Micas - Crystal Chemistry and Metamorphic Petrology 0939950588

Citation preview

REVIEWS in MINERALOGY and GEOCHEMISTRY Volume 46 2002

MICAS: CRYSTAL CHEMISTRY AND METAMORPHIC PETROLOGY Editors Annibale Mottana Francesco Paolo Sassi James B. Thompson, Jr. Stephen Guggenheim

Università degli Studi Roma Tre Università di Padova Harvard University University of Illinois at Chicago

FRONT COVER: Perspective view of TOT layers in Biotite down [100] ([001] is vertical), produced by CrystalMaker, Red tetrahedra contain Si and A1, green and white octahedra contain Mg and Fe, respectively, and yellow spheres represents the K interlayer cations. Courtesy of Mickey Gunter, University of Idaho, Moscow. [Data: S.R. Bohlen et al. (1980) Crystal chemistry of a metamorphic biotite and its significance in water barometry. Am Mineral 65: 55-62] BACK COVER: A view down [001] of lepitdolite-2M2, showing tetrahedrally coordinated Si,A1 (blue) joined with bridging oxygens (red thermal ellipsoids) in the T-Layer and ordered, octahedrally coordinated A1 (gray) and Li (yellow) in the O-layer. The interlayer cation I s12-coordinator K (green). Courtesy of Bob Downs, University of Arizona, Tucson. [Data: S. Guggenheim (1981) Cation ordering in lepidolite. Am Mineral 66: 1221-1232]

Series Editor for MSA: Paul H. Ribbe Virginia Polytechnic Institute and State University

MINERALOGICAL SOCIETY of AMERICA Washington, D.C. ACCADEMIA NATIONALE dei LINCEI Roma, Italia

COPYRIGHT 2002

MINERALOGICAL SOCIETY OF AMERICA The appearance of the code at the bottom of the first page of each chapter in this volume indicates the copyright owner’s consent that copies of the article can be made for personal use or internal use or for the personal use or internal use of specific clients, provided the original publication is cited. The consent is given on the condition, however, that the copier pay the stated per-copy fee through the Copyright Clearance Center, Inc. for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other types of copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. For permission to reprint entire articles in these cases and the like, consult the Administrator of the Mineralogical Society of America as to the royalty due to the Society.

REVIEWS IN MINERALOGY AND GEOCHEMISTRY ( Formerly: REVIEWS IN MINERALOGY )

ISSN 1529-6466

Volume 46

MICAS: Crystal Chemistry and Metamorphic Petrology ISBN 0-939950-58-8 ** This volume is the eighth of a series of review volumes published jointly under the banner of the Mineralogical Society of America and the Geochemical Society. The newly titled Reviews in Mineralogy and Geochemistry has been numbered contiguously with the previous series, Reviews in Mineralogy. Additional copies of this volume as well as others in this series may be obtained at moderate cost from: THE MINERALOGICAL SOCIETY OF AMERICA 1015 EIGHTEENTH STREET, NW, SUITE 601 WASHINGTON, DC 20036 U.S.A.

MICAS: Crystal Chemistry and Metamorphic Petrology Reviews in Mineralogy and Geochemistry Volume 46 2002 FORWARD The editors and contributing editors of this volume participated in a short course on micas in Rome late in the year 2000. It was organized by Prof. Annibale Mottana and several colleagues (details in the Preface below) and underwritten by the Italian National Acadmey, Accademai Nationale dei Lincei (ANL). The Academy subsequently joined with the Mineralogical Society of America (MSA) in publishing this volume. MSA is grateful for their generous involvement. I am particularly thankful to Prof. Mottana for Herculean efforts in supervising the editing of twelve manuscripts from six countries and submitting a single package containing everything needed to compile this volume! This was a uniquely positive experience fro me as Series editor for MSA. Assembling this volume was made tolerable by the exceptional efforts of Steve Guggenheim. During recovery from spinal surgery he spent three weeks painstakingly (no pun) correcting grammar and wording of the many authors from whom English is not their first language. Special thanks to him and the gracious and patient authors who suffered the extra work of assimilating both Steve’s suggestions and mine, above and beyond those of their reviewers and the editors. MSA’s Executive Director, Alex Speer, made all the contractual arrangements with ANL. This is the second of what we hope will be many co-operative projects with international colleagues and members of MSA. The first was in the year 2000: “Transformation Processes in Minerals,” RiMG 39, the proceedings of a short course at Cambridge University in partnership with four European scientific societies. Paul H. Ribbe, Series editor Blacksburg, Virginia April 20, 2002

PREFACE Micas are among the most common minerals in the Earth crust: 4.5% by volume. They are widespread in most if not all metamorphic rocks (abundance: 11%), and common also in sediment and sedimentary and igneous rocks. Characteristically, micas form in the uppermost greenschist facies and remain stable to the lower crust, including anatectic rocks (the only exception: granulite facies racks). Moreover, some micas are stable in sediments and diagenetic rocks and crystallize in many types of lavas. In contrast, they are also present in association with minerals originating from the very deepest parts of the mantle—they are the most common minerals accompanying diamond in kimberlites. The number of research papers dedicated to micas is enormous, but knowledge of them is limited and not as extensive as that of other rock-forming minerals, for reasons mostly relating to their complex layer texture that makes obtaining crystals suitable for careful studies with the modern methods time-consuming, painstaking work. Micas were reviewed extensively in 1984 (Reviews in Mineralogy 13, S.W. Bailey, editor). At that time, “Micas” volume covered most if not all aspects of mica knowledge, thus producing a long shelf-life for this book. Yet, or perhaps because of that iii 1529-6466/02/0046-0000$05.00

DOI: 10.2138/rmg.2002.46.0

excellent review, mica research was vigorously renewed, and a vast array of new data has been gathered over the past 15 years. These data now need to be organized and reviewed. Furthermore, a Committee nominated by the International Mineralogical Association in the late 1970s concluded its long-lasting work (Rieder et al. 1998) by suggesting a new classification scheme which has stimulated a new chemical and structural research on micas. To make a very long story short: -

-

-

the extraordinarily large, but intrinsically vague, micas nomenclature developed during the past two centuries has been reduced from >300 to just 37 species names and 6 series (see page xiii, preceding Chapter 1); the new nomenclature shows wide gaps that require data involving new chemical and structural work; the suggestion of using adjectival modifiers for those varieties that deviate away from end-member compositions requires the need fro new and accurate measurements, particularly fro certain light elements and volatiles; the use of polytype suffixes based on the modified Gard symbolism created better ways of determining precise stacking sequences. This resulted in new polytypes being discovered.

Indeed, all this has happened over the past few years in an almost tumultuous way. It was on the basis of these developments that four scientists (B. Zanettin, A. Mottana, F.P. Sassi and C. Cipriani) applied to Accademia Nazionale dei Lincei—the Italian National Academy—for a meeting on micas. An international meeting was convened in Rome on November 2-3, 2000 with the title Advances on Micas (Problems, Methods, Applications in Geodynamics). The topics of this meeting were the crystalchemical, petrological, and historical aspects if the micas. The organizers were both Academy members (C. Cipriani, A. Mottana, F.P. Sassi, W. Schreyer, J.B. Thompson Jr., and B. Zanettin) and Italian scientist well-known for their studies on layer silicates (Professors M.F. Brigatti and G. Ferraris). Financial support in addition to that by the Academy was provided by C.N.R. (the Italian National Research Council), M.U.R.S.T. (the Italian Ministry for University, Scientific Research and Technology) and the University of Rome III. Approximately 200 scientists attended the meeting, most of them Italians, but, with a sizeable international participation. Thirteen invited plenary lectures and six oral presentations were given, and fourteen posters were displayed. The amount of information presented was large, although the organizers made it very clear that the meeting was to be limited to only a few of the major topics of micas studies. Other studies are promised for a later meeting. Oral and poster presentations on novel aspects of mica research are being printed in the European Journal of Mineralogy, as apart of an individual thematic issue: indeed thirteen papers have appeared in the November 2001 issue. The plenary lectures, which consisted mostly of reviews, are presented in expanded detail in this volume. This book is the first a co-operative project between Accademia Nazionale dei Lincei and Mineralogical Society of America. Hopefully, future projects will involve reviews of the remaining aspects of mica research, and other aspects of mineralogy and geochemistry. The entire meeting was made successful through a co-operative effort. The editing of this book was achieved by a co-operative effort of two Italian Academy members from one side, and by two American scientists from the other side, one of them (JBT) being also a member of Lincei Academy. The entire editing process benefited from the goodwill of many referees, both from those attending Rome meeting and from several who did not. In all the reviewers were distinguished expert of the international iv

community of mica scholars. Their work, as well as our editing work, were aided greatly by RiMG Series Editor, Professor Paul Ribbe, who continuously supported the efforts with all his professional experience and friendly advice. We, the co-editors, thank them all very warmly, but take upon ourselves all remaining shortcomings: we are aware that some shortcomings may be present in spite of all our efforts to avoid them Moreover, we are aware that there are puzzling aspects of micas that are unresolved. Please consider all these possible avenues for future research! Annibale Mottana (Rome) Francesco Paolo Sassi (Padua) James B. Thompson, Jr. (Cambridge, Mass.) Stephen Guggenheim (Chicago)

v

This page left blank.

MICAS: CRYSTAL CHEMISTRY and METAMORPHIC PETROLOGY Editors: A Mottana, F P Sassi, J B Thompson, Jr & S Guggenheim

Table of Contents

1

Mica Crystal Chemistry and the Influence of Pressure, Temperature, and Solid Solution on Atonlistic Models Maria Franca Brigatti, Stephen Guggenheim

OVERVIEW Treatment of the data and definition of the parameters used End-member crystal chemistry: New end members and new data since 1984 Synthetic micas with unusual properties EFFECT OF COMPOSITION ON STRUCTURE ., Tetrahedral sheet Tetrahedral rotation and interlayer region Tetrahedral cation ordering Octahedral coordination and long-range octahedral ordering Crystal chemistry of micas in plutonic rocks ATOMISTIC MODELS INVOLVING HIGH-TEMPERATURE STUDIES OF THE MICAS Studies of samples having undergone heat treatment Dehydroxylation process for dioctahedral phyllosilicates Dehydroxylation models for trans-vacant 2: 1 layers Dehydroxy lation models for cis-vacant 2: 1 layers Compalison of Na-rich vs. K-rich dioctahedral forms Heat-treated trioctahedral samples: The O,OH,F site and in situ high-temperature studies Heat-treated trioctahedral samples: Polytype comparisons ACKNOWLEDGMENTS APPE~DIX I: DERIV ATIONS Derivation of "tetrahedral cation displacement", T di sp Derivation of f1E 1, f1E 2 , f1E 3 Derivation of ex Explanation of O[eor Explanation of E M - o(4) APPENDIX II: TABLES 1-4 Table 1a. Structural details of trioctahedral true micas-l M, space group C2/m Table 1b. Structural details of trioctahedral true micas-1M, space group C2 Table Ie. Structural details of trioctahedral true micas-2M], space group C2/c Table Id. Structural details oftrioctahedral true micas-2M J , space groups Ce. Cl Table Ie. Structural details of trioctahedral true micas-2M 2 , space group C2!c Table I f. Structural details of trioctahedral true micas-3T, space group P3,12 Table 2a. Structural details of trioctahedral true micas-I M, Mspace groups C2/m and C2 Table 2b. Structural details of trioctahedral true micas-1M, space group C2/c Table 2c. Structural details of trioctahedral true micas-2M, space group C2/e Table 2d. Structural details of trioctahedral true micas-3T, space group P3 J 12 Table 3a. Structural details of trioctahedral brittle micas Table 3b. Structural details of dioctahedral brittle micas Table 4. Structural details of boromuscovite-I M and -2M) calculated from the Rietveld structure refinement by Liang et al. (1995) REFERENCES

Vll

1 3 .4 11 1I 11 19 25 27 37 39 39 .41 43 44 .49 50 51 51 52 52 52 53 54 54 55

55 70 72 74 74 74 76 78 84 84 86 88 88 90

2

Behavior of Micas at High Pressure and High Temperature Pier Francesco Zanazzi, Alessandro Pavese

INTRODUCTION Investigati ve techniques for the study of the thennoelastic behav ior of mi cas p- V and P- V- T equations of state Dioc tahed ral micas Tri oc tahedral mi cas ACKNOWLEDGMENTS REFERENCES

3

99 100 10 1 103 108 ] 14 114

Structural Features of Micas Giovanni Ferraris, Gabriella Ivaldi

INTRODUCTION NOMENCLATURE AND NOTATION MODULARITY OF MICA STRUCTURE The mica module CLOSEST-PACKING aspects Closest-packing and polytypism COMPOSITIONAL ASPECTS SYMMETRY ASPECTS Metric (lattice) symmetry Structural symmetry Symmetry and cation sites Two kinds of mica layer: Ml and M2Iayers The interlayer configuration Possible ordering schemes in the MDO polytypes The phengite case DISTORTIONS The misfit Geo metric parameters describing distortions Ditngonal rotation Other distortions Effects of the distortions on the stacking mode FURTHER STRUCTURAL MODIFICATION Pressure, temperature and chemical influence Thickness of the mica module Ditrigonal rotation and interlayer coordination Effective coordination number (ECoN) CONCLUSIONS APPENDIX I: MICA STRUCTURE AND POLYSOMATIC SERIES Layer silicates as members of modular series ? Mica modules in polysomatic series The heterophyllosicate polysomatic series The palysepiole polysomatic series Conclusions APPENDIX II : OBLIQUE TEXTURE ELECTRON DIFFRACTION (OTED) ACKNOWLEDGMENTS REFERENCES

Vlll

117 1] 7 118 118 ]20 121 122 124 ] 24 124 125 127 128 129 130 130 130 131 131 132 133 135 135 135 137 13 8 138 140 140 140 140 142 143 144 148 148

4

Crystallographic Basis of Polytypism and Twinning in Micas Massimo Nespolo, SlavomiJ Durovic

IN1RODUCTION NOTATION AND DEFINITIONS The mica layer and its constituents Axial settings, indices and lattice parameters Symbols Symmetry and symmetry operations THE UNIT LAYERS OF MICA Alternative unit layers MICA POLYTYPES AND THEIR CHARACTERIZATION Micas as 0D structures SYMBOLIC DESCRIPTION OF MICA POLYTYPES Orientational symbols Rotational symbols RETICULAR CLASSIFICATION OF POLYTYPES: SPACE ORIENTATION AND SYMBOL DEFINITION LOCAL AND GLOBAL SYMME1RY OF MICA POLYTYPES FROM THEIR STACKING SyMBOLS Derivation of MDO polytypes The symmetry analysis from a polytype symbol RELATIONS OF HOMOMORPHY AND CLASSIFICATION OF MDO POLYTYPES BASIC S1RUCTURES AND POLYTYPOIDS. SIZE LIMIT FOR THE DEFINITION OF "POLYTYPE" Abstract polytypes Basic structures _ H1REM observations and some implications IDEAL SPACE-GROUP TYPES OF MICA POLYTYPES AND DESYMME1RIZATION OF LAYERS IN POLYTYPES CHOICE OF THE AXIAL SETTING GEOME1RICAL CLASSIFICATION OF RECIPROCAL LATTICE ROWS SUPERPOSITION S1RUCTURES, FAMILY S1RUCTURE AND FAMILY REFLECTIONS Family structure and family reflections of mica polytypes REFLECTION CONDITIONS NON-FAMILY REFLECTIONS AND ORTHOGONAL PLANES HIDDEN SYMME1RY OF THE MICAS: THE RHOMBOHEDRAL LATTICE TWINNING OF MICAS: THEORY Choice of the twin elements Effect of twinning by selective merohedry on the diffraction pattern Diffraction patterns from twins Allotwinning Tessellation of the hp lattice Plesiotwinning TWINNING OF MICAS. ANALYSIS OF THE GEOME1RY OF THE DIFFRACTION PATTERN Symbolic description of orientation of twinned mica individuals. Limiting symmetry Derivation of twin diffraction patterns Derivation of allotwin diffraction patterns IDENTIFICATION OF MDO POLYTYPES FROM THEIR DIFFRACTION PATTERNS Theoretical background Identification procedure IDENTIFICATION OF NON-MOO POLYTYPES: THE PERIODIC INTENSITY DISTRIBUTION FUNCTION PID in tenns of TS unit layers Derivation of PID from the diffraction pattern

ix

155 156 157 158 158 159 159 160 164 164 172 172 175 178 178 180 180 184 189 191 192 193 193 193 204 206 209 212 213 214 216 217 219 220 223 224 224 230 233 235 237 243 244 244 245 247 249 251

EXPERIMENTAL INVESTIGATION OF MICA SINGLE CRYSTALS FOR TWIN I POLYTYPE IDENTIFICATION Morphological study Surface microtopography Two-dimensional XRD study Diffractometer study APPLICATIONS AND EXAMPLES 24-layer subfamily: A Series I Class b biotite from Ambulawa, Ceylon 8A 2 (subfamily ~ Series O.Class a3) oxybiotit~ from Ruiz Peak, .New Mexico 1M-2M] oxyblOtlte allotwm ZT = 4 from RUiZ Peak, New Mexlco {3,6}[7 {3,6}] biotite plesiotwin from Sambagawa, Japan APPENDIX A. TWINNING: DEFINITION AND CLASSIFICATION APPENDIX B. COMPUTATION OF THE PID FROM A STACKING SEQUENCE CANDIDATE Symlnetry of the PID ACKNOWLEDGMENTS REFERENCES

5

252 252 252 254 256 257 257 258 262 262 267 270 271 272 272

Investigations of Micas Using Advanced Transmission Electron Microscopy Toshihiro Kogure

INTRODUCTION TEMS AND RELATED TECHNIQUES FOR THE INVESTIGATION OF MICA Transmission electron microscopy New recording media for beam-sensitive specimens Sample preparation techniques Image processing and simulation ANALYSES OF POLYTYPES , DEFECT STRUCTURES CONCLUSION ACKNOWLEDGMENTS REFERENCES

6

,

,.281 281 281 286 ,.287 ,288 289 299 310 31 0 310

Optical and Mossbauer Spectroscopy of Iron in Micas M. Darby Dyar

INTRODUCTION OPTICAL SPECTROSCOPY Current instrumentation Review of existing work Sunlmary MOSSBAUER SPECTROSCOPY (MS) Current instrumentation Recoil-free fraction effects Thickness effects Texture effects and other sources of error Techniques for fitting Mossbauer spectra Review of existing Mossbauer data Sumlnary COMPARISON OF TECHNIQUES CONCLUSIONS ACKNOWLEDGMENTS APPENDIX: Other techniques for measurement of Fe 3+/LFe in Micas X-ray ray photoelectron spectroscopy (XPS) Electron energy-loss spectroscopy (EELS) X-Ray absorption spectroscopy (XAS) REFERENCES ,

x

313 315 315 316 320 320 320 320 321 322 323 325 333 334 336 337 337 337 338 338 340

7

Infrared Spectroscopy of Micas Anton Beran

INTRODUCTION LATTICE VIBRATIONS Far-IR region Mid-IR regi on OH STRETCHING VIERATIONS Polarized measurements Quantitative water determination Hydrogen bonding Cation ordering OH-F replacement Dehydroxylati on mechanisms Excess hydroxyl NH4 groups ACKNOWLEDGMENTS REFERENCES

8

351 352 352 353 359 359 360 360 362 365 366 367 367 367 367

X-Ray Absorption Spectroscopy of the Micas Annibale Mottana, Augusto Marcelli, Giannantonio Cibin, and M. Darby Dyar

INTRODUCTION OVERVIEW OF THE XAS METHOD EXAFS XANES Experimental spectra recording Optimizati on of spectra Systematics AC KNOWLEDGMENTS REFERENCES

371 373 375 37 6 384 387 395 .404 .405

9 Constraints on Studies of Metamorphic K-Na White Micas Charles V. Guidotti, Francesco P. Sassi INTRODUCTION EFFECTS OF PETROLOGIC FACTORS ON WHITE MICA CHEMISTRy Important compOSitional vari ations Controls of mica composition by petrologic factors MAXIMIZING INFORMATION FROM MICA STUDIES : SAMPLE SELECTION CONSTRAINTS Petrologic studies Mine ralogic studies DISCUSSION Common failing s in petrology studies Common failings in mineralogy studies "Standard starting points" for the compositional variations of rock-forming dioctahedral and trioctahedral micas ACKNOWLEDGMENTS REFERENCES

Xl

41 3 .41 4 41 4 .41 8 .4 23 4 24 .42 8 440 .44 0 44I 44 1 443 444

10

Modal Spaces for Pelitic Schists James B. Thompson, Jr.

INTRODUCTION NOTATIONS AND CONVENTIONS THE ASSEMBLAGE QUARTZ-MUSCOVITE-BIOTITE-CHLORITE-GARNET. THE ASSEMBLAGE QUARTZ-MUSCOVITE-CHLORITEGARNET-CHLORITOID ASSEMBLAGES CONTAINING CHLORITOID AND BIOTITE OTHER MODAL SPACES ACKNOWLEDGMENTS APPENDIX : INDEPENDENT NET-TRANSFER REACTIONS REFERENCES

11

449 .450 .451 4 54 .455 .458 .458 .460 462

Phyllosilicates in Very Low-Grade Metamorphism: Transformation to Micas Peter Arkai

I.NTRODUCTION MAIN METHODS OF STUDYING LOW-TEMPERATURE TRANS FORMATIONS OF PHYLLOSILICATES XRD techniques TEM techniques ~AIN TRENDS OF PHYLLOSILICATE EVOLUTION AT LOW TEMPERATURE CURRENT PROBLEMS IN STUDYING PHYLLOSILICATE EVOLUTION AT THE LOWER CRYSTALLITE-SIZE LIMITS OF MINERALS REACTION PROGRESS OF PHYLLOSILICATES THROUGH SERIES OF METASTABLE STAGES CONCLUDING REMARKS ACKNOWLEDGMENTS REFERENCES

12

463 464 465 .466 .467 .469 472 .473 .474 .474

Micas: Historical Perspective Curzio Cipriani

INTRODUCTION PRESCIENTIFIC ERA THE EIGHTEENTH CENTURy THE NINETEENTH CENTURy Physical properties Crystallography Chemical composition THE TWENTIETH CENTURY Crystal chemistry Synthesis POLYTYPES SYSTEMATICS CONCLUSIONS REFERENCES APPENDIX I Present-day nomenclature of the mica group and its derivation APPENDIX II Other works consulted in preparation of this historical review XlI

4 79 4 79 .480 .483 4 83 485 .486 491 491 494 494 49 5 .496 497 .498 .499

1

Mica Crystal Chemistry and the Influence of Pressure, Temperature, and Solid Solution on Atomistic Models Maria Franca Brigatti Dipartimento di Scienze della Terra Università di Modena e Reggio Emilia, Via S. Eufemia, 19 I-41100 Modena, Italy [email protected]

Stephen Guggenheim Department of Earth and Environmental Sciences University of Illinois at Chicago 845 West Taylor Street, M/C 186 Chicago, Illinois 60607 [email protected]

OVERVIEW The 2:1 mica layer is composed of two opposing tetrahedral (T) sheets with an octahedral (M) sheet between to form a “TMT” layer (Fig. 1a). The mica structure has a general formula of A M2-3 T4 O10 X2 [in natural micas: A = interlayer cations, usually K, Na, Ca, Ba, and rarely Rb, Cs, NH4, H3O, and Sr; M = octahedral cations, generally Mg, Fe2+, Al, and Fe3+, but other cations such as Li, Ti, V, Cr, Mn, Co, Ni, Cu, and Zn can occur also in mica species; T = tetrahedral cations, generally Si, Al and Fe3+ and rarely B and Be; X = (OH), F, Cl, O, S]. Vacant positions (symbol: †) are also common in the mica structure (Rieder et al. 1998). In the tetrahedral sheet, individual TO4 tetrahedra are linked with neighboring TO4 by sharing three corners each (i.e., the basal oxygen atoms) to form an infinite two-dimensional “hexagonal” mesh pattern (Fig. 1b). The fourth oxygen atom (i.e., the apical oxygen atom) forms a corner of the octahedral coordination unit around the M cations. Thus, each octahedral anion atom-coordination unit is comprised of four apical oxygen atoms (two from the upper and two from the lower tetrahedral sheet) and by two (OH) or F, Cl, O and S anions [the X anions, usually indicated as the OH or O(4) site]. The OH site is at the same level as the apical oxygen but not shared with tetrahedra. In the octahedral sheet, individual octahedra are linked laterally by sharing octahedral edges (Fig. 1c). The smallest structural unit contains three octahedral sites. Structures with all three sites occupied are known as trioctahedral, whereas, if only two octahedra are occupied [usually M(2)] and one is vacant [usually M(1)], the structure is defined as dioctahedral. The 2:1 layers, which are negatively charged, are compensated and bonded together by positively charged interlayer cations of the A site. The layer charge ideally is -1.0 for true micas and -2.0 for brittle micas. Thus, in true micas, the layer charge is compensated by monovalent A cations, whereas in brittle micas it is compensated primarily by divalent A cations. In this section, we consider and discuss the structural and chemical features of more than 200 micas. Most are true micas (146 trioctahedral and 55 dioctahedral). Brittle-mica crystal-structure refinements number about twenty, of which only three are dioctahedral (Tables 1-4, at the end of the chapter). Of the six simple polytypes first derived by Smith and Yoder (1956) and reported by Bailey (1984a, p. 7), only five (i.e., 1M, 2M1, 3T, 2M2, and 2O) have been found and studied by three-dimensional crystal-structure refinements.

1529-6466/02/0046-0001$10.00

DOI: 10.2138/rmg.2002.46.01

2

Brigatti & Guggenheim

Figure 1. (a) The 2:1 layer; (b) the “hexagonal” tetrahedral ring; (c) the octahedral sheet. For site nomenclature see text. a and b are unit cell parameters.

Most of the trioctahedral true-mica structures are 1M polytypes and a few are 2M1, 2M2, and 3T polytypes. In dioctahedral micas, the 2M1 sequence dominates, although 3T and 1M structures have been found. Brittle mica crystal-structure refinements indicate that the 1M polytype is generally trioctahedral whereas the 2M1 polytype is dioctahedral. The 2O structure has been found for the trioctahedral brittle mica, anandite (Giuseppetti and Tadini 1972; Filut et al. 1985) and recently was reported for a phlogopite from Kola Peninsula (Ferraris et al. 2000). The greatest number of the reported structures were refined from single-crystal X-ray diffraction data, with only a few obtained from electron and neutron diffraction experiments. Subsequent sections of this paper present short reviews pertaining to the description of phyllosilicates, an emphasis of the literature since the publication of MICAS, Reviews in Mineralogy, Volume 13, edited by S.W. Bailey (1984a), and a new analysis of the crystal chemistry of the micas. New formulae are presented to clarify how crystal chemistry affects the mica structure. Derivations of these formulae are provided in Appendix I. Also, please refer to other chapters in this volume that cover related topics. For example, see Zanazzi and Pavese for the behavior of micas at high pressure and high temperature.

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

3

Treatment of the data and definition of the parameters used To achieve standardization, all data in Tables 1-4 (Appendix II) were re-calculated from unit-cell parameters and atomic coordinates reported by the authors of the original articles. Information concerning rock type and sample composition was obtained from the original works as well. Suspect refinements are discussed separately or not reported. Of more than 200 reported crystal-structure refinements, about twenty refinements show an agreement factor, R, greater that 9.0%. These structures are considered of poor quality and are not considered further. Several authors used symbols and orientations that differ from convention to describe geometric arrangements of the layer and the stacking sequence of mica polytypes (e.g., Radoslovich 1961; Durovíc 1994; Dornberger-Schiff et al. 1982). To make inter-structure comparisons of features easier, however, it is advantageous to define briefly the site nomenclature adopted and the parameters used to describe and characterize layer geometry. The direction defined by the stacking of 2:1 units defines the [001] direction (i.e., the c axis), whereas the periodicity of the infinite two-dimensional sheets defines [100] and [010] directions (i.e., a and b translations). The actual value of the repeat distance in the [001] direction, as well as lateral a and b parameters, depends on several factors, such as the layer stereochemistry and polytypism (i.e., c ∼ 10 Å × n, where n identifies the number of layers involved in the stacking sequence). The sitenomenclature scheme adopted here starts from the nomenclature generally used for the 2:1 layer of the 1M polytype in C2/m symmetry: T denotes the four-coordinated site, M(1) and M(2) indicate six-coordinated sites with (OH) groups in trans- and cisorientation, respectively, A refers to the interlayer cation, O(1) and O(2) represent the basal tetrahedral oxygen atoms, O(3) is the apical oxygen atom, and O(4) refers to the (OH), F, Cl, S or O anions (Fig. 1a). The number of sites per unit cell is: T = 8; M(1) = 2; M(2) = 4; A = 2; O(1) = 8; O(2) = 4; O(3) = 8; O(4) = 4. The site nomenclature for other structural variants can be derived from this nomenclature by changes that relate to spacegroup differences and to the number of 2:1 layers per unit cell. The definition of parameters reported in Tables 1-4 (Appendix II) follows. For a more extensive review on definition and structural significance of these parameters, see Bailey (1984b) and references therein. Cation-anion bond lengths: (i) tetrahedral 〈T–O〉; (ii) octahedral 〈M–O,OH,F,Cl,S〉 for both M(1) and M(2) sites; and (iii) interlayer 〈A–O〉. Mean bond lengths were compared to those of the original papers and vacant-site distances determined (i.e., vacancy-to-anion distances). The tetrahedral Oapical–T–Obasal angles were used to obtain the tetrahedral flattening angle, τ = ∑3i=1 Oapical–T–Obasal)i/3. The internal angles of the tetrahedral ring were used to determine the tetrahedral rotation angle, α = ∑6i=1 α i / 6 where αi = |120° – φi|/2 and φi is the angle between basal edges of neighboring tetrahedra articulated in the ring. Basal oxygen-plane corrugation, Δz, was determined by Δz = (zObasal(max) – zObasal(min)) × c sinβ. The thickness of the tetrahedral and octahedral sheets was calculated from oxygen z coordinates of each polyhedron, including the OH group (or other X anions). The interlayer separation was obtained by considering the tetrahedral basal oxygen z coordinates of adjacent 2:1 layers. The octahedral flattening angle ψ was calculated from

4

Brigatti & Guggenheim ⎛ octahedral thickness ⎞ ψ = cos −1 ⎜ ⎟ ⎝ 2 × M − O, OH, F, Cl, S ⎠

Tetrahedral cation atomic coordinates, taken from the original reference, were transformed from fractional to Cartesian to calculate the Layer Offset, the Intralayer Shift, and the Overall Shift. The Layer Offset is based on the displacement of the tetrahedral sheet across the interlayer from one 2:1 layer to the next, which should be equal to zero in the ideal mica structure. The Intralayer Shift is the over-shift of the upper tetrahedral sheet relative to the lower tetrahedral sheet of the same 2:1 layer. The Overall Shift relates to both effects. In true micas, the tetrahedral mean bond distance varies from 1.57(1) Å in boromuscovite-2M1 (Liang et al. 1995; Table 4) to 1.750(2) Å in an ordered (Al vs. Si) ephesite-2M1 (Slade et al. 1987; Table 1d); in brittle micas, the 〈T–O〉 mean bond distance varies from 1.620(2) to 1.799(2) Å, both values are from anandite-2O (Filut et al. 1985; Table 3a). Octahedral mean bond length ranges from about 1.882(1) Å in an ordered ferroan polylithionite-1M (Guggenheim and Bailey 1977; Brigatti et al. 2000b; Table 1b) to 2.236(1) Å in anandite 2O (Filut et al. 1985; Table 3a). The radius of the vacant M(1) site in dioctahedral micas (〈M(1)–O〉) varies from 2.190 to 2.259 Å. The shortest 〈A–O〉inner distance occurs in clintonite (〈A–O〉inner = 2.397(2) Å; Alietti et al. 1997, Table 3a), whereas the longest distance occurs in nanpingite and synthetic Cs-tetra-ferri-annite (〈A– O〉 inner of ∼ 3.370 Å; Ni and Hughes 1996 and Mellini et al. 1996, Tables 1c and 1a, respectively). These data show the great variability in bond distances which may be ascribed not only to the local composition but also to the constraints of closest packing within the layer and the confinement of the octahedra between two opposing tetrahedral sheets. We consider the compositional and topological relationships in the following analysis. End-member crystal chemistry: New end members and new data since 1984 Boromuscovite. Boromuscovite was first reported by Foord et al. (1991). The mineral, precipitated from a late-stage hydrothermal fluid (T: 350-400°C; P: 1-2 kbar), occurred in the New Spaulding Pocket, Little Three Mine pegmatite dike (Ramona district, San Diego County, California), as a fine-grained coating of quartz, polylithionite, microcline and topaz. The mineral was found also in elbaite pegmatite at Recice near Mové Mesto na Morave, western Moravia, Czech Republic (Liang et al. 1995; Novák et al. 1999). Relatively high B contents were also reported for muscovite and polylithionite from polylithionite-rich pegmatites of Rozná and Dobrá Voda, Czech Republic (Cerny et al. 1995), for polylithionite-2M1 from Recice (Novák et al. 1999), and for muscovite from metapegmatite at Stoffhütte, Koralpe, Austria (Ertl and Brandstätter 1998). Boromuscovite (Foord et al. 1991) has the general structural formula of KAl2 (Si3B) O10(OH)2, in which [4]Al is replaced by [4]B relative to muscovite. The composition of Little Three Mine boromuscovite is (K0.89Rb0.02Ca0.01)(Al1.93Li0.01Mg0.01)(Si3.06B0.77Al0.17) O9.82F0.16(OH)2.02, whereas the composition of Recice boromuscovite shows a slightly lower [4]B content: (K0.89Na0.01)(Al1.99Li0.01)(Si3.10B0.68Al0.22)O10F0.02(OH)1.98. The unit cell parameters, very similar in natural and synthetic crystals (Schreyer and Jung 1997), are significantly smaller than those reported for muscovite (a = 5.075(1), b = 8.794(4), c = 19.82(3) Å, β = 95.59(3)° and a = 5.077(1), b = 8.775(3), c = 10.061(2) Å, β = 101.31(2)° for Little Three Mine boromuscovite-2M1 and boromuscovite-1M, respectively). A boromuscovite structure determination is complicated by the fine-grained nature

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

5

of the mineral and by the presence of the mixture of 1M and 2M1 polytypes. Nonetheless the crystal-structure determination of a mixture of 83 wt % boromuscovite-2M1 and 17 wt % boromuscovite-1M from Recice was attempted using a coupled Rietveld-staticstructure energy-minimization method (Liang et al. 1995). Although the high standard deviation for calculated parameters suggests caution in the analysis of crystal chemical details, Liang et al. (1995) indicated that: (i) boron is uniformly distributed between the two polytypes, (ii) 〈T–O〉 distances correspond well with the B-content at the corresponding T-sites, namely 〈T–O〉 distances linearly decrease as B occupancy increases, and (iii) in the 2M1 polytype, slight differences between 〈T(1)–O〉 and 〈T(2)– O〉 distances may imply a B preference for the T(1) site (Table 4). The 11B MAS NMR spectra showed a single, symmetric and narrow line (about 150 Hz wide) at 20.7 ppm. The width was interpreted as possibly relating to the coordination for B with a nearsymmetrical disposition of anions (Novák et al. 1999). Clintonite. Clintonite is the trioctahedral brittle mica with ideal composition of Ca(Mg2Al)(SiAl3)O10(OH)2. This structure violates the Al-avoidance principle of Loewenstein (1954). It crystallizes in H2O-saturated Ca-, Al-rich, Si-poor systems under wide P-T conditions. Clintonite, usually found in metasomatic aureoles of carbonate rocks, is rare in nature because crystallization is limited to environments characterized by both alumina-rich and silica-poor bulk-rock chemistry and very low CO2 and K activities (Bucher-Nurminen 1976; Olesch and Seifert 1976; Kato et al. 1997; Grew et al. 1999). The 1M polytype and 1Md sequences are the most common forms. The 2M1 form is rare (Akhundov et al. 1961) and no 3T forms have been reported. Many 1M crystals are twinned by ±120° rotation about the normal to the {001} cleavage. Such twinning causes extra spots on precession photographs that simulate an apparent three-layer periodicity (MacKinney et al. 1988). Subsequent to an extensive review of brittle micas (Guggenheim 1984), additional crystal-chemical details of clintonite-1M (space group C2/m) were reported by MacKinney et al. (1988) and Alietti et al. (1997). These studies confirmed that natural clintonite crystals do not vary extensively in composition: (i) the octahedral sites contain predominant Mg and Al with Fe2+ to ≤7% of the octahedral-site occupancy; (ii) the extent of the substitution [4]Al-1 [6]Mg-2 [4]Si [6](Al, ), which involves the solid solution of trioctahedral with dioctahedral Ca-bearing brittle micas, is very limited; (iii) Fe3+ content involves tetrahedral site occupancy, but at low ( NH4 and with (001) spacing values intermediate between illite and tobelite are referred to as “NH4-rich illite.” They occur in hydrothermal environments (Sterne et al. 1982; Higashi 1982; Von Damm et al. 1985; Wilson et al. 1992; Bobos and Ghergari 1999); in black-shales (Sterne et al. 1984); in regionally metamorphosed carbonaceous pelites (Juster et al. 1987; Daniels et al. 1996; Liu et al. 1996) and in diagenetic environments (Duit et al. 1986; Lindgreen et al. 1991; Drits et al. 1997). Tobelite-like layers are often found in interstratified dioctahedral minerals having non-expandable (mica-like) and expandable (smectite-like and/or vermiculite-like) layers. Drits et al. (1997) demonstrated that, in interstratified illite-smectite minerals from North

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

11

Sea oil-source rocks, the mica-like component contains both K-rich end-member (illite) and NH4-rich end-member (tobelite) layers. The amount and the distribution of fixed K and NH4 was determined by a peak profile-fitting procedure on experimental powder (X-ray) diffraction features (Drits et al. 1997; Sakharov et al. 1999). Synthetic micas with unusual properties Cesian tetra-ferri-annite and cesian annite. Fe-rich micas have the capacity to contain radioisotopes, such as 135Cs and 137Cs. The study of these materials has been a promising direction of mica research over the last few years; see, for example, Mellini et al. (1996), Drábek et al. (1998), and Comodi et al. (1999). The cesian-tetra-ferri-annite crystal structure was studied by Mellini et al. (1996) and by Comodi et al. (1999) at ambient conditions and at high P-T conditions. Cs-tetra-ferri-annite crystallizes in the 1M polytype (C2/m space group). It has the largest unit-cell volume reported to date for 1M micas and coordination polyhedra are undistorted (Table 1a). The tetrahedral rotation angle (α = 0.2°), and the octahedral-distortion parameter, δ, involved with the counterrotation of upper and lower oxygen triads are near 0° (δM(1) = 0; δM(2) = 0.2°), thus suggesting a nearly undistorted layer with limited internal strain. No detectable internal strain based on such parameters (e.g., α and δ) was observed at high pressure (to 47 Kbar) and temperature (to 582°C). Above 450°C, in air, the reduction of the unit cell volume is related to the loss of H atoms required to balance the layer charge after oxidation of octahedral iron in the M(2)-cis site. Li for K exchange in interlayer sites. Volfiger and Robert (1979, 1980) and Robert et al. (1983) suggested that, in synthetic trioctahedral micas, anhydrous Li can exchange for K in interlayer sites. Although the crystal quality obtained from the run products did not allow a complete crystal structure determination, they indicated, on the basis of the results obtained by infrared and powder X-ray analyses, that Li is located in the interlayer in a pseudo-octahedral cavity. This cavity is partly defined by the hexagonal ring of one layer and by the basal oxygen atoms of two tetrahedra in the adjacent layer. The Li solubility limit was estimated to be a function of the relation: Li/(Li+K)max = 2 [4][Al/(Al+Si)]2. Tetrahedral Al for Si substitution is essential to minimize the electrostatic repulsion between tetrahedral cations and Li, and therefore to create favorable cavities to host Li. Robert et al. (1983) found that the unit cell parameter, c, decreases with K for Li substitution whereas the b parameter slightly increases. In Li-exchanged synthetic paragonite-2M1 and muscovite-2M1, repulsive forces between O atoms across the interlayer region cause an interlayer overshift, resulting in an anomalously high basal spacing and smaller monoclinic β angle (Keppler 1990). Complete and rapid Li exchange in the interlayer sites was obtained for natural phlogopite, ferroan phlogopite and muscovite using “cryptand [222]” as a complexing agent, and dioxane as a solvent (Bracke et al. 1995). Powder X-ray diffraction suggests that the interlayer spacing changes with replacement of K by Li + H2O. The original reflection at 9.93 Å loses intensity progressively and an additional reflection at 11.78 Å appears. EFFECT OF COMPOSITION ON STRUCTURE Tetrahedral sheet In some naturally occurring true micas, Si nearly fills all the tetrahedral sites (e.g., polylithionite, tainiolite, norrishite, and celadonite), whereas in the most common mica species (i.e., muscovite and phlogopite) Al substitutes for Si in a ratio near 1:3. In some true micas and brittle micas, the Al for Si substitution corresponds to a ratio of Al:Si = 1:1 (e.g., ephesite, preiswerkite, siderophyllite, margarite, and kinoshitalite), whereas the

12

Brigatti & Guggenheim

trioctahedral brittle mica, clintonite, has an unusually high Al content with a ratio of Al:Si of 3:1 (Bailey 1984a,b). Evidence of Fe3+ tetrahedral substitution was reported on the basis of optical observations (e.g., Farmer and Boettcher 1981; Neal and Taylor 1989), spectroscopic studies (e.g., Dyar 1990; Rancourt et al. 1992; Cruciani et al. 1995) and crystal-structure refinement (Guggenheim and Kato 1984; Joswig et al. 1986; Cruciani and Zanazzi 1994; Brigatti et al 1996a, 1999; Medici 1996). However, only in tetra-ferriphlogopite, tetra-ferri-annite and anandite is Fe3+ the only Si-substituting cation, with a Fe3+:Si ratio near 1:3 (e.g., Giuseppetti and Tadini 1972; Semenova et al. 1977; Hazen et al. 1981; Filut et al. 1985; Brigatti et al.1996a,b, 1999; Mellini et al. 1996). Thus, the 1:3 ratio appears to be the greatest Fe3+ tetrahedral substitution possible for the micas. Two mica end-members contain B (boromuscovite; Liang et al. 1995) and Be (bityite; Lin and Guggenheim 1983), and some synthetic micas contain Ge in the T site (Toraya and Marumo 1981; Toraya et al. 1978a,c). Most mica structures display a disordered distribution of tetrahedral cations, with the exception of some brittle mica species, such as margarite (Guggenheim and Bailey 1975, 1978; Kassner et al. 1993), anandite (Giuseppetti and Tadini 1972; Filut et al. 1985) and bityite (Lin and Guggenheim 1983) and a few true micas (e.g., polylithionite-3T, Brown, 1978; muscovite-3T, Güven and Burnham 1967). Some true micas with an apparent ordered distribution of cations in the tetrahedra are those with a high R value and therefore these structures should be considered tentative. Hazen and Burnham (1973) related 〈T–O〉 distances of trisilicic micas to tetrahedral composition by the linear relationship (xAl and xSi represent Al and Si apfu, respectively) ⎛ x Al 〈T − O〉 ( A ) = 0.163 ⋅ ⎜ ⎝ x Al + xSi

⎞ ⎟ + 1.608 ⎠

A more general relationship derived here including both trioctahedral and dioctahedral true and brittle micas (Tables 1-4, Appendix II) between tetrahedral mean bond distances 〈T–O〉 and tetrahedral chemistry (in apfu) is: T − O (Å) = 1.607 + 4.201 ⋅ 10 −2

[4]

Al + 7.68 ⋅10−2[4] Fe

(correlation coefficient, r = 0.965) In the regression analysis, structures containing B, Be, and Ge in tetrahedral sites were not considered, as well as structures with symmetry lower than ideal owing to tetrahedral cation ordering (differences in 〈T–O〉 values greater than 5σ). Only structures containing tetrahedral Si, Al, and Fe were examined. Geometrical considerations of tetrahedral distortion parameters have been considered earlier (e.g., Drits 1969, 1975; Takéuchi 1975; Appelo 1978; Lee and Guggenheim 1981; Weiss et al. 1992). We further discuss these relationships here and relate them to layer composition on the basis of data from a large number of structure determinations. A crystal chemical study of the τ parameter is complex. In an ideal tetrahedron τ is equal to arcos (-1/3) ≅ 109.47°. For non-ideal cases, however, τ was found to be affected by tetrahedral content, increasing as Si increases (Takéuchi 1975) relative to Al. The τ value can deviate from its ideal value as a function of the relative position along c for the basal oxygen atoms with respect to the tetrahedral cation and with respect to the mean basal-edge length and the mean tetrahedral-edge value. These conclusions are based on the linearized topology of the tetrahedron. Several simple models of deformation are considered here (Fig. 2) and only modes (3) and (4) were found to affect the τ value. All dependences (over displacement from an ideal undeformed configuration) of order

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

13

Figure 2. Geometrical considerations over the dependence of τ from tetrahedron vertex and center displacement. The relationships in the legend have been obtained from a linearized geometrical model. k and e indicate the displacement and the tetrahedron edge length, respectively.

greater than one are ignored. The model, thus, provides results in good agreement with structural data only if displacements are small relative to the characteristic length of the system (i.e., the tetrahedral edge). Figure 3 shows the variations of τ vs. [4]Si content. Although the increase of τ with Si is confirmed, there are two different linear trends, one trend for true and one trend for brittle micas. Brittle micas show τ values greater than expected if just the composition of the tetrahedron is considered. Although this simple model ignores cation ordering, on the basis of geometrical considerations derived before (Fig. 2), the higher τ values may be explained by the increase in the electrostatic attraction of basal oxygen atoms by the high-charge interlayer cation and by the concomitant increase in repulsion between the interlayer cation and the tetrahedral cation. Note, for example, that kinoshitalite usually tends to approach true micas in composition. Samples of kinoshitalite and ferrokinoshitalite (Guggenheim and Kato 1984; Brigatti and Poppi 1993; Guggenheim and Frimmel 1999) contain significant amounts of monovalent K in substitution for Ba, whereas, kinoshitalite refined by Gnos and Armbruster (2000), marked by an arrow in Figure 3, has nearly complete interlayer Ba occupancy and a larger τ value. [4]

To better relate how the interlayer cation affects τ, we have developed a simple electrostatic model. The model is comprised of four tetrahedral oxygen atoms, with the tetrahedral and the interlayer cations located at the center of the tetrahedron and in the

14

Brigatti & Guggenheim

Figure 3. Relationships between the tetrahedral flattening angle, τ, and Si content in tetrahedral coordination as determined by microprobe analysis. Symbols used: filled circle = annite; filled circle, x-hair = magnesian annite; open circle = phlogopite; open circle, x-hair = ferroan phlogopite; filled circle, dotted = tetra-ferri-annite; open circle, dotted = tetra-ferriphlogopite; open square = polylithionite; filled square = trilithionite; filled square, x-hair = siderophyllite; open square, x-hair = ferroan polylithionite; filled hexagon, x-hair = norrishite; crosses = preiswerkite; open diamond = muscovite; open diamond, xhair = nanpingite; filled diamond = paragonite; filled diamond x-hair = boromuscovite; open triangle up = clintonite; filled triangle up, x-hair = ferrokinoshitalite; filled triangle up = kinoshitalite. The sample arrowed is kinoshitalite by Gnos and Armbruster (2000). For details see text.

middle of the interlayer, respectively. The oxygen atoms were placed at the vertices of an undistorted tetrahedron with a tetrahedral volume equal to that as considered above. A uniform displacement along the [001] direction was then imposed on the basal oxygen atom plane and the electrostatic energy associated with the system was then derived as a function of this displacement. Finally, the displacement which minimizes the electrostatic energy of the system was calculated and compared with the value obtained for a system identical to that described, but differing in the formal charge of the interlayer cation which was arbitrarily set equal to one. Therefore, the model takes into consideration the differences in energy between the two configurations described, not the total energy. The displacement obtained was used to “isolate” the τ value from the influence of the divalent interlayer cation. The τ values of tetrahedrally disordered brittle micas which was thus “isolated” (i.e., τ*) follow the same trend defined for true micas, confirming the influence of interlayer cations on τ (Fig. 4). Unlike other models reported in the literature (e.g., Giese 1984), our model introduces only the Coulombic term and does not consider the repulsive energy or van der Waals interactions. This simplification, as Giese (1984) correctly noted, does not produce correct energy values. For this reason, energy differences between structural systems, which are characterized by the same repulsive energy, were considered. The charge at each position was determined from chemical data and from structural constraints.

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

Figure 4. Relationship between τ* and Si tetrahedral content . τ* refers to the τ value “isolated” from the influence of the interlayer cation for the brittle micas clintonite and kinoshitalite. Regression equation: τ* (°) = 2.920 × [4]Si + 101.98, r = 0.950. Symbols and samples as in Figure 3.

Figure 5. Bond energy between tetrahedral cation and tetrahedral basal oxygen atoms compared with the bond energy between interlayer cation and tetrahedral basal oxygen atoms. Symbols and samples as in Figure 3.

15

16

Brigatti & Guggenheim

Figure 5 relates the bond energy between the tetrahedral basal oxygen atoms vs. tetrahedral cations (〈T–Obasal〉) and between the basal oxygen atoms vs. interlayer cations (〈A–Obasal〉), respectively. In brittle mica species, the distance between the tetrahedral cation and the basal oxygen atom plane increases, owing to the interaction with the interlayer cation. In this way the increase in 〈T–Obasal〉 bond energy is partly compensated by a decrease in bond energy between the cation and the oxygen atoms of the basal plane. The displacement of the tetrahedral cation from its ideal position can be evaluated (see Appendix I for derivation) from the tetrahedral displacement parameter, Tdisp.: Tdisp. =

T − Obasal

2

O − Obasal − 3

2

−

(T − O ) apical

3

Tdisp. was calculated for all structures starting from observed distances, and then plotted against the τ value observed (Fig. 6).

Figure 6. Mean τ value vs. the displacement of the T cation from the center of the tetrahedron mass (Tdisp.). Symbols and samples as in Figure 3.

The plane of basal oxygen atoms approaches the tetrahedral cation in flattened tetrahedra (the distance between the tetrahedral cation and the basal oxygen-atom plane decreases with respect to the T–Oapical distance), whereas the tetrahedral cation shifts toward the tetrahedral apex (the distance between the tetrahedral cation and basal-oxygen atom plane increases with respect to the T–Oapical distance) in elongated tetrahedra. In preiswerkite and in boromuscovite the tetrahedral cation shifts from its ideal position toward the plane of basal oxygen atoms (τ < 109.47°). In the brittle mica clintonite, the tetrahedral cation more closely approaches the center of the tetrahedron (τ ≈ 109.47°), whereas in other micas the cation shifts toward the tetrahedral apex (τ > 109.47°). The maximum shift was observed in norrishite (Tyrna and Guggenheim 1991) and in polylithionite (Takeda and Burnham 1969).

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

17

Figure 7. Plot of τ vs. 〈O-O〉basal. Symbols and samples as in Figure 3.

In addition, τ reflects an adjustment for the misfit between the tetrahedral sheet and the octahedral sheet (the regression coefficient, r, of τ vs. the difference between mean basal tetrahedral edges and mean octahedral triads is r = 0.92). Furthermore, as the mean 〈O–O〉 basal distance decreases, the tetrahedral cation moves away from the basal oxygen-atom plane. Thus, τ increases in value (Fig. 7). The deviation of the parameters for clintonite and kinoshitalite from the trend for true micas further suggests that there is a significant influence of the interlayer cation on the value of τ. In conclusion (i) τ increases as the distance between the tetrahedral cation and the basal oxygen-atom plane increases from its ideal value; (ii) τ increases as 〈O–O〉basal decreases, thus reflecting a dimensional adjustment between the tetrahedral sheet and octahedral sheet; and (iii) τ increases with [4]Si content. Differences between τ values of brittle micas from the true micas are related in part to electrostatic features. It is useful to understand why the tetrahedral cation moves from its ideal position. Drits (1969) stated that “the position of the tetrahedral cation depends not only on the degree of substitution of Si by Al in the tetrahedra (Brown and Bailey 1963), but also in the position and distribution in compensating positive charges.” This assumption is related to electrostatic forces in the following way (see Appendix I for derivation): ΔE1 =

−3 ⋅ q T ⎛ −9⋅ q T ⎞ −⎜ ⎟ d TπOb ⎝ T − O apical ⎠

⎛ ΔE 2 = ⎜⎜ ⎜ ⎝

ΔE3 =

q T ⋅ (q A / 4 )

(IS / 2 + d TπOb )2 +

O apical − Oapical

2

⎞ ⎛ ⎟ −⎜ ⎟ ⎜ ⎟ ⎜ ⎠ ⎝

⎛ q 2T q 2T −⎜ IS + 2 ⋅ d TπOb ⎜⎝ IS + 2 / 3 ⋅ T − Oapical

(

⎞ ⎟ ⎟ ⎠

)

q T ⋅(q A / 4)

(IS / 2 + (T − O )/ 3) + O 2

apical

apical

− Oapical

2

⎞ ⎟ ⎟ ⎟ ⎠

18

Brigatti & Guggenheim

where qT and qA are the tetrahedral and interlayer charges, respectively; IS is the interlayer separation; dTπOb is the distance of the tetrahedral cation from the basal oxygen atom plane; 〈Oapical–Oapical〉 is the distance between apical oxygen atoms; and T–Oapical is the distance between the tetrahedral cation and apical oxygen atom. E1 relates the electrostatic energy between the tetrahedral cation and the basal oxygen atoms. E2 is the electrostatic energy between the tetrahedral cation and interlayer cation. E3 considers the repulsion between tetrahedral cations of two opposing tetrahedra across the interlayer (Fig. 8). ΔE1 (ΔE2, ΔE3) is the variation of E1 (E2, E3) values in the actual structure and in an ideal structure with the tetrahedral cation ideally spaced from the basal and apical oxygen atoms. ΔE1, ΔE2, and ΔE3 were derived by considering the set of charges represented in Figure 9. This specific arrangement of charges was developed to describe the electrostatic interactions between the basal oxygen atoms of the tetrahedron and interlayer cation. All planes of atoms (i.e., the plane of interlayer cations, the plane of basal oxygen atoms and the plane of tetrahedral cations) can be described through a rigid displacement of the simple charge distribution in Figure 9, thus the energy involving the oxygen-atom plane differs, to a first approximation, from the energy related to the distribution in Figure 9 by just a scale factor. The objective of our model is to describe the factors influencing the interlayer cation displacement from its “ideal” position. However, we consider the difference in energy between the actual structure configuration and that characterized by a tetrahedral cation-basal oxygen atom plane distance, which is equal to (T–Oapical)/3. All terms in energy which do not include that distance, are therefore excluded in this derivation because they must be equal in both the configurations considered. In conclusion, differences in energy among configurations which vary for very small displacements of charge can be very useful. Our model considers van der Waals and repulsion energies equal in both configurations to simplify the calculation.

Figure 8. Relationship between ΔE2 + ΔE3 vs. ΔE1. For the definition of energy E1, E2, and E3, see text. Regression equation [(ΔΕ2 + ΔΕ3) = -1.099 ΔΕ1 + 1.26 × 10-3 ; r = 0.997). Symbols and samples as in Figure 3.

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

19

Figure 9. The set of charges used to derive ΔE1, ΔE2, and ΔE3.

Figure 8 clearly shows that an increase in the electrostatic energy associated with an increase in the tetrahedral cation-basal oxygen atom distance is compensated by a reduction in the repulsion between the interlayer cation and the tetrahedral cation and the tetrahedral-tetrahedral cations (sited in adjacent layers). Given the high correlation coefficient (r = 0.997), the relation may be useful as a predictive tool. The basal oxygen atom plane corrugation effect (Δz) produces an out-of-plane twisting of tetrahedra about the bridging basal oxygen atom in the [110] tetrahedral chain and a shortening of the distance between apical oxygens along the octahedral edge parallel to the (001) plane. Lee and Guggenheim (1981) demonstrated that the corrugation of the basal oxygen atom plane reflects differences in distance between apical oxygen atoms linked to octahedra of different size. Thus Δz is limited in trioctahedral micas with M(1) ≈ M(2) in size, whereas it shows higher values in dioctahedral micas with M(1) >> M(2) in size. Differences in Δz are related to the linkage of the tetrahedral sheet by apical oxygen with octahedral sites different in size. A strong relationship between Δz and ΔM [ΔM = 〈M–O〉max – 〈M–O〉min] for a structure is evident in Figure 10. This result confirms that differences in octahedral site dimensions play an important role over tetrahedral basal oxygen-plane corrugation [regression equation: Δz (Å) = 0.647 × ΔM; r = 0.984]. Figure 11 shows the effect of Al octahedral content ([6]Al) on Δz. Where [6]Al occupancy is less than 1 apfu, Δz is approximately zero (trioctahedral true and trioctahedral brittle micas). In trioctahedral Li-rich micas (polylithionite, trilithionite and siderophillite) and in preiswerkite, [6]Al occupancy is nearly 1 apfu and Δz is as large as 0.15 Å. A Δz of ≤0.24 Å is observed for dioctahedral micas for which [6]Al occupancy reaches 2 apfu. Al is a cation of relatively small size. For micas with significant amounts of octahedral Al and where Al ordering occurs, differences in size between octahedral sites are enhanced and the value of Δz increases. Such differences also occur for micas with a low charge cation (e.g., Li+ in trioctahedral polylithionite) or by vacancies (i.e., in dioctahedral micas), where charge balance occurs within the octahedral sheet only. Tetrahedral rotation and interlayer region The dimensions of an ideal octahedral sheet in the (001) plane are commonly less than those of an ideal and unconstrained tetrahedral sheet. Thus, to obtain congruence, the difference in size of the tetrahedral and octahedral sheets must be adjusted by any one or more of the following: (i) in-plane rotation of adjacent tetrahedra in opposite directions about c* (parameter α); (ii) thickening of the tetrahedra (parameter τ), and (iii) a flat-

20

Brigatti & Guggenheim

Figure 10. Relationship between the tilting of the basal oxygen plane, Δz and ΔM [ = 〈M–O〉max – 〈M–O〉min ]. Regression equation: Δz (Å) = 0.647 × ΔM; r = 0.984. Symbols and samples as in Figure 3.

Figure 11. Δz (Å) vs. the octahedral Al content determined by microprobe analysis. The arrow indicates the dioctahedral chromiumrich mica (Evsyunin et al. 1997) which presents an unusual chemical composition characterized by an important [6]Cr for [6]Al substitution. Symbols and samples as in Figure 3.

tening of the octahedra (parameter ψ) to lengthen the octahedral edges (Mathieson and Walker 1954, Newnham and Brindley 1956; Zvyagin 1957; Bradley 1959; Radoslovich 1961; Radoslovich and Norrish 1962; Brown and Bailey 1963; Donnay et al. 1964, Bailey 1984b, Lee and Guggenheim 1981). McCauley and Newnham (1971) specified by multiple regression analysis that, although the α value is largely controlled by the tetrahedral-octahedral sheet lateral misfit (∼90%), it also reflects the field strength of the interlayer cation. Toraya (1981) observed

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

21

two linear relationships, between α and the difference in length of the octahedral and tetrahedral sheets along the b axis, i.e., α = c1 (2√3eb -3√2do) + c2 (where do is the mean octahedral cation-anion distance, eb is the mean basal edge length of a tetrahedron, 2√3eb and 3√2do are the lengths along the b axis, in the idealized form, of the tetrahedral and octahedral sheet, respectively; c1 and c2 are the regression coefficients, c1 = 35.44 and 12.58, c2 = -11.09 and 4.30 for silicate and germanate micas, respectively). Weiss et al. (1992) used a different dataset and the same assumption of Toraya and found, for Si-rich micas, different values for c1 and c2 (c1 = 25.9 and c2 = -5.0).

Figure 12. α determined by structure refinement vs. α calculated by regression equation α = 25.9 (2√3eb -3√2do) – 5.0 (Weiss et al. 1992). Symbols as in Figure 2. The plot reports only structures published after 1992, i.e., structures not considered in the predictive equation of Weiss et al. (1992).

The calculated α values using the equation of Weiss et al. (1992) and data published after 1992 (i.e., not used to derive the equation) vs. observed α values are consistent mostly in the range of 7-9°, whereas the correspondence is lower for smaller and larger angles (Fig. 12). Weiss et al. (1992) also predicted the α value from sheet composition using a vector-representation grid. They calculated a mean tetrahedral distance, d (T–O), and a mean octahedral bond distances, d (M–A) (where A is any anion), from equations d (T–O) = Σ di (T–O)calc × xI d (M–O) = Σ di (M–O)calc × xi d (M–OH) = Σ dI (M–OH)calc × xi where di (T–O), di (M–O) and di (M–OH) are the calculated mean bond lengths for cations in tetrahedral and octahedral coordination, respectively, and xi represents the atomic fraction of each cation. To better understand the role of tetrahedral-octahedral lateral misfit for 1M, 2M1,

22

Brigatti & Guggenheim

2M2 and 3T polytypes, we have developed a geometric model (see Appendix I). According to this model, α is equal to ⎛1 / 3 + k ⋅ 4 / 3 − k 2 ⎞ α = tan −1 ⎜ ⎟ − 60 k 2 −1 ⎝ ⎠

and simplifying ⎛ 3 ⎞ α = cos −1 ⎜ ⋅ k⎟ ⎝ 2 ⎠

where k is the ratio between the 〈O–O〉 octahedral triads (〈O–O〉unshared) and 〈O–O〉 tetrahedral basal edges, 〈O–O〉basal. 〈O–O〉unshared very closely corresponds to b/3, b√3/3 and a/3 for trioctahedral-1M (and -2M1), -3T, and -2M2 polytypes, respectively, thus indicating that the “rigid” octahedral sheet primarily determines the unit-cell lateral dimensions of trioctahedral micas. This relationship is obtained with “rigid” tetrahedra and deformation involves only the “hexagonal” silicate ring. Therefore, the deformation of the octahedron and tetrahedron influences the α value only by affecting the tetrahedral and octahedral lengths as given in the formula above. Figure 13 reports α values observed vs. α values thus calculated. The correspondence appears excellent (r = 0.994), although the model could be improved by calculating all the six-ring tetrahedral angles and then averaging. The relationship between αcalculated and k is not linear and that structures which primarily deviate are Li-rich micas with octahedral ordering in the M(2) and M(3) sites. This geometric relationship is useful also to evaluate the influence of composition over α. The mean basal tetrahedral edge depends on tetrahedral cation stereochemistry

Figure 13. α determined by structure refinement vs. α calculated from the equation: α = cos-1 ( 3 2 ⋅k ) where the k is the ratio between octahedral triads (〈O–O〉unshared) and tetrahedral basal edges(〈O–O〉basal). Symbols and samples as in Figure 3.

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

23

([4]Al and [4]Fe in apfu, r = 0.970) by: 〈O − O〉 basal = 2.581+ 8.836 ⋅10 −2 ×

[4 ]

Al + 0.164

[4 ]

Fe

whereas the mean length of the octahedral triads is well fitted by the following expression ([6]Al, [6]Fe2+ in apfu, r = 0.940): 〈O − O〉 unshared = 3.072 − 4.24 ×10 −2 ×

[ 6]

Al + 2.14 ×10 −2 ×

[ 6]

Fe 2+ − 3.88 ×10 −2 × Ifs

where Ifs is the increase of the interlayer cation field strength (i.e., the charge of the interlayer cation divided by radius) in brittle micas with respect to true micas (both regression equations were obtained using chemical data reported in Tables 1-4, Appendix II). Note that the octahedral site composition is represented in a less accurate way than the tetrahedral composition because of the greater variability in the chemical composition of the octahedron. The relationship between α-observed and that calculated from composition is shown in Figure 14. The fit is fair (r = 0.922) and this low correlation is related to the influence of octahedral, tetrahedral, and interlayer composition on α. In particular, the interlayer composition appears to affect the mean value of the octahedral triads (unshared O–O distances). This result confirms the influence of the interlayer site composition on the tetrahedral in-plane rotation.

Figure 14. α determined by single crystal structure refinement vs. α calculated by the formula α = cos-1 ( 3 2 ⋅k ) where k value was obtained by calculating 〈O–O〉 octahedral unshared edges (〈O–O〉unshared) and 〈O–O〉 tetrahedral basal edges (〈O–O〉basal) from chemical composition (see text). Symbols and samples as in Figure 3.

For large cations such as Cs and Rb (e.g., Cs-tetra-ferri-annite, Rb-,Cs-rich phlogopite, and nanpingite) the small α value corresponds to a large interlayer separation, whereas for small cations such as Na and Ca (e.g., preiswerkite, paragonite, and

24

Brigatti & Guggenheim

Figure 15. Relationships between α and interlayer separation. Symbols and samples as in Figure 3.

clintonite) large α values correspond to small interlayer separations (Fig. 15). Thus, as previously noted (Radoslovich and Norrish 1962), the shape of the interlayer-cation cavity reflects the field strength of the interlayer cation. The cavity adjusts in size by tetrahedral rotation or by a shift in the cation toward or away from the plane defined by the three basal oxygen atoms (i.e., the basal plane). In K-rich trioctahedral micas, both α and interlayer separation increase from norrishite to tetra-ferriphlogopite (and aluminian phlogopite) toward values for Fe-rich polylithionite, Fe-rich phlogopite, Mg-rich annite, and phlogopite. Annite deviates from the trend of trioctahedral true micas owing to a larger interlayer separation. In the Ba-rich brittle mica, ferrokinoshitalite (M sites mainly occupied by Fe2+), α- and interlayerseparation values are smaller with respect to those of kinoshitalite (M sites mostly occupied by Mg). With respect to trioctahedral micas, the interlayer separation in both muscovite and celadonitic muscovite is smaller, but α values are similar. To explain this behavior, the octahedral, tetrahedral, and O(4) site chemistry must be considered. Compared to kinoshitalite, ferrokinoshitalite shows an enlargement of the octahedral sheet produced by the relatively large size of Fe2+ with respect to Mg and by F for OH substitution on O(4). Therefore, the smaller α value is attributed to the large size of Fe in the octahedra, which allows a better fit to the Al-rich tetrahedral sheet. Less rotation of the tetrahedra produces a larger size of the silicate ring, which allows Ba to better fit within the ring, thus reducing interlayer separation (Guggenheim and Frimmel 1999). In norrishite, the combined effects of a Si-rich tetrahedral sheet, which produces smaller individual tetrahedra within (001), and octahedral flattening owing to the relatively large Li and Mn3+, reduce the tetrahedral-octahedral sheet misfit, thus requiring limited tetrahedral rotation. In addition, the narrow interlayer region is partly related to the increase in the Coulombic interactions of O2- [in the O(4) site] and the interlayer K (Tyrna and Guggenheim 1991). In tetra-ferriphlogopite, the lateral extension of the

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

25

tetrahedral sheet is related to Fe3+ and involves a large α value to fit the Mg-rich octahedra. The O(4) site is mostly occupied by OH-, which produces H+–K+ repulsion, thus requiring a greater interlayer separation (Brigatti et al. 1996a). A similar adjustment occurs also in aluminian phlogopite (Alietti et al. 1995) because the composition 3+ [ 4] involving an exchange vector of [6 ] Al 3+ Mn1.96 Al 3+ [4 ] Si4− +1 creates larger tetrahedral-sheet and smaller octahedral-sheet dimensions with respect to phlogopite. With respect to trioctahedral micas, dioctahedral muscovite and celadonitic muscovite have smaller interlayer separations but similar α values. In dioctahedral micas, the proton position results in part from repulsion by the interlayer cation and the cations in the M(2) sites. Thus, the proton is located in that portion of the structure with minimal local positive-charge concentration, near the M(1) site (Radoslovich 1960; Guggenheim et al. 1987). The six-fold coordination of the interlayer cation with the basal inner O atom is distorted and elongated parallel to c*. Both effects (i.e., the distorted coordination of the interlayer cation and the smaller H+–K+ repulsion) thus control the interlayer separation. McCauley and Newman (1971) and Weiss et al. (1992) related α to the coordination of the interlayer cation. In an ideal structure α = 0° and the interlayer cation is in 12-fold coordination. In non-ideal structures, α values of greater than 0° reduce the interlayercation coordination number from 12 to 6. Weiss et al. (1992) determined the coordination number of the interlayer cation using the equation of Hoppe (1979): ECoN = ∑ j=12 j=1 C j , where ECoN is the Effective Coordination Number; (C j = exp[1.0 −( FIR j / MEFIR)6 ]); FIRj was calculated by dividing the A–Oj distance by the sum of anion and cation radii and then multiplying by the cation radii; MEFIR is a weighted mean of FIR, i.e., MEFIR =

j =12 ∑ j =1 w jFIR j ; j =12 ∑j=1 w j

w j = exp(1 − (FIR j FIR min ) 6 ;

FIRmin is the smallest FIRj in the interlayer cation coordination. They found that ECoN is close to 12 in tainiolite and annite, usually varies from 11 to 9 in polylithionite, ferroan polylithionite, phlogopite, and ferroan phlogopite, and is between 9 and 8 in muscovite and celadonitic muscovite, whereas paragonite and most brittle micas have the lowest ECoN (ECoN = 6). In addition to tetrahedral and octahedral site composition, the coordination of the interlayer cation was found to be affected by the stacking of the layers. In the most common polytypes (e.g., 1M, 2M1 and 3T, the polytypes of subfamily A, as defined by Backhaus and Durovíc 1984 and Durovíc et al. 1984) the coordination polyhedron of the interlayer cation varies from ditrigonal antiprism to octahedral, whereas in polytypes of subfamily B (e.g., 2M2 and 6H polytypes) it varies from ditrigonal to trigonal prismatic. Tetrahedral cation ordering Ordering of tetrahedral cations is quite unusual in the common mica species such as muscovite-2M1, phlogopite-1M and annite-1M (Bailey 1975, 1984c), whereas it is common in brittle micas. Margarite, bityite and anandite are examples of minerals with Si,Al (or Fe3+) tetrahedral ordering (Guggenheim 1984). Bailey (1984b) concluded that ordering of tetrahedral cations is favored for 3T structures (Güven and Burnham 1967; Brown 1978, Sidorenko et al. 1977b), for Si:Al ratios near to 1:1 (Guggenheim and Bailey 1975, 1978; Joswig et al. 1983; Lin and Guggenheim 1983) and for muscovite-1M, -2M1 and -2M2 crystals with a significant

26

Brigatti & Guggenheim

celadonite component (Güven 1971b; Zhoukhlistov et al. 1973; Sidorenko et al. 1975). In contrast, Amisano-Canesi et al. (1994) suggested that no long-range ordering of tetrahedral cations is present in muscovite-3T crystals and concluded that the tetrahedral cation ordering previously found by Güven and Burnham (1967) may be an artifact produced by the small number of independent reflections used in the crystal-structure refinement. It is very unusual to obtain increased cation order at high temperature, where, in most cases increasing disorder is the norm, however the results of neutron powder diffraction studies suggest tetrahedral Si-Al ordering for celadonitic muscovite (referred as “phengite”) at high temperature (Pavese et al. 1997, 1999, 2000). Guggenheim (1984) noted the importance of two factors in determining the degree of Si,Al ordering in crystals with Si:Al ratios of 1:1 that relate to octahedral- and the interlayer-site composition: (i) the charge of an apical oxygen that coordinates two Al3+ octahedral cations is undersaturated with respect to positive charge if the tetrahedral cation is Al3+, whereas it is balanced if the tetrahedral cation is Si4+; (ii) large cations in interlayer sites prop apart two adjacent 2:1 layers, thus minimizing electrostatic repulsions across the interlayer. Therefore Si,Al tetrahedral ordering seems to be favored in species with small, high-charged octahedral cations and small cations in interlayer sites. To date, long-range tetrahedral ordering has not been determined for preiswerkite and clintonite, but was found for ephesite, margarite, bityite and anandite. Although Raman spectra suggest the presence of strong short-range ordering in preiswerkite-1M [NaMg2AlAl2Si2O10(OH)2], long range ordering in tetrahedral sites was not found by crystal-structure refinement (Oberti et al. 1993). In contrast, ephesite [NaLiAl2Al2Si2O10(OH)2], which differs in composition from preiswerkite only for octahedral composition, is strongly ordered in space group C1 (Slade et al. 1987). In the latter case, perhaps, the presence of monovalent and trivalent octahedral cations requires ordering of the tetrahedral cations to achieve a suitable local charge balance on shared apical oxygen atoms. The possibility of tetrahedral cation ordering in kinoshitalite, characterized by a Si:Al ratio close to 1, was addressed by Guggenheim and Kato (1984), Guggenheim (1984), and Gnos and Armbruster (2000). Guggenheim (1984) related the lack of tetrahedral Si,Al ordering in kinoshitalite to the large interlayer separation (3.328 Å; Guggenheim and Kato 1984) caused by the large Ba interlayer cation which increases the separation between adjacent 2:1 layers, thus reducing any T–T electrostatic interactions across the interlayer. Lack of tetrahedral Si,Al ordering was also confirmed for ferrokinoshitalite (3.129 Å; Guggenheim and Frimmel 1999) which also has large interlayer separation but less than that of kinoshitalite. According to Gnos and Armbruster (2000), Si,Al ordering in kinoshitalite may be masked by twinning. They assumed different twin models to explain the average structure of this brittle mica in space group C2/m starting from complete Si,Al tetrahedral ordering in C2 and C 1 symmetries. The C2-space group model assumes that each Si tetrahedron is surrounded by three Al tetrahedra and vice-versa as consistent with Loewenstein’s (1954) Al-avoidance rule. The tetrahedral sheets of two adjacent 2:1 layers are arranged above and below the interlayer to produce the pattern along the c-axis for which Si is always adjacent to Si and Al adjacent to Al tetrahedra. The C 1 -space group model maintains the same Si,Al distribution within the tetrahedral sheets (i.e., one Si atom surrounded by three Al atoms and vice-versa), but differs for the Si,Al distribution along the c-axis. In this latter model, each Si tetrahedron is always opposed to an Al tetrahedron. Gnos and Armbruster (2000) concluded that the crystal-structure refinement is inconsistent with the twinning models that involve completely ordered Si,Al sheets. In contrast, the crystal-structure refinements of two disordered models to

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

27

produce C2/m symmetry (i.e., three-dimensional Si,Al disorder and one-dimensional disorder along the c axis) suggested a pattern of one-dimensional disorder along the [001] direction of completely Si,Al ordered tetrahedral sheets. In margarite [CaAl2 Al2Si2O10(OH)2], Al preferentially occupies two of the four symmetrically independent tetrahedra (Guggenheim and Bailey 1975, 1978; Joswig et al. 1983; Kassner et al. 1993). Kassner et al. (1993) found that mean tetrahedral Al–O and Si-O distances are identical in the two crystallographically independent tetrahedral sheets. Thus there is no asymmetry in the distribution of tetrahedral Al in these sheets as indicated by Guggenheim and Bailey (1975, 1978) based on an incompletely refined model. An ordering pattern similar to that of margarite occurs for tetrahedral sites of bityite with nearly complete ordering of Al, Be relative to Si (Lin and Guggenheim 1983). Octahedral coordination and long-range octahedral ordering Three translationally independent octahedral cation sites characterize the 2:1 layer. One site is trans coordinated by OH (or by F and/or Cl, and rarely by S) and is called M(1), the remaining two sites are cis-coordinated and are referred to as M(2) where the layer contains a symmetry plane which relates the two M(2) sites. Otherwise, the two cissites are labeled M(2) and M(3), respectively. M(1) is usually vacant in dioctahedral micas, whereas all three octahedral sites are occupied in trioctahedral micas. The cation distribution in the octahedral sites may be summarized as: (i) all the octahedra are occupied by the same kind of “crystallographic entity” (i.e., the same kind of ion or by a statistical average of different kinds of ions, including voids, referred to as homooctahedral micas by Durovíc 1981, 1994), (ii) two octahedra are occupied by the same kind of “crystallographic entity” and the third by a different entity in an ordered way (meso-octahedral micas), or (iii) each of the three sites is occupied by a different “crystallographic entity” in an ordered way (hetero-octahedral micas). The location of the origin of the octahedral sheet corresponds to: (i) the M(1) site for homo-octahedral micas; (ii) the site with different occupation for meso-octahedral micas; and (iii) the site with the smallest electron density for hetero-octahedral micas (Durovíc et al. 1984). As a consequence, two kinds of layers can be defined, namely, the “M(1) layer” and the “M(2) layer,” the first with the origin of the octahedral sheet in M(1), the latter in either the M(2) or M(3) site (Zvyagin 1967). The “M(1) layer” is the more common. Weiss et al. (1992) identified eight possible geometries of the octahedral sheet based on the size of octahedral sites. In particular, they derived four- and three-different geometries for mesooctahedral and for hetero-octahedral micas, respectively. Toraya (1981) noted that the M(1) site is usually occupied by a cation of lower charge or by a vacancy. He explained this characteristic by considering the effect on the linkages of the polyhedra. An increase in the size of M(2) is energetically unfavorable because the O–O shared edge between two adjacent M(2) cations would be enlarged [increasing the repulsion between octahedral M(2) cations], the O–OH,F edge between M(1) and M(2) would be reduced [thus decreasing repulsion between octahedral M(1) and M(2) cations], and the increased repulsion between oxygen atoms on the unshared lateral edges of M(1) would occur owing to the smaller size of this site. In contrast, the only unfavorable factor created by an increase in M(1) would be an increase in repulsion between M(1) and M(2) cations, which would be mitigated by the decrease in charge of M(1). However, examples where three sites are all equal or each site differs are not unusual in trioctahedral micas. Several phlogopite and tetra-ferriphlogopite crystals (space group C2/m) show the same kind of cations (or a disordered cation distribution) in M(1) and M(2) octahedra, i.e., the difference between the mean bond lengths and mean electron counts (m.e.c.) of M(1) and M(2) sites are equal within the standard deviations

28

Brigatti & Guggenheim

(Δ 〈M–O〉 = |〈M(1)–O〉 – 〈M(2)–O〉| < 0.004 Å; Δ m.e.c = |m.e.c.M(1) – m.e.c.M(2)| < 1.0 e-; see, for example, in Tables 1-3 (end of chapter) the data by Semenova et al. 1977; Hazen et al. 1981; Brigatti et al. 1996a; Gnos and Armbruster 2000), whereas some Li-rich micas (space group C2) have different cation ordering in M(1), M(2) and M(3) sites; e.g., zinnwaldite-1M (Guggenheim and Bailey 1977), lepidolite-1M (Backhaus 1983), zinnwaldite-2M1 (Rieder et al. 1996), ferroan polylithionite-1M and lithian “siderophyllite”-1M (Brigatti et al. 2000b). The octahedral sheet may show different cation distributions in M(1), M(2), and M(3), but the size of each octahedron need not differ. For example, the zinnwaldite-1M (polylithionite-siderophyllite intermediate) structure refined by Guggenheim and Bailey (1977) shows M(1) ≠ M(2) ≠ M(3) on the basis of the site scattering power, whereas M(1) = M(3) ≠ M(2) on the basis of size of the polyhedra. Verification of ordering requires not only the analysis of cation-anion bond length but also the refinement of octahedral-site occupancies because mean bond lengths of octahedra with different occupancies may be similar. Therefore, this discussion of octahedral ordering is based only on samples for which the m.e.c. of each octahedral site is available. For trioctahedral true micas of the phlogopite-annite join, the m.e.c. of both M(1) and M(2) sites increases from phlogopite to annite through ferroan phlogopite and magnesian annite. This suggests that an increase in the Mg-1 Fe exchange occurs and that Fe occupies both the M(1) and M(2) sites (Fig. 16). However, ferroan-phlogopite and magnesian-annite samples (Tables 1a and 1b) have differences in mean bond lengths (to 0.036 Å) and in m.e.c. (to 2.5 e-) for the M(1) and M(2) sites. Thus, a slight preference for cations with larger radii and atomic numbers for M(1) occurs. The greatest differences between M(1) and M(2) octahedral mean bond distances occur in Al-bearing magnesian annite from peraluminous granites where Al is ordered in M(2) (Brigatti et al. 2000a).

Figure 16. Mean electron count (m.e.c.) of M(2) [M(2) = M(3)] vs. m.e.c. of M(1) site. Symbols: filled circles = annite; filled circles, xhair =magnesian annite; open circle, x-hair = ferroan phlogopite; open circles = phlogopite. Estimated average standard deviation: ±0.3 e-.

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

29

Figure 17. Difference between M(1) and M(2) site mean bond distance in mica crystals of the phlogopite-annite join. Symbols and samples as in Figure 16. The average standard deviation on 〈M(1)–O〉 and 〈M(2)– O〉 bond distances was evaluated as ±0.002 Å.

Thus, ordering along this join seems to be enhanced where, in addition to Mg2+ and Fe2+, cations of different size and charge occur in octahedral coordination (Fig. 17). Although the m.e.c. of M(1) and M(2) increases with the exchange vector of Mg-1Fe (Fig. 16), annite crystals have Δ 〈M–O〉 values much smaller than those for magnesian annite. For compositions intermediate between those of phlogopite and annite, exchange vectors that introduce cations of different charge (or vacancies) in octahedral sites significantly affect the layer topology. In fact, in phlogopite the octahedral sites are equal in size and m.e.c., annite shows octahedral sites with similar m.e.c. (Δ m.e.c. < 0.4 e ) and differences in Δ 〈M–O〉 bond lengths (Δ 〈M–O〉 < 0.02), whereas crystals of phlogopiteannite with intermediate compositions always have one larger octahedron and two smaller octahedra and usually differences in m.e.c. for M(1) and M(2). The reduction of interlayer separation with Ti content (Fig. 18) is related to the decrease in the K-O(4) [ O(4) = OH, O, F, Cl ] distance, which is ascribed to “Ti-oxy” 1− substitution, Ti-oxy = [6] Ti4 +O22 − [6] Mg 2+ −1 (OH) −2 ). The interlayer cation is shifted deeper into the interlayer cavity owing to the deprotonation of the O(4) site; thus, the K–O(4) distance decreases with a decrease in interlayer separation. Cruciani and Zanazzi (1994) observed that the off-center shift of the cation at M(2) is associated with an increase in the proportion of [6]Ti, and this reveals a [6]Ti preference for the M(2) site. Li-rich micas in the siderophlyllite-polylithionite join (Fig. 19) show different patterns of octahedral order. For example, in a synthetic polylithionite (space group C2/m) with octahedral composition Li2Al (Takeda and Burnham 1969) the ordering pattern results in a large M(1) site of composition Li0.89Al0.11 and two equivalent M(2) sites of composition (Li0.55Al0.45). A similar ordering pattern was observed in natural

30

Brigatti & Guggenheim

Figure 18. Interlayer separation vs. octahedral Ti4+ content for mica crystals along the phlogopite–annite join. Symbols and samples as in Figure 16. The average standard deviation on the interlayer separation was evaluated as ±0.004 Å.

Figure 19. Ternary [6]Al3+ – [6]Li+ -[6]Fe2+ diagram showing compositional data for Li-rich micas. Symbols: filled circles = Li-containing annite crystals; open squares, x-hair = ferroan polylithionite and crystals with composition intermediate between polylithionite and siderophyllite; open squares=polylithionite; filled squares, x-hair = siderophyllite; filled square = trilithionite. The open circles indicate the composition of the end members (from Brigatti et al. 2000b).

Mica Crystal Chemistry and Influence of P-T-X on Atomistic Models

31

trilithionite-1M (Sartori 1976; Guggenheim 1981) and in polylithionite-2M1 and 2M2 (Takeda et al. 1971; Sartori et al. 1973; Swanson and Bailey 1981). In some Li-rich micas, the ideal layer symmetry is reduced from C2/m to C2, as a result of a different pattern in octahedral ordering in the cis-octahedral sites (Guggenheim and Bailey 1977; Guggenheim 1981; Backhaus 1983; Mizota et al. 1986; Rieder et al. 1996; Brigatti et al. 2000b). These minerals have 〈M(1)–O〉 ≅ 〈M(3)–O〉 > 〈M(2)–O〉 and occasionally 〈M(1)–O〉 ≅ 〈M(2)–O〉 > 〈M(3)–O〉 (Backhaus 1983; Brigatti et al. 2000b). The scattering efficiency for the M(1), M(2) and M(3) sites implies ordering with M(1) ≠ M(2) ≠ M(3), M(1) = M(3) 1 kinds and their MDO polytypes. Acta Crystallogr A38:491-498 Ďurovič S (1974) Notion of “packets” in the theory of OD structure of M>1 kinds of layer. Examples: Kaolinites and MoS2. Acta Crystallogr B30:76-78 Ďurovič S (1979) Desymmetrization of OD structures. Kristall und Technik 14:1047-1053

274 

Nespolo & Ďurovič 

Ďurovič S (1981) OD-Charakter, Polytypie und Identifickation von Schichtsilikaten. Fortschr Mineral 59:191226 Ďurovič S (1982) Ordered and disordered polytypes of sheet silicates and their diffraction apttern assuming ideal ditrigonalization of tetrahedral sheets. 9th Conf on Clay Mineralogy and Petrology, Zvolen, 127-134 Ďurovič S (1994) Significance of superposition structures in the polytypism of phyllosilicates. In Aperiodic ’94 Proceedings of the International Conference on Aperiodic Crystals. G Chapuis, W Paciorek (ed) World Scientific, Singapore / New Jersey / London / Hong Kong, p 595-599 Ďurovič S (1997) Fundamentals of OD theory. In Modular aspects of minerals / EMU Notes in Mineralogy, vol. 1. S Merlino (ed) Eötvös University press, Budapest, p 1-28 Ďurovič S (1999) Layer stacking in general polytypic structures. Sect. 9.2.2 in International Tables for Crystallography, Vol. C (Ed. ACJ Wilson, E Prince) Dordrecht / Boston / London: Kluwer Academic Publishers, 752-765 Ďurovič S, Dornberger-Schiff K (1979) New fully descriptive polytype symbols for basic polytypes of clay minerals. Eight Conference on Clay Mineralogy and Petrology, Teplice, 19-25 Ďurovič S, Weiss Z (1986) OD structures and polytypes. Bull Minéral 109:15-29 Ďurovič S, Weiss Z, Backhaus K-O (1984) Polytypism of micas. II. Classification and abundance of MDO polytypes. Clays Clay Miner 32:454-474 Endo Y (1968) Growth Spirals and Polytypism of Fluor-phlogopite. J Mineral Soc Jpn 8, Spec. issue 2:39-41 (in Japanese) Engel P, Matsumoto T, Steinamann G, Wondratschek H (1984) The non-characteristic orbits of the space groups. Z Kristallogr Supplement Issue No. 1 Farkas-Jahnke M (1966) Direct method for the determination of polytype structures. Acta Crystallogr 21:A173 Farkas-Jahnke M, Dornberger-Schiff K (1970) A Direct Method for the Determination of Polytype Structures. II. Determination of a 66R Structure. Acta Crystallogr A26:35-41 Ferraris G Gula A, Ivaldi G, Nespolo M Sokolova E, Uvarova Yu, Khomyakov AP (2001) First structure determination of an MDO-2O mica polytype associated with a 1M polytype. Eur J Miner 13: 1013-1023 Fichtner K (1965) Zur Existenz von Gruppoiden verschiedener Ordnungsgrade bei OD-Strukturen aus gleichartigen Schichten. Wiss Z TU Dresden 14:1-13 Fichtner K (1977) Zur Symmetriebeschreibung von OD-Kristallstrukturen durch Brandtsche und Ehresmannsche Gruppoide. Beitr z algebra u Geometrie 6:71-99 Fichtner K (1980) On groupoid in crystallography. MATCH, commun math comput chem 9:21-40 Filut MA, Rule AC, Bailey SW (1985) Crystal structure refinement of anandite-2Or, a barium- and sulfurbearing trioctahedral mica. Am Mineral 70:1298-1308 Frank FC (1951) The growth of carborundum; dislocations and polytypism. Phil Mag 42:1014-1021 Franzini M (1966) Nuovi dati sulla struttura delle miche triottaedriche. Atti Soc Tosc Sci Nat, Mem Ser A 73:620-631 Franzini M (1969) The A and B mica layers and the crystal structure of sheet silicates. Contrib Mineral Petrol 21:203-224 Franzini M, Schiaffino L (1963a) Polimorfismo e leggi di geminazione delle biotiti. Atti Soc Tosc Sci Nat, Mem Ser A 70:60-98 Franzini M, Schiaffino L (1963b) On the crystal structure of biotite. Z Kristallogr 119:297-309 Friedel G (1904) Étude sur les groupements cristallins. Extrait du Bullettin de la Société de l'Industrie minérale, Quatrième série, Tomes III e IV. Saint-Étienne, Societè de l’Imprimerie Thèolier J. Thomas et C., 485 p Friedel G. (1926) Leçons de Cristallographie, Berger-Levrault, Nancy, Paris, Strasbourg, 602 p Gaubert P (1896) Sur la production artificielle de la macle des spinelles dans les cristaux d’azotate de plomb. Bull Soc franç Minér 19:431-434 Giuseppetti G, Tadini C (1972) The Crystal Structure of 2O brittle mica: Anandite. Tschermaks Mineral Petrol Mitt 18:169-184 Goldschmidt V (1918) Atlas der Krystallformen, band IV. Heidelberg: Carl Winters Universitätsbuchhandlung Grell H (1984) How to choose OD layers. Acta Crystallogr A40:95-99 Griffen DT (1992) Silicate Crystal Chemistry. Oxford University Press, Oxford Guggenheim S, Bailey SW (1975) Refinement of the margarite structure in subgroup symmetry. Am Mineral 60:1023-1029 Guggenheim S, Bailey SW (1978) The refinement of the margarite structure in subgroup symmetry: corrections, further refinement and comments. Am Mineral 63:186-187 Guinier A, Bokij GB, Boll-Dornberger K, Cowley JM, Ďurovič S, Jagodzinski H, Khrisna P, DeWolff PM, Zvyagin BB, Cox DE, Goodman P, Hahn Th, Kuchitsu K, Abrahams SC (1984) Nomenclature of

 

Crystallographic Basis of Polytypism and Twinning in Micas 

275 

Polytype Structures. Report of the International Union of crystallography Ad-Hoc Committee on the Nomenclature of Disordered, Modulated and Polytype Structures Acta Crystallogr A40:399-404 Güven N (1971) Structural Factors Controlling Stacking Sequences in Dioctahedral Micas. Clays Clay Miner 19:159-165 Hahn Th, Vos A (2002) Reflection conditions. Sect. 2.13 in International Tables for Crystallography, Vol. A 5th edition. Th Hahn (ed) Dordrecht / Boston: London: Kluwer Academic Publishers (in press) Hahn T, Janovec V, Klapper H (1999) Bicrystals, twins and domain structures – A Comparison. Ferroelectrics 222:11-21 Hendricks SB, Jefferson ME (1939) Polymorphism of the micas with optical measurements. Am Mineral 24:729-771 Iijima S, Buseck PR (1978) Experimental study of disordered mica structure by high-resolution electron microscopy. Acta Crystallogr A34:709-719 International Tables for Crystallography Vol. A. (2002) 5th edition. Th Hahn (ed) Dordrecht / Boston / London: KLuwer Academic Publishers (in Press) Ito T (1935) On the symmetry of rhombic pyroxenes. Z Kristallogr 90:151-162 Ito T (1938) Theory of twinned space groups. J Japan Assoc Mineral Petr Econ Geol 20:201-210 (in Japanese) Ito T (1950) X-ray studies on polymorphism. Maruzen Co., Tokyo, 231 pp Ito T, Sadanaga R (1976) On the crystallographic space groupoids. Proc Jpn Acad 52:119-121 Iwasaki H (1972) On the diffraction enhancement of symmetry. Acta Crystallogr A28:253-261 Jackson WW, West J (1931) The crystal structure of muscovite KAl2(AlSi3)O10(OH)2. Z Kristallogr 76:211227 Joswig W, Takéuchi Y, Fuess H (1983) Neutron-diffraction study on the orientation of hydroxyl groups in margarite. Z. Kristallogr 165:295-303 Kassner D, Baur WH, Joswig W, Eichhorn K, Wendschuh-Josties M, Kupčik V (1993) A test of the importance of weak reflections in resolving a space-group ambiguity involving the presence or absence of an inversion centre. Acta Crystallogr B49:646-654 Kogure T, Nespolo M (1999a) First occurrence of a disordered stacking sequence including (±60º 180º) in Mg-rich annite. Clays Clay Miner 48:784-792 Kogure T, Nespolo M (1999b) A TEM study of long-period mica polytypes: determination of the stacking sequence of oxybiotite by means of atomic-resolution images and Periodic Intensity Distribution (PID) Acta Crystallogr B55:507-516 Kokscharow NV (1875) Materialen zur Mineralogie Russlands. Vol. 7. St. Petersburg Konishi H, Akai J (1990) HRTEM observation of new complex polytype of biotite from dacites in Higashiyama hills, Niigata, Central Japan. Clay Science 8:25-30 Knurr RA, Bailey SW (1986) Refinement of Mn-substituted muscovite and phlogopite. Am Mineral 34:7-16 Kuwahara Y, Uehara S, Aoki Y (1998) Surface microtopography of lath-shaped hydrothermal illite by tapping-mode™ and contact-mode AFM. Clays Clay Miner 46:574-582 Lipson H, Taylor CA (1958) Fourier Transforms and X-Ray Diffraction, Bell, London Mackovicky E (1997) Modularity – different types and approaches. In Modular aspects of minerals / EMU Notes in Mineralogy, vol. 1. S Merlino (ed) Eötvös University press, Budapest, p 315-343 Mallard E (1879) Traité de Cristallographie geometrique et physique, Vol. I. Paris: Dunod. 372pp Marignac C (1847) Notices minéralogiques. Suppl Bibl Universe Genève, arch Sci Phys Nat 6:293-304 Matsumoto T, Kihara K, Iwasaki H (1974) Conditions for the diffraction enhancement of symmetry of types 1 and 2. Acta Crystallogr A30:107-108 Matsumoto T, Wondratschek H (1979) Possible superlattices of extraordinary orbits in 3-dimensional space. Z Kristallogr 150:181-198 Mauguin MCh (1927) Êtude du mica muscovite au moyen des rayons X. CR Acad Sci Paris 185:288-291 Mauguin MCh (1928) Etude de Micas au moyen du rayons X. Bull Soc franç Minér Crist 51:285-332 McLarnan TJ (1981) The number of polytypes in sheet silicates. Z Kristallogr 155:247-268 Merlino S (1990) OD structures in mineralogy. Per Mineral 59:69-92 Mogami K, Nomura K, Miyamoto M, Takeda H, Sadanaga R (1978) On the number of distinct polytypes of mica and SiC with a prime layer-number. Can Mineral 16:427-435 Mügge O (1898) Über Translationen und verwandte Erscheinungen in Krystallen. N Jb Miner Geol Paläontol 1:71-158 Nespolo M (1999) Analysis of family reflections of OD-mica polytypes, and its application to twin identification. Mineral J 21:53-85 Nespolo M (2001) Perturbative theory of mica polytypism. Role of the M2 layer in the formation of inhomogeneous polytypes. Clays Clay Miner 49:1-23

276 

Nespolo & Ďurovič 

Nespolo M, Ferraris G (2000) Twinning by syngonic and metric merohedry. Analysis, classification and effects on the diffraction pattern. Z Kristallogr 215:77-81 Nespolo M, Kogure T (1998) On the indexing of 3T mica polytype. Z Kristallogr 213:4-12 Nespolo M, Kuwahara Y (2001) Apparent polytypism in the Ruiz Peak ferric phlogopite. Eur J Mineral 13 (in press) Nespolo M, Takeda H (1999) Inhomogeneous mica polytypes: 8-layer polytype of the 2M1 structural series determined by the Periodic Intensity Distribution (PID) analysis of the X-ray diffraction pattern. Mineral J 21:103-118 Nespolo M, Takeda H, Ferraris G (1997a) Crystallography of mica polytypes. In Modular aspects of minerals / EMU Notes in Mineralogy, vol. 1. S Merlino (ed) Eötvös University press, Budapest, p 81-118 Nespolo M, Takeda H, Ferraris G, Kogure T (1997b) Composite twins of 1M mica: derivation and Identification. Mineral J 19:173-186 Nespolo M, Takeda H, Ferraris G (1998) Representation of the axial settings of mica polytypes. Acta Crystallogr A54:348-356 Nespolo M, Ferraris G, Ďurovič S (1999a) OD character and twinning – Selective merohedry in class II merohedric twins of OD polytypes. Z Kristallogr 214:776-779 Nespolo M, Ferraris G, Takeda H, Takéuchi Y (1999b) Plesiotwinning: oriented crystal associations based on a large coincidence-site lattice. Z Kristallogr 214:378-382 Nespolo M, Kogure T, Ferraris G (1999c) Allotwinning: oriented crystal association of polytypes – Some warnings on consequences. Z Kristallogr 214:1-4 Nespolo M, Takeda H, Kogure T, Ferraris G (1999d) Periodic Intensity Distribution (PID) of mica polytypes: symbols, structural model orientation and axial settings. Acta Crystallogr A55:659-676 Nespolo M, Ferraris G, Takeda H (2000a) Twins and allotwins of basic mica polytypes: theoretical derivation and identification in the reciprocal space. Acta Crystallogr A56:132-148 Nespolo M, Ferraris G, Takeda H (2000b) Identification of two allotwins of mica polytypes in reciprocal space through the minimal rhombus unit. Acta Crystallogr B56:639-647 Ohta T, Takeda H, Takéuchi Y (1982) Mica polytypism: similarities in the crystal structures of coexisting 1M and 2M1 oxybiotite. Am Mineral 67:298-310 Pabst A (1955) Redescription of the single layer structure of the micas. Am Mineral 40:967-974 Pauling L (1930) The structure of micas and related minerals. Proc Nat Ac Sci 16:123-129 Pavese A, Ferraris G, Prencipe M, Ibberson R (1997) Cation site ordering in phengite 3T from the Dora-Maira massif (western Alps): a variable-temperature neutron powder diffraction study. Eur J Mineral 9:11831190 Pavlishin VI, Semenova TF, Rozhdesvenskaya IV (1981) Protolithionite-3T: structure, typomorphism and practical importance. Mineral Zh 3:47-60 (in Russian) Peacock MA, Ferguson RB (1943) The morphology of muscovite in relation to the crystal lattice. Univ Toronto, Studies in Mineral 48:65-82 Pleasants PA, Baake M, Roth J (1996) Planar coincidences for N-fold symmetry. J Math Phys 1029-1058 Radoslovich EW (1960) The structure of muscovite KAl2(Si3Al)O10(OH)2. Acta Crystallogr 13:919-932 Ramsdell LS (1947) Studies on silicon carbide. Am Mineral 32:64-82 Ranganathan S (1966) On the geometry of coincidence-site lattices. Acta Crystallogr 21:197-199 Rieder M (1968) Zinnwaldite: Octahedral ordering in lithium-iron micas. Science 160:338-1340 Rieder M (1970) Lithium-iron micas from the Krušné hory Mountains (Erzgebirge): Twins, epitactic overgrowths and polytypes, Z Kristallogr 132:161-184 Rieder M, Weiss Z (1991) Oblique-texture photographs: more information from powder diffraction. Z Kristallogr 197:107-114 Rieder M, Hybler J, Smrčok L, Weiss Z (1996) Refinement of the crystal structure of zinnwaldite 2M1. Eur J Mineral 8:1241-1248 Ross M, Takeda H, Wones DR (1966) Mica polytypes: systematic description and identification. Science 151:191-193 Rieder M, Cavazzini G, D’yakonov YuS, Frank-Kamenetskii VA, Gottardi G, Guggenheim S, Koval’ PV, Müller G, Neiva AMR, Radoslowich EW, Robert JL, Sassi FP, Takeda H, Weiss Z, Wones DR (1998) Nomenclature of the micas. Clays and Clay Miner 46:586-595 Royer L (1928) Recherches expérimentales sur l’épitaxie ou orientation mutuelle de cristaux d’espèces differentes. Bull Soc franç Minér Crist 51:7-159 Royer L (1954) De l’épitaxie; quelques remarques sur le problèmes qu’elle soulève. Bull Soc franç Minér Crist 77:1004-1028 Sadanaga R (1978) omplex structures and space groupoids. Rec Progr Nat Sci Jpn 3:143-151 Sadanaga R, Ohsumi K (1979) Basic theorems of vector symmetry in crystallography. Acta Crystallogr A35:115-122

 

Crystallographic Basis of Polytypism and Twinning in Micas 

277 

Sadanaga R, Sawada T, Ohsumi K, Kamiya K (1980) Classification of superstructures by symmetry. J Jpn Assoc Min Petr Econ Geol, Spec. Issue No. 2:23-29 Sadanaga R, Takeda H (1968) Monoclinic diffraction patterns produced by certain triclinic crystals and diffraction enhancement of symmetry. Acta Crystallogr B24:144-149 Sadanaga R, Takeda H (1969) Description of mica polytypes by new unit layers. J Mineral Soc Japan 9:177184 (in Japanese) Sadanaga R, Takéuchi Y (1961) Polysynthetic twinning of micas. Z Kristallogr 116:406-429 Sartori F (1977) The crystal structure of a 2M1 lepidolite. Tschermaks Min Petr Mitt 24:23-37 Sartori F, Franzini M, Merlino S (1973) Crystal Structure of a 2M2 Lepidolite. Acta Crystallogr B29:573-578 Schaskolsky M, Schubnikow A (1933) Über die künstliche herstellung gesetzmäβiger kristallverwachsungen des kalialauns. Z Kristallogr 85:1-16 Schläfli L (1950) Gesammelte mathematische Abhabndlungen (Vol. 1) Birkhäuser, Basel Sidorenko OV, Zvyagin BB, Soboleva SV (1975) Crystal structure refinement for 1M dioctahedral mica. Sov Phys Crystallogr 20:332-335 Sidorenko OV, Zvyagin BB, Soboleva SV (1977a) Refinement of the crystal structure of 2M1 paragonite by the high-voltage electron diffraction method. Sov Phys Crystallogr 22:554-556 Sidorenko OV, Zvyagin BB, Soboleva SV (1977b) The crystal structure of 3T paragonite. Sov Phys Crystallogr 22:557-560 Smith JV, Yoder HS (1956) Experimental and theoretical studies of the mica polymorphs, Mineral Mag 31:209-235 Slade PG, Schultz PK, Dean C (1987) Refinement of the Ephesite structure in C1 symmetry. N Jb Mineral Mh 1987:275-287 Smrčok Ĺ, Ďurovič S, Petříček V, Weiss Z (1994) Refinement of the crystal structure of cronstedtite-3T. Clays Clay Miner 42:544-551 Smrčok L, Weiss Z (1993) DIFK91: a program for the modelling of powder diffraction patterns on a PC. J Appl Cryst 26:140-141 Smyth JR, Jacobsen SD, Swope RJ, Angel RJ, Arlt T, Domanik K, Holloway JR (2000) Crystal structures and compressibilities of synthetic 2M1 and 3T phengite micas. Eur J Mineral 12:955-963 Sokolova GV, Aleksandrova VA, Drits VA, Bairakov VV (1979) Crystal structures of two brittle Lithia micas. In Kristallokhimiya i Struktura Mineralov. Frank-Kamenetskii VA (ed) Nauka, Moscow, p 55-66 (in Russian) Sorokin ND, Tairov YuM, Tsvetkov VF, Chernov MA (1982a) The laws governing the changes of some properties of different silicon carbide polytypes. Dokl Akad Nauk SSSR 262:1380-1383 (in Russian) Sorokin ND, Tairov YuM, Tsvetkov VF, Chernov MA (1982b) Crystal-chemical properties of the polytypes of silicon carbide. Sov Phys Crystallogr 28:539-542 Sueno S, Takeda H, Sadanaga R (1971) Two-dimensional regular aggregates of layered crystals. Mineral J 6:172-185 Sunagawa I (1960) Mechanism of crystal growth, etching and twin formation of hematite. Mineral J 3:59-89 Sunagawa I (1964) Growth spirals on phlogopite crystals. Am Mineral 49:1427-1434 Sunagawa I (1977) Natural crystallization. J Crystal Growth 42:214-223 Sunagawa I (1978) Vapour growth and epitaxy of minerals and synthetic crystals. J Crystal Growth 45:3-12 Sunagawa I (1984) Growth of crystal in nature. In Material Science of the Earth’s Interior. I Sunagawa (ed) Terra Publishing Company, Tokyo - D. Reidel Publishing Company, Dordrecht / Boston / Lancaster, p 63-105 Sunagawa I, Endo J, Daimon N, Tate I (1968) Nucleation, growth and polytypism of fluor-phlogopite from the vapour phase. J Crystal Growth 3,4:751 Sunagawa I, Koshino Y (1975) Growth Spirals on Kaolin Group Minerals. Am Mineral 60:407-412 Sunagawa I, Koshino Y, Asakura M, Yamamoto T (1975) Growth mechanism of some clay minerals. Fortschr Miner 52:217-224 Sunagawa I, Tomura S (1976) Twinning in phlogopite. Am Mineral 61:939-943 Tairov YuM, Tsvetkov VF (1983) Progress in controlling the growth of polytypic crystals. In Crystal Growth and Characterization of Polytype Structures. P Krishna (ed) Pergamon Press, Oxford / New York / Toronto / Sydney / Paris / Frankfurt, p 111-162 Takano Y, Takano K (1958) Apparent polytypism and Apparent Cleavage of the Micas. J Mineral Soc Jpn 3:674-692 (in Japanese) Takeda H (1967) Determination of the layer stacking sequence of a new complex mica polytype: A 4-layer Lithium Fluorophlogophite. Acta Crystallogr 22:845-853 Takeda H (1969) Existence of complex mica polytype series based on 2M1 sequence. Jpn Crystallogr Assoc Autumn Meeting, Iwate, 2-3 (in Japanese) Takeda H (1971) Distribution of mica polytypes among space gGroups. Am Mineral 56:1042-1056

278 

Nespolo & Ďurovič 

Takeda H, Donnay JDH (1965) Compound tessellations in crystal structures. Acta Crystallogr 19:474-476 Takeda H, Haga N, Sadanaga R (1971) Structural investigation of polymorphic transition between 2M2-, 1MLepidolite and 2M1 Muscovite. Mineral J 6:203-215 Takeda H, Ross M (1975) Mica polytypism: dissimilarities in the crystal structures of coexisting 1M and 2M1 biotite. Am Mineral 60:1030-1040 Takeda H, Ross M (1995) Mica polytypism: identification and origin. Am Mineral 80:715-724 Takeda H, Sadanaga R (1969) New unit layers for micas. Mineral J 5:434-449 Takéuchi Y (1965) Structures of brittle micas. Proc. 13th Natl. Conf. Madison, Wisconsin, 1964, Clays Clay minerals. Pergamon Press, 1-25 Takéuchi Y. (1971) Polymorphic or polytypic changes in biotites, pyroxenes, and wollastonites. J Mineral Soc Jpn 10:Spec. Issue No. 2:87-99 (in Japanese) Takéuchi Y, Haga N (1971) Structural Transformation of Trioctahedral Sheet Silicates. Slip mechanism of octahedral sheets and polytypic changes of micas. Mineral Soc Japan Spec Pap 1:74-87 (Proc. IMAIAGOD Meetings '70, IMA Vol.) Takéuchi Y, Sadanaga R (1966) Structural studies of brittle micas. I. The structure of xantophyllite refined. Mineral J 4:424-437 Takéuchi Y, Sadanaga R, Aikawa N (1972) Common lattices and image sets of hexagonal lattices, and their application to composite electron-diffraction patterns of biotite. Z Kristallogr 136:207-225 Thompson JB Jr (1981) Polytypism in complex crystals: contrast between mica and classical polytypes. In Structure and Bonding vol. II. M O'Keefe, A Navrotsky (ed), Academic Press, San Diego / London / Burlington, p 167-196 Tokonami M (1966) The structure determination of the 96R polytype of SiC by a direct method. Mineral J 4:401-423 Tokonami M, Hosoya S (1965) A systematic method for unravelling a periodic vector set. Acta Crystallogr 18:908-916 Tolansky S, Morris PG (1947a) An interferometric survey of the mica. Mineral Mag 28:137-145 Tolansky S, Morris PG (1947b) An interferometric examination of synthetic mica. Mineral Mag 28:146-150 Tomura S, Kitamura M, Sunagawa I (1979) Surface microtopography of metamorphic white micas. Phys Chem Miner 5:65-81 Tschermak G (1878) Die Glimmergruppe (I. Theil) Z Kristallogr 2:14-50 Tsvetkov V F (1982) Problems and prospects of growing large silicon carbide crystals. Izv Leningr Elektrotekh Inst 302:14-19 (in Russian) Udagawa S, Urabe K, Hasu H (1974) The crystal structure of muscvite dehydroxylate. J Japan Assoc Mineral Petr Econ Geol 58:381-389 (in Japanese, with English Abstract) Ungemach H (1935) Sur la Syntaxie et la Polytypie. Z Kristallogr 91:1-22 Verma AR (1953) Crystal Growth and Dislocations. London: Butterworths, 182p Weiss Z, Ďurovič S (1980) OD interpretation of Mg-vermiculite. Symbolism and X-ray identification of its polytypes. Acta Crystallogr A36:633-640 Weiss Z, Ďurovič S (1989) A united classification and X-ray identification of phyllosilicate polytypes. Collected abstracts, 9th International Clay Conference, Strasbourg (France), p. 430 Weiss Z, Wiewióra A (1986) Polytypism of micas. III. X-ray Diffraction Identification. Clays Clay Minerals 34:53-68 Wondratschek H (1976) Extraordinary orbits of space groups. Theoretical considerations. Z Kristallogr 143:460-470 Wondratschek H (2002) Introduction tp space-groups. Sect. 8 in international Tables for Crystallography, Vol. A,5th edition. The Hahn (ed) Dordrecht / Boston: London: Kluwer Academic Publishers (in press) Zhukhlistov AP, Zvyagin BB, Shuriga TN (1983) Electron-diffraction investigation of the crystal structure of di-trioctahedral Li,Fe-phengite 1M. Sov Phys Crystallogr 28:518-521 Zhukhlistov AP, Zvyagin BB, Pavlishin VI (1990) Polytypic 4M modification of Ti-biotite with nonuniform alternation of layers, and its appearance in electron-diffraction patterns from textures. Sov Phys Crystallogr 35:232-236 Zussman J (1979) The crystal chemistry of micas. Bull Mineral 102:5-13 Zvyagin BB (1962) A theory of polymorphism of micas. Sov Phys Crystallogr 6:571-580 Zvyagin BB (1967) Electron diffraction analysis of clay mineral structures. New York: Plenum Press, 364 p Zvyagin BB (1985) Polytypism in contemporary crystallography. Sov Phys Crystallogr 32:394-399 Zvyagin BB (1988) Polytypism of crystal structures. Comput Math Applic 16:569-591 Zvyagin BB (1993) A contribution to polytype systematics. Phase Trans 43:21-25 Zvyagin BB (1997) Modular analysis of crystal structures. In Modular aspects of minerals / EMU Notes in Mineralogy, vol. 1. S Merlino (ed) Eötvös University press, Budapest, p 345-372 Zvyagin BB, Drits VA (1996) Interrelated features of structure and stacking of kaolin mineral layers Clays Clay Miner 44:297-303

 

Crystallographic Basis of Polytypism and Twinning in Micas 

279 

Zvyagin BB, Gorshkov AI (1966) Effects of secondary diffraction in selected area patterns of mineral crystals superimposed with a relative rotation around the primary beam. Sixth Internat. Congress for Electron Microscopy, Kyoto. In Electron Microscopy 1966 Vol. I. R Uyeda (ed) Maruzen, Tokyo, p 603-604 Zvyagin BB, Vrublevskaya ZV, Zhukhlistov AP, Sidorenko OV, Soboleva SV, Fedotov AF (1979) Highvoltage electron diffraction in the study of layered minerals. Moscow: Nauka Press, 224 pp. (in Russian)

This page left blank.

5

Investigations of Micas Using Advanced Transmission Electron Microscopy Toshihiro Kogure Department of Earth and Planetary Science Graduate School of Science, The University of Tokyo 7-3-1 Hongo, Bunkyo-ku Tokyo 113-0033 Japan [email protected]

INTRODUCTION After a long history of development and improvement, recent transmission electron microscopes (TEMs) with various analytical functions have become important in material science and engineering. These functions include not only obtaining magnified images of specimens, but also electron-diffraction patterns, chemical analyses, and chemical-state analyses with spacial resolution far greater than other methodologies. It is impossible to cover all of these functions considering page limitations and, more importantly, considering the author’s knowledge and ability even for topics limited to studies of mica. This chapter focuses on the investigations of mica using high-resolution transmission electron microscopy (HRTEM). HRTEM is generally defined as a technique to obtain information about atomic structures in crystals from TEM images formed by phase contrast at high magnifications. Although HRTEM is just one of many functions in TEMs, several examples in sections below demonstrate that HRTEM often plays a decisive role in determining the local atomic arrangements in mica. An early study of mica by HRTEM was reported by Buseck and Iijima (1974). They clearly observed three dark lines representing a mica layer (the lines correspond to the two tetrahedral sheets and one octahedral sheet) and that cleavage was formed at the interlayer. During a quarter century after this pioneering work, many HRTEM studies for mica and related phyllosilicates have been reported (for instance, see the references in Baronnet 1992). These included many studies of mica, e.g., polytypism, transformations, defects and interface research. In the following section, recent HRTEM and related techniques are briefly reviewed. Next, two topics of HRTEM investigation, polytype and defect analyses are presented based on studies, mainly by the author and his colleagues. TEMS AND RELATED TECHNIQUES FOR THE INVESTIGATION OF MICA Transmission electron microscopy After the invention of TEM by E.E. Ruska in 1932, this apparatus was improved rapidly in response to requests from many fields of science. In the 1950s, lattice fringes in crystals were recorded (Menter 1956), which indicated an exciting possibility that a tool was possible to observe atomic arrangements in a crystal directly on a screen. Imaging theory for HRTEM developed in the 1960s showed that contrast in magnified images can be observed, which corresponds to the projection of the electrostatic potential in specimens with a resolution (referred to as “point resolution”: δ) defined by the following equation: δ = 0.66 Cs1/4 λ3/4 1529-6466/02/0046-0005$05.00

DOI:10.2138/rmg.2002.46.05

282

Kogure

where Cs is the spherical-aberration coefficient of the objective lens and λ is the wave length of the electron beam determined by accelerating voltage (Spence 1981). Following this equation, TEMs with an objective lens of low aberration and a small wavelength (a high accelerating voltage) were developed to achieve high resolution. In the 1970s, TEMs with accelerating voltages of ∼1 MV, and the resultant point resolution of 1 harmonics than for the fundamental one. The higher-order harmonic content in the synchrotron radiation beam is due to the intense continuum of the primary beam extending towards high energies, and it represents a significant contribution in all the high-energy third-generation synchrotron sources. Actually, rejection of higher-order harmonics may be obtained using either mirrors behaving as low-band pass filters, and/or by detuning crystals, or even by means of undulator sources. Optimization of spectra Orientation effect. Most experimental XANES spectra on micas were measured on powders, obtained by grinding hand-picked grains that had been gently settled on a flat sample-holder after dispersion in a liquid. The resulting mounts were considered to be randomly oriented, regardless of their grain-size homogeneity and distribution. However, experience gathered on other sheet-silicates (e.g., Manceau 1990; Manceau et al. 1988, 1990, 1998) has shown that, even in fine-grained powders, crystallite orientation strongly affects the shape of the final spectrum: primarily, it changes peak intensity, which is a significant component of the information and certainly reflects onto its quality (see above). If this is indeed the case, then among the mica XANES spectra performed so far (Table 2) only a few can be considered to be reliable. These include work by Osuka et al. (1988, 1990), Mottana et al. (1997) and Sakane et al. (1997), in which no special care was taken, but the investigated micas, being synthetic, were so homogeneously finegrained (1 μm) as to certainly lie on the sample-holder with their c axis more or less orthogonal to its surface and with their a and b axes oriented at random on it. A theoretical study of the orientation effect has been recently presented for selfsupporting clay-mineral thin films by Manceau et al. (1998), who also propose a tridimensional system of coordinates to record spectra in a standard setting. Their method, slightly modified by Cibin et al. (2001), has been adopted by Mottana et al. (in preparation) for single crystal mica blades (Fig. 6). Another approach used by Dyar et al. (in prep.) uses mica single crystals mounted on fibers in goniometer heads, which are then fitted onto a spindle stage mounted with the plane of rotation perpendicular to the path of the beam.

Sample surface

Figure 6. The coordinate system applicable to angular measurements on self-supporting phyllosilicate films as used for micas (Cibin et al. 2001; cf. Manceau et al. 1998, Fig. 2). Z-Y is the plane onto which the sample lies, with the X-ray beam impinging along X and linearly polarized on X-Y; α is the incidence (rotation) angle between the electric field vector ε and Y.

z

y hν

α x

ε

α

ε

For a perfectly random distribution of very small crystals (powder) there would be no angular variation effect on the experimental XAS spectra; however, for a fully oriented crystal structure such as that of a mica blade lying flat on the sample-holder, the amplitude of the scattered photoelectron wave depends on the angle α between the

388

Mottana, Marcelli, Cibin & Dyar

electric field vector ε of the impinging beam and the layers in the structure. This angle can be determined either by rotating the sample-holder on its vertical axis, or by preparing suitably oriented thin sections to be glued on the sample-holder in its routine setting orthogonal to the X-ray beam (α = 0°). Mottana et al. (in preparation) operated at SSRL at the 3-3 beamline (Hussain et al. 1982; Cerino et al. 1984), which is equipped with a double-crystal monochromator made Table 2. Published XAS data on mica species materials.

X-Ray Absorption Spectroscopy of the Micas

389

of efficient crystals such as YB66 (Wong et al. 1990 1999). They scanned single-crystal mica blades lying flat on the vertical sample-holder and optically-oriented in such a way as to have a ≅ b // Z. Here Z is an axis lying parallel to the mica surface (Fig. 6). The synchrotron beam first impinges the mica at right angle (α = 0°); then the blade is rotated and α increased up to 60∼80°, this being the maximum angle allowed by the mechanics of the sample compartment and the geometry of detection, which uses channeltrons. Therefore, the electric vector ε always lies on the horizontal plane, but it impinges two almost perpendicular sections of the mica structure so as to scan its atoms under different angles, with their atomic bonds and angles geometrically modified. The orientation effects observed in this way are clearly visible in a natural muscovite compositionally close to the end member (Fig. 7). It is quite clear that orientation dramatically affects the intensity of all peaks, including the white-line, but also— although to a much lesser extent—the positions of some of them, by as much a 5 eV. A comparison between Figure 7 and the Al K-edge spectrum reported by Mottana et al. (1997; cf. Fig. 4) for synthetic muscovite, which is expected to be randomly oriented owing to its very fine-grained powdery nature, shows that best agreement is attained for a rotation angle α in between 45 and 70°.

E (eV)

Figure 7. Changes in an Antarctica muscovite Al K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the white-line intensity as a function of the α rotation angle.

A similar comparison between the Fe K-edge spectra of a phlogopite single crystal rotated in the same way (Fig. 8) and the several Fe XANES spectra of phlogopites in the literature (Table 1) confirms that best agreement is obtained when the crystal is rotated at α ca. 45°. Indeed, later work (unpublished) showed that best agreement for the same sample, when scanned as both single crystal (at various angles) and as a settled homogeneous powder having a grain size of ca. 5 μm, is obtained when α is equal or very close to the “magic angle” value 54.7° (Pettifer et al. 1990). Changes with orientation are also clearly evident in the XANES spectra of a number of di- and tri-octahedral micas and one brittle mica, respectively at the Mg (phlogopite: Fig. 9), Si (muscovite: Fig. 10, and tetra-ferriphlogopite: Fig. 11), K (muscovite: Fig. 12), and Fe (clintonite: Fig. 13, and tetra-ferriphlogopite: Fig. 14) K edges. Such changes

390

Mottana, Marcelli, Cibin & Dyar

imply

Figure 8. Changes in a Franklin phlogopite Fe K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the changes undergone as a function of the α rotation angle by the edge and FMS regions (top) and by the IMS region (bottom). Philgopite Mg K edge

Figure 9. Changes in the FMS region of a Franklin phlogopite Mg K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam.

Absorption (Arb. Un.)

0º 30º 45º 70º

1300

1310

1320

1330

1340

Energy (eV) E (eV)

imply displacements in the peak positions from 0 up to 5 eV, and variation in the intensities by as much as 50%, with even reversals in the intensity of the edge top (Fig. 9) or appearance viz. disappearance (Fig. 11) of certain features. Most commonly, these changes occur gradually and trend always in the same direction, thus demonstrating their dependence upon the gradual rotation applied to the crystal. In turn, this rotation mostly reflects changes in the lengths of the bonds lying in the polarization plane, excited in the photoabsorption process, or in the lengths of multiple-scattering paths which are also probed in that geometry. Such spectral changes affect both the FMS and IMS regions, thus showing their dependence mostly upon the geometry of the section of the crystal that is being scanned by the synchrotron beam, as cosα. However, unexpected changes such as the one at the white-line in the phlogopite Mg K-edge spectrum (Fig. 9), or the sudden

X-Ray Absorption Spectroscopy of the Micas

391

Figure 10. Changes in an Antarctica muscovite Si K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the changes undergone by the edge and FMS regions.

Figure 11. Changes in a Tapira tetra-ferriphlogopite Si K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the changes undergone by the edge and FMS regions.

appearance of a new low-energy peak, as in the muscovite Al and Si K-edge spectra (Figs. 7 and 10) and in the tetra-ferriphlogopite Si K-edge spectrum (Fig. 11), demonstrate the possibility that the electronic properties of the absorbing atom are also involved. We have to underline here that this interpretation of the near-edge structure is fully equivalent to the interpretation that is based on local geometrical distributions, such

392

Mottana, Marcelli, Cibin & Dyar

lk

Figure 12. Changes in an Antarctica muscovite K K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the changes undergone by the edge and FMS regions.

Figure 13. Changes in a Lago della Vacca clintonite Fe K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the changes undergone by the edge (top) and IMS regions (bottom).

as those expected when the different local atomic distributions in the micas are being compared. To summarize, in order to obtain XANES spectra that may be meaningfully compared, we recommend orienting the sample, when a single crystal, always at the same angle of rotation α = 54.7°. This is essentially the same conclusion reached by Manceau et al. (1998) for the self-supporting clay films they experimentally investigated by

X-Ray Absorption Spectroscopy of the Micas

393

fffffffff

Figure 14. Changes in a Tapira tetra-ferriphlogopite Fe K-edge spectrum due to changing the orientation of the same crystal blade against the impinging, horizontally-polarized synchrotron radiation beam. In the right panel a magnified view of the changes undergone by the edge (top) and FMS and IMS regions (bottom).

polarized EXAFS and theoretically interpreted by performing full multiple-scattering calculations. Furthermore, we also recommend recording a full XAS spectrum of the same sample, after grinding it and settling in water for precisely determined times so as to obtain a well-classified powder possibly in the grain size range 1 to 2 μm. Dyar et al. (2000) used a different method of studying the orientation effects on the pre-edge region of Fe-bearing micas, with similar results. In that study, the microXANES probe at the National Synchrotron Light Source (NSLS), Brookhaven, NY, was used, allowing a beam size of 10 × 15 μm. Because the beam is smaller, samples on the order of 30 × 30 × 100 μm (orders of magnitude smaller than those used by other workers) could be studied, and concerns about sample homogeneity lessened. Each crystal was oriented with its cleavage perpendicular to a glass thin section, and then UVhardening epoxy was used to maintain it in that geometry. The mica+epoxy was removed from the thin section, and two mutually parallel faces were polished on each sample perpendicular to cleavage (though in an unknown orientation relative to the a and b axes: see Fig. 15). This preparation permitted acquisition of spectra in two important directions

Figure 15. Optical orientation of a model mica crystal showing the random position of the thin section cut across cleavage and used for microXANES measurements (Dyar et al. 2001).

394

Mottana, Marcelli, Cibin & Dyar

perpendicular and parallel to cleavage by simple rotation of the sample. A further advantage of this method is that parallel studies of the optical and IR absorption spectra of the identical crystals could be made. In more recent work (Dyar et al. in prep.) single crystals were analyzed while mounted on goniometer heads, so the beam could be polarized along the X, Y, and Z optical orientations. As with the work of Mottana et al. (in preparation), changes in peak intensity and, to a lesser extent, energy, were observed by Dyar et al. (2001) as a function of sample orientation. At the main edge, the difference in the intensity of the highest energy peak relative to the other prominent peak or peaks is generally greatest when the synchrotron beam is polarized in the direction of the cleavage plane, with a few exceptions. In the preedge region, intensity variations were also observed, but the maxima and minima were not necessarily parallel or perpendicular to cleavage, and the orientation at which maximum intensity occurred was different for various samples. This implies that there are variations in peak intensity not only perpendicular and parallel to the mica cleavages, but also within the sheets themselves as a function of orientation with respect to the unconstrained position in the XY plane. Such a conclusion is not surprising in a monoclinic mineral species: the XANES probe is sampling different bonds at different orientations relative to noncentrsymmetric Fe sites (Dyar et al. 2001; in prep.). Spectrum fitting. In standard XAS experiments, signal to noise (S/N) ratios in the range 103∼104 can be achieved. However, to fully enhance XANES potentials, these are not enough, especially in the soft-X-ray energy range where such ratios are only achieved after a perfect preparation of the sample. Consequently, with a lower S/N ratio, the best understanding of XANES critically depends upon a careful fitting of the experimental spectrum during which no fine details get lost. The standard procedure in XAS spectrum analysis follows two steps: the experimental spectrum is (1) corrected for background contributions from lower energy absorption edges by linear or polynomial fitting of the base line, then (2) normalized at high energy, i.e., close to the upper end of the XANES region at an energy position where no obvious features can be seen. In addition, for pre-edge analysis the contribution of the absorption jump is subtracted by an arctangent function. This procedure leaves a profile of the entire K-edge region that consists of a number of features, occasionally partially superimposed, that can be either evaluated visually or fitted by Gaussian or Lorentzian curves. The numerical values of the fitted curves (energy and intensity, with errors and significance bars) can then be used as solid data for interpretation. This standard procedure assures accuracy in energy position ±0.1 eV for the pre-edge, and ±0.03 eV for all other regions of the XANES spectrum. Both values are well within resolution, which increases with energy from ca. 0.3 to ca. 1.5 eV on going from the Na K-edge to the Fe one (Schaefers et al. 1992). Accuracy in the intensity measurements is estimated to be better than 10%. However, such intense structures as the "white line" are affected mainly by the harmonics content. At all synchrotron sources, a step preliminary to all this standard procedure consists of calibrating the energy positions of all peaks against standards (usually metal foils). An alternative way is to calibrate them against a “glitch”, i.e., a spurious absorption at constant energy in the spectra that is due to a planar defect present in the monochromator crystal (cf. Wong et al. 1999, for YB66). When high thermal loads heat the monochromator crystals, a further systematic correction is applied that takes into account the decrease of the ring current (and heat load) with time. As a matter of fact, in most mica studies a careful fitting procedure is seldom applied, and the “fingerprinting” method of evaluation is still predominant (Table 1). A

X-Ray Absorption Spectroscopy of the Micas

395

recent improvement in the fitting procedure is based upon a novel software (Benfatto et al. 2001). However minor the error induced by such evaluation may be, any concomitant carelessness in taking into account orientation effects would, at the end, result in crowding the literature with spectra useless for interlaboratory comparisons. Systematics XAS studies on micas: a catalogue. Table 1 lists all XAS studies carried out on micas that could be retrieved in the relevant literature. They are presented in the alphabetical order of the di- and tri-octahedral mica species nomenclature approved recently (Rieder et al. 1998) and are further subdivided on the basis of the investigated atom. Almost all investigated samples are natural and are therefore intermediate in composition. However, some of them are close enough to end member compositions as to make it possible to classify them accordingly. Only seven true end members corresponding to natural mica species have been studied so far by XAS, i.e., the Tapira tetra-ferriphlogopite (Giuli et al. 2001) and the six synthetic micas investigated by Mottana et al. (1997). Even all other synthetic micas (Osuka et al. 1988 1990; Sakane et al. 1997) are intermediate, as they are doped crystals obtained for technological purposes. Furthermore, among the synthetic micas quite a few have no natural counterpart (Soma et al. 1990; Han et al. 2001). XAS studies on otherwise insufficiently characterized samples, or on samples with composition being complex solid solutions from the crystal-chemical viewpoint, are listed at the bottom of Table 1, in the section that accounts for the approved series names (cf. Rieder et al. 1998 Table 4). The first XAS spectra ever recorded on micas were those by Brytov et al. (1979) at the Si and Al K edges. However, as all these spectra were recorded in the late 1970s and early 1980s using a conventional X-ray tube as the source, they are practically useless for present-days studies because of the limited resolution: in practice, only the general shape is worth examining (e.g., Jain et al. 1980 Fig. 1). Nevertheless, these early attempts deserve to be remembered, for both the pioneering effort they record and their historical significance. The earliest synchrotron-activated experimental XAS spectrum for any mica was Calas et al.’s (1984) chromium muscovite at the Cr K edge. Although noisy, particularly in the pre-edge region, this spectrum satisfactorily compares with the recent spectrum of a similar mica at the same edge (Brigatti et al. 2001; see below), thus suggesting not only the high level of technical skill of the operators, but also that comparison of power spectra collected at very different times and on widely different synchrotron storage rings can be confidently made, provided the basic requirements of energy calibration and background subtraction were carefully applied (see above). Occasionally, mica has been used also to support epitaxially-grown layers that have been investigated by XAS (e.g., Blum et al. 1986; Drozdov et al. 1997). Although reported in Table 1, these XAS studies actually do not belong to mica studies. Finally, there has never been a spectrum published so far but those presented above to which the above-given precautions on orientation effects were applied (see also Dyar et al. 2001 and in prep.). Even the spectra that will be described in the following were obtained on ground powders, presumed to be homogeneous in their grain-size and randomly oriented, but never tested for those conditions. Determination of the oxidation state. Determining the effective charge on the absorbing atom from the chemical shift of the X-ray absorption threshold is a fundamental issue for XANES. However, a direct measure of the "ionization threshold" or "continuum threshold" (i.e., the energy at which the electron is excited in the

396

Mottana, Marcelli, Cibin & Dyar

continuum: e.g., the Fermi level in metals) is not possible because of the lack of any signature of it. Therefore, XANES is not a direct probe of core-level binding energy as other methods are (e.g., XPS or ESCA). However, there is evidence in both gas molecules and solid compounds that the energy shift of the first bound excited state at the absorption threshold follows the binding energy shift of the core level. Moreover, a linear dependence between core-level binding energy and atomic effective charge has been measured (Belli et al. 1980). By contrast, no linear relationship between the measured shift of the first strong multiple-scattering resonance and the effective atomic charge on the ion exists. The energy of multiple-scattering resonances is strongly dependent on interatomic distance, so their chemical shifts are much larger than that of the core excitation. Actually, the variation of the effective charge on an atom is often increased and a linear correlation with core-level binding energy indeed exists; however, this effect is always system-dependent. Moreover, within the same structure any correlation among the parameters of the potential is certainly confined only to small changes of the interatomic distances (e.g., less than 10%). Correct identification of the oxidation state of 3d transition metals is indeed important, but the quantification of the oxidation ratio is even more important in the case of potentially multivalent minerals such as the micas, a group where the number of elements occurring with more than one oxidation state is significant (Fe, Mn, Cr, V and possibly Ti: cf. Table 1) and their amounts may be so large as to even become essential and determine new end members. All transition element K-edge spectra display a preedge (Belli et al. 1980) and, mostly, all features of the pre-edge are strong enough to be easily recorded experimentally. Position and intensity of the peaks occurring in the preedge region can be reliably used to determine the oxidation state(s) of the absorbing atom (e.g., Waychunas 1987). However, as already seen (Fig. 2, above), coordination too plays a role, so that care must be made in discriminating the two effects, and to this purpose spectra need to be properly deconvoluted. As discussed above, the energy position of the peaks in the pre-edge region may be directly related to the increase in the oxidation state of the absorber atom: e.g., the preedge feature of Fe3+ is generally ca. 2∼3 eV higher in energy than the corresponding feature for Fe2+ (Waychunas et al. 1983; cf. Petit et al. 2001). The amount of such a “chemical shift” is different for the different transition elements, and depends on the final state reached by the electron. Implicitly, this weakens the possibility of reliably determining the oxidation state of a given atom when it occurs in different coordination sites of the same compound. However, when a significant part of the atom occurs in a tetrahedrally-coordinated site, the relevant pre-edge is strongly intensified owing to d-p mixing, and the determination of the oxidation state of the tetrahedral atom is made fairly easy to measure: e.g., amounts of Cr3+ in tetrahedral coordination as small as 0.5% could be detected even in the presence of a significant amount of Cr3+ in octahedral coordination (Brigatti et al. 2001; see below). Consequently, subtraction of the tetrahedrally-coordinated component can be made. The residual pre-edge spectrum of the octahedrally-coordinated atoms is then de-convoluted into its components to determine their oxidation state(s). Bajt et al. (1994 1995) and Sutton et al. (1995) have pushed the practice of pre-edge examination further to reach an effective quantification of the oxidation states for Fe, the atom which most frequently occurs in two oxidation states in the same site of minerals. They have developed, and Galoisy et al. (2001) and Petit et al. (2001) have recently improved upon, a procedure that makes use of the known positions of pre-edge peaks of Fe K-XANES spectra in mineral standards to fit a calibration line giving the Fe3+/ΣFe ratios of various minerals (Fig. 16).

X-Ray Absorption Spectroscopy of the Micas

397

Figure 16. Plot of “pre-edge peak energy” vs. “Fe3+/ΣFe” for well characterized standards. The trend is linear with a correlation coefficient of 0.99 (after Sutton et al. 1995, p. 1465, Fig. 3).

However, extensive additional work by Dyar et al. (2001) on suites of Fe3+ and Fe2+ end members confirms that the energies of the end-member pre-edges vary considerably for several different mineral groups, and thus no single mineral species can be used to model all cases of any type of Fe (Fig. 17). Because different mineral groups have variably distorted coordination polyhedra, use of mineral group-specific standard end members will ultimately be necessary to interpret pre-edge positions assigned to different transitions. Examples of using this method to determine of the Fe3+/ΣFe ratios of a number of rock-forming micas are given elsewhere (Dyar et al. 2001). Determination of local coordination geometry. The position and intensity of the peaks in the pre-edge region do not solely depend upon the oxidation state of the absorber transition metal, but also upon the shape of the site (coordination polyhedron) where the absorber is located in the structure (Calas and Petiau 1983). An increase in coordination number provokes a positive energy shift, while the intensity of the peak is proportionally reduced (Waychunas et al. 1983). The first attempt at using the pre-edge features to determine quantitatively site geometry is Waychunas’ (1987) for the Ti K-edge of a suite of silicate and oxide minerals, including a biotite from Antarctica. He fitted Gaussian features to the entire edge region, and found that individual features are insensitive to changes in the Ti-O bond length, but sensitive to valence, with Ti3+ at ca. 2.0 eV lower energy than Ti4+. Moreover, the intensity of the second pre-edge feature at ca. 4969 eV turned out to be sensitive to both octahedral site distortion and to presence of tetrahedral Ti4+. A correlation was found for silicates between intensity and bond-angle variance σ2 in the octahedral Ti site, and for biotite σ2 could be quantified to be ca. 30 deg2, in fair agreement with the value computed from the X-ray diffraction crystal structure determination (Ohta et al. 1982). Cruciani et al. (1995) essentially followed the same

398

Mottana, Marcelli, Cibin & Dyar

gggg

Figure 17. Variation of the absolute pre-peak energy vs. Fe3+ content in the endmembers of several mineral groups; after Delaney et al. (in preparation).

SrCrO4

Absorption (arb.units)

Anatoki river

Westland Uvarovite

5990

6000

6010

6020

6030

6040

6050

Figure 18. Experimental Cr K-edge spectra for the Anatoki River and Westland E (eV) chromium muscovites, a synthetic SrCrO4 standard for tetrahedral Cr6+ (top) and an Outukumpu uvarovite standard for octahedral Cr3+ (bottom). See text for discussion (Brigatti et al. 2001, Fig. 6).

X-Ray Absorption Spectroscopy of the Micas

399

Figure 19. Pre-edge fit of the Westland chromium muscovite Cr K-edge spectrum of Figure 18 and (inset) its de-convolution in two Gaussian components (Brigatti et al. 2001 Fig. 7).

procedure when trying to determine the [4]Fe3+ contents of a series of natural phlogopites, but came to a purely speculative result owing to the insufficient resolution of the monochromator crystal and the extremely low amount of sample available. As an example of successful evaluation, we report the case of two chromium muscovites worked out by Brigatti et al. (2001) at the Cr K pre-edge; the procedure they followed is the one developed by Peterson et al. (1997) for oxides. The Anatoki River and Westland chromium muscovites Cr K-edge spectra were compared with a synthetic SrCrO4 standard, for tetrahedral Cr6+, and a natural uvarovite, for octahedral Cr3+ (Fig. 18). The Anatoki River muscovite Cr K-edge spectrum proved to be too noisy for further evaluation, but the Westland one, after subtraction of the edge contribution by a pseudoVoigt function, had its pre-edge resolved in two Gaussian components: at 5991.3 eV and 5994.0 eV, respectively (Fig. 19). The second Gaussian component appears in the experimental spectrum only as a skew tail at the end of the pre-edge, owing to interference with the rapidly rising slope leading to the edge. However, after subtracting this interference, it can be reliably measured for both energy and intensity. The evaluation step that follows involves interpretation. If the second-component intensity is assumed to be the same as that of the single, symmetrical Gaussian pre-edge feature of a SrCrO4 standard in which the Cr6+ is entirely in tetrahedral coordination, then it can be appraised that amount of [4]Cr in muscovite, if any, cannot exceed 0.4-0.5% of total Cr (cf. Lee et al. 1995). By contrast, if both Gaussian components are considered to be due to [6]Cr3+, as in the uvarovite standard, and interpreted as a way to measure the distortions of the muscovite octahedral sites where Cr3+ is possibly hosted, then their relative

400

Mottana, Marcelli, Cibin & Dyar

intensities (1.3 and 1.2% nau [= normalized absorption units]) show that these two sites are very similar. Indeed, this is nothing more than an extension to Cr of the method for quantitatively determination of site distortion for octahedra centered by Ti4+ calibrated by Waychunas (1987). In the case of the already-mentioned Fe K pre-edge of tetra-ferriphlogopite, where Fe3+ is entirely in the tetrahedral site, the pre-edge is twice as strong and shifted to higher energy (ca. 2 eV) relative to annite, where Fe is mostly in the octahedral site (Fig. 2). This apparent irregularity can be explained by comparing the sharp single peak of tetraferriphlogopite, a synthetic endmember, and the broad, probably double peak of the Pikes Peak annite, the Fe of which is entirely octahedral, but partly Fe2+ and partly Fe3+. Clearly, the oxidation effect is more important than the coordination effect in determining the position of the Fe K pre-edge. However, the strong intensity of the tetraferriphlogopite peak also suggests that its Fe is constrained in a more tightly-bound coordination polyhedron than the annite one. Note, however, that there is an underlying problem in the pre-edge region that needs a more careful evaluation, and not only in these systems: this problem is the amount of quadrupolar effects present (see Giuli et al. 2001, for additional evaluation). A1 K edge

Absorption (Arb. Un.)

Grossularia

Polilithionite

Phlogopite

Albite

1560

1565

1570

1575 1580 E (eV)

1585

1590

1595

Figure 20. Shift of the white-line in the FMS region of the Al K-edge spectra of two synthetic micas as a result of two different coordination geometries: in phlogopite the Al atoms are entirely in a tetrahedral site geometry, and in polylithionite in an octahedral site geometry, as they are in the reference albite and grossular natural standards, respectively (Mottana et al. 1997, Fig. 3).

Coordination geometry also plays a role in shaping the FMS region of a XANES spectrum. This effect was clearly documented for the Al K edges of certain synthetic micas by Mottana et al. (1997), who showed that there is a shift of at least 2 eV between [4] Al as in phlogopite and albite, and [6]Al as in polylithionite and grossular (Fig. 20). Moreover, they found that it is possible, although difficult, to recognize the concomitant presence in the spectra of two white-line features arising from contributions of the same atom occurring in two different geometries ([4]Al and [6]Al in zinnwaldite and

X-Ray Absorption Spectroscopy of the Micas

401

preiswerkite: Mottana et al. 1997 Fig. 4). Thus, the FMS region of the XANES spectrum of a mineral with Al in two coordinations can be seen as the weighted combination of the contributions arising from the two Al atoms, although the general appearance of the spectrum (and its ensuing evaluation) is somewhat blurred by next nearest neighbor effects due to the presence of other atoms in the same sites substituting for the absorber Al (cf. the muscovite vs. bityite spectra: Mottana et al. 1997 Fig. 4). In the following we will document visually and sparingly comment upon a series of XANES spectra obtained at different K edges for the powders of a number of natural micas close to the end members. The present state of our investigation, which is still under way, compels us to defer to a later moment for drawing conclusions (Mottana et al., in preparation): micas are no simple systems, and XAS literature is already cluttered by faulty reasoning and wrong conclusions reached when hastily evaluating even simpler systems!

Mg K edge

Absorption (Arb. Un.)

Phlogopite

Tetra-ferriphlogopite Biotite Clintonite

1290

1300

1310

1320

1330

1340

E (eV)

Figure 21. Experimental Mg K-edge spectra for the powders of four natural tri-octahedral micas.

Figure 21 shows the experimental Mg K-edge spectra of three tri-octahedral micas (phlogopite, tetra-ferriphlogopite, and biotite) and one brittle mica (clintonite). All spectra are very similar and have no pre-edges, as magnesium is not a transition element. The FMS regions consist of three features, like the K edge of talc (Wong et al. 1995). However, the relative intensities of the three features differ significantly among the four spectra suggesting that there are substantial differences in the local order of their Mg that may be resolved via comparison with spectra taken for other absorbers. Note, moreover, that the three features in the clintonite spectrum are possibly doubled. Figure 22 shows the experimental Al K edge spectra of three tri-octahedral micas (phlogopite, annite, biotite) and one di-octahedral mica (muscovite). Again, Al is not a transition element, therefore the spectra have no distinct pre-edges. The FMS regions are apparently simpler than the ones occurring in the Mg K-edge spectra above, but in fact

402

Mottana, Marcelli, Cibin & Dyar

they Al K edge

Absorption (Arb. Un.)

Phlogopite

Annite

Biotite

Muscovite

1550

1560

1570

1580

1590

1600

1610

1620

E (eV)

Figure 22. Experimental Al K-edge spectra for the powders of three natural tri-octahedral and one natural di-octahedral mica.

they contain the same three features, although with strongly different intensities and energies (cf. Mottana et al. 1999). Possibly, the fact that non-precisely oriented powders were used affects the recorded features (cp. this muscovite spectrum with that in Fig. 7). The IMS regions are poor in features, but they display shifts and relative differences that are enormous, considering the similarity of the local structures that originate such differences. The significant role of the outer shells around the Al absorber appears to be well depicted here, but it will create great problems when interpreting the spectra from a quantitative viewpoint. Figure 23 shows the experimental Si K-edge spectra of five micas: four tri- and one di-octahedral one. Nowhere is there a pre-edge, and the entire XANES spectrum is dominated by the strong white-line of Si in tetrahedral coordination (cf. Li et al. 1994; Li et al. 1995a). The regions in between FMS and IMS (inset) undergo subtle but significant variations as a result of changes in the local and medium-range ordering occurring in the relevant structures for the volumes that surround the Si tetrahedra. Such variations may also occur in the energies of certain peaks, but this variation is also certainly due to the tri- vs. di-octahedral structure of the investigated mica (inset: cf. muscovite with the other micas). The experimental K K-edge spectra of the same five micas are shown in Figure 24. These XANES spectra are rather complex, both to record experimentally and to reckon with. The FMS regions have no strong white-lines, and only small differences show up in the intensities of their IMS regions (inset). However, their analysis suggests that the K coordination number is less than the expected 12, possibly 8 or even 6. In a case like this, only XANES simulations by the multiple-scattering code may be able to reveal safely the actual site geometry around the potassium atom. Finally, Figure 25 shows the experimental Fe K-edge spectra of two trioctahedral

X-Ray Absorption Spectroscopy of the Micas

403

mu

Figure 23. Experimental Si K-edge spectra for the powders of five natural micas. The strong differences displayed by a portion of their FMS and IMS regions is shown as inset.

micas (biotite and tetra-ferriphlogopite) and one brittle mica (clintonite). Iron is a transition element, therefore all spectra exhibit significant pre-edges (inset), each one of them having properties of its own. In particular, the tetra-ferriphlogopite pre-edge is a singlet (cf. Fig. 2), as is the clintonite one, but at 1 eV lower energy. Fe is tetrahedrallycoordinated in both micas, but in the former one it is Fe3+ and in the latter one an additional contribution arising from Fe2+ is likely. The biotite pre-edge is weak, because it mostly arises from octahedral Fe2+. The three pre-edges require a deconvolution of the same sort as the one previously demonstrated for the Cr pre-edge of muscovite (Fig. 19) in order to reveal all the information they contain. The FMS regions of these spectra are dominated by the Fe white-line, which undergoes energy variations accounting for differences in both coordination and oxidation state. The presence of significant variations in the medium- to long-range ordering occurring in these mica structures is made evident by their greatly different IMS regions (and also by their EXAFS regions: cf. Giuli et al. 2001).

404

Mottana, Marcelli, Cibin & Dyar

Figure 24. Experimental K K-edge spectra for the powders of five natural micas. The strong differences displayed by a portion of their FMS and IMS regions are shown as an inset.

ACKNOWLEDGMENTS Our XAS work on minerals has enjoyed the support of numerous suggestions, discussions and contributions in many stages and levels over a number of years, the five more recent ones dedicated mostly to the micas. We thank all these colleagues, since it is by this form of synergy that we could carry out and develop our project over the years. A special thank goes to Maria Franca Brigatti, Jesús Chaboy, Paola De Cecco, Giancarlo Della Ventura, Gabriele Giuli, Antonio Grilli, Cristina Lugli, Jeff Moore, Takatoshi Murata, Eleonora Paris, Marco Poppi, Agostino Raco, Jean-Louis Robert, Claudia Romano, Michael Rowen, Francesca Tombolini, Hal Tompkins, Curtis Troxel, Joe Wong, Ziyu Wu and all others who allowed us to use for this review some of the data recorded together during painstaking sessions at the source. Most experimental XAS was carried out at SSRL, which is operated by Stanford University on behalf of D.O.E. Furthermore, M.D.D. acknowledges the insight and assistance of her collaborators at the N.S.L.S., Brookhaven National Laboratory: Jeremy Delaney, Tony Lanzirotti and Steve Sutton. Financial supports for our experimental work and for its evaluation and interpretation were granted by M.U.R.S.T. (Project COFIN 1999 “Phyllosilicates: crystalchemical, structural and petrologic aspects”), C.N.R. (Project 99.00688.CT05 “Igneous and metamorphic micas”), and I.N.F.N. (Project “DAΦNE-Light”) in Italy, and by N.S.F.

X-Ray Absorption Spectroscopy of the Micas

405

EAR-9909587 and EAR-9806182, and D.O.E.-Geosciences DE-FG02.92ER14244 in U.S.A. Critical readings by C.R. Natoli and a unknown referee improved the quality of this paper in a substantial manner.

Figure 25. Experimental Fe K-edge spectra for the powders of two natural trioctahedral micas and one natural brittle mica. The pre-edge regions are shown as inset.

REFERENCES Bair RA, Goddard WA (1980) Ab initio studies of the X-ray absorption edge in Cu complexes. I. Atomic Cu2+ and Cu(II)Cl2. Phys Rev (Condens Matt) B22:2767-2776 Bajt S, Sutton SR, Delaney JS (1994) Microanalysis of iron oxidation states in silicates and oxides using X-ray absorption near-edge structure (XANES). Geochim Cosmochim Acta 58:5209-5214 Bajt S, Sutton SR, Delaney JS (1995) Microanalysis of iron oxidation states in Earth and planetary materials. Physica B208&209:243-244 Balzarotti A, Comin F, Incoccia L, Piacentini M, Mobilio S, Savoia A (1980) K-edge absorption of titanium in the perovskites SrTiO3, BaTiO3, and TiO2. Solid State Commun 35:145-149 Bassett WA, Brown GE Jr (1990) Synchrotron radiation in the Earth sciences. Ann Rev Earth Planet Sci 18:387-447 Belli M, Scafati A, Bianconi A, Mobilio S, Palladino L, Reale A, Burattini E (1980) X-ray absorption nearedge structures (XANES) in simple and complex Mn compounds. Solid State Commun 35:355-361 Benfatto M, Congiu Castellano A, Daniele A, Della Longa S (2001) MXAN: A new software procedure to perform geometrical fitting of experimental XANES spectra. J Synchrotron Rad 8:267-269 Benfatto M, Natoli CR, Brouder C, Pettifer RF, Ruiz Lopez MF (1989) Polarized curved wave EXAFS: Theory and application. Phys Rev B39:1936-1939 Benfatto M, Natoli CR, Bianconi A, Garcia J, Marcelli A, Fanfoni M, Davoli I (1986) Multiple scattering regime and higher order correlations in X-ray absorption spectra of liquid solutions. Phys Rev B34:5774-5781 Bianconi A (1988) XANES spectroscopy. In DC Koningsberger, R Prins (eds) X-ray Absorption: Principles, applications, techniques of EXAFS, SEXAFS, and XANES. Wiley, New York, p 573-662

406

Mottana, Marcelli, Cibin & Dyar

Bianconi A, Dell'Ariccia M, Durham PJ, Pendry JB (1982) Multiple-scattering resonances and structural effects in the X-ray absorption near-edge spectra of FeII and FeIII hexacyanide complexes. Phys Rev B26:6502-6508 Bianconi A, Fritsch E, Calas G, Petiau J (1985) X-ray absorption near-edge structure of 3d transition elements in tetrahedral coordination. The effects of bond-length variation. Phys Rev B32:4292-4295 Bianconi A, Incoccia L, Stipcich S (1983) EXAFS and Near-edge Structure (Springer Ser Chem Phys Vol. 27). Berlin, Springer-Verlag Blum L, Abruna HD, White J, Gordon JG II, Borges GL, Samant MG, Melroy OR (1986) Study of underpotentially deposited copper on gold by fluorescence detected surface EXAFS. J Chem Phys 85:6732-6738 Bonnin D, Calas G, Suquet H, Pezerat H (1985) Site occupancy of Fe3+ in Garfield nontronite: a spectroscopic study. Phys Chem Minerals 12:55-64 Borowski M, Bowron DT, De Panfilis S (1999) High-energy X-ray absorption spectroscopy at ESRF BM29. J Synchrotron Rad 6:179-181 Brigatti MF, Galli E, Medici L, Poppi L, Cibin G, Marcelli A, Mottana A (2001) Chromium-containing muscovite: crystal chemistry and XANES spectroscopy. Eur J Mineral 13:377-389 Brown GE Jr, Calas G, Waychunas GA, Petiau J (1988) X-ray absorption spectroscopy: Applications in Mineralogy and Geochemistry. Rev Mineral 18:431-512 Brown GE Jr, Farges F, Calas G (1995) X-ray scattering and X-ray absorption studies of silicate melts. Rev Mineral 32:317-410 Brown GE Jr, Keefer KD, Fenn PM (1978) Extended X-ray absorption fine structure (EXAFS): study of iron-bearing silicate glasses: iron coordination environment and oxidation state. Geol Soc Am Ann Meet Abstr Progr 10:373 Brown GE Jr, Parks GA (1989) Synchrotron-based X-ray absorption studies of cation environments in Earth materials. Rev Geophys 27:519-533 Brytov IA, Konashenok KI, Romashchenko YuN (1979) Crystallochemical effects on Al K and Si K emission and absorption spectra for silicate and aluminosilicate minerals. Geochem Int’l 16:142-154 Cabaret D, Joly Y, Renevier H, Natoli CR (1999) Pre-edge structure analysis of Ti K-edge polarized X-ray absorption spectra in TiO2 by full-potential XANES calculations. J Synchrotron Rad 6:258-260 Cabaret D, Sainctavit P, Ildefonse Ph, Flank A-M (1996) Full multiple-scattering calculations on silicates and oxides at the Al K edge. J Phys: Condens Matt 8:3691-3704 Calas G, Brown GE Jr, Waychunas GA, Petiau J (1984) X-ray absorption spectroscopic studies of silicate glasses and minerals. Phys Chem Minerals 15:19-29 Calas G, Levitz P, Petiau J, Bondot P, Loupias G (1980) Ètude de l’ordre local autour du fer dans les verres silicatés naturels et synthétiques à l’aide de la spectrométrie d’absorption X. Rev Phys Appl 15: 1161-1167 Calas G, Manceau A, Novikoff A, Boukili H (1984) Comportement du chrome dans les minéraux d’altération du gisement de Campo Formoso (Bahia, Brésil). Bull Minéral 107:755-766 Calas G, Petiau J (1983) Coordination of iron in oxide glasses through high-resolution K-edge spectra: information from the pre-edge. Solid State Commun 48:625-629 Cerino J, Stöhr J, Hower N, Bachrach R (1984) An ultra-high-vacuum double crystal monochromator beam line for studies in the spectral range 500-4000 eV. Nucl Instr Meth Phys Res 172:227-236 Chaboy J, Marcelli A, Tyson TA (1994) Influence of double-electron transitions on the EXAFS I edges of rare-earth systems. Phys Rev B49:11652-11661 Cibin G, Marcelli A, Mottana A (2001) Polarised X-ray absorption spectra at the Al K edge in Na+ β- and β”-aluminas. Europhys Lett (submitted) Clementi E, Roetti C (1974) Roothan-Hartree-Fock atomic wavefunctions. In Atomic Data and Nuclear Data Tables 14:177-478. Academic Press, New York Cressey G, Henderson CMB, van der Laan G (1993) Use of L-edge X-ray absorption spectroscopy to characterize multiple valence states of 3d transition metals; a new probe for mineralogical and geochemical research. Phys Chem Minerals 20:111-119 Cruciani G, Zanazzi PF, Quartieri S (1995) Tetrahedral ferric iron in phlogopite: XANES and Mössbauer compared to single-crystal X-ray data. Eur J Mineral 7:255-265 Dartyge E, Fontaine A, Tourillon G, Cortes R, Jucha A (1986) X-ray absorption spectroscopy in dispersive mode and by total reflection. Phys Lett 113A:384-388 Davoli I, Paris E (1990) Principles and recent developments of XANES spectroscopy. In A Mottana, F Burragato (eds) Absorption Spectroscopy in Mineralogy. Elsevier, Amsterdam, p 206-226 Dehmer JL, Dill D (1976) Molecular effects on inner-shell photoabsorption. K-shell spectrum of N2. J Chem Phys 65:5327-5334 Delaney JS, Sutton SR, Newville M, Jones JH, Hanson B, Dyar MD, Schreiber H (2000) Synchrotron micro-XANES measurements of vanadium oxidation state in glasses as a function of oxygen fugacity:

X-Ray Absorption Spectroscopy of the Micas

407

experimental calibration of data relevant to partition coefficient determination. 31st Ann Lunar Planet Sci Confe Dell'Ariccia M, Gargano A, Natoli CR, Bianconi A (1984) A calculation of C K-shell X-ray absorption spectra of C2Hn (n = 2, 4, 6) oriented molecules: correlation between position of the multiple scattering resonance in the continuum and the C-C bond length. LNF-INFN Report LNF-84-51 (P) Deutsch M, Hart M (1984) Single X-ray photon excitation of two and three electrons in nickel. Phys Rev A29: 2946-2948 Díaz-Moreno S, Muñoz-Páez A, Chaboy J (2000) X-ray absorption spectroscopy (XAS) study of the hydratation structure of yttrium(III) cations in liquid and glassy states: eight or nine-fold coordination? J Phys Chem A2000:1278-1286 Doyle CS, Traina SJ, Ruppert H, Kandelewicz T, Rehr JJ, Brown GE Jr (1999) XANES studies at the Al K-edge of aluminum-rich surface phases in the soil environment. J Synchrotron Rad 6:621-623 Dräger G, Frahm R, Materlik G, Brümmer O (1988) On the multiplet character of the X-ray transitions in the pre-edge structure of Fe K absorption spectra. Phys Status Solidi 146:287-294 Drits VA, Dainyak LG, Muller F, Besson G, Manceau A (1997) Isomorphous cation distribution in celadonites, glauconites and Fe-illites determined by infrared, Mössbauer and EXAFS spectroscopies. Clay Minerals 32:153-179 Drozdov YuN, Klyuenkov EB, Lentovskaya NE, Mazo LA, Suslov LA (1997) Fullerene C60 layers on mica and sapphire. Poverkhnost (3):8-13 528 Durham PJ, Pendry JB, Hodges VH (1982) Calculation of X-ray absorption spectra based on the multiple scattering theory. Comput Phys Commun 25:193-205 Dyar MD, Delaney JS, Sutton SR (2000) Advances in interpretation of Fe XANES pre-edge spectra, and resultant improvements in microanalysis of ferric/ferrous ratios on thin sections. 31st Ann Lunar Planet Sci Conf Dyar MD, Delaney JS, Sutton SR (2001) Fe XANES spectra of iron-rich micas. Eur J Mineral 13: 1079-1098 Dyar MD, McCammon C, Schaefer MW (eds) (1996) Mineral Spectroscopy: A Tribute to Roger G. Burns. Geochem Soc Spec Publ No. 5, 465 p Filipponi A, Di Cicco A (2000) GNXAS: a software package for advanced EXAFS multiple-scattering calculations and data-analysis. Task Quarterly 4:575-669 Filipponi A, Di Cicco A, Benfatto M, Natoli CR (1990) The three-body correlation function in amorphous silicon probed by X-ray absorption spectroscopy. Europhys Lett 13:319-325 Filipponi A, Di Cicco A, Tyson TA, Natoli CR (1991) Ab initio modeling of X-ray absorption spectra. Solid State Commun 78:265-275 Filipponi A, Evangelisti F, Bernieri E, Mobilio S (1987) Double-electron excitation at the Si K-edge of amorphous silicon. J Phys Colloq 48, C9:961-964 Fricke H (1920) K-characteristic absorption frequencies for the chemical elements magnesium to chromium. Phys Rev 16:202-215 Fuhrmann M, Bajt S, Schoonen MAA (1998) Sorption of iodine on minerals investigated by X-ray absorption near-edge structure (XANES) and 125I tracer sorption experiments. Appl Geochem 13: 127-141 Galoisy L, Calas G, Arrio MA (2001) High-resolution XANES spectra of iron in minerals and glasses: structural information from the pre-edge region. Chem Geol 174:307-319 Garcia J, Benfatto M, Natoli CR, Bianconi A, Davoli I, Marcelli A (1986) Three particle correlation function of metal ions in tetrahedral coordination determined by XANES. Solid State Commun 58:595-599 Giuli G, Paris E, Wu ZY, Brigatti MF, Cibin G, Mottana A, Marcelli A (2001) Experimental and theoretical XANES and EXAFS study of tetra-ferriphlogopite. Eur J Mineral 13 (in press) Greaves GN, Durham PJ, Diakun G, Quinn P, Hart M, Siddons DP (1983) An order-sorting monochromator for synchrotron radiation. Nucl Instr Meth Phys Res 208:335-339 Gudat W, Kunz C (1972) Close similarity between photoelectric yield and photoabsorption spectra in the soft-X-ray range. Phys Rev Lett 29:169-172 Gurman SJ (1989) Structural information in extended X-ray absorption fine structure (EXAFS).In SS Hasnain (ed) Synchrotron Radiation and Biophysics. Ellis Horwood, Chichester, UK, p 9-42 Guttler B, Niemann W, Redfern SAT (1989) EXAFS and XANES spectroscopy study of the oxidation and deprotonation of biotite. Mineral Mag 53:591-602 Hahn JE, Scott RA, Hodgson KO, Doniach S, Desjardins SE, Solomon EI (1982) Observation of an electric quadrupole transition in the X-ray absorption spectrum of a Cu(II) complex. Chem Phys Lett 88: 595-598 Han Y-S, Choi S-H, Kim D-K (2001) Ion exchange and fixation of rare-earth cation into expandable tetrasilicic fluorine mica. J Synchrotron Rad 8:731-733

408

Mottana, Marcelli, Cibin & Dyar

Hasnain SS (1989) Synchrotron Radiation and Biophysics. Ellis Horwood, Chichester, UK, 368 p Hasnain SS (1991) X-ray Absorption Fine Structure. Ellis Horwood, Chichester, UK, 792 p Hawthorne FC (ed) (1988) Spectroscopic methods in Mineralogy and Geology. Rev Mineral 18, 698 pp Hayes TM, Boyces JB (1982) Extended X-ray absorption fine structure. Solid State Phys 37:173-351 Heald SM, Amonette JE, Turner GD, Scott AD (1995) An XAFS study of the oxidation of structural iron in biotite mica by solutions containing Br2 or H2O2. Physica B208&209:604-606 Heald SM, Stern EA (1977) Anisotropic X-ray absorption in layered compounds. Phys Rev B16:5549-5559 Hedin L, Lundqvist BI (1971) Explicit local exchange-correlation potentials. J Phys C4:2064-2083 Henderson CMB, Charnock JM, Helz GR, Kohn SC, Patrick RAD, Vaughan DJ (1991) EXAFS in Earth science research. In SS Hasnain (ed) X-ray Absorption Fine Structure. Ellis Horwood, Chichester, UK, p 573-578 Hertz G (1920) Über Absorptionslinien im Röntgenspektrum. Physik Z 21:630-632 Holmes DJ, Batchelor DR, King DA (1988) Surface structure determination by SEXAFS: The reliability of bond lengths and coordination numbers from multi-shell analyses. Surface Sci 230:476-492 Huggins FE, Huffman GP (1996) Application of XAFS spectroscopy to coal geochemistry. In MD Dyar, C McCammon, MW Schaefer (eds) Mineral Spectroscopy: A tribute to Roger G. Burns. Geochem Soc Spec Publ No. 5:11-23 Hussain Z, Umbach E, Shirley DA, Stöhr J, Feldhaus J (1982) Performance and application of a double crystal monochromator in the energy region 800