Hemical Ngineering Inetics: Based On CHEM - ENG 408 at Northwestern University

CHEMICAL ENGINEERING KINETICS
Based on CHEM_ENG 408 at Northwestern University

REACTOR ARCHETYPES | 1
TABLE OF CONTENTS
1 Reactor Archetypes ............................................................................................................... 3
1.1 The Mass Balance ......................................................................................................................... 3
1.2 Batch Reactor ................................................................................................................................ 3
1.3 Continuous-Stirred Tank Reactor ................................................................................................. 4
1.4 Plug-Flow Reactor ........................................................................................................................ 4
2 Arrhenius Expression ............................................................................................................ 5
2.1 Arrhenius Equation ....................................................................................................................... 5
2.2 Van’t Hoff Equation ..................................................................................................................... 5
3 Mass-Action Kinetics............................................................................................................. 7
3.1.1 Overview .................................................................................................................................................7
3.1.2 First Order Kinetics .................................................................................................................................7
3.1.3 N-th Order Kinetics .................................................................................................................................7
3.1.4 Reversible Reactions................................................................................................................................8
3.1.5 Non-Integer Rate Laws ............................................................................................................................9
3.2 Determining the Rate Law and Rate Constants ............................................................................ 9
3.2.1 Differential Method .................................................................................................................................9
3.2.2 Integral Method ..................................................................................................................................... 10
3.2.3 Regression Method ................................................................................................................................ 10
3.2.4 Working with Pseudo Orders ................................................................................................................. 10
4 Working with Multiple Elementary Steps ........................................................................ 11
4.1 General Approach ....................................................................................................................... 11
4.2 Synthesis of HBr ......................................................................................................................... 11
4.2.1 Elementary Steps ................................................................................................................................... 11
4.2.2 Rate Expressions .................................................................................................................................... 11
4.2.3 The Pseudo-Steady State Hypothesis..................................................................................................... 12
4.2.4 Bond Dissociation Energies ................................................................................................................... 12
4.3 Complex Reactions: Cracking .................................................................................................... 12
4.3.1 Terminology .......................................................................................................................................... 12
4.3.2 Mechanism and Mass-Action Kinetics .................................................................................................. 13
4.3.3 Simplifications Based on Concentrations .............................................................................................. 14
4.3.4 Simplifications Based on Statistical Termination .................................................................................. 14
4.3.5 Simplifications Based on the Long-Chain Approximation .................................................................... 14
4.3.6 Ethane Cracking Rate Law .................................................................................................................... 14
4.3.7 Determining Observed Activation Energies .......................................................................................... 15
4.3.8 General Overview of Simplification Process ......................................................................................... 15
4.3.9 Complex Reactions: Additives .............................................................................................................. 16
4.4 Complex Reactions: Radical Chain Autoxidation ...................................................................... 16
5 Consequences of Chemical Equilibria ............................................................................... 18
5.1 Relationships Between Thermodynamics and Equilibrium ........................................................ 18
5.1.1 Temperature-Dependence of Thermodynamic Quantities ..................................................................... 18
5.2 The Equilibrium Constant ........................................................................................................... 18
5.2.1 Activity Equilibrium Constant ............................................................................................................... 18
5.2.2 Activities for Gases ................................................................................................................................ 19
5.2.3 Activities for Liquids ............................................................................................................................. 19
5.2.4 Relationship Between Various Equilibrium Constants .......................................................................... 20
5.3 Enzyme Kinetics ......................................................................................................................... 21
5.3.1 Derivation of the Michaelis-Menten Equation ....................................................................................... 21
5.3.2 Plotting Michaelis-Menten Data ............................................................................................................ 22
5.3.3 Reversible Product Binding ................................................................................................................... 23
5.3.4 Competitive Inhibition ........................................................................................................................... 24
5.3.5 Non-Competitive Inhibition ................................................................................................................... 24
6 Reaction Networks............................................................................................................... 25
6.1 Introduction to Reaction Networks ............................................................................................. 25

6.2 Delplots ....................................................................................................................................... 25
7 Kinetic Theory ..................................................................................................................... 27
7.1 Collision Theory ......................................................................................................................... 27
7.1.1 Distribution Laws .................................................................................................................................. 27
7.1.2 Collision Frequencies ............................................................................................................................ 27
7.1.3 Rate Constants ....................................................................................................................................... 28
7.2 Lindemann Theory ...................................................................................................................... 28
7.3 Transition State Theory............................................................................................................... 30
7.3.1 Partition Functions ................................................................................................................................. 30
7.3.2 Computing Rates of Reaction ................................................................................................................ 32
7.3.3 Thermodynamic Analysis ...................................................................................................................... 33
7.3.4 Hammett Equation ................................................................................................................................. 34
7.3.5 Taft Equation ......................................................................................................................................... 35
8 Surface Catalysis ................................................................................................................. 36
8.1 Adsorption Rate Laws................................................................................................................. 36
8.1.1 Molecular Adsorption ............................................................................................................................ 36
8.1.2 Dissociative Adsorption ......................................................................................................................... 37
8.1.3 Competitive Adsorption ......................................................................................................................... 38
8.2 Surface Reaction Rate Laws ....................................................................................................... 39
8.2.1 Single Site .............................................................................................................................................. 39
8.2.2 Dual Site ................................................................................................................................................ 39
8.2.3 Reaction with Unbound Species ............................................................................................................ 40
8.3 Desorption Rate Laws ................................................................................................................. 40
8.4 Determining the Reaction Mechanism and Rate-Limiting Step ................................................. 40
8.5 Nonidealities ............................................................................................................................... 41
8.5.1 Nonideal Surfaces .................................................................................................................................. 41
8.5.2 Sticking Probability ............................................................................................................................... 41
9 Reactions in Heterogeneous Systems ................................................................................. 43
9.1 Definitions................................................................................................................................... 43
9.1.1 Diffusivity .............................................................................................................................................. 43
9.1.2 Thiele Modulus ...................................................................................................................................... 43
9.1.3 Effectiveness Factor ............................................................................................................................... 43
9.2 Limiting Cases ............................................................................................................................ 44
9.2.1 No Diffusion Limitations ....................................................................................................................... 44
9.2.2 Diffusion Limitations ............................................................................................................................. 44
9.3 Determining if Diffusion Limitations are Dominant .................................................................. 45
9.3.1 Changing Particle Size ........................................................................................................................... 45
9.3.2 Weisz-Prater Criterion ........................................................................................................................... 46
9.4 External Mass Transfer ............................................................................................................... 46
9.4.1 Mass Transfer Coefficient ..................................................................................................................... 46
9.4.2 Mass Transfer in Reactor Engineering................................................................................................... 47
9.4.3 Nonisothermal Theory ........................................................................................................................... 47
9.4.4 Thin-Film Diffusion Reaction................................................................................................................ 47
1 REACTOR ARCHETYPES
1.1 THE MASS BALANCE
The key equation governing processes on the reactor level is the mass balance. In order to inherently account
for the proper stoichiometry, this is most typically written as a mole balance. The general mole balance for
a species 𝑖 is given as
𝑑𝑁𝑖
𝐹𝑖0 − 𝐹𝑖 + 𝐺𝑖 =
𝑑𝑡
where 𝐹𝑖0 is the input molar flow rate, 𝐹𝑖 is the output molar flow rate, 𝐺𝑖 is the generation term, and the
differential term is the accumulation.
If the system variables are uniform throughout the system volume, then
𝐺𝑖 = 𝑟𝑖 𝑉
where 𝑟𝑖 is the reaction rate of species 𝑖 and 𝑉 is the system volume. More generally speaking, if 𝑟𝑖 changes
with position in the system volume, then
𝐺𝑖 = ∫ 𝑟𝑖 𝑑𝑉
such that the mole balance can be written as

𝑑𝑁𝑖
𝐹𝑖0 − 𝐹𝑖 + ∫ 𝑟𝑖 𝑑𝑉 =
𝑑𝑡
Two other useful expressions that should be kept in mind are as follows. For a uniform concentration of
𝑖 across the system volume
𝑁𝑖 = 𝐶𝑖 𝑉
Additionally, for a given flow rate
𝐹𝑖 = 𝐶𝑖 𝑣̇
where 𝑣 is volumetric flow rate.
1.2 BATCH REACTOR

A batch reactor is a constant volume reactor has no input or output when the chemical reaction is occurring.
The batch reactor is often a good reactor archetype for slow reactions. With this information, it is clear that
the batch reactor has 𝐹𝑖0 = 𝐹𝑖 = 0. As such, the mole balance is
𝑑𝑁𝑖
= ∫ 𝑟𝑖 𝑑𝑉
𝑑𝑡
If the reaction mixture is perfectly mixed (i.e. spatially uniform) so that 𝑟𝑖 is independent of position (a
common assumption for the batch reactor), then we can state
𝑑𝑁𝑖
= 𝑟𝑖 𝑉
𝑑𝑡
Solving for the rate of reaction of species 𝑖, we see that
1 𝑑𝑁𝑖 𝑑𝐶𝑖
𝑟𝑖 = =
𝑉 𝑑𝑡 𝑑𝑡
where 𝐶𝑖 is the concentration of species 𝑖 and the expression 𝑁𝑖 = 𝐶𝑖 𝑉 was utilized.
Occasionally, batch reactors can be operated at a constant pressure but with a system volume that changes
as a function of time. In this special case,
1 𝑑𝑁𝑖 1 𝑑 𝑑𝐶𝑖 𝐶𝑖 𝑑𝑉
𝑟𝑖 = = (𝐶𝑖 𝑉(𝑡)) = +
𝑉(𝑡) 𝑑𝑡 𝑉(𝑡) 𝑑𝑡 𝑑𝑡 𝑉(𝑡) 𝑑𝑡
1.3 CONTINUOUS-STIRRED TANK REACTOR

The continuous-stirred tank reactor (CSTR) has an inlet and outlet flow of chemicals. CSTRs are operated
at steady state (such that the accumulation term is zero) and are assumed to be perfectly mixed. As such,
the mole balance for the CSTR can be written as
𝐹𝑖0 − 𝐹𝑖 + 𝑟𝑖 𝑉 = 0
Solving for the reaction rate yields
𝐹𝑖0 − 𝐹𝑖
𝑟𝑖 = −
𝑉
Utilizing the relationship of 𝐹𝑖 = 𝐶𝑖 𝑣̇ , we can then say
𝐶𝑖0 𝑣0̇ − 𝐶𝑖 𝑣̇
𝑟𝑖 = −
𝑉
Since the system is at steady state, we know that 𝑣0̇ = 𝑣̇ such that the expression becomes
𝑣̇ (𝐶𝑖0 − 𝐶𝑖 )
𝑟𝑖 = −
𝑉
Noting that the residence time is defined as
𝑉
𝜏≡
𝑣̇
we can simplify the rate expression as
𝐶𝑖0 − 𝐶𝑖
𝑟𝑖 = −
𝜏
1.4 PLUG-FLOW REACTOR
The plug-flow reactor (PFR) is a tubular reactor operated at steady state and has axial gradients but no
radial gradients. These types of reactors are useful for fast reactions that could not be as easily observed in
a batch environment. Since the concentration varies continuously down the reactor tube, so does the reaction
rate (except for zeroth order reactions).
For a PFR, the design equation can be solved by differentiating the mole balance with respect to volume,
but an easier way is to perform a mole balance on species 𝑖 in a differential segment of the reactor volume,
Δ𝑉. This differential balance can be written as
𝐹𝑖 (𝑉) − 𝐹𝑖 (𝑉 + Δ𝑉) + 𝑟𝑖 Δ𝑉 = 0
ARRHENIUS EXPRESSION | 5
since the system is in steady state. Solving for the rate, dividing by Δ𝑉, and letting it approach zero (while
applying the definition of the derivative) yields
𝑑𝐹𝑖
𝑟𝑖 =
𝑑𝑉
Once again substituting in for 𝜏 ≡ 𝑉/𝑣̇ and 𝐹𝑖 = 𝐶𝑖 𝑣̇ (assuming steady state conditions such that 𝑣̇ is
constant) yields
𝑑𝐶𝑖
𝑟𝑖 =
𝑑𝜏
Note that the design equation can be written in terms of the length of the reactor, 𝑧, and the cross-sectional
area, 𝐴𝑐 , if 𝑉 ≡ 𝐴𝑐 𝑧. Occasionally, with PFRs, the superficial velocity will be referred to, which is simply
𝑢 ≡ 𝑣̇ /𝐴𝑐 .
2 ARRHENIUS EXPRESSION
2.1 ARRHENIUS EQUATION
The main assumption behind the Arrhenius expression is that 𝑟𝑖 = 𝑓(𝑇) ⋅ 𝑓(𝐶𝑖 ). This is an approximation,
but it works quite well. The rate coefficient is the term that is a function of temperature but may also depend
on things like catalyst, solvent, and such. Empirically, the Arrhenius expression states that
𝐸
𝑘 = 𝑘0 𝑒 −𝑅𝑇
where 𝑘 is the rate coefficient, 𝑘0 is the pre-exponential factor, 𝐸 is the activation energy, 𝑇 is the
temperature, and 𝑅 is the ideal gas constant. By linearizing the equation, one finds that
𝐸 1
ln(𝑘) = − ( ) + ln(𝐴)
𝑅 𝑇
such that plotting ln(𝑘) vs. 1/𝑇 should yield a straight line of slope −𝐸/𝑅 and 𝑦-intercept of ln(𝐴).
To find the ratio of two rate coefficients and two temperatures,
𝑘2 𝐸 1 1
= exp (− ( − ))
𝑘1 𝑅 𝑇2 𝑇1
2.2 VAN’T HOFF EQUATION

The Arrhenius expression is based on the Van’t Hoff expression which states that
𝑑 ln 𝑘 Δ𝐻rxn
=
𝑑𝑇 𝑅𝑇 2
If we state that 𝑘 is the rate coefficient, replace Δ𝐻rxn with 𝐸 for the activation energy, assume that 𝐸 ≠
𝑓(𝑇) and 𝑘0 ≠ 𝑓(𝑇), and integrate we arrive at the Arrhenius equation. This will be shown a little more
rigorously below.
Assume that we have a unimolecular reaction such that Δ𝑛 = 0, such as the isomerization reaction 𝐴 ↔ 𝑅
with forward reaction rate 𝑘𝑓 , reverse reaction rate 𝑘𝑟 , and equilibrium constant 𝐾𝑐 ≡ 𝑘𝑓 /𝑘𝑟 .
ARRHENIUS EXPRESSION | 6
The Van’t Hoff equation can be written as

𝑑 ln 𝐾𝑐 Δ𝐻rxn
=
𝑑𝑇 𝑅𝑇 2
which can also be expressed as
𝑘𝑓
𝑑 ln ( ) Δ𝐻
𝑘𝑟 rxn
=
𝑑𝑇 𝑅𝑇 2
and thereby
𝑑(ln 𝑘𝑓 ) 𝑑(ln 𝑘𝑟 ) Δ𝐻rxn

− =
𝑑𝑇 𝑑𝑇 𝑅𝑇 2
Arbitrarily, we set Δ𝐻rxn = Δ𝐻𝑓 − Δ𝐻𝑟 such that the expression becomes
𝑑(ln 𝑘𝑓 ) 𝑑(ln 𝑘𝑟 ) Δ𝐻𝑓 Δ𝐻𝑟

− = −
𝑑𝑇 𝑑𝑇 𝑅𝑇 2 𝑅𝑇 2
Separating the forward and reverse components and integrating yields the Arrhenius expression in the
forward and reverse directions.
MASS-ACTION KINETICS | 7
3 MASS-ACTION KINETICS
3.1.1 OVERVIEW
We assume that rates can be described by
Rate = 𝑓(𝑇) ⋅ 𝑔(𝐶𝑖 )
where 𝑓(𝑇) is the rate constant from Arrhenius’ law and 𝑔(𝐶𝑖 ) is the rate law. Empirically, we say that
𝛽 𝛾
𝑔(𝐶𝑖 ) = 𝐶𝐴𝛼 𝐶𝐵 𝐶𝐶 …
which is typical rate law kinetics. In this expression, if 𝛼, 𝛽, 𝛾, … are the stoichiometric coefficients of a
reaction, it may be an elementary step. Typically, we find that for elementary steps ∑ 𝛼𝛽𝛾 … < 3.
3.1.2 FIRST ORDER KINETICS

We will consider the elementary reaction
𝐴→𝑃
The rate law can be given by
𝑑𝐶𝐴
− = 𝑘𝐶𝐴
𝑑𝑡
Integrating this expression yields
𝐶𝐴 𝑡
1
− ∫ 𝑑𝐶 = 𝑘 ∫ 𝑑𝑡
𝐶𝐴 𝐴
𝐶𝐴0 0
which becomes
𝐶𝐴
− ln ( ) = 𝑘𝑡
𝐶𝐴0
and thereby
𝐶𝐴 = 𝐶𝐴0 𝑒 −𝑘𝑡
and a plot of ln(𝐶𝐴 ) vs. 𝑡 should be linear for first order kinetics. The half-life can be given by
ln(2)
𝑡1 =
2 𝑘
3.1.3 N-TH ORDER KINETICS

The above process can be done for any integer 𝑛. We will consider the general reaction
𝑛𝐴 → 𝑃
𝑑𝐶𝐴
− = 𝑘𝐶𝐴𝑛
𝑑𝑡
Integrating this expression yields
𝐶𝐴 𝑡
1
− ∫ 𝑛 𝑑𝐶𝐴 = 𝑘 ∫ 𝑑𝑡
𝐶𝐴
𝐶𝐴0 0
which becomes
𝐶𝐴−𝑛+1 −𝑛+1
𝐶𝐴0
− = 𝑘𝑡
−𝑛 + 1 −𝑛 + 1
For 𝑛 ≠ 1 and thereby
1 1
= + (𝑛 − 1)𝑘𝑡
𝐶𝐴𝑛−1 𝑛−1
𝐶𝐴0
and a plot of 1/𝐶𝐴𝑛−1 vs. 𝑡 should be linear. The half-life can be given by
2𝑛−1 − 1
𝑡1 = 𝑛−1
2 (𝑛 − 1)𝑘𝐶𝐴0
The units of the rate constant for an 𝑛th order reaction can be given by 𝑘[=]concentration1−𝑛 s −1.
3.1.4 REVERSIBLE REACTIONS

We will consider the reaction
𝐴↔𝑃
𝑑𝐶𝐴
− = 𝑘𝑓 𝐶𝐴 − 𝑘𝑟 𝐶𝑃
𝑑𝑡
where 𝑘𝑓 and 𝑘𝑟 represent the rate constants of the forward and reverse reactions, respectively. For the case
of no initial concentration of 𝑃, one can state that
𝐶𝐴0 − 𝐶𝐴 = 𝐶𝑃
Therefore
𝑑𝐶𝐴
− = 𝑘𝑓 𝐶𝐴 − 𝑘𝑟 (𝐶𝐴0 − 𝐶𝐴 )
𝑑𝑡
This can be rearranged to

𝑑𝐶𝐴 𝑘𝑟
− = (𝑘𝑓 + 𝑘𝑟 ) (𝐶𝐴 − 𝐶 )
𝑑𝑡 𝑘𝑓 + 𝑘𝑟 𝐴0
To simplify the algebra

𝑘𝑓 𝐶𝑃,𝑒𝑞 𝐶𝐴0 − 𝐶𝐴,𝑒𝑞
𝑘𝑒𝑞 ≡ = =
𝑘𝑟 𝐶𝐴,𝑒𝑞 𝐶𝐴,𝑒𝑞
This implies that

𝑘𝑟
𝐶𝐴,𝑒𝑞 = 𝐶
𝑘𝑓 + 𝑘𝑟 𝐴0
Therefore,
𝑑𝐶𝐴
− = (𝑘𝑓 + 𝑘𝑟 )(𝐶𝐴 − 𝐶𝐴,𝑒𝑞 )
𝑑𝑡
which can be integrated to
𝐶𝐴 𝑡
𝑑𝐶𝐴
− ∫ = (𝑘𝑓 + 𝑘𝑟 ) ∫ 𝑑𝑡
𝐶𝐴 − 𝐶𝐴,𝑒𝑞
𝐶𝐴0 0
which is equal to
𝐶𝐴0 − 𝐶𝐴,𝑒𝑞
ln ( ) = (𝑘𝑓 + 𝑘𝑟 )𝑡
𝐶𝐴 − 𝐶𝐴,𝑒𝑞
3.1.5 NON-INTEGER RATE LAWS

The previous procedures can be used to find non-integer rate laws. Generally speaking, if the reaction order
is an integer or half-integer, then it is possible that it tells us something physical about the system and the
reaction step is elementary. However, if it is not the case, we can be rather certain that this is just a
generalization of the system and doesn’t mean anything chemically significant about the reaction
mechanism. It should also be noted that for a rate law of 1/2 order, it is likely a multistep process (such as
the production of radicals) since 𝐴 → 2𝐵 is not typically reasonable. We should also note that the opposite
of a second order reaction is a half order reaction, so if the reaction is reversible this is where a half order
can come into play as well.
3.2 DETERMINING THE RATE LAW AND RATE CONSTANTS

3.2.1 DIFFERENTIAL METHOD
The differential method says that we should propose a rate law of the type
𝑑𝐶𝐴
− = 𝑘𝐶𝐴𝛼
𝑑𝑡
We then can rewrite this as
𝑑𝐶𝐴
ln (− ) = ln 𝑘 + 𝛼 ln 𝐶𝐴
𝑑𝑡
𝑑𝐶𝐴
which can be used to fit a function to find the slope and order 𝛼. It is also possible to plot ln (− | )
𝑑𝑡 𝑡=0
vs. ln(𝐶𝐴,0 ) for many initial concentrations to back out the rate constant and order. Of course, any
conclusion from either method depends on the fit and is only as good as the data. Oftentimes, we are only
given the concentration as a function of time, not the rate of the reaction. There are three main ways to get
the rate from the dataset.
The first method is the graphical method, wherein a smooth curve is drawn through the experimental data
points on a 𝐶𝐴 vs. 𝑡 graph. At each time instant of interest, tangents are drawn to this curve, the slope of the
tangent line is the derivative value at that time instant and is therefore the rate.
The second method is the polynomial method, wherein a polynomial of suitable order is fitted to the data.
The derivative can then be evaluated by differentiating the polynomial expression (typically, the simplest
monotonic function is the best choice).
The third method is the finite difference method, wherein we can approximation the derivative via a
numerical finite difference equation. This, however, requires a sufficient data size to do accurately.
3.2.2 INTEGRAL METHOD

The integral method simply involves deriving the rate laws for a hypothetical reaction order and seeing if
the data fits the linearized equations. This is essentially just a guess and check method.
3.2.3 REGRESSION METHOD

If we assume a rate law to the 𝑛th power, we can use the general integrated rate law of
1 1
= + (𝑛 − 1)𝑘𝑡
𝐶𝐴𝑛−1 𝑛−1
𝐶𝐴0
to solve for 𝑛. This can be done through a non-linear least-squares analysis (a typical fitting toolbox would
suffice).
3.2.4 WORKING WITH PSEUDO ORDERS

Sometimes it’s easier to assume an order and use pseudo rate constants. For instance, if we assume a rate
law of the form
rate = 𝑘𝐶𝐴𝑛
We can arbitrarily assume pseudo first order to make it
rate = 𝑘1st 𝐶𝐴
where
𝑘1st ≡ 𝑘rxn 𝐶𝐴𝑛−1

We can linearize this to get
ln(𝑘1𝑠𝑡 ) = ln(𝑘rxn ) + (𝑛 − 1) ln(𝐶𝐴 )
so that the slope of a ln(𝑘1𝑠𝑡 ) vs. ln(𝐶𝐴 ) curve would be equal to 𝑛 − 1.
WORKING WITH MULTIPLE ELEMENTARY STEPS | 11
4 WORKING WITH MULTIPLE ELEMENTARY STEPS

4.1 GENERAL APPROACH
When dealing with a reaction that has multiple elementary steps, the following procedure should be used
in general:
1) Postulate elementary steps (in particular, bond making and bond breaking events)
2) Postulate a mechanism (determining fast vs. slow reactions as well as reversible vs. equilibrium
reactions)
3) Derive a closed-form rate law that describes the rate as a function of observables
4) Test the rate law against available data and refine as necessary
4.2 SYNTHESIS OF HBr

4.2.1 ELEMENTARY STEPS
Consider the reaction
H2 + Br2 → HBr
We will see that the rate expression for this reaction is not as simple as one would initially anticipate. For
HBr we can write the following possible elementary steps:
Initiation – these are reactions that turn the reactants into radical species
1. Br2 → Br· + Br· (denoted 𝑘1 )
2. H2 → H· + H· (denoted 𝑘4 )
Degenerate reactions – these reactions produce the same products as reactants and are important if
isotopically labeled for determining kinetically relevant steps
1. Br· + Br2 → Br2 + Br·
2. H· + H2 → H2 + H·
Propagation – these are reactions that produce at least as many radicals as are consumed to propagate the
chain reaction
1. Br· + H2 → HBr + H· (denoted 𝑘2 )
2. H· + Br2 → HBr + Br· (denoted 𝑘3 )
3. HBr + H· → Br· + H2 (denoted 𝑘−2)
4. HBr + Br· → H· + Br2 (denote 𝑘−3)
Termination – these are reactions that convert radical species into stable species
1. Br· + H· → HBr (denoted 𝑘5 )
2. H· + H· → H2 (denoted 𝑘−4)
3. Br· + Br· → Br2 (denoted 𝑘−1)
4.2.2 RATE EXPRESSIONS

With the reactions listed, we now write out the mass action kinetics.
The overall balance for Br2 is
𝑑[Br2 ] 1
𝑟Br2 = = −𝑘1 [Br2 ] + 𝑘−1 [Br ∙]2 − 𝑘3 [H ∙][Br2 ] + 𝑘3 [HBr][Br ∙]
dt 2
The overall balance for HBr is

𝑑[HBr]
𝑟HBr = = 𝑘2 [Br ∙][H2 ] + 𝑘3 [H ∙][Br2 ] − 𝑘−2 [HBr][H ∙] − 𝑘−3 [HBr][Br ∙] + 𝑘5 [Br ∙][H ∙]
dt
The overall balance for Br· is
𝑑[Br ∙]
𝑟Br. = = 2𝑘1 [Br2 ] − 𝑘2 [Br ∙][H2 ] + 𝑘3 [H ∙][Br2 ] + 𝑘−2 [HBr][H ∙] − 𝑘−3 [HBr][Br ∙]
dt
− 𝑘5 [Br ∙][H ∙] − 2𝑘−1 [Br ∙]2
And we can do an analogous thing for the other species, but you get the idea.
4.2.3 THE PSEUDO-STEADY STATE HYPOTHESIS

Once the rate expressions are written, they can either be solved simultaneously without any simplification
using a standard computational ODE-solving package, or algebraic simplifications can be made to make
the expression more manageable. The most common assumption to make is the pseudo-steady-state
hypothesis (PSSH), which states that the rate of appearance of a given species is essentially 0. This is most
commonly applied to intermediate radical species, in this case Br· and H·. It is also important to check that
for all chain reactions, the rate of initiation should equal the rate of termination if PSSH is applied.
4.2.4 BOND DISSOCIATION ENERGIES

Another simplification that can be made is neglecting small terms. To make this decision, it is easiest to
look at bond dissociation energies (BDE) to predict the size of rate constants and relative concentrations.
In this reaction, the BDE for H2 is 104.2 kcal/mol, the BDE for HBr is 87.5 kcal/mol, and the BDE for Br2
is 46 kcal/mol. We then note that the BDE is proportional to the activation energy via the Evans-Polanyi
relationship. The Evans-Polanyi relationship states that
𝐸𝑎 = 𝐸0 + 𝛼Δ𝐻rxn
where 𝐸0 is a constant for a given family of reactions and 𝛼 is the position of the transition state along the
reaction coordinate (from 0 to 1). Also, it should be noted that 𝛼𝑓 = 1 − 𝛼𝑟 .
Therefore, based on the BDEs listed, we can assume that 𝑘1 ≫ 𝑘4 due to the low BDE of Br2 compared to
H2. This BDE analysis also means that [Br ∙] ≫ [H ∙]. As such, we will remove the 𝑘4 term. We also note
that since there is so little [H ∙] in the system, 𝑟−4 ≈ 0. We will also consider reaction 3 only in the forward
direction because of the relatively high BDE of HBr. Applying the PSSH and including these assumptions
results in the following solution
1
𝑘[H2 ][Br2 ]2
𝑟=
[HBr]
𝑘′ +
[Br2 ]
At low conversion (i.e. [HBr] ≈ 0), we see that the equation becomes first order in H2 and half order in Br2.
Looking at limiting cases like this is helpful in accurately describing the kinetics.
4.3 COMPLEX REACTIONS: CRACKING

4.3.1 TERMINOLOGY
Before getting into the mechanism for ethane cracking, it is important to go over some terminology used in
this field of research. The two most common propagation steps in hydrocarbon pyrolysis and cracking are
hydrogen abstraction and 𝛽-scission.
Hydrogen abstraction takes the general form of X· + H-Y → X-H + Y·. The radical that is produced from
a hydrogen abstraction process is called a 𝜇 species. It is literally the abstraction of a hydrogen from an
otherwise stable substrate.
𝛽-scission takes the following general form:
The radical that is produced from a 𝛽-scission reaction is called a 𝛽 species.
4.3.2 MECHANISM AND MASS-ACTION KINETICS

In this example, we will consider ethane cracking:
C2 H6 → C2 H4 + H2
Writing out the possible elementary steps:
Initiation – in cracking, initiation occurs with a single molecule breaking down into two identical free
radicals
𝛼
1) C2H6 → 2 CH3· (𝐴 → 2 𝛽2 )
(note that CH3CH3 → C2H5· + H· has a significantly higher BDE)
Propagation – The two main reactions are H-abstraction and 𝛽-scission
𝑘21
1) CH3· + C2H6 → C2H5· + CH4 (𝛽2 + 𝐴 → 𝜇 + CH4 ) (H-abstraction)
𝑘1
2) C2H5· → C2H4 + H· (𝜇 → C2H4 + 𝛽1 ) (𝛽-scission)
𝑘11
3) H· + C2H6 → H2 + C2H5· (𝛽1 + 𝐴 → H2 + 𝜇) (H-abstraction)
Termination
𝑘𝑡
1) H· + C2H5· → C2H6 (𝛽1 + 𝜇 → 𝐴)
𝑘𝑡′
2) 2 H· → H2 (2 𝛽1 → H2)
𝑘𝑡′′
3) 2 C2H5· → C4H10 (2 𝜇 → C4H10)
𝑘𝑡′′′
4) 2 CH3· → C2H6 (2 𝛽2 → 𝐴)
𝑘𝑡𝐼𝑉
5) CH3· + H· → CH4 (𝛽2 + 𝛽1 → CH4)
𝑘𝑡𝑉
6) C2H5· + CH3· → C3H8 (𝜇 + 𝛽2 → propane)
A couple of things should be noted before continuing. Of course, other reactions can be written but they are
not especially likely to occur. Also, it should be noted that the methyl C-H bond is one of the strongest
aliphatic C-H bonds and therefore CH4 does not tend to form CH3· radicals very frequently. Finally, the
propagation steps are conventionally written in the spontaneous (i.e. exothermic) direction in the literature.
Writing out the mass action expressions:

𝑑𝐴 1
𝑟𝐴 = = −𝛼𝐴 − 𝑘21 𝛽2 𝐴 − 𝑘11 𝛽1 𝐴 + 𝑘𝑡 𝛽1 𝜇 + 𝑘𝑡′′′ 𝛽22 + ⋯
𝑑𝑡 2
𝑑𝛽1
𝑟𝛽1 = = 𝑘1 𝜇 − 𝑘11 𝛽1 𝐴 − 𝑘𝑡 𝛽1 𝜇 − 2𝑘𝑡′ 𝛽12 − 𝑘𝑡𝐼𝑉 𝛽1 𝛽2 + ⋯
𝑑𝑡
𝑑𝛽2
𝑟𝛽2 = = 2𝛼𝐴 − 𝑘21 𝛽2 𝐴 − 2𝑘𝑡′′′ 𝛽22 − 𝑘𝑡𝐼𝑉 𝛽1 𝛽2 + ⋯
𝑑𝑡
𝑑𝜇
𝑟𝜇 = = 𝑘21 𝐴𝛽2 − 𝑘1 𝜇 + 𝑘11 𝛽1 𝐴 − 𝑘𝑡 𝛽1 𝜇 − 𝑘𝑡𝑉 𝜇 − 2𝑘𝑡′′ 𝜇2 + ⋯
𝑑𝑡
4.3.3 SIMPLIFICATIONS BASED ON CONCENTRATIONS
These expressions can be simplified by employing the PSSH on the radical species 𝛽1 , 𝛽2 , and 𝜇. We can
also trim the expressions based on low rate coefficients and/or concentrations. The only low rate coefficient
is in the initiation step, but you cannot delete the initiation term since then the reaction will never occur.
However, we will remove any reactions of 𝛽2 with another radical based on experimental data that shows
that no significant amounts of CH4, butane, propane, or other larger hydrocarbons are observed because
CH4 is so stable that it is unlikely it will dissociate.
4.3.4 SIMPLIFICATIONS BASED ON STATISTICAL TERMINATION

Additional simplifications can be made based on the assumption of statistical termination. Take the 𝛽1 and
𝜇 radicals as an example. A table can be made as follows showing the rate coefficients for each possible
combination:
𝛽1 𝜇
𝛽1 𝑘𝑡 ′ 𝑘𝑡
𝜇 𝑘𝑡 𝑘𝑡 ′′
If we assume statistical termination, then the 𝑘𝑡 reaction should occur twice as likely as the 𝑘𝑡 ′ and 𝑘𝑡 ′′
reaction. As such, if we let 𝜔 be the rate constant for termination, we can state that
1
𝜔 = 𝑘𝑡′ = 𝑘𝑡′′ = 𝑘𝑡
2
4.3.5 SIMPLIFICATIONS BASED ON THE LONG-CHAIN APPROXIMATION
The long-chain approximation can be made when the rate of propagation is significantly faster than the rate
of initiation and/or termination.
4.3.6 ETHANE CRACKING RATE LAW

Combining all of the previous assumptions allows us to arrive at the rate law for ethane cracking:
1
𝛼𝐴 2
𝑘11 ( 𝜔 ) 𝐴
𝑟𝐴 =
𝑘 𝐴
1 + 11
𝑘1
This rate law shows that at low partial pressure of 𝐴, the reaction is 3/2 order in 𝐴 whereas at high partial
pressure of 𝐴 the reaction is 1/2 order in 𝐴.
In order to better understand what is going on in this mechanism, a diagram can be drawn like the one
below. We see that there is a radical ratcheting cycle between the ethyl radical and hydrogen radical. This
controls most of the kinetics of the process (hence the presence of 𝑘1 and 𝑘11).
4.3.7 DETERMINING OBSERVED ACTIVATION ENERGIES

At low partial pressure of 𝐴 we have
1 3
𝑘11 𝛼 2 𝐴2
𝑟𝐴 = 1
𝜔2
To calculate the overall or observed activation energy, we can rewrite the expression as
1
𝐸11 𝐸𝛼 2
(𝐴11 exp (− 𝑅𝑇 )) (𝐴𝛼 exp (− 𝑅𝑇 ))
1
𝐸𝜔 2
(𝐴𝜔 exp (− 𝑅𝑇 ))
This shows us that the overall activation energy for this reaction is
1 1
𝐸𝑜𝑏𝑠 = 𝐸11 + 𝐸𝛼 − 𝐸𝜔
2 2
This approach can be used anytime rate coefficients are multiplied and/or divided together. If rate
coefficients are added or subtracted, the only way to find an observed activation energy is to break it into
different regimes where the addition/subtraction disappears.
4.3.8 GENERAL OVERVIEW OF SIMPLIFICATION PROCESS

1. Postulate steps
2. Write out mass-action kinetics
3. Apply PSSH on intermediate radicals
a. This, in turn, sets the rate of initiation equal to the rate of termination
4. Use BDEs and/or experimental data about which species are not observed to further reduce the
mechanism
5. If applicable, apply the long-chain approximation such that the rate of propagation is much greater
than the rate of initiation and termination
6. The goal is to write the rate in change of reactant (or product) as a function of only experimentally
observable (i.e. stable) species and the corresponding rate coefficients
4.3.9 COMPLEX REACTIONS: ADDITIVES

Reactions can be controlled not only by changing pressure and temperature but also by introducing
additives. These may include things such as initiators for polymerization, reactive species for the stable
operation of combustion engines, anti-oxidants (to prevent oxidation), vitamins (i.e. radical cofactors), and
so on. In these cases, the ratio that the rate has change is called the enhancement factor. A value greater
than 1 means the reaction is enhanced whereas a value less than 1 means the reaction is impeded.
One common class of additives is radical initiators. These are species that can produce radical species under
mild conditions and therefore promote radical reactions. They generally possess weak bonds (i.e. bonds
with a small BDE). Organic peroxides (e.g. benzoyl peroxide) are therefore oftentimes used as radical
initiators since the BDE of the O-O bond in the peroxides is very weak, and the resulting oxyl radicals are
unstable and preferentially abstract hydrogen.
4.4 COMPLEX REACTIONS: RADICAL CHAIN AUTOXIDATION

Assume we have the following autoxidation mechanism:
𝑘𝑑
ROOH → RO· + HO·
𝑘1
RO· + RH → ROH + R·
𝑘2
OH· + RH → H2O + R·
𝑘3
R· + O2 → ROO·
𝑘4
ROO· + RH → ROOH + R·
𝑘𝑡
ROO· + ROO· → ROH + R=O + O2
Mass action kinetics and PSSH
𝑑[RH]
= −𝑘1 [RO ∙][RH] − 𝑘2 [OH ∙][RH] + 𝑘4 [ROO ∙][RH]
𝑑𝑡
𝑑[RO ∙]
= 0 = −𝑘1 [RO ∙][RH] + 𝑘𝑑 [ROOH]
𝑑𝑡
𝑑[OH ∙]
= 0 = −𝑘2 [OH ∙][RH] + 𝑘𝑑 [ROOH]
𝑑𝑡
𝑑[R ∙]
= 0 = 𝑘1 [RO ∙][RH] + 𝑘2 [OH][RH] − 𝑘3 [R ∙][O2 ] + 𝑘4 [ROO ∙][RH]
𝑑𝑡
𝑑[ROO ∙]
= 0 = 𝑘3 [R ∙][O2 ] − 𝑘4 [ROO ∙][RH] − 2𝑘𝑡 [ROO ∙]2
𝑑𝑡
Add together 𝑑[R ∙]/𝑑𝑡 and 𝑑[ROO ∙]/𝑑𝑡 to get
0 = 𝑘1 [RO ∙][RH] + 𝑘2 [OH][RH] − 2𝑘𝑡 [ROO ∙]2

We note that from 𝑑[RO ∙]/𝑑𝑡
𝑘1 [RO ∙][RH] = 𝑘𝑑 [ROOH]
We note that from 𝑑[OH ∙]/𝑑𝑡

𝑘2 [OH ∙][RH] = 𝑘𝑑 [ROOH]
so
0 = 2𝑘𝑑 [ROOH] − 2𝑘𝑡 [ROO ∙]2

This shows that the rate of initiation equals the rate of termination, as it should from the PSSH. We can
then use this to rewrite the rate in change of RH as
1
𝑑[RH] 𝑘𝑑 2
= 2𝑘𝑑 [ROOH] + 𝑘4 [RH] ( [ROOH])
𝑑𝑡 𝑘𝑡
We can combine rate coefficients to rewrite this as
𝑑[RH] 1
= 2𝑘𝑑 [ROOH] + 𝑘 ′ [RH][ROOH]2
𝑑𝑡
If we know in advance that we can apply the long-chain approximation, then the left-hand term becomes
negligible. To make an educated decision, we should look at BDEs. They are as follows:
ROO-H, 366 kJ/mol | RO-H, 441 kJ/mol | HO-H, 497 kJ/mol | R-H, 416 kJ/mol
We can now calculate the Δ𝐻rxn associated with reaction 4.
Δ𝐻rxn,4 = BDEbonds broken − BDEbonds formed = BDERH − BDEROOH ≈ 50 kJ/mol
From this, we can see that 𝑘4 is likely a small quantity, and so the long-chain approximation probably isn’t
best to use here.
If we want to find the selectivity of ROH versus R=O, we note that ROH is produced in both reaction 1 and
the termination step while R=O only is produced in the termination step. The selectivity is then
𝑑[ROH]
𝑑𝑡 step 1 + step 𝑡 𝑘1 [RO ∙][RH] + 𝑘𝑡 [ROO ∙]2 𝑘𝑑 [ROOH]
= = = +1=2
𝑑[R = O] step t 𝑘𝑡 [ROO ∙]2 𝑘𝑡 [ROO ∙]2
𝑑𝑡
CONSEQUENCES OF CHEMICAL EQUILIBRIA | 18
5 CONSEQUENCES OF CHEMICAL EQUILIBRIA

5.1 RELATIONSHIPS BETWEEN THERMODYNAMICS AND EQUILIBRIUM
5.1.1 TEMPERATURE-DEPENDENCE OF THERMODYNAMIC QUANTITIES
∘ ∘
In this section, we will make great use of Δ𝐺rxn and Δ𝐻rxn . The superscript ∘ indicates a standard reference
state, most typically 1 atm in this class (or 1 bar in many tables) when dealing with gases. The standard
state does not refer to a temperature, although most tabulated values are at 298.15 K. Since both Gibbs free
energy and enthalpy are functions of temperature, this can complicate equilibrium calculations because
tabulated quantities are most often reported at temperatures other than the reaction conditions (and
oftentimes the temperature of the system changes throughout the course of a given reaction).
∘
To find the Δ𝐻rxn at a particular temperature 𝑇, one can use the following relationship:
𝑇
∘
∘ (𝑇)
Δ𝐻rxn = Δ𝐻rxn,𝑇𝑟𝑒𝑓
+ ∫ Δ𝐶𝑃∘ (𝑇) 𝑑𝑇
𝑇𝑟𝑒𝑓
∘
The Δ𝐻rxn,𝑇𝑟𝑒𝑓
term refers to the standard enthalpy of reaction at a tabulated reference temperature. This is
typically known from the heats of formation of each species at the same reference temperature (Δ𝐻𝑓∘𝑖,𝑇𝑟𝑒𝑓 ).
The relationship is
∘
Δ𝐻rxn,𝑇𝑟𝑒𝑓
= ∑ 𝜈𝑖 Δ𝐻𝑓∘𝑖,𝑇𝑟𝑒𝑓
𝑖
For species in the most stable state (e.g. monatomic carbon or diatomic oxygen), the enthalpies of formation
are assigned to be zero.
The Δ𝐶𝑃∘ term is change heat capacity term weighted by stoichiometric ratios and is also a function of
temperature. This is
∘
Δ𝐶𝑃∘ (𝑇) = ∑ 𝜈𝑖 𝐶𝑃,𝑖
𝑖
∘
The value of the heat capacity for a given species (𝐶𝑃,𝑖 ) can be found in reference tables as well. It is often
reported as a power series of temperature with empirically derived coefficients.
An analogous expression can be written for the standard Gibbs free energy change of reaction at a reference
temperature:
∘
Δ𝐺rxn,𝑇𝑟𝑒𝑓
= ∑ 𝜈𝑖 Δ𝐺𝑓∘𝑖,𝑇𝑟𝑒𝑓
𝑖
5.2 THE EQUILIBRIUM CONSTANT

5.2.1 ACTIVITY EQUILIBRIUM CONSTANT
The equilibrium constant, using activities, is defined as
𝜈
𝐾𝑎 = ∏ 𝑎𝑖 𝑖
𝑖
where 𝑎𝑖 is the activity of a given species.

The activity equilibrium constant can be related to Δ𝐺 by

∘
Δ𝐺rxn
𝐾𝑎 = 𝑒 − 𝑅𝑇
∘ ∘
However, it is critical to note that since Δ𝐺rxn are typically found from Δ𝐺𝑓,𝑖 at a reference temperature,
𝑇ref, then the resulting 𝐾𝑎 is only valid at 𝑇ref and is
∘
𝛥𝐺rxn,𝑇𝑟𝑒𝑓
𝐾𝑎,𝑇𝑟𝑒𝑓 = exp (− )
𝑅𝑇𝑟𝑒𝑓
In order to determine the 𝐾𝑎 at a temperature other than the reference temperature, one must use the
following relationship
𝑇2
𝐾𝑎,𝑇2 ∘ (𝑇)
Δ𝐻rxn
ln ( )= ∫ 𝑑𝑇
𝐾𝑎,𝑇1 𝑅𝑇 2
𝑇1
∘
where 𝑇1 typically is 𝑇ref. Since Δ𝐻rxn is a function of temperature, this must usually be accounted for
using the aforementioned temperature-dependence relationships. However, if this temperature effect can be
ignored, then the equation simplifies to
𝐾𝑎,𝑇2 Δ𝐻∘rxn 1 1
ln ( )=− ( − )
𝐾𝑎,𝑇1 𝑅 𝑇2 𝑇1
5.2.2 ACTIVITIES FOR GASES

The most general way to compute the activity of an arbitrary mixture of gases is by using fugacities:
𝑓𝑖 𝑃𝑖 𝑃
𝑎𝑖 = = 𝛾𝑖 = 𝜑𝑖 𝑦𝑖
𝑃ref 𝑃ref 𝑃ref
where 𝜑𝑖 is the dimensionless fugacity coefficient and 𝛾𝑖 is the activity coefficient. For ideal species, 𝜑𝑖 =
𝛾𝑖 = 1. Therefore, for a gas mix at low enough pressure that the system can be assumed to be ideal, we can
simplify this expression by saying
𝑃
𝑎𝑖 = 𝑦𝑖
𝑃ref
where 𝑃 is the total pressure of the system.1 This then means that
𝑃 𝛿 𝜈
𝐾𝑎 = ( ) ∏ 𝑦𝑖 𝑖
𝑃ref
𝑖
where 𝛿 is the change in (stoichiometric) moles.
5.2.3 ACTIVITIES FOR LIQUIDS

Activities for liquids can be written as
𝑎𝑖 = 𝛾𝑖 𝑥𝑖
1
Since partial pressure is just 𝑃𝑖 = 𝑃𝑦𝑖 , we can optionally rewrite activity as 𝑎𝑖 = 𝑃𝑖 /𝑃ref
where 𝛾𝑖 is specifically the Raoult’s law activity coefficient and indicates ideal behavior according to
Raoult’s law. As 𝑥𝑖 → 1, the value of the activity coefficient also approaches 1. As 𝑥𝑖 → 0, the activity
coefficient approaches an infinitely dilute case denoted 𝛾𝑖∞ and can be found by divided the Henry’s law
constant by the saturation vapor pressure. The value of 𝛾𝑖∞ is large for poorly soluble, low volatility species.
∞ ∞
If 𝛾product > 𝛾reactants then there is an equilibrium shift toward products.
Dilute solutions typically follow Henry’s law instead of Raoult’s law, and it is more conventional to use
concentration. As such,
𝐶𝑖
𝑎𝑖 = 𝛾𝑖
𝐶ref
where 𝐶𝑖 refers to the concentration of a species 𝑖, 𝐶ref is typically 1 M. For an ideal liquid mixture, the
activity coefficient approaches 1 and
𝐶𝑖
𝑎𝑖 =
𝐶ref
Therefore, for ideal solutions
1 𝛿 𝜈
𝐾𝑎 = ( ) ∏ 𝐶𝑖 𝑖
𝐶ref
𝑖
5.2.4 RELATIONSHIP BETWEEN VARIOUS EQUILIBRIUM CONSTANTS

Oftentimes, equilibrium constants in terms of pressure, concentration, and mole fraction are used. These
are as follows, respectively:
𝜈𝑖
𝐾𝑃 = ∏ 𝑃𝑖
𝑖
𝜈
𝐾𝐶 = ∏ 𝐶𝑖 𝑖
𝑖
𝜈𝑖
𝐾𝑦 = ∏ 𝑦𝑖
𝑖
For an ideal gas, the following relationships hold:
1 𝛿 𝑃 𝛿
𝐾𝑎 = ( ) 𝐾𝑃 = ( ) 𝐾𝑦
𝑃ref 𝑃ref
𝐾𝑃 = 𝐾𝐶 (𝑅𝑇)𝛿
For an ideal liquid mixture, we have that
1 𝛿
𝐾𝑎 = ( ) 𝐾𝐶
𝐶ref
We therefore see that for ideal solutions, 𝐾𝑎 ≈ 𝐾𝑃 ≈ 𝐾𝐶 (taking into account reference terms to take care
of units).
5.3 ENZYME KINETICS

5.3.1 DERIVATION OF THE MICHAELIS-MENTEN EQUATION
Catalysts are defined as agents that decrease the activation energy of a reaction without being consumed.
Therefore, at a given temperature, the reaction rate increases in the presence of a catalyst.
We will use the following nomenclature when discussing both physical and biological catalysts: E (free
catalyst), A (reactant), P (product), and all other lets (e.g. X, Y) are catalytic intermediates (e.g. an enzyme-
substrate complex).
Let us consider the reaction mechanism A + E ⟷ X → P + E, which I will enumerate as follows
𝑘1 𝑘2
A+E→X X→A+E
𝑘3
X→E+P
The goal is to find the rate of appearance of P in terms of observable quantities and constants. If we apply
PSSH to our intermediate X, we get
𝑑X
= 0 = 𝑘1 [𝐴][𝐸] − 𝑘2 [X] − 𝑘3 [X]
𝑑𝑡
Applying a mass (i.e. site) balance to the catalyst yields
𝐸𝑇 = [E] + [X]
where 𝐸𝑇 is the total amount of catalyst. It is ideal to write [X] in terms of 𝐸𝑇 instead of [E] because 𝐸𝑇 is
a constant with time. Therefore,
𝑘1 [𝐴]𝐸𝑇
[X] =
𝑘1 [A] + 𝑘2 + 𝑘3
We can write the rate in change of product as
𝑑[P] 𝑘1 𝑘3 [A]𝐸𝑇
= 𝑘3 [X] =
𝑑𝑡 𝑘1 [A] + 𝑘2 + 𝑘3
Dividing by 𝑘1 yields
𝑑[P] 𝑘3 [A]𝐸𝑇
=
𝑑𝑡 𝑘 + 𝑘3
[A] + 2
𝑘1
We say that 𝑉max ≡ 𝑘3 𝐸𝑇 and 𝐾M ≡ (𝑘2 + 𝑘3 )/𝑘1 to get the typical form of the Michaelis-Menten equation
𝑑[P] 𝑉max [A]
=
𝑑𝑡 [A] + 𝐾M
The value of 𝐾M is the concentration of A at which the rate is half of its maximum. The value of 𝑉max is the
maximum rate at [A] ≫ 𝐾M . From this equation, it is clear that at [A] ≪ 𝐾M , the rate appears to be first
order in A. At [A] ≫ K 𝑀 , the rate is pseudo zeroth order in A (and therefore independent of A). At all [A],
the rate is proportional to the total enzyme concentration.
5.3.2 PLOTTING MICHAELIS-MENTEN DATA

Oftentimes, a linearized plot is desired. This can be done using a Lineweaver-Burk plot, where the inverse
of the rate is plotted against the inverse concentration of the substrate. This is merely a linearization of the
Michaelis-Menten equation to
1 𝐾M 1 1
= +
𝑟𝑃 𝑉max [A] 𝑉max
This then leads to the following linearized for (here, the substrate is denoted S)
The Eadie-Hofstee plot is another way to linearize the Michaelis-Menten equation, which uses the following
equation
𝑟P
𝑟P = −𝐾M + 𝑉max
[𝐴]
and has a plot with characteristics shown below
Since the Lineweaver-Burke plot uses inverse values, it unevenly weights data points in concentration and
reaction rate whereas Eadie-Hofstee plots do not. However, the Eadie-Hofstee plot’s abscissa and ordinate
are dependent on the reaction rate, and therefore any experimental error will be present in both axes. Any
uncertainty also propagates unevenly and becomes larger over the abscissa, giving weight to smaller values
of 𝑟P /[A].
In a batch reactor, one often wants to plot [A] as a function of time. To obtain this, the Michaelis-Menten
equation must be integrated. Doing so yields
𝐶𝐴0
𝐾M ln ( ) + (𝐶𝐴0 − 𝐶𝐴 ) = 𝑘3 𝐸T 𝑡
𝐶𝐴
Plotting concentration as a function of time yields a function that clearly changes slope as it switches from
first order to zeroth order regimes
The equation can also be linearized in the following form
𝐶𝐴0 − 𝐶𝐴 𝑡
= 𝑉max ( ) − 𝐾M
𝐶 𝐶
ln ( 𝐶𝐴0 ) ln ( 𝐶𝐴0 )
𝐴 𝐴
5.3.3 REVERSIBLE PRODUCT BINDING

Instead of irreversible product binding, we will now consider reversible product binding given by A + E
⟷ X ⟷ P + E, which I will enumerate as follows
𝑘1 𝑘2
A+E→X X→A+E
𝑘3 𝑘4
X→P+E P+E→X
Applying PSSH on species X yields
0 = (𝑘1 [A] + 𝑘4 [P])[𝐸] − (𝑘2 + 𝑘3 )[X] = 0
The site balance can be given by
𝐸𝑇 = [X] + [E]
Substituting in for [E] allows us to write the following (I will drop the brackets for brevity)
(𝑘1 𝐴 + 𝑘4 𝑃)𝐸𝑇
𝑋=
𝑘1 𝐴 + 𝑘4 𝑃 + 𝑘2 + 𝑘3
We can write the mass-action equation for the product as
𝑑P
= 𝑘3 X − 𝑘4 PE
𝑑𝑡
Substituting in for [X] and making use of stoichiometry (i.e. 𝑃 = 𝐴0 − 𝐴) yields
𝑑𝑃 (𝑘1 + 𝑘3 + 𝑘2 𝑘4 )𝐴 − 𝑘2 𝑘4 𝐴0
= 𝐸𝑇 ( )
𝑑𝑡 (𝑘1 − 𝑘4 )𝐴 + 𝑘2 + 𝑘3 + 𝑘4 𝐴0
In the case of no reversible binding, at 𝑡 → ∞ we will have [A] → 0. However, for reversible binding, at
𝑡 → ∞ we will have [A] approach some constant non-zero equilibrium value.
5.3.4 COMPETITIVE INHIBITION

Competitive inhibition occurs when there are two competing substrates for the same enzyme. This can be
generally given by the following two equations: A + E ↔ X → P + E and B + E ↔ Y. I will enumerate these
reactions as the same as before but now with a 𝑘4 and 𝑘5 for the new set of reactions.
The site balance in this case is now given by
𝐸𝑇 = E + X + Y
The resulting rate equation can be found to be
𝑑𝑃 𝑘3 𝐸𝑇 𝐴
=
𝑑𝑡 𝐾M (1 + 𝑁𝐵0 ) + 𝐴
where
𝑘2 + 𝑘3
𝐾M ≡
𝑘1
and
𝑘4
𝑁≡
𝑘5
5.3.5 NON-COMPETITIVE INHIBITION

Non-competitive inhibition occurs when a substrate, B, can bind to a remote site on an enzyme that prevents
A from binding or leaving the site that converts A to P. This is given by the following series of three
reactions: A + E ↔ X → P + E and B + E ↔ Y and B + X ↔ Z. As can be seen from this set of equations,
B can bind to either the free enzyme or the bound enzyme, both of which prevent the formation of P. We
will enumerate these reactions the same as before but with 𝑘6 and 𝑘7 for the new set of reactions.
The site balance can be written as

ET = E + X + Y + Z
The rate expression then becomes
𝑘
𝑑P 𝑘3 𝐸𝑇 𝐴 (1 + 3𝐿𝐵 ) 𝐸𝑇 𝐴
0
= =
𝑑𝑡 𝐾M + 𝐴 + 𝑁𝐵0 𝐾M + 𝐿𝐴𝐵0 𝐾 (1 + 𝑁𝐵0 ) + 𝐴
M 1 + 𝐿𝐵
0
where the new constant 𝐿 is given by

𝑘6
𝐿≡
𝑘7
REACTION NETWORKS | 25
6 REACTION NETWORKS
6.1 INTRODUCTION TO REACTION NETWORKS
Previously, we have discussed rate laws for a single reaction of the type 𝐴 → 𝐵. However, oftentimes we
have to deal with a complex reaction network, where there are many products produced from a reactant (or
multiple reactants), but little information is known a priori about how the products are actually produced.
Specifically, the reaction topology is often desired. In this section, we will refer to the “rank” (denote 𝑟) of
a species as the numerical order in which a product appears. If it comes first, it will have rank 1 and be
called primary; if second, it will have rank 2 and be called secondary, and so on. This is distinct from the
order of that species, which we will reserve for the value of 𝑛 in rate = 𝑘𝐶𝐴𝑛 (i.e. the dependence of rate
on concentration).
We will define yield of a species 𝑖 (compared to a reactant 𝐴) as
𝑁𝑖 − 𝑁𝑖0
𝑌𝑖 ≡
𝑁𝐴0
We will define the conversion of a reactant 𝐴 as
𝑁𝐴
𝑋𝐴 ≡ 1 −
𝑁𝐴0
We will define selectivity of a species 𝑖 (compared to a reactant 𝐴) as
𝑌𝑖 𝑁𝑖 − 𝑁𝑖0
𝑆𝑖 ≡ =
𝑋𝐴 𝑁𝐴0 − 𝑁𝐴
The differential selectivity will therefore be defined as
𝜕𝑌𝑖
𝑆𝑑𝑖 ≡
𝜕𝑋𝐴
6.2 DELPLOTS
To determine the rank of a reaction product, the method of delplots can be used. The first-rank delplot is
used to distinguish primary products from non-primary products. Higher rank delplots allow for the
discernment of products that are secondary, tertiary, and so on. The following notation will be used to
describe delplots: 1PA . Here, the left-hand superscript is a number representing the rank of the delplot (in
this case 1), P represents a product P, and the subscript A means that it is based on the conversion of species
A.
The first-rank delplot is a plot of selectivity versus conversion. In other words, it is a plot of 𝑌𝑃 /𝑋𝐴 vs. 𝑋𝐴 .
For each product P, the delplot is extrapolated to 𝑋𝐴 → 0 (i.e. when 𝑡 → 0 or 𝜏 → 0). This y-intercept,
denoted as 1PA , can be evaluated as
1 𝑃/𝐴0 𝑃
PA = lim = lim
𝜏→0 (𝐴0 − 𝐴)/𝐴0 𝜏→0 𝐴0 − 𝐴
This can be evaluated algebraically from given expressions of the concentrations of each product as a
function of time or numerically from the actual plots of selectivity versus conversion.
REACTION NETWORKS | 26
If a finite y-intercept is found in the first-rank delplot, then the product is primary because the initial rate
of formation of a primary product is always finite. If the y-intercept is zero, then the product is non-primary
(i.e. the rank is greater than 1) because the initial rate of a non-primary product is always zero (i.e. it lags).
Higher rank delplots allows for the sorting of products of rank greater than 1. The second-rank delplot
consists of a plot of 𝑌𝑃 /𝑋𝐴2 vs. 𝑋𝐴 . The y-intercept of the second-rank delplot for a product P is
2 𝑃/𝐴0 𝐴0 𝑃
PA = lim 2 = lim
𝜏→0 𝐴 −𝐴 𝜏→0 (𝐴0 − 𝐴)2
( 0𝐴 )
0
If a finite intercept is found, the product is secondary. If a zero intercept is found, the product’s rank is
greater than secondary. If a divergence is found as 𝜏 → 0 (i.e. no y-intercept), the product’s rank is primary.
This can be summarized generally. The delplot of rank 𝑚 is a plot of 𝑌𝑃 /(𝑋𝐴 )𝑚 vs. 𝑋𝐴 . The y-intercept can
be found via
𝑚 𝑃/𝐴0
PA = lim
𝜏→0 𝐴0 − 𝐴 𝑟
( )
𝐴0
The following general procedure can then be employed, which has been extended to include orders other
than first order. In the table, 𝑟 is the rank of the species and 𝑚 is the rank of the plot. Note that it is often
too difficult to notice changes from 0 intercepts to finite intercepts for 𝑛 > 1 when the rank is high. If you
encounter a species that has a 0 intercept no matter the rank of the Delplot, it is likely either 𝑛 = 1 and a
really high rank species (perhaps it is produced from intermediate species you are not aware of) or 𝑛 > 1
and some rank at the end of the chain. The specific order can often be determined by looking at
concentration versus time plots. In general, Delplots are limited at ranks greater than about 2 or 3 due to
the large propagations of error that occur.
Order Rank = 1 Rank > 1

1 𝑚
𝑛=1 PA = finite (otherwise 0) PA = 0 for 𝑟 > 𝑚
𝑚
PA = finite for 𝑟 = 𝑚
𝑚
PA = diverges for 𝑟 < 𝑚
1 𝑚
𝑛<1 PA = finite (otherwise 0) PA = 0 for 𝑟 > 𝑚
𝑚
PA = diverges for 𝑟 = 𝑚
1 𝑚
𝑛>1 PA = finite (otherwise 0) PA = 0 for 𝑟 ≥ 𝑚
𝑚
PA = finite for 𝑟 < 𝑚
KINETIC THEORY | 27
7 KINETIC THEORY
7.1 COLLISION THEORY
7.1.1 DISTRIBUTION LAWS
Collision theory can be used to estimate kinetic parameters from first-principles. It assumes that the
molecules are hard spheres and react by collision with one another. It also assumes that the attractive forces
between the species are negligible.
Although it will not be derived in this guide, it can be found that the probability distribution for a given
speed, 𝑐, is given by
4𝑐 2 𝑐2
𝑃(𝑐) 𝑑𝑐 = exp (− ) 𝑑𝑐
𝛼 3 √𝜋 𝛼2
where 𝛼 is the most probable speed can be found via
2𝑘𝐵 𝑇
𝛼=√
𝑚
The distribution of kinetic energy can be derived from this expression and can be given by
4 2𝐸 2𝐸
𝑃(𝐸) 𝑑𝐸 = 3
√ 3 exp (− ) 𝑑𝐸
𝛼 𝜋𝑚 𝑚𝛼 2
7.1.2 COLLISION FREQUENCIES

With the hard-sphere assumption and the distribution laws established, we can write an expression for the
number of collisions experienced per molecule per second. First, let us define the reaction cross section as
1
𝜎𝐴𝐵 = (𝜎𝐴 + 𝜎𝐵 )
2
The reaction cross section 𝜎𝐴𝐵 represents the distance between the centers of species 𝐴 and 𝐵 below which
collisions can occur. If we consider the trajectory of a single molecule 𝐴 through a stationary matrix of 𝐵
moving at a fixed speed, then the collision frequency is then given by
2
𝑍𝐴𝐵 = 𝜋𝜎𝐴𝐵 𝑐𝐴 𝑛𝐵
where 𝑐𝐴 is the speed of species 𝐴 and 𝑛𝐵 is the number density of 𝐵. The average speed from the Maxwell
distribution can be given as
8𝑘𝐵 𝑇
𝑐avg = √
𝜋𝜇𝐴𝐵
where 𝜇𝐴𝐵 is the reduced mass of 𝐴𝐵, given by

𝑚𝐴 𝑚𝐵
𝜇𝐴𝐵 ≡
𝑚𝐴 + 𝑚𝐵
KINETIC THEORY | 28
For the case of identical molecules, the reduced mass simplifies to half the mass of the molecule. This
relative speed is useful because now we can write the collision frequency of 𝐴 moving through a matrix of
𝐵 molecules when 𝐴 has a Maxwell distribution of speeds
2
8𝑘𝐵 𝑇
𝑍𝐴𝐵 = 𝜋𝜎𝐴𝐵 𝑛𝐵 √
𝜋𝜇𝐴𝐵
If both molecules 𝐴 and 𝐵 have Maxwellian distributions, then
2
8𝑘𝐵 𝑇
𝑍𝐴𝐵 = 𝜋𝜎𝐴𝐵 𝑛𝐴 𝑛𝐵 √
𝜋𝜇𝐴𝐵
If there is only one molecular species 𝐴 then the collision frequency is given by
2 2√ 𝐵
𝑘 𝑇
𝑍𝐴𝐴 = 4𝜋𝜎𝐴𝐴 𝑛𝐴
𝜋𝑚𝐴
7.1.3 RATE CONSTANTS

Naively, one could say that since 𝑍 is a collision frequency, then it is a rate and then 𝑍𝐴𝐵 = 𝑟𝐴 = −𝑘𝑛𝐴 𝑛𝐵
2 8𝑘 𝑇
for a bimolecular reaction of 𝐴 and 𝐵. This would imply that 𝑘 = 𝜋𝜎𝐴𝐵 √𝜋𝜇𝐵 . While this is reasonable for
𝐴𝐵
the pure collision of two species, it not reflective of the kinetics of reactions because there is no exponential
dependence on temperature. The missing link here is that molecules only react if they collide with sufficient
energy transfer. We therefore must define a threshold energy, 𝜂 ∗ , above which reactions can occur. We then
find that the reactive collision number is
2 √
8𝑘𝐵 𝑇 𝜂∗
𝑍 = 𝑝𝜋𝜎𝐴𝐵 exp (− )𝑛 𝑛
𝜋𝜇𝐴𝐵 𝑘𝐵 𝑇 𝐴 𝐵
Here, 𝑝 is a steric factor which is effectively a correction factor (unless otherwise defined, it can be taken
as a value of 1). For 𝑟𝐴 = −𝑘𝑛𝐴 𝑛𝐵 , we then have that
2 √
8𝑘𝐵 𝑇 𝜂∗
𝑘 = 𝑝𝜋𝜎𝐴𝐵 exp (− )
𝜋𝜇𝐴𝐵 𝑘𝐵 𝑇
From the above expression, we see that there is an exponential temperature dependence as would be
expected from Arrhenius’ law, but there is also a √𝑇 dependence in what would be analogous to the pre-
exponential factor, which is otherwise assumed to be independent of temperature in the Arrhenius equation.
The collision theory equation is fairly accurate for low values of 𝜂 ∗ , such as very exothermic elementary
steps like radical recombination.
7.2 LINDEMANN THEORY

The prior equation for 𝑘 from collision theory is not suitable for unimolecular reactions because it assumes
that bonds are formed from the collision of two molecules and does not capture decomposition that isn’t
from this relative kinetic energy of translational motion. We will present here a simplified version of
Lindemann theory in which collisions are instantaneous. Let us assume a reaction mechanism of
KINETIC THEORY | 29
𝑘1 ,𝑘−1
𝐴+𝐴↔ 𝐴 + 𝐴∗
𝑘2 ,𝑘−2
𝐴∗ ↔ products
If we apply PSSH on the activated intermediate, we get
𝑑𝐶𝐴∗
= 0 = 𝑘1 𝐶𝐴2 − 𝑘−1 𝐶𝐴 𝐶𝐴∗ − 𝑘2 𝐶𝐴∗
𝑑𝑡
Solving for 𝐶𝐴∗ yields
𝑘1 𝐶𝐴2
𝐶𝐴∗ =
𝑘2 + 𝑘−1 𝐶𝐴
The rate of product formation is then given by
𝑘1 𝑘2 𝐶𝐴2
𝑟 = 𝑘2 𝐶𝐴∗ =
𝑘2 + 𝑘−1 𝐶𝐴
At high pressures (i.e. high concentrations), we get
𝑘1 𝑘2
𝑟≈ 𝐶 = 𝑘∞ 𝐶𝐴
𝑘−1 𝐴
where 𝑘∞ is the observable rate constant. At low pressures (i.e. low concentrations) we get
𝑟 ≈ 𝑘1 𝐶𝐴2
We know that collision theory works very well for radical decomposition, so we can assume that 𝑘−1 can
be estimated well from collision theory and is therefore
2 2√
8𝑘𝐵 𝑇
𝑘−1 = 𝜋𝜎𝐴𝐴 𝑛𝐴
𝜋𝜇𝐴𝐵
where the exponential energy term can be ignored because 𝜂 ∗ ≈ 0 for this reaction. The probability
distribution function of energies in this system can also be found to be
𝐸∗ 𝐸 ∗ 𝑛−1
exp (− )( )
𝑘𝐵 𝑇 𝑘𝐵 𝑇
𝑓(𝐸 ∗ ) =
(𝑛 − 1)!
where 𝑛 is the number of degrees of vibrational freedom. For a nonlinear polyatomic molecule, it is given
by 𝑛 = 3𝑁 − 6 and a linear molecule it is given by 𝑛 = 3𝑁 − 5 where 𝑁 is the number of atoms in the
molecule. We can say that the equilibrium constant for the first reaction is given by
𝑘1 𝐶𝐴∗
𝐾1 = =
𝑘−1 𝐶𝐴
The term 𝐶𝐴∗ /𝐶𝐴 is essentially the fraction of the total number of molecules with energy greater than 𝐸 ∗ , so
𝑘1
= 𝑓(𝐸 ∗ )
𝑘−1
KINETIC THEORY | 30
We then have that
𝐸∗ 𝐸 ∗ 𝑛−1
8𝑘𝐵 𝑇 exp (− ) ( )
2 2√ 𝑘𝐵 𝑇 𝑘𝐵 𝑇
𝑘1 = 𝜋𝜎𝐴𝐴 𝑛𝐴
𝜋𝜇𝐴𝐵 (𝑛 − 1)!
If we look back at the rate constant for the high concentration case,
𝑘1 𝑘2
𝑘∞ = = 𝑘2 𝑓(𝐸)
𝑘−1
If we want to know what the observable rate constant is, we need to find 𝑘2 . The value of 𝑘2 is the
probability that the product decays and is given by
𝐸 ∗ 1−𝑛
𝑘2 = 𝜈̅ (1 − )
𝑛𝑘𝐵 𝑇
After a lot of math that is omitted here, we can arrive at the following expression for the observable rate
constant in the limit of high 𝐸 ∗ , which is
𝐸∗
𝑘∞ = 𝜈̅ exp (− )
𝑘𝐵 𝑇
which has the familiar form of the Arrhenius expression (𝜈̅ is an intrinsic frequency factor and thus has the
units of the pre-exponential factor in the Arrhenius equation). We then see that Lindemann theory succeeds
where collision theory fails – in the high 𝐸 ∗ limit.
7.3 TRANSITION STATE THEORY

7.3.1 PARTITION FUNCTIONS
Transition state theory (TST) treats the activated intermediate as a real species which means that we can
use thermodynamics to estimate its properties. Let us consider the follow equilibrium of the transition state
complex
𝐾𝐶‡
𝐴𝐵 + 𝐶 ↔ (𝐴𝐵𝐶)‡ → 𝐴 + 𝐵𝐶
We then have that
[𝐴𝐵𝐶 ‡ ]
𝐾𝐶‡ =
[𝐴𝐵][𝐶]
This can be determined from statistical mechanics via the following expression2
𝑁𝐴𝑛−1 𝑄 ‡ 𝛥𝑈 ‡
𝐾𝐶‡ = exp (− )
𝑄𝐴𝐵 𝑄𝐶 𝑅𝑇
where 𝑛 is the molecularity. The Δ𝑈 ‡ can only be obtained from calculations (typically the difference of
zero point energies of reactants and the transition state), the value of 𝑛 is the molecularity of the reaction,
𝛥𝑈 ‡ 𝜈
2
More generally, for the reaction 𝑎𝐴 + 𝑏𝐵 ↔ 𝑐𝐶 + 𝑑𝐷, we have that 𝐾𝐶 = 𝑁𝐴𝑛−1 exp (− 𝑅𝑇
) ∏𝑖 𝑄𝑖 𝑖 .
KINETIC THEORY | 31
and 𝑁𝐴 is Avogadro’s number (to convert the units properly to a mole basis). The 𝑄 values here represent
partition functions with units of inverse volume and are the products of vibrational, rotational, translational,
and electronic partition functions. Written mathematically,
𝑄 = 𝑄𝑡 𝑄𝑟 𝑄𝑣𝑁 𝑄𝑒
where 𝑁 is the number of vibrational modes. It is 3𝑁0 − 5 for linear molecules and 3𝑁0 − 6 for nonlinear
polyatomic molecules where 𝑁0 is the number of atoms. However, a key point must be addressed for
transition states. The power of 𝑁 must be decreased by one for transition states because there is one less
degree of freedom due to the pseudo-bond formed. For this,
𝑄 ‡ = 𝑄𝑡 𝑄𝑟 𝑄𝑣𝑁−1 𝑄𝑒
We now must define each of the partition functions. The translational partition function is defined as
3
2𝜋𝑚𝑘𝐵 𝑇 2
𝑄𝑡 ≡ ( )
ℎ2
and has units of inverse volume. It typically has a value of about 1024 cm-3. Importantly, if we can constrain
the system to be in a 2D space instead of 3D space, the exponent drops by a factor of 1/2.
The rotational partition function is different depending on the shape of the molecule. For a linear molecule,
8𝜋 2 𝐼𝑘𝐵 𝑇
𝑄𝑟 =
𝜎ℎ2
and for a nonlinear molecule
1 3
8𝜋 2 (8𝜋 3 𝐼1 𝐼2 𝐼3 )2 (𝑘𝐵 𝑇)2
𝑄𝑟 =
ℎ3 𝜎
In these equations, 𝜎 represents the symmetry number and is determined by the number of spatial
orientations of the subject molecule that are identical. For easy reference, it is a value of 2 for linear
molecules with a center of symmetry and 1 for linear molecules without a center of symmetry. The quantity
𝐼 is the moment of inertia, and for the nonlinear case they are the three principal moments. The moment of
inertia is defined as
𝐼 = ∑ 𝑚𝑖 𝑟𝑖2
𝑖
where 𝑟 is the distance to the axis of rotation. For a diatomic molecule, the moment of inertia is 𝐼 = 𝜇𝑅 2
where 𝜇 is the reduced mass and 𝑅 is the distance between the two atoms. For a linear, symmetric molecule
2
like CO2, the moment of inertia is 𝐼 = 2𝑚O 𝑟CO , where 𝑚O is the mass of the oxygen atom and 𝑟CO is the
C-O bond length. For a triatomic linear molecule, such as a D-D-H transition state, the moment of inertia
2 2
would approximately be 𝐼 = 𝑚𝐷 𝑟𝐷𝐷 + 𝑚𝐻 𝑟𝐻𝐷 (this assumes the center of the molecule is the center of
mass, which is a reasonable approximation). The value of 𝑄𝑟 is unitless and approximately 102 − 104 for
linear molecules and 103 − 106 for nonlinear molecules.
The vibrational partition function is given by
KINETIC THEORY | 32
𝑛
ℎ𝜈𝑖 −1
𝑄𝑣 = ∏ (1 − exp (− ))
𝑘𝐵 𝑇
𝑖
where 𝑛 is the degrees of vibrational freedom and 𝜈𝑖 is the vibrational frequency from IR or Raman
spectroscopy. The value of 𝑄𝑣 is unitless and approximately 1 to 10. Note that spectra normally yield
wavenumbers with units of inverse length. To convert a wavenumber 𝜈̃ to frequency, use 𝜈𝑖 = 𝑐𝜈̃𝑖 .
Finally, the electronic partition function is given by
𝐸𝑖
𝑄𝑒 = ∑ 𝑔𝑖 exp (− )
𝑘𝐵 𝑇
𝑖
where 𝑔𝑖 is the degeneracy and 𝐸𝑖 is the electronic energy above the ground state.
If it is impossible or too inconvenient to calculate the partition functions directly, one can still make order
of magnitude arguments. Generally, we can say that
𝑄𝑖 ≈ 𝑓𝑡3 𝑓𝑟𝑎 𝑓𝑣𝑁 𝑓𝑒
where 𝑎 = 2 and 𝑁 = 3𝑁0 − 5 for a linear molecule and 𝑎 = 3 and 𝑁 = 3𝑁0 − 6 for a polyatomic
nonlinear molecule. For the transition state,
𝑄 ‡ ≈ 𝑓𝑡3 𝑓𝑟𝑎 𝑓𝑣𝑁−1 𝑓𝑒
We then can plug in 𝑓𝑡3 ≈ 1024 cm−3, 𝑓𝑟2 ≈ 102 − 104 (linear) or 𝑓𝑟3 ≈ 103 − 106 (nonlinear), 𝑓𝑣 ≈ 1 −
3, and 𝑓𝑒 ≈ 1.
When using TST, oftentimes a number of approximations are made to make the math easier. Generally
speaking, 𝑄𝑒 is neglected unless dealing with transition metals in their excited states. Also, for species with
large vibrational frequencies, 𝑄𝑣 ≈ 1. Another common thing to realize is that 𝑄𝑟 = 1 if the species is
monatomic. When dealing with a heterogeneous catalyst surface, the following assumptions can typically
be made: the surface has all partition functions being approximately 1, and the transition state has 𝑄𝑡 ≈ 1.
7.3.2 COMPUTING RATES OF REACTION

Keeping with the prior example, the rate of reaction is simply
𝑟 = 𝑣‡ [𝐴𝐵𝐶 ‡ ]
where 𝑣‡ is the frequency of passage over the energy barrier and [𝐴𝐵𝐶 ‡ ] is the concentration of the
transition state. Substituting in for 𝐶‡ using the previously defined 𝐾𝐶‡ yields
𝑟 = 𝑣‡ 𝐾𝐶‡ [𝐴𝐵][𝐶]
We can define the frequency 𝑣‡ as the thermal energy provided such that
𝑅𝑇 𝑘𝐵 𝑇
𝑣‡ = =
𝑁𝐴 ℎ ℎ
so that we can say
𝑘𝐵 𝑇 ‡
𝑟= 𝐾 [𝐴𝐵][𝐶]
ℎ 𝐶
KINETIC THEORY | 33
Plugging in for 𝐾𝐶‡ yields
𝑘𝐵 𝑇 𝑁𝐴𝑛−1 𝑄 ‡ 𝛥𝑈 ‡
𝑟= exp (− ) [𝐴𝐵][𝐶]
ℎ 𝑄𝐴𝐵 𝑄𝐶 𝑅𝑇
We should note that this takes the form of an Arrhenius-like expression. We can say that the pre-exponential
factor is
𝑘𝐵 𝑇 𝑁𝐴𝑛−1 𝑄 ‡
𝑘0 =
ℎ 𝑄𝐴𝐵 𝑄𝐶
such that
𝛥𝑈 ‡
𝑘 = 𝑘0 exp (− )
𝑅𝑇
and then we get a functional form of

𝑟 = 𝑘[𝐴𝐵][𝐶]
like we would expect. For reference, a loose TST – something like bond dissociation – has 𝑘0 ≈ 1016 s−1
and a tight TST – something like associative bonding – has 𝑘0 ≈ 109 M −1 s−1.
𝐸𝐴
We have therefore seen that for the general equation 𝑘 = 𝑘0 𝑇 𝑚 exp (− ), 𝑚 = 0 for Arrhenius (which
𝑅𝑇
is empirical and works over small 𝑇 ranges), 𝑚 = 1/2 for collision theory (which works best for small 𝐸𝑎 ),
and 𝑚 = 1 for large transition state theory (which works best for large 𝐸𝑎 ).
7.3.3 THERMODYNAMIC ANALYSIS

Recall that we can say that
Δ𝐺 ∘‡ = −𝑅𝑇 ln(𝐾𝑎‡ )
We can then of course use the fact that
Δ𝐺 ∘‡ = Δ𝐻 ∘‡ − 𝑇Δ𝑆 ∘ ‡
to say
𝛥𝑆 ∘ ‡ 𝛥𝐻 ∘‡
𝐾𝑎‡ = exp ( ) exp (− )
𝑅 𝑅𝑇
We can then write the rate expression as
𝑘𝐵 𝑇 𝛥𝑆 ∘ ‡ 𝛥𝐻 ∘‡
exp ( 𝑅 ) exp (− 𝑅𝑇 )
ℎ
𝑟= [𝐴𝐵][𝐶]
𝐶 ∘ 𝑛−1
where 𝐶 ∘ is a reference concentration (usually given by 𝑃/𝑅𝑇 at 1 atm and 𝑇 for gases or as the molal
concentration of pure components for liquids) and 𝑛 is the molecularity.
KINETIC THEORY | 34
7.3.4 HAMMETT EQUATION

Recall that we have the Evans-Polanyi relationship of 𝐸 = 𝐸0 + 𝛼Δ𝐻rxn for a family of reactions. We note
that, from TST, a family of reactions is similar if it has a similar Δ𝑆 ‡ . Oftentimes, other quantities aside
from the Evans-Polanyi relationships are used, especially when dealing with organic molecules. They use
thermochemical or structural properties.
One of the most common set of parameters are the Hammett constants. Consider the reaction of aromatic
species with various substituent groups, denote 𝑥. An example is shown below (the group is indicated by a
𝑌 though)
The Hammett equation relates the substituted group 𝑥 with the rate of reaction and takes into account the
effect of inductive and resonance effects. It can be given by
𝑘𝑥
log ( ) = 𝛾𝐻 𝜎𝑥
𝑘𝐻
𝐾
where 𝜎𝑥 ≡ log (𝐾𝑥 ) and is a tabulated quantity3. The lower case 𝑘 values refer to reaction rate constants,
𝐻
and the capital 𝐾 values refer to equilibrium (acid) dissociation constants. For the Hammett equation, a
substituted group of 𝑥 = 𝐻 (i.e. the group is a hydrogen atom) is often used as a reference.
Higher values of 𝜎𝑥 indicate that the equilibrium constant is increased for a chemical group 𝑥 with respect
to the hydrogen reference, meaning that the reaction is shifted more in the forward direction (specifically,
via a stabilization of negative charge via resonance and inductive effects). Therefore, the acidity of the
reactant molecule increases when 𝜎𝑥 > 0. Such a substituent is considered to be an electron-withdrawing
group because it will stabilize the product anion by withdrawing negative charge away from the reaction
site. Conversely, molecules with 𝜎𝑥 < 0 indicate an electron-donating group where the acid dissociation is
disfavored since the electron density is increased near the reaction site.
There are also trends that can be seen with the 𝛾𝐻 values, which are easier to see when you make the
equation 𝑘𝑥 = 𝑘𝐻 10𝛾𝐻 𝜎𝑥 . For 𝛾𝐻 > 1, the reaction is more sensitive to substituents than the reference
molecule, and negative charge is built up in the transition state of the reaction. The reaction is assisted by
electron-withdrawing groups. For 𝛾𝐻 = 0, there is no sensitivity to substituents, and no charge is built or
lost. For 𝛾𝐻 < 0, the reaction builds up positive charge in the transition state of the reaction. The reaction
is assisted by electron-donating groups.
It should also be noted that 𝜎𝑥 values are approximately additive, so if an aromatic ring has 2 Cl groups,
then the 𝜎𝑥 value can be approximated as 2𝜎Cl .
There are alternate forms of the Hammett parameter that are parameterized for the reactions that require
positive and negative charge stabilization, given by 𝜎 + and 𝜎 − , respectively. For pure inductive effects, 𝜎′
can be used.
3
Some references use log10 whereas others use ln.
KINETIC THEORY | 35
7.3.5 TAFT EQUATION

The Taft equation is another linear free energy relationship and was developed as a modification to the
Hammett relationship. The Taft equation describes the effect of sterics and induction. The general equation
is
𝑘𝑥
log ( ) = 𝛾 ∗ 𝜎 ∗ + 𝐸𝑆
𝑘CH3
where 𝜎 ∗ is a purely inductive effect and 𝐸𝑠 is a purely steric effect. In the limit of steric effects being
negligible, it simplifies to the same functional form as the Hammett equation. The trends in 𝛾 ∗ are then the
same as the trends for 𝛾𝐻 .
SURFACE CATALYSIS | 36
8 SURFACE CATALYSIS
8.1 ADSORPTION RATE LAWS
8.1.1 MOLECULAR ADSORPTION
We will now consider surface catalysis, wherein a reactant in a continuum diffuses into a catalyst particle,
binds to the surface, reacts to form a new species, desorbs, diffuses out of the catalyst particle, and enters
the continuum. As can be seen from this outline, heterogeneous catalysis incorporates adsorption, transport
phenomena, reactions, and bulk reactor models. We will focus on developing a basic kinetic model first
from elementary steps including adsorption, desorption, and surface reaction.
Consider the first step: adsorption. We will use an asterisk to denote a surface site. As such, the adsorption
step can be given by
𝑘1 ,𝑘−1
𝐴 +∗↔ 𝐴∗
For example, this could be CO + S ∗ ↔ CO∗ (note that when dealing with gases, such as CO, it is standard
to use a partial pressure instead of concentration). This specific type of adsorption is referred to molecular
(or non-dissociative) adsorption. The equation for the net rate of adsorption can then be written as
𝐶𝐴∗
𝑟ad = 𝑘1 𝐶𝐴 𝐶∗ − 𝑘−1 𝐶𝐴∗ = 𝑘1 (𝐶𝐴 𝐶∗ − )
𝐾𝐴
where 𝐾𝐴 ≡ 𝑘1 /𝑘−1 and is referred to as the adsorption equilibrium constant. The adsorption rate constant
𝑘1 is relatively independent of temperature whereas the desorption constant 𝑘−1 increases exponentially
with increasing temperature, such that the equilibrium adsorption constant 𝐾𝐴 decreases exponentially with
increasing temperature. A site balance can generally be written as
𝐶𝑇 = 𝐶∗ + ∑ 𝐶𝑖∗
𝑖
where the summation term accounts for any other adsorbed species to the surface sites. In this example,
𝐶𝑇 = 𝐶∗ + 𝐶𝐴∗
We know that at equilibrium, the net rate of adsorption should be zero. Employing this condition yields
𝐶𝐴∗ = 𝐾𝐴 𝐶𝐴 𝐶∗
Substituting in for 𝐶∗ from the site balance yields
𝐶𝐴∗ = 𝐾𝐴 𝐶𝐴 (𝐶𝑇 − 𝐶𝐴∗ )
which can be rearranged to
𝐾𝐴 𝐶𝐴 𝐶𝑇
𝐶𝐴∗ =
1 + 𝐾𝐴 𝐶𝐴
This equation is specifically called the Langmuir isotherm. This equation can be linearized to the following
form for data-fitting purposes, as shown below
𝐶𝐴 1 𝐶𝐴
= +
𝐶𝐴∗ 𝐾𝐴 𝐶𝑇 𝐶𝑇
Typically, the parameter 𝜃 is defined as the number of moles of species adsorbed divided by the number of
moles in the monolayer. This would be
𝐶𝐴∗ 𝐾𝐴 𝐶𝐴
𝜃= =
𝐶𝑇 1 + 𝐾𝐴 𝐶𝐴
The linearized form is
1 1
= +1
𝜃 𝐾𝐴 𝐶𝐴
Oftentimes, 𝜃 is defined as the volume adsorbed divided by the volume adsorbed at saturation: 𝑉/𝑉𝑀 . In
this definition, the above expression simply becomes
1 1 1 1
= ( )( )+
𝑉 𝐶𝐴 𝐾𝐴 𝑉𝑀 𝑉𝑀
8.1.2 DISSOCIATIVE ADSORPTION

It is possible to have types of adsorption that differ from the previous scenario. For instance, consider a
molecule AB. It can either adsorb molecularly, as described before, or it can dissociate and have A bound
to one surface site and B bound to a different surface site. This is described below
𝑘1 ,𝑘−1
AB + 2 ∗↔ A∗ + B ∗
This could be CO + 2 ∗↔ C ∗ + O∗ , for instance. In this case, the adsorption rate can be written as
𝐶𝐴∗ 𝐶𝐵∗
𝑟ad = 𝑘1 𝐶𝐴𝐵 𝐶∗2 − 𝑘−1 𝐶𝐴∗ 𝐶𝐵∗ = 𝑘1 (𝐶𝐴𝐵 𝐶∗2 − )
𝐾𝐴
where 𝐾𝐴 ≡ 𝑘1 /𝑘−1 once again. For dissociative adsorption, both 𝑘1 and 𝑘−1 increase exponentially with
temperature, unlike the case with molecular adsorption. We can then employ equilibrium conditions such
that
𝐶𝐴∗ 𝐶𝐵∗ = 𝐾𝐴 𝐶𝐴𝐵 𝐶∗2

The site balance is now
𝐶𝑇 = 𝐶∗ + 𝐶𝐴∗ + 𝐶𝐵∗
Substituting in for 𝐶∗ into our rate equation at equilibrium yields
𝐶𝐴∗ 𝐶𝐵∗ = 𝐾𝐴 𝐶𝐴𝐵 (𝐶𝑇 − 𝐶𝐴∗ − 𝐶𝐵∗ )2

For the case of 𝐶𝐴∗ = 𝐶𝐵∗ (which would be anticipated), we can arrive at
1
(𝐾𝐴 𝐶𝐴𝐵 )2 𝐶𝑇
𝐶𝐴∗ = 𝐶𝐵∗ = 1
1 + 2(𝐾𝐴 𝐶𝐴𝐵 )2
We now have a Langmuir isotherm for this system. Using the 𝜃 nomenclature, this is the same as
1
(𝐾𝐴 𝐶𝐴𝐵 )2
𝜃= 1
1 + 2(𝐾𝐴 𝐶𝐴𝐵 )2
The equation can be linearized by
√𝐶𝐴𝐵 1 2√𝐶𝐴𝐵
= +
𝐶𝐴∗ 𝐶𝑇 √𝐾𝐴 𝐶𝑇
which equivalently is the following when 𝐶𝑇 is multiplied through and √𝐶𝐴𝐵 is divided through
1 1
= +2
𝜃 √𝐾𝐴 𝐶𝐴𝐵
It can be easily shown that the dissociative adsorption process of A2 + 2 ∗↔ 2𝐴∗ is identical to the case
1
before but the denominator in the expression for 𝐶𝐴∗ is 1 + (𝐾𝐴 𝐶𝐴2 )2 where the factor of 2 disappears
because the site balance is just 𝐶𝑇 = 𝐶∗ + 𝐶𝐴∗ . Therefore, in this case
1
(𝐾𝐴 𝐶𝐴2 )2 𝐶𝑇
𝐶𝐴∗ = 1
1 + (𝐾𝐴 𝐶𝐴2 )2
We can define the 𝜃 parameter then such that

1
(𝐾𝐴 𝐶𝐴2 )2
𝜃= 1
1 + (𝐾𝐴 𝐶𝐴2 )2
which can be linearized to

1 1
= +1
𝜃 √𝐾𝐴 𝐶𝐴2
Oftentimes, a plot of 1/𝜃 vs. 1/√𝐶𝐴2 will have deviations at high and low coverages (and by extension
high and low partial pressures) due to adsorbate-adsorbate interactions and site heterogeneity.
8.1.3 COMPETITIVE ADSORPTION

It is possible to have two species (A and B) competing for surface sites. The site balance would then be
written as
𝐶𝑇 = 𝐶∗ + 𝐶𝐴∗ + 𝐶𝐵∗
We start at
𝐶𝐴∗ = 𝐾𝐴 𝐶𝐴 𝐶∗
Plugging in the new site balance yields
𝐶𝐴∗ = 𝐾𝐴 𝐶𝐴 (𝐶𝑇 − 𝐶𝐴∗ − 𝐶𝐵∗ )
We know that
𝐶𝐵∗ = 𝐾𝐵 𝐶𝐵 𝐶∗
So
𝐶𝐴∗ = 𝐾𝐴 𝐶𝐴 (𝐶𝑇 − 𝐶𝐴∗ − 𝐾𝐵 𝐶𝐵 𝐶𝐴∗ )
which can be rearranged to
𝐾𝐴 𝐶𝐴 𝐶𝑇
𝐶𝐴∗ =
1 + 𝐾𝐴 𝐶𝐴 + 𝐾𝐵 𝐶𝐵
An analogous procedure for B would yield
𝐾𝐵 𝐶𝐵 𝐶𝑇
𝐶𝐵∗ =
1 + 𝐾𝐴 𝐶𝐴 + 𝐾𝐵 𝐶𝐵
8.2 SURFACE REACTION RATE LAWS

8.2.1 SINGLE SITE
After a reactant has been adsorbed, it can then react. A single site mechanism is one in which the site where
the reactant is adsorbed is the only one involved in the reaction. Written out, this would be
𝑘2 ,𝑘−2
A∗ ↔ B∗
for the transformation of species A to B on the catalyst surface. The rate law for this is given by
𝐶𝐵∗
𝑟SR = 𝑘2 (𝐶𝐴∗ − )
𝐾SR
where 𝐾SR ≡ 𝑘2 /𝑘−2 .
8.2.2 DUAL SITE

A dual-site mechanism is one in which a reactant adsorbed to a site reacts with a separate site. One way this
can be written is by the following equation:
𝑘2 ,𝑘−2
A∗ +∗↔ B ∗ +∗
The rate law for this is given by
𝐶𝐵∗ 𝐶∗
𝑟SR = 𝑘2 (𝐶𝐴∗ 𝐶∗ − )
𝐾SR
A dual site reaction may also consist of a reaction between two adsorbed species, given by
𝑘2 ,𝑘−2
A∗ + B ∗ ↔ C ∗ + D∗
The rate law for this is given by
𝐶𝐶∗ 𝐶𝐷∗
𝑟SR = 𝑘1 (𝐶𝐴∗ 𝐶𝐵∗ − )
𝐾SR
8.2.3 REACTION WITH UNBOUND SPECIES

A reaction can also occur between an adsorbed molecule and a molecule in the continuum (when the
adsorbed molecule reacts with a molecule in the gas phase, this is typically referred to as an Eley-Rideal
mechanism). The general equation is
𝑘2 ,𝑘−2
A∗ + B ↔ C∗
The corresponding rate equation is
𝐶𝐶∗
𝑟SR = 𝑘2 (𝐶𝐴∗ 𝐶𝐵 − )
𝐾SR
8.3 DESORPTION RATE LAWS

The process of desorption is given by
𝑘3 ,𝑘−3
C∗ ↔ C +∗
The rate of desorption can be given by
𝐶𝐶 𝐶∗
𝑟Des = 𝑘3 (𝐶𝐶∗ − )
𝐾Des
where 𝐾Des ≡ 𝑘3 /𝑘−3 . We can easily see that the desorption equilibrium constant is simply the inverse of
the adsorption equilibrium constant. As such, we can arrive at
𝑟Des = 𝑘3 (𝐶𝐶∗ − 𝐾Ad 𝐶𝐶 𝐶∗ )
or
𝐶𝐶∗
𝑟Des = 𝑘−3 ( − 𝐶𝐶 𝐶∗ )
𝐾Ad
if we divide through by the adsorption equilibrium constant.
8.4 DETERMINING THE REACTION MECHANISM AND RATE-LIMITING STEP

The following procedure can generally be employed when trying to determine the reaction mechanism for
a given reaction in terms of measurable:
1. Propose a mechanism consisting of an adsorption step, surface reaction step, and desorption step
2. Assume a rate-limiting step. The first guess should generally be the surface reaction step.
3. Write out the elementary rate law for the rate-limiting step
4. There will be immeasurable species in Step 3 that you will want removed. Write out the elementary
rate laws for the relevant reactions and apply equilibrium as needed to solve for the concentrations
of adsorbed species (𝐶𝑖∗ ). If the reactions are not reversible, then instead say that the net rate of
generation of 𝐶𝑖∗ is zero and apply PSSH.
5. Write out the site balance and solve for the concentration of vacant sites (plugging in quantities
from Step 4). You don’t want the concentrations of empty sites in the overall rate expression, so
this will help get rid of them.
6. Derive the rate law by combining Steps 2-5.
7. Compare the results with experimental data. If the data does not agree, choose a different rate-
limiting step and/or propose a different mechanism.
8.5 NONIDEALITIES
8.5.1 NONIDEAL SURFACES
Real surfaces are not perfect and have defects/deformations that cause Δ𝐻ads to not be constant over the
entire surface. We know that the equilibrium constant is
𝛥𝐺𝑎𝑑𝑠 𝛥𝑆𝑎𝑑𝑠 Δ𝐻ads
𝐾eq = exp (− ) = exp ( − )
𝑅𝑇 𝑅 𝑅𝑇
From thermodynamics and the above expression, we then see that Δ𝐻ads < 0 is favorable and Δ𝑆ads > 0
is favorable. In reality, Δ𝐻ads is a function of 𝜃. The Langmuir models, derived in the following subsections,
assume Δ𝐻ads is a constant and is only good for isolated sites.
Additional models have been developed. The Temkin model states that
0 (1
Δ𝐻ads (𝜃) = Δ𝐻𝑎𝑑𝑠 − 𝛼𝜃)
0
where Δ𝐻ads is the heat of adsorption at zero surface coverage and 𝛼 is a fitting parameter. This expression
is an empirical relationship that simply makes Δ𝐻ads linearly decrease with increasing 𝜃. For molecular
adsorption, the Temkin isotherm is
𝑅𝑇
𝜃≈ 0 ln(𝐾𝐴 𝐶𝐴 )
Δ𝐻ads 𝛼
The Freundlich model states that

0
Δ𝐻ads (𝜃) = −Δ𝐻ads ln(𝜃)
This model completely breaks down as 𝜃 → 0. It can be shown via the Clausius-Clapeyron equation that
the Freundlich isotherm for molecular adsorption takes the form of
𝜃 = 𝛼𝐶𝐴𝑚
where 𝛼 and 𝑚 fitting parameters. The value of 𝑚 is directly related to the intensity of adsorption. The
value of 𝛼 is an indicator of the adsorption capacity.
8.5.2 STICKING PROBABILITY

The sticking probability is defined
# molecules that stick
𝑆(𝜃) =
# molecules hitting surface
This is often reported as 𝑆(𝜃)/𝑆(0) in the literature. Ideally, 𝑆(𝜃) linearly decreases with increasing 𝜃 since
the loss of available sites makes it less likely that a molecule will stick when it hits the surface. However,
in reality it can take on other shapes. For instance, let us consider
𝑘𝑎
𝐴2 ↔ 𝐴2,surface + 2 ∗→ 2𝐴∗
𝑘𝑑
This is showing that a diatomic gas A2 comes in contact with the surface but does not always bind it. We
can apply PSSH to the A2,surface intermediate such that
𝑑[𝐴2,surfrace ]
= 𝜙𝐹𝐴 − 𝑘𝑑 [𝐴2,surface ] − 𝑘𝑎 [𝐴2,surface ](1 − 𝜃)2 ≈ 0
𝑑𝑡
Here, 𝜙 is the trapping ratio, 𝐹𝐴 is the flux of 𝐴, and 𝜃 is the fractional coverage. Note that the (1 − 𝜃)2
term comes in because that is effectively 𝐶∗2 /𝐶𝑇2 . We know that the sticking coefficient is
𝑘𝑎 [𝐴2,surface ](1 − 𝜃)2

𝑆(𝜃) =
𝐹𝐴
since this represents the number of molecules that stick to the surface divided by the number of molecules
that hit. Plugging in the PSSH equation solved for 𝐹𝐴 into the above definition of the sticking coefficient
yields
𝑘𝑎 [𝐴2,surface ](1 − 𝜃)2 𝜙 𝑘𝑎 (1 − 𝜃)2 𝜙

𝑆(𝜃) = =
𝑘𝑑 [𝐴2,surface ] + 𝑘𝑎 [𝐴2,surface ](1 − 𝜃)2 𝑘𝑑 + 𝑘𝑎 (1 − 𝜃)2
At zero coverage,
𝑘𝑎 𝜙
𝑆(0) =
𝑘𝑑 + 𝑘𝑎
so that
𝑆(𝜃) (𝑘𝑑 + 𝑘𝑎 )(1 − 𝜃)2
=
𝑆(0) 𝑘𝑑 + 𝑘𝑎 (1 − 𝜃)2
If we define 𝐾 ≡ 𝑘𝑎 /𝑘𝑑 then we arrive at
𝑆(𝜃) (1 + 𝐾)(1 − 𝜃)2
=
𝑆(0) 1 + 𝐾(1 − 𝜃)2
Recall from the molecular adsorption section that 𝐾 decreases exponentially with increasing temperature.
As such, at the low temperature limit and low values of 𝜃, we find that 𝑆(𝜃)/𝑆(0) ≈ 1. At the high
temperature limit, we find that 𝑆(𝜃)/𝑆(0) ≈ (1 − 𝜃)2.
REACTIONS IN HETEROGENEOUS SYSTEMS | 43
9 REACTIONS IN HETEROGENEOUS SYSTEMS

9.1 DEFINITIONS
9.1.1 DIFFUSIVITY
In kinetic processes that occur with heterogeneous catalysts, the reactant must diffuse through the porous
catalyst in order to react. Therefore, there are oftentimes diffusion limitations that must be considered. We
will first define a few relevant terms when dealing with transport in porous media.
The void fraction and tortuosity of a porous particle are defined as
volume of void space
𝜀 ≡ void fraction =
total volume (voids + solids)
actual distance between two points
𝜏 ≡ tortuosity =
shortest distance between two points
There is also the molecular diffusivity, often denote 𝐷𝑚 , which is a strong function of the particle diameter.
The effective diffusivity (for a species 𝐴) is defined as
𝜀𝐷𝑚,𝐴
𝐷𝐸𝐴 =
𝜏
9.1.2 THIELE MODULUS
The Thiele modulus for reaction order 𝑛 in a spherical geometry is defined as4
𝑛−1
𝑅 2 𝑘𝑛 𝐶𝐴,𝑠 surface reaction
𝜙𝑛2 = ∼
𝐷𝐸𝐴 diffusivity
When the Thiele modulus is small, the kinetics are reaction limited whereas when it is large, the reaction is
diffusion limited. Note that when 𝑛 = 1, there is no dependence on 𝐶𝐴,𝑠 . The general definition of the Thiele
modulus for an arbitrary geometry is given by
𝑛−1
𝑛 + 1 𝑘𝑛 𝐶𝐴,𝑠
𝜙𝑛2 = 𝑊 2 ( ⋅ )
2 𝐷𝐸𝐴
where 𝑊 is an arbitrary volume/surface area expression for the particular shape. It should be noted that this
definition of 𝜙𝑛2 is different by a numerical factor. Generally, just use the first definition for spheres and
the second one for more complicated geometries. Everything is an order of magnitude argument anyway.
9.1.3 EFFECTIVENESS FACTOR

The effectiveness factor is defined as
observed reaction rate
𝜂=
reaction rate if all 𝐶𝐴 = 𝐶𝐴,𝑠
4
Note that the definition of the Thiele modulus oftentimes has the 𝑘𝑛 term replaced with a 𝑘𝑛 𝜌𝐵 𝑆𝐴 term. In this, 𝜌𝐵
is the bulk density and 𝑆𝐴 is the surface area divided by the mass of the catalyst. This is because oftentimes rates are
reported in moles/(cm2 s) and therefore 𝑘𝑛 has different units than usual. If one is dealing with a rate that has
moles/(cm3 s) or equivalent dimensions, then simply omit the 𝑆𝐴 𝜌𝐵 unit correction term in the definition. If one is
dealing with a rate that has units of 1/s then keep the 𝜌𝐵 term but omit the 𝑆𝐴 term.
In the limit of 𝜂 → 1 (𝜙𝑛 → 0), there are no diffusion limits. In the limits of 𝜂 → 0 (for 𝜙𝑛 → ∞), there are
significant diffusion limits. For a first-order reaction in a sphere,
3(𝜙1 coth(𝜙1 ) − 1) 3
𝜂= 2 = 2 (𝜙𝑛 coth(𝜙𝑛 ) − 1)
𝑅 𝑘1 𝜌𝐵 𝑆𝐴 𝜙𝑛
𝐷𝐸𝐴
It turns out the right-hand form of this expression is true for other orders in spherical catalysts as well since
the reaction order only changes the power of 𝑛 − 1 in the Thiele modulus.
For a flat plate,
tanh(𝜙)
𝜂=
𝜙
The observed rate is
𝐸𝑎
−𝑟𝐴,obs = −𝜂𝑟𝐴,𝑠 = 𝜂𝑘0 𝑒 −𝑅𝑇 𝐶𝐴,𝑠
𝑛 𝑛
= 𝜂𝑘𝑛 𝐶𝐴,𝑠
where 𝑘𝑛 is the intrinsic rate constant. Also note that 𝜂𝑘𝑛 = 𝑘obs .
9.2 LIMITING CASES

9.2.1 NO DIFFUSION LIMITATIONS
There are negligible diffusion limitations when 𝜂 is close to 1 and when 𝜙 is close to zero. When the Thiele
modulus is small,
𝜙 → 0, 𝜂→1
For small 𝜙, the observed reaction rate is the surface rate (no diffusion limits), so
𝑟obs
𝜙 = small, 𝜂 → 1, 𝑟𝐴,obs = 𝑟𝐴,𝑠 , 𝑘𝑛 = − 𝑛
𝐶𝐴,𝑠
This can happen at small particle radius, small rate constants, high effective diffusivity, and/or low
temperatures. Therefore, when running at these conditions, one is measuring the real (i.e. intrinsic rate
constant).
9.2.2 DIFFUSION LIMITATIONS

There are diffusion limitations when 𝜂 is less than 1 and when 𝜙 is greater than zero. It is always true that
for large Thiele modulus,
1
𝜙 → ∞, 𝜂∝
𝜙
The above expression becomes an equality of 𝜂 → 3/𝜙 for a sphere and 𝜂 → 1/𝜙 for a plate. Then,
𝜂1 𝜙𝑛,2
=
𝜂2 𝜙𝑛,1
For large 𝜙, we know that for a sphere is the following
𝑛
3 𝑛
3
𝜙 = large, − 𝑟𝐴,obs = 𝜂𝑘𝑛 𝐶𝐴,𝑠 = 𝑘𝑛 𝐶𝐴,𝑠 = √𝐷𝐸𝐴 √𝑘𝑛 𝐶𝐴,𝑠
𝜙𝑛 𝑅
𝐸𝑎
Recall that 𝑘𝑛 is an intrinsic rate constant and can be expressed as √𝑘𝑛 = √𝑘𝑛,0 exp (− ) where 𝑘𝑛,0 is
2𝑅𝑇
the pre-exponential factor. This therefore says that for large 𝜙 (diffusion limitations), we have
𝐸𝐴
𝜙 = large, 𝐸𝐴,obs =
2
If one includes the temperature dependence of 𝐷𝐸𝐴 , it can be shown that
𝐸𝐴 + 𝐸diffusion
𝐸𝐴,obs =
2
The apparent order can be expressed as the following (for 𝑛 ≠ 1)
𝑛 − 1 𝑑 ln 𝜂
𝑛app = 𝑛 +
2 𝑑 ln 𝜙𝑛
such that the observed order is between the true order and 1.
One can find the diffusivity in the diffusion limiting regime by backing out 𝜂 from
𝑛
𝑟𝐴,obs = 𝜂𝑘𝑛 𝐶𝐴,𝑠
then using this value of 𝜂 to back out 𝐷𝐸𝐴 from

−1
𝑛−1
3 𝑘𝑛 𝐶𝐴,𝑠
𝜂 = = 3 (𝑅√ )
𝜙 𝐷𝐸𝐴
9.3 DETERMINING IF DIFFUSION LIMITATIONS ARE DOMINANT

9.3.1 CHANGING PARTICLE SIZE
It is generally true that the ratio of any two rates is
𝑟𝐴,obs,1 𝜂1
=
𝑟𝐴,obs,2 𝜂2
provided 𝑘𝑛 and 𝐶𝐴,𝑠 is the same between the two rates (e.g. if you change the catalyst size). If all the
parameters are the same except catalyst size, one can also state that
𝜙1 𝑅1
=
𝜙2 𝑅2
For the case of diffusion limitations, since 𝜂 ∝ 1/𝜙, one can state that
𝜂1 𝜙𝑛,2
=
𝜂2 𝜙𝑛,1
Therefore, when in a diffusion limiting regime,

𝑟𝐴,obs,1 𝑅2
=
𝑟𝐴,obs,2 𝑅1
The above expression can be tested to see if one is operating in the diffusion-limiting regime. Conversely,
there are no diffusion limitations if changing the radius has no impact on the observed rate.
9.3.2 WEISZ-PRATER CRITERION

Start with the definition of the Thiele modulus rearranged slightly as
𝜙𝑛2
𝑘𝑛 = 𝑛−1 𝐷𝐸𝐴
𝑅 2 𝐶𝐴,𝑠
Substituting this into the rate expression yields

𝜙𝑛2
𝑟𝐴,obs = 𝜂 𝐷 𝐶
𝑅 2 𝐸𝐴 𝐴,𝑠
This can be rearranged to
𝑟𝐴,obs 𝑅2
𝜂𝜙𝑛2 =
𝐷𝐸𝐴 𝐶𝐴,𝑆
For 𝜙 ≪ 1, 𝜂 = 1 there is no pore diffusion limitations and so we expect 𝜂𝜙𝑛2 ≪ 1. For 𝜙 ≫ 1, 𝜂 ∝ 1/𝜙𝑛
there is strong pore diffusion limitations and 𝜂𝜙𝑛2 ≫ 1. Note that for first order, the above expression
simplifies to
𝑘obs 𝑅2
𝜂𝜙12 =
𝐷𝐸𝐴
Oftentimes, the above expression is written with a 𝑊 2 instead of 𝑅 2 such that for a sphere 𝑊 2 = (𝑅/3)2 .
This causes a difference of a factor of 9 but should not change the overall trend.
9.4 EXTERNAL MASS TRANSFER

9.4.1 MASS TRANSFER COEFFICIENT
In this section, we will consider the case where 𝐶𝐴,bulk ≠ 𝐶𝐴,surf . In reality, there is a boundary layer
between the bulk and the pellet surface where the concentration varies. The mass flux at the surface of the
catalyst can be given by
mass flux = 𝑘𝑔 (𝐶𝐴,b − 𝐶𝐴,𝑠 )
where 𝑘𝑔 is the mass transfer coefficient. Of course, the rate of reaction at the catalyst surface is given by
𝑛
𝑟𝐴 = 𝑘𝐶𝐴,𝑠 . Let us assume, for example, 𝑛 = 2. We can then equate these two expressions, solve for 𝐶𝐴,s ,
and plug back into 𝑟𝐴 to arrive at
𝑘𝑔 𝑘𝑔
𝑟𝐴 = 𝑘𝑔 ((1 + ) − √(1 + ) − 1) 𝐶𝐴,b
2𝑘𝐶𝐴,b 2𝑘𝐶𝐴,b
This is clearly neither 1st nor 2nd order but some intermediate. In the limit of 𝑘 ≫ 𝑘𝑔 , we arrive at
𝑟𝐴 = 𝑘𝑔 𝐶𝐴,s
which is 1st order and limited due to mass transfer. Contrastingly, in the limit of 𝑘𝑔 ≫ 𝑘, we arrive at
2
𝑟𝐴 = 𝑘𝐶𝐴,s
which is what we’d expect – 2nd order and a rate limited by reaction.
9.4.2 MASS TRANSFER IN REACTOR ENGINEERING

In reactors, one typically does not measure the rate directly. Instead, inputs and outputs are measured. For
example, the differential conversion is usually found, which is defined as
𝐹𝐴0 − 𝐹𝐴
𝑋𝐴 =
𝐹𝐴0
The rate is then found from
𝑋𝐴 𝐹𝐴0
𝑟𝐴 = −
𝑊
where 𝑊 is the catalyst weight. Oftentimes, it is best to keep 𝑋𝐴 < 10% since the kinetics can’t be read
easily if it is higher.
9.4.3 NONISOTHERMAL THEORY

A number of catalytic reactions are accompanied by thermal effects from the heat of reaction, so it is
important to consider a combined mass and energy balance approach. The result of such an approach yields
𝐷𝑒𝐴 (−Δ𝐻)
𝑇𝑠 − 𝑇𝑠,𝑠 = (𝐶𝐴,𝑠 − 𝐶𝐴 )
𝜆𝑒
where 𝑇𝑠 is the temperature on the outer surface of the particle, 𝑇𝑠,𝑠 is the surface of a pore in the particle,
and 𝜆𝑒 is the effective thermal conductivity in the pore. The large possible temperature difference is when
𝐶𝐴 → 0. In this case we can say that
(Δ𝑇𝑠 ) 𝐷𝑒𝐴 (−Δ𝐻)𝐶𝐴,𝑠
= ≡𝛽
𝑇𝑠,𝑠 𝜆𝑒 𝑇𝑠,𝑠
where 𝛽 is the reaction heat parameter.
9.4.4 THIN-FILM DIFFUSION REACTION

We now want to know how a reaction can speed up the mass transfer, which is incorporated in an
enhancement factor, 𝐹. Consider the following concentration profile for diffusion and reaction into a thin
film from the bulk. The leftmost side represents the bulk gas phase whereas the rightmost side represents
the bulk liquid phase. The length 𝑦𝑔 is the gas film length, and 𝑦𝐿 is the film length. In the film, Henry’s
Law applies, and 𝑃𝐴,𝑖 = 𝐻𝐶𝐴,𝑖 .
𝑃𝐴
𝑃𝐴,𝑖
𝐶𝐴,𝑖
𝐶𝐴
𝐶𝐴,b
𝑦𝑔 𝑦𝐿
We will define the following gas film and liquid film mass transfer coefficients
𝐷𝐴,𝑔 𝐷𝐴,𝐿
𝑘𝑔 = , 𝑘𝐿 =
𝑦𝑔 𝑦𝐿
For a pseudo first order, irreversible reaction,
𝑦 𝑦
𝐶𝐴,𝑖 sinh (𝛾 (1 − 𝑦 )) + 𝐶𝐴,b sinh (𝛾 𝑦 )
𝐿 𝐿
𝐶𝐴 =
sinh(𝛾)
where 𝛾 is the Hatta parameter defined as
𝑘 √𝑘𝐷𝐴,𝐿
𝛾 ≡ 𝑦𝐿 √ =
𝐷𝐴,𝐿 𝑘𝐿
The flux at the interface can be given by

𝛾 𝐶𝐴,𝑏 1
𝑁𝐴,𝑖 = (1 − )𝑘 𝐶
tanh(𝛾) 𝐶𝐴,𝑖 cosh(𝛾) 𝐿 𝐴,𝑖
The effectiveness factor is

𝑁𝐴,𝑖 𝐴𝑣
𝜂𝐿 =
𝑘𝐶𝐴,𝑖
where 𝑁𝐴,𝑖 is the flux at the interface, 𝐴𝑣 is the surface area per volume, and 𝑘𝐶𝐴,𝑖 is the reaction rate at the
interface. If we define the Sherwood number as
𝑘𝐿
Sh =
𝐴𝑣 𝐷𝐴
It can be shown that the effectiveness factor is
1 𝐶𝐴,𝑏 1
𝜂𝐿 = (1 − )
Sh γ tanh(𝛾) 𝐶𝐴,𝑖 cosh(𝛾)
For very fast reactions (𝛾 > 5 or so and 𝐶𝐴,𝑏 → 0), we get
1 𝐷𝐴
𝜂𝐿 = = 𝐴𝑣 √
𝛾 Sh 𝑘
We can also define the enhancement factor as

𝑁𝐴,obs 𝛾 𝐶𝐴,𝑏 1
𝐹= = (1 − )
𝑘𝐿 (𝐶𝐴,𝑖 − 𝐶𝐴,𝑏 ) tanh(𝛾) 𝐶𝐴,𝑖 cosh(𝛾)
For very fast reactions,

𝛾
𝐹= ≈𝛾
tanh(𝛾)

Hemical Ngineering Inetics: Based On CHEM - ENG 408 at Northwestern University

Uploaded by

Copyright:

Available Formats

Hemical Ngineering Inetics: Based On CHEM - ENG 408 at Northwestern University

Uploaded by

Document Information

Original Title

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

Hemical Ngineering Inetics: Based On CHEM - ENG 408 at Northwestern University

Uploaded by

Copyright:

Available Formats

CHEMICAL ENGINEERING KINETICS

Based on CHEM_ENG 408 at Northwestern University

6.1 Introduction to Reaction Networks ............................................................................................. 25

such that the mole balance can be written as

1.2 BATCH REACTOR

1.3 CONTINUOUS-STIRRED TANK REACTOR

2.2 VAN’T HOFF EQUATION

The Van’t Hoff equation can be written as

𝑑(ln 𝑘𝑓 ) 𝑑(ln 𝑘𝑟 ) Δ𝐻rxn

𝑑(ln 𝑘𝑓 ) 𝑑(ln 𝑘𝑟 ) Δ𝐻𝑓 Δ𝐻𝑟

3.1.2 FIRST ORDER KINETICS

3.1.3 N-TH ORDER KINETICS

3.1.4 REVERSIBLE REACTIONS

This can be rearranged to

To simplify the algebra

This implies that

3.1.5 NON-INTEGER RATE LAWS

3.2 DETERMINING THE RATE LAW AND RATE CONSTANTS

3.2.2 INTEGRAL METHOD

3.2.3 REGRESSION METHOD

3.2.4 WORKING WITH PSEUDO ORDERS

𝑘1st ≡ 𝑘rxn 𝐶𝐴𝑛−1

4 WORKING WITH MULTIPLE ELEMENTARY STEPS

4.2 SYNTHESIS OF HBr

4.2.2 RATE EXPRESSIONS

The overall balance for HBr is

4.2.3 THE PSEUDO-STEADY STATE HYPOTHESIS

4.2.4 BOND DISSOCIATION ENERGIES

4.3 COMPLEX REACTIONS: CRACKING

The radical that is produced from a 𝛽-scission reaction is called a 𝛽 species.

4.3.2 MECHANISM AND MASS-ACTION KINETICS

Writing out the mass action expressions:

4.3.4 SIMPLIFICATIONS BASED ON STATISTICAL TERMINATION

4.3.6 ETHANE CRACKING RATE LAW

4.3.7 DETERMINING OBSERVED ACTIVATION ENERGIES

4.3.8 GENERAL OVERVIEW OF SIMPLIFICATION PROCESS

4.3.9 COMPLEX REACTIONS: ADDITIVES

4.4 COMPLEX REACTIONS: RADICAL CHAIN AUTOXIDATION

0 = 𝑘1 [RO ∙][RH] + 𝑘2 [OH][RH] − 2𝑘𝑡 [ROO ∙]2

We note that from 𝑑[OH ∙]/𝑑𝑡

0 = 2𝑘𝑑 [ROOH] − 2𝑘𝑡 [ROO ∙]2

5 CONSEQUENCES OF CHEMICAL EQUILIBRIA

5.2 THE EQUILIBRIUM CONSTANT

where 𝑎𝑖 is the activity of a given species.

The activity equilibrium constant can be related to Δ𝐺 by

5.2.2 ACTIVITIES FOR GASES

where 𝛿 is the change in (stoichiometric) moles.

5.2.3 ACTIVITIES FOR LIQUIDS

5.2.4 RELATIONSHIP BETWEEN VARIOUS EQUILIBRIUM CONSTANTS

For an ideal gas, the following relationships hold:

5.3 ENZYME KINETICS

5.3.2 PLOTTING MICHAELIS-MENTEN DATA

The equation can also be linearized in the following form

5.3.3 REVERSIBLE PRODUCT BINDING

5.3.4 COMPETITIVE INHIBITION

5.3.5 NON-COMPETITIVE INHIBITION

The site balance can be written as

where the new constant 𝐿 is given by

Order Rank = 1 Rank > 1