution as a whole has a negative charge. In the other
reservoir, a small number of cupric ions dissolve, but there is no increase in the number of counter ions, so this solution acquires a
positive charge. These net charges polarize the metal that connects them; the metal has an excess of positive charge at the copper-
solution end and an excess of negative charge at the silver-solution end. This polarization opposes the motion of a negatively
charged electron from the copper-solution end toward the silver-solution end. When the polarization becomes sufficiently great,
electron flow ceases and no further reaction can occur.
By this analysis, the anions that the cupric solution needs in order to achieve electroneutrality are present in the silver-ion solution.
The reaction stops because the anions have no way to get from one solution to the other. Evidently, the way to make the reaction
proceed is to modify the two-reservoir experiment so that nitrate ions can move from the silver-solution reservoir to the copper-
solution reservoir. Alternatively, we could introduce a modification that allows copper ions to move in the opposite direction or one
that allows both kinds of movement. We can achieve the latter by connecting the two solutions with a tube containing sodium
nitrate solution, as diagrammed in Figure 2. Now, nitrate ions can move between the reservoirs and maintain electroneutrality in
both of them. However, silver ions can also move between the reservoirs. When we do this experiment, we observe that electrons
do flow through the wire, indicating that silver-ion reduction and copper-atom oxidation are occurring at the separated sites.
However, after a short time, the solutions mix; silver ions migrate through the aqueous medium and react directly with the copper
metal. Because the mixing is poorly controllable, the reproducibility of this experiment is poor.
Figure 2. Electron transport and fluid flow between compartments are possible.
Evidently, we need a way to permit the exchange of ions between the two reservoirs that does not permit the wholesale transfer of
reactive species. One device that accomplishes this is called a salt bridge. The requirement we face is that ions should be able to
migrate from reservoir to reservoir so as to maintain electroneutrality. However, we do not want ions that participate in electrode
reactions to migrate. A salt bridge is simply a salt solution that we use to connect the two reservoirs. To avoid introducing
unwanted ions into the reservoir solutions, we prepare the salt-bridge solution using a salt whose ions are not readily oxidized or
reduced. Alkali metal salts with nitrate, perchlorate, or halide anions are often used. To avoid mixing the reservoir solutions with
the salt bridge solution, we plug each end of the salt bridge with a porous material that permits diffusion of ions but inhibits bulk
movement of solution in or out of the bridge. The inhibition of bulk movement can be made much more effective by filling the
bridge with a gel, so that the solution is unable to undergo bulk motion in any part of the bridge.
With a salt bridge in place, inert ions can move from one reservoir to the other to maintain electroneutrality. Under these
conditions, we see an electrical current through the external circuit and a compensating diffusion of ions through the salt bridge.
The salt bridge completes the circuit. Transport of electrons from one electrode to the other carries charge in one direction; motion
of ionic species through the salt bridge carries negative charge through the solution in the opposite direction. This compensating
ionic motion has anions moving opposite to the electron motion and cations moving in the same direction as the electrons.
Figure 3. Electron transport and ion transport between compartments are possible.
We have just described one kind of electrochemical cell. As diagrammed in Figure 3, it has four principal features: two reservoirs
within which reactions can occur, a wire through which electrons can pass from one reservoir to the other, and a salt bridge through
which ionic species can pass. Many similar electrochemical cells can be constructed. The reservoirs can contain a wide variety of
reagents. Because each reservoir must be able to exchange electrons with the connecting wire, each must contain an electrically
conducting solid that serves as a terminal and a current collector, and often participates in the chemical change as a reactant or as a
catalyst. The combination of reagents and current collector is called a half-cell. The current collector itself is called an electrode,
although this term is often applied to the whole half-cell as well. In this case, the wire is the external circuit. In applications of
chemical interest, the external circuit typically contains devices to measure the electrochemical cell’s properties as a circuit
element.
If we view this electrochemical cell as a device for producing an electrical current, we see that it has a number of practical
limitations. Two of the most important relate to the performance of the salt bridge. Whenever electrons move through the external
circuit, the salt bridge must accept a charge-compensating number of ions from one reservoir and release the same quantity of ionic
charge to the other reservoir. We construct the salt bridge so that ions can pass into and out of it only by diffusion. Consequently,
the rate at which ions can diffuse through the salt bridge limits the rate at which electrons can flow through the external circuit.
Since diffusion is a slow process over the macroscopic dimensions of the bridge, the cell can pass only a small current. From an
electrical perspective, slow diffusion of ions through the salt bridge causes a surplus of positively charged species to develop at one
end of the salt bridge and a surplus of negatively charged species to develop at the other. This charge imbalance means that there is
a potential gradient across the salt bridge, whose effect is to oppose the flow of further current.
The second limitation of the cell attributable to the properties of the salt bridge is that the amount of current the cell can produce
before its performance characteristics change dramatically is limited by the amount of inert salt in the bridge. After a relatively
small charge passes through the cell, migration of reactive species from one reservoir to the other becomes significant. Effective
electrochemical power sources must use other methods to separate reactants and products while allowing for the transport of ions
between half-cells.
Despite these limitations, such electrochemical cells are very effective tools for the study of the thermodynamics of electrochemical
reactions. The principle interaction between electrochemistry and thermodynamics revolves around the relationship between the
free energy change for the reaction and the properties of the electrochemical cell viewed as a circuit element. In Section 17.14, we
see that the Gibbs free energy change for the chemical reaction is proportional to the electrical potential that develops across the
terminals of the corresponding electrochemical cell.
In experiments, we find that the electrical potential difference across a cell depends on the amount of current that is being drawn
from the cell. Because the movement of ions and other substances within the cell is slow compared to the rate at which a wire can
transfer electrons from one terminal to another, potential differences that develop within an operating cell decrease the electrical
potential across the terminals. Only when the current being drawn from the cell is zero does the electrical potential correspond
precisely to the Gibbs free energy change of the chemical reaction occurring in the cell. This should not surprise us. The
experimental measurement of any entropy-dependent thermodynamic function must be made on a system that is undergoing
reversible change. A reversible change in an electrochemical cell is a change in which the current flow is zero.
Measuring the electrical potential at zero current is experimentally straightforward, at least in principle. We connect the cell to
some reference device that provides a known and variable electrical potential. The connection is made such that the electrical
potential from the reference device opposes the potential from the electrochemical cell; that is, we connect the positive terminal of
the reference device to the positive terminal of the cell, and the negative terminal of the device to the negative terminal of the cell.
(See Section 17.7.) We then vary the potential of the reference device until current flow in the circuit stops. When this occurs the
potential drop being supplied by the reference device must be precisely equal to the potential drop across the electrochemical cell,
which is the datum we want.
In practice, the reference device is another “standard” electrochemical cell, whose potential drop is defined to have a particular
value at specified conditions. Modern electronics make it possible to do the actual measurements with great sophistication. The
necessary measurements can also be done with very basic equipment. The principles remain the same. In the basic experiment, a
variable resistor is used to adjust the potential drop across the standard cell until it exactly matches that of the cell being studied.
When this potential is reached, current flow ceases. Current flow is monitored using a sensitive galvanometer. It is not necessary to
actually measure the current. Since we are interested in locating the potential drop at which the current flow is zero, it is sufficient
to find the potential drop at which the galvanometer detects no current. The accuracy of the potential measurement depends on the
stability of the standard cell potential, the accuracy of the variable resistor, and the sensitivity of the galvanometer.
This page titled 17.2: Electrochemical Cells is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
We introduce oxidation states to organize our thinking about oxidation–reduction reactions and electrochemical cells. When we
define oxidation states, we create a set of rules for allocating the electrons in a molecule or ion to the individual atoms that make it
up. The definition of oxidation states is therefore an accounting exercise. The definition of oxidation states predates our ability to
estimate electron densities through quantum mechanical calculations. As it turns out, however, the ideas that led to the oxidation
state formalism are directionally correct; atoms that have high positive oxidation states according to the formalism also have
relatively high positive charges by quantum mechanical calculation. In general, the absolute values of oxidation states are
substantially larger than the absolute values of the partial charges found by quantum-mechanical calculation; however, there is no
simple quantitative relationship between oxidation states and the actual distribution of electrons in real chemical moieties. It is a
serious mistake to think that our accounting system provides a quantitative description of actual electron densities.
It is a serious mistake to think that the Oxidation State system provides a quantitative
description of actual electron densities.
The rules for assigning oxidation states grow out of the primitive (and quantitatively incorrect) idea that oxygen atoms usually
acquire two electrons and hydrogen atoms usually lose one electron in forming chemical compounds and ionic moieties. The rest of
the rules derive from a need to recognize some exceptional cases and from applying the basic ideas to additional elements. The
rules of the oxidation state formalism are these:
For any element in any of its allotropic forms, the oxidation state of its atoms is zero.
In any of its compounds, the oxidation state of an oxygen atom is 2–, except in compounds that contain an oxygen–oxygen
bond, where the oxidation state of oxygen is 1–. The excepted compounds are named peroxides. Examples include sodium
peroxide, N a O , and hydrogen peroxide, H O .
2 2 2 2
In any of its compounds, the oxidation state of a hydrogen atom is 1+, except in compounds that contain a metal–hydrogen
bond, where the oxidation state of hydrogen is 1–. The excepted compounds are named hydrides. Examples include sodium
hydride, N aH , and calcium hydride, C aH . 2
In any of their compounds, the oxidation states of alkali metal atoms (lithium, sodium, potassium, rubidium, cesium, and
francium) are 1+. (There are exceptional cases, but we do not consider them.)
In any of their compounds, the oxidation states of halogen atoms (fluorine, chlorine, bromine, iodine, and astatine) are 1–,
except in compounds that contain a halogen–oxygen bond.
The oxidation states of any other atoms in a compound are chosen so as to make the sum of the oxidation states in the chemical
moiety equal to its charge. So, for a neutral molecule, the oxidation states sum to zero. For a monovalent anion, they sum to 1–,
etc.
This page titled 17.3: Defining Oxidation States is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
Having defined oxidation states, we can now redefine an oxidation–reduction reaction as one in which at least one element
undergoes a change of oxidation state. For example, in the reaction between permanganate ion and oxalate ion, the oxidation states
of manganese and carbon atoms change. In the reactants, the oxidation state of manganese is 7+; in the products, it is 2+. In the
reactants, the oxidation state of carbon is 3+; in the products, it is 4+.
7+ 3+ 2+ 4+
− 2− 2+
M nO + C2 O → Mn + C O2
4 4
These oxidation state changes determine the stoichiometry of the reaction. In terms of the oxidation state formalism, each
manganese atom gains five electrons and each carbon atom loses one electron. Thus the reaction must involve five times as many
carbon atoms as manganese atoms. Allowing for the presence of two carbon atoms in the oxalate ion, conservation of electrons
requires that the stoichiometric coefficients be
− − 2+
2 M nO + 5 C2 O → 2 Mn + 10 C O2
4 4
Written this way, two M nO moieties gain ten electrons, and five C O
−
4
moieties lose ten electrons. When we fix the
2
2−
4
coefficients of the redox reactants, we also fix the coefficients of the redox products. However, inspection shows that both charge
and the number of oxygen atoms are out of balance in this equation.
The reaction occurs in acidic aqueous solution. This means that enough water molecules must participate in the reaction to achieve
oxygen-atom balance. Adding eight water molecules to the product brings oxygen into balance. Now, however, charge and
hydrogen atoms
− − 2+
2 M nO + 5 C2 O → 2 Mn + 10 C O2 + 8 H2 O
4 4
do not balance. Since the solution is acidic, we can bring hydrogen into balance by adding sixteen protons to the reactants. When
we do so, we find that charge balances also.
− − + 2+
2 M nO + 5 C2 O + 16 H → 2 Mn + 10 C O2 + 8 H2 O
4 4
Evidently, this procedure achieves charge balance because the oxidation state formalism enables us to find the correct
stoichiometric ratio between oxidant and reductant.
We can formalize this thought process in a series of rules for balancing oxidation–reduction reactions. In doing this, we can derive
some advantage from splitting the overall chemical change into two parts, which we call half-reactions. It is certainly not necessary
to introduce half-reactions just to balance equations; the real advantage is that a half-reaction describes the chemical change in an
individual half-cell. The rules for balancing oxidation–reduction reactions using half-cell reactions are these:
1. Find the oxidation state of every atom in every reactant and every product.
2. Write skeletal equations showing:
1. the oxidizing agent → its reduced product
2. the reducing agent → its oxidized product
3. Balance the skeletal equations with respect to all elements other than oxygen and hydrogen.
4. Add electrons to each equation to balance those gained or lost by the atoms undergoing oxidation-state changes.
5. For a reaction occurring in acidic aqueous solution:
1. balance oxygen atoms by adding water to each equation.
2. balance hydrogen atoms by adding protons to each equation
6. For a reaction occurring in basic aqueous solution, balance oxygen and hydrogen atoms by adding water to one side of each
equation and hydroxide ion to the other.
7. The net effect of adding one water and one hydroxide is to increase by one the number of hydrogen atoms on the side to which
the water is added.
8. Adding two hydroxide ions to one side and a water molecule to the other increases by one the number of oxygen atoms on the
side to which hydroxide is added.
1. Multiply each half-reaction by a factor chosen to make each of the resulting half-reactions contain the same number of
electrons.
2. Add the half-reactions to get a balanced equation for the overall chemical change. The electrons cancel. Often, some of the
water molecules, hydrogen atoms, or hydroxide ions cancel also.
When we apply this method to the permanganate–oxalate reaction, we have
− + − 2+
2 M nO + 16 H + 10 e → 2 Mn + 8 H2 O
4
reduction half-reaction
− −
5 C2 O → 10 C O2 + 10 e
4
oxidation half-reaction
− − + 2+
2 M nO + 5 C2 O + 16 H → 2 Mn + 10 C O2 + 8 H2 O
4 4
balanced reaction
The half-reactions sum to the previously obtained result; the electrons cancel. For an example of a reaction in basic solution,
consider the disproportionation of chloride dioxide to chlorite and chlorate ions:
4+ 3+ 5+
− −
C lO2 → C lO + C lO
2 3
skeletal reaction
− −
C l O2 + e → C lO
2
− − −
C lO2 + 2OH → C lO + H2 O + e
3
− − −
2 C lO2 + 2OH → C lO + C lO + H2 O
2 3
balanced equation
This page titled 17.4: Balancing Oxidation-reduction Reactions is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Electrical potential is measured in volts. If a system comprising one coulomb of charge passes through a potential difference of one
volt, one joule of work is done on the system. The work done on the system is equal to the change in the energy of the system. For
Q coulombs passing through a potential difference of E volts, we have ΔE = w = QE . Whether this represents an increase or
elec
a decrease in the energy of the system depends on the sign of the charge and on the sign of the potential difference.
Electrical potential and gravitational potential are analogous. The energy change associated with moving a mass from one elevation
to another in the earth’s gravitational field is
ΔE = wgrav = mghf inal − mghinitial = m Φgrav
where Φ grav = g (hf inal − hinitial ) , which is the gravitational potential difference.
The role played by charge in the electrical case is played by mass in the gravitational case. The energies of these systems change
because charge or mass moves in response to the application of a force. In the electrical case, the force is the electrical force that
arises from the interaction between charges. In the gravitational case, the force is the gravitational force that arises from the
interaction between masses. A notable difference is that mass is always a positive quantity, whereas charge can be positive or
negative.
The distinguishing feature of an electrochemical cell is that there is an electrical potential difference between the two terminals. For
any given cell, the magnitude of the potential difference depends on the magnitude of the current that is flowing. (Making the
general problem even more challenging, we find that it depends also on the detailed history of the conditions under which electrical
current has been drawn from the cell.) Fortunately, if we keep the cell’s temperature constant and measure the potential at zero
current, the electrical potential is constant. Under these conditions, the cell’s characteristics are fixed, and potential measurements
give reproducible results. We want to understand the origin and magnitude of this potential difference. Experimentally, we find:
1. If we measure the zero-current electrical potential of the same cell at different temperatures, we find that this potential depends
on temperature.
2. If we prepare two cells with different chemical species, they exhibit different electrical potentials—except possibly for an
occasional coincidence.
3. If we prepare two cells with the same chemical species at different concentrations and measure their zero-current electrical
potentials at the same temperature, we find that they exhibit different potentials.
4. If we draw current from a given cell over a period of time, we find that there is a change in the relative amounts of the reagents
present in the cell. Overall, a chemical reaction occurs; some reagents are consumed, while others are produced.
We can summarize these experimental observations by saying that the central issue in electrochemistry is the interrelation of three
characteristics of an electrochemical cell: the electrical-potential difference between the terminals of the cell, the flow of electrons
in the external circuit, and the chemical changes inside the cell that accompany this electron flow.
This page titled 17.5: Electrical Potential is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
Suppose we use a wire to connect the terminals of the cell built from the silver–silver ion half-cell and the copper–cupric ion half-
cell. This wire then constitutes the external circuit, the path that the electrons follow as chemical change occurs within the cell.
When the external circuit is simply a low-resistance wire, the cell is short-circuited. The external circuit can be more complex. For
example, when we want to know the direction of electron flow, we incorporate a galvanometer.
If the reaction between silver ions and copper metal is to occur, electrons must pass through the external circuit from the copper
terminal to the silver terminal. An electron that is free to move in the presence of an electrical potential must move away from a
region of more negative electrical potential and toward a region of more positive electrical potential. Since the electron-flow is
away from the copper terminal and toward the silver terminal, the copper terminal must be electrically negative and the silver
terminal must be electrically positive. Evidently, if we know the chemical reaction that occurs in an electrical cell, we can
immediately deduce the direction of electron flow in the external circuit. Knowing the direction of electron flow in the external
circuit immediately tells us which is the negative and which the positive terminal of the cell.
The converse is also true. If we know which cell terminal is positive, we know that electrons in the external circuit flow toward this
terminal. Even if we know nothing about the composition of the cell, the fact that electrons are flowing toward a particular terminal
tells us that the reaction occurring in that half-cell is one in which a solution species, or the electrode material, takes up electrons.
That is to say, some chemical entity is reduced in a half-cell whose potential is positive. It can happen that we know the half-
reaction that occurs in a given half-cell, but that we do not know which direction the reaction goes. For example, if we replace the
silver–silver ion half cell with a similar cell containing an aqueous zinc nitrate solution and a zinc electrode, we are confident that
the half-cell reaction is either
0 2 + −
Zn → Zn +2 e
or
2 + − 0
Zn +2 e → Zn
When we determine experimentally that the copper electrode is electrically positive with respect to the zinc electrode, we know that
electrons are leaving the zinc electrode and flowing to the copper electrode. Therefore, the cell reaction must be
0 2 + 2 + 0
Zn + Cu → Zn + Cu
It is convenient to have names for the terminals of an electrochemical cell. One naming convention is to call one terminal the
anode and the other terminal the cathode. The definition is that the cathode is the electrode at which a reacting species is reduced.
In the silver–silver ion containing cell, the silver electrode is the cathode. In the zinc–zinc ion containing cell, the copper electrode
is the cathode. In these cells, the cathode is the electrically positive electrode. An important feature of these experiments is that the
direction of the electrical potential in the external circuit is established by the reactions that occur spontaneously in the cells. The
cells are sources of electrical current. Cells that operate to produce current are called galvanic cells.
This page titled 17.6: Electrochemical Cells as Circuit Elements is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Figure 4. Applied potential augments the cell potential.

We can incorporate another potential source into the external circuit of an electrochemical cell. If we do so in such a way that the
two electrical potentials augment one another, as diagrammed in Figure 4, the potential drop around the new external circuit is the
sum of the potential drops of the two sources taken independently. The direction of electron flow is unchanged. An electron
anywhere in the external circuit is propelled in the same direction by either potential source. The effective potential difference in
the composite circuit is the sum of the potentials that the sources exhibit when each acts alone.
Figure 5. Applied potential opposes the cell potential.

Alternatively, we can connect the two potential sources so that they oppose one another, as diagrammed in Figure 5. Now an
electron in the external circuit is pushed in one direction by one of the potential sources and in the opposite direction by the other
potential source. The effective potential difference in the composite circuit is the difference between the potentials that the sources
exhibit when each acts alone. In the composite circuit, the direction of electron flow is determined by the potential source whose
potential difference is greater.
This has a dramatic effect on the direction of the reaction occurring in the weaker cell. In the composite cell, the direction of
electron flow through the weaker cell is opposite to the direction of electron flow when the weaker cell is operating as a galvanic
cell. Since the direction of electron flow in the external circuit determines the directions in which the half-reactions occur, the
chemical reaction that occurs in the cell must occur in the opposite direction also. When the direction of current flow through a cell
is determined by connection to a greater potential difference in this fashion, the cell is called an electrolytic cell. Reduction occurs
at the negative terminal of an electrolytic cell. In an electrolytic cell, the cathode is the electrically negative electrode. The direction
of current flow in any cell can be reversed by the application of a sufficiently large counter-potential.
When a cell operates as a source of current (that is, as a galvanic cell), the cell reaction is a spontaneous process. Since, as the cell
reaction proceeds, electrons move through a potential difference in the external circuit, the reaction releases energy in the cell’s
surroundings. If the external circuit is simply a resistor, as when the terminals are short-circuited, the energy is released as heat. Let
q be the heat released and let Q be the amount of charge passed through the external circuit in a time interval Δt . The heat-release
rate is given by
q ΔE QE
= =
Δt Δt Δt
The electrical current is I = Q/Δt . If the resistor follows Ohm’s law, E = IR , where R is the magnitude of the resistance, the
heat release rate becomes
q 2
=I R
Δt
As the reaction proceeds and energy is dissipated in the external circuit, the ability of the cell to supply further energy is
continuously diminished. The energy delivered to the surroundings through the external circuit comes at the expense of the cell’s
internal energy and corresponds to the depletion of the cell reactants.
When the chemical reaction occurring within a cell is driven by the application of an externally supplied potential difference, the
opposite occurs. In the driven (electrolytic) cell, the direction of the cell reaction is opposite the direction of the spontaneous
reaction that occurs when the cell operates galvanically. The electrolytic process produces the chemical reagents that are consumed
in the spontaneous cell reaction. The external circuit delivers energy to the electrolytic cell, increasing its content of spontaneous-
direction reactants and thereby increasing its ability to do work.
In summary, the essential difference between electrolytic and galvanic cells lies in the factor that determines the direction of current
flow and, correspondingly, the direction in which the cell reaction occurs. In a galvanic cell, a spontaneous chemical reaction
occurs and this reaction determines the direction of current flow and the signs of the electrode potentials. In an electrolytic cell, the
sign of the electrode potentials is determined by an applied potential source, which determines the direction of current flow; the cell
reaction proceeds in the non-spontaneous direction.
This page titled 17.7: The Direction of Electron Flow and its Implications is shared under a CC BY-SA 4.0 license and was authored, remixed,
Electrolytic cells are very important in the manufacture of products essential to our technology-intensive civilization. Only
electrolytic processes can produce many materials, notably metals that are strong reducing agents. Aluminum and the alkali metals
are conspicuous examples. Many manufacturing processes that are not themselves electrolytic utilize materials that are produced in
electrolytic cells. These processes would not be possible if the electrolytic products were not available. For example, elemental
silicon, the essential precursor of most contemporary computer chips, is produced from silicon tetrachloride by reduction with
sodium.
0 0
SiC l4 + 4 N a → Si + 4 N aC l
(The silicon so produced is intensively refined, formed into large single crystals, and sliced into wafers before the chip-
manufacturing process begins.) Elemental sodium is produced by the electrolysis of molten sodium chloride.
Successful electrolytic processes involve artful selection of the current-collector material and the reaction conditions. The design of
the cell is often crucial. Since sodium metal reacts violently with water, we recognize immediately that electrolysis of aqueous
sodium chloride solutions cannot produce sodium metal. What products are obtained depends on numerous factors, notably the
composition of the electrodes, the concentration of the salt solution, and the potential that is applied to the cell.
Electrolysis of concentrated, aqueous, sodium chloride solutions is used on a vast scale in the chlor-alkali process for the co-
production of chlorine and sodium hydroxide, both of which are essential for the manufacture of many common products.
2 N aC l (aq) + 2 H2 O (ℓ) → 2 N aOH (aq) + C l2 (g) + H2 (g)
Hydrogen is a by-product. The overall process does not involve sodium ion; rather, the overall reaction is an oxidation of chloride
ion and a reduction of water.
− −
2 Cl (aq) → C l2 (g) + 2 e
− −
2 H2 O (ℓ) + 2 e → 2 OH (aq) + H2 (g)
The engineering difficulties associated with the chlor-alkali process are substantial. They occur because hydroxide ion reacts with
chlorine gas; a practical cell must be designed to keep these two products separate. Commercially, two different designs have been
successful. The diaphragm-cell process uses a porous barrier to separate the anodic and cathodic cell compartments. The mercury-
cell process uses elemental mercury as the cathodic current collector; in this case, sodium ion is reduced, but the product is sodium
amalgam (sodium–mercury alloy) not elemental sodium. Like metallic sodium, sodium amalgam reduces water, but the amalgam
reaction is much slower. The amalgam is removed from the cell and reacted with water to produce sodium hydroxide and
regenerate mercury for recycle to the electrolytic cell.
Elemental sodium is manufactured by the electrolysis of molten sodium chloride. This is effected commercially using an iron
cathode and a carbon anode. The reaction is
0
N aC l (ℓ) → N a + C l2 (g)
Figure 6. An electrolytic cell for the production of sodium metal.
Such a cell is diagrammed in Figure 6. A mechanical barrier suffices to keep the products separate and prevent their spontaneous
reaction back to the salt. A more significant problem in the design of the cell was to find an anode material that did not react with
the chlorine produced. From the cell reaction, we see that one electron passes through the external circuit for every sodium atom
that is produced. The charge that passes through the external circuit during the production of one mole of sodium metal is,
therefore, the charge on one mole of electrons.
In honor of Michael Faraday, the magnitude of the charge carried by a mole of electrons is called the faraday. The faraday constant
is denoted by the symbol “F .” That is,
23 −19
6.02214 × 10 electrons 1.602187 × 10 C
−1
1F = × = 96, 485 C mol
mol electron
The faraday is a useful unit in electrochemical calculations. The unit of electrical current, the ampere, is defined as the passage of
one coulomb per second. Knowing the current in a circuit and the time for which it is passed, we can calculate the number of
coulombs that are passed. Remembering the value of one faraday enables us to do stoichiometric calculations without bringing in
Avogadro’s number and the electron charge every time.
Tabulated information about the thermodynamic characteristics of half-reactions enable us to make useful predictions about what
can and cannot occur in various cells that we might think of building. This information can be used to predict the potential
difference that will be observed in a galvanic cell made by connecting two arbitrarily selected half-cells. In any electrolytic cell,
more than one electron-transfer reaction can usually occur. In the chlor-alkali process, for example, water rather than chloride ion
might be oxidized at the anode. In such cases, tabulated half-cell data enable us to predict which species can react at a particular
applied potential.
This page titled 17.8: Electrolysis and the Faraday is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
From the considerations we have discussed, it is evident that any electrolytic cell involves a flow of electrons in an external circuit
and a flow of ions within the materials comprising the cell. The function of the current collectors is to transfer electrons back and
forth between the external circuit and the cell reagents.
The measurement of solution conductivity is a useful technique for determining the concentrations and mobilities of ions in
solution. Since conductivity measurements involve the passage of electrical current through a liquid medium, the process must
involve electrode reactions as well as motion of ions through the liquid. Normally, the electrode reactions are of little concern in
conductivity measurements. The applied potential is made large enough to ensure that some electrode reaction occurs. When the
liquid medium is water, the electrode reactions are usually the reduction of water at the cathode and its oxidation at the anode. The
conductivity attributable to a given ionic species is approximately proportional to its concentration. In the absence of dissolved
ions, little current is passed. For aqueous solutions, this just restates the familiar observation that pure water is a poor electrical
conductor. When few ions are present, it is not possible to move charge through the cell quickly enough to support a significant
current in the external circuit.
This page titled 17.9: Electrochemistry and Conductivity is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
In Section 17.4, we introduce the idea of a half-reaction and a half-cell in the context of balancing equations for oxidation–
reduction reactions. The real utility of these ideas is that they correspond to distinguishable parts of actual electrochemical cells.
Information about the direction of a spontaneous reaction enables us to predict the relative electrical potentials of the half-cells that
make up the corresponding electrochemical cell. Conversely, given information about the characteristic electrical potentials of half-
cells, we can predict what chemical reactions can occur spontaneously. In short, there is a relationship between the electrical
potential of an electrochemical cell at a particular temperature and pressure and the Gibbs free energy change for the corresponding
oxidation–reduction reaction.
Since cell potentials vary with the concentrations of the reactive components, we can simplify our record-keeping requirements by
defining standard reference conditions that apply to a standard electrode of any type. We adopt the convention that a standard
electrochemical cell contains all reactive components at unit activity. The vast majority of electrochemical cells that have been
studied contain aqueous solutions. In data tables, the activity standard state for solute species is nearly always the hypothetical one-
molal solution. For many purposes, it is an adequate approximation to say that all solutes are present at a concentration of one mole
per liter, and all reactive gases at a pressure of one bar. (In Section 17.15, we see that the dependence of cell potential on reagent
concentration is logarithmic.) In Sections 17.2 and 17.7, we discuss the silver–silver ion electrode; in this approximation, a
standard silver–silver ion electrode is one in which the silver ion is present in the solution at a concentration of one mole per liter.
Likewise, a standard copper–cupric ion electrode is one in which cupric ion is present in the solution at one mole per liter.
We also need to choose an arbitrary reference half-cell. The choice that has been adopted is the Standard Hydrogen Electrode,
often abbreviated the S.H.E. The S.H.E. is defined as a piece of platinum metal, immersed in a unit-activity aqueous solution of a
protonic acid, and over whose surface hydrogen gas, at unit fugacity, is passed continuously. These concentration choices make the
electrode a standard electrode. Frequently, it is adequate to approximate the S.H.E. composition by assuming that the hydrogen ion
concentration is one molar and the hydrogen gas pressure is one bar. The half-reaction associated with the S.H.E. is
+ − 1
H +e → H
2 2
We define the electrical potential of this half-cell to be zero volts.
This page titled 17.10: The Standard Hydrogen Electrode (S.H.E) is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
Let us consider some standard electrochemical cells we could construct using the S.H.E. Two possibilities are electrochemical cells
in which the second electrode is the standard silver–silver ion electrode or the standard copper–cupric ion electrode. The diagrams
in Figure 7 summarize the half-reactions and the electrical potentials that we find when we construct these cells.
Figure 7. Two cells, each with a standard hydrogen electrode.

We can also connect these cells so that the two S.H.E. are joined by one wire, while a second wire joints the silver and copper
electrodes. This configuration is sketched in Figure 8. Whatever happens at one S.H.E. happens in the exact reverse at the other
S.H.E. The net effect is essentially the same as connecting the silver– silver ion half-cell to the copper–cupric ion half-cell by a
single salt bridge. If we did not already know what reaction occurs, we could figure it out from the information we have about how
each of these two cells performs when it operates against the S.H.E.
Figure 8. After connecting the two standard hydrogen electrodes, the combination becomes equivalent to two half cells connected
to a salt bridge.
This page titled 17.11: Half-reactions and Half-cells is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
We adopt a very useful convention to tabulate the potential drops across standard electrochemical cells, in which one half-cell is the
S.H.E. Since the potential of the S.H.E. is zero, we define the standard electrode potential, E , of any other standard half-cell (and
o
its associated half-reaction) to be the potential difference when the half-cell operates spontaneously versus the S.H.E. The electrical
potential of the standard half-cell determines both the magnitude and sign of the standard half-cell potential.
If the process that occurs in the half-cell reduces a solution species or the electrode material, electrons traverse the external circuit
toward the half-cell. Hence, the electrical sign of the half-cell terminal is positive. By the convention, the algebraic sign of the cell
potential is positive (E > 0) . If the process that occurs in the half-cell oxidizes a solution species or the electrode, electrons
o
traverse the external circuit away from the half-cell and toward the S.H.E. The electrical sign of the half-cell is negative, and the
algebraic sign of the cell potential is negative (E < 0) .
o
If we know the standard half-cell potential, we know the essential electrical properties of the standard half-cell operating
spontaneously versus the S.H.E. at zero current. In particular, the algebraic sign of the standard half-cell potential tells us the
direction of current flow and hence the direction of the reaction that occurs spontaneously.
An older convention associates the sign of the standard electrode potential with the direction in which an associated half-reaction is
written. This convention is compatible with the definition we have chosen; however, it creates two ways of expressing the same
information. The difference is whether we write the direction of the half-reaction with the electrons appearing on the right or on the
left side of the equation.
When the half-reaction is written as a reduction process, with the electrons appearing on the left, the associated half-cell potential is
called the reduction potential of the half-cell. Thus we would convey the information we have developed about the silver–silver
ion and the copper–copper ion half cells by presenting the reactions and their associated potentials as
+ − 0 0
Ag +e → Ag E = + 0.7992 volts
2+ − 0 0
Cu +2 e → Cu E = + 0.3394 volts
When the half-reaction is written as a reduction process, the sign of the electrode potential is the same as the sign of the electrical
potential of the half-cell when the half-cell operates spontaneously versus the S.H.E. Thus, the reduction potential has the same
algebraic sign as the electrode potential of our definition.
We can convey the same information by writing the half-reaction in the reverse direction; that is, as an oxidation process in the left-
to-right direction so that the electrons appear on the right. The agreed-upon convention is that we reverse the sign of the half-cell
potential when we reverse the direction in which we write the equation. When the half-reaction is written as an oxidation process,
the associated half-cell potential is called the oxidation potential of the half-cell. Older tabulations of electrochemical data often
present half-reactions written as oxidation processes, with the electrons on the right, and present the potential information using the
oxidation potential convention.
+ − 0 0
Ag +e → Ag E = + 0.7992 volts
reduction potential
0 + − 0
Ag → Ag +e E = − 0.7992 volts
oxidation potential
Note that, in the convention that we have adopted, the term half-cell potential always denotes the potential of the half-cell when it
operates spontaneously versus the S.H.E. In this convention, we do not need to write the half-reaction in order to specify the
standard potential. It is sufficient to specify the chemical constituents of the half-cell. This is achieved using another
representational convention.
Cell Notation
This cell-describing convention lists the active components of a half-cell, using a vertical line to indicate the presence of a phase
boundary like that separating silver metal from an aqueous solution containing silver ion. The silver–silver ion cell is denoted
∣ Ag . (Using the superscript zero on the symbol for elemental silver is redundant; however, it does promote clarity.) The
+ 0
Ag
copper–cupric ion cell is denoted
2+ 0
Cu ∣ Cu .
The S.H.E. is denoted H ∣ H ∣ P t , reflecting the presence of three distinct phases in the operating electrode. A complete
+
2
0
electrochemical cell can be described using this convention. When the complete cell contains a salt bridge, this is indicated with a
pair of vertical lines, ⃦. A cell composed of a silver–silver ion half-cell and a S.H.E. is denoted
0 + + 0
Pt ∣ H2 ∣ H ⃦ Ag ∣ Ag .
A further convention stipulates that the half-cell with the more positive electrode potential is written on the right. Under this
convention, spontaneous operation of the standard full cell transfers electrons through the external circuit from the terminal shown
on the left to the terminal shown on the right.
We can now present our information about the behavior of the silver–silver ion half-cell versus the S.H.E. by writing that the
standard potential of the Ag ∣ Ag half-cell is +0.7792 volts. The standard potential of the C u ∣ C u half-cell is +0.3394 volts.
+ 0 2+ 0
The standard potential of the H ∣ H ∣ P t (the S.H.E.) half-cell is 0.0000 volts. Again, our definition of the standard electrode
+
2
0
potential makes the sign of the standard electrode potential independent of the direction in which the equation of the corresponding
half-reaction is written.
This page titled 17.12: Standard Electrode Potentials is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
While our convention does not use the equation that we write for the half-reaction to establish the algebraic sign of the standard
electrode potential, it is useful to associate the standard electrode potential with the half-reaction written as a reduction, that is, with
the electrons written on the left side of the equation. We also establish the convention that reversing the direction of the half-
reaction reverses the algebraic sign of its potential. When these conventions are followed, the overall reaction and the full-cell
potential can be obtained by adding the corresponding half-cell information. If the resulting full-cell potential is greater than zero,
the spontaneous overall reaction proceeds in the direction it is written, from left to right. If the full-cell potential is negative, the
direction of spontaneous reaction is opposite to that written; that is, a negative full cell potential corresponds to the spontaneous
reaction occurring from right to left. For example,
+ − 0 0
2 Ag +2 e → 2 Ag E = + 0.7992 volts
0 2+ − 0
Cu → Cu +2 e E = − 0.3394 volts
+ 0 0 2+
2 Ag + Cu → 2 Ag + Cu
0
E = + 0.4598 volts
yields the equation corresponding to the spontaneous reaction and a positive full-cell potential. Writing
0 + − 0
2 Ag → 2 Ag +2 e E = − 0.7992 volts
2+ − 0 0
Cu +2 e → Cu E = +0.3394 volts
0 2+ + 0
2 Ag + Cu → 2 Ag + Cu
0
E = − 0.4598 volts
yields the equation for the non-spontaneous reaction and, correspondingly, the full-cell potential is less than zero.
Note that when we multiply a chemical equation by some factor, we do not apply the same factor to the corresponding potential.
The electrical potential of the corresponding electrochemical cell is independent of the number of moles of reactants and products
that we choose to write. The cell potential is an intensive property. It has the same value for a small cell as for a large one, so long
as the other intensive properties (temperature, pressure, and concentrations) are the same.
This page titled 17.13: Predicting the Direction of Spontaneous Change is shared under a CC BY-SA 4.0 license and was authored, remixed,
In Section 17.11, we see that the electrical potential drop across the standard cell P t ∣ H ∣ H ⃦ Ag ∣ Ag is 0.7992 volts.
0
2
+ + 0
We measure this potential under conditions in which no current is flowing. That is, we find the counter-potential at which no
current flows through the cell in either direction. An arbitrarily small change in the counter-potential away from this value, in either
direction, is sufficient to initiate current flow. This means that the standard potential is measured when the cell is operating
reversibly. By the definition of a standard cell, all of the reactants are at the standard condition of unit activity. If any finite current
is drawn from a cell of finite size, the concentrations of the reagents will no longer be exactly the correct values for a standard cell.
Nevertheless, we can calculate the energy that would be dissipated in the surroundings if the cell were to pass one mole of electrons
(corresponding to consuming one mole of silver ions and one-half mole of hydrogen gas) through the external circuit while the cell
conditions remain exactly those of the standard cell. This energy is
−1 −1
96, 485 C mol × 0.7992 V = 77, 110 J mol
The form in which this energy appears in the surroundings depends on the details of the external circuit. However, we know that
this energy represents the reversible work done on electrons in the external circuit as they traverse the path from the anode to the
cathode. We call this the electrical work. Above we describe this as the energy change for a hypothetical reversible process in
which the composition of the cell does not change. We can also view it as the energy change per electron for one electron-worth of
real process, multiplied by the number of electrons in a mole. Finally, we can also describe it as the reversible work done on
electrons during the reaction of one mole of silver ions in an infinitely large standard cell.
The Gibbs free energy change for an incremental reversible process is dG = V dP + SdT + dw , where dw is the
N PV N PV
increment of non-pressure–volume work. In the case of an electrochemical cell, the electrical work is non-pressure–volume work.
In the particular case of an electrochemical cell operated at constant temperature and pressure, dP = dT = 0 , and
dG = dw N PV= dw . elect
The electrical work is just the charge times the potential drop. Letting n be the number of moles of electrons that pass through the
external circuit for one unit of reaction, the total charge is Q = −nF , where F is one faraday. For a standard cell, the potential
drop is E , so the work done on the electrons is Q E = −nF E . Since the standard conditions for Gibbs free energies are the
0 0 0
same as those for electrical cell potentials, we have

rev o o
w = Δr G = −nF E
elect
If the reaction occurs spontaneously when all of the reagents are in their standard states, we have E > 0 . For a spontaneous
o
rev
process, the work done on the system is less than zero, w < 0 ; the work done on the surroundings is w
rev
elect
^ = −w > 0 ; and
elect
rev
elect
the energy of the surroundings increases as the cell reaction proceeds. The standard potential is an intensive property; it is
independent of the size of the cell and of the way we write the equation for the chemical reaction. However, the work and the Gibbs
free energy change depend on the number of electrons that pass through the external circuit. We usually specify the number of
electrons by specifying the chemical equation to which the Gibbs free energy change applies. That is, if the associated reaction is
written as
1
+ 0 +
Ag + H2 → Ag +H
2
we understand that one mole of silver ions are reduced and one mole of electrons are transferred; n = 1 and
ΔG
o
= −F E
o
. If the reaction is written
+ 0 +
2 Ag + H2 → 2 Ag +2 H
we understand that two moles of silver ions are reduced and two moles of electrons are transferred, so that n =2 and
ΔG = −2F E .
o o
The same considerations apply to measurement of the potential of electrochemical cells whose component are not at the standard
condition of unit activity. If the cell is not a standard cell, we can still measure its potential. We use the same symbol to denote the
potential, but we omit the superscript zero that denotes standard conditions. These are, of course, just the conventions we have been
using to distinguish the changes in other thermodynamic functions that occur at standard conditions from those that do not. We
have therefore, for the Gibbs free energy change for the reaction occurring in an electrochemical cell that is not at standard
conditions,
rev
w = Δr G = −nF E
elect
This page titled 17.14: Cell Potentials and the Gibbs Free Energy is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
In Chapter 14, we find that the Gibbs free energy change is a function of the activities of the reactants and products. For the general
reaction aA + bB → cC + dD
we have
~c ~d
aC aD
o
~a ~b
aA aB
Using the relationship between cell potentials and the Gibbs free energy, we find
~c ~d
aC aD
o
−nF E = −nF E + RT ln
~a ~b
aA aB
or
~c ~d
RT aC aD
o
E =E − ln
nF ~a ~b
aA aB
This is the Nernst equation. We derive it from our previous results for the activity dependence of the Gibbs free energy, which
makes no explicit reference to electrochemical measurements at all. When we make the appropriate experimental measurements,
we find that the Nernst equation accurately represents the temperature and concentration dependence of electrochemical-cell
potentials.
Reagent activities are often approximated adequately by molalities or molarities, for solute species, and by partial pressures—
expressed in bars—for gases. The activities of pure solid and liquid phases can be taken as unity. For example, if we consider the
reaction
1
+ 0 +
Ag + H2 → Ag +H
2
it is often sufficiently accurate to approximate the Nernst equation as

+
RT [H ]
o
E =E − ln
+ 1/2
nF
[ Ag ]P
H2
This page titled 17.15: The Nernst Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
If the S.H.E. is one of the half-cells, the corresponding Nernst equation can be viewed as a description of the other half-cell. Using
the cell in which the silver–silver ion electrode opposes the S.H.E., as in the preceding example, the cell potential is the algebraic
sum of the potential of the silver terminal and the potential of the platinum terminal. We can represent the potential of the silver–
silver ion electrode as E Ag∣Ag
+. Since the S.H.E. is always at standard conditions, its potential, which we can represent as
E
o
, is zero by definition. The cell potential is
P t∣ H2 ∣ H
+
o
E = EAg∣Ag + + E +
P t∣ H2 ∣ H
The potential of the cell with both half-cells at standard conditions is

o o o
E =E + +E
Ag∣Ag +
P t∣ H2 ∣ H
and, again since the S.H.E. is at standard conditions, a

~
H
+ =1 and P H2 =1 . Substituting into the Nernst equation for the full cell,
we have
o o o
RT 1
EAg∣Ag + + E + = EAg∣Ag + + E + − ln
P t∣ H2 ∣ H P t∣ H2 ∣ H ~+
F a
Ag
or
o
RT 1
EAg∣Ag + = E − ln
Ag∣Ag
+
~
F aAg +
where the algebraic signs of E Ag∣Ag

+ and E o
Ag∣Ag
+ correspond to writing the half-reaction in the direction Ag +
+e
−
→ Ag
0
. Note
that this is precisely the equation that we would obtain by writing out the Nernst equation corresponding to the chemical equation
Ag
+
+e
−
→ Ag
0
.
To see how these various conventions work together, let us consider the oxidation of hydroquinone (H 2 Q) to quinone (Q) by ferric
ion in acidic aqueous solutions:
3+ 2+ +
2 Fe + H2 Q ⇌ 2 F e + Q + 2H
The quinone–hydroquinone couple is
and the ferric ion–ferrous ion couple is

3+ − 2+
Fe +e ⇌ Fe
The standard electrode potentials are E = +0.699 v and E

P t∣Q, H2 Q, H
+ = +0.783 v . In each case, the numerical value
P t∣F e
3+
,F e
2+
is the potential of a full cell in which the other electrode is the S.H.E. The algebraic sign of the half-cell potential is equal to the
sign of the half-cell’s electrical potential when it operates versus the S.H.E.
To carry out this reaction in an electrochemical cell, we can use a salt bridge to join a P t ∣ F e , F e cell to a P t ∣ Q, H Q, H 3+ 2+
2
+
cell. To construct a standard P t ∣ F e , F e cell, we need only insert a platinum wire into a solution containing ferric and ferrous
3+ 2+
ions, both at unit activity. To construct a standard P t ∣ Q, H Q, H cell, we insert a platinum wire into a solution containing
2
+
quinone, hydroquinone, and hydronium ion, all at unit activity. For standard half-cells, the cathode and anode reactions are
3+ − 2+
Fe +e ⇌ Fe
and
+ −
H2 Q ⇌ Q + 2 H + 2e
We can immediately write the Nernst equation for each of these half-reactions as
~
o
RT aF e2+
EP t∣F e3+ ,F e2+ = E − ln
P t∣F e
3+
,F e
2+
~
F aF e3+
and
~ ~2
RT aQ a +
o H
(−EP t∣Q, H + ) = (−E + )− ln
2 Q, H P t∣Q, H2 Q, H ~
2F aH2 Q
If we add the equations for these half-reactions, the result does not correspond to the original full-cell reaction, because the number
of electrons does not cancel. This can be overcome by multiplying the ferric ion–ferrous ion half-reaction by two. What do we then
do about the corresponding half-cell Nernst equation? Clearly, the values of E and E do not depend on the
P t∣ F e
3+
,F e
2+
o
P t∣ F e
3+
,F e
2+
stoichiometric coefficients in the half-reaction equation. However, the activity terms in the logarithm’s argument do, as does the
number of electrons taking part in the half-reaction. We have
3+ − 2+
2F e + 2e ⇌ 2F e
with
~2
RT a
o F e2+
EP t∣F e3+ ,F e2+ =E 3+ 2+ − ln
P t∣F e ,F e ~2
2F a
F e3+
~
RT aF e2+
o
= EP t∣F e3+ ,F e2+ − ln
~
F aF e3+
We see that we can apply any factor we please to the half-reaction. The Nernst equation gives the same dependence of the half-cell
potential on reagent concentrations no matter what factor we choose. This is true also of the Nernst equation for any full-cell
reaction. In the present example, adding the appropriate half-cell equations and their corresponding Nernst equations gives
3+ 2+ +
2 Fe + H2 Q ⇌ 2 F e + Q + 2H
and
E = EP t∣F e3+ ,F e2+ − EP t∣Q, H +

2 Q, H
~2 ~ ~2
RT a RT aQ a +
o o F e2+ H
=E −E − ln − ln
P t∣F e
3+
,F e
2+
P t∣Q, H2 Q, H
+
~2 ~
2F a 2F aH Q
F e3+ 2
~ ~2 ~2
RT aQ a + a
0 H F e2+
=E − ln
2F ~ ~2
aH2 Q a 3+
Fe
This page titled 17.16: The Nernst Equation for Half-cells is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
The same chemical species can be a reactant or product in many different half-cells. Frequently, data on two different half-cells can
be combined to give information about a third half-cell. Let us consider two half-cells that involve the ferrous ion, F e . Ferrous 2+
ion and elemental iron form a redox couple. The half-cell consists of a piece of pure ion in contact with aqueous ferrous ion at unit
activity. Our notation for this half-cell and its potential are F e ∣ F e and E . The corresponding half-reaction and its
2+
F e∣F e
2+
potential are
2+ − 0
Fe + 2e ⇌ Fe
and
RT 1
o
EF e∣F e2+ = E − ln
F e∣F e2+ ~
2F aF e2+
Ferrous ion can also give up an electron at an inert electrode, forming ferric ion, F e . This process is reversible. Depending on 3+
the potential of the half-cell with which it is paired, the inert electrode can either accept an electron from the external circuit and
deliver it to a ferric ion, or take an electron from a ferrous ion and deliver it to the external circuit. Thus, ferrous and ferric ions
form a redox couple. Platinum metal functions as an inert electrode in this reaction. The half-cell consists of a piece of pure
platinum in contact with aqueous ferrous and ferric ions, both present at unit activity. Our notation for this half-cell and potential
are P t ∣ F e , F e and E
2+ 3+
. The corresponding half-reaction and its potential are
P t∣F e2+ ,F e3+
3+ − 2+
Fe +e ⇌ Fe
and
~
o
RT aF e2+
EP t∣F e2+ ,F e3+ = E − ln
P t∣F e
2+
,F e
3+
~
F aF e3+
We can add these two half-reactions, to obtain

3+ − 0
Fe + 3e ⇌ Fe
The Nernst equation for this half-reaction is

RT 1
o
EF e∣F e3+ = E 3+ − ln
F e∣F e ~
3F aF e3+
From our past considerations, both of these equations are clearly correct. However, in this case, the Nernst equation of the sum is
not the sum of the Nernst equations. Nor should we expect it to be. The half-cell Nernst equations are really shorthand notation for
the behavior of the half-cell when it is operated against a S.H.E. Adding half-cell Nernst equations corresponds to creating a new
system by connecting the two S.H.E. electrodes of two separate full cells, as we illustrate in Figure 8. In the present instance, we
are manipulating two half-reactions to obtain a new half-reaction; this manipulation does not correspond to any possible way of
interconnecting the corresponding half-cells.
Nevertheless, if we know the standard potentials for the first two reactions (E and E ), we can obtain the standard
o
F e∣F e
2+
o
P t∣F e
2+
,F e
3+
potential for their sum (E o
F e∣F e
). To do so, we exploit the relationship we found between electrical potential and Gibbs free
3+
energy. The first two reactions represent sequential steps that jointly achieve the same net change as the third reaction. Therefore,
the sum of the Gibbs free energy changes for the first two reactions must be the same as the Gibbs free energy change for the third
reaction. The standard potentials are not additive, but the Gibbs free energy changes are. We have
3+ − 2+ o o
Fe +e ⇌ Fe ΔG 3+ 2+
= FE 2+ 3+
Fe →F e P t∣F e ,F e
2+ − 0 o o
Fe + 2e ⇌ Fe ΔG = 2F E
F e2+ →F e0 F e∣F e2+
3+ − 0 o o
Fe + 3e ⇌ Fe ΔG 3+ 0
= 3F E 3+
Fe →F e F e∣F e
Since also
o o o
ΔG 3+ 2+
+ ΔG 2+ 0
= ΔG 3+ 0
Fe →F e Fe →F e Fe →F e
we have
o o o
FE 2+ 3+ + 2F E 2+ = 3F E 3+
P t∣F e ,F e F e∣F e F e∣F e
and
o o
E + 2E
P t∣F e2+ ,F e3+ F e∣F e2+
o
E 3+ =
F e∣F e
3
This page titled 17.17: Combining two Half-cell Equations to Obtain a new Half-cell Equation is shared under a CC BY-SA 4.0 license and was
In Section 17.15 we find for the general reaction aA + bB → cD + dD that the Nernst equation is
~c ~d
RT a a
o C D
E =E − ln
nF ~a ~b
a a
A B
We now want to consider the relationship between the potential of an electrochemical cell and the equilibrium position of the cell
reaction. If the potential of the cell is not zero, short-circuiting the terminals of the cell will cause electrons to flow in the external
circuit and reaction to proceed spontaneously in the cell. Since a spontaneous reaction occurs, the cell is not at equilibrium with
respect to the cell reaction.
As we draw current from any electrochemical cell, cell reactants are consumed and cell products are produced. Experimentally, we
see that the cell voltage decreases continuously, and inspection of the Nernst equation shows that it predicts a potential decrease.
Eventually, the voltage of a short-circuited cell decreases to zero. No further current is passed. The cell reaction stops; it has
reached chemical equilibrium. If the cell potential is zero, the cell reaction must be at equilibrium, and vice versa.
We also know that, at equilibrium, the activity ratio that appears as the argument of the logarithmic term is a constant—the
equilibrium constant. So when E = 0 , we have also that
~c ~d
a a
C D
Ka =
~a ~b
a aB
A
Substituting these conditions into the Nernst equation, we obtain
o
RT
0 =E − ln Ka
nF
or
o
(nF E )
Ka = exp
RT
We can obtain this same result if we recall that ΔG = −RT ln K and that ΔG = −nF E . We can determine equilibrium
o
a
o o
constants by measuring the potentials of standard cells. Alternatively, we can measure an equilibrium constant and determine the
potential of the corresponding cell without actually constructing it. Standard potentials and equilibrium constants are both measures
of the Gibbs free energy change when the reaction occurs under standard conditions.
This page titled 17.18: The Nernst Equation and the Criterion for Equilibrium is shared under a CC BY-SA 4.0 license and was authored, remixed,
17.19: Problems
1. Balance the following chemical equations assuming that they occur in aqueous solution.
(a) C u 0
+ Ag
+
→ Cu
2+
+ Ag
0
(b) F e 2+
+ C r2 O
2−
7
→ Fe
3+
+ Cr
3+
(c) C r 2+
+ C r2 O
2−
7
→ Cr
3+
(d) C l 2 + Br
−
→ Cl
−
+ Br2
(e) C lO 3 → Cl
−
+ C lO
−
4
(f) I −
+ IO
−
3
→ I2
(g) I −
+ O2 → I2 + OH
−
(basic solution)
(h) H 2 C2 O4
2−
+ M nO
−
4
→ C O2 + M n
2+
(i) F e 2+
+ M nO
−
4
→ Fe
3+
+ Mn
2+
(j) H 2 O2 + M nO
−
4
→ Mn
2+
+ O2
(k) P bO 2 +Pb
0
+ H2 S O4 → P bSO4
(l) F e 2+
→ Fe
0
+Fe
3+
(m) C u 2+
+Fe
0
→ Cu
0
+Fe
3+
−
(n) Al 0
+ OH
−
→ H2 + Al(OH ) 4 (basic solution)
(o) Au 0
+ CN
−
+ O2 → Au (C N )
−
4
(basic solution)
(p) C u 0
+ H N O3 → C u
2+
+ N O2
(q) Al 0
+ F e2 O3 → Al2 O3 + F e
0
(r) I −
+ H2 O2 → I2 + OH
−
(basic solution)
(s) H F eO −
4
+ Mn
2+
→ Fe
3+
+ M nO2
(t) F e 2+
+ S2 O
2−
8
→ Fe
3+
+ SO
2−
(u) C u +
+ O2 → C u
2+
+ OH
−
(basic solution)
2. Calculate the equilibrium constant,K , for the reaction C u + 2Ag → C u + 2Ag . An excess of clean copper wire is
a
0 + 2+ 0
placed in a 10 –– M silver nitrate solution. Assuming that molarities adequately approximate the activities of the ions, find the
−1
equilibrium concentrations of Ag and C u . + 2+
3. The standard potentials for reduction of F e 2+

and F e 3+
to F e are0
2+ − 0 o
Fe + 2e → Fe E = −0.447 v
3+ − 0 o
Fe + 3e → Fe E = −0.037 v
3+
(a) Find the standard potential for the disproportionation of F e 2+
to F e 3+
and F e : 0
Fe
2+
→ Fe
0
+Fe .
(b) Find the standard half-cell potential for the reduction of F e 3+
to F e : F e2+ 3+
+e
−
→ Fe
2+
.
4. The standard potential for reduction of tris-ethylenediamineruthenium (III) to tris-ethylenediamineruthenium (II) is
3+ − 2+ o
[Ru (en) ] +e → [Ru (en) ] E = +0.210 v
3 3
Half-cell potential data are given below for several oxidants. Which of them can oxidize [Ru (en) 3
2+
] to [Ru (en) 3
3+
] in acidic (
+
[H ] ≈ a
~
= 10 ) aqueous solution?
H
+
−1
(a) U O 2+
2
+e
−
→ UO
+
2
E
o
= +0.062 v
(b) [Ru (N H 3 )6 ]
3+
+e
−
→ [Ru (N H3 ) ]
6
2+
o
E = +0.10 v
(c) C u2+
+e
−
→ Cu
+
E
o
= +0.153 v
(d) AgC l + e −
→ Ag
0
+ Cl
−
E
o
= +0.222 v
(e) H g 2
C l2 + 2 e
−
→ 2H g
0
+ 2C l
−
o
E = +0.268 v
(f) AgC N + e −
→ Ag
0
+ CN
−
E
o
= −0.017 v
(g) SnO 2 + 4H
+
+ 4e
−
→ Sn
0
+ 2 H2 O
o
E = −0.117 v
5. An electrochemical cell is constructed in which one half cell is a standard hydrogen electrode and the other is a hydrogen
electrode immersed in a solution of pH = 7 ([H ] ≈ a~
= 10 ) . What is the potential difference between the terminals of the
+
H
+
−7
cell? What chemical change occurs in this cell?

6. The standard half-cell potential for the reduction of oxygen gas at an inert electrode (like platinum metal) is
+ − o
O2 + 4 H + 4e → 2 H2 OE = +1.229 v
An electrochemical cell is constructed in which one half cell is a standard hydrogen electrode and the other cell is a piece of
platinum metal, immersed in a 1 –M
––
solution of H C lO , which is continuously in contact with bubbles of oxygen gas at a pressure
4
of 1 bar.
(a) What is the potential difference between the terminals of the cell? What chemical change occurs in this cell?
(b) The 1 –M
––
H C lO solution in part (a) is replaced with pure water
4 ([ H
+ ~
] ≈ aH + = 10
−7
) . What is the potential difference
between the terminals of this cell?
(c) The 1 –M
––
H C lO solution in part (a) is replaced with 1 M
4
–––
N aOH ([ H
+ ~
] ≈ aH + = 10
−14
) . What is the potential difference
between the terminals of this cell?
7. A variable electrical potential source is introduced into the external circuit of the cell in part (a) of problem
6. The negative terminal of the potential source is connected to the oxygen electrode and the positive terminal of the potential
source is connected to the standard hydrogen electrode. If the applied electrical potential is 1.3 v, what chemical change occurs?
What is the minimum electrical potential that must be applied to electrolyze water if the oxygen electrode contains a 1 –M
––
H C lO 4
solution? A neutral (pH = 7 ) solution? A 1 –M

––
N aOH solution?
8. Two platinum electrodes are immersed in 1 M H C lO . What potential difference must be applied between these electrodes in
4
–––
order to electrolyze water? (Assume that P = 1 bar and P = 1 bar at their respective electrodes, as will be the case as soon
O2 H2
as a few bubbles of gas have accumulated at each electrode.) What potential difference is required if the electrodes are immersed in
pure water? In 1 M N aOH ?
–––
CHAPTER OVERVIEW
18: Quantum Mechanics and Molecular Energy Levels

This page titled 18: Quantum Mechanics and Molecular Energy Levels is shared under a CC BY-SA 4.0 license and was authored, remixed,
1
Beginning in Chapter 20, we turn our attention to the distribution of energy among the molecules in a closed system that is
immersed in a constant-temperature bath, that is at equilibrium, and that contains a large number of molecules. We want to find the
probability that the energy of a molecule in such a system is in a particular interval of energy values. This probability is also the
fraction of the molecules whose energies are in the specified interval, since we assume that these statements mean the same thing
for a system at equilibrium.
The probability that the energy of a particular molecule is in a particular interval is intimately related to the energies that it is
possible for a molecule to have. Before we can make further progress in describing molecular energy distributions, we must discuss
atomic and molecular energies. For our development of the Boltzmann equation, we need to introduce the idea of quantized energy
states. This requires a short digression on the basic ideas of quantum mechanics and the quantized energy levels of atoms and
molecules.
We have derived two expressions that relate the energy of a molecule to the probability that the molecule will have that energy. One
follows from the barometric formula
−mgh
η (h) = η (0) exp ( )
kT
−ϵpotential
= η (0) exp ( )
kT
in which the number density of molecules depends exponentially on their gravitational potential energy, mgh , and the reciprocal of
the temperature. From the barometric formula, we can find the probability density function
df mg −mgh
= exp ( )
dh kT kT
(See problem 3.22) The other is the Maxwell-Boltzmann distribution function

3/2 2 3/2
df m 2
−mv m 2
−ϵkinetic
= 4π ( ) v exp ( ) = 4π ( ) v exp ( )
dv 2πkT 2kT 2πkT 2kT
in which the probability density of molecular velocities depends exponentially on their kinetic energies, mv /2, and the reciprocal
2
of the temperature. We will see that this dependence is very general. Any time the molecules in a system can have a range of
energies, the probability that a molecule has energy ϵ is proportional to exp (−ϵ/kT ) . The exponential term, exp (−ϵ/kT ) , is
often called the Boltzmann factor.
We might try to develop a more general version of the Maxwell-Boltzmann distribution function by an argument that somehow
parallels our derivation of the Maxwell-Boltzmann equation. It turns out that any such attempt is doomed to failure, because it is
based on a fundamentally incorrect view of nature. In developing the barometric formula and the Maxwell-Boltzmann distribution,
we assume that the possible energies are continuous; a molecule can be at any height above the surface of the earth, and its
translational velocity can have any value. When we turn to the distribution of other ways in which molecules can have energy, we
find that this assumption produces erroneous predictions about the behavior of macroscopic collections of molecules.
The failure of such attempts led Max Planck to the first formulation of the idea that energy is quantized. The spectrum of light
emitted from glowing-hot objects (so-called “black bodies”) depends on the temperature of the emitting object. Much of the
experimentally observable behavior of light can be explained by the hypothesis that light behaves like a wave. Mechanical (matter-
displacement) waves carry energy; the greater the amplitude of the wave, the more energy it carries. Now, light is a form of energy,
and a spectrum is an energy distribution. It was a challenge to late nineteenth century physics to use the wave model for the
behavior of light to predict experimentally observed emission spectra. This challenge went unmet until Planck introduced the
postulate that such “black-body radiators” absorb or emit electromagnetic radiation only in discrete quantities, called quanta.
Planck proposed that the energy of one such quantum is related to the frequency, ν , of the radiation by the equation E = hν , where
the proportionality constant, h , is now called Planck’s constant. In Planck’s model, the energy of an electromagnetic wave depends
on its frequency, not its amplitude.
In the years following Planck’s hypothesis, it became clear that many properties of atoms and molecules are incompatible with the
idea that an atom or molecule can have any arbitrary energy. We obtain agreement between experimental observations and
theoretical models only if we assume that atoms and molecules can have only very particular energies. This is observed most
conspicuously in the interactions of atoms and molecules with electromagnetic radiation. One such interaction gives rise to a series
of experimental observations known as the photoelectric effect. In order to explain the photoelectric effect, Albert Einstein showed
that it is necessary to extend Planck’s concept to assume that light itself is a stream of discrete energy quanta, called photons. In our
present understanding, it is necessary to describe some of the properties of light as wave-like and some as particle-like.
In many absorption and emission spectra, we find that a given atom or molecule can emit or absorb electromagnetic radiation only
at very particular frequencies. For example, the light emitted by atoms excited by an electrical discharge contains a series of
discrete emission lines. When it is exposed to a continuous spectrum of frequencies, an atom is observed to absorb light at precisely
the discrete frequencies that are observed in emission. Niels Bohr explained these observations by postulating that the electrons in
atoms can have only particular energies. The absorption of visible light by atoms and molecules occurs when an electron takes up
electromagnetic energy and moves from one discrete energy level to a second, higher, one. (Absorption of a continuous range of
frequencies begins to occur only when the light absorbed provides sufficient energy to separate an electron from the original
chemical species, producing a free electron and a positively charged ion. At the onset frequency, neither of the product species has
any kinetic energy. Above the onset frequency, spectra are no longer discrete, and the species produced have increasingly greater
kinetic energies.) Similar discrete absorption lines are observed for the absorption of infrared light and microwave radiation by
diatomic or polyatomic gas molecules. Infrared absorptions are associated with vibrational motions, and microwave absorptions are
associated with rotational motions of the molecule about its center of mass. These phenomena are explained by the quantum theory.
The quantized energy levels of atoms and molecules can be found by solving the Schrödinger equation for the system at hand. To
see the basic ideas that are involved, we discuss the Schrödinger equation and some of the most basic approximations that are made
in applying it to the description of atomic and molecular systems. But first, we should consider one more preliminary question: If
the quantum hypothesis is so important to obtaining valid equations for the distribution of energies, why are the derivations of the
Maxwell-Boltzmann equation and the barometric formula successful? Maxwell’s derivation is successful because the quantum
mechanical description of a molecule’s translational kinetic energy is very well approximated by the assumption that the molecule’s
kinetic energy can have any value. In the language of quantum mechanics, the number of translational energy levels available to a
molecule is very large, and successive energy levels are very close together—so close together that it is a good approximation to
say that they are continuous. Similarly, the gravitational potential energies available to a molecule in the earth’s atmosphere are
well approximated by the assumption that they belong to a continuous distribution.
This page titled 18.1: Energy Distributions and Energy Levels is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
Subsequent to Planck’s proposal that energy is quantized, the introduction of two further concepts led to the theory of quantum
mechanics. The first was Einstein’s relativity theory, and his deduction from it of the equivalence of matter and energy. The
relativistic energy of a particle is given by
2 2 2 2 4
E =p c +m c
0
where p is the momentum and m is the mass of the particle when it is at rest. The second was de Broglie’s hypothesis that any
0
particle of mass m moving at velocity v , behaves like a wave. De Broglie’s hypothesis is an independent postulate about the
structure of nature. In this respect, its status is the same as that of Newton’s laws or the laws of thermodynamics. Nonetheless, we
can construct a line of thought that is probably similar to de Broglie’s, recognizing that these are heuristic arguments and not
logical deductions.
We can suppose that de Broglie’s thinking went something as follows: Planck and Einstein have proposed that electromagnetic
radiation—a wave-like phenomenon—has the particle-like property that it comes in discrete lumps (photons). This means that
things we think of as waves can behave like particles. Conversely, the lump-like photons behave like waves. Is it possible that other
lump-like things can behave like waves? In particular is it possible that material particles might have wave-like properties? If a
material particle behaves like a wave, what wave-like properties should it exhibit?
Well, if we are going to call something a wave, it must have a wavelength, λ , a frequency, ν , and a propagation velocity, v , and
these must be related by the equation v = λν . The velocity of propagation of light is conventionally given the symbol c , so
c = λν . The Planck-Einstein hypothesis says that the energy of a particle (photon) is E = hν = hc/λ . Einstein proposes that the
energy of a particle is given by E = p c + m c . A photon travels at the speed of light. This is compatible with other relativistic
2 2 2 2
0
4
equations only if the rest mass of a photon is zero. Therefore, for a photon, we must have E = pc . Equating these energy
equations, we find that the momentum of a photon is
p = h/λ
Now in a further exercise of imagination, we can suppose that this equation applies also to any mass moving with any velocity.
Then we can replace p with mv, and write
mv = h/λ
We interpret this to mean that any mass, m, moving with velocity, v , has a wavelength, λ , given by
λ = h/mv
This is de Broglie’s hypothesis. We have imagined that de Broglie found it by a series of imaginative—and not entirely logical—
guesses and suppositions. The illogical parts are the reason we call the result a hypothesis rather than a derivation, and the
originality of the guesses and suppositions is the reason de Broglie’s hypothesis was new. It is important physics, because it turns
out to be experimentally valid. Very small particles do exhibit wave-like properties, and de Broglie’s hypothesis correctly predicts
their wavelengths.
In a similar vein, we can imagine that Schrödinger followed a line of thought something like this: de Broglie proposes that any
moving particle behaves like a wave whose wavelength depends on its mass and velocity. If a particle behaves as a wave, it should
have another wave property; it should have an amplitude. In general, the amplitude of a wave depends on location and time, but we
are thinking about a rather particular kind of wave, a wave that—so to speak—stays where we put it. That is, our wave is supposed
to describe a particle, and particles do not dissipate themselves in all directions like the waves we get when we throw a rock in a
pond. We call a wave that stays put a standing wave; it is distinguished by the fact that its amplitude depends on location but not on
time.
Mathematically, the amplitude of any wave can be described as a sum of (possibly many) sine and cosine terms. A single sine term
describes a simple wave. If it is a standing wave, its amplitude depends only on distance, and its amplitude is the same for any two
points separated by one wavelength. Letting the amplitude be ψ , this standing wave is described by ψ (x) = Asin (ax) , where x
is the location, expressed as a distance from the origin at x = 0 . In this wave equation, A and a are parameters that fix the
maximum amplitude and the wavelength, respectively. Requiring the wavelength to be λ means that aλ = 2π . (Since ψ is a sine
function, it repeats every time its argument increases by 2π radians. We require that ψ repeat every time its argument increases by
aλ radians, which requires that aλ = 2π .) Therefore, we have
a = 2π/λ
and the wave equation must be
ψ (x) = Asin (2πx/λ)
Equations, ψ , that describe standing waves satisfy the differential equation

2
d ψ
= −C ψ
dx2
where C is a constant. In the present instance, we see that

2 2 2
d ψ 2π 2πx 2π
= −( ) Asin ( ) = −( ) ψ
2
dx λ λ λ
From de Broglie’s hypothesis, we have λ = h/mv , so that the constant C can be written as
2 2 2 2 2 2
2π 2πmv 2π mv 8π m mv
C =( ) =( ) =( ) (2m) ( ) =( )( )
2
λ h h 2 h 2
Let T be the kinetic energy, mv /2, and let

2
V be the potential energy of our wave-like particle. Then its energy is E = T +V ,
and we have mv /2 = T = E − V .
2
The constant C becomes

2 2
8π m 8π m
C =( )T =( ) (E − V )
2 2
h h
Making this substitution for C , we find a differential equation that describes a standing wave, whose wavelength satisfies the de
Broglie equation. This is the time-independent Schrödinger equation in one dimension:
2 2
d ψ 8π m
= −( ) (E − V ) ψ
2 2
dx h
or
2 2
h d ψ
−( ) + V ψ = Eψ
2
8π m dx2
Often the latter equation is written as

2 2
h d
[− ( ) + V ] ψ = Eψ
2 2
8π m dx
where the expression in square brackets is called the Hamiltonian operator and abbreviated to H , so that the Schrödinger equation
becomes simply, if cryptically,
H ψ = Eψ
If we know how the potential energy of a particle, V , depends on its location, we can write down the Hamiltonian operator and the
Schrödinger equation that describe the wave properties of the particle. Then we need to find the wave equations that satisfy this
differential equation. This can be difficult even when the Schrödinger equation involves only one particle. When we write the
Schrödinger equation for a system containing multiple particles that interact with one another, as for example an atom containing
two or more electrons, analytical solutions become unattainable; only approximate solutions are possible. Fortunately, a great deal
can be done with approximate solutions.
The Schrödinger equation identifies the value of the wavefunction, ψ (x), with the amplitude of the particle wave at the location x.
Unfortunately, there is no physical interpretation for ψ (x); that is, no measurable quantity corresponds to the value of ψ (x). There
is, however, a physical interpretation for the product ψ (x) ψ (x) or ψ (x). [More accurately, the product ψ (x) ψ (x), where
2 ∗
ψ
∗
(x)is the complex conjugate of ψ (x). In general, x is a complex variable.] ψ (x) is the probability density function for the
2
particle whose wavefunction is ψ (x). That is, the product ψ (x) dx is the probability of finding the particle within a small
2
distance, dx, of the location x. Since the particle must be somewhere, we also have
+∞
2
1 =∫ ψ (x)dx
−∞
This page titled 18.2: Quantized Energy - De Broglie's Hypothesis and the Schroedinger Equation is shared under a CC BY-SA 4.0 license and
A problem usually called the particle in a box provides a convenient illustration of the principles involved in setting up and solving
the Schrödinger equation. Besides being a good illustration, the problem also proves to be a useful approximation to many physical
systems. The statement of the problem is simple. We have a particle of mass m that is constrained to move only in one dimension.
For locations in the interval 0 ≤ x ≤ ℓ , the particle has zero potential energy. For locations outside this range, the particle has
infinite potential energy. Since the particle cannot have infinite energy, this means that it can never find its way into locations
outside of the interval 0 ≤ x ≤ ℓ . We can think of this particle as a bead moving on a wire, with stops located on the wire at x = 0
and at x = ℓ . We can also think of it as being confined to a one-dimensional box of length ℓ , which is the viewpoint represented by
the name. The particle in a box model is diagrammed in Figure 1.
Figure 1. Potential energy versus distance for a particle in a box.

The potential energy constraints mean that the amplitude of the particle’s wavefunction must be zero, ψ (x) = 0 , when the value of
x lies in the interval
−∞x < 0
or
ℓ < x < +∞
We assume that the probability of finding the particle cannot change abruptly when its location changes by an arbitrarily small
amount. This means that the wavefunction must be continuous, and it follows that ψ (0) = 0 and ψ (ℓ) = 0 . Inside the box, the
particle’s Schrödinger equation is
2 2
h d ψ
−( ) = Eψ
2 2
8π m dx
and we seek those functions ψ (x) that satisfy both this differential equation and the constraint equations ψ (0) = 0 and ψ (ℓ) = 0 .
It turns out that there are infinitely many such solutions, ψ , each of which corresponds to a unique energy level, E .
n n
To find these solutions, we first guess—guided by our considerations in §2—that solutions will be of the form
ψ (x) = Asin (ax) + Bcos (bx)
A solution must satisfy
ψ (0) = Asin (0) + Bcos (0) = Bcos (0) = 0
so that B = 0 . At the other end of the box, we must have
ψ (ℓ) = Asin (aℓ) =0
which means that aℓ = nπ , where n is any integer: n = 1, 2, …. Hence, we have
a = nπ/ℓ
and the only equations of the proposed form that satisfy the conditions at the ends of the box are
ψn (x) = Asin (nπx/ℓ)
To test whether these equations satisfy the Schrödinger equation, we check

2 2
h d nπx
−( ) [Asin ( ) ] = En ψn
8π 2 m dx2 ℓ
and find
2 2 2 2
h nπ nπx n h
( )( ) [Asin ( ) ] =( ) ψn = En ψn
2 2
8π m ℓ ℓ 8mℓ
so that the wavefunctions ψn (x) = Asin (nπx/ℓ) are indeed solutions and the energy, En , associated with the wavefunction
ψ (x) is
n
2 2
n h
En =
2
8mℓ
We see that the energy values are quantized; although there are infinitely many energy levels, E , only very particular real numbers n
—those given by the equation above—correspond to energies that the particle can have. If we sketch the first few wavefunctions,
ψ (x), we see that there are always n − 1 locations inside the box at which ψ (x) is zero. These locations are called nodes. Once
n n
we know n , we know the number of nodes, and we can sketch the general shape of the corresponding wavefunction. The first three
wavefunctions and their squares are sketched in Figure 2.
Figure 2. Wave functions for a particle in a box.

At this point, we have found a complete set of infinitely many solutions, except for the parameter A . To determine A , we interpret
ψ (x) as a probability density function, and we require that the probability of finding the particle in the box be equal to unity. This
2
means that
ℓ ℓ
nπx 1 1 2nπx ℓ
2 2 2 2
1 =∫ A sin ( ) dx = A ∫ [ − cos ( ) ] dx = A ( )
0
ℓ 0
2 2 ℓ 2
−−
−
so that A = √2/ℓ , and the final wavefunctions are
−
−
2 nπx
ψn (x) = √ sin ( )
ℓ ℓ
This page titled 18.3: The Schrödinger Equation for A Particle in A Box is shared under a CC BY-SA 4.0 license and was authored, remixed,
Molecules are composed of atoms, and atoms are composed of nuclei and electrons. When we consider the internal motions of
molecules, we have to consider the motions of a large number of charged particles with respect to one another. In principle, we can
write down the potential function (the V in the Schrödinger equation) that describes the Coulomb’s law based potential energy of
the system of charged particles. In principle, we can then solve the Schrödinger equation and obtain a series of wavefunctions,
ψ (x), and their corresponding energies, E , that completely characterize the motions of the molecule’s constituent particles. Each
n n
of the E is an energy value that the molecule can have. Often we say that it is an energy level that the molecule can occupy.
n
Since every distance between two charged particles is a variable in the Schrödinger equation, the number of variables increases
dramatically as the size of the molecule increases. The two-particle hydrogen-atom problem has been solved analytically. For any
chemical species larger than the hydrogen atom, only approximate solutions are possible. Nevertheless, approximate results can be
obtained to very high accuracy. Greater accuracy comes at the expense of more extensive calculations.
Let us look briefly at the more fundamental approximations that are made. One is called the Born-Oppenheimer approximation; it
states that the motions of the nuclei in a molecule are too slow to affect the motions of the electrons. This occurs because nuclei are
much more massive than electrons. The Born-Oppenheimer approximation assumes that the electronic motions can be calculated as
if the nuclei are fixed at their equilibrium positions without introducing significant error into the result. That is, there is an
approximate wavefunction describing the motions of the electrons that is independent of a second wavefunction that describes the
motions of the nuclei.
The mathematical description of the nuclear motions can be further simplified using additional approximations; we can separate the
nuclear motions into translational, rotational, and vibrational modes. Translational motion is the three-dimensional displacement of
an entire molecule. It can be described by specifying the motion of the molecule’s center of mass. The motions of the constituent
nuclei with respect to one another can be further subdivided: rotational motions change the orientation of the whole molecule in
space; vibrational motions change distances between constituent nuclei.
The result is that the wavefunction for the molecule as a whole can be approximated as a product of a wavefunction (ψ or electronic
ψ ) for the electronic motions, a wavefunction (ψ

e
or ψ ) for the vibrational motions, a wavefunction (ψ
vibration v
or ψ ) for
rotation r
the rotational motions, and a wavefunction (ψ or ψ ) for the translational motion of the center of mass. We can write
translation t
ψmolecule = ψe ψv ψr ψt
(None of this is supposed to be obvious. We are merely describing the essential results of a considerably more extensive
development.)
When we write the Hamiltonian for a molecule under the approximation that the electronic, vibrational, rotational, and translational
motions are independent of each other, we find that the Hamiltonian is a sum of terms. In some of these terms, the only
independent variables are those that specify the locations of the electrons. We call these variables electronic coordinates. Some of
the remaining terms involve only vibrational coordinates, some involve only rotational coordinates, and some involve only
translational coordinates. That is, we find that the Hamiltonian for the molecule can be expressed as a sum of terms, each of which
is the Hamiltonian for one of the kinds of motion:
Hmolecule = He + Hv + Hr +H t
where we have again abbreviated the subscripts denoting the various categories of motion.
Consequently, when we write the Schoedinger equation for the molecule in this approximation, we have
Hmolecule ψmolecule = (He + Hv + Hr +H t ) ψe ψv ψr ψt
= ψ ψ ψ He ψ + ψ ψ ψ Hv ψ + ψ ψ ψ Hr ψ + ψ ψ ψ Ht ψ
v r t e e r t v e v t r e v r t
= ψv ψr ψt Ee ψe + ψe ψr ψt Ev ψv + ψe ψv ψt Er ψr + ψe ψv ψr Et ψt
= (Ee + Ev + Er +E t ) ψe ψv ψr ψt
We find that the energy of the molecule as a whole is simply the sum of the energies associated with the several kinds of motion
Emolecule = Ee + Ev + Er +E t
ψt , ψ , ψ , and ψ can be further approximated as products of wavefunctions involving still smaller numbers of coordinates. We
v r e
can have a component wavefunction for every distinguishable coordinate that describes a possible motion of a portion of the
molecule. The three translational modes are independent of one another. It is a good approximation to assume that they are also
independent of the rotational and vibrational modes. Frequently, it is a good approximation to assume that the vibrational and
rotational modes are independent of one another. We can deduce the number of one-dimensional wavefunctions that are required to
give an approximate wavefunction that describes all of the molecular motions, because this will be the same as the number of
coordinates required to describe the nuclear motions. If we have a collection of N atoms that are not bonded to one another, each
atom is free to move in three dimensions. The number of coordinates required to describe their motion is 3N . When the same
atoms are bonded to one another in a molecule, the total number of motions remains the same, but it becomes convenient to
reorganize the way we describe them.
First, we recognize that the atomic nuclei in a molecule occupy positions that are approximately fixed relative to one another.
Therefore, to a good approximation, the motion of the center of mass is independent of the way that the atoms move relative to one
another or relative to the center of mass. It takes three coordinates to describe the motion of the center of mass, so there are 3N − 3
coordinates left over after this is done.
The number of rotational motions available to a molecule depends upon the number of independent axes about which it can rotate.
We can imagine a rotation of a molecule about any axis we choose. In general, in three dimensions, we can choose any three non-
parallel axes and imagine that the molecule rotates about each of them independently of its rotation about the others. If we consider
a set of more than three non-parallel axes, we find that any of the axes can be expressed as a combination of any three of the others.
This means that the maximum number of independent rotational motions for the molecule as a whole is three.
If the molecule is linear, we can take the molecular axis as one of the axes of rotation. Most conveniently, we can then choose the
other two axes to be perpendicular to the molecular axis and perpendicular to each other. However, rotation about the molecular
axis does not change anything about the molecule’s orientation in space. If the molecule is linear, rotation about the molecular axis
is not a rotation at all! So, if the molecule is linear, only two coordinates are required to describe all of the rotational motions, and
there are 3N − 5 coordinates left over after we allocate those needed to describe the translational and rotational motions.
The coordinates left over after we describe the translational and rotational motions must be used to describe the motion of the
atoms with respect to one another. These motions are called vibrations, and hence the number of coordinates needed to describe the
vibrations of a non-linear molecule is 3N − 6 . For a linear molecule, 3N − 5 coordinates are needed to describe the vibrations.
This page titled 18.4: The Schrödinger Equation for a Molecule is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
We can approximate the wavefunction for a molecule by partitioning it into wavefunctions for individual translational, rotational,
vibrational, and electronic modes. The wavefunctions for each of these modes can be approximated by solutions to a Schrödinger
equation that approximates that mode. Our objective in this chapter is to introduce the quantized energy levels that are found.
Translational modes are approximated by the particle in a box model that we discuss above.
Vibrational modes are approximated by the solutions of the Schrödinger equation for coupled harmonic oscillators. The vibrational
motion of a diatomic molecule is approximated by the solutions of the Schrödinger equation for the vibration of two masses linked
by a spring. Let the distance between the masses be r and the equilibrium distance be r . Let the reduced mass of the molecule be
0
μ , and let the force constant for the spring be λ . From classical mechanics, the potential energy of the system is
2
λ (r − r0 )
V (r) =
2
and the vibrational frequency of the classical oscillator is

−
−
1 λ
ν = √
2π μ
The Schrödinger equation is

2 2 2
h d ψ λ (r − r0 )
−( ) + ψ = Eψ
8π 2 μ dr2 2
The solutions to this equation are wavefunctions and energy levels that constitute the quantum mechanical description of the
classical harmonic oscillator. The energy levels are given by
1
En = hν (n + )
2
where the quantum numbers, n , can have any of the values n = 0, 1, 2, 3, …. The lowest energy level, that for which n =0 ,
has a non-zero energy; that is,
E0 = hν /2
The quantum mechanical oscillator can have infinitely many energies, each of which is a half-integral multiple of the classical
frequency, ν . Each quantum mechanical energy corresponds to a quantum mechanical frequency:
1
νn = ν (n + )
2
A classical rigid rotor consists of two masses that are connected by a weightless rigid rod. The rigid rotor is a dumbbell. The
masses rotate about their center of mass. Each two-dimensional rotational motion of a diatomic molecule is approximated by the
solutions of the Schrödinger equation for the motion of a rigid rotor in a plane. The simplest model assumes that the potential term
is zero for all angles of rotation. Letting I be the molecule’s moment of inertia and φ be the rotation angle, the Schrödinger
equation is
2 2
h d ψ
−( ) = Eψ
2 2
8π I dφ
The energy levels are given by

2 2
m h
Em =
2
8π I
where the quantum numbers, m, can have any of the values m = 1, 2, 3, … . , (but not zero). Each of these energy levels is two-
fold degenerate. That is, two quantum mechanical states of the molecule have the energy E . m
The three-dimensional rotational motion of a diatomic molecule is approximated by the solutions of the Schrödinger equation for
the motion of a rigid rotor in three dimensions. Again, the simplest model assumes that the potential term is zero for all angles of
rotation. Letting θ and φ be the two rotation angles required to describe the orientation in three dimensions, the Schrödinger
equation is
2 2
h 1 ∂ ∂ψ 1 d ψ
− ( (sinθ )+ ) = Eψ
2 2 2
8π I sinθ ∂θ ∂θ sin θ dφ
The energy levels are given by

2
h
EJ = J (J + 1)
2
8π I
where the quantum numbers, J , can have any of the values J = 0, 1, 2 , 3, … . E is (2J + 1) -fold degenerate. That is, there
J
are 2J + 1 quantum mechanical states of the molecule all of which have the same energy, E . J
Equations for the rotational energy levels of larger molecules are more complex.
This page titled 18.5: Solutions to Schroedinger Equations for Harmonic Oscillators and Rigid Rotors is shared under a CC BY-SA 4.0 license
We approximate the wavefunction for a molecule by using a product of approximate wavefunctions, each of which models some
subset of the motions that the molecule undergoes. In general, the wavefunctions that satisfy the molecule’s Schrödinger equation
are degenerate; that is, two or more of these wavefunctions have the same energy. (The one-dimensional particle in a box and the
one-dimensional harmonic oscillator have non-degenerate solutions. The rigid-rotor in a plane has doubly degenerate solutions; two
wavefunctions have the same energy. The J -th energy level of the three-dimensional rigid rotor is (2J + 1) -fold degenerate; there
are (2J + 1) wavefunctions whose energy is E .) We use doubly subscripted symbols to represent the wavefunctions that satisfy
J
the molecule’s Schrödinger equation. We write ψ to represent all of the molecular wavefunctions whose energy is ϵ . We let g be
i,j i i
the number of wavefunctions whose energy is ϵ . We say that the energy level ϵ is g -fold degenerate. The wavefunctions
i i i
ψ , ψ , …, ψ ,…, ψ
i,1 i,2 i,j i,gi
are all solution to the molecule’s Schrödinger equation; we have
Hmolecule ψi,j = ϵi ψi,j
for j = 1, 2, … , g . Every energy level ϵ is associated with g quantum states. For simplicity, we can think of each of the g
i i i i
wavefunctions, ψ , as a quantum state; however, the molecule’s Schrödinger equation is also satisfied by any set of g
i,j i
independent linear combinations of the ψ . For present purposes, all that matters is that there are g quantum-mechanical
i,j i
descriptions—quantum states—all of which have energy ϵ . i
This page titled 18.6: Wave Functions, Quantum States, Energy Levels, and Degeneracies is shared under a CC BY-SA 4.0 license and was
Our goal is to develop the theory of statistical thermodynamics from Boltzmann statistics. In this chapter, we explore the rudiments
of quantum mechanics in order to become familiar with the idea that we can describe a series of discrete energy levels for any
given molecule. For our purposes, that is all we need. We should note, however, that we are not developing the full story about the
relationship between quantum mechanics and statistical thermodynamics. The spin of a particle is an important quantum
mechanical property. It turns out that quantum mechanical solutions depend on the spin of the particle being described. Particles
with integral spins behave differently from particles with half-integral spins. When we treat the statistical distribution of these
particles, we need to treat particles with integral spins differently from particles with half-integral spins. Particles with integral
spins are said to obey Bose-Einstein statistics; particles with half-integral spins obey Fermi-Dirac statistics.
Fortunately, both of these treatments converge to the Boltzmann distribution if the number of quantum states available to the
particles is much larger than the number of particles. For macroscopic systems at ordinary temperatures, this is the case. In
Chapters 19 and 20, we introduce the ideas underlying the theory of statistical mechanics. In Chapter 21, we derive the Boltzmann
distribution from a set of assumptions that does not correspond to either the Bose-Einstein or the Fermi-Dirac requirement. In
Chapter 25, we derive the Bose-Einstein and Fermi-Dirac distributions and show how they become equivalent to the Boltzmann
distribution for most systems of interest in chemistry.
This page titled 18.7: Particle Spins and Statistics- Bose-Einstein and Fermi-Dirac Statistics is shared under a CC BY-SA 4.0 license and was
CHAPTER OVERVIEW
19: The Distribution of Outcomes for Multiple Trials

19.5: Problems
This page titled 19: The Distribution of Outcomes for Multiple Trials is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
1
Suppose that we have two coins, one minted in 2001 and one minted in 2002. Let the probabilities of getting a head and a tail in a
toss of the 2001 coin be P and P , respectively. We assume that these outcomes exhaust the possibilities. From the laws of
H ,1 T ,1
probability, we have: 1 = (P + P ) . For the 2002 coin, we have 1 = (P + P ) . The product of these two probabilities
H ,1 T ,1 H ,2 T ,2
must also be unity. Expanding this product gives

1 = (PH ,1 + PT ,1 ) (PH ,2 + PT ,2 )
= PH ,1 PH ,2 + PH ,1 PT ,2 + PT ,1 PH ,2 + PT ,1 PT ,2
This equation represents the probability of a trial in which we toss the 2001 coin first and the 2002 coin second. The individual
terms are the probabilities of the possible outcomes of such a trial. It is convenient to give a name to this latter representation of the
product; we will call it the expanded representation of the total probability sum.
Our procedure for multiplying two binomials generates a sum of four terms. Each term contains two factors. The first factor comes
from the first binomial; the second term comes from the second binomial. Each of the four terms corresponds to a combination of
an outcome from tossing the 2001 coin and an outcome from tossing the 2002 coin. Conversely, every possible combination of
outcomes from tossing the two coins is represented in the sum. P P represents the probability of getting a head from tossing
H ,1 H ,2
the 2001 coin and a head from tossing the 2002 coin. P P represents the probability of getting a head from tossing the 2001
H ,1 T ,2
coin and a tail from tossing the 2002 coin, etc. In short, there is a one-to-one correspondence between the terms in this sum and the
possible combinations of the outcomes of tossing these two coins.
This analysis depends on our ability to tell the two coins apart. For this, the mint date is sufficient. If we toss the two coins
simultaneously, the four possible outcomes remain the same. Moreover, if we distinguish the result of a first toss from the result of
a second toss, etc., we can generate the same outcomes by using a single coin. If we use a single coin, we can represent the possible
outcomes from two tosses by the ordered sequences H H , H T , T H, and T T , where the first symbol in each sequence is the result
of the first toss and the second symbol is the result of the second toss. The ordered sequences H T and T H differ only in the order
in which the symbols appear. We call such ordered sequences permutations.
Now let us consider a new problem. Suppose that we have two coin-like slugs that we can tell apart because we have scratched a “
1 ” onto the surface of one and a “2 ” onto the surface of the other. Suppose that we also have two cups, one marked “H ” and the
other marked “T .” We want to figure out how many different ways we can put the two slugs into the two cups. We can also
describe this as the problem of finding the number of ways we can assign two distinguishable slugs (objects) to two different cups
(categories). There are four such ways: Cup H contains slugs 1 and 2; Cup H contains slug 1 and Cup T contains slug 2; Cup H
contains slug 2 and Cup T contains slug 1; Cup T contains slugs 1 and 2 .
We note that, given all of the ordered sequences for tossing two coins, we can immediately generate all of the ways that two
distinguishable objects (numbered slugs) can be assigned to two categories (Cups H and T ). For each ordered sequence, we assign
the first object to the category corresponding to the first symbol in the sequence, and we assign the second object to the category
corresponding to the second symbol in the sequence.
In short, there are one-to-one correspondences between the sequences of probability factors in the total probability sum, the
possible outcomes from tossing two distinguishable coins, the possible sequences of outcomes from two tosses of a single coin, and
the number of ways we can assign two distinguishable objects to two categories. (See Table 1.)
Table 1.
Problems Correspondences
Sequences of probability
factors in the total PH ,1 PH ,2 PH ,1 PT ,2 PT ,1 PH ,2 PT ,1 PT ,2
probability sum
Probability factors for

coins distinguished by PH PH PH PT PT PH PT PT
identification numbers
Sequences from toss of a

HH HT TH TT
single coin
Assignments of two
Cup H holds slug 1 and Cup H holds slug 2 and
distinguishable objects to Cup H holds slugs 1 & 2 Cup T holds slugs 1 & 2
Cup T holds slug 2 Cup T holds slug 1
two categories
If the probability of tossing a head is constant, we have P =P =P and P = P = P . Note that we are not assuming
H ,1 H ,2 H T ,1 T ,2 T
PH = P . If we do not care about the order in which the heads and tails appear, we can simplify our equation for the product of
T
probabilities to
2 2
1 =P + 2 PH PT + P
H T
P
H
2
is the probability of tossing two heads, P P is the probability of tossing one head and one tail, and P is the probability of
H T
2
T
tossing two tails. We must multiply the P P -term by two, because there are two two-coin outcomes and correspondingly two
H T
combinations, P P and P P , that have the same probability, P P . Completely equivalently, we can say that the reason
H ,1 T ,2 T ,1 H ,2 H T
for multiplying the P P -term by two is that there are two permutations, H T and T H, which correspond to one head and one tail
H T
in successive tosses of a single coin.

We have lavished considerable attention on four related but very simple problems. Now, we want to extend this analysis—first to
tosses of multiple coins and then to situations in which multiple outcomes are possible for each of many independent events.
Eventually we will find that understanding these problems enables us to build a model for the behavior of molecules that explains
the observations of classical thermodynamics.
If we extend our analysis to tossing n coins, which we label coins 1, 2, etc., we find:
1 = (PH ,1 + PT ,1 ) (PH ,2 + PT ,2 ) … (PH ,n + PT ,n )
= (PH ,1 PH ,2 … PH ,n ) + (PH ,1 PH ,2 … PH ,i … PT ,n ) + ⋯ + (PT ,1 PT ,2 … PT ,i … PT ,n )
We write each of the product terms in this expanded representation of the total-probability sum with the second index, r, increasing
from 1 to n as we read through the factors, P , from left to right. Just as for tossing only two coins:
X,r
1. Each product term is a sequence of probability factors that appears in the total probability sum.
2. Each product term corresponds to a possible outcome from tossing n coins that are distinguished from one another by
identification numbers.
3. Each product term is equivalent to a possible outcome from repeated tosses of a single coin: the r factor is P or P th
H T
according as the r toss produces a head or a tail.

th
4. Each product term is equivalent to a possible assignment of n distinguishable objects to the two categories H and T .
In Section 3.9, we introduce the term population set to denote a set of numbers that represents a possible combination of outcomes.
Here the possible combinations of outcomes are the numbers of heads and tails. If in five tosses we obtain 3 heads and 2 tails, we
say that this group of outcomes belongs to the population set {3, 2}. If in n tosses, we obtain n heads and n tails, this group of H T
outcomes belongs to the population set {n , n }. For five tosses, the possible population sets are {5, 0}, {4, 1}, {3, 2}, {2, 3},
H T
{1, 4}, and {5, 0}. Beginning in the next chapter, we focus on the energy levels that are available to a set of particles and on the
number of particles that has each of the available energies. Then the number of particles, N , that have energy ϵ is the population i i
of the ϵ -energy level. The set of all such numbers is the energy-level population set for the set of particles.
i
If we cannot distinguish one coin from another, the sequence P P P P becomes P P P P . We say that P P P P H ,1 T ,2 H ,3 H ,4 H T H H H T H H
is distinguishable from P P P P because the tails-outcome appears in the second position in P P P P and in the third
H H T H H T H H
position in P P P P . We say that P P P P

H H T H and P P P P are indistinguishable, because both become
H ,1 T ,2 H ,3 H ,4 H ,3 T ,2 H ,1 H ,4
P P P P . In general, many terms in the expanded form of the total probability sum belong to the population set corresponding
H T H H
to n heads and n tails. Each such term corresponds to a distinguishable permutation of n heads and n tails and the
H T H T
corresponding distinguishable permutation of P and P terms. H T
We use the notation C (n , n ) to denote the number of terms in the expanded form of the total probability sum in which there are
H T
nH heads and n tails. C (n , n ) is also the number of distinguishable permutations of n heads and n tails or of n P -
T H T H T H H
terms and n P -terms. The principal goal of our analysis is to find a general formula for C (n , n ). To do so, we make use of
T T H T
the fact that C (n , n ) is also the number of ways that we can assign n objects (coins) to two categories (heads or tails) in such a
H T
way that n objects are in one category (heads) and n objects are in the other category (tails). We also call C (n , n ) the
H T H T
number of combinations possible for distinguishable coins in the population set {n , n }. H T
The importance of C (n , n ) is evident when we recognize that, if we do not care about the sequence (permutation) in which a
H T
particular number of heads and tails occurs, we can represent the total-probability sum in a much compressed form:
n n−1 nH nT n−1 n
1 =P + nP PT + ⋯ + C (nH , nT ) P P + nPH P +P
H H H T T T
In this representation, there are n terms in the total-probability sum that have n H = n−1 and n T =1 . These are the terms
PH ,1 PH ,2 PH ,3 … PH ,i … PH ,n−1 PT ,n
PH ,1 PH ,2 PH ,3 … PH ,i … PT ,n−1 PH ,n
PH ,1 PH ,2 PH ,3 … PT ,i … PH ,n−1 PH ,n
PH ,1 PH ,2 PT ,3 … PH ,i … PH ,n−1 PH ,n
PH ,1 PT ,2 PH ,3 … PH ,i … PH ,n−1 PH ,n
PT ,1 PH ,2 PH ,3 … PH ,i … PH ,n−1 PH ,n
Each of these terms represents the probability that n − 1 heads and one tail will occur in the order shown. Each of these terms has
the same value. Each of these terms is a distinguishable permutation of n − 1 P terms and one P term. Each of these terms
H T
corresponds to a combination in which one of n numbered slugs is assigned to Cup T , while the remaining n − 1 numbered slugs
are assigned to Cup H . It is easy to see that there are n such terms, because each term is the product of n probabilities, and the tail
can occur at any of the n positions in the product. If we do not care about the order in which heads and tails occur and are
interested only in the value of the sum of these n terms, we can replace these n terms by the one term nP P . We see that
n−1
H T
nP
n−1
H
P Tis the probability of tossing n − 1 heads and one tail, irrespective of which toss produces the tail.
There is another way to show that there must be n terms in the total-probability sum in which there are n − 1 heads and one tail.
This method relies on the fact that the number of such terms is the same as the number of combinations in which n distinguishable
things are assigned to two categories, with n − 1 of the things in one category and the remaining thing in the other category,
C (n − 1, 1) . This method is a little more complicated, but it offers the great advantage that it can be generalized.
The new method requires that we think about all of the permutations we can create by reordering the results from any particular
series of n tosses. To see what we have in mind when we say all of the permutations, let P represent the probability of toss
X,k
number k , where for the moment we do not care whether the outcome was a head or a tail. When we say all of the permutations,
we mean the number of different ways we can order (permute) n different values P . It is important to recognize that one and
X,k
only one of these permutations is a term in the total-probability sum, specifically:
PX,1 PX,2 PX,3 … PX,k … PX,n
in which the values of the second subscript are in numerical order. When we set out to construct all of these permutations, we see
that there are n ways to choose the toss to put first and n − 1 ways to choose the toss to put second, so there are n (n − 1) ways to
choose the first two tosses. There are n − 2 ways to choose the third toss, so there are n (n − 1) (n − 2) ways to choose the first
three tosses. Continuing in this way through all n tosses, we see that the total number of ways to order the results of n tosses is
n (n − 1) (n − 2) (n − 3) … (3) (2) (1) = n!
Next, we need to think about the number of ways we can permute n values P if n − 1 of them are P , P ,…, P
X,k , …, H ,1 H ,2 H ,r−1
PH ,r+1 ,…,P H ,n and one of them is P , and we always keep the one factor P
T ,r in the same position. By the argument above,
T ,r
there are (n − 1)! ways to permute the values P in a set containing n − 1 members. So for every term (product of factors P )
H ,s X,k
that occurs in the total-probability sum, there are (n − 1)! other products (other permutations of the same factors) that differ only
in the order in which the P H ,sappear. The single tail outcome occupies the same position in each of these permutations. If the r th
factor in the term in the total probability sum is P , then P is the r factor in each of the (n − 1)! permutations of this term.
T ,r T ,r
th
This is an important point, let us repeat it in slightly different words: For every term that occurs in the total-probability sum, there
are (n − 1)! permutations of the same factors that leave the heads positions occupied by heads and the tails position occupied by
tails.
Equivalently, for every assignment of n − 1 distinguishable objects to one of two categories, there are (n − 1)! permutations of
these objects. There are C (n − 1, 1) such assignments. Accordingly, there are a total of (n − 1)!C (n − 1, 1) permutations of the
n distinguishable objects. Since we also know that the total number of permutations of n distinguishable objects is n!, we have
n! = (n − 1)!C (n − 1, 1)
so that
n!
C (n − 1, 1) =
(n − 1)!
which is the same result that we obtained by our first and more obvious method.
The distinguishable objects within a category in a particular assignment can be permuted. We give these within-category
permutations another name; we call them indistinguishable permutations. (This terminology reflects our intended application,
which is to find the number of ways n identical molecules can be assigned to a set of energy levels. We can tell two isolated
molecules of the same substance apart only if they have different energies. We can distinguish molecules in different energy levels
from one another. We cannot distinguish two molecules in the same energy level from one another. Two different permutations of
the molecules within any one energy level are indistinguishable from one another.) For every term in the expanded representation
of the total probability sum, indistinguishable permutations can be obtained by exchanging P factors with one another, or by H
exchanging P factors with one another, but not by exchanging P factors with P factors. That is, heads are exchanged with
T H T
heads; tails are exchanged with tails; but heads are not exchanged with tails.
Now we can consider the general case. We let C (n , n H T ) be the number of terms in the total-probability sum in which there are
nH heads and n tails. We want to find the value of C (n
T H , nT ) . Let’s suppose that one of the terms with n heads and n tails is H T
(PH ,a PH ,b … PH ,m ) (PT ,r PT ,s … PT ,z )
where there are n indices in the set {a, b, … , m} and n indices in the set {r, s, … , z}. There are n ! ways to order the
H T H
heads outcomes and n ! ways to order the tails outcomes. So, there are n !n ! possible ways to order n heads and n tails
T H T H T
outcomes. This is true for any sequence in which there are n heads and n tails; there will always be n !n ! permutations of n
H T H T H
heads and n tails, whatever the order in which the heads and tails appear. This is also true for every term in the total-probability
T
sum that contains n heads factors and n tails factors. The number of such terms is C (n , n ). For every such term, there are
H T H T
n ! n ! permutations of the same factors that leave the heads positions occupied by heads and the tails positions occupied by tails.
H T
Accordingly, there are a total of n !n !C (n , n

H T H T ) permutations of the n distinguishable objects. The total number of
permutations of n distinguishable objects is n!, so that
n! = nH ! nT !C (nH , nT )
and
n!
C (nH , nT ) =
nH ! nT !
Equivalently, we can construct a sum of terms, R , in which the terms are all of the n! permutations of PH ,i factors for nH heads
and P factors for n tails. The value of each term in R is P P . So we have
T ,j T
nH
H
nT
nH nT
R = n!P P
H T
R contains all C (n , n ) of the P P -valued terms that appear in the total-probability sum. For each of these P P -
H T
nH
H
nT
T H
nH
T
nT
valued terms there are n !n ! indistinguishable permutations that leave heads positions occupied by heads and tails positions
H T
occupied by tails. R will also contain all of the n !n ! permutations of each of these P P -valued terms. That is, every term in
H T H
nH nT
R is either a term in the expanded representation of the total probability sum or an indistinguishable permutation of such a term. It
follows that R is also given by

nH nT
R = nH ! nT !C (nH , nT ) P P
H T
Equating these equations for R, we have

nH nT nH nT
n! P P = nH ! nT !C (nH , nT ) P P
H T H T
and, again,
n!
C (nH , nT ) =
nH ! nT !
In summary: The total number of permutations is n! The number of combinations of n distinguishable things in which n of them H
are assigned to category H and n = n − n are assigned to category T is C (n , n ). (Every combination is a distinguishable
T H H T
permutation.) The number of indistinguishable permutations of the objects in each such combination is n !n !. The relationship H T
among these quantities is

total number of permutations = (number of distinguishable combinations) × (number of indistinguishable permutations for
each distinguishable combination)
We noted earlier that C (n , n ) is the formula for the binomial coefficients. If we do not care about the order in which the heads
H T
and tails arise, the probability of tossing n tails and n = n − n heads is

T H T
nH nT
n! nH nT
C (nH , nT ) P P =( )P P
H T H T
nH ! nT !
and the sum of such terms for all n + 1 possible values of n in the interval 0 ≤ n ≤ n is the total probability for all possible
T T
outcomes from n tosses of a coin. This total probability must be unity. That is, we have
n n
n nH nT
n! nH nT
1 = (PH + PT ) = ∑ C (nH , nT ) P P = ∑ ( )P P
H T H T
nH ! nT !
nT =0 nT =0
n
For an unbiased coin, P = P = 1/2 , and P P = (1/2) , for all n . This means that the probability of tossing n heads
H T H
nH
T
nT
T H
n
and n tails is proportional to C (n , n ) where the proportionality constant is (1/2) . The probability of n heads and n − n
T H T
■ ■
tails is the same as the probability of n − n heads and n tails.

■ ■
Nothing in our development of the equation for the total probability requires that we set P = P , and in fact, the binomial H T
probability relationship applies to any situation in which there are repeated trials, where each trial has two possible outcomes, and
where the probability of each outcome is constant. If P ≠ P , the symmetry observed for tossing coins does not apply, because
H T
■ ■ ■ ■
n−n n n n−n
P P ≠P P
H T H T
This condition corresponds to a biased coin.

Another example is provided by a spinner mounted at the center of a circle painted on a horizontal surface. Suppose that a pie-
shaped section accounting for 25% of the circle’s area is painted white and the rest is painted black. If the spinner’s stopping point
is unbiased, it will stop in the white zone with probability P = 0.25 and in the black zone with probability P = 0.75. After n
W B
spins, the probability of n white outcomes and n black outcomes is

W B
n! nW nB
( ) (0.25) (0.75)
nW ! nB !
After n spins, the sum of the probabilities for all possible combinations of white and black outcomes is
n
n nW nB
1 = (PW + PB ) = ∑ C (nW , nB ) P P
W B
nB =0
n
n! nW nB
= ∑ ( )P P
W B
nW ! nB !
nB =0
n
n! nW nB
= ∑( ) (0.25) (0.75)
nW ! nB !
nB=0
This page titled 19.1: Distribution of Results for Multiple Trials with Two Possible Outcomes is shared under a CC BY-SA 4.0 license and was
Let us extend the ideas we have developed for binomial probabilities to the case where there are three possible outcomes for any
given trial. To be specific, suppose we have a coin-sized object in the shape of a truncated right-circular cone, whose circular faces
are parallel to each other. The circular faces have different diameters. When we toss such an object, allowing it to land on a smooth
hard surface, it can wind up resting on the big circular face (H eads), the small circular face (T ails), or on the conical surface (
C one-side). Let the probabilities of these outcomes in a single toss be P , P , and P , respectively. In general, we expect these
H T C
probabilities to be different from one another; although, of course, we require 1 = (P + P + P ) . H T C
Figure 1: A truncated right-circular cone (CC BY-SA Unported; Dirk Hünniger via Wikipedia).
Following our development for the binomial case, we want to write an equation for the total probability sum after n tosses. Let n , H
n , and n
T be the number of H , T , and C outcomes exhibited in n + n + n = n trials. We let the probability coefficients be
C H T C
C (n , n , n ). The probability of n , n , n
H T C outcomes in n trials is
H T C
nH nT nC
C (nH , nT , nC ) P P P
H T C
and the total probability is

n nH nT nC
1 = (PH + PT + PC ) = ∑ C (nH , nT , nC ) P P P
H T C
nH , nT , nC
where the summation is to be carried out over all combinations of integer values for nH , nT , and nC , consistent with
nH +n +n T=n . C
To find C (n H , we proceed as before. We suppose that one of the terms with n heads, n tails, and n cone-sides is
, nT , nC ) H T C
(PH ,a PH ,b … PH ,f ) (PT ,g PT ,h … PT ,m ) (PC ,p PC ,q … PC ,z )
where there are n indices in the set {a, b, … , f } , n indices in the set {g, h, … , m} , and n indices in the set {p, q,…, z
H T C
} . There are n ! ways to order the heads outcomes, n ! ways to order the tails outcomes, and n ! ways to order the cone-sides
H T C
outcomes. So, there are n !n !n ! possible ways to order n heads, n tails, and n cone-sides. There will also be n !n !n !
H T C H T C H T C
indistinguishable permutations of any combination (particular assignment) of n heads, n tails, and n cone-sides. There are n!
H T C
possible permutations of n probability factors and C (n , n , n ) distinguishable combinations with n heads, n tails, and n
H T C H T C
cone-sides. As before, we have

so that
n! = nH ! nT ! nC !C (nH , nT , nC )
and hence,
n!
C (nH , nT , nC ) =
nH ! nT ! nC !
Equivalently, we can construct a sum of terms, S , in which the terms are all of the n! permutations of P factors for n heads, H ,r H
P factors for n tails, and P factors for n cone-sides. The value of each term in S will be P P P . Thus, we have
T ,s T C ,t C H
nH nT
T
nC
nH nT nC
S = n!P P P
H T C
S will contain all C (n , n , n ) of the distinguishable combinations n heads, n tails, and n cone-sides outcomes that give
H T C H T C
rise to P P P -valued terms. Moreover, S will also include all of the n !n !n ! indistinguishable permutations of each of
H
nH
T
nT
C
nC
H T C
these P P P -valued terms, and we also have

nH
H
nT
T
nC
nH nT nC
S = nH ! nT ! nC !C (nH , nT , nC ) P P P
H T C
Equating these two expressions for S gives us the number of P

H
nH
P
T
nT
P
nC
C
-valued terms in the total-probability product,
C (n , n , n ). That is,
H T C
nH nT nC nH nT nC
S = n! P P P = nH ! nT ! nC !C (nH , nT , nC ) P P P
H T C H T C
and, again,
n!
C (nH , nT , nC ) =
nH ! nT ! nC !
In the special case that P H = PT = PC = 1/3 , all of the products P P P will have the value (1/3) . Then the probability
nH
H T
nT
C
nC n
n
of any set of outcomes, {n H , nT , nC , }, is proportional to C (n , n , n ) with the proportionality constant (1/3) .
H T C
This page titled 19.2: Distribution of Results for Multiple Trials with Three Possible Outcomes is shared under a CC BY-SA 4.0 license and was
It is now easy to extend our results to multiple trials with any number of outcomes. Let the outcomes be A , B , C , …., Z , for which
the probabilities in a single trial are P , P , P ,…P . We again want to write an equation for the total probability after n trials.
A B C Z
We let n , n , n ,…n be the number of A , B , C ,…, Z outcomes exhibited in n + n + n +. . . +n = n trials. If we do

A B C Z A B C Z
not care about the order in which the outcomes are obtained, the probability of n , n , n ,…, n outcomes in n trials is A B C Z
nA nB nC nZ
C (nA , nB , nC , … , nZ ) P P P …P
A B C Z
and the total probability sum is

n nA nB nC nZ
1 = (PA + PB + PC + ⋯ + PZ ) = ∑ C (nA , nB , nC , … , nZ ) P P P …P
A B C Z
nI
where the summation is to be carried out over all combinations of integer values for nA , nB , nC ,…, nZ consistent with
n +nA + n +. . . +n
B =n .
C Z
Let one of the terms for n A A -outcomes, n B B -outcomes, n C C -outcomes, …, n Z Z -outcomes, be
(PA,a PA,b … PA,f ) (PB,g PB,h … PB,m ) × (PC ,p PC ,q … PC ,t ) … (PZ,u PZ,v … PZ,z )
where there are n indices in the set {a, b, … , f } , n indices in the set {g, h, … , m} , n indices in the set {p, q, … , t},
A B C
…, and n indices in the set {u, v, … , z} . There are n ! ways to order the A -outcomes, n ! ways to order the B -outcomes,
Z A B
n ! ways to order the C -outcomes, …, and n ! ways to order the Z -outcomes. So, there are n ! n ! n ! … n ! ways to order n
C Z A B C Z A
A -outcomes, n B -outcomes, n
B C -outcomes, …, and n Z -outcomes. The same is true for any other distinguishable
C Z
combination; for every distinguishable combination belonging to the population set {n , n , n ,…, n } there are A B C Z
n ! n ! n ! … n ! indistinguishable permutations. Again, we can express this result as the general relationship:
A B C Z
so that
n! = nA ! nB ! nC ! … nZ !C (nA , nB , nC , … , nZ )
and
n!
C (nA , nB , nC , … , nZ ) =
nA ! nB ! nC ! … nZ !
Equivalently, we can construct a sum, T , in which we add up all of the n! permutations of P factors for n A -outcomes, P A,a A B,b
factors for n B -outcomes, P factors for n C -outcomes, …, and P factors for n Z -outcomes. The value of each term in
B C ,c C Z,z Z
T will be P . So we have
nA nB nC nZ
P P …P
A B C Z
nA nB nC nZ
T = n! P P P …P
A B C Z
nA nC
T will contain all C (n , n , n , … , n ) of the P P P … P -valued products (distinguishable combinations) that are a
A B C Z A B
nB
C Z
nZ
part of the total-probability sum. Moreover, T will also include all of the n !n !n ! … n ! indistinguishable permutations of A B C Z
each of these P P P … P -valued products. Then we also have

nA
A
nB
B C
nC
Z
nZ
T = nA ! nB ! nC ! … nZ !C (nA , nB , nC , … , nZ )
nA nB nC nZ
×P P P …P
A B C Z
nA nC
Equating these two expressions for T gives us the number of P A
P
B
nB
P
C
…P
Z
nZ
-valued products
nA nB nC nZ
n! P P P …P = nA ! nB ! nC ! … nZ !
A B C Z
nA nB nC nZ
×C (nA , nB , nC , … , nZ ) P P P …P
A B C Z
and hence,
n!
C (nA , nB , nC , … , nZ ) =
nA ! nB ! nC ! … nZ !
In the special case that P = P = P = ⋯ = P , all of the products P P

A B C Z
A
nA nB
B
P
C
nC
…P
Z
nZ
have the same value. Then, the
probability of any set of outcomes, {n , n , n , … , n }, is proportional to C (n
A B C Z A, nB , nC , … , nZ ) .
This page titled 19.3: Distribution of Results for Multiple Trials with Many Possible Outcomes is shared under a CC BY-SA 4.0 license and was
The polynomial coefficient, C , is a function of the factorials of large numbers. Since N! quickly becomes very large as N
increases, it is often impractical to evaluate N ! from the definition,
N ! = (N ) (N − 1) (N − 2) … (3) (2) (1)
Fortunately, an approximation, known as Stirling’s formula or Stirling’s approximation is available. Stirling’s approximation is a
product of factors. Depending on the application and the required accuracy, one or two of these factors can often be taken as unity.
Stirling’s approximation is
1/2
1 1/2
N N N
N! ≈ N (2πN ) exp (−N ) exp ( ) ≈N (2πN ) exp (−N ) ≈ N exp (−N )
12N
In many statistical thermodynamic arguments, the important quantity is the natural logarithm of N ! or its derivative, d ln N ! /dN .
In such cases, the last version of Stirling’s approximation is usually adequate, even though it affords a rather poor approximation
for N ! itself.
This page titled 19.4: Stirling's Approximation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
19.5: Problems
1. Leland got a train set for Christmas. It came with seven rail cars. (We say that all seven cars are “distinguishable.”) Four of the
rail cars are box cars and three are tank cars. If we distinguish between permutations in which the box cars are coupled (lined up)
differently but not between permutations in which tank cars are coupled differently, how many ways can the seven cars be coupled
so that all of the tank cars are together? What are they? What formula can we use to compute this number?
(Hint: We can represent one of the possibilities as b b b b T . This is one of the possibilities in which the first four cars behind the
1 2 3 4
engine are all box cars. There are 4! such possibilities; that is, there are 4! possible permutations for placing the four box cars.)
2. If we don’t care about the order in which the box cars are coupled, and we don’t care about the order in which the tank cars are
coupled, how many ways can the rail cars in problem 1 be coupled so that all of the tank cars are together? What are they? What
formula can we use to compute this number?
3. If we distinguish between permutations in which either the box cars or the tank cars in problem 1 are ordered differently, how
many ways can the rail cars be coupled so that all of the tank cars are together? What formula can we use to compute this number?
4. How many ways can all seven rail cars in problem 1 be coupled if the tank cars need not be together?
5. If, as in the previous problem, we distinguish between permutations in which any of the rail cars are ordered differently, how
many ways can the rail cars be coupled so that not all of the tank cars are together?
6. If we distinguish between box cars and tank cars, but we do not distinguish one box car from another box car, and we do not
distinguish one tank car from another tank car, how many ways can the rail cars in problem 1 be coupled?
7. If Leland gets five flat cars for his birthday, he will have four box cars, three tank cars and five flat cars. How many ways will
Leland be able to couple (permute) these twelve rail cars?
8. If we distinguish between box cars and tank cars, between box cars and flat cars, and between tank cars and flat cars, but we do
not distinguish one box car from another box car, and we do not distinguish one tank car from another tank car, and we do not
distinguish one flat car from another flat car, how many ways can the rail cars in problem seven be coupled? What formula can we
use to compute this number?
9. We are given four distinguishable marbles, labeled A − −D , and two cups, labeled 1 and 2. We want to explore the number of
ways we can put two marbles in cup 1 and two marbles in cup 2. This is the number of combinations, C (2, 2), for the population
set N = 2 , N = 2 .
1 2
(a) One combination is [AB] 1

[C D]
2
. Find the remaining combinations. What is C (2, 2)?
(b) There are four permutations for the combination given in (a): [AB] 1
[C D]
2
; [BA]
1
[C D]
2
; [AB]1
[DC ]
2
; [BA]
1
[DC ]
2
. Find all
of the permutations for each of the remaining combinations.
(c) How many permutations are there for each combination?
(d) Write down all of the possible permutations of marbles A − −D . Show that there is a one-to-one correspondence with the
permutations in (b).
(e) Show that the total number of permutations is equal to the number of combinations times the number of permutations possible
for each combination.
10. We are given seven distinguishable marbles, labeled A − −G , and two cups, labeled 1 and 2. We want to find the number of
ways we can put three marbles in cup 1 and four marbles in cup 2 . That is, we seek C (3, 4), the number of combinations in which
N = 3 and N = 4 . [ABC ] [DEF G] is one such combination.
1 2 1 2
(a) How many different ways can these marbles be placed in different orders without exchanging any marbles between cup 1 and
cup 2? (This is the number of permutations associated with this combination.)
(b) Find a different combination with N 1 =3 and N 2 =4 .
(c) How many permutations are possible for the marbles in (b)? How many permutations are possible for any combination with
N = 3 and N = 4 ?
1 2
(d) If C (3, 4) is the number of combinations in which N = 3 and N = 4 , and if P is the number of permutations for each such
1 2
combination, what is the total number of permutations possible for 7 marbles?
(e) How else can one express the number of permutations possible for 7 marbles?
(f) Equate your conclusions in (d) and (e). Find C (3, 4).
11.
(a) Calculate the probabilities of 0, 1, 2, 3, and 4 heads in a series of four tosses of an unbiased coin. The event of 2 heads is 20%
of these five events. Note particularly the probability of the event: 2 heads in 4 tosses.
(b) Calculate the probabilities of 0, 1, 2, 3,…, 8, and 9 heads in a series of nine tosses of an unbiased coin. The events of 4 heads
and 5 heads comprise 20% of these ten cases. Calculate the probability of 4 heads or 5 heads; i.e., the probability of being in the
middle 20% of the possible events.
(c) Calculate the probabilities of 0, 1, 2, 3,…, 13, and 14 heads in a series of fourteen tosses of an unbiased coin. The events of 6
heads, 7 heads, and 8 heads comprise 20% of these fifteen cases. Calculate the probability of 6, 7, or 8 heads; i.e., the probability of
being in the middle 20% of the possible events.
(d) What happens to the probabilities for the middle 20% of possible events as the number of tosses becomes very large? How does
this relate to the fraction heads in a series of tosses when the total number of tosses becomes very large?
12. Let the value of the outcome heads be one and the value of the outcome tails be zero. Let the “score” from a particular
simultaneous toss of n coins be
number of heads number of tails
score = 1 × ( ) +0 ×( )
number of coins number of coins
Let us refer to the distribution of scores from tosses of n coins as the “S distribution.”
n
(a) The S distribution comprises two outcomes: {1 head, 0 tail} and {0 head, 1 tail}.
1
What is the mean of the S distribution?

1
(b) What is the variance of the S distribution?

1
(c) What is the mean of the S distribution?

n
(d) What is the variance of the S distribution?

n
13. Fifty unbiased coins are tossed simultaneously.

(a) Calculate the probability of 25 heads and 25 tails.
(b) Calculate the probability of 23 heads and 27 tails.
(c) Calculate the probability of 3 heads and 47 tails.
(d) Calculate the ratio of your results for parts (a) and (b).
(e) Calculate the ratio of your results for parts (a) and (c).
14. For N = 3, 6 and 10, calculate
(a) The exact value of N !
(b) The value of N ! according to the approximation
1/2
1
N
N! ≈ N (2πN ) exp (−N ) exp ( )
12N
(c) The value of N! according to the approximation

N 1/2
N! ≈ N (2πN ) exp (−N )
(d) The value of N! according to the approximation

N
N! ≈ N exp (−N )
(e) The ratio of the value in (b) to the corresponding value in (a).
(f) The ratio of the value in (c) to the corresponding value in (a).
(g) The ratio of the value in (d) to the corresponding value in (a).
(h) Comment.
15. Find , d ln N !/dN using each of the approximations
N 1/2 1 N 1/2 N
N! ≈ N (2πN ) exp (−N ) exp ( ) ≈N (2πN ) exp (−N ) ≈ N exp (−N )
12N
How do the resulting approximations for d ln N !/dN compare to one another as N becomes very large?
16. There are three energy levels available to any one molecule in a crystal of the substance. Consider a crystal containing 1000
molecules. These molecules are distinguishable because each occupies a unique site in the crystalline lattice. How many
combinations (microstates) are associated with the population set N = 800 , N = 150 , N = 50 ?
1 2 3
CHAPTER OVERVIEW
20: Boltzmann Statistics

20.15: Problems
This page titled 20: Boltzmann Statistics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
1
In Chapter 18, our survey of quantum mechanics introduces the idea that a molecule can have any of an infinite number of discrete
energies, which we can put in order starting with the smallest. We now turn our attention to the properties of a system composed of
a large number of molecules. This multi-molecule system must obey the laws of quantum mechanics. Therefore, there exists a
Schrödinger equation, whose variables include all of the inter-nucleus, inter-electron, and electron-nucleus distance and potential
terms in the entire multi-molecule system. The relevant boundary conditions apply at the physical boundaries of the macroscopic
system. The solutions of this equation include a set of infinitely many wavefunctions, Ψ , each describing a quantum mechanical
i,j
state of the entire multi-molecule system. In general, the collection of elementary particles that can be assembled into a particular
multi-molecule system can also be assembled into many other multi-molecule systems. For example, an equimolar mixture of C O
and H O can be reassembled into a system comprised of equimolar C O and H , or into many other systems containing mixtures
2 2 2
of C O, H O, C O , and H . Infinitely many quantum-mechanical states are available to each of these multi-molecule systems.
2 2 2
For every such multi-molecule wavefunction, ψ , there is a corresponding system energy, E . In general, the system energy, E , is
i,j i i
Ω -fold degenerate; there are Ω wavefunctions, Ψ

i i , Ψ , …, Ψ , whose energy is E . The wavefunctions include all of the
i,1 i,2 i,Ωi i
interactions among the molecules of the system, and the energy levels of the system reflect all of these interactions. While
generating and solving this multi-molecule Schrödinger equation is straightforward in principle, it is completely impossible in
practice.
Fortunately, we can model multi-molecule systems in another way. The primary focus of chemistry is the study of the properties
and reactions of molecules. Indeed, the science of chemistry exists, as we know it, only because the atoms comprising a molecule
stick together more tenaciously than molecules stick to one another. (Where this is not true, we get macromolecular materials like
metals, crystalline salts, etc.) This occurs because the energies that characterize the interactions of atoms within a molecule are
much greater than the energies that characterize the interaction of one molecule with another. Consequently, the energy of the
system can be viewed as the sum of two terms. One term is a sum of the energies that the component molecules would have if they
were all infinitely far apart. The other term is a sum of the energies of all of the intermolecular interactions, which is the energy
change that would occur if the molecules were brought from a state of infinite separation to the state of interest.
In principle, we can describe a multi-molecule system in this way with complete accuracy. This description has the advantage that
it breaks a very large and complex problem into two smaller problems, one of which we have already solved: In Chapter 18, we see
that we can approximate the quantum-mechanical description of a molecule and its energy levels by factoring molecular motions
into translational, rotational, vibrational, and electronic components. It remains only to describe the intermolecular interactions.
When intramolecular energies are much greater than intermolecular-interaction energies, it may be a good approximation to ignore
the intermolecular interactions altogether. This occurs when we describe ideal gas molecules; in the limit that a gas behaves ideally,
the force between any two of its molecules is nil.
In Chapter 23, we return to the idea of multi-molecule wavefunctions and energy levels. Meanwhile we assume that intermolecular
interactions can be ignored. This is a poor approximation for many systems. However, it is a good approximation for many others,
and it enables us to keep our description of the system simple while we use molecular properties in our development of the
essential ideas of statistical thermodynamics.
We focus on developing a theory that gives the macroscopic thermodynamic properties of a pure substance in terms of the energy
levels available to its individual molecules. To begin, we suppose that we solve the Schrödinger equation for an isolated molecule.
In this Schrödinger equation, the variables include the inter-nucleus, inter-electron, and electron-nucleus distance and potential
terms that are necessary to describe the molecule. The solutions are a set of infinitely many wavefunctions, ψ , each describing a
i,j
different quantum-mechanical state of an isolated molecule. We refer to each of the possible wavefunctions as quantum state of the
molecule. For every such wavefunction, there is a corresponding molecular energy, ϵ . Every unique molecular energy, ϵ , is called
i i
an energy level. Several quantum states can have the same energy. When two or more quantum states have the same energy, we say
that they belong to the same energy level, and the energy level is said to be degenerate. In general, there are g quantum states that
i
we can represent by the g wavefunctions, ψ , , ψ , ..., , ψ , each of whose energy is ϵ . The number of quantum states that
i i,1 i,2 i,g
i
i
have the same energy is called the degeneracy of the energy level. Figure 1 illustrates the terms we use to describe the quantum
states and energy levels available to a molecule.
Figure 1. Quantum states and degenerate energy levels.
In our development of classical thermodynamics, we find it convenient to express the value of a thermodynamic property of a pure
substance as the change that occurs during a formal process that forms one mole of the substance, in its standard state, from its
unmixed constituent elements, in their standard states. In developing statistical thermodynamics, we find it convenient to express
the value of a molecular energy, ϵ , as the change that occurs during a formal process that forms a molecule of the substance, in one
i
of its quantum states, ψ , from its infinitely separated, stationary, constituent atoms. That is, we let the isolated constituent atoms
i,j
be the reference state for the thermodynamic properties of a pure substance.
This page titled 20.1: The Independent-Molecule Approximation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
If only pressure–volume work is possible, the state of a closed, reversible system can be specified by specifying its volume and
temperature. Since the system is closed, the number, N , of molecules is constant. Let us consider a closed, equilibrated, constant-
volume, constant-temperature system in which the total number of molecules is very large. Let us imagine that we can monitor the
quantum state of one particular molecule over a very long time. Eventually, we are able to calculate the fraction of the elapsed time
that the molecule spends in each of the quantum states. We label the available quantum states with the wavefunction symbols, ψ . i,j
We assume that the fraction of the time that a molecule spends in the quantum state ψ is the same thing as the probability of
i,j
finding the molecule in quantum state ψ . We denote this probability as ρ (ψ ). To develop the theory of statistical
i,j i,j
thermodynamics, we assume that this probability depends on the energy, and only on the energy, of the quantum state ψ . i,j
Consequently, any two quantum states whose energies are the same have the same probability, and the g -fold degenerate quantum i
states, ψ , whose energies are ϵ , all have the same probability. In our imaginary monitoring of the state of a particular molecule,
i,j i
we observe that the probabilities of two quantum states are the same if and only if their energies are the same; that is, we observe
ρ (ψ ) = ρ (ψ
i,j
) if and only if i = k .
k,m
The justification for this assumption is that the resulting theory successfully models experimental observations. We can ask,
however, why we might be led to make this assumption in the first place. We can reason as follows: The fact that we observe a
definite value for the energy of the macroscopic system implies that quantum states whose energies are much greater than the
average molecular energy must be less probable than quantum states whose energies are smaller. Otherwise, the sum of the energies
of high-energy molecules would exceed the energy of the system. Therefore, we can reasonably infer that the probability of a
quantum state depends on its energy. On the other hand, we can think of no plausible reason for a given molecule to prefer one
quantum state to another quantum state that has the same energy.
This assumption means that a single function suffices to specify the probability of finding a given molecule in any quantum state,
ψi,j
, and the only independent variable is the quantum-state energy, ϵ . We denote the probability of a single quantum state, ψ ,
i i,j
whose energy is ϵ , as ρ (ϵ ). Since this is the probability of each of the g -fold degenerate quantum states, ψ , that have energy ϵ ,
i i i i,j i
the probability of finding a given molecule in any energy level, ϵ , is P (ϵ ) = g ρ (ϵ ) . We find it convenient to introduce “P ” to
i i i i i
abbreviate this probability; that is, we let

gi
Pi = ∑ ρ (ψi,j ) = P (ϵi ) = gi ρ (ϵi )
j=1
(the probability of energy level ϵ ) i
There is a P for every energy level ϵ . P must be the same for any molecule, since every molecule has the same properties. If the
i i i
population set {N , N , … , N , …} characterizes the equilibrium system, the fraction of the molecules that have energy ϵ is
1 2 i i
N /N . (Elsewhere, an energy-level population set is often called a “distribution.” Since we define a distribution somewhat
i
differently, we avoid this usage.) Since the fraction of the molecules in an energy level at any instant of time is the same as the
fraction of the time that one molecule spends in that energy level, we have
N
i
Pi = P (ϵi ) = gi ρ (ϵi ) =
N
As long as the system is at equilibrium, this fraction is constant. In Chapter 21, we find an explicit equation for the probability
function, ρ (ϵ ). i
The energy levels, ϵ , depend on the properties of the molecules. In developing Boltzmann statistics for non-interacting molecules,
i
we assume that the probability of finding a molecule in a particular energy level is independent of the number of molecules present
in the system. While P and ρ (ϵ ) depend on the energy level, ϵ , neither depends on the number of molecules, N . If we imagine
i i i
inserting a barrier that converts an equilibrated collection of molecules into two half-size collections, each of the new collections is
still at equilibrium. Each contains half as many molecules and has half the total energy of the original. In our model, the fraction of
the molecules in any given energy level remains constant. Consequently, the probabilities associated with each energy level remain
constant. (In Chapter 25, we introduce Fermi-Dirac and Bose-Einstein statistics. When we must use either of these models to
describe the system, P is affected by rules for the number of molecules that can occupy an energy level.)
i
The number of molecules and the total energy are extensive properties and vary in direct proportion to the size of the system. The
probability, P , is an intensive variable that is a characteristic property of the macroscopic system. P is a state function. P
i i i
depends on ϵ . So long as the thermodynamic variables that determine the state of the system remain constant, the ϵ are constant.
i i
For a given macroscopic system in which only pressure–volume work is possible, the quantum mechanical energy levels, ϵ , are i
constant so long as the system volume and temperature are constant. However, the ϵ are quantum-mechanical quantities that
i
depend on our specification of the molecule and on the boundary values in our specification of the system. If we change any
molecular properties or the dimensions of the system, the probabilities, P , change.
i
This page titled 20.2: The Probability of An Energy Level at Constant N, V, and T is shared under a CC BY-SA 4.0 license and was authored,
In developing Boltzmann statistics, we assume that we can tell different molecules of the same substance apart. We say that the
molecules are distinguishable. This assumption is valid for molecules that occupy lattice sites in a crystal. In a crystal, we can
specify a particular molecule by specifying its position in the lattice. In other systems, we may be unable to distinguish between
different molecules of the same substance. Most notably, we cannot distinguish between two molecules of the same substance in
the gas phase. The fact that gas molecules are indistinguishable, while we assume otherwise in developing Boltzmann statistics,
turns out to be a problem that is readily overcome. We discuss this in Section 24.2.
We want to model properties of a system that contains N , identical, distinguishable, non-interacting molecules. The solutions of
the Schrödinger equation presume fixed boundary conditions. This means that the volume of this N -molecule system is constant.
We assume also that the temperature of the N -molecule system is constant. Thus, our goal is a theory that predicts the properties of
a system when N , V , and T are specified. When there are no intermolecular interactions, the energy of the system is just the sum
of the energies of the individual molecules. If we know how the molecules are allocated among the energy levels, we can find the
energy of the system. Letting N be the population of the energy level ϵ , any such allocation is a population set
i i
{ N , N , … , N , …}. We have
1 2 i
N = ∑ Ni
i=1
and the system energy is

∞
E = ∑ Ni ϵi
i=1
Let us imagine that we can assemble a system with the molecules allocated among the energy levels in any way we please. Let
{ N , N , … , N , …} represent an initial population set that describes a system that we assemble in this way. This population
o o o
1 2 i
set corresponds to a well-defined system energy. We imagine immersing the container in a constant-temperature bath. Since the
system can exchange energy with the bath, the molecules of the system gain or lose energy until the system attains the temperature
of the bath in which it is immersed. As this occurs, the populations of the energy levels change. A series of different population sets
characterizes the state of the system as it evolves toward thermal equilibrium. When the system reaches equilibrium, the population
sets that characterize it are different from the initial one, {N , N , … , N , …}.
o
1
o
2
o
i
Evidently, the macroscopic properties of such a system also change with time. The changes in the macroscopic properties of the
system parallel the changing energy-level populations. At thermal equilibrium, macroscopic properties of the system cease to
undergo any further change. In Section 3.9, we introduce the idea that the most probable population set, which we denote as
{N , N , … , N , …}
1 2 i
or its proxy,
{N P (ϵ1 ) , N P (ϵ2 ) , … , N P (ϵi ) , …}
(where N = N + N +. . . +N +. . . ), is the best prediction we can make about the outcomes in a future set of experiments in
1 2 i
which we find the energy of each of N different molecules at a particular instant. We hypothesize that the most probable
population set specifies all of the properties of the macroscopic system in its equilibrium state. When we develop the logical
consequences of this hypothesis, we find a theory that expresses macroscopic thermodynamic properties in terms of the energy
levels available to individual molecules. In the end, the justification of this hypothesis is that it enables us to calculate
thermodynamic properties that agree with experimental measurements made on macroscopic systems.
Our hypothesis asserts that the properties of the equilibrium state are the same as the properties of the system when it is described
by the most probable population set. Evidently, we can predict the system’s equilibrium state if we can find the equilibrium N i
values, and vice versa. To within an arbitrary factor representing its size, an equilibrated system can be completely described by its
intensive properties. In the present instance, the fractions N /N , N /N , …, N /N , … describe the equilibrated system to within
1 2 i
the factor, N , that specifies its size. Since we infer that Pi = P (ϵi ) = N /N
i
, the equilibrated system is also described by the
probabilities (P , P , … , P , …).
1 2 i
Our hypothesis does not assert that the most-probable population set is the only population set possible at equilibrium. A very large
number of other population sets may describe an equilibrium system at different instants of time. However, when its state is
specified by any such population set, the macroscopic properties of the system are indistinguishable from the macroscopic
properties of the system when its state is specified by the most-probable population set. The most-probable population set
characterizes the equilibrium state of the system in the sense that we can calculate the properties of the equilibrium state of the
macroscopic system by using the single-molecule energy levels and the most probable population set—or its proxy. The
relationship between a molecular energy level, ϵ , and its equilibrium population, N , is called the Boltzmann equation. From
i i
P = N /N , we see that the Boltzmann equation specifies the probability of finding a given molecule in energy level ϵ .
i i
i
Although we calculate thermodynamic properties from the most probable population set, the population set that describes the
system can vary from instant to instant while the system remains at equilibrium. The central limit theorem enables us to
characterize the amount of variation that can occur. When N is comparable to the number of molecules in a macroscopic system,
the probability that variation among population sets can result in a macroscopically observable effect is vanishingly small. The
hypothesis is successful because the most probable population set is an excellent proxy for any other population set that the
equilibrium system is remotely likely to attain.
We develop the theory of statistical thermodynamics for N -molecule systems by considering the energy levels, ϵ , available to a
i
single molecule that does not interact with other molecules. Thereafter, we develop a parallel set of statistical thermodynamic
results by considering the energy levels, E
^
, available to a system of N molecules. These N -molecule-system energies can reflect
i
the effects of any amount of intermolecular interaction. We can apply the same arguments to find that the Boltzmann equation also
describes the equilibrium properties of systems in which intermolecular interactions are important. That is, the probability,
^
Pi (E i ) , that an N -molecule system has energy E
^
is the same function of E
i
^
as the molecular-energy probability, P
i i = P (ϵi ) , is
of ϵ .
i
When we finish our development based on single-molecule energy levels, we understand nearly all of the ideas that we need in
order to complete the development for the energies of an N -molecule system. This development is an elegant augmentation of the
basic argument called the ensemble treatment or the ensemble method. The ensemble treatment is due to J. Willard Gibbs; we
discuss it in Chapter 23. For now, we simply note that our approach involves no wasted effort. When we discuss the ensemble
method, we use all of the ideas that we develop in this chapter and the next. The extension of these arguments that is required for
the ensemble treatment is so straightforward as to be (almost) painless.
This page titled 20.3: The Population Sets of a System at Equilibrium at Constant N, V, and T is shared under a CC BY-SA 4.0 license and was
There are an infinite number of successively greater energies for a quantum mechanical system. We infer that the probability that a
given energy level is occupied is a property of the energy level. Each of the probabilities must be between 0 and 1. When we sum
the fixed probabilities associated with the energy levels, the sum contains an infinite number of terms. By the nature of probability,
the sum of this infinite number of terms must be one:
1 = P1 + P2 + ⋯ + Pi + …
= P (ϵ1 ) + P (ϵ2 ) + ⋯ + P (ϵi ) + …
= ∑ P (ϵi )
i=1
That is, the sum of the probabilities is an infinite series, which must converge: The sum of all of the occupancy probabilities must
be unity. This can happen only if all later members of the series are very small. In the remainder of this chapter, we explore some of
the thermodynamic ramifications of these facts. In the next chapter, we use this relationship to find the functional dependence of
the P on the energy levels, ϵ . To obtain these results, we need to think further about the probabilities associated with the various
i i
population sets that can occur. Also, we need to introduce a new fundamental postulate.
To focus on the implications of this sum of probabilities, let us review geometric series. A geometric series is a sum of terms, in
which each successive term is a multiple of its predecessor. A geometric series is an infinite sum that can converge:
∞
2 i 2 i i
T = a + ar + ar + ⋯ + ar ⋯ = a (1 + r + r + ⋯ + r + …) = a + a ∑ r
i=1
Successive terms approach zero if |r| < 1 . If |r| ≥ 1 , successive terms do not become smaller, and the sum does not have a finite
limit. If |r| ≥ 1 , we say that the infinite series diverges.
We can multiply an infinite geometric series by its constant factor to obtain
2 3 i
rT = ar + ar + ar + ⋯ + ar + …
2 3 i
= a (r + r +r + ⋯ + r + …)
i
= a∑r
i=1
If |r| < 1 , we can subtract and find the value of the infinite sum:
T − rT = a
so that
T = a/(1 − r)
In a geometric series, the ratio of two successive terms is r

n+1
/r
n
=r The condition of convergence for a geometric series can
also be written as
∣ rn+1 ∣
∣ ∣ <1
n
∣ r ∣
We might anticipate that any other series also converges if its successive terms become smaller at least as fast as those of a
geometric series. In fact, this is true and is the basis for the ratio test for convergence of an infinite series. If we represent
successive terms in an infinite series as t , their sum is
i
T = ∑ ti
i=0
The ratio test is a theorem which states that the series converges, and T has a finite value, if
∣ tn+1 ∣
lim ∣ ∣ <1
n→∞∣ tn ∣
One of our goals is to discover the relationship between the energy, ϵ , of a quantum state and the probability that a molecule will
i
occupy one of the quantum states that have this energy, P = g ρ (ϵ ) . When we do so, we find that the probabilities for all of the
i i i
quantum mechanical systems that we discuss in Chapter 18 satisfy the ratio test.
This page titled 20.4: How can Infinitely Many Probabilities Sum to Unity? is shared under a CC BY-SA 4.0 license and was authored, remixed,
In a collection of distinguishable independent molecules at constant N , V , and T , the probability that a randomly selected
molecule has energy ϵ is P ; we have 1 = P + P + ⋯ + P + … . At any instant, every molecule in the N -molecule system
i i 1 2 i
has a specific energy, and the state of the system is described by a population set, {N , N , … , N , … . }, wherein N can have
1 2 i i
any value in the range 0 ≤ N ≤ N , subject to the condition that

i
N = ∑ Ni
i=1
The probabilities that we assume for this system of molecules have the properties we assume in Chapter 19 where we find the total
probability sum by raising the sum of the energy-level probabilities to the N power. th
N
N! N1 N2 Ni
1 = (P1 + P2 + ⋯ + Pi + …) = ∑ P P …P …
1 2 i
N1 ! N2 ! … Ni ! …
{ Ni }
The total-probability sum is over all possible population sets, {N , N , … , N , … . }, which we abbreviate to {N }, in indicating
1 2 i i
the range of the summation. Each term in this sum represents the probability of the corresponding population set
{ N , N , … , N , … . },. At any given instant, one of the possible population sets describes the way that the molecules of the
1 2 i
physical system are apportioned among the energy levels. The corresponding term in the total probability sum represents the
probability of this apportionment. It is not necessary that all of the energy levels be occupied. We can have N = 0 , in which case k
= 1 and N ! = 1 . Energy levels that are not occupied have no effect on the probability of a population set. The unique
Nk 0
P =P k
k k
population set
{N , N , … , N , … . }
1 2 i
that we conjecture to characterize the equilibrium state is represented by one of the terms in this total probability sum. We want to
focus on the relationship between a term in the total probability sum and the corresponding state of the physical system.
Each term in the total probability sum includes a probability factor, P P … P … This factor is the probability that N
N1
1
N2
2
Ni
i i
molecules occupy each of the energy levels ϵ . This term is not affected by our assumption that the molecules are distinguishable.
i
The probability factor is multiplied by the polynomial coefficient

N!
N1 ! N2 ! … Ni ! …
This factor is the number of combinations of distinguishable molecules that arise from the population set {N , N , … , N , …}. It 1 2 i
is the number of ways that the N distinguishable molecules can be assigned to the available energy levels so that N of them are in 1
energy level, ϵ , etc.

1
The combinations for the population set {3,2} are shown in Figure 2.
Figure 2. Combinations for the population set {3,2}.

The expression for the number of combinations takes the form it does only because the molecules can be distinguished from one
another. To emphasize this point, let us find the number of combinations using the method we develop in Chapter 19. Briefly
recapitulated, the argument is this:
1. We can permute the N molecules in N ! ways. If we were to distinguish (as different combinations) any two permutations of all
of the molecules, this would also be the number of combinations.
2. In fact, however, we do not distinguish between different permutations of those molecules that are assigned to the same energy
level. If the N molecules assigned to the first energy level are B , C , Q,…, X, we do not distinguish the permutation
1
BC Q … X from the permutation C BQ … X or from any other permutation of these N molecules. Then the complete set of
1
N ! permutations contains a subset of N ! permutations, all of which are equivalent because they have the same molecules in
1
the first energy level. So the total number of permutations, N !, over-counts the number of combinations by a factor of N ! We 1
can correct for this over-count by dividing by N ! That is, after correcting for the over-counting for the N molecules in the
1 1
first energy level, the number of combinations is N !/N ! (If all N of the molecules were in the first energy level, there would
1
be only one combination. We would have N = N , and the number of combinations calculated from this formula would be
1
N !/N ! = 1 , as required.)
3. The complete set of N ! permutations also includes N ! permutations of the N molecules in the second energy level. In finding
2 2
the number of combinations, we want to include only one of these permutations, so correcting for the over-counting due to both
the N molecules in the first energy level and the N molecules in the second energy level gives
1 2
N!
N1 ! N2 !
4. Continuing this argument through all of the occupied energy levels, we see that the total number of combinations is
N!
C (N1 , N2 , … , Ni , …) =
N1 ! N2 ! … Ni ! …
Because there are infinitely many energy levels and probabilities, P , there are infinitely many terms in the total-probability sum.
i
Every energy available to the macroscopic system is represented by one or more terms in this total-probability sum. Since there is
no restriction on the energy levels that can be occupied, there are an infinite number of such system energies. There is an
enormously large number of terms each of which corresponds to an enormously large system energy. Nevertheless, the sum of all
of these terms must be one. The P form a convergent series, and the total probability sum must sum to unity.
i
Just as the P series can converge only if the probabilities of high molecular energies become very small, so the total probability
i
sum can converge only if the probabilities of high system energies become very small. If a population set has N molecules in the i
th
i energy level, the probability of that population set is proportional to P . We see therefore, that the probability of a population
Ni
set in which there are many molecules in high energy levels must be very small. Terms in the total probability sum that correspond
to population sets with many molecules in high energy levels must be negligible. Equivalently, at a particular temperature,
macroscopic states in which the system energy is anomalously great must be exceedingly improbable.
What terms in the total probability sum do we need to consider? Evidently from among the infinitely many terms that occur, we can
select a finite subset whose sum is very nearly one. If there are many terms that are small and nearly equal to one another, the
number of terms in this finite subset could be large. Nevertheless, we can see that terms in this subset must involve the largest
possible P values raised to the smallest possible powers, N , consistent with the requirement that the N sum to N .
i i i
If an equilibrium macroscopic system could have only one population set, the probability of that population set would be unity.
Could an equilibrium system be characterized by two or more population sets for appreciable fractions of an observation period?
Would this require that the macroscopic system change its properties with time as it jumps from one population set to another?
Evidently, it would not, since our observations of macroscopic systems show that the equilibrium properties are unique. A system
that wanders between two (or more) macroscopically distinguishable states cannot be at equilibrium. We are forced to the
conclusion that, if a macroscopic equilibrium system has multiple population sets with non-negligible probabilities, the
macroscopic properties associated with each of these population sets must be indistinguishably similar. (The alternative is to
abandon the theory, which is useful only if its microscopic description of a system makes useful predictions about the system’s
macroscopic behavior.)
To be a bit more precise about this, we recognize that our theory also rests on another premise: Any intensive macroscopic property
of many independent molecules depends on the energy levels available to an individual molecule and the fraction of the molecules
that populate each energy level. The average energy is a prime example. For the population set {N , N , … , N , … . }, the
1 2 i
average molecular energy is
∞
Ni
¯
ϵ = ∑( ) ϵi
N
i=1
We recognize that many population sets may contribute to the total probability sum at equilibrium. If we calculate essentially the
same ϵ̄ from each of these contributing population sets, then all of the contributing population sets correspond to indistinguishably
different macroscopic energies. We see in the next section that the central limit theorem guarantees that this happens whenever N
is as large as the number of molecules in a macroscopic system.
This page titled 20.5: The Total Probability Sum at Constant N, V, and T is shared under a CC BY-SA 4.0 license and was authored, remixed,
We are imagining that we can examine a collection of N distinguishable molecules and determine the energy of each molecule in
the collection at any particular instant. If we do so, we find the population set, {N , N , … , N , … . }, that characterizes the
1 2 i
system at that instant. In Section 3.9, we introduce the idea that the most probable population set, {N , N , … N , . , , , }, or its 1 2 i
proxy, {N P (ϵ ) , N P (ϵ ) , … , N P (ϵ ) , … . }, is the best prediction we can make about the outcome of a future replication of
1 2 i
this measurement. In Section 20.2, we hypothesize that the properties of the system when it is characterized by the most probable
population set are indistinguishable from the properties of the system at equilibrium.
Now let us show that this hypothesis is implied by the central limit theorem. We suppose that the population set that characterizes
the system varies from instant to instant and that we can find this population set at any given instant. The population set that we
find at a particular instant comprises a random sample of N molecular energies. For this sample, we can find the average energy
from
∞
Ni
ϵ̄ = ∑ ( ) ϵi
N
i=1
The expected value of the molecular energy is

∞
⟨ϵ⟩ = ∑ Pi ϵi
i=1
It is important that we remember that ϵ̄ and ⟨ϵ⟩ are not the same thing. There is a distribution of ϵ̄ values, one ϵ̄ value for each of
the possible population sets {N , N , … , N , … . }. In contrast, when N , V , and T are fixed, the expected value, ⟨ϵ⟩, is a
1 2 i
constant; the value of ⟨ϵ⟩ is completely determined by the values of the variables that determine the state of the system and fix the
probabilities P . If our theory is to be useful, the value of ⟨ϵ⟩ must be the per-molecule energy that we observe for the macroscopic
i
system we are modeling.

According to the central limit theorem, the average energy of a randomly selected sample, ϵ̄, approaches the expected value for the
distribution, ⟨ϵ⟩, as the number of molecules in the sample becomes arbitrarily large. In the present instance, we hypothesize that
the most probable population set, or its proxy, characterizes the equilibrium system. When N is sufficiently large, this hypothesis
implies that the probability of the i th
energy level is given by P i = N /N
i
. Then the expected value of a molecular energy is
∞ ∞
N
i
⟨ϵ⟩ = ∑ Pi ϵi = ∑ ( ) ϵi
N
i=1 i=1
Since the central limit theorem asserts that ϵ̄ approaches ⟨ϵ⟩ as N becomes arbitrarily large:
∞ ∞
Ni Ni N
i
¯
0 = lim (ϵ − ⟨ϵ⟩ ) = lim ∑ ( − Pi ) εi = lim ∑ ( − ) ϵi
N →∞ N →∞ N N →∞ N N
i=1 i=1
One way for the limit of this sum to be zero is for the limit of every individual term to be zero. If the ϵ were arbitrary, this would i
be the only way that the sum could always be zero. However, the ϵ and the P are related, so we might think that the sum is zero
i i
because of these relationships.

To see that the limit of every individual term must in fact be zero, we devise a new distribution. We assign a completely arbitrary
number, X , to each energy level. Now the i energy level is associated with an X as well as an ϵ . We have an X distribution as
i
th
i i
well as an energy distribution. We can immediately calculate the expected value of X. It is

∞
⟨X⟩ = ∑ Pi Xi
i=1
When we find the population set {N 1, N2 , … , Ni , … . } , we can calculate the corresponding average value of X. It is
∞
¯¯¯
¯ Ni
X = ∑( ) Xi
N
i=1
The central limit theorem applies to any distribution. So, it certainly applies to the X distribution; the average value of X
approaches the expected value of X as N becomes arbitrarily large:

∞ ∞
Ni Ni N
¯¯¯
¯ i
0 = lim (X − ⟨X⟩ ) = lim ∑ ( − Pi ) Xi = lim ∑ ( − ) Xi
N →∞ N →∞ N N →∞ N N
i=1 i=1
Now, because the X can be chosen completely arbitrarily, the only way that the limit of this sum can always be zero is that every
i
individual term becomes zero.

In the limit as N → ∞ , we find that
Ni /N → N /N
i
As the number of molecules in the equilibrium system becomes arbitrarily large, the fraction of the molecules in each energy level
at an arbitrarily selected instant approaches the fraction in that energy level in the equilibrium-characterizing most-probable
population set, {N 1
,N ,…N
2 i
…} . In other words, the only population sets that we have any significant chance of observing in a
large equilibrium system are population sets whose occupation fractions, N /N , are all very close to those, N /N , in the i i
equilibrium-characterizing population set. Estimating P as the ratio N /N gives essentially the same result whichever of these
i i
population sets we use. Below, we see that the ϵ and the P determine the thermodynamic properties of the system. Consequently,
i i
when we calculate any observable property of the macroscopic system, each of these population sets gives the same result.
Since the only population sets that we have a significant chance of observing are those for which
Ni /N ≈ N /N
i
we frequently say that we can ignore all but the most probable population set. What we have in mind is that the most probable
population set is the only one we need in order to calculate the macroscopic properties of the equilibrium system. We are incorrect,
however, if we allow ourselves to think that the most probable population set is necessarily much more probable than any of the
others. Nor does the fact that the N /N are all very close to the N /N mean that the N are all very close to the N . Suppose that
i i i i
the difference between the two ratios is 10 . If N = 10 , the difference between N and N is 10 , which probably falls
−10 20
i i
10
outside the range of values that we usually understand by the words “very close.”
We develop a theory that includes a mathematical model for the probability that a molecule has any one of its quantum-
mechanically possible energies. It turns out that we are frequently interested in macroscopic systems in which the number of energy
levels greatly exceeds the number of molecules. For such systems, we find N P ≪ 1 , and it is no longer possible to say that a i
single most-probable population set, {N , N , … N , …}, describes the equilibrium state of the system. When it is very unlikely
1 2 i
that any energy level is occupied by more than one molecule, the probability of any population set in which any N is greater than i
one becomes negligibly small. We can approximate the total probability sum as
N N1 N2 Ni
1 = (P1 + P2 + ⋯ + Pi + …) ≈ ∑ N !P P …P …
1 2 i
{ Ni }
However, the idea that the proxy, {N P (ϵ ) , N P (ϵ ) , … , N P (ϵ ) , … . }, describes the equilibrium state of the system remains
1 2 i
valid. In these circumstances, a great many population sets can have essentially identical properties; the properties calculated from
any of these are indistinguishable from each other and indistinguishable from the properties calculated from the proxy. Since the
equilibrium properties are fixed, the value of these extended products is fixed. For any of the population sets available to such a
system at equilibrium, we have
N1 N2 Ni N P1 N P2 N Pi
P P …P ⋯ =P P …P ⋯ = constant
1 2 i 1 2 i
It follows that, for some constant, c , we have

∞ ∞
c = ∑ N Pi ln Pi = N ∑ Pi ln Pi
i=1 i=1
As it evolves, we see that the probability of finding a molecule in an energy level is the central feature of our theory.
This page titled 20.6: The Most Probable Population Set at Constant N, V, and T is shared under a CC BY-SA 4.0 license and was authored,
Thus far, we have considered only the probabilities associated with the assignments of distinguishable molecules to the allowed
energy levels. In Section 20.2, we introduce the hypothesis that all of the g degenerate quantum states with energy ϵ are equally i i
probable, so that the probability that a molecule has energy ϵ is P = P (ϵ ) = g ρ (ϵ ) . Making this substitution, the total
i i i i i
probability sum becomes

N
1 = (P1 + P2 + ⋯ + Pi + …)
N! N1 N2 Ni
= ∑ P P …P …
1 2 i
N1 ! N2 ! … Ni ! …
{ Ni }
N1 N2 Ni
N !g g …g …
1 2 i N1 N2 Ni
= ∑ ρ (ϵ1 ) ρ (ϵ2 ) … ρ (ϵi ) ….
N1 ! N2 ! … Ni ! …
{ Ni }
∞ Ni
g
i Ni
= ∑ N! ∏ ( ) ρ(ϵi )
Ni !
{ Ni } i=1
∞
Ni
= ∑ W ∏ ρ(ϵi )
{ Ni } i=1
where we use the notation

ω
a1 × a2 × … ai × … aω × = ∏ ai
i=1
for extended products and introduce the function

W = W (Ni , gi )
= W (N1 , g1 , N2 , g2 , … , Ni , gi , … . )
∞ Ni
g
i
= N! ∏ ( )
Ni !
i=1
∞
Ni
= C (N1 , N2 , … , Ni , …) ∏ g
i
i=1
For reasons that become clear later, W is traditionally called the thermodynamic probability. This name is somewhat unfortunate,
because W is distinctly different from an ordinary probability.
In Section 20.5, we note that P P … P is the probability that N molecules occupy each of the energy levels ϵ and that
N1
1 2
N2 Ni
i i i
N !/(N ! N ! … N ! …) is the number of combinations of distinguishable molecules that arise from the population set
1 2 i
{ N , N , … , N , …}. Now we observe that the extended product

1 2 i
N1 N2 Ni
ρ (ϵ1 ) ρ (ϵ2 ) … ρ (ϵi ) ….
is the probability of any one assignment of the distinguishable molecules to quantum states such that N molecules are in quantum i
states whose energies are ϵ . Since a given molecule of energy ϵ can be in any of the g degenerate quantum states, the probability
i i i
that it is in the energy level ϵ is g -fold greater that the probability that it is in any one of these quantum states.
i i
Microstates
We call a particular assignment of distinguishable molecules to the available quantum states a microstate. For any population set,
there are many combinations. When energy levels are degenerate, each combination gives rise to many microstates. The factor
… . is the probability of any one microstate of the population set { N , N , … , N , …}. Evidently,
N1 N2 Ni
ρ (ϵ )
1 ρ (ϵ )
2 … ρ (ϵ ) i 1 2 i
the thermodynamic probability
∞ Ni
g
i
W = N! ∏ ( ) (20.7.1)
Ni !
i=1
is the total number of microstates of that population set.

To see directly that the number of microstates is dictated by Equation 20.7.1, let us consider the number of ways we can assign N
distinguishable molecules to the quantum states when the population set is {N , N , … , N , …} and energy level ϵ is g -fold
1 2 i i i
degenerate. We begin by assigning the N molecules in energy level ϵ . We can choose the first molecule from among any of the
1 1
N distinguishable molecules and can choose to place it in any of the g quantum states whose energy is ϵ . The number of ways
1 1
we can make these choices is N g . We can choose the second molecule from among the N − 1 remaining distinguishable
1
molecules. In Boltzmann statistics, we can place any number of molecules in any quantum state, so there are again g quantum 1
states in which we can place the second molecule. The total number of ways we can place the second molecule is (N − 1) g . 1
The number of ways the first and second molecules can be chosen and placed is therefore N (N − 1) g . We find the number of 2
1
ways that successive molecules can be placed in the quantum states of energy ϵ by the same argument. The last molecule whose
1
energy is ϵ can be chosen from among the (N − N + 1) remaining molecules and placed in any of the g quantum states. The
1 1 1
total number of ways of placing the N molecules in energy level ϵ is N (N − 1) (N − 2) … (N − N + 1) g

1 1 . 1
N1
This total includes all possible orders for placing every set of N distinguishable molecules into every possible set of quantum
1
states. However, the order doesn’t matter; the only thing that affects the state of the system is which molecules go into which
quantum state. (When we consider all of the ways our procedure puts all of the molecules into any of the quantum states, we find
that any assignment of molecules A , B , and C to any particular set of quantum states occurs six times. Selections in the orders A ,
B ,C ; A ,C ,B ; B ,A ,C ; B ,C ,A ; C ,A ,B ; and C ,B ,A all put the same molecules in the same quantum states.) There are N ! orders 1
in which our procedure chooses the N molecules; to correct for this, we must divide by N !, so that the total number of
1 1
assignments we want to include in our count is

N1
N (N − 1) (N − 2) … (N − N1 + 1) g / N1 !
1
The first molecule that we assign to the second energy level can be chosen from among the N − N remaining molecules and 1
placed into any of the g quantum states whose energy is ϵ . The last one can be chosen from among the remaining
2 2
(N − N − N + 1)
1 2 molecules. The number of assignments of the N molecules to g -fold degenerate quantum states whose
2 2
energy is ϵ is
2
N2
(N − N1 ) (N − N1 − 1) … (N − N1 − N2 + 1) g / N2 !
2
When we consider the number of assignments of molecules to quantum states with energies ϵ and ϵ we have 1 2
N (N − 1) … (N − N1 + 1) (N − N1 ) (N − N1 − 1) …
N1 N2
g g
1 2
× (N − N1 − N2 + 1) ( )( )
N1 ! N2 !
Let the last energy level to contain any molecules be ϵ . The number of ways that the N molecules can be assigned to the
ω ω
quantum states with energy ϵ is N (N − 1) … (1) g /N ! The total number of microstates for the population set
ω ω ω
Nω
ω ω
{ N , N , … , N , …} becomes
1 2 i
N (N − 1) … (N − N1 ) (N − N1 − 1) …
∞ Ni ∞ Ni
g g
i i
× (Nω ) (Nω − 1) … (1) ∏ ( ) = N! ∏ ( )
Ni ! Ni !
i=1 i=1
When we consider Fermi-Dirac and Bose-Einstein statistics, it is no longer true that the molecules are distinguishable. For Fermi-
Dirac statistics, no more than one molecule can be assigned to a particular quantum state. For a given population set, Boltzmann,
Fermi-Dirac, and Bose-Einstein statistics produce different numbers of microstates.
It is helpful to have notation that enables us to specify different combinations and different microstates. If ϵ is the energy i
associated with the wave equation that describes a particular molecule, it is convenient to say that the molecule is in energy level
ϵ ; that is, its quantum state is one of those that has energy ϵ . Using capital letters to represent molecules, we indicate that
i i
molecule A is in energy level ϵ by writing ϵ (A) . To indicate that A , B , and C are in ϵ , we write ϵ (A, B, C ). Similarly, to
i i i i
indicate that molecules D and E are in ϵ , we write ϵ (D, E). For this system of five molecules, the assignment
k k
ϵ (A, B, C ) ϵ (D, E) represents one of the possible combinations. The order in which we present the molecules that have a given
i k
energy is immaterial: ϵ (A, B, C ) ϵ (D, E) and ϵ (C , B, A) ϵ (E, D) represent the same combination. When any one molecule
i k i k
is distinguishable from others of the same substance, assignments in which a given molecule has different energies are physically
different and represent different combinations. The assignments ϵ (A, B, C ) ϵ (D, E) and ϵ (D, B, C ) ϵ (A, E) represent
i k i k
different combinations. In Figure 2, we represent these assignments more schematically.

Any two assignments in which a particular molecule occupies different quantum states give rise to different microstates. If the i th
energy level is three-fold degenerate, a molecule in any of the quantum states ψ , ψ , or ψ has energy ϵ . Let us write
i,1 i,2 i,3 i
ψi,1 (A, B) ψi,2 (C ) ψk,1 (DE)
to indicate the microstate arising from the combination ϵ (A, B, C ) ϵ

i k (D, E) in which molecules A and B occupy ψ , molecule i,1
C occupies ψ , and molecules D and E occupy ψ . Then,

i,2 k,1
ψi,1 (A, B) ψi,2 (C ) ψk,1 (DE)
ψi,1 (B, C ) ψi,2 (A) ψk,1 (DE)
ψi,1 (A) ψi,2 (B, C ) ψk,1 (DE)
are three of the many microstates arising from the combination ϵ (A, B, C ) ϵ i k (D, E) . Figure 3 shows all of the microstates
possible for the population set {2, 1} when the quantum states of a molecule are ψ 1,1
,ψ 1,2
, and ψ . 2,1
Figure 3. Microstates for {2,1} with quantum states Ψ , Ψ , and Ψ .

1,1 1,2 2,1
This page titled 20.7: The Microstates of a Given Population Set is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
In Section 20.2, we introduce the assumption that, for a molecule in a constant-N-V-T system, for which the g and ϵ are fixed, the i i
probability of a quantum state, ρ (ϵ ), depends only on its energy. It follows that two or more quantum states that have the same
i
energy must have equal probabilities. We accept the idea that the probability depends only on energy primarily because we cannot
see any reason for a molecule to prefer one state to another if both states have the same energy.
We extend this thinking to multi-molecule systems. If two microstates have the same energy, we cannot see any reason for the
system to prefer one rather than the other. In a constant-N-V-T system, in which the total energy is not otherwise restricted, each
microstate of {N , N , … , N , … . } occurs with probability ρ (ϵ ) ρ (ϵ ) … ρ (ϵ ) … ., and each microstate of
1 2 i 1
N1
2
N2
i
Ni
# # #
# # # N N N
{N
1
,N
2
, …,N
I
occurs with probability ρ (ϵ ) ρ (ϵ ) … ρ (ϵ ) … . When the energies of these population
,… } 1
1
2
2
i
i
sets are equal, we infer that these probabilities are equal, and their value is a constant of the system. That is,
N1 N2 Ni
ρ (ϵ1 ) ρ (ϵ2 ) … ρ (ϵi ) ….
# # #
N N N
= ρ (ϵ1 ) 1
ρ (ϵ2 ) 2
… ρ (ϵi ) i
…. =ρ = constant
MS,N ,E
where we introduce ρ
MS,N ,E
to represent the probability of a microstate of a system of N molecules that has total energy E . If
E =E , then ρ
#
MS,N ,E
=ρ
MS,N ,E
. #
When we think about it critically, the logical basis for this equal-probability idea is not very impressive. While the idea is plausible,
it is not securely rooted in any particular empirical observation or prior postulate. The equal-probability idea is useful only if it
leads us to theoretical models that successfully mirror the behavior of real macroscopic systems. This it does. Accordingly, we
recognize that the equal-probability idea is really a fundamental postulate about the behavior of quantum-mechanical systems. It is
often called the principle of equal a priori probabilities:
Definition: principle of equal a priori probabilities

For a particular system, all microstates that have the same energy have the same probability.
Our development of statistical thermodynamics relies on the principle of equal a priori probabilities. For now, let us summarize the
important relationships that the principle of equal a priori probabilities imposes on our microscopic model for the probabilities of
two population sets of a constant-N-V-T system that have the same energy:
A given population set {N 1, N2 , … , Ni , … . } gives rise to W (N i, gi ) microstates, and each of these microstates has energy
∞
E = ∑ Ni ϵi
i=1
A second population set, {N , N , … , N , … }, that has the same energy need not—and usually will not—give rise to
#
1
#
2 I
#
the same number of microstates. In general, for two such population sets,
#
W (Ni , gi ) ≠ W (N , gi )
i
However, because each microstate of either population set has the same energy, we have
∞ ∞
#
E = ∑ Ni ϵi = ∑ N ϵi
i
i=1 i=1
The probability of a microstate of a given population set {N 1, N2 , … , Ni , … . } depends only on its energy:
N1 N2 Ni
ρ (ϵ1 ) ρ (ϵ2 ) … ρ (ϵi ) ….=ρ = constant
MS,N ,E
This page titled 20.8: The Probabilities of Microstates that Have the Same Energy is shared under a CC BY-SA 4.0 license and was authored,
In principle, the energy of an equilibrium system that is in contact with a constant-temperature heat reservoir can vary slightly with
time. In contrast, the energy of an isolated system is constant. A more traditional and less general statement of the equal a priori
probability principle focuses on isolated systems, for which all possible microstates necessarily have the same energy:
All microstates of an isolated (constant energy) system occur with equal probability.
If we look at the fraction of the molecules of an isolated system that are in each microstate, we expect to find that these fractions
are approximately equal. In consequence, for an isolated system, the probability of a population set, {N , N , … , N , … . }, is 1 2 i
proportional to the number of microstates, W (N , g ), to which that population set gives rise.
i i
In principle, the population sets of a constant-N-V-T system can be significantly different from those of a constant-N-V-E system.
That is, if we move an isolated system, whose temperature is T, into thermal contact with a heat reservoir at constant-temperature T,
the population sets that characterize the system can change. In practice, however, for a system containing a large number of
molecules, the population sets that contribute to the macroscopic properties of the system must be essentially the same.
The fact that the same population sets are important in both systems enables us to make two further assumptions that become
important in our development. We assume that the proportionality between the probability of a population set and W (N , g ), i i
which is strictly true only for a constant-N-V-E system, is also true for the corresponding constant-N-V-T system. We also assume
that the probabilities of a quantum state, ρ (ϵ ), and a microstate, ρ
i , which we defined for the constant-N-V-T system, are
MS,N ,E
the same for the corresponding constant-N-V-E system.

Let us see why we expect the same population sets to dominate the macroscopic properties of otherwise identical constant-energy
and constant-temperature systems. Suppose that we isolate a constant-N-V-T system in such a way that the total energy,
N ϵ , of the isolated system is exactly equal to the expected value, ⟨E⟩ = N ∑ P ϵ , of the energy of the system
∞ ∞
E =∑ i i i i
i=1 i=1
when its temperature is constant. What we have in mind is a gedanken experiment, in which we monitor the energy of the
thermostatted system as a function of time, waiting for an instant in which the system energy, E = ∑ N ϵ , is equal to the ∞
i=1 i i
expected value of the system energy, ⟨E⟩. When this occurs, we instantaneously isolate the system.
We suppose that the isolation process is accomplished before any molecule can experience an energy change, so that the population
set that characterizes the system immediately afterwards is the same as the one that characterizes it before. After isolation, of
course, the molecules can exchange energy with one another, and many population sets may be available to the system.
Clearly, the value of every macroscopic property of the isolated system must be the same as its observable value in the original
constant-temperature system. Our microscopic description of it is different. Every population set that is available to the isolated
system has energy E = ⟨E⟩ , and gives rise to
∞ Ni
g
i
W (Ni , gi ) = N ! ∏ ( )
Ni !
i=1
microstates. At the same temperature, each of these microstates occurs with the same probability. Since the isolated-system energy
is ⟨E⟩, this probability is ρ . The probability of an available population set is W (N , g ) ρ
MS,N ,⟨E⟩
. i i MS,N ,⟨E⟩
Since the temperature can span a range of values centered on ⟨T ⟩, where ⟨T ⟩ is equal to the temperature of the original constant-N-
V-T system, there is a range of ρ values spanning the (small) range of temperatures available to the constant-energy
MS,N ,⟨E⟩
system. Summing over all of the population sets that are available to the isolated system, we find
1 = ∑ W (Ni , gi ) ρMS,N ,⟨E⟩ + ∑ W (Ni , gi ) ρMS,N ,⟨E⟩
{ Ni }, E=⟨E⟩,T =⟨T ⟩ { Ni }, E=⟨E⟩,T ≠⟨T ⟩
The addition of “E = ⟨E⟩ ” beneath the summation sign emphasizes that the summation is to be carried out over the population
sets that are consistent with both the molecule-number and total-energy constraints and no others. The total probability sum breaks
into two terms, one spanning population sets whose temperature is exactly ⟨T ⟩ and another spanning all of the other population
sets. (Remember that the ρ (ϵ ) are temperature dependent.)
i
The population sets available to the isolated system are slightly different from those available to the constant-temperature system.
In our microscopic model, only population sets that have exactly the right total energy can occur in the isolated system. Only
population sets that have exactly the right temperature can occur in the constant-temperature system.
Summing over all of the population sets that are available to the constant-temperature system, we partition the total probability sum
into two terms:
1 = ∑ W (Ni , gi ) ρ + ∑ W (Ni , gi ) ρ
MS,N ,⟨E⟩ MS,N ,⟨E⟩
{ Ni },E=⟨E⟩,T =⟨T ⟩ { Ni }, E≠⟨E⟩,T =⟨T ⟩
From the central limit theorem, we expect the constant-energy system to have (relatively) few population that fail to meet the
condition E = ⟨E⟩ . Likewise, we expect the constant temperature system to have (relatively) few population sets that fail to meet
the condition T = ⟨T ⟩ . The population sets that satisfy both of these criteria must dominate both sums. For the number of
molecules in macroscopic systems, we expect the approximation to the total probability sum
1 = ∑ W (Ni , gi ) ρ ≈ ∑ W (Ni , gi ) ρ
MS,N ,⟨E⟩ MS,N ,⟨E⟩
{ Ni },E { Ni },E=⟨E⟩,T =⟨T ⟩
to be very good. The same population sets dominate both the constant-temperature and constant-energy systems. Each system must
have a most probable population set, {N , N , … N , . , , , }. If these are not identically the same set, they must be so close that
1 2 i
the same macroscopic properties are calculated using either one.

Thus, the central limit theorem implies that the total probability sum, which we develop for the constant-temperature system, also
describes the constant-energy system, so long as the number of molecules in the system is sufficiently large.
Now, two aspects of this development warrant elaboration. The first is that the probability of population sets that have energies and
temperature that satisfy E = ⟨E⟩ and T = ⟨T ⟩ exactly may actually be much less than one. The second is that constant-energy and
constant-temperature systems are creatures of theory. No real system can actually have an absolutely constant energy or
temperature.
Recognizing these facts, we see that when we stipulate E = ⟨E⟩ or T = ⟨T ⟩ , what we really mean is that E = ⟨E⟩ ± δE and
T = ⟨T ⟩ ± δT , where the intervals ±δE and ±δT are vastly smaller than any differences we could actually measure
experimentally. When we write E ≠ ⟨E⟩ and T ≠ ⟨T ⟩ , we really intend to specify energies and temperatures that fall outside the
intervals E = ⟨E⟩ ± δE and T = ⟨T ⟩ ± δT . If the system contains sufficiently many molecules, the population sets whose
energies and temperatures fall within the intervals E = ⟨E⟩ ± δE and T = ⟨T ⟩ ± δT account for nearly all of the probability—no
matter how small we choose δE and δT . All of the population sets whose energies and temperatures fall within the intervals
E = ⟨E⟩ ± δE and T = ⟨T ⟩ ± δT correspond to the same macroscopically observable properties.
This page titled 20.9: The Probabilities of the Population Sets of an Isolated System is shared under a CC BY-SA 4.0 license and was authored,
In an isolated system, the probability of population set {N , N , … , N , …} is W (N , g ) ρ 1 , where ρ
2 is a
i i i MS,N ,⟨E⟩ MS,N ,⟨E⟩
constant. It follows that W = W (N , g ) is proportional to the probability that the system is in one of the microstates associated
i i
with the population set {N 1, N2 , … , Ni , …} . Likewise, W #

= W (N
i
#
, gi ) is proportional to the probability that the system is
in one of the microstates associated with the population set {N , N , … N , …}. Suppose that we observe the isolated system
1
# #
2 i
#
for a long time. Let F be the fraction of the time that the system is in microstates of population set {N , N , … , N , …} and F 1 2 i
#
# # #
be the fraction of the time that the system is in microstates of the population set {N , N , … N , …}. The principle of equal a 1 2 i
priori probabilities implies that we would find

# #
F W
=
F W
# # #
Suppose that W is much larger than W . This means there are many more microstates for {N , N , … N , …} than there are
#
1 2 i
for {N , N , … , N , …}. The fraction of the time that the population set {N , N , … N , …} characterizes the system is
1 2 i 1
# #
2 i
#
much greater than the fraction of the time {N , N , … , N , …} characterizes it. Alternatively, if we examine the system at an
1 2 i
arbitrary instant, we are much more likely to find the population set {N , N , … N , …} than the population set #
1
#
2 i
#
{ N , N , … , N , …}. The larger W (N , g , N , g , … , N , g , …), the more likely it is that the system will be in one of the
1 2 i 1 1 2 2 i i
microstates associated with the population set {N , N , … , N , …}. In short, W predicts the state of the system; it is a measure
1 2 i
of the probability that the macroscopic properties of the system are those of the population set {N , N , … , N , …}. 1 2 i
If an isolated system can undergo change, and we re-examine it at after a few molecules have moved to different energy levels, we
expect to find it in one of the microstates of a more-probable population set; that is, in one of the microstates of a population set for
which W is larger. At still later times, we expect to see a more-or-less smooth progression: the system is in microstates of
population sets for which the values of W are increasingly larger. This can continue only until the system occupies one of the
microstates of the population set for which W is a maximum or a microstate of one of the population sets whose macroscopic
properties are essentially the same as those of the constant-N -V -E population set for which W is a maximum.
Once this occurs, later inspection may find the system in other microstates, but it is overwhelmingly probable that the new
microstate will still be one of those belonging to the largest-W population set or one of those that are macroscopically
indistinguishable from it. Any of these microstates will belong to a population set for which W is very well approximated by
W ( N , g1 , N , g2 , … , N , gi , …)
1 2 i
. Evidently, the largest-W population set characterizes the equilibrium state of the either the
constant-N -V -T system or the constant–N -V -E system. Either system can undergo change until W reaches a maximum.
Thereafter, it is at equilibrium and can undergo no further macroscopically observable change.
Boltzmann recognized this relationship between W , the thermodynamic probability, and equilibrium. He noted that the
unidirectional behavior of W in an isolated system undergoing spontaneous change is like the behavior we found for the entropy
function. Boltzmann proposed that, for an isolated (constant energy) system, S and W are related by the equation S = kln W ,
where k is Boltzmann’s constant. This relationship associates an entropy value with every population set. For an isolated
macroscopic system, equilibrium corresponds to a state of maximum entropy. In our microscopic model, equilibrium corresponds
to the population set for which W is a maximum. By the argument we make in §6, this population set must be well approximated
by the most probable population set, {N , N , … N , . , , , }. That is, the entropy of the equilibrium state of the macroscopic
1 2 i
system is
S = k ln Wmax
∞
N!
= k ln + k ∑ N ln gi
i
N !N ! … N ! … i=1
i i i
This equation can be taken as the definition of entropy. Clearly, this definition is different from the thermochemical definition,
/T . We can characterize—imperfectly—the situation by saying that the two definitions provide alternative scales for
rev
S =q
measuring the same physical property. As we see below, our statistical theory enables us to define entropy in still more ways, all of
which prove to be functionally equivalent. Gibbs characterized these alternatives as “entropy analogues;” that is, functions whose
properties parallel those of the thermochemically defined entropy.
We infer that the most probable population set characterizes the equilibrium state of either the constant-temperature or the constant-
energy system. Since our procedure for isolating the constant-temperature system affects only the thermal interaction between the
system and its surroundings, the entropy of the constant-temperature system must be the same as that of the constant-energy
system. Using N = N P = N g ρ (ϵ ) and assuming that the approximation ln N ! = N ln N − N is adequate for all of the
i i i i i i i i
energy levels that make a significant contribution to S , substitution shows that the entropy of either system depends only on
probabilities:
∞ ∞
S = kN ln N − kN − k ∑ [N Pi ln(N Pi ) − N Pi ] + k ∑ N Pi ln gi
i=1 i=1
= kN ln N −kN − kN ∑ [ Pi ln(N ) + Pi ln Pi − Pi − Pi ln gi ]
i=1
∞ ∞ ∞
= k (N ln N − N ) − k (N ln N − N ) ∑ Pi − kN ∑ Pi [ln Pi − ln gi ] = −kN ∑ Pi ln ρ (ϵi )
i=1 i=1 i=1
The entropy per molecule, S/N , is proportional to the expected value of ln ρ (ϵ ) ; Boltzmann’s constant is the proportionality
i
constant. At constant temperature, ρ (ϵ ) depends only on ϵ . The entropy per molecule depends only on the quantum state
i i
properties, g and ϵ .
i i
This page titled 20.10: Entropy and Equilibrium in an Isolated System is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
To relate these ideas to a change in a more specific macroscopic system, let us consider isomeric substances A and B . (We consider
this example further in Chapter 21.) In principle, we can solve the Schrödinger equation for a molecule of isomer A and for a
molecule of isomer B . We obtain all possible energy levels for a molecule of each isomer. If we list these energy levels in order,
1
beginning with the lowest, some of these levels belong to isomer A and the others belong to isomer B .
Now let us consider a mixture of N molecules of A and N molecules of B . We suppose that individual molecules are
A B
distinguishable and that intermolecular interactions can be ignored. Since a group of atoms that can form an A molecule can also
form a B molecule, every energy level is accessible to this group of atoms; that is, we can view both sets of energy levels as being
available to the atoms that make up the molecules. For a given system energy, there will be many population sets in which only the
energy levels belonging to isomer A are occupied. For each of these population sets, there is a corresponding thermodynamic
probability, W . Let W be the largest of these thermodynamic probabilities. Similarly, there will be many population sets in
max
which only the energy levels corresponding to isomer B are occupied. Let W be the largest of the thermodynamic probabilities
B
max
associated with these population sets. Finally, there will be many population sets in which the occupied energy levels belong to
both isomer A and isomer B . Let W be the largest of the thermodynamic probabilities associated with this group of population
max
A,B
sets.
Now, W
max
A
is a good approximation to the number of ways that the atoms of the system can come together to form isomer A .
W
B
max
is a good approximation to the the number of ways that the atoms of the system can come together to form isomer B . At
equilibrium, therefore, we expect
max
NB W
B
K = =
max
NA W
A
If we consider the illustrative—if somewhat unrealistic—case of isomeric molecules whose energy levels all have the same
degeneracy (g = g for all i), we can readily see that the equilibrium system must contain some amount of each isomer. For a
i
system containing N molecules, N !g is the numerator in each of the thermodynamic probabilities W

N
,W max
A
, and Wmax
B
. max
A,B
The denominators are different. The denominator of W must contain terms, N !, for essentially all of the levels represented in
max
A,B i
the denominator of W . Likewise, it must contain terms, N !, for essentially all of the energy levels represented in the
max
A j
denominator of W max
B
. Then the denominator of W is a product of N ! terms that are generally smaller than the corresponding
max
A,B
k
factorial terms in the denominators of W and W

max
A
. As a result, the denominators of W
max
B
and W
max
A
are larger than the
max
B
denominator of W max
A,B
. In consequence, W >W
max
A,B
and W
max
A
>W
max
A,B
. (See problems 5 and 6.)
B
max
If we create the system as a collection of A molecules, or as a collection of B molecules, redistribution of the sets of atoms among
all of the available energy levels must eventually produce a mixture of A molecules and B molecules. Viewed as a consequence of
the principle of equal a priori probabilities, this occurs because there are necessarily more microstates of the same energy available
to some mixture of A and B molecules than there are microstates available to either A molecules alone or B molecules alone.
Viewed as a consequence of the tendency of the isolated system to attain the state of maximum entropy, this occurs because
kln W
A,B
max
> kln W
max
A
and kln W max
A,B
> kln W
max
B
.
This page titled 20.11: Thermodynamic Probability and Equilibrium in an Isomerization Reaction is shared under a CC BY-SA 4.0 license and
In Section 20.9, we find that the sum of the probabilities of the population sets of an isolated system is
1 = ∑ W (Ni , gi ) ρMS,N ,E .
{ Ni },E
By the principle of equal a priori probabilities, ρ MS,N ,E

is a constant, and it can be factored out of the sum. We have
1 = ρMS,N ,E ∑ W (Ni , gi )
{ Ni },E
Moreover, the sum of the thermodynamic probabilities over all allowed population sets is just the number of microstates that have
energy E . This sum is just the degeneracy of the system energy, E . The symbol Ω is often given to this system-energy
E
degeneracy. That is,
ΩE = ∑ W (Ni , gi )
{ Ni },E
The sum of the probabilities of the population sets of an isolated system becomes
1 =ρ ΩE
MS,N ,E
In Section 20.9, we infer that

∞
Ni
ρMS,N ,E = ∏ ρ(ϵi )
i=1
so we have
∞
Ni
1 = ΩE ∏ ρ(ϵi )
i=1
This page titled 20.12: The Degeneracy of an Isolated System and Its Entropy is shared under a CC BY-SA 4.0 license and was authored,
In Section 20.10, we observe that the entropy of an isolated equilibrium system can be defined as S = kln W max . In Section 20.12,
we see that the system-energy degeneracy is a sum of terms, one of which is W max = W (N , gi )
i
. That is, we have
ΩE = Wmax + ∑ W (Ni , gi )
{ Ni }≠{N }, Etota l
i
where the last sum is taken over all energy-qualifying population sets other than the most-probable population set.
Let us now consider the relative magnitude of Ω and W E . Clearly, Ω ≥ W
max . If only one population set is consistent with
E max
the total-molecule and total-energy constraints of the isolated system, then Ω = W . In general, however, we must expect that
E max
there will be many, possibly an enormous number, of other population sets that meet the constraints. Ultimately, the relative
magnitude of Ω and W
E maxdepends on the energy levels available to the molecules and the number of molecules in the system
and so could be almost anything. However, rather simple considerations lead us to expect that, for most macroscopic collections of
molecules, the ratio α = Ω /W E will be much less than W
max . That is, although the value of α may be very large, for
max
macroscopic systems we expect to find α ≪ W . If Ω = W

max E , then α = 1 , and ln α = 0 .
max
Because W for any population set that contributes to Ω must be less than or equal to W
E , the maximum value of α must be
max
less than the number of population sets which satisfy the system constraints. For macroscopic systems whose molecules have even
a modest number of accessible energy levels, calculations show that W is a very large number indeed. Calculation of α for even
max
a small collection of molecules is intractable unless the number of accessible molecular energy levels is small. Numerical
experimentation on small systems, with small numbers of energy levels, shows that the number of qualifying population sets
increases much less rapidly than W as the total number of molecules increases. Moreover, the contribution that most qualifying
max
population sets make to Ω is much less than W

E .
max
For macroscopic systems, we can be confident that W is enormously greater than α . Hence Ω is enormously greater than α .
max E
When we substitute for W in the isolated-system entropy equation, we find

max
S = k ln Wmax
= k ln(ΩE /α)
= klnΩE − k ln α
≈ k ln ΩE
where the last approximation is usually very good.

In many developments, the entropy of an isolated system is defined by the equation S = kln Ω rather than the equation weE
introduced first, S = kln W . From the considerations above, we expect the practical consequences to be the same. In Section
max
20.14, we see that the approximate equality of ln W and ln Ω is a mathematical consequence of our other assumptions and
max E
approximations.
This page titled 20.13: The Degeneracy of an Isolated System and its Entropy is shared under a CC BY-SA 4.0 license and was authored, remixed,
In principle, an isolated system is different from a system with identical macroscopic properties that is in equilibrium with its
surroundings. We emphasize this point, because this distinction is important in the logic of our development. However, our
development also depends on the assumption that, when N is a number that approximates the number molecules in a macroscopic
system, the constant-temperature and constant-energy systems are functionally equivalent.
In Section 20.9, we find that any calculation of macroscopic properties must produce the same result whether we consider the
constant-temperature or the constant-energy system. The most probable population set, {N , N , … N , …}, provides an
1 2 i
adequate description of the macroscopic state of the constant-temperature system precisely because it is representative of all the
population sets that contribute significantly to the total probability of the constant-temperature system. The effective equivalence of
the constant-temperature and constant-energy systems ensures that the most probable population set is also representative of all the
population sets that contribute significantly to the total probability of the constant-energy system.
In Section 20.12, we see that the essential equivalence of the isothermal and constant-energy systems means that we have
∞
N
i
1 = ΩE ∏ ρ(ϵi )
i=1
Taking logarithms of both sides, we find

∞
ln ΩE = − ∑ N ln ρ (ϵi )
i
i=1
From S = kln Ω , it follows that

E
S = −k ∑ N ln ρ (ϵi )
i
i=1
For the constant-temperature system, we have N = N P . When we assume that the equilibrium constant-temperature and
i i
constant-energy systems are essentially equivalent, the entropy of the N-molecule system becomes
∞
S = −k ∑ N lnρ (ϵi )
i
i=1
= −kN ∑ Pi ln ρ (ϵi )
i=1
so that we obtain the same result from assuming that S = kln Ω as we do in Section 20.10 from assuming that
E S = kln Wmax .
Under the approximations we introduce, ln Ω and ln W
E evaluate to the same thing.
max
This page titled 20.14: Effective Equivalence of the Isothermal and Constant-energy Conditions is shared under a CC BY-SA 4.0 license and was
20.15: Problems
1. Three non-degenerate energy levels are available to a set of five distinguishable molecules, {A, B, C , D, E} . The energies of
these levels are 1, 2, and 3, in arbitrary units. Find all of the population sets that are possible in this system. For each population
set, find the system energy, E , and the number of microstates, W . For each system energy, E , list the associated population sets
and the total number of microstates. How many population sets are there? What is W ? If this system is isolated with E = 10 ,
max
how many population sets are possible? What is Ω for E = 10 ? E
2. For the particle in a box, the allowed energies are proportional to the squares of the successive integers. What population sets are
possible for the distinguishable molecules, {A, B, C , D, E} , if they can occupy three quantum states whose energies are 1, 4,
and 9? For each population set, find the system energy, E , and the number of microstates. For each system energy, E , list the
associated population sets and the total number of microstates. How many population sets are there? What is W ? If this system
max
is isolated with E = 24 , how many population sets are possible? What is Ω for E = 24 ? E
3. Consider the results you obtained in problem 2. In general, when the allowed energies are proportional to the squares of
successive integers, how many population sets do you think will be associated with each system energy?
4.
(a) Compare W for the population set {3, 3, 3} to W for the population set {2, 5, 2}. The energy levels are non-degenerate.
(b) Consider an N -molecule system that has a finite number, M , of quantum states. Show that W is (at least locally) a maximum
when N = N = ⋯ = N = N /M . (Hint: Let U = N /M , and assume that N can be chosen so that U is an integer. Let
1 2 M
i=M−2
WU = N !/[U !U ! ∏ U !]
i=1
and let
i=M−2
WO = N !/[(U + 1)! (U − 1)! ∏ U !]
i=1
Show that W O / WU <1 .)

5. The energy levels available to isomer A are ϵ = 1 , ϵ = 2 , and ϵ = 3 , in arbitrary units. The energy levels available to isomer
0 2 4
B are ϵ = 2 , ϵ = 3 , and ϵ = 4 . The energy levels are non-degenerate.

1 3 5
(a) A system contains five molecules. The energy of the system is 10. List the population sets that are consistent with N = 5 and
E = 10 . Find W for each of these population sets. What are W ,W , and W ? What is the total number of microstates,
max max max
A,B A B
=Ω , available to the system in all of the cases in which A and B molecules are present? What is the ratio Ω /W
A,B ? A,B
max
A,B
(b) Repeat this analysis for a system that contains six molecules and whose energy is 12.
(c) Would the ratio Ω max
A,B / WA,B be larger or smaller for a system with N = 50 and E = 100 ?
(d) What would happen to this ratio if the number of molecules became very large, while the average energy per molecule
remained the same?
6. In Section 20.11, we assume that all of the energy levels available to an isomeric pair of molecules have the same degeneracy.
We then argue that the thermodynamic probabilities of a mixture of the isomers must be greater than the thermodynamic
probability of either pure isomer: W >W
max
A,B
and W
max
A
>W . Implicitly, we assume that many energy levels are
A,B
max max
B
multiply occupied: N > 1 for many energy levels ϵ . Now consider the case that g > 1 for most ϵ , but that nearly all energy
i i i i
levels are either unoccupied or contain only one molecule: N = 0 or N = 1 . Show that under this assumption also, we must have
i i
W
max
A,B
>W and W
max
A
>W
max
A,B
. B
max
Notes
1
The statistical-mechanical procedures that have been developed for finding the energy levels available to a molecule express
molecular energies as the difference between the molecule energy and the energy that its constituent atoms have when they are
motionless. This is usually effected in two steps. The molecular energy levels are first expressed relative to the energy of the
molecule’s own lowest energy state. The energy released when the molecules is formed in its lowest energy state from the isolated
constituent atoms is then added. The energy of each level is then equal to the work done on the component atoms when they are
brought together from infinite separation to form the molecule in that energy level. (Since energy is released in the formation of a
stable molecule, the work done on the atoms and the energy of the resulting molecule are less than zero.) In our present discussion,
we suppose that we can solve the Schrödinger equation to find the energies of the allowed quantum states. This corresponds to
choosing the isolated constituent electrons and nuclei as the zero of energy for both isomers.
CHAPTER OVERVIEW
21: The Boltzmann Distribution Function

21.10: Problems
This page titled 21: The Boltzmann Distribution Function is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
1
The probabilities of the energy levels of a constant-temperature system at equilibrium must depend only on the intensive variables
that serve to characterize the equilibrium state. In Section 20.8, we introduce the principle of equal a priori probabilities, which
asserts that any two microstates of an isolated system have the same probability. From the central limit theorem, we infer that an
isolated system is functionally equivalent to a constant-temperature system when the system contains a sufficiently large number of
molecules. From these ideas, we can now find the relationship between the energy values, ϵ , and the corresponding probabilities,
i
Pi = P (ϵi ) = gi ρ (ϵi ) .
Let us consider the microstates of an isolated system whose energy is E . For any population set, {N
#
1, N2 , … , Ni , …} , that has
energy E , the following relationships apply.
#
1. The sum of the energy-level populations is the total number of molecules:

∞
N = N1 + N2 + ⋯ + Ni = ∑ Nj
j=1
2. The energy of the system is the sum of the energies of its constituent molecules:
∞
#
E = N1 ϵ1 + N2 ϵ2 + ⋯ + Ni ϵi = ∑ Nj ϵj
j=1
3. The product of powers of quantum-state probabilities is a constant:

N1 N1 N1
ρ (ϵ1 ) ρ (ϵ1 ) … ρ (ϵ1 ) ⋯ = ĸ
or, equivalently,
∞
N1 ln ρ (ϵ1 ) + N2 ln ρ (ϵ2 ) + ⋯ + Ni ln ρ (ϵi ) +… = ∑ Ni ln ρ (ϵi )
i=1
= ln ĸ
4. For the system at constant temperature, the sum of the energy-level probabilities is one. When we infer that the constant-
temperature system and the isolated system are functionally equivalent, we assume that this is true also for the isolated system:
∞
1 = P (ϵ1 ) + P (ϵ2 ) + ⋯ + P (ϵi ) + ⋯ = ∑ P (ϵj )
j=1
We want to find a function, ρ (ϵ), that satisfies all four of these conditions. One way is to keep trying functions that look like they
might work until we find one that does. A slightly more sophisticated version of this approach is to try the most general possible
version of each such function and see if any set of restrictions will make it work. We could even try an infinite series. Suppose that
we are clever (or lucky) enough to try the series solution
∞
i k
ln ρ (ϵ) = c0 + c1 ϵ + ⋯ + ci ϵ + ⋯ = ∑ ck ϵ
k=0
Then the third condition becomes
∞
ln ĸ = ∑ Ni ln ρ (ϵi )
i=1
∞ ∞
k
= ∑ Ni ∑ [ ck ϵ ]
i
i=1 k=0
∞ ∞ ∞ ∞ ∞ ∞
k 0 1 k
= ∑ ∑ ck Ni ϵ = c0 ∑ Ni ϵ + c1 ∑ Ni ϵ + ⋯ + ck ∑ ∑ Ni ϵ +…
i i i i
k=0 i=1 i=1 i=1 k=2 i=1
∞ ∞
# k
= c0 N + c1 E + ⋯ + ck ∑ ∑ Ni ϵ +…
i
k=2 i=1
We see that the coefficient of c is N and the coefficient of c is the total energy, E . Therefore, the sum of the first two terms is a
0 1
#
constant. We can make the trial function satisfy the third condition if we set c = 0 for all k > 1 . We find k
∞ ∞
ln ĸ = ∑ Ni ln ρ (ϵi ) = ∑ Ni (c0 + c1 ϵi )
i=1 i=1
The last equality is satisfied if, for each quantum state, we have
ln ρ (ϵi ) = c0 + c1 ϵi
or
ρ (ϵi ) = α exp (c1 ϵi )
where α = exp (c ) . Since the ϵ are positive and the probabilities ρ (ϵ ) lie in the interval 0 < ρ (ϵ
0 i i i) <1 , we must have c1 < 0 .
Following custom, we let c = −β , where β is a constant, and β > 0 . Then,
1
ρ (ϵi ) = α exp (−β ϵi )
and
Pi = gi ρ (ϵi ) = α gi exp (−β ϵi )
The fourth condition is that the energy-level probabilities sum to one. Using this, we have
∞ ∞
1 = ∑ P (ϵi ) = α ∑ gi exp (−β ϵi )
i=1 i=1
The sum of exponential terms is so important that it is given a name. It is called the molecular partition function. It is often
represented by the letter “z .” Letting
∞
z = ∑ gi exp (−β ϵi )
i=1
we have
1 −1
α = =z
∞
∑ gi exp (−β ϵi )
i=1
Thus, we have the Boltzmann probability:

P (ϵi ) = gi ρ (ϵi )
gi exp (−β ϵi )
=
∞
∑ gi exp (−β ϵi )
i=1
gi
= exp (−β ϵi )
z
The probability of an energy level depends only on its degeneracy, gi , its energy, ϵi , and the constant β. Since the equilibrium-
characterizing population set is determined by the probabilities, we have P i = N /N
i
, and
N gi
i
= exp (−β ϵi )
N z
In Section 21.2, we develop Lagrange’s method of undetermined multipliers. In Section 21.3, we develop the same result by
applying Lagrange’s method to our model for the probabilities of the microstates of an isolated system. That is, we find the
Boltzmann probability equation by applying Lagrange’s method to the entropy relationship,
∞
S = −N k ∑ Pi ln ρ (ϵi )
i=1
that we first develop in §20-11. In §4, we find the Boltzmann probability equation by using Lagrange’s method to find the values of
N
i
that produce the largest possible value for W in an isolated system. This argument requires us to assume that there is a very
max
large number of molecules in each of the occupied energy levels of the most probable population set. Since our other arguments do
not assume anything about the magnitude of the various N , it is evident that some of the assumptions we make when we apply
i
Lagrange’s method to find the N are not inherent characteristics of our microscopic model.
i
This page titled 21.1: Finding the Boltzmann Equation is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul
Lagrange’s method of undetermined multipliers is a method for finding the minimum or maximum value of a function subject to
one or more constraints. A simple example serves to clarify the general problem. Consider the function
2 2
z = z0 exp (x +y )
where z is a constant. This function is a surface of revolution, which is tangent to the plane z = z at (0, 0, z ). The point of
0 0 0
tangency is the minimum value of z . At any other point in the xy-plane, z (x, y) is greater than z . If either x or y becomes
0
arbitrarily large, z does also. If we project a contour of constant z onto the xy-plane, the projection is a circle of radius
1/2
2 2
r = (x +y ) .
Suppose that we introduce an additional condition; we require y = 1 − x . Then we ask for the smallest value of z consistent with
this constraint. In the xy-plane the constraint is a line of slope −1 and intercept 1. A plane that includes this line and is parallel to
the z -axis intersects the function z . As sketched in Figure 1, this intersection is a curve. Far away from the origin, the value of z at
which the intersection occurs is large. Nearer the origin, the value of z is smaller, and there is some (x, y) at which it is a
minimum. Our objective is to find this minimum.
Figure 1: A surface and a constraint equation.

There is a straightforward solution of this problem; we can substitute the constraint equation for y into the equation for z , making z
a function of only one variable, x. We have
2 2
z = z0 exp (x + (1 − x) )
2
= z0 exp (2 x − 2x + 1)
To find the minimum, we equate the derivative to zero, giving

dz 2
0 = = (4x − 2) z0 exp (2 x − 2x + 1)
dx
so that the minimum occurs at x = 1/2, y= 1/2, and
z = z0 exp (1/2)
Solving such problems by elimination of variables can become difficult. Lagrange’s method of undetermined multipliers is a
general method, which is usually easy to apply and which is readily extended to cases in which there are multiple constraints. We
can see how Lagrange’s method arises by thinking further about our particular example. We can imagine that we “walk” along the
constraint line in the xy-plane and measure the z that is directly overhead as we progress. The problem is to find the minimum
value of z that we encounter as we proceed along the line. This perspective highlights the central feature of the problem: While it is
formally a problem in three dimensions (x, y , and z ), the introduction of the constraint makes it a two-dimensional problem. We
can think of one dimension as a displacement along the line y = 1 − x , from some arbitrary starting point on the line. The other
dimension is the perpendicular distance from the xy-plane to the intersection with the surface z .
The relevant part of the xy-plane is just the one-dimensional constraint line. We can recognize this by parameterizing the line. Let t
measure location on the line relative to some initial point at which t = 0 . Then we have x = x (t) and y = y (t) and
z (x, y) = z (x (t) , y (t)) = z (t) .
The point we seek is the one at which dz/dt = 0 .

Now let us examine a somewhat more general problem. We want a general way to find the values (x, y) that minimize (or
maximize) a function h = h (x, y) subject to a constraint of the form c = g (x, y) , where c is a constant. As in our example, this
constraint requires a solution in which (x, y) are on a particular line. If we parameterize this problem, we have
h = h (x, y) = h (x (t) , y (t)) = h (t)
and
c = g (x, y) = g (x (t) , y (t)) = g (t)
Because cis a constant, dc/dt = dg/dt = 0 . The solution we seek is the point at which h is an extremum. At this point,
dh/dt = 0 . Therefore, at the point we seek, we have
dh ∂h dx ∂h dy
=( ) +( ) =0
dt ∂x dt ∂y dt
y x
and
dg ∂g dx ∂g dy
=( ) +( ) =0
dt ∂x dt ∂y dt
y x
We can multiply either of these equations by any factor, and the product will be zero. We multiply dg/dt by λ (where λ ≠ 0 ) and
subtract the result from dh/dt . Then, at the point we seek,
dh dg ∂h ∂g dx ∂h ∂g dy
0 = −λ = ( −λ ) +( −λ )
dt dt ∂x ∂x dt ∂y ∂y dt
y x
Since we can choose x (t) and y (t) any way we please, we can insure that dx/dt ≠ 0 and dy/dt ≠ 0 at the solution point. If we
do so, the terms in parentheses must be zero at the solution point.
Conversely, setting
∂h ∂g
( −λ ) =0
∂x ∂x
y
and
∂h ∂g
( −λ ) =0
∂y ∂y
x
is sufficient to insure that

dh dg
=λ
dt dt
Since dg/dt = 0 , these conditions insure that dh/dt = 0 . This means that, if we can find a set {x, y, λ} satisfying
∂h ∂g
( −λ ) =0
∂x ∂x
y
and
∂h ∂g
( −λ ) =0
∂y ∂y
x
and
c − g (x, y) = 0
then the values of x and y must be those make h (x, y) an extremum, subject to the constraint that c = g (x, y) . We have not shown
that the set {x, y, λ} exists, but we have shown that if it exists, it is the desired solution.
A useful mnemonic simplifies the task of generating the family of equations that we need to use Lagrange’s method. The
mnemonic calls upon us to form a new function, which is a sum of the function whose extremum we seek and a series of additional
terms. There is one additional term for each constraint equation. We generate this term by putting the constraint equation in the
form c − g (x, y) = 0 and multiplying by an undetermined parameter. For the case we just considered, the mnemonic function is
Fmn = h (x, y) + λ (c − g (x, y))
We can generate the set of equations that describe the solution set, {x, y, λ}, by equating the partial derivatives of Fmn with
respect to x, y , and λ to zero. That is, the solution set satisfies the simultaneous equations
∂Fmn
=0
∂x
∂Fmn
=0
∂y
and
∂Fmn
=0
∂λ
If there are multiple constraint equations, cλ − gλ (x, y) = 0 , cα − gα (x, y) = 0 , and cβ − gβ (x, y) = 0 , then the mnemonic
function is
Fmn = h (x, y) + λ (cλ − gλ (x, y)) + α (cα − gα (x, y)) + β (cβ − gβ (x, y))
and the simultaneous equations that represent the constrained extremum are
∂ Fmn /∂x = 0 ,
∂ Fmn /∂y = 0 ,
∂ Fmn /∂λ = 0 ,
∂ Fmn /∂α = 0 , and
∂ Fmn /∂β = 0 .
To illustrate the use of the mnemonic, let us return to the example with which we began. The mnemonic equation is
2 2
Fmn = z0 exp (x + y ) + λ (1 − x − y)
so that
∂Fmn
2 2
= 2x z0 exp (x + y ) − λ = 0,
∂x
∂Fmn 2 2
= 2y z0 exp (x +y )−λ = 0
∂y
and
∂Fmn
= 1 −x −y = 0
∂λ
which yield x = 1/2, y= 1/2, and λ = z 0 exp (1/2) .
This page titled 21.2: Lagrange's Method of Undetermined Multipliers is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
In Sections 20-10 and 20-14, we develop the relationship between the system entropy and the probabilities of a microstate, ρ (ϵ ), i
and an energy level, P = g ρ (ϵ ) , in our microscopic model. We find

i i i
i=1
= −N k ∑ gi ρ (ϵi )ln ρ (ϵi )
i=1
For an isolated system at equilibrium, the entropy must be a maximum, and hence
∞
− ∑ gi ρ (ϵi )ln ρ (ϵi ) (21.3.1)
i=1
must be a maximum. We can use Lagrange’s method to find the dependence of the quantum-state probability on its energy. The
ρ (ϵ ) must be such as to maximize entropy (Equation 21.3.1) subject to the constraints
i
∞ ∞
1 = ∑ Pi = ∑ gi ρ (ϵi )
i=1 i=1
and
∞ ∞
⟨ϵ⟩ = ∑ Pi ϵi = ∑ gi εi ρ (ϵi )
i=1 i=1
where ⟨ϵ⟩ is the expected value of the energy of one molecule. The mnemonic function becomes
∞ ∞ ∞
∗
Fmn = − ∑ gi ρ (ϵi )ln ρ (ϵi ) +α (1 − ∑ gi ρ (ϵi )) + β (⟨ϵ⟩ − ∑ gi εi ρ (ϵi ))
i=1 i=1 i=1
Equating the partial derivative with respect to ρ (ϵ ) to zero,

i
∂Fmn
∗
= −gi ln ρ (ϵi ) − gi − α gi − β gi ϵi = 0
∂ρ (ϵi )
so that
∗
ρ (ϵi ) = exp (−α − 1) exp (−β ϵi )
From
∞ ∞
1 = ∑ Pi = ∑ gi ρ (ϵi )
i=1 i=1
the argument we use in Section 21.1 again leads to the partition function, z , and the Boltzmann equation
−1
Pi = gi ρ (ϵi ) = z gi exp (−β ϵi )
This page titled 21.3: Deriving the Boltzmann Equation I is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
In Section 20-9, we find that the probability of the population set {N 1, N2 , … , Ni , …} in an isolated system is
∞ Ni
g
i
ρ = N! ∏
MS,N ,E
Ni !
i=1
The thermodynamic probability

∞ Ni
g
i
W (Ni , gi ) = N ! ∏
Ni !
i=1
is the number of microstates of the population set. ρ is the constant probability of any one microstate. In consequence, as we
MS,N ,E
see in Section 20.10, the probability of a population set is proportional to its thermodynamic probability, W (N , g ). It follows that i i
the most probable population set is that for which W (N , g ) is a maximum. Our microscopic model asserts that the most probable
i i
population set, {N , N , … , N , …}, characterizes the equilibrium state, because the equilibrium system always occupies the
1 2 i
either the most probable population set or another population set whose macroscopic properties are indistinguishable from those of
the most probable one.
Evidently, the equilibrium-characterizing population set is the one for which W (N , g ), or ln W (N , g ) , is a maximum. Let us i i i i
assume that the N are very large so that we can treat them as continuous variables, and we can use Stirling’s approximation for
i
N ! . Then we can use Lagrange’s method of undetermined multipliers to find the most probable population set by finding the set,
i
N , N , … , N , …, for which ln W (N , g ) is a maximum, subject to the constraints

1 2 i i i
N = ∑ Ni
i=1
and
∞
E = ∑ Ni ϵi .
i=1
From our definition of the system, both N and E are constant. The mnemonic function is
N1 N2 Ni ∞ ∞
N !g g …g …
1 2 i
Fmn = ln( ) + α (N − ∑ Ni ) + β (E − ∑ Ni ϵi )
N1 ! N2 ! … Ni ! …
i=1 i=1
∞ ∞ ∞ ∞ ∞
≈ N ln N − N − ∑ Ni ln Ni + ∑ Ni + ∑ Ni ln gi + α (N − ∑ Ni ) + β (E − ∑ Ni ϵi )
i=1 i=1 i=1 i=1 i=1
Taking the partial derivative with respect to N gives i
∂Fmn 1
= −Ni ( ) − ln Ni + 1 + ln gi − α − β ϵi = −ln Ni + ln gi − α − β ϵi
∂Ni N1
from which we have, for the population set with the largest possible thermodynamic probability,
−ln N + ln gi − α − β ϵi = 0
i
or
N = gi exp (−α) exp (−β ϵi )

i
We can again make use of the constraint on the total number of molecules to find exp (−α) :
∞ ∞
N = ∑N = exp (−α) ∑ gi exp (−β ϵi )

i
i=1 i=1
so that exp (−α) = N z , where z is the partition function,
−1
z =∑
∞
i=1
gi exp (−β ϵi ) . Therefore, in the most probable
population set, the number of molecules having energy ϵ is i
−1
N = Nz gi exp (−β ϵi )
i
The fraction with this energy is
N
i −1
=z gi exp (−β ϵi )
N
This fraction is also the probability of finding an arbitrary molecule in one of the quantum states whose energy is ϵ . When the i
isolated system and the corresponding constant-temperature system are functionally equivalent, this probability is P . As in the two i
previous analyses, we have

Pi = gi ρ (ϵi )
−1
=z gi exp (−β ϵi ) .
This derivation of Boltzmann’s equation from W is the most common introductory treatment. It relies on the assumption that all
max
of the N are large enough to justify treating them as continuous variables. This assumption proves to be invalid for many
i
important systems. (For ideal gases, we find that N = 0 or N = 1 for nearly all of the very large number of energy levels that are
i i
available to a given molecule.) Nevertheless, the result obtained is clearly correct; not only is it the same as the result of our two
previous arguments, but also it leads to satisfactory agreement between microscopic models and the macroscopic properties of a
wide variety of systems.
This page titled 21.4: Deriving the Boltzmann Equation II is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
In Section 20.11, we discuss chemical equilibrium between isomers from the perspective afforded by Boltzmann’s definition of
entropy. Now, let us consider equilibrium in this system from the perspective afforded by the energy-level probabilities. Let us
assign even-integer labels to energy levels of isomer A and odd-integer labels to energy levels of isomer B . A group of atoms that
can arrange itself into either a molecule of A or a molecule of B can occupy any of these energy levels. The partition function for
this group of molecules to which all energy levels are available is
∞
zA+B = ∑ gi exp (−β ϵi )
i=1
The fraction of molecules in the first (odd) energy level associated with molecules of isomer B is
N
1 −1
= g1 (zA+B ) exp (−β ϵ1 )
NA+B
and the fraction in the next is
N
3 −1
= g3 (zA+B ) exp (−β ϵ3 )
NA+B
The total number of B molecules is
N = ∑ Ni
B
i odd
so that the fraction of all of the molecules that are B molecules is
N
B −1
= (zA+B ) ∑ gi exp (−β ϵi ) = zB / zA+B
NA+B
i odd
Likewise, the fraction that is A molecules is
N
A −1
= (zA+B ) ∑ gi exp (−β ϵi ) = zA / zA+B
NA+B
i even
The equilibrium constant for the equilibrium between A and B is
N zB
B
Keq = =
zA
N
A
We see that the equilibrium constant for the isomerization reaction is simply equal to the ratio of the partition functions of the
isomers.
It is always true that the equilibrium constant is a product of partition functions for reaction-product molecules divided by a product
of partition functions for reactant molecules. However, the partition functions for the various molecules must be expressed with a
common zero of energy. Choosing the infinitely separated component atoms as the zero-energy state for every molecule assures
that this is the case. However, it is often convenient to express the partition function for a molecule by measuring each molecular
energy level, ϵ , relative to the lowest energy state of that isolated molecule. When we do this, the zero of energy is different for
i
each molecule.
To adjust the energies in a molecule’s partition function so that they are expressed relative to the energy of the molecule’s infinitely
separated atoms, we must add to each molecular energy the energy required to take the molecule from its lowest energy state to its
isolated component atoms. If z is the partition function when the ϵ are measured relative to the lowest energy state of the isolated
i
molecule, Δϵ is the energy released when the isolated molecule is formed from its component atoms, and z is the partition ∗
function when the ϵ are measured relative to the molecule’s separated atoms, we have z = z exp (+Δϵ/kT ) .
i
∗
This page titled 21.5: Partition Functions and Equilibrium - Isomeric Molecules is shared under a CC BY-SA 4.0 license and was authored,
All of a substance’s thermodynamic functions can be derived from the molecular partition function. We begin with the entropy. We
consider closed (constant N) systems of independent, distinguishable molecules in which only pressure–volume work is possible.
In Sections 20.10 and 20.14, we find that two different approaches give the entropy of this system,
∞
S = −N k ∑ Pi ln ρ (ϵi ) .
i=1
In Sections 20.1, 20.3, and 20.4, we find that three different approaches give the Boltzmann equation,
−1
Pi = gi ρ (ϵi ) = z gi exp (−β ϵi ) .
We have
ln ρ (ϵi ) = − ln z − β ϵi
Substituting, and recognizing that the energy of the N-molecule system is E = N ⟨ϵ⟩ , we find that the entropy of the system is
∞ ∞ ∞
S = kN ∑ Pi [ln z + β ϵi ] = kN ln z ∑ Pi + kβN ∑ Pi ϵi = kN ln z + kβE
i=1 i=1 i=1
In Section 10.1, we find that the fundamental equation implies that

∂E
( ) =T
∂S
V
Since the ϵ are fixed when the volume and temperature of the system are fixed, ln z is constant when the volume and temperature
i
of the system are constant. Differentiating S = kN ln z + kβE with respect to S at constant V , we find
∂E
1 = kβ ( ) = kβT
∂S
V
so that
1
β =
kT
This is an important result: Because we have now identified all of the parameters in our microscopic model, we can write the
results we have found in forms that are more useful:
∞
−ϵi
1. z = ∑ gi exp ( )
kT
i=1

molecular partition function
−ϵi
2. Pi = gi ρ (ϵi ) = z
−1
gi exp ( )
kT

Boltzmann’s equation
E
3. S = kN ln z +
T

Entropy of an N-molecule system
To express the system energy in terms of the molecular partition function, we first observe that
∞ ∞
−1
−ϵi
E = N ⟨ϵ⟩ = N ∑ Pi ϵi = N z ∑ gi ϵi exp ( )
kT
i=1 i=1
Then we observe that
∞
∂ ln z −1
ϵi −ϵi
( ) =z ∑ gi ( ) exp ( )
2
∂T kT kT
V i=1
∞
1 −1
−ϵi
=( ) Nz ∑ gi ϵi exp ( )
2
N kT kT
i=1
E
=
2
N kT
The system energy becomes

∂ln z
2
E = N kT ( )
∂T
V

energy of an N-molecule system
By definition, A = E − T S . Rearranging our entropy result, S = kN ln z + E/T , we have E − T S = −N kT ln z . Thus,
A = −N kT ln z
(Helmholtz free energy of an N-molecule system)

From dA = −SdT − P dV , we have
∂A
( ) = −P
∂V
T
(Here, of course, P is the pressure of the system, not a probability.) Differentiating A = −N kT ln z with respect to V at constant
T , we find
∂ln z
P = N kT ( )
∂V
T
(pressure of an N-molecule system)

The pressure–volume product becomes
∂ln z
P V = N kT V ( )
∂V T
Substituting into H = E + P V , the enthalpy becomes

∂ln z ∂ln z
H = N kT [T ( ) +V ( ) ]
∂T ∂V
V T
(enthalpy of an N-molecule system)

The Gibbs free energy is given by G = A + P V . Substituting, we find
∂ln z
G = −N kT ln z + N kT V ( )
∂V
T
(Gibbs free energy of an N-molecule system)

The chemical potential can be found from
∂A
μ =( )
∂n
V ,T
At constant volume and temperature, kT ln z is constant.

¯
¯¯¯
¯
Substituting N = nN into A = −N kT ln z and taking the partial derivative, we find
¯
¯¯¯
¯
μ = −N kT ln z = −RT ln z

chemical potential of distinguishable molecules
In statistical thermodynamics we frequently express the chemical potential per molecule, rather than per mole; then,
∂A
μ =( )
∂N
V ,T
and
μ = −kT ln z
(chemical potential per molecule)
This page titled 21.6: Finding ß and the Thermodynamic Functions for Distinguishable Molecules is shared under a CC BY-SA 4.0 license and
Now let us return to the closed (constant-N ) system to develop another perspective on the dependence of its macroscopic
thermodynamic properties on the molecular energy levels and their probabilities. We undertake to describe the system using
volume and temperature as the independent variables. In thinking about the energy-level probabilities, we stipulate that any
parameters that affect the state of the system remain constant. Specifically, we mean that any parameters that appear in the
Schrödinger equation remain constant. For example, the energy levels of a particle in a box depend on the mass of the particle and
the length of the box. Any such parameter is called an exogenous variable. If we change an exogenous variable (say the length of
the box) by a small amount, all of the energy levels change by a small amount, and all of the probabilities change by a small
amount. The energy levels and their probabilities are smooth functions of the exogenous variable. If ξ is the exogenous variable, we
have
Pi = P (ϵi ) = gi ρ (ϵi (ξ))
A change in the exogenous variable corresponds to a reversible macroscopic process.

For a particle in a box, the successive ψ are functions that depend on the quantum number, i, and the length of the box, ℓ . When
i
we change the length of the box, the wavefunction and its associated energy both change. Both are continuous functions of the
length of the box. The energy is
2 2
i h
ϵi =
2
8mℓ
Changing the length of the box is analogous to changing the volume of a system. A reversible volume change entails work. We see
that changing the length of the box does work on the particle-in-a-box, just as changing the volume of a three-dimensional system
does work on the system.
Temperature plays a central role in the description of equilibrium from the macroscopic perspective. We can see that temperature
enters the description of equilibrium from the microscopic perspective through its effect on the probability factors. When we
increase the temperature of a system, its energy increases. The average energy of its molecules increases. The probability of an
energy level must depend on temperature. Evidently, the probabilities of energy levels that are higher than the original average
energy increase when the temperature increases. The probabilities of energy levels that are lower than the original average energy
decrease when the temperature increases. The effects of heat and work on the energy levels and their equilibrium populations are
diagrammed in Figure 21.7.1.
Figure 21.7.1 : The effects of heat and work on energy levels and their populations.
If our theory is to be useful, the energy we measure for a macroscopic system must be indistinguishably close to the expected value
of the system energy as calculated from our microscopic model:
∞
Eexperiment ≈ ⟨E⟩ = N ⟨ϵ⟩ = N ∑ Pi ϵi
i=1
We can use this equation to relate the probabilities, P , to other thermodynamic functions. Dropping the distinction between the
i
experimental and expected energies, and assuming that the ϵ and the P are continuous variables, we find the total differential
i i
∞ ∞
dE = N ∑ ϵi dPi + N ∑ Pi dϵi
i=1 i=1
This equation is important because it describes a reversible macroscopic process in terms of the microscopic variables ϵ and P . i i
Let us consider the first term. Since N is a constant, we have from N i

= Pi N that dN i
= N dPi . Substituting, we have
∞ ∞
(dE) = N ∑ ϵi dPi = ∑ ϵi dN
ϵi i
i=1 i=1
This asserts that the energy of the system changes if we redistribute the molecules among the various energy levels. If the
redistribution takes molecules out of lower energy levels and puts them into higher energy levels, the energy of the system
increases. This is our statistical-mechanical picture of the shift in the equilibrium position that occurs when we heat a system of
independent molecules; the allocation of molecules among the available energy levels shifts to put more molecules in higher energy
levels and fewer in lower ones. This corresponds to an increase in the temperature of the macroscopic system.
In terms of the macroscopic system, the first term represents an increment of heat added to the system in a reversible process; that
is,
∞
rev
dq = N ∑ ϵi dPi
i=1
∞
The second term, N ∑ P dϵ , is a contribution to the change in the energy of the system from reversible changes in the energy
i=1 i i
of the various quantum states, while the number of molecules in each quantum state remains constant. This term corresponds to a
process in which the quantum states (and their energies) evolve in a continuous way as the state of the system changes. The second
term represents an increment of work done on the system in a reversible process; that is
∞
rev
dw = N ∑ Pi dϵi
i=1
Evidently, the total differential expression for dE is the fundamental equation of thermodynamics expressed in terms of the
variables we use to characterize the molecular system. It enables us to relate the variables that characterize our microscopic model
of the molecular system to the variables that characterize the macroscopic system.
For a system in which the reversible work is pressure–volume work, the energy levels depend on the volume. At constant
temperature we have
∞ ∞
rev
∂ϵi
dw = −P dV = N ∑ Pi dϵi = N ∑ Pi ( ) dV
∂V
i=1 i=1 T
so that the system pressure, P , is related to the energy-level probabilities, P , as i
∞
∂ϵi
P = −N ∑ Pi ( )
∂V
i=1 T
To evaluate the pressure, we must know how the energy levels depend on the volume of the system.
∞
The first term relates the entropy to the energy-level probabilities. Since dq rev
= T dS = N ∑
i=1
ϵi dPi , we have
∞
N
dS = ∑ ϵi dPi
T
i=1
From the Boltzmann distribution function we have

Pi = z
−1
gi exp (−ϵi /kT ) , or
ϵi = −kT ln Pi + kT ln gi − kT ln z
Substituting into our expression for dS , we find
∞ ∞ ∞
dS = −N k ∑ (ln Pi )dPi + N k ∑ (ln gi )dPi − N k (ln z) ∑ dPi
i=1 i=1 i=1
Since ∑ ∞
i=1
Pi = 1 , we have ∑ ∞
i=1
dPi = 0 , and the last term vanishes. Also,
∞ ∞ ∞ ∞
∑ d (Pi ln Pi ) = ∑ (ln Pi ) dPi + ∑ dPi = ∑ (ln Pi ) dPi
i=1 i=1 i=1 i=1
so that
∞ ∞
dS = −N k ∑ d (Pi ln Pi ) + N k ∑ (ln gi )dPi
i=1 i=1
At any temperature, the probability ratio for any two successive energy levels is
Pi+1 (T ) Pi+1 gi+1 − (ϵi+1 − ϵi )
= = exp ( )
Pi (T ) Pi gi kT
In the limit as the temperature goes to zero,

Pi+1
→ 0
Pi
It follows that P (0) = 1 and P (0) = 0 for i > 1 . Integrating from T = 0 to T , the entropy of the system goes from S (0) = S
1 i 0
to S (T ) , and the energy-level probabilities go from P (0) to P (T ) . We have i i
S(T ) ∞ Pi (T ) ∞ Pi (T )
∫ dS = −N k ∑ ∫ d (Pi ln Pi ) + N k ∑ ∫ (ln gi ) dPi

S0 Pi (0) Pi (0)
i=1 i=1
so that
∞ ∞
S (T ) − S0 = −N k ∑ Pi (T ) ln Pi (T ) + N kP1 (0) ln P1 (0) + N k ∑ (ln gi ) Pi (T ) − N k (ln g1 ) P1 (0)
i=1 i=1
Since P 1 (0) = 1 , ln P
1 (0) vanishes. The entropy change becomes
∞ ∞
S (T ) − S0 = −N k ∑ Pi [ln Pi − ln gi ] − N k ln g1 = −N k ∑ Pi ln ρ (ϵi ) − N k ln g1
i=1 i=1
We have S 0 = N k ln g1 . If g1 =1 , the lowest energy level is non-degenerate, and S 0 =0 ; then we have

∞
i=1
This is the entropy of an N -molecule, constant-volume, constant-temperature system that is in thermal contact with its
surroundings at the same temperature. We obtain this same result in Sections 20.10 and 20.14 by arguments in which we assume
that the system is isolated. In all of these arguments, we assume that the constant-temperature system and its isolated counterpart
are functionally equivalent; that is, a group of population sets that accounts for nearly all of the probability in one system also
accounts for nearly all of the probability in the other.
Because we obtain this result by assuming that the system is composed of N , independent, non-interacting, distinguishable
molecules, the entropy of this is system is N times the entropy contribution of an individual molecule. We can write
∞
Smolecule = −k ∑ Pi ln ρ (ϵi )
i=1
This page titled 21.7: The Microscopic Model for Reversible Change is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
In Section 21.7, we obtain the entropy by a definite integration. We take the lower limits of integration, at T = 0 , as P (0) = 1 1
and P (0) = 0 , for i > 1 . In doing so, we apply the third law of thermodynamics, which states that the entropy of a perfect crystal
i
can be chosen to be zero when the temperature is at absolute zero. The idea behind the third law is that, at absolute zero, the
molecules of a crystalline substance all are in the lowest energy level that is available to them. The probability that a molecule is in
the lowest energy state is, therefore, P = 1 , and the probability that it is any higher energy level, i > 1 , is P = 0 .
1 i
While the fact is not relevant to the present development, we note in passing that the energy of a perfect crystal is not zero at
absolute zero. While all of the constituent particles will be in their lowest vibrational energy levels at absolute zero, the energies of
these lowest vibrational levels are not zero. In the harmonic oscillator approximation, the lowest energy possible for each oscillator
is hν /2. (See Section 18.5).
By a perfect crystalline substance we mean one in which the lowest energy level is non-degenerate; that is, for which g 1 =1 . We
see that our entropy equation conforms to the third law when we let
S0 = N k ln g1
so that S 0 =0 when g
1 =1 .
Let us consider a crystalline substance in which the lowest energy level is degenerate; that is, one for which g 1 >1 . This substance
is not a perfect crystal. In this case, the temperature-zero entropy is
S0 = N k ln g1 > 0
The question arises: How can we determine whether a crystalline substance is a perfect crystal? In Chapter 11, we discuss the use
of the third law to determine the absolute entropy of substances at ordinary temperatures. If we assume that the substance is a
perfect crystal at zero degrees when it is not, our theory predicts a value for the absolute entropy at higher temperatures that is too
small, because it does not include the term S = N k ln g . When we use this too-small absolute entropy value to calculate entropy
0 1
changes for processes involving the substance, the results do not agree with experiment.
Absolute entropies based on the third law have been experimentally determined for many substances. As a rule, the resulting
entropies are consistent with other experimentally observed entropy changes. In some cases, however, the assumption that the
entropy is zero at absolute zero leads to absolute entropy values that are not consistent with other experiments. In these cases, the
absolute entropies can be brought into agreement with other entropy measurements by assuming that, indeed, g > 1 for such 1
substances. In any particular case, the value of g that must be used is readily reconciled with other information about the
1
substance.
For example, the third law entropy for carbon monoxide must be calculated taking g = 2 in order to obtain a value that is
1
consistent with other entropy measurements. This observation is readily rationalized. In perfectly crystalline carbon monoxide, all
of the carbon monoxide molecules point in the same direction, as sketched in Figure 11-2. However, the two ends of the carbon
monoxide molecule are very similar, with the consequence that the carbon monoxide molecules in the crystal point randomly in
either of two directions. Thus there are two (approximately) equally energetic states for a carbon monoxide molecule in a carbon
monoxide crystal at absolute zero, and we can take g = 2 . (We are over-simplifying here. We explore this issue further in Section
1
22-7.)
This page titled 21.8: The Third Law of Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
At a given temperature, the Boltzmann equation gives the probability of finding a molecule in any of the energy levels that the
molecule can occupy. Throughout our development, we assume that there are no energies of interaction among the molecules of the
system. The molecular partition function contains information about the energy levels of only one molecule. We obtain equations
for the thermodynamic functions of an N -molecule system in terms of this molecular partition function. However, since these
results are based on assigning the same isolated-molecule energy levels to each of the molecules, they do not address the real-
system situation in which intermolecular interactions make important contributions to the total energy of the system.
As we mention in Sections 20.1 and 20.3, the ensemble theory of statistical thermodynamics extends our arguments to express the
thermodynamic properties of a macroscopic system in terms of all of the total energies that are available to the macroscopic
system. The molecular origins of the energies of the system enter the ensemble treatment only indirectly. The theory deals with the
relationships between the possible values of the energy of the system and its thermodynamic state. How molecular energy levels
and intermolecular interactions give rise to these values of the system energy becomes a separate issue. Fortunately, ensemble
theory just reuses—from a different perspective—all of the ideas we have just studied. The result is just the Boltzmann equation,
again, but now the energies that appear in the partition function are the possible energies for the collection of N molecules, not the
energies available to a single molecule.
This page titled 21.9: The Partition Function for a System of N Molecules is shared under a CC BY-SA 4.0 license and was authored, remixed,
21.10: Problems
1. Consider a system with three non-degenerate quantum states having energies ϵ = 0.9 kT , ϵ = 1.0 kT , and ϵ = 1.1 kT . The
1 2 3
system contains N = 3 × 10 molecules. Calculate the partition function and the number of molecules in each quantum state
10
when the system is at equilibrium. This is the equilibrium population set {N , N , N }. Let W be the number of microstates
1 2 3 mp
associated with the equilibrium population set. Consider the population set when 10 of the molecules in ϵ are moved to each of
−5
2
ϵ1 and ϵ . This is the population set {N + 10 N , N − 2 × 10 , N + 10 N } . Let W be the number of microstates

3
1
−5
2 2
−5
3
−5
associated with this non-equilibrium population set.

(a) What percentage of the molecules are moved in converting the first population set into the second?
(b) How do the energies of these two populations sets differ from one another?
(c) Find W mp /W . Use Stirling’s approximation and carry as many significant figures as your calculator will allow. You need at
least six.
(d) What does this calculation demonstrate?
2. Find the approximate number of energy levels for which ϵ < kt > for a molecule of molecular weight 40 in a box of volume
m at 300 K.
−6 3
10
3. The partition function plays a central role in relating the probability of finding a molecule in a particular quantum state to the
energy of that state. The energy levels available to a particle in a one-dimensional box are
2 2
n h
ϵn =
2
8mℓ
where m is the mass of the particle and ℓ is the length of the box. For molecular masses and boxes of macroscopic lengths, the
factor h /8mℓ is a very small number. Consequently, the energy levels available to a molecule in such a box can be considered to
2 2
be effectively continuous in the quantum number, n . That is, the partition function sum can be closely approximated by an integral
in which the variable of integration, n , runs from 0 to ∞.
(a) Obtain a formula for the partition function of a particle in a one-dimensional box. Integral tables give
∞
−−−−
2
∫ exp (−an ) dn = √π/4a
0
(b) The expected value of the energy of a molecule is given by
2
∂ln z
⟨ϵ⟩ = kT ( )
∂T
V
What is ⟨ϵ⟩ for a particle in a box?

(c) The relationship between the partition function and the per-molecule Helmholtz free energy is A = −kT ln z . For a molecule
in a one-dimensional box, we have dA = −SdT − ρℓ , where ρ is the per-molecule “pressure” on the ends of the box and ℓ is the
length of the box. (The increment of work associated with changing the length of the box is dw = −ρ dℓ . In this relationship, dℓ is
the incremental change in the length of the box and ρ is the one-dimensional “pressure” contribution from each molecule. ρ is, of
course, just the force required to push the end of the box outward by a distance dℓ . ρdℓ is the one-dimensional analog of P dV .)
For the one-dimensional system, it follows that
∂A
ρ = −( )
∂ℓ
T
Use this information to find ρ for a molecule in a one-dimensional box.

(d) We can find ρ for a molecule in a one-dimensional box in another way. The per-molecule contribution to the pressure of a three-
dimensional system is related to the energy-level probabilities, P , by i
∞
system ∂ϵn
P = − ∑ Pn ( )
molecule
∂V
n=1 T
By the same argument we use for the three-dimensional case, we find that the per-molecule contribution to the “pressure” inside a
one-dimensional box is
∞
∂ϵn
ρ = − ∑ Pn ( )
∂ℓ
n=1 T
From the equation for the energy levels of a particle in a one dimensional box, find an equation for
∂ϵn
( )
∂ℓ
T
(Hint: We can express this derivative as a simple multiple of ϵ .)n
(e) Using your result from part (d), show that the per molecule contribution, ρ, to the “one-dimensional pressure” of N molecules
in a one-dimensional box is
ρ = 2 ⟨ϵ⟩/ℓ
(f) Use your results from parts (b) and (e) to express ρ as a function of k , T , and ℓ .
(g) Let Π be the pressure of a system of N molecules in a one-dimensional box. From your result in part (c) or part (f), give an
equation for Π . Show how this equation is analogous to the ideal gas equation.
CHAPTER OVERVIEW
22: Some Basic Applications of Statistical Thermodynamics

22.3: Probability Density Functions from the Energies of Classical-mechanical Models
22.8: Problems
This page titled 22: Some Basic Applications of Statistical Thermodynamics is shared under a CC BY-SA 4.0 license and was authored, remixed,
1
When it is a good approximation to say that the energy of a molecule is the sum of translational, rotational, vibrational, and
electronic components, we have
ϵi,j,k,m = ϵt,i + ϵr,j + ϵv,k + ϵe,m
where the indices i, j , k , and m run over all possible translational, rotational, vibrational, and electronic quantum states,
respectively. Then the partition function for the molecule can be expressed as a product of the individual partition functions z , z , t r
z , and z ; that is,

v e
−ϵi,j,k,m
zmolecule = ∑ ∑ ∑ ∑ gt,i gr,j gv,kge,m exp ( )
kT
t r v e
−ϵt,i −ϵr,j −ϵv,k −ϵe,m

= ∑ gt,i exp ( ) ∑ gr,j exp ( ) ∑ gv,kexp ( ) ∑ ge,m exp ( )
kT kT kT kT
t r v e
= zt zr zv ze
The magnitude of an individual partition function depends on the magnitudes of the energy levels associated with that kind of
motion. Table 1 gives the contributions made to their partition functions by levels that have various energy values.
Table 1:
−ϵi −ϵi
ϵi
kT
exp (
kT
) Type of Motion
−2 −2
10 kT −10 0.990
−1 −1
10 kT −10 0.905
kT −1 0.365 translational
5 kT −5 0.0067 rotational
10 kT −10 4.5 × 10
−5
vibration
100 kT −100 3.7 × 10

−44
electronic
We see that only quantum states whose energy is less than kT can make substantial contributions to the magnitude of a partition
function. Very approximately, we can say that the partition function is equal to the number of quantum states for which the energy
is less than kT . Each such quantum state will contribute approximately one to the sum that comprises the partition function; the
contribution of the corresponding energy level will be approximately equal to its degeneracy. If the energy of a quantum state is
large compared to kT , the fraction of molecules occupying that quantum state will be small. This idea is often expressed by saying
that such states are “unavailable” to the molecule. It is then said that the value of the partition function is approximately equal to
the number of available quantum states. When most energy levels are non-degenerate, we can also say that the value of the
partition function is approximately equal to the number of available energy levels.
This page titled 22.1: Interpreting the Partition Function is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
The Boltzmann equation gives the equilibrium fraction of particles in the i th
energy level, ϵ , as
i
N gi −ϵi
i
= exp ( )
N z kT
so the fraction of particles in energy levels less than ϵ is n
n−1
−1
−ϵi
f (ϵn ) = z ∑ gi exp ( )
kT
i=1
where z = ∑ g exp (ϵ /kT ) . We can represent either of these sums as the area under a bar graph, where the height and width
∞
i=1 i i
of each bar are g exp (ϵ /kT ) and unity, respectively. If g and ϵ can be approximated as continuous functions, this area can be
i i i i
approximated as the area under the continuous function y (i) = g exp (ϵ /kT ) . That is, i i
n−1 n
−ϵi −ϵi
∑ gi exp ( ) ≈∫ gi exp ( )di
kT i=0
kT
i=1
To evaluate this integral, we must know how both g and ϵ depend on the quantum number, i.
i i
Let us consider the case in which g = 1 and look at the constraints that the ϵ must satisfy in order to make the integral a good
i i
approximation to the sum. The graphical description of this case is sketched in Figure 1. Since ϵ > ϵ > 0 , we have i i−1
−ϵi−1 /kT −ϵi /kT

e −e >0
For the integral to be a good approximation, we must have

−ϵi−1 /kT −ϵi−1 /kT −ϵi /kT
e ≫ e −e > 0,
which means that

−Δϵ/kT
1 ≫ 1 −e >0
where Δϵ = ϵ i − ϵi−1 . Now,

2 3
x x
x
e ≈ 1 +x + + +…
2! 3!
so that the approximation will be good if

Δϵ
1 ≫ 1 − (1 − + …)
kT
or
Δϵ
1 ≫
kT
or
kT ≫ Δϵ
We can be confident that the integral is a good approximation to the exact sum whenever there are many pairs of energy levels, ϵ i
and ϵ , that satisfy the condition

i−1
Δϵ = ϵi − ϵi−1 ≪ kT .
If there are many energy levels that satisfy ϵ ≪ kT , there are necessarily many intervals, Δϵ, that satisfy Δϵ ≪ kT . In short, if a
i
large number of the energy levels of a system satisfy the criterion ϵ ≪ kT , we can use integration to approximate the sums that
appear in the Boltzmann equation. In Section 24.3, we use this approach and the energy levels for a particle in a box to find the
partition function for an ideal gas.
This page titled 22.2: Conditions under which Integrals Approximate Partition Functions is shared under a CC BY-SA 4.0 license and was
22.3: Probability Density Functions from the Energies of Classical-mechanical
Models
Guided by our development of the Maxwell-Boltzmann probability density function for molecular velocities, we could postulate
that similar probability density functions apply to other energies derived from classical-mechanical models for molecular motion.
We will see that this can indeed be done. The results correspond to the results that we get from the Boltzmann equation, where we
assume for both derivations that many energy levels satisfy ϵ ≪ kT . The essential point is that, at a sufficiently high temperature,
the behavior predicted by the quantum mechanical model and that predicted from classical mechanics converge. This high-
temperature approximation is a good one for translational motions but a very poor one for vibrational motions. These results further
illuminate the differences between the classical-mechanical and the quantum-mechanical models for the behavior of molecules.
Let us look at how we can generate probability density functions based on the energies of classical-mechanical models for
molecular motions. In the classical mechanical model, a particle moving in one dimension with velocity v has kinetic energy
m v /2. From the discussion above, if many velocities satisfy kT ≫ m v /2 , we can postulate a probability density function of the
2 2
form
2
df −mv
= Btrans exp ( )
dv 2kT
where B trans is fixed by the condition

∞ ∞ 2
df −mv
∫ ( ) dv = Btrans ∫ exp ( )dv = 1
−∞
dv −∞
2kT
Evidently, this postulate assumes that each velocity constitutes a quantum state and that the degeneracy is the same for all
velocities. This assumption is successful for one-dimensional translation, but not for translational motion in two or three
dimensions. The definite integral is given in Appendix D. We find
1/2
Btrans = (m/2πkT )
and
1/2 2
df m −mv
=( ) exp ( )
dv 2πkT 2kT
With m/kT = λ , this is the same as the result that we obtain in Section 4.4. With Btrans in hand, we can calculate the average
energy associated with the motion of a gas molecule in one dimension
∞ 1/2 ∞
2 3 2
mv df m −mv
2
⟨ϵ⟩ = ∫ ( )( ) dv = ( ) ∫ v exp ( )dv
−∞ 2 dv 8πkT −∞ 2kT
This definite integral is also given in Appendix D. We find

kT
⟨ϵtrans ⟩ =
2
We see that we can obtain the average kinetic energy for one degree of translational motion by a simple argument that uses
classical-mechanical energies in the Boltzmann equation. We can make the same argument for each of the other two degrees of
translational motion. We conclude that each degree of translational freedom contributes kT /2 to the average energy of a gas
molecule. For three degrees of translational freedom, the total contribution is 3kT /2, which is the result that we first obtained in
Section 2.10.
Now let us consider a classical-mechanical model for a rigid molecule rotating in a plane. The classical kinetic energy is
= I ω /2 , where I is the molecule’s moment of inertia about the axis of rotation, and ω is the angular rotation rate. This has
2
ϵrot
the same form as the translational kinetic energy, so if we assume kT ≫ I ω /2 and a probability density function of the form
2
2
df −I ω
= Brot exp ( )
dω 2kT
finding B rot and ⟨ϵrot ⟩ follows exactly as before, and the average rotational kinetic energy is
⟨ϵrot ⟩ = kT /2
for a molecule with one degree of rotational freedom.

For a classical harmonic oscillator, the vibrational energy has both kinetic and potential energy components. They are mv /2 and 2
kx /2 where v is the oscillator’s instantaneous velocity, x is its instantaneous location, and k is the force constant. Both of these
2
have the same form as the translational kinetic energy equation. If we can assume that kT ≫ mv /2 , that kT ≫ kx /2 , and that
2 2
the probability density functions are

2
df −mv
kinetic
=B exp ( )
vib
dv 2kT
and
2
df potential
−kx
=B exp ( )
vib
dx 2kT
the same arguments show that the average kinetic energy and the average potential energy are both kT /2:
kinetic
⟨ϵ ⟩ = kT /2
vib
and
potential
⟨ϵ ⟩ = kT /2
vib
so that the average total vibrational energy is

total
⟨ϵ ⟩ = kT
vib
In summary, because the energy for translational motion in one dimension, the energy for rotational motion about one axis, the
energy for vibrational kinetic energy in one dimension, and the energy for vibrational potential energy in one dimension all have
the same form (ϵ = Xu ) each of these modes can contribute kT /2 to the average energy of a molecule. For translation and
2
rotation, the total is kT /2 for each degree of translational or rotational freedom. For vibration, because there is both a kinetic and a
potential energy contribution, the total is kT per degree of vibrational freedom.
Let us illustrate this for the particular case of a non-linear, triatomic molecule. From our discussion in Section 18.4, we see that
there are three degrees of translational freedom, three degrees of rotational freedom, and three degrees of vibrational freedom. The
contributions to the average molecular energy are
3 (kT /2)from translation
+3 (kT /2) from rotation
+3kT from vibration
= 6kT in total
Since the heat capacity is

∂ϵ
CV = ( )
∂T
v
each translational degree of freedom can contribute k/2 to the heat capacity. Each rotational degree of freedom can also contribute
k/2 to the heat capacity. Each vibrational degree of freedom can contribute k to the heat capacity. It is important to remember that
these results represent upper limits for real molecules. These limits are realized at high temperatures, or more precisely, at
temperatures where many energy levels, ϵ , satisfy ϵ ≪ kT
i i
This page titled 22.3: Probability Density Functions from the Energies of Classical-mechanical Models is shared under a CC BY-SA 4.0 license
It is enlightening to find the integral approximations to the partition functions and average energies for our simple quantum-
mechanical models of translational, rotational, and vibrational motions. In doing so, however, it is important to remember that the
use of integrals to approximate Boltzmann-equation sums assumes that there are a large number of energy levels, ϵ , for which i
ϵ ≪ kT . If we select a high enough temperature, the energy levels for any motion will always satisfy this condition. The energy
i
levels for translational motion satisfy this condition even at sub-ambient temperatures. This is the reason that Maxwell’s derivation
of the probability density function for translational motion is successful.
Rotational motion is an intermediate case. At sub-ambient temperatures, the classical-mechanical derivation can be inadequate; at
ordinary temperatures, it is a good approximation. This can be seen by comparing the classical-theory prediction to experimental
values for diatomic molecules. For diatomic molecules, the classical model predicts a constant-volume heat capacity of 5k/2 from
3 degrees of translational and 2 degrees of rotational freedom. Since this does not include the contributions from vibrational
motions, constant-volume heat capacities for diatomic molecules must be greater than 5k/2 if both the translational and rotational
contributions are accounted for by the classical model. For diatomic molecules at 298 K, the experimental values are indeed
somewhat larger than 5k/2. (Hydrogen is an exception; its value is 2.47 k.)
Vibrational energies are usually so big that only a minor fraction of the molecules can be in higher vibrational levels at reasonable
temperatures. If we try to increase the temperature enough to make the high-temperature approximation describe vibrational
motions, most molecules decompose. Likewise, electronic partition functions must be evaluated from the defining equation.
The high-temperature limiting average energies can also be calculated from the Boltzmann equation and the appropriate quantum-
mechanical energies. Recall that we find the following quantum-mechanical energies for simple models of translational, rotational,
and vibrational motions:
Translation
2 2
(n) n h
ϵ =
trans 2
8mℓ
(n = 1, 2, 3, … . Derived for a particle in a box)

Rotation
2 2
(m) m h
ϵrot =
2
8π I
(m = 1, 2, 3, …. Derived for rotation about one axis—each energy level is doubly degenerate)
Vibration
(n) 1
ϵ = hν (n + )
vibration
2
(n = 0, 1, 2, 3, … . Derived for simple harmonic motion in one dimension)

When we assume that the temperature is so high that many ϵi are small compared to kT , we find the following high-temperature
limiting partition functions for these motions:
∞ 1/2
2 2 ∞ 2 2 2
−n h −n h 2πmkT ℓ
ztranslation = ∑ exp ( ) ≈∫ exp ( ) dn = ( )
2 2 2
8m ℓ kT 0 8m ℓ kT h
n=1
∞ ∞ 1/2
2 2 2 2 3
−m h −m h 8 π I kT
zrotation = ∑ 2 exp ( ) ≈2∫ exp ( ) dn = ( )
8 π 2 I kT 0
8 π 2 I kT h2
m=1
∞ ∞
−hν 1 −hν 1 kT −hν
zvibration = ∑ exp ( (n + )) ≈ ∫ exp ( (n + )) dn = exp ( )
kT 2 0
kT 2 hν 2kT
n=0
We can then calculate the average energy for each mode as
∞
−ϵn
−1
⟨ϵ⟩ = z ∫ ϵn exp ( ) dn
0
kT
and find
∞ 2 2 2 2
n h −n h
−1
⟨ϵtranslation⟩ = z ∫ ( ) exp ( ) dn
translation 2 2
0 8mℓ 8m ℓ kT
kT
=
2
∞ 2 2 2 2
m h −m h
−1
⟨ϵrotation⟩ = z ∫ 2( ) exp ( ) dm
rotation 2 2
0 8π I 8 π I kT
kT
=
2
∞
−1
1 −hν 1
⟨ϵvibration⟩ =z ×∫ hν (n + ) exp ( (n + )) dn
vibration
0
2 kT 2
hν
= kT +
2
≈ kT
where the last approximation assumes that hν /2 ≪ kT . In the limit as T → 0 , the average energy of the vibrational mode
becomes just hν /2. This is just the energy of the lowest vibrational state, implying that all of the molecules are in the lowest
vibrational energy level at absolute zero.
This page titled 22.4: Partition Functions and Average Energies at High Temperatures is shared under a CC BY-SA 4.0 license and was authored,
One of the earliest applications of quantum mechanics was Einstein’s demonstration that the union of statistical mechanics and
quantum mechanics explains the temperature variation of the heat capacities of solid materials. In Section 7.14, we note that the
heat capacities of solid materials approach zero as the temperature approaches absolute zero. We also review the law of Dulong and
Petit, which describes the limiting heat capacity of many solid elements at high (ambient) temperatures. The Einstein model
accounts for both of these observations.
The physical model underlying Einstein’s development is that a monatomic solid consists of atoms vibrating about fixed points in a
lattice. The particles of this solid are distinguishable from one another, because the location of each lattice point is uniquely
specified. We suppose that the vibration of any one atom is independent of the vibrations of the other atoms in the lattice. We
assume that the vibration results from a Hooke’s Law restoring force
⇀ ⇀ ⇀ ⇀ ⇀
F = −λ r = −λ (x i + y j + z k )
that is zero when the atom is at its lattice point, for which ⇀
r = (0, 0, 0) . The potential energy change when the atom, of mass m, is
driven from its lattice point to the point ⇀r = (x, y, x) is
⇀
r x y z 2 2 2
⇀ x y z
⇀
V =∫ −F ∙ d r = λ ∫ xdx + λ ∫ ydy + λ ∫ zdz = λ +λ +λ
⇀ ⇀ 2 2 2
r = 0 x=0 y=0 z=0
The Schrödinger equation for this motion is

2 2 2 2 2 2 2
h ∂ ψ ∂ ψ ∂ ψ x y z
− [ + + ] +λ [ + + ] ψ = ϵψ
2 2 2 2
8π m ∂x ∂y ∂z 2 2 2
where ψ is a function of the three displacement coordinates; that is ψ = ψ (x, y, z) . We assume that motions in the x-, y -, and z -
directions are completely independent of one another. When we do so, it turns out that we can express the three-dimensional
Schrödinger equation as the sum of three one-dimensional Schrödinger equations
2 2 2
h ∂ ψx x ψx
[− +λ ]
2 2
8π m ∂x 2
2 2
2 ∂ ψy y ψy
h
+ [− +λ ]
2 2
8π m ∂y 2
2 2 2
h ∂ ψz z ψz
+ [− +λ ]
2 2
8π m ∂z 2
= ϵ ψx + ϵ ψ y + ϵ ψz
(n) (n) (n) (n) (n) (n)

where any wavefunction ψ is the same function as ψ and ψ , and the corresponding energies ϵ , ϵ , and ϵ have the
x y z x y z
same values. The energy of the three-dimensional atomic motion is simply the sum of the energies for the three one-dimensional
motions. That is,
(n) (m) (p)
ϵn,m,p = ϵx + ϵy + ϵz ,
which, for simplicity, we also write as
ϵn,m,p = ϵn + ϵm + ϵp .
This page titled 22.5: Energy Levels for a Three-dimensional Harmonic Oscillator is shared under a CC BY-SA 4.0 license and was authored,
In Section 22.4, we find an approximate partition function for the harmonic oscillator at high temperatures. Because it is a
geometric series, the partition function for the harmonic oscillator can also be obtained exactly at any temperature. By definition,
the partition function for the harmonic oscillator is
∞ ∞ ∞ n
−hν 1 −hν −nhν −hν −hν
z = ∑ exp ( (n + )) = exp ( ) ∑ exp ( ) = exp ( ) ∑ [exp ( ) ]
kT 2 2kT kT 2kT kT
n=0 n=0 n=0
This is just the infinite sum

∞
a
n
z = a∑r =
1 −r
n=0
with
−hν
a = exp ( )
2kT
and
−hν
r = exp ( )
kT
Hence, the exact partition function for the one-dimensional harmonic oscillator is
exp (−hν /2kT )
z =
1 − exp (−hν /kT )
The partition function for vibration in each of the other two dimensions is the same. To get the partition function for oscillation in
all three dimensions, we must sum over all possible combinations of the three energies. Distinguishing the energies associated with
motion in the x-, y -, and z -directions by the subscripts n , m, and p, respectively, we have for the three-dimensional harmonic
oscillator:
∞ ∞ ∞
− (ϵn + ϵm + ϵp )
z3D = ∑ ∑ ∑ exp [ ]
kT
p=0 m=0 n=0
∞ ∞ ∞
−ϵp −ϵm −ϵn
= ∑ exp ∑ exp ∑ exp
kT kT kT
p=0 m=0 n=0
3
=z
Hence,
3
exp (−hν /2kT )
z3D = [ ]
1 − exp (−hν /kT )
and the energy of a crystal of N , independent, distinguishable atoms is

E = N ⟨ϵ⟩
2
∂ ln z3D
= N kT ( )
∂T
V
3N hν 3N hν exp (−hν /kT )

= +
2 1 − exp (−hν /kT )
Taking the partial derivative with respect to temperature gives the heat capacity of this crystal. The molar heat capacity can be
expressed in two ways that are useful for our purposes:
¯
¯¯¯
∂E
CV = ( )
∂T
V
2
¯
¯¯¯
¯ hν exp (−hν /kT )
= 3 N k( ) [ ]
2
kT (1 − exp (−hν /kT ))
2
¯
¯¯¯
¯ hν exp (hν /kT )
= 3 N k( ) [ ]
2
kT (exp (hν /kT ) − 1)
Consider the heat capacity at high temperatures. As the temperature becomes large, hν /kT approaches zero. Then
hν hν
exp ( ) ≈1+
kT kT
Using this approximation in the second representation of C gives for the high temperature limit
V
2
¯
¯¯¯
¯ hν 1 + hν /kT
CV ≈ 3 N k( ) [ ]
2
kT (1 + hν /kT − 1)
¯
¯¯¯
¯ hν
≈ 3 N k (1 + )
kT
¯
¯¯¯
¯
≈ 3 N k = 3R
Since C and C are about the same for solids at ordinary temperatures, this result is essentially equivalent to the law stated by
V P
Dulong and Petit. Indeed, it suggests that the law would be more accurate if stated as a condition on C rather than C , and this V P
proves to be the case.

At low temperatures, hν /kT becomes arbitrarily large and exp (−hν /kT ) approaches zero. From the first representation of CV ,
we see that
¯
¯¯¯
∂E
lim ( ) = CV = 0
T →0 ∂T
V
In Section 10.9, we see that C P − CV → 0 as T → 0 . Hence, the theory also predicts that C P → 0 as T → 0 , in agreement with
experimental results.
The Einstein model assumes that energy variations in a solid near absolute zero are entirely due to variations in the vibrational
energy. From the assumption that all of these vibrational motions are characterized by a single frequency, it predicts the limiting
values for the heat capacity of a solid at high and low temperatures. At intermediate temperatures, the quantitative predictions of
the Einstein model leave room for improvement. An important refinement developed by Peter Debye assumes a spectrum of
vibrational frequencies and results in excellent quantitative agreement with experimental values at all temperatures.
We can give a simple qualitative interpretation for the result that heat capacities decrease to zero as the temperature goes to
absolute zero. The basic idea is that, at a sufficiently low temperature, essentially all of the molecules in the system are in the
lowest available energy level. Once essentially all of the molecules are in the lowest energy level, the energy of the system can no
longer decrease in response to a further temperature decrease. Therefore, in this temperature range, the heat capacity is essentially
zero. Alternatively, we can say that as the temperature approaches zero, the fraction of the molecules that are in the lowest energy
level approaches one, and the energy of the system of N molecules approaches the smallest value it can have.
The weakness in this qualitative view is that there is always a non-zero probability of finding molecules in a higher energy level,
and this probability changes as the temperature changes. To firm up the simple picture, we need a way to show that the energy
decreases more rapidly than the temperature near absolute zero. More precisely, we need a way to show that
¯
¯¯¯
∂E
lim ( ) = CV = 0
T →0 ∂T
V
Since the Einstein model produces this result, it constitutes a quantitative validation of our qualitative model.
This page titled 22.6: Energy and Heat Capacity of the "Einstein Crystal" is shared under a CC BY-SA 4.0 license and was authored, remixed,
In most cases, calculation of the entropy from information about the energy levels of a system is best accomplished using the
partition function. Occasionally other entropy relationships are useful. We illustrate this by using the entropy relationship
∞
S = −N k ∑ gi ρ (ϵi )ln ρ (ϵi ) + N kln g1
i=1
to find the entropy of an N -molecule disordered crystal at absolute zero. To be specific, let us consider a crystal of carbon
monoxide.
We can calculate the entropy of carbon monoxide at absolute zero from either of two perspectives. Let us first assume that the
energy of a molecule is almost completely independent of the orientations of its neighbors in the crystal. Then the energy of any
molecule in the crystal is essentially the same in either of the two orientations available to it. In this model for the system, we
consider that there are two, non-degenerate, low-energy quantum states available to the molecule. We suppose that all other
quantum states lie at energy levels whose probabilities are very small when the temperature is near absolute zero. We have
g = g = 1 , ϵ ≈ ϵ . Near absolute zero, we have ρ (ϵ ) ≈ ρ (ϵ ) ≈ 1/2 ; for i > 2 , ρ (ϵ ) ≈ 0 . The entropy becomes
1 2 2 1 2 1 i
S = −N k ∑ gi ρ (ϵi )ln ρ (ϵi ) + N kln g1
i=1
1 1 1 1
= −N k ( ) ln( ) −Nk ( ) ln( )
2 2 2 2
1
= −N kln( )
2
= N kln 2
Alternatively, we can consider that there is just one low-energy quantum state available to the molecule but that this quantum state
is doubly degenerate. In this model, the energy of the molecule is the same in either of the two orientations available to it. We have
g = 2 . Near absolute zero, we have ρ (ϵ ) ≈ 1 ; for i > 1 , ρ (ϵ ) ≈ 0 . The summation term vanishes, and the entropy becomes
1 1 i
S = N kln g1 = N kln 2
Either perspective implies the same value for the zero-temperature entropy of the N -molecule crystal.
Either of these treatments involves a subtle oversimplification. In our first model, we recognize that the carbon monoxide molecule
must have a different energy in each of its two possible orientations in an otherwise perfect crystal. The energy of the orientation
that makes the crystal perfect is slightly less than the energy of the other orientation. We introduce an approximation when we say
that ρ (ϵ ) ≈ ρ (ϵ ) ≈ 1/2 . However, if ϵ is not exactly equal to ϵ , this approximation cannot be valid at an arbitrarily low
2 1 2 1
temperature. To see this, we let the energy difference between these orientations be ϵ − ϵ = Δϵ > 0 . At relatively high 2 1
temperatures, at which Δϵ ≪ kT , we have

ρ (ϵ2 ) −Δϵ
= exp ( ) ≈1
ρ (ϵ1 ) kT
and . At such temperatures, the system behaves as if the lowest energy level were doubly degenerate, with
ρ (ϵ2 ) ≈ ρ (ϵ1 ) ≈ 1/2
ϵ2 = ϵ1 . However, since T can be arbitrarily close to zero, this condition cannot always apply. No matter how small Δϵ may be,
there are always temperatures at which Δϵ ≫ kT and at which we have
ρ (ϵ2 )
≈0
ρ (ϵ1 )
This implies that the molecule should always adopt the orientation that makes the crystal perfectly ordered when the temperature
becomes sufficiently close to zero. This conclusion disagrees with the experimental observations.
Our second model assumes that the energy of a carbon monoxide molecule is the same in either of its two possible orientations.
However, its interactions with the surrounding molecules cannot be exactly the same in each orientation; consequently, its energy
cannot be exactly the same. From first principles, therefore, our second model cannot be strictly correct.
To resolve these apparent contradictions, we assume that the rate at which a carbon monoxide molecule can change its orientation
within the lattice depends on temperature. For some temperature at which Δϵ ≪ kT , the reorientation process occurs rapidly, and
the two orientations are equally probable. As the temperature decreases, the rate of reorientation becomes very slow. If the
reorientation process effectively ceases to occur while the condition Δϵ ≪ kT applies, the orientations of the component
molecules remain those that occur at higher temperatures no matter how much the temperature decreases thereafter. This is often
described by saying that molecular orientations become “frozen.” The zero-temperature entropy of the system is determined by the
energy-level probabilities that describe the system at the temperature at which reorientation effectively ceases to occur.
This page titled 22.7: Applications of Other Entropy Relationships is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
22.8: Problems
1. The gravitational potential energies available to a molecule near the surface of the earth are ϵ (h) = mgh . Each height, h ,
corresponds to a unique energy, so we can infer that the degeneracy of ϵ (h) is unity. Derive the probability density function for the
distribution of molecules in the earth’s atmosphere. (See Problem 19 in Chapter 3.)
2. The value of the molecular partition function approximates the number of quantum states that are available to the molecule and
whose energy is less than kT . How many such quantum states are available to a molecule of molecular weight 40 that is confined
in a volume of 10 m at 300 K?
−6 3
CHAPTER OVERVIEW
23: The Ensemble Treatment

This page titled 23: The Ensemble Treatment is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen
1
When we begin our discussion of Boltzmann statistics in Chapter 20, we note that there exists, in principle, a Schrödinger equation
for an N -molecule system. For any particular set of boundary conditions, the solutions of this equation are a set of infinitely many
wavefunctions, Ψ , for the N -molecule system. For every such wavefunction, there is a corresponding system energy, E . The
i,j i
wavefunctions reflect all of the attractive and repulsive interactions among the molecules of the system. Likewise, the energy levels
of the system reflect all of these interactions.
In Section 20.12, we introduce the symbol Ω to denote the degeneracy of the energy,
E E , of an N -molecule system. Because the
constituent molecules are assumed to be distinguishable and non-interacting, we have
ΩE = ∑ W (Ni , gi )
{ Ni },E
In the solution of the Schrödinger equation for a system of N interacting molecules, each system-energy level, E , can be i
degenerate. We again let Ω denote the degeneracy of an energy level of the system. We use Ω (rather than Ω ) to represent the
i Ei
degeneracy of E . It is important to recognize that the symbol “Ω ” now denotes an intrinsic quantum-mechanical property of the
i i
N-particle system.
In Chapters 21 and 22, we denote the parallel properties of an individual molecule by ψ for the molecular wavefunctions, ϵ for
i,j i
the corresponding energy levels, and g for the degeneracy of the i energy level. We imagine creating an N -molecule system by
i
th
collecting N non-interacting molecules in a fixed volume and at a fixed temperature.

In exactly the same way, we now imagine collecting N^ of these N -molecule, constant-volume, constant-temperature systems. An
aggregate of many multi-molecule systems is called an ensemble. Just as we assume that no forces act among the non-interacting
molecules we consider earlier, we assume that no forces act among the systems of the ensemble. However, as we emphasize above,
our model for the systems of an ensemble recognizes that intermolecular forces among the molecules of an individual system can
be important. We can imagine specifying the properties of the individual systems in a variety of ways. A collection is called a
canonical ensemble if each of the systems in the ensemble has the same values of N , V , and T . (The sense of this name is that by
specifying constant N , V , and T , we create the ensemble that can be described most simply.)
The canonical ensemble is a collection of N^ identical systems, just as the N -molecule system is a collection of N identical
molecules. We imagine piling the systems that comprise the ensemble into a gigantic three-dimensional stack. We then immerse the
entire stack—the ensemble—in a constant temperature bath. The ensemble and its constituent systems are at the constant
temperature T . The volume of the ensemble is N^ V . Because we can specify the location of any system in the ensemble by
specifying its x-, y -, and z -coordinates in the stack, the individual systems that comprise the ensemble are distinguishable from one
another. Thus the ensemble is analogous to a crystalline N -molecule system, in which the individual molecules are distinguishable
from one another because each occupies a particular location in the crystal lattice, the entire crystal is at the constant temperature,
T , and the crystal volume is N Vmolecule .
Since the ensemble is a conceptual construct, we can make the number of systems in the ensemble, N^ , as large as we please. Each
system in the ensemble will have one of the quantum-mechanically allowed energies, E . We let the number of systems that have
i
energy E be N^ . Similarly, we let the number with energy E be N^ , and the number with energy E be N^ . Thus at any given
1 1 2 2 i i
instant, the ensemble is characterized by a population set, {N^ , N^ , … , N^ , …} , in exactly the same way that an N -molecule
1 2 i
system is characterized by a population set, {N , N , … , N , …}. We have

1 2 i
^ ^
N = ∑ Ni
i=1
While all of the systems in the ensemble are immersed in the same constant-temperature bath, the energy of any one system in the
ensemble is completely independent of the energy of any other system. This means that the total energy of the ensemble, E ^
, is
given by
∞
^ ^
E = ∑ N i Ei
i=1
Property System Ensemble
System comprising a collection of N molecules

Quantum entity Molecule at fixed volume and temperature
at fixed volume and temperature
System comprising a collection of N Ensemble comprising N^ systems each of

Aggregate of quantum entities
molecules at fixed volume and temperature which contains N molecules
Number of quantum entities in aggregate N

^
N
Wave functions/quantum states ψ

i
Ψi
Energy levels ϵi Ei
Energy level degeneracies gi Ωi
Probability that an energy level is occupied Pi

^
Pi
Number of quantum entities in the i th

energy ^
Ni Ni
level
Probability that a quantum state is occupied ρ (ϵi ) ρ̂ (Ei )
Energy of the aggregate’s k th

population set Ek = ∑ Nk,i ϵi
^ ^
E k = ∑ N k,i ϵi
Expected value of the energy of the aggregate ⟨E⟩ = N ∑ Pi ϵi

^ ^ ^
⟨E ⟩ = N ∑ P i Ei
The population set, {N^ , N^ , … , N^ , …} , that characterizes the ensemble is not constant in time. However, by the same
1 2 i
arguments that we apply to the N-molecule system, there is a population set
^ ^ ^
{ N 1 , N 2 , … , N i , …}
which characterizes the ensemble when it is at equilibrium in the constant-temperature bath.

We define the probability, P^ , that a system of the ensemble has energy E to be the fraction of the systems in the ensemble with
i i
this energy, when the ensemble is at equilibrium at the specified temperature. Thus, by definition,
^
Ni
^
Pi = .
^
N
We define the probability that a system is in one of the states, Ψ , with energy E , as
i,j i
^
Pi
ρ̂ (Ei ) =
Ωi
The method we have used to construct the canonical ensemble insures that the entire ensemble is always at the specified
temperature. If the component systems are at equilibrium, the ensemble is at equilibrium. The expected value of the ensemble
energy is
∞ ∞
^ ^ ^ ^
⟨E ⟩ = N ∑ P i Ei = ∑ N Ei
i
i=1 i=1
Because the number of systems in the ensemble, N^ , is very large, we know from the central limit theorem that any observed value
for the ensemble energy will be indistinguishable from the expected value. To an excellent approximation, we have at any time,
^ ^
E = ⟨E ⟩
and
^ ^
Ni = Ni
The table above summarizes the terminology that we have developed to characterize molecules, N -molecule systems, and ^
N -
system ensembles of N -molecule systems.
We can now apply to an ensemble of N^ , distinguishable, non-interacting systems the same logic that we applied to a system of N ,
distinguishable, non-interacting molecules. The probability that a system is in one of the energy levels is
^ ^ ^
1 = P 1 +P 2 +⋯ +P i +…
The total probability sum for the constant-temperature ensemble is

^
N ^ ^ ^
N1 N2 Ni
^ ^ ^ ^ ^
1 = (P 1 + P 2 + ⋯ + P i + …) = ∑ W (N i , Ωi )ρ̂ (E1 ) ρ̂ (E2 ) … ρ̂ (Ei ) …
^
{N i }
where
^
∞ Ni
Ω
^ ^ ^ i
W (N i , Ωi ) = N ! ∏
^
i=1 Ni!
Moreover, we can imagine instantaneously isolating the ensemble from the temperature bath in which it is immersed. This is a
wholly conceptual change, which we effect by replacing the fluid of the constant-temperature bath with a solid blanket of
insulation. The ensemble is then an isolated system whose energy, E ^ , is constant. Every system of the isolated ensemble is
immersed in a constant-temperature bath, where the constant-temperature bath consists of the N^ − 1 systems that make up the rest
of the ensemble. This is an important feature of the ensemble treatment. It means that any conclusion we reach about the systems of
the constant-energy ensemble is also a conclusion about each of the N^ identical, constant-temperature systems that comprise the
isolated, constant-energy ensemble.
Only certain population sets, {N^ 1,
^
N2,
^
… , N i , …} , are consistent with the fixed value, E
^
, of the isolated ensemble. For each
of these population sets, there are ^ ^
W (N i , Ωi ) system states. The probability of each of these system states is proportional to
^ ^ ^
…. By the principle of equal a priori probability, every system state of the fixed-energy ensemble
N1 N2 Ni
ρ̂ (E1 ) ρ̂ (E )2 … ρ̂ (E ) i
occurs with equal probability. We again conclude that the population set that characterizes the equilibrium state of the constant-
energy ensemble, {N^ 1
^ ^
, N 2 , … , N i , …} , is the one for which W
^
or ln W
^
is a maximum, subject to the constraints
∞
^ ^
N = ∑ Ni
i=1
and
∞
^ ^
E = ∑ N i Ei
i=1
The fact that we can make ^

N arbitrarily large ensures that any term, ^
Ni , in the equilibrium-characterizing population set can be
very large, so that N^ can be found using Stirling’s approximation and Lagrange’s method of undetermined multipliers. We have
i
the mnemonic function

∞ ∞ ∞
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^
Fmn = N ln N − N + ∑ (N i ln Ωi − N i ln N i + N i ) + α ( N − ∑ N i ) + β ( E − ∑ N i Ei )
i=1 i=1 i=1
so that
⎛ ∂F ⎞ ^
mn Ni
= ln Ωi − ^
− ln N + 1 − α − β Ei = 0
i
⎝ ^ ⎠ ^
∂N i Ni
j≠i
and
^
ln N = ln Ωi − α − β Ei
i
or
^
N i = Ωi exp(−α) exp −β Ei
When we make use of the constraint on the total number of systems in the ensemble, we have
∞ ∞
^ ^
N = ∑ N i = exp(−α) ∑ Ωi exp(−β Ei )
i=1 i=1
so that
^ Z −1
exp(−α) = N
where the partition function for a system of N possibly-interacting molecules is

∞
Z = ∑ Ωi exp(−β Ei )
i=1
The probability that a system has energy E is equal to the equilibrium fraction of systems in the ensemble that have energy E , so
i i
that
^
Ni Ωi exp(−β Ei )
^
Pi = =
^ Z
N
This page titled 23.1: Ensembles of N-molecule Systems is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by
request.
At equilibrium, the entropy of the N^ -system ensemble, S , must be a maximum. By arguments that parallel those in Chapter
ensemble
20, W^
is a maximum for the ensemble population set that characterizes this equilibrium state. Applying the Boltzmann definition
to the ensemble, the ensemble entropy is S = kln W
ensemble
^
. Since all N^ systems in the ensemble have effectively the same
max
entropy, S , we have S = NS .
ensemble
^
When we assume that W ^
occurs for the equilibrium population set,
max
^ ^ ^
{N 1 , N 2 , … , N i , …} , we have
^
N
∞ i
Ω
i
^ ^
W max = N ! ∏
i=1 ^
Ni!
so that
∞ ∞
^ ^ ^ ^
Sensemble = N S = k ln N ! + k ∑ N i ln Ωi − k ∑ ln( N i !)
i=1 i=1
From the Boltzmann distribution function, N^ i

^ =Z
/N
−1
Ωi exp (−β Ei ) , we have
^ ^
ln Ωi = ln Z + ln N i + β Ei − ln N
Substituting, and introducing Stirling’s approximation, we find

∞ ∞
^S
N ^ ln N
= kN ^ − kN
^ +k∑N
^ (ln Z + ln N
^ ^ ) − k ∑ (N
+ β Ei − ln N ^ ln N
^ −N
^ )
i i i i i
i=1 i=1
^ kln Z + kβ ∑ N
=N ^ E
i i
i=1
∞
Since ∑ i=1
^
N i Ei is the energy of the N^ -system ensemble and the energy of each system is the same, we have
∞
^ ^
∑ N Ei = Eensemble = N E
i
i=1
Substituting, we find
S = kβE + kln Z
where S , E , and Z are the entropy, energy, and partition function for the N -molecule system. From the fundamental equation, we
have
∂E
( ) =T
∂S
V
Differentiating S = kβE + kln Z with respect to entropy at constant volume, we find

∂E
1 = kβ( )
∂S
V
and it follows that

1
β =
kT
We have, for the N -molecule system

∞
−Ei
Z = ∑ Ωi exp ( )
kT
i=1
(System partition function)
−1
−Ei
^
Pi =Z Ωi exp ( )
kT
(Boltzmann’s equation)
E
S = + kln Z
T
(Entropy of the N-molecule system)
This page titled 23.2: The Ensemble Entropy and the Value of ß is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
With the results of Section 23.2 in hand, we can find the other thermodynamic functions for the N -molecule system from the
equations for Z and P^ by the arguments we use in Chapters 20 and 21. Let us summarize these arguments. From
i
^
E = ∑ P i Ei
i=1
we have
∞ ∞
^ ^
dE = ∑ Ei dP i + ∑ P i dEi
i=1 i=1
We associate the first term with dq rev

and the second term with dw = −P dV ; that is,
∞ ∞ ∞
^
P
rev i
dq ^ = −kT ∑ ln(
= T dS = ∑ Ei dP ^ − kT ln Z ∑ dP
)dP ^
i i i
Ωi
i=1 i=1 i=1
Where we substitute
^
Pi
Ei = −kT ln( ) − kT ln Z
Ωi
which we obtain by taking the natural logarithm of the partition function. Since ∑ ∞
i=1
^
dP i = 0 , we have for each system,
∞ ^ ∞ ^ ^ ∞ ^
Pi Pi Pi Pi
^ ^ ^
dS = −k ∑ ln( )dP i = −k ∑ { Ωi d ( ln ) − dP i } = −k ∑ d ( P i ln )
Ωi Ωi Ωi Ωi
i=1 i=1 i=1
The system entropy, S , and the system-energy-level probabilities, P^ , are functions of temperature. Integrating from
i T =0 to T
and choosing the lower limits for the integrations on the right to be P^ (0) = 1 and P^ (0) = 0 for i > 1 , we have
1 i
∞ ^
S P i (T ) ^
Pi
^
∫ dS = −k ∑ ∫ d ( P i ln )
S0
^
P i (0)
Ωi
i=1
Letting P^ i
^
(T ) = P i , the result is
∞
^ ^
P1 1 Pi
^ ^
S − S0 = −kP 1 ln + k ln − k ∑ P i ln
Ω1 Ω1 Ωi
i=2
∞ ^
Pi
^
= −k ∑ P i ln − k ln Ω1
Ωi
i=1
From the partition function, we have
^
P Ei
i
ln( ) =− + ln Z
Ωi kT
so that
∞
Ei
^
S − S0 = −k ∑ P i (− + ln Z ) − kln Ω1
kT
i=1
∞ ∞
1
^ ^
= ∑ P i Ei + kln Z ∑ P i − kln Ω1
T
i=1 i=1
E
= + kln Z − kln Ω1
T
We take the system entropy at absolute zero, S , to be
0
S0 = kln Ω1
If the lowest energy state is non-degenerate, Ω 1 =1 , and S 0 =0 , so that

E
S (T ) = + kln Z
T
As in Section 21.6, we observe that

∞ ∞
−1
−Ei
^
E = ∑ P i Ei = Z ∑ Ωi Ei exp ( )
kT
i=1 i=1
and that
∞
∂ln Z Ei −Ei E
−1
( ) =Z ∑ Ωi ( ) exp ( ) =
2 2
∂T kT kT kT
V i=1
so that
2
∂ln Z
E = kT ( )
∂T
V
From A = E − T S and the entropy equation, S = E/T + kln Z , the Helmholtz free energy of the system is
A = −kT ln Z
For the system pressure, we find from

∂A
P = −( )
∂V
T
that
∂ln Z
P = kT ( )
∂V
T
From H = E + P V , we find
2
∂ln Z ∂ln Z
H = kT ( ) + V kT ( )
∂T ∂V
V T
and from G = E + P V −TS , we find

∂ln Z
G = V kT ( ) − kT ln Z
∂V
T
For the chemical potential per molecule in the N -molecule system, we obtain
∂A ∂ln Z
μ =( ) = −kT ( )
∂N ∂N
VT VT
Thus, we have found the principle thermodynamic functions for the N -molecule system expressed in terms of ln Z and its
derivatives. The system partition function, Z , depends on the energy levels available to the N -molecule system. The
thermodynamic functions we have obtained are valid for any system, including systems in which intermolecular forces make large
contributions to the system energy. Of course, the system partition function, Z , must accurately reflect the effects of these forces.
In Chapter 24 we find that the partition function, Z , for a system of N , distinguishable, non-interacting molecules is related in a
simple way to the molecular partition function, z . We find Z = z . When we substitute this result for Z into the system partition
N
functions developed above, we recover the same results that we developed in Chapters 20 and 21 for the thermodynamic properties
of a system of N , distinguishable, non-interacting molecules.
This page titled 23.3: The Thermodynamic Functions of the N-molecule System is shared under a CC BY-SA 4.0 license and was authored,
CHAPTER OVERVIEW
24: Indistinguishable Molecules - Statistical Thermodynamics of Ideal Gases

The ensemble analysis shows that the thermodynamic functions for an N -molecule system can be developed from the principles of
statistical mechanics whether the molecules of the system interact or not. The theory is valid irrespective of the strengths of inter-
molecular attractions and repulsions. However, to carry out numerical calculations, it is necessary to know the energy levels for the
N -molecule system. For systems in which the molecules interact, obtaining useful approximations to these levels is a difficult
problem. As a result, many applications assume that the molecules do not interact with one another. In this chapter we apply the
results from the ensemble theory to the particular case of ideal gases.
24.5: The Partition Function for A Gas of Indistinguishable, Non-interacting, Separable-modes Molecules
24.14: Problems
This page titled 24: Indistinguishable Molecules - Statistical Thermodynamics of Ideal Gases is shared under a CC BY-SA 4.0 license and was
1
In Chapter 21, our analysis of a system of N distinguishable and non-interacting molecules finds that the system entropy is given
by
E E
N
S = + N kln z = + kln z
T T
where E is the system energy and z is the molecular partition function. From ensemble theory, we found
E
S = + kln Z
T
where Z is the partition function for the N -molecule system. Comparison implies that, for a system of N , distinguishable, non-
interacting molecules, we have
N
Z =z
We can obtain this same result by writing out the energy levels for the system in terms of the energy levels of the distinguishable
molecules that make up the system. First we develop the obvious notation for the energy levels of the individual molecules. We let
the energy levels of the first molecule be the set {ϵ }, the energy levels of the second molecule be the set {ϵ }, and so forth to
1,i 2,i
the last molecule for which the energy levels are the set {ϵ }. Thus, the i energy level of the r molecule is ϵ . We let the
N ,i
th th
r,i
corresponding energy-level degeneracy be g and the partition function for the r molecule be z . Since all of the molecules are
r,i
th
r
identical, each has the same set of energy levels; that is, we have ϵ = ϵ and g = g for any two molecules, p and r, and any
p,i r,i p,i r,i
energy level, i. It follows that the partition function is the same for every molecule
∞
−ϵr,i
z1 = z2 = ⋯ = zj = ⋯ = zN = z = ∑ gr,i exp ( )
kT
i=1
so that
N
z1 z2 … zr … zN = z
We can write down the energy levels available to the system of N distinguishable, non-interacting molecules. The energy of the
system is just the sum of the energies of the constituent molecules, so the possible system energies consist of all of the possible
sums of the distinguishable-molecule energies. Since there are infinitely many molecular energies, there are infinitely many system
energies.
E1 = ϵ1,1 + ϵ2,1 + ⋯ + ϵr,1 + ⋯ + ϵN ,1
E2 = ϵ1,2 + ϵ2,1 + ⋯ + ϵr,1 + ⋯ + ϵN ,1
E3 = ϵ1,3 + ϵ2,1 + ⋯ + ϵr,1 + ⋯ + ϵN ,1
Em = ϵ1,i + ϵ2,j + ⋯ + ϵr,k + ⋯ + ϵN ,p
The product of the N molecular partition functions is

∞
−ϵ1,i
z1 z2 … zr … zN = ∑ g1,i exp ( )
kT
i=1
∞ ∞
−ϵ2,j −ϵr,k
× ∑ g2,j exp ( ) × ⋯ × ∑ gr,k exp ( ) ×
kT kT
j=1 k=1
∞
−ϵN ,p
⋯ × ∑ gN ,p exp ( )
kT
p=1
∞ ∞ ∞ ∞
= ∑ ∑ … ∑ … ∑ g1,i g2,j … gr,k … gN ,p
i=1 j=1 k=1 p=1
− (ϵ1,i + ϵ2,j + ⋯ + ϵr,k + ⋯ + ϵN ,p )

×exp [ ]
kT
The sum in each exponential term is just the sum of N single-molecule energies. Moreover, every possible combination of N
single-molecule energies occurs in one of the exponential terms. Each of these possible combinations is a separate energy level
available to the system of N distinguishable molecules.
The system partition function is
∞
−E i
Z = ∑ Ωi exp ( )
kT
i=1
The ith
energy level of the system is the sum
Ei = ϵ1,v + ϵ2,w + ⋯ + ϵr,k + ⋯ + ϵN ,y
The degeneracy of the i th

energy level of the system is the product of the degeneracies of the molecular energy levels that belong
to it. We have
Ωi = g1,v g2,w … gr,k … gN ,y
Thus, by a second, independent argument, we see that

N
z1 z2 … zr … zN = z =Z
(N distinguishable, non-interacting molecules)
This page titled 24.1: The Partition Function for N Distinguishable, Non-interacting Molecules is shared under a CC BY-SA 4.0 license and was
In all of our considerations to this point, we focus on systems in which the molecules are distinguishable. This effectively confines
the practical applications to crystalline solids. Since there is no way to distinguish one molecule of a given substance from another
in the gas phase, it is evident that the assumptions we have used so far do not apply to gaseous systems. The number and
importance of practical applications increases dramatically if we can extend the theory to describe the behavior of ideal gases.
We might suppose that distinguishability is immaterial—that there is no difference between the behavior of a system of
distinguishable particles and an otherwise-identical system of indistinguishable particles. Indeed, this is an idea well worth testing.
We know the partition function for a particle in box, and we have every reason to believe that this should be a good model for the
partition function describing the translational motion of a gas particle. If an ideal gas behaves as a collection of N distinguishable
particles-in-a-box, the translational partition of the gas is just z . Thermodynamic properties calculated on this basis for, say, argon
N
should agree with those observed experimentally. Indeed, when the comparison is made, this theory gives some properties
correctly. The energy is correct; however, the entropy is not.
Thus, experiment demonstrates that the partition function for a system of indistinguishable molecules is different from that of an
otherwise-identical system of distinguishable molecules. The reason for this becomes evident when we compare the microstates
available to a system of distinguishable molecules to those available to a system of otherwise-identical indistinguishable molecules.
Consider the distinguishable-molecule microstate whose energy is
Ei = ϵ1,v + ϵ2,w + ⋯ + ϵr,k + ⋯ + ϵN ,y
As a starting point, we assume that every molecule is in a different energy level. That is, all of the N energy levels, ϵ , that appear i,j
in this sum are different. For the case in which the molecules are distinguishable, we can write down additional microstates that
have this same energy just by permuting the energy values among the N molecules. (A second microstate with this energy is
E =ϵ
i 1,w +ϵ 2,v+ ⋯ +ϵ r,k+ ⋯ +ϵ .) Since there are N ! such permutations, there are a total of N ! quantum states that have
N ,y
this same energy, and each of them appears as an exponential term in the product z z … z … z = z .
1 2 r N
N
If, however, the N molecules are indistinguishable, there is no way to tell one of these N ! assignments from another. They all
become the same thing. All we know is that some one of the N molecules has the energy ϵ , another has the energy ϵ , etc. This
w v
means that there is only one way that the indistinguishable molecules can have the energy E . It means also that the difference
i
between the distinguishable-molecules case and the indistinguishable-molecules case is that, while they contain the same system
energy levels, each level appears N ! more times in the distinguishable-molecules partition function than it does in the
indistinguishable-molecules partition function. We have
1 1
N
Zindistinguishable = Zdistinguishable = z
N! N!
In the next section, we see that nearly all of the molecules in a sample of gas must have different energies, so that this relationship
correctly relates the partition function for a single gas molecule to the partition function for a system of N indistinguishable gas
molecules.
Before seeing that nearly all of the molecules in a macroscopic sample of gas actually do have different energies, however, let us
see what happens if they do not. Suppose that just two of the indistinguishable molecules have the same energy. Then there are not
N ! permutations of the energies among the distinguishable molecules; rather there are only N !/2! such permutations. In this case,
the relationship between the system and the molecular partition functions is
2! 2! N
Zindistinguishable = Zdistinguishable = z
N! N!
For the population set {N 1, N2 , … , Nr , … , Nω } the relationship is

N1 ! N2 ! … Nr ! … Nω ! N
Zindistinguishable = z
N!
which is much more complex than the case in which all molecules have different energies. Of course, if we extend the latter case,
so that the population set consists of N energy levels, each occupied by at most one molecule, the relationship reverts to the one
with which we began.
∞ ∞
1 1 1
N N N
Zindistinguishable = ( ∏ Ni !) z = ( ∏ 1) z = z
N! N! N!
i=1 i=1
This page titled 24.2: The Partition Function for N Indistinguishable, Non-interacting Molecules is shared under a CC BY-SA 4.0 license and was
The particle in a box is a quantum mechanical model for the motion of a point mass in one dimension. In Section 18.3, we find that
the energy levels are
2 2
n h
ϵn =
2
8mℓ
so that the partition function for a particle in a one-dimensional box is

∞ 2 2
−n h
z = ∑ exp ( )
2
n=1
8mkT ℓ
When the mass approximates that of a molecule, the length of the box is macroscopic, and the temperature is not extremely low,
there are a very large number of energy levels for which ϵ < kt >. When this is the case, we find in Section 22-4 that this sum can
n
be approximated by an integral to obtain an expression for z in closed form:

∞ 2 2 1/2
−n h 2πmkT
z ≈∫ exp ( ) dn = ( ) ℓ
2 2
0 8mkT ℓ h
A particle in a three-dimensional rectangular box is a quantum mechanical model for an ideal gas molecule. The molecule moves in
three dimensions, but the component of its motion parallel to any one coordinate axis is independent of its motion parallel to the
others. This being the case, the kinetic energy of a particle in a three-dimensional box can be modeled as the sum of the energies
for motion along each of the three independent coordinate axes that describe the translational motion of the particle. Taking the
coordinate axes parallel to the faces of the box and labeling the lengths of the sides ℓ , ℓ , and ℓ , the energy of the particle in the x y z
three-dimensional box becomes
ϵ = ϵx + ϵy + ϵz
and the three-dimensional partition function becomes

∞ ∞ ∞ 2 2 2 2
−h nx ny nz
zt = ∑ ∑ ∑ exp [( )( + + )]
2 2 2
8mkT ℓ ℓ ℓ
nx=1 ny=1 nz=1 x y z
∞ 2 2 ∞ 2 2 ∞ 2 2
−nx h −ny h −nz h
= ∑ exp ( ) ∑ exp ( ) ∑ exp ( )
2 2 2
nx =1
8mkT ℓx ny =1
8mkT ℓy nz =1
8mkT ℓz
or, recognizing this as the product of three one-dimensional partition functions,
zt = zx zy zz .
Approximating each molecular partition function as integrals gives

3/2 3/2
2πmkT 2πmkT
zt = ( ) ℓx ℓy ℓz = ( ) V
h2 h2
where the volume of the container is V = ℓx ℓy ℓz .

Let us estimate a lower limit for the molecular partition function for the translational motion of a typical gas at ambient
temperature. The partition function increases with volume, V , so we want to select a volume that is near the smallest volume a gas
can have. We can estimate this as the volume of the corresponding liquid at the same temperature. Let us calculate the molecular
translational partition function for a gas whose molar mass is 0.040 kg in a volume of 0.020 L at 300 K. We find z = 5 × 10 . t
27
Given z , we can estimate the probability that any one of the energy levels available to this molecule is occupied. For any energy
t
level, the upper limit to the term exp (−ϵ /kT ) is one. If the quantum numbers n , n , and n are different from one another, the
i x y z
corresponding molecular energy is non-degenerate. To a good approximation, we have g = 1 . We find i
Ni gi exp (−ϵi /kT ) 1

−28
= < = 2 × 10
¯
¯¯¯
¯
zt zt
N
We calculate N ≈ 1 × 10 . When a mole of this gas occupies 0.020 L, the system density approximates that of a liquid.
i
−4
Therefore, even in circumstances selected to minimize the number of energy levels, there is less than one gas molecule per ten
thousand energy levels.
For translational energy levels of gas molecules, it is an excellent approximation to say that each molecule occupies a different
translational energy level. This is a welcome result, because it assures us that the translational partition function for a system
containing a gas of N indistinguishable non-interacting molecules is just
3N /2
1 2πmkT
N
Zt = ( ) V
2
N! h
So that Z is the translational partition function for a system of N ideal gas molecules.
t
We derive Z from the assumption that every equilibrium population number, N , for the molecular energy levels satisfies N ≤ 1 .
t i i
We use Z and the ensemble-treatment results that we develop in Chapter 23 to find thermodynamic functions for the N -molecule
t
ideal-gas system. The ensemble development assumes that the number of systems, N^ , in the ensemble that have energy E is very
i i
large. Since the ensemble is a creature of our imaginations, we can imagine that N^ is as big as it needs to be in order that N^ be big i
enough. The population sets N

i
and ^
Ni are independent; they characterize different distributions. The fact that N
i
≤1 is
irrelevant when we apply Lagrange’s method to find the distribution function for , the partition function Z , and the
^
N
i t
thermodynamic functions for the system. Consequently, the ensemble treatment enables us to find the partition function for an ideal
gas, Z , by arguments that avoid the questions that arise when we apply Lagrange’s method to the distribution of molecular
IG
translational energies.
This page titled 24.3: Occupancy Probabilities for Translational Energy Levels is shared under a CC BY-SA 4.0 license and was authored,
At this point in our development, we have a theory that gives the thermodynamic properties of a polyatomic ideal gas molecule. To
proceed, however, we must know the energy of every quantum state that is available to the molecule. There is more than one way
to obtain this information. We will examine one important method—one that involves a further idealization of molecular behavior.
We have made great progress by using the ideal gas model, and as we have noted repeatedly, the essential feature of the ideal gas
model is that there are no attractive or repulsive forces between its molecules. Now we assume that the molecule’s translational,
rotational, vibrational, and electronic motions are independent of one another. We could say that this idealization defines super-
ideal gas molecules; not only does one molecule not interact with another molecule, an internal motion of one of these molecules
does not interact with the other internal motions of the same molecule!
The approximation that a molecule’s translational motion is independent of its rotational, vibrational, and electronic motions is
usually excellent. The approximation that its intramolecular rotational, vibrational and electronic motions are also independent
proves to be surprisingly good. Moreover, the very simple quantum mechanical systems that we describe in Chapter 18 prove to be
surprisingly good models for the individual kinds of intramolecular motion. The remainder of this chapter illustrates these points.
In Chapter 18, we note that a molecule’s wavefunction can be approximated as a product of a wavefunction for rotations, a
wavefunction for vibrations, and a wavefunction for electronic motions. (As always, we are simply quoting quantum mechanical
results that we make no effort to derive; we begin with the knowledge that the quantum mechanical problems have been solved and
that the appropriate energy levels are available for our use.) Our goal is to see how we can apply the statistical mechanical results
we have obtained to calculate the thermodynamic properties of ideal gases. To illustrate the essential features, we consider diatomic
molecules. The same considerations apply to polyatomic molecules; there are additional complications, but none that introduce new
principles.
For diatomic molecules, we need to consider the energy levels for translational motion in three dimensions, the energy levels for
rotation in three dimensions, the energy levels for vibration along the inter-nuclear axis, and the electronic energy states.
This page titled 24.4: The Separable-modes molecular Model is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
24.5: The Partition Function for A Gas of Indistinguishable, Non-interacting,
Separable-modes Molecules
We represent the successive molecular energy levels as ϵ and the successive translational, rotational, vibrational, and electronic
i
energy levels as ϵ , ϵ , ϵ , and ϵ . Now the first subscript specifies the energy mode; the second specifies the energy level. We
t,a r,b v,c e,d
approximate the successive energy levels of a diatomic molecule as
ϵ1 = ϵt,1 + ϵr,1 + ϵv,1 + ϵe,1
ϵ2 = ϵt,2 + ϵr,1 + ϵv,1 + ϵe,1
ϵi = ϵt,a + ϵr,b + ϵv,c + ϵe,d
In Section 22.1, we find that the partition function for the molecule becomes
∞ ∞ ∞ ∞
− (ϵt,a + ϵr,b + ϵv,c + ϵe,d )
z = ∑ ∑ ∑ ∑ gt,a gr,b gv,cge,d × exp [ ]
kT
a=1 b=1 c=1 d=1
= zt zr zv ze
where z , z , z , and z are the partition functions for the individual kinds of motion that the molecule undergoes; they are sums
t r v e
over the corresponding energy levels for the molecule. This is essentially the same argument that we use in Section 22.1 to show
that the partition function for an N -molecule system is a product of N molecular partition functions:
N
Z =z .
We are now able to write the partition function for a gas containing N molecules of the same substance. Since the molecules of a
gas are indistinguishable, we use the relationship
1 1 N
N
Zindistinguishable = z = (zt zr zv ze )
N! N!
To make the notation more compact and to emphasize that we have specialized the discussion to the case of an ideal gas, let us
replace “Z ” with “Z ”. Also, recognizing that N ! enters the relationship because of molecular indistinguishability,
indistinguishable IG
and molecular indistinguishability arises because of translational motion, we regroup the terms, writing
N
(zt ) N N N
ZIG = [ ] (zr ) (zv ) (ze )
N!
Our goal is to calculate the thermodynamic properties of the ideal gas. These properties depend on the natural logarithm of the
ideal-gas partition function. This is a sum of terms:
N
(zt )
ln ZIG = ln[ ] + N ln zr + N ln zv + N ln ze
N!
In our development of classical thermodynamics, we find it convenient to express the properties of substance on a per-mole basis.
¯
¯¯¯
¯
For the same reasons, we focus on evaluating ln Z for one mole of gas; that is, for the case that N is Avogadro’s number, N .
IG
We now examine the relationships that enable us to evaluate each of these contributions to ln Z . IG
This page titled 24.5: The Partition Function for A Gas of Indistinguishable, Non-interacting, Separable-modes Molecules is shared under a CC
BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source content that was edited to the style and standards of the
LibreTexts platform; a detailed edit history is available upon request.
We can make use of Stirling’s approximation to write the translational contribution to ln Z IG per mole of ideal gas. This is
¯
¯¯¯
¯
N
⎡ (zt ) ⎤ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ zt
ln = N ln zt − N ln N + N = N + N ln
¯
¯¯¯
¯ ¯
¯¯¯
¯
⎣ N! ⎦ N
(We omit the other factors in Stirling’s approximation. Their contribution to the thermodynamic values we calculate is less than the
uncertainty introduced by the measurement errors in the molecular parameters we use.) In Section 24.3 we find the molecular
partition function for translation:
3/2
2πmkT
zt = ( ) V
2
h
For one mole of an ideal gas, ¯

¯¯
V
¯ ¯
¯¯¯
¯
= N kT /P . The translational contribution to the partition function for one mole of an ideal gas
becomes
¯
¯¯¯
¯
N 3/2 ¯
¯¯¯ 3/2
⎡ (zt ) ⎤ 2πmkT V 2πmkT kT
¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
ln = N + N ln[ ( ) ] = N + N ln[ ( ) ]
¯
¯¯¯
¯
⎣ ⎦ h2 ¯
¯¯¯
¯
h2 P
N! N
This page titled 24.6: The Translational Partition Function of An Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed,
Our quantum-mechanical model for a diatomic molecule takes the zero of energy to be the infinitely separated atoms at rest—that
is, with no kinetic energy. The electrical interactions among the nuclei and electrons are such that, as the atoms approach one
another, a bond forms and the energy of the two-atom system decreases. At some inter-nuclear distance, the energy reaches a
minimum; at shorter inter-nuclear distances, the repulsive interactions between nuclei begin to dominate, and the energy increases.
We can use quantum mechanics to find the wavefunction and energy of the molecule when the nuclei are separated to any fixed
distance. By repeating the calculation at a series of inter-nuclear distances, we can find the distance at which the molecular energy
is a minimum. We take this minimum energy as the electronic energy of the molecule, and the corresponding inter-nuclear distance
as the bond length. This is the energy of the lowest electronic state of the molecule. The lowest electronic state is called the ground
state.
Excited electronic states exist, and their energies can be estimated from spectroscopic measurements or by quantum mechanical
calculation. For most molecules, these excited electronic states are at much higher energy than the ground state. When we compare
the terms in the electronic partition function, we see that
exp (−ϵe,1 /kT ) ≫ exp (−ϵe,2 /kT )
The term for any higher energy level is insignificant compared to the term for the ground state. The electronic partition function
becomes just
ze = g1 exp (−ϵe,1 /kT )
The ground-state degeneracy, g , is one for most molecules. For unusual molecules the ground-state degeneracy can be greater; for
1
molecules with one unpaired electron, it is two.

The energy of the electronic ground state that we obtain by direct quantum mechanical calculation includes the energy effects of the
motions of the electrons and the energy effects from the electrical interactions among the electrons and the stationary nuclei.
Because we calculate it for stationary nuclei, the electronic energy does not include the energy of nuclear motions. The ground state
electronic energy is the energy released when the atoms come together from infinite separation to a state in which they are at rest at
the equilibrium inter-nuclear separation. This is just minus one times the work required to separate the atoms to an infinite distance,
starting from the inter-nuclear separation with the smallest energy. On a graph of electronic (or potential) energy versus inter-
nuclear distance, the ground state energy is just the depth of the energy well measured from the top down (ϵ < 0) . The work
e,1
required to separate one mole of these molecules into their constituent atoms is called the equilibrium dissociation energy, and
conventionally given the symbol D . These definitions mean that D > 0 and D = −N ϵ .
e e e
¯
¯¯¯
¯
e,1
Figure 1. Equilibrium dissociation energy, D , and spectroscopic dissociation energy, D .

e 0
In practice, the energy of the electronic ground state is often estimated from spectroscopic measurements. By careful study of its
spectra, it is possible to find out how much energy must be added, as a photon, to cause a molecule to dissociate into atoms.
Expressed per mole, this energy is called the spectroscopic dissociation energy, and it is conventionally given the symbol D . 0
These spectroscopic measurements involve the absorption of photons by real molecules. Before they absorb the photon, these
molecules already have energy in the form of vibrational and rotational motions. So the real molecules that are involved in any
spectroscopic measurement have energies that are greater than the energies of the hypothetical motionless-atom molecules at the
bottom of the potential energy well. This means that less energy is required to separate the real molecule than is required to
separate the hypothetical molecule at the bottom of the well. For any molecule, D >D . e 0
To have the lowest possible energy, a real molecule must be in its lowest rotational and lowest vibrational energy levels. As turns
out, a molecule can have zero rotational energy, but its vibrational energy can never be zero. In Section 24.8 we review the
harmonic oscillator approximation. In its lowest vibrational energy level (n = 0) , a diatomic molecule’s minimum vibrational
energy is hν /2. D and ν can be estimated from spectroscopic experiments. We estimate
0
De D0 hν
ϵe,1 = − = −( + )
¯
¯¯¯
¯ ¯
¯¯¯
¯ 2
N N
and the molecular electronic partition function becomes

D0 hν
ze = g1 exp ( + )
¯
¯¯¯
¯ 2kT
N kT
or
D0 hν
ze = g1 exp ( + )
RT 2kT
This page titled 24.7: The Electronic Partition Function of an Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed,
We base the electronic potential energy for a diatomic molecule on a model in which the nuclei are stationary at the bottom of the
electronic potential energy well. We now want to expand this model to include vibrational motion of the atoms along the line
connecting their nuclei. It is simple, logical, and effective to model this motion using the quantum mechanical treatment of the
classical (Hooke’s law) harmonic oscillator.
A Hooke’s law oscillator has a location, r , at which the restoring force, F (r ), and the potential energy, ϵ (r ) , are zero. As it is
0 0 0
displaced from r , the oscillator experiences a restoring force that is proportional to the magnitude of the displacement,
0
dF = −λ dr . Then, we have
r r
∫ dF = −λ ∫ dr
r0 r0
so that F (r) − F (r ) = −λ (r − r ) . Since F (r ) = 0 , we have

o 0 o F (r) = −λ (r − r0 ) . The change in the oscillator’s potential
energy is proportional to the square of the displacement,
r r
λ 2
ϵ (r) − ϵ (ro ) = ∫ −F dr = λ ∫ (r − r0 ) dr = (r − r0 )
r0 r0
2
Since we take ϵ (r o) =0 , we have ϵ (r) = λ(r − r 0)

2
/2 . Taking the second derivative, we find
2
d ϵ
=λ
2
dr
Therefore, if we determine the electronic potential energy function accurately near r , we can find λ from its curvature at r .
0 0
In Chapter 18, we note that the Schrödinger equation for such an oscillator can be solved and that the resulting energy levels are
given by ϵ = hν (n + 1/2) where ν is the vibrational frequency. The relationship between frequency and force constant is
n
−
−−
1 λ
ν = √
2π m
where the oscillator consists of a single moving mass, m. In the case where masses m and m oscillate along the line joining their
1 2
centers, it turns out that the same equations describe the relative motion, if the mass, m, is replaced by the reduced mass
m1 m2
μ =
m1 + m2
Therefore, in principle, we can find the characteristic frequency, ν , of a diatomic molecule by accurately calculating the
dependence of the electronic potential energy on r in the vicinity of r . When we know ν , we know the vibrational energy levels
0
available to the molecule. Alternatively, as discussed in Section 24.7, we can obtain information about the molecule’s vibrational
energy levels from its infrared absorption spectrum and use these data to find ν . Either way, once we know ν , we can evaluate the
vibrational partition function. We have
∞
hν 1 exp (−hν /2kT )
zv = ∑ exp [− (n + )] =
kT 2 1 − exp (−hν /kT )
n=0
where we take advantage of the fact that the vibrational partition function is the sum of a geometric series, as we show in Section
22.6.
This page titled 24.8: The Vibrational Partition Function of A Diatomic Ideal Gas is shared under a CC BY-SA 4.0 license and was authored,
For a diatomic molecule that is free to rotate in three dimensions, we can distinguish two rotational motions; however, their wave
equations are intertwined, and the quantum mechanical result is that there is one set of degenerate rotational energy levels. The
energy levels are
2
J (J + 1) h
ϵr,J =
2
8π I
with degeneracies g J = 2J + 1 , where J = 0, 1, 2, 3, … .

(Recall that I is the moment of inertia, defined as I = ∑ m r , where r is the distance of the i nucleus from the molecule’s
i
2
i i
th
center of mass. For a diatomic molecule, XY , whose internuclear distance is r , the values of r and r must satisfy the XY X Y
conditions r + r = r
X Y and m r = m r . From these relationships, it follows that the moment of inertia is I = μr ,
XY X X Y Y
2
XY
where μ is the reduced mass.) For heteronuclear diatomic molecules, the rotational partition function is
∞ 2
J (J + 1) h
zr = ∑ (2J + 1)exp [ ]
2
8 π I kT
J=0
For homonuclear diatomic molecules, there is a complication. This complication occurs in the quantum mechanical description of
the rotation of any molecule for which there is more than one indistinguishable orientation in space. When we specify the locations
of the atoms in a homonuclear diatomic molecule, like H , we must specify the coordinates of each atom. If we rotate this
2
molecule by 360 in a plane, the molecule and the coordinates are unaffected. If we rotate it by only 180 in a plane, the
o o
coordinates of the nuclei change, but the rotated molecule is indistinguishable from the original molecule. Our mathematical model
distinguishes the 180 -rotated molecule from the original, unrotated molecule, but nature does not.
o
This means that there are twice as many energy levels in the mathematical model as actually occur in nature. The rotational
partition function for a homonuclear diatomic molecule is exactly one-half of the rotational partition function for an “otherwise
identical” heteronuclear diatomic molecule. To cope with this complication in general, it proves to be useful to define a quantity
that we call the symmetry number for any molecule. The symmetry number is usually given the symbol σ; it is just the number of
ways that the molecule can be rotated into indistinguishable orientations. For a homonuclear diatomic molecule, σ = 2 ; for a
heteronuclear diatomic molecule, σ = 1 .
Making use of the symmetry number, the rotational partition function for any diatomic molecule becomes
∞ 2
1 J (J + 1) h
zr = ( ) ∑ (2J + 1)exp [ ] (24.9.1)
2
σ 8 π I kT
J=0
For most molecules at ordinary temperatures, the lowest rotational energy level is much less than kT , and this infinite sum can be
approximated to good accuracy as the corresponding integral. That is
∞ 2
1 J (J + 1) h
zr ≈ ( )∫ (2J + 1) exp [ ] dJ
2
σ J=0 8 π I kT
Initial impressions notwithstanding, this integral is easily evaluated. The substitutions a = h 2 2

/8 π I kT and u = J (J + 1) yield
∞
1
zr ≈ ( )∫ exp (−au) du (24.9.2)
σ u=0
2
1 1 8 π I kT
≈( )( ) = (24.9.3)
2
σ a σh
To see that this is a good approximation for most molecules at ordinary temperatures, we calculate the successive terms in the
partition function of the hydrogen molecule at 25 C. The results are shown in Table 1. We choose hydrogen because the energy
difference between successive rotational energy levels becomes greater the smaller the values of I and T . Since hydrogen has the
smallest angular momentum of any molecule, the integral approximation will be less accurate for hydrogen than for any other
molecule at the same temperature. For hydrogen, summing the first seven terms in the exact calculation (Equation 24.9.1) gives
zrotation= 1.87989, whereas the approximate calculation (Equation 24.9.3) gives 1.70284. This difference corresponds to a
difference of 245 J in the rotational contribution to the standard Gibbs free energy of molecular hydrogen.
Table 1: Rotational Partition Function Contributions for Molecular Hydrogen at 298 K
2
(2J+1) J(J+1)h Zr
J =
σ
ZJ
ex p (−
8π 2 I kT
) ≈ ∑ ZJ
0 0.50000 0.50000
1 0.83378 1.33378
2 0.42935 1.76313
3 0.10323 1.86637
4 0.01267 1.87904
5 0.00082 1.87986
6 0.00003 1.87989
This page titled 24.9: The Rotational Partition Function of A Diatomic Ideal Gas is shared under a CC BY-SA 4.0 license and was authored,
In our discussion of ensembles, we find that the thermodynamic functions for a system can be expressed as functions of the
system’s partition function. Now that we have found the molecular partition function for a diatomic ideal gas molecule, we can find
the partition function, Z , for a gas of N such molecules. From this system partition function, we can find all of the
IG
thermodynamic functions for this N -molecule ideal-gas system. The system entropy, energy, and partition function are related to
each other by the equation
E
S = + kln Z IG
T
Rearranging, and adding (P V ) system

to both sides, we find the Gibbs free energy
G = E − T S + (P V )system = (P V )system − kT ln ZIG
For a system of one mole of an ideal gas, we have (P V )

system
= N kT
¯
¯¯¯
¯
. If the ideal gas is diatomic, we can substitute the
molecular partition functions developed above to find
¯
¯¯¯
¯
GIG = N kT − kT ln ZIG
¯
¯¯¯
¯
N
⎡ (zt ) ⎤
¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯
= N kT − kTln − N kT ln zr − N kT ln zv − N kT ln ze
¯
¯¯¯
¯
⎣ N! ⎦
3/2 2
¯
¯¯¯
¯ ¯
¯¯¯
¯ ¯
¯¯¯
¯ 2πmkT kT ¯
¯¯¯
¯ 8 π I kT ¯
¯¯¯
¯
exp (−hν /2kT )
= N kT − N kT − N kT ln[ ( ) ] − N kT ln( ) − N kT ln( )
2 2
h P σh 1 − exp (−hν /kT )
¯
¯¯¯
¯ D0 hν
− N kT ln( + )
RT 2kT
For the standard Gibbs free energy of an ideal gas, we define the pressure to be one bar. Introduction of this condition
= 1 bar = 10 Pa) and further simplification gives
o 5
(P = P
3/2 2
2πmkT kT 8 π I kT exp (−hν /2kT ) D0 hν
o
G = −RT ln[ ( ) ] − RT ln( ) − RT ln( ) − RT ( + )
IG 2 o 2
h P σh 1 − exp (−hν /kT ) RT 2kT
In this form, the successive terms represent, respectively, the translational, rotational, vibrational, and electronic contributions to
the Gibbs free energy. Further simplification results because vibrational and electronic contributions from terms involving hν /2kT
cancel. This is a computational convenience. Factoring out RT ,
3/2 2
2πmkT kT 8 π I kT D0
o
G = −RT {ln[ ( ) ] + ln( ) − ln(1 − exp (−hν /kT )) + }
IG 2 o 2
h P σh RT
This page titled 24.10: The Gibbs Free Energy for One Mole of An Ideal Gas is shared under a CC BY-SA 4.0 license and was authored, remixed,
For many diatomic molecules, the data needed to calculate G are readily available in various compilations. For illustration, we
o
IG
consider the molecules H , I , and H I . The necessary experimental data are summarized in Table 2.
2 2
Table 2: Data1 for the calculation of partition functions for H 2 ,

(g) I2 (g) , and H I (g)
Compound Molar mass, g D0 , kJ mol-1 ν , hertz rX Y ,m

14 −11
H2 2.016 432.073 1.31948 × 10 7.4144 × 10
12 −10
I2 253.82 148.81 6.43071 × 10 2.666 × 10
13 −10
HI 127.918 294.67 6.69227 × 10 1.60916 × 10
The terms in the simplified equation for the standard Gibbs free energy at 298.15 K are given in Table 3.
Table 3: Gibbs free energy components
3/2 2
2π mkT kT 8π I kT D0
Compound ln[(
h
2
)
po
] ln(
σh
2
) − ln(1 − e
−hν/kT
)
RT
H2 126.23929 0.6312* 0.0000 174.295
I2 133.49256 7.932 0.4388 60.0289
HI 132.46470 3.4604 0.00002 118.868
*Calculated as a sum of terms (see Table 1) rather than as the integral approximation.
Finally, the standard molar Gibbs Free Energies at 298.15 K are summarized in Table 4.
Table 4: Calculated Gibbs free energies
Compound G
o
298 K
, kJ mole
−1
H2 −746.577
I2 −500.471
HI −631.622
These results can be used to calculate the standard Gibbs free energy change, at 298.15 K, for the reaction
H2 (g) + I2 (g) → 2H I (g) .
We find
o o o o
Δr G = 2G (H I , g, 298.15 K) − G (H2 , g, 298.15 K) − G (I2 , g, 298.15 K) = −16.20 kJ
298
This page titled 24.11: The Standard Gibbs Free Energy for H₂(g), I₂(g), and HI(g) is shared under a CC BY-SA 4.0 license and was authored,
The standard Gibbs free energies of formation for H I (g) and I (g) are 1.7 kJ mol and 19.3 kJ mol , respectively.
1
2
−1 −1
Calculation of the Gibbs free energy of this reaction from thermochemical data gives Δ G (298.15 K) = −15.9 kJ. The
r
o
difference between this value and the value calculated above is 0.3 kJ. The magnitude of this difference is consistent with the
number of significant figures given for the tabulated thermochemical data. However, some error results because we have used the
simplest possible quantum mechanical models for rotational and vibrational motions. The accuracy of the statistical–mechanical
calculation can be increased by using models in which the vibrational oscillator does not follow Hooke’s law exactly and in which
the rotating molecule is not strictly rigid.
This page titled 24.12: The Gibbs Free Energy Change for Forming HI(g) from H₂(g) and I₂(g) is shared under a CC BY-SA 4.0 license and was
In Sections 24.11 and 24.12, we see that the standard Gibbs free energy, G , that we calculate from our statistical thermodynamic
o
model is not the same quantity as the Gibbs free energy of formation, Δ G . Nevertheless, these calculations show that we can use
f
o
the statistical-thermodynamic Gibbs free energies of the reacting species to calculate the Gibbs free energy change for a reaction in
exactly the same way that we use the corresponding Gibbs free energies of formation.
The use of Gibbs free energies of formation for these calculations is successful because we measure all Gibbs free energies of
formation relative to the Gibbs free energies of the constituent elements in their standard states. By convention, we set the standard-
state Gibbs free energies of the elements equal to zero, but this is incidental; our method is successful because the Gibbs free
energies of the constituent elements cancel out when we calculate the Gibbs free energy change for a reaction from the Gibbs free
energies of formation of the reacting species.
Our statistical-mechanical Gibbs free energies represent the Gibbs free energy change for a different process. They correspond to
the formation of the molecule from its isolated constituent atoms. The isolated constituent atoms are the reference state for our
statistical-mechanical calculation of standard molar Gibbs free energies. We choose the Gibbs free energies of the isolated atoms to
be zero. (Whatever Gibbs free energies we might assign to the isolated atoms, they cancel out when we calculate the Gibbs free
energy change for a reaction from the statistical-thermodynamic Gibbs free energies of the reacting species.)
When we sum the component energies of our model for a diatomic molecule, we have
ϵmolecule = ϵt + ϵr + ϵv + ϵe .
The smallest of these quantum mechanically allowed values for ϵ is particularly significant in our present considerations.
molecule
Once we have created this molecule in its lowest energy state, we can consider that we can get it into any other state just by adding
energy to it. When the isolated constituent atoms are the reference state, the value of the lowest-energy state of the molecule is the
energy exchanged with the surroundings when the molecule is formed in this state from the constituent atoms.
Reviewing our models for the motions that a diatomic molecule can undergo, we see that the translational and rotational energies
can be zero. The smallest vibrational energy is hν /2, and the smallest electronic energy is
D0 hν
−( + )
¯
¯¯¯
¯ 2
N
The minimum molecular energy is ϵ = −D / N < 0 . Since D / N is the energy required to just separate the diatomic
minimum
molecule
0
¯
¯¯¯
¯
0
¯
¯¯¯
¯
molecule into its constituent elements, the end product of this process is two stationary atoms, situated at an infinite distance from
one another. Conversely, ϵ is the energy released when the stationary constituent atoms approach one another from infinite
minimum
molecule
separation to form the diatomic molecule in its lowest energy state. The reference state for the statistical-mechanical calculation of
molecular thermodynamic properties is a set of isolated constituent atoms that have no kinetic energy. The stipulation that the
reference-state atoms have no kinetic energy is often expressed by saying that the reference state is the constituent atoms at the
absolute zero of temperature.
This page titled 24.13: The Reference State for Molecular Partition Functions is shared under a CC BY-SA 4.0 license and was authored, remixed,
24.14: Problems
1. The partition function, Z , for a system of N , distinguishable, non-interacting molecules is Z = z , where z is the molecular N
partition function, z = ∑ g exp (−ϵ /kT ) , and the ϵ and g are the energy levels available to the molecule and their
i i i i
degeneracies. Show that the thermodynamic functions for the N -molecule system depend on the molecular partition function as
follows:
(a) E = N kT 2
(
∂ln z
∂T
)
V
(b) S = N kT ( ∂ln z
∂T
) + N kln z
V
(c) A = −N kT ln z
(d) P system = N kT (
∂ln z
∂V
)
T
(e) H = N kT 2
(
∂ln z
∂T
) + N kT V (
∂ln z
∂V
)
V T
(f) G = −N kT ln z + N kT V (
∂ln z
∂V
)
T
2. When the number of available quantum states is much larger than the number of molecules, the partition function, Z , for a
system of N , indistinguishable, non-interacting molecules is Z = z /N ! , where z is the molecular partition function, N
z = ∑ g exp (−ϵ /kT ) , and the ϵ and g are the energy levels available to the molecule and their degeneracies. Show that the
i i i i
thermodynamic functions for the N-molecule system depend on the molecular partition function as follows:
(a) E = N kT 2
(
∂ln z
∂T
)
V
(b) S = N k [T ( ∂ln z
∂T
) + ln
z
N
+ 1]
V
(c) A = −N kT [1 + ln N
z
]
(d) P system = N kT (
∂ln z
∂V
)
T
(e) H = N kT 2
(
∂ln z
∂T
) + N kT V (
∂ln z
∂V
)
V T
(f) G = −N kT [1 + ln z
N
+V (
∂ln z
∂V
) ]
T
3. The molecular partition function for the translational motion of an ideal gas is
3/2
2πmkT
zt = ( ) V
2
h
The partition function for a gas of N , monatomic, ideal-gas molecules is Z = z N

t
/N ! . Show that the thermodynamic functions are
as follows:
(a) E = 3
2
N kT
(b) S = N k [ 5
2
+ ln
z
N
]
(c) A = −N kT [1 + ln N
z
]
(d) P system =
N kT
(e) H = 5
2
N kT
(f) G = −N kT ln N
z
4. Find E , S , A , H , and G for one mole of Xenon at 300 K and 1 bar.

Notes
1
Data from the Handbook of Chemistry and Physics, 79 th
Ed., David R. Linde, Ed., CRC Press, New York, 1998.
CHAPTER OVERVIEW
25: Bose-Einstein and Fermi-Dirac Statistics

In developing the theory of statistical thermodynamics and the Boltzmann distribution function, we assume that molecules are
distinguishable and that any number of molecules in a system can have the same quantum mechanical description. These
assumptions are not valid for many chemical systems. Fortunately, it turns out that more rigorous treatment of the conditions
imposed by quantum mechanics usually leads to the same conclusions as the Boltzmann treatment. The Boltzmann treatment can
become inadequate when the system consists of low-mass particles (like electrons) or when the system temperature is near absolute
zero.
Thumbnail: Comparison of average occupancy of the ground state for three statistics. (CC BY-SA 4.0; Victor Blacus via
Wikipedia)
This page titled 25: Bose-Einstein and Fermi-Dirac Statistics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated
request.
1
In developing the theory of statistical thermodynamics and the Boltzmann distribution function, we assume that molecules are
distinguishable and that any number of molecules in a system can have the same quantum mechanical description. These
assumptions are not valid for many chemical systems. Fortunately, it turns out that more rigorous treatment of the conditions
1,2
imposed by quantum mechanics usually leads to the same conclusions as the Boltzmann treatment. The Boltzmann treatment can
become inadequate when the system consists of low-mass particles (like electrons) or when the system temperature is near absolute
zero.
In this chapter, we introduce some modifications that make our statistical model more rigorous. We consider systems that contain
large numbers of particles. We address the effects that the principles of quantum mechanics have on the equilibrium states that are
possible, but we continue to assume that the particles do not otherwise exert forces on one another. We derive distribution functions
for statistical models that satisfy quantum-mechanical restrictions on the number of particles that can occupy a particular quantum
state. Our primary objective is to demonstrate that the more rigorous models reduce to the Boltzmann distribution function for most
chemical systems at common laboratory conditions.
We have been using the quantum mechanical result that the discrete energy levels of a molecule or other particle can be labeled ϵ , 1
ϵ ,…, ϵ ,…. We have assumed that we can put any number of identifiable particles into any of these energy levels. We have
2 i
assumed also that we can distinguish one particle from another, so that we can know the energy of any particular particle. In fact,
we may not be able to tell the particles apart. In this case, we can know how many particles have a given energy, but we cannot
distinguish the particles that have this energy from one another. Moreover, there is a quantum-mechanical theorem about the
number of particles that can occupy a quantum state. If the particles have integral (0, 1, 2, … . ) spin, any number of them can
occupy the same quantum state. Such particles are said to follow Bose-Einstein statistics. If on the other hand, the particles have
half-integral (1/2, 3/2, 5/2, … . ) spin, then only one of them can occupy a given quantum state. Such particles are said to
follow Fermi-Dirac statistics.
Protons, neutrons, and electrons all have spin 1/2. The spin of an atom or molecule is just the sum of the spins of its constituent
elementary particles. If the number of protons, neutrons, and electrons is odd, the atom or molecule obeys Fermi-Dirac statistics. If
it is even, the atom or molecule obeys Bose-Einstein statistics. For most molecules at temperatures that are not too close to absolute
zero, the predicted difference in behavior is negligible. However, the isotopes of helium provide an important test of the theory.
Near absolute zero, the behavior of H e differs markedly from that of H e . The difference is consistent with the expected
3 4
difference between the behavior of a spin-1/2 particle (H e ) and that of a spin-0 particle (H e ).
3 4
This page titled 25.1: Quantum Statistics is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via
Let us consider the total probability sum for a system of particles that follows Fermi-Dirac statistics. As before, we let ϵ , ϵ ,…, 1 2
ϵ ,…. be the energies of the successive energy levels. We let g , g ,…, g ,…. be the degeneracies of these levels. We let N ,
i 1 2 i 1
N ,…, N ,…. be the number of particles in all of the degenerate quantum states of a given energy level. The probability of finding
2 i
a particle in a quantum state depends on the number of particles in the system; we have ρ (N , ϵ ) rather than ρ (ϵ ). Consequently, i i i
we cannot generate the total probability sum by expanding an equation like

N
1 = (P1 + P2 + ⋯ + Pi + …) .
However, we continue to assume:

1. A finite subset of the population sets available to the system accounts for nearly all of the probability when the system is held in
a constant-temperature environment.
2. Essentially the same finite subset of population sets accounts for nearly all of the probability when the system is isolated.
3. All of the microstates that have a given energy have the same probability. We let this probability be ρ . FD
MS,N ,E
As before, the total probability sum will be of the form

FD FD
1 = ∑W (Ni , ϵi )ρ
MS,N ,E
{ Ni }
Each such term reflects the fact that there are W (N , ϵ ) ways to put N particles in the g
FD
i i quantum states of energy level ϵ ,
1 1 1
and N particles in the g quantum states of energy level ϵ , and, in general, N particles in the g quantum states of energy level
2 2 2 i i
ϵ . Unlike Boltzmann statistics, however, the probabilities are different for successive particles, so the coefficient W is different FD
i
from the polynomial coefficient, or thermodynamic probability, W . Instead, we must discover the number of ways to put N i
indistinguishable particles into the g -fold degenerate quantum states of energy ϵ when a given quantum state can contain at most
i i
one particle.
These conditions can be satisfied only if g i ≥ Ni . If we put N of the particles into quantum states of energy ϵ , there are
i i
1. g ways to place the first particle, but only

i
2. g − 1 ways to place the second, and

i
3. g − 2 ways to place the third, and

i
4. …
5. g − (N − 1) ways to place the last one of the N particles.
i i i
This means that there are
(gi ) (gi − 1) (gi − 2) … (gi − (Ni + 1)) =
(gi ) (gi − 1) (gi − 2) … (gi − (Ni + 1)) (gi − Ni ) … (1) gi !

= =
(gi − Ni )! (gi − Ni )!
ways to place the N particles. Because the particles cannot be distinguished from one another, we must exclude assignments
i
which differ only by the way that the N particles are permuted. To do so, we must divide by N !. The number of ways to put N
i i i
indistinguishable particles into g quantum states with no more than one particle in a quantum state is
i
gi !
(gi − Ni )! Ni !
The number of ways to put indistinguishable Fermi-Dirac particles of the population set {N 1, N2 , … , Ni , …} into the available
energy states is
∞
g1 ! g2 ! gi ! gi !
FD
W (Ni , gi ) = [ ]×[ ]×⋯ ×[ ]×⋯ = ∏[ ]
(g1 − N1 )! N1 ! (g2 − N2 )! N2 ! (gi − Ni )! Ni ! (gi − Ni )! Ni !
i=1
so that the total probability sum for a Fermi-Dirac system becomes
∞
gi ! Ni
FD
1 = ∑ ∏[ ][ ρ (ϵi )]
(gi − Ni )! Ni !
{ Nj } i=1
To find the Fermi-Dirac distribution function, we seek the population set {N 1, N2 , … , Ni , …} for which W
FD
is a maximum,
subject to the constraints
∞
N = ∑ Ni
i=1
and
∞
E = ∑ Ni ϵi
i=1
The mnemonic function becomes

∞ ∞ ∞ ∞
FD
Fmn = ∑ ln gi ! − ∑ [(gi − Ni ) ln(gi − Ni ) − (gi − Ni )] − ∑ [ Ni ln Ni − Ni ] + α [N − ∑ Ni ]
i=1 i=1 i=1 i=1
+ β [E − ∑ Ni ϵi ]
i=1
We seek the N for which F

i
FD
mn is an extremum; that is, the N satisfying i
FD
∂Fmn gi − N N
i i
0 = = + ln(gi − N ) − 1 − − ln N + 1 − α − β ϵi
i i
∂Ni
gi − N N
i i
= ln(gi − N ) − ln N − α − β ϵi
i i
Solving for N , we find

i
−α −βϵi
gi e e
N =
i
1 + e−α e−βϵi
or, equivalently,
N 1
i
=
gi 1 + eα eβϵi
If 1 ≫ e e
−α −βϵi
(or 1 ≪ e e α βϵi
), the Fermi-Dirac distribution function reduces to the Boltzmann distribution function. It is easy to
see that this is the case. From
−α −βϵi
gi e e
−α −βϵi
N = ≈ gi e e
i
1 + e−α e−βϵi
and N =∑
∞
i=1
N
i
, we have
∞
−α −βϵi −α
N =e ∑ gi e =e z
i=1
It follows that e . With β = 1/kT , we recognize that N /N is the Boltzmann distribution. For occupied energy levels,
α
= z/N
i
/kT ≈ 1 ; otherwise, e < 1 . This means that the Fermi-Dirac distribution simplifies to the Boltzmann
−βϵi −ϵi −βϵi −ϵi /kT
e =e =e
distribution whenever 1 ≫ e . We can illustrate that this is typically the case by considering the partition function for an ideal
−α
gas.
Using the translational partition function for one mole of a monatomic ideal gas from Section 24.3, we have
3/2 ¯
¯¯¯
zt 2πmkT V
α
e = =[ ]
2 ¯
¯¯¯
¯
N h
N
3/2
2πmkT kT
=[ ]
2 0
h P
For an ideal gas of molecular weight 40 at 300 K and 1 bar, we find e = 1.02 × 10 and e = 9.77 × 10 . Clearly, the
α 7 −α −8
condition we assume in demonstrating that the Fermi-Dirac distribution simplifies to the Boltzmann distribution is satisfied by
molecular gases at ordinary temperatures. The value of e decreases as the temperature and the molecular weight decrease. To find
α
≈ 1 for a molecular gas, it is necessary to consider very low temperatures.

α
e
Nevertheless, the Fermi-Dirac distribution has important applications. The behavior of electrons in a conductor can be modeled on
the assumption that the electrons behave as a Fermi-Dirac gas whose energy levels are described by a particle-in-a-box model.
This page titled 25.2: Fermi-Dirac Statistics and the Fermi-Dirac Distribution Function is shared under a CC BY-SA 4.0 license and was authored,
For particles that follow Bose-Einstein statistics, we let the probability of a microstate of energy E in an N -particle system be
ρ
BE
MS,N ,E
. For an isolated system of Bose-Einstein particles, the total probability sum is
BE BE
1 = ∑W (Ni , gi ) ρ (25.3.1)
MS,N ,E
{ Ni }
We need to find W (N , g ), the number of ways to assign indistinguishable particles to the quantum states, if any number of
BE
i i
particles can occupy the same quantum state.

We begin by considering the number of ways that N particles can be assigned to the g quantum states associated with the energy
i i
level ϵ . We see that the fewest number of quantum states that can be used is one; we can put all of the particles into one quantum
i
state. At the other extreme, we cannot use more than the N quantum states that we use when we give each particle its own
i
quantum state. We can view this problem as finding the number of way we can draw as many as g boxes around N points. Let us i i
create a scheme for drawing such boxes. Suppose we have a linear frame on which there is a row of locations. Each location can
hold one particle. The frame is closed at both ends. Between each successive pair of particle-holding locations, there is a slot, into
which a wall can be inserted. This frame is sketched in Figure 1.
Figure 1. Scheme to assign Bose-Einstein particles to degenerate energy levels.

When we insert (g − 1) walls into these slots, the frame contains g boxes. We want to be able to insert the walls so that the N
i i i
particles are distributed among the g boxes in such a way that we can have any desired number of particles in any desired number
i
of boxes. (Of course, placement of the walls is subject to the constraints that we use at most g boxes and exactly N particles.) We
i i
can achieve this by constructing the frame to have (N + g − 1) particle-holding locations. To see this, we think about the case
i i
that requires the largest number of particle-holding locations. This is the case in which all N particles are in one box. (See Figure
i
2.) For this case, we need N occupied locations and (g − 1) unoccupied locations.
i i
Figure 2. Maximum size frame for N particles in g locations.

i i
Now we consider the number of ways that we can insert (g − 1) walls into the (N + g − 1) slots. The first wall can go into any
i i i
of (N + g − 1) slots. The second can go into any of (N + g − 1 − (−1)) or (N + g − 2) slots. The last wall can go into any
i i i i i i
of (N + g − 1 − (g − 2)) or (N + 1) slots. The total number of ways of inserting the (g − 1) walls is therefore
i i i i i
(Ni + gi − 1) (Ni + gi − 2) … (Ni + 1) (Ni ) … (2) (1)
(Ni + gi − 1) (Ni + gi − 2) … (Ni + 1) =
Ni !
(Ni + gi − 1)!
=
Ni !
This total is greater than the answer we seek, because it includes all permutations of the walls. It does not matter whether the first,
the second, or the last wall occupies a given slot. Therefore, the expression we have obtained over-counts the quantity we seek by
the factor (g − 1)! , which is the number of ways of permuting the (g − 1) walls. We have therefore that the N particles can be
i i i
assigned to g quantum states in

i
(Ni + gi − 1)!
Ni ! (gi − 1)!
ways, and hence

(N1 + g1 − 1)! (N2 + g2 − 1)! (Ni + gi − 1)!
BE
W (Ni , gi ) =[ ]×[ ]×⋯ ×[ ]×…
(g1 − 1)! N1 ! (g2 − 1)! N2 ! (gi − 1)! Ni !
∞
(Ni + gi − 1)!
= ∏[ ]
(gi − 1)! Ni !
i=1
so that the total probability sum for a Bose-Einstein system becomes (via Equation 25.3.1):
∞
(Ni + gi − 1)! Ni
BE
1 = ∑ ∏[ ][ ρ (ϵi )]
(gi − 1)! Ni !
{ Nj } i=1
To find the Bose-Einstein distribution function, we seek the population set {N 1, N2 , … , Ni , …} for which W BE
is a maximum,
subject to the constraints
∞
N = ∑ Ni
i=1
and
∞
E = ∑ Ni ϵi
i=1
The mnemonic function is

∞
BE
Fmn = ∑ [(Ni + gi − 1) ln(Ni + gi − 1) − (Ni + gi − 1) − Ni ln N1 + Ni − (gi − 1) ln(gi − 1) + (gi − 1) ]
i=1
∞ ∞
+ α (N − ∑ Ni ) + β (E − ∑ Ni ϵi )
i=1 i=1
We seek the N for which F

i
BE
mn is an extremum; that is, the N satisfying i
BE
∂Fmn
0 =
∂N
i
N + gi − 1 N
i i
= + ln(N + gi − 1) − 1 − − ln N + 1 − α − β ϵi
i i
N + gi − 1 N
i i
N
i
= −ln − α − β ϵi
N + gi − 1
i
Solving for N , we find

i
−α −βϵi −α −βϵi
(gi − 1) e e gi e e
N = ≈
i −α −βϵi −α −βϵi
1 −e e 1 −e e
where the last expression takes advantage of the fact that g is usually a very large number, so the error introduced by replacing
i
(g − 1) by g is usually negligible. If 1 ≫ e , the Bose-Einstein distribution function reduces to the Boltzmann distribution
−α −βϵi
i i e
function. As we find in Section 25.2, this is always the case for a molecular gas at ambient temperatures.
Notes
1
Richard C. Tolman, The Principles of Statistical Mechanics, Dover Publications, Inc., New York, 1979, pp 367-378. (This is a
republication of the book originally published in 1938 by Oxford University Press.)
2
Malcom Dole, Introduction to Statistical Thermodynamics, Prentice Hall, Inc., New York, 1954, pp 206-215.
This page titled 25.3: Bose-Einstein Statistics and the Bose-Einstein Distribution Function is shared under a CC BY-SA 4.0 license and was
CHAPTER OVERVIEW
26: Appendices
26.2: Appendix B.
26.4: Appendix C.
26.6: Appendix D.
26: Appendices is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by LibreTexts.
1
[Scaled to Ar(12C) = 12, where 12C is a neutral atom in its nuclear and electronic ground state]
Atomic Number Name Symbol Atomic Weight
1 Hydrogen H 1.00794
2 Helium He 4.002602
3 Lithium Li [6.941(2)]
4 Beryllium Be 9.012182
5 Boron B 10.811
6 Carbon C 12.0107
7 Nitrogen N 14.0067
8 Oxygen O 15.9994
9 Fluorine F 18.9984032
10 Neon Ne 20.1797
11 Sodium Na 22.989770
12 Magnesium Mg 24.3050
13 Aluminum Al 26.981538
14 Silicon Si 28.0855
15 Phosphorus P 30.973761
16 Sulfur S 32.065
17 Chlorine Cl 35.453
18 Argon Ar 39.948
19 Potassium K 39.0983
20 Calcium Ca 40.078
21 Scandium Sc 44.955910
22 Titanium Ti 47.867
23 Vanadium V 50.9415
24 Chromium Cr 51.9961
25 Manganese Mn 54.938049
26 Iron Fe 55.845
27 Cobalt Co 58.933200
28 Nickel Ni 58.6934
29 Copper Cu 63.546
30 Zinc Zn 65.39
31 Gallium Ga 69.723
32 Germanium Ge 72.64
33 Arsenic As 74.92160
34 Selenium Se 78.96
35 Bromine Br 79.904
36 Krypton Kr 83.80
37 Rubidium Rb 85.4678
38 Strontium Sr 87.62
39 Yttrium Y 88.90585
40 Zirconium Zr 91.224
41 Niobium Nb 92.90638
42 Molybdenum Mo 95.94
98
43 Technetium* Tc 97.9072
44 Ruthenium Ru 101 07
44 Ruthenium Ru 101.07
45 Rhodium Rh 102.90550
46 Palladium Pd 106.42
47 Silver Ag 107.8682
48 Cadmium Cd 112.411
49 Indium In 114.818
50 Tin Sn 118.710
51 Antimony Sb 121.760
52 Tellurium Te 127.60
53 Iodine I 126.90447
54 Xenon Xe 131.293
55 Cesium Cs 132.90545
56 Barium Ba 137.327
57 Lanthanum La 138.9055
58 Cerium Ce 140.116
59 Praseodymium Pr 140.90765
60 Neodymium Nd 144.24
145
61 Promethium Pm 144.9127
62 Samarium Sm 150.36
63 Europium Eu 151.964
64 Gadolinium Gd 157.25
65 Terbium Tb 158.92534
66 Dysprosium Dy 162.50
67 Holmium Ho 164.93032
68 Erbium Er 167.259
69 Thulium Tm 168.93421
70 Ytterbium Yb 173.04
71 Lutetium Lu 174.967
72 Hafnium Hf 178.49
73 Tantalum Ta 180.9479
74 Tungsten W 183.84
75 Rhenium Re 186.207
76 Osmium Os 190.23
77 Iridium Ir 192.217
78 Platinum Pt 195.078
79 Gold Au 196.96655
80 Mercury Hg 200.59
81 Thallium Tl 204.3833
82 Lead Pb 207.2
83 Bismuth Bi 208.98038
210
84 Polonium* Po 209.9829
210
85 Astatine* At 209.9871
222
86 Radon* Rn 222.0176
223
87 Francium* Fr 223.0197
226
88 Radium* Ra 226.0254
227
89 Actinium* Ac 227.0277
90 Thorium* Th 232.0381
91 Protactinium* Pa 231.03588
92 U i * U 238 02891
92 Uranium* U 238.02891
237
93 Neptunium* Np 237.0482
244
94 Plutonium* Pu 244.0642
243
95 Americium* Am 243.0614
247
96 Curium* Cm 247.0704
247
97 Berkelium* Bk 247.0703
251
98 Californium* Cf 251.0796
252
99 Einsteinium* Es 252.0830
257
100 Fermium* Fm 257.0951
258
101 Mendelevium* Md 258.0984
259
102 Nobelium* No 259.1010
262
103 Lawrencium* Lr 262.1097
261
104 Rutherfordium* Rf 261.1088
†
This table is slightly modified from that given in “ATOMIC WEIGHTS OF THE ELEMENTS, 1999” published by the
International Union of Pure and Applied Chemistry, Inorganic Chemistry Division, Commission on Atomic Weights and Isotopic
Abundances. Elements 105 – 118 are omitted. The Commission’s report was prepared for publication by T. B. Coplen, U.S.
Geological Survey, 431 National Center, Reston, Virginia 20192, USA. See
http://www.physics.curtin.edu.au/iupac/docs/Atwt1999.doc
*Element has no stable nuclides. Three such elements (Th, Pa, and U ) have a characteristic terrestrial isotopic composition, and
for these an atomic weight is tabulated. When the element symbol is listed with an atomic number, that isotope has the longest half-
life and its atomic weight is tabulated.
This page titled 26.1: Appendix A. Standard Atomic Weights 1999† is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or
upon request.
26.2: Appendix B.
This page titled 26.2: Appendix B. is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
Quantity Symbol Value Unit
–1
speed of light in a vacuum c 299792458 m s
–34
Planck constant h 6.62606876 × 10 J s
–19
elementary charge e 1.602176462 × 10 C
–31
electron mass me 9.10938188 × 10 kg
–27
proton mass mp 1.67262158 × 10 kg
¯
¯¯¯ 23 –1
Avogadro constant N 6.02214199 × 10 mol
–1
Faraday constant F 96485.3415 C mol
–1 –1
molar gas constant R 8.314472 J mol K
–2 –1 –1
R 8.314472 × 10 bar L mol K
–2 –1 –1
R 8.205745 × 10 atm L mol K
–1 –1
R 1.987206 cal mol K
–23 –1
Boltzmann constant k 1.3806503 × 10 J K
standard acceleration of free

–2
g 9.806650 m s
fall at the earth’s surface†
†
Source: 1998 CODATA recommended values. Peter J. Mohr and Barry N. Taylor, Reviews of Modern Physics, Vol. 72, No. 2, pp.
351-495, 2000. See www.physics.nist.gov/constants.
†
While it is included here for convenience, g is not a Fundamental Constant. Value from David R. Lide, CRC Handbook of
Chemistry and Physics, 79th Ed., CRC Press, 1999-2000, pp. 1-27, reproduced from NIST Special Publication 811, Guide for the
Use of the International System of Units (Superintendent of Documents, U. S. Government Printing Office, 1991).
26.3: Appendix B. Fundamental Constants† is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.
26.4: Appendix C.
This page titled 26.4: Appendix C. is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
1 m = 10
–10
Å
= 3.280840 ft
= 39.37008 in
1 kg = 2.204622 lb
1 lb = 453.5925 g
3 3
1 m = 10 L
3
= 35.31467 ft
= 264.1720 US liquid gallon
–2
1 N = 1 kg m s
2
= 1 Pa m
–5 2
= 10 bar m
5 –2
= 10 dyne (g cm s )
= 0.1382550 poundal
5
1 bar = 10 Pa
5 –2
= 10 N m
= 0.98692 atm
= 750.064 torr
1 atm = 101325 Pa
= 1.01325 bar
1 J = 1 N m
3
= 1 Pa m
= 1 W s
= 2 C V
7 2 –2
= 10 erg (g cm s )
–5 3
= 10 bar m
–2
= 10 bar L
−3
= 9.86923 × 10 L atm
= 0.2390 cal
1 cal = 4.184 J
–19
1 eV = 1.602176462 × 10 J
(Energy released when one electron
experiences a potential change of one volt.)
Conversion factors in this table are given by, or calculated from, values given by David R. Lide, CRC Handbook of Chemistry and
Physics, 79th Ed., CRC Press, 1999-2000, pp. 1-24 to 1-31, reproduced from NIST Special Publication 811, Guide for the Use of
the International System of Units (Superintendent of Documents, U. S. Government Printing Office, 1991).
26.5: Units and Conversion Factors is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.
26.6: Appendix D.
This page titled 26.6: Appendix D. is shared under a CC BY-SA 4.0 license and was authored, remixed, and/or curated by Paul Ellgen via source
We frequently need the values of the definite integrals below. These values are available in standard tables. Note that integrands
involving even powers of the argument are even functions; integrands involving odd powers are odd functions. (A function, f (x),
is even if f (x) = f (−x); it is odd if f (x) = −f (−x).) The integrals are given over the interval 0 < x < ∞ . For integrands that
are even functions, the integrals over the interval −∞ < x < ∞ are twice the integrals over the interval 0 < x < ∞ . For
integrands that are odd functions, the integrals over the interval −∞ < x < ∞ are zero.
−
−
∞ 2 1 π
∫ exp (−ax ) dx = √
0 2 a
∞ 1
2
∫ x exp (−ax ) dx =
0 2a
∞
−−
2 2 1 π
∫ x exp (−ax ) dx = √ 3
0 4 a
∞ 3 2 1
∫ x exp (−ax ) dx = 2
0 2a
−−
∞ 4 2 3 π
∫ x exp (−ax ) dx = √
0 8 a
5
∞
−−
6 2 15 π
∫ x exp (−ax ) dx = √ 7
0 16 a
26.7: Appendix D. Some Important Definite Integrals is shared under a not declared license and was authored, remixed, and/or curated by
LibreTexts.
Index
A C CSTR
absolute entropy calorie (unit) 5.11: Rate Laws from Experiments in a Continuous
Stirred Tank Reactor
11.5: Absolute Entropy 7.9: Measuring Heat
Cumulative Distribution Function
absolute zero calorimetry
3.8: A Heuristic View of the Cumulative Distribution
2.5: The Kelvin Temperature Scale 8.9: Calorimetry Function
11.13: Absolute Zero is Unattainable canonical ensemble cumulative probability distribution
activated complex 23.1: Ensembles of N-molecule Systems
5.6: Mechanisms and Elementary Processes function
Carnot cycle
activation energy 3.5: Bar Graphs and Histograms
9.2: The Carnot Cycle for an Ideal Gas and the
5.4: The Effect of Temperature on Reaction Rates Entropy Concept
activity Carnot engine D
14.10: Chemical Activity 9.2: The Carnot Cycle for an Ideal Gas and the Dalton’s law of partial pressures
16.19: Finding Solute Activity Using the Entropy Concept 2.14: Gas Mixtures - Dalton's Law of Partial
Hypothetical One-molal Standard State catalyst Pressures
activity coefficient 5.5: Other Factors that Affect Reaction Rates de Broglie Wavelength
14.10: Chemical Activity cathode 18.2: Quantized Energy - De Broglie's Hypothesis
16.3: Expressing the Activity Coefficient as a and the Schroedinger Equation
Deviation from Raoult's Law degeneracy
Adiabatic Changes Cell Notation
6.10: Duhem's Theorem - Specifying Reversible detailed balance
Change in A Closed System center of mass frame
adiabatic demagnetization 4.10: Collisions between Gas Molecules Relative
Velocity Coordinates distribution coefficient
Amagat's Law Central Limit Theorem 6.4: Distribution Equilibria
3.11: The Variance of the Average- The Central Distribution Equilibria
2.15: Gas Mixtures - Amagat's Law of Partial
Limit Theorem 6.4: Distribution Equilibria
Volums
anode chemical potential Duhem's Theorem
14: Chemical Potential - Extending the Scope of the 6.10: Duhem's Theorem - Specifying Reversible
Fundamental Equation Change in A Closed System
Arrhenius equation 14.1: Dependence of the Internal Energy on the
5.4: The Effect of Temperature on Reaction Rates Composition of the System
average speed in gas Clapeyron Equation
E
4.8: Statistics for Molecular Speeds 12.6: Phase Equilibria - Temperature Dependence of
electrical potential
the Boiling Point 17.5: Electrical Potential
Avogadro’s number
12.8: The Clapeyron Equation electrical work
Avogadro's number closed system 7.8: Measuring Work- Non-Pressure-Volume Work
6.1: The Thermodynamic Perspective electrode
coefficient of isothermal compressibility 17.2: Electrochemical Cells
B 10.4: Expressing Thermodynamic Functions with electrolytic cell
Independent Variables V and T 17.7: The Direction of Electron Flow and its
Balancing Redox Equations coefficient of thermal expansion Implications
17.4: Balancing Oxidation-reduction Reactions 8.3: How Enthalpy Depends on Pressure elementary process
Bar Graphs 10.4: Expressing Thermodynamic Functions with 5.6: Mechanisms and Elementary Processes
3.5: Bar Graphs and Histograms Independent Variables V and T
10.9: The Relationship Between Cv and Cp for Any ensemble
barometer Substance 23.1: Ensembles of N-molecule Systems
2.11: The Barometric Formula ensemble treatment
cognitive dissonance
barometric formula 1.9: A Few Ideas from the Philosophy of Science 20.3: The Population Sets of a System at Equilibrium
2.11: The Barometric Formula at Constant N, V, and T
colligative property
batch reaction 16.9: Properties of Ideal Solutions enthalpy
5.2: Reaction Rates and Rate Laws 7.15: Defining Enthalpy, H
collision radius
boiling point elevation 4.12: The Frequency of Collisions between Unlike
8.1: Enthalpy
16.10: Colligative Properties - Boiling-point Gas Molecules entropy of mixing
Elevation 13.3: The Thermodynamics of Mixing Ideal Gases
collision theory
boiling point elevation constant 5.7: Rate Laws for Elementary Processes equal a priori probabilities
16.10: Colligative Properties - Boiling-point 20.8: The Probabilities of Microstates that Have the
Elevation
Compound Interest
Same Energy
Boltzmann constant equation of state
2.7: The Ideal Gas Constant and Boltzmann's
compressibility factor
6.7: The Number of Variables Required to Specify
Constant 2.13: Virial Equations Some Familiar Systems
Boltzmann equation Conservation of Energy equilibrium constant
20.3: The Population Sets of a System at Equilibrium 7.10: The First Law of Thermodynamics 5.17: Chemical Equilibrium as the Equality of Rates
at Constant N, V, and T Continuous Distribution Functions for Opposing Reactions
21.4: Deriving the Boltzmann Equation II 3.6: Continuous Distribution Functions - the equilibrium dissociation energy
Boltzmann Statistics Envelope Function is the Derivative of the Area 24.7: The Electronic Partition Function of an Ideal
20: Boltzmann Statistics contrapositive Gas
1.10: A Few Ideas from Formal Logic exact differential
Exact Differentials ideal solution Method of Initial Rates
7.4: Exact Differentials and State Functions 16.2: Raoult's Law and Ideal Solutions 5.8: Experimental Determination of Rate Laws
exactness 16.9: Properties of Ideal Solutions microscopic reversibility
7.5: Determining Whether an Expression is an Exact inexact differential 5.18: The Principle of Microscopic Reversibility
Differential 7.2: The Total Differential 5.19: Microscopic Reversibility and the Second Law
exogenous variable inhibitor microstate
21.7: The Microscopic Model for Reversible Change 5.5: Other Factors that Affect Reaction Rates 20.7: The Microstates of a Given Population Set
expectation value ionic strength molecular partition function
3.9: Random Variables, Expected Values, and 16.18: Activities of Electrolytes - The Debye-Hückel 21.1: Finding the Boltzmann Equation
Population Sets Theory most probable speed in gas
expected value irreversible work 4.8: Statistics for Molecular Speeds
3.9: Random Variables, Expected Values, and 7.18: Reversible vs. Irreversible Pressure-Volume
Population Sets Work
N
extent of reaction isolated system
5.2: Reaction Rates and Rate Laws 6.1: The Thermodynamic Perspective
Nernst heat theorem
11.4: Genesis of the Third Law - the Nernst Heat
isomeric molecules Theorem
F 21.5: Partition Functions and Equilibrium - Isomeric
normal distribution
Molecules
faraday (unit) 3.12: The Normal Distribution
isothermal compressibility 3.13: The Expected Value of a Function of Several
17.8: Electrolysis and the Faraday 10.4: Expressing Thermodynamic Functions with Variables and the Central Limit Theorem
Independent Variables V and T
first law of thermodynamics 10.9: The Relationship Between Cv and Cp for Any
7.10: The First Law of Thermodynamics Substance O
flooding isothermal process open system
5.8: Experimental Determination of Rate Laws 7.17: Free Expansion of a Gas 6.1: The Thermodynamic Perspective
5.12: Predicting Rate Laws from Proposed
Mechanisms
osmotic pressure
free expansion J 16.12: Colligative Properties - Osmotic Pressure
7.17: Free Expansion of a Gas joule (unit) Oxidation Potential

7.9: Measuring Heat 17.12: Standard Electrode Potentials
fugacity
oxidation states
fugacity coefficient K 17.3: Defining Oxidation States
11.7: The Fugacity of a Gas kinetic isotope effect

P
G partial molar quantities
galvanic cell L 14.3: Partial Molar Quantities
17.6: Electrochemical Cells as Circuit Elements Lagrange’s method of undetermined partial pressure
2.14: Gas Mixtures - Dalton's Law of Partial
Gibbs Energy of Formation multipliers Pressures
11.9: The Standard Entropy and the Gibbs Free 21.2: Lagrange's Method of Undetermined
Energy of Formation
partial volume
Multipliers
Gibbs' Phase Rule Law of Dulong and Petit Pressures
6.8: Gibbs' Phase Rule 7.14: Heat Capacities of Solids- the Law of Dulong partition function
Gibbsian equilibrium manifold and Petit
6.12: Equilibria and Reversible Processes Law of Mass Action permutations
5.17: Chemical Equilibrium as the Equality of Rates
19.1: Distribution of Results for Multiple Trials with
H for Opposing Reactions
Two Possible Outcomes
half cell Le Chatelier's Principle Phase Equilibria
17.11: Half-reactions and Half-cells 12.7: Phase Equilibria - Temperature Dependence of
half life Lindemann mechanism the Melting Point
5.14: The Lindemann-Hinshelwood Mechanism for principle of microscopic reversibility
First-order Decay
Hamiltonian 1.2: Chemical Kinetics
line integrals probability density
18.2: Quantized Energy - De Broglie's Hypothesis
7.3: Line Integrals
and the Schroedinger Equation 3.6: Continuous Distribution Functions - the
Henry's law Envelope Function is the Derivative of the Area
16.6: Henry's Law and the Hypothetical One-molal
M Probability of Alternative Events
Standard State mean 3.3: Some Important Properties of Events
Hess’s Law 3.10: Statistics - the Mean and the Variance of a Probability of Compound Events
8.6: Standard Enthalpies of Reaction Distribution 3.3: Some Important Properties of Events
histograms mean activity coefficient proposition
3.5: Bar Graphs and Histograms 16.17: Activities of Electrolytes - The Mean Activity 1.10: A Few Ideas from Formal Logic
Coefficient
mean free path
I 4.12: The Frequency of Collisions between Unlike
R
ideal gas constant Gas Molecules rank probability
2.7: The Ideal Gas Constant and Boltzmann's medium effect 3.8: A Heuristic View of the Cumulative Distribution
Constant Function
ideal gas thermometer melting point Rankine scale
2.9: Temperature and the Ideal Gas Thermometer 2.5: The Kelvin Temperature Scale
12.7: Phase Equilibria - Temperature Dependence of
ideal gases the Melting Point Raoult's Law
2.8: Real Gases Versus Ideal Gases 16.2: Raoult's Law and Ideal Solutions
Rapid Equilibrium Approximation SN1CB symmetry number
5.12: Predicting Rate Laws from Proposed 5.16: The Mechanism of the Base Hydrolysis of 24.9: The Rotational Partition Function of A
Mechanisms Co(NH₃)₅Xⁿ⁺ Diatomic Ideal Gas
rate constant Solubility of gases
5.2: Reaction Rates and Rate Laws 16.14: Colligative Properties - Solubility of a Gas T
rate law solvent cage test for exactness
5.2: Reaction Rates and Rate Laws 5.7: Rate Laws for Elementary Processes 7.5: Determining Whether an Expression is an Exact
reaction mechanism spectroscopic dissociation energy Differential
5.6: Mechanisms and Elementary Processes 24.7: The Electronic Partition Function of an Ideal the second moment about the mean
Reaction order Gas 3.10: Statistics - the Mean and the Variance of a
5.2: Reaction Rates and Rate Laws standard deviation Distribution
reaction rate 3.10: Statistics - the Mean and the Variance of a Thermochemical Cycles
Distribution 11.12: Evaluating Entropy Changes Using
redox reaction standard electrode potential Thermochemical Cycles
17.12: Standard Electrode Potentials thermodynamic probability
reduced mass standard entropy 20.7: The Microstates of a Given Population Set
11.9: The Standard Entropy and the Gibbs Free Third Law of Thermodynamics
4.10: Collisions between Gas Molecules Relative
Energy of Formation 11.3: The Third Law
Velocity Coordinates
Reduction Potential standard hydrogen electrode Total Differentials
17.10: The Standard Hydrogen Electrode (S.H.E) 7.2: The Total Differential
reversible work transition state
state function 5.6: Mechanisms and Elementary Processes
7.18: Reversible vs. Irreversible Pressure-Volume
Work
7.1: Changes in a State Function are Independent of
Path V
S state functions van der Waals equation
salt bridge 7.4: Exact Differentials and State Functions 2.12: Van der Waals' Equation
17.2: Electrochemical Cells Steric Factor Variance
second order reaction 5.7: Rate Laws for Elementary Processes 3.10: Statistics - the Mean and the Variance of a
5.7: Rate Laws for Elementary Processes Stirling’s Approximation Distribution
selective oxidation 19.4: Stirling's Approximation virial equation
8.6: Standard Enthalpies of Reaction 2.13: Virial Equations
Index
A C CSTR
absolute entropy calorie (unit) 5.11: Rate Laws from Experiments in a Continuous
Stirred Tank Reactor
11.5: Absolute Entropy 7.9: Measuring Heat
Cumulative Distribution Function
absolute zero calorimetry
3.8: A Heuristic View of the Cumulative Distribution
2.5: The Kelvin Temperature Scale 8.9: Calorimetry Function
11.13: Absolute Zero is Unattainable canonical ensemble cumulative probability distribution
activated complex 23.1: Ensembles of N-molecule Systems
5.6: Mechanisms and Elementary Processes function
Carnot cycle
activation energy 3.5: Bar Graphs and Histograms
9.2: The Carnot Cycle for an Ideal Gas and the
5.4: The Effect of Temperature on Reaction Rates Entropy Concept
activity Carnot engine D
14.10: Chemical Activity 9.2: The Carnot Cycle for an Ideal Gas and the Dalton’s law of partial pressures
16.19: Finding Solute Activity Using the Entropy Concept 2.14: Gas Mixtures - Dalton's Law of Partial
Hypothetical One-molal Standard State catalyst Pressures
activity coefficient 5.5: Other Factors that Affect Reaction Rates de Broglie Wavelength
14.10: Chemical Activity cathode 18.2: Quantized Energy - De Broglie's Hypothesis
16.3: Expressing the Activity Coefficient as a and the Schroedinger Equation
Deviation from Raoult's Law degeneracy
Adiabatic Changes Cell Notation
6.10: Duhem's Theorem - Specifying Reversible detailed balance
Change in A Closed System center of mass frame
adiabatic demagnetization 4.10: Collisions between Gas Molecules Relative
Velocity Coordinates distribution coefficient
Amagat's Law Central Limit Theorem 6.4: Distribution Equilibria
3.11: The Variance of the Average- The Central Distribution Equilibria
2.15: Gas Mixtures - Amagat's Law of Partial
Limit Theorem 6.4: Distribution Equilibria
Volums
anode chemical potential Duhem's Theorem
14: Chemical Potential - Extending the Scope of the 6.10: Duhem's Theorem - Specifying Reversible
Fundamental Equation Change in A Closed System
Arrhenius equation 14.1: Dependence of the Internal Energy on the
5.4: The Effect of Temperature on Reaction Rates Composition of the System
average speed in gas Clapeyron Equation
E
4.8: Statistics for Molecular Speeds 12.6: Phase Equilibria - Temperature Dependence of
electrical potential
the Boiling Point 17.5: Electrical Potential
Avogadro’s number
12.8: The Clapeyron Equation electrical work
Avogadro's number closed system 7.8: Measuring Work- Non-Pressure-Volume Work
6.1: The Thermodynamic Perspective electrode
coefficient of isothermal compressibility 17.2: Electrochemical Cells
B 10.4: Expressing Thermodynamic Functions with electrolytic cell
Independent Variables V and T 17.7: The Direction of Electron Flow and its
Balancing Redox Equations coefficient of thermal expansion Implications
17.4: Balancing Oxidation-reduction Reactions 8.3: How Enthalpy Depends on Pressure elementary process
Bar Graphs 10.4: Expressing Thermodynamic Functions with 5.6: Mechanisms and Elementary Processes
3.5: Bar Graphs and Histograms Independent Variables V and T
10.9: The Relationship Between Cv and Cp for Any ensemble
barometer Substance 23.1: Ensembles of N-molecule Systems
2.11: The Barometric Formula ensemble treatment
cognitive dissonance
barometric formula 1.9: A Few Ideas from the Philosophy of Science 20.3: The Population Sets of a System at Equilibrium
2.11: The Barometric Formula at Constant N, V, and T
colligative property
batch reaction 16.9: Properties of Ideal Solutions enthalpy
5.2: Reaction Rates and Rate Laws 7.15: Defining Enthalpy, H
collision radius
boiling point elevation 4.12: The Frequency of Collisions between Unlike
8.1: Enthalpy
16.10: Colligative Properties - Boiling-point Gas Molecules entropy of mixing
Elevation 13.3: The Thermodynamics of Mixing Ideal Gases
collision theory
boiling point elevation constant 5.7: Rate Laws for Elementary Processes equal a priori probabilities
16.10: Colligative Properties - Boiling-point 20.8: The Probabilities of Microstates that Have the
Elevation
Compound Interest
Same Energy
Boltzmann constant equation of state
2.7: The Ideal Gas Constant and Boltzmann's
compressibility factor
6.7: The Number of Variables Required to Specify
Constant 2.13: Virial Equations Some Familiar Systems
Boltzmann equation Conservation of Energy equilibrium constant
20.3: The Population Sets of a System at Equilibrium 7.10: The First Law of Thermodynamics 5.17: Chemical Equilibrium as the Equality of Rates
at Constant N, V, and T Continuous Distribution Functions for Opposing Reactions
21.4: Deriving the Boltzmann Equation II 3.6: Continuous Distribution Functions - the equilibrium dissociation energy
Boltzmann Statistics Envelope Function is the Derivative of the Area 24.7: The Electronic Partition Function of an Ideal
20: Boltzmann Statistics contrapositive Gas
1.10: A Few Ideas from Formal Logic exact differential
Exact Differentials ideal solution Method of Initial Rates
7.4: Exact Differentials and State Functions 16.2: Raoult's Law and Ideal Solutions 5.8: Experimental Determination of Rate Laws
exactness 16.9: Properties of Ideal Solutions microscopic reversibility
7.5: Determining Whether an Expression is an Exact inexact differential 5.18: The Principle of Microscopic Reversibility
Differential 7.2: The Total Differential 5.19: Microscopic Reversibility and the Second Law
exogenous variable inhibitor microstate
21.7: The Microscopic Model for Reversible Change 5.5: Other Factors that Affect Reaction Rates 20.7: The Microstates of a Given Population Set
expectation value ionic strength molecular partition function
3.9: Random Variables, Expected Values, and 16.18: Activities of Electrolytes - The Debye-Hückel 21.1: Finding the Boltzmann Equation
Population Sets Theory most probable speed in gas
expected value irreversible work 4.8: Statistics for Molecular Speeds
3.9: Random Variables, Expected Values, and 7.18: Reversible vs. Irreversible Pressure-Volume
Population Sets Work
N
extent of reaction isolated system
5.2: Reaction Rates and Rate Laws 6.1: The Thermodynamic Perspective
Nernst heat theorem
11.4: Genesis of the Third Law - the Nernst Heat
isomeric molecules Theorem
F 21.5: Partition Functions and Equilibrium - Isomeric
normal distribution
Molecules
faraday (unit) 3.12: The Normal Distribution
isothermal compressibility 3.13: The Expected Value of a Function of Several
17.8: Electrolysis and the Faraday 10.4: Expressing Thermodynamic Functions with Variables and the Central Limit Theorem
Independent Variables V and T
first law of thermodynamics 10.9: The Relationship Between Cv and Cp for Any
7.10: The First Law of Thermodynamics Substance O
flooding isothermal process open system
5.8: Experimental Determination of Rate Laws 7.17: Free Expansion of a Gas 6.1: The Thermodynamic Perspective
5.12: Predicting Rate Laws from Proposed
Mechanisms
osmotic pressure
free expansion J 16.12: Colligative Properties - Osmotic Pressure
7.17: Free Expansion of a Gas joule (unit) Oxidation Potential

7.9: Measuring Heat 17.12: Standard Electrode Potentials
fugacity
oxidation states
fugacity coefficient K 17.3: Defining Oxidation States
11.7: The Fugacity of a Gas kinetic isotope effect

P
G partial molar quantities
galvanic cell L 14.3: Partial Molar Quantities
17.6: Electrochemical Cells as Circuit Elements Lagrange’s method of undetermined partial pressure
Gibbs Energy of Formation multipliers Pressures
11.9: The Standard Entropy and the Gibbs Free 21.2: Lagrange's Method of Undetermined
Energy of Formation
partial volume
Multipliers
Gibbs' Phase Rule Law of Dulong and Petit Pressures
6.8: Gibbs' Phase Rule 7.14: Heat Capacities of Solids- the Law of Dulong partition function
Gibbsian equilibrium manifold and Petit
6.12: Equilibria and Reversible Processes Law of Mass Action permutations
5.17: Chemical Equilibrium as the Equality of Rates
H for Opposing Reactions
Two Possible Outcomes
half cell Le Chatelier's Principle Phase Equilibria
17.11: Half-reactions and Half-cells 12.7: Phase Equilibria - Temperature Dependence of
half life Lindemann mechanism the Melting Point
5.14: The Lindemann-Hinshelwood Mechanism for principle of microscopic reversibility
First-order Decay
Hamiltonian 1.2: Chemical Kinetics
line integrals probability density
18.2: Quantized Energy - De Broglie's Hypothesis
7.3: Line Integrals
and the Schroedinger Equation 3.6: Continuous Distribution Functions - the
Henry's law Envelope Function is the Derivative of the Area
16.6: Henry's Law and the Hypothetical One-molal
M Probability of Alternative Events
Standard State mean 3.3: Some Important Properties of Events
Hess’s Law 3.10: Statistics - the Mean and the Variance of a Probability of Compound Events
8.6: Standard Enthalpies of Reaction Distribution 3.3: Some Important Properties of Events
histograms mean activity coefficient proposition
3.5: Bar Graphs and Histograms 16.17: Activities of Electrolytes - The Mean Activity 1.10: A Few Ideas from Formal Logic
Coefficient
mean free path
I 4.12: The Frequency of Collisions between Unlike
R
ideal gas constant Gas Molecules rank probability
2.7: The Ideal Gas Constant and Boltzmann's medium effect 3.8: A Heuristic View of the Cumulative Distribution
Constant Function
ideal gas thermometer melting point Rankine scale
2.9: Temperature and the Ideal Gas Thermometer 2.5: The Kelvin Temperature Scale
12.7: Phase Equilibria - Temperature Dependence of
ideal gases the Melting Point Raoult's Law
2.8: Real Gases Versus Ideal Gases 16.2: Raoult's Law and Ideal Solutions
Rapid Equilibrium Approximation SN1CB symmetry number
5.12: Predicting Rate Laws from Proposed 5.16: The Mechanism of the Base Hydrolysis of 24.9: The Rotational Partition Function of A
Mechanisms Co(NH₃)₅Xⁿ⁺ Diatomic Ideal Gas
rate constant Solubility of gases
5.2: Reaction Rates and Rate Laws 16.14: Colligative Properties - Solubility of a Gas T
rate law solvent cage test for exactness
5.2: Reaction Rates and Rate Laws 5.7: Rate Laws for Elementary Processes 7.5: Determining Whether an Expression is an Exact
reaction mechanism spectroscopic dissociation energy Differential
5.6: Mechanisms and Elementary Processes 24.7: The Electronic Partition Function of an Ideal the second moment about the mean
Reaction order Gas 3.10: Statistics - the Mean and the Variance of a
5.2: Reaction Rates and Rate Laws standard deviation Distribution
reaction rate 3.10: Statistics - the Mean and the Variance of a Thermochemical Cycles
Distribution 11.12: Evaluating Entropy Changes Using
redox reaction standard electrode potential Thermochemical Cycles
17.12: Standard Electrode Potentials thermodynamic probability
reduced mass standard entropy 20.7: The Microstates of a Given Population Set
11.9: The Standard Entropy and the Gibbs Free Third Law of Thermodynamics
4.10: Collisions between Gas Molecules Relative
Energy of Formation 11.3: The Third Law
Velocity Coordinates
Reduction Potential standard hydrogen electrode Total Differentials
17.10: The Standard Hydrogen Electrode (S.H.E) 7.2: The Total Differential
reversible work transition state
state function 5.6: Mechanisms and Elementary Processes
7.18: Reversible vs. Irreversible Pressure-Volume
Work
7.1: Changes in a State Function are Independent of
Path V
S state functions van der Waals equation
salt bridge 7.4: Exact Differentials and State Functions 2.12: Van der Waals' Equation
17.2: Electrochemical Cells Steric Factor Variance
second order reaction 5.7: Rate Laws for Elementary Processes 3.10: Statistics - the Mean and the Variance of a
5.7: Rate Laws for Elementary Processes Stirling’s Approximation Distribution
selective oxidation 19.4: Stirling's Approximation virial equation
8.6: Standard Enthalpies of Reaction 2.13: Virial Equations
Detailed Licensing
Overview
Title: Thermodynamics and Chemical Equilibrium (Ellgen)
Webpages: 374
Applicable Restrictions: Noncommercial
All licenses found:
CC BY-SA 4.0: 96% (359 pages)
Undeclared: 3.7% (14 pages)
CC BY-NC-SA 4.0: 0.3% (1 page)
By Page
Thermodynamics and Chemical Equilibrium (Ellgen) - CC 2.8: Real Gases Versus Ideal Gases - CC BY-SA 4.0
BY-SA 4.0 2.9: Temperature and the Ideal Gas Thermometer -
Front Matter - Undeclared CC BY-SA 4.0
TitlePage - Undeclared 2.10: Deriving Boyle's Law from Newtonian
InfoPage - Undeclared Mechanics - CC BY-SA 4.0
Table of Contents - Undeclared 2.11: The Barometric Formula - CC BY-SA 4.0
Licensing - Undeclared 2.12: Van der Waals' Equation - CC BY-SA 4.0
About the Author - Undeclared 2.13: Virial Equations - CC BY-SA 4.0
1: Introduction - Background and a Look Ahead - CC
Pressures - CC BY-SA 4.0
BY-SA 4.0
1.1: The Role of the Ideal Gas - CC BY-SA 4.0 - CC BY-SA 4.0
1.2: Chemical Kinetics - CC BY-SA 4.0 2.16: Problems - CC BY-SA 4.0
1.3: Classical Thermodynamics - CC BY-SA 4.0
3: Distributions, Probability, and Expected Values - CC
1.4: Statistical Thermodynamics - CC BY-SA 4.0
BY-SA 4.0
1.5: Heat Transfer in Practical Devices - CC BY-SA
3.1: The Distribution Function as a Summary of
4.0
Experimental Results - CC BY-SA 4.0
1.6: The Concept of Equilibrium - CC BY-SA 4.0
3.2: Outcomes, Events, and Probability - CC BY-SA
1.7: Chemical Equilibrium and Predicting Chemical
4.0
Change - CC BY-SA 4.0
3.3: Some Important Properties of Events - CC BY-SA
1.8: Equilibrium and Classical Thermodynamics - CC
4.0
BY-SA 4.0
3.4: Applying the Laws of Probability - CC BY-SA 4.0
1.9: A Few Ideas from the Philosophy of Science -
3.5: Bar Graphs and Histograms - CC BY-SA 4.0
CC BY-SA 4.0
3.6: Continuous Distribution Functions - the
1.10: A Few Ideas from Formal Logic - CC BY-SA
Envelope Function is the Derivative of the Area - CC
4.0
BY-SA 4.0
1.11: Problems - CC BY-SA 4.0
3.7: A Heuristic View of the Probability Density
2: Gas Laws - CC BY-SA 4.0
Function - CC BY-SA 4.0
2.1: Boyle's Law - CC BY-SA 4.0 3.8: A Heuristic View of the Cumulative Distribution
2.2: Charles' Law - CC BY-SA 4.0 Function - CC BY-SA 4.0
2.3: Avogadro's Hypothesis - CC BY-SA 4.0 3.9: Random Variables, Expected Values, and
2.4: Finding Avogadro's Number - CC BY-SA 4.0 Population Sets - CC BY-SA 4.0
2.5: The Kelvin Temperature Scale - CC BY-SA 4.0 3.10: Statistics - the Mean and the Variance of a
2.6: Deriving the Ideal Gas Law from Boyle's and Distribution - CC BY-SA 4.0
Charles' Laws - CC BY-SA 4.0 3.11: The Variance of the Average- The Central Limit
2.7: The Ideal Gas Constant and Boltzmann's Theorem - CC BY-SA 4.0
Constant - CC BY-SA 4.0 3.12: The Normal Distribution - CC BY-SA 4.0
3.13: The Expected Value of a Function of Several 5.8: Experimental Determination of Rate Laws - CC
Variables and the Central Limit Theorem - CC BY-SA BY-SA 4.0
4.0 5.9: First-order Rate Processes - CC BY-SA 4.0
3.14: Where Does the N - 1 Come from? - CC BY-SA 5.10: Rate Laws by the Study of Initial Rates - CC
4.0 BY-SA 4.0
3.15: Problems - CC BY-SA 4.0 5.11: Rate Laws from Experiments in a Continuous
4: The Distribution of Gas Velocities - CC BY-SA 4.0 Stirred Tank Reactor - CC BY-SA 4.0
4.1: Distribution Functions for Gas-velocity 5.12: Predicting Rate Laws from Proposed
Components - CC BY-SA 4.0 Mechanisms - CC BY-SA 4.0
4.2: Probability Density Functions for Velocity 5.13: The Michaelis-Menten Mechanism for Enzyme-
Components in Spherical Coordinates - CC BY-SA 4.0 catalyzed Reactions - CC BY-SA 4.0
4.3: Maxwell's Derivation of the Gas-velocity 5.14: The Lindemann-Hinshelwood Mechanism for
Probability-density Function - CC BY-SA 4.0 First-order Decay - CC BY-SA 4.0
4.4: The Probability-density Function for Gas 5.15: Why Unimolecular Reactions are First Order -
Velocities in One Dimension - CC BY-SA 4.0 CC BY-SA 4.0
4.5: Combining the One-dimensional Probability 5.16: The Mechanism of the Base Hydrolysis of
Density Functions - CC BY-SA 4.0 Co(NH₃)₅Xⁿ⁺ - CC BY-SA 4.0
4.6: Boyle's Law from the Maxwell-Boltzmann 5.17: Chemical Equilibrium as the Equality of Rates
Probability Density - CC BY-SA 4.0 for Opposing Reactions - CC BY-SA 4.0
4.7: Experimental Test of the Maxwell-Boltzmann 5.18: The Principle of Microscopic Reversibility - CC
Probability Density - CC BY-SA 4.0 BY-SA 4.0
4.8: Statistics for Molecular Speeds - CC BY-SA 4.0 5.19: Microscopic Reversibility and the Second Law -
4.9: Pressure Variations for Macroscopic Samples - CC BY-SA 4.0
CC BY-SA 4.0 5.20: Problems - CC BY-SA 4.0
4.10: Collisions between Gas Molecules Relative 6: Equilibrium States and Reversible Processes - CC BY-
Velocity Coordinates - CC BY-SA 4.0 SA 4.0
4.11: The Probability Density Function for the 6.1: The Thermodynamic Perspective - CC BY-SA 4.0
Relative Velocity - CC BY-SA 4.0 6.2: Thermodynamic Systems and Variables - CC BY-
4.12: The Frequency of Collisions between Unlike SA 4.0
Gas Molecules - CC BY-SA 4.0 6.3: Equilibrium and Reversibility - Phase Equilibria
4.13: The Rate of Collisions between Unlike Gas - CC BY-SA 4.0
Molecules - CC BY-SA 4.0 6.4: Distribution Equilibria - CC BY-SA 4.0
4.14: Collisions between like Gas Molecules - CC 6.5: Equilibria in Chemical Reactions - CC BY-SA 4.0
BY-SA 4.0 6.6: Le Chatelier's Principle - CC BY-SA 4.0
4.15: The Geometry of A Collision between Spherical 6.7: The Number of Variables Required to Specify
Molecules - CC BY-SA 4.0 Some Familiar Systems - CC BY-SA 4.0
4.16: The Energy of A Collision between Gas 6.8: Gibbs' Phase Rule - CC BY-SA 4.0
Molecules - CC BY-SA 4.0 6.9: Reversible vs. Irreversible Processes - CC BY-SA
4.17: Problems - CC BY-SA 4.0 4.0
5: Chemical Kinetics, Reaction Mechanisms, and 6.10: Duhem's Theorem - Specifying Reversible
Chemical Equilibrium - CC BY-SA 4.0 Change in A Closed System - CC BY-SA 4.0
6.11: Reversible Motion of A Mass in A Constant
5.1: Chemical Kinetics - CC BY-SA 4.0
Gravitational Field - CC BY-SA 4.0
5.2: Reaction Rates and Rate Laws - CC BY-SA 4.0
6.12: Equilibria and Reversible Processes - CC BY-SA
5.3: Simultaneous Processes - CC BY-SA 4.0
4.0
5.4: The Effect of Temperature on Reaction Rates -
6.13: The Laws of Thermodynamics - CC BY-SA 4.0
CC BY-SA 4.0
6.14: Thermodynamic Criteria for Change - CC BY-
5.5: Other Factors that Affect Reaction Rates - CC
SA 4.0
BY-SA 4.0
6.15: State Functions in Systems Undergoing
5.6: Mechanisms and Elementary Processes - CC BY-
Spontaneous Change - CC BY-SA 4.0
SA 4.0
5.7: Rate Laws for Elementary Processes - CC BY-SA
4.0 7: State Functions and The First Law - CC BY-SA 4.0
7.1: Changes in a State Function are Independent of 9: The Second Law - Entropy and Spontaneous Change -
Path - CC BY-SA 4.0 CC BY-SA 4.0
7.2: The Total Differential - CC BY-SA 4.0 9.1: The Second Law of Thermodynamics - CC BY-
7.3: Line Integrals - CC BY-SA 4.0 SA 4.0
7.4: Exact Differentials and State Functions - CC BY- 9.2: The Carnot Cycle for an Ideal Gas and the
SA 4.0 Entropy Concept - CC BY-SA 4.0
7.5: Determining Whether an Expression is an Exact 9.3: The Carnot Cycle for Any Reversible System -
Differential - CC BY-SA 4.0 CC BY-SA 4.0
7.6: The Chain Rule and the Divide-through Rule - 9.4: The Entropy Change around Any Cycle for Any
CC BY-SA 4.0 Reversible System - CC BY-SA 4.0
7.7: Measuring Pressure-Volume Work - CC BY-SA 9.5: The Tiling Theorem and the Paths of Cyclic
4.0 Process in Other Spaces - CC BY-SA 4.0
7.8: Measuring Work- Non-Pressure-Volume Work - 9.6: Entropy Changes for A Reversible Process - CC
CC BY-SA 4.0 BY-SA 4.0
7.9: Measuring Heat - CC BY-SA 4.0 9.7: Entropy Changes for A Spontaneous Process in
7.10: The First Law of Thermodynamics - CC BY-SA An Isolated System - CC BY-SA 4.0
4.0 9.8: The Entropy of the Universe - CC BY-SA 4.0
7.11: Other Statements of the First Law - CC BY-SA 9.9: The Significance of The Machine-based
4.0 Statement of The Second Law - CC BY-SA 4.0
7.12: Notation for Changes in Thermodynamic 9.10: A Slightly Philosophical Digression on Energy
Quantities - E vs. ∆E - CC BY-SA 4.0 and Entropy - CC BY-SA 4.0
7.13: Heat Capacities for Gases- Cv, Cp - CC BY-SA 9.11: A Third Statement of the Second Law - CC BY-
4.0 SA 4.0
7.14: Heat Capacities of Solids- the Law of Dulong 9.12: Entropy and Predicting Change - CC BY-SA 4.0
and Petit - CC BY-SA 4.0 9.13: Defining the Helmholtz and Gibbs Free
7.15: Defining Enthalpy, H - CC BY-SA 4.0 Energies - CC BY-SA 4.0
7.16: Heat Transfer in Reversible Processes - CC BY- 9.14: The Fundamental Equation and Other Criteria
SA 4.0 for Reversible Change - CC BY-SA 4.0
7.17: Free Expansion of a Gas - CC BY-SA 4.0 9.15: Entropy and Spontaneous Change - CC BY-SA
7.18: Reversible vs. Irreversible Pressure-Volume 4.0
Work - CC BY-SA 4.0 9.16: Internal Entropy and the Second Law - CC BY-
7.19: Isothermal Expansions of An Ideal Gas - CC SA 4.0
BY-SA 4.0 9.17: Notation and Terminology- Conventions for
7.20: Adiabatic Expansions of An Ideal Gas - CC BY- Spontaneous Processes - CC BY-SA 4.0
SA 4.0 9.18: The Heat Exchanged by A Spontaneous Process
7.21: Problems - CC BY-SA 4.0 at Constant Entropy - CC BY-SA 4.0
8: Enthalpy and Thermochemical Cycles - CC BY-SA 4.0 9.19: The Energy Change for A Spontaneous Process
8.1: Enthalpy - CC BY-SA 4.0 at Constant S and V - CC BY-SA 4.0
8.2: Using Thermochemical Cycles to Find Enthalpy 9.20: The Enthalpy Change for A Spontaneous
Changes - CC BY-SA 4.0 Process at Constant S and P - CC BY-SA 4.0
8.3: How Enthalpy Depends on Pressure - CC BY-SA 9.21: The Entropy Change for A Spontaneous Process
4.0 at Constant E and V - CC BY-SA 4.0
8.4: Standard States and Enthalpies of Formation - 9.22: The Entropy Change for A Spontaneous Process
CC BY-SA 4.0 at Constant H and P - CC BY-SA 4.0
8.5: The Ideal Gas Standard State - CC BY-SA 4.0 9.23: The Reversible Work is the Minimum Work at
8.6: Standard Enthalpies of Reaction - CC BY-SA 4.0 Constant Tˆ - CC BY-SA 4.0
8.7: Standard State Heat Capacities - CC BY-SA 4.0 9.24: The Free Energy Changes for A Spontaneous
8.8: How The Enthalpy Change for a Reaction Process at Constant Tˆ - CC BY-SA 4.0
Depends on Temperature - CC BY-SA 4.0 9.25: Summary- Thermodynamic Functions as
8.9: Calorimetry - CC BY-SA 4.0 Criteria for Change - CC BY-SA 4.0
8.10: Problems - CC BY-SA 4.0 9.26: Problems - CC BY-SA 4.0
10: Some Mathematical Consequences of the 11.12: Evaluating Entropy Changes Using
Fundamental Equation - CC BY-SA 4.0 Thermochemical Cycles - CC BY-SA 4.0
10.1: Thermodynamic Relationships from dE, dH, dA 11.13: Absolute Zero is Unattainable - CC BY-SA 4.0
and dG - CC BY-SA 4.0 11.14: Problems - CC BY-SA 4.0
10.2: dE = TdS - PdV and Internal consistency - CC 12: Applications of the Thermodynamic Criteria for
BY-SA 4.0 Change - CC BY-SA 4.0
10.3: Expressing Thermodynamic Functions with 12.1: Mechanical Processes - CC BY-SA 4.0
Other Independent Variables - CC BY-SA 4.0 12.2: The Direction of Spontaneous Heat Transfer -
10.4: Expressing Thermodynamic Functions with CC BY-SA 4.0
Independent Variables V and T - CC BY-SA 4.0 12.3: Phase Changes - the Fusion of Ice - CC BY-SA
10.5: Expressing Thermodynamic Functions with 4.0
Independent Variables P and T - CC BY-SA 4.0 12.4: Measuring the Entropy Change for Any
10.6: The Transformation of Thermodynamic Reversible Process - CC BY-SA 4.0
Variables in General - CC BY-SA 4.0 12.5: Another Perspective on the Principle of Le
10.7: Reversibility and Thermodynamic Variables in Chatelier - CC BY-SA 4.0
Gerneral - CC BY-SA 4.0 12.6: Phase Equilibria - Temperature Dependence of
10.8: Using the Pair (V, P) or the Pair (T, S) as the Boiling Point - CC BY-SA 4.0
Independent Variables - CC BY-SA 4.0 12.7: Phase Equilibria - Temperature Dependence of
10.9: The Relationship Between Cv and Cp for Any the Melting Point - CC BY-SA 4.0
Substance - CC BY-SA 4.0 12.8: The Clapeyron Equation - CC BY-SA 4.0
10.10: The Dependence of Cv on Volume and of Cp 12.9: The Clausius-Clapeyron Equation - CC BY-SA
on Pressure - CC BY-SA 4.0 4.0
10.11: The Gibbs-Helmholtz Equation - CC BY-SA 12.10: Problems - CC BY-SA 4.0
4.0
13: Equilibria in Reactions of Ideal Gases - CC BY-SA
10.12: The Second Law and the Properties of Ideal
4.0
Gases - CC BY-SA 4.0
13.1: The Gibbs Free Energy of an Ideal Gas - CC
10.13: The second-dependence of the Energy and
BY-SA 4.0
Enthalpy of A Real Gas - CC BY-SA 4.0
13.2: The Gibbs Free Energy Change for A Reaction
10.14: The Joule-Thomson Effect - CC BY-SA 4.0
of Ideal Gases - CC BY-SA 4.0
13.3: The Thermodynamics of Mixing Ideal Gases -
11: The Third Law, Absolute Entropy, and the Gibbs Free
CC BY-SA 4.0
Energy of Formation - CC BY-SA 4.0
13.4: The Gibbs Free Energy Change for Reaction at
11.1: Heat Capacity as a Function of Temperature - Constant Partial Pressures - CC BY-SA 4.0
CC BY-SA 4.0 13.5: ∆rG is the rate at which the Gibbs Free Energy
11.2: Enthalpy as a function of Temperature - CC BY- Changes with The Extent of Reaction - CC BY-SA 4.0
SA 4.0 13.6: The Standard Gibbs Free Energy Change and
11.3: The Third Law - CC BY-SA 4.0 Equilibrium in Ideal Gas Reactions - CC BY-SA 4.0
11.4: Genesis of the Third Law - the Nernst Heat 13.7: The Gibbs Free Energy of Formation and
Theorem - CC BY-SA 4.0 Equilibrium in Ideal Gas Reactions - CC BY-SA 4.0
11.5: Absolute Entropy - CC BY-SA 4.0 13.8: Equilibrium When A Component is Also
11.6: The Standard State for Third-law Entropies - Present as A Condensed Phase - CC BY-SA 4.0
CC BY-SA 4.0 13.9: Equilibrium When An Ideal Gas Component is
11.7: The Fugacity of a Gas - CC BY-SA 4.0 Also Present as A solute - CC BY-SA 4.0
11.8: A General Strategy for Expressing the 13.10: Problems - CC BY-SA 4.0
Thermodynamic Properties of a Substance - CC BY- 14: Chemical Potential - Extending the Scope of the
SA 4.0 Fundamental Equation - CC BY-SA 4.0
11.9: The Standard Entropy and the Gibbs Free
14.1: Dependence of the Internal Energy on the
Energy of Formation - CC BY-SA 4.0
Composition of the System - CC BY-SA 4.0
11.10: The Nature of Hypothetical States - CC BY-SA
14.2: Dependence of Other Thermodynamic
4.0
Functions on the Composition - CC BY-SA 4.0
11.11: The Fugacity and Gibbs Free Energy of A
14.3: Partial Molar Quantities - CC BY-SA 4.0
Substance in Any System - CC BY-SA 4.0
14.4: Chemical Potentials and Stoichiometry - CC 16.6: Henry's Law and the Hypothetical One-molal
BY-SA 4.0 Standard State - CC BY-SA 4.0
14.5: ∑µjdnj = 0 and Primitive Vs. Gibbsian 16.7: Finding the Activity of a Solute from the
Equilibrium - CC BY-SA 4.0 Activity of the Solvent - CC BY-SA 4.0
14.6: The Change Criteria in A System Composed of 16.8: When the Solute Obeys Henry's Law, the
Subsystems - CC BY-SA 4.0 Solvent Obeys Raoult's Law - CC BY-SA 4.0
14.7: At Constant P and T, ∆rµ is the Change in 16.9: Properties of Ideal Solutions - CC BY-SA 4.0
Gibbs Free Energy - CC BY-SA 4.0 16.10: Colligative Properties - Boiling-point
14.8: Gibbs-Duhem Equation - CC BY-SA 4.0 Elevation - CC BY-SA 4.0
14.9: The Dependence of Chemical Potential on 16.11: Colligative Properties - Freezing-point
Other Variables - CC BY-SA 4.0 Depression - CC BY-SA 4.0
14.10: Chemical Activity - CC BY-SA 4.0 16.12: Colligative Properties - Osmotic Pressure - CC
14.11: Back to the Fugacity- the Fugacity of A BY-SA 4.0
Component of A Gas Mixture - CC BY-SA 4.0 16.13: Colligative Properties - Solubility of a Solute
14.12: Relating Fugacity and Chemical Activity - CC in an Ideal Solution - CC BY-SA 4.0
BY-SA 4.0 16.14: Colligative Properties - Solubility of a Gas -
14.13: Relating the Differentials of Chemical CC BY-SA 4.0
Potential and Activity - CC BY-SA 4.0 16.15: Solvent Activity Coefficients from Freezing-
14.14: Dependence of Activity on Temperature- point Depression Measurements - CC BY-SA 4.0
Relative Partial Molar Enthalpies - CC BY-SA 4.0 16.16: Electrolytic Solutions - CC BY-SA 4.0
14.15: Problems - CC BY-SA 4.0 16.17: Activities of Electrolytes - The Mean Activity
15: Chemical Potential, Fugacity, Activity, and Coefficient - CC BY-SA 4.0
Equilibrium - CC BY-SA 4.0 16.18: Activities of Electrolytes - The Debye-Hückel
15.1: The Chemical Potential and Fugacity of a Gas - Theory - CC BY-SA 4.0
CC BY-SA 4.0 16.19: Finding Solute Activity Using the
15.2: The Chemical Potential and Activity of a Gas - Hypothetical One-molal Standard State - CC BY-SA
CC BY-SA 4.0 4.0
15.3: The Pressure-dependence of the Fugacity and 16.20: Problems - CC BY-SA 4.0
Activity of a Condensed Phase - CC BY-SA 4.0 17: Electrochemistry - CC BY-SA 4.0
15.4: Standard States for the Fugacity and Activity of 17.1: Oxidation-reduction Reactions - CC BY-SA 4.0
a Pure Solid - CC BY-SA 4.0 17.2: Electrochemical Cells - CC BY-SA 4.0
15.5: The Chemical Potential, Fugacity, and Activity 17.3: Defining Oxidation States - CC BY-SA 4.0
of a Pure Solid - CC BY-SA 4.0 17.4: Balancing Oxidation-reduction Reactions - CC
15.6: Chemical Potential, Fugacity, and Equilibrium - BY-SA 4.0
CC BY-SA 4.0 17.5: Electrical Potential - CC BY-SA 4.0
15.7: Chemical Potential, Activity, and Equilibrium - 17.6: Electrochemical Cells as Circuit Elements - CC
CC BY-SA 4.0 BY-SA 4.0
15.8: The Rate of Gibbs Free Energy Change with 17.7: The Direction of Electron Flow and its
Extent of Reaction - CC BY-SA 4.0 Implications - CC BY-SA 4.0
15.9: Problems - CC BY-SA 4.0 17.8: Electrolysis and the Faraday - CC BY-SA 4.0
16: The Chemical Activity of the Components of a 17.9: Electrochemistry and Conductivity - CC BY-SA
Solution - CC BY-SA 4.0 4.0
17.10: The Standard Hydrogen Electrode (S.H.E) -
16.1: Solutions Whose Components are in
CC BY-SA 4.0
Equilibrium with Their Own Gases - CC BY-SA 4.0
17.11: Half-reactions and Half-cells - CC BY-SA 4.0
16.2: Raoult's Law and Ideal Solutions - CC BY-SA
17.12: Standard Electrode Potentials - CC BY-SA 4.0
4.0
17.13: Predicting the Direction of Spontaneous
16.3: Expressing the Activity Coefficient as a
Change - CC BY-SA 4.0
Deviation from Raoult's Law - CC BY-SA 4.0
17.14: Cell Potentials and the Gibbs Free Energy -
16.4: Henry's Law and the Fugacity and Activity of A
CC BY-SA 4.0
Solution Component - CC BY-SA 4.0
17.15: The Nernst Equation - CC BY-SA 4.0
16.5: Expressing the Activity Coefficient as A
Deviation from Henry's Law - CC BY-SA 4.0
17.16: The Nernst Equation for Half-cells - CC BY-SA 20.9: The Probabilities of the Population Sets of an
4.0 Isolated System - CC BY-SA 4.0
17.17: Combining two Half-cell Equations to Obtain 20.10: Entropy and Equilibrium in an Isolated System
a new Half-cell Equation - CC BY-SA 4.0 - CC BY-SA 4.0
17.18: The Nernst Equation and the Criterion for 20.11: Thermodynamic Probability and Equilibrium
Equilibrium - CC BY-SA 4.0 in an Isomerization Reaction - CC BY-SA 4.0
17.19: Problems - CC BY-SA 4.0 20.12: The Degeneracy of an Isolated System and Its
18: Quantum Mechanics and Molecular Energy Levels - Entropy - CC BY-SA 4.0
CC BY-SA 4.0 20.13: The Degeneracy of an Isolated System and its
18.1: Energy Distributions and Energy Levels - CC Entropy - CC BY-SA 4.0
BY-SA 4.0 20.14: Effective Equivalence of the Isothermal and
18.2: Quantized Energy - De Broglie's Hypothesis Constant-energy Conditions - CC BY-SA 4.0
and the Schroedinger Equation - CC BY-SA 4.0 20.15: Problems - CC BY-SA 4.0
18.3: The Schrödinger Equation for A Particle in A 21: The Boltzmann Distribution Function - CC BY-SA 4.0
Box - CC BY-SA 4.0 21.1: Finding the Boltzmann Equation - CC BY-SA
18.4: The Schrödinger Equation for a Molecule - CC 4.0
BY-SA 4.0 21.2: Lagrange's Method of Undetermined
18.5: Solutions to Schroedinger Equations for Multipliers - CC BY-SA 4.0
Harmonic Oscillators and Rigid Rotors - CC BY-SA 21.3: Deriving the Boltzmann Equation I - CC BY-SA
4.0 4.0
18.6: Wave Functions, Quantum States, Energy 21.4: Deriving the Boltzmann Equation II - CC BY-SA
Levels, and Degeneracies - CC BY-SA 4.0 4.0
18.7: Particle Spins and Statistics- Bose-Einstein and 21.5: Partition Functions and Equilibrium - Isomeric
Fermi-Dirac Statistics - CC BY-SA 4.0 Molecules - CC BY-SA 4.0
19: The Distribution of Outcomes for Multiple Trials - 21.6: Finding ß and the Thermodynamic Functions
CC BY-SA 4.0 for Distinguishable Molecules - CC BY-SA 4.0
- CC BY-SA 4.0
Two Possible Outcomes - CC BY-SA 4.0
21.8: The Third Law of Thermodynamics - CC BY-SA
4.0
Three Possible Outcomes - CC BY-SA 4.0
21.9: The Partition Function for a System of N
Molecules - CC BY-SA 4.0
Many Possible Outcomes - CC BY-SA 4.0
19.4: Stirling's Approximation - CC BY-SA 4.0
19.5: Problems - CC BY-SA 4.0 22: Some Basic Applications of Statistical
20: Boltzmann Statistics - CC BY-SA 4.0 Thermodynamics - CC BY-SA 4.0
20.1: The Independent-Molecule Approximation - CC 22.1: Interpreting the Partition Function - CC BY-SA
BY-SA 4.0 4.0
20.2: The Probability of An Energy Level at Constant 22.2: Conditions under which Integrals Approximate
N, V, and T - CC BY-SA 4.0 Partition Functions - CC BY-SA 4.0
20.3: The Population Sets of a System at Equilibrium 22.3: Probability Density Functions from the Energies
at Constant N, V, and T - CC BY-SA 4.0 of Classical-mechanical Models - CC BY-SA 4.0
20.4: How can Infinitely Many Probabilities Sum to 22.4: Partition Functions and Average Energies at
Unity? - CC BY-SA 4.0 High Temperatures - CC BY-SA 4.0
20.5: The Total Probability Sum at Constant N, V, 22.5: Energy Levels for a Three-dimensional
and T - CC BY-SA 4.0 Harmonic Oscillator - CC BY-SA 4.0
20.6: The Most Probable Population Set at Constant 22.6: Energy and Heat Capacity of the "Einstein
N, V, and T - CC BY-SA 4.0 Crystal" - CC BY-SA 4.0
20.7: The Microstates of a Given Population Set - CC 22.7: Applications of Other Entropy Relationships -
BY-SA 4.0 CC BY-SA 4.0
20.8: The Probabilities of Microstates that Have the 22.8: Problems - CC BY-SA 4.0
Same Energy - CC BY-SA 4.0 23: The Ensemble Treatment - CC BY-SA 4.0
23.1: Ensembles of N-molecule Systems - CC BY-SA 24.12: The Gibbs Free Energy Change for Forming
4.0 HI(g) from H₂(g) and I₂(g) - CC BY-SA 4.0
23.2: The Ensemble Entropy and the Value of ß - CC 24.13: The Reference State for Molecular Partition
BY-SA 4.0 Functions - CC BY-SA 4.0
23.3: The Thermodynamic Functions of the N- 24.14: Problems - CC BY-SA 4.0
molecule System - CC BY-SA 4.0 25: Bose-Einstein and Fermi-Dirac Statistics - CC BY-SA
24: Indistinguishable Molecules - Statistical 4.0
Thermodynamics of Ideal Gases - CC BY-SA 4.0 25.1: Quantum Statistics - CC BY-SA 4.0
24.1: The Partition Function for N Distinguishable, 25.2: Fermi-Dirac Statistics and the Fermi-Dirac
Non-interacting Molecules - CC BY-SA 4.0 Distribution Function - CC BY-SA 4.0
24.2: The Partition Function for N Indistinguishable, 25.3: Bose-Einstein Statistics and the Bose-Einstein
Non-interacting Molecules - CC BY-SA 4.0 Distribution Function - CC BY-SA 4.0
24.3: Occupancy Probabilities for Translational 26: Appendices - CC BY-NC-SA 4.0
Energy Levels - CC BY-SA 4.0
26.1: Appendix A. Standard Atomic Weights 1999† -
24.4: The Separable-modes molecular Model - CC
CC BY-SA 4.0
BY-SA 4.0
26.2: Appendix B. - CC BY-SA 4.0
24.5: The Partition Function for A Gas of
26.3: Appendix B. Fundamental Constants† -
Indistinguishable, Non-interacting, Separable-modes
Undeclared
Molecules - CC BY-SA 4.0
26.4: Appendix C. - CC BY-SA 4.0
24.6: The Translational Partition Function of An Ideal
26.5: Units and Conversion Factors - Undeclared
Gas - CC BY-SA 4.0
26.6: Appendix D. - CC BY-SA 4.0
24.7: The Electronic Partition Function of an Ideal
Gas - CC BY-SA 4.0
- Undeclared
24.8: The Vibrational Partition Function of A
Back Matter - Undeclared
Diatomic Ideal Gas - CC BY-SA 4.0
24.9: The Rotational Partition Function of A Index - Undeclared
Diatomic Ideal Gas - CC BY-SA 4.0 Index - Undeclared
24.10: The Gibbs Free Energy for One Mole of An Glossary - Undeclared
Ideal Gas - CC BY-SA 4.0 Detailed Licensing - Undeclared
24.11: The Standard Gibbs Free Energy for H₂(g),
I₂(g), and HI(g) - CC BY-SA 4.0

Thermodynamics & Chemical Equilibrium (P. Ellgen)

Uploaded by

Copyright:

Available Formats

Thermodynamics & Chemical Equilibrium (P. Ellgen)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thermodynamics & Chemical Equilibrium (P. Ellgen)

Uploaded by

Copyright:

Available Formats

THERMODYNAMICS

1: Introduction - Background and a Look Ahead

3: Distributions, Probability, and Expected Values

4: The Distribution of Gas Velocities

5: Chemical Kinetics, Reaction Mechanisms, and Chemical Equilibrium

6: Equilibrium States and Reversible Processes

7: State Functions and The First Law

8: Enthalpy and Thermochemical Cycles

9: The Second Law - Entropy and Spontaneous Change

10: Some Mathematical Consequences of the Fundamental Equation

12: Applications of the Thermodynamic Criteria for Change

13: Equilibria in Reactions of Ideal Gases

14: Chemical Potential - Extending the Scope of the Fundamental

16: The Chemical Activity of the Components of a Solution

18: Quantum Mechanics and Molecular Energy Levels

19: The Distribution of Outcomes for Multiple Trials

20: Boltzmann Statistics

21: The Boltzmann Distribution Function

23: The Ensemble Treatment

24: Indistinguishable Molecules - Statistical Thermodynamics of Ideal

25: Bose-Einstein and Fermi-Dirac Statistics

1: Introduction - Background and a Look Ahead

In general, for the reaction

Figure 1. Venn diagram of a classical syllogism

Figure 3. Mapping a physical system onto a logical construct.

sufficient condition ⇒ necessary condition

Equivalently, we can write

Are propositions (a) and (b) true or false?

Figure 1. Gas volume versus pressure.

Figure 2: Gas volume versus temperature.

compound whose molecular formula is known.

electron is 1.6022 × 10 −19

96, 485 C 1 electron electrons

volume-temperature relationship, we have

The number of moles, n , cancels. Rearranging gives

Figure 3. Simplified model for velocities of gas molecules in a cubic box.

The magnitude of the momentum change for a molecule in one collision is

The magnitude of the force on the wall from one collision is

from the collision of one molecule with one wall.

Since the ideal gas equation can be written as P V = N kT we see that N mu 2

Figure 4. Atmospheric pressure versus altitude.

The weight of the gas is M g . The pressure difference is ΔP = −M g/A . We have

P (h + Δh) − P (h) ΔP −MS g −MS g

where we introduce nS as the number of moles of gas in the sample, ¯¯¯¯

Either of the latter relationships is frequently called the barometric formula.

The volume occupied by a molecule, V molecule , becomes

and volume of a hypothetical ideal gas. That is

Vreal gas = Videal gas + nb

only be determined by experiment.

The ideal-gas mixture contains n mixture = nA + nB + nC + … moles , so that

exact. For an ideal gas, the partial volume is

Since n mixture = nA + nB + nC + … , we have, for a mixture of ideal gases,

and that of gaseous water is 0.59021 kg m −−3

temperature and pressure? In L mol ? −1

you calculated in Problem 9?

the value you calculated in Problem 9?

water at this temperature and pressure? In L mol ? −1

as a percentage of the value you calculated in Problem 15?

Acetic acid 60.05 1.0491 0.10680

Acetone 58.08 0.7908 0.09940