Metallurgical Thermodynamics - Ghosh

•
Textbook of Materials and

Metallurgical Thermodynrunics
AIIlNDRA GHOSH
Formerly Professor
Materials and Metallurgical Engineering
Indian Institute of Technology Kanpur
PHI Learning [;)LJ1WOO llikmBGGGJ

Delhi-110092
201 5
t 250.00
TEXTBOOK OF MATERIALS AND METALLURGICAL THERMODYNAMICS

Ahindra Ghosh
C 2003 by PHI Leaming Private Limited, Delhi. All rights reserved. No part of this book may be
reproduced in any form , by mimeograph or any other means, without pennission in writing from
the publisher.
ISBN-978-81 -203-2091-8
The export rights of this book are vested solely with the publisher.
Seventh Printing September, 2015
Published by Asoke K. Ghosh, PHI Leaming Private Limited, Rimjhim House, 111 , Patparganj
Industrial Estate, Delhi-110092 and Printed by V.K. Batra at Peart Offset Press Private Limited,
New Delhi-110015.
ThxTBooK OF ivlATERIAI.S AND
METAILURGICAL 'fHERMODYNAMia>
To
(Late) Shri Sajoy Panchanan Ghosh
(Late) Shri Gobinda Lal Sarkar
Shrimati Binapani Biswas
Contents
Preface xv
List of Symbols with Units xvii
1. INTRODUCTION 1-8
1.1 Historical Perspective 1
1.2 History of Classical Thermodynamics 2
1.2.1 General 2
1.2.2 Chemical Thermodynamics 2
1.2.3 MetaJlurgical Thermodynamics 3
1.3 Introductory Concepts and Definitions 3
1.3.1 System and Surrounding 4
1.3.2 State of a System 4
1.3.3 Thermodynamic Equilibrium 5
1.3.4 Characterization of Systems-Some elucidations 6
1.4 Values of Some Physical Constants 7
1.5 Summary 8
2. FIRST LAW OF THERMODYNAMICS 9-25

2.1 Internal Energy 9
2.2 Statement of the First Law 10
2.2. l Significance of the First Law IO
2.2.2 Proof of the First Law 10
2.3 Reversible Processes 12
2.4 Calculation of Wand q for Reversible Processes 15
2.5 Enthalpy 16
2.6 Heat Capacity 18
2.7 Ideal Gases 19
2.7. l Heat Capacity Relations 19
2. 7.2 Adiabatic Expansion and Compression 19
2.7.3 Work Done in Reversible Adiabatic Expansion 20
2.8 Summary 24
Problems 24
vii
Contents
3. HEAT CAPACITY AND ENTHALPY-AUXILIARY RELATIONS

AND APPLICATIONS 26-43
3.1 Heat Capacity of Ideal Gases 26
3.2 Heat Capacity of Solids 26
3.2.1 Einstein's Theory of Heat Capacity of Solids 27
3.2.2 Debye's Theory of Heat Capacity of Solids 28
3.2.3 Empirical Cp vs. T Relations 29
3.3 Enthalpy Changes 29
3.3. l Classification of Enthalpy Changes 29
3.3.2 Conventions for t1ll and H 34
3.3.3 Experimental Measurements of Enthalpy Changes 35
3.3.4 Hess' Law 35
3.3.5 Kirchhoff's Law 37
3.3.6 Concluding Remarks 41
3.4 Summary 41
Problems 42
4. SECOND LAW OF THERMODYNAMICS AND ENTROPY 44-63

4.1 Introduction 44
4.1.1 Various Statements of the Second Law 44
4.2 Carnot Cycle 45
4.2. l Introduction 45
4 .2.2 Carnot Cycle with Ideal Gas as Working Substance 46
4.2.3 Efficiency of Generalized Carnot Cycle 48
4.2.4 Thermodynamic Temperature Scale 50
4 .3 Entropy 51
4.3. l Entropy-A state property 51
4.3.2 Calculation of Entropy Changes 53
4.4 Second Law and Entropy-Significance and consequences 57
4.4.l General Comments 57
4.4.2 Various Interpretations of Entropy 57
4.4.3 Entropy Changes for Reversible and Irreversible Processes 58
4.4.4 Entropy Changes of Isolated Systems 61
4.5 Summary 61
Problems 62
5. AUXILIARY FUNCTIONS AND RELATIONS, CRITERIA FOR

EQUILIBRIUM 64-74
5.1 Free Energy 64
5.2 Combined Expressions of First and Second Laws of Thermodynamics 64
5.3 Maxwell's Relations 66
5.3. l Cp, Cv Relations for Thermodynamic Substances 67
5.3.2 Thermoelastic Effect 68
Contents
5.4 Criteria for Thermodynamic Equilibria 70

5.5 Criteria for Maximum Work 70
5.6 Criteria for Irreversible Processes 71
5.7 Minimum Free Energy at Equilibrium 72
5.8 Summary 73
6. GIBBS FREE ENERGY AND ONE-COMPONENT SYSTEMS 75-95

6.1 Introductory Comments 75
6.1.1 Variables Influencing Gibbs Free Energy 75
6.1.2 A Note on Subsequent Topics 76
6.2 Isothermal Processes 76
6.2.1 Some General Relations; Fugacity 76
6.2.2 Standard State-Fugacity relationship 78
6.2.3 Fugacity-Pressure Relationship in Gases 78
6.3 Variation of L1G with Temperature for an Isothermal Process 80
6.3.1 Gibbs-Helmholtz (G-H) Equations 80
6.4 Phase Equilibria in One-component (i.e. Unary) Systems 81
6.4. I Clapeyron Equation 82
6.4.2 Phase Equilibria with Vapour: Clausius-Clapeyron Equation 86
6.4.3 Effect of Total Pressure on Vapour Pressure 88
6.4.4 Integration of Clausius-Clapeyron Equation 89
6.4.5 Solid-Liquid-Vapour Equilibria: Triple point and metastability 91
6.5 Summary 93
Problems 94
7. ACTIVITY, EQUILIBRIUM CONSTANT AND STANDARD

FREE ENERGY OF REACTIONS 96-114
7 .1 Introduction 96
7 .1.1 Multicomponent Systems-Solutions and mixtures 96
7 .1.2 Physical, Chemical and Physico-chemical. Processes 97
7 .2 Thermodynamics of Solutions-Some introductory definitions
and concepts 98
7.2. l Composition Parameters- Mole Fraction 98
7.2.2 Partial and Integral Molar Properties 98
7 .2.3 Fugacity and Activity of a Component in a Solution 100
7.3 Equilibrium Constant (K) For .a Chemical Reaction 100
7.3.l Relationship of K with Standard Free Energy of Reaction 100
7. 3.2 Comments on Application 101
7.4 Variation of L1G° and K with Temperature 102
7 .4.1 Derivation of Relationship of L1<J'? with Temperature I 02
7.4.2 Alternative Derivation Procedure of L1<J'? vs. T Relation 104
7.4.3 Variation of Equilibrium Constant with Temperature 106
Contents
7.5 Ellingham Diagrams 107

7 .5.1 Basic Features 107
7.5.2 Nature of Slopes 109
7.5.3 Assessment of Relative Stabilities of Compounds 110
7.6 Summary I 12
Problems I I 3
8. EQUILIBRIA INVOLVING IDEAL GASES AND

PURE CONDENSED PHASES 115-129
8.1 Introduction 115
8.2 Equilibrium Calculations for Reactions Involving Ideal Gases 116
8.3 Chemical Equilibria Involving Pure Condensed Phases and Gas 120
8.3. l Clausius-Clapeyron Equation-Alternative derivation J20
8.3.2 Oxidation-reduction Equilibria Involving Pure Metal (M)
and Pure Metal Oxide (MO) 12 l
8.3.3 Phase Stability Diagrams of Stoichiometric Compounds
Dependent on Gas Phase 126
8.4 Summary 128
Problems 128
9. THERMODYNAMICS OF SOLUTIONS 130-158

9 .1 Introduction 130
9.2 Ideal and Nonideal Solutions 131
9.3 Partial and Integral Molar Properties; Gibbs-Duhem equation 132
9.3. l Relations amongst Partial Molar Properties 132
9.3.2 Gibbs-Duhem Equation 133
9.3.3 Integral Molar Quantities for Mixing 134
9.4 Some More Properties of Solutions 135
9.4.1 General Features 135
9.4.2 Ideal Solutions 136
9.5 Binary Solutions 137
9.5.l Activity vs Mole Fraction; Henry's Law 137
9.5.2 Variation of L1Gm, dJlffi and L1Sm with Composition for
Ideal Binary Solutions 140
9.5.3 Relationship of Gf. c:r with L1Gm in Binary Solutions 141
9.6 Gibbs-Duhem Integration for Binary Solutions 142
9.6. l Method I 143
9.6.2 Method 2 144
9.6.3 Method 3 (Darken's Method) 145
9.7 Regular Solutions 147
9. 7. l The Excess Function 148
9 .7.2 Regular Binary Solutions-Additional features 149
Contents
9.8 Quasi-chemical Theory of Solution 152

9 .8.1 Application to Ideal and Regular Solutions 153
9.8.2 Non-regular Solutions 154
9.9 Summary I 54
Problems 156
10. CHEMICAL POTENTIAL AND EQUILIBRIA AMONGST

PHASES OF VARIABLE COMPOSITIONS 159-185
10.1 Chemical Potential 159
10.l.l Definition 159
10. 1.2 Equality of Chemical Potentials amongst Phases at
Equilibrium at Constant Temperature and Pressure 161
10.1.3 Some Additional Remarks on Chemical Potential 162
10.2 Gibbs Phase Rule and Applications 164
10.2.1 Derivation of G:bbs Phase Rule 164
10.2.2 Application of Phase Rule 165
10.3 Thermodynamic Basis for Binary Alloy Phase Diagrams 169
10.3.l Free Energy vs. Composition Diagrams-General discussions 169
10.3.2 Phase Diagrams with Miscibility Gap 171
10.3.3 Free Energy vs. Composition Diagrams Involving Change
of Standard State 173
10.4 Some Typical Free Energy vs. Composition Diagrams 177
10.4.l Complete Solubility in Solid and Liquid Phases 177
10.4.2 Eutectic Phase Diagram with Finite Solid Solubilities 177
10.4.3 Phase Diagram with an Intermediate Phase 177
10.5 Solubility of Solute in a Phase in Binary System 177
10.5.1 Solubility of Metastable Phases 183
10.6 Summary 183
Problems 184
11. REACTION EQUILIBRIA INVOLVING CONDENSED PHASES

WITH VARIABLE COMPOSITIONS 186-212
11. l Analysis of Phase Stabilities of Reactive Stoichiometric
Compounds-The Si-C-0 system 186
11.2 Phase Diagram with Reactive Non-stoichiometric Compounds-
The iron-oxygen system 189
11 .2. l Phase Diagram 189
11.2.2 Activities in the Wustite Phase 190
11 .3 Reaction Equilibria Involving Solutions of Condensed Phases 192
11 .3. l General Considerations 193
11.4 Interactions amongst Solutes in Dilute Multicomponent Solutions 197
11.4. l Interaction Coefficients 198
Contents
11.4.2 Equality of ef and ej 199

11.4.3 Practical Utility of Interaction Coefficients 200
11.5 One Weight Per cent Standard State 201
11 .5 .1 Introduction 201
11.5.2 Definition and Relations for 1 wt.% Standard State 201
11.5.3 Evaluation of t1cfJ for Reaction Involving 1 wt.% SS 203
11.5.4 Dilute Multicomponent Solution in 1 wt.% SS 204
11.6 Sieverts' Law 209
11. 7 Summary 209
Problems 2 JO
12. TillRD LAW OF THERMODYNAMICS, STATISTICAL

THERMODYNAMICS, AND ENTROPY 213-227
12.1 The Third Law of Thermodynamics 213
12.1. 1 Nemst Heat Theorem 213
12.1.2 Entropy of a Pure Substance 214
12.1.3 Fundamental Significance of The Third Law 216
12.1.4 Experimental Verification of the Third Law 216
12. 2 Introduction to Statistical Thermodynamics 218
12.2.1 Kinetic Theory and Ideal Gas 218
12.2.2 Statistical Mechanics 219
12.2.3 Quantum Mechanics and Statistical Thermodynamics 221
12.3 Basic Concepts and Relations in Statistical Thermodynamics 221
12.3.1 Basic Concepts and Postulates 2'21
12.3.2 The Most Probable Macrostate 222
12.3.3 Internal Energy and Entropy at Most Probable Macrostate 223
12.3.4 Entropy of Mixing for a Binary Solution 225
12.4 Summary 226
13. THERMODYNAMICS OF ELECTROCHEMICAL CELLS 228-247

13.1 Introduction 228
13.1.1 The Daniel Cell 229
13.2 Thermodynamics of Reversible Galvanic Cells 230
13.2.1 Relation between Cell EMF (E) and Free Energy of
Cell Reaction (t1G) 231
13.2.2 Sign Convention for EMF 232
13.2.3 Cell EMF and Chemical Potentials at Electrodes 233
13.2.4 Temperature Coefficient of EMF 234
13.3 Examples of EMF Cells 234 ,
13.3.1 Examples of Reaction Cell 234
13.3.2 Examples of Concentration Cell 236
13.3.3 EMF Cells with Solid Oxide Electrolyte 237
Contents
13.4 Single Electrode Potentials and Electrochemical Series in

Aqueous Systems 241
13.5 Pourbaix Diagrams 243
13.6 Summary 245
Problems 246
14. THERMODYNAMICS OF SURFACES, INTERFACES

AND DEFECTS 248-265
14. l
Introduction 248
14.2
Thermodynamics of Surfaces 249
14.2. l General Considerations 249
14.2.2 Adsorption Isotherms 250
14.2.3 Effect of Surface Energy on Pressure 254
14.2.4 Effect of Surface/Interface Energy on Equilibrium Phase
Transformation Temperature 256
14.3 Thermodynamics of Defects in Solids 257
14.3.l Surfaces and Interfaces of Solids 257
14.3.2 Vacancies and Interstitials in Solid Metals 258
14.3.3 Point Defects in Ionic Compounds 261
14.4 Summary 264
Appendix Selected Thermodynamic Data 267-270
Bibliography 271-272
Answers to Problems 273-276
Index 277-280
Preface
Metallurgical thermodynamics, as also its modified version, 'Thermodynamics of Materials',

constitutes one of the principal scientific foundations of the disciplines of Metallurgical
Engineering and Materials Science and Engineering. Hence, all metallurgical and materials
science departments teach it as a compulsory course in their Bachelor's degree programme.
There are only a few books published so far, which are suited as textbooks for such a
course. These foreign-authored books are well-written, but somewhat voluminous and expensive.
I have taught this subject at the undergraduate level at the Indian Institute of Technology
Kanpur for several years. My experience is that both students and teachers would be
considerably benefited if a more concise text is available in the market. It will make
teaching as well as understanding of the basics better. Moreover, students will be able to
afford it for personal acquisition.
I started writing this book with these objectives in mind. The book contains essential
figures and tables, solved examples and problems to illustrate the concepts discussed. Data
for solving problems as well as answers to problems have also been provided.
The laws of thermodynamics are general. However, their applications are specific to
various disciplines. That is why a general book on thermodynamics does not meet the
requirements of all branches of science and engineering, and specialized texts are required.
We may classify thermodynamics into:
(a) Thermodynamics of Non-reactive Systems: Branches such as physics and mechanical
engineering are primarily concerned with this; books on Engineering Thermodynamics have
major coverage on this.
(b) Thermodynamics of Reactive Systems (popularly called Chemical Thermodynamics):
Here, physico-chemical processes and chemical reactions are primarily dealt with, and the
Gibbs Free Energy criterion is extensively employed for equilibrium calculations, thermodynamic
properties of phases and solutions; calculation of heats and entropies of reactions, and
physico-chemical processes are other important features .
Broadly speaking, metallurgical/materials thermodynamics belongs to chemical
thermodynamics. However, the major emphasis is on high temperature systems and processes
involving solid or liquid solutions of metals and inorganic compounds. For this, large
number of ewerimental thermodynamic measurements have been carried out in the last
6-7 decades, and some special relations and procedures have also been evolved to deal with
these solutions. Hence, the subject attained maturity and became an Engineering Science in
its own merit only from the decade of 1960 or so, although the general thermodynamic laws
and relations were well-established by 1930s.
xv
...
Preface
In the beginning of the 20th Century, metallurgy was primarily an art. Application of
pure and engineering sciences gave it recognition as a branch of engineering by 1950s, and
contributed to the development of metallurgical sciences, including metallurgical thermo-
dynamics. Atomic and solid state physics also matured by 1930s/1940s. In the decade of
1940s, major advances were also initiated in high technology areas such as nuclear, space
and aviation technologies, and solid state electronics. These required a variety of new
special purpose materials. Metallurgical sciences, solid state physics etc. provided scientific
foundation for the same. Earlier, metallurgical engineers were primarily concerned with
metals and alloys, and ceramists with traditional oxide ceramics, such as refractories. However,
the new materials, both metallic and non-metallic, changed the character of academic
programmes. It was realized that the fundamentals of all these are the same, and hence can
be taught in a generalized way. This led to the growth of Materials Science and Engineering.
In the llTs and engineering colleges in India, Bachelor's degree programme is offered
in metallurgical engineering or in a combined progranpne of metallurgical and materials
engineering. This text aims to cater to the basic one-semester undergraduate course on
thermodynamics. There are examples and problems on both metallic and inorganic non-
metallic materials. Chapter 14 has an additional brief ,,resentation on thermodynamics of
surfaces as well as interfaces and defects in solids of importance to materials science in
general.
Chapters 1-5 cover general thermodynamics and thermochemistry. Chapter 12 presents
some introductory discussions on statistical thermodynamics. The remaining chapters
(i.e. Chapters 6-11 and 13) deal with standard topics of chemical and metallurgical/materials
thermodynamics. As stated earlier, the principal emphasis is on high temperature systems
and processes of interest to metallurgy and materials science.
I wish to express my gratitude to the All India Council for Technical Education
(AICTE) for providing Emeritus Fellowship and miscellaneous financial assistance during
preparation of the manuscript. I am aJso indebted to the Indian Institute of Technology
Kanpur for making infrastructural facilities available, which has made this work possible.
I wish to thank Mr. J.L. Kuril for typing the manuscript, Mr. B.K. Jain for tracing the
diagrams, and Mr. A. Sharma for assistance in many ways. Finally, I sincerely thank my
wife, Radha and other family members without whose cooperation and encouragement the
writing of this book would not have been possible.
AHINDRA GHOSH
List of Symbols with Units
("-" indicates a dimensionless quantity; mo! means gram mole (i.e. mole); J means Joule)
A Helmholtz free energy, surface area J or J mo1- 1; m 2

ai activity of component i in a solution
C number of components in a system
Cy, Cp heat capacities at constant volume and constant pressure,
respectively J mo1-1 K- 1
E EMF of an electrolytic cell volt
ef interaction coefficient describing influence of solute j on
f. , in 1 wt.% standard state
Ii fugacity of component i in a solution; also activity
coefficient of i in l wt. % standard state
F Faraday's constant J vblC 1 g-equC 1
G Gibbs free energy J, J mo1- 1
cP Gibbs free energy at standard state J, J mo1-1
L1G change of G due to a finite process J, J mo1- 1
H enthalpy J, J mo1- 1
h Planck's constant J. s
hi activity of component i in a solution in 1 wt. %
standard state
general symbol of a component in a solution
J activity quotient
K equilibrium constant
Boltzmann's constant J K- 1 molecule- 1
equilibrium constant in 1 wt. % standard state
molecular/atomic mass g mor 1
xvii
m mass of one molecule/atom g molecule- 1

11 number of gram moles
N number of molecules/atoms
No Avogadro· s number
p Pr total pressure in a system atm
p number Jf phases
P1 partial pressure of co mponent in a gas mixture atm
q quantity of heat transferred J
Q general symbol of an extensive state property of a J mo1- 1, m3 rno1- 1•
solution (G, H. U, S, V, etc.) per mole etc.
value of Q for the entire solution J, m3 , etc.
partial molar value of Q in a solution (e.g. GI' Hi ) J mo1- 1, etc.
partial molar value for mixing of Q in a solution
(e.g. G- im , H
-111 s-m t )
i , i , e c. J mo1- 1, etc.
D.Q/11 integral molar value of Q for mixing, i.e. change of Q
upon mixing (e.g. llS"', D.G111 , etc.) J mo1- 1, etc.
Qxs excess molar value of Q (e.g. cxs, etc.) J mo1- 1, etc.
R Universal gas constant J mo1- 1K- 1, etc.
s entropy J K - 1, J mo1- 1K- 1
T temperature K
v internal energy J, J mo1- 1
v volume m3
w work done J, m3 atm

Wi weight percent of component i in a solution
Z valency, coordination number
Greek Symbols
a coefficient of volumetric thermal expansion
a. Darken 's a-function for component i in a solution
/3 isothermal compressibility of a material
r ratio of Cp/Cv
Yi activity coefficient of component i in a solution
rP Henry's Law constant for solute i in a binary solution
surface excess of component i
8 symbol for partial differential (e.g. 8q, SW)

L1 symbol indicating finite change of a state property
(e.g. 6G, 6H, 65)
E single electrode potential volt
energy of a particle at ith level J
e/ interaction coefficient giving influence of solute j on r.
in a solution
efficiency of an engine
chemical potential of component i in a solution J mo1- 1
oscillation frequency of atoms in a crystal s-1
p density of a material g cm-3, kg m-3
(f surface tension N m- 1
summation sign
electric potential volt
a parameter J mo1- 1
Superscripts
0 denotes values at standard state for extensive state
properties (e.g. SJ, d, JfJ)
m indicates change of value of a quantity due to mixing
(e.g. G{" =Gi - GP, fl.Gm =G - G°)
XS value of a state property in excess of that in ideal
solution (e.g. c;xs = Gi - Giid)
Subscripts
i, j, for component i, j ,
m for melting
f for formation of a compound from elements
T at temperature T
Tr for phase transformation
v for vapourization
Others
[ ] metallic phase
( ) nonmetallic condensed phase (e.g. oxide, etc.)
Introduction
1.1 Historical Perspective

Thermodynamics is a major discipline in science with widespread application in all branches
of science and engineering. The subject was developed in the 19th and early 20th Century
through the efforts of engineers, physicists and chemists. The laws of thermodynamics are
general, but their applications are specific to various disciplines. Hence the emphasis, content
and approach differ considerably from one branch to another in science and engineering disciplines.
Thermodynamics may be broadly classified into three: classical, statistical, and irreversible.
1. Classical (I.e. macroscopic or phenomenological) thermodynamics

Classical thermodynamics consists of the First, Second and Third Laws of thermodynamics.
It was developed in the 19th Century and in the first decade of the 20th Century. The Zeroth
Law is a later addition not recognised in many thermodynamics texts. Classical thermodynamics
treats a substance as a continuum, ignoring behaviour of atoms and molecules.
2. Statistical thermodynamics
Statistical thermodynamics originated from the Kinetic Theory of Gases, which related
pressure to average kinetic energy of molecules in an ideal gas. The application of probability
theory, quantum theory and statistical mechanics allowed it to arrive at macroscopic
thermodynamic relations from atomistic (i.e. microscopic) point of view. It was founded by
Maxwell, Gibbs and Boltzmann in the late 19th Century, and was developed further in the
1920s and 1930s.
3. In-eversible thermodynamics
Irreversible thermodynamics deals with the application of thermodynamics to irreversible
processes, and was first proposed by I. Prigogine in 1942.
Textbook of Materials and Metallurgical Thermodynamics
Another way of classifying thermodynamics is:

1. Thermodynamics of nonreactive systems.
2 . Thermodynamics of reactive systems, also known as chemical thermodynamics. It is
often considered as a part of physical chemistry, which includes subjects such as
reaction kinetics and diffusion besides chemical thermodynamics.
Since it is only the classical thermodynamics which has widespread application in
Metallurgical Engineering and Materials Science and Engineering, we shall confine ourselves
mostly to this, except for one chapter on statistical thermodynamics. For this reason, the
term thermodynamics generally means classical thermodynamics. Thus, in Section 1.2, we
will briefly review the historical development of classical thermodynamics.
1.2 History of Classical Thermodynamics
1.2.1 General
The literal meaning of thennodynamics is that it is a subject dealing with the relation
between heat and motion. The invention of the steam engine in the 18th Century triggered
the issue of the relation between heat and work amongst engineers and physicists. In 1798,
Count Rumford observed that the heat generated during boring of cannons was approximately
proportional to the work done. In 1840. Joules' classic and precise experiments provided
confirmation of the above and led to the concept of Mechanical Equivalent of Heat. It was
later generalized into Law of Conservation of Energy, which states that "energy can neither
be created nor destroyed; it can be only transfonneJ from one fonn into another". These
developments eventually led to the generalized definition of thermodynamics as a subject
dealing with energy and its transformation from one form to another. This constituted che
basis for the First Law of Thermodynamics.
Although mechanical energy can be completely converted into heat, it was observed that
the reverse is not true, i.e. heat cannot be completely converted into work in a straightforward
fashion. Quantitative explanation came first fiom the famous Carnot's cycle, derived by
S. Carnot in 1824. These Jed to the enunciation of the Second Law of Thermodynamics,
employing the concept of reversible and irreversible processes. Later in the mid-19th Century,
Clausius and J.J. Thomson (Lord Kelvin) proposed entropy as the thermodynamic parameter
for quantitative application of the Second Law.
In 1906, W. Nernst proposed •his Heat Theorem on the basis of experimental evidence.
It was later generalized by Max Planck as "the entropy of any homogeneous substance
(which is at comp!P.te internal equilibrium), may be taken as zero at 0 K". This is known
as the Third Law of Thenrwdynamics.
1.2.2 Chemical Thermodynamics

From experimental evidence, in the late 18th Century, Lavoisier concluded that matter is
conserved even when chemical reaction occurs. Thus the Law of Conservation of Mass came
Introduction
into existence. The 19th Century further witnessed major developments in chemistry,
such as
• Dalton's atomic theory and reaction stoichiometry
• Avogadro's hypothesis and concept of g-mole
• The concept of chemical equilibrium
• The heats of reaction
• Faraday's Laws of .Electrolysis
These provided the foundation for Physical Chemistry and Chemical Thermodynamics.
In the late 19th Century, J.W. Gibbs proposed his famous free energy function, known as
Gibbs free energy. The systematic application of this to thermodynamics by Gibbs and
others firmly established the foundation of chemical thermodynamics. The subject attained
maturity by early 20th Century.
Chemical thermodynamics is also based on the three laws of thermodynamics. However,
it employs many other auxilliary relations also, mostly derived from Gibbs free energy
co~siderations as well as other physico-chemical relations.
1.2.3 Metallurgical Thermodynamics
Broadly speaking, the application of primarily chemical thermodynamics to metals and

materials has led to the development and growth of Metallurgical Thermodyrtamics or its
later generalization as Themwdynamics of Materials.
The application of laws and relationships of thermodynamics requires experimental data
such as enthalpies and free energies of reactions and processes, activity versus composition
relationships in solutions, phase equilibria and phase diagrams. Processing of metals and
ceramics is canied out primarily at high temperatures. Experimental measurements at high
temperatures pose certain problems, not encountered at room temperature. Overcoming
these require other technological development!., e.g. availability of high temperature materials,
techniques, apparatus as well as instruments, many of which were not available earlier.
Therefore, experimental measurements and assessment of thermodynamic data in high
temperature metallurgical and materials area required several decades of efforts by a large
number of scientists. This is the reason why the subject of Metallurgical Thermodynamics
attained maturity as late as 1960/1970, although chemical thermodynamics was well established
by 1930.
In the early days, several physical chemists took interest in metals and inorganic compounds
as well as their behaviour at high temperatures. It is they who did pioneering work in
thermodynamics of metals and materials.
1.3 Introductory Concepts and Definitions

As already stated. thermodynamics may be classified into: classical (i.e. macroscopic),
statistical , and irreversible. However, it is classical thermodynamics that has general and
large-scale applications. Therefore, this text will be concerned almost exclusively with this.
The following introductory concepts and definitions are the starting points for classical
thermodynamics.
1.3.1 System and Surrounding

Any portion of the Universe selected for consideration is known as the system. The rest of
the Universe is its surrounding. Depending on the nature of interaction, systems have been
classified into open, closed and isolated. An open system exchanges both matter and energy
with its surrounding. A closed system exchanges only energy with its surrounding. An
isolated system does not exchange either matter or energy with its surrounding. Hence the
mass is fixed for a closed or isolated system during occurrence of a process. Laws and
equations of thermodynamics are taught and comprehended more easily for the above.
Hence, the readers are advised to visualize and keep in mind a closed system unless
otherwise mentioned.
Classical thermodynamics deals with a system of macroscopic size (say, contafaing at
least a million molecules), which we treat as a continuum, and ignore the details of structure.
Because of these, the laws and equations of classical thermodynamics are independent of
structure (i.e. structure insensitive). This is where its power and usefulness ties. However,
the generalizations are valid under certain ass umptions. In addition, some definitions have
been proposed. It is necessary to understand these properly.
1.3.2 State of a System

By 'state' we shall mean the macrostate of a system in classical thermodynamics since this
is what we shall be concerned with. The state of a system at any instant is defined by
specifying &JI its state variables (which also include state properties), such as temperature,
pressure, volume, and density. Thermodynamics deals with energy and its transformation
(i.e. conversion) from one form to another. Simplifications are achieved by ignoring forms
of energy which do not contribute significantly to the process under consideration. For
example, in the absence of a significantly strong magnetic field of force, magnetic energy
can be ignored. Surface energy is ignored if the specific surface area is small and, hence,
surface energy term is not significant.
Pressure P, volume V, and temperature Tare the most common state variables. The Jaws
of thermodynamics have been proposed and expounded on the basis of the above. The
resulting relations were then amended to include other variables as necessary.
For a substance of fixed mass, P, V and T are interrelated. For example, for an ideal
gas,
PV = nRT ( 1. 1)
i.e.
V RT
v=-=- (1.2)
11 p
Introduction
where n is number of moles (i.e. g-moles), R the universal gas constant, T the absolute
temperature, and v the molar volume.
Now we have to distinguish between property and variable. The molar volume is a
property of the substance. Of course, it can be varied. In that sense, it is a variable too.
But pressure and temperature are not properties. They are variables, generally imposed by
the surrounding. Therefore, variables may or may not be properties. It may also be noted
here that state properties are also known as state functions .
The volume of a substance is proportional to its mass (or number of moles) . On the
other hand, P and T are independent of the mass of the system, and are known as intensive
variables. Volume V is an extensive property. It can be made intensive by dividing it by
mass or number of moles. Hence, the molar volume (v) is an intensive property. Density
is an intensive property.
It is obvious that, if an equation contains both extensive property and intensive variable,
then there must be a term denoting mass or number of moles. If the latter are missing, then
there is an implicit assumption that the extensive property is for a fixed mass (say, a closed
system). The general convention in chemical thermodynamics is to go for molar properties,
which are intensive. This way, the relationships and functions become independent of the
quantity of matter, and hence of more general applicability [e.g. Eq. (1.2)].
1.3.3 Thermodynamic Equilibrium

When we wish to characterize a state by specifying the values of the important intensive
variables, it is assumed in thermodynamics that the system has come into equilibrium with
respect to those variables. Then and only then the state can be properly defined in terms
of these variables. A substance which is at thermodynamic equilibrium is known as a
thermodynamic substance.
For a thermodynamic substance of fixed mass, the equation of state is
V = f(P, 7) (1.3)
Equation ( 1.2) is a specific case of Eq. (1.3) for ideal gas only. Another, and more general
form of Eq. ( 1.3), is
v = -v =f(P, T) (1.4)
n
It should be emphasized here th.at Eqs. (1.3) and (1.4) are applicable to a substance of
constant composition and structure, i.e. its nature remains the same during change of
pressure and temperature. For every substance, except for ideal gas, the function is different.
When the system is at equilibrium with respect to an intensive variable, the magnitude
of the variable throughout the system is the same (i.e. uniform). The term equilibrium has
originated from the Latin word "acquilibrium", meaning equal weight (e.g. as in a weighing
balance). In general, it refers to a state, which is not changing with time.
The mechanical interaction of the system with the surrounding is represented by pressure,
in the absence of a field of force (electric, magnetic etc.). Here, mechanical equilibrium
means pressure equilibrium, i.e. uniformity of pressure throughout the ~ystem. Thus, there
is no tendency for the pressure to change with time anywhere in the system.
Similarly, at themial equilibrium, temperature should be uniform throughout the "System.
Only then there will be no tendency for heat to flow from one part of the system to another,
i.e. no thermal change takes place in the system.
For purely pl )Sica] processes, such as expansion/compression of a substance, thermodynamic
equilibrium conusts of mechanical and thermal equilibria. But for physico-chemical processes
and chemically reactive systems, thermodynamic equilibrium would also require attainment
of physico-chemicallchemical equilibrium in addition to mechanical and thermal equilibria.
This means attainment of uniform chemical potential besides uniformity of pressure and
temperature in the system. Detailed discussions of chemical potential will be taken up in
Chapter 10. ,
T he system is at its most stable state when it is at thermodynamic equil ibrium. It will
stay that way unless disturbed. An open or a closed system interacts with its surrounding.
There, the equilibrium is always with respect to the surrounding. If the surrounding changes,
the state of equilibrium of the system will also change.
Sometimes a system is not at complete thermodynamic equil ibrium, but at partial or
pseudo equilibrium. Thermodynamics is capable of handling such cases. One example is
metastable state, where the system undergoes change so slowly that there will be no perceptible
change in a long period of time (say, several years/centuries/millennia). For instance, if a
mixture of hydrogen and oxygen gas is kept at room temperature, very little detectable
reaction would occur even in years, although it is not at chemical equilibrium.
Cementite (Fe3C) in iron-carbon system is metastable, but can stay for perhaps several
centuries or millennia at room temperature. In both the above cases, raising the temperature
increases the rate enormously, and allows attainment of chemical equilibria in a short period
of time.
Another example is partial chemical equi libria in some complex reactive systems, where
some reactions attain equilibrium whereas some others do not.
1.3.4 Characterization of Systems-Some elucidations
We have broadl y classified system-as open, closed and isolated already. It has also been
stated that processes may be purely physical or physico-chemical/chemical. In the former
case, we are dealing with a nonreactive system, whereas, when physico-chemical processes
and/or chemical reactions occur, the system is a reactive one. Some further comments are
now noted to elaborate the above observations.
A system consisting of one phase only (say, water) is a single-phase system, which is
known as a homogeneous system. At 0°C, ice and water can co-exist at equil ibrium. Then
it is a heterogeneous system. However, the above classifications do not provide full description
of a system. Further classifications are required for application of the laws of thermodynamics.
A substance in a closed or isolated system not only has fixed mass, but has fixed overall
composition as well. Consider water in a closed vessel. On cooling below 0°C, it is
Introduction
transformed into ice. Again, on heating above 100°C, it is transformed into a gas. Therefore,
the state of aggregation of H 20 (broadly speak ing, structure of H2 0 ) depends on temperature.
Similarly, above 910°C, the structure of pure iron changes from a-Fe (i.e. BCC) to y-Fe
(i.e. FCC). These changes are known as phase clianges, and are physico-chemical processes.
The system has one-compo11e11t only (H 20 or Fe), and is also referred to as Unary system.
A multicompo11e11r system obviously has more than one component. An aqueous solution
of NaCl, for example, has two components, (viz. NaCl and H 20 ), and is known as a binary
system. Similarly, we have ternary, quartemary etc., all coming under multicomponent
system. If the NaCl solution is cooled below 0°C, pure ice crystals would start forming. As
temperature is lowered further, more and more ice forms. This increases concentration of
NaCl in residual water. Hence, in this two-phase system, composition o; water is changing.
However, composition of the entire system in terms of number of moles of H 20 and NaCl
does not change, if it is a closed system.
In contrast, in an open system, both mass and composition of the systein can change.
For example, if we consider only the aqueous NaCl solution as our system (even in a closed
vessel) in the two-phase ice + aqueous solution situation, then both mass and composition
of the solution are changing with change of temperature. If pure water is heated in an open
beaker, its mass will keep decreasing with vaporization. If it is an aqueous solution , then
besides mass, concentration of NaCl in solution will keep increasing with vaporization.
If water is heated to a very high temperature, the steam dissociates into hydrogen and
oxygen, and we have to be concerned with this chemical reaction in the gaseous state. Then,
if the system is closed, its total mass will remain constant. Not only that, its composition
in terms of total number of moles of hydrogen and oxygen (in elemental form and combined
as H 20) will remain constant. Therefore, the meaning of the term composition has to be
interpreted that way.
To sum up, in classical thermodynamics, a system can be classified as:
1. Open, closed or isolated
2. Unary or multicomponent
3. Reactive or nonreactive
4. Homogeneous or heterogeneous
These are required for application of thermodynamics. Some of these will again be
elucidated in subsequent chapters. However, it is hoped that these introductory comments
would be useful to readers.
1.4 Values of Some Physical Constants

. Absolute temperature of ice point (0°C) = 273.15 K
1 calorie = 4.184 joules
Avogadro's number (No) = 6.023 x 10 23 per mole
Boltzmann's constant (k8 ) = 1.3805 x 10- 23 joules/K/molecule
Faraday's constant (F) = 96,500 coulombs/g-equivalent (i.e. joules/volt/g-equivalent)

l Standard atmosphere = 760 mm Hg
= 1.013 bar
= 1.013 x 105 pascal
Universal Gas Constant (R) = 8.314 joules/mol/K
= 82.06 cm3-atm/mol/K
= 82.06 x 10-6 m3 -atm/mol/K
1.5 Summary
1. Thermodynamics may be broadly classified as: classical (i.e. macroscopic or
phenomenological), statistical, and irreversible.
2. Most applications are based on classical thermodynamics. Hence, the present text is
primarily concerned with it. Also, generally, the term themwdynamics denotes classical
thermodynamics.
3 . The three laws of thermodynamics constitute the foundations of thermodynamics.
4. Thermodynamics deals with energy and its transformation from one form to another.
5. Metallurgical Thermodynamics, or its later generalization, Thermodynamics of Materials,
belongs to chemical thermodynamics, which deals with reactive systems.
6. A system is classified as: open, closed and isolated. The rest of the Universe is its
surrounding.
7 . State variables are extensive or intensive.
8. Molar properties are intensive properties (or variables).
9. Pressure, temperature and volume are common state variables.
10. A state can be defined by state variables only at thermodynamic equilibrium. It then
constitutes a thermodynamic substance.
11. A system needs characterization before its thermodynamic treatment.
Chapter2
First Law of Thermodynamics
Two important thermodynamic quantities have been proposed on the basis of the First Law
of Thermodynamics, viz. Internal Energy U and Enthalpy H. Both are energy terms and are
state properties.
2.1 Internal Energy

Imagine a system. Suppose a quantity of heat q is supplied to it from the surrounding, and
as a consequence the system does an amount of work W on the surrounding. If these are
the only two energy terms involved in the process, then the Law of Conservation of Energy
demands that q be equal to W. However, experiments have revealed that,'commonly,
q - W;it:O
In order to save the Law of Conservation of Energy, it was proposed that any system
possess an internal energy U, which is a hidden form of energy stored in it. The difference
between q and W was attributed to change in U (i.e. ,1U) during the process , i.e.
t.U =q - W (2.1)
It should be noted that according to mechanics, it is actually a force and not a system
that does work. The concept that a system does work arose from the historical context of
development of classical thermodynamics, viz. conversfon of heat into mechanical work by
a heat engine. Also, it should be further noted that in school text books and modern
thermodynamics texts in some areas, W is taken as positive, if the work is done on the
system, which makes t.U = q + W. However, in traditional thermodynamics, physical
chemistry and chemical thermodynamics texts, the convention, as in Eq. (2.1), is followed,
and, therefore, we are retaining the same.
Equation (2.1) is the mathematical statement of the First Law of Thermodynamics. On
the basis of our present day knowledge, the internal energy consists of
1. macroscopic kinetic energy due to motion of the system as a whole,
9
2. potential energy of the system due to its position in a force field,

3. kinetic energies of atoms and molecules in the form of translation, rotation and
vibration (i.e. thermal energy),
4. energies of interaction amongst atoms and molecules (a form of potential energy),
5. coulombic energy of interaction amongst electrons and nucleii in atoms, and
6. energy contents of electrons and nucleii of atoms.
In conventional chemical thermodynamics which we shall be concerned with, only
kinetic energies of atoms and molecules, and interactions amongst atoms and molecules, i.e.
items (3) and (4) above, are considered to be important since changes occurring in them
principally contribute to LJU. However, changes in other forms of internal energy also
assume importance in non-conventional processes. For example, (5) and (6) would be
significant in nuclear reactions.
It is also to be understood that the absolute value of U for a system is not known. All
we can determine is LW. It may also be noted that U of a substance (i.e. material) depends
on its mass, nature (i.e. chemical composition and structure) as well as temperature. For a
material of fixed mass, composition and structure, U is a function of temperature only.
2.2 Statement of the First Law

Equation (2.1 ) is the mathematical statement of the first law for a finite process. For an
infinitesimal (i.e. differential) process, the statement is
dU = fxJ - OW (2 .2)
2.2.1 Significance of the First Law

l. It is based on the Law of Conservation of Energy .
2. It brought in the concept of internal energy.
3. It separates heat interaction and work interaction between the system and the surrounding
as two different terms.
4. It treats internal energy as a state property, i.e. dU in Eq.,(2.2) is an exact differential.
Hence, LJU or dU would depend only on the initial and final states of the system,
and not on the path by which the system has moved from one state to another. In
contrast, q and W are energy in transition, and therefore, are not state properties/
variables but are dependent on the path. Since the path may be known or unknown,
and is likely to depend on several variables (known and unknown), infinitesimal
changes in q and W in Eq. (2.2) are designated by /jq and OW, not dq and dW. The
readers are advised to exercise particular care to maintain this distinction.
2.2.2 Proof of the First Law
The concept of internal energy was proposed assuming the validity of the Law of Conservation
of Energy. AU is just the difference between q and W. There is no direct proof of existence
of U as a state property. However, there are plenty of indirect experimental evidence of its
existence and is a well-established quantity.
In the early days of thermodynamics in the 19th Century, when the First Law was proposed,
the nature of U as a state property was inferred from experiments on cyclic processes, where
the system was changed but ultimately brought back to the same initial state. It was found
that for such cyclic processes,
1:.q- 1:.W=O (2.3)

cycle cycle
provided q and W are expressed in the same energy unit (say, joule). This means that
:E AU= O, i.e. U is a state property.
cycle As discussed in Chapter 1, in conventional classical thermodynamics, we are generally
concerned with P, V, T as state variables. In chemical thermodynamics, the chemical
composition and structure of substance in the system also are variables. However, variation
of thermodynamic quantities with composition and structure will be taken up later. For the
time being, we shall assume the substance to retain its initial composition and structure all
throughout the process (i.e. a nonreactive system).
In Chapter 1, we have further stated that P, V, T for a system of fixed mass are
interdependent, with only two of them acting as independent variables. Hence, for a system
of fixed mass (imagine a closed system for easy comprehension),
U = f(V, D (2.4)
or
U = f(P , D (2.5)
or
U = f(P, V) (2.6)
Invoking properties of exact differentials, from Eqs. (2.4)-(2.6), respectively,
du = (au)
av r
dV + (au)
ar v
dT (2.7)
or
du = (auJ
aP T
dP + (auJ
ar p
dT (2.8)
or
dU = (auJ
aP v
dP + (auJ
av p
dV (2.9)
2.3 Reversible Processes

In a reversible process, the system is displaced from equilibrium infinitesimally, and then
allowed enough time to attain a new equilibrium; and again displaced infinitesimally, and
thus the process continues. Thus, a reversible process may be defined as the hypothetical
passage of a system through a series of equilibrium stages. A reversible process is very slow
and mostly impracticable. No going process is reversible in a strict sense. The terminology
reversible process arises from the fact that it can be reversed by an infinitesimal change of
a variable in the opposite direction.
The concept of reversible process is at the heart of thermodynamics. It may be explained
by the iraditional cylinder-piston analogy. The gas inside the cylinder constitutes the system
and everything else is the surrounding. One way to expand the gas is to decrease the external
pressure. Imagine lowering of the external pressure from P 1 to P 2• During expansion, the
temperature of gas goes down, causing heat flow into it from outside through the cylinder
wall. Finally, a new equilibrium is attained when the gas pressure inside is P 2 and the
temperature is again equal to that of the surrounding (Tsur).
The overall process is isothermal since both initial and final temperatures are equal to
Tsur However, during expansion, the gas temperature will be lower than Tsur Figure 2.1
depicts the process as temperature vs. time plot. If pressure is lowered rapidly to P - AP,
where AP is a finite change in P, the process is irreversible and variation of temperature
with time is unknown, although a hypothetical curve is drawn in Fig. 2.1, just for visualization.
t State 1
Idealized
reversible
(T _ T ) State 2
_c:~;·~--
r sur
Reversible Irreversible
Time---+
Fig. 2.1 Temperature vs. time curves for isothermal processes (schematic); path is hypothetical
for irreversible process.
Process, on the other hand, occurs reversibly if the pressure is decreased a little to P - dP,
where dP is an infinitesimally small change in pressure. Then sufficient time is allowed for
the gas to attain Tsur• and again the pressure is decreased a little, and time allowed. Here
we are carrying out the process slowly through a series of equilibrium stages. It should be
noted that the temperature is always close to T5ur in the reversible process. The mathematically
limiting situation for curve 2 is the constant temperature line at T = Tsur• which is the
idealized version of an isothermal reversible process. Figure 2.2 illustrates these.
(a) Irreversible
expansion
(b) Reversible
expansion
Fig. 2.2 Irreversible and reversible expansion of a gas at any instant of time.
Now, consider isothermal expansion-compression of n moles of an ideal gas. Figure 2.3

shows the process on a P-V diagram. The equation of state is
PV = nRT
(State 1)
A Irreversible
I
I Idealized reversible path
I (PV ::;; nRT ::;; constant)
I
I
I
I
I
---r----~-p~;(..~ B
: Irreversible : (State 2)
1
contraction I
V1 V2
Volume---.
Fig. 2.3 Pressure-volume diagram for isothermal expansion-compression cycle of an ideal gas
in a cylinder-piston system (schematic); P is inside pressure; path is hypothetical for
irreversible cycle.
which is the same as Eq. (1.1), and the corresponding curve is for isothermal reversible
process at temperature T. The irreversible expansion and compres~ion cannot be depicted on
the P-V diagram since their paths are undefined. Moreover, for irreversible process, the
pressure of gas in the cylinder is not expected to be uniform. However, it is qualitatively
correct to say that, during expansion, the gas pressure inside the cylinder (i.e. in the system)
will be higher than the external pressure. The reverse is true for irreversible compression.
On this basis, some hypothetical curves are shown.
When the gas expands, it performs work against external pressure. Suppose the cylinder
moves by a distance dl. Then,
OW = Pext • A • dl = Pe,. 1 • dV (2.10)

where A is the cross-sectional area of the cylinder and dV is the increase in volume of the
gas in the cylinder. For reversible expansion,
(2. 11)
where P denotes gas pressure inside the system. Hence,
OW= PdV (2.12)

For compression, Eq. (2.12) is applicable also. Only, the sign of bW would be negative.
Consider the isothermal reversible process again. Here,
(i) The work done by the gas during expansion state A ~ B (as in Fig. 2.3) = Wi.
where
f dV y;
f
V2 V2
W1 = P dV =nRT - =nRT lo ---1. (2.13)
~ ~ v Vi
(ii) The work done by the gas during compression (B ~ A) = W2, where
v, v;
J~
~
W2 = P dV =nRT In -1.. = - nRT In ---1. = - W1 (2.14)
~ Vi
Thus, net work done in the cycle A ~ B ~ A is
L
cycle
w =WI+ W2 = 0 (2.15)
Since the system comes back to the initial state, as stated in section 2.2.2,
I: i1U = O (2.16)
cycle
because U is a state property. From Eqs. (2.1), (2.15), and (2.16), I: q = 0.

cycle
This means that there is no net interaction of the system with the surrounding for a
cyclic reversible process. Hence, it does not leave any permanent change in system or
surrounding during reversing. This is an important conclusion.
It is obvious that for irreversible processes, the above is not true and some net heat and
work interaction with the surrounding would occur for cycle A ~ B ~ A. The shaded area
is the net work done during the irreversible cycle A ~ B ~ A. The presence of friction
between the piston and the cylinder will introduce irreversibility as well. Therefore, for
reversibility, friction should be absent.
It should be noted again that the path is not known for an irreversible process. Hence,
it cannot be shown in a diagram. Therefore, the irreversible processes shown in Figs. 2.1
and 2.3 are hypothetical.
2.4 Calculation of Wand q for Reversible Processes

Both W and q are dependent on path. Hence, they can be calculated for a process if the
entire path is known, and this is possible only for reversible processes, as explained in
Section 2.3. In general,
OW= PdV +OW' (2.17)
where
OW =infinitesimal total work done by the system, as in Eq. (2.2)
P dV = infinitesimal work done by the system against pressure
OW' =infinitesimal quantity of other forms of work done by the system (electric,
magnetic, etc.).
We shall consider bW' when we discuss electrochemical cells or surfaces (in Chapters 13
and 14). Till then, in all our .subsequent discussions, we shall assume OW' = 0. Hence,
OW = P dV, and combining this with Eq. (2.2),
dU = liq - P dV, i.e. liq = dU + P dV (2.18)
For a finite reversible process from state 1 to state 2,
(2.19)
and would depend on the P-V relationship. Let us consider the following reversible process
paths:
(i) At constant volume (i.e. isochoric), W = O; hence from Eq . (2.1),
LiU = q (2.20)
(ii) At constant pressure (i.e. isobaric),
W = P(V2 - V1) (2.21)

(iii) At constant temperature (i.e. isothennal)
Textbook of Materials and Metallurgical Therm~ynamics
On the basis of experimental measurements by Joule and others in the 19th Century, it
was proposed that U is a function of temperature only for an ideal gas. Later, kinetic theory
of gases also predicted the same. Hence for the isothermal expansion-compression of an
ideal gas,
11U = 0, i.e. W =q (2.22)
It should be kept in mind that Eq. (2.22) is applicable only to a fixed mass of gas since
any change in mass will change internal energy even at a constant temperature.
(iv) Adiabatic:
q = 0, i.e. 11U =- W (2.23)
2.5 Enthalpy
Enthalpy H is defined as
H = U + PV (2.24)
Hence, by definition, H is a state property, since U and V are state properties and P is a
state variable. Differentiating Eq. (2.24), we get
dH = dU + P dV + V dP (2 .25)
i.e.
dH = dU + P dV at constant P (2.26)
If bW = 0, then by combining Eq. (2.26) with Eq. (2.18), we obtain
dH = (liq - P dV) + P dV = liq (2.27)

Hence, for a finite process at constant pressure from state 1 to state 2,
sl.ale 2
q
I
=
st ale I
dH =H 2 - HI =.&1 (2.28)
Since most processes occur or are carried out under conditions where Eq. (2.28) is
exactly or approximately valid, Eq. (2.28) constitutes the basis for process heat balance,
which is very important for manufacturing industries as well as analysis of processes.
Energy is costly, and should be accounted for and saved as much as possible. The bulk of
the energy supply goes to meet process heat requirements. Without process heat balance,
energy audit is not possible. Further, Eq. (2.28) allows experimental determination of L1H
of a process by measurement of q by a calorimeter. Since H is a state property, LJH
(i.e. H 2 - H 1) can be tabulated in thermodynamic data sources and used by all.
Many texts, especially the older ones, therefore, refer to enthalpy as heat content of a
substance due to the above equality. However, the latter terminology is not scientifically
correct since heat is energy in transition and not a state property.
First Law of The"nodynamics
As stated in Section 2.4, internal energy of an ideal gas is a function of temperature

only. At constant temperature (i.e. dT = 0), therefore, dU = 0, and Eq. (2.8) gets simplified
into
(au)r =O
dP
(2.29)
Differentiating Eq. (2.24) with respect to P, we obtain
(aH) ='(au)
dP r aP T
+ [a(PV)]
ap T
(2.30)
In ideal gas at constant T, PV = constant; so, d(PV) = 0. From all these above,
(aH) _(au) _0
ap T ap T
(2.31)
In other words, enthalpy of an ideal gas is independent of pressure at constant temperature.

Similarly, enthalpy is independent of volume. Hence, H is a function of T only for a fixed
mass of a . substance. This is of great help in calculating the process heat balance since in
- · most metallurgical processes, the gas phase is essentially ideal.
EXAMPLE 2.1 The Equation of State of a gas is given by the expression
(E.2.1 )
where a, b are constants. Prove that
8T) (8P) (8V)

(8P = _l
v 8V r or p
Solution The partial differentiation of Eq. (E.2. 1) yields
(;~l = v; b (E.2.2)
(~~l = V ~ b [2a(~3- b) - (p + :2 )] (E.2.3)
(E .2.4)
Textbook of Materials and Metallurgical Thermodyn~mics
Combining Eqs. (E.2.2)-(E.2.4), we obtain
OT) (OP) (sv) - -1

(oP v sv r or p
2.6 Heat Capacity

It is the quantity of heat required to raise the temperature of a substance by 1°C (i.e. 1 K).
In thermodynamics, the molar heat capacity (C), i.e. heat capacity of 1 g-mole of a
substance is most widely employed. Thus,
C= 8q (2.32)
dT
0q is heat required to raise the temperature of l mole by dT. Since 0q depends on path, C
is a meaningless quantity unless the path is specified. The following have proved to be
useful.
The molar heat capacity at constant volume is given by
cv -(oq) -(au)
- dT v - dT v (2.33)
since at a constant volume, 0q = dU. Similarly, molar heat capacity at constant pressure
(Cp) is given by
C -("dq)
P -dT
-("dH)
CJT
P - p
(2.34)
since at constant P, &, = dH. In Eqs. (2.33) and (2.34), U and H are molar values of
internal energy, and enthalpy, respectively.
From Eq. (2.34),
dH = Cp dT (2.35)
Integrating Eq. (2.35) between temperatures T1 and T2, we get
(2.36)
It should be noted that U and H are molar qµantities here. Earlier, they referred to
internal energy and enthalpy of the entire system. Both conventions will be followed. So, U
and H may be molar values or for the entire system, depending on what is being discussed.
Cp > Cv since Cp includes heat required to do work against pressure also, besides raising
temperature. From Eqs. (2.33) and (2.34),
Cp _ Cv = (aH)
ar
_(au)
ar v
p
= [a<u + PV)] _ (au)
ar ar v
p
_(au)
ar
_(au)
ar
+P
p
(av)
ar v p
(2.37)
Again, differentiating Eq. (2.7) with respect to T at constant P, we get
(2.38)
Combining Eqs. (2.37) and (2.38), we get
cp _ Cv = (au) (av) + P(av) _(av) [(au) + P]

av r ar p ar p - ar p av r
(2.39)
2.7 Ideal Gases
2. 7 .1 Heat Capacity Relations
For I mole of an ideal gas, from Eq. (I.I),
(2.40)
Also, from Section 2.5,
(au)
av
=O
r
(2.41)
Combining Eqs. (2.39)-(2.41), therefore, we get
Cp - Cv= R (2.42)
2.7 .2 Adiabatic Expansion and Compression

For an adiabatic process, liq = 0. Hence,
dU = - bW = - P dV (2.43)
Since PV = 11RT for n moles of ideal gas,
dU= - nKr dV (2.44)

v
Again, from Eq. (2 .33),
dU = nCvdT (2.45)
Combining Eqs. (2.44) and (2.45), and integrating between T = T 1, V = VI> and T = Ti,
V = Vi. we get
i.e.
R/Cv
i = ( ~: )
(2.46)
Let Cp/Cv = r Then, from Eq. (2.42),
R
-=r-1 (2.47)
Cv
(2.48)
From Gas' Laws,

T2 P2V2
-=-- (2.49)
Ti FW1
Combining Eqs. (2.48) and (2.49), we obtain
P2
Pi
=(V1
V2
)r, i.e. pyr =m =a constant (2.50)
Figure 2.4 compares P-V relationships for isothermal and adiabatic reversible expansion
of an ideal gas on the P-V diagram, starting from the same initial state A(P1> V 1, T1) to the
same final pressure P2 . The final volume is larger for the isothermal process.
2.7.3 Work Done in Reversible Adiabatic Expansion

For the reversible adiabatic expansion (Fig. 2.4),
(2.51)
I
I
I Isothermal
I / reversible
I
I Adiabatic
I reversible
I
I
I
---T------
1 (P2, V2, T2) C
I I
I I
V2 V3
Volume--.
Fig. 2.4 P-V diagrams for isothermal and adiabatic reversible expansion of an ideal gas from
the same initial state (schematic).
=m
vt-r -
2
v,1-r)
I
( (2.52)
1-r
Noting from Eq. (2.50) that
(2.53)
we obtain
W = P2 V2 - Pi \'i (2.54)
1- r
From Eq. (2.45),
LJU = U2 - U1 = nCv<.T2 - Ti) (2.55)

Also, LJU = -W for adiabatic process (Eq. 2.23).
Sometimes, an equation of the type:
PV" = a constant (2.56)
is employed. Here, n is not equal to y. It is known as a polytropic process.
EXAMPLE 2.2 Consider a three-step cyclic process, as shown in Fig. 2.5. The working
substance is 1 kg-mot of a diatomic ideal gas initially at 300 K and 1 atm. The gas is heated
at constant volume to 500 K, then expanded adiabatically to the initial temperature of
300 K, and finally compressed isothermally to initial pressure of 1 atm. Each step occurs
reversibly.
State 1
P 1 = 1 atm State 2
T1 = 300 K LIV= 0 Ti= 500 K
State 3
T3 = 300 K
Fig. 2.5 Stage-wise cyclic process corresponding to the data given in Example 2.2.
(a) Find out the initial as well as the other two states after each step, in terms of
pressure, volume and temperature.
(b) Verify the following data table.
Path dU &I w q
(all values in kilojoules)
I 4157 5820 0 4157

II -4157 -5820 4157 0
III 0 0 -3 185 -3185
Cycle 0 0 972 972
(I + II + III)
Solution (a) The P, V, T of various states:

State 1: P1 = 1 atm, T1 = 300 K
300
V1 = 103 x (22.4 x 10- 3) x
273
m3 = 24.62 m3
500
State 2: T2 = 500 K, V2 = 24.62 m 3, P 2 = 1 x
300
= 1.6 atm
State 3: T3 = 300 K
From Eq. (2.48),
7
For a diatomic gas, y = - = 1.4
5
Fi. st Law of Thermodynamics
Hence,
T2 )11cr-1> ( 500) 11c1.4-1>

V3 = V2
( T = 24.62 = 88.29 m 3
3 300
P 3 V3 = nRT3 , where 11 = No. of moles = 103

Thus,
3
P3 -- 10 x (82.06 x 10-{j) x 300 -- 0 •279 atm
88.29
[Note: R = 82.06 x 10-{j m 3atm/mol.]
(b) Step I: LlU = q - W = q - 0 (isochoric) = q
= nCv (T2 - T1) = 103 x (2.5R) x (500 - 300) x 10- 3

= 103 x 2.5 x 8.314 x 200 x 10- 3 = 4157 kJ
[Note: R = 8.314 J/mol.]
Lili = nC,,(T2 - T1) = 103 x 3.5 x 8.314 x WO x 10- 3 = 5820 kJ
Step II: LlU = q - W = 0 (adiabatic) - W = - W
= nC.)..T3 - T2) = 103 x (2.5 R) (300 - 500) x 10- 3

=- 4157 kJ
Therefore, W = 4157 kJ, Lili = - 5820 kJ
Step Ill: LlU =q - W =0 (isothermal)

Thus,
P dV = nRT In -1..
v. (from Eq. (2.13))
V3
24 62
= 103 x 8.314 x 3oo(tn · ) x 10-3
88.29
= - 3185 kJ
From Eq. (2 .24), H =U + PV. Hence,

LlH = LlU + Ll(PV)
For isothermal process and an ideal gas, LlU = 0, and Ll(PV) = Ll(nR7) = 0. Thus, LlH = 0.
2.8 Summary
1. The statements of the First Law of Thermodynamics are:
L1U = q - W, for a finite process
dU = liq - bW, for an infinitesimal process.
2. U is internal energy of the system, and is a state property. U is a function of
temperature only for a fixed mass of an ideal gas.
3. q is heat absorbed by the system from the surrounding, and W is work done by the
system on the surrounding.
4. q and W are energy in transition, are path dependent and, not state properties/state
variables.
S. bW = P dV + bW', where P dV is work done against pressure, and bW' is any other
form of work (electric, magnetic, surface etc.). For all topics up to Chapter 12,
bW' = 0, and OW= PdV.
6. For a reversible process, the path is known. Hence, W and q can be calculated. For
a finite reversible process going from state 1 to state 2, if bW' = 0, then
V2
W =
f
v,
P dV, where P is the pressure in the system.
7. Enthalpy H is defined as H = U + PV, and is a state property. At constant pressure,

liq = dH, i.e. q = &.
8. For an ideal gas of fixed mass, H is a function of temperature only.
9. (a) Molar heat capacity at constant volume = Cv = (~l
(b) Molar heat capacity at constant pressure = C P = ( ~) P
10. For an ideal gas, Cp - Cv = R.

11. For adiabatic expansion-compression of an ideal gas, P\fY =a constant, where
r= Cp/Cy.
PROBLEMS
2.1 Calculate the difference between ,1H and L1U per mole (i.e. g-mole) of the reaction:
2H2 (g) + 02 (g) = 2H20 (g)
at 500°C and 1 atm pressure.
2.2 One mole of an ideal monatomic gas is compressed from an initial volume of 20
litres to 10 litres by a reversible polytropic process, following the law:
PVl. 2 = constant. The initial temperature of the gas was 100°C.
Calculate:
(a) W, q for the process
(b) LJU, &I between the final and initial states of the gas.
2.3 3.5 moles of an ideal diatomic gas is initially at 500 K and 18 atm pressure. It
undergoes reversible isothermal expansion from 18 to 8 atm pressure, and then
further reversible adiabatic expansion from 8 atm to 1.6 atm. Calculate: W, q, LJU,
&I for each stage.
2.4 Calculate W, q, .dU, &I for reversible expansion of 2 moles of oxygen gas by the
adiabatic and isobaric paths. Assume oxygen to behave as ideal gas. At the initial
state, P = 2 atm, T = 273 K. The final volume is double that of the initial volume.
2.S A system comprises one mole of an ideal monatomic gas at 0°C and 1 atrn. The
system is subjected to the following processes, each of which is conducted reversibly:
(a) 10-fold increase in volume at constant temperature
(b) 100-fold adiabatic increase in pressure
(c) return to the initial state along a straight line path in the P-V. diagram.
Calculate the work done by the system in step (c), and the total heat added to the
system as a result of the cyclic process.
Heat Capacity and Enthalpy-
Auxiliary relations
and applications
3.1 Heat Capacity of Ideal Gases

According to the Kinetic Theory of Gases, a monatomic gas molecule (He, Ar etc.) has three
translational degrees of freedom. Each degree of freedom contributes 1/2R to Cv per mole
according to the Law of Equipartition of Energy. Hence, Cv of one g-mole of a monatomic
gas is equal to 312R, which is approximately 12.5 I mo1-1K- 1 Again, Cp = Cv + R,
according to Eq. (2.42). Hence, Cp of a monatomic ga~ should be 512R, and should be
independent of temperature. This has been verified by experimental measurements.
A diatomic gas molecule (H2 , 0 2 , N 2 etc.) has two additional degrees of freedom arising
from rotation. This gives a value of Cv as 512R .and Cp as 7/2R. This generally holds good
at room temperature. However, experiments have demonstrated that Cp increases to almost
9/2R as temperature is raised. This is due to oscillation along the line joining the two atoms
of the molecule. Such oscillation is associated with two additional vibrational degrees of
freedom since it has both potential and kinetic energy.
For polyatomic gases, the situation is more complex. Hence, it is customary to employ
experimentally measured (i.e. empirical) heat capacity vs. temperature relation for diatomic
and polyatomic gas. Such relations have been proposed for solids and liquids as well, and
are discussed in section 3.2.3.
3.2 Heat Capacity of Solids

In 1819, Dulong and Petit, on the basis of experimental data available then, proposed that
26
Heat Capacity and Enthalpy-Auxiliary relations and applications
molar heat capacities of all solid elements may be taken as 3R per°C. Further measurements
subsequently showed this rule to be approximate and very limited in scope. Figure 3.1
presents Cv vs. T experimental data for a few solid elements at low temperature. It shows
that:
(a) Cv --+ 0 as T --+ 0

(b) Cv increases with increase in T, but curves are different for different elements.
(c) In general, at high temperatures, Cv --+ 3R for some elements.
25
20
,,.....
~0 15
s
-;;;
u
:; 10
0
0
:::..
t>
5
Temperature (K)
Fig. 3.1 Constant-volume molar heat capacities of Pb, Cu, Si and diamond as function of
temperature.
3.2.1 Einstein's Theory of Heat Capacity of Solids

In the early 20th Century, Einstein applied quantum theory to explain experimental data on
C v of solids elements. Each atom in the crystal was assumed to behave as independent
harmonic oscil1ator with oscillations along the three perpendicular directions. The energy of
the ith level of harmonic oscillation is given as
(3.1)
where vis oscillation frequency and, h is Planck's constant. Again,

U = 3:En;E; (3.2)
where n; = number of atoms in the ith energy level.
Combining Eqs. (3.1) and (3.2) and using the relation between n; and Avogadro's
number (N0), we get
(3.3)
where k8 is Boltzmann's constant. If exp (-hvlk 8 T) = x, then
1
I: exp (-hvilk 8 T) = I:x; = 1 +x +x 2 + ... = - - if x « (3.4)
i 1 - x
Following Einstein's derivation steps (being skipped), we get
3 3N hv
U =-Nohv
2
+ (e h vlk T0
B 1)-
(3.5)
and
c -(au)
v - oT v -
-3R(BE)z_e_Br;tr_
T (elJelT - 1)2 (3.6)
(since R = N0 k8 ), where
hv
lk = Einstein temperature = (3.7)
ks
3.2.2 Debye's Theory of Heat Capacity of Solids
Einstein's derivation was to be later modified by Debye who assumed a spectrum of

vibration frequency in contrast to a single frequency of oscillation of Einstein. Debye
arrived at a different equation in terms of Debye temperature (0o), where
At very low temperatures (near 0° K), the Debye equation may be simplified as
Cv = 464.5(T/B0 ) 3 (3 .8)
The Debye equation gives better fit with experimental data, provided the best fitted
values of 8 0 are employed. Even then, it is not satisfactory. Moreover, it is concerned with
elements only, that too in the crystalline state. Hence, due to its limited applicability, it is
a normal practice to experimentally determine heat capacity vs. temperature relationships
for solids and liquids.
3.2.3 Empirical Cp vs. T Relations
Experimental data consist of Cp as a function of temperature. For solids and liquids, the
PV term is very small. Hence, H is taken as equal to U, and Cp as equal to Cv. In other
words, no distinction is made between Cp and Cv so far as applications are concerned. It
has been found that experimental Cp vs. T data for elements and compounds fit best with
an equation of the type:
Cp =a + bT + cr2 (3.9)
where a, b, c are empirically fitted constants and differ from substance to substance.
The last term is the smallest and, therefore, often ignored. In some cases, such as liquid
metals, both bT and cr-2 are usually ignored. Equation (3.9) is employed for correlation
of experimental Cp data of diatomic and polyatomic gases as well. For alloys, where
experimental data are not available, approximate estimates are done assuming heat capacity
to vary linearly with atom fraction.
3.3 Enthalpy Changes

As already presented in Section 2.5, change of enthalpy during the course of a process (,1//)
is equal to the heat supplied to the system (q) at constant pressure, and that ·this is the basis
for process heat balance and thermal energy calculations. Energy is costly. There is continuous
attempt to save energy in all kinds of activity, including technological processing. By
energy conservation, we essentially mean saving of fuels such as coal and petroleum. These
are burnt to produce thermal energy, which is either utilized directly or converted into
electrical energy.
The constant pressure restriction is mostly not important. H is not a function of P for
ideal gases (see section 2.5.1 ). Energies of solids and liquids are hardly affected by some
changes in pressure due to their very small molar volumes. In other words, the V dP term
in Eq. (2.25) is negligible. In most metallurgical and materials processing, gases are ideal
and pressure is maximum a few atmospheres. Therefore, Ml = q approximation is quite
all right for almost all applications of our interest in the area of materials and metallurgical
processing.
3.3.1 Classification of Enthalpy Changes

The absolute value of enthalpy of a substance is not known. All we can measure are changes
of enthalpy. Again, enthalpy is an extensive property, and hence depends on the mass of
the substance. By Universal convention, standard thermochemical data books provide enthalpy
changes per mole of a substance.
Enthalpy changes occur due to various causes, and are classified as follows . Since heat
is to be supplied to the substance or removed from it for enthalpy changes, the term heat
is traditionally employed, although heat is not a state property, but energy in transition.
Sensible heat (i.e. sensible enthalpy)

Enthalpy change due to change of temperature of a substance is known as sensible heat. It
is Subdivided into:
(i) Change in enthalpy without any change in state of aggregation of the substance. It
has already been presented in Eq. (2.36). As a universal convention, thermochemical data
books take sensible heat at 298 K (25°C) as zero for any substance. Hence, sensible heat
at temperature T per mole of a substance, is given as
HT - H298 = f T
298
Cp dT (3.10)
298 K (strictly 298.15) is known as reference temperature. Combining Eq. (3.10) with
Eq. (3.9), we get
HT - H 298 = fT (a+ bT + cT-2 ) dT

298
= a(T - 298) + -b (T 2 - 298 2 ) - c - - -

2 T 298
(1 1)
=AT+ BT2 - CT1 + D (3.11)
where A, B, C, D are lumped parameters and functions of empirical constants a, b, c.

(ii) Enthalpy changes due to changes in state of aggregation of substance. These
are isothermal processes. By convention, enthalpy changes for all isothermal processes are
designated by LJ.H. For example,
LJ.Hm =enthalpy change of one mole of a solid due to melting (i.e. latent heat of fusion
per mole)
LJ.Hv =enthalpy change per mole of a liquid due to vaporization (i.e. latent heat of
vaporization per mole)
Consider a pure substance A, which is undergoing the following changes during heating
from 298 K to T K .
A (solid) A (liquid) ~ A (gas) A (gas)
at 298 K at Tm at Tb at T
A is either a pure element or a pure compound. Tm and Tb are respectively, the melting
and boiling points of A. Then,
HT - H 29s f
= T,,.
298
Cp(S) dT + &lm + JTb Cp(l) dT +&Iv +
T,,.
JT
Tb
Cp(g) dT (3.12)
where Cp(s), Cp(I) and Cp(g) are Cp's for solid, liquid and gaseous A, respectively. If the
substance is not a pure element or compound, then it does not have a single melting or
boiling point, and hence Eq. (3.12) is not applicable.
Sometimes an element or a compound undergoes phase transfonnation at the solid state.
Then these terms are also to be added. An example is iron which exhibits the following
structures at the solid state.
Phase Structure Temperature range (K ) of stability

a Body-centred cubic (BCC) up to 1033
Nonmagnetic BCC 1033-1186
Face-centred cubic (FCC) 1186-1665
Body-centred cubic 1665- 1809
8-iron melts at 1809 K. Therefore, the sensible heat of iron at 2000 K is given as:
1033 f 1186
H'lJXXJ -H298 =
f 298
Cp(a)dT+L1111r(a ~ /3)+
1033
Cp(/3)dT
+ tlH m + f 2000
1809
Cp(l) dT (3. 13)
EXAMPLE 3.1 Calculate L1ll and L1.U when one mole of water at 25 °C at 1 atm pressure
is conv.erted into steam at 130°C and 2 atm.
Some relevant data (in J/mol) are now noted:
Cp of water = 75.3 per K, Latent heat of vaporization of water (L111v) = 40.64 x 103,
Cp of steam between 100 - 130°C = 30.11 + 9.937 x 10- 3 i-per K.
Solution L1ll = (L111hs•c-+ 130°C at I atm + (,111) I atm-+ 2 atm at 130°C·

Since steam is at low pressure, it may be treated as an ideal gas. Hence, (,111) steam
for going from 1 atm to 2 atm at l 30°C is zero, and
,111 = (,1/1)25•c-+ 130°C (at 1 atm)
100+273 f 130+273
= f 25+273
Cp(water) dT + L111v +
100+273
Cp(steam) dT
403
=75.3 x 75 + 40.64 x 10 + [30.11 T + 9·937 x 10-3r 2]
3
2 373
= 47 .3 1 x 103 J/mol
LlU = Mi - Ll(PV) [from Eq. (2.24)]

Since PV is a state variable, Ll(PV) can be calculated along an y path. Let us choose the
followi ng path:
Ll( P V) = [Ll(PV)hs c-+ 130°C at l

0
atm + [Ll(PV)]1 atm-+2 atm at 130°C
Ll( PV) at l 30°C = Ll(R1) at l 30°C =0 s ince T is constant.
Hence,
Ll(PV) = [Ll(PV)hs•c-+ t30°C at 1 atm
= [Ll(PV)] for heating water from 25°C to 100°C
+ [Ll(PV)] for vapori zation of water at l00°C
+ [Ll(PV)] for heating steam from 100°C ~ 130°C at l atm
Since PV change is negligible for heating of water,
Ll(PV) =P(Vs - Vw) + [Ll(PV)]~~

where
P =1 atm,
V5 = molar volume of steam at 100°C (i .e. 373 K )
373
=22.4 x 10- 3 x - =30.6 x 10-3 m 3
273
Vw "' 0 since it is very small.
Hence,
P(V5 - Vw) = 1 X 30.6 X 10-3 m3 atm
= 30.6 x 10- 3 x 1.013 x 105 = 3100 joules
[Ll(PV)]m = [Ll(RT)]~~ =8.314(403 - 373) =250 joules
Therefore,
LlU = 47 .31 x 103 - 3100 - 250 = 43.96 x 103 J/mol
Heat of reaction (L1H)

Thi s is the change of enthalpy that occurs when a reaction takes place. By convention,
the reaction is considered to be isothermal. Consider the followin g reaction occurring at
temperature, T:
A (pure) + BC (pure) = AB (pure) + C (pure) (3. 14)
where A, B, C are elements and AB, BC are compounds. The n,
&/(at 7) = HA 8 (at 7) + Hc(at 7) - H A(at 7) - H8 c(at 7) (3. 15)
where HAB. H e. HA, H 8 c are molar enthalpies of pure AB , C, A, BC, respecti vely.
Heat of mixing (L1Hm1x)

The process of dissolution of a substance in a solvent (e.g. NaCl dissolution into water) is
generalized now.
A (pure) =A (in solution) (3.16)
This process is also accompanied by a change of enthalpy (L1Hmix), where
L1Hmix = H (in solution) - HA (pure) (3.17)

Again, by co11ventio11, the process is assumed to be isothermal for thermodynamic calculations.
Some comments
For calculation of enthalpy changes, reactions, dissolutions and phase transformations have
been assumed to occur isothermally. Since enthalpy is a state property, L1H depends only
on initial and final states, and not the path. The above assumption provides the most
convenient path for calculation. Figure 2.1 has schematically shown variation of temperature
with time when an isothermal process occurs. For a reversible isothermal process, the
temperature remains constant all through . If the process is not reversible, then temperature
at beginning and end of a process would be the same. In between, the temperature can vary
significantly.
For calculation of enthalpy changes, the isothennal processes are general (i.e. irreversible
or reversible). Only the initial and final temperatures are assumed to be the same. This is
illustrated by Fig. 3.2 with the example of chemical reaction (3.14).
t
u
3
(U A+ BC AB+ C
:U Ti----.....---
c..
E
t
u Initial state Exothermic Final state
E-< (state l) reaction (state 2)
begins
Process ~
Time---•
Fig. 3.2 Variation of temperature with progress of an irreversible exothermic isothermal reaction
(schematic).
3.3.2 Conventions for !JH and H

It has already been mentioned that thermochemical data sources provide (Hr- Hi 98) and t1H
values per mole (i.e. molar values).
The sign convention for tJH

(a) Processes accompanied by liberation of heat are called exothermic. This happens if
the enthalpy in the final state (state 2) is lower than that in the initial state (state 1), i.e.
Hi < H 1• So for the process
State l ~ State 2,
we have
t1H = Hi - H1 < 0
Therefore, t1H is negative. Figure 3.2 illustrates an exothermic reaction. The opposite is an
endothemiic process which is characterized by absorption of heat and a positive value of
tJH.
(b) Sensible heat (H) as well as t1H for isothermal processes are functions of temperature.
Hence, the temperature for which the value is being used is to be indicated. This is done
in several ways. The first approach is to state the temperature separately, and not to use any
symbol. Alternatively, these may be written as Hr. '1Hr. t1H(1) to denote that these are at
temperature T K.
Standard state for enthalpy

A substance may be pure or it may be present in a solution. Again, the stable state of a
pure substance changes with temperature. The stable state of Hi O is ice below 0°C, liquid
water from 0-100°C, and a stable gas at 1 atm pressure above 100°C. Considering all these
points, a standard state has been defined as a pure element or compound at its stablest state
at the temperature under consideration and at one arm pressure. Thus, the standard state
of HiO at 50°C is pure water under 1 atmospheric pressure (i.e. standard atmosphere =
760 mm Hg).
By convention, enthalpies and enthalpy changes at standard state are denoted by superscript
'O', e.g. H~ , Lili~.
As already mentioned, 298 K is the universal reference temperature for compilation of
sensible heats. By this convention, sensible heat at standard state of a substance (element
or compound) is arbitrarily taken as zero at 298 K. However, this is solely for calculation
of sensible heats, not MfJ for a process occurring at 298 K. For example, MfJ (298) for
reaction (3.14) is not zero. The enthalpy of a substance is largely or almost wholly due to
its in~ernal ·energy. In Section 2.1 , we have mentioned the various sources of internal
energy. Temperature, which is a measure of thermal potential , contributes to kinetic energy
of atoms and molecules. This part comes under sensible heat. Other forms of internal energy
are not very much influenced by temperature unless we go to extremely high temperature.
Mic refers to heat effects when the reaction is concerned with formation of a compound
from its elements. Standard heat of formation at a temperature T, i.e. [L111f(Dl is nothing
but L1111 at T, when elements and their compound are aJI at their respective standard states.
3.3.3 Experimental Measurements of Enthalpy Changes
Enthalpy changes of a substance are associated with heat effects-either absorption or

evolution of heat. It has been stated in Chapter 1 that application of thermodynamics
requires experimental data. Measurement of heat effects is one such requirement. It is done
using a calorimeter. There are many designs of calorimeters.*
A calorimeter is an apparatus with known heat capacity. Therefore, from measurement
of change of its temperature as a result of heat absorption/evolution due to the reaction/
process, the quantity of heat can be estimated. For accurate measurements, as is the norm
in thermochemistry, te!Ilperature changes have to be measured precisely (always within
±0.01 °C and even lower). Mercury-in-glass thermometers, platinum resistance thermometers
as well as thermocouples have been employed for this purpose.
The most widely used type of calorimeter is the isoperibol Calorimeter, in which the
enclosure is held at constant temperature and the temperature of the substance undergoing
a reaction/process is measured as a function of time. Since the calorimeter cannot be isolated
from the surrounding completely, there will be some heat exchanges between them, causing
errors. Such errors have to be eliminated through proper corrections.
In early days, calorimeters were essentially at room temperature. Even now that is true.
However, over the last few decades, various designs of high temperature calorimeters going
up to 2000°C have been develo~ and employed for measurements.
For measurement of specific heat of a substance, the latter is heated to a high temperature
in a furnace and suddenly dropped into the calorimeter bath (typically water). From rise of
bath temperature, specific heat is estimated.
3.3.4 Hess' Law

There are literally millions of chemical reactions. It is extremely difficult, if not impossible,
to experimentaJly determine heats of so many reactions. Hess proposed a law, which enormously
simplified the situation. Consider reaction (3.14). It may be considered as consisting of the
following formation reactions at temperature T.
A(pure) + B(pure) = AB(pure); L1Hf. AB (T) (3 .18)
B(pure) + C(pure) = BC(pure); L1Hf 8 c (D (3.19)
•See 0 . Kubaschewski, P.J. Spencer. and C.B. Alcock, Materials Thermochemistry, 6th ed., Pergammon
Press, Oxford, 1993; see also earl ier editions by 0 . Kubaschewski, E.LI. Evans, and C.B. Alcock). Some
standard data sources have been given in the Appendix to the book.
Subtraction of Eq. (3.19) from Eq. (3.18) yields Eq. (3. 14), i.e.
A (pure) + BC (pure) = AB (pure) + C (pure); L1l-1°(7) (3.20)
According to Hess' Law, therefore,
L1JJO (T) =L1H?, AB (T) - L1H?, BC (T) (3 .21 )
Hence, L1l-I° of any reaction can be calculated, if the experimental values of m? of the
relevant compounds are available.
Enthalpy is a state property. Hence, we may write
f dH = 0, over a cycle (differential form)
or
:E L1H
cycle
=O for finite stages (3.22)
For reaction (3.14), application of Eq. (3.22), which is the basis of Hess' Law, is
illustrated by Fig. 3.3, where
L L1Ho
cycle
=M-1~ BC (T) + mo (T) -
'
m~ AB (T)
'
=0 (3.23)
i.e. L1H 0 (T)=t1H?,AB(T) - L1H~Bc (T) , which is Hess' Law [given in Eq. (3.21)].
A+B+C
(at 1)
A+ BC AB+ C
(at 1) (at 7)
Fig. 3.3 Illustration of thermodynamic basis of Hess' Law; A, B, C, AB, BC are pur& and one
mole each .
Cyclic processes and the concept of cycle have been employed in thermodynamics and
other subjects. In connection with computer programming, the term loop is more popular.
Hence, the latter term is also being advocated now. Johnson and Stracher (1995)* have
widely employed loops for problem solving.
EXAMPLE 3.2 Calculate .1H~98 for the reaction: 2Ca0 + Si02 = Ca 2Si04
Given: t1H~ for the formation of CaO, Si02 and Ca2Si04 from their elements at 298 K
are -634, - 910.9 and - 2305.3 kilojoules per mole (kJ mole- 1), respectively.
Solutio11 Mi~98 = Ml?, 298 (Ca2Si04) - 2 Ml?, 298 (CaO) - Ml?, 298 (Si0 2)
= -2305 .3 - (-2 x 634 - 910.9] = - 126.4 kJ per mole of reaction
3.3.5 Kirchhoff's Law
t1H for an isothermal process depends on the temperature at which the process occurs. The
relationship of t1Ff of a reaction with temperature and other quantities is known as Kirchhoff's
Law. Again, like Hess' law, it is based on the definition of enthalpy as a state property.
The derivation is being illustrated for reaction (3.14) by application of the loop in Fig. 3.4.
(3.24)
or
(3.25)
Equation (3.25) is the generalized version of Kirchhoff s Law. If the reactants and products
do not have any phase changes between T1 and T2, then Eq. (3.25) can be simplified as
M1~2 = m~1 + J Ti
7j
[ roducl:
p t
c~ - l:
reactant
c~J dT = m~ I
+ J L1C~
r
7j
2
dT (3.26)
If the reactants and products are pot pure and not in their standard states, then also
Kirchhoffs Law can be employed. However, the compositions of the solutions should not
change with temperature. For melting and other phase transformations, the equation gets
simplified. For example, for melting of solid,
[Cp(l) - Cp(s)] dT (3.27)
*D.L. Johnson and G.B. Stracher, Thermodynamic Loop Applications in Materials Systems, Minerals, Metals
Materials Society, Warrendale (Pennsylvania), 1995.
T-----r
Mio
A + BC Tz AB + c
T2
t
...
t>
(Ho
T2 (Ho -
T2 H~.>c
......
:J
"' (H~2 - H~1)A Loop

(Ho - H T1JAB
o'
8. T2
J______
E
t>
E-o
r,
__ J_____
A + BC Mio
T1
AB + c
Process path - - +
Fig. 3.4 Illustration for derivation of Kirchhoff's Law.
EXAMPLE 3.3 Calculate the adiabatic flame temperature when C 2H4 gas is ignited at
298 K with stoichiometric amount of oxygen.
Given: (i) Heat of combustion (&/) of C 2H 4 at 298 K is -1411 .3 kJ per mole of C 2H4
(with gaseous reference state of H 20 at 298 K).
(ii) Cp of gases (J/mol/K)
C02 26.76 + 42.26 x 10-3 ' T

H20 30.2 + 9.93 x 10-3 T
02 25.5 + 13.61 x 10- 3 T
Solution If a fuel is assumed to burn under adiabatic condition (i.e. no heat loss to
surrounding), then the resulting temperature of the products of combustion is known as
adiabatic flame temperature (Ta). The consequent heat balance equation may be written as:
Sensible heats of reactants at 298 K + hear released due to combustion at 298 K = heat
required to raise temperature of products from 298 K to Ta. The sensible heats of reactants
(i.e. C 2H 4 and 0 2) at 298 K are zero by definition of sensible heat. The combustion
reaction is:
(E.3.1)
Therefore, the heat balance equation (per mole C2 ~) is
Ta
1411.3 x 103 + O =
f
298
[2Cp(COi) + 2Cp (H20)] dT (E.3 .2)
Substituting the values and rearranging the terms in Eq. (E.3.2), and on integration, we get
3
10
1411.3 x 103 = 113.9 (T" - 298) + l04.4 x - (T"2 - 2982 ) (E.3.3)
2
Equation (E.3.3) is a quadratic equation. By rearranging the terms, it becomes
52.19 x 10- 3 r; + 11 3.93Ta - 1450 x 103 = 0 (E.3.4)
The valid solution is
T = - 113.93 + [(113.93) 2 + 4 x 52.19 x 1450]112

(I 2 x 52.19 x 10-3
= 4290 K
[Note: H 20 is a stable liquid at 298 K. If that is employed as reference state at 298 K,
the value· of Ml for combustion is to be adjusted as: Ml =-
1411.3 - 2Mlv, where Mlv
is the molar heat of vaporization of H 20. On the RHS also 2M:lv is to be subtracted. Thus,
these will automatically get cancelled, and the answer will be the same.]
EXAMPLE 3.4 An adiabatic vessel contains 1 kg of liquid aluminium at 700°C. Some

Cr20 3 at 25°C was added into it, which caused the following exothermic reaction to occur:
The final temperature was 1000°C, and the fina1 product mixture consisted of AI20 3(s),
Cr(s), and some unreacted Cr20 3(s). Calculate the weight of Cr20 3 addition.
Solution Heat balance (with 298 K as reference temperature):
Sensible heat of reactants + heat released in reaction at 298 K

= sensible heat of products (since adiabatic)
(a) Since Cr20 3 was added at 298 K, its sensible heat is zero. Hence,
Sensible heat of reactants = sensible heat of liquid Al at 973 K
= lOOO
27
[J
298
932
Cp [Al(s)]dT+.1Hm(Al)+f
913
932
Cp [Al(l)]dT]
(E.3 .5)
Since the atomic mass of Al = 27, and M.P. of Al = 932 K.
Noting
Cp[AI (s)] = 20.67 + 12.38 x 10-3 T J/mol/K, Cp[AI (I) ] = 29.29 J/mol/K,
and
Lllim(AI) = 10.46 x 103 J/mol ,
we have sensible heat of reactants = 10.96 x 105 j oules
(b) Heat of reaction at 298

=!:ili~98 =!:ill~ [Al 2 0 3 (s)] 298 - !:ili~[Cr2 0 3 (s)]i98
= - 16.74 x 105 + 11.17 x 105 = - 5.57 x 105 joules

Hence,
Heat released = 5.57 x 105 joules/mot
Therefore,
Total heat released = nAI x 5.57 x 105

2
= 103.15 x 105 joules
where nA 1 = initial no. of g. moles of Al
1000
=- =37.04
27
(c) The product has excess Cr20 3, but no excess Al. Hence,
1
nc,.,., 0
.• 3
(added)= x > -2 nA1
So,
. od 1 37.04
nA 1 0 (m pr uct) = - nA1 = - - = 18.52
2 3 2 2
ncr (in product) = nA1 = 37.04

Therefore,
ncrio3 (in product)= x - ~ nA1 = x - 18.52
The sensible heat of products at 1000°C (1273 K ) is

l273
J298
{ 18.52 Cp[Al 2 0 3 (s)] + 37.04 Cp[Cr(s)] + (x - 18.52)jCp[Cr20 3(s)] }dT
(E.3.6)
2
Cp = a + bT + c1 J/mol/K, with the following values of a, b, c:
a bx 103 C X 10-S
106.6 17.78 -28.53

24.43 9 .87 -3.68
119.4 9.2 -15.65
Heat Capacity and Enthalpy- Auxiliary relations and applications
Putting in the values in Eq. (E.3.6), and carrying out integration, we obtain
Sensible heat of products = 1.23 x 105 x + 9.58 x 105
Hence, the heat balance is
10.96 x 105 + 103. 15 x 105 = 1.23 x I0 5x + 9.58 x 105
i.e. 10.96 + 103.15 - 9.58 = l.23x or x = 85 g-moles
or weight of Cr20 3 added = 85 x (2 x 52 + 3 x 16) x 10- kg 3
= 13.0 kg.
3.3.6 Concluding Remarks

Some topics, concepts and conventions introduced in this chapter, such as standard state,
isothermal process , loop application of state properties leading to Hess' Law and
Kirchhoff's Law, would be used again for other important state properties, which will be
introduced later.
Books on chemistry deal with chemical bonds. These bonds hold the atoms together in
a material. The energy associated with a bond is known as bond energy. During the
formation of a compound from elements, bonds are formed. This causes lowering of energy
of the material and liberation of heat (i.e. Afl is negative) . Dissociation of a compound
requires extra energy and, hence, heat is absorbed (i.e. ilH is positive). Therefore, Afl is
often explained in terms of changes in bond energies. This is a simple way to look at
enthalpy changes of a substance in chemical thermodynamics. However, it is only an
approximation.
3.4 Summary
1. For ideal monatomic gases and ideal diatomic gases at not too high temperature,
molar heat capacities at constant volume (Cv) and at constant pressure (Cp) are
given by the Kinetic Theory of Gases.
2. For all other substances, the values of Cp as a function of temperature are determined
experimentally, and are expressed by the relation
Cp =a + bT + cr-2
where a, b and c are empirical constants.
3. For solids and liquids, Cp and Cv are approximately the same.
4. Enthalpy changes are due to
(a) change of temperature of a substance, and
(b) other reactions and processes, e.g. phase transformations (i ncluding melting and
vapourization), reaction, and mixing; these are all treated as isothermal processes.
5. Sensible heat (i.e. sensible enthalpy) of a substance is denoted as Hr - H 298 ,

298 K being the Universal reference temperature adopted by thermochemical data
books. For isothermal processes, enthalpy change is denoted by .&/.
6. At standard states of substances, H is designated as JfJ, e.g. H~ - H~, &fJ

7. Calorimeters are employed for measurement of specific heats and enthalpy changes.
8. Hess' Law allows estimation of LJJfJ of any reaction from LJH? (i.e. heats of
formation) of the compounds involved.
9. Kirchhoffs Law aJlows estimation of &fJ of a reaction/process at any temperature
from its value at another temperature.
10. Both Hess' and Kirchhoff's Laws are based on the fact that enthalpy is a state
property.
PROBLEMS
3.1 Assuming no heat is lost to the surroundings, calculate the amount of heat required
to just melt 1 kg of Pb initiaJly at 15°C.
3.2 Consider oxidation of tungsten carbide (WC) as follows:
WC(s) + 5/2 0 2(g) = W03 (s) + C02(g), L1f/~ = - 1195.8 kJ mo1- 1
Calculate .&1? of 1 mole of WC from elements at 298 K by Hess' Law.

3.3 Calculate the standard molar enthalpy of formation of titanium carbide (TiC) at
1200 K.
3.4 Calculate .&1?800 per mole of the reaction:
1
Mn (I) + 2 0 2 (g) = MnO (s)
Note the following transformation steps for Mn from 298 K:
Mn(a) ~ Mn(,8) ~ Mn(n ~ Mn(b) ~ Mn(l),
3.5 Calculate .11.f' and LJv° for the following reaction at 1000 K:
3.6 Which of the following processes releases more heat at 1000 K and by how much?
(a) Oxidation of graphite to CO
(b) Oxidation of diamond to CO
I
Heat Capacity and Enthalpy-Auxiliary relations and applications ,,.

3,7 Calculate MfJ per mole of the following reaction, occurring at 900 K.
2Al (s) + Cri0 3 (s) = Ali0 3 (s) + 2Cr (s)
3.8 Calculate the adi4batic flame temperature for combustion of the following gas
mixture: 25% CO, 12.5% COi. and 62.5% Ni (by volume). The initial temperatures
of the gas and air are 298 K. The theoretical amount of air was used. Assume that
air contains 21 % by volume Oi, rest Ni. 1i
3.9 Calculate the heat of reaction at 1200 K per mole for t}\e following reaction:
I
ZnO (s) + C (graphite) = Zn (g) + CO (g)
[Note: From this chapter onwards, for problem solving, data are to be picked
up from the Data Tables in Appendix, as required.]
J
I
Textbook of Materials and Metaffurgicaf Thermodynamics
S. Carnot (1824) conceptualized an ideal heat engine which operates reversibly between
a heat source of constant temperature T2 and heat sink of constant temperature T 1• Such an
ideal reversible cycle is known as Carnot Cycle (CS), which laid the fou11datio11 of the
Second Law of Thermodynamics.
It may be emphasized here that, for reversibility, not only the process should be carried
out very slowly, but there should be no friction between the cylinder and the piston since
frictional heat dissipation is irreversible.
4.2.2 Carnot Cycle with Ideal Gas as Working Substance
Figure 4.2 shows the Carnot cycle. It consists of four stages: two isothermal and two
adiabatic. The temperature of isothermal stage l ---+ 2 is T2, and that of 3 ---+ 4 is Ti.
T 2 > Ti. All stages are reversible.
t
Adiabatic
reversible
Volume-
Fig. 4.2 P-V diagram for Carnot cycle (schematic).
Stage 1 ~ 2 (isothermal)
L1U = 0 for isothermal expansion/contraction of an ideal gas (section 2.5.1). Hence, from
the First Law, and Eq. (2.13),
q1~2 =: q2 = W1 ~2 = nRT2 In CV2/Vi) (4.4)

where n = No. of moles of gas.
Stage 2 ~ 3 (adiabatic)
q2~3 = O; hence,
(4.5)
3,7 Calculate .l!JJ° per mole of the following reaction, occurring at 900 K.
2Al (s) + Cr20 3 (s) = Al20 3 (s) + 2Cr (s)
3.8 Calculate the adiabatic flame temperature for combustion of the following gas
mixture: 25% CO, 12.5% C0 2, and 62.5% N2 (by volume). The initial tem?eratures
of the gas and air are 298 K. The theoretical amount of air was used. Assume that
air contains 21 % by volume 0 2, rest N2•
3.9 Calculate the heat of reaction at 1200 K per mole for t~e following reaction:
ZnO (s) + C (graphite) = Zn (g) + CO (g)
[Note: From this chapter onwards, for problem solving, data are to be picked
up from the Data Tables in Appendix, as required.]
Second Law of Thermodynamics
and Entropy
4.1 Introduction
The First Law of Thermodynamics merely states that if a process occurs, an exact equivalence
exists amongst various forms of energy changes. It provides no information regarding the
feasibility of the process. The Second Law of Thermodynamics provides means to predict
whether a particular process/reaction would take place under certain specified conditions,
and thus is of great importance. Another important aspect of the Second Law, which is
really fundamental to the problem enunciated above, is concerned with conversion of heat
into work.
4.1.1 Various Statements of the Second Law

The Second Law of Thermodynamics has been stated in several equivalent forms since the
time it was enunciated, implicitly by Carnot (1824), and explicitly first by Clausius (1850)
and later independently by Kelvin (1851). According to Clausius and Kelvin, and as reformulate.cl
by Max Planck, the following statement ·is significant:
It is impossible to construct a cyclic engine that can convert heat from a reservoir at a
uniform temperature completely into mechanical energy without leaving any effect elsewhere.
The engine must be cyclic, i.e. it returns to its initial position (i.e. state) after each
stroke or revolution.
Some alternative statements are:
1. Heat absorbed at any one temperature cannot be completely transformed into work
without leaving some change in the system or its surroundings.
2. Spontaneous processes are not thermodynamically reversible:
44
Second Law of Thermodynamics and Entropy
A spontaneous process occurs without external intervention of any kind. Examples are
flow of heat from higher to lower temperature, diffusion of a species from higher concentration
to lower concentration, mixing, and acid-base reactions. All natural processes, i.e. processes
occurring in nature without external intervention, are spontaneous. In a refrigerator, heat is
pumped out from lower to higher temperature by an artificial device. This device consumes
electrical energy provided from outside.
4.2 Carnot Cycle
4.2.1 Introduction
A device utilizing heat to generate mechanical work and operating in a cycle is called a heat
engine. The substance contained in it is called the working substance. Its operating feature
is illustrated in Fig. 4.1.
work obtained per cycle W

Efficiency of the engine (7]) = =- (4.1)
heat absorbed per cycle q2
q)
Heat q2 Heat Heat
source at -
~
engine sink at
•
T2 -
q•
Ti
•
Work(W)
Fig. 4.1 Heat engine (schematic).
From the First Law of Thermodynamics,

1: &.!
cycle
=0 =net heat absorbed - work done
(4.2)
(Note: q denotes heat absorbed by the system, i.e. the engine; hence by convention
q) = - q1)
From Eqs. (4.1) and (4.2),
w q2 + q.
1] =-=~-~
q2 q2
(4.3)
.•
S. Carnot (1824) conceptuali zed an ideal heat engine which operates reversibly between
a heat source of constant temperature T2 and heat sink of constant temperature T 1• Such an
ideal reversible cycle is known as Carnot Cycle (CS), which laid the foundation of the
Second Law of Thermodynamics.
It may be emphasized here that, for reversibility, not only the process should be carried
out very slowly, but there should be no friction between the cylinder and the piston since
frictional heat dissipation is irreversible.
4.2.2 Carnot Cycle with Ideal Gas as Working Substance

Figure 4.2 shows the Carnot cycle. It consists of four stages: two isothermal and two
adiabatic . The temperature of isothermal stage 1 ~ 2 is T2 , and that of 3 ~ 4 is T 1•
T2 > T 1• All stages are reversible.
Isothermal
i
Adiabatic
reversible
Volume-
Fig. 4.2 P-V diagram for Carnot cycle (schematic) .
Stage 1 --+ 2 (isothermal)

LJU = 0 for isothermal expansion/contraction of an ideal gas (section 2.5.1). Hence, from
the First Law, and Eq. (2. 13),
(4.4)
where n = No. of moles of gas.
Stage 2 --+ 3 (adiabatic)
q 2 _,. 3 = O; hence,
W2...,.3 = - (L1Uh...,.3 = - nCv(T1 - T2) (4.5)
Stage 3 --+ 4 {isothermal)

(4 .6)
Stage 4 --+ 1 {adiabatic)
W4;1 = - (L1U)4;1 = 11C1.(T1 - T"J.) (4.7)

Total work done in the cycle
W = W1 ;2 + W 1;3 + W3;4 + W 4;1
Hence,
(4 .9)
For adiabatic expansions, from Eq. (2.48),
(~r
T,
= ~ (4. 10)
r.
(~J' = Ti
T,
(4.11)
Hence,
( 4.12)
W =11R In (~) (T2 - T

1)
(4 .13)
From Eqs. (4.13) and (4.4),
W T2 - T1
17 =-= ~-~ (4.14)
qi T2
Equation (4.14) shows that the efficiency of this reversible cycle depends only on T2
and r,.
------
4.2.3 Efficiency of Generalized Carnot Cycle
In a generalized Carnot cycle, the working substance is any thermodynamic substance. It is
not restricted to ideal gas only. Section 1.3.3 has defined a thermodynamic substance which
is at thermodynamic equilibrium and obeys Eq. ( 1.3). Since a reversible process proceeds
through equilibrium states, it essentially has a working thermodynamic substance.
One of the statements of Generalized Carnot Theorem is:
All reversible cycles operating between the same upper and lower temperatures must
have the same efficiency. The proof of this statement can be given in the following way.
Figure 4.3 shows two engines working in cycles between the same heat reservoirs.
Heat reservoir at T2
Heat reservoir at T1
Fig. 4.3 A hypothetical system for proof of Carnot theorem.
Engine I is working with ideal gas and transferring heat from reservoir at T2 to that at T1•
Engine II is working with another working substance and transferring heat in the reverse
direction (i.e. T1 ~ T2 ). Both engines are reversible.
Assume that
Efficiency of engine I, 711 > efficiency of engine II, 7111 (4.15)
Then ,
W1 qi (I) + qi (I)
711 =--
qi (I)
= qi (I)
(4.16)
Wn qi (II) +qt (II)

1111 = - - = (4.17)
qi (II) q2 (II)
Assume further that
(4.18)
From Eqs. (4.16)-(4.18),
Wi + Wu = q2 (I) + qi (I) + q1 (II) + qi (II)
= q1(I) + q1(Il) (4.19)

Again,
or
W1 + W11 > 0 (4.20)
Therefore, combining Eqs. (4.19) and (4.20), we obtain
W, + Wu = qi (I) + qi (II) > 0 (4.21)

In view of Eq. (4.18), there is no net heat flow into or out of reservoir at T2 . Since
W11 is negative, (W1 + W11 ) is the net work done by engines I and IL Similarly, q 1(I) and
q 1(II) have opposite signs, q 1(I) is negative and q 1(II) is positive. Therefore, [q 1(I) +
q 1(II)] is the net heat transferred from the reservoir at T1.
These mean that all the heat taken out of reservoir at T 1 is being converted into work,
without leaving any change elsewhere (i.e. reservoir at T 2). This is not possible according
to the Second Law of Thermodynamics. Hence, 1]1 cannot be different from 1JJI> and so,
1/1 = T/u (4.22)

Since engine I works with ideal gas, 1JI is given by Eq. (4.14). Therefore,
(4.23)
EXAMPLE 4.1 Two bodies ( 1 and 2) of equal heat capacity and initial temperatures T1
and T 2 form an adiabatically closed system. What would be the final temperature if one lets
the system come to thermal equilibrium (i) freely, and (ii) reversibly? What is the maximum
work that can be obtained from the system?
Solution (i) Since the bodies are in adiabatic enclosure, no heat is exchanged with the
surrounding. Therefore,
(E.4.1)
where m is mass of a body and mCp its heat capacity, and Tl is final temperature after
attainment of thermal equilibrium . .
From Eq. (E.4.1),
_1j+T2
TI -
2
(ii) Reversible attainment of thermal equilibrium is not possible without intervention of

a reversible engine which absorbs heat from the body at higher temperature (say, at initially
T 1) and rejects a part of it to the body at lower temperature (say, initially at T2) reversibly.
Let Ta, Tb be instantaneous temperatures of body 1 and 2 at any instant. Let Ta change
to Ta + dTa, and Tb change to Tb + dTb due to transfer of an infinitesimal quantity of heat
from body 1 to body 2. Now, at any instant,
Efficiency of the reversible engine ( T/) = Ta - Tb (E.4.2)

Ta
Again,
Work done = -mCp(dT 0 + dTb) = dT
0 + dTb
(E.4.3)
Heat absorbed - mCp dT0 dT0
(Note: dTa is negative and dTb is positive).

Combining Eqs . (E.4.2) aq·d (E.4.3), we get
(E.4.4)
Integrating Eq. (E.4.4) between initial and final states, we get
(E.4.5)
or
T1 = (T1T2)112
Work done by the reversible engine (Wrev) is the maximum work that can be obtained
from the system.
Since the system is adiabatic,
Total heat converted into work = Wrev = mCp(T1 - T1) - mCp(T1 - T2)
= mCp[(T1 + T2) - 2(T1Ti) 112 ]
= mCp(VT1 - VT2) 2
4.2.4 Thermodynamic Temperature Scale
In Carnot cycle for ideal gas, the temperature Twas the absolute temperature as arrived at
from Ideal Gas Laws. Since the efficiency of all reversible engines operating in cycle
between the same temperatures are the same, Kelvin proposed that the absolute temperature
scale be based on this thermodynamic conclusion. He called it the thennodynamic temperature
scale. However, it was made the same as that based on ideal gas by fixing the freezing
temperature of pure water (0°C) as 273.16 and that of boiling water as 373.16. In his
honour, the name of the unit is known as Kelvin (K).
4.3 Entropy
4.3.1 Entropy-A state property
For a Carnot cycle (CS), Eqs. (4.3) and (4.14) yield
q2 + q1 = T2 - 1j
(4.24)
q2 T2
i.e.
(4.25)
for two isothermal stages. For adiabatic stages, q = 0. Hence for the entire cycle,
L 51..=0 (4.26)
cycle T
Consider any arbitrary cycle. If it is reversible, then it may be regarded as consisting
of a large number of tiny CS, as shown in Fig. 4.4. A ~ B ~A is the arbitrary reversible
cycle. The dashed lines divide it into many small Carnot cycles. The arrows indicate path
directions for the small CS. In an interior cycle, its own path directions and those of its
surrounding neighbours are opposite. Hence they cancel one another. Consider another CS
at periphery. Here, the peripherial arrows are not opposed. Therefore, the net uncancelled
lines of all the Carnot cycles are shown by solid lines.
Illustration of cancellation
effect for interior tiny
Carnot cycles
- Solid peripheral curves

remain uncancelled
Volume--•
Fig. 4.4 P-V diagram for an arbitrary reversible cycle, divided into several tiny Carnot cycles.
For the net cycle described by the solid lines,
(4.27)
since it is the sum of several Carnot cycles. If CS are made infinitesimally small (i.e.
differential CS), then the solid lines would merge with the cycle ABA as the limiting case,
and Eq. (4.27) may be written as
f
cycle
i =O (4.28)
for the cycle ABA. Since it is a reversible cycle, heat is absorbed reversibly. Hence
Eq. (4.28) may be rewritten as
(4.29)
or
f
cycle
dS =0 (4.30)
where
dS = 8qrev (4.31)
T
for an infinitesimal reversible isothermal process at temperature T. In view of Eq. (4.30),

the net change of entropy of a system for a cyclic process is zero, and therefore, S is a state
property. Clausius called it Entropy (a Greek word, meaning "change"). For a finite process
from state 1 ~ state 2,
Af =S 2 - S 1 = l:; Dqrev (4.32)

1-+2 T
Since entropy is a state property, (S2 -S1) is independent ofpath (reversible or irreversible).
However, for calculation of (S2 - S 1), a reversible path is to be chosen.
EXAMPLE 4.2 Calculate Af for each step of the three step cyclic process of expansion/
compression of the ideal gas of Example 2.2.
Solution Step I: Reversible isocltoric
= 10.617 x 103 J K- 1
Step II: Reversible adiabatic: Afu = 0, since q =0

Step III: Reversible isothermal
6.Sm =~ = - 318~~ 103 = -10.617 x 103 J/K-1
Hence,
(M)cycle = .1S1 + Af11 + Afm = 103 (10.617 - 10.617) = 0
This is due to the fact that entropy is a state property.
4.3.2 Calculation of Entropy Changes

Entropy is a state property. Hence the basis and procedure for calculations of entropy
changes are similar to those for enthalpy changes, as discussed in Section 3.3. Hess' Law
and Kirchhoff s Law are applicable here too.
A pure substance at its stablest state also constitutes standard state for entropy at that
temperature which is designated as S~ (i.e. with superscript 0). The entropy values for pure
substances and many non-standard substances (i.e. solutions) are available in thermodynamic
data sources . Again, these were basically obtained from experimental measurements. However,
the following significant differences are to be noted between entropy and enthalpy:
1. Entropy changes are to be calculated only through a reversible path. This restriction
is not there for any other state property, including enthalpy.
2. T he absolute val ue of entropy of a substance (S) can be determined. Thermodynamic
data sources provide the values of S~ for pure substances. This is a consequence of
the Third L~w of Thermodynamics, which will be presented in Chapter 12. This is
in contrast with energy where only changes are available in data sources.
The e ntropy changes are to be calculated in the following ways, which are analogous
to those for enthalpy.
Entropy changes for temperature and phase changes

(a) Heating and cooling can be done reversibly by carrying them out vl!ry slowly. Therefore,
for 1 mole:
At constant volume,
dS = Sqrev = Cv dT = C d In T (4.33)
T T v
At constant pressure,
dS = Sqrev = Cp dT =Cpd In T (4.34)

T T
(b) Phase transformations can be made to occur slowly and reversibly. For a pure
substance, reversible phase changes (i.e. melting, boiling etc.) at a constant pressure occurs
at a constant temperature . Equation (4.31) is for an infinitesimal prcx:ess at const T. It can
be easily converted into a finite difference form at constant T. Then dS becomes /.l.S and
&Jrev becomes qrev (i.e. 6.H). Therefore, for melting of a pure substance,
• A uO
L1So = _LJJ1_m (4.35)
,,, T
m
For boiling of a pure substance,
(4.36)
In general, for phase transformation of a pure substance,
(4.37)
(c) Combining the above, for substance A and for the process
A (solid) A (liquid) A (gas) A (gas)
at 298 K at Tm at Tb at T
we have
s~ - S~9s = J Tm
298
c~T(s) dT + /.l.S~ + J Tb
T.,
c~T(I) dT + 6.S~ + J
T
Tb
c~T(g) dT (4.38)
(d) For a solution, L1Sm etc. depend on the composition of the solution as well. Moreover,
equilibrium (i.e. reversible) transformation temperature keeps changing as transformation
progresses (see Section 10.4). Therefore, simple equations like Eqs. (4.35)-(4.38) are to be
modified.
EXAMPLE 4.3 Calculate entropy of one mole of liquid Pb at 1000 K from that at 298 K.
Solution
S~(lOOO) =S~(298) + f Tm(Pb) Cp[Pb(s)] dT + Af/~(Pb) + f 1000 Cp[Pb (1)) dT (E.4.6)

298 T Tm(Pb) T., (Pb) T
on the basis of Eq. (4.38), where Tnr(Pb) = melting point of Pb = 600 K.

Substituting other values from data sources into Eq. (E.4.6), we obtain
S~(IOOO) = 64 .9 +f600 [23.6 + 9.75 x 10- T dT + 4810 +

298 T
3
600
f 1000
600
3
[32.4 - 3.1 x 10- T] dT
T
600
= 64.9 + 23.6 In + 9.75 x 10- 3(600 - 298)
298
1000
+ 8.017 + 32.4 In - - 3.1 x 10-3(1 000 - 600)
600
= 107 .69 J K- 1 mo1- 1•
Entropy changes for reaction (L15) and mixing (L1~ 1 x)
By nature, reactions and mixing are irreversible processes. Like enthalpy changes, L1S and
L1Smix are to be considered only for isothermal processes. However, they cannot be calculated
from L1ll and L111mix using equations of the type (4.35)-(4.37) in view of irreversibility. For
example, consider reaction (3.14) occurring isothermally at temperature T. For this,
L1S~ = S~ 8 (at D + oSg (at T) - ~ (at T) - 5gc (at T) (4.39)
But taking L1S~ =fll! 0 IT will not be correct.

The change of AS'~ with temperature can be calculated as
(4.40)
which is analogous to Eq. (3.25).
EXAMPLE 4.4 Calculate the entropy per mole of the following reaction at 1000 K from
that at 298 K.
Pb(J) + z1 0 2(g) = PbO(s)
Solution From Eq. (4.40), AS'° at 1000 K for the above reaction is related to AS'~98
as follows:
I
= 67.36 - 64.85 - 2x 205.0 (from data sources)
= - 100.0 J K - 1
Noting that Pb melts at 600 K, and fllf~ of Pb = 4810 J/mol,

(S?ooo - S~9s)Pb0 - (S?ooo - S~9s)Pb - ~ (S?ooo - S~9s)0i
-
- f
298
600 6.C;
--dT
T
4810
-- -+
600
f 1000
600
6.C"
__
T
Pr!T (E.4.8)
where
LC;= cP, PbO(s) - CP,Pb(s) - cP.02(g) i

= - 0.67 + 14.94 x 10-3r + 0.84 x io5r-2
= -9.54 + 27.78 x 10- 3r + o.84 x 10 51 2
with Cp being molar heat capacity.

Integration of Eq. (E.4.8) and numerical calculations yield
L1S?000 = -100.0 + 2.9 = - 97.1 J K- 1

Note: Analogous calculation can be performed on the basis of Kirchhoff s Law to find
out heat of reaction (LlH°) at 1000 K for the above reaction. Since the procedure has been
illustrated well in Chapter 3, we shall not deal with the details here. In fact, L1Ef> is the
molar heat of formation of PbO ( LJHJ ), and L'.1.S° also is nothing but molar entropy of
formation of PbO( LJSJ ).
L1H?000 =L1H~98 +
f600
LJC; dT - &!~(Pb) +
f 1000
LJC;: dT (E.4.9)
298 600
Now, L1H~98 = L1HJ(Pb0) at 298 K = - 219.2 x 103 J/mol PbO (from data sources).
Other data have already been listed above. Carrying out calculations, it is found that
L1H?000 = -217 x 10 3 J
As stated in this section, chemical reactions are irreversible. Therefore, L1S~ ~ &!~IT .
This is confirmed, since here,
&1?000 217 x 103 = -217 J/K

1000
=- 1000 '
and not equal to L1S?000 .

4.4 Second Law and Entropy-Significance and

consequences
4.4.1 General Comments

The statements of the Second Law of Thermodynamics are based on experience, and there
is no direct proof for them. So many physicists in the 19th Century were reluctant to accept
it as a part of physical science. The fundamental significance of entropy was also not
known. General acceptance of the law by scientists required a few decades after Clausius'
proposal of entropy in 1850. This came about when many experiments confirmed the
correctness of predictions on the basis of the Second Law.
It took several more decades to fully establish the fundamental meaning of entropy
qualitatively and quantitatively. This ended with the statistical interpretation of entropy and
experimental verification of the same. The stepping stones were:
(a) Relation of thermal energy with random motions (translation, rotation and vibration)
of atoms and molecules
(b) Statistical nature of macroscopic thermodynamic relations
(c) Concept of order and disorder
(d) Third Law of Thermodynamics
Developments in Physics in the 20th Century have shown that the basic laws and
concepts, which constituted the foundation of physics upto 19th Century, are approximations
only. These laws and concepts include:
• Law of Conservation of Matter
• Law of Conservation of Energy
• Separate entities for particle and wave
It has been established that matter can be converted into energy and vice versa. The
elementary particles can be considered both as particle and wave. The only law proposed
upto 19th Century, which is still held to be valid, is the Second Law of Thermodynamics.
4.4.2 Various Interpretations of Entropy
The various interpretations of entropy evolved gradually over a period of several decades.
These are enumerated as follows:
1. dS = /jqrevlT for an infinitesimal , isothermal, reversible process.
2. Entropy is Time's Arrow, i.e. a fundamental indicator of time.
3. Entropy has relation with heat not available for work.
4. Entropy is a measure of disorder of a system or a substance, which is under consideration.
The first interpretation has already been dealt with . The second and third interpretations
will be arri ved at following further discussions in this chapter. The fourth interpretation will be
elaborated in relation to the Third Law and statistical interpretati on of entropy in Chapter 12.
4.4.3 Entropy Changes for Reversible and Irreversible Processes
Conclusions regarding entropy changes for reversible and irreversible processes have been
arrived at in various ways. Two procedures for derivation are presented here.
Derivation procedure I
Consider absorption of an infinitesimal quantity of heat (8q) by a closed system at a
temperature T syst from its surroundings. Assume that both system and surrounding have
large heat capacities. Their temperatures remain essentially constant even after transfer of &,.
Hence, the system absorbs heat isothermally and reversibly. The surrounding also gives out
heat isothermally and reversibly. Thus,
oq (4.41)
(dS)system = +
~yst
8q
(dS)surrounding =- T (4.42)
SUIT
So,
1 1
(dS)syst + surr = 8q (-- - - -) (4.43)
Tsy st Tsurr
The reversible exchange of&, requires that the rate of heat transfer be very slow. This
is possible if
(4.44)
Hence, for reversible heat exchange, from Eqs. (4.43) and (4.44),
(dS)syst + surr =0 (4.45)

i.e.
(..1S)syst + surr = 0 for finite process (4.46)
If Tsurr is significantly higher than Tsyst• then the heat transfer is irreversible. From
Eq. (4.43) since - 1 --

( ~yst
1
~urr
J> 0 '
(dS)syst + surr > 0 (4.47)

i.e.
(..1S)syst + surr > 0 for finite process (4.48)
Derivation procedure II
Consider the heat engine in Fig. 4.1 as the system and the two heat reservoirs as its surrounding.
SBCOnd Law of Thermodynamics and Entropy
That is, the engine and reservoirs are insulated from outside except for the heat and work
interactions indicated in the figure. Since entropy is a state property, for one cycle,
(M)engine =0 (4.49)
(L1S)heat reservoin -
_ _ qz + q1 _ _ q2 + q,
T
2
T., - T
I
T. ( )2 I
(4.50)
For a reversible process, from Eq. (4.25),
(M)heat reservoin =0 (4.51)

Hence, from Eqs. (4.49) and (4.51), for a reversible process,
(LiS)syst + surr = 0 (4.52)

The cycle would be irreversible if a portion of the energy is dissipated as heat due to
friction, rather than resulting into work. In that case, the efficiency of the engine would be
lower than that of a reversible engine. In other words,
T2 - 1i
T/irr < T/rev• i.e. T/irr < (4.53)
Tz
i.e .
Tz - 1i
-W = qz + qi < -=--____.. (4.54)
q2 q2 T2
or
(4.55)
i.e.
q2 + !11.. < 0
(4.56)
T2 1i
Combining Eq. (4.56) with Eq. (4.50), we get
(M)heat resevoin > 0 (4.57)

For one cycle, (M)engine = 0 [Eq. (4.49)]. Hence, for an irreversible cycle,
(LIS)syst + surr > 0 (4.58)

Here, the positive value of total entropy change for system and surrounding is essentially
due to the irreversibility of the engine, which is making less heat available for work.
Therefore, the larger the increase in (LiS)syst + surrt the lesser the heat available for work.
Thus, these two are interrelated. This provides explanation for the third interpretation of
entropy, as described in Section 4.4.2.
EXAMPLE 4.5 Calculate the entropy change of the Universe in isothermal freezing of
1 g-mole of supercooled liquid gold at 1250 K, from the following data for gold.
Tm = 1336 K, &I~ (at 1336 K) = 12.36 x 103 J/mol

Cp(s) = 23.68 + 5.19 x io-3T J/mol/K
Cp(l) = 29.29 J/moVK
Solution
(i) LtSsyst
Actual freezing process is isothermal but irreversible, since it is not occurring at the equilibrium
freezing temperature. But ..dSsyst is to be calculated only along a reversible path. It makes
no difference as to which reversible path we choose since entropy is a state property.
The simplest reversible path is
Liquid Au beating > liquid Au freezing ) solid Au cooling > solid Au

At 1250 K at 1336 K at 1336 K at 1336 K at 1250 K
(state 1) (state 2)
!;.Ssyst = S2
- SI = f,1336 Cp(l) dT - Ml~ +
12S0 T 1336
f 12SO Cp(S) dT
1336 T
1336
29.69 12.36 X 10 3 + f (23.68 + 5.19 10- 3 T) dT
=
f
mo
- - dT- - - - -
T 1336 1336
12SO
T
X
= - 9.327 J K- 1
(ii) ..dSsurr
Both the initial and final states of the system have been assumed to be 1250 K. This
is possible only if the surrounding is at 1250 K, and it has infinite heat capacity (i.e. the
surrounding absorbs heat released during freezing reversibly at 1250 K). Hence,
..dS = _ t1H syst

surr 1250
Lllisyst =H2 - H1 =f
1336
29.69 dT - 12.36 x 103 +
f1250
(23.68 + 5.19 x 10-
3
T) dT
12.50 1336
= - 12.46 x 103 joules

So,
..dSsurr =+ 12.46 X 103 =9.967 J/K

1250
or
LiS'universc = - 9.327 + 9.967 = + 0.64 J/ K
LiS'universe is positi ve since the process is irre versible, as the derivation in Eq. (4.48) shows.
Note: In Example 4.4 also, the reaction is isothermal at 1000 K, but irreversible. Hence,
the surrounding is at 1000 K and has large heat capacity. Hence the surrounding may be
assumed to absorb heat released by the reaction reversibly at 1000 K.
L1Ssurr = - L1H~ooo =+217 J K-1

1000
i.e.
LiS'univ = LiS'syst + LiS'surr = 217 - 97.l = + 119.9 J K-I
LiS'univ has a very high positive value for this reaction in contrast to that of freezing of
supercooled gold. This is because a chemical reaction , in general, is more irreversible than
irreversible freezing/melting.
4.4.4 Entropy Changes of Isolated Systems
An isolated system has no interaction with outside environment. From a thermodynamic

point of view, therefore, the system contains the surroundings within itself. Therefore, from
Eqs. (4.45) and (4.47), for an isolated system,
(dS)syst =0 for reversible process (4.59)
(dS)syst > 0 for irreversible process (4.60)
The Universe is an isolated system. Moreover, spontaneous (i.e. irreversible) processes
are occurring all the time. Hence, the entropy of the Universe is always increasing with the
passage of time. There is no other property that behaves this way. Thus, entropy is an
exception from others and is considered to be a fundamental indicator of time (Time's
arrow). This explains the second interpretation of entropy (see section 4.4.2).
4.5 Summary
1. A heat engine converts heat into work and operates in cycle.
Efficiency of an engine = Tf = Wlq 2, where W is work done by the engine and q2
is heat absorbed by it from the source, in one cycle.
2. For a reversible Carnot cycle operating between source temperature (T2) and sink
temperature (T1), for any working substance,
T2 - 7;
TJ =~-~
T2
This is the basis for the thermodynamic temperature scale, proposed by Kelvin.
3. There exists a state property, entropy (S), defined as
dS = 8qttv
T
where 8qrt v is an infinitesimal quantity of heat absorbed by the system reversibly at
temperature T. For a finite process from state 1 to state 2,
S2 - S1 = f 2 8qrev
1 T
Hence, although S is a state property, change of entropy due to change of state has
to be calculated along reversible path only. For an isothennal process at T, AS = qrejT.
4. From the Third Law of Thermodynamics (not discussed yet), entropy of a perfectly
ordered crystalline solid is zero at 0 K. This allows determination and tabulation of
absolu~e values of entropy of substances at various temperatures.
S. Like enthalpy, entropy changes may be classified into:

(a) changes due to change of temperature; and
(b) changes due to isothermal processes, such as melting, phase transformations,
reactions, and mixing.
Also, Hess' and Kirchhoffs Laws are applicable.
=0 for reversible process
6. (Af)syst+surr . .
> 0 for lfrevers1ble process
7. For an isolated system,
=0 for reversible processes
(Af)system
> 0 for irreversible processes
since the system itself contains surrounding.
8. Efficiency of an irreversible engine (T/irr) is less than that of a reversible engine (T/rev).
PROBLEMS
4.1 Calculate AS for each stage for Problem 3.7.

4.2 The initial state of one mole of an ideal monatomic gas is P = 10 atrn and T = 300 K.
Calculate the entropy change in the gas for (a) reversible isothermal decrease in the
pressure to l atm; (b) reversible adiabatic decrease in pressure to l atm; and
(c) reversible constant volume decrease in pressure to 1 atm.
4.3 Calculate $ at 298 K and l atm for the reaction:
W (s) + 0 2 (g) = W02 (s)
Sscond Law of Thermodynamics and Entropy
4.4 Calculate the standard entropy of formation of TiC from titanium and graphite at
1200 K.
4.5 Calculate LiS for both system and surrounding per mole of the following isothermal
process at 800 K: Al (I) = Al (s). Assume aluminium to be pure. Is the process
irreversible?
4.6 Calculate LJS> at 1000 K for the reaction in Problem 3.5. What is the value of LiS
of Universe?
4.7 One gram of liquid Th02 at 2900°C is mixed with 5 g of Th0 2 at 3400°C adiabatica1ly.
(a) What is the final temperature?
(b) What is the entropy change of the system and the surrounding combined?
(c) Is the process spontaneous?
Assume Cp to be independent of temperature.
4.8 1 kg of a metallic powder initially at 25°C is mixed with 1 kg of the same liquid
metal, initially at 150°C, in a container insulated from the surroundings. Cp of both
liquid and powder are 4.0 J/g/K. Calculate the entropy change of the Universe for
this process.
Auxiliary Functions and Relations,
Criteria for Equilibrium
5.1 Free Energy

As discussed in section 4.4.3, more the increase of entropy of the system during a process,
less is the work done when a quantity of heat is supplied. The unit of entropy is joules per
degree Kelvin (J K- 1). Hence the quantity TS has unit of energy (J). Approximately speaking,
TS basically represents bound energy which cannot be utilized for doing work, but gets
dissipated as heat. For knowing the extent of work the system is capable of doing, we ought
to fipd out the free energy, i.e. the energy available for work.
The following two free energy functions are employed in thermodynamics, according
to the names of scientists who proposed them:
Helmholtz Free Energy A =U - TS (5 .1)

Gibbs Free Energy G = H - TS = U + PV - TS (5.2)
Since U, S, H are state properties and T is a state variable, A and G are also state properties
by definition.
5.2 Combined Expressions of First and Second Laws

of Thermodynamics
Assume: (i) closed system (i.e. fixed mass and fixed composition) and (ii) reversible
process.
Then, from Eqs. (2.2) and (2. 17),
dU = 8q - OW = 8q - P dV - OW' (5 .3)
64
Auxiliary Functions and Relations, Criteria for Equilibrium
From Eq . (4.3 1),

(jq = T dS (5.4)
dU = T dS - P dV - bW' (5.5)
Again,
dH = dU + P dV + V dP (2.25)
Combining Eqs. (5.5) and (2.25), we have
dH = (T dS - P dV - OW) + P dV + V dP
= TdS+ VdP- OW' (5.6)
Again, differentiating Eq. (5.1). we obtain
dA = dU - T dS - S dT (5.7)
dA = T dS - POV - OW' - T dS - S dT
= - S dT - P dV - OW' (5.8)
Differentiating Eq. (5.2), we obtain
dG = dH - T dS - S dT (5.9)
Combining Eq. (5.9) with Eq. (5.6), we have
dG = (T dS + V dP - OW) - T dS - S dT
= - S dT + V dP - OW' (5.10)
Equations (5.5), (5.6), (5.8) and (5.10) have been derived by combining the First and
Second Laws of Thermodynamics with the assumption of reversible process, and for fixed
mass and composition. If it is further assumed that the only work done is against pressure
(i.e. bW' = 0), then these equations get further simplified into the following equations:
dU = T · dS - P dV (5.11)
dH = T dS + V dP (5.12)
dA = - S dT - P dV (5.13)
dG =- S dT + V dP (5.14)
Equations (5.11)-(5.14) are four basic differential equations in thermodynamics, and
are valid only under the assumptions stated above. Closed and isolated systems have fixed
mass and composition. The equations are applicable to open systems as well, provided we
consider a fixed mass (1 mole or 1 kg, for example), and constant composition.
5.3 Maxwell's Relations

Consider a variable Z which is a function of two variables x, y, i.e.
Z = j{x, y) (5.15)
Then,
dZ = (~~)Y dx + (~~}. dy = M dx + N dy (say) (5 .16)
From the properties of differential equations,
(5.17)
Applying formula (5.17), we obtain from Eqs. (5. 11)-(5. 14) the following four differential
equations:
,(5.18)
(;:)s = (~~t (5.19)
(;i )T = (~;)V (5.20)
(5.21)
Equations (5.18)-(5.21) are known as Maxwell's relations. These are the four basic
equations. By combining these with other equations, it is possible to arrive at many other
thermodynamic relations in the form of differential equations. They have been found to be
very valuable in science and engineering.
Another property of function (5.15) is
(oz)
8x Y
(ox) (oy) __ 1
oy z oz x
(5.22)
Some examples of derivation of useful equations employing some of the foregoing

relations are being worked out now.
5.3.1 Cp, Cv Relations for Thermodynamic Substances

We have earlier derived
(2.39)
From Eq. (5.11) ,
(8U)
8V T
=r(8S)
8V
_ T
p (5.23)
Combining Eq. (5.23) with the Maxwell equation (5.20), we get
(8U)
8V
= r(8P) _p
T 8T v
(5.24)
From Eqs. (2.39) and (5.24),
C - Cv =r(8V)
P 8T
(8T8p) p v
(5.25)
Since V = f(P, 1), from Eq. (5.22), we have
(8V) (8P) (8T) _- l

8P T 8T v 8V p
(5.26)
From Eqs. (5.25) and (5.26),
(5.27)
Now, isobaric coefficient of volumetric thermal expansion of a. material (a) is given as
a=..!_
v 8T
(8V) p
(5.28)
Similarly, isothermal compressibility of a material ({J) is given as
f3 = _ ..!_ (8V) (5.29)

V 8P r
Combining Eqs. (5.27) to (5.29), we get
vra 2
Cp -Cv =p (5.30)
Equation (5.30) is valid only for a thermodynamic substance since Eq. (5.11), which has been
used for derivation, is applicable to reversible processes (i.e. for a thermodynamic substance).
5.3.2 Thermoelastic Effect
Consider an insulated (i.e. adiabatic) system. If a hydrostatic pressure P is applied, there

will be a change in temperature of the substance. To make the process reversible, the
pressure should be increased slowly. Also, deformation should be elastic. Under the above
circumstances, we wish to find the relationship of (:;:)q with physical variables. The
constant q restriction denotes that the system is adiabatic. When JW' = 0,

dll = liq - P dV
which is the same as Eq. (2.18). Again,
dU = (~l dV + (~l dT
which is the same as Eq. (2.7). Hence, at liq = 0,
(5.31)
Noting that (~)v = Cv• and differentiating Eq. (5.31) with respect to P, Eq. (5.31) can
be rewritten as
(5.32)
(Cp - Cv)l
(!;:). ~ (~),
= [ (:),
(5.33)
From Eqs. (5.33) and (5.30),
(5.34)
Noting that a = .!..

v
(bV)
Sfp
,
(5.35)
Again,
dV = (bV)
Sf
dT + (bV)
,5P
p
dP T
(5.36)
On the basis of definitions of a and p [see Eqs. (5.28) and (5.29)], we have
dV = Va dT - Vp dP (5.37)
or
bV)
(5P q
= va(sr)
fipq
- V/J (5.38)
Substituting Eq. (5.38) into Eq. (5.35), we obtain
(Sf)
fipq =[
vra vra
vra2] = Cp
(5.39)
Cv+p
u. Since
If we consider the effect of uniaxial tensile stress (o). then P may be substituted by
the coefficient of linear expansion is 1!3a. and the sign of tensile stress is opposite to
pressure, which is compressive in nature, Eq. (5.39) gets modified into
vra (5 .40)
3Cp
for an isotropic substance.
Equation (5.40) gives the thermoelastic coefficient (:J in terms of physical variables.
Experimental measurements have shown approximate agreement with this equation. The
relationship between Cp and Cv, and the derivation of thermoelastic coefficient are two
examples of the usefulness of Maxwell' s relations. Many other relations have been obtained,
and are listed in some standard texts.*
5.4 Criteria for Thermodynamic Equilibria

At equilibrium, Mere is no change in the system, i.e. it remains stable. Again, a reversible
process occurs through a series of equilibrium stages. Therefore, if we keep the independent
variables on the right-hand side of Eqs. (5.11 )-(5.14) fixed, then we arrive at the no
change, i.e. equilibrium criteria. Again, Eqs. (5.11)-(5.14) are for differential (i.e. infinitesimal)
processes. The equilibrium criteria can be expressed in differential as well as in finite
difference forms. Table 5.1 lists the criteria which constitute the equation given there
viz. Eqs. [(5.41)-(5.44)].
Table 5.1 Equilibrium Criteria
Basis Differential form Finite difference fonn Eq. Nos.
Eq. (5.11) (dU)s.v =0 (LlU)s,v =0 (5.41)

Eq. (5.12) (dH)s,P =0 (LlH)s,P = 0 (5.42)
Eq. (5.13) (dA)r. v = 0 (M)r. v = 0 (5.43)
Eq. (5.14) (dG)T.P = 0 (LlG)r.P = 0 (5.44)
Since it is easy to maintain temperature and pressure constant, the Gibbs free energy
criterion { Eq. (5.44 )] is ernployed in chemical and metallurgical thermodynamics. However,
in other areas, other criteria are also employed.
5.5 Criteria for Maximum Work

If c5W is not zero, then other kinds of work are involved (e.g. electrical, magnetic, and
surface etc.) besides P dV type work. Then, from Eqs. (5.5), (5.6), (5.8) and (5.10),
respectively, we arrive at the following equations:
(dU)s,v = - c5W, i.e. (LlU)s,v =- W (5.45)
(dH)S,P = - 8W', i.e. (LlH)s,P = - W (5.46)
(dA)r.v = - c5W, i.e. (LlA)r.v = - W (5.47)
(dG)r.P = - c5W, i.e. (LlG)T,P = - \.V (5.48)
*See, for example, R.A. Swalin, Thermodynamics of Solids, Wiley, New York, 1964, p. 32.
In Section 2.3, we have illustrated the concept of irreversible processes. Consider

expansion of gas in the cylinder-piston system. For irreversible expansion, external pressure
is significantly lower than the pressure in the system (say, P - Li.P, where P is pressure
inside). Then, bW = (P - Li.P) dV. However, for a reversible process, it is P dV. Therefore,
the work done by a reversible process is maximum.
Since we are considering reversible processes here, the work done would be maximum.
Hence, W' in Eqs. (5.45)-(5.48) is also designated as W'max· Hence (LiU)s, v etc. denote the
maximum capacity of the system to do work, and are loosely termed as work content. After
the work is performed, the energy decreases. This is the significance of the negative sign.
Application of these equations, especially Eq. (5.48), will be demonstrated in Chapters 13
and 14.
5.6 Criteria for Irreversible Processes

From Eqs. (4.3) and (4.24), for Carnot cycle,
(5.49)
Suppose q2 is absorbed irreversibly. Then, from Eqs. (4.53) and (5.49),
(5.50)
or
(5.51)
Rearranging Eq. (5.51), we get
q (rev) q (irr)
-1 - + -2- < 0 (5.52)
Ti Tz
In general, for a cyclic process A -7 B -7 A, let A -7 B be irreversible and B -7 A
be reversible. Then, with similar logic as given above,
(5.53)
Combining this equation with Eq. (4.32), we obtain
(5.54)
i.e.
S - S > :E 8q(irr)
D A A --+B T (5.55)
or
dS > 8q(irr) (5.56)
T
i.e. for an irreversible process,
8q < T dS (5.57)
Combining Eq. (5 .57) with Eq. (2.2) and Eq. (2.17) and for OW = 0, for irreversible process,
dU < T dS - P dV (5.58)
i.e.
(dU)sv < 0, i.e. (~U)s, v < 0 for finite process (5.59)
Similarly,
(dH)s,P < 0 or (~s.P < 0 (5.60)
(dA)r,v < 0 or (M)r,v < 0 (5.61)
(dG)T,P < 0 or (~G)r,p < 0 (5.62)

It slwuld be noted that Eqs. (5.59H5.62) are qualitative in nature. Quantitative calculations
are possible only for equilibrium {see Eqs. ( 5.4 I H 5.44 )} or for reversible processes, which
proceed through a series of equilibrium stages.
5.7 Minimum Free Energy at Equilibrium

As stated in Section 5.4, it is relatively easy to maintain temperature and pressure constant.
Hence the Gibbs free energy criteria are generally employed in chemical, metallurgical and
materials thermodynamics. For this, J.W. Gibbs is considered as the father of chemical
thermodynamics.
Let us recapitulate the criterion for equilibrium,
i.e.
(dG)T,P =0 or (~G)T,P =0 (5.44)
and, for an irreversible process, viz.

(dG)T,P < 0 or (~G)T,P < 0 (5.62)
For a process to occur spontaneously, Eq. (5.62) is to be satisfied. Therefore, Eq. (5.62)
provides the criterion for feasibility of a process.
If, on the other hand, for a process,
(dG)r,p > 0 or (~Gh.P > 0 (5.63)
·~ ..
then the process under consideration is not feasible. However, it may be noted that the
process would tend to proceed spontaneously in the backward (i.e. reverse) direction, since
for reverse process, the sign of (dG)r,P will be reverse, i.e. (dG)r,p < 0.
The above considerations have led to the minimum free energy criterion for equilibrium.
It is illustrated schematically in Fig. 5.1. It is evident from Fig. 5.1 that the system would
tend towards state 2 either from the left or from the right. At state 2, (dG)r,P = 0 since
it is at minimum.
Forward Backward
process process
l I
I
I
I
I
I
l
I
------ I
I
(dG)r,P < 0
for 1
(dG)r,P < 0
for
lI
I I I
1 State 1 ~ State 2 1 State 2 ~ State 3 1
State 1 State 2 State 3

(Equilibrium)
Process path ---+
Fig. 5.1 Minimum free energy criterion for equilibrium.
Figure 5.1 constitutes the basis for the free energy minimization technique of searching
equilibrium state of a system. This is the general method for complex equilibrium calculations.
Figure 5.1 is two dimensional. For a three-dimensional arrangement, (G)T,P -is a surface
(like a bowl), and the minimum is at the bottom of the bowl.
5.8 Summary
1. Helmholtz free energy A = U - TS
Gibbs free energy, G =H - TS = U + PV - TS
By definition, free energy is a state propert) .
2. For a system of fixed mass and composition, and OW' = 0, the following criteria were
arrived at by combining the First and Second Laws of Thermodynamics.
Reversible processes Equilibrium Spontaneous processes

dU = T dS - P dV (dlf)s.v=0 (dlf)s,v < 0
dH = T dS + V dP (dH)s,P = 0 (dH)s,P < Q
dA = - S dT - P dY (dA)r.v = 0 (dA)r.v < 0
dG = - S dT + V dP (dG)r.P = 0 (dG)r,P < 0
Of these, the criteria based on Gibbs free energy are employed in chemical and
metallurgical thermodynamics. For finite processes, dG etc. are to be replaced by L1G
etc. At equilibrium, (G)r,P is at its minimum.
3. The four differential equations for reversible processes constitute the basis for the
four basic Maxwell's relations.
2
4. Cp - Cv = vra , where V is molar volume, a is the coefficient of volumetric
f3
thermal expansion, and f3 is the isothermal compressibility of a substance.
5. Thermoelastic coefficient of a solid = ( uu

~~)q = - VTa
C , where
3 p
<J is uniaxial tensile
stress.
Cbapta 6
Gibbs Free Energy and
One-component Systems
6.1 Introductory Comments
6.1.1 Variables Influencing ·Gibbs Free Energy
As stated in Section 1.3, the state of a system decides its energy content. Again, the state
depends on many variables. Free energy is no exception. This issue is again being elaborated
for Gibbs free energy (G) with the intention of eventual understanding of its various
applications in chemical thermodynamics.
Gibbs free energy G of a substance is a function of temperature and pressu,re. But the
L1G criterion is useful for application only at constant temperature (T) and pressure (P) for
fixed mass of a system of constant composition (refer Chapter 5). Thus, let us first keep
T and P constant, and consider the dependence of G of a substance on other variables.
Free energy is an extensive property and is proportional to mass (or mole) of the
substance. Therefore, we should have a convention about the quantity of matter we are
talking about. By Universal convention in thermodynamics, tht!'tsymbol G generally refers io
Gibbs free energy per g-mole of a substance. However, G also refers to free energy of the
entire system. These will be specified in subsequent discussions. However, it is understood
that if someone wishes to adopt some basis other than g-mole, it would not make any
difference to conclusions in most cases.
The free energy of a substance of fixed mass also depends on
(a) chemical composition of the substance;
(b) state of ·aggregation of the substance (solid, liquid, gas or plasma);
(c) struc.~ure of the material, if it is a solid, such as crystal structure, microstructure,
substructure, macrostructure;
75
95
= 0 +RT In t. +0
2.149
= - 212.5 J mo1- 1 at T = 263 K
Note: L1G 1 and L1G3 are zero because these steps are at equilibrium.
6.2.2 Standard State- Fugacity relationship
The standard state of a substance has already been defined in Section 3.3 in connection with
enthalpy . The same statement is applicable to free energy, i.e. the standard state is pure
element or compound at I atm pressure and at its stablest state at the temperature under
consideration (i.e. at ambient temperature). It may be stated further that it should be ideal,
if it is gaseous.
It has also been mentioned in Section 3.3 that the thermodynamic quantities at standard
state of a substance are distinguished by superscript zero (JfJ, SJ, G°, ...).
Now, in Eq. (6.6), if P 1 = 1 atm and P2 = P (i.e. a variable pressure), and G 2 = G,
then
L1G = G - G° = RT In P (6.8)
Simil arly, for nonideal gas, Eq. (6.7) may be integrated to yield
G - Go = RT ln(flt°) (6.9)
where J° is fugacity at standard state.
It is to be noted further that any gas tends to behave ideally if P ~ 0. i.e.
if P ~ 0, then f ~ P , flP ~ 1 (6. 10)
6.2.3 Fugacity-Pressure Relationship in Gases
Gases tend to deviate from ideal behaviour more if pressure is increased and/or temperature
is decreased. The permanent gases exhibit significant departure from ideality only at extre mely
low (sub-zero) temperatures and/or very high pressures. In metals and materials processing,
these are rare. Most processing are at high temperature and moderate pressure. Hence
departures from ideal behaviour are small.
For small deviations from ideality,
V= RT - a
p ( 6.11)
where a is a temperature dependent parameter. Hence, from Eqs. (6.7) and (6. 11),
dG = V dP = RTd(ln f) = RTd In P - a dP (6.12)

Gibbs Free Energy and
One-component Systems
6. 1 Introductory Comments
6.1.1 Variables Influencing ·Gibbs Free Energy
As stated in Section 1.3, the state of a system decides its energy content. Again, the state
depends on many variables. Free energy is no exception. This issue is again b_e ing elaborated
for Gibbs free energy (G) with the intention of eventual understanding of its various
applications in chemical thermodynamics.
Gibbs free energy G of a substance is a function of temperature and pressu,re. But tht
LiG criterion is useful for application only at constant temperature (7) and pressure (P) for
fixed mass of a system of constant composition (refer Chapter 5). Thus, let us first keep
T and P constant, and consider the dependence of G of a substance on other variables.
Free energy is an extensive property and is proportional to mass (or mole) of the
substance. Therefore, we should have a convention about the quantity of matter we are
talking about. By Universal convention in thennodynamics, thtasymbol G generally refers io
Gibbs free energy per g-mole of a substance. However, G also refers to free energy of the
entire system. These will be specified in subsequent discussions. However, it is understood
that if someone wishes to adopt some basis other than g-mole, it would not make any
difference to conclusions in most cases.
The free energy of a substance of fixed mass also depends on
(a) chemical composition of the substance;
(b) state of ·aggregation of the substance (solid, liquid, gas or plasma);
(c) struc.~ure of the material, if it is a solid, such as crystal structure, microstructure,
substructure, macrostructure;
75
(d) strengths of other energy fields (gravitational, electric, magnetic); and

(e) specific surface area (i.e. surface area per unit volume) .
6.1.2 A Note on Subsequent Topics
It is a lso to be kept in mind that we are concerned with change of free energy (i.e. L1G),
and not its absolute value. In chemical thermodynamics, strengths of other energy fields
[i.e. item (d) above] are assu med to remain constant during a process, and hence, its
contribution to L1G is negligible.
With the above discussion in mind, we shall consider the application of Gibbs free
energy in thermodynamics of materials, which is based on chemical thermodynamics as
stated in Chapter 1. It is to be further noted that quantitative application is possible only
for thermodynamic equilibria in view of the criterion (L1G)r. p = 0. Of course, this criterion
is also applicable to reversible processes. However, it is beyond the scope of any introductory
text on thermodynamics.
For irreversible processes, the criterion (L1G)r,P < 0 is qualitative in nature. It can be
employed for assessment of feasibility of a process only, and the same will be occasionally
illustrated. Hence, for the reasons cited above, any text on chemical/metallurgical/material
thermodynamics is basically concerned with the study of thermodynamic equilibria by
which, we mean (a) phase equilibria in heterogeneous system, or (b) chemical equilibria in
either homogeneous or heterogeneous systems.
In this chapter, we shall take up the si mplest system of constant composition , as generally
applicable to the one-component system (a pure element or pure compound). In subsequent
chapters, more complex situations involving chemical and phase equilibria in multicomponent
systems, where compositions of phases are variable, will be taken up one after another.
Systems involving changes of electrical and surface e nergy, i.e. thermodynamics of electro-
chemical cells and thermodynamics of surfaces, will be dealt with in Chapters 13 and 14.
As illustrated in Chapter 5, in solid state processes where volume remains approximately
constant, internal energy (U) and Helmholtz free energy (A) criteria are more appropriate.
In a way, these topics also belong to thermodynamics of materials but traditionally they are
not covered in a subject like the present one.
6.2 Isothermal Processes
6.2.1 Some General Relations; Fugacity

For application of L1G criterion, the process has to be isothermal. It can be reversible or
irreversible. Only the initial and final temperatures should be the same (see Section 3.3).
Consider the process
State l (at temp 7) ~ State 2 (at temp 7) (6.1 )
(G1 . Hi. S 1) (G2. H 2. S2)
Gibbs Free Energy and One-Component Systems
G i • Hi · Si and Gi, H-1, Si are free energy, enthalpy and entropy values at states 1 and 2,
respectively. Then,
(6.2)
Again, from the definition of G (Eq. 5.2),

L1G = Gi - Gi = (Hi - TS2) - (Hi - TSi)
= (H2 - Hi) - T(Si - Si)
=L1H-TL1S (6.3)
In Eq. (6.1), the process from state I to state 2 may be any isothermal process viz.
phase transformations, mixing, chemical reactions (see Section 3.3). Suppose the process is
purely a change of pressure for a pure substance A, i.e.
(6.4)
then what should be L1G? If A is solid or liquid, L1G .. ,.,, . 0, unless the change of pressure
is very large since pressure does not have much effect on their energy (see Section 3.3).
However, if A is a gas, then change of pressure would have a significant effect on G. From
Eq. (5.14), at const T, dG = V dP. If A is an ideal gas, then
dG = V dP = (RTIP) dP = RT d(ln P) (6.5)
Integrating Eq. (6.5) from Pi to Pi. for the process (6.4),
L1G = RT ln (Pi/Pi) (6.6)
The unit of pressure in chemical thennodynamics is one standard atmosphere (760 mm'Hg},
simply designated as atm.
If the gas is not ideal, then, in analogy with Eq. (6.5), a .function known as .fugacity
(/) has been defined as follows:
dG = RT d(Jn f) (6.7)
It may be noted here that the above definition of .fugacity is a general one applicable to
all thennodynamic substances, and not restricted to nonideal gases only.
EXAMPLE 6.1 The equilibrium vapour pressure of ice at - 10°C is 1.95 mmHg, and that
of supercooled water at the same temp~rature is 2.149 mmHg. Calculate the free energy
change accompanying the change of 1 mole of supercool'ed water to ice at -10°C.
Solution Free energy is a state property. Hence we can choose any path for calculation
with the same result. Let us choose the following path:
mole supercooled <1Gi Water vapour <1Gi Water vapour

water (-10°C) (2.149 mm Hg, - 10°C) (1.95 mm Hg, - 10°C)
! L1G3
ice (-10°C)
L1G = L1G I + L1G2 + L1G3

95
=0 + RT In I. +0
2.149
= - 212.5 J mo1- 1 at T = 263 K
Note: LIG 1 and LIG 3 are zero because these steps are at equilibrium.
6.2.2 Standard State- Fugacity relationship
The standard state of a substance has already been defined in Section 3.3 in connection with
enthalpy. The same statement is applicable to free energy, i.e. the standard state is pure
element or compound at I atm pressure and at its stablest state at the temperature under
consideration (i.e. at ambient temperature). It may be stated farther that it should be ideal,
if it is gaseous.
It has also been mentioned in Section 3.3 that the thermodynamic quantities at standard
state of a substance are distinguished by superscript zero (I.fl, SJ, G°, ... ).
Now, in Eq. (6.6), if P 1 = 1 atm and P 2 = P (i.e. a variable pressure), and G2 = G,
then
L1G = G - G° = RT In P (6.8)
Similarly, for nonideal gas, Eq. (6.7) may be integrated to yield
G - G° = RT lnlfl.f) (6.9)
where .f is fugacity at standard state.
It is to be noted further that any gas tends to behave ideally if P ~ 0. i.e.
if P ~ 0, then f ~ P, PP ~ 1 (6.10)
6.2.3 Fugacity-Pressure Relationship in Gases
Gases tend to deviate from ideal behaviour more if pressure is increased and/or temperature
is decreased. The permanent gases exhibit significant departure from ideality only at extremely
low (sub-zero) temperatures and/or very high pressures. In metals and materials processing,
these are rare. Most processing are at high temperature and moderate pressure. Hence
departures from ideal behaviour are small.
For small deviations from ideality,
V =RT - a (6.11 )
p
where a is a temperature dependent parameter. Hence, from Eqs. (6. 7) and (6.11),
dG =V dP = RTd(ln f) = RTd In P - a dP (6.12)

i.e.
!!._ dP (6.13)
RT
Integrating Eq. (6.13), we get
/) aP
In ( p = - RT + C (6.14)
where C is the integration constant. If integration is done from P = 0 to P = P, then at

= = =
P 0, PP 1, and hence, C 0. This yields
(6.15)
Again, for small deviation from ideaJity, aPIRT « l. So,
f. : : : 1 - d' ::::: Pl [from Eq. (6.11)] (6.16)

P RT RT
Again,
RT =P (6.17)
v I
where P; is pressure of ideal gas corresponding to the values of V and T in Eq. (6.16).
f ::: p
(6.18)
p Pi
From experimental values of a, some values of f were calculated for H 2 , 0 2 , CH4 at
P = 50 atm at a few temperatures. These are given in Table 6.1. It may be noted that the
gases are fairly close to ideal, especially at 200°C. So at lower pressure, they may be
assumed to be ideal without any significant error.
Table 6.1 Values off at P = 50 atm
Gas 0°C l00°C 200°C

H2 51.5 51.2 51.0
02 48.5 50 .0 50.5
c~ 45.2 4 8.5 50.0
6.3 Variation of .1Gwith Temperature for an Isothermal

Process
For this purpose, mass, composition and pressure are assumed to be constant. Only the
temperature change is being considered.
6.3.1 Gibbs-Helmholtz (G-H) Equations
From Eq . (5.14),
(8G)
oT p
= -S (6.19)
G =H - TS =H + r( 0G)
8T P
(6.20)
Dividing LHS and RHS of Eq. (6.20) by T 2 , we obtain
i.e.
1
T 8T
(8G) P -
G= -
T2
H
T2
(6.21)
or
[8(~T)l = - -~ (6.22)
Equation (6.22) is one form of G-H equation. The alternative form is derived through
differentiation by parts as follows:
8(G/T)1- _!_ [___§!!__] + G_ _!_ (oG) _!!.!.__ + G (6.23)

[o(~) P - T 8(1/T) P
T 8T p d(l/T)
So,
.8(GIT)]
[ 8(1/T) r
= -T2 _!_
T
(8G) + G= _
8T P
T2 [_!_
T
(8G) _!}__]
8T y2
P (6.24)
o(GIT)]
[ 0(1/T) p
= _ yi (- !!_)
T2
=H (6.25)
Equation (6.25) is the alternative fonn of G-H equation.

For the process, as in Eq. (6.1), if dT, dP are common to both states I and 2, then
dG = dG2 - dG 1 = d(L1G)
= CV2 - V1)dP - (S2 - S1) dT
= L1V dP - L1S dT (6.26)
So,
(6.27)
Since Eq. (6.27) is analogous to Eq. (6. 19), the following relations can be derived:
(6.28)
8(iJG/T)] =L1H (6.29)

[ 8(1/T) P
where L1G is change of Gibbs free energy for an isothermal process. Similarly, L1V, L1S, nH
also refer to changes in V, S, H due to occurrence of the isothermal process.
6.4 Phase Equilibria in One-component

(i.e. Unary) Systems
In Section 1.3.4, we have briefly discussed about classification and characterization of
systems for application of thermodynamics. We have mentioned there that a system is
homogeneous if it consists of a single phase and heterogeneous if it has more than one
phase. For example, water and ice are different phases of the same compound, H2 0.
The question is: what is a phase? It may be ~tated that a phase is characterized by phase
boundary which is a surface where properties change abruptly (i.e. discontinuously). As
already discussed in Section 1.3.4 and Section 6 .1, heterogeneous equilibria may be classified
into:
l. Phase equilibria for nonreactive systems, where no chemical reaction occurs
2. Phase equilibria for reactive systems, which are accompanied by chemical reactions
(By convention, it is mostly treated as chemical equilibria rather than phase equilibria).
,1;
"
M:tf Textbook of Materials and Metallurgical Thermodynamics
Thus, by phase equilibria, we generally mean heterogeneous equilibria in a nonreactive

' system. In this section, we shall discuss about one-component system which consists of a
single element or compound (also known as Unary system). The question of chemical
reaction, of course, does not arise at all here, as undoubtedly it is a nonreactive system.
'·
6.4.1 Clapeyron Equation

Consider the phase transformation of an element or compound (designated as A).
A (phase I) = A (phase 11) (6.30)
At equilibrium phase transformation temperature (Tu-) and pressure <Pu-),
(6.31)
where GA(I) and GA(II) are Gibbs free energy per mole of A in phases I and II, respectively.
Since transformation occurs at constant T and P, .1G = 0, as per the criterion of equilibrium.
Now, let the pressure be changed to Pu- + dP. Then allow sufficient time for the phase
equilibrium to be re-established at Pu- + dP. Then Tu- will also change to Tu- + dT, GA(I)
will change to GA(I) + dGA(I), GA(ll) will change to GA(II) + dGA(II), at the new equilibrium.
As already stated in Chapter 5, these changes will be related to other variables, as
dG = V dP - S dT (5.14)
Accordingly,
dGA(I) = VA(I) dP - SA(I) dT (6.32)
'
dGA(II) = VA(II) dP - SA(II) dT (6.33)
Again, at the new equilibrium,
(6.34)
From Eqs. (6.31) and (6.34),

dGA(I) = dGA(II) (6.35)
. Combining Eqs. (6.32), (6.33) and (6.35), we get
VA(I) dP - SA(I) dT = VA(II) dP - SA(II) dT (6.36)
i.e.
(6.37)
Again, from Chapter 4,
(4.37)
j
Therefore,
(6.38)
Equation (6.38) is known as the Clapeyron equation, and is applicable to phase
transfomzations involving only condensed phases (i.e. solids and liquids). ( ~~ )eq is the rate
of change of equilibrium phase transformation temperature with change of pressure . ..1V1r>
L1S1r, ..1H1r are difference, respectively, of molar volume, entropy and enthalpy of A between
phases II and I.
Change of pressure has negligible effect on L:1V1r and ..1H1r. Also, change of temperature
is typically very small. Hence, these are taken as constants and equal to their values at the
standard pressure of 1 atm., i.e. at standard states of A in both phases. As discussed in
Section 3.3.2, values at standard state are designated by superscript 'O'. Therefore,
Eq. (6.38) may also be written as
(dT)
dP cq
..1V~· ~
= ..111~ (6.39)
For example, the Clapeyron equation for melting/freezing of a pure element or compound
may be written as
dTm) = ..1V~·~ (6.40)

( dP cq ..11/~
From Eq. (6.19), at constant pressure, dG/dT =-

S. According to the Third Law of
Thermodynamics, entropy of a substance is always positive. This makes the slope of dG/dT
negative, and decreases free energy of a substance with increase of temperature. This is
illustrated schematically in Fig. 6.1 for GA(I) and GA(II). It also shows that
..1G > 0 if T < T1r

=0 at T =Tr 1 (6.41)
< 0 at T > T1r
Thus, phase II is stabler than phase I at T > Tu. and reverse is true at T < T1r.
The sign of (dT)

dP eq
can be either positive or negative. Let us consider melting of solid.
During melting, a substance absorbs heat, i.e. ..111~ is positive. For most solids, the volume
per mole increases upon melting, i.e. ..1\1! is positive. These make the sign of ( ";; )eq
positive as per Eq. (6.40), i.e. Tm increases with increase of pressure. Exceptions to this are
Textbook of Materials and Metallurgical Thennodynamics
Tu
Temperature - - - -
(b)
Fig. 6.1 Variation of: (a) GA (I), GA (II), and (b) LiG, with temperature (schematic).
H20 , bismuth etc., where Li~ is negative and the contraction of volume occurs upon melting.
Tm will decrease with increase of P. Figure 6.2 presents the phase diagram for H20 ; S, L
.•••
l
I •
v .•
•'
----~1H
· i~~---t11~~1~
+0.0024°C +0.0098°C 100°c 374°C
Temperature
Fig. 6.2 Phase diagram for H2 0 ; dashed lines show metastable extensions. (P. Saha, in
Elements of Ceramic Science, D. Ganguly and S. Kumar (Eds.), Indian Institute of
Ceramics, Calcutta, 1982, p. 94.)
and V denote regions of stability of solid, liquid and vapour, respectively. Line SL is solid-
liquid transition boundary obeying the Clapeyron equation. It may be noted that increasing
P lowers T for SL.
EXAMPLE 6.2 Calculate the change of equilibrium transformation temperature of a-iron

into r-iron for raising pressure from 1 to 100 atm, with the help of the following data:
(i) 910°C is the equilibrium transformation temperature for this at 1 atm pressure.
(ii) At 910°C, LiH~~r = 912 J moi- 1, densities of a- and yiron are 7.571 and 7.633
g cm-3, respectively.
(iii) Ignore change of volume with pressure.
Solution
dT)
(dP = LiV~. ~ (6.39)
eq LiH2
~ = 910 + 273 = 1183 K.

55.85
Since the atomic mass of Fe is 55 .85, the molar volume of a - Fe = . = 7.377 ·
7 571
55.85
cm3 mo1- 1, and that of r- Fe =- - = 7 .317 cm3 mo1- 1• Hence,
7.633
LIV~ = 7.317 - 7.377 =- 0.06 cm3 mo1- 1 = -0.06 x 10-6 m 3 moi- 1

Again,
~ =912 J mo1-1 = 912 x 82 06

· x
10
-6 =9.0 x 10-3 m 3 . atmlmol
8.314
since R = 82.06 cm3.atm/(mol. K). Substituting the values in Eq. (6.39), we get
6
dT) = .,_ 0.06 x 1183 x 10- = _7 .88 x 10_3 K/atm
( dP eq 9.0 x 10- 3
Thus, change of T1r (LiT1r) due to change of pressure from 1 atm to 100 atm is
dT
LiTir = - (100 - 1) = - 0.78 K
dP
EXAMPLE 6.3 Carbon has two allotropes, graphite and diamond. At 25°C and I atm
pressure, graphite is the stable form . Calculate the minimum pressure which must be applied
to graphite at 25°C in order to bring about its transformation to diamond.
Given:
(i) Hf98 (graphite) - Hf98 (diamond) = - 1900 J/mol
.
1:11 Textbook of Materials and Metallurgical Thermodynamics
(ii) ~98 (J/mol.K) are 5.735 and 2.428 for graphite and diamond, r~spectively.
(iii) Densities of graphite and diamond at 25°C are 2.22 and 3.515 gm/cm3, respectively.
Solution Transformation of graphite into diamond would be thermodynamically feasible,

if for the process: graphite --) diamond, LIG < 0. Now,
dG = V dP - S dT
dG =V dP, i.e. d(LIG) = LlV • dP at constant T, (E.6.1)
where LlV = Vdiamond - Vgraphite• where V is molar volume. This is how change of pressure
affects relative stabilities of graphite and diamond.
At 1 atm and 25°C (i.e. 298 K),
= 1900 - 298 (2.428 - 5.735)

'
'
82.06 x io-6
= 2886 J mo1-1 = 2886 x 8 .314
= 0.0285 m3.atm/mol
At the minimum pressure required for transformation (Pmin), LIG = 0. Hence, the integral
form of Eq. (E.6.1) is
0
f d (LIG) = JPmin (LIV) dP (E.6.2)
Jo.028.5 I
Now,
1 1
(
LIV= 12 x 10-6(--
3.515
- - -)
2.22
= -1.99 x lo- 6 m3 .mo1-1
Solving Eq. (E.6.2), Pmin = 14.3 x 103 atm.
6.4.2 Phase Equilibria with Vapour: Clausius-Clapeyron Equation

The transformation of a liquid into its vapour is known as vaporization. The corresponding
'
l phase equilibrium is vaporization equilibrium. The transformation of a solid directly into
' its vapour is called sublimation, an'ci its equilibrium is sublimation equilibrium. Both vaporization
and sublimation equilibria will be discussed jointly since their basic Lquations are simi\ar
in form.
Since the molar volume of vapour (Vv) of a substance is orders of magnitude larger than
those in condensed states (Ve), we have
(6.42)
j
Again, vaporization and sublimation can occur over a wide range of temperature. For
example, water vapour co-exists at equilibrium with water from 0°C to 100°C. The equilibrium
vapour pressure (i.e. saturated vapour pressure) is a function of temperature. As Figure 6.2
shows, it is about 5 mm Hg at 0°C and increases with rise in temperature. Therefore, ,1J/1r
and T1r are variables, and hence would be denoted as L1H and T, respectively. In normal
phase equilibria involving vapours, the total pressure in the system is not very large and is
of the order of a few atmospheres at the most. Therefore, ,1J/ may be taken as equal to L1H°,
i.e. ,1H at standard state of 1 atm.
With the above modifications, from Eqs. (6.39) and (6.42),
dP) - -
L1H- 0
(6.43)
(-
dT cq TVv
where L1H° is either &-t;, i.e. heat of vaporization or ~, i.e. heat of sublimation, of a
pure substance.
Assuming the vapour to behave as an ideal gas, we have
V
v
= RT0 (6.44)
PA
where p~ is vapour pressure of pure A in atm. Unit of gas constant R would depend on
the unit of volume. Suppose Vv is in m3, then the unit of R would be m 3.atm mo1-1 K- 1.
The assumption of vapour as ideal gas is reasonable for metallurgy and materials science
since vapour pressures are typicaJly very low.
From Eqs. (6.43) and (6.44),
1 (dP) (6.45)
p~ dT cq
Imagine a closed vessel at temperature T, containing a solid or a liquid. If sufficient

time is allowed, the vapour pressure inside the vessel will keep increasing till equilibrium
is attained and it becomes constant. If the vessel is initially evacuated, then the total pressure
P would be the same as p~. On the other hand, if the vessel was not evacuated to start
with, then P > p~ .
In either case, however, P should be substituted by p~ since it is only p~ , which is
the common pressure to both the vapour and the condensed phase. Hence, Eq. (6.45) should
be further modified as
1 (dp~) L1Ho (6.46)

-po
A
dT
cq
= RT2
i.e .
d(ln p~ )eq L1Ho

dT
= RT2 (6.47)
Textbook of Mate1ials and Metallurgical Thermodynamics
An alternative form of Eq. (6.47) is
d(ln p~)cq
---~= (6.48)
d(l/T)
Either Eq. (6.47) or Eq. (6.48) is Clausius-Clapeyron (CC) equation i11 differential
form. The subscript "eq" on LHS can be omitted as Jong as we understand that we are
referring to equilibrium vapour pressure in the CC equation.
6.4.3 Effect of Total Pressure on Vapour Pressure
The total pressure P in the system should have effect on the value of p~ in the following
ways:
(a) Properties of condensed phase depend on P.
(b) If the system is at a very high pressure (say, abov~ 50 atm or so), then the vapour
would not behave as an ideal gas (see section 6.2.3). In that case, p~ should be
replaced by fugacity of pure A (i.e. J2 ).
This effec t can be neglected for chang._
of P in normal ranges.
Although properties of condensed phase depend on P, they have riegligible effect on
vapour pressure, as the following analysis shows.
Consider the situation where liquid A is kept in a closed chamber at temperature T and
inert gas pressure P. After some time, the pressure of vapour of A in the empty space would
attain the equilibrium value ( p~ ). Now, let us change inert gas pressure P to P + dP in
the chamber. Then after some time, a new vapour-liquid equilibrium will be established
with saturated vapour pressure as P~ + d P~ .
For change of P to P + dP, at new equilibrium, dG of condensed phase, viz.
dGcond = dGvap (6.49)
where dG refers to change of G as a resu lt of change of P. At constant T,
dGcond = Vcond dP, dGvap = Vvap dp~ (6.50)
(6.5 1)
Since Vvap is approximately 103 to 105 times that of Vcond per mole of A, dp~ ldP 1s
0
neghgibly small. i.e. change of total pressure has negligible effect on PA.
6.4.4 Integration of Clausius-Clapeyron Equation

LJ.H° is a function of temperature. However, for many situations, M-fJ may be approximately
taken as constant. Then integration of Eq. (6.48) leads to
!1H 0
lnp~ = - - - + / (6.52)
RT
where I is integration constant. At normal boi ling point, P~ = 1 atm, i.e. In P~ = 0 and
T = ~. Substitutjon of these in Eq. (6.52) yields
1np~ = - fjz 0
(~ - ~) (6.53)
Suppose we are considering vaporization of a liquid. From Eq . (3.26),
(6.54)
where
L).~ =~(v) - ~(!) (6.55 )
In general, for both liquid A and vapour of A , Eq. (3 .11 ) is applicable indjviduall y, i.e.
H~ - H~98 = AT + BT2 - C/ 1 + D (3. 11)
where A, 8, C, D are lumped parameters, and are empirical constants. Equation (3. 11 ) has
been deri ved from
H$ - H~98 = J T ~ ·dT=JT (a+bT+cr 2 )dT (6. 56)

298 298
L1~ = (a + bT + cr 2
)vap - (a + bT + cr-2) 1iq
=& + & · T + LJ.c • r-2 (6.57)
On this basis. we may rewrite Eq. (3.11) as
m 0 =L1H$ =LJ.H~98 + M • T + .1B • T 2 - LJ.C • r- 1 + ,10' (6.58)
=M • T + Af3 • T 2 - LJ.C • r- 1 + ,10

since .1H~98 = a constant and its value is known.
The simplest situation is the assumption L).~ = 0. Then M-fJ = .1D = constant, and
Eq. (6.53) is applicable. If L1~ "" &, the n
(6.59)
d (In p~ )eq = LIA • T + LID = .!_ (LIA + LID) (6.60)

dT RT2 R T T2
Integrating Eq. (6.60), we obtain
In (p~ )eq = ~ (LlA · In T - Li:) + constant = ; + n In T + I (6.61)
where m, n and I are lumped constants obtained from empirical Cp, Lill and vapour pressure
data. Equation (6.61) is the most widely employed equation. Since we are dealing with equil\-
brium vapour pressure, the subscript "eq" is generally omitted. Figure 6.3 shows vapour
pressure vs. temperature curves for some liquid metals. The unit of vapour pressure is atm.
-2
-3
-4
Oil
::i::
E 10-
2
-5 ~
E t(
...::sa) -6 ~
..."'"'
Q)
p..
10--4 -7
-8
-9
- 10
io-s 100 200 400 600 1000 2000 3000

Temperature, °C
Fig. 6.3 Equilibrium vapour pressures of some metals as functions of temperature. (J.F. Elliott and
M. Gleiser, Thermochemistry for Stee/making, Vol. 1, Addision-Wesley (Mass.), 1960, p. 209.)
It may be mentioned that, if all terms in Eq. (6.58) are retained, then we obtain the
most general vapour pressure vs. temperature relation as
n1 k
In p~ = - + n In T +IT+ - +I (6.62)
T T2
where l and k are two additional empirical constants. However, this equation is rarely
employed.
6.4.5 Solid-Liquid-Vapour Equilibria: Triple point and metastability
Figure 6.2 has presented phase diagram for H 20. There, the point 'O' is the common
intersection of the lines for all three equilibria: (i) solid-vapour (SY), (ii) solid-liquid (SL),
and (iii) liquid-vapour (LV). 0 is known as the triple point. Figure 6.4 presents the phase
diagram for zinc. Both in Figs. 6.2 and 6.4, pressure means vapour pressure for SV and
LV equilibria which involve the vapour. For solid-liquid equilibria, pressure means total
pressure P above 0. At triple point, it is vapour pressure for SL also. The existence of a
unique triple point for unary systems can be predicted on the basis of Gibbs Phase Rule,
which will be discussed in Chapter 10.
l atm
t Solid
Vapour
Temperature ___.
Fig. 6.4 Phase diagram for zinc (schematic}.
Quite often, phase transformations occur very slowly. There are situations where
thermodynamically stable states (i.e. ultimate equilibrium) are not attained even in centuries
or even in long geologic periods. These are called metastable phases. They have stability,
but not the ultimate stability. Some relevant examples are now given:
l. Cementite (Fe 3C) is most common in microstructures of iron-carbon alloys. However,
cementite is metastable, and graphite is the stablest state of carbon.
2. Glass is an amorphous supercooled liquid. The crystalline state is the stablest. But
glass can remain as such for ages at room temperature. Only upon heating for a
prolonged period above a critical temperature does it crystaJlize.
3. A mixture of hydrogen and oxygen can be kept at room temperature for a very very
long period without reaction. However, it is metastable. If it is ignited by a match
stick, H2 and 0 2 react with explosive speed to form H20.
Thermodynamics is capable of quantitative treatment of metastable phases just like
stable phases. In Fig. 6.2, for example, the dashed lines are extensions of solid lines, and
represent metastable equilibria.
EXAMPLE 6.4 The vapour pressure of liquid zinc above its melting point (692 K) is
given by the approximate formula:
ln~n = - 14.76~ X 103 + 12.805
The heat of fusion of zinc is equal to 6673 J mo1- 1. Derive a formula for the vapour
pressure over solid zinc below 692 K.
Solution At 692 K, solid zinc is at equilibrium with liquid zinc. Hence, at 692 K,
P~n (over solid Zn) = ~(over liquid Zn)

In other words, 692 K is really the triple point. From the data provided, at 692 K,
0
In Pin = - 8.532 (E.6.3)
0
From Eq. (6.52), the Pin vs. T data is to be interpreted as
L1Ho
In P~n(l) = - __ +I
v (E.6.4)
RT
Also,
&lo
ln'4,(s)= - __s +I' (E.6.5)
RT
Now, heat of sublimation = L1H~ = L1H~ + LlH~·. where L1H~ and L1H~ are heats of vaporization
and fusion, respectively. Hence,
L1H~ = 14.765 x 103 +

6673
= 15.568 x 103
R 8314
Equating relations (E.6.3) and (E.6. 5) at T = 692, we get
15.568 x 103
-8.532 = - +I'
692
i.e.
I' = l 3.965
Thus,
3
0 ()
InpZn s =- 15.568 x 10 + 13.965
T
6.5 Summary
1. For an isothermal process (reversible or irreversible) at temperature T,
= =
(a) L1G G 2 - G 1 (H2 - TS2) - (H 1 - TS 1) =
L1H - T • L1S, where 2, I designate
state 2 and state l, respectively .
(b) For change of pressure,
dG = RTd (ln P), for ideal gas
= RTd (ln j), for nonideal gas, and condensed substances
where f is known as fugacity. A gas tends to be more and more ideal, as pressure
is lowered. Then, f ~ P. At room temperature and above, and pressure not too high
(say, lower than 10 atm) , the common permanent gases may be taken as ideal.
2. The two alternative forms of Gibbs-Helmholtz (G-H) equation may be stated as:
8(L1G/T)] = _ L1H or . [8(L1GIT)] =L1H

[ 8T . p T2 8(1/T) P
where L1G, L1H are changes of G and H, respectively, during isothermal processes.
3. Change of equilibrium phase transformation temperature of a pure substance in
condensed state, due to change of pressure, is given by Clapeyron equation, viz.
where L1V1r and L1H1r are changes in molar volume and enthalpy, respectively due to
transformation.
4. For phase equilibria involving a pure condensed phase and its vapour, the Clausius-
Clapeyron (CC) equation is applicable. The two forms of CC equation are
d(ln p~ )eq L1Ho d(ln p~)cq L1Ho

dT
= RT2 ' d( l/T)
-
R
where (p~ )eq is equilibrium (i.e. saturated) vapour pressure of substance A. at

temperature T, a nd L1H° is change of molar e nthalpy due to vapour formation.
5. The commonly used integrated form of CC equation is
ln(p~)eq = m + n In T + I
T
where m, 11, I are empirical constants.
6. The temperature at which solid, liquid and vapour of a pure substance co-exist at
equilibr·1m, is invariant and is known as the triple point.
PROBLEMS
6.1 Find the change in free energy associated with isothermal expansion of 1 mole of
an ideal gas from 2 atm to 1 atm pressure at 300 K.
6.2 Vapour pressure of water at 25°C is 23.76 mm Hg. Calculate the free energy change
for transfer of 1 mole of water to vapour at 1 atm pressure at 25°C.
6.3 The variation of the volume of a liquid with pressure is given by: V = V0(1 - ftP),
where pis the compressibility coefficient and V0 is the volume at low (i.e. virtually
zero) pressure. Derive an expression of L1G for a liquid due to change of pressure
from P 1 to P 2. If, for water at 25 °C, V = 1.0029 cm 3 g- 1 at 1 atm, and p = 49 x
10-6 atm- 1, calculate L1G per mole of water due to compression from 1 to 10 atm
at 25°C. Would it make appreciable difference if pis assumed as zero?
6.4 Solid bismuth has a density of 9.80 g cm-3 at 25°C. Its coefficient of linear thermal
expansion is 14.6 x 10-6/°C. Density of liquid bismuth is 10.07 g cm- 3. The atomic
mass of Bi is 209. What would be the change in melting temperature at a pressure
of 1000 atm?
6.5 Calculate the minimum hydrostatic pressure required to stabilize Fe3C at 298 K.
Given:
(a) 3Fe(s) + C(graphite) = Fe 3C(s), LIG° = 19,045 J mo1-1 at 298 K.
(b) The molar volumes of Fe3C, graphite and Fe are 24.24, 5.34 and 7.10 cm3 • mo1- 1,
respectively.
(c) Assume the molar volumes to be independent of pressure.
6.6 From the vapour pressure vs. temperature relationship for liquid silver, calculate
(a) the heat of vaporization of liquid silver at the normal boiling temperature of
2147 K, and (b) heat capacity difference between liquid and gaseous silver.
6.7 Below the triple point (-56.2°C), the vapour pressure of solid C02 is given as
3ll6
In p(atm) =--T
- + 16.01
The molar heat of melting of C02 is 8330 joules. Calculate the vapour pressure of
liquid C02 at 25°C, and explain why solid C02 evaporates rather than melts in an
open~_£onJainer.
6.8 The chemical reaction inside the retort for ex.traction of zinc at high temperature,
above boiling point of Zn, may be written as
ZnO(s) + C(s) = Zn(g) + CO(g)
The gaseous products of reaction enter a condenser (for condensing Zn(g) into
liquid Zn) at 950°C and leave the condenser at 450°C and l atm. Assuming that
the Zn vapour coming out of the condenser is at equilibrium with liquid Zn at
450°C, calculate the efficiency of recovery of Zn in the cond~nser.
------
Chapter 7
Activity, Equilibrium Constant and
Standard Free Energy of Reactions
7.1 Introduction
7.1.1 Multicomponent Systems-Solutions and mixtures

The basic equations of thermodynamics, as derived and presented earlier, are general in
nature. However, for the sake of convenience and systematic elucidation, we have restricted
applications to: (a) physical processes, and (b) physico-chemjcal process of phase transformation
in the one-component system.
A multicomponent system co nsists of more than one element or compound. Such a
system may be classified as (a) mixture of pure elements and compounds and (b) solution.
These are briefly described now.
Mixture of pure elements and compounds

An example of this is pure metal, pure oxygen gas and pure metal oxide contained in a closed
vessel. It has three components. But each of them is present as pure, separate phases. In other
words, it is a three-phase system consisting of mixture of pure elements and compound.
Thus, a m ixture is a heterogeneous system by defi nition.
Solution
A solution is a phase containing more than one component, where the components are mixed
(i.e. dispersed) at atomic/molecular level. Aqueous solution with dissolved salts, acids etc.
is an example. A NaCl solution in water has two components: NaCl, H 2 0 . It is a binary
solution. If the sol ution contains H 20, NaCl and KN0 3, then it is a ternary solution, and so
on.
96
Activity, Equilibrium Constant and Standard Free Energy of Reactions
A solution has tl:e following characteristics:

(i) lts composition can be changed (i.e. variable composition)
(ii) It is a single-phase substance (also known as a homogeneous substance).
In realily, we generally deal with a mixture of different solutions, which is a heterogeneous
system consisting of more than one phase.
Some examples of such systems are:
• Aqueous solution and oi l since they stay separately, there are two phases, but each
of them is a solution.
• Mercury amalgam and aqueous solution co-existing as separate phases, but each of
them is a solution.
• Air and aqueous solution, air being a solution of 0 2, N2 , etc.
• Metallic solution, oxide solution and gas mixture co-existing as separate phases
(3-phase system).
7.1.2 Physical, Chemical and Physico-chemical Processes
In a physical process, such as expansion/contraction of a gas, the composition of the substance

remains the same. The system may be homogeneous or heterogeneous. substances may be
pure or solutions. It does not matter. Here, the entire substance behaves uniformly.
In contrast, in chemical reactions, and physico-chemical processes such as phase trans-
formalion, diffusion, behaviours of individual components would be different, if it is a
multicomponent system.
Some examples are as follows:
1. If an alloy of copper and gold is exposed to oxygen at high temperature, copper is
preferentially oxidized to copper oxide. If HCl is added to an aqueous solution of
NaCl and KOH, it will react only with KOH.
2. If a salt solution is cooled, water preferentially freezes out as pure ice. The aqueous
solution, as a result, becomes more and more concentrated, and reaches solubility
limit for salt, which then starts precipitating out as solid.
3. Similarly, when a liquid silver-copper solution is cooled below 780°C, it forms two
separate phases. i.e. one solid solution rich in silver (designated as a-phase), and
another solid solution rich in copper (designated as {3-phase).
Thus, for applying the laws of thermodynamics to chemical and physico-chemical
processes, each component is to be considered separately since it behaves differently from
another. For a system consisting of pure substance. it is straightforward and can be done on
the basis of what we know already. Thi s nas been illustrated in Chapter 6. However,
application to solutions requires additional information in the form of additional terms,
concepts, equ:itions and data.
7.2 Thermodynamics of Solutions-Som~ introductory

definitions and concepts
7.2.1 Composition Parameters-Mole fraction
Amongst various composition parameters, the mole fraction has been adopted as the fundamental
and generally l1sed one in chemical thermodynamics for nonaqueous solutions. Components
in a solution (e.g. H20 , NaCl etc. in aqueous solution) would be designated by subscripts
l, 2, ... , i, j , ... . The general symbol is i.
xi = _ __ __,ni_ __ _ = _!l_ = !i.

(7 .1)
n1 + '1i + ... + ni + nj + ... :E n; "r
I
where n; denotes the number of moles of component i in the solution, nr is the total number
of moles in the solution, and X; is the mole fraction of component i in the solution. By
definition,
:E Xi= 1 (7.2)
I
A gas mixture is also a solution, and an ideal gas mixture obeys Dalton ~s Law of partial
pressures, viz.
Pt + P2 + · · · + P; + Pj + ··· - 7Pi =Pr (7 .3)
x. = J!.L (7.4)
' Pr
where p; is partial pressure of component i. As already' stated, the unit of pressure is

conventionally standard atmosphere (i.e. 760 mm Hg) in chemical thermodynamics.
7 .2.2 Partial and Integral Molar Properties

Consider an extensive property of a solution and let it be designated by the general symbol
Q. Then Q may refer to U, H, G , V, S, A etc. For further discussion, let us restrict the
symbol Q to mean only molar properties of a solution. Then,
(7 .5)
where Q' = the value of the quantity Q for the entire solution and Q = integral molar value
of Q in the solution. At constant P and T,
Q' = f(n., nz, ... rl;, .. .) (7.6)

'
Differentiating Eq. (7.6), we get

.
I
.
dQ' 80.')
= (--
on1 P.T.ni ....,n;.
(0<2'
dn 1 + - -
0"1,
) n1, ...,n;,
(7 .7)
except n1 except "2
Let
(80.')
on; P. T , n , •••,except n;
1
=Q; (7.8)
Then at constant P and T,

'
(7.9)
Combining Eq. (7.9) with Eqs. (7.1) and (7.5), we obtain II'
II•
I :I: Q•dn.
dQ = -dQ
nr
= I
-'----
nr
I I
= :I:1 Q1 dX1 -
(7.10)
For Gibbs free energy, we may write
(dG')P,r = ~ G1 dn1,
I
(dG)p, r = ~ G1 dX1
I
(7.11)
Q1 is called partial molar value of Q of component i in the solution. It is a function of

composition of the solution. Its physical significance can be explained as follows.
Imagine a very large quantity of a solution. Add 1 g-mole of pure i into it. Upon
dissolution, the total quantity of the solution increases by 1 g-mole. As a result, the value
of Q' also increases. But the composition of the solution does not change significantly by
addition, due to its large quantity. In this case, if the final pressure and temperature after
dissolution is the same as the initial one, then the increase of Q' is equal to ~. Therefore,
Q1 is the value of Q per mole of i in the solution. Again,
(7.12)
where (f = Q per mole for pure i, and Q,m = change of Q upon dissolution of 1 mole of ,.
pure i. From Eq. (7.12), for Gibbs free energy per g-mole of i,
(7 .13)
where G1 = partial molar free energy of i in the solution and G;" =partial molar free energy
of mixing of i in the solution.
7 .2.3 Fugacity and Activity of a Component in a Solution
From the above discussions, it is evident that G, is nothing but G per mole of component
i in a solution. Therefore, in analogy with Eq. (6.9), we may write
(7.14)
where
f; = fugacity of component i in solution
]; = fugacity of pure i(i.e. at standard state of I)
Activity of component i in a solution ( ai) is defined as
0
ai = f;ffi (7.15)
From Eqs. (7.14) and (7. 15),
(7.16)
In an ideal gas mixture, Pi = f;. Hence,
G;" = RT In (p/p?) (7.17)

. 0
Smce P; = 1 atm by convention, Eq. (7.17) becomes
G[" = RT lnp; (7 .18)
Therefore, from Eqs. (7.16) and (7.18), for ideal gas mixture,
a;= P; (7.19)
However, it is to be noted that ai is dimensionless and the unit of Pi is standard atmosphere.
Therefo re, a; is only numerically equal to the value of Pi in atm, for ideal gases.
The validity of Eq. (7.18) has also been verified by experiments. A closed container was
partitioned into two sections with a semi-permeable membrane, which allowed passage to
oxygen only. One section was evacuated. Air at I atmosphere was filled into the other
section. Then sufficient time was allowed for attainment of equilibrium by permeation of
oxygen through the membrane. It was then found that P0z in 0 2 + N 2 mixture was equal
to pressure of pure oxygen <P0i) in the section, which was initially evacuated. In other
words, in an ideal gas mixture, it is the partial pressure, which constitutes the thennodynamic
parameter, not the total pressure.
7 .3 Equilibrium Constant ( K) For a Chemical Reaction
7.3.1 Relationship of K with Standard Free Energy of Reaction

Consider the following reaction at constant temperature and pressure:
LL + mM + . . . = qQ + rR + ... (7.20)
where L, M, ... and Q, R, ... are general symbols for reactants and products and l, m, ...
and q, r, ... denote number of moles. Then,
LIG =(qGQ + rGR + ...) - (IGL + mGM + ...) (7.21)
Lid =(qGg +~ + ...) - (lGf +m~ + ...) (7.22)
where LIG = actual change of free energy as a result of the reaction and Lid = value of LIG
if reactants and products are at their respective standard states at the temperature T of
reaction. From Eqs. (7.21) and (7.22),
L1G - £ =l q(GQ - Gg) + r( GR - ~) + ... J- lL(GL - (if)+ m(Gm - ~) + ... J

(7.23)
Combining with Eq. (7.16), Eq. (7.23) may be rewritten as
LIG - Lia° = RT[(q In aQ + r In aR + ...) - (l In aL + m Jn aM + ...)]
=RT In ~I .aR···]
[
m =RT In J (7.24)
aL • t2M .•.
The parameter J is known as activity quotient. At equilibrium,
(L1G)p,T = 0 (5.44)
and hence,
Lid = - RT In [1) 31 eq = - RT In K (7 .25)
where K is the equilibrium constant for reaction (7 .20).

Again, combining Eqs. (7.24) and (7.25), we get
LIG = RT In (JIK) (7 .26)
7 .3.2 Comments on Application

Equation (7.25) constitutes the basis for equilibrium calculations at constant T and P.
Equation (7.24) or (7.26) is employed for predicting fea.sibiliiy of a reaction.
For an irreversible (i.e. spontaneous) process,
(LIG)T.P < 0 (5.62)
This equation is satisfied only when
J/K < 1 (7.27)
It should be emphasized again that, by constant T and P in Eqs. (5.44) and (5.62), we
mean that the initial and final temperature and pressure are the same. During the course of
reaction, both T and P may vary in any manner without affecting conclusions.
Since the standard state of a substance is at l atm pressure, L1G° of a reaction is function
of temperature only. Therefore, the equilibrium constant K is also a function of temperature
only.
Free energy is a state property. Hence, Hess ' Law (see Section 3.3.4) is applicable. Thus,
Lid =(for reaction 7.20) = l: L1Gj> - l: L1Gj>

products rcactams
= l qL1dr(Q) + rL1Gj>(R) + ... j - llL1Gj>(L) + mL1dr(M) + ...j (7 .28)
where tJd/ is standard Gibbs free energy of formation of a compound from its elements at
temperature T. Obviously, L1dr for an element is zero at any temperature, by definition.
Therefore, if we have values of Liq> s, we can calculate L1G° of a reaction. Hence, we shall
concentrate on variation of L1Gj> with T in the following sections, and also present a
graphical method of representation of the same, known as the Ellingham diagrams.
7.4 Variation of L1G0 and K with Temperature
7.4.1 Derivation of Relationship of AG? with Temperat ure
On the basis of Gibbs-Helmholtz equation,
8(L1GIT)] = __
fill_
[ (6.28)
8T P T2
we may write, for formation reactions at constant pressure,
d(L1Gj>IT) __ L1H?
(7 .29)
dT T2
where L1H? = standard enthalpy change for formation of a compound from its elements at
temperature T. Since pressure is constant, Eq. (7.29) has been simplified and written as exact
differential equation.
If there is no phase change, then on the basis of Kirchhoffs Law, we may write
till? (T) =till? (298) + JT

298
L1C,: dT (7.30)
where
:E
elements
cp0 (7.31)
By Eq. (6.57), we have
Li~ = L1a + .1b • T + Lie • r 2 (7.32)

b

,,,,,
d(~) Aflf(298) +I: (Lia+ L1b. T +~) d'T

= - ~-----~---------~
2
dT T
= - ; 2
[ L1Hf (298) + Lia(T - 298) ll
•
+ -L1b
2
(T 2 - 2982 ) - L1c - (1T 1)]
- -
298
(7.33)
Rearranging terms, we obtain
.I
1[
- 2 &lf (298)
r 2
1]
- L1a • 298 - -L1b • 2982 + L1c • -
298
L1a L1b L1c k Lia L1b L1c

- - + -3 = - -
2
--+- (7.34)
T 2 T T T 2 T3
where k .is a constant. This is because Aflf (298) is also a constant. Integrating Eq. (7.34), IT•
we get
L1Gf k L1b L1c

- - = - - L1a In T - - T - - - + e (7.35)
T T 2 2~
where e is integration constant.

Arrangements of terms and generalization yield an equation of the following form:
'
L1Gf = I + mT + gT In T + nT2 + dT 1 (7.36)
where /, m, n, d, g are constants. They are lumped parameters based on the values of
L1Hf (298), Lla, L1b and & .
Equation (7.36) is the most general form of the variation of Lld/ with temperature.
However, it is not normally employed since not all terms are significant. Terms which are
small rea11y lie within the error Jimits of experimental data on L1Gf or L1Hf. Ignoring the
terms L1b • T and L1c • r 2 in Eq. (7.32), since they have been mostly found to be small, we
have
1
2 [,1//f (298) +Lia • T - Lia • 298] (7 .37)
T
'
"
{
l
,,,, Textbook of Materials and Metallurgical Thermodynamics
Following the derivation procedure of Eq. (7 .36), we may write
L'.1Gf = I + mT + gT In T (7 .38)
Most thermochemical data books express temperature dependence of u:ft in the form of
Eq. (7 .38). For many reactions, ~ is approximately independent of temperature (i.e.
Li~ .., o) over a limited temperature range. Then Eq. (7.38) gets further simplified into
dft =I+ mT (7.39)
7 .4.2 Alternative Derivation Procedure of L1Gf vs. T Relation

For an isothermal process, from Chapter 6,
I
LiG = &I - T LlS (6.3)

Therefore,
(7.40)
Noting the expressions for the variation of entropy with temperature from Section 4.3.2,
if there is no phase change, then for a pure substance,
~
_!!_ dT (7.41)
T
I
On the basis of the above, we may write
~ (T) = LiS~ (298) + f

T
293
LiCO
_
T
P dT (7.42)
Hence, Eq. (7.40) becomes
I
I
£JG?cn =[&?<298) + J: ~ ar] - r[ &1<298) + J: '7 ar] (7.43)
Combining Eq. (7.43) with Eq. (7.32), we can arrive at Eq. (7.36) as well. This is the
alternative procedure for derivation of Eq. (7.36). If Li~ = 0, then both ~ and L1.S'? are
independent of temperature, and Eq. (7 .39) is obtained. Then, combining Eqs. (7 .39) and
(7 .40), we get
,' L1G? =&I~ - T· ~=I +mT (7 .44)
I
EXAMPLE 7.1 Calculate LiG for the process: Al(s) = Al(!) at 1000 K.
f
Gived: for aluminium, melting point = 933 K; heat of fusion at melting point = 10.9
x 103 J mo1- 1; Cp = 32.5 and 29.3 J mo1- 1K- 1 for solid and liquid, respectively.
'.:
l
Solution
LIG1000 = LIH1000 - lOOO(LIS)1000 (E.7. 1)
1000
Llli1000 = Llli933 +
J.933 (LICp) dT
= 10.9 x 103 + (29.3 - 32.5)(1000 - 933)

= 10,686 J mo1- 1
1000 LICp
LlS1000 = LlS933 + f933 - - dT
T
= 10.9 x 103 + (29.3 - 32.5) In 1000

933 933
= 11.461 J mo1- 1 K- 1
Hence, from Eq. (E.7.1), we have
LIG 1000 = 10,686 - 1000 x 11.461 =- 775 J mo1-1
EXAMPLE 7.2 (i) Calculate Lie° at 600 K for one mole of the reaction:
NiO (s) + Co (s) = Coo (s) + Ni (s)
(i i) How much error is involved if LICp for this reaction is assumed to be zero?
Given: (i) LIHJ at 298 K for CoO and NiO are: -239 and -244.6 kJ mo1- 1, respectively.
(ii) ~98 values are 30.0, 52.9, 29.8, 38.l J mo1-1K- 1 for Co, CoO, Ni, NiO, respectively.
(ii i) The values of Cp (J mo1- 1K- 1) are as follows:
Co: 21.4 + 14.3 x 10- 3 T - 0.88 x 105 T - 2
CoO: 48.28 + 8.54 x 10- 3 T + 1.7 x 105 T-2
Ni: 32.6 - 1.97 x 10- 3 T - 5.59 x 105 T-2
NiO: -20.9 + 157.2 x 10-3 T + 16.3 x 105 T- 2
Solution (i) From Eq. (6.3),
LIG°(i()() = L1H2oo - 600 x LIS°(,()() (E.7.2)

From Eq. (7.30),
(E.7.3)
From Eq. (7.42),
o o J (i()() LICp dT
LIS(,()() =L1S29s + 298 T (E.7 .4)
[Note: If there are phase transformations, then the more general form of Kirchhoff equation,
as given in Chapters 3 and 4, are to be employed in Eqs. (E.7.3) and (E.7.4).]
Now,
L1H~s =L1HJ. 29g(CoO) - L1HJ. 298(Ni0), (from Hess' Law)
= (-239 + 244.6) x 103 = 5600 J mo1- 1
l
~98 = ~(Co0)+~93(Ni)j - l~(Ni0)+~98 (Co)j
= 52.9 + 29.8 - 38.1 - 30.0 = 14.6 J moi- 1K-1
LiCp = [Cp(CoO) + Cp(Ni)] - [Cp(NiO) + Cp(Co)] (E.7 .5)
From the data provided and from Eq. (E.7.5),
LiCp = 80.38 - 164.9 x 10-3T - 19.31 x H>5r2 J mo1- 1K- 1
Solving Eqs. (E.7.3) and (E.7.4), we get
L1H~ = 4250 J, &'~ = 12.99 J K- 1
L1G°600 = 4250 - 600 x 12.99 = - 3544 joules

(ii) If LiCp of the reaction is assumed to be zero, then
AU0 _ AU0 u.0 _ AC.0
'1n600 - £J1 1 298, Ll.)6oo - ~
Therefore,
Lid' = 5600 - 600 x 14.6 = - 3160 joules
7 .4.3 Variation of Equilibrium Constant with Temperature

On the basis of Gibbs-Helmholtz Equations (6.28) and (6.29), we may write
d(LJG 0 IT) LJH 0 (7.45)

dT =-7
0
d(LJG IT) =LJHo (7.46)
d(l/T)
Combining these with the relation: Lid' =- RT In K, the variation of K with T is given by
either
d(ln K ) LJHO
(7.47)
dT = RT2
or
d(ln K) LJHo
d ( l/T)
= - -R- (7.48)
The above are alternative forms of the Van ' t Hoff equation.
7.5 Ellingham Diagrams

Ellingham diagrams are basically graphical representations (i.e. nomograms} of tJ.d vs. T
relations for chemical reactions. They have some more features, which will be explained in
Chapter 8. They are so called after C.J.T. Ellingham who first constructed them in 1944.
Most common Ellingham diagrams are concerned with free energies of formation of
compounds like oxides, sulph ides, and chlorides. Figure 7.1 presents the Ellingham diagram
for formation of oxides, and is the most common and widely used diagram in metallurgy
and materials science.
7.5.1 Basic Features

(i) The formation reactions for oxides in Fig. 7 .1 may be generalized as
(7.49)
where the values of x and y will depend on the specific compound. M and MxOy are general
symbols for metal and metal oxide, respec ti vely. They can be solid, liquid or gaseous,
depending on temperature. Specific examples are:
2Ni + 0 2(g) = 2Ni0 (7 .50)
Si + 0 2 (g) = Si02 (7.51)

4 2
3 AI + 0 2(g) = 3Al203 (7.52)
Free energy is an extensive property. Hence, the value of t!J.G of a reaction would depend
on the number of moles involved. In the Ellingham diagram, I mole of oxygen constitutes
the basis in the oxide formation reaction. Thus, the values of !J(j/ are per mole of 0 2.
(ii) The tJ.d/ vs. T curve for a reaction consists of some interconnected straight lines.
As long as there is no phase change either in the element or its oxide, !J(j/ is assumed to
vary linearly with temperature in accordance with Eq. (7.44). The slope of a line changes
if a phase change occurs in the element or the oxide. M, B denote, respectively melting and
boiling points of the element. For the oxides, these are denoted by [M], [B}.
Let us take the example of the formation of ZnO, as shown in Fig. 7.1. Liquid zinc boils
.6 at 907°C i.e. 1180 K (at point B), where the slope of the line changes upwards. At point
B, liquid Zn and gaseo us Zn at 1 atm pressure co-exist at equilibrium, and from the
equilibrium criteria, i.e. Eqs. (5.44) and (6.31), for
,.
Zn(I) = Zn(g, l atm), /JGJ = dzn(g) - dzn (l) =0 (7.53)
Hence, at point B, !Jd for the reaction

2 Zn + 0 2(g) = 2 ZnO(s) (7.54)
(7 6])
Textbook of Materials and Metallurgical Thermrv'
0 200 400 ~
or-----oH:'~u.;., r>
_.,,.
-l
-~~
-lt 0 ~
ABSOLUTE ZEf\
Fig. 7.1 Ellingham 1

respective e,
J. Iron Steel
7.5 Ellingham Diagrams

Ellingham diagrams are basically graphical representations (i.e. nomograms) of L1G° vs. T
relations for chemical reactions. They have some more features. which will be explained in
Chapter 8. They are so called after C.J.T. Ellingham who first constructed them in 1944.
Most common Ellingham diagrams are concerned with free energies of formation of
compounds li ke oxides, sulphides, and chlorides. Figure 7.1 presents the Ellingham diagram
for formation of oxides, and is the most common and widely used diagram in metallurgy
and materials science.
7.5.1 Basic Features

(i) The formation reactions for oxides in Fig. 7. l may be generalized as
(7.49)
where the values of x and y will depend on the specific compound. Mand MxOy are general
symbols for metal and metal oxide, respectively. They can be solid, liquid or gaseous,
depending on temperature. Specific examples are:
2Ni + 0 2(g) = 2Ni0 (7 .50)
Si + 0 2 (g) = Si0 2 (7 .51)
4 2
3Al + 02(g) = 3Al203 (7 .52)
Free energy is an extensive property. Hence, the value of L1G of a reaction would depend
on the number of moles involved. In the Ellingham diagram, 1 mole of oxygen constitutes
the basis in the oxide formation reaction. Thus, the values of L1Cf/ are per mole of 0 2.
(ii) The L1d/ vs. T curve for a reaction consists of some interconnected straight lines.
As long as there is no phase change either in the element or its oxide, L1cft is assumed to
vary linearly with temperature in accordance with Eq. (7.44). The slope of a line changes
if a phase change occurs in the element or the oxide. M, B denote, respectively melting and
boiling points of the element. For the oxides, these are denoted by [M], [B].
Let us take the example of the formation of ZnO, as shown in Fig. 7.1. Liquid zinc boils
at 907°C i.e. 1180 K (at point B), where the slope of the line changes upwards. At point
B, liquid Zn and gaseous Zn at 1 atm pressure co-exist at equilibrium, and from the
equi\i.brium criteria, i.e. Eqs. (5.44) and (6.31), for
Zn(!) = Zn(g, 1atm), (7.53)
Hence, at point B, L1G° for the reaction

2 Zn + 0 2(g) = 2 ZnO(s) (7.54)
-·
I
-·
CHANCiU 0# ITATC nEMCNT OXIO(

MCI.TING POINT M
IOK.ING ,OINT I
2eoV SUklMATICHI Pt
TONSITION Pt
S
T •
-
0 200 .00 M>O IOOO
ABSOLUTE ZERO TEMPERATURE
Fig. 7.1 Ellingham diagram giving standard free energies for formation of oxides from their
respective elements, as function of temperature [F.D. Richardson and J.H.E. Jeffes,
J. Iron Steel Inst. , 160, 261 (1948)].
is the same for both Zn(!) and Zn(g). Thus, there is 1w discontinuity in the curve at B. With
similar arguments, there will be no discontinuity at points M, [M] , [B].
(iii) The changes in slopes of lines at M, B etc. can be explained as follows. From
Eq. (7 .44), the slope of the line of .LiGf
vs. T is - ~ and the intercept of the line at
absolute zero (i.e. 0 K) is AH?. For ZnO formation below 1180 K, let A.s1 =Li.sf
and
LJH? = &I~ . Then,
LLS'? = 2 ~nO(s) - 2 ~n(I) - sg2 (g)
(7.55)
(7.56)
Above 1180 K, let A.s1 = ~ and &I~ = ~. Then,
~ =2 s~nO(s) - 2 ~(g) - s&i (g)

(7.57)
Lili~ =2 H~<s> - 2 H~n<g> - Hg2 (g) (7 .58)
Combining Eq. (7.55) with Eq. (7.57), and Eq. (7.56) with Eq. (7.58), we get
~ - &'? =2l.st(I) - .st(g) J= - 2 ~(Zn) (7.59)
(7 .60)
where ~ and &I; are entropy and enthalpy of vaporization of liquid zinc. Since ~ is
a positive quantity, ~ - &f <0, i.e. ~< &f.
As, the slope of .LiGf
vs. T curve is
-A.s1 and - ~ > - &f, the change of slope is positive at point B. Similarly, at point M,
the slopes of lines change in the upward direction, whereas for points [M] and [B], they are
opposite.
7.5.2 Nature of Slopes
In Fig. 7 .1, all lines slope upwards except those for the formation of CO and C02• The
explanations for this are as follows: For formation of a metal oxide such as
2M (s, I) + 0 2 (g) = 2 MO (s, I) (7.61)
LiS~ =2 S~Ols, I) - 2 S~(s, I) - sgl (g) (7 .62)
Now, entropy is a measure of disorder of a substance (see Section 4.4.2). More the
disorder, the higher is its entropy. Since the gas is a highly disordered. state in comparison
to solid or liquid, entropy of oxygen gas is much larger than the entropy of metal and oxide
in Eq. (7.62). Therefore, L15'° "" -S~<&>' and hence the slope(= -LIS°} is positive. For,
C (s) + 0 2(g) = C02(g) (7 63)
one mole of 0 2 disappears, but 1 mole of gaseous C0 2 forms. Hence, L1s° is very small and
the line is horizontal in the Ellingham diagram. For,
2C(s) + Oi(g) = 2CO(g) (7.64)
The situation is reverse of Eq. (7 .62). One extra mole of gas is produced. Hence, L1s° for
the reac tion of CO is positive and the slope is downwards.
The above features of CO and C0 2 have very important consequence in extraction and
refining of metals as well as in chemical industries.
7 .5.3 Assessment of Relative Stabilities of Compounds

Graphical representations are not precise. Hence, for getting the precise values of free
energies of formation, thermochemical data tables or equations such as Eq. (7.38) should be
used. The Ellingham diagram is for finding out approximate values of L1d, quickly. In
addition, it is a quick guide for prediction of relative stabilities of compounds. For example,
at 1000°C, from Fig. 7.1, for the reaction
Si (pure, s) + 0 2 (g, 1 atm) = Si02 (pure, s) (7.65)
L1G65 = L1G25 = G2;0z - G2; - c&i = - 678 kJ per mole 02 (approximate) (7.66)
(after conversion of kcal into kJ)

Similarly, for reaction (7.64) and with C, CO, 0 2 at their respective standard states, viz.
pure solid carbon, CO and 0 2 at 1 atm each,
L1G64 =L1~ = - 448 kJ per mole 0 2 (approximate)
As points of clarification, L1G65, L1G64 denote L1G for reactions (7.64) and (7.65), respectively.
They are equal to L1~5 and L1~ respectively, since all reactants and products are at their
respective standard states at the temperature under consideration. That is why solids and
liquids have to be pure and gases at 1 atm.
Subtracting reaction (7 .65) from (7 .64), we obtain
2C(s, pure) + Si02 (s, pure) = 2CO (g, 1 atm) + Si (pure, s) (7.67)
for which L1~7 = L1~ -L1~5 = + 230 kJ per .mole of reaction (7.67).
As discussed in Section 5.6, a positive value of L1G (in this case, L1G°) means that
reaction (7 .67) is not feasible at constant' temperature and pressure. Therefore, pure carbon
cannot reduce pure Si0 2 at 1000°C and Pco = 1 atm. In other words, Si02 is more stable
as compared to CO at their respective standard states at 1000°C. This is what is meant by
relative stabilities of compounds.
In the Ellingham diagram, the line for Si02 is below that of CO at 1000°C. To
generalize, the line for a stabler oxide would be below that of a less stable oxide. Therefore,
Al 2 0 3 is stabler than Si02 .
At 2000°C, Fig. 7.1 shows that CO is stabler than Si02. Hence, carbon can reduce Si02•
This interesting feature of the oxide system allows carbon to reduce any oxide into metal
if we go to a sufficiently high temperature. This is the thermodynamic basis for extraction
of metals from oxides, with carbon as reducing agent.
It is to be specially noted that the actual criterion for feasibility of a process/reaction is
(L1G)r. p < 0 [see Eq. (5.62)]. L1G° is equal to L1G only when all reactants and products are
at their respective standard states, as illustrated above. More general criterion (i.e. L1G) for
aJl states, both non-standard and standard, will be taken up in subsequent chapters.
EXAMPLE 7.3 Predict the relative thermodynamic stabilities of solid Si 3N 4 and BN in a

mixture of Si 3N 4 and B at 1 atm pressure. Assume that all components are pure. BN and
Si 3N4 are high temperature ceramic materials having technological applications (Johnson and
Stratcher, Chapter 5).
Solution Consider the reaction:

Si 3N4 (s) + 4B(s) = 4BN(s) + 3Si(s,l) (E.7.6)
Since all reactants and products are pure and under 1 atm pressure, L1G for reaction (E.7.6)
is the same as L1G°. Hence, if the reaction is at equilibrium at a temperature T, then
(E.7.7)
The forward reaction would occur if L1(j/. < 0 , in which case, BN would have more stability
than Si 3N4 • On the other hand, if L1(j/. > O, then Si 3N4 will be stabler than BN.
From Hess' Law, at temperature T,
Nfi. = 4 Nf/ (BN) -L1Gf (Si3N 4 ) (E.7 .8)

(i) Situation 1 Assume Si is solid.
B(s) + .!_ N 2(g) = BN(s), L1G° = - 108, 800 + 40.6T J mo1-1 at 1200-2300 K.
2
3Si(s) + 2N2(g) = Si 3N.(s), L1G° = -753, 200 + 336.4T J mo1-1 at 298-1680 K.
Substituting the above values into Eq. (E. 7 .8), the equilibrium temperature (Teq) is
obtained from the following equation:
0 = 4(-108, 800 + 40.6Teq) - (-753, 800 + 336.4Teq)
yielding Teq =
1831 K. However, Si melts at 1680 K, and this temperature is above that.
Therefore, this is an invalid solution.
(ii) Situation 2 Assume Si to be a liquid.
3Si(l) + 2N2(g) = Si~.(s), for which L1d = -892, 950 + 419.3T J mo1- 1 at 1680-1800 K
Combining this with L1G,.> of BN, we get
0 = 4(-108, 800 + 40.6Tcq) -(- 892, 950 + 419.3Teq) = +457, 750- 256.9Teq
yielding Tcq = 1782 K. This is a valid solution. If T > Tcq• then .1dj. < 0. Hence, BN would
be relatively stable above 1782 K. Below 1782 K, Si 3N4 would be relatively stable.
7.6 Summary
1. The free energy criterion is useful for application at constant pressure and temperature
(see Chapter 5). Of these, the constant pressure restriction is not significant for
condensed phases. But the process must be isothermal. Hence, from now on, constant
temperature will be assumed in all derivations in this and subsequent chapters, unless
otherwise stated.
2. In a solution, various components have been designated as I, 2, ... , i, j, ... , the
general symbol being i. Mole fraction is employed as the general composition parameter
in thermodynamics of solution of non-aqueous systems.
n. n.
Mole fraction of component i -- X; -- -1:n1- -
-
-'
"r
1
I
where n; is the number of moles of i.

3. Some symbols and definitions of state properties for solution thermodynamics have
been discussed. These are being summarized below with Gibbs free energy as example.
G' = value of G for the entire solution
G'
G = - = integral molar free energy
n.,-
- = (8G'
G; on. ) ' P,T,n 1,n2 , ..., except llj
= partial molar free energy of component i
G;" =G1 - d/ = partial molar free energy of mixing of i
where d/ is free energy per mole of i in standard state

.1Gm = G - cfJ = in.tegral molar free energy of mixing
4. (Jm
I
= RT In (iL_)
/;0 =RT Ina. I
where f; is fugacity and a; is the activity of component i in a solution. For an ideal

gas mixture, a; = p;, where p 1 is partial pressure of i in atm.
5. For the chemical reaction
lL + mM + .. . = qQ + rR + .. ., at constant T
we have
a3 x a~ x ...
L1G = fjd + RT In J, where J = acti vity quotie nt = a~ x aZ x
At equilibrium at constant P and fj,d = - /ff In CJJeq = - RT In K, where K is the
equilibrium constant.
6. According to Hess' Law,
L1G° I: .16'? -
= products I: L1G'?
reactants
where .16'? denotes free energy of formation of a compound from its elements.
7. .1<J? is a function of temperature and the most common form of the relation is:
.1<J? = I + mT + gT In T
where / , m, g are empirical constants.
8. Ellingham diagrams are specially constructed nomograms, giving variation of Lid
with temperature. Their properties and usefulness have been discussed by tiling the
example of formation of inorganic oxides from elements.
PROBLEMS
7.1 Calculate the equilibrium constant for the following reaction at 298 K:
C (graphite) + 2H2 (g) = CH4 (g)

Use the following data for the reaction at 298 K:
&f1 = - 74.85 x 103 J mo1- 1, L1S° = - 80.25 J moi- 1K- 1
7.2 Calculate the molar value of L1G for Si(s) ~ Si(l) at 1500°C. The equilibrium melting
point of solid silicon is 1410°C. Assume L1Cp 0. =
7.3 Calculate: (i) the entropy of transformation of Ti( a ) to Ti(/3) at the normal transformation
temperature of 882°C, and (ii) L1G of transformation at 930°C.
7.4 For 1 mole of the reaction: Ni(s) + 1/2 0 2(g) = NiO(s), calculate &fJ, tj,(}J, L1S° and
tj,d at 500 K.
7 .5 The standard free energy of formation of one mole of MgO is given as
.1<J? = - 603.9 x 103 - 5.355T In T + 393. l T joules
Calc~late the enthalpy and entropy of formation of MgO at 300 K.
7.6 Using the Ellingham diagram for oxides, show which metal will not get oxidized in
superheated steam at 1000°C- Ni or Cr? The conclusions should be based on quantitative
values of Lid.
7.7 Predict the relative stability of Cr(s) and Cr23 C6 (s) at 1500 K.
12 2 3
Given: for Cr2 0 3 (s) + - C(s) = - Cr23 C6 (s) + - 0 2 (g)
23 23 2
.tKfJ = 1084.5 x 103 - 263. l 7T J moi- 1
7.8 For dissociation of silver oxide, i.e. Ag20(s) = 2Ag(s) + 21 0 2(g), the equilibrium
constant is 0.8 1 at 450 K. Calculate &I°, ..1S° and .tKf' for the reaction at 300 K for
one mole of the dissociation reaction .
7.9 The free energy of formation ( ..1dr') of NiO is given as - 150.25 kJ/mol, . and
- 133.43 kJ/mol at 1000 K and 1200 K, respectively. From these data, calculate
(a) equilibrium constant at each temperature
(b) ..1H~ and~ of NiO.
Equilibria Involving Ideal Gases
and Pure Condensed Phases
8.1 Introduction
In Section 7 .3, we have deri ved the basic equations for equilibrium calculations as well as
prediction of feasibility for chemical reactions. For reaction (7 .20),
q r
QQ_•aR ••.
l = (8 .1)
aL·~ ...
Also, K = [1Ja1 eq (8.2)
where a; is activity of component i in a substance (pure or solution).

Ac tivity is an abstract concept and has no ultimate utility to us. What we are concerned
with are calculations of compositions of phases at chemical equilibrium. This requires
knowledge of activity vs. composition relation for a solution. This topic will be taken up
and elaborately discussed in Chapter 9. To sum up, generaJly speaking, activity vs. composition
relations in solid and liquid solutions are based on experimental measurements, which constitute
a major programme of thermodynamic measurements, as we shall see later.
However, we do not require experimental data on activity vs. comj>osition relation for
ideal gases and gas mixtures as well as pure condensed phases. Hence, in this chapter, we
are taking up only equilibrium calculations involving these. Fyrther reaction equilibrium
calculations will be analyzed only in a later chapter after thermodynamics of solutions is
presented.
115
8.2 Equilibrium Calculations for Reactions Involving

Ideal Gases
In metallurgical and materials engineering, we are primarily concerned with systems and
processes occurring at high temperatures. Pressures of gases are also generally not too high.
As discussed in section 6.2.3, the gases and gas mixtures in the above situations may be
treated as ideal without any significant errors. Further, as pointed out in section 7 .2.3, in
ideal gas mixture, ai =Pi• where Pi is partial pressure of component i. It has also been derived
in section 7 .2.3 that, for thermodynamic calculations on the above basis, the unit of pressure
is to be taken as atm, which is the symbol for standard atmosphere (1 standard atmosphere
= 760 mm Hg). Moreover, for an ideal gas mixture,
X. = P; (7 .4)
I f'..
r
where
Xi = mole fraction of component i in the mixture
= volume fraction of component i in the mixture,
Pr = total pressure of all components, i.e.
Pr= P1 + P2 + ··· +Pi +Pi + ··· = I;.P; (7.3)

I
In view of the above, we do not require any special knowledge of activity-composition

relations for equilibrium calculations involving ideal gases. ~t is a straightforward application
of equilibrium relations already derived in Section 7.3. In this chapter, calculation procedures
shall be illustrated through a few examples.
It may also be noted that, at high temperatures (say, above 700-800°C), gaseous reactions
an~ very fast and, hence, chemical equilibria are attained easily. Therefore, the actual composition
of a gas mixture inside a high temperature furnace is generally assumed the same as that at
equilibrium. This makes such equilibrium calculations a very useful tool for applications to
real processes in industries and laboratories.
EXAMPLE 8.1 Hr H20 and CO-C02 gas mixtures are frequently employed in high temperature
expenments in laboratory to obtain a gas phase with known partial pressure of oxygen <P0i.).
These gaseous compounds along with others are also common in metallurgical and materials
processing as well as combustion of fuels at high temperature.
Solution For the reaction

(E.8.1)
we have
£i = - 239.534 x 103 + 8.139T In T - 9.247T, J mo1- 1 (E.8.2)
(Note: mo1- 1 means per mole of reaction as written.)

Equilibria Involving Ideal Gases and Pure Condensed Phases
Again, from Eq. (7.25),
LiG? = - RT In K1 = - RT In [ °" 20
au2 x aoz 112 ]

eq
= - RT lo [ Pu o 2
Pu x p,'12
2
l
Oi eq
(E.8.3)
From Eqs. (E.8.2) and (E.8.3), and noting that R = 8.316 J mo1-1K-1, we have
lo [ Pu2o ] = 239.534 x 1C>3 8.139 ln T

----+--
9.247
(E.8.4)
112
Pu2 x p02 eq 8.316T 8.316 8.316
As a sample calculation, at 2000 K, if we rr.quire POz = 10- 10 atm, then from Eq. (E.8.4),
the corresponding l~: J

f Pu ratio should be 3.4 x 10- 2•
Similarly, for the reaction
1
CO (g) + 20 2 (g) = C02 (g) (E.8.5)
~ = - 282.42 x 103 + 86.82T, J mo1- 1
= - RT In K5 = - RT In Pcai
112
[ Pco x Pai
l eq
(E.8.6)
Therefore,
In [ Peoi ] = _ l _ [282.42 x 1<>3 _ 86 _82] (E.8.7)

Pco X P:J;
eq.
8.316 T
EXAMPLE 8.2 A gas mixture consisting of 1 mole S0 2 and 0.6 mole 0 2 is introduced
into a furnace at 1000 K. The total pressure (Pr) = 1 atm. Calculate the composition after
the mixture attains equilibrium in the furnace.
Solution For the reaction
S02 (g) + 112 02 (g) = S0 3 (g) (E.8.8)
we have
~ =- 94.558 x 103 + 89.37T, J mo1- 1 (E.8.9)

Thus, at 1000 K,
Ps<>J
Ks = 1.87 = (E.8.10)
Ps0i X P~
Let y mole of S0 3 be present in the equilibrium gas mixture. Then, the gas would consist of
(1-y) mole S0 2, (0.6 - 0.5y) mole 0 2, y mole S03
Hence,
11r =1 - y + 0.6 - 0.5y + y = 1.6 - 0.5y mole (E.8.11 )
Combining Eq. (E.8 .11) with Eqs. (7.1), (7.3) and (7.4), we get
ns<>i I-y (E.8.12)

Ps0i =Xs0i· Pr= - - Pr= x 1
n,- 1.6 - 0.5y
Similarly,
0.6 - 0.5y - y (E.8.13)
Po2 = 1.6 - 0.5y' Pso3 - 1.6 _ O.Sy
From Eqs. (E.8.10), (E.8.12), and (E.8.13),
L.6 ! o.5y]
---~----..;;;...._ ___ = 1.87
112
(E.8.14)
- y ] [0.6 - 0.5y]
0.5y 1.6 - 0.5y
Solving Eq. (E.8.14), y = 0.49 mole, and hence the equilibrium gas mixture will consist
of
37.5 vol% S02. 26.5 vol% 0 2 and 36.0 vol % 503 (Ans.)
[Note: In a gas mixture, composition is expressed in volume per cent, which is 100 x
vol. fraction (i.e. 100 x mole fraction as per Gas, Laws).]
EXAMPLE 8.3 A gas mixture consisting of 20% CO, 20% C02, 10% H 2, and 50% N 2
is fed into a furnace at 900°C. Find the equilibrium composition of the gas inside the
furnace.
Solution As already mentioned, percentage composition in a gas mixture is by convention
volume per cent, which is the same as mole percent.
At high temperature, chemical reaction would generate H20(g) as seen from the following
relation:
C02 (g) + H2 (g) = CO (g) + H20 (g) (E.8.15)
For solution, more equations are required. These are obtained from materials balance,
e.g.
(E.8.16)
Let us consider 1 mole of the gas mixture at room temperature. Therefore, the number
of moles at room temperature (i.e. initial) are
(E.8.17)
From the Law of Conservation of Elements, the initial number of moles of each element is
the same as that after equilibrium. Hence, ·
1lc = nco + ncai = 0.4 (from carbon balance) (E.8.18)
no =11cO + nH20 + 2nco2 =0.6 (from oxygen balance) (E.8.19)
nH =2nH2 + 2nH2o =0.2 (from hydrogen balance) (E.8.20)
nN = 0.5 (from nitrogen balance) (E.8.21)

2
From Eqs. (E.8. l 8)-(E.8.20), through the process of elimination,
nco =0.4 - ncai , ~20 =0.2 - ncai , nH2 =ncai - 0.1 (E.8.22)
r..o (for E.8.15)

L1v- =- RT In K 15 =- RT In [Pco x PH2 o
Pc0i x PH2
l
eq
(E.8.23)
Again,
L1G° (for E.8.15) = L1G0 (E.8.1) - L1G° (E.8.5) (E.8.24)
from Hess' Law.
From the values in Eqs. (E.8.2) and (E.8.6) at 900°C,
Ltd (E.8.15) = -1.05 kJ mo1- 1, giving K15 = 1.11

Hence,
Pco x PH20
[ Peoz x PH2
l
eq
= l.ll (E.8 .25)
Again, from Eqs. (7.1) and (7.4),
Tl;
P; = X;Pr = -·Pr (E.8.26)
nr
Combining Eqs. (E.8.22), (E.8.25) and (E.8.26), we obtain
(0.4 - 1lc0i) (0.2 - 1lc0i )]

[ 1lc0i x <11c0i - 0.1)
=1.11 (E.8.27)
eq
Equation (E.8.27) is a quadratic equation. Its solution yields ncai

= 0.157 mole at
equilibrium, i.e. the gas contains 15.7% of C02 by volume. From Eqs. (E.8.21) and (E.8.22),
volume per cent of other constituents at equilibrium at 9oocc is obtained, as follows:
C02 : 15.7%, CO: 24.3%,
8.3 Chemical Equilibria Involving Pure Condensed

Phases and Gas
From the definition of activity in Eq. (7.15), a; = I if the component i is at its standard
state. For a solid or liquid, the standard state is pure substance at its stablest state at the
temperature where the process takes place. For example, aH20 = 1 for pure ice (below 0°C),
and a.t20 = 1 for pure water between 0°C and 100°C. Again, for this situation, we do not
require any special knowledge of activity vs. composition relationship. In this section, a few
examples of equilibria involving such pure condensed phases and gas will be presented as
illustrations.
8.3.1 Clausius-Clapeyron Equation-Alternative derivation

The two alternative forms of this equation for vaporization/sublimation of a pure substance A
(element or compound) have been derived in Section 6.4.2. Here, these will be derived by
employing the equilibrium constant.
Consider the following equilibrium at temperature T:
A (liquid) = A (vapour) (8.3)
~=-RT lnK =- RT In [~:L

3 (8 .4)
where PA is vapour pressure of A at temperature T. By convention, [pA]eq is designated as

P~ for pure liquid A. aA is activity of A in liquid and, by definition, aA = I for pure liquid
A. Hence,
(8.5)
Variation of equilibrium constant with temperature i& given by Van't Hoff equations, viz.
Eqs. (7.47) and (7.48). Combining Eq. (8.5) with these, we obtain
d(lnp~) &I° (8.6)

dT = RT 2
d(l np~) &I°
=--- (8.7)
d(l/T) R
Equations (8.6) and (8.7) are the same as the alternative forms of Clausius-Clapeyron
equation, as already derived in section 6.4.2.
8.3.2 Oxidation-reduction Equilibria Involving Pure Metal (M)

and Pure Metal Oxide (MO)
Consider the following reaction at temperature T:
2M(s, l) + 0 2 (g) = 2MO(s, l) (8.8)
where (s, I) denotes that M and MO may be present as solid or liquid.

By the convention already adopted in this text, LJd' and K for reaction (8.8) would be
designated as ~, K 8.
4 = 2sf/ (MO)= - RT In K8 = - RT In [ 2 Qf..,.

aM. Pai
0
l
eq
(8.9)
where ~(MO) is standard free energy of formation of MO.

If M, MO are pure, then aM = 1 = aMo· Hence, from Eq. (8.9)
(8.10)
where l Pai jeq is Pai in equilibrium with pure M and pure MO.
Phase stability diagram for Cu-Cu 2 0-02

For the reaction
4Cu(s) + 0 2(g) = 2Cu 20(s) (8.11)
&f1 = - 3.39 x 1<>5 - 14.25T In T + 247T joules (8.12)
From Eqs. (8.10) and (8.12),
4 076 104 (8.13)

1n [ Pai ]eq -- - · Tx - 1.713 ln T + 29.69
Figure 8.1 presents this relation graphically. On the equilibrium Pai line, both pure Cu
and pure Cu20 co-exist with 0 2• Below the line only Cu is stable solid, and above the line,
Cu 20 is stable solid. This can be predicted from common sense. However, for quantitative
prediction, Eq. (7 .26) is to be employed. For reaction (8.11 ),
l= - l ]
[
Pai actual
-20
&'-40
-60
-80
4 8 12 16 20
l/T x 104 , K- 1
Fig. 8.1 Relative stabilities of Cu and Cu20 as function of Po2 and temperature.
Substituting these in Eq. (7.27), we get
J (pOi. )cq
= ---=-~ (8.14)
If (pOi. )actual < (po2 )cq , JIK > 1, and oxidation of Cu is not possible. So, Cu is stable solid
below the equilibrium line. The reverse is true above the line.
EXAMPLE 8.4 Determine the lowest temperature at which copper oxide (Cu 20) can
dissociate in a vacuum of 10-5 mm Hg.
Solution For dissociation of Cu 20, LiG for reaction (8.11) should be positive.
Noting that: L1G = L1G° + RT ln J
and J for reac tion (8.11) is
and combining this with Eq. (8.12), the thermodynamic criterion for dissociation is
-3.39 x 105 - 14.25T ln T + 247T + 8.314T In (760 x 105) > 0
i.e.
-3 .39 x 105 - 14.25T In T + 397.87T > 0 (E.8.28)
In Eq. (E.8.28), the coefficient of T is (397 .87 - 14.25 In 7). Since In T is of the order
of 7 to 8 maximum, the coefficient of Tis positive (actually, it has to be positive for a valid
solution). Hence, the higher the value of T, the more positive will be the LHS. Hence, there
will be a lowest temperature (TmiJ above which dissociation of Cu 20 into Cu and 0 2 occurs.
At T = Tmin•
LHS = -3.39 x 105 - 14.25Tmin In Tmin + 397.87 Tmin = 0 (E.8.29)
This equation is implicit and can be solved graphically, or numerically through an iterative
procedure, or simply by trial and error. The solution of Eq. (E.8.29) yields a value of
Tmin = 1140 K . Therefore, Cu 20 will dissociate at T > 1140 K.
EXAMPLE 8.5 The thin layer of silicon can be deposited over a metallic substrate through
vapour deposition technique at high temperature, e .g.
SiCl 4 (g) + 2H2 (g) = Si (s) + 4HCI (g) (8.15)
Assume the temperature is 1200°C and pressure 1 atm. Calculate partial pressures of SiC14 ,
H2 and HCI gas at equilibrium with pure solid silicon. Assume that the molar ratio of
SiC1 4 : H2 in inlet gas is 1:2, as per stoichiometry of the reactants . .tK f' of the reaction is
- 4183 J mo1- 1 at this temperature (Johnson and Stracher, Chapter 3).
Solution
(E.8.30)
Assume the initial number of moles introduced as 1 mole SiC1 4 gas and 2 moles of H2 gas.
Let, upon attainment of equilibrium, n moles of SiC14 the reacted. Then the situation would
be as follows: ·
Gas ~ SiC1 4 H2 HCl

Initial moles l 2 0
Moles reacting or produced n 211 4n
Moles remaining l - /1 2-2n 4n
Since the total number of moles remaining = (1 - n) + (2 - 2n) + 4n = 3 + n, and
PT = 1 atm, we have, at equilibrium,
1- /1 2 - 211 4n
Psic14 =-3-+n-, PH2 = 3 +fl ' PHc1 =- -
3+n
(E.8.31)
Combining Eqs. (E.8.30) and (E.8.31), we get
4
4n )
( 3+n
_ _,,____.:___ = 1.4
2 (E.8.32)
2 - 2n] 1- n
[ 3+n 3+n
i.e.
65.4n4 - 8.4n2 + l 1.2n - 4.2 = 0 (E.8.33)
Solving graphically or by iteration, n = 0.368.

Putting in the value of n into Eq. (E.8.31), we obtain
P SiCl4 = 0.188 atm, PHz = 0.375 atm, PHa = 0.437 atm, Pr= 1 atm (check)
Thermodynamic basis of finding AJ2, Pc.ofPoo2 ratios in equlllbrlum with metal +

metal oxide from Ellingham diagram
In the oxide Ellingham diagram (Fig. 7 1), there are three scales around the basic diagram.
These are marked as P0i • PH2 IPH20 • PcolPc0i. These are logarithmic scales and are employed
to quickly find out the above gas composition parameters in equilibrium with a metal and
its oxide.
The procedure is illustrated with the help of Fig. 8.2 for P0i and PcolPCOi. Let us
find out the gas compositions in equilibrium with Mn (s) + MnO (s) at 1200°C, as an
example.
There are two points, marked 0, C on 0 K vertical line in the Ellingham diagram. The
point M is on Mn-MnO line at the desired temperature. If the straight line joining 0 and
M is extrapolated, it intersects at point A on Po 2 scale. Point A gives the value of Po 2
(in atm) at equilibrium with Mn + MnO at 1200°C. It is approximately 5 x 10-20 atm.
Similarly, extrapolation of CM yields point B, which gives the value of pcrJPc0i ratio in
equilibrium with Mn + MnO at 1200°C. Its value is approximately 3 x 104 • Similarly, for
PH/PH2o ratio, point Hin Fig. 7.1 is to be used with a similar procedure. The thermodynamic
basis for finding Po 2 is Eq. (8.10), from which it follows that
log (pn...)
L1d,
oc - - oc -
PM AQ
oc - (8. 16)
~L. eq T OP OQ
Since OQ is fixed, AQ oc log ( Po 2 )cq, and with a calibrated scale, AQ gives the actual
value of p 0 2 in equilibrium with Mn + MnO.
Next, consider the reaction
2Mn (s) + 2C02 (g) = 2Mn0 (s) + 2CO (g) (8.17)
Equation (8.17) is obtained by combining the following equations:

2Mn (s) + 0 2 (g) = 2Mn0 (s) (8.18)
2CO (g) + 0 2 (g) = 2C02 (g) (8.19)
10-2
-200
10~
N
104
B
io-8
0 .1
u c 10-10
-b0
e
..,),(
....
800
1<>6
1
10-12
10-14
~
1<>8
1 lo-•6
1000
' 10-18
A
1200 10-20
10-22
400 800 1200 1600 2000
Temperature, °C 10-24
Absolute zero Kelvin
Fig. 8.2 Procedure for finding out Po2 and PcolPc0i in equilibrium with a metal and its oxide,
from the Ellingham diagram.
From Hess' Law,

(8.20)
.£1Gf7 is MN in Fig. 7 .2. Again,
L1Gf7 = - RT In ( Pco ) =- 2RT In ( Pco J (8.21)

Pcai cq Pcai cq
Noting that point C lies on CO-C0 2 line,
log ( Pco J oc L1Gf1 oc MN oc BR

(8.22)
Pcai T OP OR'
cq
Since OR' is fixed, with proper calibration, BR becomes equal to log [p00/Pcaileq·
8.3.3 Phase Stability Diagrams of Stoichiometric Compounds

Dependent on Gas Phase
Stoichiometric compounds are those in which the constituent elements are present in simple
ratios, e.g. Si0 2• Al 20 3• These conform to the traditional concept of compounds, as proposed
by Dalton in the early nineteenth century. Here, the elements exhibit a single, fixed valency.
However, many compounds were subsequently found, where elements are not present in
simple ratios. M >reover, the ratio depends on temperature and pressure. These are known as
11011-stoichiometric compounds. The most well-known example is Wustite, whose nominal
composition is FeO, but in reality it is Fe.to, where x varies from 0.86 to 0.97, dependi ng
on temperature and partial pressure of oxygen in the surrounding.
Actually, Fe exhibits two valencies, Fe 2+ and Fe3+, in Wustite. The ratio Fe2+/Fe3+ ions
depends on T and Poi- In reality, therefore, it is a solid solution from thermodynamic point
of view. In this section, we shall deal only with stoichiometric compounds. Activity of such
compounds in pure state is 1, as per convention. The thermodynamic stabilities of compounds
depend on temperature as well as the nature of surrounding gas. A phase stability diagram
is an isothermal plot showing stabilities of such compounds as function of gas composition.
As an example, the construction of phase stability diagram for Ni-S-0 system is illustrated.
Phase stability diagram for Ni-S- 0 system

Several metals are found in ores as sulphide minerals (e.g. Ni, Cu, Pb, Zn). Roasting is
required before extraction. In such roasting, the ore is heated in solid state at a high
temperature in an oxidizing atmosphere. This allows conversion of sulphides into oxides,
sulphates etc. In roasting of nickel sulphide, the expected solid products are Ni, NiO, NiS,
Ni 3S 2, NiS04 . These would be present as pure compounds since solid-state diffusion is very
slow, and solid solution formation does not occur in the finite processing time. The gas is
predominantly a mixture of S02 and 0 2 besides N 2 and which is inert.
The presence or absence of N 2 in gas would not affect the diagram at all. Hence, it is
being ignored. The four independent reactions may be written as
1
Ni(s) + z ·0 2(g) = NiO(s) (8.23)
NiS(s) + z
3
0 2(g) = NiO(s) + S02(g) (8.24)
1
NiO(s) + S02(g) + 20 2(g) = NiSOis) (8.25)
3NiS(s) + 0 2(g) = Ni 3S 2(s) + S02(g) (8.26)

The reason these are independent is because each one contains one solid, which is not there
in other reactions. Hence, none of the above relations can be obtained by combining the
other three reactions.
Since solids are pure, their activities are 1. Hence, the equilibrium relations for the
above reactions can be written as
Ni- NiO: log Po 2 =- 2 log K13 (8.27)
3
NiS-NiO: log Ps0z = log K 14 + 2log Po2 (8.28)
NiO-NiS04 : log Ps0z =- log K 2s - 21 log Po 2 (8.29)
NiS-Ni3S2: log Ps0z = log K 26 + log Po 2 (8.30)
Here, Ps0z , p 0 2 are values in equilibrium with respective solids as shown above. Figure 8.3
presents the phase stability diagram for Ni-S-0 system at 1000 K. The plotting is based
on Eqs. (8.27)- (8.30). The straight lines correspond to two-phase equilibria. The areas
correspond to zones of stability of single solids. Such diagrams were termed as predominance
area diagram by T.R. Ingraham (1967) who first constructed some of these, including the
Ni-S-0 system.
+8
T=lOOOK
+4
-8
-12'--~~__........_~~--"-~~~....i......~~~.L-~~__J
- 20 - 16 - 12 -8 -4 0
log Po2
Fig. 8.3 Phase stability diagram for nickel-sulphur-oxygen system at 1000 K [T.R. Ingraham, in
Application of Fundamental Thermodynamics to Metallurgical Processes, Gordon &
Breach, New York, 1967, p. 187].
Textbook of Materials and Metal/urgi.::al Thermodynamics
8.4 Summary
1. For an ideal gas mixture, X; = p;fPr , where p; is the partial pressure of component
i, and Pr is the total pressure as given by Pr = :E P;. Again, as derived in Chapter 7,
a; = p;, in an ideal gas mixture. '
2. On the basis of the above, and equilibrium relations derived in Chapter 7, some
numeric· I examples of equilibrium calculations involving reactions of ideal gases
have been presented. It is to be noted that, in reality too, gaseous equilibria are
attained rapidly at high temperatures.
=
3. From the definition of activity in Chapter 7, a; l if component i is at its standard
state, which is pure i at its stablest state at the process temperature, for solids and
liquids (i.e. condensed phases). On this basis, some examples of equilibrium relations
have been derived and discussed. These are:
(i) The derivation of Clausius-Clapeyron equation
(ii) Pure M + pure metal oxide + 0 2 equilibria, for which
NJt per mole 0 2 l 1q

= RT ln P0i,
where l P<>i J,q is P0i, in equilibrium with metal + metal oxide mixture at temperature
T.
(iii) The thermodynamic basis of determining P0i,, PcofPeoi and PH2 /pH2o in eqilibrium
with a metal and metal oxide mixture, from the Ellingham diagram for oxide
(iv) The derivation of phase stability diagrams of stoichiometric compounds dependent
on gas phase, illustrated with the example of Ni-S-0 system.
PROBLEMS
8.1 Predict whether a sheet of metallic iron, being annealed in a furnace at 760°C, will
get oxidized by the reaction:
Fe (s) + C02 (g) = FeO (s) + CO (g)
if the atmosphere contains 10% CO, 2% C0 2 and 88% N 2.
8.2 A gas mixture of 50% H2S and 50% 0 2 is introduced into a furnace at 1000 K.
Calculate the gas composition after the equilibrium for the following reaction is attained:
1 1
H2S (g) + 202 (g) = H20 (g) + 2S2 (g)
The total pressure is 3 atm.
8.3 (a) A gas mixture containing 50% C02 and 50% H 2 is introduced into a furnace at
1600°C. What is the value of partial pressure of oxygen in the gas after attainment
of equilibrium?
(b) Using the Ellingham diagram, predict whether this gas will oxidize a piece of
titanium kept in the furnace. Note down the procedure you employed.
8.4 How much heat is evolved when 1 mole of S02 and 0.5 mole of 0 2, each at 1 atm
pressure, react to form the equilibrium SOrS02-0 2 mixture at 1000 K and 1 atm
pressure?
8.5 A partial pressure of Cl 2 of 0. i atrn is to be obtained at 500 K and total pressure
of l atm by establishing equilibrium of the reaction:
PCl 5 (g) = PCl3 (g) + Cl2 (g)

PCI 5
What ratio of - - should the initial gas mixture contain?
PCl 3
Given:
PCl3(g) + Cl2(g) = PCl5(g)
!J.cfJ = - 95, 600 - 7.94T In T + 235.2T J mo1- 1
8.6 Calculate the pressure of inert gas which must be applied to liquid lead at 1000°C
in order to triple the equilibrium vapour pressure of Pb. The density of liquid Pb
at 1000°C is 9.79 g-cm-3.
8.7 Solid Mo0 2 is to be reduced by dry hydrogen gas at 1400°<: to metallic molybdenum.
The total pressure 1s 1 atm.
(a) What percentage of H 2 will be utilized for reduction assuming attainment of
equilibrium of gas with Mo and Mo02 mixture?
(b) Would there be any change if PT 10 atm? =
8.8 Predict whether a gas mixture consisting of 10% H2S and 90% H 20 can convert
solid Cn20 into solid Cu:!S at 1200 K.
8.9 Predict whether a gas mixture containing 50% H 2, 7% HCl, and 43% Ar at 900 K
and 1 atm total pressure is at equilibrium with a mixture of liquid Sn and liquid SnCJ 2.
8.10 One kilogram of solid calcium carbonate is kept in an evacuated chamber of one
cubic meter volume at room temperature, and the system is heated. Calculate:
(a) The highest temperature up to which solid CaC0 3 would be present
(b) The pressure inside the chamber at 1000 K
(c) The pressure inside the chamber at 1500 K.
8.11 Two of the oxides of iron are FeO and Fe30 4 . Solid iron exists in equilibrium with
one of these oxides at low temperatures and with the other one at high temperatures.
Predict which oxide would be at equilibrium with Fe at room temperature and the
maximum temperature for this equilibrium co-existence.
8.12 Liquid Mn is kept in a furnace at 1900 K. The furnace atmosphere consists of
2.5% H 2S, 50% H2 , and the remaining N2 . Predict whether MnS will form.
8.13 Solid ZnO and solid ZnS are equilibrated at 2000 K with a H 2S-H20-H 2 gas mixture
having P11~0 =0.5 and P11 2 = 0.0421 atm. Calculate the equilibrium partial pressures
of 0 2, H2S, S 2 and Zn gas in the atmosphere.
[Note: Gas corr,po~icion is given in volume per cent as per standard convention.]
Thermodynamics of Solutions
9.1 Introduction
For thermodynamic analysis, we can broadly classify solutions into aqueous and nonaqueous
solutions. Aqueous solutions are extensively dealt with in inorganic and physical chemistry
texts. Also, as stated in Chapter l, metallurgical and materials processing of interest to us
are mostly carried out at high temperatures. Here, the solutions are all nonaqueous and
inorganic, and hence, these are of special interest to us. Therefore, this chapter covers
inorganic nonaqueous solutions, especially for high temperature systems. These may be
further classified inco:
1. Metallic solutions
2. Non-metallic solutions-oxide solutions, sulphide solutions, halide solutions, etc.
Again , all these may be further sub-classified into solid solutions and liquid solutions.
Single-phase alloys (Ag-Cu, Fe-Ni etc.) are metallic solutions. A molten slag is an oxide
solution, and may contain compounds like Si02, CaO, and Al 20 3 •
No material (either element or compound) is absolutely pure. It will have some impurities,
may be in trace quantities. Even at such low concentrations they may affect some properties
significantly. Some examples are now given below:
l . Liquid steel dissolves hydrogen to the extent of few parts per million (ppm). Even
this low concentration of H tends to cause fine cracks on surfaces of solid steel
during hot forging, thus damaging its properties.
2. Liquid copper dissolves some oxygen during processing. Even a concentration less
than O. l % decreases ductility and electrical conductivity of Cu. Hence, for high
conductivity copper, oxygen should be removed to a very low level.
3. In semiconductors, even traces of impurities of the order of parts per billion may
affect performance.
130
9.2 Ideal and Nonideal Solutions

Consider vaporization of component A from a liquid solution at temperature T. If the vapour
behaves as an ideal gas, then from Eqs. (8.4) and (8.5), the equilibrium constant is given
as
K -_PA -po
- A (9.1)
aA
i.e. '
aA_ -PA
-
(9.2)
p~
where aA is the activity of component A in the liquid solution . PA and P1 are equilibrium
vapour pressures over the liquid solution and pure A respectively at T.
Equation (9.2) can be understocxl by imagining two closed containers, kept at temperature T.
One contains a liquid solution having species A as a component. The other has pure A. If
the containers are kept at T for a long time, vapour-liquid equilibrium will be attained in
each of them. In one, vapour pressure of A will be PA• and in the other, it will be p~. If,
for a solution,
(9.3)
where XA is mole fraction of A in the solution, then the solution obeys Raoult's Law, and
is called an ideal solution. Solutions which do not obey Raoult 's Law are known as
nonideal solutions. To handle thermodynamics of such solutions, a parameter known as
activity coeff icient (y) is employed.
Y; =-X;Oi , 1.e.
. Oi =r, x; (9.4)
where
Y; =
Activity coefficient of component i in a solution, and X; is mole fraction of i,
defined in Eq. (7. 1).
I f,
r; = 1, solution is ideal.
Y; > 1, solution exhibits positive departure from Raoult's Law.
Y; < l , solution exhibits negative departure from Raoul t's Law.
In a solution , a; < I , in contrast to pure i , where a; = l. This difference is due to the
following two effects:
l. Dilution of component i in a solution due to the presence of other components
(dilution effect).
2. Interaction of component i with other components in a solution (interaction effect).
Roughly speaking, activity is a measure of free concentration, i.e. concentration available

for reaction. In an ideal solution, the entire concentration is available, i.e. there is only
dilution effect.
Interaction effect can be followed more easily if we consider a binary solution A-B. We
may visualize three types of bonds between atoms (or molecules) of A and B, viz. A-A,
B-B and A-B. Qualitatively speaking, if A-B bonds are stronger than average of A-A and
B-B bonds, i.e. A and B have tendency to fonn a compound, then less of A and B atoms
would be free. In other words, the entire concentrations of A and B would not be available
for reaction. In this case, the solution would exhibit negative departure from Raoult' s Law,
i.e.rA < 1, J:B < 1. In the reverse case, there is tendency towards formation of clusters of
A and B, and YA > 1 and J:B > 1.
Cu-Ag, Fe-Ni, Fe-Mn, FeO-MnO solutions are approximately ideal. Silicates, such as
Ca0-Si02, and Mn0-Si0 2 show negative departures due to their tendency for compound
formation. Cu-Fe shows positive departure. These are just some examples. As a generalization,
it may be stated that most real solid and liquid solutions are nonideal.
9.3 Partial and Integral Molar Properties; Gibbs-Duhem

equation
9.3.1 Relations amongst Partial Molar Properties
Partial molar properties have already been defined and explained in section 7 .2.2. Partial
molar quantities of mixing also have been defined there. It may be noted here further that
all thermodynamic relations derived amongst extensive properties in Chapters 2-5, are applicable
here also. For example, in analogy with Eqs. (5.12) and (5.14), we may write
dii; = T d°S; + V; dP (9 .5)
dG; = - ~ err + V; dP (9.6)
dfi'!ll = T dS.Im + V,m dP (9.7)

I I
(9.8)
Similarly,
- - -
G; =H; - TS; (9.9)
Gj =iii - rsr (9.10)

From section 7 .2.2,
- (OG'1
G. = - ·
I Jn.
1 ,P , ni , 112, ..., exceptn,
, - (JS'1
s. = -Jn.
I
1 , P,n-.n2 , ..., except n;
(9.11)
- (8H'l
H.=
I
-8n.
' , P,n1, "2· .... except n;
' -=(8V'
v;
I
- l 8n.
1 • P,llJ ·"2· .... except n;
(9.12)
(9.13)
where ni. 11 2, ... , n; denote number of moles of components 1, 2, ... , i, ... etc. in the solution.
9.3.2 Gibbs-Duhem Equation
At constant T and P, as shown in (Eq. (7 .7)),
dQ ' =(OQ'
-8 l
n1 . "3 ..... "l except "1
dfli + - (OQ'
8ni
) ni. II). ····"l except nz
dni + ... + - (OQ'
8nk
J "1 • "3· ... except "t
dnk
(9.14)
From Eq. (7. 1),
nr = n 1 + n2 + .. . + n; + .. . + nk = l:n;j
(9.15)
Imagine preparation of this solution. One method of preparation is to keep adding small
quantities of components (dni. dn 2 , • .• ) in stages at constant T and P such that the overall
composition of the solution remains the same in all stages. Since the composition is always
constant, the values of Q1, Q2 , .. ., Qk would also remain constant. This is because these are
functions of T, P and composition of solution only.
Therefore, after the additions are complete, we get
(9.16)
Since Q is a state property, it does not matter as to how the solution is actually prepared.
It is the final state that matters. Hence, Eq. (9.16) will be always valid. Differentiating
Eq. (9.16), we get
(9.17)
Rearranging Eq. (9.17), we obtain
dQ' =(Q1 d"1. + Q2 dni + ... + G. dn1:) + (n1dij + ni d~ + ...+ nk dG.)

= l:i ~ dn; + l:; n; d~ (9.18)
Equating Eqs. (9.14) and (9.18),

r. n· dQ. =0
• I I (9. 19)
l
Dividing Eq.(9.19) by nr, we get
r. x.dn. =o
. I lC.j (9.20)
I
Equations (9.1 9 and (9.20) are alternative forms of the Gibbs-Duhem equation (G-H
equation), which provide the principal foundation to thermodynamics of solutions.
For Gibbs free energy, the G-H equation may be written as
r. ni dG; =0
l
or ~ X; dG;
I
(9.21)
For pure components, the values of Q, viz. cf., dJ, ..., di. are constants at constant T and
P. So, d(f, ddj , ... , ddi_ are zero. Hence,
"'£, Tl; dQo

j
=0, ~ X; d(}f = 0
I
(9.22)
Subtracting Eq. (9.22) from Eq. (9.19) or (9.20), and from definition of Q;m in Eq. (7.12),
we get
- 0
!.1i;d(Q; - Q; )
j
=0, i.e.
-
r.j n; dQt•
=0, r. x .dQ!"
I
. I I
=o (9.23)
9.3.3 Integral Molar Quantities for Mixing

Dividing Eq. (9.16) by nT, we obtain
(9.24)
where Q is i11tegral molar value, as already stated (Eq. 7.5).

Suppose we take X1 moles of pure component I , X2 moles of pure component 2, and
so on. Then we prepare a total of I mole of mechanical mixture of these pure components.
In such a case,
(9.25)
where the superscript 'O' denotes properties at the respective standard states of components,
as already noted in Section 7.2.2.
Subtracting Q0 from Q, we obtain
,1Qm =Q - Qo = ~X;(Q; l
- Qo)= ~X;Qt
l
(9.26)
LiQm is known as the integral molar value for mixing of the property Q. This represents
the change in the value of Q per mole of solution when pure components are mjxed to form
the solution. Again, L1Q"' is a function of composition of the solution besides temperature
and pressure.
On the basis of Eq. (9.26), we may write
AGm =G - d' = r.. X.G!"

I
. I I
' Q.27)
where L1Gm = integral molar .Gibbs free energy of mixing of the solution.
EXAMPLE 9.1 If the integral molar enthalpy of mixing (L1/f") of a binary solution is
LX 1X2, where Lis a constant independent of composition, find the expressions for Hf1 and Hi
in terms of mole fractions .
Solution
(E.9.1)
(E.9.2)
From Eqs. (E. 9 .1) and (E. 9.2) ,
1
Hf1 =L [ 'Ii + n1ni(- l) x 1]
n1 +ni (n, + ni)2
since X1 + X2 = I. From symmetry, Hi = LXf.
9.4 Some More Properties of Solutions
9 .4.1 General Features

From Section 7 .2.2,
(7.16)
AGm =RT .I:i x. In a.

I I
(9.28)
Since a; = y;X;,
(9.29)
and,
iJ:n
~ = R(ln X; + In y;) (9.30)
The application of Gibbs- Helmholtz equation [Eq. (6.28)] to partial molar quantities
leads to
jj-rn
=- -·-
T2 (9.31)
In Eq. (9.31), the composition as also pressure is fixed since partial molar properties
depend on composition as well. Since X; is a composition parameter, and is independent of
temperature, we have
8(ln X; )] =O (9.32)
[
OT P.com
8(ln Y;)] = _ _ii,_-rn

[ 2 (9.33)
8T P,com RT
9.4.2 Ideal Solutions

Ideal solutions constitute a special class, and possess some unique characteristic properties.
These are derived as follows:
For an ideal solution, a; = Xi. Hence, from Eq. (7.16)
G;" =RT lnX; (9.34)
From Eqs. (9.27) and (9.34), therefore,
llGm =RT ~X; In X; (9.35)

I
Again, from Eq. (9.8),
[8GilOP T,comp
=vm 1 (9.36)
It is being emphasized again that Gr', V;m are functions of composition besides T and P.
Thus, constant composition restriction is required in Eq. (9.36), besides temperature.
For ideal solution, from Eq. (9.34), it follows that GI'1 is not a function of pressure.
So, from Eq. (9.36),
V;m =O , and hence,
LlVm = l:. x.' vm' =0 (9.37)

'
Again , from Eq. (9.34),
G!'l
-'-
T =R In X. I
-:t:. f(T) (9.38)
Therefore, for an ideal solution, application of Eq. (9.31) leads to
H;m =O , and hence,
LlHm = l:. X.I H!'l = 0

I (9.39)
'
Further, applying the above relation to Eq. (9.10) in an ideal solution, we get
(9.40)
and, hence,
LlSm = - R ~ X; In X; (9.41)
'
9.5 Binary Solutions

Binary solutions have two components, designated here as A and B. Due to its relative
simplicity, and the fact that thermodynamics of ternary and multicomponent solutions are
often extrapolated from respective binaries, all introductory thermodynamics texts are primarily
concerned with thermodynamic analysis of binary solutions on the basis of general equations
already presented.
9.5.1 Activity vs Mole Fraction; Henry's Law
Figure 9.1 presents activity vs mole fraction curves schematically for binary A-B solution
at constant temperature. It shows Raoult's Law (RL) lines for components A and B. Positive
and negative departures from Raoult"s Law are illustrated for a 8 . It may be noted that at
small values of X 8 , the variation of a 8 with X8 is linear. This is the basis for Henry's Law
(HL) for binary solutions. Henry 's Law may be rigorously stated as follows: If X8 ~ 0, then
~ ~ a constant ('fa). In other words, in Henry 's Law region for B,
Raoult's Law lines

for B
\
Raoult' s Law lines
o _____________
for B _
A ___.. B B
Mole fraction of B(X8 ) XB XB = 1
(a) Positive deviation (b) Negative deviation
Fig. 9.1 (a) Positive, and (b) negative deviations from the Raoult's Law for component B in the
binary A-B solution; Henry's Law lines for B are also ,shown (schematic).
(9.42)
where 'fa
is a constant. B is the solute, and A is known as solvent.
Similarly, if XA --+ 0,
(9.43)
One way of classifying solution is division into dilute (low concentration of solute) and
concentrated solution. On this basis, it may also be stated that the solute in binary dilute
solution obeys Henry 's lAw. There is no fixed demarcation between dilute and concentrated
solutions. From a fund.amental point of view, Henry ' s Law is expected to be obeyed if solute
atoms are so far apart from one another ~hat we can ignore solute-solute interactions, i.e.
B-B or A-A interactions, as the case may be.
It may also be observed in Fig. 9.1 that the curves for a 8 merge with Raoult's Law line
if X8 --+ l. Similarly curves for aA merge with Raoult's Law if XA --+ l. In other words,
in the region where B obeys Henry's Law, A tends to obey Raoult's Law, and vice versa.
The thermodynamic basis for the above observation is now derived.
From the Gibbs- Duhem equation, viz. Eq. (9.23),
(9.44)
i.e.
· R1lXAd(ln aA) + X 8 d(ln a 8 )] =0 (9.45)
by combining Eq. (9.44) with Eq . (7.16).
.
'"
Thennodynamics of Solutions
If B obeys Henry's Law, then
d(ln a8 ) = d[ln( feXe)] = d[ln fe +In Xe] = d(ln Xe) (9.46)
since fe is a constant. Therefore, from Eqs. (9.45) and (9.46),
(9.47)
Again, XA + Xe = 1, and so
dXA + dXa = 0 (9.48)
(9.49)
Integrating Eq. (9.49), we obtain

In aA = In XA + constant, i.e. aA = IXA (9.50)
where I is a constant.
At XA = 1, aA = l. Hence, I= l, and aA =XA, i.e. A obeys Raoult's Law, when B obeys
Henry's law.
EXAMPLE 9.2 The following partial pressures of Zn have been determined in Cu-Zn
alloys at 1060°C:
Xz.u 1.0 0.45 0.3 0.2 0.15 0.1 0.05
Pz.u (mm Hg) 3040 970 456 180 90 45 22
(i) Does the system obey Raoult's Law?
(ii) In what ranges of composition, does Zn obey Henry's Law?
(iii) What is the free energy change when 1 gm. atom of pure Zn at 1060°C dissolves
in a large quantity of the alloy at X'ln = 0.3? ·
Solution
J4i = 3040 mm Hg; Yz.u = ~

Based on this, the following table is prepared from the data given in the problem:
Xzn 1.0 0.45 0.3 0.2 0.15 0.1 0.05
azn 1.0 0.3 19 0.15 0.0592 0.0296 0.0148 0.0073
Yzn l.O 0.709 0.50 0.296 0.197 0.148 0.145
(i) Since :>Zn is not 1 throughout the solution, Raoult' s Law is not obeyed.
(ii) In Henry's Law region, Yzn =constant. It is valid for XZn less than 0.1 to 0 .15 (XZn
< 0.12 as obtained by the graph plotting of Yz0 ).
(iii) When l g-atom of Zn dissolves in a large quantity of the alloy at 1060°C,
the free energy change = Gu, -
~ = GEt
(see Section 7.2.2 and Section 9.3).
Gz:i = RT In aZn = RT In (0 . 15) = 5025 J mo1- 1 at Xzn = 0.3.
9.5.2 Variation of .16"1, .1JITI and .15"1 with Composition for Ideal
Binary Solutions
For an ideal solution, Af-f" = 0 from Eq . (9.39). Further, if the solution is a binary A-B,
then from Eqs. (9.35) and (9.4 1),
L1Gm = RT(XA In XA + Xs In Xs) (9.51)
L1Sm = - R(XA In XA + Xs In Xa) (9.52)
Figure 9.2 shows variations of Af-f", L1S"1 and L1G"1 with composition at constant temperature
for an ideal binary solution. The curves of L1Gm, L1Sm, are symmetric with respect to A and
B. L1Sm is positive and is at maximum at XA = X8 = 0.5 . L1Gm is negative and is at minimum
at XA = X8 = 0.5.
A XA = X8 = 0.5 B
Xa-
Fig. 9.2 Variations of e.Hm, aS"' , aG"' with composition in an ideal binary solution at constant
temperature (schematic).
9.5.3 Relationship of GX, Gg1 with LIGTI in Binary Solutions

For a binary solution, Eq. (9.27) may be rewritten as
.1G111 =XAGr;: +X8 G8 (9.53)
Again, on the basis of Eq. (9.14),
d(,1Gm) = G:J:dXA +Gs dXs (9.54)
(9.55)
Multiplying both sides of Eq. (9.55) by X8 , then adding LiGm, and finally combining with
Eq. (9.53), we obtain
d(Li<?) - -
..1(1lll + Xs = Xs(Gr;: - 0:) + L1Gm
dXA
(9.56)
Similarly,
(9.57)
Figure 9.3 presents a schematic curve of LiGm vs. composition for a binary solution at
constant temperature. It also demonstrates the graphical procedure to find out {J; and Gg1
at a certain composition from this curve.
=
By definition, N?1 0 at XA = I and X8 = 1 (i.e. for pure A and B). The zero of LtG"1 is
at the top of the diagram since LiGm for a solution is negative. The point q corresponds to &11
at X8 = X8 , i.e. the composition for which Gr; and GB1 are to .be determined, sqt is tangent
to the curve at q. The graphical ·procedure gives values of Gr; and Ga
as intercepts of the
tangent at XA = l and X8 = l , respectively. The basis for the same is explained as follows:
os = or + rs =pq + rs
- Ar'm X d(L1Gm) - ;:;m [from Eq. (9.56)]
-LA.I + B -VA
dXA
vt = vy - ty = pq - (- y t) = pq + yt
[from Eq. (9 .57))

A B
Xa-+
Fig. 9.3 Graphical procedure for determination of (}:_, ~ from LIG"' vs. composition curve in
a binary solution.
9.6 Gibbs-Duhem Integration for Binary Solutions

Except for the vapour pressure technique with mass spectrometer, experimental arrangement
allows measurement of activity of one component only. Let us consider iron-silicon alloy.
Let the technique consist of equilibration of the alloy with a gas mixture of CO + C0 2 at
temperature T. It is found that only Si02 is present as the product. It is pure Si02 and exists
as a separate phase.
On the basis of the above observations, we consider only the following equation:
[Si] in Fe-Si alloy + 2C02 (g) = (Si02)purc solid + 2CO (g) (9.58)
.,..o = - RT In K = - RT In
LK.Jss 58
[ 1x P~o 2
]
(9.59)
[Gsi la11oy Pco, cq
since Osio = 1 .
'
Rearrangement of Eq. (9.59) yields
In [Gsi 1anoy = ~8
RT
+ 2 In ( Pco )
Pco
(9.60)
I CC)
"
Th6rmodynamics of Solutions ,,
The value of ~8 is available in thermodynamic data books. ( Pco,

Pco )
eq
is Pco ratio in gas
Pco,
at equilibrium with [Si] and Si0 2, and is measured experimentally. Thus, asi in the alloy
is determined with the help of Eq. (9.60). ll
The question is how do we find out aFe in the aJloy? The answer is that the values of
aFe in the alloy at various concentrations of Si are to be theoretically evaluated from
experimental values of asi as a function of composition of the alloy. The thermodynamic
basis for this is Gibbs-Duhem equation.
The G-D equation is a differential equation and is to be integrated. Such integration is
straightforward for binary solutions, but involves elaborate procedure for ternary solutions.
Since experimental activity versus composition data generally cannot be expressed by equations,
the technique involves graph plotting and graphical integration. Nowadays, computer-oriented
numerical methods are generally employed.
The method of integration based on Darken's a-function is generally employed. Before
discussing this method, the reason for its adoption rather than other simpler methods would
be discussed first.
9.6.1 Method 1
Equation (9.44) may be rewritten as the following equation using standard relation between
Gr and~:
(9.61)
A common logarithm has been chosen in the above equation for convenience of graph
plotting.
Integrating Eq. (9.61) for calculating aA from experimental values of a 8 , we get
(9.62)
The reason for this lower limit is because at XA = 1, aA = 1 and log aA = 0. Hence the
integration constant becomes zero.
Figure 9.4 is a graph (schematic) of X8/XA as function of log a 8. The figure is self-
explanatory. At the base line, XA = l, X8 = 0. Hence a 8 = 0, i.e. log a 8 ~ oo. This poses
difficulty for proper evaluation of the shaded area. The curve also tends to ~ asymptotic
to the y-axis towards the left-hand side, since XA ~ 0, X8 /XA ~ oo. For these reasons, this
:
method is not used.
log aA at XA = XA is
given by the shaded area
-log as__...
Jog as at XA = XA
Fig. 9.4 Variation of log a8 with Xef XA in a binary solution, and calculation of log aA by Gibbs-
Duhem integration (schematic}.
9.6.2 Method 2
Equation (9.48) may be rewritten as
(9.63)
Combining Eq. (9.61) with Eq. (9.63), we obtain
XAd(log aA - log XA) + Xsd(log as - Jog XA) = 0 (9.64)

i.e.
XA d log (;:) + Xsd( log i:) = XAd(log rAJ + Xsd(log )h) = 0 (9.65)
Integrating Eq. (9.65), we get
(9.66)
Xs/XA plot against log y8 is similar to Fig. 9.4 having the following features:
• At base line, XA = 1, Xs = 0. In view of Henry 's Law, log JB = log ~, and hence
has a finite value. Therefore, there is no problem at the lower limit.
• However, for XA ~ 0, X8 /XA ~ oo, and we shall have asymptotic behaviour as in
Fig. 9.4. Thus the shaded area will be somewhat erroneous for small values of XA-
9.6.3 Method 3 (Darken's Method)

Darken 's method eliminates the asymptotic behaviour of curve entirely, and so it allows a
reliable method of evaluation 'Of the shaded area. It is based on the a-function, defined as
a. = In r·
I (9.67)
I (1 - X;)2
For binary solution A-B,
a - In r· In YA
ae - In Ye (9.68)
- -xi-
I -
A - (1 - X;)2 - -X-~-,
A
If XA ~ 0, Xe ~ I, re~ 1, In re~ 0. It can ~ shown that as

would have finite
value for the above, using the L' Hospital theorem. If X8 ~ 0, XA ~ 1, re ~ Hence, 'fs.
a 8 is finite at X8 ~ 0. Similar would be the behaviour of aA-
Differentiating Eq. (9.68) by parts, we get
d(ln rA) = 2aAXadXe + X~ daA (9.69)
d(ln re) = 2aaXAdXA + Xi dae (9.70)
Xe
d(ln YA) = - - d(ln re)
XA
Xe 2
= - -XA (2aeXA dXA + XA daa)
= - (2X8 a 8 dXA + X8 XA da8 ) (9.71)
Integrating Eq. (9.71) between XA = XA to XA = 1, we get
(9.72)
The formula for integration by parts is
(9.73)
where x and y are two variables. On the basis of Eq. (9.73), we have
(9.74)
Substituting Eq. (9.74) into Eq. (9.72), and for integration limits XA = 1 to XA = XA,
we get
(9.75)
Figure 9.5 shows the variation of log Yai (i.e.

(1 - Xai)2 . 2~303)
with XPb for Cd-Pb binary at
500°C. The finite values of £XccJ at XPb = 0 and XPb = 1 are demonstrated. The integral in
Eq. (9.75) is the shaded area, and can be reliably found out in the entire composition range.
Thus, this method is generally employed for G-D integration.
1.2
1.0
,.......
"'
~ 0.8
-l
........
0.6
0.4
00
~
0.2
0
0 0.2 0.4 0.6 0.8 1.0
XPb
Fig. 9.5 Determination of lt>b from J'cd for the Cd-Pb system at S00°C by use of a-Function; the
shaded area represents contribution of the final term in Eq. (9.75).
Procedures are also available for G-D integration in ternary solutions, and even quartemaries.
However, they are beyond the scope of this text.
EXAMPLE 9.3 In liquid Fe-Ni solution at 1873 K, the activity of nickel as a function of
composition is known (noted in the solution that follows).
(i) Calculate the activity of iron at mole fraction of Fe of 0.6 by the Gibbs-Duhem
integration.
(ii) Calculate the following at XFe = 0.6: ~. C't ..dam, Gxs.
SolutWn (i) Given the values of aNi at various XNi as follows:
XNi = 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
a Ni = 0.067 0.137 0.208 0.287 0.376 0.492 0.620 0.776 0.89
We have the following calculations:
lNi = 0.67 0.685 0.69 0.72 0.75 0.82 0.986 0.96 0.99
(l-XNi)2= 0.81 0.64 0.49 0.36 0.25 0.16 0.09 0.04 0.01
llt.r; = -0.49 -0.59 -0.75 -0.92 -1.14 -1.24 -1.35 -1.07 -1.01
By method 3 of the Gibbs-Duhem integration (Eq. 9.75),
The graphical evaluation of the integral (as in Fig. 9.5) yields the value of integral as -0.08.
Hence,
In Jf;e = - (-
0.92 x 0.6 x 0.4 - 0.08) = 0.301, i.e.
Jf;e = 1.35, aFe = 0.6 X 1.35 = 0.81
(iii) GN\ =RT lo ~i = 8.314 x 1873 In (0.237) = -19438 J mo! at XFe = 0.6
Gfe =R:r In life= 8.314 x 1873 In (0.81) = -3281 J mo1-1
£m =(XNiGNl + XFe~) = 0.4 X (-19438) + 0.6 (-3281) = -9744 J mo1- 1
[as per Eq. (9.80)]

= RT(XNi In /N; + XFe In X:J [see Eq. (9.89]
= 8.314 x 1873 [0.4 In (0.72) + 0.6 In (1.35)] = 758 J mo1- 1•
9.7 Regular Solutions

Ideal solutions provide some equations relating activity to composition, .t1Hm, L1Sm, L1Gm
expressions. Most of the solutions, however, are 11onideal. There is a great need of expressing
thermodynamic quantities in these solutions in the form of analytical equations as function
of composition and temperature.
In this connection, the regular solution model, first proposed by Hildebrand, is of great
practical utility for various uses. Again, it is an approximation, but is much more consistent
with behaviours of a large number of solutions.
Hildebrand defined a regular solution as the one in which
s.m=s.
l I
01
• id ' but 11~
I
*o (9.76)
i.e.
(9.77)
where the superscipt 'id' refers to ideal solution. In other words, a regular solution has the
same entropy as an ideal solution of the same composition. But enthalpy is different from
ideal solution.
9.7.1 The Excess Function

For dealing with nonideal solutions, a function known as excess function (or molar excess
property) is also employed. It is only for extensive properties, and is defined as
(9.78)
where,
Q = some extensive molar property (G, H, S, ...) of the actual solution
Qid = value of Q for ideal solution at same composition and temperature
Qxs = value of Q in excess of that of the ideal solution.
Similarly, we can define partial molar excess properties as
(9.79)
Since the values of Q and Q; are the same (i.e. Q0 , (f) for standard states for both ideal
and nonideal solutions, we may also write
Qxs = (Q _ Q°) _ (Qid _ QO) = LlQm _ LlQm.id (9.80)
o;s= (Q - (f) - cQ;d - (f) =Qr ... Qm, id (9.81)
Therefore, on the basis of the above equations, for a regular solution,

S;xs = 0, sxs =0 (9.82)
Also,
(9.83)
since H;m , LlHm for ideal solutions are zero.
Hence, for a regular solution ,
e;~s
I
= 11xs
l
_ rs.xs = jj~s = H!l'
I I I
(9.84)
(9.85 )
Again,
axs = LiGm - LiGm,id = RT"LX; In a; - RT"LX; In X; = RT"LX; In 11 (9.86)
i j j
9.7.2 Regular Binary Solutions-Additional features
Hildebrand showed that a;, as defined in Eq. (9.67), is not a function of composition in a
binary regular solution. If, in Eq. (9.75) a8 is a constant, then
In YA = - XaXAaB - aa(XA - 1)
= - XaXAaa + aaXa
= aaXa(l - XA) = aa x~ (9.87)
Again, from Eq. (9.68), In YA = aA X~ . Hence, for a binary regular solution,

aA = a 8 = a = a constant (9.88)
Combining Eq. (9.88) with Eq. (9.85) and (9.86), we get
&Im = Gxs = RT(XA In YA + Xa In Jb)

= RTcef.,X~XA + X1Xa)
= RTaxAXa(Xa + XA)
= RTo.XAXa
= nxAxa (9.89) , .,
where fl is a parameter related to interaction energies in the solution. Again,
ii~ = a;: =RT In YA = RTaX~ = flX~ (9.90)

and,
H
-m
B -- G-XS
s -- RT ln Ya -- RTa XA2 nx 2
-- :.~. A (9.91)
As discussed earlier, '1H of reactions and processes mostly do not vary much with
temperature. Similar has been the observations on &m, ii;" of solutions, which have been
found to be approx.imately constant in a limited temperature range (of course, for a fixed
composition). Hence, from the above equations, fl may be assumed to be independent of
temperature. Therefore,
a= -
n cc
1
(9.92)
RT T
Equation (9.92) is often employed for estimating the effect of temperature on activity etc.
where experimental data are not available.
From Eq. (9.89), if Lilr is positive, then a is positive. Then, on the basis of Eqs. (9.90'
and (9.91), In YA and In 1h are positive, i.e. YA > 1, 1h > I, i.e. we have positive departures
from Raoult's Law. Similarly, negative &F1 means negative a and negative departures of
aA, a 8 from Raoult's Law. Again, an ideal solution may be looked upon as a regular solution
with a= 0 .
Figure 9.6 presents variations of LJJfD, L1Gm and LI.Sm for some hypothetical regular
binary solutions with a = + 1 and a = - 1. It also compares these with the ideal solution.
Of course, all these curves are at the same temperature.
L1.S"1 = LIS'11-id
- ve
A B
XA =Xe= 0.5
Xe--+
Fig. 9.6 LJHm, LJSm, LJGm as functions of composition at constant temperature for regular binary
solutions with different values of a (schematic).
EXAMPLE 9.4 Liquid lead-tin solution exhibits regular solution behaviour. At 746 K, it
is given that
In 1f>ti = - 0.737(1 - XPb) 2
(i) What is the corresponding expression for ~n at 746 K?
(ii) If 1 mole of solid Pb at 298 K is added to a large quantity of this liquid alloy of
composition XPb = 0.5, kept in a thermostat at 746 K, calculate:
(a) The quantity of heat required to be transferred from the thermostat to bring the
final temperature of alloy back to 746 K
(b) Entropy changes of system and surrounding
(c) Activity of Pb at 1000 K.
Solution (i) Since it is a regular solution, lXsn = aPb = - 0.737, and hence,
In ~n = - 0.737(1 - Xs 0 ) 2
We can also derive it by the Gibbs-Duhem integration as follows:
(E.9.3)
Xsn
=- (-0.737)XsaX'Pb -
f
1
(-0.737)dXsn
= 0.737[Xs XPb + 0 (Xso - I)]

= 0.737[X5 0 (1 - Xs0 ) - (1 - Xso)J
=- 0.737(1 - Xs0 ) 2
(ii) (a) Heat to be supplied from the thermostat (&/)
=sensible heat required to raise temperature of solid lead + heat of dissolution
=l~(l) (at 746) - ~ (s) (at 298) +ii~ J (E.9.4)
ii~ = RT In Jf'b = 8.314 x 746 x (-0.737) x 0.5 2 = - 1142 J, at XPb = 0.5

Since the melting point (TnJ of Pb = 600 K, we have
(E.9.5)
= f 600
298
[23.55 + 9.75 x 10-3 T] dT + 4212
746
+
f 600
[32.43 - 3.1 x 10-3 T] dT =17 ,673 J
Therefore,
Llli = 17,673 - 1142 = 16,531 J
(b) Similarly, L1S of the alloy as a result of addition is given by
L1S = l s>l'b(I). 746 - s>l'b(s), 298 J+ s~ (E.9.6)
~ =Spt; id = - R In XPb since it is a regular solution.

Hence,
~ = - 8.314 In (0.5) = 5.76 J K- 1
s2
l'b(ll. 746
- {'() -
"'Pb(s>. 298 -
f 600
298
CP, l'b(s) dT + L1Hm (Pb) +
T Tm (Pb)
f146
600
CP,l'b(I) dT
T (E.9.7)
= 34.78 J K- 1 (after solving)

Therefore,
(.1S)alloy = (LlS)syst = 34.78 + 5.76 = 40.54 J K- 1
(L\S) = _ (L11/)syst = _ 16,531 = _ 22.17 J K-1

surr 746 746
i.e. (.1S)syst + surr = 40.54 - 22.17 =+ 18.37 J K- 1
(c) For a regular binary solution, a oc l/T (9.92)
Again, at fixed composition, In Y; oc a. Hence,
ln(yl'b)1000 746
=
In <rl'bh46 1000
Since at XPb = 0.5, Ji'b = 0.83, we have Ji'b at 1000 K = 0.88 and al'b = 0.~.
9.8 Quasi-chemical Theory of Solution

Thermodynamic properties of solution are a consequence of interactions amongst atoms/
molecules. There have been several attempts to explain the properties from atomistic point
of view with varied degrees of success. However, none has been found to be universally
acceptable since all of them have some simplifying assumptions. Quasi-chemical theory is the
simplest and is able to provide explanations to some behaviour patterns of solutions approximately.
Therefore, this is being presented here as an example.
In a metallic solution, the components are assumed to be present as atoms and not
molecules due to the nature of metallic bonds. The interaction energy amongst atoms is the
sum total of several effects. In quasi-chemical theory, the assumptions are:
1. Atoms have chemical bonds amongst them.
2. The bonds are associated with some enthalpy, which alone contribute to enthalpy of
solution.
3. Only bonds amongst nearest neighbouring atoms are energetically significant.
In the binary A-B system, the bonds are A-A, B-B and A-B type. Assume that the
enthalpies associated with these bonds are HAA• H88 and HAB. These are enthalpy changes
to bring two atoms together from hypothetical gaseous state (i.e. a state where atoms are
free) . For example, HAA is enthalpy change associated with the following isothermal process:
atom A (g) + atom A (g) = A-A pair in solution
Let us now consider one gram-atom of the solution. Let it have PAA • P 88 and PAB
numbers of A-A, B-B and A-B bonds, respectively. Then
(9.93)
where 11 is the enthalpy of formation of the solution for one gram-atom (i.e. per mole) from
gaseous state at temperature T.
Let us assume that the solution contains NA atoms of A and N8 atoms of B. Each A-B bond
has one atom of A, whereas each A-A bond has two atoms of A. If Z is the coordinati on
number, then the number of A type atoms in A-B bonds is PAa!Z. On the other hand, the
number of A atoms in A-A bonds is 2PAA/Z. Hence,
N =PAS + 2PAA (9.94)

A Z Z
Similarly,
(9.95)
Thus, PAA and Pas can be expressed in terms of PAB• NA> Na and Z. Substituting these into
Eq. (9.93), we obtain
H =i ZNAHAA + i ZNsHaa + PAa [ HAs - i(HAA +Has)] (9.96)
Since absolute thermodynamic quantities cannot be measured, we are interested only in

enthalpy of mixing (L1/-F) of the solution, where
&F1 =H - (enthalpy of NA atoms in pure A + enthalpy of Na atoms in pure B) (9.97)
NA atoms in pure A contain only A-A bonds, whose numbers are 1/2ZNA since each bond
has two A atoms. Similarly, N8 atoms in pure B have 1/2ZN8 B-B bonds. Hence,
Lllim =H - 112ZNAHAA - 112ZNaHaa (9.98)
L1I-F = p AB [HAB - 1/2(H4-.A + Has)] (9.99)
9.8.1 Application to Ideal and Regular Solutions

Since .&Im =0 for an ideal solution, from Eq. (9.99),
HAa = 1/2(HAA + Hae) (9.100)

The physical meaning of this is that the energy of A-B type bond is the arithmetic average
of those of A-A and B-B bonds.
The atomistic interpretation of entropy of mixing of ideal solution (L1.Sm,id) shall be
provided in Chapter 12 on the basis of statistical thermodynamics. This will lead to the
conclusion that there is random mixing of atoms in an ideal solution. Since entropy of a
regular solution is the same as that of an ideal solution at the same composition and
temperature, we may assume random mixing of atoms in a regular solution as well.
If the total number of atoms (i.e. Avogadro's number) in a mole of solution is N0 , then
x =NA - Na
Xs- (9.101)
A N ' No
0
The probability of an A atom to occupy one atomic site is XA. The probability of a B atom
to occupy a specific site is X8 . For random mixing, the joint probability of an A atom and
a B atom to occupy neighbouring sites forming A-B bond would, therefore, be 2XAX8 . The
factor 2 occurs because A and B atoms can interchange sites mutuaJly also. Since the total
number of bonds is 1/2ZN0 , we have
1
PAB = 2XAXa 2ZN0 = XAX8 ZN0 (9.102)
Substitution of this in Eq. (9.99) yields
1
Lilf" = XAXaZNo [HAB - 2<HAA + Haa)J (9.103)
Comparison of Eq. (9.103) with Eq. (9.89) gives
(9.104)
Equation (9.104) thus provides interpretation of n parameter of a regular binary solution

in terms of atomic parameters.
9.8.2 Non-regular Solutions

In a solid solution, there may be significant strain energy in the lattice due to differences
in atomic sizes. In a liquid, strain energy is insignificant and, therefore, quasi-chemical
theory is more applicable. For a non-regular solution, Guggenhein and others derived the
following equation:
(9.105)
Expanding the exponential and ignoring higher order terms, we get
(9.106)
sxs "" - xi x~d l(ZRT2 ) (9.107)

axs "" XAXa£ll - XAXai11(ZR1)J (9.108)
9.9 Summary
1. This chapter deals with thermodynamics of condensed solutions, i.e. solid solutions
or liquid solutions. Temperature is assumed constant, unless otherwise stated.
2. A solution, which obeys Raoult's Law, i.e. a; = X;, is known as an ideal solution.
3. Activity coefficient is defined as: y1 = !!!.__ • '}'; = 1 for ideal solutions, but °* 1 for
nonideal solutions. x,
4. All properties of solutions are functions of composition at a constant temperature.
5. All standard thermodynamic relations derived earlier for pure substances are applicable
to properties of solution. For example, in analogy with equation : dG = - S dT +
V dP, we may write
dG; = - s; dT + Yi dP, d(;;' = - s;m dT + Y;m dP
d(L1Gm) = - ..1Sm dT + AV"' dP

6. The alternative forms of the Gibbs- Duhem (G-D) equation are:
~ n1 dQ=0,
I
L.X1 dQ
I
=0, r.i n1 dQt =0, L.X,dQr
I
=0
where Q is an extensive property of state (e.g. H, G, V, S).
7. L1Qm = L.X-nm, such as ..1Gm = L.X G~

I
• I~
I
• I I
8. (i) 8(1n r,) = - ii,m at fixed pressure and composition.

8T T2
(ii) ..1Gm = RT r, X 1 In a; at fixed temperature and composition.

I
9. In an ideal solution, a;= X;; hence, G~

I
=RT In X1 and ..1Gm =RT L.X;
j
In X;. Also,
..11-f" = 0, L1V"' = 0. Therefore, ..1Sm =- R L.X; In X 1•

i
10. In a binary solution A-B,
(i) if B is a solute, then at X8 ~ 0. )B ~ yg (a constant). This is known as Henry' s
Law;
(ii) if the solute obeys Henry's Law, then the solvent obeys Raoult's Law;
(iv) Integration of the Gibbs-Duhem equation allows evaluation of aA(or YA) from
experimental values of a 8 (or rs) as a function of composition at constant
temperature. Darken's method is generally employed. It uses a-function, where
11. In a regular solution, entropy of mixing is equal to that of an ideal solution, i.e.
-s-m.id
S-m
; - ,. '
AC"IJI __ AC"IJl.id
LJt.) LlJ
However,
ii;n "# 0, mm "# 0
12. Excess property of a solution is defined as
Following this, for a regular solution,
5XS = 0, Gxs = MF= RT :EX; In

j
r;
13. For a regular binary solution A-B,

(i) aA = a 8 = a = a constant, independent of composition,
(ii) Gxs = MF = RTaXAXa = !JXAXa.
14. From the quasi-chemical theory of solution,
D = ZNo [HAa - l/2(HAA + Hae)]

for binary regular solution, where Z is coordination number, No is Avogadro's
number, and HAB, HAA, Hee are enthalpies of A-B, A-A, B-B bonds respectively.
PROBLEMS
9.1 Show that in a dilute binary solution A-B at a given temperature, ii;: and Yem are
independent of composition. Here, the subscript B refers to the solute component.
9.2 One mole of solid Cr at 1873 K is added to a large quantity of liquid Fe-Cr solution
at 1873 K with Xer = 0.2. If Fe and Cr form ideal solution, calculate the heat and
entropy changes in the system resulting from the addition. Ignore the difference of
Cp between solid and liquid chromium.
9.3 In a liquid Cu-Sn alloy at 1400 K and at X50 = 0.4, the values of activity of Sn
and Cu are 0 .333 and 0.362, respectively. Calculate PZn• Peu; Ysn• reu; G;!. ~.
L1Gm, and Gxs.
9.4 Evaluate the thermodynamic feasibility of removal of aluminium from liquid silver
under vacuum at 1000°C. The maximum residual Al in the solution should be
1 mole percent. Henry's Law constant for Al( /1i) = 0.11.
9.5 The following data are given for liquid Cu-Sn alloys at 800°C:
Xsn = 0.1 0.2 0.3 0.4 0.5 0 .6 0.7 0.8 0.9

asn = 0.007 0.055 0.213 0.353 0.474 0.574 0.683 0.805 0 .909
acu = 0.851 0.602 0.388 0.34 0.267 0.2 1 0.15., 0.098 0.047
Note: aSn values were experimental; aeu values calculated by Gibbs-Duhem integration with hypothetical
liquid Cu at 800°C as standard state.
(i) Do Sn and Cu obey Raoult's Law?
(ii) Find out the composition ranges for the validity of Henry's Law for Sn and Cu.
(iii) Calculate the following at Xsn = 0.6: PSn• Pcu; iJSn, L1Gm, Gxs. Gc:i.
(iv) By the Gibbs-Duhem integration, verify the value of acu at Xsn = 0.6.
(v) Is it a regular solution?
9.6 The following data are provided for liquid Cu-Ag alloys at 1300 K:
Xcu = 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

rcu = 2.54 2.09 1.74 1.50 1.36 1.22 1.14 1.04
YAg = 1.01 1.05 1.11 1.20 1.32 1.50 1.75 2.16
Note: Like Problem 5. YA& calculated by GD integration.
The following additional data are also provided at Xeu = 0.5:

(i) Ycu = 1.4, YAg = 1.37, at 1200 K.
(ii) Ycu = 1.32, YAg = 1.29, at 1400 K.
Calculate: (i) Yeu by GD integration at any of the above values of Xeu (as per your
choice) at 1300 K, and verify against data.
(ii) the foJlowing at Xcu = 0.5 and T = 1200 K, 1300 K , 1400 K:
G~U' G~g. L1Gm, cXS; a.a.. aAg;
( 1·1·1·) -m -m -xs - xs A11m HXS X _ 0 .

Heu• HAg• Heu• HAg • LJIJ • at Cu - .5 ,
-m s-m sNCS sNCS
(iv) S L1Sm sXS at Xeu = 0.5·
Cu • Ag• Cu • Ag ' ' '
(vi) Compare the characteristics of this solution with various characteristics of a

regular solution.
9.7 In liquid Cu-Zn solution ,
tXz,, =- - -,
19,250 w here R 1
. s .in J mo 1-t K-'
RT
Textbook of Materials and Metallurgical The""odynamics
(i) by GD integration, show that CXcu = az0;
(ii) calculate cxs, ..1Gm at 1400 K and Xeu = 0.6.

9.8 Assume that, in a binary solution A-B, the activity coefficients can be expressed as
In YA = a 1X8 + 112 aix~ + 113 Cl:JX~ + ...
In re = /31XA + 112 Pi X~ + 1/3 AXl + .. .

in the entire composition. Prove:
(i) a 1 = /3 1 = 0
(ii) if only the quadratic terms are retained, ai = fh.
9.9 In the zinc-cadmium solution at 700 K,
In y01 = 0.967(1 - X01) 2 + 0.691 (I - Xcd) 3
(i) Express In l'zn as a function of mole fraction.

(ii) Find out Henry's Law constant for zinc.
(iii) If In rai = 0.967(1 - Xru)2, calculate H0J at Xai = 0.3.
Chemical Potential and Equilibria
amongst Phases of Variable
Compositions
10.1 Chemical Potential
10.1.1 Definition
In Section 1.3, we discussed about thermodynamic equilibrium. It was mentioned there that
for physico-chemical processes and chemically reactive systems, thermodynamic equilibrium
requires attainment of physico-chemical equilibrium (also referred to as chemical equilibrium)
besides mechanical and thermal equilibria. This means that chemical potential should be
uniform in the entire system in addition to uniformity of pressure and temperature.
J.W. Gibbs first proposed the term 'chemical potential' in analogy with other potentials
(e.g. thermal potential and gravitational potential). The meaning of the term will become
clear as we go along.
In Chapter 7, a distinction was made between a molar quantity (Q) and the same for
system as a whole (Q') vide Eq. (7.5), for convenience of derivation. By adopting this
convention, let us consider a system with variable composition. Then Eqs. (5.11)-(5.14) may
be modified as follows:
dU' = T dS' - P dV + -
8n1(oU') 5• v•
• ·"2' ···
dn1 + - (8U')
0"2 S'. v·.ni . ...
except "2
(oU')
d"2 + ... + -
On; S', V'.n1• ...
except n;
dn, + ...
except n 1
= T dS' - (oU')s·.v-.,,. ....

P dV' + L -
; 01\ cxccptllj
dn; (10.1)
159
Similarly,
dH' = T dS' + V dP + :t(8H')

i 8n, S', P,"t , ... except "I
dn1 (10.2)
dA' = - S' dT - P dV' + ~( 8811;A'l

I ,
d"'i (10.3)
,V •"t• ... except "I
dG'= - S' dT + V dP + '8G'l

I 811; ,P."t , ... except "I
dn, (10.4)
As defined in section 7.2.l, n 1, n2, • • ., n;, . .. are numbers of moles of components 1, 2, .. .,

i .. . in the solution. The last term in Eqs. (10.1)-(10.4) is related to variation of U', H', A'
and G' with change of nl> n 2 etc. As usual, i is the general symbol of a component.
Gibbs' original definition of chemical potential of a component i(µ;) was
/.4 = -
(8U'
8n;
J S', v'.ni . nz ... except n;
(1 0.5)
Therefore,
dU'=TdS'-PdV+ :tµ;dn; (10.6)
i
Again,
dH' = dU' + P dV + Y dP (2.25)
dH' = T dS' + V dP + 1: µ; dn; (1 0 .7)

j
Equating relations (10.2) and (10.7), we get
/.4 = -
[8H'
8n,
)
S',P."t, "2• ... except n;
(10.8)
Similarly, by combining Eqs. (10.3) and (10.4) with Eq. (10.6), the following relations can
be obtained:
/.4 (8A'l
= -8n.
1
,
,V ,n1, "2 ···· exceptn;
(10.9)
11; =(-8G'l
on, ( 10.10)
,P. n1, "2, ... except n;
In accordance with the above equations, therefore, there are four different definitions of
chemical potemial. In chemical thermodynamics, we employ Gibbs free energy criterion for
equilibrium. Hence, in this text as well as in other texts dealing with metallurgical thermodynamics,
the definition of chemical potential, as given in Eq. (10.10), will be employed. This way,
µ; becomes the same as the partial molar free energy of component i (i.e. q) in a solution,
as defined and discussed in section 7.2.2.
10.1.2 Equal ity of Chemical Potentials amongst Phases at

Equilibrium at Constant Temperature and Pressure
At constant T and P, Eq. (10.4) may be rewritten as
(10.11)
Consider 2 phases (I and II) in the system. Then,
dG ,,l _
-µI
I dn'I + ... +J.liI d11;I + ... (10.1 2)
II d
. n,II + ... +J.liII d n;11 + ...
dG ,,ll -µ,
_
(10. 13)
Consider exchange of dn 1 moles of only component 1 between phases. Then,
(10.14)
Since,
(10.15)
Therefore,
dG' = dG'· 1 + dG'· II = dn1 (µ~ - µf) (10.16)
where dG' is the total free energy change in the system as a result of exchange of dn 1
between phases I and II. For equilibrium at constant T and P,
dG' = 0, and hence from Eq. (10.16),
µf =Jlill (10.17)
Equation (10.17) can be generalizeo for all components with similar arguments. Hence,
at constant T and P, if phases I and II are at equilibrium, then
µ,1_11
- µj ' J..121-_J..12
u 1_n
' ... , µ; - µ; ' ... (10.18)
Similarly, Eq. (10.17) can be generalized for any number of phases at equilibrium as
µ II -_ µIII -_ µIll _ _ µP
I - ... - I
µJ =µJ• =µJll =... =14.
(10.19)
where P is the total number of phases in the system.

Equation (10.19) is a statement of equality of chemical potentials of all components
amongst all phases at equilibrium at the same pressure and temperature.
10.1.3 Some Additional Remarks on Chemical Potential

Elucidation of Eq. {10.19)
Equation (10.19) is the most general one. For application purposes, it does not matter whether
a phase is a solution or a pure substance. However, it is necessary to clarify some confusions.
Let us consider equilibrium between liquid copper and oxygen gas, and find out the chemical
potential of 0 2( J1<>i,) in liquid copper. The confusion may be due to the following reason.
Oxygen dissolves in metals in the atomic state. How can we imagine Jl<>i in liquid copper then?
The answer lies in the fact that, according to equations of chemical equilibria, the
concentration of molecular 0 2 in the melt may be extremely small, but it can never be zero.
This follows from the equilibrium
[£1oi]
K - -- (10.20)
20 - [~]2
Here, [ ) denotes components in metallic solution. K20 has a nonzero value. If it is zero, then
..1~ = - RT In K20 would be infinity, which is not possible. Thus, [ lloil can not be zero.
To take another example, suppose a piece of pure Si0 2 is brought to equilibrium with
a molten Cu-Ag alloy. How can we think of µAg in Si0 2 and µSi in the alloy? The
explanation can be given with the help of the following reactions:
Si02(s) + 4[Ag) = [Si) + 2(Ag 20) (10.21)
(Ag20) = 2(Ag) + (0) (10.22)
Si02(s) + 4[Cu] = [Si] + 2(Cu20) (10.23)
In the above equations, ()denotes oxide phase. Some Si would dissolve in the alloy, through
reactions (10.21) and (10.23). Its concentration would be very small. Similarly, through
reactions (10.2 1) and (10.22), a very very small concentration of Ag in oxide phase can be
predicted from thermodynamic equilibria considerations as in the case of 0 2 in Eq. (10.20).
µ 1 criteria for phases not at equilibrium

If a process is to occur, then (dG')r, p < 0. Combining this with Eq. (10.16), the process
would take place if
Chemical Potential and Equilibria amongst Phases of Variable Compositions
(10.24)
For Eq. (10.16), the proqess under consideration is transfer of component 1 from phase II to
phase I. [Note: dn 1 has been taken as positive. Hence dni is positive, i.e. phase I receives
dn 1 and phase II loses it.]
Equation (10.24) can be generalized to state that a component would tend to be transferred
from a higher to a lower chemical potential. This makes µ; analogous to thermal potential
etc. For example, heat flows from higher thermal potential (i.e. higher temperature) to lower
thermal potential.
Application of chemical potential

OveraJI chemical equilibrium calculations are carried out through equilibrium constant as
illustrated in Chapters 7 and 8. However, there are areas of application where chemical
potential approach offers advantages. These are as follows:
1. Studies of phase equilibria will be illustrated in this chapter itself.

2. In multicomponent systems, often we are interested in one component only. Hence,
an easy approach is to consider chemical potential of that component alone.
3. Equation (10.24) allows easy visualization as to which way a species (element or
compound) would tend to get transferred.
Chemical potential and activity

Equation (7.16) may be rewritten in terms of chemical potential as
Jl; - Jl;o = RT In a; (10.25)
where, Jf = µ; at standard state of i.

By universal convention, Jf is taken as zero for all phases in the system. It also means
that we adopt the same standard state in all phases. This simplifies Eq. (10.25) into
µ;= RT In a; (10.26)
For example, for Eq. (10.20), chemical potential of 0 2 (also known as oxygen potential)
in both the gas and liquid copper co-existing at equilibrium may be written as
J1<>z =RT In °Oi =RT In P0i_ (10.27)

where P0i = partial pressure 0 2 in gas et equilibrium.
From Eq. (10.26), we may also conclude that at equilibrium (provided standard states
are same in all phases, (say) pure components 1, 2, ... , i, ... ),
a: =a/I =... =ai (10.28)
aJ =aJI =... =a{'
Also, species i will be transferred from phase I to phase II, if aJ > llin .
10.2 Gibbs Phase Rule and Applications
10.2.1 Derivation of Gibbs Phase Rule
Consider a system at equilibrium with the following specifications:

l. Constant temperature and pressure
2. Components are 1, 2, .. . , i, j , ...
3. No. of components = C
4. All components are independent, i.e. they do not enter into chemical reactions with
one another
5. No. of phases = P.
The thermodynamic state of each phase would be determined by temperature, pressure

and composition variables (mole fraction, weight % or anything else). The number of
composition variables to be fixed independently is (C - 1).
Therefore, the thermodynamic state of a single phase would be determined by (C - l + 2),
i.e. C + 1 variables. The number "2" stands for T and P, which constitute two indepen-dent
variables. For the entire system of P phases, the total number of variables would be P(C + 1).
All the above variables, however, would not be independent since the requirements for
thermodynamic equilibrium impose the following restrictions (i.e. co11s1rai11ts):
yI = yll = .. . = yP for thermal equilibrium

pl = pll = ... = pP for pressure equilibrium
,) - ,)I - - ,,P
?I -1-'1 - ... -?I
;{ = µf = ... = J.( (C-equations for chemical equilibrium) ( 10.29)
where I, II, ... are phases. From Eq. (10.29),
Total No. of constraints = (P - 1) (C + 2) (10.30)
since there are (P - 1) equalities in each line of Eq. (10.29). Hence,
Total No. of independent variables = F = P(C + 1) - (P - 1) (C + 2)

=C-P+2 (10.3 1)
Equation (10.31) is the famous Gibbs phase rule.

10.2.2 Application of Phase Rule

Non-reacting components
EXAMPLE 10.1 Solid-Liquid-Vapour Equilibria (Section 6.4). If all three-phases co-

exist, then P = 3. For a pure substance, C = 1. From Eq. (10.31), therefore, F = 1 - 3 +
2 = 0 . This means an Invariant equilibrium with no degree of freedom at all , i.e. we cannot
vary Tor P independently. This explains the triple point phenomena (see Fig. 6.4).
EXAMPLE 10.2 Alloy Phase Diagrams. Here, the system is non-reactive, the environment
of the alloy being either vacuum or an inert gas (Ar, He etc.). Again, pressure has hardly any
influence. So, Eq. (10.31) gets simplified into
F=C - P+l (E.10.1)
For a binary system (Cu-Ag, Fe-Ni , etc .), C 2. Hence, F =

3 - P. If only liquid or =
= =
solid solution is present, then P 1 and F 2. This means both temperature and composition
can be varied independently. If there ·is a mixture of solid solution + liquid solution, then
= =
P 2 and F 1. This means that if temperature is fixed, compositions of solid and liquid
are automatically fixed, and vice versa.
Figure 10. l shows the binary phase diagram of Pb-Sn system. At 200°C, compositions
of liquid as well as a and p solid solutions can be varied over some ranges. Bui composition
°C Tin (atomic%) op
10 20 30 40 50 60 70 80 90
350 I I -
' ' I I I
300
~ - 600
'\· ~ -
L
I\ ~ - 500
250 ............. 232°
200 -a\, a+L

"
lx~0
i'...
.....
-
61.9
.......-
L---
B+I
~I'
97.5.,
400
/ 1 9.2
150 300
100
I a+P
-
- 200
50
I -
- 100
j
0 -
Pb 10 20 30 40 50 60 70 80 90 Sn
Tin (wt.%)
Fig. 10.1 Lead-tin binary phase diagram.
of the liquid solution on liquidus and that of the solid solution on the solidus are fixed since
these correspond to equilibrium co-existence of two-phases.
At the eutectic point, there are thr~e-phases co-existing at equilibrium. Therefore,
F = 0 here, and is an invariant.
EXAMPLE 10.3 Nonreacting Oxide Phase Diagrams. A good example of non-reacting

oxide phase diagram is the Si0i-A1 20 3 binary system which constitutes the principal phase
diagram for oxk e refractory systems (Fig. 10.2). Like alloy systems, here also the gas phase
and pressure are neglected (i.e. F = C - P + l).
Again, both temperature and composition of liquid phase can be varied, but at a fixed
temperature, liquidus composition is fixed. Solid phases are pure. Point P is an invariant
known as peritectic. Here, three phases co-exist at equilibrium (liquid, corundum, mullite).
Reading components
If components are reacting amongst themselves, then the reaction equilibria constitute additional
constraints. However, all possible reactions are not independent. Only the number of independent
reactions are to be taken into account. In such a situation,
C=N-R (10.32)
where N is the number of species (elements and compounds) and R is the number of
independent reactions.
EXAMPLE 10.4 Consider a system of pure metal (M), oxygen and pure metal oxide.
There is only one possible ·reaction, viz.
l
M(s) + -02(g)
2
= MO(s) (E.10.2)
N =3 (M, 0 2, MO), R = 1, and P = 3. The gas phase is involved. So both T and P are
variables. Hence, F = (3 - 1) - 3 + 2 = l. If T is fixed ,. then P0i. at equilibrium is fixed.
M and MO are pure anyway. Therefore, the state of the system gets fixed completely by
fixing T. Figure 8.1 on the Cu-Cu 20-02 system illust~tes this.
Suppose, in the above system, M is present as a liquid solution M-X, where X is a noble
metal. Also, MO is present as a separate oxide solution, MO-BO, where BO is more stable
than MO. Then, in this case too, the only significant reaction is oxidation of M to MO. X
would not form oxide significantly, and BO would not be reduced to B significantly. Here,
N = 5 (M, X, 0 2 , MO, BO), R = 1, P = 3 (liquid metallic solution, gas, liquid oxide
solution). Hence, F = 3. This means that even after fixing temperature and pressure, compos.itions
of phases cannot be fixed.
Now, in addition to fixing T, let us fix compositions of metallic solution and oxide
solution independently. Then, all three degrees of freedom are utilized, and P0i of the gas
phase also gets fixed at equilibrium. This may also be explained from the use of equilibrium
constants as follows. For Eq. (E.10.2),
Chemical Potential and Equilibria amongst Phases of V~riable Compositions
Si02 (wt. %)
100 90 80 70 60 50 40 30 20 10 0
MELTING POINT OF SILICA (CRISTOBALITE), l 724uc , 3135-r
"''" .Tll'I r. POINT OF ALUMINA !CORUNDUM), 2050"C, 3722°F
3800
3600 ·-
Cl
3
0
"] ~
...--- 2050
- 2000
~
+
LIQ um VCORUNDUM
....
-;
3400 ,___ <
/I LIO+u111 - 1900
0
"«)
~
i2
a:l
I -
/ p ~' 17°F
-
'O
...
c<S
t:
... 3200 - ~
0 __... ~I I I E-5.i' 1800 .~
~
u
Ix/ I ..J c0
3Al2~ 2Si°'2
c<S ..J
I.I.. 3135
- 1700 u
y
::::>
Vl
0 / • (MUU.ITE)+ ,!QUID 6 Vl
0
0
@i
3000 ·v 2903°F I
Q"'
CORUNl>UM ...
0
t)I)
0 ' en +
JA~L· 2Si02
- 1600 0
5 .5%A~ca
0
2800
- 94.5% i ,, I I N
L TE) 0
CRISTOBAUrE + 3Al2 0 3 • 2Si0:2 (MUU.ITE)
•2578°Fl I I I I~ <(
- 1500
I
2600
i "'
,YMITE i 3Al, O,, . r 0:2 (MrITE) 71.8 %Al 10,
/8.2% S102
- 1400
2400
0 10 20 30 40 50 60 70 80 90 100
Al20 3 (wt. %)
Fig. 10.2 Silica-alumina binary phase diagram.
(E.10.3)
K is fixed by temperature, aMo and aM are fixed by temperature and composition. Thus, from
Eq. (E.10.3), (PO]. )eq is automatically fixed.
EXAMPLE 10.5 Dissociation of limestone

CaC03 (s) = CaO (s) + C02 (g) (E.10.4)
N =3 (CaC03, CaO, C0 2), R =1 (i.e. Eq . (E.10.4), P =3 (CaC03, CaO, gas).
Hence, F = (3 - 1) - 3 + 2 = l. If temperature is fixed, then ( Pc0i )eq is automatically fixed ,

thus fixing the state of the entire system.
Figures. 10.l and 10.2 have presented phase diagrams of Pb-Sn and Si0i-Al 20 3 binary
systems, respectively. These are examples of condensed phase systems (i.e. solids and liquids),
which are nonreactive. The influence of pressure is also negligible (unless the pressure is
very high, of the order of thousands of atmospheres). In contrast, construction of phase

stability diagrams for reactive systems has to consider the gas phase if there are gaseous
reactant (s) and/or product (s). In Section 8.3.2, we have already provided some discussions
on this. The thermodynamic basis for finding out P0i_, PcdPc0i. and PH/PH2o ratios at
equilibrium with metal and metal oxide also have been explained. Further examples will be
presented in Chapter 11 .
EXAMPLE 10.~ (i) Calculate the chemical potential of carbon of a gas mixture of CO and
C0 2 at 1000 K and 2 atm. total pressure. Assume PcdPc0i. = 103 .
(ii) Suppose this gas mixture is brought in contact with zirconium metal, would ZrC form?
Solution (i) For the reaction:
2CO (g) = C02 (g) + C (s) (E.10.5)
..1ds can be obtained by combining the following reactions:
I
C(s) + 202(g) = CO(g); ..1~ = -199.4 x 103 J mo1- 1 (E.10.6)
C(s) + 02(g) = C02(g); L1cfi = -395.1 x 103 J mo1- 1 (E.10.7)
The values of ..1~, ..1cfi have been obtained from the data table in the Appendix. Hence,
..161 = ..1cfi - 2..1~ = + 3.7 x 103 J mo1- 1 =- RT In K5
=- RT In (
Pc0i. x
2
ac) (E.10.8)
Pco cq
2 1
Noting that PcdPc0i. = 103, Pco = 2 atm, Pc0i. =
1000
= 500 atm., and substituting these
values in Eq. (E.10.8), we get
ac = activity of carbon at equilibrium of reaction (E.10.5) = 7.78 x 10-4
Ile = RT In ac = -59.52 kJ mo1- 1
[Note : Actually, RT In ac =
fie - µg.
However, by convention, is zero [see µg
Eq. (10.26)). Also, as elucidated in section 10.1.3, although carbon is not present in the
gas, there will be a finite value of Ile in the gas.)
(ii) For, Zr(s) + C(s) = ZrC(s); ..1cft = - 41.5 kJ mo1- 1 at 1000 K (E.10.9)
Again, ..1G'f (ZrC) = - RT In ( llz,c ) (E.10.10)

l1zr X Cle cq
Combining the above equations, and noting that the activities of ZrC and Zr are 1 since they
are pure, we get
1
RT ln(ac)eq = (µc)at eq with zr+ZrC = - 41.5 kJ mol-
Since
{µc)c0-co 2 gas < {µc)zr+ZrC• ZrC will not form.
10.3 Thermodynamic Basis for Binary Alloy Phase

Diagrams
Figure 10. l shows the binary Pb-Sn phase diagram. In section 10.2.2, we have made some
preliminary observations on these diagrams. Phase diagrams provide the basis for understanding
the behaviour of alloys, as well as ceramic materials. Hence, they are very important in
metallurgy and materials science, and these are typically taught in detail in separate courses.
Therefore, the thermodynamic basis would be presented briefly for binary systems only,
along with some examples. Ternary diagrams etc. will be omitted.
The terminology 'alloy' is traditionally reserved for metallic systems. However, all
discussions are applicable to nonmetallic condensed systems as well. Therefore, for the sake
of clarity the word 'alloy ' is used in the heading of this section. Incidentally, nowadays, it
is being used for nonmetallic materials also, e.g. ceramic alloys.
10.3.1 Free Energy vs. Composition Diagrams-General discussions
As elucidated in Section 5.7, any system tends to lower its Gibbs free energy at constant T, P.
The equilibrium state corresponds to minimum (G)r.P· Therefore, the nature of a phase
diagram would depend on variation G with composition at various temperatures. As discussed
several times earlier, the constant pressure restriction is not important.
Section 9.5 has dealt with thermodynamics of binary solutions. Figure 9.2 has schematically
shown variation of L1Gm, Mf11, L1Sm with mole fraction for an ideal binary solution at
constant temperature. Figure 9.6 has schematically shown variation of L1Gm for regular
binary solution for both positive and negative values of a.
For a regular binary solution A-B, where A and B are general symbols of the two
components. From Eqs. (9.86) and (9.89), we have
(10.33)
i.e.
( 10.34)
From Eq. (10.34), L1Gm/RT as function of mole fraction can be calculated for different
assumed values of a. This shows that, if a > 2.0, the curve exhibits the maxima-minima
behaviour. This is shown in Fig. 10.3(a). The curve is symmetric with respect to X8 =
0.5 .
XA = X8 = 0.5
+ve
.-----------*----------.. .
a b
stable I
-ve (a)
ly
0 --.--
I
I Xa
I
I
I
(b)
(c)
0
A B
Xa - -~
Fig. 10.3 Variation of: (a) free energy, (b) chemical potentials of A and B, and (c) activities of
A and B, at constant temperature, for binary system A-B exhibiting maxima-minima
type behaviour in .i:IG"' (schematic).
Chemical Potential and Equilibria amongst Phases of Variable Compositions II.ii
Positive a means positive departure from Raoult' s Law (see section 9.7.2). An arbitrary
solution, which exhibits such positive departure but is not a regular solution, would not be
symmetric with respect to X8 = 0.5. Figure 10.3(a) also shows such a behaviour schematically.
Let us now construct the common tangent amnb to the curve mpn. Let us consider an
arbitrary composition, X8 . The point p is L1Gm for the solution. Point q is L1Gm for a mixture
of two phases, whose compositions are fixed at m and n, respectively. Since q is lower than
p , the two-phase mixture a + f3 would be stabler than the solution in this region.
The intercepts of the tangent at pure A and pure B provide the values of G:,:: and GB''
respectively (see Section 9.5.3). Therefore,
=
G;: for a-phase at composition corresponding to m G;:: for ,8-phase at composition
corresponding to 11, i.e.
(G: )a at = (G:: )pat "
Ill (10.35)
Similarly,
'
(10.36)
As discussed in Section 10.1.3, µ~ =0 =~. Hence,
(];:=µA , GB' =µa (10.37)

(µA)a at 111 = (µA){J at n (10.38)

'
' (µa)a at m = (µa)p ai 11 (10.39)
This is in conformity with the principle of equality of chemical potentials for phases at
equilibrium (Section 10.1.2). Hence, it is confirmed that the two-phase mixture would have
fixed compositions of a and {3. It also follows that only a or f3 single phase solutions would
be stable at the two-terminal regions since the L1Gm curves are lower than straight lines
joining any two points on the curve in the regions 'am' and ' nb' .
From Eqs. (10.38) and (10.39), µA and µ 8 should be constant in two-phase mixture.
Similarly, on the basis of Eq. (10.28), aA and a 8 also should be constant in the two-phase
field a + {3. These are shown in Figures '10.3(b) and 10.3(c). It may be noted that µA = 0
at XA = 1, µ 8 = 0 at X8 = 1, as .per convention already discussed in section 10.1.3.
10.3.2 Phase Diagrams with Miscibility Gap
The free energy vs. c.omposition diagram of Fig. 10.3 schematically illustrates the thermodynamic
basis regarding stabilities of the two-terminal solutions a and f3 and formation of a+ f3 two-
'
phase mixtures. There are phase diagrams which exhibit what is known as 'miscibility gap'.
i This is schematically illustrated in Fig. 10.4(a). Here, above the critical temperature Tw
there is only one phase. But below Tw the single phase is transformed into two-phase
mixture, which is the miscibility gap.
Assuming the solutions to be regular, L1Gm is given by Eq. (10.33). Differentiating .1cm
w.r.t. X8 , we obtain (noting XA = 1 - X 8 , and dXA = - dX 8 )
Liquid
solution
(a)
~(s)
+ve
i
!
-ve
(b)
jAt T0 j1
I
I
(c)
I (At T0 I
I
I
i (d)
A B
Xa-----+
Fig. 10.5 (a) Phase diagram, (b) free en1rgy-composition diagram at T = T0 , and (c), (d) activity-
composition diagrams of B and A, respectively at T = T0 , for system with complete
miscibility in liquid and solid phase (schematic).
Positive a means positive departure from Raoult's Law (see section 9.7.2). An arbitrary
solution, which exhibits such positive departure but is not a regular solution, would not be
=
symmetric with respect to Xa 0.5. Figure 10.3(a) also shows suc h a behaviour schematically.
Let us now construct the common tangent amnb to the curve mpn. Let us consider an
arbitrary composition, X 6 • The point p is L1Gm for the solution . Point q is L1Gm for a mixture
of two phases, whose compositions are fixed at m and n, respectively. Since q is lower than
p, the two-phase mixture a + f3 would be stabler than the solution in this region.
The intercepts of the tangent at pure A and pure B provide the values of G~ and G:;1,
respectively (see Section 9.5.3). Therefore,
a~ for a-phase at composition corresponding to m = a;: for /J.-phase at composition
corresponding to n, i.e.
( G~ )a at = ( G~ )13 at
/11 11 (10.35)
Similarly,
(10.36)
As discussed in Section 10.1.3, µ~ = 0 = µg. Hence,
G;: =µA, G:; =µB (10.37)
(µA)a at 111 = (µA){J at 11 (10.38)
(µa)a at 111 = (µa)13 ai /1 (10.39)
This is in conformity with the principle of equality of chemical potentials for phases at
equilibrium (Section 10.1.2). Hence, it is confirmed that the two-phase mixture would have
fixed compositions of a and f3. It also follows that only a or f3 single phase solutions would
be stable at the two-terminal regions since the L1Gm curves are lower than straight lines
j oining any two points on the curve in the regions 'am' and ' nb' .
From Eqs. (10.38) and (10.39), µA and µ 8 should be constant in two-phase mixture.
Similarly, on the basis of Eq. ( 10.28), aA and a 8 also should be constant in the two-phase
field a+ f3. These are shown in Figures ·10.3(b) and 10.3(c). It may be noted that µA =
0
at XA = 1, µ 8 = 0 at X8 = 1, as .per convention already discussed in section 10.1.3.
10.3.2 Phase Diagrams with Miscibility Gap
The free energy vs. c'omposition diagram of Fig. 10.3 schematically illustrates the thermodynamic
basis regarding stabilities of the two-terminal solutions a and f3 and formation of a+ f3 two-
phase mixtures. There are phase diagrams which exhibit what is known as 'miscibility gap'.
This is schematically illustrated in Fig. 10.4(a). Here, above the critical temperature Tw
there is only one phase. But below Tero the single phase is transformed into two-phase
mixture, which is the miscibility gap.
Assuming the solutions to be regular, L1Gm is given by Eq. (10.33). Differentiating L1Gm
w.r.t. X8 , we obtain (noting XA = l - X8 , and dXA = - dX8 )
(a)
I
I
I
i I•
I
----------------
Single phase
.
I
Phase I Phase II
Phase I
+
Phase 11
A B
A B
(b)
+ve
i
£D 0
-ve
at T = T 1
0
XA = X8 = 0.5
Xs
Fig. 10.4 (a) Phase diagram, ilnd (b) corresponding free energy-composition diagrams tor a
binary system exhibiting miscibility gap, assuming regular solution behaviour (schematic).
(10.40)
(10.41)
(10.42)
83(.1(Jm)
From Eq. (10.42), we get s:
uX83
=0 at XA = X8 = 0.5.
.
Smce at T = Tcr• b oth 83(.1Gm) and 82(.1Gm) s h ou Id be zero, th e maxima
. of t he nusc1
. 'b"l '
1 1ty
8X83 8X82
gap would occur at X8 = 0.5, provided the solution is a regular one.
Putting XA = X8 = 0.5 in Eq. (10.41), we get
52(L.\Gm)
8
X2 = 0 at 2 a = 4, i.e. a =2 (10.43)
B
Since RTa = .Q,
.Q
Tcr = (10.44)
2R
for a binary regular solution.
From Eq. (9.89), we see that, since L.\lf11 is approximately independent of temperature
at a fixed composition, .Q may be taken as a constant, and not as a function of either
temperature or composition for a particular binary regular solution. It has already been
emphasized once, and concluded that
n
a = - oc
1
RT T
which is the same as equation (9.92).
Thus, at T > Tw a < 2, and the solution is single phase. On the other hand, at T < Tcr•
a> 2, and the miscibility gap ap'pears. Figure 10.4(b) schematically shows the L.\Gm vs. X8
curves at different temperatures.
10.3.3 Free Energy vs. Composition Diagrams Involving

Change of Standard State
In Figs. 10.3 and 10.4, in the free energy vs. composition curves, the standard states for both
A and B were taken as the same (either both solid or both liquid). In actual phase diagrams,
Liquid Tm, (B)

solution
f To (a)
T If
1
Solid
solution
B
~(S)
dscn
+ve
f
AGm ds(s)
!
- ve
(b)
jAt T0 ,l 1
I
I
(c)
I [At T0 I
I
i
I
(d)
A B
Xa---+
Fig. 10.5 (a) Phase diagram, (b) free en1rgy-composition diagram at T = T0 , and (c), (d) activity-
composition diagrams of B and A, respectively at T = T0 , for system with complete
miscibility in liquid and solid phase (schematic).
Chemical Potential and Eqwlibria amongst Phases of Variable Compositions
we often encounter situations where one of them is liquid and the other solid at the same
temperature. An example is Fig. 10.5(a), where at temperature T0 , pure A is liquid, but pure
B is solid. In such a situation, for construction of free energy-composition diagrams, conversion
of standard state either from liquid to solid or vice-versa would be required. The thermodynamic
basis for doing the same is derived now.
Suppose free energy of pure solid i is d/
(s) and that of pure liquid i is d/
(I) at
temperature T. Then,
RT In [a ; (s)] = G; - d/ (s) (10.45)
RT In [a; (I)] = G; - d, (I) (10.46)
where a; (s) and a; (I) refer to activity of i in the solution with respect to solid and liquid
standard states, respectively.
RT In [a;(s)]
. a;(l)
=d;(I) - d,(s) =L1G2. (10.47)
where .1G~ is the free energy change of pure i upon melting at temperature T. Now,
(10.48)
At equilibrium melting temperature T,,,,
.1~ = .1H~ (10.49)

m T
m
Again,
•
.1~ (at 1) - L1~ (at Tm)= L1~
=Ml~ (at 1) - Mt;. (at Tm)
(10.50)
i.e.
.1~ = J~ [Cp(l) - CP(s)] dT +Tm~ (at Tm)
-r[ ~ (at Tm)+ J~ [CP(I) ; CP(s)] dT] (10.51)
If it is assumed that Cp(l) = CP(s), then Eq. (10.51) gets simplified into
L1~ :::: ~,(at Tm)• (Tm - T) = Li~ iat T,n) • (T,,, - T) (10.52)
m
i.e. LlG~ changes linearly with change of temperature.
EXAMPLE 10.7 Gold and silicon are mutually insoluble in the solid state and form a
binary eutectic system with eutectic temperature of 636 Kand composition of Xsi = 0.186
=
and XAu 0.814. Calculate the free energy of the eutectic melt relative to (i) unmixed (i.e.
pure) solid Au and solid Si, and (ii) unmixed liquid Au and liquid Si. Assume Cp is the same
for solid and liquid for both Au and Si.
Solution (i) Free energy of the eutectic melt, relative to pure Au and pure Si solids
at 636 K, is .dGm, where
l j l
LiGm = XAu GAu (eut) - ~u (s) + Xsi Gsi (eut) - OJi (s) j
(E.10.11)
Since the eutectic melt co-exists at equilibrium with both solid gold and solid silicon at
636 K,
(E.10.12)
Therefore, .dGm = 0.
(ii) For free energy of the eutectic melt relative to pure liquid Au and pure liquid Si
at 636 K,
(E.10.13)
Combining Eq. (E.10.13) with Eq. (E.10.12),
LiGm =XAu l ~u(s) - J

~u(I) +~Si ldJi(s) - dJi(I) J (E.10.14)
Since Cp of solid and liquid are assumed as the same, from Eq. (10.52),
G'°(s) - G'°(I) = - £.m = - .dll;,(Tm) (T.. - 636)

T.. m
(E.10.15)
m
For Au,
Tm = 1336 K, ~ = 12,760 J mol- 1
For Si,
Tm = 1683 K, ~ = 50,630 J mo1- 1
Substituting the values in Eqs. (E.10.15) and (E.10.14), we get
12 760 50 630
.dGm = 0.814 x • (636 - 1336) + 0.186 X • (636 - 1683) =- 11300 J mol1.
1336 1683
10.4 Some Typical Free Energy vs. Composition Diagrams
10.4.1 Complete Solubility in Solid and Liquid Phases
Figure 10.S(a) shows the phase diagram schematically. An example is iron-nickel system.
Tm(A) and Tm(B) are melting points of pure A and pure B, respectively. At T = T0 , pure
A is liquid, but pure B is solid.
Figure 10.S(b) presents the free energy diagrams at T0 . It is to be noted that, in reality,
we do not have either liquid or solid solution in the entire composition range. Hence, both
curves are partly hypothetical and partly real at T = T0 . Since the stablest states of A and
B are liquid and solid, respectively, ~(I) and da(s) have been arbitrarily taken as zero.
~ (s) and daO) are values at metastable states. Hence, they correspond to hypothetical
standard states of A and B, and the values can be calculated by employing Eqs. (10.47) and
(10.52); ef is the common tangent.
.. '
Figures 10.5(c) and 10.S(d) show activity vs. composition relations for Band A, respectively
1
at T0 . Curves corresponding to liquid solution have liquid A and B have standard states,
whereas curves corresponding to those of solid solutions have solid standard states.
10.4.2 Eutectic Phase Diagram with Finite Solid Solubilities
Figures 10.6(a) and 10.7(a) are the schematic phase diagrams. The terminal solid solutions
are a and /J. Figures 10.6(b) and 10.6(c) present free energy vs. composition and activity
vs. composition diagrams at T = Ti. Figures 10.7(b) and 10.7(c) are those at T = T2. Curves
in both Figs. 10.6(b) and 10.7(b) for the liquid solution have liquid standard states (real or
hypothetical, as the case may be). Curves for a and /3 solid solutions, respectively, have solid
A and B as standard states. The activity vs. composition curves have employed the real (i.e.
stablest) standard states for each component.
10.4.3 Phase Diagram with an Intermediate Phase

Figure 10.S(a) shows the nature of free-energy composition curves at temperature Ti.
Figure 10.S(b) is a schematic of a phase diagram with an intermediate phase; a and rare the
two terminal solid solutions. In addition, a high-melting intermediate solid phase fl is present.
10.5 Solubility of Solute in a Phase in Binary System

The solubility of a component in a solution refers to the concentration of the same in the
solution co-existing with another phase at equilibrium. In this section, this will be illustrated
and some quantitative relations derived using simple binary system A-B.
Consider Fig. 10.7(a). The solubilities of component B at T = T2 are x:(sat) and x: (sat)
Textbcok of Materials and Metallurgical Thermodynamics
A B
Liquid
(a)
T
+ve at T 1 ]
t b
d
LIG'11
a c
l
- ve
(b)
a
(c)
0 ........
~...._~~~~ ........
A B
Fig. 10.6 (a) Phase diagram, (b) free energy-composition diagram at T =

T1 , and (c) activity-
composition diagram at T =
T,, for a binary eutectic system with terminal solid
solubilities (schematic) .
A B
Liquid
t
T
(a)
b
a t T2
+ve
c
t
LlG'11 a d
(b)
i
- ve I
I
I
aJ--a+p--1p
QA
as
(c)
a
at T2
OA B
Xe ._..
Fig. 10.7 (a) Phase diagram , (b) free energy-composition diagram at T = T2 , and (c) activity-
composition diagram at T = T2 , for a binary eutectic system with terminal solid
solubilities (schematic) .
a Primary Inter-
solid mediate
0 solution phase 0
(/J)
t (a)
t
(b)
a+P
/J+-r
0 0 .5 1.0
Xs
Fig. 10.8 (a) LiG'"-composition diagram at temperature T1 (schematic), for (b) binary phase
diagram with an intermediate phase.
in a and p phases, respectively. Since at these compositions a and P phases are co-existing
at equilibrium, according to Eq. (10. 19), we have
~ [at x: (sat)] = ~ [at xC (sat)] (10.53)
Let us assume that B obeys Henry' s Law in a-phase since its concentration is small.
Therefore, based on the discussion in section 9.5.l , B will obey Raoult's Law in ,B'.phase.
Hence,
14 [at x: (sat)] = ~ +RT In a:[at x: (sat)]
=~ + RT In [fa · x: (sat)]
= µ& + RT In fa + RT In [X: (sat)] (10.54)
J4 [at x: (sat)] =µ& +RT In c4lat xC (sat)]

= µ& + RT In [XC (sat)] (10.55)
Equating Eqs. (10.54) and (10.55), we get
In lx: (sat)J= In lx: (sat)J+In fa (10.56)

i.e.
fa= xecsat)
x:csat) (10.57)
Suppose now that there is no solid solubility in ,8-phase, as shown in Fig. 10.9. Then
x: (sat)= 1. Hence,
l
x:csat) =fa (10.58)
L
rTJA)
L+ B
T
Metastable extensions
a+B
A Xa X:(sat) B
Xe ---+
Fig. 10.9 Hypothetical binary phase diagram illustrating derivation of solubility equation for B
in a-phase; dashed lines show metastable extensions.
Let us consider a composition X8 in a-phase, and let a 8 be the activity of B at X8 . Then,

xB (10.59)
aa = faXe =
x: (sat)
From Eq. (10.25),
(10.60)
At fixed composition X8 , on the basis of Gibbs-Helmholtz equation i.e. Eqs. (6.28), (6.29)
and their applications to partial molar properties [Eqs. (9.31), (9.33))
J n - II?. =nm = - R[d In [X~ (sat)l]

d[ (µB - µg)IT =
8 8 8 (10.61)
d(l/T) d(l/T)
. O 60 d[(µB - µg)IT] R[d In [X~ (sat)]] (10.62)

Again, from Eq. (1 . ), d(l/T) = - d(l/T)
Combining Eqs. (10.61) and (10.62), and integrating
X~ (sat) =A exp (- !~) (10.63)
where A is a constant.
Equation (10.63) thus provides a quantitative relation between solubility of B in a-phase
with temperature. Figure 10.9 shows the metastable extensions of curves by dashed lines.
It should be noted that the solubility vs. temperature relation would be governed by
Eq. (10.62) even in the metastable state.
EXAMPLE 10.8 Melting point of lithium is 453 K . Solubility of iron in liquid Li is

minimum at melting point of Li, and increases with increasing temperature. Given that
solubilities of Fe in Li CXFe (sat)] are 4.36 x 10-4 at 1473 K and 4.97 x 10~ at 673 K,
determine the minimum solubility.
Solution From Eq. (10.63),
XF,(sat) = A exp (- ~~) (E.10.16)
where n~ = partial heat of mixing of Fe in liquid Li. Hence,
In [XFc (sat)] = In A - ~
nm (E.10.17)
RT
Putting in the values of XFe(sat) at 1473 Kand 673 Kin Eq. (E.10.17), we obtain
ii~
-7.738 =In A - (E.10.18)
Rx 1473
n~ (E.10.19)
- 12 ..212 = In A -
Rx 673
Solving the above equations, In A = -3.97, n~ = 46143 J mo1- 1• Putting the values of
the above equations, XFe (sat) at 453 K = 9 x 10-8 •
[Note: A may be considered as equal to exp (Sf'e/R) which gives Sr:"! = -40.65 J mo1- 1 K- 11
10.5.1 Solubility of Metastable Phases

Figure 10.9 shows that solubilities of B in a-phase in metastable extensions are larger than
those in the stable regions of the bounding curves. The schematic free-energy composition
diagram for the iron-carbon system at a temperature above the eutectoid temperature is
shown in Fig. 10.10. Here austenite is the stable phase at the Fe-end. For carbon, graphite
is the stable and cementite is the metastable state. This means that the chemical potential of
carbon in cementite is higher than that in graphite. As Fig.10.10 shows, this leads to a higher
solubility of carbon in austenite in equilibrium with cementite as compared to that in
equilibrium with graphite. Experimental phase diagram data have confirmed this.
Chemical potential of carbon

for metastable equilibrium of /
I
austenite and cementite /
+ve /
/
Cementite /
""' /Graphite
/'/ ,,/::- 0 t
I / ,,,.,..,,."' I
,,,.,."" /le
l
..1.ve
r'
I
I
I
1- Solubility of
cementite
Fe
Xe~
c
Fig. 10.10 Schematic diagram of Fe·C system explaining why solubility of carbon in austenite
is higher at equilibrium with cementite as compared to that for graphite.
10.6 Summary
1. There are several definitions of chemical potential (µ;) of a component i in a
solution. However, the commonly employed definition in chemical and metallurgical
thermodynamics is in terms of Gibbs free energy, as follows:
(g(;'
Ji;= -
on.
)
r P,T , 11] , n2 • ...• except n;
This way, /1; becomes the same as the partial molar free energy of component i (G;).
2. For phases at equilibrium at constant T and P, the chemical potential of a component

is the same in all phases. If they are unequal, then the component may be transferred
from higher to lower chemical potential.
3. By convention, the chemical potential of a component at its standard state is assumed
to be zero. Hence,
µ, - ~ =µ, : : : G[" = RT In a;
4. Gibbs Phase Rule has been derived as:

Degree of freedom (F) = C - P + 2
where C = No. of components, and P = of phases. For equilibria of nonreacting

condensed phases, such as in alloy phase diagrams, pressure has negligible effect.
Hence, F = C - P + l.
For reacting systems, C = N - R, where N =No. of species, R =No. of independent
reactions.
5. Binary alloy phase diagrams (say A - B) can be theoretically constructed from
variation of .1Gm with composition at various temperatures. The basis for the same
is:
(a) intersections of the tangent to the curve at any composition at pure A and pure
B give values of µA and µ 8, respectively.
(b) The values of µA will be the same for all phases co-existing at equilibrium; the
same is true for µ 8 .
6. Thus, from .1Gm vs. composition curve, we can also construct µA and µ 8, as well
as aA and a 8 vs. composition curves.
7. Often, at some temperatures, the standard state of A may be solid and that of B
liquid, or vice versa. In these cases, construction of free energy composition curves
requires changing of standard state from solid to liquid or vice versa. Equations have
been derived for the same in this chapter.
8. The procedure for construction of phase diagrams etc. from .1Gm vs. composition
curves has been illustrated for some typical phase diagrams.
9. Finally, the solubility of a solute in a phase in binary system has been derived as
function of temperature.
PROBLEMS
10.1 Calculate the chemical potential of carbon in a gas mixture containing 30% CO,
20% C02 , 50% N2 at 1300 K. Will this gas saturate a piece of steel with carbon?
The total pressure is 5 atm.
Chemical Potentiar and Equilibria amongst Phases of Variable Compositions
10.2 Pure solids of Si, Si02 and Si 3N4 are equilibrated with a 0 2 + N2 gas mixture at
1000 K. How many degrees of freedom does this equilibrium have? What is the
value of chemical potential of N 2 at equilibrium? While maintaining the gas composition
constant, if the temperature is raised to 1200 K, what happens?
10.3 How many degrees of freedom does the system consisting of FeO-MnO solution +
Fe-Mn solution + OrN2 gas mixture have?
10.4 Calculate the chemical potential of S2 in a gas mixture consisting of 2% H2S, 50%
H 2 , and the rest N 2 at 1200 K. From chemical potential considerations, will this
mixture convert metallic copper into Cu2S?
10.S Calculate the chemical potential of oxygen gas (i.e. Oi) in a system consisting of ( .
pure liquid iron and a liquid slag containing FeO at equilibrium at 1600°C; the
activity of FeO in the slag is given as 0.45.
10.6 Derive an expression for the free energy of fusion of iron (LiGm) as function of
temperature. Assume that Cp of liquid Fe is larger than that of solid Fe by
1.3 J mo1-1 K- 1• Heat of fusion of Fe is 15,360 J mo1-1 at its melting point of 1535°C.
10.7 The solubility of carbon in liquid aluminium is 6 ppm at 960°C and 12.5 ppm at
1000°c.
(i) Calculate the solubility at melting point of Al, i.e. at 660°C. Assume dilute
solution behaviour.
(ii) Predict whether solid aluminium carbide (Al 4C 3) will form at carbon saturation
of liquid Al at 660°C. It is given that, for the reaction:
Lie° = - 266,521 + 96.23 T J moi- 1

10.8 The Ag-Cu alloy has a eutectic phase diagram with eutectic temperature of 780°C.
Below the eutectic temperature, the eutectic liquid decomposes into a mixture of
a(Cu-rich) and /1(Ag-rich) solid solutions . At· 500°C, the saturated a- and
,8-phases respectively contain 98.8 and 3.35 atom % Cu. Assuming Henrian behaviour
of Cu in ,8-phase and Raoultian behaviour in the a -phase, calculate Henry's Law
constant of Cu in ,8-phase.
Reaction Equilibria Involving
Condensed Phases with
Variable Compositions
In Chapter 8, we presented techniques of calculations of reaction equilibria involving ideal

gases and pure condensed phases. These were illustrated with worked-out examples as well.
This chapter is concerned with equilibrium calculations where the solid and liquid are not
necessarily pure, but are solutions. However, before doing that, we shall give some examples
of construction and/or use of phase diagrams for reactive systems.
In section 8.3.3, the phase stability diagram of Ni-S-0 system has been derived on the
basis of gas-solid reaction equilibria. There, all solid phases were taken as pure, and hence,
activities of solids are 1. In this chapter, first we shall give some further examples of pure
solids.
11.1 Analysis of Phase Stabilities of Reactive

Stoichiometric Compounds-The Si-C-0 system
In the Si-C-0 system at temperatures of interest, there are four stable condensed phases-
silicon, carbon, sil ica, and silicon carbide. C and SiC remain solids at all temperatures of
interest in manufacture of SiC, and may be assumed to be pure. Si melts at 1683 K. It
dissolves some carbon whose concentration is very low, and liquid Si may be assumed to
be pure. Si0 2 melts at 2001 K and remains essentially a stoichiometric compound.
Besides the condensed phases, there would be a gas phase containing CO, C0 2 and SiO
at high temperature. The following four reactions may be considered:
C(s) + C02(g) = 2CO(g) (11.1)
186
Reaction Equilibria Involving Condensed Phases with Variable Compositions
Si0 2(s, I) + CO(g) = SiO(g) + C0 2(g) ( 11.2)

SiC(s) + 2C02(g) = SiO(g) + 3CO(g) (11.3)
Si(s, I) + C02(g) = SiO(g) + CO(g) (11.4)
The above reactions are all independent since none of them can be arrived at by combining
the others. Section 10.2 has discussed phase rule and its application to reaction systems as:
Degree of freedom F =C - P+2
C=N-R
which are the same as Eqs. (10.31) and (10.32). In this example,
N = 7 (Si, Si02, SiC, C, CO, C02, SiO)
R = No. of independent reactions = 4
Hence, if P and Tare fixed, then the maximum number of phases co-existing at equilibrium
will be 3 (i.e. for F = 0).
The relations for equilibrium constants for the above reactions are
_ (pcoi _ Pc0i • Ps;o

KI - ' K2 -
Peoi. Pco
K4 -_ PcoPs;o (11.5)
Pc0i
where p refers to partial pressures at equilibrium. Again,

Pco + Pc0i + Ps;o = PT = total pressure (l 1.6)
Figure 11.1 presents calculated log Pc0i vs. log Ps;o curves at equilibria with reactions
=
(11.1)-(11.4), at 1700 K and P 1 atm on the basis of Eqs. (11.5) and (11.6) (Ref.: A.
Ghosh and G.R St. Pierre, Trans. Met. Soc. AIME, 245, 1969, p. 2106), and from the
standard free energies of formation of compounds.
Derivation of the ternary phase diagram from Fig. 11.1 is based on the following
principles:
1. Phase rule allows equilibrium co-existence of a maximum of three phases at fixed
P and T.
2. The gas compositions in equilibrium with single condensed phases are given by the
curves in Fig. 11. l.
3. At constant Ps;o. Si0 2 is unstable with respect to a gas composition which contains
less C02 than that corresponding to the SiO:reurve. On the other hand, at a constant
PSiO• C, Si and SiC are unstable if the gas contains more C02 than those corresponding
to their equilibrium curves. These statements follow directly from Eqs. (11.1)-{1 l.4).
-2
-3
B
Carbon
-5
-6
- 7 L-.--...L...'-----'-----"'----'-----'---'
-5 -4 -3 -2 -1 0
log PSiO
Fig. 11.1 Log Pco-i vs. log PsiO in the gas phase in equilibrium with various condensed phases at
1700 K, and at Pr= 1 atm. [A. Ghosh and G.R. St. Pierre, Trans. AIME, 245, 2106, (1969.)J
Using the above principles, all regions (A, B, C etc.) and all points of intersection in
Fig. 11.1 were examined to find out the stable co-existing phases in the system. Thus we
arrive at Table 11.1.
Table 11.1 Derivation of Phase Fields
Phase Fields (Fig.11.2) Phase field derived from Fig. 11.1

Gas Region B
Si0 2 +gas SiOr curve
C +gas Carbon-curve
C + Si02 +gas Carbon-Si02 Intersection
C + Si02 Region C
SiC + C + Si02 Region D
SiC + Si02 Region E
Si + SiC + Si02 Region F
As an explanation of Table 11.1, let us consider a gas composition in region C of
Fig. 11.1. Carbon will be stable, since Pco-i is below that at equilibrium of reaction (11.1),
meaning that the reaction will tend to proceed from RHS to LHS, thus depositing carbon.
Si02 will be stable since Pco-i in region C is above that of equilibrium for SiOrgas. Thus
the reaction (11.2) shall proceed from RHS to LHS . However, SiC and Si will not be stable
if similar arguments are employed for their reactions.
Figure 11.2 presents the calculated Si-C-0 ternary phase diagram based on the stable
phases of Table 11. l.
____ _ Boundary line not

shown by Krivsky and
Schuhmann.
Si + SiC + Si02
Si SiO 0
Flg.11 .2 Si-C-0 phase diagram at 1700 K and 1 atm [A. Ghosh and G.R. St. Pierre, Trans.
AIME, 245, 2106 (1969)] .
11.2 Phase Diagram with Reactive Non-stoichiometric

Compounds-The iron-oxygen system
11.2.1 Phase Diagram
Oxides of iron, viz. Magnetite (Fe30 4) and Hematite (Fe2 0 3), are minerals from which iron
is extracted. They have many other industrial uses also. The phase diagram of iron-oxygen
system is presented in Fig. 11.3. Its special features are as follows:
• There are three solid oxides, viz. FeO (Wustite), Fe30 4 (Magnetite), and Fe20 3 (Hematite).
• Wustite is a non-stoichiometric compound. Hence it is designated either as "FeO" or
FexO, where x ranges between the overall limits of 0.835-0.945. Moreover, it is stable
only above 570°C.
Atom % oxygen
50 52 54 56 58 60
Liquid iron Liquid o xid
1600 + oxygen
R Magnetite
+ oxygen
y z
1400
c:
0
><
0
1200 +
u
0
y-lron B
";:J
<IS
0
.....
+ e
0
::I wustite :I:
.....
~
.....
4) 1000
c..
E Wustite Magnetite
4)
L + +
E-< magnetite hematite
800
a-Iron
+
wustite
600
Q a-Iron + magnetite
FeO Fe304 Fe203
400
0 0.422 24 26 28 30
Wt.% oxygen
Flg.11.3 Iron-oxygen phase diagram [L.S. Darken and R.W. Gurry, J.Am. Chem. Soc., 68, 798
(1946)].
• The oxygen-to-iron ratio in the system at a fixed temperature depends on partial

pressure (i.e. chemical potential) of oxygen in the gas phase with which the solid
phases are brought to equilibrium.
There are several other phase diagrams of oxide system etc., where such features are
present, i.e. non-stoichiometric compounds and dependence of composition and nature of
phases of the condensed system on gas composition. But the Fe-0 system constitutes the most
common one.
11.2.2 Activities in the Wustite Phase

Wustite phase may be considered as a solid solution of FeO and Fe 30 4. It is a ionic solid
having Fe2+, Fe 3+ and 0 2- ions. There is appreciable concentration of vacancies in the

cationic lattice due to Fe deficiency. Therefore, the activity of FeO would continuously vary
in the Wustite phase field. The choice of standard state poses problem since stoichiometric
FeO is not available. Hence, it is appropriate to talk of an "FeO'', where "FeO" refers to
non-stoichiometric Wustite. By convention, "FeO" co-existing with metallic Fe at equilibrium
is taken as its standard state (e.g. point 'L') at any temperature, T.
The choice of standard state for ions poses problem. Let us take the example of aqueous
solution and consider the case of CuS04 dissolved in water. The standard state of Cu2+ and
sof- are unknown since they are not available in pure form. The difficulty is overcome by
assuming
2+ 2-
llCu ' 0504 =£lcuS04 ( aq) (11.7)
If the solution is saturated with pure solid CuS04 , then as per accepted convention,
(11.8)
With similar approach, it may be written that in Wustite phase,

(11.9)
At a temperature T and the corresponding point on the line QU, aFc0 = 1 as assumed earlier.
Since aFc0 is at equilibrium with pure Fe, aFe = 1. Hence, from Eq. (11.9), a0 = 1 in Wustite
co-existing with metallic iron.
The reaction of oxygen may be written as
(O)w = 21 02 (g) (11.10)
wh~re the subscript 'w' denotes Wustite, for which
(11.11)
Combining Eq. (11.11) with equation at standard state, where a 0 = 1, we obtain
(11.12)
where P°oi_ represents P0z in equilibrium with Fe + FeO.

The procedure adopted for experimental determination of a 0 , aFe• aFcO in Wustite field
is as follows:
1. Temperature and P0z of gas were fixed. The solids were allowed to come to equilibrium
with the gas. Then the solid sample was rapidly quenched to room temperature, and
analyzed for Fe and 0 content.
2. The solid phases were identified by X-ray diffraction studies.

3. a 0 in Wustite field was calculated at various compositions from Eq. (11.12).
4. aFe was calculated by the Gibbs-Duhem integration on the basis of Eq. (9.62) from
a0 vs. composition data.
5. aFeO was calculated from the above using Eq. (11.9).
Figure 11.4 presents activity vs. composition data in the Fe-0 system at 1100°C. It may
be noted that tre minimum value of a 0 is l. This is due to the procedure adopted.
0.8 40
0.6 30
~
~
i::s
0
i::s
tf
i::s
0.4 20
0 .2 10
0 _ _ ___,.....___._ _ _ . . _- - 4 _ - - & ._ ......L_ __.
0 .51 0.52 0 .53 0.54 0.55

Xo
Fig. 11.4 Activities of iron, oxygen, and Wustite in the Wustite phase at 1100°C.
11.3 Reaction Equilibria Involving Solutions of

Condensed Phases
The thermodynamic behaviour of solutions has been discussed in Chapter 9. Now we shall
present some examples of reaction equilibria calculations involving solutions of condensed
phases. However, first we shall discuss the general considerations.
11.3.1 General Considerations

For this, let us reconsider oxidation of Mn , viz.
2Mn(s) + 0 2(g) = 2Mn0(s)
which is the same as Eq. (8. 18).
In section 8.3.2, both Mn and MnO were assumed to be pure; that is a special case. The
general situation is that Mn is present in metallic sol ution, and MnO is present as an oxide
solution in association with other oxides. The solutions may be solid or liquid. By convention,
we write the above reaction as
2[Mn] + 0 2(g) = 2(Mn0) (11.13)
or
2Mn + 0 2 (g) = 2(Mn0) (11.14)
The round brackets denotes a compound present in a non-metallic solution (oxide, sulphide,
chloride etc.), while the square bracket or underline denotes an element in a metallic
solution. It should be mentioned here that all components are present as elements in a
metallic solution. In other words, it is an atomic solution. Therefore, solutions containing
compounds will exist as separate phases.
The equilibrium constant for the above reaction is
Ku= {,~~;~L (11.15)
For pure Mn and pure MnO, if p°o.i_ is equilibrium Pai, then
(11.16)
From Eqs. (11.15) and ( 11.16),
(11.17)
From the Ellingham diagram (Fig. 8.2), at 1200 K, P°oz = 5 x 10-20 atm. Let [aMn1
= 0.3 in a Mn-Fe solid solution and (aMno) =·0.45 in a Mn0-Si02 solid solution. Then, Pai
at equilibrium with these will be 11.25 x 10- 20 atm.
A major part of thermodynamic measurements on metals, alloys and other materials
consists of experimental determination of activities of components in various solid and liquid
solutions at high temperature. With extensive measurements by physical chemists, metallurgists
and materials scientists over a period of 60 to 70 years, activity versus composition data are
available in most solutions of interest.
EXAMPLE 11.1 Consider the reaction:
[Fe] + H20(g) = FeO(s) + H 2(g), at 910°C and 1 atm. total pressure.

When the iron is pure, the gas mixture at equilibrium contains 60.2% H2 and 39.8% H20(by
volume, as per standard convention for gases). If Fe is present as a solid solution of Fe and
Ni(XFc = 0.721), the equilibrium gas composition is 51.9% H 2, and 48.1 %H 20.
CaJculate the activity of Fe in the alloy, assuming FeO to be pure.
Solution The equilibrium constant for the above reaction is
£li.eo x PH2 )
K = ~ x PH20 eq
( (E.11.1)
Since FeO is pure, aFe0 = 1. Hence the above equation may be rewritten as
K-[:,~)., =~:: for pure Fe
=( :.~ ~)., = ~:~~~ .~ for Fe-Ni (E. 11.2)
From Eq. (E.11.2), aFc in the alloy= 0.713. Since XFc = 0.721, it is an ideal solution (within
error limit).
EXAMPLE 11.2 The PH/PH2o ratio in an HrH20 gas mixture in equilibrium with pure
liquid Pb and lead silicate melt, at XPbO = 0.7, is 5.66 x 10-4 at 900°C. The corresponding
value at 1100°C is 1.2 x 10-3. Calculate the partial molar heat of mixing of liquid PbO in
this lead silicate melt.
Solution 900°C = 1173 K, 1100°c = 1373 K
The reaction is:
Pb(l) + H20(g) = (Pb0); 0 melt + H2(g) (E.11.3)
for which
~=- RT ln K3 = - RT In (~ x PHz J (E.11.4)

~ XPHzo
Again,
~ =£r(Pb0) - £r(H20) = (-191,750 + 79.IT) - (-246,000 + 54.8T)
= 54,250 + 24.3T J mo1- 1
Hence,
uJJ, = 82,754 J , K3 = 2.06 x 10- 4 at 1173 K,
LYJ1 = 87,614 J, K3 = 4.64 x 10- 4 at 1373 K,
Noting that a Pb = 1,
(E.11.5)
Putting in the values given in Eq. (E.11.5), we get aPbO = 0.364 and 0.387, respectively at
1173 K and 1373 K.
Now,
GM =RT In (aPbO) (E.11.6)

Hence,
G;:;,0 = - 9856 and - 10,837 J mo1- 1, respectively at 1173 K and 1373 K.
Partial molar heat of mixing of liquid PbO in the lead silicate melt (HM) is related
to GM by Gibbs-Helmholtz equation as
8(~) (E.11.7)
8(~)
P, comp
Assuming HM to be independent of temperature, integration of Eq. (E.11.7) leads to
(E.11.8)
Putting in values, we get HM = - 4803 J mo1- 1•
EXAMPLE 11.3 Calculate the partial pressure of nitrogen gas over an eutectic liquid
solution of Al-Si alloy at the eutectic temperature, at equilibrium with solid Si 3N 4 (Johnson
and Stracher, 1995, p. 134).
Solution The relevant reaction is
(E.11.9)
for which
(E.11.10)
Since Si3N4 is pure, the activity of Si 3 N4 is l.

From Al-Si phase diagram, the eutectic temperature = 850 K, and at eutectic point,
Xsi = 0.122. At this temperature, pure Si is a solid. But for application of Eq. (E.11.10),
liquid Si is to be employed as the standard state, and ~ is to have the appropriate value
for this.
~ can be calculated from the following data obtained from literature sources .
..1Gf1 = -753,190 + 336.43T J mo1-1 for 3Si(s) + 2N2(g) = Si3N 4(s) (E.11.11)
..1~ = 50,630 - 29.91 T J mo1- 1 for Si(s) = Si(l) (E.11.12)
From Hess' Law,
~ =.16?. - 3~ = (-753,190 + 336.43 x 850)
-3(50,630 - 29.91 x 850) = - 542,844 J moi- 1

Solving Eq. (E.11.10), we get
{[asd3 di 2 } =4·.36 x l 0-34

x cq
From literature sources, at 850 K and Xsi = 0.122 in liquid Al-Si solution, with reference
to liquid Si as standard state,
GS: = RT In asi = - 24,324 J

This yields
asi = 3.28 x 10-5
Solving, PN2 at equilibrium with Al-Si melt and Si3 N4 (s) = 3.7 x 10-• 5 atm.
If formation of Si3N4 is not desired, then PN2 in the atomosphere should be kept below this value.
[Note: If solid Si is used as standard state, then asi in Al-Si. eutectic with reference to solid
Si is to be emloyed.]
EXAMPLE 11.4 The reaction:

2[Cu]in liq. Pb + PbS(s) = Cu2S(s) + Pb(!) (E.11.13)
allows removal of Cu present as impurity in molten lead. The solid sulphides are present as
pure compounds, and Pb is insoluble in solid Cu. The solubility of Cu in liquid Pb is given as
8060.5
In [Xculsai =- - T
- + 5.207 (E .11.14)
Calculate the maximum extent of removal of copper by this reaction at 800°C.

Solution For the reaction (E.11.13),
('1cu2s> x [aPb]}
~3 =- RT In K13 =- RT In { 12 (E.11.15)
["cuJ X (~ ) eq
Since Cu 2S and PbS are pure, Ocu2s = aPbS = 1. Concentration of Cu in liquid Pb would
be very small (can be verified later). Hence Pb is almost pure, and the activity of Pb may
be assumed as 1. Therefore,
L!Gf, =- RT In {r~J' L (E.11.16)
Assuming that Henry' s Llw is applicable to Cu in liquid Pb, we have
'1cu =r°eu •Xeu = xeu (E.11.17)

[Xeu lsat
on the basis of (Eq. 10.59).
From Eq. (E.11.14),
[Xailsat = 0.10 at 800°C (i.e. 1073 K)
~3 = lNf/ (Cu S) - Nf/ (PbS)j

2
= [(-142,900 - lt.3T In T + 120.2n - c-151,750 + son1

= 14,350 - l l.3T In T + 40.2T = -27,126 J mo1- 1 at 1073 K.
Combining Eqs. (E.11.16) and (E.11.17) and putting in values, X01 = 0.0216.
This is the minimum copper concentration attainable. In practice, it may be somewhat
higher if equilibrium is not attained.
Since XPb = 1 - Xai = 0.9784 = 1, the assumption that aPb = 1 is all right here without
much error. If it were not the case, then the calculation had to be repeated once more by
assuming Raoult's Law for Pb (i.e. aPb =1 - X01) .
11.4 Interactions amongst Solutes in Dilute

Multicomponent Solutions
In a binary A-B solution, the interactions amongst components may be classified into the
following three categories : A-B, A-A, B-B. In a ternary A-B-C, the interaction pairs would
be: A-B, A-C, B-C, A-A, B-B, C-C. These additional interactions will have some effect on
activity coefficients. Now take a binary solution. Suppose the activity coefficient of B is
1fi; add a component C to the solution; the value of 1B will change to some other value
(say r8) due to additional interaction of component C on 1fi.
Textbook of Materials and Metallurg1cal Thermodynamics
In metallurgy and materials science. we have to ofren deal with multicomponent solutions-
an important issue. In concentrated multicomponent solutions, the mathematical procedure
for dealing with such interactions is lengthy. and 1s beyond the scope of an introductory
thermodynamics course. For example, during extraction and refining of liquid metals, we
mostly deal with dilute multicomponent solutions. An important example 1s liquid steel,
where Fe is the solvent, and several solutes (C, Si, Mn, S, P etc.) are present in very low
concentrations. Fnr such dilute multicomponent solutions, Carl Wagner ( 1952) proposed a simple
but elegant ana {tical method of handling solute-solute interactions, which is discussed now.
11.4.1 Interaction Coefficients
Let A designate the solvent, and l, 2, ... i, j, ... are the symbols for the solutes. Wagner
expressed In y, (i = general symbol of solute) as function of XI> X2, .•. by the Taylor series
expansion around pure solvent A. The equation is as follows:
= In y, + [ X1 (8[1ny,]) + ··· + X1 [8[1ny,]l + ···]

. 0
In }'; (mult1component) DX DX
I XA-+ I l XA-+ I
2 2
+ -l X12 (8 [lny,]) + ... + -1 X1 (8 [1ny,Jl + ...] (1 J.18)
2
2 :!
[2 8X1 xA-+ I 2 8X) XA4 I
Here, 'f. = Henry's Law constant in A - i binary.

For dilute solution, XI> X.2 , ... Xj ... are very small. Hence, Wagner ignored all higher
order terms. Then Eq. (l l.18) becomes
In Y; -_ 1n Y;0 + X1 (8[ln Y, ]) + ··· + X, (8lln Y, ]) + X1 [8lln Y; ]] + ...

8XI XA-+I
8X 1
XA-+I
8X1 XA-+I
(ll.19)
Although XA ~ l is the mathematically rigorous constraint, it is valid only 1f mole
fractions of all solutes tend to zero. On the basis of this constraint, Eq. (l l.19) may be
rewritten as
In Y; =In r? + X1 (8[ln Y, ])
8X1 x1-+ 0
+ ... + X, (8[1n Y, ])
8X, x,-+ 0
+ X1 [ 8lln Y,
8X1
]l + ...
X 1 -+0
(l l.20)
The influence of one solute on y of another solute was termed as interaction coefficient
by Wagner. This allows us to rewrite of Eq. ( 11.20) as
(1 1.21)
where
e! = (blln Y;]J (11.22)

I ox.
1 Xr-+ 0
and cf is known as interaction coefficient describing influence of solute j 011 In r;.
11.4.2 Equality of e1 and ej

In analogy with the definition of cf in Eq. (1 1.22),
(11.23)
where ~ = interaction coefficient describing the influence of solute i on In IJ·

From Eq. (9.11) we may write, at constant P and T,
- (/j(;'
G. = -
I t5i
) n, "I .... excep1 n;
' - (/j(;']
G1= -
on . (11.24)
J "! .... CXcepl nJ
Again,
cm =(o(G'Ol- G~)J
I '
G'!1
J
=(o(G' - G~)J
on . (11.25)
n; "I .. ., except n; 1 "I .... excep1 nj
where G~ is G' when components 1, 2, ... , i, j, ... are present as pure in mixture of the
same overall composition.
From Eq. (11.25),
oGf" _ $1-(G' - q;) _ <>Gj

(11.26)
on1 - Oflt on1 - Oflt
Again,
(oGf")
& .
J Xr -+ 0
=RT [o(ln a;)]
& .
J Xr-+O
= RT {<>(In X;) + o(ln r;)}
~ ~.
J Xr-+O
(1 1.27)
since X; and Xi are independent variables. Similarly,
oGmJ
_ J_ RT .
=-E'·
( on. 1
X;-+0
n,. J
(11.28)
Combining Eqs. (11.26)-( 11.28), we get
(11.29)
This is a very important conclusion. It means that if experimental value of Ef is

available, we automatically know Ej .
11.4.3 Practical Utility of Interaction Coefficients
As already stated, activity versus composition data are to be basically obtained from experimental
measurements. For a binary solution, it is the simplest. The activity of one component is
measured, and from G-D integration, the activity of the other can be calculated. This
approach works for ternary also. For multicomponent solutions, however, much larger
experimental programme is required since there are several composition variables. This is
where the concept of Interaction coefficient for dilute solutions comes in handy.
Let us consider the multicomponent A, 1, 2, ... i, j, .. . again. For knowing y,, what we
have to do now is to make measurements in binary A-i, and all ternaries, i.e. A-i-1 ,
A-i-2, .. . A-i-j , .... Experimental data in A-i allows determination of Y/. If experimental
data in A-i and A-i-j, are combined, we can find out cf since in the ternary A-i-j ,
(11.30)
Si milarly, from data on other ternaries, tf, if' ...

are evaluated. Then these can be
employed to calculate r. in multicomponent solution by Eq. (11.21). This is a tremendous
saving of experimental efforts.
t4 is known as self-interaction coefficient and is a measure of deviation from Henry 's
Law in A-i binary. If HL is obeyed, then t4= 0.
Although derivation of Eq. (11.21) assumes mole fractions of solutes as tending to zero,
it has been found that often the equation can be satisfactorily employed upto a reasonable
concentration. For still higher concentration, retention of the second order terms is necessary.
For example,
(11.31)
is the second order interaction coefficient describing influence of j on In y,.

11.5 One Weight Per cent Standard State
11.S.1 Introduction
The conventional standard state is also known as Raoultian standard state, where the
standard state (SS) is pure solid or pure liquid, whichever is the stablest at the temperature
under consideration. Moreover, Raoult's Law serves as the reference.
Thermodynamics permits use of any other alternative standard state.· Such a state may
be real or hypothetical. However, for such a change of SS, the thermodynamic data employed
for calculation should conform to the new SS. For example, ..16'° for reactions in general
data sources are all for the conventional SS. They are to be changed if a new SS is employed.
In extraction and refining of liquid metals as well as for a variety of other uses, we deal
with dilute solutions. Also, industries prefer to find out answers directly in weight percent.
In order to take care of these, Chipman proposed the " I wt.% Standard State". Its features
are:
• Composition scale is in wt.%, not in mole fraction.
• Henry' s Law, rather than Raoult's Law, is the basis.
• The answer for equilibrium calculation would be directly in wt.%.
Consider the multicomponent solution with 1, 2, ... , i, j , ... as components. Then
conversion of wt.% into mole fraction can be done with the help of the following equation:
_ WJM;
X1 - (11.32)
l:(W;IM;)
where
W; = wt.% of i,
M; = molecular mass of i.
The reverse conversion is as follows:
(11.33)
11.5.2 Definition and Relations for 1 wt.O/o Standard State

Consider a dilute binary solution A-B, where A is the solvent and B the solute. Then the
activity of B in 1 wt.% standard state is defined as
8 (11.34)
h = (::. ""L
s -+ 0
where h8 is activity of B in the new SS.

When Ws ~ 0, Henry's Law (HL) is obeyed. Hence, (~ l is constant and is

B~~o
related to HL constant (~). Now, if we consider the application of the above equation on
the HL line, then as in both the numerator and the denominator in Eq. (11.34) are the same,
and Eq. (11.34) gets simplified into
Ira = Wa at W8 ~0 (11.35)
In other words, on HL line, hs = W8 .

Now, consider the point on the HL line at W8 1. From Eq. (11.35), hs 1 on that = =
point. This is the 1 wt.% standard state. The SS may be real or hypothetical. Figure 11.5(a)
and 11.S(b) illustrate these.
In real SS, HL extends beyond W8 = 1 (Fig. 11.5a). In the hypothetical SS, HL ends
below W8 = 1 (Fig. 11.Sb). At point S, Jz 8 = 1, and hence it corresponds to 1 wt.% standard
state.
Raoult's
t as vs X8
curve /
/
/
t
as. ha
/
as /
/
J,, . .
1fa
/
as = as ------7--- ,., ...
/ 'P .... h8 = 1 ---<
/
,,,,,
I .,..
""1S
as = a8 - - ; 7- -.:. ... ::..-17 .... .... .... I
.... I
0
§/
A Ws = 1
~
t
:
B
l A
I
I
Wa = 1
We= Wa X8 = 1 Wa-
X8 = 0
W8 = 0 We= 100
(a) (b)
Real 1 wt.% SS at point S Hypothetical l wt.% SS at point S
Fig. 11.5 Illustration for 1 wt. % SS: (a) real 1 wt.% SS, and (b) hypothetical 1 wt.% SS.
In Fig. 11.S(a), at W8 = W8 , pd = h8 , qd = Ws for 1 wt% SS. Hence,
(11.36)
where fa = activity coefficient of B in 1 wt.% SS.

fleACliOfl [Qfli/i/Jria Involving Condensed Phases with
From the value of yg shown in Fig. 1l.5(a). at point q,
( 11.37)
Since at p, a 13 = raX8 , the combination of Eqs. (11.36) and (11.37) yields
(l l.38)
Again, from Eq . ( 11.38),
(l l.39)
For binary A-B, and W8 ~ 0. WA ~ 100, Eq. (l 1.32) gets simplified into
(11.40)
Combining the above, we get

hp, JOO Me
-=--~
(1 1.41)
aB yg MA
11.5.3 Evaluation of L'.1G 0 for Reaction Involving 1 wt.o/o SS

Consider the following reaction at temperature T for a dilure binary metallic solution A-B,
where B is the solute.
(l l.42)
The equ ilibrium relation is to be rewritten, for l wt.% SS for B, as
Ll~ = - RT In Kh = - RT In [ (aee>i)
[he] P0z
leq
(11.43)
where Ll~, K1r are LlG° and K for l wt.% SS for component B. The subscript 'h' denotes
Hennan since HL is reference here. The corresponding equation in the Raoultian SS are:
LlG° =- RT In K =- RT In { (aeo.,) } ( 11.44)

o:::.- [aeJ P0i cq
From Eqs. (l l.43) and (l l.44),
Ll~ - Lld =- RT In ( ~) = - RT In ( ~:) =RT In ( ~:) (I 1.45)
(~Jj
-<1)
Textbook of Materials and Metallurgical Therm
Equation ( 11.46) r -
thermochemical ·
only addit=
are -
f;' is
Eq. (1 .
The commm
The relationship of
Combining Eq. l
i.e.
A term-by-term comparison wih

From the value of yg s hown in Fig. 11.S(a}. at point q,
(l l.37)
Smee at p. aB = )UX8 , the combinacion of Eqs. (11.36) and ( 11 .37) yields
(l l.38)
Again , from Eq. ( l l .38),
(11.39)
For binary A- B, and W8 ~ 0, WA ~ 100, Eq. (l l.32) gets simplified into
(11.40)
Combining che above, we get
-118 = -100M
-=-8
(l 1.41)
aa ygMA
11.5.3 Evaluation of .1G 0 for Reaction Involving 1 wt. 0/o SS
Consider the following reaction at temperature T for a dilute binary metallic solution A-B,
where B is the solute.
(l l.42)
The equilibrium relation is to be rewritten, for l wt.% SS for B, as
Lle'2 =- RT In K 11 =- RT In [ (aaai)
[ha) Pai
l
eq
(1 1.43)
where Lle'2, K 11 are LlG° and K for 1 wt.% SS for component B. The subscript 'h ' denotes
Henrian since HL 1s reference here. The corresponding equation in the Raoultian SS are:
LlG° =- RT In K =- RT In { (aacn) } (l l.44)

fanJ Pai eq
From Eqs. ( l l.43) and ( l l.44 ),
RT In (Kh)
K
=- 0
RT In ( a) =RT In
ha
(!ta )
l an (ll.45)
LC2 =KfJ +RT In (100 Ms] (11.46)

fa MA
Equation (11.46) provides the basis for evaluation of 4 ·
L1G0 is available in standard
thermochemical data sources. Hence, ex,.perimental determination of fa in binary A-B is the
only additional information required. rB depends on both B and A. Such experimental data
are available with important solvents such as liquid Fe, Cu, Pb, and a variety of solutes.
Tables are also available with compilation of 4for the formation of common compounds.
11.5.4 Dilute Multicomponent Solution in 1 wt. 0/o SS

In multicomponent solutions, additional solute-solute interactions are to be considered. It has
been done elaborately for Raoultian standard state in Section 11 .4. For I wt.% SS,
Eq. (11.21) has been modified as
log/;I = log t' + L.j W.e~

I ) ) (11.47)
f is Henry's Law constant for I wt. % SS. Since f = 1 by definition, log t' = 0. Hence,
Eq. (11 .47) gets simplified into
(11.48)
The common logarithm was chosen for convenience, and
(11.49)
The relationship of e/ with ej may be derived on the basis of Section 11.4, as follows:
Combining Eq. (11.48) with Eq. (11.38), the former may be rewritten as
'Yi] .
In ft =In
(~ =2.303 7W;ef (11.50)
i.e.
In 'Y; = In i/ + 2.303 L. Wie/
)
(ll.51)
A term-by-term comparison with Eq. (1 l.21) leads to
(l l.52)
Combining Eq. (11.52) with (11.40), we obtain
(11.53)
This allows evaluation of e( from e( and vice versa. Similarly,
. MA .
e'.1 = t 1. (11.54)
230.3 M;
Since Ef = ~, from Eq. (8.104), we have
230.3 e( Mj M1 j
e'. = MA el = MA ----=-e ; (11.55)
} 230.3 M; I 230.3 M, MA M;
For a ternary A-i-j, Eq. ( l l.48) reduces to: log f; = "') e/ .

Hence, plots of Jog f; vs. Wj
would be straight lines passing through the origin. Figure 11.6 shows such plots for activity
of nitrogen dissolved in liquid iron at 1833 K for several ternaries such as Fe-N-Ni and
Fe-N-Cr. Nitrides of Cr, V etc. are stronger than nitride of Fe. In other words, Cr-N,
V-N bonds are stronger than Fe-N bonds, whereas Ni-N bond is weaker than Fe-N bond.
This explains why fN decreases upon addition of Cr etc., and increases upon addition of
elements like Ni.
Tables of e/, e/ in metallic solutions for common solvents (Fe, Cu etc.) at high temperature
are available in literature.
EXAMPLE 11.5 A Hr H2S gas mixture with H 2/lI2S ratio of 10 was equilibrated with
pure liquid iron at 1850 K. On analysis of quenched sample, the iron was found to contain
0.8 wt.% sulphur. Calculate the free energy change for the following reaction at 1850 K:
z
l
S2 (g, 1 atm) = [S]1 wt.% ss in liq. Fe (E.11.18)
Given at 1850 K : Jog fs =- 0.028Ws
(E. l l.19)
Solution
H2(g) + [S]1 wt%SS = H2S(g); Liejl0 (E. 11.20)
Substracing Eq. (E. l 1.20) from (E.11.19), we get
(E.ll.2 1)
l.O
0.8
0.6
0.4
0.2 Co, Ni
~ Sn w
e.o
0
0
Mo
-0.2
- 0.4
-0.6
-0.8
-1.0
0 2 4 6 8 10 12 14
Alloying element, wt.%
Fig.11.6 Activity coefficients of nitrogen in Fe-N-j ternary liquid solutions at 1600°C, with iron
as the solvent (A.G. Ward, Physical Chemistry of Iron and Steelmaking, Edward Arnold,
London, 1962).
Now.
A~ 20
= - RT In {
PH2
2
PH s
x [/is]
}
eq
=- 1
RT ln ( -- )
10 [/ts] eq
(E.11 .22)
Again,
hs = fsWs = (10--0.ois x
08
· ) • 0.8 = 0.76
Putting the above value of hs in Eq. (E.11.22), we obtain
A~ = 31,190 J mo1- 1, ,1~ 1 = 690 - 31,190 = - 30,500 J mo1- 1
EXAMPLE 11.6 From experimental measurements, it has been determined that, for Si (pure, I)
= (Si]1 wt% SS in liq. Fe•
G~ = - 119.24 x 103 - 24T J mo1- 1 (E.11.23)

At 1600°C (i.e. 1873 K), rs; relative to pu;e Si as standard state is 0.0014 at Xs; = 0.0 l.
Calculate /si at Xsi = 0.01. Also calculate fiSt at 1 wt.% standard state of Si in liquid Fe,
at infinite dilution of Si.
(i) (E.11.24)
since 1 wt.% SS is, by definition, on Henry's Law line. Again, ~om Eq. (11.40), at Wsi = 1,
Xs·
'
= Ms;MFe =-56- =0.02.
x 100 2800
Solving Eq. (E.11.24), we get J1; = 1.25 x io-3. At Xsi· = 0 .01, /§i = 0 .0014; from
Eq. (11.50),
fi. = Ysi = 0.0014 =1 12

S1 11i 1.25 X 10-3 .
[Note: Here, Henry's Law is not obeyed at Xsi = 0.01 since/si::;; 1. Hence, obviously HL
would not be obeyed at 1 wt.% (i.e. Xsi = 0.02). This means that the 1 wt.% SS is not real
bur hypothetical.]
(ii) H-m
Si
-
- (E.11.25)
according to Gibbs-Helmholtz equation.

At infinite d.iJution, HL is obeyed and GS: is given by the Eq. (E.11.23). Hence,
fill! =
1 119.24 x 10 + 24.0T
T
3
1
= - 119 24 x 1<>3 J mo1- 1
& ~~) .
EXAMPLE 11.7 Liquid iron containing 2 wt. % silicon, 2 wt.% chromium, and 0.5 wt.%
vanadium was kept in a graphite crucible at 1600?C for a long time till equilibrium was
established between the liquid and the crucible. Calculate the wt.% of carbon in iron after
the experiment is over.
Given:
(i) Solubility of graphite in pure liquid iron = 5.4 wt.% at 1600°C
(ii) For Fe-C binary, log /~ = 4350 [l + 4 x 10-4(T - 1770)) (l - XFJ 2 (E.11.26)
T
(iii) e~i =0.1, ~r = - 0.024, ~ = -0.033
Solution T= 1600 + 273 = 1873 K

At graphite saturation, [ac] = 1 in liquid Fe, i.e., Xere = Xe 'if:Je = 1 (E.11.27)
(i) In binary Fe-C solution,
X _ Wc/12 5.4/12
c - We WFe = 5.4 100 - 5.4 = 0. 2 l
- +- - + - -- -
12 56 12 56
Noting that
4350
1 - XFe =Xe. log/cc= [1 + 4 x 10-4 (1873 - 1770)] (0.21)2 = 0.1067
1873
From Eq. (E.11.27), we have, yg

= 3.716.
(i) For liquid Fe-Cr-V-Si-C solution,
logfe = logf'f + ~i"'si + ~rwer + ~Wv

= Jogf'f + 0.1 x 2 - 0.024 x 2 - 0.033 x 0.5
=log~ + 0.1355 (E.11.28)
[Note: If in Fe-C binary and in the multicomponent solution are different since the value
of XFe will be different. Since the interaction effects of Si, Cr, V have been separated, from
a thermodynamic point of view, approximately speaking, (1 - XFe) can be substituted by Xe
in Eq. (E.11.26)]. Hence, from Eq.(E.11.26),
log/~ = 2.322[1 + 412 x 10-4] x2_ = 2.418 x2_ (E.11.29)
Combining all the above equations, we get
3 _716 Xe l0(2.418X~ +0.1355) =l (E. 11 .30)
Graphical solution yields Xe = 0.169. Again,
We/12
Xe = --------'=------- (E.11.31)
Wsi Wcr
- + - +We
- +Wv WFe
- +--
28 52 12 51 56
Since Wsi = 2, Wer = 2 and Wv = 0.5, we have
WFe = 96.5 - We (E.11.32)
Combining the above equations, and further graphical solution yields We = 4.2 wt.% after
crucible-melt equilibrium is attained.
11.6 Sieverts' Law

Sieverts carried out extensive measurements of solubilities of diatomic gases (H 2, N2 etc.)
in metals. Based on these, he proposed that solubilities of these gases are proportional to
square roots of partial pressures at a constant temperature. This is readily explained if we
recognize that all solutes are present in atomic state (not as molecules) in a metallic solution.
Therefore, dissolution of nitrogen follows the reaction
(11.56)
for which
K
h
= ([/;.]l112
PN2
(11.57)
Since solubility means concentration of N in metallic solution at equilibrium with a certain

value of PN2 , Eq. (11.57) is applicable. For binary Metal-N solutions, solubilities of nitrogen
is very small. Hence Henry's Law is applicable, and on the basis of Eq. (11.35), hN = WN·
Therefore,
(11.58)
11.7 Summary
1. This chapter contains discussions of some additional topics to supplement the earlier
chapters so that the readers have complete information for solution of aJJ types of
equilibrium calculations.
2. As an extension of Chapter 8, a further analysis of phase stabilities of reactive
stoichiometric compounds has been presented with the example of construction of
Si-C-0 ternary diagram. Wustite (FexO) is the best metallurgical example of a
nonstoichiometric compound. Thermodynamic analysis of this has been discussed in
the chapter.
3. Reaction equilibria calculations involving solutions of condensed phases have been
illustrated by solved examples.
4. In a multicomponent solution, components interact amongst themselves. For a dilute
solution, in connection with interactions amongst solutes, Wagner derived analytical
equation to quantify such interactions, and the equation is
Jn y,.I = Jn r? + I: X1e!
I . I
J
where Ii is activity coefficient of i in the multicomponent solution with solvent A

and solutes, l , 2, .. ., i, j, ... f; is Henry 's Law constant in binary A-i solution.
E( is the interaction coefficient describing the influence of j on r;: and is given as
el = (<5[1nY; ])
I DX
1 X1 --+ 0
Moreover, ~ =cf.
5. Weight% is commonly used as the composition parameter for practical applications.
In order to obtain answers of calculations directly in wt.%, the 'one weight percent
standard state' was proposed. It is based on Henry's Law (HL) as reference. On HL
line, in the binary A-B, where B is the solute,
Activity of B in l wt.% SS (h 8 ) = W8 , where W 8 is wt.% B.
At a composition where HL is not obeyed,
where / 8 = activity coefficient of B in l wt.% SS.

6. Use of l wt. % SS requires employment of a revised value of free energy of reaction
(~), which can be calculated from experimental values of L1G° and (i.e. HL yg
constant in binary A-8). Thermodynamic data tables are available for ~ most
extensively with liquid iron as solvent.
7. For multicomponent dilute solution , in 1 wt.% SS,
log f; = log /? + l: W.) elI

J
where
~ = (<5Q:.f;]) , and log/? = 0 by definition.

1 Wj-+0
PROBLEMS
11.1 A methane-hydrogen gas mixture at l atm. totaJ pressure is allowed to come to
equilibrium with a steel containing 0.60 wt.% carbon at 925°C: The gas was found
to contain 0.64% CH4 and the rest H 2 by volume. Calculate
(i) the activity of carbon in steel;
(ii) the expected analysis of a CO-C02 gas mixture which would be at equilibrium
with carbon in steel.
11.2 An alloy with 20 atom% Ni and 80 atom%. Au is allowed to react with a gas
mixture of H 20-H'.! at 1000 K to form pure solid NiO. At equilibrium the gas
contains 0.175% H 2 , 50% N'.!, and the rest H 20. Find the activity of Ni in the alloy.
11.3 A Cr-M solid solution contains 15 atom% Cr, and behaves ideally. Upon reaction
with a H '.!0-H'.! gas mixture, only pure solid Cr'.!03 forms. Calculate the Ptt/Ptt 2o
ratio in the gas mixture at equilibrium with the solid solution and Cr20 3 at 1000°C.
11.4 Consider the reaction between a liquid sulphide solution of Cu 2S-FeS (known as
Matte) and a molten slag of FeO-SiOrCu 2 0 at l250°C. Calculate the activity of
Cu 20 in the slag upon attainment of equilibrium. The matte contains 40 wt.% Cu 2S
and the rest FeS, and behaves as an ideal solution. The activity of FeO in the slag
may be assumed as 0.5 relative to pure liquid FeO.
11.5 Zinc, present as impurity in molten lead, can be selectively removed as ZnCl2 by
treatment with chlorine gas. Find the residual concentration of Zn in liquid Pb upon
attainment of the following reaction equilibrium at 663 K:
[Zn] + (PbCI 2) = (ZnCl 2) + Pb(!)
Given: (i) ZnCli-PbCl 2 forms an ideal liquid solution.
(ii) ..1G° for formation ~f PbCl 2 and ZnCl 2 are -257.5 kJ mo1- 1
and -320.7 kJ mo1- 1, respectively at 663 K.
(iii) XZnc12 in salt phase = 0 .983 .
0 2798
(iv) In [YZn Jin Pb= - -- 0.962
T
11.6 It is also possible to lower Zn content in liquid Pb by vacuum distillation. Calculate
the maximum pressure that can be tolerated in the vacuum system at 900 K for the
following conditions:
( i) Zn as only solute in liquid Pb, and Zn concentration is to be lowered to 0.01 atom%.
(ii) Liquid Pb contains 2 wt.% cadmium and 2 wt.% copper, besides Zn. The Zn
content is to be lowered to 0.3 wt.%.
Given: (i) ~~ = -0.04, ei;: = -0.07, ~~ =O in liquid Pb
(ii) Y~n in liquid Pb as in Problem 11.5.
11.7 An Fe-Mn solid solution containing XMn = 0.001 is in equilibrium with an FeO-MnO
solid solution and an oxygen containing gaseous atmosphere at 1000 K. Calculate:
(i) Composition of the oxide solid solution, and (ii) P0z in the gaseous atmosphere.
Both the metallic and oxide solution are ideal.
11.8 A mixture of solid CaO, MgO, 3Ca0.AI20 3 and liquid aluminium exerts an equilibrium
vapour pressure of magnesium of 0 .035 atm. at 1300 K. Write the equation for the
appropriate reaction equilibrium. Calculate the standard free energy of formation of
3Ca0.Al:!03 from CaO and Al 2 0 3, and the activity of Af 20 3 in CaO-saturated
3Ca0.Al 2 0 3 at 1300 K.
11.9 A liquid iron solution contains 0.5 wt.% silicon and 2 wt.% carbon. Temperature is
1900 K. Calculate (i) hsi> (ii) µ Si· Given: Y~i in binary Fe-Si solution = l.37 x 10- 3;
4 =0.18. ~i =0. ll
11.10 (i) In the binary Fe-Cu liquid solution at 1550°C with Fe as the solvent. it is gi ven
that:
log Ycu = 1.45 X~c - 1.86 X~e + 1.41 xte

Calculate the free energy change accompanying the change of standard state of
Cu from pure liquid Cu to 1 wt.% SS in liquid Fe.
(ii) What would be the corresponding free energy change for Mn in Fe-Mn liquid
solution at 1550°C, which behaves as an ideal solution?
11.11 An iron - l wt.% vanadium liquid alloy is equilibrated at 1900 K with a steam-
hydrogen mixture containing 5% steam. On analysis, the metal was found to concain
0.033 wt.% oxygen. Assuming that oxygen dissolved in molten iron obeys Henry's
Law, calculate ~.
11.12 An iron-titanium liquid solution with Fe as the sol vent is at equilibrium with solid
pure Ti02 and a H 20 -H 2 gas mixture having PH2olfJH2 ratio of 5.56 x 10- 3• Temperature
is 1600°C, and hTi = l. Calculate J4..
11.13 Calculate the residual oxygen content of liquid iron containing 0. I wt. % Si, at
equilibrium with pure solid silica at 1600°C. ~! = + 0.32, ~i = - 0.24, eg = - 0.20
(since eJ; can be calculated, it is not given).
11.14 When an iron-phosphorus liquid solution is equilibrated at 1900 K with solid CaO,
solid 3Ca0. P 20 5, and a gas phase containing P0i. = 10- 10 atm, the activity of P
in the iron wi th res pect to 1 wt.% ss in Fe is 20. Calculate 900 for the reaction:£.
5
P2(g) + -02(g) = P20 5(g)
2
Given: (i) 3Ca0 (s) + P20 5(g) = 3Ca0.P20 5 (s)
£. 900 = - 564,000 J mo1- 1
(ii) 112 P2(g) = (P)1 w1% SS in Fe
L1G° = - 122,200 - 19.2 T J mo1- 1
11.15 Calculate the chemical potential of nitrogen gas (i.e. N 2) in liquid steel at 1600°C.
The steel has the following composition:
WN = 0.01 , We = 0.5, Wp = 0 .2, WMn = 0 .5
188 1
Given: (i) (NJ1 wt% SS = K N /Ji..•2'
112 log K N -- - · -
-- l . 246
T
(ii) e~ = 0.25, e~ =0, e~ =0.051 , e~" = - 0.02
Chapt~ 12
Third Law of Thermodynamics,
Statistical Thermodynamics,
and Entropy
As discussed in Section 4.4, although Clausius proposed entropy in 1850, it took several
more decades to fully establish the fundamental significance of entropy qualitatively and
quantitatively from the atomistic viewpoint. Early work leading to the Third Law of
Thermodynamics was done by T.W. Richards (1902). W. Nemst (1906) generalized these
findings and formulated the statement of the Third Law, which he called as Heat Theorem.
Subsequently, the statement was refined by Max Planck. The Third Law provided some
atomistic interpretation of entropy qualitatively. It specifically dealt with the behaviour of
entropies and specific heats of substances at absolute zero of temperature (i.e. T = 0).
The subject of Statistical Thermodynamics started developing from the late
19th Century and matured in the decade of 1930s. It is a major subject in its own right, and
provides quantitative relations of internal energy, entropy etc. from atomistic considerations.
However, in common applications in the metals and materials science and engineering, it is
not widely employed. Hence, in this chapter, only introductory concepts and a few quantitative
relations would be briefly presented, primary attention being devoted to interpretation of
internal energy and entropy.
12.1 The Third Law of Thermodynamics
12.1.1 Nernst Heat Theorem

Based on experimental evidence, Nemst postulated that for chemical reaction between pure
solids or liquids, at temperature approaching absolute zero (i.e. T ~ 0),
213
(12.1)
Now,
which is the same as Eq. (6.27). Hence, from Eq. ( 12. 1) and (6.27),
$ ~ 0 as T ~ 0 (12.2)
Again,
(6.3)
• for isothermal processes. Differentiating Eq. (6.3) with respect to T at constant P, we get
(12.3)
At T ~ 0, combining Eqs. (12.1)-(12.3), we obtain
[o<;:o)t ~ o (12.4)
Now, from Eq. (4.34),
(12.5)
where
~=L~-L~ (3.26)
product reactant
Hence, from Eqs. (12.4) and (12.5), for reactions involving pure solids and liquids,
~ ~ 0, asT ~ 0 (12.6)
In accordance with Eq. (12.2), Nernst's Heat Theorem states that "for a\\ reactions
involving pure substances in condensed state, .dS is zero at the absolute zero of temperature".
12.1.2 Entropy of a Pure Substance
A pure substance is either a pure element or a pure compound. Consider the formation of
a compound AB from elements A and B, i.e.
A(s) + B (s) = AB(s) (12.7)
Third Law of Thermodynamics, Statistical Thermodynamics, and Entropy
for which
$:~B-~-~ (12.8)
At absolute zero temperature, LiS° = 0. It is possible provided either
(i) ~B =~ +~
or
(ii) ~B• ~. ~ are all individually zero.
From a probability point of view, the first alternative is very unlikely as a general feature.
It may be true, by chance, in a few cases. Hence, alternative (ii) is accepted as of general
=
validity, and it may be concluded that entropies of pure solids at T 0 are zero. Since all
substances are solids at T =
0, we are omitting liquids. Therefore, the absolute value of
entropy of a pure substance at any temperature can be determined by taking T 0, SJ 0= =
as the lower limit. This is in contrast to energy, whose absolute value cannot be found out
within the scope of conventional thermodynamics. This is the basis on which we find as to
how thermochemical data tables provide values of entropy for pure substances. Figure 12.1
presents variation of SJ with temperature for a few substances.
Figure 12.1 shows some increase in entropy during melting. In this connection, it is
worthwhile to note down the following rules which were proposed long back from experimental
observations.
200
~
0 150
e
;::::;
50
Graphite
Diamond
0-=..~--'-~~-'-~~L-.~-L~~-'-~--1.__J
300 500 700 900 1100 1300 1500
Temperature, K
Fig. 12.1 Variation of molar entropies of some elements and compounds with temperature.
(i) Richard's rule: ~ :::: 10 - 20 J K- 1moi-1, for melting
(ii) Trouton 's rule: ~ "" 100 J K- 1mo1- 1, for vaporization

These rules are empirical guidelines and are approximate only. They have been explained
from atomistic viewpoint. However, it is being omitted for the sake of conciseness.
12.1.3 Fundamental Significance of The Third Law
After Nernst, Max Planck modified Nernst's statement as follows:

"The entropy of any homogeneous substance, which is in complete internal equilibrium,
may be taken as zero at 0 K '.
As stated in section 4.4.2, one of the interpretations of entropy is that it is a measure
of disorder in a substance. Its concrete proof is based on deductions in statistical thermodynamics,
and will be presented in Section 12.3. Right now, we shall assume it to be correct and
proceed further.
When disorder in a substance completely vanishes, it becomes perfectly ordered and has
only one possible arrangement of atoms. Then and only then a complete (i.e. absolute)
internal equilibrium is established. The causes for deviation from perfect order are:
1. Thermal energy and resulting motions of atoms and molecules
2. Noncrystallinity, as in amorphous solids and liquids
3. Crystal defects- vacancies, interstitials, dislocations
4. Disordered compound structure
5. Disordered solid solution.
At T = 0, thermal energy is zero. Thus this source of dis9rder gets eliminated. But other
sources may remain. Hence, in reality, it is difficult to have entropy exactly zero at T = 0.
However, for pure crystalline, homogeneous substances, entropy at T = 0 is fairly close to
zero.
Actually, all disorders are caused by thermal energy and, hence, exist at normal and high
temperatures. As temperature is lowered, the disorders should decrease continuously and
vanish at T = 0, provided internal equilibrium exists. However, at low temperatures, the rates
are very low. This causes lack of internal equilibrium. Disorders of higher temperatures, so
to say, get 'frozen' at lower temperatures.
12.1.4 Experimental Verification of the Third Law
There are three general methods of evaluation of entropy of a substance:
(i) The Third Law method. Thi s consists of integration of experimental heat capacities
as function of temperature from 0 K to T, assuming SJ to be zero at T = 0. From
Eq. (4.38),
50 =1Tm ~(s) dT + ~ + JTb ~(I) dT + ~ + JT ~(g) dT (12.9)

T 0 T m Tm T v Tv T
This is the most general form of the equation which takes into account transformation
of the solid into liquid and gas. If at temperature T, it remains a solid, then
Eq. (12.9) gets simplified into:
(12.10)
(ii) The spectroscopic method. In this method, we evaluate entropy of an ideal gas
from spectroscopic data. The procedure for calculation is based on statistical mechanics.
(iii) The Second law method. This consists of evaluation of LIG° of reactions over a
range of temperature. &-' is evaluated from variation of LIG° with temperature on
the basis of Eq. (6.27).
Comparison of predictions of these methods provided experimental verification of the
Third Law. Comparison of results of all three methods is not common. One example is
entropy of chlorine gas at room temperature (i.e. 298 K).
Method S' per mol Cl 2, J mo1-1K-1

Third Law 223 .2
Spectroscopic 223 .2
Second Law 221.9
The close agreement provides verification of the Third Law. It may be noted that the
agreement between the Third Law and the spectroscopic method is good. Similar very close
agreements have been found for several ideal gases by these two methods.
The oft-quoted illustrations of experimental verification of Third Law are through
studies on the following allotropic transformations:
(i) Transformation of monoclinic sulphur into rhombohedral sulphur at 368.5 K
(ii) Transformation of white tin into gray tin at 292 K.
The procedure of verification based on the first case is illustrated in Fig. 12.2. Entropy
is a state property and independent of path. But changes of entropy can be found out only
by following reversible paths (see Section 4.3). For Fig. 12.2,
(12.l l)
So,
(12.12)
The experimental values are:
~ = 1.093 J mo1- 1K- 1
~ + L1.S?. 1 =0.963 J mo1- 1K- 1

II=--
Monoclinic Ttr Rhombohedral
,.
sulphur sulphur
Ttr = 368.5 K
f
e::s
......
..."'
G)
E
0.
G)
f-e
&f
G j\ ~l
OK -
L'.1.S?v
Fig. 12.2 An example of experimental verification of the Third Law of Thermodynamics.
Hence,
L1Sfv = entropy of transformation at 0 K = 1.093 - 0.963 = 0.13 J mo1-1 K- 1 = O

This is the verification of the Third Law within experimental errors.
12.2 Introduction to Statistical Thermodynamics
12.2.1 Kinetic Theory and Ideal Gas
In traditional mechanics, one form or the other of Newton's laws is employed to analyze the
dynamics of system, and to follow motions of individual particles or elements (e.g. fluid
elements) in a system as function of time. However, there are systems which consist of a
large number of particles in motion and are too complex to analyze in terms of individual
particles. These systems usually appear disordered because of the large number of particles
involved, and the innumerable ways of sharing energy of the particles amongst themselves.
Thermodynamics deals with such systems more easily.
The basic laws of thermodynamics deal with relationships amongst macroscopic variables,
such as pressure, temperature, volume and internal energy. Such relations are simplest for
ideal gases. These laws and relations do not, however, consider the behaviour of atoms and
molecules which constitute a system.
The alternative approach to explain these macroscopic thermodynamic variables is the
so-called Kinetic Theory. It developed gradually from the 17th Century and matured jn .the
19th Century. It was successfully applied to ideal gases on the postulate that the atoms/
molecules of a gas are always in motion due to their thermal energy. The thermodynamic
variables are related to this and are macroscopic averages. Newtonian mechanics is the basis
of dealing with the velocities and kinetic energies of individual molecules.
The equations derived by the kinetic theory of ideal gases are the following:
1-
Pressure =P = - 2
pu (12.13)
3
where p is density of gas and u 2 is mean square (i.e. rms) of linear velocities of the
molecules. Again,
(12.14)
where M is molecular mass. That is, the average translational kinetic energy per mole of an
ideal gas is proportional to its temperature. Dividing Eq. (12.14) by Avogadro' s number
(N0), we get
I - 3
- mu2
2
=-k
2 B
T (12. 15)
where m is mass of one molecule, and ks =

R/N0 =
Boltzmann's constant.
The internal energy of a monatomic ideal gas was assumed to be only due to translational
motion of molecules. Hence, for I mole of gas, from Eq. (12.14),
3
U= -RT (12.16)
2
Classical statistical mechanics (section 12.2.2) arrived at the Principle of Equipartition
of Energy amongst all independent degrees of freedom. In a monatomic gas, there are three
degrees of freedom of translational motion along the three perpendicular coordinates. Hence,
each is associated with energy of 112 RT per mole. In a diatomic gas, there are five degrees
of freedom (viz. three translational and two rotational). Hence,
I 5
U =5 x - RT= - RT
2 2
(12.17)
12.2.2 Statistical Mechanics
Classical formulations of statistical mechanics were done in the late 19th century by Maxwell,
Gibbs and Boltzmann. The kinetic theory · predicted average translational kinetic energy of
a gas. However, it could not predict distribution of speed and energy amongst molecules. We
take recourse to statistical mechanics for this purpose. Besides mechanics, it is founded on
the mathematical science of probability and statistics, an emrging field at that period.
Maxwell first solved the problem of the distribution of speeds in a gas containing a large
number of molecules in a box. The Maxwellian speed distribution for a sample of gas at
temperature T containing N molecules, each of mass m is
2
n(u)=4trN(-m-)3f2 u2 exp (- mu ) (12.18)
2trk8 T 2k8 T
where ri(u) is the probability distribution function of speed, n(u). du is the number of
molecules in the gas sample having speed between u and u + du. Therefore,
0
N = fo° n(u) du (12.19)
Figure 12.3 presents Maxwellian distribution of n(u) as function of u at T = 300 K. The

distribution of speeds of molecules in a liquid also resembles the curve in Fig. 12.3.
However, very few molecules have large enough speed to break bond and escape as vapour
molecules. It is also interesting to note that Maxwell's equation could be experjmentally
verified almost 100 years later (1955), when sufficiently precise experimental techniques
could be employed.
2.0
E
......
,..,"'
-
0
x
,,......
1.0
::t
-.._;
s:::
200 400 600 800 1000 1200

Speed u, mis
Fig. 12.3 Maxwellian speed distribution for the molecules of oxygen gas as function of u at
T = 300 K. The number of molecules with speeds in any interval du is n(u)du. The
number with speeds between u1 and u2 is given by the curve between these limits.
Following the derivation of speed distribution, energy distribution function was subsequently
derived as
n e - -2N
( )- ~
I
(k5T)312
e12 exp ( -ksT
e )
- (12.20)
where n(e) ·de is number of molecules with energy between e and e + de.
U =J: n(E)dE (12.21)
upon integration of Eq. (12.21) by combining it with Eq. (12.20), we get
U =-3 N0 k8 T =-3 RT for 1 mole of gas

2 2
which is the same as Eq. (12.16) derived from the kinetic theory of gases.
12.2.3 Quantum Mechanics and Statistical Thermodynamics
Internal energy and energy distribution in an ideal gas were successfully predicted by
statistical mechanics. Here the molecules were free and moved like distinct particles. However,
in solids and liquids, the atoms/molecules are close and interacting. This gives rise to
complex interaction and energy exchanges between them. In this, the electrons also participate.
It required development of the subject of quantum mechanics to tackle such situations.
In the late 19th Century, there were two empirical laws of radiation, viz. Wein's Law
and Rayleigh-Jean 's Law. They could not be reconciled on the basis of the wave nature of
radiation. In 1901, Planck reconciled them by postulating that radiation is emitted discretely
as energy packets. It was a revolutionary concept. In 1905, Einstein caJled these packets as
quanta. Thus Planck's theory came to be known as Quantum Theory.
The postulate of the quantum theory can be stated as follows: "If a particle is confined
to move within a fixed volume, then its energy is quantized, i.e. discretized". Application
of quantum theory later explained the structure of an atom by assuming that electrons can
only be at some discrete energy levels.
On the basis of quantum theory and the techniques of statistical mechanics, the subjects
of quantum mechanics and statistical thermodynamics were developed by several 20th Century
physicists. An important postulate is that all particles are exactly alike (i.e. indistinguishable
from one another).
12.3 Basic Concepts and Relations in Statistical

Thermodynamics
12.3.1 Basic Concepts and Postulates

In classical thermodynamics, we deal with Macrostates of substances. According to statistical
interpretation of thermodynamics, each macrostate consists of a very large number of Microstates
(also known as quantum .states).
Chemical and physico-chemical equilibria are dynamic in nature, and are attained when
the forward and backward processes have equal rates, thus balancing each other and resulting
into zero net rate. There are always fluctuations in microscale. But these fluctuations are not
perceptible for a large assembly of molecules. Macroscopic quantities (pressure, temperature,
energy etc.) are only averages. Such averaging is to be done only statistically. For such
statistical analysis, the basic postulates of statistical thermodynamks are as follows:
1. Each of the accessible and distinguishable quantum states (i.e. each microstate) of a
system of fi,xed e11ergy is equally probable.
2. The equi ibrium state corresponds to the most probable macrostate. It is a state
which contains the largest number of microstates.
3. As has already been stated, all particles are exactly alike.
Equation (12.20) provides the distribution function for energy of molecules in ideal gas.
It is known as Maxwell-Boltvnann (MB) statistics.
On the basis of the above postulates (i.e. utilizing quantum theory), the following two
energy distribution functions at the most probable macrostate were derived later by physicists
whose names they bear
1. Bose-Einstein quantum statistics
2. Fermi-Dirac quantum statistics.
Each of these has played a very important role, not only in statistical thermodynamics
but also in quantum mechanics quantum physics.
All the above three statistics are applicable to systems, where mass, composition, internal
energy and volume are fixed. They differ in some other assumptions, the discussion of which
is beyond the scope of this text.
12.3.2 The. Most Probable Macrostate

As already stated, the most probable macrostate corresponds to the ~quilibrium state with
which classical thermodynamics is concerned with. This state corresponds to the largest
number of microstates, which means the largest possible number of arrangements of particles
(W) amongst the different discrete energy levels of the system. In other words, at equilibrium,
W = Wmax.
The mathematical procedure of derivation of Wmax is illustrated now. Let n0 , n" ... , n;,
••• 11r be the number of particles at energy levels eo.
E1, ... E;, .... er. Then,
Total No. of particles = no + n 1 + ... + n; + ... + nr = .tn. = N = a constant

.
I
I
(12.22)
From mathematics, the number of .ways N particles can be arranged amongst the various
energy levels are given as:
(12.23)
where '!' is the symbol for factorial .

Since a system consists of a large number of particles (i.e. atoms etc.), say of the order
of Avogadro's number, Stirling 's approximation is valid and, accordingly,
In W = (N In N - N) - .t (11; In n ; - 11;)
;
(12.24)
or
d(ln W) = -.t(ln
I
n; dn1 + !i. dn;
n
- dn;) = - ~In n; dn;
1
(12.25)
1
since N = a constant. Again,

d(ln W) =0 at W = Wmax (12.26)
12.3.3 Internal Energy and Entropy at Most Probable Macrostate

Internal energy of the system of section 12.3.2 is given as
U = 'lot-0 + n1E1+ ... + n;E; + ... + n,£,. = ~ n;&i (12.27)

'
Since U is assumed to be a constant,
d U = ~ £i dn1 + ~ ll; d E1 = 0 (12.28)

' '
Since tQ, £ 1... , E;, ... Er are all constants (i.e. fixed discrete energy levels as per quantum
theory)
.tn.d£.. =0 (12.29)
.
I ' '
dU = LE1 dn; =0 (12.30)

'
Again from Eq. (12.22), since N is a constant,
dN = I:.dn.
j I
=0 (12.31)
Combining Eqs. (12.25), (12.26), (12.30) and (12.31), we may write
.t(ln n1 + a + /k;) dn1 =0 (12.32)

I
where a and f3 are multipliers.

Equation (12.32) is satisfied by substituting the following relation into it (detaHs skipped)
at the most probable macrostate (i.e. equilibrium state):
Nexp( - ~)
n =- - --'--'- ( 12.33)
' p
'**'
' Textbook of Materials and Metallurgical Thermodynamics
1
~ where
'
(12.34)
'-'
l P = partition function = l: exp ( - /3E;)

i
(12.35)
The above ec ·1ations give a distribution function of n; as proportional to exp (-E/ k8 7),
and is analogous to Maxwell-Boltzmann distribution function [Eq. (12.20)].
It can be derived further (derivation skipped) that, at the most probable macrostate,
In W=N lnP+-
u (12.36)
k8T
or
;
(12.37)
since both N and P are constants.

At constant volume, from Eq. (2.18),
dU =liq (12.38)
where q is heat flow from surrounding into the system. Hence,
. 8q
".
d(ln W) =- (12.39)
k8 T
At constant temperature and for reversible processes, from Eq. (4.31),
oq =dS (12.40)
T
Combining Eqs. (12.39) and (12.40) and integrating, we get
S = k8 In W + constant (12.41)
From the Third Law of Thermodynamics, S = 0 at T = 0 for a perfectly ordered solid

(see section 12.1.3). As mentioned there, perfect order means only one unique arrangement
of particle in the system, i.e. W = I (i.e. In W = 0). Therefore, the constant in Eq. (12.41)
drops out and,
S = k8 In W (12.42)
This equation was proposed by Planck in 1906, and is known as Boltvnann equation.
It relates entropy with the number of arrangements of particles amongst different energy
levels available in a system of fixed mass, volume, number of particles, as al so of fixed ·
'
internal energy and composition.
12.3.4 Entropy of Mixing for a Binary Solution
Consider mixing of atoms A and B to form a binary solution. The process may be represented
as
A + B (unmixed) ~ A + B (mixed) (12.43)
(state 1) (state 2)
Integral molar entropy of mixing is
Li.Sm = S2 - S1 = (Sth + Sconch - (Sth + Sconf)1 (12.44)

where subscripts 2 and 1 denote state 2 and 1, respectively.
Entropy is a measure of disorder, qualitatively speaking, and such disorder in the
solution primarily originates from
(i) motion and vibration of atoms due to their thermal energy, entropy from this source
being known as thermal entropy (Sm); and
(ii) arrangements of atoms in various lattice sites, entropy from this source being known
as co11figurational entropy (Sconf).
Assuming (Sm)2 = (S1J1> combination of Eq. (12.44) with Eq. (12.42) yields
(12.45)
In unmixed state, configurations of both A and B are unique since they exist separately.
Hence, Wconf,t = 1 and, therefore,
(12.46)
Let us take 1 mole of solution. The total number of atoms is Avogadro's number (N0 ).
Assume the total number of sites = N0 , and also assume the number of A and B atoms as
"A and n8, respectively. From the theory of combination, nA atoms can be arranged in No
sites in NOc ways assuming random mixing of A and B atoms. Once A atoms are arranged,
DA
B atoms occupy the remaining sites (i.e. one arrangement only). Therefore,
(12.47)
From Eqs. (12.46) and (12.47),
(12.48)
Employing Stirling's approximation,
(12.49)
Noting that nA + n8 = N0 , rearrangement of Eq. (12.49) leads to
(12.50)
Since nAIN0 and 11 8 1N0 are mole fractions of A and B (i.e. XA and X8 ), and k8 N0 = R,
Eq. (12.50) may be rewritten as
(12.51)
Equation (12.51) is identical with Eq. (9.52) for ideal binary solution. Hence, it may
be concluded that an ideal solution is characterized by
(i) random mixing of atoms/molecules, and
(ii) no change in thermal entropy when the solution forms from pure components.
12.4 Summary
1. Max Planck stated the Third Law of Thermodynamics as
The entropy of any homogeneous substance, which is in complete internal equilibrium,
may be taken as zero at 0 K.
This is valid only for a perfectly ordered, pure, crystalline solid.
2. The Third Law was originally proposed on the basis of experimental measurements
at temperatures close to 0 K. It was subsequently verified for several substances
systematically from experimental data.
3. For ideal gases, the following equations were derived on the basis of the kinetic
theory:
Pressure P = 1/3 pU2

where p is density of gas, and u2 is the rms of linear velocities of molecules.
U = 312RT, for a monatomic gas due to translational motion of molecules.
4. Subsequently, distribution functions for speed and energy of molecules in ideal gas
were successfully derived on the basis of statistical mechanics.
5. In condensed phases, atoms/molecules are close and mutually interacting, in contrast
to ideal gases where they are free. Energy distributions could be derived in these
only after the advent of the Quantum Theory. The subjects of Quantum Mechanics
and Statistical Thermodynamics, then, were developed on the basis of statistical
mechanics and quantum theory.
6. Some basic postulates and relations of statistical thermodynamics, such as microstate,
macrostate, most probable macrostate etc., have been briefly presented in this chapter.
7. At the most probable macrostate, which corresponds to equilibrium state in classical

thermodynamics, it has been derived that:
(i) n. = N exp (-,&i)

I p
where p = llkaT, P = partition function = I:i exp (-/JE;),

where ni is number of particles at energy level 4, and N is total number of particles.
(ii) S = ks In W
where W is number of ways particles in an assemblage can be arranged.
8. Finally, entropy of mixing of a binary solution A-B has been derived on the basis
of above as
L1Sm = - R(XA In XA + Xa ln Xa)
This is the same as that of an ideal solution. Assumptions in derivation are random
mixing, and no change of thermal entropy upon mixing.
:.
Chapter 13
Thermodynamics of
Electrochemical Cells
13.1 Introduction
An electrochemical reaction involves coupling of chemical reaction with flow of electric
current. It is a very important field in metaJlurgy and materials science, as the following
examples will illustrate:
• Many metals are extracted and/or refined by electrolytic process (Zn, Al, Mg etc.).
• Electroplating and anodizing are widely employed for surface protection of metals
and alloys from corrosion.
• Electrochemical reactions occur in corrosion, hydrometallurgy, and slag-metal reactions.
• Batteries are electrochemical cells.
• The electrochemical method is an important tool for thermodynamic measurements,
especially at high temperatures.
Electrochemical cells are broadly classified into:
(i) Galvanic cells (i.e. batteries), where stored chemical energy is converted into electrical
energy
(ii) Electrolytic cells, i.e. cells for electrolysis, where electrical energy is used to do
chemical work.
Thennodynamic studies/predictions/measurements can be properly done only for galvanic
cells since these can be made to operate reversibly. It may be recalled that the term 'galvanic'
originated from the famous Italian scientist Galvani, who was a pioneer in this field.
228
Thermodynamics of Electrochemical Cells
13.1.1 The Daniel Cell

The most common example of a galvanic cell is a Daniel cell, shown in Fig. 13.1.
Overall reaction: Zn(s) + CuS04 (aq) = Cu(s) + ZnS04 (aq) (13.l)
where "aq" means aqueous solution.
Cu electrode
Porous
diaphragm
Insulated
container
0 0 0 0
000000
Solid ZnS04 Solid CuS0 4

particles particles
Fig. 13.1 Sketch of a Daniel cell.
Figure 13. 1 shows that both CuS04 and ZnS04 solutions are saturated in a Daniel cell.
The overall reaction may also be written in ionic form as
Zn(s) + Cu 2+(aq) = Cu(s) + Zn 2+(aq) (13 .2)
Reaction (13.2) consists of:

Anodic reaction: Zn(s) = Zn 2+(aq) + 2e- (13.3)
Cathodic reaction: Cu2+(aq) + 2e- = Cu(s) (13.4)
The aqueous solutions constitute the electrolytes in the Daniel cell. Electrolytes are ionic
liquids or solids and may be classified as given in Table 13.1, along with examples of some
ions, which are mobile and carry electric current in the electrolytes.
It is to be emphasized here that an electrolyte may have several current carrying ions,
but not all of them participate in electrochemical reaction. For example, during electrolysis
of aqueous ZnS04 solution, some H 2S04 is added to improve its electrical conductivity.
Although H+ will carry the major fraction of current through the aqueous solution due to
its high mobility, the reaction at cathode would be primarily discharge of Zn 2+ and deposition
of Zn. Similarly, a molten salt at high temperature may consist of NaCl, KC! and CdC1 2,
but it is only Cd2+ which will participate in electrochemical reaction.
Table 13.1 Classification of Electrolytes
Electrolytes Examples of current carrying ions

Aqueous solutions Na+, H+, c1- etc.
Molten salt., Na+, K+, Cd 2+, c1-, r etc.
Molten sla?<> Ca 2+, Mn 2+ etc.
Solid oxidl electrolytes 02-
Other solid electrolytes
You are fami liar with aqueous solutions and their chemical and physico-chemical behaviour.
Electrochemical cells with aqueous electrolytes are operated around room temperature. In
contrast, all other electrolytes listed in Table 13. l are for cells operating primarily at higher
temperatures. It is difficult to specify the exact temperature ranges for them. Some approximate
idea only may be given. Molten salts typically consist of solutions of chlorides or fluorides
and are suited to a temperature range 400-900°C. Molten slags consist of solutions of metal
silicates or borates, and are ideal for electrolytes in the temperature range of 1000-1500°C.
Solid electrolytes (SE) are developments in the last 50 years. Electrochemical cells with
these are also operated primarily at high temperatures, the ranges varying with the nature of
the electrolyte. A large number of such electrolytes have been developed. The best known
and the most widely investigated one is stabilized zirconia. Zr02 is an excellent high temperature
ceramic material. But it undergoes some phase transformations during heating and cooling.
This causes thermal stresses and consequent failure in service. Addition of oxides such as
CaO, MgO, and Y20 3 stabilizes the cubic fluorite (i.e. CaF2) structure from room temperature
upto its melting temperature (approx. 2400°C) and thus prevents failure.
Let us take the example of Zr02-Ca0 solid solution. In pure Zr02 , Zr4+ ions occupy
cationic lattice sites and 02- ions anionic sites. On addition of CaO, divalent Ca 2+ occupies
some cationic sites in place of tetravalent zr4+. For maintainance of local electrical neutrality,
replacement of a zr4+ ion by one Ca2+ ion leads to removal of one 0 2- from anionic site.
This creates an oxygen ion vacancy and allows movement of 0 2- ion by vacancy mechanism.
Thus, 0 2- is the only current carrying species since Zr4+ and Ca2+ are almost immobile.
However, since the material is solid, a reasonably large mobility of 0 2- can be attained only
above 700-800°C. Thus the cell is to be operated only at high temperature.
For thermodynamic measurements, ZrOz-CaO solid electrolyte has been widely used.
Other important ones are molten salts· based on KCI and Li Cl mixture as base electrolyte
(as H 20 is base for aqueous solutions), and SEs such as Zr02-Y 20 3, "ThOr Y 20 3, CaF2 doped
with YF3.
13.2 Thermodynamics of Reversible Galvanic Cells

In practice, the same electrochemical cell can be made to behave either as a galvanic cell
or as an electrolytic cell. Let us take the example of Daniel cell (Fig. 13. 1). Let us connect
it to an external DC source such that the externally imposed voltage (Vext) opposes that of the
Danjel cell (Vceii). ·If Vext < Vcell> then the cell behaves as a galvanic cell with reaction (13.1)
occurring as shown. However. if Vex•> Vcell> then the direction of reaction (13. 1) is reversed,
i.e. the reaction becomes
Cu(s) + ZnS04(aq) = Zn(s) + CuS04(aq) (13.5)
In this situation, external electrical energy is being consumed to force reaction (13.5) to take
place, and the cell behaves as an electrolytic cell.
In Section 2.3, we have discussed the concept of a reversible proce~s. We may recall that
(a) it is very slow and (b) it occurs near equilibrium, and the process can be reversed by
shifting it from equilibrium in the other direction.
In an electrochemjcal cell, chemjcal reaction occurs only if the current is allowed to flow
through the cell. Ideally, no current flows and, hence, we have cell equilibrium, if the circuit
is open. The voltage at thi s condition is known as Electromotive force (i .e. emf) of a
galvanic cell. However, this emf can not be measured without a voltmeter. Use of a voltmeter
of very high resistance makes the current flow negligible and, hence the cell behaves as a
reversible galvanic cell. Another method is to impose voltage from an external source such
that Vex• is very close to the cell emf, and current flow is negligible.
13.2.1 Relation between Cell EMF(£) and Free Energy of Cell

Reaction (LIG)
For a reversible process, from Section 5.1,

dG = VdP - SdT- OW' (5.10)
where W' is any form of work other than that done against pressure. In this chapter, W'
would refer to electrical work performed by the galvanic cell.
At constant temperature and pressure,
(dG)T.P = - OW', i.e. (LlG)T.P = - W' (5.48)

As discussed in Section 5.5, here W' is also known as maximum work, since work is
maximum for a reverc;ible process only.
Since in a reversible galvanic cell the voltage across the cell is its EMF, we have
OW'= E/fq (13.6)
where E = cell emf and /fq = infinitesimal quantity of electrical charge transferred across
the cell due to chemjcal reaction in the cel l.
Again, from Faraday's Laws of electrolysis,
/fq = ZFdn (13.7)
where
Z = valency involved in the chemical reaction
F = Faraday's constant = 96,520 joules/volt/gm-equivalent

dn = an infinitesimally small quantity of species (in gm.moles) electrochemically transferred
between electrode and electrolyte as a result of 8q quantity of charge.
Combining Eq. (5.48) with Eqs. (13.6) and (13.7), at constant T and P, we get
dG = - 8W' = - ZFE dn ( l 3.8)
For a finite proc..ss, if n = 1 gm-mole, then

LlG = -ZFE (13.9)
If the species transferred between electrodes and the electrolyte are at their respective standard
states, then we write Eq. (13.9) as
(13.10)
where !fJ is known as the standard emf of the cell .
Agai n, from Section 7 .3,
LlG = L1G° + RT In J (7.24)

where
J = activity quotient.
From Eqs. (13.9), (13.10) and (7.24),
RT
E=f!J - - Inf (13.ll )
ZF
13.2.2 Sign Convention for EMF
We need not sketch an electrochemical cell in order to describe it. There is a standard
convention for its short representation. Accordingly, for example, the Daniel cell may be
represented as
Zn(s) I ZnS04 (aq) II C uS04Caq,. I Cu(s) (13.12)
(sat) (sat)
Single vertical li nes separate electrode from electrolyte. The double vertical line means
porous diaphragm.
An important question is: What should be the sign of the emf of the above cell-positive
or negative? For this, some convention is required. There is the European convention and
also the US convention. We shall adopt the US convention. Accordingly, if the reaction
proceeds spontaneously from left to right, the sign of E is positive. For the Daniel cell,
reaction proceeds spontaneously in the forward direction of Eq. (13.1) if the circuit is closed ,
i.e. Zn dissolves as ZnS04 and Cu is deposited at cathode from CuS04 . Therefore, in the
representation of Daniel cell (13. 12), reaction proceeds spontaneously from left to right, and
hence the emf is positive.
13.2.3 Cell EMF and Chemical Potent ials at Electrodes

In Section 10.1, we have disc ussed chemical potential. It has presented the general statement
of equality of chemical potentials of components in phases at equilibrium. If there are only
two phases (I and II) at equilibrium, then at constant T and P,
I_ II I_ II I_ IJ
µ. - µ. ' J1i - J.11. .... , µ; - µi ' ... (10.18)
where the subscripts denote components in solution (elements or compounds).

In an electrochemical cell, there is electrical work also. Let us consider the two electrodes
as phases I and II. Then each electrode will have not only chemical potential, but also
electrical potential. Let us consider exchange of component i between electrodes I and II.
Then Eq. (10.14) is to be modified as follows:
(13.13)
dG'· n = Jlin dn;u + ZF<P" dn;u = ()1f + 7Ff) dn;" (13.14)
q} and q}' are electrical potentials of electrodes I and II respectively. (ZFl/J) dn; represents
the electrical work to be performed for transfer of dn; moles of component i. The quantity
{µ; + ZF¢) is the sum of chemical and electrical potential energy per mole of i, and is known
as the electrochemical potential of component i.
Proceeding further as in Chapter 10, for exchange of component i between the electrodes,
we have
dn,n = - dn~I = - dnI ( 13.15)
( 13.16)
The two electrodes are at electrochemical equilibrium with one another (not chemical
equilibrium) when there is no current flow (i.e. cell voltage =emf= E). Under this situation,
dG' = 0. Therefore, from Eq. (13.16), at constant T and P,
(13.17)
In other words, the electrochemical potentials are equal at the two electrodes for reversible
exchange of component i between the electrodes. Again, under this condition,
q} 1 - ¢1 = cell emf = E (13.18)
µ.l 1 - µ: = -ZF(¢ 11
- ¢1) = -ZFE (13.19)
Equation (13.9) relating E with L1G is of general applicability to any emf cell.
Equation (13.19) is applicable only to concentration cells. Examples of both will be presented
in the following section.
13.2.4 Temperature Coefficient of EMF
Let us recall the following equations from Sections 6.2 and 6.3,
L1G = Llli - TL1S (6.4)

for an isothemzal process. Also,
[oc:')l = -& (6.27)
Combining the above equations with Eq. (13.9), we obtain
L1S =ZF(~)P (13.20)
Llli =- ZFE + zn{~l (13.21)
13.3 Examples of EMF Cells

Electrochemical cells may be classified into:
(i) reaction cells, where the source of cell emf is a chemical reaction,
(ii) concelltration cells, where the source of emf is the difference in concentration of a
component i at the two electrodes.
13.3.1 Examples of Reaction Cell

EXAMPLE 13.1 Find out F!1 for the Daniel cell at 298 K. Also express E as function of
activities if the electrolytes are not saturated solutions.
Given: LtG° for reaction (13.1) is -213.0 kJ rno1- 1 at 298 K.
Solution From Eq. (13.10),
103
F!1 = - £ = - C- 213 x ) = 1.104 volts (E.13.1)
ZF 2 x 96520
From Eq. (13.1 1),
8 314 298
E =1.104 - · x In J = 1.104 - 0.0128 In J (E. l 3.2)
2 x 96500
where
Cacu l Ctiznso )
J = - - - - ' -4- (E. 13.3)
[az~ l Cacuso4 )
Thermodynamics of Electrochemical Cells fih1
acu and az11 are activities of Cu and Zn at respective electrodes. Since they are pure, acu =
a Zn= 1. In a Daniel cell, the electrolytes are saturated solutions. It means that the ZnS04
solution co-exists at equilibrium with solid ZnS04 , and the CuS0 4 solution with solid
CuS04 . In other words, in Daniel cell, the activities of ZnS04 and CuS04 in the electrolytes
are also 1. However, in unsaturated electrolytes, the activities in electrolytes will be less
than l. Hence, from Eqs. (E.13.2) and (E.13.3), at 298 K,
E = 1.104 - 0.0128 In (llznso4 ) (E.13.4)

(£lcuS04)
EXAMPLE 13.2 At 1300 K, the emf of the following cell is 0.60 V.

Pb Liquid I Pt(s), 02(g)
(pure liquid) PbO - Si02 (p02 = 1 atm) (E.13.5)
electro! yte
(XpbO = 0.5)
Given: ~G 0 = - 88,800 J moi- 1 at 1300 K for the reaction
Pb(!) + 112 0 2 (g) = (PbO) (E.13.6)
Calculate the activity and the activity coefficient of PbO in the Pb0-Si0 2 electrolyte.
Solution In the above cell, solid platinum serves as the oxygen electrode since its
surface comes to equilibrium with the gas phase quickly without forming oxide. The cell
reaction is reaction (E.13.6), i.e. formation of PbO from Pb and 0 2.
For the emf cell (E.13 .5), therefore,
E5 =EJ - -RT lnJ =EJ - -RT In [ (a ) ]

PbOJn (E.13.7)
ZF ZF [apb] Poi
Now, ~ = - Ni'IZF, where Z =2 for reaction (E.13 .6). Hence,
EJ5 = - (- 88 •800) =+ 0.460 volt

2 x 96500
aPb =l since Pb is pure and P0z = l. Hence,
Es = 0.460 - 0.056 In (apbQ) (E.13.8)

Since Es = 0.60 volts, we have
GpbQ = 0.083; lf>bO = ap~XPbO = 0.166 [Ans.]
l
I
- 12
10
11
p ~
v ~
ii ~
II
II
-
6
v r..
~
~
";
~ ~ I;
~
~ ~ II
v ~ II
.
~
i,. ~ II
~
i,. ~ II
~ c
II ~ II
~ ii ii
~
II ~ ii ii II
~ ii ~ II
~
II ~
"'
II
~
~
~ II
~ II
[,
ii
..,._ I;
i,.
I
;
''
'
II
II
ii
II
I
• I I
. ;
- 2
7 ; II
~ "'
' ;
-... - -- ----
~
----
-
ii
ii
~
~
I; - - -
I I : ::_ ------ ---
~
--- - -----
- -- ---- ---- ~ ~
3
-- - -- = === =-11
~ ~ - --- ~ ~ ~
8 ~
-·- -- ....,._
-- - - ---- -
- - ------
ii
II - :;;. _ ~ ~ ii
-
~
- - ~ - -- -----·~ ~
4
9 ,_ - ~ ~
ii
~
~ II
~· J J I I I I
- I I I I I
-
I I
-
-·11
I
i,
' ' ' ' ..... .... ..... ' ......... ' "' ' ' .... ii
'
;
;
ii
ii l. Solid electrolyte crucible
;
ii
;
ii
5
ii
~
[, 2. Recrystallized alumina crucible ii
~
~
ii [,
; ii
3. Outer alumina crucible ii
ii [,
ii
[,
4. Reaction tube-inconel [,
[,
~
~ ii
~
5. S upporting al umina tube ii
i,.
ii 6. Thermocouple protection tube ii
ii ii
7. Cermets
ii
;
;
ii 8. Tin-melt + Sn02(S) "I
; 8. Ag-Sn alloy melt+ Sn02(S) ii "
;
; 10 . Kanthal lead wires v"
v LI
,I LI
v 11. Double bore alumina sheath ii
i,.
I'
II
LI'
12. Chrome) alumel thermocouple ii
ii
ii
..
LI' ii ~
Fig . E.1 3.1 EMF cell for thermodynamic measurements in molten tin-silver alloys.
Ti n is much more reacti ve than sil ver. He nce, the chemical reaction in both electrodes is
(E. 13. 19)
S ufficie nt time was allowed fu r attainment of this equilibri um at each electrode, and
acu and azn are activities of Cu and Zn at respective electrodes. Since they are pure, acu =
aZn =1. In a Daniel cell, the electrolytes are saturated solutions. It means that the ZnS04
solution co-exists at equilibrium with solid ZnS04 , and the CuS0 4 solution with solid
CuS04 . In other words, in Daniel cell, the activities of ZnS04 and CuS04 in the electrolytes
are also 1. However, in unsaturated electrolytes, the activities in electrolytes will be less
than 1. Hence, from Eqs. (E.13.2) and (E.13.3), at 298 K,
(Gzn 4 )
E = 1.104 - 0.0128 In so (E.13.4)
(GcuS04)
EXAMPLE 13.2 At 1300 K, the emf of the following cell is 0.60 V.

Pb Liquid I Pt(s), 02(g)
(pure liquid) PbO - Si02 (po2 = l atm) (E. 13.5)
electro} yte
(XpbO = 0.5)
Given: .!lG0 = - 88,800 J mo1- 1 at 1300 K for the reaction
Pb(I) + 1/2 0 2(g) = (PbO) (E. 13.6)
Calculate the activity and the activity coefficient of PbO in the Pb0-Si0 2 electrolyte.
Solution In the above cell, solid platinum serves as the oxygen electrode since its
surface comes to equilibrium with the gas phase quickly without forming oxide. The cell
reaction is reaction (E.13.6), i.e. formation of PbO from Pb and 0 2 •
For the emf cell (E.13.5), therefore,
E
s
=£1
s
- RT In J
ZF
=FJ -
s
RT In [ (apbO)
ZF [aPb
]'nP0i,
l (E.13.7)
Now, £'l =-£/ZF, where Z =2 for reaction (E.13.6). Hence,
( -88,800) -
r-O -
c,5 - -
- + 0 ·460 VO
It
2 x 96500
aPb =l since Pb is pure and P0i, = I. Hence,
£5 = 0.460 - 0.056 Jn (apbo) (E.13.8)
Since £ 5 = 0.60 volts, we have
apbQ = 0.083; lf>bO = apbd'Xrbo = 0.166 [Ans.]

13.3.2 Examples of Concentration Cell
EXAMPLE 13.3 Chlorides, fluorides etc. are low melting solids. Hence, for thermodynamic
measurements, emf cells, with these molten salts as electrolytes, can be fabricated with glass.
An example is the following cell:
W I Cd (pure liquid) I CdC1 2 dissolved in I [Cd-Pb] I W

(electrode I) molten LiCl-KCl (liquid alloy) (E.13.9)
solution as (electrode II)
electrolyte
Here, W means tungsten lead wires. Calculate the activity of Cd in Cd-Pb alloy at 800 K,
if the cell EMF is 0.1 volt.
Solution It is a concentration cell. The emf is due to difference in concentration of

cadmium at the two electrodes. Electrochemical exchange of Cd between the two electrodes
occurs via Cd2 + ion in electrolyte. Tungsten serves as solid lead wire for connection with
the outside circuit. Employing Eq. (13.19), we may write
/.1oJ (alloy) - ~(pure Cd) = - ZFE (E.13.10)

Again, from Eq. ( 10.25),
(E.13.11)
Combining the above equations and noting that Z = 2, we obtain
RT
E= -2F ln[llcdlauoy (E.13.12)
Putting E = 0.1 V and T = 800 K, we get

aCd in alloy = 0.055 (Ans.)
We can also employ Eq.(13.11). f.!J is emf of a hypothetical cell where cadmium at
standard state (i.e. pure Cd) constitutes both electrodes. Obviously then, f.!J = 0 for this
system.
Again, the cell reaction (actual process) is
Cd(pure) = [CdJanoy (E.13.13)
For this, L1G = partial molar free energy of mixing of Cd in the alloy = GQi
J
= Cllcd Luoy =[Ocd Jalloy
(Ocd] pure
Combining all these, Eq. (13.11) yields
- 2EF = L1G = Gcd = RT In [aCdla11oy (E.13.14)

which is the same as Eq. (E.1 3.12). Therefore, this example demonstrates how either
Eq. (13.19) or Eq. (13.11 ) can be employed for solving problems of concentration cells.
13.3.3 EMF Cells with Solid Oxide Electrolyte
In the solid oxide electrolyte, 0 2- is the current carrying ion. The emf of such a cell is due
to difference in chemical potentials of oxygen in the two electrodes. Hence, it is a concentration
cell, with concentration difference of oxygen of the electrodes being the cause of emf.
A general representation of these cells with ZrOrCaO electrolyte is
02(g, at pbi_ ), Pt I ZrOrCaO I Pt, 0 2(g, at p~ ),

2
(electrode I) electrolyte (0 -) (electrode II) (E. 13.15)
On the basis of Eq. (13.1 9),
~ - µb 2 = -ZFE (E.13. 16)
Here Z =4 since 0 2 molecule has a valency of 4. Again, from Chapter 10,
J.1oi =RT In P0z (10.27)
Combining Eqs. (E.13. 16) and (10.27), we get
11!.. - 11!! RT p1
E = rvi rvi =- In ~ (E. 13..17)
4F 4F P8i
Such cells have found widespread use in industry for measurement of oxygen dissolved .•
in liquid steel, copper, furnace atmosphere control, combustion control and so on. These emf
cells have also been and are still being used widely for precise thermodynamic measurements
at high temperature. In fact, such cells act as oxygen meters (i.e. oxygen sensors) with
capabilities to measure oxygen concentration or partial pressure in a wide range of temperature
and P0z . These will be illustrated by the following examples.
EXAMPLE 13.4 Roy Chowdhury and Ghosh (Metallurgical Transactions, 2, 217 1-74,
1971) made thermodynamic measurements in liquid tin-silver alloys usi ng solid electrolyte
cell as follows:
Cermet I Sn(l) + Sn02(s) I Zr02-CaO I Sn-Ag(l) + Sn02 (s) I Cermet
(electrode I, P0i =Pb2 ) electrolyte pg2 )
(electrode II, P0i = (E. 13. 18)
The emf cell is shown in Fig. E.13.1. The apparatus and technique had been employed
by Ghosh and co-workers earlier. The electrolyte was in the form of a crucible. Pure Sn melt
with Sn02 was in outer alumina crucible. The alloy melt with Sn02 was inside the zirconia
crucible. The cermets served as nonreacting leads, and the entire assembly was at constant
temperature inside a furnace.
- 12
10
11
" ~
"
II
~
~
~
"' ~
6
,, ~
~
"'
I\
I\
' ~
'
'' ~ p '' ;
~
I
v v ' ~
~
,, ' ~ ~
' ,."
~
' ; ~
' I ~ ~
,, ''
I .I ~ ~
I '' ,. I I I I 1
-
~
~
' . - 2
' ,,
~
7 ~ ' ~
"
~
1::=
- - - - -. ------ --- -- - -- -
-
-
- - -... - - -- "-- :_,,.:_~_':.
- ____
~
v
3
- -- --- --- -
~
- - ---
~
- - ----
II ~ ~
~
8
--- - -·- -- - --
-~ ~
-- _--- ~ ~
v
I\
~
I - --
- ----- ----- -· ~ 4
-,,,,.,,,,,,,,,,,, , ,,,,,,,. ,,,,,,.,,- ,, ,,.,,._,
~
9 ~ ~
;
~
I\ ~ ~-·
k-" _,,, _,,_,_,,,, ~
.I
I\
' ... , ........ , , , .... ,, ' '' .............. """.,. ..... "" .... ' ..
~
1. Solid e lectrolyte crucible I ~
5
~ v
,,
; 2. Recrystallized alumina crucible
;
~
~
3. Outer alumina crucible ~
..v
~
.I
~
4. Reaction tube-inconel ....

~
~
.I
;
~
v 5. Supporting alumina tube
.. ~
.I
.I 6. Thermocouple protection tube .. II..v
.I ,.
.I
v v
~
II
.I 7. Cermets v ..
; v
II 8. Tin-melt + Sn02(S) II
..v
II
"v ..
II
8. Ag-S n alloy melt + Sn02(S) k-"
II ..
II 10. Kanthal lead wires "
II
II
II v 11. Double bore alumina sheath ,,
LI
1,.1 ~
12. ,,
II
LI
.. Chrome! a lumel thermocouple
.. .I
LI
Fig. E.13.1 EMF cell for thermodynamic measurements in molten tin-silver alloys.
Tin is much more reactive than silver. Hence, the chemical reaction in both electrodes is
(E.1 3.1 9)
Sufficient time was allowed fc,r attainment of this equili brium at each e lectrode, and
I
E qui·1·b .
1 num constant
K = { (llsoOz)
] } (E.13.20)
Cllsn P0z eq
Oso0z = 1 since it is pure solid, a 50 =1 for pure liquid tin. Hence, from Eq. (E.13.20),
K=-1-= __ 1_ (E.13.21)
P&i [llsn] P~
i.e. ll
P,I
i
P0z
= Casnla11oy (E.13.22)
Combining Eq. (E.13.22) with Eq. (E.13.17), we get
RT
E =4 F In Casolanoy (E.13.23)
(Note that the sign of E is negative. The sign of measured E depends on how the voltmeter
is connected. It has to be taken as negative for calculation purposes.)
Figure E.13.2 presents activity vs. composition relationship in liquid Sn-Ag solution
at 1100 K. The activity of tin was experimentally measured. The activity of silver was
0 .8
0.6
0.4
0.2
0.2 0.4 0.6 0.8 1.0

Xsn
Fig. E.13.2 Activity-composition plots for liquid tin-silver alloys at 1100 K [P.J. Roy Chowdhury
and A. Ghosh, Met. Trans, 2, 2171 , 1971 ).
calculated from experimental data by the Gibbs-Duhem integration using Darken 's method
(see Section 9.6.3). Pure 'silver is a solid at 1100 K. For GD integration, hypothetical liquid
Ag standard state was employed. The procedure and equations for this conversion have been
presented in section 10.3.3.
EXAMPLE 13.5 During steel making, measurement of dissolved oxygen content of liquid
steel is often required. It is done by oxygen sensor, which is an emf cell with SE electrolyte.
The sensor is immersed into liquid steel. Since it gets damaged, it cannot be reused. The
sensor also carries a thermocouple. The readings of emf and temperature are to be completed
within a minute.
Several commercial designs are available. A common design is:
Mo I Cr + Cr20 3 I Zr0 2- Mg0 I liquid steel I Fe (E.13.24)

(reference)
Molybdenum and iron serve as leads. Since they are dissimilar, emf reading requires correction
of the themwcouple effect. The solid electrolyte is in the form of a small tube open at one
end.
From Eq. (E.13.16),
/1oz (liquid steel) - /1oz (reference)= - 4FE (E.13.25)
Since Cr and Cr20 3 are pure, we obtain the following relation on the basis of Eq.(8 .10):
/1oz (reference)= Irr In P0i, (reference)

=L1G.f for formation of 2/3Cr2 0 3(s) (i.e. per mol.02) (E.13.26)
For liquid steel, the relevant reaction is
1
202(g) = [O]dissolved in liquid steel (E.13.27)
for which the equilibrium relation is
K 27 ={[ho]} 112 (E.13.28)

Po-i eq
where K 27 is equilibrium constant for reaction (E.13 .27) and ho is the activity of dissolved
oxygen in steel in 1 wt.% standard state (see Section 11.5). Hence,
/1oz (liquid steel)= RT In Po-i (in equilibrium with liquid steel)
=2RT In [ho]
K 11
(E.13.29)
1
<
Combining these with Eq. (E.13.25), we get
2RT In [ ~7 ] L1drG Cr 0 = -4FE

- 2 3)
(E.13.30)
At 1873 K, !ff/ of Cr20 3 = -422.7 x 103 J/mol 0 2, K27 = 2615. If E = - 0.153 V, then
\
from Eq. (E.13.30), ho = 0.0005. If the behaviour of oxygen in steel is assumed to be
Henrian, then
weight % 0 = W0 = ho = 0.0005 wt.% (i.e. 5 ppm) [Ans.] (E.13.31)
13.4 Single Electrode Potentials and Electrochemical

Series in Aqueous Systems
The electrical potential of a single electrode is unknown. It cannot be found out since any
measurement woulq require two electrodes. Therefore, the term single electrode potefll ial t;
as used in literature, is not literally correct from the above point of view, and requires
clarification.
H+ is the common ion in aqueous solutions. Hence, the single electrode potential of a
standard hydrogen electrode has been arbitrarily assumed to be zero, and is employed as
reference electrode for measurements of E of other electrodes. For example, Jet us consider
0
the following cell set-up:
Zn(s) I ZnCl 2(aq) II HCl(aq) I Pt(s), H 2(g) (13.22)

(aZni+ = 1) (aH+ = 1) at 1 atm
:
The emf of this cell is taken as the standard single electrode potential of Z11/Zn 2+ couple.
It is designated as tfJ 2+. Again, platinum serves as the hydrogen electrode since its surface
comes to equilibriu~ith the gas fairly fast.
There are two conventions for single electrode potential, viz. oxidation potential and
reduction potential.
(i) Oxidation potential. For oxidation reaction such as
Zn(s) = Zn 2+ (aq) + 2e-, tP = + 0.763 v (13.23)
(ii) Reduction potential. For reduction reaction such as
Zn 2+(aq) + ze- = Zn(s), tfJ = - 0.763 v (13.24)
In other words, reduction potential of a system is equal and opposite to that of its
oxidation potential.
In conformity with other standard text books, Table 13.2 presents values of standard
reduction potelllials for some metals at 25°C in aqueous solutions of their respective ions.
The standard single electrode potential (i>) is the value of E, when both metal and Its ion
' "
are at their respective standard states, as in cell (13.22), where a~11 , az112+, a"+ , PH2 are all
equal to 1. Here, the question arises- what do we mean by a i+ = 1 in an aqueous solution ?
Zn
az112+ simply c an never by measured. Hence, by convention, if the concentration of the salt
(ZnCl2) is 1 molal (i .e . 1 gm-mole in 1000 g water), then Eis taken as E0. This is indicated
in Table 13.2. If the ion exhibits Henrian behaviour, i.e. activity is proportional to molality,
then the calculation of E at concentrations other than l molal is simple. However, it is not
true in most cases and, hence, specific information of behaviour is required. Further discussion
of this topic is beyond the scope of this text.
Table 13.2 Standard single electrode potentials in aqueous solution at 298 K, 1 atm
(standard state is 1 molal)
Electrode reaction e 0, volts

F2 + 2e- = 2r 2.65
Co3- + e- = Co 2- 1.82
Cl2 + 2e- = 2Cl- 1.3595
0 3 + H20 + 2e- = 0 2 + 20H- 1.24
B r2(1) + 2e- = 2Br- 1.0652
2Hg2+ + 2e- = Hg;f+ 0.92
Hg2+ + 2e- = Hg 0.854
Ag++ e- =Ag 0.7991
Fe 3+ + e- = Fe2+ 0.77 1
12 + 2e- = 2i- 0.5355
Cu2+ + 2e- = Cu 0 .337
Cu2+ + e- = Cu+ 0.153
2H+ + 2e- =Hi 0.000
Fe3+ + 3e- =
Fe -0.036
Pb2+ + 2e- = Pb -0.126
Sn 2+ + 2e- = Sn - 0.136
Cd 2+ + 2e- Cd = -0.403
Cr3+ + e- = cr2+ -0.41
Fe 2+ + 2e- = Fe -0.440
Zn 2+ + 2e- = Zn - 0.763
2H 20 + 2e- =
H2 + 20H- -0.828
Zndz- + 2H20 + 2e- = Zn + 40H- - 1.216
Al3+ + 3e- = Al - 1.66
Mg 2+ + 2e- Mg = -2 .37
Na+ + e- =Na - 2 .714
C a 2+ + 2e- = Ca - 2.87
K+ + e- = K - 2 .925
Li + + e- = Li - 3.045
The overall reaction in a Daniel cell is given by Eq. (13.2). It is the sum of Eqs. (13.3)
and (13.4). Hence, for the Daniel cell,
Standard cell emf ('/f>) = t>Znfln2+ + ~2+ieu =teu2+1eu - ~2+/Zn

= 0.337 - (- 0.763) = 1.10 V at 25°C (13.25)
If the electrode and electrolyte are not at their standard states then, on the basis of
Eq. (13.11), the reduction potential for Zn 2+1zn is given as
E=Efl - -
RT
2F
In [-az,.-
aZn2+
l (13.26)
In Eq. (13.26), Zn is the reduced state and Zn2+ is the oxidized state. We may generalize
it as
Oxidized state + 2e- = reduced state (13.27)
On the basis of the above equations, the generalized equation is
E =t - RT In (aR
ZF a0
l (13.28)
where a 0 = activity of oxidized species and aR = activity of reduced species. Equation

(1 3.28) is the famous Nemst equation in electrochemistry.
13.5 Pourbaix Diagrams

Pourbaix diagrams, or potential pH diagrams, are graphical representations of electrochemical
equilibria occurring in aqueous systems, where pH = - log [H+]. It was first proposed by
Pourbaix* and are very useful in the areas of corrosion, electrometallurgy and hydrometallurgy.
They are essentially electrochemical analogues of phase stability diagrams discussed in
Chapters 10 and 11.
As an example, again consider dissolution of Zn in aqueous solution. At 25°C, in acidic
solution of pH < 5, the metal dissolves in aqueous solution as Zn2+. At higher pH, Zn(OHh
is the thermodynamically stable phase, and the equilibrium activity of Zn 2 + will be governed
by the solubility product of Zn(OHh and the pH of the solution, i.e. by the equilibrium of
the reaction
(13.29)
At still higher pH, Zn02 ions become stable and both Zn 2+ and Zn(OH) 2 are unstable.
*M. Pourbaix, J. Electrochem, Soc., 101, p. 213, 1954.

The above discussion is applicable to a situation where there is no externally imposed

electric potential on the system. By imposing a potential, however, pH ranges for stability
of phases can be altered. Here, the guiding criterion is the Nemst equation. For a Zn 2+tzn
couple, it is Eq. (13.26). If the activity of Zn 2+ in aqueous solution and temperature are
known, e for Zn 2+tzn can be calculated from Eq. (13.26). If the externally imposed potential
(fex1) is more negative than E, then metallic zinc is stable. If fext is less negative than e, then
Zn 2+ is stable. Also, it should be noted that the activities in various phases would govern
stabilities.
Figure 13.2 gives the potential-pH diagram for Zn/H20 system at 25°C. The stable
phases are indicated. The reactions and equations for construction of Fig. 13.2 are as follows:
Zn 2+ + 2e- = Zn, e = tfJ + 0.03 log (a i+)
Zn
(13.30)
Zn(OH)i + 2H+ + 2e- = Zn + 2H 20, e = IP - 0.059 pH (13.3 1)
Zn<Yi- + 4H+ + 2e- = Zn(OHh + H2, e = tP - 0.118 pH + 0.03 log ( azno2-)

(13.32)
where "log" is the common logarithm.
Zno~-(aq)
a
Zn02_
= io-yr1'..~
Zn 2+(aq)
y=6 2 0
I Ozn2+ = 10-x J
Zn(OH}i
x=O 4 6
Eo = -0.763 v
/ x=O
-1
Zn
0 2 4 6 8 10 12 14 16
pH
Fig. 13.2 Simplified potential-pH diagram for Zn/H20 system at 298 K.
From Table 13.2, £o = - 0.763. For construction of Fig. 13.2, three values of ai~ and
az002- were assumed as 1, 104 and I 0-6 and 1, 10-2 and l 0-6 res pee ti vel y.
13.6 Summary
1. Electrochemical cells may be classified into galvanic cells (i.e. batteries) and electrolytic
cells. Only galvanic cells can be operated reversibly.
2. Besides aqueous solutions at room temperature, molten salts, molten slags (i.e.
oxides) and solid electrolytes are employed as electrolytes in metallurgy and materials
applications at high temperatures.
3. For a reversible galvanic cell, the cell emf(E) is related to thermodynamic quantities
by the two alternate forms of the equation:
(i) L1G = - ZFE
where L1G is free energy of cell reaction, Z is valency involved, and F is Faraday's
constant.
(ii) II I - - ZFE
Jlj-Jlj-
where i is the species exchanged between electrodes I and II, and µ; is the chemical
potential of i. 'i'he first form is used for reaction cells and the second form is
preferred for concentration cells.
RT
4. E =f.!J - -- In J
ZF
where E° is standard emf of the cell, and J is the activity quotient.
5. Some examples of emf cells have been discussed in this chapter. It may be specially
mentioned that, for a solid oxide electrolyte cell, equation in 3(ii) above may be
rewritten as
I
E = RT In Po2
11
4F p 02
where Pbz and p~ are partial pressures of oxygen at electrodes I and II, respectively.
The US Sign convention for emf has been adopted in this text.
6. The single electrode potential of an electrode (£), and electrochemical series in
aqueous solution have been discussed. £ can be taken as oxidation potential or
reduction potential, which are equal and opposite for a specific system. The reduction
potential has been c hosen for application in this text.
7. The fundamentals of Pourbaix diagrams or Potential-pH diagrams, have been illustrated
through the example of Zn-aqueous solution system.
PROBLEMS
13.1 The emf of the cell
Pt, H2 (g) I H2S04 I Pb02(s) + PbS04 (s), Pt
aqueous
solution
was measured as 1.63195 and 1.62950 volts, respectively at 45°C and 35°C. Calculate
LIG, LIH and LIS of the cell reaction at 25°C.
13.2 The emf of the cell: Ag(s) I AgCl(s) I Cl 2(1 atm), Pt, was found to be:
£(volts) = 0.977 + 5.7 x 10-4 (350 - t) - 4.8 x 10-7 (350 - t) 2
in the temperature range t = 100°C to 450°C. Calculate the value of LICp for the cell
reaction .
13.3 For commercial production of magnesium, MgC12 is decomposed by electrolysis
according to the following reaction at high temperature:
Calculate the minimum theoretical voltage (to ov'!rcome cell emt) to be applied for
decomposition at 750°C. Assume MgCl 2, Mg as pure.
13.4 The emf of the cell: Al(s) I AlClrNaCl melt I Al-Zn alloy(s)
was measured as 7.43 millivolt at 380°C, and the temperature coefficient of the emf
as 2.9 x 10-5 volt/°C. Calculate aAi. c;:I and ii;:I in the alloy.
13.S Consider the high temperature emf cell in Eq. (E.13.18).
At Xsn = 0.5 in the Sn-Ag alloy, a measured cell emf was fitted with the equation:
£(volt) = 14.69 x 10- 3 - 30.4 x 10-6 T, where Tis temperature in K. Calculate (at
Xsn = 0.5):
(i)
(ii)
13.6 Calculate the emf of the following electrochemical cell at 1200 K:

Pt I Ni(s) + NiO(s) I Zr0 2 - CaO I Mn(s) + MnO(s) I Pt
13.7 Consider the following solid electrolyte cell:
Pt I S0 2 + S03 gas mixture I Zr02 - CaO I M(s) + MO(s) I Pt
Calculate:
(i) LiG° of formation of MO
(ii) chemical potential of S2 in the gas mixutre.
Given: T = 1500 K; total pressure of 502 + 503 = 1 atm;

Ps0z IPs03 ratio = 500; cell emf = 0.46 volt
13.8 The emf of the following cell:
Cu(s) + Cu20(s) I Zr02 - Cao I Cu - Ag alloy (1) + Cu 20(s)
was measured as: E(mV) = -177.8 + 0.1657 T,
at Xeu = 0.5 and temperature range 1090--1300 K.
Calculate (at 1200 K):
(i) acu with solid Cu as standard state
(ii) aeu with liquid Cu as standard state.
(If the sign of emf causes confusion, use your judgement. For example, the calculated
activity of a component in alloy should be less than 1.)
Chapter 14
Thermodynamics of Surfaces,
Interfaces and Defects
14.1 Introduction
By surface, we mean a free surface which separates a solid or liquid from surrounding gas
or vacuum. Ideally, the surrounding should be a vacuum. However, a gas phase has negligible
concentration of atoms/molecules pe r unit volume as compared to that in solid or liquid.
Hence, approximately any interface with gas is also treated as a free surface.
Surfaces play an important role in a variety of phenomena encountered in metallurgy and
materials science. A surface is only a few atomic layers thick. On the surface, there are some
free bonds projecting normal to surface. This causes a difference in structure of the surface
layer from that of the bulk of the solid/liquid.
A surface possesses some extra energy due to the work done to create it. Specific surface
energy is this excess energy per unit surface area (say, in joules per m2 of :mrface). From
an atomistic viewpoint, work is to be done to create free bonds on the surface (i.e. equivalent
to bond breaking). Bond energy is of the nature of enthalpy. Strictly speaking, it is internal
energy, which is approximately the same as enthalpy for condensed phases. Since H = G +TS,
excess surface energy is due to excess surface free energy and excess surface entropy. But
specific surface energy is usually taken as free energy (either Gibbs or Helmholtz free
energy, dependi ng on application) per unit surface area, in excess of that of the bulk.
An interface, by convention, is 'the surface separating two condensed phases. Here, we
do not have the concept of free bonds. Solid surfaces are rough. Hence, the interface between
two solids has only some contact points. Moreover, such contacts depend on surface roughness,
and hence, are unpredictable and irreproducible. Therefore, thermodynamics is not capable
of predicting its properties. Polished interfaces joined by bonding such as adhesives, soldering,
brazing etc., are more amenable to thermodynamic treatment. In contrast, solid/liquid and
liquid/liquid interfaces, especially the latter, can be more defi nitely reproducible. Excess
248
Thermodynamics of Surfaces, Interfaces and Defects
energy at the interface is influenced by the extent of composition mismatch and/or structural
mismatch (such as water and ice). At electrode/electrolyte, slag/metal, metal/aqueous solution
interfaces, we have the electric double layer also, which modifies interfacial energy. Hence,
the phenomena here are more complex than those at free ~urfaces. Grain boundaries in solids
are also interfaces. There, the mismatch in grain orientations gives rise to excess interfacial
energy (i.e. grain boundary energy).
Grain boundaries, stacking faults , point defects (i.e. vacancies and interstitials) and line
defects (i.e. dislocations) play very important roles in phase transformations, diffusion etc.
and in control of mechanical and other properties of solids, as well as their processing. Fine
particles have very high surface energy per unit volume, which are responsible for their
special behaviour pattern.
In all the above arl!as, it is thermodynamics again, which constitutes a principal scientific
foundation. However, other courses in the area of metallurgy and materials are offered ,
especially on structure, properties and processing of solid materials, where thermodynamic
aspects are also dealt with in detail. Hence, for the sake of basic information and understanding,
in this introductory text, onl y a brief mention will be made about thermodynamic aspects.
14.2 Thermodynamics of Surfaces
14.2.1 General Considerations

From Chapter 10, we recall that, at constant temperature and pressure,
dG' =µ 1 dn1 + J1?. d"'i. + ... + µ; dn; + ... = "f.i µ, dn; (10. 11)
where 1, 2, ... are components, in the absence of any other work. In the presence of any
other work (OW'),
dG' = "f.µ;
i
dn; - 8W' (14.1)
The work done to create an infinitesimal small surface dA is given as
c5W' = - adA (14.2)

where the negative sign is due to the fact that work is to be done on the surface by an outside
agency to create a surface; a is specific surface energy, as defined in Section 14.1. Therefore,
dG' = "f.µ;
i
dn, + adA (14.3)
8W' is nothing but [- dG;]. which is the change in surface free energy. Hence,
dG; = adA (14.4)
If O" is independent of surface area, then integration of Eq. (14.4) leads to
G'
G~ = aA, i.e. a= ; = GsA (14.5)
where G~ is surface free energy and GsA is specific surface energy per unit area. O" is a
constant for a surface, provided the surface is isotropic, i.e. it has no directionality. Liquid
surfaces are isotropic, whereas solid surfaces may or may not be isotropic.
You are familiar with surface tension, which is force per unit length of a surface. This
originated from Young's model, which considered the mechanical behaviour of a surface.
The actual interface is replaced by an imaginary infinitesimally thin elastic membrane, called
surface of tension. Dimensions of specific surface energy and surface tension are the same
(Mr2) . For isotropic surfaces, numerically also they are equal. Table 14.1 presents some
values of <1.
Table 14.1 Specific Surface Energies of Some Materials
Material Temperature, K GsA x 103, J m-2
Water (I) 298 72

NaCl (s) · 298 300
Al203 (s) 2123 905
Al203 (I) 2353 700
TiC (s) 1373 1190
Al (s) 723 980
Al (I) 933 866
Cu (s) 1198 1780
Cu (I) 1573 1300
Fe (I) 1823 1835
Pb (I) 600 450
Hg (I) 298 465
14.2.2 Adsorption Isotherms

Surface free energy is lowered by the phenomenon of adsorption, and this lowering constitutes
the driving force for the same. Adsorption means preferential incorporation of a component
i (element or compound) in the surface layer. In other words, there is an excess concentration
of i on the surface as compared to that in the bulk, if component i is adsorbed on the surface.
Adsorption is broadly classified into physical adsorption and chemisorption.
(i) Physical adsorption. Here, the forces between the surface and the adsorbed species
is physical in nature (van der Waals type forces)
(ii) C~emisorption. Where, chemical bonds form between surface and adsorbed species.
Physical adsorption is dominant at low temperatures. Above room temperature. chemisorption

is dominant. Hence we are concerned primarily with the latter, and subsequent discussions
will be restricted to only chemisorption.
Adsorption isotherm means adsorption equilibrium relatio11s at co11sta111 temperature. For
chemisorption, we are concerned with the following adsorption isotherms:
(i) Langmuir adsorption isotherm
(ii) Gibbs adsorption isotherm
These are according to the names of the scientists who proposed them.
Langmuir adsorption isotherm

This was proposed by Langmuir for adsorption of gases on solid/liquid surfaces. The assumptions
are:
1. The adsorbed molecules/atoms are present on the surface as a single layer (i.e.
monolayer adsorption)
2. The gas is ideal
3. Adsorbed layer is ideaJ, i.e. there is no interaction amongst adsorbed atoms/molecules.
The process of chemisorption was considered as a chemical reaction by Langmuir, and
was written in the form
G
I I
G + S =S (14.6)
I
where G is the symbol for a monatomic gas molecule, S indicates a free surface
G
I
site, S is a surface site with adsorbed gas atom.
If K0 is adsorption equilibrium constant at a temperature, then
9 (14.7)
-
1 --8 -K
- a p,G
where
1- 0 = fraction of free surface
0 = fraction of surface site with adsorbed atom.
9, (1-0) are measures of surface concentration, i.e. mole fractions of surface sites. Since
the atoms in the adsorbed layer behave ideally, mole fractions are the same as respective
activities. These considerations constitute the basis of Eq . (14.7), which is the Langmuir
adsorption isothenn.
For adsorption of diatomic gases, adsorption is preceded by dissociation of molecules, i.e.
1
2G2 (g) = G(g), Ks= li2
PG
(14.8)
Poi
From Eqs. (14.7) and (14.8),
_O_ -K K 112 (14.9)

l _ (} - n 8 PGJ.
It may be noted that Langmuir isotherm can also be applied to adsorption of a component
from the bulk of the solution. This requires invoking some additional equilibrium relations.
Gibbs adsorption isotherm

The Gibbs adsorption isotherm is applicable to adsorption of a component i (element/
compound) of a solution onto the surface of the solution. Diffusion of atoms and molecules
are several orders of magnitude faster in liquid as compared to those in solids. Moreover,
convection in a liquid enhances mass transfer significantly. Hence, attainment of adsorption
equilibrium between the surface and the bulk of a solution is possible only for liquid
solutions in a realistic timeframe.
For Gibbs free energy, Eq. (9.16) may be written as
G' = I.G;n;
I
(14.10)
Since we are considering a surface, the surface energy term is to be included in Eq. (14.10).
Hence,
G' =· I: G;n; + Aa (14.11)

i
Differentiating Eq. (14.11), we get
dG' = (7 G; dn, +a dA) + (7 n; dG; +Ada) (14.12)
Replacing G, by IJi, we get
dG' = (7J..4 dn; + a dA) + (7n; dJ..4 +A da) (14.13)
Combining Eqs. (14.13) and (14.3), we obtain the modified Gibbs-Duhem equation for
surfaces as
I:n;dµ; +Ada= 0 (14.14)

i
Consider adsorption of component i only. Then ,

11; dµ; + A da =0 (14.15)
i.e.
n; da
-=-- (14.16)
A dJL;
By definition,
Surface excess of i = r..

i =
(n;)
A surface -
(n;)
A bulk
(14.17)
Since, for a surface active species,
(14.18)
r..1
== (n;) = da
(14.19)
A surface dJli
Equation (14.19) is the mathematical statement of Gibbs adsorption isothenn. One of the
best examples of its application is for adsorption of oxygen atoms on surface of liquid iron
at high temperature, if we djssolve some oxygen in the liquid.
Figure 14.1 shows variation of awith wt.% oxygen dissolved in liquid iron, as measured
experimentally. Since d<Jldµ; is negative in this case, I'; is positive from Eq. (14.19). Thus,
oxygen is preferentially adsorbed on surface of liquid iron, i.e. it is surface active.
[wt.%, 0)
0.001 0.01 0.1
1600
...
I
0
E
......
,.., 1400
0
0
x
b
1200
1000.__~~--4-~~--'"~~~-'-~~-'-~~---'~~~
-8 -7 -6 -5 -4 -3 -2
In [wt.%, 0]
Fig. 14.1 Variation of surface tension of liquid iron with dissolved oxygen content at 1550°C
[F.A. Holden and W.D. Kingery, J. Phys. Chem., 59, 557, 1955).
Figure 14.2 shows fraction of surface covered by adsorbed oxygen as function of wt.%
oxygen in the iron-oxygen solution. It was calculated from the data given in Fig. 14. l
assuming monolayer adsorption. It may be noted that the surface is almost saturated with
oxygen· even at as low as 0 .03% oxygen. In other words, oxygen is strongly surface active
in liquid iron . Sulphur also shows a similar behaviour. It may be mentioned that these
phenomena significantly influence steel making processes.
u
II)
1.0
0.8
Ideal
mono layer
I
/
/
--
~
....
::J I
"'
...... I
'
0
0.6
II)
on
eo;S
I
....
II) t
> I
0
u 0.4
e;;
c:
0
·p
u
.... 0.2
~""
OL-~~_._~~~~~~'--~~-'-~~~
0 0.01 0.02 0.03 0.04 0.05

Oxygen, wt.% (in bulk)
Fig. 14.2 Fractional coverage of surface of liquid iron by adsorbed oxygen atoms as function
of dissolved oxygen content at 1600°C.
14.2.3 Effect of Surface Energy on Pressure
It had been known for centuries that the level of water is higher inside a small diameter glass
tube. Not only that, the surface of water in the tube is concave upwards. This was attributed
to water 'wetting' glass. This is the capillarity phenomenon which you are familiar with. All
these, viz. the height of water inside the tube and the radius of curvature of water surface,
can be quantitatively calculated by force balance involving surface tension values. The
behaviour of mercury is opposite to that of water since it is ' non-wetting' to glass.
In a liquid-gas system, gas pres~ure and c urvature of interface are interrelated due to
surface energy effects. The general tendency of a gas bubble inside a liquid or a liquid
droplet in a gas is to assume spherical shape since spheres have lowest specific surface area
(i.e. surface area/volume ratio ), and hence the lowest surface energy per unit volume of the
liquid (i.e. per mole of liquid).
Thermodynamics of Surfaces, Interlaces and Defects
Excess pressure (L1P) inside a gas bubble in liquid

Consider a gas bubble of radius r inside a liquid. Let L1Gxs be the excess free energy due
to the presence of the bubble. Then ,
(14.20)
where L1Gp/V is the excess free energy of bubble per unit volume due to excess pressure (L1P)
inside the bubble. The excess pressure is due to squeezing action of surface tension force on
the bubble. Hence, the 2nd term in Eq. (14.20) containi ng L1GP is negative. At constant
temperature, dG = VdP. Hence,
(14.21)
L1Gxs = 4nr 2<1 - 4/3m3. L1P (14.22)
The system would have the minimum L1Gxs at bubble-liquid equilibrium. Therefore, at
equilibrium,
d(L1Gxs) =0 =8nr0' - 4nr2 • L1P (14.23)

dr
i.e.
The value of L1P can be quite large for tiny bubbles. For example, if the bubble radius
is 0.1 mm (10-4m), then L1P = 3.2 x 104 Pa (0.3 atrn) in liquid steel with O' = 1.6 J m-2 ,
at bubble-steel equilibrium.
Equilibrium vapour pressure of a liquid droplet

Consider a pure liquid A. For a large volume with flat su~ace, its saturated vapour pressure
( P~) is given by the Clausius-Clapeyron equation (section 6.4.2). Proceeding with derivation
along the line as for gas bubble in liquid, it can be shown that the equilibrium vapour
pressure of the droplet (p A) will be larger than P~ , and would follow the relation
2a
PA =pA0 + - ( 14.25)
r
where r is the droplet radius.

Figure 14.3 ·shows variation of vapour pressure of liquid zinc as function of droplet
radius at 900 K.
~~~~~~~~~~~
0.031 - - - - - - - - - - - - - -
0.030
E 0.029 '
....
"';:,
N
~
0.028
0.027
pfln - - - - - - - - _-:_:-:_:-::_::-=---...J
0.026
0.01 0.1 10
r, micron
Fig. 14.3 Equilibrium vapour pressure of zinc at 900 K as function of radius of liquid Zn droplet.
14.2.4 Effect of Surface/Interface Energy on Equilibrium Phase

Transformation Temperature
In Section 6.4, we discussed equilibrium phase transformation in condensed systems.

Figure 6.1 provided basic classifications by a schematic presentation of GA(I), GA(ll) and
L1G[=GA(II) - GA(I)] at constant pressure as function of temperature. In the figure,
phase II is stabler than phase I above the equilibrium transformation temperature, T1r
Let us now suppose that A is present in phase I in fine sizes with a large specific surface
area. Then GA(I) will change to G,, (I), where
G,, (I) = GA(I) + Gs (14.26)
with Gs as surface energy per mole of A in phase I. Then,
(14.27)
Figure 14.4 schematically presents the features of Eq. (14.27). As may be noted from
the figure, T1r gets lowered to ~ due to surface energy of phase I. This phenomenon is
encountered during melting of fine particles of solid, or crystals with a highly curved
surface, such as at tips of dendrites, where melting occurs at a lower temperature. It is
obvious that the situation will be reverse and T1r will increase if phase II is of fine size.
(a)
I
t I
I c•
-~(/) ~ G
- - A(l)..,_G,
-- s
+ve (b)
t
L1G 0
........ ......
- - -- °i'~
.......
I 1 ......
I I ....... .......
I I <Jc.........
T~r Trr
Temperature --+
Fig. 14.4 Illustration of effect of surface energy on equilibrium phase transformation temperature
at constant pressure (schematic).
14.3 Thermodynamics of Defects in Solids

As stated in Section 14.1, defects in solids govern a variety of properties of solids-
mechanical, electrical, electronic, and so on. Hence, they are elaborately taught in metallurgy
and materials science programme. Thermodynamics of defects are also covered there. Therefore,
we shall make only very brief mention of some important points here to understand the basics.
It should be mentioned here that concentration 1lf dislocations is orders of magnitude
larger than what thermodynamic equi librium predicts. It is a highly noncquilibrium situation
and thermodynamics is not of much help. Hence, we shall not discuss about dislocations.
14.3.1 Surfaces and Interfaces of Solids
Liquid surfaces are isotropic. Also, atoms at liquid surfaces are mobile, and hence, equilibrium
is attained easily. But these are not generally applicable to solid surfaces. Therefore, the
meaning of surface tension is not clearcut for solids. Also, specific surface energy and
surface tension would not have the same numerical value. Consider a single crystal. It does
not possess a unique specific surface energy since the latter has different values for different
crystal j)lanes. However, for a polycrystalline solid with random orientations at the surface,
the statistically averaged values would be isotropic, as in the case of a liquid surface. The
surface energy values of some solids are also included in Table 14.1.
The internal bou.n daries in crystalline solids may be divided into the following:
l. Internal boundaries between crystals of the same phase, resulting from orientation
differences (hereafter referred to as grain boundaries)
2. Internal boundaries separating phases of different structure or compositions or both.
The specific interfacial energy of a grain boundary is known as its grain boundary energy
Ge. Ge would depend on the misorientation, which is also known as the tilt angle (9) of
neighbouring grains. Ge is maximum at 9= 45° since it corresponds to maximum misorientation.
Figure 14.5 presents experimental values of Ge. as a function of 9 in copper at 1338 K.
Besides Ge. the figure also presents values of Ge/GsA• where GsA is surface energy per unit
area, as defined in Eq. (14.5). It may be noted that Ge/GsA is less than 0.4. It has been found
to be less than 0.5 for several metals. GsA corresponds to increase of free energy due to the
presence of free bonds at surface. But Ge is due to misorientation of bonds only. Thus, Ge
is expected to be lower than GsA·
0.40
600
0.32
500
' N
I
< 0.24
Cl)
c.:> '' 400

E
.......
'
-...
al
M
c.:> 300 0
0.16 \
\ x
200 C)
0.08
\ c.:>
100
0 0
0 10 20 30 40 50 60 70 80 90
9, degrees
Fig . 14.5 Experimental value of grain boundary energy as function of misorientation for a
simple t ilt boundary in copper at 1338 K [N.A. Gjostein and F.N. Rhines, Acta Met.,
7, 319, 1959].
14.3.2 Vacancies and Interstitials in Solid Metals

Derivation of equilibrium relations
Vacancies and interstitials are point defects. As an illustration, let us consider vacancies in
a crystalline metal. In N atoms of a perfect crystal, the number of atoms as well as the
number of lattice sites are also N. If nv number of atoms are removed from lattice points
and transferred to the surface, we create nv vacancies. Hence, mole fraction of vacancie:; (X,.)
would be equal to n .,l(N+n ,.), and mole fraction of atoms = Nl(N+nv) = l - X,..
•·1
~· '
f
'
'

*fiil :,
=
Now, N + nv corresponds to (N+nv)IN0 y moles, where N0 is Avogadro's number. Hence, ,,
the number of moles of vacancies = yXv. Let, 'i•
.&Iv = Enthalpy of formation of 1 mole of vacancies

L1Sv = Change in vibrational entropy due to vacancy formation, per mole of vacancy l
Ll.S'c = Configurational entropy change due to mixing of 1tv vacancies in lattice sites.
Then, from Chapter 12, J
(14.28)
.
:
For a perfect crystal, the number of arrangements, W = 1. W' is the number of arrangements
of nv vacancies in N + nv lattice sites. From Eq. (12.23) and Stirling's approximation
[Eq.(12.24)],
L1Sc = ks In W' = ks In (N + nv) ! = -k8 [ N ln -N- + n,, In --'--

nv J (14.29)
N! n,, ! N+n,, N+n,,
In terms of mole fractions,

..
~
(14.30)
If L1G is the free energy change due to formation of nv number of vacancies, then !
(14.31)
L1G = -n,, (.&Iv - T L1Sv) + -RT(N In -N- + n,, In -n,,-) (14.32)

No No N + nv N + nv
Figure 14.6 presents the various terms of Eq. (14.32) schematically as function of
vacancy concentration nv· R and No are universal constants. T, N, .&Iv, L1Sv are also assumed
constant in Fig. 14.6.
It may be noted that L1G goes through a minimum. The equilibrium vacancy concentration,
I
ny{eq), corresponds to this minimum. At this concentration, d(L1G)/dnv = 0.
Differentiation of Eq. (14.32) leads to
d(L1G) &Iv - TL1Sv -

No·--= RT[ - -N - +(In nv - ln(N +n,,)) +nv • -1 - n,,· -1-]
dnv N + nv nv N + nv
(14.33)
+ve
.1 !2_ (LJH
No v
VJ
...
E
...
~
VJ
::I
0 LJG = ft.n)
·c:
>"'
_R_T (N In N + nv In _nv-'---)
-ve No N + nv N + n.
0
No. of vacancies (nv) --+
Fig. 14.6 Various terms of Eq. (14.32) as function of number of vacancies at constant temperature
(schematic).
Upon rearrangement,
x.(eq) = n.(eq)
N +n_(eq)
=exp (&v) R
exp (- m.)
RT
=exp (- LJG.
RT J
I (14.34)
where '1Gv = free energy of formation of 1 mole of vacancy and, Xv(eq) = equilibrium
vacancy fraction in lattice at temperature r.
Equilibrium concentration of interstitials can also be arrived at following the procedure
adopted above, and, in analogy with Eq. (14.34), we may write
X1 (eq) = n1 (eq) =exp (..1S·)

- ' exp (- -
LJH-)
- ' =exp ( - --' LJG.) (14.35)
N + n;(eq) R RT RT
where the subscript i denotes interstitial.
Additional comments
Equations (14.34) and (14.35) are applicable to vacancies and interstitials originating from
thermal motions of atoms. These are known as thennal or intriTlsic vacancies and interstitials.
There is another type, which is fixed by composition and is independent of temperature, i.e.
the extrinsic type. In metals, the best known example is carbon dissolved in iron as extrinsic
interstitial. The carbon atoms are located at interstitial sites in iron lattice due to their small
atomic radii .
..1Sv, ..1S1 values are very small due to very low concentrations of vacancies and interstitials.
-
Henc.;_e, Eqs.
. .. (14.34)
. and (14.35) may be approximated as
(14.36)
X; (eq) = exp ( - '::::) (14.37)
& v and &; were calculated theoretically for materials like copper. It was found that
for copper, &/&v is approximately 7 . At 1000 K, from Eqs. (14.36) and (14.37),
Xv(eq) 39
- - ::10
X; (eq)
From this result we may make a generalized statement that, for a close-packed crystal such
as copper, the concentration of interstitials is negligible as compared to that of vacancies.
=
It was also calculated that, for Cu at 1000 K, Xv (eq) 3 x 10-7 This is indeed small.
Hence, vacancies will obey Henry's Law, and all properties related to the same will be
proportional to vacancy concentration. One such property is change of electrical resistivity
(L1p) of a metal due to vacancies. From Eq. (14.36), we obtain
(14.38)
Figure 14.7 presents L1p for gold as function of l/T. From the slope, &v was determined
as 92.5 kilojoules per mole of vacancy. For some other metals also, the values are near about
this. These values also agree reasonably with theoretically calculated ones.
14.3.3 Point Defects in Ionic Compounds
Amongst ionic compounds, the oxides have been most widely studied in view of the fact that
their importance is next only to metals and alloys in metallurgy and materials science.
Compounds may be classified into stoichiometric compounds and non-stoichiometric
compounds.
(i) Stoichiometric compounds. Examples are Si02 , Al 2 0 3 , Zr02 , CaO, ZnO, where
metal/oxygen ratios are simple as in their chemical symbols,
(ii) Nonstoichiometric compounds. FexO is an example, where 0.85 < x < 0.97, and
Fe/O ratio is thus not a simple one. Another example is VOx-
In stoichiometric compounds, point defects, as mentioned in section 14.3.2, originate
either from composition effect or due to thermal disturbances. In ionic crystals, there are two
sub-lattices, viz. anionic and cationic. An anion cannot occupy a cationic sub-lattice, and vice
versa, ·since these require very high energies. Moreover, neighbouring positive charges and
negative charges have to be equal for local charge neutrality.
T, °C
950 750 650 550 450
10-S
E
uI
E
.s:::
0
~ 10-9
10-10
0.8 1.0 1.2 1.4

103/T, K- 1
Fig. 14.7 Change of electrical resistivity of quenched gold wires as function of quench temperature
[J.E. Bauerie and J.S. Kochler, Phys. Rev.. 107, 1493, 1957).
Stoichiometric compounds
In Section 13.1, we have discussed ZrO:z-CaO solid electrolyte. It was mentioned there that,
when Ca2+ replaces Zr4+ in cationic sub-lattice, two + charges are less in the cationic site
occupied by Ca 2+. For local charge neutrality, an adjacent 0 2- ion site remains vacant, thus
creating 0 2- vacancy. This is an example of a stoichiometric compound where the point
defects are generated by composition effect.
Point defects arising from thermal effects will always occur in pairs-one positive and
another negative. These may be vacancy-interstitial pair, vacancy-vacancy pair, or pairs
where one is an electronic defect (excess electron or electron hole). For example, NaCl and
KCl exhibit Schottky defects, i.e. equal number of vacancies in M sub-lattice and X sub-
lattice, where M and X denote metal and nonmetal respectively.
As an example, let us consider a crystal MX, where M+ and x- are singly charged. Then
the reaction generating Schottky defects may be written as
MX = Mo-orc1-5) + bVM + OVx (14.39)
The activity of MX is 1 for pure MX. Since o« 1, the activity of M (J-0,• X{l-5) also may
be set equal to l. Therefore, the equilibrium relation for reaction (14.39) may be expressed
as
(14.40)
where [VM] and [Vx] are concentrations of vacancies.

Again, for the reaction
(14.41)
we have (14.42)
The basis of Eq. (14.41) is that a vacancy in cationic sub-lattice is equivalent to a

negative charge. e• denotes an electron hole. Similarly, for
(14.43)
(14.44)
In addition to the above, further relations are obtained from conditions of electroneutrality,
stoichiometry and relation between [e+] and [e- ], as follows:
l vM J + l e- J = l Vx J +le+J (14.45)
[vM] + lvMJ = [vx ) + lvxJ (14.46)
(14.47)
J,
In principle, the six unknown concentrations, viz. [VM], [ VM [Vx]. [ Vx].
[e-] and [e+]
can be solved from the four equilibrium relations and two charge and stoichiometric balances.
Non-stoichiometric compounds
Let us take the example of Wustite (FexO) again. In section l l .2.2, we discussed general
issues related to the activities of Fe, 0, "FeO" in the Wustite field and how these are related
to partial pressure of oxygen in the surrounding atmosphere, with which it was assumed to
be at equilibrium. To sum up, the value of x would be governed both by temperature and
p0z . Therefore, defect concentrations in Wustite also will depend on the above variables.
Experiments have established that the dominant point defect is cation vacancy. Moreover, it
is a p-type semiconductor, i.e. it has electron holes. Some other examples of this type of
defect are NiO, Cu2 0, FeS. Figure 14.8 presents log X vs. log P0i in Wustite at 1223 and
1273 K. From the slope, it is found that log X oc p~ approximately, where X is electrical
J
conductivity. Recognizing that Xis proportional to concentration of electron holes, le+ oc p~.
Textbook of Materials and Metallurg1cal Thermodynamics
-0.6
-0.7
~
bl)
0 ........ 8
-0.9
-1.01.___ _.__ _.....__ _.....__ _.....__ _...__ _ ~ _ __._ __,

-15 -14 -13
log Poi.
Fig. 14.8 Dependence of electrical conductivity of Wustite on oxygen pressure [K. Hauffe and
H. Pfeiffer, Z. Metallik, 44, 27, 1953].
A thermodynamic analysis of the above is as follows:
(14.48)
where 0 0 is oxygen in lattice site, V/c- is vacancy in Fe2+ site, and hence is equivalent to
two nega~ive charges. Since 0 0 is very large compared to defect concentration, it hardly
changes and can be assumed as constant. Hence, we may write
( 14.49)
From charge balance,

[e+) = 2CVlc-l (14.50)
(14.5.1)
14.4 Summary
1. A surface possesses some extra energy due to the work done to create it. The
relationship is
-SW= dG; = adA
where G~ is excess surface free energy, O' is specific surface energy per unit surface
area, and A is surface area.
2. a is lowered by the phenomenon of adsorption at the surface. The adsorption
equilibrium relations at constant temperature are known as adsorption isotherms. We
are primarily concerned with chemisorption. For this, the following two are standard
isotherms:
(i) The Langmuir adsorption isotherm deaJs with monolayer adsorption of ideal
gases on a surface.
(ii) The Gibbs adsorption isotherm deals with adsorption of a component of a
solution onto its surface, and is given as
r.,-- - dO'
dfl;
where I'; is known as surface excess, and µ; is chemical potential of component i in

the bulk solution. The application of this isotherm has been discussed for chemisorption
of oxygen dissolved in liquid iron, onto its surface.
3. A phase of a small size has high surface area-to-volume ratio, and hence, a high
surface energy per unit volume. This causes excess pressure in small gas bubbles in
liquid, higher equilibrium vapour pressure of a liquid droplet, and change of phase
transformation temperatures for fine particles. The appropriate equations have been
derived in this chapter.
4. In crystalline metallic solids, the defects are surfaces/interfaces, line defects and
point defects (vacancies and interstitials). Brief discussions on thermodynamics of
grain boundary, which is a special type of interface, and vacancies and interstitials
have been given in this chapter.
5. We have given derivations for dependence of equilibrium concentrations of vacancies
[ X, (eq)] and interstitials [X; (eq)] on temperature, i.e.
Xv(eq) =exp (- ~i)
X,(eq) =exp (- : )
where ,1J/v and .dHi are enthalpies of formation per mole of vacancies and interstitials,
respectively. We have also shown that Xv (eq) » X; (eq).
6. In ionic compounds, all the above defects are present. However, additional features
arise due to electrical charges associated with vacancies and interstitials, as well as
existence of electronic defects. In nonstoichiometric oxides, again, defect concentrations
depend on oxygen potential. In this chapter we have presented some thermodynamic
derivations.
Selected Thermodynamic Data
Table A-1 Molar Heat Capacities at Constant Pressure
Cp = a + bT + cr-2, joules/mol/K
Substance a bx 103 c x 10-5 Range, K
Al(s) 20.7 12.4 298-932

Al(I) 29.0 932-1273
Al203(S) 106.6 17.8 -28.5 298-1800
C(dia) 9.12 13.2 -6.19 298-1200
C(gr) 17.2 4.27 -8.79 298-2300
CO(g) 28.42 4.10 298-2500
C02(g) 44.16 9.04 298-2500
Cr(s) 24.4 9.87 -3.7 298- Tm
Cr203(s) 119.4 9.2 - 15.6 350-1800 l .... . J
FeO(s) 48.4 8.37 298-1200

Fe304(S) 91.6 201.8 298-900
Mn(a.) 21.6 15.9 298-993
Mn(~) 34.9 2.8 993-1373
Mn(y) 44.8 1373-1409
Mn(o) 47.3 1409-1517
Mn(I) 46.0 1517-Tb
MnO(s) 46.48 8.12 -3.7 298-1800
N1(g) 27.88 4.27 298-2500
Ni(s) 32.6 - 1.97 -5 .59 298-630
NiO(s) -20.9 157.2 16.3 298-525
(Comd.)
267
Table A-1 (Contd.)

Cp = a + bT + c12, joulcs/mol/K
Substance a b x 103 c x 10-5 Range, K
02(g) 30.0 4.18 -1.7 298-3000

Pb(s) 23.6 9.75 298-600
Ti( a) 22.09 10.04 298-1155
Ti(~) 28.91 1155-1350
TiC(s) 49.5 3.35 14.98 298-1800
Zn(s) 22.4 10.0 298-692
Zn(I) 31.4 692-1180
Zn(g) 20.8 1180-Tb
ZnO(s) 49.02 5.11 298-1600
Table A-2 Standard Molar Heats of Formation and Molar Entropies at 298 K
Substance L111f (298), joules ~98 , joules/K
Al(s) 0 28.34
Al203(s) -1,674,000 51.1
C(diamond) 1,900 2.44
C(graphite) 0 5.7
CO(g) - 111,700 198.0
C02(g) -394,100 213.9
Cr(s) 0 23.78
Cr203(s) -1,120,300 81.2
FeO(s) -259,600 58.81
Fe304Cs) -1,091,000 151.53
MnO(s) -384,700 59.12
Ni(s) 0 29.8
NiO(s) -244,600 38.0
02(g) 0 205.l
Ti(a) 0 30.56
TiC(s) -183,700 24.3
W(s) 0 33.48
W02(s) -589,800 66.95
W03(s) -844,300 83.3
ZnO(s) -348,300 43.53
Appendix
Table A-3 Molar Heats of Fusion and Transformations
Substance Transformation &f~, joules Tuans• K

Al s~I 10,500 932
Bi s~I 10,900 544
Cr s~ I 21,000 2173
Cu s ~I 12,970 1356
Mn a~ /J 2,010 993
Mn /J ~ r 2,300 1373
Mn r~ o 1,800 1409
Mn o~I 13,400 1517
Pb s~ I 4,810 600
Si s~I 50,630 1683
Ti a~ /J 3,473 1155
Zn s~I 7,388 692
Zn l~g 114,280 1180
Table A-4 Standard Free Energy Changes for Some Reactions
Reaction LtG°, joules Range, K

2Al(l) + 3/2 0 2(g) = Al203(s) - 1,676,000 + 320T 923-1800
CaC03(s) = CaO(s) + C02(g) - 126,400 + 144T 449-1500
C(s) + 2H2(g) = CH4(g) -69,120 + 22.25T In T - 65.34T 298-1200
C(s) + ~0 2(g) = CO(g) -11 1,700 - 87.65T 298-2500
C(s) + 02(g) = COz(g) -394,100 - 0.84T 298-2000
2Cr(s) + 3/2 0 2(g) = Cr20 3(s) -1, 120,800 + 260T 298-2100
2Cu(s) + Yz02(g) = Gu 20(s) - 169,000 - 7.12T In T + 123T 298-1356
2Cu(s) + ¥2S 2(g) = Cu 2S(s) - 142,900 - l l.3T In T + 120.2T 623-1360
Fe(s) + \/202 (g) = FeO(s) -259,600 + 62.55T 298-1642
Fe(s) + \/20 2(g) = FeO(I) -232,700 + 45.31T 1808-2000
3Fe(s) + 202(g) = Fe304(s) - 1,091,060 + 312.75T 298-1642
Fe(s) + \/2S 2(g) = FeS(I) - 113,400 + 26.45T 1465-1809
H2(g) + Clz(g) = 2HCJ(g) - 182,200 + 3.60T In T - 43.68T 298-2100
H2(g) + Yi02(g) = H20(g) -246,000 + 54.8T 298-2500
H2(g) + YiS2(g) = H 2S(g) -90,290 + 49.39T 298-2000
Mg(I) + Cl2(g) = MgCl2(1) -605,000 + 125.4T 973-1133
Mg(g) + Yi02(g) = MgO(s) -759,800 - 13.39T In T + 316.7T 1380-2500
Mn(s) + \/20 2(g) = MnO(s) -384,700 + 72.8T 298-1500
(Contd.)
.. . .
Table A-4 (Contd.)
Reaction L1G°, joules Range, K
Mn(!) + YlS 2(g) = MnS(I) -262,600 + 64.4T 1803-2000

Mo(s) + 0 2(g) = Mo0 2(s) -547,270 + 143T 298-1300
Yl02(g) = [0)1 wt% in Fe -116,740 - 6.llT 1823-1923
Si(!) = [Si)i wt% in Fe -119,240 - 52.5T 1823-1923
3Si(s) + 2N2(g) = Si3N4(s) -741,000 - 10.5T In T + 403T 298-1686
Si(s) + 02(g) = Si02(s) -902,000 + 174T 700-1700
Si(!) + 0 2(g) = Si02(s) -947,700 + 198.7T 1700-2000
Sn(!) + Cl 2(g) = SnCl 2(1) -333,000 + l 18T 520-925
Y2S2(g) + 02(g) = S02(g) -364,400 + 12.43T 718-2273
S02(g) + Y202(g) = S03(g) -94,560 + 89.37T 318-1800
Ti(s) + 0 2(g) = Ti02(s) -910,000 + 173T 298-2080
Zn(g) + Yl02(g) = ZnO(s) -482,920 - 18.8T In T + 344.7T 1170-2000
Zn(g) + YlS 2(g) = ZnS(s) -397,400 - 14.62T In T + 313.5T 1120-2000
Table A-5 Vapour Pressures of Some Metals
In p(atm) = -AIT + B In T + CT + D
Metal A B c x 103 D Range, K
Ag(l) 33,200 -0.85 20.31 Tm-Tb
Al(l) 37,884 -1.023 21.83 Tm-Tb
Cu(l) 40,350 -1.21 23.7 Tm-Tb
Sn(l) 35,697 12.32 Tm-Tb
Zn(I) 15,250 - 1.255 21.79 693- 1180
Table A-6 Atomic Mass of Some Elements
Al 27.0 c 12.0 Cd 112.4

Cr 52.0 Cu 63.5 Fe 55.8
Mn 54.94 0 16.0 Pb 207.2
Si 28.1 Ti 47.9 Zn 65.4
~
'
•
Bibliography
A. TEXTS
Darken, L.S., and Gurry, R.W., Physical Chemistry of Metals, McGraw-Hill, New York
(1953).
DeHoff, R.T., Thennodynamics in Materials Science, McGraw-Hill, New York (1993).
Denbigh, K., The Principles of Chemical Equilibrium, Cambridge University Press, Cambridge
(1963).
Gaskell, D.R., Introduction to the Thennodynamics of Materials, 3ro ed., Taylor & Francis,
Washington DC (1995).
Johnson, D.L., and Stracher, G.B., Thennodynamic Loop Applications in Materials Systems,
The Minerals, Metals and Materials Society, Warrendale, Pennsylvania (1995).
Kubaschewski, 0., Spencer, P.J. and Alcock, C.B., Materials Thennochemistry, 6th ed.,
Pergamon Press, Oxford (1993).
Swalin, R.A., Themwdynamics of Solids, John Wiley & Sons, New York (1964).
Upadhyaya, G.S., and Dube, R.K., Problems in Metallurgical Thennodynamics and Kinetics,
Pergamon Press, New York (1977).
B. DATA SOURCES
Elliott, J.F. and Gleiser, M., Themwchemistry for Steelmaking, Vol.I, Addison Wesley,
Reading, Mass, (1960).
Elliott, J.F., Gleiser, M., and Ramakrishna, V., Thennochemistry for Steelmaking, Vol. II,
Addison Wesley, Reading, Mass, (1963).
Hultgren, R., Orr, R.L., Anderson, P.O., and Kelley, K.K., Selected Values of Thennodynamic
Properties of Metals and Alloys, John Wiley & Sons, New York (1963).
JANAF Thermochemical Data, The Dow Chemical Co., Midland, Michigan (1962-63).
271
Kelley, K.K., Co11tributions to the Data 011 Theoretical Metallurgy XIII, Bulletin 584, U.S.
Govt. Printing Office, Washington DC (1960).
Knacke, 0., Kubaschewski , 0., and Hesselman, K., Thennochemical Properties of Inorganic
Substances, Vols. I and II, 2nd ed., Verlag Stahleisen mbH, Dusseldorf, Germany (1991).
Kubschewski, 0., Spencer, P.J., and Alcock, C.B., (already noted in TEXTS Section)
Schick, Harold L. Thennodynamics of Certain Refractory Compounds, Vol. 2, Academic
Press, New York (1966).
The Japan Society for the Promotion of Science, Steelmaking Data Source Book, The 19th
Committee on Steelmaking, The Gordon and Breach Science Publishers, New York (1984).
Answers to Problems
[Note: Unit of energy is either J (i.e. Joules) or kJ (i.e. kilojoules). Wherever in the
problem, the amount of substance is not mentioned, the values are per mole (i.e. per gm-
mole) of either a substance or per mole of reaction, as the case may be]
CHAPTER 2
1. Lili - L1U = -6424 J
2. (i) w = - 2331 J, q = - 1639 J
(ii) L1U = 692 J, Lili = 1153 J
3. (i) W = q = 11825 J, L1U = Lili = 0 for 151 stage
(ii) W = 13317 J, q = 0, L1U = - 13317 J, Lili= -18800 J
4. (a) W = 2748 J, q = 0, L1U = - 2748 J, Lili = - 3848 J
(b) W = 4539 J, q = Lili = 15888 J, L1U = 11349 J
S. W(step iii) = 4.51 kJ, q(for cycle) = - 8.5 kJ
CHAPTER 3
1. 59.9 kJ
2. -42.2 kJ
3. - 186.7 kJ
4. -404.9 kJ
S. Llli0 = - 288 .2 kJ, L1U0 = - 284.0 kJ
6. Process (ii) releases 906 J more heat.
7. - 562.5 kJ
8. 1589 K
9. 349.4 kJ
CHAPTER 4
1. L1S(syst) = - 53.26 J/K, L1S(surr) = 625 J/K
2. (a) 18.1 J/K; (b) O; (c) -28.72 J/K
273
3. -171.61/K
4. - 13.64 J/K
5. L1S(syst) =- 11.64 J/K, L1S(surr)= 12.84 J/K
6. .d.S'°(reaction) = &(syst) = -89.39 J/K, L1S(universe) = L1S(syst + surr) = 198.8 J/K
7. (a) 3589 K; (b) 0.0075 Cp; (c) process spontaneous
8. 118.0 J/K.
CHAPTER 6
1. - 1729 J
2. 8586 J
3. 16.46 J, No
4. -2.42 K
5. 78450 atm
6. (i) 260.6 kJ; (ii) 7.07 J/K
7. 908 atm; as the triple point pressure is 5.14 atrn, the 1-atm isobar does not pass
through the liquid field .
8. 99.95%.
CHAPTER 7
1. 8.5 x 108
2. - 2707 J
3. (i) 3.0 J/K; (ii) - 131.1 J
4. &J° = - 242.7 kJ, L1U0 = - 240.6 kJ, L1S0 = -89.73 J/K, L1G0 = -197.8 kJ
5. L1H° =- 602.4 kJ, L1S° =- 357 .2 J/K
6. Ni shall not be oxidized; L1Gj (NiO) = 108.8 kJ,
L1Gj (Cr20 3) =- 171 .5 kJ per mole of 0 2
7. Cr23C 6 stabler than Cr
8. L1H° = 30.42 kJ; & 0 = 65.9 J/K; L1G0 = 7.35 kJ
9. (i) 7.1 x 107 , 6.4 x HP; (ii) Lllij = - 234.4 kJ, L1Sj = - 84.1 J/K
CHAPTER 8
1. No oxidation
2. H 2S "' 0, 0 2 = 25, H;?O = 50, S 2 = 25 volume %
3. (a) 4.56 x 10-8 atm; (b) Titanium will get oxidized.
4. 43,800 J heat evolved
5. 0 .371
6. 5250 atm
7. (i) 51 %; (ii) No change
8. Conversion of Cu 20 into Cu2S possible
9. Not at equilibrium
10. (i) 1166 K; (ii) 0 .053 atm; (iii) 1.23 atm
(Note: Since 1500 K > 1166 K, no CaC03 will be present and Pc0z. at 1500 K will
be governed by Gas Law.)
Answers to Problems
11. Fe304 , 836 K

12. MnS will form.
13. Po 2 = 1.06 x 10-s atm, PH2s = 0.0151 atm, Psi = 0.356 atm, pz.n = 2.61 atm
CHAPTER 9
V,m = Kf[b'(ln
B b'P
rB)]
T, caqi
1e = Henry' s Law constant, and is not a function of composition.

2. 21000 J, 23.0 J!K
3. (i) Psn = 6.3 x 10-7 atm, Peu. = 3.68 x 10-7 atm
(ii) ~n = 0.833, 1tu = 0.603
(iii) (;;0 = - 12800 J, %
= -13046 J
(iv) .dGm = -12950 J, G = - 5ll7 J
4. PAI = 2.64 x 10-10 atm, PAg = 7.1 x 10~ atm; since PAI « PAg• removal of Al not
feasible.
S. (i) No
(ii) yshould be constant in Henry's Law region; X80 < 0.1, Xeu < 0.1 for validity
of HL for Sn, Cu, respectively.
(iii) and (v) Pso = 4.6 x 10- 10 atm, Peu. = 4.5 x 10-11 atm
GEi = - 4955 J, ~ = - 13922 J, .dGm = - 8542 J, Gxs = - 2630 J
Not regular solution since as0 , QQi depend on composition.
6. (ii)
Values in J/mole
Gxs
T, K ~ G'Ag .6Gm acu a Ag
1200 - 3558 - 3775 - 3666 3249 1.346 1.26
1300 - 4168 - 4491 - 4329 3162 1.23 1.11
1400 - 4836 -5104 -4970 3097 1.11 1.02
(iii) and (iv) Values in J/mole for H , in J/mole/K for S

fl~ = fleu = 980, ii'A.8 = 785, .&r = Hxs = 882
SC:, = 1.52, S,.\l! = 1.48, M m = 1.50

S&,5 = -1.36, s;: = - 1.40, S XS = - 1.38
(v) Since sxs * 0, not a regular solution .
7. (ii) GXS = -4620 J, L1G111 = - 12453 J
9. (i) In /Zn = 2.00 (I - Xz0 ) 2 - 0.691 (1 - Xz.n) 3 ; (ii) 3.70; (iii) 506 J
CHAPTER 10
1. }lc(gas) = - 46.24 kJ/mole C ; cannet saturate steel with carbon.
2. F = l ; 11N2 = - 205.3 kJ; stable phases at 1200 K: Si(s) + gas
3. 3
4. Psi = - 126.3 kJ; Cu will be converted to CuiS.
s. - 320.5 kJ
6. L1G';. = 13f 10 - 1.3T Jn T + 2.554T J
7. (i) 0.003 p1Jm (i.e. 3 ppb); (ii) Al4C3 will form.
8. 1eu (/J) =29.5
CHAPTER 11
1. (i) 0.396; (ii) Proi = 0.049 atm, Pco = 0.951 atm
2. 0.846
3. 1002
4. 1.35 x 10-5
S. [XZn] in Pb = 2.4 x 10-5
6. (i) 2.13 x 10-5 atm; (ii) 1.14 x 10- 3 atm
7. (i) XFe0 = 0.001, XMoO = 0.999; (ii) 2.6 x 10-27 atm
8. -50.0 kJ per mole 3Ca0.Al20 3; 9.79 x 10- 3
9. (i) 1.3; (ii) -163 kJ
10. (i) -36.8 kJ/mole Cu; (ii) -69.6 kJ/mole Mn
11. -0.341
12. 0.0125
13. 0.09 wt.%
14. -567.8 kJ
15. -36.39 kJ
CHAPTER 13
1. LlG = -314.l kJ, &I = -30.l kJ, LlS = 44.l J/K
2. -0.022T J/mole/K
3. 2.47 volts
4. aA1 = 0.673, GAj = -2152 J, H'fi =3315 J
S. (i) asn = 0.43, Gfn = -8414 J, Gl-5 = -1499 I
0
(ii) Ss~ = 11.74 J/K, .SS";; = 5.98 J/K, Hg:, = B:; = 5673 I
6. 0.231 volt
7. (i) - 127.0 kJ; (ii) -177.6 kJ
8. (i) 0.813; (ii) 0.70
Index
A Chemical thermodynamics, 2
Clapeyron equation, 82
Activity, 96 Classical thermodynamics, 1-3
coefficient, 131 Clausius-Clapeyron equation, 87, 120
composition and, 131 , 137 Closed system equations, 65
free energy and, 100-101 Composition parameters, 98
fugacity and, 100 Compressibility, 67
Henrian, 201-202 Configurational entropy, 225
ionic, 191 Constant pressure processes, 15-18
quotient, 101 Constant volume processes, 15-18
Adiabatic processes, 16 Cyclic processes, 11
expansion in, 19
reversible, 20 D
Adsorption isotherms, 250-254
Alloy phase diagrams, 169 Dalton's Law, 98
Alpha function , 145 Daniel cell, 229
Avogadro number, 7, 28, 219-226, 259 Debye equation, 28
Defect thermodynamics, 256-264
B Degrees of freedom, 165
Binary phase diagrams, 169-183 E

Binary solutions
entropy of mixing, 225 Efficiency of heat engines, 45-50
general, 137-147 Einstein equation, 27
regular, 149- 150 Electrical work, 231
Boltz mann constant, 7, 28, 219-226 Electrochemical reaction, 229
Boltzmann equation, 224 Electrochemical series, 242
Electrolyte, 230
c Electromotive force (EMF), 211 -240
Ellingham diagram, 107-110
Calorimeter, 35 Enthalpy
Carnot cycle, 45- 50 bond energies and, 152
Chemical potential, 159-163 changes in, 27-35
.]
fi:I Index
entropy and, 54, 64 First Law, 9

First Law and, 16 Second Law and, 64
free energy and, 64, 80 Free energy
heat capacity and, 30 activity and, 100-101
heat effects and, 29 chemical potential and, 160
phase changes and, 30, 175 condensed solutions and, 136-141
temperature and, 30 definition, 64
Entropy enthalpy and, 64
Boltzmann equation, 224 equilibrium and, 70, 72
,,'
changes in, 53-56 fugacity and, 76, 100
configurational, 225 phase diagram and, 173
derivation, 51 pressure and, 55
l
disorder and, 57, 216 stability and, 72, 109
enthalpy and, 54 standard states for, 78, 100
freezing and, 54, 60 surface, 248
heat capacity and, 53 temperature and, 80, 102
heat effects and, 52 Free energy change for
irreversibility and, 58-59 isothermal processes, 77, 80
mixing and, 55 melting, 173-175
physical significance of. 56-57, 223 mixing, 99
reversibility and, 52, 58-59 reactions, 101
Second Law and, 57-61 Fugacity, 76-79
thermal, 225
Third Law and, 215 G
Equilibrium
chemical, 101 Galvanic cell, 230-238
chemical potential and, 161 Gases
criteria for, 70 gas constant, 8
electrochemical, 233 ideal gas law, 4
free enrgy and, 70, 72 non-ideal, 78
gaseous, 116 reactions in, 116-119
gas-solid/liquid, 120-127 Gibbs adsorption isotherm, 252
oxidation-reduction, 121 Gibbs-Duhem equation
phase transformation and, 82, 161 derivation, 133
solid-liquid, 82 integration, 142
solid-solid, 169-171 Gibbs free energy (see free energy)
temperature and, 101, 107 Gibbs-Helmholtz equation, 80, 132
thermodynamic, 5 Gibbs ph:ise rule, 164
vapour-liquid/solid, 87
. Equilibrium constant, 106
Equilibrium state, 5, 70
H
Excess function , 148 Heat capacity, 18, 26-29

Expansion work, 13 enthalpy and, 30
Extensive property/variable, 5 entropy and, 53
relations, 67
F relations for ideal gases, 19-20
temperature and, 27-29
Faraday constant, 7 Heat engines, 45-50
Index
Heat of formation, 35 Melting

Helmholtz free energy, 64 enthalpy of, 30
Henrian standard state, 202 entropy of, 54
Henry's Law, 137-138 Metal-oxygen sysetm, 107, 121, 189
Hess' Law, 33 Metal
carbon-oxygen system, 123, 186
I gas solubility in, 209
oxides, 107, 121, 189
Ideal gas, 4, 19, 26, 46, 116, 218 Metallurgical thermodynamics, 3
Integral molar properties, 98, 132 Metastable system, 91
Intensive properties, 5 Microstate, 222
Interaction coefficients, 197, 204 Mixing in solutions
Interfaces, 257 enthalpy of, 33, 133-156
Internal energy, 9, 219-223 entropy of, 55, 133-140, 229
Interstitials, 258-264 free energy of, 133-150, 173-181
Ionic compounds, 261 Mixture, 96
Iron-oxygen system, 189 Mole fraction, 98
Irreversible processes, 12 Multi-component system, 96, 197
criteria for, 71 Multi-phase system, 162-168, 204
entropy and, 57-59
Isobaric processes, 15 N
Isochoric processes, 15
Isothermal processes Natural processes, 45
irreversible, 33 Nernst equation, 243
reversible, 12- 15 Nernst theorem, 213
Non-ideal solution, 131
J Non-stoichiometric compound, 189, 263
Joule, J.P., 16 0
K One-component system, 75-95

One wt.pct. standard state, 201
Kelvin, 44, 50 Oxidation potential, 241
Kinetic theory, 218-219 Oxidation reactions
Kirchhoffs Law, 37 Ellingham diagram, 107-111
for carbon, 109, 124, 186
L Oxygen potential, 163, 237-243
Langmuir adsorption isotherm, 251 p

Latent heat, 30
Loop concept, 37 Partial molar properties, 98, 132
Partition function , 224
M Phase
diagrams, 84, 168
Macrostate, 222 equilibria, 81
Maximum work, 70 stability diagrams, 126, 186
Maxwell equation , 66 transformations, 3 J
Index
Phase rule, 164 quasi-chemical theory of, 152

Physico-chemical processes, 97 regular, 147, 153
Planck, M., 44, 215, 219 statistical model of, 225
Polytropic processes, 15, 21 Spontaneous processes, 44-45
Pourbaix diagrams, 243 Standard atmosphere, 8
Pressure Standard state
enthalpy and, 16 alternative, 173-175
entropy and, 65 free energy and, 203
equilibrium and, 70, 100 Henrian, 201
free energy and, 64-65 one wt.pct., 201
Process paths, 12, 16 Raoultian, 131
P-V-T relations, 4, 79 State properties/variables, 5
Statistical mechanics, 219
Q Statistical thermody namics, 1, 218, 222
Stirling theorem, 223 , 227
Quantum theory, 221 Sublimation , 86
Quasi-chemical theory, 152 Supercooled liquid, 59-60
Surface thermodynamics, 249-256
R Surrounding- definition, 4
System
Rao ult's Law, 131 characterization, 5
standard state and, 131 definition, 4
Reaction equilibria, 101 , 186- 197
Reduction potentials, 241 T
Regular solution, 147-154
Temperature
binary system, 149, 171
absolute, 50
Reversible processes, 12-16
EMF and, 234
elec:ric potential of, 23 1
enthalpy and, 30
equations for, 65
entropy and, 54
free energy and, 65 equilibrium and, 101, 107
free energy and, 80, 102-105
s heat capacity and, 27-29
Thermal expansion coefficient, 67-69
Saturation, 87, 177-182, 191, 230 Thermo-elasticity, 68
Second Law, 44 Third Law, 213-218
and entropy, 51-61 en tropy and, 215
Sievcrts' Law, 209 verificati on of, 217
Si-C-0 sys tem, 186 Triple point, 97
Single electrode potential, 241
Solid elctrol ytes, 230, 237 v
Solutions, 96, 130-154
atomic bonds in, 132, 152 Vacancies, 258-264
binary, 137-146, 149, 225 Van't Hoff equation, 106
excess function, 148 Vapour pressure, 87
free energy of, 132-1 42
Henry's Law,. 137 w
ideal, 131 , 136
non-ideal, 131 Wustite, 189-192

Metallurgical Thermodynamics - Ghosh

Uploaded by

Copyright:

Available Formats

Metallurgical Thermodynamics - Ghosh

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Metallurgical Thermodynamics - Ghosh

Uploaded by

Copyright:

Available Formats

•

Textbook of Materials and

PHI Learning [;)LJ1WOO llikmBGGGJ

TEXTBOOK OF MATERIALS AND METALLURGICAL THERMODYNAMICS

Seventh Printing September, 2015

2. FIRST LAW OF THERMODYNAMICS 9-25

3. HEAT CAPACITY AND ENTHALPY-AUXILIARY RELATIONS

4. SECOND LAW OF THERMODYNAMICS AND ENTROPY 44-63

5. AUXILIARY FUNCTIONS AND RELATIONS, CRITERIA FOR

5.4 Criteria for Thermodynamic Equilibria 70

6. GIBBS FREE ENERGY AND ONE-COMPONENT SYSTEMS 75-95

7. ACTIVITY, EQUILIBRIUM CONSTANT AND STANDARD

7.5 Ellingham Diagrams 107

8. EQUILIBRIA INVOLVING IDEAL GASES AND

9. THERMODYNAMICS OF SOLUTIONS 130-158

9.8 Quasi-chemical Theory of Solution 152

10. CHEMICAL POTENTIAL AND EQUILIBRIA AMONGST

11. REACTION EQUILIBRIA INVOLVING CONDENSED PHASES

11.4.2 Equality of ef and ej 199

12. TillRD LAW OF THERMODYNAMICS, STATISTICAL

13. THERMODYNAMICS OF ELECTROCHEMICAL CELLS 228-247

13.4 Single Electrode Potentials and Electrochemical Series in

14. THERMODYNAMICS OF SURFACES, INTERFACES

Appendix Selected Thermodynamic Data 267-270

Answers to Problems 273-276

Metallurgical thermodynamics, as also its modified version, 'Thermodynamics of Materials',

A Helmholtz free energy, surface area J or J mo1- 1; m 2

m mass of one molecule/atom g molecule- 1

w work done J, m3 atm

8 symbol for partial differential (e.g. 8q, SW)

1.1 Historical Perspective

1. Classical (I.e. macroscopic or phenomenological) thermodynamics

Another way of classifying thermodynamics is:

1.2 History of Classical Thermodynamics

1.2.2 Chemical Thermodynamics

1.2.3 Metallurgical Thermodynamics

Broadly speaking, the application of primarily chemical thermodynamics to metals and

1.3 Introductory Concepts and Definitions

1.3.1 System and Surrounding

1.3.2 State of a System

1.3.3 Thermodynamic Equilibrium

1.3.4 Characterization of Systems-Some elucidations

1.4 Values of Some Physical Constants

Faraday's constant (F) = 96,500 coulombs/g-equivalent (i.e. joules/volt/g-equivalent)

2.1 Internal Energy

2. potential energy of the system due to its position in a force field,

2.2 Statement of the First Law

2.2.1 Significance of the First Law

2.2.2 Proof of the First Law

1:.q- 1:.W=O (2.3)

Invoking properties of exact differentials, from Eqs. (2.4)-(2.6), respectively,

2.3 Reversible Processes

Now, consider isothermal expansion-compression of n moles of an ideal gas. Figure 2.3

OW = Pext • A • dl = Pe,. 1 • dV (2.10)

where P denotes gas pressure inside the system. Hence,

OW= PdV (2.12)

because U is a state property. From Eqs. (2.1), (2.15), and (2.16), I: q = 0.

2.4 Calculation of Wand q for Reversible Processes

W = P(V2 - V1) (2.21)