Boron: Mineralogy, Petrology, and Geochemistry 9781501509223, 9780939950416

Citation preview

BORON MINERALOGY, PETROLOGY AND GEOCHEMISTRY Second Printing with corrections and additions Edited by

E . S. Grew University of Maine

L. M. Anovits Oak Ridge National

Laboratory

Series Editor: Paul H. Ribbe Department of Geological Sciences Virginia Polytechnic Institute & State University Blacksburg, Virginia 24061 U.SA.

Mineralogical Society of America Washington,

D.C.

COPYRIGHT 1996 Second Printing with corrections and additions: 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 ( F o r m e r l y : SHORT COURSE NOTES ) ISSN 0275-0279

Volume 33 BORON: Mineralogy, Petrology and Geochemistry ISBN 0-939950-41-3 ADDITIONAL COPIES of this volume may be obtained at moderate cost from: THE M I N E R A L O G I C A L S O C I E T Y OF A M E R I C A 1 0 1 5 EIGHTEENTH STREET, N W , SUITE 6 0 1 WASHINGTON, D C 2 0 0 3 6 U . S . A .

BORON: MINERALOGY,

PETROLOGY AND GEOCHEMISTRY

FOREWORD to the Second Printing This is Volume 33 in the Reviews in Mineralogy series. It held the distinction of being the largest and most exhaustively encyclopedic of all the Reviews published before 1998 ("Planetary Materials" has since outdone it). It was edited by Ed Grew (University of Maine) and Larry Anovitz (Oak Ridge National Laboratory), who supervised 19 contributions from 31 authors. No short course was associated with this book, and this, the Second Printing has benefited from addenda (especially of reference lists) and numerous refinements of the text, additions and corrections to tables (especially in Chapters 1 and 9), and the addition of Figure 15 that was missing in the First Printing on page 619. PaulH. Ribbe, Series Editor Blacksburg, VA April 24, 2002 COVER: The cover illustrates the entry for boron in the Periodic Table. From upper left to right, it gives the atomic number, valence state (the only one known), chemical symbol, name, atomic weight and electronic configuration. The color on the cover is a close match to the diagnostic flame color for boron obtained when a mixture of boric acid and methanol is ignited (George Rossman, pers. comm., 1996). An exact match is difficult because the flame is dominated by a series of emission bands near 518, 548 and 580 nm (spectrum shown below), whereas color chips used to make the match represent broad-band absorption phenomena. Important physical and chemical properties of elemental boron, a typical non-metal, are melting point 2075°C, boiling point 4000°C. Crystalline boron is exceedingly complex, and the number of polymorphs could be "a round dozen," in addition to an amorphous form (Hoard and Hughes, 1967, p. 95; Sharpe, 1986). The density range is 2.35 to 2.52 g/cm3 (Greenwood, 1973). Boron's ionization potentials are 8.298, 25.155, 37.931, 259.38 and 340.23 eV. An ionic radius cannot be given for boron: Shannon's (1976) crystal radii for oxygen compounds are 0.15A for I^B, 0.25A for Mb. There are two naturally occurring isotopes: 10 B and ^ B (natural abundances = 19.9(2)% and 80.1(2)%, respectively. Synthetic, short-lived, radioactive isotopes ranging from masses 7 to 9, and masses 12 to 15 are known (CRS Handbook, 1994). Some nuclear properties of boron are (Kidd, 1983; CRC, 1994): Spin

10B 11B

3 3/2

400

Quadrupole Magnetogyric moment ratio (107 rad T-V 1 ) (10" 2 8 m 2 ) 0.074 2.87 8.58 0.036

500 600 Wavelength, nm

Receptivity referred to 13 C 22 754

700

Thermal Neutron cross section (barns = 10"24 cm 2 ) 0.5(3) barns 5.(3) mb

Sources of data: CRC Handbook of Chemistry and Physics, 1994 edition, CRC Press, Boca Raton, Florida; Greenwood NN (1973) Boron. In JC Bailar Jr, HJ Emeldus, R Nyholm, AF Trotman-Dickenson (eds) Comprehensive Inorganic Chemistry, p 665-991, Pergamon, Oxford; Hoard JL, Hughes, RE (1967) Elemental boron and compounds of high boron content: Structure, properties, and poly-morphism. In EL Muetterties (ed) The Chemistry of Boron and its Compounds, p 25-154 Wiley, New York, NY; Kidd RG (1983) Boron-11. In: P Laszlo (ed) NMR of Newly Accessible Nuclei, p 50-78 Academic Press, New York; Shannon RD (1976) Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallographica A32:751-767; Sharp AG (1986) Inorganic Chemistry. 2nd edition. Longmans, London.

iii

PREFACE TO SECOND PRINTING (2002) At the time of the first printing (1996), interest in the element boron was growing rapidly. We felt that it was an opportune moment to ask investigators active in research on boron to review developments in their respective fields so that readers could learn what was—and wasn't—known about boron and its minerals, geochemistry and petrology. Since 1996, interest in boron has, if anything, increased, and continued demand for the Reviews in Mineralogy "boron bible" has motivated the Mineralogical Society of America to reprint the volume. Demand is reflected in citations, and according to ISI's Science Citation Index, the number of citations since publication to the volume is about 380, with some individual chapters having been cited as many as 44 times. In preparation for this printing, authors of 15 of the 19 original chapters have updated, corrected or added to their chapters within the constraints that no pages be added. Most addenda are bibliographies of literature published since 1996; a few also include summaries of significant findings. Addenda for each chapter follow the chapter, except for those for Chapters 1 and 2, which are merged onto pages 115-116 and 385. A table of new B-minerals since 1996 is given on p. 28, and many modifications were made to the table (p. 7-27) of B-minerals known prior to 1996 (corrections to formulae, mineral names, localities, etc.). Similar up-datings of Table 1 (p. 223) in Chapter 5 and numerous tables in Chapter 9 (p. 387) were undertaken, and Figure 15 in Chapter 11 (p. 619), which—embarrassingly—was missing from the first printing, has been supplied. Addenda to Chapter 13 are introduced on p. 744 and completed on p. 863 and 864. The following salient developments in research related to B are mentioned in the addenda: New minerals. Twenty-two boron minerals have been or are about to be described, and four more have been approved by the International Mineralogical Association, representing an increase of 10%, comparable to the increase in the number of all new minerals described during the same period (Anovitz and Grew, Chapter 1) Tourmaline group. In addition to four new tourmaline species, a new classification has been proposed. Another tourmaline, olenite, has been shown to contain substantial amounts of excess B in tetrahedral coordination, a finding that has revolutionized our view of tourmaline crystal chemistry (Werding and Schreyer, Chapter 3; references in addendum to Henry and Dutrow, Chapter 10). Boron isotopes. New techniques for measuring isotope ratios using secondary ion mass spectroscopy (SIMS) with the ion microprobe open up new opportunities for in situ analyses of individual grains and fluid inclusions (Hervig, Chapter 16). Boron isotopes have found applications in paleoceanography and thus add to the tools available for the study of past climates (Palmer and Swihart, Chapter 13). We thank all the contributors who responded to our solicitation on such short notice and provided material for inclusion in the reprinted volume. We also thank James Bird (Fogler Library, University of Maine) for the Science Citation Index data. EdwardS. Grew University of Maine Orono, Maine

April 25, 2002

iv

Lawrence M. Anovitz Oak Ridge National Laboratory Oak Ridge, Tennessee

B O R O N : MINERALOGY, PETROLOGY AND GEOCHEMISTRY TABLE OF CONTENTS, VOLUME 3 3

Tables of General Interest The following is a list of tables in which information of more general interest is summarized. Some of these are cross-referenced, and some will serve as a sort of "index" to the book. Chapter 1, Table 1, pp. 8-29: Boron minerals: names, mineral groups, formulas, environments, localities. Chapter 5 Table 1, pp. 223-234: Formulae, end-member structures, and crystallographic properties of boron minerals Chapter 5 Table 2, pp. 235-251: Measured molar volumes of boron minerals Chapter 6. Table 2, pp. 2285-289: Major sources of boron in the world today Chapter 6. Table 3, pp. 290-298: Other significant occurrences of boron in the world Chapter 9. Table 5, pp. 400-401: Pressures and temperatures estimated for formation of dumortierite, grandidierite, and kornerupine Chapter 9. Table 6, pp. 402-403: Occurrences of grandidierite Chapter 9. Table 13, pp. 441-443: Occurrences of kornerupine sensu stricto (K) and prismatine (P) Chapter 9. Table 17, pp. 461-462: B2O3 contents of rock-forming silicate minerals Chapter 9. Table 18, pp. 472-473: Selected bulk analyses of borosilicate-bearing rocks Chapter 10. Table 1, p. 506: List of IMA-approved and synthetic tourmaline species Chapter 11. Table 1, p. 561: Selected tourmaline-bearing tin +/- tungsten greisens, veins, skarns and replacements Chapter 11. Table 2, p. 569: Selected tourmaline-bearing breccia pipes and porphyry deposits Chapter 11. Table 3, p. 575: Selected tourmaline-bearing vein deposits in metamorphic rocks Chapter 11. Table 4, p. 581: Selected tourmaline-bearing stratabound deposits Chapter 11. Table 5, p. 592: Selected stratiform and/or stratabound tourmalinites

v

Chapter 12. Table 6, p. 589: Inventory of B and selected elements in Earth reservoirs Chapter 14. Table 1, pp. 751-752: Boron and gadolinium concentrations in terrestrial rocks Chapter 14. Table 2, p. 758: Boron in sediments and sedimentary rocks Page

CONTENTS BY CHAPTER Chapter 1

L. M. Anovitz & E. S. Grew MINERALOGY, PETROLOGY AND GEOCHEMISTRY OF B O R O N : A N INTRODUCTION

Introduction Chemistry and Mineralogy Analytical Methods Geological Applications Thermodynamic and Experimental Data Other Topics Conclusion Acknowledgments References Chapter 2

1 1 3 4 4 5 5 6 30

F. C. Hawthorne, P. C. Burns & J. D. Grice THE CRYSTAL CHEMISTRY OF BORON

Introduction Models of Chemical Bonding in Borates

41 41

S t e r e o c h e m i s t r y of B3 and B4 P o l y h e d r a in Minerals

Variation in (B-) distances Variation in B- distances General polyhedral distortions in borate minerals Molecular-orbital Studies of Borate Polyhedra Prediction of equilibrium geometries B3 a n d B4 g r o u p s

Finite polynuclear clusters Heteropolyhedral clusters Calculated deformation-electron-density maps Polyhedral stability and reaction energies Interpretation of spectroscopic data Orbital energies X-ray spectra NMR spectra Vibrational spectra Calculation of electronic dipole polarizabilities Hierarchical Organization of Crystal Structures P o l y m e r i z a t i o n of B3 a n d B4 P o l y h e d r a

Polyhedral Clusters and Fundamental Building Blocks in Borate Minerals B-B graphs Algebraic descriptor Polyhedral linkage Enumeration of Possible Clusters vi

42

42 43 44 45 46 46

48 50 50 52 52 52 53 53 54 54 54 55

56 56 56 56 59

A Structural Hierarchy for Borate Minerals Structures Based on Isolated Polyhedra Structures Based on Finite Clusters of Polyhedra Structures Based on Infinite Chains of Polyhedra Structures Based on Infinite Sheets of Polyhedra Structures Based on Infinite Frameworks of Polyhedra Mixed Oxyanion Borates Sulphate-borates Phosphate-borates Arsenate-borate Carbonate-borates Beryllate borates Silicate-borates The Occurrence of FBBs in Borate Minerals Some General Observations on Borate Mineral Structures Acknowledgments References Chapter 3

65 65 68 77 80 88 90 93 93 94 95 96 97 106 108 109 110

G. Werding & W. Schreyer EXPERIMENTAL STUDIES ON BOROSILICATES AND SELECTED BORATES

Introduction Experimental Techniques Experiments in Boron-bearing Systems The system B2O3-SÌO2-H2O The system Na20-B 2 03-Si02-H 2 0 The system Ca0-B203-Si02-H20 The system Mg0-B 2 0 3 -Si02-H 2 0 The system Al203-B 2 03-H 2 0 The system AI2O3-B2O3-SÌO2-H2O "Boron-mullites" Dumortierite Alkali-free Al-tourmaline The system Mg0-Al 2 03-B203-H 2 0 Sinhalite Al analogue of magnesiohulsite(?) "Pseudosinhalite" The system Mg0-Al203-B 2 03-Si02-H 2 0 Grandidierite Werdingite Kornerupine and prismatine (komerupine s.l.) Magnesiodumortierite Alkali-free dravite Complex tourmalines Dravite Uvite Elbaite Highly aluminous tourmalines, olenite Tourmalines with iron and other transition elements Serendibite Axinite Boron micas Ludwigite-vonsenite vii

117 118 119 120 120 122 123 125 127 128 128 131 132 133 134 134 138 139 140 143 147 148 149 150 152 153 153 154 154 155 155 156

Boron in framework silicates Concluding Remarks Acknowledgments References

157 158 159 159

Chapter 4

A. Navrotsky

THERMOCHEMISTRY OF BOROSILICATE MELTS AND GLASSES— FROM PYREX TO PEGMATITES Introduction Binary and Pseudobinary Borate Systems Alkali and alkaline earth borates The B203-Si02 system Charge-coupled substitutions The NaBSi30g-NaAlSi308 join Multicomponent Systems with Boron as a Major Element Na20-B203-Si02 Borosilicate glass for commercial use and for radioactive waste containment Boron in Magmas Unanswered Questions, Missing Data, and Future Directions.. References

Chapter 5

165 168 168 171 172 173 174 174 175 176 177 178

L. M. Anovitz & B. S. Hemingway

THERMODYNAMICS OF BORON MINERALS: SUMMARY OF STRUCTURAL, VOLUMETRIC AND THERMOCHEMICAL DATA Introduction Formulas and Structures of Boron-bearing Phases Ammonioborite Bakerite Charlesite Diomignite Ginorite/strontioginorite Harkerite Hellandite Holtite Hulsite/magnesiohulsite Kornerupine Manandonite-2//2 Melanocerite-(Ce) Nagashimalite Peprosiite-(Ce) Rhodizite Sakhaite Sturmanite Sussexite Tadzhikite Taramellite/titantaramellite Tienshanite Tinzenite viii

181 182 182 183 183 183 183 183 183 183 183 184 184 184 184 185 185 185 185 185 185 185 186 187

Tourmaline Werdingite Wiserite Volumetric Properties Standard state volume Thermal expansion and compressibility Enthalpy and Gibbs Energy Experimentally derived values Estimating enthalpy and Gibbs energy of formation The system Ca-Mg-B-Si-O-H-Cl-S The system K-Na-B-Si-O-H-F-N Other chemical systems Heat-capacity and Entropy Experimentally derived values Estimating heat-capacity and entropy Conclusions Acknowledgments References Chapter 6

187 187 187 188 188 188 193 193 193 194 195 198 198 198 198 202 203 203

G. I. Smith & M. D. Medrano

CONTINENTAL BORATE DEPOSITS OF CENOZOIC A G E Introduction Mineralogy of Borate Deposits Diagenetic Reactions between Borate-mineral Assemblages Thermal diagenesis Reaction diagenesis Mechanics of diagenesis Chemical Activity Diagrams Types of Borate Deposits in Sedimentary Rocks of Cenozoic Age Neogene magnesium borate deposits Neogene calcium borate deposits Neogene sodium-calcium borate deposits Neogene sodium borate deposits Quaternary sodium-calcium borate deposits Quaternary sodium borate deposits Quaternary borate-brine deposits Origin of Borate Deposits Summary Acknowledgments References Chapter 7

263 264 265 266 268 269 269 272 273 273 274 274 275 276 276 277 279 280 280

D. London, G. B. Morgan VI & M. B. Wolf

BORON IN GRANITIC ROCKS AND THEIR CONTACT AUREOLES Introduction Why boron is important Getting Boron into Magmas Tectonic settings and source lithologies of boron-rich magmas Boron in sediments Boron in regional metamorphic rocks Sources of boron for granitic magmas: The importance of tourmaline Stability of tourmaline at subsolidus conditions Reaction of tourmaline at the onset of anatexis ix

299 299 300 300 301 302 303 303 304

Boron in Magmas 305 Melt fractionation and boron concentration 305 Tourmaline solubility in granitic melts 306 The buffer capacity of reactions between tourmaline, biotite and cordierite 307 Petrology of Boron in Peraluminous Silicic Magmas 310 Tourmaline in the borders of granites and pegmatites 310 Occurrence and composition 310 Significance 313 Disseminated tourmaline in granites and pegmatites 314 Occurrence and composition 314 Significance 314 314 Tourmaline in miarolitic cavities Occurrence and composition 314 Significance 317 Breccias and veins within granite or pegmatite 318 Occurrence and composition 318 Significance 318 Petrology of Boron in Peralkaline Magmas 319 Metasomatic Reactions with Wallrocks 320 Occurrence and composition 320 General characteristics 320 Around pegmatites 321 Around granitic plutons 321 Wallrock metasomatism as a record of boron in magma 322 Boron and Ore Deposits 323 Boron Partitioning and Isotope Geochemistry in Granite Systems 323 Concluding Remarks 324 Acknowledgments 324 References 325 Chapter 8

D. B. Dingwell, M. Pichavant & F. Holtz

EXPERIMENTAL STUDIES OF BORON IN GRANITIC M E L T S Introduction Structural Role of Boron in Melts Boron oxide melt Borate melts Borosilicate and boroaluminosilicate melts Hydrous granitic melts Physical Properties of Boron-bearing Melts Synthesis of boron-bearing haplogranitic melts Density B2O3 liquid Borate and borosilicate melts Volume behavior of B2O3 in granitic melts Influence of water on the partial molar volume of B2O3 Effects of pressure-compressibility Viscosity B2O3 melt Borate melts Borosilicate melts Boron-bearing granitic melts Boron diffusion Isotopic versus elemental homogenization rates x

331 333 333 333 334 337 338 339 339 339 339 340 343 344 345 345 345 346 346 349 352

Chemical Properties of Boron-bearing Melts 353 Dry boron-bearing systems 353 Hydrous boron-bearing systems 355 General representation of phase equilibria in silicate-H20-B203 .... 355 Effect of B2O3 on granite phase relations: Solidus 360 Effect of B2O3 on granite phase relations: Liquidus 363 Effect of B2O3 on reciprocal melt/H20 miscibilities: H2O solubility 366 Effect of B2O3 on reciprocal melt/H20 miscibilities: Silicate components in the magmatic fluid phase 368 Boron melt/fluid partitioning 370 B2O3 in hydrothermal fluid phases: Effect on metasomatic processes 371 B2O3 in hydrothermal fluid phases: Buffering the boron concentration of natural fluids 373 Boron incorporation in feldspars 374 Tourmaline stability in magmas: Dravite 376 Tourmaline stability in magmas: Quartz-tourmaline assemblages 376 Tourmaline stability in magmas: Schorl-dravite 377 Acknowledgments 379 References 379 Chapter 9

E. S. Grew BOROSILICATES (EXCLUSIVE OF TOURMALINE) AND BORON IN ROCK-FORMING MINERALS IN METAMORPHIC ENVIRONMENTS

Introduction Borosilicates with Stoichiometric Trigonal Boron Dumortierite group Description Chemical composition Occurrence Grandidierite Description Chemical composition Occurrence Wiserite Borosilicates with Non-stoichiometric Trigonal Boron Harkerite and sakhaite Distinction between harkerite and sakhaite Crystallography and chemical composition Occurrence Occurrence—Skarns Occurrence—Mn-rich rocks Occurrence—Deep-seated rocks Concluding statement Werdingite Description Chemical composition Occurrence Borosilicates with Stoichiometric Tetrahedral Boron Axinite group Description

xi

387 389 389 389 389 394 398 398 398 399 407 407 407 411 411 413 413 414 415 415 415 415 416 416 418 418 418

Crystal structure and chemical composition Occurrence Danburite Crystallography and chemical composition Occurrence Datolite Crystallography and chemical composition Occurrence Bakerite Taramellite group Description Crystal structure and chemical composition Occurrence Searlesite and reedmergnerite Borosilicates with Non-stoichiometric Tetrahedral Boron Kornerupine group: Kornerupine sensu stricto and prismatine Description Cell parameters Chemical composition Mineral associations Physical conditions of formation Interpretation of the assemblages and textures Serendibite Description Chemical composition Occurrence Leucosphenite Hyalotekite Manandonite Borosilicates with Non-stoichiometric Trigonal and Tetrahedral Boron Borian vesuvianite Borosilicates with Unknown Boron Coordination Oyelite Chemical composition Occurrence Charlesite, wawayandaite and vistepite Boron as a Minor Constituent in Rock-forming Minerals Tectosilicates Layer silicates Chain silicates Olivine and humite-group minerals Garnet Melilite group Aluminosilicates Aenigmatite group and sapphirine Miscellaneous minerals Origin of the Borosilicates Exclusive of Tourmaline Borosilicates in aluminous settings "Quartzites" Tourmalinites Aluminous rocks Anatexis Pegmatites Calcareous setting Settings with mafic rocks and meta-pelagites Settings enriched in Mn and Ba Conclusion xii

418 421 424 425 426 428 429 429 431 431 431 432 434 435 436 436 436 437 437 440 445 446 448 449 449 450 452 454 455 455 455 457 457 458 458 458 459 459 463 464 464 465 466 466 467 468 469 470 470 471 471 475 475 476 476 477 478

Acknowledgments Appendix References

478 479 480

Chapter 10

D. J. Henry & B. L. Dutrow METAMORPHIC TOURMALINE AND ITS PETROLOGIC APPLICATIONS

Introduction Tourmaline Crystallography and Crystal Chemistry Tourmaline species and solid solutions Minor and trace elements Potassium Vanadium and nickel Zinc Analytical procedures Electron microprobe analysis and normalization procedures Polar Properties of Tourmaline and Its Petrologic Implications Polar asymmetry in tourmaline Compositional polarity Petrologic implications of compositional polarity Examples of compositional polarity Possible causes for compositional polarity Metamorphic Tourmaline Compositions Element fractionation in tourmaline relative to coexisting minerals Metapelitic rocks and metaquartzites Diagenesis and low grade metamorphism Medium grade metamorphism High grade metamorphism High pressure metamorphism Calcareous metasediments Metamorphosed tourmalinite metasediments Metamorphosed granites and pegmatites Metamorphosed mafic rocks Blueschist facies Eclogite facies Meta-ultramafic and Cr-rich rocks Metamorphosed ironstones Meta-evaporite rocks Hydrothermally altered rocks Tourmaline as a Monitor and Reservoir of Boron Boron contents in minerals and rocks B in typical pelitic and psammitic sedimentary and metasedimentary minerals and rocks B in calcareous sediments, metasediments and skarns B in mafic and ultramafic igneous and metamorphic minerals and rocks P-T-X stability of tourmaline Low temperature and pressure conditions High temperature stability High pressure stability Stability with respect to fluid compositions Closed-system behavior and development of tourmaline

xiii

503 503 507 508 511 511 511 511 512 514 514 516 516 517 521 521 522 523 524 525 526 526 526 527 527 528 528 528 528 529 529 530 530 530 531 532 532 534 534 534 535 536 536

Open system behavior Boron mobility in hydrothermal and metamorphic Tourmaline growth and external fluids Detrital Tourmaline and Provenance Geothermometry Involving Tourmaline Intermineral element partitioning Intramineral polar partitioning Isotope thermometry Hydrogen isotopes Oxygen isotopes Tourmaline as a Geochemical Probe Boron isotopes Oxygen and hydrogen isotopes Rb-Sr isotopes Silicon isotopes Trace elements Tourmaline major element chemistry as an exploration guide Geochronology Involving Tourmaline K-Ar and ^Ar/S^Ar systematics Rb-Sr geochronology Sm-Nd geochronology Fission track thermochronology Tourmaline as a Kinematic Indicator Acknowledgments References Appendix Chapter 11

fluids

537 537 538 539 540 540 542 543 543 543 543 543 544 544 545 545 545 545 545 546 546 547 547 547 547 557

J. F. Slack TOURMALINE ASSOCIATIONS WITH HYDROTHERMAL ORE DEPOSITS

Introduction Granitoid-related Deposits Tin ± tungsten greisens, veins, skarns, and replacements Greisen and vein deposits Skarn deposits Replacement deposits Copper ± gold breccia pipes and copper + molybdenum porphyries Gold-bearing veins Copper-bearing veins and replacements Uranium-bearing veins Lead-zinc veins and replacements Veins in Volcanic Rocks Uranium-molybdenum-zinc deposits Silver-gold-zinc deposits Modern geothermal analogues Veins in Metamorphic Rocks Gold deposits Copper-gold deposits Lead-zinc deposits Mercury deposits Cobalt-copper deposits Stratabound Deposits Boron-bearing seafloor hydrothermal systems xiv

559 560 560 560 566 566 568 571 571 572 572 572 572 573 573 574 574 577 578 579 579 580 580

Lead-zinc deposits Copper-zinc deposits Copper-zinc-cobalt-nickel deposits Copper-cobalt deposits Gold deposits Tungsten deposits Uranium deposits Tourmalinites Field relations Mineralogy and petrography Whole-rock chemistry Origin Historical review Premetamorphic replacement Syngenetic-exhalative processes Submarine-hydrothermal leaching Colloids and gels Evaporitic processes Contact metasomatism Regional metasomatism Metallogeny Tourmaline Chemistry Granitoid-related deposits Veins in volcanic rocks Veins in metamorphic rocks Stratabound massive sulfides Miscellaneous stratabound deposits Exploration applications Isotopic Compositions Oxygen and hydrogen Boron Silicon Strontium and neodymium Summary Future Research Acknowledgments References Chapter 12

581 585 586 586 587 589 590 590 591 593 593 598 598 598 600 601 602 602 603 603 604 605 605 608 609 610 613 615 616 616 618 620 620 621 621 622 623 W. P. Leeman & V. B. Sisson

GEOCHEMISTRY OF BORON AND ITS IMPLICATIONS FOR CRUSTAL AND MANTLE PROCESSES Introduction Early studies—An historical perspective Recent developments—A broadening of scope General Inventory of Boron in the Major Reservoirs Boron in fluids and the hydrosphere The oceans Ocean-floor hydrothermal fluids Subaerial geothermal fluids Fluid inclusions as relicts of deep crustal or magmatic fluids Formation waters Atmospheric boron Anthropogenic contributions XV

645 647 650 650 651 651 653 653 655 656 656 656

Boron in crustal rocks and average continental crust Sediments and sedimentary rocks Diagenetic effects Weathering Evaporative borate deposits Tourmalinites Summary of inventories Temporal evolution of the oceans and continental crust Metamorphic rocks Hydrothermal metamorphism Contact metamorphism Regional metamorphism Granulites Migmatites Impact metamorphism Distribution of B in metamorphic and igneous minerals Igneous rocks Ocean floor basalts and oceanic crust Granitic rocks Silicic volcanic rocks Petrogenetic implications Estimated inventories in major reservoirs Implications of Boron Geochemistry for Subduction Processes Boron systematics in arc lavas Petrologie implications Reflections on Global Geochemical Fluxes Cosmochemistry of Boron Meteorites Lunar samples Solar abundances and condensation temperatures Acknowledgments Appendix. Brief Overview of Modern Analytical Methods Macro-analytical methods Prompt-gamma neutron activation Plasma emission spectroscopy Isotope dilution mass spectrometry Micro-analytical methods Alpha-track mapping Secondary ionization mass spectrometry (ion microprobe) Electron microprobe Geochemical reference samples for boron analysis References

657 657 659 659 660 660 660 661 662 662 662 667 670 672 673 673 675 675 677 680 681 682 682 682 687 688 690 690 690 691 692 692 692 692 693 693 693 693 693 694 695 695

M. R. Palmer & G. H. Swihart

Chapter 13

BORON ISOTOPE GEOCHEMISTRY: AN OVERVIEW Introduction Isotope Fractionation Theoretical Tourmaline-aqueous vapor Geothermal processes Carbonate-water Borate-water Mantle and Cosmochemistry Cosmochemistry

709 710 710 711 713 714 715 717 717 xvi

Tektites Mantle geochemistry Interaction of Seawater with Oceanic Crust Marine Sediments Sediment geochemistry Ancient sediments and paleoceanography Sediment-hosted hydrothermal systems Island Arcs and Subduction Zones Continental Geothermal Systems Evaporites Tourmaline General Broken Hill (Australia) Barberton (South Africa) Evolution of the continental crust and seawater Metamorphic Environments Future Developments Acknowledgments References Chapter 14

717 717 719 721 721 723 723 725 727 728 733 733 735 736 738 738 739 740 740 D. M. Shaw

SIMILARITIES AND CONTRASTS IN LUNAR AND TERRESTRIAL BORON GEOCHEMISTRY Introduction Boron in Lunar Rocks Bulk concentrations Boron and lithium sites in lunar rocks Boron and Water in Terrestrial Rocks Incompatibility and complications Igneous rocks Basaltic rocks Mantle rocks Differentiated igneous rocks Magmatic volatiles Sedimentary environments Metamorphism Mass transfer Tourmaline breakdown Boron within terrestrial rocks Summary Addendum Acknowledgments References

:

745 745 745 747 750 750 754 754 755 755 756 758 759 759 761 762 763 764 764 764

J. J. McGee & L. M. Anovitz

Chapter 15

ELECTRON PROBE MICROANALYSIS OF GEOLOGIC MATERIALS FOR BORON Introduction Instrumental and Analytical Developments Energy-dispersive spectrometry (EDS) xvii

771 771 772

Wavelength-dispersive spectrometry (WDS) Layered synthetic microstructures (LSM) Analytical conditions Accelerating voltage Beam current Counting times Spectral properties Standards Matrix correction procedures Sample preparation EPMA Analysis of Minerals for Boron Kornerupine Vesuvianite Tourmaline Summary and Future Considerations Acknowledgments References

772 772 775 775 776 776 776 778 779 781 781 782 782 782 785 786 786

Chapter 16

R. L. Hervig

ANALYSES OF GEOLOGICAL MATERIALS FOR BORON BY SECONDARY ION MASS SPECTROMETRY Introduction Analytical Techniques Instrumentation General description of analysis conditions Sample charging Secondary ion characteristics Quantification Working curves Matrix effects Ion implantation Limits on B analysis by SIMS Detection level Contamination Applications Microanalyses Diffusion measurements Isotope ratio measurements Ion imaging Future Work Acknowledgments References Chapter 17

789 789 789 789 790 791 792 792 792 795 796 796 796 796 796 798 798 798 799 800 800 J. D. Robertson & M. D. Dyar

NUCLEAR METHODS FOR ANALYSIS OF BORON IN MINERALS Introduction Nuclear Methods compared to EMPA and SIMS Particle-induced Nuclear Reaction Analysis Prompt-gamma Neutron Activation Analysis xviii

805 805 806 812

Fast Neutron Activation Analysis Summary Acknowledgments References

815 818 818 819

Chapter 18

L. A. J. Garvie & P. R. Buseck

PARALLEL ELECTRON ENERGY-LOSS SPECTROSCOPY OF BORON IN MINERALS Introduction Experimental EELS Relation to Other Analytical Methods Interactions of Electrons with Matter The EELS spectrum EELS Spectra of Minerals containing Boron Quantification The coordination fingerprint Determination of site occupancy Beam damage High-spatial-resolution EELS Interpretation of the B K ELNES Molecular Orbital (MO) approach Multiple scattering Band structure Additional bonding information Site symmetry and structure Bond character Summary Acknowledgments References Chapter 19

821 821 823 824 824 826 826 828 830 832 833 833 833 834 836 836 837 838 838 839 839 G. H. Swihart

INSTRUMENTAL TECHNIQUES FOR BORON ISOTOPE ANALYSIS Introduction Mass Spectrometry of Boron Magnetic sector mass spectrometry Thermal ionization—positive ions Thermal ionization—negative ions Comparison of the thermal ionization mass spectrometric methods Electron impact ionization Primary beam sputtering Glow discharge Quadrupole mass spectrometry Chemical ionization Electron impact ionization Inductively coupled plasma ionization Other instrumental techniques Sample Preparation Methods xix

845 845 845 845 849 851 852 853 853 853 853 854 854 854 855

Sample decomposition and dissolution Fluxed fusion Acid dissolution Pyrohydrolysis Boron separation methods Methyl borate distillation Ion exchange Organic removal General precautions Isotope Standards for Boron General Summary Acknowledgments References

xx

856 856 856 856 857 857 857 858 858 859 859 859 860

Chapter 1 MINERALOGY, PETROLOGY AND GEOCHEMISTRY OF BORON: AN INTRODUCTION Lawrence M. Anovitz Chemical and Analytical Sciences Division P.O. Box 2008, M.S. 6110, Bldg. 4500-S Oak Ridge National Laboratory Oak Ridge, Tennessee 37831 U.S.A.

Edward S. Grew Department of Geological Sciences The University of Maine 5711 Boardman Hall Orono, Maine 04469 U.S.A.

INTRODUCTION Boron is a widespread constituent of crustal rocks, being found in sedimentary, volcanic, plutonic and metamorphic environments. The Earth's upper continental crust (average of 15 ppm B, 27th in abundance) is highly enriched in boron relative to the primitive mantle, which is inferred to have had 0.6 ppm B prior to creation of the crust (Taylor and McLennan, 1985; 1995; Leeman and Sisson, Chapter 12, cite 16.8 ppm, 0.10 ppm B, respectively). Thus, boron can be considered as a quintessential crustal element. Boron is also a major industrial material that has wide use in ceramics, glasses, metallurgy, insulation and textile grade fibers and lightweight fiber-strengthened materials, medicine, soaps, pyrotechnics, rocketry, and fuel technology. A few boron minerals have applications in their own right, such as tourmaline, kornerupine, sinhalite, and danburite as gemstones. Boric acid is used as a mild antiseptic and insecticide for cockroaches, and borax as a cleaning flux in welding and as a water softener. Because of the numerous analytical difficulties of working with boron, researchers in the earth sciences have given it less attention than other minor elements in the Earth's crust. In the last decade, however, earth scientists in a broad spectrum of specializations have taken up boron as a research topic, and considerable mineralogical, thermochemical and geochemical data are being generated. Eventually these should provide both an understanding of the chemistry of boron in geological environments and tools by which boron can be used to further our understanding of geologic processes. At the present time, these data are widely dispersed in the literature. Neither the current state of research, nor future directions are obvious to investigators not directly involved in boron research. We therefore decided that the moment was opportune to invite investigators active in research on boron to contribute chapters to a Reviews in Mineralogy volume. CHEMISTRY AND MINERALOGY Borax (tincal) was known in the ancient world (Persian burah, Arabic baurach, Sanskrit tincana), where it was used in the working of gold and preparation of hard glasses and some glazes. Whether tourmaline was used as a decorative stone by the time 0275-0279/96/0033-0001 $05.00

2

Anovitz & Grew: Introduction

of the Roman Empire is conjecture, but its use for gems and carvings is known from the eighteenth century onwards (Kunz, 1913; Dietrich, 1985). The earliest modern chemical investigations of boron involved preparation of boric acid from borax. In the eighteenth century natural borax, know as Tincar or Tincal was imported from the East Indies by the Dutch who purified if for metal fusion into a compound known as borax (Weeks and Leicester, 1968). Boric acid, called sal sedativum (sedative salt) was first prepared from borax in 1702 by Wilhelm Homberg by reaction with iron vitriol (ferrous sulfate). Natural boric acid (sassolite) was first described from a boiling hot spring in Tuscany, Italy in 1778 and boracite was the first more complex mineral in which the presence of boric acid as a constituent was recognized in 1788 (Weeks and Leicester, 1968). Elemental boron was first isolated by Humphry Davy (reported June 30), and independently by J. L Gay-Lussac and L. J. Thenard (reported June 21) in 1808 (GayLussac and Thenard, 1808; Davy, 1808; Weeks and Leicester, 1968; Greenwood and Earnshaw, 1984). In both experiments, metallic potassium was used to reduce boric acid. Davy originally named this new material boracium, while Gay-Lussac and Thenard called it bore, both etymologically derived from borax. In 1812, Davy realized, as did Gay-Lussac and Thenard before him, that boron was more akin to carbon than a metal, and so proposed the name now used in English from borax + carbon. Gay-Lussac and Thenard's name bore has found its way into other European languages as bor. Some 224 naturally occurring boron compounds have been or are being described. Of these, 217 are recognized as valid mineral species by the International Mineralogical Association (IMA), and 7 have yet to be approved. These phases, together with their formulas, mineral groups, environments of formation and some important localities are listed in the Table at the end of this chapter. With the exception of three fluorides (avogadrite, barberiite, ferruccite), the boron minerals are oxygen compounds in which boron is linked only to oxygen or hydroxyl either trigonally or tetrahedrally. Electron loss and cation formation plays no role in boron crystal chemistry because its first oxidation potential is relatively high (8.296 eV). Rather, covalent bond formation is of major importance. In this boron compounds are similar to other non-metals such as silicon (Cotton and Wilkinson, 1980). The crystal chemistry of these compounds is reviewed by Hawthorne et al. (Chapter 2). The oxygen-bearing boron minerals can be grouped into two broad categories: borates and borosilicates. In borates, boron polyhedra are isolated or share vertices with one another, or, in rare cases with Be tetrahedra (berborite, hambergite, rhodizite). In borosilicates boron polyhedra share vertices with B (MB-O- WB bridges only) or Si tetrahedra, and, rarely with Al (serendibite) or Be (hyalotekite) tetrahedra. Other types of polyhedra, including: CO3, SO4, PO4, and ASO4 are also present in boron minerals. CO3 and SO4 polyhedra are not linked to B polyhedra in minerals, but BO-C bridges are reported in organoborates (e.g. Ross and Edwards, 1967; Greenwood, 1973) and mB-O-WS and WB-O-WS bridges have been inferred in some synthetic sulfoborates (Gillespie and Robinson, 1962). Isomorphic substitutions for B are not common; substitution of t3lB invariably involves a change in coordination, whereas substitution of t 4 lB by another tetrahedral cation, generally Si, is more the rule (Grew, Chapter 9). Boron minerals form in a wide variety of geologic environments (see the Table at the end of this chapter). These range from sublimates formed in volcanic fumeroles to

Anovitz & Grew: Introduction

3

soluble "salts" deposited in the driest climates and boric acid lagoons to highly refractory materials formed under granulite-facies conditions. The mineralogies of these environments are strikingly different. Borates with variable and substantial water and hydroxyl contents characterize saline deposits (Smith and Medrano, Chapter 6), while borosilicates and borates in plutonic systems (London et al., Chapter 7; Dingwell et al., Chapter 8) and metamorphic rocks (Grew, Chapter 9) are either anhydrous or contain minor hydroxyl. Tourmalines are by far the most widespread boron minerals, occurring in a number of geologic environments. Two chapters is this volume are devoted to this mineral group: Henry and Dutrow (Chapter 10) on tourmaline in metamorphic rocks, and Slack (Chapter 11) on its association with hydrothermal ore deposits. Several others cover specific aspects of the data on this group (Anovitz and Hemingway, Chapter 5; Dingwell et al., Chapter 8, London et al., Chapter 7). The complexity of tourmaline chemistry is due to the wide variety of solid solutions tolerated on four distinct cation sites. To date, 15 distinct naturally occurring tourmaline species have been recognized and named, but even these do not suffice to describe the known compositional range (Burt, 1989; Hawthorne et al., Chapter 2; Henry and Dutrow, 1990, Chapter 10). It is quite likely that additional tourmaline species will be recognized and named in the future. Boron for industrial use is largely obtained and most readily extracted from borates in (1) continental saline deposits such as Furnace Creek and Boron (Kramer), California and Turkey; (2) brine deposits, notably Searles Lake, California; (3) salar deposits in the Andes of Argentina, Chile and Peru; and (4) marine evaporites, the largest of which is the Inder deposit, Kazakhstan (Smith and Medrano, Chapter 6). Another source of boron is skarns and boron-enriched metamorphic rocks, which are worked, respectively, for datolite at Dal'negorsk, Russian Far East, and Mg-Fe borates in the Liaoning Province, China. In compiling the Table, whenever possible, we have consulted the original papers on crystal structure refinements and parageneses, as well as summary volumes such as Palache et al. (1944, 1951), Anthony et al. (1995), Hey's Mineral Index (Third Edition, Clark, 1993), Glossary of Mineral Species (Fleischer and Mandarino, 1995), Mineral Reference Manual (Nickel and Nichols, 1991), the Encyclopedia of Minerals (Roberts et al., 1990) and Malinko et al. (1991). We have also included references to abstracts in the American Mineralogist in which data on new minerals are summarized or updated. In cases where the formulas given for a phase differed from one source to another, the formula derived from the structural analysis was used if available. Several phases for which the formula remains uncertain are discussed in the chapters by Anovitz and Hemingway (Chapter 5), and Grew (Chapter 9). The Table is intended to serve as a guide to the literature on boron minerals, particularly those minerals not covered elsewhere in this volume. ANALYTICAL METHODS Although boron was recognized as a constituent of minerals as early as 1770 (Hey, 1973), reliable analyses did not become possible until some 100 years later (e.g. Whitfield, 1887), and borosilicates remained a challenge well into the 20th century (e.g. Hey et al., 1941). Consequently, boron was overlooked in the early descriptions of several borosilicates. Measurement of boron as a trace element was also deterred by analytical difficulties; Goldschmidt and Peters (1932) were among the first to apply emission spectroscopy to analyze trace B in rocks and minerals. These analytical difficulties persist to the present. Boron is not detected at all with X-ray fluorescence

4

Anovitz & Grew: Introduction

spectrometers and only using special spectrometer crystals with electron microprobes (McGee and Anovitz, Chapter 15). Consequently, when these instruments displaced wet chemistry in geochemical, mineralogical, and petrological research, boron was virtually ignored. Newer technologies have begun to reverse the trend to omit or avoid analysis of boron. Analysis of boron has now become feasible for many researchers, if not exactly routine. For minerals and rocks, advances in electron microprobe instrumentation have put boron within reach, and other technologies, ranging from ion microprobe to particleinduced gamma-ray emission, have opened many new avenues, especially for microanalysis (McGee and Anovitz, Hervig, Robertson and Dyar, Garvie and Buseck, Chapters 15-18). Despite these advances, instrumental boron analysis is associated with more uncertainty than analysis of routinely measured elements, and has not entirely displaced wet chemistry. In our opinion, there is no reason not to include B or B2O3 in tabulations of rock and mineral analyses accompanying geochemical, mineralogical, and petrological studies. The potential importance of systematic analytical data on boron cannot be overemphasized. Isotopic analysis of boron has also been revolutionized in recent years. Highly precise measurement of the two naturally occurring l 0 B and U B isotopes is due to both improved chemical preparation methods and vastly improved methodology and resolution in mass spectroscopy (Hervig, Chapter 16; Swihart, Chapter 19). GEOLOGICAL APPLICATIONS Expanded interest and new analytical capabilities have led to numerous systematic studies on the behavior of boron in a variety of geologic systems. This includes studies of boron during metamorphism (Grew, Chapter 9, Henry and Dutrow, Chapter 10), in subducted slabs (Leeman and Sisson, Chapter 12), in sedimentary environments (Smith and Medrano, Chapter 6), and in granites (Dingwell et al., Chapter 8 and London et al., Chapter 7), as well as the application of boron isotopes as a sensitive geochemical indicator in a number of environments. (Palmer and Swihart, Chapter 13). Comparison of terrestrial and lunar boron geochemistry (Shaw, Chapter 14) and meteorites (Leeman and Sisson, Chapter 12) gives us a perspective from outside the earth, as the ultimate origin of boron is an active topic of research in astrophysics (e.g. Casse et al., 1995; Cameron, 1995). THERMODYNAMIC AND EXPERIMENTAL DATA Despite the industrial importance of borosilicate glasses and ceramics, experimental studies of boron-bearing systems relevant to nature have been surprisingly few in the earth science community. Nonetheless, subsystems of N a 2 0 - C a 0 - M g 0 - A l 2 0 3 B 2 0 3 - S i 0 2 - H 2 0 have been explored. This experimental work is reviewed by Werding and Schreyer (Chapter 3). More complex systems relevant to granite are discussed by Dingwell et al. (Chapter 8) and London et al. (Chapter 7). A number of experimental studies are also reviewed or listed by Anovitz and Hemingway (Chapter 5). Boron-bearing glasses are among the most important industrial uses of boron. Beyond the common and laboratory uses of these materials, they have recently become even more significant as potential materials for the long term storage of nuclear wastes. Navrotsky (Chapter 4) reviews the thermochemical and structural aspects of several of these systems.

Anovitz & Grew: Introduction

5

There have been few attempts to directly measure the thermodynamic properties of boron minerals. For instance, the compilation of Robie et al. (1978) lists only borax. None the less, significant data do exist scattered throughout the literature. These include molar volume data for nearly all known boron phases, experimental data including that summarized by Werding and Schreyer (Chapter 3) and a scattering of heat capacity, entropy, enthalpy and Gibbs energy data. Anovitz and Hemingway (Chapter 5) have summarized these scattered data, reviewing earlier work and suggesting directions for future analysis and experiment. They also present newly obtained data on a number of boron minerals, and discuss the possibilities of estimating data for boron minerals for which there are currently no measured data available. OTHER TOPICS No volume such as this can lay claim to completeness. There are bound to be omissions. Space is limited, and individual authors emphasize their own particular perspective, experience and expertise. We particularly regret that the following topics were either left out altogether or treated only in passing; each deserved a full-scale review because a sizable literature exists: (1) parageneses of high-temperature metamorphic borates; (2) molecular orbital theory of boron, (3) the economic geology of boron other than continental borate deposits, (4) application of infra-red (IR), visible-light and nuclear magnetic resonance (NMR) spectroscopy to boron minerals (e.g. IR work was most recently reviewed by Ross, 1974, and NMR, by Fyfe, 1983; Kidd, 1983; Kirkpatrick, 1988) and (5) aqueous geochemistry and hydrogeology of boron. CONCLUSION This volume is a progress report. Our objectives have been to assess the current state of research on boron in the earth science community, and to identify directions for future research. A common theme in many of the papers presented is work on boron is far from complete and many basic questions remain unanswered despite the progress to date. Basic questions which remain unanswered include: (1) The 7 unapproved minerals listed in the Table are only the tip the iceberg on the potential for new boron minerals. Werding and Schreyer (Chapter 3) report the synthesis of several compounds that could occur as minerals, including a family of sillimanite- and mullite-like Al-B-Si phases that have also appeared in some experiments on B-rich granitic melts (Dingwell et al., Chapter 8). One such phase, boralsilite, has recently been discovered in pegmatites (Grew et al., 1998; Peacor et al. 1999; Grew, Chapter 8; illustrated by McGee and Anovitz, Chapter 15, Fig. 5). Other new boron minerals have recently been approved or are under consideration by the IMA Commission on New Minerals and New Mineral Names. The potential contributions of these new minerals to our understanding of boron geochemistry and petrology are exciting lines of future research. Also, some B minerals need more study, e.g. the veatchites. (2) The thermodynamic properties and stability relations of boron minerals are virtually terra incognita. Systematic experimental studies on the stability ranges of boron minerals have only begun, as have accumulation of calorimetric and other thermodynamic data. Additional measurements and experimental studies, as well as careful evaluation of those already available, are badly needed if the phase equilibria of the various boron systems are to be understood. (3) The development of the vast, economically important sedimentary accumulations of boron remains incompletely understood (Smith and Medrano, Chapter 6). In addition, virtually all of the important continental borate deposits are all less

6

Anovitz & Grew: Introduction than 20 Ma old. Are such boron accumulations geologically recent phenomena, or do other factors play a role in their preservation? One scenario for the Early Proterozoic deposit in Liaoning Province, China is that the deposit formed by metamorphism of continental borates. (4) The boron budget during prograde regional metamorphism is the subject of considerable debate. In some cases, boron concentrations clearly decrease with increasing metamorphic grade, but in others, they do not. The extraordinary boron enrichments in some granulite-facies metamorphic rocks raise questions as to the source of the boron and its retention under high-grade conditions. The fate of boron expelled from metasediments during metamorphism is yet another problem awaiting further research. In this volume Grew (Chapter 9), Slack (Chapter 11), Leeman and Sisson (Chapter 12) and Shaw (Chapter 14) present several perspectives on boron behavior in metamorphic systems.

With the advances in analytical technology now available to researchers, questions such as the above can be more easily addressed, and may yield important insights into geologic processes. Just as Deer, Howie and Zussman's five-volume set Rock Forming Minerals (Deer et al., 1962) and previous volumes in the Reviews in Mineralogy series have stimulated research in other aspects of mineralogy, our volume is intended to be a stimulus to research on boron in the earth science community. ACKNOWLEDGMENTS We thank a number of individuals and organizations whose help was invaluable in our work to create both this paper and this volume. First, we would like to thank the authors of the individual chapters who did such a wonderful job of writing and exceeded our expectations in the quality of their work. The efforts of the series editor, Paul Ribbe were critical to the completion of this project. We also thank the following individuals for their thoughtful and prompt reviews of individual chapters: Timothy J. Barrett, Peter C. Burns, Donald M. Burt, Marc Chaussidon, Joseph V. Chernosky, Jr., Alan H. Clark, Francis D. Correll, Don B. Dingwell, Robert F. Dymek, Eric J. Essene, Karen Geisinger, Michael D. Glascock, Lee A. Groat, Charles V. Guidotti, Bruce S. Hemingway, N. Gary Hemming, C. Michael B. Henderson, Paul C. Hess, David A. Hewitt, Richard W. Hinton, François Holtz, John M. Hughes, Ian D. Hutcheon, William P. Leeman, Valérie Michaud, George B. Morgan VI, Peter I. Nabelek, Nikolai N. Pertsev, Michel Pichavant, Mati Raudsepp, Philip E. Rosenberg, Werner Schreyer, Denis M. Shaw, Charles K. Shearer, Virginia B. Sisson, John F. Slack, Michael N. Spilde, Garrison Spositio, George H. Swihart, John A. Tossell, John W. Valley, David J. Waters. Assistance from John Jambor, Christian Chopin and Frank Mazdab in the creation and updating of the mineral table and in providing preliminary data on new boron-bearing minerals is greatly appreciated. Comments on the text of this manuscript by Priscilla C. Grew greatly improved it. We also thank George Rossman, who generously took the time to run a flame test and measure emission spectra of boric acid, and provided valuable information on the boron flame color used on the cover of this volume (see text on inside cover). L.M. Anovitz is grateful to the Geochemistry Group, Chemical and Analytical Sciences Division, Oak Ridge National Laboratory, and the Department of Geological Sciences, University of Tennessee, without whose support his work on this project would not have been possible. E.S. Grew acknowledges the financial support of U.S. National Science Foundation grants EAR-9118408 and EAR-95-26403 to the University of Maine.

A n o v i t z

&

»

;

j : e K : > U CD 1. a 3a • ^ u

a ^ Kii

>

• o . S | S

>

>

n

: s - i 8-

co j i O u

•X i " 7> > 1

i

-•a

si

ID o

XM

o -J

7

Introduction

G r e w :

60 e DO q} CO •5 "

J

™ Co r il 3 » " H i J •S « I 3 co ^ » t ) A - 1 - (U 3 J « H

3 a

3 b

CD « O cj

3 O ^ S 3

>• c

B

o o Q. O o. & c

O

60 la p - S 3 w

M

w

03 M

S O,1" Q J

S >•

8 S < 2

m

6 0 .a5 °M

.

I - a - S

" t e a " « c Jo• 3 HS • '> 5 _ o 5 £ ¿ 2 -r) - H O .

o

< in

S « o O

• 3 .is 3 S o O.

; . „ £ M 55 ¡ U S a, « ^ i f I Q . co ¿ r ^ Y S ' s Cfl T 3 co e o ^ o CQ

o *s

03

00 3 CQ

'c CO

U

o u

s

s

S

Introduction

Anovitz & Grew:

.2 i s *

^ ^ S

o o d

ON ^

40

—

X? -2 o 21 C, -5 U M§ .ts « 2 » o T2. cn < a;>

T3 I W Bj

Q O

ca co 2 S

v

~

r rS ^ 8 c £

^o o &

" P 5 U & .. C O

Ill

i

o ^ Z .

o ^

1°.

a 3


I

„

S w

I

b A3

s -S flj «

— cd „ y S O ^ «

§ js ° -5 ' 2\

y a CO

S C.T3 ed fl C

g ) § .S -o a

.2 p

? . s | fa . 2 m c

S ¡2 ti? w !T] a

« " w - o o co e xi co

l

l i l bo O , CO

3

£

to

O

I

g

-^¡i

I J= | ! S h K i

c3

D

•s e S

l

2P „

t^- T3

g s >

"

I

«

j

^

S3 U : «SC. (0 SA! e

3

§ s 1

l

O

S

I

M

5

3

l

g i

.5

^ S i CJ CO O

O K O

u

o

ca

oCM X

ffiO O K

K O

o »

X

o

EC O O

tt. oi

oa

s

t cd O

o O

a

e

«

3 < m

o

° m

o

>

C

£ S §•£

S K t :

ao • c « e

-o S3

C

s

" S I „-

i l l E - T3

2

TO Ov

OJ CO

n S • - if"O 5< § 2 ,fU UL » « £ "8 ¡2 2 a

> «B gG feb a.

JJ M S" *4> ON 00 ONON ~ |N vU ^eo> OON — S g . S i .3 ~ I I o /— ...\' — bo C3 /N NO £N£ON -s P U c -5 CU w wffi

s R;» ¿a

Pryce (1971); Hoskins et al. Voloshin et a

3 £ca .2o [I.

«ca s0\a0\a ¡6 1 | a" oo s al iiD. iZUS S o i z o

U w

CO o < E 2 CQ & ^ eoc7S 4> 8 ^ Li 6Z B W a 5 C 5 .« Qco"O•£ bflt 32 0) :3 r . o S i OS TJ 2M ca" • O ^ ^ c 3 J; n C Q to r^ ^ ^ t a, ^ S 5 S < ¡3 s 5" oo U •SHcaSumXtsi

£ .2

03

i

Ss s

•O

B.-s «"J "S S E c ao ex

a o

g ™ J ™ " Ji ^

ca

g< Ju o ^

g c f2 13 _cf T3 •J 61 si

Introduction

U

o X

O(N X

/-C X o o ®o

X o

n Ph (N 5OO C

o

X

X

o X •r,

X o rl oro C Q M s

ocs ffi 00 o

^ X U

>o o CQ ca CJ

U| ^bo iN S ' zrn o

Anovitz &

Introduction

Grew:

2 1 - o\ r - s^ ; > — «0 m w •3« « 2

O .S

W (d TO M CU 2

fl

> ~ ]H . . CL,

3I o • ¡ bg

«_

.„

>

.s .s u £ >•> . - o >.c 2 5 .2 3

'n j"C j

¡^•rf 0 « ts

U :

e D

J D,—

E . «

3 c J H a-S

N .0 e.S O e J ! « •3 I < . ¡ S i t 2 "w c/0

_: - • S

C >O

O (3

Introduction

^ 6 - ' O p .X! 0\ o

O O >

4)

¡2

•

rrv

o\

ON ~

bo

K ' J J> T3 CgL- 13 ^

' zl•

£ M

«

Os ; r-1. ^u ^* S ; O. CO > bo bo O > >-. < N < ° J > (2 Io I

• S j= -S •o S o ^ £g > £

< e « a< S T3 « illS

^

m Ch

o (3s S co 2 ^ w *C0 m^ 4) S 73

6 s j o ^ u?

' "3 ¿a » u c

, a S3 g £ u c -g c Q O M
o >

% o

•S « £ « co 'C -"S « « ^ S "C ft, U M O

-

«

S-s -c 2 | 2 < g I 8 s"§ o _ .5 e S g -H

s e : 1 & as cd

s: ! ^

It" « u S o -

60

? §

M

« 'g C S

= o 2 w -

1)

g s j 1 - i

e

— 3 a B

i^m

o

CU «

S ^ .2 s

-ti

b p a g

E o j2

.s

P E "' 8 6 o ifl "O ^ a? 3 2 ^

O

a

ca

W CO • s ^ - g « 60 S •w ¡ . g - s ^ i- c « rW •— 13 «5 4) t - K) « . 5 Si o «$5

-S

-o

S ~ J >>

g w C ""

£ —

S S. S 5 P M cu S

C-

A n o v i t z

&

« g CO

—

e

DOT

.•2o a3 O -

O U C O

~s CO N « ^

5 Jr

W CO

B Z O i M

Ul aj c 0» of ° fa o

l

o3 >> o

o ;

io

i £ S

M r?

0\

S

Is

-

19

Introduction

G r e w :

„ c O

CO g !D c -o «

2 - 8

2

> B 3 ^ z " T i g U

J r g

5

I

3 fc o

t o od

N C fflS ¡ ¿ 2 a a ."S O -S £ 'C 2 w CL, C 1 -o H

•Si.

§

R l t ' c ^ ffl >

•C TCIS 3 sO - g | « s > « S 8 cs > •£3 b. in o 3 O 3 O U U Q H o - c o

CL,

u s M a s g S S c c o

B 1

ex

i - a

o.

I

S

I

H< , O

10

o

X

(N •O ST o o (B1 (0 O

O

m


r o

O 00

a o

m 00 S

o «

Anovitz & Grew: Introduction

24

o\

s •s

°

W CO

~ B

•3 .2

^. o W

»®a| gc 73 ® cd — I •—-2 S | ] cd o o §•1 « £ » o s.s a g '•SE

•S^Z'-g a | S3 c I c-rs?J .2 „ e O SfS 2 u S > .S g 5 c » 2 • £ oi S re cS < ;

Cd Cd

» g-a« •a — I.&« • u-

c/D cd t> > •o O ¿ 2 | 2 cd3 o. a *8 -3 oS rO J * § g= e o U u -o b 3 "o a •S s S J 3 ¡ 1 ^ BM2 c .SO a J) o c-

J3 C/3 03> H

Iv> P

cfl w 13 C 4J _ S au j j (0i < U N |J3 £ ¡2 J M gHfOT3C/3 XI B-5

E

g B•

3 rt - »N . c i2

U< ^3 S « < .3w JS TX

5u ™ a> me fe 3cdo 2 c ° S i x c COo. O , cd JB >P B to• O M

S E E.S

e£

2 3

s'^ c Srt ^ , .2 -S P rC a. .S m S 9o 2 £§£ ¿•a i> s -s a .is

Z B,

W) c S

CN

g

ca ™ u = ?

73 i -C • JS « OJ

o.

j

co •

S

b 3 § 2 « a - U '3

-2 3

2

£

MU

•a « o fc

s

(Q o

ti

M

= B ! CL,

Q, > S j V •O Z

E g - S 01 C 3 CQ 03 H

2

g

> O

. - CO C T3 co

IT o CC ca" z

a U

a o

o

o

a

O

?

o

o CO v-l O oa

+

cd

e -o 3 and B4 polyhedra, and observed that it is important to have a low cluster-charge in order to reproduce experimental bond-lengths. A summary of their results is given in Table 1. A minimal basis-set, STO-3G (Slater-Type Orbitals expanded by three Gaussians: see

Table 1. Optimized bond-lengths (A) using various basis sets for ab inito SCF HartreeFock-Roothaan calculations* STO-3G

6-31G*

B(OH)3

1.364

1.358

1.361

B(OH)4

1.480

1.474

1.478

•from Gupta & Tossell (1981, 1983)

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

47

• * • M• v

°s 01 m

115

118 2 ( ° )

123

*.

' * .

L

•

109 113 ,H

5 . Variation in bond-overlap population with the average of the O-T-O angles common to the bond: (a) BO3 3 " group; (b) BO4 5 " group. Variation in B-O bond-lengths with the average of the O-T-O angles common to the bond: (d) BO33" group; (c) B0 4 5 " group; after Schlenker et al. (1978). Figure

Tossell and Vaughan, 1992, for a readable summary of MO methods), reproduces experimental distances quite closely for B(OH)3° and B(OH)4 polyhedra. Using more extensive basis-sets improves the agreement with experiment even more. Gupta and Tossell (1983) also obtained results consistent with the correlation (Figs. 5a,c) of Schlenker et al. (1978), but note that the calculated effect of bond angle on bond length is smaller than that observed experimentally. Zhang et al. (1985) reported analogous calculations to those of Gupta and Tossell (1983) but with less restrictive symmetry conditions and larger basissets. For B(OH)3, they used a 6-31G* basis-set and no symmetry restrictions. The agreement with the experimental values (Table 1) is very close, and the group optimized to a planar geometry; although the three calculated B-OH distances are identical, the calculated angles differ slightly from 120°, in agreement with the experimental values. Zhang et al. (1985) noted that the calculated geometry conforms almost exactly to C ^ symmetry. Similar calculations for B(OH) 4 with D 2 d symmetry and identical O-B-O angles (Zhang et al., 1985) gave a B - 0 value very close to the experimental average and very close to that obtained for a STO-3G basis-set (Gupta et al., 1981).

Hawthorne, Bums & Grice: Crystal Chemistry of Boron

48

Finite polynuclear clusters. Gupta and Tossell (1983) reported HF calculations done using the STO-3G and 431G basis sets on H 2 BOBH 2 and (OH) 2 BOB(OH) 2 (= [ B 2 0 ( 0 H ) 4 H , both of which contain the [3] B-0- [3] B linkage. The calculated [31 B-0P'B angle is 130°, in close agreement with the average observed value of 134°. Zhang et al. (1985) optimized the geometry of B 2 0(0H)4 assuming planar symmetry and using both STO-3G and split-valence 6-31G basis-sets with and without polarization functions. The former calculations gave geometries that agree with observed values quite closely, whereas the latter calculations only agree with the observed values when polarization functions are used. Releasing the constraint of planar geometry and using the 6-31G and 6-31G* basissets gives a dihedral angle of about 60° (as compared with 0° for the planar arrangement). The barrier to twisting is only about 5 kcal/mol, perhaps accounting for the wide range of observed dihedral angles in B2C>3. Tossell and Vaughan (1992) also note the strong correlation between B-O-B angle and dihedral angle. Zhang et al. (1985) also report HF calculations using the STO-3G basis set with geometry optimization for the dimer [B 2 0(0H)5]; the calculated (observed) values are B-O-B = 122.8(120.5)°, P ' B - 0 = 1.33(1.36) A, ^ B - 0 = 1.46(1.49) A, in quite close agreement. They also examined the energetics of bond bending and showed that the variation in cluster energy as a function of [ 3 lB-0- [41 B angle matches exactly the distribution of experimental angles (Fig. 6).

b -494.9074— ID

o

e

—>

UJ -*) in synthetic Na g [B 12 O20(OH) 4 ] (Menchetti and Sabelli, 1979); after Burns et al. (1995a).

ENUMERATION OF POSSIBLE CLUSTERS In general, FBBs have six or fewer borate polyhedra; there are exceptions to this statement, but they are sufficiently few that we can consider them on an individual basis. Examination of all possible clusters of six or less borate polyhedra (Burns et al., 1995a) indicates that there are many possible clusters that are not observed, whereas others are very common FBBs in borate mineral structures. This suggests some fascinating questions concerning the paragenesis and mechanisms of crystallization of borate minerals. Do FBBs exist as complexes in aqueous brines and hydrothermal solutions? Do borates crystallize by direct condensation of such complexes? Do some complexes change their topological characteristics on crystallization? Are specific borate minerals indicators of the complexes (and hence pH) of their precursor fluids? What controls the occurrence and stability of clusters as FBBs? The borate minerals are an ideal natural laboratory to consider such questions. All clusters of B ^ tetrahedra (3 < n < 6) derived by Burns et al. (1995a) are shown in Figures 17 and 18, together with their corresponding B-B graphs and algebraic descriptors. Such clusters do not include (1) edge-sharing or face-sharing between polyhedra; (2) one-

60

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

connected polyhedra (these are considered as decorated examples of smaller clusters [e.g. 2A2D:(A2n)A (Fig. 13g) is considered as a decorated example of 1A2D(A2D) (Fig. 13f)]. Some of the clusters of Figure 17 and 18 allow some flexibility in selecting their algebraic descriptor. In these cases, the smallest rings within the cluster are emphasized, subject to the restriction that the descriptor gives a complete description of the cluster. Clusters containing both B I ( ==

6D:(=)=

)g \

/

X

B

/

o.

S

V

A< \

S

^

I 0

>/

. A 1

I,

6a:(] I = I A 0

\

/

°

/ V \• / \

"NT

y°

60:[00 % \ -i cn /•>< lo r

•

\>

v \ M II A _ CO /•; co / V X v CD

*

10

v ^

V- IAt

A • 8 V lO.

1/

/>

V

v!" A CD y CO V • 3 1/ 1/

63

A V " _

(T\t

/ • > x

l " -

CO

V

/

> •




V "

V A CD CO V I A CD CO V m lO

V /

/I

& •i

f />

V

•

eo s

v > -

V - •

V *

£ i x i

A

V *—•

< V CD

CO < \

•< •

v - -

4

A "

A:

V-

A"

ao •a &

< —

N.

V >

r

Na[B,0,(0H) < ]

Dal Negro et al. (1975)

1A2D

inderite

(A2D)

Mg[B 3 0 3 (0H) 5 ](H 2 O) 5

Corazza (1976)

1A2D

kurnakovite

(A2D)

Mg[B 3 0 3 (0H) 5 ](H 2 0) s

Corazza (1974)

1A2Q

inyoite

(A2D)

Ca[B 3 0 3 (0H) 5 ](H 2 0) 4

Clark et al. (1964)

1A2D

meyerhofferite

(A2D)

Ca[B 3 0 3 (0H) 5 ](H 2 0)

Burns & Hawthorne (1993b)

1A20

soiongoite

(A20)

Ca 2 [B 3 0 4 (0H) 4 ]CI

Yamnova et al. (1977)

1A20

inderborite

!A2D>

CaMg[B 3 0 3 (0H) 5 ] 2 (H 2 0) 6

Burns & Hawthorne (1994c)

3D

nilontovite

(3D)

Ca 3 [B 3 0 3 (0H)J(H 2 0) 2

Simonov et al. (1978)

2A2D

hydrochlorborite

(A2DJA

Ca 2 [B 3 0 3 (0H) 4 ][B0(0H) 3 ]CI(H 2 0) 7

Brown & Clark (1978)

4D

uralborite

!3D)n

Ca 2 [B 4 0 4 (0H),)

Simonov et al. (1977)

4A1D

sborgite

;2AD)-;2AD>

Na[B 5 0 6 (0H) 4 ](H 2 0) 3

Merlino & Sartori (1972)

2A3D

ulexite

!A2D;-(A2D;

NaCa[B 5 0 s (0H) 6 ](H 2 0)s

2A2D

borax

;A2D)=iA2D)

Na2[B405(OH)4](H2O)e

Ghose et al. (1978) Levy& Lisensky (1978)

2A2D

tincalconite

(A2D)=iA2D)

Na 2 [B 4 0 5 (0H) 4 ](H 2 0) 3

Powell et al. (1991)

2A2D

hungchaoite

;A2n)=;A2n)

Mg[B 4 0 5 (0H) 4 ](H 2 0) 7

Wan & Ghose (1977)

2A2D

fedorovskite

(A2D)=;A2D)

Ca2Mg2(OH)4[B4Or(OH)2]

Malinkoefa/. (1976)

2A2D

roweite

(A2n;=(A2n)

Ca 2 Mn 2 (0H) 4 [B 4 0,(0H) 2 ]

Moore & Araki (1974b)

3A3D

mcaiiisterite

[|iA2

Mg[B s 0 ; (0H)J(H 2 0) 2

Dal Negro et al. (1971)

3A3D

rivadavite

[4>](A2D) | (A2D) J (A2

Na6Mg[B6O;(OH)s]4(H2O)l0

Dal Negro & Ungaretti (1973)

12A3D

ammonioborite

3(!2AD)-(2AD))

(NH4)3[Bt5O20{OH)8J(H2O)4

Merlino & Sartori (1971)

F B B = 2A Two of this group are wallpaper structures and show very strong affinities with the F B B = A wallpaper structures (Hawthorne 1986). Suanite, Mg2[B205], consists of ribbons of parallel edge-sharing octahedra, four octahedra wide, that extend in the [010] direction and are cross-linked in the (101) plane by [B2O5] groups (Fig. 23a). Szaibelyite, Mg2(OH)[B 2 C>4(OH)], and sussexite, M n + ( 0 H ) [ B 2 0 4 ( 0 H ) ] , are isostructural and consist of 1 x 2 ribbons of edge-sharing octahedra that link by sharing vertices to form corrugated octahedral sheets perpendicular to [100] (Fig. 23b); these sheets are cross-linked into a framework by [B2O5] groups. In kurchatovite, CaMgP^Os], Mg octahedra share corners to form a checkerboard pattern resembling a slice through the perovskite structure (Simonov et al., 1980). These [MgOzJ sheets are linked in the [100] direction by [B2O5] groups. One (BO3) group links to two apical octahedral vertices of adjacent sheets, and the other (BO3) group links to two meridional vertices of octahedra from the same sheet (Fig. 23c). Seven-coordinated Ca further links the [MgO^ sheets to form a fairly dense-packed structure. FBB =

•-•

There are only two minerals in this group, and both have [E^CKOH)^] (2D) as their

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

71

Figure 2 3 . Finite-cluster borate structures containing the FBBs 2A, 2 0 and (2Ad): (a) suanite; (b) szaibelyite; (c) kurchatovite; (d) pentahydroborite; (e) pinnoite; (f) ameghinite. In ameghinite, triangles at the front of the figure are shaded black whereas triangles at the back of the figure are left unshaded for clarity.

FBB. Pentahydroborite, Ca[B20(0H) 6 ](H 2 0) 2 , consists of pairs of edge-sharing Ca7 polyhedra cross-linked by [B27] groups into sheets parallel to (100) (Fig. 23d). These sheets are linked directly via H-bonds involving the (OH) groups of the pyroborate group

72

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

and the (H2O) groups that bond directly to the Ca atoms. Pinnoite, Mg[B 2 0(0H)£], consists of octahedrally coordinated Mg linked into a framework by [B27] pyroborate groups. Each (OH) ligand of the borate group links directly to one (Mgg) polyhedron by corner-sharing (Fig. 24d). Interchain linkage is by H-bonding directly between the chains, as there are no interstitial (H 2 0) groups. Solongoite, Ca 2 [B304(0H) 4 ]Cl, consists of sheets of edge-sharing (Cag) polyhedra parallel to (010) (Fig. 24e) that are cross-linked by (A2D) rings. Both (BO4) tetrahedra share edges with the (Cag) polyhedra, and the (BO3) triangle extends outward from the sheet to link by corner-sharing to the adjacent sheet. Inderborite, CaMg[B303(0H)5] 2 (H 2 0)6, is the most complex structure of this particular group. The borate tetrahedra of two (A2D) rings each share a comer with the (Mg^g) octahedron (Fig. 24f); one tetrahedron shares a non-bridging anion, and both tetrahedra share a bridging anion with a single (Cag) polyedron, forming infinite rods parallel to [001] that cross-link via corner-sharing between (BO3) and (Cag polyhedra share edges to form chains extending along [101]. These chains are cross-linked

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

c sin 3 -H

73

|c sin

Figure 24. Finite-cluster borate structures containing the FBB (A20): (a) inderite; (b) kurnakovite; (c) inyoite; (d) meyerhofferite; (e) solongoite; (f) inderborite.

by (3D) rings which share both edges and corners with Cag polyhedra, together with a network of H-bonds. FBB = , (3d)D, Sborgite, Na[B 5 0 6 (0H) 4 ](H 2 0)3, is the only known example of this FBB (Fig. 25d), which consists of two (2AD) rings that have a common (B(t>4) tetrahedron. There are two distinct Na atoms: Na(l) is coordinated by two (OH) groups and four (H2O) groups, and Na(2) is coordinated by two (OH) groups, two (H2O) groups at 2.298 A and a further two (H 2 0) groups at 2.998 A (Merlino and Sartori, 1972). The (Na6) polyhedra and the (2AD)-(2AD) clusters link to form a three-dimensional network that is strengthened by extensive H-bonding. FBB = 4] chains (see end-on in Fig. 25e) that link the polyborate anions into sheets parallel to (110), and these sheets are linked through interstitial Ca and via hydrogen bonding. FBB = in

o m

EE, CM

X

ca a

o

O

•O «

£ o

°

o •O

oo ®

CO

CO >-

O m

X

05

" o 08 CD

CO CO CT>

S CO

X

a ® 0) (/)

»8

(0

a

m

a

c

m

O O

mo

X

o>

•o CD n .>. o

~




91

£ §

a (n •

•

S


, (B4) and (Si04) groups. There are five symmetrically distinct borate groups, one triangle and four tetrahedra, and one distinct (Si04) group. One ( B ^ ) group and two (B4> groups link to form a (A2D) ring of the form [ B j O ^ O H ^ that constitutes an FBB of the structure. These rings then link via corner-sharing between A and • to form chains along the c axis (Fig. 39a) that are topologically (but not geometrically) identical to the chains in colemanite (Fig. 28b). The two remaining ( B ^ ) groups share one vertex to form a [B27] group. Two (Si04) groups link by vertex sharing to (B4) tetrahedra of two [B27] groups to form a four-membered ring of alternating (Si0 4 ) and (B4) tetrahedra (Fig. 39a) with appended (B4) groups. These appended (B4) tetrahedra link to (Si0 4 ) tetrahedra of other four-membered silicate-borate rings to form corrugated walls parallel to (100) (seen in cross-section in Fig. 39b). These walls are linked into a complex framework by vertex sharing with the borate chains parallel to the c axis. Two distinct Ca cations occupy [8]-coordinated cavities within this framework, and hydrogen bonding further strengthens the intra-framework linkage. Hyalotekite, (Ba,Pb,K)4Ca2(B,Si)2(Si,Be)io02gF, has a chemically complex framework of (B0 4 ), (Be0 4 ) and (Si04) tetrahedra. There is one symmetrically distinct (BO4) group that connects by vertex sharing to two (Si04) groups and to two ({Si,Be 104) groups. The (BO4) group links to another (BO4) group and two ({Si,Be}04) groups to form a four-membered ring (Fig. 39c). Four symmetrically distinct (Si04) groups link to form another four-membered ring at a different height along the c-axis. These two types of four-membered rings link by sharing tetrahedral corners to form a framework (not completely connected) with prominent puckered eight-membered rings that share edges with the four-membered rings (Fig. 39c). There is one symmetrically distinct Ca cation, two distinct Pb cations and two distinct Ba cations in interstices of the tetrahedral framework, and both Pb 2+ cations show the asymmetric coordination typical of stereoactive lone-pair behavior (Moore et al., 1982).

102

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

Figure 39. Mixed silicate-borate minerals: (a) howlite; (b) howlite; (c) hyalotekite; (d) kalborsite; (e,f) leucosphenite.

Kalborsite, K^ALtSi^C^oHEHOH^Cl, has a complex aluminosilicate framework with isolated [B(OH)4] groups in the interstices of the framework. There are two distinct (SiC^) groups and one distinct (AIO4) group. These tetrahedra link by sharing vertices to form a 4.8 2 net parallel to (001), and there is alternation of (SiC^) and (AIO4) tetrahedra around all rings in the net (Fig. 39d). The tetrahedra within the rings point {udud} (Up-Down-Up-

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

103

Down) and {uudduudd}, respectively. The nets stack along [001] with a slight angular displacement of adjacent nets, and adjacent nets are linked through an additional (Si0 4 ) group that shares vertices with pairs of tetrahedra belonging to opposing four-membered rings in each net (Fig. 39e). This results in a framework with very prominent channels parallel to [001]; [B(OH)4] groups, [9]-coordinated K cations and CI anions alternate along each channel (Fig. 39d). There is an additional K cation forming the walls of the channels though the tetrahedral framework. Leucosphenite, Na4BaTi2[B2Siio028]02, has three symmetrically distinct tetrahedrally coordinated cation sites, two of which are occupied by Si only and one of which is occupied by Si and B in a 1:1 ratio. The {(Si,B)04] tetrahedra link by sharing corners to form a four-membered ring (Fig. 39f), but because of the disordered nature of the cations, we do not know if there is short-range order involving isolated (BO4) groups (i.e. rings of the form Si-B-Si-B), pyroborate groups (i.e. rings of the form Si-Si-B-B) or even borate rings (i.e. rings of the form B-B-B-B and Si-Si-Si-Si). There are also four-membered rings of (SiC>4) groups, and these four-membered rings link into a framework of tetrahedra. There is one distinct Ti cation octahedrally coordinated by O atoms, and the resulting octahedra form edge-sharing dimers (Fig. 39f). Within the interstices of the framework is one distinct Ba cation in [9]-coordination and two distinct Na cations in [8]- and [9]-coordination, respectively. Taramellite, Ba4(Fe 3+ ,Ti) 4 [Sig02o(B207)]02Cl x , has one symmetrically distinct borate group that polymerizes to form a [B2O7] group. There are two distinct (SiC^) groups that link to form a four-membered ring orthogonal to [100] (Fig. 40a), and adjacent rings are linked by sharing vertices with the [B2O7] group. There is one distinct octahedrally coordinated site occupied by (Fe3+,Ti), and the octahedra share trans edges to form an [(Fe3+,Ti)((>4] chain extending along [100]. These chains are cross-linked into a heteropolyhedral framework by the [B2O7] and [Si40j2] groups, and three distinct Ba cations and a CI anion occupy the interstices. Poudretteite, I G ^ ^ S i ^ C ^ o L is a member of the milarite group of minerals (Hawthorne et al., 1991). It is a framework structure with hexagonal rings of (Si04) tetrahedra cross-linked into slabs orthogonal to [001] by (BO4) tetrahedra (Fig. 40b). These slabs fuse along [001] by sharing opposing vertices of the [Si^Ojg] rings to form [Sij203o] cages (Fig. 40c). The K cation is sandwiched between two [Si^C^o] cages and the Na cation occurs in the channel bounded by both (Si04) and (BO4) groups. It should be noted here that the tetrahedrally coordinated site occupied by B in poudretteite is occupied by a wide variety of cations (i.e. Li, Be, B, Mg, Al, Si, Mn 2+ and Zn) in the minerals of the milarite group. Searlesite, Na[BSi205(0H)2], is based on complex borosilicate sheets. There is one symmetrically distinct ( B ^ ) tetrahedron and two distinct (Si04) tetrahedra. The simplest way to describe the sheet of tetrahedra is as corner-linked [SiC>3] chains parallel to the caxis and cross-linked by [B0 2 (0H)2] tetrahedra (Fig. 40d) in which only two of the four anions of the borate group are involved in the linkage. The resulting sheet is actually a 5 3 net which accounts for its very contorted nature, an aspect that is very apparent in Figure 39e. Stillwellite, ideally (REE^BSiCy, has one symmetrically distinct (BO4) group, and this group polymerizes to form spiral [BO3] chains parallel to [001] (Fig. 41a). There is one distinct (Si04) group, and this decorates the periphery of the [BO3] chain by sharing two of its vertices with adjacent borate tetrahedra (Fig. 41a), producing spiral silicoborate columns parallel to [001] (Fig. 41b); these columns are linked through [9]-coordinated

104

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

REEs. There is significant positional disorder of the anion that links the borate groups into a chain. Burns et al. (1993) suggest that stillwellite crystallizes in a higher symmetry (P3i21) and inverts to P3j on cooling via a ferroic transition.

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

105

Tienshanite, approximately Na2BaMn 2+ TiB 2 Si 6 02o, has one symmetrically distinct (BO4) group and three distinct (Si0 4 ) groups. The (Si0 4 ) and (BO4) groups link into a sheet parallel to (001) by sharing three of their four vertices with other tetrahedra (Fig. 41c). The result is a 6 3 net with three types of rings, an Si-Si-Si-Si-Si-Si ring, an SiB-Si-B-Si-B ring and Si-Si-B-Si-Si-B ring. Note that all (Si0 4 ) tetrahedra point upward and all (BO4) tetrahedra point downward in Figure 41c. The anions of (Si04) groups form the ligands for the octahedrally coordinated (Ti0 4 ) and (MnOj) polyhedra that form a sheet orthogonal to [001]. The tetrahedral-'octahedral'-tetrahedral slabs link along [001] by sharing the opposing vertices of the (B0 4 ) groups, forming [B2O7] groups aligned along [001], The alkali and alkaline-earth cations occur between the back-to-back (Si0 4 ) tetrahedra of adjacent slabs, essentially at the same level as the bridging anion of the [B2O7] group.

(a)

IJJ iri V

106

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

Werdingite, (Mg,Fe2+)2Al14B4Si4037, has two symmetrically distinct (B0 3 ) groups, one of which shows some replacement (BO.77AI023) by (AIO4). There are six distinct (AlOg) groups that share trans edges to form two distinct [AIO4] chains extending parallel to the c-axis (Fig. 41d). There are two (AIO5) and one (Mg,Fe 2+ )05 triangular-bipyramidal groups, and these decorate the [AIO4] chains in a staggered fashion that resembles the arrangement in mullite. In one of the chains, (BO3) groups bridge pairs of adjacent vertices in a staggered fashion on either side of the [AIO4] chain. The resulting heteropolyhedral framework is strongly related to the corresponding arrangement in sillimanite (Niven et al., 1991). Wiserite, (Mn 2+ ,Mg) 14 [B 2 0 5 ]4(0H) 8 [ {Si^Mg*} {0,_ X (0H) X } 4 ]Cl 2x , is a 3-A wallpaper structure. There are two symmetrically distinct (BO3) groups and these share a vertex to form a [B2O5] group (cf., suanite, szaibelyite and sussexite, Table 4). There are four symmetrically distinct (Mng) octahedra and these share trans edges to form the [Mn=(3B> and 5B:. A m Mineral 59:1005-1015 Corazza E, Menchetti S, Sabelli C (1975) The crystal structure of nasinite, Na2[B508(0H)]-2H20. Acta Cryst B31:2405-2410 Dal Negro A, Tadini C (1974) Refinement of the crystal structure of fluoborite, M g 3 ( F , 0 H ) 3 ( B 0 3 ) . Tschermaks Mineral Petrogr Mitt 21:94-100 Dal Negro A, Ungaretti L (1973) Crystal structure of rivadavite. Naturwissenschaften 60:350 Dal Negro A, Sabelli C, Ungaretti L (1969) The crystal structure of macallisterite, Mg2[B6C>7(0H)6]2'9H20. Accademia Nazionale dei Lincei, Rend Classe Sci Fis, Mat Nat XLVII:353-364 Dal Negro A, Ungaretti L, Sabelli C (1971) The crystal structure of aksaite. Am Mineral 56:1553-1566 Dal Negro A, Kumbasar I, Ungaretti L (1973) The crystal structure of teruggite. A m Mineral 58:1034-1043 Dal Negro A, Pozas J M M , Ungaretti L (1975) The crystal structure of ameghinite. A m Mineral 60:879-883 DeWaal SA, Viljoen EA, Calk LC (1974) Nickel minerals from Barberton, South Africa. VII Bonaccordite, the nickel analogue of ludwigite. Trans Geol Soc S Africa 77:373 Dowty E, Clark JR (1973) Crystal-structure refinements for orthorhombic boracite, M g 3 C l B 7 0 i 3 , and a trigonal, iron-rich analogue. Z Krist 138:64-99 Edwards JO, Ross V F (1960) Structural principles of the hydrated polyborates. J Inorg Nucl Chem 15:329-337 Effenberger H (1982) Verfeinerung der Kristallstruktur von synthetischem Teepleit. Acta Cryst B38:82-85 Effenberger H, Pertlik F (1984) Verfeinerung der Kristallstrukturen der Isotypen Verbindungen M3(BC>3)2 mit M = Mg, Co, Ni (Structurtyp: Kotoit). Z Krist 166:129-140 Effenberger H, Zemann J (1986) The detailed crystal structure of nordenskioldine, CaSn(B03)2- N Jahrb Mineral Monatsh 111-114 Egorov-Tismenko YK, Simonov MA, Belov N V (1980) Crystal structures of calciborite Ca2[B03B0]2 and synthetic calciboraluminate 2CaAl[B03]sCa2[A103B0]2. Sov Phys Dokl 25:226-227 Erd RC, Foord EE (1988) Chestermanite, a new member of the ludwigite-pinakiolite group from Fresno County, California. Can Mineral 26:911-916 Fang JH, Newnham RE (1965) The crystal structure of sinhalite. Mineral Mag 35:196-199 Fleet M E (1992) Tetrahedral-site occupancies in reedmergnerite and synthetic boron albite (NaBSi30g). Am Mineral 77:76-84 Foit FF Jr., Phillips M W , Gibbs GV (1973) A refinement of the crystal structure of datolite, C a B S i 0 4 ( 0 H ) . A m Mineral 58:909-914 Geisinger KL, Gibbs GV, Navrotsky A (1985) A molecular orbital study of bond length and angle variations in framework silicates. Phys Chem Minerals 11:266-283 Ghose S (1982) Stereoisomerism of the pentaborate polyanion [B5012]^", polymorphism and piezoelectricity in the hilgardite group of minerals: a novel class of polar borate zeolites. Am Mineral 67:1265-1272

112

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

Ghose S (1985) A new nomenclature for the borate minerals in the hilgardite (Ca2B509ClH20)-tyretskite ( C a 2 B 5 0 9 0 H H 2 0 ) group. A m Mineral 70 6 3 6 - 6 3 7 Ghose S, W a n C (1976) Structural chemistry of borosilicates, part II: Searlesite, N a B S i 2 0 5 ( 0 H ) : Absolute configuration, hydrogen locations, and refinement of the structure. A m Mineral 61:123-129 Ghose S, Wan C (1977) Aristarainite: Na2Mg[B60g(0H>4]2'4H20: a sheet structure with chains of hexaborate polyanions. A m Mineral 62:979-989 G h o s e S, Wan C (1979) Hilgardite, C a 2 [ B 5 0 9 ] C l - H 2 0 : a piezoelectric zeolite-type pentaborate. A m Mineral 64:187-195 Ghose S, W a n C, Ulbrich H H (1976) Structural chemistry of borosilicates. I. Garrelsite, NaBa3Si 2 B7C>i6(OH)4: a silicoborate with the pentaborate [ B 5 O 1 2 ] ' - polyanion. Acat Cryst B32:824 G h o s e S, Wan C, Clark JR (1978) Ulexite, N a C a B 5 0 6 ( 0 H ) 6 ' 5 H 2 0 : structure refinement polyanion configuration, hydrogen bonding, and fiber optics. A m Mineral 63:160-171 Gibbs G V (1982) Molecules as models for bonding in solids. A m Mineral 67:421—450 Giese R F , Penna G (1983) The crystal structure of sulfoborite, M g 3 S 0 4 ( B ( 0 H ) 4 ) 2 ( 0 H ) F . A m Mineral 68:255-261 Giuseppetti G, Mazzi F, Tadini C (1977) The crystal structure of harkerite. A m Mineral 62:263-272 Giuseppetti G , Mazzi F, Tadini C, Larsen AO, Asheim A, Raade G (1990) Berborite polytypes. N Jahrb Mineral Abh 162:101-116 Grice JD, Ercit TS (1986) The crystal structure of moydite. Can Mineral 24:675-678 Grice JD, Ercit TS, Van Velthuisen J, Dunn PJ (1987) Poudretteite, K N a 2 B 3 S i i 2 O 3 0 , a new member of the osumilite group from M o n t Saint-Hilaire, Quebec,and its crystal structure. Can Mineral 25:763-766 Grice JD, Burns PC, Hawthorne FC (1994) Determination of the megastructures of the borate polymorphs pringleite and ruitenbergite. Can Mineral 32:1-14 Grice JD, Bums P C , Hawthorne FC (1996) Borate minerals II. Structural classification. Can Mineral (submitted) Griffen D T (1988) Howlite, C a 2 S i B 5 0 9 ( 0 H ) 5 : Structure refinement and hydrogen bonding. A m Mineral 73:1138-1144 Griffen DT, Ribbe P H (1979) Distortions in the tetrahedral oxyanions of crystalline substances. N Jahrb Mineral A b h 137:54-73 Guo GC, Cheng W D , Chen JT, Zhuang HH, Huang JS, Zhang Q E (1995) Monoclinic M g 2 B 2 0 5 . Acta Cryst C51:2469-2471 Gupta A, Tossell JA (1981) A theoretical study of bond distances, X-ray spectra and electron density distributions in borate polyhedra. Phys Chem Minerals 7:159-164 Gupta A, Tossell JA (1983) Quantum mechanical studies of distortions and polymerization of borate polyhedra. A m Mineral 68:989-995 Gupta A , Swanson DK, Tossell JA, Gibbs G V (1981) Calculation of bond distances, one-electron energies and electron density distributions in first-row tetrahedral hydroxy and oxyanions. A m Mineral 66:601-609 Hawthorne FC (1983) Enumeration of polyhedral clusters. Acta Cryst A39:724-736 Hawthorne FC (1984) The crystal structure of stenonite and the classification of the aluminofluoride minerals. Can Mineral 22:245-251 Hawthorne FC (1985) Towards a structural classification of minerals: the v l M l v T 2 0 n minerals. A m Mineral 70:455^173 Hawthorne FC (1986) Structural hierarchy in V I M x n l T y z minerals. Can Mineral 24:625-642 Hawthorne FC (1990) Structural hierarchy in M(T4) minerals. Z Krist 192:1-52 Hawthorne FC (1994) Structural aspects of oxide and oxysalt crystals. Acta Cryst B 5 0 : 4 8 1 - 5 1 0 Hawthorne FC, Groat LA, Raudsepp M , Ercit TS (1987) Kieserite, a titanite-group mineral. N Jahrb Mineral A b h 157:121-132 Hawthorne FC, Kimata M , Cerny P, Ball N, Rossman, GR, Grice JD (1991) The crystal chemistry of the milarite-group minerals. A m Mineral 76:1836-1856 Heller G (1970) Darstellung und Systematisierung von Boraten und Polyboraten. Fortschr Chem Forschung 15:206-280 Heller G, Pickardt J (1985) Über ein Ikosaborat-ion in hydratisierten Kalium-und Natriumkupferpolyboraten. Z Naturforsch 40b:462-466 Honea RM, Beck FR (1962) Chambersite, a new mineral. A m Mineral 47:665-671 Hoskins BF, M u m m e W G , Pryce M W (1989) Holtite, (Si2.25Sbo.75)B[Al6(Alo.43Tao.27Go.3o)0 i 5 ( 0 , 0 H ) 2 . 2 5 ] : crystal structure and crystal chemistry. Mineral M a g 53:457-463 Ingri N (1963) Equilibrium studies of polyanions containing B ' " , S i ' v , G e ' v and V v . Svensk Kem Tidskr 75:3-34 Kazanskaya EV, Chemodina TN, Egorov-Tismenko YK, Simonov M A , Belov N V (1977) Refined crystal structure of pentahydroborite C a [ B 2 0 ( 0 H ) 6 ] - 2 H 2 0 . Sov Phys Cryst 22:35-36

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

113

Konnert JA, Clark JR, Christ CL (1970a) Crystal structure of fabianite, CaB305(0H), a and comparison with the structure of its synthetic dimorph. Z Krist 132:241-252 Konnert JA, Clark JR, Christ CL (1970b) Crystal structure of strontioginorite, (Sr.Ca^B 14O20(0H) 6 -5H 2 0. Am Mineral 55:1911-1931 Konnert JA, Clark JR, Christ CL (1972) Gowerite, CaB 5 08(0H) B(0H)3'3H20: Crystal structure and comparison with related borates. Am Mineral 57:381-396 Krogh-Moe J (1962) The crystal structure of lithium diborate, Li22B203. Acta Cryst 15:190-193 Krogh-Moe J (1967) A note on the structure of pinnoite. Acta Cryst 23:500-501 Kiihn R, Schaacke I (1955) Vorkommen und Analyse der Boracit- und Ericaitkrystalle aus dem Salzhorst von Wathlingen-Hanigsen. Kali und Steinsalz 11:33-42 Levy HA, Lisensky GC (1978) Crystal structures of sodium sulfate decahydrate (Glauber's salt) and sodium tetraborate decahydrate (borax). Redetermination by neutron diffiaction. Acta Cryst B34:3502-3510 Liebau F (1985) Structural Chemistry of Silicates. Springer-Verlag, Berlin Ma Z, Shi N, Shen J, Peng Z (1981) The refinement of the crystal structure of carboborite MgCa 2 [C03]2[B(0H)4]2-4H 2 0. Bull Mineral 104:578-581 Malinko SV, Shashkin DP, Yurkina KV (1976) Fedorovskite, a new boron mineral, and the isomorphous series roweite-fedorovskite. Zapiski Vses Mineralog Obshch 105:71-85 (in Russian) Malinovskii YA, Belov NV (1980) Crystal structure of kalborsite. Dokl Akad Nauk SSSR 252:611-615 Malinovskii YA, Pobedimskaya EA, Belov NV (1977) Crystal structure of tienshanite. Dokl Akad Nauk SSSR 236:863-865 Malinovskii YA, Jamnova NA, Belov NV (1981) The refined crystal structure of leucosphenite. Dokl Akad Nauk SSSR 257:1128-1132 Matsubara S (1980) The crystal structure of nagashimalite, Ba4(V3+,Ti)4[(0,0H)2]|Cl|Si8B2C>27]. Mineral J 10:131-142 Mazzi F, Rossi G (1980) The crystal structure of taramellite. Am Mineral 65:123-128 Mellini M, Merlino S (1977) Hellandite: a new type of silicoborate chain. Am Mineral 62: 89-99 Menchetti S, Sabelli C (1979) A new borate polyanion in the structure of Nag[B]2O20(OH)4]. Acta Cryst B35:2488-2493 Menchetti S, Sabelli C, Trosti-Ferroni R (1982) Probertite CaNa[B 5 07(0H) 4 ]-3H20: A refinement. Acta Cryst B38:3072-3075 Merlino S, Sartori F (1969) The crystal structure of larderellite, NH4B 5 07(0H) 2 H20. Acta Cryst B25:2264-2270 Merlino S, Sartori F (1971) Ammonioborite: New borate polyion and its structure. Science 171:377-379 Merlino S, Sartori F (1972) The crystal structure of sborgite, NaB506(0H)4'3H 2 0. Acta Cryst B28:3559-3567 Miyawaki R, Nakai I, Nagashima K (1985) Structure of homilite, Ca2 00(F e 0 9()Mno 3o)~ B2.00Si2.00O9.86(OH)0.i4-Acta Cryst C41:13-15 Moore PB (1965) A structural classification of Fe-Mn orthophosphate hydrates. Am Mineral 50:2052-2062 Moore PB (1973a) Pegmatite phosphates. Descriptive mineralogy and crystal chemistry. Mineral Rec 4:103-130 Moore PB (1973b) Bracelets and pinwheels: a topological-geometrical approach to the calcium orthosilicate and alkali sulfate structures. Am Mineral 58:32^12 Moore PB, Araki, T (1972a) Wightmanite, Mg 5 (0)(0H) 5 [B03] «H20, a natural drainpipe. Nature Phys Sci 239:25-26 Moore PB, Araki T (1972b) Johachidolite, CaAl[B30y], a borate with very dense atomic structure. Nature Phys Sci 240:63-65 Moore PB, Araki T (1974a) Pinakiolite, Mg2Mn3+02[BC>3]; warwickite, Mg(Mgo.5Tio.5)0[BC>3]; wightmanite, Mg5(0)(0H)5[B03] Î1H2O: crystal chemistry of complex 3 $ wallpaper structures. Am Mineral 59:985-1004 Moore PB, Araki T ( 1974b) Roweite, Ca2Mn 2+ 2(0H)[B 4 07(0H) 2 ]: its atomic arrangement. Am Mineral 59:60-65 Moore PB, Araki T (1976) Painite, CaZrB[AlgO 1 g]: its crystal structure and relation to jeremejevite, B 5 [G 3 A1 6 (0H)30i 5 ], and fluoborite, B3[Mg9(F,0H)çi09]. Am Mineral 61:88-94 Moore PB, Araki T (1978) Dumortierite, Si3B{Alg 75O17 25(OH)o 75}: A detailed structure analysis. N Jahrb Mineral Abh 132:231-241 Moore PB, Araki T, Ghose S (1982) Hyalotekite, a complex lead borosilicate: its crystal structure and the lone-pair effect of Pb(II). Am Mineral 67:1012-1020 Moore PB, Ghose S (1971) A novel face-sharing octahedral trimer in the crystal structure of seamanite. Am Mineral 56:1527-1538 Nakai I, Okada H, Masutomi K, Koyama E, Nagashima K (1986) Henmilite, Ca2Cu(OH)4[B(OH)4]2, a new mineral from Fuka, Okayama Prefecture, Japan. Am Mineral 71:1234-1239

114

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

Niven ML, Waters DJ, Moore, JM (1991) The crystal structure of werdingite, (Mg,Fe)2Al]2(Al,Fe)2Si4(B,Al)4C>37, and its relationship to sillimanite, mullite, and grandidierite. Am Mineral 76:246-256 Pauling L (1929) The principles determining the structure of complex ionic crystals. Am Chem Soc J 51:1010-1026 Pertlik F, Dunn PJ (1989) Crystal structure of wiserite. Am Mineral 74:1351-1354 Phillips MW, Gibbs GV, Ribbe PH (1974) The crystal structure of danburite: a comparison with anorthite, albite and reedmergnite. Am Mineral 59:79-85 Powell DR, Gaines DF, Zerella DJ, Smith RA (1991) Refinement of the structure of tincalconite. Acta Cryst C47:2279-2282 Prewitt CT, Buerger MJ (1961) The crystal structure of cahnite, Ca2BAs04(0H)4. Am Mineral 46:1077-1084 Pring A, Din VK, Jefferson DA, Thomas JM (1986) The crystal chemistry of rhodizite: a re-examination. Mineral Mag 50:163-172 Raade G, Mladeck MH, Din VK, Criddle AJ, Stanley CJ (1988) Blatterite, a new Sb-bearing Mn 2 + -Mn 3 + member of the pinakiolite group, from Nordmark, Sweden. N Jahrb Mineral Monatsh 121-136 Rastsvetaeva RK, Andrianov VI, Genkina EA, Sokolova TN, Kasbaev AA (1992) Crystal structure of volkovskite. Kristallografiya 37:326-333 (in Russian) Ross VF, Edwards JO (1967) The structural chemistry of the borates. In The chemistry of boron and its compounds. Muetterties, EL (Ed) John Wiley, New York Rumanova IM, Gandymov O (1971) The crystal structure of the natural strontium borate, p-veatchite, Sr2[B 5 08(0H)] 2 B(0H)3 H20. Sov Phys Cryst 16:75-81 Sabelli C, Stoppioni A (1978) Refinement of the structure of hydroboracite. Can Mineral 16:75-80 Schlenker JL, Griffen DT, Phillips MW, Gibbs GV (1978) A population analysis for Be and B oxyanions. Contrib Mineral Petrol 65:347-350 Sen Gupta PK, Swihart GH, Dimitrijevic, Hossain MB (1991) The crystal structure of liineburgite, Mg3(H20) 6 [B 2 (0H)6(P04)2], Am Mineral 76:1400-1407 Shannon RD (1976) Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Cryst A32:751-767 Shashkin DP, Simonov MA, Belov NV (1968) Crystalline structure of the recently discovered natural borate vimsite Ca[B 2 0 2 (0H) 4 ]. Sov Phys Dokl 13:974-977 Shen J, Moore, PB (1984) Crystal structure of cappelenite, Ba(Y,RE)6[Si3Bg024]F2: a silicoborate sheet structure. Am Mineral 69:190-195 Simonov MA, Kazanskaya EV, Egorov-Tismenko YK, Zhelezin EP, Belov NV (1976a) Refinement of the crystal structure of frolovite Ca[B(OH) 4 ] 2 . Sov Phys Dokl 21:471^173 Simonov MA, Yamanova NA, Kazanskaya EV, Egorov-Tismenko YK, Belov NV (1976b) Crystal structure of a new natural calcium borate, hexahydroborite CaB204'6H20sCa[B(0H)]2'2H20. Sov Phys Dokl 21:314-316 Simonov MA, Egorov-Tismenko YK, Belov NV (1977) Accurate crystal structure of uralborite, Ca 2 [B40 4 (0H) 8 ]. Sov Phys Dokl 22:277-279 Simonov MA, Egorov-Tismenko YK, Kazanskaya EV, Belokoneva EL, Belov NV (1978) Hydrogen bonds in the crystal structure of nifontovite Ca2/B503(0H)6/2'2H20. Sov Phys Dokl 23:159-161 Simonov MA, Egorov-Tismenko YK, Yamnova MA, Belokoneva EL, Belov NV (1980) Crystal structure of natural monoclinic kurchatovite Ca2(Mgo 86Feo n K M g o 92Feo 08[B2Os]2- Sov Phys Dokl 25:228-230 Skakibaie-Moghadam M, Heller G, Timper U (1990) Die Kristallstruktur von Ag6[Bi 2 0i8(0H) 6 ].3H20, einem neuen Dokekaborat. Z Krist 190:85-96 Smith JV (1977) Enumeration of 4-connected 3-dimensional nets and classification of framework silicates. I. Perpendicular linkage from simple hexagonal net. Am Mineral 62:703-709 Smith JV (1988) Topochemistry of zeolites and related materials. 1. Topology and geometry. Chem Rev 188:149-182 Snyder LC, Basch H (1969) Heats of reaction from self-consistent field energies of closed-shell molecules. J Am Chem Soc 91:2189-2198 Snyder LC, Petersen GE, Kurkjian CR (1976) Molecular orbital calculations of quadrupolar coupling of 1 ' b in molecular models of glasses. J Chem Phys 64:1569-1573 Sokolova EV, Egorov-Tismenko YK, Kargal_tsev SV, Klyakhin VA, Urusov VS (1987) Refinement of the crystal structure of synthetic fluorian jeremejevite, A l f t f B O j ^ F j . Vesta Mosk Univ Geol 87:82-84 (in Russian) Stephenson DA, Moore PB (1968) The crystal structure of grandidierite, (Mg,Fe)Al3SiBC>9. Acta Cryst B24:1518 Sueno S, Clark JR, Papike JJ, Konnert JA (1973) Crystal-structure refinement of cubic boracite. Am Mineral 58:691-697

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

115

Swinnea JS, Steinfink H (1983) Crystal structure and Mossbauer spectrum of vonsenite, 2Fe0FeB03- Am Mineral 68:827-832 Swinnea JS, Steinfink H, Rendon-Diaz-Miron LE (1981) The crystal structure of Mexican axinite. Am Mineral 66:428^132 Takeuchi Y, Ozawa T, Ito T, Araki T, Zoltai T, Finney JJ (1974) The B2SigO30 groups of tetrahedra in axinite and comments on the deformation of Si tetrahedra in silicates. Z Krist 140:289-312 Takeuchi Y, Kudoh Y (1975) Szaibelyite, Mg2(0H)[B204(0H)]: Crystal structure, pseudosymmetry, and polymorphism. Am Mineral 60:273-279 Takeuchi Y, Haga N, Kato T, Miura Y (1978) Orthopinakiolite, Me2.9502[BC>3]: Its crystal structure and relationship to pinakiolite, Me2.9o02[BC>3]. Can Mineral 16:475-485 Tennyson C (1963) Eine Systematik der Borate auf Kristallchemischer Grundlage. Fortschr Mineral 41:64-91 Tossell JA (1986) Studies of unoccupied molecular orbitals of the B-O bond by molecular orbital calculations, X-ray absorption near edge, electron transmission, and NMR spectroscopy. Am Mineral 71:1170-1177 Tossell JA (1990) First principles quantum mechanical calculation of the electric dipole polarizabilities of the borate, carbonate, nitrate and related ions. Phys Chem Minerals 17:133-141 Tossell JA, Lazzeretti P (1988) Ab initio calculation of B and O nuclear quadrupole coupling constants and NMR shielding constants in molecular models for B2O3 glass. J Non-Cryst Solids 99:267-275 Tossell JA, Vaughan DJ (1992) Theoretical Geochemistry. Oxford University Press, New York Uchida N, MaeKawa T, Yokokawa T (1985) An application of MNDO calculation to borate polyhedra. J Non-Cryst Solids 74:25-36 Vaughan DJ, Tossell JA (1973) Molecular orbital calculations on beryllium and boron oxyanions: Interpretation of X-ray emission, ESCA, and NQR spectra and of the geochemistry of beryllium and boron. Am Mineral 58:765-770 Wan C, Ghose S (1977) Hungchaoite, Mg(H20)5B405(0H)4'2H20: A hydrogen-bonded molecular complex. Am Mineral 62:1135-1143 Wan C, Ghose S (1983) Parahilgardite, Cag[B509]3Cl3-3H20: a triclinic piezoelectric zeolite-type pentaborate. Am Mineral 68:604-613 Yakubovich OV, Simonov BM, Belov NV (1975) Structure refinement for gaudefroyite. Sov Phys Cryst 20:87-88 Yakubovich OV, Yamnova NA, Shchedrin BM, Simonov MA, Belov NV (1976) The crystal structure of magnesium kurchatovite, C a M g ^ O s ] . Sov Phys Dokl 21:294-296 Yakubovich, OV, Egorov-Tismenko YK, Simonov MA, Belov NV (1978) Crystal structure of natural sakhaite Ca 3 Mg[B03]2[C0 3 ] 0.36H20. Sov Phys Dokl 23:225-226 Yamnova NA, Simonov MA, Belov NV (1977) Refined crystal structure of solongoite Ca 2 [B 3 04(0H)4]Cl. Kristallografiya 22:624-626 Yamnova NA, Simonov MA, Belov NV (1978) The distribution of the cations, Sn, Fe 3 + , Fe 2 + , Mg in the crystal structure of hulsite. Dokl Akad Nauk SSSR 238:1094-1097 Yang G, Peng Z, Pan Z (1985) Magnesiohulsite—a new tin-rich borate mineral. Acta Mineral Sinica 5:97-101 (in Chinese with English abstract) Zachariasen WH (1954) The precise structure of orthoboric acid. Acta Cryst 7:305-310 Zha F, Xie X, Lin C, Wang Y (1993) Molecular orbital study of borate polyhedra—an effect of 6-3IG* basis set. J Mineral Petrol 13:1-3 Zhang ZG, Boisen MBJr, Finger LW, Gibbs GV (1985) Molecular mimicry of the geometry and charge density distribution of polyanions in borate minerals. Am Mineral 70:1238-1247 Zoltai T (1960) Classification of silicates and other minerals with tetrahedral structures. Am Mineral 45:960-973

REFERENCES FOR OTHER STUDIES (1996-2002) of minerals and crystal structures found in Chapters 1 and 2 Alexandrov SM, Troneva MA (2000) Isomorphism in borates of the ludwigite-vonsenite series from magnesian skarns of North America. Geochem Int'l 38:144-158 Alexandrov SM, Troneva MA, Kuril'chikova GE (2000a) Tin-bearing borates of hulsite-pageite series from skarn deposits of northeastern Russia: Composition and geochemical evidence for genesis. Geochem Int'l 38:676-688 ["paigeite" refers to hulsite with 75-100% Fe2+ end member] Alexandrov SM, Troneva MA, Kuril'chikova GE (2000b) Boron-tin mineralization in contact aureole at Brooks Mountain, Alaska, the USA: Composition and geochemical evidence for genesis. Geochem Int'l 38:772-787 Andreozzi GB, Lucchesi S, Graziani G (2000a) Structural study of magnesioaxinite and its crystal-chemical relations with axinite-group minerals. Eur J Mineral 12:1185-1194

116

Hawthorne, Burns & Grice: Crystal Chemistry of Boron

Andreozzi GB, Ottolini L, Lucchesi S, Graziani G, Russo U (2000b) Crystal chemistry of the axinite-group minerals: A multi-analytical approach. Am Mineral 85:698-706 Appel PWU, Brigatti MF (1999) Ludwigite from central Sweden: new data and crystal structure refinement. Mineral Mag 63:511-518 Appel PWU, Bigi S, Brigatti MF (1999) Crystal structure and chemistry of yuanfuliite and its relationships with warwickite. Eur J Mineral 11:483-491 Belokoneva EL, Pletnev PA, Spiridonov EM (1997) Crystal structure of low-manganese tinzenite (severginite). Crystallogr Rep 42:934-937 Belokoneva EL, Goryunova AN, Pletnev PA, Spiridonov EM (2001) Crystal structure of high-manganese tinzenite from the Falotta Deposit in Switzerland. Crystallogr Rep. 46:30-32 Burns PC, Carpenter MA (1996) Phase transitions in the series boracite-trembathite-congolite: Phase relations. Can Mineral 34:881-892 Bums PC, Carpenter MA (1997) Phase transitions in the series boracite-trembathite-congolite: An infrared spectroscopic study. Can Mineral 35:189-202 Cámara F, Ottolini L (2000) New data on the crystal-chemistry of fluoborite by means of SREF, SIMS, and EMP analysis. Am Mineral 85:103-107 Cooper MA, Hawthorne FC (1998) The crystal structure of blatterite, Sb 5 VMn 3 \Fe 3+ ),(Mn : \Mg) 35 (B0 3 )| 6 0 32 , and structural hierarchy in Mn3+-bearing zigzag borates. Can Mineral 36:1171-1193 Cooper MA, Hawthorne FC, Grew ES (1998) Refinement of the crystal structure of tienshanite: Shortrange-order constraints on chemical composition. Can Mineral 36:1305-1310 [Am Mineral 84:1467] Ferro O, Pushcharovskii DYu, Teat S, Vinogradova SA, Lovskaya EV, Pekov IV (2000) Crystal structure of strontium hilgardite. Crystallogr Rep 45:410-415 ["strontium hilgardite" is kurgantaite] Fleet ME, Muthupari S (2000) Boron X-edge XANES of borate and borosilicate minerals. Am Mineral 85:1009-1021 Galuskin EV, Galuskina IO (2000) Wiluite, Ca„(Al,Mg,Fe,Ti)13(B,Al,D)5Si,8O68(O,OH)10, a new mineral species isostructural with vesuvianite, from the Sakha Republic, Russian Federation: Discussion. Can Mineral 38:763-764 Garrett DE (1998) Borates Handbook of Deposits, Processing, Properties, and Use. Academic Press, San Diego Giuli G, Bindi L, Bonazzi P (2000) Rietveld refinement of okayamalite, Ca 2 SiB 2 0 7 : Structural evidence for the B/Si ordered distribution. Am Mineral 85:1512-1515 Grew ES, Redhammer GJ, Amthauer G, Cooper MA, Hawthorne FC, Schmetzer K (1999) Iron in komerupine: A " F e Móssbauer spectroscopic study and comparison with single-crystal structure refinement. Am Mineral 84:536-549 Grice JD, Ferraris G (2002) New minerals approved in 2001 by the Commission on New Minerals and Mineral Names, International Mineralogical Association. Can Mineral 40 (in press) Grice JD, Burns PC, Hawthorne FC (1999) Borate minerals. II. A hierarchy of structures based upon the borate fundamental building block. Can Mineral 37:731-762 Groat LA, Hawthorne FC, Ercit TS, Grice JD (2000) Wiluite, Ca ]9 (Al,Mg,Fe,Ti) ]3 (B,Al,D) 5 Si ]8 0 68 (O,OH) l0 , a new mineral species isostructural with vesuvianite, from the Sakha Republic, Russian Federation: Reply. Can Mineral 38:765-766 Hassan I, Duane MJ (1999) The differential thermal analysis of gaudefroyite. Can Mineral 37:1363-1368 Hawthorne FC, Cooper MA, Taylor MC (1998) Refinement of the crystal structure of tadzhikite. Can Mineral 36:817-822 [Am Mineral 84:994] Hoffmann C, Armbruster T, Kunz M (1997) Structure refinement of (001) disordered gaudefroyite Ca 4 Mn 3+ 3 [(B0 3 ) 3 (C0 3 )0 3 ]: Jahn-Teller-distortion in edge-sharing chains of Mn 3+ 0 6 octahedra. Eur J Mineral 9:7-19 Hybler J, Petrícek V, Jurek K, Skála R, Císarová I (1997) Structure determination of vistepite SnMn 4 B 2 Si 4 0 16 (0H) 2 : Isotypism with bustamite, revised crystallographic data and composi-tion. Can Mineral 35:1283-1292 [Am Mineral 83:1120-1121] Irwin MB, Peterson RC (1999) The crystal structure of ludwigite. Can Mineral 37:939-943 Li Y, Burns PC (2000) Refinement of the structure of bandylite. Can Mineral 38:713-716 Ma C, Goreva JS, Rossman GR (2002) Fibrous nanoinclusions in massive rose quartz: HRTEM and AEM investigations. Am Mineral 87:269-276 Marincea S (1999) Ludwigite from the type locality, Ocna de Fier, Romania: New data and review. Can Mineral 37:1343-1362 Marincea S (2000) The influence of Al on the physical and crystallographic properties of ludwigite in three Romanian occurrences. Eur J Mineral 12:809-823 Marincea S (2001) New data on szaibelyite from the type locality, Baita Bihor, Romania. Can Mineral 39:111-127

Continued

on p.

385

Chapter 3 EXPERIMENTAL STUDIES ON BOROSILICATES AND SELECTED BORATES G. Werding and W. Schreyer Research Group on High Pressure Metamorphism Institut flir Mineralogie Ruhr-Universität Bochum D-44780 Bochum, Germany

1. INTRODUCTION Boron, like its two light-element companions lithium and beryllium, is a rare element in the cosmos. Following common belief, it is mainly due to its high volatility that boron is enriched in the Earth's continental crust, where it forms borates mainly in near-surface environments or borosilicates at depth. Thus, taking the thickness of normal continental crust, boron-bearing minerals are restricted approximately to the upper 30 kilometers of our globe, while mantle rocks are extremely poor in boron. This is both caused and enhanced by the fact that during melting in the mantle boron fractionates strongly into the liquid (Chaussidon and Libourel, 1993). Under these circumstances, it was accepted until recently that experimental studies on borosilicates and borates should be limited to pressures below 10 kbar. However, following the discovery of ultrahigh-pressure metamorphism of crustal rocks occurring at depths of 100 km or more (Chopin, 1984), the fate of boron in deeply subducted continental crust became of considerable petrological and geochemical interest as well. As with water, the degree of boron retention in subducted slabs is important for any discussion of the question as to whether all boron present in the mantle has to be regarded as primordial in the sense of a chondritic Earth or whether there is any contribution by crustal recycling. This chapter reviews experimental data on syntheses and stability relations of borosilicates and some selected borates generally up to pressures of 30 to 50 kbar, in a few cases up to 100 kbar. The high-pressure data were largely obtained by our group at Bochum as a contribution to the topic "High-pressure Metamorphism in Nature and Experiment" which is supported by Deutsche Forschungsgemeinschaft. In the literature, experimental mineralogical studies on boron-bearing minerals are scarce compared to those on other rock-forming systems such as boron-free silicates. Therefore, the review to be presented is by no means exhaustive relative to the multitude of borate and borosilicate species known in nature. High-pressure experiments on boron minerals are also of considerable crystal chemical interest Borosilicates and borates contain boron either in planar triangular or in tetrahedral coordination with oxygen and hydroxyl groups (Hawthorne et al.; Grew; this volume). Application of the pressure/coordination rule would suggest that minerals with BO3 groups are stable preferably at relatively low pressures, while those with BO4 tetrahedra are stable at higher pressures. 0275-0279/96/0033-0003$05.00

118

Werding & Schreyer: Experimental Studies

There is a large number of natural borates, most of which contain hydroxyl plus abundant molecular H2O so that their occurrences are limited to near-surface and sedimentary environments. These minerals are largely ignored in this review, but many of them will be found as phases in the model systems treated in §3. We concentrate here on some important borosilicates and borate minerals that occur in more deep-seated geological environments, either igneous or metamorphic. 2. EXPERIMENTAL TECHNIQUES The following conventional techniques are used to investigate the synthesis and stability of boron-bearing systems at elevated pressures and temperatures: (1) hydrothermal bombs up to about 6 kbar, 650° to 800°C; (2) internally heated gas-pressure vessels up to about 8 kbar, 1000°C; and (3) several types of piston-cylinder apparatus up to about 60 kbar, 1100°C. More detailed descriptions were given by Werding and Schreyer (1978; 1990). Experiments at one atmosphere are conducted in conventional furnaces with the charges generally enclosed within noble metal capsules to avoid preferential loss of boron. Some „multianvil" experiments at extreme pressures were performed at Bayerisches Geoinstitut, Bayreuth (see Acknowledgments). At present there are essentially two ways of controlling the component B2O3 in experimental systems: 1. By weighing in stoichiometric amounts of boric oxide or hydroxide relative to the other oxide components necessary for a specific mineral phase, e.g. sinhalite, MgAlB0 4 = 2MgOAl 2 03-B 2 03. 2. By adding hydrous solutions to form fluids under run conditions with specified molar ratios Xg 2 03 = B203/(B203+H20). In the specific case of using solid boric acid, H3BO3 (= sassolite), the resulting X B2 03 is 0.25. Unfortunately, no technique has been developed for buffering the activity of B2O3. The limited potential in the system Ca0-B203-Si02~H20 for such a buffer will be discussed later (§3.3). Future efforts will have to be seen in the light of thermodynamic data such as those derived by Anovitz and Hemingway (this volume). In the experiments conducted at Bochum (see parts of §3 below), the amounts of boron optimal for the synthesis of a particular borosilicate were found empirically. Excess B2O3 of 100% proved to be successful for alkali-free tourmaline (Werding and Schreyer, 1984), whereas for kornerupine the amount of B2O3 added had to be limited to prevent formation of phases with higher B-contents than kornerupine (Werding and Schreyer, 1978). Another problem with boron-bearing systems is that the boron and water contents of the solid phases may not be constant as given in their formulae. Kornerupine and its highboron variety prismatine are known to contain variable amounts of boron (see Grew, this volume). For tourmaline boron is usually assumed to be present in ideal stoichiometric amounts, but water may vary. This problem can only be overcome by careful analyses of the boron and water contents of the synthetic borosilicates. Successful analytical techniques for single-phase products were initially described by Werding and Schreyer (1978). More recently, improved methods were developed for boron determination using alkalimetric titration (Werding and Schreyer, 1990) and plasma emission spectrometry (Werding and Schreyer, 1992). In both cases the relative errors are +2.0% B2O3. For water determination an improved Karl-Fischer titration method was described by Werding and Schreyer (1990) with a relative error of +3.0% H 2 0. Moreover, procedures of analyzing

Werding & Schleyer: Experimental Studies

119

boron by electron microprobe were developed that allow the boron contents of individual synthetic crystals to be determined. Nuclear methods for boron analysis in minerals are described by Robertson and Dyar (this volume). 3. EXPERIMENTS IN BORON-BEARING SYSTEMS Numerous natural borates and borosilicates have been described in the mineralogical literature, and the ceramic literature (Levin et al., 1964, 1969; Levin and McMurdie, 1975; Roth et al., 1981,1983) contains many additional boron-bearing phases that have not been found in nature. For the geoscientific community the study of boron-bearing systems has thus far been a largely neglected field. In this contribution, we will present and subdivide the existing data, as far as possible, on the basis of chemical model systems with increasing complexity. These systems show large inventories of solid phases including borates and borosilicates. Most of the former are low-temperature, low-pressure, near-surface minerals that are identified by their mineral names. They will not be dealt with further in the text. For more information the reader is referred to the Mineral Reference Manual (Nickel and Nichols, 1991) and to the chapter by Smith and Medrano in this volume. Other solids shown in the phase triangles (Figs. 1, 2, 3, 4, 6,9, and 15) by their chemical or structural formulae are purely synthetic phases. Chemically more complicated borosilicates and borates are discussed, after the model systems, in separate subsections towards the end of the section. The text concentrates on those borosilicates and borates that are of relevance to more deep-seated igneous and metamorphic environments. It should also be noted that experimental results recently obtained at Bochum that were made known to the authors in personal communications, or that are described in theses as yet unpublished, will be treated here in relative detail compared to data to be found in the literature. This may necessarily lead to the impression of an over-emphasis of the Bochum work in this review. The main goal was to present to the reader as much of the available knowledge as possible. Before going through the details of the various chemical systems the following general points should be raised: 1. An interesting finding during the preparation of this review was the differential behavior of the component B2O3 as a function of the chemical system involved, or even parts of such a system. Thus, it became quite clear that boron can by no means always be regarded as a „perfectly mobile component" in the sense of Korzhinskii (1959), but that it may often behave as "inert component". A more detailed discussion will be given by Schreyer and Werding (in press), and some conclusions will be drawn in the final section of this chapter. 2. One of the important goals of experimental mineralogy and petrology is to determine the maximum PT-field of stability for a mineral. Just as hydrous minerals have varying stability fields as a function of water activity, boron-bearing minerals may depend in their stability on boron-activity. As boron activity cannot be controlled by a buffer (see Experimental Techniques), the FT-stability of a boron mineral can only be defined for a particular bulk B203-content of the mixture studied by experiment. For the mineral dumortierite an example will be given (see §3.6.2). In order to recognize the maximum (physical) stability, the phase rule can be of help. In a system of n components (including B2O3 and H2O) the breakdown curve is univariant—and thus represents maximum stability —if n+1 phases, that is generally n solid phases plus the at least binary fluid with B2O3 + H2O, coexist. A simple example for the mineral jeiemejevite will be given in §3.5.

120

Werding & Schreyer: Experimental Studies

Importantly, this reasoning applies only to the closed-system behavior as studied in the Bochum experiments. If the systems were open to B2O3 and H2O and controlled by external reservoirs such as perhaps in some natural environments, the breakdown reactions could, in fact, be univariant when involving only n-2 solid phases for a fixed though numerically unknown boron activity. 3. True thermodynamic equilibrium can only be demonstrated by reversed experiments of phase assemblages, one of which is that with the lowest free energy. As many of the experimental results to be reported are of a reconnaissance nature, they may indeed not always represent true equilibrium, the less so the more components and phases there are in the system. Yet most of the curves to be shown were obtained by reversed experiments. For pure synthesis runs equilibrium cannot be claimed. 3.1. The system B 2 0 3 - S i 0 2 - H 2 0 This system (BSH) is an essential limiting system for nearly all other more complex systems to follow. Most importantly, there is no binary B 2 03-Si02 phase nor any hydrous ternary phase. The melting points at 1 atmosphere for Si0 2 (1723°C; Lide, 1990) and B2O3 (450°C; Kracek et al., 1938) differ drastically, so that at the elevated temperatures relevant for deep-seated pedogenesis only quartz or other silica polymorphs will appear as crystalline phases. The minerals sassolite (H3BO3) and metabolite (HBO2) crystallize in the subsystem B2O3-H2O and are limited to temperatures below the melting point of B2O3. Pichavant (1983) reports the phase relations in the BSH-system for 500°C at a fluid pressure of 1 kbar (see Dingwell et al.; this volume). Quartz is the only solid and coexists over most of the compositional space with fluids or melts virtually limited to the binary subsystems B2O3-H2O and B 2 03-Si02, respectively. Thus, in the more complex systems to follow, quartz and other silica polymorphs may also coexist, at elevated PT-conditions, direcdy with the requisite SiC^-saturated borosilicates, and all excess boron will be fractionated into hydrous fluids or melts. In the BSH-system B2O3 certainly behaves as a mobile component. As an aside, framework cage structures containing only the components B2O3 and SiC>2 can be synthesized at temperatures up to 220°C, provided templates of appropriate organic compounds are available that can be incorporated as guest molecules inside the cages. An example of such clathrate compounds is the phase RUB-10 with the formula [(CH3)4N]4*[Si32B4072] (Oberhagemann et al., 1994). These cage structures are also called porosilicates. RUB-10 is thus a poro-borosilicate. Nothing is known about the thermodynamic stability of such phases, which are essentially structural and compositional derivatives of the SiC>2 mineral melanophlogite, (SiC>2 areguest molecules); it is likely that they are metastable. 3.2. The system ^ O - B j C b - S i C ^ - H j O Crystalline phases of this system (NBSH) are shown in Figure 1. Like sassolite, H3BO3, and metaborite, HBO2, the hydrous Na-borate minerals are confined to hydrothermal boron-rich exhalations and lakes in arid climate. Nothing is known about their occurrence and stability at elevated temperatures and pressures. Searlesite, a phyllosilicate with the formula NaB[Si 2 05](0H) 2 , occurs both in this environment and in peralkaline plutonic rocks (Khomyakov and Rogachev, 1991; Wight and Chao, 1995). It was first synthesized by Eugster and Mclver (1959) and found to melt incongruently to reedmergnerite and glass at 300°C and 2 kbar fluid pressure. Hubert et al. (1973) prepared searlesite at 1 atmosphere, 100° and 250°C.

Werding & Schreyer: Experimental Studies

121

F i g u r e 1 . Crystalline phases of the system Na 2 0-B 2 03-Si0 2 -H20 projected from H 2 0 . Minerals are given by their names; for formulae see Nickel and Nichols (1991). Structural and oxide formulae indicate synthetic phases taken from the ceramic literature (Levin et al„ 1964 and 1969; Levin and McMurdie, 1975). Numbers indicate oxide proportions in the order Na 2 0-B 2 0 3 -Si0 2 . For reedmergnerite and searlesite see text The 1:1:2 phase appears to form two polymorphs.

A very interesting ternary phase is reedmergnerite, Na[BSi30g], the boron analogue of the feldspar albite, Na[AlSi30g], Reedmergnerite was also found both in a sedimentary diagenetic environment and in peralkaline pegmatites (for details see Grew et al., 1993). It has been synthesized under hydrothermal conditions between 0.1 and 2 kbar at temperatures up to 500°C (Eugster and Mclver, 1959; Kimata, 1977; Mason, 1980a,b; Pichavant et al., 1984; Fleet, 1992). According to Fleet (1992) there is an order-disorder transformation in this phase at 500° to 550°C at a fluid pressure of 1 kbar affecting the tetrahedral sites similarly as in the case of of albite. At one atmosphere reedmergnerite melts incongruently at 831°C to Si02-phases + liquid (B.J. Skinner cited by Appleman and Clarke, 1965). With increasing water pressure the incongruent melting is lowered drastically to 567°C at 1 kbar and 516°C at 2 kbar (Eugster and Mclver, 1959). At low temperatures, reedmergnerite becomes unstable, in the presence of water, due to the reaction searlesite + quartz = reedmergnerite + water, which was located by Eugster and Mclver (1959) at 285°C for 2 kbar. Because of its isostructural relationship to albite the stability of reedmergnerite at high pressures is of interest. Recent experimentation at Bochum by Ufer (personal communication) indicates that reedmergnerite synthesized at 1.8 kbar not only persists up to 35 kbar, but also grows in the range 25 to 35 kbar, 600° to 700°C, from a mixture of quartz, Na2CC>3 and H3BO3 seeded with previously synthesized crystals of reedmergnerite. Direct synthesis with good yields was achieved at 30 kbar, 500°C using Na2B4(>7 ( N a 2 0 2 B 2 0 3 ; Fig. 1) and aerosil powder as a reactive source of SiC>2. At

122

Werding & Schreyer: Experimental Studies

pressures in excess of 35 kbar, reedmergnerite was found to break down, with or without excess H2O, to coesite plus an unknown phase lying along the join NaBSÌ30g-NaBC>2, and this upper pressure limit could be confirmed by seeded experiments. A preliminary phase diagram is given by Schreyer and Werding (in press). It is thus clear that reedmergnerite is thermodynamically stable up to high pressures, by about 20 kbar in excess of those limiting the stability of albite. In a preliminary fashion Pichavant et al. (1984) determined that there is probably a miscibility gap between reedmergnerite and albite at 1 kbar, 500°C with the coexisting phases having the compositions Rdg7Ab3 and AbgoRdjo- This is compatible with the reedmergnerite-albite assemblage described by Grew et al. (1993) from a peralkaline pegmatite. The latest finding in the NBSH-system is that the unknown phase formed at highpressures by the breakdown of reedmergnerite is in fact the isostructural boron analogue of nepheline with the formula NafBSiO^ (Ufer, pers. comm.). This phase can perhaps be regarded as a high-pressure polymorph of the phase Na20B2C>3,2Si02 shown in Figure 1 that was synthesized at 1 atmosphere by Morey (1951) and found to be isotropic. In nature, several minerals with the formula NaBSi04 are among the unnamed species from Mont Saint-Hilaire (Wight and Chao, 1995; U53; 53A; 53B). 3.3. The system C a 0 - B 2 0 3 - S i 0 2 - H 2 0 The system CBSH (Fig. 2) contains numerous anhydrous and hydrous borates of which perhaps only calciborite, CaB2C>4, may be relevant for elevated PT-conditions. However, the ternary calcioborosilicates danburite, Ca[B2SÌ20g], the stoichiometric boron-analogue of anorthite but with a paracelsian structure, and datolite, CaBSiO^OH), certainly occur in deep-seated environments such as limestone skams and contacts. Morey and Ingerson (1937) successfully synthesized danburite hydrothermally. They also recognized that danburite melts at atmospheric pressure above 996°C to form two coexisting liquids, one practically lying within the limiting system B203-SiC>2and the other having a ternary composition with about equal amounts of all three components. Concerning the relationship between danburite and datolite Eugster and Wise (1963) bracketed the reaction : datolite + quartz = danburite + wollastonite + H2O for 1 to 2 kbar near 500°C, with datolite lying on the low-temperature side. Semenov et al. (1988) in a thermodynamic analysis of the system including CO2 conclude that the breakdown of datolite alone to wollastonite and boron-bearing fluid is strongly dependent on boron-activity, while danburite stability relative to datolite is also influenced by the acidity. The reaction relationship between datolite and wollastonite may open up the possibility of calibrating a boron-buffer for temperatures below 500°C. However, difficulties will arise due to the instability of wollastonite at low temperatures relative to hydrous phases such as xonotlite and tobermorite (Fig. 2), or—in the presence of CO2— relative to calcite + quartz. No experimental data are available for bakerite, Ca4B4(B0 3 0H)[Si04]3(0H)4, and howlite, Ca2B5Si09(0H)5_ which occur not only in sedimentary to diagenetic environments, but also, in the case of bakerite, as a hydrothermal minerei in metamorphic environments (see Grew, this volume). Just as for reedmergnerite, questions arise concerning the high-pressure stability of

Werding & Schreyer: Experimental Studies

123

datolite and danburite, which both contain tetrahedrally coordinated boron. For the framework silicate danburite preliminary experiments (Ufer, pers. comm.) demonstrate that danburite previously synthesized with low yields at low pressures showed strong growth when the low-pressure run product was rerun at high pressures up to 60 kbar. TTiis again suggests that the boron-bearing framework silicate is a stable phase at high pressures, even within the stability field of coesite, and it may exceed its aluminous counterpart anorthite stability-wise by at least 30 kbar. The CBSH-system does not contain boron analogues of the aluminous breakdown products of anorthite, that is of zoisite, kyanite, and grossular, all of which have solely octahedral aluminum.

Urolborite

Colemonile

Vimsite

Meverhofferite

Calcibonte Pentohvdroborite

Invoite

Figure 2. Crystalline phases of the system CaO-BjOj-SiOj-HjO projected from H 2 0. Minerals are given by their names; for formulae see Nickel and Nichols (1991). Structural and oxide formulae indicate synthetic phases taken from the ceramic literature (Levin et al. 1964; Levin and McMurdie 1975). Numbers indicate oxide proportions in the order CaO^C^SiC^. For anhydrous ternary and quaternary minerals see text. For the new mineral takedaite see Kusachi et al. (1995).

3.4. The system M g 0 - B 2 0 3 - S i 0 2 - H 2 0 The system MBSH (Fig. 3) does not contain anhydrous ternary magnesioborosilicates (Kuzel, 1963), nor quaternary phases although forsterite may incorporate small amounts of boron replacing Si (Grew, this volume). All of the hydrous borates except perhaps wightmanite, Mg50(B03)(0H)5.2H20, are near-surface minerals. Of interest here are the anhydrous borate minerals kotoite, Mg3(BC>3)2, and suanite, Mg2B2C>5, because they occur in contact pneumatolytic deposits. Both contain boron in triangular coordination. Suanite has been synthesized by Berdesinski (1955) at atmospheric pressure between 350°C and 850°C. Werding et al. (1981) in an experimental study on sinhalite, MgAlB0 4 , report the synthesis of suanite under hydrothermal conditions as an additional phase

124

Werding & Schreyer: Experimental Studies

between 3 and 20 kbar at temperatures of 400° to 900°C. Kotoite was also synthesized in one run at 5 kbar, 650°C (Werding et al., 1981). More recently, kotoite was obtained up to 50 kbar from mixtures in the system Mg0-B203-H20 (Poter, pers. communication). A surprising result obtained by Krosse (1995) is that szaibelyite, MgBC^OH), was synthesized in a high-pressure run at 40 kbar, 790°C, from a starting material in the system Mg0-Al 2 03-B203-H 2 0 (see §3.7).

Figure 3. Crystalline phases of the system Mg0-B 2 03-Si0 2 -H 2 0 projected from H 2 0. Minerals are given by their names; for formulae see Nickel and Nichols (1991). For references concerning the synthetic Mgborate phases (MgB 4 0 7 and Mg0.B 2 0 3 ) see Werding et al. (1981).The solids in the Mg0-Si0 2 -H 2 0 partial system labelled Phase A etc. are synthetic phases which are listed by Wunder and Schreyer (1992).

Due to the lack of anhydrous ternary or quaternary phases, the system MBSH could principally be an example for a synthetic system, in which quartz (or another SiC>2polymorph) could coexist with borate minerals. Alternatively, the well-known hydrous or anhydrous Mg-silicates such as talc, enstatite and forsterite might coexist with boron-rich siliceous fluids (Fig. 3). Preliminary experiments by the authors show complicated compatibility relations as a function of pressure and temperature, but kotoite and suanite were never found to coexist with an Si02-phase. At pressures up to 10 kbar at about 750°C mixtures of suanite + quartz + water reacted to produce talc or enstatite with some suanite remaining and B-rich fluid developing. The tie line enstatite + suanite only becomes unstable at still higher pressures and temperatures, so that under those conditions forsterite may also coexist with B-rich fluids (Fig. 3). A preliminary compatibility diagram is given by Schreyer and Werding (in press). As usual for Mg-silicate systems the experiments are hampered by quench problems. An interesting experimental project to study would be the question of boron introduction into forsterite, for which Sykes et al. (1994) propose the substitution

Werding & Schreyer: Experimental Studies

125

B+(F,OH) for Si+O. Grew et al. (1991) found that in a natural olivine up to 5 % of the Si can be replaced by B at an estimated pressure of 4 to 5 kbar. The behavior of forsterite in the purely hydrous MBSH system at much higher pressures has important implications for the geochemistry of the Earth's mantle. 3.5. The system A l 2 0 3 - B 2 0 3 - H 2 0 In striking contrast to BSH (§3.1), the ABH-system (Fig. 4) contains a multitude of binary anhydrous Al-borates (which do not occur in nature) and a ternary phase which is known as the mineral jeremejevite, AlgiBC^iF.OH^. According to Foord et al. (1981) natural jeremejevite is predominantiy fluorine-bearing, so that the hydroxyl endmember of interest here would be "jeremejevite-OH." Relevant literature data including synthesis conditions for the anhydrous Al-borates are listed by Stachowiak and Schreyer (1996).

B2O3

AI2O3

Diaspore Boehmite

Gibbsite etc

H2O

Figure 4. Crystalline phases of the system AI2O3-B2O3-H2O. Minerals are given by their names; for formulae see Nickel and Nichols (1991). Oxide formulae in the A^Oj-I^Oj-system stand for synthetic phases, those maiked (S) are structural derivatives of sillimanite. For references see Werding et al. (1981), Reynaud (1977), Mazza et al. (1992), Stachowiak and Schreyer (1996).

Concerning the largely unknown phase relations in the limiting AB-system and—even more—those of the anhydrous silica-bearing system ABS (see Fig. 6 below), Scholze (1956) made the important discovery that the phase 9Al203'2B203 (see Fig. 4) is similar in its physical properties to mullite, 3Al 2 03'2Si02. Meanwhile it has been established that other Al-borates also have mullite- or sillimanite-type structures (Werding and Schreyer, 1992 and Fig. 4 here). Mazza et al. (1992) synthesized at 1 atm., 580° to 900°C from amorphous solids an extensive series of metastable solid solutions between 5A1 2 03 - B203 (= AI5BO9) and Al 2 03-B 2 03 (= AI3B3O9; see Fig. 4) all exhibiting mullite-like X-ray powder diffraction patterns indicating considerable disorder. When heated to 1000°C these

126

Werding & Schreyer: Experimental Studies

metastable precursor phases gradually transformed into the more stable, well defined Alborates 9Al203 - 2B203 and 2A1203-B203, both having more ordered sillimanite-type structures. No detailed crystallographic studies dealing with the ordering scheme involved can be cited.

Figure 5. PT-plot of reversed experiments on the bulk composition 60 mol % "jeremejevite-OH" plus 40 mol % H 2 0 in the ABH-system (see Fig. 4) after Stachowiak and Schreyer (1996) showing the lowtemperature stability field for "jeremejevite-OH" under these compositional constraints. Note that for other bulk compositions the fluid phase contains different amounts of B 2 0 3 in addition to H 2 0 , so that the nature of the assemblages and the location of reaction boundaries may change. This is true for the dasbed lines shown in the low pressure region. The solid line above 15 kbar, however, represents the maximum thermal stability of "jeremejevite-OH". For further discussion see text.

Both the fluorine and the hydroxyl endmembers of the only natural Al-borate mineral jeremejevite, have been synthesized. Sokolova et al. (1987) obtained A l ^ B C ^ F g at 555° to 665°C and 800 to 1000 bar. Capponi et al. (1972) reported the hydrothermal synthesis of the hydroxyl endmember at 800°C and the amazingly high pressure of 35 kbar. In recent experiments at Bochum a preliminary stability field for jeremejevite-OH) was determined (Stachowiak and Schreyer, 1996) which is redrawn here as Figure 5. It extends to at least 50 kbar and shows a maximum thermal stability of about 750°C at pressures near 35 kbar. An important result was that the stability of jeremejevite-OH varies considerably as a function of the bulk compositions studied, that means also of the amounts of H2O and B2O3 present. The stability field given in Figure 5 was determined for the bulk composition Jego-(H20)4o (mol %). Nevertheless, it depicts true thermodynamic stability of jeremejevite-OH for pressures above 15 kbar, because the phase assemblage along the breakdown curve is univariant (3 solids + 1 fluid). For a given P and T the composition of the B2O3-H2O fluid is invariant and can be determined from the ratios of the solid phases present. Stachowiak and Schreyer (1996) conclude that these coexisting fluids become increasingly hydrous with rising pressure so that at 50 kbar, 750°C the molar ratio Xjj2Q3 (= B203/(B203+H20)) is as low as 0.07. This means that boron is strongly fractionated into the solid phases. Fluid composition is buffered by the solids. B2O3 behaves as an inert component. For pressures at and below 15 kbar the breakdown curve shown is divariant and applies only to the bulk composition investigated. Nevertheless, three-phase fields A L ^ O g + jeremejevite-OH + fluid can be constructed for temperatures below the curve, and they show that Xg2Q3 of the invariant fluids decreases with sinking temperatures: at 3 kbar, 450°C it is 0.01 (Stachowiak and Schreyer, 1996).

Werding & Schreyer: Experimental Studies

127

The high pressure stability of jeremejevite-OH (Fig. 5) is of crystal chemical interest as it contains planar BC^-groups in its crystal structure. Thus PlB cannot be principally restricted to low pressures. The occurrence of the phase AI4B2O9 at low pressures only may indicate that this sillimanite-type structure is indeed restricted in its stability to low pressures as are other sillimanite-type phases (see §3.8). Note that the synthesis conditions for jeremejevite-OH given by Capponi et al. (1972), that is 35 kbar, 800°C, lie just outside the stability field for this phase (Fig. 5). The phase must have formed metastably. 3.6. The system AI2O3-B2O3-S1O2-H2O Figure 6 summarizes the crystalline phases appearing in the complex, still largely unknown system ABSH. A particularly intriguing crystal chemical aspect is that there may be a wide range of ternary Al-borosilicates that span the gap between pure Al-silicates and the Al-borates as discussed in §3.5. In addition, the system contains an important rockforming mineral, dumortierite, and, most surprisingly, a tourmaline phase that has most recently been synthesized at high pressures (§3.6.3). Thus, in contrast to the systems discussed previously (Figs. 1, 2 and 3), the Al-bearing borosilicate system encompasses the largest number of solid phases stable at elevated temperatures and pressures. It becomes evident that the presence of Al favors the fractionation of boron into the solids.

Figure 6. Crystalline phases of the system A^Oj-I^C^-SiC^-l^O projected from H 2 0. Minerals are given by their names; for formulae see Nickel and Nichols (1991) except for "boron-mullites" (see §3.6.1) and the new synthetic phase "alkali-free Al-touimaline" (see §3.6.3). For phases in the AI2O3-B2O3-H2Osystem, see Figure 4. The two phases "topaz-OH" and phase Pi were synthesized by Wunder et al. (1993a,b). Numbers indicate oxide proportions AI2O3-B2O3 or A^CVSiC^. For phase Al 8 Si 2 B 2 0 19 see Werding and Schreyer (1992). §3:1:4 represents the bulk composition used for the synthesis of "alkali-free Al-tourmaline."

128

Werding & Schreyer: Experimental

Studies

3.6.1. "Boron-mullites." In his classical paper Scholze (1956) recognized that the Al-borate 9A1 2 0 3 *2B 2 03 was not only similar to the high-temperature Al-silicate mullite, but also that there is mutual solid solution between these phases. Therefore, Werding and Schreyer (1984, 1990) used the term "boron-mullites" for such intermediate members obtained in synthesis experiments. Together with the recent discovery of a wide range of binary mullite-type Al-borates (§3.5; Mazza et al„ 1992) this seems to indicate the existence of an extended ternary miscibility range as indicated in Figure 6, although large parts of it may well be metastable. Rather vaguely defined "boron-mullites" perhaps falling into this category were synthesized by Werding and Schreyer (1984) as high-temperature breakdown products of alkali-free dravite (see §3.8.5). "Boron-mullite" is also among the breakdown products of dumortierite at high temperatures and low pressures (§3.6.2). As a step toward better characterization, the discrete phase AlgSi2B 2 0 1 9 (Fig. 6), whose composition represents a structural unit in the mineral werdingite (Niven et al., 1991; see §3.8.2), has been synthesized at fluid pressures of 1 to 4 kbar, 800° to 830°C (Werding and Schreyer, 1992). Much more experimental and structural work will be necessary to resolve the interesting crystal chemical and miscibility relations of „boron-mullites" which so far have not been found as natural minerals. Importantly, however, Grew and Hinthorne (1983) as well as Grew and Rossman (1985) have found small amounts of boron in natural sillimanites of granulite-facies rocks. In these sillimanites, boron incorporation is linked with introduction of Mg according to the scheme 2 B + Mg for 2 Si + Al + 1.5 O, which is identical to the relationship between werdingite and sillimanite (Niven et al., 1991; Werding and Schreyer, 1992) to be discussed in §3.8.2. 3.6.2. Dumortierite. The geologically most important ternary phase of the ABSHsystem (Fig. 6) is undoubtedly the mineral dumortierite, which occurs in pegmatites as well as in metasomatic and metamorphic rocks up to very high grades. It was long held to be anhydrous (A^BSigOjg), but spectroscopic (Ono, 1981; Werding and Schreyer, 1983) analytical, structural and experimental work proves that dumortierite contains essential hydroxyl groups. Moore and Araki (1978) proposed the following structural formula for dumortierite: OH5.75ao.25)( B0 3)[ Si 3 0 14.25(° H )o.75]- Werding and Schreyer (1990) synthesized dumortierite in the ABSH-system between 3 and 20 kbar, 650° and 800°C, using starting materials with a range of Al/Si ratios and analyzed the single-phase products for all components including H 2 0 . They concluded that there is a ternary range of solid solutions due to the substitution schemes H j ^ A l - j and H^AlSi-j, and that the composition is strongly dependent on the PT-conditions of synthesis. The latter substitution introducing Al into tetrahedral sites is mainly operative at low pressures, whereas the former substitution involves only octahedral Al and predominates at high pressures. Thus a typical structural formula for low P dumortierites formed at 3 to 5 kbar is (Al 6 6 7O 0 33)(BO 3 )[Al 0 49Si2 5 1 Oi3 53(011)! 47], while at high pressures (15 to 20 kbar) it changes to (A1^6gQ032)(BO3)[Ai009^'2.9lÖl3.54(OH)] 06]. Considering both substitutions, a general formula for dumortierite can be given as: MAl 7 . x (B03)[[ 4 ]Al y Si3. y 0 1 5 .3 x . y (0H)3 x + y ] with x > 0.1 and y < 0.5. Figure 7 shows the strong dependence of tetrahedral Al in dumortierite on pressure with only little variation in the total hydrogen contents. As a whole, water contents were found to vary between 1.32 and 2.30 wt % equalling 0.85 to 1.47 hydrogen atoms per formula unit. It is important to note here that this range is only valid for the experimental conditions chosen by Werding and Schreyer (1990), with B2O3 (added as H3BO3) in excess of 200 mol % over the ideal dumortierite stoichiometry and with a rather constant molar B2O3/H2O

Werding & Schreyer: Experimental

Studies

129

ratio of about 0.06 to 0.08. It can be anticipated that the (OH)-contents of dumortierites will decrease with increasing B2O3/H2O of the total system and of the coexisting fluid phase. Htot p f u.

o o

o

Data of Synthesis • 650°C, 5 kbar O 650°C, 3kbar ® 700 °C. 5 kbar • 800 °C. 20 kbar O 800°C, 15 kbar

00

00

1

1

01

1

1

02

1 AI

U1

1

03

1

p.f.u.

1

01

1

05

1

1

06

Figure 7. Relationship between total hydrogen contents (Htot) and tetrahedral A1 in synthetic dumortierite as a function of synthesis conditions as indicated. For discussion see text [Used by permission of the editor of Contributions to Mineralogy and Petrology, from Werding and Schreyer (1990), Fig. 7, p. 21],

The experimental study by Werding and Schreyer (1990) also showed that all their synthetic dumortierites invariably contained 1.0 boron atom per formula unit. This is in contrast to the assumption by Ono (1981) of excess boron replacing A1 in octahedral sites. Another important problem is the PT-stability field of dumortierite, especially with the possibility in mind that the planar BO3 group may not be able to withstand high pressures. Ono (1981) had obtained dumortierite between 9.9 and 17.7 kbar as well, but never as a single-phase product, the unstable pair quartz + corundum being present in addition. In an attempt at synthesizing the hexagonal dumortierite polymorph postulated theoretically by Moore and Araki (1978), Werding and Schreyer (1990) conducted synthesis experiments at 30 to 45 kbar and obtained dumortierite having normal orthorhombic symmetry. In a reconnaissance experiment at 80 kbar, 800°C, using the multianvil apparatus at Bayerisches Geoinstitut Bayreuth, Germany, dumortierite was found to break down into "topaz-OH," Al2SiC>4(OH)2, (Wunder et al., 1993a) plus at least one additional unknown phase (unpublished data). Thus, there is an upper pressure limit to dumortierite stability between 45 and 80 kbar. Concerning the thermal stability of dumortierite unpublished data of the authors are presented in Figure 8 for two different bulk compositions concerning B2O3 and H2O. For details see legend of Figure 8. The experiments at 20 kbar show that the presence of excess B2O3 extends the upper thermal stability of dumortierite by about 30°C relative to that for the stoichiometric mixture, although the B203/H20-ratio of the total bulk composition with excess B2O3 is actually lower than in the stoichiometric mixture. This seeming anomaly may be explained by an inert behavior of the component B2O3 as also encountered in the ABH-system (§3.5). In the stoichiometric mixture virtually all B2O3 present is fractionated into the solid phase below the reaction curve at P = 20 kbar. Thus dumortierite coexists

Werding & Schreyer: Experimental Studies

130

with a very H^O-rich fluid. Only with excess boron present in the runs the fluids become more boron-rich thus causing the slight thermal stabilization of dumortierite as observed at P = 20 kbar. All the experiments indicated by open and filled symbols in Figure 8 represent seeded run reversals. At pressures up to about 8 kbar the assemblage quartz plus a "boronmullite" (see §3.6.1) + fluid appears; at higher pressures (>11 kbar) only the boron-free phases kyanite + corundum are formed, so that all boron is expected to be dissolved in the fluid or melt. The composition of the "boron-mullite" is unknown, but judging from the assemblage it seems to be poorer in boron than the phase AlgSi2B20i9 (see Fig. 6). Note in

200

i

¿00

1

i

600

1

i

800

1

i

.

1000 T I

, CJ

Figure 8. Preliminary PT stability of dumortierite (heavy line), for two different conditions of fluid chemistry below and above 20 kbar. Starting materials seeded with synthetic dumortierite were as follows: Circles and diamonds = kyanite + corundum; squares = sillimanite + corundum; triangle = andalusite + corundum; crosses = kaolinite + diaspore. Solid symbols and crosses = growth of dumortierite. Open symbols = breakdown of dumortierite. At P < 20 kbar (circles) boric acid (200 mol % in excess) plus extra water were added so that B 2 0 3 / H 2 0 = 0.10 [mol/mol], At > 20kbar (diamonds) only boric acid in stoichiometric amount was added giving B 2 0 3 / H 2 0 = 0.33 [mol/mol]. As shown by the break of the curve near 20 kbar excess boron in the bulk composition extends dumortierite stability to somewhat higher temperatures. For further discussion see text. The Al 2 Si0 5 equilibrium diagram shown for comparison is after Holdaway (1971, thin dashed lines) with the sillimanite/kyanite curve extrapolated to higher temperatures.

Figure 8 that the phase relations for dumortierite breakdown between 8 and 11 kbar are not clear at this stage, as they depend on the composition and the pressure stability of the illdefined "boron-mullites". Theoretically two invariant points along the dumortierite breakdown curve are required to interconnect the different assemblages. At any rate, it

Werding & Schleyer: Experimental Studies

131

seems that "B-mullites" are confined to low pressures, which would be in line with the behavior of other boron-bearing derivatives of sillimanite (see §3.5 and §3.8). Note that both at high and low pressures the upper thermal breakdown curves of dumortierite of Figure 8 do not represent univariant reactions and can thus not be taken as true maximum stability limits of dumortierite in the thermodynamic sense. Moreover, the fluids coexisting with the breakdown products are not of invariant composition for a given P and T, which is in contrast to the relations found for jeiemejevite-OH stability above 15 kbar (see §3.5 and Fig. 5). However, judging from the results at 20 kbar (see above and Fig. 8) the authors suggest that the curve located at higher temperatures approaches the maximum thermal stability of dumortierite rather closely. In an attempt to determine a possible lower-temperature stability limit of dumortierite, the authors investigated mixtures of kaolinite + diaspore + H3BO3 seeded with synthetic dumortierite (see crosses in Fig. 8). Growth of dumortierite took place at 3 kbar, 380°C and 5 kbar, 360°C. At 3 kbar, 350°C, no reaction could be observed after 65 days. These results indicate that dumortierite is a stable mineral already at low-grade metamorphic conditions, which is in agreement with observations from nature. Klawa (1996) describes dumortierite in hydrothermally altered granite containing pyrophyllite. Whereas the experimental results on dumortierite reported thus far are strictly confined to the ABSH-system, recent analytical data on natural dumortierite indicate that titanium as well as magnesium may also appear as major components (e.g. Schertl et al., 1991 and Grew, this volume). The incorporation of each of these constituents was also studied separately by experiment. Magnesiodumortierite (Ferraris et al., 1995) has meanwhile been accepted as a new mineral species and represents a phase of the quinary system MgO-AIO3B2O3-SÌO2-H2O (see §3.8.4). Ti-incorporation alone was studied experimentally by the present authors (unpublished data) between 2 and 45 kbar, 650° and 850°C. The synthetic dumortierites showed maximum Ti-contents of about 0.30 Ti per formula unit (18 oxygens) according to the substitution scheme ®Ti + for '^Al + ^ S i . No dependence on the PT-conditions of synthesis could be established. Although - as in pure Al-dumortierite - the lattice constants of the orthorhombic phase approach those of an hexagonal cell with increasing pressure (Werding and Schreyer, 1990), this symmetry was not attained even at the highest pressure. 3.6.3. Alkali-free Al-tourmaline. The presence of an alkali- and Mg-free tourmaline phase in the ABSH-system has been searched for since some time. Werding and Schreyer (1984) have performed experiments using a composition modeled after the hypothetical formula •A^Al^BC^fSi^OigJCtyOH^, which would have a deficiency of water. This composition is shown as 3:1:4 in Figure 6. Runs between 2 and 20 kbar, 750° and 800°C were, however, unsuccessful. Similar attempts by Rosenberg et al. (1986) in the presence of excess B2O3 and H2O at low pressures and 500°C and above also failed. After these experiences, it was a considerable surprise that very recent experimentation at Bochum by Wodara (personal communication) has produced a tourmaline phase in the ABSH-system in the pressure range 20 to 40 kbar and at temperatures of 550° to 650°C. Apparendy the temperatures applied previously by Werding and Schreyer (1984) were too high. In the synthesis attempts by Wodara on the composition previously studied by Werding and Schreyer (1984; see above) tourmaline was not obtained as a single-phase product; quartz (or coesite) and little dumortierite were always observed. Therefore, the composition of the new tourmaline endmember phase should be more aluminous than 3:1:4. In Figure 6 a possible candidate with the oxide ratio 11:3:8 corresponding to a theoretical formula A l j A l ^ B C ^ f A ^ S i ^ i g K O H ^ is indicated. Although a tourmaline

132

Werding & Schreyer: Experimental Studies

with such high amounts of tetrahedral A1 may seem unlikely to exist at high pressures, microprobe analyses of the extremely fine grained felt of tourmaline needles synthesized appear to support compositions at least close to 11:3:8 (Wodara, pers. comm.). Because all peaks of the powder X-ray diffraction pattern of the new tourmaline are shifted to higher angles 2 relative to those of alkali-free dravite, •(Mg 2 Al)Al 6 (B03)3[Si 6 0ig](0H)4 (see §3.8.5), it seems clear that the smaller A1 atom occupies octahedral sites otherwise occupied by Mg. Another open question is whether or not the new endmember tourmaline is a stable phase. Time studies on the 3:1:4 composition (Fig. 6) show that in long runs the tourmaline phase grows at the expense of dumortierite. If then this tourmaline were a stable phase in the ABSH-system, why has it not been observed in nature? At any rate, a possible stability field can only lie at relatively low temperatures. Since dumortierite is a stable phase over a large PT-range (Fig. 8), a possible stable breakdown of the ABSH-tourmaline with increasing temperature should lead to the assemblage dumortierite + corundum + boronwater fluid (see Fig. 6). The existence of the Na-bearing Al-tourmaline olenite (see §3.9.4) and the yet unconfirmed finding of a natural Na-poor Al-tourmaline by Ertl (1995) may open up new aspects of tourmaline crystal chemistry. 3.7. The system M g O - A l j C V B j C V H j O All the phases of the ABH system (§3.5) as well as the silica-free ones of MBSH (3.4) are shown again in the respective limiting systems of the MABH quaternary displayed in a projection from H2O (Fig. 9). In the following, only the ternary anhydrous phases will be considered.

Figure 9. Crystalline phases of the system MgO-A^Oj-I^Oj-I^O projected from H 2 0 . Minerals are given by their names; for formulae see Nickel and Nichols (1991). For identification of synthetic phases given by formulae see legends to Figures 3 and 4. All ternary MgAl-borates are discussed in the text (see §3.7.1-§3.7.3). Numbers indicate oxide proportions in the order Mg0-Al 2 0 3 -B 2 03.

Werding & Schreyer: Experimental Studies

133

20

1*1

0

I

I

t

'BreakI down i products

mo

600

_

Tit]

c

i - O d d on 1000

T2M

Figure 10. Experimental data bearing on the PT stability range of sinhalite. Solid dots indicate direct sinhalite syntheses. Open and half filled circles represent sinhalite growth in seeded runs. Solid square: breakdown of sinhalite. For discussion see text. [Used by permission of the editor of Neues Jahrbuch Jilr Mineralogie Abh., from Werding et al. (1981), Fig. 5, p. 213],

3.7.1 Sinhalite. The only accepted ternary anhydrous phase of the MABH-system is the mineral sinhalite, MgAlBO^ a borate occurring in high-grade skarns and gneisses. It has an olivine-type structure with boron in tetrahedral coordination. Whereas initially direct synthesis of this phase at one atmosphere had failed (Claringbull and Hey, 1952), Werding et al. (1981) were able to synthesize sinhalite at one atmosphere from seeded runs in sealed capsules up to 1050°C. At 1100°C sinhalite was found to break down into suanite, spinel, and probably a I^C^-rich melt (see Fig. 9). In Figure 10 the results of Werding et al. (1981) are reproduced suggesting that sinhalite is a very stable phase also at higher fluid pressures up to 20 kbar, and at temperatures between 400° or even lower and some 1075°C. A possible indication for the existence of a low-temperature stability limit of sinhalite under hydrous conditions and at very high pressure is the result of a synthesis run by Krosse (1995) at 40 kbar, 790°C, on sinhalite composition, which yielded "pseudosinhalite" (see §3.7.3) and the hydrous phase szaibelyite, MgBC^OH). However, due to the short run duration of 24 hours this cannot be taken as evidence. Concerning the high-pressure behavior of sinhalite, Capponi et al. (1973) had already synthesized sinhalite at high pressures between 20 and 80 kbar, 1000° to 1200°C and had concluded that it is "a remarkably stable high pressure phase." Werding et al. (1981) demonstrated experimentally that sinhalite does not exhibit notable solid solution towards spinel, MgA^O^, in the MAB(H)-system which would require substitution of boron by aluminum. However, minor deviations from stoichiometry such as those detected by Hayward et al. (1994) in natural sinhalite were not investigated. For compositions extending toward the isostructural phase forsterite, Mg2SiC>4, again no solid solution could be detected in sinhalite. Instead, other phases like kotoite, suanite and a tourmaline phase formed, with or without coexisting sinhalite (Werding et al., 1981). These relations must be seen as compatibility problems within the quinary system with S i 0 2 (MABSH; see §3.8).

134

Werding & Schreyer: Experimental

Studies

3.7.2. Al analogue of magnesiohulsite(?). Werding et al. (1981) had obtained an unknown phase called X in runs at or near one atmosphere and high temperatures. From similarities in the powder X-ray diffraction lines with the iron borate mineral hulsite, they hypothesized that phase X could be the MgAl-analogue of hulsite with a formula Mg2Al[02/BC>3] equalling 4 MgO-A^C^-I^C^ as shown in Figure 9. For further notes on hulsite see §3.13. 3.7.3 "Pseudosinhalite." Recent experimentation at our Institute at Bochum led to a rather well-defined quaternary phase in die MABH-system. It was discovered through studies in the MABSH-system on the mineral grandidierite (Heide, 1992) as well as by work on dravite, an Na endmember of the tourmaline series (Krosse et al., 1993). In synthesis runs for grandidierite, MgA^BSiOg, at relatively low temperatures and pressures Heide (1992) discovered a phase which showed an X-ray powder diffraction pattern that differed in several respects (additional peaks; and displacement of others) from that of sinhalite. Therefore, he called this phase sinhalite/B. Krosse et al. (1993) detected in their high-pressure synthesis products on the composition of dravite, NaMgjAlgiBC^^fSigOigKOH^, euhedral crystals of an unknown phase, that could be analyzed by microprobe. Neglecting small amounts of silica, the composition was found to be close to or identical with an oxide formula of 4Mg03Al203-2B 2 03 equalling N ^ A ^ ^ O ^ or—normalized to four oxygens as in s i n h a l i t e — M g o ^ A l i ^ B o ^ O ^ This composition lies exactly on the join sinhalitecorundum in the MABH-system (Fig. 9). With this knowledge the new phase could be synthesized from appropriate silica-free starting materials in the MABH-system, often as a single-phase product (Krosse, 1995). Because of its kinship to sinhalite the new phase was named in a preliminary fashion "pseudosinhalite" (Krosse et al., 1993). Because the totals of the microprobe analyses (including B2O3) were close to 100 %, and because IR-spectra of "pseudosinhalite" do not exhibit typical OH-stretehing bands in the range of about 2800 to 3500 cm"1, it was assumed by Krosse et al. (1993) and Krosse (1995) that "pseudosinhalite" is an anhydrous phase with the formula as given above. However, more recent water determinations and thermogravimetric work on the synthetic single-phase material clearly showed the presence of about 2.9 wt % water, which corresponds either to one mole H2O added to the oxides 4Mg03Al203'2B 2 03, or to one (OH)-group for a total oxygen content of 10. Based on the crystal structure of "pseudosinhalite" determined by Daniels (personal communication) and to be discussed later, water must be present as hydroxyl-groups. Thus the formula of "pseudosinhalite" is actually Mg 2 Al 3 B 2 d)(OH). In Figure 11 the X-ray powder diffraction patterns of single-phase sinhalite and "pseudosinhalite" are presented. Considering the similarities between the two patterns, the danger is evident that "pseudosinhalite," especially in mixtures with other phases, can easily be mistaken for true sinhalite. The X-ray diffraction peaks of "pseudosinhalite" can be indexed on the basis of a monoclinic cell with the cell dimensions given in Table 1. This table also summarizes other physical properties of the new phase and compares them with those of sinhalite. X-ray powder diffraction data of "pseudosinhalite" are listed in Table 2. The synthetic crystals obtained (Krosse et al., 1993; Krosse, 1995) are always twinned, (101) being the twin plane (Fig. 12). Salient features of the crystal structure of "pseudosinhalite" were already reported in an abstract (Krosse et al., 1993). Like the olivine-type sinhalite, "pseudosinhalite" is based on

Werding & Schreyer: Experimental Studies

135

"Pseudosinhalite "

JJLJ

IaaWU^.

Sinha/ite

10

J l

20

30 40 ° 2 0 CuKa

50

60

Figure 11. Powder X-ray diffraction diagrams of sinhalite, MgAlB04, and of the new phase "pseudosinhalite," Mg2Al30[B04]2(0H).

Table 1. Physical properties of sinhalite, MgAlBO, and "pseudosinhalite," Mg2Al30(B04)2(0H). Sinhalite (Werding etal. 1981) a b c V

5.676(3) A 4.329(1) A 9.868(3) A 242.4(1) A3

Z = 4 1.664(3) nn* y 1.694(3) 1.702(3) nz A 0.038 2VX 53°

"Pseudosinhalite" (Krosse etal. 1993) a 7.461(4) A b 4.335(2) A c 9.832(5) A V 297.6(1) A3 P 110.63(1)° Z == 2 n* 1.691(1) % 1.713(1) ny 1.730(1) 0.039 A 2VX 80°

Figure 12. SEM-photograph of a twinned crystal of the new phase "pseudosinhalite" taken from Krosse (1995).

hexagonal close packing of oxygens. The occupancy of octahedral and tetrahedral sites is different, however, for the two structures, which are compared in Figure 13 in polyhedral models projected along b. In both structures boron occupies tetrahedra, Mg and A1 occupy octahedra. One octahedral site in "pseudosinhalite" is fully occupied by Al, the remaining two contain Mg and Al disordered. Based on the same oxygen contents of 20 for both structures, the cation sum for sinhalite, M ^ A J ^ f B O ^ , is 15.0, whereas for "pseudosinhalite," IV^AlgCyBO^^OH^ it is only 14 with one boron tetrahedron unoccupied. The substitution scheme leading from sinhalite to "pseudosinhalite" is thus Al 3+ + 2H + for Mg 2+ + B 3+ . The anomaly that no OH-stretching vibrations can be observed in the IR-spectra of "pseudosinhalite" can be explained by some very short O-O distances

Werding & Schreyer: Experimental

136

Studies

Table 2. X-ray powder data of „pseudosinhalite", synthesized at 40 kbar, 803°C, 114 hours (Krosse 1995). h

k

I

I obs.

2 © obs.

2 © cale

d obs.

d calc.

-1

0

2

30

18.98

18.979

4.672

4.6719

0

0

2

5

19.25

19.281

4.607

4.5993

0.031

0

1

1

27

22.64

22.660

3.924

3.9206

0.020

1

1

0

3.6813

-0.045

10

24.20

24.155 }

2 © obs " 2 © -0.001

3.675

-1

1

1

3.6611

-0.090

-2

0

2

4

26.00

24.290 26.026

3.424

3.4207

0.026

1

0

2

11

26.62

26.695

3.346

3.3365

0.075

1

1

1

70

27.70

27.704

3.218

3.2172

0.004

-2

1

1

68

31.68

31.638

2.822

2.8256

-0.042

2

1

0

3

32.92

32.935

2.718

2.7172

0.015

-2

1

2

4

33.35

33.340

2.684

2.6851

-0.010

-1

1

3

70

34.38

34.370

2.606

2.6069

-0.009

0

1

3

45

35.83

35.843

2.504

2.5031

0.013

-1

0

4

2.4572

0.433

2

1

1

2.4306

-0.019

36.537 J-

48

2.429

36.97 36.951

2

0

2

10

-2

1

3

62

37.42

37.426

2.401

2.4008

0.006

37.70

37.682

2.384

2.3851

-0.018

-2

0

4

8

38.50

38.506

2.336

2.3359

0.006

3

0

0

8

38.63

38.691

2.329

2.3252

0.061

0

0

4

4

39.10

39.138

2.302

2.2997

0.038

1

1

3

2.1659

0.023

-3

1

1

-1

1

4

2

1

2

75

43.03

43.032

2.100

2.1001

0.002

-3

1

3

8

44.90

44.874

2.017

2.0181

-0.026

41.663 ]•

100

41.64

42.055

2.167

42.243

2.1467

-0.415

2.1375

-0.603

3

1

1

5

48.40

48.381

1.879

1.8797

-0.019

1

2

2

6

50.19

50.155

1.816

1.8173

-0.035

-4

0

4

5

53.50

53.532

1.711

1.7103

0.032

-4

1

1

18

54.20

54.206

1.691

1.6907

0.006

-3

1

5

6

55.79

55.762

1.646

1.6471

-0.028

-3

2

2

88

56.62

56.627

1.624

1.6240

0.007

2

2

2

65

57.21

57.218

1.609

1.6086

0.008 -0.030

0

2

4

8

58.50

58.470

1.576

1.5771

2

1

4

9

59.30

59.305

1.557

1.5569

0.005

1

1

5

21

60.25

60.226

1.535

1.5353

-0.024

4

1

1

8

61.31

61.287

1.511

1.5112

-0.023

3

1

3

10

61.72

61.729

1.502

1.5014

0.009

-4

1

5

17

62.80

62.792

1.478

1.4785

-0.008

0

3

1

36

65.30

65.328

1.428

1.4272

0.028

-4

2

2

5

66.04

66.077

1.413

1 4128

0.037

-4

2

1

18

66.70

66.712

1.401

1.4009

0.012

5

0

0

54

67.03

67.022

1.395

1.3951

-0.008

* = not used for refinement Lattice constants (A): a« = 7.456(1), bD = 4.334(1), c0 =9.832(2), (3 = 110.67(1)°, V0 = 297.23(7) (A3)

Werding & Schreyer: Experimental Studies

137

of only about 2.50 A. This indicates strong hydrogen bridging which in turn leads to displacement of the OH-stretching frequencies towards lower values. According to Novak (1974; Fig. 6) an O-O distance of 2.50 A is correlated with OH-stretching frequencies near 1300 cm"1. However, in this range the IR-spectrum of "pseudosinhalite" exhibits numerous bands attributed to lattice vibrations. A manuscript summarizing structural, physical and chemical properties of "pseudosinhalite" is presendy being prepared by Daniels et al. •

SINHALITE

"PSEUDOSINHALITE" c

Figure 13. Comparison of portions of the crystal structures of sinhalite and the new phase "pseudosinhalite" projected down [010] (after Daniels, pers. comm., 1995). Dark ruling of octahedra in sinhalite indicates occupancy by Al, lighter ruling by Mg. In "pseudosinhalite" light ruling shows also MgO e octahedra, the darker one those containing mainly Al, but also some Mg. Unit cells are outlined.

500

600

700

800

T[°C]

900

1000

Figure 14. PT-diagram after Krosse (1995) showing the reversed dehydration curve "pseudosinhalite" = sinhalite + corundum + H 2 0 marking the upper temperature and lower pressure limit of the former phase. Solid squares = growth of "pseudosinhalite;" open squares = breakdown of "pseudosinhalite."

"Pseudosinhalite" exhibits a huge field of pressure stability pressure ranging from syntheses by Heide (1992) at 4 kbar to others by Krosse (1995) at 40 kbar. The experimentally reversed dehydration curve determined by Krosse (1995) in the presence of excess water is given in Figure 14. Breakdown products of "pseudosinhalite" are sinhalite + corundum + H 2 0. Below the curve of Figure 14 "pseudosinhalite" and sinhalite seem to coexist over a wide PT-range (compare Fig. 10). Just as for sinhalite, a lower thermal stability limit of "pseudosinhalite" is not known.

Werding & Schreyer: Experimental

138

Studies

It seems that the phase called "pseudosinhalite" here has not been found in nature as yet. However, it may have been misidentified as sinhalite. A possible candidate might be the sinhalite described by Shabynin (1956) with the formula M

g0.95 Fe2+ 0.15 A1 1.54Si0.06 B 0.3lO4-

provided its analytical boron value is too low. Otherwise, its (Mg+Fe 2+ )/Al-ratio is remarkably similar to that in "pseudosinhalite". Even the low Si contents might be a positive indication, because the initial crystals synthesized by Krosse (1995) from dravite composition contained up to 0.08 Si per formula unit sinhalite as well. The locality of this sinhalite is the Tayozhnoye deposit in the Aldan shield, where conditions for boron metasomatism are estimated to have been 4 to 5 kbar, 600° to 700°C (Grew et al., 1991). This would be in the range, where Heide (1992) synthesized "pseudosinhalite;" see also Figure 14.

Si02 Na20)] o B - f r e e phases • B - b e a r i n g phases H 2 0-bearing phases underlined

"Pseudo- ^Spinel sinholite" 1 *1 -4:31:21

Figure 15. Crystalline phases of the system MgO-A^Oj-I^C^-SiC^-HjO projected from B 2 0 3 and H 2 0. The mineral dravite containing additional sodium is also indicated and projected from Na 2 0 as well. For the boron-bearing minerals of limiting systems see Figures 6 and 9, for boron-free ones Schreyer (1988). The boron-bearing MgAl-silicates are discussed in §3.8. Numbers indicate oxide proportions in the sequence Mg0:Al203:Si02(:B203).

3.8. T h e system

MgO-A^Os-BjOa-SiC^-K^O

This system contains by far the most numerous and also some of the most common borosilicates and borates occurring as minerals in nature. For the probably best known borosilicates, the tourmalines, an endmember phase exists in this system, which will be named alkali-free dravite here. All the relevant boron-bearing phases are shown by solid dots in the phase triangle of Figure 15, which includes also some important phases of the MASH-system. The phases are projected from B2O3 and H2O, and, only for the sake of showing the composition of dravite, from Na20. Other sodium-bearing phases were neglected.

Werding & Schleyer: Experimental Studies

139

There are five "pseudoternary" (= quaternary and quinary) borosilicate phases present in the MABSH-system (Fig. 15). These are in the order of increasing SiC^-contents: grandidierite, werdingite, kornerupine, magnesiodumortierite, and the alkali-free dravite endmember. Experimental results bearing on these phases will be reported in the above order in the following §3.8.1 to §3.8.5. 3.8.1. Grandidierite. Structurally, the mineral grandidierite belongs to a family of phases that can be regarded as derivatives of sillimanite, in which SiO^tetrahedra of the chains are replaced by BO3 groups with oxygen being lost. At the same time in grandidierite Mg replaces Al, again in connection with oxygen loss. Thus, grandidierite, MgAl 3 BSiOo, is derived from sillimanite, Al4Si20jo, by the combined substitutions B 3 + = S r + + 0.5 Cr and Mg 2 + = Al 3 + + 0.5 O2" in the ratio 1:1. Actually, grandidierite contains Mg and part of its Al in five-fold coordination and is thus closer to andalusite than to sillimanite (Stephenson and Moore, 1968), but this does not invalidate the overall picture. In Figure 16 the three-dimensional plot introduced by Werding and Schreyer(1992) summarizes the mutual relationships of all the sillimanite-derivatives. Note that werdingite (see §3.8.2), the new phase AlgSi2B20i9, other "boron-mullites" (see §3.6), many Alborates (§3.5) as well as normal mullites also belong to this family. Endmember grandidierite as well as iron-bearing members, (Mg,Fe)Al 3 BSi09 with XFe up to 0.3, were synthesized under hydrothermal conditions at relatively low pressures and high temperatures by Olesch and Seifert (1976) as well as by Werding and Schreyer (1978, 1984, 1992). A more detailed study including preliminary work on the stability of endmember grandidierite was performed at Bochum by Heide (1992). He found the phase to melt incongruently at atmospheric pressure at temperatures between 1200° and 1225°C to spinel, a "boron-mullite" and silica-rich melt (compare Fig. 15). Concerning a possible low-temperature stability limit of grandidierite, Heide (1992) found the synthetic Mg endmember to persist at a fluid pressure of 4 kbar between 300° and 500°C, but growth from a seeded gel could not be achieved under these conditions even after some 500 hours. At 4 kbar, 600°C grandidierite grew, so that the PT-stability of this phase could lie at about 550°C or even lower. As a possible low-temperature breakdown assemblage the phases alkali-free dravite, "boron-mullite," and little chlorite plus spinel were obtained from the gel. It seems likely that the "boron-mullite" phase formed here showing very broad X-ray powder diffraction lines contains Mg in solid solution in addition to B, Al, and Si. This expands the ternary solid solution range of "boron-mullites" shown in Figure 6 into a quaternary one by virtue of the mechanism for Mg-incorporation Mg for Al + 0.5 O (see Fig. 16). Further discussion on Mg-bearing "boron-mullites" will follow in connection with werdingite syntheses (§3.8.2). Certainly, the problem of a lowtemperature stability limit of grandidierite under hydrous conditions requires additional studies. An important result of the Heide (1992) study is that grandidierite becomes unstable already in a medium-pressure range, between about 5 and 10 kbar depending on temperature and water activity (Fig. 17). In the presence of excess water and H3BO3 the high-pressure breakdown assemblage was found to be alkali-free dravite + corundum + "pseudosinhalite" (see §3.7.3) + spinel. The requisite univariant reaction curve determined by reversed runs (Heide, 1992) is shown by the solid line in Figure 17. Note, however, that at temperatures in excess of 700°C "pseudosinhalite" is replaced by sinhalite, without any noticeable change in slope. Importantly, grandidierite breakdown towards higher pressures is not confined to excess-fluid conditions; in water-poor mixtures, using B2O3 instead of H3BO3, thus containing only some 1.7 wt % of water due to the hygroscopic

140

Werding & Schreyer: Experimental Studies

Corundum, i - A l 2 0 3

Spinel

Figure 16. Three-dimensional diagram to demonstrate mutual compositional and substitutional relationships between sillimanite and its structural derivatives in the system MgO-AljOj-SiC^-BjC^. The sum of cations fixed at 24 is used as a common basis. Solid dots indicate known crystalline phases; open circles: no such phase exists, formula in brackets. Minerals shown in boxes have sillimanite-type structures, those without boxes are of other structural types. The Al-borates shown are probably also sillimanite-derivatives, and so is mullite-like i-A1 2 0 3 (Foster 1959). [Used by permission of the editor of the European Journal of Mineralogy, from Werding and Schreyer (1992), Fig. 1, p. 194],

nature of B2O3, grandidierite was found to decompose in seeded runs into dumortierite, kyanite, and sinhalite at 15 kbar, 775°C and above, while at 10 kbar, 775° to 840°C, no reaction was detected. This seems to indicate that grandidierite as one of the sillimanitederived phases cannot persist into PT-regions in which the parent phase sillimanite is not stable itself (Fig. 17). Although the composition of grandidierite is ideally located along the binary join A^SiOj-sinhalite (see Fig. 15), and although this even holds in quaternary space (MgA^BSiOp = MgAlBC^+A^SiOs), dumortierite was always obtained as an additional breakdown product of grandidierite. This may be due to the unavoidable presence of some water stabilizing the hydrous phase dumortierite. In fact, dumortierite may contain some Mg (see Fig. 15 and §3.8.4), so that this phase can interfere with the kyanite-sinhalite assemblage. It should be noted that this grandidierite breakdown reaction (dashed line in Fig. 17) in the quinary system is at least divariant and may shift as a function of fluid composition. This may even hold for an inert behavior of B2O3. Unfortunately, contrary to §3.5, fluid compositions could not be determined in this complex system. 3.8.2. Werdingite. This newly discovered mineral (Moore et al., 1990; Niven et al., 1991) also belongs structurally to the family of sillimanite-derived phases. Following Werding and Schreyer (1992), werdingite is obtained from sillimanite by the combined

Werding & Schleyer: Experimental Studies

141

15-

ueuKuuwu

/

^

Products

. . nearly anhydrous conditions

excess-H,0 conditions

5 •

0 500

'

600

11

700

i

800

1

900

1

1000

1

1100

I Incongr. I Melting

1

1200

_

T TC]

Figure 17. Preliminary PT stability field of grandidierite, MgAl 3 BSi0 9 , after Heide (1992). Note the relationship of the fields delineated for hydrous as well as nearly anhydrous conditions to that of sillimanite as extrapolated from Holdaway (1971).

substitutions B 3+ for Si 4+ + = 0.5 O2" and Mg2+ for Al3+ + 0.5 O2" in the ratio 2:1 (see Fig. 16). It is interesting to note that it is exactly this mechanism which is also responsible for the introduction of trace amounts of B and Mg into sillimanite (see §3.6.1). Within this newly recognized structural family, werdingite, Mg2Ali4Si4B4037, was found to be composed of structural units of grandidierite alternating with others of the composition AlgB2Si20i9 in the ratio 1:1 (Niven et al., 1991). The latter composition was crystallized as a distinct phase as described in §3.6.1. Werding and Schreyer (1992) show that the substitutional relationship between AlgB2Si20j9 and both werdingite and grandidierite is Mg2+ for Al3+ + 0.5 O2" (see also Fig. 16). Indeed, werdingite was synthesized from AlgB2Si20i9 and grandidierite as the discrete 1:1 compound between these two phases (Werding and Schreyer, 1992). For the synthesis of the Mg endmember of werdingite from gel it was necessary to apply the least hydrous conditions possible (for details see Werding and Schreyer, 1992) at relatively high temperatures (800° to 875°C) and intermediate pressures. At lower pressure, phases exhibiting broad mullite-like X-ray diffraction lines were obtained that seem to be Mg-bearing "boron-mullites." Werding and Schreyer (1992) hypothesized that these phases could be disordered "mixed-layers" of grandidierite and Al8B2Si20i9 units. They may be metastable just like many of the "boron-mullites" in the ABSH-system (see §3.6.1), but much more work is necessary to clarify these relations. For the PT-stability of endmember werdingite two limiting cases must be distinguished: (1) Under nearly anhydrous conditions werdingite was found by Werding and Schreyer (1992) to break down at pressures above about 10 tol2 kbar and T > 800°C to dumortierite + sinhalite + corundum (see Fig. 18 and compare Figs. 14 and 15). (2) Under conditions with an excess hydrous fluid werdingite exhibits a lower-temperature stability limit, below which it breaks down into grandidierite + corundum + alkali-free dravite (see Fig. 19 and compare Fig. 15), an assemblage that contains less boron but is hydrous. At fluid pressures greater than about 9 kbar this lower stability limit of werdingite is probably intersected by the hydrous melting curve (Fig. 19). Combining these results from (1) and (2), it is clear that Mg-werdingite is also confined in its PT-stability to the field of sillimanite, its structural parent phase, similarly as it holds for grandidierite (§3.8.1) and probably also for the "boron-mullites" (§3.6.1).

142

Werding & Schreyer: Experimental

Temperature

Studies

(°C)

Figure 18. Preliminary PT stability of werdingite (heavy line) for nearly water-free conditions, compared to the Al 2 Si0 5 equilibrium diagram after Holdaway (1971, thin solid lines). Sillimanite/kyanite curve was extrapolated to higher temperatures (thin dashed line). Sillimanite/mullite relationships are neglected. For the nature of the high-pressure breakdown assemblage see text. [Used by permission of the editor of the European Journal of Mineralogy, from Werding and Schreyer (1992), Fig. 5, p. 204],

\

•

Werdingite

X

B r e a k d o w n of werdingite

. x x

\

' \

growth

(Corundum)

/ • - T o u r maline \ \ • Corundum /

i

> \

/

\ \

Crystals \ \

Melt

Grandidierite 'A \ • Corundum / * • ID-Tourmaline / ? •B,03 / * \

/

/

^

x < • \ w, Werdingite^ x ' , / qos

u. 600

800

1000

T e m p e r a t u r e (°C)

\ 1200

Figure 19. Preliminary PT stability of werdingite in the presence of excess hydrous fluid. For further discussion see text. [Used by permission of the editor of the European Journal of Mineralogy, from Werding and Schreyer (1992), Fig. 6, p. 205]

Werding & Schreyer: Experimental Studies

143

3.8.3. Kornerupine and prismatine (kornerupine s.l.). These minerals with the general formula (•,Mg)Mg 3 (Al,Mg) 6 (Si,Al,B) 5 (0,0H) 22 (Grew, this volume) are unique among the B-bearing phases of the MABSH-system (Fig. 15) in that they may also be boron-free. According to most recent mineralogical terminology (Grew, this volume), kornerupine sensu lato stands for the group name comprising kornerupine sensu stricto with boron contents less than 0.5 per formula unit and prismatine with 0.5 < B < 1.0 p.f.u. Schreyer and Seifert (1969) first synthesized boron-free kornerupine as a surprise phase in the MASH-system. Seifert (1975) showed experimentally that it becomes stable only at water pressures above about 4.5 kbar, and Wegge and Schreyer (1994) determined that its stability ends between about 12 and 13 kbar through breakdown into enstatite + corundum + H2O (Fig. 20).

750

800

Temperature (°C)

850

Figure 20. Water-pressure temperature plot outlining the stability field of boronfiee kornerupine. Arrows labelled H 2 0 indicate the requisite directions of dehydration. Abbreviations: Chi = chlorite Cord = cordierite C = corundum En = Enstatite Pyr = pyrope Sa = sapphirine Tc = talc. [Used by permission of the editor of the European Journal of Mineralogy, from Wegge and Schreyer (1994) Fig. 7, p. 73],

While only two localities of boron-free kornerupine are known from nature (Schreyer and Abraham, 1976; Droop, 1989), most natural kornerupines sensu lato contain boron in variable amounts (for details see Grew, this volume). Therefore, the compositional and stability relations of boron-bearing kornerupines are of predominant interest in this context. The composition of Mg endmember boron-free kornerupine was found by Wegge and Schreyer (1994) to be (Mg 3 . 72 Al 5 . 21 )[6](Al 1 . 2 8Si3. 72 )[4]0 19J6 (OH) 2 . 2 4. This projects exactly between the two ideal anhydrous oxide ratios 1:1:1 and 4:3:4 (= MgOAl 2 03'Si0 2 ), that were proposed for kornerupine (Schreyer and Seifert, 1969). On the other hand, Werding and Schreyer (1978), in a detailed experimental study on the crystal chemistry of boron-bearing kornerupines at the fixed condition 830°C and 7 kbar fluid pressure, showed that, with increasing boron introduction, the above oxide ratio is increasingly shifted towards 4:3:4. Figure 21 summarizes the analytical findings, that might be best explained by a substitution B for Al. However, this is rather unlikely in view that boron largely replaces silicon in the T(3) site of the kornerupine structure (Grew and Klaska, 1991). Moreover, it was found by Werding and Schreyer (1978, Fig. 8) that increasing amounts of boron in the synthetic kornerupines were also linked with decreasing

144

Werding & Schreyer: Experimental Studies

Figure 21.. Section through the quaternary system Mg0-Al 2 0 3 -Si0 2 -B 2 03 along the enstatite-corundumB 2 0 3 join (see inset) exhibiting approximated solid solution limits of B-bearing komerupine. Proportions 4:3:4 and 1:1:1 give the oxide ratios Mg0:Al 2 0 3 :Si0 2 . The shaded area represents the possible ternary range of komerupine solid solution synthesized at 7 kbar, 830°C, as based on the analytical findings; see text as well as Werding and Schreyer (1978, modified Fig. 10).

OH-contents, crudely according to B 3 + for 3H + . Therefore, it seemed more likely that the Al-decrease in B-bearing kornerupines is due to the inverse Tschermak-substitution Mg + Si for 2 Al, and that B introduction into the Si-site initiates a chain mechanism, whereby the released Si replaces Al in a neighboring tetrahedron, Al in turn moves to an octahedral site where it replaces Mg, and finally Mg is accommodated in the largely vacant X-site: I^B MSi -> MAI -> IfflMg -> • (Werding and Schreyer, 1978). More recent data, however, show that the compositional variations must be still more complicated. Boron-bearing kornerupines s.l. are not always restricted to the plane enstatite-corundum-B203 as shown in Figure 21, and the relation B = 3H is modified for other conditions of komerupine synthesis. Details derived from the work by Krosse (1995) will be given below. A petrologically important question for kornerupine-bearing rocks was also addressed in the study of Werding and Schreyer (1978), who determined quantitatively the fractionation of boron and water between komerupine and the coexisting hydrous fluid. They found that B2O3 is enriched over H2O in the solid by a factor of about 10 as compared with the gas phase. The Ko-value defined as (B2O3/H2O) in komerupine over (B2O3/H2O) in the fluid was determined as 18.7 and indicates that the mineral komerupine is indeed a sink for boron in fluid/rock systems of appropriate bulk compositions. On the other hand, kornerupines or prismatines can only accept a limited amount of boron in their structure (see later). This explains why Werding and Schreyer (1978) found additional, more boron-rich phases like grandidierite and sinhalite to coexist with komerupine when they increased the B2O3/H2O in their runs at constant PT-conditions. A point of major concern is the question as to whether or not boron-incorporation has any influence on the PT-stability of kornerupines. Whereas no data are available thus far on the lower pressure limit, Krosse (1995) made the exciting discovery that the upper pressure limit of boron-rich kornerupines (prismatines) is extended from some 12 kbar for the boron-free phase (Fig. 20) to about 38 kbar. A preliminary stability diagram of B-rich prismatine was worked out by Krosse (1995) using the seeding technique and a fixed bulk composition 3.67MgO-3.02Al 2 03-3.74Si02-0.43B 2 03-1.3H20. A prismatine of this

Werding & Schreyer: Experimental Studies

145

composition (except for H2O) had previously been obtained as a breakdown product of dravite (see §3.9.1). The X B 2 0 3 of this bulk composition (0.25) is that of H3BO3 which was used as the boron source. The results are shown in Figure 22. Note that in the pressure range 10 to 30 kbar prismatine is stable at temperatures as much as 100°C lower than in the B-free system for lower fluid pressures (640°C at 21 kbar versus 740°C at 7 kbar; compare Fig. 20 and Seifert, 1975). Thus the stability field of boron-bearing kornerupines s.l. clearly expands with increasing fluid pressure. The various breakdown assemblages shown in Figure 22 are those obtained initially by Krosse (1995) in synthesis runs and, although persistent in subsequent bracketing runs seeded with kornerupine, may not represent the true state of equilibrium. It should also be noted that the sequence of breakdown assemblages shown in Figure 22 is not even in agreement with a theoretical Schreinemakers' analysis simplified for a ternary system with excess fluid (Krosse, 1995) which requires several more intermediate breakdown assemblages, e.g. tourmaline + chlorite + "pseudosinhalite" for the range 30 to 35 kbar, 750°C. These reactions have not been studied. The same holds for the phase relations at fluid pressures below 10 kbar and the possible links with the stability field of boron-free kornerupine (Fig. 20). Importantly, at pressures between 20 and 30 kbar, prismatine remains stable up to the surprisingly high temperature of 1050°C. P [kbar]

py+stau+pss+v

40 35 30 ..

Tu+Chl+Stau

25 20 -

•

15 10

Tu+Chl+Co

5

Q

•

Tu+En+Spin

500

600

700

800 TfC]

900

1000

1100

Figure 22. Preliminary PT stability field for boron-rich prismatine containing about 0.9 Batoms per formula unit based on reversal runs after Krosse (1995). Solid symbols indicate growth, open ones breakdown into the assemblages indicated. Abbreviations: Chi = Mg-chlorite Co = corundum, En = enstatite, Ky = kyanite, PSS = "pseudosinhalite," Py = pyrope, Sap = sapphirine, Spin = spinel, Stau = staurolite, Tu = alkali-free dravite, V = fluid. For compositions of these phases see Figure 15.

It must be emphasized that the data presented in Figure 22 can only be considered a first approach to an enormously complex problem in a multisystem with 12 solid phases and 5 components. All the reaction boundaries of Figure 22 mark at least divariant equilibria. For other bulk compositions than the one chosen here with Xg 2 03 - 0-25 these boundaries may vary. Note that Xg203 - 0.25 does not apply to the fluid, as B2O3 is strongly fractionated into primatine (Werding and Schreyer, 1978). The microprobe analyses of Krosse's (1995) prismatines (see below) confirm that practically all the boron in the bulk composition used is accommodated in this phase. Thus, the coexisting fluid must be practically pure H 2 0 , just as in many cases of the jeremejevite assemblages (see §3.5); B2O3 is again an inert component. These considerations lead to the conclusion that the PT-field outlined in Figure 22 may be close to the maximum stability field of B-rich prismatine. It is interesting to speculate about the reasons behind the spectacular high-pressure stabilization of kornerupine in the presence of boron. One reason might be that boron cannot be incorporated by any other crystalline phase, but inspection of Figure 22 shows

146

Werding & Schreyer: Experimental Studies

that below some 900°C this is not the case. Both alkali-free dravite (see §3.8.5) and "pseudosinhalite" (see §3.7.3) are among the breakdown products of boron-rich prismatine. At higher temperatures there are no longer any B-bearing solid breakdown phases, so that boron must enter the fluid or melt. Therefore, it seems that the persistence of the phase kornerupine to high pressures is solely due to the incorporation of boron into its crystal structure. Apparently, it is the boron replacing Si in the tetrahedral T(3) site (Klaska and Grew, 1991) that stabilizes the kornerupine structure as a whole. The chemistry of the synthetic high-pressure B-rich prismatine obtained by Krosse (1995) is characterized by boron contents equalling or even exceeding the highest values known for this mineral species in nature and experiment thus far (Grew et al., 1990 and Grew, this volume). A typical mean composition as obtained by microprobe analyses and H20-determination on the bulk sample of single-phase kornerupine synthesized at 30 kbar, 942°C, for 69 hours is M

g3.92A15.92B0.90Si3.62O20.77(OH)l.23-

However, there are also individual microprobe analyses in this sample such as M

g3.82A15.93B103Si3.56O20.77(OH)l.23'

where the T(3)-site is apparently fully occupied by boron. In either case, the Mg/Si-ratio is >1. so that these kornerupines do no longer lie along the join enstatite-A^C^^Og as implied by Figure 21. The relationship found by Werding and Schreyer (1978) of decreasing hydroxyl with increasing boron for 7 kbar, 830°C holds for the high-pressure prismatines of Krosse (1995) as well, but the array is shifted for a given amount of hydrogen towards higher boron contents (Fig. 23). This is obviously due to the vastly different conditions of synthesis applied in the Krosse experiments (25 to 37 kbar, 700° to 950°C), under which boron-free kornerupine is no longer stable (compare Fig. 20). Extrapolation of the dashed line of Figure 23 towards lower B-contents would finally lead to a boron-free kornerupine with impossibly high water contents. Future high-pressure experiments on kornerupines with lower boron-contents ( 50 kbar). Solid symbols = growth of dravite; open symbols = breakdown of dravite. For the change of the high-temperature assemblage from prismatine to pyrope compare Figure 22. The runs at 80 and 100 kbar showed breakdown of dravite seeds into the assemblages indicated. Hie breakdown phases are shown in Figure 15 except for "topaz-OH" (see Fig. 6) and MgMgAlpumpellyite (see Schreyer 1988).

Pyrope +" Topaz-OH" + unknown phase

90

MgMgAI- Pumpellyite + Kyanite

80 70

60 ¥

50

r

40 +

.o

30

» O

• •

«o

Dravite

4

• • •

20

•

10 +

•Or

yO

Pyrope +Corundum + melt

k> Prismatine + Corundum + mel'

Robbins ^ Yoder (1^62)

600

700

800 900 TfC]

1000

1100

More than 30 years later Krosse (1995) extended the stability investigations on dravite composition towards high and ultrahigh pressures, because dravite had been found to occur as inclusions within pyrope of the coesite-bearing ultrahigh-pressure rocks of the Western Alps (Schreyer, 1985). Her results on the upper thermal stability of dravite up to a fluid pressure of 50 kbar are shown in Figure 25. The data link up rather smoothly with those of Robbins and Yoder (1962), and—like in the earlier work—the decomposition products contain only phases of the MABSH system. Thus, the entire sodium component of dravite enters the phases liquid + gas or, at high pressures, perhaps a single-phase supercritical fluid. Compared with the 5 kbar result of Robbins and Yoder (1962) corundum rather than sapphirine was found by Krosse (1995) to coexist with kornerupine over the pressure range 15 to 25 kbar. Between 30 and 50 kbar kornerupine is replaced by pyrope so that at these high pressures and temperatures boron, like sodium, is completely fractionated into melt or fluid as well. It is important to note, however, that dravite containing considerable amounts of the volatile elements H, B, and Na is stable to temperatures near 1000°C. Concerning an upper pressure limit of dravite, three experiments were performed using the multi-anvil press at Bayreuth at 900°C and pressures of 60, 80, and 100 kbar. Whereas dravite seeds remained unchanged at 60 kbar, they were found to break down at 80 and 100 kbar, and the typical high-pressure phases "MgMgAl-pumpellyite" (Schreyer, 1988) and "topaz-OH" (Wunder et al., 1993a) formed instead plus at least one additional unknown phase. This indicates that dravite breakdown occurs between 60 and 80 kbar,

Werding & Schreyer: Experimental Studies

151

which for a crystal structure with boron in triangular coordination is a surprisingly high pressure stability limit. Electron microprobe analyses of the synthetic dravites prepared by Krosse (1995) from stoichiometric mixtures show that they are practically never of the ideal composition. The boron values obtained by microprobe vary between about 9.77 and 12.08 wt %. Compared to the ideal value of 10.89 % for ideal dravite this variation is considered to be within the analytical error. Therefore, in the subsequent recalculation procedure, boron was taken to have the ideal value of 3.0 p.f.u. in all cases. Similarly, water determinations on four batches of single-phase dravite gave a mean value of 4.39 wt % H2O. Because this is fairly close to the ideal value of 3.76 wt % for dravite, ideal 4 (OH) p.f.u. were assumed. Thus, typical dravite formulae recalculated from the microprobe analyses are: Na0J5(Mg2.62Al0.38)Al6[A]0.i0Si5.90Oi8](BO3)3(OH)4 synthesized at 15 kbar, 818°C; Na 0 . 82 (Mg 25 4Al 0 4 6 )Al5 95 [Al0. 11 Si5. 89 O 18 ](BO 3 )3(OH)4 synthesized at 50 kbar, 952°C; Na

0.90( M g2.79 A 1 0.2l) A 1 5.97[ A 1 0.03Si5.97Ol8](BO 3 ) 3 (OH) 4

synthesized at 40 kbar, 807°C. Perhaps most notably, the X-site contains only 75 to 90% Na and is never completely occupied as is also common for many natural dravites (see Henry and Dutrow, this volume). Statistically, the highest Na occupancies were measured when a starting material with excess Na (Naj 2 rather than Naj q) was used. In addition, increasing pressures seem to have a favourable influence on Na incorporation. The sodium deficiencies are always connected with deficiencies in Mg and excess amounts of A1 p.f.u. ideal Dravite

4.00 3.90 3.80 3.70

a.

01

3.60 3.50 3.40 3.30 + 3.20 3.10

Na05(Mg25AI05)AI6[Si6Olal(BO3)3(OH)4

3.00 5.9

Figure 26. Plot of microprobe analytical data of synthetic dravites obtained in high-pressure experiments by Krosse (1995). The solid line connecting ideal dravite, NaMg3Al6[Si60lg](B03)3(0H)4, with the composition given on the lower right end follows the substitution (Q+Al) for (Na+Mg) and would finally lead to the formula of "alkali-free dravite" (see Fig. 15), •(Mg 2 Al)Al 6 [Si 6 0 18 ](B0 3 ) 3 (0H) 4 . The data points were corrected for the influence of the relationship shown in Figure 27.

Figure 26 shows that the Na deficiency is mainly due to the substitution scheme A1 for Na+Mg. Note that the Mg- and Al-values were corrected for the effect of the Tschermak

152

Werding & Schreyer: Experimental Studies

substitution [41A1+MA1 for Si+Mg. In Figure 27 the sum of Mg+Si is plotted versus AW, this time correcting Mg and A1 for the effect of the A1 = (Na+Mg) substitution. Again a linear correlation is indicated around the ideal Tschermak substitution line. There is a weak but significant correlation of the Tschermak substitution with pressure of synthesis: ^ A l decreases with increasing pressure (Krosse, 1995).

Si + Mg p.f.u. Figure 27. Plot of microprobe analytical data of synthetic dravites obtained in high-pressure experiments by Krosse (1995). The solid line follows the substitution as indicated from ideal dravite to the hypothetical Tschermak's member of the dravite series. The data points shown were corrected for the influence of the relationship shown in Figure 26.

The substitution of hydroxyl by fluorine in dravite was studied by Gourdant et al. (1994) at 2 kbar fluid pressure and 500°C. Single-phase tourmaline was only obtained up to XF = F/(F+OH) = 0.25. Higher XF in the starting composition yielded the additional phases albite, quartz, and sellaite (MgF2), while dravite maintained its constant composition with XF = 0.25. This is in agreement with the findings on natural tourmalines (Deer et al., 1986) that only one out of four hydroxyl groups is replaced by fluorine. A lower temperature limit to dravite stability has not been investigated as yet. Vorbach (1989) reports syntheses as low as 300°C, 1 to 4 kbar, but suspects that dravite may also form at still lower temperatures. 3.9.2. Uvite. No experimental work has been performed on uvite, the most important Ca-bearing endmember of the tourmalines with the ideal formula CaMg3(MgAl5)(B0 3 )3[Si 6 0 18 ](0H)4. It is related to dravite by the substitution CaMg for NaAl just as diopside is to jadeite in the pyroxene series. Uvite may actually be more common in nature than dravite, and analyses of natural tourmalines demonstrate that there is a complete series of solid solutions between the two endmembers. The only experimental knowledge on a Ca-bearing tourmaline is that from a single run at 20 kbar, 870°C, reported by Henry and Dutrow (1990) on a bulk composition similar to the ideal uvite endmember. These authors had derived, based on extrapolations of a natural CaNa-tourmaline occurrence, a hypothetical formula of a new Ca-endmember tourmaline,

Werding & Schreyer: Experimental Studies

153

CaMg 3 Al 6 (B03)3[Si 6 0 18 ]0(0H)3 ) which they call "deprotonated" relative to the normal hydroxyl content (4.0) of tourmaline. Since single-phase tourmaline was obtained in their experiment, Henry and Dutrow (1990) assume the above formula to apply to their product. If this is true, the new Ca-tourmaline endmember is related to dravite by the substitution Ca for NaH, and to uvite by A1 for MgH. The problem with this experimental result is that the composition of the reaction product was neither analyzed for its major element chemistry by microprobe nor was a water determination performed on the bulk sample to prove deprotonation. More experimental work on Ca-bearing tourmaline compositions is overdue. 3.9.3. Elbaite. This Li-bearing tourmaline is generally given the ideal formula Na(Al,Li) 3 A] 6 (B0 3 )3[Si 6 0 18 ](0H)4, where for stoichiometric reasons the Y-position of the structure must be occupied by equal amounts of A1 and Li. Vorbach (1989) has performed synthesis experiments on elbaite using starting materials with theoretical Y-occupancies of LiAl2, Li2Al, and Li 3 . Elbaite was obtained from all three starting materials, but often together with other phases such as eucryptite, quartz, albite, and a cookeite-like phase. No conclusions were drawn concerning the true composition of elbaite, especially as considerable amounts of Na and Li were found to be present in the coexisting gas phase. Synthesis conditions ranged from 300° to 700°C in the pressure range 1 to 4 kbar. At 750°C elbaite melted incongruently to corundum + liquid. Apparently, the lower temperature limit of elbaite growth increases with the amount of excess boron present in the runs. Further investigations are necessary in this system to resolve the many questions pending. 3.9.4. Highly aluminous tourmalines, olenite. Ever since Donnay et al. (1966) discovered the new tourmaline species buergerite, NaFe 3 3*Al 6 (B03)3[Si 6 0 18 ]0 3 F, there have been speculations that ferric iron in this phase could be replaced by additional Al so that another endmember, "aluminobuergerite," could be obtained. Indeed, Sokolov et al. (1986) described a natural tourmaline close to the endmember and named it olenite. It has the idealized formula Na 1 . x Al 3 Al 9 B 3 Si 6 027(0,0H)4. The synthesis of tourmaline phases in the system Na20-Al20 3 -B203-Si02-H20 first reported by Voskresenskaya and Barsukova (1968) and later confirmed by Rosenberg et al. (1986) is thus of direct relevance. Unfortunately, the latter results are not easy to interpret. Considerable trial and error, using different starting compositions and long run durations, was necessary to obtain only small yields of tourmaline, which were then difficult to characterize chemically. Synthesis conditions were 500° and 550°C, 1 kbar for about 2 months. Analytical data obtained on the minute crystals by both electron microprobe and electron microscope for Na, Al and Si led to the following two structural formulae: (Nao.9400 06>A13 Al 6 (Al! .78Si4.22)(B03)3019.2(0H)2.8 (Nao.76ao.24)Al3Al6(Alo.18Si5.82)(B03)302o.6(OH)1.4. The number of protons given is calculated in order to attain charge balance. Thus there is again no direct analytical evidence for proton-deficiency of tourmaline synthesized under excess water conditions. Rosenberg et al. (1986) consider their synthetic tourmalines as proton-enriched and possibly metastable relative to ideal "aluminobuergerite," now olenite endmember. The problem of the existence of highly aluminous tourmalines has to be seen in a new

Werding & Schreyer: Experimental Studies

154

perspective after the recent synthesis of a Na-free tourmaline-phase in the system ABSH (§3.6.3 and Fig. 6). If this phase turns out to be stable, further investigations on tourmaline miscibility in the system Na20-Al203-B203-Si02-H20, particularly at high pressures, will be quite important. 3.9.5. Tourmalines with iron and other transition elements. It is one of the strange facts in the experimental mineralogy of tourmaline that the most common members in nature, the iron-rich species with the ideal endmember schorl, NaFe2+ 3 Al 6 (B0 3 ) 3 [Si 6 0 1 8](0H)4, are the ones least studied under laboratory conditions. We suspect that the additional experimental task of controlling the oxygen fugacity is the main reason for this abstinence. Nevertheless, Taylor and Tenell (1967) emphasize that NaFe-tourmalines can be readily synthesized hydrothermally between 400° and 600°C, and Tomisaka (1968) reports synthesis at 650° to 700°C, 0.7 to 2 kbar. An experimental investigation of endmember schorl under different conditions of oxygen buffering would be highly desirable. We suspect that with increasing J02 schorl will be oxidized and may finally form the endmember "hydroxy-buergerite" with a formula NaFe 3 3+Al 6 (B03)3[Si 6 0 18 ]0 3 (0H), which could also be named "oxy-schorl". This would be analogous to the case of oxidation and dehydrogenation of the Fe mica annite, KFe32+[AlSÌ30io](OH)2, to oxy-annite, KFe 2+ Fe 3+ 2[AlSi 3 Oio]02, as it has been discussed by Wones (1963), or, of hornblendes to oxy-hornblendes (Clowe et al., 1988). The significant aspect of these considerations is that here is a mechanism for hydroxyl- or proton-deficiencies in tourmaline that is solely linked to the oxidation state of iron and that may occur extensively in nature. On the other hand, as pointed out in earlier sections, there is still no analytical evidence that iron-free synthetic tourmalines do indeed exhibit proton-deficiencies or deprotonation (§3.8.5; §3.9.2; §3.9.4) relative to the ideal 4.0 hydrogens per formula unit. Although perhaps not very important for natural tourmalines, reference is made here again to the study of Taylor and Terrell (1967), who were able to synthesize hydrothermally a great variety of tourmalines containing the transition elements Mn, Co, Ni, Cu, Zn as divalent cations and V, Cr as trivalent ones. Similar syntheses are reported by Voskresenskaya et al. (1975). 3.10.

Serendibite

This rare mineral occurring in Ca-bearing, silica-undersaturated skarns has the approximate formula Ca2(Mg,Fe2+)3(Al,Fe3+)45B15SÌ3023 and belongs to the aenigmatite group (Grew et al., 1991). It contains boron in tetrahedral coordination in a crystal structure similar to that of sapphirine (Van Derveer et al., 1993). The only experimental work performed thus far is the very preliminary study by Werding et al. (1990) on the iron-free serendibite endmember in the system CaO-Mg(> AI2O3-B2O3-SÌO2-H2O. Primarily synthesis runs without seeds were performed, and serendibite could be obtained with high yields at high temperature (830°C) and fluid pressures between 10 and 20 kbar. In a few seeded runs at 500°C serendibite was found to grow at 10 kbar, but to break down at 20 kbar. This and other results at 20 and 30 kbar (Fig. 28) seem to indicate that there is an upper pressure limit to serendibite stability, similarly as in the case of the related phase sapphirine (Ackermand et al., 1975).

Werding & Schreyer: Experimental Studies

155

Concerning serendibite stability at pressures below 10 kbar, no clear evidence can be derived from the synthesis runs. In the range between about 2 and 5 kbar serendibite could only be grown at relatively high temperatures (~600°C), but failure to grow does not mean instability. The PT-conditions estimated for the nine serendibite localities known in the world (Grew et al., 1991) all fall within the empirical growth field of serendibite as outlined by Werding et al. (1990) and reproduced here as Figure 28. Phases found to coexist with serendibite in these experiments are spinel, sinhalite, a diopsidic pyroxene, and a tourmaline that may be close to or identical with uvite (see §3.9.2).

< ¿00

500

600

700

800

900 T(°C)

Figure 28. PT-plot showing conditions of growth (solid dots) of the MgAl end member of serendibite, mainly in synthesis runs. Open circles indicate the formation of other phases; at 500°C, 20 kbar serendibite breakdown was achieved in a seeded run. The field outlined by dashed lines is not a stability field. For further comments see text. Redrawn after Werding et al. (1990)

3.11. Axinite Axinites form a chemically and structurally complex mineral group with the general formula H(Ca,Fe,Mn,Mg) 3 [Al 2 BSi 4 0 16 ] as given by Deer et al. (1986); for details see Grew (this volume). The only relevant experimental study seems to be that of Nekrasov and Kashirtseva (1975), who synthesized axinites hydrothermally in the system Ca0-Mn0(Fe0)-Al 2 03-Si02-B203-H20, but never as single-phase products. Synthesis conditions were 0.3 to 1 kbar between 300° and 500°C. The upper temperature stability limit was found to be near 525°C, the lower limit was expected to lie below 200°C. Phases coexisting with axinite were anorthite and danburite (§3.3). At high B2O3 concentrations axinite was replaced by the assemblage danburite and anorthite. Addition of tin as SnO to the starting mixtures led to incorporation of less than 1 % tin into the axinite structure. As suggested by Mossbauer spectroscopy, Sn 2+ is assumed to replace Si. Conversely, Sonnet and Verkaeren (1989) applying the same technique on natural axinite report Sn replacing (Al,Fe +). More experimental work at higher pressures would be of interest to check the stability of axinite with its boron in tetrahedral coordination.

3.12. Boron micas Unfortunately, very litde is known thus far on the incorporation of boron into micas and other phyllosilicates, on possible stabilities and on boron fractionation between such solids and the coexisting fluids. In natural micas boron analyses are usually not performed. If so, as much as 2000 ppm may be detected (Harder, 1959). Nevertheless, micas in which all tetrahedral Al is replaced by boron have been synthesized long ago: Boron phlogopite, KMg3[BSi3O10](OH)2 was grown by Eugster and Wright (1960) as well as by Stubican and Roy (1962). Synthesis conditions range from 1 to 2 kbar, 250° to 800°C. Boron muscovite,

156

Werding & Schreyer: Experimental Studies

KAl2[BSi3O10](OH)2, was reported by Stubican and Roy (1962) to form at 2 kbar, 500°C. This is of interest, as the new mineral boromuscovite with 0.77 B p.f.u. was recently discovered by Foord et al. (1991) in a late-stage portion of a pegmatite. In a current study at Bochum by Jung (pers. comm.) stoichiometric boromuscovite was readily synthesized at high pressures (30 kbar, 700°C), but was found to decompose when rerun at low pressures (2 kbar). This seems to indicate that endmember boromuscovite is a high-pressure phase. Jung (pers. comm.) has also synthesized boron-bearing phlogopite both at low and high pressures. Here a compositional problem arose as the synthetic products showed variable physical constants that are probably due to an at least ternary solid solution range between normal phlogopite, KMgjfAlSijOioKOH^, boron phlogopite, KMg3[BSi30jo](OH)2, and Al-free phlogopites such as KMg2 5[Si40jo](OH)2. Phlogopite solid solubility in the Al- and B-free system as well as between normal and Alfree phlogopites was studied by Seifert and Schreyer (1971). Normal phlogopite is an important stable phase in the upper mantle, and muscovites (or phengites) are common constituents of deeply subducted rocks of the continental crust (Chopin, 1984). Like the boron-feldspar reedmergnerite, Na[BSi30g], and the framework silicate danburite, Ca[B2Si20g], it is likely that the boron micas remain stable up to rather high pressures. Thus, they may represent effective sinks for boron in these high- to ultrahigh-pressure environments. However, the crucial question is whether boron at these high pressures is preferentially enriched in micas or in fluids. Experimental determination of boron distribution coefficients as a function of pressure and temperature are required to resolve this question. It may well be that B2O3 is an inert component, as in the case of kornerupine (§3.8.3). 3.13.

Ludwigite-vonsenite

Ludwigite, Mg2Fe3+02[BC>3] and vonsenite, Fe 2 Fe 0 2 [B0 3 ], form an lsomorphous series and occur predominantly in contact skarns. The first successful hydrothermal syntheses were carried out by Grigoriev and Nekrasov (1963) at about 0.85 kbar and temperatures between 500° and 680°C. They obtained small amounts of a probably nearly pure ludwigite from a mixture of magnetite, dolomite and borax; ludwigites with about 20% vonsenite endmember were synthesized from oxide mixtures containing both Fe and Mg; and endmember vonsenite grew from a Mg-free starting material. Subsequently, Kravchuk et al. (1966) synthesized other members of the solid solution series with 25, 50, 75 and 100 mol % vonsenite. With increasing Fe2+ contents the phase formed at successively lower temperatures, that is at 650°, 450°, 350°, and 200°C for the respective compositions listed above. Barrese et al. (1984) obtained pure vonsenite in the pressure range 10"2 to 1000 bars, 300° to 500°C. In all these studies the oxygen fugacity was not controlled, but the presence of di- and divalent iron was insured by using magnetite or hematite plus metallic iron in the starting materials. The work by Xie et al. (1985) shows that ludwigite and vonsenite may also be stable at high pressures. They obtained the pure endmembers at 30 kbar and 800°C. Grigoriev (1968) reports the synthesis of a borate phase with a hulsite structure but with the composition of vonsenite. Hulsite and magnesiohulsite from natural rocks contain additional tin. On the other hand, an unknown phase in the system Mg0-Al 2 03-B 2 03 (-H 2 0) (see Fig. 9) may represent an Al analogue of magnesiohulsite (§3.7.2). This might indicate that the monoclinic hulsites do not require tin as an essential element, and that they

Werding & Schleyer: Experimental Studies

157

may form a considerable number of endmembers not known as minerals thus far. They might even appear as polymorphs of the minerals of the ludwigite-vonsenite series. A cobalt analogue of vonsenite with the formula Co22+Co3+02[BC>3] was synthesized by Goetz and Herrmann (1966) at atmospheric pressure and 1000°C. 3.14. Boron in framework silicates The introduction of four-coordinated boron into tetrahedral silicate framework structures is well established. Prominent examples are the feldspar structure of reedmergnerite, NafBSijOg], and the paracelsian structure of danburite, Ca[B2Si20g] (§3.2 and §3.3), which occur as natural minerals. KBSi33 and K2O-B2O3 at 985°C, PbO-B2C>3 at 800°C. o.o

0.2

0.4

0.6

0.8

1.0

Mole Fraction Boron Oxide

Figure 3. Thermodynamic activity of B2O3 in several M2O-B2O3 and M O - B 2 O 3 melts. Modified from Itoh et al. (1976), reproduced with permission.

0.0 '0.6

0.7

0.8

0.9

1.0

Mole Fraction of B 2 0 3

Clearly these deviations from ideal, or even simple, thermodynamic behavior, characteristic in general for borate, silicate, and phosphate melts, reflect changes in structure and speciation. Itoh et al. (1984) have shown that the major characteristics of the activity curve for the Na20-B203 system can be modeled by considering the boroxol group (a ring of three BO3 triangles sharing common oxygens, which gives the stoichiometry B3O4.5) to be a polybasic acid, with two major neutralization steps involving oxide ions produced by adding Na20: ( 8 / 3 ) B 3 0 4 5 + 0 2 ~ = B 8^13' 8 0 ? 3 , p K = 15.1 (boroxol)

(tetraborate)

Navrotsky: Thermochemistry of Borosilicate Melts and Glasses

170 followed by

Bg0i3 + 0 2 - = 2 B 4 0 ? (tetraborate)

pK = 11.3

(diborate)

c

o

F i g u r e 4. Variation of fraction of boroxol, tetraborate, and diborate species in N a 2 0 B 2 O 3 melts as a function of ratio of N a 2 0 to B 2 O 3 . Solid curves indicate calculated values from values of equilibrium constants deduced from measured activities in melts. Dashed curves indicate speciation deduced from boron N M R studies on glasses. Dot dashed lines indicate maximum (quantitative) conversion that would occur if equilibria were to lie far to the right (K = Reproduced with permission from Itoh et al.

(1984).

The speciation predicted by this model (see Fig. 4) generally falls between that predicted by quantitative conversion (K = equilibria lie to the right) and that inferred from NMR data on quenched glasses. The values of AE(B 2 0 3 ) obtained from the temperature dependence of e.m.f. data (Itoh et al., 1984) are in generally good agreement with those from direct calorimetry (Ostvold and Kleppa, 1970). Enthalpies of mixing in Pb0-B203 and Ca0-B203 melts, based on calorimetric data (Holm and Kleppa, 1967; Klein and Müller, 1987), but corrected to liquid oxide standard states, are shown in Figure 5. The minimum is both deeper and sharper for CaO than for PbO, consistent with greater basicity and more complete transfer of oxide ions in speciation reactions for the former system. Macedo and Simmons (1974) proposed a thermodynamic model which was fairly successful at symmetrizing the strongly asymmetric metastable subliquidus miscibility gaps in the alkali borates and in lead borate. They redefined components as a polymer of boron oxide, (B203) m and the metal borate phase nearest the metal oxide-rich limb of the miscibility gap, M2O • nB2C>3 (see Table 4). Redefining new mole fractions in terms of these components, they applied a model consisting of a regular solution enthalpy, Wh and excess (vibrational) entropy, W$, such that, with X = mole fraction of (B203) m , Y = mole fraction M2O • nB2C>3, the free energy of mixing is given by 4 G m i x = XY(W H - TW S ) - RT[XAiX + Y&iY]

Navrotsky: Thermochemistry of Borosilicate Melts and Glasses

171

Figure 5. Enthalpies of mixing in melts in the system PbO-B203 at 800 °C and CaOB2O3 at 1450°C. The values are calculated using the calorimetric data of Holm and Kleppa (1967) and Klein and Miiller (1987), respectively, with corrections for the heat of fusion of PbO taken to be 25.5 kJ/mol at 800°C and of CaO estimated to be 60 kJ/mol at1450°C.

Mole Fraction Boron Oxide

Table 4. Parameters in thermodynamic model of Macedo and Simmons (1979) for miscibility gaps in borate systems Cation

Metal borate component

Li Na K Rb Cs Pb

L i 2 0 • 4B2O3 K 2 0 • 3B203 Rb20 • 3B203 Cs20 • 3B203 PbO • 4 B 2 0 3

Critical temperature (°Q 660 590 590 580 580 775

Mixing parameters W H (kJ/mol) W s (J/mol«K) 15.5 0 model appears inapplicable 18.0 +4.16 10.6 -4.16 8.5 -6.65 105.0 +83.14

They found, for lithium, potassium, cesium, rubidium, and lead borates, a good fit to the experimental data with m = 5. This redefinition of components decreases the configurational entropy of mixing, and does so more severely as B2C>3-content increases, relative to a regular solution model with M2O and B2O3 as components. The excess entropy parameter, Ws, further adjusts the width of the miscibility gap. The obvious question is whether the ability to fit several borate systems with the same (6203)5 polymer implies any physical reality or structural identity to such a speciation model. This question has not been addressed in the two decades since this paper was published. Intriguingly, a similar approach has been successfully applied to the sodium, lithium, and barium silicates (Haller et al., 1974). The B203-S102 system The phase diagram for the B203-SiC>2 system is discussed by Dingwell et al. (this volume). Pichavant (1978) suggested that a positive regular solution interaction parameter of 13.8 kJ/mol in the liquid can fit the sigmoid liquidus and produce a metastable subliquidus glass-glass immiscibility dome which closes at 560°C, in good agreement with experiments of Haller et al. (1970), (see Fig. 6). Calorimetric determination of heats of solution in molten 2PbOB24 (Wight and Chao, 1995) has not been considered. Even in the binaries some doubts still exist concerning polymorphic transitions, glass transition temperatures, and a possible primary phase field for Na2CMB203. Metastable immiscibility in ternary sodium borosilicate glasses has been studied (Charles and Wagstaff, 1968; Charles and Turkalo, 1969; Haller et al., 1970). Immiscibility along all three binary joins in the low sodium oxide region expands rapidly with decreasing temperature. These regions may join below 700°C into a three-phase triangle defining three immiscible glasses (see Fig. 9c). Enthalpies of solution in molten 2PbOB2C>3 at 701 °C of amorphous materials (liquids or glasses, depending on composition, since Tg can be above or below calorimetric temperature) were measured by Hervig and Navrotsky (1985). They suggest that a simple Kohler-type polynomial formalism, with essentially zero ternary excess enthalpies of mixing, adequately describes the data, namely (with x = mole fraction Na20, y = mole fraction B2O3, z = mole fraction Si02):

s

2°3

AHmix = -559.6 xy + 13.0 yz + (-267.5 - 384.9 x(x+y)-!)xz

Clearly the major exothermic enthalpy contributions in the ternary stem from acid base reactions (speciation changes) along the Na20-B203 and Na20-SiC>2 binary joins with small endothermic contributions from the B2C>3-Si02 join. Such a balance of interactions is entirely consistent with a tendency toward clustering into borate-rich and silicate-rich domains and metastable glass-glass immiscibility.

Borosilicate glass for commercial use and for radioactive waste containment Table 5 shows some typical compositions of some commercial glasses. These typically have softening points in the 400°-500°C range and liquidus temperatures below 1000°C. Although food container and window glass tend to be boron-free or have low boron contents (to avoid leaching), other commercial glasses are borosilicates. Vycor is an almost pure silica containing some B2O3. Pyrex, glass fibers for construction, and optical glass represent complex borosilicates (see Table 5). Borosilicate glass is the presently accepted storage medium for commercial nuclear waste in the U.S.A.; the plan is to melt the waste with appropriate glass components, cool to a glass or glass/crystal composite, and store these logs encapsulated in steel cylinders in a repository (Krauskopf, 1988; Lutze, 1988; Bourcier, 1994). The thermochemical stability of borosilicate glasses is relevant to their use for longterm storage of radioactive waste (Ellison et al., 1990). Contributions to the enthalpies of

176

Navrotsky: Thermochemistry of Borosilicate Melts and Glasses

mixing arise both from the mixing properties of major cation oxides and from contributions by the radionuclides themselves. The principal cation components of radioactive waste, actinides and lanthanides, have high oxidation states. Glasses containing other high-valence cations (e.g. La, Zr, Mo) are used as "proxies" for the thermochemical behavior of the radionuclides in waste glasses. Table 5. Typical compositions of commercial borosilicate glasses Glass Si0 2 Vycor Pyrex E-glass for fibers Nuclear waste disposal

96 81 55 55

AI2Q3

0 2 15 4

wt % oxide CaO Na 2 0 0 1 20 1

0 4 0 9

B203

Other

4 12 10 5

trace trace trace 26a

a. Including 1% TiC>2, 7% Fe2C>3, 3% FeO, 3% MnO, 1% MgO, 4% U2O, 3% K 2 0 and 4% radionuclides (Ellison et al., 1990).

The composition of a proposed waste glass composition is shown in Table 5. Silica, trivalent network-forming cation oxides (AI2O3, B2O3, and Fe2C>3), alkali oxides, and oxides of alkaline earth and other divalent cations comprise more than 96 wt % of the parent glass. Waste loadings are at levels of several percent (oxides). Variations in major oxide component ratios among different proposed waste glass compositions are relatively minor. These glasses are generally relatively near multicomponent eutectic compositions. Heats of mixing among silicate and borate components in these systems, with their relatively restricted composition range, are expected to be small, but can be either positive or negative (Ellison et al„ 1990). Thus a small compositional fluctuation in major components is not expected to have large energetic consequences. If enthalpies of mixing are small and exothermic, glass-glass phase separation is unlikely. If enthalpies of mixing are slightly endothermic, however, the solution behavior of high-valence cations (lanthanides and actinides) may exacerbate an existing tendancy towards unmixing. Very small endothermic heats of mixing are sufficient to drive glass-glass phase separation, forming silica-rich and network-modifierrich regions (Henry et al., 1982; Ellison et al., 1990). Such macroscopic phase-separation may be preceded by analogous clustering on a microscopic scale (phase-ordering). Clustering and phase separation may cause glasses to be unstable (thermodynamically and kinetically) relative to crystallization, hydration, leaching, and other reactions. Given renewed interest in glass waste forms for high level civil and military waste, including weapons plutonium (Holdren et al., 1994), additional study of the thermodynamics of both model and real systems would be very useful. BORON IN MAGMAS In natural silicate melts, boron is a trace rather than a major element (see also Ding well et al. and London et al., this volume). What can be gleaned from the above discussion about the thermodynamic and structural state of boron in magmatic systems?

Navrotsky: Thermochemistry of Borosilicate Melts and Glasses

177

The NaBSÍ308-NaAlSÍ308 studies (Geisinger et al., 1988) have some implications for the behavior of boron in natural magmatic systems. Anhydrous albite glass is a relatively simple three-dimensional framework of silicate and aluminate tetrahedra with only a defect level of non-bridging oxygens (Taylor and Brown, 1979; Riebling, 1966; Seifert et al., 1982; Oestrike et al., 1987), and has a structure broadly similar to that of anhydrous rhyolite glass. Adding Rd component to Ab glass produces a much more complex clustered structure with substantial amounts of trigonal B (Geisinger et al., 1988). Boron concentrations in fractionated silica melts are at most only 1 wt % B2O3 (Pichavant and Manning, 1984). Based on the Ab-Rd glass NMR data (Geisinger et al., 1988), the majority of this boron would occur in trigonal coordination, possibly segregated from the aluminosilicate tetrahedral framework. However, some boron may also occur as WB, which appears to be incorporated in the silicate network in a manner analogous to t^B in reedmergnerite. The Ab-Rd calorimetric and structural data do not support the idea that boron occurs in a sodium-tetraborate-like melt species, suggested as a thermodynamic component by Bumham and Nekvasil (1986), in the application of the "quasi-crystalline model" to granite pegmatite magmas. However, the effects of increasing temperature, which might stabilize t ^ B (Araujo, 1980); the effects of increasing pressure, which might stabilize ^ B , the presence of volátiles in the melt, whose interaction with borate species are not well known; and other variations in composition may be important factors. In general, the Ab-Rd work identifies mechanisms which boron may partially disrupt an aluminosilicate tetrahedral framework, depolymerizing some regions and causing significant changes in the physical and chemical properties. It is likely that a substantial proportion of the boron in magmas may be in trigonal coordination but that whatever tetrahedral boron is present would occur mostly in the alumino(boro)silicate framework. Pichavant (1981) and London et al. (1988) studied the effect of B2O3 on phase relations in water saturated haplogranite and peraluminous granite pegmatite (macusanite) systems, respectively. In both systems, boron partitions preferentially into the H20-rich fluid phase relative to the melt. Large solidus depression, without the formation of new crystalline phases, is seen. London (1987) and London et al. (1988) argue that crystallization of tourmaline depletes the melt of boron, both by crystallization of a boron-bearing mineral and the production of additional H20-rich fluid phase into which remaining boron can partition, driving a "boron quench" at the end-stages of primary crystallization in pegmatites. It is clear that additional quantitative thermodynamic data for both hydrous fluids and wet silicate melts containing boron are needed. The addition of B2O3 lowers the viscosity of haplogranite liquids; this effect is most pronounced at low levels of B2O3 addition (Dingwell et al., 1992; Dingwell et a l , this volume). The discussion above, which suggests that the fraction of boron which assumes trigonal coordination, and therefore is most disruptive of the tetrahedral framework, is highest at low levels of boron substitution, suggests a microscopic reason for this observation. UNANSWERED QUESTIONS, MISSING DATA, AND FUTURE DIRECTIONS The thermochemistry of simple binary and ternary borate systems is relatively well known. As mentioned above, proposed nuclear waste glasses, which contain B2O3 as a major component in a complex and somewhat variable multicomponent system, need additional fundamental study. Do the structural state and associated energetics vary with quench rate? Does radiation damage affect the energetics? How are composition, KlB/I^B ratios, and incipient immiscibility linked, and how do high field strength cations

178

Navrotsky: Thermochemistry of Borosilicate Melts and Glasses

(including actinides) influence this behavior? How are fundamental properties and durability related? In magmatic systems (see Dingwell et al. and London et al., this volume), many of the above questions also need to be answered, although both the boron and the high field strength cations are present at minor to trace concentrations. The role of water and other volatiles needs more quantitative study. In addition to trace element partition coefficients, quantitative measures of complexation reactions using spectroscopic techniques, solubility studies, and high temperature solution calorimetry are needed. These should be done in both dry and wet systems, and the properties of the aqueous fluid coexisting with the melt must also be better constrained. The thermodynamic properties of boron-bearing minerals, e.g. tourmaline solid solution series, need to be determined (see Anovitz and Hemingway (this volume) for an appraisal of present data). Once a more complete data set is available, thermodynamic modelling of phase relations and crystallization sequences can proceed. It is likely that there are complex interactions among species in melts and fluids which contain boron, hydroxyl, fluorine, aluminum, and high field strength cations, even when these are present at low concentrations in a magma. Such interactions can be exothermic (strong complexation), endothermic (clustering into microscopically heterogeneous domains or macroscopically phase-separated regions), or a complex interplay of both these types of energetics. When kinetic complexity is superimposed on this thermodynamic and structural variability, it becomes clear that the evolution of boron-bearing silicate melts can follow many different paths. Though we now have a qualitative picture of these possibilities, the next decade should see stronger quantitation of thermodynamic, structural, and kinetic properties, with the development of theoretically sound numerical models based on established data for both natural and synthetic multicomponent systems. REFERENCES Araujo RJ (1980) Statistical mechanical model of boron coordination. J Non-Cryst Solids 42:209-230 Bourcier WL (1994) Review of Glass Performance Modeling. Chemical Technology Division, Argonne National Laboratory 94/17, July Bumham CW, Nekvasil H (1986) Equilibrium properties of granite pegmatite magmas. Am Mineral 71:239-263 Charles RJ, Wagstaff FE (1968) Metastable immiscibility in the B2O3-SÍO2 system. J Am Ceram Soc 51:16-20 Charles RJ, Turkalo AM (1969) The question of three-liquid immiscibility in the Na20-B203-SiC>2 system. General Electric Research and Development Center, Schenectady, New York, Report No. 69C-213 Chorlton LB, Martin RF (1978) The effect of boron on the granite solidus. Canadian Mineral 16:239-244 DeYoreo JJ, Navrotsky A, Dingwell DB (1990) Energetics of the charge-coupled substitution S i 4 + Na+ T 3 + in the glasses NaTC>2-Si02 (T = Al, Fe, Ga, B). J Am Ceram Soc 73:2068-2072 Dingwell DB, Knoche R, Webb SL, Pichavant M (1992) The effect of B2O3 on the viscosity of haplogranitic liquids. Am Mineral 77:457-461 Ellison AJG, Navrotsky A (1990) Thermochemistry and structure of model waste glass compositions. In: Scientific Basis for Nuclear Waste Management XIII, V.M. Oversby, P.W. Brown (Eds) Mat Res Soc Symp Proc 176:193-207 Fasolino LG (1965) Heats of solution of crystalline and amorphous boron oxide and boric acid. J Chem Engineering Data 10:373-374 Geisinger KL, Gibbs GV, Navrotsky A (1985) A molecular orbital study of bond length and angle variation in framework silicates. Phys Chem Minerals 11:266-283 Geisinger KL, Oestrike R, Navrotsky A, Turner GL, Kirkpatrick RJ (1988) Thermochemistry and structure of glasses along the join NaAlSi308-NaBSi30g. Geochim Cosmochim Acta 52:2405-2414 Ghanbari-Ahari K, Cameron AM (1993) Phase diagram of Na20-B203-SiC>2 system. J Am Ceram Soc 76:2017-2022

Navrotsky: Thermochemistry of Borosilicate Melts and Glasses

179

Haller W, Blackburn DH, Wagstaff FE, Charles RJ (1970) Metastable immiscibility surface in the system Na20-B203-Si02. J Am Ceram Soc 1:34-39 Haller W, Blackburn DH, Simmons JH (1974) Miscibility gaps in alkali-silicate binaries—data and thermodynamic interpretation. J Am Ceram Soc 57:120-126 Hattori M, Taniguchi A (1977) Heats of solution of mixed alkali borate glasses. Kogyo Kyokai Shi 85:33-37 Henry DJ, Navrotsky A, Zimmermann HD (1982) Thermodynamics of plagioclase-melt equilibria in the system albite-anorthite-diopside. Geochim Cosmochim Acta 46:381-391 Hervig RL, Navrotsky A (1985) Thermochemistry of glasses in the system Na20-B203-Si02. J Am Ceram Soc 11:284-298 Holdren JP, Kelleher CM, Panofsky WKH, Baldeschweiler JD, Doty PM, Flax AH, Garwin RL, Jones DC, Keeny SM, Lederberg J, May MM, Patel CKN, Pollack JD, Steinbruner JD, Wertheim RH, Wiesner JB (1994) Committee on Int'l Security and Arms Control; Nat'l Academy of Sciences, Management and Disposition of Excess Weapons Plutonium, Washington, DC: National Academy Press Holm JL, Kleppa OJ (1967) Thermochemistry of the liquid system lead (II) oxide boron oxide at 800°C. Inorg Chem 6:645-648 Itoh M, Sato S, Yokokawa T (1976) E.m.f. measurements of molten mixtures of lithium oxide +, sodium oxide +, and potassium oxide + boron oxide. J Chem Thermodynamics 8:4834-4847 Itoh H, Sasahira A, Maekawa T, Yokokawa T (1984) Electromotive-force measurements of molten oxide mixture. J Chem Soc, Faraday Trans 1, 80:473-487 Klein J, Miiller F (1987) Measurement of the enthalpy of mixing of the liquid system CaO-B2C>3 by drop calorimetry. High Temperatures-High Pressures 19:201-209. Konijnendijk WL, Stevels JM (1976) The structure of borosilicate glasses studied by Raman scattering. J Non-Cryst Solids 20:193-224 Krauskopf KB (1988) Radioactive Waste Disposal and Geology. New York: Chapman and Hall Levin EM, Robbins CR, McMurdie HF (1964) Phase diagrams for ceramists. Am Ceram Soc. Compiled at the National Bureau of Standards, MK. Reser (ed) London D (1987) Internal differentiation of rare-element pegmatites: effects of boron, phosphorus, and fluorine. Geochim Cosmochim Acta 51:403-420 London D, Hervig RL, Morgan GB (1988) Melt-vapor solubilities and elemental partitioning in peraluminous granite-pegmatite systems: experimental results with Macusani glass at 200 MPa. Contrib Mineral Petrol 99:360-373 Lutze W (1988) Silicate glasses. In: Radioactive Waste Forms for the Future, 3. New York: North Holland Macedo PB, Simmons JH (1974) Theoretical analysis of miscibility gaps in the alkali-borates. J Res National Bureau Standards—A. Physics and Chemistry 78A:53-59 Navrotsky A. Hon R, Weill DF, Henry DJ (1980) Thermochemistry of glasses and liquids in the systems CaMgSi206-CaAl2Si208-NaAlS¡308, Si02-CaAl2Si2C>8-NaAlSi308 and Si02-Al203-Ca0-Nç0. Geochim Cosmochim Acta 44:1409-1423 Navrotsky A, Geisinger KL, McMillan P, Gibbs GV (1985) The tetrahedral framework in glasses and melts - inferences from molecular orbital calculations and implications for structure, thermodynamics, and physical properties. Phys Chem Minerals 11:284-298 Navrotsky A (1986) Thermodynamics of silicate glasses and melts. Mineral Assoc Canada Short Course in Silicate Melts, CM Scarfe (ed) 12:130-153 Navrotsky A (1994) Thermochemistry of crystalline and amorphous silica. Rev Mineral 29:309-329 Navrotsky A (1995) Energetics of silicate melts. Rev Mineral 32:121-144 Oestrike R, Yang W-H, Kirkpatrick RJ, Hervig RL, Navrotsky A, Montez B (1987) High-resolution 23 Na, 27 Al and 2 9 Si NMR spectroscopy of framework aluminosilicate glasses. Geochim Cosmochim Acta 51:2199-2209 Ostvold T, Kleppa OJ (1970) Thermochemistry of liquid borates. II. Parital enthalpies of solution of boric oxide in its liquid mixtures with lithium, sodium, and potassium oxides. Inorg Chem 9:1395-1400 Pichavant M (1978) Etude du système SÍO2-B2O3 à un kb. Diagramme de phase et interprétation thermodynamique. Bull Minéral 101:498-502 Pichavant M (1981) An experimental study of the effect of boron on a water saturated haplogranite at 1 kbar vapour pressure. Contrib Mineral Petrol 76:430-439 Pichavant M, Manning D (1984) Petrogenesis of tourmaline granites and topaz granites; the contribution of experimental data. Phys Earth Planet Interiors 35:31-50 Pichavant M (1987) Effects of B and H2O on liquidus phase relations in the haplogranite system at 1 kbar. Am Mineral 72:1056-1070 Riebling EF (1966) Structure of sodium aluminosilicate melts containing at least 50 mole % SÍO2 at 1500°C. J. Chem. Phys 44:2857-2865 Roy BN, Navrotsky A (1984) Thermochemistry of charge-coupled substitutions in silicate glasses: the systems Ml n + /nA10 2 -Si0 2 (M = Li, Na, K, Rb, Cs, Mg, Ca, Sr, Ba, Pb). J Am Ceram Soc 67:606-610

180

Navrotsky: Thermochemistry

of Borosilicate

Melts and

Glasses

Seifert FA, Mysen BO, Virgo D (1982) Three-dimensional network structure of quenched melts (glass) in the systems Si02-NaAlC>2, Si02-CaAl2C>4 and SiC>2-MgAl204. Am Mineral 67:696-717 Taylor M, Brown GE Jr. (1979) Structure of mineral glasses—I. The feldspar glasses NaAlSi3C>8, KAlSi308, CaAl2Si208. Geocbim Cosmochim Acta 43:61-75 Wight Q, Chao GY (1995) Mont Sainte-Hilaire revisited. Rocks and Minerals 70:90-138

Chapter 5 THERMODYNAMICS OF BORON MINERALS: SUMMARY OF STRUCTURAL, VOLUMETRIC AND THERMOCHEMICAL DATA Lawrence M. Anovitz Chemical and Analytical Sciences Division P.O. Box 2008 MS 6110 Building 4500-S Oak Ridge National Laboratory Oak Ridge, Tennessee 37831 U.S.A.

Bruce S. Hemingway U.S. Geological Surrey National Center MS 953 Reston, Virginia 20192 U.S.A. INTRODUCTION Boron minerals are widespread in the upper crust, occurring in a wide variety of igneous, sedimentary and metamorphic environments (see Anovitz and Grew, this volume). The high ionization potential of boron leads to an important covalent character in its bonding, and to its incorporation in a wide variety of minerals and other compounds. In this respect, it is somewhat similar to silicon (Cotton and Wilkinson, 1980). In addition, the preference of boron for the fluid phase and its isotopic fractionation make boron a useful tracer for a wide range of hydrologic and magma/hydrothermal processes (see Dingwell et al.; London et al.; Palmer and Swihart, Leeman and Sisson; and Shaw this volume). A quantitative understanding of the geology and geochemistry of boron requires that the thermodynamic properties of boron-bearing minerals be known. Three types of information are needed: the volume of each phase as a function of pressure and temperature, its entropy as a function of temperature, and an enthalpy or Gibbs energy at some pressure and temperature to serve as a starting point. In addition, if solid solutions are to be understood, activity-composition relationships for the important solid solution must be available. Unfortunately, such information has been largely unavailable. Standard physicalchemical reference sources contain data on few boron minerals (cf. Robie et al., 1979; Wagman et al., 1982; Chase et al., 1985; Berman, 1988; Cox et al., 1989; Holland and Powell, 1990; Robie and Hemingway, 1995). While some data are available, they are scattered in the literature. Earlier reviews by Sims (1966) and Bassett (1976) concentrated on low-temperature phases, and there has been no attempt to summarize or evaluate all of the available data. In this paper we have collated and systematized the available data to provide a convenient source for this information, and these data have been used to provide a method to estimate values for properties not yet measured. We have thereby identified where gaps in the data base exist, which should encourage others to obtain these data. This paper is divided into several sections. The first summarizes the structures and formulas of the boron minerals to clarify the compositions of the phases to be investigated 0275-0279/96/0033-0005$05.00

182

Anovitz & Hemingway: Thermodynamics of Boron Minerals

and to provide data on cation coordination for later estimates. The next summarizes available volume, entropy, heat-capacity, enthalpy and Gibbs energy data, and experimental data from which enthalpy and Gibbs energy may be derived. Finally, methods are examined to estimate missing thermodynamic properties for boron-bearing phases. FORMULAS AND STRUCTURES OF BORON-BEARING PHASES The formulas and cation coordinations in boron minerals for which such data are available are summarized in Table 1 (at the end of this chapter). There are currently 198 known boron minerals (and 10 more reported phases not yet approved by the IMA, Table 1). Of these, we have been able to locate structural determinations or estimate structures for all but 32. Except where otherwise stated, all of the thermodynamic properties for the phases summarized in this paper refer to the formulas in Table 1. In a few cases coordinations are inferred by analogy to a similar phase, as noted. Thermodynamics is a macroscopic description. While thermodynamic properties are ultimately determined by microstructural properties, knowledge of microstructures is not necessary for the derivation of accurate, precise and useful thermodynamic data. A summary of the structures of boron minerals is, however, useful for our purposes in two respects. The formulas typically presented for complex phases, indicating solid solutions, are of limited thermodynamic use. Thermodynamic properties must be defined for a single composition, although the properties of solid solutions can obviously be represented. Structural refinements have therefore been used to define, where possible, an appropriate end-member composition for each phase. In many cases this was done unambiguously, but in some cases arbitrary choices had to be made. The second reason for summarizing the structures of boron minerals is that crystallographic data provide a means of estimating thermodynamic properties for phases for which no thermodynamic data exist. Several authors (e.g. Robinson and Haas, 1983; Chermak and Rimstidt, 1989) have shown that it is possible to express thermodynamic data in terms of (Active) additive properties (e.g. heat-capacity) of cation coordination polyhedra. Consequently, the coordination of each cation must be known. One caveat involves the coordinations of large cations. Whereas the coordination numbers of small cations can usually be determined unambiguously, this is often not the case for larger cations. The coordinations given reflect some arbitrary cutoff value for the cation-anion distance. The coordinations listed in Table 1 are those chosen by the workers who determined the structure of each phase. Therefore, these values are not necessarily consistent, and the reader is referred to the original source for more detail. Table 1 also gives the symmetry and space group of each mineral. No attempt has been made to summarize all of the information on the structures of boron minerals, to show the coordinations of the anions or anion groups, or to show the details of the mineral structures. Interested readers should see the summary of Hawthorne et al. (this volume), or the individual structural refinements listed in Table 1. In addition, end-member formulas and coordinations are not given for phases for which no structure is known. Those for which arbitrary choices had to be made are discussed below. Ammonioborite. The formula obtained by Merlino and Sartori (1971) (NH4)3Bi5O20(OH)8-4H2O, contains more hydrogen (16 H per 3 NH4) than listed by Fleischer and Mandarino (1995) (NH4) 2 BioOi6-5H 2 0 (15 H per 3 NH4) and originally proposed by Schaller (1933). The formula given by Clark and Christ (1959) (NH4)20-5B203-5l/3H20 also yields 16 H per 3 NH4. We have therefore accepted the total H and (0H)/H20 partitioning of Merlino and Sartori (1971).

Anovitz & Hemingway: Thermodynamics of Boron Minerals

183

Bakerite. The nature, and possibly the amount, of water in bakerite remains uncertain. Some authors (Giles, 1903; Kramer and Allen, 1956; Baysal and Dilekoz, 1975) suggest 4Ca0-2.5B203-3Si02-3H20, while others (Palache et al., 1951; Kusachi et al., 1994) give 4Ca0-2.5B203-3Si02-2.5H20. No data exist on the hydroxyl/molecular water distribution in this phase. The results of Kusachi et al. (1994) have been accepted, and a molecular water-free formula adopted, but more work is needed. Charlesite. Charlesite is a member of the ettringite group, with the formula Ca28 (OH)Cl. We have adopted this suggestion, but it needs to be tested. The situation is more complicated for titantaramellite, as the Ti 4 + /Fe 3 + /Fe 2 + ratios can also vary and few water analyses are available. Titantaramellite is related to taramellite by the exchange 2 Fe 3+ = Ti 4+ + Fe 2+ . Thus the most reasonable formula is Ba4Ti2Fe2B2Sig028(0H)Cl. Most of the samples yield CI > 0.5, and this formula has been adopted here. The one available water analysis shows no water present (Alfors and Pabst, 1984; Table 3), but this specimen contains significant ferric iron. This suggests the endmember Ba4Ti2Fe2+Fe3+B2Si8C>29Cl. Finally, one sample, from the Victor Claim, California, has almost no CI, suggesting Ba4Ti2Fe2B2Sis029. It thus appears that available titantaramellite analyses may represent as many as three end-members, two currently unnamed. Available analyses also suggest that Mg end-members may exist. While chemical zoning may complicate the interpretation, analyses complete with water contents are needed. Tienshanite. The formula of tienshanite remains uncertain. The normalized analysis given by Dusmatov et al. (1967) was (cf. new formula by Cooper et al., 1998; see p. 262). (Na,Ca) i .g7(Ba,K) i .34(Mn,Fe) i .o l (Ti,Nb,Ta) i .ooB i ,78Sis .95O20, which suggested the end-member Na2BaMnTiB2Si6O20- Refinement by Malinovskii et al. (1977, 1979) showed that the structure is partially mica-like, with six-membered rings of Mn five-point half octahedra, Ti,Nb octahedra, tetrahedral boron and tetrahedral Si. The structure contains distinct K, Ba and Ca sites, two different Na sites, and an OH anion, leading to the formula: KNa9Ca 2 Ba6(Mn,Fe) 6 (Ti,Nb,Ta) 6 Si3 6 Bi20i23(0H)2. The problem is that, for the Mn 2+ and Ti end-member, this does not charge balance. A number of possible explanations exist. Replacement of Ti 4+ by Nb 5 + is not sufficient to explain the discrepancy unless all of the Ti is replaced, which does not correspond to the analysis. Mn could be trivalent, although the light color of this mineral suggests this is not the case. Recalculation of the original analysis on the basis of 124 oxygens (on an anhydrous basis) assuming all iron ferric yields (Grew, pers. comm.) (K2.308Na9.798Cai.78i)i3.887Ba6.01l(Mn5.630Fe3+0.64l)6.271 (Ti4.422Nbi.59iTa0.195)6.208Si36.85Bll.03Ol23(OH)2. This must, by definition, charge balance. However, with the exception of boron, all of the sites so derived are overfilled, unlikely if the sites defined by the structural refinement are correct. It is possible that the problem with the formula for tienshanite lies in the structural refinement. Guggenheim (pers. comm.) notes that Malinovskii et al. (1977, 1979) do not present standard deviations for the atomic coordinates, and that the R factor is based on the isotropic case, suggesting that they could not continue with an anisotropic refinement. This suggests that the structural model is in need of further assessment. As no resolution can be obtained from the available data, a new refinement of the structure of tienshanite is needed before its true formula can be defined. The formula given in Table 1 suggests the possibility of Mn 3 + but, as noted above, this seems unlikely.

Anovitz & Hemingway: Thermodynamics of Boron Minerals

187

Tinzenite. Tinzenite is a member of the axinite group in which one calcium site is filled with manganese (Milton et al., 1953). Sanero and Gottardi (1968) defined tinzenite as "those axinites with Ca < 1.5 and Mn > Fe (but usually Mn » Fe)". Basso et al. (1973) showed that, of the two independent calcium sites in axinites, manganese occupies only one. We have therefore defined end-member tinzenite as the ordered compound CaMnMnAl 2 BSi40i 5 (0H). Tourmaline (buergerite, chromdravite, dravite, elbaite, ferruvite, foitite, liddicoatite, olenite, povondraite, schorl, uvite). The tourmaline group, discussed in more detail elsewhere in this volume, is one of the most compositionally complex groups of boron minerals. A number of end-member minerals have been named, although it is clear (see Burt, 1989; Hawthorne et al., this volume; Henry and Dutrow, 1990, this volume) that these are not sufficient to describe the range of tourmaline compositions observed in nature. For our purposes, we have defined end-members for accredited mineral species on the basis of simple exchanges and the limits allowable by charge balanced If we begin with schorl as (NaFe2+3Al6B3Si6027(0H)4) those exchanges are: NaFe 3+ 3 Al 6 B 3 Si6027(03Fi) Fe 2+ + (OH) = Fe 3+ + O; (OH) = F chromdravite: NaMg3Cr6B3Sig027(0H)4 Fe 2+ = Mg ; Al = Cr 3+ dravite: NaMg 3 Al 6 B3Si60 2 7(0H)4 3 Fe 2+ = 3 Mg 2+ elbaite: Na(Lii.5Ali.5)Al6B3Si6027(0H)4 3 Fe 2+ =1.5 Li + +1.5 Al 3+ ferruvite: CaFe 3 (Al5Mgi)B3Si60 2 7(0H)4 Na + Al = Ca + Mg foitite: •(Fe 2+ 2 Al)Al 6 B 3 Si 6 0 2 7(0H)4 Na + Fe 2+ = • + Al liddicoatite: Ca(Li 2 Al)Al 6 B 3 Si 6 0 2 7(0H) 4 Na + 3 Fe 2+ = Ca + 2 Li + Al olenite: NaAl 3 Al 6 B 3 Si 6 0 2 7 (0 3 (0H)i) Fe 2+ + (OH) = Al + O 3+ povondraite: NaFe 3 Al 6 B 3 Si 6 0 2 7(0 3 (0H)i) Fe 2+ + (OH) = Fe 3+ + O uvite: CaMg 3 (Al 5 Mgi)B3Si 6 027(0H)4 Na + Al = Ca + Mg; Fe 2+ = Mg buergerite:

Werdingite. Werdingite contains several complex solid solutions (Niven et al., 1991). To define an end-member we have filled the Mg/Fe trigonal bipyramidal site with Mg, as in the refined specimen that site contains 84% Mg. Both Al and Fe 3 + occur in the tetrahedrally coordinated Fe(l) site. Despite the label by Niven et al. (1991), Fe(l) contains 79% Al, and we have defined the end-member as Fe 3 + free. The refinement showed 77% occupancy by B at the B(2) boron site. We have defined this site as full for the endmember. Finally, the second tetrahedrally coordinated Al site Al(4)2 was only 23% occupied. We have defined the end-member with this site vacant. This gives Mg2Ali2Al2B4 Si4037 (cf. Werding and Schreyer, 1992, this volume) but a number of alternatives exist. Wiserite. The tetrahedral site in wiserite can contain both Si and Mg, which Pertlik and Dunn (1989) suggested are present in a 1:1 ratio. We have accepted this suggestion, and defined the O, (OH) and CI concentrations to charge balance the formula.

188

Anovitz & Hemingway: Thermodynamics of Boron Minerals VOLUMETRIC PROPERTIES

The volumetric properties of minerals are, in part, some of the most complete and precise thermodynamic data available. Both structure refinements and powder-diffraction data give V°(298), with individual lattice parameters commonly measured to better than one part in a thousand. There are only six reported natural boron phases for which V°(298) data are not available, and three of these are not yet approved, named minerals. On the other hand, data on the thermal expansivity and compressibility of boron minerals are extremely limited. Only twenty four data sets on the thermal expansivity of sixteen phases (counting all tourmalines as one) have been located. Compressibility data are even less available. Only seventeen data sets have been located, providing data for only three minerals (again counting all tourmalines as one). Whereas this information is of lesser importance in low pressure environments and for reactions involving a fluid phase, the effects of thermal expansivity and compressibility on solid-solid reactions, and on the calculated stabilities of boron minerals at high pressures is probably significant. Additional thermal expansivity and compressibility data are therefore needed. Standard state volume The standard state (1 bar, 25°C) volumes for boron minerals are listed in Table 2. Where lattice parameters are not available, molar volumes were calculated from formula weights and measured densities. For many minerals, this list is a reasonably complete summary of available volumetric data, although for more complex phases such as tourmalines only a subset of the data is given. The compositional variation common in many phases means that volumetric data, while accurate, may require correction to represent the properties of the end-member phase. In addition, volume as a function of composition can be modeled to yield partial molar volumes, and the pressure dependence of activity/composition relations. Unfortunately, analytical data were not provided for a number of such measurements. Nonetheless, the data in Table 2 provide a good reference for molar volumes for most boron minerals. Analysis of these data to provide end-member values, or to model volumes of mixing is beyond the scope of this paper, but sufficient data exist to begin this task for several of the complex solid solutions in this system. Thermal expansion and compressibility Available data on thermal expansion and compressibility of boron minerals are listed in Tables 3 and 4 (above). We have chosen to present the results in the original form because refitting to some standard equation would decrease the precision of the presented values, and because a number of different equations are used for these data. Thus, thermal expansion data are presented in terms of constant and temperature-dependent values as well as temperature-dependent cell parameter equations. Compressibility data are given as compressibilities, bulk moduli and normalized volumes as a function of pressure. Tables 3 and 4 clearly show the limited amount of data available. Only for tourmalines does a significant data base exist The data of Hemingway et al. (1996) have expanded the number of minerals for which expansivities are available, but a systematic examination of the thermal expansion and compressibility of boron minerals is needed before the more volume-sensitive reactions can be calculated accurately.

Anovitz & Hemingway: Thermodynamics of Boron Minerals

6 > T3

J33 I

II a

0 H H U + ''o1 E-

§

1

H «

a. x t> m E 3 O > tri O È9 U a

i n c s c s 3

3

7 cn o o\ CNCS{S cs

o

o

^ +

+

2

>9

• ci VO

vo C-4

'

"o cn _ vo o

o

H "

H « no r~ ¿ 9 » o

..

o

n

o

O +

. o o +

o £ >n

.S 3

ca

I

» o

o

o +

O +

r-

vo

eK ts w-i

g «o

>

«

o n

f> H • 1-H

r H ® . o ¿ 5 1—t . o o S W

v

« 2 o\ VO ! o t s oo

T"

(m

h

s o vo r ~ o Ti

.10 ts §

d

— —

+ +

>n r-

i-i o

>

ts

r~

ON T f i J »1

T

H

o i-i i-i S > o w-1 o

>

«

Ò

H Ò

o VO >

o\

>ri vo m TJ- ON t— v o CS ( S i-i' v o 0 0 o cs i-i to

a\

d

A

S

-r

+ H H J - riT "V H O f i . ^—. „ in vo g . I*-, * Oft « 3 5 m 0 1 ! d P ° 2 + o • 0 0 + 0 0 3 N O o M - 1-1 v o

m

o d d + + +

tr, o

H Ò

5

fn

Óv -\ '

P

cn

^Ì-ÌT'

2 " C T- Q o OO < s i—t

l ! ^ (lì H ci O 1-* O O 1-1 i—i oo T1 J vò 00 Tt ) OO i-i T cs Ti o

m d

Q

;

J r ^

Tt n

1

>

fs

I

Jr"?

o 1-1

rcm

o

•t ^ i-i lI

C*"1 ^ r-1 O O t s P P i n « o > o

1

> o v ' • mS *

T* r i

» ! t-

_ o

y

" v o

J3 o o

*

f - P h faP f—i , 2 i vó M t N Ó O O S O S 1 > 1 CT\ 1

1 o « n

m m m f i r - t— rt s t s t s

7

0\ o \

o


-

m

00 «

£

O CQ.

>

Anovitz &

/- N /—s ON NO NO NO t s 00 (S ^ N O NO ON NO Os © ON / — * ON t—i >n —1 O r—1 v ' On w On < J 3 OO ¿3 S S

, NO NO on OS

Thermodynamics of Boron Minerals

Hemingway:

¡ 1

c3

I

.a 00 PQ Cu . .

I

i

ft

I

oo o 00 s o f ^H © ©

On OO I O f i 'O g 'w « I l i - i l p S a'a s-a jì

P

¿3 N CU C/5

•a

ug c* 9]Cl-H20: a piezoelectric zeolite-type pentaborate. Am Mineral 64:187-195 Ghose S, Wan C (1977) Hilgardite and parahilgardite piezoelectric zeolite type phases. Nature (London) 270(5638):594-595 Ghose S, Wan C, Ulbrich HH (1976) Structural chemistry of borosilicates. I. Garrelsite, NaBa3SÍ2B70ig(0H)4: a silicoborate with the pentaborate [B5O12]9" polyanion. Acta Crystallogr B32:824-832 Giacovazzo C, Menchetti S, Scordari F (1973) Crystal structure of tincalconite. Am Mineral 58:523-530 Giese RF Jr. (1966) Crystal structure of kemite, Na2B406(0H)2-3H20. Science 154:1453-1454 Giese RF Jr., Penna G (1983) The crystal structure of sulfoborite, Mg3S04(B(0H)4)2(0H)F. Am Mineral 68:255-261 Giles WB (1903) Bakerite (a new borosilicate of calcium) and howlite from California. Mineral Mag 13:353-355 Giuseppetti G, Mazzi F, Tadini C (1977) The crystal structure of harkerite. Am Mineral 62:263-272 Giuseppetti G, Mazzi F, Tadini C, Larsen AO, Aasheim A, Raade G (1990) Berborite polytypes. N Jahrb Mineral Abh 162:101-116 Gode H, Cenkus R, Mazurov Yul, Osokina RM, Sizov IV (1971) Transition of noncrystalline calcium borate in inyoite studied from boron isotopic composition. Latv PSR Zinat Akad Vestís Kim Ser 6:752 Gode H, Majore I, Sokolova EV, Yamnova NA (1986) Identity of synthetic magnesium borate 2Mg0-3B2C>3-15H20 to the mineral inderite (Mg23.15). Latv PSR Zinat Akad Vestís Kim Ser 5:532533 Gorbov AF (1973) The composition of wiserite and mineral X - two manganese borates. Trudy Vsesoyuznogo Nauchno-Issledovatel'skogo Proyektnogo Inst Galurgii 64:54-62 (in Russian) Gorshenin AD, Pertsev NN, Organova NI, Laputina IP, Nikitina IB (1977) Discovery of monoclinic kurchatovite and cubic harkerite with low silicon content in the Balkhash district Dokl Akad Nauk SSSR 236:1203-1206 Gorskaya MG, Frank-Kamenetskaya OV, Rozhdestvenskaya IV, Frank-Kamenetskii VA (1982) Refinement of the crystal structure of Al-rich elbaite, and some aspects of the crystal chemistry of tourmalines. Kristallografiya 27:107-112 Gourdant JP, Robert JL, Linnen RL (1994) Controle structural de la correlation Na-F dans les tourmalines: étude experimentale et spectroscopique. Bull Liason Soc Fr Mineral Crist 6:49 Grew ES, Chemosky JV, Werding G, Abraham K, Marquez N, Hinthome JR (1990) Chemistry of kornerupine and associated minerals, a wet chemical, ion microprobe, and X-ray study emphasizing Li, Be, B and F contents. J Petrol 31:1025-1070 Grew ES, Pertsev NN, Boronikhin VA, Borisovskiy SYe, Yates MG, Marquez N (1991) Serendibite in the Tayozhnoye deposit of the Aldan Shield, eastern Siberia, USSR. Am Mineral 76:1061-1080 Grew ES, Belakovskiy DI, Fleet ME, Yates MG, McGee JJ, Marquez N (1993) Reedmergnerite and associated minerals from peralkaline pegmatite, Dara-i-Pioz, southern Tien Shan, Tajikistan. Eur J Mineral 5:971-984 Grew ES, Yates MG, Belakovskiy DI, Rouse RC, Su S-C, Marquez N (1994) Hyalotekite from reedmergnerite-bearing peralkaline pegmatite, Dara-i-Pioz, Tajikistan and from Mn skarn, Lângban, Varmland, Sweden: a new look at an old mineral. Mineral Mag 58:285-297 Grew ES, Cooper MA, Hawthorne FC (1996) Prismatine: revalidation for boron-rich compositions in the kornerupine group. Mineral Mag 60:483-491

210

Anovitz & Hemingway: Thermodynamics

of Boron

Minerals

Grice JD, Robinson GW (1989) Feruvite, a new member of the tourmaline group and its crystal structure. Can Mineral 27:199-203 Grice JD, Ercit TS (1986) The crystal structure of moydite. Can Mineral 24:675-678 Grice JD, Ercit TS (1993) Ordering of Fe and Mg in the tourmaline crystal structure: the correct formula. N Jahrb Mineral Abh 165:245-263 Grice JD, Van Velthuizen J, Dunn PJ, Newbury DE, Etz ES, Nielsen CH (1986) Moydite (yttrium, rare earth elements) [B(OH)4](CC>3). A new mineral species from the Evans-Lou pegmatite, Quebec. Can Mineral 24:665-673 Grice JD, Ercit TS, Van Velthuizen J, Dunn PJ (1987) Poudretteite, KNa2B3Sii2C>3o. A new member of the osumilite group from Mont Saint-Hilaire, Quebec, and its crystal structure. Can Mineral 25:763766 Grice JD, Ercit TS, Hawthorne FC (1993) Povondraite, a redefinition of the tourmaline ferridiavite. Am Mineral 78:433-436 Grice JD, Bums PC, Hawthorne FC (1994) Determination of the megastructures of the borate polymorphs pringleite and niitenbergite. Can Mineral 32:1-14. See also Jambor JL, Roberts AC, Vanko DA (1994) New Mineral Names. Am Mineral 79:1009-1014 Griffen DT (1988) Howlite, Ca2SiBs09(0H)5: Structure refinement and hydrogen bonding. Am Mineral 73:1138-1144 Grigor'ev AP (1971) Synthesis and study of high-temperature borates. In VV Lapin (ed) Tr Soveshch Eksp Tekh Mineral Petrogr 8th, Meeting Date 1968, p 160-163, Nauka, Moscow Grigor'ev AP, Brovkin AA (1969) Sellaite-kotoite binary section of the magnesium oxide-magnesium fluoride-boron oxide system. Dokl AkadNauk SSSR 186:1387-1390 Grigor'ev AP, Nekrasov IY (1963) Hydrothermal synthesis of minerals of the ludwigite-vonsenite series. Dokl Akad Nauk SSSR 151:671-674 Guggenheim S, Schulze WA, Harris GA, Lin J-C (1983) Noncentric layer silicates: An optical second harmonic generation, chemical, and X-ray study. Clay Clay Minerals 31(4):251-260 Gulensoy H, Gedikbey T (1980) Studies on production of borax from colemanite, ulexite, pandermite and tunellite minerals with solid phase reactions. Doga Sen B 4:33-37 Gulensoy H, Teberdar T (1972) Solid phase reactions with tunellite mineral. Bull Min Res Explor Inst Turk 79:50-62 Gurvich VM, Sokolov VA (1976) Enthalpies of inyoite and colemanite formation. Geokhimiia 3:455-457 Gurvich VM, Semenov YuV, Aksenova TL, Gorbunov VE, Gavrichev KS, Zhdanov VM, Turdakin VA, Khodakovsky IL (1980) Thermodynamical properties of inyoite and colemanite and their formation under natural conditions. Geokhimiia 5:671-679 Gurevich VM, Gorbunov VE, Aksenova TD, Gavrichev KS, Khodakovski IL (1988) Thermodynamic properties of hydroboracite, CaMg[B303(0H)5]2-H20. Zh Fiz Khim 62:3110-3113 Hackwell TP, Angel RJ (1992) The comparative compressibility of reedmergnerite, danburite and their aluminium analogues. Eur J Mineral 4:1221-1227 Hansen S, Haalenius U, Lindqvist B (1988) Antimony-rich pinakiolite from Längban, Sweden: A new structural variety. N Jahrb Mineral Mh 5:231-239 Hawthorne FC, Grice JD (1990) Crystal-structure analysis as a chemical analytical method: Application to light elements. Can Mineral 28:693-702 Hawthorne FC, MacDonald DJ, Burns PC (1993) Reassignment of cation site occupancies in tourmaline: Aluminum-magnesium disorder in the crystal structure of dravite. Am Mineral 78:265-270 Heide F (1955) Über bemerkenswerte Borazitvorkommon in den Kalilagern des Südharzbezirkes. Chemie der Ende 17:214-216. See also: Am Mineral 41:372, 1956 Heide M (1992) Synthese und Stabilität von Grandidierite, MgA^BSiOg. Diplomarbeit, Ruhr-Universität, Bochum, Germany Heinrich EW (1946) A second discovery of inderite. Am Mineral 31:71 Helgeson HC, Delany JM, Nesbitt HW, Bird DK (1978) Summary and critique of the thermodynamic properties of rock-forming minerals. Am J Sei 278A, 229 pp Helme BG, King PJ (1978) The elastic constants of iron tourmaline (schorl). J Mater Sei 13:1487-1489 Helvaci C (1978) A review of the mineralogy of the Turkish borate deposits. Mercian Geol 6:257-270 Hemingway BS (1982) Thermodynamic properties of calcium aluminates. J Phys Chem 86:2802-2803 Hemingway BS, Anovitz LM, Robie RA, McGee JJ (1990) The thermodynamic properties of dumortierite Si3B[Al6.75D0.25Oi7.25(OH)0.75]. Am Mineral 75:1370-1377 Hemingway BS, Evans HT Jr., Mazdab FK, Anovitz LM (1996) Thermal expansion of fluoborite, danburite, sinhalite, datolite, elbaite, dravite, kornerupine, dumortierite, ferro-axinite, and manganaxinite between 25 and about 1200°C US Geological Survey Open-File Report, in prep Henry, DJ, Dutrow BL (1990) Ca substitution in Li-poor aluminous tourmaline. Can Mineral 28:111-124 Hoffmann C, Armbruster T (1995) Crystal structure of a (001) twinned sussexite Mn2B2Ü4(OH)2 from the Kalahari Manganese Field South Africa. Schwiez Mineral Petrogr Mitt 75:123-133

Anovitz & Hemingway: Thermodynamics

of Boron Minerals

211

Hogarth DD, Chao GY, Harris DC (1972) New data on hellandite. Can Mineral 1:760-776 Holland TJB (1989) Dependence of entropy on volume for silicate and oxide minerals: a review and predictive model. Am Mineral 74:5-13 Holland TJB, Powell R (1990) An enlarged and updated internally consistent thermodynamic dataset with uncertainties and correlations: the system K20-Na20-Ca0-Mg0-Mn0-Fe0-Fe2C>3-Al203-Ti02-Si02C-H-0 2 . J Met Geol 8:89-124 Honea RM, Beck FR (1962) Chambersite, a new mineral. Am Mineral 47:665-671 Horn WD, Van Wagner E (1903) The solubility-curves of sodium tetraborate. Am Chem Soc J 30:344-350 Hoskins BF, Mumme WG, Pryce MW (1989) Holtite, (Si2.25Sbo.75)B[Al6(Alo.43Ta0.27 •0.30)Oi5(O,OH)2.25]: Crystal structure and crystal chemistry. Mineral Mag 53:457-463 Huang S, Zhang Q, Li Y, Chen P (1988) Specific heat measurements of kumakovite (2Mg03B2C>3-15H20) from 65 to 310 K and calculation of its thermodynamic properties. Huaxue Xuebao 46:967-971 Huang Z, Wang P (1994) Yuanfuliite-A new borate mineral. Acta Petrol Mineral 13(4): 328-334 (in Chinese). Referenced from Jambor JL, Pertsev NN, Roberts AC (1996) New Mineral Names. Am Mineral 81:252-253 Hurlbut CS Jr., Aristarain LF (1967) Rivadavite, NagMgB2404o-22H20, a new borate from Argentina. Am Mineral 52:326-335 Hurlbut CS Jr., Aristarain LF, Erd RC (1973) Kernite from Tincalayu, Salta, Argentina. Am Mineral 58:308-313 Hurlbut CS Jr., Erd RC (1974) Aristarainite, Na20-Mg0-6B203-10H20, a new mineral from Salta, Argentina. Am Mineral 59:647-651 Hurlbut CS Jr., Aristarain LF, Erd RC (1977) Hydrochlorborite from Antofagasta, Chile. Am Mineral 62:147-150 Huteheon I, Gunter AE, Lecheminant AN (1977) Serendibite from Penrhyn Group marble, Melville Peninsula, District of Franklin. Can Mineral 15:108-112 Iorysh ZI, Rumanova IM, Belov NV (1976) Crystal structure of sulfoborite Mg3(0H)2[S04][B(0H)4]2. Dokl Akad Nauk SSSR 228:1076-1079 Ito T, Mori H (1953) The crystal structure of datolite. Acta Crystallogr 6:24-32 Jaffe HW, Molinski VJ (1962) Spencite, the yttrium analogue of tritomite from Sussex County, New Jersey. Am Mineral 47:9-25 Jobbins, EA Tresham AE, Young BR (1975) Magnesioaxinite, a new mineral found as a blue gemstone from Tanzania. J Gemmol 14:368-375 Jouravsky G Permingeat F (1964) La gaudefroyite, une nouvelle espèce minéral. Soc Franç Minéral Cristall Bull 87:216-219 Kalacheva LV, Yamnova NA, Pusharovskii DYu, Konovalinko SD (1992) Crystal structure of the new natural borate, tusionite, manganese tin borate (MnSn(BC>3)2). Kristallografiya 37:1557-1558 Kan X, Xie H, Li Y, Zhang E (1984) The magnetic ordering of ludwigite and synthetic vonsenite and their Moessbauer spectra. Kexue Tongbao (Foreign Lang Ed) 29:928-931 Kato A, Matsubara S (1980) Manganese borate minerals from Japan. Kobutsugaku Zasshi 14 (Tokubetsugo 3):86-97 (in Japanese) Kazanskaya EV, Chemodina TN, Egorov-Tismenko YuK, Simonov MA, Belov NV (1977) Refined crystal structure of pentahydroborite Ca[B20(0H)6]-2H20. Kristallografiya 22:66-68 Kelley KK (1960) Contributions to the data on theoretical metallurgy. XIII High-temperature heat-content, heat-capacity, and entropy data for the elements and inorganic compounds. US Bur Mines Bull 584, 232 p Khomyakov AP, Proshchenko EG (1974) Find of stillwellite in alkaline pegmatites of the Synnyrsk Massif. Tr Mineral Muz Akad Nauk SSSR 23:229-230 Khomyakov AP, Sandomirskaya SM, Malinovskii YuA (1980) The new mineral kalborsite (K6BAl4Si6O20(OH)4Cl). Dokl Akad Nauk SSSR 252:1465-1468 Khomyakov AP, Rogachev DL (1991) Searlesite and reedmegnerite from the Lovozero alkali massif. Izv Akad Nauk SSSR Ser Geol 1991:149-152 Kimata M (1977) Synthesis and properties of reedmergnerite. Ganseki Kobutsu Kosho Gakkaishi 72:162172 King GSD, Sengier-Roberts L (1988) Drugmanite, Pb2(Feo.78Alo.22)H(P04)2(C)H)2: its crystal structure and place in the datolite group. Bull Mineral 111:431-437 King EG, Torgeson DR, Cook OA (1948) High-temperature heat contents of 3CaO-B2C>3, 2Ca0 B 2 03, Ca0-B 2 03, and Ca0-2B 2 03. J Am Chem Soc 70:2160-2163 Klaska R, Grew ES (1991) The crystal structure of boron-free kornerupine: conditions favoring the incorporation of variable amounts of boron through MB O MSi substitution in kornerupine. Am Mineral 76:1824-1835

212

Anovitz & Hemingway: Thermodynamics of Boron Minerals

Kondrat'eva VV (1977) Change in natural calcium hexaborates during heating. Zap Vses Mineral Obshchestva 106:490-495 Kondrat'eva VV, Zorina ML (1976a) Behavior of inyoite during heating. In VA Frank-Kamenetskii (ed), Kristallokhim Stmkt Osob Mineral, p 113-116, "Nauka", Leningr Otd, Leningrad Kondrat'eva W , Zorina ML (1976b) Changes in pinnoite aid szaibelyite during heating. . In VA FrankKamenetskii (ed), Kristallokhim Strukt Osob Mineral, p 108-113, "Nauka", Leningr Otd, Leningrad Kondrat'eva VV, Filatov W (1986) Thermal expansion of hydroboracite CaMg[B304(0H)3]2-3H20. Izv Akad Nauk SSSR Neorg Mater 22:273-276 Konev AA, Lebedeva VS, Kashaev AA, Ushchapovskaya ZF (1970) Azoproite, a new mineral of the ludwigite type. Zap Vses Mineral Obshchestva 99:225-231 Koneva AA, Ushchapovskaya ZF (1991) Harkerite and bultfonteinite from the skarns of Tazheran alkaline intrusion (southwestern Baikal region). Geol Geofiz 2:74-78 Konnert JA, Clark JR, Christ CL (1970) Crystal structure of strontioginorite, (Sr,Ca)2Bi402o(OH)6 •5H 2 0. Am Mineral 55:1911-1931 Konnert JA, Clark JR, Christ CL (1970) Crystal structure of fabianite, CaB30s(0H), and comparison with the structure of its synthetic dimorph. Z Kristallogr 132:241-254 Konnert JA, Clark JR, Christ CL (1972) Borate minerals. XVn. CaB50 8 (0H) B(0H)3-3H 2 0. Crystal structure and comparison with related borates. Am Mineral 57:381-396 Konnert JA, Appleman DE, Clark JR, Finger LW, Kato T, Miura Y (1976) Crystal structure and cation distribution of hulsite, a tin-iron borate. Am Mineral 61:116-122 Konovalenko SI, Rossovskii LN, Anan'ev SA (1983) Jeremejevite as a mineral recently found in the USSR. Zap Vses Mineral Obshchestva 112:212-217 Konovalenko SI, Voloshin AV, Pakhomovskii YaA, Anan'ev SA, Perlina GA, Rogachev DL, Kuznetsov, VYa (1983) Tusionite (MnSn(BÛ3)2)—a new borate from granite pegmatites of the southwest Pamir. Dokl Akad Nauk SSSR 272:1449-1453 Kopp, H. (1864) Ueber die specifische Wärme der starren Körper. Ann Chem Pharm Suppl 3(l):289-342 Kovalev AV, Andreeva GT (1979) Low temperature x-ray diffraction investigation of boracite Ni3B70i3l. Fiz Tveid Tela (Leningrad) 21:1744-1749 Kramer, H and Allen, RD (1956) A restudy of bakerite, priceite, and veatchite. Am Mineral 41:689-700 Kravchuk TA, Nekrasov IY, Grigoriev AP (1966) Mineral formation conditions of the ludwigite-vonsenite series according to experimental data. Zap Vses Mineral Obshchestva 95:272-286 Krogh-Moe J (1962) The crystal structure of lithium diborate, Li20-2B203. Acta Crystallogr 15:190-193 Krogh-Moe J (1967) Structure of pinnoite. Acta Crystallogr 23:500-501 Krosse S (1995) Hochdrucksynthese, Stabilität und Eigenschaften der Borsilikate Dravite und Komerupin sowie Darstellung und Stabilitätsverhalten eines neuen Mg-Al-borates. PhD Dissertation, RuhrUniversität, Bochum Krosse S, Daniels P, Mendenbach O, Schreyer W (1993) "Pseudo-Sinhalit": ein neues Hochdruckborat. Z Kristallogr Suppl Issue S7:117 Kubaschewski O, Evans ELL (1958) Metallurgical Thermochemistry, Pergamon Press, New York Kudoh Y, Takeuchi Y (1973) Ascharite. Crystal Struct Commun 2:595-598 Kühn R, Schaacke I (1955) Vorkommen und analyse der Boracit- und Ericaitkrystalle aus dem Salzhorst von Wathlingen-Hänigsen. Kali und Steinsalz 11:33-42. See also: Am Mineral 41:372, 1956 Kumbasar I (1979) Veatchite-A, a new modification of veatchite. Am Mineral 64:362-366 Kurash VV, Urusov VS, Malysheva TV, Aleksandrov SM (1972) Causes of magnesium and iron ordering in the ludwigite-vonsenite isomorphic series. Geokhimiia 2:148-157 Kurkutova EN (1966) The crystal structure of inderborite. Acta Crystallogr 21:A61 Kurshakova LD (1975) Equilibrium of natural axinite in relation to temperature and oxygen pressure. Ocherki Fiz-Khim Petrol 5:111-120 Kusachi I, Henmi C, Henmi K (1986) An oyelite-bearing vein at Fuka, the town of Bitchu, Okayama Prefecture. Am Mineral 71:230 Kusachi I (1992) New data on mineralogical properties of henmilite. Kobutsugaku Zasshi 21:127-130 Kusachi I, Henmi C, Kobayashi S (1994) Chemical composition of bakerite from Fuka, Okayama Prefecture, Japan. Mineral J 17:111-117 Kusachi I, Henmi C, Kobayashi S (1995) Takedaite, a new mineral from Fuka, Okayama Prefecture, Japan. Mineral Mag 59:549-552 Kuyunko NS, Semenov YuV, Gurevich VM, Kuz'min VI, Topor ND, Gorbunov VE (1984) Experimental determination of thermodynamical properties of dravite tourmaline. Geokhimiia 10:1458-1465 Lacroix A (1912) Sur une nouvelle espèce minérale (manandonite) des pegmatites de Madagascar. Bull Soc Franç Minéral 35:223-226 Lacroix A (1922) Minéralogie de Madagascar, vol 1 and 2. Challamel, Paris Latimer, WM (1938) The oxidation states of the elements and their potentials in aqueous solutions. Prentice-Hall Inc., New York. 352 pp

Anovitz & Hemingway: Thermodynamics of Boron Minerals

213

Levy HA, Lisensky GC (1978) Crystal structures of sodium sulfate decahydrate (glauber's salt) and sodium tetraborate decahydrate (borax). Redetermination by neutron diffraction. Acta Crystallogr B34:35023510 Liang J, Hawthorne FC (1994) Rietveld structure-refinement of muscovite and boron-muscovite using highresolution powder-diffiaction data. GAC/MAC Prog Abstr 19:A66 Liberman A, Gandall WB (1952) Design and construction of a self-calibrating dilatometer for high temperature use. J Am Ceram Soc 35:304-308 Lindemann CL (1916) Om turmalines termiska dilatation. Acta Soc Sci Fennicae 46(6):1-18 Linke WF (1965) Solubilities of Inorganic and Metal Organic Compounds. American Chemical Society (4th ed.), Washington, D.C Lisitsyn AE, Rudnev VV, Gaft AL, Dobrovol'skaya NV, Dara OM, Tkacheva TV (1985a) New data on ludwigite and szaibelyite of the Taiga skam-magnetite deposit (southern Yakutia). Zap Vses Mineral Obshchestva 114:62-73 Lisitsyn AE, Rudnev VV, Yurkina KV (1985b) Suanite and its mineral associations in the Taezhnoe borate-magnetite deposit (South Yakutia). Mineral Zh 7:32 Lonappan MA (1955) Thermal expansion of boric acid. Proc Indian Acad Sci A42:10-21 London D, Zolensky ME, Roedder E (1987) Diomignite: natural Li2B407 from the Tanco pegmatite, Bernic Lake, Manitoba. Can Mineral 25:173-180 London D, Morgan GB VI, Hervig RL (1989) Vapor-undersaturated experiments with macusanite glass + H2O at 200 MPa, and the internal differentiation of rare-element pegmatites. Contrib Mineral Petrol 102:1-17 Lovanova VV (1962) A new borate, halurgite. Dokl Akad Nauk SSSR 143:693-696. See also Fleischer M (1962) New Mineral Names. Am Mineral 47:1217-1218 Lovanova VV, Avrova NP (1964) The new mineral metaborate—natural metaboric acid. Zap Vses Mineral Obshchestva 93:329-334. See also: Fleischer M (1965) New Mineral Names. Am Mineral 50:261-262 Luo S, Lu J, Wang L, Zhu J (1993) Qilianshanite: A new boric carbonate mineral. Acta Mineral Sinica 13(2): 97-101 (in Chinese). See also Jambor JL, Puziewicz J, Roberts AC (1994) New Mineral Names. Am Mineral 79:763-767 Ma Z, Shi N, Shen J, Peng Z (1981) The refinement of the crystal structure of carboborite MgCa 2 [C03]2[B(0H)4]24H 2 0. Bull Mineral 104:578-581 Ma R, Zhang J, Yang F (1986) Hellandite of Quyang, Hebei, China. Dizhi Xuebao 60:68-77 MacDonald DJ, Hawthorne FC, Grice JD (1993) Foitite, [Fe 2 2+ (Al,Fe 3+ )]Al 6 Si60i8(B03)3(0H)4, a new alkali-deficient tourmaline: Description and crystal structure. Am Mineral 78:1299-1303 Machin MP, Suesse P (1974) Serendibite. New member of the aenigmatite structure group. N Jahrb Mineral Mh 435441 Machin MP, Miehe G (1976) Aluminum tetrasilicate [Al(Si04)4]13" tetrahedral pentamers in harkerite. N Jahrb Mineral Mh 5:228-232 Majling J, Figusch V, Hanic F, Wiglasz V, Corba J (1974) Crystal data and thermal expansion of tricalcium borate. Mater Sci Bull 9:1379-1382 Malinko SV (1974) New boron mineral, solongoite. Zap Vses Mineral Obshchestva 103:117-121 Malinko SV, Kuznetsova NN (1973) New find of sakhaite. Zap Vses Mineral Obshchestva 102:164-170 Malinko SV, Pertsev NN (1979) Rhombic and monoclinic kurchatovite (new discoveries and data). Zap Vses Mineral Obshchestva 108:595-599 Malinko SV, Pertsev NN (1983) Clinokurchatovite—a new structural modification of kurchatovite. Zap Vses Mineral Obshchestva 112:483-487 Malinko SV, Stolyarova TI, Shashkin DP (1972) First find of roweite, its paragenesis, and replacement products. Zap Vses Mineral Obshchestva 101:465473 Malinko SV, Shashkin DP, Yurkina KV, Bykov VP (1973) New data on kurchatovite. Zap Vses Mineral Obshchestva 102:696-702 Malinko SV, Shashkin DP, Yurkina KV, Vinogradov BK (1976) New data on minerals of the roweite group. In SI Lebedeva, VI Kuz'min, VV Matias, (eds) Nov Mineral Issled, [Mater Nauchn Konf Mosk Otd Vses Mineral Obshchestva], 1st, Meeting Date 1974, 48-51. Vses Nauchno-Issled Inst Mineral Syr'ya: Moscow Malinko SV, Shashkin DP, Yurkina KV (1977) Federovskite, a new boron mineral, and the isomorphous series roweite-federovskite. Am Mineral 62:173 Malinko SV, Kuznetsova NN, Yurkina KV (1978) Borcarite from Solongo, its parageneses, composition and properties. Tr. Mineral Muz Akad Nauk SSSR 26:94-104 Malinko SV, Stolyarova TI, Yurkina KV, Lisitsyn AE (1978) Composition and properties of kurchatovite and sakhaite in Sayak ore-field skarns. Zap Vses Mineral Obshchestva 107:271-279 Malinko SV, Lisitsyn AE, Naumova IS, Purusova SP (1981) Strontiohilgardite from the Korshunovskoe iron ore deposit. Zap Vses Mineral Obshchestva 110:588-593

214

Anovitz & Hemingway: Thermodynamics

of Boron

Minerals

Malinko SV, Fitsev BP, Kuznetsova NN, Cheikasova LE (1980) Ekaterinite—a new boron mineral. Zap Vses Mineral Obshchestva 109:469-476 Malinko SV, Yamnova NA, Pushcharovskii DYu, Lisitsin AE, Rudnev VV, Yurkina KV (1986) Iron-rich warwickite from the Taiga ore deposit (southern Yakutia). Zap Vses Mineral Obshchestva 115:713-719 Malinovskii YuA, Pobedimskaya EA, Belov NV (1977) Crystal structure of tienshanite. Dokl Akad Nauk SSSR 236:863-865 Malinovskii YuA, Belov NV (1980) Crystal structure of kalborsite. Dokl Akad Nauk SSSR 252:611-615 Malinovskii YuA, Yamnova NA, Belov NV (1981) Refined crystal structure of leucosphenite. Dokl Akad Nauk SSSR 257:1128-1132 Mandarino JA, Rachlin AL, Dunn PJ, Le Page Y, Blade ME, Muiowchick BL, Ramik RA, Falls RB (1990) Redefinition of volkovskite and its description from Sussex, New Brunswick. Can Mineral 28:351-356 Mandarino JA (1992) New minerals recently approved by the commission on new minerals and mineral names International Mineralogical Association. Can Mineral 29:1177-1179 Maras A, Parodi GC, Delia Ventura G, Ohnenstetter D (1995) Vicanite-(Ce): a new Ca-Th-REE borosilicate from the Vico volcanic district (Latium, Italy). Eur J Mineral 7:439-446 Marezio M, Plettinger HA, Zachariasen WH (1963) Refinement of the calcium metaborate structure. Acta Crystallogr 16:390-392 Marshukova NK, Sirina TN, Pavlovskii AB (1968) First find of nordenskioldine in the USSR. Zap Vses Mineral Obshchestva 97:695-698 Mason WP (1950) Piezoelectric crystals and their application to ultrasonics. Van Nostrand, New York Mason RA (1980a) The ordering behavior of reedmergnerite, NaBSi3Û8. Contrib Mineral Petrol 72:329333 Mason RA (1980b) Changes in the crystal morphology of synthetic reedmergnerite (NaBSi30g) during ordering experiments. Mineral Mag 43:905-908 Mattigod SV (1983) A method for estimating the standard free energy of formation of borate minerals. Soil Sei Soc Am J 47:654-655 Mattigod SV, Sposito G (1978) Improved method for estimating the standard free energies (AGf,298.15) of smectites. Geochim Cosmochim Acta 42:1753-1762 . Matsubara S (1980) The crystal structure of nagashimalite, Ba4(V +,Ti)4[(0,0H)2lCllSi8B2027]. Mineral J 10:131-142 Matsubara S, Kato A (1980) Nagashimalite, Ba4(V ,Ti)4[(0,0H)2ICllSigB2027], a new mineral from the Mogurazawa mine, Gumma Prefecture, Japan. Mineral J 10:122-30 Mazdab FK, Anovitz LM, Hemingway BS, Robie RA, Navrotsky A (1992) Thermodynamic properties of some boron-bearing minerals. Geol Soc Am Abstr with Prog, 1992 Ann Meeting, Cincinnati, Ohio, A258 Mazdab FK, Anovitz LM, Hemingway BS (1994) Heat capacities and derived thermodynamic functions for several borosilicate and borate minerals. MS Thesis, Univ. of Arizona, 74 p Mazzi F, Rossi G (1980) The crystal structure of taramellite. Am Mineral 65:123-128 Meixner H (1952) Einige Boratminerale (Colemanit und Tertschit, ein neues Mineral) aus der Türkei. Fort Mineral 31:39-42 Meixner H (1953) Mineralogische Beobachtungen an Colemanit, Inyoit, Meyerhofferit, Tertschit und Ulexit aus neuen türkischen Boratlagerstätten. Heidelberger Beit zur Mineral und Petrol 3:445-455 Mellini M, Merlino S (1977) Hellandite: a new type of silicoborate chain. Am Mineral 62:89-99 Menchetti S, Sabelli C, Trosti-Ferroni R (1982) Probertite, CaNa[B507(0H)4]-3H20: a refinement Acta Crystallogr B38:3072-3075 Menzel H, Shultz H (1940) D a Kernit (Rasorit) Na 2 B 4 074H 2 0; Zur Kenntnis der Borsäuren und borsauren Alkalisalze. Zeit Anorg All Chem 245:157-220 Mer'kov, AN Nedorezova AP, Busson IV, Kondrat'eva VV, Kul'chitskaya EA, Latysheva LG, Men'shikov YuP (1975) Leucosphenite as the first boron mineral from the Lovozero massif. Zap Vses Mineral Obshchestva 104:343-346 Merlino S, Sartori, F (1969) Crystal structure of Iarderellite, NH4B507(0H)2 H 2 0 Acta Crystallogr B25:2264-2270 Merlino S, Sartori F (1970) Santite, a new mineral phase from Larderello, Tuscany. Contrib Mineral Petrol 27:159-165 Merlino S, Sartori F (1971) Ammonioborite: new borate polyion and its structure. Science 171:377-379 Merlino S, Sartori F (1971) Santite, a new mineral phase from Larderello, Tuscany. Am Mineral 56:636 Merlino S, Sartori F (1972) Crystal structure of sborgite, NaB506(0H) 4 -3H 2 0. Acta Crystallogr B28:3559-3567 Michel-Levy M (1953) Reproduction artificielle de minéraux qui apparaissent dans le métamorphisme dé contract du granite. Bull Soc Fr Minéral Crist 76:236-293

Anovitz & Hemingway: Thermodynamics

of Boron Minerals

215

Milton C, Hildebrand FA, Sherwood AM (1953) The identity of tinzenite with manganoan axinite. Am Mineral 38:1148-1158 Milton C, Chao ECT, Axelrod J, Grimaldi FS (1960) Reedmergnerite, NaBSi3Os, the boron analogue of albite, from the Green River formation, Utah. Am Mineral 45:188-199 Minakawa T, Inaba S, Noto S (1986) Oyelite from Suisho-dani, Ise, Mie Prefecture [Japan]. Ganseki Kobutsu Kosho Gakkaishi 81:138-142 Miyawaki R, Nakai I, Nagashima K (1985) Structure of homilite, Ca2.00CFe0.90Mn0.03) B2.00Si2.00O9.86(OH)0.i4. Acta Crystallogr C41:13-15 Miyawaki R, Nakai I, Nagashima K, Okamoto A, Isobe T (1987) The first occurrences of hingganite, hellandite and wodginite in Japan. Kobutsugaku Zasshi 18:17-30 Mokeeva VI, Aleksandrov SM (1969) Magnesium and iron distribution in the structure of borates of the ludwigite -vonsenite series. Geokhimiia 4:428-435 Mokeeva VI (1969) Refinement of the structure of ludwigite (Mgi.85Fe2+o.is)(Fe3+o.6oAlo.4o)BC>302 and the distributions of Mg 2+ and Fe 2+ in cation sites. Geokhimiia 8:975-979 Moore AE, Taylor HFW (1970) Crystal structure of ettringite. Acta Crystallogr B26:386-393 Moore PB, Ghose S (1971) Novel face-sharing octahedral trimer in the crystal structure of seamanite. Am Mineral 56:1527-1538 Moore PB, Araki T (1972) Johachidolite, CaAl[B3C>7], a borate with very dense atomic structure. Nature (London), Phys Sci 240:63-65 Moore PB, Araki T (1972) Wightmanite, Mg5(0)(0H)5[B03]-nH20, a natural drainpipe. Nature 239:25-26 Moore PB, Araki T (1974) Pinakiolite, Mg2Mn3+02[B03]. Warwickite, Mg(Mg0.5Ti0.5)O[BC>3]. Wightmanite, Mg5(0)(0H)5[B03]-nH20 Crystal chemistry of complex 3 angstrom wallpaper structures. Am Mineral 59:985-1004 Moore PB, Araki T (1974) Roweite, Ca2Mn2+2(0H)4[B407(0H)2]. Its atomic arrangement. Am Mineral 59:60-65 Moore, PB, Araki T (1976) Painite, CaZrB[AI9O18]: Its crystal structure and relation to jeremejevite, B5[03Al6(0H)30i5], and fluoborite, B3[Mg9(F,0H)909]. Am Mineral 61:88-94 Moore PB, Araki T (1978) Dumortierite, Si3B[AIe 75^0 25O17 25(OH)o 75]: a detailed structure analysis. N Jahrb Mineral Abh 132:231-241 Moore PB, Araki T (1979) Konierupine: a detailed crystal-chemical study. N Jahrb Mineral Abh 134:317336 Moore PB, Araki T, Ghose S (1982) Hyalotekite, a complex lead borosilicate: its crystal structure and the lone-pair effect of lead(D). Am Mineral 67:1012-1020 Moore PB, Gupta PKS, Schlemper EO (1989) Kornerupine: Chemical crystallography, comparative crystallography, and its cation relation to olivine and to Ni2ln intermetallic. Am Mineral 74:642-655 Moore JM, Waters DJ, Niven ML (1990) Werdingite, a new borosilicate mineral from the granulite fades of the western Namaqualand metamorphic complex, South Africa. Am Mineral 75:415-420 Morey GW, Ingerson E (1937) The melting of danburite: a study of liquid immiscibility in the system Ca0-B203-Si02. Am Mineral 22:37-47 Morgan GB VI, London D (1989) Experimental reactions of amphibolite with boron-bearing aqueous fluids at 200 MPa: implications for tourmaline stability and partial melting in mafic rocks. Contrib Mineral Petrol 102:281-297 Morimoto N (1956) The crystal structure of borax. Mineral J Japan 2:1 Mrose ME, Fleischer M (1963) The probably identity of magnesioborite with suanite. Am Mineral 48:915 Mrose ME, Rose HJ Jr. (1961) Behierite, (Ta,Nb)B04, a new mineral from Manjaka, Madagascar. Geol Soc Am Abstr 1961 Ann Meeting:lllA Mrose ME, Schaller WT (1965) The identity of patemoite with kaliborite. Am Mineral 50:1079-1083 Murdoch J (1962) Bakerite crystals. Am Mineral 47:919-923 Murnaghan FD (1967) Finite Deformation of an Elastic Solid. New York. Dover Nakai I (1986) Henmilite, Ca2Cu(OH)4[B(OH)4]2, a new mineral from Fuka, Okayama Prefecture, Japan. II. Crystal structure. Am Mineral 71:1236-1239 Nakai I, Okada H, Masutomi K, Koyama E, Nagashima K (1986) Henmilite, Ca2Cu(OH)4[B(OH)4]2, a new mineral from Fuka, Okayama Prefecture, Japan. I Occurrence and description. Am Mineral 71:1234-1236 Nakamura T, Kubo T (1992) Tourmaline group crystals reaction with water. Ferroelectrics 137:13-31 Naumov GB, Ryzhenko BN, Khodakovsky IL (1971) Handbook of thermodynamic data. Atomizdat, Moscow, 327 p Nekrasov IY, Kashirtseva GA (1975) Hydrothermal synthesis of stanniferous axinite. Dokl Acad Sci USSR Earth Sci Sect 222:121-123 Nekrasov IYa, Nekrasova RA (1995) A new borosilicate of thorium from the Tommot rare-earth deposit. Dokl Akad Nauk SSSR 342:361-364. See also Jambor JL, Pertsev NN, Roberts AC (1996) New Mineral Names. Am Mineral 81:766-770

216

Anovitz & Hemingway: Thermodynamics of Boron Minerals

Nickel EH (1993) Standardization of polytype suffixes. Am Mineral 78:1313. Nies NP, Hulbert RW (1967) Solubility isotherms in the system sodium oxide-boric oxide-water. J Chem Eng Data 12:303-313 Nikolaichuk GV, Nekrasov IYa, Shashkin DP (1970) Hydrothetmal synthesis of kurchatovite in the calcium oxide-magnesium oxide-boron oxide-water system. Dokl Akad Nauk SSSR 190:936-939 Niven ML, Waters DJ, Moore JM (1991) The crystal structure of weidingite (Mg,Fe)2Ali2(Al,Fe)2Si4(B,Al)4037, and its relationship to sillimanite, mullite, and grandidierite. Am Mineral 76:246-256 Norrestam R, Bovin JO (1987) The crystal structure of takeuchiite, Mgi.7iMni.29B05. A combined single crystal x-ray and HRTEM study. Z Kristallogr 181:135-149 Norrestam R, Dahl S, Bovin JO (1989) The crystal structure of magnesium-aluminium ludwigite, Mg2 i i A l 0 3i Feo 53T10 05Sb 0 01BO5, a combined single crystal X-ray and HREM study. Z Kristallogr ¿87:201-211 Norrestam R, Hansen S (1990) Structural investigation of an antimony-rich pinakiolite, Mg1.90Mn0.91Sb0.19 O2BO3, from Ldngban, Sweden. Z Kristallogr 191:105-116 Novoselova LN, Sosedko TA (1991) Authigenic tourmaline from carbonate rocks of the Bateneuv Ridge (Altai-Sayan folded region). Zap Vses Mineral Obshchestva 120:79-83 Nriagu JO (1975) Thermochemical approximations for clay minerals. Am Mineral 60:834-839 Nuber B, Schmetzer K (1979) Die Gitterposition des Cr 3+ in Turmalin: Strukturverfeinerung eines Cr-reichen Mg-Al Turmalins. N Jahrb Mineral Abh 137:184-197 Nuber B, Schmetzer K (1981) Structural refinement of liddicoatite. N Jahrb Mineral Mh 5:215-219 Olesch M, Seifert F (1976) Synthesis, powder data and lattice constants of grandidierite, (Mg,Fe)Al3BSi09. N Jahrb Mineral Mh 1976:513-518 Ono A (1981) Synthesis of dumortierite in the system Al20j-Si02-B203-H20. J Japan Assoc Mineral Petr Econ Geol 76:21-25 Ostrovskaya IV (1969) Crystal structure of sakhaite and haikerite. Trudy Mineral Muz Akad Nauk SSSR 19:197-201 Ostrovskaya IV, Pertsev NN, Nikitina IB (1966) Sakhaite, a new caibonate-borate of calcium and magnesium. Zap Vses Mineral Obshchestva 95:193-202 Ostrovskaya IV, Nikitina IB (1969) Formula of preobrazhenskite. Tr Mineral Muz Akad Nauk SSSR 19:206-209 Ozaki M (1972) Chemical composition and occurrence of axinite. Kumamoto J Sci Ser B Sect 1 9:1-34 Pabst A, Milton C (1972) Leucosphenite, and its occurrence in the Green River formation of Utah and Wyoming. Am Mineral 57:1801-1822 Palache C, Foshag WF (1938) Antofagasite and bandylite, two new copper minerals from Chile. Am Mineral 23:85-90 Palache C, Berman H, Frondel C (1951) Dana's System of Mineralogy, 7th ed. Wiley, New York Pankratz LB, Fenante MJ (1971) Thermodynamic properties of four crystalline sodium borates. US Bur Mines Rpt Inves 7551. 8 p Pautov LA, Bellakovskiy DI, Skala R, Sokolova YeV, Ignatenko KI, Mokhov AV (1992) Vistepite Mn5SnB2Si5O20—a new borosilicate of manganese and tin. Zap Vses Mineral Obshchestva 121(4):107-112 Peacor DR, Dunn PJ, Duggan M (1983) Stunnanite, a ferric iron, boron analog of ettringite. Can Mineral 21:705-709 Pertlik F, Dunn PJ (1989) Crystal structure of wiserite. Am Mineral 74:1351-1354 Pertsev NN (1961) Haikerite and kotoite in skarns of Polar Yakutia. Geologiya i Geofizika 1961(7):102105 Pertsev NN, Ostravskaya IV, Nikitina IB (1965) The new mineral borcarite. Zap Vses Mineral Obshchestva 94:180-186. See also Fleischer M (1965) New Mineral Names. Am Mineral 50:2097 Pertsev NN, Malinko SV, Vakhrushev VA, Fitsev BP, Sokolova EV, Nikitina IB (1980) Shabynite, a new magnesium borate chloride hydrate. Zap Vses Mineral Obshchestva 109:569-573 Pertsev NN, Malinko SV, Vakhrushev VA, Fitsev BP, Sokolova EV, Nikitina IB (1981) Shabynite, a new hydrated borate-chloride of magnesium. Am Mineral 66:1101 Petersen OV, Johnsen O, Jensen A (1980) Giant crystals of kornerupine. Mineral Rec 11:93-96 Philips MW, Gibbs GV, Ribbe PH (1974) The crystal structure of danburite: a comparison with anorthite, albite, and reedmergnerite. Am Mineral 59:79-85 Portnov AM, Siderenko GA, Dubinchuk VT, Kuznetsova NN, Ziborova TA (1969) Melanocerite from the north Baikal region. Dokl Akad Nauk Earth Sci Sect 185:107-109 Povondra P, Cech F (1988) Lattice parameters of tourmalines of the dravite-schorl series and their relation to chemical composition. Acta Univ Carol Geol Volume Date 1985 3:235-245 Powell DR, Gaines DF, Zerella PJ, Smith RA (1991) Refinement of the structure of tincalconite. Acta Crystallogr C47:2279-2282

Anovitz & Hemingway: Thermodynamics

of Boron Minerals

217

Prewitt CT, Buerger MJ (1961) The crystal structure of cahnite, Ca2BAs04(0H)4_ Am Mineral 46:10771085 Pring A, Din VK, Jefferson DA, Thomas JM (1986) The crystal chemistry of ihodizite: a re-examination. Mineral Mag 50:163-172 Pryce MW (1971) Holtite, a new mineral allied to dumortierite. Mineral Mag 38:21-25 Raade G, Mladeck MH, Din VK, Criddle AJ, Stanley CJ (1988) Blatterite, a new antimony-bearing manganese ion(2+)-manganese ion(3+) member of the pinakiolite group, from Nordmark, Sweden. N Jahrb Mineral Mh 3:121-136 Rachlin AL, Mandarino JA, Murowchick BL, Ramik RA, Dunn PJ, Bade ME (1986) Mineralogy of hilgardite-4Af from evaporites in New Brunswick. Can Mineral 24:689-693 Ramdohr P (1934) NordenskiCtldin in einer Zinnerzlagerstatte. N Jahrb Mineral Geol Palaontol Abh Bd Abt A 68:288-295 Randmets R (1960) Orthopinakiolite, a new modification of Mg3Mn2+ Mn 3+ 2B2O10 from LSngban, Sweden. Aikiv Mineral Geol 2(42):551-555 Ranorosoa N, Fontan F, Fransolet AM (1989) Rediscovery of manandonite in the Sahatany Valley, Madagascar. Eur J Mineral 1:633-638 Rastsvetaeva RK, Andrianov VI, Genkina EA, Sokolova TN, Kashaev AA (1992) Crystal structure of volkovskite, KCa4B22O32(OH)i0Cl-4H2O, from the Nepa deposit in eastern Siberia. Kristallografiya 37:326-333 Rastsvetaeva RK, Khomyakov AP, Sokolova TN (1993) Crntal structure of calcium-rich p-Witchite and its position in a series of complex borates with BsOsOH*" radicals. Kristallografiya 38:86-93 Rastsvetayeva RK, Pushcharovskiy DYu, Pekov IV, Voloshin AV (1996) The crystal structure of "calcybeboiosilite" and its position in the isomorphic series datolite-gadolinite. Kristallographiya 41(2):235-239 Rau VG, Kuikutova EN (1971) Determination of the symmetry center of seamanite according to the high moments of the intensities. Uch Zap Vladimirskii Gos Pedagog Inst Ser "Fiz" 32:23-24 Rau VG, Rumanova IM, Kurkutova EN (1972) Determination of the positions of light atoms in a seamanite structure. Uch Zap Vladimirskii Gos Pedagog Inst 40:3-8 Raup OB, Gude AJ ID, Dwomik E J, Cuttitta F, Rose HJ Jr. (1968) Braitschite, a new hydrous calcium rare-earth borate mineral from the Paradox Basin, Grand County, Utah. Am Mineral 53:1081-1095 Razmanova ZP, Rumanova IM, Belov NV (1969) Crystal structure of kurnakovite, Mg2Bg0n l5H20 = 2Mg[B303(0H) 5 ]-5H 2 0. Dokl Akad Nauk SSSR 189:1003-1006 Reif J, Paulis P, Sane I (1991) Ferroaxinite from the Alpine vein in amphibolite at Libodrice (near Kolin), central Bohemia, Czechoslovakia. Cas Mineral Geol 36:243-246 Roberts AC, Stirling JAR, Grice JD, Burns PC, Roulston BV, Curtis JD, Jambor JL (1993) Pringleite and ruitenbergite, polymorphs of Ca9B26034(0H)24Cl4«13H20, two new mineral species from Sussex, New Brunswick. Can Mineral 31:795-800 Roberts WL, Campbell TJ, Rapp GR Jr. (1990) Encyclopedia of Minerals, 2nd ed. Van Nostrand Reinhold., New York Robie RA, Hemingway BS, Fisher JR (1979) Thermodynamic properties of minerals and related substances at 298.15 K and 1 bar (103 pascals) pressure and at higher temperatures. US Geol Surv Bull 1452 Robie RA, Hemingway BS (1995) Thermodynamic properties of minerals and related substances at 298.15 K and 1 bar (103 pascals) pressure and at higher temperatures. US Geol Surv Bull 2131 Robins CR, Yoder, HS Jr. (1962) Stability relations of dravite, a tourmaline. Carnegie Inst Wash Yearbook 61:106-107 Robinson GR, Jr., Haas, JL Jr. (1983) Heat capacity, relative enthalpy, and calorimetric entropy of silicate minerals: An empirical method of prediction. Am Mineral 68:541-553 Rodellas C, Garcia-Blanco S, Vegas A (1983) Crystal structure refinement of jeremejevite (AI6B5F3O15). Z Kristallogr 165:255-260 Rosenberg PE, Foit FF Jr. (1975) Alkali-free tourmalines in the system Mg0-Al203-Si02-H20-B203. Geol Soc Am Abstr with Prog 1250-1251 Rosenberg PE, Foit FF Jr. (1979) Synthesis and characterization of alkali-free tourmaline. Am Mineral 64:180-186 Rosenberg PE, Foit FF Jr., Ekambaram V (1986) Synthesis and characterization of tourmaline in the system Na20-Al203-Si02-B2C>3-H20. Am Mineral 71:971-976 Ross M (1964) Crystal chemistry of beryllium. US Geol Surv Prof Pap 468 Ross V, Edwards JO (1959) Tetrahedral boron in teepleite and bandylite. Am Mineral 44:875-877 Rumanova IM, Kurbanov KhM, Belov NV (1965) Determination of the crystal structure of probertite CaNa[B 5 0 7 (0H)4]-3H 2 0. Kristallografiya 10:601-613 Rumanova IM, Gandymov O (1971) Crystal structure of natural strontium borate, paraveatchite Sr2[B508(0H)]2B(0H)3-H20. Kristallografiya 16:99-106

218

Anovitz & Hemingway: Thermodynamics

of Boron

Minerals

Rumanova IM, Kurkutova EN (1971) Determination of crystal structures of inderborite CaMg[B303(0H)5]2-6H20 by using the method of partial synthesis of weighted Fourier projections. Probl. Kristallologii 102-15 From: Ref Zh Khim 1972, Abstr. No. 9B476 Rumanova IM, Razmanova ZP, Belov NV (1971) Crystal structure of preobrazhenskite 3Mg0-5.5B 2 0 3 -4.5H20 = Mg3[B u 0i 4 (0H) 8 H02]. Dokl Akad Nauk SSSR 199:592-595 Rumanova IM, Iorysh ZI, Belov NV (1977) Crystal structure of triclinic hilgardite Ca2[B509]Cl H 2 0 = Ca 2 [B3 t B 2 0 9 ]Cl-H 2 0. Dokl Akad Nauk SSSR 236:91-94 Rumanova IM, Genkina EA (1981) Precise structure of inyoite Ca(H20)3[B2B03(0H)5] H2O Study of hydrogen bonds. Latv PSR Zinat Akad Vestis Kim Ser 6:643-653 Rumyantseva EV (1983) The new mineral chromdravite from Karelia. Zap Vses Mineral Obshchestva 112:222-226 Sabelli C, Stoppioni A (1978) Refinement of the crystal structure of hydroboracite. Can Mineral 16:75-80 Sadanaga R, Nishimura T, Watanabe T (1965) Structure of jimboite and relation with the structure of kotoite. Mineral J 4:380-388 Saiko IG, Kononova GN, Burlyaev VV, Tavrovskaya AYa (1985a) Study of thermal transformations of ulexite NaCa[B 5 06(0H) 6 ]-5H 2 0. Zh Neorg Khim 30:1149-1153 Saiko IG, Kononova GN, Petrov KI, Agafonov VN (1982a) Thermal transformations of inderborite. Zh Neorg Khim 27:2433-2436 Saiko IG, Kononova GN, Petrov KI, Tavrovskaya AYa (1982b) Thermal transformations of kaliborite. Zh Neorg Khim 27:335-338 Saiko IG, Kononova GN, Zakgeim AYu (1985b) Study of the dehydration kinetics of natural borates. Zh Prikl Khim (Leningrad) 58:654-656 Sanero E, Gottardi G (1968) Nomenclature and crystal chemistry of axinites. Am Mineral 53:1407-1411 Schaller WT, Mrose ME (1960) The naming of the hydrous magnesium borate minerals from Boron, California—a preliminary note. Am Mineral 454:732-734 Schaller WT, Vlisidis AC, Mrose ME (1965) McCallisterite, 2Mg0-6B2C>3-15H20, a new hydrous magnesium borate mineral from the Death Valley region, Inyo County, California. Am Mineral 50:629-640 Schmetzer K, Nuber B, Abraham K (1979) Zur Kristallchemie Magnesium-reicher Tunnaline. N Jahrb. Mineral Abh 136:93-112 Schreyer W (1988) Esperimental studies on metamorphism of crustal rocks under mantle pressures. Mineral Mag 52:1-26 Schreyer W, Abraham K (1976) Natural boron-free kornerupine and its breakdown products in a sapphirine rock of the Limpopo Belt, southern Africa. Contrib Mineral Petrol 54:109-126 Schreyer W, Seifert F (1969) High pressure phases in the system Mg0-Al 2 03-Si0 2 -B203-H 2 0. Am J Sei 267A:407-443 Sedlacek P, Domberger-Schiff K (1971) An OD-disondered modification of the calcium monoborate dihydrate Ca[B(0H)4]2-2H20. Acta Crystallogr B27:1532-1541 Seifert F (1975) Boron-free kornerupine: a high pressure phase. Am J Sei 275:57-87 Semenov YuV, Malinko SV, Kiseleva IA, Khodakovsky EL (1987) The formation conditions of endogenous calcium borates and borosilicates. Geokhimiia 8:1182-1190 Semenov Yel, Dusmatov VD, Samsonova NS (1963) Yttrium-beryllium minerals of the datolite group. Kristallografiya 8(4):677-679 See also Fleischer, M (1964) New Mineral Names. Am Mineral 49:443-444) Sen Gupta PK, Swihart GH, Dimitrijevic R, Hossain MB (1991) The crystal structure of luneburgite, Mg,(H20)6[B2(0H)l5(P04)2]. Am Mineral 76:1400-1407 Shaller WT (1933) Ammonioborite, a new mineral. Am Mineral 18:480-492 Shashkin DP, Simonov MA, Chemova NI, Malinko SV, Stolyarova TI, Belov NV (1968) A new native borate, vimsite. Dokl Akad Nauk SSSR 182:1402-1405 Shashkin DP, Simonov MA, Belov NV (1969) Crystal stnicture of uralborite, Ca2[B4C>4(OH)8]. Dokl Akad Nauk SSSR 189:532-535 Shashkin DP, Simonov MA, Belov NV (1970) Crystal structure of calciborite CaB2C>4=Ca2[B03BO]2. Dokl Akad Nauk SSSR 195:345-348 Shashkin DP, Simonov MA, Belov NV (1971) X-ray diffraction study of natural calcium metaborates. Kristallografiya 16:231-235 Shen J, Moore PB (1984) Crystal structure of cappelenite, Ba(Y,RE)6[Si3B6C>24]F2: a silicoborate sheet structure. Am Mineral 69:190-195 Shipovalov YuV, Avrova NP (1971) X-Ray studies of new borates, aldzhanite and chelkarite. Issled Obi Khim Fiz Metod Anal Mineral Syr'ya 176-9 From: Ref Zh Khim 1972, Abstr No 13B445 Shumyatskaya NG, Voronkov AA, Pyatenko YuA (1968) The atomic structure of leucosphenite Na8B2Ti4O4[B4Si20O56]. Kristallografiya 13:165-167

Anovitz & Hemingway: Thermodynamics of Boron Minerals

219

Shumyatskaya NG, Voronkov AA, Belov NV (1971) X-ray diffraction study of leucosphenite. Kristallografiya 16:492^99 Sieh S-T, Chen M-P (1963) A study of kumakovite crystals. Sei Sinica 12(4):575 Simonov MA, Shashkin DP, Egorov-Tismenko YuK (1971) X-ray structural analysis of endogenic calcium metaborates. Probl. Kristallologü 84-101 From: Ref Zh Khim 1972, Abstr. No. 9B475 Simonov MA, Kazanskaya EV, Egorov-Tismenko YuK, Zhelezin EP, Belov NV (1976) Redetermination of the crystal structure of frolovite (Ca[B(OH) 4 ] 2 ). Dokl Akad Nauk SSSR 230:91-94 Simonov MA, Yamnova NA, Kazanskaya EV, Egorov-Tismenko YuK, Belov NV (1976) Crystal structure of a new natural calcium borate hexahydroborite C a B 2 0 4 - 6 H 2 0 = Ca[B(OH) 4 ]2-2H2Ö. Dokl Akad Nauk SSSR 228:1337-1340 Simonov MA, Egorov-Tismenko YuK, Belov NV (1976) Refined crystal structure of vimsite, C a [ B 2 0 2 ( 0 H ) 4 ] . Kristallografiya 21:592-594 Simonov MA, Egorov-Tismenko YuK, Belov NV (1977) More precise definition of the crystal structure of uralborite, C a 2 [ B 4 0 4 ( 0 H ) 8 ] . Dokl. Akad Nauk SSSR 234:822-825 Simonov MA, Malinko SV, Belov NV, Kazanskaya EV, Egorov-Tismenko YuK, Fedorenko MB, Belokoneva EL, Yamnova NA, Kuznetsova N N (1977) The new-mineral hexahydroborite C a [ B ( 0 H ) 4 ] 2 - 2 H 2 0 . Zap Vses Mineral Obshchestva 106:691-697 Simonov MA, Egorov-Tismenko YuK, Kazanskaya EV, Belokoneva EL, Belov NV (1978) Hydrogen bonds in the crystal structure of nifontovite Ca3[B303(0H)ö] 2 - 2 H 2 0 . Dokl Akad Nauk SSSR 239:326-329 Simonov MA, Egorov-Tismenko YuK, Belov NV (1979) Hydrogen bonds in crystal structures of endogene calcium metaborates. Kristallografiya 24:439-445 Simonov MA, Egorov-Tismenko YuK, Yamnova NA, Belokoneva EL, Belov NV (1980) Crystal structure of natural monoclinic kurchatovite Ca 2 (Mgo 86Feo 14)( M g092 F^O 08)[B2Os] 2 . Dokl Akad Nauk SSSR 251:1125-1128 Sims JR (1966) Retention of boron by layer silicates, sesquioxides and soil minerals. PhD Dissertation, Univ. of California, Riverside Skvortsov VG, Tsekhanskii RS, Gavrilov AM (1976) The sodium metaborate-sodium chloride-water system at 20°C. Russ J Inorg Chem 21:315-316 Smith GS, Rindone GE (1961) High-temperature energy relations in the alkali borates: Binary alkali borate compounds and their glasses. J Am Ceram Soc 44:72-78 Sokolov PB, Groskaya MG, Gordienko VV, Petrova MG, Kretser YuL, Frank-Kamenitskii VA (1986) Olenite, Nai_ x Al3Al6B3Si60 2 7(0,0H) 4 , a new high-alumina mineral from the tourmaline group. Zap Vses Mineral Obshchestva 115:119-123 Sokolova EV, Egorov-Tismenko YuK, Kargal'tsev SV, Klyakhin VA, Umsov VS (1987) Refinement of the crystal structure of synthetic fluorian jeremejevite Alg[B03]5F3. Vestn Mosk Univ Ser 4 Geol 3:82-84 Sokolova YeV, Khomyakov AP (1991) The structure of a new natural Na-borosilicate with a framework of the tridymite type. Doklady Akad Nauk SSSR 319(4):879-883 (in Russian), See also Jambor JL, Grew ES (1993) New Mineral Names. Am Mineral 78:233-238 Speer JA (1982) Zircon. In PH Ribbe (ed) Orthosilicates, Mineralogical Society of America Reviews in Mineralogy, 5, 2nd edition, Ch 3:67-112 Sposito G (1986) The polymer model of thermochemical clay stability. Clays and Clay Minerals 34:198203 Stachowiak A (1995) Jeremejevit in Experiment und Natur. Diplomarbeit, Ruhr-Universität, Bochum Stephenson DA, Moore PB (1968) Crystal structure of grandidierite, (Mg,Fe)Al3SiB09. Acta Crystallogr B24:1518-1522 Stoch L, Waclawska I (1990) Thermal decomposition of ulexite. J Therm Anal 36:2045-2054 Stoch L, Waclawska I (1993a) Thermal decomposition of hydrated borates. Part 1. Thermal decomposition of pandermite C a 2 B s 0 g ( 0 H ) 3 - 2 H 2 0 . Thermochim Acta 215:255-264 Stoch L, Waclawska, I (1993b) Thermal decomposition of hydrated borates. Part 2. Thermal decomposition of kaliborite HKMg 2 Bi 2 Oi6(OH)i0-4H2O. Thermochim Acta 215:265-271 Stoch L, Waclawska I (1993c) Thermal decomposition of hydrated borates. Part 3. Structure mechanism of thermal decomposition of borates. Thermochim Acta 215:273-279 Strunz, H. (1957) Bor und Beryllium in Phyllosilikaten. Rendiconti della Societa Italiana di Mineralogia e Petrologia 13:372 Stubican V, Roy R (1962) Boron substitution in synthetic micas and clays. Am Mineral 47:1166-1173 Sueno S, Clark JR, Papike JJ, Konnert JA (1973) Crystal-structure refinement of cubic boracite. Am Mineral 58:691-697 Swinnea JS, Steinfink H, Rendon-DiazMiron LE, Enciso de la Vega S (1981) The crystal structure of a Mexican axinite. Am Mineral 66:428-431

220

Anovitz & Hemingway: Thermodynamics

of Boron

Minerals

Swinnea JS, Steinfink H (1983) Crystal structure and Mossbauer spectrum of vonsenite, 2Fe0-FeB03- Am Mineral 68:827-832 Switzer G, Clark RS Jr, Sinkankas J, Worthing HW (1965) Fluorine in hambergite. Am Mineral 50:85-95 Taguchi S, Parmentier PP, Yamasaki T (1981) Sassolite sublimated in a steam well at the Yunotani Geothermal Field, Aso Caldera, Kyushu. Mineral J 10:338-343 Takiuchi Y (1952) The crystal structure of magnesium pyroborate. Acta Crystallogr 5:574-581 TakSuchi Y, Ozawa T, Ito T, Araki T, Zoltai T, Finney JJ (1974) Boron silicon oxide (B2Sig03o) groups of tetrahedra in axinite and comments on the deformation of silicon tetrahedra in silicates. Z Kristallogr 140:289-312 Takduchi Y, Kudoh Y (1975) Szaibelyite. Crystal structure, pseudosymmetry, and polymorphism. Am Mineral 60:273-279 Takduchi Y, Haga N, Kato T, Miura Y (1978) Orthopinakiolite, Me2.9502[B03]: its crystal structure and relationship to pinakiolite, Me 2 9502(603] [Me = magnesium, manganese(2+), manganese(3+)]. Can Mineral 16:475-85 Tak6uchi Y, Kogure T (1992) The structure type of ludwigite. Z Kristallogr 200:161-167 Tardy Y (1979) Relationships among Gibbs energies of formation of compounds. Am J Sci 279:217-227 Tardy Y, Duplay J (1992) A method of estimating the Gibbs free energies of formation of hydrated and dehydrated clay minerals. Geochim Cosmochim Acta 56:3007-3012 Tardy Y, Garrels RM (1974) A method of estimating the Gibbs energies of formation of layer silicates. Geochim Cosmochim Acta 38:1101-1116 Tardy Y, Garrels RM (1976) Prediction of Gibbs energies of formation—I Relationships among Gibbs energies of formation of hydroxides, oxides and aqueous ions. Geochim Cosmochim Acta 40:10511056 Tardy Y, Garrels RM (1977) Prediction of Gibbs energies of formation—II Monovalent and divalent metal silicates. Geochim Cosmochim Acta 41:87-92 Tatli A (1985) Zero-field elastic constants of uvite. J Phys Chem Solids 46:1015-1018 Tatli A, Ozkan H (1987) Variation of the elastic constants of tourmaline with chemical composition. Phys Chem Mineral 14:172-176 Taxer KJ, Buerger MJ (1967) Crystal structure of rhodizite. Z Kristallogr 125:423^136 Taylor HFW (1973) Crystal structures of some double hydroxide minerals. Mineral Mag 39:377-389 Taylor AM, Terrell BC (1967) Synthetic tourmalines containing elements of the first transition series. J Crystal Growth 1:238-244 Teeple JE (1929) The industrial development of Searles Lake brines, with equilibrium data. Am Chem Soc Monograph Ser 49, The Chemical Catalog Co, Inc, New York t'Hart J (1979) On the structural morphology of two olivine—type minerals: chrysoberyl and sinhalite. N Jahrb Mineral Abh 134:117-146 Timper U, Heller G, Shakibaie-Moghadam M (1990) Sborgite and beta-sborgite—a second synthetic modification of Na[B506(0H)4]-3H20. Z Naturforsch B Chem Sci 45:1155-1166 Tippe A, Hamilton WC (1971) A neutron-diffraction study of the ferric tourmaline, buergerite. Am Mineral 56:101-113 Tissot P, Painot J, Rivera JP, Schmid H (1982) Thermal decomposition of some boracites. Thermochim Acta 56:359-369 Tkachev KV, Leont'eva IA (1989) Thermal analysis of inorganic borates. Latv PSR Zinat Akad Vestis Kim Ser 1:30-32 Tomisaka T (1968) Syntheses of some end-members of the tourmaline group. Mineral J (Tokyo) 5:355-364 Tung C-H (1974) X-ray powder photographic indexing and cell constant determination of hungchaoite. Ti Ch'iu Hua Hsueh 3:204-212 Ulbrich HH, Ghose D (1973) A note on the chemical composition of the silicoborate garrelsite. Schweiz Mineral Petrogr Mitt 53:199-202 Van Derveer DG, Swihart GH, Sen Gupta PK, Grew ES (1993) Cation occupancies in serendibite: a crystal structure study. Am Mineral 78:195-203 Vasil'eva ZV, Sveshnikova EV (1972) New data on melanocerite. Tr Mineral Muz Akad Nauk SSSR 21:12-16 Vasilkova NN (1963) A new calcium borate, siberskite. Am Mineral 48:433 Vieillard Ph, Tardy Y (1988) Estimation on enthalpies of formation of minerals based on their refined crystal structures. Am J Sci 288:997-1040 Vlisdis AC, Schaller WT (1974) The identity of paigeite with vonsenite, and chemical analyses of vonsenite, ludwigite, and hulsite. N Jahrb Mineral Mh 3-4:95-105 Voigt W (1890) Bestimmung der Elasticitatsconstanten des brasilianischen Turmalines. Annales der Physik undChemie 41:712-717

Anovitz & Hemingway: Thermodynamics

of Boron Minerals

221

Voloshin AV, Gordienko VV, Gel'man EM, Zorina ML, Elina NA, Kul'chitskaya EA, Men'shikov YuP, Polezhaeva LI, Ryzhova RI, et al. (1977) Holtite (first find in the USSR) and its interaction with other tantalum minerals in tare metal pegmatites. Zap Vses Mineral Obshchestva 106:337-347 Von Hodenberg R, Kuehn R (1972) Discussion of the names ascharite and szaibelyite based on recent studies on ascharite. Kali Steinsalz 6:104-111 Von Hodenberg R, Kuehn R (1977) A new mineral of the hilgaidite group: the chloride-tyretskite from Boulby [England]. Kali Steinsalz 7:165-170 Vorbach A (1989) Experimental examinations on the stability of synthetic tourmaline in temperatures from 250 to 750°C and pressures up to 4 kb. N Jahrb Mineral Abh 161:69-83 Voskresenskaya IE, Barsukova ML (1968) Synthesis and properties of some ferruginous and nonferruginous tourmalines. In AN Lobachev (ed), Gidroterm Sin Krist, p 175-92, Izd "Nauka": Moscow Voskresenskaya IE, Plyusnina II (1977) Synthesis of boracite in highly concentrated chloride media and their IR-spectroscopic study. Vestn Mosk Univ Ser 4: Geol 4:51-54 Voskresenskaya IE, Kovyzhenko NA, Shtemberg AA (1975) Synthesis of tourmaline in chloride media II. Sov Phys Crystallogr 20:129-131 Vry J (1993) Mg- and Al-rich rocks from the Reynolds Range, NT, Australia:constraints on the uplift history of a Proterozoic low-pressure terrain. Geol Soc Am Abstr with Prog 25:A-422 Wagman DD, Evans WH, Parker VB, Schümm RH, Halow I, Bailey SM, Chumey KL, Nuttall RL (1982) The NBS tables of chemical thermodynamic properties. Selected values for inorganic and Ci and C2 organic substances in SI units. J Phys. Chem. Ref. Data 11, Suppl. 2 Walenta K (1982) Chemical study on admontite and machatschkiite. Schweiz Mineral Petrogr Mitt 62:177183 Walenta K (1979) Admontite, a new borate mineral from the gypsum deposit Schildmauer bei Admont in Styria (Austria). Tschermaks Mineral Petrogr Mitt 26:69-77 Wan C, Ghose S (1977) Hungchaoite, Mg(H20)5B405(0H)4-2H20: a hydrogen-bonded molecular complex. Am Mineral 62:1135-1143 Wan C, Ghose S (1983) Parahilgaidite, Ca6[B509]3Cl3-3H20: a triclinic piezoelectric zeolite-type pentaborate. Am Mineral 68:604-613 Watanabe J, Hasegawa K (1986) Borosilicates (datolite, schorl) and aluminosilicates (andalusite, sillimanite) in the Oketo Rhyolite. Hokkaido Fac Sei Hokkaido Univ Ser 4 21:583-598 Watanabe T (1955) Suanite, a new magnesium borate from Hol Kol, Suan, North Korea. Am Mineral 40:941 Wegge S, Schreyer W (1994) Boron-free kornerupine: its upper pressure stability limit in the system MgOAl 2 03-Si02-H 2 0 (MASH). Eur J Mineral 6:67-75 Weir CE (1966) Infrared spectra of the hydrated borates. J Res Nat Bur Stds A Phys Chem 70A:153-164 Weisbrod A, Polak C, Roy D (1986) Experimental study of tourmaline stability in the system Na-Mg-AlSi-B-O-H Applications to the boron content of natural hydrothermal fluids and tourmalinization processes. In: Experimental Mineralogy and Geochemistry Intern. Symp., Nancy. Abstr vol 140 Wendling E, Von Hodenberg R, Kuehn R (1972) Congolite, trigonal iron boracite. Kali Steinsalz 6:1-3 Werding G, Schreyer W (1978) Synthesis and crystal stability of kornerupine in the system Mg0-Al203Si02-B 2 03-H20. Contrib Mineral Petrol 67:247-259 Werding G, Schreyer W (1984) Alkali-free tourmaline in the system Mg0-Al203-Si02-B 2 03-H 2 0. Geochim Cosmochim Acta 48:1331-1344 Werding G Schreyer W (1990) Synthetic dumortierite: its PTX-dependent compositional variations in the system Al203-B203-Si02-H 2 0. Contrib Mineral Petrol 105:11-24 Werding G, Schreyer W (1992) Synthesis and stability of werdingite, a new phase in the system magnesium oxide-aluminum oxide-boron oxide-silica (MABS), and another new phase in the ABSsystem. Eur J Mineral 4:193-207 Werding G, Alsumady K, Schreyer W, Medenbach O (1981) Low pressure synthesis, physical properties, miscibility, and preliminary stability of sinhalite, MgAlBO/j. N Jahrb Mineral Abh 141:201-216 Werding G, Pertsev NN, Schreyer W (1992) Synthesis and phase relationships of serendibite. Dokl Akad Nauk SSSR, Earth Sei Sect 315:243-247 Westrum EF Jr., Grenier G (1957) The heat capacities and thermodynamic properties of crystalline and vitreous anhydrous sodium tetraborate from 6 to 350°K. J Am Chem Soc 79:1799-1801 Wight Q, Chao GY (1995) Mont Saint-Hilaire Revisited, Part 2. Rocks and Minerals. 70:90-103,131-138 Wuhan Geologic College; X-Ray Lab. (1977) X-ray crystallographic investigation of hungchaoite and hydrochlorborite. K'o Hsueh Tung Pao, 22(10):451-453,422 Wunder B, Rubie DC, Ross CR n, Mendenbach O, Seifert F, Schreyer W (1993) Synthesis, stability, a*l properties of Al2SiC>4(OH)2: a fully hydrated analogue of topaz. Am Mineral 78:285-297 Xie H, Peng W, Zhang Y, Xu H, Lai Z (1985) A study of the synthetic ludwigite series. Acta Mineral Sinica 5:102-108

222

Anovitz & Hemingway: Thermodynamics

of Boron

Minerals

Yakubovich OV, Simonov MA, Belov NV (1975) Structure refinement of gaudefroyite. Kristallografiya 20:152-155 Yakubovich OV, Simonov MA, Belokoneva EL, Egorov-Tismenko YuK, Belov NV (1976a) Crystal structure of calcium magnesium diorthotri borate (pyroborate) of kurchatovite CaMgft^OsL Dokl Akad Nauk SSSR 230:837-840 Yakubovich OV, Yamnova NA, Shchedrin BM, Simonov MA, Belov NV (1976b) Crystal structure of magnesium kurchatovite CaMg[B205]. Dokl Akad Nauk SSSR 228:842-845 Yakubovich OV, Egorov-Tismenko YuK, Simonov MA, Belov NV (1978a) Crystal structure of natural sakhaite Ca3Mg[B03]2[C03] 0.36H20. Dokl Akad Nauk SSSR 239:1103-1106 Yakubovich OV, Simonov MA, Belov NV (1978b) Crystal structure of synthetic manganese-kurchatovite CaMn[B 2 0 5 ], Dokl Akad Nauk SSSR 238:98-100 Yakubovich OV, Egorov-Tismenko YuK, Simonov MA, Belov NV (1981) Crystal structure of natural saccite [sakhaite] Ca3Mg[B03]2[C03]-0.36H20. Sov Phys Dokl 23:225-226 Yamnova NA, Egorov-Tismenko YuK, Simonov MA, Belov NV (1974) Crystal structure of solongoite Ca2[B 3 0 4 (0H)4]Cl. Dokl Akad Nauk SSSR 216:1281-1284 Yamnova NA, Simonov MA, Belov NV (1975) Crystal structure of roweite Ca2(Mno 52Mgo 48)2 [B 4 0 7 (0H)2](0H)4. Dokl Akad Nauk SSSR 221:1326-1329 Yamnova NA, Simonov MA, Kazanskaya EV, Belov NV (1975) Refined structure of calcium magnesium carbonate borate, borcarite Ca4Mg[B406(0H)6](C03)2. Dokl Akad Nauk SSSR 225:823-826 Yamnova NA, Simonov MA, Belov NV (1975) Crystal structure of iron, magnesium borate hulsite [(Fe ,Mg,Fe ,Sn) 3 B030 2 ]. Kristallografiya 20:156-159 Yamnova NA, Egorov-Tismenko YuK, Simonov MA, Belov NV (1976) Structure of nifontovite Ca 3 [B 3 0 3 (0H)6] 2 -2H 2 0 and solongoite Ca 2 [B 3 0 4 (0H) 4 ]Cl calcium borates. In VA FrankKamenetskii (ed), Kristallokhim Strukt Osob Mineral, p 104-108 "Nauka", Leningr. Otd.: Leningrad Yamnova NA, Pushcharovskii DYu, Malinko SV (1988) Cation distribution in the crystal structure of iron-warwickite. Kristallografiya 33:349-352 Yamnova NA, Egorov-Tismenko YuK, Pushcharovskii DYu, Malinko SV, Dorokhova GI (1993) Crystal structure of a new natural Na,Ca-hydroborate Ca[B304(0H)4], Kristallografiya 39:991-993. Crystallography Reports 39:905-907 Yamnova NA, Egorov-Tismenko YuK, Malinko SV, Pushcharovskii DYu, Dorokhova GI (1994) Crystal structure of a new natural calcium hydroborate NaCa2[B0j4(0H)4]-2H20. Kristallografiya 38:71-76. Crystallography Reports 38:749-752 Yang G, Peng Z (1986) X-ray diffraction studies on the regular intergrowth of tin-rich borates and its genesis. Kuangwu Xuebao 6:289-297 Yang W, Rosenberg PE (1992) The free energy of formation of searlesite, NaBSi205(0H)2, and its implications. Am Mineral 77:1182-1190 Yang W, Rosenberg PE (1995) The free energy of formation of datolite. Am Mineral 80:576-584 Yang X, Sun Z, Tang Y (1991) DSC studies on the kinetics of decomposition of some magnesiumcontaining borates under high pressures. Thennochim Acta 191:13-20 Yarzhemskii YaYa (1956) Preobrazhenskite, a new borate of the saliferous strata of the Into uplift. Dokl Akad Nauk SSSR 111:1087-1090. Referenced from Fleischer, M (1957) New Mineral Names. Am Mineral 42:704 Zachariasen WH (1954) The precise structure of orthoboric acid. Acta Crystallogr 7:305-310 Zachariasen WH (1963a) The crystal structure of cubic metabolic acid. Acta Crystallogr 16:380-384 Zachariasen WH (1963b) The crystal structure of monoclinic metaboric acid. Acta Crystallogr 16:385-389 Zachariasen WH, Plettinger HA, Marezio M (1963) The structure and birefringence of hambergite, Be 2 B0 3 (0H). Acta Crystallogr 16:1144-1146 Zeeman J, Effenberger H, Pertlik F (1982) Verfeinerung der Kristallstrucktur des Kotoits, Mg 3 (B0 3 ) 2 : Ein Beitrag zur Aplanarität von B03-Gruppen. Öster Akad Wiss Mathem-Natur. Klasse Anzeiger 119:61-62 Zeng Y, Liou JG (1982) Synthesis of chambersite and its geological implications. N Jahrb Mineral Mh 2:69-83 Zhan W, Guggenheim S (1995) The dehydroxylation of chlorite and the formation of topotactic product phases. Clay Clay Minerals 43:622-629 Zhao W, Yang G, Pan Z (1991) Electron diffraction analysis of magnesiohulsite twinning. Yanshi Kuangwuxue Zashi 10:339-344 Zhdanov VM, Turdakin VA, Arutyunov VS, Semenov YuV, Malinko SV, Khodakovsky IL (1977) Lowtemperature specific heat, standard entropy, and elastic parameters of datolite and danburite. Geokhimiia 12:1817-1824 Zheng H, Bailey SW (1995) The crystal structure of manandonite—2H2. Am Mineral 80:387-393

CQ ffl a ¿¡P ^ 3? g

a o *

OO ÛvOn* ^ «Mffl »? m M aS S S Z o

O W

t a .2.

s ^

fe

03 i.

g ¥>

A OqQ feo

€ a o

"âù S

oo M *—' ï O

a o

2. cT

P •j Oa

B

M ^ "3

F-

« K> a m «' HF 3 u o z s - s Pnma

y

*

q O r a fflg-g •—i u

g -

^ H

a o

1

ooq' s O

o

a HH "oo

q

S o' o m

ffl

o o "W

nT «T

Pp S

X a asM

3 3 -•> o O O z. m m e

£

03 %

«r

a z

¿C2 -

S , S 2 - 03" oT oT oT pT m 03 o z

«fm 0" y S= s - í - f o ä «c i è Z S

q q q a a a o Gr ET 5T a a a a ^ g ^ O ^O ^O w o o o o" O

°os ? 1

•t q e K 1 O

8 0" ? £ B S . i « I I

O

a o o B

„ q

n

u Ö" B

M 1

£

%

Pu

Alla

cubic high T (>265°C), CI coordinates Mg

oí od a F43c

r-T CU CU O. CU

Pnnm or after pinakiolite Pnn2 Pbam(?) after ludwigite Pca2, orthorhombic low T

Na coordinations irregular

after zircon

no structure refinement

P4/nmm Cu coord by 4 (OH) and 2 CI

from datolite

K coord could be 10-fold no structure refinement

P2,/c

Pnma Pbam

Reference

Dal Negro et al. (1975)

Dal Negro et al. (1971)

Dal Negro et al. (1976)

Levy & Lisensky (1978) Morimoto (1956)

Sueno et al. (1973)

Dowty & Clark (1973)

Sueno et al. (1973)

Dowty & Clark (1973)

De Waal et al. (1975)

Raade et al. (1988)

Corazza et al. (1974)

Giuseppetti et al. (1990)

Giuseppetti et al. (1990)

Giuseppetti et al. (1990)

cf. Speer (1982)

Mrose & Rose (1961);

Ross & Edwards (1959) Collin (1951) Fomaseri (1950, 1951)

Christ (1959) Kramer & Allen (1956)

Branton(1969)

Ghose & Wan (1977)

structural H20 differs from Merlino & Sartori (1971) Fleischer & Mandarino (1995)

Comments

O (2 m rn

Borax

B f f l n » =•„=•„2; m (S S

Boracite alpha

n

q s

1

of Boron

Bonaccordite Boracite beta

œ c? ^

Blatterite

œ hC

Berborite-17 Berborite-27" Berborite-2/í Biringuccite

~ s ^ OC s

Behierite

M

Barberiite

/-S

S O o

Bandylite

Space End-member structural formula Sym group S o S S CN~ PL,

Bakerite

Formula ^ B) ü ü (s" Í M N Oh Ou, U U

Avogadrite Azoproite

Admontite Aksaite Ameghinite Ammonioborite Aristarainite

Mineral

A n o v i t z & Hemingway: Thermodynamics Minerals 223

Table 1

£

O

a0 0"

a 0 w 0, ® m" a ,f ?

0 a~

Ü= q

,-î

O w 0;

S

Z

«'2 y « 2 W

CS 8 g ^ SB £ S 0 2 s O «g, M 5 ci < 0 a

i f O A" « Ï= 'a B z

3

S P - 2/m no structure refinement class P3 F coordinates Y P2,/n Pca2, P31c Pbca Pbam R3m a sf

x £ O S O oi O O

e£ S

» Tm m

Z?

.fr S f oT 5

Z

O

% o

s? §

a

S 0 ^

9 0' < g 5 " O m ci* cd y y

A

S

OIS

X

0' S0 «/— S

X VO

cfiP — m « 0 a B 2 eo j^S W oî* & y

o S g «>: * S o ¿-T u U ^ £ 0

S o

o f

_T £ « J? y «1» 0 « S

?
I 1

a

Tu3 S a" 2 E2 q*

oc» X cs

n" S o < m ¡2. oT 00 s J? 9> io cT o£ 3 o~

^

j

c,

-S O S X oa ¿re

fe

a

1 S

• 1 ^ 1 pq n, b n,

a

Swinneaetal. (1981) Takiuchi et al. (1974) Brunton (1968) Grice & Robinson (1989) Grice 4 Ercit (1993) Dal Negro &. Tadini (1974) Moore & Araki (1974) MacDonald et al. (1993)

Christ (1953a) Christ et al. (1958) Burns & Hawthorne (1993) polymorph of boracite ? Wendling et al. (1972) Dowty and Clark (1973) Phillips et al. (1974) Ito & Mori (1953) Foit et al. (1973) King&SengLer-Roberts (1988) Li located only approximately Krogh-Moe (1962) Hawthorne et al. (1993) Buerger et al. (1962) Nuber & Schmetzer (1979) Grice & Ercit (1993) Moore & Araki (1978) Malinko et al. (1980) Donnay & Barton (1972) Gorskaya et al. (1982) Grice & Ercit (1993) Hade (1955) Ktthn & Schaacke (1955) Cannillo et al. (1973) Konnert et al. (1970)

Anovitz & Hemingway: Thermodynamics of Boron Minerals

a 1 e I"a

'2 £ il 'E £ S o £ o a? ^ a p 5 s ^ S o o O"^ OK § 5® «

£ £"c K

o, 1 3 © 0 tS

6

l l

s J c^ Bi o Oi

H.5 no,

'S b

S K O, < S3 ¿jo"© ^ O g Z o s • 0>

CL.

S> s§

225

Table 1

|| .§ £ "g II CO

11 z 4 ens vo 6 Pi Pi

o,

a

5 —i < m B »H a" O" o O ¡2, J?¡2,& « ¡r [L i CO — — ci* cf ci* S3 «z ya ya y« 7z gjBjf

'8 5§ 2

Hi

S

£

I5

o

® m —
i U VI

o

S

ffi o

o B B

o oa ffi O CO

K>

ffi

S

oT cq

O

S

o

m

' fa

o

y

B B

u

u

X

o tc ^

ffi O pi" « CJ

o~ OQ

o o

S

°

ffi pa

S pa, (d' etT u u

s o

•S "S .3 o O

s

B en

ffi

o 00

o

m

o CQ

<


& s r .. S5 | g

•5 1 o

•B• S i3

1= eS

« s "5 ctf

§

o •o

g

c

\

J

¡ 1 ' S o «2 >3 3 4> %* s: P3

OO ^ ^ ^ « s 0 è ^ £ S m < 5 5 © S

o ttf o ^ o £

_ O o ^

Ä

.