CT-CH 1-Lecture Slides For Academic Version

CT Chapter 1
Basic Thermodynamics
1.1 First law of thermodynamics
States that energy can be neither created nor

destroyed
Concerns changes in the internal energy – not the

absolute value
If the system receives an amount of heat dQ and if an

amount of work dW is performed on the system, then
the internal energy has increased by dQ+dW i.e.
dU=dQ+dW
We consider only pressure-volume work, i.e. dW=-PdV
dU = dQ − PdV
1 Computerized Thermodynamics for Materials Scientists and Engineers

Hillert and Selleby, 2018
1.1 First law of thermodynamics
dU = dQ - PdV
Under constant volume (i.e. dV=0 ): dQ = dU
Under constant pressure (rewrite):
dQ = dU + PdV
dQ = dU + PdV + VdP − VdP = dU + d ( PV ) − VdP
dQ = d (U + PV ) − VdP = dH
i.e. dQ=dH
where we have introduced enthalpy, H=U+PV
In thermodynamics we cannot distinguish between

heat and work
- they are just different ways to transfer energy

1.2 Second law of thermodynamics
Introduces the quantity entropy, S. It requires that a
process or reaction occurring spontaneously inside a
system must increase S of the system if there is no
heat exchange with the surroundings.
d ip S > 0 for spontaneous internal processes
An internal process will be infinitely slow at d ip S = 0

and may be regarded as a reversible process
A heat exchange itself will change the entropy by dQ / T
dQ dQ
Total change of S: dS = + d ip S >
T T
1.2 Second law of thermodynamics
∆Q
Tb Ta
Heat only goes spontaneously from high to low temperature
∆Q ∆Q ∆Q(Tb − Ta )
∆Stot =− + = > 0 where Tb > Ta
Tb Ta TaTb
Entropy produced!

1.3 Combined law
1st law dU = dQ − PdV
2nd law rewritten TdS = dQ + Td ip S
Combined law TdS = dU + PdV + Td ip S
Rewrite − Td ip S = dU + PdV − TdS = d (U + PV − ST ) − VdP + SdT
Introduce G = U + PV − TS
Rewrite dG = VdP − SdT − Td ip S < VdP − SdT
At constant P and T: dG = −Td ip S < 0
since dipS always
or − dG = Td ip S > 0 positive

1.4 Driving force and dissipation
Equilibrium condition
(∂G / ∂ξ ) P ,T = 0
G
G = G ( P, T , ξ ) (ξ is the extent of
the internal process)
The slope, the rate of
decrease of G may be regarded
as the driving force, D.
D = −(∂G / ∂ξ )T , P = Td ip S / dξ 0 ξeq ξ
FIG.1.1.
Identify D with TdipS/dξ i.e.
TdipS=Ddξ in previous eqs.
dG = VdP − SdT − Ddξ

Equilibrium condition
D = −(∂G / ∂ξ )T , P = Td ip S / dξ
For a phase transformation between two states,
α −> β, one obtains by integration
D = −∆Gm where ∆Gm = Gmβ − Gmα

Internal process
What is an internal process?
Everything that happens spontaneously inside a

system as the system approaches equilibrium,
i.e. processes that are not ”directly controllable”.
Example: solidification of a supercooled liquid.

Internal process – Example 1
Imagine for instance isothermal solidification of a
pure supercooled liquid metal. The only thing that
changes during the process is the amounts of solid
and liquid metal. Let ξ denote the number of moles
of solid metal, n(solid).
External variable = can be directly controlled during an experiment

Internal variable = property of the system that adjusts as equilibrium
is approached and cannot be directly controlled from the outside

The combined law: dG = VdP − SdT − Ddξ
During the solidification P and T are constant
and we get D = -dG/dξ = -∆G/∆ξ = -∆G/∆Nsol
The process:
liquid  solid ∆G = Gsol-Gliq
and the driving force:

D = (Gliq-Gsol)/∆Nsol = Gmliq - Gmsol

Here an example where -97

FCC
Gm for solid and liquid LIQUID
-98
Ni have been calculated.
-99
Gm(kJ)
Solidification: liq  fcc -100
D = Gmliq − Gmfcc -101
-102 The lines cross at the

melting temperature.
i.e. positive below Tm
-103
and negative above 1700 1725 1750
T

Here same calculation but 300

with D on the y-axis. 200
D=Gm(liq)-Gm(fcc)
100
-100
-200
-300
1700 1725 1750
T

T, V1 and V2 are fixed.
Suppose we blow in nA moles of component A, what are
then the equilibrium content of nA’ and nA’’ ?
A
nA’ nA’’
V1 V2
External variables are T, V1, V2 and nA and the

internal variables are nA’ and nA’’.

The internal variable nA’ and nA’’ are not independent:
nA = nA’ + nA’’
We introduce: ξ = nA’ and get nA’’ = nA - ξ
At equilibrium: P’ = P’’
The gas law, PV = nRT, gives
P’ = ξRT/V1 = P’’=(nA-ξ)RT/V2
ξ/V1 = (nA-ξ)/V2  ξ = nA/(1+V2/V1)

Reversible process and frozen process
dG = VdP − SdT − Ddξ

An internal process occurring under D=0 would not
produce any entropy or dissipate any Gibbs energy
and it would be infinitely slow (sometimes called a
reversible process – does not exist)
For a frozen-in process dξ = 0
In both cases: dG = VdP − SdT

1.5 Variable composition
Consider systems where also the content may
change by exchange of matter with the
surroundings, then the combined law becomes:
dG = VdP − SdT + ∑ µi dN i − Ddξ
V = (∂G / ∂P) T , N i ,ξ S = −(∂G / ∂T ) P , N i ,ξ
µ k = (∂G / ∂N k ) P ,T , N j ,ξ D = −(∂G / ∂ξ ) P ,T , N i
If no internal entropy production, we get
dG = VdP − SdT + ∑ µi dN i
and if P, T and Nj are constant
µ k = (∂G / ∂N k ) P ,T , N j

External variables and potentials
µk is the partial Gibbs energy of component k
P, T, µk, V, S, Ni are all external variables
P, T, µk are also potentials as they must have

uniform values in the whole system at equilibrium
Specifically, µk is the chemical potential of k

Chemical potential
Using x k ≡ N k / ΣN i = N k / N we can get

dG = VdP − SdT + (Σµi xi )dN − Ddξ
Consider a large amount of homogenous matter
with uniform P, T and composition and let the
system be a very small volume of it. Extend its
limits gradually i.e. integrate
G = (Σµi xi ) N = Σµi N i
the molar Gibbs energy is defined as
Gm ≡ G / N = Σµi xi

1.5 Variable Composition
Definition of chemical potential illustrated
Definition: µ k = (∂G / ∂N k ) P ,T , N j
Binary case: µ B = (∂G / ∂N B ) P ,T , N A
GB ≡ µ B = Gm + x A
dGm Gm .G o
B
dxB
.µ
G A ≡ µ A = Gm − xB
dGm oG . . .G (x )
B
(1 − xB )
dGm
µ .
A
dxB m B
dxB
A
xB 1-xB= xA
A xB B
FIG.1.2.

1.6 Expressing chemical potentials
through the molar Gibbs energy
Binary system
We may derive the following equation for the slope of
the tangent in Fig.1.2:
µB − µ A = Gm .G o
B
 ∂G ∂G   ∂G ∂G 
Gm + x A  m − m  − Gm − xB  m − m  = .µ
 ∂xB ∂x A   ∂x A ∂xB  oG . . .G (x )
B
(1 − xB )
dGm
µ .
A
dxB
∂Gm ∂Gm dGm m B
− = A
∂xB ∂x A dxB xB 1-xB= xA
A xB B
FIG.1.2

Multicomponent system
Chemical potential of B
∂Gm ∂Gm
G B ≡ µ B = Gm + − ∑ xi
∂x B ∂xi
may also be regarded as the partial molar Gibbs
energy of component B.
Generally we can express any partial molar quantity:
 ∂A  ∂A ∂A
Ak ≡   = Am + m − ∑ xi m
 ∂N k  P ,T , N j ∂x B ∂xi
e.g. if A is exchanged with H we get the partial enthalpy

of component k usually called heat of solution of
component k
Heat of solution – Example 1
Enthalpy versus mole fraction Pt in an Fe-Pt fcc solution

at 700 K
Reference phase for both
components is fcc.
Enthalpy of solution may be
calculated as
dH m
H Pt = H m + xFe
dxPt
or
 ∂H 
H Pt =  
 ∂N Pt  P ,T , N Fe

Two phase equilibrium – common tangent
construction
Phases at equilibrium must have the same values for
T, P and all chemical potentials.
Gm
α
β
o
G Aα .. .G
µ α+ β
A
. . .µ
o
B
β
α+ β
B
A xαB / β xBβ / α B
FIG.1.3.

Stoichiometric phase, ϕ
Here the tangents can be drawn arbitrarily i.e. the
chemical potentials are not define for such a phase
alone. Gm ϕ µ"
B
µA'
Gmϕ
µ" µ 'B
A
xϕB xϕA
A xϕB B
FIG.1.4.

1.7 Gibbs-Duhem relation
Different possiblities (pairs) for the chemical potentials

in Fig. 1.4. Can express the molar Gibbs energy of ϕ
using any pairs and the relation
Gmϕ = Σµiϕ xiϕ = µ ϕA xϕA + µ Bϕ xϕB
and since the xi:s and Gm are constants we get
dGmϕ = xϕA dµ ϕA + xϕB dµ Bϕ = 0
Can be proven that it holds also for solution phases

with small compositional variations. The general form is
∑ xi dµi − Vm dP + S m dT = 0

1.8 Phases
Phases
• are often denoted by a Greek letter e.g. α, β, γ.
• are identified by their structure (not composition)
• may be
• stoichiometric i.e. fixed composition e.g. Al2O3, CaO
• line compounds e.g. Al2O3 – Fe2O3
• solution phases e.g. fcc, liquid
In thermodynamic databases phases usually have

generic phase names, named by their structure e.g.
FCC_A1, BCC_A2, BCC_B2.

Crystallographic data
A1 B1 L10 L12
A2 B2 D03 L21

1.9 Gibbs phase rule
In Gibbs-Duhem relation
∑ xi dµi − Vm dP + S m dT = 0
there are c+2 terms (c =# of independent components),
each term consisting of one extensive quantity and one
potential. The two quantitites in a pair are conjugate
variables.
In a one phase system we may vary the potentials in
c+2-1 ways since G-D gives a relation between the
potentials.
In a system with p phases we get the variance, v, (or
degrees of freedom): v = c + 2 - p

1.9 Gibbs phase rule
Example
Gibbs phase rule was derived for potentials and can
thus only be applied to phase diagrams with potential
axes, e.g. P vs T or T vs µi or as below, T vs activity:
Ag-Cu system at constant P.

c=2
v = c+1- p (c+1 since P const.)
p=1: v = 2 + 1 – 1 = 2 (area)
p=2: v = 2 + 1 – 2 = 1 (line)
p=3: v = 2 + 1 – 3 = 0 (point)
FIG 1.5.

1.10 Introduction of new components
Sometimes convenient to use other components than

the elements and at equilibrium we have
µ A B = aµ A + bµ B
a b
For the H-O system we might want H2O

µ H O = 2µ H + µO
2
or expressed differently
µ H O = µ H + 0.5µO
2 2 2

1.10 Introduction of new components
For the reaction H 2 + 0.5O 2 → H 2 O we have,

before equilibrium has been established, a driving
force, D
D = µ H 2 + 0.5µO2 − µ H 2O
Any new set of components may be used, but the

number of independent components never changes.

1.11 Gibbs energy of formation
The change of the molar Gibbs energy when one mole
of phase ϕ is formed from the correct amounts of A
and B. A and B are insoluble in one another.
∆Gm = Gmϕ − xϕA µ A − xϕB µ B = Gmϕ − xϕA oG αA − xϕB oGBβ
The driving force, D = − ∆Gm , is here positive.
FIG 1.6.

Standard Gibbs energy of formation
A compound is usually give as AaBb and Gibbs energy is
given per mole of formula units
G ϕAa Bb = (a + b)Gmϕ
and the change per formula unit becomes
∆ f o G ϕAa Bb = o G ϕAa Bb −a o G αA −b o G Bβ = −(a + b) D
o denotes pure component (i.e. pure element or

stoichiometric compound)
∆ f oG ϕAa Bb is the standard Gibbs energy of formation

of ϕ if pure α and β have been chosen as the standard
states for A and B, respectively, at the same P and T.

Precipitation from a solution phase
When ϕ precipitates from a solution phase the
maximum driving force is obtained with a parallel
tangent construction
FIG 1.7a.
D = x ϕA µ αο
A + x ϕ αο
µ
B B − G ϕ
m
(a + b) D = a ( µ αο
A − o α
G A ) + b ( µ αο o β
B − G B ) − ∆ o ϕ
f G Aa Bb

Driving force for the complete process
The driving force decreases as the process proceeds
and will reach zero when equilibrium is reached.
The total, integrated or average driving force is the
total change in molar Gibbs energy between the final
state and the initial state
FIG 1.7b.

1.12 Properties of the Gibbs energy
function
All thermodynamic properties can be expressed through

its Gibbs energy, e.g.
V = (∂G / ∂P)T , N i
S = −(∂G / ∂T ) P , Ni
H ≡ U + PV = G + TS = G − T (∂G / ∂T ) P , N i = (∂ (G / T ) / ∂ (1 / T )) P , N i

function
Heat capacity
Heat capacity is defined as the capability of a system to
receive heat under a given increase of T, dQ/dT
Under constant volume: dQ = dU

 ∂Q   ∂U 
CV ≡   = 
 ∂T V  ∂T V
Under constant pressure: dQ=dH
 ∂Q   ∂H 
CP ≡   = 
 ∂T  P  ∂T  P

function
Isothermal compressibility and thermal
expansion
We can also obtain κT and α from Gibbs energy
κ T = −(∂V / ∂P) T , N / V = −(∂ 2 G / ∂P 2 ) T , N /(∂G / ∂P) T , N

i i i
α = (∂V / ∂T ) P , N / V = (∂ 2 G / ∂T∂P) N /(∂G / ∂P) T , N

i i i
Such important properties can be determined

experimentally and are stored in tables in books.
Nowadays one instead stores G(P,T,Ni) in

thermodynamic databases and obtain all
quantities discussed above by computer calculations.

1.13 Adiabatic changes
During an adiabatic process dQ = 0

During a reversible process dξ = 0
We may rearrange the second law into

dS = dQ / T + d ip S = dQ / T + (1 / T ) Ddξ
and for an adiabatic reversible process we thus obtain

dS = 0, i.e. an isentropic process.

1.14 State of reference and standard
state
Model-based reference
Chemical potentials of pure components are usually given as
o α
G A i.e. molar Gibbs energy of pure (o) A in the same
crystal structure as the solution (α). oG αA is thus the
reference state for A in the α-solution. We may call this a
model-based reference.
Comparing the chemical potential of A in α and β one
α
µ
should remember that A is unaffected by the choice of
reference
µ αA = oG αA + f (comp)= oG Astd +( oG αA − oG Astd ) + f (comp)

1.14 State of reference and standard
state
State of reference
Different choices e.g.
GkNPT = Gibbs energy of the equilibrium state at 25 oC and

1 atm for element k. NPT = Normal P and T
H kSER= enthalpy of the equilibrium state at 25 oC and 1 bar

and entropy at 0 K which by agreement is set to zero, for
the element k.
SER = Standard Element Reference

1.15 Duhem’s theorem
For a closed system, i.e. all Ni are constant, the

equilibrium state is uniquely defined by giving P and T
(or their conjugate variables) regardless of how many
phases are involved. We thus have to give c+2
conditions to define the equilibrium.

1.16 Characteristic state function and
Gibbs energy model
The characteristic state function for fixed P, T and Ni
is Gibbs energy.
In order to use it we need to know how it varies for all

the different phases as function of P, T, Ni and internal
variables
Gα ( P, T , N iα , ξ1 , ξ 2 ...)
Such an analytical expression is regarded as a
thermodynamic model from which all thermodynamic
information may be obtained.


CT-CH 1-Lecture Slides For Academic Version

Uploaded by

Copyright:

Available Formats

CT-CH 1-Lecture Slides For Academic Version

Uploaded by

Document Information

Original Description:

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

CT-CH 1-Lecture Slides For Academic Version

Uploaded by

Copyright:

Available Formats

CT Chapter 1

States that energy can be neither created nor

Concerns changes in the internal energy – not the

If the system receives an amount of heat dQ and if an

We consider only pressure-volume work, i.e. dW=-PdV

1 Computerized Thermodynamics for Materials Scientists and Engineers

In thermodynamics we cannot distinguish between

2 Computerized Thermodynamics for Materials Scientists and Engineers

d ip S > 0 for spontaneous internal processes

An internal process will be infinitely slow at d ip S = 0

A heat exchange itself will change the entropy by dQ / T

Heat only goes spontaneously from high to low temperature

4 Computerized Thermodynamics for Materials Scientists and Engineers

2nd law rewritten TdS = dQ + Td ip S

Combined law TdS = dU + PdV + Td ip S

Rewrite − Td ip S = dU + PdV − TdS = d (U + PV − ST ) − VdP + SdT

5 Computerized Thermodynamics for Materials Scientists and Engineers

dG = VdP − SdT − Ddξ

6 Computerized Thermodynamics for Materials Scientists and Engineers

D = −∆Gm where ∆Gm = Gmβ − Gmα

7 Computerized Thermodynamics for Materials Scientists and Engineers

Everything that happens spontaneously inside a

Example: solidification of a supercooled liquid.

8 Computerized Thermodynamics for Materials Scientists and Engineers

External variable = can be directly controlled during an experiment

9 Computerized Thermodynamics for Materials Scientists and Engineers

and the driving force:

10 Computerized Thermodynamics for Materials Scientists and Engineers

Here an example where -97

D = Gmliq − Gmfcc -101

-102 The lines cross at the

11 Computerized Thermodynamics for Materials Scientists and Engineers

Here same calculation but 300

12 Computerized Thermodynamics for Materials Scientists and Engineers

External variables are T, V1, V2 and nA and the

13 Computerized Thermodynamics for Materials Scientists and Engineers

The gas law, PV = nRT, gives

ξ/V1 = (nA-ξ)/V2  ξ = nA/(1+V2/V1)

14 Computerized Thermodynamics for Materials Scientists and Engineers

dG = VdP − SdT − Ddξ

In both cases: dG = VdP − SdT

15 Computerized Thermodynamics for Materials Scientists and Engineers

16 Computerized Thermodynamics for Materials Scientists and Engineers

µk is the partial Gibbs energy of component k

P, T, µk, V, S, Ni are all external variables

P, T, µk are also potentials as they must have

Specifically, µk is the chemical potential of k

17 Computerized Thermodynamics for Materials Scientists and Engineers

Using x k ≡ N k / ΣN i = N k / N we can get

18 Computerized Thermodynamics for Materials Scientists and Engineers

Binary case: µ B = (∂G / ∂N B ) P ,T , N A

19 Computerized Thermodynamics for Materials Scientists and Engineers

20 Computerized Thermodynamics for Materials Scientists and Engineers

e.g. if A is exchanged with H we get the partial enthalpy

Enthalpy versus mole fraction Pt in an Fe-Pt fcc solution

22 Computerized Thermodynamics for Materials Scientists and Engineers

23 Computerized Thermodynamics for Materials Scientists and Engineers

24 Computerized Thermodynamics for Materials Scientists and Engineers