M ain

/ ' g r o u p s 5 IS

1A , '^
1 J \^ >
H 2
1.00794
2A ?cS a THE UNIVERSITY Of ucEKbU svD ARY
3 4
1'
PHYSICAL SCIENCES &ENGINEERING LIBRARY
Li Be ^ -------------------------------- Transition m etals --
6.941 9.01218
11 12 / 3 4 5 6 7 8 9
Na Mg
22.98977 24.305
3B 4B 5B 6B 7B ^ ------- 8B
19 j 20 y 21 22 23 24 25 26 27
K Ca Sc Ti V Cr Mn Fe Co
39.0983 40.078 44.9559 47.88 50.9415 51.996 54.9380 55.847 58.9332
37 38 39 40 41 42 43 44 45
Rb Sr Y Zr Nb Mo Tc Ru Rh
85.4678 87.62 88.9059 91.224 92.9064 95.94 (98) 101.07 102.9055
55 56 57 72 73 74 75 76 77
Cs Ba *La Hf Ta w Re Os Ir
132.9054 137.33 138.9055 178.49 180.9479 183.85 186.207 190.2 192.22
87 88 89 104 105 106 107 108 109
Fr Ra ^Ac Rf Db Sg Bh Hs Mt
(223) 226.0254 227.0278 (261) (262) (263) (262) (265) (266)
58 59 60 61 62

*Lanthanide series Ce Pr Nd Pm Sm
140.12 140.9077 144.24 (145) 150.36
90 91 92 93 94
^Actinide series Th Pa u Np Pu
232.0381 231.0359 238.0289 237.048 (244)
 3 4 0 6 7 0 2 1 7 8 598 4

/ ---------------- M ain groups

18
8A
13 14 15 16 17 2
He
4.00260
3A 4A 5A 6A 7A
5 6 7 8 9 ^ 10
BA c
1 2 .0 1 1
N o F Ne
10.81 14.0067 15.9994 18.998403 20.1797
13 14 15 16 17^ 18
10 11 12 A1 Si ^ P S Cl Ar
26.98154 28.0855 30.97376 32.066 35.453 39.948
IB 2B
28 29 30 31 32 33 . 34 35 36
Ni/ CV Zn y Ga Ge As Se B r / Kr
58.69 63.546 65.39 69.72 72.61 74.9216 78.96 79.904 83.80
46 47 48 49 50 51 52 53 54
Pd Ag Cd In Sn Sb Te I Xe
106.42 107.8682 112.41114.82118.710 121.757127.60 126.9045 131.29
78 79 80 81 82 83 84 85 86
Pt Au Hg T1 P Bi Po At Rn
b/
195.08 196.9665 200.59 204.383 207.2 208.9804 (209) (2 1 0 ) (2 2 2 )
110 111 112 k

(269) (272)(277)

63 64 65 66 67 68 69 70 71
Eu Gd Tb Dy Ho Er Tm Yb Lu
151.96 157.25 158.9254 162.50 164.9304 167.26 168.9342 173.04174.967
95 96 97 98 99 100 101 102 103
Am Cm Bk Cf Es Fm Md No Lr
(243) (247) (247) (251) (252) (257) (258) (259) (260)
nirffl
TT
Principles and Applications of
Geochemistry
PRINCIPLES AND APPLICATIONS OF

GEOCHEMISTRY
A Comprehensive Textbook for Geology Students

Second Edition

GUNTER FAURE
T H E O H IO S TA TE U N IV E R S IT Y

Prentice Hall
Upper Saddle River, New Jersey 07458
L ib rary o f C ongress C ataloging -in -P ublication D a ta

Faure, Gunter.
Principles and applications of geochemistry : a comprehensive
textbook for geology students/G unter Faure.2nd ed.
p. cm.
Rev. ed. of: Principles and applications of inorganic
geochemistry. cl991.
Includes bibliographical references and index, ISBN
0-02-336450-5
1. Geochemistry I. Faure, Gunter. Principles and applications
of inorganic geochemistry. II. Title.
QE515.F28
551.9 dc21
1998
L 97-44563
.
C IP
O ( . QL/bOo?
THE UNIVERSITY OF QUEENSLAND LIBRASW
PSCE.

Executive Editor: Robert A. McConnin

Total Concept Coordinator: Kimberly P. Karpovich
Art Director: Jayne Conte
Cover Designer: Karen Salzbach
Manufacturing Manager: Trudy Pisciotti
Production Supervision/Composition: WestWords, Inc.

1998,1991 by Prentice-Hall, Inc. Simon

& Schuster/A Viacom Company Upper
Saddle River, New Jersey 07458

All rights reserved. No part of this book may be

reproduced, in any form or by any means, without
permission in writing from the publisher.

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

ISBN Q-DS-33b4SD-5

ISBN 0-02-33b450-5

Prentice-Hall International (UK) Limited, London

Prentice-Hall of Australia Pty. Limited, Sydney
Prentice-Hall Canada Inc., Toronto
Prentice-Hall Hispanoamericana, S.A., Mexico
Prentice-Hall of India Private Limited, New Delhi
Prentice-Hall of Japan, Inc., Tokyo
Simon & Schuster Asia Pte. Ltd., Singapore
Editora Prentice-Hall do Brasil, Ltda., Rio de Janeiro
For Terri
Contents
 Preface xiii 4.3 D ifferentiation of Igneous and
Sedim entary R ocks 49
4.4 DDiffferentiationofoftheHHydrosphere 5252
Part I 4.5 Summmary 56
Problem s 57
PLANET EARTH IN THE SOLAR R eferences 57
SYSTEM

1
What Is Geochemistry? 2 Part II
2
1.1 Early History
PRINCIPLES OF INORGANIC
1.2 G eochem istry in the U.S.S.R. 3 GEOCHEM ISTRY
1.3 V. M. G oldschm idt 4
1.4 M odern G eochem istry 5
R eferences 7
5
2 The Electronic Structure
of Atoms 60
In the Beginning 8
8
5.1 The A tom of Thom son and
2.1 The Big Bang
1 1
R utherford 60
2.2 Stellar Evolution
5.2 B o h rs Theory of the H ydrogen
2.3 Nucleosynthesis 13
2 0 A tom 61
2.4 Sum m ary
2 0 5.3 Em ission of X -rays 63
Problem s
2 0 5.4 Schrdingers M odel of the A tom 64
References
5.5 The A ufbau Principle 67
3
5.6 Sum m ary 69
Problem s 72
The Solar System 22 R eferences 72
2 2
3.1 Origin of the Solar System

3.2 O rigin of the E arthlike Planets 25

3.3 Satellites of the O uter Planets 26
3.4 Pictures of O ur Solar System 29
6
3.5 Summary 42 The Periodic Table and Atomic
References 42
Weights 74
4 6 .1
M endeleevs Periodic Table 74

Chemical Differentiation
6 .2
The M odern Periodic Table 74
of the Earth 43 6.3 Basic Principles of A tom ic Physics 77
4.1 Internal Structure of the E arth 43 6.4 A tom ic W eights 78
6.5 Sum m ary 81
4.2 The C ontinental Crust: M ajor Problem s 82
Elem ents 46 R eferences 82

VII

V III CO N TEN TS

7 9.5 pH Control of Dissociation

Chemical Bonds , Jonic Radii, Equilibria 123
9.6 Solubility of A m orphous Silica 124
and Crystals 83
83 9.7 Summary 127
7.1 Electron D onors Versus A cceptors Problem s 128
7.2 M easures of M etallic C haracter 84 R eferences 128
7.3 Bonding in M olecules 87
7.4 Ionic Crystals 89 10
7.5 Ionic Radii 91
Salts and Their Ions 130
7.6 Summary 97
Problem s R 97 10.1 Solubility of Salts 130
eferences 98 10.2 Hydrolysis 137
10.3 Activities and Concentrations 139
10.4 Solubility of Calcium Carbonate 142
8
10.5 Chemical W eathering 148
Ionic Substitution in 10.6 Transform ation of Potassium
Crystals 99 Feldspar to Kaolinite 149
8.1 G oldschm idts Rules of 10.7 Summary 152
Substitution 99 Problem s 153
8.2 Camouflage, C apture, and R eferences 153
1 0 0

Adm ission
8.3 Coupled Substitution: Key to the 11
, Thermodynamics 155
1 0 2
Feldspars
8.4 D istribution Coefficients and 103
1 1 .1
Definitions 155
G eotherm om eters
1 . 2 The First Law 157
8.5 G eochem ical Classification of 1

the Elem ents 105 11.3 Enthalpy 157

8 . 6 Summary 106 11.4 H eats of Reaction 159
Problem s R 108 11.5 H eat Capacity 160
eferences 108
1 1 .6 The Second Law 161
11.7 G ibbs Free Energy 163
1
1 .8
D erivation of the Law of Mass
A ction 165
Part III 11.9 Fugacity and Activity 166
1 1 .1 0
The vant H off Equation 168
A QUEO US GEOCHEM ISTRY AND 11.11 Solubility of Am orphous Silica
THE STABILITY OF MINERALS B etw een 0 and 100 C 168
1
1 .1 2
Summary 169
- 9 Problem s 170
References 171
Acids and Bases 110 1 1 0
12
9.1 Chem ical Reactions and E quilibria
9.2 The Law of Mass A ction
Mineral Stability Diagrams 172
1
1 2

9.3 Dissociation of W eak A cids

114 12.1 Chemical W eathering of Feldspars 172
and Bases
9.4 Solubility of Sparingly Soluble 12.2 Form ation of Zeolites 182
Bases 1 2 1
12.3 M agnesium Silicates 186
 CONTENTS ix
190 Problem s 271
12.4 Solubility D iagram s
12.5 Fugacity D iagram s 194 References 273
12.6 Sum m ary 198
Problem s R 199
eferences 199
Part IV
13
ISOTOPE GEOCHEM ISTRY AND
Clay Minerals 200
MIXING
13.1 Crystal Structure 2 0 0

13.2 Classification and Chem ical 16

Com position
Isotopic Geochronometers 276
2 0 2

13.3 Gibbs Free Energies of Form ation 208

2 1 1
13.4 Stability D iagram s 16.1 D ecay M odes 276
13.5 Colloidal Suspensions and Ion 16.2 Law of Radioactivity 281
Exchange 217 16.3 M ethods of D ating 284
2 2 1
13.6 D ating of Clay M inerals 16.4 Cosmogenic R adionuclides 292
2 2 2
13.7 Sum m ary 16.5 Summary 297
Problem s R 223 Problem s 297
eferences 224 References 299
17
14
Oxidation-Reduction 0 Isotope Fractionation 301
Reactions 226 17.1 Principles of Isotope Fractionation 301
14.1 Balancing E quations of 17.2 M athem atical R elations 304
O xidation -R eduction R eactions 226 17.3 Isotope Fractionation in the
14.2 The E lectrom otive Series 229 H ydrosphere 306
14.3 The E m f of Electrochem ical Cells 234 17.4 Oxygen Isotope Com position of 308
14.4 Stability Limits of W ater in Terms of E 235 Calcite
h and pH 17.5 Oxygen and H ydrogen in Clay
14.5 Stability of Iron C om pounds 238 M inerals 311
14.6 Summary 249 17.6 G roundw ater and G eotherm al 313
Problem s R 251 Brines
eferences 252 17.7 Isotope Fractionation of Carbon 315
17.8 Isotope Com positions of Strontium
15 in C arbonate Rocks 317
Rates of Geochemical 17.9 Isotope Fractionation of Sulfur 319
17.10 Summary 323
Processes 253
Problem s 325
15.1 R ates of Chem ical R eactions 253 R eferences 325
15.2 Transport of M atter: A dvection 257
15.3 Transport of M atter: Diffusion 259 18
15.4 G row th of C oncretions D uring 263 Mixing and Dilution 328
Diagenesis
328
15.5 G row th of M onom ineralic Layers 268 18.1 Binary M ixtures
15.6 Sum m ary 271 18.2 Dilution 330
X CONTENTS
18.3 Evaporative C oncentration 331
20.3 Chemical Com position of M eteoric
18.4 Ternary M ixtures 332 Precipitation 375
18.5 Isotopic M ixtures of O ne E lem ent 335 20.4 Norm ative M inerals from W ater
18.6 Isotopic M ixtures of Two Elem ents 337 Compositions 376
18.7 Sum m ary 338 20.5 Evaporative C oncentration 383
Problem s 339 20.6 W ater Quality 392
References 340 20.7 Summary 395
Problem s 396
References 397

Part V
21
APPLICATIONS O F GEOCHEM ISTRY
Chemical Weathering of Mineral
TO THE SOLUTION OF GLOBAL
Deposits 400
PROBLEMS 2 1.1
M etallic M ineral D eposits 400
2 1
Oxidation of Iron Sulfides and the
.2
19
Role of Bacteria 401
Consequences of Chemical
21.3 E h -pH D iagram for C opper
Weathering 342 M inerals 403
19.1 Changes in Chem ical Com position 21.4 Supergene E nrichm ent of F e -C u
Sulfide D eposits 407
of Rocks 343
21.5 R eplacem ent of Pyrite by
19.2 N orm ative M ineral Com position
Chalcocite 408
of W eathering Products 345 2 1 .6
Oxidation of O re M inerals of
19.3 Susceptibility of M inerals to
O ther M etals 411
W eathering 347
21.7 Geochem ical Exploration 413
19.4 Form ation of Placer D eposits 351 2 1 .8
Production and C onsum ption of
19.5 Provenance D eterm ination by
M ineral Resources 417
Isotopic D ating 353
21.9 Summary 421
19.6 Form ation of Soils 354
Problem s 422
19.7 G eom icrobiology 358
R eferences 423
19.8 Food Production and Population
22
G row th 361
19.9 Summary 363
Problem s 364
R eferences 365 Geochemical Cycles 425
2 2 .1
The Principle of Mass Balance 426
2 2 .2
Mass Balance for M ajor Elem ents
in the Ocean 428
20 22.3 Mass Balance for Trace Elem ents
in the Ocean 431
Chemical Composition of Surface
22.4 The Cycles of C -H -O -N 433
Water 368 22.5 The Sulfur Cycle 443
20.1 Chem ical Analysis of W ater in 2 2 .6
Summary 446
Stream s 368 Problem s 447
20.2 Chem ical Com position of Stream s 370 R eferences 447
 CO N TEN TS
25
23
Chemistry of the Atmosphere 449 Effect of Environmental Lead on
450 Human Health 485
23.1 Structure and Composition
23.2 U ltraviolet R adiation 451 25.1 Isotope C om position of
452 Environm ental Lead
23.3 O zone in the Stratosphere 453 25.2 L ead in the E nvironm ent
23.4 O zone H ole over A ntarctica 25.3 L ead in Plants
23.5 Infrared R adiation and the G 455 25.4 L ead Poisoning of Cows and H
orses
reenhouse Effect 457 25.5 H um an B ones and Tissues
23.6 Prediction of the Future G lobal 458 25.6 Lead in the B ones of A ncient
459 Peoples
Climate
459 25.7 Sum m ary
23.7 Sum m ary Problem s R
Problem s R eferences
eferences

Appendix A
24 461 Compilations of Geochemical D ata
Environmental Geochemistry: 462
Disposal of Radioactive Waste 461 465 Appendix B
24.1 The Big Picture
468 Standard Gibbs Free E nergies ( G f)
24.2 H igh-Level N uclear W aste D isposal
24.3 G eological Disposal of Radioactive W 473 and Standard Enthalpies of
aste Form ation (H f )
24.4 G eochem istry of Plutonium 476
24.5 E h - p H Diagrams of N eptunium
478 Author Index 575
and Plutonium
24.6 A nalog Studies: The N atural R
481
eactors at Oklo, G abon
482
24.7 R eactor Accidents: Chernobyl, U
483 Subject Index 589
kraine
24.8 Sum m ary
Problem s R
eferences
Preface
This textbook is intended as an introduction to geo
chemistry for geology students in their senior year or in
their first year of graduate work. A t that time in their
education, students are ready to releam those principles
of chemistry that are especially applicable to the study of
geological processes. G eochem istry can enhance their
understanding of these processes and can teach them to
apply principles of chem istry to the solution of geolog
ical problem s. Geologists in virtually all branches of the
science can benefit from an introductory course in
geochem istry because it can help them to m ake
quantitative predictions about the o u t com e of chem
ical reactions that occur in the con text of m any
geological processes.
The subject m atter of this book is presented in five
parts. P art I: P lanet E arth in the Solar System (C hapters
1 through 4) presents the big picture starting w ith the Big
Bang, stellar evolu tion, nucleosynthesis, the solar
system, and the
geochem istry of the E arth . This p art also con
tains a brief history of geochem istry leading up
to a statem ent of its goals and stating the them e of this
book: Studies of geochem istry convert idle speculation
into understanding.
The second part, Principles of Inorganic
Geochemistry (Chapters 5 through 8 ), starts with the
electronic structure of atom s and dem on strates how w
ave mechanics and the Aufbau prin ciple lead directly to
the periodicity of the chemical properties of the elements.
Once the periodic table has been constructed, systematic
variations in bonding, ionic radii, and the structure of
crystals becom e apparent, thus preparing the way for a
discussion of ionic substitution based on the rules of G
oldschm idt and Ringwood. Part II concludes w ith
discussions of coupled substitution in feldspar,
distribution coefficients, and the geo chemical
classification of the elements.
Part III is entitled A queous Geochemistry and the
Stability off M inerals (Chapters 9 through 15).
These chapters explain the concept of chemical
equilibrium and use the Law of Mass A ction to study the
dissociation of weak acids and bases and to calculate the
solubility of am orphous silica as a function of the pH .
Next, the dissociation of salts, the hydrolysis of their
ions, and the necessity of using the activities of ions
rather than their molal concentrations in M ass A ction
problem s are pre sented. A ll of these concepts com e
into play in the carbonate equilibria involving calcite, its
ions in aqueous solution, and C 0 2 in the atm osphere.
However, the incongruent dissolution of micro-cline to
form koalinite plus ions leads to an impasse because the
value of the equilibrium con stant for that reaction has not
been m easured.
Therefore, C hapter 11 contains a guided tour of the
principles of therm odynam ics leading up to a derivation
of the Law of M ass A ction for reacting mixtures of ideal
gases at equilibrium . In addition, this chapter dem
onstrates that equilib rium constants at the standard tem
perature and pressure can be calculated from the Gibbs
free energy change of chem ical reactions. This skill is im
m ediately used to construct ion -activity dia grams for
alum inosilicate m inerals including zeo lites (C hapter
12) and clay m inerals (C hapter 13).
Finally, oxidation -reduction reactions are
introduced in C hapter 14, and the previously acquired
know ledge of therm odynam ics is used to explain some
basic principles of electrochem istry. These insights are
then used to define the stability limits of w ater on the E
arth and to con struct E h -p H diagram s for the oxides of
iron and their ions. C hapter 15 deals with the kinetics of
chemical reactions, and with the transport of ions by
diffusion and advection. These concepts are used to
model the grow th of concretions during diagenesis and
the form ation of m onom ineralic layers at the sedim ent-
w ater interface.
P art IV is entitled Isotope G eochem istry and M
ixing (C hapters 16 through 18). Following
X III
X iv PREFACE
the p resentation of the different m odes o f decay
and a statem ent of the Law of Radioactivity,
C hapter 16 continues with the derivation of the
geochronom etry equation based on the accum u lation of
stable radiogenic daughter nuclides. This equation is then
used to explain the principles and assum ptions of dating
by the K -A r, R b-Sr, Sm -N d, R e -O s, and U, T h -P b
m ethods. In addi
tion, the chapter contains descriptions of the
2 3 0 T h /238U m ethod of m easuring sedim entation
rates and of dating by the cosm ogenic radionu
clides 1 4 C, 1 0 Be, and 2 6 A1.
C hapter 17 is devoted to the fractionation of the
stable isotopes of H, O, C, and S in nature. This chapter
includes derivations of the basic m athe m atical
relationships used to in terp ret variations
of the isotope com positions of m eteoric water,
m arine calcite, clay m inerals, oilfield brines,
hydrocarbons, and sulfide minerals. In addition, C hapter
17 presents a discussion of the time-dependent variation
of the isotope com positions of Sr and S in m arine
carbonates of Phanerozoic age and the use of 8 7 S r/86Sr
ratios for dating Cenozoic m arine carbonates.
P art IV concludes with C hapter 18 on mixing and
dilution of w ater and sedim entary rocks that are m
ixtures of two or three com ponents having different
chemical compositions. The chapter con tinues with tw o
-com ponent mixtures containing elem ents having
distinctive isotope com positions (Sr, Pb, O, H , etc.). The
systematics of such mix tures are briefly illustrated using
oilfield brines form ed by mixing two com ponents having
differ en t ^ S r/^ S r ratios and Sr concentrations.
The fifth and last part of this book is entitled A
pplications of G eochem istry to the Solution of G lobal
Problem s (C hapters 19 through 25). This p art treats the
consequences of chemical weath ering of rocks, such as
the form ation of soils and their agricultural productivity
(C hapter 19); the chem ical com position of surface w
ater, its evolu tion by evaporative concentration, and its
conta m ination by natural and anthropogenic causes (C
hapter 20); and the w eathering of m etallic min eral
deposits based on E h - p H diagram s and including a
section on the im portance of bacteria in the oxidation of
iron sulfides (C hapter 21). The
resulting release of m etals into the environm ent is the
basis for geochem ical m ethods of prospecting. The
chapter ends with a discussion of mineral economics
including King H u b b erts prediction of the ultim ate
depletion of finite and nonrenew able resources such as
petroleum .
The m igration of ions released by weathering of
rocks and m etallic m ineral deposits on the con tinents
leads to a consideration of geochemical cycles in C hapter
22. A fter presenting the concept of mass balance, this
chapter contains summaries of the geochem ical cycles C
-H -O -N, including water. The links betw een the
reservoirs of these and other elem ents play an im portant
role in sta bilizing the chemical com position of the atm
os phere and the oceans.
C hapter 23 on the chem istry of the atm os phere
continues this train of thought by present ing the basic
facts about the structure and chemical com position of the
atm osphere. This inform ation is then used to explain the
form ation of ozone in the stratosphere and its destruction
by anthropogenic emissions of C l-bearing gases,
especially over A ntarctica. The chapter also con tains a
presentation of the radiation balance of the E arth and the
perturbation of this balance by emissions of gases that
enhance the absorption of infrared radiation in the atm
osphere. C hapter 23 ends with a long-range forecast of
the effects of global warm ing on the tim e scale of 1 0 4
years based on the w ork of W. S. Broecker.
The dangers of environm ental contam ination
are illustrated in C hapter 24 by consideration of
the problem s associated with high-level radioac
tive waste generated by nuclear reactors. The chapter
explains how this waste is generated and how it is (or
should be) processed prior to perm a nent storage in
underground repositories. The geo chemical properties of
the transuranium elements, which play an im portant role
in the long-term safety of such repositories, are briefly
discussed using E h -p H diagrams for neptunium and
plutoni um. Chapter 24 ends with a sum m ary of the
events and consequences of the explosion of a nuclear
reactor in Chernobyl, U kraine, on A pril 26,1986.
The final chapter sum m arizes the geochem istry of
Pb in the environm ent and its effect on
 PREFACE XV
M iami U niversity (O hio); D onald I. Siegel,
hum an health. Lead is dispersed primarily
through the atm osphere in the form of aerosol Syracuse University, R obert D. Shuster, U niv
particles. Some civilizations have suffered the con ersity of Nebraska, O m aha; and A . Russell Flegal, U
sequences of Pb toxicity because this m etal was used as a niversity of California, Santa Cruz. The second edition has
food additive and in household utensils. T he effects of Pb benefited significantly from com m ents by colleagues
poisoning on hum ans are subtle and difficult to identify who have pointed out errors of various kinds in the first
because all hum ans alive today are contam inated with Pb edition and suggested im provem ents in the presentation
released prim ari ly because of its use as an additive in of the subject m atter. I thank these individuals by nam e
gasoline. The in alphabetical order: W alther M . B arnard, G eorges
im pairm ent of m ental functions as well as the Beaudoin, Phillip Boger, D aniel M arcos Bonotto, Steven
toxic effects of Pb were tragically dem onstrated E . Bushnell, C hristopher Conaway, E than L. G rossm
by the m em bers of Sir John Franklins expedition to the an, G iehyeon Lee, Peter C. Lichtner, James O Neil, R
C anadian A rctic in 1845-47. isto Piispanen, R obert D. Shuster, and Terri L. Woods. I
The presentation of the subject m atter in this book also thank R obert A. M cConnin, Executive E ditor of
starts with basic principles and emphasizes quantitative m Prentice Hall, for his interest in this book, Elisabeth H . B
ethods of problem solving in order to gain better elfer for
understanding of natural phenom e na. A lthough this
book is intended to be an intro duction to geochemistry, her role in its production, and Susan M.
the references at the end of each chapter will perm it M cM ullen for drafting the diagram s of the first
students to pursue topics of their choice in the library and edition. I am grateful to B etty H eath for typing
thereby approach the state o f the art. In addition, the end- part of the m anuscript for the second edition.
of-chapter problem s enable students to test their The ultim ate purpose of this book is to
understanding of the principles. The book ends encourage students of the E arth to adopt the
quantitative approach to problem solving that is exem
with a com pilation of geochemical data in plified by geochem istry, and to m ake them aw are of the
A ppendix A and of therm odynam ic data for a im portant role E arth scientists m ust play in the effort to
large num ber of elem ents and their com pounds in A preserve our h ab itat on the E arth .
ppendix B. In addition, m ost chapters contain data tables
and diagram s that will m ake the book a useful source o f G U N T E R FA U R E
geochemical data. The diagram s have lengthy captions Postscript: A m anual with solutions to the
that should encourage browsing and will help users to
extract inform a tion from them w ithout necessarily
having to reread the entire chapter. end-of-chapter problem s is available to instruc
tors from the publisher by writing on school sta tionery to
This geochem istry text dem onstrates th at the the Geology E ditor, Prentice Hall, 1 Lake Street, U pper
principles and applications of geochem istry are internally Saddle River, NJ 07458.
consistent, useful for understanding the world around us,
and accessible to anyone willing to m ake th e effort. In
addition, this book m akes the point th at geochem ists
have an im portant task to perform in society by assisting
in the prudent use of natural resources and in the safe
disposal of the resulting waste products.

In conclusion, I thank the reviewers who read the m

anuscript of this book and m ade m any sug gestions
which helped me to im prove it: Jam es H. Crocket, M cM
aster University; W illiam J. G reen,
 I
PLANET EARTH IN THE
SOLAR SYSTEM

F uture generations may rem em ber the 20th century prim arily for the advances in tech nology that
have enabled us to begin the exploration of the solar system . O ur result ing aw areness of the celestial
environm ent of the E arth has raised questions about the origin and evolution of the solar system, of
the M ilky Way Galaxy, and of the U niverse at large. The start of an introductory geochem istry
course is an appropriate tim e to sat isfy our desire to know the origins of the world in which we live.
 1
What Is Geochemistry?
Geology began when early man first picked up a stone, considered its qualities,
and decided that it was better than the stone he already had. Good stones were
useful and they were collected, mined, and traded.

H ENRY FA U L AND C AROL FA U L (1983, p. 1)

G eochem istry is based on the urge to understand why
som e stones are good and how they form ed. It
involves applications of the principles of chem istry to the
solution of geological prob lems and therefore could not
develop until chem istry and geology had been
established as scientific disciplines. These applications
were practiced first during the 16th century in the mines
of E urope w here know ledge of m inerals and their
chemical com positions was used to rec ognize and follow
veins of ore. For exam ple, in 1574 Lazarus E rcker, the
su p erin ten d en t of m ines for King R udolf II of A
ustria, published a m anual in which he described ores
and outlined procedures for their analysis (E rcker, 1951).
How ever, several centuries passed before geo chem istry
in the present sense of th e word becam e established as
an integral com ponent of E arth and Planetary Science.
1.1 Early History
Two essential prerequisites for the grow th of geochem
istry were the discovery of the chemical elem ents and the
developm ent of sensitive and accurate m ethods for the
analysis of rocks and minerals. U ntil the end of the 16th
century m at
ter was thought to consist of earth, w ater, fire,
and air, whose basic qualities w ere described as
warm, cold, dry, and wet. The m etals gold, silver,
copper, iron, tin, mercury, and lead together with sulfur
and carbon had been known for thousands of years but
were not recognized as elem ents in the present sense of
the word. Antimony, arsenic, bism uth, and phosphorus
were studied by the alchemists during the Middle Ages.
The developm ent of analytical chemistry during
the 18th century resulted in the discovery of 46 chemical
elements between 1720 and 1850. A fter B unsen and
Kirchhoff invented the optical emission spectrograph, 30
additional elem ents w ere added from 1850 to 1925.The
transuranium elem ents and certain other radioactive
elem ents
w ere discovered later in the 2 0 th century
(Correns, 1969).
As the num ber of known chemical elem ents
increased during the 19th century, chem ists
becam e aware that they could be organized into groups
based on similarities of their chemical properties. These
tendencies becam e the founda tion of the periodic table
of the elem ents p ro posed independently by D. I. M
endeleev in 1869 and by J. L. M eyer in 1870 (Asimov,
1965). The cause for the periodicity of the chemical p ro
p er ties of the elem ents was not known then and was
recognized only after the internal structure of atom s was
worked out during the first 30 years of the 2 0 th century.
2
 1.2 GEOCHEMISTRY IN THE U.S.S.R. 3
The periodic table is a m anifestation of the
1.2 Geochemistry
relationship betw een the internal structure of
in the U.S.S.R.
atom s and the chem ical and physical properties
of the elem ents. It facilitates a rational explana G eochem istry becam e an im portant subject in
tion of the chem ical properties of the elem ents the U.S.S.R. because of the need to develop the
and of their observed distribution in nature. m ineral resources of the country. In 1932 Soviet
The roots of geochem istry lie mainly in the geologists began to analyze systematically col
19th century, b u t it acquired its present m ultifac lected soil samples using optical spectroscopy
eted com plexity during the 20th century. The word (Fersm an, 1939). In subsequent years the geolo
geochem istry was first used in 1838 by Christian gists of the Central G eological and Prospecting
Friedrich Schnbein, who was a chem istry profes Institute in Moscow carried out extensive m etal-
sor at the U niversity of B asel in Switzerland. The lom etric surveys based on soil samples in search
term was used again in 1908,70 years later, w hen of geochemical anomalies.
Frank W. C larke of the U.S. G eological Survey G eochem istry p rospered in th e U.S.S.R.
published the first edition of his book, The Data because of the popular support it received partly
o f Geochemistry. in response to the rem arkable work of A lexander
C larke was the Chief Chemist of the U.S. Fersm an (1883-1945). Fersm an studied m iner
Geological Survey from 1884 until 1925. H e started alogy at the U niversity of M oscow under V. I.
the tradition for excellence in the analysis of rocks Vernadsky, who took his students on m any field
and m inerals that has helped to m ake the Survey trips and encouraged them to regard m inerals as
one of the largest geochemical research centers in the products of chemical reactions. These were
the world. The first edition of C larkes book was radical departures from conventional m ineralogy
amazingly m odem in its approach to geochemistry. practiced at the turn of the century. Fersm an
H e opened with chapters on the chemical elements, graduated in 1907 and was elected to a professor
the atm osphere, lakes and rivers, and the ocean fol ship in m ineralogy in 1912 at the age of 27. H e
lowed by chapters on saline lakes and springs and em barked on an extraordinary career of teaching
on evaporite deposits. N ext he took up volcanic and research in geochem istry with em phasis on
gases, magma, rock-forming minerals, and igneous applications. H e traveled widely in the U.S.S.R.
rocks and then weathering, sedim entary rocks, and explored regions rarely visited before. His
m etam orphic rocks, and ore deposits. The book sustained interest in the exploration of the Kola
ends with chapters on fossil fuels and their origin. Peninsula in the north ultim ately led to the dis
That is pretty much how we would present the data covery of deposits of apatite and nickel ore. As a
of geochemistry today, but Clarke did it this way in result, a rem ote wilderness area was transform ed
1908! M ore than 15 years passed before the great into an im portant center of mining and industry.
Soviet geochemist V. I. Vernadsky published his Fersm an also explored central Asia and discov
geochemistry book in 1924. Clarke was the first ered a native sulfur deposit at K ara Kum that was
geochemist in the m odem sense of the word. actively mined for decades.
C larkes book w ent through five editions, the B etw een 1934 and 1939 Fersm an published
last of which appeared in 1924. It is now being a com prehensive four-volum e w ork entitled
updated and expanded by the U.S. G eological Geochemistry in which he applied the principles
Survey under the editorship of M ichael Fleischer. of physical chemistry to the distribution of the
H ow ever, the H an d b o o k o f Geochemistry, edited chemical elem ents in nature. Fersm an, above all,
by K. H . W edepohl from 1969 to 1978, is the m ost aroused public interest in geology and geochem
com plete source of geochem ical data currently istry by his inspiring lectures and popular books.
available. A lexander K. Tolstoy called him a poet of
4 WHAT IS GEOCHEMISTRY?
stones. Fersm an earned this accolade by writing books
intended for the layman: M ineralogy fo r
Everyone (1928), Recollections about a Stone (1940), M
y Travels, Stories about Precious Stones,
and Geochemistry fo r Everyone (1958). The last
two w ere published posthum ously. F ersm ans
conviction that geochem istry should serve his
country is reflected in his words: We cannot be m erely
idle adm irers of our vast country; we m ust actively help
reshape it and create a new life (Shcherbakov, 1958).
The scientific legacy of Fersm ans form er teacher,
Vladim ir Ivanovich Vernadsky (1863-1945), is similarly
impressive, but it has not been
appreciated in the w estern world. Vernadsky
em phasized the im portance of the activities of
living organisms in geological and geochemical processes.
In fact, he regarded living m atter as the m ost pow erful
geological force and w rote that the E a rth s crust
originated from the biosphere. These convictions about
the im portance of life on E arth were not widely accepted
by geologists in N orth A m erica, partly because
Vernadsky wrote in m any languages (Russian, German,
French, Czech, Serbo-Croatian, and Japanese). His most
im portant books (Vernadsky, 1924, 1929) were written in
French and published in Paris. However, Vernadsky was
also far ahead of his contem po raries by concluding that
the biosphere may have controlled the environm ent on
the surface of the E arth from the very beginning, long
before the existence of life in Early Precam brian time was
known (Lapo, 1986). Vernadskys views on this subject
are now reflected in the Gaia Hypothesis of Lovelock
(1979), made popular on American television in the
Planet E arth series by the Public Broadcasting System.
Vernadsky also iden tified the biosphere as the principal
transform er of solar energy into chemical energy. This
idea is now widely accepted. For exam ple, Cloud (1983)
described the biosphere as a huge metabolic device for
the capture, storage and transfer of energy. Fossil fuels
(coal, oil shale, petroleum , and natural gas) are now
regarded as deposits of solar energy collected by the
biosphere.
V ernadskys views on the im portance of the
biosphere (V ernadsky, 1945) are becom ing
ir : -asingly relevant as we attem pt to predict the
cilimntic consequences of the greenhouse effect caused
by the discharge of carbon dioxide and o ther gases into
the atm osphere. We are carrying omt an experiment on a
global scale the outcome o f which could be detrim ental
to life. The response of the biosphere to this perturbation
of
tlhe environm ent may be crucial to the ultimate
outcom e of this experim ent.
V. I. Vernadsky is revered in the U.S.S.R. as
oine of the giants of science in the 2 0 th cen
tury. His teachings still m otivate the geochem
ists working at the V ernadsky Institute 'of
G eochem istry and A nalytical C hem istry in
Moscow, which is am ong the forem ost geochem i cal
research institutes in the world.
1.3 V. M. Goldschmidt
T he Institute of G eochem istry at the University
o>f G ttingen in G erm any was m ade fam ous by
tlhe work of Victor M. G oldschm idt (1888-1947). G
oldschm idt earned his d o cto rate at the U niversity of
Oslo in 1911 with a study of contact nnetam orphism
based on the therm odynam ic p'hase rule. In the
following year Max von Laue discovered diffraction of x-
rays by crystals and thereby provided a m ethod for
determ ining the crystal structure of a m ineral and the
radii of the ions of which it is com posed. From 1922 to
1926 G oldschm idt and his associates at the University of
Oslo used x-ray diffraction to determ ine the
crystal structures of many m inerals (Goldschmidt,
1930). In 1930 G oldschm idt m oved to the
Unrversity of G ttingen, w here he studied the
distribution of the chemical elem ents using an
optical spectrograph. From this body of data he deduced
that the chemical com position of m iner als is determined
by the requirem ents of closest packing of ions. M
oreover, the substitution of the ions of a major elem ent by
the ions of a trace ele m ent depends on the similarity of
their radii and charges. These generalizations provided a
rational explanation for the observed distribution of the
elem ents in the minerals of the crust of the E arth .
Goldschmidt returned to Oslo in 1935 b u t
w as unable to continue his w ork after th e
 1.4 MODERN GEOCHEMISTRY 5

G erm an invasion of Norway in 1940. H e fled to N. L. Bowen and his colleagues, and by their suc cessors,
Sw eden in 1942 and ultim ately made his way to at that laboratory have becom e the foun dation of m
E ngland, w here he w orked at the M acaulay Institute for odern igneous petrology. B ow ens book, The Evolution o
Soil R esearch. His health had seri ously deteriorated as a f the Igneous Rocks, pub lished in 1928, turned igneous
consequence of im prison petrology from a preoccupation with the description and
m en t in co n cen tratio n cam ps in Norway. classifica tion of igneous rocks tow ard a concern for their
G oldschm idt returned to Oslo in 1946 but died origin and the geochemical differentiation of the E arth by
th ere prem aturely in 1947 at the age of 59. H e left m agm atic activity.
behind the m anuscript of a partially com pleted book
entitled G eochem istry, which was com plet ed by A lex In 1952 B rian M ason published the first edi
M uir and published posthum ously in 1954. A special tion of Principles o f Geochemistry, which was
com m em orative issue of A p p lie d widely used as a textbook at universities and helped to
G e o c h e m istry (1988) was devoted to establish geochem istry as a legitim ate com ponent of E
G oldschm idt on the 100th anniversary of his arth Science. In the next decade R obert M. Gar'rels and K
birthday. onrad B. K rauskopf used therm odynam ics and solution
G oldschm idts principal contribution to geo chem istry to determ ine the stability of m inerals and the
chem istry is the rational explanation of isomor- m obil ity of their ions at the surface of the E arth . They
phous substitution in crystals based on the com patibility trained the m odern generation of geochem ists by their
of the radii and charges of the ions. H e thereby achieved own research and through their popular
one of the m ajor goals of geo chemistry. Nevertheless,
the urge to know and to understand the distribution of the textbooks (G arrels, 1960; G arrels and Christ,
chemical ele m ents has continued to m otivate the work 1965; Krauskopf, 1967).
of sev G eochem ical prospecting, as we know it
eral outstan d in g geochem ists, notably L. H. today, was initiated in 1947 by the U.S. Geological
A h ren s in South A frica, S. R. N ockolds in Survey (Hawkes and Lakin, 1949) and was based on
England, S. R . Taylor in Australia, K. R ankam a in colorimetry. The techniques developed in the
Finland, K. H . W edepohl in Germany, A . R U nited States w ere used in 1953 for soil surveys
V inogradov in the U.S.S.R., D. M. Shaw in in the U nited Kingdom and in A frica (Webb,
Canada, and K. K. Turekian in the U nited States. 1953). Subsequently, the Geochemical Prospecting
R esearch C enter at the Im perial College of
Science and Technology in London becam e a cen ter of
activity in this aspect of geochemistry. Geochemical
1.4 Modern Geochemistry prospecting continues to be one of the most im portant
Beginning in the 1950s, geochem ists tu rn ed practical applications of geo chemistry and now relies on
increasingly to the study of chemical reactions and a wide range of sophisticated analytical techniques.
processes. The roots of this line of research can be found
in the work of A . Fersman, who used the concepts of E nvironm ental geochem istry has arisen recently
therm odynam ics to study the stability of m inerals in because of the need to m onitor the dis persion of m etals
their natural environment. In addi tion, J. H . vant H off and various organic com pounds that are introduced into
(1852-1911) studied the crys tallization of salts by the environm ent as anthropogenic contam inants. This
evaporating seaw ater in Berlin and, in 1904, the new applica tion of geochem istry to the welfare of hum
Geophysical Laboratory of the Carnegie Institution of W ankind is closely related to pollutant hydrogeology and
ashington was founded in the U nited States. The principal medical geochemistry. Geochem ical prospecting and
objec tive of the Geophysical Laboratory has been to environmental geochemistry are of param ount im
study the origin of igneous rocks and ore deposits by portance because they contribute to the con tinued well-
experim ental methods. The results achieved by being of the hum an species on Earth .
6 W HAT IS GEOCHEMISTRY?
Since about the middle of the 20th century
geochemistry has becom e highly diversified into many
subdivisions, among them inorganic geo chemistry,
organic geochemistry, isotope geochem istry, geochem
ical prospecting, m edical geo chemistry, aqueous
geochem istry, trace-elem ent geochemistry, and
cosmochemistry. Each of these subdivisions has its own
rapidly growing body of literature. However, the geochem
ists of the world have organized scientific societies that
transcend the boundaries of specialization and help to
unify the field. These societies also publish journals that
prom ote the free flow of inform ation and ideas am ong
their m em bers. For example, the Geochem ical Society
and the M eteoritical Society both sponsor Geochimica et
Cosmochimica Acta.
The E uropean A ssociation for G eochem istry has
adopted the journal Chemical Geology, and the
International A ssociation of G eochem istry and Cosm
ochem istry sponsors A p p lied Geochemistry. A m ong
other journals that publish papers dealing with geochem
istry are
Earth and Planetary Science Letters
Contributions to M ineralogy and Petrology
Journal o f Geochemical Exploration
E conom ic Geology
A m erican Journal o f Science
Nature
Science
Some journals are closely associated with certain regions,
for example,
G eokhim ia
D oklady (E arth
Science Section)
Lithology and Russia
M ineral Resources
Geochemistry
International
Geochemical Journal Japan
Chinese Journal P eoples Republic
o f Geochemistry of China
In fact, m ost E arth Science journals now publish papers
in which geochemical data or principles are used to
explain a geological process or to solve a particular
problem . The strength of geo chem istry lies in the fact
that most geological processes involve chemical reactions
in some sig nificant way. G eochem istry really is for
everyone, as Fersm an claimed in his popular book.
W hat then is geochemistry? There is no easy answ
er to this question because geochemists con cern them
selves with a wide range of natural phe nom ena and use
many different techniques to study them . According to
Fersman, Geochemistry studies the history of chemical
elements in the
E a rth s crust and their behavior under different
therm odynam ic and physicochem ical natural
conditions. G oldschm idts definition of geo
chem istry was G eochem istry is concerned with the
laws governing the distribution of the chem i cal elem ents
and their isotopes throughout the E a rth (Correns,
1969). We could go on and on. G eochem istry is a highly
diversified and con stantly evolving subject that can be
described in m any different ways.
For our purposes the m ajor goals of geo chem ists
are
1. To know the distribution of the chemical elem
ents in the E arth and in the solar system .
2. To discover the causes for the observed chemical
composition of terrestrial and extraterrestrial
materials.
3. To study chemical reactions on the surface of the
Earth, in its interior, and in the solar system
around us.
4. To assemble this inform ation into geo chemical
cycles and to learn how these cycles have
operated in the geologic past and how they may
be altered in the future.
These are the lofty goals of a scientific disci pline. T
heir intrinsic m erit is apparent. How ever, geochem ists
practicing their craft today also have an obligation to hum
ankind to assist in the devel
 REFERENCES 7

Em ery and J. M. H unt (1974, p. 586) put it this way:

Studies of geochem istry ... convert idle speculation
into ... understanding ... .T h a t will be the them e of
this book. We can no longer afford idle speculation in the
E arth Sciences. We m ust have understanding . G
eochem istry can show us how to achieve it.

O
WEN

opm ent and m anagem ent of natural resources, to m

onitor the quality of the environm ent both locally and on
a global scale, and to warn hum an ity against dangerous
practices that may threaten the quality of life in the future.

W ith so much at stake it is not easy to express

succinctly w hat geochemistry is all about. K. O.
References

A S IM O V , I., 1965. A Short
History o f Chemistry.
Doubleday,
Garden City, NY, 263 pp.
, N. L ., 1928. The
B Language
Evolution o f the Igneous
Rocks.
Princeton University
Press, Princeton, NJ, 34
pp.
C L A R K E , F. W., 1908. The Data
o f Geochemistry. U.S. Geol.
Surv. Bull. 330. (The fifth
edition in 1924 is
U.S.G.S. Bull.
770 .)
C L O U D , P., 1983. The
biosphere. Sei. Amer.,
249:132-144.
C O R R E N S , C. W., 1969. The
discovery of the chemical
ele
ments. The history of
geochemistry. Definitions
of geo
chemistry. In K. H.
Wedepohl (Ed.),
Handbook
of
Geochemistry, vol. 1,1-11.
Springer-Verlag, Berlin,
442 pp.
E M E R Y , K. O., and J. M. H U N
T , 1974. Summary of
Black Sea investigations.
In E. T. Degens and D. A.
Ross (Eds.), The Black
Seageology, chemistry,
and biology, 575-590.
Amer. Assoc. Petrol. Geol.,
Mem. 20,633 pp.
E R C K E R , L., 1951. Treatise on
Ores and Assaying. From the
German Edition of 1580.
A Grnhaidt Sisco and
C. Stanley Smith, trans.
and ann. University of
Chicago Press.
F A U L , H., and C. F A U L , 1983.
It Began with a Stone. Wiley,
New York, 270 pp.
F e rs m a n , A. Y E ., 1939.
Geochemical and
mineralogical meth ods of
prospecting. Chapter IV
in Special Methods o f
Prospecting. Akad. Nauk
S.S.R., Moscow.
Translated by L.
Hartsock and A. P.
Pierce, U.S. Geol. Surv. 62:321-348.
Circ. 127,1952 . W E D E P O H L , K . H. (Ed.), 1969. Handbook o f
G A R R E L S , R. M., and C. L. C H
F e rs m a n , A., 1958. vol. I. Springer-Verlag,
, 1965. Solutions,
Geochemistry for
RIST
Berlin. Vol. II (1970).
Minerals and Equilibria.
Everyone. Foreign
Harper & Row, New York
Publishing
(later Freeman
House, Moscow, 454 pp.
and Cooper, San
G A R R E L S , R. M., 1960.
Francisco), 450 pp.
Mineral Equilibria. Harper &
Row, G O L D S C H M ID T , V. M., 1930.
Geochemische
New York.
Verteilungsgesetze
und kosmische
Hufigkeit
der
Elemente.
Naturwissenschaften,
18:999-1013.
H A W K E S , H. E., and H. W. L A
K I N , 1949. Vestigial zinc in
surface residuum
associated with primary
zinc ore in east
Tennessee. Econ. Geol.,
44:286-295.
K RAU S KO PF , K . B., 1967 .
Introduction to
Geochemistry.
McGraw-Hill, New York.
L A P O , A. V., 1986. V. I.
Vernadskys ideas on the
leading role of life in the
generation of the Earths
crust. Earth Sei. Hist.,
5:124-127.
L O V E L O C K , J. E., 1979. Gaia. A
New Look at Life on Earth.
Oxford University Press,
Oxford, England, 158 pp.
M A S O N , B 1952. Principles o
f Geochemistry. Wiley, New
York.
S H C H E R BAK O V , D., 1958.
Foreword. In A.
Fersman,
Geochemistry for
Everyone. Translated by
D. A. Myshne,
Foreign Languages
Publishing House,
Moscow, 453 pp.
V E R N A D S K Y , V. I., 1924. La
Geochimie. Alcan, Paris, 404
pp.
V E R N A D S K Y , V. I., 1929. La
Biosphere. Alcan, Paris, 232
pp.
Vernadsky, V. I., 1945. The
biosphere and the noosphere.
Amer. Scientist, 33:1-12.
Webb, J. S., 1953. A review of American progress in geo
chemical prospecting. Inst.
 2
In the Beginning
The urge to trace the history of the universe back to its beginnings is irresistible.

C ertain questions about our existence on E arth are so
fundam ental that they have been incorpo rated into
religious mythologies. These questions not only concern
the origin of the E arth and the evolution of life but also
extend to the origin of the universe and to the nature of
space and time.
D id the universe have a beginning and will it
ever end? W hat existed before the universe
form ed? D oes the universe have limits and what exists
beyond those limits? It is proper to raise these questions
at the beginning of a geochem istry course because they
are within the scope of cosmochemistry.
2.1 The Big Bang
The universe started like a bubble in a stream . At first it
was not there, and suddenly it form ed and
expanded rapidly as though it w ere exploding
(G ott, 1982). Science has its share of practical
jokers who im m ediately referred to the start of the
expansion of the universe as the Big Bang (Gamow,
1952). From the very beginning the universe had all of
the mass and energy it con tains today. As a result, its
pressure and tem per ature, say 1CT3 2 sec after the Big
Bang, w ere so high that m atter existed in its m ost
fundam ental form as quark soup. As the universe
expanded and cooled, the quarks com bined to form more
fam iliar nuclear particles th at ultim ately becam e
organized into nuclei of hydrogen and helium .
ST E V E N W E IN B E R G (1977, p. 1)
Form ation of atomic nuclei began about
13.8 sec after the Big Bang when the tem pera
ture of the universe had decreased to 3 X 109 K. This
process continued for about 30 min, but did not go
beyond helium because the nuclear reac tions could not
bridge a gap in the stabilities of the nuclei of lithium,
beryllium, and boron. A t that time the universe was an
intensely hot and rapidly expanding fireball.
Some 700,000 years later, when the tem per ature
had decreased to about 3 X 103 K, elec trons becam e
attached to the nuclei of hydrogen and helium . M atter
and radiation were thereby separated from each other,
and the universe becam e transparent to light.
Subsequently, m at ter began to be organized into stars,
galaxies, and galactic clusters as the universe continued
to expand to the present time (W einberg, 1977).
But how do we know all this? The answ er is that
the expansion of the universe can be seen in the red
shift of spectral lines of light em itted by distant
galaxies, and it can be heard as the cosmic microwave
radiation, which is the rem nant of the fireball, that still
fills the universe. In addition, the properties of the
universe im m edi ately after the Big Bang were similar
to those of atom ic nuclei. Therefore, a very fruitful
collabo ration has developed among nuclear physicists
and cosmologists that has enabled them to reconstruct
the history of the universe back to about 1CT3 2 sec after
the Big Bang. These studies
8

have shown that the forces we recognize at low tem
perature are, at least in p art, unified at extrem ely high
tem peratures and densities. There is hope that a G rand
Unified Theory (G U T ) will eventually em erge th at
may perm it us to approach even closer to understanding
the start of the universe.
W hat about the future? Will the universe continue to
expand forever? The answer is that the future of the
universe can be predicted only if we know the total am
ount of m atter it contains. The m atter th at is detectable
at the present tim e is not sufficient to perm it gravity to
overcom e the expansion. If expansion continues, the
universe will becom e colder and em ptier with no
prospect of an end. However, a large fraction of the mass
of the universe is hidden from view in the form of gas and
dust in interstellar and intergalactic space, and in the
bodies of stars that no longer em it light. In addition, we
still cannot rule out the pos sibility that neutrinos have
mass even when they are at rest. If the mass of the
universe is sufficient to slow the expansion and ultim
ately to reverse it, then the universe will eventually
contract until it disappears again in the stream of time.
Since the universe had a beginning and is
still expanding, it cannot be infinite in size.
How ever, the edge of the universe cannot be
seen with telescopes because it takes too long for the light
to reach us. As the universe expands, space expands with
it. In other words, it seems to be im possible to exceed the
physical limits of the universe. We are trapped in our
expanding bu b ble. If other universes exist, we cannot
com m uni cate with them .
Now that we have seen the big picture, let us review
certain events in the history of the stan dard m odel of
cosmology to show that progress in
Science is som etim es accidental.
In 1929 the Am erican astronom er Edw in
H ubble reported that eighteen galaxies in the
Virgo cluster are receding from E arth at different rates th
at increase with their distances from Earth. H e calculated
the recessional velocities of these galaxies by m eans of
the D oppler effect from observed increases of the
wavelengths of
2.1 THE BIG BANG 9
characteristic spectral lines of light they emit. This red
shift is related to the recessional veloc ity by an equation
derived in 1842 by Johann Christian D oppler in Prague:
where A' is the wavelength of a spectral line of light em
itted by a moving source, A is the wave length of the same
line em itted by a stationary source, c is the velocity of
light, and v is the reces sional velocity. H ubbles estim
ates of the dis tances to the galaxies were based on the p
roper
ties of the C epheid Variables studied previously
by H . S. Leavitt and H. Shapley at H arvard
University. The C epheid Variables are bright stars in the
constellation C epheus whose period of variation depends
on their absolute luminosity, which is the total radiant
energy em itted by an astronom ical body. H ubble found
such variable stars in the galaxies he was studying and
deter m ined their absolute luminosities from their peri
ods. The intensity of light em itted by a star decreases as
the square of the distance increases.
Therefore, the distance to a star can be deter
m ined from a com parison of its absolute and its
apparent luminosity, w here the latter is defined as the
radiant pow er received by the telescope per square
centim eter. In this way, H ubble deter m ined the
recessional velocities and distances of the galaxies in the
Virgo cluster and expressed their relationship as:
v=Hd (2.2)
where v is the recessional velocity in km /sec, d is the
distance in 106 light years, and H is the H ubble constant
(H ubble, 1936).
The H ubble constant can be used to place a limit on
the age of the universe. If two objects are moving apart
with velocity v, the tim e (t) required for them to becom e
separated by a distance d is:
The initial results indicated that the H ubble con stant had
a value of 170 k m /sec/1 0 6 light years,
10 IN THE BEGINNING
which corresponds to an expansion tim e of less than 2 X
109 years. This result was very awkward because age
determ inations based on radioactivi ty had established
that the E arth is older than this date. Eventually W alter B
aade discovered an error in the calibration of the C epheid
Variables, and the value of the H ubble constant was
revised (B aade, 1968). The presently accepted value is
15 k m /sec/1 0 6 light years, which indicates an
expansion time for the universe of less than
20 X 109 years. This date is com patible with inde
p endent estim ates of its age based on considera tion of
nucleosynthesis and the evolution of stars. By com bining
all three m ethods H ainebach et al.
(1978) refined th e age of the universe to
(14.5 1.0) X 109 years.
The Big Bang theory of cosmology was not
accepted for m any years for a variety of reasons.
The turning point came in 1964 w hen A rno A.
Penzias and R o b ert W. W ilson discovered a of blackbody radiation at a particular tem perature
m icrow ave background rad iatio n th at corre
sponds to a blackbody tem perature of about 3 K. The
discovery of this radiation was accidental,
even though its existence had been predicted
tw enty years earlier by G eorge G am ow and his
colleagues R alph A . A lpher and R obert Herm an.
Because they w ere unaw are of G am ow s work, Penzias
and W ilson w ere skeptical about the phenom enon they
had discovered and took great care to elim inate all
extraneous sources of the background radiation. For
example, they noticed th a t two pigeons had been nesting
in the throat of the antenna they were using at H olm del,
New Jersey. The pigeons w ere caught and taken to a
distant location, but prom ptly returned. Therefore, they w
ere caught again and dealt with by more decisive
means. The pigeons had also coated the antenna with a
white dielectric m aterial, which was carefully removed.
However, the intensity of the background radiation rem
ained constant and independent of time in the course of a
year.
W ord of this phenom enon reached a group of
astrophysicists at nearby Princeton University who w ere
working on models of the early history of the universe
under the guidance of R obert H. Dicke. Eventually,
Penzias called Dicke, and it
was agreed that they would publish two com pan ion
letters in the Astrophysical Journal. Penzias and Wilson
announced the discovery, and Dicke and his colleagues
explained the cosmological significance of the microwave
background radia tion (Penzias and Wilson, 1965). In
1978 Penzias and Wilson shared the N obel Prize in
physics for their discovery.
The radiation discovered by Penzias and
Wilson is a rem nant of the radiation that filled
the universe for about 700,000 years w hen its
tem perature was greater than about 3000 K.
During this early period, m atter consisted of a
m ixture of nuclear particles and photons in ther
mal equilibrium with each other. U nder these
conditions the energy of radiation at a specific
w avelength is inversely p ro p o rtio n al to the
absolute tem perature. A ccording to an equation derived
by Max Planck at the start of the 20th century, the energy
increases rapidly with increasing wavelength to a m axim
um and then decreases at longer wavelengths. R adiation
in therm al equilibrium with m atter has the same
properties as radiation inside a black box with opaque
walls. Therefore, the energy distribution of radiation in the
early universe is related to the wavelength and to the
absolute tem perature by Plancks equation . The
wavelength near which most of the energy of blackbody
radiation is con centrated (Amax) is approxim ately equal
to:
where Amax is m easured in centim eters and T is in
Kelvins (W einberg, 1977).
The original m easurem ent of Penzias and Wilson
was at a wavelength of 7.35 cm, which is much greater
than the typical w avelength of radi
ation at 3 K. Since 1965, m any additional m ea
surem ents at different wavelengths have con
firm ed th at the cosmic background radiation
does fit Plancks form ula for blackbody radiation . The
characteristic tem perature of this radiation is about 3 K,
indicating th at the typical w avelength of photons has
increased by a factor of about 1 0 0 0

because of the expansion of the universe since its tem
perature was 3000 K (Weinberg, 1977).
2.2 Stellar Evolution
M atter in the universe is organized into a hier archy of
heavenly bodies listed below in order of decreasing size.
clusters of galaxies comets
galaxies asteroids
stars, pulsars, and m eteoroids
black holes dust particles
planets molecules
satellites atom s of H and H e
O n a subatom ic scale, space between stars and galaxies
is filled with cosmic rays (energetic nuclear particles) and
photons (light).
Stars are the basic units in the hierarchy of heavenly
bodies within which m atter continues to evolve by
nuclear reactions. M any billions of stars are grouped
together to form a galaxy, and large num bers of such
galaxies are associated into galac tic clusters. Stars may
have stellar companions or they may have orbiting
planets, including ghostly comets that flare briefly when
they approach the star on their eccentric orbits. The
planets in our solar system have their own retenue of
satellites. The space betw een M ars and Jupiter contains
the asteroids, m ost of which are fragments of larger
bodies that have been broken up by collisions and by the
gravitational forces of Jupiter and Mars. Pieces of the
asteroids have impacted as meteorites on the surfaces of
the planets and their satellites and have left a record of
these events in craters.
O n an even sm aller scale, space betw een stars
contains clouds of gas and solid particles. The gas is com
posed prim arily of hydrogen and of helium th at w ere
produced during the initial expansion of the universe. In
addition, the interstellar m edi um contains elem ents of
higher atomic num ber th at w ere synthesized by nuclear
reactions in the interiors o f stars that have since
exploded. A third com ponent consists of com pounds of
hydrogen and carbon th at are the precursors of life. These
2.2 STELLAR EVOLUTION 11
clouds of gas and dust may contract to form new stars
whose evolution depends on their masses and on the H /H
e ratio of the gas cloud from which they form ed.
The evolution of stars can be described by specifying
their luminosities and surface tem pera tures. The lum
inosity of a star is proportional to its mass, and its surface
tem perature or color is an indicator of its volume. W hen
a cloud of interstel lar gas contracts, its tem perature
increases, and it begins to radiate energy in the infrared
and visible parts of the spectrum . A s the tem perature in
the core of the gas cloud approaches 20 X 10 6 K, energy
production by hydrogen fusion becomes possible, and a
star is born. M ost of the stars of a typical galaxy derive
energy from this process and therefore plot in a band,
called the main sequence, on the H ertzsprung -R ussell
diagram shown in Figure 2.1. M assive stars, called blue
giants, have high luminosities and high surface tem
peratures. The Sun is a star of interm ediate mass and has
a surface tem perature of about 5800 K. Stars that are less
massive than the Sun are called red dwarfs and plot at the
low er end of the main sequence.
As a star five times m ore massive than the
Sun converts hydrogen to helium while on the
m ain sequence, the density of the core increases, causing
the interior of the star to contract. The core tem perature
therefore rises slowly during the hydrogen-burning phase.
This higher tem perature accelerates the fusion reaction
and causes the outer envelope of the star to expand.
However, when the core becom es depleted in hydrogen,
the rate of energy production declines and the star
contracts, raising the core tem perature still fur ther. The
site of energy production now shifts from the core to the
surrounding shell. The result ing changes in lum inosity
and surface tem pera ture cause the star to move off the
main sequence tow ard the realm of the red giants (Figure
2.1).
The helium produced by hydrogen fusion in the shell
accum ulates in the core, which continues to contract and
therefore gets still hotter. The resulting expansion of the
envelope lowers the surface tem perature and causes the
color to turn red. A t the sam e tim e, the shell in which
hydrogen
12 IN THE BEGINNING

log Surface Temperature, K

Figure 2.1 Stellar evolution on a Hertzsprung-Russell diagram for stars ranging from 1 to 9 solar masses. When a star
has used up the hydrogen in its core, it contracts and then moves off the main sequence and enters the realm of the red
giants, which generate energy by helium fusion. The evolutionary track and the life expectancy of stars are strongly
dependent on their masses. Stars five times more massive than the sun are nearly 1000 times brighter, have surface
temperatures of about 18.000 Kcompared to 5800 K for the Sunand remain on the main sequence only about 68
million years. Their evolution to the end of the major phase of helium burning takes only about 87 million years (Iben,
1967).
is reacting gradually thins
as it moves toward the
surface, and the
luminosity of the star
declines. These changes
transform a m ain-
sequence star into a
bloated red giant. For
example, the radius of a
star five tim es m ore
massive than the Sun
increases about 30-fold
just before helium burn
ing in the core begins.
W hen the core tem perature
1 0 0 X 106 K, helium fusion byas
alpha process begins and
converts three helium
nuclei into the nucleus of
carbon-12. A t the same
time, hydrogen fusion in
the shell around the core
continues. The
luminosities and surface
tem pera tures (color) of
red giants become
increasingly variable as
they evolve, reflecting
changes in the rates of
energy production in the
core and shell. The
evolutionary tracks in
Figure 2.1 illustrate the
importance of the mass of
a star to its evolution. A
approaches
star five times as massive
m the
eansSun
of the
is triple
1000 times
brighter while on the main
sequence and has a m ore
eventful life as a red giant
than stars below about two
solar masses (Iben,
1967,1974).

The length of tim e a star spends on the main
sequence depends primarily on its mass and to a lesser
extent on the H /H e ratio of its ancestral gas cloud. In
general, massive stars (blue giants) consum e their fuel
rapidly and may spend only 10 X 106 years on the main
sequence. Small stars (red dwarfs) have much slower m
etabolic rates and rem ain on the main sequence for very
long periods of tim e exceeding 10 X 109 years. The Sun,
being a star of m odest magnitude, has enough hydrogen in
its core to last about 9 X 109 years at th e present rate of
consum ption. Since it form ed about 4.5 X 109 years ago,
the Sun has achieved m iddle age and will provide energy
to the planets of the solar system for a very long tim e to
come. H ow ever, ultim ately its lum inosity will increase,
and it will expand to becom e a red giant, as show n in
Figure 2.1. The tem perature on the sur face of the E arth
will then rise and becom e intol erable to life forms. The
expansion of the Sun m ay engulf the terrestrial planets,
including Earth, and vaporize them . W hen all of its
nuclear fuel has been consum ed, the Sun will assum e the
end stage of stellar evolution th at is appropriate for a star
of its mass and chemical com position.
Toward the end of the giant stage, stars becom e
increasingly unstable. W hen the fuel for a particular
energy-producing reaction is exhaust ed, the star contracts
and its internal tem perature rises. The increase in tem
perature may trigger a new set of nuclear reactions. In
stars of sufficient mass, this activity culminates in a
gigantic explo sion (supernova) as a result of which a
large frac tion of the outer envelope of the star is blown
away. The debris from such explosions mixes with hydro
gen and helium in interstellar space to form clouds of gas
and dust from which new stars may form.
A s stars reach the end of their evolution they
turn into white dwarfs, or neutron stars (pulsars),
or black holes, depending on their masses
(W heeler, 1973). Stars whose mass is less than
about 1 . 2 solar masses contract until their radius is
only about 1 X 104 km and their density is
betw een 104 and 108 g/cm 3. Stars in this configu
ration have low luminosities but high surface tem
peratures and are therefore called white dwarfs (Figure
2.1). They gradually cool and fade from
2.3 NUCLEOSYNTHESIS13
view as their lum inosities and surface tem pera tures
diminish with time. Stars that are apprecia bly m ore
massive than the Sun develop dense cores because of the
synthesis of heavy chemical elem ents by nuclear
reactions. Eventually, such stars becom e unstable and
explode as supernovas. The core then collapses until its
radius is reduced
to about 1 0 km and its density is of the order of
10n to 101 5 g/cm 3. Such stars are com posed of a
neutron gas because electrons and protons are forced to
com bine under the enorm ous pressure and the abundance
of neutrons greatly increases as a result. N eutron stars
have very rapid rates of rotation and em it pulsed radio
waves that were first observed in 1965 by Jocelyn Bell, a
graduate student working with A . Hewish at the
Cavendish Laboratory of Cam bridge University in
England (Hewish, 1975). The Crab N ebula contains such
a pulsar, which is the rem nant of a supernova observed
by Chinese astronom ers in 1054 A .D . (Fowler, 1967). The
cores of the most massive stars collapse to form black
holes in accordance with Einsteins theory of general
relativity. Black holes have radii of only a few kilom eters
and densities in excess of 10 1 6 g/cm 3. Their gravitational
field is so great that neither light nor m atter can escape
from them, hence the nam e black hole. O bservational
evidence supporting the existence of black holes is
growing and they are believed to be an im portant phenom
enon in the evolution of galaxies.
Stars, it seems, have predictable evolutionary life
cycles. They form, shine brightly for a while, and then
die. H ans Bethe (1968, p. 547) put it this way:
If all this is true, stars have a life cycle much like
animals. They are born, they grow, they go through
a definite internal development, and finally die, to
give back the material of which they are made so
that new stars may live.
2.3 Nucleosynthesis
The origin of the chem ical elem ents is intim ately linked
to the evolution of stars because the ele m ents are
synthesized by the nuclear reactions from which stars
derive the energy they radiate into space. O nly helium and
deuterium , the
14 IN THE BEGINNING
heavy isotope of hydrogen, w ere synthesized during the
initial expansion of the universe. The entire theory of
nucleosynthesis was presented in a detailed paper by B
urbidge, B urbidge, Fowler and H oyle (1957) referred to
affectionately as B2 FH . Subsequent contributions (e.g.,
Schram m and A rn ett, 1973) have been concerned with
spe cific aspects of the theory and its application to the
evolution of stars of different m asses and ini tial
compositions.
The theory p resen ted by B 2F H (1957) evolved
from the work of several other scientists, am ong whom G
eorge G am ow deserves special recognition. G am ow
received his doctorate from
the U niversity of Leningrad and cam e to the
U niversity of G ttingen in th e spring of 1928
for som e postdoctoral studies. T here he m et
Fritz H outerm ans, with w hom he spent many
hours in a cafe making calculations (Gamow,
1963). Subsequently, H outerm ans m oved to the
U niversity of Berlin w here he m et the British
astronom er R o b ert A tkinson. H outerm ans and
A tkinson used G am ow s theory of alpha decay to
propose that stars generate energy in their interi ors by the
form ation of helium nuclei from four protons captured by
the nucleus of another light elem ent. The initial title of
their paper was Wie kann m an einen H elium kern im
Potentialtopf kochen. (H ow one can cook a helium
nucleus in a pressure cooker.) Needless to say, the editors
of the Zeitschrift f r P hysik insisted that the title be
changed (A tkinson and H outerm ans, 1929).
Ten years later, in A pril of 1938, Gam ow orga nized
a conference in Washington, D.C., to discuss the internal
constitution of stars. The conference stimulated one
attendee, H ans Bethe, to examine the possible nuclear
reactions betw een protons and the nuclei of the light
elements in order of increas ing atomic number. Bethe
(1939) found that the pressure cooker of A tkinson and
H outerm ans was the nucleus of n6C and m ade it the
basis of his famous CNO cycle for hydrogen fusion in
stars (to be presented later in this section) (Bethe, 1968).
G am ow influenced the evolution of the theo ry of
nucleosynthesis in m any other ways. In 1935 he
published a paper in the Ohio Journal o f
Science (an unlikely place for a nuclear astro physics
paper) on the buildup of heavy elements
by neutron capture and subsequently cham pi
oned the idea th at the chemical elem ents were
synthesized during the first 30 min after the Big Bang. In
the mid-1940s A lpher and Gamow wrote a paper
detailing the origin of the chemical ele
m ents based on that assumption. A t G am ow s
suggestion, H ans B ethes name was added in
absentia, thus creating the famous trium virate
A lpher, Bethe, and Gam ow (1948).
The abundance of the chemical elem ents and
their naturally occurring isotopes is the blueprint for all
theories of nucleosynthesis. For this rea son, geochem
ists and stellar spectroscopists have devoted m uch tim e
and effort to obtaining accu rate analytical data on the
concentrations of the elem ents in the Sun and other
nearby stars from the wavelength spectra of the light they
emit. Inform ation on the abundances of nonvolatile elem
ents has com e also from chemical analyses of stony m
eteorites, especially the carbonaceous chondrites, which
are the most undifferentiated
sam ples of m atter in the solar system available
to us (M ason, 1962). Table 2.1 lists the abun
dances of th e elem ents in the solar system com piled by
A nders and E bihara (1982). The abun dances are
expressed in term s of the num ber of atom s relative to
106 atom s of silicon. Figure 2.2 is a plot of these data and
illustrates several im portant observations about the
abundances of the elem ents.
1. H ydrogen and helium are by far the most
abundant elem ents in the solar system, and the
atomic H /H e ratio is about 12.5.
2. The abundances of the first 50 elem ents
decrease exponentially.
3. The abundances of the elem ents having
atom ic num bers greater than 50 are very low
and do not vary appreciably with increasing
atomic number.
4. Elem ents having even atom ic num bers are
m ore abundant than their im m ediate
neighbors with odd atom ic num bers
(O ddo -H arkins rule).
T ab le 2.1 A b u n d a n c e s o f th e E le m e n ts in th e S o lar S ystem in U n its o f N u m b e r o f A to m s p e r
106 A to m s o f Silicon
Atom ic no. Element Symbol Abundance Atomic no. Element Symbol Abundance"
1 hydrogen H 2.72 x 1010 53 iodine I 9.0 x KT1

2 helium He 2.18 X 109 54 xenon Xe 4.35 x 10

3 lithium Li 5.97 x 10* 55 cesium Cs 3.72 X 101
4 beryllium Be 7.8 x HT1 56 barium Ba 4.36 X 10
5 boron B 2.4 X 10' 57 lanthanum La 4.48 X 101
6 carbon C 1.21' X 107 58 cerium Ce 1.16 X 10
7 nitrogen N 2.48 x 106 59 praseodymium Pr 1.74 x 10 1
8 oxygen O 2.01 X 107 60 neodymium Nd 8.36 X 10-*
9 fluorine F 8.43 X 102 61 promethium Pm 0
10 neon Ne 3.76 X 106 62 samarium Sm 2.61 X 101
11 sodium Na 5.70 X 104 63 europium Eu 9.72 X 10~2
12 magnesium Mg 1.075 x 106 64 gadolinium Gd 3.31 X 101
13 aluminum A1 8.49 X 104 65 terbium Tb 5.89 X 10 2
14 silicon Si 1.00 X 106 66 dysprosium Dy 3.98 X 10~
15 phosphorus P 1.04 X 104 67 holmium Ho 8.75 x 102
16 sulfur s 5.15 X 105 68 erbium Er 2.53 X 10 1
17 chlorine Cl 5.240 x 103 69 thullium Tm 3.86 X 10~2
18 argon Ar 1.04 X 105 70 ytterbium Yb 2.43 x 101
19 potassium K 3.770 X 103 71 lutetium Lu 3.69 X 102
20 calcium Ca 6.11 X 104 72 hafnium Hf 1.76 X 10 -1
21 scandium Sc 3.38 x 101 73 tantalum Ta 2.26 X 10"2
22 titanium Ti 2.400 x 103 74 tungsten W 1.37 X 10"1
23 vanadium V 2.95 x 102 75 rhenium Re 5.07 X 102
24 chromium Cr 1.34 X 104 76 osmium Os 7.17 x HT1
25 manganese Mn 9.510 x 103 77 iridium Ir 6.60 x 101
26 iron Fe 9.00 x 105 78 platinum Pt 1.37 x 10
27 cobalt Co 2.250 x 103 79 gold Au 1.86 X 101
28 nickel Ni 4.93 x 10" 80 mercury Hg 5.2 X 101
29 copper Cu 5.14 X 102 81 thallium Tl 1.84 X 101
30 zinc Zn 1.260 X 103 82 lead Pb 3.15 X 10
31 gallium Ga 3.78 X 102 83 bismuth Bi 1.44 X 101
32 germanium Ge 1.18 X 102 84 polonium Po ~0
33 arsenic As 6.79 x 10 85 astatine At ~0
34 selenium Se 6.21 x 101 86 radon Rn ~0
35 bromine Br 1.18 X 101 87 francium Fr ~0
36 krypton Kr 4.53 X 101 88 radium Ra ~0
37 rubidium Rb 7.09 X 10 89 actinium Ac ~0
38 strontium Sr 2.38 X 101 90 thorium Th 3.35 x io -2
39 yttrium Y 4.64 x 10 91 protactinium Pa ~0
40 zirconium Zr 1.07 x 101 92 uranium U 9.00 x 10~3
41 niobium Nb 7.1 X 10 1 93 neptunium Np 0
42 molybdenum Mo 2.52 X 10 94 plutonium Pu ~0
43 technetium Tc 0 95 americium Am 0
44 ruthenium Ru 1.86 x 10 96 curium Cm 0
45 rhodium Rh 3.44 X 101 97 berkelium Bk 0
46 palladium Pd 1.39 x 10 98 californium Cf 0
47 silver Ag 5.29 X 101 99 einsteinium Es 0
48 cadmium Cd 1.69 x 10 100 fermium Fm 0
49 indium In 1.84 X 101 101 mendelevium Md 0
50 tin Sn 3.82 X 10 102 nobelium No 0
51 antimony Sb 3.52 X 10"1 103 lawrencium Lr 0
52 tellurium Te 4.91 x 10

The terrestrial abundances of the radioactive daughters of uranium and thorium are very low and are therefore indicated as ~0 . In
addition, neptunium and plutonium are produced in nuclear reactors and therefore occur on the Earth. However, the transuranium
elements having atomic numbers of 95 or greater do not occur in the solar system. Their abundances are therefore stated as 0. S O U R C E :
Anders and Ebihara (1982).
16 IN THE BEGINNING

Atomic Number (Z)

Figure 2.2 Abundances of the chemical elements in the solar system in terms of atoms per 106 atoms of Si. The data
were derived primarily by analysis of carbonaceous chondrite meteorites and by optical spectroscopy of light from the
Sun and nearby stars (Anders and Ebihara, 1982).
5. T he abundances of ters of long-
lithium , lived radioactive
beryllium, and isotopes of
boron are anom uranium and
alously low com thorium .
pared to o ther
The nucleosynthesis
elem ents of low
m odel of B 2F H (1957)
atom ic num ber.
includes eight different
6 . The abundance of
kinds of nuclear
iron is notably
reactions that occur at
higher than those specified tem peratures
of other elem ents in the course of the
with similar atom evolution of a star.
ic numbers. Several of these reac
tions m ay take place sim
7. Two elem ents, ultaneously in the cores
technetium and and outer shells of
prom ethi um , do massive stars. As a
not occur in the result, the core of a star
solar system m ay have a different
because all of their chemical com position
isotopes are than the shell
surrounding it. M
unstable and decay
oreover, not all of the
rapidly.
nuclear processes take
place in all stars.
8 . T he elem ents
Consequently, other stars
having atom ic
in the M ilky Way
num bers greater
Galaxy do n o t
than 83 (Bi) have
necessarily have the
no stable iso topes,
same chem i cal com
but occur naturally position as the Sun and
at very low her planets.
abundances
because they are
the daugh
 2.3 NUCLEOSYNTHESIS17

A ll stars on the main sequence generate ener gy by ed, the interstellar gas clouds contained elem ents of
hydrogen fusion reactions. This process results in the higher atom ic num ber. The presence of carbon-12 (gC)
synthesis of helium either by the synthesized by the ancestral stars has m ade it easier for
direct p ro to n -p ro to n chain (equations 2.5-2.8) subsequent generations of stars to generate energy by
or by the CN O cycle (equations 2.9 -2.14). The hydrogen fusion. This alter native m ode of hydrogen
proton -proton chain works as follows. Two nuclei fusion was discovered by H ans B ethe and is know n as
of hydrogen, consisting of one proton each, collide to the CN O cycle:
form the nucleus of deuterium ((H ) plus a positron ( +) 12r + !H 13N + y (2.9)
and a neutrino (v). (The designation of atomic species is 0
'(N >13r + + + V (2 . 1 )
presented in Chapter 6 .) Each reaction of this kind
1
liberates 0.422 million electron volts (M eV) of energy. l3r + IH >(N + y (2 . 1 )
The positron (positively charged electron) is annihilated 2
'(N + 15o + y (2 . 1 )
by interacting with a negatively charged electron giving
S'* +
off additional energy of 1.02 MeV. The deuterium nucleus N

col lides with another proton to form the nucleus of
helium-3 QHe) plus a gamma ray (y) and 5.493 M eV of
energy. Finally, two helium-3 nuclei must collide to form
helium-4 ((H e), two protons, and 12.859 MeV. The end
result is that four hydrogen nuclei fuse to form one
nucleus of helium-4, a gamma ray, a neutrino, and 19.794
M eV of energy.
The entire process can be described by a series of
equations in which the nuclei of hydrogen and helium are
represented by the symbols of the appropriate isotopes
(see C hapter 6 ), even though these isotopes do not
actually exist in atomic form in stellar interiors w here
their electrons are rem oved
from them because of the high tem perature:
|H + JH -> (H + + + v + 0.422 M eV (2.5)
+ + ~ 7 -> 1.02 M eV (annihilation) (2.6)
(H + }H -> (H e + y + 5.493 M eV (2.7)
(H e + (H e - (H e + }H + |H + 12.859 M eV
| o 13 ++ V (2.13)
13N + + (H e (2.14)
!H >12r
The end result is that four protons are fused to form one
nucleus of (H e, as in the direct proton - proton chain. The
nucleus of l\C acts as a sort of catalyst and is released at
the end. It can then be reused for another revolution of the
CN O cycle.
The Sun contains elem ents of higher atomic num ber
than helium including *gC and therefore carries on
hydrogen fusion by th e C N O cycle. In fact, m ost stars
in our M ilky W ay G alaxy are second-generation stars
because our G alaxy is so old that only the very sm allest
first-generation stars could have survived to the present
time.
The low reaction cross section of the p ro to n -
proton chain by which the ancestral stars gener ated
energy has been a source of concern to nuclear
astrophysicists. W hen this difficulty was pointed out to
Sir A rthur E ddington, who pro posed hydrogen fusion in
stars in 1920, he replied (Fowler, 1967):

(2.8)
The direct proton -proton fusion to form helium -4
can only take place at a tem perature of about
1 0 X 106 K, and even then the probability of its
occurrence (or its reaction cross section) is very small.
Nevertheless, this process was the only source of nuclear
energy for first-generation stars th at form ed from the
prim ordial m ixture of hydrogen and helium after the Big
Bang.
Once the first generation of stars had run through
their evolutionary cycles and had explod
We do not argue with the critic who urges that the
stars are not hot enough for this process; we tell him
to go and find a hotter place.
A fter the hydrogen in the core has been convert ed to
helium ash, hydrogen fusion ends, and the core
contracts under the influence of gravity. The core tem
perature rises tow ard 100 x 10 6 K, and the helium ash
becom es the fuel for the next set of energy-producing
nuclear reactions. The critical reaction for helium burning
is the fusion of three
18 IN THE BEGINNING

alpha particles (triple-alpha process) to form a nucleus of are the neutron-capture reactions, which produce a large
l\C\ num ber of the atoms having atomic numbers greater than
26 (Fe). These reactions involve the addition of a neutron
jH e + gHe - gBe (2.15)
to the nucleus of an atom to produce an isotope having
IjBe + gHe - l2C + y (2.16)
the same atomic number but a large mass number. For
This is the critical link in the chain of nucleosyn thesis example, the equation:
because it bridges the gap in nuclear stabili ty of the ggNi + in -> Ni + y (2.21)
isotopes of lithium, beryllium, and boron. The crux of the neutron

problem is that the nucleus of gBe is very unstable and indicates that the nucleus of ggNi absorbs a neu tron,
decays quickly with a half-life of about 10 " 1 6 sec. which changes it to an excited state of ggNi, which then
Therefore, gBe m ust absorb a third helium nucleus very deexcites by em itting a gam m a ray. Nickel-63 is
soon after its form ation to make it safely to stable 'gC radioactive and decays to stable ggCu by em itting a ~
(Fowler, 1967). An alternative reaction involving the particle:
addition of a pro ton to the nucleus of gHe has an even
smaller chance to succeed because the product, gLi, has a ggNi -> ggCu + - + v + 0.0659 M eV (2.22)
half-life of only 1 (T 2 1 sec and decom poses to form helium w here v is an antineutrino. ggCu is a stable iso
and hydrogen: tope of copper and can absorb another neutron to form
IgCu:
jH e + }H - 53Li (2.17) ggCu + in -> IgCu + y (2.23)
gLi - gHe + |H (2.18)
Copper-64 is radioactive and undergoes branched decay
The triple-alpha process is indeed the key to to form g^Zn and ggNi, both of which are stable:
the synthesis of all elem ents beyond helium .
W ithout it, stellar evolution w ould be short- ggCu -> ggZn + ~ + v + 0.575 M eV (2.24)
circuited, and the universe would be com posed ggCu -> g N i + + + v + 1.678 MeV (2.25)
only of hydrogen and helium .
H elium burning sustains red giants only for a few This process of successive additions of neutrons is
tens of millions of years or less. W ith increas illustrated in Figure 2.3. It takes place during the red giant
ing tem perature in the core, alpha particles fuse stage of stellar evolution when the neu tron flux is low
with the nuclei of 'gC to produce nuclei of still enough to perm it the product nucleus to decay before the
higher atomic num ber: next neutron is added. This process therefore is
12C + gHe -> 'gO (2.19) characteristically slow and is therefore referred to as the s-
'gO + gHe -> 2"Ne (2.20) process.
By exam ining Figure 2.3 closely we see that the
etc.
track of the s-process bypasses Zn, which is
However, electrostatic repulsion betw een posi tively one of the stable isotopes of . zinc. In order to
charged nuclei and alpha particles limits the size of the m ake this isotope by neutron capture reactions,
atom s th at can form in this way. The heaviest atom the pace m ust be speeded up so that unstable
produced by the addition of alpha particles is ggNi, which goZn can pick up a neutron to form ggZn before it decays
decays to ggCo and then to stable ggFe. These nuclear
to stable 3 ?Ga:
reactions there fore cause the enhanced abundance of the
ele m ents in the iron group illustrated in Figure 2.2. goZn + In - ]0Z n + y (2.26)
A n even m ore rapid rate of neutron capture is
During the final stages of the evolution of red giants, required to m ake ggZn from ggCu by addition of five
several other kinds of nuclear reactions occur (B 2 FH , neutrons in succession to form 2 9 CU, which then decays
1957). The most im portant of these by ~ emission to stable ggZn:
 2.3 NUCLEOSYNTHESIS 19

Neutron Number (N)

Figure 2.3 Nucleosynthesis in red giants by neutron capture on a slow time scale (s-process) followed by
beta decay. The dark squares are stable isotopes, whereas the unshaded ones are radioactive. The process
starts with stable (|Ni, which absorbs a neutron to form unstable Ni, which decays to stable ))Cu by
emitting a ~ particle. The main line of the s-process, as indicated by arrows, proceeds from N i to ((Se
and beyond but bypasses ([}Zn and ((Se. (jjZn is produced by neutron capture on a rapid time scale (r-
process) from fgZn via unstable (fjZn and from ))Cu, which captures five neutrons in rapid succession to
form (Cu followed by decay to stable (JjZn. ((Se is a proton-rich nuclide that cannot form by either the
s-process or the r-process and requires the addition of two protons (p-process) to stable ((Ge.

65Cu + 5 in -> JgCu +
29
70
lC u ~ + v +
3oZn +
C apture of neutrons at
such rapid rates is charac
teristic of the r-process,
which requires a much
greater neutron flux than
the s-process and th ere
fore takes place only
during the last few m
inutes in the life of a red
giant w hen it explodes as
a supernova.
How ever, no neutron
capture on any tim e scale
can account for the form
ation of some atom s such
as stable ((Se shown also
in Figure 2.3. This nuclide
is synthesized by the
addition of two p ro tons
to stable ((G e in the so-
called p-process:
IjGe + 2 }H
black holes have high
This process also takes
place at the very end of
the giant stage of stellar
evolution.
The system of
nuclear reactions
originally pro posed by B
2F H (1957) can account
for the observed
abundances of the
chemical elem ents in the
solar system and in nearby
stars. Nucleo synthesis is
taking place at the present
time in the stars of our
galaxy and in the stars of
other galaxies throughout
the universe. We have
good evidence in the
wavelength spectra of
light from distant galaxies
that the chemical elements
we find on E arth also
occur everywhere else in
the universe.
20 IN THE
BEGINNING
However, pulsars and
hydrogen fusion. Stars
inter nal pressures and tem
peratures that cause
atomic nuclei to
disintegrate into m ore
primitive con stituents.
The relative proportions of
the chemical elements in
other stars are different
because local conditions m
ay affect the yields of the
many nuclear reactions
that contribute to their
synthesis.
2.4 Summary
We live in an expanding
universe whose future is
uncertain. The universe
started with a Big Bang
about 15 X 109 years ago
and has evolved since then
in accordance with the
laws of physics.
Stars are the basic
units in the hierarchy of
heavenly bodies. They
form by contraction of
clouds of interstellar gas
and dust until their core
tem peratures are
sufficient to cause
evolve through predictable
stages depending on their
masses and initial
compositions. They
generate energy by nuclear
reactions that synthesize
other elem ents from prim
ordial hydro gen and
helium . In the end, stars
explode, and the rem nants
becom e solid objects of
great density.
The abundances of
the chemical elem ents in
the solar system can be
explained by the nuclear
reactions that energize the
stars. These reactions
progress from fusion of
hydrogen and helium to
neutron capture and to
other reactions, m ost of
which occur only for a
short time at the end of the
active life of a star. The
chemical elem ents we
know on E arth occur
throughout the universe,
but their abundances vary
because local conditions
affect the yields of the
nucleosynthesis processes.
Problems*
 1. How has the abundance of hydrogen (H) in the
universe changed since the Big Bang?
2. Why are elements with even atomic numbers more
abundant than their neighbors with odd atomic numbers?
3. How do we know that the Sun is at least a sec ond-
generation star?
4. Why do technetium (Tc) and promethium (Pm) lack
stable isotopes?
5. What other elements also lack stable isotopes?
6. Why do the elements of Problem 5 exist on the
Earth, whereas Tc and Pm do not?
*Not all of these questions have unequivocal answers.
7. Why is lead (Pb) more abundant than we might have
expected?
8. Check the abundance of argon (Ar) and deter mine
whether it is greater than expected. If so, suggest an
explanation.
9. How did lithium (Li), beryllium (Be), and boron
(B) form?
10. Compare the abundances of the rare earths
(lanthanum to lutetium) to such well-known metals as Ta,
W, Pt, Au, Hg, and Pb. Are the rare earths really all
thatrare?
References
A L P H E R , R. A., H. B E T H E , and A T K IN S O N , R. dE., and F. G. H
G. G A M O W , 1948. The origin O U T E R M A N S , 1929. Zur
of
Frage der
the chemical elements. Phys. Rev.,
Aufbaumglichkeit der
A N D E R S , E ., and M. E B I H A R A , 1982. Solar-system
Elemente in Sternen. Zeit.
dances of the elements. Geochim. Cosmochim.
Phys., 54:656-665.
46:2363-2380.
B A A D E , W., 1968. Evolution o f
Stars and Galaxies. Harvard
University Press,
Cambridge, MA.
 REFERENCES 21
B E T H E , H. A., 1939. Energy production in stars. Phys. Rev., H E W IS H , A., 1975. Pulsars and high density physics. Science,
55:434456. 188:1079-1083.
B ethe, H. A., 1968. Energy production in stars. Science, H U B B L E , E., 1936. The Realm o f the Nebulae. Yale University
161:541-547. Press, New Haven, CT. Reprinted 1958 by Dover Pub., Inc.,
B U R B R I D G E , E . M., G. R. B U R B R I D G E , W. A. F O W L E R , and F. New York.
H O Y L E , 1957. Synthesis of the elements in stars. Rev. Mod. I B E N , I., Jr., 1967. Stellar evolution: Comparison of theory with
Phys., 29:547-650. observation. Science, 155:785796.
F O W L E R , W. A., 1967. Nuclear astrophysics. Amer. Phil. Soc. I B E N , I., Jr., 1974. Post main sequence evolution of single stars.
Mem. 67,109 pp. Ann. Rev. Astron. and Astrophys., 12:215-256.
Gamow, G., 1935. Nuclear transformations and the origin of M A S O N , B., 1962. Meteorites. Wiley, New York, 274 pp.
the chemical elements. Ohio J. Sei., 35:406-413. P E N Z IA S , A. A., and R. W. W IL S O N , 1965. A measurement of
Gamow, G., 1952. The Creation o f the Universe. Viking Press, excess antenna temperature at 4,080 Mc/s. Astrophys. J.,
New York, 144 pp. 142:419.
G A M O W , G., 1963. My early memories of Fritz Houtermans. In S C H R A M M , D . N., and W. D. A R N E T T , 1973. Explosive
J. Geiss and E. D. Goldberg (Eds.), Earth Science and Nucleosynthesis. University of Texas Press, Austin and
Meteoritics. North-Holland Publ. Co., Amsterdam, 312 pp. London, 301 pp.
G O T T , J. R., Ill, 1982. Creation of open universes from de W E IN B E R G , S., 1977. The First Three Minutes. Bantam Books,
Sitter space. Nature, 295:304-307. New York, 177 pp.
H A IN E B A C H , K ., D . K A Z A N A S , and D . N. SC H R A M M , 1978. A W H E E L E R , J. C., 1973. After the supernova, what? Amer.
consistent age for the universe. Short papers of the Fourth Scientist, 61:42-51.
International Conference, Geochronology, Cosmochronology
and Isotope Geology. U.S. Geol. Sur. Open-File Rept., 78-
701,159-161.
 3
The Solar System
The Sun form ed from a cloud of gas and dust par ticles,
as did all other stars in the Milky Way G alaxy and
elsewhere in the universe. Sometimes the process leads to
the form ation of two com panion stars. However, in the
case of the Sun a small fraction of the original cloud
accreted to form a set of nine planets, including the Earth.
C om pared to stars, these planets are insignificant
objects, but to us they are the very basis of exis tence.
The hum an race appeared on the Earth only two or three
million years ago and, after a slow start, learned to fly
and to explore the solar system within a span of a m ere
century.
G eochem istry today not only encompasses
the study of the com position and chemical
processes occurring on the E arth, but is also con cerned
with all of the planets and their satellites. Inform ation
for geochemical studies of the solar system is derived by
analysis of m eteorites and rock samples from the M oon
and by remote sensing of planetary surfaces. The
exploration of the solar system has expanded our horizon
and has provided the basis for com parative planetary
geochemistry. The satellites of the large gaseous planets
are of special interest in this new field of study because
som e of them are larger than our M oon and have very
different chemical compo sitions and surface features
than the E arth .
We therefore need to becom e acquainted with the
new worlds we m ust explore before we concentrate our
attention on the conventional geochem istry of the E arth.
3.1 Origin of the Solar System
The origin of the planets of the solar system is intim ately
linked to the form ation of the Sun. In the beginning there
was a diffuse mass of inter
stellar gas and dust know n as the solar nebula. It had
form ed about six billion years ago as a result
of the term inal explosions of ancestral stars,
which added the elem ents they had synthesized
to the prim ordial hydrogen and helium that orig inated
from the Big Bang. The chemical composi tion of the
solar nebula was given in Chapter 2 (Table 2.1). The dust
cloud was rotating in the same sense of the M ilky Way
Galaxy and was acted upon by gravitational, magnetic,
and elec trical forces.
As soon as the m ain mass of the solar nebu la
began to contract, order began to be imposed on it by the
physical and chemical conditions that existed during this
phase of star formation. These included the developm ent
of pressure and tem perature gradients and an increase in
the rate of rotation. Certain kinds of solid particles that
had form ed in the nebula evaporated as the tem pera
ture increased in order to m aintain equilibrium
betw een solids and gases. As a result, only the
m ost refractory particles (Fe-N i alloys, A12 0 3,
CaO, etc.) survived in the hottest part of the con
tracting nebula, w hereas in the cooler outer regions a
larger variety of com pounds remained in the solid state.
The increase in the rate of rota tion caused p art of the
nebula outside of the pro-tosun to form a central disk.
The solid particles congregated in this disk and m ade it
sufficiently opaque to absorb infrared radiation. The tem
per ature in the central disk therefore increased until it
ranged from about 2000 K at the center to about 40 K at
approxim ately 7.5 X 109 km from the protosun. The
pressure ranged from less than 0 . 1 atm to about 1 0 - 7 atm
near the edge of the disk (Cam eron, 1978; Cam eron and
Pine, 1973).
The developm ent of pressure and tem pera ture
gradients within the disk caused the first
22
3.1 ORIGIN OF THE SOLAR SYSTEM 23
m ajor chemical differentiation of the solar nebu la. C om
pounds with low vapor pressures persisted throughout the
nebula and form ed dust parti cles, w hereas com pounds
with high vapor pres sures could exist only in the cooler
outer regions. The condensation tem peratures of various
com pounds that existed in the solar nebula are listed in
Table 3.1. The condensates accreted to form larger bodies
as a result of selective adhesion caused by electrostatic and
magnetic forces. The resulting solid bodies, called
planetesimals, had diam eters ranging from about 1 0 m to
m ore than 1 0 0 0 km, and their chemical compositions varied
with distance from the center of the planetary disk.The
planetesim als close to the protosun were com posed of
refractory com pounds dom inated by oxides and m etallic
iron and nickel; farther out were Mg and Fe silicates and
farthest out, ices com posed of water, ammonia, m ethane,
and other
volatiles.
The initial rate of evolution of the solar sys
tem was rem arkably fast. The tim e required for
the Sun to reach the ignition tem perature for
hydrogen fusion was less than 100,000 years. The Suns
initial luminosity was two to three times greater than is
consistent with the main sequence because the Sun
contained excess therm al energy g enerated during the
initial contraction . This superlum inous phase of the
Sun lasted about 1 0 million years and resulted in the
expulsion of about 25% of its original mass in the form of
solar w ind com posed of a proton/electron plas ma.
This is called the T-Tauri stage of stellar evolution after
the star th at is the prototype for this process. All gaseous
m atter in the vicinity of the Sun was blown away during
this period, and only the solid planetesim als having diam
eters larger than about 1 0 m rem ained.
The planetesim als in the inner region of the
planetary disk subsequently accreted to form the
so -called earthlike planets M ercury, Venus,
E arth, and M ars and the parent bodies of the
m eteo rites now rep resen ted by asteroids.
Instabilities in the outer p art of the disk resulted in the
form ation of the gaseous outer planets: Jupiter, Saturn, U
ranus, N eptune, and Pluto.
Table 3.1 Condensates from the Solar Nebula at Different Temperatures
Temperature, C Condensates
 1325 refractory oxides: CaO, A120 3, T i02, REE oxides
1025 metallic Fe and Ni
925 enstatite (MgSiO,)
925-220 Fe forms FeO, which reacts with enstatite to form olivine [(Fe,Mg)2S i0 4] Na
725 reacts with A120 3 and silicates to form feldspar and related minerals.
Condensation of K and other alkali metals F12S
400 reacts with metallic Fe to form troilite (FeS)
280 H20 vapor reacts with Ca-bearing minerals to make tremolite
150 H20 vapor reacts with olivine to form serpentine
- 100 HzO vapor condenses to form ice
- 125 NH3 gas reacts with water ice to form solid NH3 H20
- 150 CH4 gas reacts with water ice to form solid CH4 7H20
1N>O Ar and excess CH4 condense to form Ar and CH4
-250" Ne, H, and He condense
These reactions probably did not occur because the temperature in the planetary disk did not decrease to such
low values.
SOURCE: Glass (1982), based on Lewis (1974).
24 THE SOLAR SYSTEM
 The origin and chem ical com position of Pluto are
still not well known because this planet is difficult to
observe from E arth . Pluto also does not conform to the
so-called Titius-B ode law, which appears to govern the
distances of the planets from the Sun when these are
expressed in astronomical units (A . U.), defined as the
average
distance betw een the E a rth and Sun. The
T itius-B ode law was published in 1772 by J. E.
Bode, director of the A stronom ical O bservatory in Berlin,
and is based on a series of num bers dis covered by J. D.
Titius of W ittenberg in 1766. The series is com posed of
the num bers 0.4, 0.7, 1.0, 1 .6 . 2 .8 . . . . , which are obtained
by writing 0 ,3 , 6 , 1 2 .2 4 ...., adding 4 to each num ber, and
dividing by 10. The resulting num bers m atch the dis
tances of the planets from the Sun rem arkably well up to
and including U ranus (M ehlin, 1968). However, the
radius of the orbit of N eptune is only 30.1 A .U ., w hereas
the T itius-B ode value is 38.8 and the discrepancy in the
case of Pluto is even greater (Table 3.2). The T itius-B ode
law predicts
a value of 77.2, but the actual orbital radius of
P luto is only 39.4 A .U . The discrepancy may
suggest th at P luto did n o t form in th e orbit it
now occupies.
The descriptive physical properties of the solar
system listed in Table 3.2 indicate that 99.87% of the total
mass of the solar system
(2.052 X 103 3 g) is concentrated in the Sun. The
rem aining 0.13% is distributed among the nine
m ajor planets, among which Jupiter is by far the largest
with about 71% of the planetary masses. The densities and
sizes of the planets, shown in Figure 3.1, vary widely and
imply the existence of large differences in their chemical
compositions. The inner planets (M ercury, Venus, E arth,
and M ars), as well as the M oon and the asteroids, are
solid objects com posed prim arily of elem ents and com
pounds having low vapor pressures. The outer planets
(Jupiter, Saturn, Uranus, and N eptune) have low densities
and are essentially gaseous, although all of them probably
have condensed cores. The inner planets resem ble the E
arth in chemical com position and are therefore referred to
as the terrestrial or earthlike planets. The outer planets
consist prim arily of hydrogen and helium
with small am ounts of the other elem ents and
resem ble the Sun in chemical com position.
The earthlike planets (plus the M oon and
asteroids), taken together, m ake up only 0.0006% of the
total mass of the solar system and only
Table 3 .2 Properties of the Sun and H er Planets
Mean solar distance
Equatorial Density Number o f
Object 106km A.U. diameter, km Mass, g g/cm3 satellites
Sun 1,391,400 1.987 x 1033 1.4

Mercury 57.9 0.387 4,878 3.30 x 1026 5.44 0

Venus 108.2 0.723 12,100 4.87 X 1027 5.25 0
Earth 149.6 1.00 12,756 5.98 x 1027 5.52 1
Moon 3,476 7.35 x 1025 3.34

Mars 227.9 1.52 6,786 6.44 x 1026 3.94 2

Ceres (asteroid) 414.0 2.77 1,020 1.17 x 1024 2.2
Jupiter 778.3 5.20 142,984 1.90 x 1030 1.33 16
Saturn 1,427.0 9.54 120,536 5.69 x 1029 0.70 21
Uranus 2,869.6 19.2 51,118 8.66 x 1028 1.30 15
Neptune 4,504 30.1 49,562 1.01 23 x 1029 1.76 8
Pluto 5,900 39.4 3,000 1.5 x 1025 1.1 1

SOURCE: Carr et al. (1984), Glass (1982), and fact sheet, Jet Propulsion Laboratory.
 3.2
0.44% of the planetary masses. A ll of these objects are so
close to the Sun that the orbital radius of the outerm ost
asteroid is only about 7% of the total radius of the solar
system. Evidently, the terrestrial planets are not typical o f
the solar system and owe their existence to the special
conditions in the planetary disk close to th e Sun. E arth is
the largest of the inner planets w ith about 50.3% of the
mass, followed by Venus (40.9% ), M ars (5.4% ), and M
ercury (2.8%).
The earthlike planets, viewed in Figure 3.1 in th e
perspective of the entire solar system, are a physical and
chemical anomaly. The E arth is unique am ong its
earthlike neighbors in having about 71% of its surface
covered by liquid w ater w ithin which life developed
early in its history and evolved into the m ultitude of
species of the p lan t and anim al kingdoms. Insofar as we
know, life form s do not exist at the present tim e any w
here else in the solar system.
ORIGIN O F THE EARTHLIKE PLANETS 25
3.2 Origin of the Earthlike
Planets
O ur current understanding of the origin of the solar
system indicates that the earthlike planets w ere hot w hen
they form ed and that their inter nal geochem ical
differentiation may have begun with the sequential
accretion of planetesim als of differing com positions (M
urray et. al., 1981). Planetesim als com posed of m etallic
iron and oxides accreted first to form a core that was sub
sequently buried by the planetesim als com posed of
silicates. The earthlike planets w ere initially m olten
because of the heat generated by the rapid capture of the
planetesim als and because of radioactive heating.
The last phase of form ation of E arth, Venus, and M
ars involved the capture of planetesim als com posed of
volatile com pounds that had form ed in the outer reaches
of the planetary disk, perhaps
 M ean Distance from th e C enter of th e Sun, 10s km
Figure 3.1 A: Variation of density of the planets with mean distance from the Sun. Note that the Earth has the
highest density among the earthlike planets, which, as a group, are more dense than the outer gaseous
planets. B: The planets of the solar system magnified 2000 times relative to the distance scale. The
earthlike planets are very small in relation to the Sun and the gaseous planets of the solar system.
26 THE SOLAR SYSTEM
beyond the orbit of Jupiter. These volatile-ric )
to form m ore than 4 X 109 years ago by conden sation of
planetesim als, also know n as com etesim ai-. w ater vapor in the atm osphere. The presence of a large
deposited solids com posed of water, ammonia, volume of w ater on the sur face of E arth perm itted
m ethane, and other volatiles on the surfaces of the geological processes to operate and created conditions
planets.The w ater and other volatiles deposited on conducive to the developm ent and evolution of life. N
the E arth prom ptly evaporated to form a dense either the E arth nor any of the earthlike planets ever had
atm osphere from which w ater ultim ately con atm ospheres com posed of the hydrogen and heli um of
densed as the surface of the E arth cooled. Mercury and the solar nebula because these gases were expelled from
the M oon do not have atm ospheres partly because they the inner region of the solar system during the T-Tauri
are too small to retain gaseous ele m ents of low atom ic stage of the Sun.
num ber and their compounds.
A ccording to this scenario, the earthlike planets
have been cooling since the time of their 3.3 Satellites of the Outer Planets
form ation. M ercury and the M oon have cooled
sufficiently to becom e geologically inactive in All of the outer planets have satellites, som e of which are
the sense th at their interiors no longer interact larger than the M oon and the planet Mercury. In addition,
w ith their surfaces. Venus and E arth, being the these satellites have a wide range of chem ical com
largest of the earthlike planets, have retained positions and have responded in very different ways to the
m ore of the initial heat than their neighbors in forces acting on them . The large satellites of Jupiter were
the solar system and are still active. M ars is inter m ediate seen by Galileo Galilei on January 7,1610, with his newly
in size and has had volcanic eruptions in the not-too- built telescope. Actually, they m ay
distance geologic past. However, the age of the last m
artian volcanism is not known. have been observed even earlier by Sim on
In spite of the similarity in size and overall com Marius. However, detailed images of their su r
position of Venus and E arth, their surfaces have evolved faces were obtained only recently during flybys of the A
very differently. Venus has a dense atm osphere com m erican Pioneer and Voyager space probes. A chronology
posed of C 0 2 that has caused its surface to becom e of these events in Table 3.3 reveals the sudden burst of
extrem ely hot and dry. The surface of E arth cooled activity in the explo ration of the solar system betw een
rapidly, allowing oceans 1973 and 1989.
Table 3.3 Chronology of Exploration of the Satellites of Jupiter and S a tu rn
Planet Encounter date Spacecraft Comments
Jupiter Dec. 3,1973 Pioneer 10 left solar system on June 14,1983

Jupiter Dec. 2,1974 Pioneer 11 continued to Saturn

Jupiter March 5,1979 Voyager 1 1408 images of jovian satellites
Jupiter July 9,1979 Voyager 2 1364 images of jovian satellites
Jupiter Dec. 7,1995 Galileo mission in progress in 1997
Saturn Sept. 1,1979 Pioneer 11 low-resolution images
Saturn Nov. 13,1980 Voyager 1 900 images of saturnian satellites
Saturn Aug. 26,1981 Voyager 2 1150 images of saturnian satellites
Uranus Jan. 25,1986 Voyager 2 images of all satellites as well as the rings
Neptune Aug. 25,1989 Voyager 2 images of the planet and its satellite Triton
After Greeley (1985).
3.3 SATELLITES OF THE OUTER PLANETS 27

The planet Jupiter has
16 satellites and a ring
guarded by two smallthesatellites.
G alilean satellites Io, E uropa,generated within Io by
Callisto w ere actually nam
m ythological lovers of the
G reek god Zeus whose
Latin nam e was Jupiter. A
fifth satellite was dis
covered 282 years later by
B arnard who nam ed it A
m alth ea after a nym ph
who once nursed Jupiter.
These satellites, together
with three small ones
found in 1979, move in
nearly circular orbits close
to the equatorial plane of
Jupiter. Four
small satellites (Leda, Him
E lara) are located about large
Jupiter and have eccentric
orbits th at are inclined
nearly 30 to its equatorial
plane. A third group (A
nanke, Carm e, Pasiphae,
and Sinope) at a dis tance
of about 22 million km
from Jupiter have
retrograde orbits whose inclination
150 and 160. The G alileandioxide satellites
effect a small-scale
planetary system of their
own with Jupiter as their
central star. Their
densities, listed in Table
3.4 and displayed in
Figure 3.2, decrease w ith
increasing orbital radius
from 3.53 (Io) to 1.79
(Callisto) g/cm3; therefore
they are believed to have
significantly different
chem ical compositions.
Io is com posed prim
arily of silicate m aterial
and m ay have an iron
sulfide core. Before the
encounters with the
Voyager spacecraft, Peale
et al. (1979) calculated
am ount The so-called
of heat
ed friction caused by
tidal
the gravita
tional pull of Jupiter and
that active volcanoes may
diction was supported by
images sent back by
Voyager 1 on M arch
9,1979, which indicated
that the surface of this
planet was n o t cratered as
had been expected.
Subsequently, M orabito et.
al. (1979) discovered a
huge volcanic eruption on
Io that was obliterating im
pact craters and all other
alia, Lysithea,
topographic features in a
11 million km
area surrounding the
volcano. Eventually, nine
active volcanoes
were identified and
deities associated w ith surface
M arduk, M asubi, M aui,
Pele, Prom etheus, Surt,
and Voland). The volcanic
plum esis contain sulfur
and the lava flows
may be com posed of
liquid sulfur, consistent
with the fact that the sur
1 0 0
km thick and is underlain by a m antle o
face of Io is yellow to red uid w ater betw een 400 and 800 km deep.
in color. The surface heat w ater may have intruded
flow of Io is about 48 m the crust locally in the
icrocalories/cm 2/sec and form of ice slush to
is therefore about 30 tim produce intrusive bodies
es greater than that of the of w ater ice.
E arth . Io is clearly the m
ost volcanically active Callisto is the outerm
object in the solar system . ost of the G alilean satel
The o th er G alilean lites. It is darker in color
(E uropa, G anym ede, and than the others and has a
densities than Io and are heavily cratered icy crust
m aterial with crusts of w about 200 km thick. A
ater ice and m antles of prom inent feature on its
liquid water. They are not surface is a very large m
volcanically active at the ultiringed basin called
present time, and their Valhalla whose diam eter
surfaces are cratered. E is nearly 2000 km. A
uropa appears to be com second ringed basin near
pletely covered by a frozen the north pole is called
ocean 75-100 km deep. Asgard. The icy crust may
The icy crust may be be
underlain by liquid w ater underlain by a liquid m antle about 1 0 0 0
that does not freeze com posed of water. Callisto apparently
because of heat generated inactive very early in its history p
nam by
ed tidal friction. The the am ount of heat generated by ti
fire of E uropa is less than that of the other G alilean
crisscrossed by a m
ultitude of curving bands,
some of which have been
traced for m ore than 1000
km . The bands appear to
be fractures in the crust
caused by internal
tectonic activity and by
fractures w ere subsequently
crustal w ater that froze to form
G anym ede is larger than the planet M ercury
and appears to be com posed of w ater and
cate m aterial in about
surface is com posed of
im purities that cause it to d arken
dark terrains are fragm ented and m ore heavily
cratered than the light-
colored terrain in which
th e y a re em b ed d
ed . T he ice c ru st is ab
out
28 THE SOLAR SYSTEM

Table 3.4 Physical

Properties of the Major Satellites of Jupiter, Saturn,
Uranus, and Neptune
Mean distance Diameter, Mass, Density,
Name from planet, km km g g/cm3
Jupiter
Io 421,600 3,640 8.91 X 1025 3.53

Europa 670,900 3,130 4.87 x 1025 3.03

Ganymede 1,070,000 5,280 1.49 x 1026 1.93
Callisto 1,883,000 4,840 1.06 X 1026 1.79
Saturn
Mimas 188,000 390 (3.7 0.1) X 1022 1.2

Enceladus 240,000 500 (7.4 3.4) X 1022 1.1

Tethys 297,000 1,050 (6.26 0.17) X 1023 1.0
Dione 379,000 1,120 (1.05 0.03) X 1024 1.42
Rhea 528,000 1,530 (2.16 0.7) X 1024 1.3
Titan 1,221,000 5,140 (1.348 0.017) X 1026 1.9(+)
Hyperion 1,502,000 -290
Iapetus 3,559,000 1,440 (2.8 0.7) X 1024 1.2
Phoebe 10,583,000 160
Uranus

Miranda 130,000 470 7.34 X 1022 1.35

Ariel 192,000 1,150 1.32 x 1024 1.66

Umbriel 267,000 1,170 1.26 X 1024 1.51
Titania 438,000 1,580 3.47 x 1024 1.68
Oberon 586,000 1,520 2.90 X 1024 1.58
Neptune

Triton 354,000 3,000 9.273 x 1025

Nereid 5,510,000 940 1.3 X 1018

Data from Glass (1982), Greeley (1985), Hunt and Moore (1981), Hamblin and Christiansen (1990),
and fact sheet, Jet Propulsion Laboratory.
addition, less heat is
generated by radioactivity
because its rocky core is
relatively small.
The satellites of
Saturn, U ranus, and N
eptune are identified in
Table 3.4. A ll of them
were seen during flybys system th at has recorded
of Voyager 2. We will m
en tion only Titan, the
largest of the satellites of
Saturn. It has a dense atm
osphere com posed prin
cipally of m ethane,
nitrogen, and sm aller am
ounts
of other gases including
ethane, acetylene, ethyl
ene, and hydrogen
cyanide. Like the Gallilean
satellites of Jupiter, each
of the satellites of Saturn,
U ranus, and N eptune is a
unique body in the solar
on its surface a history of
events caused by internal
and external processes. Im
ages and rem ote-sensing
data of these satellites w
ere received only recently
during
 3.4 PICTURES OF O UR SOLAR SYSTEM 29

Distance from the Center of Jupiter, 105 km

Figure 3.2 A: Variation of the density of the Galilean satellites of Jupiter with increasing distance from the planet.
The decrease in density is caused by increases in the proportion of water relative to silicate material. B: The Galilean
satellites magnified 50 times relative to the distance scale. Amalthea is much smaller than Io but appears to be a
silicate object.
the flyby of the Voyager
space probes in 1981
(Saturn), 1986 (U ranus),
and 1989 (N eptune).
3.4 Pictures
of Our
Solar
System
The exploration of the
solar system relies prim
ar ily on unm anned space
probes th at have the
capa bility of taking
pictures of planetary
surfaces and of returning
the images to E arth . The
two Viking landers and
the Pathfinder mission on
Mars, as well as the two
Voyager spacecraft and
the space craft G alileo in
the Jupiter system are
spectacular examples of
this technique, w hereas
the M oon has been
explored prim arily by A
m erican astro nauts who
actually w alked on its
surface.
The exploration of
the new worlds, which are
suddenly w ithin our
reach, will becom e an im
por tant task of the com m
unity of E arth Scientists.
A lthough an
understanding of these
new worlds m ust ultim
ately be based on studies
of the chem ical com
positions of m atter and of
reactions and processes
that take place on them ,
images of the surfaces of
planets help to identify the
problem s that need to be
solved.
For this reason, we
now exam ine som e of
the images of our solar
system to confront the
chal lenge that lies ahead .
Excellent photographs of
landform s on planetary
surfaces appear in text
books by H am blin and
Christiansen (1990) and
by G reeley (1985).
30 THE SOLAR SYSTEM

The Moon is a familiar image in the sky. Its surface is pockmarked with craters formed by impacts of meteoroids, which
continue to fall, though at a greatly reduced rate. The lunar landscape consists of dark plains, called mare (singular) and
maria or mares (plural), and bright mountainous highlands. The mare basins, formed by impacts of large objects during
the early history of the Moon, are filled with sheetlike basalt flows. The highlands are older than the mare basins and are
composed of anorthositic gabbro. The surface of the Moon is covered by a layer of regolith (colloquially called soil) that
consists of rock and mineral particles, beads of impact glass, and chunks of regolith breccia.

This view of the Moon was taken from space by the astronauts of Apollo 17 in December 1972. It shows the roughly
circular, black Mare Crisium in the left upper quadrant. Southwest of Mare Crisium is the Mare Fecunditatis, and
directly west of Crisium is the Mare Tranquilitatis. The Mare Serenitatis is located northwest of Tranquilitatis and
extends northwesterly beyond the horizon. (Photo by NASA.)
 3.4 PICTURES OF OUR SOLAR SYSTEM 31

In the late evening of July 20,1969 (EDT on Earth), Neil A. Armstrong and Edwin E. Aldrin descended from their
spacecraft and set foot on the surface of the Moon. They had landed near the southwest margin of the Sea of Tranquility
(Mare Tranquilitatis). The picture shows Edwin Aldrin on his way down just before he stepped onto the lunar surface.
The dusty plain in the background contains scattered boulders ejected from craters exca vated by impact of meteoroids.
(Photo by NASA.)
32 THE SOLAR SYSTEM

The surface of the planet Mars comes closest among all the planets in the solar system to the landscapes of Earth. Mars
has an atmosphere composed of N2 and CO, with a small amount of water. It also has roughly circu lar plains called
planitia and highly cratered highlands resembling those of the Moon. Mars has been an active planet, as indicated by
large shield volcanoes and rift valleys. In some places on Mars the surface is dissected by valleys in dendritic patterns
similar to stream valleys on Earth. Therefore, there is reason to believe that liquid water has existed on the surface of
Mars and that ice, in the form of permafrost, may still occur on Mars at the present time.

The picture shows the summit of the volcano Olympus Mons protruding through clouds on a frosty morning on Mars
much like Mauna Loa on the island of Hawaii. The summit contains several overlapping calderas whose presence
suggests a long history of volcanic activity. The volcano is 550 km in diameter at its base, and it rises 25 km above the
surrounding plain far higher than any mountain on Earth. (Photo by NASA.)
 3.4 PICTURES OF O U R SOLAR SYSTEM

The Valles Marineris on Mars, which are probably rift valleys, extend more than 2400 km in
an east-west direction near the martian equator. Some of the valleys are up to 200 km wide
and 7 km deep. The walls of the valleys have been extensively modified by slides and by
erosional channels. The rocks into which the valleys are cut are layered and may be sheetlike
flows of basalt. (Photo by NASA.)
34 THE SOLAR SYSTEM

The first Viking lander touched down on Mars in Chryse Planitia on July 20,1976. As shown, this area is a stony desert
containing small sand dunes and angular boulders. The view at the landing site of Viking 2 in the Utopia Planitia is quite
similar, indicating the importance of wind in shaping the surface of Mars. Some of the boulders are pitted or vesicular,
perhaps because they originated from underlying lava flows. (Photo by NASA.)
 3.4 PICTURES OF OUR SOLAR SYSTEM 35

Jupiter is by far the largest and most massive of the planets in the solar system. It has a turbulent atmosphere that
contains several cyclonic storm centers, including the Great Red Spot. The atmosphere is composed pri marily of
hydrogen and helium with small amounts of other elements of higher atomic number. Jupiter is not massive enough to
initiate hydrogen fusion in its core. Nevertheless, it has several sets of satellites, and some of the satellites are
comparable in size to the Moon of the Earth. (Photo by NASA.)
36 THE SOLAR SYSTEM

^ 4 fit

<& if

Io is the innermost of the so-called Galilean satellites of Jupiter. It is probably composed of silicate rocks like the Moon
and Mars, but its many active volcanoes emit sulfur-rich lavas and gases. The surface of Io is not cratered because impact
craters are quickly buried by lava flows and volcanic ash. The heat that causes the volcanic activi ty is generated by tidal
forces caused by the gravitational fields of Jupiter and Europa, the neighbor of Io. (Photo by NASA.)
 3.4 PICTURES OF O UR SOLAR SYSTEM 37

Europa is the second of the Galilean satellites of Jupiter. Unlike Io, Europa is encased in a layer of water ice beneath
which liquid water may be present. The icy surface is cut by sets of lines that may be fractures filled with ice-dikes
from below. Impact craters are present but are not common, presumably because of periodic renewal of the surface of
extrusion of water through the system of fractures. The interior of Europa consists of silicate rocks and a small dense
core. Ganymede and Callisto, the other two Galilean satellites, are also ice-covered, but each is distinctly different from
its neighbors. (Photo by NASA.)
38 THE SOLAR SYSTEM

Saturn is justly famous for its intricate rings, although all of the gaseous planets are now known to have rings. Saturn,
like Jupiter, has a retinue of satellites, the largest of which is Titan. This satellite is only slightly smaller than Ganymede
and has an atmosphere composed mainly of nitrogen with some methane and other hydrocar bon gases. Because of the
low surface temperature on Titan, most of the methane (or ethane, CfjHg) may form oceans of liquid hydrocarbons.
(Photo by NASA.)
 3.4 PICTURES OF O UR SOLAR SYSTEM 39

Neptune, the last of the great gaseous planets, was photographed in August 1989 by Voyager 2. It is blue in color and has
a turbulent atmosphere like that of Jupiter in spite of its great distance from the Sun (more than 4 x IO9 km) and its
much smaller size compared to Jupiter (about 6%). Neptune has at least eight satellites, Triton being the most massive.
The surface temperature of Triton is only 37 K or -2 3 6 C. Nevertheless, cryovol-canic activity is occurring on Triton
involving the eruption of liquid N2 and perhaps methane. Consequently, the surface of Triton contains few impact
craters and is composed of landforms never seen before on any satellite. (Photo by NASA.)
40 THE SOLAR SYSTEM

Planet Earth is our home in the solar system. It is the only planet or satellite with liquid water on its surface and with an
atmosphere containing molecular oxygen. Earth is also the only place in the solar system that can sustain life as we know
it.
This view of the southern hemisphere showing Africa, the Indian Ocean with Madagascar, and the Atlantic Ocean
was taken from space by the astronauts of Apollo 17, December 1972. (Photo by NASA.)
 3.4 PICTURES OF O UR SOLAR SYSTEM 41

Manicouagen-Mushalagen Lakes, north shore, St. Lawrence River, Quebec, Canada. These two lakes form a cir cular
structure because they are the remnants of a deeply eroded impact crater. This and hundreds of other such impact craters
testify to the fact that the Earth has been bombarded by meteoroids and comets just as have the Moon, Mars, and all
other bodies in the solar system. In fact, such impacts are now recognized as an important geological process that has
disrupted the surface environment by causing short-term catastrophic climate changes and biological extinctions. (Photo
by NASA.)
42 THE SOLAR SYSTEM
3.5 Summary
The sequence of events leading to the form ation of the
solar system can be reconstructed as a direct extension of
stellar evolution by applying the laws of physics and chem
istry to a diffuse cloud of gas and dust particles in
interstellar space.
The earthlike planets constitute a very small fraction
of the total m ass of the solar system and are dw arfed even
by the outer gaseous planets. Nevertheless, E arth is the
only planet in the entire solar system on which the surface
environ
m ent is conducive to the developm ent and evolu
tion of life forms.
The satellites of Jupiter form a m iniature
planetary system of their own. The four largest
satellites are similar in size to M ercury and the
M oon but differ significantly in their chemical
com positions and surface features.
The satellites of Saturn, Uranus, and N eptune are
likewise of great interest in the study of the solar system
but are less well know n than the satellites of Jupiter.
References
ANONYMOUS, 1989. Voyager
Fact Sheet; Neptune
Encounter.
Jet Propulsion Laboratory,
Pasadena, CA.
C A M E R O N , A . G . W ., 1978.
Physics of the primitive solar
accre
tion disk. Moon Planets,
18:540.
C A M E R O N , A. G. W., and M. R.
Pine, 1973. Numerical models
of the primitive solar
nebula. Icarus, 18:377-106.
C A R R , M. H., R. S. S A U N D E R S ,
W IL H E L M S , 1984. The Geology
NASA SP-469,317 pp.
G L A S S , B. P., 1982 . IntroductionM u r r a y , B., M. C. M a u n ,
Cambridge University
Press, Cambridge,
England, 469 pp.
G R E E L E Y , R., 1985. Planetary
Landscapes. Allen and Unwin,
London, 265 pp.
H A M B L IN , W. K., and E. H. C H R
IS T IA N S E N , 1990. Exploring the
Planets. Macmillan, New
York, 449 pp.
H U N T , G , and P. M oore,
1981. Jupiter. Rand McNally,
New
York, 96 pp.
L E W IS , J. S., 1974. The
chemistry of the solar system.
Sei.
Amer., 230(3):51-65.
M E H L I N ,T . G., 1968.
Astronomy and the Origin o f
the Earth,
Brown Co., Dubuque, IA,
131 pp.
M o r a b i t o , L. A., S. P. S y n n o t t ,
C o llin s , 1979. Discovery of currently active extraterres
trial volcanism. Science.
204:972.
and R. G r e e l e y , 1981.
Earthlike Planets. W. H.
Freeman, San Francisco,
387 pp .
P E A L E , S. J., P. C A S S E N , and R .
T. R E Y N O L D S , 1979. Melting o f
Io by tidal dissipation.
Science, 203:892 -894 .
 4
Chemical Differentiation
of the Earth
The E arth is a highly differentiated planet, and its
geological activity continues to produce diver
sified suites of igneous and sedim entary rocks.
A lthough som e of the other terrestrial planets
m ay also have rem ained active in certain limited and
specific ways, the E arth still has fire in her belly and is
cloaked in envelopes of corrosive w ater and atm
ospheric gases.
4.1 Internal Structure of the Earth
The E arth was probably initially m olten and dif
ferentiated very early in its history into several layers or
shells having different chemical com po
sitions. The resulting internal structure of the
E arth has been determ ined by seismologists
based on the reflection and refraction of com-pressional
(P) and shear (5) seismic waves. On the basis of these
results, the interior of the E arth has been divided into the
core, the mantle, and the crust, as shown in Figure 4.1.
The densities of
these m ajor interior units of the E arth range
.8
from 2 g/cm 3 for rocks in the crust to m ore than
12 g/cm 3 for the core (Bullen, 1963).The increase
of the density and of the velocities of seismic waves with
depth is caused both by changes in the chemical com
position and by the recrystal lization of m inerals into m
ore closely packed structures. Figure 4.1 shows th at the
seismic velocities and densities have two m ajor disconti
nuities m arking the boundaries betw een the crust and m
antle and betw een the m antle and core. The crust-m
antle boundary, located at a depth of
about 40 km under the continents, is known as the M
ohorovicic discontinuity or M oho, whereas the m
antle-core boundary, at a depth of 2883 km, is referred to
as the W iechert-G utenberg discon tinuity, and both are
nam ed after em inent seis mologists. There are
additional, but m ore subtle, transition regions within the
E arth caused by other im portant changes in its physical
properties. Table 4.1 contains a sum m ary of the physical
properties of the E arth.
The m antle of the E arth has been subdivided into
three parts based on the presence of bound aries at depths
of 413 and 984 km . The upper m an tle, between the M
oho and the boundary at about 400 km, is probably
heterogeneous because of the form ation of m agm a by
partial melting. Xenoliths of ultram afic rocks derived
from this region indi cate that the upper m antle is com
posed of olivine, pyroxene, and garnet with lesser am
ounts of spinel, amphibole, and phlogopite. The transition
zone, located betw een depths of about 400 and
1 0 0 0
km, is a region of pressure-induced phase
transform ations of the orthosilicates of Fe and Mg from
the structure of olivine to that of spinel. The lower
mantle, below a depth of about 1 0 0 0 km, is probably hom
ogeneous and m ay be richer in Fe than the upper m antle
(H enderson, 1982).
The chem ical com position of the m antle is difficult
to determ ine with certainty because it is inaccessible and
heterogeneous. N evertheless, ultram afic xenoliths, as
well as the stony m ete orites, have provided useful clues.
In addition, large masses of alpine-type peridotite m ay
have originated from the upper m antle at convergent
43
 44 CHEM ICAL DIFFERENTIATION OF THE EARTH
Figure 4.1 A: Variation of seis

mic velocities with depth in the

Earth. The velocity of P waves
increases abruptly at the base of
the crust (Mohorovicic disconti
nuity or Moho) from about
6 km /sec to nearly 8 km/sec. It
increases with depth in the man
tle but drops sharply again at the
mantle-core boundary. A third
discontinuity in the E-wave
velocities within the core marks
the transition from the liquid
outer core to the solid inner core.
The velocity of S waves also
increases at the Moho and con
tinues to rise with depth in the
mantle. However, S waves do not
penetrate the mantle-core
Depth, km boundary, which indicates that
the outer core is liquid.
B: The density of rocks rises
abruptly at the Moho and contin
ues to increase with depth in the
mantle. At the mantle-core
boundary the density rises
sharply from about 6 g/cm3 to
nearly 10 g/cm 3, and then contin
ues to increase to values in
excess of 12 g/cm 3 at the bound
ary between the outer and inner
core. The variation of these phys
ical properties leads to the con
clusion that the mantle is solid
and is composed primarily of sili
cates and oxides of Mg and Fe.
The core consists of an alloy of
Fe, Ni, and other siderophile ele
ments and is liquid except for the
Depth, km inner part which is solid (after
Bullen, 1963).

plate boundaries. E stim ates of the chem ical basaltic m agm as by partial m elting under the
com position of the m antle m ust m eet certain physical conditions likely to prevail in th at
geophysical req u irem en ts regarding density, region of the E arth .
seism ic velocities, and h eat production by the These requirem ents are generally satisfied by
decay of U, Th, and K. In addition, the m antle a rock calledp y r o lite invented by Ringw ood (1966)
m ust be capable of form ing th e m ajor types of by combining peridotite and basalt in the ratio of 3
 S i0 2

Table 4 .2 E s tim a te s o f th e C h e m ic a l

C o m p o sitio n o f th e M a n tle o f th e E a r th

MgO
FeO
A I 2O 3

CaO
NazO
Cr20 3
MnO
P2Os
K 2O

T i02
NiO
Sum

Pyrolite (Ringwood, 1966).

bMantle plus crust based on meteorites (Mason, 1966).

'Undepleted mantle based on lherzolites (Hutchison, 1974).

to 1. The resulting chemical com position of the m antle

is similar to one proposed by M ason (1966), which was
based on m eteorites, and also resem bles the estim ate m
ade by H utchison (1974), which was derived from
ultram afic lherzolite inclusions in igneous rocks. The
estim ates of the chemical com position of the m antle
are compiled in Table 4.2.
Table 4.1
C ru st

Whole Earth
Core
Mantle
Crustb
Hydrosphere
Atmosphere
Radius.

bContinental crust.
SO U R C E : Parker (1967) and Taylor and McLennan (1985).

Sample 13

45.2
37.5
8.0
3.5
3.1
0.57
0.43
0.14
0.06
0.13
0.17

98.8
46 CHEMICAL DIFFERENTIATION OF THE EARTH
 up the continents but also addition, transport at the
upper mantle could be, and occurs under the oceans. but their occurrence at th e base of the ice sheet and
in fact had been, recov The oceanic crust is only surface of the conti subsequent deposition by
5 - 8 km thick nents enhances their im portance.rnelt-w ater caused glacial
ered by dredging.
and differs from th e The chem ical
continental crust in chem i com position clay to be well mixed,
cal com position and origin . T he tal crust has been estim making it a representative
ated
large num bers of chem sam ple of the rocks in the
ical
4.2 The crust is com posed prim arily of igneous
path of the ice sheet.The
m etam o rp h ic rocks. bining
M etam o rp h ic chem ical analyses
Co concentrated in orogenic types w eighted in term s average chemical com
of
nti belts, but the distinc (3) analyzing sedim ent derived position of 77 sam ples of
tion betw een igneous nent, (4) com bining the com positions glacial clay from sou th
nen is not always obvious and m afic rocks in varying ern Norway (colum n 3,
tal igneous rocks in the (5) by m odeling. The Table 4.3) is, in fact, quite
results
continental crust m ethods are surprisingly sim ilar to the estim ates
Cru have of C larke and W
sim ilar as indicated in
st: been m etam orphosed . S edim Tableentary
4.3. ashington. We note,
canic rocks form at th e however, that the
Maj and th erefore tend to
C larke and W
concentrations of C aO
ashington (1924) estim
or igneous and m etam orphicated the chemical com and
and volcanic rocks m ay
Ele position of the crust not
into elongated m asses by only by averaging a large
me to n ic basins in w hich num ber of chemical
deposited . C larke and W ashington Na20 of the glacial clay are som ew h
nts m ated that the continental
analyses of igneous rocks
crust their values in colum ns 1 and 2
from all continents and
The crust of the E arth is 95% igneous and m etam o rp h ic because these elem ents w ere parti
from the ocean basins, but
im portant because it con 5% sed im en tary rocks. also by com bining the the clay by leaching. In
tains all of the natural 4% shale, 0.75% sandstone, and 0.25%
average chemical com addition, the glacial clays
resources th at sustain us. stone. Evidently, sedim entary
position of igneous rocks have a low er T i0 2
It only a m inor p o rtio n of with those of shale, content than the crustal
includes the atm osphere, h sandstone, and lim estone aver ages of Clarke and W
ydrosphere, and bio s ashington, which m ay be
in accordance with their
phere, as well as part of the a regional characteristic of
observed abundances. The
lithosphere. The term the rocks of Scandinavia.
results in columns 1 and 2
lithosphere is now defined
of Table 4.3 are very
as the rigid outer por tion
similar. An enorm ous am
of the E arth consisting of
ount of work was required
the crust and upper m antle
to collect and analyze the
in contrast to the
well over 5000 rock sam
underlying asthenos-phere,
ples on which th e estim
which deform s plastically
ate in colum n 1 is
in response to tectonic
based.Therefore, V. M. G
stress. The study of the
oldschm idt decided in the
chem ical com po sition of
early 1930s to analyze the
the continental crust, and
clay-size fraction of till in
of the different types of
southern Norway because,
rocks of which it is com
he argued, glacial sedim
posed, has been a
ent of Pleistocene age was
m ajo r objective of geochem istry
produced by m echanical
1969; Taylor and M cLennan, 1985).
ero sion of bedrock in
The crust is the o u ter shell of
Scandinavia with little
w hich lies above the m antle. It not only
chemical w eathering. In
 4.2 THE CONTINENTAL CRUST: MAJOR ELEMENTS47
Table 4.3 E s tim a te s o f th e A v e ra g e C h em ical C o m p o sitio n o f th e C o n tin e n ta l C ru st in

W e ig h t P e rc e n ts
r 2b 3C 4d 5e 6< 7% <Sh

S i0 2 59.12 59.07 59.19 60.06 59.4 59.3 57.3 68.4

T i0 2 1.05 1.03 0.79 0.90 1.2 0.7 0.9 0.4

A I 2O 3 15.34 15.22 15.82 15.52 15.6 15.0 15.9 14.8
Fe20 3 3.08 3.10 3.41 3.55 2.3 2.4 1.3
FeO 3.80 3.71 3.58 4.06 5.0 5.6 9.1 3.2
MnO 0.12 0.11 0.11 0.21 0.1 0.1
MgO 3.49 3.45 3.30 3.56 4.2 4.9 5.3 1.7
CaO 5.08 5.10 3.07 5.62 6.6 7.2 7.4 3.4
NazO 3.84 3.71 2.05 3.28 3.1 2.5 3.1 3.1
K 2O 3.13 3.11 3.93 2.88 2.3 2.1 1.1 3.6
P 2O 5 0.30 0.30 0.22 0.36 0.2 0.2 0.1
co2 0.35 0.54
H 2O 1.15 1.30 3.02
Sum 99.50 99.56 99.03 100.00 100.0 100.0 100.1 100.0

Clarke and Washington (1924) based on the average of 5159 analyses of igneous rocks from all continents and the oceans. bClarke
and Washington (1924) based on 95% igneous rocks, 4% shale, 0.75% sandstone, 0.25% limestone.
'Goldschmidt (1954) based on 77 analyses of glacial clay from southern Norway. dDaly
(1914) based on 1 :1 mixture of average granite and basalt.
'Poldervaart (1955), based on the average of young folded belts and continental shield region shown in Table 4.4. fRonov
and Yaroshevsky (1976) based on a detailed model like that of Poldervaart (1955).
eTaylor and McLennan (1985,Table 3.4, No. 10) based on 75% Archean crust and 25% andesite crust.
hSederholm (1925) based on the Precambrian rocks of Finland.
S O U R C E : Parker (1967);Taylor and McLennan (1985).

continental crust
 Columns 4 -8 of crust as a whole.
Table 4.3 contain Poldervaarts estimates of
estimates of the chemical the chemical com position
composition of the of each of the four
continental crust derived structural units are shown
by various methods. Daly separately in Table 4.4.
(1914) estim ated its These results illustrate the
composition by combining differences in the
chemical composi
analyses of granite and tion
(continental
of the

basalt in equal proportions

(column 4,Table 4.3) shield + young folded belts)
because these rock types and the oceanic crust (deep
were thought to originate oceanic region). It is
from fundam ental evident that the oceanic
magmas of granitic and crust has higher average
basaltic compositions. concentrations of Fe2 0 3 +
Poldervaart (1955) FeO, MgO, CaO, MnO, P2
O s, and T i0 2 and lower
constructed a detailed
concentrations of S i0 2,
model by dividing the
A12 0 3, N a2 0 , and KzO
crust into the con ti nental
than the continental crust
shields, young folded belts,
(column 5, Table 4.4).
the suboceanic region, and These differences arise
the deep oceanic region. H because the continental
e assigned each region crust is composed primarily
certain lithologic of granitic rocks
compositions (indicat ed (granodior-ite?), w hereas
in Table 4.4) and the oceanic crust was
combined the chemical assumed by Poldervaart
com po sitions of the (1955) to consist largely of
different rock types olivine basalt. However, the
according to their study of ophiolite
abundances. Column 5 of complexes,
Table 4.3 contains his esti
m ate of the chemical
composition of the
continental
48 CHEM ICAL DIFFERENTIATION OF THE EARTH
Table 4 .4 E s tim a te s o f th e C h em ical C o m p o sitio n o f F o u r M a jo r S tru c tu ra l U n its o f

th e L ith ic C ru st o f th e E a r th in W eight P e rc e n t
Poldervaart Ronov and Yaroshevsky

1 2 3 4 5
Continental Young Suboceanic Deep Oceanic
shieldsa folded beltsb regions: oceanicd crust
S i0 2 59.8 58.4 49.4 46.6 49.4

T i0 2 1.2 1.1 1.9 2.9 1.4

AI O 3 15.5 15.6 15.1 15.0 15.4
2

Fe20 3 2.1 2.8 3.4 3.8 2.7

FeO 5.1 4.8 6.4 8.0 7.6
MnO 0.1 0.2 0.2 0.2 0.3
MgO 4.1 4.3 6.2 7.8 7.6
CaO 6.4 7.2 13.2 11.9 12.5
Na20 3.1 3.1 2.5 2.5 2.6
K 2O 2.4 2.2 1.3 1.0 0.3
P 2O 5 0.2 0.3 0.3 0.3 0.2
Sum 100.0 100.0 100.0 100.1 100.0

Sedimentary cover 0.5 km: 41% shale, 43% sandstone, 16% limestone. Igneous rocks vary in composition with depth
such that there are 22 km of granodiorite, 3 km of diorite, and 10.5 km of basalt.
bSedimentary cover 5 km: 52% shale, 13% sandstone, 22% limestone, 5% greywacke, 6% andesite, 2% rhyolite. Igneous
rocks: 22.5 km of granodiorite, 2.7 km of diorite, and 7.3 km of basalt. In addition, 40% of the sediment layer is assumed to be
granodiorite based on the outcrop area of granitic batholiths.
cShelf sediment 4 km, similar in composition to that of folded mountain belts. In addition: 74% terrigenous mud, 22%
coral mud, 4% volcanic mud. Igneous rocks: 4 km of diorite, 2 km of tholeiite, and 5 km of olvine basalt.
dSediment thickness 0.3 km: 72% calcareous sands and oozes, 19% red clay, 9% siliceous ooze. Igneous rocks: 5.57
km of olivine basalt.
S O U R C E : Poldervaart (1955);Ronov and Yaroshevsky (1976).
summarized by M oores
(1982), indicates that the
oceanic crust is actually
com posed of mafic and
ultramafic rocks, whose
average composition may
approach that of basalt,
although trace elem ent
con centrations have been
altered by interaction with
heated seawater. Ronov
and Yaroshevsky (1976)
also m odeled the
continental and oceanic
crust and arrived at
compositions that are
similar to those of
Poldervaart (1955). Their
estim ate of the chemical
composition of the
continental crust is listed
in col um n 6 of Table 4.3,
w hereas column 5 of
Table 4.4 contains their
estim ate for the oceanic
crust.
Taylor (1964) revived
D alys m ethod of esti (1985) is som ew hat m
mating the chemical com
position of the continental
crust by dem onstrating
that the average concentra
tions of rare earth elem
ents (R E E s) in sedim
entary
rocks can be duplicated by
combining chemical
analyses of mafic and
felsic igneous rocks in the
proportions 1:5. The
resulting chemical com
posi tion resembles that of
diorite or andesite, which
is
why this crustal m odel is generally referred
the andesite m odel. Subsequently,
M cLennan (1985) proposed that the
crust consists of 75 %
average A rchean rocks
and 25% andesite crust.
The resulting com position
is shown in column 7 of
Table 4.3. It differs from
all other estim ates listed
in Table 4.3 by having som
e
what lower concentrations of S i0
K20 and higher concentrations
FeO, MgO, and CaO.
Accordingly, the chemical
composition of the
continental crust proposed
by Taylor and M cLennan
ore
4.3 DIFFERENTIATION OF IGNEOUS AND SEDIMENTARY ROCKS 49
 45.5 high-Ca granites
mafic than those of their
o on the extent of melting, represent igneous rocks of
Si 26.8
predecessors. However, it which is controlled by the interm ediate com
Al 8.40
does m eet the following pressure and tem perature position, which may not
Fe 7.06 at the source. The com actu ally be granites in the
im portant constraints.
Ca 5.3 position of magmas may strict sense of th at term .
1. It satisfies the All of the rem aining elembe
entsmodified subsequently The sedim entary
observed rate of by assimilation of country rocks are just as highly
heat flow based rocks or by mixing with diversified in th eir chem
very small fraction of the
on the decay of crust of the Earth. magmas derived from ical com positions as
U ,Th, and K. different sources. The igneous rocks. H ow ever,
2. It is capable of crys tallization of magmas the processes th at cause
generating the m depends on their chemical this diversification operate
ore highly
4.3 Differ compositions and on the on the surface of the E arth
differentiated entiat physical conditions in the rath er than at depth and
include weather
granodiorites of ion of magm a chamber. The
ing, transport,
the upper crust by rocks produced by
partial melting.
Igneo fractional crystallization of
deposition, and
3. It restricts crustal us m agm a may vary widely lithification. The
in their m ineral com
grow th by and position from ultram afic
study of these processes encom pas
andesitic island portion of the subject of geochem is
-arc volcanism to
Sedim cum ulates of early-form occupy the rem ainder of this book.
ed ferrom agnesian m
post-A rchean entar Sedim entary rocks can be divided
inerals (olivine or groups depending on w
time. y pyroxene) to quartz-rich hether they are com posed
4. It recognizes th at
75% of th e Rocks late-stage differentiates. of m ineral and rock
continental particles of preexisting
The chemical The most com rocks or w hether they
crust is com compositions of igneous prehensive com pilation of precipitated from aqueous
posed of rocks rocks vary widely because the chem ical com solu
th at are m ore of geochemical positions of the highly tion. This division is n o t com pletely
than 2500 differentiation that takes diverse igneous rocks can because clastic sedim entary rocks ma
million years place during their form be found in the H a n chemically precipitated m
ation. The differentia tion dbook o f G eochemistry,
old, which did ineral cements, w hereas
starts during m agm a edited by K. H . W
not the chemical precipitates m
form ation by partial m edepohl from 1969
necessarily form by elt ing of rocks in the ay contain a com po nent
to 1978. The com pilation of data
magm atic processes of detrital sedim ent. M
lower crust or upper based on earlier com pilations by
tant in post-A rchean oreover, the chemical com
mantle. The chemical com W edepohl (1961) andV
positions of rocks in both
In conclusion, we position of the resulting inogradov (1962).The data
are grouped into ultram groups are modified
note that the chemical magm a depends on the
com position of the composition of the source afic rocks, basalt, high-Ca
continental crust of the E rocks and granite, and low -Ca
arth is very different from granite and therefore rep
th at of the solar nebula re sent m ost of the com
and positional spectrum of
stony m eteorites. The nine m ost abundant igneous rocks. The
m ents and their average concentrations category basalt includes
continental crust are: both volcanic and plutonic
Wt.% rocks of basaltic com
position, w here as the
50 CHEM ICAL DIFFERENTIATION OF THE EARTH

Table 4.5 C h e m ic a l C o m p o sitio n o f Ig n e o u s a n d S e d im e n ta ry R o c k s (in p a rts p e r m illio n

u n le ss o th e rw ise in d ic a te d )

Element, Z Ultra-mafic* High-Ca Low-Ca Sandstoneb Carbonate Deep-se

Basalf granitesb granitesb Shaleb rocks0 clay0
3 Li 0.5 16 24 40 66 15 5 57
4 Be 0.2 0.7 2 3 3 2.6
5B 2 5 9 10 100 35 20 230
9F 100 385 520 850 740 270 330 1300
11 Na (%) 0.49 1.87 2.84 2.58 0.96 0.33 0.04 4
12 Mg (%) 23.2 4.55 0.94 0.16 1.50 0.70 4.70 2.10
13 A1 (%) 1.2 8.28 8.20 7.20 8 2.50 0.42 8.40
14 Si (%) 19.8 23.5 31.40 34.70 7.30e 36.80 2.40 25
15 P 195 1130 920 600 700 170 400 1500
16 S 200 300 300 300 2400 240 1200 1300
17 Cl 45 55 130 200 180 10 150 21000
19 K (%) 0.017 0.83 2.52 4.20 2.66 1.07 0.27 2.50
20 Ca (%) 1.6 7.2 2.53 0.51 2.21 3.91 30.23 2.90
21 Sc 10 27 14 7 13 1 1 19
22 Ti 300 11400 3400 1200 4600 1500 400 4600
23 V 40 225 88 44 130 20 20 120
24 Cr 1800 185 22 4.1 90 35 11 90
25 Mn 1560 1750 540 390 850 1100 6700
26 Fe (%) 9.64 8.60 2.96 1.42 4.72 0.98 0.33 6.50
27 Co 175 47 7 1 19 0.3 0.1 74
28 Ni 2000 145 15 4.5 68 2 20 225
29 Cu 15 94 30 10 45 4 250
30 Zn 40 118 60 39 95 16 20 165
31 Ga 1.8 18 17 17 19 12 4 20
32 Ge 1.3 1.4 1.3 1.3 1.6 0.8 0.2 2
33 As 0.8 2.2 1.9 1.5 13 1 1 13
34 Se 0.05 0.05 0.05 0.05 0.6 0.05 0.08 0.17
35 Br 0.8 3.3 4.5 1.3 4 1 6.2 70
37 Rb 1.1 38 110 170 140 60 3 100
38 Sr 5.5 452 440 100 300 20 610 180
39 Y 21 35 40 26 40 30 90
40 Zr 38 120 140 175 160 220 19 150
41 Nb 9 20 20 21 11 0.3 14
42 Mo 0.3 1.5 1 1.3 2.6 0.2 0.4 27
47 Ag 0.05 0.11 0.051 0.037 0.07 0.01 0.01 0.11
48 Cd 0.05 0.21 0.13 0.13 0.3 0.035 0.42
49 In 0.012 0.22 0.26 0.1 0.08
50 Sn 0.5 1.5 1.5 3 6 0.1 0.1 1.5
51 Sb 0.1 0.6 0.2 0.2 1.5 0.01 0.2 1
53 I 0.3 0.5 0.5 0.5 2.2 1.7 1.2 0.05
55 Cs 0.1 1.1 2 4 5 0.1 0.1 6
56 Ba 0.7 315 420 840 580 10 10 2300
57 La 1.3 6.1d 45 55 92 30 1 115
58 Ce 3.5C 16d 81 92 59 92 11.5 345
59 Pr 0.49 2.7d 7.7 8.8 5.6 8.8 1.1 33
 4.3 DIFFERENTIATION OF IGNEOUS AND SEDIMENTARY ROCKS 51
Table 4 .5 (co ntinued)

High-Ca Low-Ca Carbonate Deep-sea

Element, Z Ultramafic Basalf granitesb granitesb Shaleb Sandstoneb rocksb clayb

60 Nd 1.9C 14d 33 37 24 37 4.7 140
62 Sm 0.42c 4.3d 8.8 10 6.4 10 1.3 38
63 Eu 0.14c 1.5d 1.4 1.6 1 1.6 0.2 6
64 Gd 0.54c 6.2d 8.8 10 6.4 10 1.3 38
65 Tb 0.12 l.ld 1.4 1.6 1 1.6 0.2 6
66 Dy 0.77c 5.9d 6.3 7.2 4.6 7.2 0.9 27
67 Ho 0.12c 1.4d 1.8 2 1.2 2 0.3 7.5
68 Er 0.30c 3.6d 3.5 4 2.5 4 0.5 15
69 Tm 0.041c 0.60d 0.3 0.3 0.2 0.3 0.04 1.2
70 Yb 0.38c 3.2d 3.5 4 2.6 4 0.5 15
71 Lu 0.036c 0.55d 1.1 1.2 0.7 1.2 0.2 4.5
72 Hf 0.4 1.5 2.3 3.9 2.8 3.9 0.3 4.1
73 Ta 0.5 0.8 3.6 4.2 0.8 0.01 0.01 0.1
74 W 0.5 0.9 1.3 2.2 1.8 1.5 0.6 1
79 Au 0.006 0.004 0.004 0.004
80 Hg 0.01 0.09 0.08 0.08 0.4 0.03 0.04 0.1
81 T1 0.04 0.21 0.72 2.3 1.4 0.82 0.01 0.8
82 Pb 0.5 7 15 19 20 7 9 80
90 Th 0.0045 3.5 8.5 17 12 1.7 1.7 7
92 U 0.002 0.75 3 3 3.7 0.45 2.2 1.3

Average of Turekian and Wedepohl (1961) and Vinogradov (1962).

bTurekian and Wedepohl (1961).
'Calculated from data listed by Herrmann (1970).
dAverage values calculated by Hermann (1970).
'Krauskopf (1979) listed a value of 23.8%.

by ion exchange processes associated with electri cal the E arth so rt the elem ents on the basis of their chem
surface charges of m ineral grains regardless of w hether ical properties. We can think of the surface environm ent
they are detrital or form ed as chemical precipitates. of the E arth as a giant m a chine that processes igneous
and m etam orphic rocks into a wide variety of sedim
The com pilation of data in Table 4.5 includes two entary rocks. The products of this m achine are highly
rock types th at are predom inantly detrital (shale and diversified because the chem ical elem ents, and the m
sandstone), one that is predom iriantly precipitated inerals they form , respond differently to the treatm ent
(carbonate rocks), and deep-sea clay, which is a m ixture they experi ence as they pass through the machine.
of detrital grains and chemical precipitates. The selection
of these rock types has an additional virtue because it The representation of geological processes
includes rocks that form in shallow m arine basins in the operating on the surface of the E arth as a
vicinity of continents as well as sedim ent deposited in m achine is helpful because it em phasizes the
deep ocean basins far from land. coherence of these processes and because it pro
m otes a global view of the geochemistry of the
The ap p aren t geochem ical differentiation indicated E a rth s surface. We return to this them e in P art V of this
by these data reflects the way in which geological book, w here we consider the geochemical cycles of
processes operating on the surface of selected elem ents (C hapter 22).
52 CHEMICAL DIFFERENTIATION OF THE EARTH
 w ater (0 .38% ).T hat oxygen bonds in the
leaves a very small minerals.
4.4 D fraction of the w ater in O nce an elem ent has
Table 4 .6 In v e n to ry o f W a te r in th
H y d ro s p h e re
if the hydrosphere in entered the hydros phere, it
Volume,
shallow ground is subjected to other kinds
f w ater (0 .30% ), lakes (0 of.01% ),
processes th at m ay Reservoir 106 few3
e (0 .0 0 0 1 % ) w here it is accessible to concentration .
affect its Oceans 1370
m unicipal and industrial use (see alsoexample,
C hapter som e ions
r For Ice sheets and glaciers 29
W hen igneous, m are selectively adsorbed on
e etam orphic, or sedim the charged surfaces of
Deep groundwater 5.3
(750-4000 m)
n entary rocks come in clay m inerals or particles
Shallow groundwater 4.2
contact w ith w ater at or of oxides and hydroxides
ti near the sur face of the E of Fe, M n, and A l. O th er (<750 m)
a arth, different kinds of elem ents m ay enter th e Lakes 0.125
chemical reac tions occur, biosphere as nutrients and Soil moisture 0.065
ti which collectively Atmosphere3 0.013
becom e associated with
o constitute chem ical w organism s and biogenic Rivers 0.001 7
eathering or w ater-rock
n interaction. The outcom e
carbon com pounds. The Biosphere 0.000 6
noble gases are released Total 1408.7
o of these reactions depends into the atm osphere, w Liquid equivalent of water vapor.
not only on the chemi cal
f hereas oxygen, carbon
Berner and Berner (1987).
and physical conditions at dioxide, and
SOURCE:

t the site but also on the

nitrogen of the atm
h properties of the minerals.
osphere dissolve in the w
M inerals vary w ide ly in ater.
e their susceptibility to Therefore, the chem ical com position
H chem ical w eathering varies widely and reflects not only the
although m ost are quite
y insoluble in w ater.
com position of the rocks it has in
but is also affected by the geochem
d Therefore, the chemical
m ent, which is determ ined by clima
elem ents do not enter the
r aqueous phase in the sam
phy, vegetation, and by the
possible discharge of
o e proportions in which industrial or m unicipal
they occur in the rocks of
s the continental crust. As a
waste. These m atters are
considered in m uch
p result, the chem ical com greater detail in P art V.
position of surface w ater
h Stream s ultim ately
or groundw ater differs transport the elem ents th
e from th at of the rocks of at dissolve in the w ater to
r the upper continental crust the oceans w here they
in ways th at reflect the reside for varying periods
e geochem ical properties of of time. The geochem i cal
T he hydrosphere consists the elem ents and of the m processes operating in the
of a num ber of reser voirs inerals they form . W oceans selectively rem ove
of w ater th at are hitfield and Turner (1979) elem ents from solution in
connected by m eans of such a way that seaw ater
actually expressed these differences
the hydrologic cycle. is not simply average river
partition coefficients (w a te r/u p p e r
Table 4.6 indicates that the w ater th at has been
showed that they are
oceans are by far the largest reservoirs of w ater concentrated by
related to the ionic
with 97.25% of the total volum e, followed by ice evaporation . This statem
character of the m etal-
sheets and glaciers (2.05% ) and deep ground ent is supported by
inspection of the average
concen trations of the
elem ents in stream s and
in the oceans (Table 4.7).
The fifth colum n in Table
4.7 contains enrichm ent
factors calculated by divid
ing the concentration of
each elem ent in the
oceans by its concentration
in average river w ater.
The resulting enrichm ent
factors vary widely from
3350 for B r to about
0.0006 for Th. Evidently,
geo chemical processes in
the oceans are changing its
 4.4 DIFFERENTIATION OF THE HYDROSPHERE

Table 4 .7 A v e ra g e C o m p o sitio n o f W a te r in S tream s a n d in th e O c e a n s in

m ic ro g ra m s p e r g ra m
 0.013
.merit Classificationa Stream Seawater
water enrichment 1.3 X 102
Li I 3 X 1(T3 M O R Tb 1.3 x 105
8
Be IV 1 x IO5 2.5 X 106
B I 1 x 1026.3 X 101
450 1.6 <4 .0 x 106
F I 1 X 1(T3
1300 7.9 4.0 X 104
Na I 6.3
1714 2.0 5.0 X 103
Mg I 4.1
0.094 7.9 X 101
A1 IV 5x 1(T2 315 5.0
0.044 3.2 X 101
Si II 6.5 0.016
P II 2 X 102 7.9 x 103
S I 3.7 4.0 X 104
Cl I 7.8 243 5.0 x
K I 2.3 2500 6.3 x
Ca I 15 1.3 x 107
Sc IV 4 X 106 1.3 x 106
Ti IV 3 X 1(T3 2.5 x 101
V IV 9 X 104 0.32 <
Cr IV 1x 1CT3
1
Mn III 7x 1(T3
Fe IV 4x 102 .
Co III 1x 10"4 6
Ni III 3x 1(T4
x
Cu III 7x 10"3
Zn III 2x 1(T2 1
Ga IV 9x 1(T5 0
Ge IV 5x 10 6 2
As IV 2x 1037.9 x 103
Se III 6x 1(T51.6 X 103
Br I 2x 1(T23.2 x 101
Rb I 1x 10"3 0.0015
Sr I 7x 102
Y III 4x 1(T5 6.9 x 10-1
Zr IV -- 2.0 x 101
Nb IV -- 1.6 x 103
Mo I 6 x 1(T41.0 x 103
Ag III 3 X 104 1.3 x 103
Cd IV 1 x 1(T57.9 x 102
In IV -- 1
Sn IV 4 x 1(T5
Sb III 7 x 10"5 2.0 x 103
I I 7 x 1031.0 x 105
Cs I 2 x 1(T56.3 X 10s
Ba III 2 X 102 3350 7.9
La III 4.8 x 1(T5 120 7.9
Ce III 7.9 x 1055.0 X 106
1.3 X 102
1.6 x 102
<2.5 x 102
3.2 x 105
0.009 2.0
8 7.9
1.0 x 103
54 CHEM ICAL DIFFERENTIATION OF THE EARTH

Table 4.7 (continued)

Element Classificationa Stream Seawater
Seawater
water enrichment M O RTb
Pr III 13 x IO '6 1.0 x 10 6
0.14 7.9 X 101
Nd III 3.8 X 1(T5 4.2 X 106 0.11 7.9 X 101
Sm III 7.8 X IO6 8.0 X 107 0.10 7.9 X 101
Eu III 1.5 X IO6 1.5 X IO7 0.10 6.3 X 101
Gd III 8.5 X IO6 1.0 X 106 0.11 1.0 x 102
Tb III 1.2 X IO"6 1.7 x 107 0.14 1.0 x 102
Dy III 7.2 X 1(T6 1.1 x 10~6 0.15 1.0 x 102
Ho III 1.4 X 1(T6 2.8 x 10 -7 0.20 1.3 X 102
Er III 4.2 X IO6 9.2 X 107 0.22 1.6 x 102
Tm III 6.1 x 107 1.3 x 107 0.21 1.6 x 102
Yb IV 3.6 x IO6 9.0 x 107 0.25 2.0 x 102
Lu IV 6.4 X 107 1.4 x IO7 0.22 2.0 x 102
Hf IV -- < 7 x 106 <1 .3 x 103
Ta IV -- <2 .5 X 10~6 < 2 .0 x 103
W IV 3 x 10 5 1 x 104 3.3 7.9 x 104
Re IV -- 4 x 106 3.2 x 106
Au III 2 x 106 4.9 x 106 2.5 1.6 x 106
Hg III 7 x IO5 1 x 106 0.14 7.9 X 103
T1 IV -- 1 x 10~5 6.3 X 103
Pb III 1 x 10^3 2x 106 0.002 5.0 X 101
Bi IV -- 2x 10"5
2.5 x 104
Th IV <1 x io-4 6x IO"8 ~ 0.0006 3.4
U I 4 x 105 3.1 x 103 2.7 1 x 106
I. Conservative element whose concentration is directly related to the salinity of seawater.
II. Nonconservative element whose concentration varies with depth or regionally within the oceans, or both,
generally because of involvement in biological activity.
III. Nonconservative element whose concentration varies irregularly and is not related to salinity, depth, or
geographic factors.
IV. Unclassified, but probably nonconservative.
bMean oceanic residence time in years.
S O U R C E : Taylor and McLennan (1985).
chem ical com position in
com parison w ith river w
ater and thereby contribute
to the geochem ical
differentiation of the
hydrosphere.
T he eventual fate of
all elem ents and com
pounds dissolved in the
ocean is to be rem oved
from it. M ost elem ents
are incorporated into the
sedim ent th at is accum ulating
the oceans, but a few elem ents
ocean into the atm osphere,
w hose com position is
sum m arized in Table 4.8.
These changes in resi
dence are p art of the m
igration of elem ents on
the surface of the E arth
and contribute to the opera
tion of geochem ical
cycles (see C hapter 22).
Every
elem ent is moved from
one reservoir to the next
by geological,
geochemical, or biological
process es operating on
the surface of the E arth .
This is the geochemical m
achine we spoke of earlier.
The dissolved
constituents in the oceans
can be divided into the
conservative (I in Table
4.7) and nonconservative
groups (II and III in Table
4.7). The conservative
elem ents occur in
constant p r o portions
throughout the oceans,
although their
concentrations m ay vary
because of dilution or
evaporative concentration.
This group includes the m
ajor elem ents and com
plex anions N a, K, Mg,
Ca, Cl, sulfate, and
borate. The
concentrations
 4.4 DIFFERENTIATION OF THE HYDROSPHERE 55
Table 4 .8 Chemical Composition of river w ater.Their average enrichm ent factors in the

Dry A ir oceans are

Concentration by Volume Alkali metals 416 (14.5-1714)

Constituent % ppm
Alkaline earths 151 (27.5-315)
N2 78.084 (except Be and Ba)
o2 20.946 Halogens 1790 (8-3350)
0.033
co2
Ar 0.934 In addition, B and S, both of which are m ajor con
Ne 18.18 servative elem ents, have large enrichm ent factors
He 5.24 indicating that they too tend to accum ulate in the
Kr 1.14 oceans. They are follow ed by P, V, Ni, Ge, Se, Mo,
Xe 0.087 Sb, W, Au, and U, w hose enrichm ent factors are all
H2 0.5 < 20 but > 1 . Only M o and U are conservative
CH4 am ong the elem ents in this group.
2
The enrichm ent factors of the rem aining non
N 2O 0.5
conservative elem ents are all < 1 , an indication that
Data from the CRC Handbook o f Chemistry and they are rapidly rem oved from seaw ater by geo
Physics (Weast et al., 1986). chemical processes.The concentrations of these ele
m ents in seaw ater are controlled prim arily by ion
of the nonconservative elem ents vary with depth exchange reactions with particulate m atter, by ionic
as well as regionally w ithin the oceans, and their substitution in calcium carbonate, calcium phos
concentration ratios w ith chloride are n o t con phate, and am orphous silica, which are in large part
stant. The nonconservative elem ents in seaw ater biogenic in origin, by direct precipitation caused by
include m ost of the trace elem ents along with dis evaporation of seawater, and by exchange reactions
solved nitrate, bicarbonate, silicic acid, phosphate, w ith volcanic rocks along spreading ridges (H art
and dissolved oxygen. T he concentrations of m any and Staudigel, 1982). The concentrations of most
nonconservative constituents vary w ith depth elem ents in seaw ater are n o t controlled by the solu
because they are involved in the biological activity bilities of their compounds.
of the oceans. A s a result, they have low concen In contrast to the conservative elem ents (and
trations in the surface layer of the oceans where a few nonconservative ones), the m ajority of the
m ost of th e biological activity is concentrated. nonconservative elem ents do n o t accum ulate in
W hen the organism s die, their bodies sink and the ocean but are rapidly rem oved from it by one
decom pose, thereby returning nutrients to the or several of the geochemical processes m en
water. Consequently, deep ocean w ater is enriched tioned earlier. The adsorption of ions on the
in nutrients com pared to surface water. charged surfaces of both inorganic and organic
The conservative constituents have large particles plays a very im portant role in this regard
enrichm ent factors, indicating that they are unreac and was described by B roecker and Peng (1982)
tive and therefore becom e concentrated in seawa as the great particulate sweep. O ther im portant
ter com pared to river water. The average enrich books on the geochem istry of seaw ater are by
m ent factor of the m ajor (conservative) con B ern er and B ern er (1987), H olland (1978),
stituents of seaw ater is 775. These elem ents are G oldberg (1972), and Riley and C hester (1971).
generally joined by their congeners, m any of which The wide variation of the enrichm ent factors
are trace elements. For example, the alkali metals, we have been discussing im plies th at the unreac
alkaline earths (except Be and Ba), and halogens tive conservative elem ents rem ain in the oceanic
all have higher concentrations in seaw ater than in reservoir m uch longer than the reactive noncon
56 CHEM ICAL DIFFERENTIATION OF THE EARTH

servative elem ents. This
phenom enon is reflected
by the residence time
defined as:
w here t is the residence
tim e in years, A x is the
total am ount of an elem
ent (x ) in solution in the
oceans, and d X /d t is the
average annua] input of
th a t elem ent into the oceans. TheActually,
an elem ent in the oceans is the
stant, which is probably
valid for the open ocean
but may not apply to the
continental shelves and
estuaries. The annual
input into the oceans can
be estim ated from the discharge
com position of the m ajor rivers
Such estim ates are uncertain, how ever, because
both discharge and elem ental concentrations of
stream s vary seasonally
and are not know n for all
rivers with equal certainty
(see also C hapter 20). In
addition, the ru n o u t of
groundw ater into the
oceans is disregarded
because of lack of inform
a tion. Furtherm ore, the
discharge of industrial and
m unicipal waste w ater
has significantly altered
the chem ical com
position of m any rivers.
For this reason Taylor and
M cL ennan (1985) estim
ated the average annual
input to the oceans from
the rate of deposition and
the chem ical com
position
of deep -sea sedim ent as originally
Fi (1982). Accordingly, the m ean
dence tim es (M O R T )
listed in Table 4.7 w ere
cal culated by this m ethod
based on sedim entation
rates provided by
Fisitsyn et al.
(1982). The
M O R Ts calculated from
the output to deep-sea
sedim ent differ from
those obtained from river-
input data, but the two sets
of results are well
correlated .
am ount the residence
assum ed to was
be con
oceans first used
the E arth . The idea was
1715 by E dm und Hailey,
but nothing was done
about it for nearly 200
years. Finally, in 1899
John Joly took up H
aileys suggestion and
calcu lated the residence
tim e of N a in the oceans.
His result was 90 m illion
years, which is less than
half the 200 million years
calculated by Taylor and
M cFennan (1985). N
evertheless, Jolys result
was an im portant m
ilestone in the history of
Evidently, the oceans
geology because it play a very im portant role
contradicted Ford K in the geochem ical
differentiation of the crust
elvins calculations based
of the E arth by storing
on th e cooling of the
som e elem ents and letting
Earth, which indicated that others pass rapidly to the
its age was less than 40 m sedim ent accu
il lion years. m ulating at the bottom of the ocea
m ents that are rem oved from the oc
The M O RTs of the form of sedim ent may
elem ents in Table 4.7 vary ultim ately reenter the
over nine orders of m rock cycle and becom e
agnitude from 0.69 years
the input for another pass
for Fe to 7.9 X 10s years
through the geochem ical
for Br. As expected, the
machine.
con servative elem ents
have long residence times,
w hereas m any 4.5 Summary
nonconservative elem ents The E arth differen tiated
have sig nificantly shorter very early in its history
residence times. The into a core, m antle, and
average residence tim es in crust w ith characteristic
the oceans of the
chem ical com positions. T
conservative elem ents are he core consists of m
etallic Fe, Ni, and re la te
d elem ents. T he m an tle
is com posed prim arily of
Alkali metals
silicates and oxides of M g
Alkaline earths (except Be and Ba)
and Fe. The chem ical com
Halogens position of the continental
crust has b e e n estim ated
in different ways with
The other two m ajor elem
generally co n co rd an t
ents (B and S) have results. T he nine
residence tim es of 16 X
106 and 500 X 106 years,
respectively.
The lithophile elem
ents have widely different
enrichm ent factors and
residence times, w hereas
a m ajority of the
chalcophile and
siderophile ele m ents are
depleted in seaw ater and
have short residence
times.
 REFERENCES 57
igneous rocks. They form by chemical w eather
m ost ab u n d an t chem ical elem ents in th e co n ti nental
crust of the E arth are O, Si, A l, Fe, Ca, Mg, Na, K, and Ti. ing of igneous, m etam orphic, and sedim entary
A ll o th er elem ents to g eth er m ake up only a very sm rocks and by the transport, deposition, and lithi-fication of
all fraction o f the co n ti nental crust. the resulting products. M any elem ents
becom e concentrated in deep-sea clay and may
The distribution of trace elem ents am ong dif ferent reenter the rock cycle w hen deep-sea sedim ent
kinds of igneous rocks depends on their ability to replace melts in subduction zones to form magma.
m ajor elem ents in the ionic crys tals that form during The oceans play an im p o rtan t role in the
crystallization of m agm a. A s a result, different kinds of chem ical differentiation of the crust because
igneous rocks are enriched in specific trace elem ents, som som e elem ents accum ulate in seaw ater, w h ere
e of which are im portant industrial metals. C ertain kinds as others pass through it rapidly. The geological and
of igneous rocks may therefore host ore deposits of these geochem ical processes associated w ith chem ical w
metals. eathering of m inerals, tran sp o rt of w eathering products,
and their ultim ate deposi
The chem ical com positions of sedim entary rocks tion in th e oceans are the subjects of P art III
are just as diversified as those of the of this book .
Problems
 1. Recalculate the chemical composition of pyrolite
(Table 4.2, column 1) in terms of the weight percent
concentrations of the elements and rank them in terms of
their abundance.
2. Recalculate the concentrations of Si, Al, Fe, Ca, Mg,
Na, K, Ti, Mn, and P of shale, sandstone, and carbonate
rocks (Table 4.5) in terms of the appro priate oxides.
3. Combine the concentrations calculated in Problem 2
in proportions of 4 parts shale, 0.75 parts sandstone, and
0.25 parts limestone.
4. Compare the results of Problem 3 to the esti mates
of the average chemical composition of the con tinental
crust (Table 4.3). Take note of discrepancies and
similarities and relate them to the geochemical properties
of the individual elements.
5. Assess the validity of the hypothesis that the sedi
mentary rocks (shale, sandstone, and limestone) collec
tively represent the weathering crust on the continents.
What additional sources or sinks are needed to ade quately
describe the redistribution of elements from the
continental crust into different kinds of marine sediment?
References
 G O L D S C H M ID T , V. M., 1954.
B E R N E R , E. K., and R. A. B E R N E R ,
Cycle. Prentice-Hall, Upper Saddle Geochemistry.
River, NJ, 397 Oxford
pp.
University
B R O E C K E R , W. S., and T.-H. Peng, 1982.
Eldigio Press, Palisades, NY, 690 pp.Press, 730 pp.
H A R T , S. R., and H. ST A U D IG E L ,
B U L L E N , K. E., 1963. An Introduction
1982. The control of
Seismology, 3rd ed. Cambridge
alkalies and uranium in
Cambridge, England, 381
pp. seawater by ocean crust
C L A R K E , F. W ., a n d H. S. W A S H alteration. Earth Planet.
IN G T O N , 1924. The Composition Sei. Lett., 58:202-212.
o f the Earths Crust. U.S. H E N D E R S O N , P., 1982.
Geol. Surv. Prof. Paper Inorganic
127,117 pp. Geochemistry.
D A L Y , R. A ., 1914. Igneous Pergamon
Rocks and Their Origin. Press, Oxford, England,
McGraw- 353 pp.
Hill, New York, 563 pp. H E R R M A N N , A. G., 1970.
G O L D B E R G , E. D. (Ed.), 1972. Yttrium and the lanthanides.
The Sea. Wiley, New York, 5 In K. H.
vols. Wedepohl (Ed.), Handbook
o f Geochemistry, vol. 2,
part 5, ch. 39, sec. E, 57-71.
Springer-Verlag, Berlin.
H O L L A N D , H. D., 1978. The
Chemistry o f the Atmosphere
and
Oceans. Wiley, New York,
351 pp.
58 CHEM ICAL DIFFERENTIATION OF THE EARTH

S EDERHOLM
 of the Earths crust.
Geochem. Int.,
H U T C H IS O N , R., 1974. The , J. J., 1925. The
13(6):89-121.
formation of the Earth.
Nature, average
250:556-568. composition of
K R A U S K O P F , K . B., 1979. the
Introduction to Geochemistry, earths crust in Finland.
2nd ed. Comm. geol.
McGraw-Hill, New York, Finlande Bull.,
617 pp. 12(70), 20 pp.
Li, Y.-H., 1982. A brief T A Y L O R , S. R., 1964. The
discussion on the mean abundance of chemical
oceanic residence time of elements in the continental
elements. Geochim. crust a new table.
Cosmochim. Acta, 46:2671- Geochim. Cosmochim.
2675. Acta, 28:1273-1285.
L IS IT S Y N , A. P., et al., 1982. T A Y L O R , S. R., and S. M. M C L
The relation between E N N A N , 1985. The Continental

element influx from rivers Crust: Its Composition and

and accumulation in ocean Evolution. Blackwell
sediments. Scientific Publ., Oxford,
Geochem. Int., 19:102-110. England, 312 pp.
M A S O N , B ., 1966. Principles o f T U R E K I A N , K. K., and K. H. W
Geochemistry, 3rd ed. Wiley, E D E P O H L , 1961. Distribution

New York, 329 pp. of the elements in some

M O O R E S , E . M ., 1982. O rig in a n major units of the Earths
d e m p la c e m e n t o f o p h io
crust. Geol. Soc. Amer.
lites .
Bull., 72:175-192.
Rev. Geophys. Space Phys.,
20:735. V I N O G R A D O V , A. P., 1962.
P A R K E R , R. L., 1967. Average contents of
Composition of the Earths chemical ele ments in the
crust. In principal types of igneous
M. Fleischer (Ed.), Data o rocks of the Earths crust.
f Geochemistry. U.S. Geol. Geochemistry, 1962(7):641-
Surv. Prof. Paper 440D, 1- 664.
19. W E A S T , R. C., M. J. A S T L E ,
P O L D E R V A A R T , A., 1955. and W. H. B E Y E R (Eds.), 1986.
Chemistry of the earths crust. CRC Handbook o f
In
Chemistry and Physics.
A. Poldervaart (Ed.), Crust
o f the EarthA CRC Press, Boca Raton,
Symposium. FL.
Geol. Soc. Amer. Spec. W E D E P O H L , K. H., 1969.
Paper 62,119-144. Handbook o f Geochemistry,
vol. 1.
R I L E Y , J. P., and R. C H E S T E R ,
1971. Introduction to Marine Springer-Verlag, Berlin,
Chemistry. Academic 442 pp.
Press, New York, 465 pp. W H I T F I E L D , M., and D. R. T U R
R IN G W O O D , A. E ., 1966. The N E R , 1979. Water-rock
chemical composition and partition coefficients and
origin of the Earth. In P. the composition of
M. Hurley (E d .), Advances seawater and river water.
in Earth Nature, 278:132-137.
Sciences, 287-356. MIT
Press, Cambridge, MA.
R O N O V , A. B., and A. A. Y A R O S
H E V S K Y , 1976. A new model

for the chemical structure

 II
PRINCIPLES OF INORGANIC
CHEMISTRY

A n im portant task of geochem ists is to understand the chem ical com positions of m inerals and
rocks based on the physical and chem ical properties of the elem ents and their atoms. The regular
pattern of variation of the electronic structures of atom s d eter m ines the chem ical properties of the
elem ents and hence their positions in the period ic table. Consequently, our understanding of the
distribution of chem ical elem ents am ong the m inerals and rocks of the E arth rests securely on the
principles of physics and chem istry em bodied in the periodic table.
 5
The Electronic Structure
of Atoms
We now end our tour of th e stars and the solar that the plum -pudding m odel of the atom could not be
system and turn to som e basic principles of correct because it did not provide for the quantization of
chem istry th at will help us to understand the electrom agnetic radiation that had been discovered by M
chemical processes taking place on the earthlike planets. ax Planck (1858-1947) in G erm any (Planck, 1900). H
The intellectual excitem ent generated by the exploration ow ever, Thomson
of the solar system cannot hide the fact that only on the E was in no m ood to argue with th e 26-year-old
arth can we live w ith out elaborate protection against the D ane who had little know ledge of the English
harsh envi ronm ents of the other planets. This lesson of language at the time. Sir John, w ho was the direc
planetary exploration should increase our in ter est in our tor of the Cavendish Laboratory, had won the
own planet and in the natural process es th at take place N obel Prize in 1906, and had been knighted in
on its surface and in its interior. We need to understand 1908, was not interested in B ohrs ideas about
these processes well in order to assure adequate living how to im prove his electron theory. Besides, he was then
conditions for working on his positive-ray apparatus, which would
soon lead to the discovery of iso topes and the developm
future generations. For these reasons w e now
ent of mass spectrom e ters (H olton, 1986).
m ove to a description of atom s whose orbiting
electrons resem ble the orbiting planets and their satellites B ohr therefore decided to leave Cambridge, and
of the solar system . went to M anchester where E rnest R utherford (1871-
1937) was the Langw orthy Professor of Physics. Perhaps
as a test of T hom sons m odel of the atom , R utherford
5.1 The Atom of Thomson and had fired a beam of alpha particles at a piece of gold foil
Rutherford and had observed to his great surprise th at the beam did
not simply pass through the foil as expected. Instead, a
Throughout the 19th century physicists accepted Sir Isaac sig nificant num ber of alpha particles w ere scattered and
N ew tons opinion th at atom s are solid, som e even seem ed to bounce straight back as though
massy, hard, im penetrable, m oveable particles they had struck a heavy object. H e said later It was
(Jastrow, 1967, p. 9). H ow ever, in 1897 J. J. quite the m ost incredible event that has ever happened to
Thom son (1856-1940) discovered the electron m e in m y life. It was alm ost as incredible as if you fired
and used it to develop a new m odel of the atom in which a 15-inch shell at a piece of tissue paper and it cam e back
negatively charged electrons w ere thought to be em and hit you. (Jastrow, 1967, p. 9). A fter thinking about
bedded in a positively charged m atrix like raisins in a it, R utherford decided that the m assive objects had to be
plum pudding. In 1911 Niels B ohr (1885-1962) cam e to very small because they w ere struck only rarely by alpha
the C avendish Laboratory at Cam bridge as a particles and concluded that most
postdoctoral fellow to work with Thom son on his
electron theory. H e argued
60
5.2 BOHR'S THEORY OF THE HYDROGEN ATOM 61

of the m ass of the atom
was concentrated in this
small body, which he nam
ed the nucleus.
R u th erfo rd s m
odel of th e atom was
inherent ly unstable
because in classical
physics an electron
orbiting a positively
charged nucleus loses
energy by em itting an
electrom agnetic wave. As
a result, the electron m ust
speed up and the radius of
its orbit m ust decrease
until the electron crashes
into the nucleus.
Nevertheless, B ohr liked
R u th erfo rd s m odel
because he had a hunch th
at he could im prove it by
extending the concept of
quantiza tion of electrom
agnetic energy to the m
echanical energy of an
electron orbiting the
nucleus.
We all know that B
ohr succeeded brilliantly,
received the N obel Prize
in physics in 1922, and
becam e one of the leaders
in science in the 2 0 th cen
tury. However, he had a
hard tim e convincing even
R utherford that his model
of the atom was right. W
hen Bohr, who had briefly
returned to D enm ark, sent
a m anuscript copy of his
paper, R utherford agreed
to subm it it for
publication but added, I
suppose you have no
objection to my using my
judgm ent to cut out any m
atter I may consider un
necessary in your paper?
(H olton, 1986, p. 240).
y
Bohr did object and came velocity v. The condition
back to M anchester from for stability of such an
Copenhagen to discuss his o atom is that the force of
paper with R utherford electrostatic attraction m
point by point. In the end, f
ust be equal to the
R utherford agreed to t centrifugal force:
submit the entire
manuscript, but grumbled
h mv
(5.1)
that B ohr had been so e
obstinate. The energy of such an
H atom is the sum of the
kinetic energy and th e
W hen this paper and y potential energy.
a second one w ere pub
lished (B ohr, 1913a, b), d
1 2
they were not well r m v
received. The prom inent
G erm an physicist O tto S o The potential energy term
tem m ay have spoken for g has a negative sign
m any of his colleagues w because the electron is
hen he reportedly said, If
e attracted tow ard the p ro
this nonsense is cor rect, I n ton. From equation 5.1 we
will give up being a obtain:
physicist. (H olton, 1986,
p. 240). It was the end of A m
rv
an era. Physics was about t
to be shaken to its roots by Substituting into equation
quantum mechanics and o 5.2 yields:
the uncertainty principle. m \( e
(5.4)
The instability of an atom E=t-r)~
com posed of one elec tron 7
5.2 B orbiting the nucleus can be
dem onstrated by m eans =
o of N ew tonian physics.
h
21
We assume that the
electron orbits the proton
r r
at a distance r and a
' We see from equation 5.4
that the energy of the atom
s is negative and that it is
inversely related to the
radius of the orbit. In the
T limit, the energy approach
h es minus infinity as the
radius goes to zero.
e
The instability arises
o because the atom is an
r energy well. Energy is
liberated as the radius of
the electron orbit
decreases. Since there are
no restrictions on the m
agnitude of the electronic
radius in this m odel, atom
s will tend to achieve a
condition of m inim um
energy and m axim um sta
bility by reducing the
radius to zero.

B ohr avoided this

problem by specifying
that the energy of the atom
is not continuously
variable but is, in fact,
quantized so th at the
electron m ust be
restricted to orbits having
specific radii. In order to
accom plish this, he
postulated that the
angular
m om entum of the
electron is restricted to m
ulti
ples of h /h r , w here h is
Plancks constant, which
is equal to 6.62517 X 10 - 2
7 erg /sec (see also the

inside back
cover).Therefore,
according to Bohr:
m vr =

w here n = 1, 2, 3, . . . ,
and is know n as the first
quantum num ber. H ow
ever, the atom m ust still
62 THE ELECTRONIC STRUCTURE OF ATOMS
 e an atom as a result of an
2
satisfy the condition square of the electronic
first quantum numtransition is:
ber. Evidently,
T he sm allest radius perm
expressed by equation 5 .1 . the quantum num ber plays a decisive role a it hc
itted by B o h rs theory is
Accordingly, the angular because it controls both the A=t
obtained from equation
m om entum of the elec radius of the elec tronic
5.10 by setting n = 1, in In other words, the energy
tron m ust also be: orbit and the energy of the
which case: of the radiation is
atom .
inversely proportional to
m vr = r = aQ = 0.529 X 10 In B ohrs m odel of
its wavelength.
the atom the electron can
w here a0 is the so-called W hen B ohr
and a stable atom m ust B o h r radius. occupy only certain orbits
developed this m odel of
satisfy both equations 5 . 5 whose radii depend on the
The energy of the the atom betw een 1912
and 5.6; thus: first quantum num ber. The
atom can now be and 1913, it was already
energy of the atom varies
e2 _ calculat ed by know n that hydrogen
v in increm ents depending
substituting equation atom s absorb and em it
on which orbit the electron
5.10 for the radius occupies. W hen the light at discrete
The velocity of the
wavelengths. These
electron is derivable from into equation 5.4 for the energy electron jum: ps from a
of th e atom
E = - f e 2 (Air2m e2\ higher orbit to a low er wavelengths had been ex
equation 5.5: pressed by the equation:
21,2 one, the energy of the atom
v= decreases. The energy
nh
m r see th at in B o h rs m odel liberated
2 v We the energyin this way is
Substituting into equation atom no longer depends on radiated away as an 4
(5.7) yields: the radius of the elec tronic electrom agnetic wave. B
w here al and a2 are
ohr used an equation
2m nre2 _ orbit b u t varies as the whole num bers, ax > a2,
derived by A lb e rt
nh reciprocal of the and R is the R ydberg
Einstein to represent this
constant, whose value was
which leads to: energy change.
know n to be 109,678.18
,21,2
nz If E 1 is the energy of cm -1. (N ote th at we use
h the atom w hen the a in equation 5.17 instead
r=
(5.10 elec of n in order to avoid con
)
4
7
tron is in a particular orbit
fusionand with the first
r w hen the electron is in a
quantum num ber.) By
2 change in the energy of the
m apply
atom as the electron
e moves from orbit 1 to orbit
2 - E2=
E quation 5.10 indicates th
w here v is the frequency
at the radius of the electron
of the electrom agnetic
now depends on the square
radiation and h is Plancks
of the first quantum num
constant. The velocity of
ber m ultiplied by a
electrom agnetic radiation
constant a w hose m
c is related to its w
agnitude is:
avelength A and the
frequency v by:
h2
(5.11) c = Xv
4
7
r Therefore, the energy of
2
m the radiation em itted by

ing quantum m echanics to R u th erfo rd s m odel of the
hydrogen atom B ohr was able to calculate the w
avelengths of light th at hydrogen atom s can em it as a
result of transitions am ong th e orbits the electron can
occupy. By com bining equations 5.13, 5.14, and 5.16 we
have:
27r2me4/ 1 1 he
(5.18)
T generat ed when the resulting vacancies are filled by other
where n x and n 2 are th e quantum num bers, and n2 > n x
as required by the R ydberg form ula. It follows that:
1 difference in the energies of the orbit in which a vacancy
1 _ 27r2me4 / 1
A h3c (5.19)
By com paring equations 5.19 (B ohrs m odel) and 5.17
(R ydberg form ula) we see that:
R= 2iT2m e4
(5.20)
h3c
B ohr substituted appropriate values into equation 5.20
and calculated a value of 109,677.76 cm - 1 for the R
ydberg constant. The result agreed very well with the
observed value and thereby confirm ed B ohrs m odel for
the hydrogen atom . Niels B ohr was justifiably pleased
with this result and subse quently devoted him self w
holeheartedly to the study of quantum mechanics.
The R oyal D anish A cadem y of Science, which
receives a large portion of its funding from the C arlsberg
Brewery, gave B ohr a grant of m oney to build an
Institute for Theoretical Physics in C openhagen. B ohrs
institute becam e the center for research in quantum m
echanics and nuclear physics. It was fam ous n o t only for
its intellectual excitem ent but also for its relaxed atm
osphere created by the Professor (Gamow, 1966).
5.3 Emission of X-rays
B ohrs m odel of the atom provided a direct expla nation
for th e emission of x-rays, which had been discovered in
1895 by W ilhelm K onrad R ntgen
(1845-1923). A lthough B ohrs m odel has been
replaced by the wave mechanics of W. K.
H eisenberg (1901-1976) and E . Schrdinger
5.3 EMISSION OF X-RAYS 63
(1887-1961), som e aspects of it are still in use in the
physics of x-rays.
In a com m ercial x-ray tube a stream of elec trons is
accelerated by a voltage difference in a vacuum before
striking a target com posed of a m etal such as chrom ium ,
copper, m olybdenum , or tungsten. The energetic
electrons interact with the electrons of the target atom s
and m ay knock them out of their orbits. X -rays are then
electrons.
The wavelengths of these x-rays depend on the
has been created and the orbit from which the replacing
electron originated, as re quired by equation 5.16. These
x-rays therefore have discrete wavelengths and give rise to
a wave length spectrum th at is characteristic of the target
elem ent. The characteristic x-ray spectra w ere dis
covered in 1911 by Charles G. B arkla (1877-1944), who
received the N obel Prize for physics in 1917. The
relationship betw een the characteristic x-ray spectrum and
the atom ic num ber of the target ele m ent was discovered
in 1913 by H . G. J. M oseley (1877-1915).
In addition to the characteristic wavelength spectrum ,
a continuous x-ray spectrum is p ro duced by electrons th
at pass through the target atom s w ithout actually colliding
with any of their electrons. N evertheless, som e energy is
trans ferred in this type of interaction, and this energy is
em itted as x-rays having continuously varying
wavelengths.
The characteristic x-ray spectrum of the tar get atom s
is divided into several series of w ave lengths referred to by
the letters K, L, M, etc. The K x-rays are em itted w hen
a vacancy in the low est energy level (K; n = 1) is filled.
The electron filling the vacancy in the K -level m ay
originate from the next higher level (L) or from others
associated with electronic orbits having still larg er radii.
The L -series of characteristic x-rays is generated by
transitions to the L-level (n = 2), and so on. The general
schem e of identifying the characteristic x-rays is m ore
com plicated than th at shown in Figure 5.1 because
electrons in each of th e K-, L-, and M -energy levels have
slightly
64 THE ELECTRONIC STRUCTURE OF ATOMS

because of the splitting of the wavelengths caused by

small energy differences of the elec trons. In general, K
-alpha x-rays are produced in greatest abundance because
vacancies in the K -energy level are m ost often filled by
electrons from the overlying L-level. T herefore, K -alpha
E
x-rays are usually selected for analytical purposes both in
o x-ray-diffraction and in x-ray-fluores-cence studies. The K
< -beta and K -gam m a x-rays are m ore energetic, b u t
CD
their intensities are less than those of the K -alpha x-rays
-C
because the respective transitions occur less frequently.
o
<v a 7

c
LU
L-series
a 7 ' e indispensable tools in geochemical research. The form er
K-series Electron orbits
Figure 5.1 Schematic energy level diagram of atoms
based on Bohrs model. X-rays of the K-series are
generated by electronic transitions to the K-energy
level (n = 1), whereas the L-series involves transitions
to the L-level (n = 2). Greek letters are used to
indicate the orbit from which the electrons originate.
In reality the characteristic wavelength spectrum is
more complicated than shown here because of small
differences in the energies of the electrons that
populate the M-, N-, and O-energy levels.
different energies. These differences are explain able in
term s of wave m echanics and are not p re dicted in B o h
rs m odel of the atom .
We see in Figure 5.1 th at the K -series of char
acteristic x-rays is subdivided according to a sim ple
scheme.
X -ray diffraction and x-ray fluorescence have becom
is based on the diffraction of
x-rays by crystals (Klug and A lexander, 1954),
w hereas the latter arises from the emission of
characteristic x-ray by atoms. X -ray fluorescence is now
used to determ ine the chemical com posi tions of rocks
including not only the m ajor ele m ents but also many
trace elem ents (N orrish and H utton, 1969).
5.4 Schrdinger's Model of the
Atom
B ohrs m odel w orked for the hydrogen atom and for
ions having only one electron such as H e+, Li2+, and
Be3+. How ever, it did not w ork well for atoms having
two or m ore electrons. A rnold Som m erfeld (1868-1951)
m ade some im provem ents by intro ducing elliptical
orbits like those of the planets in the solar system . In
addition, he allowed the planes of the orbits to include all
the free space around
Electronic transition Characteristic x-ray the nucleus, thereby m aking the atom look m ore like a
L to K K-alpha sphere than a flat disk. These refinem ents of B ohrs m
M to K K-beta odel required the introduction of a sec ond quantum num
N to K ber but still did not m ake it applicable to m ultielectron
K-gamma
atoms.
O to K K-delta
The problem with B ohrs m odel is that is treats
electrons as particles whose positions in space can
The L -series is subdivided similarly, but the details in both be determined. The validity of this idea was ques tioned
series get m ore com plicated by Louis-Victor 7th D uke de Broglie
5.4 SCHRDINGER'S M ODEL OF THE ATOM 65

(L892-1987) in 1923 who
dem onstrated that parti
cles such as electrons also
have the properties of
waves. The wavelength A
of particles of mass m and
velocity v is:
gen erally stated in the
form:
d2ljj d2ip d2lp 87T2m .
^? +^7- +jr2+~ ir{E~v),p=0
where ip is a wave
function, m is the mass of
the electron, E is the total
bers are needed to
describe the m otions of
an electron in the space
around the nucleus of an
atom . These quantum
num bers are:
n = 1, 2, 3 , . . . , 0
0
o rb its of the B ohr-Som
m erfeld m odel. The x-
orbital has spherical sym m
etry about the nucle us. T
he p -type orbital is sym m
etrical about the x-, y-, and
z-axes, and the d-orbitals
are sym m etrical about the
xy, xz, and y z planes.

(principal These orbitals are pictured

where h is Plancks
constant. For exam ple, the
w avelength of an electron
having a kinetic energy of 1
0 electron volts is 0 . 1 2 X 10 8
cm, correspond ing to
fairly energetic x-rays (M
oore, 1955, p. 271).
If electrons have
wave properties, then it m
akes no sense to think of
one as being in a p ar
ticular place at a particular
time. W erner H eisenberg
expressed this condition in
the form of the uncertainty
principle, which implies
that the
m otion of an electron around
atom cannot be described quantum
orbits regardless of w hether they
elliptical.
These strange ideas w
ere incorporated into a
new m odel of the atom
proposed in 1926 by W
erner H eisenberg and Erw
in Schrdinger.
eisenbergs m ethod was
abstract and relied on m
atrix algebra, w hereas
Schrdinger (1926) used
de Broglie waves to
represent the distribution
of electrons in three-dim
ensional space.
Schrdingers equation is
energy of the atom , and V
A - A is the potential energy quantum in Figure 5.2. The /-type
num bers) orbitals cannot be pictured
(Fyfe, 1964).
The wave equation / = n 1, n 2, in three dim ensions. In the
case of the x-type wave
can be understood by com n 3, . . . , 0
function
paring it to a violin string (azim uthal quantum num there
ber) is only one
vibrating betw een two m =0 ,1 , 2 , . .solution;
. , (/ - l),
fixed points. Such a string (m agnetic quantum num ber)
can develop nodes whose s =i (spin quantum num ber)
but th ere are three p -orbitals corr
num ber m ay be 0 ,1 ,2 ,
the principal axes, five d-orbital
etc. The num ber of nodes The fourth quantum num
/-orbitals.
in the one-dim ensional ber, 5, indicates the direc
T he wave functions
case is determ ined by the tion of spin of the electron,
of the different types of
quantum num ber n. If n = which controls the polarity
orbitals convey the
1, the num ber of nodes is of the m agnetic m om ent
probability th at the
zero; when n = 2 there is
exerted by the electron . N
electron of the hydrogen
one node, and so on. In ote that the electron is
again regarded as a atom will be found in a
the three-dim ensional
charged particle in space certain volum e of space. C
electron wave, nodes may
in this context. onsequently, we can
form along the three
principal directions; Solutions to the wave illustrate the distribution of
therefore at least three equation are obtained for the electron around the
wave functions having nucleus by plotting th e
in num
certain types of sym m function Airr2^ 2 versus
etry in space. For exam
ple, the wave function can
have spherical sym m etry
around the nucleus, or it m
ay be sym m etrical with
respect to the x, y, z
directions or the xy, xz, y z
H
planes. In general, four
different types of wave
functions are possible,
which are identified by the
letters s, p, d, and / These
wave functions describe
the m otions of the electron
to which they apply and
are therefore called
orbitals in order to
distinguish them from the
Figure 5 .2 Electron clouds representing the s-, p-, and d-orbitals of the hydrogen atom according to the wave
mechanical model of Schrdinger.

the distance r from th e nucleus. Figure 5.3 is such a cross
section of an ^-orbital for the hydrogen atom . We see th at
the profile has a m axim um at a distance r = a0, the radius
of the first electron orbit in B o h rs model. However, we
also see that the electron can be found closer to or farther
away from the nucleus som e of the time. In other words,
the electron actually occupies all of the space around the
nucleus b u t can be found m ost often at a distance of a0
from th e nucleus. The same is true for all orbitals, and the
orbital balloons shown in Figure 5.2 therefore have a
fuzzy skin. The elec trons occupying these orbitals do n
o t spend all of their tim e on the skins of the balloons but
m ove in and out in a random pattern predictable only in
term s of probability. This aspect of quantum m echanics
was very disturbing to A lbert E instein and even to Erw in
Schrdinger. E instein objected to the elem ent of chance
im plied by the proba-
5.5 THE AUFBAU PRINCIPLE 67
bilistic description of the electronic structure of atom s by
saying D er liebe G o tt wrfelt nicht! (the good Lord
does not throw dice).
5.5 The Aufbau Principle
We are now ready to begin the construction of atom s of
increasing atom ic num ber by adding electrons to the
possible orbitals of the hydrogen atom . This procedure is
know n as the A u fb a u principle by m eans of which the
orbitals of the hydrogen atom s are used to m ake m
ultielectron atoms. In doing so we assum e th at the num
ber of protons in th e nucleus increases in step with the
electrons.
The energy of an atom having a certain num ber of
electrons depends on the orbitals the elec trons occupy.
Therefore, the electrons occupy only those orbitals th at m
inim ize the resulting energy of the atom because the
lowest energy state of an atom is also its m ost stable or
station ary state. This requirem ent controls the
sequence in which orbitals are filled. The sequence can be
derived by calculating the energy of an atom for different
electron configurations. H ow ever, for m ultielectron atom
s these calculations becom e very difficult and m ust be
supplem ented by spec troscopic studies.

The four quantum num bers play an im p o r

tan t role in this process because, according to
P au lis exclusion principle, no two electrons in
one atom can have the sam e set of quantum
num bers. This principle th erefo re lim its th e
num ber of electrons th at can en te r a particular orbital. T
Distance from the Nucleus (r)
he relationship b etw een th e quantum
Figure 5.3 Variation of the radial distribution function num bers and the differen t orbitals is indicated
(4Trr2ip2) for the y-orbital with increasing distance in Table 5.1.
from the nucleus (r).The radial distribu tion function is We see there th at the quantum num bers
a measure of the probability of finding the electron at allow only the s-orbital for n = 1 ; s-, and p-orbitals
a certain distance from the nucleus. In the s-orbital this for n = 2; s-, p-, d -orbitals for n = 3; and s-, p-, d-,
probability has a maxi mum at a distance r = a0,
where a0 is the radius and /-orbitals w hen n = 4. The table also indi
of the smallest orbit (n = 1) permitted by Bohrs cates th at th ere are three different p -orbitals,
model. We see that the electron in an s-orbital can five d -orbitals (the seven /-orbitals are not list ed). M
also be found inside as well as outside that radius. oreover, each individual orbital can accom m odate tw o
The electron in effect occupies all of the space around electrons provided th eir spin num bers are different. A s a
the nucleus but can be found most often at the result, only two elec trons can be accom m odated in
distance from the nucleus predicted by Bohr. orbitals having
68 THE ELECTRONIC STRUCTURE OF ATOMS
Table 5.1 Q u a n tu m N u m b e rs a n d P o ssib le O Subsequently, a second electron may enter an
rb ita ls o f th e H y d ro g e n A to m orbital that already contains an electron provided
its spin is opposite (rule 2 ).
 Quantum numbers 4. A tom s achieve a state of decreased energy and
n I m s Type o f orbital enhanced stability w hen the available orbitals are
1 0 0 +> 2 either com pletely filled, half filled, or empty.
Is
2
2 0 0 +I 2 2s
'2i
2 1 0 +- 2 2p We are now ready to construct the atom s of the elem
T 2

2 1 1 +i _> ents in order of increasing atomic num ber by specifying

2> 2 2p
2 1 -1 + 1 _1 their electronic structures. We do this by stating the value
2p
T 2> 2 of the first quantum num ber, the type of orbital, and the
3 0 0 + i _ Jl 3s num ber of electrons in each of those orbitals. For
2 2
3 1 0 f1 2 2
3p example, ls22s22p3 is the electronic structure of nitrogen
3 1 1 + 1 _I 3p (atom ic num ber 7). It has two electrons in the Is orbital,
2 2
3 1 -1 +' -1 3p two in the 2s orbital, and three in the 2p orbitals.
T 2 2
3 2 0 +1 - Evidently, each of the three p-orbitals contains one
3d
3 2
2 2
electron.
1 -f1 2 3d
3 2
' 2 We can anticipate from rule 4 th at nitrogen
-1 +1 -1 3d
2 2
atom s may attract three additional electrons to
3 2 2 +1 _! 3d
T 2 2 them selves in order to fill the p -orbitals. In addi tion,
3 2 -2 +! _I 3d
'2i 2 they may also em pty the p -orbitals by giving up three
etc. electrons, or lose five electrons by em p tying both the p -
and s-orbitals having n = 2 .
n = 1 , eight electrons for n Consequently, nitrogen is
= 2 , and 18 electrons for n likely to have different
= 3. oxidation states or
valences: - 3 , 0, +3, and
The sequence in +5. Evidently, we are
which electronic already able to m ake
orbitals are predictions about the
actually filled is: chemical properties of
I s, 2s, 2p, 3s, 3p, 4s, 3d, 4p, 5s, based entirely on
nitrogen
its electronic structure and
4 d, 5p, 6s,
the princi ples of wave
The A ufbau principle, by mechanics.
m eans of which we can Table 5.2 lists the
predict the electronic electronic form ulas of
configuration of the ele m the elem ents in order of
ents in the periodic table, increasing atom ic num ber
is governed by a set of from 1 to 96. The electrons
rules we state here in plain are inserted into the
language w ithout regard orbitals in the proper
to their historical origins. sequence. The highest
posi tive oxidation state of
each elem ent can be antic
1. The first electron
ipated by m eans of rule 4.
enters the orbital
We have no difficulty in
having th e lowest
assigning hydrogen an
energy. A dditional
oxidation state or valence
electrons en ter
of + 1 , although - 1 is also
orbitals that m
possible and occurs in
inim ize the
hydrides such as LiH . In
energy of the atom natural systems hydro gen
(see above). has a valence of + 1 , or 0 w
hen it occurs as the diatom
2. E ach orbital can ic m olecule H 2.
accom m odate
two elec trons Helium (Is2) has
provided their filled the only orbital
spins are opposite. avail able to it, and
3. W hen electrons therefore it neither accepts
enter a particular nor gives up additional
electrons. Its valence is 0.
set of orbitals,
Lithium (ls 22sl)
they distribute
begins a series of elem
themselves such
that each orbital at ents that fill orbitals
first aquires only having n = 2. The lone
one electron. electron in

th e 25 orbital of lithium exists outside the closed lx 2
orbital and is therefore readily lost. H ence lithium has a
valence of + 1 .
Beryllium (l52252) has two electrons in the 2s
orbital, but it gives them up easily, thereby acquiring a
valence of + 2 .
B oron (1 s22s22 p x) starts a sequence of six elem
ents in which the 2p orbitals are filled. The positive
valences of these elem ents follow logi cally: boron + 3,
carbon + 4, nitrogen +5 . The next elem ent, oxygen ( l s
22s22p4), should have a
valence of + 6 ; but its valence in n atu re actually
is - 2 because, after the p -orbitals are half-full
(nitrogen), atom s of oxygen attract electrons to them
selves in ord er to fill their orbitals. E ven nitrogen can
acquire three electrons and have a
valence of - 3 as, for exam ple, in am m onia
(N H 3). Fluorine ( l s 22s22 p 5) has th e strongest
tendency of all the elem ents to attract electrons to itself
and, consequently, has a valence of - 1 . N eon ( l s
22s22p':) com pletes the sequence by fill ing th e 2 p
orbital and thereby ends the series of elem ents that started
w ith lithium . In Table 5.2
square brackets indicate th at the noble gases
have filled orbitals. N eon is an inert or n oble
chem ical elem ent that does n o t form bonds with oth er
elem ents.
In the next sequence the 5 - and p-orbitals of the third
shell are filled. The valences of these ele m ents (sodium,
magnesium , aluminum, silicon, phosphorus, sulfur,
chlorine, and argon) follow predictably from their
electronic structures. Sodium has a 3s1 electron outside of
a neon core and therefore has a valence of + 1 , just like
lithi um and hydrogen. M agnesium follows with +2, like
beryllium before it, and so on. The electronic
configuration of sulfur ([Ne]3s2 3p4) resem bles th at of
oxygen, which has a valence of - 2 because it strongly
attracts electrons to itself. Sulfur atoms likewise attract
two electrons to form S2_, but they can also release their
p-electrons and then
assum e a valence of + 6 , as in the sulfate ion
(SO 4 - ). Chlorine strongly prefers the valence of
- 1 but takes a valence of + 7 in the perchlorate
(C IO 4 ) ion. A rgon ([N e]3 5 2 3 p 6) closes this se
quence with a valence of 0 .
5.5 THE AUFBAU PRINCIPLE 69
The third energy shell includes af-type orbitals;
however, potassium ([A r]4 5 J) puts the next elec tron into
the 4x orbital rather than one of the 3d orbitals. Calcium
([Ar]4x2) does likewise, so these elem ents have valences
of + 1 and + 2 , respect ively. The d -orbitals are filled next,
as we progress from scandium to zinc. In this se quence,
chrom ium and copper are starred in Table 5.2 to draw
attention to a significant anom aly in their electronic
structures. Chrom ium should be [A r]3d4 4x2 but is
actually \A r \3 d 2As{, and cop per is [A rc/'H ?1. B oth
exemplify the benefit of half-filling or filling a set of
electronic orbitals. Irregularity also occurs in the cases of
iron, cobalt, and nickel because their valences do not
follow the expected pattern . A ll three elem ents have
valences of + 2 and +3, even though we m ight have
predicted + 4 for cobalt and + 5 for nickel.
Next, we have another sequence of six ele ments,
starting with gallium and ending with krypton, in which
the Ap orbitals are filled. These
elem ents are sim ilar to the alum inum -argon
series in which 3p orbitals are filled.
Table 5.2 continues w ith another 18 ele
m ents: rubidium and strontium (5x), yttrium to
cadm ium (Ad), and indium to xenon (5p ). A fter
that com e elem ents with electrons in 65, Af, 5d,
and 6 p orbitals ending with radon (atom ic num ber 8 6 ).
The sequence continues with Is, 6 d, and 5 / through the
transuranium series. In general, the chemical properties of
elem ents having partially filled d- and /-orbitals do not
vary as m uch as those in which s- and p -orbitals are
being filled.
5.6 Summary
O ur concept of the internal structure of the atom has
undergone m any changes since J. J. Thomson form ulated
the plum -pudding model. Some of the im portant m
ilestones in this evolution are the
scattering of alpha particles, discovered by
E. R utherford; the quantization of the energy of
the hydrogen atom by N. Bohr; and the dem onstra tion
by L. V. de Broglie that small particles have wave
properties. U ltim ately W. K. H eisenberg and E.
Schrdinger described the energy states of
 Electronic Structure and Valences of the Elements
za Element Electronic structure Valences
1
Hydrogen Is1 + 1 ,- 1
2 Helium Is2 0
3 Lithium l^Zs1 +1
4 Beryllium ls 22s2 +2
5 Boron ls 22s22pl +3
6 Carbon ls22s22p2 + 4 ,+ 2 , - 4
7 Nitrogen ls22s22p3 +5, +4, +3, +2, 3
8 Oxygen ls 22s22p4 +6,-2
9 Fluorine ls z2s22p5 +7,-1
10 Neon 1s22s22p6 0
11 Sodium [Ne^s1 +1
12 Magnesium [Ne]3s2 +2
13 Aluminum [Nejss^p1 +3
14 Silicon [Ne]3s23p2 +4,-1
15 Phosphorus [Ne]3s23p3 +5, +4, +3, 3
16 Sulfur [Ne]3s23p4 +6, +4, +2, 2
17 Chlorine [Ne]3s23p5 +7, +5, +3, +1, 1
18 Argon [Ne]3s23p6 0
19 Potassium [Ar]4s1 +1
20 Calcium [Ar]4s2 +2
21 Scandium [Ar]3d14s2 +3
22 Titanium [Ar]3rf24s2 + 4 ,+ 3
23 Vanadium [Ar]3dMs2 +5, +4, +3, +2
24= Chromium [A+ dW +6, +3, +2
25 Manganese [Ar]3d54s2 +7, +6, +4, +3, +2
26 Iron [Ar]3d64s2 + 3 ,+ 2
27 Cobalt [Ar]3d74.y2 + 3 ,+ 2
28 Nickel [Ar]3rf8452 +3, +2
29= Copper [Ar]3d4sl + 2 ,+ 1
30 Zinc [Ar]3d104s2 +2
31 Gallium [Ar]3d104sa4p1 +3, +1
32 Germanium [Ar]3<i104s24p2 + 4 ,+ 2
33 Arsenic [Ar]3<i104s24p3 +5, +3, 3
34 Selenium [Ar]3d104s24p4 +6, +4, - 2
35 Bromine [Ar]3rf104s24p5 +7, +5, +1, 1
36 Krypton [Ar]3d104524p6 0
37 Rubidium [R+s1 +1
38 Strontium [Kr]5s2 +2
39 Yttrium [Kr]4rf15s2 +3
40 Zirconium [Kr]4d25i2 +4
41 Niobium [Kr]4d35s2 + 5 ,+ 3
42= Molybdenum [Kr]4ds5s1 +6, +5, +4, +3, +2
43 Technetium [Kr]4ds5s2 +7
44 Ruthenium [Kr]4d65s2 +8, +6, +4, +3, +2
45 Rhodium [Kr]4d75s2 + 4 ,+ 3 ,+ 2
Palladium [Kr]4d85s2 +4, +2
Silver [Kr]4d10551 +1
Cadmium [Kr]4d105s2 +2
Indium [Kr]4d105s25p1 +3, +1
Table 5 .2 (co ntinued)
za Element Electronic structure Valences

50 Tin [Kr]4d105s25p2 +4, +2

51 Antimony [Kr]4du)5s25p3 +5,+3,-3
52 Tellurium [Kr]4d105s25p4 + 6 ,+ 4 , - 2
53 Iodine [Kr]4d105s25p5 +7, +5, +1, 1
54 Xenon [Kr]4rf105s25 / 0
55 Cesium [Xe^1 +1
56 Barium [Xe]6s2 +2
57 Lanthanum [Xe]5dl6s2 +3
58* Cerium [Xe]4f25d6s2 +3, +4
59 Praseodymium [X e]4/35d6s2 +3,+4
60 Neodymium [X e]4f5d6s2 +3
61 Promethium [Xe]4f55d6s2 +3
62 Samarium [X e]4/65d6s2 + 3 ,+ 2
63 Europium [X e]4f5d6s2 + 3 ,+ 2
64* Gadolinium [Xe]4/75d16s2 +3
65 Terbium [Xe]4/95o!06s2 +3, +4
66 Dysprosium [X e]4/105d06s2 +3
67 Holmium [X e]4/n5rf6s2 +3
68 Erbium [X e]4/125d6s2 +3
69 Thullium [X e]4/135d6s2 + 3 ,+ 2
70 Ytterbium [X e]4/145d6s2 +3, +2
71 Lutetium [X e]4/145rf16s2 +3
72 Hafnium [X e]4/145d26s2 +4
73 Tantalum [X e]4fH5d36s2 +5
74 Tungsten [X e]4/145d46s2 +6, +5, +4, +3, +2
75 Rhenium [X e]4/145 2 +7, +6, +4, +2, - 1
76 Osmium [Xc]4f'45d66s2 +8, +6, +4, +3, +2
77 Iridium [X e]4/145d76s2 +6, +4, +3, +2
78* Platinum [X e]4/145d96s1 +4, +2
79 Gold [Xe]4fu5d106sl + 3 ,+ 1
80 Mercury [Xe]4fu5dw6s2 + 2 ,+ 1
81 Thallium [Xe]4/ iA5d106sl6p' + 3 ,+ 1
82 Lead [Xe]4fH5d106s26p2 +4, +2
83 Bismuth [X e]4/145rf106s26p3 + 5 ,+ 3
84 Polonium [Xe] 4 /145 d16s26p4 +4, +2
85 Astatine [Xe]4/,45d106s26p5 +7, +5, +3, +1, 1
86 Radon [Xe]4/145d106s26p6 0
87 Francium [Rn]7s1 +1
88 Radium [Rn]7s2 +2
89 Actinium [Rnjd^s2 +3
90 Thorium [Rn]5/6d27.52 +4
91* Protactinium [Rn]5/26d17s2 +5,+4
92 Uranium [Rn]5/36rf17i2 +6, +5, +4, +3
93 Neptunium [Rn]5/46rf17i2 +6,+5, + 4 ,+ 3
94 Plutonium [Rn]5/66d7s2 +6, +5, +4, +3
95 Americium [Rn]5/76d7s2 +6, +5, +4, +3
96 Curium [Rn]5/76d17.s2 +3
aThe electron configuration of starred elements is irregular.
72 THE ELECTRONIC STRUCTURE OF ATOMS
atoms in terms of electron orbitals based on the wave-
mechanical treatm ent of the m otions of elec trons in the
space around the nucleus. As a result, we can only
describe the m otion of an electron in terms of the
probability of finding it at a particular point in space. B
ohrs original model of the atom has been replaced by the
wave-mechanical treat m ent except in the explanation for
the generation of the characteristic x-ray spectra of the
elements.
The electronic structure of the elem ents can be
deduced from the possible orbitals of the hydrogen atom
by m eans of the A ufbau principle together with a set of
rules that tell us, in effect, how electrons distribute them
selves in the avail able orbitals. We find th at atom s
change the num
ber of electrons in such a way that partially filled orbitals
are em ptied completely, filled com plete ly, or are
occupied by only one electron per orbital in a set. This
property of atom s can be used to explain the valences of
the elem ents and hence their chemical properties.
We conclude that the chemical and physical
properties of the elem ents depend on the num ber of
protons in the nuclei of their atom s (atom ic num ber) and
on the distribution of the electrons around the available
orbitals. Knowledge of the electronic structure of the elem
ents therefore enhances our understanding of their
chemical properties and provides a rational explanation
for the construction of the periodic table.
Problems
 1. Familiarize yourself with the names, chemical 6. Without consulting Table 5.2, write the electron
symbols, atomic numbers, and electronic structures of the formula for a neutral atom having 14 electrons and
first 36 elements. predict its highest positive valence.
2. How many electrons can be accommodated in 7. Table 5.2 suggests that the valences of the ele ments
orbitals having the quantum number n = 4? (Answer: 32) vary systematically with increasing atomic num ber.
Determine how many sequences there are that end with a
3. Write a complete set of quantum numbers for = 4. valence of zero.
8. Elements having similar chemical properties tend to
4. Using the electronic formulas in Table 5.2, explain
be associated in nature. For the following pairs of
the valences of the following elements: Co, S, Br, V, and
elements indicate whether or not you expect them to be
Cu.
associated with each other. Base your answer on the
5. What do the elements of each group have in common electronic structures of these elements.
(a) K and Rb
(b) Al and Ge
(a) B, Al, Ga, and In (c) Sc and Cu
(b) Be, Mg, Ca, Sr, Ba, and Ra (d) S and Se
(e) Li and Mg
(c) N, P, As, Sb, and Bi
(d) K+,Ca2+,Sc3+,Ti4+,V 5+,Cr6+
 References
B O H R , N ., 1913a. Binding of electrons by positive nuclei. Phil. F Y F E , W. S., 1964. Geochemistry o f Solids. McGraw-Hill, New
Mag. J. Sei., Ser. 6, 26(l51):l-25. York, 199 pp.
B O H R , N ., 1913b. Systems containing only a single nucleus. G A M O W , G., 1966. Thirty Years That Shook Physics.
Phil. Mag. J. Sei., Ser. 6, 26(153):476-507. Doubleday, Garden City, NY, 224 pp.
 REFERENCES

H O L T O N . G., 1986. Niels Bohr and the integrity of science. 73

Amer. Scientist, 74:237-243.
J A S T R O W , R., 1967. Red Giants and White Dwarfs. Harper & P L A N C K , M., 1900. Zur Theorie des Gesetzes der Energie
Row, New York, 176 pp. verteilung im Normalspectrum. Verh. Deutschen Phys.
K L U G , H. P., and L. E . A L E X A N D E R , 1954. X-ray Diffraction Gesellschaft, 2(170):237-245.
Procedures. Wiley, New York, 716 pp. S C H R D I N G E R , E., 1926. Quantisierung als Eigenwert problem,
M O O R E , W. J., 1955. Physical Chemistry, 2nd ed. Prentice-Hall, Part 1. Ann. Phys., 79(4):361-376; Part 2, 79(6):489-527.
Upper Saddle River, NJ, 633 pp.
N O R R IS H , K., a n d J. T. H U T T O N , 1969. A n a c c u ra te x -ray sp ec - tro g ra
p h ic m e th o d fo r th e an aly sis o f a w id e ra n g e o f g e o logical sam
p les . Geochim. Cosmochim. Acta, 33:431^ -53 .
 6
The Periodic Table and
Atomic Weights
In the early p art of th e 19th century chemists
w ere alm ost overw helm ed by the discovery of
new elem ents. By 1830 the num ber of elem ents had
increased to 55, and there was no way of know ing how m
any m ore rem ained to be discov ered . This uncom
fortable situation m otivated efforts to organize the elem
ents in som e system atic ways th at m ight bring ord er
to this increas ingly chaotic state of affairs.
6.1 Mendeleev's Periodic Table
A lthough m any chem ists contributed to the effort to
organize the elem ents, D m itri Ivanovich M endeleev
(1834-1907) deserves to be recog nized as the inventor of
the periodic table (Asimov, 1965; van Sprosen, 1969).
M endeleev was a graduate student in
G erm any when Friedrich A ugust Kekule von
Stradonitz (known as K ekule) organized a m eet ing of
the leading chemists of E urope in 1860 in Karlsruhe. The
purpose of this First International Chemical Congress
was to debate the merits of
the structural form ulas of organic compounds
K ekule had proposed in 1858. These formulas
w ere based on the idea that carbon has a valence of 4 and
bonds with hydrogen, and even with itself,
to form increasingly com plex m olecular chains.
D uring the Congress in K arlsruhe the Italian
chem ist Stanislao Cannizzaro urged the delegates to
recognize the im portance of the atom ic weights of the
elem ents and to distinguish them from the equivalent
weights and the m olecular weights
based on an hypothesis his colleague A m adeo Avogadro
had advanced in 1811.
W hen M endeleev returned to Russia he used both
the valences and the atomic weights to orga nize the elem
ents into periods. In 1869 he pub lished a table in the
Journal o f the Russian Chemical Society in which he
identified several periods within which the valences rose
and fell in regular patterns. In the following year (1870),
the G erm an chemist Julius Lothar M eyer published a
graph of atomic volumes plotted versus atomic weights
that also revealed the periodicity of these properties.
However, M endeleev (1871) boldly reordered some elem
ents in his table on the basis of their valences and left
gaps where, he claimed, elem ents existed that had not yet
been discovered. For example, he predicted the existence
of an ele m ent similar to silicon and specified many of its
physical and chemical properties by comparison with
those of its neighbors in his table. His predic tions were
confirmed when C. A . W inkler discov ered germ anium
in 1886 during the analysis of the rare m ineral argyrodite
(Ag8 G eS6) (Weeks, 1956).
6.2 The Modem Periodic Table
The periodic table has assumed many forms since it was
first presented by M endeleev. M azurs (1974) described
over 700 different graphic rep resentations. Today we use
the so-called long form depicted in Figure 6.1.
The elem ents are arranged into horizontal rows
called periods and vertical columns called
74
 g i
.2 c

to
X .2

spouaj
k.
m
_2
c

oc
k_
cu
E

re
CD
cu

<U

0
*

tN
<55

59 68

CO
vo a
vo Q

vO Z
VO
1% o

LO U
3
i-

E
E

5
vo I

iO h -
u-> J3

T -O
vO >-

v> (X
iO LU

Pr Er
<
c

[c

u
ai

-0
4-*

cu
cu
E
QJ
WO

^
rx

U cri z 95 96 100 101 1 0 21 0 3

92
0^ 0-

O JZ
y~
3

^ Q.
Fm N o Lr
G> <Si

LU
co '*-

cn &

A mC m
CQ

t ffl
Md
n

u
Figure 6.1 The periodic table of the elements. The numbering of the groups has been changed to run consecutively from 1 to 18 across the table.
The old method is used here to emphasize the relationship between the electronic structures and highest positive valences of atoms.
76 THE PERIODIC TABLE AND ATOMIC W EIGHTS

groups. The periods are
num bered in accordance
with the first quantum num
ber of the orbitals that are
being filled with increasing
atom ic num ber. For
example, the elem ents of
the first period con tain
electrons in the Is orbital,
the elem ents of the second
period have electrons in
the 2s and 2 p orbitals,
those in third period are
filing 3s and 3p orbitals,
and in the fourth period
electrons enter 4s and 4p
orbitals. H ow ever, the
fourth period contains an
additional group of 1 0 elem
ents in which the 3d
orbitals are filling up.
Similarly, the fifth and
sixth periods each contain
elem ents in which d-
orbitals are being filled.
These elem ents together
form the so-called d-block
in the perio d ic table. The
sixth period contains 14
extra ele m ents known as
the rare earths with
electrons in 4/orbitals. The
rare earths are pulled out
of the table and placed
separately below it
together with the actinide
elements in which
5/orbitals are being filled
and which include the
transuranium or super
heavy elements. We see,
therefore, that the num ber
of the elem ents in each
period is a direct conse
quence of the progressive
filling of electronic
orbitals. These
relationships are evident in
Table 6.1.
attract an electron and
T he num bering of m ents in th at group. In actually form ions with a
the vertical colum ns or this version of the periodic charge of - 1 . N
groups has been confusing table the elem ents in the evertheless, they have
to say the least. They used s- and p-blocks form ed electronic structures th a t
to be num bered from left the A groups, and the m ake a valence of + 7 at
to right in accor dance least technically possible.
elem ents in the J-block
with the highest positive The last group of elem
com posed the B groups.
valence of the ele- ents are the noble gases H
How ever, the identifi
e, Ne, A r, Kr, Xe, and R
cation of A and B groups
n, all of which have filled
has not been consistent.
Table 6.1 R e la tio n sh ip BThe
e twvalences
e e n P e ofrio the
d s elem orbitals. Consequently,
a n d E le c tro n O rb ita ls o f Aents
to mins othe
f th A
e groups, as they neither gain n o r lose
E le m e n ts electrons under n a tu r al
used in Figure 6.1, can be
Period Orbitals being conditions and th erefo re
deduced directly from the
have a valence of 0 . In
filled electronic configurations
spite of th at fact, this
1 \s of the atoms. For example,
group used to be given the
group IA consists of H, Li,
2 2s,2p num ber 8 , although 0 is m
N a, K, Rb, Cs, and Fr,
3 3s, 3p ore appropriate.
each of which has one
4 4s, 3d, 4p electron in an s-orbital
Partly because of
5 5s, 4 d, 5p outside of a noble gas core
these kinds of inconsisten
6 6s, 4f 5d, 6p and thus a valence of +1.
cies, chemists have
7 Is, 5 f 6d, Ip Excluding hydrogen, the
decided to num ber the
elem ents in group IA are
Only 17 are known. groups consecutively from
know n as the alkali
left to right. In this new
metals. The elem ents in
scheme the alkali m etals
group IIA , called th e
are still in group 1 but the
alkaline earths, include
noble gases are now in
Be, Mg, Ca, Sr, Ba, and
group 18 (Figure 6.1).
Ra. These elem ents have
Nevertheless, we will
two electrons in an s-
continue the old num
orbital outside a noble
bering system outlined
aboveof +because it em
gas core and therefore have valences
The num bering of the A phasizes the
above continues w ith
p -block on the right side
O nce again the num bering
highest positive valence
consistent with the elec
tronic configuration . Thus
the elem ents in group 7A
can have valences of + 7
because they all have five
electrons in p -orbitals and
tw o elec trons in y-
orbitals. T he elem ents in
this group are
called th e halogens and include F,
and A t. In n atu re th e atom s of these
6.3 BASIC PRINCIPLES OF ATOMIC PHYSICS 77

close relationship betw een
electronic structure of the elem ents.
The elem ents w ith
d -orbitals are referred to
as the transition metals.
The groups in this block of
elem ents are num bered in
accordance with th e
valences of the ele m ents
just like the A groups. T
herefore, group IB consists
of the precious m etals Cu,
A g, and Au, which have
electron configurations of
dws 1 and therefore can
have a valence of +1. H ow
ever, cop per can also be
+2, and gold can be + 3 .
The num bering of the B
groups continues w ith IIB
com posed of Zn, Cd, and
H g with d ws2 electron
configurations that form +
2 ions. N ext, we move to
the left to IIIB (Sc, Y, and
La, d 1s2, valence + 3) and
progress from there to
group V IIB (M n, Tc, and
Re, d5s2, valence + 7). H
ow ever, th e next three
groups are anom alous. The iron group (Fe, Ru,
and O s) should have a valence of
dance w ith their electronic structure
iron actually has valences of + 2 and +3. Cobalt
and nickel also have valences of + 2
Therefore, these three
groups are lum ped togeth
er as group VIIIB . This is
n o t a very elegant solu
tion, but it has tradition on
its side. The congeners of
Fe, Co, and Ni form the
platinum group ele m ents
(P G E ) consisting of R u,
Os, R h, Ir, Pd, and Pt.
These elem ents can occur
in the native state and, in
addition, have valences of
+2, +3, + 4, + 6 , and even +
8 in the case of R u and Os.
The periodic table is
founded on the recur rence
of similar physical and
chem ical properties of the
elem ents with increasing
atom ic num ber. This
statem ent is know n as the
periodic law of the elem
ents. In addition, elem ents
in groups have the same
valences because they
have similar elec tronic
structures. A s a result,
their physical and
chemical properties tend
to be similar.
In geochem istry this
m eans that the elem ents
of a group tend to occur
together and display
geochemical coherence. For
m etals, the alkaline earths,
noble gases, and the
precious m etals display
well-developed geochem
ical coherence in their
distrb bution in nature.
the sam e place in the periodic table
6.3 Basic they have the sam e (or very sim ilar
properties.
Princip The internal
les of composition of the nuclei
of iso topes is indicated by
Atomi the shorthand notation
c used to identify them .
This notation consists of
Physic the chemi cal symbol of
s the elem ent, preceded by
the atomic
We know that atom s have
num ber (Z) w ritten as a subscript, and the
nuclei com posed of
num ber (A ) w ritten as a superscript, where:
protons and neutrons. The
A =Z+N
num ber of protons in the
nucleus determ ines the T herefore, the notation nbC indicates
positive charge and
therefore the num ber of atom or nuclide is an isotope of carbo
extranuclear electrons nucleus contains 6 protons, and th at the sum
required to form a neutral
protons and neutrons is 12.
atom . The num ber of
From equation 6.1 we find
electrons and their th at the nucleus also
distribution in the contains 6 neutrons.
available orbitals give M ost chemical elem
atom s their characteristic ents have two or m ore sta
chemical properties. ble isotopes. H owever,
Therefore, the atom ic num some elements, including
ber (Z) Be, F,Na, Al, P, Mn, Co,
and As have only one
indirectly determ ines the
stable
an atom . Since all atom s
m ust have the sam e
(or nearly the sam e)
chemical properties, it
follows that all atom s
of a given ele
m ent m ust have the sam e atom ic num ber.
The num ber (N
atom s affects atom ic m asses
bearing on chem ical
atom s of the sam e chem ical
tain different num bers
nuclei. A tom s th at have
ber b u t different n eu tro n num bers
isotopes. The w ord m
eans sam e place and
indi cates th a t the
isotopes of an elem ent
all occupy
78 THE PERIODIC TABLE AND ATOMIC W EIGHTS
isotope. Several other elem ents, including K, Rb, Sm, Lu, expressed has changed over the years. Originally, chemists
Re,T h, and U, have long-lived radioactive isotopes that used oxygen as the standard and gave it an atom ic weight
decay so slowly th at they still occur in the solar system of 16.0000. H ow ever, after the discovery of isotopes,
even though they form ed m ore physicists adopted a differ ent scale based on '(JO =
than 6 X 109 years ago by nucleosynthesis reac 16.0000. A ll of the resulting confusion has now been elim
tions in ancestral stars. In addition, m ore than inated by the adoption of l\C as the standard . Therefore,
2 0 0 0
short-lived radioactive isotopes are known the masses of atom s are now expressed in the atom ic
that do not occur in nature because they have decayed mass unit (am u), defined as 1 / 1 2 of the mass of C , which
since the tim e of nucleosynthesis. is the m ost abundant stable isotope of carbon. The masses
A fairly large group of short-lived radioactive of the naturally occurring iso topes have also been determ
isotopes does occur in nature because these iso topes are ined by m ass spec trom etry, and the results are tabulated
continuously produced either by the decay of long-lived in standard
parents (U and Th) or by nuclear reactions involving
cosmic rays. The unstable daughters of uranium and reference books such as the C R C H a ndbook o f
thorium are polonium , astatine, radon, francium , radium , Chemistry and Physics (W east et al., 1986).
actinium , and protactinium . A n o th er large num ber of We are now ready to define the atom ic
radioactive isotopes is produced by nuclear reactions in the weight of an elem ent.
atm osphere: tritium , ,H , ' Be, '^C , 1 7 CI, and m any
others. The atomic weight o f an element is the sum of the
masses of its naturally occurring isotopes weight ed
The abundances of the naturally occurring iso topes in accordance with their abundances.
(stable or long-lived radioactive) are d eter m ined with
mass spectrom eters and are expressed in percent by num A n exam ple will clarify what this means. Silicon has
ber. For exam ple, boron has two stable isotopes whose three naturally occurring stable isotopes whose
abundances a r e 'B = 19.8% abundances and masses are
and " B = 80.2% . This m eans th at out of 1000
Isotope Abundance, % Mass, amu
boron atom s 198 are the isotope 'B and 802 are
?|Si 92.23 27.976927
'jB . Actually, it turns out th at the isotopic com po sitions
of boron and of other elem ents of low atom ic num ber S si 4.67 28.976495
(H , C, N, O, and S) m ay be changed by isotope Si 3.10 29.973 770
fractionation because mass differences am ong the isotopes
of an elem ent affect the kinet ics of chemical reactions and The atom ic weight of silicon is found by m ultiply ing the
the strengths of chem ical bonds. These effects are p art of masses of the isotopes by their abun dances expressed as
the sub ject of isotope geoscience presented in detail in a decim al fractions and adding the resulting products:
textbook by Faure (1986) and will be discussed in C hapter
18 of this book. $Si: 0.9223 X 27.976927 = 25.803 12
fJSi: 0.0467 X 28.976495 = 1.353 20
iSi: 0.0310X 29.973770 = 0.929 18
Sum = 28.085 50
6.4 Atomic Weights
The concept of the atom ic
weight is fundam ental to
all form s of chem istry but
is n o t always clearly
understood . O ne reason
for the uncertainty is that
the scale by m eans of
which atom ic weights are
Therefore, the atomic
weight of silicon, rounded
to five significant figures,
is 28.086. Chemists regard
the atomic weights of the
elem ents as dimension-
 6.4 ATOMIC W EIGHTS 79
less num bers because they are expressed relative to the The relationship betw een m oles and the cor
mass of (G reenw ood and Eamshaw, 1984). responding num ber of atom s and m olecules enables us to
The im portance of the atom ic weights of the elem represent chem ical reactions by m eans of equations. For
ents arises from their use in the definition of the gram exam ple, the equation:
-atom ic weight. 2 H , + O , - 2 FLO (6.2)
The gram-atomic weight of an element is equal to the indicates that two m olecules of hydrogen (H 2)
atomic weight in grams.
m ust com bine w ith one m olecule of oxygen (0 2) in
Similarly, the gram -m olecular weight or the gram - form order to form two m olecules of w ater (H 2 0 ) . If
ula w eight of a com pound is defined as we mix 2 g of H 2 with 1 g of 0 2 we would not
achieve the desired proportions of molecules.
The gram-molecular weight (or the gram-formula H ow ever, w hen the am ounts of H 2 and 0 2 are
weight) of a compound is the molecular weight (or expressed in term s of moles, then the H 2 and 0 2
formula weight) in grams. m olecules will be present with the necessary 2 : 1 ratio
required for the reaction.
B oth the gram -atom ic weight and gram -m olecular The atom ic weights of the elem ents listed in Table
weight are referred to as the m ole (mol). This is the basic 6.2 enable us to calculate m olecular weights of com
unit of mass of elem ents and com pounds in chemistry. pounds based on their chem ical form ulas and to convert
am ounts of such com pounds from grams to the
The reason why the m ole is so im portant in
corresponding num ber of moles. For example, the m
chem istry is that one m ole of an elem ent or a
olecular weight o f B a S 0 4 (barite) is calculated from the
com pound always contains a fixed num ber of
atom ic weights of the ele m ents:
atom s or molecules. T hat num ber is:

N a = 6.022 045 X 102 3 atom s or molecules per Ba: 137.33

mole S: 32.06
O: 15.9994
which is know n as A vogadros number. It is a con
sequence of the hypothesis proposed by A m adeo The m olecular weight of barite is 233.366, and one m ole
Avogadro in 1811 that equal volumes of gases at the same of barite weighs 233.366 g. The solubility of barite in cold
pressure and tem perature contain equal num bers of atom w ater is 2.22 X 10- 4 g/100 m L (W east et al., 1986).
s or molecules. Therefore, the concept of the mole Therefore, one liter of a satu
together with A vogadros num ber enables us to convert rated solution contains 2.22 X 10- 3 g of barite,
an am ount of an ele m ent or com pound from grams to which corresponds to:
the correspond ing num ber of atom s or molecules. In 2.22 X 10~ mol
addition, we can now appreciate the form al definition of = 9.513 X 10~ 6

the m ole stated on the inside back cover of this book. 233.366
W hen barite dissolves in w ater, it dissociates

The mole is the amount of a system that contains as into ions:

many elementary entities as there are atoms in 0.012 BaSCK - B a2+ + SCE (6.3)
kg of carbon-12.
The equation indicates that each m ole of barite that
In other words, 6.022 X 102 3 atoms, molecules, ions,
dissolves produces one mole of Ba2+ and one mole
electrons, or other specified particles m ake up one mole.
of SO2 - . Since 9.512 X 1(T6 m ol of barite dissolves
in 1 L of a saturated solution, the concentration of
Table 6 .2 A to m ic W eig h ts o f th e E le m e n ts R e la tiv e to 'j?C
z Element Atomic Weight Z Element Atomic Weight

1 H 1.00794 48 Cd 112.41
2 He 4.00260 49 In 114.82
3 Li 6.941 50 Sn 118.71
4 Be 9.01218 51 Sb 121.75
5 B 10.81 52 Te 127.60
6 C 12.011 53 I 126.905
7 N 14.0067 54 Xe 131.29
8 O 15.9994 55 Cs 132.905
9 F 18.9984 56 Ba 137.33
10 Ne 20.179 57 La 138.906
11 Na 22.9898 58 Ce 140.12
12 Mg 24.305 59 Pr 140.908
13 A1 26.9815 60 Nd 144.24
14 Si 28.0855 61 Pm (145)a
15 P 30.9738 62 Sm 150.36
16 S 32.06 63 Eu 151.96
17 Cl 35.453 64 Gd 157.25
18 Ar 39.948 65 Tb 158.925
19 K 39.0983 66 Dy 162.50
20 Ca 40.08 67 Ho 164.930
21 Sc 44.9559 68 Er 167.26
22 Ti 47.88 69 Tm 168.934
23 V 50.9415 70 Yb 173.04
24 Cr 51.996 71 Lu 174.967
25 Mn 54.9380 72 Hf 178.49
26 Fe 55.847 73 Ta 180.948
27 Co 58.9332 74 W 183.85
28 Ni 58.69 75 Re 186.207
29 Cu 63.546 76 Os 190.2
30 ' Zn 65.39 77 Ir 192.22
31 Ga 69.72 78 Pt 195.08
32 Ge 72.59 79 Au 196.967
33 As 74.9216 80 Hg 200.59
34 Se 78.96 81 Tl 204.383
35 Br 79.904 82 Pb 207.2
36 Kr 83.80 83 Bi 208.980
37 Rb 85.4678 84 Po (209)a
38 Sr 87.62 85 At (210)a
39 Y 88.9059 86 Rn (222)a
40 Zr 91.224 87 Fr (223)a
41 Nb 92.9064 88 Ra 226.025
42 Mo .95.94 89 Ac 227.028
43 Tc (98 )a 90 Th 232.038
44 Ru 101.07 91 Pa 231.036
45 Rh 102.906 92 U 238.029
46 Pd 106.42 93 Np (237)a
47 Ag 107.868 94 Pu (239)a
aMass number of longest-lived isotope in nature.
 6.5 SUMMARY 81

both Ba2+ and SO2- is also 9.512 X 10- 6 m ol/L .The of hydrogen ions per m ole of the com pound, and
concentration of Ba2+ in 1 L of saturated barite solution therefore its equivalent weight is equal to its m olecular
can also be expressed as: weight divided by three. In this case, it m akes no
difference that phosphoric acid actual ly gives up its
9.512 X 1CT6 X 137.33 = 1.306 X 10 3 g /L hydrogen ions reluctantly and releases all three of them
A lternatively, we calculate the num ber of B a2+
only u n d er special con ditions.
ions in one liter of a satu rated barite solution by m eans of
A vogadros num ber:

9.512 X 1(T6 X 6.022045 X 102 3

6.5 Summary
= 5.728 X 101 8 ions/L
The periodic table was originally invented by
Such com putations involving conversions of am D. I. M endeleev and other chem ists in the 19th
ounts of elem ents or com pounds in grams to the century based on the relationship betw een the
corresponding num ber of m oles by m eans of the atom ic atom ic weights of the elem ents and their chemical and
or m olecular weights occur com m only in geochemistry. physical properties. We now know that these relationships
are caused by the com position of the nuclei of the atom s
B efore we close this chapter we define the gram and by the distribution of their electrons in the available
-equivalent weight. orbitals.
The so-called long form of the periodic table is
A gram-equivalent weight o f an ion is the gram-
organized into seven horizontal rows called periods and
atomic or gram-molecular weight divided by the
eighteen vertical colum ns called groups. The physical and
valence. For acids and bases the gram-equivalent
weight is the gram-molecular weight of the com chemical properties of
pound divided by the number of hydrogen ions or elem ents in the periods change system atically
hydroxyl ions the acid or base can produce when w ith increasing atom ic num ber. The elem ents in
dissolved in water. the groups tend to have sim ilar properties
because they have sim ilar electronic structures.
Equivalent weights are used in geochem istry The concept of the atom ic w eight is now
w hen electrical charges if ions are an im portant based on the masses and abundances of the natu rally
consideration. In addition, they are traditionally used to occurring isotopes of the elem ents. The atom ic weights
express the concentrations of acids or bases in a way th at are used to define the gram -atom ic weight of elem ents
indicates the concentrations of hydrogen or hydroxyl ions or the gram -m olecular
that may be available for reaction. w eight of com pounds, also know n collectively, as
the mole. The im portance of the m ole in chem
In practice, the equivalent weight of an ele m ent is istry arises from the fact that one mole of an ele
calculated by dividing its atom ic weight by the charge the m ent or com pound always contains the same
atom s of that elem ent acquire num ber of atom s or molecules. This num ber is
w hen the elem ent is dissolved in water. know n as A vogadros num ber.
T herefore, the equivalent weight of a divalent Chemical reactions are represented by equa tions in
elem ent like calcium is equal to the atomic which the quantities of reactants and products are specified
w eight divided by two. Conversely, one mole of Ca2+ in term s of moles, which can be converted to the num bers
contains two equivalents of Ca2+. In the case of m of atom s or m ol ecules by virtue of A vogadros num ber.
olecules of an acid such as H 2 S 0 4, the equivalent weight The gram -equivalent w eight takes into account the
is equal to one half of the m ol ecular weight because one electrical charges of ions or the num ber of hydrogen or
m ole of H 2 S 0 4 con tains two equivalents of hydrogen hydroxyl ions th at acids or bases can release into an
ions. Similarly, phosphoric acid (H 3 P 0 4) can yield three aqueous solution.
moles
82 THE PERIODIC TABLE AND ATOMIC W EIGHTS

Problems
 1. Familiarize yourself with the names, atomic num 5. What is the weight of one atom of 2g2U in grams
bers, and positions in the periodic table of the elements given that the mass of this isotope is 238.050 784 amu?
having atomic numbers from 1 to 36. (Answer: 3.953 x 10~22 g)
2. Calculate the atomic weight of magnesium given the 6. How many atoms of iron are present in 5.00 g of
abundances and masses of its naturally occurring isotopes. hematite (Fe20 3)? (Answer: 3.77 X 1022 atoms)
7. What is the concentration of Cr in the mineral
chromite (FeCr20 4)? (Answer: 46.5%)
Abundance, % Mass, amu
?2Mg 78.99 23.985 042 8. If the concentration of S 0 4~ is 2.5 x 103 mol/L,
how many grams of SO|" are present in one liter?
Mg 10.00 24.985 837 (Answer: 0.24 g)
?^Mg 11.01 25.982 593
9. A solution of A12(S 0 4)3 contains 2 X 10 4 mol/L of
Al3+. H ow many grams of SO2- does it contain per liter?
(Answer: 24.305)
(Answer: 0.029 g/L)
3. Calculate the formula weight of orthoclase feld spar
10. How many gram-equivalent weights of Al3+ and
(KAlSi30 8) using the atomic weights in Table 6.2.
S 0 4' are in 2 mol of A12(S 0 4)3?
(Answer: 278.332)
(Answer: Al3+: 12; SO;". 12)
4. How many moles of Na+ are present in one liter of an
aqueous solution of Na2S 0 4 containing 4.760 g of the
compound? (Answer: 0.0670 mol)
References
A S IM O V , I., 1965. A Short VAN SPRO SE N, J. W., 1969. The
History o f Chemistry. Periodic System o f the
Doubleday, Elements.
Garden City, NY, 263 pp. Elsevier, Amsterdam, 368 pp.
F A U R E , G., 1986. Principles o f W E A S T , R. C., M. J. A S T L E , and W. H.
Isotope Geology. 2nd ed. Wiley, Handbook o f Chemistry and
New York, 589 pp. Raton, FL.
G R E E N W O O D , N. N., and A. E A R W E E K S , M . E ., 1956. Discovery
N S H A W , 1984. Chemistry o f the
Chem. Ed. Pub., 910 pp.
Elements. Pergamon Press,
Oxford, England, 1542 pp.
M A Z U R S , E . G ., 1974. Graphic
Representation o f the Periodic
System During One
Hundred Years. University
of Alabama
Press, Tuscaloosa.
 7
Chemical Bonds, Ionic
Radii, and Crystals
The physical and chem ical properties of the ele m ents th e nonm etals. If a com plete transfer of elec
vary with increasing atom ic num ber in accordance with trons is achieved, the electron -donor atom s
the periodic law. Som e of these properties are neatly sum becom e positively charged ions called cations,
m arized in periodic tables available in university book w hereas the electron -acceptor atom s acquire a
stores. It will be helpful to use such a table to study the negative charge and form anions. O ppositely
topics presented in this chapter. charged ions are attracted to each other and
m ay form an electrostatic or ionic bond. M ost
m inerals are chem ical com pounds com posed of
cations (m etals) and anions (nonm etals) held
7.1 Electron Donors Versus together by ionic bonds.
Acceptors W hen two elem ents having sim ilar m etallic
or nonm etallic character interact, electrons are
We can divide the elem ents in the periodic table into two not actually transferred, but are shared by the atoms. The
groups based on their tendency to give up electrons or to sharing of electrons am ong the atom s of two or m ore
attract them . In general, the elem ents on the left side of different elem ents establishes a covalent bond betw een
the periodic table are the electron donors because their them . The elem ents do not necessarily share the
electronic orbitals are largely empty. The elem ents on the electrons equally, which causes covalent bonds to have a
right side of the periodic table have orbitals that are certain am ount of ionic character. Similarly, the transfer
nearly full, which causes them to be electron acceptors. of elec trons am ong atom s of elem ents on opposite sides
The elem ents that are located betw een the alkali m etals of the periodic table is never com plete. As a
on the left and the halogens on the right vary result, ionic bonds have covalent character
progressively in their tendencies to be electron donors or w hose m agnitude depends on the difference in the
acceptors, depending on how close they are to filling their tendency of the atom s to attract electrons to them selves.
orbitals with As atom s becom e m ore equal in their ability to attract
electrons, the covalent character of the bond increases.
electrons. The electron donors are the metals,
The distinction betw een ionic and covalent bonds is th
w hereas the electron acceptors are the non-
erefore an over-sim plification of a very com plex
metals. The m etallic character of the elem ents
interaction am ong atom s of different elem ents (Pauling,
generally decreases from left to right in the peri ods as
1960). The only truly covalent bonds occur when tw o
the orbitals are filling up with electrons.
atom s of the sam e elem ent com bine to form diatom ic
The properties of m etals and nonm etals are com
m olecules such as H 2 , 0 2, or N 2.
plem entary in the sense that the electrons given up by m
etallic elem ents are picked up by
8
3
84 CHEM ICAL BONDS, IONIC RADII, AND CRYSTALS
 that, in fact, perm it us to
do this. They are
7.2 M identified and defined as
e follows.
a reflects the reluctance of
s First ionization
potential is the
u energy required to
r remove one electron
from a neutral atom
e in a vacuum and to
s place it at rest an
infinite dis tance
o away.
f Electronegativity is a
measure of the
M ability of an atom in
a molecule to attract
e electrons to itself
t (Pauling, quoted in
Cotton and
a Wilkinson, 1962, p.
ll 89).
Standard electrode
i potential is the
c voltage gener ated
when one mole of
C electrons is removed
h from one mole of an
element or ion in
a the stan dard state.
r elem ents with elec trons
a We deal with the standard
c electrode potential in
Section 14.3 and consider
t here how the first ion
period . The increase
coincides with the progres
sive filling of the
electronic orbitals and
atom s to give up electrons
from orbitals th at are
nearly full. For exam ple,
we see that in periods 2
and 3 the first ionization p
o ten tials increase sharply
w ith increasing atom ic
num ber tow ard th e noble
gas elem ents th at te r m
inate each period . W e can
even recognize the effect
on the first ionization
potential of filling the s-
orbitals or half-filling the
p-orbitals. This explains
why the first ionization
potential of B is less than
th at of Be and why it
takes less energy to rem
ove an electron from O
than from N. The
same effect can be seen in
period 3 in the first ion
ization potentials of M g
and A1 and of P and S.
The first ionization
potentials of the transi tion
m etals of perio d 4, which
have partially filled 3d
orbitals, vary m uch less
than those in the first three
periods. P eriod 4 starts out
well enough with K and
Ca, b u t th en th e first
ioniza tion potentials
change little as we com
plete the
sequence ending w ith Zn . G allium
the 4p orbitals, which causes it to
ionization potential than Z
n . H ow ever, the first
ionization potentials th en
rise rapidly with
increasing atom ic num
ber until period 4 ends w
ith Kr. The next period, 5,
starting w ith Rb and
ending with Xe, likew ise
illustrates the lack of
differentiation am ong the
in the Ad orbitals. The first
ionization potentials for
these and all other elem
ents are listed in Table 7.1.

e ization potentials and the

r used to com pare the m
A lthough we can
appreciate the difference
betw een electron donors
and electron acceptors
qualitatively, we need to m
easure these tenden cies
quantitatively in order to
be able to predict w hat
kind of a bond the atom s
of two elem ents are likely
to form . Several param
eters have been defined
Evidently, the system
electronegativities can be
atic variation of the first
ionization potentials
etallic character of dif
confirms that the metallic
ferent elements.
character of the elem ents
in each period decreases
The first ionization
with increasing atom ic
potential indicates how
num ber. The first ioniza
strongly the nucleus of a
tion potentials even reflect
neutral atom attracts an
the reduction in the energy
electron in a partially filled
of atom s that have com
orbital. Figure 7.1 A
pletely filled or half-filled
indicates th at the first
orbitals and, in general,
ionization potentials
perm it us to rank the elem
increase with increasing
ents throughout the
atom ic num ber in each
periodic table in terms of
their metallic character,
which arises from their
tendency to be electron
donors.

Ionization potentials
required to remove
additional electrons rise to
high values because it
takes extraordinary am
ounts of energy to rem ove
electrons from atom s that
have acquired a positive
charge because they have
already lost one elec tron
or because the rem aining
electrons com plete ly fill
the orbital they occupy.
Therefore, during chemical
reactions in nature elem
ents generally lose only
their valence electrons,
that is, those that reside
outside of filled orbitals.

The problem with

ionization potentials as
predictors of chemical
properties is that the
chemical reactions we w
ant to study do not occur
in a vacuum b u t take
place in m ixtures of
gases, in aqueous
solutions, in silicate melts,
and even in the solid state.
Therefore, we need a
param eter that is
appropriate for the
environm ents in which
chemical reactions
actually take place.
7.2 MEASURES OF METALLIC CHARACTER 85
 Periods and Orbitals Figure 7.1 A: Variation of the first ionization
3 potential with increasing atomic number in the
first four periods. The first ionization potentials
increase as the electronic orbitals are filling up in
First Ionization Potential each period. Note that Be and N have exception
ally high ionization potentials consistent with fill
ing or half-filling of electronic orbitals. The
ionization potentials of elements in period 4 with
,/! partially filled T-orbitals vary little with atomic
number. In general, the data show that the
c 15- * / Kr
charac ter of the elements in each period changes
r Br/i from metallic (low ionization potential, hence
As
o 10 - I Be '/ y Zn / 1 electron donor) to nonmetallic (high ionization
A/ I
1/ B I Mg
f*Ge 1
potential, hence electron acceptor).
I J
Na J Al y ea Rb/ B:Variation of the electronegativity of the ele
ments with increasing atomic number. The dia
Electronegativity
gram shows that the electronegativity increases as
the available orbitals become filled with electrons.
/I /
/ i This pattern, therefore, reflects the increasing ten
dency of elements in the periods to fill their
/ / Zn~
I
/ orbitals by attracting electrons to themselves. Note
that metallic elements have low electronega
*Li Na
1 K I *Rb tivities, whereas nonmetallic elements have high
I Ar I Kr values. Therefore, the electronegativities also
c L Halogens reflect the progressive decrease of the metallic
A character of elements with increasing atomic
Covalent Bond
number in each period.
C:Variation of the covalent character of single
J5 60 - Character bonds with oxygen based on electronegativity dif
U ferences (Table 7.2). As the elements lose their

J.-AV/ metallic character and become increasingly reluc

tant electron donors, the bonds they form with
Alkali metals / oxygen are increasingly covalent in character. In
Na y
Rb general, the alkali metal-oxygen bonds are highly
10 I
r-| i r 1 _r r 35 r~rr ionic, whereas the halogen-oxygen bonds are
20 25 30 40 highly covalent. All other elements in the periodic
Atomic Number table form bonds with oxygen whose character is
intermediate between these extremes.
 The
electronegativities,
proposed by Pauling
(1960), are a set of dim
ensionless num bers
calcu lated from the
known strengths of bonds
between atom s in
molecules. They are a m
easure of the ionic
character of covalent
bonds and therefore
indicate the extent to
which two atom s in a m
ole cule actually share
their valence electrons
equal ly. Elem ents having
a low electronegativity act
as electron donors, w
hereas elem ents with
high elec tronegativity act
as electron acceptors.
Cesium has the lowest
electronegativity (0.7) and
th ere fore has the
strongest tendency to be
an electron donor, w
hereas fluorine has the
highest elec-
tronegativity (4.0) and is
the strongest electron
acceptor. The
electronegativities of all
naturally occurring elem
ents are listed in Table 7.1
and are displayed in
Figure 7.1B.
A s expected, we find
th at the electronegativi
ties rise with increasing
atom ic num ber in each
period. The
electronegativities of the
noble gases are shown as
zero because these elem
ents do not attract
electrons to themselves.
The alkali m et als and the
halogens dem onstrate that
the elec tronegativities of
group A elem ents
decrease with increasing
atom ic num ber. The first
ionization potentials of
these elem ents also
decrease with
Table 7.1 F irst Io n iz a tio n P o te n tia ls a n d E le c tro n e g ativ ities o f th e E le m e n ts
 st et al.
(1986).
bPaulin
z
g (1960).
Element Electronegativity b
1 2.1

2
3 1.0
4 1.5
5 2.0
6 2.5
7 3.0
8 3.5
9 4.0
10
11 0.9
12 1.2
13 1.5
14 1.8
15 2.1
16 2.5
17 3.0
18
19 0.8
20 1.0
21 1.3
22 1.5
23 1.6
24 1.6
25 1.5
26 1.8
27 1.8
28 1.8
29 1.9
30 1.6
31 1.6
32 1.8
33 2.0
34 2.4
35 2.8
36
37 0.8
38 1.0
39 1.3
40 1.4
41 1.6
42 1.8
43 1.9
44 2.2
45 2.2
46 2.2
47 1.9

W
e
a

increasing atom ic num ber in each group.This indi cates
that the binding energy of the first valence electron of
elem ents in a given group decreases as the volume of the
atom s in th at group increases. In other words, the larger
an atom , the m ore loosely it holds its valence electrons.
Consequently, the
metallic character of the group A elem ents
increases with increasing atom ic num ber.
In the B groups, w here d- and /-orbitals are
being filled, the electronegativities change m uch
less dram atically with increasing atom ic num ber
than they do in the A groups. A s a result, the
chemical properties of these elem ents are m ore
uniform than those of the elem ents in the
A groups. M oreover, the electronegativities of
elem ents in B groups do not decrease w ith
increasing atom ic num ber as they generally do in the A
groups. For exam ple, the electronegativities of the
congeners of group V IIIB (Fe, Co, and Ni) actually rise
with increasing atom ic num ber.
The concept of electronegativity has m erit because
the differences in the electronegativities of elem ents are
related to the ionic character of the bond form ed by their
atoms. Accordingly, two elem ents with similar
electronegativities form covalent bonds with little ionic
character, whereas elem ents having different
electronegativities form bonds whose ionic character is
proportional to the m agnitude of the electronegativity
difference. The relationship betw een the electronegativity
differ ence and the ionic character of chemical bonds is
expressed quantitatively by the data in Table 7.2. In Figure
7.1C we see how the bond character
varies with increasing atom ic num ber when ele
m ents form bonds with oxygen. N ote that the
graph shows the percent covalent character in order to em
phasize the relationship between bond character and
electronegativity (Figure 7.1B) or first ionization potential
(Figure 7.1A). We see that the oxygen bonds of the elem
ents become increasingly covalent in each period. The
halo gens form the m ost covalent oxygen bonds, w hereas
the alkali m etals form oxygen bonds that are largely ionic.
The transition m etals in the fourth period show less
variation in bond charac ter than elem ents with valence
electrons in s- and p-orbitals.
7.3 BONDING IN MOLECULES 87
The data in Tables 7.1 and 7.2 perm it us to
predict the ionic character of the bond betw een any two
elem ents in the periodic table. In general, we find that
bonds betw een neighboring elem ents in the periodic table
are highly covalent. For example, the nonm etals on the
right side of the periodic table form m any fam iliar com
pounds and anion com plexes that are covalently bonded.
The H O bond in the H 2 0 , the C O bond in
CO 2 - , the N O bond in N O 3 , the S O bond
in S 0 4~, and the Si O bond in S i0 4_ are all
strongly covalent.
In addition, we can dem onstrate th at bonds betw een
elem ents from opposite ends of the peri odic table are
highly ionic. The best exam ples are the alkali halides (e.g.,
NaCl, KC1, and CsF). In fact, the Cs F bond is the m
ost ionic bond we know because Cs has the strongest
tendency to be an electron donor (lowest electronegativity)
and F is the strongest electron acceptor (highest elec
tronegativity). B onds with w ell-developed ionic character
also occur betw een the cations of m et als and com plex
anions exem plified by C aC O ,, N a N 0 3, M g S 0 4, and
Fe2 S i0 4, all of which occur naturally as minerals.
A lthough the electronegativities of complex anions
are not defined, we nevertheless know that the bonds betw
een the cation of the m etals and the complex anions are
highly ionic because these com pounds dissociate into ions
w hen they dis solve in water.
7.3 Bonding in Molecules
We now com e to the question: W hat difference does it m
ake in geochem istry w hether chemical bonds are covalent
or ionic? The answ er is that the physical and chem ical
properties of all com pounds depend on the character of
the bonds that hold them together.
A n im portant property of most m inerals is that they
dissociate into ions when they dissolve in water. For
example, w hen calcite (C a C 0 3) dis solves in water, it
forms Ca2+ ions and C O 2- ions:
C a C 0 3 -> Ca2+ + C O f" (7.1)
88 CHEM ICAL BONDS, IONIC RADII, AND CRYSTALS

Table 7.2 Percent Ionic Character of a Single Chemical Bond with

Oxygen
Difference in Ionic Difference in Ionic
electronegativity character, % electronegativity character, %
0.1 0.5 1.7 51

0.2 1 1.8 55
0.3 2 1.9 59
0.4 4 2.0 63
0.5 6 2.1 67
0.6 9 2.2 70
0.7 12 2.3 74
0.8 15 2.4 76
0.9 19 2.5 79
1.0 22 2.6 82
1.1 26 2.7 84
1.2 30 2.8 86
1.3 34 2.9 88
1.4 39 3.0 89
1.5 43 3.1 91
1.6 47 3.2 92

SOURCE: Sargent-Welch (1980).

The carb o n ate anion does electrical polarity and is th erefo re said to
n o t dissociate into carbon po lar.
and oxygen ions because th The m anner in which the
e C O bonds are highly bonds form also determ ines
covalent, w hereas the C a w ater molecule. A quick review of the electronic
C 0 3 bond is largely structure of O indicates
ionic. W hy should th at m that it has four electrons in
ake a difference? T he answ the 2p orbitals, whereas H
er has to do w ith th e has one electron in the Is
interactions b etw een th e orbital. The bond betw een
m olecules of the solvent them form s by sharing of
(w ater in this case) and th the unpaired electrons. In
e calcite m o l ecules. accordance with rule 3 of
the Aufbau principle (C
hapter 5), one of the p
T he H O b o n d -orbitals of oxygen
ab o u t 60% covalent as contains two elec trons
Figure 7.1C. In ad d itio n , with paired spins, whereas
of oxygen (3.5) is larg er ththe
a nother two p-orbitals of
(2.1). C onsequently, weoxygen contain one
attracts th e electro n s o felectron
th e each. We can
T he u n eq u al sharing o f represent the p-orbitals of
th e electro n b etw een H oxygen as cir cles and the
and O th erefo re causes a electrons they contain by
n et negative charge to arrow s that point up or
develop on th e O atom s down depending on the
w hereas th e H atom s spin.
charge. In o th e r w ords, th e w ater m olecule has

(ID
Px Py
 7.4 IONIC CRYSTALS 89
The angle betw een these p -orbitals of oxygen is 90 strongly th at they are in co rp o rated into crystals
because they are sym m etrical about the x-, y-, and e-axes. as w ater of hydration, as, for exam ple, in
T herefore, we expect that the angle betw een the H O C a S 0 4 2H20 (gypsum ).
bonds in the w ater m olecule should also be 90. H ow Therefore, the answ er to the original question
ever, the observed angle is th at w ater is able to dissolve ionic solids
is 104.45 (Fyfe, 1964). because it dissociates them into ions, which are then
T he discrepancy arises because the s- and separated from each other by becom ing
p -o rb itals of O are m erged in such a way th at hydrated. H ow ever, covalent bonds do not disso
th e angles b etw een the h y b rid ized orbitals are ciate in w ater or in other polar solvents.
no longer 90. H ybridization of orbitals is a Therefore, w ater is generally not an effective sol vent for
com m on ph en o m en o n in covalently bonded covalently bonded com pounds.
m olecules and is th e key to understanding the
shapes or ste reo ch em istry o f covalently bonded m
olecules (G reenw ood and E arnshaw , 1984; Pauling, 7.4 Ionic Crystals
1960).
The polarity of the w ater m olecule therefore arises W hen ions com e together to form a crystal they assem ble
not only because of th e unequal sharing of the electrons in regular patterns dictated by the requirem ent of closest
betw een H and O but also because one of the hybridized p packing. This require m ent arises because the strength of
-orbitals of O containing the two paired electrons points the electro static bond (F) betw een cations and anions is
away from the other two, which are involved in the inversely proportional to the square of the dis tance (r)
bonding. In addition, the H zO m olecule is inherently betw een them:
polar because the O atom has m ore electrons than the H
atoms.

Because of their electrical polarity, w ater

m olecules are attracted to electrical charges on w here e x and e2 are the charges of the ions.
the surface of an ionic crystal such as C a C 0 3. The Evidently, th e m agnitude of F increases as
resulting electrostatic interactions w eaken the ionic bonds r decreases. T herefore, the ions in a crystal
betw een the C a2+ and the CO 2- ions, but have no effect arrange them selves in such a way th a t the in teri onic
on the covalent C O bonds of the carbonate ions. distances are m inim ized, hence the te n d e n cy for
closest packing. In this context, ions are assum ed to be
The ions th a t form by dissociation attract incom pressible spheres like bil liard balls.
w ater m olecules to them selves and thereby
becom e h y d rated . The num ber of w ater m ole cules A second requirem ent is th at crystals m ust be
bound to an ion depends prim arily on its radius b u t also electrically neutral. This requirem ent determ ines the ratio
on its charge. For exam ple, Be2+ of cations to anions and is reflected in
(radius = 0.35 ; A hrens, 1952) holds four the chemical form ulas of m inerals. In addition,
w ater m olecules to form the hydrated ion the requirem ent for electrical neutrality m ust be
B e(O H 2)2+ w hereas A l3+ (radius = 0.51 ) obeyed w henever substitutions occur in ionic crystals.
holds six w ater m olecules and form s A l(O H 2 )g+. N ote
th a t we rep resen t the hydrated B e2+ ion as B e(O H 2 )4 The arrangem ent of ions in a crystal can be
+ because the B e2+ is attracted to the negatively charged understood in term s of the m agnitudes of the radii of the
O atom s of the w ater m ole cule. A nions are also h y d cations and anions expressed by the radius ratio, which is
rated by attracting the positively charged hydrogen atom s defined as the ratio of the radius of the cation to th at of the
of water. In som e cases th e w ater m olecules are held so anion. In calculating radius ratios one com m only uses
O2- as the anion
90 CHEM ICAL BONDS, IONIC RADII, AND CRYSTALS

Coordination

f - -'55
 radius ratio exceeds a structure when the radius
critical value, ratio is

Figure 7.3 Derivation of

the critical radius ratio for
threefold coordination. Rc
and R3 are the ionic radii
* of the cation and anion,
respectively. The critical
relation ship is sin 60 = R
0.414 j( R a + Rc) from which it
R. follows that Rc/R . = 0.155.

Figure 7.2 Examples of the coordination num ber

closest packing of anions rises to 3. A t a still high er
around cations at value of the radius ratio,
increasing values of the four anions can be packed
radius ratio. When R c/R
around the cation.
a > 0.155, three anions fit
The critical value of
around the cation and
when Rc/R r > 0.414, the radius ratio at which
fourfold coordination threefold coordination is
occurs. Coordination possible in a planar
numbers 6,8, and 12 are structure can be derived
also important in mineral from geom etrical consid
structures and are
erations pictured in Figure
pictured in Figure 7.4.
7.3. L et R e and R a be the
radii of the cation and
anion, respectively. By
examining Figure 7.3 we
because oxygen is the m ost abundant find that:
elem ent in the crust of the Earth .
The geom etrical arrangem ents of
crystal are described by the sin t) =
coordination number, Rc+Ra
which is the num ber of since 0 = 60 and sin 60 = 0.8660:
anions th at surround a
R'
particular cation in an ionic
crystal. It turns out th at the
Rc + R a
coordination num bers of
elem ents in ionic crystals By cross multiplying and
can be predicted based on collecting terms we
their radius ratios. We see obtain:
from Figure 7.2 th at only
two large anions can fit (7.5)
around a very sm all R.
cation. The sm allest
0.866
coordination num ber
therefore is 2. If the cation Therefore, threefold
is larger than in the case coordination becom es
just consid ered and the possi ble in a planar
 7.5 IONIC RADII 91
Radius Coordination Arrangement of ions
equal to or greater than 0.155. A similar argum ent
indicates th at fourfold coordination occurs when the ratio
O --# -O Linear
radius ratio is 0.414. <0.155 2
All of the packing arrangem ents shown in Figure 7.2
are two dim ensional or planar. Since crystals are actually o
three-dim ensional bodies, we m ust also consider three-
0.155-0.225 3 Trigonal

cA o
dim ensional patterns of fitting anions around cations. For
planar
example, four fold coordination occurs when the anions
occupy the corners of a tetrahedron with the cation in its
center. Similarly, sixfold coordination is realized in an
octahedron w here six anions occupy the cor 0.225-0.414 4 O A \ Tetrahedral

ners and the cation resides in the center. H igher

coordination num bers of 8 and 1 2 are common
and involve the cubic structures shown in \7 ~
Figure 7.4. O ther coordination num bers, for o-
example, 5, 7, 9,10, and 11, occur only rarely and are not
0.414-0 .732 4 Square
im portant in the crystal structures of m in erals (Fyfe,
1964).
planar
7.5 Ionic
Radii
The radii of ions in crystals were first determ ined
by V. M. G oldschm idt and Linus 0.414-0.732
Pauling in
Octahedral
late 1920s by means of x-ray diffraction and
expressed in angstrom s (; 1 = 10
radii are determ ined
betw een ions in crystals
structure of the crystal in
which they w ere m ea
sured. M ost ionic radii w 0.732-1.00
ere originally derived
from crystals in which
each cation is surrounded
by six anions (H eydem ann, 1969). A com plete
set of such radii was published by A hrens
O-
(1952), but these have
> 1.00

y
now been superseded by 12

the work of Shannon and Io

Prew itt (1969).
Subsequently, W
hittacker and M untus C atio n , OAnion
(1970) published a new Figure 7.4 Possible
set of ionic radii (listed arrangements of ions
in in crys tals for
Table 7.3) th a t takes into account the different
limiting values of the
coordination num bers th at may occur in differ
ent kinds of
radius ratios and dif
crystals. A nother set of radii was
later published by ferent values of the
Shannon (1976). The d coordination
epen dence of ionic radii number.
on the coordination num
ber occurs because the
electron cloud of an ion
can
be deform ed by the electrical charges that sur
round it in a crystal lattice. For exam ple,
radius of N a + varies from 1.07 to 1.40
Table 7 .3 R a d ii o f th e Io n s o f th e E le m e n ts in A n g stro m U n its ( ) fo r D iffe re n t C h arg es
a n d C o o rd in a tio n N u m b e rs
Coordination numbers

z Symbol Valence II III IV V

vrh VII VIII IX X XII

1 H +1 extremely small
3 Li +1 0.68 0.82
4 Be +2 0.25 0.35
5 B +3 0.10 0.20
6 C -4 2.60 (P)
7 N +4 0.16 (A)
+3 0.16 (A)
8 +5 0.13 (A) 1.34
O -2 1.27 1.28 1.30 1.32
9 F -1 1.21 1.22 1.23 1.25 1.21 1.24 1.40
11 Na +1 1.07 1.08 1.10
12 Mg +2 0.66 0.75 0.80 0.97
13 Al +3 0.47 0.56 0.61
14 Si +4 0.34 0.48
15 P +3 0.44 (A)
16 +5 0.25 0.35 (A) 1.78
S -2 1.56 1.72
17 +6 0.20 0.30 (A) 1.65
Cl -1 1.67 1.72
+5 0.20 0.34 (A)
19 K +7 0.28 0.27 (A) 1.54 1.59 1.63 1.67 1.68
+1 1.46
20 Ca +2 1.08 1.15 1.20 1.26 1.36 1.43
21 Sc +3 0.83 0.95
22 Ti +2 0.94
+3 0.75
23 V -1-4 0.61 0.69
+2 0.87
+3 0.72
+4 0.67
24 Cr +5 0.44 0.54 0.62
+2 0.81 (L)
0.90 (H)
+3 0.70
+4 0.52 0.63
+5 0.43 0.52 (A)
25 +6 0.38 1.01
Mn +2 0.75 (L)
0.92 (H)
+3 0.66 0.66 (L)
0.73 (H)
+4 0.62
+6 0.35
+7 0.34 0.46 (A)
Table 7 .3 (continued)

Coordination numbers
z Symbol Valence // III IVs V
vr-h VII VIII IX
X
26 Fe +2 0.71 (H) 0.69 (L)
+3 0.57 (H) 0.86 (H)
0.63 (L)
27 Co +2 0.65 (H) 0.73 (H)
0.73 (L)
+3 0.83 (H)
0.61 (L)
28 Ni +2 0.69 (H)
0.77
+3 0.64 (L)
29 Cu +1 0.54 0.68 (H)

+2 0.70 0.73 0.81

30 Zn +2 0.68 0.76 0.83 0.98
31 Ga +3 0.55 0.63 0.70
32 Ge +4 0.48 0.62
33 As +5 0.42 0.58
34 Se -2 1.88 1.90
+6 0.37 0.42 (A)
35 Br -1 1.88 1.95 (P) 1.84
+7 0.34 1.39 (A)
37 Rb +1 1.57 1.64 1.68 1.74
38 Sr +2 1.21 1.29 1.33 1.40
39 Y +3 0.98 1.10 1.18
40 Zr +4 0.80 0.86 0.92
41 Nb +2 0.79
+3 0.78
+4 0.77
42 +5 0.40 0.72 0.74
Mo +3 0.75
+4 0.73
+5 0.71 0.79
+6 0.50 0.58 0.68
43 Tc +4 0.72
44 Ru +3 0.76
+4 0.70
45 Rh +3 0.75
+4 0.71
46 Pd +1 0.67
+2 0.72 0.94
+3 0.84
+4 0.70
47 Ag +1 0.75 1.10 1.20 1.23 1.32 1.38
+3 0.73
48 Cd +2 0.88 0.95 1.03 1.08 1.15
49 In +3 0.88 1.00
50 Sn +2 0.93 (A) 1.30
+4 0.77
Table 7 .3 (co ntinued)
Coordination numbers
z Symbol Valence II III IV* V vpb VII VIII IX X XII

51 Sb +3 0.85 0.88 0.76 (A)

+5 0.69
52 Te +4 0.60 0.70 (A)
+6 0.56 (A) 1.97
53 I -1 2.13
+5 1.03 1.82 1.86 1.89 1.96
55 Cs +1 1.78
56 Ba +2 1.44 1.47 1.50 1.55 1.60 1.68
57 La +3 1.13 1.18 1.26 1.28 1.36 1.40
58 Ce +3 1.09 1.22 1.23 1.37
+4 0.88 1.05
59 Pr +3 1.08 1.22
+4 0.86 1.07 1.17
60 Nd +3 1.06 1.20
61 Pm +3 1.04 1.17
62 Sm +3 1.04
63 Eu +2 1.25 1.33
E3 1.03 1.11 1.15
64 Gd +3 1.02 1.12 1.14
65 Tb +3 1.00 1.10 1.12
+4 0.84 0.96
66 Dy +3 0.99 1.11
67 Ho +3 0.98 1.10
68 Er +3 0.97 1.08
69 Tm +3 0.96 1.07
70 Yb +3 0.95 1.06
71 Lu +3 0.94 1.05
72 Hf +4 0.79 0.91
73 Ta +3 0.75
+4 0.74 0.77
+5 0.72
74 W +4 0.73
+6 0.50 0.68
75 Re +4 0.71
+5 0.60
+6 0.60
+7 0.48 0.65
76 Os +4 0.71
77 Ir +3 0.81
+4 0.71
78 Pt +2 0.68 0.80 (A)
+4 0.71
79 Au +1 1.37 (A)
+3 0.78 0.85 (A)
80 Hg +1 1.05 1.10 1.22
+2 0.77 1.04 1.84
81 TI +1 1.58 1.68
+3 0.97 1.08
 7.5 IONIC RADII 95
Table 7 .3 (continued)

Coordination numbers

z Sym bol Valence II III IV3 V V7a,b VII VIII IX X

XII
82 Pb +2C 1.02 1.26 1.37 1.41
1.57
+4 0.86 1.02
83 Bi +3 1.07 1.10 1.19
84 Po +4 1.16
+6 0.67 (A)
85 At +7 0.62 (A)
86 Fr +1 1.80 (A) 1.56
87 Ra +2 1.43 (A) 1.72
89 Ac +3 1.18 (A) 1.12 1.17
90 Th +4 1.08
91 Pa +3 1.13(A) 1.09
+4 0.98 (A) 1.03
+5 0.89 (A) 0.99
92 U +3 1.12 1.06 1.08 1.13
+4 0.97 (A)
+5 0.84 1.04
+6 0.53 0.56 0.81 0.96
93 Np +2 1.18
+3 1.10 1.06
+4 0.95 (A)
94 Pu +3 1.09 1.04
+4 0.88

Low spin and high spin are indicated by L and H, respectively. bA

= Ahrens (1952); P = Pauling (1927).
cPb2+, coordination number II, radius = 1.47 . S O
U R C E : Whittacker and Muntus (1970).
depending on the
coordination num ber on
the lattice site. Evidently,
ions are not actually hard
spheres, although they are
treated this way in som e
aspects of crystal
chemistry.
T he ionic radii listed
in Table 7.3 vary in reg
ular patterns depending
on the atom ic num ber of
the elem ent and the
charge of the ion as
shown in Figure 7.5,
provided all radii are
based on the sam e
coordination num ber. We
take note first of the
existence of isoelectronic
series con sisting of
sequences of ions in
which the charge
increases w ith increasing
atom ic num ber and
which th erefore contain
the same num ber of
electrons. Such
isoelectronic series exist
in all
but the first period . F or
exam ple, in the second
period we find:
L i+, B e2+, B 3+,
C4+, N 5+, 0 6+,
and F 7+
E ach of these ions has
only two electrons.
Similarly, the ions:
C4 , N 3
, O 2 ,
and F"
form an isoelectronic
series in which each ion
contains 10 electrons. O th
er isoelectronic series in
periods 3,4, and 5 are
indicated in Figure 7.5.
We will find it
helpful later to have a m
ental im age of the ionic
radii and to recall som e
gener
96 CHEM ICAL BO NDS, IONIC RADII, AND CRYSTALS

<
~o
nJ
Q

20 25

Atom ic Num ber

Figure 7.5 Variation of ionic radii with atomic number and valence. Note that ions having negative charges are larger
than those having positive charges. Note also that the radii of ions forming isoelectronic series decrease with increasing
atomic number (e.g., Li+ to F7+, Na+ to Cl7+, etc). The radii of ions in the same group that have the same charge increase
with atomic number (e.g., Li+ to Rb+, F to Br~, etc) [based on the ionic radii reported by Ahrens (1952) for sixfold
coordination].
alizations about the way 3. The radii of ions of
they vary within the p eri the same elem ent
odic table. We therefore decrease with
derive a num ber of con increasing positive
clusions about the ionic charge and
radii of the elem ents by increase with
exam ining Figure 7.5. increasing negative
charge.
1. The ionic radii of
isoelectronic
4. The radii of ions with charges of
series decrease
+3 am ong the transition m etals
with increasing
fourth period decrease
atom ic num ber
atom ic num ber, thereby
for both
traction of the
positively and
electron cloud as
negatively
the 3d orbitals are
charged ions.
filled.
2. The radii of ions 5. Ions of different
with the sam e elem ents may
charge in a group
have simi lar
increase with
increasing atom ionic radii.
ic num ber, th at
is, dow nw ard in
the periodic
table.
 PROBLEMS 97
Exam ples to illustrate the last observation can be 7.6 Summary
found in Figure 7.5 or by consulting Table 7.3.
The chemical elem ents can be divided into electron
Na", Ca2+, Cd2+, Y3+, ~ 1.00 donors and electron acceptors, depending on
Tl3+, R EE(+3),Th4+ (sixfold coordination) w hether their valence orbitals are nearly em pty or
H f4+,Zr4+ ~ 0.80 A nearly full. The tendencies of elem ents to be elec
(sixfold coordination) tron donors, and therefore to act like metals, is
reflected by their first ionization potentials and
Fe3+, Co3+, Ni3+ ~ 0.65 A
electronegativities. Elem ents with low ionization
(sixfold coordination)
potentials and low electronegativities have strongly
developed m etallic character. In general, the m etal
The sim ilarity of the ionic radii of different ele lic character of the elem ents decreases from left to
m ents is an im portant phenom enon because it right across the periodic table and increases dow n
m ay perm it the ion of one elem ent to substitute ward in those groups in which the elem ents have
for an ion of another elem ent in ionic crystals valence electrons in s- and p-orbitals.
(G oldschm idt et al., 1926). Chemical bonds can be classified into ionic or
The radii of ions in aqueous solution may be covalent bonds, depending on whether the valence
significantly changed by hydration. In this connec electrons are effectively transferred or shared,
tion it is helpful to define the ionic potential as the respectively. In reality, all ionic bonds have varying
ratio of the charge of an ion in electronic charges degrees of covalent character as indicated by the
to its radius m easured in angstrom units. The ionic magnitude of their electronegativity difference.
potential is a m easure of the density of the electri Polar solvents such as w ater are able to break ionic
cal charge on the surface of an ion and reflects the bonds because of electrostatic interactions betw een
strength of the electrostatic bond the ion can form ions in the solid phase and charges on the w ater
with another ion of the opposite charge or with a molecules. The hydration of cations and anions fur
polar molecule like water. Ions with high ionic ther enhances the solubility of ionic solids in water.
potential attract m ore w ater molecules and form The radius ratio of a cation with respect to
hydrated ions that are larger than those of ions the surrounding anions determ ines its coordina
having lower ionic potentials. For example, the tion num ber and hence affects the three -dim en
radii of the hydrated ions of the alkali m etals sional patterns the ions form in ionic crystals. The
decrease from Li+ to Cs+ even though their ionic ionic radii of the elem ents in the periodic table
radii in crystals increase with atomic num ber vary predictably, depending on the atom ic num
(Table 7.3). This can be explained by considering bers of the elem ents, the electrical charges of the
the ionic potentials of the alkali metals: Li+ = 1.2. ions, and on the coordination num ber.
N a+ = 0.91. K + = 0.68. R b + = 0.64. and Cs+ = T he m etallic ch aracter of the elem ents, the
0.56. Therefore, even though L i+ (radius = 0.82 ) kinds of bonds they form with each other, and
is much smaller than Cs+ (radius = 1.78 ), it the radii of ions are all related to their position
forms a larger hydrated ion than Cs+ because of its in th e periodic table and thus to their electronic
higher ionic potential. configurations.

Problems
 1. Determine the percent
ionic character of the fol-
lowing bonds, based on the
electronegativity differ-
ences: (a) NaCl; (b) FeCl2;
(c) CuCl; (d) A1C13; (e)
CC14.
2. Look up the ionization
potentials of silicon in a
recent edition of the CRC
Handbook o f Chemistry
and Physics (Weast et al.,
1986) and plot them in
coordi-
98 CHEM ICAL BONDS, IONIC RADII, AND CRYSTALS
nates of electron volts (eV) versus the number of elec trons
removed.
3. Explain why the third and fifth ionization poten tials
of silicon are significantly higher than those that
immediately precede them.
4. Explain why the first ionization potential of sulfur
(Z = 16) is less than that of phosphorus (Z = 15).
5. Write an equation to represent the dissociation of
(NH4)2S 0 4 into ions in an aqueous solution.
6. Calculate the volume of an ion of Na+ in cubic
meters (assume sixfold coordination).
7. Calculate the radius ratio for a planar structure
with fourfold coordination.
8. Determine the ionic potentials of the ions in the
isoelectronic series K+ to Mn7+ and derive a conclu sion
about the size of the hydrated ions in the iso electronic
series (use sixfold and fourfold coordination depending on
the availability of data in Table 7.3).
References
A H R E N S , L . H., 1952. The use of
ionization potentials Part 1.
Ionic radii of the elements.
Geochim. Cosmochim.
Acta,
2:155.
C O T T O N , F. A., and G. W IL K IN S O
N , 1962. Advanced Inorganic
Chemistry. Wiley, New
York, 959 pp.
F y f e , W. S., 1964.
Geochemistry o f Solids.
McGraw-Hill, New
York, 199 pp.
G O L D S C H M ID T , V. M., T.
B AR T H , G. L U N D E ,
and W.
Z A C H A R I A S E N , 1926.
Geochemische
Verteilungsgesetze der
Elemente. VII. Skrifter
Norske Videnskaps.-Akad.
Oslo, I. Mat.-Naturv.
Klasse, No. 2.
G R E E N W O O D , N. N., and A. E A R
N S H A W , 1984. Chemistry o f the
Elements. Pergamon Press,
Oxford, England, 1542 pp.
H E Y D E M A N N , A., 1969. Tables.
H andbook o f Geochemistry,
Verlag, Berlin, 442 pp.
P A U L IN G , L., 1927. The sizes of
ions and the structure o f
ionic crystals. J. Amer.
Chem. Soc., 49:763.
PA U L IN G , L., 1960. Nature o f
the Chemical Bond, 3rd ed.
Cornell University Press,
Ithaca, NY.
S A R G E N T -W E L C H SC IE N T IF IC C
O ., 1980. Periodic Table o f the
Elements. 7300 Linder
Ave., Skokie, IL 60076.
SH A N N O N , R. D., and C.T. P R E W
IT T , 1969. Effective crystal
radii in oxides and
fluorides. A eta
Crystallogr., Sect. B,
25:925-946.
SH A N N O N , R. D., 1976. Revised
effective ionic radii and
sys tematic studies of
interatomic distances in
halides and chalcogenides.
Acta Crystallogr. Sect. A,
32:751-767.
W E A S T , R. C., M. J. A S T L E ,
and W. H. B E Y E R , 1986. CRC
Handbook o f Chemistry
and Physics. CRC Press,
Boca Raton, FL.
W H IT T A C K E R , E. J. W., and R.
M U N T U S , 1970. Ionic radii
for use in geochemistry.
Geochim. Cosmochim.
Acta, 34:945-956.
 8
Ionic Substitutions in
Crystals
O ne of the principal objectives of geochemistry has been
to explain the distribution of chemical elem ents in the E
arth (Chapter 1). Before this objective could be achieved,
it was first necessary to determ ine how the elements are
actually dis tributed by analyzing large numbers of rocks
and minerals from all over the Earth. F. W. Clarke took a
giant step in that direction by his own efforts as an
analytical chemist and by presenting the data of
geochemistry in his book, first published in 1908. In
subsequent years, geochemists continued to ana lyze
geological materials in order to determ ine the chemical
composition of the E arth and to under stand the reasons
for the observed variation in chemical composition of
different kinds of rocks.
This was by no m eans an idle exercise
because the inform ation contributed directly to
the understanding of the origin of m etallic and
nonm etallic ore deposits. The search for ore
deposits of all kinds continues to be one of the principal
tasks of geologists in society. In a very real sense,
geologists are still expected to know where to find the
good rocks that are the foun dation of our m odern
industrial society.
8.1 Goldschmidt's Rules of
Substitution
A n im portant objective of G oldschm idts w ork in
geochemistry was to discover the laws of dis tribution of
the chemical elem ents (Goldschm idt, 1937). H e and
Pauling (1927) saw perhaps m ore clearly than their
contem poraries that the internal
structure of crystals could be understood in term s of the
size and charge of the ions and that an ion of one elem ent
can replace the ion of another if it is similar in size and
charge. These insights are sum m arized by a set of
generalizations know n as
G oldschm idts rules o f substitution.
1. The ions of one elem ent can extensively replace
those of another in ionic crystals if their radii
differ by less than about 15% .
2. Ions w hose charges differ by one unit sub stitute
readily for one another provided electrical
neutrality of the crystal is m ain tained. If the
charges of the ions differ by m ore than one unit,
substitution is gener ally slight.
3. W hen two different ions can occupy a p ar ticular
position in a crystal lattice, the ion with the
higher ionic potential form s a
stronger bond with the anions surround ing the
site.
4. Substitution m ay be limited, even w hen the size
and charge criteria are satisfied, w hen the com
peting ions have different electronegatives and
form bonds of dif ferent ionic character.
The fourth rule was actually form ulated by Ringwood
(1955) in order to explain discrepancies that arose from
the three rules proposed by Goldschmidt. For example, N
a+ and Cu+ have the same charge and their radii are
virtually identical (A hrens, 1952). H ence, according to G
oldschm idts rules, C u+ should replace N a+ in sodium
minerals
99
100 IONIC SUBSTITUTIONS IN CRYSTALS
such as albite (NaAlSi3 0 8) or halite (NaCl). We know, of
course, that this substitution does not occur. The reason is
that copper forms m ore cova lent bonds than sodium, as
indicated by their elec tronegatives.
The occurrence of elem ents in minerals and rocks
either as major or m inor constituents depends on their
abundances and chemical properties. In general, the most
abundant elem ents form the m in eral com pounds within
which the m inor elem ents may be accom m odated by
ionic substitution, in interstitial lattice positions, in fluid
inclusions, or as exsolved m ineral phases. Some m inor
elem ents also occur in accessory minerals such as zircon
or apatite.
The tendencies of elem ents to form m inerals are m
ost clearly displayed during the crystallization of a cooling
magma. Therefore, G oldschm idts rules prim arily apply
during this process, which was thought to so rt the ions
according to their size and charge. Those ions th at do not
fit into the m ajor rock-form ing m inerals are said to be in
c o m pa tib le and therefore accum ulate in the residual m
agm a. Consequently, the incom patible elem ents are
concentrated in late-stage differentiates of magmas,
including aplite dikes, pegm atites, and hydrotherm al
veins. The elem ents in this category include K +, R b +, C
s+, Sr2+, B a2+, the rare earth elem ents (R E E ), Z r4+, H
f 4 +,N b 5 +,Ta5 +,T h4+, and U 4+. A ll of these elem ents
ultim ately associate them selves with rocks com posed of
silicate m iner als and, for that reason, are also referred to
as the large ion lithophile (LIL) group.
G oldschm idts rules w ere critically review ed by m
any geochem ists, including Shaw (1953) and Burns and
Fyfe (1967). The rules are at best a first
approxim ation to which there are m any excep
tions th at arise because the replacem ent of one
ion by another is generally site-specific, especially in com
plex com pounds. The problem is m agnified because the
ionic radii, used as a criterion in G oldschm idts rules,
depend on the site w here the replacem ent takes place. In
spite of these lim ita tions, G oldschm idts rules still serve
a useful p u r pose by identifying som e of the m ajor
factors that m ust be considered in the possible replacem
ent of ions in crystals form ing in a cooling m agm a or
from an aqueous solution.
8.2 Camouflage, Capture, and
Admission
A ccording to G oldschm idts first rule, ions that have sim
ilar radii and equal charges replace each o th er extensively
in ionic crystals. The extent of substitution that actually
takes place depends on
th e concentration of the ions in the m edium in
which th e crystals are form ing, on the tem pera
ture, and on the com patibility of their bonds and
coordination num bers. Crystals form ing at high
tem p eratu re are m ore tolerant of foreign ions
than crystals form ing at a low tem perature.
T herefore, the concentration of trace elem ents in crystals
can be used to estim ate the tem perature of form ation of
certain m inerals. For example, th e concentration of Fe2+
in sphalerite (Fe, ZnS) increases with the tem perature
provided that enough iron was available to saturate the
spha lerite (K ullerud, 1959).
The second rule applies to substitution of ions of sim
ilar size but having different charges. W hen th e charge
difference is greater than one, substitu
tion is limited because of the difficulty in m ain
taining electrical neutrality. C harge deficiencies
th at result from substitution of ions of unequal
charge m ust be com pensated by a second substitu tion
involving an ion having a different charge. This process of
coupled substitution contributes to the diversification of
chemical com positions of m any minerals. A n alternative
to coupled substitu tion displayed by the clay m inerals is
adsorption of ions on the charged surfaces of small
crystals.
The third rule describes the effect of the ionic
potential on the relative strengths of ionic bonds of ions
com peting for the sam e site. Ions having a higher ionic
potential (charge/radius ratio) form a stronger bond than
their com petitors and are therefore preferentially
incorporated into the crystal. Similarly, com peting ions
with lower ionic potentials are discrim inated against and
are ini tially excluded from crystals forming from a m agm
a. Consequently, ions with high ionic poten tials are
concentrated in early-form ed crystals in a cooling magma,
w hereas those with low ionic potentials are concentrated
in the residual magma and enter late-form ing crystals.
8.2 CAM O UFLAGE, CAPTURE, AND ADMISSION 101

A ccording to the
fourth rule, a m inor ion
can replace a m ajor ion
only if their
electronegativi ties are
similar. W hen two m inor
ions th at are sim ilar in
size and charge com pete
for the sam e site, the one
w hose electronegativity is
m ore sim ilar to that of
the m ajor ion is preferred
because the bonds are m
ore com patible.
We see th at G
oldschm idts rules of
substitu tion provide
criteria for predicting the
extent to which m inor
elem ents can replace m
ajor ele m ents in the m
inerals they form . A s a
result, we can predict how
m inor elem ents distribute
th em selves w hen m
inerals crystallize from a
cooling m agm a or from a
supersaturated aqueous
solu tion. Som e elem
ents are selectively
concentrated into the solid
phase, whereas others rem
ain in the liquid phase.
The different ways in
which the ions
of m inor elem ents are p a rtitioned
solid and liquid phases are described
term s th at form the heading of rep
Camouflage occurs
when the minor
element has the
same charge and a
similar ionic radius
as the major
element it is
replacing.
In this case, the m inor
elem ent does not form its
own m inerals but is hidden in the
another elem ent. Cam ouflage is displayed
Z r4+ (0.80 ) and H f4
+(0.79 A ) because
hafnium rarely form s its
own minerals and is
always present in the m
ineral zircon (Z rS i0 4).
We could say that zircon
crystals do not distinguish
between Z r4+ and H f4+
ions and accept both with
equal ease.
Capture takes place
when a minor
element enters a
crystal
preferentially
because it has a
higher ionic
potential than the
ions of the major
element.
E xam ples of capture
occur in the form ation of
feldspar crystals, which m which ions are adm itted
ay capture B a2+ (1.44
) or Sr2+ (1.21 ) in
place of K + (1.46 A ). A s
a result, the
concentrations of these
elem ents in the residual
m agm a decrease during
th e crys tallization of K
by
-feldspar. H ow ever, the
this section.
lace m ent of th e
univalent K + ion by a
divalent ion requires a
coupled substitution of A
l3+ for Si4+
the relatio n sh ip betw een
in o rd er to preserve the th e ionic potentials of
electrical neutrality of the cations and the stren g th
crystal crystals
lattice. of the bonds they form w
ith anions. This
Admission involves relationship m anifests
the entry of a itself in som e cases in th e
foreign ion that has m elting tem p eratu res o f
a lower ionic the com pounds they
potential than the form . For exam ple, the M
major ion because it g sili
has either a lower cate fo rsterite (M g2 S i0 4) has a h
charge or a larger tem p eratu re (1910 C) th an th e F
radius, or both. alite (Fe2 S i0 4), w hich m elts at 1503
m inerals form a solid solution k
m ineral olivine ((M g, F e ) 2 S i0 4)
The occurrence of R b + in lizes from cooling m agm
K -feldspar and other as of basaltic com posi
tion . E arly form ed olivine is enrich
potassium m inerals is an
fo rsterite (M g) end m em ber, w her
exam ple of admission
form ing at low er tem p
because R b + (1.57 )
eratu res is enriched in the
has a sm aller ionic
fayalite (Fe) end m em
potential than K + (1.46
ber. A possible
). O ther exam ples are
explanation is that early
the replace m ent of Ca2+
form ed olivine captures
(1.08 ) by Sr2+ (1.21 A )
Mg2+ (0.67 A)
in calcite, and the
in favor of Fe2+ (0.74 A ) because
substitution of Cl- (1.72
ionic p o ten tial (A hrens, 1952).
A ) by B r (1.88 ) in
feldspar an o rth ite (C aA l2 Si2 O g)
chlorides. T he extent to
m elting tem p eratu re than albite (
presum ably because C
into a particular lattice site
a2+ (1.08 A ) has a higher
decreases as the differ
ionic p o ten tial th an N a + (1.10 A
ence in the radii of com
1952). A s a result, early -form ed p
peting ions increases. For
example, I (2.13 )
replaces C1 (1.72 A )
much less than B r (1.88
A ) and B a2+ (1.44 A ) is
less abundant in calcite
than Sr2+ (1.21 A).
Evidently, adm ission of
foreign ions into the
crystal of a m ajor elem
ent is ultim ately
controlled by the size
criteri on expressed in G
oldschm idts first rule.
H ie third ru le m
akes a statem en t about
102 IONIC SUBSTITUTIONS IN CRYSTALS

Table 8.1 C lassificatio n o f S ilicate M inerals B ased o n T h e ir S tru c tu re s as Illu stra te d in

F ig u re 8.1

Shared Silicate S i/O Examples

Class corners anion
Neso 0 sioy 1 :4 fayalite Fe2S i04

Soro 1 si2o r 2 :7 akermanite Ca2MgSi20 7

Cyclo 2 S iO t 1:3 benitoite BaTiSi3OCJ
Ino single 2 sior 1:3 enstatite MgSiO,
Ino double 2j si4o?7 4:11 anthophyllite Mg7(Si40 11)2(0H ):
Phyllo 3 s.2o r 2 :5 kaolinite A l2Si20 5(0 H )4
Tekto 4 S i02 1 :2 quartz S i0 2

initially enriched in
calcium relative to sodium
. The ionic potential by
itself, how ever, is n o t a
reliable p redictor of the m
elting tem peratures of
solid com pounds and the
exam ples presented above
are m ore the exception
than the rule (Burns and
Fyfe, 1967).
8.3 Coupled
Substituti
on: Key
to the
Feldspars
W hen the ion of a m ajor
elem ent is replaced by a
foreign ion having a
different charge, the
electri cal neutrality of the
crystal m ust be preserved
by a com plem entary
substitution elsewhere in
the crystal lattice. This
phenom enon helps to
explain the chemical
diversity of silicate m
inerals such as the
feldspars, zeolites, and
micas.
Silicate m inerals
contain a fram ew ork of
sili ca tetrahedra linked
together to form different
kinds of structures. The
silicate anion (S iO ^ )
form s a tetrahedron in
which the Si atom is locat
ed in the center and is
bonded to the four O
atoms that occupy the
corners of the
tetrahedron . The Si O
bonds are about 50%
covalent based on the
difference in their
electronegativities. Each
O atom at the corner of a
tetrahedron has an
unpaired electron in a
hydridized p -orbital that
can form ionic bonds with
cations in orthosili cates,
such as M g2 S i0 4, Fe2 S
i0 4, or Li4 S i0 4. The
silica tetrah ed ra can also
form bonds with each
other by sharing O atom s
at the corners to form
discrete form because the
lattice extends contin
uously in all directions. H
rings, chains, double owever, the form ula of
chains, sheets, and three- the alum inosilicate anion
dim ensional netw orks
allows us to recognize
listed in Table 8.1 and
that the replacem ent of
illus trated in Figure 8.1.
one out of every four Si
The feldspars are
atom s by A l3+ causes a
tektosilicates charge im balance of - 1
composed of a three-
dim ensional netw ork
of silica tetrahe
dra linked at their corners. In this
the two unpaired electrons of O
bonds with different Si
atoms and thereby link
the tetrahedra together. W
hen all of the O atoms
are shared in this m anner,
ow ns only half of each of
located at the corners of)
its tetrahedron. Therefore,
in a three-dim ensional
netw ork the ratio of Si to
O is reduced from 1 : 4 to
1 : 2 and its form ula is S
i0 2.
This fram ew ork of
linked silica tetrahedra can
adm it a lim ited num ber
of A l3+ ions in place of
Si4+. The substitution is
restricted because in
fourfold coordination A
l3+ (0.47 A ) is signifi
cantly larger than Si4+
(0.34 ) and actually vio
lates G oldschm idts first
rule. O n the other hand,
the radius ratio of A l3+ to
O 2- is 0.36, which allows
it to have a coordination
num ber of 4 like
Si4+, whose radius ratio
Figure 7.4). As a result, one in
replaced to form the
AlSijOg. This anion does
not actually exist in
8.4 DISTRIBUTION COEFFICIENTS AND GEOTHERMOMETERS 103
 in the lattice. This excess negative charge is n eu tralized
by th e introduction of N a + or K +. The resulting com
pounds are know n to us as the alkali feldspars:

A"IX K A lSi3 0 8 N aA lSi3 0 8

orthoclase, microcline albite
W hen two out of four Si atom s are replaced

by A l3+, the alum inosilicate anion has a

charge of - 2 , which is neutralized by the
addition of Ca2+ or B a2+ to form:
C aA l2 Si2 O s B aA l2 Si2 O s

anorthite celsian

A lbite and anorthite form a solid solution known as

plagioclase, which is subdivided into six m iner al species,
depending on the m olar concentrations of albite (A b) or
anorthite (An).

Mineral Ab, % An, %

Albite 100-90 0-10

Oligoclase 90-70 10-30

Andesine 70-50 30-50
Labradorite 50-30 50-70
Bytownite 30-10 70-90
Anorthite 10-0 90-100

The series is divided into 20% increm ents of A b or A n,

except at the ends, w here the increm ents
are 10% . The nam es of the plagioclase series
can be rem em bered by noting that the first let
ters spell the nam e of the fictitious geochem ist
A . O. A lba.

8.4 Distribution Coefficients and

Geothermometers
Figure 8.1 Two-
G oldschm idts rules
dimensional networks of
silica tetrahe-dra provide a rational basis
connected in different for predicting how m
ways by sharing oxygen inor elem ents m ay
atoms at their corners, (a)
Nesosilicates, (b) enter crystals form ing
sorosilicates, (c) from a cooling m elt or
cyclosilicates, (d) from an aqueous
inosilicates (single chain),
(e) inosili-dates (double solution. T hese effects
chain), and (f) are described
phyllosilicates. The qualitatively
tektosili-cate structure,
consisting of a three-
as cam ouflage, capture,
dimensional network of How ever, words alone are
tetrahedra in which all describe the partitioning of
four oxygen atoms are
shared, is not shown. More
details are given in Table
8.1.
104 IONIC SUBSTITUTIONS IN CRYSTALS

betw een crystals and the liquid from which they E quation 8.3 indicates that the ratio of the con
form ed. centrations of the trace elem ent y in m inerals A
We therefore define the distribution coeffi and B is equal to a constant K, which is equal to
cient (D ): the ratio of the distribution coefficients of y in
Cx those m inerals at the tem perature of form ation.
(8.1) In principle, the tem perature dependence of
D =c'
the distribution coefficients of a trace elem ent in
w here C* is the concentration of a m inor elem ent different m inerals can be determ ined experim en
in the crystal (x) of a m ineral and C 1 is the con tally. The tem perature of form ation of rock sam
centration of that elem ent in the liquid (1 ) from ples can then be determ ined from the ratio of the
which the crystal form ed under equilibrium con m easured concentrations of the selected trace
ditions. The distribution coefficient of a particular elem ent in these two m inerals (M cIntyre, 1963).
elem ent in a specific m ineral m ay be > 1 , < 1 , or Consider, for exam ple, the replacem ent of
= 1. T he m agnitude of D is related to the verbal Z n2+ (0.68 , fourfold coordination) in spha
descriptors as follows: lerite (ZnS) and Pb2+ (1.26 , sixfold coordina
D> 1 capture tion) in galena by Cd2+ (0.88 , fourfold
coordination, 1.03 A, sixfold coordination). The
1
D< adm ission electronegativities of the three elem ents are simi
D= 1 cam ouflage lar and the bonds they form with sulfur are about
85 % covalent. A n experim ental study by B ethke
The num erical values of D m ust be determ ined and B arton (1971) indicated th at Cd is strongly
experim entally in the laboratory or by analysis of partitioned into sphalerite (sp) coexisting with
natural systems in which both crystals and liquid galena (gn) betw een 600 and 800 C and th at the
can be sampled. tem perature dependence of the distribution coef
T he observed distribution of trace elem ents ficients is given by:
in crystals can be used to estim ate th e tem p era (2080 - 0.0264P)
ture of form ation of coexisting m inerals. R ecall lo g tf(C d ) = - 1.08 (8.4)
th a t the extent of substitution is tem p eratu re
d ep en d en t because crystals becom e m ore to le r where:
ant of foreign ions as the tem p eratu re increases. r vn
T herefore, distribution coefficients are, in g en er K (C d) =
al, tem p eratu re d ep en d en t and are also affected - f 2 for Cd
by the com positions of th e liquid and th e crys C= concentration of CdS in weight percent
tals, as well as by pressure. If two m inerals A and
T=
B coprecipitate or crystallize from the sam e absolute tem perature in kelvins
solution or m agm a, a m inor elem en t (y) can P = pressure in atm ospheres
e n te r b oth m inerals A and B at a p articu lar te m
The relationship betw een P'(C d) and the tem
peratu re . The distribution coefficients of ele
m ent y are: p eratu re is found to be highly nonlinear w hen
cx equation 8.4 is plotted in linear coordinates in
and B= ( ^) ( 8 .2 ) Figure 8.2. H ow ever, in coordinates of log K (y-
C1/ coordinate) and 1 /T (x -coordinate) equation 8.4
W/A is a straight line in the slope-intercept form:
T he ratio of the distribution coefficients for ele
m ent y is: y=mx+b (8.5)

(COnStant) The slope m = 2080 - 0.0264P and the intercept

on the y-axis is b = -1 .08 . The pressure can be dis-

^= =
K
8. 5 GEOCHEM ICAL CLASSIFICATION O F THE ELEMENTS 105

Te
mp
era
tur
e,
C serious lim itations
Figure 8.2
Geothermometer based on
the distribution of CdS
between sphalerite and
galena (equation 8.4). The
constant K is the ratio of
the concentrations of CdS
in sphalerite and galena in
weight percent. The curve
implies that the Cd content
of sphalerite decreases
rela tive to that of galena
with increasing
temperature of formation
of these minerals. The solid
part of the line is based on
the experimental data of
Bethke and Barton (1971).
The dashed line is an
extrapolation of those
data.
regarded for low values of
m om eters studied by Bethke
are based on the
distributions of M n (600-
800 C)
and Se (390-595 C) betw een sphalerite, galena,
and chalcopyrite. The results indicate than M n2+,
like Cd2+, is
preferentially
incorporated into spha
lerite rather than galena, 8.5 Geoche
whereas the order of
preference for Se2- is mical
galena, chalcopyrite, and Classific
sphalerite, listed in order
of decreasing preference. ation of
The applicability of the
these trace elem ent geo
therm om eters is lim ited Element
by the requirem ent that s of form ation of the E arth
the m inerals m ust have
The geochem ical
form ed at the same tem
classification of the elem
perature from the same
ents is based on the way in
solution under equilibri
which the elem ents
um conditions and that
actually distribute them
interfering effects caused
selves betw een different
by the presence of other
kinds of liquids and a gas
ions or pressure varia
phase. D uring the sm
tions w ere either negligible
elting of oxide and sulfide
because
ores th ree different liquids
precipitate sequentially
are encountered th at are
ously, which m eans that
immiscible in each other
not necessarily form
and th at segregate into
from the sam e solution.
layers depending on their
density. These liquids are
com posed of m olten Fe,
m olten sulfides (m atte),
and m olten silicates (slag).
G oldschm idt and his
contem poraries con
sidered it likely th at th e E
arth was initially com
pletely m olten and th at
these liquids separated
from each other under the
influence of gravity to
form an Fe core, a sulfide
layer, and a silicate layer in
the interior of the E arth .
The gases form ed the atm
osphere, which
subsequently produced the
hydrosphere by
condensation of w ater
vapor.
The heat required for
m elting was provided by
the im pacts of the
planetesim als, by com
pression caused by the
gravitational contraction of
the E arth, by the m
igration of dense phases
tow ard the center of the E
arth, and by the decay of
the radioac tive isotopes of
U, Th, and K, which were
m ore abundant at the tim e
4.5 x 109 years ago than
they are today and th ere
fore generated m ore heat
than they do at present. A
lthough the E arth does not
actually contain a sul fide
shell, G oldschm idts
geochemical classification
still conveys useful inform
ation about the tenden cies
of the elem ents to enter
liquids of different com
position or to be
concentrated in the gas
phase.
The inform ation on
which this classification is
based cam e from the
study of m eteorites and
from
the sm elting of the K upferschiefer ore
slate) at M ansfeld in Germ any. M
fragm ents of larger parent bodies tha
betw een the orbits of M ars and Jupiter
1962) (see Table 3.2).
Som e of the parent
bodies were large
enough to retain
sufficient heat for m
elting and differentiated
into m etallic Fe, silicate
rocks, and sulfide m
inerals. The parent
bodies w ere
subsequently broken up
by tides caused by the
gravitational fields of
Jupiter and M ars and
by collisions am ong
them selves. The rem
nants of the
106 IONIC SUBSTITUTIONS IN CRYSTALS
parent bodies still occupy the space betw een M ars (group IIA ), the halogens (group V IIA ), as well as

and Jupiter. Fragments resulting from collisions B, Al, O, Si, and som e of the transition m etals such
am ong them continue to be deflected into E arth - as Sc, Ti, V, Cr, Mn, and som e of their congeners.
crossing orbits and impact on the E arth as m ete The noble gases, H, and N m ake up the a tm o p h ile
orites. Goldschmidt, as well as Ida and W alter group.
N oddack in Berlin, analyzed the m etallic, sulfide, T here are a few surprises in this classification.
and silicate phases of many m eteorites and d eter For example, O is lithophile rather than atmophile,
m ined from the results how the elem ents had been and C, as well as P, dissolves in m etallic Fe in the
partitioned into the three immiscible liquids dur absence of O. Thallium is a chalcophile elem ent
ing the geochemical differentiation of the parent although it com m only substitutes for K + in silicate
bodies of m eteorites (N oddack and N oddack, minerals. We note also that Au, Sn, and M o are
1930). The inform ation derived from the study of siderophile, and presum ably have been carried
m eteorites was consistent with the chem ical com away by metallic Fe to form the core of the E arth.
positions of the silicate slag, F e -C u sulfide m atte, The same is true for Co and Ni, which occur as sul
and m etallic Fe, all of which form during the fides in ore deposits but prefer liquid Fe when
sm elting of ore. given a choice. O n the other hand, Ti, Cr, and Mn,
The resulting classification of the elem ents in which are com m only associated with Fe in igneous
Table 8.2 contains four groups th at G oldschm idt rocks, are not siderophile but lithophile elements.
nam ed siderophile (iron liquid), chalcophile (sul
fide liquid), lithophile (silicate liquid), and 8.6 .Summary
a tm o p h ile (gas phase). N ote th at several elem ents
occur in m ore than one group and that such sec The ions of different elem ents can substitute for
ondary affinity is indicated by parentheses. In gen each other provided their radii, charges, and elec
eral, we see from Figure 8.3 th at the elem ents in tronegativities are similar. In addition, they must
group V IIIB are siderophile together with C, Mo, be com patible with the lattice site in term s of
Au, Ge, Sn, and P. The congeners of groups IB and their radius ratios and coordination numbers.
IIB are joined by the elem ents G a, In, TI, Pb, As, G oldschm idts rules provide a rational basis for
Sb, and Bi, as well as by S, Se, and Te, to form the understanding the distribution of trace elem ents
chalcophile group. The lithophile elem ents include but do n o t give enough em phasis to the character
the alkali m etals (group IA ), alkaline earths of the site of replacem ent in the crystal lattice.

Table 8.2 G e o c h e m ic al C la ssific a tio n o f th e E le m e n ts 3

Siderophile Chalcophile Lithophile Atmophile
Fe, Co, Ni Cu, Zn, Ga Li, Be, B, (C), , O, F H, He

Ru, Rh, Pd Ag, Cd, In Na, Mg, Al, Si, P, , Cl (C), N, (O), Ne
Os, Ir, Pt Hg, TI K, Ca, (Ga), (Ge), , , Br Ar, Kr, Rn
plus plus Rb, Sr, , I
Mo, Re, Au (Ge), As, S Cs, Ba, (TI), plus
(W), and (Sn), Sb, Se Sc, Ti, V, Cr, Mn, (Fe)
C, P, Ge Pb, Bi, Te Y, Zr, Nb
Sn, (As), (Pb) and (Cr), (Fe) La, Hf, Ta, W
(Mo) REE, Th, , U, and
(H)
"The elements in parentheses occur primarily in another class.
 8.6 SUMMARY 107

II----------------------------A G r o u p s -------------------------- III IV V VI VII VIII

Siderophile C h alco p h ile Lithophile

Atm ophile N onexistent

Figure 8.3 Geochemical classification of the elements in the periodic table.The classification is based on the way the
elements distribute themselves between an iron liquid (siderophile), a sulfide liquid (chalcophile), a silicate liquid
(lithophile), and a gas phase (atmophile).
 The extent to which
replacem ent of ions takes
place is described by the
terms camouflage, capture,
and admission. Trace elem
ents whose ions closely
resem ble those of m ajor
elements in terms of size,
charge, and
electronegativity are cam
ouflaged in the crystals
formed by the major
elements. Trace elem ents
that are captured during
the crystalliza tion of a
magma are concentrated
into early-formed crystals
and are depleted in the
residual liquid. Ions that
are adm itted into the
crystals of a m ajor ele m
ent are initially enriched in
the residual magma and
subsequently enter late-
forming crystals.
Substitution of A l3+
for Si4+ in silicate struc
tures causes a charge im
balance in the lattice that
is neutralized by the com
plem entary introduction
of certain cations that are
com patible with the par
ticular lattice sites. This
phenom enon of coupled
substitution accounts for
the chemical composi
BEL
SON
tion of the feldspars and of
108 IONIC
SUBSTITUTIONS IN
other aluminosilicate m
CRYSTALS from the chem ical com
ineral groups.
The distribution positions of m eteorites
coefficient is a quantitative m etallic Fe, transition m and from m etallurgical
m easure of ionic etal sulfides, and silicates. studies in th e smelting of
substitution. The num The geochem ical sulfide and oxide ores.
erical val ues of classification of the elem
distribution coefficients ents, which is based on
must be d ete r m ined this phenom enon, was
experim entally as a derived
function of tem perature,
chem ical compositions of
liquids and crystals, and of
pressure. The distribution
of trace elem ents betw
een two coexisting m
inerals th at form ed under
equilibrium conditions
from the sam e liquid can
be used to determ ine the
tem p eratu re of form
ation . Such geotherm om
eters are of great interest
in geochem istry but their
application is restricted by
the conditions that m ust
be im posed.
The geochem ical
properties of the elem ents
are reflected by th eir
distribution betw een a gas
phase and am ong natural
liquids com posed of
Problems
 1. Determine the coordination numbers of Ca2+ and 8. The melting temperatures of the fluorides of the
SF+ relative to O2-, and use the result to predict the alkali metals and alkaline earths are listed below. Consider
substitution of Ca2+ by Sr2' in calcite (coordination : 6) the observed variation in terms of both the ionic character
and aragonite (coordinate : 8). of the bonds and the ionic potentials. How should the
melting temperatures vary according to Goldschmidts
2. Predict whether Hg2+ can replace Sr2+ in the min rules? Which compounds conform to this prediction and
eral strontianite (SrC03). which do not?
3. Explain the formula of leucite (feldspathoid,
Melting
KAlSi20 6) in terms of coupled substitution.
Com pound temperature C
4. Examine the formula of the feldspathoid sodalite LiF 845
(Na8A l6Si60 24Cl2) and explain it on the basis of cou pled
substitution. NaF 993
KF 858
5. Silver (A g+) is a common trace element in galena
where it replaces Pb2+. Identify the ion of a chalcophile RbF 795
element that is best suited to enter galena with A g+ in a CsF 682
coupled substitution. BeF2 800 (sublimes)
6. Lithium (Li+) andMg2+ have similar radii and elec MgF2 1261
tronegativities, yet Li+ does not replace Mg2+ in olivine. CaF2 1423
Explain the reason for this occurrence and suggest anoth SrF2 1573
er host mineral for Li+ in which it does replace Mg2+. BaF2 1355
7. Zircon crystals (ZrSi04) commonly admit urani um
(U4+) but strongly exclude Pb. Deduce the valence of Pb in
magma from this observation.
References
A H R E N S , L. H ., 1952. The use
of ionization potentials Part 1.
Ionic
radii of the elements.
Geochim. Cosmochim.
Acta, 2:155.
B E T H K E , P. M., and P. B. B A R T
O N , J R ., 1971. Distribution
of some minor elements
between coexisting sulfide
minerals.
Econ. Geol., 66:140-163.
B U R N S , R. G., and W. S. F Y F E ,
1967.Trace element
distribution rules and their
significance. Chem. Geol.,
2:89-104.
G O L D S C H M ID T , V. M., 1937. The
principles of distribution
of chemical elements in
minerals and rocks. J.
Chem. Soc. London,
1937:655-673.
K U L L E R U D , G., 1959. Sulfide
systems as geological
thermome
ters. In P. H. A
(Ed.), Researches in
Geochemistry,
vol. 1,301-335. Wiley, New
York, 511 pp.
M a s o n , B., 1962. Meteorites.
Wiley, New York, 274 pp.
M C I N T Y R E , W. L., 1963. Trace
element partition coefficients
a review of theory and
applications to geology.
Geochim.
Cosmochim. Acta, 27:1209-
1264.
N O D D A C K , I., and W. N O D D A C K ,
1930. Die Hufigkeit der
chemischen Elemente.
Naturwissenschaften,
18:757-764.
P A U L IN G , L., 1927. The sizes of
ions and the structure of
ionic crystals, J. Amer.
Chem. Soc., 49:765-790.
R IN G W O O D , A . E ., 1955. T h e p
rin c ip le s g o v e rn in g tra c e
e le m e n t d is trib u tio n d u
rin g m a g m a tic cry stalliza tio
n.
Geochim. Cosmochim.
Acta, 7:189-202.
S H A W , D .M ., 1953 .T h e c am o u
fla g e p rin c ip le a n d tra c e e
le m e n t d is trib u tio n in m a g
m a tic m in erals . J. Geol.,
61:142 -151 .
 Ill
AQUEOUS GEOCHEMISTRY
AND THE STABILITY OF
MINERALS

Im portant chemical reactions occur at the surface of the E arth when m inerals are exposed to w
ater, carbon dioxide, and oxygen. The principles th at govern such reac tions can be used to explain the
interactions of ions in aqueous solutions and to con
strain the stabilities of solid com pounds with which the ions are in equilibrium . The
inform ation derivable from such studies is conveniently presented in graphical form
because E arth Scientists are accustom ed to using m aps and diagram s in their work.
 9
Acids and Bases
Chemical reactions in nature commonly take place in the
presence of water, which acts as the m edium within
which ions and molecules can interact. W hen two ions or
m olecules approach each other closely, they have an
opportunity to establish a bond betw een them . W hen
they do so, we say that a chemical reaction has taken
place and a product has form ed. Such chemical reac
tions are represented by m eans of algebraic equations
that express the balance of masses and charges of the
reactants and products. We use such equations to
understand chemical reactions in nature, and therefore
need to develop the skill of writing and interpreting
them . For this reason, we begin the third part of this
book with a very simple exercise.
9.1 Chemical Reactions and
Equilibria
In order to understand how chemical reactions
work we first consider what happens when sodi
um chloride is dissolved in pure w ater at room
tem perature and atm ospheric pressure. The
apparatus consists of a beaker containing contin uously
stirred water, into which is placed an ion-sensitive
electrode capable of m easuring the concentration of N a+
ions. The electrode is con nected to a strip-chart recorder
operating at a known chart speed. We ignore all practical
limi tations and com plications and concentrate our
attention entirely on the events that occur when
we place a quantity of solid NaCl into the
w ater in the beaker. The results of this experi
m ent are entirely predictable based on past
experience: the salt will dissolve in the water.
H owever, we deliberately added a larger quanti ty of salt
to the beaker than will dissolve in the w ater it contains.
As soon as the solid NaCl comes in contact
with the w ater in the beaker, it begins to dissolve
and the concentration of N a+ in the w ater
increases with time. A fter a few m inutes the N a +
concentration becom es constant regardless of
how long we continue the experim ent. A t this
tim e the beaker still contains som e solid NaCl
that did not dissolve in the water. In order to interpret the
results of this experim ent we turn to Figure 9.1, which is
the recorder trace of our experim ent.
We see that the N a+ concentration increased rapidly
at first but eventually leveled out and becam e constant.
W hen no m ore NaCl would dis solve in the w ater at the
tem perature and pressure of the experim ent, the solution
becam e saturated with respect to NaCl. W hen we
reached that con dition, the reaction achieved a state of
equilibrium in which the rate of solution of NaCl is equal
to the rate at which it precipitate s. Therefore, at equi
librium the am ount of excess solid NaCl and the
concentrations of N a+ and Cl- ions in the saturat ed
solution are invariant with respect to time. Once the
reaction reaches equilibrium, it stops being productive
because the forward and back ward reactions are
occurring simultaneously and at the same rates. We can
indicate this condition by means of equations. A t
equilibrium the forward reaction:
N aCl(s) ^ N a+ + Cl (9.1)
and the backw ard reaction:
N a + + CI N aCl(s) (9.2)
110
9.1 CHEMICAL REACTIONS AND EQUILIBRIA 111

Figure 9.1 Increase of the
concentration of Na+ with
time after the addition of
solid NaCl to pure water
at 25 C. This experiment
illustrates the point that
chemi cal reactions
progress toward a state of
equilibrium in which the
concentrations of
reactants and products are
invariant with time. In this
illustration the reaction
reaches equilibrium when
the solution becomes
saturat ed with respect to
NaCl. While reactions are
at disequi librium they
effectively convert
reactants into products
When reactions reach
equilibrium, the rate at
which products are
converted back to
reactants equals the for
ward reaction rate and, as
a result, the net change in
the concentrations of
reactants and products is
zero.
occur at the same rates.
We represent this state by
m eans of double arrows:
N a C l( s ) ^ N a + + Cl
That is, double arrows
mean that the reaction
expressed by an equation
is in a state of equilibrium
. and chemical conditions, it
O ur experim ent
illustrates the im portant
point that chemical
reactions have a natural
ten dency to progress tow
ard equilibrium . In order
to reach this condition th e
reactants are consum ed
and products are form ed
such th at the am ounts of
both change with time.
We can think of this dise
quilibrium condition as
the active m ode of a reac
tion. A fter a reaction has
reached equilibrium it
becom es inactive in the
sense th at the am ounts of
+ 2 HC1 - words,
reactants and products per
unit w eight or unit
volume of w ater becom e
constant. In fact, w hen a
chemical reaction at
equilibrium is disturbed
by
changes in the physical
responds by reestablishing
a new state of equilib
rium . This property of
chemical reactions was
expressed by H enry Le
Chtelier in 1888 in these
words: A ny change in one
of the variables that
determ ine the state of a
system in equilibrium
causes a shift in the
position of equilibrium in
a direction that tends to
counteract the change in
the variable under
consideration (M oore,
1955, pp. 79-80). In other
Ca2+ +when
2 C1a+reaction
H zO + Cat 0 2 (9.4)
equilibrium is disturbed, it
reestablishes equilibri um
by counteracting the
disturbance. This is the
essence of Le Chteliers
principle. This principle
will be very useful in the
interpretation of equa
tions representing
chemical reactions.
O ur experim ent
leads us also to understand
the m eaning of the
concept of solubility,
which is defined as the am
ount of a com pound th at
dis solves to form a this case, the reaction is
saturated solution.
Therefore, in order to m
easure or calculate the
solubility of a com pound
the reaction by which it
dissolves m ust be at
equilibrium . This is an im
portant point to which we
will return when we
discuss solubility
calculations.
The reaction we used
in our experim ent
involves the dissociation
of a solid into ions.
Therefore, in this reaction
no chemical bonds are
actually form ed by close
encounters betw een ions
or molecules. We chose
that reaction mainly
because it is familiar to all
of us. A nother reaction
that is familiar to
geologists is the decom
position of calcium
carbonate by hydrochloric
acid. This reac tion can be
represented by the
following equation:
The carbon dioxide that is
produced escapes in the
form of bubbles when the
reaction is carried out in
contact with the atm
osphere. U nder these
conditions the reaction
cannot reach equilibrium
because the C 0 2 does not
accum ulate. Therefore,
the reaction will continue
in the direction indicat ed
by the arrow in equation
9.4 until either the HC1 or
the C a C 0 3 is used up. In
prevented from reaching
equilibrium and instead
runs to completion.
112 ACIDS AND BASES

Here we encounter the first of many differ ences The rate of the forw ard reaction is:
between geochemistry and chemistry. In the chemistry v f = /cf(A )(B ) (9.6)
laboratory the dissolution of C a C 0 3 by HCl can be
and the rate of th e backw ard reaction is:
carried out in such a way that the COz cannot escape. In
this case, the reaction will achieve equilibrium and excess = k b(C )(D ) (9.7)
C a C 0 3 and H C l will then coexist without appearing to
w here (A ), (B), (C), and (D ) are the m olar
react with each other. However, when this reaction occurs
under natural conditions, it does not achieve equilibrium am ounts or concentrations of the com pounds or elem
because the CO z escapes into the atm osphere. Similarly, ents sym bolized by A, B, C, and D and k f and
chemical reactions on the surface of the E arth may not
k b are proportionality constants. A t equilibrium
achieve equilibrium because some of the prod ucts v f = v b. Therefore:
continually escape either because they are gases or
kf(A )(B ) = k b(C )(D ) (9.8)
because ions and molecules are carried away by the
movement of groundwater. which can be rew ritten as:

kf (9.9)
A nother reason why some reactions in nature fail to (C )(D ) = K
achieve equilibrium is th at their reaction rates are slow.
kh (A )(B )
This is especially true of reactions involving the transform
ation of a solid com pound into another solid. Such w here K is the equilibrium constant. In the gener al case
reactions require a com plex chain of events and their of a reaction at equilibrium represented by:
rates are deter m ined by the slowest or the rate-determ aA+bB^cC+dD (9.10)
ining step in the process. However, we will see th at the the Law of M ass Action, first form ulated by
fail ure of some reactions in nature to achieve equi librium
does not diminish the im portance of this condition as an G uldberg and W aage in 1863, takes the form:
aid to understanding the geo chem istry of the surface of K (9.11)
the E arth . (CT(D)d =
(A )fl(B )
w here a, b, c, and d are the m olar coefficients
taken from a balanced equation representing the reaction,
9.2 The Law of Mass Action and (A ), (B), (C), and (D ) are the con centrations of
reactants and products of the reac tion at equilibrium . The
The alchem ists originally believed that chemical reactions left side of equation 9.11 is the reaction quotient, which
occur becuse certain elem ents and com pounds love each changes continu ously as a reaction proceeds tow ard
other. R o b ert Boyle expressed reservations about this equilibrium .
idea in 1661 and Claude Louis de B erthollet pointed out W hen a reaction has achieved equilibrium, the
about 150 years later that the direction of chem ical concentrations of reactants and products do not change
reactions can be reversed by adding excess am ounts of and the reaction quotient becom es a con stant know n as
one of
the equilibrium constant K (M oore, 1955).
the products (M oore, 1955). The apparent
reversibility of chemical reactions ultim ately led A lthough it is true that at equilibrium the rate
to the conclusion that reactions achieve a state of of the forw ard reaction is equal to that of the
equilibrium w hen the rate of forw ard reaction is equal to backw ard reaction, the Law of Mass A ction
the rate of the backw ard reaction . For a simple reaction expressed in equation 9.11 cannot be derived from
such as: kinetics. The reason is th at chem ical reactions con sist of
sequences of discrete steps occurring at their own rates.
(9.5)
Therefore, the overall rate of a

reaction is not necessarily a linear function of the m olar
concentrations of reactants and products, but m ay vary as
the square, the cube, or even the square root of the
concentrations. T herefore, the Law of M ass A ction was
initially stated only as a generalization of experim ental
results and did not have the force of a scientific law. It
turns out, how ever, th at the Law of Mass A ction can be
derived from the first and second laws of therm odynam ics
for reactions am ong ideal gases at equilibrium . In order to
apply it to reactions am ong ions and m ol ecules in
aqueous solutions we m ust replace their m olar
concentrations by their activities. The activ ity a of an ion
in solution is related to its m olar concentration c by the
activity coefficient y such that:
a = yc (9.12)
The activity coefficients correct the m olar con
centration of ions for the interference by other ions in real
solutions. In m ost cases, the values of activity coefficients
are less than one, which indi cates th at the effective
concentration (or activity)
of ions is less than their actual concentration .
N um erical values of activity coefficients can be
calculated from the D ebye -H ckel theory and
its extensions, which we will discuss later.
T he Law of M ass A ctio n sta te d in e q u a tion
9.11 m ust now be restated in term s of activ ities: If a
reactio n rep resen ted by equ atio n 9.10 is at equilibrium ,
then:
(9.13)
[C]c[D ]' = K easured. On the other hand, it is m ore convenient to m
[A)[B]b
where the brackets [ ] symbolize activities, wheras
parentheses ( ) symbolize concentrations of ions and
molecules. U nfortunately, there is no uniform convention
to represent activities. We use brackets for activities and
parentheses for concentrations following the exam ple of
G arrels and Christ (1965) and K rauskopf (1979), w
hereas Stumm and M organ (1970) used the opposite
convention and D rever (1988) symbolized activities by a
low er case a with appropriate subscripts. The difference
between activity and concentration is unim por
9.2 THE LAW OF M ASSACTION 113
tant in very dilute solutions and vanishes at infinite
dilution when y = 1.0. We will treat the activity coefficient
as a dimensionless num ber, which enables us to express
activities in the same units as concentration.
The units of concentration of ions and m ole cules in
aqueous solutions are based on am ounts expressed in m
oles but differ in term s of weight or volum e of solvent or
solution.
Molality (m) is the number of moles of solute per kg
of water.
Formality (F) is the number of moles of solute per kg
of solution.
Molarity (M) is the number of moles of solute per
liter of solution.
Normality (N) is the number of equivalent weights of
solute per liter of solution.
The m olarity and norm ality are expressed in terms of
volumes of solution, whereas m olality and for mality are
based on weight of the solvent or solu tion, respectively.
The latter are to be preferred because they are independent
of the tem perature, whereas m olarity and normality vary
with tem per ature because of the expansion of w ater with
increasing tem perature. The data in Table 9.1 indi
cate that the volum e of 1 g of w ater increases from
1.000 13 m L at 0C to 1.043 42 m L at 100C or
about 4.3% . M oreover, the specific volume of
water has a m inim um value of 1.000 00 m L at
3.98C. Therefore, concentrations based on vol umes m
ust specify the tem perature at which the volume was m
easure volumes of liquid than it is to weigh them .
Therefore, both kinds of concentration units are in use.
W hen a chem ical reaction betw een a solid and its
ions and molecules in solution has achieved equilibrium ,
the Law of Mass A ction applies with the following
conventions:
1. The activities of ions and molecules must be
expressed in terms of moles but may be referred
to a unit weight of solvent (molal ity), a unit of
weight of solution (formality), or a unit of volume
of solution (molarity).
114 ACIDS AND BASES

Table 9.1 D e n sity a n d S pecific V o lu m e o f solutions in which the ratio of w ater to solute is sig
P u re W a te r F re e o f A ir nificantly reduced. The convention to express the partial
pressures of gases in units of atm ospheres is arbitrary but
T , C D e n sity , g / m L V o lu m e , m L / g
affects the num erical values of equi librium constants.
0 0.99987 1.00013
The tem parature and pressure at equilibrium m ust be
3.98 1.00000 1.00000 specified because equilibri-ium constants vary with tem
5 0.99999 1.00001 perature. The effect of pressure on equilibrium constants is
10 0.99973 1.00027
negligible under Earth-surface conditions.
15 0.99913 1.00087
20 0.99823 1.00177
25 0.99707 1.00293
30 0.99567 1.00434
9.3 Dissociation of Weak Acids
35 0.99406 1.005 97
and Bases
40 0.99224 1.00782
45 0.99025 1.00984 Following the tradition of the Swedish chemist Svante
50 0.98807 1.01207 Arrhenius, we define an acid as a com pound that releases
55 0.98573 1.01447 hydrogen ions when it is dissolved in water. Similarly, a
base is a com pound that releases hydroxyl ions in aqueous
60 0.98324 1.01704
solution. These defini tions are no longer accepted by
65 0.98059 1.01979
chemists who now prefer B rpnsteds definition that an
70 0.97781 1.02269
acid is any sub stance that can donate a proton to another
75 0.97489 1.02575
sub stance and a base is a substance that can accept a
80 0.97183 1.02898
proton . Strictly speaking, protons or hydrogen ions (H +)
85 0.96865 1.03236 cannot exist unattached in w ater and interact with w ater
90 0.96534 1.035 90 molecules to form hydronium ions (H 3 0 +). We adopt the
95 0.96192 1.03958 A rrhenius concept of acids and bases because in
100 0.95838 1.04342 geochemistry we deal only with aqueous solutions of
eletrolytes and we will represent protons as H + regardless
SO U R C E : Weast et al. (1986), p. F-10.
of whether they actually occur in this form.

2. The activities of pure solids and of w ater are equal

to one.
3. The concentration of gases are expressed as
partial pressures in atm ospheres.
4. R eactions are assum ed to take place at the
standard tem perature (25 C) and pressure ( 1
atm ) unless otherw ise indicated.
The use of m oles as the unit of mass is required
because chemical reactions result from the interac tions of
individual ions or molecules whose num ber per mole is
specified by A vogadros num ber. The convention that
solids and w ater have unit activities eliminates them from
the equation of the Law of Mass Action. The activity of w
ater may, however, be less than unity in very concentrated
Some acids are said to be strong because they release
all or m ost of the available hydrogen ions when they are
dissolved in water, w hereas weak acids release only a
small fraction of their hydro gen ions. Bases are classified
similarly by the extent to which they release hydroxyl ions
(O H - ) in aqueous solutions. How ever, the behavior of
bases is m ore com plicated because som e bases do
not readily dissolve in water. For example,
M g(O H ) 2 is so insoluble in w ater th at it occurs
naturally as the m ineral brucite. T herefore, a sat
u rated solution of brucite has a low concentra
tion of O H - even though M g(O H ) 2 is a strong
base and dissociates extensively into M g2+ and
O H - . Som e com m on strong acids and bases are:
 9.3 DISSOCIATION OF W EAK ACIDS AND BASES 115
Strong acids Strong bases (i.e., y = 1 .0 ) and that an unknow n fraction x

Hydrochloric acid (HC1) hydroxides of alkali
metals
Nitric acid (H N 03) hydroxides of alkaline
earths (except Be)
dissociated into ions. T herefore, at equilibrium :
[CH 3 C O O H ] = 0.1 - x
[CH 3 C O O -] = [H +] = x
Sulfuric acid (H2S 0 4)
lanthanum hydroxide
The com m on w eak
acids include acetic acid
(C H 3 C O O H ), carbonic
acid (H 2 C 0 3), phosphor
ic acid (H 3 P O 4 ), and
silicic acid (H 4 S i0 4). A
m ong the weak bases we
m ention am m onium
hydrox ide (N H 4 O H ),
nickel hydroxide (N i(O
H )2), cop p er hydroxide
(C u (O H )2) and the
hydroxides of the R E E s
except La. A clear
distinction m ust be m ade
betw een the strength of an
acid or base and its co n
centration. For exam ple,
hydrochloric
acid rem ains
tion because it is com pletely dissociated . Sim i
larly, acetic acid rem ains
a w eak acid even in a co n
cen trated solution because
it is only partially
dissociated.
M any elem ents
form hydroxides that can
release either H + or O H -
depending on the con
centration of H + in the w
ater in which they dis
solve. These hydroxides
are said to be am photeric.
A m ong the m ore that 20
elem ents w hose hydrox
ides are am photeric we
find Be, A l, Si, Ti, V, Fe,
Co, Zn, Ag, Au, Sn, Pb,
and U.
W hen a w eak acid
such as acetic acid is dis
solved in water, it
dissociates into ions:
CH3COOH Substituting into equation equation 9.16, we note that
-> C H 3 C O O - 9.15 we obtain a qua K is a very small num ber,
(La(OH)3) + FL dratic equation: which im plies th at only a
(9.14) sm all fraction of the
x2
D issociation reactions = 1.76
0.1 x
like this one reach equilib
rium very quickly and are This equation can be
then subject to the Law of solved by putting it into
Mass Action: the standard form:
[H +][CH 3 COQ~ ax2 + bx
which then yields the quadratic formula:
[CH 3 C O O H ]
w here K = 1.76 X 10~ 5 at 25 C and -b (b2 -
small value of K indicates that at x=
m ost of the acetic acid will rem ain in the undis
sociated m olecular form . In ordInerour
to case, we find th at a = 1,
tion 9.15 we now specify thatand c = - 1 . 7 6 x l 0 -6 and that
com pound was dissolved in 1 Lyields x l = +1.32 X IO- 3 and
We discard the negative
root of the quadratic
equation and conclude
m olar solution of acetic acid contains:
[H +] = 1.32 x
a strong acid even in dilute solu
[CH 3 C O O '] = 1.32 X
[CH 3 C O O H ] = 0.1 - 1.32
The degree of dissociation
solution of acetic acid is:
3 2
D- 1 X /
Evidently, acetic acid is
a weak acid because
only 1.32% of the m
olecules in solution are
dissociat
ed into ions.
The first application
A ction has already taught us
problem s can result in
equations th at are tedious
to solve. We therefore
resolve to m ake approxim
ations w henever possi ble
in ord er to reduce the
effort required to work
through to the answer. For
example, going back to
116 ACIDS AND BASES
acetic acid m olecules will dissociate. Therefore,
we assume that x 0 . 1and that 0 . 1
Equation 9.16 now reduces to:
= 1.76 X 10 -5
0.1
from which it follows im m ediately that:
x = (1.76 X 10- f ) 1 / 2 = 1.32 X 10- 3
Therefore, x really is much less than 0.1 and our assum
ption was justified. The final result is indis tinguishable
from the value we obtained by solv ing a quadratic
equation.
A nother lesson we can learn here is that m ost num
erical problem s in geochem istry are based on conceptual
m odels whose validity as representa tions of reality is
limited. For this reason, it is inappropriate to extend num
erical solutions of problem s to m ore than two or three
significant figures. Similarly, a difference of 10% betw
een approxim ate and exact solutions is acceptable in m
any cases. In fact, approxim ate solutions have a special
virtue because they force the investigator to develop som e
insight into a problem before attem pting to solve it.
progress of chemical reactions th at may be occur ring in
x ~ 0 .1 . an aqueous solution and because it can be m easured by
eletrical m ethods (Stum m and M organ, 1970). Therefore,
(9.20) the activity of hydro gen ions is com m only expressed as
the pH value, defined by:
pH = - lo g 1 0 [H +] (9.25)
(9.21)
It follows from equation 9.24 that the pH of pure w ater at
25 C is equal to 7.0. W hen the hydrogen
ion activity of a solution is greater than
10- 7 m ol/L , its pH is less than 7, and when its
activity is less than 10- 7 m o l/L , the pH is greater th at 7.
These statem ents therefore define the p H scale according
to which pH < 7.0 at 25 C indi cates acidic conditions and
pH > 7.0 indicates basic conditions.
In conclusion, we can now calculate the pH of
a 0.1 m olar solution of acetic acid. A ccording to
equation 9.21, [H +] = 1.32 X 10- 3 m o l/L from which we
calculate the pH by applying equa tion 9.25:
pH = - lo g (1.32 X 10-3) = 2.88 (9.26)

The hydrogen ions released by the acetic acid affect the

We have seen that acetic acid releases hydro gen ions
activity of hydroxyl ions th at m ust rem ain in equilibrium
w hen it is dissolved in water. However,
with water. Therefore, from equation 9.23:
w ater is itself dissociated into ions:
H 2CT H + + O H " (9.22)
0 -14
[O H "] = 1 x1
1.32 x 10- 3
and the dissociation constant at 25 C is, to a very good
approxim ation, K w = 1 X 1 0 '14. A ccording to eq u atio = 0.757 X 10- 1 1 m o l/L (9.27)
n 9.22, pure w ater contains equal con centrations of The hydroxyl ion activity can be expressed as
hydrogen and hydroxyl ions whose activities are derivable
from the Law of M ass A ction: p O H defined by:
PO H = - lo g 1 0 [O H - (9.28)

[H +][O H ~ -14 (9.23) In the solution of acetic acid pO H = -log(0 .757 x

[h2o] =1 x 10

10- n ) = 11.12. Evidently, it follows from equa

Since [H2O j = 1.0 by convention and since
tion 9.23 that at 25 C:
[H +] = [O H - ], we find that in pure water: p H + pO H = 14.0 (9.29)
[H +] = 1 X 10- 7 m o l/L (9.24)

The activity of hydrogen ions is a useful p ara The dissociation constant of water, like all other
equilibrium constants, varies with tem pera ture, as shown
m eter because it reflects directly or indirectly the in Table 9.2. Therefore, th e neutral
9.3 DISSOCIATION OF W EAK ACIDS AND BASES 117
 1979). The reactions [H 2
Table 9.2 D isso c ia tio nrapidly
C o n sachieve
ta n t have the sam e S(aq)] =
o f W a te r a t D iffe re n t T equilibrium
e m p e ra tuand
re sthen concentration in both of 0.1 - X
T, C - lo g K w
satisfy the Law of Mass the equi libria in which [H
Action: they occur. However, the +]
0 14.9435 =x
am ount of H + produced
[H +][H S ~] by the first equilibrium
5 14.7338 [HS
10 14.5346 (9.30) is m uch greater ~] =
[ h +][s2 ]
15 14.3463
than th at contributed by x
14.1669
the second equilibrium
20 [HS] (9.31) because K x is From the Law of Mass A
24 14.0000 ction we obtain:
H ere we encounter for the about 1 million tim es
25 13.9965
first tim e two sim ulta larger than K v
neous equilibria th at have (9.34)
S O U R C E : Weast et al. 0 .1 -
(1986), p.D-164. (H + and H S - ). Obviously, The problem is to x
calculate the activities of
Since K x is a very small
all ionic and m olecular
num ber, we predict that
point of the pH scale is at pH species in a solution con
only a small fraction of H
Actually, the data in Table taining 0.1 mol of H 2S
2 S(aq) will be dissociat
K w = IO- 1 4 0 at 24C and at 25 C it is slightly dissolved in 1 L of pure w
er than this value. ater at 25 C. We will dem ed at equilibrium ,
Several weak acids onstrate two strategies for which justifies the
are capable of yielding solving problem s assum ption
two or m ore hydrogen involving two or m ore th at x 0.1 and that 0.1 - x 0
ions or protons per m simul taneous equilibria: equation 9.34 becomes:
olecule of acid. W hen 7.0

such acids dissolve in w

1. Stepwise solution.
ater they disso ciate from which it follows that
2. Sim ultaneous equations.
stepwise such that each
The stepwise m ethod applies to
step has a different degree [H +] = [HS~] = 10~ 4 0 m ol/L , and pH
of dissociation. M any of weak acids and bases, w heredegree of dissociation of H 2 S(aq) is:
bases behave exactly the constants of successive
same way when they steps differ by orders of m
dissociate in w ater to agnitude. H ow ever, the
form O H ~. In order to use of sim ultaneous
illustrate the application of equations is m ore exact
the Law of Mass Action to and applies to all situa
weak diprotic acids we tions in which several
choose hydrosulfuric acid reactions are involved.
(H 2 S).This com pound is The stepwise m ethod
a gas th at dissolves in w is based on simplifying
ater to form H 2S m assumptions that the
olecules reactions reach
in aqueous solution, represented equilibrium in sequence.
which then dissociate in two steps: Therefore, we turn first to
H 2 S(aq); H + + reaction 9.30 and let x be
H S "; ; H + + the num ber of moles of H
2 S(aq) that dissociate. A t

w here K } = IO- 7 0 and K 2 equilibrium we then have:

118 ACIDS AND BASES
io -40 x 100 = . 0 1 % (9.36)
(equation 9.30) in accordance with Le C hteliers
o .i principle. H ow ever, the stepwise solution disre gards this
T herefore, the assum ption that x 0.1 was justi interdependence of sim ultaneous equi libria having com m
on ions.
fied in this case.
The dissociation of H 2 S(aq) can be represent
A certain num ber of HS m olecules dissoci
ed exactly by m eans of sim ultaneous equations
ate fu rth er to form S2- and additional H +. We
derived from applications of the Law of Mass
let y be the num ber of moles of H S - th at dissoci
A ction and from the requirem ent of electrical
ate and set:
neutrality and mass balance. The num ber of equa tions
[HS- ] = 10- 4 - y
required to solve the problem m ust be equal to the num
[H +] = 10- 4 + y ber of unknow n param eters. In this case we have five
unknowns: H 2 S(aq), H S", S2 -, H +, and O H - , and
[S21 = y
therefore we require five equa tions. The first two arise
Then, from the equation 9.33: from the application of the Law of Mass A ction to the two
dissociation equilibria and w ere stated previously as equa
(1 0 - 4 + y )(y ) (9.37)
10' 12.9 tions 9.30 and 9.31. To these we add an equation based on
(10-4 - y)
=
the dissociation of water:

We again simplify the problem w ith the assum p

[H +][O H - ] = 10" (9.39)
tion th at y . IO- 4 0 and that IO- 4 , 0 + y = IO- 4 0
and IO- 4 0 - y ~ IO-40. It follows th at eq u a Next, we recall th at all sulfur species in the solu tion arise
tion 9.37 reduces to y = 10-129, which confirms by dissociation of H 2 S(aq) whose initial concentration
0
the validity of the assum ption th at y IO- 4 was 0.1 m ol/L . Therefore:
and indicates th at the degree of dissociation is:
2 (H 2 S) + (H S - ) + (S2-) = 0.1 (9.40)
IO- 1 9 X 100 = 10-6-9% (9.38)
N ote th at the mass balance equation m ust be stat
D >= ^To
1 0 ed in term s of concentrations rather than activi
T hese calculations therefore indicate th a t a ties. However, in dilute solutions the activity
0.1 m olar solution of H 2 S(aq) contains the follow ing coefficients of all ions are close to unity and the
difference betw een activities and concentration
ions and molecules:
therefore vanishes. The fifth equation is derived
4
0
[H 2 S] = 0.1 - 10- = 0.1 m o l/L from the requirem ent of electrical neutrality:
0 2'9 0
[H S - ] = IO- 4 - IO- 1 = 10- 4 m o l/L (H +) = 2(S2-) + (H S - ) + (O H - ) (9.41)
0 29 0
[H +] = IO- 4 + 10- 1 = IO- 4 m o l/L
which states that the sum of the positive charges in the
2
[S2-] = 10- 1 9 m o l/L solution m ust be equal to the sum of the nega tive charges.
The m olar concentration of S2- is m ultiplied by 2
The second dissociation of H 2S contributes only a because each S2- ion contributes two charges. O nce again
negligible am ount of H +, but it provides all of the S2- we see the im portance of A vogadros law because it
ions w hose activity is equal to the value of the second allows us to relate the quantity of an ion or molecule,
dissociation constant. expressed in moles, to the corresponding num ber of ions
The stepw ise m ethod presented above is inexact or molecules.
because it treats the two sim ultaneous equilibria as though We now have five independent equations that express
they were independent of each other. If we look again at relationships among the five unknowns and therefore we
equations 9.30 and 9.31, we can see th at the dissociation can solve the problem . The pro cedure is to express the
of H S - (equation 9.31) causes m ore H 2S to dissociate activities of the ions in term s of one of the ions, and then
to substitute
9.3 DISSOCIATION OF W EAK ACIDS AND BASES 119

these relationships (after
converting them to con
centrations) into the
equation for electrical neu
trality because that
equation contains all of the
ions in the solution. We
then solve this equation,
by trial and error if
necessary, and use the
result to calculate the
activities of all of the other
ions and molecules in the
solution.
Once again, the
sensible approach is to m
ake simplifying assum
ptions that reduce the com
puta tional labor w ithout
com prom ising the
validity of the results. In
the case at hand, we
assum e that [O H ] [H
+] because we are
discussing the dis
sociation of an acid.
Therefore, we may om it
the hydroxyl ion in sum m
ations such as equation
9.41 but n o t from m
ultiplications such as
equation 9.39. In addition,
we know from past
experience that [s2-]
[HS ] because K 2 K v
This assum p tion
simplifies both equations
9.40 and 9.41. From
equation 9.40 we obtain:
(H 2 S) + (H S - ) = 0.1
and from 9.41:
(H +) = (H S ")
The latter can be used to
revise equations 9.32 and
9.33. From equation 9.32
we obtain:
[H S - ] 2
[H2 S] C d(O H )2 (s) + OH-^Cd(OH)3
and from equation 9.33: which reduces to [H S - ]
K (am photeric)
[S2-] = IO- 1 2 -9 = 1 X IO- 4 0 m ol/L .
Evidently, in this The two dissociation steps
Next, we use equation 9.44 to expresscase, the stepwise
may be com bined by
[H S - ]2 /1 0 - 7 0 and substitute approach adding equations 9.52 and
equation 9.42: yields the sam e results as 9.53. In this operation C d
[H S - 1 2 + H Sthe - solution of sim ulta (O H )+ cancels, and we
neous equations because
1 0 -
obtain:
the dissociation con stants
', -7 .0 andare
rearranging
not onlyterm s yields:
small num C d(O H )2 (s) ^ Cd2 +
Multiplying by 10 bers but also differ (aq) + 2 O H - Kr
[H S - ] 2 + 10- 7 0 [H S - ] from - (9.56)
each other by a factor of nearly
We solve this equation by These circum stances do not apply to
acids and bases whose w here K T = K x X K 2. E
m eans of the quadratic
dissociation constants are quation 9.54 represents the
form ula (equation 9.18) solution of solid C d(O H )
listed in Table 9.3. M
and obtain: 2 to form m olecular
oreover, we will encounter
- I O -7 0 C d(O H ) 2 in aqueous solution. We can i
sets of sim ultaneous
[IO- 1 4 0 - rate th at step into the dissociation eq
equilibria in w hich the
sequential treatm ent is
4 ( - 1 0 -s
o )]1/2 inappropriate and which
[H S - ] can be solved only by m
(9.48)eans of a set of indepen
dent equations.
The dissociation
constants listed in Table
9.3 apply o nly to specific
equations. In the case of a
polyprotic acid such as
arsenic acid the appropri
ate equations are:
H 3 A s 0 4 (aq) H 2 A s 0 4- +
H 2 A s O / ^ H A sO 2 - +
H A sO 2 - ^ A sO 3- + H +
N ote that the acid molecule H
dissolved in w ater in the
The situation is m ore
because som e of them are
insoluble and because
they m ay also be am photeric (A ). We
C d(O H ) 2 to define the appropriate equations:
C d(O H )2 ( s ) ^ C d (O H )+(aq) +
C d(O H ) + ( a q ) ^ C d 2+ + O H -
C d(O H )2 ( s ) ^ C d ( O H )
120 ACIDS AND BASES
Table 9.3 Dissociation Constants of Acids and Bases at 25Ca

Name Formula pK, VK 2 P*3 PKA P^aq R e fb

A dds
Acetic CH3COOH 4.75 1
Arsenious H3A s 0 3(aq) 9.2 2
Arsenic H3A s 0 4(aq) 2.2 7.0 11.5 2
Boric H3B 0 3(aq) 9.2 2
Carbonic H2C 0 3(aq) 6.35 10.3 2
Hydrofluoric HF(aq) 3.2 2
Phosphoric h 3p o 4 2.1 7.2 12.4 2
Hydrosulfuric H 2S(aq) 7.0 12.9? 2
Sulfuric H2S 0 4(aq) 2.0 2
Hydroselenic H 2Se(aq) 3.9 15.0 2
Selenic H 2S e 0 4(aq) 1.9 2
Silicic H4S i0 4 9.71 13.28 9.86 13.1 3
(pK 4)
Bases (Hydroxides)
Ammonium NH4OH(aq) 4.7 ; 2
Aluminum, amorph. Al(OH)3 12.3 10.3 9.0 2
Aluminum, gibbsite a i ( o h )3 14.8 10.3 9.0 1.4 2
Beryllium, amorph. Be(OH)2 2.2 2
Cadmium Cd(OH)2 10.5 3.9 4.1? 6.7 2
Cobalt Co(OH)2 10.6 4.3 5.2 6.5 2
Copper Cu(OH)2 13.0 6.3 2.9 2
Ferrous Fe(OH)2 10.6 4.5 5.1 8.4? 2
Ferric, amorph. Fe(OH)3 16.5 10.5 11.8 4.4 2
Lead, red PbO + H20 9.0 6.3 1.4 4.4 2
Magnesium Mg(OH)2 8.6 2.6 2
Manganese Mn(OH)2 9.4 3.4 5.1 2
Mercurous, red HgO + H20 14.8 10.6 4.5 3.6 2
Nickel Ni(OH)2 11.1 4.1 4 7 2
Silver 2 Ag20 + 5 H20 3.4 5.7 2
Thorium, amorph. Th(OH)4 10.3 5.8 10.8 2
uo2 u o 2( o h )2 14.2 8.2 3.6 (P*4) 2

Vanadium v ( o h )3 11.7 2
VO v o ( o h )2 15.2 8.3 2
Zinc, amorph. Zn(OH), 10.5 5.0 1.9 4.4? 2

Explanations: pK = - log K
Zn(OH)2(s) Zn(OH) + + OH- K,
Zn(OH)+ ^ Zn2+ + OH K2
Zn(OH)2(s)^Zn(OH)2(aq) Kaq
Zn(OH)2(s) + OH- =^Zn(OH)3 KA (amphoteric)
means does not apply or unknown.
b(l)Weast et al. (1986); (2) Krauskopf (1979); (3) calculated from thermodynamic data of Lindsay (1979).
9.4 SO LUBILITY OF SPARINGLY SOLUBLE BASES 121
combining equation 9.54 and 9.52. In ord er to accomplish ous equilibria existing in a given environm ent. Therefore,
this, we first invert equation 9.54: the degree of dissociation of w eak acids and bases and the
resulting solubility of o th erwise insoluble weak bases are
C d(O H )2 ( a q ) ^ C d ( O H ) 2 (s) 1 / K aq (9.57) controlled by the activity of H + ions in the environm ent.

and we then add it to equation 9.52, thereby elim inating C

d(O H )2 (s):
9.4 Solubility of Sparingly
C d(O H )2 (aq) ^ C d (O H )+ + O H " K J K aq
(9.58)
Soluble Bases
Evidently, these equations 2. The solubility of We have defined the equal to one and thereby
can be m anipulated sparingly soluble solubility of a com pound equate activities and
algebraically w ith the bases is pH as the am ount of the com concentrations.
understanding that: dependent. pound that dissolves to The necessary
form a saturated solution. equations are derived from
1. W hen equations In geochem istry the p H is the dissociation equilibria
Since m any bases are only
are added, their an environm ental para
sparingly soluble, we may of M g(O H ) 2 and H 20
equilib rium meter that is determ ined
use the Law of M ass A and from the requirem
constants are m by all of the sim ultane ction to calculate their ent of electrical
ultiplied. solubilities. We choose M
2. W hen equations neutrality:
g(O H ) 2 for this dem
are subtracted, onstration because it has a
their equi librium low solubility and
constants are therefore occurs as the m
divided. ineral brucite.
3. W hen an equation
The first and m ost im portant step tow ard the
is reversed, its solution of a problem involving the Law of M ass A ction is
to w rite the equations that represent the relevant equilibria.

equilib rium The dissociation of M g(O H )2

constant is proceeds in two steps:
inverted. M g(O H )2 (s) ^ M g(O H )+(aq) + O H -
Finally, we n o te th a t
the am photeric n atu re
M g (O H )+(aq) ^ Mg^aq)+
w hen it takes on an additional hydroxyl ion
form C d (O H )/, a hydrogen ion is
A ccording to Table 9.3,
the dissociation of w ater. T herefore, C d (O H
this case acts as an acid because it releases H + K 2 = IO-2,6. Evidently, in
into solution. We predict, in accordance with this case, K 2 is 1 million
Le C hteliers principle, tim es larger than K v
character of C d (O H ) 2 is which may cause a
en h a n ced in basic solu problem if we attem pt a
tions because a high stepwise solution. T
activity of O H - causes the herefore, we decide to treat
equilibrium to m ake m ore the problem by m eans of a
C d (O H )/ in ord er to set of sim ultaneous
reduce the activity of O H equations. We have four
- . C onversely, in acidic unknow ns [M g(O H )+,
solutions, which have a M g2+, O H - , and H +]
low activity of O H - , the and therefore need four
reaction shifts to the left by independent equations to
converting C d (O H ) 2 determ ine the activities of
to O H - and C d (O H )2. T herefore, all ions in a saturated
hydroxides act like
solution of brucite in pure
and like bases in acidic environm ents.
water. We elect not to
The behavior of
specify the pH of the
teaches us two im portant lessons:
solution at this tim e in
1. W eak acids and bases do
order to dem on
pH of natural environm ents
respond to it. strate the pH dependence of the solubility of
M g(O H )2, but we do set all
 122 ACIDS AND BASES
[M g(O H )+][O H - ] = 1(T8 6 (9.61)
= io -3-94
10 - 11.2

[Mg2+] =
[Mg2+] [ O H l i o -726
= 11.48 X 10- 5 m o l/L (9.71)
[M g (O H *)] - 10~ <9-62>
[H +][O H - ] = IO 1 4 0 (9.63) T he solubility of M g(O H ) 2 in pure w ater at

25 C is the s u m of the concentrations of the

(O H ") = 2(M g2+) + (M gO H +) + (H +) (9.64)
two M g -bearing ions and th erefo re equals
12.55 X 10- 5 m o l/L . Since the m olecular w eight
In order to simplify the problem , we assum e that
of M g (O H ) 2 is 58.3267, we can express its solu
(H +) (O H - ) and therefore drop (H +) from bility as 58.3267 X 12.55 X 10- 5 = 7.32 X 10- 3
equation 9.64. Equations 9.61 and 9.62 allow us to g /L , w hich is n o t a large am ount. The activity of
relate the M g-bearing ions to [O H - ]: H + in a satu rated solution of M g(O H ) 2 in pure
(9.65) w ater is 10- l 4 0 /1 0 - 3 '6 3 = 1 0 - 1 0 ' 3 7 m o l/L and the
10- 8.6 pH = 10.37. E vidently a satu rated solution of
[M g(O H )+] =
[O H - ] M g (O H ) 2 is strongly basic.
10- 2 6 [M g(Q H +)] The relationship betw een th e activities of
[Mg2+] = th e M g ions in a satu rated solution of M g(O H ), and
[O H - ] the pH is indicated by equations 9.65 and 9.66. By taking
lO -i i-2 (9.66) logarithm s and converting pO H to pH by m eans of eq u
[O H - ] 2 atio n 9.69, we obtain from equation 9.65:
Substituting into equation 9.64 yields:

r 2 X IO-11'2 IO-86 log [M g(O H )+] = - 8 . 6 + pO H (9.72)

or
[ H ] [O H - ] 2 + [O H - ] (9 67)
2 log [M g(O H )+] = 5.4 - pH (9.73)
A fter m ultiplying by [O H - ] and rearranging
Similarly, from equation 9.66:
terms, we obtain one equation in one unknown: [O H - ] 3
log [Mg2+] = 16.8 - 2pH (9.74)
- 10- 8 6 [O H - ] = 2 X 10- 1 1 2

= 1.26 x 10- n (9.68) E ach of these equations has been plotted in

Figure 9.2 in coordinates of the log of the activity and pH .

The equation can be solved by trial and error. We assume
The total solubility of M g(O H ) 2 is the sum of the
that [O H - ] > 10- 7 because M g(O H ) 2 is a base. W hen
concentrations of M g(O H )+ and Mg2+ at different
we try [O H - ] = IO- 4 0 in equation 9.68, we obtain:
values of the pH.
E quations 9.73 and 9.74 are straight lines that
- 12.0 -13 (9.69) intersect at p H = 11.4. We see by inspection of
10 - IO- 1 2 6 = 7.5 x 10
Figure 9.2 th at M g2+ is the dom inant ion at pH
which is less than 1.26 X 10 n . F u rth er trials indi values <11.4, w hereas M g(O H )+ becom es dom
cate th at equation 9.68 is solved satisfactorily
'63
inant in m ore basic solutions. Such extrem ely
when [O H - ] = 10- 3 m ol/L . T herefore, the
basic environm ents occur only in highly saline
activities of the M g-bearing ions from the equa
lakes such as L ake N atron in E ast Africa.
tions 9.65 and 9.66 are:
T herefore, Mg2+ is the dom inant ion in most geo
[M g(O H )+] = = IO- 4 9 7
logical environm ents. Brucite is very soluble in
acidic solutions and can occur in contact with
= 1.07 X 10- 5 m o l/L (9.70) w ater only w hen the activity of Mg2+ is large
 9.5 pH CONTROL OF DISSOCIATION EQUILIBRIA 123

PH
Figure 9.2 Solubility of brucite (Mg(OH)2) in pure water at 25C as a function of pH.
Mg(OH)2 dissociates to form Mg(OH)+ and Mg2+ ions, whose activities in a saturated
solution are strongly pH dependent, as given by equations 9.73 and 9.74. The resulting
solubility of brucite is the sum of the concentrations of the Mg-bearing ions in the
solution and is indicated by the dashed line. The coordinates of point P on the diagram
represent conditions that will cause brucite to precipitate. The straight lines representing
activities of Mg2+ and M g(OH)+ have been arbitrarily terminated at unit activities
because in nature other Mg compounds will precipi tate at elevated concentrations of
Mg ions.
enough to m aintain
equilibrium . T herefore,
the solubility of brucite in
w ater can be used to
define its stability as a
solid phase in a geological
environm ent.
C onsider point P in
Figure 9.2 w hose coordi
nates are [Mg2+] = 10 2 m
o l/L and pH = 10. The
coordinates of P do not
satisfy equation 9.74,
which implies that brucite
cannot be in equilibri um
with a solution of that com
position because, at
this value of [Mg2+],
high. The application
to equations 9.59 and 9.60
indicates th at the pres
ence of excess O H will
result in the precipitation
of solid M g(O H ) 2 until
both [Mg2+] and [O H - ]
have been reduced
sufficiently to place point
P on
the equilibrium line. We
conclude, therefore, that
brucite will precipitate in
all environm ents that lie
to the right of the
solubility limits in Figure
9.2. In
other words, the activities of the
M g(O H ) 2 constrain its stability
w ater containing M g2+
and M g(O H ) + .
9.5 pH
Control
of
Dissoci
ation
Equilibr
ia
The dissociation of weak
acids and bases is con
trolled by the pH of the
geochem ical environ m
ent. In order to exam ine
this concept we return to
the dissociation of H 2
S(aq) represented by
124 ACIDS AND BASES

equations 9.30 and 9.31. H ow ever, here we rew rite the rro m m e mass oaianee:
m ass-action equations (9.32 and 9.33) as follows: (H 2 S) + (H S - ) + (S2-) = 1 X 10- 2 m o l/L

[HS] (9.75) (9.79)

10 -7.0
If all activity coefficients are equal to one, we can
[ h 2 s] [ h +]
substitute equations 9.77 and 9.78 into 9.79, obtaining:
and:
129
[S2-] IQ- (9.76)
10[HS- ] + [HS- ] + 10- 6 9 [H S - ] (9.80)
[HS] [H +] = 1X 10 -2 m o l/L
We see th at the activity ratios of [H S - ]/[H 2 S]
which reduces to:
and [S2 ]/[H S -] depend on the activity of the 11[HS- ] = 1 X 10- 2 m o l/L (9.81)
hydrogen ion. A ccording to equation 9.75,
and hence [HS- ] = 9.09 X 10- 4 m ol/L . From
[H S "]/[H 2 S] = 1.0 when [H +] = IO- 7 0 m ol/L . If
[H + ] > IO-70, say 10-60,th en [H S - ]/[H 2 S] = 0.1, equation 9.77 we determ ine that [H 2 S] =
which m eans that the activity of H 2S is 10 times 9.09 X 10- 3 m o l/L and from 9.78 we have
greater than the activity of [H S - ]. Similarly, if [S2-] = 1.14 X 10- 1 0 m ol/L . Therefore, H 2S is the
[H +] < IO70, say IO80, then [H S - ]/[H 2 S] = 10 m ost abundant species at pH = 6.0 and constitutes 90.9%
and [HS ] is 10 tim es greater than that of H 2 S. of all S-bearing ions or molecules in the solution. The
Evidently, [H +] = 10-7 (pH = 7.0) is an im por bisulfide ion (H S - ) is next in abun dance w ith 9.09%,
tant boundary at which the abundances of H 2S which leaves only 0.01% for S2 -. W hen we repeat this
and H S - are equal. H 2S dom inates at pH < 7.0, calculation for different values of the pH, we generate
w hereas H S - dom inates at pH > 7.0 three curves that represent the changing abundances of H 2
Similar arguments, applied to equation 9.76, indicate S, H S - , and S2- in the solution. As expected, we see
that pH = 12.9 is another boundary in from Figure 9.3 that H 2S dom inates at pH < 7.0, that H S
the system . A t pH < 12.9 we find th at H S - is - is the dom inant ion for pH values from 7.0 to 12.9, and
m ore abundant than S2-, w hereas at p H > 12.9 that S2- is dom inant at pH > 12.9. A lthough all of the
the sulfide ion (S2 -) dom inates. Such relation ions or molecules are present throughout the range of pH
ships occur in all w eak acids and bases because their values, m ost of the S in the solution is associated with a
dissociation into ions is controlled by the pH of th e particular species except at the boundaries at which the
environm ent. abundances of two species are equal.
The control of dissociation equilibria by the pH has
an im portant consequence because it means that we can
identify the dom inant ions at dif ferent values of the pH .
In order to dem onstrate how this works we fix the total
am ount of sulfur-bearing species in the solution at 1CT 2 9.6 Solubility of Amorphous Silica
m o l/L and calculate the activities of H 2 S, H S - , and S2-
as a function of pH . A t pH = 6.0 we obtain from equa The oxides of m any m etals react with w ater to
tion 9.75: form bases and the oxides of nonm etals react with
w ater to form acids. For example, C 0 2 reacts with
w ater to form carbonic acid:
[ h 2 s] - 1 0 2
1
] = QtH S ~] (9 -77) CO,+ h,o->h2co3 (9.82)
1
0

and from equation 9.76:

and S 0 2 form s hydrosulfurous acid:
0
10 [HS ] = 1 6.9 [H S -] (9 78)
[S2-] = S02 + h2o->h2so3 (9.83)
1 0 - 6.0
io ro
 9.6 SOLUBILITY OF AM ORPHOUS SILICA 125

PH
Figure 9.3 The pH dependence of the abundances of S-bearing ions and molecules in a solution of H2S(aq)
containing a fixed amount of S in solution (equation 9.79). Note that H2S is dominant at pH < 7.0, HS" domi nates
between pH = 7.0 and 12.9, and that S2_ is dominant only at pH > 12.9. However, all of the ions are pre sent at all pH
values even though their abundances may be small compared to that of the dominant ion.
In a very similar m anner,
silicon dioxide reacts with
w ater to form silicic acid,
such that at equilibrium:
S i0 2 (am orph.) + 2 H 20
S i0 2 is a solid com pound
that can either be am or
phous or assum e several
different polym orphic
crystalline forms,
including quartz,
cristobalite, and tridymite.
Silicic acid actually forms
as a result of chemical w
eathering of the com m on
rock-forming
aluminosilicate minerals
such as the feldspars and
micas rather than by
solution of crys talline or
am orphous S i0 2.
However, when its solu
bility is exceeded, it forms
am orphous silica, which
settles out slowly as a
gelatinous precipitate (Iler.
1979). Given sufficient
time, the precipitate expels
w ater and begins to
crystallize, forming opal A
and opal C T as interm
ediate phases (K astner et
al., 1977).The process
ends with the
crystallization of a
cryptocrystalline variety
of quartz called chal
cedony, which forms chert
or flint and occurs in geo-
des as agate. Q uartz is
highly insoluble, and
dissolves in w ater only
sparingly, even during
long periods of geologic
time. Am orphous silica,
howev er, is much m ore
reactive and generally
maintains equilibrium
with silicic acid.
We are now in a
position to discuss the
solu bility of am
orphous silica based on
equation 9.84 and on
the dissociation T the
silicic acid:
126 ACIDS AND BASES
H4Si04 ^ H 3 S i0 4 +H + (9.85)
K x = 10 -9.71 of the concentrations of the Si-bearing ions in solu
H 3 S i0 4" ^ H 2 S iO ^ + H + K 2 = 10
1128
(9.86) tion. The first dissociation step (equation 9.85)
m akes the largest contribution to the increase in
9S6
H* 2,S^ i0 24- ^ HSiO] + H + ,v3K -=. u10 (9.87) the solubility of S i0 2 (am orphous) with increasing pH .T
he other ions come into play only in extrem e ly basic
H SiO 3- ^ SiO4 + H + K 4 = IO 1 3 1 0 (9.88) solution.
The solubility of am orphous silica as a func
W hen am orphous silica and p u re w ater have reached tion of pH is shown graphically in Figure 9.4.
equilibrium at 25 C (equation 9.84), the activity of silicic N ote that the concentration is expressed in m il
acid is given by the Law of Mass A ction: ligrams per liter (m g/L ) of SiOz because it is
com m only stated this way in chemical analyses of w ater.
The S i0 2 is not actually present in m olecu lar form but
[H 4 S i0 4] = IO 2 7 4 m o l/L (9.89) represents H 4 S i0 4 to which it can be related by equation
9.84. Therefore, the concen tration of S i0 2 stated in
The equilibrium (equation 9.84) is independent of the pH;
milligrams per liter can be recalculated as H 4 S i0 4 in
therefore, the activity of [H 4 Si 4] in a sat urated
moles per liter by the
solution of silicic acid in contact w ith solid am orphous S
following consideration. If S i0 2 = 25.0 m g/L , its
i0 2 is constant and varies only with tem perature. The
m olar concentration is 25.0/60.08 m m ol/L or
dissociation constants of silicic acid are all very small,
0.416 X IO- 3 m ol/L , w here 60.08 is the m olecular
making this a very weak acid. We can calculate the degree
weight of S i0 2. E quation 9.84 indicates that for
of dissociation
every m ole of am orphous S i0 2 that dissolves
of silicic acid at pH = 7.0 by proceeding stepwise
1 m ol of H 4 S i0 4 is produced. Therefore, since
from equation 9.85: 3
0.416 X 10 mol of S i0 2 dissolved to give a con
[H +][H 3 SiQ4 ] IO 9 7 1 (9.90) centration of 25.0 m g/L , the concentration of
=
H 4 S i0 4 in that solution is also 0.416 X 10- 3 or
[H4 S i0 4]
4.16 X 10~ 4 m ol/L .
[H 3 S i0 4] = ic r 971 x n r 2-74
The concentration of SiO z in Figure 9.4 is
io- p lotted on a linear rath er than a logarithm ic scale
= IO" 5 4 5 m o l/L (9.91) in o rd er to em phasize th at the solubility of am or
phous S i0 2 at 25 C increases steeply at pH > 8 because of
1(T5-45 X 102
the dissociation of H 4 S i0 4 to H 3 S i0 4 , which introduces
2 -7 4
IO additional Si-bearing ions into the w ater. T herefore, even
= 0.19% (9.92) a small decrease in the pH of basic solutions containing H
4 S i0 4 may reduce the solubility of S i0 2 enough to
Therefore, the activity of H 3 S i0 4 at pH = 7.0 is
saturate the solution and to cause am orphous S i0 2 to be
about 500 tim es less than that of H 4 Si 4 and does deposited . For exam ple, a decrease of 0.1 pH units from
not contribute appreciably to the solubility of 8.5 to 8.4 of a saturated solution can result in the
am orphous S i0 2. deposition of 1.37 mg of am orphous SiO z p er liter of
A t pH = 8.0 the activity of H 3 S i 0 4 increas solution. Such seem ingly sm all am ounts quickly increase
es to 10 4 4 5 m o l/L (equation 9.91) and D x rises w hen we apply them to th e large volum e of groundw ater
to 1.95% . A t pH = 9.0 [H 3 S i 0 4] increases to that m ay flow through an aquifer in the course of geologic
10- 3 4 5 m o l/L and D x reaches 19.5%. We see that the time. Subtle changes in the geochem ical environm ent can
degree of dissociation of silicic acid is very low at pH < be m agnified by tim e into large-scale tran s form ations.
8.0 and rises rapidly as the pH increases above th at value.
A s a result, the solubility of am or phous silica also rises
because it is based on the sum

pH
Figure 9.4 Solubility of amorphous SiO, expressed in
units of mg of SiO, per liter of pure water at 25 C based
on equations 9.84 and 9.85. Note that H4S i0 4 begins to
dissociate appreciably at about pH = 8.0 and that HjSiO^
becomes dominant at pH = 9.71.The sec ond dissociation
(equation 9.85) is negligible in the pH range shown here
and, at pH = 10.0, it increases the solubility of amorphous
SiO, by only 0.03%. Silicic acid is derived primarily by
chemical weathering of the common rock-forming Al-
silicate and silicate min erals, but may be precipitated
from saturated solutions as gelatinous amorphous silica
when the pH decreases. As amorphous silica ages, it
recrystallizes to form opal A and opal CT until it
ultimately forms chalcedony, which is a variety of quartz.
Surprisingly, quartz is not an im portant source of
silicic acid in natural solution because it is quite
insoluble.The equilibrium constant of the reaction:
S i0 2 + 2 H 20 : H 4 SiQ 4 (9.93)
Q uartz
is K = 10 . As a result, the activity of H 4 S i0 4 in
a saturated solution in contact with quartz at 25 C is 9.76 X
lO- ^ m ol/L ; which is equivalent to T5-9 m g/L of S i0 2
in solutiotf/The concentrations of Si in average river water
and seawater (Table 4.7) are 6.5 and 2.8 /xg/g,
respectively.These concentra tions can be transform ed into
SiOz in milligrams
9.7 SUMMARY 127
per liter by converting first to micromoles of Si and then
to m icrograms of SiO,. Fof average river water:
6.5 X 60.08: x IQ3
SiO , = = 13.9 m g /L (9.94)
28.086 X 103
Therefore, average river w ater is supersaturated
with respect to quartz but is undersaturated with respect to
am orphous silica. However, quartz does not precipitate
from aqueous solutions directly but forms only by
recrystallization of am orphous S i0 2 as discussed
previously. The con centration of SiO , in seaw ater is 5.8
m g/L (assuming a density of 1.025 g/cm 3), which is
close to the solubility of quartz at 25 C. Seaw ater
has a lower concentration of S i0 2 than river
w ater because certain organisms (sponges,
diatoms, and radiolarians) form skeletons com
posed of opal A . W hen these organisms die, their
siliceous skeletons accum ulate on the bottom of the ocean
and may ultim ately form either chert, com posed of
cryptocrystalline quartz, or deposits of diatom ite (Cressm
an, 1962; W edepohl, 1972).
The geochemistry of silica illustrates the point that
some reactions in nature are essentially irre versible. W
hen NaCl dissolves in w ater the reaction is reversible
because the com pound does precipi tate from
supersaturated solutions. W hen am or phous silica
dissolves in water, it can be made to precipitate again,
although with some difficulty. However, when quartz
dissolves in w ater at 25 C to form silicic acid, the
reaction cannot be reversed at the same tem perature. We
will encounter this phe nom enon again in Chapter 10
when we discuss the solubility of minerals, some of which
form only by crystallization of magma at elevated tem
peratures.
9.7 Summary
Chem ical reactions have a natural tendency to achieve a
state of equilibrium in which the rates of the forw ard and
backw ard reactions are equal. As a result, the am ounts of
reactants and products that coexist at equilibrium are
invariant with time. This observational evidence was
originally used to
128 ACIDS AND BASES

form ulate the Law of Mass A ction, which was later
derived from the principles of therm odynam ics.
Chem ical reactions in the natural environ
m ent may fail to reach equilibrium because the
products escape from the site of the reaction or because
reactions involving the transform ation of one solid into
another are very slow. In addition, som e natural reactions
are irreversible.
Acids and bases are defined as com pounds th at
release H + or O H - ions in aqueous solutions,
respectively. The bases of som e elem ents are am photeric
and act like bases in acidic environ m ents and like acids
in basic environm ents. W eak acids and bases dissociate
only partially depend ing on the p H of the environm ent.
In general, the p H of natural environm ents results from
the com
bined effect of all chemical reactions that are tak ing place
simultaneously.
Som e bases are insoluble and actually occur as m
inerals in nature. For exam ple, M g(O H ) 2 form s the m
ineral brucite. The solubility of such insoluble bases can
be calculated by solving a set of sim ultaneous equations
derived from the Law of Mass A ction and from the
requirem ent of con servation of m ass and electrical
charge.
W eak acids play an im portant role in geo
chem istry because the abundances of the ions
they form by dissociation are controlled by the pH
of the environm ent. This phenom enon affects the
solubility of am orphous silica, which increases
rapidly at p H > 8 because of the dissociation of
silicic acid.

I LE R
Problems
 1. If the concentration of an ion in a solution is 5.0 X 10-2 bottle. Assay: 98.0%; net weight: 4.08 kg; molecular
m ol/L at 25 C, what is its concentration in the same weight: 98.08; specific gravity: 1.84. (Answer 36.8)
solution at 45 C? Use the data in Table 9.1.
5. What volume of 35 N (normal) H2S 0 4 is required
(Answer: 4.96 x 10-2 mol/L)
to make 250 mL of 1.5 N H2S 0 4?
2. Derive a relationship between the molarity (M), the
6. Calculate the pH of hydrofluoric acid containing
formality (F), and the density (d) of a solution.
0.1 mol of HF per liter of solution. Find the dissociation
3. Calculate the normality of hydrochloric acid (HC1) constant in Table 9.3.
from the following information provided by
7. Calculate the activities of all ions and the pH of a
the label on the bottle. Assay: 37%; net weight: 2.72 kg;
solution containing 0.1 mol of phosphoric acid per liter of
molecular weight: 36.46; specific gravity: 1.18.
(Answer: 12.0) solution. Find the equilibrium constants in Table 9.3.
8. Calculate the solubility of gibbsite (Al(OH)3) at pH
4. Calculate the normality of sulfuric acid (H2S 0 4)
= 5.0. Find the dissociation constants in Table 9.3.
based on the information provided by the label on the
References
C R E S S M A N , E. R., 1962. , R. K., 1979. The
Nondetrital siliceous Chemistry o f Silica.
sediments. Wiley, New York,
Chapter T, 1-23. In M. 866 pp.
Fleischer (Ed.), Data o f K S T N E R , M.. J. B. K EEN
Geochemistry, U.S. Geol. E , and J.
M. G IESKES ,
Surv. Paper 440-T.
1977.
D R E V E R , J. I., 1988. The
Diagenesis of siliceous
Geochemistry o f Natural
oozes I. Chemical
Waters, 2nd ed. Prentice-
controls on the rate of opal
Hall, Upper Saddle River,
A and opal CT
NJ, 435 pp.
transformation: an experi
G A R R E L S , R. M ., and C. L . C H R
mental study. Geochim.
IS T , 1965. Minerals,
Cosmochim. Acta, 41:1041-
Solutions and Equilibria.
1060.
Harper & Row, New York,
450 pp.
 REFERENCES 129
K R A U S K O P F , K . B., 1979. Introduction to Geochemistry, 2nd ed. W E A S T , R. C., M. J. A S T L E , and W. H. B E Y E R (Eds.), 1986. CRC
McGraw-Hill, New York, 617 pp. Handbook o f Chemistry and Physics, 66th ed. CRC Press, Boca
L IN D S A Y , W. L ., 1979. Chemical Equilibria in Soils. Wiley, New Raton, FL.
York, 449 pp. W EDE POHL , K. H. (Ed,), 1972. Silicon. Handbook of
M O O R E , W. J., 1955. Physical Chemistry, 2nd ed. Prentice-Hall, Geochemistry, vol. II-2. Springer-Verlag, Berlin.
Upper Saddle River, NJ, 633 pp.
ST U M M , W., a n d J. J. M O R G A N , 1970. Aquatic Chemistry. An
Introduction Emphasizing Chemical Equilibria in Natural
Waters. Wiley, New York, 583 pp.
 10
Salts and Their Ions
W hen an acid and a base are m ixed, the hydro gen ion of sulfuric acid and alum inum hydroxide, is quite soluble in
the acid com bines with the hydroxyl ion of the base to w ater and dissociates readily into ions:
form w ater. The rem aining anion of the acid and the
A l2 (S 0 4 )3 ( s ) ^ 2 A l3+ + 3 SO 2- (10.1)
cation of the base form a salt, which may precipitate or
rem ain in solu tion depending on its solubility. Salts are We assum e for the tim e being that these ions do
nam ed n o t interact with w ater, that this reaction is the
after the acids th at provided the anions. For
only source of A l3+ and S 0 4_ ions, that no other
exam ple, hydrochloric acid form s chlorides, sul
furic acid form s sulfates, carbonic acid form s car
ions are present, and that the w ater tem perature
bonates, and silicic acid form s silicates. The naturally is 25 C. We see from equation 10.1 that each
occurring salts of these and other acids are m inerals th a t m ole of alum inum sulfate that dissolves p ro
form the rocks of the crust of the E arth . The only m duces two moles of A l3+ and three m oles of S 0 4_ in
inerals th at are not salts are the oxides, hydroxides, and the solution. Therefore, if x is the num ber of m oles of
native elements. salt that dissolve in a saturated solution, the
Evidently, the study of salts and their properties concentrations of the resulting ions will be
is an im portant subject in geochem istry. In this (A l3+) = 2x and ( S 0 4~) = 3x. A ccording to the
chapter we will study the solubility of salts in Law of Mass A ction (equation 9.13):
w ater and the interactions of th e resulting ions [A13 +]2 [S 0 4 - ] 3 = K sp (10.2)
w ith m olecules of water.
w here K sp is the solubility product constant, and
the activity of excess solid alum inum sulfate is equal to
10.1 Solubility of Salts one. If the concentrations of the ions
W hen salts dissolve in w ater, they dissociate into the are equal to the activities in the solution,
anions and cations of the acid and the base from which [Al3+] = 2x and [SO2 -] = 3x. Substituting into
the equation 1 0 . 2 we obtain:
they w ere derived. N eutral m ole cules of salts m ay also
(2x)2(3 x f = K sp (10.3)
form , but their abundance is low in m ost cases. W hen a
salt has form ed a sat urated solution, a state of w here K sp = 69.19 (calculated from its solubility
equilibrium exists betw een the ions in the solution and
given by W east et al., 1986).The solubility product
any excess salt rem aining in the solid state. Therefore, the
constant is a large num ber, indicating th at alu
Law of Mass A ction applies and can be used to calculate
m inum sulfate is, in fact, a very soluble salt. For
the activities of the ions in a saturated solution and hence
this reason, it does not occur naturally as a m iner al.
the solubility of the salt.
Equation 10.3 reduces to 108x5 = 69.19, which yields x =
We choose alum inum sulfate [A12 (S 0 4)3] to 0.9147. Therefore, a saturated solution of alum inum
illustrate how the Law of M ass A ction can be used for sulfate contains 2x = 1.829 m o l/L of A l3+ and 3x =
2.744 m ol/L of SO;;- . The A l3+ reacts with w ater to
this purpose. A lum inum sulfate, a salt of
form insoluble A l hydroxide.
130
 10.1 SO LU BILITY OF SALTS 131

The am ount of alum inum sulfate that dissolves in a
saturated solution is equal to x = 0.9147 m ol/L . Since the
m olecular weight of aluminum sulfate is
342.1478, its solubility is 0.9147 X 342.1478 =
3.13 x 102 g /L of water.
M ost m inerals are m uch less soluble in w ater than
alum inum sulfate, and their solubility p ro d uct constants
have correspondingly sm aller val ues. Table 10.1 contains
the solubility product constants of som e com m on m
inerals taken p ri
m arily from a com pilation by K rauskopf (1979)
but augm ented with data from Lindsay (1979).
These constants can be used to calculate the activ ities of
the ions in saturated solutions of m inerals by m eans of
the Law of Mass Action.
We are also interested in finding out w hether a
solution is actually saturated with respect to a specific m
ineral. For example, if a sample of w ater contains 5.00 X
10- 2 m o l/L of Ca2+ and 7.00 x 10 - 3 m o l/L of S 0 4~, we
may w ant to know w hether this solution is saturated with
respect to calcium sulfate (anhydrite). In order to find out,
we first write the equation to represent the disso ciation of
C a S 0 4 into its ions:
C a S 0 4 ^ C a 2+ + SO2- (10.4)
constant, the solution is undersaturated and th ere fore can
dissolve the m ineral. If IA P = K , the solution is saturated
and a state of equilibrium exists betw een the solid and its
ions.
In the case we are considering, it turns out
that C a S 0 4 does not actually precipitate from a
supersaturated solution of its ions. Instead, g yp
sum (C a S 0 4 2H 2 0 ) form s and subsequently
crystallizes to anhydrite. This is another exam ple
of the irreversibility of som e geochem ical reac
tions we first m entioned in Section 9.6. In this
case, C a S 0 4 dissolves in w ater but does no t p re
cipitate at 25 C. Similarly, m ost silicate m inerals
dissolve in w ater b u t do n o t precipitate from
aqueous solutions o f their ions at low tem p era
ture. Som e silifcate m inerals precipitate from
aqueous solutions only at elevated tem peratures or require
a specific starting m aterial like volcanic glass, and the
silicate m inerals of igneous rocks crystallize only from
silicate m elts at tem p era tures close to 1000 C.
In the case of the calcium sulfate solution we are
considering, gypsum precipitates until the activities of the
ions are reduced sufficiently so that the IA P is equal to K
sp and solid gypsum is in equilibrium with its ions:

A t equilibrium the Law of Mass A ction applies; thus: C a S 0 4 -2 H 2 0 ( s ) ^ C a 2+ + S 0 A~2 + 2H 20

(10.7)
[Ca2 +][S 0 2 -] = IO- 4 5 (10.5) Therefore, at equilibrium betw een gypsum and

If the solution is saturated with respect to anhy drite, the its ions:
product of activities of the ions in the solution m ust equal [Ca2 +][S 0 2_] = IO- 4 6 (10.8)
10-4,5. We assume for the sake of argum ent th at the m
easured concentrations If x m oles of gypsum precipitate, the activities of the ions
are equal to the activities of the ions and th ere at equilibrium will be:
fore calculate the ion activity product (IA P):
5
[Ca2+] = 5.00 X IO- 2 - x (10.9)
(5.00 X 10- 2 )(7.00 X 10~3) = 35.00 X 10 3

[SO2 -] = 7.00 X 10 - x (10.10)

= IO- 3 4 5 (1 0 .6 ) Substituting into equation 10.8 yields:
(5.00 X 10- 2 - x)(7.00 X 10 3 - x) = IO- 4 6
Since the IA P (IO-345) in this case is larger than
the solubility product constant, the solution is (10.11)
supersaturated with respect to calcium sulfate. Therefore, which reduces to a quadratic equation:
anhydrite should precipitate. If the IA P is found to be 4

x2 - (5.7 X 10-2)x + 3.25 X 10- = 0 (10.12)

less than the solubility product
132 SALTS AND THEIR IONS

Table 10.1 Solubility Product Constants for Different Salts at 25 C

p/0 'b
Carbonates

Ag2C 0 3 11.09 (L)

B aC 03 witherite 8.3 (K)c
C aC 03 calcite 8.35 (K)
CaC03 aragonite 8.22 (K)
C aC 03 6HzO ikaite 6.38 (L)
CdC03 otavite 13.7 (K), 12.0 (L)
C oC 03 sphaerocobaltite 10.0 (K)
C uC 03 9.63 (L)
Cu2(0 H )2C 0 3 malachite 33.8 (K), 33.2 (L)
Cu3(0 H )2(C 0 3)2 azurite 66.5 (L)
FeC 03 siderite 10.7 (K), 10.2 (L)
HgCOj 22.6 (L)
Hg2C 0 3 [Hg^+][C 03- ] 14.0 (L)
M gC 03 magnesite 7.5 (K),7.46 (L)
M gC 03 3H20 nesquehonite 5.6 (K),4.67 (L)
-M gC 03 5H20 lansfordite 4.54 (L)
M gCa(C03)2 dolomite 17.9 (L)
M nC03 rhodochrosite 9.3 (K), 10.1 (L)
N iC 03 cerussite 6.9 (K)
PbC 03 13.1 (K), 13.5 (L)
Pb2C 0 3Cl2 phosgenite 20.0 (L)
Pb3(C 0 3)2(0 H )2 46.8 (L)
SrC03 strontianite 9.0 (K)c
uo2co3 [uoHtcor] 10.6 (K)
Z nC 03 smithsonite 10.0 (20C, K), 10.2 (L)
Sulfates

Ag2S 0 4 4.81 (K, L)

k a i3( s o 4)2( o h )6 alunite 80.95 (L)
B aS04 barite 10.0 (K)
C aS04 anhydrite 4.5 (K),4.41 (L)
C aS04 2HzO gypsum 4.6 (K),4.63 (L)
C dS04 0.044 (L)
C dS04 2H20 1.59 (L)
C uS04 -3 .72 (L)
C uS04 5H20 2.61 (L)
Cu4(0 H )6S 0 4 brochantite 68.6 (L)
F eS04 -2 .65 (L)
F eS04 7H20 2.46 (L)
Fe2(S 0 4)3 -2 .89 (L)
KFe3(S 0 4)2(0 H )6 jarosite 96.5 (L)
HgSO, 3.34 (L)
Hg2S 0 4 6.20 (L)
MgSO, -8 .18 (L)
M nS04 -3 .43 (L)
M nS04 H20 -0 .47 (L)
 10.1 SO LUBILITY OF SALTS 133
Table 10.1 (co ntinued)

p K*

Mn2( S 0 4)3 11.6 (L)

PbS04 anglesite 7.8 (K), 7.79 (L)
S rS 0 4 celestite 6.5 (K)
ZnS04 zinkosite 3.41 (L)

Sulfides
a- e S a ip o 4 A1P04
Ag S 2H20
2S y H6K3A15(P 0 4)S -18H20
-A F - C aH P04 2HzO C aH
g e P04 a~Ca3(P 0 4)2 /3-
N
2S S Ca3(P 0 4)2
i
Ag2 2
S
S
B F
e P
i
S b
2
2 S
S
3
F S
e b
C
2 2
a S
S S
3 3
C a
d - S
S H n
g S
C S
o -K gS a
S H -
g Z
( 2 n
a S S
) -ZnS
M
C g P
o S h
S o
( M s
/ n p
3 S h
) a
M t
C n e
u S s
2
A
S M g
n
C 3
S
u P
2
S
a_^ 0
M
e0.9 o
5^ 4
S
a
2
-
F
e a
S -
N
F i
 galena
stibnite

bismuthinitesphalerite
oldhamite wurtzite
greenockite

chalcocite berlinite
covellite variscite
pyrrhotite K-taranakite
troilite brushite
troilite monetite
pyrite
marcasite whitlockite

cinnabar (red)
metacinnabar (black)
[Hg2+][S2-]

(green)
(pink) alabandite
haurite
molybdenite
134 SALTS AND THEIR IONS
Table 10.1 (continued)

p tfab

Ca8H2(P 0 4)6 5HzO octacalcium phosphate 80.6 (L)

Ca10F2(PO4)6 fluorapatite 119.2 (L)
Ca10(OH)2(PO4)6 hydroxyapatite 116.4 (L)
Cd3(P 0 4)2 38.1 (L)
Cu3(P 0 4)2 36.9 (K, L)
Cu(P04)2 2HjO 38.8 (L)
FeP04 (amorph.) 21.6 (K)
F eP04 2HzO strengite 26.4 (K,L)
Fe3(P 0 4)2 8HzO vivianite 36.0 (L)
Hg2H P 04 12.4 (L)
k h 2p o 4 0.21 (L)
k 2h p o 4 -3 .60 (L)
M gHP04 3H20 newberyite 5.82 (L)
MgKP04 6H ,0 10.6 (L)
MgNH4P 0 4 6H20 struvite 13.2 (L)
Mg3(P 0 4)2 25.2 (K), 14.6 (L)
Mg3(P 04)2 8HzO bobierrite 25.0 (L)
Mg3(P 04)2 22H20 23.1 (L)
M nHP04 12.9 (L)
Mn3(P 0 4)2 27.4 (L)
PbHP04 11.4 (K), 11.5 (L)
Pb(H2P 0 4)2 114.7 (L)
Pb3(P 0 4)2 43.5 at 38C (K),44.4 (L)
Pb,(P04)3Br bromopyromorphite 78.2 (L)
Pb5(P 0 4)3Cl chloropyromorphite 83.7 (L)
Pb5(P 0 4)3F fluoropyromorphite 71.6 (L)
Pb5(P 0 4)30 H hydroxypyromorphite 76.8 (L)
( u o 2)3(p o 4)2 [U O f ]3[PO^]2 49.7 (K)
u o 2h p o 4 [ u o 2+][h p o J _] 12.2 (K)
Zn3(P 0 4)2 4HzO hopeite 35.3 (L)
Zn3(P 0 4)2 35.3 (K)
Molybdates

Ag2M o04 11.55 (L)

CaM o04 powellite 7.94 (L)
CUMO04 6.48 (L)
FeM o04 7.70 (L)
MgM o04 0.62 (L)
MnM o04 4.13 (L)
PbM o04 wulfenite 16.0 (L)
ZnM o04 4.49 (L)
Chlorides

AgCl cerargyrite 9.75 (K, L)

CuCI 6.7 (K)
PbCl2 cotunnite 4.8 (K)
Hg2Cl2 [Hg2+][c r ]2 17.9 (K)
 10.1 SOLUBILITY OF SALTS 135
Table 10.1 (continued)

p iC-b

Fluorides
AgF 100.0 (L)
BaF, 5.8 (K)
CaF2 fluorite 10.4 (K)
MgFz sellaite 8.2 (K)c
PbF, 7.5 (K)
SrF2 8.5 (K)

Expressed as log K sp.

bK = Krauskopf (1979), pp. 552-553; L = Lindsay (1979), calculated from standard free energies.
Uncertain by more than 0.8 (Krauskopf, 1979).
 C a2 +/SC>4 _ ratio was operate when other salts
greater than one. If C a2 +/ precipitate from isolated
We solve this equation by S 0 2 < 1.0, precipitation volumes of w ater
m eans of the quadratic of gypsum causes undergoing evaporative
form ula (equation 9.18) depletion in Ca2+ and concentration in a desert
x , = 5.05 X 10~ 2 and enrichm ent in sulfate. lake (H ardie and Eugster,
choose x 2 because x, Therefore, the value of the 1970).
reduces [SO 4 - ] in m olar Ca2 +/ S 0 2_ ratio
equation 10.10 to a of a solution from which
negative value. Therefore, gypsum precipitates deter-
the am ount of gypsum th
at will precipitate p er
liter of solu tion is 6.45 X
10~ 3 x 172.17 = 1.11 g.
W hen equi librium has Gypsum
been established after Precipitated,
gypsum has moles per
precipitated, the IA P (5.00 0.645) liter
(7.00 - 6.45) x 10- 3 = IO-46, as required Figure 10.1 Changes in
equilibrium . N ote that the molar Ca2+/S 0 2_
has increased by m ore than a factor of ratio of a solution from
initial value of (5.00 X which gypsum is
10_2 )/(7 .00 X 10-3) = 7.1 precipitating based on
to (4.35 X 10~2)/(0.055 X equations 10.9 and 10.10.
dently, the precipitation of a salt from a supersat In this example, the solu
urated solution not only tion contains more Ca2+
reduces the concentrations than SO2 -, and its
Ca2+/S 0 2' ratio
of the ions but actually
therefore increases as
changes the chemical com
equal amounts of Ca2+
position of the remaining and SO2- are removed
solution. from it. As a result, the
If gypsum continues solution becomes
to precipitate from this enriched in Ca2+ and
solution, the activity of depleted in SO2 -. If the
SO 2 ~ approaches zero solution initially contains
and the [Ca2 +] /[ S 0 2_] more SO2- than Ca2+,it
ratio rises toward infinity. becomes enriched in SOj~
and depleted in Ca2+ as a
This phenom enon, which
result of progressive
is illustrated in Figure
gypsum precipitation.
10.1, occurs during the This phe nomenon
evolution of brines by constitutes a geochemical
evapora tive concentration divide that affects the
and constitutes a chemical evolution of
geochemical divide. The brines by progressive
solution we have been evapo rative
discussing is depleted in concentration.
sulfate and is enriched in
Ca2+ by the progressive mines w hether the
precipitation of gypsum solution becomes enriched
because its ini tial m olar in Ca2+ or SO4. Similar
geochemical divides
136 SALTS AND THEIR IONS

N atural w ater samples are m uch m ore com plex w here y 2 is the activity coefficient of Ca2+ and Ba2+
than the solution we have been discussing because they defined by equation 9.12. By substituting into equation
contain a m ixture of cations and anions and therefore may 10.18, setting y 2 = 1 .0 , and drop ping (H +) and (O H
precipitate several dif ferent salts. For example, a solution ), we obtain:
m ay be satu rated with respect to both gypsum and
barium sulfate (barite). In this case we have two sim ulta 2X10- + 2>O0^ = 2(son ( ia 2 1 )
[SO2 ]
neous equilibria with a com m on ion: [so2-]
If we stipulate that [S 0 4 ] = (SO 2 ), equation
C a S 0 4 2H 2 O ^ C a 2+ + SO 2 + 2H zO (10.13) 10 .21
reduces to:
B a S 0 4 ^ B a 2+ + SO2 (10.14) 10 -4.6 - 10.0 _ (10.22)
1 0

Both reactions contribute S 0 4 ions to the solu = (SO 2 ) 2

Evidently, the barite contributes a negligible
tion, but the resulting activities must be the same for both am ount of sulfate to the solution and [S 0 4] =
equilibria. We can calculate the [Ca2 +]/[B a2+] ratio of a ( S O D = 10 23 m ol/L -
solution that is saturated with respect to both gypsum and The results of this calculation indicate that
barite by applying the Law of M ass A ction to equations gypsum is able to force barite to precipitate when a
10.13 and 10.14: saturated solution of barite comes in contact with gypsum.
[Ca2 +][S 0 2 ] = IO 4 6 (10.15) In other words, barite can replace gypsum or anhydrite
[Ba2 +][S 0 4 j = (10.16) because barite is less soluble than gyp sum or anhydrite.
10 - 10.0 The calcium sulfate can dissolve in a solution that is
already saturated with respect to barite and thereby it can
Since [S 0 4 ] has the sam e value in both equilibria, we
increase the sulfate concentration. As a result, the solution
elim inate it by substituting [SO2 ] = 10 4 6 /[C a 2+] into
becomes supersaturated with respect to barite. Therefore,
equation 10.16:
barite precipitates as gypsum or anhydrite dis solves until
[Ba2+] X IO " 46 the [Ca2 +]/[B a2+] ratio of the solution approaches
= IO 1 0 0 (10.17)
[Ca2+] 250,000 which is required for equilibri um . W hen that
value is reached, the replacem ent of gypsum or anhydrite
from which it follows that [Ba 2 +]/[C a2+] = 10 5 4 or [Ca2 by barite stops.
+]/[B a2+] = 2.5 X 105. Evidently, the activity of Ca2+ is
250,000 times that of Ba2+ w hen the solution is saturated R eplacem ent of one m ineral by another is a com m
with respect to both gypsum and barite. on phenom enon in geology and affects not only sulfates
We can determ ine th e[S 0 2'] concentration of such but also sulfide minerals, carbonates, and other types of
a solution from the requirem ent for elec trical neutrality: compounds. In each case, the introduction of a common
ion causes the solution to becom e supersaturated with
respect to the less soluble com pound. The effective ion m
2(B a2+) + 2(Ca2+) + (H +) = 2(SO^ ) + (O H ) (10.18) ay be the anion, as we have seen, or the cation itself. For
example, if BaCl2, which is quite soluble, is added to a
From equations 10.15 and 10.16 we have, in term s saturated solution of B a S 0 4, the sulfate p re cipitates
until m ost of it is rem oved from the solu tion. In this case
of concentrations:
barite replaces BaCl2. The process always involves the
(10.19)
10-4.6 replacem ent of the m ore soluble com pound by the less
(Ca2+) = 4 ~] soluble one.
and:
Solutions in nature may become supersaturat ed with
- 10.0
respect to a specific com pound in several different ways,
1 0
(10.20) including (1 ) the introduction of a
(Ba2+) =
y2[so2]

com m on ion, as just described; (2) a change in the pH ,
exemplified in C hapter 9 by the precipitation of am
orphous silica; (3) evaporative concentration of w ater in
closed basins and in soil or sedim ent drying in the sun;
and (4) tem perature variations.
The solubility of com pounds does depend on
the tem perature, but not always as we might
expect. In general, solubilities increase with
increasing tem perature, but for som e com pounds the
opposite is true. For example, the solubility of
C a C 0 3 and other carbonates decreases with
increasing tem perature because the solubility
product constants of calcite and aragonite actually
decrease with increasing tem perature. We will take up the
tem perature dependence of the solu bilities of m inerals in
C hapter 11 after we learn to calculate equilibrium
constants from therm ody nam ic data.
10.2 Hydrolysis
W eak acids are only partly dissociated in w ater because
the anion of the acid has a strong affini ty for hydrogen
ions. W hen a salt of a weak acid is dissolved in water, it
releases anions into solu tion that have the same strong
affinity for hydro gen ions. Therefore, these anions im m
ediately bond to hydrogen ions and thereby form the
parental acid. The hydrogen ions that bond with anions
may already exist in the environm ent, or they may
originate from the dissociation of water. In either case,
when the salt of a weak acid is dissolved in water, the rem
oval of H + causes the solution to becom e m ore basic.
The process we have just described is hydrol ysis. It
is defined as the interaction betw een w ater
and one or both ions of a salt th at results in the
form ation of the parental acid or base, or both.
We can classify salts by the strength of the acid and base
from which they form:
1. Strong acid + strong base.
2. Strong acid + weak base.
3. W eak acid + strong base.
4. W eak acid + weak base.
10.2 HYDROLYSIS 137
Salts derived from strong acids and strong bases (category
1), such as NaCl or C a S 0 4, do not hydrolyze because
neither the anion nor the cation has a particular affinity for
H + or for O H - . Salts of strong acids and weak bases
(category 2), such as FeCl3 or C u S 0 4, release cations
into solu tion that com bine with O H - to form the parental
base. The rem oval of O H - causes the solution to becom e
m ore acidic. Salts of w eak acids and strong bases
(category 3) release anions that hydrolyze and m ake the
solution m ore basic. Salts of weak acids and weak bases
(category 4), like CuC O s or FeS2, release anions and
cations both of which hydrolyze. In this case, the effect on
the pH of a neutral solution must be determ ined by calcu
lation because it depends on a com parison of the strengths
of the parental acid and base.
H ydrolysis is a very im portant process because m
ost of the com m on rock-form ing m in erals of the crust
of the E arth are salts of weak acids and strong bases. For
exam ple, the carbon ates and silicates of the alkali m etals
and alkaline earths are salts of this type. T herefore, the
anions
of these m inerals split w ater m olecules in order
to form the w eak parental acid and thereby
cause the hydroxyl ion concentration of the solu tion to
increase. This is why groundw ater in car bonate aquifers
is commonly basic. Even silicate minerals, such as the
feldspars, hydrolyze w hen they interact with w ater by
releasing N a+ or K + from the surfaces of m ineral grains
and replacing it w ith H +.
The dissociation of salts into ions and subse quent
hydrolysis of the ions takes place very rapid ly. As a
result, these kinds of reactions are usually at equilibrium .
Therefore, if we dissolve a small quan tity of potassium
carbonate in pure w ater at 25 C, it will dissolve
completely and dissociate into ions:
K2C 0 3 2K + + C O 2- (10.23)
The carbonate ions react with w ater to form bicarbonate:
co2- + h 2o ^ h c o j + O H " a:h1 (1 0 .2 4 )
The bicarbonate ion ( H C 0 3) is itself an acid because it
can release H + into solution. H ow ever, it also
hydrolyzes:
138 SALTS AND THEIR IONS

HCOJ+H20^H 2C03 +OH Km (10.25) K2C03 in one liter of pure w ater at 25 C. In addi
2
tion, we specify th at C 0 gas does not escape from
By applying the Law of M ass A ction to reaction 10.24
the solution. Since i^ A 1 = 10- 6 , 3 5 and K A2 = 10- 1 0 3
we have: (Table 9.3) and K w = 10-14, we find that
[H C O j][O H - ] /CH] = 10- 3 7 and K m = 10 - 7 65. Since 7CH 1 is n ear
(10.26)
= Ku ly 9000 tim es larger than K m , the first reaction is m uch
[C O M m ore productive than the second. We th ere fore attem pt
and from equation 10.25: a stepwise solution. If x m oles of the carbonate ion
[h2co3][oh-] (10.27) hydrolyze in the first step, we have at equilibrium :

= K,H2
[H C O j] [CO3 -] = 0.1 - x
To find the value of the first hydrolysis constant (7fH1) we [H C O j] = [O H - ] =
replace [O H - ] in equation 10.26 by: Substituting into equation 10.26 yields:

[O H - ] = (10.28)
[H +]
= io -3-7 (10.35)
w here K w is the dissociation constant of water. 0.1
Therefore, from equation 10.26: If x 0 .1 , 0 . 1 x ~ 0 . 1 and, from equation
10.35, x = 10-235, which is only 4.5% of 0.1.
[H C O 3 K (10.29)
K, Therefore, the approxim ation that x 0.1 is jus
[ c o r ] [ n +] tified. The second hydrolysis step produces a neg ligible
Since the dissociation of carbonic acid proceeds stepwise: am ount of additional hydroxyl ion equal to [O H - ] = 10 - 7 6
5 m ol/L . The pH of a solution

obtained by dissolving 0.1 m ole of K 2 C 0 3 in one

H 2 C 0 3 ^ H C 0 3 + H +K (10.30) liter of pure w ater at T = 25 C is:
Al
H C O J ^ C O 2- + H + (10.31) pH = 14 - 2.35 = 11.65 (10.36)

we see that in equation 10.29:

Clearly, the solution is quite basic, as predicted. The
problem can also be solved by m eans of a
[H C O j] (10.32)
set of independent equations th at arise from the
sim ultaneous equilibria in the solution, from the
[COM[H+] K A2
requirem ent of electrical neutrality, and from the
By substituting equation 10.32 into equation 10.29 we stipulation th at the total am ount of m olecular and ionic
obtain: carbonate species in the solution is 0 . 1 mol. We could have
set ourselves a m ore general pro b lem by allowing the
solution to be open to the
v ~ H1 (10.33)
atm osphere, perm itting the carbonate species in
A2 the solution to equilibrate with the C 0 2 of the
The sam e procedure yields:
atm osphere. We choose not to do so here because we will
K H2 (10.34)
treat carbonate equilibria in Section 10.4.
Salts derived from strong acids and weak bases also
A1 hydrolyze, and the necessary hydrolysis constants can be
from equation 10.27. derived as dem onstrated above. For example, C u (O H ) 2
We are now able to calculate the pH of a solu tion that is a w eak base th at disso ciates in two steps:
was prepared by dissolving 0 . 1 m ol of
10.3 ACTIVITIES AND CONCENTRATIONS 1 39

C u (O H )2(s ) ^ C u ( O H ) +
C u (O H )+( a q ) ^ C u 2+ + O H
C u(O H ) 2 form s the salt C u S 0
H 2 S 0 4. W hen C u S 0 4
is dissolved in w ater,
Cu2+ is released and
hydrolyzes, as shown:
Cu2+ + H 2 O ^ C u ( O H ) + + H +
C u (O H )+ + H 2O ^ C u ( O H ) 2(s)
w here K m = K j K B2 and
Therefore, we predict that
a solution of C u S 0 4 in
pure w ater becom es
acidic because of the
hydrol ysis of the cupric
ion (Cu2+).
Hydrolysis is an im
portant phenom enon
because it enhances the
solubility of salts form ed
from w eak acids and bases
and because it tends to
stabilize the pH of salt
solutions. Hydrolysis
affects the solubility of
salts by producing addi
tional ionic and m olecular species
m ore of the salt to dissolve. In
of salts derived from w eak
enhanced w hen they are
dissolved in strong acids
that provide an am ple
supply of hydrogen ions.
Therefore, all carbonates,
silicates, phosphates, and
sulfides are m ore soluble
in acids than in pure
w ater because of the affinity
hydrogen ions. The sam e
salts derived from weak
bases they form m ay
them selves be quite
insolu ble. For exam ple,
if we place ferric chloride
into a
basic solution, the Fe3+ ion reacts with charges. Similarly, the
form insoluble F e(O H )3: increases in accordance effect of a given suite of
FeC l 3 F e3+
with + 3 C
Le P C hteliers ions on a particular ion
principle. T herefore, the depends on the charge of
Fe3+ + 3 O H - F e(O H )
solution has effectively that ion and its radius,
Solutions th at contain stored hydrogen ions, and which together determ ine
they are released w hen a the charge den sity on the
th at acid are buffered
reaction occurs in this surface of the ion. In
changes in their pH . C onsider, for
solution th at con sum es addition, the molecules of
solution containing carbonic acid
H +. Similarly, if the solvent play a role
The carbonic acid is only
additional hydrogen ions because, if they have
partially dissociated, as
w ere to be released electrical polarity, they
indicated by equations also interact with both
would be consum ed by reacting
10.30 and 10.31. By anions and
bonate and bicarbonate ions to cations. All car
form m ore of
adding K 2 C 0 3, which these factors are included
bonic acid.
releases additional C O 2 - in the D ebye -H iickel
, the dissociation of the theory, on the basis on
acid is even m ore inhibit which the relationship
ed and the concentration betw een activity and
of m olecular H 2 C 0 3 10.3 Activities concentration of ions in a
and solution of elec trolytes
can be estim ated.
Concentr
ations
We have assum ed so far th The concentrations and charg
at the ions in elec trolyte solution are expressed by m eans
solutions do n o t interfere strength, defined as:
with each other. This can I = \^ m - z ]
only be true in very dilute
solutions because, in
where m, are the
general, oppositely
concentrations expressed
charged ions in a solution
attract each other. in moles and z-t are the
Therefore, cations in an charges of the ions. The
electrolyte solution are ionic strength of
surrounded by anions and
aqueous solutions on
vice versa. The result is th
at the ions of a p ar ticular the surface of
com pound are less able to
interact with each other
than expected from their
concentra tions. We
express this condition by
saying th at the activity of
the ions in electrolyte
solutions is less than their
concentration. It is
plausible to expect that the
interference by the other
ions increases with their
concentrations and
140 SALTS AND THEIR IONS
the E arth generally ranges from about 1 X IO- 3
For dilute solutions having ionic strengths
3
in rivers and lakes to 1 X 10 1 in old groundw a < 5 X 10~ the relationship is, to good
ter. Seaw ater has an ionic strength of 7 X 10- 1 approxim ation:
and oilfield brines and brine lakes m ay have ionic - lo g y = A z 2I l/2 (10.45)
strengths of 5 or m ore. In order to calculate the ionic w here A = 0.5085 for w ater at 25 C, z is the
strength of a w ater sample, we m ust have a com plete
chemical analysis, such as the analysis of w ater from the charge of the ion w hose activity coefficient is
M ississippi River at New Orleans in Table 10.2. The being calculated, and I is the ionic strength of the solution
concentrations of the ions are converted from parts per as defined by equation 10.43. N ote that the activity
million (ppm ) to moles per kilogram before they are coefficient of m olecules is equal to one because their
entered into equa charge is equal to zero.
tion 10.43. The result is I = 4.4 X IO-3. N ote that A m ore com plete statem ent, valid for / < 0.1, is
S i0 2 does not contribute to the ionic strength of given by:
this w ater because its charge is zero and th at H + and O H
- are om itted because their concentra tions are very low
(about IO- 7 m ol/kg) unless the w ater is highly acidic or
basic. N ote also that the w here A and B are constants w hose values depend on the
concentrations of H 2 C 0 3 and C O 3- w ere not dielectric constant of the solvent and the tem perature, and
reported in the analysis in Table 10.2 because a is the effective diam eter of the ion in the solution in A
H C O ] is the dom inant ion when the pH is ngstrom units. Values of A and B for w ater are listed in
betw een 6.35 and 10.3. Table 10.3 as a func tion of tem perature, and values of a
The ionic strength is used in the of several ions in aqueous solution are com piled in Table
D ebye -H iickel theory to estim ate values of the 10.4. E quation 10.46 was used to calculate the activity
activity coefficient y defined by the equation:
[activity] = y (concentration) (10.44)

Table 10 .3 List
of Values for Constants Used
Table 10.2 Chemical Composition of Water in the D ebye-H ckel Theory Expressed in
of the Mississippi River at New Orleans Equation 10.46
Concentration Temperature, C A B
0 0.4883 0.3241
Ion ppm m o l/k g

HCOj 116 1.90 x 103 5 0.4921 0.3249

10 0.4960 0.3258
s o \- 25.5 2.65 x 10~4 15 0.5000 0.3262
cr 10.3 2.90 x 10~4
20 0.5042 0.3273
N0 7 2.7 4.3 x 105 25 0.5085 0.3281
Ca2+ 34 8.5 X 10"4 30 0.5130 0.3290
Mg2+ 8.9 3.7 X 10^4 35 0.5175 0.3297
Na + 11.9 5.17 x 10"4 40 0.5221 0.3305
K+ 1.9a 4.9 x 10' 5 45 0.5271 0.3314
Fe2+ 0.14 2.5 x 10-6 50 0.5319 0.3321
SiO, 11.7 1.95 x IO4 55 0.5371 0.3329
Based on Na+/K + = 6.4 (concentration ratio) in average North 60 0.5425 0.3338
American river water (Livingstone, 1963,Table 81, p. G41). s o u r c e : Garrels a n d Christ (1965).
S O U R C E : Livingstone (1963), Table 15, entry H, p. G15.
 10.3 ACTIVITIES AND CONCENTRATIONS 141
Table 10 .4 V alues of Param eter a of Ion s in A q u eo us Solution

Ions a,

Rb+, Cs+,N H 4,T l+,A g+ 2.5

K+,C L ,B r ,r ,N O ; 3
OH , F,H S - , B rO J,IO ;,M n 04 3.5
N aLH C O j, H2POL HSOj, Hg2+, SO2 -, SeOf- , CrOj , HPO2 -, PO3- 4.0-4.5
Sr2+, Ba2+, Ra2+, Cd2+, Hg2+, S2 -, WO2- 5.0
Li+, Ca2+, Cu2+, Zn2+, Sn2+, Mn2+, Fe2+, Ni2+, Co2+ 6
Mg2+, Be2+ 8
H+, A l3+, Cr3+, REE3+ 9
Th4+,Zr4+,C e4+,Sn4+ 11
SOURCE: Garrels and Christ (1965).

coefficients for ions of different charges and electrolyte solutions can be attributed to crow d

hydrated radii shown in Figure 10.2. A third equa ing of the ions, which counteracts the interfer
tion developed by Davies yields reliable results up ence th at occurs at low er ionic strengths.
to / = 0.5: The interference of other ions in the solution
fl/2 with the ions of a salt causes th e solubility of the
- lo g y = A t 0. 2/ (10.47) salt to increase. For exam ple, the solubility of
1 + I 1' 2 C a S 0 4 in pure w ater at 25 C ( 7 = 1.0, K sp =
25
F urther extensions of these m ethods by Pitzer 10-4,5) is IO- 2 or 5.6 X 10- 3 m ol/L . H ow ever, in
(1973) enabled H arvie and W eare (1980) to calcu w ater of the Mississippi R iver ( / = 4.4 X 10- 3
late the solubilities of m inerals in brines having y 2 = 0.759) the concentration of Ca2+ in equilib
ionic strengths greater than 2 0 . rium with anhydrite is:
Activity coefficients for ions of different (C a2+) = [Ca2 10-2.25 7.4 X 10- 3 m o l/L
charges in solutions of varying ionic strengths, as 0.759 =
?2
determ ined from the Davies equation (10.47), are (10.48)
listed in Table 10.5. N ote that the values of the
which implies an increase of the solubility of
activity coefficients in Table 10.5 decrease with
increasing ionic strength of the solution and with anhydrite of 32.1% .
increasing charge of the ion. For example, we find This result illustrates the point that the Law
by interpolating in Table 10.5 that the activity of of M ass A ction yields inaccurate results unless the
N a+ in w ater of the Mississippi River concentrations of reactants and products of a
(/ = 4.4 X 10-3) is about 0.936, w hereas that of reaction at equilibrium are converted to activities
Ca2+ is 0.759, and that of L a3+ is only 0.543. In by m eans of activity coefficients in order to cor
other words, the presence of other ions reduces the rect them for the nonideality of electrolyte solu
effective concentration of N a+ in the Mississippi tions. Conversely, the results of calculations based
River by 6.4%, that of Ca2+ by 24.1%, and that of on the Law of Mass A ction are expressed in term s
La3+ by 45.7% . In general, activity coefficients of activities and m ust be converted to concentra
decrease with increasing ionic strength, but tions before they are interpreted . The solubility
increase again at high values of I and may becom e calculation of anhydrite in M ississippi R iver
larger than one (G arrels and Christ, 1965). The w ater indicates that the presence of other ions in
increase of the activity coefficient in highly saline a solution m ust not be ignored.
142 SALTS AND THEIR IONS

Ionic Strength (/)

Figure 10.2 Variation of the activity coefficients of ions having different charges and hydrated radii with the ionic
strength of aqueous solutions at 25C.The curves were calculated from the Debye-Hiickel theory expressed
10.46 with data from Tables 10.3 and 10.4. Note that the magnitude of hydrated radii of ions of equal
charge becomes important at I > 10 2.
Table 10.5 10.4 Solubi
Activity
lity of
Coefficients for
Ions of Different Calciu
Charges in m
Electrolyte Carbo
Solutions of
Varying Ionic nate
Strengths Based We are now ready to
on Equation calculate the solubility
10.47 by Davies of calcite, which is
one of the m ost com
m on m inerals on the
/ 1 face of the E arth . The
calculation requires all
1 x 1 0 '3 0.97
of the skills we have
5 x IO'3 0.93 developed in applying
1 x IO'2 0.90 the Law of Mass A
5 x IO'2 0.82 ction and is by no m
1 x 1 0 '1 0.78 eans trivial (Butler,
2 x 1 0 '1 0.73 1982; G arrels and
5 x 1 0 '1 0.69
Christ, 1965). We start
with the dissociation
s o u r c e : Rrauskopf of calcite into ions at
(1979). 25 C and the
subsequent hydrolysis
of the carbonate and
bicarbonate ions to
form carbonic acid:

C a C 0 3 ^ C a 2+ + C O 2'
CO2 + H20^HC03 +

HC03 + H20^H2C0
10.4 SOLUBILITY OF CALCIUM CARBONATE 143
 [H +][O H _] = [O H " ] 2
If the solution is open to We notice th at according
The sixth equation arises from the
to equation 10.56: Finally, from equation 10.57:
bonic acid is in equilibrium w ith
ide, which has a p artial for electrical neutrality: = [C P 2] = 3 [H +] = 10 140
atm osphere: [ h 2 c o 3] 32.2
2(C a2+) + (H +) = 2(C O ]-) + ( H C 0 3) + O H ") (10.58) [O H ]
H 2 C 0 3 ^ C 0 2 (g) + H 20
= 9.3
We now convert the
W hen calcite crystals are activities of the ions to
Evidently, the activity of
in equilibrium with their their corresponding
carbonic acid is fixed by
ions in pure w ater open to concentrations for
the partial pressure of C 0
the atm osphere at 25 C, substitution into equation
2. We can use this result in
seven ions and m olecules 10.58. In order to do so we
equation 10.55 to express [
are present: must know the ionic
H C 0 3] as a func tion of strength of the solution,
Ca2+, C O ]", H C O J, [O H - ]: which depends on the
H 2 C 0 3, C 0 2, H +, O
concentrations of the ions
H
[ it con tains. Strictly
H ow ever, the partial H speaking, we cannot
pressure of C O z is fixed 2 determ ine the ionic
by strength w ithout first com
c
the com position of pleting the calcu lation. H
o
soil gases may contain up to ten ow ever, we cannot com
3
C02 as the open atm osphere. plete the calcula tion w
]
need only six independent ithout knowing the ionic
[
equations in order to strength. We deal w ith this
O
specify the concentrations Catch-22 situation by
H
of all ions and m ole cules iteration. T hat is, we
-]
in a saturated solution of assum e that all activity
[H C O 3 ] =
calcite. We obtain
(9.3 coefficients are equal
five equations by applying
A ction to equations 10.49 through (10.60)
the dissociation equilibrium io
5

equation 10.49:
=
[Ca2 +][C O ]1 = 102 6 1
8[O
From equations 10.50 and 10.51: H ]
[hco3][oh-] From equation 10.54 we
[C O ]-] ~~ derive a relationship betw
een [C O ]- ] and [O H - ]:

[ h 2 c o 3 ][o h ~ ] [C O ]] = [ h c o 3 ] [ o h ]
= 1 0 - 37

[H C O 3]
= 106 3 1 8 [O H " ] 2

Finally, from equation 10.52: From equations 10.61 and

10.53 we obtain:
[c o 2]
1 0 - 8 ,35
[Ca2+] =
[h2c o 3] [CO ] ]
and for water: IQ 14.668
144 SALTS AND THEIR IONS

to one, m ake the calculation, and then use the results to We can now calculate the activities of the
determ ine the ionic strength and hence the activity other ions in the solution by substituting
66
coefficents from the D avies equation (10.47) or from [O H - ] = 10- 5 into equations 10.60-10.63. The
Table 10.5. We then return to equation 10.58, convert the results are:
activities of all ions to the appropriate concentrations, and [H C O - ] = 102 6 1 8 [O H - ] = IO2 6 1 8
solve the problem a second time. If necessary, the process X 10 -5.66
can be repeated several tim es to achieve a desired level of = io -3042 = 9.08 X 10- 4 m o l/L
accuracy. [C O |- ] = io 6 -3 1 8 [ o h - ] 2 = IO6 3 1 8 x io -11-32

T herefore, we now proceed w ith the solu tion of the = IO- 5 0 0 2 = 9.95 X 10- 6 m o l/L
problem by substituting equations 10.60, 10.61, 10.62, and IQ -14 .668 14.668
10.63 into equation 10.58, assum ing th at all activity
[Ca2 [O H - ] 2 10- 1 L 3 2 = 10-3-348
coefficients are equal to one:
4.48 X 10- 4 m o l/L
2
x i c r 14-668 io -140 , , , 1R. r ,, [JJ+] =
+ r Ti = (2 X 10 )[O H _] 10-14- IO- 1 4 0 - 8 .3 4

[O H ] 2 [O H _ (10.64) _____ _____ -

= 10 -8 34 m o l/L

[O H - ] 10
+ 102 6 1 8 [O H - ] + [O H ]

We simplify equation 10.64 by elim inating the We check the correctness of these results by cal culating
the IA P of calcite in this solution:
term for (H +) because the solution will be basic
and by com bining 102 6 IS[O H _] + [ O H ] as IA P = (9.95 X 10-6) X (4.48 X 10-4) = 10- 8 3 5
102 6 1 9 [ o h - ]. N ext, we clear the fractions by m ul ( 10.6 8 )
tiplying all rem aining term s by [O H - ] 2 and collect
term s: which is identical to for calcite and therefore
106 6 1 9 [O H - ] 4 + 102 6 1 9 [O H - ] 3 = 10- 1 4 3 6 5(10.65) confirms the correctness of the calculation. The ionic
619
strength of the solution is I = 1.4 X 10-3, and the activity
D ividing each term by 106 yields: coefficients of the ions from
]3 Table 10.5 are y 1 = 0.966 and y 2 = 0.857.
[O H - ] 4 + 10- 4 0 0 0 [O H - = IO- 2 0 9 8 5 (10.66)
R eturning to equations 10.60-10.63, we now
convert the activities to concentrations:
W e solve this equation by trial and erro r starting w ith [O 1 0 2 6 1 8 [ o h - ] _ 102 6 1 8 [O H - ]

H - ] = 10- 6 Qbecause we predict that the

(H C O j) = Ti 0.966
solution will be basic. Substituting into equation
0 -2
1 0 . 6 6 yields 1 4 + 1 0 -22, which is about 1 0 times 102.633[O H - ]
sm aller than IO-20'985. Therefore, [O H - ] m ust be
106 3 1 8 [O H - ] 2 106 3 1 8 [O H - ] 2
larger than IO- 6 0. We note th at 10-24 10-22and
therefore elim inate the first term from equation (CO 2 -) r2 0.857
1 0 . 6 6 to get:
1
-4.000 1 0 -20.985 (10.67) io 6 3 8 5 [ o h - ] 2
0 [O H - ] 3
(Ca24) = -^Q -14.668 10 -14 .668

w hich yields [O H - ] = IO- 5 , 6 6 m ol/L . By substitut ing 0.857[O H - ]2

y 2 [O H - ] 2
this value into equation 1 0 . 6 6 we find that the term we om ^ Q 14.600
itted is only 2.2% of 10- 4 0 0 0 [O H - ] 3, w hich is
[O H - ] 2
acceptable.
10.4 SOLUBILITY OF CALCIUM CARBONATE 145
 8.36, thus confirm ing our
S ubstituting into equation
the effect of hydrolysis of
(H +), and com bining the carbonate ion of
before, w e obtain: calcite.
2 X 10 -14.600 = ( 2 x 106
The chem ical
[O H - 1 2
analysis in Table 10.2
+ indicates that the w ater in
1 the Mississippi R iver has
02 a Ca concentration of 34
fi3 ppm , which is nearly 75%
4
discrepancy betw een the C
the predicted Ca a2+ concentration of a
concentration in the saturated calcite solution
Mississippi River only to and the calcium
22.0 m g /L and therefore concentration in the w ater
does not resolve the of the Mississippi River to
discrepancy. consider how the solubility
The w ater in the M of calcite is affected by
ississippi River, like all variations in the partial
natural solutions, contains pressure C 0 2, the pH ,
a m ixture of ions and and the tem perature.

[ larger than the value we molecules derived from

O obtained above. The dif different sources. Its Ca2+
H ference is partly due to the content originates from the The reaction of calcite
-] higher ionic strength of w solution of calcium w ith carbonic acid betw
ater in the M ississippi R carbonate (calcite and een pH values of 6.35 and
N ote th at this equation is
iver (I = 4.4 X 10-3) and aragonite), calcium sul fate 10.3 can be rep re sented
stated in term s of the
to the correspondingly (anhydrite and gypsum ), by the equation:
activities of O H - , which
lower activity coeffi cients and calcium phos phate CaC03 +H2C03^
express the concentra
(y 1 = 0.936, y 2 = (apatite) and from th e C a 2+ + 2H C O J
tions of the other ions. By
0.759). However, transform ation of (10.70)
solving equation 10.69 as
allowance for the higher plagioclase to kaolinite or w here carbonic acid is
before, we obtain [O H - ]
ionic strength increases o th er clay m inerals. In
= io - 5 6 4 4 m ol/L . This also in equilibrium with
addition, C a2+ m ay be
value yields for the other aqueous C 0 2, which in
adsorbed on clay m in
ions: turn is in equilibrium with
erals and on the surfaces of
o th er sm all m ineral C 0 2 gas:
[H C O j] = 9.42 X 10
[CO 2 -] = 1.07 X particles in suspension H 2 C 0 3 ^ C 0 2 (aq) + H 20
cium occurs in m any C 0 2 ( a q ) - C 0 2 (g)
[Ca2+] = 4.17 X cules w hose presence
[H +] = 10 - 8 3 5 6 or concentration of n atu ral
concentration of C a2+ based on the solubility of
The ionic strength, based on these results, is its principal m inerals. A
partial list of the com
I = 1.3 X 10-3, which does not differ enough from
plex ions of C a includes
the previous estim ate to
C aC l+, CaCl2, C a N O j,
justify a repetition of the
Ca(O H ), C a P O /,
calculation. Therefore, the
CaP2 0 2 - , C aH P 2 O j,
concentration of Ca2+ in
(Lindsay, 1979). T he point
the solution is:
is that natural w aters are
(C a2+) 4.17 X IQ- 4 very com plex and their
0.857 com positions are not
explainable by the
or 19.5 m g/L . The
solubility of a few com m
solubility (S) of calcite in
water on minerals. We therefore
in equilibrium with C 0 2 abandon the effort to
25 C is: explain the concentration
5 = 4.86 x 10- 4 X 100.0787 of calcium or any other
elem ent in the M ississippi
In addition, note th at the River. Instead, we use the
146 SALTS AND THEIR IONS

N ote that equation 10.52 is the sum of 10.71 and 10.72
and th at m olecular C 0 2 dissolved in w ater is thereby
elim inated from consideration. We can use these
equations to predict qualitatively how the solubility of
calcite is affected by changes in the partial pressure of C 0
2. A n increase in the partial pressure of C O z at constant
tem perature increases th e concentration of carbonic acid
in the solution. Consequently, m ore calcite dissolves by
equation 10.70 in accordance with Le C hteliers
principle. Similarly, a decrease in the partial pres sure of
CO z causes a saturated solution of calcite to becom e
supersaturated and results in the p re cipitation of calcite
until equilibrium is restored.
The partial pressure of C 0 2 at a site on the surface of
the E arth m ay decrease as a result of photosynthesis of
aquatic plants, which has the effect of com bining C 0 2 and
H20 to produce glu cose (C 6 H 1 2 0 6) and 0 2 with the help
of ultraviolet radiation . Consequently, com m unities of
aquatic plants can low er the concentration of C 0 2 (aq) in
the w ater during daylight hours. This may cause calcite to
precipitate if the w ater becom es super saturated with
respect to calcite. C ertain colonial algae are capable of
precipitating calcite by this m echanism , thereby form ing
thinly lam inated cal cite m ounds called strom atolites (W
alter, 1977).
to form a stalactite hanging from the roof of the cavern. W
hen the rate of flow of groundw ater is so rapid that the
solution does not reach equilib rium on the roof of the
cavern, additional calcite m ay precipitate after the w ater
has dripped onto the floor of the cavern, thereby forming a
stalag mite. Ultim ately, stalactites and stalagm ites may
join to becom e pillars or curtains of calcite.
The effect of the partial pressure of C O , on the
solubility of calcite is shown quantitatively in Figure 10.3.
H ere we see that the concentration of Ca2+ in a saturated
solution of calcite at 25 C increases from about 20 m g /L
to about 45 m g /L w hen the partial pressure of C 0 2
increases from 3 X 10~ 4 atm in the air to 3 X IO- 3 atm in
soil gases. Consequently, each liter of w ater appearing on
the roof of a cavern could, under ideal condi tions, deposit
25 mg of calcite. The am ount of cal cite th at is actually
deposited depends on the partial pressure of C 0 2 in the
soil above a given cavern, on the tem perature, and on the
establish m ent of equilibrium betw een calcite and the
groundw ater before the w ater reaches the cavern. N
evertheless, stalactites and stalagmites can grow

A sim ilar process in lim estone caverns results in the

_1 \ /
deposition of calcite speleothem s in the form of 'oo s / /
stalactites and stalagmites. Lim estone cav erns form by E 100 - X X /
the solution of calcite by water con taining carbonic acid c 80 - X / - 8
o
in accordance with equation +- PH /
03
X /
10.70. This process occurs prim arily below the c 60 - / / X
X
w ater table, and the solution cavities are initially O)
u
- 7
40 -
filled with w ater. H ow ever, if the groundw ater c
o
table is low ered because of uplift of the crust or a u
+ 20 -
decrease in m eteoric precipitation, the solution To -6
cavities above the w ater table becom e filled with air. M u 0-
eteoric w ater percolating through the soil m ay equilibrate 3 x 10-* 3 x io-3 3 x io -2
with C 0 2 at a higher partial pres sure than exists in the Partial Pressure of C 0 2, atm
atm osphere and thus becom es a m ore effective solvent of
Figure 10.3 Increase of the solubility of calcite in water
calcite than it w ould have been had it equilibrated with C
at 25 C with increasing partial pressure of C 0 2 from 3 X
0 2 of th e atm osphere. W hen this w ater form s a drop on 1CT4 to 3 X 1CT2 atm. The solubility of calcite is
the ceiling of a lim estone cavern, the excess CO z escapes, expressed as the concentration of Ca2+in a saturated
the solution therefore m ay becom e supersaturated, and solution based on ionic strengths, which rise from
calcite may then precipitate 1.3 X 1CT3 to 6.3 x 1CT3, whereas the pH decreases, as
more C 0 2 dissolves.
10.4 SOLUBILITY OF CALCIUM CARBONATE 147

to large size over
thousands or tens of
thousands of years.
The solubility of
calcite increases very
signifi cantly with
increasing acidity of the
solution, if the pH is
controlled independent of
the carbon ate equilibria.
E quation 10.62 provides a
relation ship betw een the
activity of Ca2+ in a
saturated solution of
calcite in equilibrium with
COz of the atm osphere of
25C and the activity of O
H . Replacing [O H ] by
/CW/[H +] gives:
1 0
-6 6 81
0 1 4 -6
6 8 [H + ]
2
a J [OH
]2 (Kj
2 increase with increasing
= io ,3-
332[ h
+]2
Taking logarithm s of both
sides and converting
[H +] to pH , we obtain:
log [Ca2+] =
13.332 -
2pH
(10.74)
This equation indicates
th a t the solubility of
calcite changes by a
factor of 100 w hen the
pH is
varied by one unit. For example,
[Ca2+] = 2.15 X 10 1 m ol/L , whereas
[Ca2+] = 2.15 X 10+ 1 m ol/L . Evidently, when cal
cite reacts with an excess
am ount of a strong acid,
its solubility is very high. Temperature,
Similarly, calcite cannot of Ca2+ at 5C but only 40 mCg /L at 30C. P*A1
Such PKA2 pA/p
persist as a stable phase in changes in the environm ental0 conditions 6.58
may 10.62 8.02
natural environm ents th a occur w hen cold w ater from the 5 6.52 10.56 8.09
t are even mildly acidic. oceans rises to the surface and 10 6.47 10.49 8.15
For exam ple, calcite in shallow lagoon. As a
15 6.42 10.43 8.22
till of W isconsin age in result, seaw ater m ay
20 6.38 10.38 8.28
the m idcontinent area of becom e supersaturated
25 6.35 10.33 8.34
N orth A m erica has been with respect to calcite (or
rem oved by leaching to 30 6.33 10.29 8.40
arago nite) because of the
depths of about one m eter 40 6.30 10.22 8.52
increase in tem perature as
or m ore in only 15,000 well as the decrease in the 50 6.29 10.17 8.63
years. partial pressure of C 0 2. H
ap^M = -log *ai;
ow ever, carbonate sedim p^a2 = -log Aa2;
The solubility of ent in the oceans is pri m pKsp = log * sp;
calcite is also affected by arily of biogenic origin pXca = log
the tem perature because Kc0_for CO, + H ,0
and results from the
of changes in the num H2CO,
secretion of carbonate bValues in
erical values of all of the skeletons by plants and parentheses were
equilibrium constants. G anim als living in the obtained by graphical
arrels and C hrist (1965) water. inter polation.
com piled the set of values SOURCE:Garrels and Christ
(1965),Table 3.2.
14 of the relevant equilibrium The opposite
constants betw een 0 and environm ental changes
50C that is listed in Table are experienced by
10.6. We see by inspec skeletal carbonate grains
tion that the dissociation sinking through the ocean
constants of carbonic acid tow ard the bottom . O
ceanic bottom w ater is
tem perature, w here as colder than surface w ater
the solubility product and contains m ore
constants of calcite and C dissolved C 0 2 produced
0 2 decrease. As a result, by decay
the solubility of calcite in of organic m aterial. Consequently, the solubility
pure w ater in equilibrium of calcite (or aragonite) rises with increasing
with C O z of the depth, and skeletal carbonate particles dissolve as
they sink in the oceans. W hether such particles
atm osphere actually decreasessurvive the trip through
tem perature, as shown in Figure
reach the bottom of the
ple, a saturated solution
of calcite in equilibrium
with C 0 2 at 3 X 10 3 atm
contains about 75 m g /L Table 10.6
Equilibrium
Constants of
Carbonate
Equilibria as a
Function of
Tem perature3
148 SALTS AND THEIR IONS
Temperature, C
Figure 10.4 Decrease of the concentration of Ca2+ in a
saturated solution of calcite in pure water with
increasing temperature at two partial pressures of
C 0 2. Activity coefficients were taken from Table 10.5
for the ionic strengths of the solutions, which range from
about 5 X IO-3 (Pco, = 3 X 103 atm, 5C) to 0.5 X
103(PCO, = 3 x 10r4 atm, 80C).
rates at which they sink and dissolve. If the w ater is deep
enough, carbonate particles dissolve before they reach the
bottom . Therefore, no car bonate sedim ent can accum
ulate in the oceans below this critical depth, which is
known as the carbonate com pensation depth (CCD ).
10.5 Chemical Weathering
The solution of calcite as lim estone or as m ineral grains
in soil is a form of chem ical weathering. The chemical
reactions th at take place in this case
are the dissociation of the com pound into ions as it
dissolves in the w ater, accom panied by hydroly sis of the
carbonate ion. The point here is that cal cite and m any
other naturally occurring chemical com pounds dissolve
congruently, th at is, w ithout forming another com pound.
How ever, the Al sili cate m inerals generally dissolve
incongruently; th at is, they react with w ater to form
another solid com pound as well as ions and molecules.
The process of chemical w eathering not only
includes the congruent and incongruent solution of
m inerals but also encom passes all kinds of
interactions of w ater and atm ospheric gases with
m inerals at or n ear the surface of the Earth.
Chemical w eathering also includes oxidation -
reduction reactions, which we will take up in
C hapter 14. The reactions that constitute chemical w
eathering are not restricted to the surface of the E arth w
here m inerals are actually exposed to the weather, but
also operate effectively in soils and below the groundw
ater table w here m inerals are continuously in contact
with water. The products of chemical w eathering consist
of ( 1 ) new m inerals such as clay minerals, oxides, and
hydroxides; (2 ) ions and m olecules th at dissolve in the
water; and
(3) grains of m inerals that are unreactive such as quartz,
garnet, zircon, muscovite, and native gold.
The solid weathering products, including new
compounds and resistant minerals, are a major source of
sedim ent that may accumulate in resid ual deposits at
the site of weathering or is transport ed to the oceans
where it is deposited and ultimately forms clastic sedim
entary rocks. The ions and m ole cules in solution determ
ine the chemical composi tion or quality of w ater and
make up the solution load that is transported to the oceans
by streams. Ultimately, as noted in C hapter 4, these ions
and molecules are rem oved from the oceans by precipi
tation of com pounds or by adsorption on solid par ticles.
Evidently, chemical weathering is part of the geochemical
machine that operates on the surface of the Earth.
Chemical and mechanical forms of weathering cause the
destruction of rocks, thereby releasing the elem ents for a
trip to the ocean. We recognize, of course, that the trip to
the ocean may be delayed by deposition of sedim ent or of
dis solved constituents in basins on the continents and
10.6 TRANSFORMATION OF POTASSIUM FELDSPAR TO KAOLINITE 149

that chemical wethering
occurs also in the oceans
themselves. Some of these
complexities will become
the subject of m ore
detailed consideration in
Part V of this book. For
the time being, we attem
pt to apply the Law of
Mass A ction to the
chemical weathering of a
common A l silicate such
as K-feldspar.
10.6 Transfor
mation
of
Potassiu
m based on the following
Feldspar
to
Kaolinite
W hen K -feldspar (orthoclase or
igneous or high-grade m etam orphic rocks
exposed to chem ical w eathering,
chalky appearance
caused by the form ation
of kaolinite. This m
ineral belongs to the
large group
of clay m inerals (C hapter
chemical w eathering of Al-silicate of the reaction.
igneous and m etam orphic
involving volcanic glass,
by hydrotherm al process
es, and by transform
ations from other clay m
iner als. In order to apply
the Law of Mass A ction
to the reaction th at causes
K -feldspar to form kaoli
nite, we m ust first write
the equation th at rep re
sents this reaction. The
construction of such
equations requires an We now have 16 oxygens on th
understanding of the chem 4. A ny H + ions
9 + 16 = 25 oxygens on the right. T
ical properties of the elem consum ed by the
add 9 H 20 to the left side
ents that may partici pate reaction originate
of the equation to bal ance
in the reaction and of the from the
environm ent in which the oxygen:
environm ent.
reaction takes place. 5. A fter the 2 KAlSi3 0 8 + 9 H 20 -> A l2 Si2 O s(O H
Therefore, we set up a list equation has first + 2 K + + 4 H 4 S i0 4
of guidelines th at will been balanced Finally, we have 18 hydrogens on th
govern the way in which with respect to Al,
we construct equations for Si, and soluble 4 + 16 = 20 hydrogens on the right. T
chemical reac tions in cations, we equation 10.77 is balanced
nature. balance oxygen in all respects by adding
by adding H 20 as two H + to the left side:
necessary.
We start by placing 2 K A lSi3 0 8 + 9 H 20
the reactant on the left and
6. We then balance
hydrogen by adding H
the know n product on the + + 2 H + -
(10.78)
right side of the equation as needed. A l2 Si2 0 5 ( 0 H ) 4 + 2
and then proceed to 7. The equation m ust K + + 4 H 4 S i0 4
balance the equa tion term s of m atter and This equation also satisfies
charges. the requirem ent for
conventions and requirem N ote th at oxygen cannot
ents, provided that the
electrical neutrality
unless the reaction because we have tw o
valence num bers of the
involves oxidation positive charges on the left
elem ents do n o t change
and two on the right side.
during the reaction. -reduction . In
balancing oxidation
1. In reactions
involving Al- -reduction reactions
silicates the A l is the
conserved; that is, electrons th at are exchanged m ust be
it does not first before a balance of m atter and
dissolve charges is accom plished.
appreciably but is The rules listed above
transferred into now perm it us to con
the solid product struct an equation for the
conversion of K -feldspar
2. Excess Si form s to kaolinite. We start by
silicic acid, which writing down the known
does not reactant and its product:
dissociate unless
the environm ent
KAlSi3 O g - A l
is very basic (pH
K -feldspar
> 9.0).
3. A lkali m etals Next, we balance A l and
and alkaline earth then take care of K and Si:
elem ents form 2 KAlSi3 O s -> A l
ions in solution. + 2K +4H
150 SALTS AND THEIR IONS
We now examine the equation we have con structed to
learn by the application of Le C hteliers principle how
the reaction at equilibri um responds to changes in certain
environm ental parameters. We recall that reactions in
nature may not achieve equilibrium but run to completion
if one of the products escapes from the system. In the case
we are considering, this could occur if K + and H 4 S i0 4
are rem oved by the m ovem ent of the water in which the
reaction is taking place. Therefore, the conversion of K
-feldspar to kaolinite is favored by the m ovem ent of
groundw ater or surface water in response to a hydraulic or
topographic gradient. We also note that nine moles of w
ater are consumed for each mole of kaolinite that is
produced. Therefore, the form ation of kaolinite is favored
by an abun dance of water, which, in addition, must be
acidic to provide the H + needed by the reaction. We con
clude, therefore, that K -feldspar is converted into kaolinite
in places where an abundance of acidified w ater is
available and where the w ater flows at a sufficient rate to
rem ove the soluble products of the reaction. Also, since
reaction rates generally double for every 10 C increase in
tem perature, the form a tion of kaolinite is likely to be
favored by tropical climatic conditions near the equator of
the Earth.
The reaction betw een K -feldspar and water involves
the hydrolysis of a salt derived from a weak acid and
strong base. We predict that such a reaction should rem ove
H + from the solution and increase its pH .This is indeed
the case here because H + is consum ed by the reaction.
However, the pre dicted increase of the pH can only be
dem onstrated in the laboratory because natural waters are
buffered (Wollast, 1967).Therefore, when this reac tion
occurs in nature, the H + originates from the environm ent
by the dissociation of carbonic or hydrosulfuric acid, or
some other reaction. If the supply of H + is large and if the
soluble products are rem oved, the reaction can run to
completion. If, however, the w ater is basic and the soluble
products rem ain at the site, the reaction can achieve
equilib rium . In that case, the conversion of K -feldspar to
kaolinite stops and the reaction becom es unpro ductive.
Therefore, the establishm ent of chemical equilibrium may
p reven t the form ation of a poten
tially valuable m ineral deposit and, in that sense, is not
desirable from our point of view.
The probability that a chemical reaction in nature,
such as equation 10.78, can achieve equi librium depends,
am ong other things, on the w ater/rock ratio. If there is an
abundance of w ater
com pared to the volum e of rock, reactions are
likely to run to com pletion because the environ
m ent controls the reaction. If, on the other hand, the w
ater/rock ratio is small, the reactions control the environm
ent and equilibrium is possible, p ro vided reaction rates
are adequate. Therefore, reac tions taking place in large
bodies of water and in aquifers with high percolation rates
having large w ater/rock ratios are m ore likely to run to
com pletion than to achieve equilibrium , whereas reac
tions in small volumes of pore w ater (small w ater/rock
ratios) may achieve a localized state of equilibrium betw
een ions and their solids. Such restricted environm ents
occur in the pore spaces of rocks in which m ineral cem
ents m ay be deposited by reactions that m aintain a state
of equilibrium .
The Law of Mass A ction enables us to deter mine the
activities of ions and molecules required for equilibrium in
a chemical reaction represented by a balanced equation.
Therefore, we return to equation 10.78 and place that
reaction at equilibri um. By applying the Law of Mass A
ction we obtain:
[H 4 SiQ 4 ]4 [K + ] 2
K (10.79)
[H + ] 2
We linearize this equation by taking logarithms:
4 log[H 4 S i0 4] + 2 log [K+] - 2 log [H +]
(10.80)
= log K
In order to reduce the num ber of variables to two, we
com bine K + and H + as the [K +]/[H +] ratio:
4 log [H4S iO j + 2 log | | J | = log K (10.81)
We rewrite this equation as:
[K 5 log K -2 log[H 4 S i0 4] (10.82)
log =
[H +]
which is the equation of a straight line in coordi nates of
log [K+]/[H +] and log [H 4 S i0 4] having a
10.6 TRANSFORMATION OF POTASSIUM FELDSPAR TO KAOLINITE 151

slope m = - 2 and an
intercept of b = ) log K . It
is the locus of all points w
hose coordinates are the
logarithm s of the [K+]/[H
+] ratio and of [H 4 S i0 4]
required for equilibrium
of reaction 10.78. Thus,
any point P on the line in
Figure 10.5 (equation
10.82) represents an
environm ent in which K-
feldspar and kaolinite
coexist in equilibrium
with the ions and m
olecules in equation
10.78. Further,
any point th at is n o t on equation 10.16 of Garrels
environm ent in which K -feldspar
are n o t at equilibrium . In
and vice versa, in the
presence of water at
25C.The line labeled
10.82 represents
conditions at which K-
feldspar and kaolin ite
coexist in equilibrium.
Point P on the line was
cho sen arbitrarily to help
determine the direction of
reaction 10.78 when [H4S
i0 4] is increased or
decreased. Line 10.83
represents the solubility of
amorphous silica, which
limits the activity of [H4S
i0 4] to values less than
IO2 74 m ol/L, as shown
in equation 9.89. (Line
10.82 was plotted from
and Christ, 1965.)
will proceed in the
direction th at moves it
closer to a state of
equilibrium . If the w
ater/ro ck ratio is small,
the environm ental param
eters can change until they
satisfy equation 10.82 and
equilibrium is established.
If the w ater/ro ck ratio is
large, the environm ental
param eters m ay not be
affected appreciably by
the reaction we are now
consider ing. The reaction
therefore runs to com
pletion
and does n o t achieve equilibrium .
The straight line we have draw n in
10.5 (equation 10.82) is a boundary th a t separates
th e
environm ents in which K -feldspar
duced by reaction 10.78
from those in which
kaolinite is produced . We
assum e h ere th at the
reaction represented by
equation 10.78 is, in fact,
reversible. A lthough this
m ay n o t be tru e in all
cases, a K -feldspar know
n as adularia does form
from kaolinite at low tem
p eratu re (M ensing and
Faure, 1983; K astner and
Siever, 1979). In any
case, if we m ove from
point P on the equilibrium
environ m ents containing
less H 4 S i0 4 than is
required for equilibrium of
reaction 10.78.
The activity of silicic
acid is lim ited by its sol
ubility, as shown in Figure
9.4. T herefore, am or
phous silica precipitates w
hen its solubility limit is
reached and the activity of
[H 4 S i0 4] in Figure 10.5
cannot rise above the value
required for equilibri
um in the reaction:
S i0 2 (am orph .) + 2 H 2 0 ^ H 4 S i0 4
A t equilibrium , [H 4 S i0 4] = 10 2 7 4 m
tion 9.89). Therefore, a line has b
Figure 10.5 at log [H 4 S i0 4] = -2 .7
im penetrable boundary
because [H4 S i0 4]
cannot rise to higher
values. The two lines in
Figure 10.5, therefore,
divide the com
positional plane into
three fields th at
represent environm
ental condi tions
needed to stabilize K
-feldspar, kaolinite, and
am orphous silica.

line in Figure 10.5 tow ard

H 4 S i0 4, the reaction
log kaolinite with H 4 S i0 4 to
[H4S Therefore, K -feldspar is
i0 4] the stable phase in all
Figure 10.5 Partially environm ents represented
completed activity in Figure 10.5 that have
diagram depicting the an excess of H 4 S i0 4
environmental conditions above th at required for
required to convert K- equi librium . Similarly,
feldspar into kaolinite,
kaolinite is stable in all
152 SALTS AND THEIR IONS
The procedure we have used to determ ine
the environm ental conditions necessary for the
conversion of K -feldspar to kaolinite can serve to define
the stabilities of all m inerals th at may be exposed to w
ater and atm ospheric gases at or near the surface of the E
arth . U nfortunately, we cannot proceed in this direction
because the equi librium constants of the reactions we w
ant to study have not been m easured owing to experi m
ental difficulties. For this reason, we next un d er take a
brief excursion into therm odynam ics because it is
possible to calculate equilibrium con stants for chem ical
reactions from the change in the standard free energy of
reactants and p ro d ucts. Therm odynam ics will enable us
to evaluate m ost chem ical reactions we m ay care to
investi gate and will also provide a proof for the validity of
the Law of Mass A ction.
10.7 Summary
W hen salts dissolve in w ater, they dissociate into cations
and anions. The solubility of a salt can be determ ined by
calculating the activities of its ions in a saturated solution
based on the Law of M ass A ction. A salt may precipitate
from a supersatu rated solution in which the ion activity
product (IA P) of its ions exceeds its solubility product
constant. W hen a salt does precipitate, the concen tration
of its ions rem aining in the solution change and their ratio
tends tow ard infinity or zero. The resulting change in the
chem ical com position of
the w ater contributes to the evolution of brines
form ed by progressive evaporation of w ater in
desert lakes or in sedim ent drying in the sun.
The precipitation of salt A m ay also be caused by the
addition of another, m ore soluble, salt B that shares a com
m on ion with salt A . A s a result, salt A precipitates as salt
B dissolves until equilibrium betw een the ions and both
salts is established. This process can account for the
replacem ent of a m in eral (salt B) by another (salt A ) th
at is less soluble.
Salts can be classified on the basis of the strength or
weakness of the acids and bases from which they
originated. W hen the ions of a weak acid or w eak base
are released into solution by the dissociation of a salt, they
react with w ater to
form the parental w eak acid or weak base. This phenom
enon is called hydrolysis and causes the pH of a salt
solution to change. Hydrolysis of car bonates, silicates,
and phosphates of the alkali m etals and alkaline earths
tends to m ake ground w ater basic, thereby enhancing
their solubility. In addition, salt solutions that also contain
the parental acid (or base) of the salt resist change in their
pH and are therefore said to be buffered.
The ability of the ions of a salt in solution to interact
with each other is dim inished by the pres ence of other
ions and by interactions with the molecules of the solvent.
Therefore, the m olar concentrations of ions m ust be
corrected by fac tors called activity coefficients derived
from the D ebye-H ckel theory before they are used in
applications of the Law of Mass A ction. Similarly,
calculations based on the Law of Mass A ction
yield the activities of ions, which m ust then be
converted to concentrations before they can be com pared
with observed concentrations d eter mined by chemical
analyses.
The solubility of calcium carbonate is enhanced by
the hydrolysis of the carbonate ion and also rises w ith
increasing partial pressure of C 0 2, with decreasing pH of
the environm ent, and decreasing tem perature. These
relationships are all derivable by applications of the Law
of Mass A ction and explain such diverse phenom e na as
the deposition of strom atolites, the form a tion of lim
estone caverns, the grow th of stalactites and stalagm ites,
leaching or deposition of calcium carbonate in soils, and
the existence of the carbonate com pensation depth (C C D
) in the oceans.
The Law of M ass A ction can also b e used to
study chem ical w eathering of A l silicates such
as K -feldspar, which is converted by incongru-
ent solution into kaolinite. T hese kinds of reac tions can
be rep resen ted by equations th at are
constructed on the basis of certain assum ptions
and conventions. The equations indicate the
necessary conditions for the com plete conver
sion of one m ineral into another, although the
decom position of A l silicates is n o t reversible at surface
tem p eratu res in m any cases. T he equa tions also yield
the activities of ions and m ole
 REFERENCES 153
cules req u ired for equilibrium and thereby environm ental conditions at which the solids are stable. T
define the stability lim its of the solids. By p lo t he num erical values of the equilibri um constants req u
ting the relevant equations derived from the Law of M ass ired for this purpose can be calculated from therm odynam
A ction in log -log coordinates, use ful diagram s can be ics, which is the topic of the next chapter.
constructed th at depict the
Problems
 1. Calculate activities of Ca2+ and F and the solu bility NOJ 0 .5 2 1 .2

of fluorite (CaF2) in pure water at 25 C without regard to Ca2+ 1 4 .1 39

possible hydrolysis. Assume that all activity coefficients Mg2+ 3 .7 8 .7

are equal to one. Express the solubility of fluorite in terms Na+ 2. 9 8.2
of m ol/L and g/100 mL. K" 0 .5 1 .4

2. A quantity of 0.50 g of crystalline NaCl is added to Fe2+ 0 .3 6 0 .0 3

one liter of a solution containing IO3,0 mol of dis solved S i0 2 4 .1 2 .1
A g+. Demonstrate that AgCl precipitates from the
solution and calculate the weight of AgCl in grams that
forms per liter. Assume that all activity coefficients are
equal to one. 6. Calculate the activity coefficient of Mg2+ in an
aqueous solution having / = 5 x 10-2 at 15C based on the
3. What is the ratio of the activities of Zn2+ and Pb2+ in a Debye-Htickel theory represented by equation 10.46.
solution that is in equilibrium with respect to sphalerite (ZnS)
and galena (PbS)? (Disregard hydrolysis effects.)
7. Calculate the solubility of strontianite (SrC03) in
4. Predict the outcome of reactions that occur when a pure water at 25 C in equilibrium with C 0 2 having a
solution having [Cu2+]/[Fe2+] = 10 comes in contact with partial pressure of 3 X 1 0 -4 atm. Specify the concentra tions
hydrotroilite (FeS). What will be the ratio of [Cu2+]/ of all ions in a saturated solution in units of mol/L.
[Fe2+] in the solution after equilibrium is estab lished?
(Disregard hydrolysis effects.)
8. Calculate the solubility of calcite at 15 C in pure
5. Calculate the ionic strengths of water in Lake water in equilibrium with C 0 2 at 3 X 1 0 -3 atm pressure.
Superior (A) and Lake Erie (B) and explain the differ ence Set up the equations based on the dissociation of car bonic
(Livingstone,1963, Table 7). The concentrations are in acid and solve for the activity of H +. Specify the
mg/L. concentrations of all ions and express (Ca2+) in mg/L.
9. Construct an equation to represent the incongru-ent
A B
5 0 .0 121
solution of K-feldspar (KAlSi3Og) to illite
HCO3
(KAl3Si3O10(OH)2).
SOj- 4 .8 28

cr
1 .5 17
References
B U T L E R , J. N., 1982. Carbon
Dioxide Equilibria and Their
Applications. Addison-
Wesley, Reading, MA, 259
pp.
G A R R E L S , R. M., and C . L. C H
R IS T , 1965. Solutions,
Minerals and Equilibria.
Harper & Row, New York,
450 pp.
HARDlE,L.A.,andH.REuGST
ER,1970.Theevolutionofclosed
-
basin brines. Mineral. Soc.
Amer. Spec. Pub. 3,273-290.
H A R V IE , C. E., and J. H . W E A R E ,
1980. The prediction of
min-eral solubilities in
natural waters: the Na-K-
Mg-Ca-Cl-S04-
154 SALTS AND THEIR IONS
H20 system from zero to high concentration at 25 C.
Geochim. Cosmochim. Acta, 4 4 :9 8 1 -9 9 7 .
K s t n e r , M., and R. S i e v e r , 1979. Low temperature feldspars in
sedimentary rocks. Amer. J. Sei., 2 7 9 :4 3 5 -4 7 9 .
K r a u s k o p f , K. B., 1979. Introduction to Geochemistry, 2nd ed.
McGraw-Hill, New York, 617 pp.
L IN D S A Y , W. L ., 1979. Chemical Equilibria in Soils. Wiley, New
Y o rk , 449 pp .
L IV IN S T O N E , D. A., 1963. Chemical composition of rivers and lakes.
In M . Fleischer (Ed.), Data of Geochemistry, 6th ed.
Ch. G., U.S. Geol. Surv. Prof. Paper 440-G, 64 pp.
M E N S I N G ,T. M ., and G. F A U R E , 1983. Identification and age of
neoformed Paleozoic feldspar (adularia) in a Precambrian
basement core from Scioto County, Ohio, USA. Contrib.
Mineral. Petrol., 82:327-333.
P I T Z E R , K. S., 1973. Thermodynamics of electrolytes. I.
Theoretical basis and general equations. J. Phys. Chem.,
77:268-277.
W A L T E R , M. R., 1977. Stromatolites. Elsevier, Amsterdam, 790 pp.
W E A S T , R. C., M. J. A S T L E , and W. H. B E Y E R (Eds.), 1986. CRC
Handbook o f Chemistry and Physics, 66th ed. CRC Press, Boca
Raton, FL.
W O L L A S T , R., 1967. Kinetics of the alteration of K-feldspar in
buffered solutions at low temperature. Geochim. Cosmochim.
Acta, 31:635-648.
 11
Thermodynamics
The science of therm odynam ics originally evolved
from a variety of observational evidence
indicating that m echanical energy is trans
form able into heat. For example, Benjamin
Thom pson (1753-1814) of W oburn, M assachu
setts, b etter known as C ount R um ford of the
H oly R om an Em pire, was im pressed by the
am ount of heat generated during the boring of
cannon at the arsenal in Munich. In 1798 he pro posed th
at the heat was form ed by conversion of the m echanical
energy expended by the boring and actually calculated the
am ount of heat gener ated by one horse turning the borer
for one hour. C ount R um fords ideas about the
equivalence of w ork and heat were supported by Sir H
um phry D avy (1778-1829), who dem onstrated in 1799
that two pieces of ice could be m ade to melt by rubbing
them together by a clockwork m echa nism in a vacuum .
Nevertheless, the scientific com m unity did not accept
this idea for nearly 50 years until James R Joule (1818-
1889), working in his fath ers brewery in M anchester,
England, carefully m easured the rise in tem perature
caused by stirring w ater with a mechanical paddle wheel.
H is research betw een 1840 and 1849 clear ly dem
onstrated that the expenditure of a certain am ount of
work always produced the same am ount of heat (M oore,
1955).
The observation that work is transform able into heat
is an extension of the principle of con servation of energy.
It can be restated by saying th at w henever w ork is done,
heat energy is con sum ed, implying that one m ust
always pay a price by the expenditure of energy if w ork
is to be accom plished.
Therm odynam ics is founded on three basic
principles, which arise from com m on experience
but cannot be proved or disproved. They m ust be
accepted as articles of faith. The science of th e r
m odynam ics has evolved from these so-called
laws by careful deductive reasoning aided by
m athem atics.
11.1 Definitions
Therm odynam ics describes the world the way it appears
to an observer in term s of certain m ea surable properties
of m atter, such as volume, pres sure, tem perature, and
chemical com position. M atter is regarded as continuous
and indivisible. In this regard therm odynam ics is the
opposite of
statistical m echanics in which m atter is treated as
a collection of particles in m otion. In order to
en ter the world of therm odynam ics we m ust first becom
e fam iliar with certain definitions and con ventions th at
may differ from com m on usage.
The basic entity for consideration in therm o
dynamics is the system, which is that p art of the universe
whose properties are under considera tion. The system is
separated from the rest of the universe, know n as the
surroundings, by a b o u n d ary whose properties can be
defined. A system is said to be open w hen m atter can
pass across the boundary and closed w hen m atter is
prevented from entering or leaving the system. The bound
ary may prevent not only m atter but also heat and any
other form of energy from passing through it. Systems that
have such im pervious boundaries are said to be isolated.
The investiga tor m ust identify the system to be studied
and specify the properties of its boundary.
The properties of a system are classified as being
either intensive or extensive. Intensive properties, such as
pressure, tem perature, and
155
156 T H ER M O D Y N A M IC S
density, are independent of the am ount of m atter that is
present, w hereas the extensive properties, such as volum e
and mass, depend on the mass and are therefore additive.
W hen the properties of a system are changed, it
experiences a change in its state. For exam ple, w hen a
quantity of gas in a cylinder is com pressed by m oving a
frictionless piston, the system undergoes a change in state.
The chem ical com position of therm odynam ic
system s is expressed in term s of com ponents and
phases. N ordstrom and M unoz (1986, p. 67)
defined a phase as a uniform , hom ogeneous,
physically distinct, and m echanically separable
portion of a system . W hen the system under con
sideration is a rock, the m inerals of which the rock is com
posed are the phases. Phases need not be solids, but can
also be liquids or gases. For exam ple, a system containing
ice, w ater, and w ater vapor has three phases (solid, liquid,
and gas).The aqueous phase in such a system m ay also
contain a num ber of ionic or m olecular species in
solution.
The com ponents of a system are the chemical
constituents by m eans of which the chemical
com positions of the phases of a system can be
com pletely described. The choice of com ponents is
arbitrary. For exam ple, the chem ical com posi tion of a
rock specim en can be expressed in term s of oxides or in
term s of chem ical elements. Flowever, the num ber of
com ponents required to describe the com positions of all
of the phases p re sent in a system at equilibrium is
constrained by the G ibbs phase rule.
The num ber of variables th at m ust be speci
fied in order to define th e state of a system is
equal to the num ber of com ponents plus the
num ber of intensive properties, such as pressure and tem
perature . T he state of the system is com - pletely defined
w hen the num ber of variables is
equal to the num ber of in dependent relations
betw een them . If th e num ber of independent
equations is less than th e n um ber of variables, the system
has variance or degrees o f freedom . In this case, th e state
of th e system cannot be com plete ly constrained . W hen
th e num ber of variables exceeds the num ber of in
dependent equations by one, th e system is univariant and
has one degree of freedom . U nivariant system s can be
con
strained by specifying the value of one of the variables. W
hen the num ber of variables exceeds the num ber of
equations by two, the system is divariant, and so on.
W hen a chem ical system is at equilibrium , the num
ber of phases (p ), the num ber of com po nents (c), and
the num ber of degrees of freedom
(/) m ust satisfy the Gibbs phase rule (N ordstrom and M
unoz, 1986):
p+f=c +2 (1 1 . 1 )
Therefore, systems in which the num ber of phases is
equal to the num ber of com ponents have two degrees of
freedom . In other words, pressure and tem perature m ust
be specified in this case in order to com pletely define the
state of the system .
All systems contain a certain am ount of heat, which
is defined as that which flows across the boundary of a
closed system during a change in its state by virtue of a
difference in tem perature between the system and its
surroundings. M ore over, heat always flows from a point
of higher to a point of lower tem perature.
Thermodynamics defines heat as though it w ere an
invisible fluid whose direction of flow is indicated by its
algebraic sign. H eat is positive w hen it flows from the
sur roundings to the system and negative when it flows
from the system to the surroundings. We can rem em ber
this im portant convention by recogniz ing that systems
are selfish. W hen a system gains heat, the added heat
has a positive sign, and when it loses heat the sign is
negative.
Work is defined in mechanics as the product of force
times distance, whereas in thermodynamics work is
defined as a quantity that flows across the boundary of a
system during a change in its state and is completely
convertible into the lifting of a weight in the surroundings.
W ork, like heat, is an algebraic quantity, but the sign
convention of work is opposite to that of heat. W hen
work flows to the surroundings it has a positive sign, and
when it flows to the system it has a negative sign. A
nother way to visualize the sign convention for work is to
say that when the system does work, the work is positive,
but when the surroundings do it, the work is negative.
W ith these definitions and conventions we have now
set the stage to introduce the first law of
 11.3 ENTHALPY 157
ut
int/ ayait/H i .
therm odynam ics. The presentation will be selec tive and
based on plausibility rather than m athe m atical rigor. M
ore com plete and rigorous presentations can be found in AE=qw (11.4)
college level text books on physical chemistry. Therm
odynam ics for The first law of therm odynam ics implies that
use by geologists is presented in books by K ern
and W eisbrod (1967), W ood and Fraser (1976), work and heat are related and th at some of the
Fraser (1977), N ordstrom and M unoz (1986), and A heat that flows into the system during a change in state is
nderson (1996). converted into w ork and therefore does not augm ent the
internal energy.
We can im agine th at a reversible change in state can
11.2 The First Law be carried out in small increm ental steps such th at the
internal energy changes by small am ounts designed by dE
Every system whose state has been defined by specifying . Now we can restate the first law in differential form:
the num erical values of its properties contains a certain
am ount of energy. This energy arises from the heat
d E = dq d w (11.5)
content of the system as well as from other internal
sources such as gravitation al, electrical, magnetic, m w here dq and dw are very small increm ents of heat and w
olecular, atomic, and even nuclear phenom ena. The am ork, respectively. The m ost com m on form of w ork a
ount of internal energy (E ) of a particular system is system can do during a change in its state is to expand
difficult to determ ine. However, its m agnitude depends against the constant pressure of the surroundings. Thus we
on the state of the system . W hen a system undergoes a can specify that:
change in its state in a reversible m anner, its internal
dw =PdV (11.6)
energy also changes. The m agnitude of the change in the
w here P is the external pressure and d V is an
internal energy AE of a system that undergoes a reversible
change in state from state 1 to state 2 is given by: increm ental change in the volum e of the system . A
simple dem onstration indicates that P d V does have the
dim ensions of w ork (force X distance):
AE=E2Ex (11.2) PdV= force X (distance) 3
= force X distance
w here E 2 and E x are the internal energies of the system (distance ) 2
in states 2 and 1, respectively. It makes no difference how (11.7)
the change in state is carried out, provided that it is Therefore, we can express the first law by the equation:
reversible, which requires that the system is never far
from being in equilibrium . If the system is returned to its
d E = dq - P d V (11.8)
original state, the net change in its internal energy is zero.
We con clude therefore th at the change in the internal
energy of a system undergoing a change in state along a
circular path is zero: 11.3 Enthalpy
The statem ent of the first law represented by equation 1 1 . 8
AE = (E 2 - E x) + (E x - E z) = 0 (11.3) enables us to predict the change in the internal energy of a
system during a change in its state. The equation indicates
The first law o f thermodynamics states that the th at the increase in the internal energy of the system is
increase in the internal energy of a system during a equal to the heat absorbed from the surroundings m inus
reversible change in its state is equal to the heat that the work done by the expansion of the system against the
flows across its boundary from the surround ings (q) pressure exerted by th e surroundings. We can
minus the work done by the system (w).
158 TH ER M O D Y N A M IC S
 This is a very im Standards, and in com
carry out such a change in portant and concrete their compounds. The pilations by Kelley (1962),
result because qP is a m reference state is defined in Robie et al. (1978), B arin
state by integrating
easurable quantity. For terms of certain standard et al. (1989), and Woods
equation 1 1 . 8 from state 1 to
example, w hen a chem properties that define the so- and G arrels (1987).
state 2 : However, it is im portant
ical reaction is carried called standard state. The
out in a well-insulated to rem em ber that
tem perature of the standard
d E = J dq vessel called a calorim standard enthalpies of
The integration yields: state is equal to 25C
eter, the heat evolved or form ation and other therm
(298.15 K ),the pres sure is 1
consum ed is indicated odynam ic constants are
atm (recently redefined as 1
E 2 - E , = (q2 - q j - by the change in tem measure m ents and are
bar, which is equal to 10 6
perature inside the subject to experim ental
W e simplify this result by replacing dyne/cm 2, or 0.987 atm,
calorim eter. Therefore, errors. Consequently, the
which is an am ount of h eat added at 29.53 in. Hg), and the
equation 11.16 enables values reported by
activity of ions and
pressure. In addition, we m us to m easure the different investigators for
ultiply the P V terms: molecules in aqueous
change in the enthalpy of a given com pound
solution is equal to one. In
E 2 - Ej = q p - a system as a result of a commonly dif fer from
addition, we specify that the
P V 2 + P chem i cal reaction, b u t each other by am ounts
V 1 (11.11) enthalpies of the pure elem
we still do not know the that may or may not be
ents in their stable state of
By rearranging the term s of am ount of enthalpy the within the error of the m
aggregation in the standard
this equation we obtain: system contains. easurem ents. Such
state are equal to zero. Thus
discrepancies in therm
(E 2 + P V 2) - we have the definition: odynam ic constants affect
In order to establish
a scale for enthalpy we the results of calculations
(E, + EV ,) = qP
The enthalpy o f in which they are used and
(11.12) now define a reference
formation of cause internal
It is advantageous at this state for chemical elem
compounds and their inconsistencies. Some of
point to define a new kind ents and ions and molecules in the
of energy called the aqueous solution is
enthalpy (H ): the heat absorbed or
given off by chemical standard enthalpies
H =E +PV reactions in which the of form ation
The enthalpy is a function of the state of the sys compounds, ions, and listed in
tem , as isthe internal energy.Therefore, the molecules form from A ppendix B are average
the elements in the values of two or m ore
enthalpy of the system in states
standard state. determ inations that
tfj = E, + P V X
appear to be com patible
We represent this im portant
H2=E2+PV2 with in the errors of m
param eter by the sym bol
easurem ent. However, no
Substituting equations 11.14 and 11.15 into equa script identifies the
assur ance can be given th
standard state and the sub
tion 1 1
.1 2
yields: script/sym bolizes form ation at . these values are
The standard enthalpies of accurate to within the
H2- Hx=qP=
other therm odynam ic constants stated num ber of
In other words, the change pounds, ions, and molecules significant figures or that
in the enthalpy of a sys tem have been m easured the listed values are
(AH ) during a reversible experim entally and are internally consistent.
change in its state at tabulated in the CRC H
constant pressure is equal andbook o f Chemistry and
to the heat absorbed by the Physics (W east et al.,
system during th at change 1986), in Technical N otes
in state. of the National B ureau of
 11.4 HEATS OF REACTION 159
4.1840 joules (J) so that 1 kcal/m ol is equal to
11.4 Heats of Reaction
4.1840 kJ/m ol. A t the present tim e both units are in use.
W hen two com pounds or elem ents, A and B, react
to form a product A 2 B3 by the reaction: W hen AHR of a reaction is positive, the reac tion is
2 A + 3 B A 2 B3 (11.17) endothermic, which m eans that h eat flows from the
surroundings to the system . Conversely, w hen AHR is
a certain am ount of heat is either used up or given off. We negative, the reaction is exotherm ic and heat flows from
can now calculate the am ount of this heat (.q P) from the the system to the surround
enthalpy change of the system when the reaction takes ings. E ndotherm ic reactions consum e heat,
place in the standard state. In other words, we treat the w hereas exotherm ic reactions produce heat.
reaction as a change in state and calculate the difference in In order to illustrate how heats of reaction
the enthalpy (AHr ) betw een the final state (A 2 B3) and are calculated, we choose th e reaction betw een
the ini tial state (2 A + 3 B). Therefore, the change in hydrogen and oxygen to form w ater in the
enthalpy of the reaction in the standard state is: gaseous state, sym bolized by (g):
2 H 2 (g) + 0 2 (g) - 2 H zO(g) (11.20)
AHR = Hf (A 2 B3) - [2Hf (A ) + 3 /T /B )] (11.18)
The am ount of h eat absorbed or given off w hen this
reaction occurs at 1 atm and 25 C is given by:
where AHR is the heat of the reaction in the stan dard AHR = 2 //} (H 2 0 (g ))
state and Hf is the standard enthalpy of for m ation of A 2
- [2H/ (H 2 (g)) + Hf ( 0 2(g))}
B 3, A, and B, respectively. Note that the heat is evolved
or absorbed as a result of, or during, the reaction and not The standard enthalpies of H 2 and 0 2 are both
when the reaction is at equilibrium . A fter a reaction has equal to zero by definition, so th at in this case:
reached equilibri um, the heat of the reaction (AH R) is
AHR = 2 //} (H 2 0 (g )) (11.22)
zero because the rates of the forward and backward
reactions are exactly equal. Thus, we can state that at The standard enthalpy of form ation of w ater vapor is
equilibrium AH R = 0. However, the reference state is a 57.80 kcal/m ol (A ppendix B ).Therefore:
defined condition in which equilibrium does not occur. AHR = 2 ( -5 7 .8 0 ) = -1 1 5 .6 kcal (11.23)
Therefore AHR is never zero. In general, AHR is
We conclude that this reaction is strongly exother mic. In
calculated by summing the standard enthalpies of the
fact, this reaction is com m only used to dem onstrate the
products and by subtracting the enthalpies of the reactants:
explosive nature of som e chem i cal reactions. N ote that
the reaction is endotherm ic w hen equation 1 1 . 2 0 is
reversed.
H eat effects are also associated with the dis sociation
AHR = 2 niHfi ( p r o d u c t s ) w , / / ) , (reactants)
of salts into ions, w ith phase transform a tions, and with
(11.19) mixing or dilution of solutions. For exam ple, heat is
absorbed from the surroundings w hen w ater evaporates to
w here n is the m olar coefficient of each reactant form w ater vapor:
and product taken from a balanced equation rep H 2 0 (l) -> H zO (g) (11.24)
resenting the reaction and i identifies the com
pounds or ions th at participate in the reaction. W hen the evaporation occurs in the standard state:
Standard enthalpies are expressed in units of AHR = H/ (H 2 0 (g )) - /T )(H 2 0 (l)) (11.25)
kilocalorie per m ole (kcal/m ol), where the nor
From A ppendix B:
m al calorie is defined as the am ount of heat
required to raise the tem perature of 1 g of w ater from AHR = (-5 7 .8 0 ) - ( - 6 8 .3 2 ) = +10.52 kcal (11.26)
14.5 to 15.5 C. The calorie is equivalent to
160 TH ER M O D YN A M IC S

Evidently, the evaporation of w ater at 25 C and 1 atm all of the heat that enters a system during a reversible
pressure is an endotherm ic process that rem oves heat from change in its state is available to do work. Similarly, the
the surroundings and thereby causes the tem perature to direction in which a reaction proceeds spontaneously is
decrease. The heat required to evaporate one mole of a not determ ined solely by the enthalpy change because
liquid is called the latent heat o f evaporation. W hen the systems have an additional property called entropy. We
vapor con denses to form the liquid, the heat given off is return to this phenom enon in Section 11.6.
called the latent heat o f condensation. In other words,
evaporation is endotherm ic, whereas con densation of w
ater vapor is exothermic, and the latent heat of evaporation
is equal in m agnitude to the heat of condensation. 11.5 Heat Capacity
Similarly, fusion or m elt ing of solids is endotherm ic,
W hen heat is added to a solid, a liquid, or a gas, the tem
whereas crystalliza tion of liquids is exotherm ic, and the
perature of the substance increases. This fam iliar
latent heat o f fusion is equal to the latent heat o f
observation can be stated as an equation:
crystallization.
dq = C d T (11.27)
N ote th at we are unable to calculate the
w here dq is an increm ent of heat added, C is a constant
latent heat of fusion or crystallization of w ater
known as the heat capacity, and d T is the corresponding
because it takes place at 0C rather than at 25 C.
increm ental increase in the tem perature expressed in
Therefore, in order to calculate the latent heat of fusion of
kelvins (K). In therm ody namics, tem peratures are
w ater, we m ust first extrapolate the enthalpies of form
always expressed on the absolute or Kelvin scale, which
ation from 25 to 0C. In gener al, chemical reactions on the
has its zero point at 273.15C, w hen all m olecular m
surface of the E arth occur at tem peratures betw een 0 and
otion stops and the heat content of pure crystalline m
100 C, and their heats of reaction therefore com m only
aterials becomes equal to zero.
differ from those calculated at 25 C.
The heat capacity is a characteristic property
The response of a reaction at equilibrium to a
of chemical elem ents and their com pounds that
change in tem perature can be predicted on the
m ust be m easured experim entally. It turns out
basis of Le C hteliers principle. If the forw ard
that heat capacities actually vary with tem pera
reaction is exotherm ic, an increase in the tem pera ture
ture and that it m akes a difference w hether the
favors the backw ard reaction because it con sum es heat
heat is added at constant pressure or at constant volume.
and therefore counteracts the increase in the tem perature.
On the surface of the E arth pressure is constant, and we
Consequently, the equilibrium will shift to the left m aking
therefore w ork with the heat capacity at constant
the equilibrium con stant smaller. Similarly, a decrease in
pressure symbolized by CP. Consequently, we restate
tem perature applied to a reaction at equilibrium in which
equation 11.27 as:
the forw ard reaction is exotherm ic causes the equilib rium
to shift to the right and increases the value of the dqP = CPd T (11.28)
equilibrium constant. Evidently, the enthalpy change of a
We now recall that the heat added to a closed system
chem ical reaction is related to the way in which the num
during a reversible change in its state at constant pressure
erical value of the equilibrium constant varies with tem
is equal to the increase in the enthalpy of the system . This
perature. This relation ship is form ally expressed by the
conclusion was expressed by equation 11.16, which we
vant H off equa tion, which will be derived in Section
now restate in terms of differentials:
11.10.

W e may also expect th at exotherm ic reactions take dqP = d H (11.29)

place spontaneously because they do not require heat from
Substituting equation 11.29 into 11.28 yields:
the surroundings. H owever, this generalization is not
correct. It turns out th at not d H = CPd T (11.30)
 11.6 THE SECOND LAW 161
+ c X K T j T d T (11.34)

By integrating this equation we obtain:

, , b X 10 3r , , s,n

HT- Hj. = a (T - T) + --- [T 2 - (T )2]

r v 1 r r 6 [T3 _ ( to)3] (n 3 5 )
+ CX^U
By carrying out the integration we obtain:

This equation can be integrated betw een appro priate tem

perature limits and therefore enables us to extrapolate
enthalpies of form ation of com
pounds from the standard tem perature (T ) to
som e other tem perature ( T ) above or below it:
dH= CP dT (11.31)

In the 19th century m uch w ork was done by m eans of

steam engines in which h o t steam expands in a cylinder
and pushes a piston. The steam cools as it expands and is
returned to the reservoir at a lower tem perature. A n im
portant scientific problem of th at tim e was to understand
the factors that limit the efficiency of steam engines. This
problem was solved in 1824 by the French engineer Sadi C
arnot (1796-1832). W ithout going into the details, we can
say that the efficiency of steam engines can never be 1 0 0 %
because som e of the h eat added to the system must be
discharged w hen the expanded steam returns to the
reservoir.

The work of C arnot was a very im portant contribution

to therm odynam ics and from it arose a new principle know
n as the second law o f thermodynamics. The essence of the
second law is
 E quation 11.35 can be applied to the reactants
H T - H r = CP( T - T) (11.32) and products of a chemical reaction as specified
E quation 11.32 enables us to calculate the by equation 11.19. As a result, we calculate Aa,
Ab, and Ac for the reaction and th en evaluate the
enthalpy of form ation of a com pound at tem pera enthalpy change of the reaction at tem perature
tu re T if we know its standard enthalpy and its T (AH R t ) from the equation:
heat capacity, provided the tem perature depen
dence of CP can be neglected. AH R T - AHR = Aa (T - T) (11.36)
In reality, use of equ atio n 11.32 to ex trap o + Ab X 10 3 [T 2 - (T )2]
late enthalpies is lim ited because the h eat capac
ity is in fact a function of th e tem p eratu re and + Ac X 10 6 r , - / \ n
therefo re cannot be treated as a constant in the ------ [T (T)]
integration . M oreover, the range of tem p era The relationship betw eenCP and T is som e
tures over which the integration is carried out
tim es expressed in ways th a tdiffer from eq u a
m ust not include phase transform ations, such as
tion 11.33, but the application of these equations
boiling or freezing, which cause a discontinuity
is lim ited by th e availability of experim ental
in th e h eat capacity and involve their own
data.
enthalpy changes. T hese restrictions generally
are not severe provided the environm ental con
ditions rem ain betw een 0 and 100 C at 1 atm
11.6 The Second Law
pressure. T herefore, the m ost straightforw ard
procedure is to recalculate the enthalpies of fo r
m ation of reactants and products to the desired
tem p eratu re (betw een 0 and 100 C) and then to
calculate the enthalpy change of th e reaction at
th at tem perature.
A m ore accurate and elegant procedure is
based on equations that express the variation of
CP as a function of tem perature in the form:
CP = a + (b X 10"3)T + (c X 106 )T 2 (11.33)

w here a, b , and c are constants derived by fitting an

algebraic equation to experim entally determ ined data in
coordinates of CP and T. Substituting equa tion (11.33)
into equation 11.31 yields:

d H = a I d T + b X 10 3 TdT
T JT T
162 TH ER M O D YN A M IC S
th at all systems possess a property called entropy, nam ed
by the G erm an physicist R udolf Clausius (1822-1888). O
ne way to explain entropy is to say that a certain fraction
of the enthalpy of a system is not convertible into w ork
because it is con sum ed by an increase in the entropy. A
nother way to explain the m eaning of entropy is to com
pare it to the property of random ness recognized in
statistical mechanics. W hen one tries to predict
the behavior of a collection of particles, one com es to the
conclusion th at they will m ost p ro b ably assum e a
condition of m axim um random ness. This conclusion
leads to the generalization th at every system that is left to
itself will, on the average, change tow ard a condition of
maximum random ness. The im plication of this statem ent
is th at the entropy of a system increases sponta neously
and that energy m ust be spent to reverse this tendency.
W hen entropy is thought of as random ness, it
can be recognized in m any natural phenom ena.
We have all learned from the com m on experi
ences of life that work m ust be done to counteract the n
atural tendency tow ard disorder. M any pop
ular science books have been w ritten about
entropy and its m anifestations in everyday life
including one by R ifkin (1980) who claim ed that
entropy is the suprem e law of nature and gov erns
everything we do.
The second law introduces entropy by the statem ent:
In any reversible process the change in the entropy
of the system (dS) is equal to the heat received by
the system (dq) divided by the absolute temperature
(T):
d S = ^jr (reversible process) (11.37)
In any spontaneous irreversible process the change in
entropy is greater than this am ount:
d S > C^ r (irreversible process) (11.38)
Entropy, like internal energy and enthalpy, is a property w
hose m agnitude depends on the state
of the system . However, unlike internal energy and
enthalpy, the absolute value of the entropy of a system can
be determ ined by virtue of the third law of therm odynam
ics.
M easurem ents indicate th at the heat capaci ties of
pure crystalline m aterials decrease with decreasing tem
perature and approach zero at absolute zero. This evidence
has been elevated into the third law o f thermodynamics,
which states:
The heat capacities of pure crystalline substances
become zero at absolute zero.
Since dq = C d T (equation 11.28) and
dS = d q /T (equation 11.37), it follows that:
dS= c(y) (11.39)
If the heat capacity is equal to zero at absolute zero,
thenafS = 0. Therefore, when a system com posed of pure
crystalline substances undergoes a reversible change in
state at absolute zero, the entropy change AS is equal to
zero. The G erm an physicist W alther H erm ann Nernst
(1864-1941) expressed these insights by stating that at
absolute zero the molar entropies of pure crystalline solids
are equal to zero.Therefore, the third law enables us to
calculate entropies by integrating equation 11.39 from
absolute zero to some higher tem perature (7):
1"-/.Mr)
(11'40)
In order to calculate the entropy of an elem ent or
compound in the standard state from equa tion 11.40, the
tem perature dependence of its heat capacity must be
determ ined experimentally. The standard molar entropies
of many elem ents and compounds have been determ ined
and are listed in reference books including the C RC H
andbook o f Chemistry and Physics (W east et al., 1986).
M olar entropies are expressed in entropy units or calo
ries per degree (cal/deg). W hen the standard m olar
entropies of the reactants and products of a chemi cal
reaction are known, the change in the entropy (ASR) of a
chemical reaction or phase transform a tion can be
calculated from the statem ent:
 11.7 GIBBS FREE ENERGY 163

ASR= 2 -^(p ro d u cts) 2 ,5(reactants) w here H is the enthalpy and S is the entropy. The change
in the Gibbs free energy of a chem ical reaction in the
(11.41) standard state is:
W hen ASR is positive, the entropy of the system AGR = AHR - TASR (11-47)
increases as a result of the change in state, w here as w
hen AS^ is negative, it decreases. For exam ple, when The Gibbs free energy is a function of the state of the
liquid w ater evaporates in the standard state to form w system like the internal energy and the enthalpy. It is
ater vapor: therefore subject to the sam e con ventions regarding the
H 2 0 ( l) -> H 2 0 (g ) (11.42) standard state as enthalpy and is m easured in kilocalories
per mole:
the change in entropy is (K rauskopf, 1979):
The standard Gibbs free energy o f formation of a
ASR = 45.10 - 16.71 = +28 .39 cal/d eg (11.43)
compound (Gf) is the change in the free energy of
the reaction by which it forms from the ele ments in
The increase in the entropy is consistent with our the standard state.
expectation that w ater m olecules in the gas phase are m
ore random ly distributed than m olecules of liquid water. N um erical values of Gf for com pounds and ions in
Similarly, w hen solid N aCl dissoci ates to form ions: aqueous solution are listed in A ppendix B together with
values of Hf . The value of AGR for a reaction is
calculated by summing the values of Gf of th e products,
N aCl(s) -> N a+ + Cl" (11.44) each m ultiplied by its m olar coefficient, and by
the change in entropy is (K rauskopf, 1979): subtracting from it the sum of the Gf values of the
reactants:
ASR = 14.0 + 13.5 - 17.2 = +10.3 cal/d eg (11.45)
AGR = 2 n , Gfi(products) (reactants)
Evidently, the entropy increases when N aCl dis solves to (11.48)
form ions in aqueous solution because the ions occupy The algebraic sign and m agnitude of AGR
fixed lattice positions in the crys tal w hereas the ions in
the solution are mobile. H ow ever, for other reactions depend on the sign and m agnitude of AH R and ASR in
ASR may be nega tive, which implies a reduction in the equation 11.47. If the forw ard reaction is exotherm ic
entropy of the system. (AHR is negative) and th e entropy increases (ASR is
positive), then the tw o term s of equation 11.47 com bine
to m ake AGR negative. If, how ever, the entropy
decreases (ASR is nega
tive), then TASR in equation 11.47 is positive. If
11.7 Gibbs Free Energy AHR is negative, A GR will be less negative than
AHR and may even becom e positive if TA S > AH .
The second law of therm odynam ics implies that the
N ote that TA S is in fact a form of enthalpy
increase in the enthalpy of a system during a reversible
because AS = A H /T at constant pressure and
change in its state at constant tem pera ture is dim inished
tem perature (equations 11.16 and 11.37).
because a certain am ount of the enthalpy is consum ed by
Therefore, TAS has the dim ensions of A H and the two
an increase in the entropy of the system . This insight was
term s can be added or subtracted from each other.
used by J. Willard Gibbs (1839-1903) to define a new
form of energy, now known as the Gibbs free energy (G): W hen AGR of a reaction is negative, the for ward
G=H-TS (11.46) reaction has excess energy w hen it occurs in the standard
state. W hen AGR is positive, there is
164 TH ERM O D YN AM ICS
a deficiency of energy for the forw ard reaction in the
standard state and the reaction will therefore run in the
backw ard direction. In general, AGR is a m easure of the
driving force of a chem ical reac tion such as equation
11.17:
2 A + 3 B > A 2 B 3
If AGR of this reaction is negative, the reaction pro ceeds
as w ritten from left to right in the standard state. W hen
the reaction ultimately reaches equilib rium , the product
is m ore abundant than the reac tants so that the
equilibrium constant K is large:
[ a 2 b 3] (11.49)
=K
[A ] 2
[B]:
If, however, A G R for reaction 11.17 in the standard
state is positive, the reaction runs the other way:
2 A + 3 BA 2 B 3 (11.50)
Therefore, th e equilibrium constant is small because the
reactants are m ore abundant than the product. Evidently,
the sign and m agnitude of AG^ of a reaction can be used
to predict both the direc tion in which a reaction runs in
the standard state and the m agnitude of the equilibrium
constant.
In conclusion, we calculate AGR for the fam il iar
reaction:
C a C 0 3 - Ca2+ + C O f (11.51)
calcite
using the data in Appendix B:
AGR = [(-132 .3) + (-126 .17)] - [(-269 .9)]
= +11.43 kcal (11.52)
Since AGR is positive, the reaction runs from right to left
in the standard state and calcite precipi tates. The
equilibrium constant for reaction 11.51 should therefore
be a small num ber, as we know it to be. It is not
surprising that the reaction (equa tion 11.51) runs to the
left in the standard state because the activities of the ions
in the standard state are each equal to one and the ion
activity product is therefore much larger than the equilib
rium constant.
We can also calculate AHR and ASR for reac
tion 11.51 using values of Hf from A ppendix B
and of Sf from K rauskopf (1979):
AHR =[(-129.7) + (-161.84)] - [(-288.6)]
= -2.94 kcal (11.53)
ASR =[(-12.7) + (-13.6)] - [(+22.2)]
= 48.5 c al/d eg (11.54)
N ote that ASR for the dissociation of calcite into ions is
negative, implying a decrease in the entropy. N ote also
that the entropy change is expressed in calories per degree
(cal/deg) rather than in kilocalories (kcal). We can now
com bine AHR and ASR for reaction 11.51 and calculate
AGr , thereby testing the internal consistency of the
therm odynam ic data:
AGR = - 2 .9 4 - 298.15(48.5)
(11.55)
1000
w here 298.15 is the tem perature of the standard state and
we divide the TAS term by 1000 in order to convert
calories to kilocalories.
The result is:
AGR = +11 .5 kcal (11.56)
which is in satisfactory agreem ent with AGR =
+ 11.43 kcal calculated from equation 11.52. N ote that
AGR for the dissociation of calcite into ions is positive
even though the reaction is exotherm ic, that is, AHR is
negative. This occurs because m ore heat energy is
consum ed to decrease the entropy than is produced by the
dissociation. Therefore, in this case AHR is not a reliable
predictor of the
direction in which the reaction runs in the stan
dard state.
The sign of AHR does, however, enable us to
use Le Chateliers principle to predict how the equilibrium
constant of reaction 11.51 varies with tem perature. Since
the reaction is exotherm ic as written, it produces heat
which increases the tem perature. Therefore, if reaction
11.51 is at equilibri um and the tem perature is decreased,
the reaction moves to the right because that tends to
increase
11.8 DERIVATION OF THE LAW OF MASS ACTION 165

the tem perature. W hen
equilibrium is
reestablished at a lower
tem perature, the
equilibrium constant is
larger than it was before.
Therefore, the solubility-
product constant of calcite
dG = dq - P d V + P d V
Substituting into 2 " A - ' 2 n iGi = R T '2 , n i \n P i
equation 11.60
Next, we assum e that the
Since dS = d q /T , dq = yields:
m ixture of ideal gases is
TdS and equation 11.59 the result of a chemical
therefore reduces to: dG = R T ^ p j
reaction and specify th at
Integrating from the standard pressure
dG = V dP the m olar coefficients (n)
som e other pressure (P): of the products are p o s
We can use equation 11.60
dG = itive and those of the

J
increases with decreas ing to calculate the change in
reactants are negative.
tem perature and calcite the Gibbs free energy of
The provision is not new
becomes m ore soluble. A one m ole of an ideal gas
or arbitrary but goes back
s we discussed in Section as a function of pressure at yields: to the convention
10.4, thermodynamics constant tem pera G P - G = R T (lnregarding
P the algebraic
provides a rational ture. Since the volume
of gases varieswith pres sign of heat. If we choose
explanation for the tem Since P = 1 atm , In P
sure, wem ust express to represent the chem ical
perature dependence of the
the volum e in = 0. Therefore, equation reac tion by the equation:
solubility of calcite and
equation 11.64 reduces to:
allows us to predict the
11.60 by an equation that
course of other chemical GP- G=RT\nP aA +bB^cC +dD
relates volume to pres
reactions provided the
sure. A ccording to the w here GP is the m olar
necessary data are
ideal gas law, for one mole free energy of the ideal
available.
of gas: gas at pressure P, G is the
m olar free energy in the
11.8 Deri standard state, and T is the
vatio V = tem perature of the
standard state in
n of where R is the gas kelvins.The m olar free
the constant. energy is also called the
Law chemical potential (/x),
and eq u a tion 11.65 can
of be w ritten as:
Mass
Actio ix = fjL + R T \ n P

n The free energy of n

Because the G ibbs free energy moles of an ideal gas can
term s of both the first and second be calculated by
modynamics, we can multiplying equation
express it in term s of 11.65 by n:
direct
ly m easurable properties. We n G P - nG = n R T
equation 11.46 by equation 11.13 and obtain: A ccording to D alto n s
G= E+PV- law, the total pressure of a
For an infinitesim al change in m ixture of ideal gases is
the sum of the partial
tem perature d T = 0 and therefore:
pressures exerted by each
dG = dE + P d V + V dP -
gas in the mixture.
Substituting equation 11.8 for Therefore, for a m ixture
of i ideal gases:
166 TH ER M O D YN A M IC S
then we can evaluate w hat equation 1 1 . 6 8 tells us
W hen the reaction reaches equilibrium , AG0, but
to do. For exam ple, 2 n ,G , means: AG^ rem ains the same.
2 ,-G,-(products) - 2 ,G ,(reactants) = AG R R eturning to equation 11.73 we can now state that
for a m ixture of ideal gases at equilibrium A G R = 0 and
(11.70) therefore:
AG* = -- R T l n K (11.74)
Similarly, 2 n ,G m eans:
where K is the equilibrium constant at the standard tem
perature. This rem arkable equation establishes the
2 n,G (products) 2 n,G (reactants) = AGR functional relationship that exists betw een AGR and K,
(11.71) both of which are constants that apply to chemical
reactions in the standard state and at equilibrium,
and R T 'Z n i ln P, m eans: respectively. In the process of deriving this relationship we
have also proved the validity of the Law of Mass Action
R T {c In P c + d In P D a In P A b In P B)
for reactions among ideal gases. These two results are
= R T (ln P cP d - In P%P'f) among the m ost im por tant contributions
(11.72) thermodynamics has m ade to science. Equation 11.74
=RTIn =RTlnQ allows us to calculate the numerical values of equilibrium
\pxpii constants at 25 C for all chemical reactions for which we
can calcu late A GR from existing m easurem ents of the
w here Q is the reaction quotient. Therefore, we
stan dard free energies of formation.
now restate equation 1 1 . 6 8 as:
AG R - AGR = R T ln Q (11.73)
We reduce equation 11.74 to a form ula by converting
the equation to logarithm s to the base 10 and by
which is valid for a reacting m ixture of ideal gases. T he substituting R = 1.987 cal/deg mol and T = 298.15 K:
reactio n q u o tien t ( Q ) varies continu ously as the
reactio n proceeds until the reaction
2.3025 X 1.987 X 298.15 log K (11.75)
reaches a state of equilibrium . A fter equilibri AGR
um has b een established, the partial pressures 1000
of th e reactan t and p ro d u ct gases becom e con sta n t = -1 .3 6 4 log K (11.76)
because th e forw ard and backw ard re a c tions are
occurring at the sam e rate . T herefore, at equilibrium th e The right side of equation 11.75 m ust be divided by 1000
reactio n q u o tien t becom es constant and is know n to us because R is expressed in calories, whereas AGR is in
as th e equilibrium constant. kilocalories. Therefore, the equilibrium constant K is
related to AGR by the formula:
N ext, we consider a very im portant point: the free K = 1 0 - 4 G 1 /1 J 6 4 (11.77)
energy change of a chem ical reaction at equi librium is
equal to zero. This statem ent becom es obvious if we This is a very useful result, but it applies only to reactions
recall th at AG R is regarded as the driving force of a chem am ong ideal gases unless we take steps to m ake it
ical reaction . W hen a reac tion is in equilibrium , th ere is applicable to real gases and to ions and molecules in
no driving force pushing it one way or the other. H ow aqueous solutions.
ever, note
th at AGR is not zero because it is the change in
11.9 Fugacity and Activity
the G ibbs free energy w hen th e reaction takes
place in the standard state. Therefore, AG R is a constant, In order to m ake equations 11.74 and 11.77 applicable to
w hereas AG R varies as the reaction pro ceeds from the reactions am ong real gases we define the fugacity as the
standard state tow ard equilibrium . partial pressure a real
 11.9 FU CA C ITY AND ACTIVITY 167
gas would have if it were ideal. Therefore, fugaci-ties The m ole fraction of A in a dilute solution can be
satisfy equation 11.74. In ord er to apply equa tion 11.74 transform ed into the m olal concentration of A (m A) by
to real solutions of any kind we define the activity (a) of a converting (B ) to kilogram s of B:
solute as: (A ) X 1000
(11.85)
L (11.78) (B ) X M W (B)
a=
r w here M W (B) is the gram -m olecular weight of B. Since
w h e re /is the fugacity of the solute in the solution a n d / 0 n (A )/ (B ) = N A in dilute solutions of A in B:
is its fugacity w hen the vapor is in equilib rium with the
pure substance in the standard state. Since fugacities are m mt ( 11.86)
M W ( B)
easured in atm os pheres or other units of pressure, the
activity as defined by equation 11.78 is a dimensionless and k ' is a constant equal to 1000/M W (B). W hen we
substitute equation 11.86 into equation 11.83, we obtain:
num ber. It also follows from equation 11.78 that
for a pure substance:
k kmA (11.87)
a = 1.0 (11.79)
a * = AC = -------
fA fAk '

because the fugacity / of a pure substance is equal to f .

By combining the three constants in equation 11.87 we
obtain:
Binary m ixtures of two nonelectrolytes A and B obey
R aoults law, which states that the fugaci 7Am A (11.88)
ty of com ponent A is given by: w here yA is the activity coefficient.
/A = / (11.80) This relationship applies to all solutions regardless of
the vapor pressures of the solutes or of the solvent. If the
w here N A is the mole fraction of A defined as: solution is a m ixture of volatile nonelectrolytes such as
acetone dissolved in w ater, we have no difficulty detecting
Wa = "<A > (1 1 .8 i)
A n( A ) + (B ) v ' the pres ence of acetone vapor and the definition of the
activity of acetone in the solution given by eq u a tion
w here (A ) and n(B ) are the num bers of m oles of A and 11.78 m akes sense. Ions produced by the dis sociation of
B, respectively. W hen the solution is dilute, salts, acids, and bases in w ater also produce vapors, but
th at is, (A ) (B ), the vapor obeys H enrys their partial pressures or
law:
/A = (11.82) fugacities are so low th at they may be u n d e
tectable. H ow ever, this experim ental difficulty
By dividing both sides of equation 11.82 by fA, we does n o t alter the validity of relationships derived
obtain: by consideration of vapors in equilibrium with
solutes and solvents regardless of w hether they
^ = ^ N A = aA (11.83) are electrolytes or nonelectrolytes.
I A / A
T he concept of activity presented above is quite
Next, we note that for a dilute solution of solute A in different from th at discussed in Sections 9.2 and 11.3 We
solvent B the num ber of m oles of A is much sm aller can reconcile the tw o representations by saying th at
than the num ber of m oles of B. Therefore, activities are introduced in th erm o
if n(A ) (B ), equation 11.81 can be restated dynam ics in ord er to extend the Law of Mass
to a very good approxim ation as: A ction from reacting m ixtures of ideal gases to
N =^ (11.84) chem ical reactions am ong m olecules and ions in
nonideal aqueous solutions. This is accom plished
A n( B)
168 TH ER M O D YN A M IC S

by m eans of activity coefficients th at convert The result of the integration is:

m olal concentrations of ions and m olecules into activities. A /A J ~ Y )(11-96)
W hen activity coefficients are regarded as dim ensionless In K T In K r = \ -
conversion factors, then activi ties have the sam e units as R
concentrations. E quation 11.96 is the im portant vant H off equa tion,
which we now restate in term s of logarithm s to the base 1 0
:
11.10 The van't Hoff Equation 1
AHR (1
Therm odynam ics enables us to calculate the equi librium log K T = log K TO 298.15)
2.30257? \ T
constants of chem ical reactions at 25 C. In order to
(11.97)
extrapolate the equilibrium constant from the standard tem
perature to other tem pera tures betw een 0 and 100 C, we N ote that equation 11.97 can be used to calculate K at any
return briefly to equation 11.74: tem perature betw een 0 and 100 C if we know the value
of K at the standard tem perature.
However, if the reference tem perature differs
AGfi = - R T ln K
from 25 C, then AH R m ust be evaluated at that
By solving this equation for ln K we obtain: tem perature by m eans of heat capacities, as
AG (11.89) shown in equation 11.32 or 11.35.
In* = We have now com pleted this selective p re
RT sentation of therm odynam ics, which has enabled us to
N ext, we differentiate equation 11.89 with prove the validity of the Law of Mass A ction and to
respect to T: calculate equilibrium constants of chem i cal reactions not
only at the standard tem perature but also at other tem
peratures betw een 0 and
<- 100 C.
Since AGR = AHR TSSR, it follows that:
A G AHR k no (11.91)
-r = = r ~ ASR
and: 11.11 Solubility of Amorphous
Silica between 0 and 100C
dsc yr _ a H R (11.92)
dT t
0
A t the end of C hapter 9 we calculated the solubil
T 1 - ity of am orphous silica in w ater as a function of pH
Substituting into equation 11.90 yields: at 25 C. We are now in a position to investi gate
1
dlnK a m AH I (11.93)
how the solubility of am orphous silica varies as a
dT \ R)\ T2 RT2 function of tem perature at different values of
M ultiplying by d T gives: the pH . The therm odynam ic constants required
,, / A //p \ / d T for this calculation are listed in A ppendix B.
(11.94) A m orphous silica reacts with w ater to form
d ln K =
/ brl silicic acid:
S i0 2 (am orph.) + 2 H zO (l) -+ H 4 SiC> 4 (11.98)
We integrate equation 11.94 from the standard tem
perature ( T ) to som e other tem perature (T), which m ay
The change in Gibbs free energy of this reaction in the
be higher or lower than T:
standard state is:
AGfi = [(-312 .66)] - [(-203 .33) + 2(-56 .687)]
R
= +4.044 kcal (11.99)
 11.12 SUMMARY 169
and the equilibrium constant K r is: C onsequently at 0C, K T = IO-1019, w hereas at 100C,
K r = 10-868. O nce again, the equilibrium constant
K r = i o - (4044h-364) = IO296 (11.100)
increases with rising tem perature because reaction 11.104
The change in enthalpy in the standard state is: is endotherm ic.
These calculations illustrate the point that therm
AHR = -349 .1 - | 215.94 + 2 (68.315)] odynam ics no t only confirms our qualitative predictions
based on Le C hteliers principle but actually perm its us
= +3.47 k c a l / (11.101)
to determ ine the num erical values of the equilibrium
The reaction is endotherm ic, suggesting that its constants of all kinds of chemical reactions betw een tem
equilibrium constant increases with tem perature. From peratures of 0 and 100 C or other appropriate tem perature
the vant H off equation (11.97): limits.
A lthough this is a very im portant advance in
log K T = geochemistry, the calculations can becom e quite tedious.
3.47 X 103 (1 1 \ , Therefore, it is helpful to sum m arize the results of therm
~~2 -9 6
~~ 2.3025 X 1.987VT ~~ 298.15/ (1L102) odynam ic calculations by m eans of suitable diagrams. In
the case at hand, we w ant to depict the concentration of S
w here R = 1.987 c a l/d e g -m o l. N ote th at AHR
i0 2 in saturated solutions of am orphous silica at different
m ust be converted to calories in order to be com patible values of the pH and tem perature. We therefore carry out
with R. E quation 11.102 reduces to: the calculations first m ade in Section 9.6 at a series of tem
peratures for selected p H values and plot the results in
\o g K T = -2 .9 6 - 7 5 8 .4 0 ; - Figure 11.1 as a set of curves in coordinates of the
0.003 35 j
concentration of S i0 2 and tem perature. The diagram
(11.103)
depicts the increase in the solubility of am orphous S i0 2
W hen T > 298.15 K, l / T - 0.003 35 is negative with increasing tem perature at specified values of the
pH . Such dia gram m atic representations of chem ical
and the algebraic sign of the second term in equa reactions are widely used in geochem istry and therefore
tion 11.103 changes to plus. Therefore, at are the subject of C hapter 12.
T = 273.15 K (0C) equation 11.103 yields
K T = 1(T3-20, whereas at 373.15 K (100C)
K T = 10-2'45. Evidently, the equilibrium constant
of reaction 11.98 increases with increasing tem per
ature as predicted.
The first dissociation of silicic acid forms 11.12 Summary
H 3 S iO / and H +:
H 4 S i0 4 -> H 3 S i0 4 + H + (11.104) Therm odynam ics is based on three statem ents regarding
the relationship betw een heat and other forms of energy
The standard free energy change is A GR = that do not require proof because they are self-evident.
Starting from these basic premises, therm odynam ics has
+13.24 kcal, and the standard enthalpy change is AHR =
grown into a large body of knowledge by rigorous
+ 7 .0 kcal. N ote th at the Gf and Hf of H + are both deductive reason ing aided by m athem atics. The products
equal to zero. The equilibrium con of this developm ent include the concepts of enthalpy and
stant of reaction 11.104 at 25 C is K = 10- 9 7 1 and Gibbs free energy and a proof of the Law of Mass A ction
varies with tem perature in accordance with the for reacting m ixtures of ideal gases.
vant H off equation:
log K j = -9 .7 1 - 1 5 3 0
.0 3 0 - 0.003 35j These results of therm odynam ics have great
practical value in geochem istry because they p er
(11.105) mit us to calculate the equilibrium constants of
170 TH ERM O D YN AM ICS
chem ical reactions at any desired tem perature,

lim ited only by the availability of the necessary

therm odynam ic param eters and constrained by
the stabilities of the reactants and products. Consequently,
we can now explore the behavior of chem ical reactions on
the surface of the E arth and use the results to specify the
stability limits of the com m on rock-form ing minerals. M
any of the reactions th at take place in the natural environ
m ent can be studied in the laboratory only with great
difficulty because these reactions are so slow. Therm
odynam ics enables us to m ake predic tions about the
outcom e of these reactions that can be tested by evidence
in the field.

Figure 11.1 Solubility of amorphous silica as a func

tion of temperature at different pH values. The equilib
rium constants at 25 C were calculated from equation
11.77 using the standard free energies of formation
from Appendix B. Equilibrium constants at other tem
peratures were derived from the vant Hoff equation
(11.97) based on K and kHR. The activity coefficients
of all ions and molecules were assumed to be equal to
one. The resulting curves indicate that the concentra
tion of S i0 2 in milligrams per liter in saturated solu
tions of amorphous silica in water increases both with Temperature, C
increasing temperature and pH.

Problems
 1. Calculate the
enthalpy change when
fluorite (CaF2) dissolves in
water in the standard state.
2. Based on the result in
Problem 1, predict how the
solubility of fluorite varies
with temperature.
3. Calculate the
standard free energy
change of the reaction by
which fluorite dissociates
into ions.
4. Calculate the
solubility of fluorite in
water at 10, 20, and 30 C
and express each in terms
of the concen tration of
Ca2+ in mg/L. (Assume y =
1 for all species.)
5. Calculate K r for the
(NaAlSijOg), water, and
(Al2Si20 5(OH)4) and silicic acid.
6. Apply the Law of
Mass Action to the
reaction in Problem 5,
take logarithms to the
base 10, and express the
resulting equation in
terms of log [Na+]/[H +]
and log [H4S i0 4].
7. The water of the
Mississippi River at
Baton Rouge, Louisiana,
has the following
chemical composi tion at
pH = 7.2.
 REFERENCES 171
ppm ppm Mississippi River at Baton Rouge; that is, determine

HCO3 101 Mg2+ 7.6 whether albite or kaolinite is produced.

9. Make the necessary calculation (assuming
sof 41 Na+ 11 T = 25 C) to determine whether the Mississippi River
cr 15 K+ 3.1 at Baton Rouge is saturated or undersaturated with
N O 3- 1.9 S i0 2 5.9 respect to (a) calcite, (b) gypsum, (c) amorphous silica,
Ca2+ 34 (d) quartz, and (e) dolomite.

10. Calculate the value of Gf for BaF2 given that the

Calculate log [Na+]/[H +] and log [H4S i04] for this water concentration of Baz+ in a saturated solution at 25 C,
using the necessary activity coefficients. having an ionic strength of 0.01, is 1.12 X 102 m/L.
8. Use the results of Problems 6 and 7 to determine the Assume that Ba2+ and F are derived entirely from BaF2
direction of the reaction in Problem 5 in the and that hydrolysis is negligible.
References
A N D E R S O N , G. M ., 1996. Thermodynamics.
Thermodynamics o f Natural Blackwell Scientific, Palo
Systems. Alto, A ll pp.
Wiley, New York, 382 pp. R IF K IN , J., 1980. Entropy.
B A R I N , I., F. S U E R T , E. SC H U L T Bantam, New York, 302 pp.
Z E - R H O N H O F , and W. S. SH E N R O B I E , R. A., B. S. H E M IN G W A Y
G , , and J. R. F I S H E R , 1978.
1989. Thermochemical Thermodynamic
Data o f Pure Substances,
properties of minerals and
VCH
related sub stances at
Publishers, New York, 1739 pp.
F R A S E R , D. G. (Ed.), 1977.
298.15 K and 1 bar
pressure and at higher
Reidel, Amsterdam, 410 pp.
K E L L E Y , K . K ., 1962. H e a ts o f tem
peratures. U.S. Geol. Survey Bull.
a n h y d ro u s silicates. U.S. Bureau o f Mines
W E A S T , R. C., M. J. A S T L E , and W. H.
5901.
CRC Handbook o f Chemistry and Physics.
K E R N , R ., and A . W E IS B R O D ,
1967. Boca Raton, FL.
Thermodynamics for W OOD , B. J., and D. G. F
Geologists. W. H. Thermodynamics for Geologists.
Freeman, New York, 304 Oxford, England, 303 pp.
pp. W O O D S , T. L., and R. S.
K R A U S K O P F , K. B., 1979.
Introduction to Geochemistry, Values at Low Temperature
2nd ed. for Natural Inorganic
Materials.
McGraw-Hill, New York,
617 pp. Oxford University Press,
New York, 242 pp.
M O O R E , W. J., 1955. Physical
Chemistry, 2nd ed. Prentice-
Hall,
Upper Saddle River, NJ,
633 pp.
N O R D S T R O M , D. K., and J. L.
M U N O Z , 1986. Geochemical
 12
Mineral Stability
Diagrams
W hen m inerals participate in chem ical reactions as
reactants or products, they are either con sum ed or
produced depending on the direction
in which the reaction is moving. We can th e re
fore say that reactant m inerals are unstable,
w hereas p ro d u c t m inerals are stable while the
reaction is in progress. If the reaction ultim ately reaches
equilibrium , b o th reactant and product m inerals becom
e stable. Thus, we can investigate the stability of m inerals
under two kinds of con ditions:
1. We can determ ine the environm ental conditions
necessary for m inerals to coex ist at
equilibrium .
2. We can determ ine which m inerals are sta ble
and which are unstable in a particular geochem
ical environm ent.
The theoretical basis for these kinds of studies is provided
by the Law of Mass A ction and by the relationship betw
een the standard free energy change of a chem ical
reaction and its equilibrium constant at 25 C.
The activity diagram s we w ant to construct are
based on the in co n g ru en t solution of certain alum
inosilicate m inerals in igneous and high-
grade m etam orphic rocks. T he reactions on
which the diagram s are based contribute to
chemical w eathering and result in the form ation of
oxides, hydroxides, clay m inerals, or zeolites, depending
on th e geochem ical environm ent. Transform ations am
ong m inerals that form only at elevated tem perature and
pressure are in ap propriate in activity diagram s draw n at
the tem
perature and pressure of the E arth s surface. However,
the stability of m inerals can also be defined by equations
arising from their c o n g ru ent solution in dilute
aqueous solutions. B oth kinds of activity diagram s p
resen t com plex chemical interactions in a form that is
especially appropriate for geologists accustom ed to using
phase diagram s in igneous petrology.
12.1 Chemical Weathering of
Feldspars
The feldspars in igneous rocks crystallize from silicate m
elts at high tem perature, but can also form on the surface
of the E arth by reactions of aqueous solutions w ith other
alum inosilicates such as the clay minerals. In Section
10.6 we dis cussed how to construct the equations th at re
p resent the reactions am ong aluminosilicates, but we
were unable to proceed because we did not know the
equilibrium constants of the reactions we w anted to
investigate.
Now we can restate the equation for the con
version of m icrocline to kaolinite:
2 KAlSi3 0 8 + 9 H 20 + 2 H + ^
A l2 Si2 05 ( 0 H ) 4 + 2 K + + 4 H 4 S i0 4(12.1)
Using the standard free energies of form ation
from A ppendix B we obtain:
A GR = [-(906 .84) + 2 ( 67.70) + 4(-312 .66)]
- [ 2( - 894.9) + 9 (56.687)]
= +7.103 kcal (12.2)
172
1 2.1 CHEM ICAL W EATHERING OF FELDSPARS 1 73
 than required
Therefore, the equilibrium equilibrium at a particular
value of log [K +]/[H +],
constant of the reac tion at
25 C is: reaction 12.1 moves to the
R = 1 0 -(7.103/1.364) =left, thereby converting
1 0 -5.21
kaolinite into microcline.
A pplying the Law of M Therefore, microcline is
ass A ction when the reac stable under these condi
tion is at equilibrium tions, w hereas kaolinite is
unstable. The same rea
gives:
soning indicates that if log
[H 4 SiQ 4 ]4 [iO
[H 4 S i0 4] is less than
[H + ] 2 required for equilibrium ,
kaolinite is stable and m
w here the square brackets
icrocline is unstable at 25
signify activities or
C.
fugacities of reactants and
products. N ext, we take
logarithm s of equation
12.4:
4 log [H 4S iO j + 2 log [K +] -
Collecting terms:
4 log [H4S iO j + 2(log [K +] -
for A HR = [(-9 8 3 .5 ) + 2 ( 60.32) + 4
We might have
chosen to write an - [ 2 ( - 948.7) + 9(68.315)]
equation for the = +14.895 kcal
conversion of microcline
to gibbsite instead of We see that the reaction is endothe
kaolinite. The resulting occurs in the standard
line plots to the left of the state, which m eans th at it
m icrocline-kaolinite consum es h e a t in th e
equilibrium in Figure 12.1. forw ard d irection .
This means that Therefore, an increase in
microcline is reacting to
tem perature favors the
form gibbsite under conversion of m icrocline
conditions in which to kaolinite and causes the
kaolinite, not microcline,
equilibrium constant to
is stable. Consequently,
increase. The rela tionship
the transform a
betw een K T and K T is
tion of microcline to
given by the vant H off
m ust be excluded from the diagram .
equation:
W hen reaction 12.1 occurs in a closed
ro n m en t having a sm all w a te r/ro c k
log [H 4 S i0 4] and log [K +]/[H +] m ay change A
th e rea ctio n p ro g resses until eq u ilib riu m H
established. In th at
R
case, kaolinite and m
icrocline coexist in X

from which it follows that: equilibrium w ith the

aqueous solution, 1
4 log [H 4S iO j + provided the tem perature 0
2 logf pj = the activities of th e ions 3
unchanged. O n the other /
and: 1
m ent is open or the w log K T = log K TC -
ater/ro ck ratio is large, the 2.30257? \T
[K
reaction m ay not achieve
l0 g [H T] = 2 l0g
equilibrium because it is
E quation 12.8 is a straight line, shown in Fig
unable to change the
ure 12.1, along w hich log activity of H 4 S i0 4 and
log [H 4 S i0 4] have the the [K +]/[H +] ratio of the
values required for environm ent enough to m
ake them com patible w ith
equilib rium betw een
equation 12.8. In this case,
m icrocline and the reaction continues until
kaolinite. If the two either kaolinite or m
param eters have values that place the environ icrocline is used up,
m ent o f f the line, the reaction is depending on the direc
um . In that case, we can tion in which it is moving; th at is, the
determ ine the direction goes to com pletion.
in which th e reaction We can investigate the effect of a
Le C h te lie rs principle tu re change on reaction 1
log [H 4 S i0 4] of the
environm ent is greater
1 74 MINERAL STABILITY DIAGRAMS

log [H4S i04]

Figure 12.1 Stabilities of microcline, muscovite (proxy for illite), kaolinite, gibbsite, and amorphous silica in
equilibrium with K+, H +, and H4S i0 4 in aqueous solution at 25 C and 1 atm pressure. The equations for the
stability boundaries are based on the standard free energy values from Appendix B and are listed in
Table 12.1. Note that K-smectite, pyrophyllite, and quartz are omitted from the diagram in order to keep it simple.
Note also that the diagram lies in the upper left quadrant of the coordinate system. The reaction path was adapted
from Steinmann et al. (1994).
12. 1 CHEM ICAL W EATHERING O F FELDSPARS 1 75

Since K T. = IO 5 21, AHR =- H C l- H 2 0 . These
T = 308.15 K (35 C):
log K T =
- 5 .2 1 -
K(35 C) = K T 4 8 5
Since 10- 4 ' 8 5 > 10-521, th e
to the right in favor of
kaolinite with increasing
tem perature, as predicted
. pyrophyllite are not shown
W hen reaction 12.1
proceeds from left to right
as w ritten, it produces K
+ ions and H 4 S i0 4 m
olecules while consum ing
H + and H 2 0 . T here
fore, if the system
is closed,
sition of the w ater changes in thesuch
log [H 4 S i0 4] and
log [K +]/[H
com patible with the
values required for
equilibri um. Therefore,
chemical w eathering
alters the chemical com
position of w ater and
hence its quality as a
natural resource. M microcline, it forms sec
oreover, reactions
such as 1 2 . 1 take place not only on the surface of
the E arth but also below
groundw ater is in contact
with m inerals that are
unstable in the presence of
w eathering is n o t restricted
exposed to the w eather, but
depth below the surface
depending on the loca tion
of the w ater table and the
porosity and p er m
eability of the rocks.
We can now com
plete the construction of
the stability diagram for
microcline (Figure 12.1)
by adding o th er
crystalline or am orphous
com pounds that can
occur in the system com
posed of Si0 2 - A l2 0 3 - K 2 0
include gibb-site, m the solids in adjoining
uscovite (proxy for illite),
14 895
K -smectite, pyro-phyllite,
and am orphous silica. N
ote that HC1 serves as a
source of hydrogen ions
and is not actually present
in natural environments. N
ote also that the stability
fields of K -smectite and
in Figure 12.1, but they do
occur betw een kaolinite
and microcline. The
equations required
define the stability limits
of the solid phases in this
system
are derivable
the chemical com po by
same
a wayprocedure
that we
dem +]
onstrated
becom eabove
m orefor
reac tion 12.1 and are
therefore listed in Table
12.1.
The boundaries in
Figure 12.1 are the loci of
points whose coordinates
have the values required
for equilibrium between
fields. Each field is
labeled to indicate which
solid is stable within its
sufficient time.
boundaries. The solids are
stable only within their
stability fields. W hen they
are exposed to an
environm ent that lies
outside their own stability
field, they are converted
into the solid that is stable
in that environm ent. For
example, if a grain of
to microcline is placed into
an environm ent within the
stability field of kaolinite,
reaction 1 2 . 1 will run from
left to right and convert
the m icro cline into
kaolinite, provided enough
time is avail able to allow
the reaction to run to
completion. Similarly, if
kaolinite is placed into an
environment within the
stability field of
ondary K-feldspar known
as adularia (Kastner, 1971;
Kastner and Siever, 1979;
Mensing and Faure, 1983).
M ost natural w aters
plot w ithin the stability
field of kaolinite in Figure
12.1, w hereas seaw ater
straddles the boundary
betw een m uscovite
(illite) and microcline. T
herefore, m icrocline and
other
feldspars in igneous and
the continents w eather
(kaolinite and sm ectite).
W hen these clay m iner
als are deposited in the
oceans, they tend to take
up K + and m ay be
converted into illite (rep re
sented by m uscovite in
Figure 12.1) or low -tem -
p erature feldspar, given
The evolution of the
chem ical com position of
w ater as a result of chem
ical w eathering of m in
erals such as m icrocline can be pre
num erical m odeling (H elgeson et al.,
progress of such reactions
produces curved reac tion
paths on activity diagram s.
A n exam ple of such a
path is show n in Figure
12.1 based on m od eling
by Steinm ann et al.
(1994). T heir calcula tions
w ere based on the assum
ptions that:
1. The system is
closed and the w
ater/ro ck ratio is
small.
2. The secondary m
inerals are in
equilibrium with
ions in the
solution even
though the
prim ary m inerals
are n o t in
equilibrium .
3. Tem perature and
pressure rem ain
constant.
4. A lum inum is
conserved in the
reactions.
5. The solutions rem
ain dilute such th
at all activity
coefficients are
close to unity.
1 76 MINERAL STABILITY DIAGRAMS
Table 12.1 S tab ility R e la tio n s A m o n g M icro clin e, K ao lin ite,

G ib b site, a n d A m o rp h o u s Silica in th e P re se n c e o f W a te r a t 25 C
a n d 1 a tm p re s s u re 3
Microcline-kaolinite

2 KAlSi3Os + 9 H20 + 2 H + ^ Al2Si2Os(OH)4 + 2 K+ + 4 H4S i0 4

AGR = +7.103 kcal K = 105-21
log | | i | = ~2 log [H4SiO j - 2.60 (1)

Kaolinite-gibbsite

Al,Si20 5(0 H )4 + 5 H20 ^ 2 Al(OH)3 + 2 H4S i0 4

AGR = +12.755 kcal K = 109-35

log [H4S i0 4] = -4 .68 (2)
Microcline-muscovite

3 KAlSi30 8 + 12 H20 + 2 H + ^ KAl3Si3O10(OH)2 + 2 K+ + 6 H4SiQ4

AGR = +16.184 kcal K = IO118

lo g j |J j = -3 1 o g [H 4S i04] - 5 .9 3 (3)

Muscovite-kaolinite

2 KAl3Si30 H1(0 H )2 + 3 H20 + 2 H + ^ 3 Al2Si20 5(0 H )4 + 2 K+

A GR = -11.059 kcal K = 10+8U

l0gf ] =+405 (4)

Muscovite-gibbsite

KAl3Si30 ]u(0 H )2 + 9 H20 + H + ^ 3 Al(OH)3 + K+ + 3 H4S i0 4

AGR = +13.603 kcal K= IO997

log j |^ j = - 3 log [H4SiO j - 9.97 (5)

Solubility limit of amorphous silica

S i0 2(amorph.) + 2 HzO ^ H4S i0 4

A GR = +3.604 kcal K = IO' 264

log [H4SiO j = -2 .64 (6)
aBased on the thermodynamic data in Appendix B.
12.1 CHEMICAL WEATHERING OF FELDSPARS 177
 incongruently to m ultiplying
W hen form kaolinite equation 12.14 by reactions
microcline is (equation 1 , Table 1/2 and adding it to represented by
exposed to an 12.1): equation 12.15
12.13. The
environ ment cause the [K+]/[H
represented by the +] ratio of the w
coordinates of point 2 K A lSi3 0 8 ater to rise along
A in + 9 H 20 + 2 H the gibbsite
Figure 12.1 (i.e., log [H + - ( 1 2 1 4 ) A l2 -kaolinite boundary
[H +] = 2.0, pH = 4), it will react to form gibbsite: from point B to
Si2 0 5 ( 0 H ) 4 + point C, which is
K A lSi3 O s + 7 H 20 + H + - A l(O H
2K++4H4S reached when all of
+3H the gibbsite has
i0 4
As a result, the been converted to
This reaction kaolinite.
the w ater both increase because
releases silicicthe acid Subsequently,
assum ed to be and also causes the m icrocline reacts w
closed and the w [K+]/[H +] ratio of ith ions in the w ater
ater/rock ratio is the w ater to to form kaolinite in
small. Therefore, increase. If accordance with
the chemical com equilibrium betw equation 12.14. A s
een gibbsite and a result, the [K +] /
position of the
kaolinite is m [ H +] ratio as well
aqueous phase
aintained, the as [H 4 S i0 4] b oth
changes along the
concentration of increase along the
reaction path from
silicic acid in the w reac
A to B. ater m ust rem ain tion path betw een points C an
constant until all of th at enough m icrocline is ava
W hen the
the previ ously reaction 12.14 to continue.
chemical com
form ed gibbsite If the chemical com position
position of the aque
has been converted reaches the stability
ous phase reaches to kaolinite. U nder boundary betw een
point B, kaolinite these conditions,
becom es the stable kaolinite and m
the reactions at
phase. Therefore, uscovite (illite),
point B in Figure
the previously form kaolinite becom es
12.1 can be
ed unstable and reacts
represented by one
gibbsite reacts with silicicequation w ritten in with the ions in the
form kaolinite: such a way that Si w ater to form m us
2 A l(O H is con covite (equation 4,
served (Steinm ann Table 12.1):
A l2 Si2 0 et al., 1994):
3 A l2 Si2 0
KAlSi3 0 8 + 2 5 ( 0 H )4
This reaction
A l(O H ) 3 + H + 2 KT
consum es silicic (12.16)
+ ->
acid, but does not ^ 2 K A l3 Si3 O 1 0
affect the [K+]/[H (O H ) 2 + 3 H
+] ratio of the water. 1.5 A l2 Si2 0 5 (O 2Q + 2 H 4
H ) 4 + K + + 0.5
A t the sam e tim e,
H zO This reaction causes
m icrocline
a decrease of the [K
continues to This equation can +]/[H +] ratio of the
dissolve be constructed by
w ater w ithout affecting
the activity of [H 4 S i0 4].
H ow ever, if m icrocline is
still avail
able, it will now form m uscovite (equation 3,
Table 12.1):
3 KAlSi3 O s + 12 H 20

K A l3 Si3 O 1 0 (O H ) 2 + 2 K + +

This reaction increases

both the [K+]/[H +] ratio
and [H 4 S i0 4]. Since the
com position of the w ater
m ust rem ain on the
stability boundary betw
een kaolinite and
muscovite until all of the
previous
ly form ed kaolinite has been converted to
covite, the [K+]/[H +] ratio of the
rem ain constant. Therefore, equations 12.16 and
12.17 can be com bined to yield:
KAlSi3 0 8 + A l2 Si2 0 5 (O H

K A l3 Si3 O 1 0 (O H ) 2 + 2 H

This reaction increases

ing the [K+]/[H +] ratio

the chemical com position of
along th e kaolinite -m uscovite
Figure 12.1 from point D tow ard point
178 MINERAL STABILITY DIAGRAMS
If sufficient microcline is available to drive reaction
12.18 until the chemical com position of the w ater
reaches point E, all three m inerals (microcline, kaolinite,
and muscovite) coexist in equilibrium w ith the ions in the
solution . Consequently, the chemical w eathering
reactions stop at point E provided that the tem perature
rem ains constant and the system rem ains closed.
The reaction path shown in Figure 12.1 is one of an
infinite num ber of possible paths whose course depends
on the initial chemical com position and p H of the water.
The work of Steinm ann et al. (1994) indicated th at the
shape of reaction paths in Figure 12.1 is especially sen
sitive to the initial pH of the water. If pH > 5, the reaction
paths proceed from point A to the gibb-site-m uscovite
boundary w ithout entering the kaolinite field.
The purpose of this discourse is to em phasize that
the reactions betw een microcline (or anoth er m ineral)
and acidified w ater not only cause transform ations
among the solid phases but also affect the chemical com
position of the water. These reactions therefore contribute
significantly to the chemical evolution of w ater both on
the surface and below the surface of the E arth . Therefore,
activity diagrams like Figure 12.1 p ro vide useful inform
ation about several geochem i cal questions:
1. W hat environmental conditions are required to
allow a particular mineral to form?
2. W hat m inerals are stable in a given geo chemical
environm ent?
3. W hat ions or molecules are consum ed or
produced when an unstable m ineral reacts in a
given geochemical environm ent?
4. How does the w ater evolve chemically
w hen reactions occur in a closed system
with a small w ater/rock ratio?
The answers to these kinds of questions provide inform
ation that is useful in hydrogeology, sedi m entary
petrology, soil science, geochem ical prospecting,
economic geology, clay mineralogy, and other im portant
applications of inorganic aqueous geochemistry. A large
num ber of activity
diagram s for many different systems at different tem
peratures and pressures was published by Bowers et al.
(1984).
The stability limits of albite (N aA lSi 3 0 8) and
anorthite (C aA l2 Si2 0 8) are shown in Figure 12.2, and
the relevant equations are listed in Table 12.2. These
diagram s can be interpreted in the same way as the
microcline diagram . They are, in fact,
com plem entary to Figure 12.1 because they not
only introduce N a+ and Ca2+, respectively, but
also involve H + and H 4 S i0 4. In m ost natural envi ronm
ents all three m inerals are likely to be p re sent.
Therefore, the activities of H + and H 4 S i0 4 of natural
solutions reflect the progress of all three chemical
reactions occurring simultaneously. In addition, other
reactions may be taking place involving other silicate
minerals, as well as car bonates, phosphates, sulfides, or
sulfates. As a result, predictions about the progress of one
par ticular reaction occurring in such com plex natural
environm ents may be affected by the other reac tions that
are occurring simultaneously.
The stability diagram of albite in Figure 12.2A
includes the Na-m ica paragonite and therefore is sim ilar
to that of microcline. The solubility limit of m agadiite
(Surdam and Eugster, 1976) is listed in Table 12.2 but is
not shown in Figure 12.2A.
The stability of analcim e is discussed in Sec
2.2
tion 1 with other zeolite minerals, w hereas
reactions am ong the clay m inerals are discussed
in C hapter 13.
The stability field of the Ca-m ica m argarite is
not shown in Figure 12.2B because it is a rare
m ineral and its standard free energy of form a
tion is not well known. In addition, the activity of
Ca2+ in aqueous solutions reacting with an o r
thite is lim ited by the solubility of calcite, which
depends on the fugacity of C 0 2 of the environ
m ent and on the H + activity. The equation that governs
the solubility lim it of calcite can be con structed as
follows. We know that calcite dis solves to form Ca2+ and
H C O j at p H values betw een 6.4 and 10.3 (Section 9.5).
H ow ever, the bicarbonate ion is also in equilibrium w ith
m ole cular H 2 CO j, which equilibrates w ith C 0 2 (g) + H
2 0 . Therefore, we express the solubility of cal cite in term
s of Ca2+ and C 0 2 (g) because we
 12.1 CHEM ICAL WEATHERING OF FELDSPARS 179
log [Na+]/[H +]

Figure 12.2 Stability of albite (A)

and anorthite (B) in the presence of acidi
fied water at 25 C and 1 atm pressure
based on equations listed in Table 12.2.
Ab = albite, K = kaolinite, G = gibbsite,
Pa = paragonite, Py = pyrophyllite,
Si = amorphous silica, An = anorthite,
C = calcite. Na- and Ca-smectite and the
solubility limits of quartz and magadiite
are not shown, but the magadiite solubility
limit is given by equation 9 in Table 12.2A.
The solubility limit of calcite (equation 7,
Table 12.2B) in B is based on a fugacity of
C 0 2 = 3 x 10 4 atm. The solubility limits
of amorphous silica and calcite restrict the
log [Ca2+]/[H +]2

compositions of natural solutions to the

lower left corner of B. Point P in that area
represents an environment in which anor
thite alters to gibbsite, a reaction that can
cause the composition of the water to
change. When the reaction path reaches
line 5 (Table 12.2B), calcite begins to pre
cipitate and the path turns to the right but
does not advance beyond the
gibbsite-kaolinite border (equation 2,
Table 12.2B). Because of the way in which
the solubility limit of calcite restricts the
[Ca2+]/[H +]2 ratio of water, authigenic
anorthite cannot form by reactions of nat
log [H4S iO J ural solutions with gibbsite, kaolinite, or
pyrophyllite at 25 C.
1 80 MINERAL STABILITY DIAGRAMS

Table 12.2 Stability

Relations for Albite and A northite in Contact
with Aqueous Solutions at 25 C and 1 atm pressure3
A. Albite
Albite-kaolinite

2 NaAlSi30 8 + 9 H20 + 2 H +^ Al2Si20 5(0 H )4 + 2 Na+ + 4 H4S i0 4

[Na+] (1)
loS Tu+T = 2 lQg [H4SiQ4] - 0.19
[H+]
Kaolinite-gibbsite

Al2Si2Os(OH)4 + 5 H20 ^ 2 Al(OH)3 + 2 H4S i0 4

log [H4S i0 4] = -4 .68 (2)

Albite-paragonite

3 NaAlSi3Os + 12 H ,0 + 2 H +^ NaAl3Si30 1(1(0 H )2 + 2 N a+ + 6 H 4S i0 4

108 W T = 3 log [Ii4Si 4] 4-10 (3)

Paragonite-kaolinite

2 NaAl3Si30 ]()(0 H )2 + 3 H20 + 2 H +^ 3 Al2Si2Os(OH)4 + 2 Na+

logWT = +7-63 ^
Paragonite-gibbsite

NaAl3Si3Oul(OH)2 + 9 H20 + H+^ 3 Al(OH)3 + Na+ + 3 H4S i0 4

log W T = 3 log [H4Si 4] 640 (5)

Albite-pyrophyllite

2 NaAlSi3Os + 4 H20 + 2 H +^ Al2Si4O10(OH)2 + 2 Na+ + 2 H4S i0 4

log ^ = - lo g [H4S i0 4] + 2.48 (6)

Pyrophyllite-Kaolinite

Al2Si4O10(OH)2 + 5 H20 ^ Al2Si2Os(OH)4 + 2 H4S i0 4

log [H4S104] = -2 .67 (7)

 12.1 CHEMICAL WEATHERING OF FELDSPARS 181

Table 1 2 .2 (c o n tin u e d )

Solubility limit of amorphous silica

S i0 2(amorph.) + 2 H20 ^ H4S i0 4

log [H4SiO j = -2 .64 (8)
Solubility limit of magadiite

NaSi70 13(0 H )3 + 12 H20 + H +^ N a + + 7 H4S i04

AGS = +18.88 kcal K = IO1184

log W T = 7 log [Ii4Si 4] 13,84 (9)

B. Anorthite
Anorthite-kaolinite

CaAl2Si2Og + HzO + 2 H +^ Al2Si2Os(OH)4 + Ca2+

logW F = + 16,17 (1)

Kaolinite-gibbsite

Al2Si20 5(0 H )4 + 5 HzO ^ 2 Al(OH)3 + 2 H4S i0 4

log [H4SiO j = -4 .68 (2)

Anorthite-gibbsite

CaAl2Si20 8 + 6 H20 + 2 H V 2 Al(OH)3 + Ca2+ + 2 H4S i0 4

log W F = ~2 log [Ii4Si 4] + 6-82 (3)

Anorthite-pyrophyllite

CaAl2Si20 8 + 2 H4S i0 4 + 2 H V Al2Si4O10(OH)2 + Ca2+ + 4 HzO

log W F = 2 log [H4Si 4] + 2L50 (4)

Solubility limit of calcite

CaCO, + 2 H V Ca + CO, + H ,0

[ca2+l
lQg VfTTIT = 13-38 if [c 2] = 3 x 10-4 atm (5)
LH J
Based on the thermodynamic data in Appendix B.
182 MINERAL STABILITY DIAGRAMS

thereby avoid having to introduce the bicarbon ate ion A nother consequence of the calcite solubility limit is
into the equations:. that neither gibbsite nor kaolinite can be converted into
anorthite because log [Ca2 +]/[H + ] 2 of the solution cannot
C a C 0 3 (s) + 2 H 3 Ca2+ + CO ,(g) + H 2 0 (l)
rise above 13.38. Therefore, the reactions betw een
(12.19)
acidified w ater and anor thite that produce either kaolinite
In the standard state, AGR = -13.441 kcal and
or gibbsite are
not reversible at 25 C. If the fugacity of C O z is
hence K = 109'85. If [C 0 2] = 3 X IO" 4 atm:
increased by a factor of 10 to 3 X 10- 3 atm ,
(12.20)
1 0 9
[Ca2+]
log [Ca2 +]/[H + ] 2 = 12.38 (equation 12.20), which
= 10 13.38

3 X 10" causes calcite to precipitate at even lower values of the

[H + 12 [Ca2 +]/[H + ] 2 ratio. Therefore, anorthite is not form ed
and log [Ca2 +]/[H + ] 2 = 13.38. Therefore, calcite
from clay m inerals or alum inum hydroxide at Earth-
precipitates w hen log [Ca2 +] /[ H + ] 2 > 13.88. surface tem peratures, w hereas albite and K -feldspar do
Consequently, the decom position of anorthite form because their car bonates, oxides, and hydroxides are
can result in the form ation of gibbsite + calcite, much m ore soluble than those of Ca or Mg and therefore
kaolinite + calcite, or kaolinite + calcite + am or do not impose limits on the activities of their ions in
phous silica, depending on the environm ent to aqueous solutions.
which anorthite is exposed. For example, if anor thite is
placed into an environm ent represented by point P in the
stability field of gibbsite, anorthite is converted to
gibbsite in accordance with reac tion 3 in Table 12.2B. If
the system is closed and 12.2 Formation of Zeolites
the w ater/rock ratio is small, the composition of
the water changes because Ca2+ and H 4 S i0 4 are The zeolites are a large and highly diversified group of
released. If the fugacity of C 0 2 is 3 X IO- 4 atm, aluminosilicates that occur in certain
calcite begins to precipitate when the reaction path igneous, m etam orphic, and sedim entary rocks
reaches the solubility boundary of calcite. As a result, the (G ottardi and Galli, 1985). They can form by
reaction path moves along the solu bility boundary of direct precipitation from aqueous solution in vesi cles and
calcite (line 5 in Figure 12.2B) because the activity of H 4 fractures of lava flow. However, zeolites
S i0 4 increases as anor also form in sedim entary rocks by reactions
thite continues to be converted to gibbsite. betw een volcanic glass, feldspars, feldspathoids,
W hen the reaction path reaches the stability and other silicate m inerals with saline solutions of high
boundary betw een gibbsite and kaolinite (line 2 pH (Lisitzina and Butuzova, 1982; Petzing and Chester,
in Figure 12.2B), the rem aining anorthite reacts 1979). The zeolites contain w ater of hydration, which is
to form kaolinite by reaction 1 in Table 12.2B. gradually forced out by increases in pressure and tem
This reaction tends to increase the [Ca 2 +]/[H + ] 2 ratio of perature. Hence, zeolites have been used to determ ine the
the solution and therefore causes m ore calcite to depth of burial of lava flows and of volcaniclastic sedim
precipitate. The reaction path does not advance beyond en tary rocks (Coom bs et al., 1959; W alker, 1960;
the gibbsite-kaolinite boundary because the conversion of Sutherland, 1977). The crystal structures of zeo lites
anorthite to kaolinite does not increase the activity of H 4 contain channels and cavities that m ake these m inerals
S i0 4 of the solution and because the [Ca2 +]/[H + ] 2 useful as m olecular sieves. Therefore, zeolites are used in
activity ratio cannot exceed the solubility limit of calcite. the purification of w ater and other liquids and gases, for
Therefore, the solubility limits of calcite and am orphous chemical separations, and for decontam ination of
silica in Figure 12.2B restrict the com positions of radioactive waste (Barrer, 1978). Because of their many
natural waters to the lower left cor ner of the diagram . industrial uses, zeolites are synthesized for specific
applica-
 12.2 FORMATION OF ZEOLITES 183
tions, but natural m aterials recovered from large sedim 1984). Hence, therm odynam ic constants of only a few
entary deposits are also used (Hay, 1966). zeolites (analcime, wairakite, laum ontite, and
Zeolites are tektosilicates (Section 8.3) com posed of leonhardite) are listed in A ppendix B.
silica tetrahedra linked at the corners by sharing oxygen We begin by modifying the stability diagram of
atoms so that the ratio of Si to O is 1 :2 . The replacem ent albite (Figure 12.2A) by including the zeolite analcim e
of som e Si atom s by Al atoms creates a deficiency of (N aA lSi2 0 6 H 2 0 ) . The reaction by which albite is
positive charge in the lattice because Al is trivalent w transform ed into analcime is easi ly constructed:
hereas Si is tetrava lent. The resulting excess negative
charge is neutral ized by incorporation of N a+ or Ca2+
N aA lSi3 O g + 3 H 20 ^
and, m ore rarely, of K +, Ba2+, and Sr2+ ions. The
NaA lSi2 0 6 H 20 + H 4 S i0 4 (12.23)
zeolites have a m ore open lattice than other tektosilicates,
such as the feldspars and feldspathoids, which allows them A GR = +5.741 kcal K= K T 4'21
to contain water and m akes them useful as m olecular
log [H4 S i0 4] = -4 .2 1 (12.24)
sieves. Some zeolites (chabazite) can be dehydrated by
heating without perm anently dam aging the crystal W hen equation 12.24 is plotted as line 1 in Fig
lattice. D ehydrated chabazite can reabsorb the same am
ure 12.3A, it becom es apparent that analcim e
ount of w ater it contained originally. O ther zeolites
m ust also react to form paragonite. Therefore, we
(phillipsite and gismon-dine) experience m ajor structural
changes during dehydration below 200 C. construct the equation to represent this reaction:
3 N aA lSi2 0 6 H 20 + 3 H zO + 2 H + +
N aA l3 Si3 O 1 0 (O H ) 2 + 2 N a+ + 3 H 4 S i0 4 (12.25)
The chemical com positions of zeolites range widely
because of variations in the num ber of Si atom s that are AGR = -6 .045 kcal K = 10+ 4 A 3
replaced by A l atoms, because the resulting charge logW T = _ 1'5
log [H4Si 4] +
2 -2 2
(12-26)
deficiency is neutralized by vary ing proportions of N a +,
K +, Ca2+, B a2+, and Sr2+,
and because the am ount of w ater may vary The rem ainder of Figure 12.3A is identical to
depending on pressure and tem p eratu re .
Figure 12.2A. We see that analcime is stable in environm
N evertheless, the chemical form ulas of zeolites
ents having high activities of N a+ but low activities of H
can be rationalized by norm alizing the A l to two atoms,
+, which is typical of saline brines form ed by
regardless of the num bers of S i0 2 units or w ater
evaporative concentration of surface water. Therefore,
molecules. For example, the form ula of stilbite (C aA l 2
analcime commonly occurs in playa-lake deposits and in
Si6 0 1 6 6 H zO ) can be understood in term s of coupled
soils and volcaniclastic sedim ent in arid regions of the
substitution:
world. It occurs in very small crystals ranging from 5 to
8 (S i0 2) + 2 A l3+ - A l2 Si6 0 2g + 2 Si4+ (12.21) 100 p,m but can be detected by x-ray diffraction (Hay,
A l2 Si6 0 2jj + Ca2+ + 6 H zO - 1966).
C aA l2 Si6 0 1 6 6 H zO (12.22)
Next, we consider the Ca-zeolites: C
The com position of zeolites in sedim entary
aA l2 Si4 0 1 2 2H 20
rocks m ay change continuously in response to wairakite
changes in the physical and chemical state of their
CaA l2 Si4 0 1 2 3.5H20
environm ent. The com plexity of the chem i
cal com positions of zeolites and their sensitivity leonhardite CaA
to environm ental change m ake it difficult to
determ ine their standard free energies of form a tion l2 Si4 0 1 2 4H 20
(Johnson et al., 1982; Hem ingw ay and Robie,
laum ontite
184 MINERAL STABILITY DIAGRAMS

log [Na+]/[H +]
log [Ca2+]/[H +]2

Figure 12.3 Stability of albite (A) and anorthite

(B) with respect to analcime and laumontite,
respectively, in the presence of acidified water at
25 C at 1 atm pressure. Ab = albite, Ac =
analcime, Pa = parago-nite, K = kaolinite, G =
gibbsite, Si = amor phous silica, An = anorthite,
L = laumontite. The equations for the numbered
boundaries are given in the text and the others
are the same as in Figure 12.2. Analcime can
form when any of the solids in Figure 12.3A
reacts with Na-rich brines of high pH. However,
the brine compositions required to form laumon
tite at 25 C are restricted by the solubility limits
of calcite and amorphous silica to the small area
in Figure 12.3B labeled L.
log [H 4S i0 4]
 12.2 FORMATION OF ZEOLITES 185

all of which have identical m olar S i/A l ratios of 25 C only in those environm ents th at lie within the
small triangular field labeled L . In addi tion, the
1
2 :
but differ in the num ber of w ater molecules
they contain. Therefore, we choose laum ontite to conversion of anorthite to laum ontite is irreversible
represent these minerals on the stability diagram in Figure because the calcite solubility b ound ary at atm ospheric
12.3B because it forms in near-surface envi ronm ents and values of the fugacity of C 0 2 prevents the [Ca2 +]/[H + ] 2
may subsequently convert to leon- ratio of natural solu tions from increasing to values within
hardite or wairakite. L eonhardite form s by the anor thite field.
dehydration of laum ontite upon exposure to the
atm osphere, whereas wairakite becom es stable We can com bine the stability diagram s of analcim e
betw een 150 and 300 C at 1 atm pressure (Bowers et al., and laum ontite (Figure 12.3) by specify ing that the
1984).The transform ation of anorthite to lau m ontite solutions are saturated with respect to am orphous silica.
proceeds as follows (line 1, Figure 12.3B): This presents no problem for laum ontite, which can be in
equilibrium with such a solution. How ever, analcime,
C aA l2 Si2 O g + 2 H 4 S i0 4^ C aA l2 Si4 0 1 2 4H 20 paragonite, and gibbsite are not stable in contact with
(12.27) silica-satu rated solutions. Therefore, we anticipate that
A GR = -1 1 .3 2 kcal K = IO8 3 0 kaolinite is the stable phase in silica-saturated solutions
having low activities of N a+ and Ca2+ and that kaolinite
log [H4S iO j = - 4 .1 5 (12.28)
can coexist in equilibrium with albite and laum ontite as
the activities of N a+ and Ca2+ in the solution increase.
The decom position of laum ontite to kaolinite is Actually, pyrophyl-lite forms from kaolinite just below
represented by the equation: the silica-sat-
C aA l2 Si4 0 ]2' 4H 20 + H 20 + 2 H +^
A l2 Si2 0 5 ( 0 H ) 4 + Ca2+ + 2 H 4 S i0 4 (12.29) uration limit (Figure 12.2). H owever, we prefer
kaolinite because it is far m ore abundant as a
AGR = -10 .73 kcal K = 107 ' 8 7 w eathering product of alum inosilicate m inerals
[Ca2+1 (1 2 3 ) than pyrophyllite. Accordingly, we begin the con struction
7
of this diagram with the conversion of kaolinite to albite
l0 g W F = ~ 2 1 o g [H 4Si4] + 7 .8
in a silica-saturated solution:
E quation 12.30 has been plotted as line 2 in
A l2 Si2 O s(O H ) 4 + 2 N a+ + 4 S i0 2 (a m o rp h .)^
Figure 12.3B. Finally, the laum ontite-pyrophyllite
equilibrium is represented by: 2 NaAlSijOg + H 20 +2H+ (12.33)
AGR = +13.899 kcal K = 1(T1 0 1 9
C aA l2 Si4 0 1 2 4H 20 + 2 H V
A l2 Si4 O 1 0 (O H ) 2 + Ca2+ + 4 H 20 (12.31) log [N a+] 12.4) (12,34)
= +5 .09 (line 1, Figure
A GR = -1 8 .0 0 kcal K = io + 1 3 -2 0 [H +]
= +13 .20 (12.32)
[Ca2 + 1 The reaction of kaolinite with Ca2+ to form lau
log + 12
m ontite is represented by:
[H
A l2 Si2 0 5 (O H ) 4 + Ca2+ + 2 S i0 2 (am orph.)
E quation 12.32 forms line 3 in Figure 12.3B. All other
boundaries are the same as in Figure 12.2B. + 3 H 2 0 ^ CaA l2 Si4 0 1 2 4H 20 + 2 H + (12.35)
We see th at laum ontite has a large stability 15
A GR = +17.941 kcal K = IO- 1 3
field in Figure 12.3B, but only a very small p art
of it lies w ithin the com positional range of n a t log ICa2+l
ural solutions. T herefore, anorthite, or any other A l- r +12 = +13.15 (line 2, Figure 12.4) (12.36)
bearing com pound, can alter to laum ontite at LH J
186 MINERAL STABILITY DIAGRAMS

Applying the Law of M ass A ction to eq u a

tion 12.37 yields:
'96
= IO- 2 (12.38)

We now divide the [N a+ ] 2 and [Ca24] in equa

tion 12.38 by [H 4]2:

]2
[N a4 ]2 /[ H 4 -2.96 (12.39)
[Ca2 +]/[H + = 1 0
]2

Taking logarithms of equation 12.39 yields the desired

equation:
[N a+] log [Ca24]
2 log = -2 .9 6 (12.40)
[h4 [H + ] 2
and hence:

, [Na+] o s ! lo
tCa2+ - 1.48
log [Ca2 +]/[H + ] 2 log T7JTT = '5

[H 4] B [H 4 ] 2
Figure 12.4 Stability diagram for albite (Ab), kaolin ite (line 3, Figure 12.4) (12 .41)
(K), laumontite (L), and calcite (C) in contact with
solutions saturated with respect to amorphous S i02 (Si) at The three lines derived from reactions 12.33, 12.35,
25 C. Analcime, paragonite, anorthite, and gibb site were and 12.37 have been p lo tted in Figure 12.4
excluded from this diagram because they are unstable in to g eth er with the solubility lim it of calcite
silica-saturated solutions as shown in Figure 12.3. (Table 12.2B, equation 5) for [C 0 2] = 3 X 10 4
However, pyrophyllite is probably stable, but was excluded
because kaolinite is more abundant as a weathering
atm . The resulting diagram delineates the con
product of Al-silicates than pyrophyl lite. The calcite- ditions under w hich albite decom poses to kaoli
saturation limit is based on nite + silica or laum ontite + silica. Its virtue
[C 02] = 3 x IO-4 atm (Table 12.2B, equation 5). The lies in the fact th at the environm ental condi
numbered lines (1,2, and 3) were derived from equa tions are described in term s of both N a 4 and
tions 12.33,12.35, and 12.37, respectively. C a24, but it is lim ited by the instability of cer
tain solids in the presence of silica-saturated solutions.
E quilibrium betw een albite and laum ontite in sil ica-
saturated solutions is based on:

2 N aA lSi3 O s + C a2+ + 4 H 2 0 ^ 12.3 Magnesium Silicates

C aA l2 Si4 0 1 2 4H 20 + 2 N a 4 M agnesium forms a variety of silicate minerals, some of
(12.37) which include Al (phlogopite, cordierite, vermiculite,
+ 2 Si0 (am orph.)
2
chlorite, and M g-smectite), whereas others lack A l
-
AGR = +4 .042 kcal 10-296 (forsterite, enstatite, serpentine, talc, and sepiolite). Some
K = 10 Mg silicates contain other cations, such as K 4
We note th at reaction 12.37 is independent of the (phlogopite), Fe24, Ca24, and K 4 (vermiculite), and Fe2 4
pH and th erefore m ust be m odified before it can (M g-montmorillonite). Therefore, these m inerals can be
be plotted in the coordinates of Figure 12.4. shown only on diagrams that assume a specified activity
of silicic

acid or require other assumptions. For this reason, Figure
12.5A includes the stability fields only of chlorite (C t)
together with pyrophyllite (Py), kaolinite (K), gibbsite
(G ), am orphous silica (Si), and magnesite (Ms). Brucite
and periclase have
solubility limits of log [Mg2 +]/[H + ] 2 = 16.71 and
log [Mg2 +]/[H + ] 2= 21.59, respectively, and
unstable in Figure 12.5A. Therefore, magnesite
effectively limits the [Mg2 +]/[H + ] 2 ratio of natural
solutions to values less than
1 0 13
52, assuming that
[C 0 2] = 3 X 10 4 atm . The equations for these
reactions are listed in Table 12.3A.
The A l-bearing silicates of Mg dissolve incon-
gruently because gibbsite, kaolinite, and pyrophyl lite are
highly insoluble. This situation does not occur when the
Mg silicates that lack Al dissolve in acidified water. These
minerals therefore dissolve congruently and form M g2+
and H 4 S i0 4. Consequently, the stabilities of the Mg-
silicates are limited by the activities of Mg2+, H +, and H 4
S i0 4 of the solution with which they come in contact. The
equations defining the solubilities of the Mg silicates in
coordinates of log [Mg2 +]/[H + ] 2 and log
[FI4 S i0 4] are listed in Table 12.3B and have been
plotted as dashed lines in Figure 12.5B.
The solubilities of m agnesite, serpentine,
sepiolite, and am orphous silica outline a region
in Figure 12.5B in which all of the m inerals are
soluble and are therefore unstable. W hen the
activity ratio [Mg2 +]/[H + ] 2 a n d /o r [H 4 S i0 4] of
solutions in that region increase, either m agne
site, serpentine, sepiolite, or am orphous silica
precipitates, depending on which solubility line is
contacted. The com position of a solution may change
either because a soluble Mg silicate such as forsterite,
talc, enstatite, or anthophyllite is dissolving in the w ater
in a closed system w ith a small w ater/rock ratio, or
because of evaporative concentration of surface w ater, or
both. For example, if enstatite dissolves in acidified w ater
represented by the point P in Figure 12.5B, the reaction
path intersects the serpentine solubility line. Therefore,
serpentine precipitates from the solution as enstatite
dissolves. In effect, serpen tine replaces enstatite by the
reactions:
12.3 MAGNESIUM SILICATES 187
M gS i0 3 + H zO + 2 H +^ Mg2 + H 4 S i0 4 (12.42)
3 Mg2+ + 2 H 4 S i0 4 + H 2 0 ^ (12.43)
M g3 Si2 O s(O H ) 4 +6 H+
If M g2+ is conserved by m ultiplying eq u a
are tion 12.42 by 3 before adding it to equation 12.43, the net
result is:
3 M gSiO, + 4 H 2 0 ^
(12.44)
M g3 Si2 0 5 ( 0 H ) 4 + H 4 S i0 4
Therefore, the conversion of enstatite to serp en tine
increases the activity of silicic acid. If en sta tite continues
to dissolve, the activity of silicic acid m ay increase until
the solubility lim it of sepiolite is reached . C ontinued
solution of en sta tite causes sepiolite to precipitate, and
so on, until the solution becom es saturated with am or
phous silica. H ow ever, the suggested sequence
of events depends on assum ptions ab o u t the
reactions (i.e., M g is conserved, enstatite contin ues to
dissolve, no o th er reactions are taking place) and about
the environm ent (closed, with small w a ter/ro ck ratio,
stable with respect to erosion, constant tem p eratu re) th at
are rarely satisfied in nature.
The replacem ent of enstatite by serpentine in reaction
12.44 can, in principle, achieve equilibri
um at a specific value of log [H 4 S i0 4]. Since
AGR = 2.63 kcal and K = IO1,93, equilibrium
betw een e n statite and serpentine occurs at
log [H4 S i0 4] = 1.93. This value exceeds the solu bility
limit of am orphous silica, which m eans that enstatite and
serpentine cannot coexist in equi librium in contact with
natural solutions at 25 C. This is true also of enstatite
-talc, serpentine -talc, forsterite -talc, serpentine
-anthophyllite, and anthophyllite-talc because, in each
case, the activity of H 4 S i0 4 required for equilibrium is
greater than that of a saturated solution of am orphous
silica. Similarly, forsterite cannot coexist in equilibrium
with enstatite, sepiolite, or anthophyllite because the [Mg 2
+]/[H + ] 2 ratios required for equilibrium in each case
exceed the solubility limit of magnesite.
188 MINERAL STABILITY DIAGRAMS

log [Mg2+]/H +] Figure 12.5 A: Stability diagram of

Mg chlorite (Ct) (aluminosilicate) with
respect to gibbsite (G),
kaolinite (K), pyrophyllite (Py), amor
phous silica (Si), and magnesite (Ms) at
[COz] = 3 x IO-4 atm in the pres ence of
acidified water at 25 C.
B: Solubility limits of pure Mg-silicates
in acidified water at 25 C, based on
equations in Table 12.3. Each of the
minerals shown here dissolves in
solutions whose [Mg2+]/[H +]z ratios and
H4S i0 4 activities are less than those of
their respective solubility lines.
Conversely, when these parame ters
' i
B increase in a solution in the lower left
\_| Brucite ___ corner of the diagram, magnesite,
+ 16 - \ \ \ - - ^ i serpentine, sepiolite, or amorphous sil ica
precipitates, depending on which
\ solubility line is reached by the solu tion.
+]2

Consequently, these are the only stable

minerals in contact with natural solutions
+ 14 - Magnesite
s V | l \
[Mg2+]/[H

on the surface of the Earth. The

combination of solution and
reprecipitation causes the replace ment of
log

Natural
forsterite, enstatite, talc, and
anthophyllite by magnesite, serpen tine,
+ 12 - solutions sepiolite, or amorphous silica. If
amorphous silica precipitates first, and if
one of the more soluble Mg silicates
continues to dissolve, the [Mg2+]/[H +]2
of the solution rises until sepiolite
+ 10 -
N\ \ Y coprecipitates with silica. Similarly, the
Mg2*, H*, H4S i0 4 -Silica
\
\ \ initial precipitation of magnesite may
1------- 1------- r t '-------- 1------- -------- r eventually result in the formation of
8 -6 -4 -2 0 serpentine, provided the activity of H4S
log [H4S iO J i0 4 increases sufficiently.
 12.3 MAGNESIUM SILICATES 189

Table 12.3 Stability Relations among Mg Silicates in the Presence of Acidified

W ater at 25 C and 1 atm pressure3
A. Mg-Al Silicates

Kaolinite-chlorite

Al2Si20 5(0 H )4 + 5 Mg2+ + H4S i0 4 + 5 H20 ^ Mg5Al2Si3O10(OH)8 + 10 H +

AG0 = +85.235 kcal K = IO"6249
fMe2+l
log = _ 0 '2 l0g [H4Si J + 12-50 (!)
Gibbsite-chlorite

2 Al(OH)3 + 5 Mg2+ + 3 H4S i0 4^ Mg5A l2Si3OI0(OH)8 + 10 H +

AG0 = +72.48 kcal K = 10~5314

rMe2+l
log I t r F = ~0'6 log [H4Si 4] + 10-62 (2)
Pyrophyllite-chlorite

Al2Si4O10(OH)2 + 5 Mg2+ + 10 H20 ^ Mg5Al2Si3O 10(OH)8 + H4S i0 4 + 10 H+

AG0 = +92.51 kcal K = IO"67-82

[Me2+1
log T S rF = + 02 log fH4Si 4^+ 1356 (3)
Solubility limit of brucite

Mg(OH)2 ^ M g 2+ + 2 O H '

AG = +15.40 kcal K = IO1129

logW Y = +16J1 (4)

Solubility limit of periclase

MgO + 2 H +^ Mg24 + H20

AG = -29.45 kcal K = 102L59

lo g M . +21.59 (5)
 We conclude th at the th e solutions m ay rise
M g-silicates w hose until th ey reach the solu
solutibilities are show n as bility lim its of m agnesite, serpentine, sepiolite,
d a sh ed lines in Fig ure or am orphous silica, which are the only stable
12.5B are u nstable in the phases th a t can
presence of acidi coexist w ith n atu ral
fied w ater and th erefo re dissolve
solutions at 25inC.it. A s a
result, th e [Mg2 +] /[ H
190 MINERAL STABILITY DIAGRAMS
Table 1 2 .3 (c o n tin u e d )

B. Solubility of Mg Silicates

Forsterite

Mg2S i0 4 + 4 H V 2 M g 2+ + H4S i0 4
AGr = -38.76 kcal K = 10+ 2 8 -4 2
fMe2+l
log = ~0'5 log [H4Si 4] + 142 (1)
Enstatite

MgSiO- + H ,0 + 2 H V Mg2+ + H4S i0 4

A GR = -15.31 kcal K = 10n 23

lMe2+l
log W Y = _log [H4Si J + 1 L 2 3 <2)
Serpentine

Mg3Si20 5(0 H )4 + 6 H+^ 3 Mg2+ 4- 2 H4S i0 4 + H20

A Gr = -43.31 kcal K = 103L75

,W U
7
!g = -6 log [H4Si4-I + 1 0 - 5 8 (3)
Anthophyllite

Mg7Sis0 22(0 H )2 + 8 H20 + 1 4 H f ^ 7 M g 2+ 4- 8 H4S i0 4

A Gr = -93.95 kcal K= IO 6888

1
r\/[p^ +
logT iF F = _1-14 log [H4Si 4] + 9-84 (4)
Talc

Mg3Si4Ou)(OH)2 4- 4 H ,0 + 6 H+^ 3 Mg2+ + 4 H4S i0 4

A GR = -29.91 kcal K = 102IM

fMe2+l (5)
log = - 2 lo [H4SiQ4] + 10.96
12.4 Solubilit A l3+ ions in the solution
increase sufficiently to
y stabilize gibbsite.
Diagram Subsequently, m icrocline
dis solves incongruently to
s form gibbsite plus K + and
A ll m inerals dissolve H 4 S i0 4. We therefore
need to know how gibbsite
congruently when placed
dissolves and under w hat
in p u r e w ater. For exam
conditions it is stable
ple, m icrocline dissolves
congruently in pure w ater
until the activities of
 12.4 SOLUBILITY DIAGRAMS 191

Table 12 .3 (continued)

Sepiolite (see Appendix B)

(a) Mg4Si60 l5(0 H)2 -6H20 + H20 + 8 H +^ 4 M g 2+ + 6 H4S i0 4

A GR = -43 .28 kcal K = IO3173

logW F = L5 log [Ki 4] + 7-93 (6)

(b) Mg2Si30 6(0 H )4 + 2 H20 + 4 H V 2 Mg2+ + 3 H4S i0 4

AGR = -21.26 kcal K = IO15'58

logl ^ F = _1'5 Iog [H4Si 4] + 7-79 (7)

Based on the thermodynamic data in Appendix B.


with respect to the
different Al ions that can
form in an aqueous
solution.
a. Gibbsite
Since gibbsite is the
insoluble hydroxide of Al,
we recall the discussion of
the dissociation of bases
and their am photeric
character from Section
9.4. A t that stage in our
discussion we w ere
restricted to known
dissociation constants, w
hereas we can now
calculate them from
values of AG^ for any
reaction we wish to study.
The A ppendix B con tains
Gf values for A l3+, A
l(O H )2+, A l(O H )+ , A
l(O H )3, A l(O H )4, and
A l( O H ) f , all of which
can form w hen an A l-
bearing m ineral dissolves
congruently at different
pH values. The reactions
by which these ions form
from gibbsite and their
equi librium constants at
25 C are listed in Table
12.4.
The equations
derived in this way
outline the stability field
of gibbsite in Figure 12.6.
O utside this area,
gibbsite dissolves
congruently to form ions
whose relative
abundances becom e
equal at certain pH values
corresponding to the
points of intersection of
the solubility lines. For
example, A l3+ is the dom
inant ion at pH < 4.65, and
A l(O H + ) 2 dom inates at
pH values from 4.65 to
5.7, as show n in Figure
12.6. T he so lu b ility of
gibbsite at a specified p H
is th e sum of the con
centrations of all A l-
bearing species th at exist
in a saturated solution in
equilibrium w ith gibb
site. The solubility of this
m um at p H from 6 to
and <
6
in accordance
ch aracter of Al.
W hen m icrocline dissolves
very dilute solutions
having a pH <4.65, it
releas es K +, A l3+, and
H 4 S i0 4 according to the
reaction:
K A lSi3 0 8 + 4 H 20 + 4 H
K + + A l3+ +
W hen microcline dissolves congruently
of com position P in Figure 12.6, both the activity
of A l3+ and th e pH rise along the reaction path . If
the system is closed and
small, the solution may
respect to gibbsite, which then precipitates:
A l3+ + 3 H zO ^ A l(O H
The net result, obtained by
adding equations 12.45
and 12.46, is that
microcline then dissolves
incon-gruently to form
gibbsite plus K + and H 4
S i0 4:
KAlSi3 O s + 7 H 20 +
H +^
A l(O H ) 3 +
K++3H4S
i0 4
(12.47)
If m icrocline continues to
dissolve, the com posi tion
of the solution m oves
across the gibbsite sta
bility field in Figure 12.1
until it reaches the gibbsite
-kaolinite boundary (line
2). A t that
192 MINERAL STABILITY DIAGRAMS

Table 12 .4 C o n g ru e n t S o lu tio n o f G ib b site

to F o rm D iffe re n t Io n ic a n d M o le c u la r S
p ec ies a t 25 C a
Al(OH)3(s) + 3 H +^ A 13+ + 3 H20
AGR = -11 .29 kcal K = IO8,28
log [Al3+] = 3pH + 8.28 (1)
A1(OH)3(s) + 2 H +^ Al(OH)2+ + 2 H 20

AG = -4 .44 kcal K = IO3 26

log [Al(OH)2+] = -2p H + 3.26 (2)
A1(OH)3(s) + H +^ Al(O H ) 2 + H20

A G = +1.39 kcal K = IO' 1"2

log [Al(O H )2] = - p H - 1.02 (3)
Al(OH)3( s )^ Al(OH) pH
AG0 = +9.16 kcal K = 10' 6 72
log [Al(OH)j] = -6 .72 (4) Figure 12.6 Congruent solubility of gibbsite as a

A1(OH)3(s) + H ,0 ^ Al(OH) 4 + H + function of the environmental pH at 25 C, based on

AG = + 2 0 .5 kcal K = IO' 15,1 the equations in Table 12.4. The lines define the stabil
ity field of gibbsite where it precipitates from saturat
log [Al(OH )4] = pH - 15.1 (5) ed solutions. The environments in which gibbsite is
A1(OH)3(s) + 2 H20 ^ Al(OH)2G + 2 H + soluble have been subdivided by identifying the domi
nant Al species depending on the pH. Note that
AG = +35.41 kcal K= 10'2fio Al(OH)2+ is not dominant at any pH according to the
log [Al(OH)T] = 2pH - 26.0 (6) standard free energy values used to construct this dia
Based on the thermodynamic data in Appendix B. gram. The amphoteric character of Al causes the solu
bility of a gibbsite to increase with both decreasing
and increasing pH. Its solubility is at a minimum at
point, m icrocline form s kaolinite and the previ pH values between 6 and 8, which justifies the
assumption that Al is conserved during the incongru
ously form ed gibbsite is converted to kaolinite: ent solution of aluminosilicates.
 2 K A lSi3 O s + 9 H 20 + 2 H + +
A l2 Si2 0 5 ( 0 H ) 4 + 2 K + +
then traverses the kaolinite
4 A l(O H ) 3 + 4 H 4 S i0 4^ field as discussed before.
Therefore the solubility
2 A l2 Si2 0 5 ( 0 H diagram of gibb site in
Figure 12.6 com plem ents
If the conversion of the stability dia gram of
gibbsite to kaolinite keeps microcline in Figure 12.1.
pace w ith the incongruent The pH dependence
solution of m icrocline to of the solubility of gibb
kaolinite, then [H 4 S i0 4] site and of other A l-
rem ains constant and only bearing m inerals causes
[K +]/[H +] increases. the abundance of Al
Consequently, the reac species in aqueous
tion path m oves u p solution to rise at both
along the gibbsite- high and low environm
kaolinite boundary until all ental pH val ues. High
of the gibbsite is converted concentrations of
to kaolinite. If m ore m dissolved A l in soil tend
icrocline dissolves, the to inhibit the growth of
path rooted plants for reasons
that are not yet fully
understood . The Al
toxicity associated with
strongly acidic or basic
 12.4 SOLUBILITY DIAGRAMS 193
environm ents contributes to the infertility of such
environm ents. In addition, we note that A l is m obile in
strongly acidic and basic environm ents because it can be
transported by the m ovem ent of w ater. M oreover,
gibbsite can precipitate when strongly acidic or basic
solutions are neutralized.
W hen other A l-bearing m inerals such as kaolinite,
pyrophyllite, feldspars, feldspathoids, zeolites, or clay m
inerals dissolve congruently in aqueous solutions, the
abundances of the aque ous A l species depend on the
environm ental pH as shown in Figure 12.6. However,
these m inerals also produce H 4 S i0 4 and the equilibria
therefore involve three variables: the A l species, pH , and
H 4 S i0 4. The stabilities of these m inerals in equi librium
with their ions can be represented either in three-dim
ensional diagram s or in coordinates of log [Al species]
and log [H4 S i0 4] at a series of pH values. pH
Figure 12.7 Congruent solubility of hematite as a function
of pH based on equations like those for gibb site in Table
12.4. The solubility of hematite, or other compounds of Fe,
b. Hematite has a minimum at pH values from 4.3 to 8.7 and increases
Iron, like Al, is am photeric and forms hydroxyl com both in more acid and in more basic solutions because Fe,
plexes whose standard free energies are list ed in A like Al, is amphoteric. As a result, Fe is mobile in acidic
ppendix B. E quations for the congruent solution of hem and basic environments and precipitates as amorphous
atite as a function of pH are easi ly derived and have been ferric hydroxide when Fe-bearing solutions are
plotted in Figure 12.7. The results define the stability field neutralized. Ferric hydroxide recrystallizes through
lepidocrocite and goethite to hematite.
of hem atite w here it precipitates from aqueous solutions.
The environm ents in which hem atite dissolves have been
subdivided according to the dom inant Fe species in
solution at different pH ranges. The sol ubility of hem atite
has a m inim um betw een pH values of about 4.3-8.7 thite, or hem atite, which is the m ost stable form of
when F e(O H ) 3 is the Fe oxide on th e surface of the E arth .
The geochem istry of Fe differs from th at of Si
dom inant ion. The Fe3+ ion is dom inant only at and Al because it has two oxidation states (+ 2
pH <2.2. and + 3) both of which occur at E arth -surface
The solubilities of all oxides, hydroxides, car
conditions. The transform ation from Fe2+ to Fe3+
bonates, and sulfides of iron are strongly depen
dent on the pH of the environm ent. A t pH <4.3 Fe becom occurs by the loss of an electron from Fe2+ in the
es increasingly m obile in Earth-surface environm ents presence of an electron acceptor. Loss of elec
and is leached from soils, regolith, and rocks exposed to trons is know n to us as oxidation, w hereas a
such conditions. W hen such acidic waters are neutralized, gain of electrons is reduction . O xidation and
they may becom e satu rated w ith respect to ferric reduction m ust always occur together by a transfer
hydroxide [Fe(O H )3], which may then precipitate. The of electrons from an electron donor to an elec
ferric hydroxide recrystallizes in tim e to form m ore sta tron acceptor. O x id atio n -red u ctio n reactions
ble crystalline phases such as lepidocrocite, goe- m ust be balanced fir s t in term s of the electrons
th at are transferred before a mass balance is
achieved . T hese m atters are the subject of
1 94 MINERAL STABILITY DIAGRAMS
C hapter 14, w here we review som e basic princi
and the gases escape. Therefore, the fugacities of 0 2 and H
ples of electrochem istry and apply them to the 2 m ust each rem ain <1 .0 atm if liquid w ater is to be
construction of the so-called E h - p H diagrams. stable at sea level on the surface of
In Section 14.5 we take up the construction the Earth . If the fugacity of H 2 alone is 1.0 atm ,
of fugacity diagram s based on equilibria betw een then the fugacity of 0 2 m ust be as low as it can
solid com pounds and certain gases, including 0 2, get in the presence of liquid w ater. Setting
C 0 2, and S2. The solid -gas equilibria of both Fe [H 2] = 1.0 in equation 12.51 yields:
and Cu include possible changes in their o xid a tion [ 0 2] = K T 8 3 ' 1 atm (12.52)
states, but the reactions are quite uncom pli cated. Thus we
can present fugacity diagram s here rather than in C hapter
Therefore, the fugacity of 0 2 in the presence of liquid w
14. ater may vary from 1 . 0 to IO 8 3 1 atm . Similarly, the fugacity
of H 2 can vary from 1.0 to IO41,6 atm, based on
equation 12.51.

12.5 Fugacity Diagrams b. Oxides and Carbonates of Iron

The oxides, carbonates, and sulfides of m etals can be M etallic Fe exposed to the air is know n to form the
considered to be in equilibrium with gaseous 0 2, C 0 2, oxides wiistite (FeO ), m agnetite (Fe 3 0 4), and hem atite
and S2, respectively. For exam ple, the fugacity of 0 2 (Fe2 0 3). Therefore, we can represent the oxidation of
ranges from about 2 X IO- atm in air at the surface of the metallic Fe by a series of equa tions starting with the form
E arth to m uch lower val ues in certain anoxic environm ation of wiistite:
ents such as peat bogs, stagnant basins, and deep -sea
Fe (m etallic) + 5 0 2^FeO (12.53)
trenches. Therefore, the first question we need to consider
For this reaction, AG^ = 58.7 kcal and K =
is the range of fugacities of 0 2 th at can occur on the surface
of the Earth . 10+43
Thereforei at equilibrium :

a. Stability Limits of Water

We know that liquid w ater dissociates into H + and O H "
ions and that the ion activity product of w ater at 25 C is and [Oz] = IO861. Evidently, wiistite can coexist in
about 1 X IO "14. H ow ever, w ater m ust also m aintain equilibrium with m etallic Fe only when the fugacity of O
an equilibrium with the gases 0 2 and H2: z is less than is perm itted in norm al terrestrial environm
ents.
2 H 2 0 ( 1 ) - 0 2 (g) + 2 H 2 (g) (12.50) We can continue either by considering the transform
ation of wiistite to m agnetite, which is a mixed oxide
The standard free energy change of this reaction is AGR = (FeO Fe2 0 3), or by relating m ag netite directly to
+113.374 kcal and its equilibrium con stant at 2 5 C is K metallic Fe. The latter is prefer able because wiistite does
= IO"831. T herefore, at equi librium : not occur in terrestrial rocks. Therefore, we consider the
reaction:
[ 0 2 ][H 2 ] 2 = IO" 8 3 1 (12.51) 3 Fe (metallic) + 2 O , ^ FeO Fe2 0 3 (12.55)
2 2 3
AGR = -2 4
The total pressure of all gases occurring naturally on the
6

2 .6 kcal K = IO1 7 7 8
surface of the E arth m ust be 1 atm or less. If the pressure
of gases rises to higher values, they expand against the atm It follows th at at equilibrium the fugacity of 0 2
osphere, which m eans th at bubbles form in w ater m ust be io - 8 8 -9 2 atm , which is also less than the
exposed to the atm osphere lowest fugacity of 0 2 in equilibrium with liquid
 12.5 FU GACITY DIAGRAMS195

w ater. Therefore, m etallic Fe cannot be stable on the surface of the E arth, w here th e fugacity of 0 2 Taking logarithm s to the base 10:
m ust be > 10 8 3 1 atm . H ow ever, m agnetite does
5 log [ 0 2] - 3 log [C 0 2] = -32 .61 (12.59)
occur in surface environm ents and, therefore, is
log [ 0 2] = 6 log [C 0 2] - 65.22 (12.60)
stable at 0 2 fugacities that are com patible with
liquid water. Similarly, h em atite reacts with C02 to form
We consider next the reaction betw een mag netite
and hem atite in the hope that this equilibri um will siderite according to:
provide a stability limit for m agnetite that is w ithin the Fe2 0 3 + 2 C 0 2^ 2 F e C 0 3 + j 0 2 (12.61)
range of perm issible 0 2 fugacities:
AGR = + 45 .52 kcal K = IO' 3 3 3 7
2 FeO Fe2 0 3 + 2 02 = 3Fe203
(12.56) and:
AG0 = - 4 7 .6 kcal K ,34.89
= 10; log [ O j = 4 log [C O J - 66.74 (12.62)
Thus, m agnetite and hem atite are in equilibrium

at [ 0 2] = IO- 6 9 8 atm, which is within the perm is
sible range of 0 2 fugacities. H em atite is the most
O -rich oxide of Fe and can coexist with w ater at 0 2
fugacities from IO- 6 9 8 to 1 . 0 atm .
Next, we add siderite (F e C 0 3) to the system by
reacting m agnetite and hem atite with CO z gas. Since Fe
in siderite is divalent, we start with m ag netite because it
also contains divalent Fe. We construct the desired
equation with m agnetite as the reactant and siderite as the
product. Then we
add C 0 2 to the reactant side to balance the C in
siderite. Finally, we com plete the equation by
adding 0 2 as needed for balance:
FeO Fe2 0 3 + 3 C 0 2 =
N ote that this is an oxidation -reduction reaction and that
the equations cannot be balanced in term s of mass only,
as we just did. In this case, two of the three divalent Fe
ions in siderite give up one electron each for a total of
two. These elec trons are accepted by one atom of 0 2 gas,
The equations representing solid -gas equilibria derived
above have been plotted in Figure 12.8 in coordinates of
log [ 0 2] and log [C 0 2], The result is a fugacity diagram
for the oxides and
carbonates of Fe th at indicates the fugacities of
the tw o gases th at are required to stabilize each
of the m inerals we have considered . T he d ia gram also
predicts the direction of solid -gas reactions th a t occur w
hen a particular m ineral is placed in an environm ent in
which a n o th er m in eral is stable.
For exam ple, so-called ironstone concretions com
posed of siderite may be exposed to the atm osphere as a
result of erosion of the enclosing sedim entary rocks.
F e C 0 3 + 5 0 2 (12.57)
Figure 12.8 indicates th at hem atite is stable in contact
with the atm osphere.
Therefore, siderite is transform ed into hem atite
in accordance w ith equation 12.61. A t equilibri
um betw een siderite and hem atite:
[o2]
-33.37 (12.63)
1 /2

which becomes one of the O 2 ions in the magnetite. 10

Equation 12.57 is reversed because we have treat ed m [co2]: =
agnetite as the reactant rather than as the product. The How ever, in the atm osphere [ 0 2] = 2 X 10 1 atm
standard free energy change of reac tion 12.57 is AGR = and [C 0 2] = 3 X 10' 4 atm . Therefore, siderite
+44.48 kcal and k = 1CT3 2 61. Therefore, at equilibrium: can coexist in equilibrium with hem atite in con
tact w ith the atm osphere only when:
1/2 r(o2)i/2i 1 /2
[o2]1/2 [co2]=[[02]1/21
-32.61

= (12.58)
K
L10-3337 J
[co2]3 10
16.51 (12.64)
= 10 atm
196 MINERAL STABILITY DIAGRAMS

log [ C 0 2] log [S2]

Figure 12.8 Fugacity diagram for the oxides and car
bonate of Fe in terms of the fugacities of 0 2 and COz,
where Mt = magnetite, H = hematite, and S = siderite. Note
that siderite and magnetite are not stable in con tact with
the atmosphere, where [0 2] = 2 x 1CT1atm and [C 02] = 3
X IO-4 atm. Therefore, when siderite or magnetite is
exposed to the atmosphere, it converts to hematite. The
diagram is based on equations 12.52 (stability limit of
water) and 12.56,12.58, and 12.61. Point P represents
conditions when siderite is stable in contact with 0 2 of the
atmosphere.
Figure 12.9 Fugacity diagram for the oxides and sul
fides of Cu. Cu = metallic Cu, Cup = cuprite (Cu20 ), Ten
= tenorite (CuO), Chal = chalcocite (Cu2S),
and Cov = covellite (CuS), based on equations in Table
12.5. Metallic Cu is stable in the presence of water, but
when it is exposed to the atmosphere, it reacts to form
tenorite. Note that G{ of S2(g) is not zero because the
stable form of S in the standard state is the rhombic
solid.
 c. Oxides and Sulfides of Copper
C opper forms the oxides cuprite (Cu2 0 ) and
which is identified by point P in the diagram . tenorite (CuO ) and the sulfides chalcocite (C u 2 S)
Since the fugacity of C O z in the atm osphere is and covellite (CuS). N ote that Cu is univalent in
only 3 X IO- 4 atm , reaction 12.62 is not in equilib cuprite and chalcocite and divalent in tenorite
rium in contact with the atm osphere but runs to and covellite. Therefore, when metallic Cu forms
the left, thereby converting siderite to hem atite. C u+ and w hen C u+ form s Cu2+, electrons are
Therefore, ironstone concretions should w eather transferred either to 0 2 or to S2, depending on
to hem atite upon exposure to the air. This predic the com pound that is produced. The equations
tion is confirm ed by the presence of a hem atite required to construct the fugacity diagram for Cu
rind around ironstone concretions w here the in Figure 12.9 are listed in Table 12.5.
thickness of the rind depends on the rate of the We see th at native Cu is stable in the pres
reaction and the duration of the exposure. ence of w ater, w hereas native Fe is not. H ow ever,
 12.5 FU G ACITY DIAGRAMS 197
Table 12.5 Fugacity Diagram of Copper
Oxides and Sulfides at 25 Ca
2 Cu (metal) + ) 0 2^ C u 20

A G = -35 .1 kcal K = 1025 73

log [Oz] = -51.5 (1)
Cu20 + ) () . . 2 CuO
AG = -26 .3 kcal K = IO1928
log [ 0 2] = -38 .6 (2)
2 Cu (metal) + \ S2^ C u 2S
A G = -30.08 kcal K = IO2205
log [S2] = -44.1 (3)
Cu20 + ) S2 Cu2S + : 0 2
A GR = + 5.02 kcal K = IO360
log [0 2] = log [S2] - 7.4 (4)
2 CuO + \ S ,. Cu2S + 0 2
A GR = +31.32 kcal K = HF2296
log [0 2] = 0.5 log [S2] - 23.0 (5)
Cu2S + \ S2; 2 CuS log [C 0 2]
AG = -14.48 kcal K= IO10 61
log [S2] = - 21.2 (6) Figure 12.10 Combined fugacity diagram for
CuO + ) S2^ C uS + | 0 2 Fe and Cu and their oxides and carbonates.
AG = +8.42 kcal K = 10~617 H = hematite (Fe20 3); Mt = magnetite (Fe30 4);
log [0 2] = log [S2] - 12.34 (7) S = siderite (FeC 03); Cu = native Cu; Cup = cuprite
Based on the thermodynamic data in Appendix B.
(Cu20);Ten = tenorite (CuO); Mai = malachite
[CuC03 Cu(OH),]; A = azurite [2CuC03- Cu(OH)2].
The boundaries between tenorite and malachite and
between malachite and azurite were drawn after
w hen m etallic Cu is exposed to 0 2 at fugaci- Garrels and Christ (1965). This diagram indicates the
ties > 1 (T 5 -5 atm , it form s th e oxide cuprite environmental conditions under which certain assem
(C u 2 0 ) or tenorite (C uO ), depending on the 0 2 blages of Fe and Cu minerals can occur together.
fugacity. Similarly, Cu reacts with S2 to form the
sulfides chalcocite (C u2 S) and covellite (CuS). For exam ple, the diagram indicates th at the
C opper also form s mixed sulfides with Fe includ
ing chalcopyrite (CuFeS2), cubanite (C uFe2 S3), carbonates of Cu (m alachite and azurite) form at
and bornite (C u5 FeS4). lower fugacities of C 0 2 than siderite and th ere
Fugacity diagram s of different m etals can be fore can coexist in equilibrium with hem atite and
superim posed on each other to form com posite tenorite. Consequently, malachite, azurite, tenorite,
diagram s that indicate the stability limits of the and hem atite are com m on w eathering products
com pounds of both metals. Such a diagram has of F e -C u sulfide m inerals and occur in the oxi
been constructed in Figure 12.10 for the oxides dized part of sulfide ore bodies (gossans), w here
and carbonates of Fe and Cu. The diagram as siderite is unstable at th e surface of the E arth
defines the fugacities of 0 2 and C 0 2 th at perm it (see also Section 21.2). Figure 12.10 also points
certain assemblages of these m inerals to coexist out that native Cu has a large stability field in the
in equilibrium in the presence of water. absence of S and can coexist with m agnetite,
198 MINERAL STABILITY DIAGRAMS

hem atite, and siderite, provided the fugacity of O z is low T herefore, certain m inerals cannot be in equilib
enough. W hen native Cu is exposed to the atm osphere, it rium in the presence of w ater and also cannot
tarnishes by form ing a layer of tenorite or m alachite or form by transform ation from other m inerals.
both, depending on the fugacity of C 0 2. This phenom enon is exem plified by anorthite,
which does not form as an authigenic m ineral at
E arth -surface tem peratures, w hereas albite and
K -feldspar do occur in places w here alkali-rich
brines can react w ith volcanic glass, clay m iner als, or
12.6 Summary
other kinds of solids.
W hen the silicate m inerals of igneous and meta-m orphic The conditions required for stability of m in erals can
rocks are exposed to acidified water, they may dissolve in also be expressed by their congruent solution or by their
co n g ru en tly to produce clay m inerals and zeolites as interaction with gases such as 0 2, C 0 2, and S2. For
well as insoluble oxides and hydroxides, depending on th e example, the non-alum inous Mg silicates dissolve
environm ental conditions. The chem ical reactions th at congruently to form Mg2+ and H 4 S i0 4, which may
take place precipitate as a different Mg silicate whose solubility is
during these transform ations can be used to exceeded. The com bination of congruent solution of a
define the conditions required for equilibrium prim ary Mg silicate and the precipitation of a
w hen both reactants and products are stable. secondary m ineral results in the replacem ent of one m
T herefore, equations can be derived from the iner al by another. The apparent conversion of ensta-tite to
Law of Mass A ction, which specify the conditions serpentine is an exam ple of such paired reactions.
required for equilibrium betw een the solids.
T he resulting m ineral stability diagram s
define the environm ental conditions in which The equilibrium betw een w ater and its con stituent
each of the solids in a given com positional sys gases, 0 2 and H 2, defines the stability range of w ater w
tem is stable. In addition, the diagram s predict ithin which all reactions on the surface of the E arth take
place. Solid oxides, car
w hat happens to a m ineral w hen it is exposed to
conditions outside of its own stability field. The reactions bonates, and sulfides m ust sim ilarly rem ain in
0
equilibrium with their constituent gases 2,
th at occur in such cases n o t only cause transform ations
am ong the solids but also affect C 0 2, and S2, respectively. These solid -gas reac
the chem ical com position, or quality, of w ater tions can be used to construct fugacity diagram s th at
b o th above and below the surface of the E arth . contain the stability fields of the solids appropriate to a
W hen the reactions occur in closed system s with sm all w particular system . The fugacity
ater/ro ck ratios, the chem ical com posi diagram s of different m etals can be superim
tion of th e w ater m ay evolve along curved reac posed on each other to form com bination dia
tion paths. Such conditions exist in the pore grams. These diagram s effectively subdivide the
spaces or fractures of rocks in subsurface w here stability field of w ater into environm ents in
reactions betw een m inerals and w ater not only which suites of m inerals of different elem ents
co n trib u te to the chem ical evolution of the can coexist in equilibrium .
w ater but also cause the deposition of m ineral The virtue of activity and fugacity diagrams is that
cem ents. they depict the relationships betw een m iner als and
The solubility lim its of certain com pounds, aqueous solutions or gases in a readily usable form . A s a
including calcite, m agnesite, am orphous silica, result, idle speculation is replaced by understanding and
and others, which can p recip itate from natural progress is made in understanding the chemical reactions
solutions, significantly restrict th e chem ical com p o sitio that con tribute to geological processes.
n s of aq u eo u s so lu tio n s in nature .
 REFERENCES 199

Problems
 1. Construct the equations necessary to introduce
pyrophyllite into the activity diagram for microcline in
Figure 12.1.
2. Construct the equations necessary to intro duce Na-
smectite (Na0.33Al2.33S>3.670lo(OEI)2 GJ - 1277.76
kcal/mol) into the activity diagram for albite in Figure
12.2.
3. Construct a solubility diagram for goethite (FeOOH)
as a function of pH.
4. What are the environmental conditions in terms of
[Mg2+]/[H +]2 and [H4S i0 4] at which enstatite and
forsterite coexist in equilibrium?
5. Is equilibrium possible between enstatite and
forsterite in the presence of aqueous solutions in nature?
6. What mineral precipitates first when forsterite
dissolves congruently in water having the composition of
point P in Figure 12.5?
7. Construct a fugacity diagram for the sulfides and
carbonates of Pb.
8. Construct a fugacity diagram for the sulfides and
carbonates of Zn.
9. Combine the two diagrams for Zn and Pb.
10. What are the fugacities of 0 2 and COz at which
native Cu, cuprite, and azurite coexist in equilibrium?
11. Can hematite and siderite both join that assem
blage in equilibrium with each other?
References

B A R R E R , R. M., 1978. Zeolites
and Clay Minerals as
Sorbents and Molecular
Sieves. Academic Press,
London, 497 pp.
B o w e r s , T. S., K . J. J a c k s o
n , and H .
C. H E L G E S O N , 1984.
Equilibrium Activity
Diagrams. Springer-
Verlag, New York,
397 pp.
C O O M B S , D. S., A. J. E L L IS , W.
S. F Y F E , and A. M. T A Y L O R ,
1959.
The zeolite facies, with
comments on the
interpretation
hydrothermal syntheses.
Geochim. Cosmochim.
Acta, 17: 52-107.
G A R R E L S , R. M., and C. L. C H R
IS T , 1965. Solutions,
Minerals and Equilibria.
Harper & Row, New York,
450 pp.
G O T T A R D I , G ., and E . G A L L I,
1985. Natural Zeolites,
Springer-Verlag, New
York, 390 pp.
H A Y , R. L., 1966. Zeolites and
zeolitic reactions in
sedimenta ry rocks. Geol.
Soc. Amer., Spec. Paper No.
85,130 pp.
H E L G E S O N , H . C., R. M . G A R R E
L S , and F. T. M A C K E N Z IE , 1969.
Evaluation of irreversible
reactions in geochemical
processes involving
minerals and aqueous
solutions. II. Application.
Geochim. Cosmochim.
Acta, 3 3 :4 5 5 -4 8 1 .
H E M IN G W A Y , B. S., and R. A. R
O B I E , 1984.
Thermodynamic
properties of zeolites: low-
temperature heat
capacities and
thermodynamic functions
for phillipsite and
clinoptilolite. Estimates of
the thermochemical
properties of zeolitic water
at low temperature. Amer.
Mineral., 69:692-700.
J O H N S O N , G . K., H. E. F L O T O W ,
P. A. G. O 'H A R E , and W. S.
W IS E , 1982.
Thermodynamic studies of
zeolites: analcime and
dehydrated analcime.
Amer. Mineral., 67: 736-
748.
of
K S T N E R , M., 1971. Authigenic
feldspars in carbonate rocks.
Amer. Mineral., 56:1413-
1442.
K S T N E R , M., and R. S I E V E R ,
1979. Low temperature
feldspars in sedimentary
rocks. Amer. J. Sei., 279:
435-479.
L IS IT Z IN A , N. A., and G. YU. B U
T U Z O V A , 1982. Authigenic
zeo lites in the sedimentary
mantle of the world ocean.
Sed. Geol., 31: 33^2 .
M E N S IN G .T . M., and G. F A U R E ,
1983. Identification and
age of neoformed Paleozoic
feldspar (adularia) in a
Precambrian basement
core from Scioto County,
Ohio, U.S.A. Contrib.
Mineral Petrol., 82: 327-
333.
P E T Z IN G , J., and R. C H E S T E R ,
1979. Authigenic marine
zeolites and their
relationship to global
volcanism. Marine Geol.,
29: 253-272.
ST E IN M A N N , P., P. C. L I C H T N E R ,
and W. SH O T Y K , 1994.
Reaction path approach to
mineral weathering
reactions. Clays and Clay
Minerals, 42:197-206.
S U R D A M , R. C., and H. P. E U G S
T E R , 1976. Mineral
reactions in the
sedimentary deposits of the
Lake Magadi region,
Kenya. Geol. Soc. Amer.
Bull., 87:1739-1752.
S U T H E R L A N D , F. L., 1977.
Zeolite minerals in the
Jurassic dolerites of
Tasmania: Their use as
possible indicators of
burial depth. Geol. Soc.
Aust. J., 24:171-178.
W A L K E R , G. P. L., 1960.
Zeolite zones and dike
distribution in relation to
the structure of the basalts
of eastern Iceland.
J. Geol., 68: 515-528.
 13
Clay Minerals
Clay m inerals form one of the largest and m ost difference to the geochem ical processes that
highly diversified groups of m inerals known. The m ay be occurring. Since we encountered the
m ajority of clay m inerals are alum inosilicates clay m inerals kaolinite, pyrophyllite, and sm ec tite in C
and they crystallize as phyllosilicates. How ever, hapter 12, it is now necessary for us to exam ine these m
som e clay m inerals are M g-silicates or Fe-sili- inerals in m ore detail.
cates, others are n o t crystalline, or they are not
phyllosilicates. The property com m on to all clay 13.1 Crystal Structure
m inerals is the small size of their crystals or
grains with diam eters less th an 2 m icrom eters Clay m inerals are divided into three groups based on
(iom). A s a result, sedim ent in th at size range is their crystallographic properties: (l) platy clay m inerals
som etim es referred to as the clay-size fraction or even as (phyllosilicates), (2 ) fibrous clay minerals, and (3) am
clay. This usage is incorrect in geochem istry because orphous clay. The platy clay m inerals are much m ore
clay m inerals are defined on the basis of their crystal abundant than the other two groups and include m ost of
structure and chem ical com position and not by the size the m inerals that traditionally m ake up the clay
of their crystals or minerals. The fibrous clays sepiolite and palygorskite
grains. M oreover, the clay-size fraction of sedi (also know n as attapulgite), though not as com m on as
m ent frequently includes sm all grains of other the platy clays, do occur in many different kinds of
m inerals (quartz, feldspar, calcite, etc.), as well as grains depositional environm ents, including m arine and
of insoluble oxides or hydroxides of Fe, Al, M n, and so lacustrine sedim ent, hydrotherm al deposits,
on, which may be partly or com plete ly am orphous. N
and soils of arid regions (Singer and G alan,
evertheless, clay m inerals are typically fine grained and
1984). So-called am orphous clay, referred to as
therefore cannot be identified solely by their physical or
allophane, is an alum inosilicate of variable chem ical
optical p ro p erties. Instead, the classification of clay m
com position that is am orphous to x-rays in m ost cases
inerals relies heavily on their crystallographic and cer tain
but does diffract electrons. It may be an interm ediate
physical properties th at are revealed by x-ray diffraction.
product in the form ation of phyllosilicate clays from
various kinds of m ateri als, but its abundance is not well
Clay m ineralogy has evolved into an im por ta n t docum ented because it is not directly detectable by x-ray
field w ith applications in geology, agrono dif fraction (Carroll, 1970).
my, ceram ic engineering, soil m echanics, and
Clay m inerals are described in m any text books on
geochem istry. We will be concerned prim arily
mineralogy and in num erous specialized books,
w ith the classification of clay m inerals, with
including those by Grim (1968), Millot (1970), W eaver
th eir stability in d ifferent geochem ical environ m ents,
and Pollard (1973), Velde (1977), Sudo and Shim oda
w ith ion exchange phenom ena, and w ith th e use of clay
(1978), Sudo et al. (1981), and Caims-Smith and H artm
m inerals for dating sedim entary rocks. The presence of
an (1987). In addition, several scientific journals are
clay m inerals in n atural environm ents frequently m akes
devoted to clay min
a significant
200

eralogy, including Clays and Clay Minerals, Clay
Minerals Bulletin, Clay Science, and Clays and
Clay Technology. Articles on clay mineralogy are also
published in the Am erican Mineralogist, the
Journal o f the Soil Science Society o f America, and the
Journal o f Sedimentary Petrology.
The phyllosilicate clay minerals consist of two kinds
of layers of different chemical composition and
coordination. One of these is the tetrahedral layer (T),
which is composed of silica tetrahedra. The second layer
consists of A l hydroxide with A l3+ ions located in the
centers of octahedra and is there fore known as the
octahedral layer (O). In some clay minerals, such as
serpentine and talc, the octahedral layer is composed of
Mg hydroxide, which forms the m ineral brucite.
Therefore, the octahedral layer is also referred to as the
brucite layer or the gibbsite layer, depending on its
composition.
W hen A l3+ occurs in the octahedral layer, only two
of every three available sites are filled. T herefore, such
clays are said to be dioctahedral. W hen A l3+ is replaced
by Fe3+ or by other triva lent ions, the m inerals rem ain
dioctahedral. H ow ever, when the two A l3+ ions are
replaced by three divalent ions of Fe2+, M g2+, Z n2+,
etc., then all positions in the octahedral layer are filled and
the clay is trioctahedral. B ased on these distinc tions we
conclude that kaolinite is dioctahedral, w hereas
serpentine is trioctahedral.
The tetrahedral layer consists of silica tetrah e dra
that are linked together by sharing O atom s at three of the
four corners. The linkages occur only at the bases of the
tetrahedra, thus forming a sheet with the unshared O atom
s all pointing in the same direction. If all four O atom s of
silica tetrahedra are shared (as in feldspar), then each Si
atom con
trols only one half of its O atoms. Therefore, the
S i/O ratio in feldspar and other tektosilicates is
1: 2 (Section 8.3). Since in phyllosilicates only the three
basal O atoms are shared, the S i/O ratio is 1:1 .5 . W hen
the unshared O atom is counted with the basal ones, the S
i/O ratio becom es 1 : 2.5.
The octahedral layer has six-hydroxyl ions located at
the corners of an octahedron. Therefore, the ratio of Al to
hydroxyl ions is basi cally 1 : 6 . However, the octahedra
that m ake up the gibbsite sheet are linked at the corners
by
13.1 CRYSTAL STRUCTURE 201
sharing hydroxyl ions and have an A l/O H ratio of 1: 3
instead of 1 : 6 .
The tetrah ed ral and octahedral layers are joined to
form tw o-layer clays (T -O ), three-layer clays (T -O -T ),
or m ixed-layer clays that are mix tures of different tw o-
layer and three-layer clays. The linking of the octahedral
and tetrahedral sheets in T - O clay m inerals takes place
by sharing of O atom s and hydroxyl radicals, which
reduces the num ber of hydroxyls by one each time such a
link is established. T herefore, tw o-layer clays, such as
kaolinite, are com posed of Si, Al, O, and O H in the
proportions S i0 2 5 A l(0 H )2. We m ultiply the form ula by
tw o in ord er to clear the decim al frac
tion and obtain the form ula of kaolinite and its
polymorphs:
A l2 Si2 0 5 ( 0 H ) 4 (13.1)
Similarly, the form ula of three-layer clays
(T -O -T ), such as pyrophyllite, is 2 (S i0 2 s)A l(O H ) or
Si2 0 5 A l(0 H ) . A gain, we m ultiply by two and obtain:
A l2 Si4 O 1 0 (O H ) 2 (13.2)
The structures of two- and three-layer clays are illustrated
in Figure 13.1.
The phyllosilicate clay m inerals have widely varying
chem ical com positions because the A l3+ ion in the
octahedral layers can be wholly or p art ly replaced by
Fe3+, C r3+, Fe2+, M g2+, Z n2+, L i+, and m any other
ions. In addition, some of the Si4+ ions in the tetrah ed ral
layer are replaced by A l3+. This results in an excess of
negative charge, which is neutralized by the adsorption of
cations to the outer surfaces of the tetrahedral layers of
adjoin ing clay units. The substitutions within the layers
are constrained by the requirem ent for electrical
neutrality, which is m et w hen th e cations in the
octahedral layer have a com bined charge of
+ 6
and the tetrahedral layers contain four m oles of
Si + A l per form ula weight. T he tw o-layer clays
perm it very little substitution of either A l or Si, w
hereas the three -layer clays, w ith the exception of
pyrophyllite, are characterized by extensive
substitutions in either the octahedral layer or the
tetrahedral layer, o r in both.
202 CLAY MINERALS
(a) Two-layer clay: A lS i0 2.5(0 H ) 2
13.2 Classification and Chemical
\ ^ ^ ^ 7 Composition
\ 1 \ / \ / . /Tetrahedral layer

V v v V The classification of clay m inerals is based on

Octahedral layer their crystal stru ctu re and chem ical com posi
tion. In practice, clay m inerals are identified by x-ray
diffraction com bined w ith treatm en t p ro cedures th at
to g eth er serve to identify certain m ineral species. In o th
er words, som e clay m in erals are defined by the p ro
(b) Three-layer clay: AlSi2O s(O H ) ced u re th a t is used to identify them .

Tetrahedral

a. Two-Layer Clays
Octahedral The phyllosilicate clays are subdivided into the two-layer,
three-layer, and m ixed-layer clays. The two-layer clays
Tetrahedral consist of two subgroups, each of which contains a num
ber of m ineral species, as listed in Table 13.1.

1. KAOLINITE The m inerals in the kaolinite

subgroup all have the fam iliar form ula
(c) Smectites (e) lllite A l2 Si2 0 5 (0 H ) 4 and include kaolinite, dickite,
r i nacrite, and halloysite. These m inerals differ from
kaolinite only by having different crystal structures. For
Cations
I f K example, halloysite form s tubular
crystals, w hereas kaolinite crystals are platy, but
or
all m em bers of the kaolinite subgroup have
water\ /
identical chemical compositions. Therefore,
7 \ halloysite and the other m ineral species in the
L
L kaolinite subgroup are polym orphs of kaolinite

\ but occur in certain geochem ical environm ents in

preference to kaolinite.
2. SERPENTINE AND GREENALITE Serpentine
is a tw o-layer phyllosilicate like kaolinite in
(d) Verm iculite (f) Chlorite
which the gibbsite layer is replaced by brucite.
Consequently, the form ula of serpentine
[Mg3 Si2 0 5 ( 0 H ) 4] contains Si and O in the
characteristic p roportion of the tetrahedral
layer, and the octahedral layer contains a
+ 6
positive charge of as a result of the
replacem ent of two A l3+ by three M g2+. The
m ineral species in th at group are chrysotile,
lizardite, and antigorite, all of which are
polym orphs of serpentine. Greenalite [Fe2+
Figure 13.1 Schematic diagrams of the structures of two- Si2 0 5 (0 FI)4 ] is a tw o-layer clay in which three
layer and three-layer clay minerals. Fe2+ ions take the place of two A l3+ ions.
13.2 CLASSIFICATION AND CHEMICAL COMPOSITION 203
Table 13.1 Classification of Clay Minerals Table 13.1 (continued)
Group Subgroup Species Group Subgroup Species
A . Two-Layer Clays C. Mixed-Layer Clays
Kaolinite kaolinite kaolinite Complex interstratified two-layer and three-layer clay
(dioctahedral) dickite minerals following either a regular or a random
nacrite pattern.
halloysite D. Fibrous Clay Minerals
Kaolinite serpentine chrysotile palygorskite
(trioctahedral) lizardite (attapulgite)
antigorite sepiolite
greenalite
Adapted from Grim (1968) and Carroll (1970).
B. Three-Layer Clays the octahedral layer and
Pyrophyllite dioctahedral the partial replacem ent of
b. Three-Layer Si4+ in the tetrahedral
trioctahedral
dioctahedral
Clays layer.
The three-layer clays can
Smectite dioctahedral
be divided into the The smectite group is
pyrophyllite, smectite, divided into dioctahedral
vermiculite, mica, brittle and trioctahedral
Smectite trioctahedralmica, and chlorite groups.
subgroups each of which
A ll of the m inerals contains
included here have the
Vermiculite dioctahedralbasic three-layer structure
trioctahedralof pyrophyllite but differ
Mica dioctahedralfrom it in chem ical com
position and physical
properties.

Mica trioctahedral1. PYROPHYLLITE

AND TALC
Pyrophyllite
Brittle mica dioctahedralitself (A l2 Si4 O 1 0 (O H )2) is
three-layer clays and
trioctahedral
the octahedral or tetrahedral layers.
Talc is related to pyrophyllite
way th at serpentine
Chlorite dioctahedralbecause in talc the o
trioctahedralctahedral layer is com
posed of brucite instead
of gibbsite. C
onsequently, talc has the
form ula M g 3 Si4 O l0 (O
H )2, w here th ree M g2+
ions replace two A l3+
and the S i/O ratio is
derived from the presence
of two tetrah ed ral layers.
M innesotaite is th e Fe m
em ber in this group in
which the octahedral
layer contains Fe2+
instead of A l3+. C
onsequently, its form ula
is Fe2 +Si4 O 1 0 (O H )2.

2. SMECTITES The
smectite group contains a
large num ber of m ineral
species that differ in chem
ical com position because
of the com plete or partial
replacem ent of A l34' in
204 CLAY MINERALS
m ineral species that differ from each other in chem ical
composition. For example, m ontm orillonite is a prom
inent m em ber of the dioctahedral subgroup of smectites
in which Mg2+ occurs in the octahedral layer, but the
replacem ent of Si4+ by A l3 4 in the tetrahedral layer is
quite limited. For example, a sample of m ontm orillonite
from G erm any has the form ula (Carroll, 1970,Table 6 ):
[(Al[.7 7 FeQ0 3 M go2 o)(Si3 7 4 AlQ2 6 )O lo(O H )2]
(octahedral) (tetrahedral)
(13.3)
C a0 .i6 ' N ao.o7 Mgo.o4
(exchangeable)
This form ula can be readily interp reted in term s of the
three-layer m odel if we determ ine the elec trical charges
in each of the layers. The octahedral layer contains 1.77 A
l3+ + 0.03 F e3+ + 0.20 M g2+, which adds up to + 5 .8
electronic charges. Since th e octahedral layer is supposed
to contain six positive charges, th ere is a deficiency of 0 . 2
posi tive charge, which is equivalent to an excess of 0.2
negative charge. In the tetrahedral layer replacem ent of
som e Si4+ by A l3+ causes a defi ciency of positive
charge equivalent to a charge of -0 .2 6 . Evidently, the
substitutions in th e two kinds of layers have caused a total
charge of 0.2 + ( - 0 .2 6 ) = 0.46. In the example
before us, the negative charges are neutralized by
adsorption of 0.16 C a2+ + 0.07 N a+ + 0.04 M g2+, which
together provide 0.47 positive charge. The agreem ent
betw een the negative charges in the layers ( 0.46) and
the positive charges of the addi tional ions (+0.47) is well
within the uncertainty of the chemical analysis of this m
ineral. The overall
charge distribution for this m ontm orillonite is
given by adding up the ionic charges of all ele
m ents that m ake up this mineral:
1.77(A13+) + 0.03(Fe3+) + 0.20(M g2+)
+ 3.74(Si4+) + 0.26(A13+)
+ 0.16(Ca2+) + 0 .07(N a+)
+ 0.04(M g2+) - 10(O 2~) - 2(O H )
= +0 .01 (13.4)
The additional cations, w hich neutralize th e negative
charges caused by im perfect su b stitu tions w ithin th e
layers, are adsorbed on the su r
faces of the m ineral grains and on the
tetrah ed ral layers of the clay units. T herefore,
they are partly or com pletely exchangeable and can be
replaced by o th er cations or by the polar
m olecules of w ater in th e environm ent. The
adsorbed ions located on the o u ter surfaces of
the tetrah ed ral layers bond th e structural units
together. W hen the bonding involves the cations of m
etals, the interlayer distances betw een ad ja cent
structural units are short. H ow ever, w hen certain clay m
inerals are im m ersed in w ater, the interlayer cations are
replaced by m olecules of w ater, which have m uch sm
aller n et positive charges than m etal ions. C onsequently,
a larger num ber of w ater m olecules is required to n e u
tralize excess negative charges g enerated w ithin the
layers. As a result, the interlayer distances increase as w
ater m olecules crow d in and the clay expands.
The swelling of clay m inerals in the sm ectite group
is a characteristic property that is used to identify them .
For example, w hen natural m ont
m orillonite is heated at 300 C for an hour or
more, the interlayer distance decreases from
about 15 angstrom s (A ) to about 9 A because adsorbed w
ater m olecules are expelled. W hen the clay is
subsequently exposed to the vapor of eth ylene glycol (H
O C H 2 C H 2 O H ) for several hours, the interlayer
distance expands to 17 or 18 .
M ontm orillonite com m only forms by alter ation of
volcanic ash and therefore is the principal m ineral of
bentonite beds, which form by alter ation of volcanic ash
deposits. The well-known swelling of bentonite, when it is
placed in water, is actually caused by the expansion of the
interlayer distances of m ontm orillonite. The m ineral also
occurs in soils of arid regions w here leaching is lim ited
by lack of w ater and the pH is above 7.
Beidellite is a dioctahedral sm ectite contain ing little
if any Fe2+ or Mg2+ and in which nega tive charges
originate prim arily by replacem ent of Si4+ by A l3+ in
the tetrahedral layer. A ccording to G rim (1968), its form
ula can be w ritten as:
13.2 CLASSIFICATION AND CHEM ICAL COM POSITION 205

[(A l2 , 7 )(S i3 .1 7 A l0 .8 3 ) O Fe2+. Negative charges
N ote that the replacem
ent of Si4+ by A l3+ in
the tetrahedral layer has
caused a net charge of -0 .
83, which is partly com
pensated by the positive
charge (+0.51) generated
by an excess of 0.17 Al3+
in the
octahedral layer. The rem aining
charge ( 0.83 + 0.51 = 0.32)
adsorption of 0.33 moles of M g2 4 and L i+ occur together
weight of beidellite.
N ontronite, the third
dioctahedral sm ectite in
Table 13.1, contains Fe3+
in the octahedral layer in
place of A l3+ with m inor
am ount of Fe2+ and M
g2+. The net charge im
balance arises from sub
stitutions of Si4 4 by Al3+
in the tetrahedral layer.
Therefore, the form ula of
nontronite is:
[(Fe2 +)(Si3 6 7 Alo3 3 ) 0 1 0 (O H )
A s in beidellite, the
negative charge generated
in the tetrahedral layer
may be partly com
pensated by the presence
of additional Fe34, Fe24,
or M g2 4 in the octahedral
layer. W hen A l3+ in the
octahe dral layer is
replaced by Cr34, the
resulting dioc tahedral sm
ectite is know n as
volkhonskoite.
The trioctahedral
smectites include the m
iner al species saponite,
hectorite, and sauconite.
In
saponite, the octahedral layer
m arily of Mg2+ in place of
am ounts of Fe3+ and
arise in the tetrahedral
layer by replacem ent of
Si4 4 by A l3+ and may be
partly com pensated by
addition al cations in the
octahedral layer. The form
ula of saponite according
to G rim (1968) is:
[(M g ^ S i^ A lg j^ O n /O H ^ ] N a
Hectorite is a Li-bearing
layer, but no substitutions
occur in the tetrahedral
layer. Thus the form ula
for hectorite is:
[(Mg2 .6 7 Li0 .3 3 )(Siclay m ineral is +0.33, that
The sm ectite m ineral
stevensite is similar to
hec torite because it is M
g-rich and lacks Al in the
tetrahedral layer. How
ever, stevensite also lacks
Li in the octahedral layer.
charge of -1 .0 8 is neutralized by ad
The sm ectite m and Ca2 4 in the interlayer position
ineral sauconite is characteristically contains both M g2 4
character ized by the molecules in the interlayer
presence of Z n 2+ in the sites. The w ater can be
octahedral layer and by expelled by heating at
lim ited replacem ent of 500C, which causes the
Si4+ by A l3 4 in th e interlayer distance to
decrease from about 14 A
tetrahedral layer. C arroll
to about 9 . A fter
(1970) recalculat ed a
heating at 500 C or less,
chem ical analysis of
verm i culite rehydrates
sauconite to obtain the
spontaneously; however,
following form ula:
heating at 700C causes
irreversible dehydration. In
[(Z n 2 4 0 Mg0 1 8 A10 2 2 F'eQ.n)addi tion, when verm
(Si3 .4 7 A l0 ,5 3 )O iculite is treated with a
solution of KC1, the
The charge im balance in interlayer Mg2 4 ions are
the octahedral layer of this replaced by K 4,
and it becom es nonexpandable. Vermicul
of the tetrahedral layer is - fers from sm ectites by having a higher charg
0 .5 3 for a net charge of the tetrahedral layer, by commonly con
-0 .2 0 , which is
neutralized by the
adsorption of
exchangeable
cations in interlayer sites.
The w ord verm iculite
L atin verb vermiculari, w hich
worm s. The clay m ineral
b o th in dio ctah ed ral and
w ith o th er clay m inerals
p ro d u ct of micas. A ccording to
verm iculite from B are Hills, M aryland, has the
form ula:
[(Feo.24 ^ 6 2 . 7 0 Fegg4 Ni001)
(Si2.73Al1 .26 ) 0 1 o(O H )
Evidently, Mg2 4 dom
inates in the octahedral
layer, and the Si4 4 in the
tetrahedral layer is
extensively replaced by A
l34. The octahedral layer
has an excess positive
charge of + 0 .2 2 , which
partly com pensates for
the strong negative charge
of - 1 .3 0 in
the tetrahedral layer. The
206 CLAY MINERALS

M g2+ in the exchangeable interlayer sites, and by being 2M polym orphs and are derived by m echanical
less expandable. weathering of igneous and m etam orphic rocks. Therefore,
the 2M illite in sedim entary rocks is detrital in origin, w
3 . 1LUTE T he pow er of K + ions to stabilize clay
hereas the lM d or 1M poly m orphs may consist of
m inerals is well illustrated by illite, which is
authigenic illite. The x-ray diffraction spectra of the
listed in Table 13.1 w ith the m ica m inerals.
different muscovite poly m orphs in the clay fraction of
A ctually, illite could also be regarded as a special
sedim entary m ateri als were w orked out by Yoder and
m em ber of the sm ectite group. H ow ever, it
Eugster (1955).
differs from typical sm ectites by having a
nonexpandable lattice because of th e presence of K + ions 4. GLAUCONITE OR GLAUCONY Glauconite is a
in interlayer sites. Illite was nam ed by G rim et al. (1937) dioctahedral clay m ineral in which Fe3+, Fe2+, and Mg2+
after the state of Illinois in the U n ited States. T hey replace A l3+ in the octahedral layer. Negative charges in
intended it to represent m ica-like clay m inerals as a the octahedral and tetrahedral layers are neutralized by
group, rath er than a specific clay m ineral, and expressed interlayer K +, Ca2+, and N a+. The concentration of K +
the form ula of illite as: in glauconite is variable and increases with time depending
on the availability of K + in the environment. Johnston and
[(A l2, F e 2, M g2, M g 3 )(S i4 _I A l,)O 1 0 (O H )2] K , Cardile (1987) reported that glauconite from Point Jackson,
Francosia, Wisconsin, has the formula:
(13.11)
The form ula indicates that the octahedral layer m ay
contain Al, Fe, or M g and th at illites may be dioctahedral [(Fe2 097 A l0 8 4 9 M g 0 442Ti 0.003 M noom)
(2 A l3+ or 2 Fe3+) o r trioctahedral (3 M g2+). The value (Si3.61lAIo.38g)010(O H )2]K 0 7 2 5 C a(,n% (13.13)
of x is less than one and usu ally varies betw een 0.5 and
In this sam ple of glauconite, Fe2+ dom inates in
0.75. The negative charge generated in the tetrahedral layer
the octahedral layer, and the deficiency of posi
is com pensated by K + ions, which fit into hexagonal
tive charges in both the octahedral and tetrah e
holes form ed by the silica tetrahedra of the tetra
dral layers is largely com pensated by interlayer
hedral layers. A s a result, the K + ions are not K + and Ca2+ ions.
exchangeable, the spacing of the interlayers is G lauconite is an authigenic clay m ineral that forms
fixed, and the m ineral is not expandable. in the m arine environm ent by transform a tion from other
H ow ever, illites m ay be interlayered with m ont- kinds of m aterial (M cRae, 1972).
m orillonite or other clay m inerals th at are It commonly occurs as greenish pellets about
1 -2

expandable. mm in diam eter com prised of aggregates of

Illite is a com m on w eathering product of
m uscovite and differs from it prim arily by the
extent of substitution of Si4+ by A l3+ in the tetra hedral
layer. M uscovite and its N a-analog parago-nite are
dioctahedral and have the form ula:
[(A l2 )(Si3 A l)O 1 0 (O H )2 ]K, Na (13.12)
The charge deficiency in the tetrahedral layer of 1 is
neutralized by K + in muscovite and by N a ' in paragonite.
The stacking of adjacent mica units in m uscovite can
occur in different ways, giving rise to polym orphs
designated as 1M, lM d,2M ,, 2M2, and 3T (Smith and
Yoder, 1956). Illite in the smallest grain-size fraction ( < 0 . 2
pm ) of sedim ent or soil is of the lM d or 1M type. L arger
grains consist of the
platy glauconite crystallites about 1 0 pm in diam
eter. G lauconite has been used for dating
unm etam orphosed sedim entary rocks by the
K -A r and R b -S r m ethods (C hapter 16) because it
contains both K and Rb and because it is authi genic in
origin.
The status of glauconite as a discrete mineral species
has become controversial, partly because of its importance
for dating sedim entary rocks and the hoped-for refinem
ent of the geological time scale. Burst (1958) and H ow er
(1961) regarded glauconite as an Fe-rich illite with
variable K content that occurs in pure form as the lM d or
1M polymorph. However, this definition included poorly
crystal lized materials with low K concentrations that are
13.2 CLASSIFICATION AND CHEM ICAL COM POSITION 207
unsuitable for dating by isotopic methods. Therefore, Odin
and M atter (1981) proposed the w ord glaucony for a
spectrum of Fe-rich clay min erals with varying K
concentrations that form the characteristic green pellets in
sedim entary rocks of m arine origin. Glaucony therefore
represents clay minerals that are interm ediate between Fe-
rich and K -poor smectite and Fe-rich K-mica. As the K
con centration of glaucony increases, the material becomes
less expandable and m ore retentive with respect to the
radiogenic daughters of K and Rb.
5. MICAS The trioctahedral micas in igneous and m etam
orphic rocks include biotite and phlogopite. These differ
from muscovite by the chem ical com position of the
octahedral layer, but both replace one Si4+ by one AI3+ in
the tetrahedral layer. Actually, biotite is interm ediate betw
een phlogopite:
[(Mg3 )(Si3 A l)O 1 0 (O H )2] K
and annite or lepidom elane:
[(F e|+)(Si3 A l)O 1 0 (O H )2] K
Therefore, the general form ula for biotite can be w ritten
as:
[(Mg, Fe 2 +)3 (Si3 A l)O ,0 (O H )2] K (13.14)
and Bailey (1962) and by Eggleston and Bailey (1967). D
ioctahedral chlorite has been identified in soil sam ples
with pH <5 .5 and may also crys tallize m etastably in
sedim entary environm ents as an authigenic m ineral
(Carroll, 1970).
The trioctahedral chlorites have been classi fied on
the basis of their chemical com positions by Hey (1954)
and by Foster (1962). The latter speci fied com positional
limits for eight varieties of chlorite including the Fe-rich
species thuringite and chamosite and the M g-rich species
clinochlore
and penninite.
The structural form ula of trioctahedral chlo
rites includes the m ica unit (Mg, Fe2 +)3 (Si,
A l)4 O 1 0 (O H )2, which carries a negative charge
because of the replacem ent of up to 1.6 Si4+ by
A l3+ in the tetrahedral layer. The negative charge is
neutralized by a brucite sheet th at carries a pos itive
charge caused by replacem ent of Mg2+ by A l3+: (Mg, A
l)3 (O H )6. The brucite layer occupies an interlayer
position and bonds two mica units together as K + ions do
in illite (Figure 13.1). The com plete structural form ula of
chlorite is obtained by com bining the com ponents of the
mica and brucite units:
[(Mg, Fe2 +)3 (Si, A l)4 O 1 0 (O H )2 ](Mg, A l)3 (O H ) 6
(13.15)
Biotite is considerably m
ore susceptible to chem
ical w eathering than m
uscovite because the Fe2+
can oxidize to Fe3+ in the
presence of an electron
acceptor. As a result, the
charge balance of the
lattice is disturbed and the
m ineral alters to other
aluminosilicates and ferric
oxide, depending on
environm ental conditions.
The so-called brittle
micas in Table 13.1
include the dioctahedral
species margarite and
other rare trioctahedral .6 O 5 (O H ) 4 (13.16)
phyllosilicates listed by
Grim (1968).
6 . CHLORITE The
chlorite group contains
pri m arily trioctahedral
species com posed of
three-layer clay units
bonded to brucite sheets.
The chemical com
positions of chlorite range
from Mg-rich to Fe-rich
varieties, but all contain
varying am ounts of Al,
which replaces Si in the
tetrahedral layers.
In addition, chlorites may
contain Fe3+ ions, which
replace A l3+ in the
brucite layer.
Chamosite, one of
the Fe-rich chlorites, is an
im portant constituent of
Fe-rich sedim entary rocks
of m arine origin
described by James
(1966).
A nother Fe-rich phyllosilicate called
has a 1 : 1 kaolinite-type lattice, but has also been
called cham osite
(Brindley, 1961, Brindley
et al., 1968). This m ineral
has the formula:
and is structurally related
to kaolinite but com po-
sitionally sim ilar to
serpentine. How ever, it
differs from serpentine by
containing Fe in the octahe
dral layer and by lim ited
A l3+ in the tetrahedral layer.
B oth cham osite and
genic m inerals th at form

The only
dioctahedral chlorite is an
Al-rich variety known as
cookeite described by
Brown
208 CLAY MINERALS

recrystallize during low -grade regional m etam or phism to
the Fe-chlorite thuringite.
c. Mixed-Layer Clays
The clay m inerals we have described occur as
interstratified layers in the so-called mixed-layer clays,
which are very com m on in soils and in about
70% of clay-rich sedim entary rocks (W eaver,
1956). To some extent the principal types of
m ixed-layer clays depend on climatic factors that affect
the environm ent within which chemical w eathering
occurs. U nder m oist and tropical con
ditions the stratification in m ixed-layer clays is
m ontm orillonite-halloysite-kaolinite, w hereas in
hum id tem perature regions the m ixed-layer clays consist
of m ontm orillonite-chlorite-m ica or m ica-interm ediate
products-illite (Carroll, 1970). The crystallographic
properties of mixed-layers
clays have been discussed by Z en (1967),
R eynolds (1967, 1980), and C orbat and
Tettenhorst (1987).
d. Fibrous Clay Minerals
The fibrous clay m inerals palygorskite and sepio-lite are
com posed of double chains of silica tetra-
hedra with an S i/O ratio of 4 : 11. The channel
betw een the double chains is occupied by w ater
m olecules that are lost stepwise during heating up
to 850C, at which tem perature the structure is
destroyed. N either palygorskite nor sepiolite
expand w hen treated w ith organic liquids.
P alygorskite is a M g-silicate with some
replacem ent of M g2+ and Si4+ by A l3+. Its struc ture
was w orked out by B radley (1940), who reported the form
ula:
5 In order to overcom e this problem, Tardy and
M g Si8 O 2 0 (O H )2 (O H 2 ) 4 4 H zO (13.17)
w here O H 2 is bound w ater and H 20 is w ater in
channels like th at of zeolites. The structure p re sented
above is electrically neutral but m ay con tain
exchangeable C a2+ to com pensate charge im balances
caused by lim ited substitutions in this structure.
Palygorskite is called attapulgite in the U nited States after
an occurrence of this clay in A ttapulgus, G eorgia.
Palygorskite is an authigenic mineral that occurs in
alkaline, unleached soils and as a product
of hydrotherm al alteration of pyroxene and amphi-bole.
One of the largest accumulations is in the Hawthorn Form
ation of M iocene age in G eorgia and Florida, which was
deposited in fresh to brack ish water lagoons.
Sepiolite is also com posed of chains of silica
tetrahedra but it differs from palygorskite in spe cific
details. G rim (1968) reported two different structural form
ulas for this m ineral based on a review of the literature
available at the time. The
form ulas currently in use (see A ppendix B) are
Mg2 Si3 0 6 ( 0 H ) 4 and Mg4 Si6 O t5 (O H ) 2 6H 2 0 .
Carroll (1970) reported that a chemical analysis of
sepiolite from L ittle C ottonw ood, U tah, yields the
formula:
[(Mg7 4 2 MnQ5 3 Feon2) (Si| 1 6 7 A l024Feo J9 ) 0 3 2]Cu() 4 5
(13.18)
Evidently, M g2+ is partly replaced by M n2+ and Fe3+, w
hereas Si4+ is replaced by A l3+ and Fe3+. The structure
has a net charge of -0 .3 7 , which is largely neutralized by
exchangeable Cu2+.
13.3 Gibbs Free Energies of
Formation
The therm odynam ic properties of the three-layer and m
ixed-layer clay m inerals are difficult to establish because
of their highly variable chem i cal compositions. As a
result, standard free ener gies or enthalpies of form ation
are available only in certain specific cases or apply only to
idealized com positions th at do not reflect the true
chemical diversity of real clay minerals.
Garrels (1974), Chen (1975), and Nriagu (1975) pro posed
empirical m ethods for estimating standard free energies of
form ation of clay minerals. Tardy and Garrels
(1976,1977),Tardy and G artner (1977), Tardy and
Vieillard (1977), and Tardy (1979) subse quently
generalized this approach to estimate the standard free
energies of form ation of other kinds of compounds and
their ions in aqueous solution.
The m ethod of Tardy and G arrels (1974) is based on
the prem ise th at the standard free ener gy of a
phyllosilicate can be treated as the sum of
13.3 GIBBS FREE ENERGIES OF FORMATION 209
th e free energies of its oxide and hydroxide com ponents. E qu atio n s 13.22-13.24 can be solved by sub
The num erical values of the free energies o f the com stitution and yield the standard free energies
ponents m ust reflect the fact that they of fo rm atio n in th e silicate lattice of th e th ree
exist in a silicate lattice and not in free form . com ponents listed in Table 13.2. O th er com po
T herefore, they m ust first be determ ined from n en ts can be ad d ed by considering the A l sili
phyllosilicate m inerals whose standard free ener cates kaolinite, pyrophyllite, and m uscovite,
gies of form ation are know n from experim ental which yield values for G^(A12 0 3), G ^(H 2 O s),
determ inations. The first step tow ard th at end is and G ^(K 2 O s).Tardy and G arrels (1974) also
to restate the structural form ulas of phyllosili- determ in ed th e stan d ard free energies of fo rm a tion in
cates in term s of oxide com ponents, except that the silicate lattice of exchangeable cations and w ater. T
M g is converted to the hydroxide until all of the hydroxyl hese values are listed in Table 13.2.
ion in the form ula has been assigned to M g(O H )2. For
example, serpentine is rew ritten: In order to illustrate their m ethod of estim at ing
standard free energies of form ation of a real
M g3 Si2 O s(O H ) 4 = 2M g(O H ) 2 M gO 2 S i0 2
(13.19)

Similarly, talc becomes: Table 13 .2 Standard Free Energies of Formation

M g 3 Si4 O 1 0 (O H ) 2 = M g(O H ) 2 2M gO 4 S i0 2
of Oxide and Hydroxide Components in the
(13.20)
Lattice of Phyllosilicate Minerals at 25 C

and sepiolite, which is not a phyllosilicate, never theless is

Com ponent Gf (silicated), kcal/mol
recast into the form: K 2O -188.0
M g2 Si3 0 6 ( 0 H ) 4 = 2M g(O H ) 2 3 S i0 2 (13.21) NazO -162.8
MgO -149.2
The three Mg-silicates are composed of M g(O H )2,
MgO, and SiOz, whose standard free energies in the FeO -64.1
silicate lattice are to be determ ined. This is done by Fe20 3 -177.7
assuming that the known standard free energies of the A i2o 3 -382.4
three m inerals can be expressed as the sums of the S i0 2 -204.6
unknow n standard free energies of the com ponents in the H 2O -59 .2
silicate lattice. Therefore, we can write appropriate Mg(OH)2 -203.3
equations for each of the three minerals. For serpentine: A l(O H )3 -280.0
FeOOH -117.0
2G (M g(O H )') + G )(M gO s) + 2Gf {SiO s2) KOH ' -123.6
NaOH - 111.0
= -965 .1 kcal/m ol (13.22)
Exchangeable Ions Expressed as Oxides
For talc: K 2O -188.0
G y(M g(O H )2) + 2G ?(M gO s) + 4G ?(SiO |) CaO -182.8
= -1320 .38 kcal/m ol (13.23) NazO -175.4
MgO -159.5
A nd for sepiolite: H 2O -58 .6
2G ?(M g(O H )s) + 3G ?(SiO |) Li2 -190.6
= -1020 .95 kcal/m ol (13.24) s o u r c e : Tardy and Garrels (1974).
210 CLAY MINERALS

clay m ineral, Tardy and G arrels (1974) chose a typical m Table 13 .4 Standard Free Energies of Formation
ontm orillonite: of Clay Minerals Based on Experimental
Determinations and Estimates by the Method of
[(A lisF eQ ^ M ggjX S ijjA lt^ O u /O H ^ ] M g04
Tardy and Garrels (1974)
(13.25)
G} G
The form ula is electrically neutral, which is im portant Mineral or (measured), (estimated),
because the exchangeable cations do contribute to th e free locality kcal/m ol kcal/mol
energy of the m ineral and therefore cannot be om itted . In Dlite
Table 13.3 the for m ula for this m ontm orillonite has been
recast in term s of m olar am ounts of the oxide com Fithian -1277.23 -1278.8
ponents. These are then m ultiplied by their respective G^ Beavers Bend 1274.7a -1276.2
(silicated) values from Table 13.2 and the products are Goose Lake -1272.1s -1273.4
summed. The result is 1282.5 kcal/m ol, which is a Montmorillonite
reasonable value for the standard free
Aberdeen -1230.6C -1228.3
Belle Fourche 1240.6b -1236.4
energy of form ation of m ontm orillonite. For
exam ple, A ppendix B contains a value of Colony, Wyoming 1241.5d -1240.6
1258.84 kcal/m ol listed by Lindsay (1979) for a Other Phyllosilicates
M g-m ontm orillonite. Annite (1) -1149.3 -1150.2
M ore specific com parisons betw een free Annite (2) -1151.7 -1149.1
energies of form ation, estim ated by the m ethod
Clinochlore -1961.8 -1958.7
of Tardy and G arrels (1974), and experim entally
Phlogopite -1410.9 -1400.7
determ ined values are presented in Table 13.4.
Greenalite -720.0
T he experim ental determ inations w ere m ade by
Minnesotaite -1055.0

Dlite formulas
Table 13 .3 Estimate
of the Standard Free Fithian,
Energy of Formation of a Typical [(A l1, 4Fea29Mg0.l9)(Si3, 1Al[..49)O
M ontmorillonite with the Formula Stated in 10(OH)2]Ka64 Beavers Bend,
Equation 13.25 [(AI] 6ftFeQ 2()Mgo i3)(Si3 ^2A1q 3g)OjQ(OH)2]K(j 53
1 2 3 Goose Lake,
Component, G (silicated), 1X2, [(AIj 5^FeQ24Mgo i5)(Si3 AIq 3 5 )0 ]q(0 H)2 ]Kq j y
moles kca l/m o l kcal
Montmorillonite formulas
1.0 a i 2o 3 -382.4 -382.4
Aberdeen,
0.1 Fe20 3 -177.7 -17.77 [(Ali.29Feo335Mg0 445)(Si3 82A10 i9)O
0.3 Mg(OH)2 -203.3 -60.99 10(OH)2]Ka415 Belle Fourche,
3.5 S i0 2 -204 .6 -716.1 [( Al i .5 lsFejj 225Mg0 29) (Si3 935Al0
0.7 H2Oa -59 .2 -41.44 065) O Colony,
0.4 MgO(exch.) -159 .5 -63 .8 [(A li,2 Fe3 l 2 Mgo.2 9 )(Si3 .8 1 Alo.i9 )Oio(OH)2 ]K0.4()
Sum -1282.5 kcal
"Routson and Kittrick (1971).
a0.3 Mg(OH)2 uses up 0.6 hydrogen atom, leaving 1.4 hydro gen bKittrick (1971a).
atoms to make 0.7 mole of H20 . Kittrick (1971b).
S O U R C E : Tardy and Garrels (1974). dWeaver et al. (1971).
 13.4 STABILITY DIAGRAMS 211
dissolving the m inerals and by analyzing the other direct com parisons in Table 13.4 are also
resulting saturated solutions. For exam ple, impressive.
W eaver et al. (1971) dissolved a sam ple of m ont-m
orillonite from Colony, Wyoming:
13.4 Stability Diagrams
[(A li.5 2 F e 0 .2 2 ^ 0 .2 9 )(^ *3 .8 lA lo .i9 )0 10( O H ) 2] M g 0 2o
The three -layer clays are m ore sensitive to the
+ 6.10 H + + 3.57 H 2 0 ^ geochem ical environm ent in which they exist
1.71 A l3+ + 3.81 H 4 S i0 4 (13'26) than m ost o th er m inerals because of the p res
+ 0.11 F e 2 0 3 (s) + 0.49 Mg2+ ence of exchangeable ions and m olecules. T heir
approach to equilibrium takes place by exchange of the
adsorbed ions with ions or m olecules from th e environm
A t equilibrium in a saturated solution the Law of Mass A
ent, by transform ations involving the chem ical com
ction applies:
position of the layers and their
[Mg2 +]0 4 9 [H 4 SiO 4 ]3 8 1 [Al3 + ] 1 -7 1 stacking, and perhaps even by th e form ation of
[H + ] 6 1 0 m ixed-layer clays. A lthough ion exchange is
rapid, reactions th at require adjustm ents in the
(13.27)
chem ical com positions of the octahedral and
where Fe2 0 3 (s) and H 2 0(1) were given activities of one. tetrah ed ral layers, and in the bonding betw een
The value of the equilibrium constant, cal culated from individual clay units, are slow. Consequently, n a t ural m
equation 13.27 using the m easured activities of the ions aterials com m only contain m ixtures of several clay m
and molecules, was found to be K = IO0" 1. Since log K = inerals th at cannot be in equilibri um with each o th er u n
-AG*/1.364, it fol lows that AGR = 1.352 kcal. The d er one set of environ m ental conditions. For this reason,
standard free energy change that takes place during the we now turn
disso ciation of the m ontm orillonite from Colony, to the construction and use of activity diagram s
Wyoming, in accordance with equation 13.26 can be for clay m inerals in o rd er to identify the stable
expressed as: phases and to indicate the direction of reactions am ong the
clay m inerals in a given geochem ical environm ent.

A GR = 1.71 G(A13+) + 3.81 G (H 4 S i0 4)

+ 0.11 G (Fe2 0 3) + 0.49G (M g2+) a. Aluminosilicate Clay Minerals

- 3.57 G ?(H 2 0 ) - 6 .1 0 G j(H +) - G (m ont.) In order to simplify the construction of stability diagrams,
we restrict ourselves to a suite of m ont-m orillonites in
(13.28) which negative charges in the tetrahedral layers are com
Since AGR for this reaction is know n from the pensated by different adsorbed cations, w hereas the
octahedral layers consist of unaltered gibbsite sheets. These
m easured equilibrium constant, G / of the m ont ideal ized m ontm orillonites and their standard free
m orillonite can be calculated from the Gf values energies of form ation w ere proposed by H elgeson (1969)
of the other constituents. Tardy and G arrels (1974) and H elgeson et al. (1978), and are listed in Table 13.5
obtained a value of Gf = 1241.5 kcal per mole for the together w ith a standardized illite. These data perm it us to
Colony m ontm orillonite from explore the stabilities of m ontm orillonites in the sam e
the solubility data, w hereas their own m ethod coordinates we used in C hapter 12 to depict the stability
yields Gf = - 1240.6 kcal per mole. The agree fields of other alum inosilicate minerals. R eal m ontm
ment is excellent. H ow ever, W eaver et al. (1971) orillonites and other sm ectites should react similarly after
obtained G ^(m ont.) = -1255 .8 0.6 kcal/m ol
for the M g-substituted m ontm orillonite. The
212 CLAY MINERALS

Table 13 .5 Formulas and. Standard Free Energies of

Form ation of Standardized Montmorillonites and Illite
Formula G-, kca l/m o l
Montmorillonite

[(A W (S i3.67A W O I0(OH)JCaai67 -1279.24

[(Al2.00)(Si3.67Ala33)O1()(OH)2]Nan.33 -1277.76
[(Al2.0o)(Si3.67AIo.33)Oio(OH)2]Koj3 -1279.60
[(AIzqo) (^ 3.67^^0.33)^ 10(^ 11)2] 167 -1275.34
Illite

[(All.8oMgo.25)(Sb.50JMo.5o)Olo(OIi)2]lC().33 -1300.98
(1307.485)

Estimated by the method ofTardy and Garrels (1974).

S O U R C E : Helgeson (1969).
allowance is m ade for
their m ore complex chem
i cal compositions.
Based on the stability
diagram s we derived in C
hapter 12, we can say that
kaolinite is the stable
alum inosilicate clay m ineral
w here product ions are
the activity of H 4 S i0 4 is
even kaolinite breaks
down to form gibbsite.
Therefore, we can reverse
the decom position of
alum inosilicate m inerals and
m orillonite and illite m ight
when the activities of their
ions increase in solu
tions that are saturated with lo
phous silica. Figure 13.2 is an g
stability diagram in [
coordinates of log [K+]/[H N
a
+] and log [N a+]/[H +],
+]
assum ing saturation with
/
respect to am orphous
[
silica. The equations for
H
the boundaries are based
+]
on the therm odynam ic
data in Table 13.5 and in A
ppendix B. The reaction to Figure 13.2 Stability of
form K -m ontm orillonite kaolinite, K-
montmorillonite, and Na-
constructed based on the conservation of Al:
montmorillonite in the
7 A l2 Si2 0 5 ( 0 H ) 4 + 8 S i0presence of amorphous
silica based on data in
6[(A12 0 0 )(Si3 6 7 A10
Table 13.5 and Appendix
+ 7 H zO B.

If we multiply kaolinite by
7 and K -montmorillonite to the left side of the
by 6, we can balance AI to equation to balance Si
a good approximation. As and 2 moles of K + to
a result, we must add 8
balance K. Finally, we
moles of S i0 2
need 7 moles
(amorphous)
of H 20 on the right side to balance O and 2 moles of
H + to balance H . Equation 13.29 yields
 13.4 STABILITY DIAGRAMS 213
AGR = +35.511 kcal and K = l ( r 2 6 0 3 . From the
Law of Mass Action:
[ I2 = | Q 26.03 (13.30)

[K + ] 2
and therefore (line 1, Figure 13.2):

rK +l
log FTTZi- = 13.02 (13.31)
[H-
Equation 13.29 i n d i c a t e s that K-montmorillonite

is f a v o r e d o v e r k a o l i n i t e i n e n v i r o n m e n t s w h e r e [ H +] is
l o w a n d [ K + ] is h i g h . S u c h c o n d i t i o n s o c c u r i n p o o r l y d r
a i n e d s o i l s u n d e r a r id c l i m a t i c c o n d i t i o n s o r in t h e o c e
log [M g2+]/ [H +]2
answhereseawaterprovidesa

source ofK+atapH o f a b o u t 8.2. T h e f o r m a t i o n Figure 13.3 Stability of selected minerals in the sys tem
ofNa-montmorillonite from kaolinite takesplace K20 - M g 0 -A l20 3- S i0 2-H 20 -H C l at 25 C at
b y a r e a c t i o n t h a t is a n a l o g o u s t o e q u a t i o n 13.29 1 atm pressure in the presence of amorphous silica. The
(line 2, F i g u r e 13.2, l o g [ N a + ] / [ H + ] = 13.32). positions of the lines were adjusted as indicated in the
The direct conversion of K-montmorillonite footnotes to Table 13.6 to overcome internal inconsis
toNa-montmorilloniteinvolvesonlyanexchangeoft
tencies in the thermodynamic data. Nevertheless, the
diagram shows that the clay minerals kaolinite (K), illite
hetwoelements:
(I), montmorillonite (Mg-mont.), and chlorite (Ct) can
3 [ ( A l 2 . 00)( S i3 . 67A l0 .3 3 )O 10( O H ) 2] K 0.33 + N a + ^ form as weathering products of microcline (Mi),
3 [ ( A 1 2.oo) ( S i 3 .67 A l 0 33) O 10(O H ) 2] N a 0 33 + K + muscovite (Mu), and phlogopite (Ph), which crystallize
from cooling magma or grow by metamorphic recrys
(13.32) tallization of clay minerals at elevated temperatures and
AG = +0.413 k c a l K = IO 0 -3 0 2 pressures.
Therefore,atequilibrium:

[K+ -0.302 (13.33) value of AGR of a reaction changes its equilibrium

1 0
constant by a factor of 1 0 and displaces the corre sponding
[Nal
stability boundary in the activity dia gram . In order to
and hence (line 3, Figure 13.2):
avoid som e of the resulting inconsistencies, the standard
lg THTT = log 7TTT -30 (13.34) free energy of form a tion of illite of H elgeson (1969) in
Table 13.5 was recalculated by the m ethod of Tardy and G
[Hi [H i arrels (1974), which increased it from -1300 .98 to -1307 .
We can also construct a stability diagram in 49 kcal/m ol. In addition, the value of Gf of M g-m ontm
coordinates of log [K+]/[H +] and log [Mg2 +]/[H +]2, orillonite was changed from 1275.34 to -1279.61 kcal/m
which includes kaolinite, illite, muscovite, m icro ol. These and other adjust m ents are listed at the bottom of
cline, phlogopite, and chlorite in contact with Table 13.6.
am orphous silica. The equations for such a dia
gram are listed in Table 13.6 and have been plotted in The resulting stability diagram in Figure 13.3
Figure 13.3. indicates that in acidic low -K environm ents p h lo gopite
During the construction of Figure 13.3 we (or biotite) alters to chlorite, M g -m ontm o rillonite, or
encountered difficulties caused by the lack of internal kaolinite, depending on th e activity of Mg2+. A lthough
consistency of som e of the therm odynam ic data. A phlogopite m ay be transform ed directly into each of the
difference of 1.364 kcal in the calculated alteration products, sys-
214 CLAY MINERALS

Table 13.6 Equations for Mineral Stability Diagram (Figure 13.3) in the System K20 - M
g 0 - A l20 3- S i0 2-H C l-H 20 at 25 C, 1 atm Pressure, and in Equilibrium with
Amorphous Silica
Kaolinite-muscovite
3 Al2Si20 5(0 H )4 + 2 K+^ 2 KAl2Si3AlO 10(OH)2 + 3 H20 +2H+

AGfi = +10.959 kcal K = IO803

log f| i] = 4.02 (1)

Illite-muscovite

1.3 [(Alj 80Mg025)(Si35QAl050)O 10(OH)2]KU6Q+ 0.22 K+ + 0.43 H ^

KAl2Si3A lO 10(OH)2 + 1.55 S i0 2(amorph.) + 0.325 Mg2+ + 0.5 H20
A GR = 1.6583 kcal K = 10L21S

log^|=1.481og^?-5.52 (2)

Kaolinite-Mg-montmorillonite

7 A l2Si20 5(0 H )4 + 8 S i0 2 (amorph.) + Mg2+^

6 [(Al20n)(Si3 67Al0 33)O||)(OH)2]Mg(l 167 + 7 HzO + 2 H

AGR = + 8.75b kcal K = IO8'42

l0g^ F =6-42 (3)

Illite-Mg-montmorillonite

1.013[(A1, 8nMg025)(Si350Al05u)O ,0(OH)2K060 + 0.1245 S i0 2(amorph.) + 0.781 H +-

[(Al2 oo)(Si3.f,7Al 33)O |o(OH)2]Mgu 167 + 0.608 K* + 0.086 Mg2+ + 0.405 H20
A GR = -3 .2 8 ab kcal K = IO241

l08f iF j = - - 141og^ r + 3-96C W

Mg-montmorillonite-chlorite

[(Al2()D)(Si36 7Alo3 3)0 1o(OH)2 ]Mg( ) 1 6 7 + 5.658 Mg2+ + 9.32

1.165 Mg5Al2Si3O l0(OH)8 + 0.175 S i0 2(amorph.) + 11.316 H +
AGR = + 87.67 kcal K = IO64'27

Inn tMgZ+] - 11 Ifi 151

S [H ]2' ^
 13.4 STABILITY DIAGRAMS 215
Table 13 .6 (continued)

Chlorite-phlogopite

Mg5Al2Si3O10(OH)8 + Mg2+ + 3 S i0 2(amorph.) + 2 K+^ 2 KMg3Si3A lO 10(OH)2 + 4 H

A GR = + 26.60e kcal K = 1(T19-50

[K+] n ^ w t Mg2+
, [H + ]- 0 .5 1 o g [H+]

lc,g f m T = - - 5 1 g T 4 ir + 9-75f (6)

Phlogopite-illite

2.30 KMg3Si3A lO 10(OH)2 + 15.0 H ^ ^ [(A li a0Mg0 25)(Si3.50Ala50)O10(OH)2]Ka60

+ 1.70 K+ + 6.65 Mg2+ + 3.40 S i0 2(amorph.) + 8.8 H20

A GR = -113.98 kcal K = 10+83-57

log W \ = ~3-91 log + 49-15 (7)

Microline-illite

2.3 KAlSi3Os + 0.25 Mg2+ + 0.4 H20 + 1.2 H+

[(Al1.80Mg0.25)(Si3.50Al0.50)O 10(OH)2]K0.60 + 1.7 K+ + 3.4 S i0 2(amorph.)

AGR = -14.67 kcal K = 10+10-76

[K+l TMe2+l
log W \ = -147 log + 633 (8)

Muscovite-microcline

KAl2Si3AlO 10(OH)2 + 6 S i0 2(amorph.) + 2 K +^ 3 KAlSi3Oa + 2 H +

AGr = +14.43 kcal K = l(T la58

1o8 | | t | = 5-29 (9)

Microcline-phlogopite
KAlSi30 8 + 3 Mg2+ + 4 H2O ^K M g3Si3A lO 10(OH)2 + 6 H +

AGR = +44.948 kcal K = IO32-95

10-98
log^ T = (1 0 )

Based on Gf (illite) = 1307.485 kcal/mol, according to Tardy and Garrels (1974).

bGj- (Mg-montmorillonite) = -1279.61 kcal/mol adjusted so that the kaolinite-Mg-montmorillonite line goes through
log [K+]/[H +] = 4.02 and log [Mg2+]/[H +]2 = 6.42.
Increased intercept to 4.86 to force line through log [K+]/[H +] = 4.02 and log [Mg2+]/[H +]2 = 6.42.
dG^(chlorite) = -1974.0 kcal/mol (Zen, 1972).
G^(phlogopite) = -1400.7 kcal/mol (Tardy and Garrels, 1974).
'Reduced to 9.0 to improve the fit.
216 CLAY MINERALS

tem atic variations in the chem ical com position of the Table 13 .7 SolubilityEquations of Pure Fe-
fluid in the im m ediate vicinity of a m ineral grain m ay Silicate Minerals in Pure Water at 25 Ca
cause a sequence of reactions to occur such th at phlogopite
Fayalite
first alters to chlorite, which changes to M g -m ontm
Fe2S i0 4 + 4 H +^ 2 F e 2+ + H4S i0 4
orillonite, which ultim ately
decom poses to kaolinite or even gibbsite. A GR = -20 .76 kcal K = 101522
Similarly, illite m ay form either m icrocline or
m uscovite by a reduction of [K +] o r decrease in pH . D log I f f = ~0'5 108 [l*SiC)J + 7-61 (1)
epending on the existence of the appropri ate chem ical Ferrosilite
gradients, m uscovite flakes m ay alter to kaolinite or illite, FeSiOj + HzO + 2 H +^ F e 2+ + H4S i0 4
which m ay transform into AGR = -7.663 kcal K = IO5 62
illite -m ontm orillonite or illite -kaolinite mixed-
layer clays. H ow ever, predictions of reaction
products based on therm odynam ics alone are not always log I f f = los [Ii4Si 4] + 5-62 (2)
reliable because unfavorable kinetics may p rev en t the Greenalite
reactions from reaching equilibrium or from going to com Fe3Si20 5(0 H )4 + 6 H+^
pletion . 3 Fe2+ + 2 H4S i0 4 + H20
AGfi = -18.577 kcal K = 1013 M
b. Solubility of Ferrous Silicates
The ferrous silicates include fayalite (Fe2 S i0 4), ferrosilite
(F e S i0 3), and the tw o-layer phyllosili cate greenalite. In log I F F =~0-67108[H4Si 4] +0,53 (3)
addition, the carbonate m ineral siderite (F e C 0 3) and the Siderite
hydroxide [Fe(OH)2] m ay form . We assum e th at Fe is F eC 03 + 2 H +^ F e 2+ + C 0 2 + H20
predom inantly in the divalent state and th a t m inerals AGR = +0.509 kcal K = IO037
containing Fe3+ do not occur in the system . loS t S J = 3 1 5 at [C J = 3 x 1 C r 4 atm (4)

The ferrous silicates dissolve congruently much like [H J

the Mg-silicates we discussed in Chapter 12 (Figure 12.5). Ferrous hydroxide
The solution of the ferrous silicates is expressed by the Fe(OH)2(s) + 2 H +^=Fe2+ + 2 HzO

(5)
equation in Table 13.7, which have been plotted in Figure A GR = -15.924 kcal K = IO1167
13.4 in coordinates of log [Fe2 +]/[H + ] 2 and log [H4 S i0 4]. lFe2+l
The lines are the boundaries betw een solutions that are iogW =1L67
undersatu rated or supersaturated with respect to each Amorphous silica
miner al. For example, fayalite dissolves in solutions S i02(amorph.) + 2 H20 ^ H 4S i0 4
represented by points lying to the left of the equilib rium A G R = +4.044 kcal K = 10~2%
line (equation 1, Table 13.7). Evidently, greenalite log [H4S i0 4] = -2 .96 (6)
(equation 3, Table 13.7) is less soluble, and therefore m ore
stable, than either fayalite or ferrosilite in contact with w Based on thermodynamic constants in Appendix B.
ater at 25 C. However, the com position of natural
solutions in contact with the atm osphere is restricted by
the solubility
The solubility data in Figure 13.4 indicate th at
,boundaries of siderite (equation 4, Table 13.7) and am greenalite could precipitate in place of siderite if log [Fe 2
orphous silica (equation 6 , Table 13.7). Therefore, even +]/[H + ] 2 is greater than about 7. The fugacity of C O z th at
greenalite dissolves in natural solu tions and precipitates w ould perm it this can be cal culated from the solubility
only from solutions that have low bicarbonate ion equation for siderite in Table 13.7:
activities.
13.4 CO LLO IDAL SUSPENSIONS AN D ION EXCHANGE 217
 T r~ m ost com m on iron silicate m ineral in unaltered
\ i parts of the G unflint F orm ation in w estern
\
IFe (O H )2
\ Ontario, C anada. It occurs as light, green to olive green
\
10 granules up to 3 m m in diam eter, except in the banded ch
- "\
\ \ ert-sid erite and tuffaceous shale facies. The absence of
X\ X\j)' greenalite from siderite - bearing rocks is explained by the
\ N - f - '- ? difference in their solubilities w hen th e solutions are in
+ \ I N% . '-v
contact with C 0 2 of the atm osphere. H ydrotherm al solu
X [C 0 2] = 10 7 4 atm \ SideriteX \ %
>sl \ \
tions discharged into th e G u nflint basin could precipitate
greenalite and am orphous silica if they had sufficiently
CN Greenalite | \
6 precipitates low carbonate ion concen trations to prevent siderite from
00
\ forming.
o
 low carbonate ion immiscible in an o th er
4- M innesotaite is com
concentrations and are liquid are know n as e m u
[C 0 2] = 10 3 5 atm m on in the Fe-silicate
2- Natural l sions, and solid particles o
close to being saturated facies of the Biwabik Iron
r liquid droplets dis persed
with respect to am Form ation of northern
in a gaseous m edium form
solutions Minnesota, but is absent
orphous silica. Such aerosols. Such colloidal
from the G unflint
solutions m ay be dis suspensions m ay appear to
-8 -6 Formation. Jam es (1966)
charged by volcanic be h om oge neous, and the
lo pointed out that this dif
particles do n o t settle o u t
g hotsprings in shallow m ference betw een the two
under the influence of
[ arine iron form ations m ay be
gravity. Som e kinds of
H basins w here cherty iron form ations
related to the fact th a t th
particles in sols attract w
4 deposited in Precam brian e Biwabik Form ation was
ater m olecules to them
S G oodw in (1956) rep o rted intruded by the E m barras
selves and are therefore
i0 G ranite over m uch of its
said to b e hyd ro p h ilic, w
4] length, w hereas the G
hereas others lack this
unflint Form ation was not
tendency and are classified
Figure 13.4 Solubility of intruded by granite.
as h y d r o p h o bic. A sol
ferrous silicates in pure Consequently, he agreed
water at 25 C. com posed of hydrophilic
with Tyler (1949) that m
Greenalite, which is a particles m ay set to form a
innesotaite is a product of
two-layer clay analogous translucent o r tran sp aren t
m eta morphism, which m
to kaolinite and gel. Som e
serpentine, is soluble in ade the Biwabik Form
natural solutions open to ation m ore susceptible to
the atmosphere. It can chem ical w eathering than
pre cipitate at 25 C, only the Gunflint Form ation
when [C 02] = IO7 4 and resulted in the form
atm or less, from ation of hem atite ore
solutions that are nearly bodies in M innesota.
saturated with respect to
amorphous silica.
Greenalite is the most
abundant primary Fe-
silicate in the Gunflint 13.5 Colloid
Iron Formation of
al
Proterozoic age in
western Ontario, Suspen
Canada, and its absence
from the siderite facies sions
implies that the siderite and
facies was deposited
where the fugacity of Ion
C 0 2 was greater than IO
7'4 atm.
Exchan
ge
W hen clay m inerals and o
[Fe.2+1 10
th er sm all particles with
[H + ] 2
[C O J diam eters of < 1 g m are
which yields [COz] = dispersed in pu re w ater,
they form a colloidal
greenalite can suspension called a sol.
precipitate only from Suspensions of sm all
solutions that have very droplets of a liquid th at is
218 CLAY MINERALS
gels can be dispersed again into a sol, w hereas in others
the process of gel form ation is irreversible. For example,
concentrated solutions of silic acid can polym erize to form
am orphous silica parti cles, and the resulting colloidal
suspension can set to form a gel. H ow ever, the silica gel
does not redisperse spontaneously, but slowly loses w ater
as it crystallizes to form quartz (variety chal cedony) with
opal as an interm ediate product (see Section 9.6).
The properties of colloidal suspensions arise from
electrical charges on the surface of the p ar ticles. The
presence of electrical charges on col loidal particles is
related to th eir small size and resultant large surface area.
The atom s on the sur faces of solid particles of all kinds of
m aterials are incom pletely bonded because they are not
sur rounded by ions of the opposite charge. Therefore, such
particles attract ions from the solution in which they are
suspended and thereby acquire a characteristic charge. W
hen clay m iner als are suspended in w ater, som e of the
adsorbed interlayer ions m ay be lost, leaving the
individual particles with a negatively charged surface. In
addition, clay particles also have surface ions with
unsatisfied charges and hence becom e charged not only
because of internal crystallographic rea sons but also
because of the surface effects that
occur in all small particles.
The surfaces of silicate, oxide, or hydroxide
particles expose atom s of Si, Al, Fe, M n, etc.,
which are bonded to O 2 or O H ' radicals within
the solid. H ow ever, O H radicals on the surface
exert an attractive force on positively charged ions in the
solution, including H +. Therefore, such particles acquire a
positive charge in strongly acidic solutions by virtue of the
reaction:
A l O H (_) + H + -> A l O H (2+) (13.36)
w here the parentheses indicate n e t positive or negative
charges th at are not necessarily equal to one electronic
unit. In strongly basic solutions the O H - radicals m ay
release H +, thereby causing a n et negative charge at the
surface:
A l O H (" } - A l 0 (_) + H + (13.37)
Similar reactions m ay take place with atom s of Si, Fe,
Mn, etc., on the surface of colloidal particles th at
contribute to their total electrical charges.
The presence of electrical charges on the sur face of
colloidal particles overcom es their te n dency to settle out
and keeps them in suspension indefinitely unless the
charges are neutralized by the addition of an electrolyte or
the passage of an electrical current. In general, colloidal
suspen sions are stabilized in dilute electrolyte solutions
because ions from the solution are needed to build up the
surface charge. H ow ever, at high electrolyte
concentrations the surface charges of the particles are
neutralized and the suspension coagulates, th at is, the
particles settle out.
The polarity of the surface charge of colloidal
particles depends on the activity of H + in the water.
Consequently, we define the isoelectric p oint as the pH of
the aqueous m edium at which the surface charge of
particles of a specific m ater ial is equal to zero. If the pH
is less than the iso electric point, the charge is positive by
virtue of equation 13.36; w hereas if the pH is greater than
the isoelectric points the charge is negative according to
equation 13.37. The isoelectric points of different kinds of
m aterials in Table 13.8 range from pH = 1.0 (am orphous
silica) to pH = 9.1 (A12 0 3, corundum ). The isoelectric
points can be used to deduce the polarity of the surface
charge of colloidal particles of each m aterial in acid and
basic media. The data in Table 13.8 indicate that colloidal
particles com posed of am orphous silica, quartz, M n-
dioxide, m ontm orillonite, and kaolin ite have a negative
surface charge, except in high ly acidic environm ents
caused, for example, by local discharge of acid m ine
waters. The oxides of Fe form particles whose surface
charges may change from positive to negative betw een pH
val ues 5 and 7. H ow ever, Fe-oxide or hydroxide and
TiOz particles in the oceans (pH = 8.2) have neg ative
charges. The oxides and hydroxides of Al stand out from
the others in Table 13.8 because they form positively
charged particles in all but highly basic environm ents.
It turns out th at the presence of electrolytes in a
colloidal suspension affects the developm ent
 13.5 CO LLO IDAL SUSPENSIONS AND ION EXCHANGE 219
Table 13 .8 Isoelectric
Points of Various Natural Materials When
Suspended in Water of Varying Acidity but Lacking Dissolved
Electrolytes
Material Polarity in Polarity in
pH acid medium basic medium
S i0 2(amorph.) 1.0-2.5 _a -

Quartz 2.0 - -
M n02 (different forms) 2.0 -4.5 - -
Albite 2.0 - -
Montmorillonite 2.5 - -
Kaolinite 4.6 - -
Hematite 5 -9 + -t-

commonly 6-7
Magnetite 6.5 + -
Goethite 6 -7 + +
Limonite (Fe20 3 H20 ) 6 -9
Anatase (T i02) 7.2 + -
Gibbsite ~9 + +
Corundum 9.1 + +
Periclase (MgO) 12.4 + +

The polarity changes from negative to positive at pH values less than the isoelectric point. Thus,
the charge is negative if the pH is greater than the isoelectric point. A cid and basic are defined
relative to pH = 7.0 at 25 C.
After Drever (1982),Table 4.2;Stumm and Morgan (1970).
of surface charges of the
particles. Therefore, the p
H at which surface charges
of colloidal particles
becom e equal to zero in
the presence of elec
trolytes differs from the
isoelectic point in ways th
at depend on the com
position of the particles
and the charges of the ions
of the electrolyte. The pH
at which the surface
charge of colloidal parti
cles in the presence of
electrolytes is equal to
zero is know n as the zero
po in t o f charge (Z P C ).
The num erical values of
the Z P C depend not only
on the com position of the
colloidal particles but also
on the com position of the
electrolytes in the sus
pension.
The charges of
colloidal particles of
different com positions
affect geochem ical
processes in the oceans, in
soils, and in the
deposition of sedim ent
because negatively
charged particles, of
colloidal
size or larger, adsorb
cations (m etals), w
hereas positively
charged particles
adsorb anions (non-
m etals). Particles of all
oceans scavenge m etal
transfer them to the sedim
ent accum ulating at the
bottom of the oceans.
Consequently, deep-sea
sedim ent is enriched in m
any m etals com pared to
norm al m arine sedim ent,
which is dom inated by
detrital sedim ent derived
from nearby continen
tal areas (Section 4.5).
Colloidal particles in
soils act as tem
porary
repositories of ions
released by chemical w
eather ing or by the
addition of fertilizers. The
resulting capacity of soils
to store plant nutrients
enhances their fertility. In
addition, the fixation of
ions in soils and sedim ent
enhances the solubility of
m inerals and results in the
purification of groundw
ater.
220 CLAY MINERALS

W hen clay m inerals or other particles of col loidal or
larger size are dispersed in w ater, their surfaces becom e
charged and attract ions of the opposite charge, which are
adsorbed by m eans of electrostatic bonds. W ith a few
exceptions, m iner al surfaces are not particularly selective
and attract ions of different elem ents w ithout m uch
discrim ination. This m eans that the adsorbed ions are
exchangeable for others in the solution depending prim
The capacity of clay m inerals to act as ion
exchangers varies widely, depending at least in part on the
charge im balances th at originate within the crystal lattice.
For example, kaolinite commonly has a low cation
exchange capacity (C E C ), w hereas m ontm orillonite
and verm iculite have large CECs. The C E C of a sam ple
of clay is determ ined by m eans of a standard procedure: 1 0
0 g of dry clay is repeatedly placed into NaCl solutions at

arily on their activity. The result ing phenom enon of ion
exchange can be rep re sented by a simple exchange
reaction:
AX+ B+ ^ B X +A+ (13.38)
which indicated that ion A + on the adsorber is exchanged
for ion B + in the solution. Such exchange reactions reach
equilibrium very rapidly and are then subject to the Law of
Mass Action; that is:
pH = 7.0 until all exchangeable cations are replaced by N
a+. The N a -saturated clay is then added to concen trated
KC1 solutions and the am ount of N a + released by the
clay at pH = 7.0 is determ ined by analysis and is
expressed in units of milliequiva-lents per 100 g of dry
clay. The C E C values of clay m inerals are related to their
chem ical com posi tions and therefore vary widely. H ow
ever, the val ues in Table 13.9 are fairly typical of the
respective clay minerals.

1BX1[A 0 =K (13 39) The C E C of a clay m ineral can be estim ated from
[A X ][B + ] " 1 the net negative charge of its structural for mula. For
where K cx is the exchange constant. If the exchang er example, the m ontm orillonite rep re sented by form ula
accepts A + and B + with equal ease, then K ex = 1.0. 13.3 has a charge of -0 .4 6 electronic units per form ula
However, in general this is not the case because A + and B weight which is neu tralized by exchangeable Ca2+,
+ do not have the same charge/radius ratio and their Mg2+, and N a+. The C E C of this m ineral caused only
hydrated ions have different radii. M oreover, divalent or by the charge in the layers is:
trivalent ions form stronger electrostatic bonds than
univalent ions and may therefore be adsorbed
preferentially.
E quation 13.39 can be restated in the form:
Table 13 .9 Cation Exchange Capacity
lA ll = ur f ( AX A n (13.40) (CEC) of Various Clay Minerals
[B +] ex\(B X )/
CEC, m eq/100 g
w here (A X ) and (BX ) are the m olar concentra tions of Mineral at pH = 7.0
A and B on the exchanger and n is an Kaolinite 3-15
exponent w ith values betw een 0 . 8 and 2 . 0
Chlorite 10-40?
(B erner, 1971; G arrels and Christ, 1965) . These
Illite 10-40
equations explain why the activities of ions in the aqueous
phase of colloidal suspension are con trolled by ion Glauconite 11- 20+
exchange. For example, if A + is added to a colloidal Palygorskite 20-30
solution containing A + and B + in ion exchange Allophane -70
equilibrium with clay m iner als in suspension, A + will Smectite (montmorillonite) 70-100
displace B + from exchange sites and thereby reestablish Vermiculite 100-150
equilibri um in accordance with Le C hteliers principle.
SOURCE: Garrels and Christ (1965).

0.46 X 100 X 1000
C EC = 36025---------- = m eq/100 g
(13.41)
This value is probably an underestim ate because it does
not include charges caused by broken bonds. Nevertheless,
the result indicates that 100 g of this clay can adsorb 64.6
mmol of Ca2+, which am ounts to 40.08 X 64.6/1000 =
2.59 g of Ca or 2.97 g of Na. These am ounts are not only
substantial, but the presence of N a+, Ca2+, or other
cations has a signifi-gant effect on the mechanical
properties of the clay.
The presence of N a+ in exchangeable positions
increases the plasticity of clay and can m ake soil
containing such clay unsuitable for agriculture. The best
soils contain clay m inerals having a mixture of Ca2+, H
+, M g2+, and K + as exchangeable cations. However, in
the ceramic industry, where plasticity is a desirable
property, clay may be converted to the Na-form by suitable
treatm ent prior to its use.
Sm ectite clays expand w hen w ater m olecules
replace exchangeable ions, and slurries of sm ec tite may
set to form a solid that can revert to a liq uid w hen it is
disturbed. This property, called thixotropy, characterizes
bentonite, which is com posed largely of sm ectite clay
minerals, and m akes it useful as a sealant of building
foundations and as a drilling m ud in the petroleum
industry.
13.6 Dating of Clay Minerals
O ne of the prim ary tasks in the E arth Sciences is to
establish an accurate chronology of the history of
the E arth, which is recorded in the sedim enta ry
rocks th at have survived to the Present. This has
turned out to be a very difficult task because
sedim entary rocks are generally not datable by
m ethods based on radioactivity. One of several
reasons for this difficulty is that sedim entary
rocks are com posed prim arily of detrital particles
of m inerals th at crystallized to form igneous or
m etam orphic rocks long before they w ere incor
porated into clastic sedim entary rocks, such as
sandstones or shales. O ne can m ake a virtue of
13.6 DATING OF CLAY MINERALS 221
this circum stance by dating detrital m inerals such as
muscovite, feldspar, or zircon to determ ine their p ro ven
a n ce by identifying their source by its age (Faure and
Taylor, 1981). However, the age of sedim entary rocks can
be determ ined only by d at ing m inerals th at form ed at
the tim e of deposition (C lauer and C haudhuri, 1995).
U nfortunately, m ost of the authigenic m iner als in
sedim entary rocks are not suitable for d at ing by isotopic
m ethods because they do not contain radioactive p aren t
elem ents in sufficient concentration or because they do
not retain the radiogenic daughter isotope quantitatively.
These shortcom ings effectively elim inate several authin-
genic m inerals th at com m only occur in sedim en tary
rocks, including calcite, dolom ite, quartz, gypsum,
anhydrite, hem atite, m agnetite, pyrite, pyrolusite, halite,
sylvite, and other evaporite m in erals. However, certain
clay m inerals are authi genic and do contain K and Rb,
which m ake them potentially datable by the K - A r and R
b -S r m eth ods described in C hapter 16.
G lauconite and illite are best suited for d at ing of
sedim entary rocks, but only under narrow ly defined
circum stances. For exam ple, glaucony
in the form of greenish pellets and encrustations
in m arine sedim entary rocks contains varying
am ounts of K, which stabilizes the lattice by
strengthening the bonds betw een clay units. A s a result,
the retentivity of glaucony with respect to radiogenic 40A
r im proves with increasing K con
ten t and the m ineral becom es closed to 40A r
w hen KzO > 6.0% (O din and M atter, 1981).
Therefore, only glauconitic micas w ith the
required K concentration can yield reliable dates by the K
-A r m ethod. The selection criteria for dating glaucony by
the R b -S r m ethod are not as well defined, partly because
radiogenic 87Sr is less
m obile than 4 0 A r. H ow ever, both m ethods are
also sensitive to gain or loss of the radioactive
p arent elem ent so that w ell-crystallized glau
conitic m ica is also preferable for dating by the
R b -S r m ethod. H ow ever, even well-crystallized
glauconies may be appreciably altered by th e r
m al m etam orphism , tectonic deform ation, and
222 CLAY MINERALS
ion exchange w ith brines in the subsurface.
in the tetrahedral layer Si4+ is partially replaced by A l3+.
T herefore, glauconies th at rem em ber their age The charge im balance th at results from this substitution in
are valuable exceptions rath er than the rule the tetrahedral layer may be partly com pensated by
(O din, 1982). additional cations in the octahedral layer. How ever, the net
The application of the R b -S r m ethod to negative charge arising from either the tetrahedral or th e
dating clay-rich sedim entary rocks has been octahedral layer, or both layers, is neutralized by cations
sum m arized by C lauer (1982). The w hole-rock that are adsorbed to the outer surfaces of the terahedral
isochron m ethod som etim es yields useful results (C layer and thereby bond individual clay units to each other.
ordani et al., 1985), especially in unfossilifer-ous shales of Potassium ions are espe cially effective in this regard
Precam brian age containing large am ounts of accum because they fit into the hexagonal rings of silica
ulated radiogenic 8 7 Sr. In gen tetrahedra th at m ake up the tetrahedral layer.
eral, how ever, it is b e tte r to use m ethods that Consequently, K -bearing clays like illite have short
rely on the properties of a specific authigenic interlayer spacings and do not expand in water. Clays that
m ineral, such as illite. M orton (1985) dem on have adsorbed N a+, Ca2+, or M g2+ ions have wider
strated th at coarse-grained illite (1 - 2 //.m) in interlayer spacings than K-clays and expand in water
U p p er D evonian black shales of Texas is com because the inter layer cations are displaced by water
posed of the 2M polym orph of detrital origin molecules.
and yielded a provenance d a te by th e R b -S r
m ethod th at p red ates the tim e of deposition . The chemical diversity of clay m inerals m akes it
H ow ever, lM d illite in the <0.2-/u,m fraction in difficult to ascertain their therm odynam ic prop erties.
this case dated the tim e of diagenetic alteration Several em pirical schemes have been p ro posed to
of the illite by m igratingbrines. The fit of data overcom e this difficulty, including one by
points to a R b -S r isochron was dram atically Tardy and G arrels (1974). Their m ethod for esti
im proved by treating fine-grained lM d illites m ating standard free energies is based on the
w ith am m onium acetate, which rem oved R b + oxides and hydroxides that m ake up the layers of two- or
and S r2+ from exchangeable sites. three-layer clays and on the adsorbed ions, including H +.
However, in order to construct m in eral stability diagram s
of general validity, stan dardized smectite-type clay m
inerals are used. The results indicate that prim ary mica m
13.7 Summary inerals (m us covite, biotite, and phlogopite) alter to a
variety of clay m inerals depending on environm ental
Clay m inerals are classified according to their crystal condi tions. M ineral stability diagram s can be used to
structure into platy clays, fibrous clays, and x-ray am understand natural geochemical processes such as the form
orphous clays. The platy clays are com posed of sheets of ation of greenalite in cherty iron form a tions and the
silica tetrahedra joined at their bases (tetrahedral layer, T) incompatibility of greenalite and siderite.
and Al hydroxide octahedra joined by sharing hydroxyl
ions at the corners (octahedral layer, O ). The phyllosilicate
clays are m ade up of tw o-layer (T -O ) and three-layer (T Clay m inerals and other small particles can form
- O - T ) packages. The chemical diversity of clay m colloidal suspensions in dilute electrolyte solutions
inerals results from com plete or partial replacem ent of A because of electrical charges on their surfaces. Such
l3+ in the octahedral layer and of Si4+ in the tetrah ed ral suspensions may be stable for long periods of tim e but are
layer. coagulated by an increase in the concentration of
electrolytes. The polarity of the electrical charges of
In the octahedral layer, A l3+ can be replaced by colloidal particles of different com positions depends on
Fe3+, Fe2+, M g2+, Z n2+, Li+, and C r3+, w hereas the pH of the

solution. Clay m inerals, quartz, feldspar, and M n oxides
are negatively charged except in strongly acidic environm
ents with pH < 2.0. The various oxides and hydroxides of
Fe can have positive or negative surface charges in the pH
range from 5 to 9, w hereas A l oxides and hydroxide
particles have positive charges except at pH > 9.
C ations th at are adsorbed on the charged surfaces of
small particles in contact with dilute electrolyte solutions
are exchangeable for ions in the solution. These cation
exchange reactions can achieve a state of equilibrium th a t
causes w ater-particle m ixtures to respond to changes in
w ater com positions in accordance w ith Le
C h teliers principle. The ability of sm all p a rti
cles of different m inerals to adsorb exchange
able ions is expressed by its cation exchange
PROBLEMS 223
capacity (C E C ), which can be m easured by m eans of a
standardized procedure . T he physical properties of clays
are strongly affected by the kinds of adsorbed cations.
C ertain authigenic K -bearing clay m inerals are
dateable by the K -A r and R b -S r m ethods and therefore
provide rare opportunities to determ ine the ages of sedim
entary rocks. How ever, only clay samples whose chemical
com position, crystallinity, or grain size has been evaluated
should be used, and the resulting dates may be affected by
therm al m etam orphism , structural deform ation, or ion
exchange with circulating brines. D ating of sedi m entary
rocks by m eans of clay m inerals requires a thorough
understanding of clay m ineralogy as well as of
geochemistry.
Problem
s
 1. Determine the
Component %
electrical charges of the
S i0 2 45.20
octahedral and tetrahedral
layers and check whether a i2o 3 37.02
the exchange able ions Fe20 3 0.27
balance the charge in the
FeO 0.06
following clay mineral.
MgO 0.47
[ (Fe j 3 4 5 Mg0 595Mn CaO 0.52
0.004)
aH20 is water removed
(Si3.g36^0.112^e0.05l) at 105 C, whereas HzO +
lo(H)2]Ko- is lost only at high
779Ca0.076 temperature.

2. A glauconite from 4. Recalculate the

Point Jackson, Francosia, chemical analysis of a clay
Wisconsin, has the miner al from
Montmorillon, France, into
following chemical
a structural formu la for a
composition (Johnston and three-layer clay (Grim,
Cardile, 1987). 1968).

Element Moles Component %

Si 3.611 S i0 2 51.14

Al 1.238 a i2o 3 19.76

Fe2+ 1.097 Fe20 3 0.83
Mg 0.442 FeO
Ti 0.003 MgO 3.22
CaO 1.62
Rewrite this analysis in the
form of a structural
formula for a three-layer
clay and test it for
electrical neutrality.
3. Recalculate the
chemical analysis of an
impure clay mineral from
Macon, Georgia, into a
structural formula for a two-
layer clay (Grim, 1968).
224 CLAY MINERALS
 5. Recalculate the chemical analysis of a chlorite mineral 7. Estimate the standard free energy of formation of
from Schmiedefeld,Thuringia, into a structural formula pure illite by the method of Tardy and Garrels (1974)
(Grim, 1968). based on the formula
Component % Com ponent % [(Ali.5oMg0 .34Fe3 ; 7 Fe2 j4)
(Si3 43A l() 57) 0 i0(OH)2]K0 80
SiOz 20.82 MgO 4.15

a i 2o 3 17.64 H20+ 10.30 8. Calculate the solubility product constant for hec-
Fe20 3 8.70 Sum 99.57 torite at 25 C, assuming congruent solution based on
FeO 37.96 the formula
6. Estimate the standard free energy of formation of [(Mg2.6? Li 0.33) (Si 4) 0 10(OH)2]Na0 33

average montmorillonite by the method of Tardy and

Garrels (1974) based on the formula

[(^1.50^go.34Fe0.17pe0.04)
(S*3.83AIO.I7)OIO(OH)2]K(M0

References

B E R N E R , R. A., 1971. Principles
o f Chemical Sedimentology.
McGraw-Hill, New York,
240 pp.
B R A D L E Y , W. F., 1940. The
structural scheme of
attapulgite.
Amer. Mineral., 25:405^110.
B R IN D L E Y , G. S., 1961. Kaolin,
serpentine, and kindred
miner als. In G. Brown
(Ed.), The X-ray
Identification and Crystal
Structures o f Clay
Minerals, 2nd ed., 51-131.
Mineralogy Society
London, Monograph.
B R IN D L E Y , G. S., S. W. B A I L E Y ,
G. T. F A U S T , S. A. F O R M A N , and
C. I. R I C H , 1968. Report of
the Nomenclature
Committee (1966-67) of the
Clay Minerals Society.
Clays and Clay Minerals,
16:322-324.
B R O W N , B. E ., a n d S. W. B A I L E Y Waters.
p a rt 1, re g u la r a n d se m i- ra n d oPrentice-Hall,
Amer. Mineral., 4 7 :8 19 -850 .
B u r s t , J. F., 1958. G la u c o n ite
p e lle ts, th e ir m in e ra l n a tu
re an d a p p lic a tio n to S tratig
ra p h ie in te rp re ta tio n s .
Amer. Assoc.
Petrol. Geologists Bull., 4 2 :
3 10 -327 .
C A IR N S -S M I T H , A . G ., and H .
Minerals and the Origin o f
Press, Cambridge, England, 202 pp.
C A R R O L L , D., 1970. Clay
minerals: a guide to their
x-ray iden tification. Geol.
Soc. Amer., Special Paper
126,80 pp.
C h e n , C.-H., 1975. A method
of estimating standard free
ener gies of formation of
silicate minerals at 298.15
K.
Amer. J. Sei., 2 7 5 :801 -817 .
C L A U E R , N., 1982. The
rubidium-strontium
method applied to
sediments: certitudes and
uncertainties. In G. S.
Odin (Ed.),
G R I M , R. E., 1968. Clay
Minerology, 2nd ed. McGraw-
Numerical Dating in Hill,
Statigraphy, vol 1, 245-276. New York, 596 pp.
Wiley,
Chichester, England, 630 pp.
C L A U E R , N., and S. C H A U D H U R I
Environments; Isotopic Dating
Verlag, Heidelberg, 359 pp.
C O R B A T O . C. E and R. T. T E T T
E N H O R S T , 1987. Analysis of
illite-smectite
interstratification. Clay
Miner., 22:269-285.
C O R D A N I , U. G., A. T H O M A Z -F IL
H O , B. B. B R I T O -N E V E S , and
K. K A W A S H IT A , 1985. On
the applicability of the Rb-
Sr
method to argillaceous
sedimentary rocks: some
examples from
Precambrian sequences of
Brazil. Giornale di
Geologia, ser. 3a., 47:253-
280.
D R E V E R , J. I., 1982. The
Geochemistry o f Natural
m Upper
Saddle River, NJ, 388 pp.
E G G L E S T O N , R. A., and S. W. B
A I L E Y , 1967. Structural
aspects of dioctahedral
chlorite. Amer. Mineral.,
52:673-689.
FA U R E , G., and K. S. T A Y L O R ,
1981. Provenance of some
glacial deposits in the
Transantarctic Mountains
based on Rb-Sr dating of
feldspars. Chem. Geol.,
32:271-290.
F O S T E R , M. D ., 1962.
Interpretation of the
composition of
trioctahedral micas. U. S.
Geol. Surv. Prof. Paper
354-B:ll-49.
G A R R E L S , R. M., and C. L. C H
R I S T , 1965. Solutions,
Minerals, and Equilibria,
Harper & Row, New York,
450 pp.
G O O D W IN , A. M., 1956. Facies
relations in the Gunflint Iron
Formation. Econ. Geol.,
51:588-595.
 REFERENCES 225
G R I M , R. E., R. H. B R A Y , and W. F. B R A D L E Y , 1937. The mica in Occurrences, Genesis and Uses. Elsevier, Amsterdam, 352 pp.
argillaceous sediments. Amer. Mineral., 22:813-827. SM I T H , J. V., and H. S. Y O D E R , 1956. Studies of mica poly
G R N E R , J. W., 1934. The structure of vermiculites and their morphs. Mineral. Mag., 31:209-235.
collapse by dehydration. Amer. Mineral., 19:557-575. ST U M M , W., and J. J. M O R G A N , 1970. Aquatic Chemistry. Wiley,
H E L G E S O N , H.C., 1969.Thermodynamics of hydrothermal sys tems New Y o rk , 583 pp.
at elevated temperatures and pressures. Amer. J. Sei., S U D O , T., and S. S H IM O D A (Eds.), 1978. Clays and Clay
267:729-804. Minerals in Japan. Elsevier, Amsterdam.
H E L G E S O N , H. C ., J. M. D E L A N Y , H. W., N E S B IT T , and D. K. SU D O , T., S. S H IM O D A , H. Y O T S U M O T O , and S. A IT A , 1981.
B I R D , 1978. Summary and critique of the thermodynamic Electron Micrographs o f Clay Minerals. Elsevier,
properties of rock-forming minerals. Amer. J. Sei., Amsterdam, 204 pp.
278A:l-229. T A R D Y , Y., and R. M. G A R R E L S , 1974. A method of estimating the

H ey, M. H., 1954. New review of the chlorites. Mineral. Mag.,
30:277-292.
H O W E R , J., 1961. Some factors concerning the nature and ori
gin of glauconite. Amer. Mineral., 46:313-334.
James, H. L., 1966. Chemistry of the iron-rich sedimentary rocks.
In M. L. Fleischer (Ed.), Data of geochemistry. U. S.
Geol. Surv. Prof. Paper, 440-W; 60 pp.
J O H N S T O N , J. H ., and C. M. C A R D I L E , 1987. Iron substitution in
montmorillonite, illite, and glauconite by 57Fe Mssbauer
spectroscopy. Clays and Clay Minerals, 35:170-176.
K I T T R I C K , J. A., 1971a. Stability of montmorillonites: I. Belle
Fourche and Clay Spur montmorillonites. Soil Sei. Soc.
Amer. Proc., 35:140-145.
K I T T R I C K , J. A., 1971b. Stability of montmorillonites: II.
Aberdeen montmorillonite. Soil Sei. Soc. Amer. Proc.,
35:820 -823 .
L IN D S A Y , W. L ., 1979. Chemical Equilibira in Soils. Wiley, New
Y o rk , 449 pp .
M C R A E , S. G., 1972. Glauconite. Earth Sei. Rev., 8(4):397^140.
M ILLOT , G., 1970. Geology o f Clays. Springer-Verlag, Berlin, 429
pp.
M O R T O N , J. P., 1985. Rb-Sr dating of diagenesis and source age of
clays in Upper Devonian black shales of Texas. Geol. Soc. Amer.
Bull., 96:1043-1049.
N R I A G U , J. O., 1975. Thermochemical approximations for clay
minerals. Amer. Mineral., 60:834-39.
O D I N , G . S. (E d .), 1982. Numerical Dating in Stratigraphy, vols.
1 and 2. Wiley, Chichester, England, 1040 pp.
O D T N , G . S., a n d A . M A T T E R , 1981. D e g la u c o n ia ru m o rig in e .
Gibbs energies of formation of layer silicates. Geochim.
Cosmochim. Acta, 38:1101-1116.
T A R D Y , Y., and R. M. G A R R E L S , 1976. Prediction of Gibbs energies
of formation I. Relationships among Gibbs energies of
formation of hydroxides, oxides and aqueous ions. Geochim.
Cosmochim. Acta, 40:1051-1056.
T A R D Y , Y., and R. M. G A R R E L S , 1977. Prediction of Gibbs energies
of formation of compounds from the elements
II. Monovalent and divalent metal silicates. Geochim.
Cosmochim. Acta, 41:87-92.
T A R D Y , Y., and L . G A R T N E R , 1977. Relationships among Gibbs
energies of formation of sulfates, nitrates, carbonates, oxide,
and aqueous ions. Contrib. Mineral. Petrol, 63:89-102.
T A R D Y , Y., and P. V I E I L L A R D , 1977. Relationships among Gibbs
free energies and enthalpies of formation of phosphates,
oxides and aqueous ions. Contrib. Mineral. Petrol.,
63:75-88.
T A R D Y , Y., 1979. Relationships among Gibbs energies of for
mation of compounds. Amer. J. Sei., 279:217-224.
T Y L E R , S. A., 1949. Development of Lake Superior soft iron ores
from metamorphosed iron formation. Geol. Soc.
Amer. Bull, 60:1101-1124.
V E L D E , B., 1977. Clays and Clay Minerals in Natural and
Synthetic Systems. Elsevier, Amsterdam, 218 pp.
W E A V E R , C. E., 1956. The distribution and identification of
mixed-layer clays in sedimentary rocks. Amer. Mineral.,
41:202-221.
W EAVE R , R . M ., M . L . JA C K S O N , a n d J. K . SY E R S , 1971.
M ag n e siu m a n d silico n activ itie s in m a trix s o lu tio n s o f

Sedimentology, 28:611-641. m o n tm o rillo n ite - c o n ta in in g so ils in re la tio n to clay m in e ra l

R EYNOLDS , R . C ., 1967. Interstratified clay systems calcula tion
of the total one-dimensional diffraction function.
Amer. Mineral., 52:661-672.
R E Y N O L D S , R . C ., 1980. Interstratified clay minerals. In G. W.
stab ility . Soil Sei. Soc. Amer. Proc., 35:823-830.
W E A V E R , C. E., and L. D. P O L L A R D , 1973. The Chemistry o f
Clay Minerals. Elsevier, Amsterdam.
Y , H. S., and H. P. E U G S T E R , 1955. Synthetic and natural
O D E R

Brindley and G. Brown (Eds.), Crystal Structures o f Clay muscovites. Geochim. Cosmochim. Acta, 8:225-280.
Minerals and Their X-ray Identification. Mineral. Soc. London, Zen, E -A n, 1967. Mixed-layer minerals as one-dimensional
Monograph 5,495 pp. crystals. Amer. Mineral., 52:635-660.
R O U T S O N , R. C., and J. A. K I T T R I C K , 1971. Illite solubility. Soil Zen, E -A n, 1972. Gibbs free energy, enthalpy and entropy of ten
Sei. Soc. Amer. Proc., 35:714718. rock-forming minerals: calculations, discrepancies,
S IN G E R , A., and E. G A L A N (Eds.), 1984. Palygorskite-Sepiolite. implications. Amer. Mineral., 57:524-553.