Alan D. Wilson, John W. Nicholson-Acid-Base Cements_ Their Biomedical and Industrial Applications (Chemistry of Solid State Materials) (2005).pdf

Although acid-base cements have been known since the mid 19th
century, and have a wide variety of applications, there has been a failure
to recognize them as constituting a single, well-defined class of material.
This book remedies the situation by unifying the subject and treating this
range of materials as a single class.
These cements are defined as materials that are formed by mixing a
basic powder with an acidic liquid, and offer an alternative to polymer-
ization as a method for forming solid substances. They are quick-setting
materials, with unusual properties, which find diverse applications as
biomaterials and in industry.

Chemistry of Solid State Materials
Acid-base cements
Their biomedical and industrial applications

Chemistry of Solid State Materials
Series Editors
A. R. West, Department of Chemistry, University of Aberdeen
H. Baxter, formerly at the Laboratory of the Government Chemist,
London
1 Segal: Chemical synthesis of advanced ceramic materials
2 Colomban: Proton conductors
3 Wilson & Nicholson: Acid-base cements

Acid-base cements
Their biomedical and industrial
applications
Alan D. Wilson
formerly Head, Materials Technology, Laboratory of the Government Chemist
Senior Research Fellow, Eastman Dental Hospital
John W. Nicholson
Head, Materials Research, Laboratory of the Government Chemist
m
0
CAMBRIDGE
UNIVERSITY PRESS

CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo
Cambridge University Press
The Edinburgh Building, Cambridge CB2 2RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521372220
© Cambridge University Press 1993
This book is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 1993
This digitally printed first paperback version 2005
A catalogue recordfor this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Wilson, Alan D.
Acid—base cements: their biomedical and industrial applications /Alan D.
Wilson, John W. Nicholson
p. cm. - (Chemistry of solid state materials; 3)
Includes bibliographical references and index.
ISBN 0-521-37222-4
1. Adhesives. 2. Dental cements. I. Nicholson, John W.
II. Title. III. Series.
TP968.W54 1993
620.1'35-dc20 91-38946 CIP
ISBN-13 978-0-521-37222-0 hardback
ISBN-10 0-521-37222-4 hardback
ISBN-13 978-0-521-67549-9 paperback
ISBN-10 0-521-67549-9 paperback

Dedicated to the past and present members of the
Materials Technology Group at the Laboratory of the
Government Chemist

Contents
Preface xvii
Acknowledgements xix
1 Introduction 1
References 4
2 Theory of acid-base cements 5
2.1 General 5
2.2 The formation of cements 7
2.2.1 Classification 7
2.2.2 Requirements for cementitious bonding 8
2.2.3 Gelation 10
2.3 Acid-base concepts 12
2.3.1 General 12
2.3.2 History of acid-base concepts 12
2.3.3 Acid-base concepts in AB cement chemistry 14
2.3.4 Relevance of acid-base theories to AB cements 19
2.3.5 Acid-base strength 20
2.3.6 Acid-base classification 22
2.3.7 Hard and soft acids and bases (HSAB) 24
References 26
3 Water and acid-base cements 30
3.1 Introduction 30
3.1.1 Water as a solvent 30
3.1.2 Water as a component 30
3.2 Water 31
3.2.1 Constitution 31
3.2.2 Water compared with other hydrides 33
3.3 The structure of water 34
3.3.1 The concept of structure in the liquid state 34
3.3.2 The structures of ice 35
3.3.3 Liquid water 36
3.4 Water as a solvent 40
IX

Contents
3.4.1 Hydrophobic interactions 40
3.4.2 Dissolution of salts 41
3.4.3 Ion-ion interactions in water 44
3.4.4 Dissolution of polymers 45
3.5 Hydration in the solid state 47
3.5.1 Coordination of water to ions 47
3.6 The role of water in acid-base cements 48
3.6.1 Water as a solvent in AB cements 48
3.6.2 Water as a component of AB cements 48
3.6.3 Water as plasticizer 51
References 52
4 Polyelectrolytes, ion binding and gelation 56
4.1 Polyelectrolytes 56
4.1.1 General 56
4.1.2 Polyion conformation 58
4.2 Ion binding 59
4.2.1 Counterion binding 59
4.2.2 The distribution of counterions 59
4.2.3 Counterion condensation 63
4.2.4 Effect of valence and size on counterion binding 65
4.2.5 Site binding - general considerations 67
4.2.6 Effect of complex formation 69
4.2.7 Effect of the polymer characteristics on ion binding 70
4.2.8 Solvation (hydration) effects 72
4.2.9 Hydration of the polyion 73
4.2.10 Hydration and ion binding 76
4.2.11 Desolvation and precipitation 77
4.2.12 Conformational changes in polyions 79
4.2.13 Interactions between polyions 82
4.2.14 Polyion extensions, interactions and precipitation 82
4.3 Gelation 83
References 85
5 Polyalkenoate cements 90
5.1 Introduction 90
5.2 Adhesion 92
5.2.1 New attitudes 92
5.2.2 The need for adhesive materials 92
5.2.3 Acid-etching 93
5.2.4 Obstacles to adhesion 93
5.2.5 The nature of the adhesion of polyalkenoates to tooth
material 94
5.3 Preparation of poly(alkenoic acid)s 97
5.4 Setting reactions 98

Contents
5.5 Molecular structures
5.6 Metal oxide polyelectrolyte cements
5.7 Zinc polycarboxylate cement
5.7.1 Historical
5.7.2 Composition
5.7.3 Setting and structure
5.7.4 Properties
5.7.5 Modified materials
5.7.6 Conclusions
5.8 Mineral ionomer cements
5.9 Glass polyalkenoate (glass-ionomer) cement
5.9.1 Introduction
5.9.2 Glasses
5.9.3 Poly(alkenoic acid)s
5.9.4 Reaction-controlling additives
5.9.5 Setting
5.9.6 Structure
5.9.7 General characteristics
5.9.8 Physical properties
5.9.9 Adhesion
5.9.10 Erosion, ion release and water absorption
5.9.11 Biocompatibility
5.9.12 Modified and improved materials
5.9.13 Applications
5.10 Resin glass polyalkenoate cements
5.10.1 General
5.10.2 Class I hybrids
5.10.3 Class II hybrids
5.10.4 Properties
References
Phosphate bonded cements
6.1 General
6.1.1 Orthophosphoric acid solutions
6.1.2 Cations in phosphoric acid solutions
6.1.3 Reactions between oxides and phosphoric acid
solutions
6.1.4 Effect of cations in phosphoric acid solutions
6.1.5 Important cement-formers
6.2 Zinc phosphate cement
6.2.1 General
6.2.2 History
6.2.3 Composition
6.2.4 Cement-forming reaction
6.2.5 Structure
99
101
103
103
103
104
106
112
113
113
116
116
117
131
133
134
143
146
147
152
156
159
162
166
169
169
170
171
173
175
197
197
197
198
201
203
204
204
204
204
205
207
212
XI

Contents
6.2.6 Properties 214
6.2.7 Factors affecting properties 218
6.2.8 Biological effects 219
6.2.9 Modified zinc phosphate cements 219
6.2.10 Hydrophosphate cements 220
6.3 Transition-metal phosphate cements 220
6.4 Magnesium phosphate cements 222
6.4.1 General 222
6.4.2 Composition 222
6.4.3 Types 222
6.4.4 Cement formation and properties 223
6.4.5 Cement formation with phosphoric acid 223
6.4.6 Cement formation with ammonium dihydrogen
phosphate 223
6.4.7 Cement formation with diammonium hydrogen
phosphate 231
6.4.8 Cement formation with ammonium polyphosphate 232
6.4.9 Cement formation with aluminium acid phosphate 232
6.4.10 Cements formed from magnesium titanates 235
6.5 Dental silicate cement 235
6.5.1 Historical 235
6.5.2 Glasses 237
6.5.3 Liquid 241
6.5.4 Cement-forming reaction 243
6.5.5 Structure 249
6.5.6 Physical properties 253
6.5.7 Dissolution and ion release 255
6.5.8 Biological aspects 260
6.5.9 Conclusions 261
6.5.10 Modified materials 262
6.6 Silicophosphate cement 263
6.7 Mineral phosphate cements 265
References 265
Oxysalt bonded cements 283
7.1 Introduction 283
7.1.1 Components of oxysalt bonded cements 284
7.1.2 Setting of oxysalt bonded cements 284
7.2 Zinc oxychloride cements 285
7.2.1 History 285
7.2.2 Recent studies 286
7.3 Magnesium oxychloride cements 290
7.3.1 Uses 290
7.3.2 Calcination of oxide 290
7.3.3 Setting chemistry 291
xn

Contents
7.3.4 Kinetics of cementation 293
7.3.5 Phase relationships in the MgO-MgCl2-H2O system 294
7.3.6 Consequences for practical magnesium oxychloride
cements 295
7.3.7 Impregnation with sulphur 297
7.4 Magnesium oxysulphate cements 299
7.4.1 Setting chemistry 299
7.4.2 Phase relationships in the MgO-MgSO4-H2O system 300
7.4.3 Mechanical properties of magnesium oxysulphate
cements 302
7.5 Other oxysalt bonded cements 304
References 305
8 Miscellaneous aqueous cements 307
8.1 General 307
8.2 Miscellaneous aluminosilicate glass cements 307
8.3 Phytic acid cements 309
8.4 Poly(vinylphosphonic acid) cements 310
8.4.1 Metal oxide polyphosphonate cements 311
8.4.2 Glass polyphosphonate cements 314
8.5 Miscellaneous copper oxide and cobalt hydroxide
cements 315
References 316
9 Non-aqueous cements 318
318
320
320
321
321
322
323
331
333
334
334
335
335
336
336
336
337
337
337
339
Xlll
9.1
9.2
9.2.1
9.2.2
9.2.3
9.2.4
9.2.5
9.2.6
9.2.7
9.2.8
9.2.9
9.2.10
9.2.11
9.3
9.3.1
9.3.2
9.4
9.4.1
9.4.2
9.4.3
General
Zinc oxide eugenol (ZOE) cements
Introduction and history
Eugenol
Zinc oxide
Cement formation
Setting
Structure
Physical properties
Biological properties
Modified cements
Impression pastes
Conclusions
Improved ZOE cements
General
Reinforced cements
2-ethoxybenzoic acid eugenol (EBA) cements
General
Development
Setting and structure

Contents
9.4.4 Properties 340
9.5 EBA-methoxyhydroxybenzoate cements 342
9.5.1 EBA-vanillate and EBA-syringate cements 342
9.5.2 EBA-divanillate and polymerized vanillate cements 344
9.5.3 EBA-HV polymer cements 345
9.5.4 Conclusions 346
9.5.5 Other zinc oxide cements 347
9.6 Calcium hydroxide chelate cements 347
9.6.1 Introduction 347
9.6.2 Composition 348
9.6.3 Setting 348
9.6.4 Physical properties 350
9.6.5 Biological properties 350
9.6.6 The calcium hydroxide dimer cement 351
References 352
10 Experimental techniques for the study of acid-base
cements 359
10.1 Introduction 359
10.2 Chemical methods 360
10.2.1 Studies of cement formation 360
10.2.2 Degradative studies 361
10.3 Infrared spectroscopic analysis 361
10.3.1 Basic principles 361
10.3.2 Applications to AB cements 362
10.3.3 Fourier transform infrared spectroscopy 364
10.4 Nuclear magnetic resonance spectroscopy 364
10.5 Electrical methods 366
10.6 X-ray diffraction 367
10.7 Electron probe microanalysis 369
10.7.2 Applications to dental silicate cements 369
10.7.3 Applications to glass-ionomer cements 369
10.8 Measurement of mechanical properties 370
10.8.1 Compressive strength 371
10.8.2 Diametral compressive strength 372
10.8.3 Flexural strength 372
10.8.4 Fracture toughness 373
10.9 Setting and rheological properties 374
10.9.1 Problems of measurement 375
10.9.2 Methods of measurement 375
xiv

Contents
10.10 Erosion and leaching 378
10.10.1 Importance in dentistry 378
10.10.2 Studies of erosion 379
10.11 Optical properties 379
10.11.1 Importance in dentistry 379
10.11.2 Measurement of opacity 380
10.12 Temperature measurement 380
10.13 Other test methods 381
References 382
Index 386
xv

Preface
The senior author first became interested in acid-base cements in 1964
when he undertook to examine the deficiencies of the dental silicate cement
with a view to improving performance. At that time there was much
concern by both dental surgeon and patient at the failure of this aesthetic
material which was used to restore front teeth. Indeed, at the time, one
correspondent commenting on this problem to a newspaper remarked that
although mankind had solved the problem of nuclear energy the same
could not be said of the restoration of front teeth. At the time it was
supposed that the dental silicate cement was, as its name implied, a silicate
cement which set by the formation of silica gel. Structural studies at the
Laboratory of the Government Chemist (LGC) soon proved that this view
was incorrect and that the cement set by formation of an amorphous
aluminium phosphate salt. Thus we became aware of and intrigued by a
class of materials that set by an acid-base reaction. It appeared that there
was endless scope for the formulation of novel materials based on this
concept. And so it proved.
Over the years, from 1964 to date, a team at the LGC, with its expertise
in Materials Chemistry, has studied many of the materials described in this
book, elucidating structures, setting reactions and behaviour. This
experience has formed a strong experimental background against which
the book was written. In addition we have maintained contact with leaders
in this field throughout the world. We should mention Professor Dennis
Smith of Toronto University, who amongst his many achievements
invented the adhesive zinc polycarboxylate cement (Chapter 5); Dr G. M.
Brauer, who was for many years at the Institute for Materials Research,
National Bureau of Standards, Washington, D.C., and is the acknowl-
edged authority on cements formed by the reaction between zinc oxide
and phenolic bodies (Chapter 9); and Dr J. H. Sharp of the University of
Sheffield, who has developed magnesium phosphate cements (Chapter 6).
xvu

Preface
In particular we thank Dr J. H. Sharp for supplying original photographs
for use in the section on magnesium phosphate cements and for critically
reading the draft manuscript and making constructive suggestions. On
clinical matters we have benefited from a 20-year collaboration with
Dr J. W. McLean OBE.
Our own research at the LGC, while not confined to, has centred on,
cements formed by the reactions between acid-decomposable glasses and
various cement-forming acids (Chapters 5, 6, 8, 9). One of these materials
invented at the LGC, the glass polyalkenoate or glass-ionomer cement,
has proved of immense importance. Indeed, so successful has this material
been in general dentistry, that the Materials Technology Group earned the
Queen's Award for Technology in 1988. This material illustrates the useful
combination of properties that can be found in the acid-base cements, for
it has the aesthetic appearance of porcelain, the ability to adhere to teeth,
and also the ability to releasefluoridewith its beneficial effect of reducing
caries.
We hope that this work will encourage, stimulate and assist others
choosing to work in this interesting field.
Alan D. Wilson
John W. Nicholson
xvin

Acknowledgements
We make a particular acknowledgement to the late Dr John Longwell
CBE, Deputy Government Chemist in 1964, who encouraged the Labor-
atory to enter the field, and to the line of Government Chemists who
supported the work over the long years; the late Dr David Lewis CB, the
late Dr Harold Egan, Dr Ron Coleman CB (who became Chief Scientist
of the Department of Trade and Industry), Mr Alex Williams CB and
Dr Richard Worswick.
We note the particular contributions of Brian Kent, present Head of the
Materials Technology Group, as co-inventor of the glass polyalkenoate
cement way back in 1968, and of Dr John McLean OBE in developing
clinical applications. It was Surgeon Rear Admiral Holgate CB, OBE,
Chief Dental Officer at the Ministry of Health in 1964, who introduced Dr
McLean to the Laboratory of the Government Chemist (LGC) to initiate
a collaboration that proved so fruitful. Since then there has been constant
support from the Department of Health and its various officers and also
from the British Technology Group, particularly from G. M. Blunt and
R. A. Lane.
Most importantly we acknowledge the contribution of those who
worked at that essential place, the laboratory bench, on which everything
depends.
Our colleagues in the Materials Technology Group (formerly the Dental
Materials Group) who have worked with one or other of us since 1964 are:
R. F. Batchelor, B. G. Lewis, Mrs B. G. Scott, J. M. Paddon, G. Abel,
Dr S. Crisp, A. J. Ferner, Dr H. J. Prosser, M. A. Jennings, Mrs S. A.
Merson, M. Ambersley, D. M. Groffman, S. M. Jerome, D. R. Powis,
Mrs P. J. Brookman (nee Brant), R. P. Scott, J. C. Skinner, Dr R. G. Hill,
G. S. Sayers, Dr C. P. Warrens, Miss A. M. Jackson, Dr J. Ellis,
Miss E. A. Wasson, Miss H. M. Anstice, Dr J. H. Braybrook, Miss S. J.
Hawkins and A. D. Akinmade.
xix

Acknowledgements
In addition we have received support from members of other divisions
at the LGC: Dr R. J. Mesley, M. A. Priguer, D. Wardleworth, Dr I. K.
O'Neill, B. Stuart, R. A. Gilhooley, Dr C. P. Richards, Dr O. M. Lacy
and Dr S. L. R. Ellison.
Guest workers to the Materials Technology Group who have con-
tributed include Professor P. Hotz (Klinik fur Zahnerhaltung der Uni-
versitat, Bern), Ms T. Folleras (NIOM, Scandinavian Institute of Dental
Materials).
Workers in other Government Research Stations and the Universities
who have collaborated with us are: R. P. Miller, D. Clinton, Dr T. I.
Barry, Dr I. Seed (National Physical Laboratory); K. E. Fletcher (Build-
ings Research Station); Miss D. Poynter (Warren Spring Laboratory);
Professor L. Holliday, Dr J.H.Elliott, Dr P. R. Hornsby, Dr K. A.
Hodd, Dr A. L. Reader (Brunei University); R. Manston, Dr B. F.
Sanson, Dr W. M. Allen, P. J. Gleed (Institute for Research on Animal
Diseases); Professor Braden (London Hospital); A. C. Shorthall (Bir-
mingham University), I. M. Brook (University of Sheffield); and
R. Billington (Institute of Dental Surgery, London).
We thank Dr L. J. Pluim of the Rijksuniversiteit te Groningen for
drawing our attention to the early and neglected work of E. van Dalen on
zinc phosphate cements.
We thank Mrs Margaret Wilson for her help in checking the proofs and
the indexing.
We acknowledge the stoic forbearance of our wives in putting up with
the disturbances and neglect of domestic routines and duties occasioned
by the writing of a book.
Alan D. Wilson
John W. Nicholson
xx

1 Introduction
Acid-base (AB) cements have been known since the mid 19th century.
They are formed by the interaction of an acid and a base, a reaction which
yields a cementitious salt hydrogel (Wilson, 1978) and offers an alter-
native route to that of polymerization for the formation of macro-
molecular materials. They are quick-setting materials, some of which have
unusual properties for cements, such as adhesion and translucency.
They find diverse applications, ranging from the biomedical to the
industrial.
Despite all this there has been a failure to recognize AB cements as
constituting a single, well-defined class ofmaterial. Compared with organic
polymers, Portland cement and metal alloys, they have been neglected and,
except in specializedfields,awareness of them is minimal. In this book we
attempt to remedy the situation by unifying the subject and treating this
range of materials as a single class.
Human interest in materials stretches back into palaeolithic times when
materials taken from nature, such as wood and stone, were fashioned into
tools, weapons and other artifacts. Carving or grinding of a material is a
slow and time-consuming process so the discovery of pottery, which does
away with the need for these laborious processes, was of the greatest
significance. Here, a soft plastic body, potter's clay, is moulded into the
desired shape before being converted into a rigid substance by firing.
Pottery is but one of a group of materials which are formed by the physical
or chemical conversion of a liquid or plastic body, which can be easily
shaped by casting or moulding, into a solid substance. Other examples of
this common method of fabrication are the casting of metals and the
injection moulding of plastics.
Into this category come the water-based plasters, mortars, cements
and concretes which set at room temperature as the result of a chemical
reaction between water and a powder. Some of these have been known
1

Introduction
since antiquity. The AB cements are related to these materials except that
water is replaced by an acidic liquid.
ThefirstAB cement, the zinc oxychloride cement, was reported by Sorel
in 1855. It was prepared by mixing zinc oxide powder with a concentrated
solution of zinc chloride. Its use in dentistry was recommended by
Feichtinger in 1858 but it did not prove to be a success (Mellor, 1929).
However, other AB cements have proved to be of the utmost value to
dentistry, and their subsequent development has been closely associated
with this art (Wilson, 1978). The AB cements, developed against the
backcloth of the severe demands of dentistry, have interesting properties.
Some possess aesthetic appeal and the ability to adhere to base metals and
other reactive substrates. Most have superior properties to plasters,
mortars, and Portland cements, being quick-setting, stronger and more
resistant to erosion. These advantageous properties make them strong
candidates for other applications. In fact, one of these cements, the
magnesium oxychloride cement of Sorel (1867), is still used to surface walls
and floors on account of its marble-like appearance (Chapter 7).
In the 1870s more effective liquid cement-formers were found: ortho-
phosphoric acid and eugenol (Wilson, 1978). It was also found that an
aluminosilicate glass could replace zinc oxide, a discovery which led to the
first translucent cement. Thereafter the subject stagnated until the late
1960s when the polyelectrolyte cements were discovered by Smith (1968)
and Wilson & Kent (1971).
In recent years Sharp and his colleagues have developed the magnesium
phosphate cements - Sharp prefers the term magnesia phosphate cement
- as a material for the rapid repair of concrete runways and motorways
(Chapter 6). These applications exploit the rapid development of strength
in AB cements. This cement can also be used for flooring in refrigerated
stores where Portland cements do not set. Interestingly, this material
appears to have started life as an investment for the casting of dental alloys.
The glass polyalkenoate, a polyelectrolyte cement, of Wilson & Kent
(Chapter 5), was originally developed as a dental material but has since
found other applications. First it was used as a splint bandage material
possessing early high-strength and resistance to water. Currently, it is
being used, as a biocompatible bone cement, with a low exothermicity on
setting and the ability to adhere to bone, for the cementation of prostheses
(Jonck, Grobbelaar & Strating, 1989).
Outside thefieldof biomaterials it has been patented for use as a cement
for underwater pipelines, as a foundry sand and as a substitute for plaster

Introduction
in the slip casting of pottery. Quite often it appears as a substitute for
plaster of Paris, for it is stronger, less brittle and more resistant to water.
There are other possibilities. Its translucent nature suggests that it could be
used for the production of porcelain-like ceramics at room temperature.
Phosphate and polyelectrolyte AB cements are resistant to attack by
boiling water, steam and mild acids and this suggests that they could be
employed in technologies where these properties are important.
The ability of the polyelectrolyte-based AB cements (Chapter 5) to bond
to a variety of substrates, combined with their rapid development of
strength - they can become load-bearing within minutes of preparation -
suggests that they have applications as rapid-repair and handyman
materials.
A current area of interest is the use of AB cements as devices for the
controlled release of biologically active species (Allen et aL, 1984). AB
cements can be formulated to be degradable and to release bioactive
elements when placed in appropriate environments. These elements can be
incorporated into the cement matrix as either the cation or the anion
cement former. Special copper/cobalt phosphates/selenates have been
prepared which, when placed as boluses in the rumens of cattle and sheep,
have the ability to decompose and release the essential trace elements
copper, cobalt and selenium in a sustained fashion over many months
(Chapter 6). Although practical examples are confined to phosphate
cements, others are known which are based on a variety of anions:
polyacrylate (Chapter 5), oxychlorides and oxysulphates (Chapter 7) and
a variety of organic chelating anions (Chapter 9). The number of cements
available for this purpose is very great.
A recent development has been the incorporation of a bioactive organic
component into the AB cement during preparation. Since AB cements are
prepared at room temperature, this can be done without causing
degradation of the organic compound. In this case, the AB cement may
merely act as a carrier for the sustained release of the added bioactive
compound.
Another development has been the advent ofthe dual-cure resin cements.
These are hybrids of glass polyalkenoate cements and methacrylates that
set both by an acid-base cementation reaction and by vinyl polymerization
(which may be initiated by light-curing). In these materials, the solvent is
not water but a mixture of water and hydroxyethylmethacrylate which is
capable of taking dimethacrylates and poly(acrylic acid)-containing vinyl
groups into solution. In the absence of light these materials set slowly and

Introduction
have extended working times, but they set in seconds when illuminated
with an intense beam of visible light. These hybrids are in their infancy but
have created great interest.
From this account we are to expect diversification of these AB cements
both for biomedical and for industrial usages. There should be further
developments of the glass polyalkenoate cements both as bone substitutes
and as bioadhesives. We also expect more types of AB cements to be
formulated as devices for the sustained release of bioactive species. These
materials would have applications in agriculture, horticulture, animal
husbandry and human health care. In industrialfieldswe expect that there
will be continued interest in developing AB cements as materials for the
rapid repair of constructural concrete, as materials for the surfacing of
floors and walls, and as adhesives and lutes for cementation in aqueous
environments. The hybrid light-cured cements also appear to be a
promising new line of development which may give us entirely novel classes
of materials.
References
Allen, W. M., Sansom, B. F., Wilson, A. D., Prosser, H. J. & Groffman, D. M.
(1984). Release cements. British Patent GB 2,123,693 B.
Jonck, L. M., Grobbelaar, C. J. & Strating, H. (1989). The biocompatibility of
glass-ionomer cement in joint replacement: bulk testing. Clinical Materials, 4,
85-107.
Mellor, J. W. (1929). A Comprehensive Treatise on Inorganic and Theoretical
Chemistry, vol. IV, p. 546. London: Longman.
Sorel, S. (1855). Procede pour la formation d'un ciment tres-solide par 1'action
d'un chlorure sur l'oxyde de zinc. Comptes rendus hebdomadaires des seances
de TAcademie des sciences, 41, 784-5.
Sorel, S. (1867). On a new magnesium cement. Comptes rendus hebdomadaires
des seances de VAcademie des sciences, 65, 102—4.
Wilson, A. D. (1978). The chemistry of dental cements. Chemical Society
Review, 7, 265-96.

2 Theory of acid-base cements
2.1 General
From the chemical point of view AB cements occupy a place in the vast
range of acid-base phenomena which occur throughout both inorganic
and organic chemistry. Like Portland cement they are prepared by mixing
a powder with a liquid. However, this liquid is not water but an acid, while
the powder, a metal oxide or silicate, is a base. Not surprisingly, the
cement-forming reaction between them is extremely rapid and a hardened
mass is formed within minutes of mixing.
AB cements may be represented by the defining equation
Base + Acid = Salt + Water
(powder) (liquid) (cement matrix)
The product of the reaction, the binding agent, is a complex salt, and
powder in excess of that required for the reaction acts as the filler. Each
cement system is a particular combination of acid and base. The number of
potential cement systems is considerable since it is a permutation of all
possible combinations of suitable acids and bases.
Cement-forming liquids are strongly hydrogen-bonded and viscous.
According to Wilson (1968), they must (1) have sufficient acidity to
decompose the basic powder and liberate cement-forming cations, (2)
contain an acid anion which forms stable complexes with these cations and
(3) act as a medium for the reaction and (4) solvate the reaction products.
Generally, cement-forming liquids are aqueous solutions of inorganic or
organic acids. These acids include phosphoric acid, multifunctional
carboxylic acids, phenolic bodies and certain metal halides and sulphates
(Table 2.1). There are also non-aqueous cement-forming liquids which are
multidentate acids with the ability to form complexes.
Potential cement-forming bases are oxides and hydroxides of di- and

Theory of acid—base cements
Table 2.1. Examples of acids usedfor cement formation
Protonic acids Aprotic acids
(used in aqueous solution) (used in aqueous solution)
Phosphoric acid Magnesium chloride
Poly(acrylic acid) Zinc chloride
Malic acid Copper(II) chloride
Tricarballylic acid Cobalt(II) chloride
Pyruvic acid Magnesium sulphate
Tartaric acid Zinc sulphate
Mellitic acid Copper(II) sulphate
Gallic acid Cobalt(II) sulphate
Tannic acid Magnesium selenate
Zinc selenate
Protonic acids Copper(II) selenate
(liquid non-aqueous) Cobalt(II) selenate
Eugenol
2-ethoxybenzoic acid
Table 2.2. Examples of bases usedfor cementformation
Copper(II) oxide
Zinc(II) oxide
Magnesium oxide
Cobalt(II) hydroxide
Cobalt(II) carbonate
Calcium aluminosilicate glasses
Gelatinizing minerals
trivalent metals, silicate minerals and aluminosilicate glasses (Table 2.2).
All cement-forming bases must be capable of releasing cations into acid
solution. The best oxides for cement formation are amphoteric (Kingery,
1950a,b) and the most versatile cement former is zinc oxide, which can
react with a wide range of aqueous solutions of acids, both inorganic and
organic, and liquid organic chelating agents. Gelatinizing minerals, that is
minerals that are decomposed by acids, can act as cement formers, as can
the acid-decomposable aluminosilicate glasses.
In this chapter the nature of the cementitious bond and the acid-base
reaction will be discussed.

Theformation ofcements
2.2 Theformation of cements
2.2.1 Classification
Before proceeding further it is well to consider the term cement, for its
definition can be the source of some confusion. Both the Oxford English
Dictionary and Webster give two alternative definitions. One defines a
cement as a paste, prepared by mixing a powder with water, that sets to a
hard mass. In the other a cement is described as a bonding agent. These two
definitions are quite different. The first leads to a classification of cements
in terms of the setting process, while the second lays emphasis on a
property. In this book the term cement follows the sense of thefirstof these
definitions.
Cements can be classified into three broad categories:
(1) Hydraulic cements. These cements are formed from two con-
stituents one of which is water. Setting comprises a hydration and
precipitation process. Into this category fall Portland cement and
plaster of Paris.
(2) Condensation cements. Here, cement formation involves a loss of
water and the condensation of two hydroxyl groups to form a
bridging oxygen:
R-OH + HO-R = R-O-R + H2O
One example is silicate cement where orthosilicic acid, chemically
generated in solution, condenses to form a silicic acid gel. Another
is refractory cement where a cementitious product is formed by
the heat treatment of an acid orthophosphate, a process which
again involves condensation to form a polyphosphate.
(3) Acid-base cements. Cement formation involves both acid-base
and hydration reactions (Wilson, Paddon & Crisp, 1979). These
cements form the subject of this book.
This classification differs from that given by Wygant (1958), who
subdivides cements into hydraulic, precipitation and reaction cements. The
advantage of the present classification is that it clearly differentiates
phosphate cements formed by condensation from those formed by an
acid-base reaction (Kingery, 1950a). Wygant includes these in the same
category, which can be confusing. Moreover, he puts silicate cements and
the heat-treated acid phosphate cements into separate categories, although
both are condensation cements.

Theory of acid—base cements
2.2.2 Requirements for cementitious bonding
The essential property of a cementitious material is that it is cohesive.
Cohesion is characteristic of a continuous structure, which in the case of a
cement implies an isotropic three-dimensional network. Moreover, the
network bonds must be attributed to attractions on the molecular level.
Increasingly, recent research tends to show that cements are not bonded by
interlocking crystallites and that the formation of crystallites is incidental
(Steinke et al., 1988; Crisp et al., 1978). The reason is that it is difficult to
form rapidly a mass which is both cohesive and highly ordered.
Cement formation requires a continuous structure to be formed in situ
from a large number ofnuclei. Moreover, this structure must be maintained
despite changes in the character of the bonds. These criteria are, obviously,
more easily satisfied by aflexiblerandom structure than by one which is
highly-ordered and rigid. Crystallinity implies well-satisfied and rigidly-
directed chemical bonds, exact stoichiometry and a highly ordered
structure. So unless crystal growth is very slow a continuous molecular
structure cannot be formed.
In random structures, stoichiometry need not be exact and adventitious
ions can be incorporated without causing disruption. Bonds are not highly
directed, and neighbouring regions of precipitation, formed around
different nuclei, can be accommodated within the structure. Continuous
networks can be formed rapidly. Thus, random structures are conducive to
cement formation and, in fact, most AB cements are essentially amorph-
ous. Indeed, it often appears that the development of crystallinity is
detrimental to cement formation.
The matrices of AB cements are gel-like, but these gels differ from the
tobermorite gel of Portland cement. In AB cements, setting is the result of
gelation by salt formation, and the cations, which cause gelation, are
extracted from an oxide or silicate by a polyacid solution. The conversion
of the sol to a gel is rapid and the cements set in 3 to 5 minutes. Two basic
processes are involved in cement formation: the release of cations from the
oxide or silicate and their interaction with polyacid. This interaction
involves ion binding and changes in the hydration state which are
associated with gelation and structure formation (Section 4.3). Thus, there
are two reaction rates to be considered: the rate of release of cations and
the rate of structure formation. These two reaction rates must be balanced.
If the rate of release of cations is too fast a non-coherent precipitate of
crystallites is formed. If too slow the gel formed will lack strength.

Theformation of cements
During cement formation, domains are formed about numerous nuclei
and there must be bonding between the domains as well as within them. In
AB cements bonding within the domains is mainly ionic, with a degree of
covalency. The attractive forces between domains are those of a colloidal
type. In random structures, residual forcefieldsexist which act in a similar
fashion to polar forces and serve to bond domains. These forces must
include hydrogen bonds, for the addition offluorideions always enhances
cement strength and the fluoride-hydrogen bond is a strong one.
The structures of cement gels bear some relationship to the structure of
glasses. Spatially, the O2
~ ion is dominant. The matrices are based on a
coordinated polyhedron of oxygen ions about a central glass-forming
cation (Pauling, 1945). In effect, these are anionic complexes where the
cations are small, highly charged, and capable of coordinating with oxygen
or hydroxyl ions. Examples of these polyhedra are [SiOJ, [POJ and [A1OJ.
Thus, wefindthat there are silicate, phosphate and aluminosilicate glasses
and gels.
There are, however, differences which are best illustrated by reference to
the simple example of silica glass and silica gel. In silica glass, Si4+
is four-
coordinate and the polymeric links are of the bridging type:
-«>Si—O—Si<^-
In aqueous solution, coordination increases to 6, Si-OH links are possible
as well as Si-O-Si, and H2O is a possible ligand. In silica cements the
condensation of silicic acid, Si(OH)4, to SiO2 is only partial. Silica gel
therefore contains both bridging oxygen and non-bridging hydroxyl
linkages. Again, in contrast to the situation in glasses the possibility of
hydrogen bond formation will also exist.
In AB cements the gel-forming cations are frequently Zn2+
, Mg2+
, Ca2+
or Al3+
. As Kingery (1950b) has pointed out, it is the amphoteric cations,
for example Zn2+
and Al3+
, that possess the most favourable cement-
forming properties. Their oxides are capable of glass formation, not by
themselves, but in conjunction with other glass formers. Kingery also
indicated that weakly basic cations, for example Mg2+
, are less effective,
and more strongly basic cations, for example Ca2+
, even less effective.
The nature of the association between cement-forming cation and anion
is important. As we shall see from theoretical considerations of the nature
of acids and bases in section 2.3, these bonds are not completely ionic in
character. Also while cement-forming cations are predominantly a-

Theory of acid-base cements
acceptors and the anions cr-donors, both have weak ^-capabilities also.
This topic is treated in more detail in the next section. Complex formation
is clearly important and this view is supported by the anomaly that B2O3
forms cements with acids, not as a result of salt formation, but because of
complex formation (Chapters 5 and 8).
A final point needs to be made. Theory has indicated that AB cements
should be amorphous. However, a degree ofcrystallization does sometimes
occur, its extent varying from cement to cement, and this often misled early
workers in the field who used X-ray diffraction as a principal method of
study. Although this technique readily identifies crystalline phases, it
cannot by its nature detect amorphous material, which may form the bulk
of the matrix. Thus, in early work too much emphasis was given to
crystalline structures and too little to amorphous ones. As we shall see, the
formation of crystallites, far from being evidence of cement formation, is
often the reverse, complete crystallinity being associated with a non-
cementitious product of an acid-base reaction.
2.2.3 Gelation
The formation of AB cements is an example of gelation, and the matrices
may be regarded as salt-like hydrogels. They are rigid and glass-like. A gel
has been defined by Bungenberg de Jong (1949) as 'a system of solid
character, in which the colloidal particles somehow constitute a coherent
structure'. A more exact definition is not possible, for gels are easier to
recognize than define; they include a diversity of substances. Coherence of
structure appears, however, to be a universal criterion for gels.
Flory (1974) classified gels into four types on the basis of their
structures:
(1) Well-ordered lamellar structures. The lamellae are arranged in
parallel, giving rise to long-range order. Examples are soaps,
phospholipids and clays.
(2) Covalent polymeric networks which are completely disordered.
Continuity of structure is provided by an irregular three-
dimensional network of covalent links, some of which are
crosslinks. The network is uninterrupted and has an infinite
molecular weight. Examples are vulcanized rubbers, condensation
polymers, vinyl-divinyl copolymers, alkyd and phenolic resins.
10

Theformation of cements
(3) Polymer networks formed through physical aggregation; these
are predominantly disordered, but have regions of local order.
Linear structures of finite length are connected by multiple-
stranded helices, which may be crystalline. Examples are gelatin
and sodium alginate gels.
(4) Particulate, disordered structures. These include flocculent pre-
cipitates where particles generally consist of fibres in brush-heap
disarray or connected in irregular networks.
Since the matrices of AB cements bear some similarity to alginate gels
they most probably fall into type 3.
The classical theory of gelation, due to Flory (1953, 1974), sees gelation
as the result of the formation of an infinite three-dimensional network.
According to Flory, the theory can be applied without ambiguity to the
type 2 (covalent) gels and is also applicable to type 3 gels. The conditions
for the formation of such an infinite network are critical. Flory conceives
the growth of a random network as a sequential condensation process
between difunctional and multifunctional units involving a branching
process. During growth, the probability of branching (a) at each potential
branching point has to reach a critical value (ac) for an infinite network to
be formed. In the case of condensation between di- and trifunctional
groups, the probability has to be more than 50 % for an infinite network to
be formed. If it is 50 % or less then an infinite network is not formed. This
theory explains why gelation occurs suddenly.
In general, the critical value for a, ac, is given by the expression
where/is the degree of functionality of the multifunctional group.
The most investigated examples are to be found in the precipitation of
polyelectrolytes by metal ions. Here, networks are formed by the random
crosslinking of linear polymer chains, and the theory requires some
modification. The condition for the formation of an infinite network is
that, on average, there must be more than two crosslinks per chain. Thus,
the greater the length of a polymer chain the fewer crosslinks in the system
as a whole are required.
11

2.3 Acid-base concepts
2.3.1 General
The cement-forming reaction is a special case of an acid-base reaction so
that concepts of acid, base and salt are central to the topic. In AB cement
theory, we are concerned with the nature of the acid-base reaction and
how the acidity and basicity of the reactants are affected by their
constitution. Thus, it is appropriate at this stage to discuss the various
definitions and theories available.
Although acids and bases have been recognized since antiquity, our
concepts of them are still the subject of debate and development (Walden,
1929; Hall, 1940; Bell, 1947, 1973; Luder, 1948; Kolthoff, 1944; Bjerrum,
1951; Day & Selbin, 1969; Jensen, 1978; Finston & Rychtman, 1982). The
history of these concepts is a long one and can be seen as a prolonged and
continuous refinement of inexact and commonsense notions into precise
scientific theories. It has been a long and difficult journey and one that is by
no means ended.
There are various definitions of acids and bases, and in discussing them
it should be emphasized that the question is not one of validity but one of
utility. Indeed, the problem of validity does not arise because of the
fundamental nature of a definition. The problem is entirely one of choosing
a definition which is of greatest use in the study of a particular field of
acid-base chemistry. One point that needs to be borne in mind is that a
concept of acids and bases is required that is neither too general nor too
restrictive for the particular field of study.
2.3.2 History of acid-base concepts
From early times acids were recognized by their properties, such as
sourness and ability to dissolve substances, often with effervescence. The
story of Cleopatra's draught of a pearl dissolved in vinegar illustrates this
point (Pattison Muir, 1883). Vinegar, known to the Greeks and Romans,
was associated with the concept of acidity and gives its name to the term
acid which comes from the Latin acetum. Boyle (1661) observed that acids
dissolve many substances, precipitate sulphur from alkaline solution,
change blue plant dyes to red and lose these properties on contact with
alkalis.
It also has been known since antiquity that aqueous extracts of the ash
12

Acid-base concepts
of certain plants have distinctive properties: slipperiness, cleansing power
in the removal of fats, oils and dirt from fabrics, and the ability to affect
plant colours (Day & Selbin, 1969; Pattison Muir, 1883). These substances
were called alkalis, a name which comes from the Arabic for plant ash, al
halja (Finston & Rychtman, 1982) or algali. The term alkali applies only
to the hydroxides and carbonates of sodium and potassium, and it was
Rouelle in 1744 who extended the concept to include the alkaline earth
analogues and used the term base to categorize them (Walden, 1929; Day
& Selbin, 1969).
Salt formation as a criterion for an acid-base interaction has a long
history (Walden, 1929). Rudolph Glauber in 1648 stated that acids and
alkalis were opposed to each other and that salts were composed of these
two components. Otto Tachenius in 1666 considered that all salts could be
broken into an acid and an alkali. Boyle (1661) and the founder of the
phlogistic theory, Stahl, observed that when an acid reacts with an alkali
the properties of both disappear and a new substance, a salt, is produced
with a new set of properties. Rouelle in 1744 and 1754 and William Lewis
in 1746 clearly defined a salt as a substance that is formed by the union of
an acid and a base.
It can be seen that these definitions are derived from experimental
observation and are no more than classifications based on a set of
properties shared by a group of substances. They are scientifically
inadequate for the interpretation of results, which requires a definition
based on concepts. Historically, the attempt to provide a model rather than
a classification comes in the form of a search for underlying universal
principles. It seems that the alchemists recognized vague principles of
acidity and alkalinity, and in the 17th century the iatrochemists made these
the basis of chemical medicine. Disease was attributed to a predominance
of one or other of these principles (Pattison Muir, 1883).
Boyle (1661) attempted to provide a more definite concept and attributed
the sour taste of acids to sharp-edged acid particles. Lemery, another
supporter of the corpuscular theory of chemistry, had similar views and
considered that acid-base reactions were the result of the penetration of
sharp acid particles into porous bases (Walden, 1929; Finston &
Rychtman, 1982). However, the first widely accepted theory was that of
Lavoisier who in 1777 pronounced that oxygen was the universal acidifying
principle (Crosland, 1973; Walden, 1929; Day & Selbin, 1969; Finston &
Rychtman, 1982). An acid was defined as a compound of oxygen with a
non-metal.
13

After this theory was disproved, other acidifying principles were
proposed. The most significant was the recognition, by Davy & Dulong
early in the 19th century, of hydrogen as the acidifying principle (Walden,
1929; Finston & Rychtman, 1982). During this period no such search was
made for a basic principle. Bases were merely regarded as a motley
collection of antiacids with little in common apart from the ability to react
with acids.
The first substantial constitutive concept of acid and bases came only in
1887 when Arrhenius applied the theory of electrolytic dissociation to
acids and bases. An acid was defined as a substance that dissociated to
hydrogen ions and anions in water (Day & Selbin, 1969). For thefirsttime,
a base was defined in terms other than that of an antiacid and was regarded
as a substance that dissociated in water into hydroxyl ions and cations. The
reaction between an acid and a base was simply the combination of
hydrogen and hydroxyl ions to form water.
This theory was a milestone in the development of acid-base concepts:
it was the first to define acids and bases in terms other than that of a
reaction between them and the first to give quantitative descriptions.
However, the theory of Arrhenius is far more narrow than both its
predecessors and its successors and, indeed, it is the most restrictive of all
acid-base theories.
Since Arrhenius, definitions have extended the scope of what we mean
by acids and bases. These theories include the proton transfer definition of
Bronsted-Lowry (Bronsted, 1923; Lowry, 1923a,b), the solvent system
concept (Day & Selbin, 1969), the Lux-Flood theory for oxide melts, the
electron pair donor and acceptor definition of Lewis (1923, 1938) and the
broad theory of Usanovich (1939). These theories are described in more
detail below.
2.3.3 Acid—base concepts in AB cement chemistry
We now review the various concepts of acids and bases in order to see how
appropriate and useful they are in the field of AB cements.
The definition ofArrhenius
This definition of acids and bases is of restricted application. The reaction
between acids and bases is seen as the combination of hydrogen and
hydroxyl ions in aqueous solution to form water.
14

Acid-base concepts
An acid is defined as a species that dissociates in aqueous solution to give
hydrogen ions and onions, and a base as a species that dissociates in aqueous
solution to give hydroxyl ions and cations.
Thus, acids and bases are defined as aqueous solutions of substances and
not as the substances themselves. It follows that ionization is a necessary
characteristic of Arrhenius acids and bases. Another restriction of this
definition is that acid-base behaviour is not recognized in non-aqueous
solution.
The Arrhenius definition is not suitable for AB cements for several
reasons. It cannot be applied to zinc oxide eugenol cements, for these are
non-aqueous, nor to the metal oxychloride and oxysulphate cements,
where the acid component is not a protonic acid. Indeed, the theory is,
strictly speaking, not applicable at all to AB cements where the base is not
a water-soluble hydroxide but either an insoluble oxide or a silicate.
The protonic Bronsted-Lowry theory
The theory of Bronsted (1923) and Lowry (1923a, b) is of more general
applicability to AB cements. Their definition of an acid as' a substance that
gives up a proton' differs little from that of Arrhenius. However, the same
is not true of their definition of a base as' a substance capable of accepting
protons' which is far wider than that of Arrhenius, which is limited to
hydroxides yielding hydroxide ions in aqueous solution. These concepts of
Bronsted and Lowry can be defined by the simple equation (Finston &
Rychtman, 1982):
Acid = Base + H+
[2.2]
Thus, the relationship between acid and base is a reciprocal one and an
acid-base reaction involves the transfer of a proton. This concept is not
restricted to aqueous solutions and it discards Arrhenius' prerequisite of
ionization.
This concept covers most situations in the theory of AB cements.
Cements based on aqueous solutions of phosphoric acid and poly(acrylic
acid), and non-aqueous cements based on eugenol, alike fall within this
definition. However, the theory does not, unfortunately, recognize salt
formation as a criterion of an acid-base reaction, and the matrices of AB
cements are conveniently described as salts. It is also uncertain whether it
covers the metal oxide/metal halide or sulphate cements. Bare cations are
not recognized as acids in the Bronsted-Lowry theory, but hydrated
15

cations are. Thus, in the case of the group III elements, the octahedral
[M(H2O)6]3+
aquo ions are quite acidic (Cotton & Wilkinson, 1966):
[M(H2O)6]3+
= [M(H2O)5 (OH)]2+
+ H+
[2.3]
However, although both zinc and magnesium ions, the cations of the oxy-
cements, are hydrated as [M(H2O)6]2+
ions, these hydrated ions hydrolyse
only slightly (Baes & Mesmer, 1976). Thus, in magnesium chloride
solutions the aquo ions, in contrast to beryllium aquo ions, are not
perceptibly acidic. So there must be some doubt as to whether these
hydrated ions can be regarded as protonic acids. But for this, the
Bronsted-Lowry theory would almost exactly define AB cements.
Aluminosilicate glasses are used in certain AB cement formulations,
and the acid-base balance in them is important. The Bronsted-Lowry
theory cannot be applied to these aluminosilicate glasses; it does not
recognize silica as an acid, because silica is an aprotic acid. However, for
most purposes the Bronsted-Lowry theory is a suitable conceptual
framework although not of universal application in AB cement theory.
The solvent system theory
Although the protonic theory is not confined to aqueous solutions, it does
not cover aprotic solvents. The solvent system theory predates that of
Bronsted-Lowry and represents an extension of the Arrhenius theory to
solvents other than water. It may be represented by the defining equation:
Acid + Base = Salt + Solvent [2.4]
This theory is associated in its early protonic form with Franklin (1905,
1924). Later it was extended by Germann (1925a,b) and then by Cady &
Elsey (1922,1928) to a more general form to include aprotic solvents. Cady
& Elsey describe an acid as a solute that, either by direct dissociation or by
reaction with an ionizing solvent, increases the concentration of the solvent
cation. In a similar fashion, a base increases the concentration of the
solvent anion. Cady & Elsey, in order to emphasize the importance of the
solvent, modified the above defining equation to:
Acidic solution + Basic solution = Salt + Solvent [2.5]
Thus, acids and bases do not react directly but as solvent cations and
anions. Since emphasis is placed upon ionization interactions, inherent
acidity and basicity is neglected, as are interactions in the non-ionic state.
The theory is a simple extension of the Arrhenius theory and suffers from
16

Acid-base concepts
the same drawbacks. The definition cannot be applied directly to the
reaction between a basic solid and acidic liquid characteristic of AB
cements.
The Lux-Flood theory
The Lux-Flood theory relates to oxide melts. Geologists have often used
acid-base concepts for the empirical classification of igneous silicate rocks
(Read, 1948). Silica is implicitly assumed to be responsible for acidity, and
the silica content of a rock is used as a measure of its acid-base balance:
Rock type
Acid
Intermediate
Basic
Ultra-basic
Silica content (SiO2) %
>66
52-66
45-52
<45
Lux (1939) developed an acid-base theory for oxide melts where the oxide
ion plays an analogous but opposite role to that of the hydrogen ion in the
Bronsted theory. A base is an oxide donor and an acid is an oxide acceptor
(Lux, 1939; Flood & Forland, 1947a,b; Flood, Forland & Roald, 1947):
Base = Acid+ O2
~ [2.6]
Thus an acid-base reaction involves the transfer of an oxide ion (compared
with the transfer of a proton in the Bronsted theory) and the theory is
particularly applicable in considering acid-base relationships in oxide,
silicate and aluminosilicate glasses. However, we shall find that it is
subsumed within the Lewis definition.
The Lewis theory
This theory was advanced by G. N. Lewis (1916, 1923, 1938) as a more
general concept. In his classic monograph of 1923 he considered and
rejected both the protonic and solvent system theories as too restrictive. An
acid-base reaction in the Lewis sense means the completion of the stable
electronic configuration of the acceptor atom of the acid by an electron
pair from the base. Thus:
A base has the ability to donate apair of electrons and an acid the ability to
accept a pair of electrons toform a covalent bond. The product of a Lewis
acid—base reaction may be called an adduct, a coordination compound or a
coordination complex (Vander Werf 1961). Neither salt nor conjugate
acid—baseformation is a requirement.
17

Although Lewis and Bronsted bases comprise the same species, the same is
not true of their acids. Lewis acids include bare metal cations, while
Bronsted-Lowry acids do not. Also, Bell (1973) and Day & Selbin (1969)
have pointed out that Bronsted or protonic acids fit awkwardly into the
Lewis definition. Protonic acids cannot accept an electron pair as is
required in the Lewis definition, and a typical Lewis protonic acid appears
to be an adduct between a base and the acid H+
(Luder, 1940; Kolthoff,
1944). Thus, a protonic acid can only be regarded as a Lewis acid in the
sense that its reaction with a base involves the transient formation of an
unstable hydrogen bond adduct. For this reason, advocates of the Lewis
theory have sometimes termed protonic acids secondary acids (Bell, 1973).
This is an unfortunate term for the traditional acids.
Lewis (1938) was not content with a purely conceptual view of acids and
bases, for he also listed certain phenomenological criteria for an acid-base
reaction. The process of neutralization is a rapid one, an acid or base
displaces a weaker acid or base from its compounds, acids and bases may
be titrated against each other using coloured indicators, and both acids and
bases have catalytic effects.
The Lewis definition covers all AB cements, including the metal
oxide/metal oxysalt systems, because the theory recognizes bare cations as
aprotic acids. It is also particularly appropriate to the chelate cements,
where it is more natural to regard the product of the reaction as a
coordination complex rather than a salt. Its disadvantages are that the
definition is really too broad and that despite this it accommodates
protonic acids only with difficulty.
The Usanovich theory
The Usanovich theory is the most general of all acid-base theories.
According to Usanovich (1939) any process leading to the formation of a
salt is an acid-base reaction. The so-called' positive-negative' definition of
Usanovich runs as follows.
An acid is a species capable of yielding cations, combining with onions or
electrons, or neutralizing a base. Likewise a base is a species capable of
yielding anions or electrons, combining with cations, or neutralizing an acid.
When developed, this theory proved to be more general than the theory of
Lewis, for it includes all the above acid-base definitions and also includes
oxidation-reduction reactions.
18

Acid—base concepts
It is better than the Lewis theory for describing acid-base cements, for
it avoids the awkwardness that the Lewis definition has with protonic
acids. However, as Day & Selbin (1969) have observed, the generality of
the theory is such that it includes nearly all chemical reactions, so that
acid-base reactions could simply be termed 'chemical reactions'.
2.3.4 Relevance of acid-base theories to AB cements
The various acid-base definitions are summarized in the Venn diagram
(Fig. 2.1). From this it can be seen that the Usanovich definition subsumes
the Lewis definition, which in turn subsumes all other definitions (i.e.
Arrhenius, Bronsted-Lowry, Germann-Cady-Elsey, Lux-Flood).
Also shown is how the topic of AB cements relates to these definitions.
An ideal definition for a subject should be one that exactlyfitsit. It should
cover all aspects of the subject while excluding all extraneous topics. Thus,
a theory should be neither too restrictive nor too general. The Arrhenius
and Germann-Cady-Elsey definitions do not relate to the subject at all as
USANOVICH
LEWIS
Electron-pair acceptor
BR0NSTED
proton-donor
any solvent
ARRHENIUS
proton-donor
in water
GERMANN
sol vent-cation
donor
Figure 2.1 Venn diagram showing the relationship between the various definitions of acids
and bases.
19

the basic component of an AB cement is a powdered solid. The
Bronsted-Lowry definition is not broad enough to include all AB cements
and excludes the concept of salt, which is unfortunate since the matrices of
AB cements are salts. Both the Lewis and Usanovich definitions cover all
aspects of AB cement theory at the cost of including topics not relevant to
this subject.
From this discussion it can be seen that there is no ideal acid-base theory
for AB cements and a pragmatic approach has to be adopted. Since the
matrix is a salt, an AB cement can be defined quite simply as the product
of the reaction of a powder and liquid component to yield a salt-like gel.
The Bronsted-Lowry theory suffices to define all the bases and the
protonic acids, and the Lewis theory to define the aprotic acids. The subject
of acid-base balance in aluminosilicate glasses is covered by the Lux-Flood
theory.
2.3.5 Acid-base strength
Ever since the formulation of the Bronsted-Lowry theory, efforts have
been made to develop a general approach to acid-base strength. The
influence of ionic charge and size of the central atom on acidity and
basicity is important. In 1926, Bronsted found that an increase in acidity
corresponded to an increase in positive charge or a decrease in negative
charge on an ion. Cartledge (1928a,b), against the background of the
protonic theory, proposed to correlate acidity or basicity with a function
he called ionicpotential, by considering acids and bases to be hydroxides of
non-metals and metals, respectively. He defined ionic potential, (/>, as
</> = Z/r (2.1)
where Z is the charge on the central atom and r its ionic radius. Cartledge
(1928b) then used values of ^05
to define acidity and basicity of a species.
f5
value
>3-2
2-2-3-2
<2-2
Acid-base status
acidic
amphoteric
basic
Thus, highly charged smaller cations are highly acidic. This point is
illustrated for the series Na+
, Mg2+
, Al3+
, Si4+
, P5+
, S6+
and Cl7+
in Table
2.3a.
Note, however, that Zn(OH)2 is not classified as amphoteric as it should
20

Acid-base concepts
Table 2.3a. Effect of cation on acidity-basicity (Cartledge, 1982a,b)
Cation
Na+
Ca2+
Zn2+
Mg2+
Al3+
Si4+
p5+
S 6 +
Cl7+
Ionic potential
102
1-42
1-64
1-76
2-45
313
3-83
4-55
5-20
Species
NaOH
Ca(OH)2
Zn(OH)2
Mg(OH)2
A1(OH)3
Si(OH)4
H3PO4
H2SO4
HC1O4
Acidity-basicity
Strong base
Weak base
Weak base
Weak base
Amphoteric
Weak acid
Intermediate acid
Strong acid
Strongest acid
Table 2.3b. Effect of cation on acidity-basicity
Cation Ionization potential In Species Acidity-basicity
Na+
Ca2+
Mg2+
Cd2+
Zn2+
Cu2+
Bi3+
Al3+
Si4+
p5+
s6+
5-14
11-87
1503
16-84
17-96
20-20
25-42
28-45
45-14
6502
88-05
NaOH
Ca(OH)2
Mg(OH)2
Cd(OH)2
Zn(OH)2
Cu(OH)2
Bi(OH)3
A1(OH)3
Si(OH)4
H3PO4
H2SO4
Strong base
Weak base
Weak base
Amphoteric
Amphoteric
Amphoteric
Amphoteric
Amphoteric
Weak acid
Intermediate acid
Strong acid
In is the nth ionization potential.
be. Clearly, ionic potential alone is not a sufficient criterion for classi-
fication. As will be shown, unlike other cations in Table 2.3a which are
classified as hard acids, Zn2+
is an intermediate because of the presence of
d orbital electrons. The effect of d electrons in increasing the polarizing
power of the cations, because of ineffective screening, has been demon-
strated by Hodd & Reader (1976). They found that Cd2+
was a more
effective cement-former than Ca2+
, because although both have a similar
ionic radius, Ca2+
has no d electrons. For these reasons, ionization
potential is a better criterion than ionic potential. As Table 2.3b shows,
Zn2+
is ranked correctly by this criterion and can be classified as
21

amphoteric. Inspection of this table throws some light on the requirements
for cement formation. If judged by strength and hydrolytic stability of
cements formed with orthophosphoric acid, poly(acrylic acid) and poly-
(vinylphosphonic acid), the common cement-forming cations can be
ranked in the following order of decreasing effectiveness.
Al3+
> Cu2+
> Zn2+
> Mg2+
> Ca2+
The first three form amphoteric oxides and are distinctly superior, as
cement-formers, to the latter two which form weakly basic oxides. Data
from Table 2.3b indicate that optimum cement formation occurs with
cations that have In values lying between 18 and 29.
2.3.6 Acid-base classification
The strength of a Lewis acid or base depends on the particular reaction,
and for this reason there is no absolute scale for the strengths of Lewis
acids and bases. However, certain qualitative features have been observed.
Ahrland, Chatt & Davies (1958) divided metal ions (which are Lewis
acids), on the basis of the stability of their complexes, into what they
termed class (a) and class (b) acceptors (Table 2.4). They stated that class
(a) acceptors form their most stable complexes with ligands of the lightest
member of a non-metal group. By contrast, class (b) acceptors form their
most stable complexes with heavier members of each group. Thus, complex
stability can be ranked according to the ligand as follows. For class (a)
acceptors O P S and for class (b) acceptors O <
^ S. Class (a) metal ions are
small and non-polarizable, whereas class (b) metal ions are large and
polarizable. The class of a given element is not constant and depends on
oxidation state; class (a) character increases with increase in the positive
charge. Chatt (1958) considers that the important feature of class (b) acids
is the presence of loosely held outer d orbital electrons which can form n-
bonds to certain ligands. These ligands would contain empty d orbitals on
the basic atom; examples are P and As.
In the context of AB cements, Al3+
, Mg2+
, Ca2+
and Zn2+
are in class (a)
while Cu2+
is in the border region. Zn2+
contains a completed 3d shell and
forms stronger complexes with O than with S ligands, as do other class (a)
cations.
22

Table 2.4. Classification of acceptor atoms in their normal valent states (Ahrland, Chatt & Davies, 1958)
H
Li
Na
K
Rb
Cs
Be
Mp
Ca I
Sr ^
Ba 1
>c
ft
^a
Class (a)
Ti V
Zr Nb
Hf Ta
Cr
Mo
W
Mn
Tc
Re
a/b border
Fe
Ru
Os
Co
Rh
Ir
Ni
Pd
Pt
a/b border
Cu
Ag
Au
B
Al
Zn Ga
Cd
Hg
In
TI
C
Si
Ge
Sn
Pb
N
P
As
Sb
Bi
Class
0
s
Se
Te
Po
(a)
F
Cl
Br
I
At
Class (a) a/b border Class (b) a/b border

2.3.7 Hard and soft acids and bases (HSAB)
This concept of Chatt and his coworkers was developed further by Pearson
(1963, 1966, 1968a,b) in his theory of hard and soft acids and bases. Hard
acids correspond with class (a) acceptors and soft acids with class (b)
acceptors.
Hard acids prefer to react with hard bases and soft acids prefer to react
with soft bases.
Hard acids are characterized by small size, high positive charge and
absence of outer electrons which are easily excited to higher states; they are
thus of low polarizability. In this class are the common protonic acids, HA,
the hydrogen-bonding molecules in the Lewis scheme and Mg2+
, which are
all acids of relevance to AB cements. The soft acids have low or zero
positive charge, large size and several easily excited outer electrons (often
d orbital electrons). These properties lead to high polarizability. The
division between these two classes is not sharp; amongst the intermediate
class are Zn2+
and Cu2+
.
Pearson (1966) defines a soft base as 'one in which the donor atom is of
high polarizability and low electronegativity and is easily oxidized or
associated with empty, low-lying orbitals'. A hard base has opposite
properties. 'The donor atom is of low polarizability and high electro-
negativity, is hard to reduce, and is associated with empty orbitals of high
energy.'
The underlying theory for hard-hard and soft-soft preferences is
obscure and no one factor is responsible (Pearson, 1966). Pearson (1963,
1968b) advanced several explanations. He stated that the ionic-covalent
theory provides the most obvious explanation. Hard acids are assumed to
bind bases primarily by ionic forces and soft acids by covalent bonds. High
positive charge and small size favour strong ionic bonding, and bases of
large negative charge and small size would be most strongly held. Soft acids
bind to bases by covalent bonding, and the atoms should be of similar size
and electronegativity for good bonding.
The classification of Lewis acids and bases relevant to AB cements is
shown below.
Hard acids: HA, H+
, Ca2+
, Mg2+
, Al3+
, Si4+
Borderline acids: Zn2+
, Cu2+
Hard bases: H2O, O H , F", POJ", SO2
", RCOO"
24

Acid-base concepts
Table 2.5. YatsimirskiVs hardness indices {Yatsimirskii, 1970)
Base Indices Acid Indices
OH-
F-
HPOJ-
CH3COO-
sor
H2O
6-3
1-7
1-7
0-8
0-5
zero
H+
In3+
Cu2+
Zn2+
La3+
9-0
1-2
10
0-2
01
Extension of HSAB theory
Yatsimirskii (1970) attempted to quantify HSAB theory and produced
hardness indices (S) for acids and bases. These indices were obtained by
plotting the logarithms of the equilibrium constants for the reactions of
bases with the proton (the hardest acid) against similar values for the
reactions with CH3Hg+
(one of the softest acids). For acids, the hydroxyl
ion (the hardest base) and the chloride ion (a soft base) were chosen.
These S indices for cations and anions relevant to AB cements are shown
in Table 2.5. Bases which add on through F or O and do not form Tr-bonds
have similar hardness values; they are hard bases. Soft bases form dative
7r-bonds with many cations. They have high-energy-level occupied orbitals
with unshared electron pairs.
Yatsimirskii considered that the hard and soft classification was too
general and proposed instead a more detailed approach. He classified
Lewis acids and bases into six groups, based on the nature of the adduct
bonding.
Group (1) Cations and anions which are incapable of donor-acceptor
interactions. These are the large univalent ions. Bonding is purely
by Coulomb and Madelung electrostatic interactions. From the
Lewis point of view these are not acids or bases. They have no
cement-forming potential.
Group (2) Strong a-acceptor acids and donor bases. Included here are
protonic acids, which are relevant to AB cements. Their adducts
can only contain one coordinate bond.
Group (3) G- and n-acceptor acids and donor bases with o-interactions
predominating. In this group acceptors are capable of adding on
electron pairs of donors in both types of interactions. Includes
cations with stable closed electron shells: Al3+
, Mg2+
, Ca2+
and
25

Zn2+
. Donors are ligands coordinated through oxygen atoms or
fluoride ions: RCOO", PO*~, OH", F" and H2O. These acceptors
and donors are of relevance to AB cements.
Group (4) Strong a- and n-acceptor acids and donor bases. Bi3+
, In3+
and Sn2+
are of some relevance to AB cements.
Group (5) Acids that are o-acceptors but capable of n-donation in
backbonding. This group includes cations with mobile d electrons
e.g. Cuw+
, Cow+
, Few+
.
Group (6) Bases that are a-donors but n-acceptors.
According to Yatsimirskii, group (2) and (3) species are equivalent to
Pearson's hard acids and bases, and group (4), (5) and (6) species
correspond to Pearson's soft acids and bases. This classification is of more
value than HSAB theory to our subject. It can be seen that all cement-
forming anions come from group (3) and cations from groups (3), (4) and
(5). Thus, the bonding in cement matrices formed from cation-anion
combinations is not purely a but contains some n character.
References
Ahrland, S., Chatt, J. & Davies, N. R. (1958). The relative affinities of ligand
atoms for acceptor molecules and ions. Quarterly Reviews, 12, 265-76.
Baes, C. F. & Mesmer, R. E. (1976). The Hydrolysis of Cations. New York:
John Wiley.
Bell, R. P. (1947). The use of the terms 'acid' and 'base'. Quarterly Reviews, 1,
113-25.
Bell, R. P. (1973). The Proton in Chemistry. Ithaca, New York: Cornell
University Press.
Bjerrum, J. (1951). Die Entwickhmgsgeschichte des Saure-Basenbegriffes und
iiber die ZweckmaBigkeit der Einfuhrung eines besonderen Antibasenbegriffes
neben dem Saurebegriff. Naturwissenschaften, 38, 461-4.
Boyle, R. (1661). The Sceptical Chymist. Everyman Library Edition, 1911.
Brensted, J. N. (1923). Einige Bemerkungen iiber den Begriff der Sauren und
Basen. Recueil des Travaux chimiques des Pays-Bas et de la Belgique, 42,
718-28.
Bronsted, J. N. (1926). The acid-base function of molecules and its dependency
on the electronic charge type. Journal of Physical Chemistry, 30, 777-90.
Bungenberg de Jong, H. G. (1949). In Kruyt, H. R. (ed.) Colloid Science II, p. 2.
Amsterdam: Elsevier Publishing Co. Inc.
Cady, H. P. & Elsey, H. M. (1922). A general conception of acids, bases and
salts. Science, 56, 27 (Lecture abstract).
Cady, H. P. & Elsey, H. M. (1928). A general definition of acids, bases and
salts. Journal of Chemical Education, 5, 1425-8.
26

References
Cartledge, G. H. (1928a). Studies on the periodic system. I. The ionic potential
as a periodic function. Journal of the American Chemical Society, 50, 2855-63.
Cartledge, G. H. (1928b). Studies on the periodic system. II. The ionic potential
and related properties. Journal of the American Chemical Society, 50, 2863-72.
Chatt, J. (1958). The stabilisation of low valent states of the transition metals.
Journal of Inorganic & Nuclear Chemistry, 8, 515-31.
Cotton, F. A. & Wilkinson, G. (1966). Advanced Inorganic Chemistry, 2nd edn.
New York, London & Sydney: Wiley Interscience.
Crisp, S., O'Neill, I. K., Prosser, H. J., Stuart, B. & Wilson, A. D. (1978).
Infrared spectroscopic studies on the development of crystallinity in dental
zinc phosphate cements. Journal of Dental Research, 57, 245-54.
Crosland, M. P. (1962). Historical Studies in the Language of Chemistry.
London: Heinemann.
Crosland, M. (1973). Lavoisier's theory of acidity. Isis, 64, 306-25.
Day, M. C. & Selbin, J. (1969). Theoretical Inorganic Chemistry. New York:
Reinhold.
Finston, H. L. & Rychtman, A. C. (1982). A New View of Current Acid-Base
Theories. New York: John Wiley & Sons.
Flood, H. & Forland, T. (1947a). The acidic and basic properties of oxides.
Ada Chemica Scandinavica, 1, 592—604.
Flood, H. & Forland, T. (1947b). The acidic and basic properties of oxides. II.
The thermal decomposition of pyrosulphates. Acta Chemica Scandinavica, 1,
781-9.
Flood, H., Forland, T. & Roald, B. (1947). The acidic and basic properties of
oxides. III. Relative acid-base strengths of some polyacids. Acta Chemica
Scandinavica, 1, 790-8.
Flory, P. J. (1953). Principles of Polymer Chemistry, Chapter 11. Ithaca, New
York: Cornell University Press.
Flory, P. J. (1974). Introductory lecture. In Gels and Gelling Processes. Faraday
Discussions of the Chemical Society, No. 57, pp. 7-18.
Franklin, E. C. (1905). Reactions in liquid ammonia. Journal of the American
Chemical Society, 27, 820-51.
Franklin, E. C. (1924). Systems of acids, bases and salts. Journal of the
American Chemical Society, 46, 2137-51.
Germann, A. F. O. (1925a). What is an acid? Science, 61, 71.
Germann, A. F. O. (1925b). A general theory of solvent systems. Journal of the
American Chemical Society, 47, 2461-8.
Hall, N. F. (1940). Systems of acids and bases. Journal of Chemical Education,
17, 124^8.
Hodd, K. A. & Reader, A. L. (1976). The formation and hydrolytic stability of
metal ion-polyacid gels. British Polymer Journal, 8, 131-9.
Jensen, W. B. (1978). The Lewis acid-base definitions: a status report. Chemical
Reviews, 78, 1-22.
Kingery, W. D. (1950a). Fundamental study of phosphate bonding in
refractories. I. Literature review. Journal of the American Ceramic Society, 33,
239-41.
27

Kingery, W. D. (1950b). Fundamental study of phosphate bonding in
refractories. II. Cold setting properties. Journal of the American Ceramic
Society, 33, 242-7.
Kolthoff, I. M. (1944). The Lewis and Bronsted-Lowry definitions of acids and
bases. Journal of Physical Chemistry, 48, 51-7.
Lewis, G. N. (1916). The atom and the molecule. Journal of the American
Lewis, G. N. (1923). Valence and the Structure of Atoms and Molecules. New
York: Chemical Catalog Co.
Lewis, G. N. (1938). Acids and bases. Journal of the Franklin Institute, 226,
293-337.
Lowry, T. M. (1923a). The uniqueness of hydrogen. Chemistry & Industry, 42,
43.
Lowry, T. M. (1923b). Co-ordination and acidity. Chemistry & Industry, 42,
1048-52.
Luder, W. F. (1940). The electronic theory of acids and bases. Chemical
Reviews, 27, 547-83.
Luder, W. F. (1948). Contemporary acid-base theory. Journal of Chemical
Education, 25, 555-8.
Lux, H. (1939). 'Sauren' und 'Basen' im Schelzfluss: Die Bestimmung der
Sauerstoffionen-Konzentration. Zeitschrift fur Elektrochemie, 45, 303-9.
Pattison Muir, M. M. (1883). Heroes of Science-Chemists, Chapter IV, pp.
171-89. London: Society for Promoting Christian Knowledge.
Pauling, L. (1945). The Nature of the Chemical Bond. Ithaca, New York:
Cornell University Press.
Pearson, R. G. (1963). Hard and soft acids and bases. Journal of the American
Pearson, R. G. (1966). Acids and bases. Science, 151, 172-7.
Pearson, R. G. (1968a). Hard and soft acids and bases, HSAB. Part I.
Fundamental principles. Journal of Chemical Education, 45, 581-7.
Pearson, R. G. (1968b). Hard and soft acids and bases, HSAB. Part II.
Underlying theories. Journal of Chemical Education, 45, 643-8.
Read, H. H. (1948). Rutle/s Elements of Mineralogy, 24th edn. London:
Thomas Murby & Co.
Smith, D. C. (1968). A new dental cement. British Dental Journal, 125, 381-4.
Steinke, R., Newcomer, P., Komarneni, S. & Roy, R. (1988). Dental cements:
investigation of chemical bonding. Materials Research Bulletin, 23, 13-22.
Usanovich, M. I. (1939). On acids and bases. Journal of General Chemistry
(USSR), 9, 182-92.
Vander Werf, A. (1961). Acids, Bases, and the Chemistry of the Covalent Bond.
New York: Reinhold.
Walden, P. (1929). Salts, Acids and Bases: Electrolytes, Stereochemistry. New
York: McGraw-Hill.
Wilson, A. D. (1968). Dental silicate cements: VII. Alternative liquid cement
formers. Journal of Dental Research, 47, 1133-6.
Wilson, A. D. & Kent, B. E. (1971). The glass-ionomer cement: a new
28

References
translucent cement for dentistry. Journal of Applied Chemistry and
Biotechnology, 21, 313.
Wilson, A. D., Paddon, J. M. & Crisp, S. (1979). The hydration of dental
cements. Journal of Dental Research, 58, 1065-71.
Wygant, J. F. (1958). Cementitious bonding in ceramic fabrication. In Kingery,
W. D. (ed.) Ceramic Fabrication Processes, pp. 171-88. New York: John
Wiley & Sons.
Yatsimirskii, K. (1970). Acid-base and donor-acceptor properties of ions and
molecules. Theoretical and Experimental Chemistry (USSR), 6, 376-80.
29

3 Water and acid-base cements
3.1 Introduction
The setting reaction for the great majority of acid-base cements takes place
in water. (The exceptions based on o-phenols are described in Chapter 9.)
This reaction does not usually proceed with formation of a precipitate but
rather yields a substance which entrains all of the water used to prepare the
original cement paste. Water thus acts as both solvent and component in
the formation of these cements. It is also one of the reaction products,
being formed in the acid-base reaction as the cements set.
3.1.1 Water as a solvent
It is widely recognized that the solvent in which any chemical reaction
takes place is not merely a passive medium in which relevant molecules
perform: the solvent itself makes an essential contribution to the reaction.
The character of the solvent will determine which chemical species are
soluble enough to enter solution and hence to react, and which species are
insoluble, and thus precipitate out of solution, thereby being prevented
from undergoing further chemical change. In the case of water, as will be
seen, polar and ionic species are the ones that most readily dissolve. But
even so, mere polarity or ionic character is not sufficient to ensure
solubility. Solubility depends on a number of subtle energetic factors, and
the possible interactions between water and silver chloride, for example, do
not fulfil the requirements despite the ionic nature of the silver salt. Hence
silver chloride is almost completely insoluble in water.
3.1.2 Water as a component
In AB cements water does not merely act as solvent for the setting reaction.
It also acts as an important component of the set cement. For example,
30

Water
glass-ionomer dental cements as generally formulated include at least
15% by mass of water, all of which becomes incorporated into the
complete cement (Wilson & McLean, 1988). Indeed, great importance is
attached to the retention of water by these cements, since if they are
allowed to dry out by storage under conditions of low humidity, they
shrink significantly, and develop cracks and crazes.
Another class of AB cement, the oxychloride cements of zinc and
magnesium, are also formulated in aqueous solution and retain substantial
amounts of water on setting (Sorrell & Armstrong, 1976; Sorrell, 1977).
Water may have a number of roles in the set versions of these cements.
It is capable of solvating the cement-forming ions, such as Ca2+
or Zn2+
,
depending on the cement. It also contributes a sheath of solvating
molecules around polyelectrolytes such as poly(acrylic acid) in glass-
ionomer and zinc polycarboxylate cements. Significant amounts of water
are known to be retained by metal polyacrylate salts at equilibrium and
this water contributes to reducing the glass transition temperature of such
materials by acting as a plasticizer (Yokoyama & Hiraoko, 1979).
These various aspects of water in AB cements are covered in the present
chapter. Its solvent character, structure and hydration behaviour are
described, and the chapter concludes with a more thorough consideration
of the precise role of water in the various AB cements.
3.2 Water
3.2.1 Constitution
Water has a deceptively simple chemical constitution, consisting as it does
of molecules containing two atoms of hydrogen and one of oxygen. It was
viewed by the ancients as one of the four 'elements', following Aristotle's
classification, the others being air, fire and earth. The modern view that it
is a compound composed of hydrogen and oxygen was first established in
1789 by two amateur chemists, Adriaan Paets van Troostwijk (1752-1837),
a merchant, and Jan Rudolph Deiman (1743-1808), a pharmacist (Hall,
1985). They were able to show by synthesis which elements combine to
make water, forming it from reaction of hydrogen gas with oxygen. Their
work was important historically for the part it played in undermining the
phlogiston theory of combustion. It was left to the great Swedish chemist
J. J. Berzelius (1779-1848) to determine that the ratio of hydrogen to
oxygen is 2:1.
31

Water and acid-base cements
Table 3.1. Molecular dimensions of normal and isotopic water in
the vapour phase {Benedict, Gailar & Plyler, 1956)
Bond length, Bond angle,
Molecule pm degrees
D2O 95-75 104-474
H2O 95-718 104-523
HDO 95-71 104-529
As a compound water is remarkable. It is the only inorganic liquid to
occur naturally on earth, and it is the only substance found in nature in all
three physical states, solid, liquid and vapour (Franks, 1983). It is the most
readily available solvent and plays a vital role in the continuation of life on
earth. Water circulates continuously in the environment by evaporation
from the hydrosphere and subsequent precipitation from the atmosphere.
This overall process is known as the hydrologic cycle. Reports estimate
that the atmosphere contains about 6 x 1015
litres of water, and this is
cycled some 37 times a year to give an annual total precipitation of
224 x 1015
litres (Franks, 1983; Nicholson, 1985).
The bond lengths and bond angle for the water molecule are known
very precisely following studies of the rotation-vibration spectra of water
vapour, and also the vapour of the deuterated analogues of water, D2 O
and HDO (Eisenberg & Kauzmann, 1969). The data for these compounds
are shown in Table 3.1. The nuclei of the water molecule, regardless of the
isotopes involved, form an isosceles triangle having a slightly obtuse angle
at the oxygen atom. All of the data in Table 3.1 refer to the equilibrium
state of the water molecules, which is formally acceptable, but is actually
a hypothetical state, since it assumes neither rotation nor vibration in the
molecule.
The equilibrium bond lengths and bond angles can be seen to differ little
between the different isotopic molecules. Such a finding agrees with the
predictions of the Born-Oppenheimer approximation, that the electronic
structure of a molecule is independent of the mass of its nuclei, it being the
electronic structure of a molecule alone which determines the geometry.
The bond angle in water is slightly less than the ideal tetrahedral angle
of 109-5°.This is attributed to the presence of lone pairs of electrons on the
oxygen atom which repel more strongly than the bonding pairs of electrons
between the oxygen and hydrogen nuclei (Speakman, 1975). The valence-
32

Water
Table 3.2. Properties of hydrides of first row elements (Weast, 1985-6)
Compound
CH4
NH3
H2O
HF
Relative
molar mass
16
17
18
19
Melting point,
°C
-182-0
-11-1
00
- 8 3 4
Boiling point,
°C
-1640
-33-4
1000
19-5
Gas
phase
dipole
moment,
debyes
000
1-47
1-85
1-82
shell electron-pair repulsion concepts of Gillespie & Nyholm (1957) show
that such increased repulsion by lone pairs closes the angle between the
bonding pairs slightly but significantly for the water molecule.
The O-H bond energy of water is taken as half the energy of formation
of the molecule, since water has two such bonds. This gives a value of
458-55 kJ mol"1
at 0 K (Eisenberg & Kauzmann, 1969). Related to the
bond energy is the dissociation energy, i.e. the energy required to break the
bond at 0 K. Neither of the O-H bonds in water has a dissociation energy
equal to the O-H bond energy. Instead, the first O-H dissociation energy
has been found experimentally to be 424-27 kJ mol"1
. From conservation
of energy considerations which lead to the requirement that the sum of the
two dissociation energies must equal the energy of formation, it is found
that the second O-H dissociation energy has to take a value of
492-83 kJ mol"1
. This has been explained (Pauling, 1960) by postulating an
electronic rearrangement on the oxygen atom of the O-H fragment left
behind after scission of the first O-H bond, and that breaking the bond
between oxygen in this new electronic configuration and the remaining
hydrogen requires greater energy.
3.2.2 Water compared with other hydrides
Water shows properties that are interestingly different compared with
hydrides of the neighbouring elements of thefirstrow of the periodic table.
Some of these properties are given in Table 3.2. From this table, water can
be seen to have a very high melting point and a very high boiling point for
its relative molar mass. Indeed, it is the only one of the hydrides of the
33

elements from this portion of the periodic table to be liquid at room
temperature and atmospheric pressure. In the gas phase it has a dipole
moment that, while only slightly greater than that of hydrogenfluoride,is
the highest for this group of hydrides. All of these properties point to water
having a structure in which its constituent molecules are more highly
associated and interact more strongly than the molecules of the closely
related hydrides.
3.3 The structure of water
At first sight the concept of a 'structure' for liquid water appears strange.
In the solid state atoms are relatively fixed in space, albeit with some
vibrational motion about equilibrium positions, and no difficulty is
associated with the idea of locating these equilibrium positions by some
appropriate physical technique, and thereby assigning a structure to the
solid.
3.3.1 The concept of structure in the liquid state
With water or any other liquid, molecules do not occupy even reasonably
fixed locations but have considerably more freedom for movement than in
the solid state. What then do we mean by the term structure applied to a
liquid?
To answer this question we need to consider the kind of physical
techniques that are used to study the solid state. The main ones are based
on diffraction, which may be of electrons, neutrons or X-rays (Moore,
1972; Franks, 1983). In all cases exposure of a crystalline solid to a beam
of the particular type gives rise to a well-defined diffraction pattern, which
by appropriate mathematical techniques can be interpreted to give
information about the structure of the solid. When a liquid such as water
is exposed to X-rays, electrons or neutrons, diffraction patterns are
produced, though they have much less regularity and detail; it is also more
difficult to interpret them than for solids. Such results are taken to show
that liquids do, in fact, have some kind of long-range order which can
justifiably be referred to as a 'structure'.
In considering the structure of a liquid, two possible conceptual
approaches exist. One is to begin from an understanding of the gaseous
34

The structure of water
state, characterized as it is by gross translational movement of the
constituent molecules and substantial disorder. The liquid is then viewed as
a gas that has been condensed and in which translational motion has
become constrained. Alternatively, consideration can start from the solid
state, with its well-characterized structure, having little or no translational
motion, but some vibrational motion of the constituent atoms or
molecules. The liquid state is then viewed as a solid in which some degree
of translational motion has become allowed, but with a structure still
recognizable as being derived from that existing in the solid state (Franks,
1983). With the growth in application of the techniques of X-ray and
neutron diffraction to the study of the liquid state, the latter approach has
become increasingly favoured in recent years.
In this section, rather than give a detailed account of theories of the
liquid state, a more qualitative approach is adopted. What follows includes
first a description of the structure of ice; then from that starting-point,
ideas concerning the structure of liquid water are explained.
3.3.2 The structures of ice
Water is capable of solidifying into a number of different structural states
or polymorphs depending, for example, on the external pressure applied
during solidification. The simplest and most common of these polymorphs
is known as ice I, whose structure was first determined by W. H. Bragg
(1922). In this structure, every oxygen atom occupies the centre of a
tetrahedron formed by four oxygen atoms, each about 0-276 nm away. The
water molecules are connected together by hydrogen bonds, each molecule
being bonded to its four nearest neighbours. The O-H bonds of a given
molecule are oriented towards the lone pairs on two of these neighbouring
molecules, and in turn, each of its lone pairs is directed towards an O-H
bond of one of the other neighbours. This arrangement gives an open
lattice in which intermolecular cohesion is large.
The arrangement of oxygen atoms in ice I is isomorphous with the
wurtzite form of zinc sulphide, and also with the silicon atoms in the
tridymite form of silicon dioxide. Hence, ice I is sometimes referred to as
the wurtzite or tridymite form of ice (Eisenberg & Kauzmann, 1969).
Location of the hydrogen atoms in ice I has caused more problems. This
is because hydrogen is less effective at scattering X-rays or electrons than
oxygen. For a long time, arguments about the position of hydrogen were
based on indirect evidence, such as vibrational spectra or estimates of
35

residual entropy at 0 K (Eisenberg & Kauzmann, 1969). Since the advent
of neutron diffraction the positions of the hydrogen atoms have become
clearer. These studies have shown that the water molecules have very
similar dimensions in ice I to those in the isolated molecule: the O-H bond
length is 0-101 nm and the bond angle 104*5°.
Ice I is one of at least nine polymorphic forms of ice. Ices II to VII
are crystalline modifications of various types, formed at high pressures;
ice VIII is a low-temperature modification of ice VII. Many of these
polymorphs exist metastably at liquid nitrogen temperature and atmos-
pheric pressure, and hence it has been possible to study their structures
without undue difficulty. In addition to these crystalline polymorphs, so-
called vitreous ice has been found within the low-temperaturefieldof ice I.
It is not a polymorph, however, since it is a glass, i.e. a highly supercooled
liquid. It is formed when water vapour condenses on surfaces cooled to
below -160°C.
It is not appropriate in this chapter to give a detailed review of the solid-
state behaviour of water in its various crystalline modifications. However,
there are some general structures which are relevant and worth high-
lighting. Firstly, water molecules in these various solids have dimensions
and bond angles which do not differ much from those of an isolated water
molecule. Secondly, the number of nearest neighbours to which each
individual molecule is hydrogen-bonded remains four, regardless of the ice
polymorph.
The differences in structure between the polymorphs, particularly the
high-pressure ones, lie in (a) the distances between the non-hydrogen
bonded molecules, and hence the amount of' free volume' in the structure,
(b) the angles of the hydrogen bonds, which may differ markedly from the
180° of ice I, and (c) the distance between nearest neighbouring oxygen
atoms, which may fall to well below the 0-276 nm value in ice I. All of these
are consistent with closer packing of the water molecules, and a closing up
of the cage structure by comparison with that found for ice I.
3.3.3 Liquid water
Before considering the details of the structure of liquid water, it is
important to define precisely what is meant by the term structure as applied
to this liquid. If we start from ice I, in which molecules are vibrating about
mean positions in a lattice, and apply heat, the molecules vibrate with
greater energy. Gradually they become free to move from their original
36

The structure of water
lattice sites and acquire significant translational energy. However, trans-
lational energy is not confined to molecules in the liquid state. There is a
finite possibility of any molecule in ice I moving from its lattice site, thus
acquiring translational energy. In principle, a given molecule can move
through the solid structure in a process that is essentially diffusion.
From this model of ice I we derive three meanings of the term structure
for the solid. We may refer to the positions of the molecules at an instant
of time. We may allow some averaging of the positions, i.e. we may have
a vibrationally averaged structure, considered over a short time-period,
during which molecules have time to undergo only minor vibrational
reorientations. Finally we may have a diffusionally averaged structure,
considered over longer time-periods, in which the minor translational
motion has been allowed to proceed to such an extent as to be significant.
These three possible structures, the instantaneous, the vibrationally
averaged and the diffusionally averaged, are referred to as I-, V- and D-
structures respectively.
Let us now turn our attention to liquid water. Just as in ice I, molecular
motions may be divided into rapid vibrations and slower diffusional
motions. In the liquid, however, vibrations are not centred on essentially
fixed lattice sites, but around temporary equilibrium positions that are
themselves subject to movement. Water at any instant may thus be
considered to have an I-structure. An instant later, this I-structure will be
modified as a result of vibrations, but not by any additional displacements
of the molecules. This, together with the first I-structure, is one of the
structures that may be averaged to allow for vibration, thereby con-
tributing to the V-structure. Lastly, if we consider the structure around an
individual water molecule over a long time-period, and realize that there is
always some order in the arrangement of adjacent molecules in a liquid
even over a reasonable duration, then we have the diffusionally averaged
D-structure.
No experimental technique exists for determining I-structures in either
the liquid or the solid state. Techniques do exist for obtaining information
on both the V- and D-structures of liquid water; the results of applying
these techniques are considered next.
Spectroscopic studies have established that for liquid water, the V-
structure has the following features.
(a) Considerable local variation between the environments of the
individual water molecules, compared with the relatively uniform
37

molecular environments in a crystal of ice I. The frequency spans
of the uncoupled O-H and O-D spectral bands indicate that some
nearest neighbours are as close as 0-275 nm, while others are
separated by 0.310 nm or more. The most probable equilibrium
separation is about 0.285 nm (Eisenberg & Kauzmann, 1969).
(b) The differences between the various molecular environments are
continuous. In other words, the V-structure does not contain
discrete types of molecular environment.
(c) The frequency of the stretching band indicates that hydrogen
bonds in the V-structure are weaker than those in ice I, though still
distinctly present.
Ideas about the D-structure have come mainly from two sources, namely
a consideration of the underlying reasons for the values of certain physical
properties, such as heat capacity or compressibility, and a study of radial
distribution functions that arise from X-ray diffraction work on liquid
water. The D-structure represents the average arrangement of molecules
around an arbitrary central water molecule. This average is either the
'space average' for several central molecules in different V-structures, or
the 'time average' for a single molecule over very long periods of time.
Near the freezing point, the D-structure is found to have relatively high
concentrations of neighbours at distances 0-29, 0-50 and 0*70 nm from the
central water molecule. This suggests that a substantial hydrogen-bonded
network is discernible, even in the liquid state. As the temperature is raised,
so the distinct concentrations at 0-50 and 0-70 nm disappear. Thermal
agitation thus distorts or destroys the hydrogen-bonded networks, and the
amount of observable long-range order decreases significantly.
Structural studies on liquid water reveal that the majority of molecules
are effectively tetrahedral, since the O-H bonds and the lone pairs are used
in hydrogen-bonding. Questions remain about the nature of these
hydrogen-bonds (Symons, 1989). Specifically: on average, how many such
hydrogen bonds are formed per molecule, how strong and how linear are
they, and what is their lifetime? One recent approach has been to consider
the possibility that, because of their weakness, some of the hydrogen bonds
in liquid water will break. This then gives concentrations of free O-H
bonds, OHfree, and free lone pairs, LPfree, on certain molecules which are
bonded to only three others (Symons, 1989). Symons (1989) also suggests
that the chemical properties of liquid water depend on the relative
concentrations of these species. Fully hydrogen-bonded water can be
38

Alan D. Wilson, John W. Nicholson-Acid-Base Cements_ Their Biomedical and Industrial Applications (Chemistry of Solid State Materials) (2005).pdf

Alan D. Wilson, John W. Nicholson-Acid-Base Cements_ Their Biomedical and Industrial Applications (Chemistry of Solid State Materials) (2005).pdf

Recommended

Recommended

More Related Content

Similar to Alan D. Wilson, John W. Nicholson-Acid-Base Cements_ Their Biomedical and Industrial Applications (Chemistry of Solid State Materials) (2005).pdf

Similar to Alan D. Wilson, John W. Nicholson-Acid-Base Cements_ Their Biomedical and Industrial Applications (Chemistry of Solid State Materials) (2005).pdf (20)

More from Muhammad565043

More from Muhammad565043 (12)

Recently uploaded

Recently uploaded (20)

Alan D. Wilson, John W. Nicholson-Acid-Base Cements_ Their Biomedical and Industrial Applications (Chemistry of Solid State Materials) (2005).pdf