Notes for

Chemical Engineering 623

APPLICATIONS OF
THERMODYNAMICS TO CHEMICAL
ENGINEERING

James C. Holste

Copyright © 2021 James C. Holste, All rights reserved

 ACKNOWLEDGMENT

Professors Kenneth R. Hall, Philip T. Eubank and David M. Ford made substantial contributions to the
philosophy and content of this course. I am most grateful to them for their contributions.

TABLE OF CONTENTS

1 INTRODUCTION ............................................................................................................... 1
1.1 Basic Definitions for Thermodynamics ................................................................................. 1

2 THERMODYNAMIC LAWS............................................................................................. 7
2.1 ZEROTH LAW OF THERMODYNAMICS ........................................................................ 8
2.2 FIRST LAW OF THERMODYNAMICS ........................................................................... 12
2.2.1 Single Particle ............................................................................................... 13
2.2.2 Collections of Particles ................................................................................. 19
2.2.3 Consequence of the First Law of Thermodynamics ..................................... 30
2.2.4 General Energy Balance ............................................................................... 31
2.3 SECOND LAW OF THERMODYNAMICS ...................................................................... 35
2.3.1 Physical Observations ................................................................................... 35
2.3.2 Formal Statements of the Second Law ......................................................... 36
2.3.3 Consequences Of Second Law...................................................................... 46
2.3.4 Summary Statement of Second Law ............................................................. 66
2.3.5 Irreversible Processes: .................................................................................. 67
2.3.6 Practical Applications - Efficiencies ............................................................. 69
2.3.7 Summary Statement of Combined First and Second Laws .......................... 74
2.4 THIRD LAW OF THERMODYNAMICS .......................................................................... 75
2.4.1 Nernst-Simon Statements of Third Law ....................................................... 75
2.4.2 Lewis-Randall Statement of Third Law ........................................................ 75
2.4.3 Consequences of Third Law ......................................................................... 76
2.4.4 Practical Thermometry.................................................................................. 78
2.5 MICROSCOPIC INTERPRETATIONS: TEMPERATURE, ENTROPY ........................ 90
2.5.1 Microscopic (Statistical) Viewpoint of Entropy ........................................... 90
2.5.2 Microscopic (Statistical) Viewpoint of Energy and Temperature .............. 103
2.6 MATHEMATICAL IMPLICATIONS OF THERMODYNAMIC LAWS ...................... 109
2.6.1 Functions ..................................................................................................... 109
2.6.2 Multivariable calculus ................................................................................. 114

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) i

 2.6.3 Applications of Mathematics to Physics Summary .................................... 124
2.6.4 Property Change Calculations..................................................................... 135
2.6.5 Equilibrium Criteria .................................................................................... 213
2.6.6 Comments On Energy Functions ................................................................ 237
2.6.8 Summary ..................................................................................................... 243

3 PROPERTIES OF PURE FLUIDS ................................................................................ 244

3.1 IDEAL GAS HEAT CAPACITIES................................................................................... 245
3.2 EQUATIONS OF STATE ................................................................................................. 247
3.2.1 General P, V, T Behavior ............................................................................ 247
3.2.2 Brief History of Equations of State ............................................................. 250
3.2.3 Important Equations of State ...................................................................... 252
3.2.4 Corresponding states ................................................................................... 268
3.3 MOLECULAR INTERACTIONS .................................................................................... 275
3.3.1 The Fundamentals ....................................................................................... 275
3.3.2 Force-field concept ..................................................................................... 277
3.3.3 Intramolecular potentials ............................................................................ 278
3.3.4 Improper out-of-plane potential .................................................................. 280
3.3.5 Van der Waals and Coulombic potentials ................................................... 282
3.3.6 Three-body potentials ................................................................................. 289
3.3.7 Equations Derived From Interatomic Potentials ......................................... 292
3.4 PURE FLUID PHASE EQUILIBRIUM ........................................................................... 305
3.4.1 Gibbs Energy of van der Waals Fluid ......................................................... 308
3.4.2 Vapor pressure from EOS ........................................................................... 312
3.4.3 Temperature dependence of DsH ................................................................ 317
3.4.4 Trouton's Rule ............................................................................................. 319
3.4.5 Types of phase transitions ........................................................................... 320
3.5 BEHAVIOR NEAR CRITICAL POINTS ........................................................................ 326
3.5.1 P-V-T near a vapor - liquid critical point: .................................................. 327
3.5.2 Physical Observations: ................................................................................ 328
3.5.3 Critical Opalescence ................................................................................... 329
3.5.4 Universality ................................................................................................. 329
3.5.5 General Behavior ........................................................................................ 330
3.5.6 Law of Rectilinear Diameters ..................................................................... 331
3.5.7 Critical Exponents ....................................................................................... 332
3.5.8 Experimental Problems Near Critical Point ................................................ 334

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) ii

4 PROPERTIES OF MIXTURES ..................................................................................... 336
4.1 GENERAL FORMAL DESCRIPTIONS .......................................................................... 336
4.1.1 Gibbs-Duhem equation ............................................................................... 340
4.1.2 Calculation of partial molar properties ....................................................... 343
4.1.3 Property change on mixing ......................................................................... 347
4.2 PHASE EQUILIBRIUM FOR MIXTURES ..................................................................... 373
4.2.1 General Phase Equilibrium ......................................................................... 373
4.2.2 Phase Equilibrium Calculation ................................................................... 378
4.2.3 Equilibrium Relationships .......................................................................... 384
4.3 CHEMICAL REACTION EQUILIBRIUM ...................................................................... 417
4.3.1 Chemical Reaction (Generalized Form) ..................................................... 417
4.3.2 Reaction coordinate .................................................................................... 418
4.3.3 Equilibrium criteria ..................................................................................... 421
4.3.4 Standard Property Change of Reaction ....................................................... 426
4.3.5 Standard and Reference State Choices ....................................................... 434
4.3.6 Reaction Equilibrium by Minimization of Gibbs Free Energy ................... 450
4.3.7 Reaction Equilbrium Using Reaction Coordinate....................................... 458
4.3.8 Heat Effects Associated with Reactions ..................................................... 461
4.4 IONIC SOLUTIONS ......................................................................................................... 464
4.4.1 Molality ....................................................................................................... 464
4.4.2 Summary of Conventions for Activities ..................................................... 468
4.4.3 Free Energy of Formation in Solution ........................................................ 469
4.4.4 Electrolytes ................................................................................................. 471
4.4.5 Equilibrium Involving Ionic Solutions ....................................................... 478

5 SURFACE EFFECTS ..................................................................................................... 496

5.1 SURFACE ENERGIES ..................................................................................................... 496
5.1.1 Surface Tension .......................................................................................... 500
5.2 APPLICATION OF FIRST AND SECOND LAWS ........................................................ 502
5.3 EQUILIBRIUM CONDITIONS ....................................................................................... 505
5.4 PHASE EQUILIBRIUM ................................................................................................... 506
5.4.1 Case I: Small liquid drops in a vapor (pure fluid) ..................................... 506
5.4.2 Case II: Small Vapor Bubbles Surrounded By a Liquid (Pure Fluid) ....... 509
Correlations for Surface Tension............................................................................................ 512
5.5 ADSORPTION ON SURFACES ...................................................................................... 513
5.5.1 Terminology ................................................................................................ 514
5.5.2 Two Dimensional Phase ............................................................................. 519

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) iii

 5.5.3 Adsorption Isotherms .................................................................................. 522
5.5.4 Heat of Adsorption ...................................................................................... 524

6 EXTERNAL FIELDS ...................................................................................................... 526

6.1 Gravitational Fields ........................................................................................................... 532
6.1.1 Hydrostatic Pressure Gradient .................................................................... 533
6.1.2 Chemical Potential Gradient ....................................................................... 534
6.2 Centrifugal Force Fields .................................................................................................... 535
6.2.1 Centrifugal Pressure Gradient ..................................................................... 535
6.2.2 Centrifugal Chemical Potential Gradient .................................................... 536
6.3 Electric Fields .................................................................................................................... 537
6.4 Magnetic Fields ................................................................................................................. 537

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) iv

First Law (Differential and Integral Forms)

d[ m(U + E k + E p )] system = Σ( dQ ) + Σ( dW ) + Σ( dm ) ⎛H + Eik + Eip ⎞ − ( dm ) c 2

t
Diff. form:
i i i⎝ ⎠ n

Δ[ m(U + E + E )] system = ΣQi + ΣWi + Σ (δt m) i ⎝H + Ei + Ei ⎠ − (δ n m) c

Integral form: k p ⎛ k p⎞ 2

dQrev Q
Second Law: d(nS) =
T
; Δ(nS) universe ≥ 0 ; ΔSsurr = −
total

Tsurr
− ∑ ni Si
all
streams

Combined First and Second Law (Differential Forms):

dU = TdS − PdV d(nU) = Td( nS ) − Pd(nV ) + ∑ µi dni

dH = TdS + VdP d(nH) = Td( nS ) + (nV )dP + ∑ µi dni

dA = −SdT − PdV d(nA) = −(nS )dT − Pd(nV ) + ∑ µi dni

dG = −SdT + VdP d(nG) = −(nS)dT + (nV )dP + ∑ µi dni

Energy Functions: H = U + PV
A = U − TS
G = U + PV − TS = H − TS = A + PV
Useful Mathematical Relationships:

⎛∂f ⎞ ⎛∂f ⎞ ⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂w ⎞
MR1. f = f ( x, y ) ↔ df = ⎜ ⎟ dx + ⎜ ⎟ dy MR4. ⎜ ⎟ = ⎜
⎝ ∂x ⎠ y ⎝ ∂y ⎠ x ⎝ ∂x ⎠ y ⎝ ∂w ⎟⎠ y ⎜⎝ ∂x ⎟⎠ y
−1
∂z ⎡ ∂x ⎤ ⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂w ⎞
MR2. ⎛⎜ ⎞⎟ = ⎢⎛⎜ ⎞⎟ ⎥ MR5. ⎜ ⎟ = ⎜ ⎟ + ⎜
⎝ ∂x ⎠ y ⎣⎝ ∂z ⎠ y ⎦ ⎝ ∂x ⎠ y ⎝ ∂x ⎠ w ⎝ ∂w ⎟⎠ x ⎜⎝ ∂x ⎟⎠ y

⎛ ∂z ⎞ ⎛ ∂x ⎞ ⎛ ∂y ⎞ ⎛ ∂2 f ⎞ ⎛ ∂2 f ⎞
MR3. ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ = −1 MR6. ⎜ =
⎝ ∂x ⎠ y ⎝ ∂y ⎠ z ⎝ ∂z ⎠ x ⎝ ∂x ∂y ⎟⎠ ⎜⎝ ∂y ∂x ⎟⎠

Residual Functions:
Ψ r = Ψ (T , ρ ) − Ψ ig (T , ρ ) Ψ R = Ψ (T , P ) − Ψ ig (T , P )
ρ ρ
⌠
Ur UR 1 ⎛ ∂Z ⎞ d ρ ⌠ ∂Z
⎛ ⎞ dρ Hr HR
= = I1 = ⎮⎮ ⎜ ⎟ = −T ⎮⎮ ⎜ ⎟ = I1 + Z − 1 =
RT RT T ⎮⌡ ⎝ ∂(1 / T ) ⎠ ρ ρ ⌡ ⎝ ∂T ⎠ ρ ρ RT RT
0 0
ρ
Ar AR ⌠ dρ Gr GR
= + ln Z = I 2 = ⎮⎮ ( Z − 1) = I2 + Z − 1 = + ln Z
RT RT ⌡ ρ RT RT
0

Sr SR P rV PV R
= I1 − I 2 = − ln Z = Z −1 =
R R RT RT

Property Changes:
Ψ ( 2 ) − Ψ (1) = Ψ R (T2 , P2 ) − Ψ R (T1, P1 ) + Ψ ig (T2 , P2 ) − Ψ ig (T1, P1 ) = Ψ R (T2 , P2 ) − Ψ R (T1, P1 ) + Δ1→2 Ψ ig
= Ψ r (T2 ,V2 ) − Ψ r (T1,V1 ) + Ψ ig (T2 ,V2 ) − Ψ ig (T1,V1 ) = Ψ r (T2 ,V2 ) − Ψ r (T1,V1 ) + Δ1→2 Ψ ig

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 1

Mixture Properties:
⎛ ∂( nM ) ⎞ ⎛ ∂M ⎞ ⎛ ∂M ⎞
M = ∑ xi M i where: Mi = ⎜ ⎟
⎝ ∂ni ⎠ T ,P,n j≠i
= M +⎜ ⎟
⎝ ∂xi ⎠ T ,P,xk≠i
−
j
xj ⎜∑ ⎟
⎝ ∂x j ⎠ T ,P,xk≠ j
Property Changes on Mixing:
Δ †M G
Δ †M M ≡ M − ∑ xi M i† = ∑ xi M i − M i† ( ) RT
= ∑ xi ln α i† = ∑ xi ln γ i† xi

Ideal Solution:
Δ ISM S Δ IS G Δ IS A Δ ISM U Δ ISM H PΔ ISM V
= − M = − M = − ∑ xi ln xi = = =0
R RT RT RT RT RT
E IS
Excess Properties: M ≡ ΔM M − ΔM M
Activity, Activity Coefficient, Fugacity, Fugacity Coefficient:
f̂i xiφ̂i αi f̂ φ̂
α i = γ i xi = = γi = = i† = i
fi † φi† xi xi fi φi
∞
⌠⎡ ⎤ dV
∂( nZ ) ⎞
P
⎮ ⎛
ln φ̂i = ⎮⎮ ( Z i − 1)
⌠ dP
= ⎮⎢ − 1⎥ − ln Z
⎮ ⎢⎜ ⎟
∂ni ⎠ T ,nV ,n
⌡ P ⎮ ⎝ ⎥ V
0 ⌡⎣ j≠i ⎦
V

( )
†
⎛ GE ⎞ GiE ⎛ ∂ nG
E
⎞
⎜⎝ RT ⎟⎠ = ∑ xi ln γ i
†
ln γ i =
RT
=⎜ ⎟
RT ⎝ ∂ni ⎠
T ,P,ni≠ j

Note: The superscript † denotes a reference state, which usually is taken to be the pure material at the same temperature and
pressure as the mixture.
Phase Equilibrium:
β β
Equilibrium Condition: µiα = µi fˆiα = ˆfi
Fugacity coefficient description: fˆ = y φˆ P
i i i
Activity coefficient description: fˆi = xi γ i† fi†
⎡ ⎤
y γ LRφ σ Pσ ⎢ P ⎥
⎢ 1 ⎥
Partition coefficient: K i ≡ i = i Vi i exp ⎢⎢ Vi L dP ⎥⎥
⌠
⎮
⎮

φ̂i P
⌡
xi ⎢ RT
⎢⎣ Piσ ⎥
⎥⎦

Chemical Reaction Equilibrium:

∑ νi Ai = 0 where: ni = stoichiometric coefficient of substance i, (ni < 0 for reactants, ni > 0 for products)

∑ν G i i =0
⎛ f̂ ⎞ ⎛ f̂ ⎞
Gi = Gi ° + RT ln ⎜ i ⎟ = Δ f Gio + RT ln ⎜ i ⎟ RHS: free elements are standard states for all substances
⎝ fi ° ⎠ ⎝ fi ° ⎠
νi
Δ r G° ⎛ f̂ ⎞
Standard Free Energy of Reaction and Equilibrium Constant: = − ln K = − ln ∏ ⎜ i ⎟
RT i ⎝ fi ° ⎠

Standard Property Change of Reaction: Δ r M ° ≡ ∑ ν i M i °

Minimization of Gibbs Energy Using Lagrange Multipliers:

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 2

 ∂F w

( )
m
= Gi + ∑ λ k aik = 0 fk = ∑ nio − ni aik m = number of substances, w = number of elements
∂ni k=1 i=1

Ionic Solutions:

1000ni
mi =
n0 M 0

⎛ f̂ ◊ ⎞
Δ f Gi◊ = Δ f Gio + RT ln ⎜ i o ⎟
⎝ fi ⎠

M ν + Aν − = ν + M z+ + ν − A z−

Δ r G ◊ = −RT ln K e = ν + Δ f G+◊ + ν − Δ f G−◊ − Δ f Gu◊

γ ν± mν±
Ionic Equilibrium Constant: Ke =
γ u mu

Mean Ionic Activity Coefficient, Mean Molality, and Ionic Strength

ν ν
γ ν± = γ + + γ ν−− m ν± = m ++ m ν− −
1
ν = ν+ + ν − I= ∑ mi zi2
2 i
ν
Solubility Product: S.P. = γ ν± m + + m ν−−

ν RT
Galvanic Cells: ∑ν G i i = −zFE Gi = ν + Δ f G+◊ + ν − Δ f G−◊ + ν RT ln γ ± m± E = E° −
zF
ln γ ± m±

I
Extended Debye-Hückel Theory: ln γ ± = −1.174 z+ z−
1+ I
−1
Physical Constants: R = 8.314 J mol ; F = 96, 500 coulombs/(mole electrons)

Pressure Difference Across a Curved Surface: P β = Pα + 2σ

r

⎡VL ⎛ V 2σ ⎞ ⎤ ⎧+ : liquid inside, vapor outside of spherically curved surface

P V ≅ Piσ exp ⎢ σ
⎜⎝ P − Pi ± ⎟ ⎨
⎣ RT r ⎠ ⎥⎦ ⎩ –: vapor inside, liquid outside of spherically curved surface

External Field:
⎡Ξ ⎤
f̂i field = f̂i no field exp ⎢ i ⎥
⎣ RT ⎦
where Xi is the potential energy of molecule i due to the presence of the external field

Gravitational Field:

dP Mg ⎡ M g ( z − z0 ) ⎤
=− ; f̂i ( z ) = f̂i no gravity exp ⎢ i ⎥
dz V ⎣ RT ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 3

 1 INTRODUCTION

1.1 Basic Definitions for Thermodynamics

Thermodynamics: a model of the behavior of collections of particles, based

upon physical observations and mathematical principles, that attempts to
describe the behavior of the universe.

Classical thermodynamics: describes the macroscopic behavior of a

system of particles in terms of macroscopic properties.

Statistical thermodynamics: describes the macroscopic behavior of a

system of particles in terms of microscopic properties (atomic and
molecular interactions).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 1

System: a region of space that is of interest and everything that is contained
within that region.

Surroundings: all parts of the universe not included in the system.

Boundary: separates the system from the surroundings. The most general
definition of a boundary is a mathematical surface.

Diathermal boundaries: allow interaction between the system and the

surroundings.

Adiathermal boundaries: prevent interaction between the system and the

surroundings.

NOTE: The term adiabatic is often used to describe situations where there
is no energy transferred as heat, but other forms of energy transfer
may occur.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 2

Systems

Homogeneous: all thermodynamic properties of a homogeneous system

vary continuously on a macroscopic scale throughout the system.

Heterogeneous: at least one ot the thermodynamic properties varies

discontinuously on a macroscopic scale at one or more points within the
system.

Isolated: there is no interaction of any type between an isolated system and

its surroundings. (An isolated system is surrounded by an adiabatic
boundary.)

Closed: the amount of material (mass) contained within a closed system

remains constant (excluding nuclear reactions and relativistic effects).
No mass transfer occurs to or from a closed system.

Open: the amount of material (mass) within an open system may vary with
time because mass transfer across the system boundary may occur.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 3

State: the condition of a system. States normally are described in terms of the
physical properties of the system.

State variables: any of the physical properties used to describe the state of the
system. Each state of the system has a unique set of state variables.

Intensive properties: properties whose values can be specified at each point in

a system and which may vary from point to point. Intensive properties are
not additive and they do not require any specification of the quantity of
sample involved.

Extensive properties: properties that are additive in the sense that the value of
the property for the whole system is the sum of the values for each of its
constituent parts. For extensive variables, there is a single aggregate value
for the entire system.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 4

Equilibrium: an isolated thermodynamic system is in equilibrium when none of
its thermodynamic properties are changing with time at a measurable rate.
At equilibrium:

dφ
=0
dt

where f represents any thermodynamic property.

Steady state: an open system is said to be at steady-state when none of the

properties of the system are changing at a measurable rate. At steady-
state

⎛ ∂φ ⎞
⎜⎝ ⎟⎠  = 0
∂t x

where f represents any thermodynamic property.

Time: a measure of the order in which events will occur in nature. Time is not a
legitimate variable in classical thermodynamics.

Law: a statement of the behavior of the physical world for which no contradiction
ever has been observed.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 5

Temperature: a mathematical variable that provides a measure of the "hotness"
or "coldness" of a body, i.e., whether energy will be transferred to or from a
body as heat if the body is placed in thermal contact with another body at a
different temperature.

Energy: a mathematical variable that is independent of time for a closed,

isolated system. This function is particularly useful for describing the
physical behavior of systems.

Heat: energy transferred across a system boundary as the result of a

temperature gradient driving force, excluding effects associated with mass
transfer.

Work: energy transferred across a system boundary due to any driving force
other than a temperature gradient, excluding effects associated with mass
transfer.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 6

 2 THERMODYNAMIC LAWS

Law: a description of the physical behavior of the universe for which no

measurable contradiction ever has been observed.

Note: it is impossible to prove conclusively the validity of scientific laws; they

are summary statements of observations of past physical behavior.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 7

2.1 ZEROTH LAW OF THERMODYNAMICS

Statement: If two bodies (closed systems) each are in thermal equilibrium with a
third body, then they are in thermal equilibrium with each other.

Consequence: There exists a mathematical variable (which we call

temperature) that provides a measure of hotness and coldness and that may
be used to determine whether or not two systems are in thermal equilibrium.

Proof:

For simplicity, consider a single-phase fluid so that only two intensive variables
need be specified to specify the state of the system completely.

Let x and y be the independent thermodynamic variables of interest, and A, B

and C the closed systems of interest.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 8

Then, if A and C are in thermal equilibrium, there exists some relation between
xA, xC, yA and yC such that

F1( x A ,y A ,x C ,y C ) = 0

Likewise, if B and C are in equilibrium:

F2 (x B ,y B ,x C ,y C ) = 0

Each of these equations may be solved for yC:

y C = f 1 ( x A ,y A ,x C ) and y C = f 2 ( x B ,y B ,x C )

But, yC is the same in both cases, therefore

f 1 ( x A ,y A ,x C ) = f 2 ( x B ,y B ,x C )

The Zeroth Law requires that A and B also are in equilibrium, therefore we also
have
F3 ( x A ,y A ,x B ,y B ) = 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 9

This equation is equivalent to the previous result only if f1 and f2 contain xC in
such a way that it cancels from the equation.

The most general forms possible for f1 and f2 therefore are

f1( x A ,y A ,x C ) = g1( x A ,y A ) Ω ( x C ) + Λ ( x C )

f2 ( x B ,y B ,x C ) = g 2 ( x B ,y B ) Ω ( x C ) + Λ ( x C )

So, for bodies in thermal equilibrium,

g1( x A ,y A ) = g 2 ( x B ,y B ) = g 3 ( x C ,y C )

This function then has the same numerical value for each body if the bodies are
in thermal equilibrium.

This variable is called temperature.

Definition: q = g (x,y)

NOTE: For systems other than single-phase pure fluids, the temperature is a
function of the variables chosen to specify completely the state of the system
(independent variables).
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 10
Many physical properties have been used to measure temperatures. Some of
the more common are:

• The pressure of a gas at constant density (constant volume gas thermometry)

• The density (volume) of a liquid or solid (liquid-in-glass thermometers,
bimetallic strip thermometers)
• Electrical resistivity (thermistors, resistance thermometers, e.g. RTD's, etc.)
• Electrical potential appearing across the junction of two dissimilar metals
(thermocouple)
• Intensity of thermal radiation (optical pyrometer)
• Vapor pressure of a pure liquid
• Breakdown voltage of an electrical diode (silicon or gallium arsenide
thermometers)
• Magnetic susceptibility
• Phase transitions (liquid crystal thermometers)
60
• Anisotropy of gamma radiation from Co
• Amplitude of random noise appearing across a resistor due to the thermal
motion of the electrons within the conductor (Johnson or shot noise)
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 11
2.2 FIRST LAW OF THERMODYNAMICS

Statement: Energy is conserved when heat is taken into account.

We examine here the consequences of Newton's law of motion to obtain a better

understanding of the concept of energy.
 
 du
Newton's Law: F = ma = m
dt

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 12

2.2.1 Single Particle

2.2.1.1 One Dimension

2.2.1.1.1 Momentum

Consider the case where the force on an object depends only upon time

F = f (t )

Then
du
f (t ) = m ⇒ f (t )dt = m du
dt

Integration yields

∫ f (t ) dx = ∫ m du = mu + Constant of Integration

The product mu is the momentum, p, and the integral term is sometimes

referred to as the impulse.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 13

2.2.1.1.2 Energy

Consider the case where the force on an object depends only upon the position
of the object,

F = f (x )

Use the chain rule of differential calculus and the definition of velocity to write

du du dx du
= =u
dt dx dt dx

Then
du
f (x ) = mu ⇒ f (x )dx = mu du
dx

Integration yields

⌠
⎮f
⌡ ( x ) dx = ⌠⎮⌡ mu du = 21 mu 2 + Constant of Integration

Rearrange to obtain

1 mu 2
2
− ⌠⎮⌡ f ( x ) dx = E ( = constant )
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 14
The combination of Newton's Law and the condition that the force depends only
on position leads to a constant of integration that does not vary with respect to
time, and

E = KE + PE
where: E = total energy

KE = 21 mu 2 = kinetic energy

PE = − ⌠⎮⌡ f ( x ) dx = potential energy

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 15

2.2.1.2 Three Dimensions

In three dimensions, force and position are vector quantities:


r = ( x,y,z )
   
(  
F = F (r ) = Fx (r ),Fy (r ),Fz (r ), )
 
 du
F = ma = m
dt

2.2.1.2.1 Momentum

The three components of the vectors are given by

Fx dt = mdu x → p x − ∫ Fx dt = ℘x = Constant x

Fy dt = mdu y → p y − ⌠⌡ Fy dt = ℘y = Constant y

Fz dt = mdu z → p z − ∫ Fz dt = ℘z = Constant z

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 16

2.2.1.2.2 Energy

The three components of the vectors are given by

du x ⎛ du ⎞ ⎛ dx ⎞ du
Fx = m = m ⎜ x ⎟ ⎜ ⎟ = mu x x
dt ⎝ dx ⎠ ⎝ dt ⎠ dx

⎛ du y ⎞ ⎛ dy ⎞
du y du y
Fy = m = m⎜ ⎜ ⎟ = mu y
dt ⎝ dy ⎟⎠ ⎝ dt ⎠ dy

du z ⎛ du ⎞ ⎛ dz ⎞ du
Fz = m = m ⎜ z ⎟ ⎜ ⎟ = mu z z
dt ⎝ dz ⎠ ⎝ dt ⎠ dz

and

Fx dx = mu x du x → 1 mu 2 − ⌠⎮⌡ Fx dx = E x
2 x

Fy dy = mu y du y → 1 mu 2
2 y − ⌠⎮⌡ Fy dy = E y

Fz dz = mu z du z → 1 mu 2 − ⌠⎮⌡ Fz dz = E z
2 z

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 17

Adding these results together produces

1m
2 (u 2
x ) ( ⌠
)
+ u y2 + u z2 − ⎮ Fx dx + Fy dy + Fz dz = E x + E y + E z
⌡
=E

The two terms on the left hand side of the equation can be expressed much more
concisely in terms of vector notation (scalar or dot product):
 
u x2 + u y2 + u z2 = u •u
 
Fx dx + Fy dy + Fz dz = F • dr

The energy equation for a single particle in three dimensions is given by:
   
  p •p ⌠  
E = 2 m u • u − ⌡ F • dr =
1 ⌠
⎮ − ⎮⌡ F • dr
2m

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 18

2.2.2 Collections of Particles

Thermodynamics applies only to collections of particles, therefore we consider

the implications of the previous result for a collection of discrete, classical
particles.

Total Energy: E = ∑ E i

where: E i is the energy possessed by the ith particle

Substituting the previous result produces:

  ⌠ 
i
⎣
1 
E = ∑ 2 m u i • u i − ⎮⌡ Fi r i • dr i ⎤
⎡
⎦ ( )
 
where: ( )
F r i is the force acting on particle i


r i is the location of particle i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 19

The particle collection has a center of mass at

∑m i ri
 1 
rcm = i = total ∑ m i r i
∑m i m system i
i

where
total
msystem = ∑m i
i

is the total mass of the particles in the system.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 20

2.2.2.1 Kinetic Energy Calculation

The position of an individual particle can be described in terms of the position of

the center of mass and of the position of the individual particle relative to the
center of mass:
  
r i = r cm + R i

where: R i is the displacement of particle i from the location of the center of mass

Differentiation with respect to time provides a similar expression for velocities:

  
u i = ucm + U i

The kinetic energy term then can be written as:

 
2∑ i
1 m
i
 
ui •ui = 2∑ i
1 m
i
( 
) (
u cm + U i • u cm + U i )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 21

so that
  
1 m u    
• u i = 2 ∑ m i u cm • u cm + ∑ m i u cm • U i + 2 ∑ m i U i • U i
2∑ i i
1 1
i i i i
total      
= 21 msystem u cm • u cm + ∑ m i u cm • U i + 2 ∑ m i U i • U i
1
i i

But, the definition of the center of mass can be rewritten as


total 
msystem r cm
 
= ∑ m i r i = ∑ m i r cm + R i
i i
( )
total 

= msystem r cm + ∑m i R i
i

Therefore

∑m i R i = 0
i

and

∑m i U i = 0
i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 22

The kinetic energy term then becomes
 
1 m u  total  
• ui = 2 msystemucm • ucm + 2 ∑ m iUi • Ui
2∑ i i
1 1
i i

Therefore, the total kinetic energy of the system contains a term that contains the
motion of the center of mass and a term containing motion relative to the
center of mass, but no terms that contain both types of motion.

This means that the kinetic energy contains a term that is associated with the
macroscopic motion of the system, and a term that results from the motion of
the particles within the system.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 23

2.2.2.2 Potential Energy Calculations

Two types of forces contribute to the potential energy calculations:

1. Internal forces: forces that arise from interactions between particles within
 int
the system, F i

2. External forces: forces that arise from interactions of the individual particles
 ext
with external force fields, F i

Examples of common external force fields:

gravity
electric fields
magnetic fields

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 24

Forces are additive, so that
   int   ext 
( )
Fi ri = F i ri + F i ri ( ) ( )

Now, if the internal force acting on particle i because of the presence of particle j
depends only upon the distance between the particles,

( ))
         
( ) 
fik ri − rk = fik rcm + R i − rcm + Rk ( (
= fik R i − Rk )

The total force on particle i resulting from its interactions with all the other
particles in the collection is obtained by summation:
 int   
Fi = ∑ fik Ri − Rk
k
( )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 25

The total effect for the collection of particles is

ri
 
Potential Energy = -∑ ⌠⎮⌡
i r
( ) 
F i ri • d ri
ref
 
ri ri
 ext   int 
= -∑ ⌠⎮⌡
i r
( )
F i ri • d ri -∑ ⌠
⎮
⌡
i r
( ) 
F i ri • d ri
ref ref

so that the potential energy term becomes:

 
ri ri
    
∑ ⌠⎮⌡

i r
( ) 
F i ri • d ri = 21 ∑ ∑
i k r
⌠
⎮ fik
⌡
( ) 
Ri − Rk • d ri
ref ref
 
rcm ri
  ext 
⌠ ext
+ ∑ ⎮⌡ F i
i r
( )
 
ri • d ri + ∑ ⌠
⎮
⌡
( ) 
F i ri • d ri
ref rcm

In this expression, the first and third terms depend upon positions relative to the
center of mass, while the second term depends primarily upon the position of the
center of mass.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 26

Combining all the previous results gives

rcm
   ext 
E=

1 mtotal u
2 system cm

• ucm + ∑ 2 mi Ui • Ui − ∑
1
i i r
⌠
⎮
⌡
( ) 
F i ri • d ri
ref
 
ri ri
 ext    
− ∑ ⌠⎮⌡
i r
( )
F i ri • d ri − 21 ∑ ∑
i k r
⌠
⎮ fik
⌡
( ) 
Ri − Rk • d ri
cm ref

 ext 
where F ( rcm ) is the total force acting upon the collection of particles due to
external fields.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 27

The total energy expression is
   ext  
E= 1 mtotal u
2 system cm
• ucm − ⌡ F ( rcm ) • d rcm
⌠
⎮
 
⎡      ri ri
 ext  ⎤
⎢ i i k r
⌠ 
( )
i r
⌠
( )
+ ⎢ 21 ∑ mi Ui • Ui − 21 ∑ ∑ ⎮⌡ fik Ri − Rk • d ri − ∑ ⎮⌡ F i ri • d ri ⎥
⎥
⎣ ref cm ⎦

The first two terms deal with the macroscopic velocity and position of the system.

The last three terms deal with the properties possessed by the individual
particles.

The term in square brackets is intractable for calculations of fluids because the
number of particles generally exceeds 1020.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 28

The terms inside the square brackets are lumped and defined as the total
internal energy of the system:


ri
    
total
Usystem = 2∑ i i
1 mU
i
• Ui + 21 ∑ ∑
i k r
⌠
⎮ fik
⌡
( 
Ri − Rk • d ri)
ref

ri
 ext 
+ ∑ ⌠⎮⌡

i r
( ) 
F i ri • d ri
cm

total
Kinetic energy E ktotal = 21 m system 2
u cm


rcm
 
Potential energy E ptotal = − ⌠⎮⌡

( ) 
F rcm • d rcm
rref

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 29

Total energy of the system

total
E system total
= Usystem + E ktotal + E ptotal = msystem
total ⎡U + E k + E p ⎤
⎣ ⎦
total

U, Ek and Ep are the internal, kinetic and potential energies per unit mass. (or
per mole).

Note that the internal energy, U, depends only on the state of the system,
therefore the internal energy is a state function.

2.2.3 Consequence of the First Law of Thermodynamics

There exists a function (which we call energy) that is constant with respect to
time for an isolated system.

(This function is quite useful for describing the physical behavior of practical
systems.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 30

2.2.4 General Energy Balance

By applying the conservation of energy to the entire universe (which is an

isolated system) and dividing the universe into two entities, the system and the
surroundings, which are separated by a boundary), we conclude that:

Net change in total energy of the system =

= net transfer of energy across the boundary

+ conversion of mass to energy within the system

(nuclear reactions)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 31

This may be stated in mathematical form in either the differential form of the
general energy balance:

⎣ (
d ⎡m U + E k + E p ⎤
⎦system
= ) ∑
all heat
dQi + ∑
all work
dWi
transfer mechanisms
mechanisms

+ ∑ (dm )i
all
xfer
(H + E
i
k
i )
+ E ip − ∑
all nuclear
(dm )nuc
i
c 2

streams reactions

or the integral form of the general energy balance:

⎣ (
Δ ⎡m U + E k + E p ⎤
⎦system
= ) ∑
all heat
Qi + ∑
all work
Wi
transfer mechanisms
mechanisms

+ ∑
all
⌠
⎮
⌡
(dm )xfer
i (H i + E k
i + E i )−
p
∑ Δm nuc
i c
all nuclear
2

streams reactions

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 32

In these equations:
U = internal energy per unit mass

Ek = kinetic energy per unit mass

Ep = potential energy per unit mass

Q = energy transferred as heat
W = energy transferred as work
mxfer = mass transferred across the system boundary
H = U + PV = enthalpy per unit mass
mnuc = change in mass due to nuclear reactions
c = velocity of light in vacuum

Note that the use of the enthalpy in the mass transfer term automatically includes
the effect of injection and ejection work.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 33

Definitions:

Heat: energy that is transferred across the system boundary because of a

temperature gradient driving force, i.e., a temperature difference between
the system and the surroundings, excluding energy transfer associated
directly with mass flow across the boundary. The mechanisms for heat
transfer are conduction, convection and radiation.

Work: energy that is transferred across the system boundary because of any
driving force other than a temperature gradient, again excluding energy
transfer associated with mass flow across the boundary.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 34

2.3 SECOND LAW OF THERMODYNAMICS

2.3.1 Physical Observations

• Not all processes that conserve energy occur.

• Energy never is transferred as heat from a cold body to a hot body with no
other change in the system or its surroundings.

• Certain physical states of an isolated system never can precede other physical
states of that system.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 35

2.3.2 Formal Statements of the Second Law

2.3.2.1 Kelvin

It is impossible to extract heat from a reservoir and convert it wholly into work
without causing other changes in the universe.

2.3.2.2 Clausius

Heat can never, of itself, flow from a lower to a higher temperature.

2.3.2.3 Planck

No cyclic process is possible that results in the flow of heat from a single heat
reservoir and the performance of an equivalent amount of work.

• The statements of Kelvin and Clausius are equivalent, but the equivalence is
not obvious.

• We prove the equivalence by demonstrating that a violation of Kelvin's state-

ment implies a violation of Clausius statement and vice versa.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 36

2.3.2.4 Heat Engine:

A heat engine is a conceptual device that is used to study Second Law effects.
It consists of a device that is capable of exchanging energy as heat with two
reservoirs, one at a higher temperature and one at a lower temperature, and of
exchanging energy as work with the surroundings. The schematic diagram for a
heat engine is shown below.

A reservoir is an entity that is capable of absorbing or delivering any amount of

energy as heat without changing its temperature.
High Temperature
Reservoir

QH

QC
Low Temperature
Reservoir

Schematic Diagram of Heat Engine

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 37

For the discussions at present, we shall continue to use the symbol q to denote
temperature, so that qH is the temperature of the hotter reservoir and qC is the
temperature of the colder reservoir.

2.3.2.5 Notation:

qH = temperature of the hotter reservoir

qC = temperature of the colder reservoir

QC = amount of energy transferred as heat from the colder reservoir to the

system

QH = amount of energy transferred as heat from the hotter reservoir to the

system

W = amount of energy transferred as work from the surroundings to the

system

This choice of signs is consistent with that used for the general energy balance,
but not consistent with discussions of Second Law effects found in many other
references. Other references, especially those discussing mechanical
engineering applications, use the opposite sign convention for W.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 38
2.3.2.6 Proof of Equivalence of Kelvin and Clausius Statements:

To prove the equivalence of these two statements, we consider two heat engines
that are coupled and that operate between a reservoir at qH and another
reservoir at qC.

The general arrangement is shown schematically below.

θH

QH
A QHB
WA WB
A B
QCA QCB

θC

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 39

 θH

H
QH
A QB
WA WB
A B
QCA QCB

θC

The net energy transferred as work for the coupled engine is:

W = WA +WB

The net energy transferred as heat from the hotter reservoir is:

Q H = Q AH + QBH

The net energy transferred as heat from the colder reservoir is:

Q C = Q AC + QBC

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 40

Case I: Assume that Engine A violates Kelvin's statement.

Then Q AC = 0 and WA < 0

The energy balance for Engine A requires:

Q AC + Q AH +WA = 0 → WA = −Q AH

The energy balance for Engine B gives:

QBC + QBH +WB = 0 → QBC + QBH = −WB

If the engines are coupled so that the net work is zero (all work delivered by one
engine is consumed by the other), then

WA = −WB

and

−Q AH = QBH + QBC → −QBC = Q AH + QBH

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 41

The net energy transfers for the coupled engines in this case are given by:

Q H = Q AH + QBH = −QBC

QC = QA
C
+ QBC = QBC

W = WA +WB = 0

When QBC > 0 , a net transfer of energy as heat occurs from the colder (qC) to the
hotter (qH) reservoir with no other change in the universe. This violates the
Clausius statement.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 42

Case II. Assume that Engine A violates the Clausius statement.

Then

WA = 0, Q AC > 0, and Q AH < 0

The energy balance for Engine A requires:

WA + Q AH + Q AC = 0 → Q AH = −Q AC

The energy balance for Engine B gives:

WB + QBC + QBH = 0 → WB = −QBC − QBH

Operate the engine in such a way that:

Q AC = −QBC = −Q AH

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 43

The net energy transfers for the coupled engines in this case are given by:

QC = QA
C
+ QBC = 0

W = WA +WB = −Q H

QH = Q AH + QBH = −W

When W < 0 and QH > 0, heat is converted entirely to work, which violates the
Kelvin statement of the Second Law.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 44

2.3.2.7 Carnot Cycle

A Carnot Cycle is a cyclic process that consists of:

two adiabatic reversible processes

two isothermal reversible processes.

Nicolas Sadi Carnot invented these as a tool for analyzing the efficiencies of heat
engines.

1824 Monograph: "Reflections on the Motive Power of Heat, and on Machines

Fitted to Develop That Power"

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 45

2.3.3 Consequences Of Second Law
We now use the concept of heat engines and Carnot cycles to prove four
implications or consequences of the observations summarized by the Second
Law of Thermodynamics.

2.3.3.1 Consequence No. 1:

QH
C
= f (θC ,θH )
Q

Proof:
Consider two Carnot cycles, one that utilizes an ideal gas as the working
substance and the other using condensing and evaporating steam as the working
substance.

The two working substances (fluids in this case) have very different properties,
therefore the amounts of energy transferred as heat per cycle are most likely
different, but it always is possible to choose numbers of cycles such that:

nG QGC ≅ nS QSC

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 46

 QH
G QH
S

P P
C
QG QC
S

V V
Ideal Gas Steam
QH
G QH
S

G WG S WS

C
QG QC
S

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 47

The net heat transfers for the combined system are:

H H
Hot Reservoir: Qnet = nGQG + nSQSH

C C
Cold Reservoir: Qnet = nGQG + nSQSC

and the net work is: Wnet = nGWG + nSWS

Choose the coupling and direction of operation such that:

nSQSC = −nGQG
C

Then:

C
Qnet =0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 48

Application of Second Law:

When no energy is transferred as heat to or from the low temperature reservoir,

C
Qnet = 0 , and the Second Law requires that
H
Wnet ≥ 0 and Qnet <0

Application of First Law:

The First Law requires that:

C H
Wnet + Qnet + Qnet =0

C
When Qnet = 0 , this becomes: Wnet = −Qnet
H

For the coupled engines,

nSWS + nGWG = Wnet = −Qnet

H
= −nSQSH − nGQGH

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 49

The Second Law requires that:

(
− nSQSH + nGQGH ≥ 0 ) ⇒ (n QS
H
S )
+ nGQGH ≤ 0

The operation of the two devices can be reversed, that has the effect of reversing
all signs, i.e., multiplying through by -1. The reverse operation leads to

( nSQSH + nGQGH ≥ 0 )
Both conditions must be satisfied, and that only is true for the equality, so

nSQSH = −nGQG
H

The original setup specified that

nSQSC = −nGQG
C

therefore:

nSQSH H
−nGQG QSH H
QG
= ⇒ =
nSQSC C
−nGQG QSC C
QG

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 50

The ratios are the same for both cycles, even though the substances have
substantially different properties.

In addition, the cycles are completely arbitrary, so that only the reservoir
temperatures can be involved as variables.

Therefore

QH
C
= f (θC ,θH )
Q

NOTE: This result is completely independent of the properties of the working

substance and the precise details of the Carnot Cycle.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 51

2.3.3.2 Consequence No. 2:

QH TH
=−
QC TC

Proof:

3
4
P θH
2 5
θI
1 6
θC

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 52

We will first show that

g (θH )
f (θC ,θH ) =
g (θC )

This will be accomplished by applying consequence 1 twice.

QH
Cycle 2345: I
= f (θI ,θH )
Q

QI
Cycle 1256: C
= f (θC ,θI )
Q

By algebra:

QH QH QI
C
= I C
⇒ f (θC ,θH ) = f (θC ,θi )f (θI ,θH )
Q Q Q

Then

g (θH ) g (θI ) g (θH )

f (θC ,θH ) = ⇒ f (θC ,θI )f (θI ,θH ) =
g (θC ) g (θC ) g (θI )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 53

No other form will cancel completely in the product.

To determine the form of the function g(q), we use the ideal gas for its simplicity
(all substances must have the same result). The equation of state is

PV = Rθ
For the isothermal steps where energy is transferred reversibly as heat:

V4
H H
V4 ⌠ RθH ⎛ V4 ⎞
Q = −W = ∫ P dV = ⎮
⎮ dV = RθH ln ⎜ ⎟
V3 ⌡ V ⎝ V3 ⎠
V3

⎛V ⎞
Likewise: Q C = RθC ln ⎜ 6 ⎟
⎝ V1 ⎠

Combining these results gives:

⎛V ⎞
ln ⎜ 4 ⎟
QH θ ⎝ V3 ⎠
= H
QC θC ⎛ V1 ⎞
ln ⎜ ⎟
⎝ V6 ⎠
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 54
The adiabatic (Q=0) steps 123 and 456 must obey the First Law of
Thermodynamics: (differential form)

dU = dW + dQ

For the ideal gas

dW = −PdV ⎫
ig ⎪ Rθ
dU = CV dθ ⎬ ⇒ CVig dθ = −PdV = − dV
⎪ V
dQ = 0 ⎭

Then
θH V3
⌠ ig dθ ⌠ RdV
⎮ CV = −⎮
⌡ θ ⌡ V
θC V1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 55

For the two adiabatic steps:

⎛V ⎞ ⎛θ ⎞
123: R ln ⎜ 3 ⎟ = −CVig ln ⎜ H ⎟
⎝ V1 ⎠ ⎝ θC ⎠

⎛ V6 ⎞ ig ⎛ θC ⎞ ig ⎛ θH ⎞
456: R ln ⎜ ⎟ = −CV ln ⎜ = CV ln ⎜
⎝ V4 ⎠ ⎝ θH ⎟⎠ ⎝ θC ⎟⎠

So

⎛V ⎞ ⎛V ⎞ ⎛V ⎞ V6 V1 V6 V4
ln ⎜ 6 ⎟ = − ln ⎜ 3 ⎟ = ln ⎜ 1 ⎟ ⇒ = ⇒ =
⎝ V4 ⎠ ⎝ V1 ⎠ ⎝ V3 ⎠ V4 V3 V1 V3

and

⎛ V4 ⎞ ⎛ V4 ⎞
ln ln
Q H θH ⎜⎝ V3 ⎟⎠ θH ⎜⎝ V3 ⎟⎠ θH
= = − = −
Q C θC ⎛ V1 ⎞ θC ⎛ V6 ⎞ θC
ln ⎜ ⎟ ln ⎜ ⎟
⎝ V6 ⎠ ⎝ V1 ⎠

Therefore consequence 2 is proved.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 56

Because the Carnot cycle is universally valid, we use this consequence of the
Second Law to refine the concept of temperature first encountered in the
observations leading to the Zeroth Law of Thermodynamics.

The absolute temperature scale is defined by defining the temperature of the

triple point temperature of water as exactly 273.16 K, and all other temperatures
according to

QH TH
= −
QC TC

where T is the thermodynamic temperature.

The equation of state for an ideal gas now is:

PV = RT
Constant volume gas thermometry with real gases provides a practical primary
device to establish the absolute temperature scale according to

lim ⎛ PV ⎞ T lim ⎛ (PV )T ⎞

T = ⎜⎝ ⎟⎠ ⇒ =
P →0 R 273.16 K P → 0 ⎜⎝ (PV )273.16 ⎟⎠
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 57
2.3.3.3 Consequence No. 3:

dQrev
dS ≡ defines a state function
T

Proof:

Carnot Cycle has

QH QC
=−
TH TC

so that

QH QC
+ =0
TH TC

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 58

For any arbitrary cycle:

P •2
1 •
V

For each Carnot cycle:

QiH QiC
+ =0
TH TC

and the total for all Carnot cycles is

⎛ QiH QiC ⎞ QiH QiC

∑⎜ + ⎟ =∑T +∑T =0
i ⎝ H
T TC ⎠ i H i C

The more Carnot cycles that are used to approximate the actual cycle, the more
accurate the approximation.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 59

In the limit as the number of cycles goes to ∞, the approximation to the arbitrary
cycle becomes accurate.

Also:
2
⎡ #cycles QiH ⎤ dQ H ⌠
lim ⎢ ∑ ⎥= ⎮
⎮
#cycles→∞ ⎢⎣ i TiH ⎥⎦ TH ⌡
1

1
⎡ #cycles QiC ⎤ ⌠ dQ C
lim ⎢ ∑ ⎥ = ⎮⎮ C
#cycles→∞ ⎢⎣ i TiC ⎥⎦ ⌡ T
2

Then
2 1
H
⌠ dQ ⌠ dQ C ⌠ dQrev
⎮
⎮
+ ⎮⎮ = ⎮⎮
⌡ TH ⌡ TC ⌡ T
1 2

The symbol ∫ denotes an integral along a closed path, i.e., the path begins and
ends at the same point.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 60

Then, if we define a new function S according to

dQrev
dS ≡
T

the previous discussion shows that

⌠ ⌠ dQ
 dS
⎮
⎮
= 
⎮
⎮
rev
=0
⌡ ⌡ T

This means that S is a state function.

In other words, the value depends only on the state of the system, not upon how
the system reached that state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 61

2.3.3.4 Consequence No. 4:

ΔS ≥ 0 for any allowed process in an isolated system, i.e., for any process that
does not violate the Second Law of Thermodynamics.

Proof:

Adiabatic, Reversible Change of State:

For the adiabatic, reversible path from state 1 to state 2, dQrev = 0 at all points
along the path, Then:

2
⌠ dQ
S2 − S1 = ⎮
⎮
rev
=0
⌡ T
1

Then, S2 = S1 ⇒ ΔS = 0 for any reversible, adiabatic process.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 62

Adiabatic, Irreversible Change of State:

For an irreversible, adiabatic path from state 1 to state 2, dQ = 0 at all points

along the path, so that:

2
1Q 2 = ∫ dQ = 0
1

dQ
Note that dS ≠ in this case
T

Complete a cycle by returning from state 2 to state 1 using a reversible process

that involves energy transferred as heat and as work. Then
1
⌠ dQ
S1 − S2 = ⎮⎮ rev
⌡ T
2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 63

First Law Application:

Adiabatic, irreversible path from state 1 to state 2: U 2 − U1 = 1W2

Non-adiabatic, reversible path from state 2 to state 1: U1 – U2 = 2Q1 + 2W1

U is a state function ⇒ − (U 2 − U1 ) = U1 − U 2

Then

− 1W2 = 2W1 + 2Q1 ⇒ 1W2 + 2W1 = − 2Q1

The net work for the cycle is

Wnet = 1W2 + 2W1 = − 2Q1

If 2Q1 > 0, then Wnet < 0, and the cycle from 1 to 2 and return would violate the
Planck statement of the Second Law.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 64

To satisfy the Second Law, we must have

1
⌠ dQ
2 Q1 < 0 ⇒ <0
⎮
rev
⎮
⌡ T
2

Therefore:

1
⌠ dQ
S1 − S2 = ⎮⎮ rev
<0 ⇒ S2 > S1
⌡ T
2

Therefore, for any process allowed by the Second Law:

ΔS ≥0
1 2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 65

2.3.4 Summary Statement of Second Law

a. The entropy, S, defined by

/ rev
dQ
dS ≡
T

is a state function.

b. The entropy of an isolated system never decreases; S remains constant

during a reversible process, and increases during an irreversible
process.

NOTE: The universe (system + surroundings) is an isolated system, therefore

the entropy of the universe either increases or remains constant during
any process that is allowed by the Second Law of Thermodynamics.

One implication of the Second Law of Thermodynamics is that there is a

fundamental difference between energy transferred as heat and energy
transferred as work.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 66

2.3.5 Irreversible Processes:

For an irreversible process that is allowed, we can have 1Q2 = 0 even though
ΔS = S2 − S1 > 0 .
2
2 /
⌠ dQ
1Q 2 / =0
= ∫ dQ ⇒ ⎮
⎮ =0
1 ⌡ T
1
2
ΔS = ⌠
⎮ dS
⌡ > 0 because the process is irreversible
1

/
dQ
But, ΔS > 1Q 2 ⇒ dS > ⇒ /
TdS ≥ dQ
T

where the equality holds for reversible processes and the inequality for
irreversible processes.

/ rev = −PdV
For a volume change of the system: dW

The First Law holds for both reversible and irreversible processes, therefore:
dU = dQ/ rev + dW
/ rev (reversible process)
/ + dW
= dQ / (irreversible process )
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 67
Second Law conclusions lead to

/ rev = TdS ⎫
dQ
/ < dQ
⎬ ⇒ dQ / rev / + δ = dQ
⇒ dQ / rev
/ < TdS ⎭
dQ

where d is always positive.

Therefore, for processes that accomplish a given change in internal energy, dU:

/ rev + dW
dU = dQ / rev
/ = (dQ
/ + dW
= dQ / rev − δ ) + dW
/

Therefore

/ rev = dW
dW / −δ ⇒ / = dW
dW / rev + δ

So, for all processes that are allowed by the Second Law of Thermodynamics:

Q ≤ Qrev
W ≥ Wrev

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 68

2.3.6 Practical Applications - Efficiencies

Thermodynamic Efficiencies:

W
Thermal efficiency: ηT ≡
QH

Wrev TC
Carnot efficiency: η ≡ = 1−
C QH TH

QC
Coefficient of Performance: COP ≡
W

TC
(COP )Carnot =
TH −TC

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 69

Pattern Efficiencies:

For descriptions of irreversible processes in practical machinery, the equality

form of the Second Law does not apply, so a second specification is required.
This is done by defining a pattern efficiency that compares the actual
performance to the performance of a reversible device. Pattern efficiencies are
determined by measuring actual perfomance characteristics.

W
Work-producing machines (W < 0): ηp ≡
Wrev

Wrev
Work-consuming machines (W > 0): ηp ≡
W

Note that the Second Law requires that 0 ≤ ηp ≤ 1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 70

Maximum Work

Consider a case where the system goes from state 1 to state 2 and energy is
transferred as heat to a reservoir. (The temperature of the reservoir remains
constant during the process.)
total
Suniverse = Ssys
total
+ Sres
total

The differential form of this equation is

total
dSuniverse = dSsys
total
+ dSres
total

Because the reservoir temperature is constant

/ total
dQ
dS total
res =−
Tres

Then
/ total
dQ
dS total
universe = dS total
sys + dS total
res = dS total
sys −
Tres

/ total = TresdSsys
dQ total
−TresdSuniverse
total

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 71

First Law: / total + dW
dU total = dQ / total → dQ
/ total = dU total − dW
/ total

/ total
Substitute the previous result for dQ

/ total = TresdSsys
dU total − dW total
−TresdSuniverse
total

/ total = dU total −TresdSsys

dW total
+TresdSuniverse
total

Integrate this equation to get:

(
W total = U 2total − U1total −Tres Ssys )
total
(
( 2 ) − Ssys
total
) (
(1) +Tres Sunitotal ( 2 ) − Sunitotal (1) )
The Second Law of Thermodynamics requires that

⎡⎣Suniverse
total
(2 ) − Suniverse
total
(1)⎤⎦ ≥ 0

for any allowed process. Therefore

(
W total ≥ U 2total − U1total −Tres Ssys )
total
(
(2 ) − Ssys
total
(1) )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 72

When useful work is obtained, W < 0. The maximum amount of work that can
obtained from the process in question is

total
Wmax = ΔU system
total
− Tres ΔSsystem
total

Note that ΔSsystem ≥ 0 for any adiabatic process.

total

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 73

2.3.7 Summary Statement of Combined First and Second Laws

For reversible processes:

/ rev = TdS ⎫
dQ
⎬ ⇒ dU = TdS − PdV
/ rev = −PdV ⎭
dW

Only state functions are involved in this result, therefore it must hold for both
reversible and irreversible processes.

The summary statement of the physics involved in the First and Second Laws of
Thermodynamics is:

dU = TdS − PdV

We will refer to this as the Physics Summary Statement.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 74

2.4 THIRD LAW OF THERMODYNAMICS

Formal Statements:

2.4.1 Nernst-Simon Statements of Third Law

The entropy change associated with any condensed system undergoing a

reversible isothermal process approaches zero as temperature approaches
0 K, where condensed system refers to liquids and solids.

It is impossible for any process, no matter how idealized, to reduce the

entropy of a system to its absolute-zero value in a finite number of
operations.

2.4.2 Lewis-Randall Statement of Third Law

If the entropy of each element in some (perfect) crystalline state be taken

as zero at the absolute zero of temperature, every substance has a finite
positive entropy; but at the absolute zero of temperature the entropy may
become zero, and does so become in the case of perfect crystalline
substances.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 75

2.4.3 Consequences of Third Law

It is impossible by any procedure, no matter how idealized, to reduce the

temperature of any system to zero temperature in a finite number of finite
operations.

2.4.3.1 Heat Capacity

lim C X = 0
T →0

Ideal gas with Fermi statistics (odd total spin): lim CV ∝T

T →0
3
Ideal gas with Bose statistics (even total spin): lim CV ∝T 2
T →0

Electrically conducting crystalline solid: lim CV ∝T

T →0

Electrically insulating crystalline solid: lim CV ∝T 3

T →0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 76

2.4.3.2 Thermal Expansion

⎡ 1 ⎛ ∂V ⎞ ⎤
lim α = lim ⎢ ⎜ ⎟⎠ ⎥ = 0
T →0 ⎝
T →0 ⎣V ∂T P ⎦

The functional dependence of the thermal expansion is the same as that of the
heat capacity.

2.4.3.3 Phase Transitions (Vapor/Liquid, Vapor/Solid, Solid/Liquid)

Only 3He and 4He have liquid as an equilibrium phase at T = 0 K. The entropy
change of a vapor/condensed phase transition goes to infinity as T → 0 ,
therefore
lim P σ = 0 , where Ps is the sublimation or vapor pressure.
T →0

For the helium isotopes, both solid and liquid phases exist at absolute zero, so
the entropies of the solid and liquid phases must be identical at the equilibrium
pressure. Then:
dP σ
lim = 0 , where Ps is the melting (freezing) pressure.
T →0 dT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 77

2.4.4 Practical Thermometry

2.4.4.1 Primary Thermometer:

A device for which the relationship between a measured physical property and
the thermodynamic temperature is known exactly from theory.

Examples: Constant volume gas thermometry

Optical pyrometry
Susceptibility of paramagnetic salts
Resistance white noise (Johnson noise)

Constant Volume Gas Thermometry Relation:

T2
lim
(PV )2
=
P→0

T1 lim
P→0 (PV )1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 78

2.4.4.2 Secondary Thermometer:

A device for which the relationship between a measured physical property and
the thermodynamic temperature is not known exactly. Secondary thermometers
must be calibrated in some manner.

Examples: Platinum resistance (PRT's)

Thermocouples
Liquid-in glass

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 79

2.4.4.3 Philosophy of Practical Temperature Scales

• Use primary thermometers to measure the thermodynamic temperatures of

fixed points.
• Assign temperatures to the fixed points based upon the best available
measurements.
• Define a standard instrument for interpolating between the assigned fixed
points. Also define a standard relationship between the measured property
and the practical temperature.

Abbreviations
ITS: International Temperature Scale
IPTS: International Practical Temperature Scale
CCT: Comité Consultatif de Thermométrie
CIPM: Comité International des Poids et Mesures

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 80

2.4.4.4 History of Temperature Scales

Early Temperature Scales

Newton (1701): Ice point = 0 °N

Human body = 12 °N

Fahrenheit (1724): Freezing point of salt + water = 0 °F

Human Body = 96 °F

Celsius (1742): Ice Point = 0 °C

Boiling point of water = 100 °C

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 81

International Temperature Scales:

CIPM - 1887: This scale was based upon constant volume gas thermometry
using normal H2

• Originally defined from -25 to 100 °C

• Fixed points: freezing point of water = 0 °C

boiling point of water = 100 °C

ITS-27:

• Fixed points: Freezing point: water

Boiling points: oxygen, water, sulfur
Melting points: silver, gold

• Interpolating instruments:
platinum resistance thermometers
platinum-rhodium thermocouple
optical pyrometer

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 82

ITS - 48:

• Change from melting point to freezing point for Ag, Au

• Adjusted the ranges of applicability for the standard instruments

ITS-48 (1960 modification):

• Change from freezing point to triple point of H2O.

• Boiling point of sulfur changed to freezing point of Zn.

IPTS - 68:

• Boiling point of H2O changed to freezing point Sn.

• Fixed points added at low temperatures

• Values changed for: Boiling point of O2

Freezing points of Zn, Au, Ag

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 83

IPTS - 68 (1975 Modification):

• Triple point of Ar added as an acceptable fixed point.

EPT - 76:

• Provisional scale adopted for 0.5 to 30 K range, utilizing mostly

superconducting transition temperatures as fixed points.

ITS - 90:

• Current international temperature scales

• Significant changes in many assigned fixed point temperatures, especially at

higher temperatures.

• Change in interpolating instrument between 600 and 1050 °C from

thermocouples to platinum resistance thermometers.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 84

2.4.4.5 Evolution of International Temperature Scales

Fixed NHS ITS-27 ITS-48 IPTS-68 IPTS-68(75) EPT-76 ITS-90

Point

Au FP ---- 1336.15 1336.15 1337.58 1337.58 --- 1337.33

Ag FP ---- 1233.65 1233.95 1235.08 1235.08 --- 1234.93

Al FP ---- --- --- --- --- --- 933.473

S BP ---- 717.75 717.75 --- --- --- ---

Zn FP ---- --- --- 692.73 692.73 --- 692.677

Sn FP ---- --- --- (505.1181) (505.1181) --- 505.078

In FP ---- --- --- --- --- --- 429.7485

H2O BP 373 373.15 373.15 373.15 373.15 --- ---

Ga TP ---- --- --- --- --- --- 302.9146

H2O TP ---- --- --- 273.16 273.16 --- 273.16

H2O FP 273 273.15 273.15 --- --- --- ---

Hg TP ---- --- --- --- --- --- 234.3156

O2 BP ---- 90.18 90.18 90.188 90.188 --- ---

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 85

 Fixed
NHS ITS-27 ITS-48 IPTS-68 IPTS-68(75) EPT-76 ITS-90
Point

Ar TP ---- --- --- --- (83.798) --- 83.8058

O2 TP ---- --- --- 54.361 54.361 --- 54.3584

Ne BP ---- --- --- 27.102 27.102 27.102 ---

Ne TP ---- --- --- --- --- 24.5591 24.5561

H2 BP ---- ---- --- 20.28 20.28 20.2734 20.3

H2 BP† ---- --- --- 17.042 17.042 17.0373 17.0

H2 TP ---- --- --- 13.81 13.81 13.8044 13.8033

Pb SP ---- --- --- --- --- 7.1999 ---

4He BP ---- --- --- --- --- 4.2221 4.2

In SP ---- --- --- --- --- 3.4145 ---

3He BP ---- --- --- --- --- --- 3.2

Al SP ---- --- --- --- --- 1.1796 ---

Zn SP ---- --- --- --- --- 0.851 ---

Cd SP ---- --- --- --- --- 0.519 ---

†
Reduced pressure boiling point (P = 33,330.6 Pa)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 86

Defining Fixed Points For ITS-90
Temperature
Material Equilibrium State
(kelvins)
He Vapor Pressure 3-5
e-H2 Triple Point 13.8033
e-H2 Vapor Pressure 17 - 20.3
Ne Triple Point 24.5561
O2 Triple Point 54.3584
Ar Triple Point 83.8058
Hg Triple Point 234.3156
H2O Triple Point 273.16
Ga Melting Point 302.9146
In Freezing Point 429.7485
Sn Freezing Point 505.078
Zn Freezing Point 692.677
Al Freezing Point 933.473
Ag Freezing Point 1234.93
Au Freezing Point 1337.33
Cu Freezing Point 1357.77

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 87

Differences Between ITS - 90 and IPTS - 68

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 88

2.4.4.6 Proposed Redefinition of Temperature Unit in SI System

The definition of the kelvin will undergo a fundamental change on May 20, 2019,
if the proposed revision of the SI system is accepted at the November 2018
meeting of the General Conference of Weights and Measures (CGPM). Rather
than using the triple point of water to fix the temperature scale the proposal
recommends that the energy equivalent as given by Boltzmann's equation be
used.

Current definition: The kelvin, unit of thermodynamic temperature, is the

fraction 1/273.16 of the thermodynamic temperature of the triple point of
water.

Proposed definition: The kelvin, K, is the unit of thermodynamic

temperature; its magnitude is set by fixing the numerical value of the
Boltzmann constant to be equal to exactly 1.380649×10−23 when it is
expressed in the unit s−2·m2·kg·K−1, which is equal to J·K−1.

One consequence of this change is that the new definition makes the definition of
the kelvin depend on the definitions of the second, the metre, and the kilogram.
Temperature no longer is an independent definition.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 89

2.5 MICROSCOPIC INTERPRETATIONS: TEMPERATURE, ENTROPY

2.5.1 Microscopic (Statistical) Viewpoint of Entropy

Entropy is related to the number of different ways that a system of constant

energy may be arranged.

2.5.1.1 Model system

As a concrete example, consider a set of N particles. Each particle may be in

one of two states; ‘up’ and ‘down’.

.....

1 2 3 4 5 ..... N

The particles are completely independent – they do not interact with one another
in any fashion. There is no external field present. Thus, all configurations of the
system have identical energy. For convenience, set the energy equal to zero.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 90

2.5.1.2 Some necessary definitions

Microstate: the most detailed level of description of a system.

Examples:

Quantum mechanics: quantum state

Classical statistical mechanics: position and momentum of every atom

Model system: the state of every particle

Macrostate: a set of microstates with something in common

Examples:

Quantum mechanics: same value of the energy

(then the macrostates may be called “levels”)

Model system: same total number of particles in ‘up’ state

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 91

2.5.1.3 Microstates and the connection with entropy

There are 2N different microstates in the example system.

Because this number represents the total number of different ways that the
system may be arranged, and all arrangements have the same energy (i.e., 0),
we expect that it will be related to entropy.

In statistical mechanical terminology, this number represents the

microcanonical partition function Ω for the model system. The term
microcanonical generally means “at constant energy, volume, and number of
particles” and so Ω is sometimes written more fully as Ω (E,V,N ). It generally is
defined as

Ω (E,V ,N ) ≡ the number of different microstates having energy E

at volume V and number of particles N

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 92

Boltzmann defined entropy as being proportional to the natural log of the
microcanonical partition function

S (E,V ,N ) = kB lnΩ

In the simple model system as defined, volume does not really exist, and energy
is always 0, so

Ω (E = 0,N ) = 2N

S = k B ln2N = Nk B ln2

For “real atoms” in three dimensions, the expression for entropy will be more
complex.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 93

2.5.1.4 Macrostates

For our model system, consider macrostates being defined based on the same
total number of particles in the ‘up’ state. There are N + 1 different macrostates.
For large N, this is considerably smaller than the number of different microstates.

The number of different microstates in a macrostate having nu particles in the

‘up’ state is
N!
(N − nu )!nu !
The total number of microstates can be calculated by taking the sum of the
number of microstates in each macrostate over all of the possible macrostates

N N!
Total # of microstates = ∑ = 2N
nu =0 (N − nu )!nu !

Using the binomial theorem with r = 1 gives the total number of microstates
N N! N N!
r = (1+ r ) ⇒
N
∑ Total # of microstates = ∑ = 2N
n=0 (N − n )!n! nu =0 (N − nu )!nu !

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 94

2.5.1.5 Connection with macroscopic property measurements

Microstate picture.
From statistical mechanics, the average value of any property B , which has a
value of Bi in any given microstate i , can be calculated from a probabilistic
average
total # of
microstates
B = ∑ Bi pi
i =1

where pi is the probability that the system is in microstate i.

Likewise, the variance in that property is

total # of
microstates
σ B2 ≡ (B − B )2 = ∑ (Bi − B )2 pi
i =1

A fundamental assumption of statistical mechanics is that for a system at

constant (macroscopic) energy, all microstates are equally likely. Therefore
1
pi =
Ω

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 95

Macrostate picture.

Alternatively, if macrostates are chosen so that all members of a macrostate

have the same value for the property B, we can use a sum over macrostates to
define the average.
total # of
macrostates
B = ∑ Bj pj
j =1

where p j is the probability of a macrostate. Again assuming that all microstates

are equally likely, we can say
# of microstates in macrostate j
pj =
Ω

The variance can be written

total # of

(B j − B )2 p j
macrostates
σ B2 ≡ (B − B )2 = ∑
i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 96

Measuring a property.

Assume that we have an experimental device that is capable of measuring the

difference between the number of up and down particles in the system, nu − nd ,
but it is not sufficiently precise to follow individual particles.

What is the average reading, and how large are the expected fluctuations?

For the model system, using the macrostate approach

N!

nu − nd = ∑ (nu − nd )
N (N − nu )!nu !
nu =0 2N

Noting that

nd = N − nu

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 97

leads to

N!

nu − nd = ∑ ( 2nu − N )
N (N − nu )!nu !
nu =0 2N

2 N N! N N N!
= N ∑ nu − N ∑
2 nu =0 (N − nu ) ! nu ! 2 nu =0 (N − nu ) ! nu !

2 ⎡ N −1⎤ N ⎡ N ⎤
= N ⎣
N2 ⎦ − N ⎣2 ⎦
2 2

=0

This result occurs because there is no energy incentive for the particles to be
either up or down.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 98

Consider (nu-np) and pi for an example of 10 particles.

10
2

0.1
6 5
4

nu - nd
2
0
p

0.01
6
4 -5
2

0.001 -10
0 2 4 6 8 10
nu

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 99

 ((n − nd ) − nu − nd )
2
The variance in nu − nd is defined as σ ud
2
≡ u

We already have shown that nu − nd = 0 , so ≡ (nu − nd )

2
σ ud
2

So
N!
2
σ ud
N
= ∑ ( 2nu − N )
2 (N − nu )!nu !
nu =0 2N

( )
1 N N!
= ∑ 4nu2 − 4Nnu + N 2
2N nu =0 (N − nu )!nu !
1 ⎧⎪ N 2 N! N N! 2
N N! ⎫⎪
= N ⎨4 ∑ nu − 4N ∑ nu +N ∑ ⎬
2 ⎪⎩ nu =0 (N − nu )!nu ! nu =0 (N − nu ) !nu ! nu =0 (N − nu )!n !
u ⎪ ⎭
1
{
= N 4 ⎡⎣N(N + 1)2N −2 ⎤⎦ − 4N ⎡⎣N 2N −1⎤⎦ + N 2 ⎡⎣2N ⎤⎦
2
}
= N(N + 1) − 2N 2 + N 2
=N

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 100

The variance in ( nu − nd ) is equal to the total number of particles in the system.
This means that the standard deviation, or typical size of a fluctuation in this
variable, is

σ ud = N

Often, it is not so much the absolute size of a fluctuation that is important, but
rather the size relative to the size of the system. To put it on a relative basis, we
can write

σ ud 1
=
N N

So for a system with ~ 1022 particles, the typical fluctuation of (nu − nd ) / N away
from 0 will be ~ 10 −11.

For a system with ~ 102 particles, the typical fluctuation of (nu − nd ) / N away from
0 will be ~ 10 −1.

For any property X, the relative size of the fluctuations is σ X ∼ 1

X N

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 101

The comparison of the probability distributions for systems of size N = 10 (•) and
N = 100 ( ) are shown in the graph below.

nu
0 20 40 60 80 100
-2
10
-6
10
-10
10
-14
10
p

-18
10
-22
10
-26
10
-30
10
0 2 4 6 8 10
nu

The behavior of relative fluctuations with system size has some implications for
nanotechnology.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 102

2.5.2 Microscopic (Statistical) Viewpoint of Energy and Temperature

In reality, all systems (including the particles in an ideal gas) are quantum
systems with discrete energy levels.

The total energy of the a system of particles is given by

∞
total
E = ∑ njE j
j =1

where: Ej = energy associated with quantum state j

nj = number of particles in quantum state j

Examples – Energy Levels:

Particle in a box (ideal gas): Ej =

(
h 2 nx2,j + ny2,j + nz2,j ) = h (n
2 2
x ,j + ny2,j + nz2,j
2
)
8mL2 8mV 3

⎛ 1 ⎞ hω
Harmonic oscillator (solids): E j = ⎜ nj + ⎟
⎝ 2 ⎠ 2π

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 103

Examples – Probability Distributions:

⎡ E ⎤
Boltzmann statistics: n Bj ∝ exp ⎢ − j ⎥
⎣ kT ⎦

1
Bose-Einstein statistics (even spin): n BE ∝
⎡E ⎤
j
exp ⎢ j ⎥ −1
⎣ kT ⎦

1
Fermi-Dirac statistics (odd spin): n FD ∝
⎡ Ej ⎤
j
exp ⎢ ⎥ +1
⎣ kT ⎦

Note:

Ej
As → 0 ("high" temperatures), n BE
j → n Bj , and n FD
j → n Bj
kT

"High" temperature is defined by the ratio (Ej/kT). For most cases this is true for
temperatures above 100 K.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 104

Temperature is related to the distribution of particles among the available
quantum states.

Heat denotes energy transfer that occurs because the distributions among the
available energy levels are different for two objects that are in thermal contact.

Work denotes energy that is transferred to the system because of changes in the
energy levels in the system.

Entropy is related to the number of different ways that a system can have the
same total energy.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 105

Example:

Consider a system with four particles distributed among two energy levels, with
energies E0 = 0 and E1 = 1. In the Boltzmann distribution (the simplest of several
distribution functions), the probability that an individual particle has energy Ei is

⎡ E ⎤ ⎡ E ⎤ ⎡ E ⎤
exp ⎢ − i ⎥ exp ⎢ − i ⎥ exp ⎢ − i ⎥
P (E i ) = ⎣ kT ⎦ = ⎣ kT ⎦ = ⎣ kT ⎦
⎡ E ⎤ ⎡ E ⎤ ⎡ E ⎤ ⎡ 1 ⎤
∑ exp ⎢ − k ⎥ exp ⎢ − 0 ⎥ + exp ⎢ − 1 ⎥ 1+ exp ⎢ −
all ⎣ kT ⎦ ⎣ kT ⎦ ⎣ kT ⎦ ⎣ kT ⎥⎦
energy
levels

The particles then are distributed as shown below for several different
temperatures.
E = 1: x xx xxx xxxx

E = 0: xxxx xxx xx x

kT: → 0+ = 0.910 → ∞ +,− = −0.910 → 0−

1/kT: →∞ = 1.099 → 0 +,− = −1.099 → −∞

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 106

Possible arrangements:

E total = 0 :
S
0 0 0 0 ⇒ Ω =1 ⇒ = lnΩ = 0
kB
E total = 1:
1 0 0 0 ⎫
⎪
0 1 0 0 ⎪ S
⎬ ⇒ Ω=4 ⇒ = 5.54
0 0 1 0 ⎪ kB
0 0 0 1 ⎪⎭
E total = 2 :
1 1 0 0 ⎫
⎪
1 0 1 0 ⎪
⎪
1 0 0 1 ⎪⎪ S
⎬ ⇒ Ω=6 ⇒ = 7.17
0 1 1 0 ⎪ kB
⎪
0 1 0 1 ⎪
⎪
0 0 1 1 ⎪⎭

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 107

E total = 3 :
1 1 1 0 ⎫
⎪
1 1 0 1 ⎪ S
⎬ ⇒ Ω=4 ⇒ = 5.54
1 0 1 1 ⎪ kB
0 1 1 1 ⎪⎭

E total = 4 :
S
1 1 1 1 ⇒ Ω =1 ⇒ =0
kB

Note that states 3 and 4 cannot be reached by isolated systems moving toward
equilibrium.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 108

2.6 MATHEMATICAL IMPLICATIONS OF THERMODYNAMIC LAWS

2.6.1 Functions

Function: y is said to be a function of x if there is a unique value of y associated

with every value of x.

Notation: y = f ( x )

where: y is the dependent variable

x is the independent variable

NOTE: the choice of independent variables often is quite arbitrary in

thermodynamics problems.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 109

 y2 •2
y = f1(x)

y1 1•
y = f2(x)

x1 x2

NOTE:

y1 = f 1 ( x1) = f 2 ( x1)

y 2 = f1 (x 2 ) = f 2 (x 2 )

but, in general, f1( x ) ≠ f 2 ( x )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 110

Observations:

Integrals with values independent of path:

2
⌠
⎮ dx
⌡ = x 2 − x1
1

2
⌠
⎮ dy
⌡ = y 2 − y 1 = f1( x 2 ) − f1( x 1) = f 2 ( x 2 ) − f 2 ( x 1)
1

Path dependent integrals:

2
/ i = si
Path length = ⌠⎮⌡ ds
1

2 2
/ i = ⌠⎮⌡ f i ( x ) dx
Area under curve = ⌠⎮⌡ dA
1 1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 111

Terminology:

dφ is an exact differential if the value of the integral does not depend upon the
choice of paths from 1 to 2.

d/ φ is an inexact differential if the value of the integral depends upon the choice
of paths from 1 to 2.

Notation: use d/ φ to denote inexact differentials.

NOTE: ∫ d φ = 0 for an exact differential for any closed path.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 112

Practical Significance:

Each state of a system is described by a unique set of state variables. As a

result, the state variables must behave as mathematical functions.

Notes:

• The differentials of all state functions are exact differentials.

• The energy transferred as heat or work depends upon the choice of paths from
/
state 1 to state 2, therefore dW / are inexact differentials.
and dQ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 113

2.6.2 Multivariable calculus

A dependent variable may be a function of several independent variables:

z = f ( x1,x 2 ,x 3,,x n )
or
F ( x1,x 2 ,x 3,,x n ,x n+1) = 0

Total Derivative:
dz = lim {f ( x1 + Δx1,x 2 + Δx 2 ,,x n + Δx n ) − f ( x1,x 2 ,,x n )}
Δx i →0

⎧f ( x 1 + Δx 1,x 2 + Δx 2 ,,x n + Δx n ) − f ( x 1,x 2 + Δx 2 ,,x n + Δx n )⎫

⎪+f ( x ,x + Δx ,,x + Δx ) − f ( x ,x ,,x + Δx ) ⎪
⎪ 1 2 2 n n 1 2 n n ⎪
= lim ⎨ ⎬
Δx →0 + − 
⎪i ⎪
⎩+f ( x 1,x 2 ,,x n + Δx n ) − f ( x 1,x 2 ,,x n )
⎪ ⎪
⎭

⎛ ∂f ⎞ ⎧ f ( x + Δx,y ) − f ( x,y ) ⎫
Definition of partial derivative: ⎜⎝ ⎟ = lim ⎨ ⎬
∂ x ⎠ y Δx →0 ⎩ Δx ⎭

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 114

 ⎛ ∂f ⎞ n
Then: dz = lim ∑ ⎜ ⎟ Δx i
Δx i →0 i =1⎝ ∂ x i ⎠
x k ≠i

Also: lim Δx i = dx i
Δx i →0

Therefore

⎛ ∂z ⎞
n
dz = ∑ ⎜ ⎟ dx i
i =1⎝ ∂ x i ⎠
x k ≠i

If there are two independent variables:

⎛ ∂z ⎞ ⎛ ∂z ⎞
z = f ( x,y ) ⇔ dz = ⎜ +
⎝ ∂ x ⎟⎠ y
dx ⎜⎝ ∂ y ⎟⎠ dy
x

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 115

For physical systems, we often use equations of the form:

f ( x,y,z ) = 0

The choice of independent variables in this case is arbitrary, so we could use any
of:
x = F ( y,z )
y = G ( x,z )
z = H ( x,y )

Then

⎛ ∂z ⎞ ⎛ ∂z ⎞
z ( x,y ) → dz = ⎜ +
⎝ ∂ x ⎟⎠ y
dx ⎜⎝ ∂ y ⎟⎠ dy
x
⎛ ∂y ⎞ ⎛ ∂y ⎞
y ( x,z ) → dy = ⎜ +
⎝ ∂ x ⎟⎠ z ⎜⎝ ⎟⎠ dz
dx
∂z x

Combine these by substituting dy from the second equation into the first
equation.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 116

Then:

⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎡⎛ ∂ y ⎞ ⎛ ∂y ⎞ ⎤
dz = ⎜ + +
⎝ ∂ x ⎟⎠ y ⎢⎜⎝ ⎟ ⎜⎝ ⎟
dx dx dz
⎜⎝ ∂ y ⎟⎠
x ⎣ ∂x ⎠z ∂ z ⎠ x ⎥⎦

After collecting like terms:

⎡ ⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎤ ⎡⎛ ∂ z ⎞ ⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎤
dz ⎢1− ⎜ ⎟ ⎜ ⎟ ⎥ = dx ⎢⎜ ⎟ + ⎜ ⎟ ⎜ ⎟ ⎥
⎢⎣ ⎝ ∂ y ⎠ x ⎝ ∂ z ⎠ x ⎥⎦ ⎢⎣⎝ ∂ x ⎠ y ⎝ ∂ y ⎠ x ⎝ ∂ x ⎠ z ⎥⎦

x,z could be chosen as the independent variables, therefore the above equation
must hold for arbitrary choices of dx, dz.

−1
⎡ ⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎤ ⎛ ∂z ⎞ ⎡⎛ ∂ y ⎞ ⎤
⎢1− ⎜ ⎟ ⎜⎝ ⎟⎠ ⎥ = 0 → ⎜⎝ ∂ y ⎟⎠ = ⎢⎜⎝ ∂ z ⎟⎠ ⎥
⎢⎣ ⎝ ∂ y ⎠ x ∂ z x ⎥⎦ x ⎣ x⎦

⎡⎛ ∂ z ⎞ ⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎤ ⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎛ ∂x ⎞
⎢⎜ ⎟ + ⎜ ⎟ ⎜ ⎟ ⎥ = 0 → ⎜⎝ ∂ y ⎟⎠ ⎜⎝ ∂ x ⎟⎠ ⎜⎝ ∂ z ⎟⎠ = −1
⎢⎣⎝ ∂ x ⎠ y ⎝ ∂ y ⎠ x ⎝ ∂ x ⎠ z ⎥⎦ x z y

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 117

Consider another variable: w = w ( x,y )
⎛ ∂z ⎞ ⎛ ∂z ⎞
Then, begin with z ( x,y ) ⇔ dz = ⎜ +
⎝ ∂ x ⎟⎠ y
dx ⎜⎝ ∂ y ⎟⎠ dy
x

dz ⎛ ∂ z ⎞ dx ⎛ ∂ z ⎞ dy
Divide by dw: =⎜ ⎟ +
dw ⎝ ∂ x ⎠ y dw ⎜⎝ ∂ y ⎟⎠ x dw

⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂x ⎞
Restrict to the case of constant y: ⎜⎝ ⎟ = ⎜ ⎟ ⎜ ⎟
∂w ⎠ y ⎝ ∂ x ⎠ y ⎝ ∂w ⎠ y

If the original equation is divided instead by dy:

dz ⎛ ∂ z ⎞ dx ⎛ ∂ z ⎞ dy ⎛ ∂ z ⎞ dx ⎛ ∂ z ⎞
=⎜ ⎟ + =⎜ ⎟ +
dy ⎝ ∂ x ⎠ y dy ⎜⎝ ∂ y ⎟⎠ x dy ⎝ ∂ x ⎠ y dy ⎜⎝ ∂ y ⎟⎠ x

⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂x ⎞ ⎛ ∂z ⎞
When this is restricted to constant w: ⎜⎝ ∂ y ⎟⎠ = ⎜⎝ ⎟⎠ ⎜ ⎟ + ⎜⎝ ∂ y ⎟⎠
w
∂ x y ⎝ ∂ y ⎠w x

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 118

For any state function, f, the value is defined uniquely at any state of the system,
therefore:

⌠
⎮
 df
⌡ =0

for any closed path in the state variable space.

Vector Notation:

F = (F1,F2 ,,Fn )

dx = (dx 1,dx 2 ,,dx n )
  n
F • dx = ∑ F i dx i
i =1
 ⎛ ∂ ∂ ∂ ⎞
∇=⎜ , ,,
⎝ ∂ x1 ∂ x 2 ∂ x n ⎟⎠
 ⎛ ∂f ∂f ∂f ⎞
∇f = ⎜ , ,,
⎝ ∂ x1 ∂ x 2 ∂ x n ⎟⎠
  ⎡⎛ ∂F ∂F ⎞ ⎛ ∂F ∂F ⎞ ⎛ ∂F1 ∂F ⎞ ⎤
∇ × F = ⎢⎜ n − 2 ⎟ ,,⎜ i −1 − i +1 ⎟ ,,⎜ − n−1 ⎟ ⎥
⎣⎝ ∂x 2 ∂x n ⎠ ⎝ ∂x i +1 ∂x i −1 ⎠ ⎝ ∂x n−1 ∂x 1 ⎠ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 119

    
From advanced calculus, Stokes Theorem gives: F •d x
⌠
⎮
⌡

= ⌠⌠
⎮⎮
⌡⌡ ( )
∇ ×F •dA

n⎛ ∂f ⎞  
Observe that df = ∑ ⎜ ⎟ dx i = ∇f • dx
i =1⎝ ∂ x i ⎠ x
k ≠i

  ⌠⌠   
Then, by Stoke's theorem: ⌠
 df
⎮
⌡
= ⌠
 ∇f
⎮
⌡ ( )
• d x = ⎮⌡⎮⌡ ∇ × ∇f • d A

  ∂2 f ∂2 f ⎡x 0 → xn ⎤
Where ( ∇ × ∇f )i = −
∂x i −1 ∂x i +1 ∂x i +1 ∂x i −1
; ⎢x ⎥
⎣ n+1 → x 1⎦

⌠ ⌠⌠ n ⎡ ⎤
⎮ ⎮⎮ ∂2 f ∂2 f
Then  df
⎮ = ⎮⎮ ∑ ⎢ − ⎥ dAi
⎮ ⎮⎮ i =1⎢ ∂x i −1 ∂x i +1 ∂x i +1 ∂x i −1 ⎥⎦
⌡ ⌡⌡ ⎣

But, because ⌠⎮⌡ df = 0 for any state function for any path and surface, the term in
square brackets in the integrand always must be zero.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 120

Therefore:

∂2 f ∂2 f
=
∂x i ∂x k ∂x k ∂x i

This requires that, if f is a state function, the order of differentiation with respect
to the independent variables does not matter.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 121

2.6.2.1 Summary of Math Results

⎛ ∂f ⎞
n
Math 1: df = ∑ ⎜ ⎟ dx i ⇔ f ( x 1,x 2 ,,x n )
i =1⎝ ∂ x i ⎠ x
k ≠i

−1
⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎡⎛ ∂ z ⎞ ⎤ ⎛ ∂y ⎞
Math 2: ⎜⎝ ∂ y ⎟⎠ ⎜⎝ ∂ z ⎟⎠ = ⇔ ⎢⎜ ⎟⎠ ⎥ =⎜
⎝ ∂ z ⎟⎠ x
1
x x ⎢⎣ ⎝ ∂ y x⎥
⎦

⎛ ∂z ⎞ ⎛ ∂y ⎞ ⎛ ∂x ⎞
Math 3: ⎜⎝ ∂ y ⎟⎠ ⎜⎝ ∂ x ⎟⎠ ⎜⎝ ∂ z ⎟⎠ = −1
x z y

⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂x ⎞
Math 4: ⎜⎝ ⎟⎠ = ⎜⎝ ⎟ ⎜ ⎟
∂w y ∂ x ⎠ y ⎝ ∂w ⎠ y

⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂z ⎞ ⎛ ∂x ⎞
Math 5: ⎜⎝ ∂ y ⎟⎠ = ⎜⎝ ∂ y ⎟⎠ + ⎜⎝ ⎟⎠ ⎜ ⎟⎠
w x
∂ x y ⎝ ∂ y w

∂2 f ∂2 f
Math 6: =
∂x i ∂x k ∂x k ∂x i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 122

2.6.2.2 Legendre Transformations

In many practical cases, we will find it convenient to change to a different set of

independent variables.
n ⎛ ⎞
Math 1 ® y = f ( x 1,x 2 ,,x n ) ⇔ dy = ∑ Ci dx i where: Ci = ⎜ ∂ f ⎟
i =1 ⎝ ∂ xi ⎠ x
k ≠i

Define a new function by: τ i = y − Ci x i

n n
Then: d τ i = dy − Ci dx i − x i dCi = ∑ Ck dx k − Ci dx i − x i dCi = ∑ Ck dx k − x i dCi
k =1 k =1
k ≠i

Math 1 ® τ i = f ( x 1,x 2 ,,Ci ,,x n )

The new function replaces xi with Ci as an independent variable.

Note that multiple transformations are allowed.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 123

2.6.3 Applications of Mathematics to Physics Summary

The fundamental differential form of the summary of the First and Second Laws
of Thermodynamics (Physics Summary Statement) for a closed system is:

dU = TdS − PdV
⎛ ∂f n ⎞
Compare to Math 1: df = ∑ ⎜ ⎟⎠ dx i ⇔ f ( x 1,,x n )
i =1⎝ ∂ x i

The form of the Physics Summary Statement indicates that the "natural"
independent variables for internal energy are entropy and volume. In general,
manipulations will be more convenient when guided by the natural independent
variables.

Result: dU = TdS − PdV ⇒ U = U (S ,V )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 124

We would prefer U = U(T,P) because pressure and temperature are convenient
from a practical point of view, and the choice of variables is arbitrary from the
mathematical point of view.

Math 1 ® ⎛ ∂U ⎞ ⎛ ∂U ⎞
Then: dU = ⎜ +
⎝ ∂T ⎟⎠ P ⎜⎝ ⎟⎠ dP
dT
∂P T

∂U ⎞ ⎛ ∂U ⎞ with other
However, we cannot easily identify ⎛⎜ and
⎝ ∂T ⎟⎠ P ⎜⎝ ⎟
∂ P ⎠T
thermodynamic quantities.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 125

2.6.3.1 Legendre Transformations of Internal Energy:

Legendre transformations of the internal energy provide energy functions that

have other natural independent variables

dU = TdS − PdV ⎫ ⎧C1 = T x1 = S

⎬ ⇒ ⎨
U = U (S,V ) ⎭ ⎩C2 = −P x 2 =V

Replace S with T:

τ 1 = U − C1x 1 = U −TS
d τ 1 = dU −TdS − SdT = TdS − PdV −TdS − SdT
d τ 1 = −SdT − PdV ⇒ τ 1 = τ 1 (T ,V )

Result: A (T ,V ) = U − TS (Helmholtz Free Energy )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 126

Replace V with P:

τ 2 = U − C2 x 2 = U − ( −P )V = U + PV
d τ 2 = dU +VdP + PdV = TdS − PdV +VdP + PdV
d τ 2 = TdS +VdP ⇒ τ 2 = τ 2 (S,P )

Result: H (S ,P ) = U + PV (Enthalpy )

Replace S with T and V with P

τ 12 = U − C1 x 1 − C2 x 2 = U −TS + PV = A + PV = H −TS
d τ 12 = −SdT +VdP ⇒ τ 12 = τ 12 (T ,P )

Result: G (T ,P ) = U + PV − TS ( Gibbs Free Energy )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 127

2.6.3.1.1 Alternative Forms of Physics Summary Statement

dU = TdS − PdV ⇔ U = U (S ,V )

dH = TdS +VdP ⇔ H = H (S ,P )

dA = −SdT − PdV ⇔ A = A (T ,V )

dG = −SdT +VdP ⇔ G = G (T ,P )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 128

2.6.3.1.2 Extension to Mixtures and Open Systems:

U total = nU = U total (nS,nV ,n1,,nm )

where: m = number of chemical species present.

Math 1 requires:
⎛ ∂ (nU ) ⎞ ⎛ ∂ (nU ) ⎞ ⎛ ∂ (nU ) ⎞
m
d (nU ) = ⎜ d (nS ) + ⎜ d (nV )+ ∑ ⎜ ⎟ dn i
⎝ ∂ (nS ) ⎟⎠ nV ,{nk }
⎝ ∂ (nV ) ⎟⎠ nS,{nk } i =1⎝ ∂ ni ⎠ nS,nV ,n
k ≠i

In the first two terms, {ni} held constant implies that n also is held constant.

⎛ ∂ (nU ) ⎞ ⎛ ∂U ⎞ ⎛ ∂ (nU ) ⎞ ⎛ ∂U ⎞
Then =⎜ =⎜
⎝ ∂ S ⎟⎠V ,n ⎝ ∂V ⎟⎠ S,n
⎜⎝ ∂ (nS ) ⎟⎠ and ⎜⎝ ∂ (nV ) ⎟⎠
{ }
nV , ni { }
nS, ni

Notation: Define: µi ≡ ⎜ ( ) ⎟
⎛ ∂ nU ⎞
where µi is the chemical potential
⎝ ∂ n i ⎠ nS,nV ,n
k ≠i

m
Then: d (nU ) = Td (nS ) − Pd (nV ) + ∑ µi dn i
i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 129

2.6.3.1.3 Physics Summary Statements for Open Systems:

m
d (nU ) = T d (nS ) − P d (nV ) + ∑ µi dn i
i =1
m
d (nH ) = T d (nS ) + (nV ) dP + ∑ µi dn i
i =1
m
d (nA ) = − (nS ) dT − P d (nV ) + ∑ µi dn i
i =1
m
d (nG ) = − (nS ) dT + (nV ) dP + ∑ µi dn i
i =1

where:

nU = nU (nS,nV ,n1,n2 ,,nm )

nH = nH (nS,P,n1,n2 ,,nm )
nA = nA (T ,nV ,n1,n2 ,,nm )
nG = nG (T ,P,n1,n2 ,,nm )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 130

2.6.3.2 Thermodynamic Definitions

Combine Math 1 with the physics summary statements to identify partial

derivatives of energy functions with physical variables:

Temperature:

⎛ ∂ (nU ) ⎞ ⎛ ∂U ⎞
T =⎜ =⎜
⎝ ∂ (nS ) ⎟⎠ ⎝ ∂ S ⎟⎠ n,V
{ }
nV , ni

⎛ ∂ (nH ) ⎞ ⎛ ∂H ⎞
=⎜ =⎜
⎝ ∂ (nS ) ⎟⎠ ⎝ ∂ S ⎟⎠ n,P
{ }
P, ni

Pressure:

⎛ ∂ (nU ) ⎞ ⎛ ∂U ⎞
−P = ⎜ =⎜
⎝ ∂ (nV ) ⎟⎠ ⎝ ∂V ⎟⎠ n,S
{ }
nS, ni

⎛ ∂ (nA ) ⎞ ⎛ ∂A⎞
=⎜ =⎜
⎝ ∂ (nV ) ⎟⎠ ⎝ ∂V ⎟⎠ n,T
{ }
T , ni

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 131

Volume:

⎛ ∂H ⎞ ⎛ ∂G ⎞
V =⎜ =⎜
⎝ ∂ P ⎠ S,n ⎝ ∂ P ⎟⎠ T ,n
⎟

Entropy:

⎛ ∂A⎞ ⎛ ∂G ⎞
−S = ⎜ =
⎝ ∂T ⎟⎠V ,n ⎜⎝ ∂T ⎟⎠ P,n

Chemical Potential:

⎛ ∂ (nU ) ⎞ ⎛ ∂ (nH ) ⎞ ⎛ ∂ (nA ) ⎞ ⎛ ∂ (nG ) ⎞

µi = ⎜ ⎟ =⎜ ⎟ =⎜ ⎟ =⎜ ⎟
⎝ ∂ n i ⎠ nV ,nS,n ⎝ ∂ n i ⎠ nS,P,n ⎝ ∂ n i ⎠ T ,nV ,n ⎝ ∂ n i ⎠ T ,P,n
k ≠i k ≠i k ≠i k ≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 132

2.6.3.3 Maxwell Relations

The Maxwell Relations follow from Math 6 applied to the differential forms of the
Physics Summary Statements:

⎛ ∂T ⎞ ⎛ ∂P ⎞
dU = TdS − PdV ⇒ ⎜⎝ ⎟ = − ⎜⎝ ⎟
∂V ⎠ S ∂ S ⎠V

⎛ ∂T ⎞ ⎛ ∂V ⎞
dH = TdS +VdP ⇒ ⎜⎝ ⎟ = ⎜ ⎟
∂P ⎠ S ⎝ ∂S ⎠ P

⎛ ∂S ⎞ ⎛ ∂P ⎞
dA = −SdT − PdV ⇒ ⎜⎝ ⎟ = ⎜ ⎟ (MxA)
∂V ⎠ T ⎝ ∂T ⎠V

⎛ ∂S ⎞ ⎛ ∂V ⎞
dG = −SdT +VdP ⇒ ⎜⎝ ⎟⎠ = − ⎜⎝ ⎟ (MxG)
∂P T ∂T ⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 133

2.6.3.4 Partial Molar Properties

Begin with the differential form of the Gibbs summary statement for an open
system:
m
d (nG ) = − (nS ) dT + (nV ) dP + ∑ µi dn i
i =1

⎛ ∂ (nG ) ⎞
µi = ⎜ ⎟
⎝ ∂ n i ⎠ T ,P,n
k ≠i

⎛ ∂µi ⎞ ⎛ ∂ (nS ) ⎞
⎜⎝ ⎟ = −⎜ ⎟
∂T ⎠ P,n ⎝ ∂ n i ⎠ T ,P,n
k ≠i

⎛ ∂µi ⎞ ⎛ ∂ (nV ) ⎞
⎜⎝ ⎟⎠ = ⎜ ⎟
∂ P T ,n ⎝ ∂ n i ⎠ T ,P,n
k ≠i

⎛ ∂ (nX ) ⎞
Definition: Xi ≡ ⎜ ⎟ where: X i is the partial molar quantity
⎝ ∂ n i ⎠ T ,P,n
k ≠i

(Partial molar properties are very useful for describing mixture properties.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 134

2.6.4 Property Change Calculations

2.6.4.1 General Approach

For practical applications of the First and Second Laws of Thermodynamics, the
change in physical properties accompanying a change of state is required. For
any physical property, Y, the change is calculated using
2
ΔΨ = Ψ ( 2 ) − Ψ (1) = ⎮dΨ
⌠
⌡
1
where: Y may be any state function, but most commonly is U, H, A, G or S.
∆Y is the change in the property Y
Y(i) is the value of property Y for state i

For pure fluids (or mixtures at fixed composition) the convenient independent
variables are either (T,P) or (T,V) as the independent variables.
2
⌠⎡ ⎤
⎮ ⎛ ∂Ψ ⎞ ⎛ ∂Ψ ⎞
Math Relation 1: Ψ ( 2 ) − Ψ (1) = ⎢⎜ ⎟ dT + ⎜⎝ ⎟ dX ⎥
⎮ ⎝
⎮
∂T ⎠ X ∂X ⎠ T
⌡⎣ ⎦
1
where X denotes either P or V.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 135

Because the physical properties are state functions, we may choose any path
between state 1 and state 2 to evaluate the integral. The most convenient path
for the calculation is to vary only one of the independent variables at a time:

T2 X2
⌠ ⌠
⎛ ∂Ψ ⎞ ⎛ ∂Ψ ⎞
Ψ (T2 ,X 2 ) − Ψ (T1,X 1) = ⎮
⎜⎝ ⎟ dT +
⎮
⎜⎝ ⎟ dX
⎮
⎮
⌡ ∂T ⎠ X
1
⎮
⎮
⌡ ∂X ⎠ T
2
T1 X1

or

X2 T2
⌠ ⌠
⎛ ∂Ψ ⎞ ⎛ ∂Ψ ⎞
Ψ (T2 ,X 2 ) − Ψ (T1,X 1) = ⎮
⎮ ⎜⎝ ⎟⎠ dX +
⎮
⎮ ⎜⎝ ⎟⎠ dT
⎮
⌡ ∂X T
1
⎮
⌡ ∂T X
2
X1 T1

However, to calculate the values of these integrals we must relate the partial
derivatives in the integrals to properties that are measurable.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 136

2.6.4.2 Manipulation of Thermodynamic Variables

Thermodynamic variables can be divided into two classes:

Physical Observables: variables that can be measured in the laboratory

Examples: P, V, T, CP, CV.

Intangibles: variables that cannot be measured or observed directly

Examples: U, H, A, G, S.

Because the intangibles cannot be measured directly, they must be expressed in

terms of the physical observables through manipulations of the thermodynamic
variables, usually in the form of partial derivatives.

These manipulations are made easier using a systematic approach as shown

below.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 137

2.6.4.2.1 Common Physical Observables:
In addition to the basic variables P, V and T, there are several derivatives that
can be measured in the laboratory:
1 ⎛ ∂V ⎞
Thermal expansion coefficient: β≡ ⎜ ⎟
V ⎝ ∂T ⎠ P
1 ⎛ ∂V ⎞
Isothermal compressibility: κT ≡ − ⎜ ⎟
V ⎝ ∂ P ⎠T
1 ⎛ ∂V ⎞
Isentropic compressibility: κS ≡ − ⎜ ⎟
V ⎝ ∂P ⎠ S

⎛ ∂P ⎞
Bulk modulus (isothermal): BT ≡ −V ⎜
⎝ ∂V ⎟⎠ T

⎛ ∂P ⎞
Bulk modulus (adiabatic): BS ≡ −V ⎜
⎝ ∂V ⎟⎠ S

1 ⎛ ∂P ⎞
Thermal pressure coefficient: γ ≡ ⎜ ⎟
P ⎝ ∂T ⎠ V
dQ
Heat capacities: CX ≡
dT Constant X

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 138

2.6.4.2.1.1 Heat Capacity at Constant Volume:

⎛ ∂ Qrev ⎞
Definition: CV ≡ ⎜
⎝ ∂T ⎟⎠V

Begin with differential form of First Law: / rev + dW

dU = dQ / rev = dQ
/ rev − PdV

Rearrange: / rev = dU + PdV

dQ

/ rev dU
dQ dV
Divide by dT: = +P
dT dT dT

⎛ ∂Qrev ⎞ ⎛ ∂U ⎞ ⎛ ∂V ⎞
Restrict to constant volume: ⎜⎝ ⎟ = ⎜ ⎟ + P ⎜⎝ ⎟
∂T ⎠V ⎝ ∂T ⎠V ∂T ⎠V

⎛ ∂U ⎞
Therefore: CV = ⎜
⎝ ∂T ⎟⎠ V

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 139

Use the Physics Summary Statement in the form:

dU = TdS − PdV

Divide by dT and restrict to constant V

⎛ ∂U ⎞ ⎛ ∂S ⎞ ⎛ ∂V ⎞
⎜⎝ ⎟ = T ⎜⎝ ⎟ − P ⎜⎝ ⎟
∂T ⎠V ∂T ⎠V ∂T ⎠V

⎛ ∂U ⎞ ⎛ ∂S ⎞
Therefore: CV = ⎜ = T
⎝ ∂T ⎟⎠ V ⎜⎝ ⎟
∂T ⎠ V

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 140

2.6.4.2.1.2 Heat Capacity at Constant Pressure:

⎛ ∂ Qrev ⎞
Definition: CP ≡ ⎜
⎝ ∂T ⎟⎠ P

Begin with differential form of First Law

/ rev + dW
dU = dQ / rev = dQ
/ rev − PdV ⇒ dQ
/ rev = dU + PdV

Divide by dT and restrict to constant pressure

⎛ ∂Qrev ⎞ ⎛ ∂U ⎞ ⎛ ∂V ⎞
⎜⎝ ⎟ = ⎜ ⎟ + P ⎜⎝ ⎟
∂T ⎠ P ⎝ ∂T ⎠ P ∂T ⎠ P

Note that
⎛ ∂H ⎞ ⎛ ∂U ⎞ ⎛ ∂V ⎞ ⎛ ∂P ⎞ ⎛ ∂Qrev ⎞
H = U + PV ⇒ ⎜ = + +V =
⎝ ∂T ⎟⎠ P ⎜⎝ ∂T ⎟⎠ P ⎜⎝ ⎟ ⎜⎝ ⎟ ⎜ ⎟
P
∂T ⎠ P ∂T ⎠ P ⎝ ∂T ⎠ P

⎛ ∂H ⎞
Therefore: CP = ⎜
⎝ ∂T ⎟⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 141

Use the Physics Summary Statement in the form:

dH = TdS +VdP

Divide by dT and restrict to constant P

⎛ ∂H ⎞ ⎛ ∂S ⎞ ⎛ ∂P ⎞
⎜⎝ ⎟⎠ = T ⎜⎝ ⎟⎠ +V ⎜⎝ ⎟
∂T P ∂T P ∂T ⎠ P

⎛ ∂H ⎞ ⎛ ∂S ⎞
Therefore: CP = ⎜ = T
⎝ ∂T ⎟⎠ P ⎜⎝ ⎟
∂T ⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 142

2.6.4.2.1.3 Relations Between Heat Capacities - (CP - CV)

Use relations between heat capacities and temperature derivatives of entropy

⎛ ∂S ⎞ ⎛ ∂S ⎞
CP − CV = T ⎜ ⎟ −T ⎜
⎝ ∂T ⎠ P ⎝ ∂T ⎟⎠V

Apply Math Relation 5 to the first term, then

⎡⎛ ∂S ⎞ ⎛ ∂S ⎞ ⎛ ∂V ⎞ ⎤ ⎛ ∂S ⎞
CP − CV = T ⎢⎜ ⎟ +⎜ ⎟⎠ ⎜⎝ ⎟⎠ ⎥ −T ⎜⎝ ⎟
⎣ ⎝ ∂T ⎠ V
⎝ ∂V T ∂T P ⎦ ∂T ⎠V

Apply the Maxwell Relation derived from the Helmholtz energy

⎛ ∂S ⎞ ⎛ ∂V ⎞ ⎛ ∂P ⎞ ⎛ ∂V ⎞
CP − CV = T ⎜ ⎟⎠ = T ⎜⎝
⎝ ∂V ⎟⎠ T ⎜⎝
∂T P
⎟
∂T ⎠V
⎜⎝ ⎟
∂T ⎠ P

Therefore

⎛ ∂P ⎞ ⎛ ∂V ⎞ ⎛ ∂P ⎞ TV β 2
C P − CV = T ⎜ = TV β ⎜ = TV β BT =
2
⎝ ∂T ⎟⎠ V ⎜⎝ ∂T ⎟⎠ P ⎝ ∂T ⎟⎠ V κT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 143

2.6.4.2.1.4 Relations Between Heat Capacities - (CP /CV):

Use relations between heat capacity and entropy, the cyclic relation (Math
Relation 3), and the Maxwell relations derived from the Helmholtz and Gibbs
energies
⎛ ∂S ⎞ ⎛ ∂S ⎞ ⎛ ∂P ⎞ ⎛ ∂V ⎞ ⎛ ∂P ⎞
T⎜ − ⎜⎝ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟
CP ⎝ ∂T ⎟⎠ P ∂P ⎠ T ⎝ ∂T ⎠ S ⎝ ∂T ⎠ P ⎝ ∂T ⎠ S
= = =
CV ⎛ ∂S ⎞ ⎛ ∂S ⎞ ⎛ ⎞ ∂V ⎛ ∂P ⎞ ⎛ ∂V ⎞
T⎜ −
⎝ ∂T ⎟⎠V ⎜⎝ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟
∂V ⎠ T ⎝ ∂T ⎠ S ⎝ ∂T ⎠V ⎝ ∂T ⎠ S

Use the chain rule for partial derivatives (Math Relation 4) in the numerator and
the cyclic relation in the denominator, so that

⎛ ∂V ⎞ ⎛ ∂P ⎞ ⎛ ∂V ⎞ ⎡⎛ ∂P ⎞ ⎛ ∂V ⎞ ⎤
⎜⎝ ⎟ ⎜ ⎟ ⎜ ⎟
CP
⎜⎝ ⎟ ⎜ ⎟
∂T ⎠ P ⎝ ∂T ⎠ S ∂T ⎠ P ⎢⎣⎝ ∂V ⎠ S ⎝ ∂T ⎠ S ⎥⎦
=− =−
CV ⎛ ∂P ⎞ ⎛ ∂V ⎞ ⎡⎛ ∂P ⎞ ⎛ ∂V ⎞ ⎤ ⎛ ∂V ⎞
⎜⎝ ⎟⎠ ⎜⎝ ⎟⎠ − ⎢⎜ ⎟⎠ ⎜⎝ ⎟⎠ ⎥ ⎜⎝ ⎟⎠
∂T V ∂T S ⎣ ⎝ ∂V T ∂T P⎦ ∂T S

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 144

Therefore

⎛ ∂P ⎞
C P ⎜⎝ ∂V ⎟⎠ S BS κ T
= = =
CV ⎛ ∂P ⎞ BT κ S
⎜⎝ ⎟⎠
∂V T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 145

2.6.4.2.2 Systematic Approach to Manipulations Involving Derivatives:

• Roman letters (a,b,c) denote the observables (P,V,T),

• Greek letters (a,b,g) denote the intangibles (U,H,A,G,S).

The usual goal is to reduce a given derivative to an expression containing only

observables.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 146

Divide the partial derivatives into six classes according to the six possible
combinations of the variables:

Class Form

⎛ ∂a ⎞
1 ⎜⎝ ⎟⎠
∂b c

⎛ ∂α ⎞
2 ⎜⎝ ⎟
∂b ⎠ c

⎛ ∂a ⎞
3 ⎜⎝ ⎟⎠
∂b γ

⎛ ∂α ⎞
4 ⎜⎝ ⎟⎠
∂β c

⎛ ∂α ⎞
5 ⎜⎝ ⎟⎠
∂b γ

⎛ ∂α ⎞
6 ⎜ ⎟
⎝ ∂β ⎠ γ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 147

Procedures for Manipulation of Variables

Class 1 Derivatives:

General form: ⎛ ∂a ⎞ Example: ⎛ ∂P ⎞

⎜⎝ ⎟⎠ ⎜⎝ ⎟
∂b c ∂V ⎠ T

Procedure: The only useful relation for this class is the cyclic relation (Math
Relation 3):

⎛ ∂a ⎞ ⎛ ∂b ⎞ ⎛ ∂c ⎞
⎜⎝ ⎟⎠ ⎜⎝ ⎟⎠ ⎜⎝ ⎟⎠ = −1
∂b c ∂c a ∂a b

⎛ ∂P ⎞ ⎛ ∂P ⎞ ⎛ ∂T ⎞ γP
Example: ⎜⎝ ⎟ = − ⎜⎝ ⎟ ⎜⎝ ⎟ = −
∂V ⎠ T ∂T ⎠V ∂V ⎠ P βV

βκ T β
Note that this also leads to the conclusion that = =P
γ γ BT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 148

Class 2 Derivatives:

∂α ⎞ ⎛ ∂H ⎞
General form: ⎛⎜ Example: ⎜⎝ ⎟
⎝ ∂b ⎟⎠ c ∂P ⎠ T

Procedures:

If a is S, then use the relations to the heat capacities and the Maxwell relations
directly.

If a is U, H, A or G, then begin with the relevant form of the physics summary

statement, divide through by the differential of interest, and eliminate dS using a
Maxwell Relation if necessary.

Examples involving S:

⎛ ∂S ⎞ CP ⎛ ∂S ⎞ ⎛ ∂P ⎞
⎜⎝ ⎟ = ⎜⎝ ⎟ = ⎜ ⎟
∂T ⎠ P T ∂V ⎠ T ⎝ ∂T ⎠V

⎛ ∂S ⎞ CV ⎛ ∂S ⎞ ⎛ ∂V ⎞
⎜⎝ ⎟⎠ = ⎜⎝ ⎟⎠ = − ⎜⎝ ⎟
∂T V T ∂P T ∂T ⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 149

Example 2.1: ⎛ ∂H ⎞
⎜⎝ ⎟
∂P ⎠ T

Begin with dH = TdS +VdP

dH dS dP
Divide through by dP: =T +V
dP dP dP

⎛ ∂H ⎞ ⎛ ∂S ⎞
Restrict to constant T: ⎜⎝ ⎟⎠ = T ⎜⎝ ⎟⎠ +V
∂P T ∂P T

Apply Maxwell Relation (MxG):

⎛ ∂H ⎞ ⎛ ∂V ⎞
⎜⎝ ⎟⎠ = −T ⎜⎝ ⎟⎠ +V
∂P T ∂T P
⎛ ∂V ⎞
=V −T ⎜
⎝ ∂T ⎟⎠ P
=V [1−T β ]

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 150

Example 2.2: ⎛ ∂H ⎞
⎜⎝ ⎟
∂V ⎠ T

Begin with dH = TdS +VdP

dH dS dP
Divide through by dV: =T +V
dV dV dV

⎛ ∂H ⎞ ⎛ ∂S ⎞ ⎛ ∂P ⎞
Restrict to constant T: ⎜⎝ ⎟ = T ⎜⎝ ⎟ +V ⎜⎝ ⎟
∂V ⎠ T ∂V ⎠ T ∂V ⎠ T

⎛ ∂H ⎞ ⎛ ∂P ⎞ ⎛ ∂P ⎞
Apply Maxwell Relation (MxA): ⎜⎝ ⎟ = T ⎜⎝ ⎟ +V ⎜⎝ ⎟
∂V ⎠ T ∂T ⎠V ∂V ⎠ T

Note that the relations are more complex when a variable is used that is not a
"natural" independent variable.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 151

Class 3 Derivatives:

∂a ∂T
General form: ⎛⎜ ⎞⎟ Example: ⎛⎜ ⎞⎟
⎝ ∂b ⎠ γ ⎝ ∂P ⎠ H

Procedure: Use the cyclic relation (Math Relation 3) to obtain two Class 2
derivatives:

⎛ ∂a ⎞ ⎛ ∂a ⎞ ⎛ ∂γ ⎞
⎜⎝ ⎟⎠ = − ⎜⎝ ∂γ ⎟⎠ ⎜⎝ ∂b ⎟⎠
∂b γ b a

Example: (Joule-Thompson coefficient)

⎛ ∂T ⎞ ⎛ ∂T ⎞ ⎛ ∂H ⎞ ⎛ 1 ⎞ ⎡ ⎛ ∂V ⎞ ⎤
⎜⎝ ⎟ = − ⎜⎝ ⎟ ⎜ ⎟ = ⎢ ⎜
T ⎟ −V ⎥
∂P ⎠ H ∂H ⎠ P ⎝ ∂P ⎠ T ⎜⎝ CP ⎟⎠ ⎣ ⎝ ∂T ⎠ P ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 152

Class 4 Derivatives:

⎛ ∂α ⎞ ∂H
General form: ⎜ Example: ⎛⎜ ⎞⎟
⎝ ∂β ⎟⎠ c ⎝ ∂A ⎠ P

Procedure: Introduce a new observable variable through Math Relation 4 to

produce two Class 2 derivatives:

⎛ ∂α ⎞ ⎛ ∂α ⎞ ⎛ ∂a ⎞
⎜⎝ ∂β ⎟⎠ = ⎜⎝ ⎟⎠ ⎜ ⎟⎠
c
∂a c ⎝ ∂ β c

⎛ ∂H ⎞ ⎛ ∂H ⎞ ⎛ ∂T ⎞ CP
Example: ⎜⎝ ⎟⎠ = ⎜⎝ ⎟⎠ ⎜⎝ ⎟⎠ = −
∂A P ∂T P ∂A P ⎛ ∂V ⎞
S +P ⎜
⎝ ∂T ⎟⎠ P

S is acceptable as a variable here because we will show later how to calculate S

from observables. (Derivatives of S are not acceptable.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 153

Class 5 Derivatives:

∂α ⎞ ∂H ⎞
General form: ⎛⎜ Example: ⎛⎜
⎝ ∂b ⎟⎠ γ ⎝ ∂P ⎟⎠ A

Procedure: Introduce a new observable and use Math Relation 5 to obtain one
Class 3 and two Class 2 derivatives:

⎛ ∂α ⎞ ⎛ ∂α ⎞ ⎛ ∂α ⎞ ⎛ ∂c ⎞
⎜⎝ ⎟ = ⎜ ⎟ + ⎜ ⎟ ⎜ ⎟
∂b ⎠ γ ⎝ ∂b ⎠ c ⎝ ∂c ⎠ b ⎝ ∂b ⎠ γ

Example:
⎛ ∂H ⎞ ⎛ ∂H ⎞ ⎛ ∂H ⎞ ⎛ ∂T ⎞
⎜⎝ ⎟⎠ = ⎜⎝ ⎟⎠ + ⎜⎝ ⎟ ⎜ ⎟
∂P A ∂P T ∂T ⎠ P ⎝ ∂P ⎠ A

⎛ ∂H ⎞ ⎛ ∂H ⎞ ⎛ ∂T ⎞ ⎛ ∂A ⎞
=⎜ −
⎝ ∂P ⎟⎠ T ⎜⎝ ∂T ⎟⎠ P ⎜⎝ ∂A ⎟⎠ P ⎜⎝ ∂P ⎟⎠ T

⎛ ∂V ⎞ ⎛ ∂T ⎞ ⎛ ∂A ⎞
=V −T ⎜ −
⎝ ∂T ⎟⎠ P P⎜
⎝ ∂A ⎟⎠ P ⎜⎝ ∂P ⎟⎠ T
C

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 154

where:

⎧⎛ ∂A ⎞ ⎛ ∂V ⎞
⎪⎜⎝ ∂T ⎟⎠ = −S − P ⎜⎝ ⎟⎠
∂T
⎪⎪ P P
dA = −SdT − PdV ⇒ ⎨
⎪ ∂A
⎪⎛⎜ ⎞ ⎛ ∂T ⎞ ⎛ ∂V ⎞
⎟ = −S ⎜⎝ ⎟ − P ⎜⎝ ⎟
⎪⎩⎝ ∂P ⎠ T ∂P ⎠ T ∂P ⎠ T

Then

⎛ ∂H ⎞ ⎛ ∂V ⎞ ⎛ ∂T ⎞ ⎛ ∂A ⎞
⎜⎝ ⎟ =V −T ⎜⎝ ⎟⎠ − P⎜
⎝ ∂A ⎟⎠ P ⎜⎝ ∂P ⎟⎠ T
C
∂P ⎠ A ∂T P

⎛ ∂V ⎞
CP P ⎜
⎛ ∂V ⎞ ⎝ ∂P ⎟⎠ T
=V −T ⎜ ⎟ −
⎝ ∂T ⎠ P ⎛ ∂V ⎞
S +P ⎜ ⎟
⎝ ∂T ⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 155

Class 6 Derivatives:

General form:
⎛ ∂α ⎞
Example: ⎛ ∂H ⎞
⎜⎝ ∂β ⎟⎠ ⎜⎝ ⎟
γ
∂G ⎠ A

Procedure: Use Math Relation 4 to introduce a new observable thereby

obtaining two Class 5 derivatives:

⎛ ∂α ⎞ ⎛ ∂α ⎞ ⎛ ∂a ⎞
⎜⎝ ∂β ⎟⎠ = ⎜⎝ ⎟⎠ ⎜ ⎟⎠
γ
∂a γ ⎝ ∂ β γ

Example:
⎛ ∂H ⎞
⎜⎝ ⎟
⎛ ∂H ⎞ ⎛ ∂H ⎞ ⎛ ∂P ⎞ ∂P ⎠ A
⎜⎝ ⎟⎠ = ⎜⎝ ⎟⎠ ⎜⎝ ⎟⎠ = ∂G
∂G A ∂P A ∂G A ⎛ ⎞
⎜⎝ ⎟
∂P ⎠ A

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 156

From the previous example

⎛ ∂V ⎞
CP P ⎜
⎛ ∂H ⎞ ⎛ ∂V ⎞ ⎝ ∂P ⎟⎠ T
⎜⎝ ⎟⎠ =V −T ⎜⎝ ⎟⎠ −
∂P A ∂T P ⎛ ∂V ⎞
S +P ⎜
⎝ ∂T ⎟⎠ P

Also:
⎛ ∂G ⎞ ⎛ ∂G ⎞ ⎛ ∂G ⎞ ⎛ ∂T ⎞
⎜⎝ ⎟ = ⎜ ⎟ + ⎜ ⎟ ⎜ ⎟
∂P ⎠ A ⎝ ∂P ⎠ T ⎝ ∂T ⎠ P ⎝ ∂P ⎠ A

⎡ ⎛ ∂T ⎞ ⎛ ∂A ⎞ ⎤ ⎛ ∂T ⎞ ⎛ ∂A ⎞
=V + ( −S ) ⎢ − ⎜ ⎟ ⎜ ⎟ ⎥ =V + S ⎜⎝ ⎟ ⎜ ⎟
⎣ ⎝ ∂A ⎠ P ⎝ ∂P ⎠ T ⎦ ∂A ⎠ P ⎝ ∂P ⎠ T

⎛ ∂V ⎞
SP ⎜
⎝ ∂P ⎟⎠ T
=V +
⎛ ∂V ⎞
S +P ⎜
⎝ ∂T ⎟⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 157

so that
⎛ ∂V ⎞
CP P ⎜
⎛ ∂V ⎞ ⎝ ∂P ⎟⎠ T
V −T ⎜ ⎟ −
⎝ ∂T ⎠ P ⎛ ∂V ⎞
S +P ⎜ ⎟
⎛ ∂H ⎞ ⎝ ∂T ⎠ P
⎜⎝ ⎟ =
∂G ⎠ A ⎛ ∂V ⎞
SP ⎜
⎝ ∂P ⎟⎠ T
V+
⎛ ∂V ⎞
S +P ⎜ ⎟
⎝ ∂T ⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 158

2.6.4.3 Residual functions

An alternative method for calculating physical property changes utilizes ideal gas
calculations and the differences between the real material and an ideal gas.

Define the concept of a residual function:

Ψresidual (T ,X ) = Ψ (T ,X ) − Ψig (T ,X )

Because the independent variable X may be either P or V, there actually are two
definitions of residual functions.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 159

2.6.4.3.1 Density Form (Definition)

Ψr (T , ρ ) ≡ Ψ (T , ρ ) − Ψig (T , ρ )
where:
Ψr (T , ρ ) = density residual function
Ψ (T , ρ ) = value of Ψ for real material at T , ρ
Ψig (T , ρ ) = value of Ψ for an ideal gas at T , ρ

2.6.4.3.2 Pressure Form (Definition)

ΨR (T ,P ) ≡ Ψ (T ,P ) − Ψig (T ,P )
where:
ΨR (T ,P ) = pressure residual function
Ψ (T ,P ) = value of Ψ for real material at T ,P
Ψig (T ,P ) = value of Ψ for an ideal gas at T ,P

Note that the pressure and density forms of the residual functions are not
necessarily equal because they involve comparisons to different ideal gas states.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 160

The density form of the residual function compares the real fluid at T, P, r with
the ideal gas at

T ig = T
ρ ig = ρ
P ⎛ P ⎞
P ig = ρ ig RT ig = ρRT = ⎜⎝ Z =
Z ρRT ⎟⎠

The pressure form of the residual function compares the real fluid at T, P, r with
the ideal gas at

T ig = T
P ig = P
ig P ig P
ρ = = = ρZ
RT RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 161

Summary Table

Temperature Pressure Density

Real Fluid T P r

Ideal Gas State: T ig P ig ρ ig

Density Form T (P/Z) r

Pressure Form T P rZ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 162

The property change between state 1 and state 2 is calculated using
Ψ ( 2 ) − Ψ (1) = Ψ (T2 ,P2 , ρ2 ) − Ψ (T1,P1, ρ1)
= Ψ (T2 ,P2 , ρ2 ) − Ψig (T2 , ρ2 ) + Ψig (T2 , ρ2 )
−Ψig (T1, ρ1) + Ψig (T1, ρ1) − Ψ (T1,P1, ρ1)

= Ψr (T2 , ρ2 ) + Ψig (T2 , ρ2 ) − Ψig (T1, ρ1) − Ψr (T1, ρ1)

or
Ψ ( 2 ) − Ψ (1) = Ψ (T2 ,P2 , ρ2 ) − Ψ (T1,P1, ρ1)

= Ψ (T2 ,P2 , ρ2 ) − Ψig (T2 ,P2 ) + Ψig (T2 ,P2 )

−Ψig (T1,P1) + Ψig (T1,P1) − Ψ (T1,P1, ρ1)

= ΨR (T2 ,P2 ) + Ψig (T2 ,P2 ) − Ψig (T1,P1) − ΨR (T1,P1)

Both of these forms are equivalent.
For practical calculations, the independent variables are chosen for convenience
in calculating the property change of interest.
Note that only ideal gas calculations involve a change of temperature, therefore
only ideal gas heat capacities are required.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 163
2.6.4.3.3 Methods for Calculation

The residual functions are related by the same expressions as the energy
functions:

Density Form

H r = U r + P rV
Ar = U r −TS r
G r = H r −TS r = U r + P rV −TS r

Pressure Form

H R = U R + PV R
AR = U R −TS R
GR = H R −TS R = U R + PV R −TS R

Therefore, only two relationships must be derived for the five U H A G S

residuals.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 164

The values for the real fluid and the ideal gas become equal in the limit as the
pressure and density go to zero, therefore

lim lim
ΨR (T ,P ) = 0 and Ψ r (T , ρ ) = 0
P →0 ρ→0

Then

ρ ρ
⌠⎡
⌠
⎮⎛ ∂Ψr ⎞ ⎮ ⎛ ∂Ψ ⎞ ⎛ ∂Ψig ⎞ ⎤
Ψr (T , ρ ) − Ψr (T ,0 ) = Ψr (T , ρ ) = ⎮⎜ d ρ = ⎮ ⎢⎜ ⎟⎠ − ⎜ ∂ ρ ⎟ ⎥ d ρ
⎮⎝ ∂ ρ ⎟⎠ ⎮⎢
⎮ ⎝ ∂ ρ T ⎝ ⎠ T ⎥⎦
⌡ T ⌡⎣
0 0

and

P P
⌠⎡
⌠
⎮⎛ ∂ΨR ⎞ ⎮ ⎛ ∂Ψ ⎞ ⎛ ∂Ψig ⎞ ⎤
ΨR (T ,P ) − ΨR (T ,0 ) = ΨR (T ,P ) = ⎮⎜ ⎟ dP = ⎮ ⎢⎜ ⎟ −⎜ ⎟ ⎥ dP
⎮⎝ ∂P ⎠ ⎝ ⎠
⎮ ⎢ ∂P T
⎮ ⎝ ∂P ⎠ T ⎥⎦
⌡ T ⌡⎣
0 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 165

2.6.4.3.3.1 Residual Expressions (Density Forms)

2.6.4.3.3.1.1 Pressure

P r (T , ρ ) = P (T , ρ ) − P ig (T , ρ ) = Z ρRT − ρRT = ( Z − 1) ρRT

Dimensionless form:

P rV
= Z −1
RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 166

2.6.4.3.3.1.2 Internal Energy

Calculate the residual internal energy at any density by integrating the partial
derivative along the isotherm from zero density (where the residual is known to
vanish) to the density of interest
ρ ρ ρ
⌠ ⌠
⎮ ⎛ ∂U ⎞ ⎮ ⎛ ∂U
⌠ ig ⎞
r
⎛ ∂ U ⎞
U r (T , ρ ) = ⎮⎜
⎮
⎟ d ρ = ⎮⎮ ⎜⎝ ∂ρ ⎟⎠ d ρ − ⎮⎮ ⎜ ∂ρ ⎟ d ρ
⎮ ⎝ ∂ρ ⎠T
⌡ ⌡ T ⌡⎝ ⎠T
0 0 0

MATH 4 gives:

⎛ ∂U ⎞ ⎛ ∂U ⎞ ⎛ ∂V ⎞ ⎛ ∂U ⎞ ⎛ 1⎞
⎜⎝ ∂ ρ ⎟⎠ = ⎜⎝ ⎟⎠ ⎜⎝ ∂ ρ ⎟⎠ = ⎜⎝ ⎟⎠ ⎜− 2⎟
T
∂V T T
∂V T ⎝ ρ ⎠

Use the physics summary statement for internal energy and MaxA to relate the
partial derivatives to observable properties

⎛ ∂U ⎞ ⎛ ∂S ⎞ ⎛ ∂V ⎞ ⎛ ∂P ⎞
dU = TdS − PdV ⇒ ⎜⎝ ⎟ = T ⎜⎝ ⎟ − P ⎜⎝ ⎟ = T ⎜⎝ ⎟ −P
∂V ⎠ T ∂V ⎠ T ∂V ⎠ T ∂T ⎠ V

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 167

so that the residual internal energy is

ρ
⌠⎧
⎤ ⎫
ig
⎮ ⎪⎡ ⎛ ∂P ⎞ ⎤ ⎡ ⎛ ∂P ⎞ ⎪dρ
U r (T ,V ) = ⎮ ⎨ ⎢T ⎜⎝ ⎟ − P −
⎥ ⎢ ⎜⎝
T ⎟⎠ − P ⎥ ⎬ 2
⎮
⎮ ⎪⎣ ∂T ⎠ V ⎦ ⎣ ∂ T ⎦ ⎪ ρ
⌡⎩ ⎭
V
0

It also is useful to derive a dimensionless form that is written in terms of the

compression factor, Z. Begin with the equation of state
ZRT ⎛ ∂P ⎞ ZR RT ⎛ ∂ Z ⎞ P RT ⎛ ∂ Z ⎞
P= ⇒ ⎜ = + = +
V ⎝ ∂T ⎟⎠V V
⎜ ⎟
V ⎝ ∂T ⎠ V T
⎜ ⎟
V ⎝ ∂T ⎠ V

Then
⎛ ∂P ⎞ RT 2 ⎛ ∂ Z ⎞ RT 2 ⎛ ∂ Z ⎞
T⎜ −P = P + ⎟ −P =
⎝ ∂T ⎟⎠V ⎜
V ⎝ ∂T ⎠ V
⎜
V ⎝ ∂T ⎠ V
⎟

so that

⎛ ∂U ⎞ RT 2 ⎛ ∂ Z ⎞ ⎡ 1 ⎤ RT 2 ⎛ ∂ Z ⎞
⎜⎝ ∂ρ ⎟⎠ = V ⎜⎝ ∂T ⎟⎠ ⎢− 2 ⎥ = −
ρ
⎜⎝
∂
⎟⎠
T V ⎣ ρ ⎦ T ρ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 168

For the ideal gas

ig
⎛ ∂Z ig ⎞
Z =1 ⇒ ⎜ ∂T ⎟ = 0
⎝ ⎠

Also, inverse temperature is a more convenient variable than temperature in

many of equations of state. Note that

⎛ ∂Z ⎞ ⎛ ∂ Z ⎞ ⎛ ∂ T1 ⎞ 1 ⎛ ∂Z ⎞
⎜⎝ ⎟ = =− 2⎜ 1⎟
∂T ⎠ ρ ⎜⎝ ∂ T1 ⎟⎠ ⎜⎝ ∂T ⎟⎠ T ⎝ ∂T ⎠ ρ
ρ ρ

so that the residual internal energy (density form) is

ρ ρ
⌠
⎮ R ⎛ ∂Z ⎞
⌠ 2
⎮ RT ⎛ ∂Z ⎞
U r
(T , ρ ) = ⎮− ⎜⎝ ⎟⎠ d ρ = ⎮ ⎜ ⎟ dρ
⎮
⌡
ρ ∂T ρ ⎮ ρ ⎝∂ ⎠
1
⌡ T ρ
0 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 169

This may be put in dimensionless form by dividing through by RT

ρ ρ ρ

U r ⌠T ∂Z
⎛ ⎞ 1 ⌠⎮ ⎛ ∂ Z ⎞ dρ 1 ⌠⎮ ⎛ ∂ Z ⎞
= − ⎮⎮ ⎜
⎝ ⎟⎠ d ρ = ⎮⎜ 1 ⎟⎠ ρ = ⎮⎜ 1 ⎟⎠ d ln ρ
RT ⌡ ρ ∂T ρ T ⎮⌡ ⎝ ∂ T ρ
T ⎮⌡ ⎝ ∂ T ρ
0 0 0

NOTATION:

For simplicity of notation in subsequent calculation, define a dimensionless

integral as follows

ρ ρ
⌠
⌠ ∂Z
⎮⎛ ⎞ d ρ 1 ⎮⎛ ∂Z ⎞ d ρ
Ι 1 = −T ⎮⎜ ⎟⎠ = ⎮⎜ 1 ⎟
⌡ ⎝ ∂T ρ ρ T ⎮⌡ ⎝ ∂ T ⎠ ρ ρ
0 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 170

2.6.4.3.3.1.3 Helmholtz Energy

One more explicit calculation is required. We choose the Helmholtz energy

because it has the variable of integration (density) as a natural independent
variable.
ρ
⌠
⎮⎛ ∂A ⎞
r
A (T , ρ ) − A (T ,0 ) =
r r
⎮⎜ ⎟ dρ
⎮ ⎝ ∂ρ ⎠T
⌡
0

The definition of the residual function provides

r ig
⎛ ∂ Ar ⎞ ⎛ ∂A⎞ ⎛ ∂ Aig ⎞
A = A−A ⇒ ⎜ ∂ρ ⎟ = ⎜⎝ ∂ρ ⎟⎠ − ⎜ ∂ρ ⎟
⎝ ⎠T T ⎝ ⎠T

Use the physics summary statement for Helmholtz energy to relate the partial
derivatives to other properties

⎛ ∂A⎞
dA = −SdT − PdV ⇒ ⎜⎝ ⎟ = −P
∂V ⎠ T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 171

so that one form for the residual Helmholtz energy is
V
A (T ,V ) = −RT ⌠⎮⌡ ( Z − 1) dV
r

Convert to density as the variable

⎛ ∂A ⎞ ⎛ ∂A ⎞ ⎛ ∂V ⎞ ⎛ 1⎞ P ZRT
⎜⎝ ∂ ρ ⎟⎠ = ⎜⎝ ⎟⎠ ⎜⎝ ∂ ρ ⎟⎠ = ( −P ) ⎜ − ⎟ = =
T
∂V T T ⎝ ρ2 ⎠ ρ2 ρ

ig
For the ideal gas, Z = 1, therefore

⎛ ∂Aig ⎞ RT
⎜ ∂ρ ⎟ =
⎝ ⎠T ρ

Then
ρ ρ ρ
⌠
⎮⎛ ∂A ⎞ ⌠⎛
r
ZRT RT ⎞ ⌠ dρ
A (T , ρ ) =
r
⎮⎜ ⎟ dρ =
⎮
⎮⎜
− d ρ = RT ⎮ ( Z − 1)
⎮ ⎝ ∂ρ
⌡ ⎠T ⌡⎝ ρ ρ ⎟⎠ ⎮
⌡ ρ
0 0 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 172

The dimensionless form is

ρ
A (T , ρ )
r ⌠ dρ
= ⎮ ( Z − 1)
RT ⎮
⌡ ρ
0

NOTATION:

This form leads to the definition of a second dimensionless integral

ρ
⌠⎛
Z − 1⎞
Ι2 = ⎮
⎮⎜
dρ
⌡⎝ ρ ⎟⎠
0

2.6.4.3.3.1.4 Enthalpy

H r U r P rV
r
H =U +P V r r
⇒ = + = Ι1 + ( Z − 1)
RT RT RT
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 173
2.6.4.3.3.1.5 Entropy

r r U r − Ar rS r U r − Ar
r
A = U −TS ⇒ S = ⇒ = = Ι1 − Ι2
T R RT

2.6.4.3.3.1.6 Gibbs

Gr
r r
G = H −TS = A + P V r r r
⇒ = Ι2 + ( Z − 1)
RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 174

2.6.4.3.3.1.7 Summary of Density Forms

Ur Ar
= Ι1 = Ι2
RT RT

Hr Gr
= Ι1 + Z − 1 = Ι2 + Z − 1
RT RT

P rV Sr
= Z −1 = Ι1 − Ι 2
RT R

ρ ρ
⌠
1 ⎮⎡ ∂Z ⎤ d ρ ⌠Z −1
Ι1 = ⎮⎢ Ι2 =⎮ dρ
T ⎮ ⎣∂
⌡
(1/ T ) ⎥⎦ ρ ρ ⎮
⌡ ρ
0 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 175

2.6.4.3.3.2 Residual Functions (Pressure Form)

Because nature forces the use of temperature and density as the independent
variables in the equation of state, the pressure forms of the residual functions are
calculated most easily using the density forms just derived and a calculation of
property changes in the ideal gas.

The residual function definitions are

ΨR (T ,P, ρ ) = Ψ (T ,P, ρ ) − Ψig (T ,P, ρ Z )

Ψr (T ,P, ρ ) = Ψ (T ,P, ρ ) − Ψ ig (T ,P / Z , ρ )

Subtracting the second equation from the first produces

ΨR (T ,P, ρ ) = Ψr (T ,P, ρ ) + Ψig (T ,P / Z , ρ ) − Ψig (T ,P, ρ Z )

Methods to calculate Ψr are shown in the previous section, and the difference
term ⎡ Ψig (T ,P / Z , ρ ) − Ψig (T ,P, ρ Z ) ⎤ is an ideal gas calculation.
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 176

2.6.4.3.3.2.1 Volume

ZRT RT RT
V R =V −V ig = − = ( Z − 1)
P P P

PV R
Dimensionless form: = ZR = Z −1
RT

2.6.4.3.3.2.2 Internal Energy

Because entropy and volume (density) are the natural independent variables for
internal energy, the change in internal energy for the ideal gas is calculated most
easily from
ρ
⌠
⎛ ∂U ig ⎞
U ig
(T , ρ ) − U (T , ρZ ) =
ig ⎮
⎮ ⎜ ∂ρ ⎟ d ρ
⎮
⌡ ⎝ ⎠T
ρZ
where
⎛ ∂U ig ⎞ ⎛ ∂U ig ⎞ ⎛ ∂V ⎞ 1 ⎛ ∂U ig ⎞
⎜ ∂ ρ ⎟ = ⎜ ∂V ⎟ ⎜⎝ ∂ ρ ⎟⎠ = − 2 ⎜ ∂V ⎟
⎝ ⎠T ⎝ ⎠T T ρ ⎝ ⎠T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 177

and
⎛ ∂U ⎞ ⎛ ∂S ⎞ ⎛ ∂P ⎞
dU = TdS − PdV ⇒ ⎜ = − = ⎟ −P
⎝ ∂V ⎟⎠ T ⎜⎝ ⎟ ⎜⎝
T P T
∂V ⎠ T ∂T ⎠V

For the ideal gas

ig RT ⎛ ∂P ig ⎞ R
P = ⇒ ⎜ ⎟ =
V ⎝ ∂T ⎠V V

so that
⎛ ∂P ⎞ RT RT
T⎜ − = − =0
⎝ ∂T ⎟⎠V
P
V V

Therefore
ρ
⌠
⎛ ∂U ig ⎞
U ig
(T , ρ ) − U (T , ρZ ) =
ig ⎮
⎮ ⎜ ∂ρ ⎟ d ρ = 0
⎮
⌡ ⎝ ⎠T
ρZ

and U R (T ,P, ρ ) = U r (T ,P, ρ )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 178

2.6.4.3.3.2.3 Helmholtz Energy

Density also is the most convenient variable for calculations involving the
Helmholtz energy

ρ
⌠
⎛ ∂A ig ⎞
A ig
(T , ρ ) − A (T , ρZ ) =
ig ⎮
⎮ ⎜ ∂ρ ⎟ d ρ
⎮
⌡ ⎝ ⎠T
ρZ

where

⎛ ∂A ig ⎞ ⎛ ∂A ig ⎞ ⎛ ∂V ⎞ 1 ⎛ ∂A ig ⎞
⎜ ∂ ρ ⎟ = ⎜ ∂V ⎟ ⎜⎝ ∂ ρ ⎟⎠ = − 2 ⎜ ∂V ⎟
⎝ ⎠T ⎝ ⎠T T ρ ⎝ ⎠T

and

⎛ ∂A ⎞
dA = −SdT − PdV ⇒ ⎜⎝ ⎟⎠ = −P
∂V T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 179

For the ideal gas

ig RT ⎛ ∂Aig ⎞ RT ⎛ ∂Aig ⎞ RT
P = ⇒⎜ ⎟ = − ⇒ ⎜ ∂ρ ⎟ =
V ⎝ ∂V ⎠ T V ⎝ ⎠T ρ

so that
ρ
⌠
⎛ ∂A ig ⎞
A ig
(T , ρ ) − A (T , ρZ ) =
ig ⎮
⎮ ⎜ ∂ρ ⎟ d ρ
⎮
⌡ ⎝ ⎠T
ρZ
ρ
⌠ RT ρ ρ
= ⎮ d ρ = RT ln ρ ρZ = RT ln = −RT lnZ
⎮
⌡ ρ ρZ
ρZ

Then

AR (T ,P, ρ ) = Ar (T ,P, ρ ) − RT lnZ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 180

Calculate the remaining functions using the definitions.

2.6.4.3.3.2.4 Enthalpy

H R = U R + PV R = U r + P rV ⇒ H R = H r

2.6.4.3.3.2.5 Entropy

SR =
r r
U R − AR U − A − RT lnZ
=
( )
⇒ S R = S r + R lnZ
T T

2.6.4.3.3.2.6 Gibbs Energy

G R = AR + P RV = Ar − RT lnZ + PV r ⇒ G R = G r − RT lnZ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 181

2.6.4.3.3.2.7 Summary of Pressure Forms

PV R P rV
= = Z (T , ρ ) − 1
RT RT
UR Ur
= = Ι1 (T , ρ )
RT RT
HR Hr
= = Ι1 (T , ρ ) + Z (T , ρ ) − 1
RT RT
AR Ar
= − ln Z = Ι2 (T , ρ ) − ln Z (T , ρ )
RT RT
GR Gr
= − ln Z = Ι2 (T , ρ ) + ⎡⎣ Z (T , ρ ) − 1⎤⎦ − ln Z (T , ρ )
RT RT
SR Sr
= + ln Z = Ι1 (T , ρ ) − Ι2 (T , ρ ) + ln Z (T , ρ )
R R

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 182

2.6.4.4 Ideal Gas Changes

When the residual function approach is used to calculate property changes, the
changes in ideal gas properties

Ψig (T2 , ρ2 ) − Ψig (T1, ρ1) or Ψig (T2 ,P2 ) − Ψig (T1,P1)

also are required. To obtain expressions for the ideal gas changes, we use the
general form
T2 X2
⌠ ⌠
⎛ ∂Ψig ⎞ ⎛ ∂Ψig ⎞
Ψ ig
(T2 ,X 2 ) − Ψ (T1,X 1) =
ig ⎮
⎮ ⎜ ∂T ⎟ dT +
⎮
⎮ ⎜ ∂X ⎟ dX
⎮
⌡ ⎝ ⎠X ⎮
⌡ ⎝ ⎠T
T1 X1

where X represents either r or P.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 183

2.6.4.4.1 Internal Energy

For the internal energy, derive the expression using density as the second
independent variable
T2 ρ2
⌠ ⌠
⎛ ∂U ig ⎞ ⎛ ∂U ig ⎞
U ig
(T2 , ρ2 ) − U (T1, ρ1) =
ig ⎮
⎮ ⎜ ∂T ⎟ dT +
⎮
⎮ ⎜ ∂ρ ⎟ d ρ
⎮
⌡
⎝ ⎠ρ ⎮
⌡ ⎝ ⎠T
T1 ρ1

For the internal energy of the ideal gas, the derivatives are

⎛ ∂U ig ⎞ ig
⎜ ∂T ⎟ = CV
⎝ ⎠ρ

⎛ ∂U ig ⎞ 1 ⎛ ∂U ig ⎞ 1 ⎡ ⎛ ∂P ig ⎞ ⎤
⎜ ∂ ρ ⎟ = − 2 ⎜ ∂V ⎟ = − 2 ⎢T ⎜ ∂T ⎟ − P ⎥ = 0
ig
⎝ ⎠T ρ ⎝ ⎠T ρ ⎢⎣ ⎝ ⎠V ⎥⎦

so that the change in internal energy is

T2
U ig
(T2 , ρ2 ) − U (T1, ρ1) = ⌠⎮⌡ CVig dT
ig

T1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 184

This result depends only upon the temperature, therefore the result is the same if
the pressure is the second independent variable
T2
U ig
(T2 ,P2 ) − U (T1,P1) = ⌠⎮⌡ CVig dT
ig

T1

If the result is made dimensionless using the final temperature

T2
U ig
(T2 ) − U (T1) =
ig
1 ⌠
⎮
CVig
dT
RT2 T2 ⎮ R
⌡
T1

This leads to the definition of a third dimensionless integral

T2 T2
1 ⌠ CVig 1 ⌠⎮ ⎛ C Pig ⎞
Ι (T1,T2 ) = ⎮⎮
ig
dT = ⎮ ⎜ − 1⎟ dT
3
T2 ⌡ R T2 ⌡ ⎝ R
⎮ ⎠
T1 T1

Then
U ig ( 2 ) − U ig (1) ig
= Ι 3 (T1,T2 )
RT2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 185

2.6.4.4.2 Enthalpy

The change in the ideal gas enthalpy is calculated using the relation between the
enthalpy and the internal energy

H ig = U ig + P igV = U ig + RT

This result depends only on temperature, so

H ig ( 2 ) − H ig (1) U ig ( 2 ) − U ig (1) R (T2 −T1)

= +
RT2 RT2 RT2

⎛ T1 ⎞
3 ( 1 2 ) ⎜ 1−
= Ιig T ,T +
⎝ T2 ⎟⎠

Alternate expression

T2
H ig
(T2 ) − H (T1) = ⌠⎮⌡ CPig dT
ig

T1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 186

2.6.4.4.3 Entropy

Using the density as the second independent variable

T2 ρ2
⌠ ⌠
⎛ ∂S ig ⎞ ⎛ ∂S ig ⎞
S ig
(T2 , ρ2 ) − S (T1, ρ1) =
ig ⎮
⎮ ⎜ ∂T ⎟ dT +
⎮
⎮ ⎜ ∂ρ ⎟ d ρ
⎮
⌡
⎝ ⎠ρ ⎮
⌡ ⎝ ⎠T
T1 ρ1

For the entropy of the ideal gas, the derivatives are

⎛ ∂S ig ⎞ CVig
⎜ ∂T ⎟ = T
⎝ ⎠ρ

⎛ ∂S ig ⎞ 1 ⎛ ∂S ig ⎞ 1 ⎛ ∂P ig ⎞ 1 R
⎜ ∂ρ ⎟ = − ⎜ ⎟ = − ⎜ ⎟ = − R ρ = −
⎝ ⎠T ρ 2 ⎝ ∂V ⎠ T ρ 2 ⎝ ∂T ⎠V ρ2 ρ

so that the change is

T2 ρ2 T2
⌠ C ig ⌠ R ⌠ CVig ρ
S ig
(T2 , ρ2 ) − S (T1, ρ1) = ⎮
ig ⎮ V
T
dT − ⎮
⎮ ρ
dρ = ⎮
⎮ T
dT − R ln 2
ρ1
⌡ ⌡ ⌡
T1 ρ1 T1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 187

If the pressure is the second independent variable
T2 P2
⌠ ⌠
⎛ ∂S ig ⎞ ⎛ ∂S ig ⎞
S ig
(T2 ,P2 ) − S (T1,P1) =
ig ⎮
⎮ ⎜ ∂T ⎟ dT +
⎮
⎮ ⎜ ∂P ⎟ dP
⎮
⌡ ⎝ ⎠P ⎮
⌡ ⎝ ⎠T
T1 P1

The derivatives are

⎛ ∂S ig ⎞ CPig
⎜ ∂T ⎟ = T
⎝ ⎠P

⎛ ∂S ig ⎞ ⎛ ∂V ig ⎞ R
⎜ ∂P ⎟ = ⎜ ⎟ = −
⎝ ⎠ T ⎝ ∂T ⎠ P P

so that the change is

T2 P2 T2
⌠ C ig ⌠ R ⌠ CPig P
S ig
(T2 ,P2 ) − S (T1,P1) = ⎮
ig ⎮ P
T
dT − ⎮
⎮
⌡ P
dP = ⎮
⎮ T
dT − R ln 2
P1
⌡ ⌡
T1 P1 T1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 188

The dimensionless form is

T2
S ig
(T2 ,P2 ) − S (T1,P1) = ⌠⎮ CPig dT
ig
− ln
P2
R ⎮ R T P1
⌡
T1

Define a fourth dimensionless integral

T2 T2
⌠C ig
dT ⌠⎮ ⎛ CVig ⎞ dT
Ι ig
(T1,T2 ) = ⎮
⎮
P
= ⎮⎜
⎮⎝ R
+ 1⎟
⎠ T
4
⌡ R T ⌡
T1 T1

Then

S ig ( 2 ) − S ig (1) P2 ρ2T2
R
= Ιig
4 (T ,T
2 1 ) − ln
P1
= Ι ig
4 (T ,T
2 1 ) − ln
ρ1T1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 189

2.6.4.4.4 Helmholtz Energy

Use the relation of the Helmholtz energy to the internal energy and entropy

Aig ( 2 ) − Aig (1) = U ig ( 2 ) −T2S ig ( 2 ) − ⎡⎣U ig (1) −T1S ig (1) ⎤⎦

Then

Aig ( 2 ) − Aig (1) U ig ( 2 ) − U ig (1) S ig ( 2 ) T1S ig (1)

= − +
RT2 RT2 R T2R

U ig ( 2 ) − U ig (1) ⎡ S ig ( 2 ) S ig (1) ⎤ ⎡ T1 ⎤ S ig (1)

= −⎢ − ⎥ − ⎢1− ⎥
RT2 ⎢⎣ R R ⎥⎦ ⎣ T2 ⎦ R

P2 ⎡ T1 ⎤ S ig (1)
= Ιig
3 ( )
T1,T2 − Ιig
4 (T1,T2 ) + ln P − ⎢1− T ⎥ R
1 ⎣ 2⎦

T2 ρ2 ⎡ T1 ⎤ S ig (1)
= Ιig
3 ( )
T1,T2 − Ιig
4 (T1,T2 ) + ln T ρ − ⎢1− T ⎥ R
1 1 ⎣ 2⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 190

2.6.4.4.5 Gibbs Energy

Use the relation of the Gibbs energy to the enthalpy and entropy

G ig ( 2 ) − G ig (1) = ⎡⎣H ig ( 2 ) −T2S ig ( 2 ) ⎤⎦ − ⎡⎣H ig (1) −T1S ig (1) ⎤⎦

Then
G ig ( 2 ) − G ig (1) H ig ( 2 ) − H ig (1) S ig ( 2 ) T1S ig (1)
= − +
RT2 RT2 R T2R

H ig ( 2 ) − H ig (1) ⎡ S ig ( 2 ) S ig (1) ⎤ ⎡ T1 ⎤ S ig (1)

= −⎢ − ⎥ − ⎢1− ⎥
RT2 ⎢⎣ R R ⎥⎦ ⎣ T2 ⎦ R

⎡ T1 ⎤ ig P2 ⎡ T1 ⎤ S ig (1)
= Ιig
3 (T1,T2 ) + ⎢1− T ⎥ − Ι 4 (T1,T2 ) + ln P − ⎢1− T ⎥ R
⎣ 2⎦ 1 ⎣ 2⎦

P2 ⎡ T1 ⎤ ⎡ S ig (1) ⎤
= Ιig
3 ( )
T1,T2 − Ιig
4 (T1,T2 ) + ln P − ⎢1− T ⎥ ⎢1+ R ⎥
1 ⎣ 2 ⎦ ⎢⎣ ⎥⎦

T2 ρ2 ⎡ T1 ⎤ ⎡ S ig (1) ⎤
= Ιig
3 ( )
T1,T2 − Ιig
4 (T1,T2 ) + ln T ρ − ⎢1− T ⎥ ⎢1+ R ⎥
1 1 ⎣ 2 ⎦ ⎢⎣ ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 191

2.6.4.5 General Formulae for Property Changes

Combine the previous results to obtain formulae for the general cases.

2.6.4.5.1 Internal Energy

Begin with the basic difference equation

U ( 2 ) − U (1) = U r ( 2 ) − U r (1) + U ig (T2 , ρ2 ) − U ig (T1, ρ1)

= Ι1 (T2 , ρ2 ) ⋅RT2 − Ι1 (T1, ρ1) ⋅RT1

3 (T1,T2 ) ⋅RT2
+Ιig

Dimensionless form

U ( 2 ) − U (1) T
= Ι 1 ( 2 ) − 1 Ι 1 (1) + Ι 3ig (T1,T2 )
RT2 T2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 192

2.6.4.5.2 Enthalpy

Begin with the basic difference equation

H ( 2 ) − H (1) = U ( 2 ) + RT2 ⋅ Z (T2 , ρ2 ) − ⎡⎣U (1) + RT1 ⋅ Z (T1, ρ1) ⎤⎦

Dimensionless form

H ( 2 ) − H (1) U ( 2 ) − U (1) T
= + Z (T2 , ρ2 ) − 1 Z (T1, ρ1)
RT2 RT2 T2
T1
= Ι1 (T2 , ρ2 ) − Ι1 (T1, ρ1)
T2
T1
3 (T1,T2 ) + Z (T2 , ρ2 ) −
+Ιig Z (T1, ρ1)
T2

Algebraic rearrangement and simplified notation gives

H ( 2 ) − H (1) T
= Ι 1 ( 2 ) + Z ( 2 ) − 1 ⎡⎣Ι 1 (1) + Z (1) ⎤⎦ + Ι 3ig (T1,T2 )
RT2 T2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 193

2.6.4.5.3 Entropy

Begin with the basic difference equation

S ( 2 ) − S (1) = S R ( 2 ) − S R (1) + S ig (T2 ,P2 ) − S ig (T1,P1)

Dimensionless form

S ( 2 ) − S (1) S R ( 2 ) S R (1) S ig (T2 ,P2 ) − S ig (T1,P1)

= − +
R R R R

= Ι1 (T2 , ρ2 ) − Ι2 (T2 , ρ2 ) + lnZ (T2 , ρ2 )

− ⎡⎣Ι1 (T1, ρ1) − Ι2 (T1, ρ1) + lnZ (T1, ρ1) ⎤⎦ + Ιig
4 (T1,T2 )

Rearrange to obtain

S ( 2 ) − S (1) ⎛ P2 Z (1) ⎞
= Ι1 ( 2 ) − Ι1 (1) − ⎡⎣Ι2 ( 2 ) − Ι2 (1) ⎤⎦ + Ι 4 (T1,T2 ) − ln ⎜
ig
R ⎝ P1Z ( 2 ) ⎟⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 194

2.6.4.5.4 Helmholtz Energy

Begin with the basic difference equation

A ( 2 ) − A (1) = U ( 2 ) −T2S ( 2 ) − ⎡⎣U (1) −T1S (1) ⎤⎦

Dimensionless form
A ( 2 ) − A (1) U ( 2 ) − U (1) T2S ( 2 ) −T1S (1)
= −
RT2 RT2 RT2

U ( 2 ) − U (1) T2S ( 2 ) −T2S (1) +T2S (1) −T1S (1)

= −
RT2 RT2

U ( 2 ) − U (1) S ( 2 ) − S (1) ⎛ T ⎞ S (1)

= − − ⎜ 1− 1 ⎟
RT2 R ⎝ T2 ⎠ R

⎧ T1 ⎫ ⎧
= ⎨Ι1 ( 2 ) − Ι1 (1) + Ι 3 (T1,T2 ) ⎬ − ⎨Ι1 ( 2 ) − Ι1 (1)
ig
⎩ T2 ⎭ ⎩
⎛ P2 Z (1) ⎞ ⎫
⎪ ⎧⎛ T1 ⎞ S (1) ⎫
− ⎡⎣Ι2 ( 2 ) − Ι2 (1) ⎤⎦ + Ιig (T1,T2 ) − ln ⎜ P Z 2 ⎟ ⎬ − ⎨⎜ 1− T ⎟ R ⎬
⎝ 1 ( )⎠ ⎪
4
⎭ ⎩⎝ 2⎠ ⎭

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 195

Rearrange to obtain

A ( 2 ) − A (1) ⎛ T ⎞⎡ S (1) ⎤
= ⎡⎣Ι2 ( 2 ) − Ι2 (1) ⎤⎦ + ⎜ 1− 1 ⎟ ⎢Ι1 (1) −
RT2 ⎝ T2 ⎠ ⎣ R ⎥⎦

⎛ P2 Z (1) ⎞
3 ( 1 2)
+Ιig 4 ( 1 2)
ig
T ,T − Ι T ,T + ln ⎜P Z 2 ⎟
⎝ 1 ( )⎠

2.6.4.5.5 Gibbs Energy

Begin with the basic difference equation

G ( 2 ) − G (1) = H ( 2 ) −T2S ( 2 ) − ⎡⎣H (1) −T1S (1) ⎤⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 196

Dimensionless form
G ( 2 ) − G (1) H ( 2 ) − H (1) T2S ( 2 ) −T1S (1)
= −
RT2 RT2 RT2
H ( 2 ) − H (1) T2S ( 2 ) −T2S (1) +T2S (1) −T1S (1)
= −
RT2 RT2
H ( 2 ) − H (1) S ( 2 ) − S (1) ⎛ T ⎞ S (1)
= − + ⎜ 1− 1 ⎟
RT2 R ⎝ T2 ⎠ R
⎧ T ⎫ ⎧⎛ T1 ⎞ S (1) ⎫
3 ( 1 2 ) ⎬ − ⎨⎜ 1−
= ⎨Ι1 ( 2 ) + Z ( 2 ) − 1 ⎡⎣Ι1 (1) + Z (1) ⎤⎦ + Ιig T ,T ⎟⎠ R ⎬
⎩ T2 ⎭ ⎩ ⎝ T2 ⎭
⎧⎪ ⎛ P2 Z (1) ⎞ ⎫⎪
− ⎨Ι1 ( 2 ) − Ι1 (1) − ⎡⎣Ι2 ( 2 ) − Ι2 (1) ⎤⎦ + Ι 4 (T1,T2 ) − ln ⎜
ig
⎟⎠ ⎬
⎪⎩ ⎝ P1Z ( 2 ) ⎪⎭
Rearrange to obtain

G ( 2 ) − G (1) ⎛ T1 ⎞ ⎡ S (1) ⎤
= ⎜ 1− ⎟ ⎢Ι1 (1) − ⎥ + ⎡⎣Ι2 ( 2 ) − Ι2 (1) ⎤⎦
RT2 ⎝ T2 ⎠ ⎣ R ⎦

⎛ P2 Z (1) ⎞ T1
+Ιig
3 (T ,T
1 2 ) − Ι ig
4 (T ,T
1 2 ) + ln ⎜⎝ P Z ( 2 ) ⎟⎠ + Z ( 2 ) −
T2
Z (1)
1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 197

2.6.4.6 Properties relative to a reference state

Because the internal energy and entropy are defined by differential equations, an
arbitrary choice of one boundary condition is required to prepare a table.
Generally take

S† = 0 H† = 0 G† = 0 at (T †
,P † , ρ † )
The functions cannot all be chosen to be zero at the reference point because

H † = U † + (PV ) = U † + Z †RT † ⇒ U † = −Z †RT †

†

G † = A† + (PV ) = A† + Z †RT † ⇒ A† = −Z †RT †

†

where

(
Z † = Z T †, ρ † )
The table entries for the various properties are calculated according to

(
Ψ = Ψ − Ψ† + Ψ† )
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 198
Begin with internal energy

U = U −U ( †
) +U †
⇒
U
RT
=
U −U† U†
RT
+
RT

Using the previous results

U
RT
= Ι1 (T , ρ ) −
T † † ig †
Ι + Ι T ,T −
T 1 3
T† †
T
Z ( )
H U
= + Z (T , ρ )
RT RT

⎡P ⋅ Z † ⎤
S S − S† S†
R
=
R
+
R
† † ig †
= Ι1 − Ι1 − Ι2 + Ι2 + Ι 4 T ,T − ln ⎢ † ⎥
⎢⎣ P ⋅ Z ⎥⎦
( )
A U S
= −
RT RT R
G U S
= − +Z
RT RT R

where: † † †
(†
)
Ι1 = Ι1 (T , ρ ) ; Z = Z (T , ρ ); Ι1 = Ι1 T , ρ ; Z = Z T , ρ
† †
( )
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 199
2.6.4.6.1 Ideal Gas Reference State

( )
If the reference state is an ideal gas at T † ,P † , then Ι1† = 0; Ι†2 = 0; Z † = 1, and

U
RT
ig †
= Ι1 + Ι 3 T ,T −
T†
T
( )
H
RT
ig †
= Ι1 + Z + Ι 3 T ,T −
T†
T
( )
S
R
= Ι1 − Ι2 + Ιig
4 T †
,T − lnZ(− ln
P
P†
)
A
RT
ig † ig †
(
= Ι2 + Ι 3 T ,T + Ι 4 T ,T −
TT
T
) P
− lnZ − ln †
P
( )
G
RT
ig † ig
(
†
= Ι2 + Ι 3 T ,T − Ι 4 T ,T + Z −
T†
T
) ( P
)
− lnZ − ln †
P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 200

2.6.4.6.2 Change of reference state
All tables are based upon reference states, but there are many choices for the
reference state. Some commonly used reference states are
Tref Pref

0.01 °C PHvapO
2

60 °F 14.696 psia
60 °F 14.73 psia
15 °C 1 bar
0 °C 1 atm
0 °F 1 atm
25 °C 1 atm
vap
-40 °C P
0K 1 bar

Because so many choices are available, a method is required for converting

values based upon one reference state to another reference state.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 201
Let:
Y = value of a property at (T,P) based upon a reference state at (T†,P†)

Y† = value of the property at the reference state (T†,P†)

Y∆ = value of the property at (T,P) based upon a reference state at (T*,P*)

Y* = value of the property at the new reference state (T*,P*)

In each case

(
Ψ (T ,P ) = Ψ † + Ψ − Ψ † )
ΨΔ (T ,P ) = Ψ * + ( Ψ Δ
−Ψ* )
The differences (Y-Y†) and (Y∆-Y*) are calculated using the procedures
developed previously. Subtracting the first equation from the second yields

Ψ Δ (T ,P ) = Ψ (T ,P ) + Ψ * −Ψ † + Ψ Δ − Ψ * − Ψ − Ψ † ( ) ( )
Table Property Property
Value Change Change

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 202

Internal Energy

Begin with the basic relation for each case

U (T ,P ) U †
= +
U −U† ( )
RT RT RT

(T ,P ) = U * + (U )
Δ
U Δ −U *
RT RT RT

Subtracting the first expression from the second provides

U Δ − U U * −U † ⎡U Δ − U * ⎤ ⎡U − U † ⎤
= +⎢ ⎥−⎢ ⎥
RT RT ⎣⎢ RT ⎥⎦ ⎢⎣ RT ⎥⎦

where
U −U†
RT
= Ι1 (T , ρ ) −
T†
T
Ι1 T † , ρ † + Ιig
3 T †
,T ( ) ( )
U Δ −U * T*
= Ι1 (T , ρ ) − Ι1 (T *, ρ * ) + Ιig
3 (T *,T )
RT T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 203

Subtraction produces

⎡U Δ − U * ⎤ ⎡U − U † ⎤ T*
⎢ ⎥−⎢ ⎥ = Ι1 (T , ρ ) − Ι1 (T *, ρ * )
⎢⎣ RT ⎥⎦ ⎢⎣ RT ⎥⎦ T
⎡ ⎤
+Ιig
3
T†
( ) ( )
(T *,T ) − ⎢Ι1(T , ρ ) − Ι1 T †, ρ † + Ιig3 T †,T ⎥
⎢⎣ T ⎥⎦

=
T†
T
(
Ι1 T † , ρ † −
T*
T
)
Ι1 (T *, ρ * ) + Ιig
3 (T *,T ) − Ι ig
3 T †
,T ( )
=
T†
T
(
Ι1 T † , ρ † −
T*
T
)
Ι1 (T *, ρ * ) + Ιig
3 T *,T †
( )
For the most common choice of standard state values
H† = 0 H* = 0
S† = 0 S* = 0
G† = 0 G* = 0
U† = -Z†RT† U* = -Z*RT*
A† = -Z†RT† U* = -Z*RT*

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 204

Then

U * −U † ( −Z *RT *) − ( −Z†RT† ) T† † T *
= = Z − Z*
RT RT T T

so that

U Δ − U U * −U † ⎡U Δ − U * ⎤ ⎡U − U † ⎤
= +⎢ ⎥−⎢ ⎥
RT RT ⎢⎣ RT ⎥⎦ ⎢⎣ RT ⎥⎦

=
T† † T *
T
Z −
T
Z *+
T † † T * * ig
T
Ι1 −
T
Ι1 + Ι 3 T *,T † ( )

Note that this also can be written as

U Δ − U U * −U † U
= +
R
(T †
,P † )−U R
(T *,P * ) + U (T *,P * ) − U (T ,P )
ig ig † †

RT RT RT RT RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 205

Entropy

The table entries are calculated using

S (T ,P ) S † ⎡ S − S † ⎤
= +⎢ ⎥
R R ⎢⎣ R ⎥⎦

=
S † S R (T ,P ) S
+ −
R
(T †
,P † )+S ig
(T ,P ) − S ig (T †,P † )
R R R R

S Δ (T ,P ) S * ⎡ S Δ − S * ⎤
= +⎢ ⎥
R R ⎢⎣ R ⎥⎦

=
S
+
*
S (T ,P )R
−
(
S R T * ,P * )+S ig
(T ,P ) − S ig (T * ,P * )
R R R R

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 206

Subtracting the first from the second produces

S Δ − S S* − S† S
= +
R
(T †
,P † ) − S (T ,P ) + S (T
R * * ig †
) (
,P † − S ig T * ,P * )
R R R R R

where

S* − S†
=0
R

(
S R T † ,P † ) = Ι (T †
) (
, ρ † − Ι2 T † , ρ † + lnZ † )
1
R

(
S R T * ,P * ) = Ι (T , ρ ) − Ι (T , ρ ) + lnZ
* * * * *
1 2
R

( )
S ig T † ,P † − S ig T * ,P * ( ) = Ι (T ,T ) − ln⎛ P
ig * †
†⎞
4 ⎜ * ⎟
R ⎝P ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 207

Enthalpy

Use the relation between enthalpy and internal energy

H (T ,P ) U (T ,P )
= + Z (T ,P )
RT RT

H Δ (T ,P ) U Δ (T ,P )
= + Z (T ,P )
RT RT

Then

H Δ − H U Δ −U
=
RT RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 208

Helmholtz Energy

Begin with the relation between the Helmholtz energy, the internal energy and
the entropy:

A (T ,P ) U (T ,P ) S
= −
RT RT R

A Δ (T ,P ) U Δ (T ,P ) S Δ
= −
RT RT R

Then

A Δ − A ⎡U Δ − U ⎤ ⎡ S Δ − S ⎤
=⎢ ⎥−⎢ ⎥
RT ⎢⎣ RT ⎥⎦ ⎢⎣ R ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 209

Gibbs Energy

Begin with the relation between the Helmholtz energy, the internal energy and
the entropy:

G (T ,P ) U (T ,P ) S
= +Z −
RT RT R

G Δ (T ,P ) U Δ (T ,P ) SΔ
= +Z −
RT RT R

Then

G Δ − G ⎡U Δ − U ⎤ ⎡ S Δ − S ⎤
=⎢ ⎥−⎢ ⎥
RT ⎢⎣ RT ⎥⎦ ⎢⎣ R ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 210

2.6.4.6.3 Alternate Form for Change of Reference State

The adjustment from one reference state to another also can be calculated as
follows. The values in each table can be expressed as
(T ,P,ρ )
Ψ = Ψ† + ⌠
⎮
⌡ dΨ
†
and
(T ,P,ρ )
Δ
Ψ = Ψ*+ ⌠
⎮
⌡ dΨ
*

where: Y = value relative to reference state †

Y∆ = value relative to reference state *
Y† = value assigned to the property at reference state †
Y* = value assigned to the property at reference state *

Subtracting the first equation from the second produces

( )+
†
Δ * †
Ψ =Ψ+ Ψ −Ψ ⌠dΨ
⎮
⌡
*

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 211

Geometric Interpretation

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 212

2.6.5 Equilibrium Criteria

2.6.5.1 Entropy and Energy Functions

Equilibrium Criterion for Entropy

Consider an isolated system. This system will be at constant volume and
constant internal energy. All processes that may occur have

(dS ) total
U total ,V total
≥0

Over a long period of time, any process that may occur eventually will occur.
The equilibrium state (the state from which the isolated system will not depart no
matter how long we wait) therefore is characterized by a state for which the
change of entropy in moving to any adjacent state is

(dS ) total
U total ,V total
≤0

Any such change would violate the second law of thermodynamics, therefore the
system remains in the equilibrium state.
At fixed internal energy and volume, entropy is at a maximum for the
equilibrium state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 213

Equilibrium Criterion for Internal Energy

The equilibrium state is the one state from which the system does not move for
specified (fixed) values of Stotal and Vtotal.

First Law:
total
dUsystem / + dW
= dQ /

Because the system is at constant volume, / =0

dW

Therefore,
total
dUsystem /
= dQ

The Second Law allows / ≤ TdSsystem

dQ total

so that, for any event allowed by the Second Law, (dU total
system )
S total ,V total
≤0

Then, if a state has (dU total

system )
S total ,V total
≥ 0 for any change to an adjacent state,
the system cannot leave that state.

At fixed entropy and volume, the internal energy is at a minimum for the
equilibrium state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 214

Equilibrium Criterion for Helmholtz Energy

Here we consider (dA total

system ) total
T ,Vsystem

Begin with the definition of the Helmholtz energy

total total total
Asystem = Usystem −TSsystem
Then
total
dAsystem total
= dUsystem total
−TdSsystem total
− Ssystem / + dW
dT = dQ / −TdSsystem
total

/ ≤ TdSsystem . Therefore events that may occur

The Second Law requires that dQ
total

/ = 0 ) must have
at constant volume ( dW
total
dAsystem / −TdSsystem
= dQ total
≤0

Then, if a state has (dA total

system ) total
T ,Vsystem
≥ 0 for any change to an adjacent state,

the isolated system cannot leave that state.

At fixed temperature and volume, the Helmholtz energy is at a minimum for

the equilibrium state.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 215
Equilibrium Criterion for Enthalpy

Here we consider (dH total

system )total
Ssystem ,P

Begin with the definition of the enthalpy

total total total

Hsystem = Usystem + PVsystem

Then, the First Law leads to

total total total total
dHsystem = dUsystem + PdVsystem +Vsystem dP

/ + dW
= dQ / + PdVsystem
total

Volume changes may be required to maintain the system at constant pressure,

which result in energy being transferred as work in the amount of

/ = −PdVsystem
dW total

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 216

 / ≤ TdSsystem .
The Second Law requires that dQ
total

Therefore events that may occur at constant pressure and entropy must have

total
dHsystem / ≤0
= dQ

Then, if a state has

(dH total
system ) total
Ssystem ,P
≥0

for any change to an adjacent state, the isolated system cannot leave that state.

At fixed entropy and pressure, the enthalpy is at a minimum for the

equilibrium state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 217

Equilibrium Criterion for Gibbs Energy

Here we consider (dG total

system )
T ,P

Begin with the definition of the Gibbs energy

total total total total
Gsystem = Usystem + PVsystem −TSsystem

Then, the First Law leads to

total total total total total total
dGsystem = dUsystem + PdVsystem +Vsystem dP −TdSsystem − Ssystem dT

/ + dW
= dQ / + PdVsystem
total total
−TdSsystem

Volume changes may be required to maintain the system at constant pressure,

which result in energy being transferred as work in the amount of

/ = −PdVsystem
dW total

/ ≤ TdSsystem .
The Second Law requires that dQ
total

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 218

Therefore events that are allowed to occur at constant pressure and temperature
must have

(dG total
system )T ,P
≤0

Then, if a state has

(dG total
system )T ,P
≥0

for any change to an adjacent state, system cannot leave that state.

At fixed temperature and pressure, the Gibbs energy is at a minimum for

the equilibrium state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 219

Summary of Equilibrium Criteria for Entropy and the Energy Functions

(dS ) total
U total ,V total
≤0

(dU total
system ) S total ,V total
≥0

(dH total
system ) total
Ssystem ,P
≥0

(dA total
system ) total
T ,Vsystem
≥0

(dG total
system )T ,P
≥0

At equilibrium, the entropy is maximized and the energy functions are

minimized.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 220

Types of Equilibrium
The variation of entropy near equilibrium points may have several general
characteristics, as shown below.

Metastable
Stable Equilibrium Equilibrium

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 221

 Neutral Equilibrium Unstable Equilibrium

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 222

2.6.5.2 First Derivatives (T, P, µi)

Consider a closed system with no external fields present, so that only PdV work
is involved in the statement of the First Law. As shown below, the system
contains two chambers, A and B, separated by a frictionless, thermally
conductive, semipermeable piston. The system is isolated from its surroundings.

A B

For this system:

S total = S Atotal + SBtotal

U total = U Atotal + UBtotal = constant ⇒ dU Atotal = −dUBtotal

V total =VAtotal +VBtotal = constant ⇒ dVAtotal = −dVBtotal

nitotal = niA + niB = constant ⇒ dniA = −dniB

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 223

The Physics Summary Statement written for each chamber gives

total
dU A = TA dS Atotal − PA dVAtotal + ∑ µiAdn Ai
i

dUBtotal = TB dSBtotal − PB dVBtotal + ∑ µiB dnBi

i

Solve each equation for the entropy:

total 1 total PA total µiA A

dS A = dU A + dVA − ∑ dn i
TA TA i T A

1 PB µiB B
dSBtotal = total
dUB + total
dVB −∑ dn i
TB TB T
i B

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 224

Then

dS total = dSAtotal + dSBtotal

1 total PA total µiA A

= dU A + dVA − ∑ dni
TA TA T
i A

1 total PB total µiB B

+ dUB + dVB −∑ dni
TB TB T
i B

⎡ 1 1 ⎤ total ⎡ PA PB ⎤ total ⎡ µiA µiB ⎤ A

=⎢ − ⎥ dU A + ⎢T − T ⎥ dVA − ∑⎢ − ⎥ dni
T
⎣ A TB ⎦ ⎣ A B ⎦ i ⎢⎣ A
T TB ⎥
⎦

At the equilibrium state

(dS total )U total ,V total = 0

{ }
Also, U Atotal . VAtotal and niA are independent variables, therefore the equilibrium
condition is satisfied in general only if all terms in square brackets are zero.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 225

Then:

1
−
1
=0 ⇒ TA = TB
TA TB

1 1
− =0 ⇒ PA = PB
PA PB

1 1
− =0 ⇒ µiA = µiB
µ Ai µ Bi

The Second Law of Thermodynamics requires that, at equilibrium, the

temperature, pressure and chemical potential of each chemical species be
constant throughout the system.

Note that no external fields are present and there is no mechanism for
doing any kind of work other than PdV work.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 226

2.6.5.3 Second Derivatives (CX, kT)

At equilibrium, the internal energy is at a minimum at fixed entropy and volume,

therefore

(dU ) total
S total
,V total
≥ 0 and (d U )
2 total
S total ,V total
≥0

The Physics Summary Statement for total internal energy has total entropy and
total volume as the natural independent variables.

⎛ ∂U ⎞ ⎛ ∂U ⎞
dU = ⎜ +
⎝ ∂S ⎟⎠V ⎜⎝ ⎟ dV
dS
∂V ⎠ S

⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞
d U = ⎜ 2 ⎟ (dS ) + 2 ⎜ ( ) ( ) ( )
2 2 2
⎟ dS dV + ⎜ 2⎟ dV
⎝ ∂S ⎠V ⎝ ∂S ∂V ⎠ ⎝ ∂V ⎠ S

Consider the temperature as a function of entropy and volume

⎛ ∂T ⎞ ⎛ ∂T ⎞
T = T (S,V ) ⇒ dT = ⎜ +
⎝ ∂S ⎟⎠V ⎜⎝ ⎟ dV
dS
∂V ⎠ S

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 227

But, the Physics Summary Statement also gives the thermodynamic definition

⎛ ∂U ⎞
T =⎜
⎝ ∂S ⎟⎠V

so that

⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞
dT = ⎜ 2 ⎟ dS + ⎜ ⎟ dV
⎝ ∂S ⎠V ⎝ ∂V ∂S ⎠

Squaring this result produces

2 2
⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞
(dT )2 =⎜ 2⎟ (dS ) + 2 ⎜ 2 ⎟ ⎜
2
⎟ (dS ) (dV ) + ⎜ ∂V ∂S ⎟ (dV )2
⎝ ∂S ⎠ ⎝ ∂S ⎠ ⎝ ∂V ∂S ⎠ ⎝ ⎠

Rearrange to obtain
−2 −1 −2 2
⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞
(dS )2 =⎜ 2⎟ (dT ) − 2 ⎜ 2 ⎟
2
⎜ ∂V ∂S ⎟ (dS ) (dV ) − ⎜ 2⎟ ⎜ ∂V ∂S ⎟ (dV )2
⎝ ∂S ⎠ ⎝ ∂S ⎠ ⎝ ⎠ ⎝ ∂S ⎠ ⎝ ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 228

Substitute into the expression for d2U

⎛ 2 ⎞ ⎧⎛ 2 ⎞ −2 ⎛ 2 ⎞ −2 ⎛ 2 ⎞
2
∂ U ⎪ ∂ U ∂ U ∂ U
d 2U = ⎜ 2 ⎟ ⎨⎜ 2 ⎟ (dT ) − ⎜ 2 ⎟ ⎜ ( )
2 2
⎟ dV
⎝ ∂S ⎠ ⎪⎝ ∂S ⎠ ⎝ ∂S ⎠ ⎝ ∂V ∂S ⎠
⎩
−1 ⎫
⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎪ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞
−2 ⎜ 2 ⎟ ⎜ ⎟ ( dS ) ( dV ) ⎬ + 2 ⎜ ∂S ∂V ⎟ ( dS ) ( dV ) + ⎜ 2⎟ ( dV )2
⎝ ∂S ⎠ ⎝ ∂V ∂S ⎠ ⎪⎭ ⎝ ⎠ ⎝ ∂V ⎠

−1 ⎡⎛ 2 ⎞ ⎛ 2 ⎞ −1 ⎛ 2 ⎞ 2 ⎤
⎛ ∂ 2U ⎞ ∂ U ∂ U ∂ U
=⎜ 2⎟ (dT )2 + ⎢⎜ 2 ⎟ − ⎜ 2 ⎟ ⎜ ⎥ (dV )2
⎝ ∂S ⎠ ⎢⎝ ∂V ⎠ ⎝ ∂S ⎠ ⎝ ∂V ∂S ⎟⎠ ⎥
⎣ ⎦

The squared differentials dT and dV always are positive, therefore the

coefficients also must be positive to satisfy the equilibrium condition. Therefore

⎛ ∂ 2U ⎞
−1 ⎡⎛ 2 ⎞ ⎛ 2 ⎞ −1 ⎛ 2 ⎞ 2 ⎤
∂ U ∂ U ∂ U
⎜ 2⎟
> 0 and ⎢⎜ ⎟ − ⎜ ⎟ ⎜ ⎟
⎥>0
⎝ ∂S ⎠ ⎢⎝ ∂V ⎠ ⎝ ∂S ⎠ ⎝ ∂V ∂S ⎠ ⎥
2 2
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 229

The derivatives involved are related to observables

−1 −1
⎛ ∂ 2U ⎞ ⎛ ∂T ⎞ ⎛ ∂S ⎞ CV
=⎜ = =
⎜ 2⎟
⎝ ∂S ⎠ ⎝ ∂S ⎟⎠V ⎜⎝ ⎟
∂T ⎠V T

⎛ ∂ 2U ⎞ ⎛ ∂P ⎞
⎜ 2⎟
= − ⎜⎝ ⎟
⎝ ∂V ⎠ ∂V ⎠ S

⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂P ⎞
=
⎜ ∂S ∂V ⎟ ⎜ ∂V ∂S ⎟ = − ⎜⎝ ⎟
⎝ ⎠ ⎝ ⎠ ∂S ⎠V

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 230

Then
−1 2
⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂ 2U ⎞ ⎛ ∂P ⎞ ⎛ ∂S ⎞ ⎛ ∂P ⎞ ⎛ ∂P ⎞
⎜ −
2⎟ ⎜ 2⎟ ⎜ ∂V ∂S ⎟ = − ⎜⎝ ⎟ − ⎜ ⎟ ⎜ ⎟ ⎜ ⎟
⎝ ∂V ⎠ ⎝ ∂S ⎠ ⎝ ⎠ ∂V ⎠ s ⎝ ∂T ⎠V ⎝ ∂S ⎠V ⎝ ∂S ⎠V

By Math 4

⎛ ∂P ⎞ ⎛ ∂P ⎞ ⎛ ∂P ⎞
= −⎜ −
⎝ ∂V ⎟⎠ s ⎜⎝ ∂T ⎟⎠V ⎜⎝ ∂S ⎟⎠V

By MaxA

⎛ ∂P ⎞ ⎛ ∂S ⎞ ⎛ ∂P ⎞
= −⎜ −
⎝ ∂V ⎟⎠ s ⎜⎝ ∂V ⎟⎠ T ⎜⎝ ∂S ⎟⎠V

and, by Math 5

⎛ ∂P ⎞
= −⎜
⎝ ∂V ⎟⎠ T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 231

Thermal stability condition

−1
⎛ ∂ 2U ⎞ CV
⎜ 2⎟
>0 ⇒ > 0, CV > 0
⎝ ∂S ⎠ T

Mechanical Stability Condition

⎡⎛ ∂2U ⎞ ⎛ ∂2U ⎞ −1 ⎛ ∂2U ⎞ 2 ⎤ ⎛ ∂P ⎞

⎢⎜ 2 ⎟ − ⎜ 2⎟ ⎜ ⎟ ⎥>0⇒ ⎜⎝ ⎟ < 0, κT > 0
⎢⎣⎝ ∂V ⎠ ⎝ ∂S ⎠ ⎝ ∂V ∂S ⎠ ⎥⎦ ∂V ⎠ T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 232

2.6.5.4 Gibbs Phase Rule
Consider a mixture that contains N different chemical species.
The number of phases that coexist at equilibrium is denoted by π.
The variables required to describe this situation are:
P α = pressure within phase a
T α = temperature within phase a
ρα = density within phase a
x iα = mole fraction of species i within phase a

The total number of variables involved is

Variable Pα Tα ρα x αi Total

No. per phase 1 1 1 N

Total p p p Np (N+3)p

However, not all of these variables are independent. Phase equilibrium

conditions, reaction equilibrium conditions, equations of state and mass balances
all provide relations between these variables.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 233
 Equilibrium Conditions:

P α = P α +1 α = 1,2,,π − 1 π −1
T α = T α +1 α = 1,2,,π − 1 π −1
⎧α = 1,2,,π − 1
µiα = µiα +1 ⎨ N (π − 1)
⎩i = 1,2,,N

Equations of State

(
F T α , ρα ,P α = 0 ) α = 1,2,,π p

Mass Balances
N
∑ x iα = 1 α = 1,2,,π p
i =1

Reactions
equilibrium
Km m = 1,2,,r r

Total Number of Constraints (4 + N )π − N − 2 + r

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 234

 F = number of degrees of freedom
= number of independent variables
= number of variables that must be specified to describe a specific
state of the system
= number of variables - number of constraints
= [(N + 3) π ] − [ (N + 4 ) π − N − 2 + r ] = N +2−π −r

Gibbs Phase Rule

F = N + 2-π -r

where: N = number of chemical species

π = number of coexisting phases
r = number independent chemical reactions

Duhem's Theorem
For any closed system formed initially from given masses of prescribed chemical
species, the equilibrium state is determined completely when any two
independent variables are specified.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 235

Assumptions Made in Deriving the Forms shown above for the Gibbs
Phase Rule and Duhem's Theorem

• The pressure is equal in every equilibrium phase. For example, this

assumption does not hold for osmotic equilibrium.

• Specification of two variables other than composition (e.g., T and P) is

sufficient to specify the state of the system. This assumption does not hold if
several work mechanisms are present simultaneously.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 236

2.6.6 Comments On Energy Functions

2.6.6.1 Internal Energy

The internal energy appears in the formulation of the First Law of

Thermodynamics

Stationary closed system: ΔU = Q +W

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 237

2.6.6.2 Enthalpy
Because enthalpy has entropy and pressure as the natural independent
variables, it is useful for processes that occur at constant pressure. (W is work
other than PV work.)
Stationary closed system (constant P): ΔH = Q +W

Enthalpy also is useful for the mass transfer term in the energy balance equation
because it automatically includes the injection work.

Steady-state flow process: ΔH = Q +W

Maximum useful work:

H 2 − H1 = Q +W
2
Q≤ ⌠
⎮T
⌡ dS
1

Therefore

2
W ≥ H 2 − H1 − ∫T dS
1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 238

2.6.6.3 Helmholtz Energy

For the change of a system from state 1 to state 2

A2 − A1 = U 2 − U1 − (T2S2 −T1S1)

First Law for a closed stationary system: ΔU = Q +W

Then

A2 − A1 = Q +W −T2S2 +T1S1

Consider the special case where

• Heat is transferred only from a reservoir at T.
• System change is isothermal with: T1 = T2 = T

Then

A2 − A1 = Q +W −T (S2 − S1)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 239

Also
2
Q≤ ⌠
⎮T
⌡ dS = T (S2 − S1) ⇒ Q −T (S2 − S1) ≤ 0
1

W = A2 − A1 +T (S2 − S1) − Q ⇒ W ≥ A2 − A1

The maximum useful work available from an isothermal process is

Wmax = A2 − A1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 240

2.6.6.4 Gibbs Energy

For the change of a system from state 1 to state 2

G2 − G1 = U 2 − U1 + P2V2 − P1V1 − (T2S2 −T1S1)

First Law for a closed stationary system: U 2 −U1 = Q +W

Then

G2 − G1 = Q +W + P2V2 − P1V1 − (T2S2 −T1S1)

For a constant temperature and pressure process

Q ≤ T (S2 − S1) ⇒ T (S2 − S1) − Q ≥ 0

G2 − G1 = Q +W + P (V2 −V1) −T (S2 − S1)

W = (G2 − G1) − P (V2 −V1) − ⎡⎣Q −T (S2 − S1) ⎤⎦

≥ (G2 − G1) − P (V2 −V1)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 241

But, W = −P (V2 −V1) +W ' , where −P (V2 −V1) is the work done on the system by
the surroundings because of a change of system volume, and W' is all other work
done on the system. Then

W '− P (V2 −V1) ≥ (G2 − G1) − P (V2 −V1) ⇒ W ' ≥ (G2 − G1)

Then, the maximum useful work possible with a constant temperature and
pressure process is

W 'max = G2 − G1 = ΔG

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 242

2.6.8 Summary
The entropy (S) of an isolated system is maximized at equilibrium.
Energies (U,H,A,G) are minimized at equilibrium, subject to the constraint
that the natural independent variables are held constant.
Equilibrium conditions:

Pα = P β

T α =T β

µiα = µiβ

Thermal Stability Criterion

CV > 0
Mechanical Stability Criterion

⎛ ∂P ⎞
κT > 0; ⎜ <0
⎝ ∂V ⎟⎠T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 243

 3 PROPERTIES OF PURE FLUIDS
In the previous derivations, we see that the only information about individual
materials required to calculate all the UHAGS is:

Equation of state: f (P,V ,T ) = 0

Ideal gas heat capacity: CPig = CPig (T )

The equation of state contains the information about interactions between the
molecules.

The ideal gas heat capacity contains information about the energy levels within
the molecules.

Many references provide coefficients for one of the following forms for the ideal
gas heat capacity:

CPig = A + BT + CT 2 + DT 3

CPig = a + bT + cT 2 + dT −2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 244

3.1 IDEAL GAS HEAT CAPACITIES

The ideal gas heat capacity is determined completely by the characteristics of the
individual molecules.

Each degree of freedom for translation contributes 21 R to the ideal gas heat
capacity at constant volume.

Each degree of freedom for rotation contributes 21 R to the ideal gas heat
capacity at constant volume.

NOTES

A degree of freedom for translation requires freedom to move along an

independent axis in a Cartesian coordinate system.

A degree of freedom for rotation requires a significant moment of inertia about a

given independent axis of rotation. Multiple atoms must be involved to satisfy
this requirement.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 245

Monatomic molecules:
Translation: three degrees of freedom
Rotation: zero degrees of freedom

CVig = 23 R

CPig = CVig + R = 23 R + R = 52 R

Diatomic molecules:
Translation: three degrees of freedom:
Rotation: two degrees of freedom:

CVig = 23 R + R = 52 R

CPig = CVig + R = 52 R + R = 72 R

Other energy mechanisms include internal vibrations and rotations in larger

molecules such as hydrocarbons.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 246

3.2 EQUATIONS OF STATE

3.2.1 General P, V, T Behavior

Ideal Gas

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 247

Real Fluid

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 248

Water

Need to develop equations of state that describe all the behavior shown above.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 249
3.2.2 Brief History of Equations of State

1662: Boyle's Law: PV = constant

(fixed temperature and mass)

1787: Charles' Law: ∆V proportional to ∆T

(fixed temperature and mass)

1801: Dalton's Law of Partial Pressures: In a mixture, each gas behaves as

though it alone occupied the entire volume of the vessel.

1802: Gay-Lussac: Verification of Charles's Law.

1822: Cagniard de la Tour: Discovery of the critical state.

1834: Clapeyron: Combined Boyle's and Charles's laws into PV = R(t+267).

1863: Andrews: Extensive investigation of the critical state.

1873: van der Waals: Dissertation on the continuity of the gas and liquid
states. The equation of state and the concept of corresponding states.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 250

1880: Amagat's Law: The volume of a mixture of gases equals the sum of the
volumes of the component gases, each at the temperature and pressure
of the mixture.

1901: Onnes: Development of the virial equation of state as an empirical

relation.

1901: G. N. Lewis: The concept of fugacity

1927: Ursell: Statistical-mechanical development of the virial equation.

1937: Mayer: Further theoretical development of the virial equation of state.

1940: Benedict, Webb and Rubin: introduced BWR equation of state.

1949: Redlich & Kwong: an improved two parameter equation of state.

1955: Pitzer: the acentric factor as a corresponding states parameter.

1970's: Soave, Peng & Robinson: cubic equations optimized for phase
equilibrium calculations

1980's: Helmholtz equations of state

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 251

3.2.3 Important Equations of State

3.2.3.1 Compression factor

PV
Z ≡ ⇔ PV = ZRT
RT

where Z is the compression factor, compressibility factor or real gas factor.

Z is dimensionless, and it describes how the fluid differs from an ideal gas.

3.2.3.2 Virial equations

There are two forms of the virial equation:
Density (Leyden) form:
∞
Z = ∑ βn ρ n = 1+ B ρ + C ρ 2 + D ρ 3 +  ( β0 = 1)
n=0

Pressure (Berlin) form:

∞
Z = ∑ α nP n = 1+ bP + cP 2 + dP 3 +  (α 0 = 1)
n=0

In general, the density form is the more useful of the two. The pressure form is
useful at low pressures (< 5 bar) or at very high reduced temperatures (>2).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 252

3.2.3.3 Cubic Equations of State

Cubic forms of the equation of state are very popular for practical calculations,
because the cubic form is the simplest volume dependence that is capable of
describing vapor-liquid phase equilibrium

3.2.3.3.1 van der Waals

The first cubic form for an equation of state was proposed by van der Waals in
his Ph.D. dissertation (1873). The parameter a is nonzero because of attractions
between molecules, and the parameter b is nonzero because each molecule
occupies a finite volume.

Original form:
⎛ a ⎞
⎜⎝ P + 2⎟
⎠
(V − b ) = RT
V

Alternative forms:
RT a ⎛ RT ⎞ 2 a ab
P= − 2; V 3 − ⎜b + ⎟ V + V − =0
V −b V ⎝ P ⎠ P P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 253

Compressibility Factor:

V a 1 aρ
Z vdw = − = −
V − b RTV 1− b ρ RT

The equation of state also may be written with the compressibility factor as the
variable:

Z 3 − (1+ A ) Z 2 + BZ − AB = 0

where

bP
A=
RT

aP
B=
(RT )2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 254

3.2.3.3.2 General Forms for Cubic Equations of State

3.2.3.3.2.1 Martin

RT α (T ) − δ (T ) / V
P= −
V (V + β ) (V + γ )
3.2.3.3.2.2 Abbott

RT θ (V − η )
P= −
V − b (V − b ) V 2 + δV + ε ( )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 255

3.2.3.3.3 Specific Forms for Cubic Equations of State

3.2.3.3.3.1 Redlich-Kwong (1949)

Ref.: O. Redlich & J. N. S. Kwong, Chem. Review, 44, 233 - 244 (1949)

RT a
P= −
V − b V (V + b ) T

where

R 2Tc2.5 ⎛ 1 ⎞
a = Ωa ⎜ Ωa = = 0.427480⎟
Pc ⎜⎝ 9 ( 3
)
2 −1 ⎟⎠

RTc ⎛ 3
2 −1 ⎞
b = Ωb ⎜ Ωb = = 0.086640⎟
Pc ⎝ 3 ⎠

and

V a 1 aρ
Z RK = − = −
V − b RT 1.5 (V + b ) 1− b ρ RT 1.5 (1+ b ρ )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 256

3.2.3.3.3.2 Martin

RT a
P= −
V − b (V + c )2

3.2.3.3.3.3 Soave-Redlich-Kwong (1972)

Ref.: G. Soave, Chem. Eng. Science, 27, 1197-1203 (1972).

RT aα
P= −
V − b V (V + b )

or
V aα 1 aαρ
Z SRK = − = −
V − b RT (V + b ) 1− b ρ RT (1+ b ρ )

where
a = 0.42747R 2Tc2 / Pc
b = 0.08664RTc / Pc

( )( )
2
α = ⎡1+ 0.480 + 1.574ω − 0.176ω 2 1− Tr ⎤
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 257

M. S. Graboski and T. E. Daubert, in a series of papers in Ind. & Eng. Chem.,
Proc. Des. & Dev., [17, 443-448 (1978); 17, 448-454 (1978); 18, 300-306
(1979)], presented more precise values for the numerical coefficients, and a
separate expression for hydrogen:

(
α = ⎡⎢1+ 0.48508 + 1.55171ω − 0.15613ω 2 1− Tr
⎣
)( )
2⎤
⎥⎦

α = 1.202exp ( −0.30288Tr ) for hydrogen

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 258

3.2.3.3.3.4 Peng-Robinson (1976)

Ref.: D. Y. Peng & D. B. Robinson, Ind. & Eng. Chem., Fundamentals, 15,
59-64 (1976)
RT a (T )
P= −
V − b V (V + b ) + b (V − b )
or
V a (T )V 1 a (T ) ρ
Z PR = − = −
V − b RT ⎡⎣V (V + b ) + b (V − b ) ⎤⎦ 1− b ρ RT ⎡1+ 2b ρ − (b ρ )2 ⎤
⎣ ⎦
where
a (Tc ) = 0.45724R 2Tc2 / Pc
b (Tc ) = 0.07780RTc / Pc
a (T ) = a (Tc )α (Tr ,ω )

( )( )
2
α (Tr ,ω ) = [1+ 0.37464 + 1.54226ω − 0.26992ω 2
1− Tr ⎤
⎦

3.2.3.3.3.5 Harmon & Knapp

RT a
P= − 2
V − b V + bcV + b 2 (c − 1)
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 259
3.2.3.4 Complex Equations of State

3.2.3.4.1 Benedict-Webb-Rubin (BWR, 1940)

Ref.: M. Benedict, G. B. Webb & L. C. Rubin, [J. Chem. Phys., 8, 334-345

(1940); 10, 747-758 (1942)], [Chem. Eng. Progress, 47, 419 (1951); 47,
449-454 (1951)]

P=
RT ⎛ C ⎞ 1 1 aα
+ ⎜ B0RT − A0 − 02 ⎟ 2 + (bRT − a ) 3 + 6 +
(
c 1+ γ / V 2 ) exp ⎛ − γ
⎞
⎝ ⎜⎝ ⎟
V T ⎠V V V T 2V 3 V2⎠

Parameters: A0 , B0 , C0 , a, b, c, α , γ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 260

3.2.3.4.2 Benedict-Webb-Rubin-Starling (BWRS, 1970)

Ref. K. E. Starling, Thermodynamic Properties for Light Petroleum

Systems, Gulf Publishing (1973).

RT ⎛ C D E ⎞ 1 ⎛ α⎞ 1
P= + ⎜ B0RT − A0 − 02 + 03 − 04 ⎟ 2 + ⎜ bRT − a − ⎟ 3
V ⎝ T T T ⎠V ⎝ T ⎠V

⎛ α ⎞ 1
+α ⎜ a + ⎟ 6 +
c (
1+ γ / V 2
) ⎛ γ ⎞
⎝
exp ⎜⎝ − 2 ⎟⎠
T ⎠V 2 3
T V V

Parameters: A0 , B0 , C0 , D0 , E 0 , a, b, c, α , γ

Compare with BWR

P=
RT ⎛ C0 ⎞ 1 1 aα c 1+ γ / V
+ ⎜ B0RT − A0 − 2 ⎟ 2 + (bRT − a ) 3 + 6 +
2
⎛ γ ⎞ ( )
⎝ ⎠ 2 3
exp ⎜⎝ − 2 ⎟⎠
V T V V V T V V

Parameters: A0 , B0 , C0 , a, b, c, α , γ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 261

3.2.3.4.3 NBS/NRC Steam Tables (1984)

Ref.: L. Haar, J. S. Gallagher and G. S. Kell, NBS/NRC Steam Tables,

Hemisphere Publishing, 1984

The working function is an expression for the Helmholtz energy

A ( ρ,T ) = Abase ( ρ,T ) + Aresidual ( ρ,T ) + Aideal gas ( ρ,T )

where
⎡
⎢
⎛ bρ ⎞ β − 1 α + β +1
Abase ( ρ,T ) = RT ⎢ − ln ⎜ 1− ⎟ − +
⎢ ⎝ 4 ⎠ 1− b ρ ⎛ bρ ⎞
2
⎢ 4 2 ⎜⎝ 1− 4 ⎟⎠
⎣
⎛B ⎞ α − β +γ ρRT ⎤
+b ρ ⎜ − γ ⎟ − + ln ⎥
⎝b ⎠ 2 P0 ⎦
k
T ⎛T ⎞
b = b1 ln + ∑ bk ⎜ 0 ⎟
T0 k =0,1,3,5 ⎝ T ⎠
k
⎛T ⎞
B = ∑ Bk ⎜ 0 ⎟
k =0,1,2,4 ⎝T ⎠
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 262
 (i )
g ⎛T ⎞ − ρ k (i )
(1− e ) (i )
( (i )
)
36
Aresidual ( ρ,T ) = ∑ i ⎜ ⎟ + ∑ gi δ i exp −α i δ i − βi τ i2
i =1 k (i ) ⎝ T0 ⎠

ρ − ρi
δi =
ρi
T −Ti
τi =
Ti

⎡ ⎛ C1 ⎞ 18
i −6 ⎤
Aideal gas (T , ρ ) = −RT ⎢ ⎜
1+ + C 2⎟ lnTR + ∑ C i R ⎥
T
⎣ ⎝ TR ⎠ i =3 ⎦

T
TR =
100 K

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 263

3.2.3.4.4 Tilner-Roth (Difluoromethane, R-32)

Ref.: R. Tilner-Roth and A. Yokozeki, J. Phys. Chem. Ref. Data, 26, 1273-
1328 (1997).

Dimensionless Helmholtz free energy

A Aig Ar
= + = Φig + Φr
RT RT RT

ρ T
where, in terms of the reduced variables, δ = ; τ= c
ρc T

the ideal gas and residual terms are given by

6
Φ ig
= lnδ + a0ig + a1ig τ + a2ig lnτ + ∑ aiig ln ⎡⎣1− exp ( −ni τ ) ⎤⎦
i =3

( )
8 19
Φ = ∑ ai δ τ + ∑ ai δ d i τ t i exp −δ ei
r di ti
i =1 i =9

There are seven terms in the ideal gas formulation and nineteen terms in the
residual formulation.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 264

3.2.3.4.5 Wagner & Pruß (Water, IAPWS-95)

Ref.: W. Wagner and A. Pruß, J. Phys. Chem. Ref. Data, 31, 375-535 (2002).

Using forms and the reduced variables defined on the previous page, the ideal
gas and residual terms are given by
8
ig
Φ = lnδ + n1o + n2oτ + n3o lnτ + ∑ nio ln ⎡⎣1− exp ( −γ i τ ) ⎤⎦
i =4

( )
7 51
Φ = ∑ ni δ τ + ∑ ni δ d i τ t i exp −δ c i
r di ti
i =1 i =8
54 56
+ ∑ ni δ τ exp ⎡ −α i (δ − ε i ) − βi (τ − γ i ) ⎤ + ∑ ni Δbi δψ
di ti 2 2
i =52 ⎣ ⎦ i =55

where
ai
Δ = θ + Bi ⎡(δ − 1) ⎤
2 2
⎣ ⎦
1/ ( 2 βi )
θ = (1− τ ) + Ai ⎡(δ − 1) ⎤
2
⎣ ⎦
ψ = exp ⎡ −Ci (δ − 1) − Di (τ − 1) ⎤
2 2
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 265

3.2.3.4.6 Ratio of Poynomials (Rational) Equation of State

Ref.: M. Gomez Osorio, et al., Molecular Physics, 115, 1417-1434 (2017).

This approach uses the same form for the ideal gas contribution as those used
previously in the high accuracy Helmholtz equations, but the residual form is
replaced by a ratio of polynomials in density which have temperature dependent
coefficients.

∑IiN=1N i (τ )δ i JNi JDi

Φ = ID
r
; N i (τ ) = ∑c ij τ ;
j
Di (τ ) = ∑ bij τ j
∑ i =0 Di (τ )δ i j =0 j =0

For argon, nitrogen and methane, IN = 6, ID = 3, JNi = 4, JDi = 4

This form is slightly less accurate near the critical temperature and pressure than
the previous form, but, because there are no exponential terms in density, the
computation time required to solve for denisties is reduced by more than a factor
of 20. This form does require about 15 times the computational time as cubic
equations, but the accuracies are much better.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 266

Properties from Helmholtz Functions:

⎛ ∂A ⎞
P = ρ2 ⎜
⎝ ∂ ρ ⎟⎠ T

⎛ ∂A ⎞
S = −⎜
⎝ ∂T ⎟⎠ ρ

U = A +TS

P
H =U +
ρ

P
G = A+
ρ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 267

3.2.4 Corresponding states

3.2.4.1 2 parameter

At the critical point

⎛ ∂P ⎞ ⎛ ∂2 P ⎞
⎜⎝ ⎟⎠ = 0 and ⎜ 2⎟
=0
∂V T c
⎝ ∂V ⎠T
c

Therefore, two constraints are imposed on the parameters of the equation of

state. For two parameter equations, the values of the parameters are fixed in
terms of the critical properties. The equation of state then can be written as
functions of reduced variables

f (Pr ,Vr ,Tr ) = 0

where
P V T
Pr = ; Vr = ; Tr =
Pc Vc Tc

The mathematical form is the same for every substance.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 268

3.2.4.2 3 parameter

Ref.: K. S. Pitzer, R. F. Curl, et al., Ind. & Eng. Chem., 50, 265-274 (1958).

The acentric factor w was introduced as a third parameter for corresponding

states. The compressibility factor was correlated in the form

Z = Z 0 + ωZ1

where

⎛ ⎞
ω ≡ − log10 ⎜ Prvap ⎟ −1
⎝ Tr =0.7 ⎠

This correlation was presented as charts to be used for graphical determination

of the properties.

Because they can be derived directly from Z and its first derivative with respect to
temperature, the energies (U, H, A, G) and the entropy (S) also may be
represented using the same form.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 269

3.2.4.2.1 Lee-Kesler (1975)

Ref.: B. I. Lee & M. G. Kesler, AIChE J., 21, 510-527 (1975).

Lee & Kesler developed an equation that provides a numerical representation of

the Pitzer-Curl three parameter corresponding states. The formulation utilizes
equations for two fluids: 1.) a composite of simple fluids (argon, krypton, xenon);
and 2.) a large molecule reference fluid (iso-octane).

Z = Z (0) +
ω
ω (r )
Z ( )
(r ) − Z (0 ) = Z (0 ) + ω Z (1)

ω (r ) = 0.3978

PrVr B C D c4 ⎛ γ ⎞ ⎛ γ ⎞
Z = = 1+ + + + ⎜ β + 2 ⎟ exp ⎜ − 2 ⎟
Tr Vr Vr2 Vr3 Tr3Vr2 ⎝ Vr ⎠ ⎝ Vr ⎠

B = b1 − b2 / Tr − b3 / Tr2 − b4 / Tr3

C = c1 − c 2 / Tr + c 3 / Tr3
D = d1 + d 2 / Tr

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 270

Lee-Kessler Parameters

Constant Simple Fluid Reference Fluid

b1 0.1181193 0.2026579
b2 0.265728 0.331511
b3 0.154790 0.027655
b4 0.030323 0.203488
c1 0.0236744 0.0313385
c2 0.0186984 0.0503618
c3 0.0 0.016901
c4 0.042724 0.041577
d1 0.155488´10-4 0.48736´10-4
d2 0.623689´10-4 0.0740336´10-4
b 0.65392 1.226
g 0.060167 0.03754

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 271

3.2.4.3 4 Parameter Corresponding States

For polar fluids, the dipole moment of the molecule (µ) often is used as a fourth
parameter in equations of state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 272

3.2.4.4 Extended corresponding states

A more sophisticated form of corresponding states, which incorporates

information from statistical mechanics calculations and from extensive
experimental values for a reference fluid, often is referred to as Extended
Corresponding States. In this representation, the compressibility factor and
Helmholtz energy are expressed in terms of reduced variables in the form

⎛ V T ⎞
Z i (V ,T ) = Z o ⎜ , ⎟
⎝ hi ,o fi ,o ⎠

r ⎛ V T ⎞
Air (V ,T ) = fi ,o Ao ⎜ , ⎟
⎝ i ,o fi ,o ⎠
h

where: fi,o and hi ,o are transformation parameters given by

Toc Voc
fi ,o = c
θ (Tr ,Vr ) and hi ,o = c
φ (Tr ,Vr )
Ti Vi

Here the functions θ (Tr ,Vr ) and φ (Tr ,Vr ) are shape factors.

The subscripts i and o refer to the fluid of interest and the reference fluid
respectively.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 273
3.2.4.4.1 Modified BWRS (NIST form)

(
P = ρRT + ρ 2 N1T + N 2T 0.5 + N 3 + N 4T −1 + N 5T −2 )
( ) (
+ ρ 3 N 6T + N 7 + N 8T −1 + N 9T −2 + ρ 4 N10T + N11 + N12T −1 + ρ 5N13 )
+ ρ6 (N T + N T
14
−1
15
−2
) + ρ (N
7
16T ) + ρ (N
−1 8
17T
−1
+ N18T −2)
+ ρ9 (N T )
19
−2

+ρ3 (N T + N T
20
−2
21
−3
) exp (−γρ )2

+ ρ5 (N T + N T
22
−2
23
−4
) exp (−γρ )2

+ ρ7 (N T + N T
24
−2
25
−3
) exp (−γρ )2

+ ρ9 (N T + N T
26
−2
27
−4
) exp (−γρ )2

+ ρ11 (N T + N T
28
−2
29
−3
) exp (−γρ )
2

+ ρ13 (N T + N T
30
−2
31
−3
+ N T ) exp ( −γρ )
32
−4 2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 274

3.3 MOLECULAR INTERACTIONS

3.3.1 The Fundamentals

Need forces (potentials) to predict molecular motions.

All forces of relevance here, e.g. interactions among electrons and nuclei, are
electrostatic in nature.

Fundamental to everything is the Schrödinger equation (any physical chemistry

textbook)

∂Ψ
HΨ = i 
∂t
where
Ψ (R,r,t ) is the wave function
⎧  ⎫
H = K + U = ⎨− ∑ ∇i2 + U ⎬ is the Hamiltonian
⎩ 2m i ⎭

Time-independent form

HΨ = EΨ
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 275
Born-Oppenheimer approximation: Electronic motion is very fast compared to
nuclear motion so that an average can be taken over the electrons for a given set
of nuclear coordinates.

{Hel +Unuc } Ψ (r;R ) = E (R ) Ψ (r;R )

By solving this for different R, you obtain the function E(R). This is the energy of
the system as a function of nuclear configuration, also known as the potential
energy surface.

The potential energy surface is everything you need to know to describe the
motion of the nuclei in the system.

In principle, the problem is solved!

In practice:
• except in the simplest cases, the Schrödinger equation cannot be solved
exactly
• even with accurate approximations, the computational cost of solving it is
high

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 276

3.3.2 Force-field concept

Define potentials (forces) using empirical formulas.

Atoms (or groups of atoms) are the fundamental units. Their nuclei (centers)
become the focal points for the empirical mathematical expressions.

Intra- and inter- molecular contributions exist.

Example: interactions characteristic of butane and methane

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 277

3.3.3 Intramolecular potentials

3.3.3.1 Bond-stretching

Example equation

Ubond = k (r − r0 )
2

3.3.3.2 Angle-bending

Example equation

Uang = k (θ − θ 0 )
2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 278

3.3.3.3 Torsional rotation

Example equation
3
Utor = ∑ k i ⎡⎣1− cos (φ − φ0i ) ⎤⎦
i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 279

3.3.4 Improper out-of-plane potential

Generally used for a set of three atoms connected to an sp2 hybridized central
atom (amide N, sp2 C).

Example equation

Uoop = k χ 2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 280

Combinations of the above (a.k.a. cross-terms) are also included in modern
force-fields.

Example of a full intra-molecular potential energy expression (MSI CFF family)

Evalence = ∑ ⎡k 2 (b − bo ) + k 3 (b − bo ) + k 4 (b − bo ) ⎤
2 3 4
b ⎣ ⎦

+ ∑ ⎡k 2 (θ − θo ) + k 3 (θ − θo ) + k 4 (θ − θo ) ⎤
2 3 4
θ ⎣ ⎦
+ ∑ ⎡⎣k1 (1− cos φ ) + k 2 (1− 2 cos φ ) + k 4 (1− cos 3φ ) ⎤⎦
φ

+ ∑ ⎡k 2 ( χ − χ o ) ⎤
2
χ⎣ ⎦
+ ∑ ⎡⎣k (b − bo ) (b '− b 'o ) ⎤⎦ + ∑ ⎡⎣k (b − bo ) (θ − θo ) ⎤⎦
b,b ' b,θ

+ ∑ ⎡⎣k (b − bo ) (1− cos φ ) ⎤⎦ + ∑ ⎡⎣k (θ − θo ) (θ '− θ 'o ) ⎤⎦

b,φ b,θ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 281

3.3.5 Van der Waals and Coulombic potentials

3.3.5.1 Van der Waals

Consider two atoms of a neutral, monatomic species (e.g. Ar)

+ - + -
instantaneous induced
fluctuation (dipole) dipole
here here

Although the time-averaged dipole for each atom is zero, the net potential is not.

The potential energy is attractive, and is given by (Reed and Gubbins,1973)

c6 c8 c10
U att = − − − −
r6 r8 r 10

The first term is generally dominant.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 282

When the atoms get too close, the energy becomes repulsive, due to non-
bonding overlap between electron clouds.

Since the physics of repulsion is not as straightforward as that of attraction,

different forms have been used to represent the repulsive component of the vdW
interaction.

Two common, mathematically convenient forms are

c12 and c9
U rep = + U rep = +
r 12 r9

A form with more physical basis is

Urep = +Be −br

(Reed and Gubbins, 1973)

The exponential repulsion has been shown to better predict VLE properties of
hydrocarbons (Errington and Panagiotopoulos, J. Phys. Chem. B 103, 6314,
1999).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 283

Van der Waals interactions always exist between any two chemical groups,
although they are not always the dominant contribution.

VdW interactions are usually handled in a pairwise fashion between sites on

different molecules.

They are usually turned off (or scaled down) for the so-called 1-2, 1-3, and 1-4
intra-molecular interactions.

4 5

2 3

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 284

The combination of the r -6 attraction and the r -12 repulsion produces the
Lennard-Jones potential (probably the most-studied model fluid!).

⎡⎛ σ ⎞ 12 ⎛ σ ⎞ 6 ⎤
ULJ = 4ε ⎢ ⎜ ⎟ − ⎜ ⎟ ⎥
⎢⎣⎝ r ⎠ ⎝r⎠ ⎥
⎦

Example: comparison of LJ and exponential-six potentials for argon

-21
15x10
Exponential-6
Lennard-Jones
Energy (J)

10

parameters from
5
Reed and Gubbins,1973

2 3 4 5 6 7 8

r (Å)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 285

3.3.5.2 Coulombic

Coulomb’s Law: the potential energy between two point charges is given by (any
physics text)
1 q1q 2
UCoul =
4πε 0 r

or
e 2 z1z 2
UCoul =
4πε 0 r

with
e = 1.602 x 10-19 C
e0 = 8.8542 x 10-12 C2/(N m2)

Obviously, ionic species have net charges that must be accounted for in this way.

Even in neutral molecules, there are generally regions that are electron-rich and
regions that are electron-poor. These regions can be represented by local
charges.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 286

Here is an example of the complexities that can arise in a model of the simplest
of molecules, N2 (Allen and Tildesly,1987).

- + - + -
3 2 1 2 3

with
z1 = -2.3794
z2 = +5.2366
z3 = -4.0469

This model captures the quadrupole moment for N2.

As in the case of vdW forces, a pairwise additivity assumption is generally made

for Coulombic forces.

For convenience (especially for MD), the locations of the charges are usually
chosen to coincide with the vdW force centers. The N2 model above is one
exception.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 287

Example of a full force-field expression
E potential = ∑ ⎡k 2 (b − bo ) + k 3 (b − bo ) + k 4 (b − bo ) ⎤
2 3 4
b ⎣ ⎦

+ ∑ ⎡H 2 (θ − θo ) + H 3 (θ − θo ) + H 4 (θ − θo ) ⎤
2 3 4
θ ⎣ ⎦

{ ( ) ( )
+ ∑ V1 ⎡1− cos φ − φ10 ⎤ +V2 ⎡1− cos 2φ − φ20 ⎤ +V3 ⎡1− cos 3φ − φ30 ⎤
φ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ( )}
+ ∑ K 2 χ 2 + ∑ ∑ ⎡⎣Fbb ' (b − bo ) (b '− b 'o ) ⎤⎦ + ∑ ∑ ⎡⎣Fθθ ' (θ − θo ) (θ '− θ 'o ) ⎤⎦
χ b b' θ θ'

+ ∑ ∑ ⎡⎣Fbθ (b − bo ) (θ − θo ) ⎤⎦
b θ
+ ∑ ∑ ⎡⎣(b − bo ) (V1 cos φ +V2 cos 2φ +V3 cos 3φ ) ⎤⎦
b φ

+ ∑ ∑ ⎡⎣(b '− b 'o ) (V1 cos φ +V2 cos 2φ +V3 cos 3φ ) ⎤⎦

b' φ

+ ∑ ∑ ⎡⎣(θ − θo ) (V1 cos φ +V2 cos 2φ +V3 cos 3φ ) ⎤⎦

θ φ

qi qk ⎡A Bik ⎤
+ ∑ ∑ ∑ ⎡⎣K φθθ ' cos φ (θ − θo ) (θ '− θ 'o ) ⎤⎦ + ∑ + ∑ ⎢ ik − 6 ⎥
φ θ θ∍ i >k ε rik r 9
i >k ⎣ ik rik ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 288

3.3.6 Three-body potentials

The assumption of pairwise vdW interactions is an approximation.

For example, the interaction between the two blue atoms is certainly different in
the two configurations,

but within the pairwise approximation, it is assumed to be the same.

For an assembly of atoms, we can write formally

UTOT = ∑ uij + ∑ w ijk + ∑ v ijkl +

i <j i < j <k i < j <k <l

This expansion is rigorous when summed to all orders. However, evaluating the
higher-order terms becomes computationally expensive.

Let’s examine the importance of 3-body contributions.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 289

3.3.6.1 Three-body vdW potential

A three-body vdW correction term was derived by Axilrod and Teller (Reed and
Gubbins,1973).

r12 q2 r23
q1 q3
r13

U TOT = upairwise +w 123

3cos θ1 cos θ 2 cos θ 3 + 1

w 123 = c123
(r12r13r23 )3
For acute triangles, w123 is positive.

For most obtuse triangles, w123 is negative.

(The three-body potential tends to stabilize near-linear configurations.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 290

In a typical condensed phase, the overall contribution of the three-body term is
positive and constitutes a correction on the order of 10% to the total energy
(Allen and Tildesley,1987).

This correction is generally deemed not sufficiently important to justify the (very
costly) rigorous sum of the three-body terms. Instead, the correction enters
through an “effective” two-body potential (Allen and Tildesley,1987). That is, the
parameters for the LJ two-body potential are chosen to reproduce the properties
of condensed phases, not isolated gas-phase pairs.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 291

3.3.7 Equations Derived From Interatomic Potentials

Statistical mechanics provides insight into the mathematical forms of equations of

state that describe the macroscopic behavior of fluids based upon the
interactions between molecules (interatomic potential). For example, the second
virial coefficient is related to the interatomic potential as follows
∞
⌠ ⎡ ⎛ u (r ) ⎞ ⎤ 2
B= 2π Na ⎮⎮ ⎢1− exp ⎜⎝ − kT ⎟⎠ ⎥ r dr
⎮
⌡ ⎣ ⎦
0

Because the interatomic potential appears in the exponential argument, only

certain select forms lead to integrals that may be evaluated analytically.

Some commonly used potentials are discussed below.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 292

Ideal Gas:

The ideal gas has no intermolecular forces, therefore the potential is

u (r ) = 0 0<r <∞

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 293

Point of Repulsion at Center of Molecules

Two such potentials are

u (r ) = ar −α

and the Born potential

⎛ r⎞
u (r ) = a exp ⎜ − ⎟
⎝ σ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 294

Hard Spheres

The hard sphere potential is

⎧∞ 0 <r <σ
u (r ) = ⎨
⎩0 σ <r <∞

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 295

Square-Well Potential

The square-well potential is

⎧∞ 0 <r <σ
⎪
u (r ) = ⎨ − ε σ < r < gσ
⎪0 gσ < r < ∞
⎩

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 296

Sutherland Potential

The Sutherland potential is

⎧∞ 0 <r <σ
⎪
u (r ) = ⎨ ⎛ σ ⎞ α
⎪− ε ⎜⎝ r ⎟⎠ σ <r <∞
⎩

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 297

Yukawa Potential

The Yukawa potential is

⎧∞ 0 <r <σ
⎪
u (r ) = ⎨ ⎛ r −σ ⎞
⎪ − ε exp ⎜⎝ − ⎟ σ <r <∞
⎩ z ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 298

Bireciprocal

The general bireciprocal form is

α β
⎛σ ⎞ ⎛σ ⎞
u (r ) = a ⎜ ⎟ − b ⎜ ⎟
⎝r⎠ ⎝r⎠

The Lennard-Jones potential is a special case of the bireciprocal potential

⎡⎛ σ ⎞ 12 ⎛ σ ⎞ 6 ⎤
uLJ (r ) = 4ε ⎢⎜ ⎟ − ⎜ ⎟ ⎥
⎢⎣⎝ r ⎠ ⎝r⎠ ⎥
⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 299

Kihara Potential

The Kihara potential is

⎧∞ 0<r <d
⎪
u (r ) = ⎨ ⎡⎛ σ − d ⎞ 12 ⎛ σ − d ⎞ 6 ⎤
⎪4ε ⎢⎜⎝ r − d ⎟⎠ − ⎜⎝ r − d ⎟⎠ ⎥ d ≤r <∞
⎩ ⎢⎣ ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 300

3.3.7.1 Perturbed Hard Spheres

3.3.7.1.1 Ree & Hoover (1967)

PR =
RT
V
(
1+ 4y + 10y 2 + 18.36y 3 + 28.3y 4 +  )
where PR is the pressure due to repulsive interactions and y = b
4V

3.3.7.1.2 Carnahan-Starling (1972)

Carnahan and Starling combined the repulsive form suggested by hard spheres
with simple attractive contributions to obtain useful equations of state.

RT ⎡ 1+ y + y 2 − y 3 ⎤ a
CS w/ van der Waals attractive term: P= ⎢ ⎥− 2
V ⎢⎣ (1− y ) 3
⎥⎦ V

RT ⎡ 1+ y + y 2 − y 3 ⎤ a
CS w/ Redlich-Kwong attractive term: P= ⎢ ⎥ −
V ⎢⎣ (1− y )3 ⎥⎦ T (V + b )V
Many other forms of perturbed hard sphere equations have been proposed.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 301

3.3.7.1.3 Boublik-Alder-Chen-Kreglewski (BACK)

Z = Z hb + Z a

where Z hb is the contribution to the compressibility factor arising from repulsive

interactions:
hb
Z
( 3α + 1)
= 1+
ξ + 3α 2
− 3α − 2 ξ 2
+ 1− (
α 2
ξ 3
) ( )
(1− ξ )3
and Z a is the contribution to the compressibility factor arising from attractive
interactions:
n
⎛ 1 ⎞
a
Z = ∑ ∑ mDnm ⎜
⎝ T * ⎟⎠
( ρ * )m
nm

The parameters and variables are

α = shape factor (due to Boublik)
ξ = 0.74048 ρ *
V°
ρ* =
V
Dnm = universal numerical constants (24 of these are used in the BACK
formulation)
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 302
3.3.7.1.4 PC-SAFT Equation of State

Perturbed Chain Statistical Associating Fluid Theory

This approach considers separately the contributions to Z that come from hard
spheres, assembly into chains, dispersion (van der Waals) forces and specific
interactions (e.g., hydrogen bonding):

Ar = Ahs
r r
 +Adisp r
 +Achain r
 +Aassoc

(The colors in this equation correspond to the colors of the arrows in the figure.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 303

Potential for Hard Sphere and Dispersion Contribution

⎧∞ r < (σ − s1); s1 = 0.12σ

⎪
⎪3ε
u (r ) = ⎨
(σ − s1) ≤ r < σ
⎪− ε σ < r < λσ
⎪0 r ≥ λσ
⎩

Fluid specific parameters for nonassociating fluids:

s = segment diameter (length dimensions)

e = depth of potential well (energy dimensions)

m = number of segments per chain

Fluid specific parameters for association contribution

κ Ai Bk = volume of association

ε Ai Bk = energy of association

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 304

3.4 PURE FLUID PHASE EQUILIBRIUM

The stability criteria resulting from the Second Law are that the entropy is
maximized and U, H, A and G are minimized at equilibrium. When the energy
functions are minimized, the natural independent variables are held constant.

For calculations of phase equilibrium, we consider one mole of fluid subject to

constraints of constant temperature and pressure, therefore the equilibrium
configuration of the system is that configuration that has the lowest Gibbs
energy.

The appropriate form of the Physics Summary Statement is

dG = −SdT +VdP

For changes that occur at constant temperature (along an isotherm)

dG =VdP (constant temperature)

Calculate the Gibbs energy along an isotherm relative to the Gibbs energy at a
reference pressure, P†, at the same temperature.
P
†
G −G = ⌠
⎮
⌡ V dP (constant temperature)
P†
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 305
Equation of state behavior

V P

P V

The behavior of V as a dependent variable is not a proper integrable

mathematical function because multiple values of V exist for certain single values
of P. Pressure cannot be used as the variable of integration, so the expression
for the Gibbs energy must be transformed to obtained a proper integral.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 306

Note: d (PV ) = PdV +VdP ⇒ VdP = d (PV ) − PdV

Then:

dG =VdP = d (PV ) − PdV

and

P PV V
G −G = † ⌠
⎮
⌡ V dP = ⌠
⎮
⌡
d (PV ) − ⌠
⎮
⌡ P dV
P† P †V † V†

( )−
V
† †
= PV − P V ⌠
⎮
⌡ P dV
†
V

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 307

3.4.1 Gibbs Energy of van der Waals Fluid

Reduced form of van der Waals equation

8Tr 3
Pr = − 2
3Vr − 1 Vr

In terms of reduced variables, the Gibbs energy expression given above is

PrVr Vr
G − G † = PcVc ⌠
⎮
⌡
d (PrVr ) − PcVc ⌠
⎮
⌡ Pr dVr
† † †
Pr Vr Vr

For the van der Waals fluid

3
PcVc = Z cvdw RTc = RTc
8

Substitution into the Gibbs energy produces

† PrVr Vr
G −G 3 3
= ⌠
⎮
⌡
d (PrVr ) − ⌠
⎮
⌡ Pr dVr
RTc 8 † † 8 †
Pr Vr Vr

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 308

The first integral is

PrVr
⌠
⎮
⌡
d (PrVr ) = PrVr − P †V †
† †
Pr Vr

The second integral is

Vr Vr
Vr ⌠ dVr ⌠ dV 8 ⎡ 3V − 1 ⎤ ⎡1 1 ⎤
⌠
⎮
⌡ Pr dVr = 8Tr ⎮⎮ − 3 ⎮ 2 = Tr ln ⎢ r† ⎥ + 3 ⎢ − † ⎥
⎮ r
† ⌡ 3Vr − 1 ⌡ Vr 3 ⎢⎣ 3Vr − 1⎥⎦ ⎣Vr Vr ⎦
Vr † †
Vr Vr

The reduced Gibbs energy then is

( )
G − G† 3 † † ⎛ 3Vr − 1⎞ 9 ⎛ 1 1⎞
= Pr Vr − Pr Vr − Tr ln ⎜ † ⎟ − ⎜ − † ⎟
RTc 8 ⎝ 3Vr − 1⎠ 8 ⎝ Vr Vr ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 309

The Gibbs energy as a function of volume is

The Gibbs energy as a function of pressure is

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 310

For fluid densities (specific volumes) between the volume of the vapor and the
liquid at the intersection on the G-P diagram, the fluid has a lower total Gibbs
energy when divided into two phases of different densities than it has as a single
phase at the same pressure.

The equilibrium criteria state that the equilibrium condition has the lowest Gibbs
energy at constant T and P, so the equilibrium condition has two phases for a
range of densities at one pressure (the intersection point on the plot shown
above.

The molar Gibbs energy (µI for a pure fluid) of the liquid and vapor phases must
be equal at the equilibrium condition.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 311

3.4.2 Vapor pressure from EOS

Compare the molar Gibbs energy at two points, 1, and 2, along the PV isotherm
V2
G ( 2 ) − G = P2V2 − P V −
† † † ⌠
⎮
⌡ P dV
V†

V1
G (1) − G = P1V1 − P V −
† † † ⌠
⎮
⌡ P dV
V†

Subtraction produces
V2
G ( 2 ) − G (1) = P2V2 − P1V1 − ⌠
⎮
⌡ P dV
V1

In the case of two coexisting phases, the Gibbs energies and pressures are
equal, therefore
V2sat
(
G ( 2 ) − G (1) = P eq V2sat −V1sat − ∫ P dV = 0
sat
) V1

For vapor-liquid equilibrium

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 312

 VVsat
GV − GL = P sat V ( sat
V
sat
)
−VLsat − ∫ P dV = 0
VL

Equilibrium States

GV − GL > 0 equilibrium state is compressed liquid

GV − GL < 0 equilibrium state is superheated vapor

GV − GL = 0 equilibrium state is two phase (vapor+liquid)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 313

3.4.2.1 Vapor Pressure Calculation Procedure

1. σ
Estimate the vapor pressure, Pest

2. Solve the equation of state for VL, VV.

σ
⎧< 0, Pest < Pσ
VV ⎪⎪
3. Calculate GV − GL = P (VV −VL ) − ∫ P dV ⎨= 0, Pest
σ σ
= Pσ
VL ⎪ σ σ
⎪⎩ > 0, Pest > P

σ (subject to the constraint that σ

4. Adjust Pest Pest > 0 ), using

GV − GL
σ
Pest (new ) = Pest
σ
(old ) −
VV −VL

5. Repeat steps 2 through 4 until convergence is achieved.

Note: The formula for the new estimate used in step 4 is based on the
observation that
d (GV − GL )
σ
= VV −VL
dPest

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 314

Clapeyron equation
One of the equilibrium criteria for two phases coexisting in equilibrium is
Gα = G β
where the superscripts a and b denote the properties of the two coexisting
phases.
For a path along the saturation curve
dGα = dG β
Apply the Physics Summary statement (dG=-SdT+VdP) to each phase
−S α dT +V α dP = −S β dT +V β dP

(V α −V β )dP = (Sα − S β )dT

This leads to the Clapeyron Equation

dP σ S α − S β Hα − H β
= α =
dT V −V β
(
T V α −V β )
This is a general result that applies to transitions in pure materials (e.g. vapor
pressures, sublimation pressures, melting pressures, solid-solid, etc.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 315

3.4.2.2 Implications for Vapor Pressure Equations

Case I:
Assumptions: ΔLV H is independent of temperature
vapor is an ideal gas
vapor volume is much larger than liquid volume

The Clapeyron equation yields

ΔLV H B
lnP σ = − + constant = A −
RT T

Case II:

Assumptions: ΔLV H = ΔLV H 0 + ( ΔLV H1)T

vapor is an ideal gas
vapor volume is much larger than liquid volume

The Clapeyron equation yields

ΔLV H 0 ΔLV H1 B
lnP σ = − + lnT + constant = A − + C lnT
RT R T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 316

3.4.3 Temperature dependence of DsH

Relate the heat of a phase transformation to other measurable quantities. For

enthalpy considered as a function of temperature and pressure. math relation 1
leads to

⎡ ∂( Δσ H ) ⎤ ⎡ ∂( Δσ H ) ⎤
d ( Δσ H ) = ⎢ ⎥ dT + ⎢ ⎥ dP
⎣ ∂T ⎦P ⎣ ∂P ⎦T

The coefficients can be related to observables

⎡ ∂( Δσ H ) ⎤
⎢
⎡ ∂ Hα − H β
⎥ =⎢
( ) ⎤⎥ = CPα − CPβ = Δσ CPσ
⎣ ∂T ⎦P ⎢⎣ ∂T ⎥
⎦P

⎡ ∂( Δσ H ) ⎤ ⎡ ∂ Hα − H β
⎥ =⎢
( ) ⎤⎥ ⎡
α ⎛ ∂V α ⎞ ⎤ ⎡ β
= ⎢V −T ⎜
⎛ ∂V β ⎞ ⎤
⎢ ⎟ ⎥ − ⎢V −T ⎜ ∂T ⎟ ⎥
⎣ ∂P ⎦T ⎢⎣ ∂P ⎥
⎦T ⎣
⎢ ⎝ ∂T ⎠ P ⎥⎦ ⎢⎣ ⎝ ⎠ P ⎥⎦

⎛ ∂Δ V ⎞
= ΔσV −T ⎜ σ ⎟
⎝ ∂T ⎠ P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 317

Then
⎡ ⎛ ∂Δ V ⎞ ⎤
d ( Δσ H ) = Δσ CP dT + ⎢ ΔσV −T ⎜ σ ⎟ ⎥ dP
⎣ ⎝ ∂T ⎠ P ⎦

and
d ( Δσ H ) ⎡ ⎛ ∂ΔσV ⎞ ⎤ ⎛ dP ⎞
σ
= Δσ CP + ⎢ ΔσV −T ⎜
dT ⎣ ⎝ ∂T ⎟⎠ P ⎥⎦ ⎜⎝ dT ⎟⎠

⎡ ⎛ ∂Δ V ⎞ ⎤ ⎛ Δ H ⎞
= Δσ CP + ⎢ ΔσV −T ⎜ σ ⎟ ⎥ ⎜ σ ⎟
⎣ ⎝ ∂T ⎠ P ⎦ ⎝ T ΔσV ⎠

Δσ H Δσ H ⎡ T ⎛ ∂ΔσV ⎞ ⎤
= Δσ CP + − ⎢ ⎜⎝ ⎟⎠ ⎥
T T Δ
⎣ σ V ∂T P⎦
or

d ( Δσ H ) Δ H ⎡ ⎛ ∂lnΔσV ⎞ ⎤
== ΔσC P + σ ⎢1− ⎜⎝ ∂lnT ⎟⎠ ⎥
dT T ⎣ P⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 318

3.4.4 Trouton's Rule

This rule is based upon the observation that the entropy of boiling is
approximately the same for all fluids:

ΔSboiling ≈ 88 J/K ⋅mol

Note also that

ΔHboiling = Tbp ΔSboiling

This observation may be understood by noting that the ratio of the vapor and
liquid molar volumes is of the same order of magnitude for all fluids.

A similar, but less general, relation exists for entropies of melting:

ΔSmelting ≈ 8 − 16 J/K ⋅mol

Note that some substances have entropies of melting as large as 40 J/K·mol

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 319

3.4.5 Types of phase transitions

For various real substances, the Gibbs surfaces in P,T space exhibit a variety of
behaviors.

General behavior:

The equilibrium phase has the lowest Gibbs energy at fixed T,P. At the
intersections of the surfaces, both phases can exist in equilibrium. Metastable
states also may exist in the vicinity of the intersection.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 320

Recall that

⎛ ∂G ⎞ ⎛ ∂G ⎞
⎜⎝ ⎟ = −S ⎜⎝ ⎟ =V
∂T ⎠ P ∂P ⎠ T

⎛ ∂2 G ⎞ ⎛ ∂S ⎞ CP ⎛ ∂2 G ⎞ ⎛ ∂V ⎞
⎜ 2⎟ = − ⎜⎝ ⎟⎠ = − ⎜ 2⎟ = − ⎜⎝ ⎟⎠ = −V κT
⎝ ∂T ⎠ P ∂T P T ⎝ ∂P ⎠ T ∂P T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 321

Then

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 322

In this case, the first derivatives are discontinuous at the transition. This is not
always the case.

Ehrenfest proposed a classification scheme that refers to tranisions of order n,

where the nth derivatives of G with respect to T and P are the lowest derivatives
that are discontinuous at the phase transition.

The actual situation is more complex. The types of behavior actually observed
for CP are shown below.

First Order:

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 323

Second order:

Third Order:

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 324

Examples:

First Order:
Solid/Liquid
Vapor/Liquid
Solid/Vapor
Solid/Solid (Crystal Structure Changes)

Second Order:
Superconducting Transition

Lamda:
Helium I/Helium II (Superfluid transition in liquid 4He)
Order/disorder in b-brass
Order/disorder in quartz

Third Order:
Curie points of ferromagnets

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 325

3.5 BEHAVIOR NEAR CRITICAL POINTS

The term critical region refers to the region in state variable space where a fluid
goes from conditions where multiple phases exist to conditions where only a
single phase exists. At the transition point, all properties of the individual phases
become identical.

Approximate range: 0.99 < Tr < 1.01

0.60 < rr < 1.40

Types of critical points involved with fluid phase behavior:

Vapor - Liquid
Liquid - Liquid
Vapor - Liquid - Liquid (Tricritical point)

The critical region was first observed by Andrews (1869) for carbon dioxide.

Note the difference between the critical point and an azeotrope:

Phases become identical (all properties are the same) at a critical point.
Phases have the same composition but different densities at an azeotrope.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 326

3.5.1 P-V-T near a vapor - liquid critical point:

⎛ ∂2P ⎞
⎜ 2⎟ < 0
⎝ ∂V ⎠ T
• ⎛ ∂P ⎞
P ⎛ ∂2P ⎞
⎜ 2⎟ = 0
⎜ ⎟ <0
⎝ ∂V ⎠ T
⎝ ∂V ⎠ T
•
•
B (T > Tc)
•
⎛ ∂2P ⎞
⎜ 2⎟ < 0 ⎛ ∂P ⎞
⎝ ∂V ⎠ T ⎜ ⎟ =0 A (T < Tc )
⎝ ∂V ⎠ T

V
Equation of state isotherms are shown. Mathematically, to pass smoothly from
behavior A (two phases can be present for some conditions) to behavior B (one
phase only for all pressures) requires that at some temperature between TA and
TB:
⎛ ∂P ⎞ ⎛ ∂2 P ⎞
⎜⎝ ⎟⎠ = 0 and ⎜ 2 ⎟ = 0
∂V T ⎝ ∂V ⎠ T
1 ⎛ ∂V ⎞
Note also that: KT = − ⎜ ⎟ → ∞ at the critical point
V ⎝ ∂P ⎠ T
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 327
3.5.2 Physical Observations:

Behavior of the meniscus (interface between phases) for a pure fluid:

ρ > ρc

Increasing Temperature

ρ < ρc

Increasing Temperature

ρ = ρc

Increasing Temperature
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 328
3.5.3 Critical Opalescence

Density fluctuations (large) in the fluid phases lead to significant light scattering.
The scattering causes the fluid to turn black, to turn milky, or to glow depending
upon the fluid and the pattern of illumination.

3.5.4 Universality

Long range order and correlation occurs in the critical region. This order extends
over many molecules, hence the specific nature of the intermolecular force
becomes unimportant. Therefore the general behavior of all fluids will be the
same.

Other transitions, such as the Curie point for ferromagnetism also involve the
same types of long range order, therefore the critical behavior of magnets is
similar to that of fluids.

Theoretical calculations are the most tractable for the Ising model for
ferromagnets; there is an analytic solution for the 2 dimensional case, but a
numerical solution is required for 3 dimensions.

The renormalized group theory treatment introduced by Kenneth Wilson, et al., is

the most current theoretical model.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 329

3.5.5 General Behavior
Supercritical Fluid
CP
•
T Vapor Liquid
Liquid + Vapor
Two Phase Region

ρ
Heat Capacity at Constant Volume:
−α +
Cv ~ T − Tc
−α −2
Cv 2 ~ T − Tc CP
•
T

Critical
Isochore

ρ
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 330
Densities
+
−γ
δ κ T ~ T − Tc
P − Pc ~ V − Vc
CP
• Tc
T
κT = ∞ ρ = 21 (ρv + ρL )
• • •
ρ = ρc
β
Δρ ~ Tc − T
ρv ρL
ρ
3.5.6 Law of Rectilinear Diameters
The mean of the saturated vapor and liquid densities varies linearly with the
temperature excursion from the critical temperature, i. e.,

ρ= 1
2 ( ρv + ρL ) = ρc + k1(T −Tc )
This often provides the best way to determine experimental values for the critical
density. The temperature at which the meniscus disappears is measured by
visual observation, and measured vapor pressures are extrapolated to the critical
temperature to get the critical pressure.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 331

3.5.7 Critical Exponents

−α
CV ~ T −Tc

β
Δρ ~ T −Tc

−γ
κ T ~ T −Tc

δ
P − Pc ~ V −Vc

d 2P σ −θ
~ T −Tc
dT 2

ε
σ ~ Tc −T

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 332

Critical Exponent Values

Classical 3-D Ising Renormalized Experimental

Exponent Property
Value Model Group Theory Values

a Cv 0 0.125 0.109 ~0.12

b r 1/2 0.312 0.325 0.32 - 0.38

g κT 1 1.250 1.241 ~1.25

e s 3/2 ~1.25

d P 3 5.05 4.82 4.5 - 5.0

d 2P σ
2
0 0.125 0.109 0.21
dT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 333

3.5.8 Experimental Problems Near Critical Point

Long time required to achieve equilibrium

Impurity effects: small amounts of impurities drastically alter the physical

properties, because the critical temperature and pressure change with
composition.
Gravity Effects: the very large values of kT in the critical region mean that the
variation of pressure with height due to hydrostatic head causes very large
density gradients within the system.
Moldover, Sengers, Gammon and Hocken (1979) determined that gravity effects
produced errors greater than 1 % in a cell of 1 cm height within the following
ranges of temperature and density:
PVT experiment: density is the measured variable
T -Tc
At rc: ΔT = = 1-Tr < 6 × 10-4
Tc

ρ -ρc
At Tc : Δρ = = 1− ρr < 0.08
ρc

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 334

Calorimetry: heat capacity is the measured variable

T -Tc
At ρc : ΔT = = 1-Tr < 3 × 10-4
Tc

ρ -ρc
At Tc : Δρ = = 1− ρr < 0.1
ρc

Water: PVT: |T-Tc| < 0.4 K CV: |T-Tc| < 0.2 K

Methane: PVT: |T-Tc| < 0.1 K CV: |T-Tc| < 0.05 K

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 335

 4 PROPERTIES OF MIXTURES
4.1 GENERAL FORMAL DESCRIPTIONS

Usually use T,P as the independent variables for mixture calculations. Then

(
M total = M total T ,P,n1,n2, ,nm )
where:
M total = total property of the mixture
n i = number of moles of specie i present
m = number of chemical species present in the mixture
Also:
M total = nM
where:
M = molar property of a mixture
m
n = ∑ n i = total number of moles of mixture
i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 336

Mathematical Note: A function is homogeneous of order h if f ( λ x ) = λ hf ( x )

At constant T,P,
M total is homogeneous of order 1 with respect to the variables {n i }.

Implications of homogeneity

M total = M total (n1,n2 ,,nm )

λM total = M total ( λn1, λn2 ,, λnm )

Note: T,P are not considered to be independent variables here because of the
restriction to constant T,P

Rewrite as: λnM = nM ( λn1, λn2 ,, λnm )

Differentiate with respect to l:

d ( λnM ) m ⎡ ∂(nM ) ⎤ ⎡ ∂( λn ) ⎤ m ⎡ ∂(nM ) ⎤

i
= nM = ∑ ⎢ ⎥⎢ ⎥ = ∑ ni ⎢ ⎥
dλ i =1⎣ ∂( λn i ) ⎦ ⎣ ∂λ ⎦ i =1 ⎣ ∂( λ n )
i ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 337

The previous result is true for any value of the parameter l. Choose l = 1 to
obtain

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤
⎢ ⎥ = ⎢ ∂n ⎥
⎣ ∂( λ n )
i ⎦T ,P,n
k ≠i
⎣ i ⎦T ,P,n k ≠i

Result:

⎡ ∂(nM ) ⎤
m m
nM = ∑ n i ⎢ ⎥ = ∑ ni M i
i =1 ⎣ ∂n i ⎦T ,P ,n i =1
k ≠i

Notation:

Mi = molar property of pure fluid i

Mi = partial molar property of specie i in the mixture

M= bulk molar property of the mixture

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 338

Observations:

lim
xi → 1 Mi =Mi
( x k ≠i → 0)

lim
xi → 0 Mi = M∞
i
( ∑ x k ≠i → 1)

where: M∞
i is the partial molar volume at infinite dilution

Note: The properties of the mixture are known if the partial molar properties of
all species are known.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 339

4.1.1 Gibbs-Duhem equation

Any mixture property depends upon the temperature, pressure, and the number
of moles of each chemical species present

M total = M total (T ,P,n1,n2 ,,nm )

Mathematics relation 1 gives

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ m ⎡ ∂(nM ) ⎤

d (nM ) = ⎢ ⎥ dT + ⎢ ⎥ dP + ∑ ⎢ ⎥ dn i
⎣ ∂T ⎦ P,n i ⎣ ∂P ⎦T ,n i i =1⎣ ∂n i ⎦T ,P,nk ≠i

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ m
=⎢ ⎥ dT + ⎢ ⎥ dP + ∑ M i dn i
⎣ ∂T ⎦ P,n i ⎣ ∂P ⎦T ,n i i =1

m m m
Also, nM = ∑ n i M i requires that d (nM ) = ∑ n i dM i + ∑ M i dn i
i =1 i =1 i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 340

Equate the two expressions for d(nM) to obtain

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ m m m
⎢ ∂T ⎥ dT + ⎢ ⎥ dP + ∑ M i dn i = ∑ n i dM i + ∑ M i dn i
⎣ ⎦ P,n i ⎣ ∂P ⎦T ,n i i =1 i =1 i =1

Eliminating common terms and rearranging produces

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ m
⎢ ∂T ⎥ dT + ⎢ ⎥ dP − ∑ n i dM i = 0
⎣ ⎦ P,n i ⎣ ∂P ⎦T ,n i i =1

For those partial derivatives where the number of moles of every species is held
constant, n also is constant, so the previous result may be written as

⎡ ∂M ⎤ ⎡ ∂M ⎤ m
n⎢ ⎥ dT + n ⎢ ⎥ dP − ∑ n i dM i = 0
⎣ ∂T ⎦ P,n i ⎣ ∂P ⎦T ,n i i =1

Divide through by n to put this equation on a molar basis.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 341

 ⎛ ∂M ⎞ ⎛ ∂M ⎞ m
⎜⎝ ⎟ dT + ⎜ dP − ∑ x i dM i = 0
∂T ⎠ P ,n i ⎝ ∂P ⎟⎠T ,n i =1
i

m
∑ x i dM i = 0 (constant T ,P )
i =1

This result is known as the Gibbs-Duhem Equation.

The Gibbs-Duhem equation provides a consistency test for thermodynamic

properties of mixtures. It often is used in phase equilibrium studies to calculate
the composition of a phase that is not amenable to direct measurement of
composition.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 342

4.1.2 Calculation of partial molar properties

The definition of a partial molar property is

⎛ ∂(nM ) ⎞
Mi = ⎜
⎝ ∂n i ⎟⎠ T ,P,n
k ≠i

However, in most cases the information about the mixture properties is given as
a function of composition (mole fraction)

M = M ( x 1,x 2 ,,x m )

An alternative method for calculating partial molar properties using partial

derivatives with respect to mole fraction can be derived as follows, beginning with
the definition of the partial molar property.

⎛ ∂(nM ) ⎞ ⎛ ∂M ⎞
Mi = ⎜ = M +n⎜
⎝ ∂n i ⎟⎠ T ,P,n ⎝ ∂n i ⎟⎠ T ,P,n
k ≠i k ≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 343

This derivative is related to partial derivatives with respect to mole fraction by

⎛ ∂M ⎞ m ⎛ ∂M ⎞ ⎛ ∂x j ⎞
⎜⎝ ∂n ⎟⎠ = ∑⎜ ⎟ ⎜⎝ ∂n ⎟⎠
i T ,P,n j =1⎝ ∂x j ⎠ i T ,P,n
k ≠i T ,P,x k≠j k ≠i

The mole fraction is related to the amount of each species by

nk nk
xk = m =
∑ nj n
j =1

Then

⎛ ∂x k ⎞ 1 ⎛ ∂nk ⎞ nk ⎛ ∂n ⎞
⎜⎝ ∂n ⎟⎠ = −
i n n ⎜⎝ ∂n i ⎟⎠ n n 2 ⎜⎝ ∂n i ⎟⎠ n
≠i ≠i ≠i

Because the {nI} are independent variables

⎛ ∂nk ⎞ ⎧0 i ≠k
⎜⎝ ∂n ⎟⎠ = δ ik = ⎨
i n ⎩1 i =k
≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 344

The second partial derivative on the right is
⎛ ∂n ⎞ ∂ ⎛m ⎞ m ⎛ ∂nk ⎞ m
⎜⎝ ∂n ⎟⎠ = ⎜ ∑ nk ⎟ = ∑ ⎜ = ∑ δ ij = 1
i n ∂n i ⎝ k =1 ⎠ k =1⎝ ∂n i ⎟⎠ n j =1
≠i ≠i

Then
⎛ ∂x k ⎞ 1 ⎛ ∂nk ⎞ nk ⎛ ∂n ⎞ δ ik nk
⎜⎝ ∂n ⎟⎠ = − = −
i n n ⎜⎝ ∂ni ⎟⎠ n 2 ⎜⎝ ∂ni ⎟⎠ n n2
≠i

so that
⎛ ∂M ⎞ m⎛ ∂M ⎞ ⎛ ∂x k ⎞
⎜⎝ ∂n ⎟⎠ = ∑⎜ ⎟ ⎜⎝ ∂n ⎟⎠
i T ,P,n k =1⎝ ∂x k ⎠ T ,P,x i T ,P,n
k ≠i j ≠k j ≠i

m⎛ ∂M ⎞ ⎡ δ ik nk ⎤
= ∑⎜ ⎟ ⎢⎣ n − n 2 ⎥⎦
k =1⎝ ∂x k ⎠ T ,P,x
j ≠k

1 ⎛ ∂M ⎞ ⎛ ∂M ⎞
m nk
= ⎜ − ∑⎜
n ⎝ ∂x i ⎟⎠ T ,P,x ⎟
k =1⎝ ∂x k ⎠ T ,P,x n2
j ≠i j ≠k

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 345

and

⎛ ∂M ⎞ ⎛ ∂M ⎞ m ⎛ ∂M ⎞
n⎜ =⎜ − ∑ xk ⎜
⎝ ∂n i ⎟⎠ T ,P,x ⎝ ∂x i ⎟⎠ T ,P,x k =1 ⎝ ∂x k ⎟⎠ T ,P,x
j ≠i j ≠i j ≠k

Substitution into the original equation produces

⎛ ∂(nM ) ⎞ ⎛ ∂M ⎞ m ⎛ ∂M ⎞
Mi = ⎜ =M +⎜ − ∑ xk ⎜
⎝ ∂n i ⎟⎠T ,P ,n ⎝ ∂x i ⎟⎠T ,P ,x k =1 ⎝ ∂x k ⎟⎠T ,P ,x
k ≠i k ≠i ≠k
l

Other methods for calculating M i exist, but in most cases they involve algebraic
elimination of xi and summations from which one or more terms are excluded.
The result given above is much more convenient for computer calculations for
mixtures with many species present. The notation here also is simpler and less
likely to cause confusion.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 346

4.1.3 Property change on mixing

4.1.3.1 General Definition

The concept of a property change on mixing is compares the properties of a
mixture to the molar average of the properties of the pure components at the
same temperature and pressure.
m
ΔmM (T ,P,{x i }) ≡ M (T ,P,{x i }) − ∑ x i M i (T ,P )
i =1

( )
m m m
= ∑ x i M i (T ,P,{x i }) − ∑ x i M i (T ,P ) = ∑ x i M i − M i
i =1 i =1 i =1
where:
Δm M (T ,P,{x i }) = property change on mixing

M (T ,P,{x i }) = molar property of the mixture at T, P and {xi}

M i (T ,P ) = molar property of pure substance i at T, P
M i (T ,P,{x i }) = partial molar property of substance i in the mixture at T, P
and {xi}

The property changes of interest include DmV, DmU, DmH, DmA, DmG and DmS
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 347
4.1.3.2 Property Changes on Mixing for Ideal Gases

The property changes on mixing for the ideal gases will provide reference points
for more general discussions of mixture properties.

Begin with three separate ideal gases, each at T, P. Then mix, with no energy
transferred as heat during the mixing process. The process is shown
schematically below.

Before mixing, at T, P:

PVAtotal PVBtotal PVCtotal

nA = nB = nC =
RT RT RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 348

After mixing, at T, P:

n = nA + nB + nC

nRT RT RT ⎡ PVAtotal PVBtotal PVCtotal ⎤

total
Vmixture = = (nA + nB + nC ) = ⎢ + + ⎥
P P P ⎣⎢ RT RT RT ⎥⎦

=VAtotal +VBtotal +VCtotal

In terms of molar volumes

total
Vmixture = nV = nAVA + nBVB + nCVC

Divide through by the number of moles

nA n n
V = VA + B VB + C VC = x AVA + x BVB + x CVC = ∑ x iVi
n n n

Then

ΔigmV =V ig − ∑ x iViig = 0

Therefore the volume change on mixing for ideal gases is zero.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 349
Note also that no PdV work is done during the mixing process because the
volume remains constant.

The change in total internal energy resulting from the mixing process is
calculated from the First Law

Δig
mU
total
= Q +W = 0

Therefore

Δig
mU = 0

The enthalpy of mixing for ideal gases follows directly from the internal energy
and volume.

Δig
m H = Δ ig
m U + PΔ ig
mV = 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 350

Because the particles in an ideal gas do not interact with each other, we may
calculate the entropy change for each component as though it expanded from its
original total volume to the total volume of the mixture.

⎛ nRT ⎞
total ⎜⎝ ⎟
Vmixture P ⎠ n 1
Note: = = =
Vitotal ⎛ ni RT ⎞ ni x i
⎜⎝ ⎟
P ⎠
Then
total
Vmixture 1
Δig
mSi
total
= n i R ln total = n i R ln = −n i R lnx i
Vi xi
and
mS = −R (nA lnx A + nB lnx B + nC lnx C )
nΔig

Divide through by n to obtain

Δ igmS = −R ( x A lnx A + x B lnx B + x C lnx C ) = −R ∑ x i lnx i

Δig
mS
m
so that the dimensionless form is = − ∑ x i lnx i
R i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 351

Use the definitions and the previous results to calculate the values for the
Helmholtz and Gibbs energies:

Δig
mA
=
( Δ U ) −T ( Δ S )
ig
m
= ∑ x lnx
ig
m m
i i
RT RT i =1

Δig
m G
=
Δig
m (
H −T Δig
mS m ) (
= ∑ x i lnx i
)
RT RT i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 352

4.1.3.3 Ideal solutions

We use the results derived for mixing of ideal gases to define a concept of ideal
solutions, which are mixtures for which the properties changes of mixing are
identical to those for mixing of ideal gases, even though the pure substances are
not ideal gases.

Δis
mU = 0 Δis m
m S = − x ln x
∑ i i
RT R i =1

P Δis
mV = 0 Δis
mA =
m
∑ x i ln x i
RT RT i =1

Δis
mH = 0 Δis
mG =
m
∑ x i ln x i
RT RT i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 353

4.1.3.3.1 Alternative Definition of Ideal Solution

The Gibbs energy of an ideal gas mixture is

m
G ig
= Δig
mG + ∑ x i Giig
i =1

The Gibbs energy change on mixing for an ideal gas is

Δis
mG = RT ∑ x i lnx i

Combining the previous expressions gives

G ig = ∑ x i Giig + RT ∑ x i lnx i = ∑ x i ⎡Giig + RT lnx i ⎤

⎣ ⎦
and

nG ig = ∑ ni ⎡Giig + RT lnx i ⎤ = ∑ nk Gkig + ∑ nk RT lnx k

⎣ ⎦
so that

Giig =⎢
(
⎡ ∂ nG ig ) ⎤⎥ = Giig
m
+ RT lnx i + RT ∑ nk
∂
( lnx k )
⎢ ∂ni ⎥ k =1 ∂ni
⎣ ⎦T ,P,nk ≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 354

The partial derivative in the last term is
∂(lnx k ) 1 ∂x k 1 ∂ ⎛ nk ⎞ 1 ⎡ δ ik nk ⎤ 1
∂ni
= = ⎜ ⎟ = − =
x k ∂ni x k ∂ni ⎝ n ⎠ x k ⎢⎣ n n 2 ⎥⎦ nx k
(δ ik − x k )
The summation in the last term is

m ∂(lnx k ) m nk m m
∑ nk
∂n
= ∑
nx
(δ ik − x k ) = ∑ δ ik − ∑ x k
k =1 i k =1 k k =1 k =1

m ⎫
∑ δ ik = 1⎪
k =1 ⎪ m ∂(lnx k )
⎬ ⇒ ∑ nk =0
m ∂n
∑ x k = 1⎪
k =1 i
k =1 ⎪⎭

so that

Giig = Giig + RT lnx i

Use this form as the basis for an alternative definition of the ideal solution

G isi = G i + RT ln x i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 355

4.1.3.4 Activity Coefficient

Preserve the same general form for real mixtures by including the activity
coefficient in the partial molar volume.

G i = G i + RT lnγ i x i

where gI is the activity coefficient.

4.1.3.5 Activity

The partial molar Gibbs energy also can be written in the form

G i = G i + RT lnα i

where aI is the activity.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 356

4.1.3.6 General Definitions

In the general case, the definition of the property change on mixing must
accommodate reference states that differ from the final state of the mixture:

( )
m m
Δ m† M = M − ∑ x i M †i = ∑ x i M i − M †i
i =1 i =1

Usually the reference state (denoted by †) is chosen to be of the same T, P and

in the same state as the mixture, but this is not always possible.
A more general definition of the ideal solution then follows as

G isi = G †i + RT ln x i

and

G i = G †i + RT lnγ i†x i = G †i + RT lnα i†

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 357

4.1.3.7 Excess Properties

Define excess properties based upon the difference between the property
changes on mixing for the real solution and for the ideal solution.

M E (T ,P,{x i }) = ΔmM (T ,P,{x i }) − ΔismM (T ,P,{x i })

Note that both the excess properties and the property changes on mixing are
functions of temperature, pressure and composition.

The partial molar excess properties are related to the real mixture and ideal
solution values by

M Ei = M i − M is
i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 358

For the Gibbs energy

G i = G †i + RT lnγ i†x i

G isi = G †i + RT lnx i

( )
†
GEi = G i − G isi = RT lnγ i†

Therefore

G Ei
= lnγ i
RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 359

Note that all information about the real mixture is available if (GE/RT) is known,
because all other excess properties and partial molar excess properties can be
obtained from the excess Gibbs function using the following relations

E
⎛ ∂GE ⎞ ⎛ ∂GEi ⎞
S = −⎜ ⎟ S Ei = −⎜ ⎟
⎝ ∂T ⎠ P ⎝ ∂T ⎠ P

E
⎛ ∂G E ⎞ ⎛ ∂GEi ⎞
V =⎜ ⎟ V Ei =⎜ ⎟
⎝ ∂P ⎠ T ⎝ ∂P ⎠ T

H E = GE +TS E H Ei = G Ei +TS Ei

AE = GE − PV E AEi = G Ei − PV iE

U E = GE − PV E +TS E U Ei = G Ei − PV iE +TS Ei

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 360

In terms of the activity coefficients

GE m
= ∑ x i lnγ i
RT i =1

PV E m ⎛ ∂lnγ i ⎞
= ∑ xi ⎜ ⎟
RT i =1 ⎝ ∂lnP ⎠ T

HE m ⎛ ∂lnγ i ⎞
= − ∑ xi ⎜ ⎟
RT i =1 ⎝ ∂lnT ⎠ P

SE m ⎡ ⎛ ∂lnγ i ⎞ ⎤
= − ∑ xi ⎢lnγ i + ⎜⎝ ⎟⎠ ⎥
R i =1 ⎣ ∂lnT P⎦

AE m ⎡ ⎛ ∂lnγ i ⎞ ⎤
= ∑ xi ⎢lnγ i + ⎜⎝ ⎟⎠ ⎥
RT i =1 ⎣ ∂lnT P⎦

UE m ⎡⎛ ∂lnγ i ⎞ ⎛ ∂lnγ i ⎞ ⎤
= − ∑ xi ⎢⎜⎝ ⎟⎠ + ⎜⎝ ⎟⎠ ⎥
RT i =1 ⎣ ∂lnT P ∂lnP T⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 361

4.1.3.7.1 Relation of Excess Properties to Residual Functions and Equation
of State
For the case where all reference states are taken as the same phase at the same
temperature and pressure, the Gibbs energy change on mixing is

ΔmG G − Σx i Gi Σx i G i − Gi
= =
( )
RT RT RT

=
1
RT { (
Σx i ⎡ G i − G ig
⎣ i + G ig
i ) (
− Giig + G ig
i − Gi ⎤
⎦) ( )}

=
1
RT { (
Σx i ⎡ G i − G ig
⎣ i + RT lnx i + G ig
i )− Gi ⎤
⎦ ( )}
⎡ GRi GR ⎤ Δis
mG
⎡ GRi GR ⎤ Δis
mG G
E
= Σx i ⎢ − i
⎥ + Σx i lnx i = + Σx i ⎢ − i
⎥= +
⎢⎣ RT RT ⎥⎦ RT ⎢⎣ RT RT ⎥⎦ RT RT

Therefore:

GE ⎡ GRi GiR ⎤
= Σx i ⎢ − ⎥
RT ⎢⎣ RT RT ⎥⎦
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 362
By the definition of a residual function:

Pure species:

P P
GiR Gi − Giig ⌠VR ⌠ dP
= = ⎮ i
⎮ RT
dP = ⎮⎮ ( Z i − 1)
RT RT ⌡ ⌡ P
0 0

Mixture:

P P
R ig ⌠VR
G G −G ⌠ dP
= = ⎮
⎮ RT
dP = ⎮⎮ ( Z − 1)
RT RT ⌡ ⌡ P
0 0

Partial Molar Gibbs function:

P P
GiR Gi − Giig ⌠VR
( Z i − 1) dP
⌠
= = ⎮ i
⎮ RT
dP = ⎮
⎮
RT RT ⌡ ⌡ P
0 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 363

Then

⎡P R P ⎤ P
G E ⎡ GRi
⎤ m ⎢⌠ V i GR ⌠V R
⎥ m ⌠ V R −V R
= Σx i ⎢ − ⎥ = ∑ x i ⎢ ⎮⎮
i
dP − ⎮⎮ i
dP = ∑ x i ⎮⎮ i i
dP
RT ⎢⎣ RT RT ⎥⎦ i =1 ⌡ RT ⌡ RT ⎥ i =1 ⌡ RT
⎢⎣ 0 0 ⎥⎦ 0
P
m ⌠ V −V
= ∑ x i ⎮⎮ i i
dP
i =1 ⌡ RT
0

Also

P
GEi ⌠ V −V
= ⎮ i
⎮
i
dP = lnγ i
RT ⌡ RT
0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 364

4.1.3.7.2 Fugacity

For an ideal gas at constant temperature

dP
dG ig = V ig dP = RT = RT (d lnP )
P
Integration from P† to P yields
G ig (P ) − G ig P † ( ) = ln P
RT P†

Define fugacity to obtain the same form for the real fluid as for the ideal gas
dG = RTd (lnf )

Then
G (P ) − G P † ( ) = ln⎛ f ⎞
⎜⎝ † ⎟⎠
RT f
Using these forms

⎛ f ⎞
dGR = dG − dG ig = RTd (lnf ) − RTd (lnP ) = RTd ⎜ ln ⎟
⎝ P⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 365

and

⎡ R †
GR ⎢ G P
=
⎛ f† ⎞⎤( ) ⎛f ⎞
− ln ⎜ † ⎟ ⎥ + ln ⎜ ⎟
RT ⎢ RT ⎝P ⎠⎥ ⎝P⎠
⎣ ⎦

In the limit as the pressure goes to zero

⎛ R ⎞
⎛f ⎞
lim ⎜ G ⎟ = ⎟ =0
lim ln ⎜
P→0 ⎜ RT ⎟ P→0 ⎝P ⎠
⎝ ⎠

Define fugacity coefficient as

f
φ= where lim
P →0
φ =1
P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 366

Then, use the following forms.

Pure Substance:

lim ⎛ fi ⎞
dGi = RTd lnfi P→0 ⎜ =1
⎝ P ⎟⎠

Mixture:

lim ⎛ f ⎞
dG = RTd lnf P→0 ⎜ =1
⎝ P ⎟⎠

Component in the mixture:

lim ⎛ fˆi ⎞
dGi = RTd lnfˆi P→0 ⎜ ⎟ =1
⎝ xiP ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 367

Relations between activities, fugacities, activity coefficients and fugacity
coefficients

fˆi x i φ̂i
α i† = =
f i† φi†

fˆi φ̂i
γ i† = =
x i f i† φ i†

fˆiis = x i f †i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 368

 0 1
x

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 369

4.1.3.7.3 Relationships between activity coefficients based upon different
reference states.

Consider activity cases based upon two different reference states, one denoted
by † and the other by ∆. The Gibbs energy change on mixing as described by
each choice of reference states is

†
Δm G
=
(
∑ x i Gi − Gi† ) Δ
Δm G
=
(
∑ x i Gi − GiΔ )
RT RT RT RT

So

†
Δm G ΔmΔ
G x i ⎡⎣GiΔ − Gi† ⎤⎦
− =∑
RT RT RT

Note:
†
Δm G Δis,†
m G G
E ,†
G E ,† m
= + = + ∑ x i lnx i
RT RT RT RT i =1
Δ
Δm G Δis,Δ
m G G
E ,Δ
G E ,Δ m
= + = + ∑ x i lnx i
RT RT RT RT i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 370

Finally

GE .† GE ,Δ Δm†
G ΔmΔ
G m ⎡ GiΔ − Gi† ⎤
− = − = ∑ xi ⎢ ⎥
RT RT RT RT i =1 ⎢⎣ RT ⎥⎦

also

G E .† G E ,Δ m
†
m
Δ
m γ i†
− = ∑ x i lnγ i − ∑ x i lnγ i = ∑ x i ln Δ
RT RT i =1 i =1 i =1 γi

Then

γ i† GiΔ − Gi† Δ
⎡ GiΔ − Gi† ⎤
ln = ⇒ γ i† = γ i exp ⎢ ⎥
γ iΔ RT ⎢⎣ RT ⎥⎦

Δ
⎡ GiΔ − Gi† ⎤
αi = γ i xi ⇒ α i† = α i exp ⎢ ⎥
⎢⎣ RT ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 371

These results are consistent with the definitions

fˆi fˆi
γ i† = γ iΔ =
x i f i† x i f iΔ

fˆi fˆi
α i† = α iΔ =
f i† f iΔ

where: fˆi = fugacity of component i in the mixture

f i† = fugacity of pure component i at the reference state denoted by †.

f iΔ = fugacity of pure component i at the reference state denoted by ∆.

Note that analogous developments could be based upon molal concentrations

(often used for electrochemical applications) or mass fractions (often used for
polymer solutions or mixtures of undetermined molecular weight) instead of mole
fraction.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 372

4.2 PHASE EQUILIBRIUM FOR MIXTURES

4.2.1 General Phase Equilibrium

Equilibrium condition is that the total Gibbs enery of the system is at a minimum
for a given T, P. Consider G(x) at fixed T, P.

Δis
mG
m
Example: = ∑ x i lnx i
RT i =1

ΔismG
For ideal solutions,such mixtures, is negative for all compositions.
RT

The behavior as a function of composition for a binary mixture is shown on the

following page.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 373

Ideal Solution:

0 1
x

The chord connecting the values for any two compositions gives the total Gibbs
energy for a two-phase mixture, where the two phases have the compositions of
the end points of the chord.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 374

Binary Mixtures:

Totally miscible:

0 1
x

At all overall compositions, ∆mG < 0, and G(x) has a lower value than any linear
combination of other compositions, therefore such a mixture always exists as a
single phase at this T, P.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 375

Totally immiscible:

0 1
x

At all overall compositions, ∆mG > 0, therefore

m
∑ x i Gi < G ( x )
i =1

The system will exist as two separate pure phases.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 376
Partially immiscible (or partially miscible):

0 1
x

For xA< x < xB, the tangent AB has a lower Gibbs energy than G(x). This system
will exist as a single phase for x < xA and x > xB, but as a two-phase mixture with
phase compositions xA and xB for xA< x < xB.

⎫ ΔmG < 0
⎪
Mixtures are completely miscible if: ∂2 Δ G ⎬ for all 0 ≤ x ≤ 1
m
2
> 0⎪
∂x ⎭
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 377
4.2.2 Phase Equilibrium Calculation

Phase equilibrium calculations may be done by direct minimization of the Gibbs

energy for a given T, P, but a more common method is to use the equilibrium
relationship between the phases and the conditions imposed by material balance
requirements.

4.2.2.1 Partition Coefficient

The partition coefficient describes the ratio of compositions in phases that

coexist at equilibrium

x iβ
(
K i ≡ α = K i T ,P ,{x iα },{x iβ }
xi
)

The equilibrium conditions resulting from the First and Second Laws of
Thermodynamics provide methods for calculating Ki for mixtures.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 378

4.2.2.2 Saturation Boundary Working Equation

A condition that often is of practical interest is to determine the boundary

between the single- and multiple-phase equilibrium conditions.

Begin with the requirement that the mole fractions in a phase must sum to unity
m
∑ xi = 1
i =1

The composition of the saturated phase is known (the specified overall

composition), and the composition of the incipient phase is unknown.

Begin with the expression for the incipient phase

m m
β
∑ x iα = 1 or ∑ xi = 1
i =1 i =1

Use the equilibrium relation in the form of the partition coefficient to eliminate the
mole fractions of the unknown phase.

x iβ m
β
m
Ki = ⇒ ∑ xi = ∑ K i x iα = 1
x iα i =1 i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 379

Determine the saturation condition by finding the T or P (whichever is not
specified originally) that satisfies the working equation

m ⎛ β⎞ m
x
∑ K i x iα = 1 or ∑ ⎜ i ⎟ = 1
i =1 i =1⎝ K i ⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 380

4.2.2.3 Phase Split Working Equation

Use the material balances to describe the state when two coexisting phases are
present. The overall composition is known but the compositions of the coexisting
phases are unknown.

Total material balance: ψ α +ψ β = 1

Species material balance: x iαψ α + x iβψ β = x i

nj
j
where: ψ =
n

x iα is the mole fraction of specie i in phase a

x i is the overall mole fraction of specie i in the mixture

Use the partition coefficient and the total material balance to eliminate phase b
from the species material balance

( )(
x iαψ α + K i x iα 1− ψ α = x i )
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 381
Rearrange to obtain

xi
x iα =
1+ ψ α (K i − 1)

Working equations:

m ⎡ xi ⎤
∑⎢ ⎥ =1
i =1 1+ ψ (K i − 1)
α
⎣ ⎦

m ⎡ K i xi ⎤
∑⎢ ⎥ =1
i =1 1+ ψ (K i − 1)
α
⎣ ⎦

m ⎡ (K i − 1) x i ⎤
∑⎢ ⎥=0
i =1 1+ ψ (K i − 1)
α
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 382

Solve the working equation for y, then compute the phase mole fractions using

xi
x iα =
1+ ψ α (K i − 1)

and

x iβ = K i x αi

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 383

4.2.3 Equilibrium Relationships

The stability criteria include

µiα = µiβ ⇔ Gαi = G βi ⇔ fˆαi = fˆ iβ

Alternative descriptions of the fugacity of a substance in the mixture

Fugacity coefficient: fˆi = x i φ̂i P

Activity coefficient: fˆi = γ i x i f i†

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 384

Computational Approaches

Gamma-Phi (G-F)

Use the fugacity coefficient description for one phase (fluid) and the activity
coefficient description for the other (solid or liquid):

α α β β † x iα γ iβf i† γ iβPisat( β )
x φ̂ P = x γ f ⇒ β = α =
φ̂i P Φi P
i i i i i
xi

where
φ̂iα Pi sat( β ) φ̂iα Pi sat( β ) φ̂iα ⎡ Vi β ⎤
Φi = =
⎡ Vi β sat( β ) ⎤
=
φisat( β )
exp (
⎢ RT i P sat( β )
− )
P ⎥
fi † sat( β ) sat( β )
φi Pi exp ⎢
⎣ RT
(P − Pi ) ⎥
⎦
⎣ ⎦

Equation of State

Use the fugacity coefficient description for both phases:

x iα φ̂iβ
x iα φ̂iα P = x iβφ̂iβ P ⇒ =
x iβ φ̂iα
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 385
4.2.3.1 Vapor Phase Description

For the vapor phase, the most common representation is the fugacity coefficient
notation

fˆiv = y i φ̂vi P

( ) dP
⌠
where lnφ̂vi = ⎮⎮ Z i − 1
⌡ P
0

If
BP
Z = 1+
RT

provides an accurate description of Z (usually true for low pressures or high

temperatures), then
P
⌠ Bi B P
lnφ̂vi ≅ ⎮
⎮
dP = i
⌡ RT RT
0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 386

4.2.3.2 Liquid/Solid Phase Description

For liquids and solids, the activity coefficient notation often is used. For liquids,

⎡ 1 P ⎤
fˆiL = γ iL x i f iL = γ iL x i φiv Pi v exp ⎢ ⎮ V dP ⎥
⌠
⌡ i
L
⎢ RT Pi v ⎥
⎣ ⎦

For solids

fˆis = γ is x i f is

where

⎧ ⎡ P ⎤
⎪φ sv P sv exp ⎢ 1 ⎮ V dP ⎥
⌠ s
T ≤ Titp
⎪ i i ⎢ RT
⌡ i
⎥
⎪ ⎣ Pi sv ⎦
⎪
f is = ⎨
⎪
⎪ v v ⎡ ⎧⎪Pi s P ⎫⎪ ⎤
⎢ 1
⎨ ⌠⎮⌡ Vi dP + ⌠⎮⌡ Vi dP ⎬ ⎥ T > Titp
L s
φ
⎪ i i P exp
⎪⎩ ⎢ RT ⎪⎩Pi v Pi s ⎪⎭ ⎥⎦
⎣

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 387

4.2.3.3 VLE Equilibrium Descriptions

4.2.3.3.1 General formulation

The general relationship for representing vapor-liquid equilibrium (VLE) using the
gamma-phi (G-F) notation is

y i γ iLφiσ Piσ ⎡ 1 P ⎤
Ki ≡ = v
exp ⎢ ⎮ V dP ⎥
⌠
⌡ i
L
xi φ̂i P ⎢ RT Piσ ⎥
⎣ ⎦

where

⎡Piσ ⎤
⎢ ⌠ dP ⎥
φiσ = exp ⎢ ⎮⎮ ( Z i − 1)
⌡ P ⎥
⎢⎣ 0 ⎥⎦

⎡P ⎤
(dP
)
⌠
φ̂iv = exp ⎢ ⎮⎮ Z i − 1 ⎥
⎢⌡ P ⎥
⎣0 ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 388

Common Assumptions

⎡ 1 P ⎤
1. exp ⎢ ⌠
⎮
⌡ Vi dP ⎥ ≅ 1
L
⎢ RT Piσ ⎥
⎣ ⎦

φiσ
2. v
≅1
φ̂i

3. γ iL = 1 (ideal solution)

⎡ 1 P ⎤
4. γ iLφiσ Piσ exp ⎢ ⌠
⎮
⌡ Vi dP ⎥ = constant
L
⎢ RT Piσ ⎥
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 389

4.2.3.3.2 Raoult's law

If assumptions 1, 2 and 3 are valid,

y i γ iLφiσ Piσ ⎡ 1 P ⎤ Pσ
Ki ≡ = v
exp ⎢ ⎮ V dP ⎥ = i
⌠
⌡ i
L
xi φ̂i P ⎢ RT Piσ ⎥ P
⎣ ⎦
Note also that

γ iLφiσ ⎡ 1 P ⎤
v
exp ⎢ ⎮ V dP ⎥ = 1 → Raoult's Law
⌠
⌡ i
L
φ̂i ⎢ RT Piσ ⎥
⎣ ⎦
(This condition does not require that 1, 2 and 3 above be satisfied individually.)

Raoult's Law is

y i P iσ
fˆ (RL ) = x i P
i
L σ
i and K i
RL
= =
xi P

This relation often provides a good approximation for estimating phase behavior.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 390

4.2.3.3.3 Henry's law

If the fugacity is linear with composition, then fˆi = Η i x i

This result is known as Henry's law, and it provides a good representation of the
fugacity of a species at nearly infinite dilution.

For vapor-liquid equilibrium calculations, it applies when assumption 4 pertains,

so that
yi ΗLi
Ki = = v
x i φ̂ i P
⎡ 1 P ⎤
where ΗLi = γ iLφiσ Piσ exp ⎢ ⌠
⎮
⌡ Vi dP ⎥
L
⎢ RT Piσ ⎥
⎣ ⎦

or, in the general case,

lim ⎛ fˆi ⎞
Ηi =
x i → 0 ⎜⎝ x i ⎟⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 391

4.2.3.4 Excess Gibbs Models for Liquid Phase
Normally, the parameters for these models are determined from experimental
data. There are methods available (UNIFAC [UNIquac group-Functional Activity
Coefficient], ASOG [Analytical Solution of Groups], etc.) for predicting model
parameters based on contributions to the excess Gibbs energy from various
chemical groups.

4.2.3.4.1 Binary Mixtures

4.2.3.4.1.1 Redlich-Kister Expansion

The Redlich-Kister expansion has the form

GE
= x 1x 2 ⎡ A + B ( x 1 − x 2 ) + C ( x 1 − x 2 ) + D ( x 1 − x 2 ) + ⎤
2 3
RT ⎣ ⎦

Special Cases
A=B=C=D=0 ® ideal solution
B=C=D=0 ® symmetric
C=D=0 ® two-suffix Margules
D=0 ® three-suffix Margules

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 392

4.2.3.4.1.2 Van Laar

This form was derived from the van der Waals equation:

GE x 1x 2
=
RT 1 1
x1 + x
A12 A21 2

4.2.3.4.1.3 Wilson

This expression is not capable of describing liquid-liquid equilibrium behavior

GE
= −x 1 ln ( x 1 + Λ12 x 2 ) − x 2 ln ( Λ21x 1 + x 2 )
RT

4.2.3.4.1.4 T-K Wilson

This modification of the Wilson equation is capable of describing liquid-liquid

equilibrium.

GE ⎡ x +V x / V ⎤ ⎡V x / V + x 2 ⎤
= x 1 ln ⎢ 1 2 2 1 ⎥ + x 2 ln ⎢ 1 1 2 ⎥
RT ⎣ x 1 + Λ12 x 2 ⎦ ⎣ Λ 21x 1 + x 2 ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 393

4.2.3.4.1.5 NRTL

NRTL is the acronym derived from Non-Random-Two-Liquid theory.

GE ⎡ τ 21G21 τ 12G12 ⎤
= x 1x 2 ⎢ + ⎥
RT ⎣ x 1 + G21x 2 G12 x 1 + x 2 ⎦

4.2.3.4.1.6 UNIQUAC

UNIQUAC is the acronym derived from UNIversal QUAsi-Chemical Activity

Coefficient model.

GE ⎡ φ qz θ ⎤
= x 1 ⎢ln 1 + 1 ln 1 − q1ln (θ1 + θ 2τ 21) ⎥
RT ⎣ x 1 2 φ1 ⎦

⎡ φ q z θ ⎤
+x 2 ⎢ln 2 + 2 ln 2 − q 2 ln (θ1τ 12 + θ 2 ) ⎥
⎣ x2 2 φ2 ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 394

4.2.3.4.1.7 Scatchard-Hildebrand

Note that the form is the same as van Laar, but the parameters, di, are
determined for pure substances, allowing for prediction of mixture behavior

GE
=
(δ1 − δ 2 )2
RT ⎛ 1 1 ⎞
RT ⎜ +
⎝ V1x 1 V2 x 2 ⎟⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 395

4.2.3.4.2 Multicomponent Mixtures

GE ⎛ ⎞
Wilson = − ∑ x i ln ⎜ ∑ x k Λik ⎟
RT i ⎝k ⎠

GE ⎡ ⎛ ⎞⎤
T-K Wilson = ∑ x i ⎢ ∑ x kVk / Vi − ln ⎜ ∑ x k Λik ⎟ ⎥
RT i ⎣k ⎝k ⎠⎦

GE ⎡ ⎤
NRTL = ∑ x i ⎢ ∑ τ ji G ji x j / ∑ Gki x k ⎥
RT i ⎣ j k ⎦

GE φ z θ ⎛ ⎞
UNIQUAC = ∑ x i ln i + ∑ qi x i ln i − ∑ qi x i ln ⎜ ∑ θk τ ki ⎟
RT i xi 2 i φi i ⎝k ⎠

GE
Scatchard-Hildebrand =
1
RT RT i
(
∑ x iVi δ i − δ ) 2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 396

4.2.3.5 Calculation of fugacities from equations of state

A difficulty here is that our previous expressions relating the fugacity to the
equation of state involve pressure as the variable of integration. This is feasible
for the vapor phase, but, as seen with the pure fluid vapor pressures, it is not
feasible for the liquid phase. An expression utilizing the volume as the variable
of integration must be derived.

Begin with

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ m
d (nM ) = ⎢ ⎥ dT + ⎢ ⎥ dP + ∑ M kdnk
⎣ ∂T ⎦P,nk ⎣ ∂P ⎦T ,nk k =1

Divide by dni

d (nM ) ⎡ ∂(nM ) ⎤ dT ⎡ ∂(nM ) ⎤ dP m dnk

=⎢ ⎥ + ⎢ ⎥ + ∑ M k
dn i ⎣ ∂T ⎦P,nk dn i ⎣ ∂P ⎦T ,nk dni k =1 dn i

⎡ ∂(nM ) ⎤ dT ⎡ ∂(nM ) ⎤ dP
=⎢ ⎥ + ⎢ ∂P ⎥ +Mi
⎣ ∂T ⎦P,nk dn i ⎣ ⎦T ,nk i
dn

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 397

Restrict to constant nk, T and nV

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ ⎡ ∂P ⎤
⎢ ∂n ⎥ = ⎢ ∂P ⎥ ⎢ ∂n ⎥ +Mi
⎣ i ⎦T ,nV ,nk ≠i ⎣ ⎦T ,nk ⎣ i ⎦T ,nV ,nk ≠i

Therefore

⎡ ∂(nM ) ⎤ ⎡ ∂(nM ) ⎤ ⎡ ∂P ⎤
Mi = ⎢ ⎥ − ⎢ ⎥ ⎢ ∂n ⎥
⎣ ∂n i ⎦T ,nV ,nk ≠i ⎣ ∂P ⎦T ,nk ⎣ i ⎦T ,nV ,nk ≠i

Now take

M = lnφ → M i = lnφ̂i

We showed previously that

P
⌠ dP
lnφ = ⎮
⎮ ( Z − 1)
⌡ P
0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 398

Convert the variable of integration using the compressibility factor equation of
state

ZRT
P=
V

RT ZRT ZR
dP = dZ − 2 dV + dT
V V V
dP dZ dV dT
= − +
P Z V T

For any isothermal situation, the last term is zero, and the integral becomes
P Z V Z V
⌠
lnφ = ( Z − 1)
dP ⌠ ( Z − 1) dZ − ⌠ dV ⌠ Z
dZ ⌠ ⌠ dV
⎮
⎮ = ⎮
⎮
⎮
⎮ ( Z − 1) = ⌡ dZ −
⎮ − ⎮⎮ ( Z − 1)
⎮
⎮
⌡ P ⌡ Z ⌡ V 1
Z ⌡ ⌡ V
0 1 ∞ 1 ∞

Then

V
⌠ dV
lnφ = ( Z − 1) − ln Z − ⎮⎮ ( Z − 1)
⌡ V
∞

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 399

Multiply by n
V V
⌠ dV ⌠ dV
n lnφ = n ( Z − 1) − n lnZ − ⎮⎮ n ( Z − 1) = nZ − n − n lnZ − ⎮⎮ n ( Z − 1)
⌡ V ⌡ V
∞ ∞

The partial derivative is

⎡ ∂(n lnφ ) ⎤ ⎛ ∂(nZ ) ⎞
⎢ ∂n ⎥ =⎜ ⎟⎠ − 1− lnZ
⎣ i ⎦T ,nV ,nk ≠i ⎝ ∂n i T ,nV ,nk ≠i
V
⌠⎡ ⎤ dV
⎡
n ∂Z ⎤ ⎮ ⎛ ∂(nZ ) ⎞
− ⎢ ⎥ − ⎮ ⎢⎜ ⎟ − 1⎥
Z ⎣ ∂n i ⎦T ,nV ,n ⎮ ⎢ ⎝ ∂n ⎠
⎮ ⎥V
⌡⎣ ⎦
k ≠i
i T ,nV ,nk ≠i
∞
Note that
⎡ ∂(nZ ) ⎤ ⎡ ∂Z ⎤
⎢ ∂n ⎥ = Z + n ⎢ ∂n ⎥
⎣ i ⎦T ,nV ,nk ≠i ⎣ i ⎦T ,nV ,nk ≠i

Therefore
n ⎡ ∂Z ⎤ 1 ⎡ ∂(nZ ) ⎤
⎢ ∂n ⎥ = ⎢ ∂n ⎥ −1
Z ⎣ i ⎦T ,nV ,nk ≠i Z ⎣ i ⎦T ,nV ,nk ≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 400

Substitution into the previous expression yields

V
⎡ ∂(n lnφ ) ⎤ Z − 1⎛ ∂(nZ ) ⎞ ⎡⎛ ∂(nZ ) ⎞
⌠
⎮
⎤ dV
⎢ ∂n ⎥ = ⎜ ⎟ − lnZ − ⎢⎜
⎮
⎟ − 1⎥
⎣ i ⎦T ,nV ,nk ≠i Z ⎝ ∂n i ⎠ T ,nV ,n ⎢⎝ ∂n i ⎠ T ,nV ,n
⎮
⎮ ⎥V
k ≠i ⎣
⌡ k ≠i ⎦
∞

⎡ ∂(nM ) ⎤
This is the actual form of the ⎢ ⎥ term.
⎣ ∂n i ⎦T ,nV ,nk ≠i
Also

⎡ ∂P ⎤ ⎡ ∂ ⎛ ZRT ⎞ ⎤ ⎡ ∂ ⎛ nZRT ⎞ ⎤
⎢ ∂n ⎥ = ⎢ ⎜ ⎟⎥ = ⎢ ⎜ ⎟⎥
⎣ i ⎦T ,nV ,nk ≠i ⎣ ∂n i ⎝ V ⎠ ⎦T ,nV ,nk ≠i ⎣ ∂n i ⎝ nV ⎠ ⎦T ,nV ,nk ≠i

RT ⎡ ∂(nZ ) ⎤
=
nV ⎢⎣ ∂n i ⎥⎦T ,nV ,n
k ≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 401

 ⎛ ∂(nM ) ⎞
The term of the form ⎜ ⎟ is
⎝ ∂P ⎠ T ,n
k

⎡ ⎧ P ⎫⎤
⎡ ∂(n lnφ ) ⎤ ∂ ⎪ ⌠ dP ⎪ ⎥ n ( Z − 1) nV ( Z − 1)
⎢ ∂P ⎥ =⎢ ⎨ ⎮
n ⎮ ( Z − 1) ⎬⎥ = =
⎣ ⎦T ,nk ⎢ ∂P P ⎪ P RT Z
⎢⎣ ⎪⎩ ⌡0 ⎥
⎭ ⎦T ,nk

Then, combining the two previous results yields

⎡ ∂(nln φ ) ⎤ ⎡ ∂P ⎤ RT ⎡ ∂(nZ) ⎤ n(Z − 1)V

⎢ ⎥ ⎢ ⎥ = ⎢ ⎥
⎣ ∂P ⎦T,nk ⎢⎣ ∂n i ⎥⎦T,nV,nk≠i nV ⎢⎣ ∂n i ⎥⎦T,nV,nk≠i ZRT

Z − 1 ⎡ ∂(nZ) ⎤
= ⎢ ⎥
Z ⎢⎣ ∂n i ⎥⎦T,nV,n
k≠ i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 402

Collecting all the results produces

lnφ̂i =
( Z − 1) ⎡ ∂(nZ ) ⎤ − lnZ
Z ⎢ ∂n ⎥
⎣ i ⎦T ,nV ,nk ≠i

V
⌠
⎮
⎧⎪ ⎡ ∂(nZ ) ⎤ ⎫⎪ dV ( Z − 1) ⎡ ∂(nZ ) ⎤
−⎮ ⎨⎢ ⎥ − 1⎬ −
⎮
⎮ ⎪⎩ ⎣ ∂n i ⎦T ,nV ,nk ≠i ⎪⎭ V Z ⎢⎣ ∂n i ⎥⎦T ,nV ,n
⌡ k ≠i
∞
which, after canceling the first and last terms, becomes

∞
⌠⎧ ⎫⎪ dV
lnφ̂i =
(
⎮ ⎪ ⎡ ∂ nZ )⎤ − 1⎬ − ln Z
⎮ ⎨⎢ ⎥
⎮ ⎣ ∂n i ⎦T ,nV ,nk ≠i V
⌡⎩⎪ ⎭⎪
V

ρ
⌠⎧ ⎫⎪ d ρ
=
(
⎮ ⎪ ⎡ ∂ nZ )⎤ − 1⎬ − ln Z
⎮ ⎨⎢ ⎥
⎮ ⎣ ∂n i ⎦T ,nV ,nk ≠i ρ
⌡⎪⎩ ⎪⎭
0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 403

Alternative notation

( Z i − 1) dV
⌠
lnφ̂i = ⎮
⎮ − lnZ
⌡ V
V
where:
⎡ ∂(nZ ) ⎤
Z i = ⎢ ⎥
⎣ ∂n i ⎦T ,nV ,nk ≠i

4.2.3.5.1 General Relationship for Cubic Equations

The more common cubic equations can be expressed in a general form as

follows

b ⎡⎛ V − b ⎞ ⎤ a 1 ⎛ ai b i ⎞ ⎡V + σ b ⎤
lnφ̂i = i ( Z − 1) − ln ⎢⎜ Z
⎟⎠ ⎥ + ⎜ 1+ − ⎟ ln ⎢
b ⎝
⎣ V ⎦ bRT ( ε − σ ) ⎝ a b ⎠ ⎣ V + εb ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 404

where

V aV
Z= −
V − b RT (V + ε b ) (V + σ b )

α ΩaR 2Tc,i
2
ai =
Pc,i

ΩbRTc,i
bi =
Pc,i

1
α RK =
T

( )( )
2
α SRK = ⎡1+ 0.48508 + 1.5517ω − 0.1561ω 2 1− Tr ⎤
⎣ ⎦

( )( )
2
α PR = ⎡1+ 0.37464 + 1.54226ω − 0.26992ω 2 1− Tr ⎤
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 405

The numerical values of e, s, Wa and Wb for the Redlich-Kwong, Soave-Redlich-
Kwong and Peng-Robinson equations of state are given in the table below.

RK SRK PR

e 0 0 1− 2

s 1 1 1+ 2

1 1
Wa
9 ( 3
2 −1) 9 ( 3
2 −1) 0.45724

3
2 −1 3
2 −1
Wb 0.07780
3 3

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 406

4.2.3.5.2 Mixing rules for cubic equations of state

The relations between the cubic equation parameters for the mixture and the
composition and parameters for the pure fluids are known as mixing rules.
These provide the key to accuracy in descriptions of mixture behavior.

An excellent reference for detailed information on this subject is:

H. Orbey and S. I. Sandler, Modeling Vapor-Liquid Equilibria: Cubic

Equations of State and Their Mixing Rules, Cambridge University Press,
1998

4.2.3.5.2.1 Conventional
The form used most frequently are known as the Lorentz-Berthelot mixing rules,
which have the following form
bm = ∑ x i b i
i
2
⎛ ⎞
am = ⎜ ∑ x i ai ⎟ = ∑ ∑ ai  x i x 
⎝ i ⎠ i 

ai  = aii a (1− k i  )

where: k i  is the binary interaction parameter

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 407

4.2.3.5.2.2 Conformal solution

For this description, the assumption is that the intermolecular potential is a

universal function, with reducing parameters fij and hij, so that

(
uij (r ) = fij u0 r / hij1/ 3 )
h = ∑ ∑ x i x j hij
i j

1
f = ∑ ∑ x i x j fij hij
h i j

In this approach, the Peng-Robinson equation assumes the form

c cd
+d +2
V
Z= − RT RT
V −b
(V + b ) + ⎛⎜⎝ ⎞⎟⎠ (V − b )
b
V

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 408

where the pure fluid values are

bii = 0.07780RTc,i / Pc,i

c ii = aii α PR = 0.45724 α PR R 2Tc,i

2
/ Pc,i

aii α PR
d ii = = 0.45724 α PR RTc,i / Pc,i
RTc

The mixture values are: and the cross values are:

3
⎡ bii1/ 3 + b1/ 3⎤
bm = ∑ ∑ x i x j bij
i j
(
bij = 1−  ij ) ⎢
⎢⎣ 2
jj
⎥
⎥⎦

( )
c ii c jj
c m = ∑ ∑ x i x j c ij c ij = 1− k ij bij
i j bii b jj
3
3 ⎡ d ii + d jj ⎤
1/ 3 1/ 3
d m = ∑ ∑ x i x j d ij
i j
(
d ij = 1− mij ⎢
⎢⎣
) 2
⎥
⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 409

Complicated mixing rules are required when using cubic equations of state to
represent the VLE behavior of strongly interacting molecules such as water and
alcohols.

4.2.3.5.2.3 Mixing Rules Based on Excess Properties

For highly non-ideal solutions, the phase behavior descriptions obtained from
equations of state can be improved significantly by incorporating known
information about the behavior of the liquid phase. For many systems, low
pressure VLE measurements at one or more temperatures provide activity
coefficient data.
The excess Gibbs function is given by
GE ⎛ φ̂ ⎞
= ∑ x i lnγ i = ∑ x i ln ⎜ i ⎟ = ∑ x i lnφ̂i − ∑ x i lnφi
RT i i ⎝ φi ⎠ i i

Also: ∑ x i lnφ̂i = lnφ

i
V
⌠ dV
and: lnφ = ( Z − 1) − lnZ − ⎮⎮ ( Z − 1)
⌡ V
∞

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 410

Therefore

G ⎡
E
V
⌠ dV ⎤ ⎡ V
⌠ dV ⎤
= ⎢ Z − lnZ − ⎮⎮ ( Z − 1) ⎥ − ∑ xi ⎢ Z i − lnZ i − ⎮⎮ ( Z i − 1) ⎥
RT ⎢ ⌡ V ⎥⎦ i ⎢⎣ ⌡ V ⎥⎦
⎣ ∞ ∞

The interest here is in cubic equations of state, which follow the general form

V a
Zk = − fk (T ,V ) k = RK, SRK, PR, etc.
V − b bRT

For this form, the excess Gibbs energy becomes

GE ⎛Z⎞ ⎡ 1− b / V ⎤
= Z − ∑ x i Z i − ∑ x i ln ⎜ ⎟ − ∑ x i ln ⎢ ⎥
RT i i ⎝ Zi ⎠ i ⎣ 1− bi / Vi ⎦

⎛ a ⎞ ⎛ ai ⎞
+⎜ ( ) − ∑ C (Vi )
⎝ bRT ⎟⎠
C V x i⎜
⎝ bi RT ⎟⎠
where
V
⌠ dV
C (V ) = ⎮ fk
⎮ (T ,V )
⌡ V
∞

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 411

 E
Note that GEOS varies with pressure. Most of the models used for GγE are
independent of pressure, so that

E
GEOS = GγE

can be used at only one pressure. There are two potential choices

Zero pressure: GγE = lim

P→0
E
GEOS

Infinite pressure: GγE = lim E

GEOS
P→∞

The available activity coeficient data usually are at low pressures, but there is no
zero pressure root for the cubic equations for reduced temperatures above 0.8 to
0.85, depending on the equation.

As the pressure becomes infinite: lim

P→∞
V =b

The infinite pressure limit yields an equation relating a and b to the activity
coefficient data. The function C assumes a numerical value, C*, in the limit,
which varies with equation of state.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 412

4.2.3.5.2.3.1 Huron-Vidal

Vidal and co-workers were first to use the infinite pressure limit, resulting in the
Huron-Vidal mixing rules

b = ∑ x i bi
i
⎡ ⎛ ai ⎞ Gγ ⎤
E
a = b ⎢∑ xi ⎜ ⎟ + ⎥
⎢⎣ i ⎝ bi ⎠ C * ⎥⎦

References:

Vidal, Chem. Eng. Sci., 33, 787 (1978)

Huron and Vidal, Fluid Phase Eq., 3, 255 (1979).

The Huron-Vidal mixing rules work well for some systems, but not for others,
especially at higher pressures and temperatures (approaching the critical region
or the cricondentherm).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 413

4.2.3.5.2.3.2 Wong-Sandler

Wong and Sandler [AIChE J., 38, 671 (1992)] proposed an alternative set of
mixing rules following this approach, but recognizing that the excess Helmholtz
energy exhibits much less variation with pressure than does the excess Gibbs
energy.

At low pressures

GE (T ,LowP,x ) = AE (T ,LowP,x ) + PV E ≅ AE (T ,LowP,x )

Furthermore, because the excess Helmholtz energy depends only weakly upon
pressure

AE (T ,P = ∞,x ) ≅ AE (T ,LowP,x ) ≅ G E (T ,LowP,x )

The excess Helmholtz energy is related to the equation of state through

E ⎡V V ⎤
A Z ⎢⌠ dV ⌠ dV ⎥
RT
= − ∑ x i ln − ⎮⎮ ( Z − 1)
Z ⎢ ⌡ V
− ∑ xi ⎮
⎮
⌡
( Z i − 1) V ⎥
i i i
⎣∞ ∞ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 414

Wong and Sandler also used the mixing rule for second virial coefficients

B = ∑ ∑ x i x k Bik
i k

to obtain the following expression

a ⎡ aik ⎤
b− = ∑ ∑ xi xk b −
⎢⎣ ik RT ⎥⎦
RT i k

The Wong-Sandler mixing rules are

1 ⎛ Gγ ⎞
E
a ai
= ⎜ ⎟ + ∑ x i
bRT C * ⎝ RT ⎠ i bi RT

a ⎛ a ⎞
b− = ∑ ∑ xi xk ⎜ b − ⎟
RT i k ⎝ RT ⎠ ik

In the original formulation, the cross term is

⎛ a ⎞ 1 ⎡⎛ ai ⎞ ⎛ a ⎞ ⎤
⎜⎝ b − ⎟⎠ = ⎢ ⎜⎝ i
b − +
⎟⎠ ⎜⎝ 
b − ⎟⎠ ⎥ (1− k i  )
RT i  2 ⎣ RT RT ⎦
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 415
The Wong-Sandler mixing rules were reformulated by Orbey and Sandler [AIChE
J., 41, 683 (1995)] to simplify the requirements for data. The reformulated cross
term is

⎛ a ⎞ 1 ai a
⎜⎝ b − ⎟ =
RT ⎠ i  2
( bi + b  ) −
RT
(1− k i  )

and a modified NRTL expression is used for the excess Gibbs function

⎛ ∑ x k gki τ ki ⎞
GγE
= ∑ xi ⎜ k ⎟ where: gki = bk exp ( −σ τ ki )
RT i ⎜ ∑ x g i ⎟
⎝  ⎠

This formulation is useful, because infinite dilution activity coefficients at a single

temperature are sufficient to obtain the parameters for the cubic equation.
Furthermore, in the absence of data, methods for estimating infinite dilution
activity coefficients (such as UNIFAC or ASOG) may be used.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 416

4.3 CHEMICAL REACTION EQUILIBRIUM

4.3.1 Chemical Reaction (Generalized Form)

ν1 A1 + ν 2 A2 + … → … + ν m−1 Am−1 + ν m Am

where: m = number of chemical species involved

ni = stoichiometric coefficient for species i

Ai = chemical compound or element (species i)

Alternate General Form:

∑ ν i Ai → 0

In both cases, ni < 0 for species on left

ni > 0 for species on right

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 417

As the reaction occurs,

Δni ν i
=
Δn j ν j

because of the stoichiometry of the reaction.

4.3.2 Reaction coordinate

Rearrange the previous result to obtain:

Δni Δn j Δnk
= = =
νi νj νk

In the limit of infinitesimal changes, ∆ni ® dni. Use the above result to define the
reaction coordinate (or extent of reaction), e:

dni dn j dnk
dε = = = =
νi νj νk

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 418

Then:

dni = ν i d ε → ni − n oi = ν i ε

Stoichiometry and compositions (single reaction):

ni = n oi + ν i ε

nT = ∑ ni = ∑ n oi + ∑ ν i ε = nTo + ε ∑ ν i
i i i i

ni n oi + ν i ε
xi = =
nT nTo + ε ∑ ν i
i

ε = 1 defines a mole of reaction

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 419

Multiple Simultaneous Reactions
For multiple reactions we have a system of equations:
∑ ν ij Ai → 0 ( j = 1, 2, 3,, r )
where: i denotes chemical species
j denotes individual reaction
r = number of independent reactions
There is a reaction coordinate, ej, for each reaction.

Stoichiometry and compositions (multiple rx):

ni = nio + ∑ ε j ν ij
j

nT = ∑ ni = ∑ nio + ∑ ∑ ε j ν ij = nTo + ∑ ε j ∑ ν ij
i i i j j i

nio + ∑ ε j ν ij
ni j
xi = = o
nT nT + ∑ ε ∑ ν ij
j
j i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 420

4.3.3 Equilibrium criteria

From the Second and First Laws of Thermodynamics we derived:

(dG ) total
T ,P
=0

so that Gtotal = nG is at a minimum when the system is at equilibrium.

(nG) is a function of composition for constant T,P. Because each composition

can be written as a function of the reaction coordinate, e, (nG) becomes a
function of e.

At equilibrium, nei = n oi + ν i εe
nG
where: ee = value of e at equilibrium

ni = 0 for inert species

• Minimum

εe ε

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 421

4.3.3.1 Standard Free Energy Change/Equilibrium Constant

The general form of the Physics Summary Statement for open systems with T,P
as the independent variables is:

d (nG ) = − (nS ) dT + (nV ) dP + ∑ µi dni

i

From the definition of the reaction coordinate

dni = ν i d ε ⇒ ∑ µi dni = ∑ µi ν i d ε
i i

So:

⎛ ⎞
d (nG ) = − (nS ) dT + (nV ) dP + ⎜ ∑ µi ν i ⎟ d ε
⎝i ⎠

Math Relation 1:

⎡ ∂(nG ) ⎤ ⎡ ∂(nG ) ⎤ ⎡ ∂(nG ) ⎤

d (nG ) = ⎢ ⎥ dT + ⎢ ∂P ⎥ dP + ⎢ ∂ε ⎥ dε
⎣ ∂T ⎦P,ε ⎣ ⎦T ,ε ⎣ ⎦T ,P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 422

Comparison of the previous two equations yields

⎡ ∂(nG ) ⎤
⎢ ∂ε ⎥ = ∑ µi ν i
⎣ ⎦T ,P

Equilibrium criteria of minimum in nG at constant T,P requires

⎡ ∂(nG ) ⎤
⎢ ∂ε ⎥ = 0 ⇒ ∑ µi ν i = 0
⎣ ⎦T ,P

Also, because

µi = G i

the equilibrium condition requires:

∑ µi ν i = ∑ ν i G i = 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 423

In our previous discussions we had

dG i = RTd lnfˆi

⎛ fˆi ⎞
so that G i = G oi + RT ln ⎜ o ⎟ = G oi + RT lnα i
⎝ fi ⎠

where: G oi , f io refer to the free energy and fugacity, respectively, at a standard

state.

Note: For reaction equilibrium calculations, the standard state usually is taken
to be an ideal gas at unit pressure.

∑νiG i = 0 ⎫
⎪
⎬ ⇒ ∑ ν i G oi + RT ∑ ν i lnα i = 0
G i = G oi + RT lnα i ⎪
⎭
ν νi
Note: ∑ ν i lnα i = ∑ ln (α i ) i = ln ∏ (α i )
i
3
where:
ν
∏ (α i ) i = (α1)
ν1
(α 2 )ν (α 3 )ν
2 3

i =1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 424

Then:
ν
∑ ν i G oi = −RT ln ∏ (α i ) i = −RT lnK eq
i
where:
ν
⎡ fˆi ⎤
i

K eq ≡ ∏ (α i ) i
ν
= ∏ ⎢ o ⎥ = equilibrium constant
i i
⎢⎣ f i ⎥⎦

The Standard Free Energy of Reaction then is defined by

Δr G o ≡ −RT ln K eq = ∑ ν i G oi

Notes:
• Keq, G oi are functions only of temperature
• The standard states need not be identical for every chemical species present,
but for each species, the standard state temperature must be the equilibrium
temperature and the same standard state must be used to calculate G oi , fˆi
and f io .

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 425

4.3.4 Standard Property Change of Reaction

Definition: Δr M o = ∑ ν i M oi

where: Δr M o = standard property change of reaction.

M oi = value of property M for pure species i at the standard state.

Examples:

Δr H o = ∑ ν i H oi is the standard heat of reaction

Δr G o = ∑ ν i G oi is the standard free energy of reaction

Δr S o = ∑ ν i S oi is the standard entropy change of reaction

Δr C po = ∑ ν i C po i is the standard heat capacity change of reaction

Note: All standard property changes of reactions are functions only of

temperature.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 426

Standard property changes of reaction obey the same thermodynamic relations
as the properties discussed previously.

Δr Ao = Δr U o −T Δr S o

Δr U o = Δr H o − Δr (PV )
o

Δr S = −o (
d Δr G o )
dT

⎡
d ⎢ Δr G (
o
) ⎤⎥ = − Δ H r
o

dT ⎢ RT ⎥ RT 2
⎣ ⎦

(
d Δr H o )=ΔC o
r p
dT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 427

4.3.4.1 Temperature dependence of equilibrium constant

d ⎛ Δr G o ⎞ Δr H o
⎜ RT ⎟ = −
dT ⎝ ⎠ RT 2

Δr G o
= − lnK eq ⇒
(
d ln K eq )= ΔH r
o

RT dT RT 2

This result is known as the van't Hoff equation.

Integrate the van't Hoff equation

K eq (T2 )
T2
Δr H o
( )
⌠
⌠
⎮
⌡
d lnK eq = ⎮
⎮ 2
dT
K eq (T1) ⌡ RT
T1

T2
⎡ K eq (T2 ) ⎤ 1 ⌠ Δr H o
to obtain ln ⎢ ⎥= ⎮ dT
⎢⎣ K eq (T1) ⎥⎦ R
⎮ 2
⌡ T
T1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 428

or

⎡ T2
⌠ Δ Ho
⎤
1
K eq (T2 ) = K eq (T1 ) exp ⎢ − ⎮⎮ r 2 dT ⎥
⎢ R⌡ T ⎥
⎣ T1 ⎦

Also

T
Δr H o
(T ) = Δr H (To ) + ⌠⎮⌡ Δr CPo dT
o

To

Necessary information: Δr G o , Δr H o at one temperature

Δr CPo (T ) for all substances

This information gives K eq at any temperature. An equation of state also is

required to calculate the various fugacities that appear within K eq .

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 429

4.3.4.2 Alternate Method for Temperature Dependence of Equilibrium
Constant

Normally, Δr H o is a function of temperature, so the value of the free energy of

formation at temperatures other than To is calculated as follows:

Δr G° (T ) = Δr H ° (T ) −T ΔS° (T )

where
T
Δr H (T ) = Δr H (To ) +
o o ⌠
⎮
⌡
Δr C po dT
To

and

T
⌠ Δr C po
Δr S o (T ) = Δr S o (To ) + ⎮
⎮
dT
⌡ T
To

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 430

Combining the above produces
T ⎡ T ⎤
Δr G° (T ) Δr H (To ) 1 ⌠ Δ Co o
o
1⎢ ⌠ Δr C p ⎥
dT − ⎢ Δr S o (To ) + ⎮⎮
r p
= + ⎮
⎮
dT ⎥
RT RT T ⌡ R R⎢ ⌡ T ⎥
To ⎣ To ⎦
T T
Δ H (To ) Δr S (To ) 1
o o ⌠ Δ Co
r p
⌠ Δr C po
= r − + ⎮
⎮
dT − ⎮
⎮
dT
RT R T ⌡ R ⌡ RT
To To
Substituting
Δr S o (To ) Δr H o (To ) Δr G o (To )
= −
R RTo RTo
produces
T T
Δr G° (T ) Δr H (To ) Δr H (To ) − Δr G° (To ) 1 ⌠ Δ Co Δr C po
o o ⌠
r p
= − + ⎮
⎮
dT − ⎮
⎮
dT
RT RT RTo T ⌡ R ⌡ RT
To To
T T
Δ G° (To ) Δr H (To ) ⎡ 1 1 ⎤ 1 o ⌠ Δ Co
r p
⌠ Δr C po
= r + ⎢ − ⎥+ ⎮
⎮
dT − ⎮
⎮
dT
RTo R ⎣T To ⎦ T ⌡ R ⌡ RT
To To

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 431

4.3.4.3 Free Energies of Formation

Consider the special case of a property change of reaction when the reaction
forms a compound from its constituent elements.

Δf M o = M substance
o
− ∑ ν i M oi
constituent
elements

The subscript, f, denotes properties of formation.

All reactions can be described as combinations of formation reactions,

Δr M o = ∑ ν i Δf M oi

Because of the known relations between thermodynamic proporties, only the

Δf M o values for two properties (usually Δf H o and Δf G o ) need be tabulated.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 432

Then:
Δr G o (To ) = ∑ ν i Δf G oi (To )

Δr H o (To ) = ∑ ν i Δf H oi (To )
Δr Ao = Δr U o −T Δr S o

Δr U o = Δr H o − Δr (PV )
o

Δr S = −o (
d Δr G o )
dT
⎡
d ⎢ Δr G (
o
) ⎤⎥ = − Δ H r
o

dT ⎢ RT ⎥ RT 2
⎣ ⎦
(
d Δr H o )=ΔC o
r p
dT
Necessary information: Δf G o and Δf H o at one temperature, ideal gas heat
capacities and the equation of state. This information allows calculation of K eq
and the fugacities contained in equilibrium constant.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 433

4.3.5 Standard and Reference State Choices

In this section, the term “standard state” denotes the base states for calculating
the standard free energy, standard entropy and standard enthalpy of reaction.
The term “reference state” denotes the base states for calculating the fugacity of
a species in the mixture. In many cases, the standard and reference states are
different physical states.

Notation: f io denotes standard states

f i† denotes reference states

The general equilibrium relation is

νi
⎡ fˆi ⎤ ⎡ Δr G o ⎤
K eq = ∏⎢ o ⎥ = exp ⎢ − ⎥
i ⎣f i ⎦ ⎢⎣ RT ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 434

Common Standard States, ( f io )

Ideal gas at T, P°
PURE liquid or solid at T, P°
Infinite dilution

Common Reference States

Fugacity coefficient: Ideal gas at T, P

Lewis- Randall: Pure liquid or solid at T, P

Henry’s Law: Solute behavior at infinite dilution

4.3.5.1 Ideal Gas Standard States

For the ideal gas standard state: f io = P o

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 435

4.3.5.1.1 Ideal Gas Reference State

The fugacity coefficient notation is based upon an ideal gas reference state,
where the ideal gas is at the same pressure as the system. The fugacity of
substance i in the mixture is

fˆi = y i φ̂i P

The equilibrium constant then is

ν ν
⎡ fˆi ⎤ i ⎡ y i φ̂i P ⎤ i ⎛ P ⎞ ν
ig
K eq = ∏⎢ o ⎥ = ∏⎢
i ⎣ P
o ⎥
= ⎜ o ⎟ ∏ y i φ̂i
⎝P ⎠ i
( ) νi

i ⎣f i ⎦ ⎦

ν ν
⎛ P ⎞
⎝P ⎠ i
ν
= ⎜ o ⎟ ∏ ( y i ) i ∏ φ̂i
i
( )
νi ⎛ P ⎞
= K y ⎜ o ⎟ ∏ φ̂i
⎝P ⎠ i
( )
νi

where
νi
K y = ∏ (yi ) and ν = ∑νi
i i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 436

4.3.5.1.2 Lewis-Randall Reference State

The fugacity of the substance in the mixture is

fˆi = x i γ iLR f iLR = x i γ iLR fi (T ,P )

where fi (T ,P ) is the fugacity of pure substance i at the same temperature and

pressure, and in the same phase as the mixture. Then

νi νi
ig ⎡ fˆi ⎤ ⎡ x i γ iLR f iLR ⎤
K eq = ∏⎢ o ⎥ = ∏⎢ ⎥
i ⎣fi ⎦ i ⎢⎣ Po ⎥⎦

νi νi
⎡ f iLR ⎤ ⎡ f (T ,P ) ⎤
=∏
i
( xi γ i )
LR ν i
∏⎢ o ⎥
i ⎢⎣ P ⎥⎦
= Kx ∏
i
( )
LR ν i
γi ∏⎢ i o ⎥
i ⎣ P ⎦

f iLR φiσ Piσ ⎡ 1 P ⎤

νi
where K x = ∏ (xi ) and = exp ⎢ ⎮ V dP ⎥
⌠
⌡ i
i Po P o
⎢ RT Piσ ⎥
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 437

4.3.5.1.3 Henry's Law Reference State

The fugacity of the substance in the mixture is

fˆi = x i γ iHL Η i (T )

Henry’s constant is based upon the behavior of the solute at infinite dilution in the
mixture

⎛ fˆi ⎞
Ηi = lim
x i →0 ⎜x ⎟
⎝ i⎠

Then

νi νi ν
⎛ fˆi ⎞ ⎛ x iγ iHL Η i ⎞ ⎛ Ηi ⎞ i
( )
ig νi
K eq = ∏⎜ o ⎟ = ∏⎜ ⎟ = ∏ ⎜ o ⎟ ∏ x i γ iHL
i ⎝ fi ⎠ i ⎝ Po ⎠ i ⎝P ⎠ i

ν
⎛ Ηi ⎞ i
( )
νi
= K x ∏ ⎜ o ⎟ ∏ γ iHL
i ⎝P ⎠ i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 438

4.3.5.2 Liquid or Solid Standard States

For liquid or solid standard states, the standard state is the fugacity of the pure
material at the standard state pressure

f io = f i P o( )
4.3.5.2.1 Ideal Gas Reference State

The fugacity coefficient notation is based upon an ideal gas reference state,
where the ideal gas is at the same pressure as the system. The fugacity of
substance i in the mixture is

fˆi = y i φ̂i P

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 439

The equilibrium constant then is
νi
νi ⎡ ⎤
⎡ fˆi ⎤ y φ̂ P
L/S
= ∏⎢ o ⎥ = ∏ ⎢ i io ⎥
( )
K eq
i ⎣f i ⎦ i ⎢f i P ⎥
⎣ ⎦

( )
−ν i
= P ∏ ⎡f i P o ⎤
ν
i ⎣ ⎦ (
∏ y i φ̂i
i
)
νi

( )
−ν i
= P ∏ ⎡f i P o ⎤
ν
i ⎣ ⎦ i
ν
∏ ( y i ) i ∏ φ̂i
i
( )νi

( )
−ν i
= K y P ∏ ⎡f i P o ⎤
ν
i ⎣ ⎦
∏ φ̂i
i
( ) νi

ν
⎡ ⎤ i
= Ky ∏ ⎢
P ⎥ ∏ φ̂i ( ) νi

i ⎢f i P
⎣
o
( ) ⎥ i
⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 440

4.3.5.2.2 Lewis-Randall Reference State

The fugacity of the substance in the mixture is

fˆi = x i γ iLR f iLR = x i γ iLR fi (T ,P )

where fi (T ,P ) is the fugacity of pure substance i at the same temperature and

pressure, and in the same phase as the mixture. Then
νi νi
νi ⎡ LR LR ⎤ ⎡ LR ⎤
⎡ fˆi ⎤ γ
( )
x f νi f
L/S
= ∏⎢ o ⎥ = ∏⎢ i i oi ⎥ = ∏ x i γ iLR ∏⎢ i o ⎥
( ) ( )
K eq
i ⎣fi ⎦ i ⎢ fi P ⎥ i i ⎢fi P ⎥
⎣ ⎦ ⎣ ⎦

νi
⎡ ⎤
f (P )
( )
νi
= K x ∏ γ iLR ∏⎢ i o ⎥
i i ⎢fi P
⎣ ( ) ⎥
⎦

f i (P ) ⎡ 1 P ⎤
= exp ⎢ ⌠Vi dP ⎥
( )
and ⎮
⌡
fi Po ⎢⎣ RT Po ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 441

4.3.5.2.3 Henry's Law Reference State

The fugacity of the substance in the mixture is

fˆi = x i γ iHL Η i (T )

Henry’s constant is based upon the behavior of the solute at infinite dilution in the
mixture
⎛ fˆi ⎞
Ηi = lim
x i →0 ⎜x ⎟
⎝ i⎠

Then
νi ν
νi ⎛ ⎞ ⎛ ⎞ i
⎛ fˆi ⎞ x iγ iHL Η i Ηi
( )
L/S νi
= ∏⎜ o ⎟ = ∏⎜ ⎟ = ∏⎜ ⎟ ∏ x i γ iHL
( ) ( )
K eq
i ⎝ fi ⎠ i ⎜⎝ f i P ⎟⎠ i ⎜⎝ f i P ⎟⎠ i
o o

ν
⎛ ⎞ i
Ηi
( )
νi
= Kx ∏⎜ ⎟ ∏ γ iHL
i ⎜⎝ f i P
o
( ) ⎟⎠ i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 442

4.3.5.3 Infinite Dilution Standard State
In some cases, such as ionic solutions, it is convenient to use a standard state
for the solute already in solution

⎛ fˆi ⎞
f oi = Ηi = lim
x i →0 ⎜x ⎟
⎝ i⎠

4.3.5.3.1 Ideal Gas Reference State

The fugacity coefficient notation is based upon an ideal gas reference state,
where the ideal gas is at the same pressure as the system. The fugacity of
substance i in the mixture is

fˆi = y i φ̂i P
The equilibrium constant then is
ν ν ν
⎡ fˆi ⎤ i ⎡ y i φ̂i P ⎤ i ⎛P⎞
∞
K eq = ∏⎢ o ⎥ = ∏⎢
i ⎣f i ⎦ i ⎣ Ηi ⎦
⎥ = ∏ ⎜ ⎟ ∏ y i φ̂i
i ⎝ Ηi ⎠ i
( )
νi

ν ν
⎛P⎞ ⎛P⎞
ν
= ∏ ⎜ ⎟ ∏ ( y i ) i ∏ φ̂i
i ⎝ Ηi ⎠ i i
( )
νi
( )
= K y ∏ ⎜ ⎟ ∏ φ̂i
i ⎝ Ηi ⎠ i
νi

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 443

4.3.5.3.2 Lewis-Randall Reference State

The fugacity of the substance in the mixture is

fˆi = x i γ iLR f iLR = x i γ iLR fi (T ,P )

where fi (T ,P ) is the fugacity of pure substance i at the same temperature and

pressure, and in the same phase as the mixture. Then

νi νi
∞ ⎡ fˆi ⎤ ⎡ x i γ iLR f iLR ⎤
K eq = ∏⎢ o ⎥ = ∏⎢ ⎥
i ⎣fi ⎦ i ⎢⎣ Ηi ⎥⎦

νi
⎡ f iLR ⎤
=∏
i
( )
LR ν i
xi γ i ∏⎢
Η
i ⎢⎣ i ⎥⎦
⎥

νi
⎡ f (P ) ⎤
= Kx ∏
i
( )
LR ν i
γi ∏⎢ i
i ⎣ Ηi ⎦
⎥

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 444

4.3.5.3.3 Henry's Law Reference State

The fugacity of the substance in the mixture is

fˆi = x i γ iHL Η i (T )

Henry’s constant is based upon the behavior of the solute at infinite dilution in the
mixture

⎛ fˆi ⎞
Ηi = lim
x i →0 ⎜x ⎟
⎝ i⎠

Then

νi νi ν
⎛ fˆi ⎞ ⎛ x iγ iHL Η i ⎞ ⎛ Ηi ⎞ i
( ) ( )
∞ νi νi
K eq = ∏⎜ o ⎟ = ∏⎜ ⎟ = ∏ ⎜ ⎟ ∏ x i γ iHL = K x ∏ γ iHL
i ⎝ fi ⎠ i ⎝ Ηi ⎠ i ⎝ Ηi ⎠ i i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 445

4.3.5.4 Summary of the Equilibrium Constant Definitions
ν
ig ⎛ fˆi ⎞ i equilibrium constant based upon ideal gas standard
K eq = ∏⎜ o ⎟
i ⎝P ⎠
states
νi
⎛ ˆ ⎞
fi equilibrium constant based upon pure liquid or solid
L/S
= ∏⎜ ⎟
( )
K eq standard state
i ⎜⎝ f i P ⎟⎠
o

ν
∞ ⎛ fˆi ⎞ i equilibrium constant based upon infinite dilution
K eq = ∏⎜ ⎟
i ⎝ Ηi ⎠
standard state
νi
K P = ∏ (Pi ) Pi = y i P = partial pressure of specie i
i
νi
K y = ∏ (yi ) y i = mole fraction of specie i in the vapor phase
i
νi
K x = ∏ (xi ) x i = mole fraction of specie i in the liquid phase
i

K c = ∏ (c i )
νi c i = concentration of specie i (usually in the liquid
i phase)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 446

4.3.5.4.1 Relations Between Various K's

( )
νi
ig
K eq ⎛ f Po ⎞
i
L/S
= ∏⎜ o
⎟
K eq i ⎜⎝ P ⎟⎠

ig νi
K eq ⎛Η ⎞
∞
= ∏ ⎜ oi ⎟
K eq i ⎝P ⎠

( )
νi
∞
K eq ⎛ f Po ⎞
i
= ∏⎜ ⎟
i ⎜⎝ Η i
L/S
K eq ⎟⎠

ν νi ν νi
⎡ fˆi ⎤ i ⎡ fˆi ⎤ ⎡ f io ⎤ i ⎡ f io ⎤
K p = ∏ [Pi ]
νi ig
= ∏⎢ V ⎥ = ∏⎢ o ⎥ ⎢ V ⎥ = K eq ∏⎢ V ⎥
i i ⎢⎣ φ̂ i ⎥⎦ i ⎢⎣ f i ⎥⎦ ⎢⎣ φ̂ i ⎥⎦ i ⎢⎣ φ̂ i ⎥⎦

ig
Note: K eq is dimensionless but KP has dimensions of pressure if n ≠ 0.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 447

Also:

ν νi νi
⎡ fˆi ⎤ i ⎡ fˆi ⎤ ⎡ f io ⎤ ⎡ Po ⎤
K y = ∏ [yi ]
νi ig
= ∏⎢ V ⎥ = ∏⎢ o ⎥ ⎢ V ⎥ = K eq ∏ ⎢ V ⎥
i i ⎢⎣ φ̂ i P ⎥⎦ i ⎢⎣ f i ⎥⎦ ⎢⎣ φ̂ i P ⎥⎦ i ⎢⎣ φ̂ i P ⎥⎦

νi
Kx ⎛ φ̂i P ⎞
= ∏⎜ ⎟
Ky i ⎝ γ if i ⎠

Note: Both KP and Ky are functions of pressure through φ̂Vi , and Ky is directly a
function of pressure if n ≠ 0.

νi
⎛ f io ⎞
( )
νi νi
K c = ∏ (c i ) = ∏ yi c = K eq ∏ ⎜ v ⎟
i i i ⎝ φ̂i cP ⎠

Note: K P , K y and K c may depend upon pressure, temperature and

composition, but K eq depends only upon temperature.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 448

Composition/Standard State/Reference State Matrix

Composition Standard State Reference State

Mole fraction Ideal gas Ideal gas

Mass fraction Pure liquid or solid Lewis-Randall

Molality Infinite dilution Henry's law

Molarity

Concentration

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 449

4.3.6 Reaction Equilibrium by Minimization of Gibbs Free Energy

Consider the case where m chemical species are present at fixed temperature
and pressure. The total Gibbs energy for this system is a function of the
amounts of each species present:

Gtotal = nG (n1,n2 ,n3 ,,nm )

When reactions occur, the number of atoms of each element present must be
conserved even though the atoms may be regrouped through the reactions to
form different compounds.

The mass balance for each atom provides an additional constraint for the
minimization problem:
m
A j − ∑ ni aij = 0
i =1

where: Aj = number of gram atoms of element j.

ni = number of moles of substance i present.

aij = number of atoms of element j present in molecule i.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 450

The total Gibbs energy for the mixture is given by:
m ⎡ m fˆi ⎤ m
nG = ∑ n i G i = ∑ n i ⎢G i + RT ln † ⎥ = ∑ ni G i + nΔM G
†
i =1 i =1 ⎢⎣ f i ⎥⎦ i =1
m
E
N ⎛n ⎞
= ∑ ni Gi + nG + ∑ ni ln ⎜ i ⎟
i =1 i =1 ⎝n⎠
† † o
If G i = 0 for the elements in their standard state, then G i = Δf G i for all
compounds, and:
⎡
m ⎛ fˆi ⎞ ⎤
nG = ∑ ni ⎢ Δf G i + RT ln ⎜ o ⎟ ⎥
o
i =1 ⎢ ⎝ f i ⎠ ⎥⎦
⎣

fˆi y i φ̂Vi P
Vapor phase: = (Po = 1 in most cases)
f io Po
L
fˆi L f i
Liquid phase: = xi γ i ≅ xi γ i
f io f io

fˆi f is
Solid phase: = ≅1 (pure solids)
f io f io
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 451
4.3.6.1 Method of Lagrange multipliers

The general objective here is to minimize a function of m variables:

g = g ( x1,x 2 ,,x m )

subject to w constraints that may be written as:

f k ( x1,x 2 ,,x m ) = 0 k = 1,,w

The objective function (the function to be minimized) has m variables, but,

because of the w constraints, only m - w variables are independent.

The minimum in a function, h = h(y1, y2, ..., ym), where the {yi} all are
independent, occurs when:
∂h
=0 i = 1, , m
∂y i

∂2 h ⎧⎪ i = 1, , m
>0 ⎨
∂y i ∂y j j = 1, , m
⎩⎪
simultaneously.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 452
In the constrained problem, all m variables {xi} are not independent. One
solution is to eliminate w variables algebraically, then solve the case where the
derivatives of g with respect to the remaining m - w variables all are zero.
However, this procedure is not suited well to general applications.

An alternative method is to define a new function that is equivalent numerically to

the function of interest, but that has N independent variables.

Constraints:

fk = 0 ⇒ λk f k = 0, where lk is an arbitrary multiplier.

w
Also: ∑ λk f k = 0
k =1

w
Define: F = g + ∑ λk f k
k =1

F is numerically equal to g for all values of {xi}, therefore F is at a minimum when

g is at a minimum and vice versa.

Note: F = F ( x 1,x 2 ,,x m , λ1,, λw )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 453

For F we have:

(m + w )variables ⎫
⎬ ⇒ m independent variables
w constraints ⎭

Now we can choose {xi} to be the independent variables. Then, the minimum
values of F and hence g, occur when:

⎛ ∂F ⎞
⎜⎝ ∂x ⎟⎠ =0 for i = 1, 2, , m
i x j ≠i

The values {xi} that yield the minimum for g are given by the solution to the
following set of simultaneous equations:

⎛ ∂F ⎞
⎜⎝ ∂x ⎟⎠ =0 for i = 1, 2, , m
i x j ≠i

fk = 0 k = 1, 2, !, w

The Lagrange multipliers {lk} have no physical significance.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 454

4.3.6.2 Application to Single Phase Reactions

Total Gibbs energy:

Gtotal = nG (n1,n2 ,,nm )

Constraints (atom balances):

( )
m m
Aj − ∑ ni aij = ∑ nio − ni aij = 0 ( j = 1, 2,!,w )
i =1 i =1

where: m = number of chemical species present

w = number of different atoms present.

Then, the Lagrangian function to be minimized is:

⎡ mw ⎤
F = nG − ∑ λ j ⎢ A j − ∑ nk akj ⎥
j =1 ⎣ k =1 ⎦
So:

⎛ ∂F ⎞ ⎡ ∂(nG ) ⎤ w w
† fˆi w
⎜⎝ ∂n ⎟⎠ =⎢ ⎥ + ∑ λ j aij = G i + ∑ λ j aij = G i + RT ln † + ∑ λ j aij
i n ⎣ ∂ni ⎦n j ≠i j =1 j =1 fi j =1
j ≠i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 455

If all the fugacities are based upon the elements in their standard states, the
simultaneous equations become:

∂F o fˆi w
= Δf G i + RT ln o + ∑ λ j aij
∂ni fi j =1

Reaction equilibrium occurs when

fˆi w
Δf G oi + RT ln + ∑ λ j aij = 0 (i = 1, 2, , m)
f io j =1

m
A j − ∑ ni aij = 0 ( j = 1, 2,,w )
i =1

{ni} are the variables to be solved.

Perfect gas standard states f io = Pio

Liquid or solid standard states f io = f iL or f is

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 456

Fugacity Expressions:

Gas phase fˆiV = y i φ̂Vi P

Liquid phase fˆLi = x i γ i†f i†

Solid phase fˆsi = x i γ isf is

Useful Reference:

J. H. Dluzniewski & S. B. Adler, "calculation of complex reaction and/or phase

equilibrium problems," Industrial & Chemical Engineering Symposium
Series, 25, Institution of Chemical Engineers, London, 1972, Pages 4:21 - 4:26.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 457

4.3.7 Reaction Equilbrium Using Reaction Coordinate

Each reaction provides one equation with one unknown (the reaction coordinate)
for each independent chemical reaction present.

4.3.7.1 Single Reaction

For the case of a single reaction, the mole fraction is related to the reaction
coordinate through stoichiometry by

n oi + ν i ε
xi =
∑ n oi + ε ∑ ν i
i i

Then, the mole fraction product becomes

νi
⎡ o
⎤
⎢ n i + νi ε ⎥
Kx = ∏ ⎢ o ⎥
i ∑ n i + ε ∑νi
⎢⎣ i i ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 458

4.3.7.1.1 Gas Phase (Fugacity Coefficient)

Use the fugacity coefficient notation

ν
⎡
⎢
Ky = ∏ ⎢ o
n o
(
i + νi ε ) ⎤i ⎧
⎥ ⎪ ⎡ f o ⎤ν i ⎫
⎪ ⎡ Δ G o⎤
⎥ = ⎨∏ ⎥ ⎬ exp ⎢ −
i r
⎢ ⎥
i ∑ n i + ε ∑νi φ̂
⎪⎩ i ⎢⎣ i ⎥⎦ ⎪⎭
P ⎢⎣ RT ⎥⎦
⎢⎣ i i ⎥⎦

This gives one equation, with the single unknown e. The equation is implicit
because the φ̂i are dependent upon {yi}.

4.3.7.1.2 Liquid Phase (Activity Coefficient)

Use the activity coefficient notation

ν
⎡
⎢
Kx = ∏ ⎢ o
n o
(
i + νi ε ) ⎤i ⎧
⎥ ⎪ ⎡ fio ⎤ i⎫
ν
⎪ ⎡ Δr G o ⎤
i ∑ n i + ε ∑νi
⎥ = ⎨∏ ⎢ † † ⎥ ⎬ exp ⎢ − ⎥
⎢⎣ i ⎥⎦ ⎪⎩ i ⎢⎣ γ i f i ⎥⎦ ⎪⎭ ⎢⎣ RT ⎥⎦
i

This gives one equation, with the single unknown e. The equation is implicit
because the γ i† are dependent upon {xi}.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 459
4.3.7.2 Multiple Independent Reactions

For multiple reactions, each reaction introduces an equation with a

corresponding unknown, usually resulting in a set of coupled simultaneous
equations in the various reaction coordinates.

The stoichiometry gives

n oi + ∑ ν ik ε k
xi = k
⎛ ⎞
∑ n oi + ∑ ⎜ ε k ∑ ν ik ⎟
i k ⎝ i ⎠

where the subscript i denotes the individual chemical species (m total) and the
subscript k denotes the independent chemical reactions (r total).

The product based upon mole fractions becomes

νi
⎡ o ⎤
⎢ + ∑ ν ε ⎥
n i ik k (i = 1,2,,m )
Kx = ∏ ⎢
k k ⎥
i ⎢
∑ n o
+ ∑
⎛
ε ∑ ν
⎞⎥ (k = 1,2,,r )
⎢ i ⎜⎝ k ik ⎟⎠ ⎥
⎣ i k i ⎦
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 460
Using the fugacity coefficient notation or the activity coefficient notation (see
εi
single reaction section) then provides a system of simultaneous equations in the
reaction coordinate variables, .

4.3.8 Heat Effects Associated with Reactions

The amount of energy that must be transferred as heat to compensate for

reaction effects is an important design parameter. This can be done
conveniently when the free elements in their standard states are used as the
zero reference point for all enthalpy calculations.

The amount of energy that is transferred as heat during a constant pressure

process (assuming that no shaft work is involved) is
total total
Q = Hfinal − Hinitial

The enthalpy of any mixture may be expressed as

m
H total
(T ,P ) = nΔm H (T ,P ) + ∑ ni Hi (T ,P )
i =1

Express the pure material enthalpies relative to their standard states as

( ) (
Hi (T ,P ) = ⎡⎣Hi (T ,P ) − Hio (T ,P ) ⎤⎦ + ⎡Hio (T ,P ) − Hio T o ,P o ⎤ + Hio T o ,P o
⎣ ⎦ )
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 461
When the enthalpies are expressed relative to the free elements in their standard
states
⎧⎪0
(T )
for free elements
Hio o
,P o
=⎨ o
⎪⎩ Δ fHi for compounds

The middle term comes from

T P
⌠
⎛ ∂Hio ⎞
(T )=
⌠
Hio (T ,P ) − Hio o
,P o ⎮
⎮
⎮
o
C p,i dT +
⎮
⎮ ⎜ ∂P ⎟ dP
⌡ ⎮
⌡ ⎝ ⎠T
To Po
P
⌠
⎮ ⎛ ∂Hio ⎞ ⎧= 0 ideal gas standard states
where ⎮ ⎜ ∂P ⎟ dP ⎨≅ 0
⎮
⌡ ⎝ ⎠T ⎩ liquid or solid standard states
Po

The first term on the right is

⎧⎪H R (T ,P ) ideal gas

H i (T ,P ) − H oi (T ,P ) = ⎨
⎪⎩0 liquid or solid

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 462

Summarize results

m
H total
(T ,P ) = nΔm H (T ,P ) + ∑ ni Hi (T ,P )
i =1

⎧ T
⎪ Δf Hio + ⌠⎮⌡ C p,i
ig
dT + H R (T ,P ) ideal gas standard states
⎪ To
⎪
where H i (T ,P ) = ⎨
⎪ T
⎪ Δ H + ⌠⎮ C /s dT
o
liquid or solid standard states
⎪ f i ⌡ p,i
⎩ To

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 463

4.4 IONIC SOLUTIONS

4.4.1 Molality
To describe the behavior of ionic solutions, compositions often are expressed
using molalities.
Definition: Molality is the number of moles of solute per kilogram of solvent.
Notation: ni = number of moles of solute (i)
no = number of moles of solvent.
Mo = molar mass of solvent
mi = molality of solute (i)
Then:
1000ni ni
mi ≡ and xi =
noM o no + ∑ nk
k
so that
⎛ ⎞
no + ∑ nk 1000 ⎜ no + ∑ nk ⎟
mi ni (1000 ) ⎝ k ⎠ 1000
= k = =
xi noM o ni noM o x oM o

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 464

 mi 1000 1000x i
As ∑ nk → 0 (a very dilute solution), → and mi ≅
k xi Mo Mo

Volume Concentration

ni
Definition: ci ≡
V

total mass = ρV = noM o + ∑ nk M k

k

Then
noM o + ∑ nk M
V= k
ρ

Substitution into the definition of concentration produces

ni ρ ρx
ci = ≅ i (dilute solution)
noM o + ∑ nk M Mo
k

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 465

4.4.1.1 Treatment of Solution Thermodynamics

The properties of the various species in the solution usually are described as
follows

⎛ fˆi ⎞
Solute: G i = G ‡i + RT ln ⎜ ‡ ⎟
⎝fi ⎠

⎛ fˆo ⎞
Solvent: Go = Go† + RT ln ⎜ † ⎟
⎝ fo ⎠

Solute: fˆi = γ i m i f i‡

Solvent: fˆo = γ o x i fo†

lim ⎛ fˆi ⎞
Solute: f i‡ = mi →0 ⎜m ⎟
⎝ i⎠

Solvent: fo† = fugacity of pure liquid solvent

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 466

Then

Solvent: Go† = Gibbs energy of pure solvent as a liquid at the same

temperature and pressure as the solution

Solute: G ‡i = Gibbs energy of the solute in a hypothetical ideal solution of

unit molality at the same temperature and pressure as the
solution

Alternate form based upon volume concentrations

Solute: fˆi = γ i◊ c i fi◊

⎛ fˆi ⎞
f i◊
lim
Solute: = ci →0 ⎜c ⎟
⎝ i⎠

Solute: Go◊ = Gibbs energy of the solute in a hypothetical ideal solution of

unit concentration at the same temperature and pressure as
the solution

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 467

4.4.2 Summary of Conventions for Activities

Form of fugacity
Convention
Mixture Solutes Solvents

LR/LR x i γ i f i† f i† = f iL f i† = f iL

lim ⎛ fˆi ⎞
HL1/LR x i γ i f i† f i† = x i →0 ⎜ x ⎟ = Ηi f i† = f iL
⎝ i⎠

x oγ ofo† lim ⎛ fˆi ⎞

HL2/LR f i‡ = mi →0 ⎜m ⎟ f o† = foL
mi γ i f i‡ ⎝ i⎠

x oγ ofo† ⎛ fˆi ⎞
f i◊
lim
HL3/LR = ⎜c ⎟ f o† = foL
c i γ ◊i f i◊
ci →0
⎝ i⎠

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 468

4.4.3 Free Energy of Formation in Solution

Consider a vapor in equilibrium with a liquid solution:

Vapor:

⎛ fˆi ⎞ ⎛ P ⎞
GvI = Gio + RT ln ⎜ o ⎟ = Gio + RT ln ⎜ y i φ̂i o ⎟
⎝ fi ⎠ ⎝ P ⎠

where: P o is the standard state pressure (at which Gio is specified), and
usually is at unit pressure.

Liquid Solution:

⎛ fˆi ⎞ ⎛ m ⎞
GLI = Gi† + RT ln ⎜ † ⎟ = Gi‡ + RT ln ⎜ γ i oi ⎟
⎝ fi ⎠ ⎝ mi ⎠

‡
where: m oi is the standard state molality (at which Gi is specified), which
usually is unit molality.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 469

Then, application of the equilibrium condition produces

⎛ P ⎞ ⎛ mi ⎞
Gio ‡
+ RT ln ⎜ y i φ̂i o ⎟ = Gi + RT ln ⎜ γ i o ⎟
⎝ P ⎠ ⎝ mi ⎠

or

⎡ y φ̂ P ⎛ m o ⎞ ⎤
Gi‡ − Gio = RT ln ⎢ i i ⎜ oi ⎟ ⎥
⎢⎣ γ i m i ⎝ P ⎠ ⎥⎦

The difference between the standard states must be independent of solution

composition because the standard states are fixed states.

Use a dilute solution and a pure vapor to evaluate the numerical value of the
difference between the standard states.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 470

4.4.4 Electrolytes

Notation: Mν + Aν − = ν +Mz + + ν − Az −

where: M = positive ion

A = negative ion
z + = charge of positive ion
z − = charge of negative ion
ν + = number of positive ions in electrolyte balance
ν − = number of negative ions in electrolyte balance

Example

H2SO 4 → 2H + + SO =4

ν+ = 2 ν− = 1

z+ = 1 z − = −2

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 471

4.4.4.1 Charge Balance

z = ν +z + + ν −z − = 0 if original molecule is neutral

Notation:

no = number of moles of solvent

mu = number of moles of undissociated electrolyte ( Mν + Aν − )

m+ = number of moles of positive ions ( Mz + )

m− = number of moles of negative ions ( Az − )

4.4.4.2 Dissociation Stoichiometry

m + = ν + (m − mu )

m − = ν − (m − mu )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 472

Classes of Electrolytes

A weak electrolyte has a small degree of dissociation:

mu >> m+ ,m−

A strong electrolyte has almost complete dissociation:

mu << m ⇒ mu ≈ 0

The Gibbs energy relative to the infinite dilution standard state is related to the
Gibbs energy relative to ideal gas standard state by

⎡ y φ̂ P ⎛ m o ⎞ ⎤
Gi‡ − Gio = RT ln ⎢ i i ⎜ oi ⎟ ⎥
⎢⎣ γ i m i ⎝ P ⎠ ⎥⎦

For a pure vapor, y i = 1

In the limit of infinite dilution, γ i = 1

If the vapor is treated as an ideal gas, φ̂i = 1

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 473

Then

⎧
⎪ lim
⎡ P ⎛ mo ⎞ ⎤ ⎫ ⎪
Gi‡ − Gio = RT ln ⎨ mi →0 ⎢ ⎜ oi ⎟ ⎥ ⎬
⎪
⎩ ⎢⎣ m i ⎝ P ⎠ ⎥⎦ ⎭
⎪

Using the above results, the free energy of formation in solution is defined as

⎧
⎪ ⎡ fˆ ⎛ m o ⎞ ⎤ ⎫ ⎪
Δf Gi‡ (solution) = Δf Gio + RT ln ⎨
lim
mi →0 ⎢ i ⎜ oi ⎟ ⎥ ⎬
⎪
⎩ ⎢⎣ γ i m i ⎝ P ⎠ ⎥⎦ ⎪
⎭

where: Δf G oi = Gibbs energy of formation of species i as a perfect gas at the

standard state pressure

Δf G ‡i = Gibbs energy of formation of species i in a hypothetical ideal

solution of unit molality

This type of approach often is used for solutions of electrolytes in water, e.g.
salts, inorganic acids, etc.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 474

4.4.4.3 Gibbs Energy of Electrolyte Solution

General form: dGtotal = −S total dT +V total dP + ∑ µi dn i

i

Using the notation defined for electrolyte solutions changes this equation to

dGtot = −S tot dT +V tot dP + µu dmu + µ+dm + + µ−dm − + µo dno

The differential stoichiometry relations are

dm + = ν + (dm − dmu )

dm − = ν − (dm − dmu )

Then
dGtot = −S tot dT +V tot dP + [ µu − ν + µ+ − ν −m− ]dmu + [ν + µ+ + ν − µ− ]dm + µo dno

In the usual restriction to constant temperature and pressure, the first two terms
on the right hand side become zero, so that

dGtot = [ µu − ν + µ+ − ν − µ− ]dmu + [ν + µ+ + ν − µ− ]dm + µo dno

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 475

In a closed system, no ,m are constant, leading to

dGtot = [ µu − ν + µ+ − ν − µ− ]dmu

The only independent variable remaining is mu , therefore equilibrium occurs

when

⎡ ∂Gtotal ⎤
⎢ ⎥ =0
⎢⎣ ∂mu ⎥⎦T ,P,m,n
o

= µu − ν + µ+ − ν − µ−

that may be put in the form

∑ ν i Gi = 0
i

where ν i = 1 for undissociated electrolyte. Note that this form is exactly the same
as that for chemical equilibrium.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 476

 ⎛ γ i mi ⎞
For the form: G i = G ‡i + RT ln ⎜ o ⎟
⎝ mi ⎠

∑ ν i G i = 0 ⇒ Gu = ν +G+ + ν −G−
i

which become, upon substitution

Gu = ν +G+‡ + ν −G−‡ + RT ln ⎡⎣γ ν++ γ ν−− mν++ mν−− ⎤⎦

Nomenclature

ν = ν+ + ν−

γ ν± = γ ν++ γ ν−− mean ionic activity coefficient

mν± = mν++ mν−− mean ionic molality

Then

Gu = ν +G+‡ + ν −G−‡ + νRT ln[γ ± m ± ]

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 477

4.4.5 Equilibrium Involving Ionic Solutions

4.4.5.1 Vapor/Liquid Equilibrium

Vapor-liquid equilibrium for electrolyte solutions most often is of interest for

determining experimentally the activity coefficient of the solvent in the solution,
although some solutes (e.g., HCl) are sufficiently volatile that ion activity
coefficients may be determined in this way.

For the solvent,

fˆov = y oφ̂ov P ≅ P ⎫
⎪ P
⎬ ⇒ γ o = σ σ
fˆoL = x oγ ofoL ≅ x oγ oφoσ Poσ ⎪
⎭ x oφo Po

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 478

4.4.5.2 Solid/Liquid Phase Equilibrium

Solvent

The activity coefficient of the solvent also can be determined by measuring the
freezing point depression. For aqueous solutions, if the solid formed is pure ice
(usually the case for ice, the solvent fugacity is

fˆoL = x oγ ofoL ⎫⎪ 1
⎬ ⇒ γ o =
ˆf s = f s = f L ⎪ xo
o o o ⎭

Note that as freezing occurs, solvent is removed from the solution resulting in an
increasing solute concentration. Therefore, the above relation must be applied to
the conditions when the amount of ice goes to zero.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 479

Solute

Assume the solid to be pure electrolyte

Gus = Gus = GuL = ν +G+‡ + ν −G−‡ + νRT ln[γ ± m ± ]

where γ ν± mν± = γ ν± mν++ mν−− = solubility product (SP)

General case: SP = constant

Dilute solution: γ± ≅1 and mν++ mν−− ≅ S.P.

∂(lnSP ) Hu∞ − hus

Temperature dependence: =
∂T RT 2

where Hu∞ is the partial molar enthalpy at infinite dilution.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 480

4.4.5.3 Equilibrium Constants for Ionic Solutions

Dissociation reaction: Mν + Aν − = ν +Mz + + ν − Az −

Equilibrium condition: Gu = ν +G+ + ν −G−

Also:
Gu = Gu‡ + RT ln (mu γ u )
G+ = G+‡ + RT ln (m +γ + )
G− = G−‡ + RT ln (m − γ − )

Substitution and algebraic manipulations produce

⎡ (m γ )ν ⎤
ν +G+‡ + ν −G−‡ − Gu‡ = −RT ln ⎢ ± ± ⎥ = −RT lnK e
⎢⎣ muγ ⎥⎦
where
ν
Ke =
( m± γ ± )
=
γ ν±mν++ mν−−
= true dissociation constant
muγ u muγ u

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 481

Also

K e = K e (T ,P ) because the states denoted by ‡ are at infinite dilution at the

same temperature and pressure as the solution.

Then

Δ r G ‡ = −RT ln K e = ν + Δf G+‡ + ν − Δf G−‡ − Δf Gu‡

4.4.5.4 Apparent Dissociation Constant

The apparent dissociation constant is a convenient quantity because it does

not require knowledge of the activity coefficient.

mν++ mν−− mν±

Ka = =
mA mA

z+
for the dissociation reaction: Mν + Aν − = ν +M + ν − Az −

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 482

Example 1
H2CO3 → H + + HCO3−
HCO3− → H + + CO3=
Simplify notation for ease in use of subscripts
H2CO3 → A ⎫
⎪ ⎧Reaction 1: A → B + C
H+ → B ⎪ ⎪
⎬ ⇒ ⎨Reaction 2: C → B + D
HCO3− → C ⎪ ⎪
CO2= → D ⎪⎭ ⎩Reaction 3: A → 2B + D
2
γ Bγ C mB mC γ BC mB mC
Equilibrium for reaction 1: K1 = =
γ Am A γ Am A

γ Bγ D mB mD γ B2γ D mB mD γ BD
3
mB mD
Equilibrium for Reaction 2: K2 = = =
γ C mC γ Bγ C mC 2
γ BC mC
2 3
where γ BC = γ Bγ C , γ BD = γ B2γ D
γ B2 γ D mB2 mD γ BD
3
mB2 mD
Equilibrium for Reaction 3: K3 = =
γ A mA γ A mA
Note: K 3 = K 1K 2
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 483
Example 2:

2Fe +++ + Sn ++ → 2Fe ++ + Sn ++++

A B C D
Equilibrium constant

γ C2 γ D mC2 mD γ CD
3
mC2 mD
Ke = 2 2
= 3 2
γ Aγ B mAmB γ AB mAmB

Temperature Dependence of Ke

∂lnK e ν +H +∞ + ν −H −∞ − H ∞
=
∂T RT 2

4.4.5.5 Free Energies of Ions in Solution

The free energies of formation of individual ions cannot be calculated on an

absolute scale, instead they must be considered in neutral pairs.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 484

Example:
1
H2 ( g ) + O2 ( g ) → H2O ( ℓ ) Δf GHo O( ℓ)
2 2

H2O ( ℓ ) → H+ + OH− (aqueous ) Δf G ‡ + (aq )

H , OH−

Free energy of ions in solution: Δf GHo+ , OH− = Δf GHo O( ) + Δf G ‡+

2 H , OH- (aq )

Example:
1
H2 ( g ) + O2 ( g ) → H2O (  ) o
Δf GH
2 2O(  )

3 1
H2 ( g ) + N2 ( g ) → NH3 (aq ) ◊
Δf GNH
2 2 3 (aq )

NH3 (aq ) + H2O (  ) → NH+4 + OH- Δf G ‡

NH+4 , OH- (aq )

Free energy of ions in solution:

o ‡ ◊ o
Δf GNH = Δ G − Δf GNH − Δ G
, OH (aq ) f OH (aq ) 3 (aq ) f H2O(  )
+ -
4 NH+4 , -

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 485

Hydrogen Ion Convention

To allow tabulation of Δf G ◊ , adopt the convention that:

Δf GHo+ = 0 for H + at unit molality at all temperatures

Also
Δf HHo+ = 0 = Δf SHo+

Galvanic Cells

DG = work done at constant T,P in a reversible reaction

= -zFE

where

z = number of Faradays passing through the cell

F = Faraday's constant = 96,500 coulombs/mole eq.

E = electromotive force (volts)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 486

Also ΔG = ∑ ν i G i → ∑ ν i G i = −zFE

where: G i = ν +G+‡ + ν −G−‡ + νRT ln[γ ± m ± ]

Examples:
1
Platinum electrode: H → H+ + e-
2
1
Mercury electrode: Hg2Cl2 + e- → Hg (  ) + Cl-
2
1 1
Total reaction: H2 + Hg2Cl2 → HCl + Hg (  )
2 2

Then
1 1
−FE = GHCl + GHg − G H 2 − G Hg 2Cl2
2 2

GHCl = G ‡+ + G ‡ - + 2RT ln(γ ±m ± )

H Cl

(The previous equation assumes that HCl is a strong electrolyte, i.e., that
essentially all HCl is dissociated.)

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 487

Because the vapor and liquid are in equilibrium, the partial molar Gibbs energy
for hydrogen in the solution is

GH2 = GHo2 + RT lnfˆH2 ≈ GHo2 + RT lnPH2 (at low pressures )

Because neither mercury nor mercurous chloride dissolve in water to a significant
extent
L
GHg = GHg
S
GHg2Cl2 = GHg2Cl2

Then
1 s 1 o

−FE = G ‡+ + G ‡ - + GHg − GHg − GH2 − RT ⎡ln PH2 − ln (γ ±m ± ) ⎤
H Cl 2 2 Cl2
2 ⎣ ⎦
or
⎛γ m ⎞
−FE = −FE + RT ln ⎜ ± ± ⎟
o
⎜⎝ PH ⎟⎠
2

 1 s 1 o
where: −FE o = G ‡+ + G ‡ - + GHg − GHg2Cl2 − GH = Δr G ‡
H Cl 2 2 2

E o = emf of an ideal solution of unit molality

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 488

Also
2RT lnm + FE = FE o − 2RT lnγ ±
(m, E are measurable [observable] variables)

Note
lim
m→0 γ± =1

lim
m→0 ⎡⎣2RT ln (m ) + F E ⎤⎦ = F E o

Therefore, E o can be obtained by extrapolating the measurements to infinite

dilution.

Again, the electrode potentials come in pairs, so the electrode potential

associated with the hydrogen ion is defined to be zero
1
Eo = 0 for H2 → H+ (aq ) + e-
2

Using this definition, various reactions can be used to generate tables of single
electrode potentials for various ions.

Note: Δf G ‡ = −zFE o

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 489

4.4.5.6 Activity Coefficient Models for Electrolyte Solutions

4.4.5.6.1 Ionic Strength

The ionic strength of the solution is

1
Ι= ∑ mi z i2
2 all ions
in solution

4.4.5.6.2 Debye Limit Law

Debye and Huckel derived a theoretical expression for the behavior of the ionic
activity coefficient in the infinite dilution limit:

lnγ ± = −1.174 z + z − Ι

In this treatment, the solvent is treated as a structureless background continuum

with a fixed permittivity (e) and dielectric strength (D), and the interactions
between electrically charged particles are used to calculate the contribution of an
individual ion to the solution.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 490

4.4.5.6.3 Extended Debye-Huckel Theory

The Debye-Huckel result can be extended to higher concentrations in the form

Ι
lnγ ± = −1.174 z + z −
1+ Ι

4.4.5.6.4 Pitzer Model

Pitzer combined intermolecular interactions and electrical charge interactions to

express the excess Gibbs energy in the form

GE 1 1
= w f ( Ι ) + ∑ ∑ λik ( Ι ) ni nk + 2 ∑ ∑ ∑ λik ni nk n
RT w i k w i k 

where w is the mass of water present, and the l's are osmotic virial
coefficients. The second osmotic virial coefficient is treated as a function of
ionic strength, while the third osmotic virial is assumed not to be affected by ionic
strength.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 491

The activity coefficient expression then is

2 2 (ν +ν − )
⎡ 2 ν +ν − ⎤ γ ⎡ 3/ 2 ⎤
γ γ
lnγ ± = z +z − f + m ⎢ B + m ⎢ ⎥ C
⎣ ν ⎥⎦
± ±
⎢⎣ ν ⎥⎦

where:

⎡ ⎤
f γ = −3Aφ ⎢
Ι 2
+ ln 1+ b Ι ⎥
⎣ 1+ b Ι b ⎦
( )
3 φ
C ±γ = C
2 ±
3/ 2
1 ⎡ e2 ⎤
Aφ = 2π (1000 )N A ρs ⎢ ⎥
3 ⎢⎣ 4πε oDs kT ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 492

For a single 1:i or j:1 electrolyte

B ±γ = 2 βo +
2 β1 ⎡
2 ⎣
α Ι
1− 1+ α Ι −(0.5α 2
) (
I exp −α Ι ⎤
⎦ )
α = 2.0 (kg/mol)
1/2

β = 1.2 (kg/mol)
1/ 2

For a single 2:2 electrolyte

2 βi ⎡
( ) ( )
2
B ±γ = 2 βo + ∑ 2 ⎣
1− 1+ α Ι − 0.5α 2
Ι exp −α i Ι ⎤
i =1 α i Ι
i i ⎦

α1 = 1.4 (kg/mol)
1/2

α 2 = 12 (kg/mol)
1/2

βo , β1 and C ±γ are adjustable parameters that are determined by fitting

experimental data.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 493

4.4.5.6.5 Meissner Model

Meissner and co-workers developed a corresponding states model for electrolyte

solutions that provides activity coefficients for single electrolytes in solution using
a reduced ionic activity coefficient

Γ oij = (γ ± )
1/ z + z −

The corresponding states prediction is made using a single Meissner parameter

(q ) for each aqueous electrolyte.
o
ik

= 1+ B (1+ 0.1Ι ) − B ⎤ ΓDH

⎡ qiko
Γ oik
⎢⎣ ⎥⎦ ik
Ι
lnΓDH
ik = −1.174
1+ C Ι
o
B = 0.75 − 0.065qik

C = 1+ 0.055qiko exp −0.023Ι 3 ( )

⎡ T −Tref ⎤
o
q ik (T ) = o
qik (Tref ) ⎢1− z z ⎥
⎣ + − ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 494

Values for the Meissner parameters are given for several electrolytes in the
following table.

4.4.5.6.5.1 Meissner Parameters for Selected Electrolytes at 25 °C

Electrolyte q oik Electrolyte q oik

HCl 6.69 FeCl2 2.16

LiCl 5.62 CaCl2 2.40
NaCl 2.23 MgSO4 0.15
KCl 0.92 CuSO4 0.00
NH4NO3 -1.15 AlCl3 1.92
Na2SO 4 -0.19 Al2 (SO4 )3 0.36
K2SO4 -0.25

A more complete table is available in Kusik & Meissner, "Electrolyte activity

coefficients in inorganic processing," in Fundamentals of Hydrometallurgical
Processing, AIChE Symposium Series, pp. 14 - 20 (1978).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 495

 5 SURFACE EFFECTS
5.1 SURFACE ENERGIES

In the previous treatments of multiple phases coexisting in equilibrium, the

molecules in a given phase all are assumed to have the same chemical potential,
including those at the phase boundary. In addition, the phase boundary is taken
to be infinitely thin, with a discontinuous transition from one equilibrium phase to
the other. In reality, the environment for a molecule near the phase boundary
differs significantly from that of a molecule far from the boundary.

Consider the edge of a close packed crystal structure

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 496

A molecule distant from the surface is completely surrounded by other
molecules. The diagram below shows first and second nearest neighbors for a
molecule in the bulk and at the surface where "1" are nearest neighbors and "2"
are second nearest neighbors.

2 1 2
1 1 1 1
2 1 2 2 1 2

Bulk Surface

For the two cases, the contributions of the intermolecular interactions between
the molecules to the energy of the system are

1
E bulk =
2
( 4ε 1 + 4ε 2 ) < 0

1
E surface =
2
( 3ε1 + 2ε 2 ) < 0

Here ε1 is the average interaction energy for nearest neighbors, and ε 2 is the
average interaction energy for second nearest neighbors. (The factor of 1/2
avoids double counting).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 497

For normal pressures, both ε1 and ε 2 usually are negative. Then, the difference
between a surface molecule and an interior molecule is

1 1
δ = E surface − E bulk = ⎡⎣ 3ε1 + 2ε 2 − ( 4ε1 + 4ε 2 ) ⎤⎦ = − ( ε1 + 2ε 2 ) ⇒ δ > 0
2 2

The net result is that a molecule in the surface layer contributes a negative
contribution of smaller magnitude to the energy than does a molecule in the
interior.

As a result, the total energy differs from the result calculated by assuming that all
molecules have the same energy as an interior molecule by a small positive
amount

E total = ∑ ni µi + nsurface δ
i

Here nsurface is the number of moles that differ from the bulk, and δ is the
average energy deviation. Clearly, nsurface will vary directly with the total surface
area.

A similar situation applies to liquid-vapor interfaces, although the neighbor

interactions are less well-defined than in the case of the crystal structure.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 498

Furthermore, at the interface, the density varies shown below, where the change
is rapid, but not discontinuous.

Position

Although the change is not purely discontinuous, it occurs over a distance of a

few nanometers, except for very near the vapor-liquid critical point.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 499

5.1.1 Surface Tension

It is most convenient to use the total area of the surface as the independent
variable to describe this effect, and the variation of energy with area is the
surface tension, which is defined thermodynamically by

⎛ ∂(nG ) ⎞ ⎛ ∂(nA ) ⎞ ⎛ ∂(nH ) ⎞ ⎛ ∂(nU ) ⎞

σ =⎜ = = =
⎝ ∂a ⎟⎠ T ,P,n ⎜⎝ ∂a ⎟⎠ T ,nV ,n ⎜⎝ ∂a ⎟⎠ nS,P,n ⎜⎝ ∂a ⎟⎠ nS,nV ,n
k k k k

where s is the surface tension, and a is the surface area.

For the case where surface effects are significant, the physics summary
statement becomes

d (nU ) = Td (nS ) − Pd (nV ) + σ da + ∑ µi dni

i

and the reversible work is described by

/ rev = −P dV + σ da
dW

where: s = surface tension

da = differential change in surface area

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 500

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 501
5.2 APPLICATION OF FIRST AND SECOND LAWS

Consider a phase b that is completely surrounded by a phase a where the two

phases are separated by a surface with surface tension s.

β
β β

α - phase β

Use an analysis similar to that used previously to describe the conventional

equilibrium situation.

We consider the a and b phases separately.

The surface contribution normally is included in the phase that is surrounded

completely by the surface (which in this case is b).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 502

Then:

( ) ( ) ( )
d nαU α = T α d nα S α − P α d nαV α + ∑ µiα dniα

d (n βU β ) = T β d (n β S β ) − P β d (n βV β ) + σ d (a β ) + ∑ µ β dn β i i

Solve each equation for the differential total entropy:

(
d nα S α = ) T
1 ⎡
α ⎣ ( ) (
d nαU α + P α d nαV α − ∑ µiα dniα ⎤
⎦ )
(
d n βS β = ) 1 ⎡
Tβ ⎣
d n β β
U (
+ P β
d n β β
V − )
σ da β
− (
∑ µi
β
)
dn β⎤
i ⎦

For the case where a and b together form a closed isolated system:

nαU α + n βU β = constant ⇒ d nαU α = −d n βU β ( ) ( )

nαV α + n βV β = constant ⇒ d (nαV α ) = −d (n βV β )

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 503

Then, the total entropy change is given by

( ) ( )
d (nS ) = d nα S α + d n β S β

⎡ µα µ β ⎤ α
= ⎢ α − β ⎥ d (n U ) − ∑ ⎢ α − β ⎥dn
⎡ 1 1 ⎤ α α i i
i
⎣T T ⎦ ⎢T T ⎥ ⎣ ⎦
⎡ α β ⎛ β ⎞⎤
⎢T
P P
T
σ
+⎢ α − β + β ⎜
da
(
T ⎜⎝ d n βV β
(
⎟ ⎥ d n βV β
)
⎟⎠ ⎥
)
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 504

5.3 EQUILIBRIUM CONDITIONS

Because the total entropy of the universe is a maximum at equilibrium, we must

have d(nS) = 0.

(nαU α ), (nαV α ) and {n } all are independent variables.

α
i

Therefore all terms in square brackets must be identically equal to zero, and the
equilibrium conditions are:

T α =T β
µiα = µiβ

α da β
β
P = P −σ
d (n βV β )

The equilibrium conditions are the same as those derived for the case without
surface effects, except that the pressures are not equal in the coexisting phases.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 505

5.4 PHASE EQUILIBRIUM

5.4.1 Case I: Small liquid drops in a vapor (pure fluid)

Here the a phase is vapor, and the b phase liquid

µiα = µiβ ⇒ f iα = f iβ

Then: f iα = φ αi P α

f iβ = f iβ P β( ) where P β ≠ Pα

⎡ 1 Pβ ⎤
and: f iβ σ σ
= φ i Pi exp ⎢ β
∫ V dP ⎥
⎢⎣ RT Piσ ⎥⎦

The equilibrium condition leads to:

φ σi Piσ ⎡ 1 Pβ β ⎤ ⎡V β σ ⎤
Pα =
φ αi
exp ⎢
⎢⎣ RT
∫
Pi σ
V dP ⎥
⎥⎦
≅ P σ
i exp ⎢
⎢⎣ RT
P (
β
− Pi ⎥)
⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 506

Use the equilibrium relation to eliminate Pb from the previous equation:

⎡ β ⎛ ⎞⎤
da β V da β
P β = Pα + σ ⇒ P α = Piσ exp ⎢ ⎜Pα + σ − σ ⎟⎥

( ) ( )
P
d n V β β ⎢ RT ⎜⎝ d n βV β
i
⎟⎠ ⎥
⎣ ⎦

This result is nonlinear; it must be solved through an iterative solution.

A good approximation usually is provided by:

α da β
− Piσ << σ
( )
P
d n βV β

so that

⎡ β ⎛ ⎞⎤
σV da β
α
≅ Piσ exp ⎢ ⎜ ⎟⎥
( )
P
⎢ RT ⎜⎝ d n βV β ⎟⎠ ⎥
⎣ ⎦

This result also provides a good initial guess for the iterative solution of the
general equation.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 507

5.4.1.1 Example: Spherical Geometry - Liquid drop

For spherical drops of radius r

β 2 da β
a = 4π r ⇒ = 8π r
dr

β β 4 3
n V = πr ⇒
d n β β
V
= 4π r 2
( )
3 dr

Combine these relations to obtain

da β ⎛ da β ⎞ ⎛ dr
⎞
8π r 2
=⎜ ⎜ ⎟= =
(
d n βV β ) ⎟
⎝ dr ⎠ ⎜⎝ d n βV β ( ) ⎟⎠ 4π r 2 r

so that

α ⎡ σV β 2 ⎤ ⎡ 2σV β ⎛ 1⎞ ⎤
P ≅ Piσ exp ⎢ σ
⎥ = Pi exp ⎢ ⎜⎝ ⎟⎠ ⎥
⎢⎣ RT r ⎥⎦ ⎢⎣ RT r ⎥⎦

and
2σ
Pα = P β −
r
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 508
5.4.2 Case II: Small Vapor Bubbles Surrounded By a Liquid (Pure
Fluid)

The derivation for this case is similar to that of Case I, except that here the b
phase is a vapor and the a phase is a liquid. Then

⎡ 1 Pα ⎤
f iα σ σ
= φ i Pi exp ⎢ α
∫ V dP ⎥
⎢⎣ RT Piσ ⎥⎦

f iβ = φi P β

The equilibrium relation leads to

φ σi ⎡
P β = β P iσ exp ⎢
1 Pα
∫ V α
dP
⎤
⎥ ≅ P σ
exp ⎢
(
⎡V α P α − P σ
i ) ⎤⎥
i
φi ⎢⎣ RT Piσ ⎥⎦ ⎢ RT ⎥
⎣ ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 509

Substitution of the relation between Pa and Pb leads to

⎡ α ⎛ ⎞⎤
V da β
P β ≅ Piσ exp ⎢ β σ
⎜ P − Pi − σ ⎟⎥
⎢ RT
⎣
⎜⎝ d n βV β ( ) ⎟⎠ ⎥
⎦

and, in the case where the surface term dominates:

⎡ ⎛ ⎞⎤
⎢ σV α da β
β σ
P ≅ Pi exp − ⎜ ⎟⎥
⎢ RT
⎣ (
⎜⎝ d n βV β ) ⎟⎠ ⎥
⎦
Note that the form is the same for Cases I and II, but that the sign of the
exponential argument is different. The pressure in a liquid drop is larger than the
equilibrium vapor pressure, while the pressure in a bubble is smaller than the
equilibrium vapor pressure.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 510

5.4.2.1 Example: Spherical geometry - Vapor bubble

da β 2
=
( )
As in the example for Case I,
d n βV β r
Then
β ⎡ 2σV α ⎛ 1⎞ ⎤
P ≅ Piσ exp ⎢ − ⎜⎝ ⎟⎠ ⎥
⎢⎣ RT r ⎥⎦
and
2σ
Pα = P β −
r

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 511

Correlations for Surface Tension

Pitzer and Curl developed a correlation for surface tensions that requires
knowledge of the liquid molar volume at one temperature, the critical temperature
and the acentric factor.

σ
= (1−Tr ) 9
11

σ0

−1
2/ 3 ⎡ ⎛ 3 ⎞ 2/3 ⎤
σ oVo ⎢ dyne cm 1⎥
= 1.86 + 1.18 ω
Tc ⎢ cm ⎜⎝ mol ⎟⎠ K ⎥
⎣ ⎦

1+ 0.85 (1−Tr ) + (1−Tr ) (1.89 + 0.91ω )

1/ 3
Vo = V (Tr )
3.74 + 0.91ω

This correlation generally provides values within 5% for "normal" fluids. It does
not provide accurate estimates for hydrogen-bonding fluids such as water,
alcohol, etc.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 512

5.5 ADSORPTION ON SURFACES

The attractive forces arising from interactions with the molecules of a solid
surface leads to an increased density of the fluid near the wall, as shown
schematically below.

Region of Adsorption

Density

Potential Energy

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 513

Even at supercritical temperatures, there is a significant enhancement of density
near the surface, and at lower temperatures, the density may approach liquid -
like densities.

Because the adsorption region is associated with the surface, and it is only a few
nanometers in thickness, it is convenient to treat this region as a separate phase
that exists in two dimensions rather than three.

5.5.1 Terminology

Adsorption refers to the capturing of gas molecules by a solid surface. There are
two classes of adsorption: chemisorption and physical adsorption. In the case of
chemisorption, the molecules are bound to specific sites on the surface, as
shown below, due to special chemical properties associated with the site. In this
case the molecules have little mobility along the surface

Solid Surface

Chemisorption

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 514

In the case of physical adsorption, the molecules are held near the surface of
the solid by the attractive forces from the solid molecules. in this case, the
molecules have almost completely free mobility along the surface.

Solid Surface

Physical Adsorption

This gives rise to fluid behavior in two dimensions rather than three. Such fluids
exhibit phase transitions between vapor and liquid and between liquids and
solids as the surface temperature is lowered. Depending upon how much gas is
adsorbed on the surface, the adsorption falls into one of the following classes:

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 515

The case where the surface is only partially covered with adsorbed molecules is
referred to as submonolayer adsorption.

Solid Surface

Solid Surface
Submonolayer Adsorption

Monolayer adsorption refers to the case the

Submonolayer entire surface is covered by a
Adsorption
single layer of adsorbed molecules.

Solid Surface

Solid Surface
Monolayer Adsorption

Monolayer Adsorption

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 516

 Solid Surface

Multilayered adsorption refers to the case

Monolayer when additional layers are adsorbed
Adsorption
onto the initial monolayer of molecules.

Solid Surface

Multilayer Adsorption

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 517

The approximate amount of substance adsorbed per square meter of surface for
monolayer coverage may be estimated by assuming the molecules are located
on a square grid with an average spacing of about 3 Å. Because 1 Å = 10-10 m,
the total number of molecules adsorbed is (3´109)2 » 1019 molecules. This
represents less than 10-4 moles, so very large surfaces are required to adsorb
significant amounts of substance from the vapor.

Special structures are required to obtain adequate surface areas. Materials such
as activated charcoal, zeolites, etc. can provide 2000 - 3000 m2 of surface area
per gram of material.

Adsorption phenomena are used for practical applications such as removal of

contaminants from air streams, long-term maintenance of high vacuum in
evacuated systems, catalysis, separations processes, etc.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 518

5.5.2 Two Dimensional Phase

The internal energy of a two dimensional material is given by

d (ηU ) = Td (ηS ) + π d (ηΘ ) + ∑ µi d ηi

i

where

η= total number of moles in the two dimensional material

a
Θ= = molar area (two dimensional analog of molar volume
η

⎛ ∂(ηU ) ⎞ ⎛ ∂(ηA ) ⎞
π =⎜ =⎜ = spreading pressure
⎝ ∂(ηΘ ) ⎟⎠ ηS,ηk
⎝ ∂(ηΘ ) ⎟⎠ T ,ηk

Legendre transformations produce

d (ηA ) = −ηSdT + π d (ηΘ ) + ∑ µi d ηi

i

d (ηG ) = −ηSdT − (ηΘ ) d π + ∑ µi d ηi

i
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 519
The two dimensional analog of the Gibbs-Duhem equation is

ηSdT − (ηΘ ) d π + ∑ ηi d µis = 0

i

which, for the isothermal case becomes

(ηΘ ) d π = ∑ ηi d µis = ∑ ηi dG is
i i

The equilibrium condition of equality of chemical potential in coexisting phases

gives
d µ s = d µ b = dG b = RTd lnfˆb =V dP
i i i i i

Substitution into the two dimensional Gibbs-Duhem gives

(ηΘ ) d π = RT ∑ ηi d lnfˆib = RT ∑ ηi d ln( y i φ̂ibP )

i i
= RT ∑ ηi d lnP + RT ∑ ηi d lny i + RT ∑ ηi d lnφ̂ib
i i i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 520

or
dP
(ηΘ ) d π = ∑ ηi V ibdP = RT ∑ ηi Z bi
i i P
Then
P
π ⌠ dP
(ηΘ ) ⎮dπ
⌠
⌡ = RT ∑ ⎮⎮ ηi (P ) Z bi (P )
0 i ⌡ P
0

P
RT ⌠ dP
so that π (P ) = ∑ ⎮⎮ ηi (P ) Z bi (P )
(ηΘ ) i ⌡ 0
P

P
RT ⌠ dP
For a pure fluid, the previous result reduces to π (P ) = ⎮ η i (P ) Z i (P )
b

(ηΘ ) ⎮⌡
0
P

Information about the adsorption isotherm (the variation of h with P) is required to

calculate the spreading pressure, which is of interest because it provides
information about surface energies.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 521

5.5.3 Adsorption Isotherms

Some commonly used descriptions are

5.5.3.1 Henrys Law

η (P ) = kP

This result applies in the low pressure limit, where the surface density of
adsorbed molecules is low.

The corresponding spreading pressure is

RT ⎡ BP ⎤ RT ⎡ BP ⎤ RT
π HL = 1+ ( k P ) = 1+ ( k P ) ≅ (k P )
(ηΘ ) ⎢⎣ RT ⎥⎦ a ⎢⎣ RT ⎥⎦ a

where the compressibility factor is truncated after the second term

BP
Z bi = 1+
RT

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 522

5.5.3.2 Langmuir isotherm

bkP
η (P ) =
b +P

The Langmuir isotherm is limited to monolayer adsorption.

The spreading pressure calculation is
P P
RT ⌠ dP RT ⌠ b k P ⎛ BP ⎞ dP
⎮ η (P ) Z i (P )
b
π Langmuir = = ⎮
⎜ 1+ ⎟
(ηΘ ) ⎮⌡ P a ⎮
⌡ b + P ⎝ RT ⎠ P
0 0
⎡P P ⎤
RT ⌠ dP B ⌠ P
= b k ⎢ ⎮⎮ + ⎮ dP ⎥
a ⎢ ⌡ b + P RT ⎮
⌡ b +P ⎥
⎣0 0 ⎦
RT ⎡ ⎛ P⎞ B ⎧ ⎛ P ⎞ ⎫⎤
= b k ⎢ln ⎜ 1+ ⎟ + ⎨P − b ln ⎜⎝ 1+ ⎟⎠ ⎬ ⎥
a ⎣ ⎝ b ⎠ RT ⎩ b ⎭⎦
RT ⎡⎛ bB ⎞ ⎛ P ⎞ BP ⎤ RT ⎛ P⎞
= b k ⎢⎜ 1− ⎟⎠ ⎜⎝
ln 1+ ⎟⎠ + ⎥ ≅ b k ln ⎜⎝ 1+ ⎟⎠
a ⎣ ⎝ RT b RT ⎦ a b

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 523

5.5.3.3 Toth Equation

mP
η (P ) =
(b + P ) t 1/t

(The Toth equation is an empirical model. Note that it becomes ill-behaved at

low pressures if t < 1.)

5.5.4 Heat of Adsorption

Because the environment of the adsorbed molecules differs significantly from

that of the molecules in the bulk vapor phase, there are significant changes in
energy associated with the adsorption process. The form of the Clapeyron
equation that applies to this case is

⎛ ∂P ⎞ HV − H ads Δav H
= =
⎜⎝ ⎟
∂T ⎠ n ads T V V −V ads (T ΔavV )
At low pressures, where the vapor phase may be considered ideal, and the
volume of the vapor is much greater than the volume of the adsorbate, the
Clapeyron equation becomes:
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 524
 ⎛ ∂lnP ⎞ Δav H
⎜⎝ ⎟ =
∂T ⎠ nads RT 2

The heat of adsorption can be calculated from isotherm data at several

temperatures, or it can be measured directly by microcalorimetry techniques.

Heats of adsorption for chemisorption are of approximately the same magnitude

as heats of reaction (60 - 170 kJ/mol), while those for physical adsorption usually
are comparable to heats of vaporization (10 - 50 kJ/mol).

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 525

 6 EXTERNAL FIELDS
In this section, we shall consider the effects of external fields such as gravity,
electricity and magnetism, and the centrifugal forces that exist in rotating
systems.

The expression for the total internal energy of a collection of particles that was
derived in the discussion of the First Law of Thermodynamics (energy balance) is
! !
total
Usystem = 21 ∑ mi Ui • Ui (motion: translation, rotation, vibration)
i
!
ri
! ! !
1
i k r!
⌠
+ 2 ∑ ∑ ⎮⌡ fik Ri − Rk ( ) !
• d ri (intermolecular interactions )
ref
!
ri
!
⌠ ext !
+ ∑ ⎮⌡ F i ri • d ri
i r!
!
( ) (external field interactions )
cm

In the previous treatments, the effect of the external field has been assumed to
be sufficiently small that the third term can be taken as zero.

The third term involves a summation over all molecules present, so it is not
tractable for calculations in this form.
CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 526
In the third term, the integral is the potential energy of particle i relative to the
position of the center of mass that is due to the external field. To simplify the
notation, X will denote the potential energy contribution resulting from the
interaction of one mole of substance i with the external field
!
!
ri
! ext ! !
Ξi ( ri ) = ⌠
⎮
!⌡
( )
Fi ri • d ri
rcm

To convert the summation to an integral, we choose a differential volume

element, dJ that is sufficiently small that X has essentially the same value
throughout, but sufficiently large that the molecules within can be treated as a
continuum. The number of moles of substance i contained within the differential
volume element is
!
niδV = ⎡⎣c i ( r ) ⎤⎦ d ϑ

where ci is the concentration (moles per volume) of substance i, and r is the

molar density. All of these variables may vary with position.

The choice of which coordinate system to use usually is based on the symmetry
of the physical situations.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 527

The contribution to the potential energy of the interaction of substance i with the
external field within the differential volume element is

Ξδi V = niδV Ξi

and the contribution of substance i to the internal energy of the system then is

ni Ξisys = ∫system niδV Ξδi V d ϑ = ∫system c i Ξi d ϑ

and the total internal energy due to the presence of the field is

nΞ = ∑ ni Ξi
all
substances

Also, the total number of moles of substance i in the system is

!
ni = ∫system c i (r ) d ϑ

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 528

The general versions of Physics Summary statements become

( )
d (nU ) = Td (nS ) − Pd (nV ) + ∑ ni dΞi + ∑ µinf + Ξi dn i
i i

d (nH ) = Td (nS ) + (nV ) dP + ∑ n dΞ + ∑ ( µ + Ξ ) dn i i

nf
i i i
i i

d (nA ) = − (nS ) dT − Pd (nV ) + ∑ n dΞ + ∑ ( µ + Ξ ) dn i i

nf
i i i
i i

d (nG ) = − (nS ) dT + (nV ) dP + ∑ n dΞ + ∑ ( µ + Ξ ) dn i i

nf
i i i
i i

The reversible work associated with imposing an external field is

d (nW )rev = ∑ ni dΞi

field

The chemical potential of substance i in the mixture with the field present is the
coefficient of dni in the physics summary statements:
µifield = µino field + Ξi

In terms of the partial molar Gibbs energy

Gifield = Gino field + Ξi

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 529

Fugacity

The fugacity of substance i in the mixture is given by

fˆi ⎡Gi − Gi† ⎤ ⎡ µi − µi† ⎤

†
= exp ⎢ ⎥ = exp ⎢ ⎥
fi ⎣ RT ⎦ ⎣ RT ⎦

fˆi field ⎡Gifield − Gi† ⎤

With field present: = exp ⎢ ⎥
fi † ⎣ RT ⎦

fˆi no field ⎡Gino field − Gi† ⎤

With no field present: †
= exp ⎢ ⎥
fi ⎣ RT ⎦

⎡Gifield − Gino field ⎤ ˆ no field ⎡ Ξi ⎤

Combine to obtain: fˆi field
= fˆi no field
exp ⎢ =
⎥ i f exp ⎢⎣ RT ⎥⎦
⎣ RT ⎦
! !
Equilibrium condition: ˆ
∇fi = 0

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 530

 ! field ! no field ⎡ Ξ ⎤ ˆno field ! ⎧ ⎡ Ξ ⎤⎫
ˆ
∇fi = ∇fi ˆ exp ⎢ + f ∇ ⎨ exp ⎬
⎣ RT ⎥
⎦
i
⎩ ⎢
⎣ RT ⎥
⎦⎭
! no field ⎡ Ξ ⎤ ˆno field ⎡ Ξ ⎤!⎡ Ξ ⎤
ˆ
= ∇fi exp ⎢ + fi exp ⎢ ∇⎢
⎣ RT ⎥
⎦ ⎣ RT ⎥
⎦ ⎣ RT ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 531

6.1 Gravitational Fields

The force of gravity on a molecule is

!
! grav ! ! ! rik ! !
Fik (rik ) = M i g ⇒ Ξik (rik ) = !∫ M i g i drik
rcm

⎛ ⎞ ! ! ! ! ! ! ! !
so that ∑ i i
n dΞ =
⎝
∑ i i
n M
⎠
g i dr = ( )
nM g i dr = mg i dr = ρ m( nV ) i dr
g
i i

The physics summary statements become

! !
d (nU ) = Td (nS ) − Pd (nV ) + ρm (nV ) g i dr + ∑ ( µi + Ξi ) dn i
i
! !
d (nH ) = Td (nS ) + (nV ) dP + ρm (nV ) g i dr + ∑ ( µi + Ξi ) dn i
i
! !
d (nA ) = − (nS ) dT − Pd (nV ) + ρm (nV ) g i dr + ∑ ( µi + Ξi ) dn i
i
! !
d (nG ) = − (nS ) dT + (nV ) dP + ρm (nV ) g i dr + ∑ ( µi + Ξi ) dn i
i

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 532

6.1.1 Hydrostatic Pressure Gradient

Application of Math Relation 6 to the Helmholtz energy expression produces

!
⎛ ∂P ⎞ ⎛ ∂( ρm (nV ) g ) ⎞
⎜⎝ ! ⎟⎠ = −⎜
∂r nT ,n ⎝ ∂(nV ) ⎟⎠
nT ,n

so that

! ! ∑ xi M i !
∇P = − ρm g = − i g
V

The pressure gradient lies in the direction of the gravitational field.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 533

6.1.2 Chemical Potential Gradient

Application of Math Relation 6 to the Physics Summary Statement produces

⎛ ∂ µifield ⎞ " field ∂ ⎡⎛ ⎞ g" ⎤ = M g"

!
⎜⎝ ∂r ⎟⎠ = ∇µ = ∑ n M
∂ni ⎢⎣⎝ k ⎠ ⎥⎦
i k k i

Again, the gradient lies along the direction of the external field. The simplest
treatment is to define a coordinate axis z in either cylindrical or Cartesian
coordinates that lies in the direction of the gravitational field. Then

Ξi = M i g ( z − z o )

and

⎡M g (z − z o ) ⎤
fˆi field = fˆi no field exp ⎢ i ⎥
⎣ RT ⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 534

6.2 Centrifugal Force Fields

Rotational motion introduces a centrifugal force of magnitude mw2r. that lies in

the direction of the radial coordinate in a cylindrical coordinate system where the
axis of rotation defins the z axis. The situation is analogous to the gravitational
case, except that the acceleration due to gravity, g, is replaced in all equations
by the centrifugal acceleration, w2r.

6.2.1 Centrifugal Pressure Gradient

The pressure gradient induced by rotational motion

(
⎛ ∂P ⎞ ⎛ ∂ mω r ⎞ ∑i x i M i 2
2
)
⎜⎝ ∂ r ⎟⎠ = ⎜ ∂(nV ) ⎟ = V ω r
⎝ ⎠

where w is the angular frequency of rotation.

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 535

6.2.2 Centrifugal Chemical Potential Gradient

The gradient in chemical potential induced by rotational motion is

⎛ ∂ µifield ⎞ ∂ ⎡⎛ ⎞rω 2 ⎤ = −M rω 2
⎜⎝ ∂r ⎟⎠ = − ∑ n M
∂ni ⎢⎣⎝ k ⎠ ⎥⎦
k k i

Again, the gradient lies along the direction of the external field. The simplest
treatment is to define a coordinate axis z in either cylindrical or Cartesian
coordinates that lies in the direction of the gravitational field. Then

1
Ξi = − M i ω 2 r 2 − ro2
2
( )
and

fˆi field
= −fˆi no field
exp ⎢
(
⎡ M i ω 2 r 2 − ro2 ) ⎤⎥
⎢⎣ 2RT ⎥⎦

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 536

6.3 Electric Fields

For applied electric fields, the reversible work term is

   
/ rev
dW electric
= ⌠
⎮
⌡ (E • dD dV = ) ⌠
⎮
⌡ ( ( ))
E • d εE dV
Volume Volume

6.4 Magnetic Fields

For applied magnetic fields, the reversible work term is

  
( ( ))

/ rev
dW magnetic
= ⌠
⎮
⌡ (
H • dB dV = ) ⌠
⎮
⌡
H • d µ H dV
Volume Volume

CHEN 623 Lecture Notes (Date of last revision: 08/30/2021) 537