Buku Kimia Fisika PDF

P1: OTA/XYZ P2: ABC
fm JWBS043-Rogers October 8, 2010 21:3 Printer Name: Yet to Come
CONCISE PHYSICAL
CHEMISTRY
DONALD W. ROGERS
Department of Chemistry and Biochemistry
The Brooklyn Center
Long Island University
Brooklyn, NY
A JOHN WILEY & SONS, INC., PUBLICATION

P1: OTA/XYZ P2: ABC
P1: OTA/XYZ P2: ABC
CONCISE PHYSICAL CHEMISTRY

P1: OTA/XYZ P2: ABC
P1: OTA/XYZ P2: ABC
CONCISE PHYSICAL
CHEMISTRY
DONALD W. ROGERS
Department of Chemistry and Biochemistry
The Brooklyn Center
Long Island University
Brooklyn, NY
A JOHN WILEY & SONS, INC., PUBLICATION

P1: OTA/XYZ P2: ABC
Copyright
C 2011 by John Wiley & Sons, Inc. All rights reserved.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or
by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as
permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior
written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to
the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400,
fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission
should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken,
NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in
preparing this book, they make no representations or warranties with respect to the accuracy or
completeness of the contents of this book and specifically disclaim any implied warranties of
merchantability or fitness for a particular purpose. No warranty may be created or extended by sales
representatives or written sales materials. The advice and strategies contained herein may not be suitable
for your situation. You should consult with a professional where appropriate. Neither the publisher nor
author shall be liable for any loss of profit or any other commercial damages, including but not limited to
special, incidental, consequential, or other damages.
For general information on our other products and services or for technical support, please contact our
Customer Care Department within the United States at (800) 762-2974, outside the United States at
(317) 572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may
not be available in electronic formats. For more information about Wiley products, visit our web site at
www.wiley.com.
Don Rogers is an amateur jazz musician and painter who lives in Greenwich Village, NY.
Library of Congress Cataloging-in-Publication Data:
Rogers, Donald W.
Concise physical chemistry / by Donald W. Rogers.
p. cm.
Includes index.
Summary: “This book is a physical chemistry textbook that presents the essentials of physical
chemistry as a logical sequence from its most modest beginning to contemporary research topics. Many
books currently on the market focus on the problem sets with a cursory treatment of the conceptual
background and theoretical material, whereas this book is concerned only with the conceptual
development of the subject. It contains mathematical background, worked examples and problemsets.
Comprised of 21 chapters, the book addresses ideal gas laws, real gases, the thermodynamics of simple
systems, thermochemistry, entropy and the second law, the Gibbs free energy, equilibrium, statistical
approaches to thermodynamics, the phase rule, chemical kinetics, liquids and solids, solution chemistry,
conductivity, electrochemical cells, atomic theory, wave mechanics of simple systems, molecular orbital
theory, experimental determination of molecular structure, and photochemistry and the theory of
chemical kinetics”– Provided by publisher.
ISBN 978-0-470-52264-6 (pbk.)
1. Chemistry, Physical and theoretical–Textbooks. I. Title.
QD453.3.R63 2010
541–dc22
2010018380
Printed in Singapore
10 9 8 7 6 5 4 3 2 1
P1: OTA/XYZ P2: ABC
CONTENTS
Foreword xxi
Preface xxiii
1 Ideal Gas Laws 1
1.1 Empirical Gas Laws, 1
1.1.1 The Combined Gas Law, 2
1.1.2 Units, 2
1.2 The Mole, 3
1.3 Equations of State, 4
1.4 Dalton’s Law, 5
Partial Pressures, 5
1.5 The Mole Fraction, 6
1.6 Extensive and Intensive Variables, 6
1.7 Graham’s Law of Effusion, 6
Molecular Weight Determination, 6
1.8 The Maxwell–Boltzmann Distribution, 7
Figure 1.1 The Probability Density for Velocities of Ideal
Gas Particles at T = 0., 8
Boltzmann’s Constant, 8
Figure 1.2 A Maxwell–Boltzmann Distribution Over
Discontinuous Energy Levels., 8
1.9 A Digression on “Space”, 9
Figure 1.3 The Gaussian Probability Density Distribution
in 3-Space., 10
The Gaussian Distribution in 2- and 3- and 4-Space, 10
v
P1: OTA/XYZ P2: ABC
vi CONTENTS
1.10 The Sum-Over-States or Partition Function, 10

Figure 1.4 The Probability Density of Molecular Velocities
in a Spherical Velocity Space., 12
Problems and Exercises, 12
Exercise 1.1, 12
Exercise 1.2, 13
Problems 1.1–1.13, 15–16
Computer Exercise 1.14, 16
Problems 1.15–1.18, 16–17
2 Real Gases: Empirical Equations 18

2.1 The van der Waals Equation, 18
2.2 The Virial Equation: A Parametric Curve Fit, 19
2.3 The Compressibility Factor, 20
Figure 2.1 A Quadratic Least-Squares Fit to an
Experimental Data Set for the Compressibility Factor of
Nitrogen at 300 K and Low Pressures (Sigmaplot 11.0C )., 21
File 2.1 Partial Output From a Quadratic Least-Squares
Curve Fit to the Compressibility Factor of Nitrogen at
300 K (SigmaPlot 11.0C )., 22
Figure 2.2 The Second Virial Coefficient of Three Gases as
a Function of Temperature., 22
2.3.1 Corresponding States, 23
Figure 2.3 The Z = f ( p) Curve for Two Different Gases or
for the Same Gas at Two Different Temperatures., 23
2.4 The Critical Temperature, 24
Figure 2.4 Three Isotherms of a van der Waals Gas., 24
Figure 2.5 Conversion of a Liquid to Its Vapor Without
Boiling (1–4)., 25
2.4.1 Subcritical Fluids, 25
2.4.2 The Critical Density, 26
Figure 2.6 Density ρ Curves for Liquid and Gaseous
Oxygen., 26
2.5 Reduced Variables, 27
2.6 The Law of Corresponding States, Another View, 27
Figure 2.7 Compressibility Factors Calculated from the
van der Waals Constants., 28
2.7 Determining the Molar Mass of a Nonideal Gas, 28
Problems and Exercises, 28
Exercise 2.1, 28
Figure 2.8 Boyle’s Law Plot for an Ideal Gas (lower curve)
and for Nitrogen (upper curve)., 29
Exercise 2.2, 30
P1: OTA/XYZ P2: ABC
CONTENTS vii
Table 2.1 Observed Real Gas Behavior from 10 to 100 bar

Expressed as ( p, pVm )., 30
Figure 2.9 Experimental Values of pVm = z( p) vs. p for a
Real Gas., 30
Table 2.2 Observed Real Gas Behavior Expressed
as ( p, pVm )., 31
Figure 2.10 Quadratic Real Gas Behavior., 32
Problems 2.1–2.15, 32–34
Figure 2.11 Cubic Real Gas Behavior., 34
3 The Thermodynamics of Simple Systems 35

3.1 Conservation Laws and Exact Differentials, 35
3.1.1 The Reciprocity Relationship, 36
3.2 Thermodynamic Cycles, 37
Figure 3.1 Different Path Transformations from A to B., 38
3.2.1 Hey, Let’s Make a Perpetual Motion Machine!, 38
3.3 Line Integrals in General, 39
Figure 3.2 Different Segments of a Curved Rod., 39
3.3.1 Mathematical Interlude: The Length of an Arc, 40
Figure 3.3 Pythagorean Approximation to the Short
Arc of a Curve., 40
3.3.2 Back to Line Integrals, 41
3.4 Thermodynamic States and Systems, 41
3.5 State Functions, 41
3.6 Reversible Processes and Path Independence, 42
Figure 3.4 The Energy Change for Reversible Expansion of
an Ideal Gas., 43
3.7 Heat Capacity, 44
3.8 Energy and Enthalpy, 44
3.9 The Joule and Joule–Thomson Experiments, 46
Figure 3.5 Inversion Temperature Ti as a Function of
Pressure., 47
3.10 The Heat Capacity of an Ideal Gas, 48
Table 3.1 Heat Capacities and γ for Selected Gases., 48
Figure 3.6. Typical Heat Capacity as a Function of
Temperature for a Simple Organic Molecule., 50
3.11 Adiabatic Work, 50
Figure 3.7 Two Expansions of an Ideal Gas., 51
Problems and Example, 52
Example 3.1, 52
Problems 3.1–3.12, 52–55
Figure 3.8 C = Diagonal Along x = 1 to y = 1., 53
Figure 3.9 C = Quarter-Circular Arc., 53
P1: OTA/XYZ P2: ABC
viii CONTENTS
4 Thermochemistry 56
4.1 Calorimetry, 56
4.2 Energies and Enthalpies of Formation, 57
4.3 Standard States, 58
4.4 Molecular Enthalpies of Formation, 58
Figure 4.1 Combustion of C(gr) and CO(g)., 59
Figure 4.2 A Thermochemical Cycle for Determining
f H 298 (methane)., 60
4.5 Enthalpies of Reaction, 60
4.6 Group Additivity, 62
4.7 f H 298 (g) from Classical Mechanics, 64
4.8 The Schrödinger Equation, 64
4.9 Variation of H with T, 65
4.10 Differential Scanning Calorimetry, 66
Figure 4.3 Schematic Diagram of the Thermal Denaturation
of a Water-Soluble Protein., 67
Example 4.1, 68
Problems 4.1–4.9, 68–70
5 Entropy and the Second Law 71

5.1 Entropy, 71
Figure 5.1 An Engine., 72
5.1.1 Heat Death and Time’s Arrow, 73
5.1.2 The Reaction Coordinate, 73
5.1.3 Disorder, 74
5.2 Entropy Changes, 74
5.2.1 Heating, 74
5.2.2 Expansion, 75
5.2.3 Heating and Expansion, 75
5.3 Spontaneous Processes, 77
5.3.1 Mixing, 77
5.3.2 Heat Transfer, 77
5.3.3 Chemical Reactions, 78
5.4 The Third Law, 78
5.4.1 Chemical Reactions (Again), 79
Example 5.1, 80
Figure 5.2 C p /T vs. T for Metallic Silver Ag(s)., 81
Problems 5.1–5.9, 81–83
6 The Gibbs Free Energy 84

6.1 Combining Enthalpy and Entropy, 84
6.2 Free Energies of Formation, 85
P1: OTA/XYZ P2: ABC
CONTENTS ix
6.3 Some Fundamental Thermodynamic Identities, 86

6.4 The Free Energy of Reaction, 87
6.5 Pressure Dependence of the Chemical Potential, 87
Figure 6.1 A Reaction Diagram for G 4 ., 88
6.5.1 The Equilibrium Constant as a Quotient of Quotients, 88
6.6 The Temperature Dependence of the Free Energy, 88
Example 6.1, 90
Problems 6.1–6.12, 90–92
7 Equilibrium 93
7.1 The Equilibrium Constant, 93
7.2 General Formulation, 94
7.3 The Extent of Reaction, 96
7.4 Fugacity and Activity, 97
7.5 Variation of the Equilibrium Constant with Temperature, 97
The van’t Hoff Equation, 98
7.5.1 Le Chatelier’s Principle, 99
7.5.2 Entropy from the van’t Hoff Equation, 99
7.6 Computational Thermochemistry, 100
7.7 Chemical Potential: Nonideal Systems, 100
7.8 Free Energy and Equilibria in Biochemical Systems, 102
7.8.1 Making ATP, the Cell’s Power Supply, 103
Problems and Examples, 104
Example 7.1, 104
Example 7.2, 105
Problems 7.1–7.7, 105–106
8 A Statistical Approach to Thermodynamics 108

8.1 Equilibrium, 108
Figure 8.1 A Two-Level Equilibrium., 109
Figure 8.2 A Two-Level Equilibrium., 109
8.2 Degeneracy and Equilibrium, 109
Figure 8.3 A Degenerate Two-Level Equilibrium., 110
Figure 8.4 A Degenerate Two-Level Equilibrium., 110
Figure 8.5 A Two-Level Equilibrium with Many A and
Many B Levels., 111
8.3 Gibbs Free Energy and the Partition Function, 112
8.4 Entropy and Probability, 113
8.5 The Thermodynamic Functions, 113
Table 8.1 Thermodynamic Functions (Irikura, 1998)., 114
8.6 The Partition Function of a Simple System, 114
8.7 The Partition Function for Different Modes of Motion, 116
P1: OTA/XYZ P2: ABC
x CONTENTS
8.8 The Equilibrium Constant: A Statistical Approach, 117

8.9 Computational Statistical Thermodynamics, 119
Table 8.2 Some Computed Partition Functions for
Molecular and Atomic Sodium., 120
Example 8.1, 120
Example 8.2, 121
Problems 8.1–8.9, 122–123
9 The Phase Rule 124

9.1 Components, Phases, and Degrees of Freedom, 124
9.2 Coexistence Curves, 125
Figure 9.1 Pure Water in One Phase (left) and Two Phases
(right)., 126
Figure 9.2 A Liquid–Vapor Coexistence Curve., 127
Figure 9.3 A Single-Component Phase Diagram., 128
9.3 The Clausius–Clapeyron Equation, 128
9.4 Partial Molar Volume, 129
Figure 9.4 Total Volume of an Ideal Binary Solution., 130
Figure 9.5 Volume Increase (or Decrease) Upon Adding
Small Amounts of Solute n 2 to Pure Solvent., 130
9.4.1 Generalization, 130
Figure 9.6 Partial Molar Volume as the Slope of
V vs. n 2 ., 131
Figure 9.7 Volume Behavior of a Nonideal Binary
Solution., 131
9.5 The Gibbs Phase Rule, 134
9.6 Two-Component Phase Diagrams, 134
9.6.1 Type 1, 135
Figure 9.8 A Type I Phase Diagram., 135
9.6.2 Type II, 136
Figure 9.9 A Type II Phase Diagram., 135
9.6.3 Type III, 137
Figure 9.10 A Type III Phase Diagram., 137
9.7 Compound Phase Diagrams, 137
Figure 9.11 A Compound Phase Diagram with a Low
Boiling Azeotrope., 138
9.8 Ternary Phase Diagrams, 138
Figure 9.12 A Ternary Phase Diagram with a Tie Line., 139
Example 9.1, 139
Figure 9.13 The Liquid–Vapor Coexistence Curve of Water
Leading to vap H (H2 O) = 44.90kJmol−1 ., 140
P1: OTA/XYZ P2: ABC
CONTENTS xi
Example 9.2, 140

Figure 9.14 A Ternary Phase Diagram in which B and C Are
Partially Miscible., 141
Problems 9.1–9.9, 141–143
10 Chemical Kinetics 144

10.1 First-Order Kinetic Rate Laws, 144
Figure 10.1 First-Order Radioactive Decay., 146
Figure 10.2 Logarithmic Decay of a Radioactive Element., 147
10.2 Second-Order Reactions, 147
10.3 Other Reaction Orders, 149
10.3.1 Mathematical Interlude: The Laplace Transform, 149
10.3.2 Back to Kinetics: Sequential Reactions, 150
10.3.3 Reversible Reactions, 151
10.4 Experimental Determination of the Rate Equation, 154
10.5 Reaction Mechanisms, 154
10.6 The Influence of Temperature on Rate, 156
Figure 10.3 An Activation Energy Barrier Between an
Unstable Position and a Stable Position., 156
Figure 10.4 Enthalpy Level Diagram for an Activated
Complex [B]., 157
Figure 10.5 An Activation Barrier., 157
Figure 10.6 A Boltzmann Distribution of Molecular Speeds., 158
10.7 Collision Theory, 158
10.8 Computational Kinetics, 159
Example 10.1, 160
Example 10.2, 160
Figure 10.7 First-Order Fluorescence Decline from
Electronically Excited Iodine in Milliseconds., 161
Figure 10.8 The Natural Logarithm of Relative Intensity vs.
Time for Radiative Decay., 161
Problems 10.1–10.10, 162–164
11 Liquids and Solids 165

11.1 Surface tension, 165
Figure 11.1 Intermolecular Attractive Forces Acting Upon
Molecules at an Air–Water Interface., 166
Figure 11.2 Stretching a Two-Dimensional Membrane by
Moving an Edge of Length l., 166
Figure 11.3 Stretching a Two-Dimensional Liquid
Bimembrane., 167
Figure 11.4 Capillary Rise in a Tube of Radius R., 167
P1: OTA/XYZ P2: ABC
xii CONTENTS
11.2 Heat Capacity of Liquids and Solids, 168

Figure 11.5 Heat Capacity as a Function of Temperature., 168
11.3 Viscosity of Liquids, 169
Figure 11.6 Approximation of Laminar Flow Inside a Tube., 169
11.4 Crystals, 170
Figure 11.7 Close Packing of Marbles Between Two Sheets., 171
Figure 11.8 A Less Efficient Packing of Marbles., 172
Figure 11.9 Bragg’s Law for Constructive Reflection., 173
11.4.1 X-Ray Diffraction: Determination of Interplanar
Distances, 173
Figure 11.10 A Face-Centered Cubic Unit Cell., 174
11.4.2 The Packing Fraction, 174
Figure 11.11 A Two-Dimensional Unit Cell for
Packing of Discs., 175
Figure 11.12 A Simple Cubic Cell., 175
11.5 Bravais Lattices, 176
Table 11.1 The Bravais Crystal Systems and Lattices., 176
11.5.1 Covalent Bond Radii, 176
11.6 Computational Geometries, 177
11.7 Lattice Energies, 177
Problems and Exercise, 178
Exercise 11.1, 178
Figure 11.13 The Born–Haber Cycle for NaI., 179
Problems 11.1–11.8, 179–181
Figure 11.14 Close Packing (left) and Simple Square Unit
Cells (right)., 180
Figure 11.15 A Body-Centered Primitive Cubic Cell., 180
12 Solution Chemistry 182

12.1 The Ideal Solution, 182
Figure 12.1 Entropy, Enthalpy, and Gibbs Free Energy
Changes for Ideal Mixing at T > 0., 183
12.2 Raoult’s Law, 183
Figure 12.2 Partial and Total Pressures for a Raoult’s
Law Solution., 184
12.3 A Digression on Concentration Units, 184
12.4 Real Solutions, 185
Figure 12.3 Consistent Positive Deviations from
Raoult’s Law., 185
12.5 Henry’s Law, 186
Figure 12.4 Henry’s Law for the Partial Pressure of
Component B as the Solute., 186
12.5.1 Henry’s Law Activities, 186
P1: OTA/XYZ P2: ABC
CONTENTS xiii
12.6 Vapor Pressure, 187

12.7 Boiling Point Elevation, 188
Figure 12.5 Boiling of Pure Solvent (left) and a Solution of
Solvent and Nonvolatile Solute (right)., 189
12.8 Osmotic Presure, 191
Figure 12.6 Osmotic Pressure, π ., 192
12.9 Colligative Properties, 194
Figure 12.7 Lowering of the Freezing Point of Water
by Ammonia., 195
Problems, Examples, and Exercise, 195
Example 12.1, 195
Table 12.1 Vapor Pressures of Acetone over Dilute Binary
Solutions of Acetone in Diethyl Ether., 196
Example 12.2, 196
Exercise 12.1, 198
Exercise 12.2, 199
Problems 12.1–12.10, 199–202
13 Coulometry and Conductivity 203

13.1 Electrical Potential, 203
13.1.1 Membrane Potentials, 203
Figure 13.1 The Potential Drop Between Charged Plates
Is V = φ(0) − φ(l)., 204
Figure 13.2 An Ion-Permeable Membrane (Schematic)., 204
13.2 Resistivity, Conductivity, and Conductance, 205
13.3 Molar Conductivity, 206
Figure 13.3 Kohlrausch’s Law for Conductance of the
Strong Electrolytes HCl and NaOAc and the Weak
Electrolyte HOAc., 207
13.4 Partial Ionization: Weak Electrolytes, 208
13.5 Ion Mobilities, 209
Figure 13.4 Moving Boundary Determination of the
Mobility of H+ ., 210
13.6 Faraday’s Laws, 211
13.7 Mobility and Conductance, 211
13.8 The Hittorf Cell, 211
Figure 13.5 A Three-Compartment Hittorf Cell., 212
13.9 Ion Activities, 213
Example 13.1, 215
Example 13.2, 216
Example 13.3, 216
Problems 13.1–13.11, 217–219
P1: OTA/XYZ P2: ABC
xiv CONTENTS
14 Electrochemical Cells 220

14.1 The Daniell Cell, 220
14.2 Half-Cells, 221
Figure 14.1 The Hydrogen Half-Cell., 222
14.3 Half-Cell Potentials, 222
Table 14.1 A Few Selected Reduction Potentials., 223
14.4 Cell Diagrams, 223
14.5 Electrical Work, 224
14.6 The Nernst Equation, 224
14.7 Concentration Cells, 225
14.8 Finding E ◦ , 226
Figure 14.2 Extrapolation to E ◦ = 0.2223 for the
Standard Hydrogen–Silver–Silver Chloride
Cell., 228
14.9 Solubility and Stability Products, 228
14.10 Mean Ionic Activity Coefficients, 229
14.11 The Calomel Electrode, 229
14.12 The Glass Electrode, 230
Example 14.1, 230
Example 14.2, 231
Figure 14.3 The Mean Activity Coefficient of HCl as a
Function of m 1/2 ., 232
Problems 14.1–14.9, 232–234
15 Early Quantum Theory: A Summary 235

15.1 The Hydrogen Spectrum, 235
Figure 15.1 The Hydrogen Emission Spectrum., 236
Figure 15.2 The First Six Solutions of the H Atom Energy
Calculated by Bohr (1913)., 236
15.2 Early Quantum Theory, 236
Schrödinger, Heisenberg, and Born: An Introduction, 237
The Hamiltonian Operator, 237
15.3 Molecular Quantum Chemistry, 238
Heitler and London, 238
Hartree and Fock, 239
Antisymmetry and Determinantal Wave Functions, 240
15.4 The Hartree Independent Electron Method, 240
15.5 A Digression on Atomic Units, 243
Example 15.1, 243
Example 15.2, 244
Problems 15.1–15.9, 246–247
P1: OTA/XYZ P2: ABC
CONTENTS xv
16 Wave Mechanics of Simple Systems 248

16.1 Wave Motion, 248
Figure 16.1 Graph of sin(x), sin(2x), and sin(3x) Shown
over the Interval 0, π ., 249
16.2 Wave equations, 249
Eigenvalues and Eigenvectors, 250
16.3 The Schrödinger Equation, 250
16.4 Quantum Mechanical Systems, 251
is a Vector, 251
The Eigenfunction Postulate, 252
16.5 The Particle in a One-Dimensional Box, 253
Figure 16.2 Wave Forms for the First Three Wave
Functions of the Particle in a Box., 254
Fundamentals and Overtones, 254
Figure 16.3 A MathcadC Sketch of the Born Probability
Densities at the First Three Levels of the Particle in
a Box., 255
16.6 The Particle in a Cubic Box, 255
Separable Equations, 256
16.6.1 Orbitals, 257
Figure 16.4 The Ground State Orbital of a
Particle Confined to a Cubic Box., 257
16.6.2 Degeneracy, 257
Figure 16.5 The First Excited State of a Particle
Confined to a Cubic Box., 257
16.6.3 Normalization, 257
Figure 16.6 The Degenerate Energy Levels for
the Hydrogen Atom., 258
16.7 The Hydrogen Atom, 258
The Radial Equation and Probability “Shells”, 258
16.8 Breaking Degeneracy, 259
Figure 16.7 Reduced Degeneracy in Energy Levels for
Hydrogen-Like Atoms., 260
16.8.1 Higher Exact Solutions for the Hydrogen Atom, 260
Table 16.1 The First Six Wave Functions for Hydrogen., 261
Table 16.2 The First Three s Wave Functions for
Hydrogen (Simplified Form)., 261
Figure 16.8 Roots of the Radial 3s Wave Function of
Atomic Hydrogen as a Function of Distance r ., 261
Figure 16.9 The Radial Probability Density for
an Electron in the 3s Orbital of Hydrogen., 262
16.9 Orthogonality and Overlap, 262
Figure 16.10 The Radial Node of the 2 p Atomic
Orbital., 262
P1: OTA/XYZ P2: ABC
xvi CONTENTS
Figure 16.11 Favorable spx and Unfavorable spz

Overlap of Orbitals Depending upon Orbital
Symmetry., 263
16.10 Many-Electron Atomic Systems, 263
The Hartree Method, 263
Problems 16.1–16.9, 264–266
17 The Variational Method: Atoms 267

17.1 More on the Variational Method, 267
17.2 The Secular Determinant, 268
17.3 A Variational Treatment for the Hydrogen Atom: The
Energy Spectrum, 271
17.3.1 Optimizing the Gaussian Function, 272
Simultaneous Minima, 272
The Exact Wave Function, 272
The Gaussian Approximation, 272
17.3.2 A GAUSSIANC HF Calculation of E atom :
Computer Files, 273
File 17.1 Gaussian gen Input for the Hydrogen Atom., 273
File 17.2 Energies Drawn from the Gaussian
gen Output File for the Hydrogen Atom., 273
17.4 Helium, 274
17.4.1 An SCF Variational Ionization Potential for
Helium, 275
17.5 Spin, 278
17.6 Bosons and Fermions, 278
17.7 Slater Determinants, 279
17.8 The Aufbau Principle, 280
17.9 The SCF Energies of First-Row Atoms and Ions, 281
Figure 17.1 Calculated IP1 for Elements 1–10., 281
17.10 Slater-Type Orbitals (STO), 282
Table 17.1 Slater’s Rules., 282
17.11 Spin–Orbit Coupling, 283
Figure 17.2 Linear and Angular Momentum Vectors., 283
Example 17.1, 283
File 17.3 MathcadC Calculation of the Ionization
Potential of Helium., 284
Example 17.2, 284
Problems 17.1–17.9, 285–286
18 Experimental Determination of Molecular Structure 287

18.1 The Harmonic Oscillator, 287
Figure 18.1 A Classical Harmonic Oscillator., 288
P1: OTA/XYZ P2: ABC
CONTENTS xvii
18.2 The Hooke’s Law Potential Well, 289

Figure 18.2 Parabolic Potential Wells for the Harmonic
Oscillator., 290
18.3 Diatomic Molecules, 290
18.4 The Quantum Rigid Rotor, 290
Figure 18.3 Energy Levels within a Simple Rotor., 291
18.5 Microwave Spectroscopy: Bond Strength and Bond
Length, 292
18.6 Electronic Spectra, 292
Figure 18.4 Electronic Promotion in Alkenes., 293
Figure 18.5 Absorption Wavelengths of Conjugated
Polyalkenes., 293
18.7 Dipole Moments, 294
Figure 18.6 A Charged Parallel Plate Capacitor., 294
Figure 18.7 A Charged Capacitor with a Dielectric., 294
Dielectric Constant, 294
Polarizability, 295
18.7.1 Bond Moments, 296
Figure 18.8 The Total Dipoles of Two
Dichloroethene Isomers., 296
18.8 Nuclear Magnetic Resonance (NMR), 297
18.8.1 Spin–Spin Coupling, 298
Figure 18.9 Schematic NMR Spectrum of
Ethanol, CH3 CH2 OH., 299
Magnetic Resonance Imaging (MRI), 299
18.9 Electron Spin Resonance, 299
Example 18.1, 299
Figure 18.10 Schematic Diagram of a Vibration–Rotation
Band., 300
Example 18.2, 301
Problems 18.1–18.13, 301–304
Figure 18.11 The Vibration–Rotation Spectrum of CO., 302
19 Classical Molecular Modeling 305

19.1 Enthalpy: Additive Methods, 305
Figure 19.1 Enthalpies of Formation of “Adjacent”
n-Alkanes., 306
Group Additivity, 306
19.2 Bond Enthalpies, 306
Bond Additivity, 306
Figure 19.2 Bond Enthalpies Calculated in CH2 , from the
Reference State of Gaseous Atoms (top), and Relative
to Elements in their Standard State (H2 (g) and
C(graphite))., 307
P1: OTA/XYZ P2: ABC
xviii CONTENTS
19.3 Structure, 307

Figure 19.3 Structurally Distinct Alkane Conformers
Resulting from the Tetrahedral Symmetry of Carbon., 308
Force Constants and Parameters, 308
Energy Equations, 309
Force Fields, 309
The Allinger MM Method, 309
19.4 Geometry and Enthalpy: Molecular Mechanics, 309
19.5 Molecular Modeling, 310
19.6 The GUI, 310
Figure 19.4 Visualization of the Output for the Ethane
Molecule (PCModel 8.0C )., 310
19.7 Finding Thermodynamic Properties, 311
File 19.1 Partial MM4 Enthalpy Output for Ethane., 311
19.8 The Outside World, 312
19.9 Transition States, 313
Example 19.1, 314
Example 19.2, 314
File 19.2 An Input File for Water., 314
File 19.3 The MM4 Geometry Output TAPE9.MM4 for
Water., 315
Example 19.3, 315
File 19.4 MM4 Input Geometry for Methane., 316
File 19.5 MM4 Output Geometry for Methane., 316
Problems 19.1–19.10, 316–317
20 Quantum Molecular Modeling 318

20.1 The Molecular Variational Method, 318
20.2 The Hydrogen Molecule Ion, 319
Figure 20.1 The Hydrogen Molecule Ion, H+ 2 ., 319
Figure 20.2 Bonding and Antibonding Orbitals for H+ 2 ., 321
Figure 20.3 Bonding and Antibonding Solutions for
the H+2 ., 322
20.3 Higher Molecular Orbital Calculations, 322
20.4 Semiempirical Methods, 323
20.5 Ab Initio Methods, 324
20.6 The Gaussian Basis Set, 324
Figure 20.4 The 1s STO (solid line) and a Gaussian
Approximation (dotted line)., 324
File 20.1 (Input) A Four-Parameter Gaussian File for
the Hydrogen Atom., 325
Figure 20.5 Comparison of the 1s STO of Hydrogen with
an Arbitrarily Parameterized Two-Gaussian Function
φ(r ) = 0.40e−1.0 r + 0.60e−0.25 r ., 326
2 2
P1: OTA/XYZ P2: ABC
CONTENTS xix
File 20.2 (Output) The STO-2G Basis Set Written as a

1s Orbital Consisting of Functions with Arbitrarily
Selected Exponents 1.00 and 0.25., 326
20.7 Stored Parameters, 326
File 20.3 (Input) An STO-2G Input File Using a Stored
Basis Set., 327
File 20.4 (Output) Stored Parameters for the STO-2G
Basis Set., 327
Figure 20.5 Approximation to the 1s Orbital of Hydrogen
by 2 Gaussians., 328
20.8 Molecular Orbitals, 330
File 20.5 (Input) A Molecular Orbital Input File for H2 ., 329
z-Matrix Format, 329
File 20.6 (Input) A GAUSSIAN Input File for H2 ., 330
20.8.1 GAMESS, 330
File 20.7 (Input) GAMESS File for Hydrogen Molecule., 330
20.9 Methane, 331
File 20.8 One of Many Possible STO-2G Optimized
Coordinates Sets for Methane., 331
20.10 Split Valence Basis Sets, 331
20.11 Polarized Basis Functions, 332
20.12 Heteroatoms: Oxygen, 332
File 20.9 Input file for a GAUSSIANC STO-3G
Calculation on Methanol., 333
File 20.10 Optimized Geometry from a GAUSSIANC
STO-3G Calculation on Methanol (Internal and
Cartesian Coordinates)., 334
20.13 Finding f H 298 of Methanol, 334
Figure 20.6 The G3MP2 Thermochemical Cycle for
Determination of f H 298 of Methanol., 335
20.14 Further Basis Set Improvements, 336
20.15 Post-Hartree–Fock Calculations, 336
20.16 Perturbation, 337
20.17 Combined or Scripted Methods, 338
Scheme 20.1 A Computational Chemical Script., 338
Figure 20.7 Additive Extrapolations in the G3(MP2)
Scripted Method., 339
File 20.11 Partial GAUSSIAN G3(MP2) Output., 339
20.18 Density Functional Theory (DFT), 339
Problems And Examples, 340
Example 20.1, 340
File 20.12 A z-Matrix Input File for Methane., 340
Example 20.2, 340
File 20.13 Control Lines for a GAMESS Calculation., 341
Example 20.3, 340
Problems 20.1–20.9, 342–343
P1: OTA/XYZ P2: ABC
xx CONTENTS
21 Photochemistry and the Theory of Chemical Reactions 344

21.1 Einstein’s Law, 344
21.2 Quantum Yields, 345
Table 21.1 Some Experimental Quantum Yields., 345
Figure 21.1 Mechanism for Fluorescent and
Phosphorescent Light Emission., 346
21.2.1 Lipid Peroxidation, 346
21.2.2 Ozone Depletion, 347
21.3 Bond Dissociation Energies (BDE), 348
21.4 Lasers, 348
21.5 Isodesmic Reactions, 349
21.6 The Eyring Theory of Reaction Rates, 349
21.7 The Potential Energy Surface, 350
Figure 21.2 Eyring Potential Energy Plot for the
Reaction H + H–H → H–H + H., 350
Figure 21.3 Activation of the Symmetrical Reaction
H + H–H → H–H + H., 351
Figure 21.4 The Enthalpy of Activation of an Exothermic
Reaction., 351
21.7.1 Optical Inversion, 352
Figure 21.5 An Optically Active Species., 352
Figure 21.6 Inversion of Optical Activity., 352
21.8 The Steady-State Pseudo-Equilibrium, 353
21.9 Entropies of Activation, 354
21.10 The Structure of the Activated Complex, 355
Example 21.1, 355
Example 21.2, 356
Problems 21.1–21.8, 357–359
References 361
Answers to Selected Odd-Numbered Problems 365
Index 369
P1: OTA/XYZ P2: ABC
FOREWORD
Among many advantages of being a professional researcher and teacher is the pleasure
of reading a new and good textbook that concisely summarizes the fundamentals and
progress in your research area. This reading not only gives you the enjoyment of
looking once more at the whole picture of the edifice that many generations of
your colleagues have meticulously build but, most importantly, also enhances your
confidence that your choice to spend your entire life to promote and contribute to
this structure is worthwhile. Clearly, the perception of the textbook by an expert in
the field is quite different, to say the least, from the perception of a junior or senior
undergraduate student who is about to register for a class. A simple look at a textbook
that is jam-packed with complex integrals and differential equations may scare any
prospective students to death. On the other hand, eliminating the mathematics entirely
will inevitably eliminate the rigor of scientific statements. In this respect, the right
compromise between simplicity and rigor in explaining complex scientific topics is
an extremely rare talent. The task is especially large given the fact that the textbook is
addressed to students for whom a particular area of science is not among their primary
interests. In this respect, Professor Rogers’s Concise Physical Chemistry is a textbook
that ideally suits all of the above-formulated criteria of a new and good textbook.
Although the fundamental laws and basic principles of physical chemistry were
formulated long ago, research in the area is continuously widening and deepening. As
a result, the original boundaries of physical chemistry as a science become more and
more vague and difficult to determine. During the last two decades, physical chemistry
has made a tremendous progress mainly boosted by a spectacular increase in our
computational capabilities. This is especially visible in quantum molecular modeling.
For instance, on my first acquaintance with physical chemistry about 30 years ago,
the only molecule that could be quantitatively treated with an accuracy close to
xxi
P1: OTA/XYZ P2: ABC
xxii FOREWORD
experimental data by wave mechanics was the hydrogen molecule. In a lifetime, I

have witnessed a complete change of the research picture in which thermodynamic
and kinetic data are theoretically obtained routinely with an accuracy often exceeding
the experimental one. Quite obviously, to keep the pace with the progress in research,
textbooks should be permanently updated and revised. In his textbook Professor
Rogers sticks to the classical topics that are conventionally considered as part of
physical chemistry. However, these classical topics are deciphered from a modern
point of view, and here lies the main strength of this textbook as well as what actually
makes this textbook different from many other similar textbooks.
Traditionally, physical chemistry is viewed as an application of physical principles
in explaining and rationalizing chemical phenomena. As such, the powerful principles
and theories that physical chemistry borrows from physics are accompanied by an
advanced and mandatory set of mathematical tools. This makes the process of learning
physical chemistry very difficult albeit challenging, exciting, and rewarding. The level
of mathematics used by Professor Rogers to formulate and prove the physicochemical
principles is remarkably consistent throughout the whole text. Thus, only the most
general algebra and calculus concepts are required to understand the essence of
the topics discussed. Professor Rogers’s way of reasoning is succinct and easy to
follow while the examples used to illustrate the theoretical developments are carefully
selected and always make a good point. There is no doubt that this textbook is a work
of great value, and I heartily recommend it for everybody who wants to enter the
wonderful world of physical chemistry.
Worcester Polytechnic Institute Ilie Fishtik

Worcester, MA
July 2010
P1: OTA/XYZ P2: ABC
PREFACE
Shall I call that wise or foolish, now; if it be really wise it has a foolish look to it; yet, if
it be really foolish, then has it a sort of wiseish look to it.
Moby-Dick (Chapter 99) —Herman Melville
Physical chemistry stands at the intersection of the power and generality of classical
and quantum physics with the minute molecular complexity of chemistry and biology.
Any molecular process that can be envisioned as a flow from a higher energy state
to a lower state is subject to analysis by the methods of classical thermodynamics.
Chemical thermodynamics tells us where a process is going. Chemical kinetics tells
us how long it will take to get there.
Evidence for and application of many of the most subtle and abstract principles
of quantum mechanics are to be found in the physical interpretation of chemical
phenomena. The vast expansion of spectroscopy from line spectra of atoms well
known in the nineteenth century to the magnetic resonance imaging (MRI) of today’s
diagnostic procedures is a result of our gradually enhanced understanding of the
quantum mechanical interactions of energy with simple atomic or complex molecular
systems.
Mathematical methods developed in the domain of physical chemistry can be
successfully applied to very different phenomena. In the study of seemingly unrelated
phenomena, we are astonished to find that electrical potential across a capacitor, the
rate of isomerization of cyclopentene, and the growth of marine larvae either as
individuals or as populations have been successfully modeled by the same first-order
differential equation.
Many people in diverse fields use physical chemistry but do not have the op-
portunity to take a rigorous three-semester course or to master one of the several
∼1000-page texts in this large and diverse field. Concise Physical Chemistry is
xxiii
P1: OTA/XYZ P2: ABC
xxiv PREFACE
intended to meet (a) the needs of professionals in fields other than physical chemistry
who need to be able to master or review a limited portion of physical chemistry or
(b) the need of instructors who require a manageable text for teaching a one-semester
course in the essentials of the subject. The present text is not, however, a diluted
form of physical chemistry. Topics are treated as brief, self-contained units, graded
in difficulty from a reintroduction to some of the concepts of general chemistry in
the first few chapters to research-level computer applications in the later chapters.
I wish to acknowledge my obligations to Anita Lekhwani and Rebekah Amos
of John Wiley and Sons, Inc. and to Tony Li of Scientific Computing, Long Island
University. I also thank the National Center for Supercomputing Applications and
the National Science Foundation for generous allocations of computer time, and the
H. R. Whiteley Foundation of the University of Washington for summer research
fellowships during which part of this book was written.
Finally, though many people have helped me in my attempts to better appreciate
the beauty of this vast and variegated subject, this book is dedicated to the memory
of my first teacher of physical chemistry, Walter Kauzmann.
Donald W. Rogers
P1: OTA/XYZ P2: ABC
c01 JWBS043-Rogers September 13, 2010 11:20 Printer Name: Yet to Come
1
IDEAL GAS LAWS
In the seventeenth and eighteenth centuries, thoughtful people, influenced by the

success of early scientists like Galileo and Newton in the fields of mechanics and
astronomy, began to look more carefully for quantitative connections among the
phenomena around them. Among these people were the chemist Robert Boyle and
the famous French balloonist Jacques Alexandre César Charles.
1.1 EMPIRICAL GAS LAWS
Many physical chemistry textbooks begin, quite properly, with a statement of Boyle’s
and Charles’s laws of ideal gases:
pV = k1 (Boyle, 1662)
and
V = k2 T (Charles, 1787)
The constants k1 and k2 can be approximated simply by averaging a series of experi-

mental measurements, first of pV at constant temperature T for the Boyle equation,
then of V/T at constant pressure p for Charles’s law. All this can be done using simple
manometers and thermometers.
Concise Physical Chemistry, by Donald W. Rogers

Copyright
C 2011 John Wiley & Sons, Inc.
1
P1: OTA/XYZ P2: ABC
2 IDEAL GAS LAWS
1.1.1 The Combined Gas Law

These two laws can be combined to give a new constant
pV
= k3
T
Subsequently, it was found that if the quantity of gas taken is the number of grams
equal to the atomic or molecular weight of the gas, the constant k3 , now written R
under the new stipulations, is given by
pV = RT
For the number of moles of a gas, n, we have
pV = n RT
The constant R is called the universal gas constant.
1.1.2 Units
The pressure of a confined gas is the sum of the force exerted by all of the gas
molecules as they impact with the container walls of area A in unit time:
f in units of N
p=
A in units of m2
The summed force f is given in units of newtons (N), and the area is in square meters
(m2 ). The N m−2 is also called the pascal (Pa). The pascal is about five or six orders
of magnitude smaller than pressures encountered in normal laboratory practice, so
the convenient unit 1 bar ≡ 105 Pa was defined.
The logical unit of volume in the MKS (meter, kilogram, second) system is the
m3 , but this also is not commensurate with routine laboratory practice where the liter
is used. One thousand liters equals 1 m3 , so the MKS name for this cubic measure is
the cubic decimeter—that is, one-tenth of a meter cubed (1 dm3 ). Because there are
1000 cubic decimeters in a cubic meter and 1000 liters in a cubic meter, it is evident
that 1 L = 1 dm3 .
The unit of temperature is the kelvin (K), and the unit of weight is the kilogram
(kg). Formally, there is a difference between weight and mass, which we shall ignore
for the most part. Chemists are fond of expressing the amount of a pure substance in
P1: OTA/XYZ P2: ABC
THE MOLE 3
terms of the number of moles n (a pure, unitless number), which is the mass in kg
divided by an experimentally determined unit molar mass M, also in kg:1
kg
n=
M
If the pressure is expressed as N m−2 and volume is in m3 , then pV has the unit N m,
which is a unit of energy called the joule (J). From this, the expression
pV
R=
nT
gives the unit of R as J K−1 mol−1 . Experiment revealed that
R = 8.314 J K−1 mol−1 = 0.08206 L atm K−1 mol−1
which also defines the atmosphere, an older unit of pressure that still pervades the
literature.
1.2 THE MOLE
The concept of the mole (gram molecular weight in early literature) arises from the
deduction by Avogadro in 1811 that equal volumes of gas at the same pressure and
temperature contain the same number of particles. This somewhat intuitive conclusion
was drawn from a picture of the gaseous state as being characterized by repulsive
forces between gaseous particles whereby doubling, tripling, and so on, the weight
of the sample taken will double, triple, and so on, its number of particles, hence its
volume. It was also known at the time that electrolysis of water produced two volumes
of hydrogen for every volume of oxygen, so Avogadro deduced the formula H2 O for
water on the basis of his hypothesis of equal volume for equal numbers of particles
in the gaseous state.
By Avogadro’s time, it was also known that the number of grams of oxygen
obtained by electrolysis of water is 8 times the number of grams of hydrogen. By
his 2-for-1 hypothesis, Avogadro reasoned that the less numerous oxygen atoms
must be 2(8) = 16 times as heavy as the more numerous hydrogen atoms. This
theoretical vision led directly to the concept of atomic and molecular weight and
to the mass of pure material equal to its atomic weight or molecular weight, which
we now call the mole.2 Various experimental methods have been used to determine
the number of particles comprising one mole of a pure substance with the result
1 General practice is to write experimentally determined quantities in italics and units in Roman letters,
but there is some overlap and we shall not be strict in this observance.
2 The word is mole, but the unit is mol.
P1: OTA/XYZ P2: ABC
4 IDEAL GAS LAWS
6.022 × 1023 , which is now appropriately called Avogadro’s number, NA . One mole
of an ideal gas contains NA particles and occupies 24.79 dm3 at 1 bar pressure and
298.15 K.
1.3 EQUATIONS OF STATE
The equation pV = RT with the stipulation of one mole of a pure gas is an equation
of state. Given that R is a constant, the combined gas law equation can be written in
a more general way:
p = f (V, T )
which suggests that there are other ways of writing an equation of state. Indeed,
many equations of state are used in various applications (Metiu, 2006). The common
feature of these equations is that only two independent variables are combined with
constants in such a way as to produce a third dependent variable. We can write the
general form as p = f (V, T ), or
V = f ( p, T )
or
T = f ( p, V )
so long as there are two independent variables and one dependent variable. One mole
of a pure substance always has two degrees of freedom. Other observable properties
of the sample can be expressed in the most general form:
z = f (x1 , x2 )
The variables in the general equation may seem unconnected to p and V, but there
always exists, in principle, an equation of state, with two and only two independent
variables, connecting them.
An infinitesimal change in a state function z for a system with two degrees of
freedom is the sum of the infinitesimal changes in the two dependent variables, each
multiplied by a sensitivity coefficient (∂z/∂ x1 )x2 or (∂z/∂ x2 )x1 which may be large
if the dependent variable is very sensitive to independent variable xi or small if dz is
insensitive to xi :

∂z ∂z
dz = d x1 + d x2
∂ x1 x2 ∂ x2 x1
P1: OTA/XYZ P2: ABC
DALTON’S LAW 5
The subscripts x1 and x2 on the parenthesized derivatives indicate that when one
degree of freedom is varied, the other is held constant. We shall investigate state
functions in more detail in the chapters that are to come.
1.4 DALTON’S LAW
At constant temperature and pressure, by Avogadro’s principle, the volume of an

ideal gas is directly proportional to the number of particles of the gas measured in
moles:

RT
V =n = n NA
p
This principle holds regardless of the nature of the particles:

RT
p=n
V const
Since the nature of the particles plays no role in determining the pressure, the total
pressure of a mixture of ideal gases3 is determined by the total number of moles of
gas present:

RT RT RT RT
p = n1 + n2 + ··· = ni = ni
V const V const i
V const V const i
Each gas acts as though it were alone in the container, which leads to the concept
of a partial pressure pi exerted by one component of a mixture relative to the total
pressure. This idea is embodied in Dalton’s law for the total pressure of a mixture as
the sum of its partial pressures:

ptotal = pi
i
Apart from emphasizing Avogadro’s idea that the ideal gaseous state is characterized
by the number of particles, not by their individual nature, Dalton’s law also leads to
the idea of a pressure fraction of one component of a mixture relative to the total
pressure exerted by all the components of the mixture:
pi
X pi =
pi
i
3 Many real gases are nearly ideal under normal room conditions.
P1: OTA/XYZ P2: ABC
6 IDEAL GAS LAWS
1.5 THE MOLE FRACTION
Recognizing that the pressure of each gas is directly proportional to the number
of moles through the same constant, we may write the pressure fraction as a mole
fraction:
ni
Xi =
ni
i
The pressure of a real gas follows Dalton’s law only as an approximation, but the
number of particles (measured in moles) is not dependent upon ideal behavior; hence
the summation of mole fractions

X total = Xi
i
is exact for ideal or nonideal gases and for other states of matter such as liquid and
solid mixtures and solutions.
1.6 EXTENSIVE AND INTENSIVE VARIABLES
Mass m is an extensive variable. Density ρ is an intensive variable. If you take twice

the amount of a sample, you have twice as many grams, but the density remains the
same at constant p and T. Molar quantities are intensive. For example, if you double
the amount of sample under at constant p and T, the molar volume (volume per mole)
Vm remains the same just as the density did.
1.7 GRAHAM’S LAW OF EFFUSION
Knowing the molar gas constant R = 8.314 J K−1 mol−1 = 0.08206 L atm K−1
mol−1 , which follows directly from measurements of p and V on known amounts of
a gas at specific values of T, one can determine the atomic or molecular weight of an
independent sample within the limits of the ideal gas approximation. Another way
of finding the molecular weight of a gas is through Graham’s law of effusion, which
states that the rate of escape of a confined gas through a very small hole is inversely
proportional to its particle weight—that is, its atomic or molecular weight. This being
the case, measuring the rate of effusion of two gases—one of known molecular weight
and the other of unknown molecular weight—gives the ratio MWknown /MWunknown
and hence easy calculation of MWunknown .
Aside from important medical applications (dialysis), Graham’s work also focused
attention on the random motions of gaseous particles and the speeds with which they
move. We can rationalize Graham’s law as the result of a very large ensemble of
P1: OTA/XYZ P2: ABC
THE MAXWELL–BOLTZMANN DISTRIBUTION 7
particles colliding with the wall of a constraining container, supposing that the wall
has a hole in it. Only a few particles escape the container because the hole is small.
Escape probability is determined by how fast the particle is moving. Fast particles
collide with the walls of the container more often than do slow ones.
By a standard derivation (Exercise 1.2), one finds
pV = 13 N A m ū 2x
where ū x is the average speed of an ensemble consisting of one mole of an ideal

gas. Notice that because pV = RT has the units J K−1 mol−1 K = J mol−1 , pV is
a molar energy. Increasing the temperature of a gas requires an input of energy. We
usually write the kinetic energy E kin of a single moving mass such as a baseball
as E kin = 12 mv 2 , where v is its speed and m is its mass. Consider a hypothetical
one-dimensional x-space along which point particles can move without interference.
If the kinetic energy of molecular particles follows the same kind of law as more
massive particles, we obtain
1
2
m ū 2x = Ē kin
where Ē kin is the average kinetic energy because kinetic energy is the only kind an
ensemble of point particles can have. Substitute 2 Ē kin for m ū 2x in
pV = 23 N A Ē kin
but pV also equals RT for one mole of a gas, so
pV = RT = 13 N A m ū 2x
This enables us to calculate ū x at any specified temperature. The calculation gives

high speeds. For example, nitrogen molecules move at about 400 m s−1 (meters per
second) at room temperature and hydrogen molecules move at an astonishing speed
of nearly 2000 m s−1 . There are different ways of calculating averages (mean, mode,
root mean square), which give slightly different results for molecular speeds.
1.8 THE MAXWELL–BOLTZMANN DISTRIBUTION
All particles of a confined gas do not move with the same velocity even if T is
constant. Rather, they move with a velocity probability density ρv which is randomly
distributed about v = 0 and which follows the familiar Gaussian distribution e−v .
2
The probability density function drops off at large values of ±v because the prob-
ability of finding particles with velocities very much different from the mean is small.
The curve is symmetrical because, picking an arbitrary axis, the particle may be going
either to the left or to the right, having a velocity v or –v. The peak at v = 0 is somewhat
P1: OTA/XYZ P2: ABC
8 IDEAL GAS LAWS
2
−v
e
0
−3 v 3
FIGURE 1.1 The probability density for velocities of ideal gas particles at T = 0.
misleading because it may suggest that the most probable velocity is zero. Not so. The
particles are not standing still at any temperature above absolute zero. The peak at
v = 0 arises because we don’t know which direction any particle is going, left or
right. In our ignorance, assuming a random distribution, the best bet is to guess zero.
We will always be wrong, but the sum of squares of our error over many trials will
be minimized. This is an example of the principle of least squares.
The Maxwell–Boltzmann distribution of molecular speeds was originally derived
assuming that particle velocities are distributed along a continuous spectrum like
Fig. 1.1. This implies that E kin can take any value in a continuum as well. The laws of
quantum mechanics, however, deny this possibility. They require a distribution over
a discontinuous energy spectrum or manifold of energy levels like that in Fig. 1.2.
The connection between Figs. 1.1 and 1.2 can be seen by tilting the page 90◦ to the
left. The number of particles at higher energies tails off according to a Gaussian dis-
tribution. The Maxwell–Boltzmann distribution over nondegenerate, discontinuous
energy levels is
Ni
= e−Ei /k B T
N0
.
..
.....
E .........
................
....................
FIGURE 1.2 A Maxwell–Boltzmann distribution over discontinuous energy levels. Particles
are not static; they exchange energy levels rapidly. The levels need not be equally spaced.
P1: OTA/XYZ P2: ABC
A DIGRESSION ON “SPACE” 9
where Ni is the number of particles at the level having energy E i . In this expression,
N0 is the number at the lowest energy, usually designated zero E0 = 0 in the absence
of a reason to do otherwise.4 Energy, being a scalar, is proportional to the square
of the speed of an ensemble of molecules. The population of the energy levels in
Fig. 1.2 drops off rapidly at higher energies.
The term degeneracy is used when two or more levels exist at the same en-
ergy, which sometimes happens under the laws of quantum mechanics. Now the
number of particles at level E i is multiplied by the number of levels gi having that
energy
Ni
= gi e−Ei /k B T
N0
The degeneracy is always an integer and it is usually small. Also, from E kin = 32 RT
for one mole, we can find the expectation value εkin of the kinetic energy per
representative or average particle
3 R
εkin = T
2 NA
This leads to the important constant
R 8.3145
= kB = = 1.381 × 10−23 J K−1
NA 6.022 × 1023
and
εkin = 32 k B T
where k B , the universal gas constant per particle, is called the Boltzmann constant. It
should be evident that k B T must have the units of energy because Ni /N0 is a unitless
(pure) number, ln (Ni /N0 ) = −E i /k B T , which is also unitless, hence the units of
k B T must be the same as E i . We are taking advantage of the fact that if y = e x , then
ln y = x.
1.9 A DIGRESSION ON “SPACE”
The terms density and probability density were used in Section 1.8. These are different
but analogous uses of the word density. In the first case, density was used in the usual
sense of weight or mass per unit volume, m/V. In the second case, the probability
density is defined as the probability in a specified space. Any variable measured
4 Like all energies, this zero point is arbitrary.

P1: OTA/XYZ P2: ABC
10 IDEAL GAS LAWS
FIGURE 1.3 The Gaussian Probability Density Distribution in 3-Space. The distribution
curve is in the fourth dimension of the space. The probability maximum is at the center of the
sphere.
along an axis defines a space. For example, plotting x along a horizontal axis defines
a one-dimensional x-space. Space in the x, y, and z dimensions is the familiar 3-space
often called a Cartesian space in honor of the seventeenth-century mathematician and
philosopher René Descartes. We usually plot functions along mutually perpendicular
or orthogonal axes for mathematical convenience. If velocity is plotted along a v
axis, we have a one-dimensional velocity space. If probability density ρ is plotted
along one axis, and velocity is plotted along another axis, the result is a probability
density–velocity space of two dimensions. If ρ(v) is plotted in vx ,v y space, the result
is a function in 3-space; or if it is thought of as a function of all three Cartesian
coordinates, the resulting function is in 4-space. That is, ρ(v) in 1-, 2-, or 3-space
gives a function in 2-, 3-, or 4-space, one dimension more. There should be nothing
terrifying about many-dimensional space or hyperspace; it is merely an algebraic
generalization of the more commonplace use of the term.
The four-dimensional surface of the Gaussian distribution in Cartesian 3-space
cannot be precisely drawn but it can be imagined as a figure with spherical symmetry,
having a maximum at the center of the sphere. Imagine that the sphere in Fig. 1.3
can be rotated any amount in any angular direction, leaving the distribution curve
unchanged.
1.10 THE SUM-OVER-STATES OR PARTITION FUNCTION
Adding up all the particles in all the states of a system gives the total number of
particles in the system:

Ni = N
P1: OTA/XYZ P2: ABC
THE SUM-OVER-STATES OR PARTITION FUNCTION 11
or

Ni = N0 gi e−εi /k B T = N
Since N0 is a number appearing in each term of the sum, it can be factored out:

N0 gi e−εi /k B T = N

Dividing Ni = N0 gi e−εi /k B T by N0 gi e−εi /k B T = N , gives
Ni N0 gi e−εi /k B T gi e−εi /k B T
= =
N N0 gi e−εi /k B T Q
−εi /k B T
where we have given the symbol Q to the summation gi e . This important
summation appears frequently and is given the name sum-over-states or partition
function. Rewriting the ratio Ni /N , we have
Ni Q = N gi e−εi /k B T

fixed at T = const
We see that, for a given number of molecules N at temperature T, the right hand
side of the equation is fixed for any specific energy state, that is, Ni Q = const.
The sum-over-states is then a scaling factor, determining the relative population of a
state, Ni = const/Q. If Q is large, the state is sparsely populated. If Q is small, it is
densely populated. Another way of looking at Q is that it is an indicator of the number
of quantum states available to a system. For many available states (large Q) a given
state is less densely populated than it would be if only a few states were available
(small Q).
The occupation number of a quantum state relative to the total number of particles
Ni /N is, strictly speaking, a probability; however, given the immense number of
particles in a mole of gas, we may treat it as a certainty. The summation of all
possible fractions Ni /N must be 1:

Ni
Ni i N
= = =1
i
N N N
The probability distribution of molecular velocities in 3-space is a collection of

randomly oriented vectors away from an origin of the zero (least squares) estimate.
Plotted in velocity-space, the probability density approaches zero near the origin.
This comes about when we use spherical polar coordinates to express our probability
densities because, near the origin, the volume of the spherical space becomes very
small. As the radial distance is increased, the space becomes larger but the probability
P1: OTA/XYZ P2: ABC
12 IDEAL GAS LAWS
P(v)
FIGURE 1.4 The probability density of molecular velocities in a spherical velocity space.
density drops off as a Gaussian function. In between these two approaches to zero,
the probability density must go through a maximum as shown in Fig. 1.4.
PROBLEMS AND EXERCISES
Exercise 1.1 The Combined Gas Law

Combine Boyle’s law and Charles’s law to obtain the combined gas law.
Solution 1.1 Take an ideal gas under the arbitrary conditions p1 V1 T1 and convert
it to p2 V2 T2 by a two-step process, varying the pressure first and the temperature
second. After the pressure is changed from p1 to p2 , according to Boyle’s law, the
volume, still at T1 , is at an intermediate value Vx
p1 V1 = p2 Vx T1 = const
p1 V1
Vx =
p2
Now change the temperature to T2 at constant p2 . By Charles’s law, the volume goes
from Vx to V2
Vx V2
= p2 = const
T1 T2
V2
Vx = T1
T2
Now equate Vx from the equations above:
p1 V1 V2
= T1
p2 T2
p1 V1 p2 V2
=
T1 T2
P1: OTA/XYZ P2: ABC
PROBLEMS AND EXERCISES 13
Since the pV/T quotients are equal to each other for any arbitrary variations in p, V,
and T, they must be equal to the same constant k:
pV
= const = k
T
which is the combined gas law.

If we accept the combined gas law, there is a gas constant for any specified quantity
of each individual gas, subject only to the restriction of ideal behavior. If we demand
that V be the molar volume Vm

pVm p Vn
= =k
T T
where n is the number of moles in the gas sample, then this equation becomes
pV
= nR
T
where the symbol R is used to denote a universal gas constant applicable to one
mole of any gas in the approximation of ideal behavior. One can obtain a value for
R by arbitrarily assigning a pressure of 1 bar at T = 298.15 K to precisely 1.0 mole
of an ideal gas. We know the molar volume to be 24.790 dm3 at this pressure and
temperature, so
V
p 1.000(24.790)
R= n
= = 0.083146
T 298.15
with units of bar dm3 K−1 mol−1 . If the volume is expressed in m3 , then R is in
bar m3 K−1 mol−1 = 100(0.0831) = 8.310 J K−1 mol−1 . Remember that the unit
J K−1 is for a molar gas constant. Notice that the numerator is an energy. The
tabulated value is 8.3144725 J K−1 mol−1 (CRC Handbook of Chemistry and Physics,
2008–2009, 89th ed.)
Exercise 1.2 The Maxwell–Boltzmann Distribution

Derive
pV = 13 N A mv x2
where N A , m, and v x are Avogadro’s number, the mass, and the average x-component
of the velocity of a collection of ideal gas particles confined to a cubic box l dm on
an edge.
P1: OTA/XYZ P2: ABC
14 IDEAL GAS LAWS
Solution 1.2 Consider a particle (molecule or atom) moving in the x direction in

a cubic box l units on an edge, oriented so that the particle path is perpendicular to
one of its faces. Consider for now only collisions with one wall. The momentum of
the particle is mv on the way into the collision, and −mv (in the opposite direction)
on the way out of the collision. The change of momentum is 2mv. The number of
collisions per second (collision frequency) is v/2l because the particle must travel
the length of the box l to collide with the opposite wall and l once again on the return
trip. Force is the change in momentum with respect to time, dp/dt (Newton’s second
law). Force per collision is F = ma = dp/dt, where a is the acceleration, a change
in speed over unit time (in seconds). The total force exerted by the particle on the
wall for many collisions is the force per collision times the number of collisions per
unit time (a frequency):
v mv 2
F = 2mv =
2l l
Pressure is force per unit area A = l 2 :

mv 2 1 mv 2 1 mv 2 mv 2
p= = 2
= 3 =
l A l l l V
where the volume of the box is the cube of one of its edges. Given that there are very
many molecules in the box, on average, only one-third of them are moving in the x
direction; or, better said, only one-third of all components of all velocity vectors are
oriented in the x direction. (The other two-thirds are oriented in the y and z directions.)
For Avogadro’s number, NA , of particles we find,
1 mv 2
p= NA
3 V
where v is an unbiased or average particle speed. This is the equation we sought.

We already know that pV is an energy and that 12 mv 2 is the kinetic energy of a
moving mass m at velocity v. One sees the proportionality in the preceding equation.
More explicitly, multiply and divide the preceding equation by 2. Now,
2 1 2
pV = NA m v̄ x2 = Ē kinetic
3 2 3
where the overbar notation for v̄ x and Ē kinetic is added to stress that the speed (scalar
magnitude of the velocity component) is an average value over all of the NA particles.
P1: OTA/XYZ P2: ABC
Problem 1.1
Calculate the volume of 50.0 gr of methane at 400 K and p = 12.0 bar. R =
8.314 J K−1 mol−1 . Discuss units.
Problem 1.2
A quantity of an ideal gas occupies 37.5 L at 1.00 atm pressure. How many liters
will it occupy when compressed to 4.50 atm pressure at constant temperature? What
is the Boyle’s law constant k? Give units. 1 bar = 0.986923 atm.
Problem 1.3
Plot the pV curves for 1.00 dm3 of three ideal gas samples, each expanded from a
volume of 1.00 dm3 in increments of 10.0 dm3 to 100 dm3 , where the three samples
are maintained at energies of 500, 1500, and 2500 joules, respectively.
Problem 1.4
Four grams of an organic liquid vaporizes to produce 1.00 dm3 of vapor at 298.15 K
and 1.00 atmosphere. Find an approximate molar mass of the liquid.

C
Problem 1.5 Mathcad Computer Exercise
(a) The volume of a fixed quantity of a real gas at 298.15 K was measured at five
different pressures, p = 0.160, 0.219, 0.310, 0.498, and 0.652 atm. The ex-
perimental results were V = 3.42, 2.48, 1.71, 1.03, 0.75 dm3 . These pres-
sures and volumes were tabulated as column vectors p and V. Calculate five
approximate Boyle’s law constants from these measurements.
(b) What would the Boyle’s law constant have been if the gas had been ideal?
(c) If the amount of sample, identified as carbon dioxide, is 1.000 g, what is its
molar volume?
Problem 1.6
The volume of a 0.5333-g sample of gas was measured at 298 K and pressures of
0.0590, 0.143, 0.288, 0.341, and 0.489 bar with the results V = 14.8, 6.07, 2.99, 2.54,
and 1.75 dm3 . What is the molar mass of the gas?
Problem 1.7
A mixture of 8.00 g of H2 and 2.00 g of D2 was allowed to effuse through a minute
orifice, and the composition of the effusing gas mixture was monitored by glc. What
was the percent composition of the first trace of gas mixture so monitored?
P1: OTA/XYZ P2: ABC
16 IDEAL GAS LAWS
Problem 1.8
We have two expressions for the molar volume of an ideal gas: (1) Vm = 22.414 L
at 0◦ C and 1 atm and (2) Vm = 24.790 dm3 at 25◦ C and 1 bar. (The atm and bar
are taken to have an indefinite number of significant digits.) Use this information to
obtain absolute zero (T = 0 K) on the Centigrade scale.
Problem 1.9
If a gas occupies 47.6 dm3 at NSTP = 298.15 K, what is its volume at 1.00 bar and
500 K? The acronym for new standard temperature and pressure NSTP replaces the
old STP for standard temperature and pressure.
Problem 1.10
(a) Suppose that 18.44 g of N2 occupies a container at new standard temperature
and pressure NSTP and 24.35 g of a sample of a different gas is introduced into
the container, keeping the temperature constant. If the pressure after addition
is 3.20 bar, what is the average molar mass Mav of the mixture?
(b) What is the molar mass of the introduced gas?
Problem 1.11
A pure gas takes twice as long as helium to effuse through a porous membrane. What
is its molar mass?
Problem 1.12
Compute the root mean square speed of H2 molecules at 1000 K.
Problem 1.13
What is the translational energy of 1 mol of an ideal gas at T = 298.15 K?
Computer Exercise 1.14

Using a standard plotting package, plot a graph of pV = k where k = 1, taking p as
the vertical axis and V as the horizontal axis. Play with your plotting package so as to
produce many different plots, thereby learning the idiosyncrasies of your particular
package.
Problem 1.15
What is the expectation value of the molecular speed among an ensemble of nitrogen
molecules at 298 K?
P1: OTA/XYZ P2: ABC
Problem 1.16
(a) Calculate the expectation value of the speed of hydrogen molecules among
an ensemble at 298 K. Give units.
(b) At the same temperature, v = 515 m s−1 for nitrogen molecules. What is the
ratio v H2 /v N2 ? Explain this ratio. Give units.
Problem 1.17
A sample of 2.50 mol of a gas was confined to a certain volume at 1.00 atm pressure
and 298 K. Assuming ideal behavior, what volume did it occupy?
Problem 1.18
If the molar volume is 24.79 dm3 at 298.15 K and 1 bar pressure, what is the universal
gas constant. Give units.
P1: OTA/XYZ P2: ABC
2
REAL GASES:
EMPIRICAL EQUATIONS
The ideal gas laws are based on two assumptions, neither of which is true. First,
atoms or molecules comprising the ideal gas are assumed to have no volume. They
are treated as mathematical point masses for convenience. Second, they are assumed
to have no interactions with each other. Attractive and repulsive forces are ignored
by setting them to zero.
2.1 THE VAN DER WAALS EQUATION
The Dutch physicist van der Waals remedied both of these failures. He treated the first
of them by simply subtracting an empirical parameter taken to represent the volume
of the particles, called the excluded volume b, from the total volume V of the gas to
leave an effective volume of (V – b). There is less space for each molecule to move in
because of the space taken up by its neighbors.
Attractive and repulsive forces often operate through an inverse square law. For
example, gravitational force on masses m 1 and m 2 at a separation of r is
m1m2
F=G
r2
where G is the gravitational constant. The constants G, m 1 , and m 2 can be grouped

as a single constant k in the numerator to give an attractive or repulsive force F that
varies as F = k/r 2 . Van der Waals reasoned that the distance between gas particles

Copyright
18
P1: OTA/XYZ P2: ABC
THE VIRIAL EQUATION: A PARAMETRIC CURVE FIT 19
increases directly with the volume occupied by the gas; thus, for the attractive force
between particles, he wrote
a
F=
V2
The conversion factor from distance to volume is not needed because it is included in
the parameter a, which is determined experimentally. Pressure is force per unit area
on the walls of a container, so van der Waals added the force term to the pressure and
rewrote the ideal gas equation for one mole of a real gas as the corrected pressure
times the volume remaining after subtracting the volume of the particles:
a
p+ (V − b) = RT
V2
The van der Waals equation is a semiempirical equation because the ideal gas law on
which it is based can be derived from pure theory (see below), but a and b are empirical
parameters found by trial and error. One can start with a pair of plausible estimates
for a and b, vary them, compare the results with measured p, V, T behavior, and select
the values of a and b that give the best agreement with experimental measurements.
Computer routines are available that make many thousands of estimates and give the
best curve fit in a matter of seconds.
A knowledge of b permits one to calculate order-of-magnitude radii of molecules.
For example, the van der Waals radius of methane is 190 pm (picometers, 10−9 m) as
compared to the spectroscopic value (obtained many years later) of 109 pm for the
C–H bond length of methane. Numerous similar calculations give comparable results.
This rough agreement supports van der Waals’s qualitative picture of the excluded
volume of real gases.
2.2 THE VIRIAL EQUATION: A PARAMETRIC CURVE FIT
The virial equation is an example of a more general and frequently more accurate
curve fitting routine than the van der Waals equation, but it that gives less insight into
possible causes of nonideal behavior than the van der Waals equation does. Any data
set can be graphed and fit by an analytical equation (an equation that can be written
out in terms of a limited number of variables and some accompanying parameters).
A parameter is a number entering into an equation that takes on a fixed value for one
system but may change to some other fixed value for another system. For example,
what is usually called the Boyle’s law “constant” pV = k is, in fact, a parameter
because it is valid only for a specified, fixed temperature. Change the temperature
and the parameter takes on a different value, but it acts like a constant as long as you
maintain the temperature fixed. The gas law constant R is a true constant; for one
mole of an ideal gas it is always the same.
Of any two equations, one will be a better fit to a given data set than the other.
Of three equations, one will be a better fit than either of the other two, and so on.
P1: OTA/XYZ P2: ABC
20 REAL GASES: EMPIRICAL EQUATIONS
Computer speed makes it possible to try very many equations and select the best one,
but there is a point of diminishing returns. An equation may have so many parameters
that no one will ever use it or it may follow random fluctuations in the data set that
tell us nothing about the physics of the actual system. If you are not prudent in your
use of curve-fitting programs, you may be calculating the characteristics of a unicorn
to many significant figures. These caveats apply to any curve-fitting problem, not just
those of real gases.
A very nice balance that avoids daunting complexity but achieves good accuracy
is the series equation
y = a + bx + cx 2 + d x 3 + · · ·
which has an infinite number of terms but which is cut off or truncated at some
reasonable number of terms, usually 3 or 4. As applied to real gases, this series is the
virial equation of state:
2 3
RT RT RT
pVm = RT + B2 [T ] + B3 [T ] + B4 [T ] + ···
Vm Vm Vm
The parameters B2 [T ], B3 [T ], and B4 [T ] are called the second, third, fourth, . . .

virial coefficients and the notation Vm = V /n is used to remind us that the volume
taken is a molar quantity. The notation B2 [T ], B3 [T ], B4 [T ], . . . is used to indicate
temperature dependence of the virial coefficients. The square brackets do not indicate
multiplication. By a simple algebraic manipulation, it is possible to express the virial
coefficients in terms of the van der Waals constants and find B2 [T ] = b − a/RT . By
another simple manipulation, one obtains the compressibility factor.
2.3 THE COMPRESSIBILITY FACTOR
The difference between ideal and real gaseous behavior can be made clearer if we
define a compressibility factor Z, a way of indicating the degree of nonideality of
a gas
pVm pVm
Z= =
RT ( pVm )ideal
If Z is less than one, nonideality is largely due to attractive forces between molecules.
If Z is greater than one, the nonideal behavior can be ascribed to the volume taken
up by individual molecules treated as hard spheres or to repulsive forces, or both.
An ideal gas would show a compressibility factor of 1.00 at all pressures. At high
temperatures, the total volume is large for any selected pressure. Molecular crowding
becomes less significant, and attractive or repulsive forces are weaker because they
act over longer distances. The gas approaches ideal behavior and Z approaches the
constant value of 1.00 as p approaches zero.
P1: OTA/XYZ P2: ABC
THE COMPRESSIBILITY FACTOR 21
0.9870
0.9865
Compressibility Factor, Z
0.9860
0.9855
0.9850
0.9845
0.9840
0 5 10 15 20 25
Pressure
FIGURE 2.1 A quadratic least-squares fit to an experimental data set for the compressibility
factor of nitrogen at 300 K and low pressures (sigmaplot 11.0
C ).
Dividing pVm by RT over a set of different pressures at a fixed temperature

gives a series of Z values that can then be plotted against p. This has been done
for nitrogen at 300 K to give Fig. 2.1. Commercial curve-fitting software can be
used to give the least-squares expression for a polynomial fit to the data points.
One needs to select the degree of the polynomial to be fit to the points. The data
set shown in Fig. 2.1 shows a little experimental scatter at the lower pressures and
(perhaps) some slight curvature. Therefore, we selected a simple quadratic fit to the
points.
Real gas law calculations like this one have considerable practical value. The
engineering literature contains data sets of a much more complicated nature, over a
much larger range than Fig. 2.1. The curve-fitting technique is the same, although one
might choose a cubic or quartic curve fit. The output for the simple nitrogen curve fit
is given in File 2.1. The Rsqr (square of the residual), being close to 1.0, indicates
a good fit, although the extrapolated intercept y0 is not as close to 1.0 as we would
like to see it.
The two virial coefficients are quite small, −0.0002 and 1.69 × 10−6 ; nitrogen
is nearly ideal at 300 K over the short pressure range 1–10 bar. The second virial
coefficient is negative, reflecting the gentle downward slope away from ideal behav-
ior. Note that in File 2.1 the notation f=y0+a∗ x+b∗ x∧ 2 is used so that the some-
what overworked parameters a and b appear in a new and different context. Now,
P1: OTA/XYZ P2: ABC
Data Source: Data 1 in N2 molar density

Equation: Polynomial, Quadratic
f=y0+a∗ x+b∗ x∧ 2
R Rsqr Adj Rsqr Standard Error of Estimate
0.9977 0.9954 0.9932 6.7995E-005
Coefficient Std. Error t P
y0 0.9869 6.0263E-005 16377.0857 <0.0001
a −0.0002 1.6559E-005 −9.6859 0.0006
b 1.6977E-006 7.5741E-007 2.2415 0.0885
FILE 2.1 Partial output from a quadratic least-squares curve fit to the compressibility factor
of nitrogen at 300 K (SigmaPlot 11.0
C ).
a = B2 [T ] is the second virial coefficient and b = B3 [T ] is the third. The first

virial coefficient (term rarely used) is 1.0 by definition. In File 2.1, the third virial
coefficient is positive, indicating a slight upward curvature. (Experiments at higher
pressures confirm the curvature.)
Nonideal gas behavior is nearly linear at low pressures. That is why the slope of the
linear function is a measure of the second virial coefficient B2 [T ]. The temperature
variation of the second virial coefficients of helium, nitrogen, and carbon dioxide are
shown schematically in Fig. 2.2. When B2 [T ] = 0, the slope of the virial equation
for Z is zero and Z = 1 over the range. If this is true, the gas is ideal. Helium shows
ideal or nearly ideal behavior over most of the temperature range. Carbon dioxide is
very nonideal over the range, and N2 is in between. This order is pretty much what we
would expect from our qualitative knowledge of the three gases. Helium is a “noble”
gas, CO2 is commonly available in the condensed state as “dry ice,” and atmospheric
N2 is in between. Nitrogen is not as easily driven into the condensed state as CO2 ,
but liquid nitrogen is far easier to produce than liquid helium.
0
He
B2[T]
N2
CO2
300
0 Temperature 600
FIGURE 2.2 The second virial coefficient of three gases as a function of temperature. Notice
the slight maximum in the curve for helium. It is not a computational error, helium really does
that. Intermolecular repulsion brings about a small positive deviation of Z from Z ideal over part
of the temperature range.
P1: OTA/XYZ P2: ABC
THE COMPRESSIBILITY FACTOR 23
2.3.1 Corresponding States

An interesting comparison from among Z factors is shown schematically in Figs. 2.3
and 2.4 which might be Z = f ( p) for two different gases or Z = f ( p) for the same
gas at two different temperatures. Simply by choosing the right two temperatures,
one can make any two gases identical to each other in their degree of nonideality,
that is, their Z factors. When different gases at different temperatures behave in the
same way, they are said to be in corresponding states. To define the state of a real
gas, we must describe it in such detail that a colleague can reproduce it in all of its
physical properties from the description. Because of the essentially infinite number
of physical properties one can measure, this would seem to be a tall order; but, if it is
pure, the number of degrees of freedom for one mole of a real gas is 2 regardless of
its degree of nonideality, so specifying any two properties is equivalent to specifying
all of them. The equation of state is written with two independent variables, but it
doesn’t matter which two.
To summarize up to this point: We are left with a van der Waals qualitative picture
of nonideal gas behavior that is quite reasonable but gives an equation that doesn’t
work very well outside of common laboratory conditions. Our alternative is to rely
upon empirical equations that work quite well in most cases but are hard to interpret.
The term “empirical” as applied to the virial equation in this context has become
somewhat of a misnomer over the years, however, because considerable progress has
been made in theoretical interpretation of the virial equation and Fig. 2.3. Indeed, the
statistical mechanics of these curves and others like them is an active research topic.
1.20
1.15
1.10
1.05
Z
1.00
0.95
0.90
0.85
0 100 200 300 400 500 600 700
p (bar)
FIGURE 2.3 The Z = f ( p) curve for two different gases or for the same gas at two different
temperatures. The unit—bar—is approximately one atmosphere.
P1: OTA/XYZ P2: ABC
p (bar)
73.8
0 1 2
Vm (dm3)
FIGURE 2.4 Three isotherms of a van der Waals gas. The top isotherm is above Tc , the
middle isotherm is at the critical temperature Tc and the bottom curve is below Tc . The critical
pressure is 73.8 bar.
The apparent paradox that there are only two degrees of freedom in the equation of
state of a pure substance which may have an infinite number of terms in an equation
of state is removed by noting that each term contains only the pressure, p, and an
adjustable parameter (not a variable) that is a function of the temperature. Hence the
only true variables in the equation are p and T.
2.4 THE CRITICAL TEMPERATURE
At lower temperatures, gas molecules occupy a smaller volume and move more
slowly than they do at higher temperatures. Attractive forces among molecules or
atoms become more important at lower temperatures. Ultimately, they become so
strong that the gas liquefies. Thus, a useful physical picture of the liquid state is that
liquefaction is the limiting behavior of an extremely nonideal gas and it results from
large interparticle attractive forces.
As the temperature of a real gas is lowered, its deviation from hyperbolic (Boyle’s
law) behavior becomes more pronounced until the p–V curve has become so distorted
that it goes through a horizontal inflection point. The temperature at which this occurs
is called the critical temperature, Tc .
The curves in Fig. 2.4 arise from plotting the van der Waals equation at each of
three temperatures above, at, and below Tc . The locus of p–V points at the same
temperature is called an isotherm. The pressure, volume, and temperature at the
inflection of the critical isotherm define a point called the critical point, which is
unique to each real gas. The coordinates of the critical point (the critical constants)
are the critical pressure pc , critical volume Vc , and critical temperature Tc . Critical
constants vary widely. For example, the critical temperature of helium is 5.19 K while
that for CO2 is 304 K.
Below the critical temperature, the system may be in the liquid or gaseous state,
or it may exist as an equilibrium between liquid and vapor. The term “vapor” means
P1: OTA/XYZ P2: ABC
THE CRITICAL TEMPERATURE 25
gas
3
p
1 4
liquid equilibrium vapor
FIGURE 2.5 Conversion of a liquid to its vapor without boiling (1–4).
gas, but it is usually applied to a gas in equilibrium with its liquid form. When a liquid
is in equilibrium with its vapor, heat can be applied with no change in temperature
but with conversion of some or all of the liquid to its vapor. When spheroids of vapor
rise from the bottom of a heated liquid to the top, we say that the liquid boils. It is
sometimes said that “no gas can be liquefied above the critical point.” This is true,
but it is a little misleading because there is no distinction between liquid and gas
above the critical isotherm. Above the critical isotherm, the system is a supercritical
fluid.
The segment of the critical isotherm forming the upper boundary of the liquid
region is particularly interesting. The system passes from liquid to gas (or back again)
with no discontinuity. That is, it goes from liquid to gas but it does not boil. Points
just below the isotherm represent liquids of low density. Those above it represent
gases of high density. On the isotherm, the liquid and gaseous states become one and
the same.
To get a better feeling for the meaning of the critical isotherm, let us heat a
subcritical liquid (1) to one of its supercritical isotherms (2), expand it (3), and cool
it to its original temperature (4) as in Fig. 2.5. At the end of the process, the liquid
has been transformed to a state that is clearly in the gaseous region, but there is no
discernible phase change (boiling or vaporization → gas) during this process.
2.4.1 Subcritical Fluids

A subcritical curve in Fig. 2.5 has three real roots, predicting three different vol-
umes for the same fluid below Tc . This seems absurd. How can a gas have three
different volumes at the same time? The answer is that the term “fluid,” meaning
that which flows, is more general than “gas.” The term fluid includes both liquids
and gases. The volume of a subcritical liquid is small and is given by the leftmost
intersection (root) of the subcritical isotherm with the horizontal. The volume of
the vapor is large and is given by the rightmost intersection (root). The middle
P1: OTA/XYZ P2: ABC
root of the subcritical isotherm is not observed experimentally but has theoretical
significance.
The constant pressure horizontal that intersects the subcritical isotherm represents
vaporization going from left to right or condensation going from right to left. It is
called an isobar. In vaporization, the volume of the system gets much larger but the
pressure stays the same. Along an isobar, liquid and vapor are in equilibrium. If the
pressure indicated by the isobar is atmospheric pressure, the subcritical isotherm is
at the normal boiling point Tb .
2.4.2 The Critical Density

Below the critical temperature, one can measure the densities of a gas and a liquid at
equilibrium in a closed container. Upon raising the temperature slightly at constant
pressure, the density of the gas increases because liquid vaporizes. The density of
the liquid decreases due to expansion. These densities, one increasing and the other
decreasing, must approach each other as in Fig. 2.6. At the point where they meet,
the densities of the gas and liquid and all other properties become identical. This is
the critical temperature. Determination of the critical temperature of O2 by density
measurements is shown in Fig. 2.6.
153.5
Density
0
115 120 125 130 135 140 145 150 155
T (K)
FIGURE 2.6 Density ρ curves for Liquid and Gaseous Oxygen. The straight line represents
the arithmetic mean density of the liquid and gas (ρl + ρg )/2. The three curves meet at T =
153.5 K. The tabulated value for Tc is 154.6 K (CRC Handbook of Chemistry and Physics
2008–2009, 89th ed.).
P1: OTA/XYZ P2: ABC
THE LAW OF CORRESPONDING STATES, ANOTHER VIEW 27
2.5 REDUCED VARIABLES
The degree of nonideality for real gases is determined by how near the temperature
is to the critical temperature. We can express the “nearness” of the temperature of
a gas to its critical temperature as the unitless ratio T /Tc . This ratio is called the
reduced temperature
T
TR =
Tc
The other reduced variables are defined in the same way. The (unitless) reduced
pressure and reduced volume are, for one mole,
p V
pR = and VR =
pc Vc
These new variables are scaling factors by which we take an entire family of isotherms
suggested by the three representative isotherms in Fig. 2.4 and move, stretch, or
compress them until their critical isotherms coincide. Having done that, at the same
TR , the p R and, VR behavior of all gases fall approximately on the same family
of curves. The gas data have been manipulated into corresponding states (Section
2.3). Knowing the behavior of one gas in terms of its reduced variables p R , VR , and
TR , we know the behavior of any real gas, provided only that we know its critical
constants pc , Vc , and Tc . Needless to say, industrial chemists and chemical engineers
are delighted by this, and they have devoted considerable effort to construct standard
tables and Z-curves in terms of the reduced variables.
2.6 THE LAW OF CORRESPONDING STATES, ANOTHER VIEW
Replacing the parameters a and b and the constant R in the van der Waals equation
for one mole
RT a
p= − 2
(V − b) V
with
Vc 8 pc Vc
b= , a = 3 pc Vc2 , and R=
3 3Tc
we get, with some algebraic manipulation (see Problems and Exercise 2.1),

3 1 8
pR + 2 VR − = TR
VR 3 3
P1: OTA/XYZ P2: ABC
2.0
T R = 1.2
T R = 1.5
T R = 2.0
Z 1.0
0.0
0 2 4 6 8 10
pR
FIGURE 2.7 Compressibility factors calculated from the van der Waals constants.
The van der Waals constants, which are characteristic of individual gases, have been
eliminated in the more general equation, which holds for any gas. Z can be calculated
as a function of p R from the equation of corresponding states as shown in Fig. 2.7.
Portrayal of extensive families of curves like those in Fig 2.7 can be found in the
chemical engineering literature. The equation of corresponding states is approximate
and holds only insofar as the van der Waals equation, on which it is based, holds.
2.7 DETERMINING THE MOLAR MASS OF A NONIDEAL GAS
A nonideal gas, even if it is pure, will generally not occupy a molar volume of 24.789
dm3 at p = 1.000 bar and T = 298.15 K; hence any molar mass computed on the
basis of this molar volume will be wrong. Depending on the temperature, the error
may be 50% or more. If, however, the weight and the volume of a pure gas sample
are known along with the pressure and temperature, the (incorrect) molar mass, often
called the effective molecular weight (EMW), can be calculated from the ideal gas
law. Historically, EMWs were measured at several different pressures, treating a real
gas as though it were ideal. The EMWs were then extrapolated as a function of p to
p = 0 to obtain the true molecular weight. It is fruitless to ask about the meaning of
the properties of a gas at zero pressure; one has simply “extrapolated out” the error
due to nonideality. Extrapolating out is a common applied mathematical device. A
number of more accurate methods for determining molar mass now exist.
Exercise 2.1 The van der Waals Cubic

Show that the van der Waals equation is a cubic.
P1: OTA/XYZ P2: ABC
Solution 2.1 Expanding the van der Waals equation and collecting terms, we get
a
p+ (V − b) = RT
V2
a ba
pV − pb + − 2 = RT
V V
pV 3 − pbV 2 + aV − ba = RT V 2
pV 3 − ( pb + RT ) V 2 + aV − ba =0
Comment: Nonideal Behavior—Another View We know that molecules are not

dimensionless points. Figure 2.8 shows that Boyle’s law, the lower curve, is not
obeyed above about 200 bars pressure for nitrogen. Positive deviation from Boyle’s
law is a high-pressure phenomenon shown by all real gases because all real molecules
occupy a nonzero volume. The real sample volume is larger than it “ought to be” on
the basis of Boyle’s law. Molecules, which may be thought of roughly as hard spheres,
are bumping into each other and refusing to invade each other’s space. The volume
in which the particles can move is the total volume minus the volume actually taken
up by the particles. As higher pressures are imposed, the particle volume becomes a
larger proportion of the total volume; hence the deviation from the Boyle’s law curve
is larger. The aggregate of molecular volume, as distinct from total volume, for all
molecules in the sample is the excluded volume.
600
500
400
T = 273.15 K
p (bar )
300
200
100
0
0.0 0.1 0.2 0.3 0.4 0.5
V (dm3)
FIGURE 2.8 Boyle’s law plot for an ideal gas (lower curve) and for nitrogen (upper curve).
P1: OTA/XYZ P2: ABC
TABLE 2.1 Observed Real Gas Behavior

from 10 to 100 bar Expressed as (p, pVm ).
p (bar) pVm (dm3 bar)
10.0000 24.6940
20.0000 24.6100
30.0000 24.5400
40.0000 24.4820
50.0000 24.4380
60.0000 24.4070
70.0000 24.3880
80.0000 24.3830
90.0000 24.3910
100.0000 24.4120
Exercise 2.2
Given the problem that experimental values of pVm between 10 and 100 bar have
been measured at intervals of 10 bar with the results in Table 2.1, find the analytical
equation that expresses these results (Fig. 2.9).
24.8
24.7
pV (dm3 bar)
24.6
24.5
24.4
24.3
0 20 40 60 80 100
p (bar)
FIGURE 2.9 Experimental values of pVm vs. p for one mole of a real gas.
P1: OTA/XYZ P2: ABC
TABLE 2.2 Observed Real Gas Behavior

Expressed as (p, pVm ).
p (bar) pVm (dm3 bar)
10.0000 24.6940
20.0000 24.6100
30.0000 24.5400
40.0000 24.4820
50.0000 24.4380
60.0000 24.4070
70.0000 24.3980
80.0000 24.3990
90.0000 24.4180
100.0000 24.4600
Solution 2.2 This problem and its solution are expressed using SigmaPlot 11.0 C
plotting software. Load the data set in the form of Table 2.2. Click on statistics →
nonlinear regression → regression wizard → quadratic, → next, specify columns as
x and y variables, and click finish.
The graph in Fig. 2.10 is shown with its fitted quadratic curve. The curve param-
eters are, as they should be, the parameters we started with in the previous problem:
y0 24.7906
a −0.0103
b 6.5341E-005
The parameters are expressed in the form y = y0 + ax + bx 2 + cx 3 + · · ·. Please

notice the slight change in notation: The y intercept is called y0 , the slope is a, and the
quadratic coefficient is b. The general independent variable is x, which is the pressure
p in our case. Translated to the terms of the problem, we have
pV = RT + B [T ] p + C [T ] p 2 + D [T ] p 3 + · · ·
= 24.7906 − 0.0103 p + 6.5341 × 10−5 p 2
where we have truncated the series at the quadratic term.

These two exercises make the important point that any data set can be expressed
as a collection of numbers (table of observations), a graph, or an analytical equation.
Tables, graphs, and empirical equations are merely different ways of saying the
same thing. Difficulties may be encountered (imaginaries, singular points, multiple
real roots, discontinuities, etc.), but they often point to interesting phenomena and
their explanation may lead to new concepts in science. Larger data sets and more
complicated functional behavior can be treated in the same way as this quadratic,
except that the curve fit may be cubic, quartic, and so on.
P1: OTA/XYZ P2: ABC
24.8
24.7
pV m, dm3 bar
24.6
24.5
24.4
24.3
0 20 40 60 80 100
p, bar
FIGURE 2.10 Quadratic real gas behavior.
Problem 2.1
The van der Waals constants for n-octane (a component of gasoline) are a = 37.81
and b = 0.2368. Find V for 1.00 mol of n-octane confined at 2.0 bar pressure and
450 K.
Problem 2.3
Find all three roots in the previous problem.
Problem 2.4
What are the units of van der Waals constants a and b?
Problem 2.5
Using commercial graphing software, produce a 3-D plot of p–V–T for the van der
Waals gas N2 , where a = 1.39 and b = 0.039. What are the units of a and b? What
happens to p as V becomes very small?
P1: OTA/XYZ P2: ABC
Problem 2.6
Find the van der Waals constant b in terms of the critical constants pc , Vc , and Tc .
Problem 2.7
Show that the van der Waals parameter a is a = 98 RTc Vc where the subscript c denotes
the critical point.
Problem 2.8
Find the molar volume of ethane at 50 bar pressure and 400 K, from the van der
Waals parameters a = 5.562 and b = 0.0638.
Problem 2.9
Draw the pV isotherms for an ideal gas at 300, 400, 600, and 800 K. The results
should resemble the corresponding isotherms in the text.
Problem 2.10
The second virial coefficient B [T ] of toluene (methylbenzene) is –1641 cm3 mol−1 at
350 K. First, convert this unit to the more modern unit of dm3 mol−1 , and then find
the compressibility factor Z at this temperature and 1.00 bar.
Problem 2.11
Continuing with the data above, what is the molar volume of toluene vapor at 1.00
bar at 350 K?
Problem 2.12
A real gas follows the equation pV = 24.79 − 0.0103 p + 6.52 × 10−5 p 2 for one
mol. Plot the curve of pV vs. p from 0 to 100 bar and locate the minimum pV
product.
Problem 2.14
Suppose the data set in the previous problem were as shown in Table 2.2. The last
four entries in Table 2.2 differ from those of the previous problem in the second and
third digits beyond the decimal point. The graph in Fig. 2.11 turns up slightly above
p = 70 bar. Fit this data set with a cubic equation.
P1: OTA/XYZ P2: ABC
24.8
24.7
pV, dm3 bar
24.6
24.5
24.4
24.3
0 20 40 60 80 100
p, bar
FIGURE 2.11 Cubic real gas behavior.
Problem 2.15
(a) What is the volume of one mole of CO2 at 366 K and p = 111 bar according
to the ideal gas law?
(b) The critical temperature of CO2 is 305.1 K. The critical pressure and volume
of CO2 are 73.8 bar and 0.0956 dm3 (Laidler and Meiser, 1999). If a compress-
ibility factor of 0.68 is read from a chart of isotherms, what is the reduced
volume of one mole of CO2 at 366 K and p = 111 bar according to the chart?
(c) The van der Waals parameters for CO2 are a = 3.66 and b = 0.0429. What
is the volume of one mole of CO2 at 366 K and p = 111 bar according to the
van der Waals equation?
(d) What is the volume of one mole of CO2 at 366 K and p = 111 bar according
to the corresponding states equation? What is the volume?
(e) The critical constants for helium are Tc = 5.3 K, pc = 2.29 bar, and Vc =
0.0577 dm3 mol−1 . What is the volume of one mole of He at 7.95 K and
p = 2.75 bar according to the corresponding states equation? What is the
volume?
P1: OTA/XYZ P2: ABC
3
THE THERMODYNAMICS
OF SIMPLE SYSTEMS
Thermodynamics, literally heat motion, is one of the pillars of physical chemistry

and one of the great achievements of modern science. Classical thermodynamics is
the study of those quantities that are conserved, like energy, enthalpy, and free energy.
Statistical thermodynamics is the link between this great theoretical edifice and that
of quantum mechanics.
3.1 CONSERVATION LAWS AND EXACT DIFFERENTIALS
Much of nineteenth-century thermodynamics was devoted to the discovery, definition,

and characterization of physical quantities that are conserved. The simple statement
energy is conserved is one of the ways of stating the first law of thermodynamics.
Conservation can be illustrated by carrying a rock up a hill and then bringing it back
down again. The potential energy of the rock in the gravitational field of the earth
increases on the way up and decreases on the way down, but it is the same at the end
of the process as it was at the beginning; it is conserved. By contrast, work is not
conserved. Carrying the rock around the circular path up and down the hill can be
easy or hard, depending on the path. Also, the work done on the way up is not the
same as the work that can be obtained by allowing the rock to drive some kind of
primitive motor as it rolls down the hill.
Work is not the only factor that has to be taken into account. The difference between
an easy path and a hard path up the hill is illustrated by the simplified model of a
mass being pushed up over the same change in height (altitude) but over two different

Copyright
35
P1: OTA/XYZ P2: ABC
36 THE THERMODYNAMICS OF SIMPLE SYSTEMS
inclined planes, one rough and the other smooth. The potential energy change is the
same, but the rough plane requires more work and it produces some amount of heat
due to friction. Heat is not conserved over the cyclic path either. We symbolize the
nonconservation of work or heat by the sum, that is, the
integral, ofinfinitesimal heat
or work increments dq or dw over the cyclic path dw = 0 or dq = 0. But we
have already said that energy U is conserved over a cyclic path so dU = 0.
One of the great discoveries of Western science is that the infinitesimal increment
of the energy of a thermodynamic system is the sum of an infinitesimal increment in
work done on the system and an infinitesimal increment of the heat put into a system1
dU = dw + dq

The statement dU = dq + dw is another of the many equivalent ways of stating the
first law of thermodynamics. Although the sum of work w and heat q is conserved,
we don’t know the ratio of w to q or even if they have the same sign, except by
experiment. The law of conservation of energy is the accumulated knowledge gained
from very many controlled observations over three centuries. It cannot be derived
from simpler principles.
3.1.1 The Reciprocity Relationship

For some functions u = u(x, y), the differential
du = M(x, y)dx + N (x, y) dy
has the property that

∂u ∂u
M(x, y) = and N (x, y) =
∂x y ∂y x
that is,

∂u ∂u
du = dx + dy
∂x y ∂y x
If this is true, then du is called an exact differential. If it is not true, then du is an

inexact differential. By the Euler reciprocity relationship, we have
∂ 2u ∂ 2u
=
∂ x∂ y ∂ y∂ x

∂ ∂u ∂ ∂ ∂u ∂
= N (x, y) = = M(x, y)
∂x ∂y x y ∂x ∂y ∂x y ∂y
x
1 In the
example of the mass being pushed up a rough plane, frictional heat is lost (goes out of the system);
hence the sign on dq is reversed: dU = dw in − dqout .
P1: OTA/XYZ P2: ABC
THERMODYNAMIC CYCLES 37
so the condition
∂ ∂
N (x, y) = M(x, y)
∂x ∂y
is a test for exactness.
3.2 THERMODYNAMIC CYCLES
The problem of inexact and exact differentials can be expressed in several ways.
Taking work as an example, work w is not a thermodynamic function. The differential
of the work dw is not exact. The work done in a thermodynamic process depends
upon the path. The integral
V2
w= f (T ) p dv
V1
is a line integral. The line integral cannot be evaluated until we know f (T ), an

arbitrary function of the temperature. There is an infinite number of paths (ways of
getting from V1 to V2 ), hence an infinite number of solutions to the work integral.
Consider two ways of taking a gas from thermodynamic state A with molar volume
V1 to state B with molar volume V2 . Let the substance be in the gaseous state. Let the
first path be an isothermal compression from 1.00 bar to 10.0 bars at 290 K followed
by an isobaric (constant pressure) temperature rise from 290 K to 310 K. The second
path will start with the isobaric temperature rise from 290 K to 310 K, followed by
an isothermal compression from 1.00 bar to 10.0 bars at 310 K. The beginning and
end points of the process are the same for both paths. For simplicity, assume that the
gas is nitrogen, which is nearly ideal over this temperature and pressure range. The
two paths are shown in Fig. 3.1.
The volume V = RT / p can be calculated after each step. The results are shown
at each corner of the rectangular diagram below. The volumes at each corner of the
diagram are larger on the left (before compression) than on the right and larger at the
top (after heating) than on the bottom.
2660 266.0
B
A
2441 244.1
P1: OTA/XYZ P2: ABC
Second
Path
B
310
T
290
A First
Path
1 10
p
FIGURE 3.1 Different path transformations from A to B.
The work done on the system in the first compression (lower horizontal) is
V2 V2 V2
RT dV V2
w =− p dV = − dV = − RT = −RT ln
V1 V1 V V1 V V1
= −RT ln 0.100 = −8.314(290) ln 0.100 = 5552 J
The work done in the heating step (rightmost vertical) is isobaric with p = 10.0 bar:
w = p(V2 − V1 ) = 10.0 (266.0 − 244.1) = 219 J
so the total over the first path is 5771 J.

Over the second path, the leftmost vertical gives w = 1.00(2660 − 2441) = 219 J;
thus the two isobaric steps require the same work into the system. The isothermal
compressions are not the same. The topmost horizontal is
−RT ln 0.100 = −8.314(320) ln 0.100 = 6126 J
so the total work done over the second path is obviously not the same as by the first
path. The difference is 574 J.
3.2.1 Hey, Let’s Make a Perpetual Motion Machine!

Let’s run the thermodynamic system in Fig. 3.1 around a cycle such that the transfor-
mation over the second path runs backwards; that is, the top horizontal is an expansion
instead of a compression and the leftmost step is a temperature decrease. The first
path remains as before.
P1: OTA/XYZ P2: ABC
LINE INTEGRALS IN GENERAL 39
The work of the leftmost and rightmost steps cancel −219 + 219, but the topmost
horizontal produces −6126 J of work and the bottom horizontal takes up only 5552 J
of work to return the machine to its original state. The cycle produces −6126 +
5552 = −574 J of work (negative because work goes from the machine into the
outside world) and has been returned to its original state, ready to produce work
−574 J, −574 J, −574 J, . . . over infinitely many work cycles, running forever.
Five or six generations of garage scientists have made the sad discovery that such
machines never work. However sad it may be for would-be millionaire inventors, the
impossibility of a perpetual motion machine leads the physical chemist to a treasure
trove of thermodynamic theory through the second law of thermodynamics to the
inspired work of the American thermodynamicists J. W. Gibbs and G. N. Lewis.
3.3 LINE INTEGRALS IN GENERAL
If you have a straight brass rod—say, 70 cm in length and 1.0 cm2 in cross section—it
is a simple matter to determine its weight even though the rod may be inaccessible
to you as part of a machine, so that you cannot simply weigh it. Multiply the density
of the brass used in its manufacture, ρ in g/cm3 , by 70 × 1.0 = 70 cm3 = 70 ρ and
you have the answer in grams. We have essentially set up an x axis and performed
the integration
70
M =ρ dx = ρ(70 − 0)
0
Most machinists (and lots of other people) would laugh at how we have made a simple
problem complicated.
But suppose the rod is bent (Fig. 3.2). Now setting up an x axis for the bar is not
the best way of finding its mass. One x increment is not the same as another because
the rod is not collinear with the axis.
The corresponding integral is
b
M =ρ ds = ρ(b − a)
a
FIGURE 3.2 Different segments of a curved rod.

P1: OTA/XYZ P2: ABC
We still use the practical method of length times cross-sectional area to get the total
volume which we then multiply byρ, but the corresponding integral is no longer a
conventional integral over dx; rather, it is the summed infinitesimal increments along
the length of the rod ds, not coincident or even parallel to the x axis.
Now suppose that the density of brass is not constant in the rod but that it varies
along the length of the rod according to some known function ρ(s). The integral is
b
M= ρ(s) ds = ρ(s) (b − a)
a
Integrals taken along some curve C with arc length s, not one of the coordinate axes,
are called curvilinear or line integrals. In general, the integral of some function of
arc length f (s) along a curve C is

I = f (s) ds
C
3.3.1 Mathematical Interlude: The Length of an Arc

By Pythagoras’s theorem for the hypotenuse of a right triangle, the length s of a
short part of a curve (arc) in x–y space is approximately (Barrante, 1998)

1/2
s ≈ x 2 + y 2
Multiply and divide by x to find

2 1/2
y
s ≈ 1 + 2
x
x
or, in the infinitesimal limit,

2 1/2
dy
ds = 1 + dx
dx
b
For the length from a to b of a finite arc, find the integral s = ds.
a
s y
x
FIGURE 3.3 Pythagorean approximation to the short arc of a curve.
P1: OTA/XYZ P2: ABC
STATE FUNCTIONS 41
3.3.2 Back to Line Integrals

Suppose some property f (x, y) is associated with a curve C in the x–y plane and we
wish to find the line integral between certain limits along that curve. The appropriate
integral is

I = f (x, y) ds
C
The integration is to be carried out, not along an axis as in the case of a simple
integral, but along the curve in x–y space. If we have y = f (x) that specifies the path
in the x–y plane, we also have ds along the curve from Section 3.3.1 so the integral is
2 1/2 2 1/2
dy dy
I = f (x, y) 1 + dx = f (x, f (x)) 1 + dx
C dx C dx
3.4 THERMODYNAMIC STATES AND SYSTEMS
A thermodynamic system is any part of the universe we want to look at. The rest of
the universe is its surroundings.
system + surroundings = universe
A system is usually defined in such a way as to be manageable; for example, a

reaction flask containing chemicals is a system. The surroundings may be a constant
temperature bath in which the system is immersed, along with the rest of the universe.
An isolated system does not exchange energy or matter with its surroundings. A closed
system exchanges energy, but not matter, with the surroundings. A piston fitted to
a cylinder that can do work and exchange heat with its surroundings, but does not
leak, is a closed system. An open system exchanges both energy and matter with its
surroundings. An animal or plant is an open system.
3.5 STATE FUNCTIONS
A thermodynamic state function is one of several conserved mathematical functions

describing a property of a system. Energy is a thermodynamic state function. The
term thermodynamic property is used in the same way. To define the state of a system,
we must describe it in all of its physical properties, but we already know that there
are only three degrees of freedom for a pure substance, two if you specify molar
properties as we normally do. In nonuniform systems, one of the variables may be
different at different locations. For example, the temperature may not be the same
everywhere in a closed room. Other variables will then be nonuniform as well. These
systems are more difficult to treat mathematically, but they have been described in a
P1: OTA/XYZ P2: ABC
series of brilliant twentieth-century researches by Prigogine (Nobel Prize, 1977) and

coworkers. Here we shall restrict ourselves to uniform systems and processes that
are carried out very slowly so that thermodynamic variables change smoothly and do
not suffer the discontinuities and multiple-valued problems encountered in sudden
processes like explosions.
Each point on a p–V graph represents a thermodynamic state. Moving from one
thermodynamic state to another impelled by a change of one or more independent
variables is described as a thermodynamic transition. Each transition from V1 to V2
brought about by a change from p1 to p2 is represented by a curve on the p–V graph.
Real transitions can be carried out very slowly. When this is done, they can be made
to approach a limiting process of an infinitely slow transition that moves from one
state to another by infinitesimal steps, each of which is an equilibrium state. A finite
transition carried out in such a way as to approach an infinite sequence of equilibrium
states is called a reversible transition.
All this talk about transitions through equilibrium states is, of course, self-
contradictory; if a state is at equilibrium, it isn’t transiting anywhere.2 The reversible
process is an idealization which is taken seriously by otherwise skeptical scientists
because it brings the tremendous power of calculus to bear on thermodynamics. With
calculus and the concept of reversible transformations, we can build a majestic frame-
work that encompasses all classical thermodynamic change, including the change in
thermodynamic properties during chemical reactions. In addition to the utility of the
idealized reversible process, it is astonishing how nearly some real chemical systems
can be brought to true reversibility (see especially electrochemistry).
3.6 REVERSIBLE PROCESSES AND PATH INDEPENDENCE
One can determine the change of a thermodynamic state function if one knows the
initial and final states of the system. Let Ui and U f be the energy of a system in its
initial and final states. The change in U is
U = U f − Ui
Neither U f nor Ui is known (classically) in an absolute sense, but this difficulty is

easily circumvented by defining Ui ≡ 0 (or U f ≡ 0) in some arbitrary state. Suppose
we define a potential energy Ui ≡ 0 in a coordinate system with the z axis in the
direction of a field. Then we increase z by 1000 meters (lift the mass). This changes
the energy in the field. The unit acceleration in the gravitational field is very roughly
10 newtons (N); hence a mass of 1.0 kg increases in potential energy by about
U = mgz = (1.0) 10(1000) = 10000 = 104 J = 10 kJ
2 We ignore quantum fluctuations in classical thermodynamics.

P1: OTA/XYZ P2: ABC
REVERSIBLE PROCESSES AND PATH INDEPENDENCE 43
V1 V2
V
FIGURE 3.4 The energy change for reversible expansion of an ideal gas. The area under the
p–V curve is the p–V work done by an expansion from V1 to V2 .
where the unit of energy is the joule (J). One does not know how much work it will
take to bring this process about because the amount of work lost to friction is not
known. Work is not a thermodynamic function.
The ideal expansion of a gas driving a piston in the absence of frictional or other
heat loss produces work equal to the energy change. The energy change can be found
by integrating over the work which now follows a defined path and which in this
respect behaves like a state function (Fig. 3.4). In the absence of heat loss, work is
defined as the integral of force F over displacement ds from position s1 to s2 :
s2
w= F ds
s1
For an idealized piston, this is the same as the integral of p over the change dV. The
amount of work done by expansion against a piston is represented by the area under
a p–V curve in Fig. 3.4. It can be written analytically as
V2
w= p dV
V1
The amount of work done by the system on the surroundings during an expansion at
constant pressure is
V2
w=p dV = pV
V1
At constant temperature (isothermal conditions) we have
V2 V2
n RT dV V2
w= dV = n RT = n RT ln
V1 V V1 V V1
P1: OTA/XYZ P2: ABC
for an ideal gas. We can also define an adiabatic process, which is a perfectly insulated
process. Before doing so, we shall need the concept of heat capacity.
3.7 HEAT CAPACITY
When heat q is put into a system its temperature rises. The change in temperature is
proportional to the amount of heat
q = CT
where C is a proportionality constant. C depends on both the nature of, say, a chemical
substance and the amount taken; it is an extensive property (Section 1.6). C is really
a parameter because it is different for different substances. It is common experience
that metals heat up faster over a flame than does water. They have different capacities
to absorb heat; hence the parameter C is called the heat capacity.
To use calculus in working with the heat capacity, it is necessary to replace the ap-
proximate macroscopic observation C = q/T with the infinitesimal C = dq/dT .
Furthermore, we chemists carry our ordinary bench reactions under conditions of
constant (atmospheric) pressure, and thermochemists carry out combustion reactions
inside a closed bomb. The heat capacity under constant volume conditions is not
exactly the same as the heat capacity under constant pressure conditions, so we
distinguish between the two heat capacities C V and C p as

∂q ∂q
CV = and Cp =
∂T V ∂T p
The infinitesimals in the heat capacity equations are partials because each parameter
is defined holding either V or p constant.
3.8 ENERGY AND ENTHALPY
Having stipulated constant volume for the first of the heat capacity expressions, the
work p dV disappears for a system that can do only work of expansion dV against a
pressure p
dU = dq + p d V
Consequently,

∂q ∂U
CV ≡ =
∂T V ∂T V
P1: OTA/XYZ P2: ABC
ENERGY AND ENTHALPY 45
It would be very convenient here to have a thermodynamic function that plays the
same role for constant pressure transformations that energy plays for constant volume
transformations. We can define such a property. It is called the enthalpy, and it is
the heat produced or absorbed under the usual constant pressure conditions that
characterize the reactions we carry out in the lab. The enthalpy has the analogous
definition

∂q ∂H
Cp ≡ =
∂T p ∂T p
Now the volume is not constant but may vary in such a way as to do work on
the surroundings or for the surroundings to do work on the system, so dU = dq +
p dVand

∂q ∂H ∂U − p dV
Cp = ≡ =
∂T p ∂T p ∂T p
We have in effect constructed a new thermodynamic variable by specifying the path
H ≡ U − pV
dH = dU − p dV − V dp
where V dp is zero over the stipulated path because p = constant. The p dV work over
a stipulated path is conserved; that is, it sums to zero over a circular path. Enthalpy
is thus the sum of a thermodynamic function and a conserved function, and therefore
it must be conserved.
Enthalpy is a thermodynamic state variable.
Enthalpy is the constant pressure analog of energy in processes like chemical

reactions. It can be handled mathematically in the same way that energy is han-
dled and should be thought of almost as its twin. In many semiquantitative dis-
cussions the distinction between energy and enthalpy is ignored but it should be
borne in mind, because it is important in rigorous treatments like the correction of
quantum mechanical values for the energy of a molecule at 0 K to the enthalpy
at 298 K.
The difference between enthalpy and energy is the pV work done on or by the
system. For ordinary chemical reactions this is often negligible but if a gas is taken
up or produced during a chemical reaction, U and H will be different because pro-
duction of a gas represents an expansion against the atmosphere and consump-
tion of a gaseous reactant is a contraction of the system driven by pressure of the
surroundings.
P1: OTA/XYZ P2: ABC
3.9 THE JOULE AND JOULE–THOMSON EXPERIMENTS
The Joule experiment is important because it showed that there are no thermal effects
arising from expansion of a gas. The Joule–Thomson experiment is important because
it showed that there are thermal effects arising from expansion of a gas. What?
Wait! There’s an explanation . . . .
The Joule series of experiments was the earlier of the two. It was carried out
with simple apparatus and relatively insensitive thermometers. The Joule apparatus
consisted of two chambers, one filled with a gas at pressure p and the other evacuated,
with the chambers being connected by a short tube with a stopcock. The apparatus
was placed in a bath and allowed to come to thermal equilibrium. The stopcock
was opened, allowing expansion of the gas into the evacuated half of the apparatus.
The temperature change was measured and a null result was recorded—hence the
conclusion that expansion of a gas produces no thermal effect. The Joule experiment
is analogous to Boyle’s law in that it is almost correct for most gases under mild
conditions (pressure change not too great). It is an ideal gas law. Joule happened to
be rich (he never had to be distracted by earning a living) and he was also smart. He
did not really believe the results of his experiment.
The Joule–Thomson experiment is a refinement of the Joule experiment intended
to find the very thing that the Joule experiment failed to find: the thermal effect of
expanding a gas. By means of a piston, a gas was driven through a porous plug from
one chamber at high pressure into a second chamber at low pressure, thus expanding
the gas (for more detail, see Klotz and Rosenberg, 2008). He then measured the
temperature difference between the two chambers and found that it was not null. The
result was more in accord with modern experience with highly pressurized gases. The
expanding gas cools. The cooling factor is called the Joule–Thomson coefficient µJT

∂T
µJT ≡
∂p H
The partial on the right-hand side is subscripted H because the process, though it in-
volves a change in T, involves no change in H . It is isenthalpic. The Joule–Thomson
coefficient is usually positive (for an expansion, dp < 0 and µ J T > 0, so the gas
cools). There are a few exceptions, which warm up on expansion starting from some-
where around room temperature. For these few gases, µ J T < 0 at room temperature;
but if they are expanded at low temperatures, µ J T changes to a positive value. As
specific examples, nitrogen has µ J T ∼ = 0.6 at 200 K; hence N2 can be cooled by
expansion starting at 200 K and brought to a temperature so low that it is liquefied.
Helium has µ J T ∼ = −0.06 and cannot be liquefied by expansion starting at 200 K.
Helium must be precooled far below 200 K to where µ J T becomes positive for it to be
liquefied by expansion. If expansion is carried out at a sufficiently low temperature,
all gases can be liquefied.3 The point at which µ J T changes from + to – is called the
inversion temperature.
3 Some special effects arise with helium isotopes.

P1: OTA/XYZ P2: ABC
THE JOULE AND JOULE–THOMSON EXPERIMENTS 47
Not surprisingly, µ J T is related to the van der Waals a and b, most importantly the
van der Waals parameter of attraction a. For gases having attractive interactions (most
of them at room temperature) expansion against their attractive forces does internal
work to separate the gaseous particles, which is why the gas cools and µ J T > 0. For
helium, neon, hydrogen, and so on, the dominant forces are repulsive hence µ J T < 0
at room temperature. At low temperatures, attractive forces become dominant for all
gases, so µ J T changes sign.
The Joule–Thomson inversion temperature Ti can be related to van der Waals a
and b by the equation
2a 3abp
− 2 2 −b =0
RTi R Ti
This equation is a quadratic in Ti ; hence double roots are possible. Indeed, two
inversion temperatures, upper and lower, are found at some pressures. At very high
pressures, the two roots approach each other and become identical as shown in
Fig. 3.5. On the high branch of Fig. 3.5, the upper inversion temperature, the second
term in the inversion temperature equation becomes unimportant because it has Ti2
in the denominator. Now
2a ∼
=b
RTi
2a
Ti ∼
=
bR
Hence the upper Ti can be calculated if a and b are known. The second equation above
is used to estimate the upper inversion temperature of real gases from a and b (which
are themselves estimates). Though approximate, this value is important in practical
problems such as production of liquefied gas for cooling certain low-temperature
experimental instruments.
Root 1
Ti
Root 2
FIGURE 3.5 Inversion temperature Ti as a function of pressure. The temperature extremum

dTi /d p for nitrogen is at about 370 atm.
P1: OTA/XYZ P2: ABC
3.10 THE HEAT CAPACITY OF AN IDEAL GAS
An ideal gas consists of point particles that cannot vibrate or rotate. It can have
only kinetic energy Ukin = 12 mv2 . For a large collection of particles, the total kinetic
energy is
Ukin = 12 N m ū 2
where N is the number of atoms or molecules and ū is their average speed. We also
know from the kinetic theory of gases that
1
pV = RT = 13 N A m ū 2 = 2
3 2
N A m ū 2 = 23 Ukin
This (slightly circular) reasoning leads to the simple statement
Ukin = 32 RT
for one mole of an ideal gas. No direction is favored over any other in Cartesian
3-space, so we can split the kinetic energy components into three equal parts of
1
2
RT per degree of freedom along any arbitrary x, y, z space coordinates. This division
is general. On a molar basis, we expect to find 12 RT of energy per mole per degree
of freedom or 12 k B T per particle (molecule or atom) per translational, rotational,
vibrational or, rarely, electronic degree of freedom. (Recall that k B is the gas constant
per particle.)
If the gas is ideal, we obtain a molar heat capacity
3
∂U ∂ RT
CV = = 2
= 32 R
∂T V ∂T
Since R ∼
= 2 cal K−1 mol−1 , the heat capacity of an ideal gas is about 3 cal K−1 mol−1 =
12.5 J K mol−1 . Table 3.1 shows that this is true for the monatomic gases helium He
−1
and mercury vapor Hg but that it is not true for more complicated molecular species.
TABLE 3.1 Heat Capacities and γ for Selected Gases.
Gas C V (J K−1 mol−1 ) C p (J K−1 mol−1 ) γ (unitless)
He 12.5 20.8 1.67

Hg 12.5 20.8 1.67
H2 20.5 28.9 1.41
NH3 27.5 36.1 1.31
Diethyl ether 57.5 66.5 1.16
P1: OTA/XYZ P2: ABC
THE HEAT CAPACITY OF AN IDEAL GAS 49
To find C p we notice that, since H = U + pV , we have

∂H ∂U
C p − CV = −
∂T p ∂T V

∂U ∂ pV ∂U
= + −
∂T p ∂T p ∂T V
For an ideal gas, we recall the Joule experiment which shows that the energy is a
function of T only U = f (T ), hence

∂U ∂U
=
∂T p ∂T V
and

∂ pV ∂ RT ∂T
C p − CV = = =R =R
∂T p ∂T p ∂T p
Now we see that
C p − C V = R = 8.3 J K−1 mol−1
and
Cp ∼
= 12.5 + 8.3 ∼
= 20.8 J K−1 mol−1
A useful unitless parameter is γ = C p /C V . From the thermodynamics we have

developed so far, γ = C p /C V = 20.8/12.5 = 1.66. Indeed, for He and Hg vapor, γ
is close to the ideal value. For more complicated molecules, however, γ begins to
fall off significantly. For ammonia vapor, γ has dropped to 1.3 and it appears to be
approaching 1.0 for the relatively complicated molecule diethyl ether.
The problem is that more complicated molecules can rotate and vibrate. The
NH3 molecule resembles a pyramid with a triangular base. It can rotate with 3
degrees of rotational freedom in 3-space. The total number of degrees of freedom for
translational plus rotational motion is 3 + 3 = 6; hence the molar kinetic energy is
Ukin = = 6(1/2RT ). From this, by the same reasoning as before, we get C V = 3R
and C p = C V + R = 3R + R = 4R. Taking R = 8.3 J K−1 mol−1 , C V = 3R ∼ = 24.9
J K−1 mol−1 , and C p = 4R = 33.2. This leads to γ = 33/25 = 1.33 (unitless) in good
agreement with the experimental value for ammonia, which is 1.31 for γ , but the
results are not so good for C V and C p . Hydrogen is an intermediate case because it
normally has enough thermal energy to rotate end-over-end (in x, y 2-space) but not
enough to spin on its axis.
Evidently, hydrogen and ammonia rotate but do not vibrate under the conditions
of Table 3.1. Most chemical bonds also stretch and bend. These motions yield extra
P1: OTA/XYZ P2: ABC
Cp
FIGURE 3.6 Typical heat capacity as a function of temperature for a simple organic
molecule. Allowed modes of motion are gradually activated as the gas is warmed and more
thermal energy becomes available. (See Klotz and Rosenberg, 2008 for more detail.)
degrees of freedom, one for each possible mode of motion. As in the hydrogen case,
not all modes of motion are activated. At any given temperature, a molecule may
have many degrees of freedom available to it but not enough thermal energy to fully
activate all modes. For this reason, heat capacity curves are sigmoidal (S-shaped)
starting from zero at 0 K, where there is no motion at all, and rising gradually, as
modes of motion are activated, to a limiting value determined by how complicated
the molecule is.
3.11 ADIABATIC WORK
Because partial derivatives like (∂U /∂ T ) p , and so on, can be handled just as though
they were algebraic variables, it is possible to develop quite an arsenal of equa-
tions relating the first law quantities described so far and to expand them to include
other variables (Klotz and Rosenberg, 2008). An important concept is that of adi-
abatic (perfectly insulated) work done on or by a gas. The work dw behaves like
a thermodynamic function because the path has been specified by setting q = 0.
Now dU = dq + dw = p dV for a system restricted to pressure–volume work. The
energy U is a state variable U = f (V, T ) for one mole, so

∂U ∂U
dU = dV + dT
∂V T ∂T V
and

∂U ∂U
dV + dT + p dV = 0
∂V T ∂T V
The first term above drops out if we consider expansion of an ideal gas because the
functional dependence on V disappears (Joule experiment). Also, (∂U /∂ T )V = C V
P1: OTA/XYZ P2: ABC
ADIABATIC WORK 51

and p = RT V so
RT dT dV
C V dT + dV = C V +R =0
V T V
Since C p − C V = R, we obtain
dT dV dT C p − C V dV dT dV
CV +R = + = + (γ − 1) =0
T V T CV V T V
Integrating between limits, with a little algebraic manipulation, yields
T V γ −1 = k and pV γ = k
where k and k are constant.

The expression pV γ = k looks like Boyle’s law except for the parameter γ ,
which is always greater than 1. The presence of γ causes the pressure p = k /V γ to
be lower at any point during the expansion than it is during the isothermal (Boyle’s
law) expansion (upper curve in Fig. 3.7).
The difference between the two expansions is in the heat that flows or does not flow
into the system to maintain T = const. In the isothermal case heat transfer is allowed.
Heat is not allowed into the system in the adiabatic case where dq = 0 by definition.
Without a compensating heat flow, the adiabatic system cools during expansion and
the pressure is always lower at any specified volume than it is in the isothermal
case ( p = k /V γ , γ > 1.0). The entire pV curve falls below the isothermal curve in
Fig. 3.7.
V1 V2
V
FIGURE 3.7 Two expansions of an ideal gas. The upper curve is isothermal and the lower
curve is adiabatic. The adiabatic expansion does less work because there is no heat flow into
the system.
P1: OTA/XYZ P2: ABC
PROBLEMS AND EXAMPLE
Example 3.1 Line Integrals

What is the line integral of the function
y) = x y over the parabolic curve
f (x,
y = f (x) = x 2 /2 from (x, y) = (0, 0) to 1, 12 ?

Solution 3.1 One way of writing a line integral of the function I = C f (x, y) ds
2 1/2
dy
over the curve C is to specify y = f (x) and ds = 1 + dx. For example,
dx

y) = x y over the parabolic curve y = f (x) = x /2
2
integrating the function f (x,
from (x, y) = (0, 0) to 1, 12 . We have (Steiner, 1996)

x2 x3
f (x, y) = x y = x =
2 2
and

x2
d
dy 2
= =x
dx dx
Thus,
2 1/2
dy 1
x3
1
I = f (x, f (x)) 1 + dx = 1 + x 2 2 dx
C dx 0 2

1 1
1/2
= x6 + x8 dx
2 0
where the limits of integration are the limits on x. Integration by Mathcad

C gives

1 1
.5
x6 + x8 dx = 0.161
2 0
Problem 3.1
One expression of a line integral is

F(x, y) dx + G(x, y) dy
C
where the subscript C indicates a line (or curve) integral. If F(x, y) = −y, G(x, y) =
x y, and the line is the diagonal from x = 1 to y = 1 (Fig. 3.8). Carry out the
integration.
P1: OTA/XYZ P2: ABC
PROBLEMS AND EXAMPLE 53
1
x
FIGURE 3.8 C = Diagonal along x = 1 to y = 1.
Problem 3.2
If F(x, y) = −y and G(x, y) = x y, evaluate the line integral over the quarter-circular
arc from x = 1 to y = 1 (Fig. 3.9). Notice that the beginning and end points are the
same and the functions are the same as in the previous problem but the path is
different.
1
x
FIGURE 3.9 C = Quarter-circular arc.
Problem 3.3
A mass m of 20.0 kg is raised to a height h of 20.0 m and allowed to drop. Ignoring
air resistance, what is its speed when it hits the ground? What is its kinetic energy?
Problem 3.4
A 20-kg rock was carried up a hill that is 20 m high.
(a) What was the energy increase in the rock?

(b) What amount of work was done on the rock when carrying it up the hill?
(c) What is the kinetic energy gained by the rock when it rolls down the hill?
P1: OTA/XYZ P2: ABC
Problem 3.5
In a Joule experiment, two 20.3-kg weights fell 1.524 m to drive a paddle wheel
immersed in 6.31 kg of water. The experiment was repeated 20 times, after which
the temperature of the water bath was found to have risen by 0.352 K. What is the
mechanical equivalent of heat in J K−1 according to this experiment?
Problem 3.6
Show that dw is an inexact differential for one mole of an ideal gas undergoing a
reversible expansion.
Problem 3.7
The definition of the calorie is that amount of heat that is necessary to raise 1 g of
water 1 K. As it stands, this definition is approximately valid for temperature changes
not too far from room temperature. Use this approximate definition to predict the
final temperature of 200 g of water at 283 K mixed with 450 g of water at 350 K.
Problem 3.8
Show that, for an ideal gas with constant C V , we have
CR
T2 V1 V
=
T1 V2
Problem 3.9
From the kinetic theory of gases, we get the expression U = 32 RT for the energy of
an ideal monatomic gas. Show that dU is an exact differential.
Problem 3.10
(a) How much energy is required to heat 1.0 mol of an ideal monatomic gas at
constant V from 25.0 to 75.0 K?
(b) What is the heat input if the process is carried out at constant pressure?
Problem 3.11
A 10.0-g piece of iron was heated to 100.0◦ C by immersing it in boiling water
and then quickly transferring it to an insulated beaker containing 1000 g of water
at 25◦ C. What was the final temperature of the water? The specific heat of iron is
0.449 Jg−1 K−1 . The specific heat of water is 4.184 Jg−1 K−1 .
P1: OTA/XYZ P2: ABC
Problem 3.12
The ratio γ = C p /C V is 1.40 for nitrogen, N2 . The speed of sound in air (mostly
nitrogen) is said to be

γ RT
v sound ∼
=
M
where M is the molar mass of nitrogen. Find the value of v sound at 273 K and compare
it with the experimental value of 334 m s−1 .
P1: OTA/XYZ P2: ABC
4
THERMOCHEMISTRY
Einstein once said, “Some things are simple but not easy.” Although he did not
have thermochemistry in mind, his comment applies to this time-honored branch
of physical chemistry. Thermochemistry is simple: Run a chemical reaction and
measure the temperature change (if any). But it is not easy. Anyone trying to do
this job at state-of-the-art precision will soon be enmeshed in technical problems
that try the patience of Job. Entire institutes of experimental science exist just for
precise measurement of the heat of chemical reactions. If governments are willing to
spend millions of dollars to support acquisition of thermochemical data, there must
be some significant advantage to be gained from them. That, in part, will be discussed
in this chapter. Today, computers play a large role in this field. The most significant
advance in thermochemistry in the last decade is the calculation of thermochemical
quantities from quantum mechanical first principles. That also will be introduced in
this chapter.
4.1 CALORIMETRY
A calorimeter is a device intended to measure heat given out or taken up when

a chemical reaction or a physical change, such as a change of state, takes place.
A Styrofoam coffee cup with a thermometer is a crude calorimeter. The word has
historical significance; it was once supposed to measure the amount of “caloric”
flowing into or out of a system. We have long since discarded the caloric theory but

Copyright
56
P1: OTA/XYZ P2: ABC
ENERGIES AND ENTHALPIES OF FORMATION 57
we still measure the amount of heat in calories or kilocalories, related through the
conversion factor 4.184 to the number of joules or kilojoules.
4.2 ENERGIES AND ENTHALPIES OF FORMATION
Many compounds can be formed by direct combination of their elements. An example

is carbon dioxide (CO2 ). When a small measured amount of carbon in the common
form of graphite, C(gr), is burned in O2 in a closed steel container called a bomb,
the process causes a small temperature rise T in the immediate surroundings called
the bath. The heat capacity of the bath having been previously determined by an
electrical calibration, one can find qV , the amount of heat given off by the combustion
at constant volume:
C(gr) + O2 (g) → CO2 (g) + qV
Straightforward proportional calculation gives the amount of heat that would have
been given off if the measured amount of C(gr) had been one gram (qV =
32.76 kJ g−1 ) or if it had been one mole (qV = U 298 = −393.51 kJ mol−1 ). The
latter qV gives the molar energy change U 298 of the system, which is negative be-
cause the system gives off heat to the surroundings. The heat is given off at constant
volume of the closed bomb; hence it is an energy change. In this reaction, the number
of moles of gas used up is the same as the number produced n gas = 0, so the energy
change is the same as the enthalpy change:
H = U + ( pV ) = U + n gas RT
We can write f U 298 (CO2 ) = f H 298 (CO2 ) = −393.5 kJ mol−1 to indicate the
energy or the enthalpy of formation of CO2 for an experiment carried out at 298 K.
The heat of combustion of a gas—for example, hydrogen—can be found using a
flame calorimetric apparatus in which a known amount of gas is burned and the heat
given off is measured by measuring the temperature rise of a suitably positioned bath.
The apparatus is a fancy Bunsen burner heating up a container of water equipped with
a thermometer. It functions at constant pressure, so the heat given out is the enthalpy
decrease of the system:
H2 (g) + 12 O2 (g) → H2 O(l) + q p
which leads to the molar enthalpy of formation of liquid water:
q p = f H 298 (H2 O(l)) = −285.6 kJ mol−1
In the formation reaction, 32 mol of gas are consumed to produce 1.0 mol of liquid
water, which has a negligible volume compared to the gas burned. Energy and enthalpy
are related by H = U + pV ; hence H = U + pV at constant pressure and
P1: OTA/XYZ P2: ABC
58 THERMOCHEMISTRY
H = U + n RT under the ideal gas assumption. For the formation of one mole
of water, n = − 32
f U 298 = f H 298 − n RT = f H 298 + 32 RT

= −285,600 J mol−1 + 32 [8.31(298)] ∼
= −281.9 kJ mol−1
For comparison with theoretical calculations, one often needs to know the
thermodynamic properties of molecules in the gaseous state—for example,
f H 298 (H2 O(g)). This is handled by adding the heat of vaporization of water to
f H 298 (H2 O(l)) to obtain
f H 298 (H2 O(g)) = −285.6 + 44.0 = −241.6 kJ mol−1
However, water vapor is not in its standard state.
4.3 STANDARD STATES
One can burn a diamond, C(dia), in an oxygen bomb calorimeter. When this is done,
the measured enthalpy of formation of CO2 (g) is about 2 kJ mol−1 more negative
than the f H 298 (CO2 (g)) found when C(gr), carbon in the standard state, is burned.
The difference is not in the CO2 (g) produced, but in the crystalline form of diamond,
which is not the standard state for carbon. Since the path from C to CO2 (g) is about
2 kJ mol−1 longer in the diamond combustion, the starting point C(dia) must have
been about 2 kJ mol−1 higher in enthalpy than C(gr). Differences like this lead us
to define the standard state of all elements as the stable form at 1.000 atm pressure.
By the nature of enthalpy and energy (thermodynamic properties), we can set any
arbitrary point to zero as a reference point. Hence we define the enthalpy of formation
of any element in its standard state as zero at all temperatures. This definition
works because elements are not converted from one to another in ordinary chemical
reactions.
Our previous observation that f H 298 (C(dia)) = 0 but is about 2 kJ mol−1 sug-
gests that, given this small enthalpy change, it might be possible to convert common
graphite into the nonstandard state of diamond. Indeed it is. Production of small
industrial diamonds for cutting tools is commercially feasible.
4.4 MOLECULAR ENTHALPIES OF FORMATION
According to the first law of thermodynamics, the enthalpy of formation of a molecule

can be determined even if the formation reaction from its elements cannot be carried
out in an actual laboratory experiment. This is illustrated by the simple example of
carbon monoxide CO(g). We know from painful examples that this poisonous gas is
produced by incomplete combustion of hydrocarbon fuels, but the simple controlled
P1: OTA/XYZ P2: ABC
MOLECULAR ENTHALPIES OF FORMATION 59
-393.5
-283.0
FIGURE 4.1 Combustion of C(gr) and CO(g).
experiment C(gr) + 12 O2 (g) in a limited supply of O2 (g) gives a mixture of products.

It is not a “clean” reaction. The reaction
CO(g) + 12 O2 (g) → CO2 (g)
is clean, however, and gives well-defined q p and qV . We find that the enthalpy change
is r H 298 = −283.0 kJ mol−1 for this reaction (Fig. 4.1). Thus we have two paths
connecting the same initial and final thermodynamic states. By a thermochemical
principle equivalent to the first law of thermodynamics known as Hess’s law, the
enthalpy change over both paths must be the same. The third leg of the triangle
must be
f H 298 (CO(g)) = −393.5 − (−283.0) = −110.5 kJ mol−1
This method of indirect determination of f H 298 is capable of wide extension.

For example, knowing f H 298 (H2 O(l)) and f H 298 (CO2 (g)), one can determine
f H 298 of a hydrocarbon like methane by burning it in a flame calorimeter and
setting up a state diagram that is a little more complicated than the triangle just
described but that works on the same principle. The combustion reaction
CH4 (g) + 2O2 (g) → CO2 (g) + 2H2 O(l) r H 298 = −890 kJ
has the same final state as
2H2 (g) + C(gr) + 2O2 (g) → CO2 (g) + 2H2 O(l) r H 298 = −966 kJ
In Fig. 4.2, the thermodynamic state of the products of combustion of methane,

CO2 (g) + 2H2 O(l) is reproduced by burning 2 mol of hydrogen and 1 mol of C(gr).
The heats of combustion are arranged in an enthalpy diagram so as to make everything
come out even except for one missing leg of the quadrangle, that of methane. The
thermodynamic state of methane is found by difference; it is 76 kJ mol−1 below
that of the elements. Formation of methane from its elements would give off 76 kJ
per mole of methane produced; therefore f H 298 (methane) = −76 kJ mol−1 . (The
tabulated NIST value is −74.9 kJ mol−1 .)
P1: OTA/XYZ P2: ABC
60 THERMOCHEMISTRY
elements
-76
2(286)=572 methane
-890
394
CO 2 + 2H2O
FIGURE 4.2 A Thermochemical cycle for determining f H 298 (methane). Not to scale.
The indirect method shown for methane has been extended to very many hydro-
carbons and other organic compounds. Although many inorganic substances do not
burn, they do react. Inorganic reaction cycles similar to Fig. 4.2 can often be set up
to obtain thermochemical data. A free thermochemical database is maintained by
the National Institutes of Standards and Technology (NIST) of the US government
(.gov). Go to webbook.nist.gov.
4.5 ENTHALPIES OF REACTION
Suppose we know the enthalpies of formation f H 298 of acetylene CH CH, ethene

CH2 CH2 , and ethane CH3 CH3 (226.7, 52.5, and −84.7 kJ mol−1 ) by the combustion
method shown in Fig. 4.1. The enthalpy change r H 298 of the reaction
CH CH(g) + 2H2 (g) → CH3 CH3 (g)
can be found by comparing the level given by f H 298 (CH CH(g)) with that of
f H 298 (CH3 CH3 (g)). (Remember that f H 298 is zero for elemental hydrogen.)
r H 298 = f H 298 (CH3 CH3 (g)) − f H 298 (CH CH(g))

= −84.7 − 226.7 = −311.4 kJ mol−1
This reaction has been carried out with the result r H 298 = −312.1 ± 0.6
kJ mol−1 .
A similar reaction, the partial hydrogenation,
CH CH(g) + H2 (g) → CH2 CH2 (g)
cannot be carried out in the laboratory. Hydrogenation doesn’t stop at CH2 CH2 (g),
but goes on to the fully hydrogenated product CH3 CH3 (g) or gives a mixed product
P1: OTA/XYZ P2: ABC
ENTHALPIES OF REACTION 61
under all known experimental conditions. That doesn’t matter because of Hess’s law.
The enthalpies of formation yield
r H 298 = f H 298 (CH2 CH2 (g)) − f H 298 (CH CH(g))

= 52.5 − 226.7 = −174.4 kJ mol−1
An equivalent way of finding r H 298 of partial hydrogenation is by measuring

the enthalpy change −136.3 kJ mol−1 for the reaction
CH2 CH2 (g) + H2 → CH3 CH3 (g)
When this is done, one has a first law triangle like that of Fig. 4.1, with only r H 298
of partial hydrogenation yet to be determined. It is
r H 298 (partial) = −312.1 − (−136.3) = −175.8 kJ mol−1
as compared to the previous value of −174.4 kJ mol−1 .

One can also find the enthalpy of hydrogenation of ethene:
r H 298 = f H 298 (CH3 CH3 (g)) − f H 298 (CH2 CH2 (g))

= −84.7 − 52.5 = −137.2 kJ mol−1
This is a complete reaction to a well-defined final state. It has been carried out in the
laboratory. The experimental value is −136.3 ± 0.2 kJ mol−1 .
Although, for simplicity, we shall use hydrocarbons to illustrate the principles
of thermochemistry, there is no such restriction in practice. For example, Pitzer
et al. (1961) give the standard enthalpies of formation f H ◦ (KCl) = −435.9 and
f H ◦ (KClO3 ) = −391.2 kJ mol−1 . (The superscript ◦ indicates the standard state in
this notation; please do not confuse it with a superscripted zero of temperature 0.)
These values lead to the enthalpy of reaction r H ◦ necessary to convert solid KCl(s)
to solid KClO3 (s):
KCl(s) + 32 O2 (g) → KClO3 (s)

r H ◦ = −391.2 − (−435.9) = 44.7 kJ mol−1
Including the notation (s) to indicate the solid state is a precaution rather than a
necessity because in the standard states, both KCl and KClO3 are solids.
Even from these few simple examples, it should be clear that with a sufficient
database of enthalpies of formation, it is possible to calculate the enthalpies of an
almost limitless array of chemical reactions of industrial, medical, and pharmaceutical
importance. This is the reason why so much effort has gone into populating the
P1: OTA/XYZ P2: ABC
62 THERMOCHEMISTRY
database with a variety of entries (webbook.nist.gov) and why the most meticulous
care is exercised to be sure that the data entered are accurate.
All of the previous examples are consistent with the equations

r U 298 = f U 298 (products) − f U 298 (reactants)
and

r H 298 = f H 298 (products) − f H 298 (reactants)
where it is to be understood that products and reactants are multiplied by their

stoichiometric coefficients and that they are in their standard states. We shall soon
see that all thermodynamic properties follow analogous equations, which is why
thermodynamics is so useful.
Although combustion thermochemistry on C(gr), CH4 (methane), and so on, is the
prime source of f H 298 data, other reactions can contribute as well. For example,
suppose we know that f H 298 (ethane) = −83.8 ± 0.4 kJ mol−1 but we don’t know
f H 298 (ethene). Measurement of the enthalpy of the hydrogenation reaction ethene
→ ethane gives −136.3 ± 0.2 kJ mol−1 ; hence f H 298 (ethene) must be higher in
enthalpy than f H 298 (ethane) by just that amount. By this reasoning,
f H 298 (ethene) = −83.8 ± 0.4 + 136.3 ± 0.2 = 52.5 ± 0.4 kJ mol−1
The tabulated value is 52.5 ± 0.4 kJ mol−1 .

Molar enthalpies of physical processes like phase changes (vaporization and melt-
ing), solution of a solute in a solvent, mixing of miscible solvents, and dilution are
treated by slight modifications of the methods shown, always in accord with the first
law of thermodynamics, which has never been violated in a controlled, reproducible
experiment.
4.6 GROUP ADDITIVITY
Upon scanning a series of heats of combustion c H 298 , one notices a regular increase
with molecular weight. For example, for the alkanes in their standard states (g), we
have the following:
Methane Ethane Propane n-Butane
c H 298 (g) −890.8 −1560.7 −2219.2 −2877.6

Difference −669.9 −658.5 −658.4
P1: OTA/XYZ P2: ABC
GROUP ADDITIVITY 63
If we were asked to predict c H 298 (n-pentane(g)), a reasonable answer would be to

add another negative change just like the last two:
c H 298 (n-pentane(g)) = −2877.6 − 658.4 = −3536 kj mol−1 .
The experimental value is 3535.4 ± 1.0.

Even simpler, we can note that, given the reference point c H 298 (ethane(g)) =
−1560.7 kJ mol−1 , the heats of combustion of n-alkanes (except for methane) obey a
linear function of the number of additional hydrogen atoms over those in the reference
compound with slope equal to –658.5/2 = −329.2 kJ mol−1 . These CH2 hydrogen
atoms are called secondary hydrogens as distinct from CH3 , hydrogens, which are
primary, and isolated C H atoms, which are tertiary. Now any c H 298 (n-alkane(g))
can be found by counting secondary hydrogen atoms and adding the count times
−329.2 to the base value of −1560.7.
Because the enthalpies of formation f H 298 of n-alkanes(g) are proportional
to c H 298 , one can estimate f H 298 values in the same way that we used in the
combustion case. Simply count hydrogen atoms and add to a base value for ethane:
Ethane Propane n-Butane
f H 298 (g) −84.0 −104.7 −125.6

Difference −20.7 −20.9
Counting secondary hydrogens for n-octane and multiplying by −20.8/2 = −10.4

gives 12(−10.4) = −124.8. Adding the base value, −84.0 gives −208.8 kJ mol−1 .
The experimental value is −208.4 ± 0.8 kJ mol−1 .
Zavitsas et al. (2008) have extended this method to cover primary, secondary, and
tertiary hydrogens by the equation

f H ◦ = −14.0n p + (−10.4n s ) + (−6.65n t ) = ci n i
where n p , n s , and n t are the numbers of primary, secondary, and tertiary hydrogen
atoms in any alkane or cycloalkane. By this system, the enthalpy of formation of
4-methylheptane, an isomer of n-octane, is
H
|
CH3 CH2 CH2 CCH2 CH2 CH3
|
CH3
f H ◦ = −14.0n p + (−10.4n s ) + (−6.65n t )

f H 298 (4−methylheptane(g)) = −14.0(9) + (−10.4(8)) + −6.65(1)
= −215.8 kJ mol−1
The experimental value is −212.1.

P1: OTA/XYZ P2: ABC
64 THERMOCHEMISTRY
4.7 f H298 (g) FROM CLASSICAL MECHANICS
One of the drawbacks of the hydrogen-atom counting method is that account is

not taken of classical mechanical properties of molecules such as mechanical strain
energy induced in distorted or crowded molecules. Other mechanical features not
accounted for are those due to the quantum mechanical influences on electron
probability densities which, in turn, influence both molecular energy and molecular
structure.
The influence of molecular strain can be seen in the discrepancy between the
hydrogen-atom counting estimate of f H 298 (cyclopentane(g)) = −24.9 kJ mol−1
as contrasted to the experimental value of −16.3 kJ mol−1 . The experimental value
is higher (less negative) than the estimate. Evidently, crowding 10 hydrogen atoms
into the small space afforded by a 5-membered ring increases the interatomic in-
terference energy that contributes to f H 298 . N. L. Allinger at the University of
Georgia has developed a series of molecular mechanics programs called MM1 to
MM4 that incorporate energies due to bond bending, bond stretching, ring distortion,
and so on, into the calculation of molecular structure and f H 298 . The method as
implemented on a computer is extremely fast and is therefore well-suited to large
molecules.
4.8 THE SCHRÖDINGER EQUATION
In 1926 Erwin Schrödinger published an equation that gives correct solutions for the
energy levels of the hydrogen atom. Shortly afterward, Heitler and London showed
that the Schrödinger equation, as it has come to be called, predicts the existence of a
chemical bond between H and H and that it gives an approximate strength of the H H
bond. A powerful refinement and extension of this molecular orbital calculation called
GAMESS is available as a site license at no cost for your microcomputer (academic
or similar affiliation must be specified):
http://www.msg.ameslab.gov/GAMESS/GAMESS.html
We shall use the GAMESS program to determine the bond energies of a number of
molecules starting with the simplest case, that of the hydrogen molecule when it is
formed from two hydrogen atoms:
2H· → H H
The GAMESS output for calculation of the total energy of H· is E = −0.4998 E h ,

where E h is the hartree, a unit of energy. A GAMESS output for the molecule H H is
E = −1.1630349978E h
P1: OTA/XYZ P2: ABC
VARIATION OF H WITH T 65
Hence the bond energy is the energy of the final state (the molecule) minus the
energy of the initial state (the two atoms):
E(H H) − 2E(H·)
= −1.1630 − (2 ×− 0.4998) = − 0.1630E h = −102.3 kcal mol−1 = −428 kJ mol−1
as compared to the experimental value of −431 kJ mol−1 . The bond energy is

negative because the molecule is more stable than the isolated atoms (energy
goes downhill). Thus our first molecular orbital calculation comes to within about
1% of the experimental value. The conversions 1.0 E h = 627.51 kcal mol−1 and
1.0 kcal mol−1 ≡ 4.184 kJ mol−1 have been used. This calculation can be carried out
on much more complicated molecules, ions, and free radicals, some of which will be
described in later chapters The advent of these powerful computer programs and the
hardware to run them has made it possible to study reactions that are not accessible
by experimental means. Computational thermochemistry is an active research area at
present.
4.9 VARIATION OF H WITH T
The definition of the heat capacity at constant pressure C p = (∂ H /∂ T ) P leads to the

infinitesimal enthalpy change with temperature of a pure substance d H = C P dT .
Over a reasonably short temperature interval T, the equation H = C P T is
approximately true. The heat capacity of a mixture is the sum of the molar heat
capacities of its components multiplied by the number of moles of each component
present. When a chemical reaction takes place, the number of moles of the reactants
decreases and the number of moles of products appears with different heat capacities.
The difference in heat capacities between the reactant state and the product state is

C P = C P (products) − C P (reactants)
Applying the definition of C p to all of the component species of a chemical reaction,

we get

∂H
C p =
∂T P
Selecting the hydrogenation of ethene to ethane for simplicity, experimen-

tal values (webbook.nist.gov) are C p = 42.90 J K−1 mol−1 for ethene and C p =
52.49 J K−1 mol−1 for ethane; and taking an experimental value (Atkins, 1994) of
C p = 28.87 J K−1 mol−1 for hydrogen, we obtain
C p = 52.49 − 42.90 − 28.87 = −19.28 J K−1 mol−1

P1: OTA/XYZ P2: ABC
66 THERMOCHEMISTRY
The change in r H for relatively small changes in T is C P T . We can test this

result against experimental values for the simple hydrogenation of ethene (Kisti-
akowsky et al., 1935). Using the computed result for C P T over the temperature
range 355–298, the range between the temperature at which his experiments were
carried out (355 K) and standard temperature, we get
hyd H 355 = −19.28 [− (355 − 298)] = 1.099 kJ mol−1
for the change in hyd H 355 , from 355 K, the temperature at which the hydrogenation
was actually carried out to room temperature 298 K. This small temperature correction
gives
hyd H 298 = −137.44 + 1.10 = −136.34 kJ mol−1
for the enthalpy of hydrogenation of ethene at 298 K. This value is to be compared to

the result of a series of experiments carried out at 298 K by a different experimental
method which led to hyd H 298 = −136.29 ± 0.21 kJ mol−1 . The computed heat
capacities are probably reliable over a temperature range of ±50 K or so. Furthermore,
the enthalpy of hydrogenation itself is insensitive to temperature, so we may take
experimental determinations of hyd H carried out under normal laboratory conditions
◦
as essentially the same as the standard state hyd H298 .
When calculations are carried out over larger temperature ranges, as they often are
in industrial applications, a polynomial approximation to the heat capacity is used.
C p = α + βT + γ T 2 + · · ·
This enables one to determine C p for the reactants and products of a chemical reaction
at some new temperature other than 298 K, thereby enabling one to determine the
new change in heat capacity C p for the reaction:

C P = C P (products) − C P (reactants)
Variation in theH of physical and chemical processes with variation in pressure

can be calculated from equations of state or by acquisition of experimental data and
curve fitting. Many reactions are less sensitive to pressure change than to temperature
change over comparable ranges. Metiu (2006) has treated both temperature and
pressure variation in the industrial production of ammonia from its elements.
4.10 DIFFERENTIAL SCANNING CALORIMETRY
When moderate amounts of heat are supplied to a solution of simple salts in water,
we expect a smooth heating curve between, say 290 and 320 K, like the lower curve
in Fig. 4.3. The heat capacity of water is nearly constant over this temperature range,
P1: OTA/XYZ P2: ABC
DIFFERENTIAL SCANNING CALORIMETRY 67
and it will be little affected by small amounts of dissolved salts. Electrical circuitry
exists that permits one to supply heat to a dilute solution in an adiabatic (insulated)
calorimeter in very small pulses which may be regarded as infinitesimals dq. We nor-
mally carry out the experiment at constant pressure, so the definition of heat capacity
at constant pressure C p = dq p /dT is satisfied. The gradual temperature rise over
many small pulses can be followed by means of a thermistor circuit or its equivalent.
If, instead of a dilute solution of simple salts, the calorimeter contains a solute
that is capable of undergoing a thermal reaction, which is a reaction brought about
by heat, the heating curve is more complicated. Thermal reactions are important
in many areas, especially in biochemistry. Proteins undergo heat denaturation. Heat
denaturation involves unfolding of the native protein and requires breaking of some
or many of the bonds holding it in its native structure. Heat denaturation may be
quite specific as to the temperature at which it occurs, and it may bring about subtle
changes in the protein, like changes in physiological activity, or it may bring about
gross changes in the form of the protein as in the cooking of an egg.
Because heat denaturation involves breaking of internal bonds in the protein, it
requires an enthalpy input at constant pressure. The reaction is endoenthalpic. A
dilute solution of salt and protein takes more heat to bring about a small temperature
change than would the solution without the protein. The difference is observed only
at or near the temperature of denaturation. Thus we have a normal temperature rise
until denaturation begins, after which the heat capacity of the solution is abnormally
large until we achieve complete thermal denaturation whereupon the temperature rise
drops back to the normal baseline of a salt solution. Plotting C p as a function of T,
we see a peak at the denaturation temperature. This is the upper line in Fig. 4.3. It
is a simple matter to interface a computer to the scanning calorimeter output and to
integrate under the experimental curve:
Tf
den H = C p dT
Ti
Cp
T
FIGURE 4.3 Schematic diagram of the thermal denaturation of a water-soluble protein. The
straight line is the baseline of salt solution without protein. The peak is due to endoenthalpic
denaturation of the protein.
P1: OTA/XYZ P2: ABC
68 THERMOCHEMISTRY
The heat capacity curve of the simple salt solution (the baseline) is subtracted from
the experimental result. There may be multiple peaks if there is more than one protein
in the test solution or if the protein is capable of unfolding in sequential steps.
Example 4.1 Oxygen Bomb Calorimetry

Exactly 0.5000 g of benzoic acid C6 H5 COOH were burned under oxygen. The
combustion produced a temperature rise of 1.236 K. The same calorimetric setup was
used to burn 0.3000 g of naphthalene (C10 H8 ) and the resulting temperature rise was
1.128 K. The heat of combustion of benzoic acid is qV = c U 298 = –3227 kJ mol−1
(exothermic). What is the heat of combustion qV = c U 298 of naphthalene?
Solution 4.1 The corresponding molar masses are: benzoic acid, 122.12 g mol−1 ;
and naphthalene, 128.19 g mol−1 . The temperature rise for each combustion was: ben-
zoic acid, 1.236/0.5000 = 2.472 K g−1 ; and naphthalene, 1.128/0.3000 = 3.760 K g−1 .
Multiplying by the molar masses in each case, one obtains 301.9 K mol−1 for benzoic
acid and 482.0 K mol−1 for naphthalene. This gives us the ratio
301.9 −3227
=
482.0 x
482.0
x = −3227 = −5152 kJ mol−1
301.9
qV = c U 298 = −5152 kJ mol−1
Notice that the units cancel on the left; thus x has the units of kJ mol−1 , not K mol−1 .
The handbook value is c U = −5156 kJ mol−1 .
Problem 4.1
A resistor of precisely 1 ohm is immersed in a liter (1 dm3 ) of water in a perfectly
insulated container. Suppose that precisely 1 ampere flows through the resistor for
precisely 1 second. What is the temperature rise of the water?
Problem 4.2
Exactly one gram of a solid substance is burned in a bomb calorimeter. The bomb
absorbs as much heat as 300 g of water would absorb. (Its water equivalent is 300 g.)
The bomb was immersed in 1700 g of water in an insulated can. During combustion
of the sample, the temperature went from 24.0◦ C to 26.35◦ C. What is the heat of
combustion per gram of the sample? What is the molar enthalpy of combustion if the
molar mass of the substance is 60.0 g mol−1 and 2 mol of gas are formed in excess
of the O2 burned?
P1: OTA/XYZ P2: ABC
Problem 4.3
The enthalpy of formation of liquid acetic acid CH3 COOH(l) is f H ◦ =
−484.5 kJ mol−1 . What is c H ?
Problem 4.4
The enthalpy of combustion of solid α-d-glucose, C6 H12 O6 (s) is −2808 kJ mol−1 .
What is its enthalpy of formation?
Problem 4.5
Estimate c H 298 (n-octane(g)) of n-octane by the hydrogen-atom counting method
for alkanes.
Problem 4.6
Find f H 298 (2,4-dimethylpentane(g)) by the hydrogen atom counting method. What
is the enthalpy of isomerization of n-heptane(g) to 2,4-dimethylpentane(g) according
to this method? Compare your answer with the experimental result of −14.6 ±
1.7 kJ mol−1 .
Problem 4.7
The input file for a Gaussian C quantum mechanical calculation will be discussed
in later chapters. Briefly, it consists of a few lines of instructions to the computer

concerning memory requirements, the number of processors to be used, and the
Gaussian procedure to be used, followed by an approximate geometry of the molecule.
In simple cases, the input geometry can be merely a guess based on what we learned
in general chemistry. The machine specifications will vary from one installation to
another. Our input file for the water molecule is
%mem=1800Mw
%nproc=1
# g3
water
0 1
H -1.012237 0.210253 0.097259
O -0.260862 0.786229 0.119544
H 0.489699 0.209212 0.142294
Adapt this input file for your system and run the water molecule. What is the optimized
geometry in the form of Cartesian coordinates (like the input file)? What is the O H
bond length? What is the H O H bond angle? What energy is given for water?
P1: OTA/XYZ P2: ABC
70 THERMOCHEMISTRY
Problem 4.8
Plot the heat capacity of ethylene from the following data set:
Temperature (K) Heat capacity, C p (J K−1 mol−1 )
300.0000 43.1000
400.0000 53.0000
500.0000 62.5000
600.0000 70.7000
700.0000 77.7000
800.0000 83.9000
900.0000 89.2000
1000.0000 93.9000
Problem 4.9
Suppose that the heat capacities C p for N2 , H2 , and NH3 in the standard state are
constant with temperature change (they aren’t) at 29.1, 28.8, and 35.6 J K−1 mol−1 .
Suppose further that r H ◦ for the reaction
N2 + 3H2 → 2NH3
is −92.2 kJ mol−1 (of N2 consumed) at 298 K. What is C p for the reaction? What
is r H ◦ at 398 K? What is the (hypothetical) r H ◦ at 0 K?
P1: OTA/XYZ P2: ABC
5
ENTROPY AND THE SECOND LAW
The second law of thermodynamics and the concept of entropy are firmly based on
Sadi Carnot’s somewhat abstract demonstration that the work done by drawing heat
from a hot reservoir and expelling it to a cold reservoir is independent of the nature of
the engine that carries out the work (Fig. 5.1). An especially readable description of
the progression of Carnot’s abstract reflections on the efficiency of steam engines to
the form of the entropy function we use today was given by Kondipudi and Prigogine
(1998). We shall not follow the historical development of this idea, interesting though
it is; rather we shall jump right to Clausius’s definition of the entropy, Ludwig
Boltzmann’s statistical interpretation, and the influence of the second law on physical
phenomena and chemical reactions.
5.1 ENTROPY

In 1865, Rudolf Clausius showed that the cyclic integral dq/T is zero for an
abstract Carnot engine operating reversibly around a cyclic path:

dq
=0
T
Thus the integrand, which Clausius named entropy and gave the symbol dS, is
a thermodynamic function. This definition is one statement of the second law of

Copyright
71
P1: OTA/XYZ P2: ABC
72 ENTROPY AND THE SECOND LAW
WORK
HOT COLD
FIGURE 5.1 An Engine.
thermodynamics, which, like the first law, can be stated in many ways. It follows that
b b
S = a dS = a dq/T is the entropy change of a system carried reversibly over an
arbitrary path from a to b and is independent of the path. This powerful definition
constitutes the second of the two great pillars of thermodynamics. If we can devise
a way of calculating S for a reversible chemical reaction, we shall know it for all
chemical reactions having the same initial and final states (reactants and products)
because of path independence.
Clausius expanded upon the concept of entropy by writing the complete statement
as
dq
dS ≥
T
which takes both reversible and irreversible changes into account. The irreversible
change dSirr > dqirr /T is the real case, a change that takes place in finite time.
If we attempt to take an engine around an irreversible cycle to reproduce its initial
state, we shall fall short. We have received a certain amount of work from the engine,
but when it comes to the payback (in heat) we see the following with regard to the
second law:
dq
dS ≥ implies that dqirr < TdSirr
T
The system will not be returned to its original state, violating the principal stipulation
that the system operate around a cyclic path. We shall have to take some heat from
the hot reservoir in Fig. 5.1 to complete the cycle and bring the entropy back to its
initial value. Where does the extra heat over and above the reversible heat eventually
end up? It can go only one place. Since it hasn’t done any work, it must have passed
through the engine and gone directly to the low-temperature reservoir. The efficiency,
work out relative to heat in, of a real engine operating irreversibly is less than 1.0
because some heat is doing work and some is not. The important concept is that, of
the heat taken from the hot reservoir, not all of it can do work. Some heat must pass
through the engine from the hot reservoir directly to the cold reservoir doing nothing
P1: OTA/XYZ P2: ABC
ENTROPY 73
but restoring the system entropy to its original state. Because of this necessity, some
of the heat drawn from the hot reservoir in an irreversible cycle is unavailable to do
work.
5.1.1 Heat Death and Time’s Arrow

Because there is always heat transferred to the surroundings in an irreversible cycle,
the entropy of the system plus the surroundings always increases. If we take the
universe as the surroundings, then since all real processes are irreversible we have
a consequence of the second law: The entropy of the universe tends to a maximum.
When the universe has reached its maximum entropy, no more irreversible change
will be possible. The driving force of change will be gone. This is called the heat
death of the universe. In case you are worried, it is calculated to be in the far distant
future.
The entropy of the universe must be greater after an irreversible change has
occurred than it was before the change, so we have a thermodynamic definition of
the direction of time (which the first law doesn’t give). Time must go from before
the change to after the change; it cannot go in the reverse direction.1 Entropy is
sometimes called “time’s arrow.”
5.1.2 The Reaction Coordinate

Prigogine has defined a reaction coordinate ξ which progresses as a chemical reaction
takes place. Starting with pure reactant A, the reaction coordinate increases as product
B is produced:
A→B
At some point, the time derivative of ξ becomes zero and the reaction stops insofar as
macroscopic concentration measurements are concerned.2 When the time derivative
of the reaction coordinate is zero, the system consisting of n A + n B is at equilibrium.
Because there are no macroscopic concentration changes, the ratio of the mole num-
bers of reactant and product n B /n A is constant. It is called the equilibrium constant
K eq :

∂ξ
=0
∂t T, p
K eq = n B /n A
It would be appealing to think that the reaction coordinate is determined solely by

the energy U or enthalpy H of the system flowing from a high level to a low level,
1 Classicalthermodynamics does not include QED.
2 There are microscopic exchanges between species A and B, but they are equal and opposite on average
so they do not bring about measurable concentration changes.
P1: OTA/XYZ P2: ABC
but this is not the case. There is more to think about in a chemical or physical change
than just minimization of U or H. There is the question of order and disorder of the
reactant state and the product state.
5.1.3 Disorder
When we look at a chemical reaction
A→B
or an analogous physical change
A(l) → A(g)
we must look at the driving force that produces the change. Part of that force comes
from the tendency to seek a minimum (water flows downhill), but part of it comes
from the universal tendency of thermodynamic systems to seek maximum disorder.
A familiar example is vaporization of a liquid such as water.
The liquid state, though not perfectly ordered, is held together by strong inter-
molecular forces. These are the very forces that we say are nonexistent or negligible
in the vapor state. The entropy change for many liquids is about 88 J K−1 mol−1 , a
rule known as Trouton’s rule, that has been verified many times over for liquids as
diverse as liquid Cl2 , HCl, chloroform, and the n-alkanes. Liquids that deviate from
this rule do so, not because of any failure of the entropy concept for vaporization, but
because of abnormal forces in the liquid state. An example is water, which deviates
a little due to hydrogen bonding, and hydrogen fluoride HF, which deviates a lot.
5.2 ENTROPY CHANGES
In general, the entropy change for any change of state, including melting and change in
crystalline form in the solid state, is given by the enthalpy change and the temperature
appropriate to the change or transition considered:
Htrans
S =
Ttrans
5.2.1 Heating
The entropy of heating of an ideal gas is positive because thermal agitation makes
the high-temperature state more disordered than the low-temperature state. From the
expression for molar heat capacity, one has

∂H
Cp =
∂T p
P1: OTA/XYZ P2: ABC
ENTROPY CHANGES 75
Hence
dqrev dH dT
dSrev ≡ = = Cp
T T T
Taking the integrals over the interval T1 to T2 , we obtain

T2 T2
dT dT T2
S = Cp = Cp = C p ln
T1 T T1 T T1
for the molar entropy change of heating at constant pressure. There is an analogous
equation for the energy change of heating at constant volume.
5.2.2 Expansion
For expansion at constant temperature, U is constant. The first law dU = dq + dw = 0
gives us dq = dw and the second law gives us
dqrev dw pdV
dSrev = = =
T T T
for reversible pressure–volume work. Taking the gas to be ideal for simplicity, we
obtain
p dV RT d V dV
dSrev = = V =R
T T V
Integrating between limits as before, we obtain

V2
dV V2
Srev = R = R ln
V1 V V1
for the reversible expansion of one mole of an ideal gas from V1 to V2 . The right-hand
side of the equation should be multiplied by n for expansion of n moles of gas. If the
gas is not ideal, a real gas equation can be substituted for the ideal gas law in these
equations. Thus, the mathematical complexity will be increased, but the principle is
the same.
5.2.3 Heating and Expansion

Because dS is an exact differential and S = f (T, p), we have

∂S ∂S
dS = dT + dp
∂T p ∂p T
P1: OTA/XYZ P2: ABC
We know from the second law that at constant pressure, we obtain
dq p Cp
dS = =
T T
so

Cp ∂S
dS = dT + dp
T ∂p T
By the Euler reciprocity relation, for exact differentials du written in differential form
du = M(x, y) + N (x, y)
we have the equality
∂ M(x, y) N (x, y)
=
dy dx
In the case of the Gibbs thermodynamic function (next chapter) µ = f (S, p), we
have
dµ = −S dT + V dp
so

∂S ∂V
− =
∂p T ∂T p
which leads to

Cp ∂S Cp ∂V
dS = dT + dp = dT − dp
T ∂p T T ∂T p
Both of these coefficients C p /T and (∂ V /∂ T ) p can be measured, so the infinitesimal

dS can be found at any T and V. The finite change S is

T2
Cp p2
∂V
S = dT − dp
T1 T p1 ∂T p
Starting with the Helmholtz free energy in place of the Gibs function, a comparable
derivation yields

T2
Cp V2
∂p
S = dT − dV
T1 T V1 ∂T V
P1: OTA/XYZ P2: ABC
SPONTANEOUS PROCESSES 77
These derivations are given in more detail in Metiu (2006) along with results cal-
culated from accurate equations of state for real gases and comparisons to National
Institutes of Standards and Technology tabulations (webbook.nist.gov).
5.3 SPONTANEOUS PROCESSES
5.3.1 Mixing
Consider two chambers of the same volume, one containing 0.500 mol of ideal gas
A and the other containing 0.500 mol of ideal gas B. The gases are at the same
temperature and 1.0 bar pressure, and the chambers are connected by a valve. When
the valve is opened, the gases mix spontaneously just as the smell of perfume gradually
permeates all regions of a closed room. The diffusion process is like an expansion
for the gases considered individually because the volume in which the gas molecules
move is doubled relative to what it was before the valve was opened. For gas A we
can say
V2 2
SA = 0.500 R ln = 0.500 R ln = 2.88 J
V1 1
The equivalent calculation for gas B gives the same result, so the total entropy change
is the sum
SA + SA = 2(2.88) = 5.76 J
If the volumes of the chambers are changed arbitrarily so that they are not equal
and if quantities of gas are taken that are not 0.500 mol, nor are they equal, the entropy
increase for gases A and B is different from what we have just calculated, but it is
always an increase because each gas sees a larger volume after the mixing process
than it did before. Mixing is always spontaneous and the entropy of mixing is always
positive because, at constant p, V2 is always larger than V1 from the point of view of
each gas.
5.3.2 Heat Transfer

Consider now two bricks in contact with one another. One is a hot brick and the other
is a cold brick. We know, from millennia of experience, that after sufficient time we
will have two warm bricks and that we never observe the reverse process, two warm
bricks spontaneously undergoing a transformation that produces a hot brick and a
cold brick. This is a crude statement of the second law. In fact an engine is merely a
device placed between a hot reservoir and a cold reservoir to siphon off some of the
heat flow and use it to do work.
Spontaneous flow of heat from hot to cold cannot be explained by the first law
because for each joule of heat lost by the hot brick there is exactly one joule gained
P1: OTA/XYZ P2: ABC
by the cold one. The process involves no change in energy or enthalpy; hence the first
law tells us nothing about it. To look at the situation more quantitatively, suppose that
the bricks are of equal size and have the same heat capacity C. Suppose further that
one brick is at 400 K and the other is at 200 K. After sufficient time, both will be at
300 K, assuming no heat is lost to the surroundings.
T2 300
S = C ln = C ln = −0.288C
T1 400
for the hot brick, but
T2 300
S = C ln = C ln = 0.405C
T1 200
for the cold brick. The result that the positive entropy change is greater than the
negative change is independent of the initial temperatures, heat capacities, sizes of
the bricks, and so on. Entropy for the spontaneous process always increases for
spontaneous heat transfer. The only way to get away from the inequalities of S for
hot and cold bricks would be to make T1 = T2 but then no heat would flow.
5.3.3 Chemical Reactions

The order or disorder within a system undergoing chemical reaction changes, some-
times dramatically. For example, combination of equal volumes of H2 (g) and O2 (g)
produces a negligible volume of H2 O(l) plus 1/2 volume of O2 (g) left over after the
hydrogen is all used up. The volume of the system goes from 2 to 1/2, so we expect
an entropy change on that basis alone. The initial volume is now larger than the
final volume at constant pressure, so, in contrast to the expansion case, the entropy
change of this reaction is negative; the final state is more ordered than the initial
state. The negative r S is in opposition to the direction of the chemical reaction. The
first and second laws operating simultaneously on the system through r H − T r S
give a term −T r S 3 which is positive but which is smaller than r H . Hence r S
is negative for this spontaneous (sometimes explosive) reaction. This is an example
of the very common tendency of reactions to go in the direction of a spontaneous
creation of order from a disordered system, provided that there is enough enthalpy
decrease to drive the entropy change “backwards.” Some creationists believe that a
spontaneous ordering process is impossible, but they are wrong.
5.4 THE THIRD LAW
The third law of thermodynamics states that the entropy, unlike the energy and
enthalpy, has a natural zero point. The entropy of a perfect crystal is zero at 0 K.
3 The product of two negative numbers.

P1: OTA/XYZ P2: ABC
THE THIRD LAW 79
Because of the third law, it is possible to obtain a standard molar entropy (often
called the “absolute” entropy) of any pure substance at any temperature. The task is
simple but not easy. One must determine the molar heat capacity at constant pressure
C p for the crystal at many temperatures until it undergoes the first phase transition.
By the methods shown in Section 5.2.1, the integral taken down to low T
T1
Cp
S= dT
0 T
gives the standard entropy at the transition temperature T1 . Since C p is a molar

heat capacity, S is an standard molar entropy. Normally one wants the entropy at
some higher temperature, say 298 K, and often the phase transition takes place at a
temperature T1 lower than 298 K. Therefore we must add the entropy contribution
from the phase transition to the value of S that we already have to obtain the standard
molar entropy after the phase change at T1 . The new phase then is heated over a
temperature interval T1 → T2 where the new temperature may be 298 K or may be
the temperature of a new phase change. Melting and vaporization are handled in the
same way as crystalline phase changes. Eventually, over a few or many phase changes,
one arrives at the desired temperature. The standard entropy is the summation of all
the contributions along the way:
T1 Htrans Thigher
Cp Cp
S= dT + + dT
0 T Ttrans Tlower T
5.4.1 Chemical Reactions (Again)

The change in entropy of a chemical reaction can be determined by carrying out a
determination of the standard molar entropies of all of the reactants and all of the
products as just described and taking the sum

Sr = S(products) − S(reactants)
The entropy of ordering or disordering that occurs when, for example, the product
state is in the gaseous phase and the reactants are in a condensed phase (liquid
or solid)
is included in this sum because terms like Hvap /Tb are included in S(products)
but not in S(reactants). It would be attractive to adopt the simplistic attitude
that all spontaneous chemical and physical reactions produce an entropy increase
for the reacting system, but, once again, things are more complicated than that. A
spontaneous change is driven both by the tendency of a system to reduce its energy
and enthalpy and by the tendency of a system to increase its disorder. A composite
function is needed that includes both the enthalpy and entropy, and this is the function
found and described in mathematical detail by the great American thermodynamicist
J. Willard Gibbs. The composite function the Gibbs free energy or, more simply, the
Gibbs function G = H − T S now bears his name. There is a comparable function
P1: OTA/XYZ P2: ABC
involving the energy and the entropy used more by engineers than by chemists; this
is called the Helmholtz free energy, A = E − T S.
Example 5.1 The Standard Entropy of Silver

The C p /T vs. T data set for silver from 15 to 300 K is
T Cp Cp /T ln T
15.0000 0.6700 2.7100 0.0447

30.0000 4.7700 3.4000 0.1590
50.0000 11.6500 3.9100 0.2330
70.0000 16.3300 4.2500 0.2333
90.0000 19.1300 4.5000 0.2126
110.0000 20.9600 4.7000 0.1905
130.0000 22.1300 4.8700 0.1702
150.0000 22.9700 5.0100 0.1531
170.0000 23.6100 5.1400 0.1389
190.0000 24.0900 5.2500 0.1268
210.0000 24.4200 5.3500 0.1163
230.0000 24.7300 5.4400 0.1075
250.0000 25.0300 5.5200 0.1001
270.0000 25.3100 5.6000 0.0937
290.0000 25.4400 5.6700 0.0877
300.0000 25.5000 5.7000 0.0850
The C p /T vs. T curve for silver is given as Fig. 5.2. Select from packaged software
(for example, SigmaPlot C ) or write a short program of your own that will enable you
to integrate the data set for silver to find the standard entropy S at 300 K. The problem
is simplified by the lack of phase transitions in solid Ag over the temperature range,
including the melting and boiling points, which are well above 298 K. The CRC
Handbook of Chemistry and Physics, 2008–2009 (89th ed.) value for the standard
−1
S 298
Ag at 298 is 42.67 J K mol−1 .
Solution 5.1 SigmaPlot C contains a macro that carries out integration under curves
that are displayed as a smooth function. First plot your function, then execute
Tools → macro → run → compute. Be sure to designate your plot be-
low the macro. The SigmaPlot output for this integration is 42.2076 over the interval
from 15 K to 298 K.
The result is pretty close to the standard entropy in the handbook, but it lacks
a contribution below 15 K. This problem is usually handled by the Debye method
(Problem 5.7), which assumes a third-power equation leading to
C p = AT 3 = 0.67 J K−1 mol−1

P1: OTA/XYZ P2: ABC
Cp / T vs. T Silver(s)
0.25
0.20
Cp/ T, J mol–1
0.15
0.10
0.05
0.00
0 50 100 150 200 250 300 350
T, K
FIGURE 5.2 C p /T vs. T for metallic silver Ag(s). There are no phase transitions for solid
Ag over this temperature range.
at 15 K and leads to
Cp 0.67
S015 = = = 0.22 J K−1 mol−1
3 3
This small addition yields S0298 = 42.43 J K−1 mol−1 , which is within 0.6% of the
handbook value.
Problem 5.1
Hexa-1,3,5-triene has a boiling point of 355 K under atmospheric pressure. Estimate
the enthalpy of vaporization of hexa-1,3,5-triene.
Problem 5.2
(a) What is the entropy change brought about by heating 2.5 mol of helium from
300 to 400 K at constant volume?
(b) What is the entropy change brought about if the same heating process takes
place at constant pressure?
P1: OTA/XYZ P2: ABC
Problem 5.3
What is the entropy of isothermal mixing of 1 mol of helium with 1 mol of argon if
the two gases start out in separate chambers, each at 1 bar pressure, and they produce
2 mol of mixed gases also at 1 bar pressure?
Problem 5.4
Industrial production of ammonia NH3 is carried out by combination of the elements
at about T = 650 K and p = 400 bar (Metiu, 2006):
3H2 + N2 → 2NH3
What is the enthalpy change for this reaction at this temperature and pressure? For
simplicity, assume ideal behavior of all three gases.
Problem 5.5
What is the enthalpy change for the pressure change from 1.0 bar to 40 MPa?
Problem 5.6
Sulfur dioxide has a heat of fusion of 7.41 kJ mol−1 at its melting point of 200 K.
Find the entropy change for the melting process
SO2 (s) → SO2 (l)
How does this compare with the entropy change for melting ice which has
Hfusion = 333.6 J g−1 (CRC Handbook of Chemistry and Physics, 2008–2009,
89th ed.)
Problem 5.7
On theoretical grounds, Peter Debye proposed what is known as the Debye third-
power law for the entropy of perfectly crystalline solids near absolute zero K.
C p = AT 3 , T < 15 K
Solid chlorine Cl2 (s) has a heat capacity of C p = 3.72 J K−1 mol−1 . What is the
entropy of Cl2 (s) at T = 15 K?
P1: OTA/XYZ P2: ABC
Problem 5.8
The C p /T vs. T data set for solid lead Pb(s) is written in BASIC as follows:
DATA 0,0,5,.061,10,.28,15,.4666,20,.54,25,.564,30,.55,50,.428,
70,.333,100,.245,150,.169 200,.129,250,.105,298,.089
N = 14
The data are in 14 pairs. The first number of the pair is T and the second is C p . Devise
a program of your own or use a canned program to estimate the standard entropy of
Pb(s).
Problem 5.9
Phase changes occur reversibly. For example, the transition from solid ice to liquid
water, which occurs at a temperature that is infinitesimally above the melting point of
273.15 K, can be reversed by lowering the ambient temperature to slightly less than
273.15 K.
(a) The standard enthalpy of fusion fus H ◦ is 6.01 kJ mol−1 . What is the entropy
of fusion, fus S ◦ ?
(b) The standard enthalpy of vaporization of water is vap H ◦ = 40.7 kJ mol−1 .
What is the entropy of vaporization of water, l → v?
(c) Why are the results so different?
P1: OTA/XYZ P2: ABC
6
THE GIBBS FREE ENERGY
Our brief consideration of engines and heat transfer showed that only part of the heat
transferred from a hot reservoir to a cold one by a real process is available to do
work. This is the available or “free” energy that we have been seeking. In the theory
of steam engines, the Helmholtz free energy A = U − TS is central. In chemical
reactions the closely analogous Gibbs free energy G = H − TS is central. We shall
be mainly concerned with the Gibbs free energy.
6.1 COMBINING ENTHALPY AND ENTROPY
We seek a combined function that expresses the spontaneous tendency of a chemical

system to undergo a change in enthalpy simultaneously with a change in its entropy.
In other words, we seek a factor that governs chemical reactions. Clearly, the signs
of enthalpy and entropy must be opposite because one function tends to a maximum
and the other tends to a minimum. We might write X = U − S for our unknown
energy function such that matter flows from a point of high potential to one of low
potential in the way that water flows downhill, but we chemists are usually interested
in reactions that are carried out at constant pressure, so we substitute H for U in the
equation. Also we notice that the units are wrong. Enthalpy has the unit J and entropy
has the unit J K−1 . To bring everything into consistent units of energy, we multiply
the entropy by T to get the central equation of chemical thermodynamics:
G ≡ H − TS

Copyright
84
P1: OTA/XYZ P2: ABC
FREE ENERGIES OF FORMATION 85
This equation implies constant p by the use of H and constant T at thermal equilibrium.
Multiplication of S by T is also consistent with the definition of entropy as d S =
dq/T , where q is a reversible thermal energy or enthalpy. The infinitesimal and small
finite expressions dG and G are also implied:
dG = dH − T dS
and
G = H − T S
The Gibbs function is the algebraic sum of a thermodynamic state function H and
a constant times a thermodynamic function –TS, hence it is a thermodynamic state
function as well.
The terms Gibbs state function, Gibbs free energy, and chemical potential are all
used for the thermodynamic property G. Usually the first term is used to stress the
mathematical properties of the function, the second is used in general descriptions,
and the third is used to stress the intensive nature of the molar free energy µ = G/n.
We shall use these terms more or less synonymously, relying on context to make the
meaning clear or reminding the reader, from time to time, of the distinction between
the molar Gibbs free energy G and the extensive Gibbs function G. The term chemical
potential, for the molar (or partial molar) quantity µ, depicts well the property of
chemical systems to flow down a gradient toward a minimum that we refer to as an
equilibrium point.
6.2 FREE ENERGIES OF FORMATION
At this point, we are able to determine the enthalpy change of a chemical reaction by
direct or indirect calorimetric measurement of f H of the participants in the reaction
and we can find the corresponding entropy change by integration of heat capacity data
for each of the participants. These results permit us to calculate Gibbs free energies
of formation. Suppose that we select combustion of C(graphite) in O2 (g) to form
CO2 (g) as the illustrative case and that we have determined the standard entropies
of these three species to be 5.74, 205.138, and 213.74 J K−1 mol−1 , respectively. We
apply the general formula

Sr298 = S(products) − S(reactants)
to the reaction
C(gr) + O2 (g) → CO2 (g)

P1: OTA/XYZ P2: ABC
86 THE GIBBS FREE ENERGY
to get
Sr298 = 213.74 − 5.74 − 205.138 = 2.862 J K−1 mol−1
as the change in standard entropy of the reacting system.

The enthalpy of formation of CO2 (g) has been measured and found to be −393.51
kJ mol−1 (Section 4.2), so from the fundamental equation for a finite change (G)
G = H − T S
we get
G f = H f − T S f = −393.51 − 298(2.862 × 10−3 )

= −393.51 − 0.852 = −394.36 kJ mol−1
which is the Gibbs function of formation at 298 K of CO2 (g).

Comparable calculations yield the Gibbs functions of many direct formation reac-
tions of elements to their compounds. Armed with these Gibbs functions of formation,
we can manipulate them to find the Gibbs functions of compounds not cleanly formed
from their elements in the same way as we did for the enthalpies of formation.
6.3 SOME FUNDAMENTAL THERMODYNAMIC IDENTITIES
For a reversible change doing only pV work, the first law gives
dU = dq − dw = TdS − p dV
whence, knowing that U = f (S,V ), dU = (∂U/∂ S)V dS + (∂U/∂ V ) S dV, we find

that

∂U ∂U
=T and = −p
∂S V ∂V S

∂G ∂G
Also, G = f ( p,T ), dG = dp + dT , and
∂p T ∂T p
dG = V dp − S dT
with the results that

∂G ∂G
=V and = −S
∂p T ∂T p
P1: OTA/XYZ P2: ABC
PRESSURE DEPENDENCE OF THE CHEMICAL POTENTIAL 87
6.4 THE FREE ENERGY OF REACTION
We have procedures that give us the enthalpies of formation of the participants in

any reaction at 1.0 atm pressure, and we can calculate the absolute entropies of the
participants. Therefore we can construct a table of Gibbs free energies of formation
for many compounds. The free energy change of reaction is found in the usual way:

G r = f G(products) − f G(reactants)
The Gibbs free energies of formation of all elements are defined as zero. This def-
inition is possible because no element can be formed from any other element by an
ordinary chemical reaction.
An insight into the difference between chemical potential and enthalpy can be
found in the stepwise hydrogenation of acetylene, first to ethene and then to ethane:
HC CH(g) + H2 (g) → H2 C CH2 (g)

HC CH2 (g) + H2 (g) → H3 C CH3 (g)
The enthalpies of these two reactions are, respectively, −174 and −137 kJ mol−1
and the Gibbs free energies of reaction are −141 and −101 kJ mol−1 , respectively.
The numbers themselves are quite different, but the difference between them is
comparable: 33 kJ mol−1 in the first case and 36 kJ mol−1 in the second. This is
because each reaction involves “tying up” two moles of gas and releasing only one.
Each reaction involves more or less the same reduction of disorder, hence there is
less energy free to seek a minimum than one might expect considering the enthalpy
change alone. The chemical potential well is less deep than it would be without the
opposing entropy factor.
6.5 PRESSURE DEPENDENCE OF THE CHEMICAL POTENTIAL
The previous calculations were carried out for reactions at 1 atm pressure. All re-
actions are not carried out at 1 atm pressure, so we need a method of finding the
change in chemical potential at any pressure. One can construct a reaction diagram
(Fig. 6.1). If we can find G 2 and G 3 for a change in pressure from 1 bar (or
1 atm) to a new pressure p2 for all the reaction components, the change in Gibbs
chemical potential for the reaction G 4 can be found at any selected pressure p2 . The
problem is already solved, however, because we have the identity (∂G/∂ p)T = V
from Section 6.3. Explicitly stipulating constant temperature, we can go from partials
to total derivatives for an ideal gas:
RT
dG = V dp = dp
p
P1: OTA/XYZ P2: ABC
ΔG4
A( g , p2 ) B( g , p2 )
→
↑ ΔG2 ↑ ΔG3
ΔG1
A( g , p1bar ) B( g , p1bar )
→
FIGURE 6.1 A reaction diagram for G 4 .
Integrating to find G 2 , we obtain

Gp2 Gp2
dp p2
G 2 = dG = RT = RT ln
G1bar G1bar p 1
where the pressure at the lower limit of integration is 1 bar. A similar equation
describes G 3 . Combination of G 1 , G 2 , and G 3 with attention to the sign
differences gives G 4 , the desired change in the Gibbs function for reaction at the
new pressure p2 . One way of getting the signs straight is to set up the diagram as a
cyclic process, and remember that the sum of changes in any thermodynamic function
must be zero around a cycle.
6.5.1 The Equilibrium Constant as a Quotient of Quotients

In writing equilibrium constant expressions such as K eq = p B / p A for a reaction
A → B in the gas phase, it should be remembered that the pressures are quotients
relative to a pressure in the standard state p/ p0 , hence they are unitless and so is K eq .
If the gas is not ideal, the fugacity f is used in place of the pressure or in the case
of liquids and solutions, the activity a may be used. Fugacities or activities are also
ratios f / f 0 or a/a0 relative to a standard state so K eq is still unitless, as it must be for
some of the mathematical manipulations to come. Changes to K eq expressed in terms
of fugacities or activities bring about complication in the mathematical expression
and in the experimental determination of the quantities involved; but once again, the
principles are the same as in the ideal case.
6.6 THE TEMPERATURE DEPENDENCE OF THE FREE ENERGY

u du dv
d v −u
From the usual expression for the derivative of a quotient, v = dx dx ,
dx v2
and stipulating constant pressure, we get

G dG dT
d T −G
T dT dT −T S − G
= =
dT T2 T2
P1: OTA/XYZ P2: ABC
THE TEMPERATURE DEPENDENCE OF THE FREE ENERGY 89
where we recall that (dG/dT ) p = −S. From G = H − TS, H = G + TS, and

−G − TS = −H , so

G
d
T −T S − G −H
= 2
= 2
dT T T
and

G
d
T −H
=
dT T2
It is convenient to remember that d(1/T )/dT = −1/T 2 or dT /d(1/T ) = −T 2 ,

so

G
d
dT T H
= −T 2
− 2 = H
1 dT T
d
T
or

G
d
T
= H
1
d
T
This is one form of the Gibbs–Helmholtz equation. It is general and applies to

chemical and physical changes. For relatively short temperature intervals, H
may be regarded as a constant. A plot of G/T vs. 1/T gives H . Con-
versely, knowing H enables one to determine G/T , hence G at tempera-
tures other than 298 K. Generally speaking, reactions are more sensitive to tem-
perature changes than to pressure changes of comparable magnitude, hence the
Gibbs–Helmholtz equation is of overarching importance in practical and industrial
chemistry.
We now have methods to determine G at any temperature and pressure from
the tabulated G 298 values under standard conditions. The usual technical com-
plications arise for real systems. For example, H may not be constant, but it
may be followed closely by a power series of the form H = a + bT + cT 2 + · · ·.
These extra terms make the equations look messy, but the principles remain the
same.
P1: OTA/XYZ P2: ABC
Example 6.1
Find the Gibbs free energy of formation for methane. The enthalpy of formation of
−1
methane is H 298
f,methane = −74.81 J mol . Standard entropies of H2 (g), C(graphite),
and methane CH4 (g) are also known. They are 130.684, 5.74, and 186.26 J K−1 mol−1 ,
respectively.
Solution 6.1 For the reaction
C(graphite) + 2H2 (g) = CH4 (g)
we have
Sr,CH
298
4
= 186.26 − 2(130.684) − 5.74 = −80.848 J K−1 mol−1
At 298.15 K, we obtain
T S = −0.080848(298.15) = −24.09 kJ mol−1
and
−1
G 298
f,CH4 = −74.81 − (−24.093) = −50.72 kJ mol
(Notice the conversion from J to kJ in the penultimate step.) In this way, we can build
up a table of Gibbs functions for as many compounds as time and money permit.
Problem 6.1
(a) Given that
G = H − TS
show that
dG = dH − T dS − S dT
(b) Given that

H = U + pV
show that
dH = dU + p dV + V dp
for p = const.
P1: OTA/XYZ P2: ABC
(c) Show that dU = T dS − p dV.

(d) Combine (a), (b), and (c) to show that
dG = V dp − S dT
Problem 6.2
(a) Calculate the change in Gibbs free energy that takes place with the isothermal
compression of water treated as an incompressible liquid at 298 K from 1.00
to 2.00 bars pressure.
(b) Calculate the change in Gibbs free energy that takes place with the isothermal
compression of water treated as an ideal gas at 298 K from 1.00 to 2.00 bars
pressure.
(c) Comment upon the difference.
Problem 6.3
The statement has been made: “Algebraic manipulation gives
dG = V dp − S dT ”
Prove this statement.
Problem 6.4
A chemical reaction at 1.0 bar pressure has G r,298p=1 = −335 kJ mol−1 . The Gibbs
function of the reactant system changes by 7.5 kJ mol−1 when the pressure is changed
from 1.0 bar to p2 . The product system changes by 8.4 kJ mol−1 over the same pressure
change. What is G r,298p2 ?
Problem 6.5
What is the entropy change of a Trouton’s rule liquid at its boiling point?
Problem 6.6
The heat capacity of carbon disulfide CS2 (s) is 6.9 J K−1 mol−1 at 15.0 K. What is its
standard entropy at 15.0 K, assuming that the solid is a perfect crystal.
Problem 6.7
Solid carbon disulfide shows the following experimental values of heat capacity as a
function of temperature. Taking the Debye contribution into account (Problem 6.4),
find the standard entropy of CS2 (s) at its normal melting point of 161 K.
P1: OTA/XYZ P2: ABC
T Cp
15.0000 6.9000
20.0000 12.0000
29.8000 20.8000
42.2000 29.1000
57.5000 35.6000
75.5000 40.0000
89.4000 43.1000
99.0000 45.9000
108.9000 48.5000
119.9000 50.5000
131.5000 52.6000
145.0000 54.3000
157.0000 56.6000
161.0000 57.4000
Problem 6.8
The enthalpy of fusion of CS2 (s) to liquid carbon disulfide CS2 (l) is 4.38 kJ mol−1
Pitzer et al. 1961. What is the entropy of fusion of CS2 (s) at its normal melting point
of 161 K?
Problem 6.9
The heat capacity at constant pressure of liquid CS2 (l) is nearly constant at 75.5
J K−1 mol−1 . What is the molar entropy increase of CS2 (l) from 161 K to 298 K?
Problem 6.10
Find the standard molar entropy of carbon disulfide at 298 K.
Problem 6.11
The Gibbs free energies of combustion of methane at 300K and 350 K are
−815 and −802 kJ mol−1 . Find the enthalpy of combustion of methane at 325 K
c H 325 (methane).
Problem 6.12
Solve the previous problem by setting up a pair of simultaneous equations and solving
them.
P1: OTA/XYZ P2: ABC
7
EQUILIBRIUM
Chemists are often depicted in the popular media as sinister fellows who pour a
solution from beaker A into beaker B with catastrophic results. In fact, we are neither
more nor less sinister than the next person and we take elaborate precautions to avoid
catastrophies. Nevertheless, the process of adding one or many components A of one
kind to a system of another kind B to establish an equilibrium mixture is central to
our art or science, and the theoretical examination of chemical equilibrium is the high
point and culmination of classical chemical thermodynamics.
7.1 THE EQUILIBRIUM CONSTANT
Suppose that gases A and B are capable of equilibration
A(g) →
← B(g)
and arbitrary amounts of A(g) and B(g) are introduced into a closed container at 298 K.
Because the amounts are arbitrary, the quotient of the partial pressures Q = pB / pA
in the container will not be the equilibrium constant K eq . The chemical potentials G A
and G B will probably not be in the standard state. Instead they will differ from G ◦A
and G ◦B , by
pA
G A = G ◦A + RT ln
1.0

Copyright
93
P1: OTA/XYZ P2: ABC
94 EQUILIBRIUM
and
pB
G B = G ◦B + RT ln
1.0
where the 1.0 in the denominators signify that the partial pressures are relative to
the standard state of 1.0 bar. The difference between the chemical potentials of the
reactant state and product state is

r G = G(prod) − G(react)
which in this simple case is
pB pA
r G = G B − G A = G ◦B − G ◦A + RT ln − RT ln
1.0 1.0
or
pB
r G = G ◦ + RT ln
pA
As the reaction progresses, r G is not zero and G of the reacting system is not constant
with time t, (∂G/∂t)T, p = 0. When the chemical reaction has come to completion,
the pressure quotient Q = pB / pA has arrived at a value such that (∂G/∂t)T, p = 0,
hence
pB
r G = G ◦ + RT ln =0
pA
The free energy change of the system has arrived at a Gibbs potential energy mini-
mum. Under these and only under these conditions, we have r G = 0, so that
pB
G ◦ = −RT ln = −RT ln K eq
pA
The expression is frequently written in the equivalent form:

◦
K eq = e−G /RT
7.2 GENERAL FORMULATION
A more general formulation of the equilibrium expressions above is given by the

reaction
aA + bB + . . . = cC + dD + . . .
P1: OTA/XYZ P2: ABC
GENERAL FORMULATION 95
where a, b, . . . c, d, . . . are the stoichiometric coefficients of the reaction. It is still

true that

r G = G(prod) − G(react)
When equilibrium has been reached, we have
G ◦ = −RT ln K eq
Now the concentration quotient Q takes the form
[C]c [D]d . . .
Q=
[A]a [B]b . . .
leading to the familiar expression of the equilibrium constant as
[C]c [D]d . . .
K eq =
[A]a [B]b . . .
which is true only after the Gibbs free energy has come to a minimum and r G = 0.
The stoichiometric coefficients become exponents, and the square brackets [ ] indicate
some kind of unitless concentration variable relative to a standard state. This notation
is often used in solution chemistry to denote a concentration in moles/liter, where the
standard state of the solute in the solvent is taken for granted.
As an example of a reaction in the gas phase, the expression
2
pNO
G ◦ = −RT ln 2
pN2 O4
can be used to find the equilibrium constant for the reaction
N2 O4 (g) →
← 2NO2 (g)
The standard state Gibbs chemical potential difference for this reaction is
r G ◦ = G ◦2(NO2 ) − G ◦N2 O4 = 2(51.31) − 97.89 = 4.73 kJ mol−1
The equilibrium constant is

◦
K eq = e−G /RT
= e−4730/8.314×298.15 = 0.148
which is in good agreement with the experimental value of 0.13.

It is difficult to obtain accurate K eq values from calorimetric determination
of r G ◦ (from r H ◦ and r S ◦ ) because of the exponential relationship K eq =
◦
e−G /RT . This mathematical form brings about a large error in K eq when r G ◦ is
in error by a small amount. To a certain degree, a “small” error or a “large” error is
in the eye of the beholder; the terms are used in the literature as influenced by the
P1: OTA/XYZ P2: ABC
96 EQUILIBRIUM
“difficulty” of the experiment and how “good” previous measurements of the same
kind have been. Clearly, these are subjective value judgments.
7.3 THE EXTENT OF REACTION
In a chemical reaction, the participant concentrations change, but they do not change
independently, rather they are related by the stoichiometric coefficients of the reaction.
For example, if I know that reactant A decreases by 0.1 mol in the reaction system
A + B = 2C
I know that reactant B also decreases by 0.1 mol and product C increases by 0.2 mol.
The infinitesimal changes in mole numbers dn A , dn B , and dn C are linearly depen-
dent. One frequently expresses the change in a reaction using the single variable ξ
which, given the linear dependence of the reaction components, suffices to express
all three in the specific case cited. In general, infinitesimal variations in the amount
of any number of components dn i can be expressed in terms of the extent of reaction
ξi where
dn i
ξi =
νi
and νi is the stoichiometric coefficient: −1, −1, and 2 in the specific example cited.
Note that the stoichiometric coefficients of the two reactants are negative because
their concentrations decrease as the reaction proceeds. This is a convention of course,
because the reaction could equally well have been written
2C = A + B
whereupon the signs of ξi would change.

Up to this point, we have stressed thermodynamic state functions for 1 mol of a
pure substance, for example, the molar energy U = f (V, T ). When we express the
energy of any mixture, specifically a mixture of reactants and products in a reacting
system, the composition of the system in terms of mole numbers n i = 1.0 influences
the energy and other state functions. We can express this dependence in terms of the
extent of reaction:
U = f (V, T, ξi )
This dependence means that along with V and T, ξi is a full-fledged degree of freedom
and we can write the exact differential

∂U ∂U ∂U
dU = dV + dT + dξi + . . .
∂ V T,ξi ∂ T V,ξi ∂ξi T,V
where i extends over all degrees of freedom.

P1: OTA/XYZ P2: ABC
VARIATION OF THE EQUILIBRIUM CONSTANT WITH TEMPERATURE 97
Notice that this implies a hyperspace in U, V, T, and ξi . Energy has been selected
here to illustrate the principle that analogous equations exist for the other state
variables—for example, G, H, and S.
7.4 FUGACITY AND ACTIVITY
For nonideal systems the concentration variable is replaced by a new variable that
expresses the effective concentration of the species in a mixture. For example, a solute
may be more chemically active in methanol solution than it is in water. Or it may be
more active in water solution than in methanol. A pure gas may behave in a nonideal
way, and its degree of nonideality may be influenced by other gases in a mixture.
These deviations from ideal behavior are expressed by a coefficient γ which yields
the activity of a solute or fugacity of a gas when multiplied into the concentration
variable, for example,
aA = γA [A] or f A = γA p A
The activity and fugacity coefficients are simply numbers telling us whether the
behavior of the species is greater or less than it would be in the standard state. They
are concentration- or pressure-dependent and are usually determined for real systems
by rather painstaking empirical methods.
7.5 VARIATION OF THE EQUILIBRIUM CONSTANT

WITH TEMPERATURE
Combining the Gibbs–Helmholtz equation for the temperature variation of free energy
with the equation connecting the free energy in the standard state to the equilibrium
constant gives
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
G ◦ −RT ln K eq

∂ ∂
⎢ ⎥ ⎢ ⎥ ⎢ ∂ R ln K eq ⎥
⎢ T ⎥ = ⎢ T
⎥ = −⎢ ⎥
⎣ 1 ⎦ ⎣ 1 ⎦ ⎣ 1 ⎦
∂ ∂ ∂
T p T p T p
but
⎡ ⎤
G ◦
∂
⎢ ⎥
⎢ T ⎥ = H ◦
⎣ 1 ⎦
∂
T p
P1: OTA/XYZ P2: ABC
98 EQUILIBRIUM
so
⎡ ⎤

⎢ ∂ R ln K eq ⎥
⎢ ⎥ ◦
⎣ 1 ⎦ = −r H
∂
T p
where r H ◦ is the standard enthalpy of reaction. Now, dT −1 /dT = −T −2 and

d(1/T ) = −T −2 dT, so

1 H ◦
◦
d R ln K eq = −H d = dT
T T2
This equation can be integrated as

r H ◦ 1
d ln K eq =− dT
R T2
to give the van’t Hoff equation

r H ◦ 1 r H ◦ 1
ln K eq =− − + const. = + const.
R T R T
This is the equation of a straight line of ln K eq plotted against (1/T ) with r H ◦ /R

as the slope. It applies so long as r H ◦ remains constant. The equation as written
implies that the slope will be positive, but this does not follow because r H ◦ may
be positive, negative, or zero for endothermic, exothermic, or null-thermal reactions.
Common examples are melting, freezing, or mixing of two isomers of a liquid alkane
which are endothermic, exothermic, and null-thermal.
The van’t Hoff equation can be integrated between limits to give an expression for
a new equilibrium constant at a new temperature from known values of r H ◦ , K eq ,
and the initial and revised temperatures, T and T :
T
K eq
r H ◦ 1
d ln K eq = − 2
dT
K eq R
T T

K eq r H ◦ 1 1 r H ◦ 1 1
ln =− − = −
K eq R T T R T T
Conversely, the integrated form is a way of determining r H ◦ from two experimental

determinations of K eq at different temperatures.
P1: OTA/XYZ P2: ABC
VARIATION OF THE EQUILIBRIUM CONSTANT WITH TEMPERATURE 99
7.5.1 Le Chatelier’s Principle

Le Chatelier’s principle states that, in a stressed chemical reaction, the equilibrium
will be displaced in such a way as to relieve the stress. First, we need to define
“stress.” An exothermic reaction
A→
←B+q
where q is heat given off, will be stressed backward by application of heat q—that
is, by a temperature rise. Application of heat will drive the reaction backward to give
more A and less B. The opposite will be true for an endothermic reaction. Looking
at the integrated van’t Hoff equation

K eq r H ◦ 1 1
ln = −
K eq R T T
we see that for a temperature rise T > T we have (1/T ) > (1/T ). For r H ◦ < 0,

ln (K eq /K eq ) < 0 and the new equilibrium constant K eq is smaller than the original
◦
one, K eq . The thermal condition r H < 0 is the characteristic of an exothermic
reaction, so Le Chatelier’s principle applied to the heat of reaction agrees with the
van’t Hoff equation. Le Chatelier’s principle is a qualitative statement giving the
sign but not the magnitude of the effect of a temperature change on the equilibrium
constant.
7.5.2 Entropy from the van’t Hoff Equation

If r H ◦ is independent of the temperature, then from dr H ◦ = r C p dT = 0. Since
dT is not zero, r C p must be zero. From

r C p = C p (prod) − C p (react)
the heat capacity of the reactant system must be equal to that of the product system. In
practical terms, we take this equality to be approximately true over short temperature
intervals. We have already found the equations
r G ◦ = −RT ln K eq
and
r G ◦ = r H ◦ − T r S ◦
so
−RT ln K eq = r H ◦ − T r S ◦
P1: OTA/XYZ P2: ABC
100 EQUILIBRIUM
and

−r H ◦ 1 r S ◦
ln K eq = +
R T R
This is a “slope–intercept” problem for the linear function ln K eq vs. 1/T , where the
slope of the plot is −r H ◦ /R and the intercept is r S ◦ /R. Two or more equilibrium
measurements at different temperatures yield both the enthalpy change of the reaction
and the entropy change. The resulting thermodynamic functions are only as good as
the input data, and this is a very sensitive experimental problem—especially in the
r S ◦ determination, which depends on what may be a long and mathematically
questionable extrapolation.
7.6 COMPUTATIONAL THERMOCHEMISTRY
Within a year of the discovery of quantum mechanics by Schrödinger and Heisenberg,

Heitler and London had demonstrated that the energy of the H H chemical bond can
be approximated by a quantum mechanical calculation. The reason this calculation
is always approximate is a reflection of one of the most profound laws of nature,
Heisenberg’s uncertainty principle. We are unable to know exactly where the moving
electrons are. Hence their Coulombic potential energy, total energy, entropy, and
Gibbs free energy are inaccessible to us in even the simplest chemical bond H H.
Nevertheless, a diverse array of approximate computational methods exists and
the results can be ranked in order of the quality of their approximations. Ranking
enables us to decide which methods are best for a given job, and it points the
way toward improved methods. Usually the quality of an approximate calculation is
measured by how well the calculated value matches well-established experimental
results. Recently, however, computed values have become so reliable that they can be
used to pick out and discard errors in the tabulated experimental results. The present
trend is from using computed results to approximate experimental results to using
them to supplant or replace experimental results, especially for reactions that occur
within hostile experimental environments (flames or explosions) or reactions that
have fleeting intermediates (free radicals). Quantum mechanical methods produce
results for molecular energy, enthalpy, entropy, Gibbs free energy, and equilibrium
constants, but they are circumscribed by limits on molecular size and they are of
varying reliability.
7.7 CHEMICAL POTENTIAL: NONIDEAL SYSTEMS
Given that the effective pressure fraction of a gas in a nonideal mixture of gases is
its fugacity f and that the effective concentration fraction of a solute in a mixture is
given by its activity a, equilibrium constants in simple A, B nonideal systems can be
P1: OTA/XYZ P2: ABC
CHEMICAL POTENTIAL: NONIDEAL SYSTEMS 101
written
fB aB
K eq = or K eq =
fA aA
The Gibbs free energy of a pure substance is a function of the temperature and
pressure only, G = g(T, p), but the free energy of a many-component system is a
function of T, p, and the amounts of each component n i , G = g(T, p, n i ). The total
change in free energy for a mixture is
dG
dG dG
dG = dT + dp + dn i
dT p,n i
dp T,n i i
dn i p,T
It is the last term that we are interested in. At constant T and p, we have
dG
dG = dn i
i
dn i p,T
where the n j are constant ( j = i). In the simple illustrative reaction mixture,
A(g) →
← B(g)
in the ideal case,
pB
G = G ◦ + RT ln Q = G ◦ + RT ln
pA
where the p values are partial pressures. In the nonideal case the form of the pressure
quotient Q still holds, but the partial pressures are replaced by the fugacities f . The
change in the Gibbs free energy is also given a different symbol µ to denote the
nonideal case:
fB
µ = µ◦ + RT ln Q = µ◦ + RT ln
fA
Evidently,
µ = µB − µA
and
µ◦ = µ◦B − µ◦A

P1: OTA/XYZ P2: ABC
102 EQUILIBRIUM
Although the Gibbs chemical potential applies to the many-component nonideal case
as well as the ideal case, it is mathematically rigorous and it is a thermodynamic
function.
From these definitions of the Gibbs chemical potentials, it follows that
µA = µ◦A + RT ln f A
and
µB = µ◦B + RT ln f B
In the general case, Gibbs chemical potentials are
µi = µ◦i + RT ln f i
for an indefinite number of fugacities.

Analogous definitions for the general case lead to
µi = µ◦i + RT ln ai
for activities ai , with

[X ]i
[C] [D] . . .
c d
i
Q= =
[A]a [B]b . . . [X ] j
j
where the [X ]i and [X ] j are relative concentration variables, pressure fractions, mole
fractions, mass fractions, and so on, symbolized by either f or a and taken to the
appropriate stoichiometric coefficient.
7.8 FREE ENERGY AND EQUILIBRIA IN BIOCHEMICAL SYSTEMS
Reactions of biochemical interest do not normally occur in the gas phase. Rather
they occur in solution, usually saline solution. Therefore the correct expression
of the free energy changes of reaction and the equilibrium constant is in terms
of the corresponding activities and changes in chemical potential. Determination of
activities and activity coefficients over a concentration range on nonideal solutions is
not a simple matter, nor is it necessary for in vitro studies of biochemical reactions.
Instead, concentrations are used with the stipulation that the “background” condi-
tions must be constant over the course of the study and must be reproduced from one
study to the next. For example, energy studies on dephosphorylation of adenosine
P1: OTA/XYZ P2: ABC
FREE ENERGY AND EQUILIBRIA IN BIOCHEMICAL SYSTEMS 103
5 -triphosphate would be carried out at a specified and constant temperature, pres-

sure, pH, pMg, and
ionic strength where
pH and pMg refer to the ion concentrations
pH = − log H+ and pMg = − log Mg2+ , and the ionic strength is constant for all
dissolved salts in the reaction solution. Ionic strength
is essentially the ionic charge
concentration in solution calculated as the sum 12 ci z i2 over all salt concentrations
ci that dissolve to give ions of charge z i .
These conditions specify a unique standard state which is not the thermodynamic
standard state but which is adhered to throughout the experiment and experiments
with which results will be combined or compared. Under these controls, it is proper
to write
G A = G ◦A + RT ln [A]
for G ◦A in the specified standard state with a concentration of reactant A, with similar
expressions for B, C, . . . . The change in free energy for a reaction quotient Q, now
in terms of initial concentrations, is
[C]ξ [D]ξ . . .
G = G ◦ + RT ln Q = G ◦ + RT ln
[A]ξ [B]ξ . . .
Different stoichiometric coefficients are all expressed using the same symbol ξ for
simplicity. It is understood that Q is a concentration quotient, which might fortuitously
be equal to the equilibrium constant but in general will not. In general, G will be
different from zero but when the reaction has arrived at equilibrium, the concentrations
will have arrived at K eq and G will have arrived at zero so that
G ◦ = −RT ln K eq
Thus the form of the free energy relation to the equilibrium constant is reproduced but
only under rigorously controlled background conditions.1 Change the ionic strength
or the pH, for example, and you can expect to find a different G ◦ and K eq .
How can G ◦ change when we usually think of it as a sum of rock-firm G ◦
values? By changing our background conditions, we have changed the standard state,
the benchmark to which we refer all G ◦ values.
7.8.1 Making ATP, the Cell’s Power Supply

The difference between r G and G ◦ in the metabolic degradation of glucose,
a physiological energy source, to lactate and ATP can be broken down into two
individual steps (Hammes, 2007), one taking up 2 mol of ATP and producing 2 mol
of the diphosphate ADP
Glucose + 2 ATP → 2 ADP + 2 glyceraldehyde-3-phosphate

1 See Treptow (1996).
P1: OTA/XYZ P2: ABC
104 EQUILIBRIUM
and the other producing 4 mol of ATP
2 Glyceraldehyde-3-phosphate + 2 phosphate + 4 ADP → 2 lactate + 4 ATP + 2 H2 O
The summed reaction is
Glucose + 2 phosphate + 2 ADP → 2 lactate + 2 ATP + 2 H2 O
for a net gain of 2 mol of ATP. The first of these two reactions is unfavorable from
the point of view of a positive G ◦ = 2.2 kJ mol−1 but at the initial concentrations
chosen to be at or near physiological concentrations,
(0.14)2 (0.019)2

Q= 2
= 0.0196 3.61 × 10−4 ≈ 4 × 10−7 1
5.0 (1.85)
This very small reaction quotient drives the metabolic conversion of ADP to ATP
which then powers other reactions within the biological system as a whole.
PROBLEMS AND EXAMPLES
Example 7.1 Solution Calorimetry

A half-liter solution calorimeter system consisted of the calorimeter itself, a tempera-
ture measuring circuit, an electrical heating circuit, and calorimeter fluid. The system
was prepared with a solution of magnesium ion at 0.001 M. The ionic strength was
brought to 0.25 M with KCl, a neutral electrolyte, and the pH was adjusted to 7.0.
When the heating circuit was activated at a current flow of I amperes for a time t,
an amount of heat energy q p = 96.5 J was delivered through a heating resistor R
(q = I 2 Rt) producing a temperature rise of T = 0.166 K.2 What was the water
equivalent (calorimeter constant) of the calorimeter system?
Solution to Example 7.1 The water equivalent of a calorimeter is the heat capacity
of the entire system as if it were all water even though it consists of various parts made
of various materials and contains a solution of reactants and products different from
pure water. The water equivalent can be found by a straightforward heat capacity
calculation even though we know that the heat capacity of the system is the sum of
many parts. We calculate
dq p qp 95.6
Cp = = = 576 J K−1 = 0.576 kJ K−1
dT T 0.166
21 amp volt sec = 1 J and E in volts = IR.

P1: OTA/XYZ P2: ABC
PROBLEMS AND EXAMPLES 105
Example 7.2 Adenosine 5 -triphosphate ATP

The hydrolysis of adenosine 5 -triphosphate ATP to adenosine 5 -diphosphate ADP
plus an inorganic phosphate ion can be written
ATP + H2 O → ADP + phosphate
If 10 mL of ATP with concentration [ATP] = 0.200 mol L−1 are pipetted into a
calorimeter with a heat capacity of 0.576 kJ K−1 (Example 7.1) and a temperature rise
of 0.107 K is found, what is the enthalpy of hydrolysis of adenosine 5 -triphosphate
to adenosine 5 -diphosphate plus an inorganic phosphate ion under these conditions?
What is the sign of r H ?
Solution to Example 7.2

0.107 K 0.576 kJ K−1 = 0.0616 kJ
10.0 mL = 0.0100 L
0.0100 L(0.200 mol L−1 ) = 0.00200 mol ATP
0.0616 kJ
= −30.8 kJ mol−1
0.00200 mol
The sign of r H is negative because the reaction is exothermic. Heat flowing out
of the reacting system at constant external pressure means that its enthalpy balance
must decrease.
Problem 7.1
What is the change in the entropy S for one mole of a pure substance for an infinites-
imal change in T and p? Given: Entropy is an exact differential

∂S ∂S
d S(T, p) = dT + dp
∂T p ∂p T
Problem 7.2
In an experiment on the nitrogen tetroxide reaction at 298 K
N2 O4 (g) →
← 2NO2 (g)
pure N2 O4 (g) was introduced into a reaction vessel maintained at constant temperature
and a total pressure of 2.500 bar. When equilibrium had been reached, the partial
P1: OTA/XYZ P2: ABC
106 EQUILIBRIUM
pressure of N2 O4 (g) had dropped to 1.975 bar. What is K eq for this reaction? Compare
your answer to the calculated value (Section 7.2). What are the units of K eq ?
Problem 7.3
Supposing that a simple A(g) →
← B(g) reaction has an equilibrium constant K eq =

0.100 at T = 200 K and K eq = 0.200 at T = 300 K. What is r H ◦ for the reaction?
Problem 7.4
For the reaction
Br2 (g) → 2Br(g)
the equilibrium constant has been measured and found to be 0.410 at 1125 K and
1.40 at 1175 K. What is K eq at 1225 K?
Problem 7.5
Data quoted in Lewis and Randall (1961) for the formation reaction of H2 S
H2 (g) + 12 S2 (g) = H2 S(g)
include the values given in Table 7.1.
TABLE 7.1 T, 1/T, K, and ln K for formation of H2 S from its elements.
T 1023 1218 1405 1667

1/T 9.775 × 10−4 8.210 × 10−4 7.117 × 10−4 5.999 × 10−4
K 105.9 20.18 6.209 1.807
ln K 4.662 3.005 1.826 0.5917
What is r H ◦ for this reaction? What is r S ◦ ?
Problem 7.6
A reaction in the gas phase has K eq = 10 at 298 K and 10−1 at 500 K. Is the reaction
endothermic or exothermic? Explain this in terms of Le Chatelier’s principle.
Problem 7.7
An experiment was set up so that conditions (pH, pMg+ , ionic strength, etc.) were
identical to those described in Exercises 7.1 and 7.2, except that, in this experiment,
P1: OTA/XYZ P2: ABC
adenosine 5 -diphosphate ADP was substituted for ATP. The hydrolysis reaction
producing adenosine 5 -monophosphate AMP
ADP + H2 O → ADP + phosphate
resulted in a temperature change of T = 0.100 K. Find H ◦ for the reaction
2ADP → AMP + ATP
(The symbol H ◦ is used to indicate a standard enthalpy change relative to a defined

concentration under specific solution conditions.)
P1: OTA/XYZ P2: ABC
8
A STATISTICAL APPROACH
TO THERMODYNAMICS
In the late nineteenth century, Ludwig Boltzmann made the connection between
Maxwell’s statistical-atomic equations and the deterministic equations of chemical
thermodynamics, which were only emerging at the time of his work. (Gibbs was not
widely read in Europe at that time.) A central concept in statistical thermodynamics,
as we now call the new science, is the partition function. We shall see the relation
between the partition function and the thermodynamic properties including the Gibbs
free energy and the equilibrium constant. Actual calculation of partition functions
falls anywhere within the range of easy to impossible. We shall calculate some of the
easy ones and approximate some of the others.
8.1 EQUILIBRIUM
If two very simple gaseous systems, A and B, are in equilibrium and each system has
only one energy level as shown in Fig. 8.1, the equilibrium constant is K eq = n B /n A =
3/5 = 0.600. Knowing K eq , we can calculate the energy separation between levels
A and B from the Boltzmann equation:
K eq = e−(εB −εA )/kB T
For example, at 298 K, in Fig. 8.1, K eq = 0.600 leads to (εB − εA ) = 2.10 × 10−21 J.

Copyright
108
P1: OTA/XYZ P2: ABC
DEGENERACY AND EQUILIBRIUM 109
A(g) B(g)
··· εB
····· εA
FIGURE 8.1 A two-level equilibrium. There are 5 molecules in A and 3 molecules in B.
Conversely, knowing the energy separation, one can calculate the equilibrium
constant. If the energy of B is lower than the energy of A by 2.10 × 10−21 J, level B
will contain more molecules at equilibrium than will level A and K eq will be larger
than 1.0 (Fig. 8.2).
nB 5
K eq = = e−(εB −εA )/kB T = = 1.67
nA 3
In these simple models, the equilibrium constants are inverses and (εB − εA ) of the
first equilibrium is equal but opposite in sign to (εB − εA ) of the second reaction. This
analysis gives a description of the relationship between energy and the equilibrium
constant that is useful as far as it goes, but it has left out an important consideration:
the influence of entropy.
8.2 DEGENERACY AND EQUILIBRIUM
If the upper level is split into two levels, each of which has the same energy, its
capacity to accommodate molecules of kind B is doubled (Fig. 8.3). The equilibrium
6
constant K eq = has been made greater than 1.0. The reaction now favors products
5
rather than reactants even though the energy change is uphill. The second law has
come into play; the product state is more disordered. There are more places for the
molecule to go. It is like having two stacks of paper on your desk rather than one.
There are now two places where that important piece of paper you are looking for may
be, whereas before there was only one. Increased disorder means increased entropy.
···
·····
A B
FIGURE 8.2 A two-level equilibrium.
P1: OTA/XYZ P2: ABC
110 A STATISTICAL APPROACH TO THERMODYNAMICS
··· ···
·····
A B
FIGURE 8.3 A degenerate two-level equilibrium.
Even if the two levels of state B are not exactly at the same energy, their capacity
to accommodate B molecules is greater than it would be if there were only one state
as in the original equilibrium. Suppose now that both state A and state B consist
of many levels. The distribution of molecules in state A will be controlled by one
Boltzmann factor and the distribution of molecules in state B will be controlled by
another Boltzmann factor, but the distribution between A and B will also be controlled
by a Boltzmann factor. We now have three factors controlling the equilibrium: the
distribution within A, the distribution within B, and the distribution between A and
B. The distribution within the levels of state A is
n Ai
= e−(εAi −εA0 )/kB T
n A0
The distribution within state B is

n Bi
= e−(εBi −εB0 )/kB T
n B0
The distribution between the lowest levels in states A and B is

n B0
= e−(εB0 −εA0 )/kB T
n A0
The total number of molecules in state A is the summation over all the states in A

nA = n Ai = n A0 e−(εAi −εA0 )/kB T
i i
···
···
·····
A B
FIGURE 8.4 A degenerate two-level equilibrium. The two energies at the B level are not
exactly the same.
P1: OTA/XYZ P2: ABC
DEGENERACY AND EQUILIBRIUM 111
==========
==========
==========
A B
FIGURE 8.5 A two-level equilibrium with many A and many B levels.
and the total number in state B is given by a similar summation

nB = n Bi = n B0 e−(εBi −εB0 )/kB T
i i
The equilibrium constant can be written as the quotient of the number of molecules
in B over the number of molecules in A:

n B0 e−(εBi −εB0 )/kB T
nB i
K eq = =
nA n A0 e−(εAi −εA0 )/kT
i
Factoring out n B0 /n A0 = e−(εB0 −εA0 )/kB T , which is common to each term in the sums,
we obtain

e−(εBi −εB0 )/kB T e−(εBi −εB0 )/kB T
n B0 i
= e−(εB0 −εA0 )/kB T
i
K eq =
n A0 e−(εAi −εA0 )/kB T e−(εAi −εA0 )/kB T
i i
which is to say
QB
K eq = e−(εB0 −εA0 )/kB T
QA
where Q A and Q B are called the partition functions for the A and B states. Please do
not confuse the partition function Q with the concentration quotients Q or Q.
P1: OTA/XYZ P2: ABC
8.3 GIBBS FREE ENERGY AND THE PARTITION FUNCTION
Taking logarithms of the previous equation, we obtain
ε0 QB
ln K eq = − + ln
RT QA
where ε0 = εB0 − εA0 . Multiplying by −RT , we get
QB
−RT ln K eq = ε0 − RT ln
QA
but −RT ln K eq = G ◦ , so
QB
G ◦ = ε0 − RT ln
QA
Comparing this equation with the definition of the Gibbs chemical potential
G ◦ ≡ H ◦ − T S ◦ , we see that the classical enthalpy change H ◦ for a re-
action is largely controlled by the energy separating the ground states ε0 , and the
entropy change T S ◦ is largely controlled by the ratio of the multiplicity of levels
available in each state RT ln (Q B /Q A ). In very many cases, the enthalpy change is
dominant in chemical reactions, which explains why some very good nineteenth-
century scientists thought that H ◦ was the only factor controlling equilibrium and
reaction spontaneity.
Given that reactions and physical processes exist that are not solely controlled or
dominated by the enthalpy change, it is the ratio of partition functions that interests
us. The partition function gives us the distribution of molecules over the energy levels
within a state. We may think of these as microstates within a state. If the enthalpy
balance is close, the reaction tends to go in the direction with the most microstates.
Such a reaction or process is maximizing the number of choices a particle (molecule)
has in which to reside within a state. A system providing a maximum number of
choices provides the maximum disorder, just as the number of places you may find
that piece of paper you are looking for is determined by the disorder of the desktop.
A disordered desktop has more microstates than an orderly one.
Looking again at the equation G ◦ = ε0 − RT ln (Q B /Q A ), we can see another
reason why thermodynamicists prior to Gibbs could have missed the importance of
entropy in equilibrium. The second term includes T, so it may be small at ordinary
temperatures. The ratio of microstates Q B /Q A may be negligible at lower tempera-
tures, but it becomes important at higher temperatures and it may favor or disfavor
the product side of the equilibrium according to which is larger, Q A or Q B . As is
true in classical thermodynamics, there are four possibilities for the two terms in
G ◦ = ε0 − RT ln (Q B /Q A ): They may be FF, FD, DF, or DD, where F denotes
a term that favors the product side of the equilibrium and D denotes a term that
disfavors the product side.
P1: OTA/XYZ P2: ABC
THE THERMODYNAMIC FUNCTIONS 113
8.4 ENTROPY AND PROBABILITY
If you drop marbles randomly into a large box with 75 compartments painted green
and 25 compartments painted red, the ratio of occupation numbers of marbles in
compartments tends toward 75/25 = 3/1 over very many trials. If you put all the
marbles into the green compartments and shake the box, the distribution of marbles
in compartments will approach 3/1, as before, even though the potential energy of
the boxes in the gravitational field is the same (for a level box). The ratio of the
number of red compartments to the number of green compartments is the ratio of
microstates for the system. It is independent of the size of the system; 7500/2500
is the same as 75/25. Call this ratio W. Boltzmann made the connection between
statistics (probability ratio) and thermodynamics by defining the proportionality
S ∝ ln W
but this is just the way we found the A, B equilibrium constant except that it doesn’t
contain the temperature T. You supply the “thermal energy” by shaking the box. If
you don’t shake the box, it does not approach equilibrium (no reaction at 0 K).
The universal constant of proportionality between S and ln W is now designated
kB and called Boltzmann’s constant in honor of this great and tragic figure1 :
S ≡ kB ln W
kB = 1.38066 × 10−23 J K−1
8.5 THE THERMODYNAMIC FUNCTIONS
If the number of particles at each energy level
ne−εi /kB T
ni =
Q
is multiplied by the energy at that level εi , then the total energy is the sum of the
individual contributions

n e−εi /kB T
i
E= εi n i =
i
Q
Having the energy and entropy, we have all that is necessary to derive the rest of
the thermodynamic functions in terms of the partition function Q. Irikura (1998) has
given the necessary equations in compact form, which we present as Table 8.1.
1 Suicide, 1906.
P1: OTA/XYZ P2: ABC
TABLE 8.1 Thermodynamic Functions

(Irikura, 1998).

∂
S = NkB (T ln Q) − ln N + 1
∂T
∂2
C V = NkB T (T ln Q)
∂T 2
C p = CV + R
T
RT 2 ∂ Q
H (t) − H (0) = C p dT = + RT
0 Q ∂T
∂ T ∂Q
(T ln Q) = ln Q +
∂T Q ∂T
2
∂2 2 ∂Q T ∂2 Q T ∂Q
(T ln Q) = + − 2
∂T 2 Q ∂T Q ∂T 2 Q ∂T
∂Q 1 −εi /kB T
= εi e
∂T kB T 2 i
∂2 Q −2 ∂ Q 1 2 −εi /kB T
= + 2 4 ε e
∂T 2 T ∂T kB T i i
The essential thermodynamic functions are expressed above the line in Table 8.1,
and the derivatives are given below the line for somewhat simpler and more practical
evaluation. Along with the equations in Table 8.1, “excess” free energy and enthalpy
functions are defined as (G ◦ − H0 )/T and (H ◦ − H0 )/T . The excess thermodynamic
property is the amount of that quantity above the value per degree K the system would
have if all its parts were in the lowest possible energy state. Excess free energy and
enthalpy functions vary in a gradual way over a wide range of temperature T. This
enables one to make accurate interpolations and to extrapolate from data one has to
get information one does not have.
Now we need to find out how to calculate the partition function.
8.6 THE PARTITION FUNCTION OF A SIMPLE SYSTEM
Spectroscopy gives information as to the intervals between energy levels in molecules,

hence it is a gateway to molecular partition functions. As a simple example, consider
a noninteracting2 system of chemical bonds treated as harmonic oscillators (Chap-
ter 18). A harmonic oscillator has the peculiar property that its energy levels are
2 Strictly, the system is weakly interacting because particle transfer from one level to another is a necessary
condition of the statistical treatment.

P1: OTA/XYZ P2: ABC
THE PARTITION FUNCTION OF A SIMPLE SYSTEM 115
equally spaced. Because of this, the sum, written without degeneracy,

Q vib = e−εi /kB T , i = 0, 1, 2, . . .
i
can be represented by
Q vib = 1 + e−ε/kB T + e−2ε/kB T + e−3ε/kB T + · · ·
where ε is the size of equal steps up an energy ladder. Multiplying both sides of this
equation by e−ε/kB T , we get
e−ε/kB T Q vib = e−ε/kB T + e−2ε/kB T + e−3ε/kB T + · · ·
but this sum is the same as the one above it except that it lacks the first term. The
difference between the two sums is just unity:
Q vib − e−ε/kB T Q vib = 1
so
1
Q vib =
1 − e−ε/kB T
An ideal harmonic oscillator has only one vibrational spectral line at frequency ν,
which gives its energy through the Planck equation ε = hν, where h is Planck’s
constant 6.626 × 10−34 J s. Knowing the energy ε for the oscillator, one knows Q vib .
Spectroscopists give the frequency of vibration as ν̃ in units of reciprocal cm, cm−1 .
Converting this unit to ν, we have ν = c ν̃ where c is the speed of electromagnetic
radiation, 3.00 × 108 m s−1 . The formula for Q vib in spectrocopist’s terms is
1 1 1
Q vib = = =
1 − e−ε/kB T 1 − e−hcν̃β 1 − e−a
where β = 1/kB T and constants have been gathered for convenience into
1.439ν̃
a=
T
The diatomic molecule Na2 (g) at 1000 K, for example, has a strong vibrational
resonance at 159.2 cm−1 . This leads to a = 1.439(159.2)/1000 = 0.229 and
1
Q vib = = 4.88
1 − e−0.229
P1: OTA/XYZ P2: ABC
Similarly, the rotational partition function Q rot , in spectroscopist’s terms is
1 0.695T
Q rot = =
hcBβ σB
where B is the rotational resonance frequency, again in units of cm−1 , and σ is

a “symmetry number” included to take into account the fact that a symmetrical
molecule has different configurations that are, from a geometric point of view, the
same. For example, the linear molecule Na–Na rotated by one half a rotation is
still just Na–Na, indistinguishable from the starting configuration. The difference
between distinguishable and indistinguishable configurations changes the statistics
of the problem. (The statistics of a poker game using a deck in which each ace has
an identical twin is not the same as the statistics using an honest deck.)
The rotational resonance frequency of interest for Na2 (g) is B = 0.1547 cm−1 .
Applying the spectroscopist’s formula to the Na–Na molecule with its symmetry
number of 2 gives
0.695T 0.695(1000)
Q rot = = = 2246
σB 2(0.1547)
This kind of calculation will be useful to us in Section 8.8. Further discussion of Q

determination is given in more detailed works (Nash, 2006; Maczek, 1998).
8.7 THE PARTITION FUNCTION FOR DIFFERENT MODES

OF MOTION
A molecule can absorb energy by increasing its translational motion (displacement

of the entire molecule), increasing its rotational motion, and increasing its vibrational
motion, as well as by electronic excitation. The total energy of a molecule is the sum
of energy contributions from all of its modes of motion:
ε = εtr + εrot + εvib + εel
In terms of its different modes of motion, the partition function is

Q= e−(εtr +εrot +εvib +εel )/kB T
i
Exponential sums can be written as products ea+b = ea eb , and so on, so Q can be

broken up into contributing partition functions, one for each mode of motion:

Q= e−(εtr )/kB T e−(εrot )/kB T e−(εvib )/kB T e−(εel )/kB T
i i i i
P1: OTA/XYZ P2: ABC
THE EQUILIBRIUM CONSTANT: A STATISTICAL APPROACH 117
This leads to
Q = Q tr Q rot Q vib Q el
where

Q tr = e−(εtr )/kB T , etc.
i
Thus, if one knows the energy absorption for each possible mode of motion or
excitation, one can write down the partition function and obtain all the thermodynamic
information for that system. In practice this may be a difficult task because systems
may not be weakly interacting; instead, they may interact in such a way that the
vibrational energy spacing for each bond in the molecule depends on its neighbors.
Other interactions may occur. Also, real chemical bonds are not truly harmonic and
their energy level spacings are equal only as an approximation.
8.8 THE EQUILIBRIUM CONSTANT: A STATISTICAL APPROACH
First taking into account the energy spacing between reactants and products and then
considering their partition functions, we can write K eq in terms of the product of
molecular partition functions Q
ξ
ε0 Qi
ln K eq = − + ln
RT NA
where ξ is the appropriate stoichiometric coefficient and N A is Avogadro’s num-

ber. Note the correspondence between the second term in this equation and the
entropy change of reaction r S ◦ as it appears in the classical van’t Hoff and related
equations.
Sodium metal can be vaporized at moderately high temperatures. The vapor
exists as diatomic sodium molecules Na2 (g) in equilibrium with atomic sodium
vapor Na(g).
Na2 (g) →
← 2Na(g)
Maczek (1998) has applied the statistical thermodynamic equation for K eq to several
reactions, including the dissociation of Na2 molecules in the gas phase at 1274 K. Not
all modes of motion apply in this case. For the product state, there is no partition func-
tion Q vib (Na) or Q rot (Na) because there can be no internuclear vibration or rotation
of atoms, which are essentially point masses. The reactant state consists of connected
masses separated in space for which vibration and rotation are quantum mechanically
P1: OTA/XYZ P2: ABC
allowed. Now, however, electronic excitation, while quantum mechanically permit-

ted, does not occur because the thermal energy at this temperature is insufficient to
drive the electrons out of their ground state. The equilibrium expression is somewhat
simpler than one might expect:
[Q tr (Na)Q el (Na)]2
K eq = e−ε0 /RT
Q tr (Na2 )Q rot (Na2 )Q vib (Na2 )
The first partition function we need is Q tr for translational motion. That simply
means the partition function for sodium molecules or atoms (or anything else) fly-
ing around and bouncing off the walls of some container. Imposition of the laws
of quantum mechanics brings about an astonishing phenomenon: Simply by being
confined to a container or “box,” the particles suddenly have allowed and forbidden
energy levels. Their energy is quantized. Energy levels of a particle in a box form
exactly the same kind of energy-level manifold we have been talking about except
that, unlike the harmonic oscillator example, they are not evenly spaced. Instead, they
go up according to the equation (Chapter 19)
n2h2
E=
8ma 2
The integers n = 1, 2, 3, . . . are called the quantum numbers, h = 6.626 × 10−34 J s

is Planck’s constant in joule seconds, m is the mass of the particle, and a is the length
of one edge of the box (taken to be cubic for simplicity).
When this energy restriction is placed on translational motion, it influences the
partition function for a cubic box of volume V = a 3 according to the equation
32 32
2π m 2π m
Q tr = a =
3
V
h2β h2β
where we have again used a convenience variable, ubiquitous in this field, β = 1/kB T .
Inverting terms in the parentheses, we obtain
− 32
h2β V V
Q tr = V = 32 ≡
3
2π m h β2
2π m
In the jargon of this field, the cube root of the denominator is called the thermal
wavelength and given the special symbol
:
12
h2β

=
2π m
P1: OTA/XYZ P2: ABC
COMPUTATIONAL STATISTICAL THERMODYNAMICS 119
The thermal wavelength has the units of length, m [or picometers (pm) in
atomic–molecular problems]. If we let the volume chosen be the molar volume
a 3 = Vm , then the volume per particle is Vm /N A , where N A is Avogadro’s number.
Now
is only a collection of constants times the inverse square root of the mass of
the particle and the temperature:
12 12 12 12
h2β 1 1 1

= =h =h
2π m 2π mkB T 2π m kB T
1
= 7.113 × 10−23 √
mT
The partition function is a measure of the accessibility of quantum states to

particles distributed within the system. Because the cube of the thermal wavelength
is inversely related to the partition function through the volume of the container in
which the particles are trapped,
gives us an idea of how big a container has to be for
its quantum states to be fully populated. For atoms and small molecules the volume

3 is of the order of 10−30 . Any imaginable laboratory container is very much larger
than this, so we conclude that all translational levels are accessible. Bear in mind
that this volume restriction relates to translational motion only and not to any other
mode of motion.
8.9 COMPUTATIONAL STATISTICAL THERMODYNAMICS
Further insight into this calculation can be gained from a computational solution
for K eq .
Several computer programs give calculated values for the thermodynamic func-
tions and the related partition functions. The functions are usually broken up into
their individual contributions as described above. An edited quantum mechanical
output of this kind for the sodium atom is given in Table 8.2. These values were taken
from a much larger output file from the program GAUSSIAN 03 C . Computed values
should be used with some caution because they often rely on approximations like the
harmonic oscillator approximation.
One can take the total partition functions, plug them into the equilibrium constant
quotient and multiply by the difference in ground state energies of the atoms relative
to the molecules to obtain K eq . The Q values chosen are from V = 0, the vibrational
ground state.
2
Q 2Na −ε0 /RT 3.44 × 108 −6.51
K eq = e = e = 18.5
Q Na2 9.531 × 1012
P1: OTA/XYZ P2: ABC
TABLE 8.2 Some Computed Partition Functions

for Molecular and Atomic Sodium.a
Na2 Molecules T = 1300 K
Q
Total Bot 0.873633D+13
Total V = 0 0.953147D+13 ←
Vib (Bot) 0.573265D+01
Vib (Bot) 1 0.573265D+01
Vib (V = 0) 0.625441D+01
Vib (V = 0) 1 0.625441D+01
Electronic 0.100000D+01
Translational 0.486489D+09
Rotational 0.313256D+04
Na atoms T = 1300
Q
Total Bot 0.344000D+09
Total V = 0 0.344000D+09 ←
Vib (Bot) 0.100000D+01
Vib (V = 0) 0.100000D+01
Electronic 0.200000D+01
Translational 0.172000D+09
Rotational 0.100000D+01
a Note that Q is unitless.
Example 8.1 The Thermal Wavelength

Calculate the thermal wavelength of sodium atoms at 1300 K.
Solution 8.1 The calculation for Na(g) is as follows:

1/2
h2β

=
2π m
Evaluation of the β constant,
β = 1/kB T = 1/1.38 × 10−23 · (1300) = 5.570 × 1019
leads to the thermal wavelength for Na(g) atoms at 1300 K.

1/2 1/2 1/2
5.570 × 1019 h 2 2.44 × 10−47 2.4 × 10−47

= = =
2π m 2π m 2π · 23.0 · 1.66 × 10−27
1/2
= 1.02 × 10−22 = 1.01 × 10−11 = 10.1 × 10−12 = 10.1 pm
P1: OTA/XYZ P2: ABC
For Na2 (g) molecules, a similar calculation gives a result that is not very different
from the result for atoms. At 1300 K,
(Na
= 7.14 pm. The dimensions for both
2 (g))
thermal wavelengths are picometers 10−12 m , so the volume
3 is of the order of
10−33 m3 . This dimension confirms our prior supposition that translational energy
levels are very close together. They can be regarded as a continuum if mathematically
convenient, and they are fully populated by a mole of particles.
Example 8.2 The Translational Equilibrium Constant

Given that the energy separation of ground states (Section 8.3) is ε0 =
70.5 kJ mol−1 (from spectroscopic measurements), calculate what the equilibrium
constant would be for translational motion only (ignoring all internal modes of
motion).
Solution 8.2 The equilibrium expression is
Q Na /NA2 −ε0 /RT

K eq = e
Q Na2 /NA
The exponents 2 and 1 arise from the stoichiometric coefficients ξ in the equilibrium
expression for the reaction written as a dissociation
( pNa )2
K eq =
pNa2
Expressing the partition functions in terms of the thermal wavelengths, we obtain

3Na2 Vm −ε0 /RT 3.64 × 10−34 0.082
K eq = e = −66
e−6.51

6Na NA 1.06 × 10 6.022 × 1023
= 4.67 × 107 e−6.51 = 6.95 × 104
The exponential value −ε0 /RT = −70.4/R (1300) = −6.51 comes from spec-
troscopic data. The molar volume is 0.0820 m3 mol−1 and e−70.4/RT = e−6.51 =
1.488 × 10−3 .
But things are not quite that simple. The Na2 molecule has internal modes of
motion, one for vibration along the molecular axis and one for rotation about its
center of mass. Both partition functions can be determined from spectroscopic data.
We have already seen how the value of Q vib arises from the resonance frequency for
vibrational motion. The quantized levels for rotation of Na2 (g) molecules are found
in essentially the same way. The two calculations give
Q vib = 6.254, Q rot = 3.132 × 103
The partition functions are unitless. One more factor comes into the calculation:
integer 4, which is the multiplicity of the sodium atom electronic structure.
P1: OTA/XYZ P2: ABC

When these two degrees of freedom and the Na multiplicity 22 are taken into
account, the equilibrium constant calculation at 1300 K is complete

1
K eq = 4 6.95 × 104 = 14.2
3.132 × 103 · (6.254)
which, considering the magnitude of the numbers revalued is in good agreement with
the previous calculation.
Problem 8.1
Find the equilibrium constant K eq at 300 K for a simple nondegenerate two-level
system
A(g) → B(g)
where the energy of level B, εB , is 1.25 kJmol−1 higher than εA . What are the
percentage occupations of levels A and B? What is the probability of selecting a
molecule from the lower level if the selection process is completely random and does
not favor either level?
Problem 8.2
Calculate the thermal wavelength of Na2 (g) molecules at 1000 K. Compare your
answer to the value 8.14 pm.
Problem 8.3
√
If we multiply the thermal wavelength of Na by 1/ 2, we get the thermal wavelength
of Na2 . Why is that?
Problem 8.4
Why is ν = c ν̃, where ν is the frequency of electromagnetic radiation, c is its speed
3.0 × 108 m s−1 , and ν̃ is the “wavenumber in spectroscopist’s terminology with
units of reciprocal seconds (s−1 )?
Problem 8.5
No doubt you have noticed that solid iodine I(s) produces a purple vapor upon heating
in a closed container. The vapor is diatomic I2 (g) which enters into an equilibrium
with atoms I(g) in a way analogous to Na2 (g) and Na(g) already discussed. Calculate
the thermal wavelength of molecular I2 (g) and I(g) atoms confined to a molar volume
at 1000 K.
P1: OTA/XYZ P2: ABC
Problem 8.6
Given that
3/2 3/2
2π m 2π m
Q tr = a =
3
V
h2β h2β
arises from space quantization in a container of volume V,
n2h2
E=
8ma 2
where does the 2π come from in the numerator?
Problem 8.7
Find the translational partition function for I2 (g) and the translational partition func-
tion for I(g).
Problem 8.8
Find the thermal wavelengths for I2 (g) and I(g). From these wavelengths and the
dissociation energy of 152.3 kJ mol−1 , find the equilibrium constant at 1000 K
ignoring internal modes of motion.
Problem 8.9
Refine the result of Problem 8.8 by taking into account the vibrational and rotational
motions of the I2 (g) molecule. The multiplicity factor is 4.
P1: OTA/XYZ P2: ABC
9
THE PHASE RULE
It is essential in science for experiments to be repeated, tested, and verified before

being accepted into the body of theory.1 To duplicate thermochemical experiments,
one must be able to duplicate the thermochemical system, and to do that we must know
how many of its infinite number of physical properties—mass m, energy, entropy, the
heat capacities C p and C V , refractive indices, and so on—must be specified.
A fundamental truth that we usually take for granted is that there is a certain
number of variables that completely describe any selected system. When we measure
some of them, we know all we can ever know about the system and we can, in
principle, calculate all the variables we haven’t measured from those that we have.
Which variables are they and how many of them are there? Must we know a substantial
portion of the infinite number of possibilities in order to define a thermochemical
system? If that were true, science as we know it would hardly be possible because we
could never really verify (or contradict) an experimental result by replication. This is
the problem that was solved by J. Willard Gibbs.
9.1 COMPONENTS, PHASES, AND DEGREES OF FREEDOM
Gibbs distinguished between components C of a system, which are chemically dis-

tinguishable, and the phases P of a system, which are physically distinct. Thus, the
1 Thereare exceptions. The science of cosmology offers interesting contradictions because we would be
hard pressed to duplicate the creation of the universe.

Copyright
124
P1: OTA/XYZ P2: ABC
COEXISTENCE CURVES 125
system ice and water has one component, identifiable as H2 O, and two phases, solid
and liquid. With his famous phase rule (see below), Gibbs specified the number
of degrees of freedom of thermochemical systems. Degrees of freedom are like the
independent variables of a system of algebraic equations or the basis vectors of a
complete set. Once a limited number of them have been specified, the others are no
longer free to take on arbitrary values. They are linear combinations of the set that is
already known. For this reason it is the number of independent variables that we get
from the phase rule, not their identity. We are free to choose which of the identifiable
quantities shall be taken as independent variables, but only up to the number specified
by the phase rule.
In the case of liquid water, one might argue that there are three components, H+ ,
OH− , and H2 O, or even four, including H3 O+ . We ignore these and similar forms
because they are obviously related to each other and are not independent. In pure
water, we know that [H+ ] = [OH− ] because the only source of the ions is dissociation:
H2 O → H+ + OH−
That makes it clear that if we measure

either one but only one, we know
the other.
When we do this, we find H+ = 10−7 ; therefore we know that OH− = 10−7 and
we can write a dissociation equilibrium constant for pure water2 :
+ − + 2 −7 2
H OH H 10
Kw = ≈ = = 10−14
[H2 O] [H2 O] 1
where the concentrations are expressed as moles per mole of solvent (water).
Pursuing these ideas just a little further, what we call “concentrations” in solution
chemistry are really activities that, as unitless numbers (ratios relative to a standard
state), give a unitless K w .
If we reformulate so as to include higher polymers such as H5 O+ 2 , and so on,
their concentrations are not independent either because dissociation constants exist
by which we could calculate their concentration. These calculations are, in principle,
possible, even though we may not have enough information to carry them out.
However you look at it, specifying one molar concentration specifies them all. There
is one component in a pure phase.
9.2 COEXISTENCE CURVES
The thermodynamic state of water vapor, H2 O(g), like that of any pure substance, has
two degrees of freedom. Any state function, including the Gibbs free energy, can be
expressed as a mathematical function of two independent variables, for example, µ =
f ( p, T ). We can also write E = f ( p, T ) or C p = f ( p, T ). In principle, equations

our conventional definition, pH ≡ − log H+ ; this leads us to say that an aqueous solution is neutral
2 By
when its pH = 7.
P1: OTA/XYZ P2: ABC
126 THE PHASE RULE
H2O(gas)
H2O(gas)
H2O(liquid)
FIGURE 9.1 Pure water in one phase (left) and two phases (right). At some temperature,
the phases coexist in equilibrium.
exist wherewith we can determine E and C p from a knowledge of p and T even

though we may not know exactly what these equations are. We could even write
p = f (C p , E). We are not restricted as to what the independent variables are, so long
as there are precisely two of them.
The number of degrees of freedom for a pure, one-phase system is easy. Identifying
some phases is easy as well.3 We can see the phase separation in mixtures of carbon
tetrachloride and water because they have different refractive indices. The two phases,
one as the top layer and one as the bottom layer, look different. The trick is to express
the behavior of a completely general system, which need not be pure, and may consist
of many chemical entities distributed over many phases.
If we choose water vapor as our example of a pure phase, examine the Gibbs free
energy as our state function, and take pressure and temperature as our independent
variables µ = f ( p, T ), the system as defined has a wide range of freedom. Within
reasonable limits, we can change p and T arbitrarily. The state of the system can
be represented as a point on a two-dimensional plot of p vs. T (or p vs. V, as in
Chapter 1). Any point represents the system in some state.
Sooner or later though, we are bound to exceed reasonable limits; and the water,
even in a closed container that previously contained only water vapor, is bound to
condense to produce some liquid water in a state of equilibrium with gaseous water
(Fig. 9.1)
We treat this physical equilibrium just as we do a chemical equilibrium. The
vaporization equilibrium is
→ H2 O(g)
H2 O(l) ←
The equilibrium constant is
pH2 O (g)
K eq = = pH2 O (g)
aH2 O (l)
where pure water is in its standard state, resulting in a denominator of aH2 O (l) ≡ 1.0
in the normal equilibrium expression. An equilibrium between pure liquid and pure
3 Identification
of some phases may not be easy. Solids can exist as phases that differ only in more subtle
ways such as heat capacity or molar volume.
P1: OTA/XYZ P2: ABC
COEXISTENCE CURVES 127
p
p*
T*
T
FIGURE 9.2 A liquid–vapor coexistence curve. Fixing the temperature at T * automatically
fixes the pressure at p* for coexisting phases of a pure substance.
gas means that their free energies must be equal:
µH2 O (l) = µH2 O (g)
An equation restricting the variables to a fixed ratio reduces the number of inde-
pendent variables by one. Now, by specifying either the temperature or the pressure,
the other variable is no longer free. There is still an infinite number of possible free
energies, but they are contiguous points on a line. Each specific T defines a coexisting
state at pressure p. The locus of points at which liquid and vapor can coexist is called
the coexistence curve. If T is changed by a small arbitrary amount, p automatically
adjusts to an appropriate value to maintain the equilibrium and stay on the curve. The
coexistence curve can now be completely described in two dimensions, p on the ver-
tical axis as a function of an independent variable T on the horizontal axis (Fig. 9.2).
The point representing the system is no longer free to move over a two-dimensional
p–V plane; it is restricted to the curve. The system has lost one degree of freedom.
The liquid–vapor coexistence curve is not unique, nor is its exponential shape.
Solids also exist in equilibrium with the vapor phase. That is why you can smell
solids like naphthalene. Solids have an exponential (or approximately exponential)
coexistence curve too, marking their equilibrium boundary with vapor, as seen at the
left of Fig. 9.3. Of course, there is also a coexistence curve between solids and liquids
(melting points), which normally has a positive slope as in Fig. 9.3. In the unusual
case of water, the slope of the solid-liquid coexistence line is negative which is why
you can ice skate.4 The three curves taken together on a p–T surface constitute a phase
diagram. As we shall see, phase diagrams are a very general way of representing all
manner of coexistence curves.
At the critical point (Section 2.4), there is no longer a distinction between a liquid
and a nonideal gas. Like the critical point, the triple point is unique to each pure
substance. It cannot be changed by altering external conditions p and T. It has no
degrees of freedom.
4 How can there be a connection between a physical chemistry coexistence curve and ice skating? Think
about it.
P1: OTA/XYZ P2: ABC
128 THE PHASE RULE
.
p
liquid
solid
vapor
T
FIGURE 9.3 A single-component phase diagram. The unusual solid–liquid coexistence
curve for water is shown as a dotted line. The terminus (•) of the curve on the right is
the critical point. The intersection of the three curves is the triple point.
9.3 THE CLAUSIUS–CLAPEYRON EQUATION
The Clapeyron equation for a phase transition is
dp trans H
=
dT T V
where trans H is the enthalpy change of the phase transition and V is the corre-
sponding volume change. It is customary to take both as molar quantities.
Clausius observed that the volume of a liquid in equilibrium with its vapor at
the boiling point is normally very much smaller than the volume of the vapor itself:
V (l) << V (g). Thus, taking the special case of vaporization of a liquid where V =
V (g) − V (l), he ignored the smaller volume V (l) and substituted the larger volume
V (g) for V . He then assumed that the vapor over the liquid is an ideal gas, which
permits the substitution
dp vap H vap H p
= =
dT T V (g) T RT
Rearranging, we obtain

dp vap H 1
= dT
p R T2
and taking the indefinite integral,

dp vap H 1
= dT
p R T2
P1: OTA/XYZ P2: ABC
PARTIAL MOLAR VOLUME 129
we get

vap H −1
ln p = + const.
R T
This result can be written
pvap = ae−vap H/RT = aeb(−1/T )
where a arises from the constant of integration. This is a form of the

Clausius–Clapeyron equation. If one can determine b in this exponential equation,
then one has vap H .
If an open container of liquid under an external pressure pext is heated to its
boiling point, bubbles form at the bottom of the container, each of which forms a
tiny closed “chamber” filled with pure vapor. That is why we observe that the normal
boiling points of pure liquids observed in an open container follow the same equation
(Clausius–Clapeyron) as that derived from the coexistence curve of the liquid in a
closed container.
9.4 PARTIAL MOLAR VOLUME
Prior to Gibbs, thermodynamics was largely about the transfer of heat in the process
of driving an engine. It was an age justly called the age of steam. Gibbs’s departure
was to focus on the transfer of matter. That is why Gibbs’s work is so important to
chemists. Our science is largely devoted to the transfer of matter from a reactant state
to a product state. In classical physical chemistry, we ask whether the transfer occurs
(thermodynamics) and, if it does, we want to know how long it takes (kinetics).
To answer the first of these questions, Gibbs used partial molar thermodynamic
state functions. To start out, we shall consider the volume of a system and we shall
restrict mixtures to two components. The volume of a system is easy to visualize,
and restricting the system to two components simplifies the arithmetic. Later we shall
release these restrictions.
The volume of an ideal two-component liquid solution at constant p and T depends
on how much of each component is present:
V = f (n 1 , n 2 )
◦
If the molar volume of pure component 1 is Vm1 and the molar volume of pure
◦
component 2 is Vm2 , then an ideal solution of equal
◦ molar amounts
of 1 and 2 will have
◦
a molar volume that is the average of the two Vm1 + Vm2 /2 (Fig. 9.4). Solutions of
other ratios of components 1 and 2 would be arrayed along a straight line connecting
◦ ◦
Vm1 to Vm2 , provided that no shrinkage or swelling takes place when 1 and 2 are mixed.
In general, for different ratios of the components, we obtain
◦ ◦
V = Vm1 n 1 + Vm2 n2
P1: OTA/XYZ P2: ABC
130 THE PHASE RULE
Vm2
Vm1
0 1
X2
FIGURE 9.4 Total volume of an ideal binary solution. X 2 is the mole fraction of compo-
nent 2.
In reality, life is not that simple. The total volume of the solution will not usually
be the sum of the molar volumes weighted by their relative amounts. The volume
actually occupied by the solute in the solution is called the partial molar volume,
which can be greater than or smaller than its molar volume in the pure state. The
volume of the solution will then be greater or smaller than the sum of its parts.
Addition of some potassium salts to one mole of pure water results in a volume
of solution that is even smaller than the initial volume of water (shrinkage occurs).
These three possibilities are shown in Fig. 9.5.
9.4.1 Generalization
We have been completely arbitrary in designating component 2 with amount n 2 as the
solute and n 1 as the number of moles of solvent, so we can switch designations just as
arbitrarily. Thus everything we have said about component 2 in a binary solution also
applies to component 1 treated as though it is the solute in solution with component 2.
In a solution of completely miscible liquids, it is conventional to take the lesser
V on2
V
n2
FIGURE 9.5 Volume increase (or decrease) upon adding small amounts of solute n 2 to pure
◦ ◦
solvent. Three cases are shown for Vn2 > Vn2 , Vn2 < Vn2 , and Vn2 < 0.
P1: OTA/XYZ P2: ABC
n2
FIGURE 9.6 Partial molar volume as the slope of V vs. n 2 . The lower line gives Vm2 as
n 2 → 0, and the upper tangent line gives Vm2 at a specific concentration n 2 = 0.
component as the solute and the greater component as the solvent. In many cases, the
choice is obvious, for example, in KCl solutions in water, KCl is clearly the solute.
Evidently, from Fig. 9.6, the partial molar volume of one component, which we
have chosen to call the solute, is the slope of one of the solid lines found by measuring
the volume increment upon adding small amounts of component 2 to large amounts
of component 1. In real solutions, these lines need not be straight; the partial molar
volume of the solute is found in the limit as dn2 → 0.
Recognizing Vm2 as the slope of the function of V vs. n 2 means that we have the
definition

∂V
Vm2 =
∂n 2 T, p,n 1
This slope can be found anywhere on the entire curve of experimentally measured
total volume vs. X 2 , where X n 2 = n 2 /(n 1 + n 2 ). Now Vm2 is defined as the volume
change found upon adding an infinitesimal amount of component 2 to a solution of
composition X 2 specified by a horizontal distance along the X 2 axis; for example,
X 2 = 0.20 in Fig. 9.7.
Vm2
Vm1
0 0.20 1
X2
FIGURE 9.7 Volume behavior of a nonideal binary solution. X 2 is the mole fraction of
component 2.
P1: OTA/XYZ P2: ABC
132 THE PHASE RULE
The definition of Vm2 should come as no surprise. It comes from the condition on
perfect differentials. In the case of the volume, we have

∂V
∂V ∂V
dV = dp + dT + dn i
∂p T,n i ,n j ∂T p,n i ,n j ni
∂n i T, p,n j
This generalization involving the sum

∂V
dn i
ni
∂n i T, p,n j
gets us away from the restriction that the solution be binary, a condition we imposed
at the beginning for simplicity. Now the concepts developed can be applied to any
number of components. Recalling that volume is a thermodynamic property, we have
V = f ( p, T, n 1 , n 2 , . . . , n j )
In view of the first two terms in the sum, the functions are sensitive to variations in
both pressure and temperature hence one or both may be held constant in the phase
diagrams discussed below. The variation of total volume with composition in curves
like Fig. 9.7 gives rise to the term excess volume as the volume above the straight
line expected from a simple sum of molar volumes in the pure state Vm◦ . The excess
volume can be negative leading to a nonideal curve below the straight line in Fig. 9.7.
The volume in a real binary system corresponds to the sum in which
◦ ◦
V = Vm1 n 1 + Vm2 n2
is replaced by
V = V̄m1 n 1 + V̄m2 n 2
where the molar volumes V̄m1 and V̄m2 are no longer volumes in the pure phase but
are partial molar volumes, unique to the ratio of n 1 and n 2 in the solution. In general,

∂V
V̄m i =
∂n i T, p,n j
Each partial molar volume must be determined experimentally. There are, of course,
simplified equations containing empirical constants that work more or less well for
real (nonideal) solutions just as there are for nonideal gases.
The greatest generalization in this field, however, is to recognize that nothing in
the arguments made is specific to volume alone. The general rule is that partial molar
P1: OTA/XYZ P2: ABC
quantities analogous to the partial molar volume exist for all thermodynamic state
variables. This rule gives us the partial molar energy,

∂U
Ūm i =
∂n i T, p,n j
the partial molar enthalpy,

∂H
H̄m i =
∂n i T, p,n j
the partial molar entropy,

∂S
S̄m i =
∂n i T, p,n j
and the partial molar Gibbs free energy

∂µ
µm i =
∂n i T, p,n j
This last partial molar quantity is so important that it is given a unique name and
symbol. It is called the Gibbs chemical potential µm i . It should be clear that in real
(nonideal) systems, each of these functions has a corresponding excess function; the
excess energy, the excess enthalpy, the excess entropy, and so on.
In generalizing thermodynamics to many-component systems, Gibbs brought
about an immense expansion of the scope of the subject. All of the classical ther-
modynamic equations apply to partial molar quantities as well. For example, by
analogy to

G
d
T
=H
1
d
T
for the Gibbs free energy of ideal components (Section 6.6), we have
µ
d
T = H̄ m
1
d
T
relating the partial molar Gibbs free energy and the partial molar enthalpy. H̄ m .
P1: OTA/XYZ P2: ABC
134 THE PHASE RULE
9.5 THE GIBBS PHASE RULE
Of the C mole fractions X i (i = 1, 2, 3, . . . , C) in a many-component one-phase mix-

ture, C − 1 are independent
because once you know C − 1 of them, you can get the
remaining one from X i = 1. The “last” concentration variable is not independent.
i
If there are P coexisting phases, there are C − 1 independent concentration variables
in each phase for a total of P(C − 1). However, all P phases must be at the same
chemical potential for equilibrium to exist. This makes P − 1 equations connecting
each component in each phase,5 and there are C components for a total of C(P − 1)
equations. The number of degrees of freedom is equal to the total number of inde-
pendent variables P(C − 1) in the system minus the number of equations C(P − 1)
connecting them. In the most general case of many components that may break up
into many phases, we have
f = P(C − 1) − C(P − 1) = PC − P − C P + C = C − P
Adding the two remaining degrees of freedom for the system, which might be p and
T, we have the celebrated Gibbs phase rule
f =C − P +2
A problem of potentially great complexity has been reduced to simple terms. The
number of degrees of freedom f of the system is equal to the number of components
minus the number of phases plus 2.
9.6 TWO-COMPONENT PHASE DIAGRAMS
Two-component phase diagrams can be complicated. Without the phase rule, they
would seem to have a bewildering array of unrelated behaviors. Many two-component
phase diagrams, however, can be broken down into combinations of three simple
diagrams that we shall label I, II, and III here. Two-component phase diagrams are
characterized by one more degree of freedom than pure one-component diagrams
because of the added composition variable
f =C − P +2=2− P +2=4− P
We can express composition in many ways, but the mole fraction of a two-component
system is most convenient for our purposes.
nB
XB =
nA + nB
5 Forexample, µ A = µ B = µC connects three variables but has only two equal signs, hence only two
equations.
P1: OTA/XYZ P2: ABC
TWO-COMPONENT PHASE DIAGRAMS 135
The mole fraction of a two-component system runs from 0 to 1.0 (Fig. 9.4) where n B
is the number of moles of component B and n A + n B is the total number of moles of
both components.
9.6.1 Type 1
In a two-component one-phase Type I diagram, pressure is taken as constant. Setting
p = const reduces the number of degrees of freedom by 1: f = 3 − P = 2 for
1 phase. (Please do not confuse p and P.) The system can be represented in two
dimensions as in Fig. 9.8. If one phase is present, the system can exist at a temperature
and mole fraction represented by any point (T, X B ) in either area above or below the
coexistance curves. The upper and lower curves represent the locus of mole fractions
of the vapor (upper) and liquid (lower), respectively, in equilibrium with one another
at a temperature specified on the vertical axis. A system at any temperature and mole
fraction between the coexistence curves splits into two phases, a vapor (upper curve)
and a liquid (lower curve).
When there are two phases in equilibrium (liquid and vapor), the number of
degrees of freedom is further reduced ( f = 3 − P = 3 − 2 = 1); and specifying X B
of the system, as the mole fraction in the liquid phase or the mole fraction in the
vapor, automatically specifies the temperature on one of the two coexistence curves.
These two temperatures will not be the same. The lower curve in Fig. 9.8 is the locus
of points at which a trace of vapor is in equilibrium with liquid. The upper curve in
Fig. 9.8 is the locus of points at which a trace of liquid is in equilibrium with vapor.
The horizontal lines connecting them represent the l → v or v → l phase change at
any specific temperature, T (K).
Fractional distillation is one of our most important laboratory and industrial meth-
ods of chemical purification. The separation between the upper and lower curves in
a Type I diagram makes fractional distillation possible. At any temperature, the mole
fraction of component B in the mixture can be read along the horizontal in Fig. 9.8.
As seen from the figure, the mole fractions are different for coexisting liquid and
vapor phases except at the end points of the curves. If we allow a liquid mixture of
T (K)
XB
FIGURE 9.8 A Type I phase diagram. Liquid–vapor equilibriums are expressed by each of
the three horizontal lines.
P1: OTA/XYZ P2: ABC
136 THE PHASE RULE
A and B to come to equilibrium with its vapor and then separate and condense the
vapor, we shall have two solutions, one richer in A than the original solution and one
richer in B. Three such equilibrations are represented in Fig. 9.8 as horizontal lines at
different temperatures T, between the two T vs. X B curves. If we start with a solution
of A and B, well to the right of the figure, and allow it to come to equilibrium with
its vapor, the composition of the vapor is given by the leftmost terminus of the top
horizontal. Now separate that vapor from the original solution and condense it. The
new liquid is richer in A than the original solution was.
A second equilibration and separation gives a solution still richer in A according
to the middle horizontal in Fig. 9.8. A third repetition yields a vapor still richer in
A at the lowest temperature. Now the purified concentration is given by the leftmost
terminus of the three equilibration steps diagrammed in Fig. 9.8.
Of course, real distillations are not carried out by such a laborious stepwise process
of equilibration and condensation. Nevertheless, a process very like this, of continual
vaporization equilibration followed by condensation, does take place in the distillation
columns we actually use. Real distilling columns may range in height from a hand’s
breadth for use in the laboratory to columns several stories tall used in the petroleum
industry. By comparing the composition of the input with that of the effluent, the
number of theoretical plates of a “still” can be calculated. Within normal practical
considerations, the more theoretical plates, the better the still.
9.6.2 Type II
Type II phase diagrams describe liquid–liquid systems in which the components are
completely miscible at some temperatures but undergo phase separation at others.
Systems with a composition and temperature above the dome-shaped coexistence
curve in Fig. 9.9 are completely miscible. Those with a temperature and composition
below the curve split into two phases. The transition can be brought about by cooling
a miscible solution so that the temperature drops along a vertical and enters the two-
phase zone or it can be brought about at constant temperature by adding one or the
T,K
A B
XB
FIGURE 9.9 A Type II phase diagram. Solutions having a composition and temperature on
the line split into an A-rich phase and a B-rich phase. The horizontal tie line is unique at each
temperature.
P1: OTA/XYZ P2: ABC
COMPOUND PHASE DIAGRAMS 137
other of the components to a miscible solution until the coexistence curve is crossed.
The transition from a clear miscible system to an opaque two-phase emulsion can
be quite dramatic in some cases, making it look as though water has turned into
milk. Like Type I diagrams, the number of degrees of freedom is 3. We usually hold
p = const at 1 atm so that the rest of the phase behavior can be represented in two
dimensions. On the dome-shaped coexistence curve, there is 1 degree of freedom.
The horizontal under the dome is called a tie line. The two intersections of the tie
line with the coexistence curve give the composition of the two coexisting phases.
The dome need not be symmetrical. Quite a variety of shapes are possible, including
one that is closed at the top and at the bottom, forming a closed irregular oval that is
essentially an island of immiscibility in a sea of miscibility. The coexistence curve
can approach the verticals representing pure A and pure B quite closely, giving rise
to the folk saying “oil and water don’t mix.” Actually, they do mix but one phase
is overwhelmingly oil-rich and the other is overwhelmingly water-rich. That is why
one does not wish to drink water that has come into contact with oil or gasoline.
9.6.3 Type III

Type III solid–liquid phase diagrams are familiar as having a eutectic point. The locus
of melting points of mixtures of A and B vs. XB follows the two curves in Fig. 9.10.
In general, mixtures of the two components have a lower melting point than either
component, A or B, alone. Ordinary electrical solder is a eutectic mixture of lead and
tin having a melting point that is lower than either pure Pb or pure Sn. We rely on
the fact that the melting temperature of a pure compound is higher than that of an
impure (mixed) sample of the same compound according to Fig. 9.10. Generations
of pre-medical students have been judged partly on the basis of the melting point
behavior of the compounds they prepare in the organic chemistry laboratory.
9.7 COMPOUND PHASE DIAGRAMS
Many complicated phase diagrams can be read quite easily as two or more simple-
phase diagrams stuck together. The principle will be illustrated using a compound
T,K
A XB B
FIGURE 9.10 A Type III phase diagram. The low melting mixture is called a eutectic
mixture.
P1: OTA/XYZ P2: ABC
138 THE PHASE RULE
T,K
.
A XB B
FIGURE 9.11 A compound phase diagram with a low boiling azeotrope.
phase diagram with an azeotrope. A phase diagram like that in Fig. 9.11 is, quite
evidently, a combination of two Type I phase diagrams, one extending slightly beyond
the midpoint of the composition axis and one somewhat shorter curve. Drawing
isotherms as in Fig. 9.11 soon convinces us that the mixture cannot be separated by
conventional fractional distillation. Even the best still, starting with a mixture to the
right of the azeotropic point (•), will produce component B and azeotrope, whereas
starting to the left of the point will produce pure A and azeotrope. One does not get
pure A and pure B by either of these methods as one did by distilling mixtures of
components having a simple Type I phase diagram. If the curve in Fig. 9.11 is turned
upside down, the azeotrope is said to be high boiling. Ethanol and water form a high
boiling azeotrope at about 95% ethanol, which is why the common laboratory ethanol
is only 95% pure while most reagent grade solvents are 99+%.
9.8 TERNARY PHASE DIAGRAMS
Any mixture of three components can be represented within an equilateral triangle.

All points within the triangle represent one and only one of the infinitely many possi-
ble mixtures. Because there are now three composition variables f = C − P + 2 =
3 − P + 2, both T and p are held constant for the three-component triangular rep-
resentation on a two-dimensional surface. If only one phase is present, all solutions
are permitted and the composition can take any point in the triangular 2-space. When
two phases are present, the composition is restricted to the locus of points on the
coexistence curve. Like the Type II phase diagram, many three-component phase
diagrams are known having a dome-shaped coexistence curve. They are widely used
for characterizing and selecting solvents used in industrial processes. Intersections of
tie lines with the coexistence curve indicate the composition of phases in two-phase
systems formed when the total composition point is below the coexistence curve.
In Fig. 9.12, solvents B and C are nearly completely immiscible, but addition of sol-
vent A brings the composition of the two coexisting phases, the termini of the tie lines,
closer together until finally there are ternary solutions of A, B, and C that are homoge-
neous. These solutions correspond to ternary mixtures above the coexistence dome.
P1: OTA/XYZ P2: ABC
B C
FIGURE 9.12 A ternary phase diagram with a tie line. At constant p and T, f = 3 −
1 = 2 within the two-dimensional triangular surface, but f = 3 − 2 = 1 on the coexistence
curve. In general, the area under the curve is smaller at higher temperatures. Numerous, more
complicated forms are known.
Addition of A may cause a milky suspension of immiscible phases to suddenly

clear. Conversely, addition of C to a clear stirred solution of A and B may cause it to
go milky at some concentration. The transition is sudden and sharp and can be used
as the end point of a phase titration.
Example 9.1 The Enthalpy of Vaporization of H2 O

Handbook values for the vapor pressures of water between 273 and 373 K are
available. Plot pvap vs. −1/T for several temperatures in the vicinity of 298 K and
use a commercial curve fitting routine to determine the enthalpy change for the
vaporization of water at that temperature.
Solution 9.1 The handbook gives pvap at temperatures from 273 to 373 K. We
have selected six values symmetrically grouped around 298 K and presented them as
1/T K −1 in Table 9.1.
TABLE 9.1 Negative Inverse Temperatures
and Vapor Pressures for Water.
−1/T pvap
−3.6600e−3 0.6110
−3.5300e−3 1.2300
−3.4100e−3 2.3400
−3.3000e−3 4.2500
−3.1900e−3 7.3800
−3.1000e−3 12.3400
The curve of pvap vs. −1/T for water is shown in Fig. 9.13.
P1: OTA/XYZ P2: ABC
140 THE PHASE RULE
14
12
10
Vapor Pressure (kPa)
0
-0.0037 -0.0036 -0.0035 -0.0034 -0.0033 -0.0032 -0.0031 -0.0030
-1/T (K-1)
FIGURE 9.13 The liquid–vapor coexistence curve of water leading to vap H (H2 O) =
44.90 kJ mol−1 .
The vapor pressures over pure water were selected at 10 K intervals

from 273 to 323 K, plotted, and submitted to a standard curve-fitting rou-
tine (SigmaPlot 11.0). A two-parameter exponential curve fit of pvap vs.
−1/T (Statistics → Nonlinear → Regression Wizard → Ex-
ponential Growth → Single, 2-Parameter) gave the empirical con-
stant b (Output b 5400.6916) K, where the unit of the empirical constant is kelvins
in order to make the exponent −b/T unitless. The curve fit is good, having a mean
residual (difference between calculated and experimental values) of 1 part per thou-
sand or 0.1% over a range of about 12 kPa. We have the relationship between the
parameter b and vap H :
vap H
b=
R
which gives us

vap H = R(b) = 8.314 5.401 × 103 = 44.90 × 103 = 44.90 kJ mol−1
This value is “averaged out” by the curve fitting technique over the temperatures
symmetrically distributed around 298 K. The handbook value of vap H of water at
298 K is 43.99 kJ mol−1 .
Example 9.2 Ternary Phase Diagrams

The ternary phase diagram ABC in which A is completely miscible with B and C but
B and C are only partially miscible in each other looks like a combination of Figs.
9.9 and 9.12 except that the tie lines are not horizontal.
P1: OTA/XYZ P2: ABC
B C
FIGURE 9.14 A ternary phase diagram in which B and C are partially miscible. Figure 9.13
might approach this form at higher temperatures.
Starting with a solution of 0.5 mol of A and 0.5 mol of B at a constant temperature
corresponding to that of phase diagram Fig. 9.14, we add component C in small
portions. What happens?
Solution 9.2 The phase behavior is somewhat complicated. At the first few small
increments of component C, a clear homogeneous solution of A, B, and C results,
corresponding to points on a straight line from the midpoint of axis AB in the
direction of apex C. Soon the coexistence curve is crossed and the solution splits into
two phases. The overall composition continues along the straight line toward C as
increments are added, but the system now consists of two phases corresponding to the
end points of the tie line. At first, a minute amount of the second phase appears; but
further along in the addition process, substantial amounts of both phases are present.
Their compositions always correspond to the end points of the tie line as it intersects
the coexistence curve.
There comes a time when the short end of the tie line approaches the overall
composition line. The amount of each phase is in the inverse ratio of the length
of the tie line cut off by the overall composition line such that the AC-rich phase
predominates over the AB rich phase. Ultimately the coexistence curve is crossed
again and the solution clears, its composition corresponding to the points on the lower
right of the overall composition curve approaching the C apex.
Problem 9.1
(a) How many components are there in a dilute solution of sodium acetate NaAc
in water?
(b) A drop of HCl is added to the solution in part a. Now the anions of the weak
acid Ac− and the strong acid Cl− are competing for the protons H+ . How
many components are there in this system of Na+ , Ac− , H+ , Cl− , H3 O+ , and
a minimal concentration of OH− ?
Problem 9.2
(a) What is the effect of a decrease in atmospheric pressure on the freezing and
boiling points of water?
P1: OTA/XYZ P2: ABC
142 THE PHASE RULE
(b) What is the effect of a decrease in atmospheric pressure on the freezing and
boiling points of benzene?
(c) Describe the behavior of a system that is carried along a horizontal line above
the critical point of the phase diagram for water (Fig. 9.3).
(d) Describe the behavior of a system that is carried along a horizontal line below
the critical point, but above the triple point of the phase diagram for water
(Fig. 9.3).
(e) Describe the behavior of a system that is carried along a horizontal line below
the triple point of the phase diagram for benzene (solid lines Fig. 9.3).
Problem 9.3
Describe the behavior of the system with a mole fraction 0.75 well above the dome-
shaped coexistence curve in Fig. 9.9 as the temperature is slowly decreased.
Problem 9.4
Describe the behavior of the system with a mole fraction 0.25 well above the coex-
istence curve in Fig. 9.11 as the temperature is slowly decreased.
Problem 9.5
The vapor pressure of liquid benzene is given by
4110
ln Pvap = − + 18.33
T
in the approximate location of the triple point (McQuarrie and Simon, 1997). The
sublimation pressure of solid benzene is given by
5319
ln Psub = − + 22.67
T
What is the triple point for benzene?
Problem 9.6
Sketch a binary phase diagram resembling Fig. 9.10.
(a) Locate the points corresponding to pure components. Call them components
X and Y.
(b) Locate a point corresponding to an equimolar mixture of X and Y.
(c) Locate a point corresponding to a mixture of 20% X and 80% Y.
(d) Locate the point corresponding to the compound X2 Y.
P1: OTA/XYZ P2: ABC
Problem 9.7
Find an area with two degrees of freedom, a coexistence curve with one degree of
freedom, and a point with no degrees of freedom on the phase diagram (Fig. 9.10).
Problem 9.8
Sketch a triangular phase diagram resembling Fig. 9.12.
(a) Locate the point corresponding to a binary solution containing components in

equal amounts.
(b) Locate the point corresponding (approximately) to a solution with three com-
ponents in the ratio 45%, 45%, 10%.
(c) Locate the point corresponding to 33.3% for each of the three components.
Problem 9.9
Sketch the three-component phase diagram for a system in which all three of the
components are miscible (soluble) in all proportions X in Y, X in Z, and Y in Z, but
phase separation takes place for some of the ternary mixtures XYZ.
P1: OTA/XYZ P2: ABC
10
CHEMICAL KINETICS
In general, chemical thermodynamics is concerned with where a chemical system

starts and where it ends up, while chemical kinetics is concerned with how long
it takes it to get there. Many analytical techniques exist for following the rate of
change of the concentration of the reactants or products of a chemical reaction, but
single atom counting is not usually done. An exception is the class of radiochemical
reactions for which counters exist to register, count, and record in computer memory
every decay particle produced. For example, the decay of radium gives α particles
(helium nuclei) and radon 222
86 Rn:
226
88 Ra → 222
86 Rn + 2 He
4
Each click of the counter registers one α particle, hence one decay. Here radium is
the radioactive element, and both radon and helium are products. Radon is referred to
as the daughter of the parent element radium. Radon is itself a radioactive gas, and it
decays to polonium and further along a radioactive series ending eventually as lead
206
82 Pb. Radiochemical reactions show the statistical nature of rate laws very clearly,
For that reason, they are a good introduction to the entire field.
10.1 FIRST-ORDER KINETIC RATE LAWS
Suppose you have a sample consisting of a hundred billion atoms of radioactive

element X. By means of a suitable counter, you observe an average of one radioactive

Copyright
144
P1: OTA/XYZ P2: ABC
FIRST-ORDER KINETIC RATE LAWS 145
decay per second. You have observed a rate of decay −d X /dt in the units (number
of atoms)/time, usually given as a frequency (s−1 ). Decay causes a decrease in the
number of atoms X, which accounts for the minus sign in front of the rate. If you
take 200 billion atoms, you find twice as many radioactive events. That only stands
to reason. Radioactive decay is a random phenomenon. Observing twice as many
atoms, one expects to see twice as many events in a specified time interval. Taking k
times as many atoms gives k times as many radioactive events: rate = kX. Now you
have a rate, a rate law, and a rate equation. The rate of disintegration is
dX
− = kX
dt
The rate of product production is equal but opposite in sign to the rate of decay.
Provided that the daughter is not radioactive, starting with a number of atoms X 0 at
a time t = t0 = 0 (an arbitrary time that you choose to start watching your counter),
the average count rate gradually decreases with time. If you watch long enough,1 the
count rate will be reduced to half of what it was at t0 . This is called the half-time of
the radioactive element.
The rate equation can be rearranged to give
dX
= −k dt
X
The ratio dX/X being unitless, −k dt must be unitless as well, so the unit of k is s−1
because t is in seconds. Integration between t0 and t is straightforward:
t t
1
d X = −k dt
t0 X t0
gives
X
ln = −k(t − t0 )
X0
In exponential form
X
= e−k(t−t0 )
X0
X 0 is the initial concentration (a constant) and t0 is usually set to zero so that

X 0 = X (0) and
X = X 0 e−kt
1 About 1600 years in the case of radium.

P1: OTA/XYZ P2: ABC
146 CHEMICAL KINETICS
Some chemical reactions follow the same mathematical laws as radioactive decay.
The rate equation in these cases is a first-order differential equation, so the reactions
that follow it are called first-order reactions. A case in point is the decomposition of
a dilute solution of hydrogen peroxide H2 O2 in contact with finely divided platinum
catalyst. The probability arguments given above apply equally to radioactive decay
and to first-order chemical reactions. In the case of the decomposition of a dilute
solution of hydrogen peroxide in contact with Pt, the reaction is first order with a
half-time of about 11 minutes.
The half-time can be determined for a first-order reaction in a simple and illus-
trative way. In the logarithmic form of the first-order equation above, set X = 12 X 0 ,
which is the definition of the half-time. Now
1
X
ln = ln 2 = −kt 12
X0 1
1
1
kt 21 = − ln 2
= ln 1
= ln 2 = 0.693
1 2
and
k = 0.693/t 12
The half-time is found by drawing a horizontal halfway between the base line and
X 0 . The horizontal intersects the experimental curve at t 1 . Notice that the con-
2
centration does not enter into calculation of the rate constant. Knowing either the
half-time or the rate constant gives you the other one. Generally the half-time is
easier to observe.
The exponential decay of X with t is seen in Fig. 10.1. The half-time is seen there
as well. The linear decrease of ln X with t is seen in Fig. 10.2.
The linear plot of ln X is usually used to obtain the rate constant because it
averages many experimental points and is, presumably, more reliable than any one
X0
X
t 1/2 t
FIGURE 10.1 First-order radioactive decay.

P1: OTA/XYZ P2: ABC
SECOND-ORDER REACTIONS 147
ln X 0
ln X
slope = – k
FIGURE 10.2 Logarithmic decay of a radioactive element.
measurement. From the form of the equation, ln X may be taken as the dependent
variable while t may be as independent, whereupon the slope is equivalent to minus
the rate constant. The constant of integration can be evaluated by setting t = 0,
C = − ln X 0
enabling us to write
− ln X = kt − ln X 0
which is equivalent to the forms obtained by integration between limits:
X
ln = −kt
X0
10.2 SECOND-ORDER REACTIONS
From the simple physical picture of a reacting mixture of reagents of two different
kinds, say A and B, in which molecules collide to produce one or more products, it is
logical to suppose that if either A or B is in short supply, the reaction must go slowly
because fewer A–B collisions take place even though the total number of collisions
may still be large. In the form of a rate law, the rate is proportional to both A and B
concentrations, hence it is proportional to the AB product2 :
dA dB
Rate = − =− = kAB
dt dt
2 The identity of the molecules is given in regular type AB, but the concentration variable referring to the
different species is italic AB.
P1: OTA/XYZ P2: ABC
If the reactants are mixed so that the initial concentrations A and B are equal, they will
remain equal throughout, because for each molecule of one that reacts, a molecule of
the other has also reacted. As a result, AB = A2 throughout, giving the special case
of the rate equation
dA
− = kAB = kA2
dt
and
dA
= −k dt
A2
Integrating

1
dA = −k dt
A2
gives
1
− = −kt + C
A
To evaluate the constant of integration, let t = 0. The concentration A is its initial

1
value A0 , so − = C and
A0
1 1
= kt +
A A0
which leads to
A0 − A
= kt
A0 A
The relation between t1/2 and k for this restricted form of the second-order reaction
is kt1/2 A0 = 1.
An example of second-order kinetics in which the rate law −dA/dt = kAB = kA2
applies is the thermal decomposition of NO2 :
2NO2 (g) → 2NO(g) + O2 (g)
The rate is controlled by collisions of NO2 molecules with other NO2 molecules,
hence the stipulation A = B throughout is met. Although in this case the order of
the reaction is the same as the stoichiometric coefficient, that need not be the case as
exemplified by the thermal decomposition of N2 O5 :
2N2 O5 (g) → 4NO2 (g) + O2 (g)
which is first order.

P1: OTA/XYZ P2: ABC
OTHER REACTION ORDERS 149
10.3 OTHER REACTION ORDERS
In general, the order of a chemical reaction is the sum of the exponents in the rate
equation a + b + c + · · ·:
dA
− = kAa B b C c . . .
dt
The order may be fractional or even decimal, but it is not a large number. Zero- and
third-order reactions exist, but they are not common. A zero-order reaction is one in
which the rate does not vary with time:
dA
=k
dt
This rate law might be followed by a reaction depending on some constant like the
surface area of a catalyst on which the reaction is taking place.
There are systematic mathematical methods for determining the order of a reaction
from a data set of concentration vs. time even when the order is not simple.
10.3.1 Mathematical Interlude: The Laplace Transform

The Laplace transform is used to simplify differential equations. For the function
F(t) where t is the time and F(t) is zero at t < 0, the Laplace transform f (s) of F(t)
is defined as
∞
f (s) = L [F(t)] = e−st F(t) dt
0
The inverse Laplace transform of f (s) is F(t). The Laplace transform L [F(t)] is a
function of a function; it is called a functional. Extensive tables of Laplace transforms
and inverse Laplace transforms are available (CRC Handbook of Chemistry and
Physics 2008–2009, 89th ed.).
First we shall convert the derivative of the function F(t) to a simple algebraic
form. By definition,
∞
dF(t) dF(t)
L = e−st dt
dt 0 dt
Integration by parts gives

∞
dF(t)
−st ∞
L = F(t)e 0
− F(t)d(e−st )
dt 0
∞
= −F(t = 0) + s F(t)e−st dt
0
= −F(t = 0) + s f (s)
P1: OTA/XYZ P2: ABC
10.3.2 Back to Kinetics: Sequential Reactions

In this section we shall obtain the concentration B in the a sequence of two first-order
reactions:
k1 k2
A→B→C
The differential equations are
dA
= −k1 A
dt
dB
= k1 A − k2 B
dt
dC
= k2 B
dt
From the first equation, we have
A = A0 e−k1 t , B0 = C0 = 0
where the subscripted 0 indicates the initial concentration. Substituting this result
into the equation for B(t), we obtain
dB(t)
= k1 A0 e−k1 t − k2 B(t)
dt
Now take the Laplace transform of both sides and solve for b(s), the transform of
B(t):
k1 A0
sb(s) − B(t = 0) = − k2 b(s)
s + k1
where B(t = 0) = B0 = 0, so the remaining terms are
k1 A0
sb(s) + k2 b(s) =
s + k1
k1 A0
b(s) (s + k2 ) =
s + k1
k1 A0 1 1 1
b(s) = = k1 A0
s + k1 (s + k2 ) (s + k1 ) (s + k2 )
P1: OTA/XYZ P2: ABC
which is our solution for the Laplace transform of both sides of the rate equation. Now
take the inverse Laplace transform. The Laplace transform b(s) inverse transforms,
1 1
of course, to B(t) and the inverse transform of is
(s + k1 ) (s + k2 )
1 1 inv transform 1 −k1 t

−→ e − e−k2 t
(s + k1 ) (s + k2 ) k1 − k2
so
1 −k1 t
B(t) = k1 A0 e − e−k2 t
k1 − k2
10.3.3 Reversible Reactions

No reaction goes to completion. Reactions we call “complete” are those in which
the concentration of, for example, reactant A is negligible or not detectable in the
reaction
A→B
This is the case when the standard free energy of A is much larger than that of B, so
that the reaction is accompanied by a significant decrease in free energy and leads to
B more stable than A.
In many cases, something closer to a free energy balance exists. We write the
reaction as an equilibrium characterized by an equilibrium constant K eq (Chapter 7).
Kinetic rate constants k f and kb for forward and back reactions compete to establish
the equilibrium balance:
kf
→B
A←
kb
kf B
= K eq =
kb A
When A and B are mixed in concentrations A and B that are not the equilibrium
concentrations, either the forward or the back reaction is faster than the other and the
system is displaced so as to approach equilibrium.
The rate of depletion of A,

dA
− = k f A − kb B
dt obs
is smaller than the rate that would be found if the reaction went to completion or if
we had some mechanism for removal of B immediately as it is formed. The rate of
P1: OTA/XYZ P2: ABC
approach to equilibrium is always less than either the forward reaction or the back
reaction because k f and kb work in opposition to one another.
For simple reversible systems,
→B
A←
→C
A+B←
→C + D
A+B←
etc.
equations can be worked out that relate the observed rate constant kobs for the approach
to equilibrium to k f and kb , the elementary reactions that contribute to it (Metiu,
2006).
The concept of equilibrium as the result of opposite forward and back reactions is
called the principle of detailed balance; for example, at equilibrium,
k f A − kb B = 0
The equations implied by the principle of detailed balance rest on the assumption that
the concentrations of reactant and product vary in the simplest possible way. This
assumption is often violated and must never be taken for granted.
Some enzyme-catalyzed reactions, including the famous Michaelis–Menten
(Houston, 2001) mechanism, are examples of more complicated reaction mecha-
nisms. Despite their complexity, they can often be broken down into elementary steps
and equilibriums. The kinetics of complex reactions can sometimes be simplified by
regarding one component of the reaction as a constant during part of the chemical
process. This is the steady-state approximation (Metiu, 2006).
An especially important class of reaction mechanisms is that of the chain reactions
in which one molecular event leads to many, possibly very many, products. The classic
example for chemists is production of HBr from the elements:
H2 + Br2 → 2HBr
for which we might guess the rate law to be
1 d [HBr]
= k [H2 ] [Br2 ] wrong
2 dt
but that guess would be wrong. Instead, the rate law is
1 d [HBr]
= k [H2 ] [Br2 ]1/2 right
2 dt
P1: OTA/XYZ P2: ABC
An acceptable mechanism for this reaction depends upon three elementary reaction
types, initiation, chain, and termination:
k1
Initiation Br2 + M → 2Br · + M
k2
Chain Br· + H2 → HBr + H·
k3
Chain H· + Br2 → HBr + Br·
k−1
Termination Br· + Br· → Br2
The point of this proposed mechanism is that the cyclic chain steps can continue
indefinitely because the Br· free radicals used in the first chain step are produced in the
second chain step. Initiation can be by a rather unlikely process such as collision with
a high-energy molecule M in the first step or impact of a photon from a flame or spark
as in chain explosions. Though initiation may not occur very often, it can have a large
effect. A hydrogen–oxygen explosion is an example. A very small spark can cause
a very large explosion. The HBr chain, though it may yield many molecules for one
initiation step, does not, of course, go on forever. Some step such as recombination of
the Br· free radicals terminates the chain or we run out of Br2 or H2 and the reaction
stops.
If we assume that initiation and termination are rare events by comparison to the
chain steps, for every Br· used up in the first step of the chain, one is produced in
the second chain step, so the amount of free radical present at any time during the
reaction is constant:
d [Br·]
=0
dt
This is an example of the steady-state hypothesis. Making this assumption, a few
lines of algebra (Houston, 2006) lead to the correct rate equation given above:
1 d [HBr]
= k [H2 ] [Br2 ]1/2
2 dt
If, by a more complicated mechanism, two or more reactive species are produced on
each step, the amount of reactive species may increase rapidly. For example, if two
reactive species are produced at each step, the geometric series, 1, 2, 4, 8, 16, . . . is
followed. If the chain steps are fast, this kind of mechanism takes place with explosive
violence. This type of mechanism is characteristic of nuclear fission bomb reactions.
Free radical and (controlled) chain reactions are also characteristic of some bio-
chemical reactions, and they can behave in ways that are beneficial or detrimen-
tal to the organism. A hydroperoxide RCOO· free radical chain may destroy lipid
molecules in a cell wall by lipid peroxidation with disastrous results for the cell.
Free radicals derived from tocopherol (vitamin E) antioxidants act as sweepers in the
blood, interfering with chain propagation, thereby slowing or preventing cell degra-
dation. Free radicals have been detected in rapidly multiplying natal or neonatal cells
P1: OTA/XYZ P2: ABC
(beneficial) but they are also found in cancers undergoing uncontrolled reproduction.
The biological role of free radicals is complex and not completely understood.
10.4 EXPERIMENTAL DETERMINATION OF THE RATE EQUATION
For some reactions, determination of the rate law and the rate constant may be as
simple as mixing the reagents, titrating a sample of one of the reactants or antic-
ipated products, waiting a while, titrating another sample, and so on, to obtain a
series of concentrations A, each at a time t. Treating this data set by an appropriate
mathematical procedure gives the rate law and k.
Some of the most interesting contemporary research in kinetics involves tracking
reaction components or intermediates by physical means, perhaps because they are
of fleeting existence, being present in the reacting system for a second or less.
A straightforward approach to fairly fast reactions involves use of a stopped-flow
reactor. One experimental design consists of two syringes connected to a mixer.
To initiate the reaction, the syringes are driven simultaneously so that reactants A
and B flow into the mixer, are mixed thoroughly, and flow out of the mixer into a
tube where the reaction mixture is monitored by a spectrophotometer. The output
from the spectrophotometer is transmitted to a microcomputer at time intervals of,
say, a few milliseconds or, perhaps, microseconds. The problem is well within the
capabilities of contemporary spectrophotometers and computer interfacing hardware,
so the principal limiting factor is speed of mixing. Reactions with a half-time of a
millisecond or less have been successfully studied in this way.
For very fast reactions, the limitation has been only the ingenuity of the researcher
and flash photolysis studies of reactions with the astonishing half-time of t1/2 =
10−15 s indicate that, for some people, this hasn’t been much of a limitation (Zewail,
Nobel Prize, 1999, see Zewail, 1994).
10.5 REACTION MECHANISMS
The mechanism by which molecules actually combine is rarely as simple as the rate
law suggests. Complex reactions may involve equilibriums,
→B
A←
consecutive reactions,
A→B→C
parallel reactions,
A→C

B→D
or combinations of any or all of them.

P1: OTA/XYZ P2: ABC
REACTION MECHANISMS 155
Because of these complications, the terms order and molecularity must be clearly
differentiated. The order is the sum of exponents in the rate equation, while the
molecularity is the number of molecules taking part in a simple reaction step.
The order and molecularity will be the same in only the simplest reactions. An
example in which the reaction order seems at first to be unrelated to the molecularity
is the important atmospheric process involving ozone
2O3 → 3O2
which is not second order. Rather, it follows the rate law:
d O3 O2
− =k 3
dt O2
This can be explained by a reaction mechanism starting with a unimolecular step
→ O2 + O
O3 ←
governed by the equilibrium constant
O2 O
K eq =
O3
The second step of the proposed reaction mechanism is bimolecular
O + O3 → 2O2
and it is the slower step of the two. As in any sequential process, the slowest step
controls the rate of the entire process. In this case there are only two steps: a relatively
fast first step and a rate-controlling second step. The rate law for the bimolecular
collision between O and O3 is
dO3
− = k2 OO3
dt
Solving for O, we obtain
K eq O3
O=
O2
Substituting for O in the previous equation,
dO3 K eq O3 O2
− = k2 OO3 = k2 O3 = k 3
dt O2 O2
we get the observed rate law and rate constant.

P1: OTA/XYZ P2: ABC
Ozone is depleted by about 4% per decade in its total volume in the Earth’s
stratosphere and to a much larger extent over Earth’s polar regions (the ozone hole).
The most important process is catalytic destruction of ozone by atomic halogens due
to photodissociation of chlorofluorocarbon compounds (commonly called freons)
and related compounds. Since the ozone layer prevents most harmful wavelengths
(270–315 nm) of ultraviolet light from passing through the Earth’s atmosphere, de-
creases in ozone have generated concern and led to banning the production of ozone
depleting chemicals. It is suspected that a variety of biological consequences such
as increases in skin cancer, damage to plants, and reduction of plankton populations
in the ocean’s photic zone may result from the increased UV exposure due to ozone
depletion (abstracted from Wikipedia).
10.6 THE INFLUENCE OF TEMPERATURE ON RATE
Variation of reaction rate with temperature is usually exponential, as you know if you
have ever been tempted to heat a reaction mixture “just a little bit more.” Observed
exponential rate curves tempt us to write equations of the same form as the van’t Hoff
equation to describe the rate constant k

a H 1
ln k = − + const
R T
The constant a H is the enthalpy of activation, often called the “energy” of activation.
It can be found empirically.
The activation process can be visualized by pushing over a block of wood, as in
Fig. 10.3. Position c is more stable than a. Thermodynamically, c corresponds to
the products in a spontaneous reaction accompanied by a decrease in free energy.
Even though the block in position a has a higher energy than c, a does not go to c
spontaneously unless a force is applied to push it over, as in b. This is the activation
process. The center of gravity of b is higher than either a or c, hence it has higher
energy. Some energy—the push—must be put into the system a, and it must be
driven to a less stable configuration b before it spontaneously goes to the most stable
configuration c. The energy liberated in going from b to c is greater than the energy
put into the push, and the process liberates energy.
a b c
FIGURE 10.3 An activation energy barrier between an unstable position and a stable
position.
P1: OTA/XYZ P2: ABC
THE INFLUENCE OF TEMPERATURE ON RATE 157
[B]
FIGURE 10.4 Enthalpy level diagram for an activated complex [B].
In a chemical system, kinetic processes can be written
A → [B] → C
where [B] is an activated complex, possibly a very fleeting intermediate like a free
radical. An enthalpy diagram for the reaction involving an intermediate is given as
Fig. 10.4. The enthalpy change for the reaction is given as the down arrow from level
A to level C. This is the measured thermochemical value. The activation energy of the
reaction is the up arrow from A to [B]. Viewed only from the enthalpy point of view,
this is a nonspontaneous process. If the activation enthalpy to get to [B] is supplied,
the reaction takes place and the enthalpy change from [B] to C more than repays the
enthalpy debt incurred in production of [B]. The enthalpy hill from A to [B] is the
activation barrier (Fig. 10.5).
The enthalpy to get over the activation barrier comes from ambient heat supplied
to the system at temperatures greater than 0 K. Evidently, the higher the temperature,
the more enthalpy is supplied to the system and the more molecules of a statistical
distribution have enough enthalpy to get over the barrier. That supplies the quali-
tative answer to the question: Why do reactions go exponentially faster at higher
temperatures than at lower temperatures? The number of activated molecules rises
exponentially with temperature.
[B]
FIGURE 10.5 An activation barrier.

P1: OTA/XYZ P2: ABC
0.541
2 − .9 ⋅ r
(.9 ⋅ r) ⋅ e
2 −r
r ⋅e
2 − 1.1 ⋅ r
[1.1 ⋅ (r)] ⋅ e
0
0 r 10
FIGURE 10.6 A Boltzmann distribution of molecular speeds. The area under the tail of the
curve beyond r increases exponentially.
The quantitative relationship between the reaction rate constant and temperature is
k = ae−a H/RT
This can be appreciated by looking at a Boltzmann distribution of reactant molecules.

The area under the Boltzmann curve that surpasses the activation barrier at r increases
sharply (exponentially) with temperature, leading to the observed exponential in-
crease in k (Fig. 10.6). The exponential relationship between k and T is called the
Arrhenius rate law.
10.7 COLLISION THEORY
One can calculate the number of molecules colliding with one another in a gaseous
sample at a specified temperature. Suppose Z AB is the collision frequency for a
collection of gas molecules capable of undergoing a bimolecular reaction. If all the
molecules were activated, the rate would be directly determined by the number of
collisions per unit time of molecules of kind A with those of kind B. Not all molecules
will be activated, however, and we must multiply Z AB by the fraction that are. This
is the Boltzmann factor e−a H/RT .
For a gas reaction taking place between molecules that are structurally compli-
cated, a probability factor P is introduced to account for collisions between activated
molecules that are not oriented toward each other in the proper way for reaction to
occur. Now
k = P Z AB e−a H/RT
where the integration factor a is included in the empirical probability factor P.

P1: OTA/XYZ P2: ABC
COMPUTATIONAL KINETICS 159
For example, if a bromide ion collided with an alkyl alcohol, in all likelihood it
would collide with the “wrong” part of the molecule.
Br− + CH3 (CH2 )n CH2 OH → No Reaction
Only rarely would it collide with the OH group, resulting in reaction.
Br−

CH3 (CH2 )n CH2 OH → OH− + CH3 (CH2 )n CH2 Br
The probability factor P would be correspondingly low.
10.8 COMPUTATIONAL KINETICS
A critical factor determining the rate of a chemical reaction is the activation enthalpy,
which is the difference a H between the reactant enthalpy and the relatively unstable
activated complex. We can write
a H = Hac − Hr
where Hac is the enthalpy of the activated complex and Hr is the enthalpy of the
reactant molecule. Hehre (2006) has computed several activation enthalpies, including
one for the isomerization of methyl isocyanide to acetonitrile:
CH3 N C → CH3 C N
The postulated activated intermediate is the three-membered ring
CH3
N C
This energetic species does not lend itself to experimental study, but it can be treated
computationally. Results from some computational kinetics studies are roughly com-
parable to measured values, but this is a difficult field because computed activation
enthalpies are very sensitive to the estimated structure of the activated complex and
the rate constant is an exponential function of a H .
P1: OTA/XYZ P2: ABC
Example 10.1 The Enthalpy of Activation

If the half-time of a first-order reaction is 20 minutes at 298 K and 4 minutes at 313 K,
what is the activation enthalpy?
Solution 10.1 The half-times of the reaction at the two temperatures are inversely
related to the rate constants k = 0.693/t1/2 ; therefore the rate constants are in the
ratio of 5:1 with temperature ratio T2 /T1 = 313/298. These values are then loaded
into the rate equation, which has been integrated between the appropriate limits:

k2 a H 1 1
ln =− −
k1 R T2 T1
The ratio of the rate constants is 5.0 and the difference of the reciprocal temperatures
is −1.60 × 10−4 , therefore a H = 83.2 kJ mol−1 :

a H 1 1
ln 5 = − −
R 313 298
a H a H
1.609 = − (0.00319 − 0.00336) = (0.0001608)
R R
a H 1.609
= = 1.00 × 104
R 0.0001608
a H = 1.00 × 104 R = 8.314 × 104 J mol−1 = 83.1 kJ mol−1
Example 10.2
The relative intensity I of fluorescence from electronically excited iodine drops off
according to the time sequence below, where the time is in milliseconds.
Time, ms 0 50 100 150 200 250 300

I, relative 1 0.61 0.42 0.30 0.21 0.08 0.02
What is the rate constant and half-time for this radiative decay? Give units. Plot
relative intensity as a function of time in milliseconds (Fig. 10.7).
Solution 10.2 The sequence, when sketched out, may be a first-order decay. If it
is, drawing a horizontal from I = 0.5 (relative) cuts the curve at about t = 70 ms, a
first estimate of the half-time. Let’s try to convert to ln I vs. time to see if the plot is
linear.
The reaction is first order, as demonstrated by the linear plot of ln I vs. time in
Fig. 10.8 (for more detail see Houston, 2001).
P1: OTA/XYZ P2: ABC
1.2
Relative Intensity
1.0
0.8
0.6
0.4
0.2
0.0
0 50 100 150 200 250 300 350

Time (ms)
FIGURE 10.7 First-order fluorescence decline from electronically excited iodine in mil-
liseconds.
The statistical routine of the SigmaPlot

C 11.0 package gives, among other things,
Equation: Polynomial, Linear
f = y0 + a * x
Coefficient Std. Error
y0 0.0889 0.1232
a −0.0098 0.0007
The intercept, given here as y0 , is close to 0 as it should be and the slope, given
here as a, is −0.0098 ms−1 . This leads to a rate constant d I /dt = −k I of k =
9.8 × 10−3 ms−1 . The half-time for radiative decay is
ln 2 .693
t1/2 = −3 −1
= = 71 ms
9.8 × 10 ms 9.8 × 10−3 ms−1
0.5
0.0
-0.5
-1.0
ln I
-1.5
-2.0
-2.5
-3.0
-3.5
0 100 200 300
Time (ms)
FIGURE 10.8 The natural logarithm of relative intensity vs. time for radiative decay. Within
experimental error, the plot is linear (Section 10.1).
P1: OTA/XYZ P2: ABC
Problem 10.1
Assume that the reaction
k
A → products
is a simple first-order decay. If the amount of A drops from an initial value at t = 0

of 0.5000 mols to 0.0625 mol in exactly 1.0 hour, what is the rate constant k and half
time for the reaction?
Problem 10.2
The curie (Ci) is a measure of radioactive decay frequency 1 Ci = 3.7 × 1010 s−1 . A
1.50 mg sample of the β emitter technetium-99 has an activity of 5.66 × 107 cpm
44 Tc and the β particle is an electron.
(counts per minute). Recall that technetium is 99
(a) What are the products of the reaction? Write the reaction.
(b) Find the frequency in units of Ci.
(c) Find the rate constant (decay constant) k, in whatever unit you choose, and
find the half-life in units of years for 99
44 Tc.
Problem 10.3
In population ecology, an exponential growth curve
P(t) = Aekt
is often assumed for the uncontrolled growth of a newly introduced species where
P(t) is the population at time t and A is a parameter. A new crab was accidentally
introduced into San Francisco bay several years go. Routine trapping and systematic
population monitoring found 12 crabs at the beginning of the survey; two years later,
24 crabs were found at the same location by the same trapping and sampling method.
What was the doubling time of the species? Find k and A. What is the estimated crab
sample after 10 years from t = 0?
Problem 10.4
Further spot checks of the population increase of the invasive species showed the
following results:
t (years) 0 2 3 5 8 10
P(t) 12 24 39 59 167 300
Reevaluate the parameters A and k using a linear least-squares method. What is the
predicted sample at the 12 year survey?
P1: OTA/XYZ P2: ABC
Problem 10.5
The half-time of radioactive decay of 14 C is 5730 years. A wood sample from an
Egyptian tomb had 14 C radioactivity of 7.3 ± 0.1 cpm (counts per minute) per gram of
sample. Freshly harvested wood has a 14 C radioactivity of 12.6 ± 0.1 cpm per gram.
How old is the wood from the tomb and what is the uncertainty of your answer?
Problem 10.6
Given that
A = A0 e−k1 t
and
1 −k1 t
B(t) = k1 A0 e − e−k2 t
k1 − k2
how does C vary with time for the reaction
k1 k2
A→B→C
Remember that A + B + C = A0 at all times t, and the initial conditions are A = A0

and B = C = 0.
Problem 10.7
A0 − A
Show that if = kt for a second-order reaction, then kt 1 A0 = 1.
A0 A 2
Problem 10.8
A sequence of irreversible first-order chemical reactions was observed for 20 hours.
The sequence of reactants and products was
k1 = 0.6 k2 = 0.7 k3 = 0.06

A → B → C → D...
The first three rate constants are 0.6, 0.7, and 0.06 h−1 in the sequence given above,
and the rate constant for decrease in component D going to another product E was
0.02. Plot the concentrations of A, B, and C. The first two equations can be found in
Sections 10.1 and 10.3.2. The third equation is a rather intimidating:

e−0.6·t e−0.7·t e−0.06·t
C(t) := 0.6 · 0.7 · + +
(0.7 − 0.6) · (0.06 − 0.6) (0.6 − 0.7) · (0.06 − 0.7) (0.6 − 0.06) · (0.7 − 0.06)
P1: OTA/XYZ P2: ABC
(There is a fourth and more equations that you can probably write down by extrapo-
lation of A, B, and C. Try it!
Problem 10.9
Given the following experimental data in terms of time t and concentration of reactant
NOBr in mol dm3 , find the rate law and the rate constant for the reaction:
NOBr(g) → NO(g) + 12 O2 (g)
t 0 6 10 15 20 25
NOBr 0.025 0.019 0.016 0.014 0.012 0.011
Problem 10.10
The decomposition of NOCl to NO and Cl2 is analogous to the reaction just studied:
NOCl(g) → NO(g) + 12 Cl2 (g)
The second-order rate constant is 9.3 × 10−5 dm3 mol−1 s−1 at 400 K and 1.0 ×
10−3 dm3 mol−1 s−1 at 430 K. What is the activation energy of this reaction?
P1: OTA/XYZ P2: ABC
11
LIQUIDS AND SOLIDS
Surface tension causes raindrops to be spherical, enables you to float a steel needle on
the surface of water (if you’re careful), and enables trees to draw liquid nourishment
from the earth by capillary attraction. These phenomena and the beautiful hexagonal
form of snowflakes are properties of liquids and solids, the condensed states of matter.
11.1 SURFACE TENSION
The surface of any liquid is enveloped by molecules that are not like other molecules
because of an imbalance of forces acting on them that is different from the balanced
forces in the interior (Fig. 11.1). An imbalance of forces on surface molecules causes
them to be drawn in so as to form an elastic film of surface molecules enclosing
the bulk molecules. The existence of such a film completely surrounding a falling
rain droplet causes it to arrive at a stable structure with the minimum surface area
surrounding the volume of the drop. The minimum surface area for a specified volume
is that of a sphere. If there were no distortions due to gravity, air resistance, and so
on, raindrops would be perfect spheres.
A model of the liquid surface film is not difficult to set up and analyze. Consider
a small rectangular frame with one movable edge enclosing an elastic film such as a
soap film, as in Fig. 11.2. The model is analogous to the three-dimensional expansion
of a gas against a movable piston of area A except that area A is replaced by length l.
The work done in the case of stretching a membrane is the product of the intensive

Copyright
165
P1: OTA/XYZ P2: ABC
166 LIQUIDS AND SOLIDS
FIGURE 11.1 Intermolecular attractive forces acting upon molecules at an air–water

interface.
variable force per unit length γ ≡ f /l times the infinitesimal increase in area dσ :

f
dw = dσ = γ dσ
l
The intensive variable force per unit length γ = f /l in the two-dimensional model
replaces the intensive variable pressure that is force per unit area of the piston face
p = f /A in the three-dimensional model of a piston and cylinder. Please do not
confuse the area of the piston face A with the area of the liquid film σ .
From the physics of vibration of stretched strings, we have the force per unit length
f /l defined as a tension. Because dσ is an infinitesimal increment in the surface area
of the membrane, the intensive variable γ = f /l is called the surface tension.
Imagine a minute device such as that in Fig. 11.3 immersed in a soap solution
and drawn up so that a film of the liquid occupies the area σ that can be expanded
an amount dσ by an upward force f on an edge of length l. Because the tension
γ arises from the surfaces on either side of the liquid membrane drawn up into the
framework, the work of expanding the bimembrane (a memebrane with two sides
or surfaces) is twice the work for each surface, dw = 2γ dσ . The expansion of the
surface is dσ = ldh, where dh is the increase in height of the movable edge of length
l. In general, work is a force times a displacement, so dw = 2γ dσ = 2γ ldh.
σ
f/l
dσ
FIGURE 11.2 Stretching a two-dimensional membrane by moving an edge of length l. A

real liquid would have two films, one facing you on the front surface of the page and one on
the back surface.
P1: OTA/XYZ P2: ABC
SURFACE TENSION 167
dh
dσ
l σ
liquid surface
FIGURE 11.3 Stretching a two-dimensional liquid bimembrane.
Now consider capillary rise caused by surface tension in a tube of radius R. The
length l of the movable edge in Fig. 11.3 is replaced by the circumference c of the tube
in which the liquid rises, c = 2π R; but there is only one circular surface, so the factor
2 drops out of 2γ dσ only to reappear as 2π R, so dw = γ cdh = γ 2π Rdh. The force
opposing capillary rise is mg due to lifting the mass of liquid m in opposition to
gravitational acceleration g (Fig. 11.4). At equilibrium, the forces are in balance:
mg = γ 2π R
The volume of suspended liquid is that of a cylinder of height h and densityρ = m/V .
From V = π R 2 h for a cylinder, we get m = ρV = ρπ R 2 h:
(ρπ R 2 h)g = γ 2π R
and
ρ Rhg
γ =
2
2γ π R
mg
FIGURE 11.4 Capillary rise in a tube of radius R. The equilibrium height of the liquid
column is determined by the balance of the capillary force and the gravitational force.
P1: OTA/XYZ P2: ABC
making it a simple matter to measure the surface tension of a liquid of density ρ that
rises to a height h in a capillary of known radius R against the acceleration due to
gravity, g = 9.807 m s−2 .
11.2 HEAT CAPACITY OF LIQUIDS AND SOLIDS
In 1907 Einstein showed that the heat capacity of many solids is 25 J K−1 mol−1
(in agreement with the prior law of Dulong and Petit) but that at some characteristic
temperature it drops off along a sigmoidal curve to zero at 0 K (Fig. 11.5). Although
his derivation is general, he used it to describe the heat capacity of diamond, an
exceptional solid because of its strong tetrahedral bonding. The resulting curve, one
of the most famous of twentieth-century science, was based on the new quantum
theory of Max Planck, which might otherwise have gone unnoticed by many in the
physics community. The success of Einstein’s theory of heat capacities is shown in a
reproduction of his original graph depicting the theoretical heat capacity of diamond
(solid curve) compared to known experimental points.
In general, the molar heat capacity of a liquid is higher than the molar heat capacity
of a gas because molecular motion in the liquid state implies distortion of the liquid
structure in the vicinity of the moving molecule. This is what we would expect if
we think of a liquid as an extremely nonideal gas. An example is mercury, which
has the heat capacity of an ideal gas, 32 R ∼= 12.5 J K−1 mol−1 in the vapor state, but
−1 −1
has C V = 23.6 J K mol in the liquid state. Notice how close liquid mercury is
5
-1
CE, cal K mol
4
-1
0
0 200 400 600 800 1000 1200 1400 1600
T, K
FIGURE 11.5 Heat capacity as a function of temperature. Einstein calculated the theoretical
curve of the heat capacity of diamond (solid line) and compared it to known experimental
points. Heat was normally measured in calories in Einstein’s time. For more detail, see Rogers
(2005).
P1: OTA/XYZ P2: ABC
VISCOSITY OF LIQUIDS 169
in this respect to the law of Dulong and Petit, which predicts a heat capacity of
25 J K−1 mol−1 for solid metals. As usual, theoretical treatment of liquids is difficult
because their behavior is between that of a perfectly ordered crystal and that of a
perfectly random (statistical) gas.
11.3 VISCOSITY OF LIQUIDS
The viscosity of a liquid often reflects the degree of entanglement of molecules as

they are moved past each other. The viscosity of heavy lubricating oil is greater than
that of gasoline because the average molecular length is greater in oil than in gasoline.
It takes some amount of work to push one molecule past its neighbors.
Consider a liquid flowing through a tube of radius R. At different radial distances
from the center of the tube, the liquid is flowing at different rates. The greatest
rate is precisely at the center of the tube, whereas the lowest rate occurs where the
liquid experiences maximum frictional drag against its inner surface. It is convenient
to consider the liquid flow as consisting of a large number of concentric laminar
“sleeves” of thickness dr, where r is the radial distance from the center of the tube
(Fig. 11.6). The viscous drag retarding the flow on one laminar sleeve relative to its
neighbor sleeve is proportional to the difference in speed of adjacent lamina dv/dr
times the surface area over which the lamina bear against each other. This is the
circumference of the laminar sleeve times its length l:
dv dv dv
drag ∝ A = ηA = η2πrl = f viscous
dr dr dr
The proportionality constant η in this equation is the viscosity coefficient. Because

pressure is force per unit area p = f /A, the force driving the liquid through the tube
under gravity flow is f grav = p A, where A is the area of the end of the sleeve in
flow
dr
dv
FIGURE 11.6 Approximation of laminar flow inside a tube. The difference in speed of flow
between the inner and outer surfaces of a lamina is dv/dr .
P1: OTA/XYZ P2: ABC
question including all sleeves within it. One can think of a pencil of liquid driven
by a force, f grav , flowing coaxially down the tube. The gravitational force on each
sleeve is f grav = p A = pπr 2 , where r is the radius of the sleeve. Under conditions
of constant flow rate, these forces are equal and opposite:
f viscous = − f grav
dv
η2πrl = − pπr 2
dr
p
dv = − r dr
η2l
Integrating the left-hand side of the equation from velocity of flow of 0 at the inner
surface of the tube to v the velocity at radius r, we get
v r
p
dv = − r dr
0 η2l R
Recall that R is the radius of the tube where v = 0 and r < R is the radius of a moving
sleeve of liquid:
p 2 p
v =− (r − R 2 ) = (R 2 − r 2 )
η4l η4l
A little further analysis along the same lines leads to the Poiseulle equation for the
volume of flow V per unit time:
π p R4
V =
8ηl
or
π p R4
η=
8V l
permitting determination of the viscosity coefficient η by measuring the volume of

flow of a fluid through a tube of known R in unit time.
11.4 CRYSTALS
In many crystals, atoms are arranged in a very regular three-dimensional pattern of

rows and columns, echelons deep. Not surprisingly, when radiation strikes a crystal, it
is reflected from planes of atoms, ions, or molecules in the way that light is reflected
from a mirror. The difference is that experimentalists use penetrating X rays to detect
planes within the crystal. The appearance of planes within a crystal is a consequence
P1: OTA/XYZ P2: ABC
CRYSTALS 171
D C
FIGURE 11.7 Close packing of marbles between two sheets.
of the regular arrangement of the atoms, ions, or molecules of the crystal, which is
itself a consequence of the regular forces holding it together.
Simply dropping marbles into a box, one gets the idea of a tendency of the
marbles to settle into a regular structure with layers separated by a distance that can
be calculated as a function of the radii of the marbles. If you shake the box gently, so
that the marbles assume a more or less compact aggregate, you may notice a repeating
structural unit of a cube or hexagon.
To simplify the picture by making a two-dimensional array, think of marbles
dropped into the space between two clear plastic sheets separated by a space equal
to the diameter of the marbles (Fig. 11.7). Now marbles are separated into alternat-
ing rows. Repeating structural units may now be rectangles. Knowing the diameters
(hence the radii) of the marbles enables one to find the distance between alternat-
ing rows.
We have extracted an isosceles triangle ABC from the marble pattern. The altitude
of the triangle DC is also the radius of the marbles. The length of a side AC is twice
the radius of the marbles. That gives us a right triangle ADC. The sum of the squares
of the two sides of ADC is equal to the square of the hypotenuse AC. Distance DC is
equal to one marble radius r, and AC is equal to 2r so AD is
√
AD = AC2 − DC2 = (2r )2 − r 2 = 3r 2
The distance AD is the distance between horizontal lines through the centers of the
alternating rows. For example, if r = 0.500 cm, the distance between the horizontal
lines through the centers of the marbles parallel to DC is 0.866 cm.
Distance DC permits us to take the inverse cosine cos−1 of the adjacent side over the
hypotenuse of angle DAC to find that it is 30◦ . The remaining angle of the right triangle
must be 60◦ . Now that we know all the distances and angles relating the centers of the
marbles, we know all that can be known about the geometry of the marble packing
everywhere the pattern in Fig. 11.7 is maintained. Dropping real marbles into a real
space, one may find irregularities and fissures in the structure. This is analogous to
real crystal structure as well. They show the same kind of irregularities and fissures.
One more thing before going to three dimensions: If we rotate the diagram in Fig. 11.7
by 60◦ , we get a pattern that is identical to the one we just analyzed. There are other
lines at other angles with the same geometric relationships as those we have found.
The marble pattern has some rotational symmetries.
P1: OTA/XYZ P2: ABC
D B
FIGURE 11.8 A less efficient packing of marbles. This packing is less efficient than close
packing because the same number of marbles take up a greater space. To see this most clearly,
notice that the interstitial spaces are larger than they are in Fig. 11.7.
With difficulty, marbles might be juggled into a less efficient but still regular
pattern like Fig. 11.8. The vertical distance between centers for 0.500 cm marbles
is now twice the radius as compared to only 0.866 r for close packing. The same
number of marbles take up more space.
Suppose we had some experimental method of determining distance between
layers DC. That would enable us to tell the difference between the packing pattern in
Figs. 11.7 and 11.8. Assuming that atoms fall into a regular array when an element
or compound crystallizes, we can picture a laminar sheet of atoms with very dense
nuclei in an array similar to that Fig. 11.7 or 11.8. If marbles were packed into a three-
dimensional box, the packing pattern determination would be very closely analogous
to the pattern determination for marbles restricted to a plane. They would pack in
a more or less regular way, and different packing patterns would be a more or less
efficient with respect to space use, just as the marbles were in the simple illustrative
case of two dimensions. Atoms, molecules, or ions pack in three-dimensional crystals
as well. We would like to have an experimental method to solve the reverse of the
problem we just solved; from an experimental value for the distance between layers
of atoms, we would like to obtain the radius of the atoms themselves. The object of
X-ray crystallography is to determine the distances and angles between atomic centers
as we have done with marbles. The problem may be much more complicated, as in
the example of proteins, or it may be nearly as simple as the method just described
extrapolated to there dimensions, as in the case of pure metals and ionic salts.
Electromagnetic waves, from radio frequency of meter wavelengths to γ rays of
λ ≈ 0.1 nm, have an electrical and a magnetic component. The radiation can be
described mathematically as two sine waves, one electrical ε and the other magnetic
H, describing two oscillating vector fields oriented at right angles to one another.
For constructive interference between two waves, the field vectors must point in the
same direction. Otherwise, the radiation of one wave dims or obliterates the other by
destructive interference.
Figure 11.9 shows that, for the radiation reflected from two adjacent horizontal
lines of atoms to be in phase (to have their arrows pointed in the same direction),
the difference in path must be an integer multiple of the wavelength λ. The path
difference is shown as a heavy line. A path difference that is twice the wavelength
of incoming radiation will interfere constructively, and one that is three times λ
P1: OTA/XYZ P2: ABC
CRYSTALS 173
θ θ
FIGURE 11.9 Bragg’s law for constructive reflection.
serves as well, but one that is some uneven multiple of λ will interfere destructively.
The condition demanded of wavelength for the emerging radiation to be in phase is
(1, 2, 3, . . . ,) λ = nλ, where n is an integer.
From this we can extract a triangle with a hypotenuse that is one-half the wave-
length of incoming radiation of wavelength λ (Fig. 11.9). The opposite side of the
triangle so extracted is d, the distance between adjacent lines of atoms in the two-
dimensional model. We measure θ , the angle of constructive reflection. All other
angles give destructive interference and we do not see the radiation reflected. The
sine of θ is its hypotenuse divided by its opposite side, which is d, so we have

nλ 1
sin θ =
2 d
or

nλ 1
d=
2 sin θ
which is Bragg’s law.
11.4.1 X-Ray Diffraction: Determination of Interplanar Distances

The earliest and by far most common diffraction studies were on reflection and
diffraction of X-radiation. When radiation is reflected from two parallel planes of
a three-dimensional crystal, the reflected beams may be in phase, out of phase, or
something in between. If θ is varied until the Bragg equation is satisfied, a sharp
maximum in reflected radiation is recorded by photographic or other means. The
(small) integer n in the Bragg equation is called the order of the reflection, and d is
the distance between reflecting planes.
On the basis of the distance between planes found at various different incident
angles, a repeating pattern is observed; from that, a repeating unit of crystal structure
in three dimensions, the unit cell, is assigned. For example, if d is found to be the
same in all three orthogonal directions, the cell is a simple cube; but if d values differ,
P1: OTA/XYZ P2: ABC
FIGURE 11.10 A face-centered cubic unit cell. The pointer indicates the face-centered Cl−
ion in the front face of the cell.
the cube may be distorted to a rhomb, tetragon, or other geometry. Let’s stick to the
simple cube to illustrate the principle.
From the angles within the unit cell and its orthogonal distances a, b, and c, one
can find its volume Vcell . A measurement of the density ρ = m/V of the crystalline
solid gives the mass of atoms in the cell. Knowing the mass of the atoms in the cell
and their molar mass, we can calculate how many of them there are—for example,
four in the case of Na+ and Cl− . The face-centered unit cell for NaCl resembles two
interpenetrating simple cubic cells (Fig. 11.10). This presents a cell with Na+ and
Cl− ions alternating along each edge. The edge of the NaCl cell is 564 pm, but it has
Na+ −Cl− −Na+ along each edge so the Na+ −Cl− distance is half that dimension,
564/2 = 282 pm. From the relative electronic structures and the fact that Na loses an
electron and Cl− gains one in the ionic bond, we can guess that the Cl− ion will be
about twice as large as Na+ . This estimate leads to an approximate ionic radius of
Na+ , rNa+ = (282/3) = 94 pm, leaving 282 − 94 = 188 pm for rCl− .
11.4.2 The Packing Fraction

One of the things we want to know about a crystal is how efficiently the atoms,
molecules, or ions are packed. This is measured by the packing fraction, the space
in the unit cell that is occupied by atoms relative to the total space in the cell. The
larger the packing fraction, the greater the space occupied by atoms and the less
space “wasted” in the interstices. This can be found by straightforward geometric
calculation.
The idea of a unit cell packing fraction can be illustrated by arranging discs on
a table top. One way is similar to the one shown for the marbles in Fig. 11.8. The
repeating planar unit is a square (Fig. 11.11). A table top covered with many discs
packed in this way can be thought of as simply a repeating pattern of square unit
cells. Once we know the length of the edge of one unit cell and we know that it
is square, we know the geometry of the entire table top full of discs. Chemists are
not usually interested in table tops covered with discs, but we are very interested in
P1: OTA/XYZ P2: ABC
CRYSTALS 175
FIGURE 11.11 A two-dimensional unit cell for packing of discs.
the three-dimensional structure of atomic, molecular, and ionic solids in the regular
arrangements we find in crystals. The concept of the unit cell is central in X-ray
crystallography.
A crystal with the simple square atomic or molecular packing arrangement in all
three directions is completely described as a repeating three-dimensional pattern of
cubic unit cells. Once we know the length of the edge of the cubic unit cell, we know
the geometry of the entire crystal (aside from impurities and structural imperfections).
The three-dimensional analog of Fig. 11.11 has a simple cubic unit cell
(Fig. 11.12).
To simplify, let us go back to the unit cell of discs in Fig. 11.11. The total area A of
discs within the cell is 4 times A/4, where A = πr 2 because, although there are four
discs, one on each corner of the cell, only one-fourth of each disc is actually inside
the cell. The dimension of the cell itself is 2r on an edge so the area of the unit cell is
(2r )2 = 4π 2 . The area of the discs within the cell is πr 2 , so the packing fraction is
Adisc πr 2 π
P= = 2 = = 0.785
Acell 4r 4
When this structure is taken to a three-dimensional simple cubic structure, the inter-
stitial space at the center of the cube takes a larger proportion of the whole and the
packing fraction is a rather inefficient 0.52. Simple cubic packing is not favored by
many crystals. An exception is the metal polonium, 84 Po.
FIGURE 11.12 A simple cubic cell.

P1: OTA/XYZ P2: ABC
Now, let us consider a different structure. It is not likely that spheres will pack,
each directly over its next lower neighbor. Instead (polonium excepted) they will slip
one-half radius to the right or left to give a more efficient packing pattern with the
central atom in the face of a cube or a slightly different pattern with the central atom
a little back of the face so that it takes a position in the center of the cube. These
are the face-centered and body-centered cubic structures fcc and bcc. You can see
what is meant by face-centered by looking at the chlorine atoms only in Fig. 11.10.
Metals packed as planes situated above each other so that atoms of one layer fit into
the interstices of the planes above and below are close-packed. Close-packed spheres
have a packing fraction of 0.740.
11.5 BRAVAIS LATTICES
At first thought, it would seem that there must be very many, perhaps infinitely
many, unit cells in real crystals. Quite the contrary is true. As early as 1850, Bravais
showed that unit cells in three dimensions can be classified into only seven sys-
tems on the basis of their rotational symmetry. These symmetries are important in
X-ray crystallographic studies in which the sample is rotated in an incident beam of
radiation. If a cell (crystal) is geometrically identical after rotation of 180◦ (π ), it has
at least a twofold axis of rotation. If that is all the symmetry it has about that axis,
then it has one C2 axis of symmetry. Bravais showed that the seven classifications
shown in Table 11.1 permit 14 distinct crystal lattices in three dimensions.
11.5.1 Covalent Bond Radii

X-ray diffraction studies can also be carried out on covalently bonded molecular
solids. The results can be augmented by comparison or combination with other kinds
of diffraction studies in the solid, liquid, or gaseous states. It is possible to use beams
of electrons or neutrons in place of X rays. These studies yield bond distances like
rC−Cl = 177 pm in CCl4 . One would like to have a bond covalent radius for the Cl
TABLE 11.1 The Bravais Crystal Systems and Latticesa .
Basic Types 14 Bravais Lattices
1. Cubic a=b=c P, I, F
2. Tetragonal a = c P, I
3. Rhombic α, β, γ = 90◦ P
4. Monoclinic α = 90◦ β, γ = 90◦ P, C
5. Hexagonal a = c P
6.Triclinic α, β, γ = 90◦ P
7. Orthorhombic a = b = c P, I, F, C
a Primitive cells are designated P, body-centered cells are I, face-centered
cells are F, and side-centered cells are C.

P1: OTA/XYZ P2: ABC
LATTICE ENERGIES 177
atom in CCl4 , for example, but there is no clear way of finding where, along the C–Cl
bond axis, one radius leaves off and the other begins.
One approach to this difficulty is by finding the homonuclear bond distances in
molecules like ethane for which the C–C bond distance is 154 pm, and Cl–Cl for which
it is 200 pm. Under the assumption that homonuclear bond distances will be carried
over into the heteronuclear cases, one has (200/2 + 154/2) = 100 + 77 = 177 pm,
which successfully reproduces the experimental value for rC−Cl . Now, by studying
other compounds involving C and other compounds involving Cl, one can gradually
build up tables of covalent bond radii such as those found in elementary textbooks
(Ebbing and Gammon, 1999).
A number of points should be considered before using bond lengths or atomic
radii. For one thing, combination of X-ray data with neutron diffraction data is risky
because X-radiation is scattered by the electron cloud surrounding the nuclei in the
molecule and neutrons are scattered by the nuclei themselves. Clearly the first is
useful in studying bonds and the second is useful for structure. Also, there is no
reason to expect ionic radii and covalent radii to agree with one another because of
the different modes of chemical bonding involved. Tables of bond distances, ionic
radii, and covalent radii should be used with some degree of reserve because of
potential inconsistencies and approximations.
11.6 COMPUTATIONAL GEOMETRIES
Most present-day molecular structure–energy computer programs contain a routine

that optimizes the geometry of the molecule under study so that each of the constituent
atoms resides at the bottom of its unique potential energy well. Once knowing the
complete molecular geometry, the bond lengths and angles can be calculated with
great precision. The results, however, are not exactly comparable to experimental data
because the energy minimum found by computational optimization for atoms in the
force field of all other nuclei and electrons does not coincide with the average position
of the vibrating atom. Current estimates of atomic radii and bond lengths are in good
agreement overall but differ slightly according to the method used to determine them.
Using the Spartan C package, the homonuclear distance r
H−H = 73.6 pm and the
heteronuclear distance rH−F = 90.0 pm are found, as compared to the experimental
values of 74.2 and 91.7 pm.
11.7 LATTICE ENERGIES
Along with their geometry, we would like to know how firmly ionic crystals are
held together. A quantitative measure of the energy or enthalpy holding the crystal
together is its lattice energy. The lattice energy is the energy necessary to draw ions
out of the crystal lattice and propel them into the gaseous state.
NaI(crystal) → Na+ (g) + I− (g)

P1: OTA/XYZ P2: ABC
This enthalpy is the negative of the enthalpy of formation of the crystal from its
gaseous ions. Please recall from Chapter 4 that this is not the standard enthalpy of
formation because Na+ (g) + I− (g) is not the standard state of elemental sodium and
iodine. The lattice energy for NaI and can be calculated from the standard state value

f H o (NaI) by a procedure called the Born–Haber cycle, which is nothing more than
an enthalpy of formation calculation like the one shown in Fig. 4.1 except that it is
a little more complicated because of the nonstandard state of the products. Example
11.1 is not only an illustration of the Born–Haber cycle, but it also should serve to
review and drive home the difference between standard and nonstandard states in
thermochemistry.
The lattice energies of crystals show logical regularities. For example, the alkali
metal iodides fall between 600 and 775 kJ mol−1 and decrease gradually from the
lithium salt (which has a short ionic bond) to cesium iodide (which has a long one).
The lattice energies of the doubly charged alkaline earth salts are very much larger
than those of the alkali metals.
Lattice energies can also be calculated from a theoretical model in which the
energy of the ionic crystal is supposed to be a function entirely of electrostatic forces.
The model is fairly successful, but it involves some infusion of empirical data. All
electrostatic attractions and repulsions can be calculated over the distances separating
ions of opposite charge or the same charge. Knowing the exact geometry of the unit
cell, these interionic distances ai can be calculated precisely, not only for ions in
the same unit cell but also for those in distant cells. The total Coulombic energy
calculated from electrostatic theory in this way is

N A Z + Z − e2 1 Z i+ Z i−
U (r ) = + + Be−r/ρ
r 4π ε0 2 ai+ ai−
where the summed terms can be either positive for repulsion of like charges or
negative for attraction between unlike charges. The constants B and ρ can be assigned
reasonable values by an essentially empirical method (Barrow, 1996). Application to
NaI gives 682 kJ mol−1 by comparison to the Born–Haber value of 658 kJ mol−1 .
The difference is about 3.6%.
Exercise 11.1 The Born–Haber Cycle

Find the lattice energy (enthalpy) of NaI by the Born–Haber cycle.
Solution 11.1 Note that the small distinction between energy and enthalpy is often
ignored in calculations involving large energies. We shall need several pieces of
information to begin. First, we need the transition enthalpy of sodium metal in the
solid form at 298 K and 1 bar, which is its standard state, to the state of the gaseous ion.
This is calculated in two steps, first vaporization and then ionization, even though the
steps may not be differentiated in the real process. We need not worry about the actual
P1: OTA/XYZ P2: ABC
I
62.4
1/2 I 2
Na+
-295.3
495.8
I-
107.3 -658
Na
∆fH
-287.8 NaI
FIGURE 11.13 The Born–Haber cycle for NaI.
path the process takes because enthalpy is a state function and is path-independent.
Vaporization takes up 107.3 kJ mol−1 and ionization takes up 495.8 kJ mol−1 (both
endothermic). These enthalpy changes are shown as vertical lines for the left-hand
side of Fig. 11.13.
Next, we sublime 0.5 mol of I2 (endothermic 62.4 kJ mol−1 ) and add an electron
(electron affinity: exothermic, – 232.9 kJ mol−1 ).
1
I (s)
2 2
→ I(g)
I(g) + e− → I−
These two steps are shown in the top middle panel in Fig. 11.13. The enthalpy of
formation of NaI is shown as the slanted line at the bottom of the diagram and after
all this is done, we have the difference between Na and I in the state of their gaseous
ions and the crystal NaI in its standard state. The formation of NaI from the gaseous
ions is shown by the downward dotted arrow in Fig. 11.13. Adding all the enthalpies,
one gets −658 kJ mol−1 . The reverse of this arrow is the vaporization of NaI into its
gaseous ions. This is the lattice energy of the crystal,
lattice H (NaI) = 658 kJ mol−1 .
Problem 11.1
Calculate the surface area of a unit volume of liquid confined within
(a) a sphere and

(b) a cube
Which is larger and what is the percentage difference?
Problem 11.2
Water is unique among common liquids because it has a surface tension of
72.0 mN m−1 , nearly double the average of the other common liquids listed in the
P1: OTA/XYZ P2: ABC
CRC Handbook of Chemistry and Physics, 2008–2009, 89th ed. The unit is millinew-
tons per meter; be careful not to confuse the two meanings for the letter m. What is
the capillary rise of water in a tube 1.000 mm in diameter?
Problem 11.3
The strongest (first-order) reflection of X rays comes when a crystal under exami-
nation is at an angle of 37.5◦ to the incident beam of radiation. The wavelength of
the X-radiation is λ = 1.54 Å = 154 pm. What is the distance between the reflecting
planes?
Problem 11.4
Suppose that discs of radius r = 1 unit are arranged on a table top in the patterns
in Fig. 11.14 as they are in Figs. 11.7 (left) and 11.8 (right). Find the unit cell and
packing fraction for these arrangements.
FIGURE 11.14 Close packing (left) and simple square unit cells (right).
Problem 11.5
The coordination number of an atom in a unit cell is the number of nearest neighbors.
Because the entire crystal is constructed of replicated unit cells, all atoms in the
system have the same coordination number. What is the coordination number of the
body-centered cubic unit cell in Fig. 11.15?
FIGURE 11.15 A body-centered primitive cubic cell.

P1: OTA/XYZ P2: ABC
Problem 11.6
How many atoms are there in the unit cell in Fig. 11.15?
Problem 11.7
(a) If r is set to 1.0 arbitrary length units, what is the volume of the unit cell in
Fig. 11.15?
(b) What is the volume of the atoms within the unit cell in Fig. 11.15?
(c) What is the packing efficiency?
Problem 11.8
What is the lattice energy (enthalpy) of NaCl?
Take information on Na+ from Example 11.1. The enthalpies of dissociation and
ionization of Cl− are 121.8 and −351.2 kJ mol−1 , respectively. The standard enthalpy
of formation of NaCl is −411.2 kJ mol−1 .
P1: OTA/XYZ P2: ABC
12
SOLUTION CHEMISTRY
The concept of the ideal solution is as useful as the concept of the ideal gas, but the
reference states must be very different because a model having no intermolecular
interactions would not be a liquid. The next best thing is to propose a model in which
no change in molecular forces takes place upon mixing two completely miscible
liquids. In this model, mixing of two ideal liquids would be essentially the same as
mixing of two ideal gases (Section 5.3.1). The enthalpy change would be zero at all
proportions, and the mixing process would be entirely driven by a positive entropy
change.
12.1 THE IDEAL SOLUTION
The enthalpy of mixing to form an ideal solution
A(pure) + B(pure) → AB(mixed)
is zero because no change takes place in attractive forces upon going from isolated
A and isolated B to AB neighbors in the mixture. On the other hand, the entropy
of a system of two pure isolated liquids A and B, in which an observer knows that
any randomly selected molecule will be either A or B according to which “beaker”
it is drawn from, is replaced by a mixed system of greater disorder where random

Copyright
182
P1: OTA/XYZ P2: ABC
RAOULT’S LAW 183
TΔ S
ΔH
0
ΔG
Α Β
XB
FIGURE 12.1 Entropy, enthalpy, and Gibbs free energy changes for ideal mixing at T > 0.
selection of a single molecule might result in either A or B. Molecular disorder has

increased, hence entropy has also increased.
Uncertainty in random selection is greater for any mixture, but the maximum
uncertainty is at equal concentrations A = B. All other ratios are less random. The
enthalpy and entropy of ideal mixing are shown as the middle and upper curves in
Fig. 12.1. The Gibbs free energy is G = H − T S, so the null result for H
causes the free energy change G to be exactly opposite to the T S curve at any
mole fraction X B and any temperature not equal to zero. Note that Fig. 12.1 could
equally well have been drawn with X A as the horizontal axis because X A + X B = 1
and X A = 1 − X B .
12.2 RAOULT’S LAW
A slightly less stringent requirement on the components of an ideal binary solution is

that the partial vapor pressure of its components be a linear function of concentration,
with the vapor pressure of the pure component p ◦A as the slope of the function
p A = X A p ◦A
The concentration is measured in terms of the mole fraction of the component labeled
A, and the variable p A is the vapor pressure of A in the solution. This rule is called
Raoult’s law. Raoult’s law is most easily understood by picking out the straight line
labeled p A in Fig. 12.2.
What is said of A is can also be said of B:
p B = X B p ◦B
P1: OTA/XYZ P2: ABC
184 SOLUTION CHEMISTRY
o
pB
pB
p Ao
pA
Α Β
XB
FIGURE 12.2 Partial and total pressures for a Raoult’s law solution.
Normally the component with the smaller mole fraction is called the solute, and the
one with the larger mole fraction is called the solvent. If we are looking at solutions
over a wide range of mole fractions, this designation becomes rather arbitrary. Real
solutions that come very close to satisfying the Raoult’s law idealization are solutions
of toluene in benzene.
The sum of two linear functions is a linear function, so the total vapor pres-
sure p contributed by both components anywhere along the concentration axis is
a linear function of the difference between the vapor pressures of the pure com-
ponents

p = p ◦A + X B p ◦B − p ◦A
12.3 A DIGRESSION ON CONCENTRATION UNITS
Concentration units in solution chemistry are simple but rather convoluted at times,
and they can cause unnecessary errors. This section is in part a summary of con-
centration units and notation already used and in part an introduction to the unit
of molality, which is favored in the study of chemistry in solvents other than the
“universal” solvent water. Molality is also used in the study of concentrated solutions
and in very precise work.
We have already been introduced to the mole fraction (Section 1.5), which is the
number of moles of a selected component of the solution relative to the total number
of moles of all components in the solution
ni
Xi =
ni + nj
P1: OTA/XYZ P2: ABC
REAL SOLUTIONS 185
We are often interested in binary solutions in which one substance is clearly the solute
and one is clearly the solvent. An example is a dilute solution of NaCl in water, in
which NaCl is clearly the solute and water is the solvent. It is conventional to use
n2 to refer to the number of moles of solute while using n 1 to refer to the number of
moles of solvent. The mole fractions for a binary solution are
n2
X2 =
n1 + n2
and
n1
X1 =
n1 + n2
12.4 REAL SOLUTIONS
Few real solutions follow Raoult’s law and none follow it exactly. (This is like saying
that there is no perfectly ideal gas.) Deviations may give vapor pressure curves
that are consistently higher than Raoult’s law as shown in Fig. 12.3, or they may be
consistently lower. Rarely they may be higher at one end and lower at the other so that
the vapor pressure curve crosses the Raoult’s law curve to give a positive deviation
at one end and a negative deviation at the other. Solutions in which attractive AB
forces are weak give a positive deviation from Raoult’s law because molecules leave
the solution more easily than from the pure liquid. Mixtures with strong AB forces
show a negative deviation.
o
pB
pB
p Ao
pA
Α Β
XB
FIGURE 12.3 Consistent positive deviations from Raoult’s law. In this case, intermolecular
attractive forces are less strong in the solution than in the pure components A and B. The
presence of B encourages A to leave the solution and go into the vapor phase.
P1: OTA/XYZ P2: ABC
12.5 HENRY’S LAW
The shape of the p B curve in Fig. 12.3 suggests another model similar to Raoult’s
law but with a different slope for the lower (solute) end of the curve. According to
Henry’s law, the partial pressure, is the linear function found at the limit of infinitely
dilute solutions (in practice, very dilute solutions). Like the ideal gas law, it is a
limiting law (Rosenberg and Peticolas, 2004).
At relatively high concentrations of solute B, Henry’s law is a poor approximation
but it is clear from Fig. 12.4 that in the very dilute region near the vertical marker,
it is a better descriptor of the partial vapor pressure of B than Raoult’s law. On the
contrary, at X B → 1 (toward the right in Fig. 12.4), B follows Raoult’s law. Now we
have a combined model: Raoult’s law for B acting as a solvent in high concentrations
and Henry’s law for B acting as a solute in low concentrations.
12.5.1 Henry’s Law Activities

We have encountered activities and activity coefficients before. A Henry’s law activity
of solute B in solvent A is shown as a vertical line well to the left on the X B axis of
Fig. 12.4, where X B is small. The activity coefficient γ is the ratio of real p B to the
ideal, as determined by Henry’s law. For the behavior of B shown in Fig. 12.4, the
activity coefficient will be γ < 1 in all cases because real behavior is always less than
ideal. It will approach γ = 1 in the limit of infinite dilution. If the partial pressure
curve in Fig. 12.4 were below the Raoult’s law dotted line, all Henry’s law activity
coefficients would be γ > 1.
The model we have described so far is for binary solutions in which the components
are mutually miscible (soluble in each other in all proportions). Everything we have
o
pB
pB
Α Β
XB
FIGURE 12.4 Henry’s law for the partial pressure of component B as the solute. The solid
line is a Henry’s law extrapolation to infinite dilution of B. The dotted line is Raoult’s law.
P1: OTA/XYZ P2: ABC
VAPOR PRESSURE 187
found for component B dissolved in solvent A is true if we change the AB ratio and
start calling A the solute and B the solvent (on the right-hand side of Fig. 12.3).
Therefore we can obtain Henry’s law activities of both components, one from the
partial pressure of very dilute solutions of A in B and the other from very dilute
solutions of B in A.
12.6 VAPOR PRESSURE
Henry’s original (1803) paper dealt with the solubility of CO2 in water, for which he
found that the partial pressure of pCO2 confined in a closed container over water is
directly proportional to the amount of CO2 dissolved in the water. It was many years
later that his law pCO2 = k X CO2 where X CO2 is the mole fraction of CO2 dissolved
in water, was extended to the binary solutions of mutually miscible components just
discussed.
Let us go back now to Henry’s point of view and look at pCO2 = k X CO2 as a rela-
tionship telling us what the partial pressure of CO2 , which we shall call component 2,
will be over a solution that we have made up to have a mole fraction of X CO2 . Notice
that we are looking at the same problem as Henry but from the opposite side. He was
thinking about the amount of gas dissolved in a solvent (water), and we are looking
at the pressure of solute leaving a solution that we have made up to some value of X 2
and going into the gas phase above the solution. At equilibrium, it’s the same thing:
Component 2(solution) →
← Component 2(g)
The critical condition for all the related colligative properties (Section 12.8) is
a balance between the Gibbs free energy functions of two systems—for example,
gas-phase CO2 and dissolved CO2 . The Gibbs free energy relationship at equilibrium
between dissolved component 2 and gas phase 2 is
µ2 (sol) = µ2 (g)
where µ2 (sol) designates the Gibbs free energy function of component 2 in solution.
We know from Chapters 6 and 7 that, for an ideal gas at constant temperature, we
have
RT
dG = V dp = dp
p
Integrating over the difference between the Gibbs function in an arbitrary state 2 and
a standard state pressure arbitrarily chosen as 1 bar, we get
Gp2 Gp2
dp p2
G 2 = dG = RT = RT ln
G1bar G1bar p 1
P1: OTA/XYZ P2: ABC
Written in terms of Gibbs chemical potentials, this gives us
p2
µ2 (g) = µ◦2 (g) + RT ln = µ◦2 (g) + RT ln p2
p◦
where µ◦2 (g) is the standard state free energy for an ideal gas at a partial pressure p2
in a mixture of gases. The analogous expression for an ideal solute at mole fraction
X 2 in a mixture of solutes is
µ2 (sol) = µ◦2 (sol) + RT ln X 2
(We shall usually treat the special case of only one solute dissolved in one solvent.)
Equality of the two Gibbs potentials at equilibrium gives
µ2 (sol) = µ2 (g)
µ◦2 (g) + RT ln p2 = µ◦2 (sol) + RT ln X 2
or
p2
RT ln = µ◦2 (sol) − µ◦2 (g)
X2
This leads to an expression with only constants on the right-hand side:
p2 µ◦ (sol) − µ◦2 (g)

ln = 2
X2 RT
or
p2 µ◦ ◦
2 (sol)−µ2 (g)
= e RT = econst = k
X2
which is Henry’s law, p2 = k X 2 . Notice that the difference between Raoult’s law
and Henry’s law is merely in the arbitrary selection of a standard state.
12.7 BOILING POINT ELEVATION
The utility of Henry’s law becomes apparent when we consider a very important
class of binary solutions, the class of nonvolatile solutes in a volatile solvent—for
example, sugar (sucrose) in water. The nonvolatile solute has no measurable vapor
pressure, hence Raoult’s law does not apply because there is no p ◦B in Figs. 12.3 and
12.4. The total vapor pressure is due to the solvent:
p = X 1 p1◦
P1: OTA/XYZ P2: ABC
BOILING POINT ELEVATION 189
Δp
1 atm
p p
ΔT b
Tb T
T
FIGURE 12.5 Boiling of pure solvent (left) and a solution of solvent and nonvolatile solute
(right). The temperature of the solution must be increased by Tb to restore its vapor pressure
to 1 atm.
Addition of a nonvolatile solute causes a change in the vapor pressure of the solution
according to the amount of solute added because (1 − X 1 ) = X 2 for a binary solution:
p = p1◦ − p = p1◦ − X 1 p1◦ = p1◦ (1 − X 1 ) = p1◦ X 2
The statement p = p1◦ X 2 is true only in the limit of very dilute solutions.
To see that addition of a small amount of nonvolatile solute causes an elevation of
the boiling point of the solution relative to the pure solvent, consider a solvent that
has an exponential variation of p with T as in Fig. 12.5 (left). Its boiling point is the
temperature at which the vapor pressure is equal to the pressure of the atmosphere
bearing down on the surface of the liquid (dotted line).
When a nonvolatile solute is added to the solvent, its entire vapor pressure curve
is displaced downward by p. An increase in temperature is necessary to restore the
vapor pressure to 1 atm. The vapor pressure of the solution moves along the lower
exponential curve with increasing temperature until it arrives once again at 1 atm and
Tb , whereupon boiling recommences. It is evident that a functional relationship
must exist among p, X 2 , and Tb . To find the relation between the amount of
nonvolatile solute and Tb we have recourse to the Clausius–Clapeyron equation

vap H 1
ln p = − + const
R T
Differentiating, we obtain

dp vap H 1
d ln p = = dT
p R T2
If we take dp/p as p/ p for a very small but finite and measurable p, brought
about by addition of the solute, we must increase the temperature by an amount
P1: OTA/XYZ P2: ABC
dT ≈ Tb to restore the system to its original vapor pressure. These small changes
are related by the equation

p vap H 1
= Tb
p R T2
where we have already shown that p = p1◦ X 2 so p/ p = X 2 p1◦ / p. For experiments
carried out at atmospheric pressure, p1◦ is the vapor pressure at the normal boiling point
of the pure solvent. The difference p = p − p1◦ is the vapor pressure depression.
The temperature rise Tb necessary to restore p to its original value is the boiling
point elevation.
In the limit of small amounts of solute, the vapor pressure change from that of the
pure solvent is small, p1◦ ≈ p, so they cancel to a good approximation and

X 2 p ◦A vap H 1
≈ X2 ≈ Tb
p R T2
or
RTb2 X 2
Tb ≈
vap H1
Writing out X 2 in terms of grams and molar mass of the solvent M1 (problems) leads
to the boiling point elevation in practical laboratory terms as
RTb2 M1
Tb ≈ m = Kbm
vap H1 · 1000
where K b is called the boiling point constant or sometimes the ebullioscopic constant.
The molality, m, is the number of moles per 1000 g of solvent, and M1 is the molar
mass of the solvent; that is, we are expressing the amount of solute in units of
molality. Typical values for K b range from about 0.5 to 5 K kg mol−1 ; for example,
K b (water) = 0.51 K kg mol−1 and K b (benzene) = 2.53 K kg mol−1 .
As we have seen, K b is derived through Gibbs chemical potentials. A simi-
lar treatment leads to the freezing point depression T f and the freezing point
constant K f :
RT f2 M1
T f ≈ m = Kfm
freez H A · 1000
where freez H A is the enthalpy of freezing of the pure solvent at 1 atm. Typical
values are K f (water) = 1.86 K kg mol−1 (that famous number we memorized in
elementary chemistry) and K f (benzene) = 5.07 K kg mol−1 . If we measure the
P1: OTA/XYZ P2: ABC
OSMOTIC PRESURE 191
freezing point depression brought about by dissolution of a known (small) weight of

solute to produce a dilute solution, we can calculate the number of moles of solute
from T f , hence its molar mass in units of kg mol−1 .
12.8 OSMOTIC PRESURE
Many membranes have the property that they pass small molecules but not large
ones. They are semipermeable. The human kidney contains semipermeable mem-
branes that separate salt and urea from the bloodstream for excretion but do not pass
blood protein. If a semipermeable membrane separates two arms of a U tube and a
solution of a protein is placed in one arm of the tube with the solvent in the other
arm, solvent molecules will pass through the membrane into the protein solution
because of the spontaneous tendency of the protein solution to evolve toward a more
dilute state of higher entropy. The process is called osmosis. The liquid level in the
protein arm of the U tube will rise until the downward pressure it exerts on the
membrane due to gravity is just sufficient to balance the entropy-driven osmotic flow
across the membrane. This osmotic pressure, π , follows a law similar to the ideal
gas law:
π V = n RT
where V is the volume of a solution containing n moles of the macromolecular

solute.
We have already made the point that liquids, contrary to gases, are subject to strong
intermolecular interactions. The internal structure of the liquid state is distinct and
different from that of the gaseous state, so why should a property of solutions follow
what appears to be an ideal gas law? The derivation of the osmotic law to arrive at
this remarkable and somewhat counterintuitive conclusion is quite distinct from ideal
gas law theory, and it reveals some of the power of the condition that, at equilibrium,
the Gibbs chemical potential must be the same over different macroscopic segments
of the same thermodynamic system.
To apply this idea to a real system, we find the osmotic pressure on the solution
in the right-hand chamber of Fig. 12.6 by a two-step process. First we shall calculate
the total pressure, which is the ambient pressure p plus an arbitrary pressure π on
pure solvent, through our knowledge of the variation of the Gibbs chemical potential
with variation in pressure. Second we shall find the variation of the Gibbs chemical
potential brought about by addition of a small amount of solute. These two steps are
then summed to find the total pressure on the solution of solvent plus solute.
Step 1
Start with pure solvent and apply some pressure π at constant temperature. The
Gibbs free energy of the system is the Gibbs free energy of the pure solvent at
P1: OTA/XYZ P2: ABC
solvent solution
membrane
FIGURE 12.6 Osmotic pressure, π .
ambient pressure µ◦1 ( p) plus the change resulting from the increase in pressure to
p + π:

p+π
∂µ1
µ◦1 ( p + π ) = µ◦1 ( p) + dp
p ∂p T
The variation of a perfect differential of a thermodynamic function is

∂µ ∂µ
dµ = dp + dT
∂p T ∂T p
but the second term on the right drops out because T = const. We shall refer to the
molar volume of solvent as Vm,1 and note that (Klotz and Rosenberg, 2008)

∂µ1
= Vm,1
∂p T
The Gibbs chemical potential of the pure solvent subjected to a pressure change from
p to p + π at constant Vm,1 is
p+π
µ◦1 (p + π) = µ◦1 ( p) + Vm,1 dp = µ◦1 ( p) + π Vm,1
p
P1: OTA/XYZ P2: ABC
OSMOTIC PRESURE 193
Step 2
In the second step, add a small amount of solute. This brings about a change in the
Gibbs free energy function1
µ◦1 ( p + π, X 2 ) = µ◦1 ( p + π ) + RT ln X 1
We already have an expression for the first term on the right, so
µ◦1 ( p + π, X 2 ) = µ◦1 ( p) + π Vm,1 + RT ln X 1
At equilibrium, the pressure on the solution is the same as the pressure on the pure
solvent:
µ◦1 ( p) = µ◦1 ( p + π, X 2 )
Substitution for µ◦1 ( p + π, X 2 ) on the right gives
µ◦1 ( p) = µ◦1 ( p) + π Vm,1 + RT ln X 1
which means that
π Vm,1 + RT ln X 1 = 0
or
π Vm,1 = −RT ln X 1
In a dilute binary solution, we have
ln X 1 ∼
= −X 2
so
π Vm,1 = RTX2
where Vm,1 is the molar volume of the solvent. This can be expressed in terms of the
total volume as Vm,1 = V /n 1 , so
V
π Vm,1 = π = RTX2
n1
1 In a binary solution, µ◦1 can be expressed either as a function of X 1 or X 2 .

P1: OTA/XYZ P2: ABC
but the mole fraction of the solute tends toward n 2 /n 1 at small concentrations, so
V n2
π = RT
n1 n1
or π V = n 2 RT , which has the form, but not the content or meaning, of the ideal gas
law. This is called the van’t Hoff equation for osmotic pressure (as distinct from the
van’t Hoff equation for equilibrium).
12.9 COLLIGATIVE PROPERTIES
If we calculate molality m from the formula weight of solute—for example,

NaCl—and use it in the equation for osmotic pressure, we shall be wrong. Os-
motic pressure π , like p, Tb , and T f , is a colligative property because it is a
property of the number of particles in solution. Our error results from two factors:
First, the number of solute particles in solution may double or triple due to ionization;
and, second, the free solvent concentration may be reduced by solvation.
If NaCl is the nonvolatile solute in water for example, there will be an effective
molality approximately two times the anticipated value because NaCl exists as Na+
ions and Cl− ions in aqueous solution. The number of particles in solution is the van’t
Hoff i factor, 2 for NaCl solutions, 1 for sucrose, which does not ionize, 3 for ZnCl2 ,
and so on. Van’t Hoff i factors are, however, integers only at infinite dilution.
In real solutions, van’t Hoff i factors show a systematic deviation from integral
values due to strong solvation (hydration) of the molecules or ions. When there is
a strong association between a solute molecule and solvent molecules, the solvent
molecules are effectively “taken away” from the solution. The amount of free solvent
is reduced and the relative amount of solute is greater than we conventionally calculate
it to be. The measured change in colligative properties is augmented.
The freezing points of aqueous solutions of NH3 , which is not ionized to any
appreciable extent, are shown in Fig. 12.7. The freezing point of water decreases
with ammonia concentration according to the van’t Hoff equation T f = −1.86 m to
about m ∼ = 4.0 but then the freezing point becomes more negative than theory predicts,
as though the solution were more concentrated than it actually is. The effective
molality (Zavitsas, 2001) is reduced to m 1 − hm 2 , where h is a parameter called
the solvation number, which gives the number of solvent molecules held so tightly
by solute as to be ineffective. In water solution, h is called the hydration number.
The solvation number h is not an integer because it is an average over many solute
molecules or ions. It is not difficult to determine h; it is just an empirical parameter
chosen to cause real colligative behavior to approach the van’t Hoff equation. In the
case of ammonia dissolved in water, the choice h = 1.8 leads to the function shown
by open circles in Fig. 12.7. These experimental freezing points differ from those
shown by solid circles only in that the molalities have been recalculated as moles of
solute per kilogram of free water remaining after the solute has been hydrated.
P1: OTA/XYZ P2: ABC
PROBLEMS, EXAMPLES, AND EXERCISE 195
Freezing Point -20 Ammonia
-40
-60
-80
-100
0 5 10 15 20 25 30 35
Molality
FIGURE 12.7 Lowering of the freezing point of water by ammonia (CRC Handbook
of Chemistry and Physics, 2008–2009). The straight line has the theoretical slope of
–1.86 K mol−1 . Open circles have had their molality corrected for hydration of NH3 .
The straight line in Fig. 12.7 was calculated from T f = −1.86 m, and the three
open circles were calculated by correction of the molality for a hydration number of
1.8 by the Zavitsas method. The small deviation of the open circles from the straight
line is due to failures in some of the approximations made in the derivation of K f (for
example, ln X 1 ∼ = −X 2 ). They can be corrected in a more rigorous derivation (see
Problems). Hydration numbers for many other solutes are known and are usually
higher than 1.8, meaning that the departure of the actual colligative property from
its ideal estimate is correspondingly greater. The approach to ideal behavior at small
NH3 concentrations in Fig. 12.7 shows the meaning of the term “limiting law” quite
graphically. The phenomenon described here is not restricted to any one colligative
property or to water as a solvent; it is general.
PROBLEMS, EXAMPLES, AND EXERCISE
Example 12.1 Partial Pressures

If we carry out experimental determinations of the partial pressure of acetone over
solutions of acetone in diethyl ether, results for small values of X 2 will be something
like Fig. 12.3. Some typical experimental data are given in Table 12.1. Examine
the data for the pressure of acetone over diethyl ether and obtain an estimate of the
activity coefficient for acetone at mole fraction X 2 = 0.20.
Solution 12.1 A first estimate of the Henry’s law constant, which is a limiting law,
can be found from the limiting slope of p2 as a function of X 2 for the first few points
P1: OTA/XYZ P2: ABC
TABLE 12.1 Vapor Pressures of

Acetone over Dilute Binary Solutions
of Acetone in Diethyl Ether.
X2 p2 (kPa)a
0.000 0.00
0.029 2.53
0.068 5.33
0.154 9.90
0.178 10.90
0.255 14.46
0.315 16.92
0.372 19.02
a The units of pressure are kilopascals.
on the curve of p2 vs. X 2 . This slope is the slope of the tangent line p2 vs. X 2 for the
most dilute solutions and is approximately 2.53/0.0290 = 87 kPa. Thus the Henry’s
law estimate of p2 at X 2 = 0.20 is 0.20 (87.2) = 17.4 kPa.
The interpolated value of p2 at X 2 = 0.20 is
0.255 − 0.200 14.5 − p2

= = 0.71
0.255 − 0.178 14.5 − 10.9
p2 = 14.5 − 0.71(3.6) = 11.9 kPa
The activity coefficient is the ratio of the actual value to the Henry’s law estimate γ =
11.9/17.4 = 0.68 and the activity is the effective concentration of acetone relative to
the Henry’s law standard a = γ X 2 = 0.68 and (0.20) = 0.14.
Example 12.2
What are the mole fractions of solvent and solute in a 0.1000 molar solution of NaCl?
Solution 12.2 The number of grams of NaCl in 1 liter of the solution is
0.1000 (22.99 + 35.45) 5.844

0.1000 = =
1.000 liter 1.000 dm3
The density ρ = m/V of pure solid NaCl is 2.17 g cm−3 , so NaCl takes up
V = m/ρ = 5.844/2.17 = 2.69 cm3
This amount of the total volume of the solution is taken up by the solid NaCl, leaving
997.3 cm3 of water in the 1.000 dm3 of solution. Now that we know the volume of
solvent, we can calculate the mole fractions. Recall that we have 0.1000 mol of NaCl
P1: OTA/XYZ P2: ABC
by definition and the molar mass of water is 18.02, so we have 997.3/18.02 = 55.345
mol of water. The mole fractions are
n2 0.1000
X2 = = = 1.804 × 10−3
n1 + n2 55.345 + 0.1000
and
n1 55.345
X1 = = = 0.9982
n1 + n2 55.345 + 0.1000
Comment: First, notice that the sum of the two mole fractions is exactly 1.000, as
it should be because there are only two components in this solution. Second, notice
that we have already made some pretty serious approximations and assumptions. We
don’t really have the right to assume that the volume taken up by NaCl in solution
is the same as the volume in the crystalline state. Also we are assuming that the
density of water is exactly 1.000 kg dm−3 and the volume of the solution is exactly
1.000 dm3 without specifying the temperature. We shall have to take these approxi-
mations more seriously later.
In precise solution chemistry, we often use molal concentrations, the amount of
solute expressed in moles per 1000 g of solvent in contrast to the ordinary stockroom
unit of moles per liter or molarity. A 0.1000 molal solution contains 1000/18.02 =
55.494 moles of water, but so does a 0.2000 molal solution because the volume of
the solute is not subtracted from that of the solvent. That is one advantage of the
molal convention. Another is that the number of grams in a molal solution does
not depend on the temperature, but the number of grams in a molar solution does
because the volume of the solution is temperature-dependent. The mole fraction of the
0.1000 molal solution is
n2 0.1000
X2 = = = 1.800 × 10−3
n1 + n2 55.494 + 0.1000
and
n1 55.494
X1 = = = 0.9982
n1 + n2 55.494 + 0.1000
with a sum of 1.000 as before.

Why specify a new concentration unit when it comes out the same as the old
one? As the following exercise shows, it does not come out exactly the same and
sometimes it is not even close. For very dilute solutions, the difference is negligible,
which is why the concentrations of solutions in working labs and elementary courses
are usually given in molarity rather than molality. In medicinal and clinical chemistry,
many other concentration units are used but they can all be reduced to the basic ones
with a little algebra.
P1: OTA/XYZ P2: ABC
Exercise 12.1
(a) Calculate the mole fractions of solute and solvent in a 1.000 molar solution
of sucrose in water.
(b) Calculate the mole fractions of solute and solvent in a 1.000 molal solution
of sucrose in water.
Solution 12.1
(a) The number of grams of sucrose C12 H22 O11 (M = 342.3) in 1.000 liter
of the solution is 342.3/1.000 dm3 . The density ρ = m/V of pure solid
sucrose is 1.580 g cm−3 so sucrose takes up V = m/ρ = 342.3/1.580 =
216.6 cm3 , leaving 783.4 cm3 of water in the 1.000 dm3 of solution. We
have 783.4/18.02 = 43.474 moles of water. The mole fractions are
n2 1.000
X2 = = = 0.02248
n1 + n2 43.474 + 1.000
and
n1 43.474
X1 = = = 0.9775
n1 + n2 43.474 + 0.100
for a total of 1.000.
Comment: These mole fractions depend on the assumption that the volume that su-
crose molecules take up in water is the same as the molecular volume in the crystalline
state. This is extremely unlikely, meaning that our mole fractions calculated in this
way are probably wrong.
(b) The mole fractions of the 1.000 molal solution are
n2 1.000
X2 = = = 0.0177
n1 + n2 55.494 + 1.000
and
n1 55.494
X1 = = = 0.9823
n1 + n2 55.494 + 1.000
With a sum of 1.000 as before. Under these circumstances, the 1.000 molar
and 1.000 molal solutions are very different.
Comment: There are no assumptions concerning the volume of sucrose molecules in

solution therefore the molality is correct as written.
P1: OTA/XYZ P2: ABC
Further Comment: Sucrose molecules (or any solute molecules) may “tie up” or hy-
drate (solvate) water molecules, rendering them inactive in any participatory physical
phenomenon like the vapor pressure over the solution, which depends on the number
of solvent molecules in the solution that are free to go into the vapor phase. Other
colligative properties (Section 12.8) are influenced in the same way by solvation.
Exercise 12.2
The freezing point of pure benzene is 5.49◦ C, and its freezing point constant is
K f = 5.07◦ C. A sample of a crystalline unknown was made up such that it contained
18.7 mg of unknown per 1.000 g of benzene. The freezing point of the resulting
solution was found to be 4.76◦ C. What was the molar mass M of the unknown?
Solution 12.2 The freezing point depression was
f T = 5.49 − 4.76 = 0.73◦ C = Kf m
0.73
m= = 0.144 g kg−1
5.07
The unknown was in a concentration of 0.0187 g per 1.000 g benzene, which is

18.7 g kg−1 . If the solution of 18.7 g kg−1 is 0.144 m, then, by direct ratio, we obtain
18.7 M
=
0.144 1.000
M = 130 g mol−1
where M is the experimental molar mass. The unknown might be naphthalene, C10 H10 .
Problem 12.1
Exactly 10.00 g of NaCl was added to sufficient water to make up 100.0 g of solution.
(a) What was the weight %?

(b) How many moles of NaCl were present?
(c) What was the molality of the solution?
(d) What is the molarity of the solution?
(e) Assuming a density of 1.071 g cm−3 at 20◦ C for this solution (CRC Handbook
of Chemistry and Physics, 2008–2009, 89th ed.), what is its volume?
(f) With this new information, find its molarity.
P1: OTA/XYZ P2: ABC
Problem 12.2
Acetone and chloroform are completely miscible in one another. The partial pressures
of chloroform as a function of the mole fraction in acetone solutions are
X2 p2 (kPa)
0.00 0.00
0.20 35
0.40 82
0.67 142
0.80 219
1.00 293
Find the Raoult’s law and Henry’s law activity coefficients of chloroform at 0.20
mole fraction.
Problem 12.3
Derive an expression for the freezing point depression from an equality of Gibbs
potentials.
Problem 12.4
Given that the enthalpy of fusion of water is about 6 kJ mol−1 , determine the value
of the freezing point constant, Kf .
Problem 12.5
Show that
M1
X2 = m
1000
where X 2 is the mole fraction of dissolved solute, M1 is the molar mass of the solvent,
and m is the solute molality, that is, show that
RTb2 M1
Tb ≈ m = Kbm
vap H1 1000
for dilute solutions.
Problem 12.6
In the section on freezing point depression, the CRC Handbook of Chemistry and
Physics, 2008–2009, 89th ed., defines the mass % as the mass of solute divided by
P1: OTA/XYZ P2: ABC
the total mass of the solution. Exactly 10.00 g of NaCl was added to sufficient water
to make up 100.0 g of solution.
(b) How many moles of NaCl were present in the solution?
(d) What was the molarity of the solution?
(e) Assuming a density of 1.0708 for this solution (CRC Handbook of Chemistry
and Physics, 2008–2009, 89th ed.), what was its volume?
(f) What was its molarity?
(g) What was the mole fraction of NaCl?
(h) What was the mole fraction of water?
Problem 12.7
Exactly 10.0 g of B dissolved in 1000 g of A which has a molar mass MA of 100 and
a density of 1.000 gave an osmotic pressure of 0.0500 atm at 300 K. Find the molar
mass of B.
Problem 12.8
In the section on freezing point depression, the CRC Handbook of Chemistry and
Physics, 2008–2009, 89th ed., defines the mass % as the mass of solute divided
by the total mass of the solution. Ammonia was infused into pure water until its
concentration was exactly 10.00 g of NH3 per 100.0 g of solution.
(b) How many moles of NH3 were present in 100 g of the solution?
(d) What was the molarity of the solution?
(e) Assuming a density of 0.9575 for this solution (CRC Handbook of Chemistry
and Physics, 2008–2009, 89th ed.), what was its volume?
(f) What was its molarity?
(g) How many moles of water were there?
(h) If 1.8 mol of water are tied up by each mole of NH3 , how many moles of free
water are there?
(i) How many grams is this?
(j) What is the molality of the solution of hydrated ammonia in water?
Problem 12.9
As part of the derivation of the osmotic pressure equation (Section 12.7), the statement
was made that “In a dilute binary solution,
ln x1 ∼
= −x2 ”
P1: OTA/XYZ P2: ABC
Verify this statement by numerical evaluation. Show that it is valid for dilute solutions
only. Hint: Go to your calculus book and search infinite series.
Problem 12.10
A solution is made up so as to have a concentration of 1.428 g of solute per dm3 in
water. The osmotic pressure exerted by the solvent across a semipermeable membrane
(Fig. 12.7) at 298 K was 0.224 atm. What is the molar mass of the solute?
P1: OTA/XYZ P2: ABC
13
COULOMETRY AND CONDUCTIVITY
Aqueous solutions of some substances conduct electricity well and some conduct it
poorly. All aqueous solutions offer resistance R to flow of electrical current I. Good
conductors offer low resistance and poor conductors offer high resistance. In what
follows, we shall assume Ohm’s law V = IR, where V is the potential difference
forcing current to flow through the resistance.
13.1 ELECTRICAL POTENTIAL
A potential difference between two points separated by a distance l can be maintained

by means of an electrochemical cell, commonly called a “battery” (Fig. 13.1):
φ = φ(0) − φ(l)
The potential difference V = φ is called the voltage or voltage drop. We shall be
concerned with the influence of a potential difference on various chemical solutions,
especially salt solutions, placed between a pair of charged parallel plates. Solutions
of salts in water that conduct electricity are called electrolyte solutions. Like other
conductors, they offer some resistance to current flow.
13.1.1 Membrane Potentials

In mammalian cells, potential differences are established by a different mechanism. If
a membrane is permeable to ions of one charge but impermeable (or less permeable) to

Copyright
203
P1: OTA/XYZ P2: ABC
204 COULOMETRY AND CONDUCTIVITY
"battery"
+ –
(0) (l )
FIGURE 13.1 The potential drop between charged plates is V = φ(0) − φ(l).
ions of the opposite charge, a potential difference is established across the membrane.
Up to this point, expressions involving the Gibbs chemical potential µ have been
used, assuming that there were no other potentials present; but in this section, a new
potential, the electrical potential φ, is introduced. The two potentials, chemical and
electrical, are additive:
µ = µ + Fφ
where F is a proportionality constant (see below).

Suppose that, for a dilute solution of KCl, the potassium ion is preferentially
passed through a membrane that blocks the Cl− anion. An electrical potential builds
up that is positive on the potassium-rich side of the membrane and negative on the Cl−
side. The situation is analogous to preferential passage of small molecules through a
membrane in opposition to an osmotic pressure, except that now it is the electrical
potential that opposes flow across the membrane. Just as in the case of osmotic flow,
when the chemical potential driving the transfer is equal and opposite to the electrical
potential, flow stops and an equilibrium has been established. This is the type of
preferential ion flow that occurs between the interior and exterior of mammalian cells
under certain conditions, and it is part of the mechanism of signal transmission along
nerve networks (Fig. 13.2).
If we designate the interior of the cell α and the exterior of the cell β and take into
account only the potassium ion, at equilibrium the sums of electrical and chemical
potentials on either side are equal:
µ+α + Fφα = µ+β + Fφβ
K+( ) K +( )
Cl – Cl –
FIGURE 13.2 An ion-permeable membrane (schematic).

P1: OTA/XYZ P2: ABC
RESISTIVITY, CONDUCTIVITY, AND CONDUCTANCE 205
We know that, in general, µ = µ◦ + RT ln a; thus, in this case, we have
µ◦+α + Fφα + RT ln a+α = µ◦+β + Fφβ + RT ln a+β
but µ◦+α = µ◦+β because we are talking about the same ion, K+ , on either side of the
membrane. Differences come, not in µ◦+ , which is fixed, but in the potentials φ and
the activities a, which are variable:
Fφα + RT ln a+α = Fφβ + RT ln a+β

a+β
F φα − φβ = RT ln
a+α
F is called the Faraday constant. It is a conversion factor from charge in coulombs

C or millicoulombs mC to moles. The faraday is the charge on 1 mol of electrons,
96,485 C mol−1 , which may in general be regarded as 1 mol of charge.
Under the stipulation that the concentrations are rather low, the activity coefficients
γ are about equal to 1.0 and we may replace the activities by molarities (or molalities)
m to find the membrane potential
RT m +β
φ α − φβ = ln
F m +α
Nerve cell walls are semipermeable, accounting for an equilibrium membrane
potential of about 70 mV. This equilibrium potential can be disturbed by an electrical
pulse from a neighbor cell, whereupon the pulse is transmitted to another neighbor
cell in a sequence that is part of the mechanism for transmission of information
along a nerve fiber.
13.2 RESISTIVITY, CONDUCTIVITY, AND CONDUCTANCE
If a charge suspended in a fluid medium is subjected to a potential difference φ,

it will move. Moving charge is called current I = d Q/dt, where Q is the charge in
coulombs. Typically, a moving charge meets resistance R.1 Resistance is proportional
to the length l of a resistor and is inversely proportional to its cross-sectional area A:
l l
R∝ =ρ
A A
The proportionality constant ρ = R A/l is a characteristic of the resistor material. It
is called the resistivity.
These definitions prompt us to define the inverse of the resistance called the
conductance L ≡ 1/R, along with the inverse of the resistivity, the conductivity
κ ≡ 1/ρ:
1 A κA
L≡ = =
R ρl l
1 We shall not consider superconductive media here.

P1: OTA/XYZ P2: ABC
When a voltage V is applied to a conductance cell with parallel plates of area A

separated by a distance l and containing an electrolyte solution, the current is
κA
I = JA = V
l
where J is the flux d Q/dt of charge Q passing through unit cross-sectional area A.
When we combine these two equations, the current becomes I = J A = L V , which
is another way of writing Ohm’s law V = I R.
With these equations, it is possible to determine the conductivity of any solution
from the resistance and a knowledge of l/A. It is not easy to measure l/A to the high
level of accuracy required for good work in this field, but the task has already been
done for solutions of KCl at several concentrations and temperatures. The results
of these measurements are available (CRC Handbook of Chemistry and Physics
2008–2009, 89th ed). A selected value is κKCl = 0.1408 ohm−1 m−1 for 1.00 × 10−2
molal KCl at 298 K.
Suppose we construct a cell looking something like the one in Fig. 13.1, fill it with
1.00 × 10−2 molal KCl, and measure the resistance R = 650 ohms. We have

l 1 l
κKCl = 0.1408 = L =
A 650 A
l
= 650(0.1408) = 91.52 m−1
A
We now know the geometric ratio l/A without having to make precise measurements
of either l or A.
13.3 MOLAR CONDUCTIVITY
For chemical applications to solutions of varying concentrations c, an additional

definition is necessary. The molar conductivity of an electrolyte solution is
κ
≡
c
where c is the concentration of the solution in mol L−1 or mol dm−3 .
We shall be concerned largely with aqueous solutions. More than 100 years ago,
F. Kohlrausch published a series of research results on conductivity of electrolytes in
aqueous solutions. The conclusions he drew included what is now called Kohlrausch’s
law. The law is a linear connection between the molar conductivity and the square
root of the electrolyte concentration:
√
= ◦ − K̃ c
where K̃ is an empirical constant. Kohlrausch’s law leads to a graphical way of

determining the parameter ◦ , a characteristic of the electrolyte alone, without
P1: OTA/XYZ P2: ABC
MOLAR CONDUCTIVITY 207
HCl
HOAc
NaOAc
c 1/2
FIGURE 13.3 Kohlrausch’s law for conductance of the strong electrolytes HCl and NaOAc
and the weak electrolyte HOAc. ◦ is the intercept on the vertical axis.
electrostatic interference from crowding by other ions that may be of the same elec-
trolyte or any other electrolytes present in the solution. For example, the Kohlrausch
curves for the NaOAc and HCl in Fig. 13.3 can be extrapolated to a linear approxi-
mation shown by the dashed lines. The intercept of this limiting straight line with the
vertical axis is ◦ . It is called the molar conductivity at infinite dilution. The shape
of the HOAc curve, however, is such that one would not wish to find ◦ in the √ same
way for this solute as for HCl and NaOAc because the curve is too steep near c = 0
to do the job with acceptable accuracy.
Electrolytes in aqueous solution generally fall into two broad classes: strong
electrolytes, which show clean linear extrapolations to ◦ , and weak electrolytes,
which do not. A collection of experimental values for ◦ for strong electrolytes runs
from about 100 to 425 ohm−1 and, other than high values for strong acids and bases,
shows no particular regularity in itself. Kohlrausch observed, however, that many
differences in ◦ are equal; for example,
◦KCl − ◦NaCl = ◦KNO3 − ◦NaNO3 = ◦KOH − ◦NaOH
and so on, for a number of other ◦ values arranged pairwise in this fashion. This
would be true, he said, if ◦ values were the result of contributions from each of the
ions produced when the strong electrolyte goes into solution; for example,
NaCl(s) → Na+ (aq) + Cl− (aq)
where (s) designates the solid compound and (aq) designates the ion in aqueous
solution. If this were true, he said, the regularity in ◦ could be explained by writing
P1: OTA/XYZ P2: ABC
the contributions to ◦ as
◦KCl = λ◦K+ + λ◦Cl−

◦NaCl = λ◦Na+ + λ◦Cl−
..
.
and so on.
In this way, the difference found for
◦KCl − ◦NaCl = ◦KNO3 − ◦NaNO3
would be merely the difference
◦KCl − ◦NaCl = λ◦Na+ + λ◦Cl− − λ◦K+ − λ◦Cl−

= λ◦Na+ − λ◦K+
and would be the same for all pairs of sodium and potassium electrolytes because the
anions, Cl− , NO− −
3 , OH . . . , would always cancel out. The same would be true for
pairs of electrolytes containing a common anion, because the cation would always
cancel out. In these and similar calculations, the integral charge on an ion must always
be taken into account so that the expression for Kohlrausch’s law of independent ion
migration is
◦ = ν+ λ+ + ν− λ−
where ν is the number of charges on each ion, ν = 1 for the examples shown.
13.4 PARTIAL IONIZATION: WEAK ELECTROLYTES
So far we have considered mainly strong electrolytes, those for which ionization
is complete in aqueous solution. This is not always the case. Ionization of weak
electrolytes is partial. The usual example chosen is that of acetic acid for which
ionization at ordinary concentrations is very small. Examples of the vast difference in
behavior can be seen by comparing the Kohlrausch conductivity curves in Fig. 13.3
for HCl, which is completely ionized, with that of acetic acid (HOAc), which is not.
Ionization of HOAc does, however, rise rapidly in dilute solutions, and Kohlrausch’s
law of independent migration holds at infinite dilution. At infinite dilution, the weak
electrolyte HOAc is completely dissociated (all electrolytes are), so we have a very
simple way of determining the approximate degree of ionization α at some finite
concentration c by comparing to ◦ . We simply take the ratio

α=
◦
P1: OTA/XYZ P2: ABC
ION MOBILITIES 209
where α runs from 0 (the trivial case of no ionization) through small numbers at most
concentrations only rising to α = 1 for complete ionization.
The only trouble with this plan is evident from Fig 13.3. Accurate data are difficult
to obtain at very low concentrations; and even if they could be found, ◦ is at the
intersection of two nearly parallel lines. A way out of this dilemma is through
Kohlrausch’s law of independent ion migration at infinite dilution. Figure 13.3 shows
that, although we cannot find an accurate value for ◦HOAc directly, we can find
accurate values for ◦HCl and ◦NaOAc . These, in combination with ◦NaCl , yield
◦HCl + ◦NaOAc − ◦NaCl

= λ◦H+ + λ◦Cl− + λ◦Na+ + λ◦OAc− − λ◦Na+ − λ◦Cl−
= λ◦H+ + λ◦OAc−
= ◦m (HOAc)
Armed with this new piece of information, we can calculate α = /◦ . The disso-
ciation reaction is
→ H+ + OAc−
HOAc+ ←
(1−α) c αc αc
where the concentrations of the individual ions at overall concentration c are con-
trolled by the degree of dissociation αc. The undissociated HOAc (1 − α)c is what
is left of the original HOAc after dissociation has taken place. Typical values might
be 17% dissociated and 83% undissociated for α = 0.17. The equilibrium constant
for this acid dissociation reaction is
[H+ ][OAc− ] (αc) (αc) 0.170 (.170)

Ka = = = c = 0.035 c
[HOAc] (1 − α)c 0.83
If these dissociation data are observed for a 5.0 × 10−4 molar solution, K a (HOAc)
is found to be about 1.7 × 10−5 (accurate value 1.754 × 10−5 (CRC Handbook of
Chemistry and Physics, 2008–2009)). This is the acidity constant of acetic acid
familiar from general chemistry.
There are now computational methods of obtaining or at least approximating
quantum mechanical gas-phase free energies for the components of the dissociation
reaction along with the necessary computed free energies of solvent interactions.
13.5 ION MOBILITIES
In ion mobility studies we wish to know how fast an ion migrates in an electrical field.
This is related to the charge flux J = dQ/dt and to the transport numbers t+ and t− ,
the proportion of charge carried through the solution by each ion. Transport numbers,
which are fractions, add up to 1 and are larger than 0.5 for an ion that carries more
P1: OTA/XYZ P2: ABC
cathode
HCl
LiCl
anode
FIGURE 13.4 Moving boundary determination of the mobility of H+ . Li+ is the follower
ion.
than its “share” of charge and less than 0.5 for an ion that carries less than its share
of the charge.
One way of measuring the mobility of a cation is to load a capillary tube with a
solution containing the ion in question on top of a second electrolyte solution that is
more dense than the measured ion solution. The cation of the more dense solution
is called a follower ion because when a field is applied to the tube, it moves upward
at the rate of the principal cation, H+ in the case shown in Fig. 13.4. The follower
ion is drawn along by H+ to prevent a charge gap from developing between the two
solutions.
Figure 13.4 shows a solution of HCl loaded into a capillary tube on top of a
solution of LiCl, where Li+ is the follower ion. Upon applying a potential difference
across the cathode and anode, H+ is attracted to the cathode (−) and Li+ follows
along. The velocity of migration is determined by the faster of the two ions. The
interface between the HCl and LiCl solutions is maintained by the follower ion and
can be visually observed because of the difference in refractive indices of the two
solutions. The velocity is x m s−1 and the mobility u is the speed x/t per volt of
potential difference:
x −1 x
u= m s−1 volt m−1 = m2 volt−1 s−1
tV tV
The rather odd-looking unit is meters per second per volt per meter of the resisting
medium. In the present case, this is unit voltage across the conducting solution. All
of the variables on the right are measurable quantities, hence the mobility can be
calculated.
P1: OTA/XYZ P2: ABC
THE HITTORF CELL 211
13.6 FARADAY’S LAWS
In the early nineteenth century, Faraday stated the laws:
1. The mass of any substance deposited or dissolved at an electrode is pro-

portional to the quantity of electricity passed through the solution.
2. One equivalent weight of any substance deposited or dissolved at an elec-
trode requires 96,485 coulombs of electricity.
(The modern value of 96,485 coulombs has been substituted for Faraday’s original
value which differed slightly.) If the quantity 96,485 coulombs is divided by the
charge on the electron (measured about a century later), one obtains the Avogadro
number. The explanation of this is that each ion carries an integral number of electrons
(negative) or lacks an integral number of electrons (positive). In honor of one of the
early geniuses of physical chemistry, the quantity 96,485 coulombs is defined as the
faraday, F.
13.7 MOBILITY AND CONDUCTANCE
If we look at transport numbers, mobility, and ionic conductance, it should be evident

that a fast-moving ion carries more of the total charge transported than a slow one.
Also, transport numbers are the fraction of electrical charge carried by each ion, so
they must add up to 1. In an aqueous solution of a univalent salt, t+ + t− = 1.0 and
the sum of the molar (or molal) ionic conductances is ◦ = λ◦+ + λ◦− . The fraction
of current carried by an ion can be given by either the transport numbers or the ratio
of ionic conductivity λ◦± to the total molar conductivity ◦ , each at infinite dilution:
λ◦+ λ◦−
t+ = , t− =
◦ ◦
In an electrolyte of faster and slower ions, the faster ion carries a greater proportion
of the total charge transfer than the slower ion in proportion to their relative speeds.
Therefore the ratio of transport numbers is the same as the ratio of ionic mobilities:
t+ u+
=
t− u−
13.8 THE HITTORF CELL
Transport numbers can be determined directly by measuring the change in concen-

tration brought about by transference in a compartmentalized cell called a Hittorf
cell (Fig. 13.5). A Hittorf cell consists of three compartments: an anode compart-
ment, a cathode compartment, and a central compartment between them. A potential
is imposed on a solution in the cell containing, for example, Ag+ ions, which are
P1: OTA/XYZ P2: ABC
Ag+
Anode Ag+ Cathode

Ag+
FIGURE 13.5 A three-compartment Hittorf cell.
deposited (electroplated) on the cathode. We can understand the use of the cell by
taking two extreme hypothetical cases:
1. If there were no transference of charge by Ag+ moving into the cathode com-
partment from the central compartment, the concentration would be reduced
by one equivalent (gram atomic weight) per Faraday passed through the cell.
2. If silver ions carried all the charge through the Hittorf apparatus, all the Ag+
ions lost at the cathode would be replaced by Ag+ ions migrating into the
cathode compartment from the central compartment. The concentration of
Ag+ ions would be unchanged.
In fact, the actual concentration in the cathode compartment is altered by less

than one equivalent per coulomb but more than zero. The number of equivalents
per coulomb transferred into the cathode compartment is the transference number
of Ag+ . One can determine the transport numbers of both ions for example, Ag+
and NO− 3 , along with their mobilities by a simple experiment (but one that requires
precise analytical technique).
The conductance of a solution is the amount of charge in coulombs passing through
it between electrodes at a given potential difference in volts per unit length of the
resistor in units of V m−1 . From our idea that current flows through a solution of
electrolyte due to passage of ions and from Faraday’s proof that each ion carries an
integral multiple of the unit charge e− , the controlling factor must be the mobility
with which ions move through the solvent, water in the case at hand. In fact the
equivalent ionic conductance at infinite dilution λ◦ and the ionic mobility are directly
proportional:
λ◦+ = Fu ◦+
and
λ◦− = Fu ◦−
Taking values for Ag+ (aq) and NO−

3 (aq) (Example 13.3), we have
λ◦+ 62.6 × 10−4

u ◦+ = = = 64.9 × 10−8 m2 V−1 s−1
F 96,485
P1: OTA/XYZ P2: ABC
ION ACTIVITIES 213
and
λ◦− 70.6 × 10−4

u ◦− = = = 73.2 × 10−8 m2 V−1 s−1
F 96,485
These unusual units deserve a note. The units of speed m s−1 can be split off:

m2 V−1 s−1 = mV−1 m s−1
and
−1 1
ms mV−1 = m s−1 −1
Vm
In other words, the unit is a typical speed in meters per second per unit voltage
drop measured in volts per meter in the MKS system for the separation between the
electrodes in the cell.
13.9 ION ACTIVITIES
One would like to have activities and activity coefficients for single ion concentrations
to use just as activities, and activity coefficients were used for single molecule
concentrations in Chapters 7 and 12. Unfortunately, this is not possible because
single ions cannot be observed in the absence of a partner ion necessary to preserve
electroneutrality. Faced with this dilemma, we make an approximation that ionic
activities are equal: a+ = a− . This is obviously not true, but it is the best we can do.
√
The best we can find is an average activity, the geometric mean, a± ≡ a+ a− . Other
√
definitions follow as they did for simple molecules: a± = γ± m, γ± ≡ γ+ γ− , and

a+ a− a±
γ± ≡ =
m m m
Debye and Huckel have given a rather involved treatment of the electrostatic forces
acting on an ion surrounded by other ions. They arrived at an expression for γ± in
the form of the Debye–Hückel limiting law:

ln γ± = −1.172 |Z + Z − | µ̃
where – 1.172 is a combination of constants, |Z + Z −√| is the product of atomic charges

(|Z + Z − | = 1.0 in the cases we consider here), and µ̃ is a concentration term which
is the summation of all the surrounding electrolytes, not just the electrolyte for which
we seek γ± . The symbol µ is almost universally used to denote both Debye–Hückel
ionic strength and Gibbs chemical potential. To diminish this source of confusion,
P1: OTA/XYZ P2: ABC
we are using the notation µ̃ for the ionic strength:

µ̃ = 1
2
m i Z i2
i
The Debye–Hückel law is applicable only in very dilute solutions; it is a typical

limiting law.
We know how to determine the ionization constant K a for weak acids, say acetic
acid, but we find upon close examination that the results are not quite constant with
different concentrations of HOAc, and we presume that this lack is due to interference
among ions. This nonideality can be expressed in terms of ionic activities:
aH+ aOAc− γ± m H+ γ± m OAc−

Ka = =
aHOAc γ m HOAc
In this model, nonideality is caused by charge interaction but HOAc is uncharged, so

we can consider it an ideal solute and take γHOAc = 1.0. The ions, however, are not
ideal. This means that γ± = 1.0, so
aH+ aOAc− γ± m H+ γ± m OAc− m2 +

Ka = = = γ±2 H = γ±2 K
aHOAc γHOAc m HOAc m HOAc
where m OAc− = m H+ because equal numbers of product ions H+ and OAc− are
produced in the dissociation reaction
→ H+ + OAc−
HOAc ←
Taking logarithms of both sides, we obtain
ln K a = 2 ln γ± + ln K
where K is a measured (nonideal) value for the acid dissociation constant and K a is
the ideal value. At infinite dilution, γ± = 1.0 and ln γ± = 0.0, so
ln K = ln K a

m→0
For a dilute acid in pure water, each ion concentration is αm, where α is the
degree of dissociation and m is the molar concentration of acid (Section 13.4). The
ionic strength is determined by the concentration of ions in the solution, but this is
by no means the bulk concentration of acid. Many acids—for example, acetic acid
HOAc—ionize to a very limited extent, but even ionization of pure water contributes
to the ionic strength. According to the Debye–Hückel limiting law, we have
√
ln γ± = −1.172 µ̃ = −1.172 αm
P1: OTA/XYZ P2: ABC
for the acid. The acidity (ionization) constant is
ln K a = 2 ln γ± + ln K
√ √
ln K a = 2 −1.172 αm + ln K = −2.34 αm + ln K
or
√
ln K = −2.34 αm + ln K a
√
A plot of ln K vs. αm gives an intercept of ln K a and has a limiting√slope of 2.34
as α → 1. As m becomes very small, the middle term of ln K = 2.34 αm + ln K a
drops out and the measured value of K approaches K a .
Example 13.1
Suppose you have access to a well-equipped electrochemistry laboratory but you do
not know the value of the faraday. How would you determine the value of the faraday
to four significant figures?
Solution 13.1 To work the problem de novo would be a pretty hard job. Remember,
Faraday was a genius. When you electrolyze different solutions, however, you will
soon find that many metals including silver can be plated out on an electrode by
passage of current through a solution of, for example, silver nitrate. A coulomb is
an ampere second. The idea is this: Measure the current flow I times the time t in
seconds to force the quantity Q of electricity in coulombs through an electrolysis cell
to plate the silver. Then measure the number of moles of silver that have been plated
out. The fraction I × t/moles Ag gives the faraday in coulombs mol−1 . Your first try
at this method will probably be unsuccessful because the faraday is a big number.
It takes more electricity than you might guess to deposit an amount of silver that is
measurable to four figures of accuracy.
Many electrochemical laboratories are equipped with a variety of potentiostats that
maintain a constant voltage across an experimental cell and amperostats that maintain
a constant current. For our purposes, an amperostat would be most convenient because
integration to find the area under an I vs. t curve would merely be multiplying constant
I into the time. Both of these measurements should be possible with high accuracy.
Silver has been known for many years to be plated out in 100% current efficiency,
so if we multiply our I measurement into the time of electrolysis, both measured
digitally, we get the number of coulombs passed through the cell. We can then weigh
the silver electrode and find out how much it has increased in weight due to the silver
deposited.
To get a rough idea of the parameters involved, suppose we have a microbalance
sensitive and accurate enough to weigh to ±0.01 mg. A silver weight of 0.2 g would
P1: OTA/XYZ P2: ABC
be within our measurement range. If the amount of current we can maintain is about
50 milliamperes (mA), we know everything in the equation except the time. Trials
show that it takes about an hour (1 h = 3600 s) to deposit 0.2 g of Ag.
I ×t 0.050 × 3600 180

F= = = = 97,110
moles Ag 0.2 0.00185
107.9
Now in our precise one-hour experiment, the real amperostat reading is 49.883 mA
over the time period of 1.000 h = 3600.0 s, and the weight of silver gained by the
(carefully dried) silver cathode is 0.20149 g
coulombs I ×t 0.049883 (3600.00) 179.5788

F= = = = = 96,367
mol mol 0.20101 0.00186534
107.868
which is a little more than 0.1% in error.
Example 13.2
Suppose you are in the unlikely circumstance of needing to know the faraday to nine
significant figures. Propose a way of calculating F from fundamental constants.
Solution 13.2 Unknown in Faraday’s time, of course, unit charge is 1.602176487 ×

10−19 C and Avogadro’s number is N A = 6.02214179 × 1023 . Since the faraday is
the charge per mole of (singly charged) particles, we have

F = 6.02214179 × 1023 1.602176487 × 10−19 = 96485.3398 C mol−1
These data are from the handbook, but you will get some argument on the last two
digits.
Example 13.3
(a) A Hittorf cell having 0.1000 mol of AgNO3 (aq) in each of its three compart-
ments had 0.0100 faradays of charge passed through it. A reduction reaction
depositing solid silver, Ag(s), took place at the cathode:
Ag+ (aq) + e− → Ag(s)
Analysis of the contents of the cathode compartment after the experiment

showed that 0.00947 mol of Ag+ (aq) remained. What are the transport num-
bers of Ag+ (aq) and NO−
3 (aq)?
P1: OTA/XYZ P2: ABC
(b) Given the additional information that ◦AgNO3 = 133.3 × 10−4 ohm−1
m2 mol−1 , what are the molar ionic conductivities and mobilities of Ag+ (aq)
and NO−
3 (aq)?
Solution 13.3
(a) If the Ag+ (aq) ions did not move, the amount of silver ion found in the
cathode compartment after electrolysis would have been 0.0900 mols because
of the removal of 0.0100 mol of Ag+ (aq) by electrodeposition. The fact that
there are 0.0947 mol in the cathode compartment after deposition means that
0.0047 mol of Ag+ (aq) have been replaced by Ag+ (aq) migration from the
center compartment. The transport number is the ratio of the charge transferred
by Ag+ (aq) transport relative to the total charge transferred by both ions:
0.0047
t+ = = 0.47
0.0100
The rest of transport must have been due to NO− 3 (aq) ions migrating in the
opposite direction, so t− = 1.00 − 0.47 = 0.53.
(b) Transport numbers are related to ionic conductances according to the equations
λ◦+ λ◦−
t+ = t− =
◦ ◦
which is reasonable because an ion that transports more than its share of
charge will have a large ionic conductivity relative to the total ◦ . Given that
◦AgNO3 = 133.3 × 10−4 ohm−1 m2 mol−1 , we obtain

λ◦+ = t+ ◦ = 0.47 133.3 × 10−4 = 62.6 × 10−4 ohm−1 m2 mol−1
and

λ◦− = t− ◦ = 0.53 133.3 × 10−4 = 70.6 × 10−4 ohm−1 m2 mol−1
Problem 13.1
A solution containing 1.155 × 10−2 mol of HCl was neutralized by adding NaOH
solution. The process brought about a 0.553 K temperature rise. An electrical heater
of 58.7-ohm resistance brought about a 0.487 K temperature rise in 15.0 min. A
potential of 6.03 volts was maintained across the resistor by a potentiostat. What is
the molar enthalpy of HCl? Remember that a joule is a volt coulomb.
P1: OTA/XYZ P2: ABC
Problem 13.2
What is the charge on the silver ions in 1.000 dm3 of dissolved Ag+ in aqueous
solution equilibrium with Cl− and solid AgCl(s) at 298 K? The solubility product
constant of AgCl is K sp = 1.77 × 10−5 . There is no outside source of Cl− ion.
Problem 13.3
l-Dopa (l-(dihydroxyphenyl)alanine) is used to treat the symptoms of Parkinson’s
disease. It is quantitatively and exhaustively reduced in an electrolysis cell, requir-
ing two electrons per molecule. Its molar mass is M = 0.1972 kg mol−1 . Suppose
a sample requires 57.5 µC (microcoulombs) for complete reduction. How much
l-dopa was in the sample?
Problem 13.4
The resistivity of copper at 298 K is ρ = 1.71 × 10−8 ohm m. A potential of 1.00 mV
was placed across a copper wire 1.00 mm in diameter and 1.00 m long. What was the
current?
Problem 13.5
(a) The resistivity of copper at 298 K is ρ = 1.71 × 10−8 ohm m. Find the
conductance and the conductivity of a copper wire 1.00 mm in diameter
and 1.00 m long.
(b) Find the conductance and the conductivity of a copper wire 2.00 mm in
diameter and .500 m long.
Problem 13.6
A conductivity cell is constructed using two platinum electrodes, precisely machined
to be a square 2 cm on an edge and 1 cm apart. A 0.1000 mol dm−3 HCl solution was
poured into the cell, and the resistance was measured and found to be 6.3882 ohms.
What is the molar conductivity of HCl at this concentration? In solving this problem,
be careful to express the units clearly at each step.
Problem 13.7
We can write the ionization of water in a somewhat simplified form as
H2 O → H+ + OH−
The conductivity of deionized water is κ = 5.50 × 10−6 ohm−1 m2 mol−1 .
(a) Find the concentration c of pure water in mol dm−3 .

(b) Find the degree of ionization α.
P1: OTA/XYZ P2: ABC
(c) Find the equilibrium constant K w for the dissociation of pure water.
(d) Find the concentration of hydrogen ions H+ in pure water.
(e) Why is it meaningful to calculate the equilibrium constant for H+ even though
we know that H3 O+ , H5 O+ , . . . also exist in pure water?
Problem 13.8
The molar conductivities of NaCl are given in the handbook as
mol dm−3 Conductivity mol dm−3 Conductivity
5.0000e-4 124.4400 1.0000e-3 123.6800

5.0000e-3 120.5900 0.0100 118.4500
0.0200 115.7000 0.0500 111.0100
0.1000 106.6900
Plot these data in the form of a Kohlrausch plot. Determine the value of ◦NaCl and
the slope of the linear extrapolation to ◦NaCl .
Problem 13.9
The molar conductivities ◦ of NH4 Cl, NaOH, and NaCl are 149.6, 247.7, and
126.4 in units of 104 S m2 mol−1 . Find ◦NH4 OH .
Problem 13.10
Show that the units of conductivity κ are ohm−1 m−1 .
Problem 13.11
One meaning of the term Gibbs free energy is that G is the maximum work that
can be obtained from a chemical reaction. Conversely, the negative of G is the
minimum work that must be put into a chemical system to reverse its direction. Given
that G = −237 kJ mol−1 for the spontaneous reaction (note the sign of G)
H2 (g) + 12 O2 (g) → H2 O(l)
and the fact that work w = charge(V) where the charge is in coulombs and V is the
voltage, find the minimum potential in volts necessary for the electrolysis of water to
form H2 (g) + 12 O2 (g).
P1: OTA/XYZ P2: ABC
14
ELECTROCHEMICAL CELLS
Advances in electrochemistry came at least a generation earlier than corresponding

advances in rigorous thermochemical theory. Even prior to Faraday’s numerous dis-
coveries, Alessandro Volta’s experiments on static electricity and electrical current
took place largely in the eighteenth century and culminated in the first true battery of
series voltaic cells. The battery was called a voltaic pile because cells were piled on
top of one another. Volta’s work excited great popular interest, especially when he
showed that Luigi Galvani’s use of a voltaic pile to activate nervous response in the
dissected leg of a dead frog depended on the electrical potential of the pile and was
not a property of the frog’s leg. Activation of a dead animal raised the question of
whether science might someday actually produce life.1 News of Volta’s experiments
became something of a fad of the day. Mary Shelley’s popular creation Frankenstein
was presumably electrochemically activated.
14.1 THE DANIELL CELL
If you put a zinc rod into a solution of zinc sulfate ZnSO4 , there will be a difference
in chemical potential between the Zn(s) in the metal rod and the Zn2+ (aq) ion in the
aqueous solution. A reaction takes place, causing a minute amount of Zn to dissolve
and go from the rod into the solution, leaving two electrons behind or, depending on
1 See Gibson, D. G. et al. 2010. Science, 329, 52–56 for a claim that science has produced life.

Copyright
220
P1: OTA/XYZ P2: ABC
HALF-CELLS 221
concentration, causing Zn2+ (aq) to go from the solution onto the rod, taking up two
electrons from the metal:
Zn(s) → Zn2+ (aq) + 2e−
or
Zn2+ (aq) + 2e− → Zn(s)
Either way, an electrical potential difference is produced between the electrode and the
solution such that the sum of the chemical potential and the electrical potential, called
the electrochemical potential, on one side balances the electrochemical potential on
the other side, and equilibrium is reached.
If the same thing is done using a copper rod in contact with a CuSO4 solution,
an electrochemical balance is quickly reached in the same way, but the electrical
potential on the Cu electrode is not the same as it was on the Zn electrode because
the chemical potential of solid Cu is not the same as it is for solid Zn(s). Zn(s) is said
to be more active than Cu(s) because it has a greater tendency to dissolve in a dilute
solution of its ions than Cu(s). If a Zn; Zn2+ electrode at potential φ is connected to a
Cu; Cu2+ electrode across a digital voltmeter and the circuit is completed by means
of a neutral salt bridge, the difference in potential, called the voltage of the cell, can
be measured. It turns out to be about 1.1 volts, depending on the concentrations of the
ZnSO4 and CuSO4 in the electrode compartments. The entire arrangement is called
a Daniell cell.2
If a resistance or small motor replaces the digital voltmeter, heat or work can be
obtained. This is the principle of all electrochemical cells. Historically, two or more
cells connected were called a battery of cells by analogy to gun batteries. What we
call the 1.5-volt AA “battery” is really a single cell.
14.2 HALF-CELLS
There are a great variety of cells available for research and practical use. The Cu; Cu2+
and Zn; Zn2+ electrodes can be replaced by other metals, leading to many metal–metal
ion combinations. These M; Mz+ systems, where z is the number of electrons in-
volved in the electrode reaction, are called half-cells because any two of them can
be combined to form a cell. The semicolon is used to denote the possibility of
electron exchange and is often, but not necessarily, a physical interface such as the
solid;solution interface in Zn; Zn2+ . The electrochemical standard is the hydrogen
half-cell (Fig. 14.1). Hydrogen is allowed to pass through the cell under 1 atm
2 Invented as early as 1837. The original Daniell cell was a copper cup containing the CuSO4 in which
a porous cup containing the Zn;Zn2+ solution was suspended. The porous cup soaked with salt solution
took the place of the salt bridge. The Daniell cell had many practical uses and played an important part in
early electrochemistry.
P1: OTA/XYZ P2: ABC
222 ELECTROCHEMICAL CELLS
H2
H+
H2
Pt
FIGURE 14.1 The hydrogen half-cell.
(or 1 bar) pressure so that the surface of the platinum electrode is alternately bathed
by H2 (g) and an acid solution containing H+ (aq)ion. The H2 (g) is adsorbed onto
the Pt surface (Chapter 11) so that what the H+ (aq) ion “sees” appears to be a solid
surface of H2 (g). The electrode reaction is comparable to a metal ion reaction, except
that now it is a hydrogen–hydrogen ion reaction:
H2 (g) → 2H+ (aq) + 2e−
Hydrogen half-cells can be connected to other half-cells, and the resulting voltage
can be measured as before.
Half-cell potentials cannot be measured by conventional means. Instead, the
hydrogen half-cell is chosen as a standard and is arbitrarily assigned a potential
of 0.000 volts. Now cells can be constructed from the hydrogen half-cell and any
other half-cell. The whole-cell potential is measured and, the contribution from
H2 (g); H+ (aq) being zero by definition, the entire cell potential is assigned to the
other half-cell. In this way, an electrochemical series of many half-cell potentials
is built up, going from active metals like Zn; Zn2+ near the top to electrodes
like Cu; Cu2+ = +0.337 volts farther down. The Cu;Cu2+ is more positive than
H2 (g); H+ (aq) and the Zn;Zn2+ = −0.763 volts is more negative; hence the
difference between them is 0.337 − (−0.763) = 1.10 volts, which accounts for the
1.1 volts observed for the Daniell cell. All this, of course, requires a definition of the
concentration of the metal (or other) ions in question. Although other conventions are
sometimes used, the standard half-cell potentials are defined at ion activities of 1.00.
14.3 HALF-CELL POTENTIALS
Many half-cells have been studied in the long and rich history of electrochemistry.
They are, by convention, listed as reduction potentials involving addition of one or
P1: OTA/XYZ P2: ABC
CELL DIAGRAMS 223
TABLE 14.1 A Few Selected Reduction Potentials.
Half-Cell Reaction E ◦ (volts)
K+ ; K(s) K+ + e− → K(s) −2.925

Zn2+ ; Zn(s) Zn2+ + 2e− → Zn(s) −0.763
Cd2+ ; Cd(s) Cd2+ + 2e− → Cd(s) −0.403
H+ ; H2 (g) 2H+ + 2e− → H2 (g) 0.000
Cl− ; AgCl(s); Ag AgCl + e− → Ag(s) + Cl− 0.222
Cu2+ ; Cu(s) Cu2+ + 2e− → Cu(s) 0.337
Fe3+ ; Fe2+ ; Pt Fe3+ + e− → Fe2+ 0.771
Ag+ ; Ag(s) Ag+ + 2e− → Ag(s) 0.799
more electrons causing reduction of the oxidation number, one for K+ (aq) + e− →
K(s), two for Cu2+ (aq) + 2e− → Cu(s), and so on (Table 14.1).
14.4 CELL DIAGRAMS
Certain conventions serve to simplify the written description of electrochemical cells.

Following convention, the half-cell that is higher in the table goes on the left and the
half-cell that is lower in the table goes on the right. We write the right half-cell as it
appears in the table and reverse the left half-cell to give a diagram for the whole cell.
In the case of the Daniell cell, this is
Zn(s); Zn2+ (aq)||Cu2+ (aq); Cu(s)
where the double line represents a salt bridge. The cell reaction is the sum of half-cell
reactions with the left-hand half-cell reversed:
Zn(s) → Zn2+ (aq) + 2e−

Cu2+ (aq) + 2e− → Cu(s)
Zn(s) + Cu2+ (aq) → Cu(s) + Zn2+ (aq)
The two electrons cancel upon addition. The electrode at which reduction occurs
(remember that half-cell potentials are listed as reductions) is defined as the cathode.
The other electrode is the anode.
The voltage of the whole cell is the sum of the half-cell potentials with the left
half-cell once again reversed:
E ◦ = E ◦1 R − E ◦1 L = 0.337 − (−0.763) = 1.10 volts

2 2
P1: OTA/XYZ P2: ABC
This is consistent with our treatment of thermodynamic potentials, for example, we

have
G ◦ = G ◦ (B) − G ◦ (A)
for the reaction
A→B
The cell reaction, which involves reduction at one electrode and oxidation at the other,
is called a reduction–oxidation reaction, or simply a redox reaction. Redox reactions
are spontaneous if written according to the conventions agreed upon.
14.5 ELECTRICAL WORK
A joule is a volt coulomb, so an electrochemical cell operating at 1.0 volt does 1.0 J
of work for every coulomb of electricity it produces. Don’t forget that the coulomb
C is an amount of charge placed in a capacitor by a specific current in a specific
time C = It. Therefore the number of coulombs is proportional to the number of
electrons driven through a resistance or motor by the cell reaction. If one electron
is exchanged in the cell reaction and one mole of reactant is used up, one mole of
electrons is exchanged, and one faraday of charge is produced. The work produced
is 96,485 joules. In general, n electrons are exchanged and the cell potential is
E volts, so
w = nFE = −G
because the amount of work done per mole of reactant consumed is the molar
decrease in the Gibbs free energy. This leads to the important connection between
thermodynamics and electrochemistry:
G = −nFE
14.6 THE NERNST EQUATION
For ideal solutions we have
G = G ◦ + RT ln Q
where Q is the equilibrium quotient, not to be confused with the charge Q. Therefore
nFE = nFE◦ − RT ln Q
P1: OTA/XYZ P2: ABC
CONCENTRATION CELLS 225
or
RT
E = E◦ − ln Q
nF
which is the Nernst equation. At 298 K, RT and F can be combined to find
0.0257
E = E◦ − ln Q
n
The meanings of G and G ◦ enable us to better understand E and E ◦ . As

G is the free energy change for any arbitrary equilibrium quotient Q, E represents
the potential obtained from a cell under these arbitrary conditions. If the reactants
and products are all in their standard states, the difference in free energy between
reactants and products is G ◦ . Hence E ◦ represents the potential of a cell in which
the reactants and products in the redox reaction are in their standard states and at unit
activity. This is just the cell potential from a table of standard half-cell potentials.
Knowing the concentrations of the constituents of any cell not at unit activity, one
can calculate E ◦ from the table and correct it to find the actual cell potential using
the Nernst equation. For example, if the activities in the electrode reaction are
Cu(s) + 2Ag+ (aq, 0.01 m) → Cu2+ (aq, 0.1 m) + Ag(s)
the equilibrium quotient is Q = Cu2+ (aq, 0.1 m)/Ag+ (aq, 0.01 m)2 = 0.1/0.012 =
1000. The standard cell potential is E ◦ = 0.799 − (0.337) = 0.462. Applying the
Nernst equation with n = 2, we obtain
0.0257
E = E◦ − ln Q = 0.462 − 0.0128(6.908) = 0.373 volts
n
Much of the older literature uses the logarithm to the base 10 in the Nernst equation.
Conversion from ln to log10 requires multiplication of the ln Q term by 2.303 to yield
0.0257 2.303 (0.0257)

E = E◦ − ln Q = E ◦ − log Q
n n
0.0592
= 0.462 − 3 = 0.343 volts
2
All of this assumes ideal behavior of course, as well as that a = m throughout. We
should be skeptical. More about this later.
14.7 CONCENTRATION CELLS
The Nernst equation suggests that we can have an electrochemical cell with half-
cells that are identical in every way except for the concentrations of the ions in
P1: OTA/XYZ P2: ABC
the electrode compartments. There will be a measurable potential that depends only
on the difference in concentrations of the aqueous solutions. Such cells are called
concentration cells.
An example would be any metal;metal ion pair with ion concentrations, say
Ag;Ag+ (aq) of 0.10 molar in one half-cell and 0.010 m in the other:
Ag(s);Ag+ (0.010 m)| |Ag+ (0.10 m); Ag(s)
The cell reaction is the half-cell potential of the cathode minus that of the anode (so
as to give a positive whole-cell potential):
Ag+ (0.10 m) → Ag+ (0.010 m)
that is, the direction of spontaneous change for any solution is to become more dilute.
(Solutions don’t spontaneously “concentrate themselves.”) Q is 0.010/0.10 = 0.10
and the Nernst equation is
E = 0 − 0.0257 ln 0.10 = 0.0592 V = 59.2 mV
If the ion concentration in one of the half-cells is known and the other is unknown,
the concentration of the unknown half-cell can be determined from the whole-cell
potential. This is the principle behind the most widely used application of electro-
chemistry, the pH meter, in which a potential is measured that is linearly related
through the Nernst equation to the logarithm of the hydrogen ion concentration in a
half-cell. The negative logarithm (log base 10) of the hydrogen ion concentration is,
as we learn in elementary chemistry, the pH. pH meters consist of an H+ -sensitive
(glass) electrode and a constant voltage (calomel) electrode with suitable electronics
to measure and display or record their algebraic sum. pH meters are calibrated to
display the hydrogen ion concentration directly in pH units, most frequently as a
digital readout.
14.8 FINDING E◦
A silver electrode immersed in a solution of HCl soon becomes covered by a coating

of AgCl(s). This half-cell, connected to a standard hydrogen electrode gives the cell
Pt(s); H2 (g); HCl(aq); AgCl(s); Ag(s)
The cell has no need for a salt bridge. One electrode is sensitive to H+ and the other
is sensitive to Cl− which is precisely equal in concentration. The cell reaction is
→ H+ (aq) + Cl− (aq) + Ag(s)

AgCl(s) + 12 H2 (g) ←
P1: OTA/XYZ P2: ABC
FINDING E◦ 227
The Nernst equation for this cell is
RT aAg aH+ aCl −

E = E◦ − ln 1
F aAgCl aH2 2
We can set three of these five activities equal to 1.0 because two of them refer to solids
AgCl(s)and Ag(s) and the other has an activity a H2 = 1.0 at a hydrogen pressure
of 1.0 atm (or 1.0 bar). The remaining activities are aH+ aCl− = γ± m H+ γ± m Cl− =
γ± 2 m HCl 2 , provided that HCl is completely ionized (which it is) and there are no
other sources of H+ or Cl− . Now the Nernst equation reads
RT RT
E = E◦ − ln aH+ aCl− = E ◦ − ln γ± 2 m HCl 2
F F
2RT 2RT
= E◦ − ln γ± − ln m HCl
F F
or
2RT 2RT
E+ ln m HCl = E ◦ − ln γ±
F F
The right-hand side of this equation approaches E ◦ as m approaches zero because

γ± → 1.0. Plotting E + (2RT /F) ln m HCl as a function of m and extrapolating to
m = 0 is appealing, but it is not quite the way the √ problem is solved. Debye–Hückel
theory (Section 13.9) says that ln γ± = −1.171 µ̃ where, in this case, µ̃ = m near
infinite dilution. Let us call the term on the left E for convenience in plotting, so that

2RT
E = E+ ln m H Cl
F
whereupon
2RT (1.171) 1/2

E = E◦ − m
F
Now we need only plot E as a function of m 1/2 to obtain an intercept of E ◦ (Fig. 14.2).
The plot will yield a straight line with a slope of −2RT (1.171)/F but only in
the limit of infinite dilution. The extrapolation has been carried out with consid-
erable precision. It yields 0.22239 volts for the silver chloride–hydrogen cell and,
since the standard hydrogen electrode has a half-cell potential of zero by definition,
this is equal to the half-cell potential of the silver–silver chloride half-cell listed in
Table 14.1.
P1: OTA/XYZ P2: ABC
0.2255
0.2250
0.2245
0.2240
E' (volts)
0.2235
0.2230
0.2225
0.2220
0.00 0.01 0.02 0.03 0.04 0.05 0.06
1/2
m
FIGURE 14.2 Extrapolation to E ◦ = 0.2223 for the standard hydrogen–silver–silver chlo-

ride Cell. (Standard value: 0.22239; data from Klotz and Rosenberg, 2008.) At higher concen-
trations the experimental points fall away from the linear function.
14.9 SOLUBILITY AND STABILITY PRODUCTS
In general chemistry we memorized the solubility product constant of silver chloride,

K sp = [Ag+ ][Cl− ] = 10−10 , which implies that the concentration of silver ion in a
solution of AgCl in pure water is [Ag+ ] = 10−5 . How do you make an accurate
measurement of the concentration of an ion that is 0.00001 molar (or molal)? The
question becomes even more daunting for the case of copper phosphate, K sp = 10−37 .
Because the Nernst equation relates the cell potential to the logarithm of the
concentration, very small metal ion concentrations such as those in saturated solutions
of sparingly soluble salts and stable complexes can be measured. For example, we can
set up a cell consisting of a silver–silver ion half-cell in opposition to a silver–silver
iodide half-cell which is analogous to the silver–silver chloride half-cell discussed in
the previous section. The cell diagram and cell reactions are
Ag(s); AgI(s); I− (aq)||Ag+ (aq); Ag(s)

Ag+ (aq) + I− (aq) ←→ AgI(s)
The equilibrium constant for the cell reaction is
aAgI 1
K eq = =
aAg aI
+ − K sp
P1: OTA/XYZ P2: ABC
THE CALOMEL ELECTRODE 229
where aAgI of the solid precipitate has been set equal to 1.0 and K sp is the solubility
product of the AgI precipitate. The potential of this cell is 0.950 volts and, solving
the Nernst equation, we find log K sp = −0.950/0.0592 = −16.0, so K sp ∼ = 10−16 .
Notice that we use 59.2 mV in the Nernst equation to calculate log10 K sp (Section
14.7). Stability constants of complexes, in which metal ion concentrations are reduced
to very low levels by ligands like EDTA, can be measured by a method analogous to
the one shown for AgI.
14.10 MEAN IONIC ACTIVITY COEFFICIENTS
Back calculating from the procedure of Section 14.8 produces γ± at molalities other
than zero. From
2RT 2RT
E = E◦ − log γ± − log m HCl
F F
E = E ◦ − 0.1184 log m HCl − 0.1184 log γ±
and
0.1184 log γ± = −E + E ◦ − 0.1184 log m HCl

−E + E ◦ − 0.1184 log m HCl
log γ± =
0.1184
At finite molality, E can be measured and everything on the right of the equal sign is
known because E ◦ is known.
14.11 THE CALOMEL ELECTRODE
The standard hydrogen half-cell has the advantage that its potential is 0.000 by
definition, but it is bulky and presents some safety hazards in the use of hydrogen.
Consequently, several half-cells that do not have these disadvantages are used as
reference electrodes. The most common one is the calomel electrode (strictly, half-
cell), which is easily set up, is light, and gives a constant voltage at a fixed temperature.
Calomel half-cells have liquid mercury in contact with a paste of Hg2 Cl2 (s) and solid
KCl in equilibrium with Cl− (aq). Connection to external circuitry is by a Pt wire
in contact with the mercury. Connection with any of a variety of other half-cells
is made through a saturated KCl solution called a salt bridge. The saturated KCl
solution does not change much upon drawing a small current, so the output voltage
is a constant 0.2444 volts at 298 K. The electrode is designated a saturated calomel
electrode (SCE) from the common name calomel for mercurous chloride (which was
once used as a medicine for children!!). The cell diagram
Pt; H2 (1.0 atm); H+ (aq)||Hg2 Cl2 (s), KCl(aq sat d); Hg; Pt
P1: OTA/XYZ P2: ABC
has a Nernst equation of
E cell = E calomel − 0.0257 ln H+ (aq)
because all of the activities other than H+ (aq) are 1.0 or are included in E calomel . This
leads to
E cell − E calomel
− ln H+ (aq) =
0.0257
or, as is more commonly written,
E cell − E calomel
pH ≡ − log H+ (aq) =
0.0592
These results will not be accurate to an indefinite number of significant figures

because the activity coefficient is not exactly 1.0 in real solutions. Cell potentials can
be measured to four or more significant figures, but pH values should be regarded with
a prudent degree of skepticism. Conversely, departure of pH values from expectation
can be used to estimate activity coefficients.
14.12 THE GLASS ELECTRODE
We still have the problem of a bulky, somewhat dangerous hydrogen half-cell to de-
termine pH in the field. For practical work, the glass electrode is used instead. Within
certain limits, the potential between opposite sides of a very thin glass membrane
varies in a linear way with hydrogen ion concentration. This is used in to determine
pH, but the nature of the hydrogen ion sensitivity is not rigorously understood beyond
knowing that there is an interchange between the aqueous hydrogen ions and the Na+
ions in the glass on one side of the membrane that is different from the interchange
on the other. Consequently, pH measurements have only relative significance. We
measure pH by the potential relative to that due to a standard buffer solution for
which we take the pH to be known. Glass electrodes can be made that are quite
small, and the hand-held pH meter is common in many branches of laboratory and
field chemistry. Other types of membranes lead to various electrodes sensitive to ions
other than H+ (aq)—for example, Ca+ (aq).
Example 14.1 pH from the Hydrogen Half-Cell

The pH meter with a glass electrode is by no means the only way of determining pH.
Suppose we reverse the method of Section 14.8 and use the half-cell potential of the
P1: OTA/XYZ P2: ABC
Ag;AgCl electrode in the cell
Pt(s); H2 (g, 1 atm); HCl(aq = ?); AgCl(s); Ag(s)
to determine the concentration of HCl(aq = ?). If the cell potential for HCl(aq = ?)
is 493 mV, what is the pH of the solution in which the electrodes are immersed?
Solution 14.1 The cell reaction is
→ H+ (aq) + Cl− (aq) + Ag(s)

AgCl(s) + 12 H2 (g) ←
The Nernst equation is
0.0592
E = E◦ − log m HCl = 0.493 volts
1
because of the unit activities of Ag(s) and AgCl(s) and pH2 = 1 atm. We know E ◦
and we take m H+ = m Cl− because each HCl breaks up to give one of each kind of ion.
We take m as the geometric mean of the ion concentrations m HCl = m H+ m Cl− = m 2H+ :
0.493 = 0.222 − 0.0592 log m 2H+

0.493 − 0.222 = 2 (0.0592) (− log m H+ )
0.271 = (0.1184) (−log m H+ ) = (0.1184) (pH)
pH = 0.271/0.1184 = 2.29
Comment: The pH is not carried out to the same five-digit accuracy that the Ag;AgCl
electrode presents because a number of approximations appear in the development of
the pH equation. One is that the ionic interference between the dissociated HCl ions
is not taken into account because the ions are very dilute and, second, dissolution of
AgCl causes a slight imbalance between the H+ and Cl− ions.
Example 14.2 A Mean Activity Coefficient

If a silver chloride–hydrogen cell with m HCl = 0.1000 produces an electrical potential
of E = 0.353, what is the mean activity coefficient of H+ at this molarity?
Solution 14.2 Notice that log m HCl = log(0.1000) = −1, so
−E + E ◦ − 0.1184 log m HCl

log γ± =
0.1184
P1: OTA/XYZ P2: ABC
1.00
0.95
0.90
γ±
0.85
0.80
0.75
0.0 0.1 0.2 0.3 0.4 0.5
m 1/2
FIGURE 14.3 The mean activity coefficient of HCl as a function of m 1/2 . The value γ± =
−0.789 calculated in Exercise 14.2 is the penultimate point on the curve reading from left to
right. The standard value is γ± = −0.797.
becomes
−E + E ◦ + 0.1184 −.353 + .2224 + 0.1184

log γ± = = = −0.10304
0.1184 0.1184
γ± = −0.789
Experimental values of E at several specific values of m lead to specific values of

γ± as a function of m 1/2 in Fig. 14.3. The values are less than 1.0 as predicted by
the Debye–Hückel theory. The function is not linear except at low concentrations.
Experimental points deviate from the linear function.
Problem 14.1
Suppose we construct a Daniell cell (Section 14.1), conventional in every way except
that we replace the Zn2+ (aq) and Zn(rod) with Cd2+ (aq) and a Cd(s) rod.
(a) Which is the cathode and which is the anode, the copper cup or the Cd rod?
(b) What is the approximate voltage? (It is approximate because we don’t know
the exact concentrations of the Cu(aq) and Cd(aq) solutions.)
Problem 14.2
Many applications of electrochemistry have been in chemical analysis. Suppose
a Zn; Zn2+ concentration cell is set up in which the Zn2+ concentration in one cell
P1: OTA/XYZ P2: ABC
compartment is 0.0100 m but the concentration of the other Zn2+ half-cell is unknown.
The whole-cell potential was measured and found to be 32.4 mV.
(a) Without doing a numerical calculation, is the unknown more or less concen-
trated than the 0.0100 m solution?
(b) What is the unknown concentration?
(c) Does your answer to part b agree with your answer to part a?
Problem 14.3
The half-cell (reduction) potential of the Ag:AgBr electrode is 0.071 volts. Use this
information with information in Section 14.3 to determine the solubility product
constant K sp of AgBr in water.
Problem 14.4
A very important extension of electrochemical of solubility determination is gener-
alization of the method to determine other kinds of equilibrium constant and fur-
ther generalization to determination of the enthalpy, entropy, and Gibbs free energy
changes for electrochemical reactions. Write or derive the equations for doing this.
Problem 14.5
Is it possible to reduce all of the Fe+3 ion in an aqueous solution containing Fe2+ and
Fe3+ by filtering the solution through finely divided Zn(s)?
Problem 14.6
What is the standard Gibbs free energy change for the preceding reaction, and does
your answer agree with the preceding answer?
Problem 14.7
A hydrogen half-cell combined with a saturated calomel electrode has a measured
cell potential of 57.3 mV. What is the pH of the aqueous solution in the hydrogen
half-cell? If the hydrogen electrode is a standard hydrogen electrode (SHE), what is
the cell potential?
Problem 14.8
From the cell
Pt(s); H2 (g, 1 atm); HCl(aq); AgCl(s); Ag(s)

P1: OTA/XYZ P2: ABC
one can obtain the following voltages at very small HCl(aq) concentrations
mol dm−3 E cell (volts)
0.00321 0.5205
0.00562 0.4926
0.00914 0.4686
0.0134 0.4497
0.0256 0.4182
Find the standard cell potential of the Ag(s);AgCl(s) half-cell.
Problem 14.9
It is proposed to use cerium(III) as a reducing agent in the redox titration of iron(III)
according to the reaction
Ce3+ (aq) + Fe3+ (aq) → Ce4+ (aq) + Fe2+ (aq) 1 MH2 SO4
The half-cell potentials written as reductions are somewhat modified in the strong
acid solution:
Ce4+ (aq), Ce3+ (aq); Pt is − 1.44 volts 1 M H2 SO4
and
Fe3+ (aq), Fe2+ (aq); Pt is 0.68 volts 1 M H2 SO4
Is this a plausible titration method for further development? If so, why? If not, why
not?
P1: OTA/XYZ P2: ABC
15
EARLY QUANTUM THEORY:
A SUMMARY
The first quarter of the twentieth century was a perplexing mix of discovery and
wonderment as physicists progressed from the seemingly irrational quantization of
energy by Max Planck (1900) to the inevitability of Werner Heisenberg’s uncertainty
principle (1925). For brevity, we shall refer the reader to more detailed works for
the early history of quantum theory (especially Barrow, 1996; Laidler and Meiser,
1999; Atkins, 1998) and take as our starting point the observations by Louis de
Broglie that small particles—in particular, electrons—have a wave nature and by
Erwin Schrödinger that the spectrum of atomic hydrogen can be deduced by solving
the wave equation that now bears his name.
15.1 THE HYDROGEN SPECTRUM
Some substances give off colors in the visible region of the electromagnetic spectrum
when excited by an energy input. An example is sodium, which glows orange when a
solution of NaCl is sprayed into a flame. In general, a multiplicity of distinct energy
levels within an atom leads to a corresponding multiplicity of wavelengths charac-
terized by its unique electromagnetic spectrum. Distinct colors or, more specifically,
wavelengths, are given off or taken up by electrons as they emit or absorb energy of
frequency ν by changing levels. Emitted radiation of different wavelengths can be
separated and recorded in the form of a line spectrum. In most cases, the result of all
these energy exchanges is of bewildering complexity, but one spectrum—that of the

Copyright
235
P1: OTA/XYZ P2: ABC
236 EARLY QUANTUM THEORY: A SUMMARY
Lyman Balmer Paschen

Wavelength
FIGURE 15.1 The hydrogen atom emission spectrum (not to scale). The wavelengths λ of
the lines correspond to differences in energy between allowed levels in the hydrogen atom
according to the equation E = hν = hc/λ, where c is the speed of light, h is Planck’s constant,
and ν is the frequency (Planck, 1901).
hydrogen atom—shows regular groupings that invite closer attention and theoretical
explanation. Three of these groupings are shown in Fig. 15.1.
The experimental facts had been known for a half century before Schrödinger’s
time and had been partly explained in 1913 by Niels Bohr, who imposed a quantum
number on the angular momentum of the orbital electron in hydrogen to arrive at a set
of energy levels corresponding to stationary states of the atom. Transitions from one
energy level to another correspond quite precisely to lines in the hydrogen spectrum,
but it was not clear where the quantum number came from and it was not possible to
generalize the Bohr system to more complicated atoms and molecules (Fig. 15.2).
15.2 EARLY QUANTUM THEORY
Early quantum theory showed that not only the hydrogen spectrum, but also several
other important problems, can be solved by taking into account a frequency ν asso-
ciated with the energy of particles according to the Planck equation E = hν. But the
method, no matter how well it worked, still lacked a theoretical base and a logically
consistent means of application, modification, and improvement. De Broglie (1924,
1926) pointed out that energy E of a moving particle implies momentum p. Frequency
ν implies a wavelength. Therefore, if the Planck equation works (which it does),
-5
E x 10-19, J
-10
-15
-20
-25
FIGURE 15.2 The first six solutions of the H atom energy calculated by Bohr (1913). The
energies are negative because the electron is bound to the proton.
P1: OTA/XYZ P2: ABC
EARLY QUANTUM THEORY 237
there should be a connection between momentum and wavelength, which he wrote as

p = h/λ. At the atomic level, particles must have a wave nature. Conversely, waves
have a particle nature. This was promptly experimentally verified. This mathematical
equivalence between waves and particles is referred to as wave-particle duality.
Schrödinger (1925, 1926) reasoned that if electrons have a wavelength, they should
follow a wave equation. He wrote an equation for the electron in the hydrogen atom
employing a wave function , and he arrived at the electromagnetic spectrum of
hydrogen just as Bohr had done but without making the arbitrary assumptions inherent
in Bohr theory. While it should be noted that the Schrödinger equation is itself an
assumption and that quantum theory is based on a set of postulates, the implications
of Schrödinger’s theory are far wider and the results are far more general than earlier
theories. Quantum mechanics now pervades virtually all of physical science from
molecular biology to string theory.
The Schrödinger equation can be written in equivalent forms with a rather in-
timidating array of notations, but they are all manifestations of the same thing: the
postulate that a state vector | or state function contains all the information we
can ever have about a mechanical system such as an atom or molecule.
The Hamiltonian function was known from classical mechanics:
E = H =T +V
where T is the kinetic energy,

written in terms of the velocities v in Cartesian 3-space
T = 12 m v x2 + v 2y + v z2 , and V (x, y, z) is the potential energy. (Be careful not to
confuse velocity v with frequency ν.)
An operator is a mathematical symbol telling you to do something. For example,
V tells you to multiply by the potential energy (a scalar) and ∂ 2 /∂ x 2 tells you to
differentiate twice with respect to x while holding some other variable constant.
(It is a partial ∂ x 2 because y and z are also variables.) An operator must operate
on something, in the case we are interested in. If a Hamiltonian operator Ĥ for
the hydrogen atom system is written in terms of mathematical operators V and
Ĥ = T̂ − V̂ in a way that is analogous to the classical Hamiltonian for the total
energy of a conservative system, we get the operator equation
Ĥ = T̂ + V̂
The Schrödinger equation is a special case of the Hamiltonian form where T̂ is

the operator corresponding to the kinetic energy of the electron:

−2 ∂2 ∂2 ∂2
T̂ = + 2+ 2
2m e ∂x 2 ∂y ∂z
and V̂ (x, y, z) is the potential energy operator that describes the Coulombic attraction
between the nucleus and the electron. (V̂ (x, y, z) is usually written simply V (x, y, z),
where we take its operator nature as obvious.)
P1: OTA/XYZ P2: ABC
Atoms in their ground states are conservative systems. A conservative system is

one that is not running down. Our solar system is a conservative system but a clock
is not. This leads to the compact form
Ĥ = E
for the time-independent Schrödinger equation for a conservative system, where

= (x, y, z) is an amplitude function, and E is the total energy of the system.
We are interested in the time-independent form of the Schrödinger equation when
we study the ground state structure and energy of atoms and molecules, which do
not change over small time intervals. This equation was generalized to cover many
electron atoms and molecules. As we shall see later, spectroscopic transitions within
atoms and molecules involve the time-dependent form of the equation.
Independently, Heisenberg was developing a parallel line of quantum reasoning
that led to his famous uncertainty principle. The Heisenberg and Schrödinger equa-
tions can be shown to be mathematically equivalent. Both Heisenberg and Schrödinger
received the Nobel Prize.
Almost immediately after publication of Schrödinger’s and Heisenberg’s initial
papers on quantum theory, Born et al. (1926) proposed that the square of the wave
function |(r )|2 dτ (strictly, its vector inner product | ) describes undulations
in the probability of finding a particle in an infinitesimal volume of space dτ . For
most of our purposes, we shall regard the square of the wave function as governing
the probable location of an electron in a region of space containing many electrons.
15.3 MOLECULAR QUANTUM CHEMISTRY
As part of an emerging theory of molecular structure, the chemical bond was

associated with a region of high relative electron density between two nuclei. Heitler
and London (1927) took the lowest solution of the Schrödinger equation for one
of the atoms in H2 and combined it with an identical solution for the other atom in
such a way as to obtain an approximate wave function for the chemical bond of the
hydrogen molecule:
ψ = c1 1 ± c2 2
The new wave function is said to be a linear combination of the two atomic wave
functions 1 and 2 , centered at their respective nuclei. By selecting appropriate
values of c1 and c2 (which turn out to be equal), an energy minimum appears for the
combined wave function. In this notation, the atomic wave functions are exact but
the H2 moleular orbitals ψ are approximate.
We are accustomed, from classical mechanics and thermodynamics, to thinking of
an energy minimum as representing a stable state. And so it is with the Heitler–London
energy minimum for the linear combination of atomic orbitals in hydrogen. The
positive combination ψ = c1 1 + c2 2 leads to a stable state for two hydrogen
atoms at approximately the experimental bond distance (74 pm). This is the first
P1: OTA/XYZ P2: ABC
MOLECULAR QUANTUM CHEMISTRY 239
quantum mechanical affirmation of the two-electron bond postulated by G. N. Lewis

and is familiar from general chemistry. We have the first instance of a calculated
orbital overlap leading to stability for a molecular system. The term orbital has been
substituted for the classical mechanical term orbit in recognition of the substitution
of a probability function for a definite point location in space.
In the negative linear combination c1 ψ1 − c2 ψ2 , we have the first example of
an antibond. The probability function for the antibond shows that electrons are
repelled from the region between the nuclei, leading to the opposite of a Lewis bond.
Antibonding orbitals lie higher in energy than bonding orbitals. In some instances,
infusion of energy into a molecule by incident radiation can cause electrons to be
promoted from a bonding to an antibonding orbital. Only some wavelengths are
absorbed. If the absorbed radiation is in the visible range (sunlight), we see the
wavelengths left over after selective absorption as color.
Neither Schrödinger’s nor Heisenberg’s work was directed toward chemical bond-
ing, so one cannot make the argument, “Well, they knew the answer before they set
up the problem.” The H H bond in molecular hydrogen is a consequence of the
quantum nature of matter. This point of view was developed into the valence bond
theory by Linus Pauling (1935). Much of theoretical chemistry was to be dominated
by valence bond theory (for better or for worse) for the next four decades.
Hartree (1928) is usually credited with the critical suggestion that atomic problems
involving many electrons can be treated as a collection of simpler problems in which
a single electron moves in an average electrostatic field created by the nucleus and all
the other electrons.1 The electrostatic attraction between a nucleus and an electron is
far greater than the energy of a chemical bond. Therefore it is reasonable to suppose, as
Hartree did, that one-electron wave functions would resemble Schrödinger’s solutions
for the hydrogen atom, being identical in the angular part and differing only in
the radial probability distribution function—that is, the probable distance from the
nucleus. This is the central field approximation.
Fock (1930) and Slater (1930), utilizing the prior concept of electron spin (Uhlen-
beck and Goudsmit, 1925), recognized that the spin of an electron can be oriented
in two ways, with or against its orbital motion. Therefore, electron spin must have
a double-valued quantum number m s = ±1. A double-valued quantum number de-
mands that there be two wave functions per orbital, identical in all respects except
for their different spins. The two wave functions must be antisymmetric just as your
right and left hands are roughly identical in shape except that they are mirror images
of one another (antisymmetric). Two trial functions ϕ(1) and ϕ(2), having opposed
spins α, and β, can combine in four ways, only one of which is antisymmetric:
ψ = ϕα(1)ϕβ(1)
ψ = ϕα(2)ϕβ(2)
ψ = ϕα(1)ϕβ(2) + ϕα(2)ϕβ(1)
ψ = ϕα(1)ϕβ(2) − ϕα(2)ϕβ(1) ←
1 Hartree himself gave credit to Bohr.

P1: OTA/XYZ P2: ABC
The antisymmetric wave function is the same as an expanded 2 × 2 determinant:

ϕα(1) ϕβ(1)
= ϕα(1)ϕβ(2) − ϕα(2)ϕβ(1)
ϕα(2) ϕβ(2)
To impose the condition of antisymmetry on two otherwise identical orbitals, the

simple linear combinations of one-electron wave functions must be replaced by a
single-determinant wave function of the form for the two-electron case (e.g., the
helium atom):

ϕα(1) ϕβ(1)
ψ(1, 2) =
ϕα(2) ϕβ(2)
Generalizing to several electrons led to more complicated determinantal wave

functions. The use of antisymmetrized orbitals, as they were called, leads to the
Hartree–Fock computational method. This method was applied to atoms and was
tentatively applied to molecules larger than H2 .
15.4 THE HARTREE INDEPENDENT ELECTRON METHOD
One can treat the electrons of a many-electron atom as though they were entirely
independent of one another moving under the influence of the nuclear charge Z,
where Z is the atomic number, an integer greater than 1. This amounts to assuming
higher-level hydrogen-like orbitals with principal quantum numbers n = Z /2 because
each filled orbital contains 2 independent electrons. The resulting atomic energy will
be wrong because electrons are, in fact, not independent.
This crude model can be substantially improved by taking an effective nuclear
charge smaller than Z, motivated by our intuitive understanding that, no matter what
the overall electronic probability distribution may be, there is some probability that
some electrons will carry a negative charge between the nucleus and the other elec-
trons, thereby reducing the influence of the nuclear charge on them. Any designated
electron will be shielded from the nuclear charge by the other electrons. Simply
plugging in arbitrary numbers for the effective nuclear charge Z eff < Z gives an im-
provement in agreement of the calculated energy of the system with the experimental
value. It turns out that Z eff = 1.6 is a pretty good guess in the example of helium, a
two-electron problem with a nuclear charge of 2.0. Unfortunately, substitution of ar-
bitrary numbers is theoretically meaningless. One would prefer to be able to calculate
a radial electron probability distribution resulting in a theoretically derived energy
and Z eff for any single electron from a calculated energy distribution of all the others.
In treating helium, we shall begin with the reasonable first guess that each of its
two independent one-electron wave functions resembles that of the hydrogen atom:
∼
= ψ1 ψ2
P1: OTA/XYZ P2: ABC
THE HARTREE INDEPENDENT ELECTRON METHOD 241
and from the Schrödinger equation,
Ĥ = E
we expect to find
ĥψi = εi j ψ j
Mathematically, the operator ĥ is a 2 × 2 transformation matrix that transforms the

vector ψi into εi j ψ j . This Hamiltonian operator is written in lowercase to show that it
is approximate and not the exact operator Ĥ . In the case of helium with two electrons,
the state vector |ψ can be expressed as a two-element column vector

ψ1
|ψ =
ψ2
We can improve upon ψ and ε by successive approximations, using an iterative

procedure. We shall assume that we know where one electron is (we don’t) and
calculate its shielding effect on the other electron. We then calculate the probability
density function for the location of the second electron as influenced by our initial
guess as to its shielding from the first. With this information we can improve upon our
initial guess as to the location of the first electron because we now have a probable
location of the second electron. This iterative process can be continued until further
iteration results in no further lowering of the calculated energy, whereupon we have
the best estimate of the energy and the electron distribution we can get from the
procedure.
The Hamiltonian operator ĥ includes the kinetic energy operator of each electron i
and the attractive electrostatic potential energy exerted on the electron by the nucleus
e2
ĥ i = T̂i − (i = 1, 2)
ri2
The potential energy of one electron, call it electron 1, is influenced by the other,
call it electron 2, according to their charge e and the distance between them, r12 .
By the uncertainty principle we do not know where electron 2 is, but we do have a
probability density function |ψ(r2 )|2 governing its location in a volume element of
space dτ2 . Hence we have a most probable charge distribution function e |ψ2 (r2 )|2 .
The probable potential energy of electron 1 shielded from the nucleus by electron 2
is, in suitable units, the product of the charges e divided by the distance between them

e |ψ2 (r2 )|2 |ψ2 (r2 )|2
V1 = e dτ2 = e2 dτ2
r12 r12
The integration is over the space occupied by electron 2 because we cannot locate it
as a point charge.
P1: OTA/XYZ P2: ABC
What can be said of the interaction between electron 1 with electron 2 can be said
of electron 2 interacting with electron 1:

|ψ1 (r1 )|2
V2 = e 2
dτ1
r21
The operator ĥ i (i = 1, 2) is the kinetic energy operator for electron 1 or 2:

−2 2 −2 ∂2 ∂2 ∂2
∇i = + 2+ 2
2m e 2m e ∂x 2 ∂y ∂z
plus the (negative) potential energy of attraction between the nucleus and the electron,
−Z e2 /ri2 :
−2 2 Z e2
ĥ i = ∇ − 2
2m e 2 ri
With these potential and kinetic energies, we can write Schrödinger equations for
electrons 1 and 2:

T̂1 + V̂1 (r2 ) ψ1 (1) = ĥ 1 ψ1 (1) = ε1 ψ1 (1)
and

T̂2 + V̂2 (r1 ) ψ2 (2) = ĥ 2 ψ2 (2) = ε2 ψ2 (2)
With these approximate Hamiltonians, we have two Schrödinger equations that con-
tain the energy of the two electrons in the helium atom εi (i = 1, 2):
ĥψi = εi j ψ j (i = 1, 2)
The two equations are integrodifferential equations because they have an integral
as part of V (r ) and second differentials in ∇ 2 . As written, they are coupled because
one equation depends on the solution of the other through the potential energies. In
the first equation, V1 (r2 ) contains the electron density of electron 2; in the second
equation, V2 (r1 ) contains the electron density of electron 1. When we assume a wave
function (any wave function) we uncouple these two equations. Though uncoupled
and therefore solvable, the equations will no doubt lead to the wrong answer for
the energies εi because the assumed ψ(r ) is only a guess. Hartree’s guess was
the most reasonable one he could make; he took ψ1 and ψ2 to be hydrogen-like
orbitals.
Allowing Z to approach an effective nuclear charge Z eff , we have the Hartree
equations. Uncoupling these equations by assuming functions ψ1 and ψ2 was a
P1: OTA/XYZ P2: ABC
brilliant step in quantum theory. It permits one to calculate a better approximation to

V1 and V2 , which leads to improved values of Z eff , which leads to improved values
of ψ1 and ψ2 , which leads to better V1 and V2 , and so on. This iterative process is
continued until the energies do not change from one iteration to the next. When that
condition has been met, the energies calculated from V1 and V2 are self-consistent.
The Hartree equations are eigenvalue equations, hence they lead to discrete (scalar)
values ε1 and ε2 for the two one-electron Schrödinger equations. The energies are
not exact, but they are the best we can get from assumed hydrogen-like orbitals ψ1
and ψ2 by the Hartree procedure.
Conveniently, the applied mathematics of self-consistent field (SCF) iterations had
been worked out by astronomers calculating the orbits of planets, prior to the advent
of quantum mechanics. They called it the variational method.
15.5 A DIGRESSION ON ATOMIC UNITS
We can measure length in any unit we want: m, mm, furlongs, and so on. It is perfectly
legitimate to use the radius of the first Bohr orbit as our unit of distance, replacing
the meter but still related to it:
1 a0 ≡ 5.292 × 10−11 m
Likewise, we can define units of mass, charge, and angular momentum:
1 m e ≡ 9.109 × 10−31 kg
1 e ≡ 1.602 × 10−19 C
1 ≡ 1.055 × 10−34 J s
These definitions lead to considerable simplification in the equations of quantum

chemistry (see Problems and Examples).
Example 15.1 A Hartree Fock Solution for He

The limiting solution for a Hartree–Fock procedure to the problem of the helium
atom is −2.862 hartrees, where the unit hartree (Eh ) is defined as exactly half the
energy of the electron in the ground state of hydrogen, E h ≡ 627.51 kcal mol−1 =
2625 kJ mol−1 . The crude unshielded solution for the helium ionization energy is
Z 2 = 22 = 4 hartrees. By our trial-and-error method, we got Z eff = 1.6 which leads
to about (1.6)2 = 2.56 E h . The experimental value is 2.903 E h .
P1: OTA/XYZ P2: ABC
The GAUSSIAN
C program, using a restricted Hartree–Fock procedure, RHF
yields
SCF Done: E(RHF) = -2.80778395662 A.U. after 1 cycles
The value after four cycles by a program named G3 (to be discussed in more detail
later) gives
A Hartree–Fock triple zeta calculation (a linear combination involving three ad-

justable Z parameters) gives
The energy – 2.860 E h agrees with the experimental value to within 1.5%.
Example 15.2
The Hamiltonian function for the ground state hydrogen atom (with angular momen-
tum equal to zero) is
2 2 e 2 2 2 ∂ 2 e2
Ĥ = − ∇ + =− r −
2m e r 2m e ∂r 2 4π 2 ε0r

2 d ∂2 e2
=− r2 2 −
2m e r dr ∂r 4π 2 ε0r
where there is no θ, φ term because there is no angular momentum in the ground

state and ∇ 2 = r 2 (∂ 2 /∂r 2 ) = r 2 (d 2 /dr 2 ) for this one-dimensional operator in which
r 2 appears because the radius vector may be oriented in any direction toward a
surface element of a sphere. The surface area of a sphere goes up as the square of the
radius A = 4πr 2 . The 4π 2 ε0 in the denominator of the potential energy arises in the
same way, where 4π 2 ε0 is the permittivity of free space (essentially a proportionality
constant between coulombs and joules).
Allowing this operator to operate on the trial function φ = e−αr gives, from the
Schrödinger equation ( Ĥ φ = Eφ),

2 d d2 e2
− r2 2 − φ(r ) = Eφ(r )
2m e r dr dr 4π 2 ε0r
P1: OTA/XYZ P2: ABC
or

d 2 e−αr 2 de−αr 2m e e2
+ + 2 E+ e−αr = 0
dr 2 r dr 4π 2 ε0r
We can carry out the differentiations, divide through by e−αr , and segregate the
terms into those that depend on r (terms 2 and 4) and those that do not (terms
1 and 3):

2 2m e e2
α − α+ 2
2
E+ =0
r 4π 2 ε0r
2 2m e E 2m e e2
α2 − α + + =0
r 2 2 4π 2 ε0r
so
2m e E 2 2m e e2
α2 + = α −
2 r 2 4π 2 ε0r
The basis! function φ = e−αr is a negative exponential, so there must be some value
of r for which the right-hand side of the equation becomes zero. But the left-hand
side of the equation is a group of constants, which add up to a single constant. If this
constant is zero for one value of r, it must be zero for all. Thus,
2m e E
α2 + =0
2
where = h
2π
The consequences of this simple equation are very important:
2m e E
α2 = −
2
2 2
E(r ) = − α
2m e
and
2 2m e e2
α− 2 =0
r 4π 2 ε0r
2 2
2 2m e e 1
α= 2
r 4π ε0
2 r
P1: OTA/XYZ P2: ABC
But α = m e e2 /2 , so
2
2 mee 2 mee 4 1 mee 4 1
E(r ) = − =− − =−
2m e 2 22 2 2 2
4
that is, E(r ) = 12 hartree in atomic units because the constants m e e/2 are all defined
as 1 in the atomic system.
Problem 15.1
The dominant line in the emission spectrum of atomic hydrogen is a beautiful crim-
son line at a wavelength of 656.1 nm. The hertz (Hz) is the unit of frequency that
corresponds to one cycle of the wave motion per second. The speed of electromag-
netic radiation is c = 2.998 × 108 m s−1 . Each unit of frequency represents one step
of wavelength λ, so the speed of wave propagation is the number of steps per second
(frequency) times the length of each step, or c = ν λ. Express the crimson hydrogen
line in terms of frequency in hertz and find its energy from the Planck equation. It
is sometimes convenient to express frequency in wave numbers, ν̃ = ν/c. Find ν̃ for
the crimson line in the hydrogen spectrum and use it to find λ. (This is a crosscheck
because you already know λ).
Problem 15.2
Einstein’s most famous equation is E = mc2 and the Planck equation is E = hν.
Since they both express the energy, combine them to obtain the de Broglie equation,
which expresses wave particle duality.
Problem 15.3 (Classical Mechanics)

A projectile of mass m was fired upward so that it traveled along a parabolic arc
with velocity v. Its initial velocity was v 0 . How high will it go? Neglect wind and air
resistance.
Problem 15.4
Expand determinant a so as to obtain a single number. Expand determinant b so as
to obtain a single number. The job is already becoming tedious, so expand b using
machine software—for example, Mathcad C . Expand c by machine. Could you have
guessed the answer simply by knowing b and looking at c?

1 2 3 1 2 3
2 3
(a) (b) 4 5 6 (c) 4 5 6
5 6
7 8 9 7 8 9.1
P1: OTA/XYZ P2: ABC
Problem 15.5
Good mathematical software contains a number of operations on matrices that will
be useful in coming chapters. Use Mathcad C or similar software to calculate the
sum, product, square, inversion, and product with inverse (under certain conditions
equivalent to division). Choosing matrix c from between the vertical lines in the
preceding problem enables one to solve most of these problems simply by inspection
and then allows one to verify (test run) using the software.
Problem 15.6
According to an Einstein theory of electromagnetic radiation, the 656.1-nm radiation
from atomic hydrogen in the first excited state is a particle (later named a photon).
What is the energy, frequency, wave number, and momentum of this putative particle?
Express and explain the units of the final result for each of the four steps.
Problem 15.7
Show that the operator ∇ 2 = (1/r 2 )(d/dr )r 2 (d/dr ) in spherical coordinates can
be written ∇ 2 = (d 2 /dr 2 ) + (2/r )(d/dr ). (This is part of the solution of the radial
Schrödinger equation for the ground state of the hydrogen atom.)
Problem 15.8
Using the operator ∇ 2 , write the radial form of the Schrödinger equation.
Problem 15.9
Allow the operator ∇ 2 = (d 2 /dr 2 ) + (2/r )(d/dr ) to operate on the function
R(r ) = e−r in spherical polar coordinates (r, θ, and φ). What is ∇ 2 R(r )? What are
(d R(r )/dθ ) and (d R(r )/dφ)?
P1: OTA/XYZ P2: ABC
16
WAVE MECHANICS OF
SIMPLE SYSTEMS
Because of the wave nature of the electron, solution of problems of atomic and
molecular structure requires solving wave equations. Problems in atomic and molec-
ular structure involve electrons that are tied to a positive nucleus or group of nuclei
by electrostatic attraction. They are bound. Therefore we shall be concerned with
mathematical boundary conditions imposed on the solutions of wave equations.
16.1 WAVE MOTION
Any mathematical description of a vibrating guitar string has to take into account
the fact that it is tied down at both ends. This puts severe restrictions on the wave
forms the string can take which are not characteristic of a free wave. A free wave
can assume any wavelength, but a bound wave is restricted to waves that come to
zero at either end of the string. The three bound wavelengths shown in Fig. 16.1 are
allowed by these two boundary conditions, but an infinite number of intermediate
wavelengths are not allowed. The wave of longest wavelength is the fundamental
(one observes half a sine wave), and the others are called overtones. A fundamental
and its overtones are well described by the sine function.
The wavelength λ of the first overtone is one-half the wavelength of the funda-
mental λ = 12 λfundamental , the next overtone has λ = 13 λfundamental , and we can imagine
many overtones at λ = n1 λfundamental . This is the source of integers n that appear for
bound waves. De Broglie’s observation that particles (specifically electrons) have a

Copyright
248
P1: OTA/XYZ P2: ABC
WAVE EQUATIONS 249
1
1
sin(x)
sin(2x)
sin(3x)
1 1
0 x
FIGURE 16.1 Graph of sin(x), sin(2x), and sin(3x) shown over the interval [0, π ]. The
fundamental mode of oscillation of a vibrating string is 12 of a sine wave, the first overtone is
a full sine wave from 0 to 2π , and the second overtone is 32 of a sine wave.
wave nature leads to quantum numbers n in atomic spectra (Bohr) and to the con-
nection between wave equations and atomic structure (Schrödinger, Hartree). Born’s
observation that the wave equation governs the probability of finding an electron,
which may contribute to or oppose formation of a chemical bond, leads to the con-
nection between wave equations and molecular structure, energy, and reactivity that
is the basis of modern quantum chemistry.
16.2 WAVE EQUATIONS
The function sin φ(x) describes the excursion away from φ (x) = 0 of an infinite
number of points along the variable x = 0 to l (Fig. 16.1). For any selected harmonic,
the fraction x/λ tells us “where we are” on the sine wave. If x = λ, we are at the very
end of the sine function. If x = λ/2, we are half way. In order to fully describe a wave,
we need one more piece of information beyond the sine function and wavelength λ.
The height of the wave is essential; a big sine wave has a large amplitude of oscillation
A and a small sine wave has a small amplitude. The complete description of the wave
2π x
is φ(x) = A sin .
λ
2π x
We can obtain the second derivative of φ(x) = A sin :
λ
d 2 φ(x) 4π 2 2π x 4π 2 4π 2
= −A sin = − A sin φ(x) = − φ(x)
dx2 λ2 λ λ2 λ2
This is typical of the wave equations that Schrödinger used:
d 2 φ(x)
= kφ(x)
dx2
P1: OTA/XYZ P2: ABC
250 WAVE MECHANICS OF SIMPLE SYSTEMS
It is also a member of the class of eigenvalue equations, in which an operator Ô

operates on the eigenfunction φ(x) to give an eigenvalue k times the same function:
Ôφ(x) = kφ(x)
In this case, the operator is d 2 /d x 2 , and the eigenvalue is k = −4π 2 /λ2 .

The eigenfunction may be represented as a vector:

ξ1 (x)
φ(x) =
ξ2 (x)
The effect of the operator is to stretch or contract the eigenvector by an amount equal
to the eigenvalue, or to change its direction, or both.
Waves in two dimensions (x, y) such as those of a vibrating plate or membrane are
described by
∂ 2 φ(x, y) ∂ 2 φ(x, y) 4π 2
+ = − φ(x, y)
∂x2 ∂ y2 λ2
and the wave equation in 3-space is
∂ 2 φ(x, y, z) ∂ 2 φ(x, y, z) ∂ 2 φ(x, y, z) 4π 2

+ + = − φ(x, y, z)
∂x2 ∂ y2 ∂z 2 λ2
16.3 THE SCHRÖDINGER EQUATION
By the de Broglie equation p = h/λ, it follows that λ2 = h 2 / p 2 . Substituting for λ2

in a three-dimensional wave equation, we obtain
∂ 2 φ(x, y, z) ∂ 2 φ(x, y, z) ∂ 2 φ(x, y, z) 4π 2 4π 2 p2

+ + = − φ(x, y, z) = − φ(x, y, z)
∂x2 ∂ y2 ∂z 2 λ2 h2
The momentum of a moving particle is its mass times its velocity, p = mv. This leads
to p 2 = m 2 v 2 = 2m( 12 mv 2 ) = 2mT , where T is the classical kinetic energy 12 mv 2 .
Now, with a slight change in symbol for notational simplicity, we obtain

∂ 2 ∂ 2 ∂ 2 ∂2 ∂2 ∂2 8π 2 mT
+ + = + 2+ 2 =−
∂x 2 ∂y 2 ∂z 2 ∂x 2 ∂y ∂z h2

The bracketed operator ∂ 2 /∂ x 2 + ∂ 2 /∂ y 2 + ∂ 2 /∂z 2 operates on the wave function
(x, y, z). It is given the shorthand notation ∇ 2 . Because its eigenvalue contains a
P1: OTA/XYZ P2: ABC
QUANTUM MECHANICAL SYSTEMS 251
group of constants times the kinetic energy T, ∇ 2 is called a kinetic energy operator.
The total energy of a classical system is the sum of the kinetic energy and the potential
energy, E = T + V , so the kinetic energy is the difference T = E − V and
8π 2 m
∇ 2 = − (E − V )
h2
which is one form of the famous Schrödinger equation.

Other notational conveniences include the definition of “h bar,” h̄ ≡ h/2π , which
gives
2m
∇ 2 = − (E − V )
h̄ 2
and
h̄ 2 2
− ∇ + V = E
2m
2
The kinetic energy operator plus the potential energy operator (− 2m h̄
∇ 2 + V )
is defined as the Hamiltonian operator Ĥ by analogy to the classical Hamiltonian
function H = T + V (Chapter 15). The kinetic and potential energy operators are T̂
and V̂ , so Ĥ = T̂ + V̂ . (Even though V̂ is an operator, it is often written V.) These
notational changes give the concise form of the Schrödinger equation:
Ĥ = E
where the total energy E is a scalar eigenvalue. It is distinguished from the opera-
tors by not having a circumflex notation. Clearly energy is a scalar because it has
magnitude but not direction.
16.4 QUANTUM MECHANICAL SYSTEMS
A system is a collection of mechanical entities governed by physical laws. If we know

the state of a system, we know every physical property it can have. It is astonishing
but true that all this information can be known by specifying a very few fundamental
variables (degrees of freedom) and a small number of postulates.
The wave function satisfies all of the properties of a vector; therefore it is written
as a vector | or |. We shall use either the vector form | or the functional form
, according to which is more convenient. When it is useful to specify the variables
as degrees of freedom, we shall do that: |(x1 , x2 , . . .) or (x1 , x2 , . . .), but usually
the simpler notation | or is preferred.
P1: OTA/XYZ P2: ABC
One of the postulates of quantum mechanics is as follows:
If a system is in a state s described by | an eigenvector of the operator Â (or,

equivalently if is an eigenfunction of Â), the corresponding observable a is an
eigenvalue:
Â | = a |
or, equivalently,
Â = a
and experiments on s in state | will always yield the observable a.
In the case of the atomic and molecular systems that concern us here, the general
operator Â is the Hamiltonian operator, Ĥ , and the observable a is the energy level
E i of the state |i . Usually there are many states leading to many energies E i .
The set{E i } represents a spectrum or multiplicity of energy levels, one for each
eigenvector:
Ĥ |i = E i |i
An operator operating on a vector produces

another vector. The operator Ĥ op-
erating on | produces the vector Ĥ | . The inner product of | premultiplied

into Ĥ | is
| Ĥ | = | E | = E |
The inner product of two vectors | is a scalar, which is why the eigenvalue E
can be moved out of the brackets on the right. The operator Ĥ cannot be factored
out. This leads to an expression for the energy
| Ĥ |
E=
|
The Born probability postulate leads to a further simplification. The inner product
| amounts to the integral over all space of the wave function squared 2 or
the product ∗ if the wave function is complex. The sum of the probabilities of
finding an electron over all space must be 1.0 (certainty) because the electron has
to be somewhere. With the inner product | = 1.0 we have the energy of an
eigenstate as
E i = i | Ĥ |i
P1: OTA/XYZ P2: ABC
THE PARTICLE IN A ONE-DIMENSIONAL BOX 253
16.5 THE PARTICLE IN A ONE-DIMENSIONAL BOX
Facility in working quantum mechanical energy problems can be gained by going

from simple problems to more complicated ones. The problem usually chosen as the
starting point is the particle in a one-dimensional box of length l:
x =0| • |x =1
Since there is only one dimension, the wave function is (x). The Schrödinger
equation in this case is
h̄ 2 2
− ∇ (x) + V (x) = E(x)
2m
We stipulate that the potential energy is zero inside the box V = 0 and that it is
infinite outside the box. These conditions mean that the particle cannot escape from
d 2 (x)
the box. We note that the one-dimensional operator is ∇ 2 = , so
dx2
h̄ 2 d 2 (x)
− = E(x)
2m d x 2
inside the box, or
d 2 (x) 2m E
= − 2 (x)
dx2 h̄
which is a typical wave equation and is an eigenvalue problem.

From our knowledge of wave equations (Section 16.2), we know that a function
2π x
like (x) = A sin will satisfy this equation. The second derivative of (x) is
λ
d 2 (x) 4π 2 2π x 4π 2
= −A sin = − (x)
dx2 λ2 λ λ2
d 2 (x) 2m E
but it is also true that 2
= − 2 (x), so
dx h̄
4π 2 2m E
= 2
λ2 h̄
We also know that bound waves must have wavelengths that go up as half-integers
of the limits placed on the oscillatory excursion—that is, the length of a vibrating
guitar string (Fig. 16.1) or the length l of a box. Wavelengths allowed by the boundary
conditions are λ = n1 λfundamental (Section 16.1), where λfundamental = 2l. It follows that
P1: OTA/XYZ P2: ABC
10
sin(x) 1 6
sin(2x) 4
sin(3x) 9
4
x
FIGURE 16.2 Wave forms for the first three wave functions of the particle in a box. The
waves are drawn at the lowest three energy levels: n = 1, 2, 3; (n 2 = 1, 4, 9).
λ = n2 l. A little algebra goes from
4π 2 2m E
− 2 = − 2
2
l h̄
n
to the energy spectrum (Fig. 16.2):
n 2h̄ 2 π 2 n2h2
E= =
2ml 2 8ml 2
Except for the lowest one, each wave function has internal values of x or nodes
at which (x) = 0. The Born probability density |(x)|2 of finding the particle
precisely at one of these nodes is zero. Not counting the nodes at the extremities of
the wave function, the number of internal nodes goes up as 0, 1, 2, . . . . Each internal
node in the wave function (x) yields an internal zero point shown as a minimum in
the probability function in Fig. 16.3. There are n − 1 internal nodes.
By elaboration of the methods used for the particle in a one-dimensional box,
the problem can be solved in two dimensions to produce solutions for the vibratory
P1: OTA/XYZ P2: ABC
THE PARTICLE IN A CUBIC BOX 255
n 1 2 3
P( n x) ( sin ( n x) ) 2 n2
10
P(1 x) 6
P(2 x)
P(3 x)
4
x
FIGURE 16.3 A Mathcad C sketch of the born probability densities at the first three levels
of the particle in a box. The lowest wave function has no internal nodes, the second wave
function has one, and the third has two.
nodes of a square plate or membrane. It can be solved in 3-space to produce solutions

to the vibratory motions in a solid cube. Other geometries are possible such as a
circle, rectangle, parallelepiped, cylinder, and so on. Each problem lends insight and
is recommended to the interested reader.
16.6 THE PARTICLE IN A CUBIC BOX
In the case of a particle confined to the interior of a cube, we have a three-dimensional

wave equation of the form
h̄ 2 2
− ∇ (x, y, z) + V (x, y, z) = E(x, y, z)
2m
P1: OTA/XYZ P2: ABC
We simplify this equation immediately by setting V (x, y, z) = 0, so the middle

term drops out. The kinetic energy operator looks complicated:
h̄ 2 d 2 (x, y, z) h̄ 2 d 2 (x, y, z) h̄ 2 d 2 (x, y, z)

− − − = E(x, y, z)
2m dx2 2m dy 2 2m dz 2
until we make the entirely reasonable observation that, in a cube, there is no reason
to prefer particle motion in any one direction over motion in any other. The energy
E x supplied by particle motion x is the same as E y supplied by y and E z supplied
by z . This being the case, we can split one complicated equation into three simple
ones:
h̄ 2 d 2 (x)
− = E x (x)
2m d x 2
h̄ 2 d 2 (y)
− = E y (y)
2m dy 2
h̄ 2 d 2 (z)
− = E z (z)
2m dz 2
Each of these equations has already been solved with the results
2π x
(x) = A sin
λ
2π y
(y) = A sin
λ
2π z
(z) = A sin
λ
and
n 2x h 2
Ex =
8ml 2
n 2y h 2
Ey =
8ml 2
n 2z h 2
Ez =
8ml 2
In the energy equations, the length of one edge of the cube l is in the denominator. If the
geometry had been other than a cube, there would be different dimensions—perhaps
a, b, and c for a parallelepiped. This would have made the solutions slightly more
complicated, but it would not have made any significant change in the result.
P1: OTA/XYZ P2: ABC
THE PARTICLE IN A CUBIC BOX 257
x
y
FIGURE 16.4 The ground state orbital of a particle confined to a cubic box.
16.6.1 Orbitals
The ground state probability density has a maximum in the center when viewed from
each of the three dimensions (Fig. 16.4). Thus there is a high probability density at the
center of the cube diminishing symmetrically in all directions (spherical symmetry).
This is the geometry of an s atomic orbital.
16.6.2 Degeneracy
If the particle is excited to the n = 2 state in the x direction while its motion in both
the y and z directions remain at the lowest energy, the resulting orbital has an internal
node in the x direction but no nodes in the y and z directions (Fig. 16.5). This gives
the geometry of the first p orbital which we denote px . What was said for the x
direction can be said for the y and z directions, so we have three orbitals with energies
that are identical. Different orbitals with the same energy are said to be degenerate
(Fig. 16.6). Perhaps we remember from elementary chemistry that the p orbitals of
hydrogen are three-fold degenerate.
16.6.3 Normalization
2π x
One can also find the constant A in (x) = A sin by normalizing the wave
λ
function. Normalization requires setting the integral of over all space equal to 1.0
2
x
y
FIGURE 16.5 The first excited state of a particle confined to a cubic box.
P1: OTA/XYZ P2: ABC
FIGURE 16.6 Degenerate energy levels in a one-dimensional box. Degeneracy increases

with n in the sequence 1, 4, 9, . . .

2
by Born’s principle. Normalization yields A = (see Problems) for the particle
l
in a one-dimensional box. Normalization constants, sometimes rather messy ones,
appear as premultipliers of wave functions.
Further elaboration of the method [see many textbooks, including Levine (2000)
and Barrow (1996)] yields solutions for the harmonic oscillator (one dimensional
vibrator) and the rigid rotor. These systems have the same kind of energy level
spectrum (Section 16.5) as the particle in a box except that the spacings are different.
The spacing is in equal steps for vibration of the harmonic oscillator but not for
rotation.
16.7 THE HYDROGEN ATOM
The kinetic energy operator of an electron moving in the vicinity of a proton (H+
nucleus) is the same as a particle in a box:
h̄
T̂ = ∇2
2m e
Many texts replace the mass of the electron m e by µ designating the reduced mass
of the electron and proton rotating about their center of gravity. The effect of this
correction is very small for the proton–electron pair, and we shall ignore it until we
reach comparable problems of molecular rotation where more nearly equal masses
are involved.
The potential energy is not zero in this problem; rather, it is
e2
V =
r
according to the electrostatic attraction between the nucleus and the electron. This
leads to
2m e
∇ 2 (r, θ, φ) = (E (r, θ, φ) − V (r, θ, φ)) (r, θ, φ) = 0
h̄ 2
where we have expressed (r, θ, φ) in spherical polar coordinates r, θ, and φ. The

problem is simpler this way because of its spherical symmetry. Routine algebraic
methods exist for conversion of problems in Cartesian coordinates to spherical polar
coordinates (Barrante 1998).
P1: OTA/XYZ P2: ABC
BREAKING DEGENERACY 259
Upon writing this equation out in full, it becomes rather intimidating, as the
reader will see by scanning several textbooks. Fortunately, expressed in r, θ, and φ,
it breaks up into three separate equations just as the particle in a cubic box did. The
equations are

∂ 2 ∂ R(r ) 2m e r 2 e2
r + +E R(r ) = R(r )β
∂r ∂r h̄ 2 r
1 ∂ ∂ (θ ) m 2e
sin θ − = −β (θ )
sin θ ∂θ ∂θ sin2 θ
and
1

(φ) = − √ eim e φ
2π
Each of these equations involves only one variable R(r ), (θ ), and

(φ). The first
R(r ) is called the radial equation and the second two are often lumped together
as the spherical harmonics Y (θ, φ) = (θ )
(φ). The name is apt; they describe
oscillations taking place on the surface of a sphere with the boundary constraints that
no wave can have a discontinuity or seam anywhere on the sphere.
The wave function or orbital for the electron in the ground state of the hydrogen
atom is (r ) = e−αr , where r is the radial distance between the proton and the
electron, and α contains several constants. Solution for α by any one of several
methods gives

1 me4 me4
E =− =−
2 (4π ε0 )2 h̄ 2 32π 2 ε02h̄ 2
This purely quantum mechanical result is in precise quantitative agreement with

Bohr’s semiclassical result for the ground state of hydrogen. When higher energy
states are considered, quantum numbers appear just as they do in Bohr’s result.
In addition to the principal quantum number n, a more general solution for the
spherical harmonics Y (θ, φ) = (θ )
(φ) leads to a quantum number l for the (θ )
equation and m for the
(φ) equation. There are also restrictions on the quantum
numbers in the complete solution. For example, n ≤ l + 1, so that l can take on a
zero value even though n cannot.
16.8 BREAKING DEGENERACY
For the simple one-electron system of hydrogen, the 2s and 2p orbitals are degenerate,
as are the more complicated 3s, 3p, and 3d orbitals. When we start putting electrons
into these orbitals to build up the atomic table by the aufbau principle, however,
some of these degeneracies are lost. For example, the probability density function
of the 2p orbital is small near the nucleus, where, in contrast, the s orbital has a
P1: OTA/XYZ P2: ABC
FIGURE 16.7 Reduced degeneracy in the energy levels for hydrogen-like atoms. Some of
the degeneracy of Fig. 16.6 has been lost.
maximum probability. Consequently, the energy of an electron drifting about in a

distant p orbital is less negative than that of a tightly held s electron. This causes the
sp degeneracy of the hydrogen atom to be broken up in beryllium, boron, carbon, and
the heavier elements. The 2s level is lower than the 2p levels in beryllium because of
its greater average radial distance from the nucleus and because of s shielding.
We see this when we traverse the series H → He → Li → Be → B in the atomic
table. The 1s level readily accommodates H and He, but the electron in Li is excluded
from the 1s level which is “filled” by two electrons.1 It could go into either the 2s or
one of 2p levels, but in fact it goes into the 2s level which has a lower energy. This
preference is even more remarkable in beryllium. The valence electron in Be has good
reason to go into the 2p orbital to escape charge and spin correlation with the electron
already in the 2s orbital, but it refuses this safe haven and goes into the lower energy
2s alternative. We know from ionization potentials that the electron in Be ionizes
from the 2s orbital, not the 2p, and that Be is a typical metal, not as metallic as Li
perhaps, but not a metalloid like boron. Atoms with orbital configurations similar to
hydrogen but with reduced degeneracy are called hydrogen-like atoms (Fig. 16.7).
16.8.1 Higher Exact Solutions for the Hydrogen Atom

The first six eigenfunctions of the hydrogen atom are shown in Table 16.1. The
geometric parameter a = 52.9 pm = 5.29 × 10−11 m is 1/α, the first orbital radius
in Bohr theory. The integer Z is the atomic number. It is 1 for hydrogen but 2 or 3
for the higher hydrogen-like atoms shown here. The orbitals fall into two different
groups, s and p. More complete listings include d, f, g, . . . orbitals of increasing
complexity. Note that the terms function and orbital are used as synonyms. The s
eigenfunctions are 1s, 2s, and 3s. They do not contain sin or cos parts because they
do not depend upon angular location. They are spherically symmetrical. Table 16.2
shows the orbitals reduced to their simple functional form. Let us concentrate on the
simple polynomial form of the s orbitals by ignoring the premultiplying constants.
The first orbital is simply a negative exponential. From the monotonic decrease of
−r
zero roots for the 1s function, but from the polynomials (1 − r ) and
e , we expect
1 − r + r 2 we expect two and three roots, respectively. A Mathcad C graph of the
3s radial wave function in Fig. 16.8 shows that the polynomial part produces two
nodes and one asymptotic approach when the function is plotted on the vertical axis
against r as the independent variable.
1 The universal restriction of two electrons with opposed spins per orbital is called the Pauli exclusion
principle.
P1: OTA/XYZ P2: ABC
BREAKING DEGENERACY 261
TABLE 16.1 The First Six Wave Functions for Hydrogen.

3/2
1 1
ψ1s = 1/2 e−Zr /a
π a

1 Zr 3/2 −Zr /2a
ψ2s = 2 − e
4 · (2π)1/2 a
5/2
1 Z
ψ2 pz = re−Zr /2a cos θ
4 · (2π)1/2 a
5/2
1 Z
ψ2 px = re−Zr /2a sin θ cos φ
4 · (2π)1/2 a
5/2
1 Z
ψ2 p y = re−Zr /2a sin θ sin φ
4 · (2π)1/2 a
3/2 3/2
1 Z Zr Z 2r 2
ψ3s = 27 − 18 + 2 e−Zr /3a
81(3π )1/2 a a a2
TABLE 16.2 The First Three s Wave Functions for

Hydrogen (Simplified Form).
1s e−r
2s (2 − r ) e−r = 2e−r − r e−r

3s 27 − 18r + 2r 2 e−r = 27e−r − 18r e−r + 2r 2 e−r
2 .63r
R(r) 27 18 r 2r e
l(r) 0
R(r)
l(r)
1
0 5 10 15
r
FIGURE 16.8 Roots of the radial 3s wave function of atomic hydrogen as a function of
distance r. Three nodes correspond to zeros of the function at about r = 2, 7, and ∼ 15. The
function l(r) simply draws a horizontal line.
P1: OTA/XYZ P2: ABC
a .529
r
2
1 r r 3a
S(r) 27 18 2 e
1.5 a a
81 3 a
2 2
R(r) r S(r)
R(r) 0.01
0
0 5 10
r
FIGURE 16.9 The radial probability density for an electron in the 3s orbital of hydrogen.
To find the probability density function for the lowest three orbitals, we square
the wave function. At each node the probability density of finding the electron at the
radial distance r is zero. Between the nodes are maxima called antinodes. For the 3s
orbital, this leads to Fig. 16.9 in which there are the antinodes at about r = 0.5, 2,
and 7. The radii of maximum probability densities correspond to the “shells” in early
atomic theory.
16.9 ORTHOGONALITY AND OVERLAP
It is not difficult to show that the probability density of the product of the pz orbital
times the px orbital is zero. Their overlap is zero. Unlike s orbitals which are positive
(+) everywhere, the 2p orbitals have one angular nodal plane. Therefore they pass
from − to + and back again as the radius vector passes through either 0 or π . The
p orbital has a plane of inversion as one of its symmetry elements. The p orbitals
have a positive lobe and a negative lobe (Fig. 16.10). If a p orbital and an s orbital
FIGURE 16.10 The radial node of the 2p atomic orbital.

P1: OTA/XYZ P2: ABC
MANY-ELECTRON ATOMIC SYSTEMS 263
pz
px + +
+
- + s
s -
(a) (b)
FIGURE 16.11 Favorable spx and unfavorable spz overlap of orbitals depending upon orbital
symmetry. Diagram a depicts bond formation but orbital overlap exactly cancels in diagram b.
are combined as basis functions (vectors) to describe a chemical bond, the result
may be favorable, as in Fig. 16.11a, resulting in enhancement of the probability
density between bonded atoms. The overlap of an s positive orbital and a px positive
orbital is positive. Or it may be unfavorable, depending on the orbital symmetry, as
in Fig. 16.11b where any + + overlap between one lobe of the 2 pz orbital with the 1s
orbital is canceled by the negative + − overlap of the other. In analyzing Fig. 16.11,
we are asking the question of whether p orbitals can or cannot be used as basis vectors
to describe a favorable overlap, thus a chemical bond. The answer is yes for the px
orbital and no for the pz . A natural question is: What is so special about the px
orbital? Couldn’t we just rotate the pz orbital a quarter turn into a favorable overlap?
The answer is that there is nothing special about the px orbital and yes, we can
rotate the pz orbital into a position of favorable overlap, but in so doing, we rotate
the px orbital out of its position of favorable overlap so we haven’t really changed
anything. We are merely saying that if one of the p orbitals is favorable, the other two
orthogonal orbitals are not. The orbitals don’t care how we label them.
16.10 MANY-ELECTRON ATOMIC SYSTEMS
If we regard each electron as obeying an orbital (function) that is independent of

all other electrons except that it moves in the potential field of the nuclei and their
average electrostatic forces, we get a wave function for the entire system that is
approximated by a product of many one-electron orbitals called the Hartree product:

N
ψHartree = ψ1 (r1 )ψ2 (r2 ), . . . , ψ N (r N ) = ψi (ri )
i=1
This product produces N integrodifferential equations:
ĥ i ψi (ri ) = εi ψi (ri ), i = 1, 2, . . . , N
where the operator ĥ i includes the kinetic energy and the potential energy of attraction
along with the potential energy of interelectronic repulsion Vi :

Z
ĥ i = − 12 ∇i2 − + Vi (ψi (r j )), j = i
ri
P1: OTA/XYZ P2: ABC
The operator ĥ i is approximate because the position of each individual electron ji is

uncertain. The potential energy of repulsion between electrons Vi is included as an
average by integrating over the Born probability function for each electron j = i and
then summing the results:

ψ j (r j )2
Vi (ψi (r j )) = dτ
j=i
ri j
The N equations are then solved using approximations to ψ1 , ψ2 , . . . , ψn and enter-

ing an iterative procedure. Each of the solutions εi for an atomic system is the energy
of an atomic orbital 1s, 2s, 2p, and so on. Each energy coincides with a distinct
set of self-consistent one-electron wave functions ψ1 , ψ2 , . . . , ψn . The radial wave
functions lead to electron probability density functions with 1, 2, 3, . . . maxima at
specific distances from the nucleus corresponding to the 1s, 2s, 2p, and so on, “shells”
of electron density familiar from elementary chemistry.
PROBLEMS
Problem 16.1
The integral
∞
xe−x d x
2
−∞
is an “improper integral because both of its limits are infinite (one would do to make
it improper). Evaluate this integral, that is, find y(x) for
∞
xe−x d x
2
y(x) =
−∞
Verify your answer using a numerical integration computer routine such as Mathcad
C .
Problem 16.2
Show that the first two orbitals of a particle in a one-dimensional box of length 1 unit
are orthogonal, that is,
1
ψ1 ψ2 dτ = 0
0
for
√
ψ1 = 2 sin (π x)
P1: OTA/XYZ P2: ABC
PROBLEMS 265
and
√
ψ2 = 2 sin (2π x)
Problem 16.3
If the operator Â = d 2 /d x 2 , find the eigenfunction φ(x) and the eigenvalue a for the
eigenvalue equation
Âφ(x) = aφ(x)
Problem 16.4
According to a theory of Niels Bohr (1913) for an electron to move in a stable
classical orbit, the centrifugal force mv 2 /r pulling away from the nucleus must be
exactly balanced by the electrostatic force of attraction between the negative electron
and the positive proton e2 /r 2 .
(a) Write an expression for the velocity of the electron.

(b) Calculate the velocity of the electron. To get your answer in SI units,
use e2 /4π ε0 for the charge on the electron, where 4π ε0 = 1.113 ×
10−10 C2 s2 kg−1 m−3 is called the permittivity in a vacuum.
(c) Give units.
(d) Suppose the reader forgets to take a square root in the last step and arrives at
the result v = 4.784 × 1012 . How can he immediately know that something
has gone wrong?
Problem 16.5
2π x
Normalize the eigenfunction (x) = A sin for the particle in a one-dimensional
λ
box of dimension a.
Problem 16.6
Suggest a simple classical (macroscopic) mechanical system for which the probability
function varies in a regular way with position. What does the probability function
look like?
Problem 16.7
Carry out the “little algebra” to go from
4π 2 2m E
− 2 = − 2
2
l h̄
n
P1: OTA/XYZ P2: ABC
to
n2h2
E=
8m2
Problem 16.8
What is the probability of finding an electron within one Bohr radius of the nucleus
of the hydrogen atom?
Problem 16.9
The normalized wave function for a particle in a one-dimensional box of unit length
is
√ 2π x
(x) = 2 sin
λ
where x tells you where you are on the excursion of the particle in the one-dimensional
x-space. What is the probability of finding a particle in the first quarter of its excursion?
P1: OTA/XYZ P2: ABC
17
THE VARIATIONAL METHOD: ATOMS
Exact orbital solutions for the hydrogen atom cannot be replicated for other atoms
or molecules, even small ones, but they are of inestimable value in the mathematical
treatment of larger systems. The energy of nuclear attraction is very much larger than
the binding energies that dominate chemistry, so one can regard the chemical bond
as a perturbation on nuclear attraction. Thus the orbitals developed for H are the
foundation stones for larger systems of ions, atoms, and the interdependent groups
of atoms that we call molecules.
17.1 MORE ON THE VARIATIONAL METHOD
A precise expression of the variational equation for approximate wavefunctions is

∞
φ0∗ (τ ) Ĥ φ0 (τ ) dτ
E 0 = −∞∞
φ0∗ (τ )φ0 (τ ) dτ
−∞
In this notation, the subscripted 0 indicates the ground state energy and wave func-
tions, the angle brackets around E 0 indicate an expectation value, the asterisk (∗ )
designates the complex conjugate of the function, and τ is a variable indicating that
the integrals are to be taken over all space.

Copyright
267
P1: OTA/XYZ P2: ABC
268 THE VARIATIONAL METHOD: ATOMS
In our treatment we shall unencumber this expression from some of its notation
by writing the expectation value of the energy simply as

φ Ĥ φ dτ
E=
φφ dτ
In order to evaluate this energy, we need a trial function φ, which we are free to
choose, and the Hamiltonian Ĥ , which is determined by the system.
17.2 THE SECULAR DETERMINANT
In very many cases, the wave function will not be expressed as a single function at
all but as a sum of functions. To illustrate, let us take the sum of two additive terms
(strictly, basis vectors in a vector space.)
φ = c1 u 1 + c2 u 2
For a variational treatment of the sum as an approximate wave function, we need the
integrals

φ Ĥ φ dτ c1 u 1 + c2 u 2 Ĥ c1 u 1 + c2 u 2 dτ
E= =
φφ dτ (c1 u 1 + c2 u 2 ) (c1 u 1 + c2 u 2 ) dτ

c12 u 1 Ĥ u 1 dτ + c1 c2 u 1 Ĥ u 2 dτ + c1 c2 u 2 Ĥ u 1 dτ + c22 u 2 Ĥ u 2 dτ
=
c1 u 1 + 2c1 c2 u 2 u 1 + c2 u 22 dτ
2 2 2
c12 H11 + 2c1 c2 H12 + c2 H22

E=
c12 S11 + 2c1 c2 S12 + c22 S22
The integration problem has been broken into sums of smaller integrals. Each of
the smaller integrals above has been given a separate symbol. The integrals in the
numerator are denoted Hi j and the integrals in the denominator are denoted Si j . The
assumption that

u 1 Ĥ u 2 dτ = u 2 Ĥ u 1 dτ
P1: OTA/XYZ P2: ABC
THE SECULAR DETERMINANT 269
has been made in order to simplify the numerator. Operators Ĥ for which this is true
are called Hermitian operators. The equivalent substitution of S12 for S21 has been
made in the denominator.
One can multiply through by the denominator to find
2
c1 S11 + 2c1 c2 S12 + c22 S22 E = c12 H11 + 2c1 c2 H12 + c2 H22
Our objective is to differentiate the energy with respect to each of the minimization
parameters c1 and c2 so as to find the simultaneous minimum with respect to both
of them. This will be the minimum energy obtainable from the sum of functions
φ = c1 u 1 + c2 u 2 . Differentiating first with respect to c1 , we obtain
∂E 2
(2c1 S11 + 2c2 S12 ) E + c1 S11 + 2c1 c2 S12 + c22 S22 = 2c1 H11 + 2c2 H12
∂c1
Differentiating with respect to c2 , we get
∂E 2
(2c1 S12 + 2c2 S22 ) E + c1 S11 + 2c1 c2 S12 + c22 S22 = 2c1 H12 + 2c2 H22
∂c2
In order to find the minimum, we set
∂E ∂E
= =0
∂c1 ∂c2
This causes the two ∂ E/∂c terms to drop out:
∂E 2 ∂E 2
c S11 + 2c1 c2 S12 + c22 S22 = c S11 + 2c1 c2 S12 + c22 S22 = 0
∂c1 1 ∂c2 1
We are left with a pair of simultaneous equations:
(2c1 S11 + 2c2 S12 ) E = 2c1 H11 + 2c2 H12
and
(2c1 S12 + 2c2 S22 ) E = 2c1 H12 + 2c2 H22
These two equations are somewhat more conformable to computer solutions if we

divide by 2 and write them in the equivalent form:
(H11 − ES11 ) c1 + (H12 − ES12 ) c2 = 0

(H12 − ES12 ) c1 + (H22 − ES22 ) c2 = 0
P1: OTA/XYZ P2: ABC
The differences H11 − ES11 , and so on, are scalar energies, so we are left with nothing
more than a simple simultaneous equation pair like the ones we solved in high
school:
ax + by = p
cx + dy = q
The only difference is that the case of p = q = 0 is considered an unusual case

in elementary presentations, whereas it is the case of interest here. We wish to
solve the simultaneous equations in the H and S integrals for the solution set c1
and c2 .
Alas, we cannot find unique values for c1 and c2 because equations with p = q = 0
are inhomogeneous. Without an extra piece of information, we can only get the ratio
of c1 to c2 . This is not as much of a limitation as it may seem because for larger sets
of N equations, we can get the ratios of N − 1 coefficients to each other, where N
may be very large. In other words, we can get almost all of the information we want.
We can obtain the ratios of nonzero solutions we seek if the second equation is a
linear combination of the first. For a linearly dependent equation pair, it must be
possible to multiply one equation by a constant k and obtain the other. In this case,
the equations are linearly related if ka = c and kb = d. We can assure ourselves that
this is true and that the ratio of c1 to c2 exists if we stipulate that the determinant of
the coefficients be zero

a b ka kb

= = kad − ckb = cd − cd = 0
c d c d
In the case of the simultaneous equations we have
(H11 − ES11 ) c1 + (H12 − ES12 ) c2 = 0

(H12 − ES12 ) c1 + (H22 − ES22 ) c2 = 0
The stipulation for a linear pair is

H − ES H12 − ES12
11 11
=0
H12 − ES12 H22 − ES22
This is called the secular determinant. When expanded, it leads to a quadratic equation
with two roots, E 1 and E 2 , a pair of energy estimates. Upon imposing the normal-
ization condition, which is the extra piece of information necessary to complete the
solution, we take the lower of the two roots as the ground state energy for the system.
P1: OTA/XYZ P2: ABC
A VARIATIONAL TREATMENT FOR THE HYDROGEN ATOM 271
The same problem for a sum of N terms demands that the general secular deter-
minant be zero. This leads to N roots:

H11 − ES11 H12 − ES12 ... H1N − ES1N

H − ES H22 − ES22 ... H2N − ES2N
21 21
=0
.. .. ..
. . ... .

H − ES HN 2 − ES N 2 ... HNN − ES NN
N1 N1
Many readers will be familiar with the use of this mathematical formalism in Hückel
molecular orbital theory.
17.3 A VARIATIONAL TREATMENT FOR THE HYDROGEN ATOM:

THE ENERGY SPECTRUM
Starting with the radial Hamiltonian

2 1 d d 2
Ĥ = − 2 ∇1 −
1 2
= − 2 r2 −
r1 2r dr dr r
and the exact wave function for the hydrogen atom φ(r ) = e−αr , the variational
method leads to
h 2α2
E= − αe2
8π 2 m e
where m e is the mass of the electron (McQuarrie, 1983). This result is similar to
the exact solution for the particle in a cubic box E = h 2 n 2 /8ml 2 except that there
is a potential energy term −αe2 . We can expect some similarities between the two
systems. One similarity is that in each there is a spectrum of specific energy levels,
each corresponding to a specific quantum number. Another is that there is no zero
energy because the lowest quantum number is 1, not 0.
We would like to carry out a systematic search for the minimum energy, so we
set the first derivative of E with respect to α equal to zero. The derivative goes to
zero at a minimum, maximum, or inflection point. If our trial function is reasonably
good,1 only the minimum is found at this level of calculation. (For more complicated
systems, maxima, saddle points, and the like may be found.)
The derivative is
dE 22 α
= − e2 = 0
dα me
1 Aschemists, we bring centuries of empirical evidence to bear on the decision of what is or is not a good
atomic or molecular wave function. A negative exponential is reasonable for the hydrogen atom.
P1: OTA/XYZ P2: ABC
This amounts to finding the wave function φ(r ) = e−αr that has a specific α selected
from among all possible values of α. The minimization gives
m e e2
α=
2
From this, the minimum energy can be found to be [Problem 7.11, McQuarrie (1983)]

1 m e e4
E =− = −2.180 × 10−18 J
2 2
which is the ground state of H originally found by Bohr (Fig 15.2). The spectrum of
allowed energies can be calculated from the quantum numbers n:

1 m e e4 1
εH = − , n = 1, 2, 3, . . .
2 2 n2
The energies are negative for a stable system. The resulting energy level differences
can be used to determine the lines in the hydrogen spectrum in agreement with
Fig. 15.1. This should come as no surprise because we started out with the exact
ground state orbital.
At this point it is reasonable to define an atomic unit of energy, the hartree
E h = m e e4 /2 , which is twice the energy of the ground state of the hydrogen atom
εH . (Please do not confuse the measured energy εH with the unit of energy E H .)
εH ≡ 2E h = 2(2.180 × 10−18 J) = 4.359 × 10−18 J = 2625 kJ mol−1

= 627.51 kcal mol−1
The process of searching out energy minima in this way is frequently referred to
as minimization for obvious reasons. It is also called optimization because in cases
where the orbital is not known, the optimum energy and the best possible values of any
parameters in the orbital expression are found. These parameters are sometimes called
optimization parameters to distinguish them from known constants. One arrives at
different optimization parameters by different optimization procedures, but universal
constants like Planck’s constant h do not change.
17.3.1 Optimizing the Gaussian Function

Let us try a function that is similar to, but not identical to, the exact function. A
variational calculation can be carried out with the approximate Gaussian function
φ(r ) = e−αr . It is of the same form as the lowest hydrogen orbital except that the
2
numerator has an r 2 , where r should be. This function can be optimized by calculus
P1: OTA/XYZ P2: ABC
A VARIATIONAL TREATMENT FOR THE HYDROGEN ATOM 273
# gen
hatom gen
0 2
h
1 0
S 1 1.0
0.280000 1.0
****
FILE 17.1 Gaussian gen input for the hydrogen atom.
and algebra, but we choose to do it now by running the Gaussian

C computer program,
using a gen input. Selecting a value of α = 0.28, one finds
E ψ = − 0.4244 E h
The computed value of the energy is less negative and is therefore higher than the
exact value of 12 E h by about 15%.
17.3.2 A GAUSSIAN© HF Calculation of Eatom : Computer Files

The gen keyword in Gaussian permits you to input a value for α as the first entry in
the penultimate line of the input file, 0.280000 in File 17.1. We have done this for
α = 0.18, 0.28, and 0.38. The energy results extracted from much larger Gaussian
output files are shown as File 17.2.
Two things are clear even from this tiny extract from the output. The value for
α = 0.28 chosen from among the three α values leads to the lowest energy, but
the optimized α = 0.28 isn’t much better than the other two choices. Evidently the
energy minimum is at the bottom of a rather shallow energy well because none of
the three energy values is obviously wrong. Changing α from .28 to .38 gives an
increase in the energy value of only 0.0107 E h . Changing the input value in the
other direction by 0.1 results in a slightly larger energy change. This small result
of 0.0107 E h is a little deceptive, however, because the hartree is such a large unit
relative to chemical problems. The conversion factor from E h to kcal mol−1 being
627.51 kcal mol−1 , the calculated value in the hydrogen atom problem is a substantial
0.0107E h (627.51) = 6.7 kcal mol−1 .
0.180000 HF=-0.4070275
0.280000 HF=-0.4244132
0.380000 HF=-0.4136982
FILE 17.2 Energies drawn from the Gaussian gen output file for the hydrogen atom.
P1: OTA/XYZ P2: ABC
17.4 HELIUM
The helium atom is similar to the hydrogen atom, with the critical difference that there
are two electrons moving in the potential field of the nucleus. The nuclear charge is
+2. The Hamiltonian for the helium atom is
2 2 1
Ĥ = − 12 ∇12 − 12 ∇22 − − +
r1 r2 r12
Regrouping, we obtain

2 2 1
Ĥ = − 12 ∇12 − + − 12 ∇22 − +
r1 r2 r12
The first two terms on the right replicate the hydrogen case, except for a different
nuclear charge. The third term on the right, 1/r12 , is due to electrostatic repulsion
of the two electrons acting over the interelectronic distance r12 . This term does
not exist in the hydrogen Hamiltonian. The sum of two nuclear and one repulsion
Hamiltonians is
1
ĤHe = Ĥ1 + Ĥ2 +
r12
If this Hamiltonian were to operate on an exact, normalized wave function for helium,
the exact energy of the system would be obtained:
∞
E He = (r1 , r2 ) ĤHe (r1 , r2 ) dτ
0
The helium atom, however, is a three-particle system for which we cannot obtain an
exact solution. The orbital and the total energy must, of necessity, be approximate.
As a naive or zero-order approximation, we can simply ignore the “r12 term” and
allow the simplified Hamiltonian to operate on the 1s orbital of the H atom. The
result is
22 22
E He = − − = − 4.00 E h
2 2
which is 8 times the exact energy of the hydrogen atom (− 12 E h ). The 2 in the
numerators are the nuclear charge Z = 2. In general, the energy of any hydrogen-like
atom or ion is −Z 2 /2 hartrees per electron, provided that we ignore interelectronic
electrostatic repulsion.
We can compare this result with the first and second ionization potentials (IP) for
helium, which are energies that must go into the system to bring about ionization:
He → He+ + e− → He2+ + e−
P1: OTA/XYZ P2: ABC
HELIUM 275
These energies can be measured experimentally with considerable accuracy. Since

they are the energy necessary to pull an electron away from the helium atom, they
are equal and opposite in sign to the binding energy of the ionized electron. In this
way we have a measure of both (a) the first or “outside” electron attracted to the He+
system and (b) the inside electron attracted to the He2+ nucleus. This second energy
IP2 can be calculated exactly because He+ is a one-electron (two-particle) system.
The calculated second ionization potential IP2 is exactly 2.0 E h .
If we compare the calculated total ionization potential, IP = 4.00 E h , with the
experimental value, IP = 2.904 hartrees, the result is very bad. The magnitude of the
disaster is even more obvious if we subtract the known second ionization potential,
IP2 = 2.0, from the total IP to find the first ionization potential, IP1 :
IP1 = IPtotal − IP2 = 2.904 − 2.000 = 0.904 E h
Under the approximations we have made, IP1 (calculated) = 2.000 E h is about 110%
in error. Clearly, we cannot ignore interelectronic repulsion.
17.4.1 An SCF Variational Ionization Potential for Helium

One approach to the problem of the r12 term is a variational self-consistent field
approximation. We shall start from Hartree’s orbital approximation, assuming that
the orbital of helium is separable into two one-electron orbitals (1, 2) = φ1 (1) φ2 (2).
It is reasonable to use the H atom as our starting point with the same kinetic energy
operator for each starting orbital φ1 (1) and φ2 (2)and −2/r as the potential energy
part for attraction of each electron to a nuclear charge of 2:
1 d 2 d 2
Ĥ = − 2
r −
2r dr dr r
Although we are solving for one-electron orbitals, φ1 and φ2 , we do not want to fall
into the trap of the last calculation. This time we shall include a potential energy term
V1 to account for the repulsion between the negative charge on the electron arbitrarily
designated electron 1, exerted by the other electron which we shall call electron 2.
We don’t know where electron 2 is, so we must integrate over all possible locations
dτ :
∞
1
V1 = φ2 φ2 dτ
0 r12
The entire Hamiltonian for electron 1 is

∞
1 d 2 d 2 1
Ĥ1 = − 2 r1 − + φ2 φ2 dτ
2r1 dr1 dr1 r1 0 r12
P1: OTA/XYZ P2: ABC
The same treatment produces a similar operator for electron 2:

∞
1 d 2 d 2 1
Ĥ2 = − r − + φ1 φ1 dτ
2r22 dr2 2 dr2 r2 0 r12
We do not know the orbitals of the electrons either. We can reasonably assume
that the ground state orbitals of electrons 1 and 2 are similar but not identical to the
1s orbital of hydrogen:

a 3 −αr1
φ1 = e
π
and

b3 −αr2
φ2 = e
π
The integral in Ĥ1 , representing the Coulombic interaction between electron 1 at r1

and electron 2 somewhere in orbital φ2 , has been evaluated for Slater-type orbitals
(Rioux, 1987; McQuarrie, 1983) and is
∞
1 1
V1 = φ2 φ2 dτ = 1 − (1 + b r1 )e−2b r1
0 r12 r1
Now the approximate Hamiltonian for electron 1 is
1 d 2 d 2 1
ĥ 1 = − r − + 1 − (1 + br1 ) e−2br1
2r12 dr1 1 dr1 r1 r1
with a similar expression for ĥ 2 involving ar2 in place of br1 in the Slater orbital.
The orbital is normalized, so the energy of electron 1 is
∞
E1 = φ1 ĥ 1 φ1 dτ
0
with a similar expression for E 2 .

Calculating E 1 requires solution of three integrals:
∞ ∞ ∞
Z
E1 = φ1 − 12 ∇12 φ1 dτ − φ1 φ1 dτ + φ1 (V1 ) φ1 dτ
0 0 r1 0
They yield (Rioux, 1987) three terms for the energy of the electron in orbital φ1 :

a2 ab a 2 + 3ab + b2
E1 = − Za +
2 (a + b)3
P1: OTA/XYZ P2: ABC
HELIUM 277
with a similar expression for E 2 except that b replaces a in the first two terms on the
right

b2 ab a 2 + 3ab + b2
E2 = − Zb +
2 (a + b)3
The parameters a and b in the Slater-type orbitals for electrons 1 and 2 are min-
imization parameters representing an effective nuclear charge as “experienced” by
each electron, partially shielded by the other electron from the full nuclear charge.
The SCF strategy is to minimize E 1 using an arbitrary starting b and to find a at
the minimum. This a value is then used to find b at the minimum E 1 . This value
then replaces the starting b value and a new minimization cycle produces a new
a and b, and so on until there is no progressive difference in E 1 . The electrical
field experienced by the electrons is now self-consistent, hence it is a self-consistent
field SCF.
In this particular case, the calculations are completely symmetrical. Everything
we have said for a we can also say for b. At self-consistency, a = b and we can
substitute a for b at any point in the iterative process, knowing that as we ap-
proach self-consistency for one, we approach the same self-consistent value for the
other.
A reasonable thing to do at the end of each iteration would be to calculate
the total energy of the atom as the sum of its two electronic energies E He =
E 1 + E 2 , but in so doing, we would be calculating the interelectronic repulsion
ab(a 2 + 3ab + b2 )/(a + b)3 twice, once as an r12 repulsive energy and once as an
r21 repulsion. The r21 repulsion should be dropped to avoid double counting, leaving
b2
E He = E 1 + E 2 = E 1 + − Zb
2
as the correct energy of the helium atom.

Although the Hartree procedure for atoms reaches self-consistency and gives a
qualitative picture of the electron probability densities, the ionization energy is still in
error by a substantial amount (0.014 h ∼= 37 kJ mol−1 on three iterations). Something
is missing. A problem arises with the Hartree product when we exchange one atomic
orbital with its adjacent neighbor
ψi (ri )ψ j (r j ) → ψ j (r j )ψi (ri )
The Hartree product is the same after exchange as it was before exchange. This is
equivalent to saying that ψi (ri ) and ψ j (r j ) are identical in all respects, even though
we recall from elementary chemistry that no two electronic wave functions can be
exactly alike and that each electron is represented by a unique set of four quantum
numbers n, l, m, and s. In three-dimensional Cartesian space, it is not difficult to
suppose that there are three quantum numbers n, l, and m in the complete solution
P1: OTA/XYZ P2: ABC
for the hydrogen atom, one for each dimension; but by using four quantum numbers,
we imply something beyond the solution in 3-space.
If helium is in the excited state, the electrons are in different spatial orbitals; but
in the ground state, both electrons have the same 1s orbital description. A distin-
guishing fourth quantum number called the spin quantum number s = ± 12 must be
introduced to give two distinct 1s orbitals in helium, one designated 1sα and the other
designated 1sβ.
17.5 SPIN
If we ionize the electrons and examine their spins, we may find that the first electron
ionized (the outer electron) has spin α implying that the inner electron (2) has spin β
and that the atomic orbital 1s is
1s = 1s(1)α + 1s(2)β
On the other hand, we may find that the first ionized electron has spin β leading to
1s = 1s(1)β + 1s(2)α
These two results have exactly the same probability. We are in a logical dilemma
that results from the Hartree independent orbital hypothesis. It cannot be said that an
electron is in either the α spin orbital or the β spin orbital, only that they are both in
an orbital that is a linear combination of equally weighted space-spin basis functions.
The two plausible basis functions
1s(1)α1s(2)β and 1s(1)β1s(2)α
can be combined in two possible ways:
1s(1)α1s(2)β + 1s(1)β1s(2)α
and
1s(1)α1s(2)β − 1s(1)β1s(2)α
Only the last antisymmetrical linear combination is acceptable for the 1s electrons in
helium.
17.6 BOSONS AND FERMIONS
There are only two kinds of elementary particles in the universe, bosons and fermions.
Bosons are symmetrical under exchange and fermions are antisymmetrical under
P1: OTA/XYZ P2: ABC
SLATER DETERMINANTS 279
exchange. Electrons are fermions. If both electrons in the 1s orbital had the same
spin, the negative (antisymmetric) combination above would integrate to zero and
there would be no probability of finding the electron. That is why electrons must have
opposite spins if they are to reside in the same space orbital.
17.7 SLATER DETERMINANTS
The linear combination selected above for ground state helium is the same as the ex-
pansion of a 2 × 2 determinant, which is the simplest example of a Slater determinant
(Section 15.3):

1s(1)α 1s(1)β

1s(1)α1s(2)β − 1s(1)β1s(2)α =
1s(2)α 1s(2)β
Larger Slater determinants also coincide with uniquely acceptable linear combina-
tions for fermions. In the general n × n case with some simplification of notation,
the Slater determinant is

φ1 α1 φ1 β1 ... φn α1 φn β1

1 φ1 α2 φ1 β2 ...

ψ(1, 2, . . . , n) = √
n! . . . ...

φ α φ1 βn ··· φn αn φn βn
1 n
√
where the premultiplier 1/ n! is a normalization constant. The many-electron wave
function for orthonormal basis functions ψi , in more concise notation, is now written
(Pople, Nobel Prize, 1998)
ψ = (n!)− 2 det [(ψ1 α) (ψ1 β) (ψ2 α) . . .]

1
The Slater determinant always produces a linear combination of one-electron orbitals

that allows for electron indistinguishability by giving equal weight to electrons of
opposing spins in each of n orbitals. The Slater determinantal molecular orbital
and only the Slater determinant satisfies the two great generalizations of quantum
chemistry, the Heisenberg principle of uncertainty (which is why you can’t tell in
advance whether the first electron ionized from He will be α or β) and the Fermi–Dirac
principle of antisymmetric fermion exchange.
In addition to his work on determinantal wave functions, Slater also fitted functions
to numerical compilations of SCF data and obtained orbitals in analytic form. These
Slater-type orbitals (STO) resemble hydrogen wave functions but have adjustable
parameters to account for field differences between hydrogen and many-electron
P1: OTA/XYZ P2: ABC
hydrogen-like atoms. A few STOs are
ψ1s = N1s e−αr

ψ2s = N2s r e−αr/2
ψ2 px = N2 px r 2 e−αr/2
etc.
The normalization constants N are N1s = (α 3 /π )1/2 , N2s = (α 3 /96π )1/2 , and so on.
Clearly, the entire system depends upon α, which is an empirically fitted parameter.
The parameter is written as
α=Z−S
where Z is the atomic number (charge on the nucleus) and S is a shielding constant
which accounts for the diminution of nuclear charge experienced by an outer electron
owing to shielding by inner electrons.
This method can, with computational difficulty, be extended to atoms larger than
helium and to a few small molecules. In a molecular problem as simple as methane,
however, the dimension of the Slater determinant is 16 × 16. Clearly, molecular
problems are daunting to anyone attempting hand calculations. Practical applications
awaited widespread availability of powerful digital computers. Although we shall
soon consider more challenging problems of correlated wave functions, Slater orbitals
and the Hartree–Fock (HF) equations contain the essence of atomic and molecular
orbital theory. Much important structural and thermochemical information can be
gleaned from them (Hehre, 2006).
17.8 THE AUFBAU PRINCIPLE
Given the orbital structure of the elements in the first three rows of the periodic
table, one can predict, with reasonable certainty, their chemical properties. Hy-
drogen ionizes to form the H+ ion (hydrated in aqueous media), and it can be
persuaded to take on one electron to form the H− hydride ion but helium does nei-
ther because its 1s orbital is “full” (Pauli). The electronic configuration of helium
is He 1s 2 .
The first full row of the periodic table starts with lithium, which, like hydrogen,
can lose an electron, this time from the 2s orbital as its characteristic reaction. The 2s
probability density antinode is considerably more distant from the nucleus than the
1s orbital of hydrogen, making this loss of an electron from Li even more facile than
the equivalent loss from H. Beryllium has two electrons in the 2s orbital. Electron
loss from Be takes place easily, but not as easily as Li because its nuclear charge is
one more than in Li. Be is a metal but is not as metallic as Li. Here the “metallic
P1: OTA/XYZ P2: ABC
THE SCF ENERGIES OF FIRST-ROW ATOMS AND IONS 281
nature” of an element is defined by its tendency to lose an electron. The next six
elements—B, C, N, O, F, and Ne—correspond to the filling of the 2 px , 2 p y , and 2 pz
orbitals with two electrons each. The 3s and 3p orbitals are filled in a similar way
presenting elements that are roughly similar to their first row counterparts.
Atomic energy levels are not energetically very far apart, and one should not
expect complete regularity in the table. Small energetic perturbations cause them to
shift around a bit. Sometimes the order of orbital filling, s, p, d, . . . is disrupted by
external fields brought in by orbital interactions, ligands, solvents, or intentionally
imposed magnetic fields.
17.9 THE SCF ENERGIES OF FIRST-ROW ATOMS AND IONS
Starting from any reasonable value of a, one can calculate a value for b (Section
17.4.1), substitute it as a parameter at the top of a Mathcad C program, and repeat
the calculation to self-consistency for other small atoms. (Ignore negligibly small
imaginary roots.) Usually only three or four iterations are necessary.
The same calculation can be carried out for the single positive ions. The energy
difference between the atom and its single + ion is the energy necessary to drive
off one electron, that is, the first ionization potential IP1 . A collection of IP1 values
calculated in this way is depicted in Fig. 17.1. Notice the sharp drop at Li (Z = 3),
B(Z = 5), and O(Z = 8). The drop at Li results from the greater distance of the
probability antinode for 2s relative to 1s. The drop at B results from shielding of the
2 px orbital by the 1s 2 electrons, and the drop at O results from the increased shielding
of the 2 px2 2 p y 2 pz configuration relative to the 2 px 2 p y 2 pz of N (Z = 7).
0.9
0.8
0.7
IP, hartrees
0.6
0.5
0.4
0.3
0.2
0.1
0 2 4 6 8 10 12
Atomic Number
FIGURE 17.1 Calculated IP1 for elements 1–10. Experimental results follow this pattern,
but they are not identical. A more complete graph of this kind is presented in most general
chemistry texts.
P1: OTA/XYZ P2: ABC
17.10 SLATER-TYPE ORBITALS (STO)
In the absence of an exact solution of the Schrödinger equation for atoms beyond
hydrogen in the periodic table, John Slater devised a set of empirical rules for writing
down approximate wave functions. Every entry in Table 17.1 is a function involving
the following: the radial distance in units of bohrs, r/a0 ; a negative exponential
e−Zr /na0 , where n is the principal quantum number, 1, 2, and 3 for H, He, and the first
two full rows of the periodic table; and a0 is the Bohr radius. Z is the nuclear charge.
Slater wrote an approximate wave function involving only the radial part for the first
two full rows in the periodic table, ignoring the spherical harmonics. The function
takes the form
−(Z −s)
−
φ(r ) = re na0
An adjustable parameter s is called the screening constant, and Z is also an adjustable

parameter called the effective quantum number. For the first-, second-, and third-row
elements, Z simply is 1, 2, and 3, though it becomes nonintegral later in the table.
What will concern us most is the screening constant, which can be arrived at by fitting
experimental data.
Slater’s rules for determining what are now called Slater-type orbitals (STO) are
simple for the first three rows of the table. They become more complicated and
less reliable lower in the table. We have already seen that the screening constant
for the second electron in the helium atom is about 0.3. For 1s electrons in higher
atoms, Slater modified this slightly to 0.35. When a 1s electron screens a p electron,
screening is more effective than simple electron–electron screening in helium because
the probability density lobes of the 2p orbitals lie outside of the 1s orbital. Slater chose
a screening constant of 0.85. Notice that the polynomial parts of the wave function
(Table 17.1) are gone. Only the “tail” of the wave function is represented because that
is the part involved in the first ionization potential and, for the most part, in chemical
bonding. For more detail on Slater’s rules, see Problem 15.79 in Levine (2000).
TABLE 17.1 Slater’s Rules
1. Array orbitals 1s, 2s, 2p, 3s, 3p, 3d, . . .

2. Consider only the orbital containing the electron in question and the one below it.
3. Electrons lower than (2) are interior orbitals.
4. Orbitals higher than (2) are exterior orbitals.
5. Electrons designated s (1s, 2s, etc) have a screening constant of 0.30.
6. Electrons in the same orbital have a screening constant of 0.85.
−(Z −s)
Electrons are in the orbital φ(r ) = re na0 . For example, in bohrs and hartrees, the orbital
−(Z −s) −(Z −s) −(1−.30)
electron in helium has φHe (r ) = re n = re n = re (1) = 1.7r due to shielding.
P1: OTA/XYZ P2: ABC
L
S
p
FIGURE 17.2 Linear and angular momentum vectors.
17.11 SPIN–ORBIT COUPLING
An object moving in a straight line tends to conserve its linear momentum p by

continuing to move in a straight line. A spinning wheel tends to conserve its angular
momentum L by continuing to spin. Although electrons cannot be correctly described
by a deterministic circular path, they have an orbital angular momentum L and, as we
have seen, electrons have a property analogous to spin, so we anticipate (correctly) a
spin angular momentum S (Fig. 17.2). These momenta are vectors.
An atom has an angular momentum that is the sum of its electronic orbital and
spin angular momenta. Orbital and spin angular momenta couple either by vector
addition or vector subtraction according to whether they are in the same direction or
are opposed. This results in anomalous spectral splitting. The appearance of a pair
of sodium D lines where one line was expected is a result of spin–orbit coupling.
The hydrogen spectrum also shows spectral splitting. For example, the 656.2 nm
(6562 Å) “line” of hydrogen is not really a line, but a doublet at 656.272 and
656.285 nm under high resolution.
In many-electron atoms, many vector combinations at different angles produce
complicated vector combinations of L and S which result in complicated spectral
splittings. These patterns are not completely understood. Partial explanations valid
for lower atomic mass elements include Russell–Saunders or LS coupling patterns
and spectral splittings.
Example 17.1 A Mathcad© SCF Calculation

A calculation of the first three approximations to the SCF energy ε (a, b) of the helium
atom is shown in File 17.3.
The first iteration of the SCF procedure in File 17.3 produces an approximation to
the first ionization potential of He −ε(a, b) = IP1 (calc) = −(−0.812) hartrees—that
is, 10.2% too small. This is not very good, but it is a great improvement over the
>100% error we found when the r12 term was ignored. Continuing the calculation
and substituting for the initial value of b, we minimize to find a new value of
IP1 (calc) = 0.925, 2.4% in error, followed by IP1 (calc) = 0.889, 1.5% in error on
P1: OTA/XYZ P2: ABC
Z := 2.000 a := 2.000 b := 2.000

2
a a · b(a + 3 · a · b + b2 )
2
ε(a, b) := −Z·a+
2 (a + b)3

d
a := root ε(a, b), a
da
a = 1.601 ε(a, b) = −0.812
b := 1.601

d
da
a = 1.712 ε(a, b) = −0.925
b := 1.712

d
da
a = 1.681 ε(a, b) = −0.889
FILE 17.3 Mathcad C calculation of the ionization potential of helium. An approximate
screening constant of 0.3 gives a = (z − s) ∼

= 1.7.
the third iteration, approaching the experimental value. Notice that the calculated IP1
on the second iteration is too large. Sometimes the solution of the iterative procedure
is approached asymptotically, and sometimes the approach is oscillatory.
Example 17.2 The Slater Orbital of Oxygen

Find the Slater orbital of oxygen.
Solution 17.2 First observe that the nuclear charge is Z = 8. The 1s 2 part of the
screening constant on the valence electron is 2(0.85) = 1.70 and the part of the
screening constant for the electrons in the 2s 2 and 2 p 3 shell is s = 5(.35) = 1.75.
The interior electrons are more effective (0.85) in screening the valence electron than
the electrons that share the second valence shell (0.35). We do not count the valence
electron because it cannot screen itself. The screening electrons are always one less
than the number of electrons in the neutral atom, in this case 2 + 2 + 3 = 7. The
total screening constant is 1.70 + 1.75 = 3.45. This gives an STO for oxygen:
s = 2(0.85) + 5(0.35) = 1.7 + 1.75 = 3.45

8 − 3.45
Z (shielded) =
2
−(z−s)/na
φ(r ) = r e = r e−(8−3.45)/2a = r e−2.28/a
Other atomic STOs can be found by applying the same routine.

P1: OTA/XYZ P2: ABC
Problem 17.1
Evaluate the following determinants:

1 0 x 1 sin θ cos θ

, ,
0 1 1 x − cos θ sin θ
Problem 17.2
What is the energy increase relative to the ground state when one of the quantum
numbers, say n z , for the particle in a cubic box is raised from 1 to 2? What is the
degeneracy of the resulting wave function and probability distribution?
Problem 17.3
1,3-Pentadiene shows a strong absorption peak at 45,000 cm−1 (Ege, 1994). What
is the wavelength of this radiation in (a) centimeters, (b) meters, (c) nanometers,
(d) picometers, and (e) angstroms? What is its frequency in hertz? What is its energy
in joules?
Problem 17.4
What is the actual energy increase for the single excitation n z = 1 → n z = 2 in
Problem 17.2 if the trapped particle is an electron and the dimension of the box is
approximately a bond length, 1.5 Å? Give your answer in joules. What wavelength of
light will promote an electron from the ground state to one of the degenerate excited
states? Give your answer in nanometers.
Problem 17.5
The length of an ethene molecule is about 153 pm according to MM3. Using Hückel
theory and the particle in a one-dimensional box as a model, in what region of
the electromagnetic spectrum (X-ray, UV, vis, IR, etc.) is the radiation necessary to
promote an electron from the highest occupied molecular orbital (HOMO) to the
lowest unoccupied molecular orbital (LUMO)?
Problem 17.6
Find the Slater-type orbital (STO) of nitrogen.
Problem 17.7
Write down the unnormalized Slater determinant for He in the ground state. The He
atom is a three-particle, two-electron system. Set the Slater determinant, including a
P1: OTA/XYZ P2: ABC
normalization constant, equal to the wave function of He. Expand the determinant.
What is the normalization constant?
Problem 17.8
We have seen that the linear combination that is the difference between terms is
equivalent to the Slater determinantal wave function for helium.
1
ψHe (1, 2) = √ (1s(1)α(1)1s(2)β(2) − 1s(2)α(2)1s(1)β(1))
n

1 1s(1)α(1) 1s(1)β(1)
≡ √
n 1s(2)α(2) 1s(2)β(2)
What is wrong with the positive combination?
1
ψHe (1, 2) √ (1s(1)α(1)1s(2)β(2) + 1s(2)α(2)1s(1)β(1))
n
Hint: Try switching electrons ψHe (1, 2) → ψHe (2, 1).
Problem 17.9
A variational treatment of atomic helium gave
5
E He = −(−2Z eff
2
+ Z eff + 4Z eff (Z eff − 2))E H
4
Find Z eff and E He given that E H = −13.6 eV. Give your answer also in electron volts.
The experimental result is E He = −79.0 eV. What is the % error of this variational
treatment?
P1: OTA/XYZ P2: ABC
18
EXPERIMENTAL DETERMINATION
OF MOLECULAR STRUCTURE
Acceptance, at least provisional acceptance, of atomic theory goes back two and
a half millennia, but it inevitably raises the question of what atoms look like. The
nature of the world around us—air, earth, fire, and water—is governed less by atomic
properties than by the physical and chemical properties of molecules. But acceptance
of molecular theory, a child of the nineteenth century, raises the same question:
What do molecules look like? Experimental evidence for the existence and physical
appearance of molecules comes from interactions of confined pure samples with their
environment, largely through energy transfer by the very broad energy spectrum of
electromagnetic radiation extending all the way from gamma and X rays to low-
energy radio waves.
18.1 THE HARMONIC OSCILLATOR
A mass attached to a fixed beam by a spring can be set into oscillation by pulling
it down and letting it go (Fig. 18.1). An ordinary spring obeys Hooke’s law, at least
approximately:
f = −k f z(t)
where f is the force exerted by the spring on the moving mass when it is displaced
by a distance z from the equilibrium distance z 0 . The proportionality constant k f is

Copyright
287
P1: OTA/XYZ P2: ABC
288 EXPERIMENTAL DETERMINATION OF MOLECULAR STRUCTURE
z0
FIGURE 18.1 A classical harmonic oscillator. The equilibrium position on the vertical axis
is z 0 .
called the force constant, and z is a function of time z(t). The sign is negative because
the force is a restoring force acting in opposition to the excursion of the mass away
from z 0 .
By Newton’s second law, f = ma where a is the acceleration, d 2 z(t)/dt 2 . These
two expressions for the force can be set equal to one another:
d 2 z(t)
m = −k f z(t)
dt 2
d 2 z(t) kf
2
= − z(t)
dt m
This is a wave equation of the kind described in Section 16.2:
d 2 φ(x) 4π 2
= − φ(x)
dx2 λ2
In the analogous harmonic oscillator case, we have
d 2 z(t) 4π 2
= − z(t)
dt 2 λ2
Comparing the two expressions for acceleration,
kf 4π 2
= 2
m λ
leads to

1 1 kf
=
λ 2π m
The speed of propagation of electromagnetic radiation is c = 2.998 × 108 m s−1 ,

which is the number of waves per second (frequency ν) times the distance covered
by each wave (wavelength λ) c = ν λ. When electromagnetic radiation of many fre-
quencies falls upon an ideal quantum harmonic oscillator, most of it bounces off but a
selected frequency is absorbed, the one that promotes the oscillator from one quantum
P1: OTA/XYZ P2: ABC
THE HOOKE’S LAW POTENTIAL WELL 289
state to the next higher quantum state. One of the most common laboratory instruments
is the spectrophotometer, which is calibrated to record selected absorbed frequencies
called resonance frequencies. These are often reported as wave numbers ν̄, which are
frequencies divided by the speed of radiation. This is just the experimental measure-
ment we want in order to calculate force constants. In units of kg m−2 , we have

ν 1 1 kf
ν̄ = = =
c λ 2π m
k f = 4π 2 m ν̄ 2
The force constant is also expressed in newtons per meter (N m−1 ) because a
newton is a kg m−1 .
18.2 THE HOOKE’S LAW POTENTIAL WELL
In the harmonic oscillator, potential energy is increased by the work done on the mass
by moving it dz against the opposing force – f of the constraining spring. This work,
dw = − f dz, is not lost. It is stored in the spring as potential energy V:
d V = dw = − f dz = k f z dz
Integrating, we obtain
V z
k f z2
dV = k f z dz =
V0 z0 2
It is convenient to define the potential energy as zero, V0 = 0, at the equilibrium

position of the oscillating mass, z 0 = 0 (Fig. 18.1). The oscillating mass passes
through z 0 = 0 on each oscillation but continues along the z axis to nonzero values
V
in the opposite direction, whereupon V0 d V = V − V0 = V and
k f z2
V =
2
The potential energy defined in this way is always positive. It is symmetrical about
the equilibrium position of a perfect Hooke’s law spring and it is steep for a stiff
spring (larger k f ) but open for a weak spring (relatively smaller k f ). The excursions z
away from z 0 are positive and negative, but the energy always goes up with z because
the square of z is always positive (Fig. 18.2).
Upon solving the Schrödinger equation for this system, one finds that the simple
harmonic oscillator has evenly spaced energy levels and that there is a half quantum
of energy at the bottom of the well (called the zero point energy). Because the energy
of transition from one level to an adjacent neighbor level is the same no matter where
P1: OTA/XYZ P2: ABC
z
z0
FIGURE 18.2 Parabolic potential wells for the harmonic oscillator. The narrow well has the
larger force constant. Only one set of quantum levels is shown.
you start, the absorption spectrum for the harmonic oscillator consists of only one
line with frequency ν = (E/ h) s−1 because E = h ν.
The simple harmonic oscillator is a fairly good model for a hydrogen atom sus-
pended from a heavy carbon atom framework (Fig. 18.1). From C–H resonance
frequencies of about 2900–3000 cm−1 , one finds the force constant to be about
500 N m−1 . Typical experimental values for molecular force constants vary consid-
erably, from about 100 N m−1 to about 800 N m−1 .
The reader with practical knowledge of infrared spectra will find a discrepancy
between the complicated structure of a real IR band spectrum and the single line
predicted by Hooke’s law. This is the result of many factors, including the failure
of Hooke’s law, energy coupling among chemical bonds, and many other motions
(bending, torsion, etc.) that are possible in a real polyatomic molecule. Nevertheless,
the presence of a strong peak near 2900–3000 cm−1 is a good indicator of a C–H
stretch lurking somewhere in the molecule. Other characteristic frequencies are used
in qualitative IR “fingerprint” analysis.
18.3 DIATOMIC MOLECULES
For diatomic molecules attached by a chemical bond, the picture is very similar to
the harmonic oscillator of one mass. The atoms vibrate harmonically relative to one
another with a natural frequency determined by their mass and the strength of the
electronic spring connecting them. One replaces the mass of the simple harmonic
oscillator with the reduced mass of the diatomic molecule µ = m 1 m 2 /m 1 + m 2 ,
where m 1 and m 2 are the atomic masses, and proceeds with the calculation. The
problem has indeed been reduced from one of two masses vibrating relative to one
another to one of a single fictitious mass µ vibrating relative to a central point.
18.4 THE QUANTUM RIGID ROTOR
A small mass on a circular orbit in a fixed plane shows quantum phenomena. The
energy level spacing shows a pattern that is similar to the energy levels of the particle
P1: OTA/XYZ P2: ABC
THE QUANTUM RIGID ROTOR 291
in a box (Chapter 16) except that the quantum number n is replaced by J and m times
the length of the constraining geometry l 2 is replaced by the moment of inertia of the
rotating mass I = mr 2
J 2h̄ 2
E=
2I
Because the mass can rotate in either direction (or stand still at J = 0), the allowed
quantum numbers are J = 0, ±1, ±2, . . . .
When diatomic molecules rotating about their center of mass (balance point) are
considered, the reduced mass of the diatomic molecule is used in the same way as it
was for the vibrating two mass problem. Diatomic molecules are not constrained to
move in a fixed plane, rather, their plane of rotation can tilt at angles from 0 to π .
This added degree of freedom changes the energy level spacing to
h̄ 2
E = J (J + 1)
2I
The energy levels diverge with increase in the quantum number J = 0, ±1, ±2, . . . ,
giving the values J (J + 1) = 0, 2, 6, 12, . . . . The problem is again reduced to one
of a fictitious mass, this time with a moment of inertia I = µ r 2 , rotating anywhere
on the surface of a sphere. For absorption of a resonance frequency to occur, the
rotational state must be changed by an increase in energy from one level to the next
J → J + 1 with
E J+1 − E J = 2, 4, 6, 8, . . . joules
The spacing between resonance frequencies is 4 − 2 = 2, 6 − 4 = 2, 8 − 6 = 2, . . .
joules, by which we predict an absorption spectrum consisting of a series of lines of
different frequencies separated by an energy of 2(h̄ 2 /2I ) (Fig. 18.3). Measuring ν̄
leads to I, which gives the bond length r through I = µ r 2 .
Vibrational and rotational spectroscopic transitions can occur simultaneously,
leading to increased complexity of the experimental spectrum. Conversely, some
FIGURE 18.3 Energy levels within a simple rotor.

P1: OTA/XYZ P2: ABC
expected transitions are conspicuously absent from experimental spectra owing to

selection rules which do not permit them. Despite these deviations from the simple
models described here, much invaluable bond distance information has been ob-
tained from microwave rotational spectra. For example, CO has a rotational spacing
of ν̄ = 3.9 cm−1 (in the far-infrared region), which leads to a bond length of 113 pm.
By the way, your microwave oven activates a rocking frequency in H2 O molecules,
thereby transmitting energy (heat) to your morning coffee.
18.5 MICROWAVE SPECTROSCOPY: BOND STRENGTH

AND BOND LENGTH
Given the input experimental

2 data of the fundamental vibrational frequency ν0 and the
h̄
line separation 2 2I usually written 2B, in a vibration–rotation band of a diatomic
molecule, the bond strength in terms of the Hooke’s law force constant k f and the
bond length in picometers (pm) can be calculated.
18.6 ELECTRONIC SPECTRA
Electronic transitions require more energy than vibrational or rotational transitions.

They produce spectral peaks that frequently fall in the UV or visible part of the
electromagnetic spectrum. The π → π ∗ transition in ethene is an electronic transition
from the highest occupied molecular orbital (HOMO) to the lowest unoccupied
molecular orbital (LUMO). A simple model of electronic transition spectroscopy of
this kind is that of a free electron trapped in a one-dimensional potential well, which
is the length of the π system in unsaturated molecules. An MM calculation (Chapter
19) of the distances between the terminal carbon atoms in ethene and 1,3-butadiene
gives the values 134 and 359 nm. The two electrons in ethene are in the lowest energy
level with E = h 2 /8m e l 2 . In the π → π ∗ transition, one of them progresses to the
next higher energy state with E ∗ = 4h 2 /8m e l 2 . The energy difference is

3h 2 3h 2 3h 2
E = = = 5.56 × 10−5
8m e l 2 8m e (134)2 8m e
The 1,3-butadiene case has 4 electrons in two π bonds (Fig. 18.4); hence the lowest
two levels are occupied and the transition is from the n = 2 to the n = 3 or n 2 = 4
to the n 2 = 9 level:

9h 2 4h 2 5h 2 5h 2 3h 2
E = − = = = 3.88 × 10−5
8m e l 2 8m e l 2 8m e l 2 8m e (359)2 8m e
Even though the quantum number change in the second case is larger (5 vs. 3), the
transition for 1,3-butadiene is the smaller of the two energy differences because of
P1: OTA/XYZ P2: ABC
ELECTRONIC SPECTRA 293
e e
4
3
e e e e
0
FIGURE 18.4 Electronic promotion in alkenes.
the square of the length in the denominator. The extended double-bond system in 1,3-
butadiene decreases the energy and frequency and, because of the inverse relationship
between ν and λ, it increases the wavelength. Experimentally, this predicted shift to
longer λ is verified.
It is true in general that extended conjugated double-bond systems absorb at
wavelengths that tend toward the longer wavelengths in the UV, and even into the
visible region of the electromagnetic spectrum. This can be seen in the orange color
of carotene from carrots and the red of lycopene, the principal natural coloring agent
in tomatoes. Both of these molecules have long conjugated π electron chains and
follow the qualitative trend predicted by the free electron model.
500
450
400
Wavelength (nm)
350
300
250
200
150
100
0 2 4 6 8 10
Double Bonds
FIGURE 18.5 Absorption wavelengths of conjugated polyalkenes. The wavelengths range

from the far ultraviolet well into the visible region. (Taken in part from Streitwieser, 1961.)
P1: OTA/XYZ P2: ABC
Field
FIGURE 18.6 A charged parallel plate capacitor.
18.7 DIPOLE MOMENTS
If a parallel plate capacitor with a vacuum between its plates is brought up to a charge
of Q coulombs by imposing a potential difference V volts between the plates, its
capacitance is C0 (Fig. 18.6):
Q
C0 =
V
where C0 , as the name implies, is analogous to the capacity of a container. It tells

how much charge the capacitor can hold per volt of potential difference.
There are two ways of increasing the capacitance of a capacitor: (i) by making
the plates larger or (ii) by allowing the space between the plates to be filled by some
substance called a dielectric (Fig. 18.7). All substances have two poles, a positive
pole and a negative electrical pole, when they are in an electrical field. These dipoles
align themselves in opposition to the field between the plates, so they reduce the field.
The potential difference between the plates is decreased, so more charge can be put
into the capacitor to reestablish the original potential. Since the capacitor can hold
more charge, its capacitance C x with substance x between the plates is always greater
than the capacitance C0 with the apparatus pumped out to a vacuum.
Field
+
– – – Dipole
+ + + Vectors
–
FIGURE 18.7 A charged capacitor with a dielectric.
P1: OTA/XYZ P2: ABC
DIPOLE MOMENTS 295
The dielectric constant of substance x is defined as the ratio
Cx
εx =
C0
which is always greater than one. Dielectric constants are very different for different
molecules; for example, the dielectric constants are 1.000272 for H2 (g), 2.283 for
benzene, and 78.0 for water at 298 K. We describe benzene as a nonpolar solvent
and we describe water as a polar solvent.
The molar polarization of a substance P is related to the dielectric constant by
ε−1M
P=
ε+2 ρ
where M is the molar mass. This is sometimes called the total molar polarization
because it is made up of two parts: the distortion polarization and the orientation
polarization, PT = Pd + Po .
Distortion polarization exists for all substances, which is the reason that ε is never
less than one. Consider a collection of atoms of a monatomic gas such as neon or
argon. The electronic charge distribution within each atom is spherically symmetrical
in the absence of a field; but in the presence of a field, the atomic charge distribution
is distorted into something like the ellipses shown in Fig. 18.7, and a dipolar nature
is induced in it.
Orientation polarization results from the permanent dipole brought about by an
unsymmetrical charge distribution in the unperturbed molecule. A familiar example
is HCl, which has a negative end (Cl) and a positive end (H).
The mathematical form of these two distinct types of polarization enables us to
determine both. Distortion polarization is not a function of temperature,
4
Pd = π N Aα
3
but orientation polarization is inversely dependent upon T:
4π N A 2
Po = µ
9kB T
where N A is the Avogadro number and k B is the Boltzmann constant. One can
rearrange the equation for the total polarization as a function of 1/T to emphasize its
linear nature:

4π N A µ2 1 4 4π N A µ2 1
PT = + π N Aα = +α
9k B T 3 3 3k B T
P1: OTA/XYZ P2: ABC
Plotting PT , which we obtain from measured values of ε, against 1/T gives a linear
function with a slope of (4π N A /3)(µ2 /3k B ). This gives us µ and enables us to subtract
the temperature-dependent term from PT at any specific temperature to determine
4
3
π N A α and thus α. (Note that µ is not the reduced mass in this context.)
The value µ is the twisting moment on a dipole in a field. The moment is the
sum of each charge

times the lever arm separating it from the center of rotation of
the molecule i qi r1 . In the simplest case of a diatomic molecule, the charges are
equal in magnitude and µ is equal to the moment of one charge times the entire bond
length:
µ = |qr1 | + |qr2 | = q (r1 + r2 ) = qr
The polarizability α is a proportionality constant telling how much of a dipole can

be induced in a given molecule by a given field. It is inversely related to the tightness
with which a system of nuclei hangs onto its electrons and is widely used in organic
chemistry to predict the ease with which electrons can be shifted to or away from a
reaction site in the course of a reaction.
18.7.1 Bond Moments

It is useful to break molecular moments into contributions from each bond. The dipole
moment of the molecule is, to a good approximation, the vector sum of individual
moments of its bonds. We expect that chlorine will draw electrons toward it and
constitute the negative end of CH3 Cl. By the principle of bond moments, we predict
that cis-1,2-dichloroethene has a permanent dipole moment while the trans isomer
has none. The values of µ given below the isomeric structures in Fig. 18.8 are
experimentally determined.
The definition of dipole moment gives us both a convenient unit to express µ and
a way of characterizing the polarization of bonds like the H–Cl bond. Since µ = qr
and we know that the electric charge is q = 4.80 × 10−10 esu, a typical bond that is
H H H Cl
C C C C
Cl Cl Cl H
D =0
FIGURE 18.8 The total dipoles of two dichloroethene isomers.

P1: OTA/XYZ P2: ABC
NUCLEAR MAGNETIC RESONANCE (NMR) 297
1 Å in length should have

µ = qr = 4.80 × 10−10 × 1.00 × 10−8 = 4.80 × 10−18 esu cm
This is called a debye, D. In SI units of coulomb meters, 1 D = 3.338 × 10−30 C m.

We have just assumed that the negative charge is at one end of the molecule and
the positive charge is at the other, that is, that the bond is completely ionic. Molecules
are not completely ionic and many of them are not completely covalent either. The
intermediate case is a compromise called a polar molecule, which has some charge
separation but which is not totally ionic. One can obtain a numerical parameter telling
where this compromise lies by computing the dipole moment expected for a perfectly
ionic bond and comparing it with the measured moment. The ratio of the measured
moment to the moment calculated assuming complete charge separation leads to the
% ionic character of the actual molecule.
18.8 NUCLEAR MAGNETIC RESONANCE (NMR)
Some atoms, including hydrogen H, have the property of nuclear spin, which produces
a small magnetic field. In the absence of an external field, the spin magnetic fields
of nuclei are randomly oriented; but in a magnetic field, they orient themselves with
or against the external field according to their spin quantum numbers, ± 12 in the
case of the proton H. One orientation is energetically favorable, whereas the opposite
orientation is unfavorable. Twofold energy splitting occurs:
No Field Field
As usual, the lower energy level is more populous but the interaction between the
spin field and the external field is weak, so energy splitting is very small (of the order
of a few thousandths of a calorie per mole). A Boltzmann calculation for this minute
energy separation shows that the lower level has a population that is only a few nuclei
per hundred thousand greater than that of the upper level.
In principle, NMR is the same as other kinds of spectroscopy, but important tech-
nical differences exist. In absorption spectroscopy, incident electromagnetic radiation
is varied until a resonant frequency coincides with the energy-level separation nec-
essary to promote the system from a lower quantum state to a higher state. Emission
spectroscopy, which is preferred for some purposes, involves the reverse process,
emission of radiation occasioned by the fall of particles from a higher state to a lower
state. If both upper and lower energy states are appreciably populated, electromag-
netic radiation can induce both absorption and emission of energy.
P1: OTA/XYZ P2: ABC
In NMR, resonance is found by varying the magnetic field until the energy separa-
tion matches an input frequency. At a resonance frequency, protons are both absorbing
and emitting radiation. In addition to a powerful electromagnet and radiation source,
an NMR spectrograph is equipped with a receiver to detect and record the reso-
nance emission frequencies from the sample. A record of the various resonances in a
molecule (there will normally be more than one) is its NMR spectrum.
Examination of the details of molecular structure by NMR is possible because of
chemical shifts. The field under which a given proton acts is primarily the external
field, but it is slightly shielded by its surroundings. The electron density around H
(or other nuclei like 13 C) is a function of its chemical environment, particularly the
electronegativity of neighboring atoms. Thus a CH2 group absorbs and emits radiation
at a different field strength than a CH3 group. Chemical shielding is normally recorded
in relative terms to minimize differences from one experimental setup to another. One
compound, tetramethylsilane TMS, is usually selected as a reference point and other
resonances are reported relative to it in units of parts per million difference between
the resonance frequency brought about by hydrogens in the sample and those in TMS.
The chemical shift is denoted δ (ppm). Tables of the approximate chemical shift of
various groups are available.
Methyl hydrogens have a chemical shift of about 1 (ppm), CH2 hydrogens have
δ∼= 2 (ppm), and COOH hydrogens have δ ∼ = 10 (ppm). NMR spectra, like IR spec-
tra, are unique and serve as “fingerprints” of compounds. If a pure unknown has
an NMR spectrum that is identical to an authentic sample, the substances are identi-
cal. NMR is, however, more than an expensive way of carrying out qualitative analysis.
NMR spectra provide information on the internal details of the molecules examined.
Ethanol CH3 CH2 OH shows three peaks at low resolution because of the three
distinct kinds of protons CH3 , CH2 , and OH. The peaks are about equally spaced,
but they can be identified because the area under each peak is directly proportional
to the number of protons producing the peak, in the ratio 3:2:1.
18.8.1 Spin–Spin Coupling

Spin–spin coupling refers to the minute interaction between the spin field of a proton
and the spin fields of adjacent protons. Normally, a proton with one immediate
neighbor can couple with the adjacent spin in two ways: ↑↓ and ↓↑.1 A proton with
two neighbors can couple in three ways: ↑↑, ↑↓ ↓↑, and ↓↓. A proton with three
neighbors can couple in four ways. We arrive at a peak multiplicity of n + 1, where
n is the number of neighboring protons. Arrow diagrams show that the intensity of
the split peaks is in the ratio 1:1, 1:2:1, 1:3:3:1, and so on (Fig. 18.9).
Extrapolating these simple examples to more complicated molecules gives an idea
of the use of NMR in experimental studies of molecular structure. Very powerful
magnetic fields are necessary to separate proton and other nuclear peaks in compli-
cated molecular structures. Combination of NMR with the times necessary for various
1 The case of ethanol is an exception because of fast proton transfer at the OH site, yielding only one peak.
P1: OTA/XYZ P2: ABC
Intensity
FIGURE 18.9 Schematic NMR spectrum of ethanol, CH3 CH2 OH. The CH3 peak on the
right is split into a triplet because of the two neighboring protons on CH2 .
nuclei to relax to their equilibrium states after being pulsed by frequency excitation
leads to powerful diagnostic methods known as magnetic resonance imaging (MRI).
18.9 ELECTRON SPIN RESONANCE
Electrons also have the property of spin and, if spins are unpaired, as in free radi-
cals, electron spin gives rise to an electron spin resonance ESR spectrum, which is
analogous to NMR. ESR spectra are useful in the characterization of free radicals,
in their detection as reactive intermediates, and in specifying the electron probability
distribution within the radical. Free radicals have been implicated in carcinogenesis
and in aging, but other free radicals appear to act as sweepers reacting with destructive
species before they can harm a host organism.
Example 18.1 The Bond Strength and Bond Length of Carbon Monoxide
Experimental results are that a low-resolution peak can be found for 12 C16 O centered
around 2142 cm−1 , which separates at high resolution to show a peak separation of
3.8 cm−1 . Find the bond length and the force constant. The superscripted 12 and
16 indicate that these data are for the predominant isotopes of C and O. Other, far
smaller peaks can be found for the other isotopes. Because this is the experimental
information for 12 C16 O, the isotopic weights can be treated as integers.
Solution 18.1
12
1. One needs the reduced mass for both calculations. For C16 O the reduced
mass is
12.00 (16.00)
µ= 1.661 × 10−27 = 1.139 × 10−26 kg
12.00 + 16.00
P1: OTA/XYZ P2: ABC
Low Resolution High Resolution
Frequency
FIGURE 18.10 Schematic diagram of a vibration–rotation band. The low-resolution vibra-
tional band contains many rotational lines that may be discernable at high resolution. The
separation between rotational lines is given the term 2B.
2. The force constant is extracted from the classical expression for simple har-
monic motion:

1 kf
ν=
2π m
Given the experimental measurement of ν̄ = 2142 cm−1 = 2.142 × 105 m−1 ,

the frequency is

ν = cν̄ = 2.998 × 108 m s−1 , 2.142 × 105 m−1 = 6.422 × 1013 s−1
From the harmonic oscillator approximation

1 kf
ν=
2π m
we get
kf
= 4π 2 ν 2 = 1.628 × 1029
m
Substituting the reduced mass of 12 C16 O for m, we obtain

k f = µ × 4π 2 ν02 = 1.138 × 10−26 × 1.628 × 1029 = 1853 N m−1
3. The bond length is found from the moment of inertia of the molecule modeled
as a rigid rotor. The energy spacing is
2h̄ 2 h2
2B = =
2I 4π 2 I
P1: OTA/XYZ P2: ABC
The spectral input to this calculation is the line separation
2B = 3.8604 cm−1 = 386.04 m−1
so B = 193.02 m−1 . This is multiplied by c to obtain the frequency in s−1 and

bond length r in pm:
h 2.799 × 10−44 2.799 × 10−44

I = = = = 1.450 × 10−46
4π 2
(2Bc) B 193.02
1.450 × 10−46
r2 = = 1.27 × 10−20
1.138 × 10−26
r = 1.129 × 10−10 m = 112.9 pm
Example 18.2 The Dipole Moment of Sulfur Dioxide

Sulfur dioxide, SO2 , has a total molar polarization of 68.2 cm3 mol−1 at 298 K and
56.0 cm3 mol−1 at 398 K. What is the dipole moment of SO2 ?
Solution 18.2 The inverse of the lower temperature is 0.00336 and the in-
verse of the upper temperature is 0.00251. The molar polarizations are 68.2 and
56.0 cm3 mol−1 . The slope of the curve of molar polarization vs. the inverse of
T is 1.44 × 104 cm3 K mol−1 . (Probably for historical reasons, the nonstandard unit
system of erg cm is often used.) The experimental slope is related to the dipole
moment as
4π N A µ2
slope = = 1.44 × 104
9k B
Solving for µ2 , we get
9k B
µ2 = slope = 1.641 × 10−47 1.44 × 104
4π N A
It is convenient at this point to change the units of k B from J K−1 to erg K−1 . This
requires multiplication by 107 to give

µ2 = 1.641 × 10−40 1.44 × 104 = 2.363 × 10−36

µ = 2.363 × 10−36 = 1.54 × 10−18 = 1.54 D
Problem 18.1
The molecule 1 H35 Cl absorbs light at λ = 2.991 × 10−4 cm = 2.991 × 10−6 m. Tak-
ing 1 H35 Cl to be a harmonic oscillator, what are the differences in energy between
P1: OTA/XYZ P2: ABC
2100.5 cm-1 2185.5 cm-1
FIGURE 18.11 The vibration–rotation spectrum of CO. Only one vibrational absorption and
the accompanying rotational absorptions are depicted.
the levels in the parabolic Hooke’s law energy well? (The superscripted 1 and 35 are
the isotopic numbers. The mass of each isotope can be taken as an integer.)
Problem 18.2
What is the force constant for the diatomic molecule 1 H35 Cl? Take advantage of the
fact that it absorbs light at λ = 2.991 × 10−4 cm = 2.991 × 10−6 m. Assume that
the molecule is a harmonic oscillator.
Problem 18.3
It is reasonable to suppose that the force constant governing a stretch has units of
force (newtons) per unit displacement (how long the stretch is). Show that the units
of k f are Nm−1 .
Problem 18.4
In early experiments on HCl, line separation in the rotational part of its spectrum was
found to be 2B̃ = 20.794 cm−1 . What is the length of the H Cl bond?
Problem 18.5
A somewhat idealized microwave spectral band of carbon monoxide CO is shown
in Fig. 18.11. The actual band would show some anharmonic distortion. Find the
vibrational frequency. Locate the transition from the 0 to the 1 rotational states.
Calculate the force constant of the CO bond.
Comment: Notice that the speed of light (electromagnetic radiation) is used in units
of cm s−1 because the frequency is given in cm−1 . The reduced mass is in kg because
the force constant is in Nm−1 .
Problem 18.6
From the information is Problem 18.5, estimate the CO bond length.
P1: OTA/XYZ P2: ABC
Problem 18.7
The rotational partition function at 298 can be written as an integral:
qrot (T ) = 2 (J + 1) e−
rot J (J +1)/T dJ
where
rot is a parameter called the rotational temperature, Show that qrot = T /
rot .
Problem 18.8
The fraction of molecules in a rotational state is given by the ratio of the partition
function for that state relative to the total partition function. But we know that
qrot = T /
rot from the previous problem, so
QJ 2 (J + 1) e−
rot J (J +1)/T dJ 2 (J + 1)
rot e−
rot J (J +1)/T dJ
= =
Q total T /
rot T
Plot Q J /Q total as a function of J for NO which has

rot = 2.34 at T = 298 K.
Problem 18.9
The dipole moment µ is a twisting moment in an electrical field between particles of
charge q separated by a distance r:
µ = qr
The unit of q r is coulomb meters. The unit charge in atomic problems is the charge
on a proton, 1.6019 × 10−19 C. The Br–F bond length is 176 pm. What would the
dipole moment for the molecule BrF be if it were completely ionic Br+ F− ?
Problem 18.10
If the actual dipole moment of Br–F is only 1.42 D, what is the percent ionic character
of the bond? Is the bond predominantly ionic or covalent?
Problem 18.11
The molar polarization of bromoethane was measured at five different temperatures
with the following results:
T 205 225 245 265 285

PT 104.2 99.1 94.5 90.5 80.8
What are the polarizability and dipole moment of bromoethane?

P1: OTA/XYZ P2: ABC
Problem 18.12
Some polyenes, . . . C C C C C C C . . . , and so on, are colored (lycopene in
tomatoes, for example). Use the particle in a one-dimensional box to predict how
long the conjugated chain must be for its absorption frequency to be in the visible
region.
Problem 18.13
A reaction product is thought to be either methyl acetate or ethyl formate. NMR
analysis produced the spectrum shown below. Which is it?
6 4 2 TMS
P1: OTA/XYZ P2: ABC
19
CLASSICAL MOLECULAR MODELING
Although the chemical bond is a result of quantum mechanical interactions of elec-

trons with nuclei and with each other, the atom itself is large enough to be treated
classically. One way of understanding and predicting chemical properties is by treat-
ing the molecule as an aggregation of atoms bound by classical bonds and interacting
with each other in a classical way. Such a study is called molecular mechanics, MM.
19.1 ENTHALPY: ADDITIVE METHODS
It has long been known that extension of an alkane chain by one CH2 group
CH3 (CH2 )n CH3 → CH3 (CH2 )n+1 CH3
brings about a decrease in the f H 298 of about 5 kcal mol−1 = 21 kJ mol−1

(Fig. 19.1). For example, if we subtract 21 kJ mol−1 from f H 298 (ethane) =
−84 kJ mol−1 , we get (correctly) f H 298 (propane) = −105 kJ mol−1 . Thus we can
call –21 kJ mol−1 a CH2 group enthalpy for the extension of ethane to propane. The
group enthalpy is transferable to give the enthalpy change for extension of other
alkanes to an alkane larger by one CH2 group.
If this works for CH2 , why not try it with other alkyl groups? The enthalpy of
the CH3 group must be one-half the enthalpy of formation of the ethane molecule
CH3 CH3 , f H 298 (ethane) = –84 kJ mol−1 . Hence we take the CH3 group enthalpy

Copyright
305
P1: OTA/XYZ P2: ABC
306 CLASSICAL MOLECULAR MODELING
Elements
-21 kJ mol -1
-21 kJ mol -1
-21 kJ mol -1
FIGURE 19.1 Enthalpies of formation of “adjacent” n-alkanes.
as –42 kJ mol−1 . Provided that these enthalpies are transferable, the enthalpy of
formation of n-pentane CH3 (CH2 )3 CH3 , for example, is estimated as 2(−42) +
3(−21) = 147 kJ mol−1 . The calculation agrees with the experimental value:
f H 298 (CH3 (CH2 )n CH3 ) = 147 kJ mol−1
to a precision of ±1 kJ mol−1 . Although the agreement is not always this nice, the
procedure is very encouraging.
What of groups other than alkyl groups? Can this group additivity strategy
be extended to include the CH group and the C atom in branched alkanes like
2-methylbutane (isopentane) and 2,2-dimethylpropane (neopentane)? The answer is
yes, and this strategy has been extended to cover many functional groups in organic
molecules including oxygenated, nitrogenous, and halogenated hydrocarbons (Cohen
and Benson, 1993).
19.2 BOND ENTHALPIES
Molecular enthalpy can be segregated in other ways. One familiar way involves
associating an enthalpy with each bond in the molecule. For example, inserting a
CH2 group into an n-alkane requires breaking the backbone of the alkane molecule
at a cost of one C C bond followed by forming two C C bonds and increasing the
number of C H bonds in the molecule by two:
R CH2 CH2 R → R CH2 CH2 CH2 R
The sum total of the process is a gain of one C C bond and two C H bonds. If we
associate –348 kJ mol−1 with the C C bond and –413 kJ mol−1 with the C H bond
relative to the isolated atoms C and H, the insertion entails a change in total bond
enthalpies of
C C + 2(C H) = −348 + 2(−413) = −1174 kJ mol−1
Why such a big number? This big number arises because theoretical folks find
it convenient to use a reference state that is different from the thermodynamicist’s
P1: OTA/XYZ P2: ABC
STRUCTURE 307
Gaseous Atoms
-1153
-1174
Elements
-21
Compounds
FIGURE 19.2 Bond enthalpies calculated in CH2 , from the reference state of gaseous atoms
(top), and relative to elements in their standard state (H2 (g) and C(graphite)). Not to scale.
standard state of the elements in their most stable form at 298 K. To translate from one
reference state to the other, we need to know their enthalpy differences. These values
have been measured with care (and some difficulty). They are atomization H 298 (H2 ) =
436 kJ mol−1 and atomization H 298 (Cgr ) = 717 kJ mol−1 . These larger enthalpies must
be subtracted from the bond enthalpies relative to the atoms to find – 1174 – (–1153)
= –21 kJ mol−1 , the sum of bond enthalpies in CH2 relative to the thermodynamic
standard state of H2 (g) and C(graphite) (Fig. 19.2).
This result, relative to a new standard state, is the same as the one used with
success in Section 19.1. The rationale for what seems to be a rather exotic standard
state will be clearer in the next chapter. It is also important to remember that without
correction factors for molecular deformations of the real molecule relative to the
additive model we have postulated here, either summing appropriate bond enthalpies
or summing simple group enthalpies gives a strainless molecule.
19.3 STRUCTURE
Given that the hydrogen atoms in methane repel one another, the only reasonable
structure we can assign to methane is that of a tetrahedron in which four bound H
atoms achieve maximum separation in 3-space about the central C atom:
Using this simple structural symmetry and an arbitrary bond length (the length
of a wooden peg, perhaps), we can construct a unique “ball-and-stick” model of
methane and, by extension, very many models of alkanes (Fig. 19.3). By the nature
P1: OTA/XYZ P2: ABC
C C C C
C
FIGURE 19.3 Structurally distinct alkane conformers resulting from the tetrahedral sym-
metry of carbon.
of tetrahedral symmetry about all C atoms, we soon find that our models include
numerous conformers of the higher members of the alkane series.
We have the force constants k f of the C H bond from spectroscopic studies
(Sections 18.1 and 18.2). Using these force constants, we can relieve the arbitrary
nature of the C H bond lengths in CH4 by starting at some reasonable value for the
bond length r and calculating the energy of the CH4 molecule. The result will be, no
doubt, very high. Making systematic small changes in r (Problem 18.4) and repeating
the process, perhaps many times, the calculated energy can be brought to a minimum
value. At the optimum bond length, each H atom is as close as it can get to the bottom
of its Hooke’s law potential well while still respecting its neighbors. In optimizing E,
we have also optimized r. This is the best value of the C H bond length in methane
we can get from the Hooke’s law force parameter k f we have chosen.
This raises the question of the best force parameter we can choose. Will it be
exactly the spectroscopic constant? What if there is no spectroscopic constant? Ex-
perience has shown that all things considered, the best MM force parameters express
knowledge derived by fitting bond lengths and stretching energies to spectroscopy,
thermodynamics, X-ray crystallography, and that old favorite “chemical intuition.”1
MM parameters have evolved away from spectroscopic force constants; the two are
not the same. MM parameters can be proposed, and then they can be revised in the
light of new experimental results. Spectroscopic force constants are determined by
a specific category of experiments. They can be made more accurate, but they do
not change.
We now have the tetrahedral bond angles and the bond lengths of CH4 . From this
we can express the structure of CH4 in Cartesian coordinate locations of all of the
atoms in the molecule. There is nothing more we can know about the geometry of CH4 .
The method can be elaborated to determine the complete MM structure of any
molecule, provided that we have the force parameters for all of the structural forces
in the molecule.
There is the rub. It turns out that there are many subtle forces operative in molecules
larger than CH4 . It is the task of anyone who wishes to find the structure of a complex
molecule to seek out all of the forces within it and to determine the force parameters
of the stretch, bend, torsional deformations, van der Waals, or other interactions. One
is not advised to attempt this daunting task. Fortunately, N. L. Allinger and coworkers
1 “Chemicalintuition” is not as silly as it sounds. Like the scriptures, it is the accumulated knowledge of
very many years of observation and experience.
P1: OTA/XYZ P2: ABC
GEOMETRY AND ENTHALPY: MOLECULAR MECHANICS 309
(Nevins et al. 1996 a, 1996 b) and other groups have devoted four decades to the job,
resulting in rather large MM computer programs that have evolved with time. Among
them is the Allinger series MM1, MM2, . . . , MM4, with some variants in between.
19.4 GEOMETRY AND ENTHALPY: MOLECULAR MECHANICS
Suppose that we have suitable force parameters for the C H and the C C bonds in
alkanes and r0 at the minima of the Hooke’s law potential energy wells (Fig. 18.2).
We should be able to construct models of ethane and higher alkanes and find the
structure and energy that best satisfy the k f values.
But not quite. Like the members of a large family, the bond–atom combinations in
a molecule do not always agree. The final structure of the molecule is a compromise
in which no bond is exactly at its ideal length r0 but no bond is too far removed from
it. The distance between the idealized bond length r0 and the real one r is assumed
to follow Hooke’s law, F = −k f (r − r0 ). The restoring force F results in a parabolic
potential energy well with the potential energy defined as V = 0 at its minimum:
V (r − r0 ) = k f 12 (r − r0 )2
Like a large family, the final compromise engenders some strain. The energy of
the real molecule is above the energy of a hypothetical strainless molecule by the
summation of the strain energy experienced by all its members at the compromise
position. The energy of the real molecule is at least

E = E strain-free + V (r − r0 )
all bonds
Bond stretching is not the only energy that enters into the final compromise.
Chemical bonds can also be bent through an angle θ . It is reasonable to assign a
Hooke’s law potential energy for bending. If the compromise structure involves some
bond bending, the potential energy of bending must be included in the sum

E = E strain-free + Vstretch (r − r0 ) + Vbend (θ − θ0 )
all bonds all simple angles
Intramolecular interactions result in forces called van der Waals forces, and molecules
may suffer dihedral angular distortion called torsional forces. These give rise to
potential energies VvdW and Vtor , respectively. The sum of all potential energies over
a hypothetical strainless enthalpy is

E = E strain-free + Vstretch + Vbend + VvdW + Vtors
P1: OTA/XYZ P2: ABC
Intramolecular interactions are not independent of one another. For example, bending
of a stretched bond is not the same as bending of a bond at its equilibrium length
r0 . Conversely, stretching of a bent bond is not the same as stretching of a bond at
θ = θ0 . A stretch-bend term Vs-b , called a cross term, is added. Other terms, especially
electrostatic terms, have been added (somewhat reluctantly) as necessary to achieve
closer and closer agreement with known experimental data. The collection of terms
and the parameters that produce them is called a force field. The MM force field has
become larger as new experimental evidence has been added to the basis set and as
a larger set of molecular problems has been attempted. Allinger et al. (1996, 2010)
have augmented the strainless energy by an 11-term equation in his programs MM3
and MM4.
19.5 MOLECULAR MODELING
The term molecular model includes ball-and-stick physical models used as visual
aids to facilitate thinking about the invisible world of the molecule. In contemporary
usage, however, the term usually denotes a mathematical model, possibly generating
a computer output set of Cartesian triples, each of which specifies the position of an
atom in the model.
The most fundamental molecular property obtained from a molecular model is its
geometry. Geometry is found by optimization of an estimated input geometry. One
of many possible geometry optimization procedures is shown in Example 19.1.
19.6 THE GUI
Input files for more complicated molecules can be constructed using a point-and-click
graphical user interface (GUI)—for example, PCModel C . Having the optimized
model in mathematical form permits use of the coordinate set to draw two- or
FIGURE 19.4 Visualization of the output for the ethane molecule (PCModel 8.0
C ).
P1: OTA/XYZ P2: ABC
FINDING THERMODYNAMIC PROPERTIES 311
three-dimensional diagrams (visualization) on the computer screen to appeal to the

pictorial thinking favored by many chemists. Indeed, the drawing may be made to
mimic the familiar wooden ball-and-stick models, with the difference that we expect
it to be more accurate.
19.7 FINDING THERMODYNAMIC PROPERTIES
Suppose that, with some labor, we have arrived at a collection of force parameters
sufficient to describe a real molecule at its optimized geometry. Suppose further that
we have arrived at a set of bond enthalpies that enables us to write idealized “heats”
of formation as (a) a sum of bond energies SBE derived from a set of molecules
taken to be strainless and (b) another sum of bond energies NBE derived from a set
of molecules taken to be normal. Some of the resulting information available from
a computational run using MM4 for ethane is shown in File 19.1. The difference
HEAT OF FORMATION AND STRAIN ENERGY CALCULATIONS
(UNIT = KCAL/MOLE)
( # = TRIPLE BOND)
NORMAL (BE) AND STRAINLESS (SBE) ENTHALPY OF INCREMENTS

(CONSTANTS AND SUMS OF INCREMENTS)
BOND OR STRUCTURE NO ----NORMAL---- --STRAINLESS--

C-C SP3-SP3 1 -89.2005 -89.2005 -84.0718 -84.0718
C-H ALIPHATIC 6 -105.7262 -634.3572 -98.4375 -590.6250
C-METHYL (ALKANE) 2 -0.0964 -0.1928 -0.5379 -1.0758
-----------------------------------------------------------------------
NBE = -723.7505 SBE = -675.7726
HEAT OF FORMATION OF ELEMENTS (HATOM) = -655.9401 KCAL/MOLE

IN STANDARD STATE FROM ATOMS
MOLAR HEAT CONTENT OF COMPOUND (MH) = 48.6328 KCAL/MOLE

(STERIC + ZPE + THERMAL ENERGIES)
PARTITION FUNCTION CONTRIBUTION (PFC)

CONFORMATIONAL POPULATION INCREMENT (POP) 0.00 (ASSUMED VALUE)
TORSIONAL CONTRIBUTION (TOR) 0.00 (ASSUMED VALUE)
----------------
PFC = 0.00 (ASSUMED VALUE)
HEAT OF FORMATION (HFN) AT 298.2 K = -19.18 KCAL/MOLE

(HFN = MH + NBE + PFC - HATOM)
STRAINLESS HEAT OF FORMATION FOR SIGMA SYSTEM (HFS) = -19.83

(HFS = SBE - HATOM)
INHERENT SIGMA STRAIN (SI) = HFN - HFS 0.65
SIGMA STRAIN ENERGY (S) = POP + TOR + SI 0.65
FILE 19.1 Partial MM4 enthalpy output for ethane. (Units are kcal/mol.)
P1: OTA/XYZ P2: ABC
between SBE and HATOM [the enthalpy input to drive hydrogen and graphite into
the atomic state (Section 19.2) for the strainless molecule] is
f H 298 (strainless) = −675.77 − (−655.94) = −19.83 kcal/mol
A similar sum of normal bond enthalpies gives a different result for the normal
molecule:
NBE + HATOM = f H 298 (normal) + MH = − 67.81 + 48.63 = − 19.18 kcal/mol.
The new term MH (molar heat content) arises from a slightly different way of defining
the normal bond energies; they should be represented by an arrow going all the way to
the bottom of the energy wells in the real molecule. Positive enthalpies—consisting
primarily of the zero point energies of vibration (Section 18.2), with a contribution
from the steric energy—and small statistical factors add up to a positive correction
of 48.63 kcal/mol. The difference between HFS and HFN is an accumulated strain
energy:
INHERENT SIGMA STRAIN (SI) = HFN - HFS 0.65
The partial MM4 output for ethane (File 19.1) illustrates some other features of the
MM procedure and the full file provides more. See STATISTICAL THERMODY-
NAMICS ANALYSIS from the MM4 output and Nevins et al. (1996 b, pp. 703–707)
for more detail.
19.8 THE OUTSIDE WORLD
In any molecular modeling enterprise, we seek to compare physical properties pre-

dicted by the model with the same properties measured by experiment. One obvious
choice is bond lengths as measured by X-ray crystallography. Agreement is usually
good but not perfect. Even assuming that the impossible task of removing all defects
from the model has been achieved, the X-ray crystallographer and the MM modeler
are not really looking at the same thing. The idea of a model as developed so far
implies an isolated gas-phase species, but molecules in crystals are subjected to pow-
erful forces holding them in the crystalline state due to the sum of forces exerted by
their neighbors.
Somewhat surprisingly, these forces do not influence bond lengths very much.
Agreement between X-ray bond lengths and MM4 lengths is usually within about
0.002 Å. Simple bond angles, being weaker, are distorted by a few degrees, and
torsional angles, being weaker still, are distorted more. X-rays are scattered by
electrons, so crystallographic structures represent the centers of electron densities
which tend to be the same as the location of the nuclei, but which may be displaced
in some chemical bonds. For example, the difference between the C H bond length
P1: OTA/XYZ P2: ABC
TRANSITION STATES 313
determined by X-ray scattering and that determined by neutron scattering (which

yields the nuclear locations) is more than 0.1 Å in some cases (Allinger, 2010).
Geometric parameters from different sources should not be used indiscriminately.
Motion of the entire molecule in the crystalline lattice is called rigid-body motion.
The amplitude of rigid-body motion is not spherically symmetrical, so the position
of the molecule is not represented as a sphere but as an ellipsoid.
As molecular spectroscopy has contributed force constants to molecular mechan-
ics, so MM contributes vibrational frequencies to spectroscopy. Because force param-
eters are not the same as force constants, MM values of bond stretching frequencies
are not the same as known experimental values but they generally agree to within
±25 cm−1 . Knowing the geometry and atomic masses, angular momenta can also
be calculated along with spectral intensities, heat capacities, entropies, Gibbs free
energy functions, and temperature variations of the thermodynamic functions.
19.9 TRANSITION STATES
Existence of several potential energy minima, one for each of the atoms in a more or
less complicated molecule, implies a potential surface analogous to a mountainous
terrain with peaks, valleys, and mountain passes. Each pass represents a potential
path from one minimum to another. In chemistry, these paths are said to be along
a possible reaction coordinate. A transition from one minimum potential energy to
another goes over a relative maximum called the transition state. A transition state
is a maximum relative to the two minima it connects but, analogous to a mountain
pass in high country, it is a least energy path between the minima, Mathematically,
all second derivatives of the energy are positive except one, leading to vibrational
frequencies all of which are real except for one which is an imaginary number.2
An example is conversion of the “chair” form of cyclohexane to the “boat” form:
For this change to take place, cyclohexane must go through a planar conformation
that is higher in energy than either the chair or the boat. The planar conformation is
the transition state. One way (not necessarily the best way) of finding the transition
state energy is by starting with a planar input geometry along with the chair and boat.
The MM4 difference between the chair and boat forms is 5.7 kcal mol−1 =
23.8 kJ mol−1 at 226 K, and the experimental value is 5.5 kcal mol−1 = 23.0 kJ mol−1 .
The transition enthalpy from chair to boat is a little less than twice this amount and
involves somewhat more complicated geometries like the twist boat conformation
(Allinger et al., 1996, p. 650 ff).
2 Pleasedo not think that there is something wrong with it: The term “imaginary” is just mathematical
jargon meaning that if you square it, you get a negative number.
P1: OTA/XYZ P2: ABC
Example 19.1 A Geometric Optimization of Water

Construct an input file for the MM4 geometry optimization of water.
Solution 19.1 First and most obvious, we name the molecule and enter the number
of atoms. In MM4, the name goes on line 1 and can occupy up to 60 columns (spaces).
In this case, we enter water in columns 1 to 5 of the first line in File 19.1 and specify
the number of atoms as 3 in column 65. On the second line or “card,” the number
0 in column 5 shows that there are no connected atoms (see Example 19.3). The 2
in column 30 gives the number of attached atoms, namely the two hydrogens. Line
3 identifies he attached atoms. Atom 2 (oxygen) is attached to atom 1 (one of the
hydrogens), and atom 2 (oxygen) is also attached to atom 3 (the other hydrogen)
Next, we input a starting geometry in Cartesian coordinates for all the atoms in the
molecule. The geometry chosen is a simple angle > with oxygen at the origin. The
position vectors of the two hydrogens in an arbitrary Cartesian coordinate space are
(−0.5, 0.5, 0.0) and (−0.5, −0.5, 0.0). The input geometry has been constrained to
the x, y plane by setting the z components to zero. The Cartesian coordinates are given
in the order x, y, and z from left to right. The O atom has an identifying number 6,
and the two hydrogens are given the number 21 in this MM4 convention. Numerical
identifiers are used to represent atoms, which may be of different kinds—for example,
sp, sp 2 , or sp 3 carbon atoms. These two parts, the descriptive cards and the Cartesian
geometry, constitute the input file. Save your input files as h20.mm4 or some similar
name with the suffix .mm4. It is to be emphasized that this is only one of very many
possible input files.
Example 19.2 A Typical Small Molecule Run

Run the MM4 input file, File 19.2. The run is initiated by typing mm4 followed by
the input filename h20.mm4. Enter 0 for parameters, 1 for the line number,
and 2 for the program choice. Several alternatives are offered. Running them
later should make most of the differences self-evident.
Solution 19.2 A successful run should give a small steric energy of <1 kcal mol−1 .
Hit enter three times to exit MM4 and to save TAPE4.MM4 and TAPE9.MM4. Enter
water 3
0 2 0 0 0 0 0 0 0
2 1 2 3
-0.50000 0.50000 0.00000 21
0.00000 0.00000 0.00000 6
-0.50000 -0.50000 0.00000 21
FILE 19.2 An input file for water. The geometry is an arbitrarily chosen estimate. The zeros
in line 2 and in the Cartesian coordinates are only place indicators to make counting the
columns easier in more complicated files (see File 19.4).
P1: OTA/XYZ P2: ABC
water 0 3 0 0 0 0 10.0
0 0 0.0000000 2 0 0 0 0 0 0 1 0
2 1 2 3
-0.76769 0.52221 0.00000 H 21( 1)
0.00000 -0.06580 0.00000 O 6( 2)
0.76769 0.52221 0.00000 H 21( 3)
FILE 19.3 The MM4 geometry output TAPE9.MM4 for water.
cat TAPE9.MM4 to display the optimized geometry of your input file. File 19.3
shows an MM4 output for water. The output File 19.3 has the same format as the
input File 19.2, but it contains some more information beyond column 65 which
we do not need at this point. These differences notwithstanding, the geometry is
clear. The simple angle > has been up-ended to a V and moved down 0.06580 Å
during the optimization. By a Pythagorean calculation, we get an O H bond length of
0.967 Å. This length is confirmed by a calculation within the program. We find, along
with much other information in TAPE4.MM4,
H( 1)- O( 2) 0.9670
O( 2)- H( 3) 0.9670
A search toward the end of the output file will reveal a computed dipole moment
(Section 18.7):
DIPOLE MOMENT = 2.0188 D
The experimental value is 1.85 D.
Example 19.3 An MM4 Calculation of the Geometry and f H298 of Methane

Convert the H2 O input file to a methane input file and run it in the MM4 protocol.
The atomic identifiers are 1 for an sp 3 carbon and 5 for an alkane hydrogen. A 1 in
column 65 of line 2 requests f H 298 .
Solution 19.3 Assuming considerable ignorance about the structure of methane,

let us place the carbon atom at the origin and suppose that the projection of each
atom upon the Cartesian coordinates centered on the carbon is 1.0 Å. For conve-
nience, it is useful to orient the molecule with two C H bonds in the x, y plane.
Examination of the TAPE4.MM4 file circumvents awkward trigonometric hand cal-
culations within the tetrahedral output geometry and gives the C H bond length
as C(1)- H(2) 1.1070 expressed in Angstroms. The experimental value is
109.3 pm = 1.093 Å. The tetrahedral H C H bond angles are H(2)- C(1)-
H(3) 109.471 expressed in degrees. The experimental value is 109.5◦ . A very
important line in TAPE4.MM4 is the f H 298 “heat” (enthalpy) of formation that was
requested in entry 2, 65 of the input file:
HEAT OF FORMATION (HFN) AT 298.2 K = -17.89 KCAL/MOLE

P1: OTA/XYZ P2: ABC
Methane MM4 5
0 4 0 0 0 0 0 0 1
1 2 1 3 1 4 1 5
0.0 0.0 0.0 1
-1.0 1.0 0.0 5
-1.0 -1.0 0.0 5
1.0 0.0 1.0 5
1.0 0.0 -1.0 5
FILE 19.4 MM4 input geometry for methane. The 1 in column 65 requests calculation of
f H 298 .
The experimental value is f H 298 = −74.8±0.3 kJ mol−1 = −17.9±0.1 kcal mol−1 .

Molecular properties beyond geometry and heat capacity, such as dipole moments,
spectra, partition functions, and entropy, are calculated in MM4.
Problem 19.1
(a) What is the enthalpy of formation f H 298 of neopentane (dimethylpropane)
by the simple group additivity method in Section 19.1?
(b) The enthalpy of isomerization of isopentane (methylbutane) to neopentane is
–14 kJ mol−1 . What is the enthalpy contribution of the C H group in alkanes?
Problem 19.2
Determine the O H bond length from the Cartesian coordinates given in File 19.3 by
hand calculation using Pythagoras’s theorem. The experimental value is 97.0 pm =
0.970 Å.
Problem 19.3
Find the H O H bond angle from the optimized geometry in File 19.3. The experi-
mental value is 104.5◦ .
Methane MM4 0 5 0 0 0 0 10.0

0 0 0.0000000 4 0 0 0 0 0 1 1 0
1 2 1 3 1 4 1 5
0.00000 0.00000 0.00000 C 1( 1)
-0.60766 0.92497 -0.02506 H 5( 2)
-0.66877 -0.88172 -0.02769 H 5( 3)
0.60089 -0.02167 0.92947 H 5( 4)
0.67554 -0.02158 -0.87672 H 5( 5)
FILE 19.5 MM4 output geometry for methane.

P1: OTA/XYZ P2: ABC
Problem 19.4
What is the dipole moment of water according to an MM4 calculation? Compare this
value with the value given in a current textbook.
Problem 19.5
Expand the input file for water, File 19.2, which has the geometry > to an input file
for ethene, which has the geometry >=<. The atom designator for an sp 2 carbon is 2
and the hydrogens are designated 112. Run the resulting program. The steric energy
for a successful run should be approximately 2 kcal mol−1 .
Problem 19.6
Modify either the input or the output files resulting from Problem 19.5 after the
manner of Example 19.3 to obtain an input file for ethane. Run the file and obtain the
full output file TAPE9.MM4 and TAPE4.MM4, part of which is given as File 19.1.
Problem 19.7
Be sure that there is a 1 in column 65 of both the ethene and ethane programs obtained
from Problems 19.5 amd 19.6 to produce values for the enthalpy of formation of
each. Determine the enthalpy of hydrogenation of ethene to ethane, each in the
standard state. The result from a successful geometry optimization should be about
–32 kcal mol−1 . The experimental value is −32.60 ± 0.05 kcal mol−1 = −136.4 ±
0.2 kJ mol−1 .
Problem 19.8
In the previous problem, why don’t you need to compute the standard state enthalpy
of formation of H2 ?
C2 H4 (g) + H2 (g) → C6 H6 (g)
Problem 19.9
If you have access to a GUI, find the enthalpies of formation of mono- di- tri- and
tetra(t-butyl)methane. Explain the curious pattern of these results. Print the structure
of tetra(t-butyl)methane.
Problem 19.10
There are three vibrational frequencies given for water in the output file TAPE4.MM4.
Two are for the symmetric and antisymmetric O H stretch. Diagram these two modes
of motion. What molecular motion corresponds to the third frequency? Why it this
frequency so different from the other two?
P1: OTA/XYZ P2: ABC
20
QUANTUM MOLECULAR MODELING
With Slater determinants and the Hartree–Fock equations, one had everything neces-
sary to solve problems in molecular structure, energy, and dynamics. But the equa-
tions could not be solved. Quantum chemistry awaited two spectacular advances.
First, Roothaan derived a way to express its apparently insoluble integrodifferential
equations as equations in linear algebra. This was accompanied by an exponential
rise in power of the digital computer, which now routinely carries out trillions of
simple mathematical operations per second.
The total energy of a molecule is almost entirely that of its constituent atoms,
leaving only a fringe energy, which might otherwise be considered trivial, to hold
the molecule together. But it is the chemical bond that dominates the world we see
around us with all its color, life, and diversity, and so it is the chemical bond that
we seek to calculate. We will be able to extend the atomic orbital concept to our
molecular calculations but, because chemical energies and energy differences are so
small, we must achieve a very high level of accuracy. Accurate programs now exist
that can be extended to all molecules (in principle at least). Finite computer speed
and memory place strict limitations on these grandiose plans, but barriers fall almost
every day in this active research field.
20.1 THE MOLECULAR VARIATIONAL METHOD

The variational method applies to molecules as well as to atoms. By it, we can
approach an optimized molecular energy E:
| Ĥ |
E=
|
Copyright
318
P1: OTA/XYZ P2: ABC
THE HYDROGEN MOLECULE ION 319
e–
rA rB
+ R
+
FIGURE 20.1 The hydrogen molecule ion, H+
2.
where | is the state vector for the entire molecule. The inner product |
is 1, so
E = | Ĥ |
In chemical applications the state vector is usually written as the equivalent wave
function:

E= Ĥ dτ
If the state vector or wave function is exact, the energy will be exact.
20.2 THE HYDROGEN MOLECULE ION
A stepping stone toward full-scale molecular structure and energy calculations is the
hydrogen molecule ion H+ 2 . Conversion of the problem to elliptic coordinates and
subsequent solution by numerical methods has been carried out, so we know the
answer before we start. The energy of the bound state of H+ −1
2 is about 268 kJ mol .
Unfortunately, this numerical method cannot be extended to larger molecules or ions,
so we shall use the known result to help us to develop of an approximate method,
which can then be applied to molecular systems large enough to be important in
chemistry.
The geometry of the H+ 2 problem is given in Fig. 20.1. Assume that the nuclei are
stationary1 at a distance R from one another. This gives us a problem of one electron
in the field of positive nuclei A and B over distances r A and rB . The Schrödinger
equation is similar to that of the hydrogen atom, except that there are two centers of
positive charge rather than just one. For any selected value of R,

1 1
−2∇ −
1 2
− = E electronic
rA rB
1 This is the Born–Oppenheimer approximation.

P1: OTA/XYZ P2: ABC
320 QUANTUM MOLECULAR MODELING
There are many possible values of R, each of which leads to a unique value of the
electronic energy E electronic . Although R is constant for any single calculation, the
total energy of the system is a function of R:
1
E total = E electronic +
R
This means that a curve of E total vs. R can be drawn. From this point on, we shall
drop the subscripts on E, taking the nature of E, electronic or total, to be clear from
context.
By the LCAO approximation, a molecular orbital can be expressed as a Linear
Combination of two hydrogenic Atomic Orbitals:
ψ = N1 e−r A ± N2 e−r B
The atomic orbitals are basis functions which define a vector space that includes the
molecular orbital. The LCAO basis set is not complete, so the molecular orbital we
obtain will not be correct. If functions N1 e−r A and N2 e−r B are normalized hydrogen
1s orbitals, we call them 1s A and 1s B and the approximate molecular orbital for H+ 2
is ψH+2 = 1sA ± 1sB . Physically, one basis function represents hydrogen atom A at
some distance from proton B, while the other basis function represents hydrogen
atom B at some distance from proton A. Neither basis function alone recognizes the
simultaneous interaction of the electron with both nuclei. It is this interaction we seek
because it brings about the chemical bond.
The energy of the positive combination ψH+2 = 1sA + 1sB is

1 1
E= ψ Ĥ ψ dτ = N 2
(1s A + 1s B ) − 12 ∇ 2 − − (1s A + 1s B ) dτ
rA rB
where N 2 is the product of the two normalization constants and the integration is
taken over the entire space τ . If we expand and simplify according to the symmetry
of the problem, the energy gives three functions often denoted J, K, and S:

1
J = 1+ e−2R
R

S
K = − 1 + R e−2R
R
and

R 2 −R
S = 1+ R+ e
3
P1: OTA/XYZ P2: ABC
THE HYDROGEN MOLECULE ION 321
0.2
EB(R)
EA(R)
0
0 5
R
FIGURE 20.2 Bonding and antibonding orbitals for H+
2 . The units of E are hartrees and the
units of R are bohr = 52.9 pm.
which are combined to give
J+K
E=
1+S
The functions J, K, and S are all dependent upon the internuclear separation R because
orbital interaction is greater when the nuclei are close together and smaller when they
are far apart. The function E = f (R) for the positive combination of atomic orbitals
is shown by the solid curve in Fig. 20.2.
Positive nuclei attracted to an electron between them are bound to each other. The
nuclei do not crash into each other because at some R, internuclear repulsion becomes
dominant over bonding and the energy begins to rise sharply. The compromise be-
tween attractive and repulsive forces results in a minimum in E = f (R). The bond
has a definite and fixed length of 2.5 bohr = 132 pm in this first approximation. To
find the energy of the H+ 2 bond at this distance, we must solve equations for S, J, and
K to give
S = 0.461, J = 0.00963, K = −0.1044, E = −0.065 E h
The hartree of energy is 1 E h = 2625 kJ mol−1 . The minimum is at E =

−0.065 E h = −171 kJ mol−1 . This is about 64% of the experimental value. The
quantitative result is not very good, but the qualitative result shows that the energy is
negative at 132 pm, which indicates bonding. It is noteworthy that the chemical bond
arises as a natural consequence of quantum mechanics without further assumptions.
There is a second solution for E in Fig. 20.2 which arises from the negative
combination ψH+2 = 1sA − 1sB . The energy does not go through a minimum with
this antibonding wave function but rises monotonically with decreasing R. In H+ 2,
bonding and antibonding orbitals are above and below the energy of the separated
system H+ and H. The higher and lower bonding and antibonding molecular orbitals
at the hond distance are often shown simply as Fig. 20.3.
P1: OTA/XYZ P2: ABC
antibonding
Atoms
bonding
FIGURE 20.3 Bonding and antibonding solutions for the H+
2 . One electron in the lower
(bonding) orbital of H+
2 is indicated by an arrow.
20.3 HIGHER MOLECULAR ORBITAL CALCULATIONS
In obtaining an approximate solution to the molecular Schrödinger equation, the

many-electron wave function (ri ), which is a function of all radial distance vectors
ri , is broken up into orbitals ψi :
(ri ) = (n!)−1/2 det[(ψ1 α) (ψ1 β) (ψ2 α) . . .]
The orbitals ψ1 α, ψ1 β, ψ2 α, . . . accommodate single electrons. The symbols α and

β designate opposite spins and “det” indicates a Slater determinant. Because of
spin pairing, each orbital can contain two electrons; hence the minimum number of
molecular orbitals ψ1 , ψ2 , ψ3 , . . . is one-half the number of electrons.
In 1951 Roothaan further divided single-electron molecular orbitals ψi into linear
combinations of basis functions χµ :

N
ψi = cµi χµ
µ=1
(µ = 1, 2, 3, . . . N ), where N > n. Having selected a basis set χµ , one wishes to find

the coefficients cµi . A large cµi means that the corresponding basis vector makes an
important contribution to the total molecular orbital, while a small cµi means that the
corresponding basis vector makes a small contribution. This gives a set of algebraic
equations in place of the set of coupled differential equations in the original problem.
Roothaan’s equations can be written in matrix form as
FC = SCE
where C is the column vector of coefficients, E is the diagonal matrix of energies,

with elements E i j = εi δi j , the elements of S are Sµν = χµ χν dτ , and F is the Fock
matrix.2
Elements in the F matrix are

Fµν = Hµν + Pλσ [ (µν|λσ ) − (µλ|νσ ) /2]
λσ
2 It
is unfortunate that F is used to represent both the force field matrix in molecular mechanics and the
Fock matrix in quantum mechanics. Be careful not to confuse the two.
P1: OTA/XYZ P2: ABC
SEMIEMPIRICAL METHODS 323
defined by

n
Hµν = χµ Ĥ χν dτ, Pµν = 2 cµi cνi
1
and

1
(µν|λσ ) = χµ (1)χν (1) ×
χλ (2)χσ (2) dτ1 dτ2
r12

1
(µλ|νσ ) = χµ (1)χλ (1) × χν (2)χσ (2) dτ1 dτ2
r12
The matrix elements Hµν are elements of the core Hamiltonian that would be imposed
by the nuclei on each electron in the absence of all other electrons, and elements εi
of the diagonal matrix E are one-electron energies. Many computer routines exist for
multiplication, inversion, and diagonalization of matrices.
The integrals (µν|λσ ) and (µλ|νσ ) are difficult to evaluate, which caused a bifur-
cation of the field of molecular orbital studies into subdisciplines called semiempirical
and ab initio. Research groups led by Dewar and by Stewart were devoted to obtaining
solutions by substituting empirical constants into Fµν . A second approach was fol-
lowed by groups led by Pople, Gordon, and others, who used very efficient computer
codes and relied upon the increasing power of contemporary computing machines to
solve the integrals in Fµν . In general, the rule of speed vs. accuracy applies. Semiem-
pirical substitution is faster, hence applicable to larger molecules. Ab initio methods
are more accurate but they are very expensive in computer resources.
20.4 SEMIEMPIRICAL METHODS
Solution of the Schrödinger equation requires evaluation of many integrals. A large

proportion of these integrals make a very small contribution to molecular energy and
enthalpy. When they are dropped, the calculation is simplified in the hope that the
sacrifice in accuracy will be small. Dropping the integrals in the equation set leaves
only Hµν in place of the Fock matrix elements Fµν . Dropping some integrals and
replacing others with empirical parameters gives legitimate Hamiltonian elements
but elements that are approximate because of the use of empirical parameters. They
are elements Hµν of a semiempirical Hamiltonian operator. The general rule is that if
you are modifying the F matrix to obtain an approximate Hamiltonian, the method is
semiempirical. If you are working with the full F matrix and attempting to approach
a complete basis set, the method is ab initio. Neither method is exact.
The most important steps in development of a computer-based semiempirical
method were in deciding which of the many integrals (µν|λσ ) and (µλ|νσ ) in a
polyatomic molecule can be dropped, which of the integrals must be retained and
parameterized, and how they should be parameterized. Neglect of differential overlap
P1: OTA/XYZ P2: ABC
(NDO) and neglect of diatomic differential overlap (NDDO) approximations led to

a series of semiempirical programs denoted CNDO (complete neglect of differential
overlap), INDO, MINDO, and so on. Initially, NDDO programs were parameterized
to reproduce ab initio values for simple molecules but later semiempirical programs
such as the AM1 programs of Dewar et al. and the PM3 method of Stewart are
parameterized against experimental thermochemical results so as to calculate energies
and enthalpies. Presently, they are both in wide use for this purpose. Along with
thermochemical data, dipole moments, geometries, and isomerization potentials are
also calculated by modern semiempirical programs.
20.5 AB INITIO METHODS
An exact solution of the Schrödinger equation would employ no empirical parameters

beyond mass and charge of the subatomic particles, and it would be an absolute
solution to the problem that could never be changed or revised. In practice, the
absolute properties of a molecule are never found, though they may be approached
by ever-improving approximate methods. Today, what we call “ab initio procedures”
contain small empirical “corrections,” but a very serious effort has been put forth
to minimize these aspects of the procedure. One condition we wish for the set of
functions {χµ } we choose as a basis set is that they be as nearly complete as possible,
that is, we hope that the set will span the vector space.
20.6 THE GAUSSIAN BASIS SET
We saw in Chapter 17 that the single Gaussian function φ(r ) = Ce−αr , where C =
2
1.0 and α = 0.2829, does not give a very good approximation to the energy of the
hydrogen atom. The result, HF = 0.4244 E h , is 84.9% of the true (defined) energy
of the hydrogen atom. The Gaussian, with r 2 in the exponent, drops off faster than
the 1s orbital, which has r in the exponent (Fig. 20.4). The Gaussian is too “thin” at
larger distances from the nucleus.
1
1
(r)
0.5
(r)
0
0 1 2 3
0 r 3
FIGURE 20.4 The 1s STO (solid line) and a Gaussian approximation (dotted line).
P1: OTA/XYZ P2: ABC
THE GAUSSIAN BASIS SET 325
The next step is to take two Gaussian functions, called primitives, parameterized
so that one fits the STO close to the nucleus and the other contributes to the part away
from the nucleus. We seek a function
STO-2G = C1 e−α1 r + C2 e−α2 r

2 2
But how shall we apportion this linear combination so that we have one tall basis
function contributing to the orbital near the nucleus and one for a “fat” tail? Let us take
α1 = 1.0 and α2 = 0.25. Both C1 e−r and C2 e−0.25 r contribute to the sum. The larger
2 2
negative exponent is tall near the nucleus but drops off faster than the small negative
exponent. Now φ(r ) will extend to larger values of r and give us the fat tail we seek:
φ(r ) = C1 e− r + C2 e−0.25 r
2 2
But we still have two parameters to worry about C1 and C2 . They control the relative
contribution of each primitive to the final wave function. Let us take a 60/40 split and
favor the fat tail:
φ(r ) = 0.40e−1.0 r + 0.60e−0.25 r

2 2
We now have a four-parameter basis set for use with the gen keyword (File 20.1).
They are entered in the format
α 1 C1
α 2 C2
The GAUSSIAN
C input file becomes File 20.1. The STO curve fit is shown in
Fig. 20.5.
STO(r ) := e−r
φ(r ) := 0.40e−1.0 r + 0.60e−0.25 r .
2 2
# gen
hatom gen
0 2
h
1
S 2
1.0 0.40
0.25 0.60
****
FILE 20.1 (Input) A four-parameter Gaussian File for the hydrogen atom. Line 8 designates
the first center (the only one in this case) and line 9 identifies it as an s orbital with 2 basis
functions.
P1: OTA/XYZ P2: ABC
STO(r)
(r)
0.5
0 2 4
r
FIGURE 20.5 Comparison of the 1s STO of hydrogen with an arbitrarily parameterized
two-Gaussian function φ(r ) := 0.40e−1.0 r + 0.60e−0.25 r .
2 2
The fit is certainly not perfect, but it is better than what we got with a single basis
function. The energy obtained from this basis set is HF=-0.4572106 which is
91.4% of the defined value of 0.5 E h . The error has been reduced from about 15%
to about 8.5%. Variational optimization in the C parameters takes place during the
program run as we see by adding GFInput to the # gen input line in File 20.1.
This change produces the output shown in File 20.2.
AO basis set in the form of general basis input:

1 0
S 2 1.00 0.000000000000
0.1000000000D+01 0.4304660143D+00
0.2500000000D+00 0.6456990214D+00
****
FILE 20.2 (Output) The STO-2G basis set written as a 1s orbital consisting of functions
with arbitrarily selected exponents 1.00 and 0.25.
20.7 STORED PARAMETERS
We can also run an STO-2G ab initio calculation on the hydrogen atom using the
GAUSSIAN stored parameters rather than supplying our own. The input file is shown
in file 20.3. We find that there are two Gaussian primitives and one unpaired electron
from the output:
1 basis functions 2 primitive gaussians

1 alpha electrons 0 beta electrons
which agrees with the picture of the STO-2G basis set that we are trying to build. Of
course we want to know what the parameters are for the two Gaussians. The keyword
P1: OTA/XYZ P2: ABC
STORED PARAMETERS 327
# sto-2g
hatom
0 2
h
FILE 20.3 (Input) An STO-2G input file using a stored basis set.
GFinput inserted after # sto-2g in the route section of the input file produces
the added information. The parameterized STO-2G basis function is
STO-2G = 0.4301e−1.309 r + 0.6789e−0.233 r

2 2
which is not too far from the arbitrary function we guessed for input file,
File 20.1. The stored function is graphed in Fig. 20.5. The coefficients 0.4301
and 0.6789 give the intercepts of the two Gaussians at r = 0. As before, the two
α parameters determine how extended the Gaussian is in the r direction (how fat the
tail of the function is), and two C parameters determine how much of a contribution
each Gaussian makes to the final STO approximation. Stored parameters have been
optimized for general application to molecular problems more complex than the hy-
drogen atom, which accounts for the “overshoot” of α1 = 1.31 arrived at in the final
compromise.
STO-2G = C1 e−α1 r + C2 e−α2 r

2 2
α1 = 1.31, α2 = 0.233
C1 = 0.430, C2 = 0.679
We now have two ways of inserting parameters into the STO-2G calculation. We can
write them out in a gen file like input File 20.1 if we know the parameters we want,
or we can use the stored parameters as in input File 20.3 if we don’t. And we can use
the GFInput keyword to find out what stored parameters were used. This process
can be carried further for the STO-3G, STO-4G, and many other atomic orbital
approximations. The stored parameters for some orbitals are quite cumbersome, and
one would not want to enter them by hand.
AO basis set in the form of general basis input:

1 0
S 2 1.00 0.000000000000
0.1309756377D+01 0.4301284983D+00
0.2331359749D+00 0.6789135305D+00
****
FILE 20.4 (Output) GAUSSIAN
C stored parameters for the STO-2G basis set.
P1: OTA/XYZ P2: ABC
1
(r)
1(r)
2(r) 0.5
1(r) 2(r)
0 2 4
r
FIGURE 20.6 Approximation to the 1s orbital of hydrogen by 2 Gaussians. The upper curve
is the sum of the lower two curves. The solid curve is the hydrogen 1s atomic orbital.
Upon extending the basis set along the series STO-2G, STO-3G, . . . , STO-6G,
and more complicated functions, one reaches a point of diminishing return, beyond
which further elaboration produces little gain. On the positive side, the basis functions
fall into natural groups, which are replicated from one calculation to the next, hence a
way of using computer resources more efficiently is by treating each group as though
it were a single function. Each natural group of basis functions, treated as a unit, is
called a contracted Gaussian-type orbital (CGTO). For example, if we start out with
36 primitives (by no means a large number in this context) and segment them into
groups of 6, we have reduced the problem six-fold.
20.8 MOLECULAR ORBITALS
Extension of the method we have developed so far to molecules is quite straightfor-

ward. One simply puts a basis set into the input file for each atom in the molecule.
In a simple illustration of the transition from atomic to molecular input files, we can
convert the input file for a single hydrogen atom using our arbitrarily chosen STO-2G
atomic basis set into an input file for the H2 molecule. The result is File 20.5.
The second atom is added to the first, in z-matrix format. This is done by selecting
the first atom as a reference point and placing the second hydrogen atom a distance r
from it. Line 7 in File 20.5 reads h 1 r meaning “the distance between the second
hydrogen h and atom number 1 is r.” The approximate distance chosen is specified
in angstrom units immediately below in line 9 as r = 0.7.
In the previous problem of the uncharged atom, the spin multiplicity M = 2S + 1,
where S = ±1/2 is the total spin of the system, was 2 (line 5 of input File 20.3), but
the multiplicity for the paired σ electrons in molecular H2 is 1. In molecular hydrogen
File 20.5, lines 11–15 are (almost exactly) repeated in lines 16–20. The first set of
P1: OTA/XYZ P2: ABC
MOLECULAR ORBITALS 329
#gen
H2mol gen
0 1
h
h 1 r
r=0.7
1
S 2
1.00 0.40
0.25 0.60
****
2
S 2
1.00 0.40
0.25 0.60
****
FILE 20.5 (Input) A molecular orbital input file for H2 . The file is in z-matrix GAUSSIAN
format.
parameters refers to H atom 1, and the second set of parameters refers to H atom 2.
The only difference is in the atom identifiers: 1 in line 11 and 2 in line 16.
The energy output is designated a restricted Hartree Fock energy E(RHF),
E(RHF) = -1.07637168255 E h
because it is the energy of the 1σ orbital calculated on the assumption that both
spin paired electrons will be in it. If the spins were unpaired and the electrons
were in different orbitals, as in an excited state, the unrestricted Hartree Fock energy
E(UHF) would be used. Unrestricted HF calculations are somewhat more demanding
on computer resources because calculations are done on each set of electrons for α
and β spin rather than on the pair. In this simple case, the difference is negligible.
The experimental dissociation energy of H2 is 0.174 Eh . The STO-2G calculated
energy is only about 44% of the experimental value in this crude approximation. The
input file for the same task using z-matrix input and stored parameters is shown in
File 20.6.
Now the energy output E(RHF) = -1.0934083 Eh is 53% of the experimen-
tal value of 0.174 E h . Neither the arbitrary parameters in File 20.5 nor the STO-2G
stored basis set gives a very convincing calculation of the total bond energy of H2
(44 and 54%, respectively). However, when compared to the energy of two H atoms
calculated using the same basis set, the result –1.0764 – 2(–0.4572) = –0.162 E h is
93% of the experimental bond energy. This is due to cancellation of error between two
P1: OTA/XYZ P2: ABC
# sto-2g
H2 molecule
0 1
H
H 1 r
variables
r=0.7
FILE 20.6 (Input) A GAUSSIAN input file for H2 . The file is in z-matrix format.
systems: atom and molecule. Cancellation of error is often helpful but cannot be relied
upon. (Bear in mind that there is some uncertainty in the experimental value as well.)
20.8.1 GAMESS
GAMESS (General Atomic and Molecular Energy Structure System) is a molecular
orbital program intended for general and academic use as open source software. The
authors do not charge for this program, but the user may not sell it to anyone else.
GAMESS gives the STO-2G energy calculated from input File 20.7. The geometry
of the molecule is given in Cartesian coordinates. Only one nonzero coordinate is
necessary to designate the geometry of a diatomic linear molecule. The bond length
is 0.7 Å (estimated).
Translating File 20.7, each statement begins with $ and ends with $END. The
statement may occupy more than one line. The $CONTROL statement specifies the
SCFTYPe as RHF and the COORDinates as CARTesian $END. The $BASIS set is
a Gaussian BASIS, which is STO and contains a Number of GAUSSians equal to
2 $END. The $DATA set follows with an identifier Hydrogen for the human reader,
the symmetry Dn1, and the coordinates in Cartesian format $END.
The single entry 0.7 Å for the bond length is an element in the 2 × 3 matrix of
locations in Cartesian space using three components for each atom. One atom is
placed at the origin and the other is placed on the x axis at a distance of 0.7.
$CONTRL SCFTYP=RHF COORD=CART $END

$BASIS GBASIS=STO NGAUSS=2 $END
$DATA
Hydrogen
Dn 1
H 1.0 0.0 0.0 0.0
H 1.0 0.7 0.0 0.0
$END
FILE 20.7 (Input) GAMESS file for hydrogen molecule. The SCF calculation is RHF, the
coordinate system is Cartesian and the stored basis set is STO-2G. The symmetry space group
is Dn 1, the charge on each hydrogen atom is 1.0, and the bond length is 0.7 Å.
P1: OTA/XYZ P2: ABC
SPLIT VALENCE BASIS SETS 331
The energy, taken from a much more detailed output file, is
TOTAL ENERGY= -1.0934083240
The value calculated here used two Gaussian basis functions STO-2G and arrived
at an energy of −1.09340 E h which replicates the output from File 20.6. The bond
energy BE is the difference between the molecular energy and the level of the atoms,
which have an energy of 2(.5000):
BE = −1.0934 − 2 (−0.5000) = −0.0934 E h = −245 kJ mol−1
This is about half of the experimental value of 435 kJ mol−1 , slightly better than
the one calculated using the estimated coefficients in File 20.5. Using six Gaussian
functions improves the result to 0.0991 E h = 260 kJ mol−1 , which is about 57% of
the experimental value:
TOTAL ENERGY= -1.0991105267
20.9 METHANE
The carbon atom can be combined with four hydrogen atoms to form input File 20.8
for methane. The geometry for methane is given in approximate coordinates, which
will be changed during the program run. Once carbon and hydrogen can be combined,
the whole vista of alkanes, alkenes, and alkynes (many thousands of molecules) is
open. Extension to heteroatomic molecules other than hydrocarbons follows by the
same method (Example 20.1).
An optimized geometry is part of the output file. Neither the basis set nor the
optimized coordinate set is unique. Many (strictly, infinitely many) sets are possible.
Center Atomic Atomic Coordinates (Angstroms)

Number Number Type X Y Z
--------------------------------------------------------------
1 6 0 0.000000 0.000000 0.000162
2 1 0 0.000000 0.000000 1.089051
3 1 0 0.000000 1.026565 -0.363340
4 1 0 -0.889032 -0.513283 -0.363340
5 1 0 0.889032 -0.513283 -0.363340
FILE 20.8 One of many possible STO-2G optimized coordinate sets for methane. The carbon
atom has been arbitrarily placed at the origin of a Cartesian coordinate system.
20.10 SPLIT VALENCE BASIS SETS
A hundred years of theory and experiment persuades us that the electrons in atoms
reside in one or more “shells” surrounding the nucleus, each at a more or less fixed
P1: OTA/XYZ P2: ABC
radius. One way to improve our basis set is recognize the shell structure for atoms Li
and higher by splitting it into a part for the inner shell core electrons and a part for
the outer shell valence electrons. An example is the 3–21G split valence basis set,
which consists of core electrons expressed by 3 basis functions followed by 2 and 1
separate orbitals for the valence electrons. The single valence basis function 1 has a
larger excursion from the nucleus than the double basis function 2. The total orbital
is the sum of all the basis functions. Addition of new basis functions leads to larger
basis sets, which we hope will be more complete and will be a better representation
of the actual orbital. Larger split valence basis sets become complicated as can be
seen from the response to GFInput in the route section of a relatively simple 6-31G
calculation for methane:
17 basis functions 38 primitive Gaussians
20.11 POLARIZED BASIS FUNCTIONS
The electron probability density of an atom participating in a chemical bond is

distorted somewhat into the region between the atoms participating in that bond. It
is not spherically symmetrical. The new probability density can be represented by
adding some kind of a directional basis function to the spherical orbitals already used.
The new orbital is said to be polarized. A convenient change in the basis set is a p
orbital added to the spherical s orbital of a bonded H atom. This would be denoted,
for example, 6–31G(p). It turns out that adding d orbitals to the p orbitals of carbon in
a C H bond has a greater effect on the energy than does the the p-hydrogen addition.
Thus we see the more common notation 6–31G(d). A step further is to add basis
functions for both distortions, resulting in the widely used 6–31G(d,p) basis set. The
notation 6–31G** is also used.
By the time split basis sets were being used, the development of ab initio molecular
orbital theory had shifted from an effort to reproduce Slater-type orbitals to an effort
to reproduce the results of experimental measurements of the chemical and physical
properties of the molecules themselves. Basis sets were further improved by adding
new functions, each representing some refined aspect of the physics of the molecular
orbital. Electrons do not have a very high probability density far from the nuclei
in a molecule, but the little probability that they do have is important in chemical
bonding, so diffuse functions, denoted + as in 6-311 + G(d,p), were added in some
very high level basis sets. While these basis set extensions were being made, the
power of computing hardware was growing to accommodate them.
20.12 HETEROATOMS: OXYGEN
We shall extend our calculations to the STO-3G methanol input file given as
File 20.9. The stored basis functions for H, C, and O are used. For this calcula-
tion, STO-3G is requested in the # route section of the input file. The keyword
P1: OTA/XYZ P2: ABC
HETEROATOMS: OXYGEN 333
# STO-3G opt=z-matrix
methanol
0 1
C
O,1,R1
H,1,R2,2,A
H,1,R2,2,A,3,D1
H,1,R2,2,A,3,D2
H,2,R2,1,A,3,D3
Variables:
R1=1.5
R2=1.1
A=110.
D1=-120.
D2=120.
D3=60.
FILE 20.9 Input file for a GAUSSIAN C STO-3G calculation on methanol. The dihedral
angle D3 accounts for the refusal of O to conform to the H C C plane.
opt=z-matrix requests optimization in internal coordinates and output in

z-matrix format. Construction of the z-matrix input geometry goes as follows:
Take carbon as the reference atom. Call it atom 1. The oxygen atom is attached to
the carbon atom at a distance R1. Three hydrogen atoms are connected to C1, each at
a distance R2. Distance R1 is specified as 1.5 Å in the Variables: section of the
input file and distance R2 is assigned the value 1.1 Å. Bond distances are guesses
taken from prior experiments, possibly spectroscopic studies.3 The simple O C H
angles are guessed to be pretty close to the tetrahedral angle of methane, A ∼= 110◦ ,
also entered into the Variables: section of the input file. Along with the simple
angles, there are two dihedral angles D1 = –120◦ and D2 = 120◦ relative to atom 2.
Finally, there is a hydrogen (atom 5) attached to O (atom 2) at a bond length of 1.1 Å
and a simple angle A ∼ = 110◦ . This is the OH hydrogen that is out of the H C O
plane at a dihedral angle of D3 = 60◦ .
C-O
There is no external coordinate system for the z-matrix input file. The position of
each atom is given only relative to some other atom or atoms in the molecule. The
input geometry is said to be given in internal coordinates.
3 Typical spectroscopic bond lengths are 1 to 2 Å (Chapter 18). Electron diffraction studies yield 1.1 Å for
the C H bond length in 1,2-dichloroethane in the vapor phase.

P1: OTA/XYZ P2: ABC
Z-MATRIX (ANGSTROMS AND DEGREES)

CD Cent Atom N1 Length/X N2 Alpha/Y N3 Beta/Z J
--------------------------------------------------------------
1 1 C
2 2 O 1 1.439156( 1)
3 3 H 1 1.066156( 2) 2 109.116( 6)
4 4 H 1 1.066156( 3) 2 109.116( 7) 3 -120.422( 10) 0
5 5 H 1 1.066156( 4) 2 109.116( 8) 3 119.800( 11) 0
6 6 H 2 1.066156( 5) 1 109.116( 9) 3 60.253( 12) 0
--------------------------------------------------------------
Z-Matrix orientation:
--------------------------------------------------------------
Center Atomic Atomic Coordinates (Angstroms)
Number Number Type X Y Z
---------------------------------------------------- ---------
1 6 0 0.000000 0.000000 0.000000
2 8 0 0.000000 0.000000 1.439156
3 1 0 1.007363 0.000000 -0.349152
4 1 0 -0.510100 0.868665 -0.349152
5 1 0 -0.500630 -0.874157 -0.349152
6 1 0 0.499824 0.874618 1.788308
FILE 20.10 Optimized Geometry from a GAUSSIAN
C STO-3G Calculation on Methanol
(internal and Cartesian coordinates).
Even this simple molecule would be a discouraging prospect for hand calculation
as seen from the printout of the number of basis functions and primitive Gaussians:
14 basis functions, 42 primitive gaussians, 14 cartesian
basis functions
9 alpha electrons 9 beta electrons
The approximate input geometry is optimized during the program run. Output files
can be obtained in either internal or Cartesian (external) coordinates. Bond distances
and angles have been corrected in small amounts by the program.
The expected distinction between the C H bond lengths and the O H length in the
optimized geometry (1.439 vs. 1.066 Å) is found in lines 2 and 3 of the z-matrix. Also,
all simple angles and dihedral angles have been shifted from their original inputs.
The presence of an oxygen atom in the molecule distorts the tetrahedral geometry of
the CH3 group slightly. The calculated energy is
TOTAL ENERGY = -113.5386327
20.13 FINDING f H298 OF METHANOL
Either the GAUSSIAN or GAMESS calculation gives the STO-3G molecular energy
as −113.538633 hartrees. This energy is relative to the separated nuclei and electrons.
P1: OTA/XYZ P2: ABC
FINDING f H 298 OF METHANOL 335
If we knew the energy released when C, O, and H atoms are formed from their
subatomic particles, and the energy released when an appropriate number of C, O,
and H atoms combine to form C(graphite), O2 (g), and H2 (g), all in the standard state
at 0 K, we could find the energy of formation of CH3 OH(g) at 0 K.
C(gr) + 12 O2 (g) + 2H2 (g) → CH3 OH(g)

f H 298 (methanol) = H 298 (methanol(g))

− H 298 (C(gr)) + 12 H 298 (O2 (g)) + 2H 298 (H2 (g))
The numbers in square brackets have been established both by computation and by
experiment. They are shown at the left of Fig. 20.7.
There are six downward steps shown on the left of Fig. 20.7. Three are large steps
given in hartrees. Their sum –114.78647 E h represents the addition of a requisite
number of electrons to a carbon nucleus, four hydrogen nuclei, and an oxygen
nucleus to produce the gaseous atoms C, H, and O. The bottom three steps on the
left represent the enthalpy in kcal mol−1 obtained when a mole of C atoms, two
moles of H2 molecules, and 12 mole of O2 molecules, all in their standard states,
are formed from the gaseous atoms. The elements in their standard states have an
enthalpy H 298 = −115.471973 E h at the short solid line.
The enthalpy change from nuclei and electrons to the methanol molecule in its
standard state is given by the long vertical line on the right-hand side of Fig. 20.7.
The difference between the calculated enthalpy of the molecule in the standard state
and that of the composite elements in their standard states is the enthalpy difference
we seek, f H 298 . Thus the whole argument depends on the length of the vertical line
on the right-hand side of Fig. 20.7. Notice that f H 298 has the opposite sign from
-37.78934 h
H 298 (methanol)
4(-.50184) h
-74.98977 h
-169.73
4(-50.62)
-57.95
FIGURE 20.7 The G3(MP2) thermochemical cycle for determination of f H 298 of

methanol. Step sizes are not to scale. Values not designated h (hartrees) are in units of
kcal mol−1 .
P1: OTA/XYZ P2: ABC
the difference in calculated enthalpies in Fig 20.7 because it is the enthalpy input for
the reaction forming the molecule from its constituent elements.
The total drop in enthalpy for all the steps on the left-hand side of Fig. 20.7 is
–115.471973 E h . When we substitute our recently found STO-3G numbers to find
f H 298 , we get −115.471973 –(-113.5386327) = −1.93334 hartrees with a sign
change to +1.93334 E h . The result is a disaster, more than 5000 kJ mol−1 for a f H 298
that we expect will be no more than 100 kJ mol−1 or so on the basis of experimental
thermochemistry done on molecules of comparable size. What went wrong?
The procedure as described involves a serious mismatch. The atomic numbers on
the left-hand side of Fig. 20.7 are state-of-the-art experimental and computational
results, while the molecular number on the right-hand side is an STO-2G approxima-
tion. A large approximate number subtracted from a large accurate number gives a
very poor result. We could go back and substitute STO-3G energies for the atoms, but
a more rewarding approach is to go forward and try to improve the molecular enthalpy.
The experimental value of f H 298 (methanol) is about –200 kJ mol−1 , that is,
∼0.0767 E h . In order to obtain 1% accuracy in the calculated result, we must
achieve a cumulative accuracy of ±0.000767E h in our calculations. This demanding
standard is the reason why computed energies and enthalpies are usually expressed
to a precision of five or six digits beyond the decimal point. In other words, we
are working in the realm of microhartrees. We must look further into the daunting
problem of finding ways to achieve 0.000767% or 767 parts per million accuracy in
our ab initio calculations.
When the STO-2G basis set is expanded to STO-6G, we get a TOTAL ENERGY
= -114.6409388873. At a higher level of 6-31G (MP2), the output file yields
a TOTAL ENERGY = -115.1928829735. Using the opt keyword, the out-
put file reads TOTAL ENERGY = -115.2040086436 OPTIMIZED. These
results lead to f H 298 0.831034 E h = 2182 kJ mol−1 and f H 298 = 0.27909 E h =
733 kJ mol−1 and f H 298 = 0.267965 E h = 704 kJ mol−1 . Evidently we are moving
in the right direction but we still have a long way to go.
20.14 FURTHER BASIS SET IMPROVEMENTS
The literature documents a long history of improved basis sets. Details of what may
be the culminating effort in this series, the basis set G3MP2Large, are given in the
original publications (Curtiss et al., 1999). The basis set itself is available on the web
(http://chemistry.anl.gov/compmat/g3theory.htm) for use with the gen keyword if
desired.
20.15 POST-HARTREE–FOCK CALCULATIONS
No matter how good the basis set is made by extension toward an infinite set,
one encounters the Hartree–Fock limit on the accuracy of molecular energy, because
the influence of one electron upon the others has not been fully accounted for in the
P1: OTA/XYZ P2: ABC
PERTURBATION 337
SCF averaging procedure. The difference between a Hartree–Fock energy and the
experimental energy is called the correlation energy. To remedy this fault, correlated
models are made up which consist of a linear combination of the Hartree–Fock
solution plus new basis functions representing singly, doubly, and so on, substituted
wave functions:

ψ = aψ0 + aia ψia + aiabj ψiab
j + ···
ia i jab
In this equation, ij . . . designate occupied spin orbitals (orbitals treated separately

according to electron spin α or β), and ab . . . designate excited orbitals of higher
energy than ij . . . called virtual orbitals. Virtual orbitals have small but nonzero
occupation numbers. The method is called configuration interaction (CI). The first
sum on the right above includes singly substituted orbitals (CIS). Inclusion of the
second sum on the right leads to doubly substituted orbitals (CID), while inclusion
of both sums is (CISD), and so on. A QCISD(T) method, with a 6-31G(d) basis set
QCISD(T)/6-31G(d), includes exponential terms in the expansion and is generally
considered to give a better estimate of the energy than do simple CI terms alone
(Pople, 1999).
20.16 PERTURBATION
Another method of progressing beyond the Hartree–Fock limit is by inclusion of

many-body perturbation terms (Atkins and Friedman, 1997):
E = E 0 + λi E i (i = 1,2, . . .)
The terms λi E i are small energies due to perturbations of the larger Hartree–Fock
calculation of the base energy E 0 . The method was described by Moeller and Plesset
(1934), long before computers existed that were powerful enough to fully exploit it.
Higher values of λi E i lead to correlated energies. Perturbations at i = 2 or 4 are
called Moeller–Plesset MP2 and MP4 energies. They are used in corrected Gaussian
calculated energies, designated, for example, MP2/6-31G.
Atomic spin–orbit coupling energies E(SO) can be added (C: 0.14 mE h ,
H: 0.0 mE h ). A “higher level correction” (HLC) and a zero-point energy E(ZPE)
are added in the powerful combined methods called G3 or G4 calculations as well.
The zero point energy arises because the ground state of a quantum harmonic oscil-
lator is one-half quantum above the bottom of its parabolic potential well (Section
18.2). The summed zero-point energies of all atoms in a molecule, oscillating about
their equilibrium positions, is E(ZPE). The HLC, 0.009279 E h per pair of valence
electrons for a neutral molecule in the ground state, is a purely empirical factor,
parameterized so as to give the minimum discrepancy between a large test set of
accurately known experimental energies and calculated energies.
P1: OTA/XYZ P2: ABC
HF/6-31G(d) FOpt HF geometry

HF/6-31G(d) Freq ZPE and vib freq
MP2(full)/6-31G(d) Opt MP2(full) geo
QCISD(T)/6-31G(d) QCI energy
MP2/GTMP2Large GTMP2Large: energy
Scheme 20.1 A computational chemical script. The G3MP2Large set also goes under the
names G3MP2large and GTMP2Large. (If one input notation gives you an error message, try
one of the others.)
The sum of all these five energy terms with a G3 basis set at the MP2 post-
Hartree–Fock level is E0 [G3(MP2)]:
E 0 [G3(MP2)] = E(QCISD(T)/6–31G(d)) + E MP2 + E(SO)

+ E(HLC) + E(ZPE)
20.17 COMBINED OR SCRIPTED METHODS
A quantum mechanical script is a list of procedures that are carried out sequentially
and automatically once the program has been started. Combined or scripted programs
like the GAUSSIAN C family and the CBS C group by Petersson (1998) are very
popular. A sample script is shown as Scheme 20.1.

Using the results from the scripted sequence, an extrapolation can be carried out
as shown in Fig. 20.8. The desired energy E[(QCISD(T)/G3MP2Large] cannot be
obtained by direct calculation, but the other three energies can. With these three
energies, extrapolation to the desired energy is carried out as diagramed in the fig-
ure. The line from E[MP2/6-31G(d)] to E[MP2/GTMP2Large] represents lowering
of the MP2/6-31G(d) energy when the basis set is expanded to MP2/GTMP2Large.
The line from E[MP2/6-31G(d)] to E[QCISD(T)/6-31G(d)] is the 6-31G(d) energy
lowering when the QCISD(T) correlation is taken into account.
All of the calculations above are carried out within the G3(MP2) script. The
output portion of the G3(MP2) printout of H 298 for methanol shown in File 20.11
contains the message G3MP2 Enthalpy = -115.547933. This value inserted
into Fig 20.7 as H 298 gives f H 298 (methanol) = −199.4 kJ mol−1 as compared to
the experimental value of –201.5 ± 0.3 kJ mol−1 . The difference is 1%. We have
reached our accuracy goal for methanol and, by implication, for any molecule we
choose, subject only to limitations on computer power.
P1: OTA/XYZ P2: ABC
DENSITY FUNCTIONAL THEORY (DFT) 339
Energy
E[MP2/6-31G(d)
E[MP2/G3MP2Large]
E[QCISD(T)/6-31G(d)]
Basis Set
E[QCISD(T)/G3MP2Large
Post
Hartree-Fock
FIGURE 20.8 Additive extrapolations in the G3(MP2) scripted method.
Temperature= 298.150000 Pressure= 1.000000

E(ZPE)= 0.049406 E(Thermal)= 0.052749
E(QCISD(T))= -115.374855 E(Empiric)= -0.064953
DE(MP2)= -0.161818
G3MP2(0 K)= -115.552220 G3MP2 Energy= 115.548877
G3MP2 Enthalpy= -115.547933 G3MP2 Free Energy= -115.574944
1\1\GINC-DOUG\Mixed\G3MP2\G3MP2\C1H4O1\DROGERS\08-Jun-2009\0\\# g3mp2\
\methanol\\0,1\C,0,0.0185989192,0.032214273,0.0391675131\O,0,0.0052594
495,0.0091096338,1.4617939406\H,0,1.0367987363,0.0123972426,-0.3682155
629\H,0,-0.5076630411,0.9040926655,-0.3682155629\H,0,-0.5013614095,-0.
8683834341,-0.2863534272\H,0,0.4651982827,0.8057470613,1.7693787775\\V
ersion=AM64L-G03RevD.01\State=1-A'\MP2/6-31G(d)=-115.3461339\QCISD(T)/
6-31G(d)=-115.3748547\MP2/GTMP2Large=-115.5079522\G3MP2=-115.5522202
FILE 20.11 Partial GAUSSIAN G3(MP2) output. (G3(MP2) is a slightly simplified version
of G3.)
20.18 DENSITY FUNCTIONAL THEORY (DFT)
DFT methods are quite fast relative to G3 and G3 (MP2) scripted molecular orbital
model chemistries. For this reason they have enjoyed popularity among computational
chemists working on practical problems. Density functional theory is based on the
fact that the average energy of an electron in an atom or a molecule is a function
of its probability density ρ in the vicinity of the nucleus or nuclei, and ρ is a
function of its position in space. A function of a function is a functional. DFT
methods are largely empirical, containing adjustable parameters, though each rests
on a reasonable theoretical model. One of the most accurate functionals, devised by
Becke (1932) and by Lee, Yang, and Parr in 1988 is denoted BLYP. The BLYP method
P1: OTA/XYZ P2: ABC
using three parameters is denoted B3LYP. At present, B3LYP calculations are quite
satisfactory for smaller hydrocarbons but do not compete successfully in studies of
larger molecules.
Example 20.1 Methane

Extend input File 20.7 to methane by including more basis functions in z-matrix
format. Use the opt keyword in the route section of File 20.12 to find the optimized
bond length.
Solution 20.1 In File 20.12, the hydrogen atoms are added in z-matrix format
sequentially below the carbon atom C, designated atom 1, the reference point. For the
first H atom, we need to know only the distance relative to C, r = 1.1 Å. To locate
the second hydrogen, we need r and the simple H C H angle, 109◦ ; and for the
third and fourth hydrogens we need the distance, the simple angle, and the dihedral
angles with the hydrogens already in place, −120◦ and 120◦ .
# STO-2G opt
Methane
0 1
C
H,1,R
H,1,R,2,A
H,1,R,2,A,3,D1
H,1,R,2,A,3,D2
Variables:
R=1.1
A=109.
D1=-120.
D2=120.0
FILE 20.12 A z-matrix input file for methane. The distances and angles are estimates.
Example 20.2 A GAMESS Calculation of Methane

For those wishing to perform the calculation using GAMESS freeware, the input
file consists of three command lines, a symmetry specification cs followed by the
z-matrix as already shown. Be sure to include the blank lines in the CONTROL file
leave a blank line for the word Variables:. GAMESS can be a little more cranky
than user friendly GAUSSIAN, but hey, . . . you can’t beat the price.
P1: OTA/XYZ P2: ABC
$CONTRL SCFTYP=RHF MULT=1 RUNTYP=OPTIMIZE COORD=ZMT $END

$DATA
Methanol
cs
C
O,1,R1
etc.
$END
FILE 20.13 Control lines for a GAMESS calculation. NGAUSS can be set to 2,3, . . . , 6 as
desired in line 2 of the control file. A z-matrix is selected here but either internal or external
coordinates can be used (see File 20.7).
Example 20.3 Symmetry Unique Atomic Coordinates

In some cases, symmetry can be used to reduce the demand on computer resources.
Although symmetry is most important in carrying out high-level calculations on large
molecules, the concept can be illustrated by a very simple example, the input file for
molecular hydrogen in the GAMESS procedure (Section 20.8.1). In this example, we
shall go from the GAMESS input file for molecular hydrogen in File 20.7 to one that
makes use of the plane of symmetry separating one hydrogen atom from an identical
partner at an equal distance from the center of symmetry.
Solution 20.3 The file presented in File 20.7 contains more information and de-
mands more computation than is necessary. Any calculation done on one H atom
must be identical to the calculation done on the other. Therefore, why not do one
calculation and be done with it?
This is accomplished by replacing the two-line Cartesian coordinate location
matrix by one line specifying the location of one of the two hydrogen atoms rel-
ative to their center of symmetry. In the CONTRL line, COORD=CART is replaced
by COORD=UNIQUE and Dn 1 is replaced by Dnh 4. The distance of the se-
lected H atom from the center of symmetry is one half the bond length, 0.35 Å.
If the bond length is not known, RUNTYP=OPTIMIZE can be inserted before
COORD=UNIQUE.
$CONTRL SCFTYP=RHF RUNTYP=OPTIMIZE COORD=UNIQUE $END

$DATA
Hydrogen
Dnh 4
H 1.0 0.0 0.0 0.35

$END
P1: OTA/XYZ P2: ABC
The energy output of this simplified file is the same as the more complicated
File 20.7.
TOTAL ENERGY = -1.0934083240
Optimizing from a bond length estimate that is 10% in error leads to
E = -1.0938179551
which is different by about 0.25 kcal mol−1 (slightly lower).
Problem 20.1
Find the energy of a particle in a one-dimensional box of length l. by the variational
method. Take (x) = A sin nπ x/l as a trial function. Note that the Hamiltonian is
−(2 /2m)(∂ 2 /∂ x 2 ).
Problem 20.2
Electron diffraction studies yield two different Cl Cl distances (not bonded) in 1,2-
dichloroethane CH2 Cl CH2 Cl. Explain why this is so. What influence does this fact
have on molecular structure and energy calculations like MM and G3(MP2)?
Problem 20.3
The benzene molecule can be modeled as a potential well with a square planar bottom
0.400 nm on each side containing six 2pπ electrons (ignoring all the other electrons
in the molecule). What is the degeneracy of the occupied orbitals and what is the
degeneracy of the first two virtual orbitals according to this model? In what respect
does this model fail to represent the true levels of the benzene molecule?
Problem 20.4
Given the model proposed in Problem 20.3, what is the minimum energy required to
promote a ground state electron to its first excited state? What is the wavelength of
electromagnetic radiation that will supply this energy? Is this in the visible region?
Is benzene colored?
Problem 20.5
(a) Add a basis function of your own to the STO-2G basis set so as to create
your personal STO-3G linear combination. Use a graphing program to plot
your function, call it ϕmySTO-3G = f (r ). STO-3G calculations were used at
the research level until the 1990s.
P1: OTA/XYZ P2: ABC
(b) What is the energy of the hydrogen atom according to your STO-3G basis
set? What is the percent difference between your value and the exact (defined)
value of E H = 0.5000 E h ?
Problem 20.6
Carry out a 3-21G split valence calculation of the energy of atomic helium, He. What
is the approximation to the first ionization potential of helium in this model? The
experimental value is 0.903 E h .
Problem 20.7
(a) Carry out a 3-21G calculation of the energy of the molecular ion HeH+ .
Do you predict that this molecular ion exists?
(b) Carry out a 3-21G calculation of the energy of the bond energy of the molecular
ion He2 + . Do you predict that this molecular ion exists?
(c) Carry out a 3-21G calculation of the energy of the bond energy of the molecule
He2 . Do you predict that this molecule exists?
(d) Carry out a full 3-21G calculation of the total energy of the lithium hydride
molecule.
Problem 20.8
Suppose, on an exam, you are given the following Hartree–Fock 3-21G total en-
thalpies H 298
Methane HF = 39.9768776 Ethene HF = −77.6009881

Ethane HF = 78.793948 1,3-Butadiene HF = −154.0594565
along with the enthalpies of formation of f H 298 (methane) = −74.4 kJ mol−1 ,

f H 298 (ethene) = 52.5 kJ mol−1 , and f H 298 (ethane) = −83.8 kJ mol−1 . No
other information is given. Find the enthalpy of formation of 1,3-butadiene. See
also Chapter 21, Section 21.4.
Problem 20.9
Recalculate the previous result, this time using the 6-31G basis set to calculate the
HF values of the four hydrocarbons in the reaction
CH2 CH CH CH2 + 2CH4 → 2CH2 CH2 + CH3 CH3
Is the agreement with the experimental result improved by using the higher-order
basis set? By how much?
P1: OTA/XYZ P2: ABC
21
PHOTOCHEMISTRY AND THE
THEORY OF CHEMICAL REACTIONS
So far we have considered chemical reactions in which the energy necessary to

surmount the activation barrier is obtained from the environment in the form of heat
translated as molecular motion. These are thermal reactions. Light of frequency ν is
also able to impart energy E = hν to reactant molecules so that they can surmount
the barrier. These are photochemical reactions.
21.1 EINSTEIN’S LAW
Like any chemical reaction, a photochemical reaction involves excitation of the

reactant species up and over an activation energy barrier before transmission to
the product state. In a photochemical reaction, the activation energy is supplied by
incident light instead of ambient heat. The Bohr–Einstein concept of a particle of
light (a photon) carrying a quantum of energy E = hν leads one to suppose that the
energy of an incident light particle is concentrated in one reactant molecule only,
rather than being distributed evenly over an entire collection of molecules. This
mechanism of energy transmission means that if the quantum E is sufficiently large,
all struck molecules will acquire the activation energy and the number of molecules
reacting will be equal to the number of photons striking and retained by the system.
This is Einstein’s law.
molecules reacting = photons absorbed

Copyright
344
P1: OTA/XYZ P2: ABC
QUANTUM YIELDS 345
Many systems follow Einstein’s law, which argues strongly in favor of the photon
concept of light and the proposed mechanism for energy transfer. Many systems,
however, deviate widely from Einstein’s law, so the true nature of photochemical
reactions must be more complicated than we have pictured so far.
21.2 QUANTUM YIELDS
There are systems for which the number of reacting molecules is smaller than the
number of incident photons and there are those for which it is larger, sometimes much
larger. A system can be characterized by its quantum yield , defined as
number of molecules reacting

=
number of photons absorbed
A system that obeys Einstein’s law has = 1.

When a quantum yield is less than 1, as in the first entry in Table 21.1, we suppose
that there are energy dissipating processes going on. In a complicated molecule, the
quantum of incident energy may be split up among several degrees of freedom and
dissipated eventually as heat. In some molecules, a part of the incoming quantum of
energy is reemitted as light. The system might emit part of the excitation energy as
light and the remainder as heat.
If the energy emitted is less than the energy taken in, the wavelength of the emitted
radiation is the longer of the two. This is the common observation in fluorescence;
beaming ultraviolet light (short λ, high energy) on a fluorescent material produces
visible light (longer λ, lower energy). If there is a time lag between absorption and
emission, the phenomenon is called phosphorescence (Fig. 21.1). Explanation of a
high quantum yield is not quite so simple and involves a new concept, the chain reac-
tion. One postulates a reaction mechanism for a photoinduced chain reaction in which
1. an energetic species is produced, which

2. brings about a chemical reaction in which it is replicated, and it
3. Passes its energy on to produce another product molecule and another energetic
species. By this mechanism, each incident photon can be responsible for many
product molecules, so a high quantum yield results. Explosions are possible.
The production of HCl from the elements H2 and Cl2 is an example of a
TABLE 21.1 Some Experimental Quantum Yieldsa .
Reaction Quantum Yield
(CH3 )2 C O → CH3 C• O + • CH3 0.17

CH3 COOH → CH4 + CO2 1
H2 + Cl2 → 2HCl ∼105
a The symbol • indicates a free radical.
P1: OTA/XYZ P2: ABC
346 PHOTOCHEMISTRY AND THE THEORY OF CHEMICAL REACTIONS
excited state
photon in photon out

intermediate state
heat
ground state
FIGURE 21.1 Mechanism for fluorescent and phosphorescent light emission.
photoinduced chain reaction. The process starts with the production of two
free radicals Cl• by absorption of a photon hν. The first step is called the
chain-initiating step:
Cl2 + hν → 2Cl•
The chain initiating step is followed by the chain-propagating steps:
Cl• + H2 → HCl + H•
H• + Cl2 → HCl + Cl•
These steps constitute the heart of the chain reaction and continue indefinitely,
which explains the high quantum yield. Ultimately, however, a free radical
collides with another free radical
Cl• + Cl• → Cl2 + heat
or loses its energy by some other transfer mechanism. These are the chain-
terminating steps.
21.2.1 Lipid Peroxidation

In lipid peroxidation or autooxidation, the chain-initiating step is the production of a
lipid free radical R• by removal of a hydrogen atom from the ground state molecule
by light, heat, or enzymatic initiation. Lipid peroxidation follows a chain sequence.
The chain-propagating steps are
R• + O2 → ROO•
ROO• + RH → ROOH + R•
and the chain-terminating step is
ROO• + ROO• → molecular products

P1: OTA/XYZ P2: ABC
QUANTUM YIELDS 347
This chain reaction is damaging to biological membranes, especially cell walls, and
has been implicated in both carcinogenesis and the aging process. Within living
systems, the unsaturated fatty acids are most vulnerable to autooxidation.
Living biological systems (as contrasted to dead ones) are protected from this
autooxidation process through the action of small amounts of exogenous antioxidants,
including vitamin E. Antioxidants form very reactive free radicals themselves. These
radicals interfere with the chain oxygenation cycle in the propagation steps by reacting
with the radical ROO• . This terminates the chain and limits damage to the organism.
Vitamin E is the name given to the group of four tocopherols. The tocopherols are
aromatic alcohols ArOH, which react with peroxyl radicals.
ROO• + ArOH → ROOH + ArO•
The phenol radical ArO• is stabilized by delocalization of electrons over the aromatic
framework. Because of their stability, they persist in the system and eventually react
with a peroxyl radical to complete the chain-breaking mechanism
ArO• + ROO• → products
21.2.2 Ozone Depletion

Neither O nor O2 absorbs radiation in the 200- to 300-nm (ultraviolet) region, but
ozone O3 absorbs at about 250 nm, shielding the earth’s surface and creatures living
on it from mutagenic DNA alteration by impact of ultraviolet photons. Among the
many photochemical reactions occurring in the earth’s stratosphere is a sequence in
which the middle two reactions below are in balance, maintaining a low but fairly
constant ozone level.
O2 + hν (< 200 nm) → 2O

O3 + hν(UV ) → O2 + O
→ O3 + M
O + O2 + M ←
O + O3 → 2O2
M is a molecule, possibly N2 , that does not enter into the reaction but exchanges
energy through collisions with those that do.
Other reactions interfere with this cycle, hence they interfere with the stratospheric
ozone level. An example is the chlorine cycle involving Cl2 O2 derived ultimately from
man-made chlorofluorocarbons. Stratospheric chlorine has increased by a factor of
about 7 in the past 50 years, and a decrease in stratospheric ozone has been found
(primarily for reasons of atmospheric turbulence) over Antarctica. This is the source
of the polar “ozone hole.” There is no reason to suppose that ozone depletion, though
it is most easily monitored over the poles, is limited to the polar stratosphere.
P1: OTA/XYZ P2: ABC
21.3 BOND DISSOCIATION ENERGIES (BDE)
In photoinduced and many other mechanisms, we are concerned with the stability
and rate of formation of free radicals, for example,
CH4 → CH•3 + H•
The enthalpy of homolytic cleavage is not easy to measure for these fleeting species,
but they can be calculated as an ordinary dissociation:
AB → A• + B•
The bond dissociation energy BDE of ethane
CH3 CH3 → CH•3 + CH•3
to produce methyl radicals is a simple example of the more general case for hydro-
carbons and their radicals:
CH3 R → CH•3 + R•

BDE [CH3 R] = f H298 CH•3 + f H298 [R• ] − f H298 [CH3 R]
The radical R may be branched, or it may contain double or triple bonds, or it may
contain two or all three of these structural features (Rogers et al., 2006).
21.4 LASERS
Some atoms and molecules can undergo a population inversion, such that an upper
energy state is more highly populated than the ground state. When this happens,
incident radiation is reemitted with its initial energy plus an energy gain obtained as
the gain medium returns to its ground state. In commercial lasers, atoms or molecules
comprising the gain medium are intentionally driven into a population inversion by
an optical pump, which emits high-energy radiation or fast electrons.
A laser consists of an optical pump that brings about a population inversion in a
gain medium which lases, bringing about a slight radiative amplification of the light
source. The radiation from this laser gain medium is reflected back and forth between
two mirrors so that amplification is increased on each pass through the medium. One
mirror is intentionally made less reflective than the other so that some of the highly
amplified radiation passes through the less reflective mirror and is emitted as a laser
beam. Einstein developed the theory of lasers long before the first commercial laser
was produced.
P1: OTA/XYZ P2: ABC
THE EYRING THEORY OF REACTION RATES 349
21.5 ISODESMIC REACTIONS
Ab initio enthalpy calculations of H 298 for hydrocarbons containing more than 2

or 3 carbon atoms are often converted to thermodynamic values by setting up an
isodesmic reaction in which an experimental f H 298 is known for all participants
but one (Hehre et al., 1970). The single remaining unknown f H 298 is calculated
from the known experimental values by difference. Computed f H 298 values for
individual molecules on the right and left of the equation may suffer considerable
error; but if the bond types are the same on both sides of the reaction, errors tend
to cancel because they arise from similar computational defects. Isodesmic reactions
are usually set up such that a relatively complicated molecule is compared to several
simple molecules like methane and ethane, for which the experimental values are
thought to be very accurate.
21.6 THE EYRING THEORY OF REACTION RATES
Henry Eyring (1935) worked out a theory of reaction rates, which explains the
Arrhenius law of temperature dependence and which leads to a conceptual picture
that has been widely used in many branches of chemistry. In Section 20.2, we based
our qualitative thinking about complicated molecules on the system of the hydrogen
molecule ion because it is the only molecule simple enough to describe completely.
So we would like to base our thinking about complicated reaction mechanisms on
the simpler system of one hydrogen atom substituting for another1 :
H+H−H→H−H+H
The attack of a hydrogen atom on a hydrogen molecule encounters an ellipsoidal

repulsive field with the minimum repulsion at either end of the molecule.
H H
Under the simplifying assumption that the attack of the hydrogen atom on H2 is an
end-on attack, only two distances are necessary to describe the reacting system, the
distance from the the attacked hydrogen molecule to the incoming H, and the distance

from the opposite end of the molecule to the outgoing H. Call these distances rHH and
1 Experimentally, deuterium atoms D were used because they can be distinguished from H by their
difference in mass.
P1: OTA/XYZ P2: ABC
rHH . At some instant, the incoming H and the departing H must be equidistant from
the central hydrogen atom so that the three atoms form an energetic intermediate:
⎡
⎤
rHH rHH

H + H − H = ⎣H · · · H · · · H⎦ = H − H + H
The bond-forming configuration is on the left and the bond-breaking configuration is

on the right of the central structure, which is the activated complex.
21.7 THE POTENTIAL ENERGY SURFACE
In the study of molecular structure, we plotted the potential energy of the molecule as
a function of internuclear distance and found that it has a minimum at the equilibrium
bond length. Some of the qualities of this potential energy curve carry over to the
activated complex, but now the potential energy is a more complicated function of two

variables rHH and rHH . It requires a three-dimensional surface or a two-dimensional
contour map for its representation. Eyring has constructed such a surface represented
by the equivalent contour map in Fig. 21.2.
This contour map is drawn in the same way that a forest service map of Western
Montana would be drawn. Each contour corresponds to a constant potential energy
that would occupy the dimension out of the plane of the paper if the three-dimensional
representation were used. The hyperbola connecting the two arrows is the locus of
potential energy minima of the system as the reaction takes place. Contour 4 bounds
a potential energy plateau.
1
2 4
3
r' HH
r HH
V
FIGURE 21.2 Eyring potential energy plot for the reaction H + H H → H H + H.
P1: OTA/XYZ P2: ABC
THE POTENTIAL ENERGY SURFACE 351
Activated Complex
a
H + H-H H-H + H
FIGURE 21.3 Activation of the symmetrical reaction H + H H → H H + H. The height

of the barrier is about 40 kJ mol−1 .
If a slice is taken at the horizontal dotted line at the top of Fig. 21.2, its shape is
that of a typical two-center molecular potential energy curve with a minimum like

those in Fig. 20.2. This is because the distance rHH is large and the potential energy
as a function of rHH is essentially that of an unperturbed H2 molecule. The same
argument holds for a slice taken at the vertical dotted line to the right of the figure;
rHH is large and the system is an H2 molecule with the H atom so far away as to have
no effect on it.
The curved line connecting the reactant H H potential energy basin with the prod-
uct H H potential energy basin is the path a system must follow if the transformation
from one H H to the other is to take place. This is the reaction coordinate, leading
from reactants, over an activation barrier to products. During the reaction, as the sys-
tem moves along the reaction coordinate, there is a rise in potential energy to a saddle
point which is a kind of “mountain pass” between the two potential energy basins.

At the top of the pass, rHH = rHH and the system exists as the activated complex. The
potential energy as a function of the reaction coordinate for the system appears as
shown in Fig. 21.3.
The potential energy of products is equal to the potential energy of the reactants
in the case of H + H H → H H + H, but they are different in the general case
of a more complicated reaction. Figure 21.4 shows the activation energy barrier
that a more complicated system must surmount if it is to go from the reactant
Activated Complex
H aH
Activation
Enthalpy Reactants
rH
Enthalpy of
Reaction Products
FIGURE 21.4 The enthalpy of activation of an exothermic reaction.

P1: OTA/XYZ P2: ABC
state to products. There are two distinct factors influencing the reaction: (1) the
thermodynamic factor, which tells where the equilibrium is going and where it will
be after infinite time, and (2) the kinetic factor, indicating the time scale necessary
to reach equilibrium—that is, whether “infinite time” is a matter of microseconds,
hours, weeks, eons, and so on. We have seen that the thermodynamic factor is
largely determined by the difference between the enthalpy levels of the reactants and
products. The kinetic factor is largely determined by the height of the enthalpy barrier
between them.
21.7.1 Optical Inversion

An example of the Eyring mechanism is the explanation of certain geometric changes
that take place during organic reactions. If a carbon atom is attached to four different
groups, it is optically active and it rotates the plane of polarized light (Fig. 21.5).
The angle of optical rotation can be measured experimentally. Such a substance can
be geometrically changed by a chemical reaction involving an attacking ion. If the
attacking ion is identical to group D, a substance is produced that is identical to the
original compound except that it rotates light in exactly the opposite direction. This
inversion of optical rotation is readily explained using the concept of the activated
complex.
B C
D
A
FIGURE 21.5 An optically active species.
The incoming D brings about a geometric inversion of groups A, B, and C

attached to the central carbon. The geometric inversion causes a reversal in the
angle of optical rotation. The enthalpy barrier to inversion, which is the control-
ling factor in the rate of inversion, is the 5-bonded carbon atom at the center
of Fig. 21.6. This is clearly an unstable species, hence it represents an enthalpy
maximum.
D+ D D D + D
FIGURE 21.6 Inversion of optical activity.

P1: OTA/XYZ P2: ABC
THE STEADY-STATE PSEUDO-EQUILIBRIUM 353
21.8 THE STEADY-STATE PSEUDO-EQUILIBRIUM
The enthalpy difference between the reactant state and the activated complex is not
a legitimate equilibrium in the thermodynamic sense; but it is useful to treat it as
though it were. It is best called a pseudo-equilibrium or steady-state equilibrium.
Suppose a second-order reaction proceeds through an activated complex [AB]*:
A + B = [AB]∗ → products
The activated complex is controlled by a pseudo-equilibrium constant:
[AB]∗
K∗ =
[A] [B]
If the reaction to form products is a first-order breakup of the activated complex, we
obtain
dA
rate = − = k1 [AB]∗
dt
The concentration of activated complex through its equilibrium constant is
[AB]∗ = K ∗ [A][B]
whence
d[A]
− = k1 K ∗ [A][B]
dt
This equation explains the observation of second-order overall kinetics for the re-
action, even though the rate constant for decomposition of the activated complex is
first order. According to this mechanism, k1 K * is the observed second-order rate
constant.
The incoming and outgoing species of the activated complex may be weakly
bound. If so, we may regard its breakup as a kind of one-cycle molecular vibration.
The complex stretches along its weakest bond, and then it falls apart to give either
products or the original reactants. The vibrational energy acquired during this one-
dimensional, one-cycle stretch is E = k B T because there is one degree of vibrational
freedom contributing 12 k B T to the kinetic energy and one contributing 12 k B T to the
potential energy. It is also true for a vibration that E = hv, hence
kB T RT
k B T = hv ν= =
h NAh
where N A is the Avogadro number.
In the limit of an activated complex that always breaks up to give products, the
number of vibrations taking place per second is the number of activated complexes
P1: OTA/XYZ P2: ABC
reacting because each one-cycle vibration leads to products. Thus k1 is ν (notice that
both are frequencies) and
d [A] RT ∗
− = K [A] [B]
dt NAh
If the assumption that all vibrations lead to products does not hold, one includes a
transmission coefficient κ, which is that fraction of activated complexes that break
up to give products. Now,
d [A] RT ∗
− =κ K [A] [B]
dt NAh
or
RT ∗
kobserved = κ K
NAh
21.9 ENTROPIES OF ACTIVATION
In general, the free energy change of a reaction varies as the natural logarithm of its
equilibrium constant, so for the steady-state activation equilibrium we write
G ◦∗ = −RT ln K ◦∗
where the superscript designates the standard state of the Gibbs free energy for the
activation pseudo-equilibrium. The equivalent form for K ◦∗ is
◦∗
K ◦∗ = e−G /RT
By analogy to a true equilibrium, for which
G ◦ = H ◦ − T S ◦
we write
∗
/R −H ∗ /RT
K ◦∗ = eS e
The observed rate constant is
RT S ∗ /R −H ∗ /RT
kobserved = κ e e
NAh
P1: OTA/XYZ P2: ABC
which can be compared with the Arrhenius equation
kobserved = se−Ha /RT
Noting that Ha in Arrhenius theory is the same thing as H ∗ in Eyring’s theory,
the preexponential s must be
RT S ∗ /R
s=κ e
NAh
This expression gives a preexponential that is a function of the temperature as con-

trasted to Arrhenius’ conclusion that s is constant. Accurate measurements have
indeed shown the temperature dependence of the supposed constant s.
21.10 THE STRUCTURE OF THE ACTIVATED COMPLEX
If the activated complex has a “tight” structure, its formation implies a reduction
of freedom in the system, therefore a negative entropy change S. If the activated
complex has a loose structure by comparison to the reactants, S is positive hence
the preexponential factor is large as well. For some reactions, e−Ha /RT may be
unfavorable but the reaction rate is appreciable because of a large preexponential.
The reactions
NO + O3 = NO2 + O2
and
CH3 I + HI = CH4 + I2
√
have been shown to√have second-order rate constants k = 6.3 × 107 T e−2300/RT
and k = 5.2 × 1010 T e−33,000/RT , respectively. Analyzing these two results, we
conclude that the first reaction has a low activation barrier relative to the second and
it has a lower preexponential than the second reaction. The entropy of activation is
more negative for the nitric oxide oxidation than the iodine abstraction, suggesting
a tightly bound structure for the first activated complex and a loose structure for the
second activated complex.
Example 21.1 A C C Bond Dissociation Enthalpy (BDE)

Find the enthalpy of dissociation (BDE) of the C C bond in ethane in the G3(MP2)
model chemistry.
P1: OTA/XYZ P2: ABC
Solution 21.1 The term “G3(MP2) model chemistry” means that G3(MP2) calcu-
lations are used for all molecular and radical species. In the case of ethane, homolytic
cleavage of the C C bond
CH3 CH3 → 2 CH•3
has a dissoc H 298 and a BDE:

BDE [CH3 CH3 ] = dissoc H 298 [CH3 CH3 ] = 2f H 298 CH•3 − f H 298 [CH3 CH3 ]

= 2H 298 CH•3 − H 298 [CH3 CH3 ]
= 2 (−39.752873) − (−79.646716) = 0.1410E h
= 88.4 kcal mol−1 = 370.0 kJ mol−1
Calculated enthalpies of formation of the molecule and radicals from the nuclei
and electrons H 298 can be substituted for f H 298 of the molecule and radicals in
Example 21.1 because the enthalpies of formation of the atoms cancel in the homolytic
cleavage reaction (they are the same for 2 CH•3 as for CH3 CH3 ). The experimental
value for this BDE is 89.7 kcal mol−1 = 375.3 kJ mol−1 , so the difference between
the experimental value and calculated value is 1.4%.
Example 21.2 An Isodesmic Reaction

Find the enthalpy of formation of tert-butylmethane (2,2-dimethylpropane) by an
isodesmic reaction using known values of f H 298 (methane) = −17.8 kcal mol−1
and f H 298 (ethane) = −20.1 kcal mol−1 .
Solution 21.2 Cheng and Li (2003) calculated f H 298 of n-tert-butyl methanes,

where n = 1, 2, 3, and 4. Find f H 298 of the simplest example tert-butylmethane
(2,2-dimethylpropane). The isodesmic reaction is
CH3
|
CH3 −C−CH3 + 3CH4 −→ 4C2 H6
|
CH3
−197.35354 −40.42210 −79.65120
−17.8 −20.1
Computed values of f H 298 are in the first line below the reaction, and experimental
values of f H 298 for CH4 and C2 H6 are in the second line.
On the reactant side, two alkanes have a total of 24 C H bonds and 4 C C bonds.
On the product side, one molecule has a total of 4(6) = 24 C H bonds and 4(1) =
4 C C bonds. The reaction is isodesmic because the number and types of bonds on
P1: OTA/XYZ P2: ABC
the left is the same as the number and types of the bonds on the right. One value is
missing from this scheme, that of f H 298 for t-butylmethane.
Their calculated enthalpy of reaction r H 298 was
r H 298 = 4(−79.64672) − (−197.35354) − 3(−40.41828)

= 0.02150 h = 13.49 kcal mol−1 = 56.44 kJ mol−1
This enthalpy change is used to calculate the remaining unknown from the experi-
mental values of f H 298 for methane and ethane:
r H 298 = 4f H 298 (ethane) − f H 298 (t-butylmethane) − 3f H 298 (methane)

13.5 = 4(−20.1) − f H298 (t-butylmethane) − 3(−17.9)
which leads to
f H 298 (t-butylmethane) = −40.2 kcal mol−1 = −168.2 kJ mol−1
The experimental value is −39.9 ± 0.2 kcal mol−1 = −166.9 kJ mol−1 .
Comments: So far, the isodesmic reaction has merely passed the test of reproducing
a known experimental result. Agreement with the experimental value is also good
for the second compound in the series, but experimental work is uncertain and
under debate for the third compound. An experimental value is nonexistent for the
last compound named, which is highly strained and has not yet been synthesized.
The authors carry these calculations on in a logical sequence to obtain f H 298 of
all four compounds di(t-butyl)methane (−59.2), tri(t-butyl)methane (−55.3), and
tetra(t-butyl) methane (−320.7 kcal mol−1 ). Example 21.2 is a good example of
verification of a method for known compounds followed by extension to unknowns
that are not amenable to experimental work.
Problem 21.1
Chlorine does not react with toluene in the dark. If light is admitted to a chlorine–
toluene mixture, a reaction occurs. Propose a mechanism for this reaction. Propose
an experiment to support or contradict your mechanism.
Problem 21.2
The number density of one mole of methane at 1 bar and 298 K is ρ = n/V =
N A p/RT = 2.43 × 1025 m−3 under the ideal gas assumption. The collision cross
section is 5.3 × 10−19 m2 , and the average speed of the methane molecule is x
=
(8RT /πM)1/2 = 630 m s−1 . What is the collision frequency? Carry all units through
your calculation and demonstrate that the answer is, indeed, a frequency.
P1: OTA/XYZ P2: ABC
Problem 21.3
Find the BDE for the following reactions within the G3(MP2) model chemistry:
CH2 = CH − CH3 → CH2 = CH• + CH•3
and
CH2 = CH − CH2 − CH3 → CH2 = CH − CH•2 + CH•3
Discuss the implications of this difference for free radical formation in unsaturated
fatty acids as contrasted to saturated fatty acids in physiological systems (Section
21.2.1). You will need to use data from Exercise 21.1.
Problem 21.4
Three coffee cups are arranged vertically A, B, C from top cup A to bottom cup C.
Cups A and B have holes in the bottom but C does not. A drains into B, and B drains
into C.
A→B→C
The hole in A is slightly larger than the hole in B. One cup of coffee is poured into
A. What happens?
1. Sketch the liquid levels in A, B, and C as a function of time.

2. The hole in A is made a little larger. Sketch A, B, and C as a function of time.
3. The hole in A is made a little larger still. Sketch A, B, and C as a function of
time.
4. If the process is continued, what happens to level B?
Problem 21.5
Nitric oxide, NO, is very destructive to ozone in the stratosphere. (Commercial
supersonic jet flight has been banned because of the potential increase in stratospheric
NO.) Propose a mechanism for this destructive reaction.
Problem 21.6
The helium–neon laser produces light at 1152.3 nm. What is the energy of one mole
(an einstein) of radiation at this wavelength?
P1: OTA/XYZ P2: ABC
Problem 21.7
Irradiation of a gaseous sample of acetone(g) by light of 313 nm brings about pho-
tochemical dissociation to give ethane(g) and CO(g). If, in an experimental run, the
total radiant energy was 200 J and 9.00 × 10−5 moles of acetone dissociated, what
was the quantum yield?
Problem 21.8
Part of chemical folklore is that heating a reaction by 10 degrees doubles the rate.
If this is true, what is the activation energy for the temperature change from 298 to
308 K? Is this rule generally true?
P1: OTA/XYZ P2: ABC
P1: OTA/XYZ P2: ABC
ref JWBS043-Rogers September 15, 2010 1:42 Printer Name: Yet to Come
REFERENCES
Allinger, N. L. 2010. Molecular Structures, John Wiley & Sons, Hoboken NJ.
Allinger, N. L.; Chen, K.; Lii, J.-H. 1996. J. Comp. Chem. 17, 642–668.
Atkins, P. W. 1998. Physical Chemistry, 6th ed., Freeman, New York.
Atkins, P. W.; Friedman, R. S. 1997. Molecular Quantum Mechanics, 3rd ed., Oxford,
New York.
Barrante, J. R. 1998. Applied Mathematics for Physical Chemistry, 2nd ed., Prentice-Hall,
Englewood Cliffs, NJ.
Becke, A. D. 1933. J. Chem. Phys. 98, 1372–1377, 5648–5652.
Barrow, G. M. 1996. Physical Chemistry, 6th ed., WCB/McGraw-Hill, New York.
Bohr, N. 1913. Philos. Mag. 26, 1–25.
Born, M.; Heisenberg, W.; Jordan, P. 1926. Z. Phys. 35, 557–615.
Cheng, M. F.; Li, W.-K. 2003. J. Phys. Chem. A 107, 5492–5498.
Cohen, N.; Benson, S. W. 1993. Chem. Rev. 93, 2419–2438.
CRC Handbook of Chemistry and Physics, 2008–2009, 89th ed., Lide, D. R.; ed., CRC Press,
Boca Raton, FL.
Curtiss, L. A., et al. 1999. J. Chem. Phys. 110, 4703–4709.
De Broglie, L. 1924. Thesis, Paris. 1926. Ann. Phys. 3, 22. See also Pauling, L.; Wilson,
E. B. 1935. Introduction to Quantum Mechanics, McGraw-Hill, New York. Reprinted
1963, Dover, New York.
Ebbing, D. D.; Gammon. S. D. 1999. General Chemistry, 6th ed., Houghton Mifflin, Boston.
Eğe, S. N. 1994. Organic Chemistry; Structure and Reactivity, 3rd ed., D. C. Heath, Lexington,
MA.

Copyright
361
P1: OTA/XYZ P2: ABC
362 REFERENCES
Eyring, H. 1935. J. Chem. Phys. 3, 107–115.

Fock, V. 1930. Z. Phys. 61, 126–148.
Gibson, D. G., et al. 2010. Science 329, 52–56.
Hammes, G. G., 2007. Physical Chemistry for the Biological Sciences, John Wiley & Sons,
Hoboken, NJ.
Hartree, D. R. 1928. Proc. Cambridge Philos. Soc. 24, 89–111, 111–132.
Hehre, W. 2006. Computational Chemistry, in Engel T., Quantum Chemistry and Spectroscopy,
Pearson-Benjamin Cummings, New York.
Hehre W. J., et al. 1970. J. Am. Chem. Soc. 92, 4766–4801.
Heisenberg, W. 1925. Z. Phys. 33, 879–893.
Heitler, W.; London, F. 1927. Z. Phys. 44, 455–472.
Henry, W. 1803. Philsophical Transactions of the Royal Society.
Houston, P. L. 2006. Chemical Kinetics and Reaction Dynamics, Dover Publications, Mineola,
New York.
Irikura, K. K. Essential Statistical Thermodynamics in Computational Thermochemistry, in
Irikura, K. K.; Furrip, D. J., eds. 1998. Computational Thermochemistry, American Chem-
ical Society, Washington, D.C.
Kistiakowsky, G. B., et al. 1935. J Am. Chem. Soc. 57, 65–75.
Klotz, I. M.; Rosenberg, R. M. 2008. Chemical Thermodynamics. Basic Theory and Methods,
7th ed., J. Wiley Interscience, New York.
Kondipudi, D.; Prigogine, I. 1998. Modern Thermodynamics: From Heat Engines to Dissipative
Structures, John Wiley & Sons, New York.
Laidler, K. J.; Meiser, J. H. 1999. Physical Chemistry, 3rd ed., Houghton Mifflin, Boston.
Lee, C. et al. Phys. Rev., 37, 785–789.
Levine, I. N. 2000. Quantum Chemistry, 5th ed., Prentice-Hall, New York.
Levine, I. 2000. Physical Chemistry 6th ed. McGraw-Hill Inc., New York.
Lewis, G. N.; Randall, M., revised by Pitzer K. S.; Brewer, L. 1961. Thermodynamics,
2nd ed., McGraw-Hill, New York.
Maczek, A. 1998. Statistical Thermodynamics, Oxford Science, New York.
McQuarrie, D. A. 1983. Quantum Chemistry, University Science Books, Mill Valley, CA.
McQuarrie, D. H.; Simon, J. D. 1997. Physical Chemistry: A Molecular Approach, University
Science Books, Sausalito, CA.
Metiu, H. 2006. Physical Chemistry, Taylor and Francis, New York.
Moeller C.; Plesset, M. S. 1934. Phys. Rev. 46, 618–622. See also Cramer, C. J. 2004. Compu-
tational Chemistry 2nd ed., John Wiley & Sons, Hoboken NJ.
Nash, L. K. 2006. Elements of Statistical Thermodynamics, 2nd ed. Dover Publications,
Mineola, New York.
Nevins, N.; Chen, K.; Allinger, N. L. 1996a. J. Comp. Chem. 17, 669–694.
Nevins, N.; Lii, J.-H.; Allinger, N. L. 1996b. J. Comp. Chem. 17, 695–729.
Pauling, L.; Wilson, E. B. 1935. Introduction to Quantum Mechanics, McGraw-Hill, New
York. Reprinted 1963, Dover Publications, New York.
Petersson, G. A. 1998. Complete basis set thermochemistry and kinetics, in Irikura, K. K.;
Furrip, D. J., eds., Computational Thermochemistry. American Chemical Society, Wash-
ington, D.C.
P1: OTA/XYZ P2: ABC
REFERENCES 363
Petersson et al. 1991. J. Chem. Phys., 94, 6081–6090, 6091–6101.

Pitzer K. S.; Brewer, L. Revision of Lewis, G. N.; Randall, M. 1961. Thermodynamics,
2nd ed., McGraw-Hill, New York.
Planck, M. 1901. Ann. Phys. 4, 553–563; 717–727. See also Kuhn, T. S. 1978. Black Body
Theory and the Quantum Discontinuity 1894–1912, The University of Chicago Press,
Chicago; and Oxford, New York.
Pople, J. A. 1999. Nobel lectures. Rev. Mod. Phys. 71, 1267–1274.
Rioux, F. 1987. Eur. J. Phys. 8, 297–299.
Rogers, D. W., et al. 2003. Org. Lett. 5, 2373–2375.
Rogers, D. W. 2005. Einstein’s Other Theory. The Planck-Bose-Einstein Theory of Heat
Capacity. Princeton University Press, NJ.
Rogers, D. W.; Zavitsas, A. A.; Matsunaga, N. 2005. J. Phys. Chem. A 109, 9169–9173.
Rogers, D. W.; Matsunaga, N.; Zavitsas, A. A. 2006. J. Org. Chem. 71, 2214–2219.
Rogers, D. W.; Zavitsas, A. A.; Matsunaga, N., 2009. J. Phys. Chem. A, 113, 12049–12055.
Rosenberg, R. M.; Peticolas, W. L. 2004. J. Chem. Ed. 81, 1647–1652.
Schrödinger, E. 1925. Z. Phys. 33, 879.
Schrödinger, E. 1926. Ann. Phys. 79, 361, Ann. Phys. 79, 734.
Silbey, R. J. et al. 2005. Physical Chemistry 4th ed., John Wiley & Sons, Hoboken NJ.
Slater, J. C. 1930. Phys. Rev. 35, 210–211.
Steiner, E. 1996. The Chemistry Maths Book, Oxford Science Publications, NY.
Streitwieser Jr., A. 1961. Molecular Orbital Theory, John Wiley & Sons, Hoboken NJ.
Treptow, R. S. 2010. J. Chem. Ed. 87, 168–171.
Uhlenbeck G. E.; Goudsmit, S. 1925. Naturwissenschaften 13, 953–954.
Webbook.nist.gov
Zavitsas, A. A. 2001. J. Chem. Ed. 78, 417–419.
Zavitsas, A. A.; Rogers, D. W.; Matsunaga, N. 2008. J. Phys. Chem. A 112, 5734–5741.
Zewail, A. H. 1994. Femptochemistry: Ultrafast Dynamics of the Chemical Bond, Vols. 1
and 2, World Scientific, Singapore.
P1: OTA/XYZ P2: ABC
P1: OTA/XYZ P2: ABC
ans JWBS043-Rogers September 15, 2010 1:22 Printer Name: Yet to Come
ANSWERS TO SELECTED
ODD-NUMBERED PROBLEMS
CHAPTER 1
1.1 8.65 m3 ; 1.3 98 g; 1.7 92.7 H2 ; 1.9 0.0798 m3 , 0.0168 kg; 1.11 0.032 kg;
1.13 3.7 kJ mol−1 ; 1.15 515 m s−1 .
CHAPTER 2
2.1 V = 17.9; 2.3 dm6 bar mol−2 , dm3 mol−2 ; 2.5 b = Vc /3; 2.7 V = 0.55 dm3 ;
2.9 z = 0.944; 2.13 pV = 24.7881 − 0.0100 p + 5.184 × 10−5 p 2 + 1.4977 ×
10−7 p 3 .
CHAPTER 3
3.1 0.667; 3.3 V = 3922 J, v = 19.8 m s−1 ; 3.5 5.46 J K−1 ; 3.11 25.08 ◦ C.
CHAPTER 4
4.1 2.39×10−4 K; 4.3 −5152 kJ mol−1 ; 4.5 −1268 kJ mol−1 ; 4.7 14.1 kJ mol−1 .

Copyright
365
P1: OTA/XYZ P2: ABC
366 ANSWERS TO SELECTED ODD-NUMBERED PROBLEMS
CHAPTER 5
5.1 31.2 kJ mol−1 ; 5.3 11.5 J; 5.5 negligible; 5.7 1.24 J K−1 mol−1 ; 5.9a 22.0 J mol−1 ;
5.9b 109.1 J mol−1 .
CHAPTER 6
6.3 88 J K−1 mol−1 ; 6.5 79.2 J K−1 mol−1 ; 6.7 46.5 J K−1 mol−1 ; 6.9 −893 kJ mol−1 .
CHAPTER 7
7.3 sol H298 > 0; 7.5 3.46 J K−1 mol−1 ; 7.7 179 kJ mol−1 , 1.93.
CHAPTER 8
8.1 62.3, 37.7%, >60/40; 8.3 8.15 pm; 8.5 3.07 pm; 8.7 2.61exp33, 1.28exp33.
CHAPTER 9
9.1a 2; 9.1b 3; 9.5 279.6 K.
CHAPTER 10
10.1 0.693, 3.46 × 10−2 min−1 ; 10.3 384 crabs; 10.5 4515 ± 86 y;
10.9 2.1 dm3 mol−1 s−1 .
CHAPTER 11
11.1a 4.84 unit unspecified; 11.1b 6 unit unspecified; 11.1c sphere: 24.1% smaller;
11.3 126 nm; 11.5 6.
CHAPTER 12
12.1a 10%; 12.1b 0.1711 mol; 12.1c 1.901 molal; 12.1d you don’t have enough
information; 12.1e 93.34 cm3 ; 12.1f 1.833 molar; 12.5 100.7; 12.7 50.
P1: OTA/XYZ P2: ABC
ANSWERS TO SELECTED ODD-NUMBERED PROBLEMS 367
CHAPTER 13
13.1 –54.8 kJ mol−1 ; 13.3 59 µg; 13.5a 45.9 S; 13.5b 368 S; 13.7 55.35 mols dm3 ,
1.81×10−6 , 1.01 ×10−14 , 1.00 ×10−7 mol dm3 .
CHAPTER 14

14.1a copper; 14.1b 0.740 volts; 14.3a K sp ∼= 5 × 10−13 ; 14.3b Ag+ = 7 × 10 −7
+3

There are no units because this is a ratio to a standard state; 14.5 Fe (aq) ≈
10−29 mol dm3 (With these approximations, we can take this as an approximate
concentration.) 14.7 pH = 5.6 (slightly acidic); 14.9 –2.12 volts.
CHAPTER 15
⎛ ⎞
30 36 42.3
−19 ⎜ ⎟
15.1 4.57 × 10 Hz, 3.03 × 10
14
J; 15.5 x · x = ⎝ 66 81 96.6 ⎠ ,
102.7 126.8 151.81
⎛ −15
⎞
1 −3.553 × 10 0
⎜ ⎟
x · x−1 = ⎝ 3.553 × 10−15 1 0⎠
4.441 × 10−15 −4.621 × 10−15 1
CHAPTER 16
∞ √ 2π x
xe−x d x = 0; 16.3 φ(x) = e−λx ; 16.5 (x) =
2
16.1 y(x) = 2 sin ;
−∞ λ
√ 2π x
16.9 (x) = 2 sin .
λ
CHAPTER 17
17.1 –2, 1, −1, 0, sin2 θ − cos2 θ ; 17.3a 22.22 × 10−6 cm; 17.3b 22.22 × 10−8 m;
17.3c 22.22 × 10−15 nm; 17.3d 22.22 × 10−18 pm, 22.22 × 10−14 A, 1.349 ×
1
1015 Hz, 8.938 × 10−19 J; 17.5 1.00 × 10−28 J; 17.7 √ ; 17.9 E He = −77.45,
n
about 2%.
CHAPTER 18
18.1 6.63 × 10−20 J; 18.5 2142.5 cm−1 , 1856 N m−1 ; 18.9 2.82 × 10−29 C
m = 8.47 D; 18.11 15.9 × 10−24 cm3 mol−1 , 1.44 D; 18.13

P1: OTA/XYZ P2: ABC
368 ANSWERS TO SELECTED ODD-NUMBERED PROBLEMS
CHAPTER 19
19.1a −168 kJ mol−1 ; 19.1b −7 kJ mol−1 ; 19.3 105.1◦ . The experimental value
is 104.5◦ ; 19.7 −31.7 kcal mol−1 = −132.7 kJ mol−1 . The experimental value is
−32.60 ± 0.05 kcal mol−1 = 136.4 ± 0.2 kJ mol−1 .
CHAPTER 20
20.3 the degeneracy pattern is 1,2,1,2, unlike the 1,2,2,1 pattern in the more common
Huckel model; 20.7a UHF=−2.8874388; 20.7b UHF=−4.8708157; 20.7c The
molecule does not exist; 20.7d The molecule is lower in energy than its atoms by
30.3 kcal mol−1 = 127 kJ mol−1 . LiH exists and is used in organic chemistry as a
reducing agent.
CHAPTER 21
21.1 search for PhCH2 CH2 Ph in the product mixture; 21.3 100.2 kcal mol−1 and
74.2 kcal mol−1 ; 21.7 0.17.
P1: OSO
ind JWBS043-Rogers September 24, 2010 12:57 Printer Name: Yet to Come
INDEX
1s STO and a Gaussian approximation, ADP, 107

324 Ag ions, 212, 216–217
6-31G basis functions, 332 Alkane conformers, 308
Allinger MM method, 308
Ab Initio methods, 324 Allinger, 64
Absorption wavelengths, 293 Ammonia, 82, 195, 201
of conjugated polyalkenes, 293 AMP, 107
Acetic acid, 69 Amperostat, 215–216
Acetone, 196, 200 Amplitude function, 238, 249
Acetonitrile, 159 Anode, 223
Acidity constant, 209, 215 Antioxidants, 347
Activated complex, 350, 355 Antisymmetric wave functions, 239
enthalpy level diagram, 157 Antisymmetrized orbitals, 240
structure, 355 Arclength, 40
Activation, 351 Arrhenius law, 158, 349
activation barrier, 156, 157, 351 Atomic units, 243
Activation energy, 344 ATP, 103–104, 107
Activity, 97, 186 Aufbau principle, 280
activity coefficient, 98, 186 Autooxidation, 346
Additive extrapolations, 339 Avogadro, 3
the G3(MP2) scripted method, Azeotrope, 138
339
Additive methods, 305 Bar (unit), 13
Adenosine 5 -triphosphate ATP, 105 BASIC, 83
Adiabatic work, 50 Basis function, 245, 320

Copyright
369
P1: OSO
370 INDEX
Battery, 221 Cathode, 217, 223

Benzene, 342 Cell diagrams, 223
Benzoic acid, 68 Cell reaction, 223
Bimembrane, 167 Central field approximation, 239
Bimolecularity, 155 Cerium, 234
Binary solutions, 184 Chain reaction, 153, 345–346
Body-centered cubic cell, 180 initiation, 346
Bohr, 236, 264 propagation, 346
Boiling point elevation, 188 termination, 346
Boltzmann 101, 108 Charge distribution, 241
Boltzmann constant, 9 Charles, 1
Boltzmann distribution of molecular speeds, Chemical kinetics, 144
158 Chemical potential, 87
Bond additivity, 306 nonideal systems, 100, 102
Bond breaking, 350 pressure dependence, 87
Bond dissociation energies (BDE), 348, temperature dependence, 88–89
355, 358 Chemical reactions, 78, 79
Bond energy, 65 Chloroform, 200
Bond enthalpies, 306 Clapeyron equation, 128
Bond length, 292, 299 Classical harmonic oscillator, 288
Bond making, 350 Classical molecular modeling, 305
Bond moments, 296 Clausius, 71
dichloroethene isomers, 296 Clausius–Clapeyron equation, 128
Bond strength, 292, 299 Close packing, 171
Born probability postulate, 252 Closed system, 41
Born, 237 Coexistance curve, 127
Born-Haber cycle, 178–179 Colligative properties, 187, 194
Bosons and Fermions, 278 Collision theory, 158
Boyle, 1 Combined gas law, 2, 12
Bragg’s law, 173 Combined or scripted methods, 338
Bravais crystal systems, 176 Combining enthalpy and entropy, 84
Bravais lattices, 176 Combustion of C(gr) and CO(g), 59
Breaking degeneracy, 259 Combustion, 92
free energy of, 92
Cadmium half-cell, 232 of methane, 92
Calculated ionization potentials, Components, 124
281 Components, phases, and degrees of
Calomel electrode, 229 freedom, 124
Calorie, 54 Compound phase diagrams, 137
Calorimeter, 104 Compressibility factor, 20, 21, 28
Calorimetry, 56 calculated from the van der Waals
Capacitance, 294 constants, 28
Capillary rise, 167 Computational chemical script, 338
Carbon dioxide, 34, 86 Computational geometries, 177
free energy of formation, 87 Computational kinetics, 159
Carbon disulfide, 91, 92 Computational statistical thermodynamics,
Carnot, 71 119
Carnot’s engine, 72 GAUSSIAN 03 C
, 119
Cartesian space, 330 Computational thermochemistry, 100
P1: OSO
INDEX 371
Computed partition functions, 120 Deviations from Raoult’s law, 185

for molecular and atomic sodium, 120 Dewar, 323
Computer files, 273 Diatomic molecules, 290
Concentration cells, 225 Dielectric constant, 294
Concentration units, 184 Dielectric, 294
Conductance cell, 206 Diethyl ether, 196
Conductance, 205, 211 Different paths of transformations, 38
Conductivity, 203, 205 Differential scanning calorimetry, 66
Conservation laws, 35 Digression on “Space”, 9
exact differentials, 35 Dipole moments, 294, 301, 303
Conservtive systems, 238 Dipole, 294
Contracted Gaussian orbitals CTGO, 328 Discontinuous energy levels, 8
Conversion of a Liquid to Its Vapor Without Disorder, 74, 109, 183
Boiling, 25 Distortion polarization, 295
Copper electrode, 221 Distribution, 110
Correlation energy, 337 DSC, 66
Corresponding states, 23, 27
Coulomb (unit), 224 Early atomic theory, 235
Coulombs, 212 Early quantum theory, 236
Coulometry, 203 Ebullioscopic constant, 190
Covalent bond radii, 176 Effective molecular weight (molar mass), 28
Critical density, 26 Effective nuclear charge, 240
Critical point, 25, 127 Effective quantum number, 282
Critical temperature, 24 Egyptian tomb, 163
Cross term, 310 Eigenfunction postulate, 252
Crystal forces in MM, 312 Eigenvalues and eigenvectors, 250
Crystals, 170 Einstein, 247
CTGO, 328 Einstein’s law, 344
Cubic (crystal) cell, 175 Electrical potential, 203, 204
Cubic equation, 28, 29 Electrical work, 224
Cubic real gas behavior, 34 Electrochemical cells, 220
Curve-fitting software, 21 Electrochemical potential, 221
Electrolyte, 203
Dalton’s Law, 5 Electromagnetic spectrum, 235
Daniell cell, 220 Electron spin resonance, 299
de Broglie, 235 Electronic promotion in alkenes, 293
Debye-Huckel law, 213 Electronic spectra, 292
Degeneracy, 257 Empirical equations, 18
and equilibrium, 109 Empirical gas laws, 1
Degenerate two-level equilibrium, 110 Energy and enthalpy, 44
Degree of dissociation, 208 Energy equations, 308, 309
Degree of ionization, 208 Energy of formation, 57
Degrees of freedom, 4, 124 indirect, 59
Density functional theory (DFT), 338 Engine, 72
Density ρ curves, 26 Ensemble, 6, 7
for liquid and gaseous oxygen, 26 Enthalpies of formation of n-alkanes, 306
Dependent variables, 4 Enthalpies of reaction, 60
Detailed balance, 152 Enthalpy level diagram for an activated
Determinantal wave functions, 240 complex, 157
P1: OSO
372 INDEX
Enthalpy of activation, 351 Failure of perpetual motion, 39

Enthalpy of formation, 57, 64 Faraday (unit), 211
from classical mechanics, 64 Faraday, 205
indirect, 59 Faraday’s laws, 211
of hydrogenation, 61, 65–66 Favorable and unfavorable orbital overlap,
Enthalpy of fusion, 92, 200 263
Enthalpy of vaporization, 129 First law of thermodynamics, 35–36
Enthalpy, 305 First-order radioactive decay, 146
Entropy, 71, 99, 113 First-order rate laws, 144–145
and probability, 113 Flash photolysis, 154
and the second Law, 71 Fluid, 25
from the van’t Hoff equation, 99 Fluorescence, 160–161, 345
Entropy changes, 74 mechanism of, 346
chemical reactions, 78 Force constants and parameters, 308
expansion, 75 Force fields, 308, 310
heat transfer, 77 Frankenstein, 220
heating and expansion, 75 Free electron model, 292
heating, 74 Free energies of formation, 85
mixing, 77 Free energy and equilibrium, 102
of transition, 75 in biochemical systems, 102
Entropy of activation, 354 Free energy, 87
Entropy, enthalpy, and Gibbs free energy of reaction, 87
changes for mixing, 183 of formation, 90
Equations of state, 4 of methane, 90
Equilibrium, 93, 94, 108 Free solvent, 194
and free energy, 99 Freezing point of water, 195
Equilibrium constant K eq , 73, 97 Fugacity, 97
as a quotient of quotients, 88 coefficient, 98
statistical approach, 117 Fundamental thermodynamic identities,
sodium vapor, 117 86
variation with temperature, 97 Further basis set improvements, 336
Equilibrium, 108
Equilibrium, 94, 108 G3(MP2), 338
general formulation, 94 Gain medium, 348
Euler’s reciprocity relationship, 36, Galvani, 220
76 GAMESS, 330, 340
Eutectic mixture, 137 calculation of methane, 340
Exact differentials, 36 control lines, 341
Excluded volume, 29 file for the hydrogen molecule, 330
Expectation value, 17 Gas constant, 13
Experimental determination of molecular Gaussian approximation, 272
structure, 287 Gaussian basis set, 324
Experimental determination of the rate Gaussian distribution, 7
equation, 154 in hyperspace, 10
Extensive variable, 6 Gaussian distribution, 7
Extent of reaction, 96 Gaussian gen input, 273
Extrapolation to E◦ , 226, 227 for the hydrogen atom, 273
Eyring reaction path, 350 energies, 273
Eyring theory of reaction rates, 349 GAUSSIAN C
input for hydrogen, 330
P1: OSO
INDEX 373
Gaussian probability density distributions, Hexa-1,3,5,-triene, 81

10 Hey, Let’s Make a Perpetual Motion
Gaussian, 69 Machine!, 38
GAUSSIAN C
, 69, 244 HF, 329
HF calculation of E, 273 Higher molecular orbital calculations, 322
Gibbs chemical potential, 133, 192 Hittorf cell, 211, 216
Gibbs free energy, 84, 112 HOMO, 285, 292
and the partition function, 112 Hooke’s law, 289
of water, 126 Hooke’s potential well, 289
Gibbs thermodynamic function, 76 Huckel, 271
Gibbs–Helmholtz equation, 89 Hydrogen atom, 244, 258, 325
Glass electrode, 230 Hydrogen emission spectrum, 236
Glucose, 69 Hydrogen half-cell, 222
Gordon, 323 Hydrogen molecule ion, 319
Graham’s Law of effusion, 6 bonding and antibonding orbitals, 321,
molecular weight (molar mass) 322
determination, 6 Hydrogen molecule, 329
Group additivity, 62, 306 Hydrogen spectrum, 235
hydrogen atom counting, 63 Hydrogen-like atoms, 260
n-octane, 69 for an electron in the 3s orbital, 262
GUI, 310 higher exact solutions, 260
radial probability density, 262
Half cells, 221 roots of the radial 3s wave functions, 261
Half-cell potentials, 222 the first six wave functions for hydrogen,
Half-time, 146 261
Hamiltonian operator, 237, 251 the first three s wave functions for
Harmonic oscillator, 287 hydrogen (simplified form), 261
Hartree (unit), 272 Hyper space, 3-, 4-, etc., 10
Hartree and Fock, 239
Hartree equations, 242 Ideal gas laws, 1
Hartree independent electron method, 240 Ideal solutions, 182
Hartree product, 263 Independent variables, 4, 125–126
Hartree-Fock limit, 336 coexistence curves, 125
Heat capacities and γ for selected real Influence of temperature on rate, 156
gases, 48 Intensive variable, 6
Heat capacity, 44, 48, 70 Intermolecular attraction at an air–water
as a function of temperature, 50, 168 interface, 166
of an ideal gas, 48 Internal and Cartesian coordinates, 334
Heat capacity of liquids and solids, 168 Interplanar (crystal) distances, 173
Heat death, 73 Invasive crabs, 163
Heat transfer, 77 Inversion of optical activity, 352
Heisenberg, 237 Inversion temperature as a function of
Heitler and London, 238, 239 pressure, 47
Helium, 81, 243, 274 Iodine vapor, 122, 123
Hemolytic cleavage, 348 Ion activities, 213
Henry’s law, 186 Ion mobilities, 209, 210, 211
Hermitian operators, 269 Ion speed, 210
Hess’s law, 61 Ionic character, 297
Heteroatoms: oxygen, 332 Ionic strength, 214
P1: OSO
374 INDEX
Ionization potential IP, 274 Maxwell-Boltzmann distribution, 7, 8, 13

of helium, 284 Mean activity coefficient of HCl, 232
Ion-permeable membrane, 204 Mean ionic activity coefficients, 229
Irreversible change, 72 Membrane potential, 203
Isobar, 26 Methane, 331
Isodesmic reactions, 349, 356 Methanol, 332
Isolated system, 41 finding the enthalpy of formation,
Isotherm, 24 334–336
of a van der Waals gas, 24 Methyl isocyanide, 159
Iterative procedure, 241 Microstates, 112
Microwave spectroscopy, 292
Joule and Joule–Thomson experiments, 46 Mixing, 77
Joule (unit), 42 MKS system, 2, 213
Joule, 46, 50 MM series, 309
input file for water, 314
KCl, 62 thermodynamic properties, 311
Kinetic energy, 7 MM4, 315
Kohlrausch’s law, 206 geometry output for water, 315
input geometry for methane, 316
L-(dihydroxyphenyl)alanine, 218 output geometry for methane, 316
Laplace transforms, 149 Mobility, 211
Lasers, 348 Models, 307
Lattice energies, 177 Moeller-Plesset energies, 337
Law of corresponding states, another view, Molality, 184, 197, 198
27 Molar conductivity, 206
LCAO, 320 at infinite dilution, 207
L-Dopa, 218 Molar heat capacities, 48
Le Chatelier’s principle, 99 Molar mass of a nonideal gas, 28
Lead, 83 Molar volume, 16, 33, 129
Length of an Arc, 40 Molarity, 198
Limiting law, 214 Mole fraction, 6, 184
Line integrals, 39, 52–53 Molecular enthalpies of formation, 58
Line spectrum, 235, 246 Molecular mechanics, 309
Linear and angular momentum vectors, 283 Molecular modeling, 310
Linear combination, 270 Molecular orbitals, 328
Linear combinations of atomic orbitals Molecular quantum chemistry, 238
LCAO, 322 Molecular variational method, 318
Lipid peroxidation, 346 Molecular velocities, 12
Liquids and solids, 165 Molecular weight (molar mass)
Liquid–vapor coexistence curve, 127 determination, 6
Logarithmic decay of a radioactive element, Moles, 3
147 Momentum, 15
LUMO, 285, 292 Moving boundary method, 210
Magnetic resonance imaging (MRI), 299 NaCl conductivity, 219

Making ATP, the cell’s power supply, 103 Naphthalene, 68
Many-electron atomic systems, 263 NDDO and related approximations, 323
MathcadC
, 52, 246 Nernst equation, 224
Maxwell, 108 Newton’s second law, 14
P1: OSO
INDEX 375
Nitric oxide, 358 Partial pressure, 5, 6, 186, 195

Nitrogen, 21, 222 Particle in a cubic box, 255
Nitrogen dioxide, 148 the first excited state, 257
Nitrogen tetroxide, 105 Particle in a one-dimensional box, 253
NMR spectrum of ethanol, 299 energy, 265
Nodes, 254 fundamentals and overtones, 254
three wave functions, 254 Partition function, 10, 108, 111, 114, 118
Born probability densities, 255 for different modes of motion, 116
Nonideal binary solutions, 131 Gibbs free energy, 112
Nonideal systems, chemical potential, 100 rotational, 116
Nonvolatile solute, 189 vibrational, 115
Normalization, 257 Path independence, 38
NSTP new standard temperature and Path, 38
pressure, 16 Pauling, 239
Nuclear magnetic resonance (NMR), 297, Perturbation, 337
304 pH meter, 226
Phase change, 84
Occupation number, 11 water, 84
Open system, 41 Phase diagram, 127
Operator, 237 water, 128
Optical activity, 352 Phase plane, 127
Optical inversion, 352 Phase rule, 124
Optimizing the Gaussian function, 272 Phases, 124
Orbitals, 257 Phosphorescence, 345
Order and molecularity, 155 mechanism of, 346
Orientation polarization, 295 Photochemistry, 344
Orthogonality, 262, 264 Photoinduction, 345
and overlap, 262 Physical equilibrium, 126
Osmotic presure, 191, 192, 202 Pitzer, 92
Oxygen bomb calorimetry, 68 Polar molecule, 297
Ozone depletion, 347 Polarizability, 295, 303
Ozone hole, 156, 347 Polarized basis functions, 332
Ozone, 155 Pople, 323
Population inversion, 348
Packing fraction, 174 Population, 8
Parabola, 52 Post-Hartree–Fock calculations, 336
Parabolic potential well, 290 Potential drop, 204
of the harmonic oscillator, 290 Potential energy surface, 350
Parallel plate capacitor, 294 Potential enrgy, 42
Parametric curve fit, 19 Pressure fraction, 5
Partial GAUSSIAN G3(MP2) output, 339 Primitives, 326
Partial ionization, 208 Principle of detailed balance, 152
Partial MM4 enthalpy output for ethane, Priogine, 42
311 Probability “shells”, 258
Partial molar thermodynamic functions, 133 Probability, 12, 113
the general case, 133 entropy and, 113
Partial molar volume, 129 Probability density, 8, 12, 241
as a slope of V vs n, 131 Probability factor, 159
the general case, 132 Pythagoras, 40
P1: OSO
376 INDEX
QCISDT, 337 SCF variational ionization potential for

Quadratic least squares, 21, 22 helium, 275
Quadratic real gas behavior, 32 Schrödinger equation, 64, 250
Quantum mechanical systems, 251 GAMESS, 64
Quantum molecular modeling, 318 Schrödinger, 237
Quantum numbers, 249, 277 Second law of thermodynamics, 39
Quantum rigid rotor, 290 Second virial coefficient, 22
energy levels, 291 as a function of temperature, 22
Quantum yields, 345 Second-order reactions, 147–148
experimental quantum yields, 345 Secular determinant, 268, 270
Self consistent field (SCF), 243, 275,
Radial equation, 258 277
Radial node of the 2p atomic orbital, Semiempirical equation, 19
262 Semiempirical methods, 323
Radiative decay, 161 Semipermiable membrane, 191
Radioactive decay, 144 Separable equations, 256
Raoult’s law, 183 Sequential reactions, 150
Rate controlling step, 155 Shell structure, 331
Ratio γ , 50–51, 55 Shelley, 220
Reaction coordinate, 73, 351 Shielding constant, 280
Reaction diagram, 88 Shielding, 240
Reaction mechanisms, 154 Silver, standard entropy, 80, 81
Reaction order, 149 Silver–Silver chloride electrode, 227, 231,
Reaction molecularity, 155 234
Real gases, 18, 30 Sin(x), sin(2x), and sin(3x) [0, π ], 249
Real solutions, 185 Single-component phase diagram, 128
Reciprocity relationship, 36 Slater determinants, 279, 286
Redox, 234 Slater orbital, 276
Reduced variables, 27 Slater, 282
Reduction potentials, 223 Slater’s rules, 282
Reference states, 307 Slater-type orbitals (STO), 279, 282
Regression wizard, 31 oxygen, 284
Reservoir, 72 nitrogen, 285
Resistivity, 205 Sodium vapor, 117
Resonance frequency, 291 Software, curve fitting, 21
Reversible expansion, 43 Solubility product constant, 228
of an ideal gas, 43 Solubility product, 228, 231
Reversible processes, 42 Solute, 184
path independence, 42 Solution chemistry, 182
Reversible reactions, 151 Solvent, 184
RHF, 329 Space, 9
Roothaan, 322 hyperspace, 3-, 4-, etc., 10
Rotational symmetry, 171 spherical velocity space, 12
Route section, 329 velocity space, 12
Speed, 7
Salt bridge, 223 Spherical polar coordinates, 259
Saturated calomel electrode SCE, 229, Spin, 278
233 Spin–orbit coupling, 283
Scalar, 251 Spin–spin coupling, 298
SCF Energies of atoms and ions, 281 Split valence basis sets, 331
P1: OSO
INDEX 377
Spontaneous processes, 77 Thermal protein denaturation, 67

chemical reactions, 78 Thermal reaction, 67
expansion, 75 Thermal wavelength, 118, 120, 123
heating and expansion, 75 electronic, 116
heating, 74 dissociation, 117
mixing, 77 translational, 116, 118
Stability products, 228 Thermochemical cycle, 60
Standard hydrogen electrode SHE, 233 Thermochemistry, 56
Standard state, 103 Thermodynamic cycles, 37
Standard states, 58 Thermodynamic functions, 37, 41
State function, 237 Thermodynamic property, 41
State functions (variables), 41, 45 Thermodynamic states and systems, 41
State of a System, 19 Thermodynamics of simple systems, 35
State vector, 241 the first law, 35
State, 19 Third law, 78
Statistical approach to thermodynamics, Tie line, 137, 139
108 Time’s arrow, 73
Statistical thermodynamic functions Tocopherols, 347
(Irikura, 1998), 114 Total volume of an ideal binary solution,
Statistical thermodynamic functions, 130
113 Transition states, 313
Statistical thermodynamics, 108 Transition, 42
Steady-state approximation, 153 Translational levels, 119
Steady-state equilibrium, 353 Translational motion, 118
Stewart, 323 quantum number, 118
STO-2G of hydrogen, 325 thermal wavelength, 118
basis set, 325, 326 Transport number, 209, 211
STO-3G calculations on methanol, 333 Triple point, 127
Stoichiometric coefficient, 96 benzene, 143
Stored parameters, 328 Trouton’s rule, 91
STO-2G basis for the 1s orbital of Two-component phase diagrams, 134
hydrogen, 326–328 Type 1, 135
Strainless molecule, 307 Type II, 136
Stretching a two-dimensional membrane, Type III, 137
166 Two-level equilibrium, 109, 111
Strong electrolytes, 208
Structure, 307
UHF, 329
Subcritical fluids, 25
Uncertainty principle, 238
Sucrose, 199
Unit cell, 174, 180
Sulfur dioxide, 82
Units, 2
Sum-over-states, 10
Universal gas constant, 13, 17
Supercritical fluid, 25
Universe, 41, 73
Surface tension, 165
Surroundings, 41
Symmetry, 340 Valence bond theory, 239
System, 19, 41 van der Waals equation, 18
van’t Hoff equation, 98
Temperature dependence of the Gibbs free Vapor pressure, 184, 187, 196
energy, 88 depression, 190
Ternary phase diagrams, 138 Raoult’s law, 184
P1: OSO
378 INDEX
Variation of enthalpy of formation with T , Water, 84, 125, 128

65 dissociation constant, 125
Variation of the equilibrium constant with enthalpy of vaporization, 139–140
temperature, 97 phase diagram, 128
Variational method, 243, 267 Wave equations, 249
Variational treatment of the hydrogen atom, two dimensions, 250
271 three dimensions, 250
the energy spectrum, 271 Wave function vector, 251
Vector space, 268, 320 Wave mechanics of simple systems, 248
Vector, 241 Wave motion, 248
Velocities of ideal gas particles, 8 amplitude, 249
Velocity space, 11, 12 fundamental, 248
Vibration-rotation band, 300, 302 overtone, 248
Vibration–rotation spectrum of CO, 302 Weak electrolytes, 207, 208
Virial equation, 19 Work, 43
Virtual orbitals, 337 adiabatic, 50
Viscosity of liquids, 169
laminar flow, 169
x-Component, 13
Poiseulle equation, 170
X-ray crystallography, 172
viscosity coefficient, 170
X-ray diffraction, 173
viscous drag, 169
x-Space, 7
Visualization, 310
the ethane molecule via PCModel 8.0 C
,
310 Z = f ( p) Curve, 23
Vitamin E, 347 for the same gas at two different
Volta, 220 temperatures, 23
Voltage, 203, 221 for two different gases, 23
Volume behavior of a nonideal binary Zero point energy, 289
solution, 131 Zinc electrode, 221
Volume increase (or decrease) upon adding z-Matrix format, 328, 329
solute, 130 z-Matrix input file for methane, 340

Buku Kimia Fisika PDF

Uploaded by

Copyright:

Available Formats

Buku Kimia Fisika PDF

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

Buku Kimia Fisika PDF

Uploaded by

Copyright:

Available Formats

P1: OTA/XYZ P2: ABC

fm JWBS043-Rogers October 8, 2010 21:3 Printer Name: Yet to Come

A JOHN WILEY & SONS, INC., PUBLICATION

CONCISE PHYSICAL CHEMISTRY

A JOHN WILEY & SONS, INC., PUBLICATION

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

1.10 The Sum-Over-States or Partition Function, 10

2 Real Gases: Empirical Equations 18

Table 2.1 Observed Real Gas Behavior from 10 to 100 bar

3 The Thermodynamics of Simple Systems 35

5 Entropy and the Second Law 71

6 The Gibbs Free Energy 84

6.3 Some Fundamental Thermodynamic Identities, 86

8 A Statistical Approach to Thermodynamics 108

8.8 The Equilibrium Constant: A Statistical Approach, 117

9 The Phase Rule 124

Example 9.2, 140

10 Chemical Kinetics 144

11 Liquids and Solids 165

11.2 Heat Capacity of Liquids and Solids, 168

12 Solution Chemistry 182

12.6 Vapor Pressure, 187

13 Coulometry and Conductivity 203

14 Electrochemical Cells 220

15 Early Quantum Theory: A Summary 235

16 Wave Mechanics of Simple Systems 248

Figure 16.11 Favorable spx and Unfavorable spz

17 The Variational Method: Atoms 267

18 Experimental Determination of Molecular Structure 287

18.2 The Hooke’s Law Potential Well, 289

19 Classical Molecular Modeling 305

19.3 Structure, 307

20 Quantum Molecular Modeling 318

File 20.2 (Output) The STO-2G Basis Set Written as a

21 Photochemistry and the Theory of Chemical Reactions 344

experimental data by wave mechanics was the hydrogen molecule. In a lifetime, I

Worcester Polytechnic Institute Ilie Fishtik

In the seventeenth and eighteenth centuries, thoughtful people, influenced by the

1.1 EMPIRICAL GAS LAWS

The constants k1 and k2 can be approximated simply by averaging a series of experi-

Concise Physical Chemistry, by Donald W. Rogers

2 IDEAL GAS LAWS

1.1.1 The Combined Gas Law

For the number of moles of a gas, n, we have

The constant R is called the universal gas constant.

gives the unit of R as J K−1 mol−1 . Experiment revealed that

R = 8.314 J K−1 mol−1 = 0.08206 L atm K−1 mol−1

1.2 THE MOLE

4 IDEAL GAS LAWS

1.3 EQUATIONS OF STATE

1.4 DALTON’S LAW

At constant temperature and pressure, by Avogadro’s principle, the volume of an

This principle holds regardless of the nature of the particles:

6 IDEAL GAS LAWS

1.5 THE MOLE FRACTION

1.6 EXTENSIVE AND INTENSIVE VARIABLES

Mass m is an extensive variable. Density ρ is an intensive variable. If you take twice

1.7 GRAHAM’S LAW OF EFFUSION

THE MAXWELL–BOLTZMANN DISTRIBUTION 7