Ultrahigh pressure metamorphism: university textbook

EUROPEAN MINERALOGICAL UNION NOTES IN MINERALOGY Series editors: Gábor Papp and Tamás G. Weiszburg

Volume 5

ULTRAHIGH PRESSURE METAMORPHISM UNIVERSITY TEXTBOOK

Edited by

Dennis A. Carswell and

Roberto Compagnoni in association with

Franco Rolfo

Eötvös University Press Budapest, 2003

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

The publication of this textbook and the organisation of the connected school were sponsored by EU Socrates/Erasmus IP on Ultrahigh Pressure Metamorphism EU Socrates/Erasmus PROG/CDA on the Preparation of a Coordinated European Curriculum in Mineral Sciences European Mineralogical Union Institute of Geosciences and Earth Resources, Italian National Research Council University of Torino IUGS Hungarian National Committee Koch Sándor Foundation Eötvös Loránd University Hungarian Natural History Museum

EMU Notes in Mineralogy A series published under the auspices of the European Mineralogical Union (EMU) in connection with the EMU Schools meetings.

Initiator of the EMU Schools and the EMU Notes in Mineralogy Prof. Giovanni Ferraris, Torino, President of the EMU 1992–1996

Executive Committee of the EMU (2000–2004) Prof. David J. Vaughan, Manchester (President), Prof. Wolfgang F. Müller, Darmstadt (Past President), Prof. Thomas Armbruster, Bern (1st Vice President), Prof. Fernando Scordari, Bari (2nd Vice President), Dr. Charles E.S. Arps, Leiden (Treasurer), Prof. Herta S. Effenberger, Vienna (Secretary) Series editors: Tamás G. Weiszburg, Gábor Papp, Budapest

Editors of this Volume Dennis A. Carswell, Sheffield & Roberto Compagnoni, Torino, in association with Franco Rolfo, Torino Technical editor: Tamás Váczi, Budapest CD-ROM compiled by Franco Rolfo, Torino Indexes prepared by Ágnes Gál, Gábor Papp and Tamás Váczi Front cover design: Mária Hodosi On the front cover: Coesite inclusion in pyrope from kyanite-phengite-pyrope whiteschist, UHPM Brossasco-Isasca Unit, Dora-Maira Massif, Western Alps (see Chapter 2). In the centre, a crystal of coesite about 3 mm across is included in pyrope. Coesite is surrounded and transected by “palisade” inversion quartz. Radial cracks in the pyrope around the inclusion appear due to the volume increase associated with the inversion of coesite to quartz during uplift. Crossed polarisers with first-order red plate (DM507). Photo by Franco Rolfo, University of Torino, Italy. ISSN 1417 2917 ISBN 963 463 646 2 © 2003 authors and EMU Published by Eötvös University Press Budapest, Szerb u. 21–23, H-1056 Hungary Printed in Hungary by Mondat Ltd., Budapest

Foreword This volume, the 5th in the EMU Notes in Mineralogy series, follows four volumes devoted to different aspects of mineralogy: Modular Aspects of Minerals, edited by S. Merlino (Vol. 1); Environmental Mineralogy, edited by D.J. Vaughan and R.A. Wogelius (Vol. 2); Solid Solutions in Silicate and Oxide Systems, edited by C.A. Geiger (Vol. 3); Energy Modelling in Minerals, edited by C.M. Gramaccioli (Vol. 4). It is the first volume dealing with a petrological subject and contains the contributions of the lectures given at the 5th School of the European Mineralogical Union (EMU) on “Ultrahigh Pressure Metamorphism” held in Budapest from 21 to 25 July 2003. The topic of UHPM was selected because this extreme type of metamorphism, initially considered as a petrographic oddity by the geologic community, has now become recognised as a normal feature of continental plate collisional orogens and important to understanding just how deep the upper part of the continental lithosphere can subduct. We note that this School took place just twenty years from the first report by Christian Chopin of coesite in exposed orogenic metamorphic rocks of the continental crust. The lectures given at this school benefited by the scientific results of the research promoted by the ILP Task Groups III-6 and III-8, active on UHPM from 1994 to 1998 and 1999 to 2004, respectively, and published in a number of monographs and special issues of international journals. It is our strong belief that this petrologic topic should be recognised to be of paramount importance in the education of students and young researchers in Earth Science. We wish to thank Giovanni Ferraris, former Past President of the EMU, for his encouragement to undertake this venture and the Executive Committee of the EMU, particularly Wolfang F. Müller Past President and David J. Vaughan present President, for their support. Special thanks also to Herta S. Effenberger, Secretary of the European Mineralogical Union, whose precious suggestions and encouragement have been very much appreciated. This publication has been made possible through the generous support of several institutions and of many individuals, who have freely devoted their time and energy. On behalf of all of those involved in the School, we should like to thank the European Mineralogical Union, and the European Commission, through the SOCRATES/ERASMUS Intensive Programme (IP), for providing scholarships to the students and support funds for the teachers. The Institute of Geosciences and Earth Resources of the Italian National Research Council, the University of Torino, Italy, the IUGS Hungarian National Committee, the Department of Mineralogy of the Eötvös Loránd University, Budapest, Hungary and the Koch Sándor Foundation, Miskolc, Hungary are all thanked for their financial support with this volume. The production of this volume, as well as a provisional version available in time for the School, was possible thanks to considerable efforts by the patient and competent EMU Notes Series Editors, Gábor Papp and Tamás G. Weiszburg, and the Technical Editor, Tamás Váczi, of the Eötvös University Press. All of the chapters underwent peer review by different authors of the volume aided by Stéphanie Duchêne, Joerg Hermann, Bruno Lombardo, Bruno Messiga, Tadao Nishiyama, Giovanni Piccardo, Marco Scambelluri, Benedetto De Vivo and Zheng Yong-Fei. We would like to thank all of those involved for their valuable contributions. Obviously, we should point out that any errors remaining in the volume are the responsibility of the authors and editors. Last, but not

least, a sincere thanks to all our colleagues, who agreed to take part in the School, thereby contributing to its success. Tamás G. Weiszburg, Coordinator of the Erasmus Intensive Program (IP), and his co-workers of the Department of Mineralogy of the Eötvös Loránd University are acknowledged for all their work in organising and operating the scientific and logistic schedule of the School within the modern streamlined lecture halls and laboratories of the new Scientific Campus of the Eötvös Loránd University, Budapest. It is our hope that this volume, which concerns the most important aspects of UHPM, will prove to be of valuable help to all people, teachers and students, interested in having an updated and comprehensive knowledge on this special subject. With this purpose in mind, the text volume has been supplemented with a CD-ROM (edited by Franco Rolfo), where all images (photomicrographs, maps and diagrams) for which colour may improve the understanding have been stored. Finally we hope that this volume will contribute to further development in recognising and understanding ultrahigh pressure metamorphism and will lead to beneficial collaborative ventures among scientists from different branches of Earth Science. Tony Carswell and Roberto Compagnoni Sheffield and Torino, August 2003

Contents Part I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

Chapter 1. Introduction with review of the definition, distribution and geotectonic significance of ultrahigh pressure metamorphism by DENNIS A. CARSWELL and ROBERTO COMPAGNONI . . . . . . . . . . . . . . . . . . . .

3

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

Part II. Reviews of representative UHPM terranes

. . . . . . . . . . . . . . . . . 11

Chapter 2. UHPM units in the Western Alps by ROBERTO COMPAGNONI and FRANCO ROLFO . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Brossasco-Isasca Unit of the Dora-Maira Massif . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Polymetamorphic Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relics of pre-Alpine metamorphic rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Variscan paragneiss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relics of Variscan augengneiss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relics of Variscan marble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basement lithologies preserving UHPM mineral assemblages . . . . . . . . . . . . . . . . Jadeite-kyanite-almandine-phengite micaschist . . . . . . . . . . . . . . . . . . . . . . Phengite-jadeite-almandine-quartz (/coesite) granofels . . . . . . . . . . . . . . . . Eclogite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unusual lithologies characterised by UHPM mineral assemblages . . . . . . . . . . . . Kyanite-pyrope talcschist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sodic whiteschist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Monometamorphic Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UHPM rocks derived from late Variscan intrusives . . . . . . . . . . . . . . . . . . . . . . . . Orthogneiss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relics of metagranitoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relics of igneous intrusive contacts with basement xenoliths . . . . . . . . . . . . Pyrope-bearing whiteschist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Garnet-jadeite-kyanite-quartz (/coesite) granofels . . . . . . . . . . . . . . . . . . . . The metamorphic evolution of the BIU and its P–T–t path . . . . . . . . . . . . . . . . . . . . . . . The age of the UHP metamorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Lago di Cignana Unit of the Piemonte zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geological and structural setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lithologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metabasites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metasediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Garnet-phengite-quartz calcschist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Piemontite-garnet-phengite-talc-quartz schist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aegirine-augite - garnet - phengite quartzite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metamorphic evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geochronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13 13 16 17 17 20 20 20 21 21 21 22 23 23 23 23 23 23 24 26 28 29 31 34 34 36 36 38 38 39 40 41 41 42 44 45 45 45

VI Chapter 3. Ultrahigh pressure metamorphism in the Western Gneiss Region of Norway by DENNIS A. CARSWELL and SIMON J. CUTHBERT . . . . . . . . . . . . . . . . . . . . . . . 51 Historical background to UHPM in western Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Most recent discoveries of UHPM rocks in western Norway . . . . . . . . . . . . . . . . . . . . . . . . . . The “foreign” versus “in situ” eclogite controversy: The influences of differential retrogression and metastability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The HPM to UHPM transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The age of the UHPM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The occurrence and interpretation of garnet peridotites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

51 53 55 60 65 66 70 70

Chapter 4. The Kokchetav massif of Kazakhstan by VLADISLAV S. SHATSKY and NIKOLAI V. SOBOLEV . . . . . . . . . . . . . . . . . . . . . . 75 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geological outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UHP rocks of the Kumdy-Kol and Barchi locations (Unit I) . . . . . . . . . . . . . . . . . . . . . . . . . . . Eclogites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biotite gneisses and garnet-pyroxene-quartz rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metasomatic rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Garnet-pyroxene rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dolomitic marbles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microdiamonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The UHP and HP rocks of Unit II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sulu-Tyube area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Enbek-Berlyk area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kulet area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P–T path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geochemistry of metamorphic rocks and the age of the UHP metamorphism . . . . . . . . . . . . . Exhumation of high pressure rocks of the Kokchetav Massif . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75 76 76 76 80 82 82 83 83 87 87 89 91 91 94 98 100

Chapter 5. The Dabie Shan–Sulu orogen by TAKAO HIRAJIMA and DAISUKE NAKAMURA . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Geological framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reconstruction of pre-HP/UHP stage geology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review of the equilibrium temperature of representative UHP rocks at peak stage . . . . . . . . . P–T history of garnet peridotite and associated UHP minerals . . . . . . . . . . . . . . . . . . . . Peak P–T conditions of eclogite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The character of Grt–Cpx geothermometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The geobarometry of eclogite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Compilation result of the Dabie Shan eclogite . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results from the Sulu eclogite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105 110 117 117 123 125 126 129 132 136 139 139

VII Chapter 6. The Bohemian Massif and the NW Himalaya by HANS-JOACHIM MASSONNE and PATRICK J. O’BRIEN . . . . . . . . . . . . . . . . . . . . 145 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geological setting and geochronological constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bohemian Massif . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Himalaya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Petrological information on HP/UHP key areas and rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bohemian Massif . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metabasites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ultramafic rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quartzofeldspathic and related rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Himalaya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kaghan and Neelum Valleys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tso Morari . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Indus suture zone, Nanga Parbat–Haramosh Massif . . . . . . . . . . . . . . . . . . . . . . . . Kharta region, east of Mount Everest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geodynamic synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

145 145 145 148 151 151 151 158 159 163 165 169 170 172 172 177 177

Part III. Mineralogy, geochemistry and tectonometamorphic evolution of UHPM terranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Chapter 7. Mineral chemistry and mineral reactions in UHPM rocks by CHRISTIAN CHOPIN and GIOVANNI FERRARIS . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Closest- and close packing structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basic features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Filling tetrahedral sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Filling octahedral sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Filling tetrahedral and octahedral sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rock-forming minerals with a complex structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Garnets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pyroxenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phengite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnesiostaurolite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnesiochloritoid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Accessory minerals with a complex structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zircon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Topaz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wagnerite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Titanite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydroxylclinohumite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ellenbergerite group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnesiodumortierite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bearthite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

191 191 191 193 195 198 202 203 203 205 207 210 212 212 213 213 213 214 215 215 216 217 218 219 219

VIII Chapter 8. Thermobarometric methodologies applicable to eclogites and garnet ultrabasites by ERLING J. KROGH RAVNA and JENS PAQUIN . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Mineral assemblages of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Practical geothermometers and geobarometers for HP/UHP rocks . . . . . . . . . . . . . . . . . 230 Some cautionary notes before proceeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 Geothermometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Geothermometers based on Fe2+–Mg exchange between garnet and other Fe–Mg minerals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Garnet–biotite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Garnet–hornblende . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Garnet–phengite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Garnet–clinopyroxene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 Garnet–olivine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Garnet–orthopyroxene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Fe-Mg geothermometry and the problem of Fe3+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Solvus thermometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Two-pyroxene thermometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Trace element thermometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 The Ni-in-garnet thermometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Partitioning of transition elements (Sc, V, Cr, Co and Mn) between orthopyroxene and clinopyroxene in peridotitic and websteritic mantle rocks . . . 241 Ca-Cr system in lherzolitic garnets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 The clinopyroxene/plagioclase symplectite geothermometer . . . . . . . . . . . . . . . . . . . . . 243 Geobarometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 The Al-in-orthopyroxene barometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 Effect of Fe3+ on the Al-in-orthopyroxene barometer . . . . . . . . . . . . . . . . . . . . . . . 245 The Cr-in-clinopyroxene barometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Geothermobarometry based on the assemblage garnet–clinopyroxene–phengite– kyanite–quartz/coesite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 Other, less common geobarometers suitable for HP/UHP rocks . . . . . . . . . . . . . . . . . . . 250 Geobarometers involving plagioclase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 Geobarometers involving zoisite/clinozoisite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Ca-Eskola molecule bearing clinopyroxenes at UHP conditions . . . . . . . . . . . . . . 252 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

Chapter 9. Coronitic reactions: Constraints to element diffusion during UHP metamorphism by MARCO RUBBO and MARCO BRUNO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kinetic theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Growth of mineral layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Symplectic reaction in olivine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theory of intergrowth spacing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An estimate of intergranular diffusion of Al in fluid undersaturated systems . . . . . . . . . . . . . . The ultrahigh pressure coronitic reactions in a metagranodiorite . . . . . . . . . . . . . . . . . . . . . . . Geology and petrography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Equilibrium thermodynamic modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Further considerations on the model and the metamorphic evolution of the eclogite facies metagranodiorite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

261 262 263 270 272 279 283 283 285 288

IX Garnet growth model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Some notions of irreversible thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Empirical laws of diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thermodynamic theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diffusion in garnet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

290 294 294 299 299 301 302 303 303

Chapter 10. Mineral assemblages in ultrahigh pressure metamorphism: A review of experimentally determined phase diagrams by STEFANO POLI and PATRIZIA FUMAGALLI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 “Fluid” phases at ultrahigh pressure conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 A few definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 How likely is fluid saturation at high pressure? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310 Fluids, melts and the 2nd critical endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314 Ultrahigh pressure rocks: The quartz–coesite transformation and the “internally consistent” thermodynamic databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318 Ultrahigh pressure rocks: The graphite–diamond transformation . . . . . . . . . . . . . . . . . . . . . . . 319 Mafic systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320 Bimineralic mafic eclogites and the relevance of minor and accessory phases . . . . . . . . 320 Phase relationships in H2O-bearing systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321 Phase relationships in a C-O-H-bearing system . . . . . . . . . . . . . . . . . . . . . . . . . . . 324 Ultramafics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Peridotite compositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 The spinel to garnet transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328 The peridotite + H2O system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 The K-peridotite system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 Metasedimentary rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336

Chapter 11. Dating UHP metamorphism by DANIELA RUBATTO, ANTHI LIATI and DIETER GEBAUER . . . . . . . . . . . . . . . . 341 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The aims and the challenge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Dora-Maira . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Kokchetav Massif . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other localities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tectonic implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exhumation rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Duration of UHP metamorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Age variations within an orogen: The case of the Western Alps . . . . . . . . . . . . . . . . . . . The challenge ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

341 341 342 343 344 344 348 351 353 353 355 355 356 359 360 360

X Chapter 12. Geochemistry and isotope tracer study of UHP metamorphic rocks by BOR-MING JAHN, DOUGLAS RUMBLE and JUHN G. LIOU . . . . . . . . . . . . . . . . . 365 I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 II. Chemical compositions of eclogites and ultramafic rocks . . . . . . . . . . . . . . . . . . . . . . . . . . 366 III. Sm–Nd and Rb–Sr isochron ages and Nd–Sr isotope tracers . . . . . . . . . . . . . . . . . . . . . . . 370 Ranges of Sm and Nd concentrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 Equilibrium partition coefficients (Kd values) between Cpx and Grt . . . . . . . . . . . . . . . 373 Meaningful Sm–Nd and Rb–Sr isochron ages: Some examples . . . . . . . . . . . . . . . . . . . 377 Failure of producing correct ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 An example from the Hong’an Block in western Dabieshan . . . . . . . . . . . . . . . . . . . . . . 382 Radiogenic isotope tracers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384 IV. Oxygen isotope tracer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 /18O values of eclogites from UHP metamorphic terranes – a summary . . . . . . . . . . . . . 389 Limited fluid activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Preservation of equilibrated high-temperature isotope fractionation . . . . . . . . . . . . . . . . 393 Conclusions from oxygen isotope tracer studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 Coupled Nd and O isotopic disequilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 V. Application of isotope constraints to tectonic evolution – example of the Dabie orogen . . . 397 Lithological and geochemical characteristics of the NDC and SDC gneisses and Cretaceous intrusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398 Isotope test of the existing tectonic models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399 Discussion and tectonic implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407

Chapter 13. Three-dimensional mechanics of UHPM terrains and resultant P–T–t paths by PETER O. KOONS, PHAEDRA UPTON and MICHAEL P. TERRY . . . . . . . . . . . . . 415 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 General mechanical considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416 Mechanical model: Constraints from natural analogues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418 Natural occurrences: Western Gneiss Region, Norway . . . . . . . . . . . . . . . . . . . . . . . . . . 419 Tectonic framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421 Kinematic framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421 Observations of the modern analogue: Central New Zealand . . . . . . . . . . . . . . . . . . . . . 423 Mechanical framework I: Orogen-scale dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Numerical results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 Role of disequilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Influence of surface processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 Strain partitioning within the orogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Mechanical framework II: Viscous mixing, local kinematics and nappe formation . . . . . . . . . 429 Thermal and petrological evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 Results of static solutions: Crustal thickening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434 Asthenospheric involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434 Dynamic model with thermal results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436 Thermal–mechanical coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

XI Chapter 14. Metamorphism and textures of dry and hydrous garnet peridotites by LAURO MORTEN and VOLKMAR TROMMSDORFF . . . . . . . . . . . . . . . . . . . . . . . 443 Central Alpine domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 Poikiloblastic and porphyroclastic garnet peridotites in the Central Alps . . . . . . . . . . . . 445 Oriented ilmenite rods in olivine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448 Lattice preferred orientation (LPO) of olivine from AA and CdG . . . . . . . . . . . . . . . . . . 449 Criteria determining the deformation conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452 Eastern Alpine domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 Nonsberg ultramafic rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 Textures of the Nonsberg peridotites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456 Coarse type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456 Fine type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461

Chapter 15. Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks by JACQUES L.R. TOURET and MARIA-LUCE FREZZOTTI . . . . . . . . . . . . . . . . . . . 467 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 The fluid inclusion approach to HP and UHP rocks: The principles . . . . . . . . . . . . . . . . . . . . . 468 Step 1: Identification of the fluid types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468 Step 2: Chronology of the different fluid types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470 Application of Steps 1 and 2: Micro-mapping at the scale of the thin section on the example of the Shuanghe coesite-bearing eclogite (Fu, 2002) . . . . . . . . . . 473 Step 3: Selection of few representative inclusions for each fluid type . . . . . . . . . . . . . . . 476 Step 4: Comparison between fluid inclusion and independent P–T mineral data . . . . . . 477 Applications: Some general considerations based on recent studies . . . . . . . . . . . . . . . . . . . . . 480 Which fluids can be expected in HP–UHP rocks? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480 Which fluids are actually found in HP–UHP rocks? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481 High grade metamorphic rocks: Some fundamental differences between high pressure (eclogites and related rocks) and high temperature (granulites) metamorphic rocks . . . . . . . . 483 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485

Name index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496

XII

Conventions, recommendations and standards used in this volume Spelling: British English Mineral names: Recommendations of the IMA CNMMN Symbols (abbreviations) for rock-forming minerals: In accordance with the Appendix of The nomenclature of minerals: A compilation of IMA reports; edited by R.F. Martin; Ottawa: Mineral. Assoc. Can., 1998. Crystallographic symbols: In accordance with the International tables of crystallography, Volume A, Space-group symmetry; edited by Th. Hahn; fourth, revised edition; Dordrecht: Kluwer; 1995. Transcription of Cyrillic characters: British Standard 2979:1958 Notes on the usage of the References lists Multiple references to an author are listed in the following order: (1) publications of the author alone, in chronological order; (2) publications of the author with a single coauthor, in alphabetical order of the co-authors; (3) publications of an author with more than one co-author in chronological order. Different authors with the same surname are not considered separately. Journal names are abbreviated according to the ISDS standards (with a few exceptions in the usage of capitals and in the transcription of Cyrillic characters).

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 1, 3–9

Introduction with review of the definition, distribution and geotectonic significance of ultrahigh pressure metamorphism DENNIS A. CARSWELL1* and ROBERTO COMPAGNONI2 1

Department of Geography, University of Sheffield, Dainton Building, Brookhill, Sheffield S3 7HF, UK 2 Dipartimento di Scienze Mineralogiche e Petrologiche, Università di Torino, Via Valperga Caluso 35, 10125 Torino, Italy; * e-mail: [email protected] Ultrahigh pressure metamorphism (UHPM) is an important type of orogenic metamorphism that over recent years has been increasingly recognised as a characteristic, though poorly preserved, feature of many Phanerozoic plate collision zones. UHPM can be defined as “a type of metamorphism that occurs at very high lithostatic pressures within the eclogite facies but above the stability field of quartz”. Figure 1 shows the currently recognised UHPM rock occurrences within exposed metamorphic terranes on a simplified global geotectonic map. The following is a list of these occurrences (number keyed to Fig. 1) together with selected primary or review references that provide further details: 1) Dora-Maira Massif, Western Alps (Chopin, 1984; Chopin et al., 1991). 2) Zermatt-Saas Zone, Western Alps (Reinecke, 1991; van der Klauw et al., 1997). 3) Leaota Massif, South Carpathians (Săbău, 2000). 4) Rhodope Metamorphic Province, Northern Greece (Mposkos & Kostopoulos, 2001). 5) Western Gneiss Region, Norwegian Caledonides (Smith, 1988; Wain, 1997; Cuthbert et al., 2000). 6) Northeast Greenland Caledonides (Gilotti & Krogh Ravna, 2002) 7) Saxonian Erzgebirge, Germany (Massonne, 1999, 2001; Nasdala & Massonne, 2000). 8) Sudetes Mountains, SW Poland (Bakun-Czubarow, 1991, 1992). 9) Maksyutov Complex, Southern Urals (Leech & Ernst, 1998). 10) Kokchetav Massif, Kazakhstan (Sobolev & Shatsky, 1990; Shatsky et al., 1995). 11) Tian Shan, Kyrghyzstan/China (Tagiri et al., 1995; Zhang et al., 2002). 12) Kaghan Valley, Pakistan Himalaya (O’Brien et al., 2001; Treloar et al., 2003). 13) Tso-Morari Complex, eastern Lakakh, Indian Himalaya (Sachan & Mukherjee, 2003). 14) Su-Lu Terrane, Eastern China (Zhang et al., 1995; Zhang & Liou, 1998) 15) Dabie Shan, Central China (Wang et al., 1989; Wang & Liou, 1991; Okay, 1993; Wang et al., 1995).

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

Fig. 1. Simplified global geotectonic map showing the recognised occurrences of UHPM rocks within exposed metamorphic terranes, with the localities number keyed to the text.

4 D.A. Carswell & R. Compagnoni

Introduction with review of UHPM 16) 17) 18) 19) 20) 21) 22)

5

North Qinling, Central China (Liu et al., 2003b). North Qaidam, NE Tibet Plateau, China (Yang et al., 2002; Song et al., 2003). South Altyn Tagh, NE China (Liu et al., 2001, 2003a). Bantimala Complex, Central Sulawezi, Indonesia (Parkinson et al., 1998). Pan-African Nappes of Northern Mali, West Africa (Caby, 1994). Pan-African Nappes of Minas Gerais State, SE Brazil (Parkinson et al., 2001). Lanterman Range, North Victoria Land, Antarctica – not shown on Figure 1 (Ghiribelli et al., 2001; Palmeri et al., 2003). Some of these occurrences (most notably in the Maksyutov Complex of the Southern Urals, the Sudetes of Poland, the South Carpathians of Romania, and the Lanterman Range of Antarctica) are currently less certain than the others and require further confirmatory evidence. The UHPM rock-bearing collision belt across central China is complex and disrupted. At its eastern end the well-documented occurrences in the Su-Lu and Dabie Shan regions are considered to be parts of the same Triassic plate collision belt that has been offset by substantial sinistral strike-slip along the NE trending Tanlu Fault. This UHPM belt may extend some 400–500 km west into the North Qinling Mountains but the separate UHPM occurrences some 1400–2400 km further west on the northern margin of the Qaidam Massif and in the South Altyn Tagh are thought to have been stabilised within an older Palaeozoic collision belt. In the majority of the well-established occurrences of UHPM rocks, mineral indicators of UHPM are best preserved in metabasic rocks (eclogites) but in a few instances also in meta-sedimentary schists and calc-silicate gneisses. In most instances it has been the identification of preserved inclusions of coesite, within relatively robust minerals such as garnet and zircon, that has led to the positive recognition that the rocks have experienced UHPM conditions. Indications are, however, that the metastable survival of coesite grains within exhumed UHPM terranes is invariably very limited. Thus it is often necessary to resort to indirect petrographic evidence that coesite was previously stable. This takes the form of inclusions of polycrystalline quartz, sometimes with a highly distinctive palisade microstructure, considered to have pseudomorphically replaced previous coesite grains (see for example Carswell & Zhang, 1999). Corroborative support for this interpretation is often provided by observations of radial expansion fractures within the mineral enclosing the polycrystalline quartz inclusion, reflecting the roughly 10% volume increase caused by the replacement of the higher-pressure silica polymorph coesite by ,-quartz. In a few UHPM terranes, most notably in the Kokchetav Massif of Kazakhstan and the Erzgebirge in Germany, preserved micro-inclusions of diamond are the most obvious mineral indicator although supportive evidence of rarer coesite preservation has also been documented. Most UHPM occurrences occur within recognised continental crust sequences that characteristically contain large volumes of granitic–granodioritic orthogneisses. Such UHPM terranes are thus interpreted to have formed in “A-type” continent–continent plate collision zones (as in the Himalayas) where the leading edge of one plate has experienced transient subduction down to at least 90 km. It may well be that in such

6

D.A. Carswell & R. Compagnoni

instances the relatively “dry” orthogneisses have largely failed to react and convert to UHPM mineralogies in response to the deep level subduction. Consequently, the relative buoyancy of the metastably preserved low pressure mineralogies in such orthogneisses may be a crucial geodynamic factor in driving the ensuing rapid exhumation of such UHPM terranes, so enabling the preservation (albeit limited) of the mineralogical evidence for UHPM in those lithologies that did respond and react to the imposed, short-lived, UHP conditions. Such UHPM terranes often even show only partial eclogitisation of metabasic igneous rocks and signal that with the absence or only limited availability of fluid to promote reactivity, the progress of prograde (subduction-related) as well as retrograde (exhumation-related) metamorphic reactions may be severely limited. It thus seems likely that kinetic controls on mineral reactivity and metastability explain why UHPM terranes invariably contain only small volumes of recognisable UHPM rocks within rock sequences dominated by much lower pressure mineral parageneses. Certainly even if felsic rocks have been transformed into UHP mineralogies during deep subduction then the mineralogical evidence is much less likely to survive their subsequent exhumation than in any associated metabasic rocks (eclogites). Exhumation-related deformation, as evidenced by widely developed mylonitic fabrics, is likely to be strongly partitioned into the quartzo-feldspathic rocks (gneisses) leading to effectively pervasive overprinting by retrograde, lower pressure mineral assemblages, mostly of amphibolite facies. Also in such rocks the dehydration breakdown during exhumation of phengitic mica, likely to have been a volumetrically important phase under UHPM conditions, can be expected to generate a hydrous fluid phase that would promote other retrograde mineral reactions. Currently only two of the recognised UHPM occurrences (namely in the Zermatt–Saas Zone of the Western Alps and in Sulawesi, Indonesia) are interpreted to have formed in deeply subducted oceanic crust sequences. Whilst expectations are that in “B-type” subduction zones some oceanic crust rocks will customarily be subducted down into the realm of UHPM, the recovery of such UHPM rocks back up to the Earth’s surface is likely to be a rare event requiring unusual, as yet unclear, geotectonic circumstances. UHPM has now been recognised as an important feature of the major Phanerozoic continental plate collision zones, represented by the Caledonian, Variscan/Hercynian, Alpine and Himalayan orogenic belts. The UHPM rocks in the correlative Pan-African nappes of Northern Mali and SE Brazil are somewhat older, i.e. Late Precambrian/ Neoproterozoic, and thought to have formed in a continental plate collision zone at around 630–640 Ma. The lack of recognised older UHPM rocks within the geological record suggests that prior to around 650 Ma the continental lithospheric plates in particular may not have been cool, thick and strong enough to enable continental crust rocks to be both subducted down to sufficient depths and then recovered back to the surface with at least partial survival of UHPM mineralogies. This volume commences with in-depth reviews of UHPM rocks in four of the best documented occurrences – the Western Alps, the Western Gneiss Region of Norway, the Kokchetav Massif of Kazakhstan and the Dabie Shan–Su Lu orogen in central China.

Introduction with review of UHPM

7

This is followed by an account of the more recently discovered occurrences in the Saxonian Erzgebirge in central Europe and in the Kaghan Valley of the Pakistan Himalaya. Much of the rest of this volume presents detailed considerations of various aspects that are important to our understanding and interpretation of UHPM rock occurrences. These, in turn, are on mineral reactions and mineral chemistry; application of appropriate thermobarometric methodologies; constraints on coronitic and pseudomorphic mineral reactions; relevant experimental petrology data; geochronological dating; geochemistry and isotopic tracer characterisation of protoliths; pressure–temperature–time paths and modelling of exhumation mechanisms; and on the identification and interpretation of fluid inclusions. There is also a separate review chapter on the occurrences and interpretation of the garnet peridotites that are conspicuous components of several UHPM terranes. We hope that this state-of-the-art review on UHPM rocks will inspire and excite future generations of Earth Science students and encourage them to tackle the fundamental research challenges imposed by the recognition of exposures of such rocks at the Earth’s surface and to appreciate their importance to an enhanced understanding of certain geodynamic processes that control the evolution of our planet.

References Bakun-Czubarow, N. (1991): On the possibility of occurrence of quartz pseudomorphs after coesite in the eclogite-granulite rock series of the Zlote Mountains in the Sudetes (SW Poland). Arch. Mineral., 47:5–16. Bakun-Czubarow, N. (1992): Quartz pseudomorphs after coesite and quartz exsolutions in eclogitic omphacites of the Zlote Mountains in the Sudetes (SW Poland). Arch. Mineral., 48:3–25. Caby, R. (1994): Precambrian coesite from northern Mali: first record and implications for plate tectonics in the trans-Saharan segment of the Pan-African belt. Eur. J. Mineral., 6:235–244. Carswell, D.A. & Zhang, R.Y. (1999): Petrographic characteristics and metamorphic evolution of ultra-high pressure eclogites in plate-collision belts. Int. Geol. Rev., 41:781–798. Chopin, C. (1984): Coesite and pure pyrope in high grade blueschists of the Western Alps: a first record and some consequences. Contrib. Mineral. Petrol., 86:107–118. Chopin, C., Henry, C. & Marchand, A. (1991): Geology and petrology of the coesite-bearing terrain, Dora Maira Massif, Western Alps. Eur. J. Mineral., 3:263–291. Cuthbert, S.J., Carswell, D.A., Krogh Ravna, E.J. & Wain, A. (2000): Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides. Lithos, 52:165–195. Ghiribelli, B., Frezzotti, M.L. & Palmeri, R. (2001): Coesite in eclogites of the Lanterman Range (Antarctica): evidence from textural and Raman spectroscopy studies. Eur. J. Mineral., 14:355–360. Gilotti, J.A. & Krogh Ravna, E.J. (2002): First evidence for ultrahigh-pressure metamorphism in the NorthEast Greenland Caledonides. Geology, 30:551–554. Leech, M.L. & Ernst, W.G. (1998): Graphite pseudomorphs after diamond? A carbon isotope and spectroscopic study of graphite cuboids from the Maksyutov Complex, south Ural Mountains, Russia. Geochim. Cosmochim. Acta, 62:2143–2154. Liu, L., Wang, Y., Sun, Y., Xiao, P., Chen, D., Luo, J. & Che, Z. (2001): The discovery and significance of magnesite-bearing garnet lherzolite from Altun Eclogite Zone, Western China. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 79–80. Liu, L., Sun, Y., Wang, Y., Chen, D.L., Luo, J.H. & Zhang, A.D. (2003a): Ultrahigh-P evidence for country rocks of the garnet-bearing lherzolite in Altyn Tagh – exsolution of clinopyroxene in the garnet of retrograde eclogite. In Alice Wain Memorial Western Norway Eclogite Field Symp., Selje, Western Norway, Abstr. Vol., 81.

8

D.A. Carswell & R. Compagnoni

Liu, L., Sun, Y., Chen, D., Zhang, A. & Luo, J. (2003b): Discovery of relic majoritic garnet in felsic metamorphic rocks of Qinling complex, north Qinling orogenic belt, China. In Alice Wain Memorial Western Norway Eclogite Field Symp., Selje, Western Norway, Abstr. Vol., 82. Massonne, H.-J. (1999): A new occurrence of microdiamonds in quartzo-feldspathic rocks of the Saxonian Erzgebirge, Germany, and their metamorphic evolution. In Gurney, J.J., Gurney, L.G., Pascoe, M.D. & Richardson, S.H. (eds.): Proc. 7th Int. Kimberlite Conf., Cape Town, Cape Town: Redroof Publ., 2 (The Nixon Vol.):533–539. Massonne, H.-J. (2001): First find of coesite in the ultrahigh-pressure metamorphic area of central Erzgebirge, Germany. Eur. J. Mineral., 13:565–570. Mposkos, E.D. & Kostopoulos, D.K. (2001): Diamond, former coesite and supersilicic garnet in metasedimentary rocks from the Greek Rhodope: a new ultrahigh-pressure metamorphic province established. Earth Planet. Sci. Lett., 192:497–506. Nasdala, L. & Massonne, H.-J. (2000): Microdiamonds from the Saxonian Erzgebirge, Germany: in situ microRaman characterisation. Eur. J. Mineral., 12:495–498. O’Brien, P.J., Zotov, N., Law, R.D., Khan, M.A. & Jan, M.Q. (2001): Coesite in Himalayan eclogite and implications for models of India-Asia collision. Geology, 29:435–438. Okay, A.I. (1993): Petrology of a diamond and coesite-bearing terrain: Dabie Shan, China. Eur. J. Mineral., 5:659–675. Palmeri, R., Ghiribelli, B., Talarico, F. & Ricci, C.A. (2003): Ultra-high-pressure metamorphism in felsic rocks: the garnet-phengite gneisses and quarzites from the Lanterman Range, Antarctica. Eur. J. Mineral., 15:513–525. Parkinson, C.D., Miyazaki, K., Wakita, K., Barber, A.J. & Carswell, D.A. (1998): An overview and tectonic synthesis of the pre-Tertiary very-high-pressure metamorphic and associated rocks of Java, Sulawesi and Kalimantan, Indonesia. Isl. Arc, 7:184–200. Parkinson, C.D., Motoki, A., Onishi, C.E. & Maruyama, S. (2001): Ultrahigh-pressure pyrope-kyanite granulites and associated eclogites in Neoproterozoic nappes of Southeast Brazil. UHPM Workshop 2001, Fluid/slab/mantle interactions and ultrahigh-P minerals, Waseda Univ., Tokyo, Abstr. Vol., 87–90. Reinecke, T. (1991): Very-high-pressure metamorphism and uplift of coesite bearing metasediments from the Zermatt-Saas zone, Western Alps. Eur. J. Mineral., 3:7–17. Săbău, G. (2000): A possible UHP-eclogite in the Leaota Mts. (South Carpathians) and its history from highpressure melting to retrograde inclusion in a subduction melange. Lithos, 52:253–276. Sachan, H.K. & Mukherjee, B.R. (2003): Metamorphic and fluid evolution of ultra-high metamorphosed (UHP) crust of Tso-Morari region, Ladakh, Himalaya, (India): constraints from mineral chemistry and fluid inclusions. In Alice Wain Memorial Western Norway Eclogite Field Symp., Selje, Western Norway, Abstr. Vol., 124–125. Shatsky, V.S., Sobolev, N.V. & Vavilov, M.A. (1995): Diamond-bearing metamorphic rocks of the Kokchetav Massif (Northern Kazakhstan). In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 427–455. Smith, D.C. (1988): A review of the peculiar mineralogy of the “Norwegian Coesite Eclogite Province”, with crystal-chemical, petrological, geochemical and geodynamic notes and an extensive bibliography. In Smith, D.C. (ed.): Eclogites and eclogite-facies rocks. Amsterdam: Elsevier, 1–206. Sobolev, N.J. & Shatsky, V.S. (1990): Diamond inclusions in garnet from metamorphic rocks: a new environment for diamond formation. Nature, 343:742–746. Song, S.G., Yang, J.S., Liou, J.G. & Shi, R.D. (2003): Metamorphic evolution of the coesite-bearing ultrahighpressure terrane in the North Qaidam, northern Tibet, NW. China. J. Metamorph. Geol., 21:613–644. Tagiri, M., Yano, T., Bakirov, A., Yakajima, T. & Uchiumi, S. (1995): Mineral parageneses and metamorphic P-T paths of ultrahigh-pressure eclogites from Kyrghyzstan Tien-Shan. Isl. Arc, 4:280–292. Treloar, P.J., O’Brien, P.J., Parrish, R.R. & Khan, M.A. (2003): Exhumation of early Tertiary, coesite-bearing eclogites from the Pakistan Himalaya. J. Geol. Soc. London, 160:367–376. Van der Klauw, S.N.G.C., Reinecke, T. & Stöckhert, B. (1997): Exhumation of ultrahigh-pressure metamorphic oceanic crust from Lago di Cignana, Piedmontese zone, western Alps: a structural record in metabasites. Lithos, 41:79–102.

Introduction with review of UHPM

9

Wain, A. (1997): New evidence for coesite in eclogite and gneisses: defining an ultrahigh-pressure province in the Western Gneiss Region of Norway. Geology, 25:927–930. Wang, X. & Liou, J.C. (1991): Regional ultrahigh-pressure coesite-bearing eclogitic terrane in central China: evidence from country rocks, gneiss, marble and metapelite. Geology, 19:933–936. Wang, X., Liou, J.G. & Mao, H.K. (1989): Coesite-bearing eclogites from the Dabie mountains, central China. Geology, 17:1085–1088. Wang. X., Zhang, R. & Liou, J.G. (1995): UHPM terrane in East Central China. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 356–390. Yang, J., Xu, Z., Zhang, J., Song, S., Wu, C., Shi, R., Li, H. & Brunel, M. (2002): Early Palaeozoic North Qaidam UHP metamorphic belt on the north-eastern Tibetan plateau and a paired subduction zone. Terra Nova, 14:397–404. Zhang, R.Y. & Liou, J.G. (1998): Ultrahigh-pressure metamorphism of the Sulu terrane, eastern China: a prospective view. Cont. Geodyn., 3:32–53. Zhang, R.Y., Hirajima, T., Banno, S., Cong, B. & Liou, J.G. (1995): Petrology of ultrahigh-pressure rocks from the southern Su-Lu region, eastern China. J. Metamorph. Geol., 13:659–675. Zhang, L., Ellis, D.J. & Jiang, W. (2002): Ultrahigh-pressure metamorphism in western Tianshan: Part I. Evidence from inclusions of coesite pseudomorphs in garnet and from quartz exsolution lamellae in omphacite in eclogites. Am. Mineral., 87:853–860.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 2, 13–49

UHPM units in the Western Alps ROBERTO COMPAGNONI* and FRANCO ROLFO Dipartimento di Scienze Mineralogiche e Petrologiche, Università di Torino, Via Valperga Caluso 35, 10125 Torino, Italy; * e-mail: [email protected]

Introduction In the Western Alps, two tectonic units have unquestionably experienced ultrahigh pressure metamorphism (UHPM): the continental Brossasco-Isasca Unit of the southern Dora-Maira Massif, in which coesite was first reported by Chopin (1984), and the oceanderived Lago di Cignana Unit of the Piemonte zone, in which coesite was first reported by Reinecke (1991). In both units the UHPM recrystallisation, acquired during the early stages of the Alpine orogeny, is largely obliterated by a late Alpine greenschist facies retrogression, more pervasive in the felsic lithologies.

The Brossasco-Isasca Unit of the Dora-Maira Massif The Dora-Maira Massif (DMM), together with Monte Rosa and Gran Paradiso, belongs to the “Internal Crystalline Massifs” of the Pennine Domain of the Western Alps (Fig. 1). The DMM is a nappe pile composed of continent-derived tectonic units, which are locally separated by thin ocean-derived units of the Piemonte Zone (Sandrone et al., 1993). The first modern geologic and petrographic study of the DMM was done by Vialon (1966) who subdivided the massif into three main lithotectonic units: the western and topmost “Sampeyre-Dronero Unit”, the “Polymetamorphic Unit”, and the underlying “Pinerolo Unit” (Fig. 2). As to the southern DMM, after Chopin’s (1984) coesite discovery, a great deal of field and petrographic work has been done (Chopin et al., 1991; Henry, 1990; Henry et al., 1993; Compagnoni & Hirajima, 1991; Compagnoni et al., 1994; Compagnoni et al., 1995; Michard et al., 1993; Turello, 1993; Chopin & Schertl, 1999; Matsumoto & Hirajima, 2000 and Groppo, 2002). These studies have shown that the massif is a pile of imbricated thrust sheets, resulting from the Alpine tectonic juxtaposition and metamorphic reworking of slices of Variscan continental crust and its Triassic cover, locally separated by thin layers of calcschists and metaophiolites of the Piemonte Zone, i.e. derived from the floor of the Mesozoic Tethys ocean (Figs. 1 and 2). Most tectonic slices are composed of the same continental lithologic association, but are characterised by different early Alpine climax overprints. The following tectonic units have been distinguished in the southern DMM (Figs. 2 and 3): x Brossasco-Isasca Unit (BIU): a portion of pre-Alpine continental crust, consisting of Variscan amphibolite facies metamorphic basement intruded by late Variscan

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

14

R. Compagnoni & F. Rolfo

Fig. 1. Tectonic sketch map of the western Alps. 1: Jura, Helvetic Domain and external Penninic Domain. The dashed line contours the External Crystalline Massifs (AR: Argentera, P: Pelvoux, BD: Belledonne, MB: Mont Blanc–AiguillesRouges, AG: Aar–Gotthard). SB: Grand St. Bernard Zone, LPN: lower Penninic nappes. 2: Internal Crystalline Massifs of the Penninic Domain (MR: Monte Rosa, GP: Gran Paradiso, DM: Dora-Maira, V: Valosio). 3: Piemonte Zone (VM: Voltri Massif) and a) main ophiolites. 4: Austroalpine Domain (DB: Dent Blanche nappe, ME: Monte Emilius, SZ: Sesia Zone) and a) South Alpine Domain (SA: Southern Alps). 5: Helminthoid Flysch nappes (EU: Embrunais–Ubaye, AM: Maritime Alps). 6: Swiss Molasse (SM), Po Plain and Piemontese–Ligurian Tertiary basin. CL: Canavese line; SVL: Sestri–Voltaggio line, SF: Subalpine frontal thrust, PF: Penninic frontal thrust. Distribution of the Alpine ultrahigh pressure metamorphism: (1) Brossasco-Isasca Unit of the southern Dora-Maira Massif; (2) Zermatt–Saas Fee nappe at Lago di Cignana, upper Valtournenche. Inset: I: Italy; F: France; CH: Switzerland; TO: town of Torino.

UHPM units in the Western Alps

15

Fig. 2. Tectonic sketch map of the southern Dora-Maira Massif and the adjoining Piemonte Zone (after Monié & Chopin, 1991, modified by Compagnoni et al., in press). 1–6: Dora-Maira Massif. 1: Pre-Alpine crystalline basement with middle-T coesite eclogite facies Alpine overprint (Brossasco–Isasca Unit); 2: Pre-Alpine crystalline basement with low-T quartz eclogite facies Alpine overprint (a: San Chiaffredo Unit; b: Rocca Solei Unit); 3: Pre-Alpine crystalline basement and Permo–Carboniferous + Permo–Triassic cover with low-T eclogite facies Alpine overprint in the northern part; 4: Carboniferous (?) graphite-rich unit (“Pinerolo unit”) with epidote-blueschist facies Alpine overprint; 5: Upper Palaeozoic + lower Triassic units with epidote-blueschist facies Alpine overprint; 6: Mesozoic cover series and epicontinental calcschists (“Schistes lustrés”) with low-grade blueschist facies Alpine overprint. 7–8: Piemonte Zone. 7: Monviso ophiolite unit with low-T eclogite facies Alpine overprint; 8: oceanic metasediments (“Schistes lustrés”) with low-grade lawsonite-blueschist facies Alpine overprint. 9: Post-orogenic sediments.

granitoids. During the Alpine orogeny this formation experienced an earlier coesite eclogite facies metamorphism (T = 730 °C, P = 3.5 GPa); x San Chiaffredo Unit: a portion of pre-Alpine continental crust, consisting of a Variscan amphibolite facies metamorphic basement intruded by late Variscan granitoids. During the Alpine orogeny this formation experienced an earlier quartz eclogite facies metamorphism (T | 550 °C, P = 1.5 GPa); x Rocca Solei Unit: a portion of pre-Alpine continental crust, consisting of a Variscan amphibolite facies metamorphic basement intruded by late Variscan granitoids. During the Alpine orogeny this formation experienced an earlier quartz eclogite facies metamorphism (T | 550 °C, P | 1.5 GPa) (Matsumoto & Hirajima, 2000); x Pinerolo Unit: this unit, first defined by Vialon (1966), consists of micaschists more or less rich in graphite, locally containing garnet and chloritoid, with interlayers of quartzite and minor paragneiss. These rocks experienced during the

16

R. Compagnoni & F. Rolfo

Alpine orogeny an earlier epidote-blueschist facies metamorphism (T | 400 °C, P | 0.8 GPa; Avigad et al., 2003). It is important to point out that the above tectonic units have been overprinted by a late Alpine greenschist facies recrystallisation, which pervasively reworked and extensively obliterated the former HP or UHP metamorphic mineral assemblages. The Brossasco-Isasca Unit (BIU), named from the two main villages of Val Varaita where it is extensively exposed, is the unit in which Chopin (1984) first reported the occurrence of coesite from continental crust. This unit was later studied by Henry (1990), Michard et al. (1993), Henry et al. (1993), Turello (1993), Compagnoni et al. (1994), Chopin & Schertl (1999) and Groppo (2002). Areal extent, internal setting and geologic interpretation of the unit, mainly derived from field and laboratory studies of Compagnoni & Hirajima (1991), Turello (1993), Compagnoni et al. (1995), Groppo (2002) and Compagnoni et al. (in press) are reported in Figure 3. The BIU, about 10×4×1 km in size, is tectonically sandwiched between the overlying Rocca Solei Unit and the underlying Pinerolo Unit (Fig. 3), which are characterised by Alpine quartz eclogite facies (ca. 550 °C and 0.5 GPa: Chopin et al., 1991; Schertl et al., 1991; Matsumoto & Hirajima, 2000) and epidote-blueschist facies (ca. 450 °C and 0.8 GPa: Chopin et al., 1991) conditions, respectively. Two lithostratigraphic complexes have been distinguished in the BIU by Compagnoni et al. (1995): a “Monometamorphic Complex”, mainly consisting of orthogneiss, derived from Alpine tectono-metamorphic reworking of late Variscan granitoids, and a “Polymetamorphic Complex”, derived from Alpine tectonometamorphic reworking of a Variscan amphibolite facies metamorphic basement (Fig. 3). The two complexes roughly correspond to the “gneiss and metagranite” and to the “varied formation” of Chopin et al. (1991). The Polymetamorphic Complex This complex has been referred to as “polymetamorphic” because it mainly consists of lithologies which underwent both Variscan and Alpine tectonic and metamorphic cycles. It consists of paragneiss and minor micaschist with eclogite and marble intercalations, and orthogneiss (Fig. 3). In this paper we will describe: 1) the paraschist, orthogneiss and marble, preserving relics of pre-Alpine mineralogies and structures; 2) those lithologies better preserving UHPM mineral assemblages; 3) a few lithologies characterised by unusual UHPM mineral assemblages. The pre-Alpine setting of the BIU is still understandable thanks to the preservation of rock portions that escaped pervasive Alpine deformation and metamorphism. These portions are relics of pre-Alpine metamorphics and of pre-Alpine intrusive contacts, which have been locally found between the Monometamorphic and Polymetamorphic Complexes (Biino & Compagnoni, 1992; Compagnoni & Hirajima, 1991; Compagnoni et al., 1994, 1995).

UHPM units in the Western Alps

17

Fig. 3. Simplified geological map of the coesite-bearing “Brossasco-Isasca Unit” (from Compagnoni et al., 1994 and Compagnoni et al., in press). Undifferentiated units: graphite-rich schists and metaclastics of the epidote-blueschist facies “Pinerolo Unit”; lower tectonic unit (“San Chiaffredo Unit”) with pre-Alpine basement rocks overprinted by Alpine low-T eclogite facies; upper tectonic unit with pre-Alpine basement rocks, overprinted by Alpine low-T eclogite facies metamorphism. Alluvial deposits in white.

Relics of pre-Alpine metamorphic rocks Variscan paragneiss. The best-preserved remnant of a pre-Alpine metamorphic rock, mostly characterised by pseudomorphic microstructures, has been discovered at Case Garneri, halfway between Brossasco and Isasca (Fig. 3). Macroscopically, the rock is very dark and includes irregular cm- to dm-sized light portions of metatects. Under the microscope, the dark portions consist of relict garnet, nematoblasts of former sillimanite pseudomorphically replaced by kyanite aggregates, red-brown biotite partly replaced by phengite r rutile and surrounded by new coronitic garnet, lens-like aggregates of patchy albite + zoisite r kyanite with jadeite relics after former plagioclase, and polygonal granoblastic quartz aggregates. Such mineral relics and pseudomorphs are

18

R. Compagnoni & F. Rolfo

embedded in a dark grey matrix, mainly consisting of fine-grained garnet + kyanite r phengite r biotite, interpreted as derived from former contact metamorphic cordierite. Accessory minerals are very abundant: graphite, ilmenite completely replaced by rutile and a garnet corona, apatite, zircon, monazite and tourmaline. In some quartzundersaturated sites, pseudomorphs consisting of corundum r rutile aggregates, surrounded by an inner kyanite and an outer garnet corona occur, which may have derived from former spinel. The rough pre-Alpine rock foliation is defined by the preferred orientation of the graphite flakes and sillimanite nematoblasts (Fig. 4a) and by the preferred dimensional orientation of the former plagioclase blasts. Pre-Alpine garnet encloses inclusions of plagioclase, quartz, biotite, rare sillimanite, and accessory apatite, tourmaline, and dusty graphite. This fine-grained graphite usually defines a closely spaced Si, often discordant with respect to Se (Fig. 4b). Pre-Alpine garnet, which usually

Fig. 4. (a) Bundle of Variscan amphibolite facies sillimanite (white, Sil), pseudomorphically replaced by a UHP kyanite aggregate, embedded in a dark matrix mainly consisting of fine grained UHP garnet, derived from late Variscan contact metamorphic cordierite. Contact metamorphosed Variscan amphibolite facies paraschist from the Polymetamorphic Complex of the Brossasco-Isasca Unit. Sample DM493. Plane polarised light. (b) Relict Variscan amphibolite-facies garnet corroded and replaced by late Variscan contact metamorphic cordierite (Crd) in turn replaced by a very fine grained aggregate (dark grey) of Alpine garnet + kyanite (Grt + Ky). Metapelite from the Polymetamorphic Complex of the UHP Brossasco-Isasca Unit. Sample DM495. Plane polarised light. (c) Metagranitoid collected close to the contact with the country metamorphic basement (“granitoid marginal facies”). Euhedral crystals of igneous plagioclase (transformed to very fine grained aggregate of jadeite + zoisite ± kyanite: dark with high relief), cordierite (replaced by a very fine grained aggregate of kyanite + garnet ± biotite: black with garnet corona), and biotite (partly replaced by phengite) are set in a matrix of polygonal granoblastic quartz derived from inversion of coesite after igneous quartz. UHP Brossasco-Isasca Unit. Sample DM 1116. Plane polarised light. (d) Marginal facies of the late Variscan metagranite, BIU, crowded with contact metamorphosed xenoliths of the country Variscan paraschists. Case Bastoneri, Gilba valley.

UHPM units in the Western Alps

19

Table 1. Simplified metamorphic evolution of the relict regionally and thermally metamorphosed Variscan paraschists from the Polymetamorphic Complex of the Brossasco-Isasca Unit, Dora-Maira Massif (modified from Compagnoni et al., 1994; Stellato, 1999). Variscan regional Late Variscan contact metamorphism metamorphism Garnet Biotite

(cordierite) (cordierite) Biotite

K-feldspar

K-feldspar

(Plagioclase)

(Plagioclase)

(Sillimanite) (Quartz) Muscovite (Ilmenite) Graphite Apatite Monazite Zircon Tourmaline

(Andalusite) (Quartz) Muscovite (Spinel?) (Ilmenite)

(+ H2O) (+ H2O)

Early Alpine UHP metamorphism

Alpine greenschist facies retrogression

Grt + Phe + Ky Grt + Phe + Ky Grt + Phe K-feldspar Phe ± Qtz Jd + Qtz + Zo Jd + Qtz + Ky Ky aggregate Ky aggregate (Coesite)

Bt + Chl sericite + Chl Bt + Chl microperthite Ab/Ol + CZo + Wm sericite sericite Quartz

Grt + Crn Grt + Rut

sericite Tit + Chl

Epidote

Epidote

Bracketed minerals are not preserved, but their former occurrence is unambiguously inferred on the basis of microstructure and/or mineralogy of pseudomorphs.

exhibits a strongly corroded shape and appears extensively transformed into contact cordierite, is in turn replaced by a dark grey aggregate of garnet and minor kyanite. The light portions, which locally preserve an igneous microstructure, consist of plagioclase replaced by jadeite (usually albitised) and zoisite, K-feldspar partly replaced by phengite-quartz intergrowths, granoblastic polygonal quartz aggregates, cordierite completely replaced by garnet and kyanite, and minor biotite and muscovite. Graphite is ubiquitous and locally occurs as large deformed flakes. Around the leucocratic portions biotite aggregates and kyanite aggregates after a former prismatic contact andalusite locally occur. Coesite has not been found: however, the ubiquitous presence of granoblastic polygonal quartz aggregates, locally pseudomorphous after a pre-Alpine single crystal, is considered to be the unambiguous evidence of the former occurrence of early Alpine coesite. On the basis of mineral compositions and microstructure, the inferred polymetamorphic evolution, summarised in Table 1, includes: 1) a pre-Alpine, most likely Variscan, medium P regional metamorphic event, which reached the upper amphibolite facies and developed the leucocratic metatects; 2) a pre-Alpine low P static recrystallisation, most likely related to the thermal metamorphism connected to the intrusion of the late Variscan granitoids, which led to the widespread development of contact biotite, andalusite, cordierite; 3) an early Alpine UHP static recrystallisation, characterised by pseudomorphous and coronitic reactions with coesite, jadeite, garnet, zoisite, phengite, kyanite, and rutile; 4) a late Alpine greenschist facies retrogression, responsible for the complete coesite inversion to polygonal granoblastic quartz and the replacement of jadeite by albite.

R. Compagnoni & F. Rolfo

20

Relics of Variscan augengneiss. Within the phengite-rich orthogneiss (Fig. 3) a relict pre-Alpine augengneiss consists of “plagioclase” and K-feldspar porphyroclasts, and of quartz aggregates several millimetres long wrapped around by a discontinuous foliation defined by large flakes of former Ti-rich biotite. The original K-feldspar is preserved, whereas biotite is pseudomorphically replaced by Ti-poor biotite + phengite + garnet r epidote r rutile, and plagioclase by albite (most likely derived from retrogression of peak jadeite) + zoisite r white mica, respectively. The quartz aggregates most likely derive from larger pre-Alpine quartz grains converted to coesite under Alpine UHPM conditions, then inverted back to quartz aggregates during decompression, such as observed in the metagranitoids of the Monometamorphic Complex (see later on). This lithology is very similar to the monometamorphic augengneiss: however, its microstructure, and particularly the presence of a foliation defined by pseudomorphs after original large Ti-rich biotite flakes and plagioclase, points to a regional medium grade (amphibolite facies) metamorphism incompatible with the greenschist facies mineralogy, and, therefore, must predate the Alpine recrystallisation. As a consequence, the medium to fine grained phengite-rich orthogneiss of the Polymetamorphic Complex derives from a pre-, or early-, Variscan granitoid intrusion, which underwent the Variscan amphibolite facies metamorphism preceding the Alpine orogeny. Relics of Variscan marble. The dolomite-rich marble consists of bluish corundum (Crn II) + clinochlore + dolomite + calcite (after aragonite) + rutile. This marble contains relics of a number of pre-Alpine minerals: bluish-green spinel, colourless corundum (Crn I), ilmenite, dolomite, and perovskite. The former occurrence of an anorthite-rich plagioclase, completely consumed by the Alpine prograde metamorphic reactions, is inferred from the composition of the Alpine minerals, and especially the significant Na content in margarite (ephesite20–40). The evolution and mineral transformation of such marble is summarised in Table 2. Basement lithologies preserving UHPM mineral assemblages In addition to the pre-Alpine relics, the Polymetamorphic Complex mainly consists of crustal lithologies totally recrystallised during the Alpine orogeny, which include Table 2. Simplified metamorphic evolution of the dolomite-rich marble with relics of pre-Alpine minerals from the Polymetamorphic Complex of the Brossasco-Isasca Unit, Dora-Maira Massif (modified from Castelli et al., 1999). Pre-Alpine metamorphic minerals

Alpine UHP peak metamorphic minerals

Alpine greenschist facies metamorphic minerals

bluish green spinel colourless corundum ilmenite (calcite) titanite (Ca-plagioclase)

bluish corundum + chlorite

Na-margarite Ÿ chlorite + calcite chlorite titanite calcite titanite

dolomite

rutile + lamellar högbomite (aragonite) rutile consumed by Na-margarite producing reactions dolomite

dolomite

Bracketed minerals are not preserved, but their former occurrence is unambiguously inferred on the basis of microstructure and/or mineralogy of pseudomorphs.

UHPM units in the Western Alps

21

gneisses, micaschists, minor marbles and eclogites (Chopin et al., 1991; Compagnoni et al., 1994, 1995). Most lithologies show mainly greenschist facies mineral assemblages, but the following still preserve traces of UHP metamorphism. Jadeite-kyanite-almandine-phengite micaschist. The most peculiar basement lithology, which best preserves the peak UHPM mineral assemblage, is a micaschist, consisting of quartz (/coesite), phengite (3T polytype with Si = 3.5 p.f.u.), porphyroblastic garnet and kyanite, jadeite, minor zoisite, yellow-greenish tourmaline, rare bearthite (Ca2AlPO4(OH)2: Chopin et al., 1993; Chopin & Ferraris, 2003), and accessory rutile and apatite. Fresh jadeite (Jd90–95) is found only as inclusions in garnet. Porphyroblastic garnet, up to 2 cm across, is zoned from Alm80Prp20 in the core to Alm60Prp40 in the rim, and consists of an inner zone crowded with mono- or bimineralic inclusions of chlorite, chloritoid (up to about 50 mol% of the Mg-chloritoid endmember), staurolite, paragonite, quartz, kyanite, rutile, and rare ilmenite, and an outer zone with only rare crystals of jadeite, kyanite, and coesite (Chopin et al., 1991). Both the distribution of mineral inclusions and garnet zoning (increasing pyrope content from core to rim) point to garnet growth during a prograde P–T trajectory from the quartz to the coesite stability field. A second generation of acicular kyanite, most likely developed during decompression, locally occurs. During retrogression, almandine is altered to biotite and/or chlorite r clinozoisite, kyanite to paragonite r margarite, jadeite to very fine grained intergrowths of albite + paragonite, and this rock type is converted to a greenschist facies micaschist, mainly consisting of quartz + albite + biotite + chlorite + white mica r clinozoisite/epidote + titanite. Phengite-jadeite-almandine-quartz (/coesite) granofels. This lithotype occurs as dm to m thick greenish layers both within the Monometamorphic and the Polymetamorphic Complexes, and is distinguished from the garnet-jadeite-kyanitequartz (/coesite) granofels for its lack of kyanite and the garnet compositionally rich in almandine. The UHPM peak assemblage includes coesite, jadeite (Jd95), almandine-rich garnet (Alm70–80 Prp10–20Grs5–20), minor phengite (Si3.25–3.41), and accessory apatite, rutile, and zircon. The granofels exhibits a polyphase retrogression, studied in detail by Hirajima & Compagnoni (1993), marked by the successive growth of Na-pyroxenes, progressively decreasing in the jadeite component from omphacite (Jd55) to aegirine (Jd0), in equilibrium with albite/oligoclase, and amphiboles ranging in composition from ferronyböite to taramite. On the grounds of composition and field occurrence, this rock most likely derives from a dyke of granitoid composition (Chopin et al., 1991). Eclogite. Numerous eclogite boudins, from a few cm to several dm across, occur throughout the paraschist. They are fine- to medium-grained and consist of garnet and omphacite with accessory rutile, apatite, rare zircon and exceptionally graphite. Locally, phengite and quartz may be very abundant, and kyanite or zoisite may be present. The eclogites are locally banded with alternating garnet-rich and omphacite-rich layers and often show a folded foliation defined by phengite and/or a strong lineation defined by omphacite. Prograde Na-Ca-amphiboles of katophoritic to taramitic and barroisitic compositions locally occur as inclusions in garnet and omphacite. A zoned poikiloblastic amphibole frequently overgrows the eclogitic foliation: the amphibole core, light

22

R. Compagnoni & F. Rolfo

pinkish-purple in thin section, is intermediate in composition between barroisite and edenite, while the greenish rim is taramitic or pargasitic. Often eclogites are cut by coarse-grained discordant veins, which consist of idioblastic omphacite, minor interstitial quartz and accessory apatite. In other eclogites, discontinuous layers or pockets of coarse-grained quartz, omphacite and accessory apatite are common, which appear to have derived from folding and transposition of former discordant metamorphic veins. In spite of the ubiquitous occurrence of quartz, the local presence of small coesite inclusions in omphacite indicates that the climax eclogite assemblage crystallised under UHP metamorphic conditions. The presence of sagenitic rutile in garnet of the phengitebearing eclogite suggests its derivation from a biotite-bearing protolith, most likely a Variscan biotite amphibolite. A detailed petrologic study of the eclogites was performed by Kienast et al. (1991). Nowlan et al. (2000), by applying conventional geothermobarometry to the growth zoning of coexisting garnet-omphacite-phengite assemblages, defined a prograde stage I constrained by three points at 1.5 GPa/500 °C, 2.5 GPa/570 °C, and 3.2 GPa/650 °C; a peak stage II constrained at 3.7 GPa/790 °C, and a retrograde stage III constrained at 2.4 GPa/650 °C and 1.4 GPa/650 °C, in good agreement with the P–T path inferred from the petrologic study of other lithologies. Most eclogites show a weak to strong retrogression, with development of greenschist facies minerals, such as blue-green to green amphiboles, chlorite, ilmenite and titanite. Marble. Marble occurs within the paraschists as lenses from a few metres to tens of metres long and from a few dm to several metres thick. The largest lens, about 800 m long and 70 m thick, is exposed at Costa Monforte (along the geologic cross-section of Fig. 3). A compositional banding may occur, which consists of alternating cm- to dmthick layers of micaschist, Ca-silicate granofels and boudinaged eclogite. Relatively common is a marble with cm-sized garnet porphyroblasts, showing a colour zoning from pink to deep red-brown. In thin section, marbles show a wide spectrum of microstructures from coarsegrained granoblastic polygonal through mortar to very fine-grained ultramylonitic. In addition to calcite and/or dolomite, marbles may also contain clinopyroxene (diopside to omphacite), phengite, garnet, epidote (zoisite, Fe-poor and Fe-rich epidote), K-feldspar, colourless to blue-green amphiboles, minor talc, Mg-chlorite, phlogopite or biotite, rare quartz, and accessory rutile, titanite, tourmaline, apatite, opaque ores, zircon and graphite flakes up to 1 mm long. The UHPM assemblage probably comprised aragonite (now calcite), dolomite, r one or more of the following minerals: coesite (now quartz), Mgchlorite, clinopyroxene, phengite, garnet, zoisite, and accessory rutile. A few Mg-rich marbles consist of forsterite r titanian clinohumite r phlogopite r talc. Locally, intergrowths of microcline + Fe-rich epidote + clinopyroxene r phengite occur, which are most likely pseudomorphs after a former, possibly pre-Alpine, unidentified mineral. Most UHPM peak minerals are retrogressed to greenschist facies mineral assemblages: diopside is replaced by tremolite, omphacite by a blue-green amphibole + albite symplectite, pink garnet by chlorite and red-brown garnet by blue-green amphibole + Fe-rich epidote, phengite by microcline + biotite, zoisite by Fe-poor to Fe-rich epidote, rutile by titanite.

UHPM units in the Western Alps

23

Rare blue-green amphibole + green-brownish biotite pseudomorphs after an unidentified mineral are surrounded by green Fe-stilpnomelane. A dolomite-poor marble from Costa Monforte, characterised by porphyroclastic garnet and green pyroxene, consists of calcite + garnets + pyroxene + quartz + highly celadonitic phengite (up to 65% Mg-Al-celadonite) + Fe-epidote (Zo45–51) + titanite with relics of rutile. Two types of garnet are evident, most likely developed within different microchemical domains: a reddish type and a pinkish type with compositions Alm53–57Grs16–26Adr14–23 and Grs36–54Alm39–52Adr0–8, respectively. The deep green pyroxene is diopside with Jd5–14 and Aeg4–16 (Castelli et al., 1999). Unusual lithologies characterised by UHPM mineral assemblages Kyanite-pyrope talcschist. East of Canova, on the left side of middle Vallone di Gilba, a light grey intercalation of talcschist is exposed within folded basement paraschists. The talcschist layer, about 30 cm thick, shows a nodular structure with cmsized closely packed nodules of fine-grained sheet silicates in a matrix of talc. Microscopically, the rock consists of roundish pseudomorphs after a pyrope-rich garnet in a scant matrix of talc (Tc I), kyanite (Ky I) and accessory rutile and opaque minerals. Skeletal Ky I locally includes talc and Mg-chlorite (Chl I). The pseudomorphs, which rarely preserve a relict garnet core, consist of a very fine grained felty aggregate of Mgchlorite (Chl II), kyanite (Ky II) and talc (Tc II). Brownish aggregates locally occur, pseudomorphous after a former unidentified UHP mineral. Sodic whiteschist. This rock has been found only as a boulder in the debris close to Case Ramello, Martiniana Po, and described by Kienast et al. (1991) as a variety of whiteschist with its composition well represented by the NMASH system. It is coarsegrained and consists of the UHPM peak assemblage: pyrope-rich garnet (Prp76Alm22Grs2) + jadeite (Jd82Di18) + glaucophane (close to the pure Mg-Al endmember) + phengite (Si3.6) + coesite + rutile. This unusual lithotype is very interesting since its mineral assemblage supports experimental data which indicate that pure glaucophane is stable in the NMASH system up to about 3.3 GPa and 750 °C (Holland, 1980). The Monometamorphic Complex This Complex has been referred to as “monometamorphic”, since it includes rocks which only experienced the Alpine metamorphic recrystallisation. It mainly consists of augengneiss, grading to medium- to fine-grained mylonitic orthogneiss, which locally contains relics of metagranitoid and lens-like layers of “pyrope-bearing whiteschist” with rare “dyke”-looking bands of a garnet-jadeite-kyanite-quartz granofels. UHPM rocks derived from late Variscan intrusives Orthogneiss. Orthogneiss makes over 90 vol% of the Monometamorphic Complex, and its appearance is very variable according to the degree of deformation and the presence or lack of porphyroclasts: augengneiss, characterised by the presence of cm-sized K-feldspar porphyroclasts, grades into medium- to fine-grained orthogneiss

24

R. Compagnoni & F. Rolfo

with a local mylonitic fabric. The orthogneiss usually exhibits a greenschist facies mineralogy, including quartz, albite, biotite, chlorite, low-celadonite phengite, epidote, titanite, and accessories. Foliated orthogneiss preserving the UHPM mineral assemblage has never been observed, the only evidence of the early coesite eclogite facies metamorphism being the local presence of porphyroclastic highly celadonitic phengite, wrapped around by the greenschist facies foliation, and aggregates of titanite (after former rutile) + grossular-rich garnet. The lack of coesite eclogite facies orthogneiss, a general feature in UHPM belts, may be explained by the following considerations: during exhumation felsic rocks undergo a pervasive reworking due to the significant volume change involved by the coesite to quartz inversion, which favours the extensive retrogression; deformed rocks, which acquired a pervasive foliation during UHP metamorphism, are preferentially deformed (and retrogressed) during later evolution. The gradual transition all over the complex from the undeformed metagranitoid to the foliated orthogneiss suggests that the latter lithology derive from the Alpine tectonic and metamorphic reworking of the former (Biino & Compagnoni, 1992). Relics of metagranitoids. The metagranitoids have been derived from late Variscan protoliths as suggested by a zircon U/Pb radiometric age of 304 r 3 Ma determined by Paquette et al. (1999), and a SHRIMP age of 275 Ma obtained by Gebauer et al. (1997) on the relict portions, evident under cathodoluminescence, of oscillatory zoned igneous zircons. Several tens of lens-like bodies of relict metagranitoids, from a few metres to several tens of metres in length, have been found in the Monometamorphic Complex (Fig. 3). They locally include aplitic and pegmatitic dykes, cognate microgranular enclaves, and centimetre- to metre-sized xenoliths of the country rocks, which indicate the local preservation of original intrusive contacts (Fig. 4d). The massive portions of the little deformed metagranitoids, several metres across, are dissected by dm-thick shear zones, which show a well-developed mylonitic foliation and a significant grain size reduction, accompanied by a general pervasive greenschist facies recrystallisation. The transition from the massive metagranite to the surrounding augengneiss is gradual and occurs over a distance from a few centimetres to several metres by progressive development of a metamorphic foliation. The metagranitoids are surrounded on all sides by orthogneisses, locally with augen structure, which are characterised by greenschist facies mineral assemblages. This suggests either that most of the augengneiss deformation was acquired during the late Alpine event or that an earlier, most likely early Alpine, eclogite facies foliation was pervasively reworked during the greenschist facies event. In the surrounding orthogneiss, the only evidence for the former UHP recrystallisation is the local occurrence of phengite porphyroclasts with high-celadonite cores, wrapped around by the main Alpine greenschist facies foliation, and the probable original occurrence of the HP assemblage grossular garnet + rutile (now titanite: Chopin et al., 1991). The metagranitoids, ranging in composition from biotite granite to biotite granodiorite, preserve the original igneous porphyritic structure, which is characterised by the presence of K-feldspar phenocrysts up to several cm long. At outcrop scale, the metagranitoids look like non-metamorphic rocks, the metamorphic recrystallisation

UHPM units in the Western Alps

25

having produced only pseudomorphous and coronitic reactions. On hand specimen, however, although K-feldspar phenocrysts are perfectly preserved, the igneous quartz, originally clear and greyish in colour, shows a milky and sugary appearance, having been replaced by a granoblastic polygonal aggregate of fine-grained quartz crystals (Fig. 5a). Under the microscope, the metagranite exhibits an igneous hypidiomorphic fabric, only partially obliterated by the Alpine metamorphic recrystallisations (Fig. 5b). The common modal composition is granodioritic, the original igneous assemblage consisting of quartz, plagioclase, K-feldspar, biotite and accessory apatite, zircon and tourmaline. Garnet xenocrysts are locally observed. Except for K-feldspar, the igneous minerals have been completely replaced by Alpine metamorphic phases. The igneous biotite is locally preserved as a microstructural relic (Fig. 5c): however, its significant increase in Mg with respect to igneous compositions indicates a metamorphic equilibration (Biino & Compagnoni, 1992). Usually, biotite is pseudomorphically replaced by a high-celadonite Ti-rich phengite + rutile and surrounded by a continuous

Fig. 5. (a) Relict late Variscan porphyritic metagranite from the Monometamorphic Complex, BIU, collected close to the Varaita bridge south of Brossasco village. The igneous structure is perfectly preserved, and no evidence of the UHP early Alpine metamorphism can be found. (b) Metagranitoid of the Brossasco–Isasca Unit. Note in the centre a fine-grained granoblastic aggregate of quartz derived from inversion of coesite, which replaced the igneous quartz during UHPM. Sample DM 496. Crossed polarisers. Symbols for minerals after Kretz (1983) with implements by Bucher & Frey (2002). (c) Detail of a metagranite, which shows the transformation of an igneous biotite (Bt) partly into a re-equilibrated Mg-enriched composition and partly into phengite. Relict biotite is surrounded by a continuous garnet corona (Grt) and a second corona against K-feldspar (Kf), which consists of a very fine-grained dactylitic intergrowth of phengite and quartz (from former coesite). UHP Brossasco-Isasca Unit. Sample DM1086. Crossed polarisers. (d) Backscattered electron SEM image of metagranodiorite from the Brossasco-Isasca Unit, southern Dora-Maira Massif. Amoeboid K-feldspar associated with kyanite crystals in a pseudomorph after plagioclase. Scale bar is 10 µm.

26

R. Compagnoni & F. Rolfo

garnet corona (Fig. 5c). At the contact with the original igneous plagioclase site, coronitic garnet is asymmetrically zoned with a pyrope- and almandine-rich composition on the biotite side and a grossular-rich composition on the plagioclase side. The igneous plagioclase has been pseudomorphically replaced by jadeite (up to Jd96) + zoisite + quartz + kyanite + K-feldspar. Because of a ubiquitous greenschist facies overprint, jadeite is mostly retrogressed to albite + white mica. Fresh jadeite is usually preserved only in pseudomorphs after igneous plagioclase, armoured within the igneous K-feldspar. Unexpectedly, in the relict jadeite about 10 mol% of the Ca-Eskola end-member has been identified (Bruno et al., 2002). A SEM-EDS study of pseudomorphous and coronitic reactions has also revealed the presence, within the pseudomorphs after plagioclase, of K-feldspar patches and of tiny kyanite crystals (Fig. 5d). Within such pseudomorphs and close to the biotite site, small grossular-rich idioblastic garnets developed instead of zoisite. Relics of igneous K-feldspar up to 3 cm long are preserved and include pseudomorphs after plagioclase, biotite and quartz. Along deformation bands or around mineral inclusions, K-feldspar recrystallises to polygonal granoblastic aggregates. Around biotite included in K-feldspar, a second reaction rim surrounds the garnet corona: it consists of an intergrowth of vermicular quartz and highly celadonitic phengite, growing from the garnet corona towards K-feldspar (Fig. 6a). Coesite relics have never been found in the metagranitoids: however, the presence of polygonal granoblastic quartz aggregates (Fig. 5b) after the igneous quartz crystals have been considered as the inversion products from peak coesite, which inverted to quartz during exhumation (Biino & Compagnoni, 1992; Compagnoni et al., 1994). The polyphase Alpine metamorphic evolution of the metagranitoids, as inferred from microstructural relationships, is summarised in Table 3. A detailed petrographic study and equilibrium thermodynamic modelling of these UHP coronitic and pseudomorphous reactions was performed by Biino & Compagnoni (1992) and by Bruno et al. (2001), respectively: the inferred metamorphic conditions are in good agreement with the peak P–T conditions estimated for the whole BIU. Relics of igneous intrusive contacts with basement xenoliths. Relics of intrusive contacts between the late Variscan granitoids and the amphibolite facies Variscan basement are locally preserved along the boundary between the Monometamorphic and the Polymetamorphic Complexes. The best example is preserved close to Case Bastoneri, about 1 km NW of Brossasco on the left side of Vallone di Gilba (see Compagnoni et al., 1994), where a porphyritic metagranitoid is exposed, which includes numerous xenoliths, from a few cm to several m across, of metapelite and metabasite (Fig. 4d). The metapelite xenoliths mainly consist of fine-grained randomly oriented phengite, which includes: kyanite aggregates pseudomorphous after prismatic sillimanite; coronitic garnet with local sagenitic rutile inclusions after equilibrated biotite; pre-Alpine porphyroblastic garnet surrounded by thin coronas of Alpine garnet; rutile aggregates rimmed with a garnet corona (most likely after primary ilmenite); jadeite-rich pyroxene (or its albite-bearing retrogressive products) with small zoisite crystals after pre-Alpine plagioclase; polygonal granoblastic quartz aggregates from

UHPM units in the Western Alps

27

Table 3. Simplified Alpine metamorphic evolution of the relict metagranitoids from the Monometamorphic Complex of the Brossasco-Isasca Unit, Dora-Maira Massif (modified from Biino & Compagnoni, 1992; Bruno et al., 2001, 2002). Pre-Alpine igneous mineralogy K-feldspar

(plagioclase)

biotite

(quartz) (Ti-[Fe]-rich phase) apatite zircon tourmaline garnet (xenocryst)

(+ H2O)

Early Alpine peak UHP mineralogy

Late Alpine greenschist facies mineralogy

K-feldspar high-Cel dactylitic Phe + (coesite) jadeite (with Ca-Esk) + zoisite + (coesite) + K-feldspar + kyanite ± Ca-rich Grt (close to Bt) high-Cel Phe + coronitic garnet + rutile + (coesite)-phengite dactylitic intergrowth

microperthite

(coesite) rutile

quartz albite/oligoclase clinozoisite quartz low-Cel Phe

chlorite red Bt Ÿ green Bt titanite

quartz (palisade and polygonal granoblastic) ilmenite Ÿ titanite

Bracketed minerals are not preserved, but their former occurrence is unambiguously inferred on textural and mineralogical grounds.

coesite inversion; aggregates of small garnets with minor phengite, most likely after preAlpine cordierite (produced by contact metamorphism?). Where the metapelite xenoliths are finely dismembered, their rims are enriched in euhedral plagioclase and peculiar pseudomorphs after euhedral cordierite (Fig. 4c), suggesting a partial assimilation of the wall rocks by the crystallising magma. In the portions more pervasively recrystallised during the early Alpine metamorphic event, a grain size coarsening of the UHP minerals progressively obliterates the pre-Alpine textures: the rock is, therefore, converted to a phengite-garnet granofels with accessory rutile and local Na-pyroxene, quartz and zoisite. The metabasite xenoliths are converted to quartz (/coesite)-phengite eclogite with accessory rutile. In the eclogite, the only recognised pre-Alpine relics are large titanite grains, partly replaced by rutile r omphacite intergrowths. The common occurrence of “sagenitic” rutile in garnet indicates its derivation from pre-Alpine Ti-bearing biotite. In a sample, rich in a pale-coloured phlogopite, a small geode was found, its walls covered with idioblastic omphacite and filled with granoblastic to “columnar” quartz, derived from former coesite.

28

R. Compagnoni & F. Rolfo

Fig. 6. (a) Detail of a metagranitoid showing an igneous biotite included in a K-feldspar which is pseudomorphically replaced by a single phengite crystal and surrounded by a thin garnet corona and a dactylitic phengite + quartz reaction rim. In the upper right, a quartz aggregate developed after euhedral igneous quartz from inversion of UHPM coesite. Sample DM 60. Crossed polarisers. (b) Idioblastic pyrope megablasts, ca. 20 cm across, from the Case Parigi whiteschists of the Monometamorphic Complex, BIU (Case Parigi). The light garnet colour is due to its unusual composition, very close to the pure pyrope end-member. (c) Retrogressed pyrope crystals from the whiteschists of the Monometamorphic Complex, BIU, are pseudomorphically replaced by two different generations of kyanite (Ky1 and Ky2). Sample DM72. Plane polarised light. (d) Garnet-jadeite-kyanite-quartz(/coesite) granofels. The presence of the jadeite-kyanite assemblage indicates that P conditions, higher than the paragonite breakdown, have been reached because of the reaction: Pg = Jd + Ky + water. In the rock matrix, jadeite is pseudomorphically retrogressed to Ab + white mica, and coesite to granoblastic quartz (white). UHPM Brossasco-Isasca Unit. Sample DM442. Plane polarised light.

Pyrope-bearing whiteschist. The pyrope-bearing whiteschist, petrographically a kyanite-phengite-pyrope-quartz (/coesite) granofels, occurs within the orthogneiss of the Monometamorphic Complex as layers from a few centimetres to ca. 20 meters thick and from a few meters to hundreds of meters long (Fig. 3). The whiteschist mainly consists of pyrope embedded in a matrix of quartz (inverted from former coesite), 3T polytype phengite (phengite-3T) with exsolutions of quartz and talc (Ferraris et al., 2000), kyanite, minor talc, rare jadeite, and accessory rutile, zircon, and monazite. The size of pyrope grains ranges from a few mm up to about 20 cm across. Pyrope developed a trapezohedron habit, but it has been frequently rounded by later post-crystalline deformation (Fig. 6b) or possibly incomplete growth, and/or partial resorption (Chopin, pers. comm.). Pyrope is usually poorly zoned with pale reddish core and pale pinkish rim, with rim composition up to Prp96–98; however, a few superzoned garnets have been locally found (see below). Pyrope megablasts are systematically crowded with prograde minerals, such as vermiculite, phlogopite, talc, Mg-chlorite, and rare phengite; minerals

UHPM units in the Western Alps

29

stable at peak conditions, such as coesite; and retrograde minerals, such as chlorite, phlogopite, kyanite and talc (the latter two formed close to a quartz-rich matrix). In addition to these relatively common minerals, pyrope megablasts also include rare or previously unknown minerals such as magnesiostaurolite, magnesiochloritoid, bearthite [(Ca2Al(PO4)2(OH)], magnesiodumortierite, ellenbergerite, phosphoellenbergerite, and wagnerite [(Mg2(PO4)(F,OH)] (for a review, see: Compagnoni et al., 1995; Chopin & Ferraris, 2003). Rare and localised “granulitic” phases, such as corundum, enstatite, sapphirine, and gedrite, have been also described by Simon & Chopin (2001) to fill pyrope cracks most likely developed during early decompression. Pyrope is full of negative crystal multiphase inclusions, which contain magnesite, Mg phosphate, sheet silicates, a chloride, and an opaque phase, without fluid preserved: these inclusions have been interpreted by Philippot et al. (1995) as decrepitated H2O + CO2 fluid inclusions formed at peak P conditions (Schertl et al., 1991; Philippot et al., 1995). The whiteschist retrogression is mainly controlled by the pyrope breakdown. Two main steps have been recognised by Schertl et al. (1991): a first one producing the assemblage phlogopite + kyanite + quartz, and a second one characterised by the assemblage chlorite + kyanite + muscovite + quartz (Fig. 6c). Phlogopite-kyanitescapolite-quartz pseudomorphs after pyrope have been also described by Schertl et al. (1991). Mass balance considerations suggest that the reaction rim formed through interaction with a metasomatising hydrous fluid. These peculiar rocks, formerly considered as the metamorphic product of original Mg-rich pelite formed in an evaporitic environment (Chopin, 1987; Schertl et al., 1991), are now accepted to have been derived from granitoid protoliths, which experienced a metasomatic process along ductile shear zones in the presence of a hydrous fluid phase (Gebauer et al., 1997; Compagnoni & Hirajima, 2001). The occurrence within a few whiteschist lenses of rare superzoned garnets, ranging in composition from Alm70Prp25Grs5 in the core to Alm14Prp84Grs2 in the rim (Fig. 8), has been interpreted as evidence that the metasomatic process involved both the granitoid and the included paraschist xenoliths and occurred during prograde UHPM. For a more detailed structural, mineralogic and chemical description of the whiteschist see the review papers by Schertl et al. (1991), Chopin & Sobolev (1995), Compagnoni et al. (1995), Michard et al. (1995) and Chopin & Schertl (1999). Garnet-jadeite-kyanite-quartz (/coesite) granofels. This peculiar lithotype was originally described as a “weathered bluish rock devoid of mica” by Chopin (1984), and later reported as a “jadeite-kyanite quartzite” by Chopin (1987) and as a “garnet-jadeitequartzite or garnet-jadeite-quartz band” by Schertl et al. (1991). It occurs within a few pyrope whiteschists as folded layers from 10 to 20 cm thick and several meters long (Chopin, 1984, 1987; Schertl et al., 1991). Under the microscope, the granofels shows the UHP peak metamorphic assemblage quartz/coesite, pyrope-rich almandine, jadeite, kyanite, very minor phengite, rare colourless glaucophane, and accessory apatite, rutile, and zircon (Fig. 6d). Garnet includes both prograde and peak minerals, such as paragonite, phengite, talc, jadeite, kyanite, rutile, zircon, apatite, and coesite or its retrograde polycrystalline quartz aggregates. Coesite is also included in jadeite and kyanite. In garnet, small polycrystalline inclusions with negative crystal shape, filled

30

R. Compagnoni & F. Rolfo

with prevailing white mica + chlorite intergrowth, have been interpreted as decrepitated UHP fluid inclusions. A second glaucophane generation locally occurs in the rock matrix, probably developed during an early retrogression stage through the reaction jadeite + kyanite + garnet + H2O = glaucophane + paragonite. Usually jadeite is extensively replaced by fibrous jadeite + paragonite + hematite + magnetite, and garnet is partly altered to green biotite and magnetite (Schertl et al., 1991). After complete greenschist facies retrogression, this rock is transformed into a fine-grained gneiss consisting of quartz, muscovite, and paragonite, albite/oligoclase, and minor clinozoisite. The presence of the kyanite + jadeite assemblage implies that pressures higher than the paragonite breakdown (about 2.5 GPa) have been exceeded. This unusual lithotype has been considered by Schreyer et al. (1987) and Sharp et al. (1993) as the possible product of partial melting of the host whiteschist protolith, but convincing evidence for this interpretation is still lacking. In summary, the local preservation of relict pre-Alpine structures and minerals in both the Polymetamorphic and Monometamorphic Complexes enlighten the pre-Alpine evolution and geologic setting of the BIU as summarised in Figure 7. The BIU is a fragment of the European continental crust, mainly consisting of Variscan amphibolite facies metamorphics, which was intruded about 275 Ma ago (Gebauer et al., 1997) by late to post-Variscan porphyritic granitoids. During the Alpine orogeny this portion of

Fig. 7. Inferred pre-Alpine geologic setting for the UHP Brossasco-Isasca Unit, Dora-Maira Massif.

UHPM units in the Western Alps

31

Variscan continental crust experienced an earlier Alpine coesite eclogite facies and a later Alpine greenschist facies recrystallisation, the Variscan amphibolite facies basement being converted to the present Polymetamorphic Complex, and the late Variscan granitoids to the Monometamorphic Complex (Compagnoni et al., 1995). The metamorphic evolution of the BIU and its P–T–t path The whole metamorphic evolution of the BIU is summarised in the P–T diagram of Figure 9. The oldest event, recognised in the paraschist of the Polymetamorphic Complex, is a pre-Alpine amphibolite facies metamorphic recrystallisation, most likely Variscan in age. This regional event is generally overprinted by a HT-LP contact metamorphism, connected to the emplacement of the late Variscan granitoids (Compagnoni & Hirajima, 1991; Compagnoni et al., 1994). The information gathered from the inclusions in the UHP climax minerals suggest a quite steep prograde Alpine P–T path (Schertl et al., 1991; Compagnoni et al., 1994). The early Alpine peak UHPM conditions are quite well constrained at P = 3.3 r 0.3 GPa and T = 750 r 30 °C (Chopin, 1984, 1987; Chopin et al., 1991; Kienast et al., 1991; Sharp et al., 1993; Compagnoni et al., 1994). Slightly higher pressures were inferred by Schertl et al. (1991). A new study applying a revised garnet-phengite thermobarometer (Coggon & Holland, 2002) estimated for the pyrope-bearing whiteschist with the assemblage garnet + kyanite + talc + phengite + coesite from the type locality (sample 2-39 of Chopin, 1984) P = 3.2 GPa and T = 730 °C and aH2O = 0.5, due to the presence of halite-saturated fluid. Other estimates of the water activity at peak conditions are either higher than 0.8 (Schertl et al., 1991) or lower than 0.4–0.7 (Sharp et al., 1993). In any case, the activity of CO2 must have been very low, otherwise pyrope could not have been stable (Sharp et al., 1993). All petrologic considerations (e.g. Kienast et al., 1991; Compagnoni et al., 1994; Simon et al., 1997) suggest that the peak pressure conditions never exceeded the graphite–diamond equilibrium curve, in agreement with the results of a mineral inclusion study in zircons (Sobolev et al., 1994), where only coesite was found. However, Hermann (2003) combined experimental results in the model KCMASH system together with petrologic information from whiteschist, metapelite and eclogite, and the resulting petrogenetic grid suggests that the BIU experienced peak metamorphic conditions at T ~ 730 °C and P ~ 4.3 GPa. The evidence that diamond has not been formed from graphite inversion, a common relict phase in most metapelites and in some eclogites, is explained by the author as a consequence of kinetic problems, due to the relatively low temperature of metamorphism, the absence of a free fluid phase, and a short residence time in the diamond stability field. The metamorphic evolution post-dating the UHPM climax is marked by a significant decompression coupled with a moderate and continuous cooling (Chopin et al., 1991; Schertl et al., 1991; Compagnoni et al., 1994). At relatively low pressures, this cooling path is followed by a moderate temperature increase; this second thermal peak, which also occurs under decompression conditions, corresponds to the second Alpine metamorphic event, and is approximately located at the boundary between the upper

32

R. Compagnoni & F. Rolfo

Fig. 8. (a) Sketch of superzoned garnet from the BIU whiteschists, with distribution of mineral inclusions, and trace of the microprobe profile. FeO wt% contour lines are also shown. (b) Microprobe profile across garnet. The Mg-rich cusp in the core (black arrow) corresponds to one of the two embayments evident in the Mg X-ray map of CD Image 8-1. Modified after Compagnoni & Hirajima (2001). The scale bar is 5 mm.

greenschist and low amphibolite facies (Compagnoni et al., 1994). A further significant cooling at lower decompression rates followed. The whole early Alpine evolution is, therefore, defined by a clockwise P–T path, characterised by an extremely flat loop, which is consistent with the rapid exhumation suggested by the new geochronological data discussed in the next chapter.

UHPM units in the Western Alps

33

Fig. 9. Petrogenetic grid showing the P–T–t path recorded in the UHP rocks of the Brossasco-Isasca Unit, modified from Compagnoni et al. (1995). Peak metamorphic conditions of the quartz eclogite and blueschist facies units overlying and underlying the BIU, respectively, are also shown. Prograde path mainly follows Schertl et al. (1991) with supplementary data from a newly found relict Cl-bearing hornblende included in eclogite garnet and from a pre-Alpine almandine garnet in paraschist. The climax stage is defined by the existence of the Jd-Prp-Gln assemblage. An early retrograde stage is defined by Fe–Mg partitioning between omphacite and garnet (Hirajima & Compagnoni, 1993). Gr l Dia after Kennedy & Kennedy (1976); Qtz l Coe after Mirwald & Massonne (1980); Gln l Jd + Tlc and the stability field of nyböite (Ny) after Carman & Gilbert (1983). Pg l Jd + Ky, Pg + Qtz l Ky + Ab, Ab l Jd + Qtz and Lws l Zo + Ky after Holland (1979). Margarite stability field after Perkins et al. (1980). Other curves are obtained with the Ge0Calc program (Brown et al., 1988) and from the database of Berman (1988). Numbers in circles are ages in Ma of the metamorphic events recorded in the BIU according to SHRIMP data (Gebauer et al., 1997; Rubatto & Hermann, 2001) and traditional radiometric data (see Hunziker et al., 1992).

34

R. Compagnoni & F. Rolfo

The age of the UHP metamorphism From the geologic lines of evidence mentioned above, it is clear that the UHP recrystallisation of the BIU is Alpine in age. However, until recently the true age of the UHP metamorphism was still unknown, as well as the duration of the related exhumation. On the one hand, Cretaceous (“early Alpine”) U–Pb zircon ages of 121 +12/–29 Ma and Rb–Sr phengite ages of 96 r 4 Ma (Paquette et al., 1989) and 40Ar–39Ar step-heating plateau ages between 110 and 100 Ma on high-celadonite phengites (Monié & Chopin, 1991) were obtained from samples best preserving the UHP mineral assemblages. 40 Ar–39Ar plateau ages of 35–41 Ma were obtained from low-celadonite phengites, separated from gneisses pervasively retrogressed to greenschist facies mineral assemblages (Monié & Chopin, 1991). Younger 40Ar–39Ar biotite ages between 28 and 35 Ma (Monié & Chopin, 1991) and a 40Ar–39Ar plateau age of 31 Ma obtained from texturally late fine-grained muscovites (Hammerschmidt et al., 1992) were interpreted as dating the cooling of the regional Alpine greenschist facies event. On the other hand, Tertiary ages were determined for the coesite eclogite facies metamorphism by Tilton et al. (1991, and references therein); U–Pb zircon ages of 38 r 1.4 Ma, U–Pb ellenbergerite and monazite ages of 30 to 34 Ma, and Sm–Nd pyrope ages of about 38 Ma were obtained from equivalent samples, best preserving the UHP mineral assemblages. These age discrepancies were in part explained by the study of Arnaud & Kelley (1995), concluding that the old 40Ar–39Ar ages in the range 60–110 Ma were the result of artefacts caused by incorporation around 40–50 Ma ago of variable excess argon. This age controversy was settled by Gebauer et al. (1997, and references therein), who dated the peak conditions of the UHPM at about 35 Ma, by combining cathodoluminescence imaging and in situ U–Pb dating of zircon (a high closure temperature mineral) single crystals with the Sensitive High Resolution Ion Microprobe (SHRIMP). These U–Pb ages, which consistently fall at the Eocene–Oligocene boundary (~ 35 Ma), were later confirmed by the results from Lu–Hf dating of garnet (Duchêne et al., 1997). The second, low pressure, Alpine metamorphic event has been recently determined by Rubatto & Hermann (2001) by in situ SHRIMP dating of a zoned titanite from a calcsilicate granofels, preserving the peak metamorphic stage and two decompressional stages well constrained by the mineral inclusions: the UHPM stage was dated at 35 Ma, and the two decompressional stages at about 33 and 32 Ma. The last age of 32 Ma is inferred to date the thermal peak of the Alpine greenschist facies metamorphism. Combining these radiometric data with a zircon fission track age of about 30 Ma, and making reasonable assumptions about the conversion of pressure to depth, maximum exhumation rates of 3.4 cm/year have been estimated (Fig. 10). Conclusions As in other UHPM belts, the tectonometamorphic evolution of the BIU shows that during exhumation the peak P mineral assemblages are extensively retrogressed to low

UHPM units in the Western Alps

35

Fig. 10. P–T–t path of the BIU (a), and depth vs. time path (b) from which mean exhumation rates have been estimated (from Rubatto & Hermann, 2001).

P upper greenschist facies parageneses. This retrogression is especially well developed in felsic rocks (Compagnoni & Rolfo, 1999), where the significant volume increase connected to the coesite to quartz inversion promotes a pervasive tectonic reworking and consequently activates metamorphic reactions. However, locally low-strain domains are found, even in marble, where relics of the pre-Alpine protoliths preserve the original microstructure and part of the primary mineralogy, and also the UHPM minerals. The best examples are the metagranitoids and the pre-Alpine amphibolite facies metamorphics, especially where overprinted by contact metamorphism. The reason for such unusual mechanical behaviour is most likely due to the absence of a free fluid phase, which lowers the rock ductility and slackens the reaction kinetics. The lack of a free fluid phase during prograde UHPM, already evident from microscopic observations for the occurrence of pseudomorphous and coronitic reactions, is supported by a thermodynamic modelling study on metagranitoids, suggesting that the UHPM recrystallisation occurred in a closed system (Bruno et al., 2001). These results are also supported by the geochronologic studies of Tilton et al. (1997) and Paquette et al. (1999), concluding that the minerals of the undeformed metagranites failed to reach isotopic equilibration during Alpine subduction in spite of the high pressures (> 3.5 GPa) and temperatures (> 700 °C) suffered, in apparent disagreement with the “closing temperature” concept.

36

R. Compagnoni & F. Rolfo

The Lago di Cignana Unit of the Piemonte zone Geological and structural setting The Lago di Cignana unit (LCU) is exposed on the southern side of the lower Conca di Cignana, upper Valtournenche, Val d’Aosta (Reinecke, 1991, 1998). The unit is part of a stack of thin nappes of meta-ophiolites and metasediments of the Piemonte zone (Pennine Domain), which are overlain by the thick continental Dent Blanche nappe, belonging to the Austroalpine Domain of the western Alps (Fig. 1). The LCU consists of eclogite facies metabasics and metasediments, which are thought to represent a coherent section through a segment of former oceanic crust because: 1) both the metasedimentary and metabasic rocks have undergone the UHP metamorphism; 2) the characteristics of non-clastic, siliceous sediments, and especially the presence of ferromanganese nodules match the features of recent oceanic crust; 3) major and trace element compositions of the eclogites have the characteristics of N-MORB (Beccaluva et al., 1984; Pfeifer et al., 1989; van der Klauw et al., 1997). The LCU is a narrow lens (Fig. 11), possibly less than 150 metres in thickness, characterised by a sequence of boudins from hundreds to tens of metres in size. Boudins may not always be apparent on a large scale as they have been truncated by late-stage shearing, faulting or scraped by glaciers. Interlayers of metasediments and metabasites preserve structures and mineral assemblages that suggest a multistage and complex exhumation history (van der Klauw et al., 1997; Reddy et al., 1999) beneath a km-scale extensional ductile shear zone in which intense fabrics are developed. Low-angle faults formed in the late stages of movement. The extensional structures have been shortened in their later history, forming regionally developed upright folds, and dome and basin structures that decorate and emphasise the UHP boudins (Lister et al., in prep.). The LCU crops out immediately south and west of Lago di Cignana in the area extending about 1.2 km in N–S and E–W directions from the dam (1950–2250 m a.s.l.). To the east and northeast, the UHP rocks disappear under scree covering the steep walls of the Valtournenche valley. In the north, west and southwest, the UHP rocks are tectonically overlain (“Combin-thrust” of Ballèvre & Merle, 1993) by the greenschist facies prasinites and calcschists of the Tsaté nappe (“Combin zone” of Bearth, 1967) and by a huge serpentinite slice (southern slope of Mt. Pancherot). Relic Mg-poor almandine–grossular garnets and poikiloblastic epidote preserve inclusions of glaucophane (Dal Piaz, 1976; Le Goff, 1986; Wagner-Zweigel, 1993) which suggest derivation from garnet-epidote-glaucophanites indicative of a former epidote-blueschist facies metamorphism. Therefore, the LCU may be considered as a UHP lens which is tectonically emplaced and located at the upper boundary of the quartz eclogite facies Zermatt–Saas Zone, directly below the epidote-blueschist facies Combin Zone (Bearth, 1967). The characters of the upper and lower contacts are different and complicated by overprinting deformational events. The Zermatt–Saas Zone is characterised by a planar layering while the UHP eclogitic metabasites are deformed boudins. The upper contact of the LCU is in contact with the Fe-Ti metagabbro of the Zermatt–Saas Zone, and can be observed on the northern boundary of the UHP Unit (below high water level) and in several locations on the upper surface of the LCU. The contacts between the UHP lens

UHPM units in the Western Alps

37

Fig. 11. Regional geological map of the Lago di Cignana region. Interpretation (a) and real outcrops are clearly distinguished. The cross-section is parallel to the dominant NW–SE trending stretching lineations. Modified after Tamagno (1999) and Lister et al. (in prep.).

38

R. Compagnoni & F. Rolfo

and Zermatt–Saas Zone at the eastern and western side have been preserved and show the Zermatt–Saas Zone wrapping the UHP boudins (Fig. 11). The upper tectonic contact has been later folded and cut by late-stage ductile and brittle deformation (Lister et al., in prep.). It is questionable whether the calcschist is part of the Combin Zone caught up in the structurally higher regions of the LCU or it has undergone the same deformation and metamorphic history as the quartz-rich metasediment layers. The LCU preserves an older deformational history and movement direction compared to the Combin Unit. The Combin Unit is the main shear zone in this region and the LCU has been exhumed beneath this large scale extensional shear zone. This shear zone is suggested to be of regional extent. It can be extrapolated northward from Lago di Cignana and beneath the Matterhorn. A comparison of the sense of shear on shear zones between the Combin and UHP Units suggests a multistage episodic exhumation of the LCU. Low angle faults formed in the late stages of movement. The extensional structures have been shortened in their later history, forming regionally developed upright folds, and dome and basin features. Lithologies The LCU is made of metabasites and metasediments ranging in composition from quartzite (locally Mn-rich) to calcschist, derived from the Mesozoic Tethyan oceanic crust. Metabasites The metabasites are medium- to fine-grained banded coesite eclogites, derived from original MORB-type basalts (Beccaluva et al., 1984; Pfeifer et al., 1989; van der Klauw et al., 1997). In the field, eclogites locally contain thin glaucophane-garnet-rich layers and streaks wrapping around metre-sized eclogite lenses and pods, interpreted by Reinecke et al. (1994) as relics of former pillow structures. Frequently, an omphacite lineation is evident, which is cut by irregular omphacite and glaucophane-rich veins and pockets. Under the microscope, most eclogites consist of the following peak assemblage: omphacite + garnet (rim), minor zoisite, phengite, quartz/coesite, local carbonate, and accessory rutile, fluorapatite, pyrite r dolomite. Clinozoisite + white mica pseudomorphs after former prograde lawsonite are ubiquitous, even as inclusions in garnet. Glaucophane, usually porphyroblastic, mainly developed after the main eclogite foliation. The presence of inclusions trails in garnet and the garnet-matrix omphacite relationships suggest a polyphase HP–UHP metamorphic evolution. Coesite occurs rarely as tiny inclusions in matrix omphacite and in narrow, inclusion-poor garnet rims. It is commonly inverted to quartz. Poikiloblastic garnet cores have inclusions of omphacite, rutile, paragonite, clinozoisite and monocrystalline, almost unstrained quartz. Omphacite inclusions in garnet tend to be less jadeitic than matrix omphacite (Fig. 12). Garnet is characterised by a continuous growth zoning with pyrope component increasing and spessartine, almandine and grossular components decreasing towards the rim (Fig. 12).

UHPM units in the Western Alps

39

Fig. 12. Compositions of UHPM garnets and clinopyroxenes from the garnet-phengite-epidote-Na-pyroxene quartzite and eclogite, Lago di Cignana. Mineral symbols according to Kretz (1983); Cald: calderite; Bly: blythite.

The eclogites display a variable degree of greenschist facies re-equilibration, which appears to be mainly related to fluid infiltration along fractures now filled by albiteactinolite-chlorite-epidote veins and quartz veins. Metasediments The metasediments are mainly composed of calcschist grading to Mn-bearing micaschist with minor impure marble and locally quartzite interlayers with cm- to dm-thick nodules or boudins of quartz-bearing garnetite and eclogite, and dm- to m-thick lenses of sheared serpentinite, “metagabbro” and “flasergabbro”. Two types of quartzite occur: garnetomphacite quartzite with dm-thick omphacitite nodules and Mn-bearing quartzite with

40

R. Compagnoni & F. Rolfo

mm- to dm-thick hematite-braunite-rich aggregates, most likely derived from original Fe-Mn-rich nodules (Reinecke, 1991, 1998). The metasediments mainly consist of two lithotypes: 1) carbonate-bearing garnetphengite-quartz schist, and 2) oxidised manganiferous quartz schists and quartzite. The oxidised manganiferous quartz schists, in turn, comprise: 2a) piemontite-garnetphengite-talc-quartz schist, and 2b) aegirine-augite - garnet - phengite quartzite. Like the calcschists, the non-felsic lithologies intercalated within the sedimentary cover, i.e. serpentinite, “metagabbro”, “flaser-gabbro”, eclogite and garnetite, may be interpreted either as ophiolitic detritus derived from erosion of the oceanic lithosphere tectonically exposed on the floor of the Mesozoic Tethys, or as portions of the adjacent Combin Zone tectonically caught up. The metasediments of this “ocean-type” cover also contain a detrital fraction derived from a continental source: this is suggested by the discovery in quartzite of zircons with Precambrian to Jurassic radiometric ages in the inherited cores (Rubatto et al., 1998). Rare relict coesite (or retrograde polycrystalline quartz aggregates), originally discovered in a tourmaline from a quartzite (Reinecke, 1991), was later found as tiny inclusions in garnet, omphacite, piemontite, and epidotes of most lithologies. Preserved coesite has also been found as inclusions in matrix apatite in a garnet-phengite-quartz schist (Reinecke, 1998). Pale green to silvery, medium-grained garnet-phengite-quartz schists and quartzites are most common among the metasediments that overlie the eclogites with an apparent thickness of a few tens of metres in the area S and SW of the lake. The dominant foliation dips to the NW–W. In structurally lower parts of the metasediments, close to the contact with the eclogites, one observes dm- to m-thick layers and boudins of garnet-rich quartzites, garnetclinopyroxene quartzites, as well as highly oxidised manganiferous rocks (piemontitephengite-quartz schists, piemontite-phengite quartzites and aegirine-augite - epidote - phengite quartzites) which may reflect hydrothermal activity in the pelagic sediments near the former basalt–sediment interface. Due to major viscosity contrasts between the different rock types, the contacts between the eclogites and the metasediments are always disturbed. Garnet-phengite-quartz calcschist This lithology derives from a carbonate-rich metapelite protolith, and consists of carbonate-rich layers alternating with phengite-quartz-rich layers. In the calcschist, the inferred UHP peak assemblage is garnet + dolomite + coesite + lawsonite (now epidote + phengite r paragonite) + phengite (Si3.35–3.34) + aragonite (now calcite) + rutile (Reinecke, 1998). The former presence of most likely prograde lawsonite is inferred from the widespread occurrence of mm-sized, lozenge-shaped pseudomorphs mainly consisting of zoisite/clinozoisite + paragonite aggregates. Other samples of garnetphengite-quartz schists at the UHP stage may have had clinozoisite/epidote (but no lawsonite), r sodic amphibole (pseudomorphed by Ca-Na-amphibole, biotite and albite), apatite and tourmaline. In the Mn-poor silicate domain, the UHP garnet composition is about Prp7Alm71Sps3Grs19 (Reinecke, 1998). Relics of the UHP stage are sparse and difficult to recognise in thin section, because (in contrast to the eclogites) the UHP matrix assemblage of the garnet-phengite-quartz

UHPM units in the Western Alps

41

calcschists has been largely erased, i.e. the matrix minerals now mainly reflect the greenschist facies overprint. From compositional zoning of garnet and phengite, inclusion relationships and pseudomorphous replacement textures, the following UHP assemblage is inferred: garnet I + phengite I (Si3.4) + coesite + epidote/clinozoisite or zoisite r dolomite r glaucophane/crossite + apatite + rutile r tourmaline. Garnet I in high-Fe2+ bulk compositions forms continuously zoned grains (up to 4 mm) with Mg and Fe increasing and Ca and Mn decreasing from core to rim [Alm63–77Prp5–8Sps1–9(Grs + Adr)14–25]. Inclusions of Fe-rich chloritoid (XMg = 0.15), paragonite and of monocrystalline, unstrained quartz in garnet I cores are considered as relics of the prograde high-pressure path. In more Mg-Mn-rich bulk compositions, garnet I cores are in the range Alm5–39Prp23–31Sps28–55(Grs + Adr)9–10. Small inclusions of partially transformed coesite or of polycrystalline, highly strained quartz, pseudomorphous after coesite, occur rarely in Mg-rich garnet cores. Piemontite-garnet-phengite-talc-quartz schist This Mn-bearing micaschist, first described by Bearth (1967) from one occurrence about 100 m SE of the southern wing of the dam, is probably derived from a manganiferous (MnO typically in the range 1 to 4 wt%), SiO2-rich metapelite, and includes scattered mm- to cm-sized hematite-braunite-rich aggregates, most likely derived from original Fe-Mn-rich nodules. Two domains have been distinguished by Reinecke (1991): “Al-rich” domains and “Al-poor” domains. In the “Al-rich” domains, the inferred UHPM peak assemblage was coesite + kyanite + talc + phengite + garnet + hematite + rutile + piemontite + zircon, with the garnet zoned from Prp19Sps67Grs14 in the core through Prp36Sps55Grs9 in the intermediate to Prp21Sps70Grs9 in the rim. In the “Al-poor” domains (the former Fe-Mn nodules), the inferred UHPM peak assemblage is coesite + phlogopite + talc + phengite + garnet + piemontite + hematite + rutile + braunite + ardennite + dravite + apatite + zircon, with the garnet zoned from Prp21Sps67Grs12 in the core through Prp35Sps55Grs10 in the intermediate to Prp20Sps71Grs9 in the rim. Two main UHP assemblages have been recognised (Reinecke et al., 1994): garnet I [rim: Prp16–36Sps71–55(Grs + Adr)9–14]+ coesite + talc + phengite-3T (Phe I) (Si3.42-3.46) + braunite I r manganoan phlogopite I, and rarely: garnet I [Prp28–41Sps62–54(Grs + Adr)4–10] + coesite + talc + phengite-3T + kyanite + piemontite I + hematite + rutile + dravite + ardennite + apatite and zircon. Higher-variance assemblages lacking talc, kyanite, phlogopite I or even garnet I are even more common. Relic coesite, or polycrystalline quartz pseudomorphous after coesite, are present as inclusions in dravite and garnet I2 (mantle of high pressure garnet; Fig. 12). Garnet I2 has also rare inclusions of all other minerals of the UHP stage. Aegirine-augite - garnet - phengite quartzite In the quartzite interlayers, chemically similar to the previous schists but higher in Na and lower in Al contents, the inferred UHPM peak assemblage is garnet I + phengite (Si3.38) + coesite + Mn-epidote + aegirine-augite (Jd26–36Aeg42–51Aug13–24) + hematite +

42

R. Compagnoni & F. Rolfo

rutile r apatite r dravite r manganoan dolomite and calcite after former aragonite. It is interesting to note that in the Al-poor and haematite-rich domains (the former Fe-Mn nodules), a calderitic (Mn3Fe2Si3O12) garnet with composition Cald44–83Adr5–14Sps47–7Prp1 (Fig. 12) was found, which according to Lattard & Schreyer (1983) is peculiar to high pressure conditions (pure calderite requires about 1.8 GPa minimum pressure at 600 °C). High pressure epidote cores locally include Si-rich phengite, manganoan dolomite, calcite (after former aragonite?) and polycrystalline quartz. In places, UHPM assemblages may lack garnet I and/or Na-pyroxene. Metamorphic evolution Peak metamorphic conditions have been estimated to be 615 r 15 °C and 2.8 r 1.0 GPa (Reinecke, 1991, 1998; Reinecke et al., 1994). A detailed petrologic study performed by Reinecke (1991, 1998) on a calcschist, a banded Mn-bearing oxidised quartz schist and a quartzite, allows reconstruction of the prograde and retrograde P–T path (Fig. 13). In eclogites, the Powell (1985) and Krogh (1988) calibrations for the Fe-Mg garnet–clinopyroxene thermometer applied to garnet (rim)–omphacite (matrix) pairs (Reinecke et al., 1994) gave 613 r 34 °C and 578 r 36 °C, respectively (at P = 2.7 GPa). Exchange temperatures calculated from the compositions of omphacites included in garnet (cores) were on average 30 °C lower than peak temperatures. In contrast to garnets from metasedimentary rocks, the eclogitic garnets do not show compositional reversals within the outermost growth zones, suggesting that garnet growth terminated near the thermal peak. In quartzites, the cores of garnet I (stage I1 of Reinecke et al., 1994) which often include monocrystalline quartz (plus hematite, rutile and rare paragonite and braunite) record some part of the prograde high P–T path in the stability field of quartz. Garnet I1 + I2 shows steep concentration profiles with continuous changes in XMg and XMn, but an almost flat XCa profile with a discontinuity at the I1–I2 interface, which may indicate a change of the limiting phase assemblage. Calculation of the dP–dT increments performed by Reinecke et al. (1994) from three garnet half-profiles in one sample (containing the phase assemblage garnet I + phengite I + talc + phlogopite + piemontite + quartz/coesite + fluid) with the GIBBS method (e.g. Spear, 1988; Spear & Menard, 1989) indicates that: I) consistent results are obtained regarding the general shape of the P–T trajectories; II) maximum pressures were attained some 30–45 °C below the thermal peak; III) decompression during early uplift was linked with significant cooling. In eclogites, metamorphic retrogression started with the growth of titanite rims around matrix rutile, growth of topotactic Ca-Na-amphibole on glaucophane and around garnet, and growth of more Fe3+-rich rims on pre-existing clinozoisite/epidote. More advanced stages of overprinting show the replacement of omphacite and glaucophane by Ca-Na-amphibole-albite symplectites and then by greenschist facies minerals with progressive alteration of eclogitic garnet to chlorite r epidote r biotite, replacement of rutile by ilmenite r titanite, and of Ca-Na-amphibole and paragonite by actinolite + albite and albite, respectively.

UHPM units in the Western Alps

43

Fig. 13. P–T–t evolution of the UHP metamorphic oceanic lithosphere at Lago di Cignana (from Reinecke, 1991; Lister et al., in prep.). Mineral abbreviations after Kretz (1983), with implements by Bucher & Frey (2002). The age of peak metamorphism is from Rubatto et al. (1998).

In metapelitic schists, the lower pressure decomposition products in the matrix comprise two generations of less magnesian, Ca- or Ca-Mn-richer garnet II and III, titanite rims on rutile, and growth of less siliceous phengite II, paragonite, chlorite, biotite, albite, epidote and calcite. Na-amphibole is pseudomorphed by intergrowths

44

R. Compagnoni & F. Rolfo

of biotite + Ca-Na-amphibole/actinolite + albite and phengite II is replaced by epidote or albite. In garnet - phengite - epidote - Na-rich pyroxene quartzites, the lower-pressure decomposition products are poikiloblastic garnet II [Sps60–84Cald0–15Prp0.5–2(Grs + Adr)12–30] which overgrows garnet I and mangano-cummingtonite, which replaces piemontite, as well as albite, calcite and phlogopite. Aegirine-jadeite/chloromelanite is partially replaced by very fine-grained symplectites consisting either of microcline (Or95Ab4Cs1) + low albite (Ab99Or1) + manganoan phlogopite + hematite + sodian tremolite or of albite + microcline + hematite + sodian tremolite and aegirine-augite (Jd4–17Aeg29–49Di42–67). In manganiferous quartz schists, during early decompression only a partial resorption of garnet I2 rims is recorded (Reinecke et al., 1994). As the stability field of clinochlore + quartz was entered, renewed growth of less magnesian garnet II [Prp12–6Sps77–83(Grs + Adr)9–13], phengite-3T II (Si3.31–3.36), ardennite II, braunite II and piemontite II started in the presence of matrix clinochlore I (AlIV = 1.2 per 10 cations), talc and quartz at about 500–550 °C, 1.0–1.6 GPa. Compositionally zoned, discontinuous garnet II rims around garnet I show an outward decrease in XMg, whereas XCa behaves variably. The second generations of phengite-3T, ardennite and braunite differ significantly in composition from those of the UHP stage. During the late greenschist facies overprint discrete grains or rims of birefringent spessartine-rich garnet III [Prp1Sps85–90(Grs + Adr)9–14] formed around garnet II, phengite III became less celadonitic (Si3.20–3.26), and clinochlore II less aluminous (AlIV = 1.1 p.f.u.). Intergrowths of clinochlore II + phlogopite II locally replace the high pressure paragenesis phengite II + talc. Replacement of phengite II by albite is common and indicates the significance of fluid in material transport on a scale of at least several 100 micrometres. A summary of the P–T evolution of the LCU is presented in Figure 13. The UHP trajectory is calculated from garnet I composition profiles (Reinecke et al., 1994), whereas for part of the uplift path a record is lacking in the mineral assemblages (piemontitephengite-quartz schists) or not yet readable (eclogites and garnet-phengite-quartz schists). The trajectory is similar to those inferred from other tectonic units of the internal western Alps, and is characterised by a post-eclogitic decompressional path coupled with a moderate cooling and a second thermal peak at low pressure (Lister et al., in prep.). Geochronology The age of the UHPM was determined at about 44 Ma by combining cathodoluminescence and the SHRIMP technique on zircon single crystals from metasediments, with sporadic rutile inclusions (Rubatto et al., 1998). A multimineral Sm–Nd isochron from an eclogite gave an age of about 40 Ma (Amato et al., 1999) and a Lu–Hf isochron on inclusion-rich garnet–whole rock gave an age of about 49 Ma (Lapen et al., 2002). In spite of the seemingly different radiometric ages, the other two age determinations are in line with the metamorphic peak determined at about 44 Ma by Rubatto et al. (1998). A more exhaustive discussion is reported in the review by Rubatto et al. (2003) on radiometric dating of UHP metamorphic rocks.

UHPM units in the Western Alps

45

The greenschist facies retrogression should have occurred between about 38 Ma (Rb–Sr whole rock–phengite isochron: Amato et al., 1999) and about 35 Ma (U–Pb age of a retrograde titanite: Barnicoat et al., 1995). Because of the uncertainty of both the UHP and retrogression metamorphic ages (Rubatto et al., 2003), it is difficult to obtain a reliable exhumation rate of the LCU, that nevertheless must be in the range 0.5–1.0 cm/year. Conclusions The ocean-derived LCU is significantly different from the BIU since it is smaller and dismembered, exhibiting a more severe and pervasive deformation that obliterated most primary features. These features prevented the study of the UHPM transformations of the primary mineral assemblages. However, in spite of the different origin and tectonic locations, both UHPM units show some characteristics in common. Both units are relatively small and thin, are juxtaposed to units with significantly different HP peak recrystallisations, and show similar P–T paths, characterised by a post-eclogitic decompressional path coupled with a moderate cooling and a second thermal peak at low pressure. Since these P–T paths are also peculiar to the other eclogite facies units of the internal western Alps, but are different from those described from other UHPM belts, it is suggested that different tectonic mechanisms were responsible for the exhumation of subducted continental crust in different collisional orogenic belts.

Acknowledgements We would like to thank Gordon Lister for sharing unpublished field and laboratory data on the Lago di Cignana Unit, and Tony Carswell for critical reading and polishing the English. Christian Chopin’s careful review and suggestions significantly improved the manuscript. Financial support was provided by the National Research Council of Italy (C.N.R.) – Institute of Geosciences and Earth Resources.

References Amato, J.M., Johnson, C.M., Baumgartner, L.P. & Beard, B.L. (1999): Rapid exhumation of the Zermatt-Saas ophiolite deduced from high-precision Sm-Nd and Rb-Sr geochronology. Earth Planet. Sci. Lett., 171:425–438. Arnaud, N. & Kelley, S.P. (1995): Evidence for excess argon during high pressure metamorphism in the DoraMaira Massif (western Alps, Italy), using an ultra-violet laser ablation microprobe 40Ar/39Ar technique. Contrib. Mineral. Petrol., 121:1–11. Avigad, D., Chopin, C. & le Bayon, R. (2003): Thrusting and extension in the southern Dora-Maira ultra-highpressure massif (Western Alps): view from below the coesite-bearing unit. J. Geol., 111:57–70. Ballèvre, M. & Merle, O. (1993): The Combin Fault; compressional reactivation of a Late Cretaceous-early Tertiary detachment, Schweiz. Mineral. Petrogr. Mitt., 73:205–227. Barnicoat, A.C., Rex, D.C., Guise, P.G. & Cliff, R.A. (1995): The timing of and nature of greenschist facies deformation and metamorphism in the upper Pennine Alps. Tectonics, 14:279–293. Bearth, P. (1967): Die Ophiolithe der Zone von Zermatt-Saas Fee. Beitr. Geol. Karte Schweiz, 130:1–132. Beccaluva, L., Dal Piaz, G.V. & Macciotta, G. (1984): Transitional to normal MORB affinities in ophiolitic metabasites from the Zermatt-Saas, Combin and Antrona units, Western Alps; implications for the paleogeographic evolution of the western Tethyan Basin. Geol. Mijn., 63(2):165–177.

46

R. Compagnoni & F. Rolfo

Berman, R.G. (1988): Internally-consistent thermodynamic data for minerals in the system Na2O-K2O-CaOMgO-FeO-Fe2O3-Al2O3-SiO2-TiO2-H2O-CO2. J. Petrol., 29:445–522. Biino, G. & Compagnoni, R. (1992): Very high-pressure metamorphism of the Brossasco coronite metagranite, southern Dora Maira Massif, western Alps. Schweiz. Mineral. Petrogr. Mitt., 72:347–363. Brown, T.H., Berman, R.G. & Perkins, E.H. (1988): Ge0-calc: Software for calculation and display of pressure-temperature-composition phase diagrams using an IBM or compatible personal computer. Comp. Geosci., 14:279–289. Bruno, M., Compagnoni, R. & Rubbo, M. (2001): The ultra-high pressure coronitic and pseudomorphous reactions in a metagranodiorite from the Brossasco-Isasca Unit, Dora-Maira Massif, western Italian Alps: a petrographic study and equilibrium thermodynamic modelling. J. Metamorph. Geol., 19:33–43. Bruno, M., Compagnoni, R., Hirajima, T. & Rubbo, M. (2002): Jadeite with Ca-Eskola molecule from an ultrahigh pressure metagranodiorite, Dora-Maira Massif, Western Alps. Contrib. Mineral. Petrol., 142:515–519. Bucher, K. & Frey, M. (2002): Petrogenesis of metamorphic rocks. 7th ed. Springer-Verlag, Berlin, 341 p. Carman, J.H. & Gilbert, M.C. (1983): Experimental studies on glaucophane stability. Am. J. Sci., 283A:414–437. Castelli, D., Compagnoni, R. & Rolfo, F. (1999): Metamorphic evolution of UHP marbles from the BrossascoIsasca Unit (Dora-Maira Massif, Western Alps). In 4th Int. Eclogite Field Symp, August 14–18, 1999, Russia-Kazakhstan, Abstr., 10–11. Chopin, C. (1984): Coesite and pure pyrope in high-grade blueschists of the westem Alps: a first record and some consequences. Contrib. Mineral. Petrol., 86:107–118. Chopin, C. (1987): Very-high-pressure metamorphism in the western Alps: implications for subduction of continental crust. Phil. Trans. R. Soc. London, 321A:183–197. Chopin, C. & Ferraris, G. (2003): Mineral chemistry and mineral reactions in UHP metamorphic rocks. In Carswell, D.A. & Compagnoni, R. (ed.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 191–227. Chopin, C. & Schertl, H.-P. (1999): The UHP unit in the Dora Maira massif, western Alps. Int. Geol. Rev., 41:765–780. Chopin, C. & Sobolev, N. (1995): Principal mineralogic indicators of UHP in crustal rocks. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 96–132. Chopin, C., Henry, C. & Michard, A. (1991): Geology and petrology of the coesite-bearing terrane, Dora Maira massif, Western Alps. Eur. J. Mineral., 3:263–291. Chopin, C., Brunet, F., Gebert, W., Medenbach, O. & Tillmanns, E. (1993): Bearthite, Ca2Al[PO4]2(OH), a new mineral from high-pressure terranes of the Western Alps. Schweiz. Mineral. Petrogr. Mitt., 73:1–9. Coggon, R. & Holland, T.J.B. (2002): Mixing properties of phengitic micas and revised garnet-phengite thermobarometers J. Metamorph. Geol., 20:683–696. Compagnoni, R. & Hirajima, T. (1991): Geology and petrology of the Brossasco-Isasca Complex, southern Dora-Maira massif, Western Alps. Terra Abstr., 3:84. Compagnoni, R. & Hirajima, T. (2001): Superzoned garnets in the coesite-bearing Brossasco-Isasca Unit, Dora-Maira Massif, Western Alps, and the origin of the whiteschists. Lithos, 57:219–236. Compagnoni, R. & Rolfo, F. (1999): Characteristics of UHP pelites, gneisses and other unusual rocks. Int. Geol. Rev., 41:552–570. Compagnoni, R., Hirajima, T., Turello, R. & Castelli, D. (1994): Ultra high pressure metamorphism in the continental crust: the Brossasco-Isasca Unit of Dora-Maira Massif. In Compagnoni, R., Messiga, B. & Castelli, D. (eds.): High pressure metamorphism in the Western Alps. Guide-book to the field excursion B1, 16th Gen. Meeting of the IMA, Pisa, 10–15 September 1994, 87–105. Compagnoni, R., Hirajima, T. & Chopin, C. (1995): Ultra-high-pressure metamorphic rocks in the Western Alps. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 206–243. Compagnoni, R., Rolfo, F., Groppo, C., Hirajima, T. & Turello, R. (in press): Mapping alpine areas characterized by “HP” to “UHP” metamorphic overprint: the Brossasco-Isasca Unit (Dora-Maira Massif, Western Alps). In Special Geologic Atlas for the 32nd International Geological Congress.

UHPM units in the Western Alps

47

Dal Piaz, G.V. (1976): Il lembo di ricoprimento del Pillonet, falda della Dent Blanche nelle Alpi occidentali. Mem. Ist. Geol. Mineral. Univ. Padova, 31:1–60. Duchêne, S., Blichert-Toft, J., Luais, B., Télouk, P., Lardeaux, J.M. & Albarède, F. (1997): The Lu-Hf dating of garnets and the ages of the Alpine high-pressure metamorphism. Nature, 387:586–589. Ferraris, C., Chopin, Ch. & Wessicken, R. (2000): Nano- to micro-scale decompression products in ultrahighpressure phengite: HRTEM and AEM study, and some petrological implications, Am. Mineral., 85:1195–1201. Gebauer, D., Schertl, H.-P., Brix, M. & Schreyer, W. (1997): 35 Ma old ultrahigh-pressure metamorphism and evidence for very rapid exhumation in the Dora Maira Massif, Western Alps. Lithos, 41:5–24. Groppo, C. (2002): Studio geologico-petrografico della terminazione nord-occidentale dell’Unità BrossascoIsasca in facies eclogitica a coesite, Massiccio Dora-Maira (Alpi Occidentali). Tesi di laurea, Univ. Torino, Italy, 175 p. Hammerschmidt, K., Schertl, H.-P., Friedrichsen, H. & Schreyer, W. (1992): Überschuss-Argon während der retrograden Metamorphose der Ultrahochdruck-Gesteine des Dora-Maira Massivs, italienische Westalpen. Ber. Dtsch. Mineral. Ges., Beih. Eur. J. Mineral., 4:112 (abstr.). Henry, C. (1990): L’unité à coesite du massif Dora-Maira dans son cadre métamorphique et structural (Alpes occidentales, Italie). Thèse de 3ème cycle, Univ. P. et M. Curie (Paris VI), France, 453 p. Henry, C., Michard, A. & Chopin, C. (1993): Geometry and structural evolution of ultra-high-pressure and high-pressure rocks from the Dora Maira massif, Western Alps, Italy. J. Struct. Geol., 15:965–982. Hermann, J. (2003): Experimental evidence for diamond-facies metamorphism in the Dora-Maira massif. Lithos, in press. Hirajima, T. & Compagnoni, R. (1993): Petrology of a jadeite-quartz/coesite-almandine-phengite fels with retrograde ferro-nyböite from the Dora-Maira Massif, Western Alps. Eur. J. Mineral., 5:943–955. Holland, T.J.B. (1979): Experimental determination of the reaction paragonite = jadeite + kyanite + water, and internally consistent thermodynamic data for part of the system Na2O-Al2O3-SiO2-H2O, with application to eclogites and blueschists. Contrib. Mineral. Petrol., 68:293–301. Holland, T.J.B. (1980): The reaction albite = jadeite + quartz determinated experimentally in the range 6001200°C. Am. Mineral., 65:129–134. Hunziker, J.C., Desmons, J. & Hurford, A.J. (1992): Thirty-two years of geochronological work in the Central and Western Alps: a review on seven maps /Mém. Géol., Lausanne, 13/. 59 p. Kennedy, C.S. & Kennedy, G.C. (1976): The equilibrium boundary between graphite and diamond. J. Geophys. Res., 81:2467–2470. Kienast, J.R., Lombardo, B., Biino, G. & Pinardon, J.L. (1991): Petrology of very high-pressure eclogitic rocks from the Brossasco-Isasca Complex, Dora-Maira Massif, Italian western Alps. J. Metamorph. Geol., 9:19–34. Kretz, R. (1983): Symbols for rock-forming minerals. Am. Mineral., 68:277–279. Krogh, E.J. (1988): The garnet-clinopyroxene Fe-Mg geothermometer: a reinterpretation of existing experimental data. Contrib. Mineral. Petrol., 99:44–48. Lattard, D. & Schreyer, W. (1983): Synthesis and stability of the garnet calderite in the system Fe-Mn-Si-O. Contrib. Mineral. Petrol., 84:199–214. Lapen, T.J., Mahlen, N.J., Johnson, C.M., Beard, B.L. & Baumgartner, L.P. (2002): Lu-Hf geochronology of UHP metamorphism in the Zermatt-Saas ophiolite, Lago di Cignana, Italy. Geochim. Cosmochim. Acta, 66:A431. Le Goff, E. (1986): Histoire structurale et métamorphique comparée de deux unitées de Schistes lustrés (Valtournanche, Alpes occidentales internes). In Mém. D.E.A., Univ. Rennes, 54 p. Lister, G., Foster, M., Compagnoni, R., Giles, D., Tamagno, E., Hills, Q., Betts, P. & Beltrando, M. (in prep.) Geology of the Ultra-High Pressure Region at Lago di Cignana, Valtournenche, Aosta Valley, Italy. Matsumoto, N. & Hirajima, T. (2000): Garnet in pelitic schists from a quartz-eclogite unit of the southern Dora-Maira massif, Western Alps. Schweiz. Mineral. Petrogr. Mitt., 80:53–62. Michard, A., Chopin, C. & Henry, C. (1993): Compression versus extension in the exhumation of the Dora Maira coesite-bearing unit, western Alps, Italy. Tectonophysics, 221:173–193.

48

R. Compagnoni & F. Rolfo

Michard, A., Henry, C. & Chopin, C. (1995): Structures in ultrahigh-pressure metamorphic rocks. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 132–158. Mirwald, P.W. & Massonne, H.-J. (1980): The low-high quartz and quartz-coesite transition to 40 kbar between 600° and 1600°C and some reconnaissance data on the effect of NaAlO2 component on the low quartzcoesite transition. J. Geophys. Res., 85:6983–6990. Monié, P. & Chopin, C. (1991): 40Ar/39Ar dating in coesite-bearing and associated units of the Dora Maira Massif, western Alps. Eur. J. Mineral., 3:239–262. Nowlan, E.U., Schertl, H.P. & Schreyer, W. (2000): Garnet-omphacite-phengite thermobarometry of eclogites from the coesite-bearing unit of the southern Dora-Maira Massif, Western Alps. Lithos, 52:197–214. Paquette, J.L., Chopin, C. & Peucat, J.J. (1989): U-Pb zircon , Rb-Sr and Sm-Nd geochronology of high- to very-high-pressure meta-acidic rocks from the western Alps. Contrib. Mineral. Petrol., 101:280–289. Paquette, J.L., Montel, J.-L. & Chopin, C. (1999): U-Th-Pb dating of the Brossasco ultrahigh-pressure metagranite, Dora-Maira Massif, western Alps. Eur. J. Mineral., 11:69–77. Perkins, D., Westrum, E.F. III & Essene, E.J. Jr. (1980): The thermodynamic properties and phase relations of some minerals in the system CaO-Al2O3-SiO2-H2O. Geochim. Cosmochim. Acta, 44:61–84. Pfeifer, H.-R., Colombi, A. & Ganguin, J. (1989): Zermatt-Saas and Antrona Zone; a petrographic and geochemical comparison of polyphase metamorphic ophiolites of the west-central Alps. Schweiz. Mineral. Petrogr. Mitt., 69:217–236. Philippot, P., Chevallier, P., Chopin, C. & Debussy, J. (1995): Fluid composition and evolution in coesitebearing rocks, Dora-Maira massif, Western Alps: implications for element recycling during subduction: Contrib. Mineral. Petrol., 121:29–44. Powell, R. (1985): Regression diagnostics and robust regression in geothermometer/geobarometer calibration: the garnet-clinopyroxene geothermometer revisited. J. Metamorph. Geol., 3:231–243. Reddy, S.M., Wheeler, J. & Cliff, R.A. (1999): The geometry and timing of orogenic extension; an example from the Western Italian Alps. J. Metamorph. Geol., 17:573–589 Reinecke, T. (1991): Very high pressure metamorphism and uplift of coesite-bearing metasediments from the Zermatt-Saas zone, Western Alps. Eur. J. Mineral., 3:7–17. Reinecke, T. (1998): Prograde high- to ultrahigh-pressure metamorphism and exhumation of oceanic sediments at Lago di Cignana, Zermatt-Saas zone, Western Alps. Lithos, 42:147–190. Reinecke, T., van der Klaw, S.N. & Stöckert, B. (1994): UHP metamorphic oceanic crust of the Zermatt-Saas Zone (Piemontese Zone) at lago di Cignana, Valtournenche, Italy. In Compagnoni, R., Messiga, B. & Castelli, D. (eds.): High pressure metamorphism in the Western Alps. Guide-book to the field excursion B1, 16th Gen. Meeting of the IMA, Pisa, 10–15 September 1994, 15–16. Rubatto, D. & Hermann, J. (2001): Exhumation as fast as subduction? Geology, 29:3–6. Rubatto, D., Gebauer, D. & Fanning, M. (1998): Jurassic formation and Eocene subduction of the ZermattSaas Fee ophiolites: implications for the geodynamic evolution of the Central and Western Alps. Contrib. Mineral. Petrol., 132:269–287. Rubatto, D., Liati, A. & Gebauer, D. (2003): Dating UHP metamorphism. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 341–363. Sandrone, R., Cadoppi, P., Sacchi, R. & Vialon P. (1993): The Dora-Maira Massif. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 317–325. Schertl, H.-P., Schreyer, W. & Chopin, C. (1991): The pyrope-coesite rocks and their country rocks at Parigi, Dora Maira massif, western Alps: detailed petrography, mineral chemistry and PT-path. Contrib. Mineral. Petrol., 108:1–21. Schreyer, W., Massonne, H.-J. & Chopin, C. (1987) Continental crust subducted to mantle depths near 100 km: implications for magma and fluid genesis in collision zones. In Mysen, O. (ed.): Magmatic processes: physicochemical principles /Geochem. Soc., Spec. Publ., 1/. St. Louis (Mo.): Geochem. Soc., 155–163. Sharp, Z.D., Essene, E.J. & Hunziker, J.C. (1993): Stable isotope geochemistry and phase equilibria of coesitewhiteschists, Dora Maira Massif, Western Alps. Contrib. Mineral. Petrol., 114:1–12.

UHPM units in the Western Alps

49

Simon, G. & Chopin, Ch. (2001): Enstatite-sapphirine crack-related assemblages in ultrahigh-pressure pyrope megablasts, Dora-Maira massif, western Alps. Contrib. Mineral. Petrol., 140:422–440. Simon, G., Chopin, C. & Schenk, V. (1997): Near-end-member magnesiochloritoid in prograde-zoned pyrope, Dora-Maira Massif, Western Alps. Lithos, 41:37–57. Sobolev, N.V., Shatsky, V.S., Vavilov, M.A. & Goryainov, S.V. (1994): Zircon from ultrahigh-pressure metamorphic rocks of folded regions as an unique container of inclusions of diamond, coesite and coexisting minerals. Dokl. Ross. Akad. Nauk., 334:488–492. Spear, F. (1988): The Gibbs method and Duhem’s theorem: The quantitative relationships among P, T, chemical potential, phase composition and reaction progress in igneous and metamorphic systems. Contrib. Mineral. Petrol., 99:249–256. Spear, F. & Menard, T. (1989): Program GIBBS, a FORTRAN program for Gibbs method calculations. Troy (N.Y.): Rensselaer Polytech. Inst. Stellato, S. (1999): Trasformazioni metamorfiche alpine delle cornubianiti Varisiche relitte nei massicci del Gran Paradiso e del Dora Maira (Alpi Occidentali). Tesi di laurea, Univ. Torino, Italy, 198 p. Tamagno, E. (1999): L’unità metaofiolitica di pressione molto alta del Lago di Cignana (Valtournenche Aosta) e le sue relazioni con le unità tettoniche vicine. Tesi di laurea, Univ. Torino, Italy, 188 p. Tilton, G.R., Schreyer W. & Schertl, H.-P. (1991): Pb-Sr-Nd isotopic behaviour of deeply subducted crustal rocks from the Dora Maira Massif., Western Alps, Italy - II: what is the age of the ultrahigh-pressure metamorphism? Contrib. Mineral. Petrol., 108:22–33. Tilton, G.R., Ames, L., Schertl, H.-P. & Schreyer, W. (1997): Reconnaissance isotopic investigations on rocks of an undeformed granite contact within the coesite-bearing unit of the Dora Maira Massif. Lithos, 41:25–36. Turello, R. (1993): Studio geologico e petrografico dell’area tra Martiniana (Valle Po) e Brossasco (Valle Varaita). Tesi di laurea, Univ. Torino, Italy, 237 p. van der Klauw, S.N.G.C., Reinecke, T. & Stockhert, B. (1997): Exhumation of ultrahigh-pressure metamorphic oceanic crust from Lago di Cignana, Piemontese zone, western Alps: the structural record in metabasites: Lithos, 41:79–102. Vialon, P. (1966): Etude géologique du massif cristallin Dora-Maira, Alpes cottiennes internes, Italie. (Thèse de Doctorat d’Etat) /Trav. Lab. Géol. Grenoble, Mém., 4/, 293 p. Wagner-Zweigel, P. (1993): Areal geology, fabrics and petrology of rocks SW of Lago di Cignana (Valtournanche, N-Italy). Diploma Thesis, Ruhr-Universität Bochum, Germany, 102 p.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 3, 51–73

Ultrahigh pressure metamorphism in the Western Gneiss Region of Norway DENNIS A. CARSWELL1* and SIMON J. CUTHBERT2 1

Department of Geography, University of Sheffield, Dainton Building, Brookhill, Sheffield S3 7HF, UK 2 Biological & Geological Sciences, School of Engineering & Science, University of Paisley, Paisley PA1 2BE, UK; * e-mail: [email protected]

Historical background to UHPM in western Norway Eskola (1921) drew attention to some of the aesthetically impressive eclogites and garnet peridotites that outcrop in the coastal region of west Norway between Bergen and Trondheim. These occurrences lie within the so-called Western Gneiss Region (WGR), the lowest exposed structural level in the southern Scandinavian Caledonides. The WGR is now recognised as a composite tectono-metamorphic terrane that mostly comprises Proterozoic autochthonous to para-autochthonous basement rocks with minor late Proterozoic cover belonging to the leading edge of the Baltic Plate, along with infolds of the main, outboard-derived Caledonian allochthon. Much of this composite edifice experienced short-lived deep level subduction beneath the Laurentian Plate during the Scandian phase of the Caledonian orogeny. Several more recent papers, including those by Andersen et al., (1991); Carswell et al. (2003a); Cuthbert et al. (1983, 2000); Cuthbert & Carswell (1990); Dewey et al. (1993); Griffin et al. (1985); Krogh & Carswell (1995); Smith (1995), have considered the stabilisation and exhumation of eclogites and other cofacial high pressure (HP) and ultrahigh pressure (UHP) rocks in this region, within the context of the tectono-metamorphic development of this segment of the Scandinavian Caledonides. Smith (1984) provided the first description, as well as confirmation by Raman spectroscopy, of an occurrence of coesite within an eclogite in the WGR. This coesite (CD Images 1 & 2) is preserved (armoured within omphacite in turn enclosed in garnet) within a small, partly retrograded, eclogite pod at Grytting, near Selje, in the SW part of the Stadlandet peninsula (Fig. 1). Interestingly, but perhaps just coincidentally, this coesite-bearing eclogite outcrops closely adjacent to the more spectacular-looking coarse orthopyroxene eclogite (CD Image 3) described by Eskola (1921). Importantly, thermobarometric evaluation by Lappin & Smith (1978) and Cuthbert et al. (2000) of samples of this orthopyroxene eclogite indicates formation under UHP conditions consistent with the stability of coesite in the nearby pod. Smith’s (1988) expansive review article on WGR eclogites, documented confirmed coesite at only one other eclogite locality named Straumen, some 14 km SW of Grytting.

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

52

D.A. Carswell & S.J. Cuthbert

Fig. 1. A generalised geological map for the Western Gneiss Region between the Sognefjord and Moldefjord areas showing the respective distributions of documented occurrences of UHP (coesite-bearing) eclogites, UHP diamond gneiss and peridotite bodies.

UHPM in the Western Gneiss Region of Norway

53

In addition, he deduced the likely previous presence of coesite from observations of polycrystalline or at least multi-crystalline quartz inclusions within garnet or omphacite in samples from five additional eclogite localities, including Årsheimneset, Drage and Liset on Stadlandet. Such polycrystalline quartz (PCQ) inclusions, sometimes with a distinctive palisade microstructure (CD Image 4), are now widely accepted to be pseudomorphs after earlier coesite crystals, some of which have a distinctive tabular form (CD Image 5). On this rather limited evidence, Smith (1988) proposed the existence within this coastal region of the WGR of a specific coesite eclogite province containing rocks that had experienced UHPM conditions. However, Smith (1988) did not establish the boundaries for this UHPM province and moreover emphasised that most, if not all, of the intervening “country-rock” gneisses enclosing the various documented eclogite occurrences in this part of the WGR lacked mineralogical evidence that they had witnessed HP (quartz eclogite stable), let alone UHP (coesite eclogite stable), conditions. Accordingly, Smith (1988) in fact concluded that, rather than the WGR incorporating a regionally extensive, structurally coherent, UHPM province or terrane, it in fact comprised a highly imbricated tectonic melange of HP and UHP eclogites enclosed within dominant lower pressure metamorphic rocks. Smith (1995) further expounded his view that the geographically scattered coesite eclogite occurrences represent tectonically dismembered fragments of an early Caledonian (ca. 440 Ma) UHPM nappe within the WGR, and in his elaborate Foreign/In situ/Foreign (FIF) geodynamic model ne introduced the notion that the quartz-stable eclogites may in fact have formed during a later Caledonian (ca. 410 Ma) HPM (Pmax ca. 2.0 GPa) event that affected a substantial segment of mixed acid + basic lithology crust. A further significant step in the recognition of a possible, regionally extensive, UHPM terrane within the WGR of Norway, was the reported recovery by Dobrzhinetskaya et al. (1993, 1995) of micro-diamonds from dissolution of samples of garnet-kyanite-biotite-rutile-quartz gneiss (CD Image 6) and of garnet-pyroxeneamphibole-biotite gneiss from the north coast of the island of Fjørtoft in Nordøyene, about 80 km to the north of the first coesite eclogite occurrences documented by Smith (1984, 1988). Consequently, in their global overview of the then recognised UHPM terranes, Coleman & Wang (1995) speculated that the UHPM terrane in western Norway might cover an area of roughly 350 × 150 km. Even given more conservative estimates of the size of the UHPM terrane, it is clear from the lithostatic pressures required for UHPM that a substantial mass of initially buoyant continental crust has been inserted (subducted) into the sub-lithospheric mantle, and subsequently exhumed, during the late Silurian to Middle Devonian Scandian plate collision episode.

Most recent discoveries of UHPM rocks in western Norway Meticulous field and petrographic studies in the outer Nordfjord and Stadlandet areas reported by Wain (1997) have greatly increased the number of eclogite localities recognised to have experienced UHPM. Actual relict coesites were identified as microinclusions in garnet, omphacite or kyanite in eclogite from five new localities and

54

D.A. Carswell & S.J. Cuthbert

petrographically distinctive PCQ pseudomorphs after coesite were recognised in eclogite minerals at a further twelve localities. Cuthbert et al. (2000) reported discoveries in the outer Nordfjord and Stadlandet areas of a further relict coesite-bearing eclogite locality at Flister and of PCQ pseudomorphs after coesite in eclogite pods at Maurstad and Sandvikneset. In addition, convincing PCQ inclusions after coesite within garnet were reported from a large eclogite body capping the peaks of Hornet and Bautene to the south of the extensive peridotite outcrops in Almklovdalen and in a smaller eclogite pod in Stigedalen, south of the peridotite outcrops in Bjørkedalen, that together extend the occurrences of UHP eclogites some 40 km east of previously recognised localities (Fig. 1). The present authors have also recently discovered convincing PCQ inclusions after coesite in garnets within a small flaser-textured eclogite body (CD Image 7) that outcrops virtually adjacent to the classic HP kyanite eclogite body (CD Image 8) at Verpeneset on the north shore of Nordfjord. This then is the most southerly UHP eclogite occurrence so far identified within the WGR. Cuthbert et al. (2000) also reported the discovery, confirmed by Terry et al. (2000b), of PCQ inclusions after coesite (CD Image 5) in a pod of kyanite-phengite eclogite (CD Image 9) at Fjørtoftvika on the island of Fjørtoft, some 85 km northeast of the UHP eclogites localities recognised in the Nordfjord and Stadlandet areas. Importantly this UHP eclogite locality is only about 2 km along strike from the outcrops of the graphite-bearing garnet-kyanite-biotite gneiss from which scarce micro-diamonds were recovered by Dobrzhinetskaya et al. (1993, 1995). Compelling new evidence that the rocks exposed along the north coast of Fjørtoft experienced UHPM conditions that extended into the diamond stability field is provided by the recent startling discovery of micro-diamonds in a garnet websterite lens within a small outcrop of peridotite at Bardane. These micro-diamonds are interpreted to have formed in response to infiltration by crustal-derived fluids during Caledonian deformation-induced recrystallisation (van Roermund et al., 2002; Brueckner et al., 2002). Smith (1988, 1995) reported finding poly- or multi-crystalline quartz thought to have replaced earlier coesite in the large Ulsteinvik–Dimnøy eclogite body on Hareidlandet and in the Hessdalen eclogite body on the opposite side of Vartdalsfjorden, these localities being roughly midway between the occurrences of UHPM rocks on Stadlandet and Fjørtoft (Fig. 1). The UHP status of the Ulsteinvik–Dimnøy eclogite has been confirmed by Carswell et al. (2003b) through the discovery of preserved inclusions of coesite within a zircon separate from this eclogite body. Further evidence that rocks in this part of the WGR also experienced UHPM conditions is provided by the reported discovery by Hacker et al. (2001) of numerous eclogite localities with relict coesite or PCQ pseudomorphs after coesite on the Sørøyane islands to the west of Hareidlandet. This steady increase over recent years in the number of recognised occurrences of UHPM rocks in the coastal region of the WGR between Nordfjord and the Nordøyene suggests that the coesite eclogite-bearing UHPM terrane in the WGR covers an area of at least 5000 km2 in a coastal strip up to 40 km wide (Fig. 1).

UHPM in the Western Gneiss Region of Norway

55

The “foreign” versus “in situ” eclogite controversy: The influences of differential retrogression and metastability The HP and UHP eclogites in the WGR occur as highly variable sized lenticular pods or tabular layers within highly deformed, dominantly quartzo-feldspathic, gneisses (e.g. CD Image 10). Typically the margins to the eclogite bodies show retrogression to amphibolite facies mineralogies (CD Image 11), co-facial with the mineralogy observed in the surrounding gneisses. Sometimes the amphibolitisation at eclogite margins is a static growth feature (e.g. CD Image 12) apparently triggered by an influx of hydrous fluid but in other instances the strong shear deformation fabric seen in the surrounding gneisses may be seen to have penetrated the margins of the eclogite bodies resulting in the development of a strongly foliated amphibolite (CD Image 13). Not uncommonly, especially towards the margins of the eclogite bodies, omphacites show at least partial replacement by a granulite facies symplectite of secondary jadeite-depleted clinopyroxene plus sodic plagioclase (e.g. CD Images 1 & 2), as a result of a retrogressive, granulite facies, decompression stage that preceded the amphibolitisation. Given the obvious major contrasts in metamorphic grade between the unretrograded eclogites and the encompassing amphibolite facies gneisses, it has remained uncertain and controversial as to whether or not the “country rock” gneisses enclosing the eclogites, garnet peridotites and other recognised HP and UHP rocks in the WGR experienced comparable HPM or UHPM conditions. Consequently, there has been a prolonged debate over whether the observed HP and UHP rocks were stabilised “in situ” within the gneisses in an essentially structurally coherent metamorphic terrane or alternatively represent “exotic” blocks or lenses of HP and UHP rocks within some sort of highly disrupted tectonic melange, as envisaged for example by Smith (1980). Krogh (1977) and Griffin et al. (1985) established the existence of a thermal gradient from around 550 °C in the SE to > 800 °C in the coastal areas to the NW (Fig. 2) for the HPM/UHPM across the WGR from consideration of Fe2+/Mg2+ partitioning between garnet and omphacite in eclogite samples. From an updated thermobarometric evaluation of phengite-bearing and orthopyroxene-bearing eclogites, Cuthbert et al. (2000) established that the regional temperature gradient is matched by a gradient of increasing lithostatic pressures, consistent with the stability of coesite to the north of Nordfjord and of diamond only in the most northwesterly exposed part of the WGR in the Nordøyene. This P–T analysis supports two important conclusions. Firstly, that the rocks in the most northwesterly part of the WGR were subducted to deepest levels during the Caledonian plate collision. Secondly, that the subduction-related P–T gradient across the WGR (Fig. 2) has not been greatly disrupted by the widely displayed late orogenic, exhumation related, amphibolite facies, proto-mylonitic fabrics that developed in response to extensional, top-to-west, shear deformation (Andersen et al., 1991). Terry et al. (2000b) have proposed the existence of a major metamorphic discontinuity on the island of Fjørtoft between a higher structural unit/plate that records UHPM conditions of ca. 820 °C and 3.4–3.9 GPa and a lower unit/plate that only records HPM conditions of ca. 780 °C and 1.8 GPa. In contrast to the melange model of Smith (1988), Terry et al. (2000b) placed the HP/UHP junction at the lower contact of a

56

D.A. Carswell & S.J. Cuthbert

Fig. 2. Regional temperature gradient across the Western Gneiss Region of Norway based on Fe2+/Mg2+ partitioning between garnet and omphacite in eclogites (after Griffin et al., 1985) plus indications of the concomitant pressure gradient based on P–T estimates for eclogite samples from various localities indicated by Cuthbert et al. (2000), Carswell et al. (2003a) and Terry et al. (2000b).

regionally extensive and coherent sequence of thrust nappes (Blåho and Saetra nappes) with para-autochthonous Baltica basement (Ulla gneiss and other migmatitic or augen gneisses). No evidence for UHPM was reported within the lower plate gneisses or their enclosed eclogites. However, pods and layers of garnet peridotite are found within these Baltica basement gneisses, including the important Bardane UHP microdiamondbearing peridotite on Fjørtoft (see below). Also, the documented UHP kyanite eclogite lies in close proximity to the contact of the upper and lower plates as defined by Terry et al. (2000b), and the high strain state of the gneisses makes exact definition of the boundary problematic, so the assignation of this eclogite to one plate or the other is problematic. When these uncertainties are considered along with the overall rarity of

UHPM in the Western Gneiss Region of Norway

57

evidence for coesite in the WGR (see next section) and the general lack of useful parageneses for geobarometry in the Baltica basement in the northwestern WGR, the HP vs. UHP status of the lower plate remains, in our view, a somewhat open question, and it is possible that both units were juxtaposed before, or during, UHPM. If the lower plate is eventually proven to have recorded UHPM, then the survival of primary, low P, igneous phases in large masses of partially eclogitised metagabbro (Mørk, 1985a) raises the possibility of extreme metastability under UHP conditions here, as it does further south in the WGR (Wain et al., 2001). Indeed, a key issue in understanding the distribution of HP, UHP and lower P facies is the operation of factors controlling the efficiency of metamorphic transformations. As we argue further below, such factors need to be considered carefully before appealing to the tectonic juxtaposition of non-cofacial rock masses, especially when direct evidence for a tectonic break is lacking. Cuthbert & Carswell (1990) and Griffin et al. (1985) have previously summarised various lines of evidence and arguments in favour of an essentially “in situ” formation of most crustal protolith eclogites within the WGR during short-lived deep subduction of a slab of continental crust. The two most compelling lines of evidence supporting this interpretation are: 1) Widespread occurrences of corona-textured metadolerites (e.g. Gjelsvik, 1952) and metagabbros (e.g. Griffin & Råheim, 1973; Mørk, 1985a,b; Krabbendam et al., 2000) that display incomplete transformation to eclogite (CD Images 14&15). Crucially, some of these preserve primary igneous contacts with granitoid gneisses (Cuthbert, 1985). It is apparent that the extent of eclogitisation is controlled by the extent of influx of fluids and/or concomitant deformation and hence is crucially dependent on reaction kinetics (Austrheim, 1998). Not only do such rocks demonstrate eclogite formation under conditions of increased P and T from lowpressure (high crustal level) protoliths but they also provide clear evidence of plagioclase metastability under eclogite facies conditions. 2) Many eclogites, especially the quartz-stable HP types in the vicinity of Nordfjord, contain large garnets that display evidence of a prograde growth history under conditions of increasing P and T (CD Image 16). Such garnets show a compositional zonation with a marked increase in Mg/Fe ratio from core to rim. Not infrequently they also show a zonation in the entrapped mineral inclusions with the blue-green amphiboles in garnet cores and omphacite inclusions within later growth garnet (CD Image 17). The growth of such garnet thus apparently commenced under amphibolite facies conditions and continued under subsequent eclogite facies conditions. As indicated earlier, the margins of eclogite and garnet peridotite bodies mostly show retrogression to amphibolite, frequently linked to the development of a deformation fabric (e.g. CD Image 13). It might be tempting to link this retrogression with tectonic emplacement of the HPM/UHPM rocks into higher crustal level, lower pressure, gneisses, but then not all eclogite body margins are the focus of such shear deformation and invariably the scale of any observed deformation is inconsistent with the notion that the HPM or UHPM rocks have been thrust upwards by some 30–90 km, to account for the confining pressure contrasts with the enclosing amphibolite facies gneisses.

58

D.A. Carswell & S.J. Cuthbert

Hence, in the almost ubiquitous absence of obvious petrographic evidence that the “country rocks” enclosing the HPM and UHPM rocks also witnessed comparable P–T conditions, the marked contrasts in observed metamorphic grade seem best explained by arguing that the late orogenic deformation associated with the exhumation of this HP/UHP terrane has been partitioned largely into the more ductile quartzo-feldspathic gneisses. This resulted in an essentially pervasive metamorphic reworking of the gneisses and the replacement of any earlier HP/UHP mineralogies by retrograde amphibolite facies assemblages. By contrast, the HP/UHP mineralogies have had much better survival rates in the more structurally competent mafic and ultramafic rocks. Hence although pervasive tectonism has undoubtedly been responsible for small-scale relative movements between eclogites and country rocks, it is unlikely to provide an adequate explanation for the large-scale relative motions required to produce the striking contrasts in metamorphic grade between the eclogite pods and the encompassing gneisses observed throughout the WGR. Rather, these contrasts result from the variable efficiency with which the HP or UHP parageneses were overprinted. Wain (1997), Krabbendam & Wain (1997) and Wain et al. (2000) have provided crucial evidence in support for this differential retrogression interpretation, through their demonstration that small volumes of schists and gneisses in low strain zones immediately adjacent to certain UHP eclogite occurrences on Stadlandet have partly preserved UHP mineralogies, including petrographic evidence of previous coesite stability. Within the WGR, pelitic paragneisses are relatively scarce but where they do occur, as in outer Nordfjord and on Fjørtoft, they do provide some good petrographic evidence that they experienced HPM or UHPM conditions, even despite their propensity and susceptibility to retrogression. Indications are that they contained Pmax assemblages of garnet + phengite ± kyanite ± zoisite ± omphacite + rutile + quartz/coesite. Only in rare instances, such as in the UHPM schists along the shore outcrops at Vetrhuset on Nordpollen (Fig. 3), has petrographic evidence been preserved for the previous stability of coesite. By comparison, finding evidence that the voluminous Proterozoic acid–intermediate orthogneisses (Lappin et al., 1979; Harvey, 1983) that outcrop extensively within the WGR also experienced HPM or UHPM conditions is a much greater challenge. Unequivocal proof that this was the case may, as with comparable rocks the Dabieshan–Sulu UHPM terrane in central China (Ye et al., 2000; Liu et al., 2002), ultimately depend on searching for preserved micro-inclusions of coesite or other HP/UHP mineral phases, within zircon mineral separates. Given the previously emphasised indications of extreme mineral metastability in metadolerite and metagabbro bodies within the WGR, metastability should also be seriously considered as an additional or alternative explanation for the apparent lack of HP or UHP mineralogies in the orthogneisses. It is feasible that such rocks may have witnessed the HPM or UHPM conditions but responded in only a limited and incomplete manner. Some little deformed granitic rocks both on Vagsøy in outer Nordfjord and further north in the Moldefjord region (Carswell & Harvey, 1985) do show limited coronitic

UHPM in the Western Gneiss Region of Norway

59

Fig. 3. Map for the Outer Nordfjord and Stadlandet area of the Western Gneiss Region showing the relative distributions of documented occurrences of UHP (coesite-stable) and HP (quartz-stable) eclogites relative to the HP–UHP Transition Zone boundaries (solid lines) indicated by Wain et al. (2000) and the modified boundaries (broken lines) proposed in this paper.

60

D.A. Carswell & S.J. Cuthbert

development of garnet. Intrusive acid igneous rocks have low inherent H2O contents bound into small amounts of micas and/or hornblende. During prograde, subductionrelated HPM or UHPM this limited H2O content is likely to become locked into newly formed phengite and zoisite, minerals that can demonstrably remain stable to UHPM conditions. Thus further reactivity may be inhibited in such rocks unless deformation provides channels for fluid ingress. The fact that many of the orthogneisses are observed to retain porphyroclasts of unmixed high-temperature feldspars is a strong pointer to frequent metastability in these rocks under HPM or UHPM conditions. More intermediate composition meta-igneous rocks, such as the mangerite at Flatraket (Krabbendam et al., 2000; Wain et al., 2001), show more extensive but still incomplete reactivity, again with residual plagioclase metastability.

The HPM to UHPM transition Wain (1997) and Wain et al. (2000) observed that coesite or PCQ-bearing UHP eclogites apparently had a set of petrographic characteristics that were distinct from eclogites lacking any evidence for coesite (interpreted as quartz-stable HP eclogites). UHP eclogites were reported to generally contain xenoblastic garnets that lack clear prograde compositional zoning and contain only eclogite facies solid inclusion suites, including coesite or PCQ. HP eclogites commonly contain idioblastic garnet, the larger grains of which enclose monocrystalline quartz inclusions throughout, exhibit strong prograde compositional zoning and have amphibolite facies inclusion suites in their cores. This latter type is exemplified by the eclogites at Verpeneset, Almenningen and Kroken along the northern side of Nordfjord (Fig. 3), so we name this petrographic group the “VAKtype” eclogites. A significant number of eclogites have ambiguous petrographic characteristics, such as those lacking prograde zoned garnet with inclusions of early amphiboles in cores but also lacking evidence of coesite stability. UHP eclogites first appear along the north shore of Nordfjord (Fig. 3) and persist northwards, being particularly in evidence around the southern part of the Stadlandet peninsula. The region stretching south from Nordfjord to Sognfjord exposes only VAK-type, HP eclogites (Krogh & Carswell, 1995; Cuthbert et al., 2000). An important observation, arising from the work of Wain (1997) and Wain et al. (2000), was that HP-type eclogites persist for a distance of about 10 km north of the coesite-in line (southern solid line in Fig. 3). Accordingly, they defined a mixed, or transitional, HP/UHP zone whose northern boundary stretched from across the southern end of the Stadlandet peninsula westwards towards Nordpollen (northern solid line on Fig. 3). Within this transition zone, Wain et al. (2000) described the discretely separate HP and UHP eclogite bodies as occurring up to a minimum of 100 metres from each other. Thermobarometric evaluation of mineral reaction equilibria, in particular for phengite-bearing eclogite samples (Wain, 1997; Cuthbert et al., 2000), has for the most part corroborated the barometric distinctions between HP and UHP eclogite samples deduced from petrographic criteria. Even allowing for generous error brackets,

UHPM in the Western Gneiss Region of Norway

61

significant non-lithostatic apparent pressure gradients have been recorded between adjacent HP and UHP eclogites (Wain et al., 2000; Cuthbert et al., 2000). Wain (1997), Krabbendam & Wain (1997) and Wain et al. (2000) thus attributed the close juxtapositioning of HP and UHP eclogites within the transition zone to the tectonic interleaving of different structural units. Accordingly it was assumed that the lithotectonic unit containing the UHP eclogites was carried to a higher lithospheric level and emplaced against a different lithotectonic unit containing only HP eclogites. In our opinion, a number of difficulties arise with this tectonic mixing interpretation of the HP–UHP transition zone in the Outer Nordfjord area. Firstly, no obvious zones of higher strain are observed between outcrops of HP and UHP eclogites within the transition zone, nor is this zone as a whole characterised by higher strains than in the rocks on either side. It is, of course, possible that the pervasive, late, exhumationrelated amphibolite facies deformation and recrystallisation in the WGR (Andersen et al., 1991; Krabbendam & Wain, 1997) has obliterated fabrics associated with such shear zones. Indeed some relative movement between adjacent eclogite bodies must have happened during development of these late fabrics, but such fabrics are not confined to the transition zone, so it cannot be regarded as a distinct displacement horizon during the later stages of exhumation. Belts of metasediments associated with meta-anorthosite, at least some of which are likely to be allochthonous, lie across the area north of Nordfjord within the predominant orthogneissic basement, (Bryhni, 1989; see Fig. 3). UHP eclogites are found in both paragneiss and orthogneiss units, and both types of eclogite can be found in the same lithological unit (Cuthbert et al., 2000). Hence we have found that HP or UHP eclogites cannot be exclusively assigned to particular lithotectonic units within the transition zone. Our second reservation concerning the tectonic mixing interpretation for the transition zone arises from the difficulty in unambiguously identifying the HP and UHP rocks. Identification of some HP eclogites effectively by default, based upon a lack evidence of coesite or PCQ, is unreliable due to the poor preservation potential of both coesite and its delicate PCQ replacement textures. Coesite is, in fact, frustratingly rarely preserved in the WGR compared to other UHP terrains such as the Dabie Shan of central China (Carswell & Zhang, 1999) but it is always possible that a single observation of coesite or PCQ in a prograde-zoned “VAK-type” (apparently HP) garnet will render the other petrographic criteria for identification of a HP eclogite invalid. Examples of prograde-zoned garnets with inclusions of coesite are certainly known from other UHP belts, such as in the Kokchetav Massif of Kazakhstan (Parkinson, 2000). The Årsheimneset UHP eclogite (Fig. 3) is known to exhibit prograde-zoned garnets with amphibole-rich inclusion suites (Carswell et al., 1985), but the same eclogite body also contains good palisade-textured PCQ (Smith, 1988; Cuthbert et al., 2000). Hence here a single body of eclogite clearly shows petrographic characteristics of both of Wain’s (1997) UHP and HP types, and potentially records development from amphibolite facies, through HP eclogite facies to UHPM. Clearly, then, a further weakness of the discriminatory HP and UHP eclogite classification of Wain (1997) is the assertion that eclogites in the transition zone are each exclusively HP or UHP in

62

D.A. Carswell & S.J. Cuthbert

character. Clearly if both types are found together in the same body, then it is difficult to envisage how they could have been brought together tectonically. In the light of these difficulties with the tectonic juxtapositioning interpretation, we have undertaken a detailed examination of certain eclogite bodies within the transitional HP/UHP zone and in the adjacent HP and UHP zones of the Nordfjord–Stadlandet region. A key locality is at Vetrhuset on the eastern shore of Nordpollen (Fig. 3), close to the northern margin of the transition zone and recognised as an UHP, coesite-bearing eclogite by Wain et al. (2000). Here, a swarm of eclogite pods lies within a belt of semipelitic schists. PCQ inclusions in both garnet and omphacite are quite common in these pods, in addition to much rarer, actual preserved, coesite. However, large, subidioblastic garnets are frequently, clearly, compositionally zoned with Fe, Mn and Ca-enriched cores and Mg-enriched rims (CD Image 18). The largest of these prograde-zoned grains contain concentrations of hornblende inclusions in the cores. Palisade-textured PCQ inclusions tend to be found in more xenoblastic garnets that may lie only a few millimetres from the zoned garnet grains. The enclosing semi-pelitic, phengite-kyanitequartz schists contain distinctive, purplish red garnets up to 5 cm in diameter displaying conspicuous prograde compositional zoning. Rarely, PCQ inclusions are found in the narrow, Ca and Mg-enriched garnet rims. Hence it is now apparent that intimately associated individual eclogite bodies and their host semi-pelitic schists may display the characteristics of both HP “VAK-type” and UHP garnets. The Vetrhuset eclogites display a range of deformation fabrics: coarser-grained eclogites with a weak linear omphacite-shape fabric tend to contain the prograde zoned “VAK-type” garnets, while more strongly lineated eclogites show PCQ inclusions in omphacite and in later grown or dynamically recrystallised garnet, especially in garnet-quartz streaks. Hence our observations at this location within the HP/UHP transition zone have revealed evidence for incomplete transformation from lower P metabasic and semi-pelitic rocks to UHP parageneses, leading to partial preservation of certain petrographic characteristics that are typical of HP eclogites. Coesite crystallisation appears to have been associated with a distinct, later stage of garnet growth that either mantles the earlier amphibolite facies and HP eclogite facies garnet, or with the recrystallisation of earlier garnet and omphacite during deformation. The coarse-grained, UHP, coesite-kyanite eclogite at Flatraket harbour (Wain et al., 2000) shows similar characteristics to the coarser-grained eclogites at Vetrhuset, with subidioblastic garnets having darker red cores that contain abundant hornblende inclusions (CD Image 19). Rare PCQ inclusions are found within the paler coloured rims of these garnets, so this eclogite clearly demonstrates evolution from amphibolite facies through HP to UHP eclogite facies. Parts of this eclogite body have spongy-textured, highly poikiloblastic garnets with only omphacite and kyanite inclusions, showing more complete recrystallisation to the UHPM paragenesis, perhaps because the original amphibole inclusions were less well-armoured within the garnets and hence more prone to decomposition. Hence the prior textural development of this eclogite seems to have had an important control on the efficiency of UHP recrystallisation in this case. Zircons from this body give TIMS U–Pb age spectra with peaks at ca. 405 Ma and ca.400 Ma

UHPM in the Western Gneiss Region of Norway

63

(Hacker et al., 2001). These ages may correspond to the HPM then UHPM growth episodes of garnet (Carswell et al., 2003a). As discussed above, the Årsheimneset eclogite (Fig. 3) also contains characteristics typical of the HP “VAK-type” eclogites, yet this body outcrops about 5 km into the UHP zone/province as recognised by Wain et al. (2000). An important observation here (CD Image 20) is that the central part of this body comprises mainly bimineralic (orthopyroxene-free) eclogite whereas adjacent to the top and bottom contacts with the country rock gneiss, the main eclogite facies rock type is an orthopyroxene-, phlogopiteand magnesite-bearing eclogite that becomes distinctly pegmatitic towards the lower margin, and is spatially associated with veins and lenses of phlogopite and/or phengitebearing glimmerites. The central, bimineralic eclogite frequently contains conspicuous, up to cm-sized, prograde zoned garnets with cores containing abundant inclusions of dark blue-green amphibole (CD Image 17). The coarser-grained orthopyroxene-bearing eclogite (CD Image 21) exhibits more irregularly-shaped, large garnets comprising early darker cores with blue-green amphibole inclusions, extensively overgrown by later UHP garnet containing PCQ inclusions after coesite, as well as frequent inclusions of clinopyroxene and phlogopite, indicating that the latter was a stable UHPM phase. Hence the Årsheimneset eclogite is another example that shows evolution of a single eclogite body from an early HP (quartz-stable), essentially bimineralic, eclogite, to a more siliceous, orthopyroxene- and phlogopite-bearing, eclogite in which substantial new growth of garnet, orthopyroxene and clinopyroxene took place under UHPM conditions and consequently trapped coesite inclusions. Thus, in this case, the UHP mineral growth is thought to have occurred in response to a substantial influx of metasomatising fluid from the enclosing continental crust gneisses. The transformation of the early, essentially bimineralic, HP eclogite into a coarser-grained, even pegmatitic, UHP eclogite can be followed along hydraulic fractures penetrating the former. Once again, field relationships give no support to an interpretation of tectonic mixing of HP and UHP eclogites as proposed by Krabbendam & Wain (1997) and Wain et al. (2000). U–Pb zircon data of Gebauer et al. (1985) for an orthopyroxene- and phlogopitebearing eclogite sample from the Årsheimneset body (labelled SAN-1 from east of “Sandviknaes”) indicate metamorphic zircon growth at ca. 400 Ma, supporting the argument for UHPM at approximately that time (Carswell et al., 2003a – see below). The evidence from the Vetrhuset, Flatraket and Årsheimneset eclogite bodies demonstrates that individual eclogites do not necessarily exhibit uniquely “HP” or “UHP” characteristics. This leads us to question the value of using prograde zoning and an apparent absence of PCQ or coesite inclusions in garnet to indicate that the rock has lacked a UHPM history. These examples demonstrate an evolution from amphibolite facies parageneses through HP to UHP eclogite during which the efficiency of the metamorphic transformation was limited. Furthermore, they indicate that a number of processes drove the transformation. The clearest manifestation of the transition to UHP parageneses in both eclogites and semi-pelitic schists is the development of new garnet, by overgrowth, recrystallisation and/or neoblast formation. Garnet growth appears to have been the result of a discontinuous series of reactions amongst its matrix phases. The temporally distinct stages of garnet growth were prompted by distinct deformation

64

D.A. Carswell & S.J. Cuthbert

events and/or triggered by influxes of externally derived fluids. In the light of these observations it is instructive to also examine some other eclogite bodies that occur within and outside the HP/UHP transition zone. The spectacularly layered coesite eclogite (CD Image 22) at Saltaneset (Fig. 3) first recognised by Wain (1997) lies in the UHP zone, some 5 km north of the northern boundary of the HP/UHP transition zone as defined by Wain (1997) and Wain et al. (2000). Carswell et al. (2003a) recognised within this eclogite body two generations of garnet growth – aggregates of deeper red, Ca-rich grains characterised by concentrations of tiny rutile needles in their cores, and overgrowths or discrete neoblasts of Ca-poorer and Mg-richer garnet. In conspicuous quartz-garnet layers, the Mg-rich garnet mantles aggregates of the earlier, Ca rich type, or exists as discrete, compositionally homogeneous grains (CD Image 23) containing remarkably abundant inclusions of PCQ after coesite, indicating that these layers were originally garnet coesitite rock. Such layers, or veins, are common in eclogites in the WGR and appear to be metamorphic segregations associated with fracturing and the infiltration of aqueous fluids. In the case of the Saltaneset eclogite the formation of the garnet-coesite layers is clearly associated with growth of the second generation, UHP garnet, so the HPM–UHPM transformation was aided by ingress of an aqueous fluid. Carefully extracted UHP garnet, omphacite plus whole rock from a sample of a garnet coesitite vein have yielded a Sm–Nd isochron age of 408.3 ± 6.7 Ma (Carswell et al., 2003a). At Flister, a swarm of eclogite pods within semi-pelitic schists and interlayers of meta-anorthosite lies close to the extreme northern edge of Wain’s (1997) transition zone (Fig. 3). Wain et al. (2000) described a typical HP eclogite at this locality. Cuthbert et al. (2000) subsequently reported the discovery of relict coesite at Flister, but it is not clear if this is the same pod as that discovered by Wain et al. (2000). The coesite-bearing body is a flaser-textured eclogite in which granular streaks of paler garnet are sometimes cored by deeper red porphyroclasts, indicating the break-up, recrystallisation and compositional adjustment of an earlier generation of amphibolite facies to HP eclogite facies garnet. Coesite or PCQ has been found only in the late-formed garnet, again indicating a distinct episode of garnet growth or recrystallisation under UHPM conditions, in this case clearly aided by deformation. Such flaser-textured eclogites form a distinctive textural type in the Nordfjord–Stadlandet area. Intriguingly, we have recently discovered the comparable flaser-textured eclogite at Verpeneset, Nordfjord (Fig. 3), that is separated by only a narrow screen of gneiss from what is one of the archetypal HP “VAK-type” eclogites, to also contain PCQ inclusions in neoblastic garnet. This pushes the southern limit of recognised UHP eclogites further south than that shown in Cuthbert et al. (2000). Wain et al., 2000 did not actually extend their boundary as far west as Verpeneset. It is clear that the location of both our modified coesite-in line and the northern limit of identified, pre-UHPM, HP eclogite relics (dashed grey lines in Fig. 3) remain only provisional until further detailed petrographic and thermobarometric studies are carried out. Nevertheless, it is evident that tectonic juxtapositioning of HP and UHP eclogites is much too simplistic a view, and we would argue that it played no more than a very minor role during the subduction and early exhumation phases of the Scandian

UHPM in the Western Gneiss Region of Norway

65

orogenic event. Instead, kinetic factors, dictated to a great extent by deformation and fluid activity, controlled the efficiency of transformation of HP and pre-HP parageneses to UHP eclogite facies parageneses as the descending continental slab passed into the stability conditions of coesite, and beyond the stability of amphibole, leading to apparent intermingling of HP and UHP eclogites (and schists) on all scales from millimetres to kilometres. The increase in modal garnet associated with the HP–UHP transition can be expected to have modified the petrophysical properties of the WGR rocks, such as their density and rheology. The general east–west trend of the coesite-in line (Figure 3) is apparently discordant to the trend of the eclogite isotherms shown in Figure 2. The significance of this is presently unclear, but is likely to be at least partly an artifact of the rather sparse geothermometry upon which the isotherms are based. A more detailed thermobarometric survey will be required to better resolve the regional variation in P and T.

The age of the UHPM It was widely assumed that the eclogite facies metamorphism in the WGR was Precambrian in age (e.g. Krogh, 1977) prior to publication by Griffin & Brueckner (1980, 1985) of Caledonian Sm–Nd ages for garnet–omphacite pairs from five eclogite samples. Subsequently the consensus view was that the HPM/UHPM event occurred at ca. 425 Ma, this corresponding to the mean Sm–Nd age obtained for these eclogite samples. For example, this age was assumed in the tectonic models for the stabilisation and exhumation for these HPM/UHPM rocks presented by Andersen et al. (1991), Cuthbert & Carswell (1990) and Dewey et al. (1993). The Sm–Nd mineral ages obtained by Griffin & Brueckner (1980, 1985) did not specifically discriminate between the timing of HPM and UHPM conditions in different parts of the WGR. Based mostly on this dataset, Smith (1988) subsequently speculated that the UHPM rocks formed at ca. 440 Ma and the HPM rocks at ca. 410 Ma, but the limited sample set and large uncertainties in the ages make this difficult to substantiate. Moreover, as documented in the preceding section, our recent petrographic observations on the relative timing of the HPM and UHPM mineralogies strongly point to the latter having been stabilised later rather than earlier than the HPM assemblages. Furthermore, as emphasised in the UHPM timing review paper by Carswell et al. (2003a), recently published U–Pb ages for zircons (Hacker et al., 2001; Carswell et al., 2003b) and for monazites (Terry et al., 2000a) in specifically identified UHPM rocks mostly fall within the rather later 400–410 Ma timeframe. Compelling evidence that the UHP mineralogies may indeed have formed at close to 400 Ma is provided by the 402 ± 2 Ma U–Pb age obtained for metamorphic zircons from the Hareidland eclogite shown to contain micro-inclusions of UHP minerals, including coesite (Carswell et al., 2003b). In addition, Carswell et al. (2003a) have reported a statistically indistinguishable Sm–Nd garnet–omphacite–whole rock isochron age of 408.3 ± 6.7 Ma for an eclogite sample from Salta (see Fig. 1) that contains abundant petrographic evidence of previous coesite stability (e.g. CD Image 4).

66

D.A. Carswell & S.J. Cuthbert

The conclusion that UHPM rocks in the WGR formed at close to 400 Ma, much later in the Scandian phase of the Caledonian orogenic cycle than was previously taken to be the case, has profound implications for the dynamic modelling of this continental plate collision belt and certainly signals extremely rapid exhumation of these UHPM rocks. Hence, Carswell et al. (2003a) have concluded that they were probably exhumed to ca. 35 km depth at a mean rate of ca. 1 cm per year. They suggested that this rapid initial exhumation could well have been driven by residual bouyancy of the deeply subducted crustal slab that resulted from incomplete eclogitisation. They particularly emphasised the survival of metastable, plagioclase-bearing assemblages in the dominant orthogneisses due to limited reactivity and the probable short duration of the UHPM event. Further exhumation to about 8–10 km apparently occurred at a much slower mean rate of ca. 1.3 mm per year with the final unroofing of the UHPM rocks attributable to the late-stage extension collapse of the Caledonian orogen (e.g. Andersen et al. 1991).

The occurrence and interpretation of garnet peridotites Occurrences of peridotite bodies of metre to kilometre-scale dimensions are a conspicuous feature of the WGR (Fig. 4). Chlorite-amphibole peridotites are most common (CD Image 24) and only a few bodies retain the pyropic garnet-olivine assemblage that is diagnostic of the stabilisation of HP/UHP eclogite facies assemblages (CD Image 25). Nonetheless, petrographic features, including the observed retrograde transitions from granoblastic garnet-bearing peridotite into foliated chlorite-amphibole peridotite within certain bodies, suggest that at least some of the designated chlorite peridotites in Figure 4 may have originally contained pyropic garnets. Carswell et al. (1983) drew attention to the existence of two fundamentally different chemical types of garnet-bearing peridotite bodies in the WGR. The Fe-Ti type garnet peridotites were interpreted to have had a prograde origin from the ultramafic portions of layered low pressure crustal intrusive bodies during Caledonian subductionrelated HPM/UHPM. The Eiksunddal eclogite complex on Hareidlandet with its subordinate garnet peridotite layers, well documented by Jamtveit (1987), is an excellent example of this garnet peridotite type. Jamtveit et al. (1991) were unable to obtain a meaningful Sm–Nd age for a garnet peridotite sample from the Eiksunddal complex, due to isotopic disequilibrium between the garnet, clinopyroxene and whole rock. However, they did obtain a Scandian age of 412 ± 12 Ma for coexisting garnet and clinopyroxene from a well-foliated eclogite sample from the same complex. By contrast, the Mg-Cr type garnet peridotites display characteristic upper mantle mineral and whole rock chemistries (Carswell, 1968) and isotopic ratios (Brueckner, 1977), consistent with derivation from mostly highly depleted sub-continental lithospheric mantle. Medaris (1984), Carswell (1986) and Medaris & Carswell (1990) have documented the complex, multistage metamorphic evolution experienced by these particular peridotite bodies. Carswell (1986) and Medaris & Carswell (1990) concluded that the earliest assemblage contained high temperature pyroxenes that, on the basis of scarce petrographic evidence, probably coexisted with spinel rather than with the oldest generation garnet.

UHPM in the Western Gneiss Region of Norway

67

Fig. 4. Map of the principal occurrences of garnet-bearing and chlorite-bearing (garnet-absent) peridotite bodies within the Western Gneiss Region of Norway.

Sm–Nd dating of garnet + clinopyroxene + whole rock in samples of Mg-Cr type garnet peridotite and associated olivine-free garnet pyroxenites by Mearns (1986) and Jamtveit et al. (1991) has importantly shown the earliest garnet-bearing assemblages in these mantle-derived peridotite bodies to be of mid-Proterozoic age and consequently to have been stabilised in the sub-continental mantle long before the Scandian, subductionrelated, HPM/UHPM event. An even more surprising discovery has been the recognition by van Roermund & Drury (1998) and van Roermund et al. (2000, 2001) of pyroxene exsolution lamellae (CD Image 26) in megacrysts of the earliest garnet generation (CD Image 27) in samples from the Mg-Cr type peridotite bodies outcropping on Otrøy. This feature has been taken to indicate that these garnets originally contained a significant majorite component. Van Roermund et al. (2000) used their deduced initial composition of this super-silicic majoritic garnet to estimate minimum pressures, at high temperature, of 6–6.5 GPa – interpreted (see also Drury et al., 2001) to indicate derivation within a rising mantle diapir originating from depths of at least 185 km. Subsequently, similar exsolution microstructures in garnets were discovered by Terry et al. (1999) in garnet peridotite/websterite bodies on Fjørtoft and Flemsøy (Fig. 4). Further petrographic observations and Sm–Nd isotopic data (Brueckner et al., 2002; van Roermund et al., 2002) on an occurrence of megacrystic garnets, with possible

68

D.A. Carswell & S.J. Cuthbert

majoritic garnet microstructures, (CD Image 28) within a small peridotite body at Bardane (CD Image 29) on the island of Fjørtoft (Fig. 4), have confirmed that these early garnets formed during the Gothian orogeny at around 1600–1700 Ma and moreover coexisted with high-temperature megacrystic aluminous pyroxenes (CD Image 30). Brueckner et al. (2002) thus considered that the earliest recognisable assemblage in the Mg-Cr type peridotite bodies at Bardane on Fjørtoft and at Ugelvik on Otrøy formed at conditions of ca. 3.0–4.5 GPa and 1300–1500 °C in the mid-Proterozoic. Accordingly, the metamorphic evolution of these peridotites as outlined by Carswell (1986) and Medaris & Carswell (1990) must be modified to eliminate the previous indication that the oldest assemblage formed at high temperature but relatively low pressure, with aluminous spinel rather than garnet coexisting with the aluminous pyroxenes. A revised, multistage metamorphic evolution for these Mg-Cr type peridotite bodies, that indicates instead the early stability of a majoritic garnet formed at UHP conditions deep in the mantle, is presented in Figure 5. Full details of the Proterozoic history and location of these peridotite bodies remain somewhat conjectural given the scarcity of preservation of stage 1 assemblages due to widespread overprinting by much later Palaeozoic/Caledonian assemblages. It seems likely that the stage 1 high P/T majoritic garnet-bearing assemblages were mostly recrystallised and reequilibrated to stage 2 assemblages, with more normal Cr-pyrope garnets, as these rocks were uplifted (perhaps within a rising mantle diapir) and cooled to the ambient Proterozoic geotherm. These mantle-derived, Mg-Cr type, garnet-bearing peridotite bodies in the WGR have been widely interpreted (e.g.. Medaris & Carswell, 1990; Krogh & Carswell, 1995) to have been tectonically emplaced into the subducted slab of continental crust during the Scandian plate collision. Brueckner (1998) and Brueckner & Medaris (2000) have presented new dynamic models for crust–mantle interaction in major continental plate collision zones and emphasised that different locations and timings are possible for the tectonic emplacement of the mantle peridotites into the crustal slab either during subduction or subsequent exhumation. A further dramatic new discovery with an important bearing on this issue and on the P–T–t evolutionary history experienced by these peridotites has been the discovery of microdiamonds within the Bardane peridotite occurrence on Fjørtoft (van Roermund et al., 2002). Rather surprisingly these microdiamonds apparently did not form in association with the UHP majoritic garnet during the mid-Proterozoic. Instead, combined petrographic, geochemical and isotopic data (Brueckner et al., 2002; van Roermund et al., 2002) provide strong pointers to the microdiamonds having formed in response to an influx of crustal-derived fluids attendant on Scandian-aged deformation and the recrystallisation of the oldest generation Proterozoic assemblage with its megacrystic majoritic garnets and aluminous pyroxenes. To date, preserved microdiamonds have only been discovered armoured within Cr-spinel grains that are in turn enclosed within later generation garnet that forms a corona network (CD Image 31) around deformed sub-grains of the original orthopyroxene megacrysts (van Roermund et al., 2002). Further ‘secondary’ generation garnet occurring as rational exsolution lamellae within the deformed orthopyroxene

UHPM in the Western Gneiss Region of Norway

69

Fig. 5. Revised outline for the deduced multi-stage metamorphic evolution of the mantle-derived, Mg–Cr type, garnet-bearing peridotite bodies in the Western Gneiss Region of Norway.

megacrysts is seen both at Bardane and in the comparable deformed and partly recrystallised megacrystal orthopyroxenite lens at Ugelvik on Otrøy (Carswell, 1973). It is difficult to extract an unquestionably pure separate of this later generation garnet. Accordingly, the rather equivocal Sm–Nd ages of 518 ± 78 Ma (Brueckner et al., 2002) and of ca. 561 Ma (Jamtveit et al., 1991) obtained for the exsolved garnet + clinopyroxene assemblages at Bardane and Ugelvik, respectively, are probably “mixed” ages but nonetheless signal the strong influence of Caledonian recrystallisation and isotopic re-equilibration. This latter interpetation is further supported by the Sm–Nd age of 437 ± 58 Ma obtained for garnet + clinopyroxene + whole rock by Jamtveit et al. (1991) from a thoroughly recrystallised, granoblastic textured, garnet pyroxenite sample from the Raudhaugene peridotite body on Otrøy. The deduction that the microdiamond-bearing assemblage within the Bardane peridotite body was probably stabilised at P–T conditions of around 3.4–4.1 GPa and 840–900 °C (Brueckner et al., 2002; van Roermund et al., 2002) during the subductionrelated Scandian-aged UHPM requires that this peridotite body must have been emplaced into the enclosing slab of continental crust either before or during its transient deep subduction. This conclusion is consistent with the nearby occurrence of a coesitebearing kyanite-phengite eclogite within the host quartzo-feldspathic gneisses (Terry et al., 2000b) and also of microdiamond-bearing garnet-kyanite-biotite-rutile-quartz gneisses that outcrop essentially along strike on the north coast of Fjørtoft (Dobrzhinetskaya et al., 1993, 1995). This surprising new interpretation that mantle-derived garnet-bearing peridotite bodies of the Mg-Cr type, similarly to the contrasting Fe-Ti type, also experienced Scandian, subduction-related, HPM/UHPM is further supported by certain previously

70

D.A. Carswell & S.J. Cuthbert

published but largely overlooked observations in other garnet-bearing peridotite bodies of the Mg-Cr type within the WGR. Griffin & Qvale (1985) documented the fact that Fe-rich eclogite lenses within the Mg-Cr type peridotite body at Raudkleivane in Almklovdalen contain garnets that show strong prograde compositional zoning and also contain ferropargasite inclusions in grain cores. This signals that this peridotite body must have been entrained into the subducting crustal slab and hence experienced a Scandian prograde metamorphic evolution from amphibolite to eclogite facies. Secondly, petrographic evidence has been recorded both in the peridotite body at Lien in Almklovdalen (Griffin & Heier, 1973) and in the Sandvika body (CD Images 32&33) on Gurskøy (Carswell et al., 1983) for the growth of new garnet in the outer parts of kelyphites composed of secondary aluminous pyroxenes and spinel around older garnet porphyroclasts. Griffin & Heier (1973) concluded that the apparent reversal of the kelyphite-forming reaction garnet + olivine o orthopyroxene + clinopyroxene + spinel was probably attributable to post-decompressional cooling but, in view of other observations, it is perhaps more appropriate to now attribute it to a later, post-kelyphite, prograde subduction-related HPM/UHPM event. Accordingly, the P–T evolution of the WGR garnet peridotites shown in Figure 5 incorporates the possibility of the existence of two separate granulite facies kelyphiteforming events, with the older one placed between the two pyropic garnet-bearing stages 2 and 4. It is conceivable that the older kelyphite-forming event occurred during the Sveconorwegian (ca. 1000–1100 Ma) orogeny, based on Sm–Nd systematics of zoned garnets from the Almklovdalen and Sandvika peridotite bodies (Brueckner et al., 1996). The long time interval between stages 2 and 4 creates uncertainly over exactly where these mantle-derived garnet-bearing peridotites resided. They conceivably may have already by this stage been emplaced in the Proterozoic continental crust. However, it is considered more likely that they resided in the uppermost sub-continental lithospheric mantle and were not incorporated into the crust until the deep subduction of the WGR during the Caledonian plate collision.

Acknowledgements The authors wish to acknowledge the support for recent studies of HP and UHP rocks in the WGR of Norway provided by the Norwegian Research Council, the Norwegian Geological Survey, the British Council, the EU “Access to Research Infrastructures” Programme, the Carnegie Trust for Scottish Universities and the Universities of Sheffield and Paisley. The section on the WGR garnet peridotites has benefited from helpful reviews by Herman van Roermund, Hannes Brueckner and Gordon Medaris.

References Andersen, T.B., Jamtveit, B., Dewey, J.F. & Swensson, E. (1991): Subduction and eduction of continental crust: major mechanisms during continent-continent collision and orogenic collapse, a model based on the south Norwegian Caledonides. Terra Nova, 3:303–310. Austrheim, H. (1998): Influence of fluid and deformation on metamorphism of the deep crust and consequences for the geodynamics of collision zones. In Hacker, B. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 297–323.

UHPM in the Western Gneiss Region of Norway

71

Brueckner, H.K. (1977): A crustal origin for eclogites and a mantle origin for garnet peridotites: strontium isotope evidence from clinopyroxenes. Contrib. Mineral. Petrol., 60:1–15. Brueckner, H.K. (1998): Sinking intrusion model for the emplacement of garnet-bearing peridotites into continent collision orogens. Geology, 26:631–634. Brueckner, H.K. & Medaris, G.L. (2000): A general model for the intrusion and evolution of “mantle” garnet peridotites in high-pressure and ultrahigh-pressure metamorphic terranes. J. Metamorph. Geol., 18:123–134. Brueckner, H.K., Blusztajn, J. & Bakun-Czubarow, N. (1996): Trace element and Sm–Nd ‘age’ zoning in garnets from peridotites of the Caledonian and Variscan Mountains and tectonic implications. J. Metamorph. Geol., 14:61–73. Brueckner, H.K., Carswell, D.A. & Griffin W.L. (2002): Paleozoic diamonds within a Precambrian peridotite lens in UHP gneisses of the Norwegian Caledonides. Earth Planet. Sci. Lett., 203:805–816. Bryhni, I. (1989): Status of the supracrustal rocks in the Western Gneiss Region, S. Norway. In Gayer, R.A. (ed.): The Caledonide geology of Scandinavia. London: Graham & Trotman, 221–228. Carswell, D.A. (1968): Possible primary upper mantle peridotite in Norwegian basal gneiss. Lithos, 1:322–355. Carswell, D.A. (1973): Garnet pyroxenite lens within the Ugelvik layered garnet peridotite. Earth Planet. Sci. Lett., 20:347–352. Carswell, D.A. (1986): The metamorphic evolution of Mg-Cr type Norwegian garnet peridotites. Lithos, 19:279–297. Carswell, D.A. & Harvey, M.A. (1985): The intrusive history and tectonometamorphic evolution of the Basal Gneiss Complex in the Moldefjord area, west Norway. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen: Scandinavia and related areas. Chichester: Wiley, 843–857. Carswell, D.A. & Zhang, R.Y. (1999): Petrographic characteristics and metamorphic evolution of ultrahighpressure eclogites in plate-collision belts. Int. Geol. Rev., 41:781–798. Carswell, D.A., Harvey, M.A. & Al-Samman, A. (1983): The petrogenesis of contrasting Fe-Ti and Mg-Cr garnet peridotite types in the high grade gneiss complex of Western Norway: Bull. Minéral., 106:727–750. Carswell, D.A., Krogh, E.J. & Griffin, W.L. (1985): Norwegian orthopyroxene eclogites: calculated equilibrium conditions and petrogenetic implications. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen: Scandinavia and related areas. Chichester: Wiley, 823–841. Carswell, D.A., Brueckner, H.K., Cuthbert, S.J., Mehta, K. & O’Brien. P.J. (2003a): The timing of stabilisation and the exhumation rate for ultra-high pressure rocks in the Western Gneiss Region of Norway. J. Metamorph. Geol., 21:602–612. Carswell, D.A., Tucker, R.D., O’Brien, P.J. & Krogh, T.E. (2003b): Coesite micro-inclusions and the U–Pb age of zircons from the Hareidland eclogite in the Western Gneiss Region of Norway. Lithos, 67:181–190. Coleman, R.G., Wang, X. (1995): Overview of the geology and tectonics of UHPM: In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 1–32. Cuthbert, S.J. (1985): Petrology and tectonic setting of relatively low temperature eclogites and related rocks in the Dalsfjord area. Sunnfjord, west Norway. PhD thesis, Univ. of Sheffield, United Kingdom. Cuthbert, S.J. & Carswell, D.A. (1990): Formation and exhumation of medium-temperature eclogites in the Scandinavian Caledonides: In Carswell, D.A. (ed.): Eclogite facies rocks. Glasgow, London: Blackie, 181–203. Cuthbert, S.J., Harvey, M.A. & Carswell, D.A. (1983): A tectonic model for the metamorphic evolution of the Basal Gneiss Complex, Western South Norway. J. Metamorph. Geol., 1:63–90. Cuthbert, S.J., Carswell, D.A., Krogh Ravna, E.J. & Wain, A. (2000): Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides. Lithos, 52:165–195. Dewey, J.F., Ryan, P.D. & Andersen, T.B. (1993): Orogenic uplift and collapse, crustal thickness, fabrics and metamorphic phase changes: the role of eclogites. In Prichard, H.M., Alabaster, A., Harris, N.B.W. & Neary, C.R. (eds.): Magmatic processes and plate tectonics /Geol. Soc. London, Spec. Publ., 76/, 325–343.

72

D.A. Carswell & S.J. Cuthbert

Dobrzhinetskaya, L., Posukhova, T., Trønnes, R., Korneliussen, A. & Sturt, B. (1993): A microdiamond from eclogite-gneiss area of Norway. 4th International Eclogite Conference, Cosenza, Italy. Terra Nova, Abstract Suppl. 5:9. Dobrzhinetskaya, L., Eide, E.A., Larsen, R.B., Sturt, B.A., Trønnes, R.G., Smith, D.C., Taylor, W.R. & Posukhova, T.V. (1995): Microdiamond in high-grade metamorphic rocks of the Western Gneiss region, Norway. Geology, 23:597–600. Drury, M.R., van Roermund, H.L.M., Carswell, D.A., De Smet, J.H., van den Berg, A.P. & Vlaar, N.J. (2001): Emplacement of deep upper-mantle rocks into cratonic lithosphere by convection and diapiric upwelling. J. Petrol., 42:131–140. Eskola, P. (1921): On the eclogites of Norway. Skr. Vidensk. Selsk. Christiania, Mat.-Naturv. Kl., I, 8:1–118. Gebauer, D., Lappin, M., Grunenfelder, M., & Wyttenbach, A. (1985): The age and origin of some Norwegian eclogites: A U-Pb zircon and REE-study. Chem. Geol., 52:227–247. Gjelsvik, T. (1952): Metamorphosed dolerites in the Gneiss area of Sunnmøre. Nor. Geol. Tidsskr., 30:33–134. Griffin, W.L. & Brueckner, H.K. (1980): Caledonian Sm–Nd ages and a crustal origin for Norwegian eclogites. Nature, 285:319–321. Griffin, W.L. & Brueckner, H.K. (1985): REE, Rb-Sr and Sm–Nd studies of Norwegian eclogites. Chem. Geol. (Isotope Geosci. Sect.), 52:249–271. Griffin, W.L. & Heier, K.S. (1973): Petrological implications of some corona structures. Lithos, 6:315–335. Griffin, W.L. & Qvale, H. (1985): Superferrian eclogites and the crustal origin of garnet peridotites, Almklovdalen, Norway. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen Scandinavia and related areas. Chichester: Wiley, 803–812. Griffin, W.L. & Råheim, A. (1973): Convergent metamorphism of eclogites and dolerites, Kristiansund area, Norway. Lithos, 6:21–40. Griffin, W.L., Austrheim, H., Brastad, K., Bryhni. I., Krill, A.G., Krogh, E.J., Mørk, M.-B.E., Qvale, H. & Torudbakken, B. (1985): High-pressure metamorphism in the Scandinavian Caledonides. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen Scandinavia and related areas. Chichester: Wiley, 783–801. Hacker, B., Root, D., Walsh, E., Young, E. & Mattinson, J. (2001): Recent progress on the Norwegian HP-UHP eclogites. In UHPM Workshop at Waseda Univ., Japan, Abstr., 4B06, 174. Harvey, M.A. (1983): A geochemical and Rb-Sr study of the Proterozoic augen orthogneisses on the Molde Peninsula, west Norway: Lithos, 16:325–338. Jamtveit, B. (1987): Metamorphic evolution of the Eiksunddal eclogite complex, Western Norway, and some tectonic implications. Contrib. Mineral. Petrol., 95:82–99. Jamtveit, B., Carswell, D.A. & Mearns, E.W. (1991): Chronology of the high-pressure metamorphism of Norwegian garnet peridotites/pyroxenites. J. Metamorph. Geol., 9:125–139. Krabbendam, M. & Wain, A. (1997): Late-Caledonian structures, differential retrogression and structural position of (ultra)high-pressure rocks in the Nordfjord-Stadlandet area, Western Gneiss Region. Nor. Geol. Unders., 432:127–139. Krabbendam, M., Wain, A. & Andersen, T.B. (2000): Pre-Caledonian granulite and gabbro enclaves in the Western Gneiss Region, Norway: indications of incomplete transition at high pressure. Geol. Mag., 137:235–255. Krogh, E.J. (1977): Evidence for a continent-continent collision in western Norway. Nature, 267:17–19. Krogh, E.J. & Carswell, D.A. (1995): HP and UHP eclogites and garnet peridotites in the Scandinavian Caledonides. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 244–298. Lappin, M.A. & Smith, D.C. (1978): Mantle-equilibrated orthopyroxene eclogite pods from the basal gneisses in the Selje district, western Norway. J. Petrol., 19:530–584. Lappin, M.A., Pidgeon, R.T. & van Breemen, O. (1979): Geochronology of basal gneisses and mangerite syenites of Stadlandet, west Norway. Nor. Geol. Tidsskr., 59:161–181. Liu, F., Xu, Z., Liou, J.G., Katayama, I., Masago, H., Maruyama, S. & Yang, J. (2002): Ultrahigh-pressure mineral inclusions in zircons from gneissic core samples of the Chinese Continental Scientific Drilling Site in eastern China. Eur. J. Mineral., 14:499–512. Mearns, E.W. (1986): Sm–Nd ages for Norwegian garnet peridotite. Lithos, 19:269–278.

UHPM in the Western Gneiss Region of Norway

73

Medaris, L.G. (1984): A geothermobarometric investigation of garnet peridotites in the Western Gneiss Region of Norway. Contrib. Mineral. Petrol., 87:72–86. Medaris, L.G. & Carswell, D.A. (1990): Petrogenesis of Mg-Cr garnet-peridotites in European metamorphic belts: In Carswell, D.A. (ed.): Eclogite facies rocks. Glasgow, London: Blackie, 260–290. Mørk, M.B.E. (1985a): A gabbro to eclogite transition on Flemsøy, Sunnmore, western Norway. Chem. Geol., 50:283–310. Mørk, M.B.E. (1985b): Incomplete high P–T metamorphic transitions within the Kvamsoy pyroxenite complex, west Norway: a case study of disequilibrium. J. Metamorph. Geol., 3:245–264. Parkinson, C.D. (2000): Coesite inclusions and prograde compositional zonation of garnet in whiteschist of the HP-UHPM Kokchetav massif, Kazakhstan: a record of progressive UHP metamorphism. Lithos, 52:215–233. Smith, D.C. (1980): A tectonic melange of foreign eclogites and ultramafics in west Norway. Nature, 287:366–367. Smith, D.C. (1984): Coesite in clinopyroxene in the Caledonides and its implications for geodynamics. Nature, 310:641–644. Smith, D.C. (1988): A review of the peculiar mineralogy of the “Norwegian coesite-eclogite province” with crystal-chemical, petrological, geochemical and geodynamical notes and an extensive bibliography: In Smith, D.C. (ed.): Eclogites and eclogite-facies rocks. Amsterdam: Elsevier, 1–206. Smith, D.C. (1995): Microcoesites and Microdiamonds in Norway: an overview. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 299–355. Terry, M.P., Robinson, P., Carswell, D.A. & Gasparik, T. (1999): Evidence for a Proterozoic mantle plume and a thermotectonic model for exhumation of garnet peridotites, Western Gneiss Region, Norway. Eos, Trans. Am. Geophys. Union, 89:S359–360. Terry, M.P., Robinson, P., Hamilton, M.A. & Jercinovic, M.J. (2000a): Monazite geochronology of UHP and HP metamorphism, deformation and exhumation, Nordøyane, Western Gneiss Region, Norway. Am. Mineral., 85:1651–1664. Terry, M.P., Robinson, P. & Krogh Ravna, E.J. (2000b): Kyanite eclogite thermometry and evidence for thrusting of UHP over HP metamorphic rocks, Nordøyane, Western Gneiss Region, Norway. Am. Mineral., 85:1637–1650. van Roermund, H.L.M. & Drury, M.R. (1998): Ultra-high pressure (P>6GPa) garnet peridotites in Western Norway: exhumation of mantle rocks from >185 km depth. Terra Nova, 10:295–301. van Roermund, H.L.M., Drury, M.R., Barnhoorn, A. & de Ronde, A. (2000): Super-silicic garnet microstructures from an orogenic garnet peridotite, evidence for an ultra-deep (>6 GPa) origin. J. Metamorph. Geol., 18:135–147. van Roermund, H.L.M., Drury, M.R., Barnhoorn, A. & de Ronde, A. (2001): Relict majoritic garnet microstructures from ultra-deep orogenic peridotites in western Norway. J. Petrol., 42:117–130. van Roermund, H.L.M., Carswell, D.A., Drury, M.R. & Heijboer, T.C. (2002): Microdiamonds in a megacrystic garnet websterite pod from Bardane on the island of Fjørtoft, western Norway: Evidence for diamond formation in mantle rocks during deep continental subduction. Geology, 30:959–962. Wain, A. (1997): New evidence for coesite in eclogite and gneisses: defining an ultrahigh-pressure province in the Western Gneiss Region of Norway. Geology, 25:927–930. Wain, A., Waters, D., Jephcoat, A. & Olijynk, H. (2000): The high-pressure to ultrahigh-pressure eclogite transition in the Western Gneiss Region, Norway. Eur. J. Mineral., 12:667–687. Wain, A.L., Waters, D.J. & Austrheim, H. (2001): Metastability of granulites and processes of eclogitisation in the UHP region of western Norway. J. Metamorph. Geol., 19:609–625. Ye, K., Yao, Y., Katayama, I., Cong, B., Wang, Q. & Maruyama, S. (2000): Large areal extent of ultrahighpressure metamorphism in the Sulu ultrahigh-pressure terrane of East China: new implications from coesite and omphacite inclusions in zircon of granitic gneiss. Lithos, 52:157–164.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 4, 75–103

The Kokchetav massif of Kazakhstan VLADISLAV S. SHATSKY* and NIKOLAI V. SOBOLEV Institute of Mineralogy and Petrography, Siberian Branch of Russian Academy of Sciences, Koptyug Ave. 3, 90–630090 Novosibirsk, Russia; * e-mail: [email protected]

Introduction The recognition of abundant microdiamonds included along with coesite in primary rock-forming minerals and zircons of metamorphic rocks from the Kokchetav massif, Northern Kazakhstan (Sobolev & Shatsky, 1990; Shatsky et al., 1991; Sobolev et al., 1991) indicates that crustal segments reached pressures of the order of at least 40 kbar (4 GPa), implying their subduction to depths greater than 100 km. The Kokchetav massif became internationally recognised as the type locality of diamondiferous metamorphic rocks; the petrological study of diamondiferous ultrahigh pressure (UHP) rocks provides a unique insight into the formation of diamond in crustal rocks at very high pressures. Along with the finding of coesite (Chopin, 1984) the discovery of diamond in supracrustal rocks has drastically changed the current ideas concerning the limits of UHP metamorphism of supracrustal rocks. The specific features and significance of such unique ultrahigh pressure metamorphism have been extensively discussed in different works (e.g. Coleman & Wang, 1995) and in numerous subsequent publications (Ernst & Liou, 2000). However, some workers preferred a hypothesis of a metastable diamond growth at low P–T parameters (Ekimova et al., 1994). The most recent collection of papers devoted specifically to petrotectonic characteristics of the Kokchetav massif is published as a special issue of Island Arc (Liou & Banno, 2000). The significance of metamorphic processes at the origin of a new type of diamond has been extensively discussed (e.g. Haggerty, 1999). It is important to note that the Kumdy-Kol microdiamond deposit covers only a small portion of a more than 200 square km large area in which diamondiferous rocks are distributed in the Kokchetav massif (Shatsky et al., 1991; Dobretsov et al., 1999a). The proven microdiamond reserves of this deposit exceed 3 billion carats (e.g. Haggerty, 1999), making it an absolutely unique phenomenon worldwide. Apart from the Kokchetav massif, occurrences of microdiamonds in other UHP metamorphic terranes elsewhere are less well documented because of the need of bulk extraction and the lack of an unambiguous confirmation of microdiamond in situ (Xu et al., 1992; Dobrzhinetskaya et al., 1995). Another microdiamond locality in gneisses, confirmed by direct observations of thin sections, is from Erzgebirge, Germany, which has been suggested to be similar to the type locality of the Kokchetav massif (Massonne, 1999; Stöckhert et al., 2001).

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

V.S. Shatsky & N.V. Sobolev

76

Geological outline The Kokchetav massif is situated within the Caledonides of the Central Asian fold belt (Fig. 1). This was formed as a single structure at the end of Paleozoic due to collision of the Siberian continent with the North China, Tarim, Tadzhik, Karakorum and Kazakhstan–North Tien Shan massifs (Zonenshain et al., 1990). The Kokchetav massif is commonly considered as a fragment of Kazakhstan–North Tien Shan massif which has been broken up in the Late Riphean–Vendian (Zonenshain et al., 1990). According to Dobretsov et al. (1995), the Kokchetav Massif is a contrasting megamelange composed of slices and blocks of ultrahigh (UHP) and high pressure (HP) (Units I and II), medium pressure (MP) (Barrow-type metamorphism, unit III), and low pressure (LP) rocks. In turn, the UHP and HP rocks are subdivided into two domains (Dobretsov et al., 1998): western (Kumdy-Kol) and eastern (Kulet) (Fig. 2), which are characterised by different P–T conditions of metamorphism and deformation types, suggesting different mechanisms of exhumation in the western and eastern blocks (Theunissen et al., 2000a,b). The ultrahigh pressure Unit I is composed of diamondbearing metasedimentary rocks and gneisses with eclogites in the core of a large antiform covered or substituted by Unit II, an UHP eclogite-bearing mélange with predominant micaschists in the western and/or orthogneisses in the central parts of megamelange. Relatively poor outcrop conditions, limited geochronological information and few metamorphic data that is largely restricted to UHP remnant sequences did not allow to present more precise information on early exhumation processes and associated structural settings.

UHP rocks of the Kumdy-Kol and Barchi locations (Unit I) According to drilling data, the diamondiferous rocks in the western block occur in the form of slices in granite-gneisses, whose thickness is within a few hundreds of meters (Fig. 3). Our studies have established considerable variations in equilibrium P–T parameters of eclogites in units I and II (Shatsky et al., 1989a). It was noticed that the highest temperature eclogites occur at the Kumdy-Kol area (T = 920–1000 °C). Further studies have revealed high temperature eclogites and diamondiferous metamorphic rocks near Pakhar village (17 km northwest of the Kumdy-Kol) (Shatsky et al., 1991) and Lake Barchi (15 km west from Kumdy-Kol) (Korsakov et al., 1998). All these types belong to unit I of the Kumdy-Kol romb-horst domain (Theunissen et al., 2000b). Microdiamonds have been found only in metasediments of unit I: garnet-biotite gneisses and schists, garnet-muscovite-kyanite schists, garnet-pyroxene rocks and dolomitic marbles. It should be emphasised that eclogites do not contain diamonds (Shatsky & Sobolev, 1993). Eclogites Eclogites occur within kyanite-mica schists, plagiogneisses and diamondiferous metasediments. The eclogites are medium-grained, pinkish to green rocks of granoblastic texture consisting of garnet (40–50%), pyroxene (30–40%), quartz (5–10%). The secondary

Fig. 1. Major tectonic features of Central Asia for the region surrounding and including the Kokchetav massif, with the location of Figure 2 shown (Dobretsov et al., 1999a).

The Kokchetav massif of Kazakhstan 77

Fig. 2. The Kokchetav Megamelange with tectonic units 1–3 and adjacent domains I–V (Dobretsov et al., 1998). 1: ultrahigh pressure/HP unit with high temperature eclogites and diamond-bearing rocks (Kumdy-Kol) and with whiteschist, coesite-bearing micaschist and relatively low temperature eclogites (Kulet); 2: medium pressure unit with Al-rich metasediments and coronite (Enbek-Berlyk); 3: low pressure unit (Daulet). Domain I: Neoproterozoic sequences mainly belonging to the Vendian–Early Cambrian island arc; Domain II: “Kokchetav Microcontinent Domain” is composed of several blocks. In various amounts each of this blocks includes: a) a gneissic basement, b) its sedimentary cover (black shale and dolomite in the lower part and metasandstone in the upper part, c) tectonic slices of Vendian–Early Cambrian (?) ophiolite and volcanics, d) Ordovician rocks and e) Devonian granites; Domain III: the “Megamelange Domain”; Domain IV: the “White Lake Domain” – amphibolites, amphibole schist and quartzites, assumed to represent fragments of an oceanic crust; Domain V: the “Granite Dome Domain” is polyphase and composed of the Imantau and Zerenda granite domes with Cambrian gabbro, Late Ordovician–Silurian diorite, granodiorite and granite, and Devonian leucocratic granite and granosyenite.

78 V.S. Shatsky & N.V. Sobolev

Fig. 3. Detailed geological map of diamond-bearing metasedimentary rocks mined, drilled and trenched (Scheshkel et al., pers. comun.). 1: granite-gneisses; 2: biotite gneisses; 3: granite-gneisses and gneisses alternation; 4: fine grained chlorite-tremolite-quartz rocks; 5: migmatites; 6: garnet-muscovite, kyanite-muscovite schists; 7: pyroxene-carbonate rocks; 8: garnet pyroxenites; 9: eclogites and amphibolites; 10: dykes of diorite porphyrites.

The Kokchetav massif of Kazakhstan 79

80

V.S. Shatsky & N.V. Sobolev

minerals are represented by amphibole (10–20%), biotite (3–5%), plagioclase (4–2%). Rutile occurs as an accessory mineral. The replacement of omphacite by plagioclasepyroxene symplectite is one of the peculiarities of eclogite. In some samples, omphacite is completely transformed into pyroxene-plagioclase symplectite. In the host rocks the following mineral associations are present: qzt + (chl) + gr, qtz + gr + bi + ph, qtz + amph + zo + ep + (chl) + gr, qtz + amph + ort + cpx + tit, qtz + fsp + bi + gr + cpx. These associations formally indicate a mixture of amphibolite, granulite, epidote–amphibolite and greenschist facies in the host rocks. But relics of UHP minerals show that the polystage history of these rocks is similar to eclogite history. Biotite gneisses and garnet-pyroxene-quartz rocks Biotite gneisses are the most common diamondiferous rocks among the metasediments (Fig. 3). Lenses and bands of calc-silicate metacarbonate rocks, garnet-pyroxene rocks, relics of cataclastic massive garnet-pyroxene-quartz and garnet-quartz rocks, and finegrained chlorite-actinolite-quartz rocks occur within biotite gneisses (Shatsky et al., 1995; Shatsky & Sobolev, 1993). Cataclastic garnet-biotite gneisses contain relics of ultrahigh pressure garnets (Shatsky et al., 1995) as well as biotite, plagioclase, potassic feldspar, muscovite and amphibole. All varieties of transitional rocks from massive garnet-pyroxene-quartz rocks through banded garnet-biotite gneisses and chlorite schists occur in both sites. Banding at both large (0.5–1 m) and small scales (0.5–1 cm) is observed in diamondiferous rocks, particularly in garnet-biotite gneisses. This banding is caused by alternating layers composed mainly of varying amounts of garnet, biotite, pyroxene, quartz and plagioclase. Clinopyroxene found in some types of biotite gneisses and garnet-pyroxene-quartz rocks could be referred to the earlier high pressure paragenesis. However, the jadeitic component in these types of rocks is lower in comparison with pyroxene from inclusions in garnet and zircon. We found that omphacite included in zircons from cataclastic garnet-biotite gneiss contains up to 50
jadeite component. It should be mentioned that clinopyroxene with 74
jadeite was identified as inclusion in zircon from pelitic gneisses by Katayama et al. (1998). This can indicate that the jadeitic component in pyroxenes of the matrix is reduced as a result of the retrograde metamorphism under the conditions of the amphibolite facies (Shatsky & Sobolev, 1985). Garnets from these types of rocks vary greatly in their composition (Figs. 4 and 5). The fields of the non-diamondiferous and diamondiferous rocks partly overlap. Garnets from eclogites plot in the field of diamondiferous rocks. The contents of Ca and Fe sometimes increase from the core to the rim when zoning is present, but in most cases garnets from diamondiferous rocks do not show zoning. Biotite intergrowth with diamonds in garnets from garnet-biotite gneisses and garnet-pyroxene-quartz cataclastic rocks has a lower Fe/(Fe + Mg) compared with biotite in the host rocks (Sobolev & Shatsky, 1990; Vavilov et al., 1991). Si contents in phengites included in zircon and garnets vary from 3.22 to 3.56 p.f.u. (Shatsky et al., 1995; Korsakov et al., 2002). Phengite from matrix, as usual, contain less celadonite component than that included in zircon (3.2–3.32 Si p.f.u.).

The Kokchetav massif of Kazakhstan

81

Fig. 4. Compositional variations of garnet from the metamorphic rocks of Kumdy-Kol. 1: diamondiferous gneiss, garnet-pyroxene-quartz rocks and schist; 2: garnet-pyroxene rocks; 3: dolomitic marbles; 4: diamondfree gneiss; 5: eclogite; 6: inclusions in zircons from dolomitic marble (K9-16); 7: garnet from matrix (K9-16); 8: diamondiferous garnet-kyanite-muscovite schist.

Fig. 5. Compositional variation of garnet from zircons. 1: dolomitic marble; 2: biotite plagiogneiss; 3: garnetpyroxene rocks; 4: pyroxene-biotite gneiss; 5: migmatites; 6: garnet from matrix of all rock types.

82

V.S. Shatsky & N.V. Sobolev

Coesite inclusions have been found in zircon grains in more than thirty samples of diamondiferous metapelites, including biotite-kyanite-zoisite gneisses and schists (Sobolev et al., 1998; Korsakov et al., 2002). These finds were confirmed by subsequent studies by Katayama et al. (2000). One bimineralic inclusion representing intergrowth of coesite and diamond was documented (Sobolev et al., 1994). Coesite was also found as inclusions in zircon grains from a high temperature eclogite from the drilling hole near Lake Barchi (Korsakov et al., 1998). In addition to zircon inclusions, single crystals of coesite occur also in garnets from three samples of zoisite gneisses of the Lake Barchi site (Korsakov et al., 2002). In all these samples, the presence of coesite was confirmed by Raman spectroscopy. The presence of coesite inclusions in garnets from zoisite gneisses is an unusual feature compared with typical relicts of coesite described from elsewhere. Metasomatic rocks Very fine-grained rocks with granoblastic structure are treated as a separate group. Great variations in the ratios of the constituent minerals are observed with dominant quartz, amphibole and chlorite. These minerals are accompanied by biotite, muscovite, zoisite and tourmaline. Some garnet grains are replaced by amphibole, chlorite and mica. Amphibole is replaced by chlorite and biotite. Zoisite forms large grains with chlorite and amphibole inclusions. Some areas in these rocks are composed of large grains of K-feldspar. Garnet-pyroxene rocks Diamondiferous and diamond-free garnet-pyroxene rocks consist mainly of garnet and clinopyroxene with a small amount of carbonate and K-feldspar. Amphibole and chlorite occur as retrograde minerals. Garnet-pyroxene rocks can be subdivided into two groups according to the composition of garnet and clinopyroxene. The first one is made up of Mg-rich clinopyroxene (12–17% MgO) and grossular-pyrope-almandine garnet (Figs. 4 and 5). High K2O and a lack of or low Na2O distinguishes the pyroxenes in this type of rock. Orthoclase lamellae (1–20 µm) and silica needles are also found in the pyroxene grains in the matrix and sometimes in clinopyroxene included in garnet (Shatsky et al., 1985, 1995; Katayama et al., 2000). When exsolution structures are absent, pyroxenes in the groundmass and those included in garnets have the same composition. Lamellae of highly aluminous titanite were also found in clinopyroxene from these rocks. Garnetpyroxene rocks of this type generally contain diamonds but samples without diamonds are also available. Microdiamonds are widely scattered as inclusions in garnet, clinopyroxene and zircon. Taking into account that diamonds are unevenly distributed within samples and even within thin sections, in specific cases it is difficult make conclusions about diamond occurrences within certain samples. In contrast, garnetpyroxene rocks of the second type never contain diamonds. These rocks are made up of grossular-almandine garnet and clinopyroxene enriched in iron. The most characteristic features are the absence of K impurities in this clinopyroxene.

The Kokchetav massif of Kazakhstan

83

Dolomitic marbles Marbles containing a variable proportion of dolomite, Mg-calcite, diopsidic pyroxene and garnet are typical of the Kumdy-Kol microdiamond deposit. As previously, two types of marbles can be recognised based on a composition of clinopyroxene. Clinopyroxene from the first type of marbles similar to clinopyroxene from the first type of pyroxenegarnet rocks, containing lamellae of K-feldspar and quartz needles. K-rich clinopyroxene is also found in the matrix. These types of rocks contain diamonds. Garnets from dolomites differ significantly in their composition from the garnets of other rock types (Figs. 4 and 5). They are notable for their high calcium content and high Mg/(Mg + Fe) ratio. Compositional profiles along the grains of garnet in most cases show that the central part is not zoned, but a change in the chemical composition occurs in the 100–125 µm wide rim (Shatsky et al., 1995). It should be noted that when zoning is absent in garnets, their composition can vary within a thin section. Strong compositional variation was found for garnets included in zircons from single samples of marbles. Calcium contents vary from 27
to 49
, and the Fe/(Fe + Mg) ratio ranges from 0.437 to 0.336. At the same time, in the garnet from the matrix the compositional variations are insignificant (Shatsky et al., 1995). New types of lamellae have been observed in pyroxene from dolomitic marbles (Shatsky & Sobolev, 2001). One of the samples (94-275) of such a rock contains pyroxene grains with garnet lamellae along with grains containing K-feldspar lamellae and no lamellae at all (Table 1). Compositions of garnet lamellae are very similar to garnet rims in matrix crystals (Grs 31.8–33.2; Prp 38.3–45.7 and Alm 19.5–15.9). Prograde zonation is typical of garnets from the matrix. Grossular and pyrope components increase from core (Grs 32, Prp 29) to rim (Grs 36.5, Prp 37.9) with decreasing almandine content (from 35.6 to 18.5). All Cpx both with and without lamellae are diopside with low FeO (2 wt%), Al2O3 (1.32–1.89 wt%) and Na2O (0.11–0.16 wt%). No detectable K2O was found. Another type of lamellae was found in clinopyroxene from sample 98-6 (Table 2). Cpx from this sample contain lamellae of phengite (3.42 Si p.f.u.) and K-feldspar. K contents in Cpx varies from 0.05 to 0.54 wt%. On the basis of textural grounds it is concluded that topotaxial growth of mica lamellae in Cpx are best explained by exsolution from a former K and OH-bearing Cpx stabilised at UHP conditions. Dolomite, magnesite and magnesian calcite were found among the inclusions in zircon. MgCO3 content in magnesium calcite is about 23.5
(Shatsky et al., 1995). Inclusions of calcite and dolomite are found in garnet. Calcite inclusions predominate. The Mg admixture is minor in matrix calcite and calcite inclusions in garnets. Radial cracks are often observed around the calcite inclusions in garnet. Microdiamonds Microdiamonds are widely scattered as inclusions in garnet, zircon, and rare in clinopyroxene, kyanite and zoisite (Sobolev & Shatsky, 1987, 1990; Korsakov et al., 2002; Ogasawara et al., 2000) They are also very typical in rounded pseudomorphs after garnet, consisting of mica + chlorite + amphibole aggregates and tourmaline (Sobolev &

V.S. Shatsky & N.V. Sobolev

84

Table 1. Representative analyses of minerals from pyroxene-carbonate rock 94–275 Name

Cpx1

Lam1

Lam2

Gt1 core

rim

Gt2

Inc

Inc2

Cpx2

SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O

53.90 0.22 1.86 0.02 2.06 0.05 16.40 24.90 0.12 0.00

40.30 0.28 21.50 0.01 10.30 0.67 10.10 15.40 0.00 0.00

40.30 0.22 21.60 0.04 8.55 0.54 12.10 13.70 0.00 0.00

40.30 0.36 21.10 0.02 10.20 0.59 9.98 15.60 0.00 0.00

40.70 0.28 21.40 0.02 10.40 0.58 10.10 16.00 0.00 0.01

41.00 0.33 21.40 0.02 10.90 0.55 9.61 15.90 0.01 0.01

53.80 0.23 1.86 0.00 2.41 0.03 16.60 24.80 0.17 0.00

38.60 0.73 17.30 0.00 4.32 0.01 22.60 0.04 0.06 9.35

53.20 0.22 1.76 0.02 2.03 0.05 16.00 25.00 0.12 0.00

Total

99.48

98.50

97.00

98.12

99.43

99.64

99.89

93.06

98.37

Si Ti Al Cr Fe2+ Fe3+ Mn Mg Ca Na K Fe tot

1.967 0.006 0.080 0.000 0.063 0.000 0.001 0.890 0.973 0.008 0.000 0.063

3.015 0.016 1.895 0.001 0.566 0.082 0.043 1.129 1.234 0.000 0.000 0.650

3.022 0.012 1.913 0.002 0.470 0.067 0.034 1.351 1.102 0.000 0.000 0.538

3.025 0.020 1.869 0.001 0.538 0.101 0.038 1.117 1.258 0.000 0.000 0.642

3.023 0.015 1.872 0.001 0.540 0.103 0.036 1.112 1.270 0.000 0.000 0.646

3.037 0.018 1.866 0.001 0.569 0.106 0.035 1.062 1.264 0.001 0.000 0.678

1.959 0.006 0.080 0.000 0.068 0.006 0.001 0.899 0.968 0.012 0.000 0.073

2.789 0.040 1.475 0.000

0.001 2.429 0.003 0.009 0.861 0.261

1.967 0.006 0.077 0.000 0.063 0.000 0.002 0.882 0.988 0.009 0.000 0.063

Total

3.990

7.980

7.974

7.968

7.973

7.959

3.997

7.867

3.993

6.60

36.55

28.49

36.50

36.73

38.98

7.55

9.70

6.64

f

Shatsky, 1990; Shatsky et al., 1995). Diamonds were not detected in intergranular space. As usual, diamonds coexist with graphite. Each analysed diamond-bearing rock type from the Kokchetav massif, (garnetclinopyroxene rocks, marbles, biotite gneisses and zoisite gneisses) contains microdiamonds with a distinctive range of morphologies (Shatsky et al., 1989b, 1998a, 1999b). Diamonds occurring in garnet-clinopyroxene calc-silicate rocks and marbles are predominantly of cuboid morphology. Biotite gneisses are characterised by a cubooctahedral diamond population, although cuboids and growth forms transitional to cuboids also exist. Octahedral crystals are only observed in zoisite gneisses. Some zoisite gneisses carry octahedral diamonds, whereas others typically contain cuboids. Detailed petrological investigation of zoisite gneiss samples demonstrates that octahedral crystals appear only in rocks which do not contain symplectitic zoisite (Korsakov et al., 2002). The greatest variety in diamond morphology within a single rock type is observed in biotite gneisses. In this rock type, the predominant crystal habit is the cubooctahedron. Intergrowths and aggregates are quite abundant. Shatsky et al. (1995) observed an octahedron and a cuboid in a single intergrowth. In addition to cuboidal and

The Kokchetav massif of Kazakhstan

85

Table 2. Representative analyses of minerals from pyroxene-carbonate rock K 98-6 Name

Grt

Inc

Cpx1 core rim

Lam1

Lam2

Cpx2

Mica

54.00 0.00 0.84 0.00 3.12 0.10 16.40 24.50 0.17 0.54

51.10 1.01 28.29 0.00 1.08 0.02 3.91 0.24 0.04 5.92

63.50 0.00 16.80 0.00 0.11 0.00 0.04 0.20 0.07 14.8

54.70 0.00 1.84 0.00 2.00 0.07 16.49 24.87 0.41 0.08

39.20 1.18 15.00 0.00 10.40 0.30 18.50 0.35 0.06 8.69

100.04

99.70

91.59

95.42

100.46

93.68

1.969 0.000 0.090 0.000 0.088 0.000 0.004 0.864 0.948 0.025 0.017 0.088

1.960 0.000 0.070 0.000 0.066 0.025 0.004 0.880 0.975 0.024 0.002 0.092

1.982 0.000 0.036 0.000 0.087 0.009 0.003 0.896 0.963 0.012 0.025 0.096

3.420 0.051 2.229 0.000

3.051 0.000 0.949 0.000

2.888 0.065 1.307 0.000

0.001 0.390 0.017 0.005 0.505 0.060

0.000 0.003 0.010 0.006 0.907 0.004

1.894 0.024 1.114 0.000 0.034 0.000 0.000 0.260 0.068 0.003 0.258 0.034

7.999

4.006

4.005

4.014

6.674

4.931

3.656

7.806

50.77

9.27

9.47

9.65

13.42

60.85

11.62

24.04

core

rim

SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O

39.30 0.40 20.80 0.02 6.55 0.99 3.37 27.60 0.05 0.01

39.60 0.38 21.30 0.01 7.00 1.08 3.82 26.30 0.04 0.02

54.10 0.00 2.11 0.00 2.90 0.14 15.90 24.30 0.36 0.36

53.80 0.00 1.62 0.00 3.01 0.12 16.20 25.00 0.34 0.05

Total

99.04

99.62

100.27

Si Ti Al Cr Fe2+ Fe3+ Mn Mg Ca Na K Fe tot

2.992 0.023 1.866 0.001 0.316 0.101 0.064 0.382 2.251 0.008 0.000 0.419

2.996 0.022 1.899 0.000 0.370 0.073 0.069 0.430 2.134 0.006 0.001 0.444

Total

8.004

f

52.27

0.019 2.030 0.028 0.009 0.817 0.642

Note: Cpx1 - clinopyroxene contains lamellae of phengite

octahedral faces, some crystals have {110} faces. They consist of alternating small {111} steps and are believed to reflect a growth form. Coated diamonds also occur. The predominant diamond type in garnet-pyroxene rocks and dolomitic marbles is the cuboid. Crystal surfaces display a large number of pits of variable configuration. Octahedral faces may be found at the edges of the cuboids. Diamonds from marbles, more likely than those of garnet-clinopyroxene rocks, often show octahedral microfaces on their cuboidal surfaces, indicating a change in the environmental conditions during diamond growth. Zoisite gneisses devoid of symplectitic zoisite occur in the Barchi area. They include dominantly octahedral diamonds (more than 80%). Diamond size varies from 10 to 50 µm. In one rock sample, sharp-edged octahedra with octahedra displaying antiskeletal growth features are present. Twins are fairly frequent. In addition to octahedra, cuboids also occur. In general, in a single sample of zoisite gneiss devoid of symplectitic zoisite, all transition forms between cuboids and octahedra can be found. The occurrence of octahedral microfaces on cuboidal surfaces is also noted, but here this process ends with the formation of octahedra.

86

V.S. Shatsky & N.V. Sobolev

Cuboids from garnet-pyroxene rocks, as well as one octahedron from zoisite gneiss, have been investigated for their internal morphology using X-ray topography (Shatsky et al., 1998a). Cuboids from garnet-pyroxene rocks show a radiated structure resulting from space-filling columnar growth, also known as “fibrous growth”. An interesting feature of the inner structure of some crystals is the presence of a core, expressed as a slight blackening on the topograms. The octahedral diamond from zoisite gneiss has predominantly cubic growth sectors and a fibrous structure. Despite the small size of this crystal (30 µm), its central zone is well marked on the X-ray topography picture. This may suggest that either the central zone has no defects, or that it is disoriented relative to the main part of crystal. X-ray topography data for the octahedron support the conclusion, based on external morphology, that octahedral microdiamonds are in fact reshaped cuboids. Diamonds from metamorphic rocks display all morphological types found for diamonds from kimberlites, lamproites, and alluvial deposits. The basic difference is however the relative abundance of the different diamond morphologies. In particular, diamonds from metamorphic rocks are dominated by cuboids, whereas in kimberlites and lamproites octahedra dominate. Moreover, the octahedra from metamorphic rocks initially grew as cuboids under conditions of high supersaturation and only at their final growth stages, when supersaturation decreased, did they acquire an octahedral shape. This is in contrast to octahedra from kimberlites and lamproites that in general grew by spiral or layer-by-layer growth mechanism at low degrees of carbon supersaturation (Sunagawa, 1984). The mean 13C and 15N values of diamonds from garnet-clinopyroxene rock are –10.5 and +5.9‰, respectively. Diamonds from dolomitic marble have a mean 13C value of –10.2 and a 15N of +8.5‰ (Cartigny et al., 2001a). The isotopic values of nitrogen and carbon rather reflect a crustal origin. For mantle diamonds there is a strict relationship between the 13C value and total nitrogen content (Cartigny et al., 2001b). The higher the nitrogen content, the higher the 13C value. When the in situ diamonds with 13C –10.5 and –10.2‰ are formed of mantle carbon, their maximum nitrogen content should be less than 1750 ppm. This is not confirmed by the nitrogen data obtained by infrared spectroscopy or combustion. Diamonds from garnet-clinopyroxene rocks are rich in nitrogen. Up to 2765 ppm of nitrogen is incorporated in their lattice (de Corte et al., 1998) and, moreover, up to 7000 ppm is present additionally as fluid inclusions of molecular nitrogen (Cartigny et al., 2001a). Diamonds from dolomitic marbles have a total nitrogen content of about 2000 ppm, of which maximum 1000 ppm is in the diamond structure and the remainder in fluid inclusions (Cartigny et al., 2001a). Therefore, the carbon source for the Kokchetav diamonds is interpreted as a mixture of crustal carbonates (13C ~ 0‰) and organic matter from carbonaceous metasediments (13C  –25‰). The similar peculiarities were earlier observed for some kimberlitic and alluvial diamonds as a possible proof of a crustal origin of their carbon (Sobolev & Sobolev, 1980). The nitrogen isotopic composition of Kokchetav diamonds reflects the moderately to highly positive 15N values of metamorphosed sedimentary rocks. Diamonds from garnet-clinopyroxene rocks, marbles and zoisite gneisses contain water and carbonate inclusions (de Corte et al., 1998, 1999). They result from entrapment

The Kokchetav massif of Kazakhstan

87

of fluid during the rapid crystallisation of the diamond fibres and hence provide evidence of diamond growth from a C-O-H fluid. A great variety in the morphology of diamond crystals even in polycrystalline aggregates (Sobolev & Shatsky, 1990) may be due to variation in the degree of carbon supersaturation in the fluid phase as a consequence of contemporaneous melt generation. The presence of melt would vary the fluid phase composition. With free carbon in the rocks, and at constant oxygen fugacity, the fluid phase composition would be a function of the reaction of carbon and water: CO2 + CH4 = 2C + 2H2O, CH4 + 2CO = 3C + 2H2O. When melt is present, the reactions given above would be shifted toward the right side resulting in precipitation of C from fluid phase. The reason is that the solubility of water in the melt is much higher compared with CO2 and CH4. Thus the appearance of a melt phase could lead to a mass crystallisation of diamonds or graphite.

The UHP and HP rocks of Unit II In Unit II lower-temperature eclogites occur among phengite-kyanite-quartz and muscovite-garnet-quartz-plagioclase schists (Kulet, Chaglinka, Sulu-Tyube) and among mylonitic granitogneisses (Enbek-Berlyk, Lake Uyaly) (Shatsky et al., 1989a). Coesite inclusions have been found only in garnet from phengite schist from the Kulet area. Sulu-Tyube area The eclogites occur as several large bodies outcropping within the mica schists and gneisses of the second unit (Fig. 6). The largest (1.5×0.5×0.5 km) body of the Sulu-Tyube Hill represents a lens deformed into a synform. Strongly zoned garnet porphyroblasts, whose Fe/(Fe + Mg) ratio varies from 0.85 to 0.57, contain quartz, rutile, titanite, zoisite and epidote inclusions. Zoisite and amphibole are in textural equilibrium with omphacite. During the detailed investigation the Sulu-Tyube eclogites were found to have a heterogeneous structure. Varieties containing porphyroblasts of garnet (up to 7 mm) are interbeds of medium-grained eclogite with garnet grains up to 1 mm. The eclogite modal composition is: garnet (25–30%), omphacite (20–25%), quartz (5–10%), amphibole (20–25%), zoisite (5–10%). Among the secondary minerals there are plagioclase, chlorite, muscovite, epidote and calcite. These rocks are massive and porphyroblastic. Drilling of the eclogite body revealed that the composition of eclogites does not change up to a depth of 550 m (A. Zayachkovsky, pers. comm.). Besides the large body of the Sulu-Tyube region, there are numerous outcrops of fine and medium-grained eclogites, forming separate boudins tracing the outline of original layers. The host rocks include both the white mica (“silver”) schists and medium-grained gneisses of dark brown colour, composed of biotite, quartz, plagioclase, muscovite, potassium feldspar and sometimes kyanite.

Fig. 6. Geological map of the Sulu-Tyube area based on geophysical, drilling and structural data (compiled by Dobretsov & Zayachovsky; Dobretsov et al., 1999b). 1: traced boundaries of eclogites; 2: blocks of fresh eclogites corresponding to highest density geophysical bodies; 3: the LP Daulet unit; 4: small eclogite bodies; 5: structural lines; 6: faults visible (a) and supposed under quaternary sediment (b).

88 V.S. Shatsky & N.V. Sobolev

The Kokchetav massif of Kazakhstan

89

Enbek-Berlyk area Amphibolised eclogite and gabbroid bodies occur in this area. According to some researchers (Dobretsov & Sobolev, 1970; Udovkina, 1985), eclogitisation reactions can be observed in gabbro. Because of these observations a detailed investigation of metabasites at this site was made. Garnet-biotite-sillimanite-kyanite schists and metagabbro-garnet amphibolite boudins are characteristic for the upper unit (Fig. 7). The lower unit contains retrograde eclogite and coarse-grained gneiss inclusions among garnet mica schists with transition to sheared orthogneisses with boudinaged metadolerite dykes. The gabbro-norites (hyperstene granulites) are medium-grained, dark grey rocks, consisting of plagioclase (25–30%), clinopyroxene (30–40%), orthopyroxene (10–15%) and ilmenite (2–3%). Secondary amphibole and biotite are found. The texture is ophitic, poikiloblastic in some places. The structure is massive. At the margins of gabbro-norite boudins, garnet amphibolites are formed: garnet (25–30%), newly formed pyroxene (10–20%), amphibolite (20–30%), quartz (5–10%) and plagioclase (3–5%). Rutile, ilmenite and titanite are presented as accessory minerals, the secondary minerals are zoisite and biotite. The texture is granoblastic, the structure is massive. Between these two rock types, transitional varieties are observed. It has been mentioned by earlier investigators that the corona gabbro in the marginal parts of the bodies is interpreted as the result of eclogitisation of gabbroid rocks (Dobretsov & Sobolev, 1970). Our investigations do not confirm this conclusion, since the newly formed pyroxene has low omphacite content and the metamorphic assemblage is represented by sodic augite, amphibole and plagioclase. Transition to garnet amphibolite involves the following stages. Corona stage: rims of garnet at the contact between plagioclase and primary pyroxene. Second stage: growth of garnet grains is observed within plagioclase. Simultaneously primary plagioclase is replaced by a very fine-grained quartz-plagioclase aggregate, the new plagioclase is more sodic (An15–20). Newly formed pyroxene replacing augite has low jadeite content (5.5% Ja). The eclogites of the Enbek-Berlyk area, as a rule, are intensely altered. Garnets (25–30%), omphacite (15–17%) and quartz (5–10%) are the primary minerals. Plagioclase (15–17%), amphibole (10–20%), zoisite, biotite and epidote also are present as secondary minerals. Accessory minerals are rutile and ilmenite. Pyroxenes from eclogite are almost completely replaced by pyroxene-plagioclase symplectite or by amphibole. The analyses of the fresh grains show that pyroxene contains 21% Jd. In symplectite pyroxene the jadeitic component decreases down to 5%. The value of Fe/(Fe + Mg) ratios of garnets from amphibolites and corona gabbro ranges from 0.80 to 0.94, while that of the eclogite garnets ranges from 0.60 to 0.80. The garnets in all metabasites are zoned. The decrease of Mg content and the increase of Fe and Mn from the core to the rim are observed in eclogites. The garnets from amphibolite have the same zoning patterns. This zoning is regressive. The garnets from corona gabbro are either non-zoned, or their Mg content increases and Ca content decreases from the core to the rim. It should be noted that the composition of garnets from the rims, surrounding pyroxene and the garnets growing inside plagioclase grains remains the same.

Fig. 7. Structural geological sketch map of the Enbek-Berlik area. 1: amphibolised dolerite; 2: amphibolite; 3: eclogite and amphibolised eclogite; 4: pyroxenite and harzburgite; 5: coronite; 6: diopside-plagioclase rock; 7: metapelites and the strike of their schistosity; 8: granite; 9: boundary of recent deposits; 10: fault; 11: dip of bedding (Reverdatto, 1999).

90 V.S. Shatsky & N.V. Sobolev

The Kokchetav massif of Kazakhstan

91

The metasedimentary schists are characterised by garnet porphyroblasts in a finegrained matrix of biotite (20–25%), quartz (15–20%), kyanite (10–15%), and muscovite (5%). Kyanite is replaced by sillimanite. Kulet area Bodies of fine and medium-grained eclogites occur among garnet-muscovite-kyanitequartz and muscovite-garnet-quartz-plagioclase schists (Fig. 8). This structure possibly includes several tectonic sheets similar to the Enbek-Berlyk area. Two eclogite varieties are interlayered on the scale of 10 to 160 metres. They are dark green and light pink fine-grained eclogites, with predominant omphacite and garnet, respectively. The main minerals are pyroxene, garnet, quartz, porphyroblastic zoisite, pale brown amphibole and rare phengite. Some areas of thin sections consist of larger foliated omphacite grains. The increase in omphacite grain size is observed near zoisite grains and quartz veins, where idiomorphic pyroxene forms. The country rocks are represented mainly by garnet-mica schists, with interbeds of talc-kyanite-garnet schists (Udovkina, 1985). Garnet and pyroxene are replaced by dark green amphibole at the eclogite boudin margins. In some cases eclogite transforms into amphibolite. Talcgarnet-kyanite rocks are intercalated with amphibole-garnet-kyanite assemblages. The other rocks present are: amphibole-garnet-zoisite, garnet-biotite-kyanite-amphibole and garnet-zoisite-kyanite rocks. A recent discovery of coesite relics in garnet from phengite schist near Lake Kulet is of particular significance (Shatsky et al., 1998b). The rocks are composed of garnet, phengite and quartz. Moreover, there are relics of kyanite after which white mica develops. Phengite in turn is replaced by biotite in the rims. Rutile and tourmaline occur as accessory minerals. Two generations of garnet are present there. The first generation garnet forms large subhedral grains often containing abundant inclusions of quartz and rutile. The second generation garnet occurs as 50–150 µm long grains. Large garnet grains display zoning, which is expressed as an increase in Fe and Mg and a decrease in Ca and Mn contents from core to rim. The Fe/Mg value decreases from 9.5 to 7.9. The second generation garnets exhibit weak zoning. In composition they are similar to the rims of large garnet porphyroblasts, though their Fe/Mg value is still lower (to 5.9). Phengite has a high content of celadonite component (3.43 p.f.u. Si). The Si content in muscovite in a previously studied mica schist from the Kulet area, however, was no more than 3.2 p.f.u., whereas phengite was found only in eclogites (3.46 p.f.u. Si). Quartz occurs in garnet both as polycrystalline inclusions and single crystals. Inclusions are often surrounded by radial cracks. Coesite inside a polycrystalline quartz aggregate was found in only one inclusion. Coesite was identified both by petrography and by the characteristic Raman spectral band at 520 cm–1. Whiteschists rich in talc and Mg-rich garnet are usual in the Kulet area. Phengite can occur instead of talc.

P–T path The results of thermobarometry are summarised in Table 3. One must keep in mind that diamondiferous rocks have minerals formed at least during three metamorphic stages

Fig. 8. Detailed geological map of the central part of the South Zheltau (Kulet) area (Dobretsov et al., 1999b).

92 V.S. Shatsky & N.V. Sobolev

The Kokchetav massif of Kazakhstan

93

Table 3. P–T conditions of western and eastern domains of the Kokchetav Massif Metamorphic stage

T [°C]

P [kbar]

Western domain Stage I

950–1000

>40

Stage II Stage III

>800 800– 820

25 10–12

Stage IV

650–700

7–8

Stage V

<350

Eastern domain Stage I Stage II

650–800 600–700

27 7–10

References Sobolev & Shatsky (1990); Shatsky et al. (1995); Zhang et al. (1997); Hermann et al. (2001); Katayama et al. (2001); de Corte et al. (2001) Ogasawara et al. (2000) Shatsky & Sobolev (1985); Hermann et al. (2001); Sobolev et al. (2001); Katayama et al. (2001) Shatsky et al. (1993); Zhang et al. (1997); Hermann et al. (2001) Shatsky et al. (1993) Shatsky et al. (1993, 1998b) Ota et al. (2000)

(UHP, amphibolite, and greenschist facies). We are using only the minerals included within garnet porhyroblasts and zircon for the estimation of P–T parameters of the UHP stage related to diamond formation. We have shown previously that eclogites of the Kumdy-Kol area have equilibrium temperature of 800–950 °C at P minimum 18 kbar (Shatsky & Sobolev, 1985). The study of diamondiferous rocks gives us evidence that the pressure of metamorphism has been significantly higher. The greatest difficulties arise in the estimation of pressure. Earlier we stressed that we can estimate only the lower limit of pressure, using the graphite–diamond equilibrium line (Sobolev & Shatsky, 1990; Shatsky et al., 1995). To substantiate higher pressures (70 kbar), the Japanese researchers (Kaneko et al., 2000) use experimental results by Luth (1997) on K impurity in pyroxene. But as shown by Harlow (1999), the partition coefficient of K between pyroxene and melt depends not only on pressure but also on melt composition. Therefore, at present there is no direct possibility to estimate pressures from K content in pyroxene. At the same time, inclusions of magnesite, along with calcite and dolomite, found in zircons from the carbonate rock, can be evidence of pressures of more than 70 kbar (Sato & Katsura, 2001). The relationships of these phases remain, however, to be understood. Provided that zircons grow at both progressive and regressive stages of metamorphism (Hermann et al., 2001; Katayama et al., 2001), another reaction can be responsible for the appearance of magnesite, for example: En + Dol = Di + Mgs. Fe–Mg partitioning between garnet and clinopyroxene, garnet and biotite (Hodges & Spear, 1981) or garnet and phengite (Green & Hellman, 1982) was used to calculate the temperatures at pressure 40 kbar for diamondiferous rocks and eclogites of Unit I. The calibration of Powell (1985) and Ellis & Green (1979) yields temperatures within the interval 760–1050 °C. For the majority of samples the estimation exceeds 950 °C. The lowest temperatures are obtained for gneisses (760–850 °C). As mentioned above, we suppose that pyroxene from the matrix recrystallised during the retrograde path. The

94

V.S. Shatsky & N.V. Sobolev

equilibrium temperature obtained by using inclusions of omphacite in zircon is much higher (970 °C). It should be mentioned that a great variety of the garnet composition in zircons from dolomitic marble can indicate a major variation in the temperature during their crystallisation. The values obtained are between 670 and 945 °C. It gives us additional evidence that zircon may be formed at the prograde path of the metamorphic evolution within a wide interval of temperature. The study of pyroxene-plagioclase symplectites, which developed after omphacite in eclogites of the Kumdy-Kol deposit, shows that these rocks experienced conditions of granulite facies metamorphism (T = 800–820 °C, P = 10 kbar; Shatsky & Sobolev, 1985). This is also confirmed by the investigation of inclusions in zircons from diamondiferous rocks (Hermann et al., 2001). The granulite facies conditions are also inferred from pyroxene-spinel symplectites developed around grossular-pyrope garnets in dolomite marbles as well as from finding of sapphirine and corundum in them (Sobolev et al., 2001). Newly formed pyroxene is characterised by a high content of the Ca-Tschermak component (up to 23%). In the Kulet domain, the high-pressure rocks contain no diamonds, but coesite inclusions are established in them (Shatsky et al., 1998b; Parkinson, 2000). The equilibrium temperature of the eclogites from Unit II indicates that metamorphic rocks from this unit were metamorphosed at temperatures not higher than 650–800 °C (Table 3). The recent discovery of coesite as inclusions in garnet from garnet-mica schist of the Kulet area indicates that some blocks of Unit II were metamorphosed at pressures higher than 26 kbar (Shatsky et al., 1998b). The estimation of the equilibrium temperature of eclogite, garnet-amphibole-zoisite and garnet-kyanite-biotite- amphibole rocks show that the temperature of the UHP metamorphism did not exceed 800 °C. The presence of primary amphibole in eclogites indicate that pressure did not exceeded 27 kbar (Poli & Schmidt, 1995). To summarise the available P–T data (Table 3), rocks exhumed from two levels can be recognised in the Kokchetav Massif. The first corresponds to diamondiferous rocks of the western block (P > 43 kbar, T  1000 °C). For the eclogites and country rocks of the eastern block the peak temperature is 650–800 °C at P = 26–28 kbar, and for amphibolites 650 °C at 10 kbar (Ota et al., 2000). There is a reason to think that amphibolites were formed during retrograde amphibolite facies (Shatsky & Sobolev, 1993; Hermann et al., 2001).

Geochemistry of metamorphic rocks and the age of the UHP metamorphism Based on major element data we can conclude that the eclogite protoliths were rocks of basic composition related to the tholeiite series (Shatsky et al., 1993). Trace and REE elements data suggests that island arc or oceanic basalts can be considered as the protoliths of eclogites. As there are no rocks in the belt of UHP and HP rocks whose protoliths are andesites, island arc basalts can be excluded from consideration as possible protoliths of eclogites. When oceanic-type basalts are considered as protoliths of eclogites, according to the existing classification they are T-type basalts. This type of

The Kokchetav massif of Kazakhstan

95

basalts occurs not only in mid-ocean ridges but also among products of basalt volcanism of passive margins. The composition of diamondiferous metamorphic rocks varies widely. Shatsky et al. (1995) concluded that diamondiferous metamorphic rocks correspond to shales mixed with carbonate. Biotite-garnet kyanite schists and biotite schists from Unit II and garnet-muscovite schists from Unit I have high Al2O3 contents and correspond to shale. Mica schists from Kulet (Unit II) have uncommonly high K2O and SiO2 contents; arkoses may be their protoliths. In most diamondiferous rocks the Th/U ratio is lower than in continental upper crust (Shatsky et al., 1999a). For the rocks from Unit II and garnet-muscovite schists and granito-gneisses from Kumdy-Kol this ratio is higher than in the continental upper crust. The distribution of REE elements in gneisses and schists of Unit II and in granitognesses from Unit I are strongly fractionated and have a larger Eu anomaly than that of typical crustal rocks (Fig. 9). The diamondiferous metamorphic rocks are depleted in REE elements compared to upper crust compositions (Fig. 10). Based on the REE distribution patterns, the Kumdy-Kol rocks can be subdivided into 4 groups. The first group exhibits a depletion of the LREE ((La/Yb)N = 0.19) and a high Sm/Nd ratio (Fig. 10a), whereas the second group shows a variable abundance flat HREE pattern with a small Eu anomaly (Fig. 10b). Both groups are depleted (with respect to the average upper

Fig. 9. Chondrite-normalised REE patterns for granite-gneisses from Kulet and Sulu-Tyube. Thick line: upper crust.

96

V.S. Shatsky & N.V. Sobolev

Fig. 10. Chondrite-normalised REE patterns for selected groups of rocks. (a) REE abundance in group 1; (b) REE abundance in group 2; (c) REE abundance in group 3; (d) REE abundance in group 4; (e) REE pattern for trondhjemite.

crust composition) in incompatible elements excluding Ti, K and Rb. The third rock group exhibits REE patterns similar to the second group but with a relatively large negative Eu anomaly (Eu*/Eu = 0.6, Fig. 10c). Compared to upper crust compositions these samples are depleted in REE, Ba, Ta and Th. The REE patterns of the fourth group are strongly fractionated and have a larger Eu anomaly (Eu*/Eu = 0.34) than that of typical crustal rocks (Fig. 10d). The lowest Sm/Nd ratios and highest Rb/Cs ratios are observed in these rocks. Rather numerous rock groups have characteristics intermediate between the groups III and IV, reflecting mixing. There is no correlation between bulk chemical composition and REE pattern. Only the group IV rocks have a rather narrow chemical compositional range including granite-gneisses and high-Mg garnetmuscovite-kyanite schists. The gneisses and schists from Unit I and II have Sm–Nd model ages within the interval 2.08–2.65 Ga (Shatsky et al., 1999a). Zircon xenocrysts from the diamondiferous

The Kokchetav massif of Kazakhstan

97

gneisses gave an age of 2 Ga (Claoue-Long et al., 1991). This age can be interpreted as the age of crust formation. The Sm–Nd model age of eclogites varies between wide limits (0.8–2.1 Ga) (Shatsky et al., 1993). Scattering in model ages of eclogites can be explained by the contamination of eclogites with Nd from country rocks during retrograde events. The Sm–Nd isochron diagrams for diamondiferous metamorphic rocks and eclogites show that whole rock, clinopyroxene and garnet do not form an isochron. This gives us additional evidence that Nd was exchanged in a later hydrothermal process. Mineral separates from the different types of rocks from Unit I and amphibole-garnet-zoisite rocks from Unit II (Kulet area) define a regression line with 524  25 Ma. This is clear evidence that before UHP metamorphism all types of rocks in Units I and II, excluding eclogites, had a similar 143Nd/144Nd ratio. However deviations of the data from the isochron are far outside analytical error (MSWD-140). Mineral disequilibrium might cause a considerable error in Nd ages. At the same time this age is supported by a four-point mineral isochron for two high-temperature eclogites from Unit I (535  3 Ma, Shatsky et al., 1999a) as well as by U–Pb dating of zircons (Claoue-Long et al., 1991; Hermann et al., 2001). The internal isochron for amphibolegarnet-zoisite rock (Kulet) gives 522 ± 27 Ma (MSWD-3.29). Whole rock analyses from Units I and II show a linear array which may represent a disturbed isochron. Diamondbearing rocks show a considerable range of Sm/Nd ratio (0.205–0.962). The Sm/Nd ratios of metamorphic rocks from Unit II vary in a remarkably narrow interval (0.14–0.194) and do not differ significantly from the average value for continental crust (Shatsky et al., 1999a). As mentioned above, matrix biotite and muscovite are found as late alteration products in diamondiferous gneisses of Kumdy-Kol. Ar–Ar age determination of micas from diamondiferous garnet-biotite gneisses yields an age of 517 Ma (Shatsky et al., 1999a). High equilibrium temperatures of diamondiferous rocks suggest their partial melting (Shatsky et al., 1995; Hermann et al., 2001). Geochemical data confirm this supposition (Shatsky et al., 1995, 1999a). Diamondiferous rocks demonstrate a considerable scatter in Sm/Nd values (from 0.2 to 0.96; Shatsky et al., 1999a). The isotope data show that prior to high pressure metamorphism all varieties of rocks, with the exception of eclogites, had similar 143Nd/144Nd ratios (Nd value of –13.3). Confirmation of partial melting is received from distribution of rare earth elements. Diamondiferous rocks are depleted with respect to light REE ((La/Yb)N = 0,19–1). Trondhjemite, which cuts an eclogite body and differs from the other granite rocks by a high Na/K ratio, has a REE distribution consistent with the equilibrium of the melt with a garnet-enriched restite. The HREE depletion suggests that trondhjemite is a result of the partial melting of eclogite (Shatsky et al., 1999a). As said above, diamondiferous rocks are interlayered with granite gneisses, and migmatised garnet-biotite gneisses are observed. It could be supposed that some part of granite gneisses and migmatite have been formed by the melting of diamondiferous rocks (Shatsky et al., 1999a).

98

V.S. Shatsky & N.V. Sobolev

Exhumation of high pressure rocks of the Kokchetav Massif In discussing exhumation models, the plotting of a regressive P–T path play a crucial role (Fig. 11). But to plot exhumation paths is rather difficult, because in many cases only fragments of this trend are documented. Thus, the granulite facies stage of metamorphism is recorded only in some specimens of eclogites and Ca-silicate rocks. On the basis of high-Mg calcite present in pyroxenes of carbonate rocks, Ogasawara et al. (2000) distinguished a step with T > 800 °C and P  25 kbar at the regressive stage of metamorphism (stage 2, Fig. 11.). Thus, at the first stage of exhumation diamondiferous rocks should be cooled from 1000 °C to 800 °C with pressure decreased from > 42 to 25 kbar. The next stage corresponds to the isothermal uplift to pressures of about 10 kbar, with the subsequent nearly isobaric cooling to 600–650 °C (stage 3, Fig. 11). Final stages of exhumation correspond to greenschist facies metamorphism (stage 5, Fig. 11). In addition to the P–T regressive trend, the following established facts must be taken into account in a model for exhumation. On the basis of SHRIMP dating of zircons (Claoue-Long et al., 1991), Sm–Nd dating of eclogites (535 ± 3 Ma), zircon domains containing diamond indusions from diamondiferous metapelites (530 ± 7 Ma) and dating of secondary biotite and muscovite from diamondiferous gneiss (517 ± 5 Ma), it was concluded that the time interval between high pressure metamorphism and the uplift of rocks to a crustal level could be no more than 10 Ma and the vertical rate of exhumation was not less than 1.2 cm/year (Dobretsov et al., 1995; Shatsky et al., 1999a). New ages for zircons from metamorphic diamondiferous rocks of the Kumdy-Kol and Barchi sites have been published recently (Hermann et al., 2001). Dating from separate zones of zircons containing mineral inclusions of both high pressure and retrograde stages of metamorphism shows that there is no systematic difference in ages between different domains in the zircons that belong to different stages of metamorphism. It means that the period of exhumation of high pressure rocks to depths corresponding to a pressure of 6–8 kbar is within the accuracy of the method, no more than 6 Ma, which connotes an exhumation rate of 1.8 cm/year. Additional evidence of rapid exhumation is a low degree of nitrogen aggregation in diamonds of metamorphic rocks. According to de Corte et al. (1999), these diamonds could not exist at 950 °C for more than 5 Ma and at 1000 °C for more than 0.1 Ma. Worthy of note is that there is some uncertainty in activation energy for nitrogen in diamond to aggregate from Ib-to-IaA. Using energies of activation of 5 eV, Finnie et al. (1994) obtained residence time of 0.2 Ma for 900 °C. Thus, the estimates obtained in different ways indicate high rates of the first stage of exhumation of the Kokchetav high pressure rocks. The time interval between the climax of metamorphism and exhumation to the level of the Earth’s crust does not exceed 6 Ma. A decrease in temperature from 1000 to 800 °C occurred during the time interval of no more than 2–5 Ma, as inferred from the degree of nitrogen aggregation in diamond and zoning in garnets. Structural studies showed that the deformation of diamondiferous rocks of the Kumdy-Kol deposit is insignificant (Theunissen et al., 2000a). The diamond inclusions in garnet are often intergrown with mica crystals carrying no traces of deformation. The low differential stress is supported by the fact that garnet aggregates in massive garnet-

The Kokchetav massif of Kazakhstan

99

Fig. 11. P–T retrograde path of Kumdy-Kol diamondiferous rocks. Calculated burial and exhumation P–T paths from Ernst & Peacock (1996); graphite–diamond transition after Bundy (1980); quartz–coesite transition after Mirwald & Massonne (1980); Bt-out, Opx-in curves after Vielzeuf & Montel (1994); Phl + Qtz = En + Sa + M after Vielzeuf & Clemens (1992). 1) biotite stability in KCMASH system (Hermann & Green, 2001); 2) biotite stability in metapelites (Vielzeuf & Holloway, 1988); the grey bend shows the field of phengite melting (Hermann & Green, 2001).

pyroxene-quartz rocks have foam structure indicating that their microstructure was controlled by grain boundary free energy. We believe all these facts could be explained by the partial melting of either metapelites or granitic rocks in the Kumdy-Kol domain. The occurrence of the melt is responsible for an essential reduction of viscosity, and a density difference () of crustal rocks and mantle material, and reduced friction between the upwelling sialic block, subducting and overriding plates.

100

V.S. Shatsky & N.V. Sobolev

Besides , the rate of exhumation seems to depend on the internal pressure in the subducting continental crustal block (Dobretsov & Kirdyashkin, 1992, 1998), which can be regarded as a viscous layer between subducting continental lithosphere and surrounding mantle.

References Bundy, F.R. (1980): The P, T phase reaction diagrams for elemental carbon. J. Geophys., 85:6930–6936. Cartigny, P., De Corte, K., Shatsky, V.S., Ader, M., De Paepe, P., Sobolev, N.V. & Javoy, M. (2001a): The origin and formation of metamorphic microdiamonds from the Kokchetav massif, Kazakhstan: a nitrogen and carbon isotopic study. Chem. Geol., 176:265–281. Cartigny, P., Harris, J.W. & Javoy, M. (2001b): Diamond genesis, mantle fractionation and mantle nitrogen content: a study of 13C-N concentrations in diamonds. Earth Planet. Sci. Lett., 185:85–98. Chopin, C. (1984): Coesite and pure pyrope in high-grade blueschists of the Western Alps: A first record and some consequences. Contrib. Miner. Petrol., 86:107–118. Claoue-Long, J.C., Sobolev, N.V., Shatsky, V.S. & Sobolev, A.V. (1991): Zircon response to diamond-pressure metamorphism in the Kokchetav massif. Geology, 19:710–713. Coleman, R.G. & Wang, X. (eds.) (1995): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 540 p. de Corte, K., Cartigny, P., Shatsky, V.S., Javoy, M. & Sobolev, N.V. (1998): Evidence of inclusions in metamorphic microdiamonds from the Kokchetav Massif, Northern Kazakhstan. Geochim. Cosmochim. Acta, 62:3765–3773. de Corte, K., Cartigny, P., Shatsky, V.S., De Paepe, P., Sobolev, N.V. & Javoy, M. (1999): Characteristics of microdiamonds from UHPM rocks of the Kokchetav massif (Kazakhstan). In Gurney, J.J., Gurney, J.L., Pascoe, M.D. & Richardson, S.H. (eds.): Proc. 7th Int. Kimberlite Conf., Cape Town. Cape Town: Redroof Publ., 1 (The Dawson Vol.): 174–182. de Corte, K., Israeli, E.S., Shatsky, V.S., de Paepe, P. & Navon, O. (2001): Garnet and clinopyroxene inclusions in metamorphic microdiamonds from the Kokchetav Massif, Kazakhstan. In Fluid/slab/mantle interactions and ultrahigh-P minerals /UHPM Workshop 2001, Tokyo, Japan/, 16–20. Dobretsov, N.L. & Kirdyashkin, A.G. (1992): Subduction zone dynamics: models of an accretionary wedge. Ofioliti, 17:155–164. Dobretsov, N.L. & Kirdyashkin, A.G. (1998): Deep-level geodynamics. Rotterdam: Balkema. Dobretsov, N.L. & Sobolev, N.V. (1970): Eclogites from metamorphic complexes of the USSR. Phys. Earth Planet. Inter., 3:401–424. Dobretsov, N.L., Sobolev, N.V., Shatsky, V.S., Coleman, R.G. & Ernst, W.G. (1995): Geotectonic evolution of diamondiferous parageneses Kokchetav Complex, Northern Kazakhstan - the geologic enigma of ultrahigh-pressure crustal rocks within Phanerozoic foldbelt. Isl. Arc, 4:267–279. Dobretsov, N.L, Theunissen, K. & Smirnova, L. (1998): Structural and geodynamic evolution of diamondbearing metamorphic rocks of Kokchetav massif, Kazakhstan. Geol. Geofiz., (39)12:1645–1666 (in Russian). English transl.: Russ. Geol. Geophys., 39:1650–1661. Dobretsov, N.L., Theunissen, K., Dobretsov, N.N., Smirnova, L.V. & Zayachkovsky, A.A. (1999a): Geological and tectonic outline of the Kokchetav M. massif. In Dobretsov et al. (ed.): Field guide book of the 4th Int. Eclogites Field Symp., Russia, Novosibirsk, 6–24. Dobretsov, N.L., Sobolev, N.V. & Shatsky, V.S. (eds.) (1999b): Diamondiferous and High-Pressure Metamorphic Rocks of the Kokchetav Massif. Field guide book (4th Int. Eclogite Field Symp., 1999, Novosibirsk, Russia). 134 p. Dobrzhinetskaya, L.F., Eide, E.A., Larsen, R.B., Sturt, B.A., Tronnes, R.G., Smith, D.C., Taylor, W.R. & Posukhova, T.V. (1995): Microdiamond in high-grade metamorphic rocks of the Western Gneiss region, Norway. Geology, 23:597–600. Ekimova, T.T., Lavrova, L.D., Nadezhdina, E.D., Petrova, M.A. & Pechnikov, V.A. (1994): Formation conditions of Kumdy Kol diamond deposit (North Kazakhstan) Geol. Rudn. Mestorozhd., 36:455–465 (in Russian).

The Kokchetav massif of Kazakhstan

101

Ellis, D.J. & Green, D.H. (1979): An experimental study of the effect of Ca upon garnet-clinopyroxene Fe-Mg exchange equilibria. Contrib. Mineral. Petrol., 71:13–22. Ernst, W.G. & Liou, J.G. (eds.) (2000): Ultrahigh-pressure metamorphism and geodynamics in collision-type orogenic belts /Int. Book Ser., 4/. Boulder (Co.): Geol. Soc. Am. & Columbia (Md.): Bellwether Publ. Ltd. Ernst, W.G. & Peacock, S.M. (1996): A thermotectonic model for preservation of ultrahigh-pressure phases in metamorphosed continental crust. In Bebout, G.E., Scholl, D.W., Kirby, S.H. & Platt J.P. (eds.): Subduction: Top to Bottom /Geophys. Monogr., 96/. Washington, D.C.: Am. Geophys. Union, 171–178. Finnie, K.S., Fisher, D., Griffin, W.L., Harris, J.W. & Sobolev, N.V. (1994): Nitrogen aggregation in metamorphic diamonds from Kazakhstan. Geochim. Cosmochim. Acta, 58:5173–5177. Green, T.H. & Hellman, P.L. (1982): Fe-Mg partitioning between coexisting garnet and phengite at high pressure, and comments on a garnet-phengite geothermometer. Lithos, 15:253–266. Haggerty, S.E. (1999): A diamond trilogy: Superplumes, supercontinents, and supernovae. Science, 285:851–860. Harlow, G.E. (1999): Interpretation of Kcpx and CaEs components in clinopyroxene from diamond inclusions and mantle samples. In Gurney, J.J., Gurney, J.L., Pascoe, M.D. & Richardson, S.H. (eds.): Proc. 7th Int. Kimberlite Conf., Cape Town. Cape Town: Redroof Publ., 1 (The Dawson Vol.): 321–331. Hermann, J. & Green, D. (2001): Experimental constraints on high pressure melting in subducted crust. Earth Planet. Sci. Lett., 188:149–168. Hermann, J., Rubatto, A., Korsakov, A. & Shatsky, V.S. (2001): Multiple zircon growth during fast exhumation of diamondiferous, deeply subducted continental crust (Kokchetav Massif, Kazakhstan). Contrib. Mineral. Petrol., 141:66–82. Hodges, K.V. & Spear, F.S. (1981): Geothermometry, geobarometry, garnet close temperatures and the Al2SiO5 triple point. Eos, Trans. Am. Geophys. Union, 62:1060. Kaneko, Y., Maruyama, S., Terabayashi, Yamamoto, H., Ishikawa, M., Anma, R., Parkinson, C.D., Ota, T., Nakajima, Y., Katayama, I., Yamamoto, J. & Yamauchi, K. (2000): Geology of the Kokchetav UHP-HP metamorphic belt, northern Kazakhstan, Isl. Arc, 9:264–283. Katayama, I., Maruyama, S., Kaneko, Y. & Liou, J.G. (1998): Mineral inclusion in zircon: A window to prograde metamorphism of the Kokchetav UHP rocks. Western Pacific Geophysical Meeting 1998, Taipei, Taiwan /EOS Trans. Am. Geophys. Union, 79(24), Suppl./, W126. Katayama, I., Parkinson, C.D., Okamoto, K., Nakajima, Y & Maruyama, S. (2000): Supersilicic clinopyroxene in ultrahigh-pressure metamorphic rocks from the Kokchetav massif, Kazakhstan. Am. Mineral., 85:1368–1374. Katayama, I., Maruyama, S., Parkinson, C.D., Terado, K. & Sano, Y. (2001): Ion micro-probe U-Pb zircon geochronology of peak and retrograde stages of ultrahigh-pressure metamorphic rocks from the Kokchetav massif, northern Kazakhstan. Earth Planet. Sci. Lett., 188:185–198. Korsakov, A.V., Shatsky, V.S. & Sobolev, N.V. (1998): The first finding of coesite in the eclogites of the Kokchetav Massif. Dokl. Ross. Akad. Nauk, 360:77–81 (in Russian). English transl.: Dokl. Earth Sci., 360:469–473. Korsakov, A.V., Shatsky, V.S. & Sobolev, N.V. & Zayachkovsky, A.A. (2002): Garnet-biotite-clinozoisite gneiss: a new type of diamondiferous metamorphic rocks from the Kokchetav massif. Eur. J. Mineral., 14:915–928. Liou, J.G. & Banno, S., (eds.) (2000): Petrotectonic characteristics of the Kokchetav Massif, Northern Kazakhstan. Isl. Arc, 9(3):259–455. Luth, R.W. (1997): Experimental study of the system phlogopite-diopside from 3.5 to 17 GPa. Am. Mineral., 82:1198–1209. Massonne, H.J. (1999): A new occurrence of microdiamonds in quartzofeldspathic rocks of the Saxonian Erzgebirge, Germany, and their metamorphic evolution. In Gurney, J.J., Gurney, L.G., Pascoe, M.D. & Richardson, S.H. (eds.): Proc. 7th Int. Kimberlite Conf., Cape Town, Cape Town: Redroof Publ., 2 (The Nixon Vol.):533–539. Mirwald, P.W. & Massonne, H.-J. (1980): The low-high quartz and quartz-coesite transition to 40 kbar between 600 °C and 1600 °C and some reconnaissance data on the effect of NaAlO2 component on the low quartzcoesite transition. J. Geophys. Res., 85:6983–6990.

102

V.S. Shatsky & N.V. Sobolev

Ogasawara, Y., Ohta, M., Fukasawa, K., Katayama, I. & Maruyama, S. (2000): Diamond-bearing and diamond-free metacarbonate rocks from Kumdy-Kol in the Kokchetav massif, northern Kazakhstan. Isl. Arc, 9:400–416. Ota, T., Terabayashi, M., Parkinson, C.D. & Masago, H. (2000): Thermobaric structure of the Kockhetav ultrahigh-pressure-high-pressure massif deduced from a north-south transect in the Kulet and Saldat-Kol regions, northern Kazakhstan. Isl. Arc, 9:328–357. Parkinson, C.D. (2000): Coesite inclusions and prograde compositional zonation of garnet in whiteschist of the HP-UHPM Kokchetav massif, Kazakhstan: a record of progressive UHP metamorphism. Lithos, 52:215–233. Poli, S. & Schmidt, M.V. (1995): H2O transport and release in subduction zones: Experimental constraints on basaltic and andesitic systems. J. Geophys. Res., 100:22,299–22,314. Powell, R. (1985): Regression diagnostic and robust regression in geothermometer/geobarometer calibration: the garnet-clinopyroxene geothermometer revisited. J. Metamorph. Geol., 3:231–243. Reverdatto, V.V. (1999): Mafic granulites in the vicinity of Enbek-Berlyk village. In Dobretsov, N.L., Sobolev, N.V. & Shatsky, V.S. (eds.): Diamondiferous and High-Pressure Metamorphic Rocks of the Kokchetav Massif. Field guide book (4th Int. Eclogite Field Symp., 1999, Novosibirsk, Russia). Sato, K. & Katsura, T. (2001): Experimental investigation on dolomite dissociation into aragonite+magnesite up to 8,5 GPa. Earth Planet. Sci. Lett., 184:529–534. Shatsky, V.S. & Sobolev, N.V. (1985): Pyroxene-plagioclase symplectites in eclogites of Kokchetav massif. Geol. Geofiz., 26(9):83–89 (in Russian). English transl.: Sov. Geol. Geophys., 26(9):76–81. Shatsky, V.S. & Sobolev, N.V. (1993): Some specific features of the origin of diamonds in metamorphic rocks. Dokl. Ross. Akad. Nauk, 331:217–219 (in Russian). Shatsky, V.S. & Sobolev, N.V. (2001): Prograde and retrograde lamellae in clinopyroxenes from UHP pyroxene-carbonate rocks of Kokchetav massif. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 147. Shatsky, V.S., Sobolev, N.V. & Stenina, N.G. (1985): Structural peculiarities of pyroxenes from eclogites. Terra Cognita, 5:436–437. Shatsky, V.S., Sobolev, N.V. & Gilbert, A.E. (1989a): Eclogites of Kokchetav massif. In Sobolev, N.V. (ed.): Eclogites and glaucophane schists in folded belts. Novosibirsk: Nauka, 54–83 (in Russian). Shatsky, V.S., Sobolev, N.V. & Yefimova, E.S. (1989b): Morphological features of accessory microdiamonds from metamorphic rocks of the earth’s crust. In 28th Int. Geol. Congr., Washington, D.C., Workshop on Diamonds, Ext. Abstr., 94–95. Shatsky, V.S., Sobolev, N.V., Zayachkovsky, A.A., Zorin, T.Y. & Vavilov, M.A. (1991): New occurence of microdiamonds in metamorphic rocks as a proof of regional character of ultra-high pressure metamorphism in Kokchetav massif. Dokl. Akad. Nauk SSSR, 321:193–198 (in Russian). Shatsky, V.S., Jagoutz, E., Kozmenko, O.A., Blinchik, T.M. & Sobolev, N.V. (1993): The age and origin of eclogites from Kokchetav massif (Northern Kazakhstan). Russ. Geol. Geophys., 34:40–49. Shatsky, V.S., Sobolev, N.V. & Vavilov, M.A. (1995): Diamond-bearing metamorphic rocks of the Kokchetav massif (Northern Kazakhstan). In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 427–455. Shatsky, V.S., Rylov, G.M., Efimova, E.S., De Corte, K. & Sobolev, N.V. (1998a): The morphology and real structure of microdiamonds from the Kokchetav massif metamorphic rocks, kimberlites and alluvial placers. Geol. Geofiz., 39:942–955 (in Russian). English transl.: Russ. Geol. Geophys., 39:949–961. Shatsky, V.S., Theunissen, K., Dobretsov, N.L. & Sobolev, N.V. (1998b): New indications of ultrahighpressure metamorphism in the mica schists of the Kulet site of the Kokchetav Massif (North Kazakhstan). Geol. Geofiz., 39:1039–1044 (in Russian). English transl.: Russ. Geol. Geophys., 39:1041–1046. Shatsky, V.S., Jagoutz, E., Sobolev, N.V., Kozmenko, O.A., Parkhomenko, V.S. & Troesch, M. (1999a): Geochemistry and age of ultrahigh pressure metamorphic rocks from the Kokchetav massif (Northern Kazakhstan). Contrib. Mineral. Petrol., 137:185–205. Shatsky, V.S., Zedgenizov, D.A., Yefimova, E.S., Rylov, G.M., De Corte, K. & Sobolev, N.V. (1999b): A comparison of morphology and physical properties of microdiamonds from the mantle and crustal environments. In Gurney, J.J., Gurney, L.G., Pascoe, M.D. & Richardson, S.H. (eds.): Proc. 7th Int. Kimberlite Conf., Cape Town, Cape Town: Redroof Publ., 2 (The Nixon Vol.):757–763.

The Kokchetav massif of Kazakhstan

103

Sobolev, N.V. & Shatsky, V.S. (1987): Inclusions of carbon minerals in the garnets from metamorphic rocks. Geol. Geofiz., 28(7):77–80 (in Russian). English translation: Sov. Geol. Geophys., 28(7):69–71. Sobolev, N.V. & Shatsky, V.S. (1990): Diamond inclusions in garnets from metamorphic rocks: a new environment for diamond formation. Nature, 343:742–746. Sobolev, V.S. & Sobolev, N.V. (1980): New proof on very deep subsidence of eclogitized crustal rocks. Dokl. Akad. Nauk SSSR, 250:683–685 (in Russian). Sobolev, N.V., Shatsky, V.S., Vavilov, M.A. & Goryainov, S.V. (1991): Coesite inclusion in zircon from diamondiferous gneiss of Kokchetav massif - first find of coesite in metamorphic rocks in the USSR territory. Dokl. Akad. Nauk SSSR, 321:184–188 (in Russian). Sobolev, N.V., Shatsky, V.S., Vavilov, M.A. & Goryainov, S.V. (1994): Zircon from ultrahigh-pressure metamorphic rocks of folded regions as a unique container of inclusions of diamond, coesite and coexisting minerals. Dokl. Ross. Akad. Nauk, 334:482–488 (in Russian). Sobolev, N.V., Shatsky, V.S. & Korsakov, A.V. (1998): Significance of coesite in UHP metamorphic rocks of Kockhetav Massif, Kazakhstan. In Int. Workshop on Ultra-High-pressure Metamorphism and Exhumation; Dec. 3–6, 1998, Stanford Univ., Abstr., A70–71. Sobolev, N.V., Schertl, H-P., Burchard, M. & Shatsky, V.S. (2001): Unusual pyrope-grossular-garnet and its paragenesis from diamond bearing carbonate-silicate rocks of the Kokchetav massif (Northern Kazakhstan). Dokl. Ross. Akad. Nauk, 380:1–5. (in Russian). Stöckhert, B., Duyster, J., Trepman, C. & Massonne, H.J. (2001): Microdiamond daughter crystals precipitated from supercritical C-O-H silicate fluids included in garnet, Erzgebirge, Germany. Geology, 29:391–394. Sunagawa, I. (1984): Morphology of natural and synthetic diamond crystals. In Sunagawa, I. (ed.): Materials science of the Earth’s interior. Tokyo: Terra Sci. Publ. Co., 303–330. Theunissen, K., Dobretsov, N.L., Shatsky, V.S., Smirnova, L. & Korsakov, A. (2000a): The diamond-bearing Kokchetav UHP massif in the Northern Kazakhstan: exhumation structure. Terra Nova, 12:181–187. Theunissen, K., Dobretsov, N.L., Korsakov, A., Travin, A., Shatsky, V.S., Smirnova, L. & Boven, A. (2000b): Two contrasting petrotectonic domains in the Kokchetav megamelange (north Kazakhstan): Difference in exhumation mechanisms of ultrahigh–pressure crustal rocks, or a result of subsequent deformation? Isl. Arc, 9:284–303. Udovkina, N.G. (1985): Eclogites of the USSR. Moscow: Nauka, 285 p. (in Russian). Vavilov, M.A., Sobolev, N.V. & Shatsky, V.S. (1991): Micas of the diamondiferous metamorphic rocks of the Northern Kazakhstan. Dokl. Akad. Nauk SSSR, 321:466–470 (in Russian). Vielzeuf, D. & Clemens, J.D. (1992): The fluid absent melting of phlogopite + quartz: experiments and models. Am. Mineral., 77:1206–1222. Vielzeuf, D. & Holloway, J.R. (1988): Experimental determination of the fluid absent melting relations in the pelitic syatem. Contrib. Mineral. Petrol., 99:257–276. Vielzeuf, D. & Montel, J.M. (1994): Partial melting of metagreywackes. 1. Fluid-absent experiments and phase relationships. Contrib. Mineral. Petrol., 117:375–393. Xu, S., Okay, A.I., Sengor, A.M.C., Su, W., Liu, Y. & Jiang, L. (1992): Diamond from the Dabie Shan metamorphic rocks and its implication for tectonic setting. Science, 256:80–82. Zhang, R.Y., Liou, J.G., Ernst, W.G., Coleman, R.G., Sobolev, N.V. & Shatsky, V.S. (1997): Metamorphic evolution of diamond-bearing rocks from the Kokchetav massif, Northern Kazakhstan. J. Metamorph. Geol., 15:479–496. Zonenshain, L.P., Kuzmin, M.I. & Natapov, L.M. (1990): Geology of the USSR: A plate tectonic synthesis /Geodyn. Ser., 21/. Washington (D.C.): Am. Geophys. Union, 240 p.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 5, 105–144

The Dabie Shan–Sulu orogen TAKAO HIRAJIMA* and DAISUKE NAKAMURA Department of Geology and Mineralogy, Graduate School of Earth Science, Kyoto University, Kyoto 606-8502, Japan; * e-mail: [email protected]

Geological framework The Qinling–Tongbai–Hong’an–Dabie Shan area is an about 2000 km long Triassic Indosinian orogenic belt produced by the collision between the Sino–Korean and the Yangtze cratons (Fig. 1). Its eastern extension, the Sulu area, occupies the southeastern side of the Shandong Peninsula, and is considered to be displaced about 500 km by the NE–SW trending left lateral Tan-Lu Fault after the Mesozoic (Fig. 1). Among these areas, most of the UHP rocks were found from Hong’an, Dabie Shan and Sulu areas, suggesting that these areas represent the most extensive UHP metamorphic belt in the world. Their UHP peak is dated around 220–230 Ma (e.g. Ames et al., 1993, 1996; Li et al., 1993; Hacker & Wang, 1995; Hacker et al., 1996; Rowley et al., 1997) and these UHP terrains are considered to be formed chiefly by attempted north-directed subduction of the Yangtze craton or a microcontinent beneath the Sino–Korean craton (e.g. Hacker et al., 1996). Many available data concerning petrology, structural geology and geochronology of the Hong’an and Dabie Shan areas have accumulated during this decade, and several geologic subdivisions were proposed in the relevant area. This paper basically follows the subdivision of Hacker et al. (1998) because the locations of UHP/HP rocks are well indicated, except for the geotectonic subdivision of the western part of the Dabie Shan area (e.g. Castelli et al., 1998). Hacker et al. (1998) subdivided the main rock units in the Dabie Shan area from south to north as follows: a fold and thrust belt, Susong Group (blueschist and high-pressure amphibolite), quartz eclogite, coesite eclogite, Northern Orthogneiss unit, Luzhenguang Group and Foziling Group. This subdivision is still supported in the eastern half of the Dabie Shan area but not in the western half by subsequent researchers (e.g. Castelli et al., 1998; Faure et al. 2003). In this paper we follow the geologic subdivision in the western half of the Dabie Shan area by Castelli et al. (1998) (see Fig. 2), where it is provisionally called amphibolite–granulite unit. The northward increase in metamorphic grade is also recognised in the Hong’an area, to the west of the Dabie Shan area bounded by the Shang-Ma Fault. Blueschist facies rocks occur in the southern part, and distinct greenschist, prograde amphibolite and eclogite retrogressed to amphibolite units have been mapped northward (Fig. 2). Quartz eclogite and coesite eclogite units occupy the north-central part. The northern part of the area is occupied by E–W trending fault bounded units, which contain HP

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

106

T. Hirajima & D. Nakamura

Fig. 1. Tectono-metamorphic sketch map of the Qinling–Dabie Shan–Sulu orogen.

rocks. Eide & Liou (2000) reported the similar metamorphic zonal mapping in the Hong’an area and considered that the high-pressure (HP) Hong’an eclogites, often preserving prograde crystallisation histories, can be directly linked in time and space to the blueschist/blueschist–greenschist rocks exposed to the south. All the units in the Hong’an–Dabie Shan area are intruded by voluminous Cretaceous granitoids, and the northern margin is overlain by Cretaceous and Cenozoic sediments, although Eide & Liou (2000) described that the Hong’an area escaped from the thermal and structural overprint during Early Cretaceous intrusions of voluminous granites and granodiorites. The Cretaceous granitoid intrusion caused the pervasive migmatization in the Northern Orthogneiss unit and the northern half of the amphibolitegranulite unit of the Dabie Shan area. Therefore, Faure et al. (2003) renamed this unit as “Gneiss or mylonitised migmatite zone”. The geologic framework of the Shandong Peninsula is rather simpler than that of the Hong’an–Dabie Shan area. The Shandong Peninsula can be divided into northwestern and southeastern areas with a boundary along Yantai–Qingdao–Wulian (YQW) Fault (Fig. 3). These areas also have been widely intruded by Cretaceous granitoids, but the nature of the basement rocks to the northwest and southeast of YQW Fault are strikingly different. The basement rocks of the northwestern area are characterised by the granulite to amphibolite facies orthogneisses and paraschists characterised by the sillimanite–garnet assemblage, locally overlain by lower-grade to unmetamorphosed rocks of the Upper Proterozoic Sinian system. Metabasic rocks in this area are amphibolite and granulite with or without garnet. The occurrence of eclogite has not been recognised through the extensive study during this decade, although some Chinese geologists had claimed to

Fig. 2. Tectono-metamorphic map of the Dabie Shan–Hong’an areas, based mainly on Hacker et al. (1998) and Castelli et al. (1998). The locations of eclogitic xenoliths in granodiorite follow Faure et al. (2003)

The Dabie Shan–Sulu orogen 107

108

T. Hirajima & D. Nakamura

Fig. 3. Tectono-metamorphic map of the Sulu area with locations of coesite eclogite and other ultrabasic and basic rocks with critical mineral assemblages, mainly after Wallis et al. (1997). YQW Fault: Yantai–Qingdao–Wulian Fault.

The Dabie Shan–Sulu orogen

109

have found eclogite (e.g. Cheng, 1986). The regional orthogneiss mainly belongs to a tonalitic series (Zhai et al., 2000) and has zircon ages of 2600–2900 Ma (Wang & Yan, 1992), Rb–Sr whole rock ages and chemical U–Th–total Pb ages of 1600–2020 Ma (Enami et al., 1993b; Ishizaka et al., 1994). The basement rocks of the southeastern area are also predominantly of tonalitic and granitic gneiss with amphibolite facies mineral assemblages, which occupies more than 90 vol% of the area. Metabasic rocks mainly occur as lenses, blocks or layers intercalated with the country gneiss. Most of the metabasic rocks are retrogressed to the amphibolite facies to various degrees, but the eclogitic mineral assemblages are still preserved in the core of less retrogressed rocks. Some eclogites also occur closely associated with peridotite or marble. Eclogites with coesite and/or its pseudomorphs are sporadic but occur throughout the southeastern area (Fig. 3). Garnet peridotite, less abundant than eclogite, also occurs in the southeast area of the YQW fault, although spinel peridotite occurs in the northwest area of the YQW fault. The similar UHP age to the Hong’an–Dabie Shan area is also reported in the area southeast of the YQW fault (218 Ma by Ames et al., 1996; 210–220 Ma by Jahn et al., 1996 and 1998). Most of the protolith ages of the country gneisses (around 700–800 Ma, Ishizaka et al., 1994; Ames et al., 1996; Hirajima & Fanning, 1999 and unpublished data) is strikingly different from those obtained in the area northwest of the YQW Fault, although an Early Proterozoic protolith age (1.7 Ga) was reported for the Weihai eclogite by the Sm–Nd isotopic data (Jahn et al., 1996). These data suggest that the southeastern area of the Shandong Peninsula was derived from the Yangtze craton and suffered UHP metamorphism but the northwestern area belongs to the Sino-Korean craton, characterised by the MP/LP type metamorphism with Archaean and early Proterozoic ages. In this paper, we call the area northwest of the YQW Fault as the Laiyang area and the southeastern area as the Sulu area. The extensive quartz eclogite unit is not recognised in the Sulu area, but the Haiyangsuo area (Fig. 3) is interpreted to be an equivalent to the quartz eclogite unit by Ye et al. (1999). The lower grade schist outcropping around Lianyuguang, located in the southernmost area of the Peninsula, is considered to be an epidote blueschist facies unit (e.g. Faure et al., 2001). The spatial extent of the UHP terrains and the position of the suture boundary between the Sino-Korean and the Yangtze cratons are still under debate in spite of the accumulation of new data. In the Dabie Shan area, Hacker et al. (1995) have proposed the Xiaotian–Mozitang Fault (Fig. 2), which bounds the Foziling and Luzhenguang Groups to the north and the Northern Orthogneiss unit and the amphibolite-granulite unit to the south. The latter is almost equivalent to the North Dabie Complex of Xu et al. (2001) or the gneiss or mylonitised migmatite zone of Faure et al. (2003) (Fig. 2). Zhang et al. (1996) have placed the suture boundary along the fault between the Northern Orthogneiss unit and the UHP eclogite unit, mainly because the UHP/HP rocks had not been found in the Northern Orthogneiss unit at that time. However, the recent finding of eclogitic xenoliths in Cretaceous granitoid rocks in the amphibolite-granulite unit (localities are shown as × in Fig. 2; Faure et al., 2003) and of eclogites and eclogitic rocks in the ultramafic belt in northern Dabie Mountains by Xu et al. (2001) suggest that the suture boundary should be the northern margin of the amphibolite-granulite unit, i.e.,

110

T. Hirajima & D. Nakamura

the Xiaotian–Mozitang fault (Xu et al., 2001) or a fault to the north of the Foziling Group (Hacker et al., 1998; Faure et al., 2003). In the Shandong Peninsula, the suture boundary between the Sino-Korean craton and the Yangtze craton is placed around the Yantai–Qingdao–Wulian (YQW) Fault (e.g. Wallis et al., 1997). Zhai et al. (2000) slightly modified the position of the suture boundary around the northern part of the YQW Fault. They proposed a new tectonic zone, Kunyushan boundary complex, bounded by the Muping Fault (northern part of YQW Fault) to the west and its sub-parallel branch fault (Mishan Fault) to the east. On the other hand, Faure et al. (2001) proposed that the suture boundary is located to the north of the Shandong Peninsula, mainly based on the same deformation style during the exhumation stage throughout the Peninsula, the lack of oceanic basin rocks around the proposed suture (i.e., the YQW fault), and the similarity of exhumation paths of granulites in the both sides of YQW Fault. From the petrological point of view, lack of late Proterozoic orthogneiss with ages of 700–800 Ma and eclogite in the northwestern area of the Shandong Peninsula do not support the idea of Faure et al. (2001).

Reconstruction of pre-HP/UHP stage geology The HP/UHP units in the Hong’an–Dabie Shan area and the Sulu area mainly consist of felsic gneiss with intercalation of eclogite, garnet pyroxenite, garnet peridotite, marble, jadeite-quartzite, and metapelite. UHP evidence, i.e., coesite, its pseudomorphs, low-Al Opx etc., is found mainly from eclogite and garnet peridotite (e.g. Okay et al., 1989; Wang et al., 1989; Hirajima et al., 1990; Yang et al., 1993). Inclusions of coesite and its pseudomorphs are also found from marble (e.g. Wang & Liou, 1991), felsic gneiss, phengite schist (Wang et al., 1992) jadeite quartzite (Cong et al., 1995; Su et al., 1996; Liou et al., 1997) and felsic whiteschist (Rolfo et al., 2000). However, the felsic gneiss is mainly composed of amphibolite facies mineral assemblages and most of them have no trace of UHP metamorphism. Such an occurrence of the UHP rocks caused an “in situ” vs. “exotic” argument even in the area in question. The reconstruction of the pre-UHP geological relationship between the UHP rocks and surrounding rocks is an important job for the geologist to settle this longterm issue. The UHP unit in the Dora-Maira, Brossasco–Isasca Unit, is one of the best examples suggesting the “in situ” origin of UHP rocks from the geological point of view, as the pre-Alpine geology is well reconstructed and various types of the constituent rocks, e.g. metapelite, eclogite, marble, metagranite and whiteschist contain various kind of UHP evidence (e.g. coesite, its pseudomorphs, pure pyrope in the felsic rocks, jadeite–kyanite assemblage after paragonite etc.; Chopin et al., 1991; Compagnoni et al., 1995; Schertl et al., 1991; see also Compagnoni & Rolfo, 2003, on the western Alps in this volume). The Hong’an–Dabie Shan–Sulu area widely suffered later stage extensive deformations mainly under the amphibolite facies conditions (e.g. Wallis et al., 1997; Faure et al., 2003), which pervasively masked pre-UHP geological relationship between the UHP rocks and the surrounding felsic gneiss. However, the coherency of eclogitebearing layers is better mapped in the Dabie Shan area than in the Sulu area.

The Dabie Shan–Sulu orogen

111

Wang et al. (1990) reported the regional scale coherency of the coesite eclogitebearing layers derived from the supracrustal metasedimentary sequence, the Chenjiahe Formation, in the coesite eclogite unit of the Dabie Shan area. The Chenjiahe Formation consists of carbonate-bearing biotite or hornblende gneisses with blocks of eclogite, marble, talc schist, amphibolite, pyroxenite and tremolite schist. Although the Chenjiahe Formation itself occurs as a discontinuous belt, Wang et al. (1990) depicted its continuity more than 40 km around Changpu–Wumiao–Shuanghe in the northern part of the coesite eclogite unit and around Shima in the southern part of the unit in Figure 2 of their paper (reconstructed in Fig. 4). Compagnoni et al. (2001) reconfirmed its coherency in the northern part of the coesite eclogite unit and reported a detailed geologic map in the relevant area. They concluded that Changpu–Pailou Unit, characterised by the close association of paragneiss, jadeite quartz granofels, marble, and eclogite, formed a NW–SE trending narrow belt about 1–2 km wide and 40 km long, is derived from a coherent supracrustal metasedimentary sequence which underwent UHP metamorphism. This suggests that the large coherent slices of the continental crust can subduct to the depths deeper than 100 km and be exhumed to the surface maintaining the pre-UHP stage geology. This is one of the concrete evidences of the “in situ” origin of UHP rocks. In Shuanghe of the Dabie Shan, Cong et al. (1995) succeeded to reconstruct the geology of UHP metamorphic slab over an area of about 1 km2. The UHP metamorphic slab is composed of the compositional layering of the following rock types (Fig. 5): 1. grey–green massive eclogite, 2. dark green foliated and retrograded eclogite, 3. epidote two-mica schist, 4. garnet-biotite gneiss with minor epidote-mica schist, 5. marble with or without eclogite nodules or boudins, 6. dark grey jadeite quartzite and 7. amphibolite. They found coesite and its pseudomorphs in garnet in eclogites, which are distributed throughout the whole metamorphic slab and concluded that the metamorphic slab itself suffered UHP metamorphism. However, they stated that the UHP metamorphic slab was tectonically juxtaposed with the country granitic gneiss at high crustal levels later in the tectonometamorphic cycle of Dabie Orogen, because no UHP evidence was found from the surrounding granitic gneiss and the regional structure in the UHP metamorphic slab is discontinuous to that of the granitic gneiss. Liou et al. (1997) found coesite and its pseudomorphs from jadeite and garnet in the jadeite quartzite from the Shuanghe UHP metamorphic slab, and deduced that the country granitic gneiss might have also suffered UHP metamorphism because systematic disposition and variation of metamorphic grade is observed in the Dabie Shan area, i.e., the UHP terrane to the north and the HP terrane to the south, although no UHP evidence was found from the granitic gneiss itself. The Hefei and Turin group is now re-evaluating the geological relationship between the Shuanghe UHP metamorphic slab and the country granitic gneiss. The reader is referred for the new interpretation to their forthcoming paper (R. Compagnoni and F. Rolfo, pers. commun.). Even though UHP data has been accumulated from many eclogites, the question whether the country granitic gneiss suffered UHP metamorphism along with the

Fig. 4. The equilibrium temperature map in central and southern Dabie Shan. The tectonic boundary follows Carswell et al. (1997) and the distribution of Chenjiahe Formation follows Wang et al. (1990). Po and Pt: Paragonite-bearing eclogite reported by Okay (1993, 1995) and Tabata et al. (1998a), respectively.

112 T. Hirajima & D. Nakamura

The Dabie Shan–Sulu orogen

113

Fig. 5. Geological map of the Shuanghe UHP metamorphic slab and its country rocks (after Cong et al., 1995).

114

T. Hirajima & D. Nakamura

enclosed eclogite or not still remains. Tabata et al. (1998b) gave an answer to this question from the extensive study of inclusions in zircon in the country gneiss of the UHP unit of the Dabie Shan. They extracted zircon grains from the orthogneiss that shows amphibolite facies mineral assemblage in the matrix, and confirmed the presence of tiny coesite and jadeite inclusions in zircon using Raman spectroscopy from several localities of the UHP unit of the Dabie Shan area. Therefore they concluded that the whole UHP unit of the Dabie Shan area subducted to the mantle depths. Further extensive studies on mineral inclusions in zircon were carried out both in the Dabie Shan and the Sulu areas (e.g. Ye et al., 2000b; Liu et al., 2001; Liu et al., 2002). They succeeded to confirm the UHP/HP evidence, such as inclusions of coesite and omphacite in zircon, from country granitic and orthogneisses. These data suggest that a large portion of the country gneiss suffered the UHP metamorphism in the relevant area. Carswell et al. (2000) re-evaluated the P–T conditions at maximum pressure and maximum temperature stages of the country orthogneiss hosting UHP eclogite in the Dabie Shan. The studied orthogneiss mainly consists of garnet with XMn = 0.18–0.45, phengite with Si = 3.20–3.35 p.f.u., zoned epidote with Ps [Fe3+/(Al + Fe3+)]·100 = 38–97, biotite, titanite, two feldspars and quartz, suggesting a typical assemblage of the amphibolite facies. However, certain orthogneiss samples preserve garnet with XCa up to 0.50, rutile inclusions within titanite or epidote and relict phengite inclusions with Si contents up to 3.49 p.f.u., overlapping with the highest value (3.49–3.62) recorded for phengite in samples of undoubted UHP schist. Further mineral composition features, such as A site deficiencies in the highest-Si phengite, negative correlation of Fe/Mg ratio and Si content of phengite, Na in garnet linked to Y + Yb substitution (e.g. Enami et al., 1995) and AlFTi-1O-1 substitution in titanite (e.g. Enami et al., 1993a), are taken to be pointers towards the orthogneisses having experienced a similar metamorphic evolution to the associated UHP schist and eclogite. The adoption of garnet–phengite and garnet–biotite Fe–Mg exchange thermometry and the 5 rutile + 3 grossular + 2 SiO2 + H2O = 5 titanite + 2 zoisite equilibrium barometry between the inclusion and the matrix phases of the orthogneiss. They also succeeded to show that inclusion phases in some orthogneisses gave UHP conditions at peak pressure stage (their Figs. 16 & 17). Then, Carswell et al. (2000) concluded that the orthogneiss may indeed have followed a common subduction related clockwise P–T path with the UHP paragneisses and eclogites, through conditions of ca. 690–715 °C, 3.6 GPa at Pmax to ca. 710–755 °C, 1.8 GPa at Tmax, to extensive recrystallisation and re-equilibration of these ductile orthogneisses at ca. 400–450 °C and 0.6 GPa. Based on these data, Carswell et al. (2000) emphasised that it is no longer necessary to resort to models of tectonic juxtapositioning to explain the spatial association of these Dabie Shan orthogneisses with undoubted UHP lithologies. In their paper, the authors re-evaluate the equilibrium compositional pairs among omphacite, garnet and phengite at Pmax and Tmax stages, mainly in order to depict the clockwise P–T path attained from the thermal modelling of the continental collision belt (e.g. England & Richardson, 1977). In the Sulu area, the occurrence of paraschist, jadeite quartzite and marble are limited to a narrow area (e.g. at Donghai by Zhang et al., 1995a; at Rongcheng by Kato et al., 1997; at Yangzhuang by Ishiwatari et al., 1992 and Nagasaki & Enami, 1998; see

(

)

The Dabie Shan–Sulu orogen

115

Fig. 3) and the several tens of kilometres scale coherency of the supracrustal rocks has not been identified yet. However, the geology of the pre-UHP igneous complex was revealed at Yangkou in the central part of the Sulu area by Hirajima et al. (1993) and Wallis et al. (1997). The Yangkou igneous complex, comprising several hundred square meters, is strongly and heterogeneously deformed at the margin but the original gabbroic and granitic textures are still preserved in the core of the complex, where no extensive syn- and post-UHP deformations were present (Fig. 6). Geochemistry (Ishiwatari et al., 1992; Chen et al., 2002) and intrusion relationships between the basic and acidic rocks suggest that the unit represents a cogenetic magmatic suite, originated by the differentiation of a matic parental magma. The parental magma was derived from the melting of an enriched mantle, and was emplaced into crustal levels during continental extension at ca. 700–800 Ma (Chen et al., 2002). The similarity of zircon SHRIMP dating (ca. 742–756 Ma, Hirajima & Fanning, 1999) extracted from the basic rock, now coesite-bearing eclogite, and the acidic rock also supports the single igneous complex origin of this body, which could have formed in the Yangtze craton. The meta-granitoid still preserves the igneous shape of K-feldspar, quartz, plagioclase, clinopyroxene, orthopyroxene and biotite. However, igneous quartz domains are now completely recrytallised to fine-grained polygonal quartz aggregates, plagioclase to the aggregate of zoisite, kyanite and albite, clinopyroxene to diopside-to-salite in the core and omphacite at the margin, and the coronitic garnet developed between biotite and other minerals. These characters are almost identical to the UHP metagranitoid in Dora-Maira massif (cf. Biino & Compagnoni, 1991; Hirajima et al., 1993). With the development of the postUHP deformations, the UHP metagranitoid was easily transformed to the streaky gneiss that forms the deformed margin of the complex. In the streaky gneiss, the original igneous texture and most of the UHP evidence was completely destroyed. The streaky gneiss is composed of fine-grained quartz, plagioclase, K-feldspar, biotite, epidote and titanite. Rutile surrounded by titanite and a tiny Ca-rich garnet in the matrix of the streaky gneiss are scarce relics of former (U)HP metamorphism, as suggested by Carswell et al. (2000) in the Dabie Shan. The Yangkou UHP complex is surrounded by the Sulu orthogneiss, which is characterised by monotonous foliation and has a typical amphibolite facies mineral assemblage without distinctive UHP evidence. However, relic coesite was found from the core of the amphibolite layer which is intercalated with the Sulu orthogneiss and is about 50 m apart from the Yangkou UHP complex (Fig. 6). Temperature estimated using garnet–clinopyroxene geothermometer (Powell, 1985) for the eclogite in the Yangkou UHP complex varies from 600 to 900 °C at 3.0 Gpa, with a strong positive correlation to the Xjd content of clinopyroxene, which ranges from Xjd = 0.45 to 0.75 (Fig. 14a). The eclogite in the amphibolite layer contains clinopyroxene with Xjd = 0.45–0.55. Comparing the estimated temperatures derived from the eclogite with similar Xjd’s of clinopyroxene in the Yangkou UHP complex, similar temperatures (600–650 °C at 3.0 GPa for Xjd = 0.45–0.55) are obtained from both of them (Hirajima, 1996). More detailed discussion on the P–T estimation in this out crop will be described later. SHRIMP U–Pb zircon ages extracted from the Sulu orthogneiss show 235±16 Ma and 801 ± 45 Ma lower and upper intercept ages, respectively (Hirajima, unpublished data). These data

Fig. 6. Detailed geologic map of the Yangkou UHP metamorphic complex and surrounding Sulu gneiss, mainly following Wallis et al. (1997).

116 T. Hirajima & D. Nakamura

The Dabie Shan–Sulu orogen

117

suggest that the Yangkou complex and its neighbouring area probably suffered UHP metamorphism as a whole, and this outcrop represents one of the best examples showing that deformation and accompanied fluid infiltration at low-pressure conditions pervasively erased the precursor UHP evidence without significant heating (e.g. Hirajima, 1998).

Review of the equilibrium temperature of representative UHP rocks at peak stage The systematic disposition and variation of metamorphic grade (e.g. Liou et al., 1997) and of the equilibrium temperature of eclogite at peak pressure conditions (e.g. Krogh, 1977) are the basic information to determine whether the relevant area suffered the regional eclogite facies metamorphism or not. The northward increase of the metamorphic grade, i.e., from the blueschist unit to the coesite eclogite unit, is established in the Hong’an–Dabie Shan area (Fig. 2), but such kind of metamorphic polarity is not clear in the Sulu area, because most of the Sulu area belongs to the coesite eclogite grade. In the early 1990’s, Wang et al. (1992) reported that the temperature and pressure of the eclogite systematically decrease from about 770 °C of the coesite eclogite in the north to 580 °C of kyanite-quartz eclogite in the south of the southern Dabie Shan terrane, and suggested that the continental crust of the southern Dabie Shan terrane has been subjected to a regional ultrahigh pressure metamorphism as part of a north-dipping subduction zone formed between the Sino–Korean and Yangtze cratons. Enami et al. (1993c) suggested that the equilibrium temperature of the UHP eclogite at the peak gradually decreases from the Sulu area to the Dabie Shan area. If the systematic disposition of the peak eclogiteforming conditions was ascertained, it is a great evidence of the “in situ” model. Some coesite eclogites in the Dabie Shan–Sulu area are closely associated with garnet peridotites, most of which recorded UHP conditions (e.g. 4.0 to 6.0 GPa or more, e.g. Yang et al., 1993) in their development history. The concordance or the discordance of P–T history between the garnet peridotite and the associated eclogite also gave important information for the development history of root zone tectonics below the continent–continent collision zone and for considering the “in situ” vs. “exotic” issue. In this section we compile the P–T histories of the garnet peridotite and other UHP rocks and then discuss the equilibrium temperatures and pressures of the eclogite. P–T history of garnet peridotite and associated UHP minerals Garnet peridotites mainly occur in two metamorphosed mafic–ultramafic complexes at Maowu and Bixiling in the coesite eclogite unit of the Dabie Shan area (e.g. Medaris, 2000; locations are marked in Fig. 4). Field evidence, mineralogical and geochemical data (Okay, 1994; Zhang & Liou, 1994; Zhang et al., 1994; Zhang et al., 1995b; Cong et al., 1996; Liou & Zhang, 1998; Zhang et al., 2000a; Jahn et al., 2001) suggest that these complexes were emplaced into quartzofeldspathic gneisses prior to Triassic UHP metamorphism and that they originated as low-pressure crystal cumulates. The Maowu mafic–ultramafic complex is a small body (250 × 100 m2) and exhibits clear magma chamber processes, e.g. characterised by three apparent rhythmic layering

118

T. Hirajima & D. Nakamura

Fig. 7. Geological sketch map of (a) the Maowu complex (Okay, 1994) and (b) the Bixiling complex (Zhang et al., 1995b).

The Dabie Shan–Sulu orogen

119

(e.g. Jahn et al., 2001). The Maowu complex consists of layers, lenses and pods of garnet orthopyroxenite, garnet clinopyroxenite, harzburgite, eclogite and omphacitite, which are intercalated on a scale of 5 cm to 1.6 m. In contrast to Bixiling, olivine is uncommon at Maowu, and the most abundant ultramafic rock type is orthopyroxenite (Fig. 7a). A zircon U–Pb age from eclogite and mineral Sm–Nd isochrons for a garnet websterite and a garnet clinopyroxenite in Maowu complex constrain the time of UHP metamorphism at 220–230 Ma (e.g. Rowley et al., 1997; Jahn et al., 2001). Pre-UHP mineral history was clearly depicted by Okay (1994) using inclusion phases in garnet. Okay (1994) observed low-pressure assemblages in garnets of a garnet orthopyroxenite as inclusions of chlorite, orthopyroxene, sapphirine, gedrite, hornblende, talc, rutile, phlogopite, corundum and zircon. He interpreted that these minerals were the result of Precambrian granulite facies metamorphism (740 ± 30 °C and 0.2–0.6 GPa) predating the Triassic UHP metamorphism. HP/UHP mineral assemblages are orthopyroxene (0.08 to 0.16 wt% Al2O3) + garnet + chlorite ± clinopyroxene ± Ti-clinohumite (0.8 to 1.3 wt% F) ± magnesite in orthopyroxenite, and olivine + orthopyroxene + garnet ± Ti-clinohumite (0.8 to 1.3 wt% F) ± magnesite ± chlorite in harzburgite. Liou & Zhang (1998) evaluated the UHP conditions of Maowu ultramafics as 650–780°C at 4.0 GPa mainly by Grt–Cpx (Ellis & Green, 1979; Powell, 1985: Krogh, 1988) and Grt–Opx (Harley, 1984a) geothermometers, and 4.5–6.5 GPa by Grt–Opx geobarometer (Brey & Köhler, 1990) (Fig. 8). Subsequent minor amphibolite facies overprint took place at P < 1.5 GPa and 650 °C. The Bixiling complex is a 1.5 km2 tectonic block within biotite gneiss and consists mostly of layered eclogite that includes about 20 lenses of ultramafic rocks, measuring from 50 to 300 m in length and from 5 to 50 m in width (Fig. 7b). The ultramafic rocks are garnet lherzolite, garnet wehrlite and garnet websterite, which contain HP/UHP mineral assemblages of olivine + orthopyroxene (0.10 to 0.22 wt% Al2O3) + clinopyroxene + garnet + ilmenite (after Ti-clinohumite), and orthopyroxene + clinopyroxene + garnet + magnesite + Ti-clinohumite (0.05 to 0.2 wt% F). Both eclogites and meta-ultramafic rocks have undergone multistage metamorphism. Zhang et al. (1995b) evaluated the peak P–T conditions of the eclogite as an average of 630 °C (range of 540–725 °C) at 3.0 GPa by the Grt–Cpx geothermometer of Powell (1985) and Krogh (1988), using rim (most cases) and core (a few cases) compositions of adjacent minerals. Meta-ultramafic rocks showed P–T conditions of 700–800 °C and 4.7–6.7 GPa by the two-pyroxene geothermometer (Wood & Banno, 1973), two Fe–Mg exchange thermometers (Opx–Grt of Lee & Ganguly, 1988 and Grt–Cpx of Powell, 1985) and the Grt–Opx geobarometer of Brey & Köhler (1990) using the rim compositions of adjacent minerals (Fig. 8). Carswell et al. (1997) estimated the peak P–T conditions for the Bixiling eclogite as 3.3–3.8 GPa and 830–870 °C (see in detail in the next section). There is a significant disagreement between two research groups on the temperature estimation at peak pressure stage of the eclogite as well as on the peak pressure of the eclogite and the metaultramafic rocks. The pressures estimated by the Grt–Opx barometer generally show relatively high values compared with those estimated by other barometers (e.g. garnet–clinopyroxene–phengite barometer; Waters & Martin, 1993) for UHP rocks. This can be due to one or a combination of the following reasons; (i) analytical error because

120

T. Hirajima & D. Nakamura

Fig. 8. P–T conditions of garnet peridotite and eclogite in the Maowu and Bixiling complexes in the Dabie Shan area. Data of the Maowu complex are from Okay (1994) and Liou & Zhang (1998), and of the Bixiling complex from Carswell et al. (1997) and Zhang et al. (1995b).

of extremely low Al content of orthopyroxene in UHP rocks, (ii) selection of calibrations on Grt–Opx barometer also cause large variation in estimated pressures (e.g. Hiramatsu et al., 1995), (iii) difficulty in judging the equilibrium compositional pair between garnet and orthopyroxene (e.g. Cuthbert et al., 2000), for example, incomplete Al diffusion during orthopyroxene formation after olivine may preserve low Al composition of orthopyroxene that was not in equilibrium with garnet (e.g. Mørk, 1985). In the Sulu region, garnet peridotite and pyroxenites are found from seven areas: Weihai, Rongcheng (locality of Chijiadian), Yangkou, Rizhao (locality of Hujialing), Junan, Yangshuang (locality of Zhoubin) and Donghai (localities of Zhimafang, Mengzhong and Jiangshuang) from north to south (Fig. 3; e.g. Yang et al., 1993; Zhang et al., 1994; Hiramatsu & Hirajima, 1995; Hiramatsu et al., 1995; Cong et al., 1996; Yang & Jahn, 2000; Yoshida et al., 2001). The ultramafic bodies generally have small dimensions but locally reach kilometre scale at Hujialing (e.g. Yang, 1991; Hiramatsu & Hirajima, 1995). They have concordant to discordant contacts with surrounding

The Dabie Shan–Sulu orogen

121

quartzofeldspathic gneisses. The garnet peridotites in the Sulu area also show the multiple P–T equilibria identified from the chemical variations along with changes in microstructure of principal constituent minerals (e.g. the porphyroblasts and the matrix phases) and the inclusion phases in them. The UHP stage mineral assemblage is characterised by the occurrence of garnet and orthopyroxene (0.11 to 0.30 wt% Al2O3) along with olivine + clinopyroxene ± phlogopite. The extremely low Al2O3 content in Opx gave pressures of diamond stability ranging from 4.0 to 7.0 GPa at 700–1000 °C (see the compilation of Medaris, 2000). Low P inclusion minerals in garnet are found from several localities in the Sulu area; e.g. chromite, chlorite, hornblende, Na-gedrite, Na-phlogopite, talc, spinel and pyrite at Zhimafang, Donghai County (Yang & Jahn, 2000), spinel, amphibole and Al-rich clinopyroxene at Hujialing (Yang, 1991; Hiramatsu & Hirajima, 1995). The existence of these low P inclusion minerals suggests many of the garnet peridotites in the Sulu area subducted during the Triassic continental collision stage. The mineral assemblage associated with garnet and low-Al orthopyroxene in the peridotite and pyroxenite in the UHP region is generally considered to represent a UHP stage chemical equilibrium. Yang & Jahn (2000), recently, proposed two UHP equilibrium conditions for garnet-bearing assemblages for the garnet peridotite at Zhimafang; higher temperature and pressure stage (1000 °C and > 5.1 GPa) defined by the Mg-rich core of porphyroblastic garnet and orthopyroxene, and lower temperature and pressure stage (760 °C and 4.2 GPa) defined by the matrix minerals included in the rims of porphyroblasts (Fig. 9a). The second UHP stage is interpreted as a result of metasomatism of the peridotites by a SiO2-rich melt at UHP conditions. Yang & Jahn (2000) proposed an isothermal decompression path or decompression with a slight temperature decrease path from the 2nd UHP stage to the lower crust (Fig. 9a). Yoshida et al. (2001) also identified two equilibrium stages for garnet-bearing assemblages for the Yangkou garnet peridotite based on the chemical variations along with changes in microstructure of principal constituent minerals: Stage I, a primary garnet lherzolite stage represented by coarse-grained (a few mm size) porphyroclastic aluminous pyroxene + chromian spinel ± garnet, and Stage II, a UHP stage defined by fine-grained matrix phases (0.1–0.3 mm size) of garnet + extremely low-Al orthopyroxene + high-Na clinopyroxene + chromite, and a subsequent Stage III, a medium pressure stage defined by fine-grained mineral aggregates (< 0.1–0.2 mm size) mainly composed of aluminous spinel + high-Al orthopyroxene in the matrix, and Stage IV, an amphibolite to greenschist facies stage defined by poikiloblastic amphibole. Orthopyroxene– clinopyroxene thermometry (Bertrand & Mercier, 1985) and an empirical spinel barometer (O’Neill, 1981) give temperatures of around 820–830 °C and pressures of 2.6 GPa for porphyroclasts of Stage I. Garnet–orthopyroxene (Harley, 1984a,b), garnet– clinopyroxene (Powell, 1985) and empirical spinel (O’Neill, 1981) geothermobarometers give relatively uniform P–T conditions for the matrix garnet–orthopyroxene– clinopyroxene–chromite assemblage of Stage II (~ 730–760 °C and 3.6–4.1 GPa). Aluminous spinel–olivine (Fabriès, 1979) pairs in the aggregates give ca. 650–700 °C at less than 1.5 GPa for Stage III. The P–T conditions of the first garnet stable stage of Yangkou garnet peridotite is quite different from that of Zhimafang complex, but the P–T

122

T. Hirajima & D. Nakamura Fig. 9. P–T paths of representative garnet peridotites in the Sulu area. (a) Zhimafang (Yang & Jahn, 2000), (b) Yangkou (Yoshida et al., 2001).

conditions of the second garnet stable stage is quite similar in both of them. The peak P–T conditions of the Yangkou eclogite, reevaluated by the combination of garnet–clinopyroxene geothermometer and garnet–omphacite– kyanite–coesite geobarometer as shown in the next section (cf. Fig. 12), are almost identical to the P–T conditions of Stage II defined from the Yangkou garnet peridotite. Furthermore, both garnet peridotite bodies show a similar decompression path, an isothermal decompression or decompression with a slight temperature decrease. To produce an isothermal decompression path, the ascending UHP body requires either thermal isolation from (or even heating by) the wall rocks. To create such a situation, rapid exhumation and large volumes of ascending hot materials are necessary. Nakamura & Hirajima (2000) estimated that the size of the exhumed mass required to prevent the loss of heat to or addition of heat from the wall rock should be > 10 km on the basis of petrological study of Rongcheng eclogites and with an assumed exhumation rate of 20 mm/year. This result requires that the small Yangkou complex (see Fig. 6) and the Zhimafang peridotite body should have been exhumed together with associated UHP rocks and the surrounding country rock gneiss. Recent transmission electron microscopy works (e.g. Su et al., 2001; Zhang et al., 2002) reveal that a certain diopside and orthoenstatite from garnet pyroxenites in the Dabie Shan–Sulu area contain clinoenstatite lamellae, which is interpreted to have formed either by inversion from orthoenstatite or by the transformation from high pressure C2/c clinoenstatite/clinopyroxene during decompression. The latter interpretation may

The Dabie Shan–Sulu orogen

123

suggest that the garnet pyroxenites containing clinoenstatite lamellae were once equilibrated under a significantly higher pressure (> 7–8 GPa; e.g. Ulmer & Stalder, 2001) than that evaluated from the conventional geobarometers depending on the Al content in orthopyroxene. Such high pressure demands that the coexisting garnet in the garnet pyroxenites be majoritic in composition (e.g. Akaogi & Akimoto, 1977; Irifune, 1987). Most of the reported compositions of garnet from the garnet-bearing ultramafics, however, have a stoichiometric formula with R3+/R4+ cation ratio around
, although Ye et al. (2000a) reported “subsilicic calculated original garnet composition”, derived from the compositions of the host garnet from an eclogite at Yangkou, and claimed a maximum pressure of Yangkou eclogite to be greater than 7 GPa. However, Yoshida et al. (2001) have not observed similar inclusion-rich porphyroclastic garnet in either eclogite or ultramafic rocks at Yangkou. After the first report on the finding of diamond from the Dabie Shan metamorphic rocks by Xu et al. (1992), the identification of the peak pressure for the Dabie Shan metamorphic rocks is one of the hot targets for geologists. As compiled in the next section, most of the peak P–T conditions estimated from the mineralogical data in eclogites fall in the graphite stability field (Table 1, 3.0–3.8 GPa in average). To explain the “lower” peak pressure than the diamond stability field, Okay (1993) considered that some assemblages of the hot eclogite terrain suffered re-equilibration during the uplift. On the other hand, the very low Al content of orthopyroxene coexisting with garnet, olivine and clinopyroxene in peridotites gave diamond stability conditions at peak pressure stage as described above (4.0–7.0 GPa). These data and the finding of C2/c clinoenstatite lamellae in host diopside and orthoenstatite by Su et al. (2001) and Zhang et al. (2002) may suggest that some garnet peridotite equilibrated under the diamond stability field once in their P–T history. The intensive search for mineral inclusions in garnet and zircon from both UHP eclogites and country gneisses was attempted by Liou & Zhang (2001) but they failed to find a positive identification of diamond. They, however, explained the lack of the occurrence of diamond in the Dabie Shan–Sulu area due to the lack or low content of C–H–O fluid with appropriate XCO and fO conditions, as such fluids essentially control the precipitation of microdiamond in gneissic rocks and dolomitic marbles, as suggested by the case studies in Kokchetav and Erzgebirge. 2

3

2

2

Peak P–T conditions of eclogite Many garnet peridotites in the Dabie Shan–Sulu area recorded UHP equilibrium conditions during their development history as mentioned above. However, their occurrence is limited at about ten localities in the huge Dabie Shan–Sulu area (Fig. 1), and their outcrop density, lower than that of eclogite, is not convenient to discuss the systematic disposition and variation of metamorphic grade at peak pressure stage. However, the eclogite, less common than the country orthogneiss, is distributed through the Dabie Shan–Sulu area and so the comparison of peak P–T conditions of the eclogites will give fundamental information defining whether the systematic disposition and/or variation of metamorphic grade at peak pressure stage exist in the relevant area or not.

average

880 845 800 680 801

870

830

22 23

36 34 40 34

30

30 31

37

33 38

750

780 -

Huangzhen east 272

Shima south 568

620

-

Dongfeng east 585 690 (at 19kb)

Shuanghe 251 Shuanghe 250

Wumiao east 229 233 230 575

720

730

760 690

690 (at 19kb) 740

Changpu 218 Changpu 223

577L

Okay (1993, 1995) T(°C, P85) T(°C, K88) Xgrs < 0.35 Xgrs > 0.35

Rock No.

Cp

Pg-Cp

C

C

C

Pg Cp

NB:

Shima 224 223 221 172

569

629 589 616 639 563 592 789

589 595 590

737 631 829

825 651 744 639

764

Tabata et al. (1998a) T(°C, P85) T(°C, P85) Cal-free Cal-bearing

Dongfeng east 235

Wumiao west 308 309 311 312 Wumiao east 336 328 302 301 290 258 248

Changpu west 407 409 303

Rock No.

C-Cal C-Cal Cal C-Pg-Cal

Pg Pg C

C Cal

C

Pg

NB:

P [kbar]*: estimated by the combination of Grt-Cpx thermometry and Phn-Grt-Cpx barometry of Waters & Martin (1993) P85: Grt-Cpx geothermometer of Powell (1985), K88: Grt-Cpx geothermometer of Krogh (1988); C: coesite, Cp: quartz pseudomorphs after coesite, Pg: paragonite in the matrix, Cal: carbonate.

Huangzhen

600 610

730

Dongfeng

Shima

680 740

870

Carswell et al. (1997) T(°C, P85) T(°C, K88) P(kb)* Xgrs < 0.35 Xgrs > 0.35

Shuanghe

Guanjialing

Bixiling

Locality

Table 1. Re-evaluated equilibrium temperatures for the coesite/quartz eclogite in the Dabie Shan area

124 T. Hirajima & D. Nakamura

The Dabie Shan–Sulu orogen

125

The temperature conditions of the eclogites are commonly obtained by several formulations of Grt–Cpx geothermometer and there are systematic discrepancies among the estimated temperatures obtained from different formulations of the Grt–Cpx geothermometer. In the following sections we will add some statement for the Grt–Cpx geothermometer and geobarometers that can be applied to eclogites. The character of Grt–Cpx geothermometer Formulations of Ellis & Green (1979), Powell (1985) and Krogh (1988) for the garnet–clinopyroxene geothermometer are favoured and frequently used especially for estimating the equilibrium temperatures of eclogite. However, many formulations have been carried out and proposed for the garnet–clinopyroxene geothermometer (cf. Krogh Ravna & Paquin, 2003 in this volume). The above three formulations are mainly based on the same experimental data (Råheim & Green, 1974; Mori & Green, 1978; Ellis & Green, 1979), and hence these formulations yield more or less similar temperatures. The synthetic experiments used in these formulations are not of the reversal type, and hence reversal experiments for Fe–Mg exchange reaction between garnet and clinopyroxene were carried out by Pattison & Newton (1989). Green & Adam (1991) performed synthetic experiments and tested the reliability of the methods of Ellis & Green (1979) and Pattison & Newton (1989), and they suggested that the method of Pattison & Newton (1989) yielded significantly lower temperatures than the experimental temperatures of Green & Adam (1991). At least in high pressure conditions (3.0 GPa), the Ellis & Green (1979) formulation yielded moderate temperatures compared to the experiments by Green & Adam (1991). Furthermore, Perkins & Vielzeuf (1992) claimed that combination of garnet–olivine (Hackler & Wood, 1989) and olivine–clinopyroxene (Perkins & Vielzeuf, 1992) reversal experiments around 1.0 GPa gave larger KD for the garnet–clinopyroxene pair than Pattison & Newton’s (1989) data. Therefore, Berman et al. (1995) re-evaluated Pattison & Newton’s (1989) reversal experiments considering compositional variations in the run products and proposed a new formulation based on the reversal experiments. However, applications of Berman et al. (1995) to eclogite also yield significantly lower temperatures than experimental temperatures (Nakamura et al., 2001). Thus, formulations of Pattison & Newton (1989) and Berman et al. (1995) are not applicable at least to eclogite. In the equilibrium experiments using natural fine-grained eclogite as starting materials (Nakamura et al., 2001), formulations of Ellis & Green (1979), Ganguly (1979), Powell (1985), Krogh (1988), Ganguly et al. (1996) and Krogh Ravna (2000) gave temperatures compatible with the experimental temperatures (1100–1200 °C). Ganguly (1979) carried out regression analysis of equilibrium experimental data (Wood, 1976) at high temperatures (1100–1400 °C) and extrapolated it to lower temperatures using thermochemical data, and Ganguly et al. (1996) revised the non-ideal mixing term of garnet. Krogh Ravna (2000) carried out regression analysis of total 311 experimental data and 49 data of natural Mn-rich assemblages and proposed an empirical formulation. These formulations show no significant divergence at high temperatures, but there is significant difference at low temperatures (< 1000 °C). For example, we compare following three formulations, Ganguly (1979), Powell (1985) and Krogh Ravna (2000) in temperature versus KD diagram (Fig. 10). Krogh Ravna (2000) yields slightly lower

T. Hirajima & D. Nakamura

126

temperatures than Powell (1985) at low temperature conditions. On the other hand, Ganguly (1979) gives significantly higher temperatures than Powell (1985). To examine the reason why such a difference exists is beyond the scope of this study. As three formulations of Ellis & Green (1979), Powell (1985) and Krogh (1988) are frequently used especially for estimating the equilibrium temperatures of eclogite, we show the systematic discrepancy in temperatures estimated by these three formulations. The systematic discrepancy is mainly caused by the different expression of the grossular component (XGrs) effect to the thermometry. For example, the application of Powell (1985) systematically gives lower temperature than that of Ellis & Green (1979) by about 20–30 °C (Fig. 11) through the wide range of KD and XGrs. The temperature obtained from Powell (1985), on the other hand, is systematically 5–30 °C higher than that obtained from Krogh (1988) for the eclogite with less calcic garnet (XGrs < 0.4), but Powell’s (1985) formulation gives a significantly higher temperature for the eclogite with calcic garnet (XGrs > 0.4); e.g. Powell (1985) gives 100 °C higher than Krogh (1988) for KD = 10.5 and XGrs = 0.55 (Fig. 11). Carswell et al. (1997) tested whether temperatures obtained from Powell (1985) or from Krogh (1988) are the most consistent with temperatures obtained by the garnet–phengite geothermometer (Green & Hellman, 1982) over the grossular range between 0.26 and 0.53 and concluded that Powell’s (1985) formulation yields significantly overestimated temperatures for samples containing the calcic garnet (XGrs > 0.35). We basically follow their suggestion; i.e., temperatures are estimated using Powell (1985) for the eclogite with XGrs < 0.35 and using Krogh (1988) for the eclogite with XGrs > 0.35. The geobarometry of eclogite The pressure estimation of eclogite-stage equilibrium is rather difficult than the temperature estimation. The garnet–clinopyroxene–phengite barometer (Waters & Martin, 1993, 1996) is one of potential tools for defining pressures of eclogites, as about 40% of eclogites contain phengite in the Sulu area (Nakamura, 2003). We also estimate the peak pressure of the representative kyanite-bearing eclogites using a garnet–clinopyroxene– kyanite–coesite (or quartz) barometer, proposed by Nakamura & Banno (1997). We add some statement for the latter geobarometer. The garnet–clinopyroxene–kyanite–coesite (or quartz) barometer is based on a reaction: pyrope + grossular + 2 coesite (or quartz) = 2 kyanite + 3 diopside. In this study, thermodynamic data set of Holland & Powell (1998) is used for this barometry. Activity calculations follow Nakamura & Banno (1997). Thermodynamic equilibrium equation for this reaction can be written as follows.

pyrope + grossular + 2coesite (or quartz) = 2kyanite + 3diopside. i is chemical potential of phase component i and can be written as follows. i = Gi + RT ln ai, Gi is Gibbs free energy of phase component i, R is the gas constant (0.0083143 kJ/K mol), T is temperature (K), and ai is activity of phase component i. Following Holland & Powell (1998), Gi is calculated as follows:

The Dabie Shan–Sulu orogen

127

Fig. 10. Comparison of three formulations (Ganguly, 1979; Powell, 1985; Krogh Ravna, 2000) for garnet–clinopyroxene thermometer. Relationships between temperature and distribution coefficient (KD) are Grt shown at X Ca (Ca/Fe + Mg + Ca) = 0.20 and 3.0 GPa. For Krogh Ravna (2000), Mg number of garnet (Mg#Grt) is fixed as 0.30. T

T

298

298

Gi =
ƒH+SmaxTc[(Q298)2–(Q298)6/3]–T[S+Smax(Q298)2]+ CpdT–T Cp/T)dT+

T[V+V(Q298)2][l+a°(T–298)–20a°(T–298)][(1+4P/T)¾–1]/3+ Smax[(T–TC )Q2+TC Q6/3] P

P

where T

CpdT = a(T–298)+0.5b(T2–2982)–c(1/T–1/298)+2d(T–298)

,

298

T

T Cp/T)dT = a(lnT–ln298)+b(T–298)–0.5c(1/T2–1/2982)–2d(1/T–1/298), 298

T =  [1 – 1.5 · 10–4 (T – 298)], Q298 = (1 – 298/TC)¼, Q = (1 – T/TCP)¼ TCP = (Vmax/Smax)P + TC.

128

T. Hirajima & D. Nakamura

Fig. 11. Isotherms based on the alternative formulations of Ellis & Green (1979; EG(79)), Powell (1985; P(85)) and Krogh (1988; K(88)) in XGrs (= Ca/(Ca + Fe + Mg + Mn)) in garnet versus KD (for Fe2+–Mg2+ partitioning between garnet and clinopyroxene) diagram. Inset data from Okay (1993, 1995). For further details see text.

In Table 5 of Holland & Powell (1998), terms of S and Smax must be multiplied by 10–3, and b and a° must be multiplied by 10–5. Following Nakamura & Banno (1997), activity of diopside (adi) can be calculated as follows: adi = [Xdi {XCaa (2 XCa – XCaa)} / XCa] exp [{– 18960 (4 XCa XCaa – 2 XCaa 2 – 2 XCa2) + + 24080 (1 – 2 XCa + XCa2)} / 8.3143 T], where Xdi = Mg/(Fetotal + Mg + Al) and XCa = Ca/(Na +Ca). XCaa is dependent on temperature, and the magnitude of XCaa should be independently determined using the following equation: – 37920 (–2 XCa + 2 XCaa) + 8.3143T {ln XCaa – ln (1 – XCaa) – ln (2 XCa – XCaa) + + ln (1 – 2 XCa + XCaa)} = 0. To calculate activity of diopside, two-step calculation is necessary in this model. Firstly, the degree of ordering (XCaa) should be determined at each composition (XCa) and temperature with this equation. This equation cannot be rewritten as XCaa = ƒ (XCa, T), so

The Dabie Shan–Sulu orogen

129

we have to look for the value of XCaa to satisfy this equation while changing the value of XCaa. For this calculation, XCaa should be as follows: 0 < XCaa < 2XCa and 2XCa – 1 < XCaa < 1. In addition, three solutions of XCaa are present in this equation when T < Tc (critical temperature at each composition). XCaa = XCa always satisfies this equation, and so one of the other two solutions should be selected for activity calculation. Thus, we should determine the XCaa value at several temperatures before activity calculation. Then, activities of pyrope and grossular components (aprp and agrs) are calculated as follows: aprp =Xprp3 exp{4528 Xgrs (1 – Xprp)/T}, agrs = Xgrs3 exp{4528 (1 – Xgrs) Xprp/T}, where Xprp = Mg/(Fe + Mn + Mg + Ca) and Xgrs = Ca/(Fe + Mn + Mg + Ca). Thus, we can draw an equilibrium P–T curve for the reaction pyrope + grossular + 2 coesite (or quartz) = 2 kyanite + 3 diopside. Combination of this P–T curve with the garnet–clinopyroxene thermometer gives P–T conditions at peak stage of the kyanitebearing eclogite. Compilation result of the Dabie Shan eclogite We intended to compile the peak P–T conditions of the eclogite using the combination of the garnet–clinopyroxene geothermometer and the garnet–clinopyroxene–kyanite– coesite (or quartz) barometer. We adopted this combination in the case of the Sulu eclogites as described in the next section. However, the kyanite-bearing eclogite is scarce in the Dabie Shan area and we do not have original petrological data for eclogites in the Dabie Shan. Therefore, we compile and examine the peak P–T conditions in Dabie Shan eclogite from the literatures. The detail is mentioned below. After the pioneering work of Wang et al. (1990; 1992) in the Dabie Shan, Okay (1993, 1995), Carswell et al. (1997) and Tabata et al. (1998a) carried out the systematic estimation of the peak eclogite-forming conditions combining of geothermometers (Grt–Cpx: Ellis & Green, 1979, EG79; Powell, 1985, P85; Krogh, 1988, K88, Grt–Phn: Green & Hellman, 1982) and geobarometers (e.g. celadonite + pyrope + grossular = muscovite + diopside reaction; Waters & Martin, 1993, zoisite/clinozoisite = grossular + kyanite + quartz/coesite + H2O; Okay, 1995). Our compiled results of their estimated peak equilibrium temperatures are shown in Table 1. The equilibrium temperatures of Carswell et al. (1997) were calculated at pressures obtained by geobarometry of celadonite + pyrope + grossular = muscovite + diopside reaction (Waters & Martin, 1993). Other equilibrium temperatures in Table 1 were calculated at 3.0 GPa. The difference in pressure between the assumed 3.0 GPa and that obtained by Carswell et al. (1997) cause small differences in estimated temperatures (less than 20 °C), which is of little influence on discussing the thermal structure of the coesite eclogite unit. Okay (1993, 1995) estimated the peak equilibrium temperatures by the Ellis & Green (1977) formulation at 3.0 GPa. We recalculated them by the Powell (1985) formulation for the eclogite with XGrs < 0.35 and the Krogh (1988) formulation for the eclogite with XGrs > 0.35, using the KD–XGrs relationship shown in Figure 7 of Okay

130

T. Hirajima & D. Nakamura

(1993). When Okay (1993, 1995) reported multiple equilibrium temperatures for the same sample number, we report the average value for each sample. Tabata et al. (1998a) estimated the peak equilibrium temperatures for 19 eclogites by Ellis & Green (1979), Powell (1985) and Berman et al. (1995) at 3.0 GPa. According to the description in the text and Figure 11 of Tabata et al. (1998a), carbonate and calcic garnet with XGrs ranging from 0.36 to 0.41 in average seems to be contained in five eclogites and less calcic garnet with XGrs ranging from 0.15 to 0.33 in average to be in the remaining 14 eclogites. Therefore, in our Table 1 we show, following equilibrium temperatures reported by Tabata et al. (1998a), the equilibrium temperatures calculated with the Powell (1985) formulation for the 14 eclogites without carbonate and those calculated with the Krogh (1988) formulation for the 5 eclogites with carbonate. Ferric/ferrous estimation from microprobe data is one of main factor to scatter the equilibrium temperatures. Generally, ferric/ferrous ratio of clinopyroxene is commonly estimated from one of following three methods: 1) based on four cations and the charge balance concentration, 2) Fe3+ = Na – (Altotal + Cr) and 3) total iron as Fe2+. Carswell et al. (1997) and Okay (1993) adopted the second method and Tabata et al. (1998a) used the first method. As far as we examined, the second method makes the least scattered value among the three methods and the first method generally makes the most scattered result, as all errors of the electron microprobe analysis were propagated to the estimation of the ferric ratio (e.g. Hirajima, 1996; Carswell et al., 1997). Carswell et al. (1997) took special care in selecting the equilibrium composition pair of garnet and clinopyroxene for the thermometry, i.e., the most calcic garnet and the most jadeite-rich clinopyroxene. This is a reasonable pairing method for estimating the peak P–T conditions of eclogite, if we can regard that the peak pressure stage is coeval with the peak temperature stage in UHP eclogite. Carswell et al. (2000), however, slightly modified their pairing policy following the idea that the peak pressure stage predated the peak temperature stage based on the expected result obtained from thermal modelling under the continent–collision zone (e.g. England & Richardson, 1977), i.e., the jadeite-richest and Si-richest compositions in clinopyroxene and phengite, respectively, and the garnet composition with the maximum aprpa2grs for the peak pressure stage, and highest Fe/Mg ratios of omphacite and phengite and lowest Fe/Mg ratio of garnet for the peak temperature stage. It is almost impossible to identify the relevant phase compositions during the peak pressure and peak temperature stages from the literature, therefore we compile the data following the former idea. The pairing policy of Okay (1993) and Tabata et al. (1998a) is unclear. As Okay (1993) described that garnet and omphacite do not show any clear zoning in some eclogite, his estimated KD may reflect the peak P–T conditions. Based on the calculation method of KD, we regard the data of Carswell et al. (1997) and Okay (1993) as a comparable data set to discuss the peak P–T conditions in the Dabie Shan area. Okay (1995) reported the occurrence of paragonite in textural equilibrium with garnet, omphacite and kyanite in the matrix of coesite-bearing eclogites and obtained the equilibrium conditions of 1.9 GPa and 710 ± 40 °C for the paragonite–omphacite–garnet association by the combination of Pg87–Jd48–Ky–Qtz isopleth and the Ellis & Green (1979) formulation, and concluded that the matrix minerals in the UHP eclogite have

The Dabie Shan–Sulu orogen

131

recrystallised during the early decompression stage and that they do not represent the peak UHP assemblages. To test whether his statement is valid or not is beyond the scope of this chapter. As the peak temperature of the paragonite-bearing eclogite (690 °C at 1.9 Gpa, recalculated by the Powell, 1985 formulation) is a bit lower than that of the coesite eclogite from which paragonite was not detected in the matrix (690–780 °C at 3.0 GPa, see column of Okay, 1993, 1995 in Table 1), we tentatively accept his conclusion, i.e., the equilibrium temperature obtained from eclogite with matrix paragonite represents a temperature condition at early decompression stage. Among the compiled data in Table 1, 2 samples of Okay (1993, 1995) and 4 samples of Tabata et al. (1998a) are in this case. Temperatures (620–640 °C at 3.0 GPa) obtained from eclogites with matrix paragonite in Tabata et al. (1998a), however, do not show significant difference from those of 11 eclogites with less calcic garnet (560–830 °C at 3.0 GPa). Among the temperatures of the 11 eclogites, some of them containing coesite, 8 samples are less than 630 °C. Recent synthetic experiments in basalt + H2O system by Schmidt & Poli (1998) indicate that lawsonite, instead of zoisite + kyanite + coesite assemblage, is a stable phase around 600–650 °C at 3.0–4.0 GPa and zoisite/epidote, paragonite and amphibole are stable at 1.5–3.0 GPa around 600–650 °C. The equilibrium temperatures reported by Tabata et al. (1998a) suggest that lawsonite is a stable phase under H2O saturated conditions at UHP stage, but they reported epidote and amphibole as coexisting phases with coesite. These data suggest that the modification of the peak compositions of garnet and clinopyroxene may have taken place during the early exhumation stage and/or the compositions used for the temperature estimation by Tabata et al. (1998a) do not represent equilibrium pairs at the peak stage. In the following discussion on the consideration of the peak P–T conditions of the Dabie eclogite, we regard 12 data of Carswell et al. (1997), 7 of Okay (1993, 1995) and 3 (* in Table 1) of Tabata et al. (1998a) compiled in Table 1 as plausible peak equilibrium temperatures. These data show: 1. the quartz eclogites distributed around Huangzhen give 600–620 °C and 2.2–2.3 Gpa; 2. the coesite eclogites at Shima give 680–880 °C (800 °C in average) and 3.4–4.0 GPa, at Dongfeng 730 °C and 3.0 GPa, at Shuanghe 680–750 °C (730 °C in average) and 3.0–3.1 GPa, at Wumiao 690–790 °C (755 °C in average) and 3.0 GPa, at Guanjialing 870 °C and 3.7 GPa, at Bixiling 720–870 °C (810 °C in average) and 3.3–3.8 GPa, and at Changpu 740–830 °C (770 °C in average) and 3.0 GPa in the order of the northward direction. Okay (1993) and Carswell et al. (1997) concluded that the central Dabie eclogites were formed under “hot” and coesite-stable P–T conditions and the southern Dabie eclogites were formed under “cold” and quartz-stable P–T conditions. The coesite eclogite terrane was considered to be structurally overlain by the southern quartz eclogite terrane at a major dipping shear zone along which late orogenic extensional collapse appears to have eliminated at least 20 km of crustal section. Okay (1993) and Carswell et al. (1997) did not support the earlier idea of the regional P–T gradient marked by decreasing peak conditions of both P and T southwestwards by Wang et al. (1990, 1992). Other petrological data, e.g. the lack of coesite in the southern Dabie eclogites and the significant difference in garnet zoning pattern between both eclogite

132

T. Hirajima & D. Nakamura

terranes, i.e., distinctive idioblastic, growth-zoned garnets containing abundant small single grains of quartz in the southern Dabie eclogites, and large garnet with homogeneous core and zoned rim in the central Dabie eclogites, support the conclusion of Okay (1993) and Carswell et al. (1997). In the coesite eclogite unit, the peak temperatures obtained around Shuanghe and Wumiao are a bit lower (20–80 °C) than those obtained in the northern (Bixiling and Changpu) and southern (Shima) parts. These data are concordant to the indication by Carswell et al. (1997), a southeastward trend of declining peak temperatures and pressures along the northern transect from Bixiling to the Shuanghe locality (Fig. 4). The equilibrium temperatures obtained from multiple eclogites collected at one locality often scatter more than 100–200°C, even though there is no clear reason, as pointed out by Carswell et al. (1997) at the Shima locality. The number of the equilibrium temperatures obtained from the quartz eclogite terrane is not sufficient to statistically discuss whether there is a continuous P–T gradient between the quartz eclogite terrane and the coesite eclogite terrane or not. The preliminary pressure estimation by the garnet–clinopyroxene–phengite barometer using the database of Holland & Powell (1998) gave about 3.0 GPa for the phengite-bearing Huangzhen eclogite. Though we do not deny the conclusions of Okay (1993) and Carswell et al. (1997), the acquisition of further petrological data, especially for indicating the peak pressure, is necessary in order to strengthen their idea further. Results from the Sulu eclogite In this section, we estimated peak P–T conditions of eclogite from the Sulu area using mainly our own published and unpublished data and partly the literature data. We estimate peak P–T conditions for eclogites collected at the following localities: Qinglongshan and Chizhuang in Donghai County, Taohang in Zhucheng County, Yangkou in Qingdao County, and Chijiadian, Xianguling and Tengjiaji in Rongcheng County. We applied geothermobarometers that were mentioned in the previous section to these eclogites; i.e., the garnet–clinopyroxene thermometer (Powell, 1985 or Krogh, 1988), garnet–clinopyroxene–phengite barometry (Waters & Martin, 1996) and garnet–clinopyroxene–kyanite–coesite (or quartz) barometry. The studied samples have mineral assemblages of garnet + omphacite + kyanite + quartz accompanied by phengite, epidote group mineral or pale green amphibole. Chemical compositions used for P–T estimation are garnet composition richest in grossular component and omphacite composition richest in jadeite component in each sample or each layer in order to avoid effects of chemical modification during decompression and/or cooling stages. Such a manner of selection of chemical compositions was adopted also in Carswell et al. (1997) and Nakamura & Banno (1997). Fe3+ in clinopyroxene was estimated as Fe3+ = Na – Al. It cannot be known whether equilibrium states were achieved and preserved or not, but we assume that garnet and omphacite preserve equilibrium compositions at peak P–T conditions. Evaluation of equilibrium states is one of the problems to be resolved in the future. Chemical compositions of garnet, omphacite and phengite are listed in Table 2, data from Rongcheng County are shown in Nakamura & Hirajima (2000).

The Dabie Shan–Sulu orogen

133

Table 2. Chemical compositions of garnet (Grt), clinopyroxene (Omp) and phengite (Phn) used for P–T estimation of eclogites from Sulu area. D: eclogites from Donghai County, T: Taohang eclogite, Y: Yangkou eclogite. Grt

D-1

D-2

D-3

D-4

T-1

Y-1

Y-2

Y-3/2 Y-3/3-2

Y-3/3

SiO2 40.13 39.44 41.15 40.12 39.81 40.36 40.24 40.44 40.09 40.58 Al2O3 21.78 21.62 23.25 22.64 22.12 22.64 22.45 22.51 22.35 22.35 FeO 21.54 20.32 15.35 16.98 21.26 18.39 18.91 18.97 19.54 19.36 MnO 0.53 0.60 0.40 0.33 0.58 0.39 0.34 0.38 0.42 0.36 MgO 6.78 6.66 10.39 8.43 6.46 6.61 5.08 8.46 7.18 5.52 CaO 10.83 11.57 10.82 12.06 10.86 13.24 14.75 10.04 11.57 14.05 Total 101.59 100.21 101.36 100.56 101.09 101.63 101.77 100.80 101.15 102.22 Si 3.029 3.012 3.025 3.006 3.021 3.023 3.032 3.037 3.019 3.042 Al 1.938 1.946 2.015 1.999 1.979 1.999 1.994 1.993 1.984 1.975 Fe 1.360 1.298 0.944 1.064 1.349 1.152 1.192 1.191 1.231 1.214 Mn 0.034 0.039 0.025 0.021 0.037 0.025 0.022 0.024 0.027 0.023 Mg 0.763 0.758 1.139 0.942 0.731 0.738 0.571 0.947 0.806 0.617 Ca 0.876 0.947 0.852 0.968 0.883 1.063 1.191 0.808 0.934 1.129 Total 8.000 8.000 8.000 8.000 8.000 8.000 8.000 8.000 8.000 8.000 D-2

D-3

D-4

SiO2 56.21 56.52 Al2O3 12.31 11.89 FeO 6.46 8.06 MgO 6.28 5.79 CaO 9.58 8.83 Na2O 9.25 9.75 Total 100.09 100.84 Si 1.978 1.978 Al 0.511 0.491 Fe 0.190 0.236 Mg 0.329 0.302 Ca 0.361 0.331 Na 0.631 0.662 Total 4.000 4.000

55.96 9.66 2.02 10.76 15.77 5.52 99.69 1.987 0.404 0.060 0.569 0.600 0.380 4.000

56.06 11.72 3.07 8.23 12.60 7.37 99.05 1.993 0.491 0.091 0.436 0.480 0.508 4.000

Omp

D-1

Phe

D-2

D-4

Y-3/2

Y-3/4

SiO2 TiO2 Al2O3 FeO MgO Na2O K2O Total Si Ti Al Fe Mg Na K Total

50.68 0.65 25.34 2.20 4.33 0.74 9.73 93.67 3.438 0.033 2.026 0.125 0.438 0.097 0.842 7.000

52.61 0.24 24.78 1.03 4.95 0.61 10.70 94.92 3.505 0.012 1.946 0.057 0.492 0.079 0.909 7.000

53.35 0.63 23.51 1.59 5.64 0.49 10.94 96.15 3.517 0.031 1.827 0.088 0.554 0.063 0.920 7.000

51.27 0.75 24.26 1.93 4.21 0.17 10.80 93.39 3.504 0.039 1.954 0.110 0.429 0.023 0.942 7.000

T1

Y-1

56.01 57.47 13.64 13.15 3.72 4.19 6.58 6.64 10.54 10.56 8.62 8.98 99.11 100.99 1.982 1.997 0.569 0.539 0.110 0.122 0.347 0.344 0.400 0.393 0.592 0.605 4.000 4.000

Y-2

Y-3/2 Y-3/3-2

Y-3/4

Y-3/5

38.75 40.06 21.94 22.34 18.83 20.51 0.38 0.44 5.11 5.72 14.11 12.88 99.12 101.95 2.998 3.016 2.000 1.983 1.218 1.292 0.025 0.028 0.589 0.642 1.170 1.039 8.000 8.000

Y-3/3

Y-3/4

Y-3/5

56.53 56.54 56.62 57.17 14.69 12.04 13.15 14.73 3.99 3.72 3.77 3.83 5.23 7.58 6.86 5.60 8.72 11.90 10.79 9.36 9.88 8.23 8.89 9.81 99.04 100.01 100.08 100.50 1.993 1.986 1.981 1.986 0.610 0.499 0.542 0.603 0.118 0.109 0.110 0.111 0.275 0.397 0.358 0.290 0.329 0.448 0.405 0.348 0.675 0.561 0.603 0.661 4.000 4.000 4.000 4.000

56.45 14.29 3.68 5.41 9.24 9.51 98.58 2.002 0.597 0.109 0.286 0.351 0.654 4.000

56.86 13.72 4.18 5.91 9.93 9.15 99.75 2.000 0.569 0.123 0.310 0.374 0.624 4.000

134

T. Hirajima & D. Nakamura

Results of calculations are shown in Figure 12. The estimated P–T conditions of eclogites from Donghai County range from 750 °C, 2.85 GPa to 880 °C, 3.6 GPa. The calculated conditions are in the coesite-stable field, consistent with the preservation of coesite in eclogite from Donghai County (e.g. Hirajima et al., 1990; Zhao et al., 1992). However, the estimated P–T conditions for the Taohang eclogite in Zhucheng County are about 790 °C, 2.65 GPa, that is in quartz-stable field, but Yao et al. (2000) estimated 3.1 GPa at 800 °C for eclogite from Taohang using garnet–clinopyroxene–phengite barometer. Coesite itself is not found in the Taohang eclogite, but peak pressure conditions of the Taohang eclogite show probably near the coesite–quartz transition pressures. The estimated P–T conditions for Yangkou eclogites range from 700 to 800 °C, 3.1–4.1 GPa. From this locality, coesite was found in eclogites (e.g. Hirajima, 1996; Wallis et al., 1997), and the calculated pressure is consistent with this observation. However, the estimated pressures are widely scattered and we cannot determine accurate peak P conditions. This is caused by the heterogeneity of mineral compositions and by the uncertainty of thermodynamic properties and data. Recently, thermodynamic properties and data have been modified and sophisticated, but quantitative problems still exist. The estimated P–T conditions of eclogites from Rongcheng County are about 790–880 °C, 2.9–4.0 GPa. In Rongcheng County, coesite was found in a Chijiadian sample among our studied kyanite-bearing eclogites (Nakamura & Hirajima, 2000), and the calculated results support regional UHP metamorphism in this county. In the Sulu area, coesite was found in several localities: Weihai (e.g. Wang et al., 1993), Rongcheng (Chijiadian: Nakamura & Hirajima, 2000), Zekou (Kurahashi et al., 2001), Yangkou (e.g. Wallis et al., 1997), Donghai (Mengzhong: Hirajima et al., 1990; Hetang: Zhao et al., 1992), although many eclogites transformed to granulite in the northeastern and to amphibolite in the central and southwestern parts of the county to various degrees. Most of the calculated P–T conditions fall in the coesite-stable field (Fig. 12), suggesting that most of the eclogite bodies from the Sulu area, except for the rocks around Haiyangsuo (Fig. 3), suffered UHP metamorphism. If eclogite bodies were exhumed as a tectonic block and were trapped into the country gneiss after decompression, some eclogite bodies would have ascended from relatively shallow levels (i.e., quartz-stable depth). The calculated results, however, indicate that they were formed under coesite stability field, suggesting that the “in situ” model is a plausible tectonics in the Sulu area. Enami et al. (1993a) proposed a southwestward decrease of metamorphic temperatures in the Sulu area, but no such tendency was identified in this study. Comparing Donghai County in the southwestern part with Rongcheng County in the northeastern part, there is no apparent difference in temperature conditions (Fig. 12). However, the estimation of temperatures has a potential difficulty: identification of equilibrium compositions, non-ideality of solid solution and analytical problems (estimation Fe3+ in omphacite). For example, the estimated temperatures for Yangkou eclogites are scattered (700–800 °C), even though they were collected from the same locality. As one of the problems, the non-ideality of the jadeite component was pointed out by Hirajima (1996), who showed that nominal calculated temperatures increase with increasing jadeite component in omphacite with high jadeite component (> 50 mol%),

The Dabie Shan–Sulu orogen

135

Fig. 12. Calculated pressure–temperature conditions for eclogites from the Sulu area. Chemical compositional data used for calculations are shown in Table 2 and Nakamura & Hirajima (2000). Fe3+ in clinopyroxene is estimated as Fe3+ = Na – Al. Temperatures are estimated by the garnet–clinopyroxene thermometer of Powell (1985) or Krogh (1988). Equilibrium curves for the garnet-clinopyroxene-kyanite-coesite assemblage (this study) are calculated using thermodynamic dataset of Holland & Powell (1998) with activity models of Nakamura & Banno (1997). Pressures are also estimated for eclogites from Donghai and Yangkou by the garnet–clinopyroxene–phengite barometer calibrated by Waters & Martin (1996) (WM96). Quartz–coesite and diamond–graphite transition curves are based on Holland & Powell (1998). Qtz: quartz, Coe: coesite, Dmn: diamond, Gr: graphite.

and suggested that the variety of the estimated temperatures for Yangkou eclogites may be due to the non-ideal effect of the jadeite component. Therefore, we investigate the relationship between the estimated temperatures and jadeite component (Al of clinopyroxene) (Figs. 13 & 14). However, our calculated data in the other localities of the Sulu area do not show a clear positive correlation, although Rongcheng eclogites, which were collected at several localities, show a weak positive correlation (Fig. 13). Zhang et al. (2000b) reported the equilibrium P–T conditions for the eclogites collected from the 558 m deep ZK703 drill hole at Mobei, Donghai County, as around 3.2 to 4.0 GPa and 770 to 880 °C using the combination of Powell’s (1985) Grt–Cpx thermometry and Grt–Cpx–Phn geobarometry for the average compositions of Grt and Omp in each sample, and by assuming Fe3+ = Na – Al in Cpx and all Fe as divalent in Grt. They estimated P–T conditions very similar to our P–T estimations in Donghai

136

T. Hirajima & D. Nakamura

(Fig. 12). However, their original data (Table 3 in Zhang et al., 2000b) suggest that the estimated temperatures positively correlate with the jadeite component, ranging from 623 to 853 °C at 3.0 GPa for XJd = 0.31 to 0.72 (Fig. 14b). Zhang et al. (2000b) regarded the temperature lower than 700 °C, obtained from two samples with lower XJd (0.31–0.38) omphacite, as a “false value” due to an incorrect estimate of Fe3+ content in omphacite and/or possible retrograde re-equilibration of mineral compositions. We think that it is not reasonable to simply abandon the lower temperature data from their description, and the ZK703 drillhole eclogite may show the second example of positive correlation between wstimated temperature and XJd. The estimated temperatures in eclogites with omphacite with low jadeite component (< 50 mol%) from Donghai and Rongcheng Counties are relatively high comparing with Yangkou eclogites. Although the problem of the effect of the jadeite component still remains, metamorphic temperatures for Yangkou eclogites may be slightly lower than those for eclogites from the other counties (cf. Figs. 12 and 13). At last, we cannot identify significant difference in metamorphic temperatures between Donghai and Rongcheng County. There is no apparent regional metamorphic gradient that was proposed by Enami et al. (1993a). In the Sulu area, most of eclogite bodies suffered UHP metamorphism that are mostly in the same P–T conditions, although eclogites from Yangkou show slightly lower temperature. At present, more precise analyses of P–T conditions are still difficult, and we have to overcome several problems to obtain accurate temperatures.

Concluding remarks Wang et al. (1992, 1995) and Enami et al., (1993a) emphasised that the peak metamorphic temperatures of eclogites systematically decrease from east to west through the Dabie Shan–Sulu area and from north to south in the Dabie Shan. Continuous variation of peak metamorphic temperatures for UHP/HP rocks is one of the positive evidences for the “in situ” school. The present compiled data suggest that the coesite eclogites show a relatively uniform peak temperature between 700–850 °C in the Dabie Shan–Sulu area, which does not support the former idea mentioned above. The equilibrium temperatures in the Dabie Shan support the conclusion of Okay (1993) and Carswell et al. (1997), i.e., that there is a distinct gap of the peak P–T conditions between the quartz eclogite terrane and the coesite eclogite terrane. Their conclusion claimed to partly modify the conclusion of Wang et al. (1992, 1995) but does not deny the “in situ” model of the UHP/HP eclogite in the study area. However we emphasise again that the number of the reported P–T estimations for the quartz eclogite terrane is not enough to discuss the tectonic relationship between the two terranes. Several tenacious field surveys guided by Chinese geologists and their co-workers succeeded to confirm that many UHP eclogites and associated UHP rocks, some of which were certainly derived from crustal materials, were exposed as coherent formations in the coesite eclogite unit of Dabie Shan and Sulu area (e.g. Shima, Changpu, Xinjian, Shuanghe in southern Dabie Shan; Donghai and Yangkou in Sulu). Among them, the areal scale reconstruction of pre-UHP geology was succeeded in

The Dabie Shan–Sulu orogen

137

Fig. 13. Relationship between estimated temperature and jadeite content of clinopyroxene (Al p.f.u.) for eclogites from the Sulu area. Temperatures are estimated by the garnet–clinopyroxene thermometer of Powell (1985) or Krogh (1988) at 3.5 GPa. For eclogites from Yangkou, there is a clear positive correlation between estimated temperature and jadeite content, but no clear correlation is observed for eclogites from other areas, except for Rongcheng.

Fig. 14. Relationship between estimated temperature and jadeite content of clinopyroxene for eclogites at (a) Yangkou (Hirajima, 1996) and (b) ZK703 drillhole at Mobei, Donghai (Zhang et al., 2002b). Temperatures are estimated by the garnet–clinopyroxene thermometer of Powell (1985) at 3.0 GPa.

138

T. Hirajima & D. Nakamura

Shuanghe (Cong et al., 1995) and Yangkou (Wallis et al., 1997) in several hundred-meter scales, and the several tens of kilometres scale continuity of the Chenjiahe Formation of Wang et al. (1990) in the central Dabie Shan was recently confirmed by Compagnoni et al. (2001) as their Changpu–Pailou unit. These geological contributions support the “in situ” school of the UHP rocks. However, these data cannot explain the existence of the apparent P–T gaps between the UHP rocks and the surrounding amphibolite facies country gneiss. Finding of tiny coesite and other UHP minerals as inclusions in zircons of the country gneisses (Tabata et al., 1998b; Ye et al., 2000b; Liu et al., 2002) encourages the “in situ” school. A further comprehensive study on the Changpu–Pailou unit will give more interesting data on this long-term argument in the near future. A precisely determined exhumation path of UHP/HP rocks can contribute to define the exhumation mechanism of UHP/HP rocks and/or to reveal the root zone tectonics of the collision belt (e.g. Hacker & Peacock, 1995). Most UHP eclogites and garnet peridotites suffered multiple re-equilibrium reactions, chiefly represented by the amphibolite and/or greenschist facies overprint during the latest stage of the exhumation. Some of them recorded the subsequent overprints during the early decompression stage, as suggested by the development of paragonite in the UHP eclogite matrix under quartz eclogite facies in Dabie Shan (Okay, 1995) or during the late decompression stage represented by granulite facies minerals in the northeastern part of the Sulu area (e.g. Wang et al., 1993; Nakamura & Hirajima, 2000). If we can estimate each re-equilibrium P–T condition precisely, we can propose the precise exhumation path. Although we do not give a detailed assessment to each proposed decompression path in this chapter, two types of exhumation paths (isothermal decompression path, ITD; e.g. Liou et al., 1997; Banno et al., 2000; Castelli et al., 1998; Nakamura & Hirajima, 2000; Compagnoni et al., 2001; Yoshida et al., 2001; or decompression with significant cooling path, DSC; e.g. Schmidt et al., 2001; Wang et al., 2001) were proposed in the Dabie Shan–Sulu area. If the UHP eclogites formed around 750–800 °C exhumed maintaining the isothermal state, they can suffer the granulite facies overprinting around the lower crustal depths. The northeastern part of the Sulu area can fit to this case. In the Dabie Shan area, both ITD and DSC paths were reported. Further efforts to delineate the precise exhumation path and collate it with the geotectonics obtained from geophysical and numerical models in the relevant area are attractive targets for the petrologists. The apparent pressure gap at the UHP stage obtained from garnet peridotite and eclogite is one of unsolved problems for the “in situ” school. Some garnet peridotite recorded the diamond stability pressure around 4–6 GPa at peak stage, which is significantly higher than the peak pressure for the eclogite. The finding of clinoenstatite lamellae in the host orthoenstatite and diopside (Su et al., 2001; Zhang et al., 2002), and the reconstructed majoritic garnet composition from a Yangkou eclogite (Ye et al., 2000a) may suggest that at least some of the coevally subducted crustal rocks have been subjected to peak metamorphism at P–T conditions much higher than that obtained from the eclogite (~ 3–4 GPa and 700–850 °C). This is another extensive target remaining for the petrologists.

The Dabie Shan–Sulu orogen

139

Acknowledgements We express our sincere thanks to D.A. Carswell and R. Compagnoni for giving a chance for writing this review, to all members of the co-operative research on Sulu area between Kyoto University and Academia Sinica originally guided by S. Banno and the late Cong Bolin for their help of carrying out the field and indoor survey. We also acknowledge Keisaku Matsumoto for his computer drawing of some geological maps of the Dabie Shan area and J.G. Liou and B.M. Jahn for their constructive and critical comments for this chapter.

References Akaogi, M. & Akimoto, S. (1977): Pyroxene-garnet solid-solution equilibria in the systems Mg4Si4O12Mg3Al2Si3O12 and Fe4Si4O12-Fe3Al2Si3O12 at high pressures and temperatures. Phys. Earth Planet. Inter., 15:90–106. Ames, L., Tilton, G.R. & Zhou, G. (1993): Timing of collision of the Sino-Korean and Yangtse cratons: U-Pb zircon dating of coesite-bearing eclogites. Geology, 21:339–342. Ames, L., Zhou, G. & Xiong, B. (1996): Geochronology and isotopic character of ultrahigh-pressure metamorphism with implications for collision of the Sino-Korean and Yangtze cratons, central China. Tectonics, 15:472–489. Banno, S., Enami, M., Hirajima, T., Ishiwatari, A. & Wang, Q. (2000): Decompression P-T path of coesite eclogite to granulite from Weihai, eastern China. Lithos, 52:97–108. Berman, R.G., Aranovich, Y.L. & Pattison, D.R. (1995): Reassessment of the garnet-clinopyroxene Fe-Mg exchange thermometer: II. Thermodynamic analysis. Contrib. Mineral. Petrol., 119:30–42. Bertrand, P. & Mercier, J.C.C. (1985): The mutual solubility of coexisting ortho- and clinopyroxene: toward an absolute geothermometer for the natural system? Earth Planet. Sci. Lett., 76:109–122. Biino, G. & Compagnoni, R. (1991): Very-high-pressure metamorphism of the Brossasco coronite metagranite, southern Dora-Maira Massif, western Alps. Schweiz. Mineral. Petrogr. Mitt., 72:347–362. Brey, G.P. & Köhler, T. (1990): Geothermobarometry in four-phase lherzolites II. New thermobarometers, and practical assessment of exiting thermobarometers. J. Petrol., 31:1353–1378. Carswell, D.A., O’Brien, P.J., Wilson, R.N. & Zhai, M. (1997): Thermobarometry of phengite-bearing eclogites in the Dabie Mountains of central China. J. Metamorph. Geol., 15:239–252. Carswell, D.A., Wilson, R.N. & Zhai, M. (2000): Metamorphic evolution, mineral chemistry and thermobarometry of schists and orthogneisses hosting ultra-high pressure eclogites in the Dabieshan of central China. Lithos, 52:121–155. Castelli, D., Rolfo, F. & Compagnoni, R. (1998): Metamorphic veins with kyanite, zoisite and quartz in the Zhu-Jia-Chong eclogite, Dabie Mountains, China. Isl. Arc, 7:159–173. Cheng, Y. (1986): Metamorphic series and metamorphic belts of China. In Yang, Z., Cheng, Y.Q. & Wang, H.Z. (eds.): The geology of China /Oxford Monogr. Geol. Geophys., 3/. Oxford: Clarendon Press, 303 p. Chen, B., Jahn, B.M., Ye, K. & Liu, J.B. (2002): Cogenetic relationship of the Yangkou gabbro-to-granulite unit, Su-Lu terrane, eastern China, and implications for UHP metamorphism. J. Geol. Soc. London, 159:457-467. Chopin, C., Henry, C. & Michard, A. (1991): Geology and petrology of the coesite-bearing terrain, Dora Maira massif, Western Alps. Eur. J. Mineral., 3:263–291. Compagnoni, R. & Rolfo, F. (2003): UHPM units in the Western Alps. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 13–49. Compagnoni, R., Hirajima, T. & Chopin, C. (1995): Ultra-high-pressure metamorphic rocks in western Alps. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 206–243. Compagnoni, R., Rolfo, F., Wu, W. & Xu, S. (2001): A coherent lithostratigraphic unit in the coesite-eclogite complex of Dabie Shan, China: geologic and petrologic evidence. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 24–25.

140

T. Hirajima & D. Nakamura

Cong, B., Zhai, M., Carswell, D.A., Wilson, R.N., Wang, Q., Zhao, Z. & Windley, B.F. (1995): Petrogenesis of ultrahigh-pressure rocks and their country rocks at Shuanghe in Dabieshan, central China. Eur. J. Mineral., 7:119–138. Cong, B., Zhang, R.Y., Liou, J.G., Ye, K. & Wang, Q. (1996): Metamorphic evolution of UHPM rocks. In Cong, B. (ed.): Ultrahigh-pressure metamorphic rocks in the Dabieshan-Sulu region of China. Beijing: Science Press & Dordrecht: Kluwer, 128–160. Cuthbert, S.J., Carswell, D.A., Krogh-Ravna, E.J. & Wain, A. (2000): Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides. Lithos, 52:165–195. Eide, E.A. & Liou, J.G. (2000): High-pressure blueschists and eclogites in Hong’an: a framework for addressing the evolution of high- and ultrahigh-pressure rocks in central China. Lithos, 52:1–22. Ellis, D.J. & Green, D.H. (1979): An experimental study of the effect of Ca upon garnet-clinopyroxene Fe-Mg exchange equilibria. Contrib. Mineral. Petrol., 71:13–22. Enami, M., Suzuki, K., Liou, J.G. & Bird, D.K. (1993a): Al-Fe3+ and F-OH substitution in titanite and constraints on their P-T dependence. Eur. J. Mineral., 5:219–231. Enami, M., Suzuki, K., Zhai, M. & Zheng, X. (1993b): The chemical Th-U-total Pb isochron ages of Jiaodong and Jiaonan metamorphic rocks in the Shandong Peninsula, eastern China. Isl. Arc, 2:104–113. Enami, M., Zang, Q. & Yin, Y. (1993c): High-pressure eclogites in northern Jiangsu – southern Shangdong province, eastern China. J. Metamorph. Geol., 11:589–603. Enami, M., Cong, B., Yoshida, T. & Kawabe, L. (1995): A mechanism for Na incorporation in garnets: an example from garnet in orthogneiss from the Su-Lu terrane, eastern China. Am. Mineral., 80:475–482. England, P.C. & Richardson, S.W. (1977): The influence of erosion upon the mineral facies of rocks from different metamorphic environments. J. Geol. Soc. London, 134:201–213. Fabriès, J. (1979): Spinel-olivine geothermometry in peridotites from ultramafic complexs. Contrib. Mineral. Petrol., 69:329–336. Faure, M., Lin, W. & Breston, N.L. (2001): Where is the North China-South China block boundary in eastern China? Geology, 29:119–122. Faure, M., Lin, W., Scharer, U., Shu, L., Sun, Y. & Arnaud, N. (2003): Continental subduction and exhumation of UHP rocks. Structural and geochronological insights from the Dabieshan (E. China). Lithos, in press. Ganguly, J. (1979): Garnet and clinopyroxene solid solutions, and geothermometry based on Fe–Mg distribution coefficient. Geochim. Cosmochim. Acta, 43:1021–1029. Ganguly, J., Cheng, W. & Tirone, M. (1996): Thermodynamics of aluminosilicate garnet solid solution: new experimental data, an optimized model, and thermometric applications. Contrib. Mineral. Petrol., 126:137–151. Green, T. H. & Adam, J. (1991): Assessment of the garnet-clinopyroxene Fe-Mg exchange thermometer using new experimental data. J. Metamorph. Geol., 9:341–347. Green, D.H. & Hellman, P.L. (1982): Fe-Mg partitioning between coexisting garnet and phengite at high pressures, and comments on a garnet-phengite geothermometer. Lithos, 15:253–266. Hacker, B.R. & Peacock, S.M. (1995): Creation, preservation, and exhumation of UHPM rocks. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 159–181. Hacker, B.R. & Wang, Q. (1995): Ar/Ar geochronology of ultra-high-pressure metamorphism in central China, Tectonics, 14:994–1006. Hacker, B.R., Ratschbacher, L., Webb, S. & Dong, S. (1995): What brought them up? Exhumation of the Dabie Shan ultrahigh-pressure rocks. Geology, 23:743–746. Hacker, B.R., Wang, X., Eide, E.A. & Ratschbacher, L. (1996): Qinling-Dabie ultrahigh-pressure collisional orogen. In Yin, A. & Harrison, T.M. (eds.): Tectonic evolution of Asia. Cambridge: Cambridge Univ. Press, 345–370. Hacker, B.R., Ratschbacher, L., Webb, L., Ireland, T., Walker, D. & Dong, S. (1998): U-Pb zircon ages constrain the architecture of the ultrahigh-pressure Qinling-Dabie Orogen, China. Earth Planet. Sci. Lett., 161:2150–2230. Hackler, R.T. & Wood, B.J. (1989): Experimental determination of Fe and Mg exchange between garnet and olivine and estimation of Fe-Mg mixing properties in garnet. Am. Mineral., 74:994–999.

The Dabie Shan–Sulu orogen

141

Harley, S.L. (1984a): An experimental study of the partitioning of Fe and Mg between garnet and orthopyroxene. Contrib. Mineral. Petrol., 86:359–373. Harley, S.L. (1984b): The solubility of alumina in orthopyroxene coexisting with garnet in FeO-MgO-Al2O3SiO2 and CaO-FeO-MgO-Al2O3-SiO2. J. Petrol., 25:665–696. Hirajima, T. (1996): Effect of jadeite-content on the garnet-clinopyroxene geothermometer for an ultrahighpressure eclogite complex. Proc. Jpn. Acad., 72:Ser. B., 208–213. Hirajima, T. (1998): Preservation and retrogression of Ultra High Pressure (UHP) rocks: Case studies of UHP metagranitoids in Western Alps and Su-Lu region. In Motoyoshi, Y. & Shiraishi, K. (eds.): Origin and evolution of continents. Proc. Int. Symp. “Origin and Evolution of Continents”, 13–14 Oct. 1997, Tokyo /Mem. Natl. Inst. Polar Res., Spec. Issue, 53/, 29–48. Hirajima, T. & Fanning, M. (1999): U-Pb age of the Yangkou UHP metamorphic complex in the Su-Lu belt, China. In Jpn. Earth Planet. Sci. Joint Meeting, Abstr., Ge-004. Hirajima, T., Ishiwatari, A., Cong, B., Zhang, R.Y., Banno, S. & Nozaka, T. (1990): Coesite from Mengzhong eclogite at Donghai county, northeastern Jiangsu province, China. Mineral. Mag., 54:579–583. Hirajima, T., Wallis, S., Zhai, M. & Ye, K. (1993): Eclogitized metagranitoid from the Su-Lu ultra-high pressure (UHP) Province, eastern China. Proc. Jpn. Acad., 69:Ser. B., 249–254. Hiramatsu, N. & Hirajima, T. (1995): Petrology of the Hujialin garnet clinopyroxenite in the Su-Lu ultrahighpressure (UHP) province, eastern China. Isl. Arc, 4:310–323. Hiramatsu, N., Banno, S., Hirajima, T. & Cong, B. (1995): Ultrahigh-pressure garnet lherzolite from Chijiadian, Rongcheng County, in the Su-Lu region of eastern China. Isl. Arc, 4:324–333. Holland, T.J.B. & Powell, R. (1998): An internally consistent thermodynamic data set for petrological interest. J. Metamorph. Geol., 16:309–343. Irifune, T. (1987): An experimental investigation of the pyroxene-garnet transformation in a pyrolite composition and its bearing on the constitution of the mantle. Phys. Earth Planet. Inter., 45:324–336. Ishiwatari, A., Hirajima, T., Enami, M., Hiramatsu, N., Zhai, Z. Wang, Q., Zhai, M., Cong, B., Zhang, R.Y., Banno, S. & Ishizaka, K. (1992): Eclogite in Shandong Province and Jiangsu Province, China: 1991 survey report. In Memorial volume of Prof. Y. Matsumoto, 393–409 (in Japanese). Ishizaka, K., Hirajima, T. & Zheng, X. (1994): Rb-Sr dating for the Jiaodong gneiss of the Su-Lu ultra-high pressure province, eastern China. Isl. Arc, 3:232–241. Jahn, B.M. (1998): Geochemical and isotopic characteristics of UHP eclogites and ultramafic rocks of the Dabie orogen. Implications for continental subduction and collision tectonics. In Hacker, B. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 203–239. Jahn, B.M., Cornichet, J., Cong, B. & Yui, T.F. (1996): Ultrahigh-Nd eclogites from an ultrahigh-pressure metamorphic terrane of China. Chem. Geol., 127:61-79. Jahn, B.-M., Fan, Q. & Yang, J.-J. (2001): Genesis of the Maowu mafic-ultramafic body from the UHP terrane of Dabieshan: Chemical and Isotopic constraints and tectonic implications. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 57. Kato, T., Enami, M. & Zhai, M. (1997): Ultra-high-pressure (UHP) marble and eclogite in the Su-Lu UHP terrane, eastern China. J. Metamorph. Geol., 15:169–182. Krogh, E.J. (1977): Evidence for a Precambrian continent-continent collision in western Norway. Nature, 267:17–19. Krogh, E.J. (1988): The garnet-clinopyroxene Fe-Mg geothermometer – a reinterpretation of existing experimental data. Contrib. Mineral. Petrol., 99:44–48. Krogh Ravna, E.J. (2000): The garnet-clinopyroxene Fe2+-Mg geothermometer: an updated calibration. J. Metamorph. Geol., 18:211–219. Kurahashi, E., Nakajima, Y. & Ogasawara, Y. (2001): Coesite inclusions and prograde compositional zonation of garnet in eclogite from Zekou in the Su-Lu ultrahigh-pressure terrane, eastern China. J. Mineral. Petrol. Sci., 96:100–108. Lee, H.Y. & Ganguly, J. (1988): Equilibrium compositions of coexisting garnet and orthopyroxene: experimental determinations in the system FeO-MgO-Al2O3-SiO2, and applications. J. Petrol., 29:93–113.

142

T. Hirajima & D. Nakamura

Li, S., Xiao, Y., Liou, D., Xhen, Y., Ge, N., Zhang, Z., Sun, S., Cong, B., Zhang, R.Y., Hart, S.R. & Wang, S. (1993): Collision of the North China and Yangtze blocks and formation of coesite-bearing eclogites: Timing and processes. Chem. Geol., 109:89–111. Liou, J.G. & Zhang, R.Y. (1998): Petrogenesis of an ultrahigh-pressure garnet-bearing ultramafic body from Maowu, Dabie Mountains, east-central China. Isl. Arc, 7:115–134. Liou, J.G. & Zhang, R.Y. (2001): Lack of in-situ occurrence of crustal diamonds in the Dabie-Sulu UHP terranes, east central China. In Annu. Meeting Geol. Soc. Am., Abstr. Progr., 33:428. Liou, J.G., Zhang, R.Y. & Jahn, B.-M. (1997): Petrology, geochemistry and isotope data on a ultrahighpressure jadeite quartzite from Shuanghe, Dabie Mountains, East-central China. Lithos, 41:59–78. Liu, J., Ye, K., Maruyama, S., Cong, B. & Fan, H. (2001): Coesite inclusions in zircon from gneisses identified by laser Raman microspectrometer in ultra-high pressure zone of Dabie Mountains, China. Chin. Sci. Bull., 46:1665–1668. Liu, F., Xu, Z., Liou, J.G., Katayama, I., Masago, H., Maruyama, S. & Yang, J. (2002): Ultrahigh-pressure mineral inclusions in zircons from gneissic core samples of the Chinese Continental Scientific Drilling Site in eastern China. Eur. J. Mineral., 14:499–512. Medaris, L.G., Jr. (2000): Garnet peridotites in Eurasian high-pressure and ultrahigh-pressure terranes: A diversity of origins and thermal histories. In Ernst, W.G. & Liou, J.G. (eds.): Ultrahigh-pressure metamorphism and geodynamics in collision-type orogenic belts /Int. Book Ser., 4/. Boulder (Co.): Geol. Soc. Am. & Columbia (Md.): Bellwether Publ. Ltd., 57–73. Mori, T. & Green, D.H. (1978): Laboratory duplication of phase equilibria observed in natural garnet lherzolites. J. Geol., 86:83–97. Mørk, M.B. (1985): A gabbro to eclogite transition on Flemsøy, Sunnmøre, western Norway. Chem. Geol., 50:283–310. Nagasaki, A. & Enami, M. (1998): Sr-bearing zoisite and epidote in ultra-high pressure (UHP) metamorphic rocks from the Su-Lu province, eastern China; an important Sr reservoir under UHP conditions. Am. Mineral., 83:240–247. Nakamura, D. (2003): Stability of phengite and biotite in eclogites and characteristics of biotite- or orthopyroxene-bearing eclogites. Contrib. Mineral. Petrol., 145:550–567. Nakamura, D. & Banno, S. (1997): Thermodynamic modelling of sodic pyroxene solid-solution and its application in a garnet–omphacite–kyanite–coesite geothermobarometer to UHP metamorphic rocks. Contrib. Mineral. Petrol., 130:93–102. Nakamura, D. & Hirajima, T. (2000): Granulite-facies overprinting of ultrahigh-pressure metamorphic rocks, northeastern Su–Lu region, eastern China. J. Petrol., 41:563–582. Nakamura, D., Hirajima, T. & Takahashi, E. (2001): Experimental evaluation of garnet–clinopyroxene thermometer. Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 97. Okay, A.I. (1993): Petrology of a diamond- and coesite-bearing metamorphic terrain: Dabie Shan, China. Eur. J. Mineral., 5:659–675. Okay, A.I. (1994): Sapphirine and Ti-clinohumite in ultra-high-pressure garnet-pyroxenite and eclogite from Dabie Shan, China. Contrib. Mineral. Petrol., 116:145–155. Okay, A.I. (1995): Paragonite eclogites from Dabie Shan, China: re-equilibrations during exhumation. J. Metamorph. Geol., 13:449–460. Okay, A.I., Xu, S. & Șengör, A.M.C. (1989): Coesite from the Dabie Shan eclogites, central China. Eur. J. Mineral., 1:595–598. O’Neill, H.St.C. (1981): The transition between spinel lherzolite and garnet lherzolite, and its use as a geobarometer. Contrib. Mineral. Petrol., 77:185–194. O'Neill, H.St.C. & Wall, V.J. (1987): The olivine-orthopyroxene-spinel oxygen geobarometer, the nickel precipitation curve, and the oxygen fugacity of Earth's upper mantle. J. Petrol., 28:1169–1191. Pattison, D.R.M. & Newton, R.C. (1989): Reversed experimental calibration of the garnet-clinopyroxene FeMg exchange thermometer. Contrib. Mineral. Petrol., 101:87–103. Perkins, D. & Vielzeuf, D. (1992): Experimental investigation of Fe-Mg distribution between olivine and clinopyroxene: Implications for mixing properties of Fe-Mg in clinopyroxene and garnet-clinopyroxene thermometry. Am. Mineral., 77:774–783.

The Dabie Shan–Sulu orogen

143

Powell, R. (1985): Regression diagnostics and robust regression in geothermometer / geobarometer calibration: the garnet-clinopyroxene geothermometer revisited. J. Metamorph. Geol., 3:231–243. Råheim, A. & Green, D.H. (1974): Experimental determination of the temperature and pressure dependence of the Fe-Mg partition coefficient for coexisting garnet and clinopyroxene. Contrib. Mineral. Petrol., 48:179–203. Rolfo, F., Compagnoni, R., Xu, S. & Jiang, L. (2000): First report of felsic whiteschist in the ultrahigh-pressure metamorphic belt of Dabie Shan, China. Eur. J. Mineral., 12:883-898. Rowley, D.B., Xue, F., Tucker, R.D., Peng, Z.X., Baker, J. & Davis, A. (1997): Ages of ultrahigh pressure metamorphism and protolith orthogneisses from the eastern Dabie Shan: U/Pb zircon geochronology. Earth Planet. Sci. Lett., 151:191–203. Schertl, H.P., Schreyer, W. & Chopin, C. (1991): The pyrope-coesite rocks and their country rocks at Parigi, Dora Maira Massif, Western Alps: detailed petrography, mineral chemistry and PT-path. Contrib. Mineral. Petrol., 108:1–21. Schmid, R., Oberhänsli, R., Franz, L., Romer, R.L., Martinotti, G., Ryberg, T., Schulze, A. & Dong, S. (2001): Geology and petrology of ultra-high-pressure basement cover sequences within the Dabie Shan, eastern China. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 143–144. Schmidt, M.W. & Poli, S. (1998): Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation. Earth Planet. Sci. Lett., 163:361–379. Su, W., Xu, S., Jiang, L. & Liu, Y. (1996): Coesite from quartz jadeiteite in the Dabie Mountains, eastern China. Mineral. Mag., 60:659–662. Su, W., You, Z., Wang, R. & Liu, X. (2001): Quartz and clinoenstatite exsolutions in clinopyroxene of garnetpyroxenite from North Dabie Mountains, eastern China. Chin. Sci. Bull., 46:1482–1485. Tabata, H., Maruyama, S. & Shi, Z. (1998a): Metamorphic zoning and thermal structure of the Dabie ultrahighpressure-high-pressure terrane, central China. Isl. Arc, 7:142–158. Tabata, H., Yamauchi, K. & Maruyama, S. (1998b): Tracing the extent of a UHP metamorphic terrane: Mineralinclusion study of zircons in gneisses from the Dabie Shan. In Hacker, B. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 262–273. Ulmer, R. & Stalder, R. (2001): The Mg(Fe)SiO3 orthoenstatite-clinoenstatite transitions at high pressures and temperatures determined by Raman-spectroscopy on quenched samples. Am. Mineral., 86:1267–1274. Wallis, S.R., Ishiwatari, A., Hirajima, T., Ye, K., Guo, J., Nakamura, D., Kato, T., Zhai, M., Enami, M., Cong, B. & Banno, S. (1997): Occurrence and field relationships of ultrahigh-pressure metagranitoid and coesite eclogite in the Su-Lu terrane, eastern China. J. Geol. Soc. London, 153:45–54. Wang, X. & Liou, J.G. (1991): Regional ultrahigh-pressure coesite-bearing eclogitic terrane in central China: evidence from country rocks, gneiss, marble and metapelite. Geology, 19:933–936. Wang, L.M. & Yan, Y.M. (1992): Archean tonalites in Qixia, Shandong. Geol. Shandong, 1:80–87 (in Chinese). Wang, X., Liou, J.G. & Mao, H.K. (1989): Coesite-bearing eclogite from the Dabie Mountains, central China. Geology, 17:1085–1088. Wang, X., Jing, Y., Liou, J.G., Pan, G., Liang, W., Xia, M. & Maruyama, S. (1990): Field occurrences and petrology of eclogites from the Dabie Mountains, Anhui, central China. Lithos, 25:119–131. Wang, X. Liou, J.G. & Maruyama, S. (1992): Coesite-bearing eclogite from the Dabie Mountains, central China; Petrogenesis, P-T paths, and implications for regional tectonics. J. Geol., 100:231–250. Wang, Q., Ishiwatari, A., Zhao, Z., Hirajima, T., Hiramatsu, N., Enami, M., Zhai, M., Li, J. & Cong, B. (1993): Coesite-bearing granulite retrograded from eclogite in Weihai, eastern China: a preliminary study. Eur. J. Mineral., 5:141–152. Wang, X., Zhang, R.Y. & Liou, J.G. (1995): UHPM terrane in east central China. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 356–390. Wang, X., Genser, J., Neubauer, F., Dachs, E. & Proyer, A. (2001): P-T deformation path and exhumation of UHP belt in the eastern Dabie Mountains, China. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 169–170. Waters, D.J. & Martin, H.N. (1993): Geobarometry of phengite-bearing eclogites. Terra Abstr., 5:410–411. Waters, D.J. & Martin, H.N. (1996): The garnet-cpx-phengite barometer. http://www.earth.ox.ac.uk/~davewa/research/ecbarcal.html. Wood, B.J. (1976): The partitioning of iron and magnesium between garnet and clinopyroxene. Year b. – Carnegie Inst. Wash., 75:571–574.

144

T. Hirajima & D. Nakamura

Wood, B.J. & Banno, S. (1973): Garnet-orthopyroxene and orthopyroxene-clinopyroxene relationships in simple and complex systems. Contrib. Mineral. Petrol., 42:109–124. Xu, S., Okay, A.I., Ji, S., Sengor, A.M.C., Su, W., Liu, Y. & Jiang, L. (1992): Diamond from the Dabie Shan metamorphic rocks and its implication for tectonic setting. Science, 256:80–82. Xu, S., Liu, Y., Wu, W. & Li, L. (2001): Eclogite in northern Dabie Mountains and its tectonic implications. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 174. Yang, J.-J. (1991): Eclogites, garnet pyroxenites and related ultrabasics in Shandong and north Jiangsu of east China. Beijing: Geol. Publ. House, 99 p. (In Chinese with English abstr.). Yang, J.-J. & Jahn, B.-M. (2000): Deep subduction of mantle-derived garnet peridotites from the Su-Lu UHP metamorphic terrane in China. J. Metamorph. Geol., 18:167–180. Yang, J.-J, Godard, G., Kienast, J.R., Lu, Y. & Sun, J. (1993): Ultrahigh-pressure (60 kbar) magnesite-bearing garnet peridotites from northeastern Jiangsu, China. J. Geol., 101:541–554. Yao, Y., Ye, K., Liu, J., Cong, B. & Wang, Q. (2000): A transitional eclogite- to high pressure granulite-facies overprint on coesite-eclogite at Taohang in the Sulu ultrahigh-pressure terrane, Eastern China. Lithos, 52:109–120. Ye, K., Cong, B., Hirajima, T. & Banno, S. (1999): Transformation from granulite to transitional eclogite at Haiyangsuo, Rushan county, eastern Shandong peninsula: the kinetic process and tectonic implications. Acta Petrol. Sin., 15:21–36 (in Chinese). Ye, K., Cong, B. & Ye, D. (2000a): The possible subduction of continental material to depths greater than 200 km. Nature, 407:734–736. Ye, K., Yao, Y., Katayama, I., Cong, B., Wang, Q. & Maruyama, S. (2000b): Large areal extent of ultrahighpressure metamorphism in the Sulu ultrahigh-pressure terrane of east China: new implications from coesite and omphacite inclusions in zircon of granitic gneiss. Lithos, 52:157–164. Yoshida, D., Hirajima, T. & Ishiwatari, A. (2001): Pressure-temperature path recorded in Yangkou garnet peridotite in the Su-Lu ultrahigh-pressure metamorphic belt, eastern China and its geological significance. In Sixth Int. Eclogite Conf., Sept. 1–7, Niihama, Japan, Abstr., 180–181. Zhai, M., Cong, B., Guo, J., Liu, W., Li, Y. & Wang, Q. (2000): Sm-Nd geochronology and petrography of garnet pyroxene granulites in the northern Sulu region of China and their geotectonic implication. Lithos, 52:23–33. Zhang, R.Y. & Liou, J.G. (1994): Significance of magnesite paragenesis in ultrahigh-pressure metamorphic rocks. Am. Mineral., 79:397–400. Zhang, R.Y., Liou, J.G. & Cong, B. (1994): Petrogenesis of garnet-bearing ultramafic rocks and associated eclogites in the Su-Lu ultrahigh-P metamorphic terrane, eastern China. J. Metamorph. Geol., 12:169–186. Zhang, R.Y., Hirajima, T., Banno, S., Cong, B. & Liou, J.G. (1995a): Petrology of ultrahigh-pressure rocks from the southern Su-Lu region, eastern China. J. Metamorph. Geol., 13:659–675. Zhang, R.Y., Liou, J.G. & Cong, B. (1995b): Talc-, magnesite- and Ti-clinohumite-bearing ultrahigh-pressure meta-mafic and ultramafic complex in the Dabie Mountains, China. J. Petrol., 36:1011–1037. Zhang, R.Y., Liou, J.G. & Tsai, C.H. (1996): Petrogenesis of a high-temperature metamorphic terrane: a new tectonic interpretation for the Northern Dabieshan, central China. J. Metamorph. Geol., 14:271–284. Zhang, R.Y., Liou, J.G., Yang, J.S. & Yui, T.-F. (2000a): Petrochemical constraints for dual origin of garnet peridotites from the Dabie-Sulu UHP terrane, eastern-central China. J. Metamorph. Geol., 18:149–166. Zhang, Z., Xu, Z. & Xu, H. (2000b): Petrology of ultrahigh-pressure eclogite from ZK703 drillhole in the Donghai, east China. Lithos, 52:35–50. Zhang, R.Y., Shau, Y.H., Liou, J.G. & Lo, C.H. (2002): Discovery of clinoenstatite in garnet pyroxenites from the Dabie-Sulu ultrahigh-pressure terrane, east-central China. Am. Mineral., 87:867–874. Zhao, Z., Wang, Q. & Cong, B. (1992): Coesite-bearing ultrahigh pressure metamorphic rocks from Donghai, northern Jiangsu province, eastern China: “foreign” or “in situ”?. Sci. Geol. Sin., 1:43–58.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 6, 145–187

The Bohemian Massif and the NW Himalaya HANS-JOACHIM MASSONNE1* and PATRICK J. O’BRIEN2 1

Institut für Mineralogie und Kristallchemie, Universität Stuttgart, Azenbergstr. 18, D-70174 Stuttgart, Germany 2 Institut für Geowissenschaften, Universität Potsdam, Postfach 60 15 53, D-14415 Potsdam, Germany; * e-mail: [email protected]

Introduction Although the occurrence of eclogites and garnet peridotites in the Bohemian Massif has been known for more than a century, evidence for ultrahigh pressure metamorphism (UHPM) by indicator minerals has been reported only very recently (diamond: Massonne, 1999; coesite: Massonne, 2001a). In contrast, although eclogites were recognised in the Tso Morari area by Berthelsen (1953), the first real petrological investigation of eclogites in the NW Himalaya followed their discovery in Pakistan in the 1980’s (Ghazanfar & Chaudhry, 1986, 1987). The finding of coesite soon after, in both Pakistan and India (O’Brien et al., 1999, 2001; Sachan et al., 2001) indicates UHP metamorphic conditions for these rocks. The timing of detection can, of course, be no criterium for treating both areas in one chapter. Rather it seems to be that both areas are very contrasting, which is certainly true in regard of the outcrop situation. In the well-mapped Bohemian Massif, natural exposures in deep valleys or as cliffs or crags at higher levels are rare and are only supplemented by a few quarries. In the poorly mapped NW Himalaya, the majestic and steep mountains provide excellent outcrops although they are less accessible and cover an enormous area. Further contrasts could also be listed, such that at first glance both areas addressed here seem to be perfect opposites. However, in the subsequent section we will outline the many common features of the HP and UHP areas of the Bohemian Massif and the NW Himalaya within a larger geographical framework. After presenting some detailed petrographic and geochronological information on key areas in both orogenic sections, we will try to interpret these in terms of a continent–continent collision model accounting for the different states of both the Bohemian Massif and NW Himalaya in terms of orogenic development.

Geological setting and geochronological constraints Bohemian Massif In spite of the extensive Permian and Mesozoic cover, several large areas of basement rocks, such as the Iberian Massif, Armorican Massif, Massif Central, and the Bohemian Massif, are exposed in western and central Europe (Fig. 1). They represent the northern

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

146

H.-J. Massonne & P.J. O’Brien

Fig. 1. Location of exposed basement areas of the Late Palaeozoic Variscan orogen in Europe (Fig. 1 of O’Brien, 2000). These are subdivided in those belonging to the Saxothuringian Zone, Moldanubian Zone and other zones. Not shown are fragments of Variscan basement involved in the Cretaceous-Tertiary Alpine orogen (south of the Alpine front).

portion of the Variscan orogen after its denudation in Late Carboniferous times. Much of the southern portion (south of the Alpine front in Fig. 1) of this orogen was again involved in an orogenic event, which took place during Late Mesozoic and Tertiary times. It formed the Mediterranean fold belt. Fragments of Alpine-overprinted Variscan rocks are abundant therein. It has become a common view that the Variscan orogen resulted from colliding continental plates. These were Gondwana and Laurentia-Baltica (or Laurussia). The latter had formed from originally separate plates during the Caledonian orogeny at the end of Early Palaeozoic times and, thus, shortly before the onset of Variscan events. It is assumed that after the end of the Caledonian orogeny (420–410 Ma) a single oceanic basin (Rheic ocean: e.g., Robardet et al., 1990; Tornquist Sea: Oliver et al., 1993) or two (e.g., Rheic ocean + Massif Central–Galician ocean, Matte, 1986; Rheic ocean + subsequently Palaeotethys, Stampfli et al., 2002) and more basins separated Gondwana and Laurentia-Baltica. The extension of the ocean(s) and the time of final closure is still a matter of debate (e.g. narrow: McKerrow et al., 2000; wide: Tait et al., 2000, see also Fig. 2), which also includes the involvement of microplates, such as Armorica (see Tait

The Bohemian Massif and the NW Himalaya

147

Fig. 2. Possible palaeogeographic situations according to von Raumer et al. (2002). (a) 490 Ma – before the amalgamation of Laurentia and Baltica (Caledonian orogeny), (b) 420 Ma – before the onset of the Variscan orogeny.

148

H.-J. Massonne & P.J. O’Brien

et al., 1994: Armorican Terrane Assemblage), Avalonia (Murphy et al., 1999) or the Hun terrane (Stampfli, 1996), between Gondwana and Laurentia-Baltica. The many microplates, except Brunia (Hegner & Kröner, 2000), were probably once contiguous to, or part of Gondwana (Tait et al., 2000; Hartz & Torsvik, 2002). In spite of the above uncertainties, the relative motion of Gondwana towards Laurentia-Baltica, implying the subduction of oceanic crust, is generally accepted. As shown in Figure 2, the subduction, also involving continental crust at a final stage, is assumed by many scientists to have been directed south or southeast (e.g., Franke & Stein, 2000) but the opposite direction is favoured by others (e.g., Matte, 1998). However, because of the bilateral symmetry of the structural elements of the Variscan orogen, two simultaneously active subduction zones with opposite directions, apart from each other or even towards each other, were also taken into consideration by several geoscientists (e.g., Matte, 1986; Franke, 2000). The basin of the Rheic Ocean had disappeared 340 Ma ago at the latest. Only small basins remained or were newly formed. In these basins up to several km thick sequences of Viséan and Late Carboniferous sediments were deposited. Contemporaneously, large volumes of granitoid magmas intruded the Variscan crust. The Bohemian Massif exposes the metamorphic core of the Variscan orogen at its eastern margin. It is a collage of several smaller basement areas differing in age and metamorphic evolution. These are termed Saxothuringian, Moldanubian, Tepla–Barrandian and in the east and northeast: Moravo–Silesian and Lugian = Sudetes (Fig. 3). Between the basement units, regions with Palaeozoic sediments exist that are predominantly only anchimetamorphic. Both sedimentary and metamorphic ages can be contemporaneous (e.g. Franke & Engel, 1986). Nappe tectonics (Matte & Burg, 1981; Tollmann, 1982, Schulmann et al., 1991) as well as major fault and shear zones (Matte et al., 1990; Rajlich, 1990; Zulauf, 1993; Krohe, 1996), which are widespread in the entire northern Variscan area, are responsible for the present collage-like aspect. High-pressure (HP) rocks in the Bohemian Massif, such as eclogites and garnet peridotites, have been known for more than a century. In fact, they are abundant (for a summary see: O’Brien & Carswell, 1993) but are restricted to specific areas (dark in Fig. 3). As pointed out by Willner et al. (2000), the age data related to HP rocks are bimodal. Ages between 400 and 380 Ma, with up to 20 Ma younger cooling ages, were reported from the Münchberg Massif, the Góry Sowie (Owl Mountains) and the Tepla–Barrandian unit (Fig. 3). Clearly younger ages (345–340 Ma) were determined for HP rocks of the Granulitgebirge (Granulite Mountains), the Erzgebirge (Ore Mountains, Krušné hory in Czech), the Śnieżnik (Snowy Mountains) and the Gföhl unit. Cooling ages are on average only a few Ma younger. As suggested by various mineral equilibria (see below), ultrahigh pressure rocks (for a summary see: Massonne, 2000; O’Brien, 2000) may occur in all units characterised by HP rocks but so far indicator minerals (see above) have only been found in the Erzgebirge. Himalaya The Himalaya and adjacent Tibetan Plateau result from collision of the Indian plate with the Asian margin and intervening microplates and arcs. The continent–continent

The Bohemian Massif and the NW Himalaya

149

Fig. 3. Simplified geological map of the Bohemian Massif according to Willner et al. (2000). In addition to major units, specific areas are named (italics and 1 = Münchberg Massif, 2 = Zone of Erbendorf-Vohenstrauß, 3 = Gföhl Unit). Grey dots point to locations with HP rocks geochronologically dated.

collision started about 50 Ma ago (le Fort, 1975; Johnson, 2002). The microplates are a consequence of the Permo–Carboniferous disintegration of Pangea whereas northward drifting, arc formation and collision resulted from the subduction of Neo- and PalaeoTethys (Dercourt et al., 1993; Șengör & Natal’in, 1996). Oceanic subduction produced an Andean-type continental arc (Trans-Himalayan batholith) as well as an oceanic island arc, the Kohistan–Ladakh arc, which collided with Asia at the end of the Cretaceous resulting in formation of minor blueschists (80–100 Ma) on its southern margin (Shams, 1972; Honegger et al., 1989). India collided at around 50 Ma (Searle et al., 1987, 1997) as documented by: 1) the youngest suture zone marine sediments (Rowley, 1996); 2) the end of major I-type magmatism in the Trans-Himalayan batholith (Honegger et al., 1982; Petterson & Windley, 1985); 3) the first S-type anatectic granites and migmatites in the Lhasa block; 4) south-directed thrusting in the Asian margin and suture zone; and 5), a change in Indian plate movement (according to palaeomagnetic data) from ca. 15–20 cm/year down to ca. 5 cm/year at around 50 Ma (Patriat & Achache, 1984; Klootwijk et al., 1992).

150

H.-J. Massonne & P.J. O’Brien

The continent–continent collision caused thrusting of the Indian plate margin under Asia (Tibet) and the Kohistan–Ladakh arc, resulting in imbrication and stacking by southward directed thrusts (e.g. Coward et al., 1986, 1988; Zhao et al., 1993). This resulted in metamorphism, magmatism and tectonism as well as in an obvious uplift producing a 2500 km long mountain belt with the world’s highest mountain peaks (e.g. Treloar & Searle, 1993). Part of the Indian plate movement is taken up by strike-slip faulting well away from the collision zone in N Tibet and SE Asia (Tapponnier et al., 1986). The Himalaya is subdivided (Fig. 4) into a series of narrow, parallel belts separated by major thrusts or faults (see reviews of Mattauer et al., 1999 and Hodges, 2000). Ophiolites mark the sutures between Asia, India and the Kohistan–Ladakh arc (Gansser, 1979). South of them lie the units of the Higher Himalaya. This mountain chain itself forms an arc between the syntaxes of Nanga Parbat in NW Pakistan and Namche Barwa in SE Tibet (Wadia, 1931; Burg et al., 1998). In the central Himalaya, the highest mountains, including Mount Everest (Qomolangma), are located south of the suture (Fig. 4) whereas in the syntaxes the crystalline rocks are being exhumed at tremendous rates in a position lying north of the former continental margin, i.e. protruding through the arc (Burg et al., 1998; Schneider et al., 1999). Centrally (e.g. in Nepal), weakly metamorphosed Palaeozoic and Mesozoic sedimentary series of the Indian margin, the Tethyan Himalaya, are separated by a normal fault, the South Tibetan Detachment (STD), from underlying high amphibolite to granulite facies gneisses belonging to the Central Crystallines or Higher Himalayan Crystalline Nappes (HHC). These form the main metamorphic unit of the Himalaya (Burchfiel et al., 1992; Hodges, 2000 and references therein). The STD is the extensional upper boundary to an extrusion wedge

Fig. 4. Simplified geological map of the Himalaya showing the location of eclogite-bearing units.

The Bohemian Massif and the NW Himalaya

151

of HHC rocks bounded on its lower side by a thrust termed the Main Central Thrust (MCT). Metamorphism in the HHC is represented by an early Barrowian staurolite + kyanite-bearing stage (6–10 kbar), yielding ages mostly older than 30 Ma, overprinted by a younger (15–22 Ma), higher temperature but lower pressure metamorphism (4–8 kbar). It produced sillimanite-bearing assemblages, migmatites and leucogranites (Guillot et al., 1999; Harris & Massey, 1994). In the region of the Kohistan–Ladakh arc, exhumed HHC rocks occur directly adjacent to the ophiolitic melange of the suture zone. It is in these regions where eclogites have been described. Tectonically below the HHC, below the MCT, are found the generally lower grade greenschist facies series forming the Lesser Himalaya (LH). This unit is thrust over underlying molasse sediments along the Main Boundary Thrust (MBT), which are themselves bounded in the south by the Main Frontal Thrust (MFT). Earthquakes reveal that MFT, MBT, MCT, and STD are linked at depth (28 to 40 km) to a single thrust (Ni & Barazangi, 1984; Zhao et al., 1993). Other geophysical studies show the Moho at ca. 70 km depth below southern Tibet (Hirn et al., 1984; Zhao et al., 1993; Nelson et al., 1996) and also that subduction of Indian lithospheric mantle (and probably also lower crust) occurs significantly further north than the trace of the original suture marked by the ophiolites. Evidence for high pressure metamorphism is scarce in the Himalaya. So far only four areas have been identified with eclogites and eclogite facies rocks. These are: 1) the Upper Kaghan and Neelum Valleys, Pakistan (Ghazanfar & Chaudhry, 1986, 1987; Spencer et al., 1990; Pognante & Spencer, 1991; Fontan, 1998; O’Brien et al., 1999, 2001; Lombardo et al., 2000; Fontan et al., 2000); 2) the Tso Morari crystalline complex, Ladakh, India (Berthelsen, 1953; Guillot et al., 1995, 1997; de Sigoyer et al., 1997, 2000; Sachan et al., 1999, 2001); 3) in the Indus suture zone at the margin of the Nanga Parbat–Haramosh massif (Le Fort et al., 1997) and 4) in the Kharta region of southern Tibet, 30 km east of the Everest–Makalu massif (Lombardo et al., 1998; Lombardo & Rolfo, 2000). These will be dealt with in turn later.

Petrological information on HP/UHP key areas and rocks Bohemian Massif Metabasites Eclogites and their equivalents after retrogression are in fact a volumetrically minor rock type but are found in several areas of the Bohemian Massif (see above). They appear as numerous lensoid bodies up to several km in length (e.g. Weissenstein of the Münchberg Massif, Meluzína–Křížová hora of the Erzgebirge). Most bodies are, however, in the tens of metres range but can also be less than one metre. The host rocks of the (retrograded) eclogites are chiefly pelitic to quartzofeldspathic rocks (felsic granulites, gneisses, mica schists) and rarely also ultramafic rocks although garnet clinopyroxenites (see subsequent section) are more common in this association. Closer investigation of the many eclogite bodies by mineral analytical studies started in the 1960’s and 1970’s (e.g. Matthes et al., 1969; Dudek & Fediuková, 1974). P–T conditions for eclogites were evaluated by later works and are now known from numerous occurrences in the Bohemian Massif (see reviews of O’Brien & Carswell,

152

H.-J. Massonne & P.J. O’Brien

1993; O’Brien, 2000). The temperature data are based almost exclusively on the Mg–Fe2+ partitioning between garnet (GT) and clinopyroxene (CPX) according to the equilibrium 3 (CaMgSi2O6)CPX + (Fe3Al2Si3O12)GT = 3 (CaFeSi2O6)CPX + (Mg3Al2Si3O12)GT. (1) Rarely, reactions such as the lawsonite breakdown to zoisite + kyanite + quartz + H2O (R1) or the equilibrium related to the Mg–Fe2+ partitioning between garnet and phengite (PHE), 3 (KMgAlSi4O10(OH)2)PHE + (Fe3Al2Si3O12)GT = 3 (KFeAlSi4O10(OH)2)PHE + + (Mg3Al2Si3O12)GT,

(2)

were considered. Most temperature estimates for eclogites from the Münchberg Massif (Franz et al., 1986; Klemd, 1989; Stosch & Lugmair, 1990; Bosbach et al., 1991; O’Brien, 1993), the Tepla–Barrandian unit (O’Brien, 1991; Medaris et al., 1995a), South Bohemia (O’Brien & Vrána, 1995), and the zone of Erbendorf–Vohenstrauß (O’Brien et al., 1992) including the area around the village of Winklarn (O’Brien, 1989) are between 600 °C and 700 °C. Typically, these eclogites contain phengite and other hydrous eclogitic minerals, such as amphiboles and zoisite (Fig. 5). A few of these eclogites might have experienced somewhat higher temperatures (see, e.g., O’Brien & Vrána, 1995) that are also confirmed by melt inclusions (Fig. 5) in eclogitic garnet from a kyanitebearing eclogite of the Münchberg Massif (Massonne, 1993). The Fig. 5. Photomicrographs (crossed nicols) of eclogites from the Münchberg Massif. (a) Phengite (in the middle) coexisting with garnet, quartz, rutile, and fresh omphacite in the Weissenstein eclogite. Amphibole is in the upper left corner. Image width is 2 mm. (b) 200 2m-sized inclusion, consisting of intergrown quartz, K-feldspar and minor plagioclase, in garnet from the eclogite near the village of Oberkotzau. It is interpreted as melt inclusion appearing in an inner garnet growth zone, which is surrounded by a zone with abundant mineral inclusions and an inclusion-free outer zone. Fresh omphacite can be seen outside the garnet. Image width is 0.8 mm.

The Bohemian Massif and the NW Himalaya

153

retrogression of the eclogites (Fig. 6) happened at isothermal conditions or at slightly decreasing temperatures during uplift (O’Brien, 1989, 1993; O’Brien et al., 1992) although increasing temperatures causing granulite facies overprint are common further south (O’Brien, 1989; Messiga & Bettini, 1990; O’Brien & Vrána, 1995). Typically, metamorphic age data (U–Pb, Sm–Nd, Rb–Sr) for the above eclogites cluster in the range 400 to 380 Ma (Münchberg Massif: Gebauer & Grünenfelder, 1979; Stosch & Lugmair, 1990; Erbendorf–Vohenstrauß: von Quadt, 1990; O’Brien et al., 1997a; Winklarn: von Quadt & Gebauer, 1993; Tepla–Barrandian: Beard et al., 1995). A few older but still Palaeozoic ages were related to the formation of protoliths. Cooling ages (Rb–Sr, K–Ar, Ar–Ar), for instance, for eclogites of the Münchberg Massif fall mainly in the range 380 to 365 Ma (Söllner et al., 1981; Kreuzer et al., 1989; Stosch & Lugmair, 1990; Hammerschmidt & Franz, 1992). The relatively short time interval between the eclogite facies stage and a stage after considerable exhumation (or at least cooling) is also confirmed by chemical zonation patterns of garnet to which the method of cation diffusion modelling was applied (O’Brien, 1997). Eclogites and their strongly retrogressed equivalents from the Granulitgebirge (Romer & Rötzler, 2001), the Erzgebirge (Schmädicke et al., 1995), Śnieżnik (Brueckner et al., 1991; Bröcker et al., 1997), and the Gföhl unit (van Breemen et al., 1982; Medaris et al., 1995b) gave considerably younger ages mainly between 345 and 340 Ma that were related to a HP event. Ar–Ar cooling ages can be surprisingly close to these HP metamorphic ages (see, e.g., Bröcker et al., 1997, Werner & Lippolt, 2000) pointing to fast uplift. This is also confirmed by strong growth zonation of garnet in eclogites that were overprinted under granulite facies conditions. The corresponding cation diffusion modelling resulted in minimum uplift rates of several cm/year (O’Brien, 1997) which are significantly higher than those for the older eclogites. Another characteristic feature of the younger eclogites, compared to the old and cold eclogites, is their bimodal peak temperature distribution. Estimated peak Fig. 6. P–T data for eclogites from the Münchberg Massif. Sources are: Klemd (1989, K1 – Weissenstein eclogite), Massonne (1993 and unpublished data, M – eclogite near Oberkotzau) and recalculated P–T values applying mineral core compositions by Franz et al. (1986, F – Weissenstein eclogite) and Equilibria 1, 5 and 7 according to Massonne (1992) and Massonne et al. (1993). In addition, data of Klemd et al. (1994, K2) for a calc-silicate rock adjacent to the Weissenstein eclogite are shown.

154

H.-J. Massonne & P.J. O’Brien

temperatures for the eclogitic event in the Erzgebirge are either between 630 °C and 760 °C and above 835 °C (Schmädicke et al., 1992) or between 600 °C and 700 °C and around 800 °C (Massonne, 1994). Those for the Śnieżnik eclogites are in the ranges 560 to 740 °C and 800 to 900 °C (Smulikowski & Smulikowski, 1985; Bakun-Czubarow, 1989). North of the Śnieżnik, further cold eclogites (570–610 °C: Achramowicz et al., 1997; Bakun-Czubarow, 1998) occur, probably of similar age. However, mainly hot eclogites were reported from the Gföhl unit (Beard et al., 1992; Medaris et al., 1995a, 1998) where only a few estimates are below 800 °C and some are even above 1000 °C. Mineral relics of a hot (> 1000 °C) eclogite facies metabasite were also reported by Rötzler & Romer (2001) from the Granulitgebirge. In case of the Erzgebirge it is obvious that cold and hot eclogites occur in different units (see also Willner et al., 2000). The medium-grade Mica-Schist/Eclogite Unit hosts only cold eclogites. A similar situation might hold for the Gföhl unit as well, because, adjacent to it, cold eclogites (around 650 °C) occur in the Letovice crystalline complex (Konopásek et al., 2002) and other areas belonging to the Monotonous and Varied unit (Medaris et al., 1995a; O’Brien, 2000). Because of the absence of plagioclase (PLAG) during eclogitic stages and before reaching the pressure regimes of the amphibolite or granulite facies, pressure estimates for eclogites on the basis of the equilibria (NaAlSi2O6)CPX + SiO2 = (NaAlSi3O8)PLAG

(3)

2 (Ca3Al2Si3O12)GT + (Mg3Al2Si3O12)GT + 3 SiO2 = = 3 (CaMgSi2O6)CPX + 3 (CaAl2Si2O8)PLAG,

(4)

and

result in minimum values. For instance, for the Münchberg Massif, minimum pressures up to 16 kbar were obtained (Franz et al., 1986; Stosch & Lugmair, 1990; Bosbach et al., 1991). However, the above equilibria proved useful in understanding the P–T range of retrogression of eclogites from the Bohemian Massif (e.g. O’Brien, 1989, 1993). Considering the mineral reaction (R1) and the paragonite breakdown to jadeite + kyanite + H2O (R2), Klemd (1989) estimated a pressure range between 20 and 26 kbar for eclogites of the Münchberg Massif (Fig. 6). Massonne (1993) even gave 34 kbar. This estimate was based on the equilibrium 3 (KAlMgSi4O10(OH)2)PHE + 2 (Ca3Al2Si3O12)GT + (Mg3Al2Si3O12)GT = = 3 (KAl3Si3O10(OH)2)PHE + 6 (CaMgSi2O6)CPX, (5) and an Si-rich phengite inclusion in garnet. However, the quality of these estimates suffers from uncertainties due to speculations on the absence of minerals (paragonite, lawsonite) and assumed mineral coexistence involving inclusions. In case of an eclogite from the Münchberg Massif that additionally contained kyanite, the equilibrium (Ca3Al2Si3O12)GT + (Mg3Al2Si3O12)GT + 2 SiO2 = 3 (CaMgSi2O6)CPX + 2 Al2SiO5,

(6)

could also be taken into account. The results for the slightly variable composition of large grains of garnet, phengite and omphacite, considering Equilibria 1, 5 and 6, ranged between 21 to 26 kbar and 680 to 750 °C (Massonne, 1993), defining a small portion of

The Bohemian Massif and the NW Himalaya

155

the uplift path (Fig. 6). A high Si phengite core composition from the Weissenstein eclogite reported by Franz et al. (1986) was used to recalculate the pressure for an early eclogitic stage considering Equilibria 5. The result of 26 kbar (Fig. 6) is confirmed by applying an additional equilibrium considering the Ti content in phengite coexisting with rutile and quartz/coesite: (KAl3Si3O10(OH)2)PHE + 2 TiO2 = 2 SiO2 + (KAl3SiTi2O10(OH)2)PHE. (7) For cold eclogites from the Erzgebirge, pressures between 22 to 26 kbar were also

Fig. 7. Photomicrographs of a cold eclogite from the Mica-Schist/Eclogite Unit (a) and hot eclogites from the Gneiss/Eclogite Unit (b–e) of the Erzgebirge. (a) Amphibole porphyroblast (Massonne, 1992) under crossed nicols. The porphyroblast hosts garnet, quartz, and minor phengite, whereas omphacite is present only in the surrounding matrix. Image width is 2.5 mm. (b) Coesite relic in garnet (sample E99-24; Massonne, 2001a) under plane polarised light. Semi-opaque phase in coesite is rutile. Image width is 0.6 mm. (c) Same object as (b) under crossed nicols. Typically, palisade quartz formed around coesite. (d) Phengite and omphacite enclosed in garnet under crossed nicols. In the matrix, omphacite is completely altered and phengite is rimmed by symplectites containing biotite. Image width is 2 mm. (e) Phengite and a pseudomorph, consisting of quartz and K-feldspar, enclosed in omphacite (Massonne et al., 2000) under crossed nicols. Image width is 0.4 mm. (f) Pseudomorph of (e) shown by a backscattered electron image. Dark: quartz, light: K-feldspar.

156

H.-J. Massonne & P.J. O’Brien

estimated (Massonne, 1992; Schmädicke et al., 1992; Klápová et al., 1998). Because of similar mineral compositions (e.g. phengite cores with Si close to 3.5 Si per formula unit), this near-UHP regime and a corresponding geotherm less than 8 °C/km seem to be characteristic for all the cold eclogites, independent of age. Trace element patterns, mainly considering rare earth elements (REE), as well as Nd and Sr isotope ratios of the cold eclogites (Schmidt et al., 1990; Stosch & Lugmair, 1990; Bosbach et al., 1991; Okrusch et al., 1991; Beard et al., 1995) point in most cases to mid-ocean ridge basalts (MORBs) as protoliths. This conclusion can probably be made for hot eclogites as well, except for those occurring in peridotites which have formed in the mantle (Beard et al., 1992; Medaris et al., 1995b). The protoliths of a minority of eclogites were also related to other types of basalts or to cumulate (gabbroic) sources (e.g. high Al basalts in case of the kyanite-bearing eclogites of the Münchberg Massif: Bosbach et al., 1991; within-plate basalts for cold and young (340 Ma) eclogites of the Sudetes: Bakun-Czubarow, 1998) to account for Nd isotopic and REE features deviating from those of MORBs. Other explanations of the deviations refer to chemical alteration of (meta-)MORBs by hydrous fluids near the surface or already deep in the subduction zone. The interaction of a hydrous fluid phase with eclogites during initial exhumation was suggested by Massonne (1992) applying geohygrometry to a cold eclogite from the Erzgebirge. The interaction had resulted in the formation of amphibole porphyroblasts (Fig. 7). This phenomenon was also reported by Smulikowski & Smulikowski (1985) and Franz et al. (1986) for the Śnieżnik and Münchberg eclogites, respectively. The peak pressure estimates for the non-peridotitic hot eclogites are problematic. At least for those from the Erzgebirge, Schmädicke (1991) reported UHP conditions derived from pseudomorphs after coesite. Similar observations were made by BakunCzubarow (1992) on Śnieżnik eclogites. However, it was shown by Massonne et al. (2000) that such pseudomorphs in hot eclogites from the Erzgebirge consist of quartz intergrown with K-feldspar (Fig. 7), whereas true coesite, as it was finally found in a hot eclogite from the Erzgebirge by Massonne (2001a), transforms to quartz only (Fig. 7). The quartz–K-feldspar intergrowths were interpreted by Massonne et al. (2000) as symplectites replacing K-cymrite (KAlSi3O8·H2O). As the lower pressure limit of the stability field of this phase is similar to that of coesite (see Massonne, 1995), minimum peak pressures for the hot eclogites might be in the range 28 to 31 kbar (see Fig. 8). Such pressures would also result from phengite barometry (Equilibria 5 and 7), considering compositions of inclusions and cores of phengites which show maximum Si contents around 3.35 per formula unit only (see Schmädicke et al., 1992; Massonne, 2001a). A further possible hint at UHP conditions are tiny quartz rods in omphacite reported by Bakun-Czubarow (1992) for a Śnieżnik eclogite and by Schmädicke & Müller (2000) for a hot eclogite from the Erzgebirge. However, the reaction phengite + garnet = phlogopite + kyanite (+ quartz) (R3), as suggested by Schmädicke & Müller (2000), is no evidence for UHP as, although in the Fe-free model system (R3) runs at about 28 kbar (Massonne, 1995), in the presence of Fe this reaction occurs at only moderately high pressures and is widespread in eclogites and HP metapelites.

The Bohemian Massif and the NW Himalaya

157

Fig. 8. P–T–d(–t) evolution for various rock types from different units of the Erzgebirge according to Willner et al. (2000). (a) Mica-Schist/Eclogite Unit including phyllite units. (b) Gneiss/Eclogite Unit.

H.-J. Massonne & P.J. O’Brien

158

Ultramafic rocks Garnet-bearing ultramafic rocks, occurring as bodies up to several km in size, are known from several areas of the Bohemian Massif. These rocks show evidence for HP and UHP conditions contrasting with other ultramafic bodies in the Bohemian Massif by mineralogical (e.g. presence of spinel and no garnet) and geological relationships. The latter ultramafic rocks, typically appearing in the non-HP units of the Bohemian Massif, are either interpreted as basic layered intrusions (Vejnar, 1986) or as part of ophiolite complexes (e.g. Werner, 1981; Jelinek et al., 1984; Mísař et al., 1984). The oceanic crust of the ophiolites could have formed as late as Lower Carboniferous (e.g. Pin et al., 1988). In spite of several hints at UHP conditions experienced by eclogites about 380–400 Ma ago, the rare and relatively small ultrabasic bodies adjacent to them, also showing similar ages (Góry Sowie: Brueckner et al., 1996; Winklarn: von Quadt & Gebauer, 1993), are spinel peridotites (Münchberg Massif: Gayk et al., 1995a; Góry Sowie: Dubińska et al., 1999; Tepla–Barrandian: Medaris et al., 1995a), giving no evidence for UHP conditions. For the serpentinite bodies near Winklarn it is assumed that the Al enrichment at the rim of spinel and orthopyroxene resulted from garnet breakdown (Okrusch et al., 1991) probably by pressure release. Fresh garnet coexists with clinopyroxene there only in pyroxenites, for which temperatures around 700 °C (pressure unconstrained) were determined (Okrusch et al., 1991), or in websterites, which yielded P–T conditions of 28 kbar at 700 °C. Small peridotite bodies in the Góry Sowie contain garnet in addition to mm-sized spinel grains (Brueckner et al., 1996) formed at P–T conditions as high as 27 kbar and 1030 °C. In HP areas with younger eclogites, numerous large peridotite bodies are embedded in gneissic rocks. In spite of serpentinisation, garnets (or easily discernable pseudomorphs) are typical constituents of the former peridotites. Garnet pyroxenites are minor but omnipresent in the ultramafic bodies. U–Pb and some Sm–Nd ages of ultrabasic rocks (Granulitgebirge: von Quadt, 1993; Erzgebirge: Schmädicke et al., 1995; Gföhl unit: Carswell & Jamtveit, 1990; Medaris et al., 1995b; Becker, 1997) scatter around 345 Ma and are, thus, similar to those of the adjacent eclogites. A few Sm–Nd ages are, however, significantly older (Carswell & Jamtveit, 1990; Brueckner et al., 1991; Beard et al., 1992; von Quadt, 1993) with no obvious systematics as also observed for eclogites. These ages are related to early metamorphic or protolith stages. The P and T conditions experienced by the peridotites were estimated from the Al content in orthopyroxene (OPX), according to the equilibrium 3 (MgSiO3)OPX + (Al2O3)OPX = (Mg3Al2Si3O12)GT,

(8)

2+

and from the Mg–Fe distribution among the pairs garnet–clinopyroxene, garnet–olivine (OL), and garnet–orthopyroxene using Equilibria 1 and equilibria 3 (Mg2SiO4)OL + 2 (Fe3Al2Si3O12)GT = 3 (Fe2SiO4)OL + 2 (Mg3Al2Si3O12)GT, (9) and (Fe3Al2Si3O12)GT + (MgSiO3)OPX = (Mg3Al2Si3O12)GT + (FeSiO3)OPX,

(10)

respectively. The data obtained for garnet-bearing ultrabasites from the Gföhl terrane (Medaris et al., 1990; Carswell, 1991) and the Erzgebirge (Schmädicke & Evans, 1997)

The Bohemian Massif and the NW Himalaya

159

scatter between 55 kbar/1200 °C and 25 kbar/900 °C (see Brueckner & Medaris, 1998). Massonne & Grosch (1995) used additionally the equilibrium 2 (Ca3Al2Si3O12)GT + (Mg3Al2Si3O12)GT = 3 (CaAl2SiO6)CPX + 3 (CaMgSi2O6)CPX, (11) estimating a pressure decrease from 40 kbar to 25 kbar for a garnet peridotite from the Erzgebirge. At the final stage, temperatures were close to 1000 °C. Although somewhat differently interpreted by Medaris et al. (1990), the above P–T data lie on an array that could represent the P–T path for the uplift of the investigated ultrabasites. This P–T path is supported by the Al zonation of orthopyroxene showing lowest Al contents in the core and highest at the rim (Fig. 9). The maximum P derived in various works might not necessarily be the peak pressure. It rather reflects the minimum Al content in orthopyroxene (accidentally) observed. The Cr zonation in garnet, as reported by Massonne & Grosch (1995) and shown in Figure 9, as well as the observation of small grains of chromian spinel in the matrix and as inclusion in Cr-poor cores of such garnet reflect not only the above discussed P–T path for the uplift but also a burial stage starting from relatively low pressures corresponding to the spinel peridotite P–T field (see also Schmädicke & Evans, 1997). The breakdown of spinel with P increase resulted in Cr-rich intermediate growth zones of garnet (Fig. 9). The T estimates for garnet pyroxenites in the peridotites led to data virtually identical to those of the peridotites (e.g., Scharbert & Carswell, 1983). Massonne & Bautsch (2002) used Equilibria1 and 11 to determine the uplift path for a garnet clinopyroxenite from the Granulitgebirge with unusual lamellar fabric (Fig. 9) originating by exsolution from former megacrysts (Reiche & Bautsch, 1984; Jekosch & Bautsch, 1991). The derived path starts at 25 kbar fitting the above P–T array. However, it is suggested from patterns of trace elements (particularly rare earth elements: (Yb/Nd)N > 10) that the rock originally consisted of majoritic garnet (Si = 3.5 p.f.u.) and was, thus, transported from the transition zone of the upper mantle to crustal levels (Massonne & Bautsch, 2002). Quartzofeldspathic and related rocks In general, the quartzofeldspathic country rocks, hosting the various eclogitic and ultramafic bodies, show the same metamorphic ages (and cooling ages) as the basic and ultrabasic HP and UHP rocks (Münchberg Massif and zone of Erbendorf–Vohenstrauß: Kreuzer et al., 1989; Góry Sowie: van Breemen et al., 1988; O’Brien et al., 1997b; Bröcker et al., 1998; Granulitgebirge: von Quadt, 1993; Kröner et al., 1998; Romer & Rötzler, 2001; Erzgebirge and adjacent areas (Ohře Crystalline Complex): Kotková et al., 1995; Kröner & Willner, 1998; Śnieżnik: Borkowska et al., 1990; Gföhl unit: see summary in Petrakakis, 1997; Kröner et al., 2000). In addition, peak temperatures of the country rocks are at least similar to those of the eclogite bodies. For instance, in the Gföhl unit, hot eclogite lenses are hosted by high temperature granulites that had experienced P–T conditions above 15 kbar and 900 °C (Carswell & O’Brien, 1993; Kröner et al., 2000). However, thermal equilibration on a regional scale happened at lower T and P close to 800 °C and 11 kbar (Petrakakis & Jawecki, 1995). Similar observations were made for felsic to intermediate granulites from the Śnieżnik and the Granulitgebirge for which pressures were estimated to be as high as 30 kbar (Kryza et al., 1996) and 22 kbar (Rötzler & Romer, 2001), respectively, at T around

160

H.-J. Massonne & P.J. O’Brien

Fig. 9 (a–f). Photomicrographs, chemical zonation patterns of minerals and P–T data of garnet-bearing ultrabasites from the Bohemian Massif. (a) Orthopyroxene from a partially serpentinised garnet peridotite (E48e) sampled SW of Seiffen, Saxonian Erzgebirge, under crossed nicols. Image width is 1.1 mm. (b) Al zonation map of (a). Scale bar is 100 µm. Increasing Al concentrations are shown by lighter grey tones. (c) Garnet (Gt), clinopyroxene (Cpx), orthopyroxene (Opx) and partially preserved olivine under plane polarised light. Same rock as (a). Image width is 1.0 mm. (d) Cr zonation map of (c); grey tones as in (b). (e) Lamellar fabric of a garnet clinopyroxenite (Massonne & Bautsch, 2002) occurring as layers in garnet serpentinite from Reinsdorf, Granulitgebirge. Crossed nicols. Image width is 4.5 mm. (f) Recrystallised fabric close to (e). Image width is 9 mm.

The Bohemian Massif and the NW Himalaya

161

Fig. 9. (g) P–T diagram showing the range for garnet peridotites of the Gföhl unit according to Brueckner & Medaris (1998; BM) as well as exhumation paths of ultramafic rocks from the Granulitgebirge (Massonne & Bautsch, 2002; MB) and the Erzgebirge (Massonne & Grosch, 1995; MG; Schmädicke & Evans, 1997; SE).

1000 °C. The regional medium pressure equilibration also took place at temperatures close to 800 °C and depths corresponding to 12 to 17 kbar (Reinhardt & Kleemann, 1994; Kryza et al., 1996). The above HP estimates were based on the equilibrium (Ca3Al2Si3O12)GT + 2 Al2SiO5 + SiO2 = = 3 (CaAl2Si2O8)PLAG, (12) and the ternary feldspar compositions (T estimate) as well as Equilibrium 1 + Equilibrium 4 in the presence of clinopyroxene. In the medium pressure decompression stages, biotite (BT) had newly formed so that equilibria such as 3(KMg3AlSi3O10(OH)2)BT + (Fe3Al2Si3O12)GT = = 3 (KFe3AlSi3O10(OH)2)BT + (Mg3Al2Si3O12)GT

(13)

2 (KMg3AlSi3O10(OH)2)BT + 3 SiO2 + 3 (CaAl2Si2O8)PLAG = = 2 (Mg3Al2Si3O12)GT + (Ca3Al2Si3O12)GT + 2 KAlSi3O8 + 2H2O

(14),

and

could be additionally considered. For the determination of the peak pressure in quartzofeldspathic rocks hosting hot eclogites in the Gneiss/Eclogite Unit of the Erzgebirge, Willner et al. (1997) used Equilibrium12 and phengite compositions (Si up to 3.35 p.f.u.) in equilibria (Mg3Al2Si3O12)GT + 2 (Ca3Al2Si3O12)GT + 3 (KAl3Si3O10(OH)2)PHE + (15) + 6 SiO2 = 3 (KMgAlSi4O10(OH)2)PHE + 6 (CaAl2Si2O8)PLAG and 3 (KMgAlSi4O10(OH)2)PHE + 4 Al2SiO5 = 4 SiO2 + + (Mg3Al2Si3O12)GT + 3 (KAl3Si3O10(OH)2)PHE,

(16)

as well as Equilibrium 7, all being independent of water activity. For the determination

162

H.-J. Massonne & P.J. O’Brien

of P–T conditions during the uplift stage, retrograde biotite could be considered again by applying Equilibrium 13, the equilibrium (Mg3Al2Si3O12)GT + 3 (KAl3Si3O10(OH)2)PHE = 6 SiO2 + + (KAl3Si3O10(OH)2)PHE + 2 (KMg3AlSi3O10(OH)2)BT,

(17)

and various water-dependent equilibria. Peak P–T values resulted being somewhat above 20 kbar and 800 °C (Fig. 8). The retrograde stage was characterised by conditions around 10 kbar and 700 °C (Willner et al., 1997). Country rocks from the MicaSchist/Eclogite unit of the Erzgebirge, hosting cold eclogites, contain abundant phengites (Si up to 3.45 p.f.u.). They had experienced P–T conditions in the range 13 kbar – 500 °C to 15 kbar – 600 °C (Konopásek, 1998; Rötzler et al., 1998). A retrograde stage occurred at about 8 kbar and 550 °C (Rötzler et al., 1998). A similar exhumation path was deduced by Klemd et al. (1991) for metasediments in contact with (UHP?) eclogites from the Münchberg Massif. Similar retrograde conditions (8 kbar) were observed for quartzofeldspathic rocks of the zone of Erbendorf–Vohenstrauß, too (Reinhardt, 1990). However, due to lack of data, it cannot be generalised that the cold eclogites of the Bohemian Massif are hosted by medium to high pressure quartzofeldspathic rocks. For the hot crustal environments of the Bohemian Massif and the medium grade example (Erzgebirge) it can be concluded that significant pressure differences between eclogites + ultrabasic slices and country rocks (e.g. HP granulites) are noticeable. Such differences amount to, at least, several kbar but also to more than 10 kbar (see e.g. Fig. 8). However, we also know several examples of quartzofeldspathic and related rocks from the Bohemian Massif that experienced (near-)UHP conditions similar to those of eclogites and ultramafic rocks. Klemd et al. (1994) reported minimum pressures of 31 kbar at 630 °C for calc-silicate rocks from the Münchberg Massif. Nearby, Gayk et al. (1995b) found quartz laths as exsolution from clinopyroxene in an intermediate granulite. These authors estimated 25 kbar but temperatures above 900 °C for this occurrence. Kotková et al. (1997) observed a peculiar relic assemblage of garnet–orthopyroxene–quartz in a rock from the Gföhl unit (Podolsko Complex). They determined 28 kbar and 830 °C considering Equilibrium 8 and 10. In the same unit, a metasediment (calc-silicate marble), containing clinopyroxene with exsolution of alkali feldspar, was found by Becker & Altherr (1992). These authors estimated minimum pressures of 30 kbar at temperatures above 1100 °C for this rock. Konopásek (2001) reported a mica schist, containing kyanite and garnet, from the Mica-Schist/Eclogite Unit of the Erzgebirge that had equilibrated at 22 kbar and 640 °C. However, most impressive are the microdiamond-bearing quartzofeldspathic rocks from the Gneiss/Eclogite Unit of the Erzgebirge detected by Massonne (1999). Microdiamonds, confirmed by RAMAN spectroscopy (Nasdala & Massonne, 2000), are enclosed in garnet, kyanite, and zircon (Fig. 10), pointing to minimum pressures of 45 kbar. Temperatures were estimated on the basis of the equilibrium (Mg3Al2Si3O12)GT + 3 TiO2 = 3 SiO2 + (Mg3Al2Ti3O12)GT,

(18)

to 1000 °C. However, the appearance of TiO2 with ,-PbO2 structure in a microdiamondbearing quartzofeldspathic rock (Hwang et al., 2000) requires P–T conditions as high as

The Bohemian Massif and the NW Himalaya

163

80 kbar and, according to Equilibrium 18, 1200 °C (Fig. 10). Further constraints for the P–T evolution came from the compositions of garnet and jadeite enclosed in zircon cores (Massonne & Nasdala, 2003), dated by SHRIMP work to 336.1 ± 2.2 Ma (Massonne et al., 2001), and from phengite inclusions (Si ~ 3.24 p.f.u.) in garnet cores (Fig. 10). An early metamorphic stage at 18 kbar and 650 °C resulted by considering Equilibrium 2, 3, 7, 16 and 18. The P–T conditions of a late stage, dated by Massonne et al. (2001) to 329.9 ± 3.6 Ma (zircon rims) and 332.5 ± 3.8 Ma (monazite), was estimated from the compositions of the matrix minerals to be close to 15 kbar and 700 °C (Massonne, 1999). This evolution path requires a fast submersion of (a fragment of) continental crust that was previously thickened to at least 60 km. The fast uplift from depths of 200 km or more is explained by anatexis in the depth and the rise of a considerably molten rock (Massonne, 2001b). Diamonds crystallised from a fluid, as proven by Dobrzhinetskaya et al. (2001), Hwang et al. (2001), and Stöckhert et al. (2001), during uplift accompanied by significant cooling (Massonne, 2001b) as indicated by decreasing Ti contents in the garnet zone (Equilibrium 18) where microdiamonds are enclosed. The hardly foliated garnet–muscovite-rich quartzofeldspathic rocks with microdiamonds form several E–W-extended lens-like bodies up to several hundred metres in length (Fig. 11). As for neighbouring eclogite and garnet peridotite bodies, they are embedded in HP gneisses of the Gneiss/Eclogite Unit (see above) from which they can be distinguished by fabric (“Fels” in the old geological maps; see Hazard, 1886) but also by geochemical signatures (Düffels & Massonne, 2001). The present position of the ca. 340 Ma old UHP rocks from the Erzgebirge on top of a normally-thick crust is the result of widespread crustal extension in E–W direction (Krohe, 1996; see also Fig. 8) following a N–S compressional regime as also suggested by Konopásek et al. (2001). This deformational event happened in the time interval 335–320 Ma for which no (near-)UHP metamorphism could be proven anymore. As shown above, quartzofeldspathic and related rocks can conserve a mineralogical record of (near-)UHP conditions in the same quality as basic and ultrabasic rocks. Thus, it seems to be rather a prejudice that HP–UHP quartzofeldspathic rocks are, in general, completely transformed to retrograde mineral assemblages. Due to the presence of rigid porphyroblasts (garnet) in these rocks, even pre-(U)HP minerals can be conserved. Thus, we must conclude for the Bohemian Massif that there is a true contrast between low-pressure areas and HP areas and that in the latter areas lenses of (near-)UHP rocks with variable lithology are embedded. The extent of (near-)UHP quartzofeldspathic rocks in the HP areas is, however, difficult to evaluate because UHP quartzofeldspathic rocks cannot as easily be distinguished from their medium-pressure equivalents as it is possible for metabasites (eclogites versus amphibolites). Himalaya As stated earlier, eclogites in the Himalaya are restricted to the Kaghan/Neelum Valleys in Pakistan, the Tso Morari area in India, the Indus suture zone at the edge of the Nanga Parbat–Haramosh massif, and the Kharta region east of Mount Everest (Fig. 4). These areas will be addressed in turn.

164

H.-J. Massonne & P.J. O’Brien

Fig. 10. Photomicrographs and P–T path of diamondiferous quartzofeldspathic rocks from the Gneiss/Eclogite Unit of the Erzgebirge. (a) Sample E97-3 showing garnet (Gt) and kyanite (Ky) in a matrix of quartz, phengitic muscovite and plagioclase. Kyanite is considerably replaced by micas and other phases. Arrows point to microdiamonds surrounded by scratches at the polished surface of the rock thinsection. Image width is 4 mm. (b) Zircon with microdiamond inclusions in sample E97-3. Image width is 0.6 mm. (c) Phengite with minor quartz enclosed in an inner growth zone of mm-sized garnet (sample St6105; Massonne & Nasdala, 2003) under plane polarised light. Open box marks multiphase inclusion with microdiamond. Image width is 0.4 mm. (d) Polyphase diamond-bearing inclusion in garnet seen by scanning electron microscopy (Stöckhert et al., 2001). Arrows point to rational mica garnet interfaces. qz = quartz, par = paragonite, phl = phlogopite, ap = apatite. (e) P–T path similar to that of Stöckhert et al. (2001). Here, the presence of TiO2 with ,-PbO2 structure in the rock and the corresponding stability curve by Akaogi et al. (1992) is newly considered.

The Bohemian Massif and the NW Himalaya

165

Fig. 11. Simplified geological map for an area with diamondiferous quartzofeldspathic rocks (originally called gneisses) and coesite-bearing eclogite (sample E99-24) at the Saidenbach reservoir, Saxonian Erzgebirge, Germany, according to Massonne (2001a). Inset shows the major crystalline units of the Erzgebirge.

Kaghan and Neelum Valleys The geological subdivision of the Upper Kaghan and Neelum Valleys (Fig. 12), derived from the studies of Ghazanfar & Chaudhry (1987), Greco et al. (1989), Greco & Spencer (1993), Spencer (1993) and Fontan et al. (2000), includes a high-grade unit (HHC equivalent) fault-bounded against mafic/ultramafic rocks of the Kohistan–Ladakh arc to the north and lower grade marbles and metasediments to the south. The HHC is subdivided into three main units: a basement composed of metagranites and paragneisses, an overlying metagreywacke unit (the first cover of Spencer, 1993) and finally a series of marbles, graphitic schists and amphibolites (the second cover of Spencer, 1993). These three are believed to be of Cambrian to Precambrian, Lower Palaeozoic and Mesozoic age, respectively, based on correlation with unmetamorphosed sequences further south (Ghazanfar & Chaudhry, 1987; Greco et al., 1989; Greco & Spencer, 1993; Fontan et al., 2000) but no firm geochronological data exist. Mafic rocks occur in all units as deformed sheets and boudins, from one metre to tens of metres in thickness and tens of metres in length. They are interpreted as lava flows (in the marble units) and their feeder dykes in the “stratigraphically older” sequences, related to the Permian Panjal Trap volcanism so typical in Indian Kashmir (Honegger et al., 1982; Greco et al., 1989; Greco & Spencer,

166

H.-J. Massonne & P.J. O’Brien

1993). It is in these bodies that eclogites have been identified (Pognante & Spencer, 1991). Magmatic relics are still preserved in many places in the mafic bodies closer to Naran but relict high grade schlieren and magmatic structures are also present in low strain domains in the granitic rocks, even where the latter are adjacent to eclogite boudins, indicating a very restricted deformational response to HP metamorphism (Treloar et al., 2003). The most obvious metamorphic events are of high amphibolite facies (with garnet and kyanite) overprinted by a strong greenschist facies typified by development of albite porphyroblasts in granitic, pelitic and mafic rocks. These blasts are themselves deformed and folded in younger structures. Structurally, the most important feature is a major southward-vergent thrust, recorded in high grade S–C’ fabrics, placing granitic rocks above marble-rich sequences close to Besal. High-grade fabrics were then overprinted by top-to-the-north extensional features at greenschist facies: later E–W open folding occurred (Treloar et al., 2003). Several different eclogite varieties exist, differing in bulk composition (Spencer, 1993; Lombardo et al., 2000; Fontan et al., 2000; O’Brien et al., 2001). The freshest eclogite is massive and fine-grained and contains phengite, rutile, quartz and relics of coesite, surrounded by palisade quartz (Fig. 13a,b), as inclusions in omphacite. Glaucophane-bearing varieties are also present (Lombardo et al., 2000; Fontan et al., 2000) as are quartz-rich and epidote-bearing ones including samples with epidote

Fig. 12. Simplified geological subdivision of the Upper Kaghan and Neelum Valleys showing eclogite locations (modified after Lombardo et al., 2000).

The Bohemian Massif and the NW Himalaya

167

containing probable pseudomorphs after coesite (Fig. 13c). As a result of penetrating hydrous fluids, randomly oriented Mg-hornblende to barroisite crystals to 1 cm in length grew at the expense of omphacite (Fig. 13b) and cm-sized clinozoisite grains grew that contain abundant inclusions of garnet and omphacite grains defining an earlier fabric. In addition, quartz-rich veins with phengite and cm-sized rutile crosscut the eclogite bodies. From Equilibrium 1, 2 and 5, peak conditions of 27–29 kbar, 690–750 °C (Fig. 14) were determined for the coesite eclogite (O’Brien et al., 2001), whereas lower temperatures of up to 650 °C at pressures up to 24 kbar were calculated for coesitefree and glaucophane-bearing samples (Pognante & Spencer, 1991; Lombardo et al., 2000; Fontan et al., 2000). Another important result is the discovery of coesite inclusions in the outer zones of composite zircons in quartzofeldspathic rocks from the neighbourhood of the coesite eclogite (Kaneko et al., 2001). Fig. 13. Photomicrographs of Kaghan eclogites. (a) Fresh eclogite with garnet (Gt), omphacite (Om), rutile (Rt), phengite (Ph) and secondary amphibole (Am). Near Gittidas, plane polarised light, image width 2 mm. (b) Close-up of the centre of (a) showing coesite (Coe) inclusion with quartz corona enclosed in omphacite (Om). Note secondary amphibole (Am) consuming the omphacite. Image width 0.4 mm, plane polarised light. (c) Palisade quartz aggregate possibly after coesite in epidote. Near Gittidas, crossed nicols, image width 0.4 mm.

168

H.-J. Massonne & P.J. O’Brien

SHRIMP dating of the coesite-bearing zircon reveals a magmatic core (253–170 Ma), overgrown by a 50 Ma quartz-bearing zone and, finally, by an outer 46 Ma coesite-bearing zone. These data correspond well with ages of 44 ± 3 Ma (rim of zircon with a Permian core: Spencer & Gebauer, 1996), 49 ± 6 Ma (Sm–Nd, garnet–omphacite), 43 ± 1 Ma (Rb–Sr, phengite) and 39–40 Ma (U–Pb, rutile). The latter data are from Tonarini et al. (1993). K–Ar and Ar–Ar data record a first cooling stage of 35–43 Ma (exhumation after the eclogite facies event) followed by a second stage of 25 Ma interpreted as the final greenschist facies extension (Chamberlain et al., 1991; Treloar, 1997; Fontan et al., 2000).

Fig. 14. P–T plot showing: peak conditions for coesite-bearing Kaghan and Tso Morari eclogites derived from garnet–omphacite–phengite geothermobarometry (Equilibria 1, 2 and 5); the P–T–t evolution of Tso Morari eclogites after Guillot et al. (1997); and typical ranges of P–T values for the Lesser Himalaya (LH) and Higher Himalaya Crystalline (HHC).

The Bohemian Massif and the NW Himalaya

169

Fig. 15. Simplified geological map of the Tso Morari area, showing eclogite locations (after Guillot et al., 1997).

Tso Morari In E Ladakh, India, crystalline rocks of the Tso Morari dome (Fig. 15) are bounded, by detachment faults, to the Indus suture and the Ladakh arc to the north and Tethyan sediments to the south (e.g. Steck et al., 1998). The high-grade sequence comprises Precambrian to Palaeozoic sediments, gneisses derived from Cambro–Ordovician granites (Girard & Bussy, 1999), and cover sequences containing fossiliferous Permian sediments. Mafic lenses in cover and basement are attributed to the Panjal Trap volcanism. In eclogites, first discovered by Berthelsen (1953), inclusion-rich garnet shows a complex two-stage growth (Fig. 16). In the matrix, omphacite has been partially consumed by glaucophane or barroisite (Fig. 16). Phengite, zoisite and rutile are also characteristic (de Sigoyer et al., 1997). According to de Sigoyer et al. (1997), the eclogite facies stage (20 ± 3 kbar, 580 ± 60 °C) was followed by isothermal decompression associated with glaucophane growth at around 11 ± 2 kbar (Fig. 14). Reworking of the data by O’Brien et al. (2001; Eqn. 5) deduced pressures in the coesite field (Fig. 14) which was subsequently confirmed by identification of coesite in similar material by

170

H.-J. Massonne & P.J. O’Brien

Fig. 16. Photomicrographs and garnet compositional variation in a single Tso Morari eclogite sample. (a) Garnet (Gt) with internal S-shaped fabric of inclusions (dashed line) in a matrix of omphacite (Om) heavily replaced by secondary amphibole (Am). Width of view is 2 mm. (b) Garnet (Gt) with inclusion-rich core and inclusion-poor rim (top) cut by a vein containing amphibole (Am) and phengite (Ph). Width of view is 2 mm. (c) Ca composition map (light is high Ca, dark is low Ca) for garnet showing the very sharp boundary between inclusion-rich core and inclusion-poor overgrowth. Width of view is 3 mm.

Sachan et al. (2001). High-pressure metamorphism is documented in metasediments of Fe-rich, Mg-rich, intermediate pelite and greywacke composition (Guillot et al., 1997). Intermediate metapelites contain staurolite + kyanite whereas jadeite + chloritoid + paragonite + garnet in Fe-rich metapelites formed at 20 ± 2 kbar and 550 ± 50 °C, with glaucophane appearing later during cooling at still high pressure (Guillot et al., 1997). Dating of different phases in the eclogites by various methods yielded ages around 55 Ma for the peak metamorphism (de Sigoyer et al., 2001). A subsequent amphibolite facies overprint at slightly higher temperature (610 ± 70 °C) was dated at 45–48 Ma. The retrogression into the greenschist facies had occurred by 30 ± 1 Ma (de Sigoyer et al., 2000). The data indicate a two stage history with early exhumation being much faster (> 4 mm/year) than the later evolution (1–2 mm/year). Diffusion modelling of garnet overgrowth composition steps (O’Brien & Sachan, 2000) predicts an even faster initial exhumation rate (23–45 mm/year) with the UHP to low greenschist stages all occuring within ca. 3 Ma (45–48 Ma): a result still consistent with the errors in the isotopic methods (Fig. 17). The rapid exhumation is further substantiated by zircon fission track ages of 35–45 Ma recording cooling through the partial annealing zone of 300–200 °C (Schlup et al., 2001). Indus suture zone, Nanga Parbat–Haramosh Massif The Indian crust was thrust northwards beneath the Kohistan–Ladakh arc. This is well documented in the Nanga Parbat–Haramosh Massif which represents a “tongue” of Indian crust (now amphibolite to granulite facies gneisses) that has forced its way up through the arc in the last 10–12 Ma creating

The Bohemian Massif and the NW Himalaya

171

the world’s 9th highest mountain peak (Zeitler et al., 1993). All around this “tongue” the boundary with the arc represents the Indus suture and it is in this marginal zone, where the Indus River cuts the massif at Stak, that retrograded, previously bimineralic eclogites occur within the gneisses (le Fort et al., 1997). No UHP relics are reported but omphacite and garnet minimum conditions of 600–650 °C, 12 kbar (Equilibria 1 and 3) were deduced from inclusions (Le Fort et al., 1997). It may well be that this is a tectonically reactivated deeper part of the suture zone mélange or even the arc crust. In the same area, HP granulite/amphibolite occurs in the high-grade gneisses of the Nanga Parbat– Haramosh Massif (Pognante et al., 1993).

Fig. 17. Depth–time plot for the Tso Morari eclogites based on the data from de Sigoyer et al. (2000), diffusion modelling of garnet zoning (O’Brien & Sachan, 2000), and zircon fission track data (Schlup et al., 2001).

172

H.-J. Massonne & P.J. O’Brien

Kharta region, east of Mount Everest Granulitised eclogites have recently been discovered within the HHC to the east of Mount Everest as deformed dykes in orthogneisses (Lombardo & Rolfo, 2000). The eclogite relics are discernable despite migmatisation of host gneisses which themselves show older MPrelics. The polyphase metamorphic evolution of pelitic rocks of the area is reflected in early kyanite + staurolite-bearing assemblages replaced by sillimanite-bearing then cordierite + spinel + andalusite-bearing ones (Pognante & Benna, 1993; Borghi et al., 2003). The mafic rocks are yet undated but their host gneisses are comparable with the many Himalayan-age gneisses dated at ca. 30 Ma in the nearby Everest section.

Geodynamic synthesis The metamorphic history of the Himalayan HP/UHP rock associations outlined above is in many ways comparable with that of the group of older HP/UHP rocks observed in the Variscan Bohemian Massif (see earlier). The common features are: (1) the lithologies, basic and sedimentary rocks, involved in the HP/UHP metamorphism, (2) the P–T climax reached by these rocks, (3) the interaction of hydrous fluids with the (near-)UHP rocks at an early stage of exhumation, (4) the juxtaposition against medium-pressure rocks afterwards, and (5) the chronological development in terms of exhumation speed. The microstructural record of the above HP/UHP rocks from both orogens is not in conflict with a similar metamorphic history. However, the character of the protoliths of the eclogites might be different. In spite of the assumptions that the protoliths of the old and cold eclogites from the Bohemian Massif are mainly MORB-derived and those of the eclogites from the NW Himalaya are related to Trap volcanism, a subducted oceanic crust (likely for the Variscan case) but also a thin continental margin with Trap volcanism adherent to a subducted oceanic crust (Indian case) can be involved in a subduction channel regime. The upwards-directed mass flow in such a geotectonic regime allows a fast ascent (several cm/year) of earlier-subducted portions on top of a downgoing lithospheric plate. The metamorphism of subducted material, sediments and basalts, is itself characterised by low geotherms resulting in the observed peak metamorphic conditions between 24 and 29 kbar (or more) around 650 °C (or somewhat higher). After partial exhumation, the insertion or tectonic mingling of previously subducted rocks in/with medium-pressure rocks of a virtually identical temperature regime is possible. It might even be that the continent–continent collision through an initial crustal thickening, by thrusting of one continental plate under another, also started to squeeze out (laterally?) material from a (at that time) deeper portion of a just-terminated subduction channel system. This can be assumed for colliding India and Asia as we know that the contact of the continental plates happened in the time interval 55 to 45 Ma. The coesite-eclogite formation so soon after ocean closure in the Himalaya, as well as the rapid exhumation, suggest a subduction zone setting with some sort of detachment of the subducting mantle lithosphere and buoyant rebound of the subducted upper continental crust as suggested in models by Chemenda et al. (1995, 2000). Subsequent to detachment there must have been a change in subduction angle. In addition, the present-day structure of crust and mantle (e.g. Owens & Zandt, 1997) points to subduction of lithosphere far to the north of the original suture (Fig. 18).

Fig. 18. A cross-section through the Himalayan orogen based on seismic velocity data from Owens & Zandt (1997), redrawn by O’Brien (2001).

The Bohemian Massif and the NW Himalaya 173

174

H.-J. Massonne & P.J. O’Brien

Such a model for the Variscan orogen is, however, in serious conflict with recent orogenic hypotheses suggesting the successive accretion of terranes (microplates) before a final continent–continent collision in Carboniferous times with only subordinate underthrusting (e.g. Winchester & PACE TMR Network Team, 2002). For instance, the assumed wide oceanic basin between Gondwana and Laurentia-Baltica (Tait et al., 2000; or Fig. 2) would not tolerate continent–continent collision already 395 Ma ago. Instead, microplate collisions between Avalonia and parts of the Armorican Terrane Assemblage can be associated with the Devonian HP metamorphism. The plate reconstruction of McKerrow et al. (2000), however, suggests much narrower oceanic realms from Devonian times onwards based on faunal continuity from Laurentia-Baltica and Gondwana. The consequence of this and the recorded palaeomagnetic data (see Tait et al., 2000) would be a landmass (continental crust) lost during the collisional events. For the India–Asia collision the “lost landmass” due to continued convergence of up to 2800 km (in the eastern section: Johnson, 2002) since the initial contact of the continental plates at around 55 to 50 Ma has been taken into consideration. As a consequence of this, thickened continental crust (see Fig. 18) spreads today over such a wide area (Himalaya, Tibet Plateau, Tien Shan) which is unique on Earth. There is no equivalent of the HT/(U)HP rock association observed in the Bohemian Massif presently exposed in the Himalaya apart from the minor HT–HP granulites in the Namche Barwa syntaxis. This is, indeed, not surprising because, if we accept the above suggestion of the beginning of continent–continent collision for the Variscan orogen 395 Ma ago, the development of the Himalayan orogen is not yet as advanced as the Variscan orogen was by about 340 to 335 Ma ago. Nevertheless, other criteria can be considered to prove a similar stage of development between the present Himalaya and the Variscan orogen shortly before 340 Ma. At least we know that crustal material, for instance from the Erzgebirge and the Granulitgebirge, was at that time deeply buried. Probably in the entire Gneiss/Eclogite Unit of the Erzgebirge (see above), the country rocks experienced pressures in the range 15 to 20 kbar. These rocks were located at or near the base of a continental crust thickened to at least 60 km. This is an important analogy with the present stage of the Himalaya. The temperature regimes (650 to 800 °C) at these crustal levels are probably also comparable, as Nelson et al. (1996) suggested the onset of anatexis (thus, possibly 650 °C) in deeper mid-crustal levels of the Himalayan orogen. A further similarity can be also assumed for the lithologies (quartzofeldspathic rocks + metabasic layers?) referring to seismic studies of the Himalaya (Fig. 18 and, e.g., Zhao et al., 1993: mean crustal velocity 6.0–6.4 km s–1). The possible future events faced by the Himalayan orogen can be tentatively deduced from the events in the Variscan orogen related to the time span 340 to 335 Ma. In particular, the diamondiferous quartzofeldspathic rocks from the Erzgebirge and the derived P–T path (Fig. 10) could give important clues to this future development. A possible explanation for fast submersion of continental crust to depths of at least 200 km and fast uplift is the subduction–collision model suggested by Matte (1998). In this model (Fig. 19) submersion of continental material to significant depths (> 100 km) starts, however, immediately after continent–continent collision because it is drawn to depth by the still adherent deeply subducted oceanic crust. After a slab break-off event,

The Bohemian Massif and the NW Himalaya

175

Fig. 19. The subduction–collision model by Matte (1998), redrawn by O’Brien (2000) for the northern Bohemian Massif.

176

H.-J. Massonne & P.J. O’Brien

the oceanic crust sinks deeper into the mantle but the lighter continental crust, driven by buoyancy forces, reaches higher levels again. Another explanation even for extremely fast submersion of continental crust comes from numerical modelling experiments recently reported by Willner et al. (2002). These are based on the previous work by Schott & Schmeling (1998), who had already predicted a delamination of the mantle lithosphere below an orogen after significant thickening of the lithosphere due to continent–continent collision. Consequences of the delamination process are subsequent magmatic activity, HT metamorphism, high heat flow and an extensional regime, all of which can be observed for the Variscan crust after 340 Ma. Referring to model A from the work by Willner et al. (2002), crustal material can also be involved in the delamination process drawing it into depths required for the diamondiferous quartzofeldspathic rocks from the Saxonian Erzgebirge. Further support to crustal delamination is given by Wittenberg et al. (2000) who argue for a considerable deficit of basic material in the present crustal column especially in the Mid-German Moldanubian zone. However, the assumption of these authors that the bulk composition of the Variscan crust was essentially basaltic andesitic may be disputed. Finally, it is worthy of note that lithospheric delamination in the India–Asia collisional belt might have started already (Kosarev et al., 1999). After the possibly cataclysmic events characterised by enormous mass flows driven by lithospheric delamination, the approach towards a crust of normal thickness began. Large volumes of HP rocks were exhumed within a short period of time (335–325 Ma). Limited ultramafic rock volumes such as the garnet pyroxenites from the Granulitgebirge described by Massonne & Bautsch (2002, see also Fig. 9) were also transported possibly by rising melts from the deeper mantle to crustal levels. The movement of Gondwana against Laurentia-Baltica continued although the compressional regime had changed direction. Due to the deformational events after 335 Ma, the Erzgebirge unit containing HP and diamondiferous quartzofeldspathic rocks came in contact with rocks representing only medium crustal levels (Red-and-Grey Gneiss Unit, see Willner et al., 2000). Probably the present collage-like aspect of the Bohemian Massif (see above) was established at this stage. Rocks of the continental crust never again experienced pressures in excess of 15 kbar, although moderate crustal thickening should be taken into account after 335 Ma. This assumption is based on the observation that large volumes of granitic melts have formed later than 335 Ma (e.g. Anthes & Reischmann, 2001). These intruded into crustal levels around 16 km deep (Massonne, 1984), which are exposed at the surface of an almost 30 km thick crust today. The base of this crust was, thus, located in depths of 45 km sometime in the interval 335 to 320 Ma, considering that only erosion was operative for subsequent crustal thinning. In the above discussion, we have not addressed the problem of which plate was thrust under the other during Mid-Devonian to Early Carboniferous times. Comparing a portion of the India–Asia collisional belt, the Variscan equivalent situation would be subduction of Laurentia-Baltica (+ accreted terranes + adherent “lost landmass”) beneath Gondwana. Hints at this view are provided by the many Gondwana-derived portions of the Bohemian Massif which have only experienced low metamorphic pressures (see, e.g., Linnemann et al., 2000) and thus represent the overlying plate in the Bohemian

The Bohemian Massif and the NW Himalaya

177

Massif. The conclusion by Mingram (1998) that the HP rocks from the Erzgebirge represent a subducted part of northern Gondwana does not really contradict the above view. It is also possible that the geochemical signatures of the metasediments investigated by Mingram (1998) point to the provenance area. The detritus of Gondwana could be transported into a basin between Gondwana and Laurentia-Baltica before continent– continent collision (395 Ma) and its sedimentary fill could have been thrust under Gondwana. On the other hand, we cannot rule out that Gondwana was the bottom plate during the process of crustal thickening due to a tremendous lack of petrological data.

Acknowledgements We thank J. von Raumer (Fribourg) and A. Willner (Bochum) for providing figures from their works. B. Lombardo (Torino) kindly reviewed the manuscript.

References Achramowicz, S., Muszynski, A. & Schliestedt, M. (1997): The northeasternmost eclogite occurrence in the Saxothuringian Zone, West Sudetes (Poland). Chem. Erde, 57:51–61. Akaogi, M., Kusaba, K., Susaki, J., Yagi, T., Matsui, M., Kikegawa, T., Yusa, H. & Ito, E. (1992): High pressure high temperature stability of ,PbO2-type TiO2 and MgSiO3 majorite: calorimetric and in situ diffraction studies. In Syono, Y. & Manghnani, M.H. (eds.): High-pressure research: Application to Earth and planetary sciences /Geophys. Monogr., 67/. Washington (D.C.): Am. Geophys. Union, 447–455. Anthes, G. & Reischmann, T. (2001): Timing of granitoid magmatism in the eastern mid-German crystalline rise. J. Geodyn., 31:119–143. Bakun-Czubarow, N. (1989): On the Sudetic eclogites, SW Poland. In 3rd Int. Eclogite Conf. Würzburg, Abstr. Vol., 1. Bakun-Czubarow, N. (1992): Quartz pseudomorphs after coesite and quartz exsolutions in eclogitic omphacites of the Złote Mountains in the Sudetes (SW Poland). Arch. Mineral. (Warszawa), 48:3–25. Bakun-Czubarow, N. (1998): Ilmenite-bearing eclogites of the West Sudetes – their geochemistry and mineral chemistry. Arch. Mineral. (Warszawa), 51:29–99. Beard, B.L., Medaris, L.G., Jr., Johnson, C.M., Brueckner, H.K. & Mísař, Z. (1992): Petrogenesis of Variscan high-temperature Group A eclogites from the Moldanubian Zone of the Bohemian Massif, Czechoslovakia. Contrib. Mineral. Petrol., 111:468–483. Beard, B.L., Medaris, L.G., Jr., Johnson, C.M., Jelinek, E., Tonika, J. & Riciputi, L.R. (1995): Geochronology and geochemistry of eclogites from the Mariánské Lázně Complex, Czech Republic: Implications for Variscan orogenesis. Geol. Rundsch., 84:552–567. Becker, H. (1997): Sm-Nd garnet ages and cooling history of high-temperature garnet peridotite massifs and high-pressure granulites from Lower Austria. Contrib. Mineral. Petrol., 127:224–236. Becker, H. & Altherr, R. (1992): Evidence from ultra-high-pressure marbles for recycling of sediments into the mantle. Nature, 358:745–748. Berthelsen, A. (1953): On the geology of the Rupshu district, NW Himalaya. Medd. Dan. Geol. Foren., 12:350–415. Borghi, A., Castelli, D., Lombardo, B. & Visonà, D. (2003): Thermal and baric evolution of garnet granulites from the Kharta region of S Tibet, E Himalaya. Eur. J. Mineral., 15:401–418. Borkowska, M., Choukroune, P., Hameurt, J. & Martineau, F. (1990): A geochemical investigation of the age, significance and structural evolution of the Caledonian-Variscan granite-gneisses of the Śnieżnik metamorphic area (Central Sudetes, Poland). Geol. Sudet., 15:1–27. Bosbach, D., Stosch, H.-G. & Seidel, E. (1991): Magmatic and metamorphic evolution of metagabbros in the Münchberg Massif, N.E. Bavaria. Contrib. Mineral. Petrol., 107:112–123.

178

H.-J. Massonne & P.J. O’Brien

Bröcker, M., Cosca, M. & Klemd, R. (1997): Geochronologie von Eklogiten und assoziierten Nebengesteinen des Orlica-Snieznik Kristallins (Sudeten, Polen): Ergebnisse von U-Pb, Sm-Nd, Rb-Sr und Ar-Ar Untersuchungen. Terra Nostra, 1997/5:29–30. Bröcker, M., Żelaźniewicz, A. & Enders, M. (1998): Rb-Sr and U-Pb geochronology of migmatitic gneisses from the Góry Sowie (West Sudetes, Poland): the importance of Mid-Late Devonian metamorphism. J. Geol. Soc. London, 155:1025–1036. Brueckner, H.K. & Medaris, L.G., Jr. (1998): A tale of two orogens: the contrasting T-P-t history and geochemical evolution of mantle in high- and ultrahigh-pressure metamorphic terranes of the Norwegian Caledonides and the Czech Variszides. Schweiz. Mineral. Petrogr. Mitt., 78:293–307. Brueckner, H.K., Medaris, L.G., Jr. & Bakun-Czubarow, N. (1991): Nd and Sr age isotope patterns from Variscan eclogites of the eastern Bohemian Massif. Neues Jahrb. Mineral. Abh., 163:169–196. Brueckner, H.K., Blusztajn, J. & Bakun-Czubarow, N. (1996): Trace element and Sm-Nd “age” zoning in garnets from peridotites of the Caledonian and Variscan Mountains and tectonic implications. J. Metamorph. Geol., 14:61–73. Burchfiel, B.C., Chen, Z., Hodges, K.V., Liu, Y., Royden, L., Deng, C. & Xu, J. (1992): The South Tibetan Detachment, Himalayan Orogen: extension contemporaneous with and parallel to shortening in a collisional mountain belt. /Geol. Soc. Am. Spec. Pap., 269/. Washington (D.C.): Geol. Soc. Am., 41 p. Burg, J.-P., Nievergelt, P., Oberli, F., Seward, D., Davy, P., Maurin, J.-C., Diao, Z. & Meier, M. (1998): The Namche Barwa syntaxis: evidence for exhumation related to compressional crustal folding. J. Asian Earth Sci., 16:239–252. Carswell, D.A. (1991): Variscan high P-T metamorphism and uplift history in the Moldanubian Zone of the Bohemian Massif in Lower Austria. Eur. J. Mineral., 3:323–342. Carswell, D.A. & Jamtveit, B. (1990): Variscan Sm-Nd ages for the high-pressure metamorphism in the Moldanubian Zone of the Bohemian Massif, Lower Austria. Neues Jahrb. Mineral. Abh., 162:69–78. Carswell, D.A. & O’Brien, P.J. (1993): Thermobarometry and geotectonic significance of high-pressure granulites: examples from the Moldanubian Zone of the Bohemian Massif in Lower Austria. J. Petrol., 34:427–459. Chamberlain, C.P., Zeitler, P.K. & Erickson, E. (1991): Constraints on the tectonic evolution of the northwestern Himalaya from geochronologic and petrologic studies of Babusar Pass, Pakistan. J. Geol., 99:829–849. Chemenda, A.I., Mattauer, M., Malavieille, J. & Bokum, A.N. (1995): A mechanism for syncollisional rock exhumation and associated normal faulting: Results from physical modelling. Earth Planet. Sci. Lett., 132:225–232. Chemenda, A.I., Burg, J.P. & Mattauer, M. (2000): Evolutionary model of the Himalaya-Tibet system: geopoem based on new modelling, geological and geophysical data. Earth Planet. Sci. Lett., 174:397–409. Coward, M.P., Windley, B.F., Broughton, R.D., Luff, I.W., Petterson, M.G., Pudsey, C.J., Rex, D.C. & Khan, M.A. (1986): Collision tectonics in the NW Himalayas. In Coward, M.P. & Ries, A. (eds.): Collision tectonics /Geol. Soc. London, Spec. Publ., 19/, 205–221. Coward, M.P., Butler, R.W.H., Chambers, A.F., Graham, R.H., Izatt, C.N., Khan, M.A., Knipe, R.J., Prior, D.J., Treloar, P.J. & Williams, M.P. (1988): Folding and imbrication of the Indian crust during Himalayan collision. Phil. Trans. R.. Soc. Lond., Ser. A, 326:89–116. Dercourt, J., Ricou, L.E. & Vrielynck, B. (1993): Atlas of Tethys palaeoenvironmental maps. Paris: GauthierVillars, 307 p. de Sigoyer, J., Guillot, S., Lardeaux, J.-M. & Mascle, G. (1997): Glaucophane-bearing eclogites in the Tso Morari dome (eastern Ladakh, NW Himalaya). Eur. J. Mineral., 9:1073–1083. de Sigoyer, J., Chavagnac, V., Blichert-Toft, J., Villa, I.M., Luais, B., Guillot, S., Cosca, M. & Mascle, G. (2000): Dating the Indian continental subduction and collisional thickening in the northwest Himalaya: Multichronology of Tso Morari eclogites. Geology, 28: 487–490. Dobrzhinetskaya, L., Green, H.W. II, Massonne, H.-J. & Stöckhert, B. (2001): Origin of microdiamonds from ultra-high-pressure terranes. In 11th Ann. V.M. Goldschmidt Conf., Hot Springs, Virginia, USA 2001, Abstr. No. 3305 (on CD). Dubińska, E., Żelaźniewicz, A., Nejbert, K. & Bylina, P. (1999): Ultramafic rocks from migmatitic gneisses of the Góry Sowie block, Sudetes. Min. Soc. Pol. Spec. Pap., 14:76–78.

The Bohemian Massif and the NW Himalaya

179

Dudek, A. & Fediuková, E. (1974): Eclogites of the Bohemian Moldanubicum. Neues Jahrb. Mineral. Abh., 121:127–159. Düffels, K. & Massonne, H.-J. (2001): Geochemical signatures of diamondiferous gneisses and their adjacent rocks from the Erzgebirge, Germany. Terra Nostra, 2001/5:24–26. Fontan, D. (1998): Regional geological mapping, metamorphic P-T-t path evaluations and ore potential assessment in Neelum valley (Azad Kashmir, NE Pakistan). PhD Thesis, Université de Louvain, Belgium, 182 p. Fontan, D., Schuppe, M., Hunziker, C.J., Martinotti, G. & Verkaeren, J. (2000): Metamorphic evolution, 40 Ar–39Ar chronology and tectonic model for the Neelum Valley, Azad Kashmir, NE Pakistan. In Khan, M.A., Treloar, P.J., Searle, M.P. & Jan, M.Q. (eds.): Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya /Geol. Soc. London, Spec. Publ., 170/, 431–454. Franke, W. (2000): The mid-European segment of the Variscides: tectonostratigraphic units, terrane boundaries and plate tectonic evolution. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 35–61. Franke, W. & Engel, W. (1986): Synorogenic sedimentation in the Variscan belt of Europe. Bull. Soc. Geol. Fr., 1986/II:25–33. Franke, W. & Stein, E. (2000): Exhumation of high-grade rocks in the Saxo-Thuringian Belt: geological constraints and geodynamic concepts. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 337–354. Franz, G., Thomas, S. & Smith, D.C. (1986): High-pressure phengite decomposition in the Weissenstein eclogite, Münchberger Gneiss Massif, Germany. Contrib. Mineral. Petrol., 92:71–85. Gansser, A. (1979): Ophiolitic belts of the Himalayan and Tibetan region. In Working Group of Ophiolite Project (ed.): International atlas of ophiolites /Geol. Soc. Am. Map Chart Ser., MC–33/. Washington (D.C.): Geol. Soc. Am. Gayk, T., Kleinschrodt, R. & Seidel, E. (1995a): Subduction-related metasomatism in the mantle wedge: an example from the Münchberg Massif. Bochumer Geol. Geotech. Arb., 44:46–48. Gayk, T., Kleinschrodt, R., Langosch, A. & Seidel, E. (1995b): Quartz exsolution in clinopyroxene of highpressure granulite from the Münchberg Massif. Eur. J. Mineral., 7:1217–1220. Ghazanfar, M. & Chaudhry, M.N. (1986): Reporting MCT in Northwest Himalaya, Pakistan. Geol. Bull. Univ. Punjab, 21:10–18. Ghazanfar, M. & Chaudhry, M.N. (1987): Geology, structure and geomorphology of Upper Kaghan Valley, northwest Himalaya, Pakistan. Geol. Bull. Univ. Punjab, 22:13–56. Gebauer, D. & Grünenfelder, M. (1979): U-Pb zircon and Rb-Sr mineral dating of eclogites and their country rocks. Example: Münchberg Gneiss Massif, northeast Bavaria. Earth Planet. Sci. Lett., 42:35–44. Girard, M. & Bussy, F. (1999): Late Pan-African magmatism in the Himalaya: new geochronological and geochemical data from the Ordovician Tso Morari metagranites (Ladakh, NW India). Schweiz. Mineral. Petrogr. Mitt., 79:399–418. Greco, A. & Spencer, D.A. (1993): A section through the Indian Plate, Kaghan Valley, NW Himalaya, Pakistan: Himalayan tectonics, the Alps. In Treloar, P.J. & Searle, M.P. (eds.): Himalayan tectonics /Geol. Soc. London, Spec. Publ., 74/, 221–236. Greco, A., Martinotti, G., Papritz, K., Ramsey, J.G. & Rey, R. (1989): The crystalline rocks of the Kaghan Valley (NE Pakistan). Eclogae Geol. Helv., 82:629–653. Guillot, S., Lardeaux, J.M., Mascle, G. & Colchen, M. (1995): Un nouveau témoin du métamorphisme de haute-pression dans la chaîne himalayenne (Dôme du Tso Morari, Est Ladakh). C. R. Scéances Acad. Sci. Paris, 320:931–936. Guillot, S., de Sigoyer, J., Lardeaux, J.M. & Mascle, G. (1997): Eclogitic metasediments from the Tso Morari area (Ladakh, Himalaya): Evidence for continental subduction during India-Asia convergence. Contrib. Mineral. Petrol., 128:197–212. Guillot, S., Cosca, M., Allemand, P. & Le Fort, P. (1999): Contrasting metamorphic and geochronologic evolution along the Himalaya belt. Geol. Soc. Am. Spec. Pap., 328:117–128. Hammerschmidt, K. & Franz, G. (1992): Retrograde evolution of eclogites: evidences from microstructures and 40Ar/39Ar white mica dates, Münchberg Massif, northern Bavaria. Contrib. Mineral. Petrol., 111:113–125.

180

H.-J. Massonne & P.J. O’Brien

Harris, N. & Massey, J. (1994): Decompression and anatexis of Himalayan metapelites. Tectonics, 13:1537–1546. Hartz, E.H. & Torsvik, T.H. (2002): Baltica upside down: A new plate tectonic model for Rodinia and the Iapetus Ocean. Geology, 30:255–258. Hazard, J. (1886): Geologische Specialkarte des Königreichs Sachsen No. 116. Section Pockau-Lengefeld. Königliches Finanz-Ministerium (ed.) under the supervision of Credner, H. Hegner, E. & Kröner, A. (2000): Review of Nd isotopic data and xenocrystic and detrital zircon ages from the pre-Variscan basement in the eastern Bohemian Massif: speculations on palinspastic reconstructions. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 113–129. Hirn, A., Nercessian, A, Sapin, M., Jobert, G., Xin, X.Z., Yuan, G.E. & Wen, T.J. (1984): Lhasa block and bordering sutures: a continuation of the 500-km Moho traverse through Tibet. Nature, 307:25–27. Hodges, K.V. (2000): Tectonics of the Himalaya and southern Tibet from two perspectives. Geol. Soc. Am. Bull., 112:324–350. Honegger, K., Dietrich, V., Frank, W., Gansser, A., Thöni, M. & Trommsdorf, V. (1982): Magmatism and metamorphism in the Ladakh Himalaya (the Indus-Tsangpo suture zone). Earth Planet. Sci. Lett., 60:253–292. Honegger, K., Le Fort, P., Mascle, G. & Zimmermann, J.-L. (1989): The blueschists along the Indus Suture Zone in Ladakh, NW Himalaya. J. Metamorph. Geol., 7:57–72. Hwang, S.-L., Shen, P., Chu, H.-T. & Yui, T.-F. (2000): Nanometer-size ,-PbO2-type TiO2 in garnet: a thermobarometer for ultrahigh-pressure metamorphism. Science, 288:321–324. Hwang, S.-L., Shen, P., Chu, H.-T., Yui, T.-F. & Lin, C.-C. (2001): Genesis of microdiamonds from melt and associated multiphase inclusions in garnet of ultrahigh-pressure gneiss from Erzgebirge, Germany. Earth Planet. Sci. Lett., 188:9–15. Jekosch, U. & Bautsch, H.-J. (1991): Orientierte Entmischungen in Pyroxenen. Z. Kristallogr., Suppl., 4:134. Jelínek, E., Pacesová, M., Mísař, Z., Martinec, P. & Weiss, Z. (1984): Geochemistry of a dismembered metaophiolite complex, Letovice, Czechoslovakia. Trans. R. Soc. Edinburgh: Earth Sci., 75:37–48. Johnson, M.R.W. (2002): Shortening budgets and the role of continental subduction during the India-Asia collision. Earth Sci. Rev., 59:101–123. Kaneko, Y., Yamamoto, H., Katayama, I., Misawa, K., Ishikawa, M., Rehman, H.U., Kausar, A.B. & Shiraishi, K. (2001): Coesite inclusions and prograde compositional zonation of zircons in Himalayan gneisses, NW Himalaya, Pakistan: evidence from SHRIMP-dating of coesite-bearing zircon. In Ogasawara, Y., Maruyama, S. & Liou, J.G. (eds.): Proc. UHPM Workshop 2001 at Waseda Univ. Tokyo: Waseda Univ., 121–123. Klápová, H., Konopásek, J. & Schulmann, K. (1998): Eclogites from the Czech part of the Erzgebirge: multistage metamorphic and structural evolution. J. Geol. Soc. London, 155:567–583. Klemd, R. (1989): P-T evolution and fluid inclusion characteristics of retrograded eclogites, Münchberg Gneiss Complex, Germany. Contrib. Mineral. Petrol., 102:221–229. Klemd, R., Matthes, S. & Okrusch, M. (1991): High-pressure relics in meta-sediments intercalated with the Weissenstein eclogite, Münchberg gneiss complex, Bavaria. Contrib. Mineral. Petrol., 107:328–342. Klemd, R., Matthes, S. & Schüssler, U. (1994): Reaction textures and fluid behaviour in very high-pressure calc-silicate rocks of the Münchberg Gneiss Complex, Bavaria, Germany. J. Metamorph. Geol., 12:735–745. Klootwijk, C.T., Gee, J.S., Pierce, J.W., Smith, G.M. & McFadden, P.L. (1992): An early India-Asia contact: palaeomagnetic constraints from Ninetyeast Ridge, ODP Leg 121. Geology, 20:395–398. Konopásek, J. (1998): Formation and destabilization of the high pressure assemblage garnet-phengiteparagonite (Krušné hory Mountains, Bohemian Massif): The significance of the Tschermak substitution in the metamorphism of pelitic rocks. Lithos, 42:269–284. Konopásek, J. (2001): Eclogitic micaschists in the central part of the Krušné hory Mountains (Bohemian Massif). Eur. J. Mineral., 13:87–100. Konopásek, J., Schulmann, K. & Lexa, O. (2001): Structural evolution of the central part of the Krušné hory (Erzgebirge) Mountains in the Czech Republic – evidence for changing stress regime during Variscan compression. J. Struct. Geol., 23:1373–1392.

The Bohemian Massif and the NW Himalaya

181

Konopásek, J., Schulmann, K. & Johan, V. (2002): Eclogite-facies metamorphism in the eastern margin of the Bohemian Massif – subduction prior to continental underthrusting? Eur. J. Mineral., 14:701–713. Kosarev, G., Kind, R., Sobolev, S.V., Yuan, X., Hanka, W. & Oreshin, S. (1999): Seismic evidence for a detached Indian lithospheric mantle beneath Tibet. Science, 283:1306–1309. Kotková, J., Kröner, A., Todt, W. & Fiala, J. (1995): Zircon dating of north Bohemian granulites, Czech Republic: Further evidence for the Lower Carboniferous high-pressure event in the Bohemian Massif. Geol. Rundsch., 85:154–161. Kotková, J., Harley, S.L. & Fisera, M. (1997): A vestige of very high-pressure (ca. 28 kbar) metamorphism in the Variscan Bohemian Massif, Czech Republic. Eur. J. Mineral., 9:1017–1033. Kreuzer, H., Seidel, E., Schüssler, U., Okrusch, M., Lenz, K.-L. & Raschka, H. (1989): K-Ar geochronology of different tectonic units at the northwestern margin of the Bohemian Massif. Tectonophysics, 157:149–178. Krohe, A. (1996): Variscan tectonics of central Europe: Postaccretionary intraplate deformation of weak continental lithosphere. Tectonics, 15:1364–1388. Kröner, A. & Willner, A.P. (1998): Time of formation and peak of Variscan HP-HT metamorphism of quartzfeldspar rocks in the central Erzgebirge, Saxony, Germany. Contrib. Mineral. Petrol., 132:1–20. Kröner, A., Jaeckel, P., Reischmann, T. & Kroner, U. (1998): Further evidence for an early Carboniferous (~ 340 Ma) age of high-grade metamorphism in the Saxonian granulite complex. Geol. Rundsch., 86:751–766. Kröner, A., O’Brien, P.J., Nemchin, A.A. & Pidgeon, R.T. (2000): Zircon ages for high pressure granulites from South Bohemia, Czech Republic, and their connection to Carboniferous high temperature processes. Contrib. Mineral. Petrol. 138:127–142. Kryza, R., Pin, C. & Vielzeuf, D. (1996): High-pressure granulites from the Sudetes (south-west Poland): evidence of crustal subduction and collisional thickening in the Variscan Belt. J. Metamorph. Geol., 14:531–546. Le Fort, P. (1975): Himalayas: the collided range, present knowledge of the continental arc. Am. J. Sci., 275A:1–44. Le Fort, P., Guillot, S. & Pecher, A. (1997): HP metamorphic belt along the Indus suture zone of NW Himalaya: New discoveries and significance. C. R. Scéances Acad. Sci. Paris, 325:773–778. Linnemann, U., Gehmlich, M., Tichomirowa, M., Buschmann, B., Nasdala, L., Jonas, P., Lützner, H. & Bombach, K. (2000): From Cadomian subduction to Early Palaeozoic rifting: the evolution of Saxo-Thuringia at the margin of Gondwana in the light of single zircon geochronology and basin development (Central European Variscides, Germany). In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt, /Geol. Soc. London, Spec. Publ., 179/, 131–153. Lombardo, B. & Rolfo, F. (2000): Two contrasting eclogite types in the Himalayas: implications for the Himalayan orogeny. J. Geodyn., 30:37–60. Lombardo, B., Pertusati, P., Rolfo, F. & Visonà, D. (1998): First reports of eclogites from the Eastern Himalaya: Implications for the Himalayan orogeny. Mem. Sci. Geol. Univ. Padova, 50:67–68. Lombardo, B., Rolfo, F. & Compagnoni, R. (2000): Glaucophane and barroisite eclogites from the Upper Kaghan Nappe: Implications for the metamorphic history of the NW Himalaya. In Khan M.A., Treloar, P.J., Searle, M.P. & Jan, M.Q. (eds.): Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya /Geol. Soc. Lond., Spec. Publ., 170/, 411–430. Massonne, H.-J. (1984): Bestimmung von Intrusionstiefen variszischer Granite Mitteleuropas und Neuschottlands anhand der Chemie ihrer Hellglimmer. Fortschr. Mineral., 62, Beih. 1:147–149. Massonne, H.-J. (1992): Thermochemical determination of water activities relevant to eclogitic rocks. In Kharaka, Y.K. & Maest, A.S. (eds.): Water-Rock Interaction, Proc. 7th Int. Symp., Park City, Utah. Rotterdam: Balkema, 1523–1526. Massonne H.-J. (1993): High-pressure melting of eclogites from the Mid-European Variscides. Terra Abstr., 4th Int. Eclogite Conf., 5:16. Massonne H.-J. (1994): P-T evolution of eclogite lenses in the crystalline complex of the Erzgebirge, Middle Europe: An example for high-pressure to ultrahigh-pressure metabasites incorporated into continental crust. In 1st Workshop UHP metamorphism and tectonics, Abstr. Vol., 29–32. Massonne, H.-J. (1995): Experimental and petrogenetic study of UHPM. In Coleman, R. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 33–95.

182

H.-J. Massonne & P.J. O’Brien

Massonne, H.-J. (1999): A new occurrence of microdiamonds in quartzofeldspathic rocks of the Saxonian Erzgebirge, Germany, and their metamorphic evolution. In Gurney, J.J., Gurney, J.L., Pascoe, M.D. & Richardson, S.H. (eds.): Proc. 7th Int. Kimberlite Conf., Cape Town, The Nixon Vol. Cape Town: Redroof Publ., 533–539. Massonne, H.-J. (2000): Ultrahigh-pressure metamorphism in the Bohemian Massif. In 31st Int. Geol. Congr. Rio de Janeiro, Brazil, Ext. Abstr., 4 p. (on CD) Massonne, H.-J. (2001a): First find of coesite in the ultrahigh-pressure metamorphic region of the Central Erzgebirge, Germany. Eur. J. Mineral., 13:565–570. Massonne, H.-J. (2001b): Origin of microdiamond-bearing quartzofeldspathic rocks (saidenbachites) from the Erzgebirge, Germany: A progress report. In UHPM workshop 2001, Fluid/slab/mantle interactions and ultrahigh-P minerals, Waseda Univ., Tokyo, Abstr. Vol., 11–15. Massonne, H.-J. & Bautsch, H.-J. (2002): An unusual garnet pyroxenite from the Granulitgebirge, Germany: origin in the transition zone (>400 km Earth’s depths) or in a shallower upper mantle region? Int. Geol. Rev., 44:779–796. Massonne, H.-J. & Burchard, M. (2000): Exotic minerals in eclogites from the Central Erzgebirge – evidence for fluid-rock interaction at UH metamorphic pressures. Ber. Dtsch. Mineral. Ges., Beih. Eur. J. Mineral., 12/1:122 Massonne, H.-J. & Grosch, U. (1995): P-T evolution of Palaeozoic garnet peridotites from the Saxonian Erzgebirge and the Aheim region, Norway. In 6th Int. Kimberlite Conf., Novosibirsk, Abstr. Vol., 353–355. Massonne, H.-J. & Nasdala, L. (2003): Characterization of an early metamorphic stage through inclusions in zircon of a diamondiferous quartzofeldspathic rock from the Erzgebirge, Germany. Am. Mineral., 88:883–889. Massonne, H.-J., Grosch, U. & Willner, A.P. (1993): Geothermobarometrie mittels Ti-Gehalten in Kalihellglimmern. Ber. Dtsch. Mineral. Ges., Beih. Eur. J. Mineral., 5/1:85. Massonne, H.-J., Dobrzhinetskaya, L. & Green, H.W. II. (2000): Quartz – K-feldspar intergrowths enclosed in eclogitic garnet and omphacite. Are they pseudomorphs after coesite? Ext. Abstract, 31st Int. Geol. Congr. Rio de Janeiro, Brazil: 4 p. (on CD, search for Massone). Massonne, H.-J., Nasdala, L. & Kennedy, A. (2001): U-Th-Pb dating of zircons and monazites from diamondiferous gneisses of the Saxonian Erzgebirge – implications for their UHP/HP evolution. In 6th Int. Eclogite Conf., Niihama, Japan, Abstr. Vol., 88. Mattauer, M., Matte, P. & Olivet, J.L. (1999): A 3D model of the India-Asia collision at plate scale. C. R. Scéances Acad. Sci. Paris, Sci. Terre, 328:499–508. Matte, P. (1986): Tectonics and plate tectonics model for the Variscan belt of Europe. Tectonophysics, 126:329–374. Matte, P. (1998): Continental subduction and exhumation of HP rocks in Paleozoic orogenic belts: Uralides and Variscides. Geol. För. Stockh. Förh., 120:209–222. Matte, P. & Burg, J.P. (1981): Sutures, thrusts and nappes in the Variscan Arc of western Europe: plate tectonic implications. In McClay, K.R. & Price, N.J. (eds.): Thrust and nappe tectonics /Geol. Soc. London, Spec. Publ., 9/, 353–358. Matte, P. Maluski, H., Rajlich, P. & Franke, W. (1990): Terrane boundaries in the Bohemian Massif: results of large-scale Variscan shearing. Tectonophysics, 177:151–170. Matthes, S., Richter, P. & Schmidt, K. (1969): Die Eklogitvorkommen des kristallinen Grundgebirges in NEBayern II. Der Klinopyroxen der Eklogite und Eklogitamphibolite des Münchberger Gneisgebietes. Neues Jahrb. Mineral. Abh., 112:1–46. McKerrow, W.S., Mac Niocaill, C., Ahlberg, P.E., Clayton, G., Cleal, C.J. & Eagar, R.M.C. (2000): The Late Palaeozoic relations between Gondwana and Laurussia. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 9–20. Medaris, L.G., Jr., Wang, H.F., Mísař, Z. & Jelínek, E. (1990): Thermobarometry, diffusion modelling and cooling rates of crustal garnet peridotites: two examples from the Moldanubian Zone of the Bohemian Massif. Lithos, 25:189–202.

The Bohemian Massif and the NW Himalaya

183

Medaris, L.G., Jr., Jelínek, E. & Mísař, Z. (1995a): Czech eclogites: Terrane settings and implications for Variscan tectonic evolution of the Bohemian Massif. Eur. J. Mineral., 7:7–28. Medaris, L.G., Jr., Beard, B.L., Johnson, C.M., Valley, J.W., Spicuzza, M.J., Jelínek, E. & Mísař, Z. (1995b): Garnet pyroxenite and eclogite in the Bohemian massif: Geochemical evidence for Variscan recycling of subducted lithosphere. Geol. Rundsch., 84:489–505. Medaris, L.G., Jr., Fournelle, J.H., Ghent, E.D., Jelínek, E. & Mísař, Z. (1998): Prograde eclogite in the Gföhl Nappe, Czech Republic: new evidence on Variscan high-pressure metamorphism. J. Metamorph. Geol., 16:563–576. Messiga, B. & Bettini, E. (1990): Reactions behaviour during kelyphite and symplectite formation: a case study of mafic granulites and eclogites from the Bohemian Massif. Eur. J. Mineral., 2:125–144. Mingram, B. (1998): The Erzgebirge, Germany – A subducted part of northern Gondwana: geochemical evidence for repetition of Early Paleozoic metasedimentary sequences in metamorphic thrust units. Geol. Mag., 135:785–801. Mísař, Z., Jelínek, E., Souček, J. & Tonica, J. (1984): The correlation of gabbro-peridotite complexes of the Bohemian Massif. Krystalinikum, 17:99–113. Murphy, J.B., Keppie, J.D., Dostal, J. & Nance, R.D. (1999): Neoproterozoic – early Paleozoic evolution of Avalonia. In Ramos, V.A. & Keppie, J.D. (eds.): Laurentia-Gondwana connections before Pangea /Geol. Soc. Am. Spec. Pap., 336/. Washington (D.C.): Geol. Soc. Am., 253–266. Nasdala, L. & Massonne, H.-J. (2000): Microdiamonds from the Saxonian Erzgebirge, Germany: in situ microRaman characterisation. Eur. J. Mineral., 12:495–498. Nelson, K.D., Zhao, W., Brown, L.D., Kuo, J., Che, J., Liu, X., Klemperer, S.L., Makovsky, Y., Meissner, R., Mechie, J., Kind, R., Wenzel, F., Ni, J., Nabelek, J., Chen, L., Tan, H., Wei, W., Jones, A.G., Booker, J., Unsworth, M., Kidd, W.S.F., Hauck, M., Alsdorf, D., Ross, A., Cogan, M., Wu, C., Sandvol, E. & Edwards, M. (1996): Partially molten middle crust beneath southern Tibet: synthesis of Project INDEPTH results. Science, 274:1684–1688. Ni, J. & Barazangi, M. (1984): Seismotectonics of the Himalayan collision zone: geometry of the underthrusting Indian plate beneath the Himalaya. J. Geophys. Res., 89:1147–1163. O’Brien, P.J. (1989): The petrology of retrograded eclogites of the Oberpfalz Forest, northeastern Bavaria, West Germany. Tectonophysics, 157:195–212. O’Brien, P.J. (1991): High pressure metamorphism in the NW Bohemian Massif: Comparisons and contrasts between the Moldanubian Zone, Münchberg Massif, ZEV, ZTT and Erzgebirge. KTB-Rep., 91–1:1–12. O’Brien, P.J. (1993): Partially retrograded eclogites of the Münchberg Massif, Germany: records of a multistage Variscan uplift history in the Bohemian Massif. J. Metamorph. Geol., 11:241–260. O’Brien, P.J. (1997): Garnet zoning and reaction textures in overprinted eclogites, Bohemian Massif, European Variscides: a record of their thermal history during exhumation. Lithos, 41:119–133. O’Brien, P.J. (2000): The fundamental Variscan problem: high-temperature metamorphism at different depths and high-pressure metamorphism at different temperatures. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 369–386. O’Brien, P.J. (2001): Subduction followed by Collision: Alpine and Himalayan examples. In Rubie, D.C. & van der Hilst, R. (eds.): Processes and consequences of deep subduction /Phys. Earth Planet. Inter., 127/, 277–291. O’Brien, P.J. & Carswell, D.A. (1993): Tectonometamorphic evolution of the Bohemian Massif: Evidence from high pressure metamorphic rocks. Geol. Rundsch., 82:531–555. O’Brien, P.J. & Sachan, H.K. (2000): Diffusion modelling in garnet from Tso Morari eclogite and implications for exhumation models. Earth Sci. Front. (China Univ. Geosci., Beijing), 7:25–27. O’Brien P.J. & Vrána, S. (1995): Eclogites with a short-lived granulite facies overprint in the Moldanubian Zone, Czech Republic: Petrology, geochemistry and diffusion modelling of garnet zoning. Geol. Rundsch., 84:473–488. O’Brien, P.J., Röhr, C., Okrusch, M. & Patzak, M. (1992): Eclogite facies relics and a multistage breakdown in metabasites of the KTB pilot hole, NE Bavaria: implications for the Variscan tectonometamorphic evolution of the NW Bohemian Massif. Contrib. Mineral. Petrol., 112:261–278.

184

H.-J. Massonne & P.J. O’Brien

O’Brien, P.J., Duyster, J., Grauert, B., Schreyer, W., Stöckhert, B. & Weber, K. (1997a): Crustal evolution of the KTB drill site: from oldest relics to the late Hercynian granites. J. Geophys. Res., 102 (B8):18,203–18,220. O’Brien P.J., Kröner, A., Jaekel, P., Hegner, E., Żelaźniewicz, A. & Kryza, R. (1997b): Petrological and isotopic studies on Palaeozoic high pressure granulites, Góry Sowie (Owl Mts.), Polish Sudetes. J. Petrol., 38:433–456. O’Brien, P.J., Zotov, N., Law, R., Khan, M.A. & Jan, M.Q. (1999): Coesite in eclogite from the Upper Kaghan Valley, Pakistan: A first record and implications. Terra Nostra, 1999/2:109–111. O’Brien, P.J., Zotov, N., Law, R., Khan, M.A. & Jan, M.Q. (2001): Coesite in Himalayan eclogite and implications for models of India-Asia collision. Geology, 29:435–438. Okrusch, M., Matthes, S., Klemd, R., O’Brien, P.J. & Schmidt, K. (1991): Eclogites at the north-western margin of the Bohemian Massif: a review. Eur. J. Mineral., 3:707–730. Oliver, G.J.H., Corfu, F. & Krogh, T.E. (1993): U-Pb ages from SW Poland: evidence for a Caledonian suture zone between Baltica and Gondwana. J. Geol. Soc. London, 150:355–369. Owens, T.J. & Zandt, G. (1997): Implications of crustal property variations for models of Tibetan plateau evolution. Nature, 387:37–43. Patriat, P. & Achache, J. (1984): India-Asia collision chronology and implications for crustal shortening and driving mechanism of plates. Nature, 311:615–621. Petrakakis, K. (1997): Evolution of Moldanubian rocks in Austria: review and synthesis. J. Metamorph. Geol., 15:203–222. Petrakakis, K. & Jawecki, C. (1995): High-grade metamorphism and retrogression of Moldanubian granulites, Austria. Eur. J. Mineral., 7:1183–1203. Petterson, M.G. & Windley, B.F. (1985): Rb-Sr dating of the Kohistan batholith in the Trans-Himalaya of N. Pakistan and tectonic implications. Earth Planet. Sci. Lett., 74:45–57. Pin, C., Majerowicz, A. & Wojciechowska, I. (1988): Upper Paleozoic oceanic crust in the Polish Sudetes: NdSr isotope and trace element evidence. Lithos, 21:195–209. Pognante, U. & Benna, P. (1993): Metamorphic zonation, migmatization, and leucogranites along the Everest transect (Eastern Nepal and Tibet): record of an exhumation history. In Treloar, P.J. & Searle, M.P. (eds.): Himalayan tectonics /Geol. Soc. London, Spec. Publ., 74/, 323–340. Pognante, U. & Spencer, D.A. (1991): First record of eclogites from the High Himalayan belt, Kaghan valley (northern Pakistan). Eur. J. Mineral., 3:613–618. Pognante, U., Benna, P. & Le Fort, P. (1993): High-pressure metamorphism in the High Himalayan Crystallines of the Stak valley, northeastern Nanga Parbat-Haramosh syntaxis, Pakistan Himalaya. In Treloar, P.J. & Searle, M.P. (eds.): Himalayan tectonics /Geol. Soc. London, Spec. Publ., 74/, 161–172. Rajlich, P. (1990): Variscan shearing tectonics in the Bohemian massif. Mineral. Slov., 22:33–40. Reiche, M. & Bautsch, H.-J. (1984): Entmischungsstrukturen in Pyroxenen aus eklogitischen Gesteinen. Freiberger Forschungsh., C393:19–33. Reinhardt, J. (1990): Metamorphism of the metapelites in the KTB drill core: reaction sequence and thermobarometric constraints. KTB Rep. 90–4:535. Reinhardt, J. & Kleemann, U. (1994): Extensional unroofing of granulitic lower crust and related low-pressure, high-temperature metamorphism in the Saxonian Granulite Massif, Germany. Tectonophysics, 238:71–94. Robardet, M., Paris, F. & Racheboeuf, P.R. (1990): Palaeogeographic evolution of southwestern Europe during Early Palaeozoic times. In McKerrow, W.S. & Scotese, C.R. (eds.): Paleozoic paleogeography and biogeography /Geol. Soc. London Mem., 12/, 411–419. Romer, R.L. & Rötzler, J. (2001): P-T-t evolution of ultrahigh-temperature granulites from the Saxon Granulite Massif, Germany. Part II: Geochronology. J. Petrol., 42:2015–2032. Rötzler, J. & Romer, R.L. (2001): P-T-t evolution of ultrahigh-temperature granulites from the Saxon Granulite Massif, Germany. Part I: Petrology. J. Petrol., 42:1995–2013. Rötzler, K., Schumacher, R., Maresch, W.V. & Willner, A.P. (1998): Characterization and geodynamic implications of contrasting metamorphic evolution in juxtaposed high-pressure units of the Western Erzgebirge (Saxony, Germany). Eur. J. Mineral., 10:261–280. Rowley, D.B. (1996): Age of initiation of collision between India and Asia: a review of stratigraphic data. Earth Planet. Sci. Lett., 145:1–13.

The Bohemian Massif and the NW Himalaya

185

Sachan, H.K., Bodnar, R.J., Islam, R., Szabó, Cs. & Law, R. (1999): Exhumation history of eclogites from the Tso Morari Crystalline Complex in eastern Ladakh: mineralogy and fluid inclusion constraints. J. Geol. Soc. India, 53:181–190. Sachan, H.K., Mukherjee, B.K., Ogasawara, Y., Maruyama, S., Pandey, A.K., Muko, A., Yoshioka, N. & Ishida, H. (2001): Discovery of coesite from Indian Himalaya: consequences on Himalayan tectonics. In UHPM workshop 2001, Fluid/slab/mantle interactions and ultrahigh-P minerals, Waseda Univ., Tokyo, Abstr. Vol., 124–128. Scharbert, H.G. & Carswell, D.A. (1983): Petrology of garnet-clinopyroxene rocks in a granulite facies environment, Bohemian Massif of Lower Austria. Bull. Minéral., 106:761–774. Schlup, M., Carter, A. & Steck, A. (2001): Exhumation history revealed by fission track cooling ages in the North Himalayan Crystalline zone of eastern Ladakh, NW Himalaya. J. Asian Earth Sci., 19:57–58. Schmädicke, E. (1991): Quartz pseudomorphs after coesite in eclogites from the Saxonian Erzgebirge. Eur. J. Mineral., 3:231–238. Schmädicke, E. & Evans, B.W. (1997): Garnet-bearing ultramafic rocks from the Erzgebirge, and their relation to other settings in the Bohemian Massif. Contrib. Mineral. Petrol., 127:57–74. Schmädicke, E. & Müller, W.F. (2000): Unusual exsolution phenomena in omphacite and partial replacement of phengite by phlogopite + kyanite in an eclogite from the Erzgebirge. Contrib. Mineral. Petrol., 139:629–642. Schmädicke, E., Okrusch, M. & Schmidt, W. (1992): Eclogite-facies rocks in the Saxonian Erzgebirge, Germany: high pressure metamorphism under contrasting P-T-conditions. Contrib. Mineral. Petrol., 110:226–241. Schmädicke, E., Cosca, M.A. & Okrusch, M. (1995): Variscan Sm-Nd and Ar-Ar ages of eclogite facies rocks from the Erzgebirge, Bohemian Massif. J. Metamorph. Geol., 13:537–552. Schmidt, W., Schmädicke, E. & Werner, C.-D. (1990): Eklogite des Erzgebirges. Ber. Dtsch. Mineral. Ges., Beih. Eur. J. Mineral., 2/2:125–151. Schneider, D.A., Edwards, M.A., Kidd, W.S.F., Khan, M.A., Seeber, L. & Zeitler, P.K. (1999): Tectonics of Nanga Parbat, western Himalaya: synkinematic plutonism within the doubly vergent shear zones of a crustal-scale pop-up structure. Geology, 27:999–1002. Schott, B. & Schmeling, H. (1998): Delamination and detachment of a lithospheric root. Tectonophysics, 296:225–247. Schulmann, K., Ledru, P., Autran, A., Melka, R., Lardeaux, J.M., Urban, M. & Lobkowicz, M. (1991): Evolution of nappes in the eastern margin of the Bohemian Massif: a kinematic interpretation. Geol. Rundsch., 80:73–92. Searle, M.P., Windley, B.F., Coward, M.P., Cooper, D.J.W., Rex, A.J., Rex, D., Li, T.D., Xiao, X.C., Jan, M.Q., Thakur, V.C. & Kumar, S. (1987): The closing of Tethys and the tectonics of the Himalaya. Geol. Soc. Am. Bull., 98:678–701. Searle, M.P., Corfield, R.I., Stephenson, B. & McCarron, J. (1997): Structure of the North Indian continental margin in the Ladakh-Zanskar Himalayas: implications for the timing of obduction of the Spontang ophiolite, India-Asia collision and deformation events in the Himalaya. Geol. Mag., 134:297–316. Șengör, A.M.C. & Natal’in, B.A. (1996): Paleotectonics of Asia: fragments of a synthesis. In Yin, A. & Harrison, M. (eds.): Tectonic evolution of Asia. Cambridge: Cambridge Univ. Press, 486–640. Shams, F.A. (1972): Glaucophane-bearing rocks from near Topsin, Swat: First record from Pakistan. Pak. J. Sci. Res., 24:343–345. Smulikowski, K. & Smulikowski, W. (1985): On the porphyroblastic eclogites of the Śnieżnik Mountains in the Polish Sudetes. Chem. Geol., 50:201–222. Söllner, F., Köhler, H. & Müller-Sohnius, D. (1981): Rb/Sr-Altersbestimmungen an Gesteinen der Münchberger Gneismasse (MM), NE-Bayern. Teil 1, Gesamtgesteinsdatierungen. Neues Jahrb. Mineral. Abh., 141:90–112. Spencer, D.A. (1993): Tectonics of the Higher and Tethyan Himalaya, Upper Kaghan Valley, NW Himalaya, Pakistan: Implications of an early collisional, high pressure (eclogite facies) metamorphism to the Himalayan belt. Ph.D. thesis, ETH Zürich, Switzerland, Dissertation no. 10194, 798 p. Spencer, D.A. & Gebauer, D. (1996): SHRIMP evidence for a Permian protolith age and a 44 Ma age for the Himalayan eclogites (Upper Kaghan, Pakistan): implications for the subduction of Tethys and the subdivision terminology of the NW Himalaya. In 11th Himalaya-Karakoram-Tibet Workshop (Flagstaff, Arizona, U.S.A.), Abstr. Vol., 147–150.

186

H.-J. Massonne & P.J. O’Brien

Spencer, D.A., Ramsay, J., Spencer-Cervato, C., Pognante, U., Chaudhry, M.N. & Ghazanfar, M. (1990): High pressure (eclogite facies) metamorphism in the Indian Plate, Northwestern Pakistan. Geol. Bull. Univ. Peshawar, 23:87–100. Stampfli, G.M. (1996): The intra-Alpine terrain: a Palaeotethyan remnant in the Alpine Variscides. Eclogae Geol. Helv., 89:12–42. Stampfli, G.M., von Raumer, J. & Borel, G.D. (2002): Paleozoic evolution of pre-Variscan terranes: From Gondwana to the Variscan collision. In Martínez Catalán, J.R., Hatcher, R.D., Jr., Arenas, R. & Díaz García, F. (eds.): Variscan-Appalachian dynamics: The building of the Late Paleozoic basement /Geol. Soc. Am. Spec. Pap., 364/. Washington (D.C.): Geol. Soc. Am., 263–280. Steck, A., Epard, J.-L., Vannay, J.-C., Hunziker, J., Giraud, M., Morard, A. & Robyr, M. (1998): Geological transect across the Tso Morari and Spiti areas: the nappe structures of the Tethys Himalaya. Eclogae Geol. Helv., 91: 103–121. Stöckhert, B., Duyster, J., Trepmann, C. & Massonne, H.-J. (2001): Microdiamond daughter crystals precipitated from supercritical COH silicate fluids included in garnet, Erzgebirge, Germany. Geology, 29:391–394. Stosch, H.-G. & Lugmair, G.W. (1990): Geochemistry and evolution of MORB-type eclogites from the Münchberg Massif, southern Germany. Earth Planet. Sci. Lett., 99:230–249. Tait, J., Bachtadse, V. & Soffel, H.C. (1994): Silurian paleogeography of Armorica: new paleomagnetic data from Central Bohemia. J. Geophys. Res., 99:2897–2907. Tait, J., Schätz, M., Bachtadse, V. & Soffel, H. (2000): Palaeomagnetism and Palaeozoic palaeogeography of Gondwana and European terranes. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 21–34. Tapponnier, P., Peltze, G. & Armijo, R. (1986): On the mechanics of the collision between India and Asia. In Coward, M.P. & Ries, A. (eds.): Collision tectonics /Geol. Soc. London, Spec. Publ., 19/, 115–157. Tollmann, A. (1982): Grossräumiger variszischer Deckenbau im Moldanubikum und neue Gedanken zum Variszikum Europas. Geotekt. Forsch., 64:1–91. Tonarini, S., Villa, I., Oberli, M., Meier, F., Spencer, D.A., Pognante, U. & Ramsey, J.G. (1993): Eocene age of eclogite metamorphism in Pakistan Himalaya: Implications for India-Eurasia collision. Terra Nova, 5:13–20. Treloar, P.J. (1997): Thermal controls on early-Tertiary, short-lived, rapid regional metamorphism in the NW Himalaya. Tectonophysics, 273:77–104. Treloar, P.J. & Searle, M.P. (eds.) (1993): Himalayan tectonics /Geol. Soc. London, Spec. Publ., 74/, 630 p. Treloar, P.J., O’Brien, P.J, Parrish, R.R. & Khan, M.A. (2003): Exhumation of early Tertiary, coesite-bearing eclogites from the Pakistan Himalaya. J. Geol. Soc. London, 160:367–376. van Breemen, O., Aftalion, M., Bowes, D.R., Dudek, A., Mísař, Z., Povondra, P. & Vrána, S. (1982): Geochronological studies of the Bohemian Massif, Czechoslovakia, and their significance in the evolution of Central Europe. Trans. R. Soc. Edinburgh: Earth Sci., 73:89–108. van Breemen, O., Bowes, D.R., Aftalion, M. & Żelaźniewicz, A. (1988): Devonian tectonothermal activity in the Góry Sowie gneissic block, Sudetes, southwestern Poland: evidence from Rb-Sr and U-Pb isotopic studies. Ann. Soc. Geol. Pol., 58:3–19. Vejnar, Z. (1986): The Kdyně Massif, South-West Bohemia – a tectonically modified basic layered intrusion. Sbor. Geol. Věd. Geol., 41:9–67. von Quadt, A. (1990): U-Pb zircon and Sm-Nd analyses on metabasites from the KTB pilot borehole. KTB Rep., 90–4:545. von Quadt, A. (1993): The Saxonian Granulite Massif: new aspects from geochronological studies. Geol. Rundsch., 82:516–530. von Quadt, A. & Gebauer, D. (1993): Sm-Nd and U-Pb dating of eclogites and granulites from the Oberpfalz, NE Bavaria, Germany. Chem. Geol., 109:317–339. von Raumer, J.F., Stampfli, G.M., Borel, G. & Bussy, F. (2002): Organization of pre-Variscan basement areas at the north-Gondwanan margin. Int. J. Earth Sci., 91:35–52. Wadia, D.N. (1931): The syntaxis of the northwest Himalayas: its rocks, tectonics and orogeny. Rec. Geol. Surv. India, 65:189–220.

The Bohemian Massif and the NW Himalaya

187

Werner, C.-D. (1981): Sächsisches Granulitgebirge – Saxon Granulite Massif. In Kramer, W. & Werner, C.-D. (eds.): Ophiolites and initialites of northern border of the Bohemian Massif. Guide book of excursions in the German Democratic Republic and Polish People’s Republic. Freiberg: Acad. Sci. GDR, 1:129–161. Werner, O. & Lippolt, H.J. (2000): White mica 40Ar/39Ar ages of Erzgebirge metamorphic rocks: simulating the chronological results by a model of Variscan crustal imbrication. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 323–336. Willner, A.P., Rötzler, K. & Maresch, W.V. (1997): Pressure-temperature and fluid evolution of quartzofeldspathic metamorphic rocks with a relic high-pressure, granulite-facies history from the Central Erzgebirge (Saxony, Germany). J. Petrol., 38:307–336. Willner, A.P., Krohe, A. & Maresch W.V. (2000): Interrelated P-T-t-d paths in the Variscan Erzgebirge dome (Saxony, Germany): Constraints on the rapid exhumation of high-pressure rocks from the root zone of a collisional orogen. Int. Geol. Rev., 42:64–85. Willner, A.P., Sebazungu, E., Gerya, T.V., Maresch, W.V. & Krohe, A. (2002): Numerical modelling of PTpaths related to rapid exhumation of high-pressure rocks from the crustal root in the Variscan Erzgebirge Dome (Saxony/Germany). J. Geodyn., 33:281–314. Winchester, J.A. & PACE TMR Network Team (2002): Palaeozoic amalgamation of Central Europe: new results from recent geological and geophysical investigations. Tectonophysics, 360:5–21. Wittenberg, A., Vellmer, C., Kern, H. & Mengel, K. (2000): The Variscan lower continental crust: evidence for crustal delamination from geochemical and petrophysical investigations. In Franke, W., Haak, V., Oncken, O. & Tanner, D. (eds.): Orogenic processes: Quantification and modelling in the Variscan Belt /Geol. Soc. London, Spec. Publ., 179/, 401–414. Zeitler, P.K., Chamberlain, C.P. & Smith, H.A. (1993): Synchronous anatexis, metamorphism, and rapid denudation at Nanga Parbat (Pakistan Himalaya). Geology, 21:347–350. Zhao, W., Nelson, K.D. & Project INDEPTH Team (1993): Deep seismic reflection evidence for continental underthrusting beneath southern Tibet. Nature, 366:557–559. Zulauf, G. (1993): The Moldanubian/Saxothuringian boundary at the western border of the Bohemian Massif – a polyphase strike-slip zone. Z. Dtsch. Geol. Ges., 144:68–87.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 7, 191–227

Mineral chemistry and mineral reactions in UHPM rocks CHRISTIAN CHOPIN1 and GIOVANNI FERRARIS2 1

Laboratoire de Géologie, École normale supérieure, UMR 8538 du CNRS, 24 rue Lhomond, 75005 Paris, France 2 Dipartimento di Scienze Mineralogiche e Petrologiche, Università di Torino, and Istituto di Geoscienze e Georisorse, CNR, via Valperga Caluso 35, 10125 Torino, Italy; e-mail: [email protected]; [email protected]

Introduction Most of, if not all, the evidence for the attainment of UHP conditions by metamorphic rocks is a mineralogical one. Some minerals by their nature, composition, texture or reactions, may be specific of such conditions and are briefly reviewed here. As pressure, an intensive parameter, has direct influence on volume, this mineralogical survey is organised on a structural basis: we emphasised the more or less efficient packing of atoms into crystal structures. Then chemical, petrological or phase stability data are presented, in the hope of offering a slightly different perspective from earlier reviews by Smith (1988), Chopin & Sobolev (1995), Liou et al. (1998), Carswell & Zhang (1999) and Zhang & Liou (2000). The reader is explicitly referred to the latter three in addition to this review, in order to have the most complete coverage.

Closest- and close packing structures Basic features Most of the index minerals occurring in high pressure metamorphic rocks have crystal structures based on more or less strict closest packings (cp) of oxygen atoms (see Ferraris, 2002). Two basic cp’s of equal spheres are known: one with cubic symmetry based on a face centred (F) lattice (it is known both as fcc = face centred cubic and ccp = cubic closest packing) and another with hexagonal symmetry based on a primitive lattice (P) (it is known as hcp = hexagonal closest packing) (Fig. 1). In both cases the packing spheres are arranged in hexagonal sheets and fill 74% of the space. The sheets are stacked along the body diagonal in fcc and along the c axis in hcp. The different stacking of a same structural layer to form different structures is known as polytypism and is well known in micas and layer silicates in general (see Ferraris & Ivaldi, 2002; Nespolo & Ďurovič, 2002). Note that spheres packed according to a body-centred cubic (bcc) lattice I fill 68% of the space (Fig. 2). In the structures of our interest, the packing spheres are oxygen ions O2– which have an ionic radius of about 1.40 Å (Shannon, 1976). The spaces left empty by the ions (interstitial sites; Fig. 1) are occupied by cations which have usually a smaller ionic radius (units in Å); e.g., IVSi4+ 0.26, IVAl3+ 0.39, VISi4+ 0.40, VIAl3+ 0.535, VIFe3+ 0.55,

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

192

C. Chopin & G. Ferraris

Ch

ah

Fig. 1. Unit cell content of the cubic ccp (top left) and hexagonal hcp (top right) closest packings. In ccp the sheets of spheres are stacked according to a sequence ABCA (dark, light grey, blank, and dark spheres, in this order) along the three-fold axes, i.e. the body diagonals db of an F cubic cell. In hcp a sequence ABA is stacked along the six-fold axis of a hexagonal P lattice with edges ah and ch. In the projection (bottom) the small blank circles and triangles represent the interstitial octahedral and tetrahedral sites, respectively. VI

Ti4+ 0.605, VIFe2+ 0.61, VIMg2+ 0.72, VIIICa2+ 1.12, VIIINa+ 1.18, VIIIK+ 1.51. For this reason the description of a structure is usually done by putting the anions bonded to a cation at the corners of the so-called coordination polyhedron. The number of anions bonded to a cation is known as the coordination number (CN, indicated by a Roman superscript as a few lines above). In both hcp and ccp, the interstitial sites are at the centre of the following regular coordination polyhedra: equilateral triangles (CN 3) lying in the plane of a sheet; tetrahedra (CN 4) formed by an equilateral triangle of spheres belonging to a sheet plus a fourth sphere belonging to the next sheet; octahedra (CN 6) formed by two equilateral triangles of spheres belonging to two consecutive sheets. Each packing sphere is at the centre of a dodecahedron, a coordination polyhedron (CN 12) which occurs in crystal structures where the coordinated and the coordinating spheres

Mineral chemistry and mineral reactions in UHPM rocks

193

Fig. 2. Body-centred cubic (bcc) close packing of spheres. Each sphere, e.g. M, coordinates 8 spheres N according to a cube and has 6 second-neighbour spheres P.

have the same size. The CN of a cation is determined by the ratio between its ionic radius and that of the anion; the CN increases with this ratio. Polymorphism Under different pressure (P) and temperature (T) conditions, the same compound may assume different crystalline structures thus showing the phenomenon of polymorphism. Typically SiO2 (silica) occurs as quartz, coesite, stishovite and other phases (Fig. 3). At least in simple structures, different polymorphs often correspond to deformations of the same basic closest packing structure. In a series of polymorphs, it is observed that the CN of a given cation may increase with P and the polymorphs with higher CN have higher density. Typically in stishovite Si has CN 6 instead of the commonly adopted CN 4. The increase of the density is due to a more compact arrangement of the anions (usually O2–), i.e. to a smaller molar volume per oxygen. The increase of the CN is related to the higher compressibility of the anions in comparison with the almost incompressible cations; consequently the ratio of the cation/anion ionic radii decreases with P and favours a higher CN. In the transition from IVSi to VISi in silica, while the radius of oxygen decreases by 0.20 Å, that of Si decreases only by 0.02 Å; the oxygen–oxygen separation is 2.63 Å in quartz and 2.51 Å in stishovite (Prewitt & Downs, 1998). Thus, the packing of anions in a structure plays a central role when minerals are investigated in connection with UHP metamorphic phenomena. Polymorphic transitions as a function of pressure are mainly related to a decrease of the interatomic distances under compression. Thus, the cation–anion bonds can attain values too short for the stability of the structure. A polymorphic transition (often even without a change in CN) practically overcomes the problem of the stability by rearranging the atoms according to a different type of structure that allows further compression. In case a different crystalline phase is not possible, amorphisation is observed (Richet & Gillet, 1997). The transition between two polymorph phases

194

C. Chopin & G. Ferraris

Fig. 3. Phase diagram showing the stability fields of some silica (SiO2) polymorphs as a function of pressure (P) and temperature (T) (modified after Zhang et al., 1996).

happens at definite P and T values; particularly useful in geothermobarometry are those phases existing in a well defined (narrow) P–T range. A non-equilibrium state of the assemblage and kinetic phenomena, like the energy of activation required for a phase transition, may favour metastable states and thus the coexistence of different polymorphs. Examples are graphite–diamond, calcite–aragonite (CaCO3), rutile–brookite–anatase (TiO2), and andalusite–sillimanite–kyanite (Al2SiO5). On geometrical ground, the polymorphic transitions are classified as follows. Displacive transitions – Only displacements of some atoms occur without breaking and/or forming bonds; no significant changes in the CN are observed. The transformation
-quartz  -quartz is an example. Some displacive transition, e.g. in silica polymorphs, can be described by tilting of coordination polyhedra only. Reconstructive transformations – The bond pattern is (partially) reconstructed through the transition. Quartz – tridymite – cristobalite and rutile – brookite – anatase are two examples. Often the coordination number changes as in the aragonite (Ca CN 9) – calcite (Ca CN 6), kyanite (Al CN 6) – andalusite (Al CN 5) – sillimanite (Al CN 4) and coesite (Si CN 4) – stishovite (Si CN 6) transitions. The mentioned shortening of the cation–anion distances under pressure leads to a more covalent character of the bond. In some cases the changes in the nature of the bond is impressive as for the diamond – graphite transition, where a part of the covalent bonds occurring in diamond (highpressure phase) is changed to van der Waals bonds between layers in graphite (lowpressure phase) (Fig. 4).

Mineral chemistry and mineral reactions in UHPM rocks

195

Fig. 4. Structure of graphite (left) and diamond (right).

Filling tetrahedral sites As reported above, the closest-packing type structures are based either on a ccp or an hcp array of large packing atoms where the interstitial sites are to some extent occupied by smaller packed atoms. The closest packing is often valid only in first approximation; in this case, the term close packing is more appropriate. Figure 5 shows where tetrahedral and octahedral sites occur in ideal ccp and hcp structures. Several structures of our interest can be formally obtained by filling (a part of) the interstitial sites. Diamond has a ccp structure where four of the eight available tetrahedral sites T are alternatively filled (Fig. 4). Each C occurs at the centre of a tetrahedron and forms four very strong covalent bonds. Thus diamond is the hardest material known and plays a unique role in the UHP metamorphic rocks because 1) its presence indicates minimum peak metamorphic pressures of about 4 GPa, and 2) its very low compressibility makes it a perfect container to preserve high-pressure phases as inclusions (see coesite). Indeed, the mechanical strength of diamond prevents the expansion of the inclusions upon decompression (as shown by direct measurement of the internal pressure of fluid inclusions in kimberlitic diamond by Raman spectroscopy: Navon, 1991; Schrauder & Navon, 1993) and prevents even more the large volume increase that would be linked to a transition to a lower-pressure phase in the inclusion. The elastic properties of diamond therefore allow high-pressure phases included in it to be preserved during decompression (e.g. Stachel et al., 2000a; Hutchison et al., 2001), by exerting on them a high confining pressure – which may differ by several GPa from the external pressure! A measure of this internal pressure through the shift of a given Raman frequency with

196

C. Chopin & G. Ferraris

respect to its room pressure value allows various barometric applications (a reappraisal of Rosenfeld’s seminal ideas: Rosenfeld & Chase, 1961; Rosenfeld, 1969), so far more or less limited to kimberlitic diamond (Izraeli et al., 1999; Sobolev et al., 2000). Given the far-reaching implications of finding metamorphic diamond (mostly microdiamond, Sobolev & Shatsky, 1990; Shatsky & Sobolev, 2003), and the ease with which its presence may be an artefact of sample preparation, it is essential to be able to establish beyond doubt its actual presence in metamorphic rocks; either by mechanical techniques (Massonne et al., 1998) or, definitely, by in situ Raman microscopy, which allows focusing below the sample surface and so prevents the suspicion of artefact (e.g. Nasdala & Massonne, 2000), or by transmission electron microscopy after focussed ionbeam milling (Dobrzhinetskaya et al., 2003; Wirth & Rocholl, 2003). Actually, microdiamond is mostly preserved as inclusions in mechanically strong minerals like zircon and garnet (CD Image 1). It was shown to have precipitated as a daughter mineral in supercritical COH fluid/melt inclusions in garnet (Erzbegirge, Germany: Stöckhert et al., 2001; Hwang et al., 2001) or spinel (W Norway: van Roermund et al., 2002), thereby discounting any role of non-hydrostatic pressure in the generation of this UHP phase. Besides, as the former presence of metamorphic diamond is important as well, it is necessary to be able to identify secondary graphite, formed by breakdown of diamond, and to distinguish it from prograde graphite that never went through the diamond state. Clearly, secondary graphite is poorly organised (Nasdala & Massonne, 2000; see also the spectrum ascribed to diamond by Mposkos & Kostopoulos, 2001, which is that of disordered graphite according to Beyssac & Chopin, 2003); it is clearly different from well-organised prograde high-grade graphite (e.g. Beyssac et al., 2002) but its distinction from the less well organised prograde “graphite” in lower-grade rocks requires further study (cf. Leech & Ernst, 1998). The cubic structure of -cristobalite (Fig. 6) is representative of the pre-stishovite silica phases. It can be derived from the diamond structure by substituting C with Si in the small cubes with edge a/2 and adding one O midway between two Si atoms. Thus Si assumes its typical tetrahedral coordination and each corner (oxygen) of a tetrahedron is shared by two Si, a situation that is common to all tectosilicates. Actually, all the prestishovite polymorphs of SiO2 (
-quartz, -quartz, -trydimite, -cristobalite and coesite) are tectosilicates and can be derived from the -cristobalite structure essentially by changing the Si–O–Si angles. Coesite (stable approximately in the range 2.5–9.0 GPa; Fig. 3) is of paramount importance in UHP metamorphic processes. As in all the prestishovite silica phases, Si is tetrahedral in coesite. In its monoclinic structure (with a pseudo-hexagonal cell; Fig. 6) the framework of tetrahedra is arranged as a threedimensional connection of four-membered rings, a feature occurring also in feldspars. The discovery of coesite by Chopin (1984) in the Dora-Maira massif (Italian Western Alps) followed by that of Smith (1984) in the Western Gneiss Region (Norway) opened a new chapter in high-pressure metamorphic petrology, proving that continental crustal rocks could experience a peak pressure of at least 3 GPa and be brought back to exposure (e.g. Chopin, 2003). The rock mineralogy tending to adapt to changing pressure (P) and temperature (T) conditions, such UHP phases normally disappear during decompression, i.e. during the movement of the rock toward the surface (= exhumation), even more so if the rock enters the stability field of the respective low pressure phase at high temperature.

Mineral chemistry and mineral reactions in UHPM rocks

Fig. 5. Octahedral (O) and tetrahedral (T) interstitial sites in a ccp (top) and hcp unit cell (bottom).

Fig. 6. Crystal structure of cubic -cristobalite (left) and monoclinic coesite (right).

197

198

C. Chopin & G. Ferraris

As a matter of fact, the metastable persistence of coesite up to surface conditions is commonly the result of its inclusion by mechanically strong host minerals like zircon (or diamond), in which it is commonly intact, or garnet, kyanite or clinopyroxene, in which it shows incipient rim transformation to quartz (CD Image 2). These strong host minerals act as pressure vessels around the inclusion, insulating it from metamorphic fluid and preventing the volume increase of the back transformation, i.e. maintaining the inclusion pressure on the quartz–coesite transition curve as long as the tensile strength of the host mineral can sustain the difference between the internal (inclusion) and external pressure. Although early anticipated (Gillet et al., 1984), the effectiveness of this mechanism was only recently demonstrated for small (10 to 20 µm) coesite inclusions in garnet and zircon, in which Raman microspectroscopy revealed residual internal pressures that could exceed 2 GPa (Parkinson & Katayama, 1999; Parkinson, 2000; Ye et al., 2001). Note that, once the host mineral is fractured (CD Image 2), or in the absence of a “container”, i.e. in the rock matrix, there is no other obstacle than kinetics to the complete transformation of the high-P phase into the lower-P polymorph, i.e., for coesite/quartz, temperatures lower than about 375–400 °C (Mosenfelder & Bohlen, 1997) or a completely dry system (Liou & Zhang, 1996). Note also that radial cracks around a quartz inclusion in a strong host mineral cannot alone be taken as evidence of pre-existing coesite. Indeed the uncommon compressibility of quartz and its high thermal expansion may combine, for quartz included at high pressure, to overcome the tensile strength of the host mineral upon decompression, in particular if accompanied by heating (Wendt et al., 1993). At about 9 GPa, coesite transforms to stishovite, which has a rutile structure (see below) where Si is playing the role of Ti. Filling octahedral sites

Fig. 7. Cubic structure of NaCl, an archetype model for close packed structures with filled octahedral sites (e.g., calcite). In this structure the large Cl– anions represent the packing spheres and the smaller Na+ cations occupy the interstitial octahedral sites.

In ccp the four octahedral interstitial sites per cell occur at the middle of the edges and at the centre of an F cubic cell (Fig. 5). The NaCl (halite) structure (Fig. 7) is the archetype of the ccp structures that have all their four octahedral sites occupied. Wüstite (FeO) and periclase (MgO) assume the NaCl structure. In spite of their isostructurality and the closeness of the ionic radii of Fe2+ and Mg2+ (two necessary, but not sufficient, conditions to form extended solid solutions), wüstite and periclase form only a very limited solid solution at high pressure. Names like “ferropericlase” (for a limited solution of Fe in periclase) and “magnesiowüstite” (for a limited solution of Mg in wüstite)

Mineral chemistry and mineral reactions in UHPM rocks

199

Fig. 8. Hexagonal structure of nickeline NiAs, an archetype structure for aragonite.

are not uncommon in petrologic literature. However, they do not correspond to approved mineral species and wordings like Fe-bearing periclase and Mg-bearing wüstite are recommended. The lack of an extended solid solution between wüstite and periclase depends of the fact that, in the two end-member structures, the packing of the oxygen atoms is strictly tailored either for Fe2+ or Mg2+ and does not admit even a limited deformation as required by the substitution Fe2+  Mg2+. In fact, this deformation would modify the cation–cation distance, which is a critical parameter for the stability of these two structures. Note that Fe2+ and Mg2+, instead, completely replace each other in more complex and flexible structures, like that of olivine (see below). A quite complex structure like that of calcite, the rhombohedral polymorph of CaCO3, can be described by a rhombohedral cell obtained from a cubic NaCl cell compressed along one body diagonal and replacing Na+ and Cl– with the ions Ca2+ and (CO3)2– respectively. The (CO3)2– groups are planar and perpendicular to the three-fold axis of the cell; each Ca2+ is coordinated by six oxygen atoms belonging to six different (CO3)2– groups. The structure of aragonite, a second polymorph of CaCO3, is instead obtained by deformation of another basic structure, that of NiAs (nickeline), which can be described as a distorted hcp array of As atoms with Ni filling the octahedral sites (Fig. 8). Because of the distortion, also the packing atom As assumes CN 6 according to a trigonal prism. By replacing As with Ca and Ni with the carbonate group, the structure of aragonite is obtained. In aragonite, the symmetry is lowered to orthorhombic by a compression perpendicular to the layers; consequently the CN of the site occupied by Ca rises from 6 to 9. The higher coordination of Ca and a density of 2.90 g/cm3 against 2.72 g/cm3 for calcite, is in agreement with aragonite being the high-pressure polymorph of CaCO3. Unlike coesite, aragonite has been seldom recorded from UHP rocks (as micrometre-size inclusions in zircon?), most likely because of the high kinetics of the transition to calcite which is operative down to low temperatures (e.g. Carlson & Rosenfeld, 1981). The rutile (one of three polymorphs of TiO2; Fig. 9) structure type can be approximately described as a hcp array of oxygen atoms with only half octahedral sites occupied by Ti. The repulsion between the high charge Ti4+ cations lowers the symmetry to tetragonal; the close packed sheets of oxygen atoms are parallel to (1¯01). Crystallographically oriented needles of rutile in garnets were reported as exsolution (Zhang & Liou, 2000; see also Belluso et al., 1998) but their significance is still unclear

200

C. Chopin & G. Ferraris

Fig. 9. Orthorhombic structure of rutile, a polymorph of TiO2 (titania).

(see below at “Garnet”). As mentioned above, a rutile-type structure is adopted by stishovite, a polymorph of SiO2 where Si has octahedral coordination. The quite complex oxides of the columbite–tantalite series (general formula AB2O6; mainly A = Mn, Fe, Mg and B = Nb, Ta) adopt a structure which is close to that of rutile, but the filled octahedral sites are differently arranged. This structure is important in the present context because, with the variant of A = B (known as
-PbO2 structure), it occurs in a polymorph of TiO2 that is stable above about 4.5 GPa (Arlt et al., 2000; Withers et al., 2003). Recently this polymorph has been identified in garnet of diamondiferous rocks from the Erzgebirge, as a nanometre-size slab in rutile (Hwang et al., 2000; Massonne & O’Brien, 2003), and as a shock-induced phase in the Ries Crater; it is expected to occur in the Earth’s upper mantle below 125 km (El Goresy et al., 2001). Some ternary compounds with general formula ABO3 are of concern here. The trigonal structure type of corundum (
-Al2O3; Fig. 10) may formally be derived from a hcp array of oxygen atoms where 2/3 of the octahedral sites is occupied by Al atoms. The perovskite (CaTiO3) structure-type (Fig. 11) can be described as a ccp array of Ca + O atoms where the small cation Ti occupies 1/4 of the octahedral sites. When the A ion in an ABO3 compound is too small to adopt the perovskite structure, a corundum-type structure is formed with both A and B ions occupying 1/3 each of the octahedral sites. This structural type is known as the ilmenite, FeTiO3, structure. Ilmenite, which shows solid solution with MgTiO3 (geikielite), is quite common in igneous and metamorphic rocks, was long known in intergrowths with clinopyroxene in kimberlite nodules, and was more recently recognised as micrometre-size inclusion rods in olivine and other minerals from various UHP terranes (Dobrzhinetskaya et al., 1996; Hacker et al., 1997; Zhang & Liou, 2000). There is still controversy as to whether these ilmenite rods are associated to pre-

Mineral chemistry and mineral reactions in UHPM rocks

201

Fig. 10. Hexagonal structure of corundum: identical (001) octahedral layers are stacked. The structure of ilmenite mainly differs for alternating a (001) layer containing Ti with one containing Fe. In the archetype structure of LiNbO3, both Li and Nb stay in the same (001) octahedral layer.

existing humite intergrowths in olivine and result from the breakdown of Ti-bearing humite layers at conditions that need not to exceed “standard” peridotite conditions of about 3 GPa (Risold et al., 2001, 2003; Morten & Trommsdorff, 2003), or whether they are evidence for the former presence of a high-pressure polymorph of olivine (wadsleyite?) in which Ti is more soluble than in olivine, implying an origin at and exhumation from huge depths (corresponding to > 6 GPa, Dobrzhinetskaya et al., 1996; Bozhilov et al., 2003). Geikielite with composition (Mg0.52Fe0.48)TiO4 has been reported by Alberico et al. (2003b) as microinclusions in Foltea (South Carpatians) pyropes (P–T estimated conditions: 2.2–2.4 GPa and 1150–1300 °C). Ilmenite- and perovskite-type structures play an important role at very high pressures (and as shock products). For example, an (Mg,Fe)SiO3 pyroxene stoichiometry is observed both with an ilmenite-type (akimotoite, in the 20–24 GPa range) and a perovskitetype (over 24 GPa) structure (cf. Fiquet, 2001). FeTiO3 itself shows a high-pressure perovskite polymorph (about 5 GPa at 1000 °C; Leinenweber et al., 1991; Navrotsky, 1998) which, according to Dobrzhinetskaya et al. (1996), was the Fe-Ti phase originally exsolved in the Alpine Arami peridotite massif. That was taken as evidence for exhumation from a depth of more than 300 km. The electron diffraction pattern attributed by Dobrzhinetskaya et al. (1996) to the perovskite polymorph of FeTiO3 has

Fig. 11. Orthorhombic structure of perovskite, CaTiO3, represented both as ccp closest packing of 3O + Ca atoms with Ti occupying the interstitial octahedral sites (left) and as packing of Ti-octahedra (right).

202

C. Chopin & G. Ferraris

Fig. 12. Cubic structure of spinel MgAl2O3. Small and intermediate spheres represent tetrahedral (Al) and octahedral (Mg) cations respectively.

been interpreted differently by Hacker et al. (1997) and Risold et al. (2001), thus denying the geological interpretation of the original authors. No ascertained natural samples of silicate perovskite have been reported so far, even if their transformation products might occur as inclusion in the “ultra-deep diamonds” sampled by some kimberlites (e.g. Stachel et al., 2000b; Hutchison et al., 2001). Filling tetrahedral and octahedral sites A large number of complex oxides AB2O4 are based on the ideally cubic spinel (MgAl2O4; Fig. 12) structure where the O2– ions form a ccp array. The A and B atoms occupy 1/8 and 1/2 of the tetrahedral and octahedral interstices, respectively. At high pressure, some simple compounds assume the spinel structure. Thus (Mg,Fe)2SiO4 occurs as olivine (
-phase; Fig. 13) up to about 12–15 GPa, depending on the Fe content and temperature, then (at over 400 km depth in the Earth’s mantle) transforms first to the -phase wadsleyite (an orthorhombic distorted ccp structure containing Si2O7 groups) and after, at 520 km depth, to the spinel-type ccp structure ringwoodite (-phase), both known from shocked meteorites. However, if the Fe/(Fe + Mg) ratio is higher than 26%, olivine transforms directly to a spinel phase. These phase transitions, as any polymorphic transition, maintain the stoichiometry of olivine but lead to compounds that are about 8% (wadsleyite) and 11% (ringwoodite) denser than olivine; besides, these high-pressure phases can be partially hydroxylated thus accounting for water in the mantle (cf. Fiquet, 2001). In the inverse spinel structure, half of the B atoms occupy tetrahedral sites and the other half is statistically distributed among the octahedral sites. Magnetite,

Mineral chemistry and mineral reactions in UHPM rocks

203

Fig. 13. Orthorhombic structure of Mg2(SiO4) forsterite. Mg occupies two independent octahedral sites, Mg1 and Mg2.

(Fe2+Fe3+2O4), which may occur as lamellae in olivine (Zhang & Liou, 2000), shows an inverse spinel structure, half of the Fe3+ being in tetrahedral sites. The orthorhombic structure of olivine, (Mg,Fe)2SiO4 (Fig. 13), can be described in terms of a distorted hcp array of oxygens where only 1/8 of the tetrahedral sites and ½ of the octahedral sites are occupied by Si and (Mg, Fe), respectively. The octahedral cations (Mg, Fe) normally occupy at random two crystallographically independent sites M1 and M2, but (partial) Mg/Fe ordering can occur depending on genetic conditions.

Rock-forming minerals with a complex structure Besides some rock-forming minerals mentioned, like olivine, other complex phases are important for the petrology of UHP metamorphic rocks. Garnets 3+ Garnets are cubic with a general formula A2+ 3 B2 [SiO4]3. A and B represent large (Ca, Mg, Fe, Mn; CN 8–12) and small (Al, Fe, Cr; CN 6) cations, respectively. Alternating octahedra and tetrahedra form a three-dimensional network by sharing corners; within this framework, large distorted cubic/dodecahedral sites are occupied by the large cation A. Natural garnets are practically always solid solutions of several end-members. The garnet Mg3MgSi[SiO4]3, end-member majorite (note the MgSiO3 pyroxene stoichiometry), is stable only at Earth’s mantle conditions (> 16 GPa). In this high-pressure garnet 25% Si occurs in octahedral coordination and the usual cubic structure deforms to tetragonal (because of ordering of the two octahedral cations). However, Fe-rich majorites are reported to be cubic and a tetragonal  cubic phase transition seems possible (cf. Fiquet,

204

C. Chopin & G. Ferraris

2001). As discussed above, octahedral coordination of Si becomes possible at high pressure (see stishovite) because the ionic radius of O2– decreases. The Si-free garnet katoite, Ca3Al2(O4H4)3, which may be formally derived from grossular, Ca3Al2(SiO4)3, by the substitution 4H  Si, is a possible candidate for incorporating OH at high pressure in the Earth’s mantle. Recently Lager et al. (2002) have shown, by single crystal X-ray diffraction, that at about 5 GPa this garnet undergoes a phase transition which lowers the symmetry to tetragonal and requires that OH occupies two independent crystallographic sites, instead of one as in the cubic phase. That may allow two independent and different mechanisms of incorporating OH in a grossular-type structure. Garnets in metamorphic rocks, including UHP ones, mostly belong to the almandine–pyrope–grossular series, i.e. A = (Fe2+, Mg, Ca) and B = Al; almost pure pyrope is typical in magnesian rocks of the Dora-Maira UHP terrane (Chopin, 1984; CD Image 3). Some incorporation of Na into the garnet structure, as reported in kimberlitic rocks, may occur also in UHP metamorphic rocks through a coupled substitution of Na+ + Ti4+ for Ca2+ + Al3+ (Shatsky et al., 1995; Enami et al., 1995; Vrána & Frýda, 2003). However, in the extremely Ti-poor, near-end-member pyrope of the Dora-Maira massif, the noticeable Na and P contents are clearly correlated and therefore evidence for the Na+ + P5+  Mg2+ + Si4+ substitution (Brunet & Lecocq, 1999). In more oxidised and Al-poor rocks, Fe3+ may replace Al in the B site. In some HP and UHP Mn-rich metasediments or microsystems, the calderite component Mn3Fe3+ 2 Si3O12 becomes significant or even dominant (Reinecke, 1987; Enami et al., 1995; Cenki & Chopin, in prep.); the endmember is itself a high-pressure phase stable above ca. 2 GPa (Lattard & Schreyer, 1983). In Mn-free systems, the ferric iron incorporation in garnet toward the ideal end3+ member Fe2+ 3 Fe2 Si3O12 (called “skiagite” and stable above 9 GPa) is strongly pressure dependent (Woodland & O’Neill, 1993, 1995) but awaits documentation in nature. Oriented precipitates (“exsolutions”, usually lamellar in shape) of pyrope–grossular garnet in pyroxene are known from eclogite xenoliths and some Alpine-type peridotites (Zhang & Liou, 2000, with references; van Roermund et al., 2002); they may result from the instability of the aluminous fraction of pyroxene (the high-temperature Tschermak component (Mg,Ca)Al[SiAlO6]) upon cooling. Much more relevant to UHP rocks is, in the peridotite body of Otrøy Island, western Norway, the microtextural evidence for exsolution of orthopyroxene in coarse garnet (CD Image 4), implying a “super-silicic” precursor garnet, i.e. the presence of several mole percent majorite component in it (van Roermund & Drury, 1998; van Roermund et al., 2000, 2001; Carswell & Cuthbert, 2003), which broke down upon decompression according to the reaction (Mg,Fe,Ca)3(Al1–2xMgxSix)2[SiO4]3 = 8x MgSiO3 + (1–2x) (Mg,Fe,Ca)3Al2[SiO4]3. garnet with 200x mol% majorite

opx

pyropic garnet

It was known from experiment (Akaogi & Akimoto, 1977) that with increasing pressure pyropic garnet can incorporate more and more MgSiO3 as majorite component. Modal estimates of the majorite component originally present in the Norwegian garnet, i.e. before exsolution, lead to formation pressures exceeding 6 GPa (> 200 km). Such sensational exsolution had been discovered in garnet from kimberlite xenoliths a few

Mineral chemistry and mineral reactions in UHPM rocks

205

years earlier (Haggerty & Sautter, 1990) but it was completely unexpected that the delicate textures (CD Image 4) could be preserved during the long ascent and slow cooling of an orogenic peridotite body, as opposed to quenching by explosive kimberlite sampling. Admittedly, mantle rather than crustal material was involved in either case, but the demonstration was made that, even with the rates attending metamorphic processes, such evidence could be preserved – opening the prospect that it could be found in UHP crustal rocks, too. Since then, indeed, three reports of related textures were made in UHP terranes. In garnet peridotite of the Su-Lu terrane, eastern China (Ye et al., 2000a), the textural evidence for pyroxene exsolution from garnet is convincing but also involves oriented micrometre-size rods of rutile and apatite. (However, it is unclear that the extreme conditions recorded by the ultramafic body were also shared by the associated coesitegrade UHP crustal rocks: as in Norway, the extreme conditions may record an earlier, deeper mantle stage before the body was amalgamated with crustal material during “normal” UHP metamorphism; cf. Brueckner et al., 2002). In two other reports from the Greek Rhodope, the crustal nature of the rocks is clear but the evidence for exsolution from a former majoritic garnet is less compelling: in garnet from kyanite-biotite gneiss, no pyroxene occurs but oriented rods of quartz and rutile (Mposkos & Kostopoulos, 2001); and in garnet from metabasite, with oriented rutile rods, clinopyroxene occurs as blebs (Liati et al., 2002), which may or may not represent recrystallised exsolution products (alternatively they could be inclusions trapped during garnet growth). A word of caution may be useful at this stage. Whereas oriented pyroxene precipitates in garnet can be simply accounted for by the breakdown of a majorite component, the occurrence of quartz, or rutile or possibly apatite precipitates in garnet points to different reactions which may result from substitutions that are unrelated to the majoritic one – and so may have quite a different meaning in terms of P and T conditions attained. The high phosphorus, titanium and sodium contents mentioned by Mposkos & Kostopoulos (2001), for instance, can be related to substitutions like Mg2+ + Al3+  Na+ + Ti4+ and Mg2+ + Si4+  Na+ + P5+, which, although known in high-pressure and/or hightemperature garnets (e.g. Thompson, 1975; Haggerty et al., 1994; Brunet & Lecocq, 1999), may not bear the same implications of extreme pressures as the majoritic substitution. In any event, the presence in garnet of precipitates that do not involve pyroxene cannot be simply equated with a majoritic component in garnet. Furthermore, oriented inclusions should not be interpreted as precipitates without discussion, since crystallographically oriented quartz rods have been repeatedly described in amphibolite facies garnet and are demonstrably an epitactic growth feature there, not an exsolution (Andersen, 1984; Burton, 1986; Rice & Mitchell, 1991). Likewise, crystallographically oriented rutile needles in garnet might not be exsolution features, but epitactic intergrowths as well (Wang et al., 1999). Pyroxenes Pyroxenes are chain silicates represented by the formula M2M1Si2O6 where the two octahedral M sites are involved in a wide range of isomorphic substitutions. In the

206

C. Chopin & G. Ferraris

orthorhombic and monoclinic pyroxenes with a P primitive Bravais lattice, two independent tetrahedral chains (A and B) occur, whereas clinopyroxenes with a C centred lattice have only one type of chain. The chains are bonded to layers of octahedra consisting of two independent sites, M1 and M2. Main octahedral cations are Mg, Fe, Ca, Al, Na. Different stacking sequences of the octahedral layers and tetrahedral chains produce different structures which may occur as submicroscopic intergrowths (or exsolved micro-inclusions? Zhang & Liou, 2000); their identification often requires transmission electron microscopy (TEM). Angel et al. (1992) have described a highpressure transition of clinoenstatite, Mg2Si2O6, from space group P21/c to C2/c occurring near 6 GPa at room temperature. The same transition in the FeSiO3 system occurs at much lower pressure, near 1.4 GPa, and could have been of relevance to UHP rocks, were it not unquenchable (Woodland & Angel, 1997). Recently Wu et al. (2002) have interpreted nano-domains of P21/n omphacite in a C2/c matrix of jadeite, Na(Al,Fe3+)Si2O6, from the Dabie UHP metamorphic terrane as the breakdown product of the primary jadeite during retrogressive decompression. These domains of P21/n omphacite, a solid solution of jadeite (Jd) and diopside (Di, CaMgSi2O6), have a composition close to Jd50Di50. Note that the octahedral cations are disordered in the C2/c structure and ordered in the P21/n structure. A remarkable feature of calc-silicate rocks from the diamond-bearing Kokchetav massif is the extremely high potassium contents of clinopyroxene (up to 1.5 wt% K2O, Sobolev & Shatsky, 1990). This feature was so far solely known from a few kimberlite eclogitic xenoliths (survey in Chopin & Sobolev, 1995) and experimentally shown to be stable at pressures of 4 to 10 GPa. The Kokchetav clinopyroxene may contain 0.17 K apfu and is thought to be a relic of the earliest high-temperature crystallisation of a Kand Al-rich silica undersaturated melt at pressure above 5 GPa. Upon decompression these potassic pyroxenes, if not included in diamond, show characteristic textures with oriented precipitates of K-feldspar (CD Image 5; also Becker & Altherr, 1992; Zhang & Liou, 2000; Perchuk et al., 2002), sometimes of a later K-bearing mica (Ogasawara et al., 2002). Nevertheless, evidence for a real structural occurrence of K in pyroxenes, and not in some included exsolution, has been recently obtained by refining the crystal structure of a natural sample from Kokchetav complex (Bindi et al., 2003) and two samples synthesised in the system CaMgSi2O6–KAlSi2O6 at 7 GPa (Bindi et al., 2002). Up to 0.23 apfu of K have been found to substitute Ca causing a deformation and increase in the volume of the structural site M2. The replacement of Ca2+ by K+ in the site M2 is charge balanced by Al3+ replacing Mg2+ in the site M1. No evidence for a replacement of Ca with the smaller Mg in order to compensate the large K+ cation was detected; as usual, all Mg occupies the site M1. At high pressure the large K+ cation can enter a smaller coordination polyhedron suitable for Ca2+ because the coordination polyhedron of K+ shows a large compressibility and at a pressure of about 5 GPa reaches dimensions typical of a Ca polyhedron (Harlow, 1997). Pyroxenes with octahedral vacancies are indicative of high-pressure conditions (Gasparik & Lindsley, 1980) but may also be obtained at high temperature (Fockenberg & Schreyer, 1997). They have been reported from various UHP terranes as summarised by Bruno et al. (2002), who describe a Ca deficient jadeite containing 0.08–0.17

Mineral chemistry and mineral reactions in UHPM rocks

207

calcium-Eskola pyroxene, Ca0.5AlSi2O6, from an UHP metagranodiorite of the DoraMaira massif. In addition, a typical feature of many UHP eclogites is the occurrence (CD Image 6) of oriented precipitates of quartz needles along the c axis of omphacite (e.g. the TEM study of Schmädicke & Müller, 2000). This may be interpreted by the presence of a HP–HT precursor with octahedral vacancies reacting to produce a vacancy-free pyroxene with some tetrahedral Al (therefore a relatively HT process, as suggested by the absence of such precipitates in “cold” UHP terranes like Dora-Maira, and their occurrence in high-pressure granulites; Gayk et al., 1995). Phengite Phengite is a series name for solid solutions (Rieder et al., 1998) involving the end-members muscovite, KAl2[AlSi3O10](OH)2, aluminoceladonite, KAl(Mg,Fe2+)[Si4O10](OH)2, and celadonite, KFe3+(Mg,Fe2+)[Si4O10](OH)2. It indicates micas with composition K(Al2–xMgx)[Al1–xSi3+xO10](OH)2 (x is roughly in the range 0.20–0.80, and Mg may be replaced by Fe2+). The substitution VI(Mg2+) + IV(Si4+)  VI(Al3+) + IV(Al3+), which formally leads from phengite to muscovite, is nothing else than the Tschermak substitution. Structurally, phengitic micas are dioctahedral TOT (or 2:1) layer silicates where the octahedral sheet O is only 2/3 occupied by (Al, Mg). They normally occur as 3T or 2M1 polytypes which, strictly speaking, should be considered polymorphs. In fact, the two structures do not differ only in the stacking of a same layer, but the occupied octahedral sites are equivalent in 2M1 (space group C2/m) and independent sets in 3T (space group P3112); instead, the tetrahedral sites belong to two independent sets in both polytypes. Consequently, ordering of the octahedral cations is allowed in 3T but not in 2M1. This thermodynamically important difference explains the different fields of stability of the two polytypes. As a matter of fact, the Si content of phengite increases with P (Massonne & Schreyer, 1986, 1989) and the 3T polytype is more stable at high P than 2M1 (Fig. 14; see Pavese et al., 1999b). A 3T  2M1 transition, accompanied by change in the pattern of order of cations, is believed to occur by decreasing pressure/temperature (P/T; Sassi et al., 1994). Tetrahedral and octahedral cation ordering has been extensively investigated by both neutron and synchrotron radiation (Pavese et al., 1997, 1999a, 1999b, 2000, 2001, 2003) and some degree of octahedral ordering has always been observed in the 3T polytype. The better is the structural fit between T and O sheets, the more hexagonal (i.e. undistorted) is the mesh of the T sheet where the interlayer cation can sink; consequently a shorter c parameter is expected. The contraction of the c parameter with the increase of the Si content reported by Massonne & Schreyer (1986, 1989) in synthetic phengites is clearly due to this sinking effect because, as mentioned above, the aluminoceladonitic substitution improves the T/O fit. The same authors proposed two phengite barometers based on the increase of Si with the metamorphic pressure (the influence of the temperature is small, but see below). Ivaldi et al. (2001) have found the same result on natural samples (see also Ferraris & Ivaldi, 2002) and obtained the following correlation (Fig. 14) between the formation pressure P (kbar) and the layer-plus-interlayer thickness tL = c sin/n (n is the number of TOT layers in a unit cell, i.e. 3 and 2 for the 3T and 2M1 polytypes, respectively).

208

C. Chopin & G. Ferraris

Fig. 14. Plot of ln P vs. the thickness of the layer tL (Å). The regression line (R2 = 0.86) has been obtained by fitting data of 21 3T (circles) and 19 2M1 (triangles) natural phengites for which the c parameter and some estimate (not based on a phengite barometer) of the formation pressure P (kbar) are known (see text).

ln P = –27(2) tL + 271(18). The equation has been obtained by fitting data of 21 3T and 19 2M1 natural phengites (R2 = 0.86; estimated standard deviations in parentheses) for which the c parameter and some estimate (not based on a phengite barometer) of the formation pressure are known (some unpublished data have been added to those of Ivaldi et al., 2001). Thus, with all the cautions of ignoring temperature and other minor effects (see below), the knowledge of the cell parameter c allows an estimate of P independently of chemical data. Note that accurate values of c can now be obtained even at nanometric scale by transmission electron microscopy. The phengitic substitution is actually one of the most powerful and flexible thermobarometers available, for a variety of rock compositions. As exemplified in Figure 15 the composition of the white mica in any low-variance assemblage (3 phases + quartz or coesite + H2O in the KMASH system, 4 phases in the KFMASH system etc.) is fixed at given P and T, i.e. independent of the bulk-rock composition. For a given phengite composition in such a divariant assemblage, we have a univariant relation between P and T. Depending on the assemblage, this relation may be more dependent on P (e.g. with biotite + K-feldspar, or talc + kyanite + garnet, or garnet + omphacite) or on T (see e.g. Wei & Powell, 2003), as can now be accurately calculated with the mixing models independently developed by Parra et al. (2002) and Coggon & Holland (2002) for the muscovite–aluminoceladonite–paragonite series. Even subtle but long-noticed features as the decrease of Na content in phengite with increasing Si content are now reproduced by the thermodynamic data, which lends confidence in them. Garnet–omphacite–phengite

Mineral chemistry and mineral reactions in UHPM rocks

209

Fig. 15. The K2O–MgO–Al2O3–SiO2–H2O system, projected from quartz and H2O. The solid solutions involved in the three-phase, divariant fields have a fixed composition at given P and T. Each of these assemblages therefore constitutes a potential geothermobarometer (e.g. Massonne & Schreyer, 1989).

thermobarometry is therefore a reliable tool, provided one can make clear what the actually coexisting mineral compositions were (Carswell & Zhang, 1999; Ravna & Paquin 2003). In this respect, it is common experience that the heterovalent Tschermak substitution is less prone to reequilibration than homovalent ion exchange reactions, making phengite a powerful recorder of metamorphic history. However, HRTEM observations of UHP phengite from Dora-Maira reveal new features that indicate a more complex behaviour, but at a sub-microscopic scale. In the pyrope-bearing rock, crystals of Si3.55 phengite-3T, which are almost free of stacking faults, contain locally monocrystalline quartz platelets 100–700 Å thick that are confined by the (001) mica planes (CD Image 7). In the vicinity of these platelets, the phengite matrix is faulted by short sequences of talc-2M. This suggests (Ferraris et al., 2000a) that both quartz and talc are precipitates in the phengite matrix, i.e. the products of a decompression reaction that leads to a less Si-rich phengite (and may also have produced the small talc flakes that are seen in optical microscopy to overgrow the large phengite crystals). The reaction is of the type 3 aluminoceladonite + 2H+  1 muscovite + 5 quartz + 1 talc + 2 H2O + 2 K+.

210

C. Chopin & G. Ferraris

In an impure marble of the same UHP unit, phengite reveals another complex nanostructure consisting of areas of exsolved lower-Si phengite (3.39–3.42 Si a.p.f.u.) whose  and c cell parameters differ from those of the higher-Si matrix (3.47–3.51 Si a.p.f.u.; Ferraris et al., 2002). The difference in the cell parameters is due to the different contents in both the octahedral and tetrahedral sites, which might result from an incomplete reaction taking place in the matrix during decompression. Another possible substitution in potassic white mica is boron for tetrahedral aluminium, leading to boromuscovite, KAl2[BSi3O10](OH)2. In contrast to muscovite, the boron end-member is not stable under normal P–T conditions of the continental crust along a 30 °C/km geotherm, and especially not during the intrusion of acidic igneous rocks including their pegmatites, which may explain its scarcity in nature. However, it may form in the waning hydrothermal stages of deep-seated (> 10.5 ºC/km) pegmatites and, provided sufficient boron is available, in HP/LT subduction zone environments, where it may carry boron to considerable depths (Jung & Schreyer, 2002). Magnesiostaurolite The early synthesised Mg end-member of the staurolite group has long been recognised as a HP–UHP phase (Schreyer & Seifert, 1969). Fockenberg (1995a, 1998) determined its compositional variations and stability field (1.2 < P < 6 GPa, 600 < T < 900 °C) in the MgO–Al2O3–SiO2–H2O system (MASH); thermodynamic extractions from these phaseequilibrium data were made by Massonne (1995) and combined with calorimetric data by Grevel et al. (2002). The structure and crystal chemistry of the staurolite group have been a persisting source of perplexity until the formidable study of Hawthorne et al. (1993), after Lonker (1983) had demonstrated the variability of the number of protons and Dutrow et al. (1986) had shown that lithium may be a major though not essential constituent of it. In brief, the staurolite structure can be described as kyanite-like and oxide-hydroxide layers alternating along [010]. Both layers contain three independent octahedra (M) and one tetrahedron (T); they are named M1A, M1B, M2 and T1 in the kyanite-like layer and M3A, M3B, M4 and T2 in the oxide-hydroxide layer. Although truly orthorhombic samples have not yet been found in nature, staurolite can be described as an order–disorder series between a completely disordered orthorhombic end-member with space group Ccmm and an ordered monoclinic end-member with space-group C2/m and  values close to 90.7° (Hawthorne et al., 1993). This transition is continuous and can be modelled as second-order; the primary order parameter (M3A vs. M3B site occupancy) being linearly related to the  angle. An uncommon feature of this silicate is the presence of a tetrahedral site (T2) occupied by divalent cations (essentially Fe2+). Whereas Fe-dominant or Zn-rich staurolite is common in intermediate-grade metamorphic rocks, Mg-rich to Mg-dominant members occur in HP and UHP rocks (e.g. in Sulu eclogite, XMg 0.68–0.74, Enami & Zang, 1988; in Cabo Ortegal ultrabasics, Spain, XMg 0.74 and Cr2O3 > 6 wt%, Gil Ibarguchi et al., 1991; in Dora-Maira pyrope, XMg 0.77–0.96, Simon et al., 1997). Magnesian staurolite offers an interesting crystal chemical paradox: its formation from low-pressure precursors like chlorite implies a

Mineral chemistry and mineral reactions in UHPM rocks

211

decreasing coordination of Mg with increasing pressure, should this element occupy the same site as Fe and Zn. However, the re-refinement of the crystal structure of Dora-Maira magnesiostaurolite (Chopin et al., 2003a) reveals that the relevant tetrahedral site T2 is occupied to less than 70% (by Mg and Li) against 90 to 95% in most staurolites; conversely the octahedral M4 site reaches near 25% occupancy (by Mg and Fe), against only a few percent occupancy in most staurolites. To emphasise this, the formula may be written T2

(Mg1.9,Li0.9,1.2)M4(Mg0.7,Fe0.2,3.1)(Al17.6Mg0.4) = 18Si8O44(OH)4.

This shift of the divalent cation from a tetrahedral to an octahedral position in the UHP phase is a significant feature, accounting for the staurolite paradox. It has important bearing on the thermodynamics and mixing properties of the staurolite group. This staurolite is highly disordered (  90°) but some crystals show a typical tweed-like appearance under crossed nicols, like cross-hatched twinning (Simon et al., 1998). At the TEM scale, distinct lamellar systems can be seen, parallel to either (001) or (100), some monoclinic, some pseudo-orthorhombic, and some showing superstructures along a or c (Simon et al., 1998). These features are evidence for incipient reorganisation of the proton and cation distribution scale, in spite of the extremely high exhumation (and cooling) rate of these metamorphic rocks (e.g. Rubatto & Hermann, 2001). Besides, the Li contents and several hundreds ppm BeO of the crystal studied by Chopin et al. (2003b) show that staurolite acts as a main carrier of light elements, also at UHP conditions. An interesting point in terms of phase relations is, in the Dora-Maira pyrope, the coexistence of magnesiostaurolite with the talc–clinochlore–kyanite assemblage (CD Image 8), which is a lower-temperature alternative both to magnesiostaurolite and to pyrope (+ H2O). The complete assemblage would therefore be invariant in the Fe- and Li-free MASH system. A key point is the existence of a stability field for the talc–staurolite pair, a rare assemblage otherwise only found in Antarctica (Grew & Sandiford, 1984, with XMg = 0.4 in staurolite). In the MASH system, the two univariant reactions bounding this field, namely clinochlore + kyanite = talc + Mg-staurolite + H2O and talc + Mg-staurolite = pyrope + kyanite + H2O are nearly indistinguishable within experimental uncertainty (from about 1.7 GPa, 780 ± 15 °C, to 30 GPa, 710 ± 20 °C, Chopin & Sobolev, 1995), leaving virtually no stability field for the talc–staurolite assemblage. Actually, the presence of lithium in natural systems will stabilise the staurolite-bearing assemblage with respect to others and extend the stability field of staurolite, the phase in which Li is preferentially incorporated. The partitioning of iron and magnesium between coexisting staurolite and garnet is another remarkable feature. In classical, relatively low-pressure metamorphic terranes, staurolite is Fe-rich but the coexisting garnet has a still higher XFe. The situation is commonly but not consistently reversed in intermediate-pressure rocks, in which staurolite, although increasingly magnesian, may be more Fe-rich than the coexisting

C. Chopin & G. Ferraris

212

garnet. Magnesiostaurolite in the Dora-Maira high-pressure rocks is consistently more Fe-rich than the coexisting garnet (Fig. 6 in Simon et al., 1997), confirming the picture obtained for staurolite proper in more Fe-rich, high-pressure rocks (Ballèvre et al., 1989; Chopin et al., 1991). This compositional dependence of the partitioning is well reproduced by the experimental study of Koch-Müller (1997), but not the apparent pressure dependence, which makes the Fe–Mg staurolite series a persisting thermodynamic nightmare (cf. Koch-Müller, 1997, with references). There are good crystal chemical reasons to this, as mentioned above, with the change in structural position of the divalent cation and order-disorder processes. In summary, the compositional features of staurolite suggestive of UHP conditions are in first place a high XMg, a high number of protons and vacancies, as well as the near absence of tetrahedral Al. Magnesiochloritoid Chloritoid, (Fe,Mg)Al2SiO4(OH)2 is structurally a layer nesosilicate (e.g. Ivaldi et al., 1988) where isolated SiO4 tetrahedra connect two different octahedral layers. The arrangement of the oxygen atoms approximates a hexagonal closest packing – at least in the (001) plane; cf. a/b  (3)½ – where tetrahedral and octahedral sites are partially occupied. This dense structure contains exclusively six-fold coordinated Al and is stable to high pressures. Whereas the Fe-rich members are common in low-grade metamorphism, Mg-rich members exclusively occur in HP rocks (cf. Chopin et al., 1992), and only near Mg endmember compositions are known as part of a stable UHP assemblage in Dora-Maira pyrope-bearing rocks (Simon et al., 1997; CD Image 9). This is in line with the experimental data on the Fe (Vidal et al., 1994) and Mg end-members (Chopin & Schreyer, 1983; Fockenberg, 1995b), showing that the latter is stable to higher P and T than the former, i.e. between 1.8 GPa and ca. 5 GPa, but limited to 750 °C by its breakdown into pyrope + corundum. Chloritoid must therefore be absent from the peak assemblages in higher-temperature UHP terranes like in the Kokchetav Massif or the Erzgebirge. In addition to this temperature indication, the stability of clinochlore + kyanite with respect to pure magnesiochloritoid + talc (no water involved) sets an upper pressure limit near 3.2 GPa for the Dora-Maira UHP rocks, whatever the fluid presence or composition (Simon et al., 1997). The limited presence of Mg in low-P chloritoid is related to the rigidity of the structure where the large octahedral sites are more suitable for Fe2+ than for Mg. The latter cation, instead, is favoured at higher P when the volume of the octahedral sites becomes smaller (Ivaldi et al., 1988).

Accessory minerals with a complex structure Beside the several accessory minerals with a simple structure mentioned above, others with a more complex structure are important for the petrology of UHP metamorphic rocks.

Mineral chemistry and mineral reactions in UHPM rocks

213

Zircon The structure of zircon, ZrSiO4, is tetragonal and is based on [100] chains of alternating Si-tetrahedra and Zr-dodecahedra which share an edge. Consequently the short distance between the high-charge cations Si4+ and Zr4+ confer to zircon a low compressibility which makes this typical accessory mineral of metamorphic rocks the best container of primary UHP minerals, especially coesite and diamond (Sobolev et al., 1994). This property is now instrumental in the mapping of the areal extent of known UHP terranes (Tabata et al., 1998; Ye et al., 2000b; Liu et al., 2002) or the identification of several new UHP terranes (e.g. Yang et al., 2001) through the systematic investigation of zircon fractions separated from country rock gneiss and eclogite lenses. Besides, experiments show that, at shock compressions exceeding 30 GPa, zircon undergoes a displacive phase transition to a CaWO4 scheelite structure type, which is known as the shockproduced mineral reidite (Glass et al., 2002). Zircon therefore also becomes instrumental in the recognition and dating of large meteorite impacts. Topaz Topaz, Al2SiO4(F,OH)2, is orthorhombic and its structure is based on [001] chains formed by alternating pairs of Al-octahedra with tetrahedra; further tetrahedra link the chains along [100]. Synthetic Al2SiO4(OH)2 is known as a high-pressure phase (P > 4 GPa, Wunder et al., 1993) but the most OH-rich natural topaz contains a maximum of 1.1 OH pfu and has been reported from the UHP unit at Hushan (Zhang et al., 2002a). The volume and IR properties along the OH–F series were studied by Wunder et al. (1999) and the crystal structure of an OH-rich topaz, Al2SiO4F1.4(OH)0.6, occurring in kyanite quartzite at Hushan (Ferrando et al., 2002) has been refined (Alberico et al., 2003a). Synthetic topaz-OH is alternative to kyanite + H2O and the OH contents of natural topaz is therefore buffered to a maximum value (at given P and T) by its coexistence with Al-silicate + H2O. In other words a given topaz OH content in this assemblage gives a univariant relation between T and P (Barton, 1982). The latter study combined with phase equilibrium data of Wunder et al. (1993) makes the OH–F topaz series amenable to thermodynamic treatment for thermobarometry. Wagnerite The magnesium phosphate wagnerite, Mg2PO4(F,OH), represents a case very similar to topaz, with the fluorine end-member having a huge stability field and the hydroxyl endmember a stability limited to high pressures (P > 0.8 GPa, Brunet et al., 1998). The structure is dense and complex, MgO4F2 octahedra and MgO4F trigonal bipyramids sharing edges or apices to form a three-dimensional network. Wagnerite is the most common Mg phosphate in nature and, with respect to polymorphs of Mg2PO4OH, is clearly stabilized by fluorine incorporation (Brunet et al., 1998). As a matter of fact, the most OHrich wagnerite known (with XOH = OH/(OH + F) near 0.6). occurs in the Dora-Maira pyrope-bearing UHP rock (Brunet et al., 1998; CD Images 11 and 12). However, wagnerite is probably a better indicator of F/OH ratio in the system than of formation pressure.

214

C. Chopin & G. Ferraris

Besides, the nature of its polytype (with five-, seven- or nine-fold superstructures along b, due to periodic faulting in the succession of the occupied F positions along b: Chopin et al., 2003b; Ren et al., 2002) is apparently more dependent on OH/F and Mg/Fe ratio than on formation conditions. Rather, the occurrence of OH-rich wagnerite + kyanite, instead of the Mg-Al phosphate lazulite, is certainly a safer indication of HP formation conditions. Titanite The titanite, CaTiOSiO4, structure consists of a corner-linked mixed octahedral–tetrahedral framework, which is also common to many phosphates, sulphates, selenates etc. Corner-sharing titanate octahedral chains extend along c and are cross-linked by silicate tetrahedra; the O atom bridging two TiO6 octahedra is underbonded, which favours its replacement by OH or F. Actually, titanite can incorporate octahedral Al via the substitution (Al,Fe3+) + (OH–,F–)  Ti4+ + O2– thus reducing the unit cell volume because VIAl (ionic radius 0.535 Å) and F– (1.33 Å) are smaller than Ti4+ (0.605 Å) and O2– (1.40 Å), respectively. The temperature dependent phase transition from space group P21/a to A2/a approaches room temperature with increasing Al content. Several findings of Al-F-rich titanite in UHP terranes suggested that this substitution toward a hypothetical more compact CaAlSiO4F end-member could be a HP index. Troitzsch & Ellis (2002) synthesised the whole series and determined its mixing parameters and the thermodynamic properties of the F end-member (cf. Tropper et al., 2002). The latter is indeed stable under UHP conditions but less stable than alternative assemblages in many rock compositions. Their petrogenetic grid shows that the Al-F content is indeed P (and T) dependent and that the highest content of the F endmember attainable in UHP rocks is near 60 mol%, in line with the highest contents found in nature. This study adequately recalls that the Al-F content is also a function of the mineral assemblage and, in the absence of a solid buffer, of the fluid composition. Besides, high Al contents in titanite may be linked to OH incorporation toward the hypothetical CaAlSiO4OH end-member (thus polymorph of vuagnatite), accounting for some low-pressure occurrences of high-Al titanite. The possibility of high-pressure replacement of Ti4+ by octahedral Si4+ in titanite was explored by Knoche et al. (1998); they could synthesise the complete series CaTiSiO5–CaSi2O5, in which the silicate end-member is stable above 8 GPa and has the titanite structure with 50% six-fold coordinated silicon. They also established the pressure dependence of VISi incorporation at silica saturation and the results did not remain of theoretical interest for long! The discovery of coesite precipitates in titanite by Ogasawara et al. (2002), in marble from the Kokchetav Massif (CD Image 10), suggests the existence of a “supersilicic” precursor titanite, thereby offering the second possible evidence for six-fold coordinated silicon in regional metamorphic rocks, after majoritic garnet. The precursor compositions reconstructed by integration of the exsolved phases (also minor calcite and apatite) point to pressures that may have exceeded 6 GPa, i.e. near 200 km burial, if one assumes that the excess Si content was linked to a CaSi2O5 compound – an assumption that is not straightforward given the non-preservation of the stoichiometry by the precipitate. In any event this finding is absolutely uncommon as,

Mineral chemistry and mineral reactions in UHPM rocks

215

paradoxically, coesite plays in this case the role of a retrograde, lower-pressure phase! This is with coesite and microdiamond a most spectacular evidence for the metamorphism of crustal material at unsuspected depths. Hydroxylclinohumite In the conclusions of their chapter “Principal mineralogic indicators of UHP in crustal rocks”, Chopin & Sobolev (1995) called the attention on possible new high-pressure minerals already known as synthetic phases, particularly the hydroxyl-bearing analogue of topaz, mentioned above, and OH-rich members of the humite series, which are known to occur also at high P–T (mantle) conditions (Thompson, 1992). Recently, a natural OH-analogue of clinohumite with more than 93% of F substituted by OH has been found in association with calcite, ferroan spinel and clinochlore and approved as the mineral species hydroxylclinohumite (Ferraris et al., 2000b). It occurs in magnesium skarns at the contact between xenoliths of dolomite marble and the metagabbro of the metamorphosed clinopyroxenite–anorthosite–gabbro Kusinsky complex in the southwestern Urals, for which no particularly high-pressure conditions are known. Therefore, although the OH end-member of clinohumite is definitely a high-pressure phase (stable at P > 1.5 GPa, Wunder, 1998), fluorine incorporation obviously stabilises clinohumite to low pressures (cf. titanian chondrodite: Engi & Lindsley, 1980). Ellenbergerite group An uncommon feature discovered in the UHP rocks of the Dora-Maira massif is the existence of a continuous solid solution series between the silicate ellenbergerite and the isostructural Mg phosphate, phosphoellenbergerite (Chopin & Sobolev, 1995; Brunet et al., 1998), the latter also discovered in a magnesite deposit near Modum, southern Norway (Raade et al., 1998). Ellenbergerite, (Mg,Ti,Zr,)2Mg6Al6Si8O28(OH)10, is a hexagonal nesosilicate (space group P63) with a dense and rigid structure (CD Image 13) based on single chains of face-sharing (Mg,Ti,Zr,)O6 octahedra and double chains of face-sharing MgO6 and AlO6 octahedra, both running parallel to c (Chopin et al., 1986; Comodi & Zanazzi, 1993; see below a comparison with magnesiodumortierite). The connection of octahedra by face-sharing instead of edge- or vertex-sharing is unusual for silicates and result in a dense structure (D = 3.15 g/cm3). Structure refinements of the Mg phosphate on material from the two localities and the synthetic end-member led to the following simplified formula for the end-member phosphate (space group P63mc): Mg14P8O30(OH)8, or (MgO3)2(Mg2O5H)6P2(OH)2P6 if one wants to stress the polyhedral arrangement in single and double chains, which ignores minor vacancies in the octahedral single chain. A more accurate formula is (Mg0.9,0.1)2Mg12P8O38H8.4 (Brunet et al., 1998). Therefore the substitution leading from the silicate to the phosphate end-member is primarily PMgSi–1Al–1 (for 6 out of 8 Si); additional mechanisms must involve protons (Brunet & Schaller, 1996), tetravalent cations, and vacancies in the octahedral single chain. This continuous series between isostructural silicate and phosphate is a unique feature in nature, even if numerous

C. Chopin & G. Ferraris

216

examples exist of isostructural silicates and phosphates (e.g. quartz–berlinite, zircon–xenotime, cerite–whitlockite, britholite–apatite). In the face-sharing octahedra of the single chain, Fe2+–Ti4+ and Fe2+–Fe3+ charge transfers along the c axis are responsible for the purple to blue colour of the silicate (CD Images 14 and 15) and the blue–green colour of the phosphate (CD Image 16; cf. Platonov et al., 2000 for an analogous mechanism in dumortierite). Experimental work (references in Brunet et al., 1998) shows both end-members to be high-pressure phases, but with very distinct stability fields. The restricted occurrence of the silicate ellenbergerite in spite of its common chemical constituents is in line with a restricted wedge-shaped stability field limited at pressures in excess of 2.7 GPa and temperatures lower than ca. 725 °C, which makes it a fine indicator of negative thermal anomaly at mantle depths. The phosphate end-member is stable above 0.85 GPa at 650 °C, above 2.25 GPa at 900 °C, and above 3 GPa at 975 °C, therefore at much lower P and higher T than the silicate, giving a barometric value to the zonations in the Si/P ratio of the natural crystals, especially in the presence of an additional Mg phosphate like wagnerite (Brunet et al., 1998). Magnesiodumortierite The borosilicate dumortierite, ideally (Al1–x,x)2Al4Al8B2Si6O36H6x, is the only silicate to bear some structural similarities with ellenbergerite, as emphasised by the above formula (actually twice the formula unit). The SiO3OH tetrahedra on the three-fold axes of ellenbergerite are replaced by planar BO3 groups (in the same way that they can be partly replaced by CO3 groups in phosphoellenbergerite); the single chain of face-sharing octahedra remains unchanged but is dominantly occupied by Al (plus some Fe, Ti or Mg); one third of the octahedral double chains also remains unchanged (with the Mg-Al pairs replaced by Al-Al) and two thirds of them are replaced by double chains in which the connection of the octahedra is achieved purely by edge-sharing (S2 slabs in Fig. 16, as opposed to the ellenbergerite-type S1 slab, both slabs alternating along [010]). In the Dora-Maira pyrope megablasts a pink dumortierite group mineral is rather common and formed at least in part near the peak pressure conditions (ca. 3 GPa). It is characterised by unusual Mg (and Ti) contents (up to 8.5 wt% MgO). The structure refinement (Ferraris et al., 1995) shows that the orthorhombic pseudohexagonal symmetry of dumortierite is retained (space group Pmcn) but that the octahedral single chain is free of Al and partly occupied by Ti and dominant Mg (just as in ellenbergerite), hence the introduction of the new species magnesiodumortierite (Chopin et al., 1995). The remaining Mg substitutes Al in the face-sharing octahedra of the ellenbergerite-type double chain. The simplified formula unit may then be written: (Mg,Ti,)(Al,Mg)2Al4BSi3O18–x(OH)x, with x between 2 and 3. In summary, all the octahedra sharing faces and normally occupied by Al in dumortierite tend to be occupied by Mg in magnesiodumortierite, as they are in ellenbergerite. The question is to what extent this feature may be related to high

Mineral chemistry and mineral reactions in UHPM rocks

217

Fig. 16. Perspective view of the magnesiodumortierite structure along [001]. The S1 slab and other features (see text) occur also in the structure of ellenbergerite, unlike the S2 slab. Independent polyhedra are differently shaded; small circles represent one of the H atoms in the structure.

pressures of formation. Admittedly, replacing high-charge cations by lower-charge ones reduces the repulsion across the shared faces and so may stabilise the structure at high pressures, which are expected to reduce the cation–cation distances. On the other hand, dumortierite with up to 3.4 wt% MgO has been reported as a relatively low-pressure secondary phase in high-grade cordierite–orthoamphibole rocks of the Bamble area, south Norway (Visser & Senior, 1991). Bearthite The structure of bearthite, Ca2Al(PO4)2OH, monoclinic, space group P21/m (Chopin et al., 1993), is that of the brackebushite group, which comprises several arsenates, vanadates and sulphates of Pb, Ba, and Cu, Mn etc., all of low-pressure and low-temperature origin. The main feature is infinite chains of edge-sharing AlO5OH octahedra extending along b. Four PO4 tetrahedra per unit cell are connected via common corners to each octahedral chain to form the fundamental building block [Al2(PO4)4(OH)2]. The same building block [Al2(SiO4)4(OH)2] is present in epidote, in the REE-Al silicate törnebohmite and their modular 1:1 combination, gatelite-(Ce) (Bonazzi et al., 2003). Since bearthite has, among the brackebushite group minerals, the smallest cation in each structural position, it could be expected to be stable to higher pressures than the other members of the group. It is actually stable at quite high pressures (at least 2.5 GPa), but also down to low pressures (Brunet & Chopin, 1995), making a fortuitous feature the

218

C. Chopin & G. Ferraris

fact that it was first discovered in HP (Monte Rosa) and UHP (Dora-Maira, CD Image 17) rocks. It has since then been reported as an accessory mineral in lower-pressure kyanite-lazulite quartzite from Sweden and Austria. Chemically, it may be pure Ca2Al(PO4)2OH, but Sr commonly substitutes part of Ca toward the Sr analogue goedkenite. The Dora-Maira material may also incorporate up to 11 wt% rare earth elements through the charge balanced substitution Ca2+ + Al3+  Mg2+ + REE3+, as is well known in allanite, but shows incipient breakdown into apatite + monazite symplectite upon decompression (Scherrer et al., 2001). These authors also report from Monte Rosa a pseudomorph of apatite + monazite + corundum (within allanite) after a HP precursor, most likely bearthite. Bearthite is therefore an important accessory mineral in HP–UHP Ca-poor rocks as it is there the main carrier of the geochemically significant elements Sr and REE. In more Ca-rich rocks, an allanitic clinozoisite may play the same role (Nagasaki & Enami, 1998).

Conclusions Undoubtedly the discovery of UHP metamorphism not only has opened a new era for petrology, and geology in general, but has given a tremendous impetus to experimental mineralogy in its broadest sense and to the petrological applications of crystal chemistry and sophisticated solid state methods. Among several indispensable aspects which are needed to modern petrology (not to mention the thermodynamic ones), this chapter particularly stressed structural, crystal chemical and microtextural topics. However, one should also bear in mind that, beside minerals taken alone as indexes, mineral assemblages do have a more considerable potential in terms of petrological constraints. As an example, the assemblages reported by Okay (2002) or by Simon & Chopin (2001) point to unusually low temperatures at near-UHP conditions or to uncommon fluid compositions, respectively, which would not be revealed by the classical UHP indexes. Several index phases of UHP metamorphism do not show new structural types but only a specific crystal chemistry due to isomorphous substitutions favoured at high pressure, like the following ones: small cation  large cation, e.g. Mg  Fe2+ and Si  Al. OH  F, a substitution which favours compression because O–H...O hydrogen bonds allow a wide variability both in configuration (angle) and distances. low-charge  high-charge cations; which favours the stability of a structure (Pauling rule) and, in particular, the compression along chains of face- or edgesharing polyhedra. Consequently we know that, e.g., an Mg-rich or/and OH-rich member of a solid solution is likely to be stable at higher P conditions than Fe or F members. In case of polymorphism, the attributes of the HP phase are, even in the absence of knowledge of the relevant stability fields, (i) higher density and (ii) higher coordination number of at least one cation. If their stability field is known, the preservation of the HP phase by some elastically suitable (hard) minerals, like garnet, diamond and zircon, makes them invaluable petrological indicators of minimum pressures, regardless of their abundance or whether they are major or accessory rock-forming minerals. The ever-

Mineral chemistry and mineral reactions in UHPM rocks

219

increasing spatial resolution of characterisation techniques also hints at one of the limits to be encountered. More and more polymorphs having a UHP thermodynamic stability field will be identified at the submicrometre scale, down to “crystals” having dimensions of a few unit cells only; thereby the limit is reached under which surface energies may become preponderant over intrinsic thermodynamic properties, casting some doubt as to the actual P–T formation conditions. The recent findings of submicrometre-size clinoenstatite in Dabie–Su-Lu (Zhang et al., 2002b) and of nanometre-size TiO2 with the
-PbO2 structure in Erzgebirge (Hwang et al., 2000) make this concern quite actual. Exsolutions (in principle an isochemical process, as opposed to the looser term precipitates) are another important source of information to the petrologist because they indicate the instability of a solid solution under changing P–T conditions. However, the detection of “heterogeneity” within a matrix is just a first step. In fact, discrimination between true exsolutions, zone growing, trapped phases and similar phenomena is often not straightforward and requires a skilled capability of interpretation based on several aspects, like stability fields, crystal growth, surface phenomena, and solid state phenomena which often are related to a modular (polysomatism, polytypism, twinning) nature of the crystal structures. For the latter aspect see the proceedings of the first EMU School (Merlino, 1997) and Ferraris et al. (2004). Clearly, electron transmission microscopy and various kinds of chemical microprobe have become indispensable tools for the UHP petrologist.

Acknowledgements G.F. is grateful to Gabriella Ivaldi, Angela Gula, and Cristiano Ferraris for figures, to MIUR (Roma) for COFIN and FIRB financial support. The help of Hans-Joachim Massonne, Hans-Peter Schertl and Esther Schmädicke through donation of photographs and rock samples to C.C. is highly appreciated.

References Akaogi, M. & Akimoto, S. (1977): Pyroxene–garnet solid-solution equilibria in the systems Mg4Si4O12–Mg3Al2Si3O12 and Fe4Si4O12–Fe3Al2Si3O12 at high pressures and temperatures. Phys. Earth Planet. Inter., 15:90–106. Alberico, A., Ferrando, S., Ivaldi, G. & Ferraris, G. (2003a): X-ray single-crystal structure refinement of an OH-rich topaz from Sulu UHP terrane (Eastern China) – Structural foundation of the correlation between cell parameters and fluorine content. Eur J. Mineral., 15:875–991. Alberico, A., Ferraris, G., Ivaldi, G. & Săbău, G. (2003b): Geikielite as microinclusion in Foltea (South Carpathians) pyropes. 5th EMU School on UHPM, Abstracts /Acta Mineral. Petrogr., Abstr. Ser., 3/: 3. Andersen, T.B. (1984): Inclusion patterns in zoned garnets from Magerøy. Mineral. Mag., 48:21–26. Angel, R.J., Chopelas, A. & Ross, N.L. (1992): Stability of high-density clinoenstatite at upper mantle pressures. Nature, 358:322–324. Arlt, T., Bermejo, M., Blanco, M.A., Gerward, L., Jiang, J.Z., Olsen, J.S. & Recio, J.M. (2000): High-pressure polymorphs of anatase TiO2. Phys. Rev. B, 61:14,414–14,419. Ballèvre, M., Pinardon, J., Kiénast, J.-R. & Vuichard, J.-P. (1989): Reversal of Fe-Mg partitioning between garnet and staurolite in eclogite-facies metapelites from the Champtoceaux Nappe (Brittany, France). J. Petrol., 30:1321–1349. Barton, M.D. (1982): The thermodynamic properties of topaz solid solutions and some petrologic applications. Am. Mineral., 67:956–974.

220

C. Chopin & G. Ferraris

Becker, H. & Altherr, R. (1992): Evidence from ultra-high-pressure marbles for recycling of sediments into the mantle. Nature, 358:745–748. Belluso, E., Castelli, D., Ferraris, G. & Săbău, G. (1998): TEM identification of microinclusions in pyrope from the Foltea garnet-peridotite (South Carpathians). Plinius, 20:43–44. Beyssac, O. & Chopin C. (2003): Comment on “Diamond, former coesite and supersilicic garnet in metasedimentary rocks from the Greek Rhodope: a new ultrahigh-pressure metamorphic province established” by Mposkos and Kostopoulos. Earth Planet. Sci. Lett., 214:669–674. Beyssac, O., Rouzaud, J.-N., Goffé, B., Brunet, F. & Chopin, C. (2002): Graphitization in a high-pressure metamorphic gradient: a micro-Raman and HRTEM study. Contrib. Mineral. Petrol., 143:19–31. Bindi, L., Safanov, O.G., Litvin, Y.A., Perchuk, L.L. & Menchetti, S. (2002): Ultrahigh potassium content in the clinopyroxene structure: an X-ray single-crystal study. Eur. J. Mineral., 14:929–934. Bindi, L., Safanov, O.G., Yapaskurt, V.O., Perchuk, L.L. & Menchetti, S. (2003): Ultrapotassic clinopyroxene from the Kumdy-Kol microdiamond mine, Kokchetav Complex, Kazakhstan: Occurrence, composition and crystal-chemical characterization. Am. Mineral., 88:464–468. Bonazzi, P., Bindi, L. & Parodi, G. (2003). Gatelite-(Ce), a new REE-bearing mineral from Trimouns, French Pyrenees: Crystal structure and polysomatic relationships with epidote and törnebohmite-(Ce). Am. Mineral., 88:223–228. Bozhilov, K.N., Green, H.W. & Dobrzhinetskaya, L.F. (2003): Quantitative 3D measurement of ilmenite abundance in Alpe Arami olivine by confocal microscopy: Confirmation of high-pressure origin. Am. Mineral., 88:596–603. Brueckner, H.K., Carswell, D.A. & Griffin, W.L. (2002): Paleozoic diamonds within a Precambrian peridotite lens in UHP gneisses of the Norwegian Caledonides. Earth Planet. Sci. Lett., 203:805–816. Brunet, F. & Chopin, C. (1995): Bearthite, Ca2Al(PO4)2OH: synthesis, stability and thermodynamic properties. Contrib. Mineral. Petrol., 121:258–266. Brunet, F. & Lecocq, D. (1999): Phosphorus incorporation in garnet: natural and experimental data. Ber. Dtsch. Mineral. Ges., 11, Beih. 1:43. Brunet, F. & Schaller, T (1996): Protons in the magnesium phosphates phosphoellenbergerite and holtedahlite: An IR and NMR study. Am. Mineral., 81:385–394. Brunet, F., Chopin, C. & Seifert, F. (1998): Phase relations in the MgO–P2O5–H2O system and the stability of phosphoellenbergerite. Petrological applications. Contrib. Mineral. Petrol., 131:54–70. Bruno, M., Compagnoni, R., Hirajima, T. & Rubbo, M. (2002): Jadeite with the Ca-Eskola molecule from an ultra-high pressure metagranodiorite, Dora-Maira Massif, Western Alps. Contrib. Mineral. Petrol., 142:515–519. Burton, K.W. (1986): Garnet–quartz intergrowths in graphitic pelites: the role of the fluid phase. Mineral. Mag., 50:611–620. Carlson, W.D. & Rosenfeld, J.L. (1981): Optical determination of topotactic aragonite-calcite growth kinetics: metamorphic implications. J. Geol., 89:915–638. Carswell, D.A. & Cuthbert, S.J. (2003): Ultrahigh pressure metamorphism in the Western Gneiss Region of Norway. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 51–73. Carswell, D.A. & Zhang, R.Y. (1999): Petrographic characteristics and metamorphic evolution of ultrahighpressure eclogites in plate-collision belts. Int. Geol. Rev., 41:781–798. Chopin, C. (1984): Coesite and pure pyrope in high-grade blueschists of the western Alps: a first record and some consequences. Contrib. Mineral. Petrol., 86:107–118. Chopin, C. (2003): Ultrahigh-pressure metamorphism: tracing continental crust into the mantle. Earth Planet. Sci. Lett., 212:1–14. Chopin, C. & Schreyer, W. (1983): Magnesiocarpholite and magnesiochloritoid: two index minerals of pelitic blueschists and their preliminary phase relations in the model system MgO-Al2O3-SiO2-H2O. Am. J. Sci., 283A:72–96. Chopin, C. & Sobolev N.V. (1995): Principal mineralogic indicators of UHP in crustal rocks. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 96–131. Chopin, C., Klaska, R., Medenbach, O. & Dron, D. (1986): Ellenbergerite, a new high-pressure Mg-Al-(Ti, Zr)-silicate with a novel structure based on face-sharing octahedra. Contrib. Mineral. Petrol., 92:316–321.

Mineral chemistry and mineral reactions in UHPM rocks

221

Chopin, C., Henry, C. & Michard, A. (1991): Geology and petrology of the coesite-bearing terrain, Dora-Maira massif, Western Alps. Eur. J. Mineral., 3:263–291. Chopin, C., Seidel, E., Theye, T., Ferraris, G., Ivaldi, G. & Catti, M. (1992): Magnesiochloritoid, and the FeMg series in the chloritoid group. Eur. J. Mineral. 4:67–76. Chopin, C., Brunet, F., Gebert, W., Medenbach, O. & Tillmanns, E. (1993): Bearthite, Ca2Al[PO4]2(OH), a new mineral from high-pressure terranes of the western Alps. Schweiz. Mineral. Petrogr. Mitt., 73:1–9. Chopin, C., Ferraris, G., Ivaldi, G., Schertl, H.-P., Schreyer, W., Compagnoni, R., Davidson, C. & Davis, A.M. (1995): Magnesiodumortierite, a new mineral from very-high-pressure rocks (western Alps). II. Crystal chemistry and petrological significance. Eur. J. Mineral., 7:525–535. Chopin, C., Goffé, B., Ungaretti, L. & Oberti, R. (2003a): Magnesiostaurolite and zincostaurolite: mineral description with a petrogenetic and crystal-chemical update. Eur. J. Mineral., 15:167–176. Chopin, C., Armbruster, T. & Leyx, C. (2003b): Polytypism in wagnerite, Mg2PO4(F,OH). Joint EUG-EGSAGU assembly, Nice, April 2003. Geophys. Res. Abstr., 5:8323. Coggon, R. & Holland, T.J.B. (2002): Mixing properties of phengitic micas and revised garnet-phengite thermobarometers. J. Metamorph. Geol., 20:683–696. Comodi, P. & Zanazzi, P.F. (1993): Structural study of ellenbergerite. Part I: effects of high temperatures. Part II: effects of high pressure. Eur. J. Mineral., 5:819–829, 831–838. Dobrzhinetskaya, L.F., Green, H.W. & Wang, S. (1996): Alpe Arami: A peridotite massif from depths of more than 300 kilometers. Science, 271:1841–1845. Dobrzhinetskaya, L.F., Green H.W., Weschler, M., Darus, M., Wang, Y.C., Massonne, H.-J. & Stöckhert, B. (2003): Focused ion beam techniques and transmission electron microscope studies of microdiamonds from the Saxonian Erzgebirge, Germany. Earth Planet. Sci. Lett., 210:399–410. Dutrow, B., Holdaway, M.J. & Hinton, R.W. (1986): Lithium in staurolite and its petrological significance. Contrib. Mineral. Petrol., 94:496–506. El Goresy, A., Chen, M., Gillet, P., Dubrovinsky, L., Graup, G. & Ahuja, R. (2001): A natural shock-induced dense polymorph of rutile with
-PbO2 structure in the suevite from the Ries crater in Germany. Earth Planet. Sci. Lett., 192:485–495. Enami, M. & Zang, Q. (1988): Magnesian staurolite in garnet-corundum rocks and eclogite from the Donghai district, Jiangsu province, east China. Am. Mineral., 73:48–56. Enami, M., Cong, B., Yoshida, T. & Kawabe, I. (1995): A mechanism for Na incorporation in garnet – an example from garnet in orthogneiss from the Su-Lu terrane, eastern China. Am. Mineral., 80:475–482. Engi, M. & Lindsley, D.H. (1980): Stability of titanian clinohumite; experiments and thermodynamic analysis. Contrib. Mineral. Petrol., 72:415–424. Ferrando, S., Frezzotti, M.L., Alberico, A. & Compagnoni, R. (2002): Petrologic and fluid inclusion study of an OH-rich topaz-bearing kyanite quartzite from Sulu UHP Terrane, Eastern China. EMPG IX, Zurich 24–27 April 2002., J. Conf. Abstr., 7:32. Ferraris, G. (2002). Mineral and inorganic crystals. In Giacovazzo, G. (ed.): Fundamentals of crystallography. Oxford: Oxford Univ. Press, 503–584. Ferraris, G. & Ivaldi, G. (2002): Structural features of micas. In Mottana, A., Sassi, F.P, Thompson, J.B. & Guggenheim, S. (eds.) Micas: Crystal chemistry and metamorphic petrology /Rev. Mineral. Geochem., 46/. Washington (D.C): Mineral. Soc. Am., 117–153. Ferraris, G., Ivaldi, G. & Chopin, C. (1995): Magnesiodumortierite, a new mineral from very-high-pressure rocks (western Alps). I. Crystal structure. Eur. J. Mineral., 7:167–174. Ferraris, C., Chopin, C. & Wessicken, R. (2000a): Nano- to microscale-decompression products in ultrahighpressure phengite: HRTEM and AEM study, and some petrological implications. Am. Mineral., 85:1195–1201. Ferraris, G., Prencipe, M., Sokolova, E.V., Gekimyants, V.M. & Spiridonov, E.M. (2000b): Hydroxylclinohumite, a new member of the humite group: Twinning, crystal structure and crystal chemistry of the clinohumite subgroup. Z. Kristallogr., 215:169–173. Ferraris, C., Castelli, D. & Lombardo, B. (2002): A new type of nano-exsolution in UHP phengite from the Dora-Maira massif, Western Alps - Italy. EMPG IX, Zürich 24–27 April 2002. J. Conf. Abstr., 7:33. Ferraris, G., Makovicky, E. & Merlino, S. (2004): Crystallography of modular materials. Oxford: Oxford Univ. Press, in press.

222

C. Chopin & G. Ferraris

Fiquet, G. (2001): Mineral phases of the Earth’s mantle. Z. Kristallogr., 216:248–271. Fockenberg, T. (1995a): Synthesis and chemical variability of Mg-staurolite in the system MgOAl2O3–SiO2–H2O as a function of water pressure. Eur. J. Mineral., 7:1373–1381. Fockenberg, T. (1995b): New experimental results up to 100 kbar in the system MgO-Al2O3–SiO2–H2O (MASH): Preliminary stability fields of chlorite, chloritoid, staurolite, MgMgAl-pumpellyite, and pyrope. Bochumer Geol. Geotech. Arb., 44:39–44. Fockenberg, T. (1998): An experimental investigation on the P-T stability of Mg-staurolite in the system MgOAl2O3–SiO2–H2O. Contrib. Mineral. Petrol., 130:187–198. Fockenberg, T. & Schreyer, W. (1997): Synthesis and chemistry of unusual excess-Si aluminous enstatite in the system MgO–Al2O3–SiO2 (MAS). Eur. J. Mineral., 8:1293–1299. Gasparik, T. & Lindsley, D.H. (1980): Phase equilibria at high pressure of pyroxenes containing monovalent and trivalent ions. In Prewitt, C.T. (ed.): Pyroxenes /Rev. Mineral., 7/. Washington (D.C): Mineral. Soc. Am., 309–339. Gayk, T., Kleinschrodt, Langosch, A. & Seidel, E. (1995): Quartz exsolution in clinopyroxene of high-pressure granulite from the Münchberg Massif. Eur. J. Mineral., 7:1217–1220. Gil Ibarguchi, J.I., Mendia, M. & Girardeau, J. (1991): Mg- and Cr-rich staurolite and Cr-rich kyanite in high pressure ultrabasic rocks (Cabo Ortegal, northwestern Spain). Am. Mineral., 76:501–511. Gillet, Ph., Ingrin, J. & Chopin, C. (1984): Coesite in subducted continental crust: P-T history deduced from an elastic model. Earth Planet. Sci. Lett., 70:426–436. Glass, B.P., Liu, S. & Leavens, P. (2002): Reidite: An impact-produced high-pressure polymorph of zircon found in marine sediments. Am. Mineral., 87:562–565. Grevel, K.-D., Navrotsky, A., Fockenberg, T. & Majzlan, J. (2002): The enthalpy of formation and internally consistent thermodynamic data of Mg-staurolite. Am. Mineral., 87:397–404. Grew, E.S. & Sandiford, M. (1984): A staurolite–talc assemblage in tourmaline-phlogopite-chlorite schist from northern Victoria Land, Antarctica, and its petrogenetic significance. Contrib. Mineral. Petrol., 87:337–350. Hacker, B.R., Sharp, T., Zhang, R.Y., Liou, J.G. & Hervig, R.L. (1997): Determining the origin of ultra-highpressure lherzolites. Science, 278:702–704. Haggerty, S.E. & Sautter, V. (1990): Ultradeep (greater than 300 kilometers), ultramafic upper mantle xenoliths. Science, 248:993–996. Haggerty, S.E, Fung, A.T. & Burt, D.M. (1994): Apatite, phosphorus and titanium in eclogitic garnet from the upper mantle. Geophys. Res. Lett., 21:1699–1702. Harlow, G.E. (1997): K in clinopyroxene at high pressure and temperature: an experimental study. Am. Mineral., 82:259–269. Hawthorne, F.C., Ungaretti, L., Oberti, R., Caucia, F. & Callegari, A. (1993): The crystal-chemistry of staurolites. I: Crystal structure and site populations. II: Order-disorder and the monoclinic – orthorhombic phase transition. III: Local order and chemical composition. Can. Mineral., 31:551–582, 583–596, 597–616. Hutchison, M.T., Hursthouse M.B. & Light M.E. (2001): Mineral inclusions in diamond: associations and chemical distinctions around the 670–km discontinuity. Contrib. Mineral. Petrol., 142:119–126. Hwang, S.L., Shen, P.Y., Chu, H.T. & Yui, T.F. (2000): Nanometer-size
-PbO2-type TiO2 in garnet: A thermobarometer for ultrahigh-pressure metamorphism. Science, 288:321–324. Hwang, S.L., Shen, P., Chu, H.T., Yui, T.F. & Lin, C.C. (2001): Genesis of microdiamonds from melt and associated multiphase inclusions in garnet of ultrahigh-pressure gneiss from Erzgebirge, Germany. Earth Planet. Sci. Lett., 188:9–15. Ivaldi, G., Catti, G. & Ferraris, G. (1988): Crystal structure at 25 and 700°C of magnesiochloritoid from a highpressure assemblage (Monte Rosa). Am. Mineral., 73:358–364. Ivaldi G., Ferraris G. & Curetti N. (2001): Crystal structure paths to a phengite geobarometer? J. Conf. Abstr., 6, 540. Izraeli, E.S., Harris, J.W. & Navon, O. (1999): Raman barometry of diamond formation. Earth Planet. Sci. Lett., 173:351–360. Jung, I. & Schreyer, W. (2002): Synthesis, properties and stability of end member boromuscovite, KAl2[BSi3O10](OH)2. Contrib. Mineral. Petrol., 143:684–693. Knoche, R. Angel, R., Seifert, F. & Fliervoet, T. (1998): Complete substitution of Si for Ti in titanite Ca(Ti1–xSix)SiO5. Am. Mineral., 83:1168–1175.

Mineral chemistry and mineral reactions in UHPM rocks

223

Koch-Müller, M. (1997): Experimentally determined Fe-Mg exchange between synthetic staurolite and garnet in the system MgO-FeO-Al2O3–SiO2–H2O. Lithos, 41:185–212. Lager, G.A., Downs, R.T., Origlieri, M. & Garoutte, R. (2002): High-pressure single-crystal X-ray diffraction – study of katoite hydrogarnet: Evidence for a phase transition from Ia3d  I43d symmetry at 5 GPa. Am. Mineral., 87:642–647. Lattard, D. & Schreyer, W. (1983): Synthesis and stability of the garnet calderite in the system Fe-Mn-Si-O. Contrib. Mineral. Petrol., 84:199–214. Leech, M.L. & Ernst, W.G. (1998): Graphite pseudomorphs after diamond? A carbon isotope and spectroscopic study of graphite cuboids from the Maksyutov Complex, south Ural Mountains, Russia. Geochim. Cosmochim. Acta, 62:2143–2154. Leinenweber, K., Utsumi, W., Tsuchida, Y., Yagi, T. & Kunita, K. (1991): Unquenchable high-pressure perovskite polymorphs of MnSnO3 and FeTiO3. Phys. Chem. Miner., 18:244–250. Liati, A., Gebauer, D. & Wysoczanski, R. (2002): U-Pb SHRIMP-dating of zircon domains from UHP garnetrich mafic rocks and late pegmatoids in the Rhodope zone (N Greece); evidence for Early Cretaceous crystallization and Late Cretaceous metamorphism. Chem. Geol., 184:281–299. Liou, J.G. & Zhang, R.Y. (1996): Occurrences of intergranular coesite in ultrahigh-P rocks from the Su-Lu region, eastern China: implications for lack of fluid during exhumation. Am. Mineral., 81:1217–1221. Liou, J.G., Zhang, R.Y., Ernst, W.G., Rumble, D. & Maruyama, S. (1998): High-pressure minerals from deeply subducted metamorphic rocks. Rev. Mineral., 37:33–96. Liu, F.L., Xu, Z., Liou, J.G., Katayama, I., Masago, H., Maruyama, S. & Yang, J. (2002): Ultrahigh-pressure mineral inclusions in zircons from gneissic core samples of the Chinese Continental Scientific Drilling Site in eastern China. Eur. J. Mineral., 14:499–512. Lonker, S.W. (1983): The hydroxyl content of staurolite. Contrib. Mineral. Petrol., 84:36–42. Massonne, H.-J. (1995): Experimental and petrogenetic study of UHPM. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 33–95. Massonne, H.-J. & O'Brien, P.J. (2003): The Bohemian Massif and the NW Himalaya. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 145–187. Massonne, H.-J. & Schreyer, W. (1986): High-pressure synthesis and X-ray properties of white micas in the system K2O-MgO-Al2O3-SiO2-H2O. Neues Jahrb. Mineral. Abh., 153:177–215. Massonne, H.-J. & Schreyer, W. (1989): Stability field of the high-pressure assemblage talc + phengite and two new phengite barometers. Eur. J. Mineral., 1:391–410. Massonne, H.-J., Bernhardt, H.-J., Dettmar, D., Kessler, E., Medenbach, O. & Westphal, T. (1998): Simple identification and quantification of microdiamonds in rock thin-sections. Eur. J. Mineral., 10:497–504. Merlino, S. (ed.) (1997): Modular aspects of minerals /EMU Notes Mineral., 1/. Budapest: Eötvös Univ. Press, 448 p. Morten, L. & Trommsdorff, V. (2003): Metamorphism and textures of dry and hydrous garnet peridotites. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 443–466. Mosenfelder, J.L. & Bohlen, S.R. (1997): Kinetics of coesite to quartz transformation. Earth Planet. Sci. Lett., 153:133–147. Mposkos, E.D. & Kostopoulos, D.K. (2001): Diamond, former coesite and supersilicic garnet in metasedimentary rocks from the Greek Rhodope: a new ultrahigh-pressure metamorphic province established. Earth Planet. Sci. Lett., 192:497–506. Nagasaki, A. & Enami, M. (1998): Sr-bearing epidote in ultra high-pressure (UHP) metamorphic rocks from the Sulu province, eastern China: An important Sr-reservoir under UHP conditions. Am. Mineral., 83:240–247. Nasdala, L. & Massonne, H.J. (2000): Microdiamonds from the Saxonian Erzgebirge, Germany: in situ microRaman characterization. Eur. J. Mineral., 12:495–498. Navon, O. (1991): High internal pressures in diamond fluid inclusions determined by infrared-absorption. Nature, 353:746–748. Navrotsky, A. (1998): Thermodynamics of high pressure phases. In Hemley, R. & Mao, D. (eds.): Ultrahighpressure mineralogy /Rev. Mineral., 37/. Washington (D.C.): Mineral. Soc. Am., 319–341.

224

C. Chopin & G. Ferraris

Nespolo, M. & Ďurovič, S. (2002): Crystallographic basis of polytypism and twinning in micas. In Mottana, A., Sassi, F.P, Thompson, J.B. & Guggenheim, S. (eds.) Micas: Crystal chemistry and metamorphic petrology /Rev. Mineral. Geochem., 46/. Washington (D.C): Mineral. Soc. Am., 155–279. Ogasawara, Y., Fukasawa, K. & Maruyama, S. (2002): Coesite exsolution from supersilicic titanite in UHP marble from the Kokchetav Massif, northern Kazakhstan. Am. Mineral., 87:454–461. Okay, A.I. (2002): Jadeite-chloritoid-glaucophane-lawsonite blueschists in northwest Turkey: unusually high P/T ratios in continental crust. J. Metamorph. Geol., 20:757–768. Parkinson, C.D. (2000): Coesite inclusions and prograde compositional zonation of garnet in whiteschist of the HP-UHPM Kokchetav massif, Kazakhstan: a record of progressive UHP metamorphism. Lithos, 52:215–233. Parkinson, C.D. & Katayama, I. (1999): Present-day ultrahigh-pressure conditions of coesite inclusions in zircon and garnet: Evidence from laser Raman microspectroscopy. Geology, 27:979–982. Parra, T., Vidal, O. & Agard, P. (2002): A thermodynamic model for Fe-Mg dioctahedral K white micas using data from phase-equilibrium experiments and natural pelitic assemblages. Contrib. Mineral. Petrol., 143:706–732. Pavese, A., Ferraris, G., Prencipe, M. & Ibberson, R. (1997): Cation site ordering in phengite 3T from the DoraMaira massif (western Alps): a variable-temperature neutron powder diffraction study. Eur. J. Mineral., 9:1183–1190. Pavese, A., Ferraris, G., Pischedda, V. & Ibberson, R. (1999a): Tetrahedral order in phengite 2M1 upon heating, from powder neutron diffraction, and thermodynamic consequences. Eur. J. Mineral., 11:309–320. Pavese A., Ferraris G., Pischedda V. & Mezouar M. (1999b): Synchrotron powder diffraction study of phengite 3T from the Dora-Maira massif: p-V-T equation of state and petrological consequences. Phys. Chem. Miner., 26:460–467. Pavese, A., Ferraris, G., Pischedda, V. & Radaelli, P. (2000): Further study of the cation ordering in phengite 3T by neutron powder diffraction. Mineral. Mag., 64:11–18. Pavese, A., Ferraris, G., Pischedda, V. & Fauth, F. (2001): M1–site occupancy in 3T and 2M1 phengites by low temperature neutron powder diffraction: reality or artefact? Eur. J. Mineral., 13:1071–1078. Pavese, A., Curetti, N., Ferraris, G., Ivaldi, G. & Russo, U. (2003): Deprotonation and order-disorder reactions as a function of temperature in a phengite 3T (Cima Pal, western Alps) by neutron diffraction and Mössbauer spectroscopy. Eur. J. Mineral., 15:357–363. Perchuk, L.L., Safonov, O.G., Yapaskurt, V.O., Barton, J.M. & Jr. (2002): Crystal-melt equilibria involving potassium-bearing clinopyroxene as indicator of mantle-derived ultrahigh-potassic liquids: an analytical review. Lithos, 60:89–111. Platonov, A.N., Langer, K., Chopin, C., Andrut, M. & Taran, M.N. (2000): Fe2+-Ti4+ charge-transfer in dumortierite. Eur. J. Mineral., 12, 521–528. Prewitt, C.T. & Downs, R.T. (1998): High-pressure crystal chemistry. In Hemley, R. & Mao, D. (eds.): Ultrahigh-pressure mineralogy /Rev. Mineral., 37/. Washington (D.C.): Mineral. Soc. Am., 283–317. Raade, G., Romming, C. & Medenbach, O. (1998): Carbonate-substituted phosphoellenbergerite from Modum, Norway; description and crystal structure. Mineral. Petrol., 62:89–101. Ravna, E.J.K. & Paquin, J. (2003): Thermobarometric methodologies applicable to eclogites and garnet ultrabasites. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 229–259. Reinecke, T. (1987): Manganoan deerite and calderitic garnet from high-pressure metamorphic Fe-Mn-rich quartzites on Andros Island, Greece. Mineral. Mag., 51:247–251. Ren, L., Grew, E.S., Xiong, M. & Ma, Z. (2002): A new polymorph of wagnerite, Mg2(PO4)(F,OH). Geol. Soc. Am. Meeting, Denver, abstr. No. 234–8. Rice, A.H.N. & Mitchell, J.I. (1991): Porphyroblast textural sector-zoning and matrix displacement. Mineral. Mag., 55:379–396. Richet, P. & Gillet, P. (1997): Pressure-induced amorphization of minerals: a review. Eur. J. Mineral., 9:907–933. Rieder, M. et al. (1998): Nomenclature of the micas. Can. Mineral., 36:905–912. Risold, A.-C., Trommsdorff, V. & Grobety, B. (2001): Genesis of ilmenite rods and palisades along humitetype defects in olivine from Alpe Arami. Contrib. Mineral. Petrol., 140:619–628.

Mineral chemistry and mineral reactions in UHPM rocks

225

Risold, A.-C., Trommsdorff, V. & Grobéty, B. (2003): Morphology of oriented ilmenite inclusions in olivine from garnet peridotites (Central Alps, Switzerland). Eur. J. Mineral., 15:289–294. Rosenfeld, J.L. (1969): Stress effects around quartz inclusions in almandine and the piezothermometry of coexisting aluminium silicates. Am. J. Sci., 267:317–351. Rosenfeld, J.L. & Chase, A.B. (1961): Pressure and temperature of crystallisation from elastic effects around solid inclusions in minerals. Am. J. Sci., 259:519–541. Rubatto, D. & Hermann, J. (2001): Exhumation as fast as subduction? Geology, 29:3–6. Sassi, F.P., Guidotti, C., Rieder, M. & De Pieri, R. (1994): On the occurrence of metamorphic 2M1 phengites: some thoughts on polytypism and crystallization condition of 3T phengites. Eur. J. Mineral., 6:151–160. Scherrer, N.C., Gnos, E. & Chopin, C. (2001): Retrograde monazite-forming reactions in bearthite-bearing high-pressure rocks. Schweiz. Mineral. Petrogr. Mitt., 81:369–378. Schmädicke, E. & Müller, W.F. (2000): Unusual exsolution phenomena in omphacite and partial replacement of phengite by phlogopite + kyanite in an eclogite from the Erzgebirge. Contrib. Mineral. Petrol., 139:629–642. Schrauder, M. & Navon, O. (1993): Solid carbon-dioxide in a natural diamond. Nature, 365:42–44. Schreyer, W. & Seifert, F. (1969): High-pressure phases in the system MgO-Al2O3-SiO2-H2O. Am. J. Sci., 267–A:407–443. Shannon, R.D. (1976): Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr., A32, 751– 767. Shatsky, V.S. & Sobolev, N.V. (2003): The Kokchetav massif of Kazakhstan. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 75–103. Shatsky, V.S., Sobolev, N.V. & Vavilov, M.A. (1995): Diamond-bearing metamorphic rocks of the Kokchetav massif (northern Kazakhstan). In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 427–455. Simon, G. & Chopin, C. (2001): Enstatite–sapphirine crack-related assemblages in ultrahigh-pressure pyrope megablasts, Dora-Maira massif, western Alps. Contrib. Mineral. Petrol., 140:422–440. Simon, G., Chopin, C. & Schenk, V. (1997): Near-end-member magnesiochloritoid in prograde-zoned pyrope, Dora-Maira massif, western Alps. Lithos, 41:37–57. Simon, G., Chopin, C. & Czank, M. (1998): Magnesiostaurolite in UHP rocks: a HP phase with some uncommon features. In 17th IMA General Meeting, Toronto, Abstract Vol., 100. Smith, D.C. (1984): Coesite in clinopyroxene in the Caledonides and its implications for geodynamics. Nature, 310:641–644. Smith, D.C. (1988): A review of the peculiar mineralogy of the Norwegian coesite-eclogite province, with crystal-chemical, petrological, geochemical and geodynamical notes and an extensive bibliography. In Smith, D.C. (ed.) Eclogites and eclogite-facies rocks. Amsterdam: Elsevier, 1–206. Sobolev, N.V. & Shatsky, V.S. (1990): Diamond inclusions in garnet from metamorphic rocks: a new environment for diamond formation. Nature, 343:742–746. Sobolev, N.V., Shatsky, V.S., Vavilov, S.V. & Goryainov, S.V. (1994): Zircon from ultra high pressure metamorphic rocks of folded regions as an unique container of inclusions of diamond, coesite and coexisting minerals. Dokl. Ross. Akad. Nauk, 334:488–492. Sobolev, N.V., Fursenko, B.A., Goryainov, S.V., Shu, J., Hemley R.J., Mao, H.K. & Boyd, F.R. (2000): Fossilized high pressure from the Earth’s deep interior: The coesite-in-diamond barometer. Proc. Natl. Acad. Sci. USA, 97:11875–11879. Stachel, T., Brey, G. & Harris, J.W. (2000a): Kankan diamonds (Guinea) I: From the lithosphere down to the transition zone. Contrib. Mineral. Petrol., 140:1–15. Stachel, T., Harris, J.W., Brey, G. & Joswig, W. (2000b): Kankan diamonds (Guinea) II: Lower mantle inclusion parageneses. Contrib. Mineral. Petrol., 140:16–27. Stöckhert, B., Duyster, J., Trepmann, C. & Massonne, H.-J. (2001): Microdiamond daughter crystals precipitated from supercritical COH plus silicate fluids included in garnet, Erzgebirge, Germany. Geology, 29:391–394. Tabata, H., Yamauchi, K. & Maruyama, S. (1998): Tracing the extent of a UHP metamorphic terrane: mineralinclusion study of zircons in gneisses from the Dabie Shan. In Hacker, B. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 261–273. Thompson, A.B. (1992): Water in the Earth upper mantle. Nature, 358:295–302.

226

C. Chopin & G. Ferraris

Thompson, R.N. (1975): Is upper-mantle phosphorus contained in sodic garnet? Earth Planet. Sci. Lett., 26:417–424. Troitzsch, U. & Ellis, D.J. (2002): Thermodynamic properties and stability of AlF-bearing titanite CaTiOSiO4–CaAlFSiO4. Contr. Mineral. Petrol., 142:543–563. Tropper, P., Manning, C.E. & Essene, E.J.. (2002): The substitution of Al and F in titanite at high pressure and temperature: Experimental constraints on phase relations and solid solution properties. J. Petrol., 43:1787–1814. van Roermund, H.L.M. & Drury, M.R. (1998): Ultra-high pressure (P > 6 GPa) garnet peridotites in Western Norway: exhumation of mantle rocks from > 185 km depth. Terra Nova, 10:295–301. van Roermund, H.L.M., Drury, M.R., Barnhoorn, A. & De Ronde, A. (2000): Super-silicic garnet microstructures from an orogenic garnet peridotite, evidence for an ultra-deep (> 6 GPa) origin. J. Metamorph. Geol., 18:135–147. van Roermund, H.L.M., Drury, M.R., Barnhoorn, A. & De Ronde, A. (2001): Relict majoritic garnet microstructures from ultra-deep orogenic peridotites in Western Norway. J. Petrol., 42:117–130. van Roermund, H.L.M., Carswell, D.A., Drury, M.R. & Heijboer, T.C. (2002): Micro-diamonds in a megacrystic garnet-websterite pod from Bardane on the island of Fjørtoft, western Norway. Geology, 30:959–962. Vidal, O., Theye, Th. & Chopin, C. (1994): Experimental study of chloritoid stability at high pressure and various fO2 conditions. Contrib. Mineral. Petrol., 118:256–270. Visser, D. & Senior, A. (1991): Mg-rich dumortierite in orthoamphibole-bearing rocks from the high-grade Bamble Sector, south Norway. Mineral. Mag., 55:563–577. Vrána S. & Frýda, J. (2003): Ultrahigh-pressure grossular-rich garnetite from the Moldanubian Zone, Czech Republic. Eur. J. Mineral., 15:43–54. Wang, L., Essene, E.J. & Zhang, Y. (1999): Mineral inclusions in pyrope crystals from Garnet Ridge, Arizona, USA: implications for processes in the upper mantle. Contrib. Mineral. Petrol., 135:164–178. Wei, C. & Powell, R. (2003): Phase relations in high-pressure metapelites in the system KFMASH (K2O–FeO–MgO–Al2O3–SiO2–H2O) with application to natural rocks. Contrib. Mineral. Petrol., 145: 301–315. Wendt, A.S., D’Arco, P., Goffé, B. & Oberhänsli, R. (1993): Radial cracks around
-quartz inclusions in almandine: constraints on the metamorphic history of the Oman mountains. Earth Planet. Sci. Lett., 114:449–461. Wirth, R. & Rocholl, A. (2003): Nanocrystalline diamond from the Earth’s mantle underneath Hawaii. Earth Planet. Sci. Lett., 211: 357–370. Withers, A.C., Essene, E.J. & Zhang Y. (2003): Rutile/TiO2II phase equilibria. Contrib. Mineral. Petrol., 145: 199–204. Woodland, A.B. & Angel, R.J. (1997): Reversal of the orthoferrosilite – high-P clinoferrosilite transition, a phase diagram for FeSiO3 ad implications for the mineralogy of the Earth’s upper mantle. Eur. J. Mineral., 9:245–254. Woodland, A.B. & O’Neill, H.St.C. (1993): Synthesis and stability of Fe2+3Fe3+2Si3O12 garnet and phase relations with Fe3Al2Si3O12–Fe2+3Fe3+2Si3O12 solutions. Am. Mineral., 78:1002–1015. Woodland, A.B. & O’Neill, H.St.C. (1995): Phase relations between Ca3Fe3+2Si3O12–Fe2+3Fe3+2Si3O12 garnet and CaFeSi2O6–Fe2Si2O6 pyroxene solid solutions. Contrib. Mineral. Petrol., 121:87–98. Wu, X., Han, Y., Meng, D. & Li, D. (2002): Discovery and implication of P21/n crystal structure on a nanoscale in single jadeite crystals. Earth Planet. Sci. Lett., 197:165–169. Wunder, B. (1998): Equilibrium experiments in the system MgO-SiO2-H2O (MSH): stability fields of clinohumite-OH [Mg9Si4O16(OH)2], chondrodite-OH [Mg5Si2O8(OH)2] and phase A (Mg7Si2O8(OH)6. Contrib. Mineral. Petrol., 132:111–120. Wunder, B., Rubie, D.C., Ross, C.R., Medenbach, O., Seifert, F. & Schreyer, W. (1993): Synthesis, stability and properties of Al2SiO4(OH)2: a fully hydrated analogue of topaz. Am. Mineral., 78:285–297. Wunder, B., Andrut, M. & Wirth, R. (1999): High-pressure synthesis and properties of OH-rich topaz. Eur. J. Mineral., 11:803–813. Yang, J., Xu, Z., Song, S., Zhang, J., Wu, C., Shi, R., Li, H. & Brunel, M. (2001): Discovery of coesite in the North Qaidam Early Palaeozoic ultrahigh pressure (UHP) metamorphic belt, NW China. C. R. Acad. Sci. Paris, 333:719–724.

Mineral chemistry and mineral reactions in UHPM rocks

227

Ye, K., Yao, Y.P., Katayama, I., Cong, B., Wang, Q. & Maruyama, S. (2000a): Large areal extent of ultrahighpressure metamorphism in the Sulu ultrahigh-pressure terrane of East China: new implications from coesite and omphacite inclusions in zircon of granitic gneiss. Lithos, 52:157–164. Ye, K., Cong, B. & Ye, D. (2000b): The possible subduction of continental material to depths greater than 200 km. Nature, 407:734–736. Ye, K., Liou, J.G., Cong, B. & Maruyama, S. (2001): Overpressures induced by coesite-quartz transition in zircon. Am. Mineral., 86:1151–1155. Zhang, R.Y. & Liou, J.G. (2000): Exsolution lamellae in minerals from ultrahigh-pressure rocks. In Ernst, W.G. & Liou, J.G. (eds.): Ultrahigh-pressure metamorphism and geodynamics in collision-type orogenic belts /Int. Book Ser., 4/. Boulder (Co.): Geol. Soc. Am. & Columbia (Md.): Bellwether Publ. Ltd., 216–228. Zhang, J., Li, B., Utsumi, W. & Liebermann, R.C. (1996): In situ X-ray observations of the coesite-stishovite transition: reversed phase boundary and kinetics. Phys. Chem. Miner., 23, 1–10. Zhang, R.Y., Liou, J.G. & Shu, J.F. (2002a): Hydroxyl-rich topaz in high-pressure and ultrahigh-pressure kyanite quartzites, with retrograde woodhouseite, from the Sulu terrane, eastern China. Am. Mineral., 87:445–453. Zhang, R.Y., Shau, Y.H., Liou, J.G. & Lo, C.H. (2002b): Discovery of clinoenstatite in garnet pyroxenites from the Dabie-Sulu ultrahigh-pressure terrane, east-central China. Am. Mineral., 87:867–874.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 8, 229–259

Thermobarometric methodologies applicable to eclogites and garnet ultrabasites ERLING J. KROGH RAVNA1* and JENS PAQUIN2 1

2

Department of Geology, University of Tromsø, Norway Mineralogisches Institut, Universität Heidelberg, Heidelberg, Germany; * e-mail: [email protected]

Introduction Thermobarometry of HP/UHP rocks is of vital importance for the understanding of tectonic and rock forming processes and large-scale vertical transport of matter at great depths within the Earth. In recent years new sets of experiments, the extraction of internally consistent thermodynamic data set, together with elaborated composition– activity models for critical minerals have taken the art of estimation of metamorphic conditions several important steps forward. In this contribution we would like to present methods we regard as important for basic and ultrabasic compositions, and address problems of vital interest in modern geothermobarometric evaluations of HP/UHP rocks. Carswell & Harley (1990) gave a comprehensive review on this task, and some of their fundamentals will not be reproduced here. For a more general review of geothermobarometry, the reader is recommended Chapter 15 of Spear (1993). Mineral assemblages of interest The essential mineral assemblage of eclogites (s. s.) is garnet + omphacite + quartz. Additional phases may be many, and the silicates phengite, amphibole, kyanite and zoisite/clinozoisite are fairly common. Rutile is apparently present in most eclogites, while Al- and F-rich titanite has been described from rather few localities. Dolomite and calcite (after primary aragonite) are not uncommon. The high-P SiO2 polymorph coesite has been found as preserved relics or as distinctive polycrystalline quartz polymorphs included in primary phases as garnet, omphacite or zircon. The principal mineral assemblage of UHP ultrabasites consists of garnet, diopside, enstatite and forsterite. Depending on both the P–T conditions and the presence of minor chemical components such as K, Ti, F and Cl additional phases like phlogopite, Krichterite, Ti-clinohumite, Mg-chlorite and antigorite are stable under HP/UHP conditions. In most cases these accessory phases are not of great interest for common geothermobarometric calculations. We will therefore not discuss the occurrence and implications of these minerals with respect to P–T determinations. The interested reader is referred to Poli & Schmidt (2002) and references therein.

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

230

E.J. Krogh Ravna & J. Paquin

Practical geothermometers and geobarometers for HP/UHP rocks Practical geothermometers are based on chemical reactions with low
V and high
S and
H. Such reactions are strongly temperature dependent, and have steep slopes in a P–T diagram. Likewise, good geobarometers can be found among mineral reactions with high
V and low to moderate
S and
H. Such reactions are highly dependent on variations in pressure and less dependent on temperature, and thus have very gentle slopes in a P–T diagram (Fig. 1). However, as most available geothermometers are, to a certain degree, dependent on pressures, and geobarometers dependent on temperatures, we commonly need a combination of at least one geothermometer and one geobarometer to constrain metamorphic conditions properly. The most frequently used thermometers for garnet-bearing basic and ultrabasic rocks can be subdivided into three categories:  Fe2+–Mg exchange thermometers between garnet and either clinopyroxene, orthopyroxene, olivine or phengite,  two-pyroxene thermometers based on the pyroxene solvus,  trace element thermometers. Practical geobarometers for HP/UHP rocks are all based on net transfer reactions between complex solid solutions. A key point is the distribution of Al between VI and IV coordination, the former being favoured by increasing pressure. Only simple equations for the most widely used thermometers and barometers will be given. For the more complicated expressions the reader is recommended to consult the original papers. Furthermore, we have not discussed the use of more elaborated thermodynamics-based methods as exemplified by THERMOCALC (Powell & Holland, 1988, 1994) or TWQ (Berman, 1991). Some cautionary notes before proceeding In calculating temperature and pressure for HP/UHP rocks such as eclogites and garnet peridotites, the careful study of the textural relationships is inevitable. At temperatures around 650–700 °C the intracrystalline diffusion of Fe2+ and Mg in most minerals except garnet is fast enough for homogenisation. Above this temperature garnet will also homogenise. One relevant assumption for calculation of peak metamorphic pressures and temperatures in many HP/UHP rocks is then that core compositions of coexisting minerals reflect the metastable peak metamorphic conditions and therefore indicate equilibrium. This presumption is valid for many orogenic UHP eclogites and garnet peridotites, and xenoliths, whose minerals show evidence for a rapid exhumation history, which is visible only in narrow zoning at the outermost rims combined with extensive homogeneous plateaus in the core region of the minerals. Most HP/UHP rocks also show intensive retrograde overprinting combined with deformation as well as generation of new minerals, which, in the case of garnet ultrabasites often are of the same phases as the primary minerals. Therefore it is important for calculating peak metamorphic pressures and temperatures in UHP rocks only to use the first generations of minerals, which in most cases occur as porphyroclasts that are in contact with each other. The timing of maximum pressure and maximum temperature in HP/UHP terranes may also

Thermobarometric methodologies

231

Fig. 1. Orientations in P–T space of a suitable geothermometer and geobarometer. The intersection between two such equilibrium curves for a single sample is commonly used to constrain the P–T conditions. In this example the system garnet–orthopyroxene is shown. The Al content of orthopyroxene coexisting with garnet is mainly a function of pressure and thus can be used as a geobarometer, while the distribution of Fe and Mg between these two phases is mainly a function of temperature and act as a geothermometer.

greatly diverge, and this can easily result in wrong combinations of mineral analyses when different thermometers and barometers are combined. This problem has also been addressed by e.g. Carswell et al. (1997, 2000). HP and UHP metamorphic rocks commonly have a complex history, often including rapid subduction followed by accelerated exhumation. Therefore it is important to be aware the problems that can arise in disequilibrium, as well as equilibrium partitioning among different elements in minerals due to significantly different diffusivities (e.g. Paquin & Altherr, 2001a). To detect such disequilibrium partitioning, detailed microanalytical (EPMA) profiling across mineral grains is needed to evaluate both inter- and intraphase partitioning. Also the fact that P–T determinations are mostly related to cation exchange equilibria, continued cation diffusion among coexisting minerals during the retrograde history must be taken into account. The degree of thermal resetting is commonly not readily identified. Moreover the different sizes of the relevant mineral grains and cutting effects of grains can also significantly influence the P–T results. In the following paragraphs we want to present the most important geothermobarometers, discuss some pitfalls in calculating metamorphic pressures and temperatures, and in addition discuss some new aspects, offering the best possible results.

232

E.J. Krogh Ravna & J. Paquin

Geothermometers The most frequently used thermometers for garnet-bearing basic and ultrabasic rocks can be subdivided into three categories:  Fe2+–Mg exchange thermometers between garnet and either clinopyroxene, orthopyroxene, olivine or phengite,  two-pyroxene thermometers based on the pyroxene solvus,  trace element thermometers. Geothermometers based on Fe2+–Mg exchange between garnet and other Fe–Mg minerals The following mineral pairs have been subject to special interest as potent geothermometers based on the strongly temperature dependent exchange of Fe2+ and Mg between garnet and common coexisting phases in HP/UHP rocks:  garnet–clinopyroxene,  garnet–orthopyroxene,  garnet–olivine,  garnet–hornblende,  garnet–phengite,  garnet–biotite. Garnet–biotite Biotite is a fairly uncommon phase in HP and UHP rocks, but has been described from low-T eclogites from e.g. the Western Gneiss Region (WGR) of Norway (Krogh, 1980), and from several Fe-Ti garnet peridotites and “orthopyroxene eclogites” from the WGR (Carswell et al., 1983, 1995). The status of biotite in the latter rocks has, however, been disputed. In pelitic systems, biotite is apparently unstable at high pressures, phengite being the stable mica. Thus, the garnet–biotite system has no important relevance to HP/UHP metamorphism. Garnet–hornblende The garnet–hornblende Fe–Mg system has been empirically calibrated (Graham & Powell, 1984; Perchuk et al., 1985; Powell, 1985; Ravna, 2000b). None of these calibrations have, however, proved to be successful for amphibole-bearing eclogites, and will not be discussed further here. Garnet–phengite The garnet–phengite Fe–Mg geothermometer was first proposed by Krogh & Råheim (1978), based on a minimum of experimental data from Råheim & Green (1974). Green & Hellman (1982) presented experimental results for this reaction for different bulk compositions, showing the dependence of this equilibrium on P, T and X. For basaltic systems with mg#  67 their expression is

Thermobarometric methodologies

T GH 82 [C] 

where KD 

233

5170  360 GPa  273 , ln K D  4.17

(Fe/Mg) Grt (CD Image 1). This thermometer has been used with generally (Fe/Mg) Phe

not some success, but is recommended due to serious uncertainties linked to the possible presence of Fe3+ in phengite, which is not readily detected by routine analyses. Recently Coggon & Holland (2002) presented a new linearised expression for this reaction, but this method is not evaluated here. Garnet–clinopyroxene This is one of the most widely used geothermometers for high-grade metamorphic basites and ultrabasites, due to the common appearance of garnet and clinopyroxene in such rocks. The first attempt to calibrate this system quantitatively was empirically done by Mysen & Heier (1972). However, Råheim & Green (1974) did the first experimental calibration of the system, expressing the relationship between temperature, pressure and the distribution coefficient grt – cpx K D(Fe/Mg) 

(Fe/Mg)grt . (Fe/Mg)cpx

Later on, many different versions of this geothermometer have been proposed (e.g. Ellis & Green, 1979; Powell, 1985; Krogh, 1988; Pattison & Newton, 1989; Ai, 1994; grt and the last three also include Ravna, 2000a). All of these contain corrections for XCa grt correction factors for XMg. Most of these versions apparently give reliable and comparable results, at least for systems containing garnet with intermediate Ca contents. For rocks with low-Ca garnets (e.g. garnet ultrabasites) or high-Ca garnets the deviation between the different calibrations are greater due to different correction factors for the non-ideality of Ca in garnet solid solution (see e.g. Carswell et al., 1997). While the expressions of Ellis & Green (1979) and Powell (1985) have rectilinear corrections for grt grt XCa , Krogh (1988) demonstrated a curvilinear relationship between lnKD and XCa , at least grt in the compositional range XCa = 0.10–0.50. Based on a variety of experimental data, Krogh (1988) derived the following geothermometric expression T K 88 [C] 

grt grt 2 1879  6731X Ca  6173( X Ca )  100 P [GPa ]  273. ln KD  1.393

Pattison & Newton (1989) presented a large set of self-consistent data on the Fe–Mg equilibria between garnet and clinopyroxene. Multiple regression of their data yield the geothermometric expression

E.J. Krogh Ravna & J. Paquin

234

T PN 89 [C] 

grt grt 2 grt grt 2 561  3395 X Ca  2388( X Ca )  9781X Mg  31026( X Mg )

ln KD+0.512 +

grt 3 26217( X Mg )  103.7 P [GPa]

ln KD+0.512

ln K  0.512

+

 273.

This expression reproduces the experimental data of Pattison & Newton (1989) well within ± 40 °C, but commonly yields unrealistic low temperatures on natural rocks. Berman et al. (1995) used the Pattison & Newton (1989) dataset to refine the thermometer through thermodynamic analysis, and application of this reformulation to a number of amphibolite to granulite facies terrains returned temperatures between 70 to 200 °C above those obtained with the original Pattison & Newton (1989) geothermometer, and compatible with independent temperature estimates. Ai (1994) combined new experimental data on Mg-rich systems with the Pattison & Newton (1989) data set, as well as with data from 15 other sources, altogether 271 pairs of garnet–clinopyroxene. Multiple regression of the data returned the expression T Ai 94 [C] 

grt grt 2 grt 1987  3648.55 X Ca  1629( X Ca )  659 X Mg  176.6 P [GPa]

ln K D+0.512 1.076

 273.

This thermometer apparently works well for most compositions, especially for systems with low-Ca/high-Mg garnet, as the calibration includes lots of experiments on ultrabasic systems. Ravna (2000a) used an extended database of 311 garnet–clinopyroxene pairs from 27 experimental sources, excluding the large self-consistent data set of Pattison & Newton (1989). He also included 49 garnet–clinopyroxene pairs from natural Mn-rich grt , and gave the expression samples to retrieve empirical correction factors for XMn T R 00 [C] 

grt grt 2 grt grt 2 1939  3270 X Ca  1396( X Ca )  3319 X Mn  3535( X Mn )  ln KD  1.223 grt grt 2 grt 3 1105 X Mg  3561( X Mg )  2324( X Mg )  169.4 P [GPa]

ln KD  1.223

 273.

This calibration (CD Image 2) includes a very wide range of compositions of both garnet and clinopyroxene, and it is expected to work for most compositional ranges covered by natural rocks. A comparison between different calibrations of the thermometer as functions of chemical variations of garnet was also shown (Ravna, 2000a). The effect of additional components. Commonly, uncertainties related to the garnet– clinopyroxene geothermometer are cited as ± 30 to 50 °C. The use of most of the abovecited expressions will for most samples produce temperatures within these limits, and they are all subject to the crucial limitation due to the ferric–ferrous iron problem (see discussion below). Moreover, none of the calibrations contain any correction for XJdcpx. Koons (1984) proposed that for more Jd-rich clinopyroxenes (XJdcpx > 0.7) e.g. in metagranitoids, KD appears to vary inversely with jadeite content. Koons (1984)

Thermobarometric methodologies

235

attributed this to non-ideality in Fe2+MgNa–1Al–1 substitution resulting in preferential ordering of Fe2+ into M2 sites in the clinopyroxene. In a geothermometric study of UHP eclogites from the Su-Lu terrane, Hirajima (1996) showed a distinct negative correlation between lnKD and XJd in the range 0.45 < XJd < 0.75, resulting in a corresponding positive correlation between calculated T and XJd. In this study calculated T shows a scatter of ca. 300 °C at one outcrop. This could apparently not be ascribed to estimation error of Fe3+/Fetot alone. He concluded that we cannot estimate the equilibrated T of UHP rocks using the common garnet–clinopyroxene geothermometers without the introduction of a new correction factor for XJd in clinopyroxene. An independent indication can be extracted from the experimental data of Schmidt (1993). At 650 °C and pressures of 20–26 kbar two pyroxenes – omphacite with about Jd50Aug50 and jadeitic pyroxene ranging from Jd65Aug35 to Jd78Aug22 – coexisted with garnet in tonalitic compositions. There is a tendency of positive correlation between XNa and Fe/(Fe + Mg) in clinopyroxene from these experiments, but not as pronounced as that shown by Hirajima (1996). Ravna (2000a) showed that for 0 < XJdcpx < 0.50 there was no obvious influence of XJd on the KD values. All these evidence points toward a certain influence of XJd on the KD value at higher (> Jd55?) XJd values, and thus on calculated temperatures. Garnet–olivine The advantage of this extensively used experimental calibration (O’Neill & Wood, 1979, corrected by O’Neill, 1980) (TOW79), which has an intrinsic calibration uncertainty of approximately ± 60 °C, lies in the simple solid solution model of olivine. The thermometric expression (CD Image 3) can be retrieved from the original papers of O’Neill & Wood (1979) and O’Neill (1980). However, although only garnet incorporates significant amounts of Fe3+, an enormous discrepancy of up to 200 °C (Canil & O’Neill, 1996) can result in calculated temperature estimates depending on whether the Fe3+ content in garnet is considered or disregarded. When Fe3+ content in garnets is ignored, the TOW79 is in best agreement with the results of the widely used two-pyroxene thermometer calibration of Brey & Köhler (1990) (TBK290), therefore Canil & O’Neill (1996) recommend ignoring possible Fe3+ contents in garnets. The faster diffusion of Fe and Mg in olivine implies rapid adjustment to changing P–T conditions compared to the Fe-Mg diffusion garnet (Ganguly & Tazzoli, 1994; Dimanov & Sautter, 2000). This implies that the Fe-Mg core compositions of olivine are in most cases not in equilibrium with the Fe-Mg core compositions in garnet. The question remains how it is possible to retain reliable T estimates. Brenker & Brey (1997) found that for olivine-rich bulk compositions (> 75 modal %) changing P–T conditions result in large changes in the Fe/Mg ratio in garnet, but only in negligible changes in the Fe/Mg ratio in olivine. This is caused by the fact that the Fe/Mg ratio in olivine reflects the Fe/Mg ratio of the bulk composition as olivine is the dominant mineral phase in the peridotitic systems and therefore acts as a buffer. Taylor (1998) found highly variable T estimates for his experimental series at T > 1150 °C with Tcalc–Texpt  200–300 °C. Moreover, Nimis & Trommsdorff (2001) suggested that the small
G0 of the Fe–Mg reaction between garnet and olivine (Ganguly

E.J. Krogh Ravna & J. Paquin

236

& Saxena, 1987) result in large uncertainties of the TOW79 formulation. However, Brey & Köhler (1990) found that the combination of TOW79 and the Al-in-orthopyroxene barometer (Brey & Köhler, 1990) (PBK90) demonstrates good reproducibility of their experiments and that the thermometer can be applied to natural peridotitic rocks in its present form. Garnet–orthopyroxene The first formulation of Mori & Green (1978) is based on their experimental phase equilibria for natural garnet lherzolites. But this thermometer version did not include a pressure correction term and hence is not of great importance. Harley (1984a) has experimentally investigated the partitioning of Fe and Mg in both the FMAS and CFMAS systems between garnet and aluminous orthopyroxene. The thermometric calibration (TH84) is expressed as TH 84 [C] 

grt 3740  1400 X grs  22.86 P

RT ln KD  1.96

 273,

grt with KD = (Fe/Mg)grt/(Fe/Mg)opx, Xgrs = Ca/(Ca + Mg + Fe) and P in kbar. Brey & Köhler (1990) noted that the TH84 slightly overestimates at low and underestimates at high temperatures, whereby the best agreement is at ~ 1000 °C for his CFMAS experiments. A new version of this thermometer (Lee & Ganguly, 1988) based on new experiments in the FMAS system yields overestimates of more than 125 °C compared to the experimental temperatures by Brey & Köhler (1990). Better results of the Lee & Ganguly (1988) thermometer version will be achieved when the correction term for Ca and Mn in garnet is neglected (Brey & Köhler, 1990). Brey & Köhler (1990) also formulated a thermometer expression based on Fe–Mg exchange equilibria for six different mineral pairs in their experiments. Their garnet–orthopyroxene thermometer is expressed as

TBK190 [C] 

1456  9.86 P  273, ln KD  0.55

with KD = (Fe/Mg)grt/(Fe/Mg)opx and P in kbar. This thermometer version gives, compared to the Harley (1984a) version, slightly higher temperatures at high T and slightly lower temperatures at low T. A general problem is that the slope of the KD lines in a P–T diagram are quite flat, resulting in a relative large error if combined with the Al-in-orthopyroxene barometer of Brey & Köhler (1990) (Fig. 1, CD Image 4). Fe-Mg geothermometry and the problem of Fe3+ The T estimates based on Fe–Mg exchange thermometry between garnet, olivine, orthopyroxene and clinopyroxene are generally not entirely reliable due to uncertainties in the Fe3+/Fetot ratios of both natural phases and the phases in the original laboratory experiments from which the thermobarometers were calibrated (Canil & O’Neill, 1996).

Thermobarometric methodologies

237

It is now well known that the experiments used to calibrate these Fe–Mg exchange thermometers produced garnet, orthopyroxene and clinopyroxene with substantial Fe3+ contents, similar to that of natural samples (e.g. Brey & Köhler, 1990; Canil & O’Neill, 1996; Smith, 1999). In natural garnet peridotites the Fe3+/Fetot ratios increase in the order olivine (~ 0.00), garnet (~ 0.03), orthopyroxene (~ 0.03–0.10) and clinopyroxene (~ 0.2–0.4), and the Fe3+ content in garnet increases significantly with increasing P and T. In common eclogites the Fe3+/Fetot ratio in garnet is expected to be low, as shown by Mössbauer studies of garnet from UHP eclogites from Dabie Shan showing a Fe3+/Fetot ratio of only 0.06 (reported by Carswell et al., 2000). This is also supported by microXANES studies (Schmid et al., 2003), where Fe3+/Fetot ratios of garnets from the same area were in the range 0.0–0.03. The inaccuracy of the Fe–Mg exchange geothermometry cannot, however, be quantified as the exact redox conditions of the original experiments are not well known (Luth & Canil, 1993). For ultrabasic systems it is recommended to treat all Fe as Fe2+ as this approximation has shown to yield the best agreement between TBK290 (as a Fe3+ independent thermometer) and various Fe–Mg thermometers. However, it is highly desirable to measure the Fe3+ contents in the mineral phases to be aware of the problems of Fe3+ in applying Fe–Mg exchange thermometry and to test to what extent the calculated temperatures are shifted towards higher or lower temperatures. Various methods have been proposed to treat this problem. A common method is to use charge balance criteria to calculate the Fe3+/Fetot ratio of minerals. For clinopyroxene 4 cations and 6 oxygens may be used. Charge balance can then be obtained by oxidising an adequate portion of Fe2+ to Fe3+ (Ryburn et al., 1976; Neumann, 1976; Droop, 1987). However, we now know that under HP/UHP conditions clinopyroxene appear to be non-stoichiometric due to the presence of the Ca-Eskola molecule Ca0.5AlSi2O6. Other attempts include calculation of Fe3+ equal Na excess over Al + Cr. Similar calculations have been proposed for amphiboles (e.g. Schuhmacher, 1991) and phengite (e.g. Schliestedt, 1980). All such calculations are, however, very sensitive to the quality of the chemical analyses of the mineral, even in the case of high-quality WDS electron microprobe analyses (Carswell & Zhang, 1999). For relatively Fe-rich clinopyroxenes this does not appear to be a very serious problem, but for Fe-poor systems an unreliable spread in calculated Fe3+/Fetot ratios is obtained. In the case of non-stoichiometry, such calculations may even appear meaningless. Sobolev et al. (1999) concluded in a detailed Mössbauer study of Fe3+ in coexisting garnet and clinopyroxene from diamondiferous eclogite from the Udachnaya kimberlite and garnet peridotite from the Mir kimberlite that values of Fe3+/Fetot calculated from EMP analyses generally are inaccurate, although they do not greatly affect temperature estimates in eclogites due to compensation effects between garnet and clinopyroxene. According to Sobolev et al. (1999), the precision in determination of SiO2 largely controls the Fe3+/Fetot values based on stoichiometry due to its abundance and high ionic charge. In addition, high-Na clinopyroxenes (> 4–5 wt% Na2O) are even more sensitive to EMP errors, and therefore can produce larger uncertainties in temperature estimates. This lead the authors to question the correctness of previously published KD values and temperatures based on any geothermometers involving Fe2+–Mg distribution.

238

E.J. Krogh Ravna & J. Paquin

Calculated the Fe3+/Fetot ratios for omphacites in 12 UHP samples from Dabie Shan varied between 0 and 0.5 Carswell et al. (1997), while Tabata et al. (1998) gave a range of 0.16–0.82 for omphacites from the same region. Mössbauer studies of the Fe3+/Fetot ratio in omphacite separates from two Dabie Shan UHP eclogite samples indicated that roughly 50% of the Fe is present as Fe3+ (reported by Carswell et al., 2000). Schmid et al. (2003) determined ferric iron in the UHP minerals garnet, omphacite and phengite by micro-XANES. Using samples of eclogites from Dabie Shan they showed that omphacite and phengite both have elevated Fe3+/Fetot ratios, while coexisting garnet had a low Fe3+/Fetot ratio. Charge balance calculations showed no Fe3+ in omphacite, while the Fe3+/Fetot ratio detected by micro-XANES was 0.25–0.30. In garnet, on the other hand, no Fe3+ was detected, while charge balance calculations gave Fe3+/Fetot = 0.027–0.054. Schmid et al. (2003) concluded that the contents of ferric iron calculated by charge balance – at least in the case of low-Fe omphacite and phengite – obviously have little bearing for P–T estimates. To retrieve confident results, one needs to determine the Fe3+ content for every specific mineral assemblage used in thermobarometry. Carswell & Zhang (1999) have highlighted the seriousness of this problem, showing a strong negative correlation between calculated Fe3+/Fetot ratios in omphacite and temperatures calculated from Fe2+–Mg partitioning between garnet and omphacite (see Fig. 2). A possible regional T gradient across the Dabie Shan UHP terrane (Wang et al., 1992) will thus be masked by the uncertainties introduced by calculation of the Fe3+/Fetot ratio in clinopyroxene. Carswell & Zhang (1999) suggested that error brackets of at least ± 100 °C should be attached to individual garnet–clinopyroxene Fe2+–Mg estimates simply because of uncertainties regarding the proportions of Fe3+ and Fe2+ present in the omphacites. As an independent test, a series of omphacite analyses (n = 39) from a single thin section from sample DAB 9872 from Dabie Shan (from the database of Schmid, 2001) was combined with a garnet containing the highest Mg/Fe ratio. Based on charge balance calculations the Fe3+/Fetot ratio in the omphacite vary from 0–0.546 (mean value 0.064), and the corresponding temperatures vary from 914–646 °C (calculated at 4.0 GPa), with a mean of 867 ± 58 °C. If Fe3+ is assumed to equal Na – (Altot + Cr), the Fe3+/Fetot ratio lies in the range 0–0.489 (mean value 0.178) and the corresponding temperatures between 914–673 °C, with a mean of 824 ± 65 °C. The negative correlation between estimated Fe3+/Fetot and calculated T (Fig. 1) fall in the middle of the apparent “regional” trend for Dabie Shan UHP eclogites, supporting the conclusions of Carswell & Zhang (1999). Schmid (2001) analysed the Fe2+ content of omphacite in the same sample by standard titration methods, and calculated a Fe3+/Fetot ratio of 0.35. Using this ratio, average temperature calculated for this sample using the same garnet composition and all 39 omphacite analyses as above, results in 758 ± 12 °C. Carswell et al. (1997) adopted a Fe3+/Fetot ratio of 0.5 for omphacites from UHP eclogites in Dabie Shan, which in this case results in 690 ± 10 °C. The results from another sample of Schmid by using micro-XANES yielded a Fe3+/Fetot ratio for omphacite of 0.25–0.30. Choosing the lower of these limits results in temperature estimates of 798 ± 13 °C. An independent estimate of 794 ± 65 °C at 3.70 ± 0.32 GPa (Ravna & Terry, 2003) for a garnet-clinopyroxene-phengite-kyanite-coesite eclogite from the same area is obtained from sample DB48 of Carswell et al. (1997). These highly diverging results should

Thermobarometric methodologies

239

Fig. 2. A plot showing the effect of calculation of Fe3+/Fetot in omphacite on estimated T (using the garnet–clinopyroxene thermometer of Ravna, 2000a) for sample DAB 9872 (from the database of Schmid, 2001). Also plotted is the trend shown by Carswell & Zhang (1999) of various samples across the Dabie Shan UHP belt, based on data from Carswell et al. (1997) and Tabata et al. (1998).

remind us about extreme caution when applying the garnet–clinopyroxene Fe–Mg thermometer without separate estimations of Fe2+ and Fe3+ in clinopyroxene. All the problems discussed above for clinopyroxene can be addressed to amphibole and phengite as well. The estimation of Fe3+/Fetot in these minerals would face similar uncertainties, and temperature estimates obtained with these methods should thus be treated as highly uncertain. Solvus thermometry Two-pyroxene thermometers Studies of the mutual solubility of clinopyroxene coexisting with orthopyroxene have shown that the transfer of enstatite and diopside components between coexisting pyroxenes are strongly temperature controlled and hence provide the basis of a thermometer (e.g. Davis & Boyd, 1966; Boyd, 1973; Wood & Banno, 1973; Wells, 1977; Bertrand & Mercier, 1985; Brey & Köhler, 1990; Nimis & Taylor, 2000). The twopyroxene solvus in a simple CMS system is controlled by a single reaction that expresses the Ca–Mg exchange in the M2 sites between the two pyroxenes, and is written as CaMgSi2O6 + Mg2Si2O6  Mg2Si2O6 + CaMgSi2O6 Di in opx

En in cpx

En in opx

Di in cpx

At least 13 different calibrations of this thermometer exist, of which we only present the most widely used versions. The Wells (1977) formulation, which is expressed as TW77 [C] 

7341  273, opx 3.355  2.44 X Fe  ln K

E.J. Krogh Ravna & J. Paquin

240

with K = a(Mg2Si2O6)cpx / a(Mg2Si2O6)opx and XFeopx = Fe2+/(Fe2++Mg2+), reproduces the experiments of Brey et al. (1990) very well, but there is an increasing underestimation at higher temperatures presumably as no pressure term is incorporated in the equation. The thermometer version of Bertrand & Mercier (1985) generally underestimates temperatures although these authors include a correction term for the minor chemical components Na and Fe in their equation (Brey & Köhler, 1990). The thermometer is expressed as TBM 90 [C] 

36273  399 P  273, (19.31  8.314 ln K  12.15Ca *cpx ) 2

M2 M2 )opx, Ca*cpx = XCaM2/(1 – XNa )cpx + where K = (1 – Ca*cpx)/(1 – Ca*opx), Ca*opx = XCaM2/(1 – XNa –3 (–0.77 + 10 T)[Fe/(Fe + Mg)] and P is in kbar. Brey & Köhler (1990) attribute the underestimation of the Bertrand & Mercier (1985) calibration to the Fe correction of Ca in clinopyroxene, which is mainly based on experiments that were carried out in a temperature range too narrow to yield a universally applicable correction factor. The Brey & Köhler (1990) experiments are best reproduced by the Bertrand & Mercier (1985) thermometer if a correction factor of –0.97 is used instead of –0.77. Brey & Köhler (1990) calibrated a new version of this two-pyroxene thermometer (TBK290) based on reversed experiments and modified the correction term for Fe. Their equation is written as

TBK290 [C] 

cpx 23664  (24.9  aX Fe )P  273, 2 opx 13.38  (ln K )  bX Fe

with K = (1 – Ca*)cpx/(1 – Ca*)opx; Ca* = CaM2/(1 – NaM2), XFeopx = Fe/(Fe + Mg) and P in kbar. An evaluation of older versions of the two-pyroxene thermometer is given by Carswell & Gibb (1987). However, is it important to note that pyroxene solvus relationships at low temperatures are not well constrained due to the steepening of the solvus limbs at successively lower temperatures. Therefore it is not recommended to use the two-pyroxene thermometers for equilibration temperatures 900 °C to yield the best possible results. The recent two-pyroxene calibration of Taylor (1998) for lherzolites and websterites has so far not been tested on many natural peridotites and will therefore not be considered in this review. Brey & Köhler (1990) formulated a Ca-in-orthopyroxene thermometer which considers only the Ca content in orthopyroxene. Their equation is as follows: TBK390 [C] 

6425  26.4 P  273,  ln Ca opx  1.843

with P in kbar. However, caution should be exercised as the Ca content in orthopyroxene is lowered in the presence of Al (Brey & Köhler, 1990), and may also be lowered through

Thermobarometric methodologies

241

the presence of Na in M2 sites in natural systems to counterbalance the incorporation of Fe. Nevertheless, the Ca-in-orthopyroxene thermometer provides insights into the closure temperature of the Ca–Mg exchange, which rules pyroxene solvus relations and hence can be used to evaluate equilibrium or disequilibrium conditions. Trace element thermometers The Ni-in-garnet thermometer Further information about equilibration temperatures can be obtained by applying the Ni-in-garnet thermometer that considers the Ni partitioning in garnet in equilibrium with olivine. Empirical (Griffin et al., 1989; Ryan et al., 1996) as well as experimental studies (Canil, 1994, 1999) have shown that the Ni content in garnet is strongly temperature dependent with Ni contents increasing with increasing temperature, whereas the Ni content in mantle-derived olivine is more or less constant at 2900 ± 360 ppm (± standard deviation) (Ryan et al., 1996). Griffin et al. (1996) suggested that the diffusion of Ni in garnet is about the same as (or a bit faster than) Fe and Mg, so the Ni thermometer seems to respond very rapidly to temperature changes, especially to heating. There is a discrepancy between the empirical and experimental calibration of the Ni-in-garnet thermometer (for discussion see Canil, 1994, 1996, 1999; Griffin & Ryan, 1996). We recommend the use of the experimental calibration of Canil (1999). The general problem is that an empirically calibrated thermometer (e.g. Ryan et al., 1996) can only be as precise as the independent P–T estimates of the used samples are. The calibration of Ryan et al. (1996) does not consider the different pressures of their samples. Since higher temperature mantle samples generally are derived from higher pressure regimes in the mantle, the linear regression from Ryan et al. (1996) in an lnDNiGrt/Ol versus 1/T diagram (see Canil, 1999, page 245) is characterised by a flatter curve than the regression of Canil (1999). This results in an underestimation at T < 900 °C and overestimation at T > 1400 °C of the calculated temperatures. Canil (1999) has clearly shown that a pressure correction for the samples used by Ryan et al. (1996) yields an almost identical regression line as the one of Canil (1999). The general advantage of this thermometer is that the calculated temperatures are reliable whether the garnet contains significant amounts of Fe3+ or not. Another advantage is the pressure independence of this thermometer. As Ni contents in peridotitic garnets are quite low, electron microprobe analyses are in most cases not accurate enough. However, numerous other analytical methods such as secondary ion mass spectrometry, laser ablation ICP-MS and proton probe work quite well. Partitioning of transition elements (Sc, V, Cr, Co and Mn) between orthopyroxene and clinopyroxene in peridotitic and websteritic mantle rocks It is well known from several studies that the partitioning of selected transition elements such as Sc, V, Cr, Co and Mn between mantle minerals is strongly controlled by temperature (Hervig & Smith, 1982; Hervig et al., 1986; Stosch, 1987; Bodinier et al., 1987). A detailed study, evaluating the partitioning of transition elements between

E.J. Krogh Ravna & J. Paquin

242

orthopyroxene and clinopyroxene for geothermobarometry yields the following five empirical equations (Seitz et al., 1999): TSc [C] 

5663  17.64 P  273, 3.25  ln DSc

TV [C] 

3975  18.06 P  273, 2.27  ln DV

TCr [C] 

2829  11.00 P  273, 1.56  ln DCr

TMn [C] 

2229  0.20 P  273,  1.37  ln DMn

TCo [C] 

2358  4.31P  273,  0.98  ln DCo

with DM = concentration [cations p.f.u.] of element M in orthopyroxene/concentration [cations p.f.u.] of element M in clinopyroxene, T in K and P in kbar. All five empirical thermometer formulations show only minor pressure dependence (CD Image 5). Moreover, compositional influences on these thermometers such as the amount of Na or tetrahedral Al in the pyroxenes have not been found. The potential of these thermometers lies in the different diffusivities of the relevant elements, which can be used to evaluate equilibrium/disequilibrium conditions (Paquin & Altherr, 2001a,b) and/or to model thermal events based on zoning profiles. As the diffusivity of Co and Mn is much faster than that for Sc, V and Cr, different thermal stages can be indicated by these elements. In the case of equilibrium all five independent thermometers should yield concordant values. At this stage these trace element thermometers have not frequently been used for temperature estimation, most likely due to the large-scale analytical methods such as LA-ICP-MS or SIMS needed for measurement of the transition elements. Therefore comparisons between more widely used thermometers and these new methods are not available. Nevertheless, experimental calibrations of these thermometers are necessary to confirm the empirical calibrations. Ca-Cr system in lherzolitic garnets Another test for the calculated peak metamorphic condition is provided by the Ca/Cr system in lherzolitic garnet coexisting with clinopyroxene, orthopyroxene and olivine. Concerning the Ca/Cr ratio in garnet, Brenker & Brey (1997) have shown that the abundance of Cr in garnet is a function of the effective bulk composition and does not depend on P and T. For a given Cr content in garnet, Ca in garnet deceases with increasing pressure and increasing temperature. At constant P and T, Ca increases linearly with Cr, so that the Ca/Cr ratio in a P–T diagram is defined by a straight line. This co-variation between Ca and Cr in garnet of lherzolitic samples was first recognised

Thermobarometric methodologies

243

by Sobolev et al. (1973) and Sobolev (1977). Based on the studies of Brey (1990) and Nickel (1983), Brenker & Brey (1997) formulated an empirical equation between P and T and the Ca/Cr ratio in lherzolitic garnets with the assumption that the slope of the CaCr isolines are constant (lherzolite trend) for any P–T value. Please note that contrary to the original equation of Brenker & Brey (1997) the terms XCaGrt and XCrGrt stand for cations per formula unit (Brenker, pers. comm.) The equation is written as Grt Grt X Ca  0.449 X Cr  107  10 6 T  146  10 4 P  0.567,

with T in K and P in GPa. In a detailed study of compositional systematics related to tectonic settings of Crpyrope garnets in the lithospheric mantle Griffin et al. (1999) have shown that the slope of the lherzolite trend varies with tectonic setting and therefore with the local geotherm. This suggests that the P/T ratio exerts a control on the Ca/Cr ratio. As high Ca garnets with high Ca/Cr ratios are not covered by the algorithm of Brenker & Brey (1997), Griffin et al. (1999) suggest not to use the algorithm for high Ca-garnets and, moreover, not to compare samples that may be derived from different tectonic settings. Nimis & Trommsdorff (2001) stated that the uncertainties of the thermobarometric formulation are very large with calculated standard errors of estimate of 1.2 GPa and 159 °C and conclude that the formula is of dubious validity. But Paquin & Altherr (2001b) pointed out that the evaluation of Nimis & Trommsdorff (2001) was misleading as they ignored the variable Cr contents in garnet in their evaluation. However, the Ca-Cr system should be considered as a test for the calculated metamorphic temperatures and pressures and not as a proper geothermobarometer. The clinopyroxene/plagioclase symplectite geothermometer According to metallurgical concepts the phase transition omphacite clinopyroxene + plagioclase is classified as a discontinuous precipitation reaction (Joanny et al., 1991). Thus the lamellar spacing (L) depends on temperature (T) according to the growth law log L = A – B/T, where A and B are constants. Therefore thin symplectites characterise lower, while coarse symplectites higher temperatures. The Jd content of the symplectitic clinopyroxene formed together with plagioclase in the presence of quartz will act as a monitor of pressure. Thus the evolution of symplectites after omphacite can be used to evaluate parts of the decompression postdating maximum P conditions of eclogites. Joanny et al. (1991) presented empirical thermometric expressions as functions of the spacing of symplectitic lamellae. This method has a potential that has not been widely used, but recently Ravna & Roux (2002) on the basis of three separate symplectite stages deducted a detailed uplift path for eclogites within the uppermost allochthon of the Scandinavian Caledonides.

E.J. Krogh Ravna & J. Paquin

244

Geobarometers As geothermometers appear to be abundant, reliable geobarometers for HP/UHP rocks are scarce. Most of them are based on reactions involving transfer of Al from tetrahedral to octahedral coordination sites. In the following we will present the more common geobarometers, and discuss some additional approaches that may be useful if they are experimentally calibrated. The Al-in-orthopyroxene barometer The octahedral Al3+ content in orthopyroxene coexisting with garnet in peridotites is known as a suitable indicator of the pressure conditions at which the rocks equilibrated. In a simple MAS system the Mg-Tschermaks net transfer reaction with Al2Mg–1Si–1 exchange between orthopyroxene in equilibrium with garnet is written as Mg2Si2O6 + MgAl2SiO6 = Mg3Al2Si3O12. En

MgTs

Prp

Several studies have been carried out to calibrate various versions of this barometer (e.g. Harley, 1984b; Nickel & Green, 1985; Gasparik, 1987; Brey & Köhler, 1990; Taylor, 1998). Here we only present the most widely used calibrations. The exact barometer equations should be looked up in the original papers, as the equations are too large to be shown here. Harley (1984b) investigated experimentally the pressure–temperature–compositional dependence of the Al3+ content in orthopyroxene coexisting with garnet in both the FMAS and CFMAS systems (PH84). He has taken XAlM2 as IVAl/2. His calibration does not consider the effect of minor components such as Cr, Mn, Na, Ti and in particular Fe3+ that can influence the pressure results significantly. Moreover, Carson & Powell (1997) addressed problems such as analytical uncertainties, the estimation of Fe3+, effects of retrograde diffusion and off-centre sectioning of garnet on P–T estimates by using both the TH84 and PH84. An evaluation of 13 different Al-in-orthopyroxene calibrations has been carried out by Carswell & Gibb (1987). These authors found out that only the calibration of Nickel & Green (1985) yielded satisfactory results. The Nickel & Green (1985) barometer version, which is based on experiments in the CMAS and Cr-bearing M1 SMACCR systems, is using a modified XAl–Cr term that allows the presence of Cr, Ti and 3+ Fe in M1 and Na in M2 sites. However, Brey & Köhler (1990) found that the Nickel & Green (1985) version works well only in the pressure range of their experiments and that extrapolation to higher pressures result in a strong inaccuracy. Brey & Köhler (1990) formulated a new version of this barometer (PBK90) which is capable of reproducing the experimental conditions for all available systems over a wide pressure range. Their barometer expression is based on thermodynamic evaluation in the MAS system by Gasparik & Newton (1984) and reproduces their own experiments within ± 0.22 GPa without any systematic dependence on either T or P. Recently Taylor (1998) calibrated a new version based on non-reversed experiments with P 3.5 GPa in which a Ti correction was applied to the activity term

Thermobarometric methodologies

245

for the Mg-Tschermaks component. However, to date the reversed experimental dataset of Brey et al. (1990) is the only one that covers pressures in excess of 3.5 GPa. Their experimental pressures can excellently be reproduced by PBK90, hence this barometer version seems most reliable in calculation pressures. Effect of Fe3+ on the Al-in-orthopyroxene barometer The effect of the oxidation state of iron on the Al-in-orthopyroxene barometry is not directly evident as it is in the Fe-Mg thermometry. The significant influence arises from the effect which Fe3+ may have on the determination of the activity of the MgAl2SiO6 component in orthopyroxene. Several studies have shown that the content of Al3+ in the octahedral sites (M1) in orthopyroxene coexisting with garnet is particularly pressure dependent and hence serve as a geobarometer for garnet-bearing peridotites (e.g. MacGregor, 1974; Harley & Green, 1982; Harley, 1984b; Finnerty & Boyd, 1984; Nickel & Green, 1985; Brey & Köhler, 1990). But in more chemically complex orthopyroxenes not only Fe3+ but also Cr3+ and Ti4+ are commonly restricted to the M1 sites, which would require equal amounts of Al in the tetrahedral sites to retain charge balance. This kind of substitution clearly has the potential to limit the amount of Al3+ available for substitution into the octahedral site M1, resulting in a considerable reduction of the calculated activity of the MgAl2SiO6 component in orthopyroxene (Canil & O’Neill, 1996). Up to date, the extent of the error produced by ignoring Fe3+ (and also Cr3+ and Ti4+) in calculating metamorphic pressures using the Al-in-orthopyroxene barometer remains completely unknown (Canil & O’Neill, 1996). The Cr-in-clinopyroxene barometer Recently a new single clinopyroxene barometer (PNT00) was calibrated based on experimental clinopyroxenes synthesised at 850–1500 °C and 0.0–6.0 GPa in the CMS, CMAS–Cr and more complex lherzolitic systems, including previously published data (Nickel, 1989; Brey et al., 1990; Taylor, 1998). This calibration thus covers a wide range of natural peridotitic compositions from fertile pyrolite to refractory, high-Cr lherzolite (Nimis & Taylor, 2000). For geobarometric evaluation they considered the Cr exchange between clinopyroxene and coexisting garnet. In their barometric equation, pressure is formulated as a function of temperature and clinopyroxene composition, expressed as PNT00 [kbar]  

 Cr # cpx T cpx  15.483 ln ln aCaCrTs 126.9  T

with acpx CaCrTs = Cr – 0.81  Cr#  (Na+K), Cr# =

 T  71.38  107.8,

Cr , with elements in atoms per 6 Cr  Al

oxygens and T in K. This formulation has a temperature dependence of around 0.12–0.24 GPa/50 °C which is less that that of the widely used Al-in-orthopyroxene barometer. However, a careful evaluation of the database for their calibration reveals a severe problem. There is a systematic divergence between experimental pressures of the reversed experiments

E.J. Krogh Ravna & J. Paquin

246

carried out by Brey et al. (1990) and calculated pressures using PNT00. The PNT00 tends to overestimate at low pressures and increasingly underestimates at high pressures (Paquin & Altherr, 2001b). One might speculate that this observation is due to incomplete equilibration in the experiments of Nickel (1989) and Taylor (1998). Therefore more independent reversed experiments over a wide pressure range are highly desirable to verify this barometer version. Geothermobarometry based on the assemblage garnet–clinopyroxene–phengite–kyanite–quartz/coesite The assemblage garnet + clinopyroxene + phengite ± kyanite ± quartz/coesite is not uncommon in some Al-rich eclogites, and equilibria between these phases have successfully been used for estimation of P and T of HP and UHP rocks. The following net transfer reactions between these phases in the KCMASH system can be written as: the phengite absent reaction [phe] 3 diopside + 2 kyanite = 1 grossular + 1 pyrope + 2 coesite/quartz,

(1a,1b)

the clinopyroxene absent reactions [di, gr] 1 pyrope + 3 muscovite + 4 coesite/quartz = 3 celadonite + 4 kyanite,

(2a, 2b)

and the SiO2–kyanite absent reaction [SiO2, ky] 6 diopside + 3 muscovite = 2 grossular + 1 pyrope + 3 celadonite.

(3)

In the KCMASH system these reactions define an invariant point in both the coesite and quartz stability field, depending on which SiO2 polymorph is stable (Fig. 3). The geothermobarometric methods based on the net transfer reactions in this system are suggested to be less affected by later thermal re-equilibration than common cation exchange thermometers, and the methods also diminish the problems related to estimation of Fe3+/Fetot in omphacite. Waters & Martin (1993) presented a new geobarometer based on the thermodynamic data set of Holland & Powell (1990) for the fairly common eclogitic mineral assemblage garnet + clinopyroxene + phengite through Equation 3 with the derived barometric expression P3WM1 [kbar] = 26.9 + 0.0159T [K] – 0.00249T lnK. This expression was later revised (Waters, 1996) to P3WM2 [kbar] = 28.05 + 0.02044T – 0.003539T lnK, where lnK = 6 lnadi – lnapy – 2 lnagr + 3 lnainvphe and ainvphe = aideal mu/aideal cel = (XAlM1)(XAlT1)/(XMgM1)(XSiT1), with XAlT1=(4 – Si) and XSiT1 = (Si – 2).

Thermobarometric methodologies

247

Fig. 3. THERMOCALC plot of reaction curve bundles in between the phases garnet-clinopyroxene-phengitekyanite-SiO2 in the coesite and quartz stability fields. Reaction numbers refer to reactions described in the text (from Ravna & Terry, 2003).

Activity models for diopside and garnet were taken from Holland (1990) and Newton & Haselton (1981), respectively. This barometer has shown to be very suitable for phengite-bearing HP and UHP eclogites from Dabie Shan, China (Carswell et al., 1997) and the Western Gneiss Complex of southern Norway (Wain, 1997, 1998; Cuthbert et al., 2000). The relation between garnet, clinopyroxene and kyanite in coesite-bearing rocks has been proposed as a potential geobarometer by Sharp et al. (1992) and Nakamura & Banno (1997). Ravna & Terry (2001, 2003) presented geothermobarometric expressions for UHP assemblages among (a) garnet-clinopyroxene-kyanite-phengite-coesite, and for the corresponding HP assemblages among (b) garnet-clinopyroxene-kyanite-phengite-

E.J. Krogh Ravna & J. Paquin

248

quartz, using the net transfer reactions 1–3 above with the equilibrium constants expressed as K1a 

K 2a 

grt grt coes 2 apyr agrs (aSiO ) 2

K1b 

ky (adicpx ) 3 (aAl )2 2SiO5 phe 3 ky (acel ) (aAl )4 2SiO5 grt phe 3 coes 4 (apyr )(amus ) (aSiO ) 2

K 2b 

grt grt qtz apyr agrs (aSiO )2 2 ky (adicpx ) 3 (aAl )2 2SiO5 phe 3 ky (acel ) (aAl )4 2SiO5 grt phe 3 qtz (apyr )(amus ) (aSiO )4 2

and K4  K3

grt grt 2 phe 3 apyr (agrs ) (acel ) phe 3 (adicpx ) 6 (amus )

.

They formulated a set of linearised barometric expressions for each of these reactions, given as Equation 1a (phe, q) P1aRT [GPa] = 7.235 – 0.000659T + 0.001162T lnK1a, Equation 1b (phe, coe) P1bRT [GPa] = 11.424 – 0.001676T + 0.002157T lnK1b, Equation 2a (di, gr, q) P2aRT [GPa] = –2.624 + 0.005741T + 0.0004549T lnK2a, Equation 2b (di, gr, coe) P2bRT [GPa] = –0.899 + 0.003929T + 0.0002962T lnK2b, Equation 3 (ky, coe/q) P3RT [GPa] = 1.801 + 0.002781T + 0.0002425T lnK4. Additional constraints demanded the iso-lnK curves for the equivalent reactions including SiO2 to intersect at the quartz–coesite transition. The present expression for reaction 4 deviates from that given by Waters & Martin (1993) and Waters (1996), probably due to a different approach in retrieving the data. Equations 1a, 2a and 3 all intersect in a single P–T point within the coesite field, while 1b, 2b and 3 intersect within the quartz stability field (Fig. 3). The intersection of any two of these sets of reactions will thus uniquely define P and T for a single sample. Calculations by THERMOCALC (Powell & Holland, 1988) on a variety of samples yield averaged standard deviations for this intersection of ± 65 °C and ± 0.32 GPa in the coesite field of and ± 82 °C, ± 0.32 GPa in the quartz stability field. The intersections of the present linear curves deviate less than 10 °C, 0.02 GPa from those obtained by THERMOCALC.

Thermobarometric methodologies

249

Fig. 4. A plot showing the compositional effect of garnet (a), clinopyroxene (b) and phengite (c) on the intersection of the reactions between the minerals garnet-clinoyproxene-phengite (gentle slope in the P–T diagram) and garnet-clinopyroxene-kyanite-SiO2 (steeper slopes). Note the change in slope of the latter equilibrium as pressures change from the quartz to coesite stability field.

250

E.J. Krogh Ravna & J. Paquin

Ravna & Terry (2001, 2003) used a combination of the ideal activity model for the phengite solid solution proposed by Holland & Powell (1998), the clinopyroxene activity model of Holland (1990), and the garnet activity model of Ganguly et al. (1996). Phengite structural formulae has been normalised to Si,Al,Ti,Cr,Fe,Mn,Mg = 12.000. Garnet is normalised to X = 3.000, Y = 2.000, and Fe3+ = 2 – (Al + Cr + Ti). Maximum recordable P conditions for a specific eclogite should, according to Equation 3, be grt 2 grt represented by garnets with maximum (agrs ) apy , omphacite with minimum adicpx (and phe correspondingly maximum XJd) and phengite with maximum a Al-cel (maximum Si content) (Carswell & Zhang, 1999; Carswell et al., 2000). The relationship between pressure, temperature and composition of garnet, clinopyroxene and phengite, respectively, is shown in Figure 4 a–c. Application of these thermobarometers to relevant eclogites from various worldwide localities shows good consistency with petrographic evidence (Terry et al., 2000; Gilotti & Ravna, 2002; Ravna & Terry, 2003). Ravna & Terry (2003) give several examples of the wide applicability of the methods, ranging from low-T blueschist type eclogites from the Franciscan and WGR, to UHP eclogites from WGR and Dabie Shan and coesite-kyanite eclogite xenoliths from kimberlites. Equation 1a may serve as a geobarometer in combination with the common garnet–clinopyroxene Fe–Mg geothermometer in phengite absent kyanite-bearing eclogites, while Equation 3 has proven to be a reliable geobarometer in phengite eclogites (Waters & Martin, 1993; Carswell et al., 1997; Wain, 1997, 1998; Cuthbert et al., 2000). Equations 2a and 2b are suggested to be useful pressure indicators for clinopyroxene-free UHP/HP pelitic schists, given that a reliable independent temperature estimate is possible. Coggon & Holland (2002) presented linearised expressions for three fluid-absent reactions, including the Waters & Martin (1993) barometer. They used Equations 2b, 3 and almandine + 3 muscovite + 4 quartz = 3 Fe-celadonite + 4 kyanite, (5) based on revised mixing properties of phengitic micas. They obtained the following barometric expressions: P3CH [GPa] = 1.978 + 0.002726T + 0.0002627T lnK, P2bCH [GPa] = –0.910 + 0.003933T + 0.0003258T lnK, P4CH [GPa] = 1.969 + 0.002597T + 0.0003474T lnK. Expressions for Equations 3 and 2b do not deviate much from the corresponding expressions of Ravna & Terry (2001, 2003). Other, less common geobarometers suitable for HP/UHP rocks Geobarometers involving plagioclase Plagioclase is not stable at UHP conditions; thus the various common geobarometers involving this phase (GASP, GAES, GADS, AbJdQ) are of no relevance for such rocks. Nevertheless, minimum pressure estimates based on the jadeite content of omphacite in

Thermobarometric methodologies

251

HP and UHP rocks devoid of plagioclase are frequently used, even in rocks lacking a SiO2 phase (quartz or coesite). The presence of coesite or polycrystalline quartz pseudomorphs after coesite adds another, yet higher minimum pressure limit, as does the less common occurrence of diamond. Geobarometers involving zoisite/clinozoisite Zoisite- and clinozoisite-bearing eclogites provide assemblages suitable for geobarometry. The reactions 6 zoisite = 4 grossular + 5 kyanite + coesite + 3 H2O and 15 diopside + 12 zoisite = 13 grossular + 5 pyrope + 12 quartz/coesite + 6H2O have both been used (e.g. Poli & Schmidt, 1998; Terry et al., 2000). The K content of clinopyroxene has been suggested as a potential geobarometer in UHP rocks, supported by experimental data of e.g. Luth (1995), Harlow (1997, 2002) and Okamoto & Maruyama (1998). Shatsky et al. (1995) have reported K2O contents in clinopyroxenes up to 1.2 wt% in diamond-bearing metamorphic rocks in the Kokchetav massif, and noted that during exhumation the K2O was exsolved as K-feldspar lamellae within diopside in carbonate-bearing rocks. There has, however, never been found any exsolved K-feldspar lamellae in omphacite in any UHP terranes, which Shatsky et al. (1995) attributed to the low-K bulk chemistry of ordinary eclogites. However, phengite in eclogites is fairly common, and should provide a K-source for the buffering of K in omphacite. Domanik & Holloway (2000) studied the phase relations of phengitic muscovite in a calcareous metapelite from 6.5–11 GPa. At 9 GPa and 900 °C the K2O content in omphacite coexisting with phengite was 0.2 wt%, and around 0.1 at P < 8 GPa. Harlow (2002) did multi-anvil experiments on a mixture of natural diopside and F-rich phlogopite in the P–T interval 3.0 to 11 GPa and 1100–1500 °C. In clinopyroxene the Kcpx (KAlSi2O6) content increases with increasing pressure at pressures above 5 GPa without any noticeable temperature effect. These experiments point to the K content of clinopyroxene as a potential geobarometer at higher pressures. The difference in the experiments of Domanik & Holloway (2000) on one side and those of e.g. Harlow (2002) may indicate that K is not as easily partitioned into Na-rich clinopyroxenes as into Ca-rich ones. If reliable geobarometers involving Kcpx should be formulated, one should look for suitable buffering assemblages as e.g. Muscovite (in phengite) = Kcpx (in clinopyroxene) + kyanite + H2O KAl2AlSi3O10(OH)2 = KAlSi2O6 + Al2SiO5 + H2O and K-feldspar = Kcpx + coesite KAlSi3O8 = KAlSi2O6 + SiO2.

252

E.J. Krogh Ravna & J. Paquin

Ca-Eskola molecule bearing clinopyroxenes at UHP conditions Eskola (1921) first reported pyroxenes with structural vacancies. These vacancies can be ascribed to the end member Ca0.5AlSi2O6, which was named calcium Eskola pyroxene (CaEs) by Khanukhova et al. (1976). Experimental work on the stability field of CaEs suggests that non-stoichiometric pyroxenes are highly P-sensitive and should be a stable component of natural clinopyroxenes, especially in the presence of excess SiO2 (Gasparik & Lindsley, 1980). Cation deficiencies in Al-rich (omphacitic) clinopyroxenes have been reported from grospydite xenoliths (Sobolev et al., 1968; Smyth & Hatton, 1977). The appearance of oriented needles of quartz in omphacite seems to be fairly common in UHP eclogites and related rocks (Smith, 1984, 1988; Bakun-Czubarow, 1992; Liou et al., 1998; Zhang & Liou, 1998; Schmädicke & Müller, 2000; Tsai & Liou, 2000; Terry & Robinson, 2001; Dobrzhinetskaya et al., 2002). Katayama et al. (2000) also described matrix clinopyroxenes with quartz rods in eclogite from the Kokchetav massif, Kazakstan, apparently recrystallised at P > 6 GPa, T > 1000 °C. Inclusions of clinopyroxenes in zircon had calculated cation totals significantly less than 4.0 per six O atoms, and showed no exsolution textures. Bruno et al. (2002) reported jadeite with a high portion (0.08–0.17) of the Ca-Eskola molecule from an UHP meta-granodiorite from the Dora-Maira massif, Western Alps. These observations have been proposed as a result of exsolution of a former non-stoichiometric omphacite through the reaction (Smyth, 1980) 2 CaEs = CaTs + 3 Qtz. Katayama et al. (2000) suggested that further experiments on the Ca-Eskola component in clinopyroxene may yield a new geobarometer for UHP metamorphic rocks. High-Al titanite has been described from a variety of eclogites and associated HP/UHP rocks (Smith, 1977, 1980; Smith & Lappin, 1982; Franz & Spear, 1985; Krogh et al., 1990; Sobolev & Shatsky, 1990; Hirajima et al., 1992; Carswell et al., 1996; Ye & Ye, 1996), and has been taken as evidence of HP/UHP metamorphism. Experimental data (Smith, 1980, 1981; Troitzsch & Ellis, 1999) show that titanite with high Al and F content can be produced at elevated pressures. However, Al- and F-rich titanites have also been reported from low-P metamorphic rocks (Enami et al., 1993; Markl & Piazolo, 1999), indicating that these compositions are not restricted to HP/UHP metamorphism, but are formed in response of the activity of F and bulk compositions (Franz & Spear, 1985; Enami et al., 1993; Carswell et al., 1996; Markl & Piazolo, 1999), as summarised by Castelli & Rubatto (2002). Troitzsch & Ellis (2002) demonstrated that petrogenetic grids based on newly derived thermodynamic properties of Al-rich titanites could explain the predominant occurrence of natural Al-rich titanite at high metamorphic grade such as eclogite facies conditions. Wide spacing of the Al-isopleths for titanite of many high-grade assemblages prevents, however, their use as geobarometers or geothermometers.

Conclusions The most widely used set of thermobarometers for garnet-bearing ultrabasites is that of Brey & Köhler (1990). The advantage of this set of thermobarometers for lherzolites

Thermobarometric methodologies

253

based on independent reactions between coexisting minerals is the possibility to evaluate whether equilibrium or disequilibrium exists between the relevant mineral phases. For common (bimineralic) eclogites, we still have to rely on the garnet–clinopyroxene Fe2+–Mg exchange thermometer with all its uncertainties, while reliable pressure estimates are impossible to obtain at present. For phengite- and kyanite-bearing eclogites, pressures and temperatures can uniquely be obtained by either combining the garnet–clinopyroxene Fe2+–Mg exchange thermometer with the garnet–clinopyroxene–phengite barometer, or combining the latter with the garnet–clinopyroxene–kyanite–SiO2 thermobarometer.

References Ai, Y. (1994): A revision of the garnet-clinopyroxene Fe2+-Mg exchange geothermometer. Contrib. Mineral. Petrol., 115:467–473. Bakun-Czubarow, N. (1992): Quartz pseudomorphs after coesite in the eclogite-granulite facies rock series of the Zlote Mountains in the Sudetes (SW Poland). Arch. Mineral. (Warszawa), 47:3–25. Berman, R.G. (1991): Thermobarometry using multi-equilibrium calculations: a new technique with petrological applications. Can. Mineral., 29:833–855. Berman, R.G., Aranovich, L.Ya. & Pattison, D.R.M. (1995): Reassessment of the garnet-clinopyroxene Fe-Mg exchange thermometer: II. Thermodynamic analysis. Contrib. Mineral. Petrol., 119:30–42. Bertrand, P. & Mercier, J.-C.C. (1985): The mutual solubility of coexisting ortho- and clinopyroxene: toward an absolute geothermometer for the natural system? Earth Planet. Sci. Lett., 76:109–122. Bodinier, J.L., Dupuy, C., Dostal, J. & Merlet, C. (1987): Distribution of trace transition elements in olivine and pyroxenes from ultramafic xenoliths: Application of microprobe analysis. Am. Mineral., 72:902–913. Boyd, F.R. (1973): A pyroxene geotherm. Geochim. Cosmochim. Acta, 37:2533–2546. Brenker, F.E. & Brey, G.P. (1997): Reconstruction of the exhumation path of the Alpe Arami garnet peridotite body from depths exceeding 160 km. J. Metamorph. Geol., 15:581–592. Brey, G.P. (1990): Geothermobarometry for lherzolites: Experiments from 10 to 60 kb, new thermobarometers and application to natural rocks. Habilitation Thesis, Technische Hochschule Darmstadt, Germany, 227 p. Brey, G.P. & Köhler, T. (1990): Geothermobarometry in four-phase lherzolites II. New thermobarometers, and practical assessment of existing thermobarometers. J. Petrol., 31:1353–1378. Brey, G.P., Köhler, T. & Nickel, K.G. (1990): Geothermobarometry in four-phase lherzolites I. Experimental results from 10 to 60 kb. J. Petrol., 31:1313–1352. Bruno, M., Compagnoni, R., Hirajima, T. & Rubbo, M. (2002): Jadeite with the Ca-Eskola molecule from an ultra-high pressure metagranodiorite, Dora-Maira Massif, Western Alps. Contrib. Mineral. Petrol., 142:515–519. Canil, D. (1994): An experimental calibration of the ‘Nickel in garnet’ geothermometer with applications. Contrib. Mineral. Petrol., 117:410–420. Canil, D. (1996): An experimental calibration of the ‘Nickel in garnet’ thermometer with applications: reply. Contrib. Mineral. Petrol., 124:219–223. Canil, D. (1999): The Ni-in-garnet geothermometer: calibration at natural abundances. Contrib. Mineral. Petrol., 136:240–246. Canil, D. & O’Neill, H.St.C. (1996): Distribution of ferric iron in some upper-mantle assemblages. J. Petrol., 37:609–635. Carson, C.J. & Powell, R. (1997): Garnet–orthopyroxene geothermometry and geobarometry: error propagation and equilibration effects. J. Metamorph. Geol., 15:679–686. Carswell, D.A. & Gibb, F.G.F. (1987): Evaluation of mineral thermometers and barometers applicable to garnet lherzolite assemblages. Contrib. Mineral. Petrol., 95:499–511. Carswell, D.A. & Harley, S.L. (1990): Mineral baromety and thermometry. In Carswell, D.A. (ed.): Eclogite facies rocks. Glasgow, London: Blackie, 83–110. Carswell, D.A. & Zhang, R.Y. (1999): Petrographic characteristics and metamorphic evolution of ultrahighpressure eclogites in plate-collision belts. Int. Geol. Rev., 4, 781–798.

254

E.J. Krogh Ravna & J. Paquin

Carswell, D.A., Harvey, M.A. & Al-Samman, A. (1983): The petrogenesis of contrasting Fe-Ti and Mg-Cr peridotite types in the high-grade gneiss complex of Western Norway. Bull. Minéral. 106:727–750. Carswell, D.A., Krogh, E.J. & Griffin, W.L. (1995): Norwegian orthopyroxene eclogites: calculated equilibrium conditions and petrogenetic implications. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen – Scandinavia and related areas. Chichester: Wiley. 823–841. Carswell, D.A., Wilson, R.N. & Zhai, M. (1996): Ultra-high pressure aluminous titanites in carbonate-bearing eclogites at Shuanhe in Dabieshan, central China. Mineral. Mag., 60:461–471. Carswell, D.A., O´Brien, P.J., Wilson, R.J. & Zhai, M. (1997): Thermobarometry of phengite-bearing eclogites in the Dabie Mountains of central China. J. Metamorph. Geol., 15:239–252. Carswell, D.A., Wilson, R.N. & Zhai, M. (2000): Metamorphic evolution, mineral chemistry and thermobaromety of schists and orthogneisses hosting ultra-high pressure eclogites in the Dabieshan of central China. Lithos, 52:121–155. Castelli, D. & Rubatto, D. (2002): Stability of Al- and F-rich titanite in metacarbonate: petrologic and isotopic constraints from a polymetamorphic eclogitic marble of the internal Sesia Zone (Western Alps). Contrib. Mineral. Petrol., 14:627–639. Coggon, R. & Holland, T.J.B. (2002): Mixing properties of phengitic micas and revised garnet-phengite thermobarometers. J. Metamorph. Geol., 20:683–696. Cuthbert, S.J., Carswell, D.A., Ravna, E.J. Krogh & Wain, A.L. (2000): Eclogites and eclogites in the Western Gneiss Region, Norwegian Caledonides. Lithos, 52:165–195. Davis, B.T.C. & Boyd, F.R. (1966): The join Mg2Si2O6–CaMgSi2O6 at 30 kilobars pressure and its application to pyroxenes from kimberlites. J. Geophys. Res., 71:3567–3576. Dimanov, A. & Sautter, V. (2000): ‘Average’ interdiffusion of (Fe, Mn)-Mg in natural diopside. Eur. J. Mineral., 12:749–760. Dobrzhinetskaya, L.F., Schwinehage, R., Massonne, H.-J. & Green, H.W. (2002): Silica precipitates in omphacite from eclogite at Alpe Arami, Switzerland: evidence of deep subduction. J. Metamorph. Geol., 20:481–492. Domanik, K.J. & Holloway, J.R. (2000): Experimental synthesis and phase relations of phengitic muscovite from 6.5 to 11 GPa in a calcareous metapelite from the Dabie mountains. Lithos, 52:51–77 Droop, G.T.R. (1987): A general equation for estimating Fe3+ concentrations in ferromagnesian silicates and oxides from microprobe analyses using stoichiometric criteria. Mineral. Mag., 51:431–435. Ellis, D.J. & Green, D.H. (1979): An experimental study of the effect of Ca upon the garnet-clinopyroxene FeMg exchange equilibria. Contrib. Mineral. Petrol., 71:13–22. Enami, M., Suzuki, K., Liou, J.G. & Bird, D.K. (1993): Al-Fe3+ and F-OH substitutions in titanite and constraints on their P-T dependence. Eur. J. Mineral., 5:219–231. Eskola, P. (1921): On the eclogites of Norway. Videnskap. skr., Kristiania, I. Mat.-Naturvid. kl., 8:1–118. Finnerty, A.A. & Boyd, F.R. (1984): Evaluation of thermobarometers for garnet peridotites. Geochim. Cosmochim. Acta, 48:15–27. Franz, G. & Spear, F.S. (1985): Aluminous titanite (sphene) from the eclogite zone, South-Central Tauern Window, Austria. Chem. Geol., 50:33–46. Ganguly, J. & Saxena, S.K. (1987): Mixtures and mineral reactions /Miner. Rocks Ser., 19/. Berlin: SpringerVerlag. Ganguly, J. & Tazzoli, V. (1994): Fe2+-Mg interdiffusion in orthopyroxene: Retrieval from the data on intercrystalline exchange reaction. Am. Mineral., 79:930–937. Ganguly, J., Cheng, W. & Tirone, M. (1996): Thermodynamics of aluminosilicate garnet solid solution: new experimental data, an optimized model, and thermometric applications. Contrib. Mineral. Petrol., 126:137–151. Gasparik, T. (1987): Orthopyroxene thermobarometry in simple and complex systems. Contrib. Mineral. Petrol., 96:357–370. Gasparik, T. & Lindsley, D.H. (1980): Phase equilibria at high pressure of pyroxenes containing monovalent and trivalent ions. In Prewitt, C.T. (ed.): Pyroxenes /Rev. Mineral., 7/. Washington (D.C): Mineral. Soc. Am., 309–339. Gasparik, T. & Newton, R.C. (1984): The reversed Al-contents of orthopyroxene in equilibrium with spinel and forsterite in the system MgO-Al2O3-SiO2. Contrib. Mineral. Petrol., 85:186–196.

Thermobarometric methodologies

255

Gilotti, J.A. & Ravna, E.J.K. (2002): First evidence for ultrahigh-pressure metamorphism in the North-East Greenland Caledonides. Geology, 30:551–554. Graham, C.M. & Powell, R. (1984): A garnet-hornblende geothermometer: calibration, testing, and application to the Pelona Schist, Southern California. J. Metamorph. Geol., 2:13–21. Green, D.H. & Hellman, P.L. (1982): Fe-Mg partitioning between coexisting garnet and phengite at high pressures, and comments on a garnet-phengite geothermometer. Lithos, 15:253–266. Griffin, W.L. & Ryan, C.G. (1996): An experimental calibration of the ‘nickel in garnet’ geothermometer with application, by D. Canil: discussion. Contrib. Mineral. Petrol., 124:216–218. Griffin, W.L., Cousens, D.R., Ryan, C.G., Sie, S.H. & Suter, G.F. (1989): Ni in chrome pyrope garnets: a new geothermometer. Contrib. Mineral. Petrol., 103:199–202 Griffin, W.L., Smith, D., Ryan, C.G., O’Reilly, S.Y. & Win, T.T. (1996): Trace-element zoning in mantle minerals: Metasomatism and thermal events in the upper mantle. Can. Mineral., 34:1179–1193. Griffin, W.L., Fischer, N.I., Friedman, J., Ryan, C.G. & O’Reilly, S.Y. (1999): Cr-pyrope garnets in the lithospheric mantle. I. Compositional systematics and relations to tectonic setting. J. Petrol., 40:679–704. Harley, S.L. (1984a): An experimental study of the partitioning of Fe and Mg between garnet and orthopyroxene. Contrib. Mineral. Petrol., 86:359–373. Harley, S.L. (1984b): The solubility of alumina in orthopyroxene coexisting with garnet in FeO-MgO-Al2O3SiO2 and CaO-FeO-MgO-Al2O3-SiO2. J. Petrol., 25:665–696. Harley, S.L. & Green, D.H. (1982): Garnet–orthopyroxene barometry for granulites and peridotites. Nature, 200:697–701. Harlow, G.E. (1997): K in clinopyroxene at high pressure and temperature: An experimental study. Am. Mineral., 82:259–269. Harlow, G.E. (2002): Diopside + F-rich phlogopite at high P and T: Systematics, crystal chemistry and the stability of KMgF3, clinohumite and chondrodite. Geol. Mater. Res., 4:1–28. Hervig, R.L. & Smith, J.V. (1982): Temperature-dependent distribution of Cr between olivine and pyroxenes in lherzolite xenoliths. Contrib. Mineral. Petrol., 81:184–189. Hervig, R.L., Smith, J.V. & Dawson, J.B. (1986): Lherzolite xenoliths in kimberlites and basalts: Petrogenetic and crystallochemical significance of some minor and trace elements in olivine, pyroxenes, garnet and spinel. Trans. R. Soc. Edinburgh: Earth Sci., 77:181–201. Hirajima, T. (1996): Effect of jadeite-content on the garnet-clinopyroxene geothermometer for an ultrahighpressure eclogite complex. Proc. Jpn. Acad., ser. B, 72(10):208–213. Hirajima, T., Zhang, R.Y., Li, J.J. & Cong, B.L. (1992): Nyböite from Donghai area, Jiangsu province, eastern China. Mineral. Mag., 56:37–46. Holland, T.J.B. (1990): Activities of components in omphacite solid solutions. Contrib. Mineral. Petrol., 105:446–453. Holland, T.J.B. & Powell, R. (1990): An enlargened and updated internally consistent thermodynamic dataset with uncertainties and correlations: The system K2O-Na2O-CaO-MgO-MnO-FeO-Fe2O3-Al2O3-TiO2SiO2-C-H2-O2. J. Metamorph. Geol., 8:89–124. Holland, T.J.B. & Powell, R. (1998): An internally consistent thermodynamic data set for phases of petrological interest. J. Metamorph. Geol., 16:309–344. Joanny, V., van Roermund, H. & Lardeaux, J.M. (1991): The clinopyroxene/plagioclase symplectite in retrograde eclogites: a potential geothermometer. Geol. Rundsch., 80:303–320. Katayama, I., Parkinson, C.D., Okamoto, K., Nakajima, Y. & Maruyama, S. (2000): Supersilicic clinopyroxene and silica exsolution in UHPM eclogite and pelitic gneiss from the Kokchetav massif, Kazakhstan. Am. Mineral., 85:1368–1374. Khanukhova, L.T., Zharikov, V.A., Isbulatov, R.A. & Litvin, Y.A. (1976): Excess silica in solid solutions of high-pressure clinopyroxenes as shown by experimental study of the system CaMgSi2O6-CaAlSi2O6-SiO2 at 35 kilobars and 1,200 °C. Dokl. Akad. Sci. USSR, 229:182–184. Koons, P.O. (1984): Implications to garnet-clinopyroxene geothermometry of non-ideal solid solution in jadeitic pyroxenes. Contrib. Mineral. Petrol., 88:340–347. Krogh, E.J. (1980): Geochemistry and petrology of glaucophane-bearing eclogites and associated rocks from Sunnfjord, Western Norway. Lithos, 13:355–380.

256

E.J. Krogh Ravna & J. Paquin

Krogh, E.J. (1988): The garnet-clinopyroxene Fe-Mg geothermometer - a reinterpretation of existing experimental data. Contrib. Mineral. Petrol., 99:44–48. Krogh, E.J. & Råheim, A. (1978): Temperature and pressure dependence of Fe-Mg partitioning between garnet and phengite, with particular reference to eclogites. Contrib. Mineral. Petrol., 66:75–80. Krogh, E.J., Andresen, A., Bryhni, I., Broks, T.M. & Kristensen, S.E. (1990): Eclogites and polyphase P-T cycling in the Caledonian uppermost allochthon in Tromsö, northern Norway. J. Metamorph. Geol., 8:289–309. Lee, H.Y. & Ganguly, J. (1988): Equilibrium compositions of coexisting garnet and orthopyroxene: Experimental determinations in the system FeO-MgO-Al2O3-SiO2, and applications. J. Petrol., 29:93–113. Liou, J.G., Zhang, R.Y., Ernst, W.G., Rumble, D. & Maruyama, S. (1998): High-pressure minerals from deeply subducted metamorphic rocks. In Hemley, R. & Mao, D. (eds.): Ultrahigh-pressure mineralogy /Rev. Mineral., 37/. Washington (D.C.): Mineral. Soc. Am., 33–96. Luth, R.W. (1995): Potassium in clinopyroxene at high pressure: Experimental constraints (abstract). Eos, Trans. Am. Geophys. Union, 76:F711. Luth, R.W. & Canil, D. (1993): Ferric iron in mantle-derived pyroxenes and a new oxybarometer for the mantle. Contrib. Mineral. Petrol., 113:236–248. MacGregor, I.D. (1974): The system MgO-Al2O3-SiO2: solubility of Al2O3 in enstatite for spinel and garnet peridotite compositions. Am. Mineral., 59:110–119. Markl, G. & Piazolo, S. (1999): Stability of high-Al titanite from low-pressure calc-silicates in light of fluid and host rock composition. Am. Mineral., 84:37–47. Mori, T. & Green, D.H. (1978): Laboratory duplication of phase equilibria observed in natural garnet lherzolites. J. Geol., 86:83–97. Mysen, B.O. & Heier, K.S. (1972): Petrogenesis of eclogites in high grade metamorphic gneisses exemplified by the Hareidland eclogite, Western Norway. Contrib. Mineral. Petrol., 36:73–94. Nakamura, D. & Banno, S. (1997): Thermodynamic modelling of sodic pyroxene solid solution and its application in a garnet-omphacite-kyanite-coesite geothermobarometer for UHP metamorphic rocks. Contrib. Mineral. Petrol., 130:93–102. Neumann, E.-R. (1976): Two refinements for the calculation of structural formulae for pyroxenes and amphiboles. Nor. Geol. Tidsskr., 56:1–6. Newton, R.C. & Haselton, H.T. (1981): Themodynamics of the garnet-plagioclase-AlSiO5-quartz geobarometer. In Newton, R.C., Navrotsky, A. & Wood, B.J. (eds.): Thermodynamics of minerals and melts /Adv. Phys. Geochem., 1/. Berlin: Springer-Verlag, 131–147. Nickel, K.G. (1983): Petrogenesis of garnet and spinel peridotites. PhD Thesis, Univ. of Tasmania, Australia. Nickel, K.G. (1989): Garnet-pyroxene equilibria in the system SMACCR (SiO2-MgO-Al2O3-CaO-Cr2O3): the Cr-geobarometer. In Ross, J. (ed.): Kimberlites and related rocks, Vol. 2, Their mantle/crust setting, diamonds and diamond exploration. Proc. 4th Int. Kimberlite Conf. /Geol. Soc. Aus. Spec. Publ., 14/, 901–912. Nickel, K.G. & Green, D.H. (1985): Empirical geothermobarometry for garnet peridotites and implications for the nature of the lithosphere, kimberlites and diamonds. Earth Planet. Sci. Lett., 73:158–170. Nimis, P. & Taylor, W.R. (2000): Single clinopyroxene thermobarometry for garnet peridotites. Part I. Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in-Cpx-thermometer. Contrib. Mineral. Petrol., 139:541–544. Nimis, P. & Trommsdorff, V. (2001): Revised thermobarometry of Alpe Arami and other garnet peridotites from the Central Alps. J. Petrol., 42:103–115. Okamoto, K. & Maruyama, S. (1998): Multi-anvil re-equilibration experiments of a Dabie Shan ultrahighpressure eclogite within the diamond stability field. Isl. Arc, 7:52–69. O’Neill, H.St.C. (1980): An experimental study of Fe-Mg partitioning between garnet and olivine and its calibration as a geothermometer: corrections. Contrib. Mineral. Petrol., 72:337. O’Neill, H.St.C. & Wood, B.J. (1979): An experimental study of Fe-Mg partitioning between garnet and olivine and its calibration as a geothermometer. Contrib. Mineral. Petrol., 70:59–70. Paquin, J. & Altherr, R. (2001a):. New constraints on the P-T evolution of the Alpe Arami garnet peridotite body (Central Alps, Switzerland). J. Petrol., 42:1119–1140.

Thermobarometric methodologies

257

Paquin, J. & Altherr, R. (2001b): ‘New constraints on the P-T evolution of the Alpe Arami garnet peridotite body (Central Alps, Switzerland)’: Reply to comment by Nimis & Trommsdorff (2001). J. Petrol., 42:1781–1787. Pattison, D.R.M. & Newton, R.C. (1989): Reversed experimental calibration of the garnet-clinopyroxene FeMg exchange thermometer. Contrib. Mineral. Petrol., 101:87–103. Perchuk, L.L., Aranovich, L.Ya., Podlesskii, K.K., Lavrent’eva, I.V, Gerasimov, V.Yu., Fed’kin, V.V., Kitsul, V.-I., Karsakov, L.P. & Berdnikov, N.V. (1985): Precambrian granulites of the Aldan shield, eastern Siberia, USSR. J. Metamorph. Geol., 3:265–310. Poli, S. & Schmidt, M.W. (1998): The high-pressure stability of zoisite and phase relationships of zoisitebearing assembalges. Contrib. Mineral. Petrol., 130:162–175. Poli, S. & Schmidt, M.W. (2002): Petrology of subducted slabs. Annu. Rev. Earth Planet. Sci., 30:207–235. Powell, R. (1985): Regression diagnostics and robust regression in geothermometer/geobarometer calibration: the garnet-clinopyroxene geothermometer revisited. J. Metamorph. Geol., 2:33–42. Powell, R. & Holland, T.J.B. (1988): An internally consistent thermodynamic dataset with uncertainties and correlations: 3. Applications to geobarometry, worked examples and a computer program. J. Metamorph. Geol., 6:173–204. Powell, R. & Holland, T.J.B. (1994): Optimal geothermometry and geobarometry. Am. Mineral., 79:120–133. Råheim, A. & Green, D.H. (1974): Experimental determination of the temperature and pressure dependence of the Fe-Mg partition coefficient for coexisting garnet and clinopyroxene. Contrib. Mineral. Petrol., 48:179–203. Ravna, E.J.K. (2000a): The garnet-clinopyroxene geothermometer: an updated calibration. J. Metamorph. Geol., 18:211–219. Ravna, E.J.K. (2000b): Distribution of Fe2+ and Mg between coexisting garnet and hornblende in synthetic and natural systems: an empirical calibration of the garnet-hornblende Fe-Mg geothermometer. Lithos, 53:265–277. Ravna, E.J.K. & Roux, M.R.M. (2002): A detailed P-T path for eclogites in the Tromsö Nappe, N Norwegian Caledonides – a complex history of subduction, exhumation, reheating and recrystallization. AGU Fall Meeting (2002), Abstr., V51B–1253. Ravna, E.J.K. & Terry, M.P (2001): Geothermobaromety of phengite-kyanite-quartz/coesite eclogites. Eleventh Annu. V.M. Goldschmidt Conf., Abstr., Houston (Fl.): Lunar and Planetary Inst., 3145 (CDROM) Ravna, E.J.K. & Terry, M.P (2003): Geothermobaromety of UHP and HP eclogites and schists – an evaluation of equilibria between garnet-clinopyroxene-kyanite-phengite-coesite/quartz. J. Metamorph. Geol., submitted. Ryan, C.G., Griffin, W.L. & Pearson, N.J. (1996): Garnet geotherms: Pressure-temperature data from Crpyrope garnet xenocrysts in volcanic rocks. J. Geophys. Res., 101:5611–5625. Ryburn, R.J., Råheim, A. & Green, D.H. (1976): Determination of the P, T paths of natural eclogites during metamorphism – record of subduction. Lithos, 9:161–164. Schliestedt, M. (1980): Phasengleichgewichte in Hochdruckgesteinen von Sifnos, Griechenland. Doctoral Thesis, Technische Univ. Braunschweig, Gemany, 142 p. Schmid, R. (2001): Geology of ultrahigh-pressure rocks from the Dabie Shan (Eastern China). Doctoral Thesis, Univ. Potsdam, Gemany, 141 p. Schmid, R. Wilke, M., Oberhänsli, R., Janssens, K., Falkenberg, G., Franz, F. & Gaab, A. (2003): MicroXANES determination of ferric iron and its application in thermobarometry. Lithos, in press. Schmidt, M.W. (1993): Phase relations and compositions in tonalite as a function of pressure: an experimental study at 650 °C. Am. J. Sci, 293:1011–1060. Schmädicke, E. & Müller, W.F. (2000): Unusual exsolution phenomena in omphacite and partial replacement of phengite by phlogopite + kyanite in an eclogite from the Erzgebirge. Contrib. Mineral. Petrol., 139:629–642. Schuhmacher, R. (1991): Compositions and phase relations of calcic amphiboles in epidote- and clinopyroxene-bearing rocks of the amphibolite and lower granulite facies, central Massachusetts, USA. Contrib. Mineral. Petrol., 108:196–211.

258

E.J. Krogh Ravna & J. Paquin

Seitz, H.-M., Altherr, R. & Ludwig, T. (1999): Partitioning of transition elements between orthopyroxene and clinopyroxene in peridotitic and websteritic xenoliths: New empirical geothermometers. Geochim. Cosmochim. Acta, 63:3967–3982. Sharp, Z.D., Essene, E.J. & Smyth, J.R. (1992): Ultra-high temperatures and pressures from oxygen isotope thermometry of a coesite-sanidine grospydite. Contrib. Mineral. Petrol., 112:358–370. Shatsky, V.S., Sobolev, N.V. & Vavilov, M.A. (1995): Diamond-bearing metamorphic rocks of the Kokchetav massif (Northern Kazakhstan). In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 427–455. Smith, D.C. (1977): Aluminium bearing sphene in eclogites from Sunnmøre (Norway) (abstr.). Geolognytt, 10:32–33. Smith, D.C. (1980): Highly aluminous sphene (titanite) in natural high pressure hydrous-eclogite facies rocks from Norway and Italy, and in experimental runs at high pressure (abstract). In 26th Int. Geol. Congr., Paris, Abstr., sect. 02.3.1, 145. Smith, D.C. (1981): The pressure and pressure dependence of Al-solubility in sphene in the system Ti-Al-CaSi-O-H. In Progr Exp. Petrol. NERC Publ. Ser., D–18:193–197. Smith, D.C. (1984): Coesite in clinopyroxene in the Caledonides and its implications for geodynamics. Nature, 310:641–644. Smith, D.C. (1988): Eclogites and eclogite facies rocks. Amsterdam: Elsevier, 524 p. Smith, D. (1999): Temperatures and pressures of mineral equilibration in peridotite xenoliths: Review, discussion, and implications. In Fei, Y., Bertka, C.M. & Mysen, B.O. (eds.): Mantle petrology: Field observations and high-pressure experimentation: A tribute to Francis R. (Joe) Boyd /Geochem. Soc. Spec. Publ., 6/. St. Louis (Mo.): Geochem. Soc., 171–188. Smith, D.C. & Lappin, M.A. (1982): Aluminium- and fluorine-rich sphene and clino-amphibole in the Liset eclogite pod, Norway. Terra Cognita, 2:317. Smyth, J.R. (1980): Cation vacancies and the crystal chemistry of breakdown reactions in kimberlitic omphacites. Am. Mineral., 65:1185–1191. Smyth, J.R. & Hatton, C.J. (1977): A coesite-sanidine grospydite from the Roberts Victor kimberlite. Earth Planet. Sci. Lett., 34:284–288. Sobolev, N.V. (1977): Deep-seated inclusions in kimberlites and the problem of the composition of the upper mantle. Washington (D.C.): Am. Geophys. Union, 279 p. (English tansl. of the Russian ed., 1974, Novosibirsk: Nauka) Sobolev, N.V. & Shatsky, V.S. (1990): Diamond inclusions in garnets from metamorphic rocks: a new environment for diamond formation. Nature, 343:742–746. Sobolev, N.V., Kuznetsova, I.K. & Zuyzin, N.I. (1968): The petrology of grospydite xenoliths from Zagodochnaya kimberlite pipe in Yakutia. J. Petrol., 9:253–280. Sobolev, N.V., Lavrent’yev, Y.G., Pokhilenko, N.P. & Usova, L.V. (1973): Chrome-rich garnets from the kimberlites of Yakutia and their paragenesis. Contrib. Mineral. Petrol., 40:39–52. Sobolev, V.N., McCammon, C.A., Taylor, L.A., Snyder, G.A. & Sobolev, N.V. (1999): Precise Mössbauer milliprobe determination of ferric iron in rock-forming minerals and limitations of electron microprobe analysis. Am. Mineral., 84:78–85. Spear, F.S. (1993): Metamorphic phase equilibria and pressure-temperature-time paths /Mineral. Soc. Am., Monogr. Ser./. Washington (D.C): Mineral. Soc. Am., 799 p. Stosch, H.-G. (1987): Constitution and evolution of subcontinental upper mantle and lower crust in areas of young volcanism: Differences and similarities between the Eifel (F.R. Germany) and Tariat depression (central Mongolia) as evidence by peridotite and granulite xenoliths. Forschr. Mineral., 65:49–86. Tabata, H., Maruyama, S. & Shi, Z. (1998): Metamorphic zoning and thermal structure of the Dabie ultrahighpressure/high-pressure terrane, central China. Isl. Arc, 7:142–158. Taylor, W.R. (1998): An experimental test of some geothermometer and geobarometer formulations for upper mantle peridotites with applications to the thermobarometry of fertile lherzolite and garnet websterite. Neues Jahrb. Mineral., Abh., 172:381–408. Terry, M.P. & Robinson, P. (2001): Evidence for supersilicic pyroxene in an UHP kyanite eclogite, western gneiss region, Norway. In Eleventh Annu. V.M. Goldschmidt Conf., Abstr., Houston (Fl.): Lunar and Planetary Inst., 3842 (CD-ROM)

Thermobarometric methodologies

259

Terry, M.P., Robinson, P. & Ravna, E.J.K. (2000): Kyanite eclogite thermobarometry and evidence for thrusting of UHP over HP metamorphic rocks, Nordøyane, Westen Gneiss Region, Norway. Am. Mineral., 85:1637–1650. Troitzsch, U. & Ellis, D.J. (1999): The synthesis and crystal structure of CaAlFSiO4, the Al-F analog of titanite. Am. Mineral., 84:1162–1169. Troitzsch, U. & Ellis, D.J. (2002): Thermodynamic properties and stability of AlF-bearing titanite CaTiOSiO4CaAlFSiO4. Contrib. Mineral. Petrol., 142:543–563. Tsai, C-H. & Liou, J.G. (2000): Eclogite-facies relics and inferred ultrahigh-pressure metamorphism in the North Dabie Complex, central-eastern China. Am. Mineral., 85:1–8. Wain, A.L. (1997): New evidence for coesite in eclogite and gneisses: Defining an ultrahigh-pressure province in the Western Gneiss Region of Norway. Geology, 25:927–930. Wain, A.L. (1998): Ultrahigh-pressure metamorphism in the Western Gneiss Region of Norway. D. Phil. thesis, Univ. of Oxford, United Kingdom. Wang, X., Liou, J.G. & Maruyama, S. (1992): Coesite-bearing eclogites from the Dabie Mountains, Central China: Petrogenesis, P-T paths, and implications for regional tectonics. J. Geol., 100:231–250. Waters, D.J. & Martin, H.N. (1993): Geobarometry of phengite-bearing eclogites. Terra Abstr., 5(1):410–411. Waters, D.J. (1996) The Garnet-Cpx-Phengite Barometer. Recommended calibration and calculation method, updated 1 March 1996. http://www.earth.ox.ac.uk/~davewa/research/ecbar.html Wells, P.R.A. (1977): Pyroxene thermometry in simple and complex systems. Contrib. Mineral. Petrol., 62:129–139. Wood, B.J. & Banno, S. (1973): Garnet-orthopyroxene and orthopyroxene-clinopyroxene relationships in simple and complex systems. Contrib. Mineral. Petrol., 42:109–124. Ye, K. & Ye, D. (1996): Significance of phosphorus (P)- and magnesium (Mg)-bearing high-Al titanite in high pressure marble from Yangguantum, Rongcheng County, Shandong Province. Chin. Sci. Bull., 41:1194–1197. Zhang, R.Y. & Liou, J.H. (1998): Ultrahigh-pressure metamorphism of the Sulu terrane, eastern China: A prospective view. Cont. Dyn., 3:3–53.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 9, 261–306

Coronitic reactions: Constraints to element diffusion during UHP metamorphism MARCO RUBBO* and MARCO BRUNO Dipartimento di Scienze Mineralogiche e Petrologiche, Università di Torino, Via Valperga Caluso 35, 10125 Torino, Italy; * e-mail: [email protected]

Introduction When nucleation and growth of minerals in rocks occur, the system as a whole approaches a lower energy state. At given temperature, pressure and composition, the stable state corresponds to the minimum value of the rock Gibbs function. However, a description in term of equilibrium is not appropriate to understand the genesis of spatially ordered dispositions of minerals, non-equilibrium minerals morphology, zonation, and all features preceding equilibrium. Very commonly the kinetics of the transformation is frozen in, and disequilibrium textures beautifully show up. These situations must be described in terms of flow of components driven by the gradients of chemical potentials. For instance when nucleation occurs, the components’ chemical potential differences are determined by the local associations of minerals in different parts of the rock (Fisher, 1973, 1977). The relative rates of intergranular diffusion and of mineral growth and dissolution determine the steepness of these gradients (Fig. 1).

Fig. 1. Schematic chemical potential (
i) gradients around a growing crystal for elements with very slow, slow, medium, and fast rates of intergranular diffusion. From Carlson (2002), modified.

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

262

M. Rubbo & M. Bruno

If the rate of intergranular diffusion controls the nucleation and growth of the new phases a spatial organisation of reactant and product minerals results. Discussion on this subject can be found in Fisher (1973, 1977) and references therein, where linear nonequilibrium thermodynamics is used to tackle the situations of disequilibrium between growing minerals and matrix. When disequilibrium prevails at thin section level, we can have access to a fragment of the metamorphic history of the rock from which the section was obtained. But what is measured is a consequence of an ensemble of kinetic processes, among which the slowest are recorded. Yet to have access to the history, models are needed in order to describe segregation of elements in minerals, inter- and intracrystalline element diffusion, growth and dissolution rates. The models are built from experiments and physical laws. If we had a complete model we could fully exploit the information from a thin section, building a bridge from measured quantities to pressure, temperature and time. The ideal situation is not the actual one and different approaches are used, taylored on the problem at hand. Some approaches couple a petrographical description of UHP rocks to petrogenetic grids, but the kinetics is somewhat overlooked. We choose to describe a few kinetic models in some detail, because they can be suitable for an analysis of rock textures formed in a great interval of temperature and pressure values. The examples refer to relatively simple textures allowing a more didactic presentation. So this chapter aims a) to illustrate some models proposed to explain the formation of textures frequently observed in rocks, including the UHP metamorphic ones; b) to report two applications of the diffusion equation to retrive relevant information from zoned minerals; c) to show how equilibrium thermodynamics can help i) understanding disequilibrium features at thin section scale and ii) to recognise ultrahigh pressure mineral associations; this can be a delicate task when associations are stable over a wide interval of pressure values.

Kinetic theory Nowadays the study of the rate of mineral transformations has a great importance in earth sciences. It lies at the intersection of many fields: chemical kinetics, transport theory, nucleation and crystal growth, structural phase transitions, dislocation theory, etc. and it benefits of the progress of the knowledge of the behaviour of organic and inorganic materials which are not minerals. A powerful phenomenological description of the processes involved in mineral transformation is furnished by the linear thermodynamics of irreversible processes. It also gives an understanding of diffusion deeper than Fick’s first and second laws alone. Moreover it can describe the coupling between transport phenomena such as thermal diffusion, where a concentration gradient sets in an initially homegeneous solution due to a temperature gradient. The literature on irreversible thermodynamics is vast: several references (De Groot, 1951; De Groot & Mazur, 1962; Fitts, 1962; Katchalsky & Curran, 1965; Prigogine, 1968; Haase, 1969; Kaiken, 1994) are suggested to the interested readers. Some notions of irreversible thermodynamics are recalled in Appendix because several subjects reviewed in this chapter are based on it.

Coronitic reactions in UHPM

263

Growth of mineral layers A typical situation often observed is the metasomatic reaction of two-mineral assemblages which produce a layered sequence of reaction products, sometimes arranged in concentric shells. The treatment of metasomatic processes dates back to the early works by Korzhinskii (1959), Thompson (1959), Fisher (1973, 1977). A review was made by Rubie (1990). We follow the treatment and notation of Joensten (1977) and its subsequent evolution due to Nishiyama (1983) and Johnson & Carlson (1990). A rich bibliography can be found in the paper by Joensten (1977); from this paper we take the following model of the genesis of a reaction band. The formation of mineral layers is often associated to complex texture, owing to the sluggishness of transport processes, such as the formation of kelyphite and symplectite (Messiga & Bettini, 1990). The model we are going to illustrate revealed useful insights in these complex cases as well. The case studied has been selected for its simplicity and because it clearly illustrates the meaning of local equilibrium and diffusion limited growth. Although the example refers to low pressure rocks the model is applicable when (constant) temperature and pressure determine diffusion rates lower than rates of mineral nucleation and growth/ dissolution and ordered sequences of mineral layers are observed. However if the minerals within a layer are zoned or some reactions are sluggish this model becomes inaccurate.

Fig. 2. Stable minerals in the system CaO – AlO1.5 – SiO2 – CO2: The minerals in (b) are stable at higher temperature and lower pressure than in (a). Block diagram in (b) shows the sequence of mineral layers that results from reaction between blocks of calcite and anorthite placed in contact in (a). From Joensten (1977), modified.

264

M. Rubbo & M. Bruno

Let us consider calcite CaCO3 adjacent to anorthite CaAl2Si2O8. Reaction bands can develop as a consequence of increasing temperature. A reaction band is a layer of product minerals separating the reactant ones, as symbolically drawn in the block diagram of Figure 2. The mineral facies in the four components system CaO – AlO1.5 – SiO2 – CO2 are seen in Figure 2. When temperature increases calcite (C) and anorthite (A) are incompatible and react with one another (Fig. 2). The assemblage C-A will initially lie at some point on the metastable line, intersection of the two potential surfaces. If the kinetics is controlled by grain boundary diffusion a reaction band is produced. Once CaSiO3 wollastonite (W) and Ca2Al2SiO7 gehlenite (G) nucleate, there may be a transient of reaction controlled growth. But growth becomes diffusion controlled as soon as W and G become large enough. The dissolution–precipitation kinetics being fast in comparison to diffusion, local equilibrium among minerals and intergranular fluid sets in. Chemical potentials, in the grain boundary where three phases meet, will shift to the invariant point. Chemical potential gradients, pinned between the two three phases invariant points (CW + G and W + GA), exist across the band of product minerals. The calculated values (Joensten, 1977) of the chemical potential differences of the mobile components, at the invariant associations are (Fig. 3):

Fig. 3. Projection of saturation surface along
CaO axis, onto
AlO –
SiO2 plane. Chemical potentials for 1027 ºC and 350 bars. From Joensten (1977), modified. 1.5

Coronitic reactions in UHPM

265

W + GA CW + G
CaO –
CaO = – 1070 cal/mole CW + G
AlWO+1.5GA –
AlO = + 535 cal/mole 1.5


SiWO+2GA –
SiCOW2 + G = + 1070 cal/mole In response to the gradients, mobile components diffuse between the reactant phases. Summarising, following Johnson & Carlson (1990), the Joensten assumptions are:  the system evolves close to equilibrium so that the linear phenomenological laws apply between driving forces and fluxes;  the local equilibrium between diffusing components and adjacent minerals is set in: then the Gibbs–Duhem equation for every mineral constrains the components’ chemical potentials in the solids and in the intergranular medium through the equilibrium condition
imineral =
imedium;  the system is set in a state of minimal rate of entropy production so that component concentrations are time independent and the flow of matter by diffusion and the reaction rates are exactly balanced (De Groot & Mazur, 1962), in this case at each reaction front;  chemical reactions occur only at the layer contacts but not within layers;  diffusion occurs within the mineral layers but non within the minerals;  the metamorphism is isochemical. The isochemical constraint is not essential to the model and in Johnson & Carlson (1990) it is removed. The isochemical steady-state diffusion model requires simultaneous solution of a set of four types of linear equations describing reactions and fluxes in the growing corona. The unknowns are c + p : c are the stoichiometric coefficients of system components and p those of the reacting phases (phase components) at every interface. The diffusing components are related, through the boundary conditions of the steady state diffusion equations, to the fluxes, in the grain boundary, within the corona (or reaction band). In the case under consideration, at the interface C  W + G there are 3 phases, four components, three of them are mobile.  Mass-balance equations are required for each chemical component at every interface. In the case we are considering, the interfaces, indicated by a vertical bar (), are: CW + GA. The diffusing components are: CaO, AlO1.5, SiO2 while it is supposed that no CO2 gradients exist. CO2 is evolved at the boundary of the reaction band with calcite and its concentration becomes homogeneous after a transient which is short, compared with the formation time of the banded structure. p

iCW + G =

 CW + Gni =1

iW + GA =

 W + GAni =1

(1)

p

(2)

where iXY is the number of mole of component i consumed (if it is negative), or evolved in the reaction at the contact XY; XY is the amount in moles of phase  involved in the reaction and ni the stoichiometric coefficient i in the formula of phase .

M. Rubbo & M. Bruno

266

So for instance, for the component CaO, the equation at the boundary CW + G reads: CW + G W + GA – CaO = CCW + G + WCW + G + 2GCW + G and at the boundary W + GA: CaO =

AW + GA + WW + GA + 2GW + GA. The opposite sign in the equations is a consequence of the closed system CW + G W + GA assumption (CaO + CaO = 0). There are 4 equations at the boundary CW + G and 3 at the boundary W + GA, in this example.  Steady diffusion equations Within the layers no chemical reaction occurs and JiW + G is constant (Nishiyama, 1983). – The amount of each component entering into or leaving the intergranular medium at each interface depends only on the fluxes of that component on either side of the interface: –

iCW + G = JiW + G – JiC, 

W + GA i

=–J

W+G i

(3) A i

+J .

(4)

The phases C and A are homogeneous and components do not enter or escape outside the reaction fronts so it is JiC = JiA = 0. If the gradient of chemical potential of component i is negative in going from calcite to anorthite, that component enters the corona W + G at the interface CW + G and it leaves the corona at W + GA. It is considered JCO2 = 0 everywhere. These conditions are in fact boundary conditions for the flux in the mineral layers and some comments are needed. It is assumed that the layers growth rate is negligible and the mass conservation equations are written in a frame fixed on the layer contact CW + G. At the steady state the mass conservation of component k in a volume around the interface CW + G is (Nishiyama, 1983):   Jk  ck  –— dV = 0 =   – –—   x  t

  + k –— Adx t

(5)

Equation 5 is obtained integrating, over a unit volume, the equation of conservation of mole d  matter with the term k –— –————— corresponding to the production of dt volume time component k by chemical reaction. After a further integration of Equation 5, on a unit time, we obtain: (Jkout – Jkin) = k.

(6)

Defining a reference state such that  = 1 (see last point) the conditions (3) and (4) are obtained.  Flux ratio equations. Within the layer W + G the fluxes are related to the chemical potential gradients by:

Coronitic reactions in UHPM d
iW + G JiW + G = – Lii –——— . dx

267

(7)

Note that cross coupling coefficients (Lij = 0) are neglected and Lii is assumed independent on composition for all components. Only c – 1 fluxes are independent ( Ji = 0), so there are c – 1 equations. Because the individual Lii are not known it is usual to write flux ratio W+G equations (in the following JSiO is the common divisor): 2 W+G W+G JCaO LCaCa d
CaO –—— –—— –—— = W+G W+G , JSiO LSiSi d
SiO 2 2

(8)

W+G W+G JAlO LAlAl d
AlO 1,5 1,5 –—— ––—— –—— = W+G W+G . JSiO d
L SiO SiSi 2 2

(9)

– Gibbs–Duhem equations. These are p equations, one for every phase in the reaction band, relating the chemical potential of the c components. c The general form is dG =  id
i = 0. For the reaction band W + G: W W d
CaO + d
SiO =2 0,

2d


G CaO

+ 2d


G AlO1.5

+ d


G SiO2

(10)

= 0.

(11)

The equilibrium betweeen minerals and intergranular fluid requires: W+G d
WCaO = d
GCaO = d
CaO ,

(12)

W+G . d
WSiO2 = d
GSiO2 = d
SiO 2

(13)

Because we are interested in c – 1 ratios of chemical potential gradients and there are p equations among them, c – 1 – p are the free ratios. Out of the c – 1 flux ratio equations c – 1 – (c – 1 – p) are independent. So there are p = 2 independent flux ratio equations in the band W + G.  one extent of reaction is fixed. In conclusion there are c + p unknowns and c – 1 + p equations among stoichiometric coefficients of system components and of phase components, because only c – 1 fluxes are independent. A coefficient must be fixed by one reaction extent at one boundary. In more general cases one extent-ofreaction equation is needed for each equilibrium boundary that has been overstepped before reaction begins (Johnson & Carlson, 1990). In the case we are dealing with, a single reaction has been overstepped. The following equations describe the reaction at the interface CW + G: CW + G CO + CCW + G = 0 2



+



+ 2

CW + G CaO CW + G AlO 1.5

CW + G C

(14)

+

CW + G G

CW + G W

+ 2

=0

SiCOW2 + G + WCW + G + GCW + G = 0

CW + G G

=0

(15) (16) (17)

M. Rubbo & M. Bruno

268

CW + G W+G CaO LCaCa d
CaO –——— C W + G = –—— –—— W+G SiO2 LSiSi d
SiO2

(18)

CW + G W+G AlO LAlAl d
AlO 1.5 i 1.5 ––—— C W + G = –—— ––—— W+G SiO2 LSiSi d
SiO 2

(19)

LCaCa, LAlAl, LSiSi are constants.

(20)

One extent of reaction is fixed: CW + G CaO ==1

(21)

The steady diffusion equations lead to: CW + G W + GA CaO + CaO =0

(22)



=0

(23)

SiCOW2 + G + SiWO+2GA = 0

(24)

CW + G AlO 1.5

+

W + GA Al O1.5

The following equations describe the reaction at the interface W + GA: W + GA CaO + WW + GA + 2GW + GA + AW + GA = 0

(25)



(26)

W + GA Al O1.5



W + GA SiO2

+

W + GA W

+

W + GA W

+ 2

W + GA G

+

W + GA G

+ 2

W + GA A

+ 2

W + GA A

=0

=0

(27)

Indicating by M, X, B the matrices of the system coefficients, of the unknowns and known terms, the equations can be written in condensed matrix form MX = B: 1 0  0  0  0  0 0  0  0  0 0  0  0

0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 1 1 2 0 0 0 0 0 0  0 1 0 0 0 2 0 0 0 0 0 0  0 0 1 0 1 1 0 0 0 0 0 0  1 0 1 0 0 0 0 0 0 0 0 0  0 1 –.5 0 0 0 0 0 0 0 0 0  1 0 0 0 0 0 0 0 0 0 0 0  1 0 0 0 0 0 0 0 0 1 0 0  0 1 0 0 0 0 0 0 0 0 1 0  0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 2 1 1 0 0  0 0 0 0 0 0 0 2 2 0 1 0  0 0 0 0 0 0 1 1 2 0 0 1

CW + G  CO2 0    CW + G  0  CaO      CW + G  0 AlO 1.5      SiCOW2 + G  0  CW + G     C 0   CW + G     W 0  CW + G   = 1  G     WW + GA  0  W + GA     G 0   W + GA     A 0    W + GA  0  CaO     AlWO+ GA  0 1.5  W + GA      SiO2  0

(28)

Coronitic reactions in UHPM

269

W+G W+G d
AlO d
CaO 1 1.5 After substituting in the appropriate equations the values ––—— W+G = — and –—— W+G = – 1 d
SiO2 d
SiO2 2 LAlAl obtained from the system (10-11), and for the tentatively assigned values –—— = LSiSi LCaCa = –—— =1, the solution (29) of the linear system (28) allows to write down the reactions LSiSi

(Joensten, 1997):  at the boundary CW + G: 2.25CaCO3 + 0.5AlO1.5 + SiO2  0.25Ca2Al2SiO7 + 0.75CaSiO3 + CaO + 2.25CO2  at the boundary W + GA: 1.125CaAl2Si2O8 + CaO  0.875Ca2Al2SiO7 + 0.375CaSiO3 + 0.5AlO1.5 + SiO2 The components produced in the reaction at CW + G are quantitatively consumed at the boundary W + GA and vice-versa. CW + G  2.2500   CO2   1.0000    CW + G    CaO    –0.5000    CW + G    AlO1.5    –1.0000   SiCOW2 + G    CW + G    –2.2500   C    CW + G    W  0.7500     CW + G  =  0.2500   G     WW + GA   0.3750   W + GA     G  0.8750    W + GA     A  –1.1250     W + GA   –1.0000   CaO     AlWO+ GA   0.5000   W +1.5GA      SiO2   1.0000 

(29)

The cycle is reported in Figure 4. With the assumed values of the ratios of the phenomenological coefficients, wollastonite and gehlenite are produced at both contacts and then the reaction band W + G is stable. With different values of these ratios other sequences of reaction bands are stable. The modal proportion of wollastonite and gehlenite produced by reaction at each contact is different and changes in a stepwise manner across the layer: the observation of this steep variation is a check of the constancy of the flux ratios. A conclusive remark: the diffusion controlled reactions at

270

M. Rubbo & M. Bruno

Fig. 4. Closed cycle of diffusion controlled reactions exchanging CaO, AlO1.5 and SiO2 between the isothermal-isobaric invariant assemblages C + W + G and W + G + A, resulting in growth of the W + G layer at the expense of C and A. From Joensten (1977), modified.

the boundary of the layers are nicely tuned and make a closed reaction cycle. The observed products within the reaction band indicate that the gradient of CO2 is absent. In the contrary case we should observe assemblages reflecting the variations of the chemical potential of carbon dioxide along the reaction band. Therefore, in the Joensten model the diffusional control is exerted by fewer components than required to describe the thermodynamic equilibrium as remarked by Johnson & Carlson (1990) who explicitly calculated the number of components controlling diffusion as a function of the observable structure of the reaction layers. If minerals are mixtures of end-members, when the rate of a reaction of the cycle changes with temperature and/or composition, not only the products’ modal distribution is affected but also their composition. In principle one should tackle the full diffusion-reaction problem in the network of the grain boundary, as well as in the growing and resorbing phases. The constraints on element migration will be used in the following section to model this problem in an approximate way.

Symplectic reaction in olivine As a second example of applications of the thermodynamics of irreversible processes, we propose the symplectic reaction in olivine studied by Ashworth & Chambers (2000). To our knowledge, it is the only study dealing quantitatively with these structures. It points out that symplectites form when a modulated flux sets in a direction parallel to the growth front. Examples of symplectite after eclogite are rewieved by Rubie (1990). The symplectites studied (Figs. 5–7) are from the Lilloise intrusion. The Lilloise is an 8  4 km layered mafic intrusion which cuts the plateau basalts of the East Greenland Tertiary province (Chambers & Brown, 1995). Lilloise was intruded at about 50 Ma, 4–5 Ma after cessation of the voluminous tholeiitic magmatism which accompanied the rifting

Coronitic reactions in UHPM

271

Fig. 5. Rod symplectites in troctolite from Belhelvie, northeast Scotland. Hornblende-spinel symplectite in which spinel (lighter) is the rod mineral. From Ashworth & Chambers (2000), modified.

of the East Greenland continental margin. The layered intrusion can be divided into three zones, the Lower, Middle and Upper, consisting of basal olivine-clinopyroxene cumulates (600 m) passing upwards into gabbroic olivine-clinopyroxene-plagioclase cumulates (1800 m) and then into a series of plagioclase-amphibole cumulates (400 m). The symplectites are found in olivine-chromo spinel cumulates (Lower zone) as thin platelets of clinopyroxene and magnetite parallel to (100) of olivine when it is less ferroan than Fo74. The symplectites are concentrated near the centres of unzoned olivine grains and tend to be associated with subgrain boundaries. For these reasons the authors think that the symplectites grew after some deformation, and in those grains where minor components were unable to diffuse out of the olivine grains. The symplectite growth is, therefore, referred to exsolution of minor components from olivine. The overall reaction can be represented by: 5(Fe,Ca,Mg)22+SiO4 + 3Fe 4– SiO4  4(Ca,Mg,Fe)2Si2O6 + 2Fe2+Fe23+O4. 3+ 3

Fig. 6. Symplectite of clinopyroxene and magnetite in Lilloise olivine. Optical micrographs, plane polarised light. Thin section cut parallel to (010) so that the symplectite platelets are viewed edgewise. From Ashworth & Chambers (2000), modified.

272

M. Rubbo & M. Bruno

The magnetite and clinopyroxene are formed from the reactants in solid solution in olivine. The previous reaction is not a redox one. The proportion of pyroxene over magnetite produced is 3 : 1 in terms of oxygen atoms and approximately in volume ratio. TEM observations indicate that the reaction is topotactic, both symplectite minerals having crystallographic orientations related to that of olivine and very a similar oxygen array (Moseley, 1984). A parental relation is found between close-packed planes in olivine and product minerals (Ashworth & Chambers, 2000). The reaction front as well as the grain boundaries are semicoherent. This implies that oxygen is an immobile component. So it is sensible to suppose that Ca, Fe, Mg can diffuse through the olivine and that Si is conserved in the reaction front. In the reaction front a flux of cations must originate between the crystallisation domains of clinopyroxene and magnetite. Theory of intergrowth spacing The model proposed by Ashworth & Chambers (2000) develops a work by Cahn (1959) from General Electric Research Laboratory, on the cellular segregation reactions. Consider a region made of a sequence of parallel lamellar phases ,  produced by the transformation of a matrix phase m (of the reactant minerals) (Fig. 8). The composition of the transformed region is equal to that of the untransformed matrix. It is made by cells whose repeating period is the lamellar spacing. Within a cell all the crystals of one type are approximately parallel and have the same lattice orientation. Concepts such as nucleation and growth rate may be applied to the cell as a whole, as well as to individual crystals within the cell. As the cell grows, the individual plates maintain an approximately constant spacing so that branching and/or nucleation of new plates must occur. It is appropriate to distinguish sideways growth, which requires the production of new lamellae by nucleation and/or branching, and edgewise growth, which is the extention of the cell at the growth front towards the matrix (Christian, 1975). Ashworth & Chambers (2000) are concerned with edgewise growth.

Fig. 7. Symplectites of clinopyroxenes and magnetite in Lilloise olivine. Optical micrographs, plane polarized light. A section parallel to the platelet plane (100) of olivine. From Ashworth & Chambers (2000), modified.

Coronitic reactions in UHPM

273

Fig. 8. Geometry of the growing symplectite:  is the recurrence distance of the lamellae of  and  phases;  is the thickness of the diffusion layer between lamellae and the reactant minerals of the matrix. P is the volume proportion of . From Ashworth & Chambers (2000), modified.

However, cases are known where the rates of sideways and edgewise growth are similar and the lamellar spacing is a sensitive function of the temperature. The situation we are considering is the one in which the solute segregation depends only on diffusion within a layer, thickness , at the interface between cells and matrix (Ashworth & Chambers, 2000). Ci and Ci are the uniform concentrations of component i in phases  and , respectively. Cim is the matrix composition up to the reaction front: transfer of matter from the matrix to the  layer determines the growth rate V of the cells. If the diffusion in the matrix were sluggish, Cim should be calculated using the appropriate boundary conditions. Concentration gradients set in the  layer, because the components’ concentration in ,  are different. Diffusion flow transfers atoms parallel to the interface between cells and matrix. At a fixed temperature, pressure and mineral composition, the affinity of the overall reaction is –G, a constant. It is made of several contributions. At first suppose a closed system whose boundary advances by z. The Gibbs function increases by

lz lz –—— Grf = Gdif –—— + 2lz, v v

(30)

where – Grf is the affinity per mole of the reaction at the reaction front, –Gdif is the affinity of the diffusion at the reaction front, lz is the new interface area formed between  as a consequence of an advancement z of the growth front,  the cell spacing, v the mean molar volume of the simplectite and  the interface energy. Then, Grf > 0 if Gdif does overcome a critical value v G*dif = – –—— . 

(31)

M. Rubbo & M. Bruno

274

v That is, –—— is the driving force –Ggb needed to form a unit area of a new interface  at the grain boundary . If the system is open the total affinity is: v –G = –Grf + (–Gext) = –Gdif + –—— + (–Gext). 

(32)

The contribution –Gext is due to transfer of components from the exterior (the matrix) towards the reaction zone. It is assumed that the growth rate edgewise depends linearly on the flux of matter from the exterior towards the reaction front of thickness , and because the fluxes are proportional to the affinity driving the symplectic growth, a linear relation holds between the affinity and the reaction front advancement velocity V: –Gext = cV,

(33)

c is a constant. To calculate Gdif we must know the diffusive flux, Ji, of the i components, parallel to the growth front and driven by difference in compositions between  and  phases. The two phases have a recurrence distance in the direction of the flux Ji so the flux has a periodicity (Fig. 9). The cell is composed by a host mineral (the thicker one, say ) and a lamella () whose volume proportion is P. Assuming that elements do not diffuse within the cell but in the reaction front only, taking the origin at half lamella, the components rejected from  (to be deposited in the host mineral ) have a maximal  concentration at n, n integer, and a minimal concentration at n — , at the middle of . 2

(2n + P) As a consequence the flux of the ith component, Ji, has a maximum at ———— 2 (2n + 1 – P) and a minimum at —————— . The Ji are linearised and represented by an 2 appropriate sequence of broken lines: by consequence in a slab thickness , parallel to the growth front, the current is dJi V(Cim – Ci) = ————— –—— dx 

mol   ––——

cm3sec

(34)

mol   ––——

cm3sec

(35)

between

  – P — ! x ! P — 2 2 and dJi V(Cim – Ci) = ————— –—— dx  between

Coronitic reactions in UHPM

275

  P — ! x !  – P — . 2 2  The variation of Ji in the interval 0 ! x ! P — , is obtained by integrating Equation 34 2 with the boundary condition x = 0, Ji = 0 V(Cim – Ci)x Ji = ————— . 

(36)

  The variation of Ji between P — ! x ! — is obtained by integrating Equation 35 with 2 2   the boundary condition x = — , Ji = 0 2 m   V(Ci – Ci )  Ji = ————— x – — (37)  2 .  According to Fick’s Law (neglecting cross coupling Dij coefficients): dCi Ji = – Di —— . dx

(38)

Fig. 9. Variation of flux (a) and concentration (b) with distance X in the reaction front. From Ashworth & Chambers (2000), modified.

 Using Equation 38, Equation 36 is integrated in the interval 0 ! x ! P — , with the 2 boundary condition Ci = Ci0 at x = 0, obtaining: V(Cim – Ci)x2 Ci = Ci0 – ————— . 2D

(39)

M. Rubbo & M. Bruno

276

  Equation 37 is integrated in the interval P — ! x ! — , with the boundary condition 2 2 (P)2 V(Cim – Ci) –——  4 given by (39): Ci = Ci0 – –——————— at x = P — , obtaining: 2D 2 V(Cim – Ci)P2 V(Cim – Ci) Ci = Ci0 – ——————— – ————— 8D 2D

2  2 — – x  – (1 – P)2 — .

 2  4

(40)

The last equation holds for “immobile” elements like Si and O. For elements diffusing from the matrix or along the interface (which is excluded here) the term containing P should be multiplied by a factor taking into account the material lost. Because Equation 40 will be used for closed reaction zone, we prefer to keep a ligther notation. We are describing the concentration and the fluxes as continuous functions of a space variable. In such a case the entropy production per unit volume and unit time is: 1 d
i "i = – — Ji —— T dx

(41)

Ji d
i —— =– — . Li dx

(42)

and

Summing up the contribution to "i due to diffusion at the growth front between 0 ! x !    P — , and between P — ! x ! — : 2 2 2 1 "i = —— [(Ji# + (Ji)$ TLi

(43)

or, with the appropriate substitutions: V 2(Cim – Ci)2x V 2(Cim – Ci)2( –2 – x) 1 ———————— "i = — ——————— + . Li2 Li2 T



(44)

 –2 In a volume   —  1 meter the entropy produced by diffusion is %0 "i dx. In the time 2 needed for the growth of a unit volume of symplectite 2 sec  t = —— —— V1 m3

the entropy produced is

Coronitic reactions in UHPM

277



– 2" joule i  2 ——  dx ——–  . 0 V

K

Multipling this quantity by the molar volume of symplectite v and by the temperature T we obtain the affinity driving the formation of one mole of symplectite by diffusion at the growth front. P  –— — 2 2 2vV  m m  2  2  2   x dx + (Ci – Ci ) ( — – x)2dx  (45) –Gi, dif = —— (Ci – Ci )  P   Li

0 2

–— 2 



After integration it we obtain: vV2 –Gi, dif = –—— [P3 (Cim – Ci)2 + (1 – P)3 (Cim – Ci)2] 12Li

(46)

Summing over the number of diffusing components and multiplying and dividing by LSi, for later convenience, one gets: & vV2 & LSi –Gdif = –—— '–——  P3  (Cim – Ci)2 + (1 – P)3  (Cim – Ci)2 ) ' (47) LSi ( 12Li i *( i The total affinity is obtained by summing all contributions: –G = –Gdif – Ggb – Gext ,

(48)

which can be cast in the form: a –G = bV2 + — + cV . 

(49)

In Equation 49 the correspondence is the following: –Gdif = bV2;

a –Ggb = — . 

(50)

The constants a, b are given by: a = 2v ,

(51)

& v & LSi m m  2 3  2 ) –——  3 (52) b = –—— LSi ' 12Li P i (Ci – Ci ) + (1 – P) i (Ci – Ci ) * ' . ( ( Equation 49 contains an implicit relationship between  and V. Equation 49 is third order in : it can be solved for fixed values of a, b, c, G and values of V increasing from zero to an upper limit. Beyond this limit no real solution for  can be found. If we assume that this limit represents the optimal value of the cell growth rate V* then the largest real root of the cubic equation is the optimal value of , for which –Gdif has the higher value. As V increases,  decreases because the faster transport of matter requires shorter diffusion distances; in turn shorter  values imply the increase

M. Rubbo & M. Bruno

278

of –Ggb. Finally –Gext increases while –Gdif decreases with increasing V so that –G is constant. Alternatively, it can be assumed that the rate of energy dissipation by transport within the reaction layer and exchange with the matrix is maximal. In this case the value of V must be found which makes a maximum V(–Gdif – Gext), whose expression is obtained using Equations 33, 47, 52: V(–Gdif – Gext) = V 2(b2 + c) .

(53)

Differentiating in respect to V the right-hand side member of Equation 53 and equating the result to zero, we obtain: c + b

 d— V +   = 0  dV 

(54)

d The expression of — is obtained by differentiating Equation 49 at constant G: dV d b2 + c — = ———— . a dV –— – 2bV 2

(55)



Substituting Equation 55 in Equation 54 and multiplying by V the resulting expression, we obtain: cV + bV2 cV + bV2 = bV –———— a 2bV – –— 2

.

(56)



So the maximum occurs when a –—3 = V . b

(57)

The relationship (57) between growth rate and spacing , holds when the rate of energy dissipation is maximal. It can be written as: a — = Vb2 . 

(58)

The condition of maximum in terms of G is (Ashworth & Chambers, 2000): (–Gdif) + (–Gext) (–Gext) + (–Gdif) = (–Gdif)  –———————— 

2(–Gdif) – (–Ggb)

(59)

which gives the relationship equivalent to Equation 58: (–Gdif) = (–Ggb)

(60)

in both cases: Gext = 0 and Gext + 0. It also indicates that half the affinity at the growth front drives diffusion and half is spent to increase the interface between host and lamella.

Coronitic reactions in UHPM

279

Equation 57 is the key one: it was used by Ashworth & Chambers (2000) to characterise the symplectites in olivine. This would be an easy task when a, b are known. Looking at Equation 51 and 52 we see that some quantities can be measured, some others like the reaction layer thickness , are related to the approximated description of the mechanism of the symplectic reaction, while the interface energy  and the phenomenological coefficients Lk must be known a priori. Unfortunately, this is not the case: researches are under way (an example will be seen in the next chapter) in order to know the relevant mineral and transport properties needed to exploit in full the constraints obtainable from rock texture and composition. To overcome these difficulties Ashworth & Chambers (2000) introduced some approximations and used independent estimates of the growth velocity V in order to obtain a lower bound to LSi. In particular, from geological information it is estimated V > 6  10–16 m3s–1; from experiments on olivine oxidation  ! 0.3 Jm–2; it is measured  , 4
m. Finally, considering the term containing LSi as the only one contributing to b, the sum in (52) reduces to: ( & ' 1 [P3 (CSim – CSi )2 + (1 – P)3 (CSim – CSi )2]' . & –— 12 ( Identifying  with magnetite, CSi = 0, CSim – CSi = 0. From Equation 57:



12–1P3 CSim mol2 LSi = V3 –———— , 10–25 –—— 2 Js

2

mol2

being 12–1P CSim , 2.8  106 –—— m3 as results from the measured P , 0.25 and the concentration of Si in olivine. There are experimental indications that the calculated value of LSi occurs in the temperature range 800–1000 ºC. It is worth to report some Authors’ considerations:  diffusion strictly parallel to the growth front can occur in fluid undersaturated systems;  high difference in composition between product minerals, particularly in the case of slow diffusion elements, favours symplectite formation. Final considerations: the work is based on the assumption that the product of the symplectite growth rate by the diffusion affinity is maximum, but from the thermodynamics of irreversible processes a minimum would be expected at the steady state: so more fundamental arguments should be desirable also to explain why a periodic concentration over the growth front is stable.

An estimate of intergranular diffusion of Al in fluid undersaturated systems Carlson (2002), using numerical simulations of coupled intergranular and intracrystalline diffusion processes in coronal textures around partially resorbed garnet crystals (Fig. 10), obtained a very precise estimate for the rate of intergranular diffusion of Al in the fluid-undersaturated system described in the following paragraph. The model is applied to partially resorbed garnets in mafic rocks of the Llano Uplift (Texas, USA).

280

M. Rubbo & M. Bruno

Fig. 10. Photomicrograph with polarised light (uncrossed polars) of layered coronal reaction zone. Relict garnet is at bottom of photo. Next layer upward is a symplectite of plagioclase and amphibole; its upper limit, marked by a solid white line, is the original boundary between the reactants garnet and omphacite. Above there is a layer of amphibole, with a narrow band of plagioclase near its centre; this borders on a brownish zone that comprises a symplectite of plagioclase, amphibole and orthopyroxene. The dashed line marks the boundary with a top zone of intergrown plagioclase and secondary sodic augite after original omphacitic clinopyroxene. From Carlson (2002), modified.

Actually these garnets are surrounded by layered coronal reaction textures and exhibit reversal intracrystalline diffusion profiles at their rims (Fig. 11). These textures are a consequence of a two-stage metamorphic event. In the first (prograde metamorphism), the garnets crystallised under conditions transitional among the amphibolite, granulite and eclogite facies, reaching peak temperature of about 750 ºC, sufficient to nearly homogenise original growth zoning profiles in all but the largest garnets (Carlson & Schwarze, 1997). In the second stage (retrograde metamorphism), recrystallisation took place at lower pressures and under static conditions that allowed development of (i) coronal textures between garnet and sodic pyroxene, and (ii) reversal zoning in garnet. The coronitic structure, shown in Figure 10, from garnet to sodic pyroxene, is made up by: symplectite of plagioclase and amphibole; layer of amphibole and plagioclase; symplectite of plagioclase, amphibole and orthopyroxene; intergrowth of plagioclase and secondary sodic augite after original omphacitic clinopyroxene. Such features are the result of the coupling of dissolution reactions and intracrystalline diffusion in garnets during retrograde metamorphism. The steepness of the compositional profiles near the garnet rim (Fig. 11), depends upon the relative rates of the dissolution reaction and the intracrystalline diffusion that results from it. The relative rates of these processes are assessed by a numerical simulation of multicomponent intracrystalline diffusion in garnet, based on the computation of garnet size variation and compositional changes in Fe, Mg, Mn, Ca in response to the variations of temperature during retrograde metamorphism. The Carlson (2002) simulation is based on the following assumptions. 1. The temperature of the rock is decreasing, at constant pressure of 3 kbar, according to the linear equation:

Coronitic reactions in UHPM

281

Fig. 11. Representative fits of modelled diffusion profiles to measured compositions in relict garnet. Circles are microprobe analyses; dotted lines: extrapolations used to estimate original profiles; solid lines: the results of the numerical simulation. From Carlson (2002), modified.

T(t) = T0 – 10t [ºC]

(61)

where T0 is the initial temperature, namely the temperature at which coronitic reaction begins and the rate is ºC expressed in –— my . 2. The garnet is a sphere whose initial radius (before dissolution) is considered to be the upper limit of plagioclase and amphibole symplectite (Carlson & Johnson, 1991). It is shown in Figure 10 by the solid white line. 3. The initial element concentration profile of the garnet is obtained by extrapolating the supposed unaffected composition of the interior portion of the crystal to the portion of the garnet dissolved (dotted line in Fig. 11). 4. The garnet dissolution rate is limited by diffusion of Al across the product zone. Therefore, diffusional control of the reaction kinetics justifies the assumption of the Arrhenius law to determine the rate of change of garnet volume dV : — dt dV dV –QIGD   –—  = –—  exp  ——— dt dt  RT(t) 

t=0

(62)

where QIGD is the activation energy for intergranular diffusion of Al: from a previous work its value results QIGD = 140

kJ –— . Because the integral of Equation 62 mol

over the dissolution interval must equal the total volume of garnet resorbed, the value of

dV [ –— dt ]

t=0

dV  –—  =

dt t=0

is determined by the volume loss of the crystal V: V —————— , t25 IGD dt ——— %0 exp –Q RT(t)

(63)

where t25 is the time at which T = 25 ºC, i.e. the time at which dissolution is considered to be complete.

M. Rubbo & M. Bruno

282

The simulation is iterative. Given an initial temperature T0 and the thermal history described by Equation 61, the garnet (with radius and composition determined as explained in points 2 and 3) is resorbed by the amount required by Equation 62, and the time for each volume decrement is computed from (61). For each increment of dissolution, the flux of material into or out of the rim of the remaining crystal is determined by a retention/loss factor. The lattice diffusion in garnet is computed solving the multicomponent diffusion equation, adapting values of the diffusion matrix from Chakraborty & Ganguly (1992) and accounting for their composition dependence following Lasaga (1979). The dissolution/diffusion cycle is repeated until the radius of the model crystal matches that of the relict crystal. The time tf resulting from this procedure gives the time required to resorb 95% of the measured volume lost by garnet. Iterative calculations are required to determine the retention/loss factors. To begin, an initial trial value for each element (Fe, Mg, Mn, Ca) is arbitrarily chosen. For instance, if at the end of the simulation the crystal contains an excess of one element with respect to the total observed amount, then for the next simulation a smaller retention factor is considered.1 Now, it is possible to calculate the effective intergranular diffusivity (strictly not the grain boundary diffusion coefficient because an intergranular medium in which Al could be soluble was possibly present) of Al (DAl eff) by equating the measured length scale for intergranular diffusion in the coronal textures with the characteristic diffusion distance, travelled by diffusing atoms: t

 25 –QIGD ——— x =  Deff dt -Al exp RT(t) 0 2

(64)

The value of x is taken as the distance between the surface of the reactants (between the edge of residual garnet and the dashed line in Fig. 10) in the corona when the reaction ceased. Fourteen compositional profiles were modelled by Carlson (2002), selected from specimens that represent a wide range of possible dissolution histories, in order to asses the uncertainties that arise from natural variations. But all fourteen modelled profiles yielded very similar results, indicating that the approach is precise. Fitting was done by adjusting only the model parameters T0 (655–686 ºC); all other parameters are identical in all 14 fits. The estimated intergranular diffusivity for Al, resulting from this procedure, is: kJ m2  –140 [— eff –13 mol]  –— D./ exp  ———— (65) Al = 1.0  10  RT  s

This equation is only applicable to hydrous but fluid undersaturated systems and in the range of temperature encompassed in these models (500–650 ºC).

1

Although in the paper there are no further “technical” indications about the retention/loss factors, we believe that they are a parameterisation of the boundary conditions for the diffusion equation in garnet, in such a way to fix the relative fluxes of each cation in or out the resorbing garnet.

Coronitic reactions in UHPM

283

Fig. 12. Plagioclase domain (site P). Aggregate of Ab + Jd + Zo + Ky + Kfs + Qtz, pseudomorphically replacing the original igneous plagioclase.

The ultrahigh pressure coronitic reactions in a metagranodiorite Geology and petrography Metagranodiorite samples from the Brossasco-Isasca Unit (Biino & Compagnoni, 1992; Bruno et al., 2001), Dora-Maira Massif, western Alps, show pseudomorphous and coronitic textures where igneous minerals were partially replaced by ultrahigh pressure (UHP) metamorphic assemblages. The Brossasco-Isasca Unit (BIU) is a slice of Variscan continental crust recrystallised under UHP metamorphic conditions during the Alpine orogeny (Henry, 1990; Chopin et al., 1991; Compagnoni et al., 1994, 1995; Compagnoni & Rolfo, 2003). The metamorphic peak is estimated by many workers at P = 33 0 3 kbar and T = 750 0 30 ºC (Chopin, 1984, 1987; Chopin et al., 1991; Kienast et al., 1991; Sharp et al., 1993; Compagnoni et al., 1994), and, recently, by Hermann (2003) at about 43 kbar and 750 ºC. The metagranodiorite originally consisted of quartz, plagioclase, K-feldspar, biotite and accessory apatite, zircon and a Ti-rich phase, most likely ilmenite. During the Alpine polyphase metamorphism, the igneous minerals were (i) replaced by polycrystalline aggregates of metamorphic minerals, (Site P; Fig. 12), (ii) reequilibrated to metamorphic compositions, e.g. biotite, or (iii) reacted and developed coronitic structures between biotite and adjacent minerals, e.g. between biotite and quartz (Site A, Fig. 13), biotite and K-feldspar (Site B, Fig. 14), and biotite and plagioclase (Site C, Fig. 15). Representative analyses of the phases are reported in Bruno et al. (2001). The plagioclase (Site P) is replaced by a polycrystalline aggregate of zoisite + jadeite + quartz/coesite + kyanite + K-feldspar to a pseudomorph (Fig. 12). Zoisite and kyanite are pure, and K-feldspar is low in albite (Or90Ab10). Na-pyroxene is a solid solution of jadeite, Ca-Tschermak and Ca-Eskola (Bruno et al., 2002), but usually it is

M. Rubbo & M. Bruno

284

partially replaced by retrograde albite or oligoclase. At the original igneous biotite–quartz contact (Site A), a single continuous corona of weakly zoned garnet, with composition Alm76–78Prp21–23Grs1–2Sps1–2, develops (Fig. 13). Usually, garnet is rimmed by a retrograde biotite. Between biotite (partially replaced by phengite I) and K-feldspar (Site B), the following composite corona is formed (Fig. 14):

Fig. 13. Site A: Biotite partly replaced by phengite. Note the development of a continuous garnet corona against quartz.

Table 1. Initial phase compositions for the three sites considered. The moles of phase components are normalised to 20 moles of cations. Phase

Phase components

Site A

Site B

Site C

Bt

Ann Phl Eas Mn-bt

0.220 0.150 0.120 0.010 14.00 0.030 0.160 0.010

0.220 0.150 0.120 0.010

0.088 0.060 0.048 0.004

Qtz Kfs

Pl

Ab Or An Ab Or An

0.600 2.390 0.010 2.448 0.072 1.080

 a continuous corona of weakly zoned garnet (Alm78–80Prp15–17Grs3Sps2);  a continuous corona of garnet (Alm78–80Prp17–19Grs2–3Sps0–1) with vermicular quartz inclusions, elongated perpendicular to the corona;  a continuous corona of a quartz-phengite (PhII) symplectite in continuity with the garnet-quartz symplectite. Between biotite and plagioclase (Site C), the following mineral sequence is observed (Fig. 15):

Coronitic reactions in UHPM

285

Fig. 14. Site B: Relict of igneous biotite included in K-feldspar. Biotite is partly replaced by phengite (PhI). Note the development of a composite corona of garnet, garnet+quartz and quartz+phengite (PhII).

Fig. 15. Site C: Composite corona developed at the original contact between igneous biotite and plagioclase. The corona consists of garnet + quartz, phengite (PhII), and jadeite + garnet. Biotite is completely replaced by phengite (PhII).

 a corona of garnet (Grs49–50Alm42Prp7–8Sps0–1) plus quartz;  a composite corona of idioblastic garnet (Grs62–84Alm16–32Prp0–6) and jadeite. From biotite towards plagioclase, the relative amount of garnet decreases and that of clinopyroxene increases. Garnet becomes richer in grossular and pyroxene richer in jadeite. Coronitic garnet is always asymmetrically zoned with Ca increasing and (Fe + Mg) decreasing, from biotite towards plagioclase. A retrograde phengite II developed outside the garnet corona. Equilibrium thermodynamic modelling Information on the metamorphic history of the metagranodiorite may be obtained by integrating petrographic observations of multivariant mineral associations with calculations of relative stability and composition of phases. To perform equilibrium calculations it is

M. Rubbo & M. Bruno

286

essential to determine the volume of the rock which reached chemical equilibrium. The bulk composition of this volume defines the effective bulk composition (EBC). In the case of the Brossasco-Isasca metagranodiorite it is impossible to define only one EBC, because the absence of a pervasive deformation during metamorphism prevented the exchange of matter among the different portions of the rock. This means that the original igneous mineral modes and mineral chemistry determined sites with different EBC (sites A, B and C), which evolved independently during metamorphism, at least as a first approximation. In order to establish the extent of equilibrium or disequilibrium and to understand the mineralogical evolution of each site during prograde metamorphism, the composition and abundance of local phases have been used as a basis for thermodynamic calculations.2 The initial compositions for the three considered sites are given in Table 1. The equilibrium calculations have been performed by minimising the Lagrangian, L(n, ) = G – (An – b),

(66)

at constant temperature and pressure (Smith & Missen, 1982). G is the Gibbs function: m

G(T, P, n) =  ni
i; i

ñ = (n1, …ni, …nm).

(67)

A(l, m) is the formula matrix, l is the number of elements, m is the number of phase components, b(l) is the element abundance vector, (l) is the row vector of Lagrange multipliers,
i is the chemical potential of each end-member component in each phase, and ni are their mole numbers which are constrained by the mass balance equation (An – b = 0) and non-negativity conditions (ni 1 0); ñ is the transpose of n. The abundance and composition of the phases have been calculated along a P–T trajectory described by the following equation: P – 4 [kbar] = 0.09(T – 430) [ºC]

(68)

which corresponds to the prograde (subduction) trajectory suggested by Schertl et al. (1991) and Compagnoni et al. (1994) for the BIU. Two hundred points were calculated for the P–T region between 3 and 40 kbar, and 420 and 830 ºC. Thermodynamic modelling was undertaken in the KNCFMnMASH system with biotite (Bt), plagioclase (Pl), K-feldspar (Kfs), phengite (Ph) , garnet (Grt), jadeite (Jd), kyanite (Ky), zoisite (Zo) , quartz (Qtz) , coesite (Coe) and water as phases. The phase components considered are: annite (Ann), phlogopite (Phl), eastonite (Eas) and Mnbiotite (Mn-bt) for biotite; albite (Ab), anorthite (An) and orthoclase (Or) for feldspar; muscovite (Ms), Fe-celadonite (Fe-cel) and Mg-celadonite (Mg-cel) for potassic white mica; almandine (Alm), grossular (Grs), pyrope (Prp), and spessartine (Sps) for garnet. The fluid is supposed to be pure H2O. The thermodynamic properties were taken from the Holland & Powell (1990) database, updated by Vance & Holland (1993). The solid solution models for biotite and phengite are from Powell (1978), Holland & Powell (1990) and Vance & Holland (1993); for garnet from Berman (1990); and for feldspar from Fuhrman & Lindsley (1988). By the equilibrium calculations the following information have been obtained.

Coronitic reactions in UHPM

287

 In each site, different reactions were overstepped at different temperatures and pressures, and produced typical mineral assemblages. In sites A and B the partial replacement of the igneous biotite by phengite (Ph I) and the development of a garnet corona suggest the following reaction: 2 Bt + 6 Qtz 2  Grt + 2 Ph,

(69)

where 0 – XEas X Eas  = n0Bt 1—————— + 2(XMs – XEas) ,

(70)

0 with n0Bt being the initial amount in moles of igneous biotite, X Eas the initial molar fraction of eastonite in igneous biotite, XEas the final molar fraction of eastonite in metamorphic biotite and XMs the final molar fraction of muscovite. Reaction (69) stops when eastonite is consumed (XEas = 0). Instead, in site C the reactions producing garnet are (69) and:

7An + Ann 2 Grs + Alm + 2Ky + Or + 2Zo

(71)

3Ms + 3Alm + 5Or + 3Zo + 4Qtz 2 16Ky + 8Grs + 9Fe-cel

(72)

 On a thermodynamic basis it has been assessed that, for peculiar bulk chemical compositions, biotite is stable all along the calculated P–T trajectory (Fig. 16). This is evident by considering the reactions producing garnet in the different sites. In sites A and B, the eastonite content determines the biotite abundance. Indeed, reaction (69) is the only net-transfer reaction involving biotite. While in the site C there are two net transfer reactions (69) and (71) able to consume completely the biotite. Likewise, both microscopic observations and thermodynamic calculations indicate that only water-constant reactions occurred, such as observed in the quartz eclogite facies metagranodiorite from Monte Mucrone, Sesia zone (Rubbo et al., 1999).  Thermobarometric estimates, obtained by comparing calculated and measured garnet compositions, indicate that the Brossasco metagranodiorite retains evidence of recrystallisation at a minimum pressure of 24 kbar at 650 ºC. As an example, the calculated garnet composition for site A is reported in Figure 17: the measured composition (Alm76–78Prp21–23Grs1–2Sps1–2) matches the calculated at 24 kbar.  Thermodynamic calculations also show that the Brossasco metagranodiorite experienced no further reactions for the P–T range between 28 kbar and 40 kbar and 690 to 830 ºC (Fig. 16).

2

A similar approach has been applied by Rebay & Powell (2002) to study eclogite facies metatroctolites (a gabbro originally consisting of the bimineralic assemblage plagioclase + olivine) from a variety of Western Alps localities, that preserve the original igneous texture. An equilibrium view of the mineral assemblages within plagioclase and olivine microdomains has been investigated, assuming local equilibrium at an appropriate scale (smaller than the original grain scale), and considering two different EBC for the microdomains.

288

M. Rubbo & M. Bruno

Fig. 16. Site A: Variations of abundance of biotite, phengite, garnet and quartz-coesite as a function of pressure and temperature.

Therefore it can be concluded that coronitic textures from eclogite facies metagranitoids preserved within greenschist to amphibolite facies orthogneiss can provide petrologic information useful to reveal the former presence of HP or UHP metamorphic recrystallisation. Further considerations on the model and the metamorphic evolution of the eclogite facies metagranodiorite Since the equilibrium calculations have only been performed along the prograde P–T path (Eqn. 68), only qualitative discussions on the retrograde metamorphism suffered by the

Fig. 17. Site A: Variation of garnet composition as a function of pressure and temperature.

Coronitic reactions in UHPM

289

metagranodiorite can be done. The major disequilibrium feature acquired during retrogression is revealed by the discrepancy between the calculated and measured compositions of biotite and phengite. The measured biotite compositions are inconsistent with that of garnet still retaining the composition of high pressure. Such discrepancies can be explained by considering that (i) the retrograde trajectory is different from the prograde one, and (ii) the sluggish volume diffusion limits the rate of the rock re-equilibration. During cooling, at subsequent steps successively large parts of garnet grains are effectively isolated from the reacting rock volume and the EBC will change. Such process is strongly stressed by the quick exhumation rate of the rock, estimated at about 2 cm/year by Gebauer et al. (1997) and 3 cm/year by Duchêne et al. (1997). (See Rubbatto et al., 2003, in this volume for further details.) As demonstrated by Bruno & Rubbo (unpublished results), the rapid cooling implies a limited volume diffusion in the coronitic garnet of the metagranodiorite, therefore, only a thin rim of garnet was in equilibrium with the adjacent phases (biotite and/or phengite) during retrogression. Therefore, re-equilibration of biotite and phengite occurred along a trajectory and with an EBC different from those of their formation. Moreover, there are some lines of evidence that metamorphism was not strictly isochemical: some matter, exchanged among sites, modified the initial EBC. Indeed, site A is characterised by Ca-free phases but garnet contains Ca; zoisite grew at the expense of the original plagioclase (site P) but hydroxyl groups diffused from close-by sites; the polyphase reaction corona at site B cannot be balanced using a simple reaction among the phases observed, but a flux of matter from outside the system is required. For this symplectite reaction the model of Ashworth & Chambers (2000) previously described can explain the formation of this corona, where a periodicity in the lamellar minerals is observed. In order to avoid misleading interpretations, all these factors should be considered when the equilibrium calculations are used to derive petrological information from the mineral assemblages. It is also worth remembering that the model system here presented lacks accurate thermodynamic knowledge on some phases especially on aqueous fluid rich in Na, K, Si and Al. As shown the Hermann’s (2002) experiments, these cations greatly affect the behaviour of fluids at high P and T in alkali-rich systems, such as the metagranodiorite studied.

Fig. 18. SEM backscattered image of the Monte Mucrone metagranodiorite. Biotite is partially replaced by phengite, and a garnet corona developed between biotite and plagioclase, now replaced by a polycrystalline aggregate of Zo + Jd(Ab) + Qtz.

290

M. Rubbo & M. Bruno

Garnet growth model Zoned coronitic garnets (Figs. 18, 19) developed between igneous biotite and plagioclase in the Monte Mucrone metagranodiorite (MMM) from the Sesia-Lanzo Zone, Western Alps. The MMM crops out in the “Eclogitic Micaschist Complex” (EMC), which is one of the main subunits making up the Sesia Zone (Dal Piaz et al., 1972; Compagnoni & Maffeo, 1973; Compagnoni, 1977; Compagnoni et al., 1977). The Sesia Zone is characterised by the widespread occurrence of eclogite facies assemblages in a wide spectrum of continental crustal lithologies. The EMC of the Sesia Zone is a fragment of Variscan continental crust, which was metamorphosed during the early Alpine HP subduction event (Oberhänsli et al., 1985). The early Alpine HP stage is constrained at T = 500–600 ºC and P = 16–20 kbar (Compagnoni, 1977; Lardeaux et al., 1982; Droop et al., 1990). The garnet corona (Fig. 18) is made up by aggregates of zoned garnets. A fingerprint of the metamorphic evolution is the Ca concentration in garnet, which is very low in the core and sharply increases towards the rim. A magnification of the X-ray compositional maps shows islands of nearly concentric iso-concentration lines (Rubbo et al., 1999), suggesting a nucleation and growth of single crystal, which later coalesced to a continuous layer. The equilibrium thermodynamic calculations show that garnet is stable over a wide range of pressure and temperature, as in the Brossasco-Isasca metagranodiorite previously described: so the isothermal isobaric model of layer formation is not applicable. Moreover, the pattern of the grossular concentration is an indication that the rate of Ca release and breakdown of plagioclase determined an increase in the garnet growth rate and Ca incorporation.3 It is also expected, as confirmed by an estimate (Rubbo et al., 1999), that the diffusion in the growth medium was faster than in garnet. Finally, it is sensible to suppose that the composition of the outer layer of garnet was close to, but not in equilibrium with, the other minerals, while temperature and pressure were changing during the metamorphism. An adjustment of the surface layer to the prevailing conditions could be achieved by an exchange of matter between garnet and the growth medium not involving its growth followed by segregation of the elements uptaken. Exchange reaction are also implied when geothermometers are used. From these considerations, it follows that thermobarometric information can be retrieved from the zoning pattern if some kinetic hypotheses can be made on the rate limiting processes. The qualitatively known P–T trajectory can be used for a consistency test. The calculation technique, used to simulate the garnet growth, is based on a work by Small & Ghez (1980), where it is shown how the equilibrium thermodynamics can be used to approximate the solution of a diffusion and growth problem. We expose this approach describing the growth of a spherical garnet (solid solution of pyrope, almandine, grossular and spessartine) and the evolution of the abundance and composition of minerals associated with it along a P–T trajectory. The calculations start at given initial amounts of phases (rock volume from which the garnet grows), determining the initial EBC, and at given initial values of temperature (T0) and pressure (P0). The initial EBC is made by the igneous phases with a composition typical of a

Coronitic reactions in UHPM

291

Fig. 19. Measured compositional pattern. Spot analyses are 2
m spaced. Bt, Core and Pl correspond to the garnet-biotite interface, the garnet core and the garnet-plagioclase interface, respectively. From Rubbo et al. (1999), modified. Table 2. Initial moles of end members in the EBC. Ann 0.1862

Phl 0.14114

Eas 0.17

Mn-bt –3

2.66

Ab

Or

An

Qtz

0.15

0.60

0.0

0.5

granodiorite while the abundance is estimated from the phase abundance in the site of the garnet (Table 2). However, plagioclase is not included in the initial EBC. Temperature and pressure are iteratively changed by small increments, T and P, along an elliptic P–T trajectory. Indeed, being interested in both prograde and retrograde paths the non linear trajectory proposed by Compagnoni et al. (1995) has been parametrised as an elliptical arc. The trajectory parameters are allowed to vary within the constraints imposed by the Compagnoni’s et al. (1995) work. Tk = T0 – Acos(3 + )

(73)

Pk = P0 – Bcos(3 +
)

(74)

T0 = 608.7K; P0 = 8.34 kbar

(75)

A = 253.51; B = 8.7497;  = 1.57,
= 2.04

(76)

4 3 = (n – 1) –—— , 1 ! 4 ! 2043, n integer 3360

(77)

Along with temperature and pressure the initial EBC progressively changes because plagioclase is progressively added and very small amounts of garnet, after its nucleation, are subtracted from the EBC, according to the following equation: 3

When the equilibrium sets in, the grossular end-member component decreases with increasing temperature and pressure, as shown in Figure 17. A similar pattern is obtained when plagioclase has the equilibrium composition but Ca is segregated in garnet.

292

M. Rubbo & M. Bruno M = E (Tk, Pk, nk) – N (Tk – T, Pk – P).

(78)

3 i k

nk stands for the vector whose elements are (n ) , where 3 runs on the phases, i runs over the phase components, and k is associated to the iteration4. M (i.e. Mg for the garnet and Mp for plagioclase) is the integral of the growth/dissolution rate of the crystal, at temperature Tk and pressure Pk over a time t. If M is positive, the phase  grows, whereas it resorbs if M is negative. To generate a spherical crystal of garnet, the moles Mg are homogeneously distributed as concentric shells. The total volume of segregated garnet is divided by a suitable number of crystals, to reproduce the size of the measured growth sector. M is calculated as a balance between the number of moles added, E(Tk, Pk, nk), and subtracted, N (Tk – T, Pk – P), to phase . E (i.e. Eg for the garnet) is the number of moles of the outer shell of phase  having the equilibrium composition at temperature Tk, pressure Pk, when the moles of end-members making the EBC are nk. E is calculated at every iteration by minimising the Gibbs function of the system. The details of the equilibrium calculations and related references, are reported in Rubbo et al. (1999). N is a function of Arrhenius type. It is calculated at (Tk – T, Pk – P) and it is the number of moles of phase  to be resorbed at Tk, Pk. The explicit expressions of N used are:

g(Tk – T0) Ng = gexp  ————— Tk 

, 

(79)

p (80) Np = pexp  –—  .  Tk  T0 = 608 K is the lowest temperature of the considered prograde trajectory and Tk the temperature of the kth iteration. To increase the input of Ca in the EBC, due to the dissolution of plagioclase, Np is made increasing with temperature. Conversely, Ng is made as a decreasing function of temperature, favouring garnet growth with increasing temperature. The small amount of garnet resorbed (and plagioclase redeposited) allows for a reequilibration of garnet (plagioclase) involving not only its surface. This also comply with the Small & Ghez (1980) calculation strategy, to represent diffusion limited growth (dissolution). There are no constraints on the values of the parameters p, g, p, g: so, they are determined by trial and error until the calculated concentrations satisfactorily fit the measured profiles. The optimal values are: g = 0.12, g = –14, p = 2.5, p = –5000. The Monte Mucrone metagranodiorite (MMM) garnet zoning reproduced with this model is reported in Figure 20. The calculated garnet zoning best compares with the measured one and is a confirmation that the kinetic assumptions made are correct. This garnet growth simulation has allowed to obtain the following information:  The MMM garnet records a typical clockwise P–T path for regional metamorphism, where pressure peak (P = 17 kbar, T = 835 K) precedes thermal peak (P = 16 kbar, T = 862 K). 4

The system phases and phase components are: plagioclase (Ab, Or, An); K-feldspar (Ab, Or, An); biotite (Ann, Eas, Phl, Mn-bt); phengite (Mg-cel, Fe-cel, Ms); garnet (Alm, Prp, Grs, Sps); Jd; Zo; Qtz; H2O.

Coronitic reactions in UHPM

293

Fig. 20. Calculated garnet zoning. It must be compared with the measured zoning (Fig. 19) of the sector from the garnet core to the face towards plagioclase. From Rubbo et al. (1999), modified.

 The maximum growth of garnet occurs close to the pressure peak and then the garnet resorbs during the retrograde path, but dissolution is interrupted by growth episodes.  It has not been possible to obtain the decrease of pyrope component towards the garnet rim by changing the parameters. This discrepancy between calculated and measured zoning may be caused by the coalescence of garnets forming the corona, which hindered the grain boundary diffusion of Mg from the garnet/biotite to the garnet/plagioclase interface. Because in the simulation the grain coalescence is not considered, the calculated molar fraction of Mg is somewhat higher than that measured in the outermost garnet layers towards the biotite/plagioclase interface. Then, the zoning simulation provides, indirectly, evidence of a factor of disequilibrium during garnet growth. Evidences of element disequilibrium in rocks have been reported in many works (e.g., Carlson, 2002; Hirsch et al., 2003). It appears that partial disequilibrium (meaning disequilibrium for some elements, but not for others) may be a common phenomenon during metamorphic mineral growth, even in ordinary prograde reactions that progress to completion. In the MMM, Mg seems to be unable to equilibrate at the scale of the coronitic structure, as well as Ca, indicating the following relation between the interdiffusion coefficients in the grain boundaries: DCa < DMg < DFe = DMn. In a recent research (Bruno & Rubbo, unpublished results) it was found that the agreement between calculated and measured pyrope profile is improved when lattice diffusion in garnet is considered. Some decrease of the pyrope component concentration at the garnet rim, could possibly be obtained by changing the retrograde P–T trajectory, but this analysis has not been undertaken.  Lattice diffusion in garnet occurred to a limited extent and was accompanied by important fluctuations of composition at the garnet growth front. The component concentration profile changes sharply near the garnet rim: this sharp change occurred at the high temperature segment of the P–T path and cation diffusion could then be active. For this reasons, a simulation of the growth accompanied by lattice diffusion in garnet has been undertaken: the correlation between

M. Rubbo & M. Bruno

294

temperature, pressure and time has been obtained and the garnet growth rate has been estimated by Bruno (2002).

Appendix Some notions of irreversible thermodynamics Let us consider a closed system in absence of external fields, capable of p–V work only. The first principle postulates the existence of a function of the state of the system, the internal energy (E), whose variations are due to the heat dQ and mechanical work –pdV exchanged with the surroundings through the system boundaries. dQ is positive if heat is received by the system. dE = dQ – pdV.

(1)

The internal energy can change by interactions between the system and the surroundings but not by transformations within the system. The internal energy of an isolated system cannot change. dE = deE;

diE = 0.

(2)

The indexes e, i, here and r in the following, will be used to indicate exchanges with the surroundings (e), exchanges within the system between phases (i), and mole variation as consequence of chemical reactions (r). For open regions can be defined as the work done on the region if it were closed (Haase, 1969), that is, in the simple situations here considered, the mechanical work dw = –pdV. Let consider a system made of two phases and let assume that no chemical reactions occur. External fields are absent. If the system is open, the “heat” adsorbed during an infinitesimal state change, by one of its open c-component phase is: c

dQ = dE + pdV –  hk(denk + dink).

(3)

k=1

hk is the partial molar enthalpy of species k and denk, dink are the amount of species k exchanged with the surroundings and with the other phases, respectively. If the system is closed and the two homogeneous phases ,  are at uniform temperature and pressure T , p and T , p, respectively, matter can be exchanged between phases only (denk = 0). dink + dink = 0;

k = 1, 2 … c.

(4)

From Equation 3 written for every phase we obtain: c

dQ = dE + pdV  –  hkdink /

(5)

k=1 c

dQ = dE + pdV  –  hkdink 5 k=1

(6)

Coronitic reactions in UHPM

295

The heat can be partitioned: that exchanged with the surroundings and that within the system between the two phases: dQ = deQ + diQ / 



(7)



dQ = deQ + diQ 5

(8) 

The heat exchanged between the whole system and the surroundings is deQ + deQ. The variation of the whole system’s internal energy is: dE = dE + dE 5

(9)

The energy conservation requires: dE = deQ – pdV  + deQ – pdV 

(10)

and thus substituting (5) and (6) in (7) and (8) and the result in (10) we find: c

c

diQ +  hkdink  + diQ +  hkdink  = 0.     k=1 k=1

(11)

This relation says that the energy received by phase  from phase  is of equal amount but of sign opposite to the energy recieved by  from . The heat transfer between the two phases is, by using Equations 4 and 11: c

diQ = –diQ +  (hk – hk)dink 5

(12)

k=1

Only if hk = hk it is diQ = – diQ (Haase, 1969). In the case the phases cannot exchange matter between them, it is: diQ + diQ = 0. The second law of thermodynamics states that the change of entropy of a system is a balance between two terms: the entropy flux due to exchanges with the surroundings (deS) and a non-negative source term due to irreversible processes occurring within the system (diS), that is dS = deS + diS,

(13)

diS 1 0.

(14)

If the system experiences only reversible modifications, then diS = 0. Let’s consider an isolated system composed of two phases  and . For the whole system the following relation holds: dS = dS + dS 1 0.

(15)

Within each phase, the production of entropy is non-negative: diS 1 0, diS 1 0.

(16) 



A situation such that dS > 0 with diS < 0 and diS > 0, is impossible (Prigogine, 1968). As an example, the entropy produced by the flux of heat caused by a difference of

M. Rubbo & M. Bruno

296

temperature is calculated for a system made by two closed phases , , kept at uniform temperatures T  and T . The system entropy is the sum dS = dS + dS 5

(17)

Every phase exchanges heat with the surroundings and with the other phase: dQ = deQ + diQ,

dQ = deQ + diQ 5

(18) 



As a consequence of the first principle, we have: diQ = – diQ and then: 1 1 deQ deQ dS = –—— + –—— + diQ  — – —  .  T T  T T

(19)

These contributions can be grouped in the following way: deQ deQ + –—— , deS = –——  T T

1 1 diS = diQ  — – —  . T T 

(20)

According to experience, if T  < T , then diQ > 0 and entropy is produced by the internal process. When the temperature is uniform then diS = 0. The rate of production of entropy is: diS diQ —— = –—— dt dt

1 1 — – — .  T T 

(21)

diS diQ Let us observe that —— is the product of the rate (or flux) –—— with the function of dt dt state

1 1  — , which is the “force” driving the heat “flux”.  – —  T T 

As a further example, let us consider two subsystems or phases  and  making a system closed to matter exchange with the surroundings. Each subsystem has homogeneous temperature T , T , pressure p, p, and chemical potential
k,
k of the k component. Let suppose that between the two phases an exchange of matter occurs. For the c-component subsystem (dropping the superscipts for the moment) the first principle reads: c

dE = deQ + diQ +  hkdink – pdV.

(22)

k=1

The Gibbs equation (obtained combining the first and second principles) is: c
k dE p dS = –— + –— dV –  –— dnk. T T T k=1

(23)

Now dnk 6 dink describes the moles variation for the exchange within the system, between the phases. The entropy balance dS = dS + dS (Eqn. 13), is calculated by Equations 23 and 22:

Coronitic reactions in UHPM deQ diQ + k=1 hkdink dS = –—— + – –———————– T T c



c k=1 

  k i k




c

k dnk +  –— T k=1

deQ diQ +  h d n + –—— + –———————– – T T 

297

c

 k=1


k –— dnk 5 T

(24)

The rate of entropy production is calculated by introducing Equation 11: c c
k
k  dink diS 1 diQ + k=1 hkdink  1 –— ––— . —— — – –—  –  –— = –———————– – T   dt dt T T  k=1  T  dt

(25)

diS Only if
i =
i and T  = T , —— is zero. dt A further example is the entropy balance for a closed phase where a chemical reaction occurs. Using the reaction advancements , it results drnk = kd. k are the stoichiometric coefficients of the reacting species. Introducing the affinity (A) and the reaction rate (v) c

A = –  (k
k), k=1

d v = –— , dt

the entropy balance reads deQ Ad dS = –— + ––— ; T T

diS A —— = — v. T dt

(26)

Being diS > 0, it ensues that A and v have the same sign; so if A < 0, then v < 0 and the reaction proceeds spontaneously towards the left (reactants). At equilibrium A = 0, v = 0. diS In the example seen and in general, each term in —— is a product of a rate or dt generalised flux (Jk) and a state function which is a generalised affinity or driving force A (Xk). In the example on the chemical reaction JK = v and XK = — . T The rate of entropy production P has therefore the following structure: diS P = —— =  JkXk 1 0. dt k

(27)

At equilibrium both fluxes and driving forces are zero for every irreversible process: Jk = 0, Xk = 0.

(28)

A central point, suggested by experience, is the assumption that close to equilibrium the fluxes depend linearly on forces. If one considers a system where two independent irreversible fluxes are produced by two independent forces, the hypothesis is made that J1 = L11X1 + L12X2

(29)

J2 = L21X1 + L22X2 ,

(30)

M. Rubbo & M. Bruno

298

or in general: Jk =  LkmXm .

(31)

m

This assumption implies that if one driving force, say X1, is zero, the flow J1 + 0, being driven by X2. The two processes are coupled. Coupling is possible between processes occurring in the same part of the system and if the forces driving them have the same tensorial character (Curie’s principle).5 For instance the affinity driving a chemical reaction (a scalar cause) may not be coupled with a temperature gradient, producing a heat flux. Substituting these expressions in Equation 27 one obtains: P = L11X12 + (L12 + L21)X1X2 + L22X22 > 0,

(32)

or in general: P =   LkiXkXi .

(33)

k

i

With reference to Equation 32 it can be shown that, P being positive for all positive and negative values of X1, X2, except X1 = X2 = 0, in which case P = 0, the following conditions hold: L11 > 0, L22 > 0; (L12 + L21)2 < 4L11L22 .

(34)

To generalise using the mathematical language, P is a definite positive quadratic form. A second central point is the theorem by Onsager, stating that the coupling coefficients are equal: Lik = Lki

(i, k = 1, …, n).

(35)

The Onsager relations are valid for the coefficients of the phenomenological equations if independent irreversible fluxes are written as linear functions of independent thermodynamic forces (see an example in Nishiyama, 1998). The calculation of the entropy production affords a means of obtaining the proper conjugate irreversible fluxes and thermodynamic forces needed to set up the phenomenological relations. We take relationship (35) as a postulate. The previous equations give deep insight in the nature of steady states. In a steady state the state functions are independent on time but fluxes can go on within the system. Particularly interesting are those states occurring when k independent affinities out of the n are constant. This will be shown with the help of Equation 33 in the case of two independent forces. In this particular case, a theorem, by Prigogine, can be stated as: if a system with two independent forces X1, X2, is kept in a state with fixed say X1, and minimum entropy production, the flux J2, conjugate to the non-fixed force, vanishes. Deriving Equation 32 with respect to X2, keeping constant X1, using Equation 35, one obtains: dP —— = 2(L12X1 + L22X2) = 2J2 = 0. dX2 5

(36)

Let us consider a scalar flow Js = LssXs + LsvXv and a vectorial flow Jv = LvsXs + LvvXv. Xs is a scalar force while Xv is a vectorial one. Lss is a scalar relating the scalar force Xs to the scalar flow Js. Lsv must be a vector relating the vectorial force Xv to the scalar flow Js. Lvs , relating a scalar force to a vectorial flow, must be a vector as well. This kind of coupling cannot occur in isotropic systems where a reversal of the sign of all coordinates axes must leave all the phenomenological coefficients invariant. The scalar coefficients Lss and Lvv are invariant but Lsv and Lvs would change sign upon inversion of the coordinate system. So it must be Lsv = 0 and Lvs = 0. A rigorous presentation can be found in Eu (1992).

Coronitic reactions in UHPM

299

Conversely if X1 is fixed and J2 = 0 the entropy production per unit time is minimal. Being P 1 0, the condition (36) identifies a minimum. To obtain (36) the Lij are assumed constant. Up to now, examples of irreversible processes in so-called discontinuous systems have been shown. These systems are made of uniform phases but differences of some intensive properties can occur at the boundaries between phases. In continuous systems the physical properties depend on time and are continuous functions of the space coordinates. The extension of the previous results to continuous systems does not require new physical principles. However, the description of these systems demands a heavy mathematical formulation. We will not derive the equations for the continuous systems, but simply introduce the particular form of these equations as needed for discussing ordinary diffusion. Diffusion Diffusion is a way of transport of matter caused by concentration gradients, occurring in a wide variety of processes in earth sciences and in materials sciences. Multicomponent diffusion is the transport process occurring when the flux of a component is affected by the concentration gradients of a second component (Cussler, 1976, 1984). Examples of applications to earth sciences are: diffusion of components in the melts feeding crystals growing in magma chambers; diffusion in solution, along mineral interfaces and within minerals promoting reactions in rocks and changes in bulk composition by metasomatic and metamorphic processes (Lasaga, 1997). A short introduction to a phenomenological description of diffusion will be presented here. Several textbooks and research papers are indicated to the interested reader: Kirkaldy & Young (1989) treats diffusion comprehensively, Crank (1999) is a good mathematical introduction; the important book edited by Ganguly (1991) is focusing on minerals; Lasaga (1997) has several chapters devoted to diffusion in minerals and to geochemical applications; Ghez (1988) is focusing on diffusion from the point of view of a crystal grower: perhaps it is the most stimulating reading for a mineralogist interested in mineral transformations in rocks. Finally, two books present a deep insight in transport and diffusion-reaction phenomena: Bird et al. (1960) and De Groot & Mazur (1962). For its high symmetry, garnet has a peculiar place in petrological works: here we would like to mention the interesting paper by Loomis (1978) on diffusion in garnet. Empirical laws of diffusion In the following the focus is on isothermal, isobaric diffusion of non-charged species, in absence of external forces and without chemical reactions, so that the driving forces are gradients of chemical potentials only. The derivation follows that by Haase (1969), but the equations will be written in a one dimension space, although the x component of vectors will be indicated with a bar, for notation clarity and to remember their origin. The first quantitative formulation of the experiments by Thomas Graham, on diffusion in gas and liquids, between 1830–1850, is due to Fick in 1855. The ingredients are:

M. Rubbo & M. Bruno

300

 the flux Ji of component i: it is the quantity of i passing through a unit area of a reference surface during unit time;  the concentration gradient of i, measured in the direction perpendicular to the reference surface. To appreciate the definition it is useful to consider that mass transport occurs by movements of atoms having different mass, volume and local averaged velocity. The local velocity, measured in a stationary reference frame, is an average over a volume large compared to the atomic dimensions, while very small compared to the measurable dimensions. 1 n v¯i = —  vij ni j=1

(37)

The weighted average flow velocity is n

 i=1

v¯ a =

n

ai = 1.  i=1

aiv¯i ;

(38)

ai can be the mole fraction of component i and examples of weights ai will be seen in the following. The diffusion motion of i is now described relative to the average flow velocity. This flow defines a frame of reference in which the diffusive motion of each species is gauged (Rosenberger, 1979). Thus the diffusion velocity of i with respect to v¯ a is defined as v¯ia = v¯i – v¯ a.

(39)

The total flux of a component relative to a stationary frame (the laboratory frame) is the product of a concentration, for instance mass density, times a velocity: J¯k = 7kv¯k = 7kv¯ a + 7k(v¯k – v¯ a).

(40)

The quantity J¯ = 7k(v¯k – v¯ ) is called the diffusion current density of species k or the diffusion flux of species k. It can be verified that a k

n

 k=1

a

ak a —— 7k J¯k = 0.

(41)

A reference velocity appropriate in diffusion experiments is the mean volume velocity of a volume element: n

v¯o =  ckVkv¯k .

(42)

k=1

ck, Vk, and v¯k are respectively the molarity, the partial molar volume and the average velocity of species k. The flux of species k is J¯ko = ck(v¯k – v¯ o). From experimental investigations it is found that

(43)

Coronitic reactions in UHPM n

J¯io = –

 k=2

ck Dik —— x

(i = 2, 3, …, n),

301

(44)

where J¯io and c1 are taken as dependent quantities. The Dik are the (n – 1)2 diffusion coefficients. They are functions of temperaure, pressure, concentrations. They do not depend on concentration gradients. The local mass balance equation is ckv¯ o ck J¯ko = – —— – —— —— x t x

(i = 2, 3, …, n).

(45)

ck In Equation 45 —— is the partial derivative with respect to time at fixed position. The t ckv¯ o convective term —— can be neglected in most diffusion experiments, so that inserting x Equation 44 in Equation 45 we obtain: n ck  ck  = ——   Dik —— —— t x  k=2 x 

(i = 2, 3, …, n).

(46)

Equations 44 and 46 are the generalisation of Fick’s first and second laws respectively. Thermodynamic theory In the case of isothermal, isobaric diffusion of non-charged species, in absence of external forces and without chemical reactions, the local entropy production is n 
k T"d = –  J¯k  ——  .  x  k=1

(47)

In Equation 47 the generalised fluxes J¯k are referred to any reference velocity. n

T"d =  J¯kXk

(48)


k Xk = – —— x

(49)

k=1

In Equation 48 both the n generalized fluxes and forces are linearly dependent. The dependent quantities can be eliminated using, for example, the relationship: n

 k=1

ck Xk = 0.

(50)

Then we obtain n

X1 = – and from

 k=2

c —k Xk c1

(51)

M. Rubbo & M. Bruno

302 n

 k=1

n

J¯kXk =

n

c

J¯k – —k J¯1  Xk =   c  k=2  k=2

n

ck(v¯k – v¯1) =

1

 k=2

J¯k1Xk .

(52)

J¯k1 is the diffusion flux of species k with respect to the reference velocity v¯1. The last expression of the local entropy contains only independent fluxes and forces. In terms of the n – 1 independent forces Xk given by Equation 49 and of the phenomenological coefficients Lik, it is: n

J¯i1 =

 k=2

LikXk

(i = 2, 3, …, n).

(53)

The last relationship can be transformed expressing Xk in terms of concentration gradients: n

J¯i1 = –  k=2

n

 l=2

cl Lik
kl —— x

(i = 2, 3, …, n),

(54)

where 
k
kl =  ——  (k, l, m = 2, 3, …, n; m + l).  cl T,P,cm

(55)

In order to compare Equation 54 with Fick’s first law, J¯i1 must be expressed in terms of J¯io: n

J¯i1 =

cV

n

i k   ik + —— J¯o =   cV  k k=2  k=2 1 1

eik J¯ko

(i = 2, 3, …, n).

(56)

By substituting in Equation 56 the expression of J¯ko given by Equation 44 we obtain n

J¯i1 = –  k=2

n

 l=2

c eikDkl –—l x

(i = 2, 3, …, n).

(57)

By comparison of relationship (57) with (54) it is seen that the phenomenological equations lead to the generalised Fick’s law. The Onsager reciprocity law Lij = Lji , (i, j = 2, 3, …, n) allows to reduce the number of the independent diffusion coefficients to –21 n(n – 1). Diffusion in garnet Garnet is one of the most studied minerals. In particular, Ca, Mg, Fe, Mn diffusion coefficients have been measured as a function of temperature and pressure, although with different degres of accuracy. The most recent measurements are reported in a series of works by Ganguly, Chakraborty and co-workers (Chakraborty & Ganguly, 1991, 1992). In a more recent one (Ganguly et al., 1998), the measurements have been tabulated as self-diffusion coefficients. These are defined in terms of the rate of transfer of components across a section fixed, so that no bulk flow occurs through it. This makes easier to interpret diffusion in terms of random molecular motion. From the self-diffusion coefficients it is possible to calculate the interdiffusion coefficients on the basis of the mean-field theory by Lasaga (1979). The Lasaga model takes into account the ionic nature of the diffusing cations in garnet.

Coronitic reactions in UHPM

303

Acknowledgements This chapter receives benefit from the criticism and suggestions by T. Nishiyama. We also acknowledge R. Compagnoni, F. Abbona and D. Aquilano, for critical comments and discussions.

References Ashworth, J.R. & Chambers, A.D. (2000): Symplectic reaction in olivine and the controls of intergrowth spacing in symplectites. J. Petrol., 41:285–304. Berman, R.G. (1990): Mixing properties of Ca-Mg-Fe-Mn garnets. Am. Mineral., 75:328–344. Biino, G.G. & Compagnoni, R. (1992): Very-high pressure metamorphism of the Brossasco coronite metagranite, southern Dora-Maira Massif, western Alps. Schweiz. Mineral. Petrogr. Mitt., 72:347–363. Bird, R.B., Stewart, W.E. & Lightfoot, E.N. (1960): Transport phenomena. New York (N.Y.): Wiley. Bruno, M. (2002): Garnet: a new growth model. From kinetics to reconstruction of the P-T-t-path. Plinius, 27:57–61. Bruno, M., Compagnoni, R. & Rubbo, M. (2001): The ultra-high pressure coronitic and pseudomorphous reactions in a metagranodiorite from the Brossasco-Isasca Unit, Dora Maira Massif, Western Italian Alps: a petrographic study and equilibrium thermodynamic modelling. J. Metamorph. Geol., 19:33–43. Bruno, M., Compagnoni, R., Hirajima, T. & Rubbo, M. (2002): Jadeite with the Ca-Eskola molecule from an ultra-high pressure metagranodiorite, Dora Maira Massif, Western Alps. Contrib. Mineral. Petrol., 142:515–519. Cahn, J.W. (1959): The kinetics of cellular segregation reactions. Acta Metall., 7:18–28. Carlson, W.D. (2002): Scales of disequilibrium and rates of equilibration during metamorphism. Am. Mineral., 87:185–204. Carlson, W.D. & Johnson, C.D. (1991): Coronal reaction textures in garnet amphibolites of the Llano Uplift. Am. Mineral., 76:756–772. Carlson, W.D. & Schwarze, E.T. (1997): Petrologic significance of prograde homogenization of growth zoning in garnet: An example from the Llano Uplift. J. Metamorph. Geol., 15:631–644. Chakraborty, S. & Ganguly, J. (1991): Compositional zoning and cation diffusion in garnets. In Ganguly, J. (ed.): Diffusion, atomic ordering, and mass transport /Adv. Phys. Geochem., 8/. Berlin: Springer-Verlag, 120–175. Chakraborty, S. & Ganguly, J. (1992): Cation diffusion in aluminosilicate garnets: Experimental determination in spessartine-almandine diffusion couples, evaluation of effective binary diffusion coefficients, and applications. Contrib. Mineral. Petrol., 111:74–86. Chambers, A.D. & Brown, P.E. (1995): The Lilloise intrusion, East Greenland: fractionation of hydrous alkalipicritic magma. J. Petrol., 36:933–963. Chopin, C. (1984): Coesite and pure pyrope in high-grade blueschists of the Western Alps: a first record and some consequences. Contrib. Mineral. Petrol., 86:107–118 Chopin, C. (1987): Very high pressure metamorphism in the western Alps: implications for subduction of continental crust. Philos. Trans. R. Soc. London, A321:183–197. Chopin, C., Henry, C. & Michard, A. (1991): Geology and petrology of the coesite-bearing terrain, Dora-Maira Massif, Western Alps. Eur. J. Mineral., 3:263–291. Christian, J.W. (1975): The theory of transformations in metals and alloys. Oxford: Pergamon Press. Compagnoni, R. (1977): The Sesia-Lanzo Zone: high pressure-low temperature metamorphism in the Austroalpine continental margin. Rend. Soc. Ital. Mineral. Petrol., 33:335–374. Compagnoni, R. & Maffeo, B. (1973): Jadeite-bearing metagranites l.s. and related rocks in the Mount Mucrone area (Sesia-Lanzo Zone, Western Italian Alps). Schweiz. Mineral. Petrogr. Mitt., 53:355–378. Compagnoni, R., Dal Piaz, G.V., Hunziker, J.C. & Gosso, G. (1977): The Sesia-Lanzo Zone, a slice of continental crust with Alpine high-pressure-low temperature assemblages in the western Italian Alps. Rend. Soc. Ital. Mineral. Petrol., 33:281–334.

304

M. Rubbo & M. Bruno

Compagnoni, R., Hirajima, T., Turello, R. & Castelli, D. (1994): The Brossasco-Isasca Unit. In Compagnoni, R., Messiga, B. & Castelli, D. (eds.): High pressure metamorphism in the Western Alps. Guide-book to the field excursion B1, 16th Gen. Meeting of the IMA, Pisa, 10–15 September 1994, 87–105. Compagnoni, R., Hirajima, T. & Chopin, C. (1995): Ultra-high-pressure metamorphic rocks in the Western Alps. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 206–243. Crank, J. (1999): The mathematics of diffusion. Oxford: Oxford Sci. Publ. Cussler, E.L. (1976): Multicomponent diffusion. Amsterdam: Elsevier. Cussler, E.L. (1984): Diffusion: Mass transfer in fluid systems. Cambridge: Cambridge Univ. Press. Dal Piaz, G.V., Hunziker, J.C. & Martinotti, G. (1972): La zona Sesia-Lanzo e l’evoluzione tettonicometamorfica delle Alpi Nord-occidentali interne. Mem. Soc. Geol. Ital., 11:433–466. De Groot, S.R. (1951): Thermodynamics of irreversible processes. Amsterdam: North-Holland Publ. Co. De Groot, S.R. & Mazur, P. (1962): Non-equilibrium thermodynamics. Amsterdam: North-Holland Publ. Co. Droop, G.T.R., Lombardo, B. & Pognante, U. (1990): Formation and distribution of eclogite facies rocks in the Alps. In Carswell, D.A. (ed.): Eclogite facies rocks. Glasgow: Blackie, 225–259. Duchêne, S., Blichert, T.J., Luais, B., Télouk, P., Lardeaux, J.M. & Albarède F. (1997): The Lu-Hf dating of garnets and the ages of the Alpine high-pressure metamorphism. Nature, 387:586–589. Eu, B.C. (1992): Kinetic theory and irreversible thermodynamics. New York (N.Y.): Wiley. Fisher, G.W. (1973): Non-equilibrium thermodynamics as a model for diffusion controlled metamorphic processes. Am. J. Sci., 273:897–924. Fisher, G.W. (1977): Non-equilibrium thermodynamics in metamorphism. In Fraser, D.G. (ed.): Thermodynamics in geology /NATO ASI Ser. C, 28/. Dordrecht: Reidel, 381–403. Fitts, D.D. (1962): Nonequilibrium thermodynamics. New York (N.Y.): McGraw-Hill. Fuhrman, M. & Lindsley, D.H. (1988): Ternary-feldspar modelling and thermometry. Am. Mineral., 73:201–215. Ganguly, J. (ed.) (1991): Diffusion, atomic ordering, and mass transport /Adv. Phys. Geochem., 8/. Berlin: Springer-Verlag. Ganguly, J., Cheng, W. & Chakraborty, S. (1998): Cation diffusion in aluminosilicate garnets: experimental determination in pyrope-almandine diffusion couple. Contrib. Mineral. Petrol., 131:171–180. Gebauer, D., Schertl, H.P., Brix, M. & Schreyer, W. (1997): 35 Ma old ultrahigh-pressure metamorphism and evidence for very rapid exhumation in the Dora-Maira Massif, Western Alps. Lithos 41:5–24. Ghez, R. (1988): A primer of diffusion problems. New York (N.Y.): Wiley. Haase, R. (1969): Thermodynamics of irreversible processes. London: Addison-Wesley Publ. Co. Henry, C. (1990): The coesite-bearing unit of the Dora-Maira Massif, western Alps, Italy: petrology and structure. Ph.D. Thesis, Univ. of Paris 6, France (in French). Hermann, J. (2002): Experimental constraints on phase relations in subducted continental crust. Contrib. Mineral. Petrol., 143:219–235. Hermann, J. (2003): Experimental evidence for diamond-facies metamorphism in the Dora-Maira Massif. Lithos, in press. Hirsch, D.M., Prior, D.J. & Carlson, W.D. (2003): An overgrowth model to explain multiple, dispersed high Mn-regions in the cores of garnet porphyroblasts. Am. Mineral., 88:131–141. Holland, T.J.B. & Powell, R. (1990): An enlarged and updated internally consistent thermodynamic dataset with uncertainties and correlation: the system K2O–Na2O–CaO–MgO–FeO–Fe2O3–Al2O3–TiO2–SiO2–C–H2–O2. J. Metamorph. Geol., 8:89–124. Joensten, R. (1977): Evolution of mineral assemblage zoning in diffusion metasomatism. Geochim. Cosmochim. Acta, 41:649–670. Johnson, C.D. & Carlson, W.D. (1990): The origin of olivine coronas in metagabbros from the Adirondack Mountains, New York. J. Metamorph. Geol., 8:697–717. Kaiken, G.D.C. (1994): Thermodynamics of irreversible processes. New York (N.Y.): Wiley. Katchalsky, A. & Curran, P.F. (1965): Nonequilibrium thermodynamics in biophysics. Cambridge (Mass.): Harvard Univ. Press.

Coronitic reactions in UHPM

305

Kienast, J.R., Lombardo, B., Biino, G. & Pinardon, J.L. (1991): Petrology of very high pressure eclogitic rocks from the Brossasco Isasca Complex, Dora Maira Massif, Italian Western Alps. J. Metamorph. Geol., 9:19–34. Kirkaldy, J.S. & Young, D.J. (1989): Diffusion in the condensed state. London: Inst. Met. Korzhinskii, D.S. (1959): Physicochemical basis of the analysis of the paragenesis of minerals. New York (N.Y.): Consultants Bureau. Lardeaux, J.M., Gosso, G., Kienast, J.R. & Lombardo, B. (1982): Relations entre le métamorphisme et la déformation dans la Zone Sesia-Lanzo (Alpes Italiennes). Bull. Soc. Géol. Fr., 24:793–800. Lasaga, A.C. (1979): Multicomponent exchange and diffusion in silicates, Geochim. Cosmochim. Acta, 43:455–469. Lasaga, A.C. (1997): Kinetic theory in the earth sciences. Princeton, New Jersey: Princeton Univ. Press. Loomis, T.P. (1978): Multicomponent diffusion in garnet: I Formulation of isothermal models. Am. J. Sci., 278:1099–1118. Messiga, B. & Bettini, E. (1990): Reactions behaviour during kelyphite and symplectite formation: a case study of mafic granulites and eclogites from the Bohemian Massif. Eur. J. Mineral., 2:125–144. Moseley, D. (1984): Symplectic exsolution in olivine. Am. Mineral., 69:139–153. Nishiyama, T. (1983): Steady diffusion model for olivine-plagioclase corona growth. Geochim. Cosmochim. Acta, 47:283–294. Nishiyama, T. (1998): Uphill diffusion and a new nonlinear diffusion equation in ternary non-electrolyte system. Phys. Earth Planet. Inter., 107:33–51. Oberhänsli, R., Hunziker, J.C., Martinotti, G. & Stern, W.B. (1985): Geochemistry, geochronology and petrology of Monte Mucrone: an example of eo-Alpine eclogitization of Permian granitoids in the SesiaLanzo Zone, Western Alps, Italy. Chem. Geol., 52:165–184. Powell, R. (1978): Equilibrium thermodynamics in petrology. An introduction. London: Harper & Row. Prigogine, I. (1968): Introduction à la thermodinamique des processes irréversibles. Paris: Dunod. Rebay, G. & Powell, R. (2002): The formation of eclogite facies metatroctolites and a general petrogenetic grid in Na2O-CaO-FeO-MgO-Al2O3-SiO2-H2O (NCFMASH). J. Metamorph. Geol., 20:813–826. Rosenberger, F. (1979): Fundamentals of crystal growth I. Berlin: Springer-Verlag. Rubbo, M., Borghi, A. & Compagnoni, R. (1999): Thermodynamic analysis of garnet growth zoning in eclogitised granodiorite from M. Mucrone, Sesia Zone, Western Alps. Contrib. Mineral. Petrol., 137:289–303. Rubie, D.C. (1990): Role of kinetics in the formation and preservation of eclogites. In Carswell, D.A. (ed.): Eclogite facies rocks. Glasgow: Blackie, 111–140. Schertl, H.P., Schreyer, W. & Chopin, C. (1991): The pyrope-coesite rocks and their country rocks at Parigi, Dora-Maira Massif, Western Alps: detailed petrography, mineral chemistry and PT-path. Contrib. Mineral. Petrol., 108:108–121. Sharp, Z.D., Essene, E.J. & Hunziker, J.C. (1993): Stable isotope geochemistry and phase equilibria of coesitebearing whiteschists, Dora Maira Massif, Western Alps. Contrib. Mineral. Petrol., 114:1–12. Small, M.B. & Ghez, R. (1980): Growth and dissolution kinetics of III-V heterostructures formed by LPE. II. Comparison between thermodynamic and kinetic models. J. Appl. Phys., 51:1589–1592. Smith, W.R. & Missen, R.W. (1982): Chemical reaction equilibrium analysis. New York (N.Y.): Wiley. Thompson, J.B., Jr (1959): Local equilibrium in metasomatic processes. In Abelson, P.H. (ed.): Researches in geochemistry. New York (N.Y.): Wiley, 427–457. Vance, D. & Holland, T.J.B. (1993): Detailed isotopic and petrological study of a single garnet from the Gassets Schist, Vermont. Contrib. Mineral. Petrol., 114:101–118. Copyright permissions Figures 1, 10 and 11 are reprinted from American Mineralogist, Vol. 87, Carlson, W.D.: Scales of disequilibrium and rates of equilibration during metamorphism, pp. 185-204, copyright 2002, with permission from the Mineralogical Society of America. Figures 2, 3 and 4 are reprinted from Geochimica et Cosmochimica Acta, Vol. 41, Joensten, R.: Evolution of mineral assemblage zoning in diffusion metasomatism, pp. 649-670, copyright 1977, with permission from Elsevier. Figures 5, 6, 7, 8 and 9 are reprinted from Journal of Petrology, Vol. 41, Ashworth, J.R. & Chambers, A.D.: Symplectic reaction in olivine and the controls of intergrowth spacing in symplectites, pp. 285-304, copyright 2000, with permission from Oxford University Press.

306

M. Rubbo & M. Bruno

Figure 16 is reprinted from Journal of Metamorphic Geology, Vol. 19, Bruno, M., Compagnoni, R. & Rubbo, M.: The ultra-high pressure coronitic and pseudomorphous reactions in a metagranodiorite from the BrossascoIsasca Unit, Dora Maira Massif, Western Italian Alps: A petrographic study and equilibrium thermodynamic modelling, pp. 33-43, copyright 2001, with permission from Blackwell Publishing. Figures 19 and 20 are reprinted from Contributions to Mineralogy and Petrology, Vol. 137, Rubbo, M., Borghi, A. & Compagnoni, R.: Thermodynamic analysis of garnet growth zoning in eclogitised granodiorite from M. Mucrone, Sesia Zone, Western Alps, pp. 289-303, copyright 1999, with permission from Springer.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 10, 307–340

Mineral assemblages in ultrahigh pressure metamorphism: A review of experimentally determined phase diagrams STEFANO POLI* and PATRIZIA FUMAGALLI Dipartimento di Scienze della Terra, Università degli Studi di Milano, Via Botticelli 23, 20133 Milano, Italy; * e-mail: [email protected]

Introduction In the last decade, the considerable amount of ultrahigh pressure terrains found around the world (Liou et al., 1998) and recent findings of mineral inclusions suggesting equilibration pressures up to more than 20 GPa (e.g. Stachel et al., 2000) have led to an increasing interest in petrological tools to be used to unravel the P–T evolution at extreme conditions. The expected resolution of such tools should provide a key for understanding tectonic processes responsible for the dynamics of UHP terrains. Even though there has been a substantial amount of experimental work in simple model systems, mainly devoted to define the maximum stability field of ultrahigh pressure phases or to retrieve thermodynamic data from phase equilibria constraints, information on phase relationships in more complex, Fe-bearing systems approaching natural compositions are still fragmentary. A brief list of references for a variety of model systems where ultrahigh pressure phases are found is given in Table 1. References cited are not intended to be exhaustive but provide a starting point for the relevant literature in some of the systems listed. Experimental data in simple model systems often provide fundamental constraints to basic mechanisms responsible for the appearance of ultrahigh pressure phases, but they are usually of limited help to resolve environmental conditions with sufficient accuracy in real rocks. Most natural systems are actually characterised by a large number of independent chemical components, therefore complex continuous reactions are expected to control most phase relationships. A number of experimental difficulties are responsible for the relatively restricted number of experimental studies in complex systems at conditions reproducing cold geotherms (high dP/dT). Most researchers observe a progressive decrease in the “reactivity” of the experimental charges with increasing pressure and with the increasing complexity of solid solutions. As a result, most subsolidus charges show sluggish transformations at UHP conditions and poorly equilibrated textures, even in long duration experiments (compare Fig. 1a and 1b). Furthermore, conventional bracketing techniques cannot be applied in complex, natural systems. Despite these limitations, such experimental systems offer a unique perspective on phases and reactions of direct application to the Earth’s interior.

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

S. Poli & P. Fumagalli

308

Table 1. Model systems where ultrahigh pressure phases are found, with a brief list of references System

UHP phases

Source for literature

FMS

Akaogi et al. (1998); Frost et al. (2001)

FMAS CMS

olivine, pyroxene, wadsleyite, ringwoodite, ilmenite, majorite, perovskite, magnesiowüstite, stishovite enstatite, pyrope, majorite, ilmenite, perovskite, “tetragonal phase” majorite diopside, Ca-perovskite, walstromite, CaSi2O5

CFS CMA

skiagite spinel, Ca-ferrite, “hexagonal Al-rich phase”

NCMAS NKAS ASH

jadeite, Na Ca-ferrite wadeite, hollandite, calcium ferrite(-structure) topaz-OH, phase Pi, phase egg

MAS

MSH

antigorite, clinohumite, phase A, phase B, superhydrous phase B (phase C), phase D (phases F and G), phase E, 10A phase MASH chlorite, Mg-sursassite (MgMgAl pumpellyite), Mg-staurolite CASH lawsonite, zoisite CFASH clinozoisite, garnet KMASH phlogopite, wadeite, phase X, phengite KFASH CM-CO2

dolomite, aragonite, magnesite

Akaogi et al. (2002); Heinemann et al. (1997) O’Neill & Jeanloz (1994) Canil (1994); Gasparik et al. (1994); Gasparik (1996a) Woodland & O’Neill (1995) Akaogi et al. (1999); Miyajima et al. (2001) Gasparik (1996b) Yagi et al. (1994) Daniels & Wunder (1996); Schmidt et al. (1998) Angel et al. (2001); Ulmer & Trommsdorff (1999); Wunder (1998); Stalder & Ulmer (2001) Bromiley & Pawley (2002); Fockenberg (1998) Poli & Schmidt (1998); Schmidt (1995) Brunsmann et al. (2002) Trønnes (2002); Massonne & Szpurka (1997); Hermann (2002b); Schmidt et al. (2001) Luth (1995, 2001)

Abbreviations (System column): A: Al2O3, C: CaO, F: FeO, H: H2O, K: K2O, M: MgO, N: Na2O, S: SiO2

Fig. 1. Backscattered electron images of run products showing textural differences between experiments performed on similar bulk compositions and run durations, but at extremely different pressure and temperature conditions. (a) graphite-bearing, fluid-saturated model MORB at 2.7 GPa, 730 °C, 224 hours (courtesy Ada Crottini); (b) graphite-bearing, fluid-saturated natural MORB at 19 GPa, 1200 °C, 120 hours (courtesy Kazuaki Okamoto). gar: garnet; cpx: clinopyroxene; ky: kyanite; maj: majorite; ca-pvs: Ca-perovskite (bright spots); sti: stishovite (dark spots). Note that magnification in (a) is 1000× and in (b) is 1800×.

Mineral assemblages in ultrahigh pressure metamorphism

309

This review will focus mainly on experimental studies both on complex model systems close to real rocks and on natural materials, in order to facilitate the approach to the world of ultrahigh pressure rocks for those who are not familiar with the variety of new crystalline phases found in the last decade.

“Fluid” phases at ultrahigh pressure conditions A few definitions Because of the global geochemical and geodynamic implications, there has been an increasing interest in the role and features of fluids at ultrahigh pressure conditions. Even though various classical textbooks have regarded the eclogite facies as a “dry” environment, there is an increasing consensus that ultrahigh pressures rather represent “wet” conditions, at least when a prograde pressure path is attained. As a first step it is important to recall definitions which will be useful in the following discussion. N H2O: it is a chemical component, it represents the stoichiometry of a pure compound (water), but it often also indicates a chemical species in complex fluids. The definition of component H2O in a system does not imply any specific value for its chemical potential. N water: a fluid composed of H2O only. The chemical potential of H2O will be maximum, i.e. it implies H2O saturation and unit activity for H2O. Though unrealistic at geological conditions, water is the reference fluid in most thermodynamic calculations and in numerical simulations. N fluid: a volatile-rich phase, showing supercritical behaviour, characterised by a poorly polymerised structure. Fluid saturation implies the appearance of a fluid, i.e. it defines a vaporous surface, but it does not imply a univocally defined chemical potential of any of the species present in it, e.g. H2O, CO2 or O2. From the previous definitions, we obtain: N “dry” conditions (“water absent”, Type I in Robertson & Wyllie, 1971): a H2O-free system. N “wet” conditions: a fluid-bearing system, where the abundance of H2O is enough to saturate also a silicate liquid, e.g. on a wet solidus (“water excess”, Type IV in Robertson & Wyllie, 1971). N H2O-deficient and fluid-absent (“water deficient” in Yoder, 1952; “water deficient and vapor-absent”, Type II in Robertson & Wyllie, 1971): an assemblage of minerals including hydrates but not a free fluid. It is a condition of H2O undersaturation. N H2O-deficient and fluid-present (“water deficient and vapor-present”, Type III in Robertson & Wyllie, 1971): an assemblage of minerals plus a free fluid, but there is insufficient H2O (and fluid) present to saturate the liquid when the assemblage is completely melted (see Figure 2 in Robertson & Wyllie, 1971). N aqueous fluid: is a fluid mainly composed of H2O plus other volatile species and/or dissolved solids. N carbonic fluid: is a fluid mainly composed of C species plus other volatiles and/or dissolved solids.

310

S. Poli & P. Fumagalli Fig. 2. Sequence of reactions involving hydrates and carbonates in the model systems MgO–SiO2– H2O (a–d) and MgO–CaO–CO2 projected from silica (e–f). Assemblages below the tie-lines brucite–antigorite–talc–quartz in (a) and (b) represent H2Oundersaturated fluid absent conditions; assemblages below the tie-lines magnesite–dolomite– aragonite are CO2-undersaturated. Points 1 and 2 stand for two hypothetical ultramafic bulk compositions. arag: aragonite; atg: antigorite; br: brucite; cc: calcite; di: diopside; dol: dolomite; en: enstatite; fo: forsterite; mag: magnesite; p: periclase; q: quartz; ta: talc; wo: wollastonite. See text for explanation.

N mixed fluid: often refers to a fluid with mixed volatile species, usually mainly H2O and CO2. How likely is fluid saturation at high pressure? Whenever hydrates or carbonates are present in a system (as in the simple model chemographies depicted in Fig. 2), subsolidus compositional diagrams are separated into two regions, the first where assemblages are fluid-saturated, i.e. a fluid occurs together with solids, and the second one where hydrates and/or carbonates occur, but no fluid is present (H2O or CO2-deficient). In Figure 2a, the mineral assemblages brucite-antigoritefluid, antigorite-talc-fluid and talc-quartz-fluid represent fluid-saturated assemblages, whereas assemblages below the tie lines brucite–antigorite, antigorite–talc, talc–quartz are H2O-deficient, fluid absent. Similarly, assemblages below the tie-lines magnesite–dolomite–calcite in Figure 2e are CO2-deficient and fluid absent. Such tielines define a “fluid-saturation surface” which separates fluid saturated from fluid absent assemblages. The fluid saturation surface is a line in simple ternary compositional diagrams while a plane in quaternary diagrams. Fluid absent (H2O-conservative) reactions occur below such tie-lines, e.g. the reaction forsterite + talc = enstatite + antigorite (Fig. 2a to 2b) or the reaction enstatite + dolomite = magnesite + diopside (Fig. 2e to 2f). These reactions do not lead to dehydration or decarbonation because no H2O or CO2 components are liberated.

Mineral assemblages in ultrahigh pressure metamorphism

311

Because low-temperature, low-pressure metamorphism commonly promotes the formation of hydrates that contain large amounts of H2O (e.g. antigorite, talc, chlorite, clay minerals, lawsonite etc.), the likelihood of H2O-deficient assemblages is maximised at such conditions. Let us assume two partially serpentinised rocks from the ocean floor, a harzburgite and an orthopyroxenite (points 1 and 2 in Fig. 2a). For this example, they will store the same amount of H2O in the assemblage antigorite + talc + forsterite. With increasing pressure and temperature, the reaction forsterite + talc = enstatite + antigorite will first lead to two different hydrate-bearing assemblages (Fig. 2b). The reaction antigorite + talc = enstatite + H2O will cause fluid saturation in the orthopyroxenite and consequently dehydration, whereas the harzburgite will remain at H2O-deficient conditions (Fig. 2c). Further dehydration will be caused by the breakdown of antigorite to forsterite + enstatite + H2O. Only at this stage does a fluid appear in the harzburgite (Fig. 2d). This reaction sequence implies that the “fluid saturation surface” progressively moves away from the volatile component, as a function of the hydrates stable at the various P–T conditions of interest. Therefore there is no straightforward relationship between “amount of volatiles” in the rock and “amount of fluid”; it is the stable phase assemblage that governs how H2O is partitioned. From Figure 2 it is evident that: N at fixed maximum H2O contents, the attainment of fluid saturation, i.e. attainment of conditions where the assemblage is on or above the fluid saturation surface, will be promoted by an increase in pressure and temperature; N fluid removal at constant P and T (and therefore phase assemblage) will never lead to H2O deficiency, simply because there is no process able to remove a fluid below the fluid saturation surface, i.e. where the fluid does not exist. However, because the appearance of a fluid is related to the thermodynamic definition of saturation, the two previous statements do not represent a paradigm. Let us first consider a schematic block diagram where the chemical potential of the volatile species is plotted on the z axis, and P and T on the other two axes (Fig. 3). The chemical potential for the volatile species, e.g. H2O, will be maximum when a fluid, composed of this species only, appears. Any phase assemblage, or reaction, occurring below the fluid saturation line previously discussed, necessarily plots in the (physically accessible) halfspace below the surface which describes the equation of state for the pure fluid, e.g. water. Figure 3 shows that with increasing pressure and temperature, along a steep geothermal gradient (as it is the case for a number of UHP terrains), the surface representing the chemical potential of H2O buffered by antigorite + forsterite + enstatite progressively approaches the fluid saturation surface. The intersection between these two surfaces corresponds to the univariant breakdown reaction of antigorite and to the appearance of a fluid, in agreement with the chemographies shown in Figure 2. As we consider CO2 fluids we observe that, similarly to the previous case, the appearance of a carbonic fluid from a CO2-deficient assemblage, e.g. magnesite + enstatite + quartz, is controlled by the intersection of the two surfaces defining the chemical potential of CO2 in the solid assemblage and in the fluid. However, in this example, as a result of the large dG/dP for a pure CO2 fluid and of the relative location of the surface for the assemblage magnesite + enstatite + quartz, a relatively “flat” dP/dT slope is obtained for the decarbonation reaction responsible for the generation of

312

S. Poli & P. Fumagalli

Fig. 3. Saturation surfaces for water and carbon dioxide compared to chemical potentials of H2O and CO2 buffered by antigorite-enstatite-forsterite and by magnesite-quartz-enstatite, respectively. Volatile saturation and therefore fluid release occurs if the P–T path drives µH O or µCO toward the thermodynamic surfaces of 2 2 H2O or CO2 fluid, respectively. Note that µi = molar G (from Poli & Schmidt, 2002).

carbonic fluid. The same prograde P–T path will therefore promote devolatilisation of the solid assemblage as far as H2O is concerned, but conversely it will favour fixation of CO2 in the solid because decarbonation would be rather driven by a moderate pressure increase relative to the temperature increase. When both H2O and CO2 are present, the free energy of the mixed fluid is not a single value at P and T, hence it cannot be plotted as a simple surface in the diagram of Figure 3. Nevertheless a similar line of reasoning can be applied in this more complex case. Again, a chemographic analysis is useful to show some counterintuitive implications of mixed volatile systems. In the triangle of Figure 4 we represent the system CaO–Al2O3–SiO2–H2O–CO2 projected from kyanite and coesite and for each three-phase and two-phase assemblage the chemical potential of H2O is reported. Such calculation demonstrates two interesting points. First, at the same pressure and temperature conditions, the chemical potential of H2O can be identical in fluid-absent conditions and in the presence of a fluid. Second, the addition of CO2 to a fluid-absent

Mineral assemblages in ultrahigh pressure metamorphism

313

Fig. 4. Composition diagram for the system CaO–Al2O3–SiO2–H2O–CO2 projected from kyanite and coesite (from Poli & Schmidt, 1998, modified) which shows occurrence of hydrates and carbonates at fluid-absent conditions, equality of chemical potential of H2O for selected fluid present and fluid absent conditions and possible appearance of an aqueous fluid by addition of CO2 to the system.

mineral assemblage (point A in Fig. 4) may lead first to the crystallisation of more hydrous minerals (e.g. lawsonite and zoisite in B) and then to the appearance of an aqueous fluid in C. Nevertheless, it has been common practice to assume identity between the chemical potential of a volatile component and the chemical potential of that volatile in the fluid, when multi-equilibrium calculations (e.g. TWEEQU, Berman, 1991) are used to derive the composition of metamorphic fluids. However, the actual presence of a mixed fluid, rather than a condition of H2O deficiency and fluid absence, should be carefully demonstrated. In conclusion, even in such very simple model systems, the prediction of fluid saturation is not an easy task. It obviously depends on the thermodynamic properties of the solid assemblages compared to the properties of the possible fluids in a system. The worst and most common condition is represented by rocks where components C, O and H are independent variables. This is the case, as an example, when hydrates, carbonates and other mineral phases coexist in a Fe-bearing system. It is beyond the goal of this contribution to review all of the complexities present in C-O-H bearing rocks (e.g. see

314

S. Poli & P. Fumagalli

Connolly, 1995). Nevertheless it is important to point out that the appearance of a fluid is related to a complex interplay of mass balance and thermodynamic constraints because variable amounts of C, H and O can be stored in carbonates, hydrates and ferrous/ferric minerals as well as in fluids. Fluids, melts and the 2nd critical endpoints Even though in most thermodynamic calculations performed to extract P–T paths, fluids are supposed to be composed of volatile species only, it has been known since a long time that fluids at geological conditions dissolve variable amounts of solids which are far from negligible. It is widely accepted that temperature and, secondarily, pressure promote dissolution. However, there is still an open debate how and how much (Stalder et al., 2001; Mibe et al., 2002) solids are dissolved. Because of intrinsic experimental difficulties, there is still a very limited amount of systematic work devoted to determine the vaporous surface in the different chemical systems. Even though petrologists are used to think about the variability of the solidus and the liquidus surfaces as a function of the buffering assemblage in P–T–composition space (e.g. the solidus for a gabbro is different from the solidus for granite), there has been a significant underestimation of the possible variability of the vaporous surface, i.e. of the surface limiting the fluid phase field (Boettcher & Wyllie, 1969). As an example, Figure 5 represents the vaporous surface in the system MgO–SiO2–H2O determined by Ryabchykov et al. (1983). At fixed P–T conditions, the amount of silica dissolved in the fluid will be maximum in assemblages coexisting with coesite, but in an assemblage with orthopyroxene will depend on the bulk composition chosen. It is therefore evident that in complex, high variance systems the amount of silica present in fluids and/or hydrous melts can be much different than in the reference system SiO2–H2O. It can be lower, as shown in Figure 5, or even higher, as in the granite–H2O system, where the hydrous liquid has a higher SiO2 content (e.g. at 0.4 GPa, 900 °C) than a liquid in the quartz–H2O system. Though it might seem a trivial statement, it should be remembered that a fluid is a solution and therefore its actual composition will depend upon the buffering phase assemblage. Because of the increasing amount of dissolved solids present in fluids at increasing pressure and temperature and because of the mutually increasing solubility of H2O in melts, it has been found that the “miscibility gap” between fluids and melts shrinks with increasing pressure (Fig. 6, after Stalder et al., 2000). A number of experimental studies have shown that the critical point between fluids and melts can be attained under geological conditions (e.g. Bureau & Keppler, 1999), even though the amount of volatile components required to encounter such a phenomenon is expected to be unusual. However, shrinking of the solvus with increasing pressure rather implies a migration of the critical point toward lower temperature, as shown in Figure 6. When the critical curve intersects the reaction describing the wet solidus, the two volatile-bearing phases (fluid and melt) are replaced by one single “supercritical” phase. As a consequence, a singularity results: the wet solidus vanishes, the concept of melting looses its definition,

Mineral assemblages in ultrahigh pressure metamorphism

315

Fig. 5. The vaporous surface in the system MgO–SiO2–H2O at 3 GPa after Ryabchykov et al. (1983).

and the breakdown of solid assemblages is replaced by a concept of continuous dissolution in a first volatile-rich, and then silicate-rich, “non-solid” phase. This singularity is called 2nd critical endpoint and it is of utmost importance in ultrahigh pressure environments. As shown by Boettcher & Wyllie in their classical paper dated 1969, in a multicomponent chemical system, there is more than one 2nd critical endpoint, and, again, the relative location of the vaporous, solidus and liquidus surfaces are related by the coexisting buffering assemblages (Fig. 7, from Boettcher & Wyllie, 1969). In the simple system SiO2–H2O the second critical endpoint is at only 1 GPa and 1100 °C (Kennedy et al., 1962), in albite–H2O and haplogranitic systems moves to ca. 1.5 GPa (Bureau & Keppler, 1999; Stalder et al., 2000), in CaO–SiO2–H2O–CO2 to 3.2 GPa and 500 °C

316

S. Poli & P. Fumagalli

Fig. 6. Schematic presentation of a binary system solid + H2O (from Stalder et al., 2000, copyright Mineralogical Society of America). P2 > P1.

(Boettcher & Wyllie, 1969), whereas in the ultramafic model system MgO–SiO2–H2O (MSH) the solidus terminates at ca.12 GPa and 1100 °C (Stalder et al., 2001). In the system quartz + H2O, silica dissolved in the fluid ranges from a few weight percent at 500–700 °C to a maximum of 12 wt% at 1 GPa and 900 °C. In MSH at 6 GPa, the amount of silicate dissolved in the fluid does not exceed ca. 10 wt% up to temperatures close to the solidus (ca. 1150 °C). Experiments performed by Schneider & Eggler (1986) suggest quantities of solute from 3 wt% in an amphibole peridotite at 1.5–2.0 GPa, 750–900 °C to 12–15 wt% in a phlogopite peridotite at 1.3–2.0 GPa, 1100°C. Addition of CO2 to the fluid strongly decreases the solubility of silicates, by ca. one order of magnitude. As a conclusion, we emphasise that the inadequacy of current knowledge on thermodynamic properties of UHP fluids strongly suggests the use of fluid-absent reactions for geothermobarometric purposes.

Mineral assemblages in ultrahigh pressure metamorphism

317

Fig. 7. Isobaric, polythermal diagrams for the hypothetical ternary system MO–SiO2–H2O (from Boettcher & Wyllie, 1969). Dashed lines are isothermal field boundaries. Dotted lines connect temperature maxima on the liquidus and vaporous field boundaries with the crystalline phase with which the liquid and vapour are in equilibrium. Reprinted with permission from Elsevier.

318

S. Poli & P. Fumagalli

Ultrahigh pressure rocks: The quartz–coesite transformation and the “internally consistent” thermodynamic databases UHP conditions refer to rocks which underwent recrystallisation within (or above) the stability field of coesite. It is therefore of primary importance that the position of the reaction quartz–coesite is still debated, especially in the relatively low temperature region. The transformation quartz–coesite is of particular interest not only as a pressure reference in high pressure experimental studies or as a lower pressure limit for UHP terrains, but it also indirectly influences the thermodynamic properties of most phases tabulated in the so-called “internally consistent” databases. In fact the concept of “internal consistency” is related to the high correlation between all of the variables contained in a very large over-determined system of non-linear equations whose dimensionality is controlled by dimensionality of the chemical components introduced. Conversely, thermodynamic properties of quartz and coesite (and therefore the calculated location of quartz–coesite transformation) will be controlled not only by the experiments performed in the pure system SiO2, but also by all of the reactions which include either quartz or coesite or both, in a variety of assemblages belonging to a very large chemical system (up to more than 12 components in the latest release of the database by Holland & Powell, 1998). Despite the rigorous approach adopted in the internally consistent databases, optimisation routines (least squares, mathematical programming, Bayes method) may often obscure the basic quality/accuracy/precision of the experimental data behind the recalculation scheme. The quartz to coesite transformation is a typical example of this problem. Figure 8 shows the experimentally derived position of this reaction from 400 °C to 1400 °C compared to the calculated equilibrium according to the database of Holland & Powell (1998) and updates. There are still major discrepancies between the different experimental studies, which cannot be simply solved by a temporal selection. Two of the most recent studies (Bose & Ganguly, 1995 and Hemingway et al., 1998) actually show the largest deviation, both in the absolute position in P–T space and in dP/dT slope. Such a deviation, as inherited by experimental difficulties, is much larger at low temperature and as large as 0.2 GPa at 600 °C. Recently, Walter et al. (2002) obtained an in situ determination of quartz to coesite at only 3 GPa and 1350 °C. Such a position would increase uncertainty to more than 0.4 GPa, a huge number when translated to geodynamic applications. The calculated curve mainly lies in the uppermost region of the experimental range, and below 600 °C it deviates significantly from currently available data. Though it is beyond the goal of this contribution to discuss the details of the experimental problems or of the limits of the optimisation routines, it should be noted that differences in the order of a few hundreds of MPa represent a major uncertainty in terms of the resolution of P–T paths from UHP terrains.

Mineral assemblages in ultrahigh pressure metamorphism

319

Fig. 8. Comparison between different experimental studies on the transformation quartz–coesite and the calculated location according to the database of Holland & Powell (1998 and recent updates). Determination by Bohlen & Boettcher (1982) and Mirwald & Massonne (1980) (reported by Hemingway et al., 1998) and Bose & Ganguly (1995) were all performed in a piston cylinder apparatus. The data by Walter et al. (2002) refer to an in situ estimate performed under a synchrotron source.

Ultrahigh pressure rocks: The graphite–diamond transformation A somewhat different uncertainty affects the transformation of graphite to diamond. The crystallisation of diamond was proved in a variety of media, from molten metals (Sung & Tai, 1997), to sulfur (Sato & Katsura, 2001), carbonates (Pal’yanov et al., 2002), carbonatitic melts (Arima et al., 2002), and C-O-H fluids (Sokol et al., 2001; Yamaoka et al., 2002 and references therein). The lowest P, T conditions for diamond nucleation were found at 5.7 GPa and 1150 °C in an alkaline carbonate C-O-H fluid (Pal’yanov et al., 1999). The currently accepted equilibrium curve for the graphite–diamond transformation still relies on the reversal experiments by Kennedy & Kennedy (1976), performed in the range 1100–1600 °C, and on physical properties of graphite and diamond summarised in Bundy et al. (1996). However, it should be noted that most UHP diamond-bearing terrains attained temperature conditions far below the experimental conditions investigated to date, because of the kinetic constraints imposed by the sluggishness of this transformation. The possible role of fullerene structures (Sundqvist, 1999) at low temperatures in geologically relevant conditions has to be entirely explored.

320

S. Poli & P. Fumagalli

Mafic systems Bimineralic mafic eclogites and the relevance of minor and accessory phases Garnet and clinopyroxene often constitute more than 90% of the mode of mafic eclogites at ultrahigh pressure conditions. Even though such abundance controls relevant physical properties, first of all density, minor and accessory phases are as important as major minerals because of their influence on a variety of geological processes. As examples, we know that physical and chemical properties of rocks, including brittle failure and creep, initiation of melting, generation of volatile-rich magmas responsible for explosive volcanism, geochemical signatures etc., are mostly influenced by the presence of minor (typically hydrates and carbonates) and accessory phases (typically rutile, allanite, zircon, phosphates, ellenbergerite). A first relevant, experimentally documented example concerns the stability field of allanite in mafic eclogites to ca. 4 GPa (Hermann, 2002a). More than 90% of LREE and Th is incorporated in allanite, and allanite is found to be a residual phase during melting. Therefore the geochemical signature of liquids eventually produced in a subducting slab will be strongly controlled by the presence of allanite and by reactions responsible for its breakdown. Furthermore, Hermann (2002a) shows that more than 95% of Ti, Nb and Ta is partitioned in rutile, 95% of Zr and Hf in zircon and P in apatite (Fig. 9, from Hermann, 2002a). Phengite, at a modal value of ca. 10%, incorporates more than 95% of the bulk rock Rb, Ba, and Cs. Such figures clearly demonstrate that careful geochemical analysis cannot leave aside such accessory phases. Chopin et al. (1986) and Brunet et al. (1998) have shown that ellenbergerites and phosphates provide important complementary information to geothermobarometric

Fig. 9. Trace element distribution among the phases of eclogites (from Hermann, 2002a). Modal amounts are 36% omphacite, 35% garnet, 14% quartz, 12% phengite, 2.5% rutile, 0.52% apatite, 0.1% allanite, 0.032% zircon. Reprinted with permission from Elsevier.

Mineral assemblages in ultrahigh pressure metamorphism

321

reconstruction of eclogites. Schmidt (1996), Ono (1998), Domanik & Holloway (1996), Okamoto & Maruyama (1998) found that melting at UHP is controlled by phengite, which is expected to be the only hydrous phase on the solidus of both subducted sediments and mafic eclogites. A number of geophysical studies proposed that intermediate depth earthquakes can be ascribed to a process of dehydration embrittlement. Despite their importance, most reactions involving minor and accessory minerals are still in a state of preliminary investigation because a clear gap exists between the chemical complexity of real rocks and model systems studied in the laboratory. Phase relationships in H2O-bearing systems Most experimental studies performed on mafic systems focussed on phase relationships in hydrated MOR basalt compositions. Because of the large number of chemical components present in these rocks, continuous reactions are the relevant mechanism controlling the appearance and the abundance of mineral phases. Hence, phase transitions observed in MORB + H2O cannot be extrapolated to the universe of gabbroic rocks found in crustal sections (continental or oceanic), often recorded in UHP terrains, due to the strong effects of variable bulk compositions. Figure 10a displays current knowledge to pressures of about 6 GPa. Most experimental studies are substantially consistent and minor, though relevant, discrepancies can be ascribed to different experimental set-ups. All of the experimental studies agree that amphibole breakdown in the system MORB + H2O, both on a simplified chemical composition (Poli, 1993; Schmidt & Poli, 1998) and on natural starting materials (Pawley & Holloway, 1993; Liu et al., 1996) is located at approximately 2.4 GPa between 650 °C and 750 °C. At such conditions amphibole is barroisitic and because of the low reaction rates in the subsolidus region, there is still no information on the mutual relationships between sodic-calcic and sodic amphiboles. Though it has been shown that, in principle, the glaucophane stability field intersects the coesite field (Carman & Gilbert, 1983), there is no experimental demonstration that amphiboles in complex, Cabearing systems extend to more than 2.4 GPa. The temperature region below 650 °C is essentially unexplored, and probably unexplorable, because, despite the presence of a free fluid, even experiments as long as one month do not show appreciable approach to equilibrium conditions. As a consequence we have only a very rough idea of reactions involving chlorite. Nevertheless, because most mafic compositions at high pressure are silica-saturated, as a result of the albite breakdown to form jadeite component and quartz, chlorite breakdown with pressure should be located at pressures comparable to amphibole breakdown and it is not expected to reach the UHP field. On the contrary, olivine-saturated compositions might show a completely different pattern for chlorite (see section devoted to ultramafic systems). Lawsonite and epidote group minerals, which have been known as typical hydrous minerals for relatively shallow hydrothermal conditions, show some of the most interesting relationships in UHP metamorphism. Even though there has been some

322

S. Poli & P. Fumagalli

(a)

(b)

Fig. 10. (a) Review of the experimentally determined phase relationships in MORB eclogites to 5 GPa in the presence of an aqueous fluid. (b) Schematic displacement of invariant point “CaAl” in (a) as a function of the stable garnet composition, which, in turn, is a function of the bulk composition adopted.

Mineral assemblages in ultrahigh pressure metamorphism

323

scepticism about the actual presence of lawsonite in ultrahigh pressure eclogites, recent findings of coesite in the famous lawsonite eclogite xenoliths from the Four-Corner Region (Colorado Plateau, Usui et al., 2003) demonstrate the relevance of these minerals even in tectonic settings that are not expected to be most suitable for such lowtemperature assemblages. The importance of lawsonite is also related to the very high amount of H2O stored in its structure, which makes it of primary interest for geophysicists and geochemists modelling the geodynamics of subduction zone environments. The presence of lawsonite in mafic rocks is controlled by a continuous reaction, with steep positive dP/dT slope, which consumes lawsonite (and clinopyroxene) to form either a relatively grossular-rich garnet (Poli & Schmidt, 1995; Okamoto & Maruyama, 1999), or zoisite (Poli & Schmidt, 1995) at pressures above or below the (pseudo-)invariant point located at ca. 3.2 GPa and 680 °C. This point, as demonstrated by the comparison between different experimental studies, is very sensitive to the bulk composition chosen (Fig. 10b). Lawsonite temperature stability and, secondarily, zoisite pressure stability apparently increase as bulk chemistry moves toward more aluminous compositions. As an example, lawsonite extends its stability to ca. 850 °C at 6 GPa in intermediate (“andesitic”) systems (Poli & Schmidt, 1995) and to more than 900 °C in peraluminous sedimentary compositions (Domanik & Holloway, 1996). It is worth to recall that the invariant point in pure CASH between lawsonite, zoisite, grossular, kyanite and silica lies at about 7 GPa and 1000 °C (Poli & Schmidt, 1998). As a consequence we might expect that the stability of lawsonite and epidote group minerals should be significantly extended in a number of troctolitic gabbros compared to normal MORB compositions. However it should be noted that Liu et al. (1996) did not find either epidote or zoisite in a large pressure range, though lawsonite was observed. Such observation does not seem to be consistent both with the other high pressure experimental studies (Pawley & Holloway, 1993; Poli, 1993) and with natural occurrences, because epidote is found in a variety of eclogite facies rocks. Absence of epidote in the experiments of Liu et al. (1996) might be related to conditions in oxygen fugacity but the inconsistency remains unresolved. Current knowledge about the stability and the phase relations involving hydrous magnesian silicates such as talc, chloritoid and Mg-staurolite is still extremely vague. Even though Pawley & Holloway (1993) and Poli (1993) observed growth of Mg-chloritoid, Liu et al. (1996) and Forneris & Holloway (2001) proposed the metastability of this phase on the basis of long duration reversal experiments (up to almost 100 days long). Since Mg-chloritoid is occasionally found in metamorphosed high pressure Mg-gabbros, one possible explanation for this apparent discrepancy is the bulk composition chosen, which is much higher in normative olivine in Poli & Schmidt (1995) and Pawley & Holloway (1993). Another possible, additional reason, is a drift in reactive bulk composition with time. Because Mg-rich minerals are present in the starting materials used (e.g. tremolite in Poli, 1993), nucleation of Mg-chloritoid might be promoted on these localised chemical systems, similarly to natural rocks, but diffusion with time might lead to resorption of previously formed crystals. Further experiments on Mg-enriched compositions should be performed to demonstrate this hypothesis and to verify the stability of Mg-chloritoid in mafic rocks at UHP conditions.

324

S. Poli & P. Fumagalli

Similarly, talc, though not ubiquitous in experimental products, is expected to play an important role in differentiated gabbros and, again, it is expected to be sensitive to the bulk Mg/(Mg + Fe) in the system. The fate of talc with pressure is closely related to the stability of the 10 Å phase chemically analogous to talc, which will be discussed in more detail in the section devoted to ultramafic systems. Finally, we mention the occurrence of Mg-staurolite in various high pressure studies at conditions close to the wet solidus of mafic systems. Despite the apparent rarity of Mg-staurolite in natural rocks, its frequency is remarkable in quite different experimental configurations, which might suggest that this mineral has been overlooked in nature. When enough potassium is present in the system (this can be the case for altered oceanic basalts or for impure mafic rocks derived from sedimentary sequences) phengite is a characteristic ultrahigh pressure mineral. Schmidt (1995) and Schmidt & Poli (1998) described the stability of phengite in both mafic and intermediate systems where the celadonite buffering assemblage is garnet + omphacite + silica. For such an assemblage a progressive increase in Si with pressure and decrease with temperature is observed. Celadonite end-member composition is achieved at ca. 9 GPa close to phengite s.s. breakdown to K-hollandite. Phengite is the hydrous mineral with the largest stability field observed in mafic systems and because it is also the only hydrate on the solidus above the epidote breakdown, phengite controls melting and geochemical signatures of first melts. Unfortunately, despite considerable effort performed both in model (e.g. Massonne & Szpurka, 1997; Hermann, 2002b) and in more complex systems (Schmidt & Poli, 1998; Schmidt, 1996; Okamoto & Maruyama, 1998), there is still no complete formulation of a phengite geothermobarometer available for mafic eclogites at ultrahigh pressure conditions. Phase relationships in a C-O-H-bearing system The effect of the presence of a C-O-H mixed fluid on phase relationships in mafic systems at high pressure subsolidus conditions are substantially unknown to date. Only a few reconnaissance works on simple systems (see Luth, 1999, and references therein) or on more complex systems approaching natural rocks (Yaxley & Green, 1994; Molina & Poli, 2000) are currently available. Experimental difficulties arise from the extremely complex interplay of mass balance constraints relating the solid assemblage (including hydrates and carbonates), the amount of fluid present, the speciation of the fluid, the iron oxidation state, the physical state of carbon (graphite/diamond vs. carbon-bearing volatiles) etc. Phase relations at UHP conditions are further complicated by the enlargement of the stability field of graphite/diamond with increasing pressure and/or decreasing temperature in a diagram log fO2 vs. P or T. The stippled field in Figure 11 shows the potential coexistence of graphite and carbonates at 680 °C and 1300 °C. This field is realistically bounded at high oxygen fugacities by the equilibrium graphite + O2 = CO2 (GCO), whereas at low oxygen fugacity by the ultimate decomposition of carbonates to form graphite + periclase + silicates. If oxygen fugacities represented by the equilibrium NNO (Ni–NiO) or FMQ (fayalite–magnetite–quartz) are assumed to be a middle course

Mineral assemblages in ultrahigh pressure metamorphism

325

Fig. 11. The stability of graphite (gph) with carbonates (stippled field) at high and low temperature conditions. FMQ: fayalite-magnetite-quartz; FFsM: fayalite-ferrosilite-magnetite; NNO: nickel-bunsenite; GCO: graphite-CO2; EMOG: enstatite-magnesite-olivine-graphite; DSDG: dolomite-quartz/coesite-diopsidegraphite. See text for explanation.

326

S. Poli & P. Fumagalli

of conditions at high pressure, calculations in Figure 11 suggest that phase assemblages of UHP terrains are expected to present either graphite/diamond only or graphite/ diamond + carbonates or carbonates only as a function of the reactive bulk composition, i.e. of the relative availability of carbon, hydrogen and oxygen. Molina & Poli (2000) performed a series of piston cylinder experiments to 2.0 GPa and temperatures to 730 °C on a tholeiite composition in the presence of a mixed fluid produced by the decomposition of oxalic acid dihydrate, at oxygen fugacities externally buffered by NNO. Amphibole was found to coexist with calcite at P @ 1.4 GPa, with dolomite at 1.4 @ P @ 1.8 GPa, and with dolomite + magnesite at pressures higher than 1.8 GPa (Fig. 12). Garnet, paragonite, kyanite and epidote participate to complex reactions with carbonates buffering the fluid. Estimates of the coexisting fluid compositions, on the basis of mass balance and thermodynamic calculations, demonstrate a continuous H2O enrichment with increasing pressure and decreasing temperature. An almost purely aqueous fluid is obtained at 2 GPa and 665 °C. This implies that carbon tends to

Fig. 12. Review of the experimentally determined phase relationships in MORB eclogites to 6 GPa in the presence of a C-O-H mixed fluid at variable oxygen fugacities and bulk C-O-H contents.

Mineral assemblages in ultrahigh pressure metamorphism

327

fractionate into the solid with increasing pressure. However, the nature of such a solid assemblage, as previously stated, cannot be straightforward predicted. For this purpose it is useful to compare the results obtained by Yaxley & Green (1994) with assemblages found by Molina & Poli (2000) and by Crottini et al. (2002) (Fig. 12). Even though the stability of hydrous phases demonstrates a substantial consistency between these different works, carbon-bearing phases are extremely variable. Amphibole breakdown in C-O-H bearing MORB is found at 2.5–2.6 GPa and the fluid saturated solidus is located at ca. 730 °C at 2.2 GPa. These observations unequivocally indicate that the coexisting fluids have to be enriched in H2O component, despite the variable amounts of carbon introduced in the system for the variable experimental strategies adopted in these studies. On the contrary, Yaxley & Green (1994) observes the appearance of dolomite only above 2.5 GPa whereas Crottini et al. (2002) in the uppermost pressure region obtained graphite only, aragonite + dolomite + graphite or aragonite + graphite as a function of both the oxygen buffer and of the amount of C-O-H added to the system. Given currently available experiments it is therefore a challenging task to predict carbonate and hydrate-bearing assemblages at UHP conditions unless specific experiments are performed. It is worth noting that theoretical calculations are of limited help in such a case, because most petrologists assume that C-O-H mixed fluids can be approximated along the join H2O–CO2 (Kerrick & Connolly, 2001) which is evidently not the case, as graphite or diamond are widely recognised around the world in ultrahigh pressure terrains. A few experimental studies in very simple model systems offer a useful reference frame, even though application to natural complex systems should be cautious. As an example the breakdown of dolomite with pressure (Luth, 2001) can be applied to a variety of bulk compositions, from ultramafic to mafic and intermediate eclogites. However, as shown in Figure 12 (compare dolomite-in in CMS and in experiments of Yaxley & Green, 1994), such model reactions may differ enormously from natural systems in P–T location because carbonate species are controlled by a variety of mass balance and thermodynamic constraints, first fluid speciation.

Ultramafics Peridotite compositions Although peridotites are typically characterised by a rather simple phase assemblage constituted of olivine, orthopyroxene, clinopyroxene and an Al-phase, i.e. plagioclase, spinel or garnet depending on pressure (see Ulmer & Trommsdorff, 1999 for a review on mantle mineralogy in simple systems), mantle phases may partake to complex solid solutions by considering both compositions approaching natural peridotites, i.e. complex systems and natural materials. As a result phase equilibria are mainly governed by continuous reactions and, again, strongly depend on bulk compositions. Model natural peridotite compositions are mainly derived by variably enriched Hawaiian xenoliths (spinel and garnet lherzolites from Hawaiian xenoliths; Mysen & Boettcher, 1975), fertile spinel lherzolites (KLB-1; Hirose & Kushiro, 1993) and peridotites which are depleted in incompatible elements, but not depleted in Ca and Al

328

S. Poli & P. Fumagalli

such as the Tinaquillo lherzolite (Robinson & Wood, 1998). Furthermore Green (1973) calculated a model theoretical peridotite composition, the MORB pyrolite, combining a primitive basalt with a harzburgite residue and used such a pyrolitic composition as starting material to investigate the amphibole stability at subsolidus and near-solidus conditions (Niida & Green, 1999). The MORB pyrolite is an intermediate member between fertile lherzolites (Hawaiian xenoliths) and the Tinaquillo lherzolite. In order to investigate the effect of potassium in mantle petrology, i.e. the mantle metasomatism due to the interaction with alkali rich fluids, natural bulk compositions have been doped with phlogopite or K-amphibole (Northern Depression Hessian peridotite + 1.5% phlogopite, Mengel & Green, 1989; Mont Briançon, Massif Central, France, spinel lherzolite plus 5 wt% phlogopite or plus 10 wt% K-amphibole, Konzett & Ulmer, 1999; KLB-1 lherzolite plus 10 wt% K-amphibole, Konzett & Fei, 2000). The spinel to garnet transition Among phase equilibria occurring at increasing pressure in peridotite systems, the spinel to garnet transition was one of the first tools to recognise UHP metamorphism in many ultramafic xenoliths and orogenic peridotites. Available data on complex ultramafic compositions are represented in Figure 13 and compared with most recent results on the spinel to garnet transition in the simple system CaO–MgO–Al2O3–SiO2 – CMAS (Klemme & O’Neill, 2000; Walter et al., 2002). Niida & Green (1999) located the spinel to garnet transition in a MORB pyrolite at 2.0 GPa and 1050 °C, which is in perfect agreement with the higher temperature, near-solidus data obtained in a dry MORB pyrolite by Robinson & Wood (1998). In the simple model system CaO–MgO–Al2O3–SiO2 the spinel–garnet transition is governed by the univariant subsolidus equilibrium orthopyroxene + clinopyroxene + spinel = forsterite + garnet. Walter et al. (2002) performed in situ X-ray experiments in order to investigate and solve many discrepancies for the high temperature position (T > 1200 °C) of the spinel–garnet boundary reported in previous results. Although Walter et al. (2002) focussed on a simple model system and on high temperature regions, the innovative approach of in situ reconnaissance of subsolidus phase assemblages is worth of mention as reference for the CMAS system. In order to investigate the effect of composition and kinetic metastability of phases, experiments were performed starting from mixtures of clinopyroxene, garnet and forsterite and orthopyroxene, clinopyroxene, garnet and forsterite, and the phase boundary was approached from both the high and low-pressure sides. Walter et al. (2002) drew two alternative curves: curve 1 and the shaded area in Figure 13 represent the phase boundary and its error while curve 2 takes into account a suspicious occurrence of spinel + orthopyroxene at about 1000 °C not solved yet in terms of possible metastability. The slope of curve 1, which is in agreement with reversals of Klemme & O’Neill (2000), should therefore represent a maximum. Adding further elements to the simple system, such as Fe2+, Cr and Na, i.e. considering a complex model peridotite system, substantially influences the position of

Mineral assemblages in ultrahigh pressure metamorphism

329

Fig. 13. The spinel to garnet transition in the simple system CaO–MgO–Al2O3–SiO2 – CMAS (Klemme & O’Neill, 2000; Walter et al., 2002) and in model peridotites (MORB-pyrolite and Tinaquillo lherzolites; Niida & Green, 1999; Robinson & Wood, 1998). Curves 1 and 2 represent two alternative phase boundaries proposed by Walter et al., 2002 (see text for further details).

330

S. Poli & P. Fumagalli

Fig. 14. Experimentally determined phase diagram for hydrous peridotites.

the spinel to garnet transition. Since Fe2+ is preferentially partitioned into garnet, its stability field is expanded towards lower pressure. For a composition with a typical mantle Mg value (XMg = Mg/(Mg + Fe) = 0.9), O’Neill (1981) found that Fe2+ tends to lower the transition of about 0.2–0.3 GPa. Chromium has the opposite effect: it expands the stability of spinel-bearing assemblages because of the preferential partitioning of Cr into spinel. As a result the spinel to garnet transition is not only shifted towards higher pressure as the ratio Cr/Cr + Al in the bulk rock increases, but the pressure and temperature field over which the reaction takes place is greatly broadened, i.e. the transition is a continuous reaction. For a ratio Cr/Cr + Al of about 0.1, Webb & Wood (1986) estimated that, if spinel and garnet are the only Cr-bearing phases, the transition should spread out over a pressure of about 1 GPa. However in natural lherzolite or pyrolite compositions (Na, Cr, and Fe-bearing systems) the occurrence of clinopyroxene may substantially change this prediction. At low mole fractions of Cr (XCr = Cr/(Cr + Al) of about 0.1) the NaCrSi2O6 component in clinopyroxene buffers the Cr content of spinel, sharpening the transition and reducing to only 0.2 GPa the pressure range over which spinel and garnet may coexist (Webb & Wood, 1986). In more refractory compositions, i.e. at higher XCr values, the chromium partitioning between

Mineral assemblages in ultrahigh pressure metamorphism

331

clinopyroxene and spinel is such that spinel composition is not affected by clinopyroxene and the Cr-spinel + garnet assemblage may survive at higher pressure. The Bardane peridotites in Western Norway testify such an ultrahigh pressure phase assemblage (Cr-spinel, garnet and diamond). Although the width of the continuous reaction has been quantitatively evaluated both in the absence and presence of clinopyroxene, the entity of shifting towards higher pressure is still poorly constrained. However, experiments performed both in natural peridotite compositions (Tinaquillo lherzolite; Robinson & Wood, 1998) and complex model peridotites (MORB pyrolite; Robinson & Wood, 1998; Niida & Green, 1999), i.e. in Fe, Cr, Na-bearing systems, suggest that the effect of chromium predominates over the effect of Fe2+, locating the spinel to garnet transition at pressures substantially higher than those observed in the simple system (CMAS). The peridotite + H2O system The hydration of peridotites along mid-ocean ridges has widely been accepted. Many efforts have therefore been devoted to unravelling phase relations in hydrated ultramafic systems, where hydrous minerals such as antigorite, chlorite, amphibole, and a variety of Dense Hydrous Magnesium Silicates may persist under subsolidus conditions at high and ultrahigh pressure (Kawamoto et al., 1996; Ulmer & Trommsdorff, 1999; Fumagalli, 2000). Phase relations in hydrous peridotites are of primary importance in depicting the dehydration evolution of a subducting slab, the metamorphic and igneous history of the mantle wedge at convergent margins and the petrology of deep mantle rocks. It is worth noting that, although many experimental studies have been devoted to investigate near-solidus and supersolidus phase relations in natural systems approaching ultramafic compositions, subsolidus equilibria are still fragmentary. Figure 14 shows a summary of H2O-saturated phase relations in ultramafic rocks. Antigorite dominates the low temperature field up to a pressure of 6.0 GPa (Ulmer & Tromsdorff, 1999). Aluminium leads to a wide occurrence of chlorite from greenschists to eclogite facies. Amphibole is stable in a wide range of temperature conditions although it shows a temperature dependent mineral chemistry: a calcic amphibole occurs at low temperatures, a relatively Al-poor, tremolitic amphibole coexists with chlorite and, as temperature further increases, above the thermal stability of chlorite (750–800 °C), a NaAl-rich pargasitic amphibole is stable up to the solidus (Niida & Green, 1999). Amphibole pressure stability is dependent on bulk compositions, mainly on the Na and Ca bulk content, i.e. on fertility. Mysen & Boettcher (1975) found that at 1000 °C amphibole breaks down at 2.2 GPa in harzburgites, at 2.5–2.8 GPa in lherzolites and at 2.8–3.0 GPa in pyrolites. Experiments performed at H2O-undersaturated conditions on MORB pyrolite by Niida & Green (1999) corroborated the hypothesis of the effect of bulk rock alkali content on the amphibole stability: in MORB pyrolites the maximum stability of pargasite lies at 2.8 GPa and 1000 °C, intermediate conditions between those of the Tinaquillo lherzolite (2.6 GPa, 1000 °C; Wallace & Green, 1991) and those of the Hawaiian pyrolite and the Northern Hessian Depression (2.9–3.0 GPa, 1000 °C and 2.8–2.9 GPa, 1000 °C, respectively). Niida & Green (1999) also showed that the

S. Poli & P. Fumagalli

332

maximum temperature stability expands from 1025 to 1150 °C with increasing bulk alkali contents. The higher temperature is related to the higher potassium content (as in the Northern Hessian Depression peridotite modified with 1.5% phlogopite, K2O = 0.43 wt%) which would stabilise amphibole to higher temperatures. At subsolidus conditions pargasitic amphibole and clinopyroxene are the only Nabearing phases. Niida & Green (1999) investigated systematically the composition of amphibole and coexisting phases in a constant bulk composition, the MORB pyrolite, from 0.4 to 3.0 GPa and between 925 and 1100 °C and found that amphibole breaks down through complex continuous reactions. Based on mass balance calculations and mineral chemistry they suggested that the ultimate breakdown of amphibole is related to the destabilisation of the richteritic component. The ultimate breakdown of amphibole at high temperature occurs as a dehydration reaction, providing therefore a release of fluid. This is not the case at temperatures lower than 800 °C. Fumagalli (2000) performed experiments at H2O saturated conditions on model lherzolite compositions (Na2O–CaO–FeO–MgO–Al2O3–SiO2–H2O, NCFMASH system) and found that, while at temperatures higher than 800 °C a dehydration reaction governs the amphibole breakdown, at temperatures lower than 800 °C, its pressure stability is controlled by the degenerate H2O conserving reaction: amphibole + olivine = orthopyroxene + clinopyroxene + chlorite. Chlorite thermal stability, related to the reaction chlorite + enstatite = garnet + olivine + water, results depressed towards lower temperatures in the peridotite model system as compared to the MASH system due to the preferential partitioning of iron into garnet (Fumagalli, 2000). The pressure stability of chlorite in complex ultramafic systems is still poorly explored and defined. Available experiments performed in the model peridotite system NCFMASH have shown that the pressure stability of chlorite is controlled by the appearance, at temperatures lower than 700 °C, of a 10 Å phase structure. The 10 Å phase, first synthesised in the MgO–SiO2–H2O system by Sclar et al. (1965), is a phyllosilicate chemically analogous to talc but with excess water: the chemical formula may be written as Mg3Si4O10(OH)2·nH2O. However, in the complex ultramafic system a 10 Å phase structure, identified on the basis of X-ray diffractometry, shows, at the electron microprobe scale, a homogeneous aluminous composition (Fumagalli & Poli, 1999; Fumagalli et al., 2001). The surprising coincidence in composition between this phase and the mixed-layered mineral kulkeite (clinochlore : talc = 1 : 1; Schreyer et al., 1982), may suggest a structural rearrangement of chlorite, intercalating the aluminiumfree 10 Å phase through a continuous reaction. The thermal stability of the 10 Å phase structure is governed by the reaction at T > 700 °C: 10 Å phase + clinopyroxene = garnet + orthopyroxene + H2O. The K-peridotite system The relevance of metasomatic processes in subduction zones, affecting peridotites of the mantle wedge, has been recognised since a long time. Even though such effects have

Mineral assemblages in ultrahigh pressure metamorphism

333

mainly been addressed on mantle xenoliths incorporated in magmatic suites, peridotites “intruded” in subducted continental crust provide a unique petrological window on processes undergoing at the slab–mantle interface. Buoyancy forces are responsible for peridotite emplacement in the subducted continental crust; the devolatilisation of felsic rocks and mass transfer toward peridotite bodies are strongly enhanced. A variety of volatile-bearing phases develop in ultramafic bodies and recent findings of Tiphlogopite and diamond as inclusions in garnet peridotites from Bardane (Norway) testify to the intensity of mass exchange at depth (van Roermund et al., 2002). Although data exist on the stability of pure phases such as phlogopite and K-amphibole (Trønnes, 2002, and references therein), available experimental data on phase relations in potassium doped model peridotites are still limited to just a few studies (Konzett & Ulmer, 1999; Konzett & Fei, 2000). Phase relations (summarised in Fig. 15) are governed by the occurrence of three potassic phases, which are, in the order of pressure stability: phlogopite, a potassic amphibole of richteritic composition, and a Krich hydrous silicate, termed phase X (K2–xMg2Si2O7Hx, x = 0–1, Luth, 1997). K-amphibole represents the breakdown product of phlogopite-bearing assemblages and phase X represents the breakdown product of K-amphibole. Sudo & Tatsumi (1990), investigating a phlogopite + diopside ± enstatite composition in the KCMASH system, suggested that phlogopite stability is governed by the reaction: phlogopite + diopside ± enstatite = K-amphibole + garnet + forsterite + fluid. In the absence of orthopyroxene the reaction is divariant in the KCMASH system and therefore a field exists where phlogopite and K-amphibole coexist: the lower boundary represents the K-amphibole-in reaction (K-amp-in ST90) while the upper limit is the phlogopite out reaction (Phl-out ST90). In the presence of orthopyroxene, however, the reaction becomes univariant in the KCMASH system and phlogopite is not stable at pressures above 6 GPa (Sudo & Tatsumi, 1990). The potassic amphibole, which forms at the expense of phlogopite, is an Al-poor potassic amphibole termed KK-richterite, KKCaMg5Si8O22(OH)2. In the Na-bearing system sodium will contribute to the formation of K-richterite, KNaCaMg5Si8O22(OH)2, as a breakdown product of phlogopite. Konzett & Ulmer (1999) investigated the KNCMASH system using bulk compositions with an excess of phlogopite over orthopyroxene, in order to determine the maximum pressure stability of phlogopite in K-richterite bearing assemblages. Their results show that K-amphibole appears between 6 and 6.5 GPa at 800 °C and between 6.5–7.0 GPa at 1100 °C (Kr-in KU99). Konzett & Ulmer (1999) underlined a discrepancy in the slope of the K-amphibolein reaction as compared with what Sudo & Tatsumi (1990) determined. Whether the slope is positive or negative is still under debate. Below the K-amphibole-in reaction the stable assemblage is constituted of phlogopite, clinopyroxene, orthopyroxene, garnet and fluid, and close to the reaction an olivine-bearing assemblage occurs. Above the K-amphibole-in reaction, phlogopite and K-amphibole coexist due to the excess of phlogopite respect to orthopyroxene up to the ultimate breakdown of phlogopite at pressures between 8 and 9 GPa, 1100 °C (Phl-out

334

S. Poli & P. Fumagalli

Fig. 15. Summary of available experimental data in K-doped peridotites. KU99: Konzett & Ulmer, 1999; ST90: Sudo & Tatsumi, 1990; KF00: Konzett & Fei, 2000; L97: Luth, 1997; I98: Inoue et al., 1998. Phl: phlogopite; K-amp: K-amphibole.

Mineral assemblages in ultrahigh pressure metamorphism

335

KU99). At pressures above 13–14 GPa (1100 °C) phase X replaces K-amphibole (Kamphibole-out reaction). As the complexity of the system investigated increases, the breakdown of K-amphibole occurs at lower pressure. The K-amphibole to phase X transition involves continuous changes in garnet compositions through Ca–Mg exchange and the limited solubility of MgSiO3 to form majoritic component (Konzett & Fei, 2000). The thermal stability of K-amphibole is determined by the appearance of the anhydrous assemblage garnet, olivine, orthopyroxene and clinopyroxene (K-amp-out KU99). In the KNCMASH system it occurs between 1300 and 1400 °C at 8.0 GPa and shows a positive slope. In an Fe-bearing system and lherzolitic composition, the effect of iron has to be taken into account. Konzett & Ulmer (1999) investigated a K-doped lherzolite modifying the composition of the Mont Briançon lherzolite (Massif Central, France) by adding phlogopite or K-richterite (see Table 1). In the lherzolitic system the K-amphibole-in reaction slightly shifts towards lower pressure (between 6.0 and 6.5 GPa at 1100 °C) due to the preferential Fe2+ partitioning into garnet, which is a product of phlogopite breakdown. The coexistence of phlogopite and K-amphibole is, however, reduced to less than 1 GPa. It is worth noting that care should be taken in dealing with Fe-bearing systems due to the not easily predictable influence of the Fe3+/Fe2+ ratio on phase relations. The K-amphibole thermal stability in the Fe-bearing lherzolitic system seems influenced again by oxidation reactions: the slope suggested by experimental data from Konzett & Ulmer (1999) is indeed opposite compared to what it was predicted in the Fefree system. An explanation given by the authors refer to the pressure and temperature dependent oxidation of graphite inner capsule which, producing CO2, would lower the H2O activity and destabilise K-amphibole. The pressure stability of K-amphibole in Fe-bearing lherzolitic systems was investigated by Konzett & Fei (2000): they run experiments by using a K-amphibole doped peridotite KLB1 from 12 to 14 GPa and 1200 °C. Potassic phases, either Kamphibole or phase X, coexist with garnet, low-Ca clinopyroxene, high-Ca clinopyroxene and Mg2SiO4. In the Fe-bearing system the K-amphibole to phase X transition, occurring between 12 and 13 GPa at 1200 °C, is shifted by about 1.0 GPa towards lower pressure as compared with what was found in the Fe-free system. Metasedimentary rocks Despite the fact that most UHP terrains have been unequivocally identified also in metasedimentary compositions, and that metapelites and metagreywackes give a unique geochemical signature to orogenic magmas, high-pressure experiments are still in a state of a preliminary investigation. Most published systematic work within the coesite stability field has been performed in the model systems KMASH, KFASH (Massonne & Szpurka, 1997) and KCMASH (Hermann, 2002b). Domanik & Holloway (1996) and Ono (1998) conversely performed experiments above 6 GPa, but on natural sedimentary compositions approaching metapelites. Ferri et al. (2000) and Poli & Schmidt (2002) presented data on experiments in KCFMASH to 2.7 GPa for bulk compositions representative of metapelites. Therefore the general picture available for the

336

S. Poli & P. Fumagalli

interpretation of natural cases is still fragmentary. The assemblage garnet-phengitekyanite-coesite-clinopyroxene is predominant in most pelitic compositions, as biotite is completely consumed by the reaction biotite + kyanite + SiO2 = garnet + phengite (Hermann, 2002b; Poli & Schmidt, 2002), unless unusually high XMg bulk compositions are considered. Appearance of the join talc–garnet at ca. 2.4–2.5 GPa in the model system KCFMASH should be responsible for the widespread occurrence of talc in most natural bulk compositions (Ferri et al., 2000; Poli & Schmidt, 2002), but the fate of talc with pressure and its relationships with chloritoid are still unknown. Most experimental studies have been performed at fluid saturation, and at such conditions K-feldspar was never found to be stable either in metapelites or in metagreywackes. It is therefore only at fluid-undersaturated conditions that K-feldspar is expected to be a possible UHP phase in metasedimentary material. However, as illustrated above, attainment of fluid-undersaturation during a prograde, subduction related P–T path for metasedimentary (initially H2O-rich) material is unlikely, and therefore the relevance of K-feldspar is expected to be very limited. In conclusion a number of questions remains unresolved: N What is the pressure–temperature stability of the join garnet–talc? N What is the pressure stability of biotite solid solution? N What is the fate of magnesium-rich chloritoid? N Which reactions control the behaviour of clinopyroxene in common metapelites and metagreywackes? N Are indeed some unusual high-pressure phases found in experimental studies (Mgsursassite, topaz-OH, phase Pi) relevant for UHP rocks and which reactions govern their appearance? Okamoto & Maruyama (1998) and Schmidt & Poli (1998) revealed that potassium enters in clinopyroxene and sodium in garnet with increasing pressure; Hermann (2002b) demonstrated significant deviation of phlogopite toward talc and of clinopyroxene toward Ca-Eskola component. Such features are responsible for the peculiar intragranular textures found in UHP rocks, such as K-feldspar lamellae in clinopyroxene or silica rods in omphacite (Liou et al., 1998). Given all of the complexities present in these bulk compositions and their importance for geodynamic processes, a particular effort has to be done in the near future to provide a reliable tool for the interpretation of metasedimentary rocks.

References Akaogi, M., Kojitani, H., Matsuzaka, K., Suzuki, T. & Ito, E. (1998): Postspinel transformations in the system Mg2SiO4-Fe2SiO4: element partitioning, calorimetry, and thermodynamic calculation. In Manghnani, M.H. & Yagi, T. (eds.): Properties of Earth and planetary materials at high pressure and temperature /Geophys. Monogr., 101/. Washington (D.C.): Am. Geophys. Union, 373–391. Akaogi, M., Hamada,Y., Suzuki, T., Kobayashi, M. & Okada, M. (1999): High pressure transitions in the system MgAl2O4-CaAl2O4: a new hexagonal aluminous phase with implication for the lower mantle. Phys. Earth Planet. Inter., 115:67–77. Akaogi, M., Tanaka, A. & Ito, E. (2002): Garnet-ilmenite-perovskite transitions in the system Mg4Si4O12Mg3Al2Si3O12 at high pressures and high temperatures: phase equilibria, calorimetry and implications for mantle structure. Phys. Earth Planet. Inter., 132:303–324. Angel, R.J., Frost, D.J., Ross, N.L. & Hemley, R. (2001): Stabilities and equations of state of dense hydrous magnesium silicates. Phys. Earth Planet. Inter., 127:181–196.

Mineral assemblages in ultrahigh pressure metamorphism

337

Arima, M., Kozai, Y. & Akaishi, M. (2002): Diamond nucleation and growth by reduction of carbonate melts under high-pressure and high-temperature conditions. Geology., 30:691–694. Berman, R.G. (1991): Thermobarometry using multi-equilibrium calculations: a new technique, with petrological applications. Can. Mineral., 29:833–855. Boettcher, A.L. & Wyllie, P.J. (1969): The system CaO-SiO2-CO2-H2O – III. Second critical end-point on the melting curve. Geochim. Cosmochim. Acta., 33:611–632. Bohlen, S.R. & Boettcher, A.R. (1982): The quartz-coesite transformation: a precise determination and the effects of other components. J. Geophys. Res., 87:7073–7078. Bose, K. & Ganguly, J. (1995): Quartz-coesite transition revisited – Reversed experimental determination at 500–1200°C and retrieved thermochemical properties. Am. Mineral., 80:231–238. Bromiley, G.D. & Pawley, A.R. (2002): The high-pressure stability of Mg-sursassite in a model hydrous peridotite: a possible mechanism for the deep subduction of significant volumes of H2O. Contrib. Mineral. Petrol., 142:714–723. Brunet, F., Chopin, C. & Seifert, F. (1998): Phase relations in the MgO-P2O5-H2O system and the stability of phosphoellenbergerite: petrological implications. Contrib. Mineral. Petrol., 131:54–70. Brunsmann, A., Franz, G., & Heinrich, W. (2002): Experimental investigation of zoisite-clinozoisite phase equilibria in the system CaO-Fe2O3-Al2O3-SiO2-H2O. Contrib. Mineral. Petrol., 143:115–130. Bundy, F.P., Bassett, W.A., Weathers, M.S., Hemley, R.J., Mao, H.K. & Goncharov, A.F. (1996): The pressuretemperature phase and transformation diagram for carbon; Updated through 1994. Carbon, 34(2):41–153. Bureau, H. & Keppler, H. (1999): Complete miscibility between silicate melts and hydrous fluids in the upper mantle; experimental evidence and geochemical implications. Earth Planet. Sci. Lett., 165:187–196. Canil, D. (1994): Stability of clinopyroxene at pressure-temperature conditions of the transition region. Phys. Earth Planet. Inter., 86:25–34. Carman, J.H. & Gilbert, C.M. (1983): Experimental studies on glaucophane stability. Am. J. Sci., 283:414–437. Chopin, C., Klaska, R., Medenbach, O. & Dron, D. (1986): Ellenbergerite, a new high pressure Mg-Al-(Ti,Zr)silicate with a novel structure based on face-sharing octahedra. Contrib. Mineral. Petrol., 92:316–321. Connolly, J.A.D. (1995): Phase diagrams methods for graphitic rocks and application to the system C-O-HFeO-TiO2-SiO2, Contrib. Mineral. Petrol., 119:94–116. Crottini, A., Molina Palma, F. & Poli, S. (2002): Carbonates vs. graphite stability in hydrous mafic systems at high pressure. J. Conf. Abstr., 7(1):26. Daniels, P. & Wunder, B. (1996): Al3Si2O7(OH)3, phase Pi (formerly piezotite): Crystal structure of a synthetic high-pressure silicate rediscovered. Eur. J. Mineral., 8:1283–1292. Domanik, K.J. & Holloway, J.R. (1996): The stability and composition of phengitic muscovite and associated phases from 5.5 to 11 GPa; implications for deeply subducted sediments. Geochim. Cosmochim. Acta., 60:4133–4150. Ferri, F., Gregnanin, A. & Poli, S. (2000): An experimentally determined petrogenetic grid for metapelites at medium-high pressure. J. Conf. Abstr., 5(1):34. Fockenberg, T. (1998): An experimental study of the pressure-temperature stability of MgMgAl-pumpellyite in the system MgO-Al2O3-SiO2-H2O. Am. Mineral., 83:220–227. Forneris, J. & Holloway, J.R. (2001): Experimental determination of chloritoid stability in subducting oceanic crust. Eos, Trans. Am. Geophys. Union., 82(47):Fall Meet. Suppl., Abstract, T22D–09. Frost, D.J., Langenhorst, F. & van Aken. P.A. (2001): Fe-Mg partitioning between ringwoodite and magnesiowüstite and the effect of pressure, temperature and oxygen fugacity. Phys. Chem. Miner., 28:455–470. Fumagalli, P. (2000): Processi di trasporto e rilascio di H2O nelle zone di subduzione: uno studio sperimentale su sistemi ultrafemici ad alta pressione. PhD thesis. Univ. Milano, Italy, 178 p. Fumagalli, P. & Poli, S. (1999): Phase relationships in hydrous peridotites at high pressure: preliminary results of multianvil experiments. Period. Mineral., 68:275–286. Fumagalli, P., Stixrude, L., Poli, S. & Snyder, D. (2001): The 10Å phase: a high-pressure expandable sheet silicate stable during subduction of hydrated lithosphere. Earth Planet. Sci. Lett., 186:125–141. Gasparik, T. (1996a): Melting experiments on the enstatite-diopside join at 70–224 kbar, including the melting of diopside. Contrib. Mineral. Petrol., 124:139–153. Gasparik, T. (1996b): Diopside-jadeite join at 16–22 GPa. Phys. Chem. Miner., 23:476–486. Gasparik, T., Wolf, K. & Smith, C.M. (1994): Experimental determination of phase relations in the CaSiO3 system from 8 to 15 GPa. Am. Mineral., 79:1219–1222.

338

S. Poli & P. Fumagalli

Green D.H. (1973): Experimental melting studies on a model upper mantle composition at high pressure under water-saturated and water-undersaturated conditions. Earth Planet. Sci. Lett., 19:37–53. Heinemann, S., Sharp, T.G., Seifert, F. & Rubie, D.C. (1997): The cubic-tetragonal phase transition in the system majorite (Mg4Si4O12) pyrope (Mg3Al2Si3O12), and garnet symmetry in the Earth’s transition zone. Phys. Chem. Miner., 24:206–221. Hemingway, B.S., Bohlen, S.R., Hankins, W.B., Westrum, E.F., Jr., & Kuskov, O.L. (1998): Heat capacity and thermodynamic properties for coesite and jadeite, reexamination of the quartz-coesite equilibrium boundary. Am. Mineral., 83:409–418. Hermann, J. (2002a): Allanite: thorium and light rare earth element carrier in subducted crust. Chem. Geol., 192:289–306. Hermann, J. (2002b): Experimental constraints on phase relations in subducted continental crust. Contrib. Mineral. Petrol., 143:219–235. Hirose, K. & Kushiro, I. (1993): Partial melting of dry peridotites at high pressures: determination of compositions of melts segregated from peridotite using aggregates of diamond. Earth Planet. Sci. Lett., 114:477–489. Holland, T.J.B. & Powell, R. (1998): An internally consistent thermodynamic data set for phases of petrological interest. J. Metamorph. Geol., 16:309–343. Inoue, T., Irifune, T., Yurimoto, H. & Miyagi, I. (1998): Decomposition of K-amphibole at high pressures and implications for subduction zones volcanism. Phys. Earth Planet. Inter., 107:221–231. Kawamoto, T., Hervig, R.L. & Holloway, J.R. (1996): Experimental evidence for a hydrous transition zone in the early Earth’s mantle. Earth Planet. Sci. Lett., 143:587–92. Kennedy, C.S. & Kennedy, G.C. (1976): The equilibrium boundary between graphite and diamond. J. Geophys. Res., 81:2467–2470. Kennedy, G.C., Wasserburg, G.J., Heard, H.C. & Newton, R.C. (1962): The upper three-phase region in the system SiO2-H2O. Am. J. Sci., 260:501–521. Kerrick, D.M. & Connolly, J.A.D. (2001): Metamorphic devolatilization of subducted oceanic metabasalts: implications for seismicity, arc magmatism and volatile recycling. Earth Planet. Sci. Lett., 189:19–29. Klemme, S. & O’Neill, H.St.C. (2000): The near-solidus transition from garnet lherzolite to spinel lherzolite. Contrib. Mineral. Petrol., 138:237–248. Konzett, J. & Fei, Y. (2000): Transport and storage of potassium in the Earth’s upper mantle and transition zone: and experimental study to 23 GPa in simplified and natural bulk compositions. J. Petrol., 41:583–603. Konzett, J. & Ulmer, P. (1999): The stability of hydrous potassic phases in lherzolite mantle; an experimental study to 9.5 GPa in simplified and natural bulk compositions. J. Petrol., 40:629–652. Liou, J.G., Zhang, R.Y., Ernst, W.G., Rumble, D. III. & Maruyama, S. (1998): High pressure minerals from deeply subducted metamorphic rocks. In Hemley, R. & Mao, D. (eds.): Ultrahigh-pressure mineralogy /Rev. Mineral., 37/. Washington (D.C.): Mineral. Soc. Am., 33–96. Liu, J., Bohlen, S.R. & Ernst, W.G. (1996): Stability of hydrous phases in subducting oceanic crust. Earth Planet. Sci. Lett., 143:161–171. Luth, R.W. (1995): Experimental determination of the reaction dolomite + 2 coesite = diopside + 2 CO2 to 6 GPa. Contrib. Mineral. Petrol., 122:152–158. Luth, R.W. (1997): Experimental study of the system phlogopite-diopside from 3.5 to 17 GPa. Am. Mineral., 82:1198–1210. Luth, R.W. (1999): Carbon and carbonates in the mantle. In Fei, Y., Bertka, C.M. & Mysen, B.O. (eds.): Mantle petrology: Field observations and high-pressure experimentation: A tribute to Francis R. (Joe) Boyd /Geochem. Soc. Spec. Publ., 6/. St. Louis (Mo.): Geochem. Soc., 297–316. Luth, R.W. (2001): Experimental determination of the reaction aragonite + magnesite = dolomite at 5 to 9 GPa. Contrib. Mineral. Petrol., 141:222–232. Massonne, H.J. & Szpurka, Z. (1997): Thermodynamic properties of white micas on the basis of high-pressure experiments in the systems K2O-MgO-Al2O3-SiO2-H2O and K2O-FeO-Al2O3-SiO2-H2O. Lithos, 41:229–250. Mengel, K. & Green, D.H. (1989): Stability of amphibole and phlogopite in metasomatized peridotite under water-saturated and water-undersaturated conditions. In Ross, J. (ed.): Kimberlites and related rocks. Proc. 4th Int. Kimberlite Conf. /Geol. Soc. Aus. Spec. Publ., 14/, 571–581. Mibe, K., Fujii, T. & Yasuda, A. (2002): Composition of aqueous fluid coexisting with mantle minerals at high pressure and its bearing on the differentiation of the Earth’s mantle. Geochim. Cosmochim. Acta., 66:2273–2285.

Mineral assemblages in ultrahigh pressure metamorphism

339

Mirwald, P.W. & Massonne, H.J. (1980): The low-high quartz and quartz-coesite transition to 40 kbar between 600 degrees C and 1600 degrees C and some reconnaissance data on the effect of NaAlO2 component on the low quartz-coesite transition. J. Geophys. Res., 85:6983–6990. Miyajima, N., Yagi. T., Hirose, K., Kondo, T., Fujino, K. & Miura, H. (2001): Potential host phase of aluminum and potassium in the Earth’s lower mantle, Am. Mineral., 86:740–746. Molina, J.F. & Poli, S. (2000): Carbonate stability and fluid composition in subducted oceanic crust: an experimental study on H2O-CO2-bearing basalts. Earth Planet. Sci. Lett., 176:295–310. Mysen, B.O. & Boettcher, A.L. (1975): Melting of a hydrous mantle: I. Phase relations of natural peridotite at high pressures and temperatures with controlled actvivities of water, carbon dioxide, and hydrogen. J. Petrol., 16:520–548. Niida, K. & Green, D.H. (1999): Stability and chemical composition of pargasitic amphibole in MORB pyrolite under upper mantle conditions. Contrib. Mineral. Petrol., 135:18–40. Okamoto, K. & Maruyama, S. (1998): Multi-anvil re-equilibration experiments of a Dabie Shan ultrahigh pressure eclogite within the diamond-stability fields. Isl. Arc, 7:52–69. Okamoto, K. & Maruyama, S. (1999): The high-pressure synthesis of lawsonite in the MORB + H2O system. Am. Mineral., 84:362–73. O’Neill, H.St.C. (1981): The transition between spinel lherzolite and garnet lherzolite and its use as a geobarometer. Contrib. Mineral. Petrol., 77:185–194. O’Neill, B. & Jeanloz, R. (1994): MgSiO3-FeSiO3-Al2O3 in the Earth’s lower mantle: perovskite and garnet at 1200 km depth. J. Geophys. Res., 99:19,901–19,915. Ono, S. (1998): Stability limits of hydrous minerals in sediment and mid-ocean ridge basalt compositions: Implications for water transport in subduction zones. J. Geophys. Res., 103:18,253–18,267. Pal’yanov, Y.N., Sokol, A.G., Borzdov, Y.M., Khokhryakov, A.F. & Sobolev, N.V. (1999): Diamond formation from mantle carbonate fluids. Nature, 400:417–418. Pal’yanov, Y.N., Sokol, A.G., Borzdov, Y.M., Khokhryakov, A.F. & Sobolev, N.V. (2002): Diamond formation through carbonate-silicate interaction. Am. Mineral., 87:1009–1013. Pawley, A.R. & Holloway, J.R. (1993): Water sources for subduction zone volcanism – new experimental constraints. Science, 260:664–667. Poli, S. (1993): The amphibolite-eclogite transformation – an experimental-study on basalt. Am. J. Sci., 293:1061–1107. Poli, S. & Schmidt, M.W. (1995): H2O transport and release in subduction zones – experimental constraints on basaltic and andesitic systems. J. Geophys. Res., 100:22,299–22,314. Poli, S. & Schmidt, M.W. (1998): The high-pressure stability of zoisite and phase relationships of zoisitebearing assemblages. Contrib. Mineral. Petrol., 130:162–175. Poli, S. & Schmidt, M.W. (2002): Petrology of subducted slabs. Annu. Rev. Earth Planet. Sci., 30:207–235. Robertson, J.K. & Wyllie, P.J. (1971): Rock-water systems, with special reference to the water-deficient region. Am. J. Sci., 271:252–277. Robinson, J.A. & Wood, B.J. (1998): The depth of the spinel to garnet transition at the peridotite solidus. Earth Planet. Sci. Lett., 164:277–284. Ryabchykov, I., Schreyer, W. & Abraham, K. (1983): Fluid compositions in Mg2SiO4-SiO2-H2O and jd-di-H2O sections at upper mantle conditions. Int. Geol. Rev., 25:777–784. Sato, K. & Katsura, T. (2001): Sulfur: a new solvent-catalyst for diamond synthesis under high-pressure and high-temperature conditions. J. Cryst. Growth., 223:189–194. Schmidt, M.W. (1995): Lawsonite: Upper pressure stability and formation of higher density hydrous phases. Am. Mineral., 80:1286–1292. Schmidt, M.W. (1996): Experimental constraints on recycling of potassium from subducted oceanic crust. Science, 272:1927–1930. Schmidt, M.W. & Poli, S. (1998): Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation. Earth Planet. Sci. Lett., 163:361–379. Schmidt, M.W., Finger, L.W., Angel, R.J. & Dinnebier, R.E. (1998): Synthesis, crystal structure, and phase relations of AlSiO3OH, a high-pressure hydrous phase. Am. Mineral., 83:881–888. Schmidt, M.W., Dugnani, M. & Artioli, G. (2001): Synthesis and characterization of white micas in the join muscovite-aluminoceladonite. Am. Mineral., 86:555–565. Schneider, M.E. & Eggler, D.H. (1986): Fluids in equilibrium with peridotite minerals; implications for mantle metasomatism. Geochim. Cosmochim. Acta., 50:711–724.

340

S. Poli & P. Fumagalli

Schreyer, W., Medenbach, O., Abraham, K., Gebert, W. & Müller W.F. (1982): Kulkeite a new metamorphic phyllosilicate mineral: ordered 1:1 chlorite/talc mixed-layer. Contrib. Mineral. Petrol., 80:103–109. Sclar, C.B., Carrison, L.C. & Schwartz, C.M. (1965): High-pressure synthesis and stability of a new hydronium-bearing layer silicate in the system MgO-SiO2-H2O. Trans. Am. Geophys. Union, 46:184. Sokol, A.G., Pal’yanov, Y.N., Pal’yanova, G.A., Khokhryakov, A.F. & Borzdov, Y.M. (2001): Diamond and graphite crystallization from C-O-H fluids under high pressure and high temperature conditions. Diamond Relat. Mater., 10:2131–2136. Stachel, T., Harris, J.W., Brey, G.P. & Joswig, W. (2000): Kankan diamonds (Guinea) II: lower mantle inclusion parageneses. Contrib. Mineral. Petrol., 140:16–27. Stalder, R. & Ulmer, P. (2001): Phase relations of a serpentine composition between 5 and 14 GPa: significance of clinohumite and phase E as water carriers into the transition zone. Contrib. Mineral. Petrol., 140:670–679. Stalder, R., Ulmer, P., Thompson, A.B. & Gunther, D. (2000): Experimental approach to constrain second critical end points in fluid/silicate systems: Near-solidus fluids and melts in the system albite-H2O. Am. Mineral., 85:68–77. Stalder, R., Ulmer, P., Thompson, A.B. & Gunther, D. (2001): High pressure fluids in the system MgO-SiO2H2O under upper mantle conditions. Contrib. Mineral. Petrol., 140:607–618. Sudo, A. & Tatsumi, Y. (1990): Phlogopite and K-amphibole in the upper mantle: implication for magma genesis in subduction zones. Geophys. Res. Lett., 17:29–32. Sundqvist, B. (1999): Fullerenes under high pressures. Adv. Phys., 48(1):1–134. Sung, C.M. & Tai, M.F. (1997): Reactivities of transition metals with carbon: Implications to the mechanism of diamond synthesis under high pressure. Int. J. Refract. Met. Hard Mater., 15:237–256. Trønnes, R.G. (2002): Stability range and decomposition of potassic richterite and phlogopite end members at 5–15 GPa. Mineral. Petrol., 74:129–148. Ulmer, P. & Trommsdorff, V. (1999): Phase relations of hydrous mantle subducting to 300 km. In Fei, Y., Bertka, C.M. & Mysen, B.O. (eds.): Mantle petrology: Field observations and high-pressure experimentation: A tribute to Francis R. (Joe) Boyd /Geochem. Soc. Spec. Publ., 6/. St. Louis (Mo.): Geochem. Soc., 259–281. Usui, T., Nakamura, E., Kobayashi, K., Maruyama S. & Helmstaedt, H. (2003): Fate of the subducted Farallon Plate inferred from eclogite xenoliths in the Colorado Plateau. Geology, 31:589–592. van Roermund, H.L.M., Carswell, D.A., Drury, M.R. & Heijboer, T.C. (2002): Microdiamonds in a megacrystic garnet websterite pod from Bardane on the island of Fjørtoft, western Norway: Evidence for diamond formation in mantle rocks during deep continental subduction. Geology, 30:959–962. Wallace, M.E. & Green, D.H. (1991): The effect of bulk rock composition on the stability of amphibole in the upper mantle: implications for solidus positions and mantle metasomatism. Mineral. Petrol., 44:1–19. Walter, M., Katsura, T., Kubo, A., Shinmei, T., Nishikawa, O., Ito, E., Lesher, C. & Funakoshi, K. (2002): Spinel-garnet lherzolite transition in the system CaO-MgO-Al2O3-SiO2 revisited: an in situ X-ray study. Geochim. Cosmochim. Acta, 66:2109–2121. Webb, S.A. & Wood, B.J. (1986): Spinel-pyroxene-garnet relationships and their dependence on Cr/Al ratio. Contrib. Mineral. Petrol., 92:471–480. Woodland, A.B. & O’Neill, H.St.C. (1995): Phase relations between Ca3Fe23+Si3O12-Fe32+Fe23+Si3O12 garnet and CaFeSi2O6-Fe2Si2O6 pyroxene solid solutions. Contrib. Mineral. Petrol., 121:87–98. Wunder, B. (1998): Equilibrium experiments in the system MgO-SiO2-H2O (MSH): stability fields of clinohumite-OH [Mg9Si4O16(OH)2], chondrodite-OH [Mg5Si2O8(OH)2] and phase A (Mg7Si2O8(OH)6). Contrib. Mineral. Petrol., 132:111–120. Yagi, T., Suzuki, T. & Akaogi, M. (1994): High pressure transitions in the system KAlSi3O8-NaAlSi3O8. Phys. Chem. Miner., 21:12–17. Yamaoka, S., Kumar, M.D.S., Kanda, H. & Akaishi, M. (2002): Crystallization of diamond from CO2 fluid at high pressure and high temperature. J. Cryst. Growth, 234:5–8. Yaxley, G.M. & Green, D.H. (1994): Experimental demonstration of refractory carbonate-bearing eclogite and siliceous melt in the subduction regime. Earth Planet. Sci. Lett., 128:313–325. Yoder, H.S. (1952): The MgO-Al2O3-H2O system and the related metamorphic facies. Am. J. Sci., 250A:569–627. Copyright permission Figure 7 (from Geochimica et Cosmochimica Acta, Vol. 33, Boettcher, A.L. & Wyllie, P.J.: The system CaO-SiO2CO2-H2O – III. Second critical end-point on the melting curve, pp. 611-632, copyright 1969) and Figure 9 (from Chemical Geology, Vol. 192, Hermann, J.: Allanite: thorium and light rare earth element carrier in subducted crust, pp. 289-306, copyright 2002) in this chapter are reprinted with permission from Elsevier.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 11, 341–363

Dating UHP metamorphism DANIELA RUBATTO1*, ANTHI LIATI2 and DIETER GEBAUER2 1

Department of Geology, Australian National University, Canberra ACT 0200, Australia; 2 Department of Earth Sciences, ETH-Zentrum, CH-8092 Zürich, Switzerland; * e-mail: [email protected]

Introduction The aims and the challenge Ultrahigh pressure (UHP) rocks are extraordinary rocks that recorded extreme metamorphic conditions. Their study has been an exciting part of geosciences and in the last few years has emerged as a distinct discipline. Only few UHP terranes have been recognised worldwide so far and this only over the last ca. 20 years. They occur within major continental collision belts, predominantly in Eurasia and Africa (see e.g. Ernst & Liou, 2000; Liou, 2000; Carswell & Compagnoni, 2003 for a general review). Even though UHP rocks have been subject to intense studies in the last two decades, many questions regarding their formation and evolution remain open. Geochronology can assist this search for information in providing the absolute time of their formation and several other parameters regarding their evolution. Geochronology determines the absolute time of formation of UHP units and nearby units within the same chain recording different pressure–temperature (PT) conditions. This can in turn serve to identify the dynamics and extension of the tectonic processes responsible for the formation and preservation of UHP terranes. Age determination of several stages of the evolution of UHP rocks provides the time of exhumation and, rarely, that of subduction. Prograde ages are difficult to obtain mainly because of later re-equilibration at higher temperatures and age resetting of the minerals of interest. The ages, in combination with petrological data, are used to infer exhumation and subduction rates. These rates are crucial for the construction of geodynamic models and to discern between the different tectonic processes responsible for bringing rocks to great depth and exhume them to the surface. A detailed geochronological investigation of UHP assemblages, applying different mineral chronometers, can provide earth scientists with an estimate of the duration of peak metamorphism. This information would have important implications for the thermal and rheological regimes present along the subduction slab at great depth. For each of these tasks the correlation between time, pressure, temperature and, ideally, fluids and deformation is crucial. This is a big challenge for geochronology, a discipline that has long operated relatively separate from petrology and to some extent even geology. It is only in the last years that increasing efforts in crossing these boundaries have been made.

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

342

D. Rubatto, A. Liati & D. Gebauer

The problems The main problem in dating UHP and HP rocks and metamorphic rocks in general is, once a suitable mineral chronometer is found, to establish the relationships between ages and metamorphic conditions under which the minerals used for dating were formed. In UHP rocks, more than in other rock types, several factors (see below) can hamper the achievement of this goal, as is also reported in the summary of Gebauer (1990) for dating of HP rocks. The best way to correlate age and PT of metamorphism is to date major metamorphic minerals for which conditions of formation are known. In HP rocks in particular, this can be best achieved by dating garnet, which can be used for thermobarometry. Micas and amphiboles are other datable minerals, together with titanite. More problems are encountered when dating accessory minerals such as zircon and monazite, which are the pillars of U–Pb dating. The metamorphic behaviour of these minerals is much less known and, thus, additional information is required to define a correlation with P and T conditions (see below). Because of the extreme conditions that UHP rocks experienced, they often display a complex evolution, during prograde and retrograde stages, as widely reported in other chapters of this volume. This is reflected in the wide presence of zoned minerals (e.g. garnet) and/or multiple generations of minerals (e.g. micas) as well as in prograde and/or peak metamorphic inclusions. Modern geochronology has to take advantage of this mineralogical complexity by dating single mineral zones or generations. This task is best achieved with beam techniques that allow isotopic measurements, either in thin section or mineral separates, of such mineral zones (domains), a discipline that has seen enormous progress in the last decade (e.g. Claoué-Long et al., 1991; Gebauer, 1990; Kelley, 2002; Müller, 2003). A consequence of the complex mineralogy of UHP rocks is the widespread occurrence of only local equilibrium, with several parageneses (e.g. peak and retrograde) being preserved in the same rock sample (e.g. Hermann et al., 2001; Reinecke, 1991; Schertl et al., 1991). This becomes a problem when using isotopic systems that rely on large mineral separates, multi-domain single minerals or whole rock data (e.g. Rb–Sr, Sm–Nd and Lu–Hf isochrons or Ar–Ar step-wise degassing). Whole rock data points always present a risk and should thus be avoided. Mineral data can be improved using careful mineral separation and chemical characterisation of the separate to check for heterogeneity or inclusions. Modern leaching techniques have sometimes produced promising results in the elimination, or minimisation, of inclusions (e.g. Amato et al., 1999; Anczkiewicz et al., 2002; Scherer et al., 2000), a problem particularly important in dating garnet. Oxygen isotopes have been proposed as a tool to establish equilibrium between HP minerals (Zheng et al., 2002). A characteristic of UHP rocks are steep PT paths in which short T intervals correspond to great depth variations, with the extreme condition being isothermal decompression (e.g. Gebauer, 1996; Gebauer et al., 1997; Hermann et al., 2001; Sánchez-Rodríguez & Gebauer, 2000; Terry et al., 2000). This feature hampers the use of thermochronometers, which assume a known temperature as closure for the diffusion

Dating UHP metamorphism

343

of a certain isotope in a given mineral. Closure temperatures are not absolute values, but are a function of cooling rates, mineral grain size, fluid–rock interaction and mineral chemistry (e.g. Kelley, 2002; Scherer et al., 2000). Therefore, closure temperatures can vary from rock to rock, and they always bear an intrinsic error. Alternatively, if radiometric mineral ages are considered to be primarily caused by fluid-induced recrystallisation (e.g. Villa, 1998), temperature is only one of a series of parameters that influence the closure of an isotopic system. Thus, a precise definition of a cooling path, with respect to absolute ages, remains difficult and the application of fission track data, mainly on zircon and apatite, is more promising. Because during fast exhumation of UHP rocks small T intervals are covered in short periods of time, even small errors in temperature–time correlation can produce large inaccuracy in cooling profiles. Together with extreme pressures, UHP rocks often display relatively high peak temperatures (> 600 °C). Such temperatures make it impossible to constrain the age of peak metamorphism with conventional thermochronometers (e.g. Ar–Ar, Rb–Sr), with closure temperatures generally below ~ 600 °C (e.g. Kelley, 2002; Scherer et al., 2000). However, closure temperatures are – amongst others – a function of cooling rates (i.e. residence time at high T), and in UHP rocks, fast exhumation and short residence time at high T can push them towards higher values (e.g. Rubatto & Hermann, 2001). There is good indication that rocks which preserve HP and UHP minerals were exhumed fast and/or have not suffered very strong deformation. This is rather the exception, since in most cases (U)HP minerals are obliterated due to metamorphic overprint and the rock preserves no memory of the (U)HP event. Fast geological processes require dating with high age resolution in order to date the different metamorphic stages that the rock experienced. This is a challenge for some geochronological techniques more than others, for example those relying on small sample sizes such as beam techniques (laser ablation ICP-MS, ion microprobe and electron microprobe U–Pb dating, and laser spot Ar–Ar dating). The method Accurate and precise dating of distinct stages during the evolution of UHP rocks requires a combination of techniques that • can take advantage of polyphase growth of minerals, • have good to high time resolution, and that are based on minerals that • can be related to metamorphic conditions, • are robust to retrogression and to HT perturbations, and • are sensitive to pressure changes. All these features are hardly met in a single chronometer. There are more chances to achieve the above requirements if mineral chronometers are combined. Moreover, techniques that help to understand the conditions of formation, the petrology, the composition, and the metamorphic behaviour of the dated minerals need to be applied. Even though the ideal chronometer that fulfils all theoretical requirements may not yet be available, a number of studies have successfully managed to describe and constrain

344

D. Rubatto, A. Liati & D. Gebauer

the P–T–t evolution of some UHP units throughout the world. These studies accomplished their goal combining different techniques, which may vary from case to case. They all present some degree of innovation that can contribute to accurate dating of UHP metamorphism. In the following, we present two case studies (the Dora-Maira and the Kokchetav massifs) to illustrate how modern geochronology can deal with the problem of dating UHP metamorphism. We go through the major burdens encountered in each case and refer particularly to the studies that introduced innovations and marked a significant advance. A brief review on studies that have dealt with other important UHP terranes is following, with the intention to stress what geochronology is still lacking to reach the goal of dating UHP metamorphism. We then discuss the tectonic and geological implications of the P–T–t paths of UHP units, which information can they provide to geologists and how this gives an insight on the tectonic processes involving UHP rocks. A final chapter is addressing the challenges that lie ahead in the geochronology of UHP rocks.

Case studies In this section we present a review of the geochronological work done on two extensively studied UHP units: the Dora-Maira Massif in the Western Alps and the Kokchetav Massif in Kazakhstan. In these localities there is a relatively complete knowledge of P–T–t evolution. Focussing particularly on the more recent works, we discuss the major improvements in dating UHP rocks, both in terms of techniques and reliability. The Dora-Maira The Dora-Maira Massif in the Western Alps contains a UHP unit (see the review of Compagnoni & Rolfo, 2003 in this volume for a complete presentation of the geological setting) that forms a coherent body (ca. 10 × 5 × 1 km) of continental crust. It consists of a lithologically heterogeneous basement (metapelites, para- and orthogneisses with intercalated eclogites, calc-silicate rocks, and marbles) that was intruded by PermoCarboniferous granites (Chopin et al., 1991; Compagnoni et al., 1995; Gebauer et al., 1997). There is evidence through the whole unit of a widespread recrystallisation in the coesite stability field during the Alpine orogeny at pressures of ~ 3.5 GPa and temperatures of ~ 750 °C (Chopin et al., 1991; Compagnoni et al., 1995; Schertl et al., 1991). In a recent work, Hermann (2003) suggests that the unit reached conditions of even 4.1 GPa, within the diamond stability field. The first geochronological investigations of the UHP rocks were carried out in the late 80’s using classical techniques, such as Ar–Ar and Rb–Sr on micas, and U–Pb isotope dilution, multigrain analysis on zircon. The first results of these studies reported ages scattering between 120 and 95 Ma (Monié & Chopin, 1991; Paquette et al., 1989). These figures fitted the view supporting a Cretaceous Alpine subduction, which was generally accepted at that time and also claimed for other Alpine areas (e.g. Hunziker et al., 1992). However, at the same time Ar–Ar data in the Dora-Maira and the nearby HP unit of Monviso (Monié & Philippot, 1989; Monié et al., 1989) already contradicted this picture reporting younger Tertiary ages around 50 Ma. Even though these were

Dating UHP metamorphism

345

interpreted as cooling ages, the authors postulated two episodes of HP metamorphism in the area. Further evidence for the possibility of Tertiary subduction came from the work of Tilton et al. (1989, 1991). These authors obtained U–Pb data on zircon, ellenbergerite and monazite interpreted to reflect metamorphic ages in the range of 30–38 Ma, not necessarily corresponding to UHP metamorphism. The same authors reported the first Sm–Nd errorchron data from a pyrope garnet of the UHP unit, interpreted to correspond to an “age” of ca. 38 Ma. Despite the clear scattering of the Tertiary ages, their dubious reliability and different significance, these works opened the possibility of Tertiary UHP metamorphism in the Dora-Maira and the Alps in general. From this set of data also arose the problem of which technique to use for best dating such complicated rocks involved in unusual metamorphic conditions. The reliability of conventional Ar–Ar stepwise heating dating technique in HP and UHP minerals was questioned by Arnaud & Kelley (1995), who documented anomalously old ages in HP minerals from several localities including the Dora-Maira. The introduction of in situ laser spot extraction techniques made possible Ar–Ar studies of within crystal age variations and distribution of argon between mineral phases. This led to the discovery of concentration of excess argon at the grain boundaries and the close correlation of excess argon with fluid mediated alteration in HP and UHP terrains (e.g. Giorgis et al., 2000; Kelley, 2002; Scaillet, 1998). The problems described remain major obstacles in dating the cooling path of HP rocks with the Ar–Ar technique. It was the work of Gebauer et al. (1997) to sign a breakthrough in the determination of the P–T–t path of the Dora-Maira UHP unit. The authors applied the technique of ion microprobe (SHRIMP) dating of single zircon domains supported by cathodoluminescence (CL) imaging of the zircons dated, a technique that was proven to be successful in several localities with polymetamorphic (U)HP rocks (e.g. Gebauer, 1996). This represented an important innovation because it allowed dating of narrow zircon domains (20–30 µm wide) that recorded UHP metamorphism separately from crystal domains that preserved pre-metamorphic ages (Fig. 1). This was most important

Fig. 1. Cathodoluminescence image of zircon crystals from Dora-Maira UHP rocks. (a) Crystal include in a 15 cm large, prograde megablast, in which the distinction between oscillatory-zoned, magmatic zircon core and metamorphic rim developing oscillatory zoning against the outer surface (increasing amounts of fluids/melts) is particularly evident. Ion microprobe analysis (analysed spots represented by the circles) is the only technique that allows dating the different domains separately. (b) Crystal, separated from a fine-grained pyrope quartzite, containing an “inclusion” of coesite partly surrounded by metamorphic zircon. For more details see text.

346

D. Rubatto, A. Liati & D. Gebauer

for zircon, a mineral that does not generally recrystallise completely during HP or UHP metamorphism, unless fluids are present in abundance. Zircon has the capacity to preserve its original U–Pb composition despite high-T overprint. With the assistance of CL imaging, Gebauer et al. (1997) could identify metamorphic domains in zircon crystals included in prograde garnets and associated with UHP minerals. These domains, generally forming crystal rims or aureoles around UHP inclusions (Fig. 1), were characterised by relatively homogeneous CL patterns contrasting with the oscillatory zoning of the older magmatic cores, and Th/U ratios lower than their magmatic precursor. Newly formed Alpine zircon developing oscillatory zoning during precipitation from a supercritical fluid/melt could also be dated (Fig. 1 in this study or Fig. 3 in Gebauer et al., 1997). Thus, based on the above information, and despite the existence of many discordant oscillatory zoned domains, the age of the metamorphic zircon domains could be well constrained at 35.4 ± 1.0 Ma (Gebauer et al., 1997). In this study the link to metamorphic conditions was possible because metamorphic zircons and/or zircon domains were extracted from large (ca. 15 cm in diameter) pyrope megablasts (Gebauer et al., 1997). These megablasts grew at the expense of chlorite + talc + kyanite at about 700 °C and 30 kbar, i.e. at UHP conditions shortly before the UHP peak (e.g. Schertl et al., 1991). The enclosed ca. 35 Ma old newly formed zircons and zircon domains (Fig. 1 this study and Fig. 3 in Gebauer et al., 1997) were thus interpreted to predate the UHP peak as they must have crystallised during or very slightly before the growth of the garnet megablasts, probably from a supercritical fluid/melt. Additionally, because of the analytical error of the SHRIMP technique, ages that may correspond to different stages of metamorphism appear as identical (e.g. Kröner et al., 2000). In this respect, minimisation of the analytical error would be a step forward in distinguishing between ages acquired during different metamorphic stages and therefore better constrain subduction and exhumation rates. In the specific case of the Dora-Maira, the zircon-bearing rocks mainly preserved a UHP assemblage and, importantly, some of the ca. 35 Ma old metamorphic zircon domains occur in contact with and around pseudoinclusions of UHP minerals such as phengite and coesite (Fig. 1). The UHP “inclusions” found within oscillatory magmatic (Permian) zircon must be secondary: they either filled fractures or replaced magmatic domains during UHP metamorphism. The process that produced these pseudo-inclusions also favoured recrystallisation of zircon during the UHP stage. These UHP “inclusions” were then armoured by the surrounding zircon and thus could survive retrogression during exhumation. Based on these observations, the age of 35.4 ± 1.0 Ma was attributed to the UHP peak and, combined with a zircon fission track age of 29.9 ± 1.4 Ma (Gebauer et al., 1997), a two-point P–T–t path was constructed. This combination of dating techniques, together with the petrological information, allowed the calculation of an average exhumation rate for the UHP DoraMaira unit on the order of 20–24 km/m.y. (2–2.4 cm/year) with a cooling rate of about 85–100 °C/m.y. (Gebauer et al., 1997). The Dora-Maira was subject of another milestone study of (U)HP rocks representing the first example of Lu–Hf dating (Duchêne et al., 1997). The Lu–Hf system is based on the same principles as the more commonly used Sm–Nd dating. The main advantage of this new technique is the high Lu/Hf ratio of garnet, the main mineral

Dating UHP metamorphism

347

used, which results in much better precision of an isochron age. Thanks to the introduction of plasma source mass spectrometry, the precision in measuring low Hf concentrations improved significantly, thus making Lu–Hf dating suitable for low Hf minerals such as garnet. For a Dora-Maira coesite-pyrope quartzite, Duchêne et al. (1997) reported an age of 32.8 ± 1.2 Ma, based on an isochron defined by two garnet and one whole rock analyses. The same authors also obtained a ca. 31.8 Ma Sm–Nd age from garnets of a similar sample. However, it is worth mentioning that Nd isotopic disequilibrium within the Dora-Maira garnets was later reported, and related to the presence of REE rich, Hf poor micro-inclusions (Luais et al., 2001). On the basis of a suggested closure temperature in excess of 600 °C for Lu–Hf in garnet, Duchêne et al. (1997) interpreted their ages as dating the early stage of exhumation after peak metamorphism at around 35 Ma. Unfortunately no pressure estimate was obtained from the dated garnet, an information that would have strengthened the importance of this age. The authors proposed an exhumation rate “of the order of 3 cm/year”, in line with the one suggested by Gebauer et al. (1997). The P–T–t path of the Dora-Maira UHP rocks, first established by Gebauer et al. (1997), was constrained in further detail by Rubatto & Hermann (2001), who applied in situ ion microprobe U–Pb dating to titanite from calc-silicates. The importance of this study lies in the fact that the above authors combined geochronological data with petrological information and dated a metamorphic mineral (titanite), for which conditions of formation are much easier to establish when compared to zircon. By using textural relationships, the composition of titanite itself, as well as that of numerous mineral inclusions in it, the existence of three generations of titanite could be established. These three generations of titanite formed at different PT conditions, based on their different major and trace element composition (Fig. 2; Rubatto & Hermann, 2001 and unpublished data). With the additional support of some thermodynamic equilibrium calculations involving titanite, it was inferred that the oldest titanite formed at the metamorphic peak at 35.1 ± 0.9 Ma, confirming the data of Gebauer et al. (1997). In addition, it was demonstrated that another two titanite generations formed during two distinct decompressional stages: at 32.9 ± 0.9 Ma and 31.8 ± 0.5 Ma. The combination of geochronology with metamorphic petrology and the addition of the fission track data from Gebauer et al. (1997) allowed the definition of a four-point P–T–t path. This represents one of the best-documented exhumation trajectories of a UHP unit (Fig. 3). More than confirming fast exhumation, Rubatto & Hermann (2001) could identify changes in exhumation rates from 3.4 to 0.5 cm/year during the evolution of the Dora-Maira. This new piece of information bears important implications for tectonic models (see below). Apart from tight constraints on the last part of the prograde, as well as on the retrograde P–T–t trajectory of these unique rocks, also a series of further indications on their pre-metamorphic history, as well as the availability of fluids during metamorphism could be gained (Gebauer et al., 1997). 1) The first order Permian protolith of granitic composition was probably metasomatised at low P–T to form Mg-rich leucophyllites during Permo-Triassic rifting. 2) Zircon formed metamorphic domains at 35 Ma only in the fluid-rich, second order protoliths leucophyllites of the UHP white-schists and not in the much drier country rocks. 3) Minor partial melting probably occurred during UHP

348

D. Rubatto, A. Liati & D. Gebauer

Fig. 2. Dora-Maira titanites. (a, b) Backscattered electron (BSE) images of dated titanite crystals from a UHP and a LP calc-silicate, respectively. Four generations of titanite are distinguishable on the base of their BSE emission, which reflects different chemical composition. (c, d) Variation in major and trace element compositions, respectively, between the four generations of titanites. Major element data from Rubatto & Hermann (2001) and unpublished trace element data.

conditions in the white-schists as suggested by rare, newly formed, metamorphic zircon domains that grade into oscillatory-zoned domains (Fig. 1). 4) Jadeite-rich layers within the white-schists probably represent former Permian aplite veins within the first order granitic protolith and are not a product of UHP melting. In summary, the geochronological investigation of the Dora-Maira UHP unit comprises a large variety of isotopic systems and methods, including ion microprobe U–Pb analysis, Sm–Nd and Lu–Hf mineral isochrons, Ar–Ar and fission track data. This is the most complete data set acquired on a UHP unit so far. The combination of different dating techniques, some more reliable than others, led to a good knowledge of the age of the UHP metamorphic peak and of the fast exhumation history (Fig. 3). The Kokchetav Massif The Kokchetav Massif in Kazakhstan forms part of the Caledonian Central Asiatic fold belt and is situated between the Siberian platform and the East European platform. This massif has become worldwide known because of the finding of metamorphic microdiamonds, providing evidence for UHP (> 40 kbar) metamorphism (see Shatsky & Sobolev, 2003 in this volume for a review). The diamondiferous rocks are present in the Zerenda Series, which consists mainly of garnet-biotite-kyanite gneisses and schists with intercalated dolomitic marbles, calc-silicates and eclogites.

Dating UHP metamorphism

349

Fig. 3. P–T–t path for the Dora-Maira UHP unit, modified after Rubatto & Hermann (2001). Geochronological data from: 1) Rubatto & Hermann (2001); 2) Gebauer et al. (1997); 3) Duchêne et al. (1997).

The first geochronological investigation of the Kokchetav Massif UHP rocks was carried out soon after the finding of diamonds in the area. ClaouéLong et al. (1991) dated zircons from a biotite schist in a work that belongs to the first generation of SHRIMP ion microprobe study of metamorphic rocks and was the first work on HP/UHP rocks. On the basis of optical microscopy only of the zircon crystals (cathodoluminescence was not yet a routine technique at the time), Claoué-Long et al. (1991) identified rounded, inherited zircon cores of likely detrital origin and zircon rims. The apparent U–Pb ages of the cores scattered between 558 and 1981 Ma, confirming their detrital origin. The rims were attributed to the UHP stage because a diamond aggregate was found at the core–rim boundary. Sixteen analyses on rims yielded a mean age of 530 ± 7 Ma, which was interpreted as the age of the UHP metamorphic peak. Several years later, the Kokchetav rocks were again investigated with the same technique (SHRIMP dating of zircon domains) by Katayama et al. (2001) and by Hermann et al. (2001). This latter study extended dating to other rock types including several gneisses and a dolomitic metacarbonate. The main innovation introduced by these studies was the use of zircon imaging to assist SHRIMP dating (CL and backscattered electron images) and the detailed investigation of mineral inclusions in zircon (Fig. 4). The inclusions, which were analysed by Raman spectroscopy (mainly Katayama et al., 2001) and by electron microprobe (Hermann et al., 2001), proved to be not exclusively of UHP origin, but also prograde and retrograde (diamond, graphite, quartz, coesite, kyanite, rutile, biotite, phengite, K-feldspar, plagioclase, chlorite, clinopyroxene). Thus, the zircon crystals in the Kokchetav rocks must have formed (or recrystallised) during different stages of the evolution of the UHP unit. With a detailed petrological study, Hermann et al. (2001) could correlate the composition of the inclusions to the variable rock mineral assemblages from which P–T estimates were obtained. The inclusion study was coupled with trace element analysis of zircon domains by laser ablation ICP-MS (Hermann et al., 2001). The trace element composition of zircon was

350

D. Rubatto, A. Liati & D. Gebauer

Fig. 4. Photomicrograph (transmitted light) (a) and CL image (b) of different zircon crystals from a Kokchetav UHP gneiss. Inclusions belonging to different P–T stages are present in crystals from the same rock and even within the same crystal, indicating that zircon formed over a range of P–T conditions. Surprisingly, the CL zoning of the domains with different inclusions is quite similar.

revealed to be another powerful tool allowing distinction between zircon domains formed at different PT conditions. In particular, zircon formed at HP/UHP had a peculiar signature characterised by a flattening of the HREE pattern and a less pronounced negative Eu anomaly, when compared to zircon formed at lower pressure (Fig. 5). The different zircon chemistry was attributed to changes in the rock paragenesis during different stages of metamorphism. The depletion in HREE is due to the presence of garnet as a HREE sink during HP metamorphism and part of the early decompression history. On the other hand, zircon domains enriched in HREE with respect to the MREE and with a marked negative Eu anomaly are related to the breakdown of garnet and the formation of feldspar (a known sink for Eu) during amphibolite facies overprint. These changes in the trace element chemistry of zircon and their correlation with different stages of metamorphism have been since investigated more in detail by Rubatto (2002) and Rubatto & Hermann (2003), who confirmed the typical signature of eclogite-facies zircon. In the Kokchetav rocks, Hermann et al. (2001) identified four different types of zircon domains. According to the composition of their mineral inclusions and trace element chemistry, three of these zircon domains were attributed to metamorphism, although most of them showed apparent similarities in CL patterns. The accurate determination of P–T conditions for zircon (re)crystallisation indicated that zircon formed during the peak of metamorphism and mainly in the early stages of the decompressional path. This observation implies caution in assuming that zircon generally forms at the metamorphic peak.

Dating UHP metamorphism

351

Fig. 5. Trace element patterns of zircon crystals from a Kokchetav UHP biotite gneiss. Note the difference in composition between zircon domains formed in different conditions. Data from Hermann et al. (2001).

Ion microprobe U–Pb dating of the different zircon domains identified by Hermann et al. (2001) on the basis of mineral inclusions and trace element chemistry, produced a narrow range of ages, confirming the data of Claoué-Long et al. (1991). Through statistical analysis it was concluded that the ages of the zircon domains formed from the UHP peak to amphibolite facies overprint were indistinguishable within analytical uncertainty (527 ± 5, 528 ± 8 and 526 ± 5 Ma; error at 95% c. l.). This provided evidence for a fast exhumation of the Kokchetav UHP rocks. On the other hand, Katayama et al. (2001) produced SHRIMP U–Pb data with large errors and spreading over a wide range (442 ± 40 to 563 ± 43 Ma). These authors argued for UHP peak at 537 ± 9 Ma, low-pressure overprint at 507 ± 8 Ma and a post orogenic thermal event at 456–461 Ma. The discrepancy between the two studies could be partly due to different sample localities (Kumdy Kol for Katayama et al., 2001, and Barchi Kol for Hermann et al., 2001) and the effect of a late thermal perturbation in the Katayama et al. (2001) samples. Additional disagreement could be related to analytical differences in the two laboratories or different statistical treatments of the data. Despite the different age results and interpretation, these studies once again pointed to a fast exhumation of UHP rocks (> 1.8 cm/year from a pressure peak of 43 kbar in Hermann et al., 2001; 0.5 cm/year from an uncertain peak of over 60 kbar in Katayama et al., 2001). The geochronological study of the Kokchetav rocks indicates how even an apparently simple zircon population, yielding an apparent single age, can bear important information about the time evolution of the host rocks. In this case, dating of crystal domains showing different CL patterns was not enough to unravel the complex history of zircon growth. The determination of metamorphic conditions of zircon formation through mineral inclusions and trace element composition could unravel the fast exhumation history of the unit. Other localities Besides the excellent examples of the Dora-Maira and the Kokchetav Massif, a number of other UHP terranes have been the subjects of geochronological investigation. In this session we will review some of these studies. The aim is to point out on one hand the

352

D. Rubatto, A. Liati & D. Gebauer

interesting approaches used, and on the other hand the problems that remain open in dating UHP rocks in general. The Dabie Shan metamorphic belt, the largest block of UHP rocks documented so far (e.g. Hirajima & Nakamura, 2003), has been dated using a variety of techniques including U–Pb dating of zircon and monazite, Sm–Nd mineral and whole rock isochrons, and Ar–Ar mineral dating (for a review see Hacker et al., 2000). In this area, early isotope dilution multigrain zircon analysis yielded an age of ~ 220 Ma, which has been later suggested to date a fluid influx during retrogression (Ayers et al., 2002; Hacker et al., 1998). The age of the metamorphic peak was determined at 230–240 Ma only when ion microprobe zircon dating was used (Ayers et al., 2002; Hacker et al., 1998; Hacker et al., 2000). Studies that applied the Sm–Nd techniques to eclogites reported both ages, but failed to correctly interpret their significance (for a review see Hacker et al., 2000). Ar–Ar dating of hornblende, micas and K-feldspar was partly compromised by excess argon and, similarly to the Sulu terrain (Giorgis et al., 2000), provided an extremely complex picture (for a review see Hacker et al., 2000). These cooling ages apparently suggest a relatively slow cooling and exhumation (only 2 mm/year) when compared to the Dora-Maira or Kokchetav massifs. Despite the large number of isotopic data, the geochronological investigation of the Dabie Shan UHP belt suffers the lack of a systematic attempt to relate ages with metamorphic conditions. The recognition and identification of a few UHP inclusions in zircon is the main ground on which the 230–240 Ma age was attributed to UHP metamorphism (Ayers et al., 2002). Neither zircon nor monazite were ever screened in detail for inclusions, or analysed for trace elements. Similarly, Sm–Nd dating was not coupled with the determination of P–T conditions of the analysed minerals. The large spectrum of Ar–Ar data would have greatly benefited from a more detailed chemical characterisation of the minerals to understand the conditions of formation. Good results in terms of P–T–t correlation have been reached for the UHP rocks of the Western Gneiss Region, Norway (e.g. Carswell & Cuthbert, 2003). Terry et al. (2000) carried out in situ dating of monazite by ion microprobe and electron microprobe. The above authors combined dating with sample textural analysis, petrography, and chemical mapping of the monazite crystals. By dating monazite in different textural relationships (crystals included in garnet and crystals in the matrix) and comparing their Th content, three monazite generations with different ages were recognised. Despite a certain degree of assumption in attributing age to metamorphic conditions, Terry et al. (2000) concluded that the peak occurred at 407 ± 2 Ma, the end of isothermal decompression at 395 ± 2 Ma, and a later shearing event at 375 ± 3 Ma. Similar to the Dora-Maira, precise ages obtained with the same technique and linked to metamorphism and deformation allowed description of the exhumation path of this unit. A new UHP terrane has been recently identified in the Rhodope zone of N Greece (Liati & Gebauer, 2001a; Mposkos & Kostopoulos, 2001). In the eastern part of this UHP terrane, U–Pb ion microprobe dating (SHRIMP), assisted by CL imaging, was carried out on zircon domains from garnet-rich rocks (Liati et al., 2002a). The zircon crystals contain Early Cretaceous magmatic cores (117.4 ± 1.9 Ma) and Late Cretaceous (73.5 ± 3.4 Ma) metamorphic domains, that were interpreted as dating the (U)HP

Dating UHP metamorphism

353

metamorphic peak. A retrograde metamorphic stage that corresponded to the intrusion of cross-cutting pegmatites was dated at 62 ± 2 Ma. According to the different tectonic scenarios and the peak pressure considered, Liati et al. (2002a) proposed average minimum exhumation rates of 2.2–2.8 mm/year or 5.2 mm/year. These figures should rather be considered as an approximation, because of the strong T-overprint of the rocks studied, which makes the calculation of precise peak UHP conditions difficult.

Tectonic implications Dating UHP rocks is not only aimed to establish the absolute age of the rocks. It should also tackle other issues such as the correlation of ages with distinct metamorphic stages along a P–T–t path, the age of nearby units, exhumation-, subduction-, heating- and cooling rates, and possibly the duration of metamorphic peak. Some of these important parameters have been determined in different localities. Their implications for tectonic and geodynamic models are reviewed in the following sections. Exhumation rates The most complete information in terms of exhumation rates resulted from recent studies that made an effort to relate ages with metamorphic conditions (Table 1). The best constrained data so far remain the ones obtained by Rubatto & Hermann (2001) for the Dora-Maira rocks. The exhumation of this unit is described by three P–T–t points, all obtained with the same dating technique (in situ ion microprobe U–Pb analysis). In this unit, average exhumation from 120 to 35 km depth has been estimated to be as fast as 3.4 cm/year. Relevant is also the fact that exhumation slowed down (1.6 cm/year) when the UHP rocks reached the base of the crust and were assimilated with other HP nappes. A further decrease of exhumation rates is recorded when the rocks reached mid crustal levels (0.5 cm/year) and exhumation probably changed from tectonic to erosion dominated. In the Kokchetav Massif, although less precisely defined, initial exhumation from depths of more than 140 km appears to be faster than 1.8 cm/year (Rubatto & Hermann, 2001). Variations in exhumation rates have been identified also in the Western Gneiss Region (Terry et al., 2000): initial fast exhumation from depths of 125 km slowed down when the unit reached 35–40 km depth, probably corresponding to the base of the crust. The same tendency is seen at shallow crustal levels where exhumation is slower. Even though in the Western Gneiss Region the absolute values are different and generally lower (1.1 cm/year, ~ 0.4 cm/year and < 0.1 cm/year; Terry et al., 2000) compared to Dora-Maira or Kokchetav they repeat the pattern already described in the Dora-Maira. A remarkably high exhumation rate of 3.1 cm/year (Sánchez-Rodríguez & Gebauer, 2000) has been suggested for rocks in and around the Ronda peridotite (Southern Spain), a unit that is considered to be the equivalent of the diamond-bearing Beni Bousera peridotite (Marocco). In this area, however, lack of good correlation between ages and metamorphic conditions, as well as extremely high cooling rates (200–340 °C/m.y.) cause considerable uncertainty on the exhumation rate and require some caution on the use of this data. In fact, a more recent study demonstrated that

354

D. Rubatto, A. Liati & D. Gebauer

metamorphic zircon in a pelitic granulite adjacent to the Ronda peridotite grew during decompression (Whitehouse & Platt, 2003). A relatively slow exhumation rate has been proposed for the large Dabie Shan UHP belt, where, however, only a minimum estimate of 0.2 cm/year (Hacker et al., 2000) was obtained because of the large scatter in cooling ages. At Lago di Cignana (Zermatt–Saas Fee ophiolites, Western Alps) exhumation rates are difficult to calculate because of the uncertainty in the age and conditions of retrogression. The best estimate on the age of the metamorphic peak, which corresponds to a depth of ~ 90 km (28–30 kbar; Reinecke, 1991), is 44.1 ± 0.7 Ma (Rubatto et al., 1998). Greenschist facies overprint probably occurred at around 38–35 Ma according to a Rb–Sr whole rock–phengite isochron (38 ± 2 Ma, Amato et al., 1999) and a U–Pb age on retrograde titanite (35 ± 3 Ma, Barnicoat et al., 1995). It results in a rate in the range 0.5–1.5 cm/year, another apparently low value. In summary, despite the numerous studies reporting exhumation rates, welldocumented estimates are only obtained where a precise age is linked to P–T conditions, a difficult task to achieve. The data available (Table 1) overall indicate fast exhumation rates of UHP rocks (on the order of centimetres per year, equivalent to tens of km per m.y.), at least for the first part of their decompressional history. These rates might be restricted to rock units of limited extent (a maximum of few km in thickness) such as the Dora-Maira. The limited dimensions are probably a crucial feature in justifying such fast movements of rocks in the upper mantle and the lower crust. Such movements of rock units on the order or a few cm/year are only comparable to the speed of plate motion. They are generally faster than average convergence rates (mm to cm/year) and much higher than average erosion rates (0.1 to 5 mm/year). This observation favours the view that the rise of UHP rocks from great depth is not related to surface erosion, but rather to the interplay at depth of tectonic processes such as buoyancy, slab break off, faults and detachment zones (Chemenda et al., 1995; England & Holland, 1979; Ernst et al., 1997; von Blanckenburg & Davies, 1995). Larger units such as the Dabie Shan and the Western Gneiss Region might have reached mid crustal levels at a slower pace. Table 1. Summary of exhumation and cooling rates of selected UHP terranes UHP terrane

Age range (Ma)

Depth range (km)

Dora Maira

35.4–32.9– 31.8–29.9 44–38 to 35

120–10 / 750–250 °C 90–30

Lago di Cignana Ronda

ca. 20–19

E' Rhodope Kokchetav WGR Dabie Shan

73.5–61.9 528 ± 3 407–401 240– ~220

WGR: Western Gneiss Region

Cooling rates Exhumation (°C/m.y.) rates (cm/year) ~ 60

340–200 >75–15 >140–35 125–60 125–

~ 60

Data source

3.4–0.5

Gebauer et al. (1997)

1–0.5

Rubatto et al. (1998) Amato et al. (1999) Barnicoat et al. (1995) Sánchez-Rodríguez & Gebauer (2000) Liati et al. (2002a) Hermann et al. (2001) Terry et al. (2000) Hacker et al. (2000)

3.1 >0.52 >1.8 <1.1 >0.2

Dating UHP metamorphism

355

Duration of UHP metamorphism We have good indication that UHP rocks rise to upper crustal levels in short time, but how long do they reside at great depth? Establishing the duration of the peak of metamorphism requires good knowledge not only of the time of distinct exhumation stages, but also of those of subduction. Constraining the timing of prograde metamorphism is made difficult by the fact that prograde minerals or structures are rarely preserved in metamorphic rocks. Armoured inclusions in garnet are the most common prograde mineral relics, also reported in UHP rocks (e.g. Chopin et al., 1991; Compagnoni et al., 1995; Hermann et al., 2001; Reinecke, 1991; Schertl et al., 1991). Even assuming that they preserved their original isotopic composition through the peak and later overprint, their dating has never been achieved. Monazite inclusions in garnet appear to generally preserve the age of peak metamorphism (e.g. Foster et al., 2000; Terry et al., 2000). However, zircon from the Dora-Maira included in a large pyrope megablast could so far be successfully dated on the prograde P–T–t path (Gebauer et al., 1997). Nevertheless, as the garnet megablasts formed at UHP conditions (ca. 700 °C and 30 kbar) shortly before the UHP peak, the early prograde path remains still to be dated. Another avenue attempted so far for dating prograde metamorphism is the determination of ages in minerals robust to HT overprint and isotopic resetting, such as zircon, that formed in prograde veins. This approach has been first used by Rubatto et al. (1999) who dated zircon domains interpreted to have grown during prograde metamorphism in quartz veins within the HP sediments of the Sesia Zone. In the Rhodope massif, a similar study determined the age of zircon contained in a prograde quartz vein concordant with the schistority of HP rocks (Liati & Gebauer, 1999). Even though the main mineral assemblage in the vein reflected rather upper amphibolite facies conditions acquired during the postpeak metamorphic overprint, the hydrothermal zircon domains were found to be ~ 3 Ma older than the domains interpreted as having formed close to the HP peak. On this basis, Liati & Gebauer (1999) proposed subduction rates of at least 1.5 cm/year (for depths between ca. 10 km and ca. 60 km). The prograde age implies that HP metamorphism was short living, a not surprising conclusion given the extreme tectonic condition in which these rocks formed. However, these are only few data and more needs to be done to understand how fast rocks can be subducted and how long they reside at depths in excess of 90 km. Age variations within an orogen: The case of the Western Alps The large number of geochronological studies in metamorphic belts where UHP rocks are present allows identification of relative differences in age of metamorphism between the UHP units and the nearby nappes. The Alps are one of the best orogens where this comparison can be made. As presented above, the Dora-Maira reached peak metamorphism at around 35 Ma (Gebauer et al., 1997; Rubatto & Hermann, 2001). To the west of the Dora-Maira Massif is the Monviso Unit, a slice of ophiolitic material comparable to the Zermatt–Saas ophiolite. The Monviso recorded eclogite facies metamorphism at lower PT conditions: ~ 600 °C and 20–24 kbar (e.g. Messiga et al., 1999; Schwartz et al., 2000). A Tertiary age of metamorphism in the Monviso was first proposed using the Ar–Ar technique

356

D. Rubatto, A. Liati & D. Gebauer

(Monié & Philippot, 1989) and then supported by Lu–Hf (49.2 ± 1.2 Ma, Duchêne et al., 1997) and Sm–Nd (60 ± 12 and 62 ± 9 Ma, Cliff et al., 1998). The discrepancy between these ages is probably due to the fact that they rely on mineral–whole rock isochrons and thus assume isotopic equilibrium, a very doubtful assumption (see above). A more recent work by Rubatto & Hermann (2003) applied ion microprobe U–Pb dating to zircon that could be proven of HP origin via mineral inclusions and trace element composition. The above authors constrained an age of 45 ± 1 Ma for the HP stage, 10 Ma older than the nearby Dora-Maira. The other Alpine unit where UHP conditions have been well documented is found at Lago di Cignana, within the Zermatt–Saas ophiolites, some 150 km to the north of the Dora-Maira (see Compagnoni & Rolfo, 2003 in this volume, for a description of the area). At Lago di Cignana, zircon crystals, some of which contained rutile inclusions, yielded an age of 44.1 ± 0.7 Ma (Rubatto et al., 1998), which was interpreted as dating the metamorphic peak. A multi-mineral Sm–Nd isochron provided an age of 40.6 ± 2.6 Ma (Amato et al., 1999) for an eclogite of the same unit. A Lu–Hf isochron on inclusion-rich garnet–whole rock yielded an age of 49 ± 3 Ma (Lapen et al., 2002). Given the relatively large errors, the presence of inclusions in garnet, and the common problem of disequilibrium and/or open system behaviour these latter ages are not straight forward. they are, however in general agreement with a metamorphic peak at ca. 44 Ma. Also in the case of Zermatt–Saas ophiolites, the nearby units, which preserve HP assemblages, recorded peak metamorphism at different times (Fig. 6). As shown in Figure 6, in the Western and Central Alps, the age of peak metamorphism in different UHP and HP nappes varies over more than 30 Ma, from the Cretaceous–Tertiary boundary in the Sesia–Lanzo Zone to the Late Eocene–Early Oligocene in the Dora-Maira and Voltri Massifs. Whatever is the tectonic scenario that can explain this variability, it implies that different UHP and HP units were subducted and exhumed at different times. For example, the UHP Zermatt–Saas ophiolite was already exhumed at upper crustal levels when the Dora-Maira was subducted to more than 120 km depth. It follows that, during the Tertiary, exhumation and subduction processes were contemporaneously active within this part of the orogen.

The challenge ahead In the last years substantial progress has been achieved in dating HP–UHP rocks, particularly in the field of correlation of time with P–T conditions. However, geochronology has only partly achieved the wishful goal of precisely dating distinct stages along UHP metamorphic paths. A number of challenges lie ahead, most of them related to improvement of techniques. As seen in the examples reported, ion microprobe U–Pb dating of single zircon, monazite and titanite domains is the most reliable and successful method to date UHP metamorphism. A better understanding of the behaviour of these minerals during metamorphism is a requisite in order to interpret better the ages obtained with U–Pb dating. Structural and textural analyses of mineral assemblages, CL and backscattered electron imaging of dated crystals, identification and petrological study of mineral

Dating UHP metamorphism

357

Fig. 6. Schematic tectonic map of the Western and Central Alps with ages of peak metamorphism in UHP and HP units. Data from: 1) Gebauer (1996); 2) Rubatto & Gebauer (1999); 3) Duchêne et al. (1997), Inger et al. (1996), Rubatto et al. (1999); 4) Rubatto et al. (1998); 5) Rubatto & Hermann (2003); 6) Duchêne et al. (1997), Gebauer et al. (1997), Rubatto & Hermann (2001); 7) Rubatto & Scambelluri (2003); 8) Liati & Gebauer (2001b); 9) Liati et al. (2002b).

inclusions, and chemical analysis (mapping and trace elements) of zircon, monazite and titanite must become routine techniques. They have the potential to allow the critical link between age and stage of metamorphism – pressure, temperature, presence of fluid and melt, and deformation. Another important step would be the generation of a consistent and reliable set of data on the trace element partitioning between zircon or monazite and key metamorphic minerals such as garnet. This data set, in part already available for zircon–garnet pairs (Rubatto, 2002; Rubatto & Hermann, 2003), would allow determining the coexistence between datable minerals and P–T indicators in a more

358

D. Rubatto, A. Liati & D. Gebauer

rigorous way. Thermobarometers involving U–Pb minerals such as the monazite–garnet proposed by Pyle et al. (2001) represent another important step toward T–t correlation. Because of the relatively large analytical errors of the ion microprobe technique, ages that correspond to different stages of metamorphism may appear as identical (e.g. Hermann et al., 2001; Kröner et al., 2000). In this respect, minimisation of the analytical errors would be a step forward in distinguishing between ages acquired during different metamorphic stages and therefore would constrain better subduction and exhumation rates. Dating prograde metamorphism remains a challenge, not only for UHP terranes. The approach of dating minerals robust to HT resetting that formed in prograde veins (e.g. Liati & Gebauer, 1999) is certainly valid, but limited by the rare occurrence and the identification of other hydrothermal minerals in such rocks. More promising for the future of dating prograde metamorphism would be inclusions of minerals such zircon and monazite armoured in garnet (e.g. Catlos et al., 2002; Foster et al., 2000; Gebauer et al., 1997; Terry et al., 2000). This is best achieved with beam techniques (ion microprobe, electron microprobe and laser ablation ICP-MS) that allow dating directly in thin section, were textural relationship can be controlled. However, the crucial point is to prove that the dated mineral really formed during prograde metamorphism and did not re-equilibrate at the peak or during later overprint. It is not enough to observe textural relationships in order to establish equilibrium between an inclusion and the host garnet. To achieve this task, a series of data acquired by means of all the methods discussed above with respect to zircon and monazite dating are necessary. Garnet is definitely the most important mineral in eclogite facies rocks, not only because it is almost always present, but also because it has the capability – under certain conditions – to record and preserve successive metamorphic events. Therefore, garnet dating is an important component for constraining prograde and cooling paths. Improvement in Sm–Nd and particularly Lu–Hf dating of garnet is thus desirable. Leaching and separation techniques need to be improved in order to eliminate or at least minimise the influence of inclusions such as zircon, monazite and apatite (e.g. Amato et al., 1999; Anczkiewicz et al., 2002; Scherer et al., 2000). Additionally, a better knowledge and modelling of the effects of inclusions in garnet, as indicated by Scherer et al. (2000), can assist the data interpretation. The closure temperature, particularly for Lu–Hf, needs to be better constrained and its variations according to chemistry, grain size and cooling history accounted for (e.g. Scherer et al., 2000; Scherer et al., 2001). Ideally, in situ dating of garnet should be achieved in order to fully exploit the potential of this mineral as chronometer. Finally, a constant effort to chemically characterise the grains dated and obtain P–T estimates directly from garnet needs to be made. Dating deformation is also a crucial task for understanding the late history of metamorphic rocks (see e.g. Müller, 2003 for a review), including UHP rocks. Micas, nearly ubiquitous in eclogite facies assemblages, are suitable for Rb–Sr and Ar–Ar dating, However, the success of these techniques in obtaining cooling ages for UHP terranes has been hampered by disequilibrium and excess argon (e.g. Giorgis et al., 2000; Kelley, 2002; Müller, 2003; Scaillet, 1998). These problems have to be carefully monitored with a combination of thermal ionisation and/or step-wise heating dating, micro-sampling or in

Dating UHP metamorphism

359

situ laser isotope analyses and chemical characterisation of the target. This should be aimed to understand its (in)homogeneity and the conditions of formation (e.g. Villa et al., 2000). Finally, for any of the dating techniques mentioned above, an effort has to be made to determine the pressure at which the mineral dated formed. In fact, in UHP rocks, which generally have a steep P–T path, determining the age of different pressure stages, and thus variations in depth, is more relevant than thermochronology.

Summary Geochronology provides important information for unravelling the evolution of UHP terranes. In order to obtain useful time constraints on UHP rocks, the geochronological information must be integrated with other geological parameters such as pressure, temperature, deformation and presence of fluid/melt. Complex mineral assemblages that include multiple generations of minerals, as well as local disequilibrium hamper the use of techniques that rely on mineral separates or bulk rock analysis. Common thermochronometers (e.g. K–Ar, Ar–Ar, Rb–Sr on micas and amphiboles) are often of little use because of relatively high peak temperatures (usually > 600ºC) and steep P–T paths recorded by UHP units. Accurate and precise geochronology of UHP rocks thus requires dating minerals that can be related to metamorphic conditions, that are robust to retrogression, and possibly sensitive to pressure changes. The dating technique (or combination of techniques) has to allow the distinction of different mineral domains, must use minerals that are robust to HT perturbations, and have good time resolution. Recent improvements have been made in sample preparation and isotopic measurements using beam techniques. Such techniques offer the best solution for achieving dating of different metamorphic stages within complex P–T paths. In key UHP terranes, most of the time constraints on the peak of metamorphism have been obtained by ion microprobe U–Pb dating of zircon, monazite and titanite. Isotopic measurements have been assisted by textural analysis, imaging techniques, study of inclusions and chemical characterisation (trace element spot analysis or mapping) of the dated mineral, in order to determine the P–T conditions at which the mineral formed. To constrain the retrograde and cooling path, U–Pb dating of minerals formed at low temperature (e.g. titanite), Sm–Nd and Lu–Hf isochrons on mineral separates (mainly garnet) and Ar–Ar dating of micas have been the major techniques used. Reliable exhumation rates of UHP units are still rare because of the problems in relating pressure, and thus depth, with age. Exhumation rates obtained so far in UHP units, such as the Dora-Maira and the Kokchetav massifs, are on the order of centimetres per year, much faster than erosion rates. These values suggest that tectonic processes including buoyancy, slab break-off, faults and detachment zones are responsible for the exhumation of UHP units of limited dimensions. The exhumation of larger units, such as the Dabie Shan and the Western Gneiss Region, might proceed at a lower pace and be related to other tectonic processes. In the Western Alps, one of the regions with the highest occurrence of UHP and HP

360

D. Rubatto, A. Liati & D. Gebauer

rocks, the large number of geochronological data allow a regional perspective on the relative timing of subduction and exhumation. It appears that subduction was diachronous across the area, and that exhumation and subduction were contemporaneously active within different parts of the orogen. The future of dating UHP metamorphism, and metamorphism in general, lies in the capability of improving metamorphism–time correlation. This requires also improvement of the dating techniques in terms of error minimisation, so that metamorphic stages with apparent identical ages can be resolved. For U–Pb dating, metamorphism–time correlation requires structural and textural analysis of rock assemblages, cathodoluminescence and backscattered electron imaging of dated crystals, characterisation of mineral inclusions, and chemical analysis (mapping and trace elements) of zircon and monazite. Sm–Nd and particularly Lu–Hf dating of garnet will have to focus on elimination of mineral inclusions, quantification of their effect on isotopic ratios, and better estimation of closure temperatures considering also the effect of fluid phases in the loss of radiogenic isotopes. In Ar–Ar and Rb–Sr dating of micas, so far the major source for cooling ages, the detection of excess argon and isotopic disequilibrium needs to be improved.

Acknowledgements The work presented in this chapter has benefited from the collaboration with a number of colleagues. The contribution of Roberto Compagnoni, Mark Fanning, Jörg Hermann, Andrei Korsakov and Ian Williams is kindly acknowledged. Stephanie Duchêne is thanked for a constructive review.

References Amato, J.M., Johnson, C.M., Baumgartner, L.P. & Beard, B.L. (1999): Rapid exhumation of the Zermatt-Saas ophiolite deduced from high-precision Sm-Nd and Rb-Sr geochronology. Earth Planet. Sci. Lett., 171:425–438. Anczkiewicz, R., Thirlwall, M. & Platt, J. (2002): Influence of inclusions and leaching techniques on Sm-Nd and Lu-Hf garnet chronology. Geochim. Cosmochim. Acta, 66:A19. Arnaud, N.O. & Kelley, S.P. (1995): Evidence for excess argon during high pressure metamorphism in the Dora Maira (Western Alps, Italy) using an ultra-violet laser ablation microprobe 40Ar/39Ar technique. Contrib. Mineral. Petrol., 121:1–11. Ayers, J.C., Dunkle, S., Gao, S. & Miller, C.F. (2002): Constraints on timing of peak and retrograde metamorphism in the Dabie Shan Ultrahigh-Pressure Metamorphic Belt, east-central China, using U-ThPb dating of zircon and monazite. Chem. Geol., 186:315–331. Barnicoat, A.C., Rex, D.C., Guise, P.G. & Cliff, R.A. (1995): The timing of and nature of greenschist facies deformation and metamorphism in the upper Pennine Alps. Tectonics, 14:279–293. Carswell, D.A. & Compagnoni, R. (2003): Introduction with review of the definition, distribution and geotectonic significance of ultrahigh pressure metamorphism. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 3–9. Catlos, E.J., Harrison, T.M., Manning, C.E., Grove, M., Rai, S.M., Hubbard, M.S. & Upreti, B.N. (2002): Records of the evolution of the Himalayan orogen from in situ Th-Pb ion microprobe dating of monazite: Eastern Nepal and western Garhwal. J. Asian Earth Sci., 20:459–479. Chemenda, A.I., Mattauer, M., Malavielle, J. & Bokun, A.N. (1995): A mechanism for syn-collisional rock exhumation and associated normal faulting: results from physical modelling. Earth Planet. Sci. Lett., 132:225–232. Chopin, C., Henry, C. & Michard, A. (1991): Geology and petrology of the coesite-bearing terrain, Dora Maira massif, Western Alps. Eur. J. Mineral., 3:263–291.

Dating UHP metamorphism

361

Claoué-Long, J.C., Sobolev, N.V., Shatsky, V.S. & Sobolev, A.V. (1991): Zircon response to diamond-pressure metamorphism in the Kokchetav massif, USSR. Geology, 19:710–713. Cliff, R.A., Barnicoat, A.C. & Inger, S. (1998): Early Tertiary eclogite facies metamorphism in the Monviso Ophiolites. J. Metamorph. Geol., 16:447–455. Compagnoni, R. & Rolfo, F. (2003): UHPM units in the Western Alps. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 13–49. Compagnoni, R., Hirajima, T. & Chopin, C. (1995): Ultra-high-pressure metamorphic rocks in the Western Alps. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism, Cambridge (U.K.): Cambridge Univ. Press, 206–243. Duchêne, S., Blichert-Toft, J., Luais, B., Télouk, P., Lardeaux, J.-M. & Albarède, F. (1997): The Lu–Hf dating of garnets and the ages of the Alpine high-pressure metamorphism. Nature, 387:586–589. England, P.C. & Holland, T.J.B. (1979): Archimedes and the Tauern eclogites: the role of buoyancy in the preservation of exotic eclogite blocks. Earth Planet. Sci. Lett., 44:287–294. Ernst, W.G. & Liou, J.G. (2000): Overview of UHP metamorphism and tectonics in well-studied collisional orogens. In Ernst, W.G. & Liou, J.G. (eds.): Ultrahigh-pressure metamorphism and geodynamics in collision-type orogenic belts /Int. Book Ser., 4/. Boulder (Co.): Geol. Soc. Am. & Columbia (Md.): Bellwether Publ. Ltd., 3–19. Ernst, W.G., Maruyama, S. & Wallis, S. (1997): Buoyancy-driven, rapid exhumation of ultrahigh-pressure metamorphosed continental crust. Proc. Natl. Acad. Sci. USA, 94:9532–9537. Foster, G., Kinny, P., Vance, D., Prince, C. & Harris, N. (2000): The significance of monazite U-Th-Pb age data in metamorphic assemblages; a combined study of monazite and garnet chronometry. Earth Planet. Sci. Lett., 181:327–340. Gebauer, D. (1990): Isotopic systems – geochronology of eclogites. In Carswell, D.A. (ed.): Eclogite facies rocks. Glasgow, London: Blackie, 141–159. Gebauer, D. (1996): A P-T-t-path for an (ultra?-)high-pressure ultramafic/mafic rock-association and its felsic country-rocks based on SHRIMP-dating of magmatic and metamorphic zircon domains. Example: Alpe Arami (Central Swiss Alps). In Basu, A. & Hart, S. (eds.): Earth processes: Reading the isotopic code. /Geophys. Monogr., 95/. Washington (D.C.): Am. Geophys. Union, 309–328. Gebauer, D., Schertl, H.-P., Brix, M. & Schreyer, W. (1997): 35 Ma old ultrahigh-pressure metamorphism and evidence for very rapid exhumation in the Dora Maira Massif, Western Alps. Lithos, 41:5–24. Giorgis, D., Cosca, M. & Li, S. (2000): Distribution and significance of extraneous argon in UHP eclogite (Sulu terrain, China): insight from in situ 40Ar/39Ar UV-laser ablation analysis. Earth Planet. Sci. Lett., 181:605–615. Hacker, R.B., Ratschbacher, L., Webb, L., Ireland, T., Walker, D. & Dong, S. (1998): U/Pb zircon ages constrain the architecture of the ultrahigh-pressure Qinling-Dabie Orogen. Earth Planet. Sci. Lett., 161:215–230. Hacker, R.B., Ratschbacher, L., Webb, L., McWilliams, M.O., Ireland, T., Calvert, A., Dong, S., Wenk, H.-R. & Chateigner, D. (2000): Exhumation of ultrahigh-pressure continental crust in east central China: Late Triassic to Early Jurassic tectonic unroofing. J. Geophys. Res., 106:13,339–13,364. Hermann, J. (2003): Experimental evidence for diamond-facies metamorphism in the Dora-Maira massif. Lithos: in press. Hermann, J., Rubatto, D., Korsakov, A. & Shatsky, V.S. (2001): Multiple zircon growth during fast exhumation of diamondiferous, deeply subducted continental crust (Kokchetav massif, Kazakhstan). Contrib. Mineral. Petrol., 141:66–82. Hirajima, T. & Nakamura, D. (2003): The Dabie Shan-Sulu orogen. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 105-144. Hunziker, J.C., Desmond, J. & Hurford, A.J. (1992): Thirty-two years of geochronological work in the Central and Western Alps: A review on seven maps. Mém. Géol. (Lausanne), 13:59 p. Inger, S., Ramsbotham, W., Cliff, R.A. & Rex, D.C. (1996): Metamorphic evolution of the Sesia-Lanzo Zone, Western Alps: time constraints from multi-system geochronology. Contrib. Mineral. Petrol., 126:152–168. Katayama, I., Maruyama, S., Parkinson, C.D., Terada, K. & Sano, Y. (2001): Ion micro-probe U–Pb zircon geochronology of peak and retrograde stages of ultrahigh-pressure metamorphic rocks from the Kokchetav massif, northern Kazakhstan. Earth Planet. Sci. Lett., 188:185–198.

362

D. Rubatto, A. Liati & D. Gebauer

Kelley, S. (2002): Excess argon in K-Ar and Ar-Ar geochronology. Chem. Geol., 188:1–22. Kröner, A., O’Brien, P.J., Nemchin, A.A. & Pidgeon, R.T. (2000): Zircon ages for high pressure granulites from South Bohemia, Czech Republic, and their connection to Carboniferous high temperature processes. Contrib. Mineral. Petrol., 138:127–142. Lapen, T.J., Mahlen, N.J., Johnson, C.M., Beard, B.L. & Baumgartner, L.P. (2002): Lu–Hf geochronology of UHP metamorphism in the Zermatt-Saas ophiolite, Lago di Cignana, Italy. Geochim. Cosmochim. Acta, 66:A431. Liati, A. & Gebauer, D. (1999): Constraining the prograde and retrograde P-T-t path of Eocene HP rocks by SHRIMP dating different zircon domains: inferred rates of heating, burial, cooling and exhumation for central Rhodope, northern Greece. Contrib. Mineral. Petrol., 135:340–354. Liati, A. & Gebauer, D. (2001a): Garnet-rich rocks recording UHP conditions in Northern Greece (eastern Rhodope zone). In Ogasawara, Y., Maruyama, S. & Liou, J.G. (eds.): Proc. UHPM Workshop 2001 at Waseda Univ. Tokyo: Waseda Univ., 275–278. Liati, A. & Gebauer, D. (2001b): U–Pb SHRIMP-dating of zircon domains from eclogites of Antrona (Western Alps): evidence for a Valais ocean origin. EUG 11 Strasbourg, J. Conf. Abstr., 6(1):600. Liati, A., Gebauer, D. & Wysoczanski, R. (2002a): U–Pb SHRIMP-dating of zircon domains from UHP mafic rocks in the Rhodope zone (N Greece); evidence for Early Cretaceous crystallization and Late Cretaceous metamorphism. Chem. Geol., 184:281–300. Liati, A., Gebauer, D. & Froitzheim, N. (2002b): Late Cretaceous basic oceanic magmatism in the Valais ocean, Western and Central Alps: Geochronological evidence and paleogeographic implications. In Annual Meeting of the Swiss Academy of Natural Sciences, Davos, Switzerland, 26. Liou, J.G. (2000): Petrotectonic summary of less intensively studied UHP. In Ernst, W.G. & Liou, J.G. (eds.): Ultrahigh-pressure metamorphism and geodynamics in collision-type orogenic belts /Int. Book Ser., 4/. Boulder (Co.): Geol. Soc. Am. & Columbia (Md.): Bellwether Publ. Ltd., 20–35. Luais, B., Duchêne, S. & de Sigoyer, J. (2001): Sm-Nd disequilibrium in high-pressure, low-temperature Himalayan and Alpine rocks. Tectonophysics, 342:1–22. Messiga, B., Kienast, J.R., Rebay, G., Riccardi, M.P. & Tribuzio, R. (1999): Cr-rich magnesiochloritoid eclogites from the Monviso ophiolites (Western Alps, Italy). J. Metamorph. Geol., 17:287–299. Monié, P. & Chopin, C. (1991): 40Ar/39Ar dating in coesite-bearing and associated units of the Dora Maira massif, Western Alps. Eur. J. Mineral., 3:239–262. Monié, P. & Philippot, P. (1989): Mise en évidence de l’age éocène moyen du métamorphisme de hautepression dans la nappe ophiolitique du Monviso (Alpes Occidentales) par la méthode 39Ar-40Ar. C.R.. Scéances Acad. Sci., Paris, Sér. 2, 309:245–251. Monié, P., Chopin, C. & Philippot, P. (1989): 39Ar-40Ar dating in the Dora Maira and Monviso massifs (Western Alps): evidence for two stages of eclogitic metamorphism. Terra Abstr., 1:264. Mposkos, E.D. & Kostopoulos, D.K. (2001): Diamond, former coesite and supersilicic garnet in metasedimentary rocks from the Greek Rhodope: a new ultrahigh-pressure metamorphic province established. Earth Planet. Sci. Lett., 192:497–506. Müller, W. (2003): Strengthening the link between geochronology, textures and petrology. Earth Planet. Sci. Lett.: 206:237–251. Paquette, J.-L., Chopin, C. & Peucat, J.J. (1989): U–Pb zircon, Rb–Sr and Sm–Nd geochronology of high- to very-high-pressure rocks from the Western Alps. Contrib. Mineral. Petrol., 101:280–289. Pyle, J.M., Spear, F.S., Rudnick, R.L. & McDonough, W.F. (2001): Monazite-xenotime-garnet equilibrium in metapelites and new monazite-garnet thermometer. J. Petrol., 42:2083–2107. Reinecke, T. (1991): Very high pressure metamorphism and uplift of coesite-bearing metasediments from the Zermatt-Saas zone, Western Alps. Eur. J. Mineral., 3:7–17. Rubatto, D. (2002): Zircon trace element geochemistry: distribution coefficients and the link between U–Pb ages and metamorphism. Chem. Geol., 184:123–138. Rubatto, D. & Gebauer, D. (1999): Eo/Oligocene (35 Ma) high-pressure metamorphism in the Monte Rosa nappe (western Alps): implications for paleogeography. Schweiz. Mineral. Petrogr. Mitt., 79:353–362. Rubatto, D. & Hermann, J. (2001): Exhumation as fast as subduction? Geology, 29:3–6. Rubatto, D. & Hermann, J. (2003): Zircon formation during fluid circulation in eclogites (Monviso, Western Alps): implications for Zr and Hf budget in subduction zones. Geochim. Cosmochim. Acta, 67:2173–2187

Dating UHP metamorphism

363

Rubatto, D. & Scambelluri, M. (2003): U–Pb dating of magmatic zircon and metamorphic baddeleyite in the Ligurian eclogites (Voltri Massif, Western Alps). Contrib. Mineral. Petrol.: in press. Rubatto, D., Gebauer, D. & Fanning, M. (1998): Jurassic formation and Eocene subduction of the Zermatt Saas-Fee ophiolites: Implications for the geodynamic evolution of the Central and Western Alps. Contrib. Mineral. Petrol., 132:269–287. Rubatto, D., Gebauer, D. & Compagnoni, R. (1999): Dating of eclogite-facies zircons: the age of Alpine metamorphism in the Sesia-Lanzo Zone (Western Alps). Earth Planet. Sci. Lett., 167:141–158. Sánchez-Rodríguez, L. & Gebauer, D. (2000): Mesozoic formation of pyroxenites and gabbros in the Ronda area (southern Spain), followed by Early Miocene subduction metamorphism and emplacement into the middle crust: U–Pb sensitive high resolution ion microprobe dating of zircon. Tectonophysics, 316:19–44. Scaillet, S. (1998): K-Ar (40Ar/39Ar) geochronology of ultrahigh pressure rocks. In Hacker, B.R. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 161–201. Scherer, E.E., Cameron, K.L. & Blichert-Toft, J. (2000): Lu–Hf garnet geochronology: Closure temperature relative to the Sm-Nd system and the effects of trace mineral inclusions. Geochim. Cosmochim. Acta, 64:3413–3432. Scherer, E.E., Munker, C. & Mezger, K. (2001): Calibration of the lutetium-hafnium clock. Science, 293:683–687. Schertl, H.-P., Schreyer, W. & Chopin, C. (1991): The pyrope-coesite rocks and their country-rocks at Parigi, Dora Maira Massif, Western Alps: Detailed petrography, mineral chemistry and PT path. Contrib. Mineral. Petrol., 108:1–21. Schwartz, S., Lardeaux, J.-M., Guillot, S. & Tricart, P. (2000): Diversité du métamophism éclogitique dans le massif ophiolitique du Monviso (Alpes occidentales, Italie). Geodin. Acta, 13:169–188. Shatsky, V.S. & Sobolev, N.V. (2003): The Kokchetav massif of Kazakhstan. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 75-103. Terry, M.P., Robinson, P., Hamilton, M.A. & Jercinovic, M.J. (2000): Monazite geochronology of UHP and HP metamorphism, deformation, and exhumation, Nordøyane, Western Gneiss region, Norway. Am. Mineral., 85:1651–1664. Tilton, G.R., Schreyer, W. & Schertl, H.-P. (1989): Pb-Sr-Nd isotopic behaviour of deeply subducted crustal rocks from the Dora Maira massif, Western Alps, Italy. Geochim. Cosmochim. Acta, 53:1391–1400. Tilton, G.R., Schreyer, W. & Schertl, H.-P. (1991): Pb-Sr-Nd isotopic behaviour of deeply subducted crustal rocks from the Dora Maira Massif, Western Alps, Italy-II: what is the age of the ultrahigh-pressure metamorphism? Contrib. Mineral. Petrol., 108:22–33. Villa, I.M. (1998): Isotopic closure. Terra Nova, 10:42–47. Villa, I.M., Hermann, J., Müntener, O. & Trommsdorff, V. (2000): 39Ar/40Ar dating of multiply zoned amphibole generations (Malenco, Italian Alps). Contrib. Mineral. Petrol., 140:363–381. von Blanckenburg, F. & Davies, J.H. (1995): Slab break-off: A model for syncollisional magmatism and tectonics in the Alps. Tectonics, 14:120–131. Whitehouse, M.J. & Platt, J.P. (2003): Dating high-grade metamorphism: constraints from rare-earth elements in zircon and garnet. Contrib. Mineral. Petrol., 145:61–74. Zheng, Y.-F., Wang, Z.-R., Li, S.-G. & Zhao, Z.-F. (2002): Oxygen isotope equilibrium between eclogite minerals and its constraints on mineral Sm-Nd chronometer. Geochim. Cosmochim. Acta, 66:625–634.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 12, 365–414

Geochemistry and isotope tracer study of UHP metamorphic rocks BOR-MING JAHN1*, DOUGLAS RUMBLE2 and JUHN G. LIOU3 1

*

Géosciences Rennes, Université de Rennes 1, 35042 Rennes Cedex, France 2 Geophysical Laboratory, C.I.W., Washington, D.C. 20015-1305, USA 3 Geological and Environmental Sciences, Stanford University, Stanford, CA 94305, USA Present address: Department of Geosciences, National Taiwan University, P.O. Box 13-318, Taipei 106, Taiwan; * e-mail: [email protected]

I. Introduction After more than a decade of research, the concept of subduction of light continental rocks into the denser upper mantle has become a well-established fact. Primary evidence for this phenomenon is found in the similarity of lithologic successions of UHP metamorphic terranes with stratigraphic sequences observed in upper continental rocks. Interlayered quartzite, marble, mica schist, and paragneiss resemble sedimentary sequences of sandstone, limestone, shale and graywacke deposited along continental margins. Concordant layers of eclogite in quartzite, marble, schist, and biotite gneiss suggest basaltic sills or lava flows. More importantly, based on chemical and isotopic analyses, eclogites from many UHP metamorphic terranes, regardless of their occurrence types, can be proven to have a continental affinity, hence a part of continental crust. Eclogites are volumetrically minor in UHP metamorphic terranes, but they are the principal rock type that bears evidence of UHP metamorphism. If eclogites were produced by metamorphism of a subducted oceanic lithosphere, as documented in the Hercynian and Alpine orogenic chains (Bernard-Griffiths & Cornichet, 1985; BernardGriffiths et al., 1985; Stosch & Lugmair, 1990; Beard et al., 1992; Thöni & Jagoutz, 1992; Miller & Thöni, 1995), their contact relation with associated continental rocks, i.e., granitic gneisses, must be tectonic. Thus, even if the eclogites possess UHP parageneses, it does not necessarily indicate that the country gneisses have also undergone the same UHP metamorphism. However, if the eclogites can be identified as part of ancient continental crust, then their presence must imply deep subduction of a continental block. This chapter will deal mainly with eclogites and particularly those from the Dabie orogen of east-central China. Eclogites from the UHP metamorphic terranes of Dabieshan, China, have three occurrence types: (1) gneiss-hosted, in which eclogites occur as enclaves within granitic gneisses, (2) marble-associated, in which eclogites form interlayers or lenses within

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

366

B.M. Jahn, D. Rumble & J.G. Liou

marble or calc-silicate rocks, and (3) layered intrusion, in which eclogites form part of a magmatic differentiation series (e.g. Bixiling and Maowu). Geochemical as well as Sr–Nd and oxygen isotopic analyses have clearly established their continental tectonic settings (Jahn, 1998, 1999; Zheng et al., 1998, 1999). On the other hand, the implication of ultramafic rocks in UHP metamorphic terranes may be more controversial. In some cases, mantle blocks could have been emplaced by an earlier tectonic process into continental crust, then subjected to subduction to mantle depths together with the host continental crust. Such a process is difficult to identify if the rocks have chemical and isotopic compositions comparable with those of upper mantle rocks. This may be represented by the Rizhao clinopyroxenites (Hiramatsu & Hirajima, 1995; Zhang et al., 2000). However, if mantle peridotites have recorded a history of crustal contamination as revealed by Nd–Sr–O isotopic analyses and/or petrological evidence, then their UHP assemblages would imply deep continental subduction. This is exemplified by garnet peridotites of Donghai-Zhimafang (Yang & Jahn, 2000). Identification of the origin of mafic and ultramafic rocks in an UHP metamorphic terrane is based much on geochemical and isotopic fingerprints. Geochemistry also plays an important role in recognising Earth surface features in subducted rocks. For example, the low G18O and low GD rocks in the Dabie and Su-Lu terranes demonstrate that UHP gneisses, eclogites and quartzites were altered by meteoric water from a cold climate in a geothermal system at Earth’s surface prior to subduction (Yui et al., 1995, 1997; Zheng et al., 1996, 1998, 1999; Baker et al., 1997; Rumble & Yui 1998; Fu et al., 1999; Rumble et al., 2002). In this chapter, emphasis will be placed on the use of chemical and isotope tracers to identify the nature of protoliths of eclogites and unravel their pre-metamorphic to post-metamorphic tectonic evolution. The tracers to be used include radiogenic (Nd–Sr) and stable (oxygen) isotopes. Some problems concerning the dating of UHP metamorphic events using the isochron methods will be discussed. Finally, an example will be given to illustrate how the isotope tracers could be used to constrain the tectonic evolution of the celebrated Dabie orogen.

II. Chemical compositions of eclogites and ultramafic rocks Eclogites are mainly of basaltic composition, but they show a wide range of major and trace element abundances. This suggests that they may have multiple origins and derived from heterogeneous mantle sources. However, unusual compositions could also be an artifact resulted from biased analyses conducted on banded rocks produced by metamorphic segregation. Eclogites and ultramafic rocks that have been recrystallised in UHP metamorphic conditions often show coarse-grained texture with distinct mineral banding of variable scale. It is often very difficult to obtain a truly representative bulk composition for the protolith of a banded eclogite. Fine-grained or homogeneous textured eclogites also occur, but they seem to be minor in comparison with heterogeneous textured facies. Eclogites from the Dabie and Su-Lu terranes have been shown to cover a wide range of SiO2 contents, from 36 to 60%, although the majority (> 70%) still have basaltic or gabbroic compositions (SiO2 = 45–52%, Fig. 1.; Jahn, 1998).

Geochemistry and isotope tracer study of UHP metamorphic rocks

367

Fig. 1. Major element variations of eclogites and associated ultramafic rocks from the Su-Lu and Dabie terranes of east-central China. Eclogites are generally basaltic and quartz-normative, with some showing the cumulative nature of their protoliths. Type I is gneiss-hosted, Type II is marble-associated, and Type III is a member of layered intrusion or associated with ultramafic rocks.

368

B.M. Jahn, D. Rumble & J.G. Liou

Higher silica eclogites (SiO2 = 53%) suggest that their protoliths are more differentiated, and lower silica ones (= 45%) imply a “cumulate” nature of their protoliths or a portion of rock rich in low SiO2 phases (e.g. garnet, epidote, rutile etc.) from a banded eclogite. Garnetite is a good example of metamorphic differentiation; it cannot be used to discuss the nature of its protolith. It is common to find that the chemical composition of an eclogite cannot be matched by any reasonable magmatic protoliths. In any case, eclogite protoliths cannot be easily classified with the commonly used classification schemes for igneous rocks because most of them, such as AFM or TAS (Middlemost, 1994), involve the use of alkali elements which are known to be mobile in metamorphic processes. This subject is discussed below. Numerous petrological studies have established that the protoliths of eclogites underwent progressive dehydration during prograde metamorphism, and many of them were later rehydrated to some extent during retrograde metamorphism when the rocks were exhumed (e.g. Liou et al., 1998; Zheng et al., 1999; Hirajima & Nakamura, 2003). These processes are not isochemical. In fact, some gneissic rocks of Dabieshan show textural and mineralogical evidence of transformation from retrograde eclogites through hydration, recrystallisation and metasomatism (Zhang et al., 2003). In this retrograde metamorphism, CO2-rich and Na+ and K+-bearing low salinity aqueous fluids reacted with the UHP eclogites (Zheng et al., 2000; Fu et al., 2001) and resulted in partial to complete transformation to biotite- and amphibole-bearing gneisses at greenschist to low amphibolite facies conditions. LIL elements (K, Rb, Cs, Sr, Ba), U and to some extent, La and Ce, could be mobilised along with volatile species such as H2O and CO2. In a normal subduction environment, dehydration of subducted oceanic crust (amphibolite and serpentinite) creates a hydrous curtain, in which hydrous fluid moves upward carrying these elements, but leave behind insoluble Nb and Ta, and metasomatizes the mantle wedge (Tatsumi & Eggins, 1995). Magmas produced by melting of such a metasomatized mantle would show enrichment in LIL elements, U and probably also La–Ce, but would be depleted in Nb and Ta. This is the characteristic geochemical signature of arc magmas formed in zones of oceanic lithosphere subduction. Subduction of a continental crust has a somewhat different chemical consequence because the continental crust is globally less hydrated, especially the lower crust which is composed mainly of anhydrous granulite facies rocks. The protolith of an eclogite could equally be a part of the upper to middle crust. It may occur as a basic dike or enclave, or as a layered intrusion (e.g. Bixiling or Maowu), within granitic gneisses; and together metamorphosed in greenschist to amphibolite facies rocks prior to the subduction to mantle depths. This scenario is often observed in Precambrian terranes. During continental subduction, greenschist or amphibolite of basaltic composition would then follow the similar route as the oceanic crust, and engaged in dehydration reactions. The pattern of elemental loss should be rather similar to that in a subducted oceanic lithosphere, except that the amount of fluid released from the continental subduction is comparatively insignificant. The present abundances of LIL elements and U, as well as REE patterns in eclogites cannot be easily used to argue for or against their mobility, because these elements show a great variation in basaltic rocks from diverse tectonic settings. Consequently, any

Geochemistry and isotope tracer study of UHP metamorphic rocks

369

difference in their concentrations could reflect the original difference in the protoliths, but not necessarily related to the gain-loss process during metamorphism. Nevertheless, the clearest evidence of elemental loss during eclogitization is from studies of the Rb–Sr isotopic systematics. Analyses on eclogites from Western Norway (Griffin & Brueckner, 1985) and from the Dabie and Su-Lu terranes (Ames et al., 1996; Jahn, 1998) indicate that many eclogites have highly radiogenic 87Sr/86Sr ratios that are “unsupported” by their Rb/Sr ratios (Fig. 2). This is particularly true for eclogites that occur as enclaves within quartzofeldspathic gneisses (Griffin & Brueckner, 1985; Jahn, 1998). The most plausible explanation is that Rb is preferentially lost relative to Sr during prograde

Fig. 2. 87Sr/86Sr vs. 87Rb/86Sr diagram for UHP eclogites from the Su-Lu and Dabie terranes. Three reference isochrons based on the protolith ages of the Weihai eclogites (1.7 Ga, Jahn et al., 1996), granitic gneisses of the Su-Lu and Dabie terranes (˜ 800 Ma) and the UHP metamorphic event (˜ 220 Ma) are shown for assessment of Rb–Sr isotopic systems in different types of eclogites. Sources: Jahn (1998), Li (2003), and Ames et al. (1996) for those occurrences “not defined”. Note that the data are highly dispersed and many show “unsupported” high Sr isotopic ratios (data left of the 1.7 Ga isochron), implying depletion of Rb and alkali elements during the metamorphic processes.

370

B.M. Jahn, D. Rumble & J.G. Liou

eclogite facies metamorphism. By analogy, elements of similar geochemical behaviour, such as K, Cs and probably Ba, could have been lost along with Rb in the same process. During the exhumation of deeply subducted continental crust, retrograde metamorphism overprints HP–UHP eclogites to variable extents. Retrograde metamorphism took place most efficiently in the presence of water. The process would rehydrate eclogites and probably rehabilitate the lost mobile elements in the course. However, the process should not be considered as reversible to prograde metamorphism. In the Dabie orogen, many examples have shown that during the entire course of continental subduction to its final exhumation, aqueous fluid phase has played little role as a whole. This is evidenced by the preservation of the world’s lowest G18O values (–10‰) in high temperature rocks (Yui et al., 1995, 1997; Zheng et al., 1996, 1998, 1999; Rumble & Yui 1998; Fu et al., 1999; Rumble et al., 2002), the world’s highest HNd values (+270) in eclogites of the Weihai area (Jahn et al., 1996) and the general absence of syntectonic granites in the orogen. This is in line with the frequent occurrence of coesite in the Dabieshan eclogites. On the contrary, if the water activity was elevated, coesite would not survive due to its easy conversion to quartz. However, localised and channelised fluid activity has also been recorded in the same metamorphic terrane (Li et al., 2001; Li, 2003). Eclogites from the European Hercynides and the Alpine chain are known to have an affinity with oceanic basalts (or MORB); whereas those from the Caledonides and the Dabie orogen have an affinity with continental basalts (see references cited by Jahn, 1999). REE patterns are invariably LREE-enriched in the eclogites of Dabieshan and SuLu (Jahn, 1998), but LREE-depleted, MORB-like patterns are abundant in the Hercynian and Alpine chains. Although LREE, especially La, might be susceptible to mobilisation in the presence of hydrous fluids, the general absence of such fluids in Dabieshan makes the identification of the protolith nature via the use of REE patterns credible (Fig. 3).

III. Sm–Nd and Rb–Sr isochron ages and Nd–Sr isotope tracers Geochronology plays a crucial role in our understanding of the processes of subduction and exhumation of continental crustal rocks. When dealing with a UHP metamorphic rock, the key questions on the timing of events include: the age of protolith, the ages of peak and retrograde metamorphism, the time of later thermal disturbance, the episode of fluid–rock interaction etc. Rate of exhumation could be estimated if precise age information can be obtained and linked to a specific stage of an orogenic process. Moreover, for the entire metamorphic terrane, comprehensive age patterns would be needed to reconstruct a scenario of tectonic evolution (e.g. Hacker et al., 1998, 2000). Different events can be dated by different chronometers. The most commonly used chronometers include U–Pb on zircon or monazite, Sm–Nd on garnet, Ar–Ar on white mica or hornblende, and Rb–Sr on biotite or white mica. Lu–Hf has also been used for garnet dating, and it was shown that this method may be superior to Sm–Nd in some cases (Duchêne et al., 1997; de Sigoyer et al., 2000; Scherer et al., 2000). However, analytical facilities of Hf isotopes are more limited, and are practiced only with MCICP-MS on eclogitic minerals (Blichert-Toft et al., 1997). Until now only a few results

Geochemistry and isotope tracer study of UHP metamorphic rocks

371

Fig. 3. REE distribution patterns of eclogites from the Su-Lu, Dabie and Hong’an terranes. All eclogites are gneiss-hosted except two from Rongcheng (Su-Lu). The common feature is the enrichment in light REE, which contrast with N-MORB but resemble continental mafic rocks (basalt, amphibolite, or basic granulite).

372

B.M. Jahn, D. Rumble & J.G. Liou

have been published. The U–Pb systems of zircon and monazite have yielded a great amount of chronological information on a variety of thermal events. However, as pointed out cogently by Thöni (2002) in his excellent summary of geochronological problems of the Alpine metamorphic rocks, even though the zircon and monazite U–Pb chronometers usually provide exact time information, they are somewhat hampered by the fact that the mechanisms of zircon and monazite (re)crystallisation are still poorly understood in the context of the P–T evolution of metamorphic rocks. In HP and UHP metamorphic terranes, zircon may grow during the prograde as well as retrograde recrystallisation (Tilton et al., 1991; Gebauer, 1996; Thöni, 2002). Metamorphic recrystallisation of zircon and resetting of its U–Pb systematics could be enhanced by fluids even at moderate temperatures (< 600 °C, Rubatto et al., 1999, 2003; Liati et al., 2000). Partial recrystallisation may even result in apparent concordant ages without geological meaning for any particular events (Pidgeon, 1992; Pidgeon et al., 1998; Li, 2003). Consequently, zircon may yield no clear petrological record in the P–T path, and it is sometimes difficult to link “punctual” U–Pb age data (e.g. SHRIMP dates) with a specific stage of metamorphic evolution. The technique of mineral isochron is based on the in situ isotopic evolution of an equilibrated mineral assemblage. In dating HP–UHP eclogite, only minerals that are formed in the eclogite facies metamorphism are useful; these include garnet, omphacite, phengitic mica etc. Presence of relict minerals or retrograde metamorphic products would likely disturb the isochron relationship between the different phases and wholerock. The mineral isochron method not only provides age information but also yields initial isotopic ratios that are very valuable in tracing the evolution of its protolith as well as the genetic relationship between UHP metamorphic rocks and spatially associated post-metamorphic rocks. Among the high-grade metamorphic minerals, garnet is known to be the most useful one in metamorphic geochronology. Garnet is ubiquitous, it occurs in a wide range of metamorphic and igneous rocks of both crustal and mantle origins. In metamorphic petrology, garnet is often used in thermobarometry, thus is most useful in the reconstruction of P–T paths. Since the mid-1980’s, garnet has become one of the most powerful metamorphic chronometers because of its high rate of radiogenic growth of Nd isotope ratios (Griffin & Brueckner, 1985; Vance & O’Nions, 1990, 1992; Mezger et al., 1992; Li et al., 1993; Chavagnac & Jahn, 1996). As emphasised by Thöni (2002), the major advantage of garnet dating is the ability to extract microstructural and thermobarometric information together. The combination of textural evidence, P–T estimates and age data allows a much better understanding of the evolution of a metamorphic terrane and hence the orogenic processes. If zircon U–Pb and Rb–Sr isochron ages are jointly used, the constraint would be much more enhanced. In any case, garnet is most useful in its Sm–Nd isotopic system, but the interpretation of garnet Sm–Nd age (or date) is not always straightforward. The most important factors that influence the garnet-based isochron chronometer are (1) isotope disequilibrium and (2) presence of high-LREE inclusions. Before the two factors are discussed, it would be instructive to understand the concentration ranges of Sm and Nd in a few most important HP–UHP minerals (garnet, omphacite, phengite etc.) and the REE partition coefficients between these minerals.

Geochemistry and isotope tracer study of UHP metamorphic rocks

373

Ranges of Sm and Nd concentrations 333 garnet analyses by isotope dilution method are compiled from the literature and our unpublished data. The ranges of Sm and Nd concentrations in garnets are shown in Fig. 4. A few observations could be made: (1) Sm and Nd concentrations are limited to 14 and 44 ppm, respectively. However, the majority of data points (250 out of 333) fall within the range of 2 ppm for both elements (Fig. 4b). (2) Overall, garnets from ultramafic rocks (solid circle and diamond) have concentrations lower than those from eclogites and granitic gneisses. (3) Sm/Nd ratios cover a wide range from 0.3 to 10. A small number of garnets (25/333) have Sm/Nd ratios less than the chondritic value of ca. 0.325; most of these have exceptionally high Nd (5 to 44 ppm), suggesting that the garnets contain significant amounts of LREE-rich inclusions, such as monazite, epidote or apatite. The same has also been underlined by Thöni (2002) in his study of eclogites from the Eastern Alps. (4) “Clean” garnets, like those from the Bixiling complex of Dabieshan, tend to have low Nd concentrations (= 0.5 ppm) and high Sm/Nd ratios (about 2). These observations suggest that LREE-rich inclusions could exert important influence on garnet geochronology (Luais et al., 2001; Thöni, 2002). Figure 4 also shows Sm–Nd concentrations of Omp–Cpx and a few Opx and Amp. Some observations: (1) Most data points have Nd d 10 ppm and Sm d 3 ppm. (2) The majority have Sm/Nd ratios falling about the chondritic value of 0.325 (0.2 to 0.5); only a few have ratios as high as 1.0. (3) Like garnets, higher Nd (t 10 ppm) Cpx tend to have lower Sm/Nd ratios, suggesting a contribution of LREE-rich inclusions. Equilibrium partition coefficients (Kd values) between Cpx and Grt Several recent studies of element abundances in eclogite minerals have established some interesting patterns of trace element partition coefficients (Harte & Kirkley, 1997; Bocchio et al., 2000; Sassi et al., 2000). These studies were mainly realised using the analytical techniques of secondary ion mass spectrometry (SIMS) and laser ablation ICP-MS. The data of the Maowu eclogite–pyroxenite body from Dabieshan is used to illustrate this point because the coherent U–Pb and Sm–Nd age data seem to indicate that REE in coexisting Cpx and Grt have attained chemical and Nd isotopic equilibrium (Rowley et al., 1997; Jahn et al., 2003). The Maowu eclogites and pyroxenites have been dated at 221 to 236 Ma by the Sm–Nd mineral isochron method (Jahn et al., 2003), and at 220 to 230 Ma by zircon and monazite U–Pb analyses (Rowley et al., 1997; Ayers et al., 2002). In addition, the metamorphic conditions have been determined at P = 40 ± 10 kbar and T = 750 ± 50 °C (Wang et al., 1990; Okay, 1994; Fan et al., 1996; Liou and Zhang, 1998). The general patterns for whole-rock and minerals observed for the Maowu rocks are very similar to those of the Norwegian eclogites (Griffin & Brueckner, 1985), the diamondiferous eclogites from Yakutia, Siberia (Jerde et al., 1993; Taylor et al., 1996) and the eclogites from the Adula Nappe of the central Alps reported by Bocchio et al. (2000). They are known to represent chemically equilibrated assemblages. Cpx/Grt partition coefficients (= Kd values) of the Maowu rocks are shown in Figure 5. The Kd’s of LREE have a variation much greater than that of HREE – up to three orders of magnitude for La, but all Kd’s decrease regularly from La (or Ce) to Lu.

374

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 4. Sm and Nd concentrations in garnet, clinopyroxene, orthopyroxene and amphibole. All data were obtained by the isotope dilution method reported in the literature (references too numerous to be cited herein). The two charts at the top represent the total ranges observed. Data of more restricted concentration ranges are shown in the two diagrams in the middle and the two at the bottom. The majority of garnet data have both Sm and Nd less than 2 ppm, and most have Sm/Nd ratios falling between 1 and 2. Higher Nd concentrations are often accompanied by lower Sm/Nd ratios, which suggest a presence of REE-rich inclusions, such as epidote, zoisite, or apatite, in garnet. Most clinopyroxenes have Sm/Nd ratios falling about the chondritic ratio of 0.325, and their Nd concentrations less than 6 ppm.

Geochemistry and isotope tracer study of UHP metamorphic rocks

375

Fig. 5. REE partition coefficients (Kd) between clinopyroxene and garnet. The data were obtained on the eclogite-pyroxenite complex of Maowu (data source: Jahn et al., 2003). Kd values for LREE vary by three orders of magnitude, whereas those for HREE by only one order.

Similar patterns have been observed for eclogites from the central Alps (Bocchio et al., 2000). The Kd’s are also quite comparable with those obtained in a high pressure (20–30 kbar) and high temperature (1300–1470 °C) experiment of Johnson (1994). Consequently, our results strongly indicate that chemical equilibrium was achieved during the UHP metamorphism. The only green amphibole shows a REE pattern almost identical to that of the coexisting Cpx, causing the REE Kd values between Amp and Cpx to be close to unity. Thus, the amphibole appears to have been transformed entirely from Cpx during retrograde metamorphism. Figure 6 shows Cpx/Grt Kd values for a few selected elements. While Mg shows little preference for Cpx or Grt, Ca and Sr are, as expected, strongly partitioned into Cpx, but Fe, Y and Zr are partitioned into Grt. The partitioning of Y is easily understood as it

376

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 6. Some selected element partition coefficients between clinopyroxene and garnet. The data were obtained on the eclogite-pyroxenite complex of Maowu (data source: Jahn et al., 2003). Note that the Kd(REE) values are positively correlated with Kd(Ca), and Zr-Y, like HREE, are preferentially partitioned in garnet.

behaves like HREE. However, the behaviour of Zr is worth some comment. It has been shown that Kd(Zr) values are consistently below 1 (0.1 to 0.6) for other Dabieshan eclogites (Sassi et al., 2000), but are greater than unity (1.6 to 2.2) for eclogites from the central Alps (Bocchio et al., 2000). In a study of mantle-derived pyroxenites and eclogites, O’Reilly & Griffin (1995) observed that Zr contents are generally higher in Grt (< 1 to 80 ppm) than in Cpx (< 1 to 50 ppm). Kd(Zr) for most samples range from 0.1 to 3, but Zr partitioning is temperature dependent (Kd proportional to 1/T). They found that Kd(Zr) decreases with increasing XJd(Cpx) and possibly XCa(Gt), but shows no P effect. Because of the relatively large Kd(Zr) in Grt, Degeling et al. (2001) suggest that metamorphic zircon could be produced by Zr-bearing Grt breakdown reaction: 2 Zrbearing Grt + 4 Sillimanite + 13 Qtz = 3 Cordierite + 4 Zircon.

Geochemistry and isotope tracer study of UHP metamorphic rocks

377

In summary, the REE Kd values for Cpx/Grt show a dependence on the Ca content of the host phases, with Kd’s decreasing with decreasing Kd(Ca), as shown by comparing Figures 5 and 6. This behaviour has also been found in mantle eclogites (O’Reilly & Griffin, 1995; Harte & Kirkley, 1997) which have much higher equilibrium temperatures of 1000–1200 °C than the Maowu UHP rocks (700–800 °C). The temperature difference appears to have little effect on the Kd values, but this view is not shared by Bocchio et al. (2000) from their analysis on alpine eclogites. Sassi et al. (2000) also observed that the patterns of REE Kd’s for all UHP eclogites from Dabieshan are similar, but the absolute values vary a great deal, from three orders of magnitude for La to about one order for Lu. Furthermore, a weak correlation between Kd(Zr) and Kd(Ca) for both mantle and Dabieshan eclogites was noticed by Harte & Kirkley (1997) and Sassi et al. (2000). Meaningful Sm–Nd and Rb–Sr isochron ages: some examples Meaningful Sm–Nd and Rb–Sr isochron ages have been obtained for the Bixiling and Maowu complexes (Figs. 7 and 8). The metamorphic ages of ca. 220 Ma for the Bixiling Complex are in perfect agreement with the SHRIMP U–Pb zircon ages obtained by Cheng et al. (2000) and Li (2003). The Maowu eclogites and pyroxenites appear to have recorded slightly older ages at ca. 230 Ma, and the results are also consistent with those obtained by the conventional (Rowley et al., 1997) as well as ion probe (CAMECA ims1270, Ayers et al., 2002) zircon analyses. Moreover, meaningful Sm–Nd and Rb–Sr isochron ages of about 620 Ma have been obtained for the Pan-African coesite-bearing eclogites from Mali (Jahn et al., 2001a). The Mali eclogites are the oldest UHP eclogites identified so far. Note that all the rocks have been recrystallised in UHP metamorphic conditions with temperatures well over 700 °C. Failure of producing correct ages Excess Ar in phengitic mica of HP–UHP rocks have been frequently documented and it produced aberrant ages (e.g. Li et al., 1994, 2000; Jahn et al., 2001). Similarly, aberrant Sm–Nd mineral isochron ages have also been obtained, particularly for “low temperature” (= 600 °C) HP–UHP metamorphic rocks. The most notorious examples are from the Alps and the Himalayas (e.g. Luais et al., 2001; Thöni, 2002). Most recently, the same phenomenon became better known to eclogitic rocks of the Hong’an Block in the western part of the Dabie orogen (Jahn & Liu, 2002). The main cause for this chronometric problem is the lack of isotopic equilibrium between garnet and its coexisting minerals. Garnet is the major Al-carrying phase in eclogite and is believed to be transformed mainly from plagioclase, whereas omphacite is derived from magmatic pyroxenes. Of course, the detail of phase reactions leading to the formation of these two minerals is much more complicated and beyond the scope of this chapter. Granting the simplified case, during the “low temperature” metamorphic transformation from plagioclase to garnet, the reconstitution of lattice-forming major elements may not be closely followed by REE’s due to their smaller diffusion coefficients. Isotopic equilibrium would not be expected to occur when chemical

378

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 7. (a) Meaningful Sm–Nd and (b) Rb–Sr (next page) mineral isochron ages were obtained for the Bixiling Complex (data source: Chavagnac & Jahn, 1996). All rocks have very similar metamorphic initial isotope ratios (HNd(T) = 0 to –2, ISr | 0.704). The amphibolite has a much younger biotite age of 179 Ma and a very different ISr of 0.707, as a result of retrograde metamorphism.

Geochemistry and isotope tracer study of UHP metamorphic rocks

379

Fig. 7. (b) See previous page.

equilibrium is not reached. In numerous cases, garnet crystals show major element zoning of prograde metamorphism. This is clear evidence for non-equilibrated growth zones. In this case, trace element and isotopic equilibrium is probably out of question. The phenomenon of Sm–Nd isotopic equilibrium/disequilibrium is most dependent on the prime factor of temperature, but little on pressure as far as goes the present knowledge. Other factors, such as the intensity of deformation, remains to be further evaluated. The temperature effect leads to the concept of blocking temperature

380

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 8. Meaningful Sm–Nd mineral isochron ages obtained for an eclogite, a garnet websterite, and a garnet clinopyroxenite from the Maowu Complex of Dabieshan (data source: Jahn et al., 2003).

Geochemistry and isotope tracer study of UHP metamorphic rocks

381

(TB; Dodson, 1973), which does not have a unique value but is also influenced by several other factors including mineral grain size, rate of cooling, duration of metamorphic reaction at temperatures higher than TB, presence of fluid phase and its composition, and penetrative deformation (e.g. Thöni, 2002). The literature data indicate that TB values range from 850 to 650 °C for garnet Sm–Nd isotope chronometer (Humphries & Cliff, 1982; Jagoutz, 1988; Mezger et al., 1992; Burton et al., 1995; Zhou & Hensen, 1995; Günther & Jagoutz, 1997; Thöni, 2002). Garnet out of isotopic equilibrium is often accompanied by non-equilibrated trace element patterns. In some cases, Sm/Nd ratios of garnets are so unusual that they are smaller than that of whole-rock samples; whereas in others, all co-existing minerals show little difference in Sm/Nd ratios so they form a cluster in an isochron diagram. In still other cases, garnet and omphacite may have established equilibrium Sm/Nd partition coefficients but not their Nd isotopic compositions, thus resulting in a negative slope “futurechron” relationship. This has been found in eclogites of Tso Morari of the Himalayas and the Sesia zone of western Alps (Duchêne et al., 1997; de Sigoyer et al., 2000; Luais et al., 2001), and the Hong’an Block of China (Jahn & Liu, 2002). Porphyroblastic garnets often contain mineral inclusions. In addition to the celebrated coesite, many of them are REE-rich phases, such as monazite, zoisite, epidote, allanite, titanite, apatite, zircon etc. Except zircon, all these phases are highly enriched in LREE with very low Sm/Nd ratios. Thus, a tiny amount of such inclusions could significantly lower the Sm/Nd ratio in garnet, hence affecting an isochron construction based on Grt–Cpx–WR and other minerals. Inclusions must have undergone the same metamorphic P–T path as garnet, Cpx and other major phases. If the metamorphic temperature exceeds the TB of the inclusions (moderate TB minerals, such as epidote, apatite and titanite), the Nd isotopic compositions of inclusions could be expected to attain isotopic equilibrium with host garnet. In this case, a correct isochron age may be expected, but the range of data spread would be reduced, and hence the statistical error in age increased. This has been most frequently observed in the “hightemperature” eclogites of the Dabie and Su-Lu terranes. On the other hand, mineral inclusions of very high TB (e.g. zircon and monazite) may not be easily reset isotopically to keep pace with growing garnet. In this case, it may produce a disequilibrated isochron relationship. In conclusion, the failure of producing a correct Sm–Nd isochron age is due to isotope disequilibrium between garnet and its coexisting minerals. The disequilibrium results from two distinct processes: (1) garnet preserves the isotopic composition of precursor mineral (plagioclase) due to the fact that the rate of Nd isotope exchange is more sluggish than that of chemical and mineral phase reconstitution; and (2) garnet contains high-temperature refractory LREE-rich inclusions, which have not been isotopically re-equilibrated with host garnet, though the garnet may have attained equilibrium with coexisting Cpx and other principal phases. The above disequilibrium does not include the open system behaviour occurring in retrograde metamorphism with strong influence of hydrothermal activity. As it will be discussed later, oxygen isotope study can provide a test of equilibrium or disequilibrium between Sm–Nd isochron minerals (Zheng et al., 2002).

382

B.M. Jahn, D. Rumble & J.G. Liou

An example from the Hong’an Block in western Dabieshan The Hong’an Block exposes a series or high P/T metamorphic rocks, with a S–N distribution from blueschist/blueschist-greenschist, amphibolite, kyanite-free and kyanite-bearing eclogites to coesite-eclogite facies rocks. The Block has been considered to provide better archives for the understanding of the tectonic evolution of the Qinling–Dabie orogen for three main reasons (e.g. Eide & Liou, 2000): (1) it is least affected by the thermal and structural overprint imposed on much of the Dabie terranes during the voluminous Cretaceous granitoid intrusion; (2) HP eclogites are widespread in the Hong’an Block and they often preserve prograde metamorphism, and can be directly linked to the blueschist/blueschist–greenschist rocks; (3) in comparison with the Dabie terrane, the better exposure of blueschist/blueschist–greenschist facies rocks offer opportunities for simultaneous structural and metamorphic analysis. Six basaltic eclogites were analysed. Two of them (Gaoqiao and Qiliping) come from the Hong’an Unit; they recorded the lowest temperatures (ca. 500 °C) and a pressure range of 16–20 kbar. The next two (Qianjinhepeng and Xuanhuadian) are from the Huwan Unit of the Sujiahe Group; their thermobarometric data are 620 °C (rim) and > 12 kbar for sample QJH01-1 and 680–700 °C and 14–18 kbar for XHD07-1. The last two were collected from within the coesite-bearing Xinxian Unit of the Dabie Group; their P–T conditions are ca. 640–680 °C and t 27 kbar. In general, the metamorphic temperatures are lower than those recorded in the Su-Lu and Dabie terranes at corresponding pressures (Zhang & Liou, 1994; Eide & Liou, 2000). The eclogites from the Hong’an and Huwan units have often been referred to as “cold eclogites” and no coesite has been identified. Published white mica Ar–Ar analyses provide cooling ages from 225 to 205 Ma for the Hong’an block, and support the peak UHPM event at about 230 Ma (Eide et al., 1994; Webb et al., 1999; Eide & Liou, 2000). Rb–Sr and Sm–Nd isotopic analyses on mineral separates (Fig. 9) yielded the following results. (1) Phengite-based Rb–Sr isochrons gave 225 ± 34 (2V) Ma for Qiliping, 212 ± 7 Ma for Tianpu, and 171 ± 17 Ma for another sample (P260) of the Xinxian Unit. The other three rocks yielded no age information as the data of coexisting minerals and WR are highly scattered. The ages of 225–212 Ma are comparable with many Ar/Ar ages obtained on a variety of rock types from the Hong’an Block (Eide et al., 1994; Webb et al., 1999; Eide & Liou, 2000). They can be interpreted as cooling ages. The meaning of the younger age of 171 Ma is not clear. (2) The pattern of Rb–Sr isotope disequilibrium is random and is independent of the petrological temperatures. (3) Sm–Nd data show even more disastrous chronological information. None of them gave meaningful ages. Negative isochron relationship is observed for two eclogites: one from the cold eclogite unit (Qiliping) and the other from the coesite-bearing unit (# P260). (4) Grt–Omp Sm–Nd tie-lines yielded “isochron ages” of 182 Ma (Gaoqiao), 260 Ma (Xuanhuadian), 420 Ma (Qianjinhepeng), and 1057 Ma (Tianpu), whereas Grt–WR tielines gave 218, 337, 250, and 820 Ma, respectively. Quite visibly, Devonian and Silurian ages, as claimed to exist by some workers, can be produced from such disequilibrated isotopic systems. (5) WR data points are always found to lie outside the Grt–Omp tielines. This indicates that isotopic compositions of coexisting minerals were not

Geochemistry and isotope tracer study of UHP metamorphic rocks

383

Fig. 9. Isotopic disequilibrium in eclogites of the Hong’an Block of western Dabieshan. The Rb–Sr systems are shown in the left diagrams and the Sm–Nd systems in the right. The non-radiogenic Sr of phengite and dispersed data points for XHD-07-1 suggests a total disequilibrium despite of its high petrological temperature of 680–700 °C. Likewise, its Sm–Nd data yield grt–omp or grt–WR ages much higher than the supposed 220 Ma. For P 260, the Rb–Sr phengite age of ca. 170 Ma is too young, and a Sm–Nd “futurechron” is obtained. Garnet is completely out of equilibrium not only for its low 143Nd/144Nd ratio, but also for its unusually low 147 Sm/144Nd (less than chondritic value). Sample TP03-2 and QLP08-1 appear to give “correct” Rb–Sr ages of 212 and 225 Ma, but no meaningful Sm–Nd ages were obtained.

384

B.M. Jahn, D. Rumble & J.G. Liou

homogenised during the metamorphism and that WR must be mass-balanced by nonanalysed accessory phases. (6) The 147Sm/144Nd ratios of garnets range from 0.17 to 0.32. This is much lower than the “normal” garnet values of 0.5 to 2.0. Besides, two garnets contain high Nd concentrations (3.5 and 6.8 ppm), which likely resulted from the presence of LREE-rich micro-inclusions (e.g. monazite minerals). The above observations raise a serious question about the validity of the published Sm–Nd mineral isochron ages on the Hong’an Block. Such disequilibrated Sm–Nd systems are rarely encountered for eclogites from the Su-Lu and Dabie terranes (e.g. Li et al., 1993, 2000; Chavagnac & Jahn, 1996; Chavagnac et al., 2001; Jahn et al., 2003) except two localities (Yakou and Yangkou) of the Su-Lu terrane (Zheng et al., 2002). In the Hong’an Block, the Xiongdian eclogite (locality identical to our Xuanhuadian sample, N31°45.14´, E114°28.50´) has been subjected to several zircon U–Pb and garnet Sm–Nd geochronological investigations. However, all the published results failed to produce consistent and interpretable ages. Clearly, the isotopic systems (Sm–Nd and U–Pb) are largely out of equilibrium in these “cold eclogites”, and some of the apparent concordant zircon dates of 300 to 480 Ma could have been produced by partial recrystallisation effect (Pidgeon, 1992; Pidgeon et al., 1998; Hoskin & Black, 2000). Radiogenic isotope tracers The protoliths of metasedimentary rocks and metagranitoids are clearly of continental origin. However, those of mafic eclogites could be controversial with regard to their oceanic or continental heritage. Yet, eclogites are the most important rock type that provides the best evidence of UHP metamorphism. Subduction of the oceanic crust leads to progressive metamorphism and formation of eclogite when the plate descends beyond the hydrous curtain. If an oceanic eclogite happens to be exhumed by any tectonic process and melanged into a continental gneiss complex, then the presence of UHP minerals such as coesite in the eclogite is not sufficient to argue for a process of continental subduction. Formation of eclogite from a subducted oceanic lithosphere is a natural consequence of plate tectonics; it is nothing special. However, if protoliths of UHP eclogite can be proven to be of continental origin, then the case of continental subduction can be established. The long-standing controversy on the tectonic relationship, “in situ” vs. “foreign”, between eclogites and country gneisses very much hinges on the identification of oceanic or continental affinity of the protoliths. Broadly, eclogites produced by metamorphism of subducted oceanic crust and mantle eclogite enclaves in kimberlites do not imply continental subduction. Whereas UHP eclogites formed from the mafic components (greenschists, amphibolites, basic granulites) of an old and stable continent would adamantly argue for subduction of a continental block. The distinction of oceanic vs. continental origin for an eclogite must rely on the analyses of trace element and radiogenic isotope compositions. Isotope tracer studies have made important contributions to the understanding of the processes involved in deep burial and exhumation of continental rocks. For instance, the very low oxygen isotope ratios of UHP rocks (eclogites and country rocks) from SuLu and Dabieshan have been taken to imply that these rocks were once exposed at the

Geochemistry and isotope tracer study of UHP metamorphic rocks

385

Earth’s surface, and the preservation of such low ratios indicate that the later deep subduction and following exhumation have not re-equilibrated the oxygen isotopes (Yui et al., 1995, 1997; Zheng et al., 1996, 1998, 1999; Baker et al., 1997; Rumble, 1998). On the other hand, very spectacular ultra-high HNd values of ca. +170 to +260 have been determined for six retrograded eclogites from Weihai (Su-Lu region; Jahn et al., 1996). These are the highest values ever measured in terrestrial rocks (not minerals). They require a very long time lapse for any reasonable Sm/Nd systems to develop, and such values allowed us to estimate rather precisely the protolith age of about 1.7 Ga (Jahn et al., 1996). The preservation of extraordinary low G18O and extremely high HNd values in UHP rocks indicates that a pervasive fluid phase was absent and the rate of subduction and exhumation was too fast to re-equilibrate oxygen isotopes (Zheng et al., 1998, 1999) and to rehomogenise the extreme Nd isotope compositions. Furthermore, isotope tracer studies also provide information on the behaviour of various elements and isotopes in rocks that have undergone UHP metamorphism. Until present, such behaviours are not well understood. A rather comprehensive survey of whole-rock Nd isotopic compositions of eclogites from worldwide occurrences was presented by Jahn (1999). A typical diagram to be presented is shown to have Nd isotopic composition, expressed as HNd value, as the ordinate, and 147Sm/144Nd ratio as the abscissa, which is roughly equivalent to the chemical nature of rocks (Fig. 10). 147Sm/144Nd ratios = 0.20 (or chondritic value of 0.1967) imply light rare earth element (LREE) depletion, whereas lower ratios indicate LREE enrichment. In each set of data, two companion plots of HNd(0) vs. 147Sm/144Nd and HNd(T) vs. 147Sm/144Nd are presented. The HNd(T) values in most cases are calculated based on the metamorphic ages, hence they are “metamorphic initial ratios”. The time interval ('T) between the protolith formation and metamorphism are not known in most cases. True HNd(T) for the protoliths require further correction based on 'T values. Figures 10 and 11 present the isotopic characteristics of eclogites from five classic orogenic belts: Alpine, Hercynian, Caledonian, and Dabie–Su-Lu. Some interesting observations: 1. The most celebrated UHP rocks, jadeite-kyanite or garnet quartzites (also termed as whiteschists) of Dora-Maira, are shown to have negative HNd(T) values of –5 to –8 (Figs. 10a, b). This is consistent with their ultimate derivation from middle Proterozoic continental crust as estimated from their TDM ages of 1.3 to 1.8 Ga (Tilton et al., 1989, 1991). Except these metasediments, the entire population of eclogites from the Alpine chain possess positive HNd(T) values, suggesting that all of them originated in subducted oceanic crust, or directly from the depleted upper mantle such as the garnet peridotites of the Alpe Arami (Becker, 1993). This conclusion is largely supported by available geochemical and petrological data (Paquette et al., 1989; Thöni & Jagoutz, 1992; Miller & Thöni, 1995; Dobrzhinetskaya et al., 1996; von Quadt et al., 1997). 2. In the Hercynian belt, most eclogites possess positive HNd(T) values (Figs. 10c, d), suggesting their dominant oceanic derivation (Bernard-Griffiths & Cornichet, 1985; Bernard-Griffiths et al., 1985; Stosch & Lugmair, 1990; Beard et al., 1992, 1995; Medaris et al., 1995). One eclogite from Münchberg and two from Schwarzwald with low HNd(T) values of about 0 to –3 are due probably to crustal contamination (Stosch &

386

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 10. HNd(0) vs. 147Sm/144Nd plots (left) and HNd(T) vs. 147Sm/144Nd plots (right) for eclogites and garnet peridotites from the Alpine, Hercynian and Caledonian belts. Data source: see Appendices of Jahn (1999). For the Alpine belt, except the UHP metasediments of Dora-Maira, all the rocks show positive HNd(T) values. Most eclogites and garnet peridotites from the Hercynian belt show positive HNd(T) values, suggesting their oceanic affinity. Those having negative values have been interpreted as due to crustal contamination by reinjection of crustal rocks (oceanic clays) into the upper mantle. As for the eclogites and garnet peridotites from the European Caledonides, the majority of the rocks show negative HNd(T) values, in strong contrast to those of the Alpine and Hercynian chains. These rocks appear to have resided in the continental crust for a lengthy period of time before their UHP metamorphism.

Geochemistry and isotope tracer study of UHP metamorphic rocks

387

Fig. 11. HNd(0) vs. 147Sm/144Nd plots (left) and HNd(T) vs. 147Sm/144Nd plots (right) for eclogites from the Su-Lu and Dabie terranes of China. All types I & II eclogites (gneiss and marble-hosted) are characterised by negative HNd(T) values, suggesting their continental affinity. The data of the Weihai eclogites are out of the scale, and their HNd and 147Sm/144Nd ratios are given in the parentheses. Data sources: see Appendices of Jahn (1999).

Lugmair, 1990; Kalt et al., 1994). A few eclogites from the Bohemian Massif also show negative HNd(T) values (–3 to –6, Medaris et al., 1995), but such isotopic signature was interpreted as due to involvement of oceanic clay in the melting of subducted oceanic lithosphere, and the eclogites were considered to represent high-pressure cumulate of the “enriched” basaltic liquid (Medaris et al., 1995). In any case, the oceanic crust origin of the eclogites from the Hercynian belt is clearly demonstrated, and the negative HNd(T) values do not imply a deep subduction of the continental crust. 3. Eclogites from the Caledonides have a much more complex evolution. Two eclogites from NW Scotland (Sanders et al., 1984) and over half of the UHP rocks (including about equal proportions of eclogites and garnet pyroxenites; Griffin & Brueckner, 1980, 1985; Mørk & Mearns, 1986; Mearns, 1986; Jamtveit et al., 1991) show negative HNd(T) values down to as low as –10 (Figs. 10e, f). It appears that the Norwegian and Scottish eclogites have both oceanic and “continental” affinities. It merits a further explanation about the meaning of “continental” affinity. From the compositional point of view, all the analysed rocks (eclogites, garnet pyroxenites and garnet peridotites) are clearly of mantle origin. Some, if not all, of them were emplaced into the continental crust in

388

B.M. Jahn, D. Rumble & J.G. Liou

Precambrian time(s) and became an integral part of the continental crust since then. This represents the first stage of evolution. However, the process of the first stage emplacement could not be precisely determined, as it is a problem apart, a controversy about the styles of the early Precambrian tectonics. Note that many eclogites have 147 Sm/144Nd ratios lower than the chondritic values of 0.1967, suggesting that they were likely emplaced first as LREE-enriched continental basalts, later metamorphosed to amphibolites, and finally became an integral part of the continental crust. This further implies that an ancient crustal segment was subducted to mantle depths and UHP metamorphic assemblages were produced during the Caledonian orogeny. In the Western Gneiss Region of Norway, eclogites recorded only the Caledonian thermal event of 410–440 Ma (Griffin & Brueckner, 1980, 1985; Mørk & Mearns, 1986; Mearns, 1986; Carswell & Cuthbert, 2003), whereas garnet peridotites and garnet pyroxenites often yielded a Proterozoic event of about 1.7 Ga (Jamtveit et al., 1991). It is possible that the isotopic systems registered at 1.7 Ga in garnet peridotites at mantle P–T conditions have not been erased or homogenised during the Caledonian event. It is equally possible that these garnets were not isotopically equilibrated hence the ages of 1.7 to 1.0 Ga render no strict geological significance. Nevertheless, newly formed garnets from basaltic protoliths appear to have faithfully recorded the Caledonian orogeny at 410–400 Ma. Thus, the evolution of the European Caledonides is clearly distinguished from that of the Alpine and Hercynian chains of western Europe. 4. The Dabie and Su-Lu terranes show quite a distinguished “continental” isotopic signature (Fig. 11). Eclogites occurring as enclaves or blocks in granitic gneisses have low HNd(T) values (–6 to –20) at the time of peak metamorphism at ca. 220–230 Ma. These values are among the lowest of the known eclogites and are significantly different from those of the Alpine and Hercynian chains in western Europe. Moreover, most rocks have 147Sm/144Nd ratios less than the chondritic value, suggesting their LREE-enriched geochemical characteristics. To some extent, only eclogites from the Caledonian belt (e.g. NW Scotland and western Norway) have Sm–Nd isotopic compositions comparable with those of the Dabie Orogen. The isotopic data and geochemical characteristics undoubtedly indicate that the eclogite protoliths of the Dabie Orogen have the “continental” affinity and could not be produced from the subducted Tethyan oceanic slab. The low HNd(T) values further require the protoliths of these eclogites to have formed long before the Triassic collision, probably in middle to late Proterozoic times. It appears that, in addition to the UHP metasedimentary rocks of Dora-Maira, Dabieshan (including Su-Lu) and the Western Gneiss Region (WGR) are two regions where coesite and diamond-bearing eclogites and ultramafic rocks show clear evidence of subduction of ancient and cold continental crustal blocks to mantle depths. The eclogite–ultramafic suites in the Su-Lu and Dabie terranes have two different tectonic origins. The first suite, represented by the Bixiling and Maowu complexes, comprises layered intrusions initially emplaced in crustal levels and later subjected to UHP metamorphism as a result of continental subduction (Okay, 1994; Zhang et al., 1995a; Chavagnac & Jahn, 1996; Fan et al., 1996; Liou et al., 1996, 1997). The second suite is made up of mantle rocks exhumed together with crustal UHP metamorphic rocks; they are considered here as “tectonic enclaves of mantle origin” within granitic gneisses.

Geochemistry and isotope tracer study of UHP metamorphic rocks

389

A probable example is the garnet pyroxenites of Rizhao. Even though these pyroxenites represent igneous cumulate rocks from a mantle derived liquid, the magma differentiation probably occurred within the mantle. The isotopic systematics of the Maowu layered intrusion (HNd(T) | –5 to –6, ISr | 0.707–0.708; Jahn et al., 2003) are distinguished from that of the Bixiling complex. Upper crustal contamination during magma chamber processes is likely to have occurred in the Maowu intrusion (Jahn et al., 2003), whereas the Bixiling complex has undergone an AFC process in the lower crustal condition (Chavagnac & Jahn, 1996; Li, 2003). In conclusion, basaltic (or gabbroic) eclogites of continental origin are best shown in the Dabie and Su-Lu terranes in east-central China and the Western Gneiss Region of Norway. Their identification provides strong evidence for subduction of large blocks of continental crust. On the other hand, most eclogites from the Alps and Hercynian belts are of oceanic origin and many have a MORB-like affinity. Their occurrence is not sufficient to indicate that the eclogite-bearing gneiss terranes have also been subducted to mantle depths, unless it can be proven that the oceanic rocks were tectonically emplaced in a continental setting, and the ensemble was subducted and metamorphosed at great depths. Nevertheless, the evidence for continental subduction in the western Alps comes from the coesite-bearing metasedimentary rocks of Dora-Maira.

IV. Oxygen isotope tracer Two major contributions have been made by studies of oxygen isotope compositions of UHP rocks and constituent minerals, particularly those from the Dabie and Su-Lu terranes. (1) Revelation of strong water–rock interactions prior to the Triassic subduction; the aqueous fluid was extremely depleted in 18O and the preservation of such low G18O isotopic signature implies a very low water mobility during the entire process of subduction and exhumation (Yui et al., 1995, 1997; Baker et al., 1997; Rumble, 1998; Rumble et al., 2002; Zheng et al., 1996, 1998, 1999; Fu et al., 1999). In fact, this has been used as a palaeoclimatic proxy, tracing the Neoproterozoic palaeoenvironment (e.g. Rumble, 1998; Rumble et al., 2002, 2003; Zheng et al., 2003). (2) Better understanding of O isotope fractionation between UHP minerals, and identification of isotopic disequilibrium in O and Nd, and hence validation of the Sm–Nd isochron chronometer (Zheng et al., 2002).

G18O values of eclogites from UHP metamorphic terranes – a summary Figure 12 summarises the oxygen isotope compositions of the principal eclogitic minerals from the Dabie and Su-Lu UHP terranes. Additional data from mantle xenoliths and eclogite facies rocks of Dora-Maira (Alps) and the Western Gneiss Region (Norway), as well as common metamorphic quartz are also shown for reference. In general, crustal rocks have a wide range of G18O values but the majority of them are greater than zero. Mantle peridotites and mantle-derived basalts have rather homogeneous values between +5 and +6‰. Seawater is defined to have both G18O and GD equal to zero. Meteoric waters (rain, groundwater, snow, ice etc.) have negative G18O and GD values. A comprehensive review of published G18O values of quartz from

390

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 12. Range of G18O values for metamorphic minerals and rocks. Modified after Rumble (1998) with additional data of Zheng et al. (1998, 1999). The Qinglongshan data show the lowest G18O values (˜ –10‰) for any magmatic and metamorphic rocks. For eclogites from other localities (Shuanghe, Bixiling, Huangzhen and Maowu), all minerals seem to have low G18O values about 0‰, but preserved equilibrium high-temperature isotope fractionation with the order of quartz > omphacite > garnet > rutile.

Geochemistry and isotope tracer study of UHP metamorphic rocks

391

metamorphic rocks worldwide by Sharp et al. (1993) shows that the lower limit of metamorphic quartz is +7‰, and the entire range extends to about +30‰ (Fig. 12). The discovery of very low G18O values for garnet and omphacite (ca. –10‰) and quartz (–7‰) in coesite-bearing eclogites of Qinglongshan, China, was most surprising because they are among the lowest ever recorded in any high-temperature rocks (igneous or metamorphic; Yui et al., 1995; Zheng et al., 1996; Blattner et al., 1997). Zircons from granitic gneisses of Qinglongshan have also recorded low G18O values, from 0 to –7.5‰ (Rumble et al., 2002). Garnets and omphacites from many localities of Dabieshan also show negative values, though they are not as depleted in 18O as in Qinglongshan eclogites. Unusually low GD values of –113 to –124‰ (VSMOW) were found in phengites of Qinglongshan eclogite and quartzite (Rumble & Yui, 1998). The GD values are not as spectacular as the low G18O, but they are at the low end of the total range of natural variation in metamorphic micas (Fig. 8 of Sharp et al., 1993). Masago et al. (2003) reported negative G18O values (–3.9‰) for minerals in eclogite and whiteschist of the Kokchetav Massif in Kazakhstan. The Kokchetav Massif is the second recognised UHP region that preserves a significant effect of such water–rock interaction prior to subduction. What do these oxygen isotopic data tell us? Many recent studies on the Dabie and Su-Lu terranes seem to have reached the following conclusions: (1) the protoliths of eclogites and associated gneisses were subjected to hydrothermal alteration involving very light meteoric waters prior to Triassic subduction; (2) the preservation of such distinctive isotopic signature for both eclogites and their host gneisses supports the hypothesis of in situ tectonic relationship between them, with a scale of structural coherence of at least 100 km; (3) the persistence of pre-metamorphic isotopic distinction for the UHPM rocks implies a very limited H2O activity, hence no or little pervasive fluid free to infiltrate the rocks that have undergone subduction, UHP metamorphism and exhumation; and (4) the preservation of high-temperature equilibrium oxygen isotope fractionation between constituent minerals, as shown by the same descending order in G18O from quartz, omphacite/garnet, to rutile (Fig. 12), also indicates a low fluid activity so that the high-temperature isotopic fractionation was not re-equilibrated after the peak metamorphism (see also below). The above points are also in support of a high rate of subduction and exhumation of these UHP terranes. However, this generality does not exclude some localised high fluid activity leading to the formation of kyanite-bearing quartz veins in Dabieshan (Li et al., 2001) and hydrous mineral veins (e.g. epidote) in eclogites of Qinglongshan (Li, 2003). Limited fluid activity Fluids of high pressure (HP) and UHP metamorphic rocks are known to exist as intergranular phases, and, indeed, must be present to facilitate the attainment of mm to cm-scale element and stable isotope exchange equilibrium between adjacent mineral grains during prograde metamorphism (Philippot & Rumble, 2000; Scambelluri & Phillipot, 2001). Fluids participate in many metamorphic reactions; they act as catalyst in solid–solid reactions, lead to compositional change (metasomatism) and control

392

B.M. Jahn, D. Rumble & J.G. Liou

rheological properties. In the Dabie and Su-Lu UHP terranes, deeply subducted rocks contain minor fluids in hydrous and carbonate phases, there is little evidence for H2O to be a separate phase (e.g. Liou et al., 1997). In such fluid-deficient environments, metamorphic reaction rates are severely retarded, and UHP index minerals and geochemical signature of the protoliths may be preserved. Fluid mobility is limited during prograde UHP metamorphism. Owing to the diversity of protoliths undergoing subduction, there is ample textural, chemical, and isotopic evidence of fluid behaviours. Evidence of limited fluid mobility during prograde HP and UHP metamorphism is provided by the failure of rocks to attain textural equilibrium. Preservation of gabbroic textures and primary igneous minerals including plagioclase, orthopyroxene and biotite in the core of a 3 m coesite-bearing eclogite block from Su-Lu (Zhang & Liou, 1997) and the occurrence of intergranular coesite (Liou & Zhang, 1996) suggest a lack of fluid to facilitate equilibration. Similarly, cover–basement relationships in the Dabie UHP terrane, which in part preserve a primary unconformity, pillow structures, and mineral fabrics, have been suggested (Dong et al., 2002; Oberhänsli et al., 2002). In addition, a magmatic mingling or partial melting texture has been preserved in a Neoproterozoic bimodal suite in the margin of the Bixiling Complex. Evidence of limited fluid mobility during prograde HP and UHP metamorphism is also seen in the failure of dissimilar protoliths to achieve stable isotope equilibrium on mm, cm, and m scales despite extreme metamorphic conditions (Rumble & Yui, 1998; Baker et al., 1997; Zheng et al., 1998, 1999; Fu et al., 1999; Philippot & Selverstone, 1991; Selverstone et al., 1992; Getty & Selverstone, 1994; Nadeau et al., 1993; FruhGreen, 1994; Barnicoat & Cartwright, 1997; Miller et al., 2001). In fact, the failure to equilibrate confers an advantage to unravel the geologic history of HP and UHP rocks because remnants of pre-metamorphic stable isotopic signatures, diagnostic of protolith origins, may have survived. Vestiges of distinctive Earth surface environments and processes have been preserved in HP and UHP metamorphic rocks, and they greatly facilitate the interpretation of the geodynamic cycle of continental collision (Rumble, 1998; Rumble et al., 2003; Zheng et al., 2003). In the European Alps, stable isotopic evidence of ocean floor and ophiolitic hydrothermal alteration has been observed in HP eclogites (Philippot et al., 1998; Barnicoat & Cartwright, 1997; Miller et al., 2001; Scambelluri & Philippot, 2001). A record of Neoproterozoic surface conditions has been found in Chinese UHP rocks including low G18O and GD values indicating a cold climate (Yui et al., 1995; Zheng et al., 1996, 1998, 1999, 2003; Baker et al., 1997; Rumble & Yui, 1998; Fu et al., 1999). Dabieshan marbles carry the high G13C values of carbonate sediments that typically accompany Neoproterozoic tillites (Yui et al., 1997; Baker et al., 1997; Zheng et al., 1998; Rumble et al., 2000). New U–Pb dating and G18O analyses of zircons from orthogneisses show that a cold-climate geothermal system covering hundreds of square kilometres existed during the Neoproterozoic, consistent with snowball Earth conditions (Rumble et al., 2002). The limited fluid activity is indicated not only by stable isotope data but also by fluid inclusion studies. Low salinity, primary fluid inclusions reported from low G18O eclogites at Qinglongshan, China, may be samples of Neoproterozoic meteoric water that survived subduction and exhumation (Fu et al., 2002).

Geochemistry and isotope tracer study of UHP metamorphic rocks

393

Preservation of equilibrated high-temperature isotope fractionation Oxygen isotopic compositions of eclogite minerals have attained isotope equilibrium at their eclogite facies temperatures. The equilibrium is indicated by the fractionation between individual minerals, as expressed by ' values (G18OA – G18OB). As shown in Figure 12, the sequence of 18O enrichment in eclogite minerals is consistent with the empirically, experimentally and theoretically determined values for equilibrium fractionation (see references cited in Zheng et al., 1998). In fact, calculated oxygen isotope temperatures have been shown to be quite comparable with petrological temperatures (Yui et al., 1995; Zheng et al., 1998, 1999). Figure 13 (Zheng et al., 1998) shows that for the Shuanghe eclogites of Dabieshan (solid circles), oxygen geothermometry yields 545–735 °C for quartz–garnet pairs and 555–680°C for quartz–omphacite pairs. These temperatures have ca. 30–50 °C of uncertainty due to analytical error and fractionation curve calibration. Generally, they are slightly lower than, but close to, the petrological temperatures of 600–750 °C (Okay, 1993; Wang et al., 1995; Cong, 1996; Cong et al., 1995). Similarly, for the Donghai eclogites (Qinglongshan included; solid triangles), the oxygen geothermometry gives 650–765 °C for quartz–garnet pairs, 655–760 °C for quartz–omphacite pairs, 620–755 °C for quartz–phengite pairs, and 665–755 °C for quartz–kyanite pairs (Yui et al., 1995; Zheng et al., 1998). These temperatures are also slightly lower than, or pretty close to, the petrological temperatures of 700–850 °C (Hirajima et al., 1990, 1992; Zhang et al., 1995b). In both cases, the peak metamorphic oxygen isotope equilibrium has been preserved and this suggests little isotope resetting during later exhumation. This, in turn, argues that water–rock interaction, which resulted in low G18O values in eclogites as mentioned above, must have taken place before the UHP metamorphism, and most likely, soon after the protoliths were formed about 700–800 Ma ago. Figure 14 illustrates oxygen isotope fractionation values ('18O) between garnet and omphacite vs. G18O of garnet in eclogites from numerous localities of the Su-Lu and Dabie terranes. The range of isotope equilibrium fractionation at eclogitic temperatures ('18O = 0 to 2) is shown by grey area for reference. Garnet covers a wide range of G18O values from +8 to –10, but the majority of eclogites (70%) fall in the area of isotope equilibrium. Other rocks of positive fractionation (' = 2.2) are out of equilibrium, whereas many eclogites (ca. 30%) show negative or reverse fractionation, resulting in quartz–omphacite isotope temperatures lower than 400 °C (Zheng et al., 1999). This phenomenon of disequilibrium has been ascribed to retrograde hydration reactions, in which the responsible fluids were probably derived from the exsolution of dissolved hydroxyl in UHP metamorphic minerals (Zheng et al., 1999), or the fluids that had equilibrated with low G18O country gneisses (Yui et al., 1997; Zheng et al., 1999). Garnet is very resistant to oxygen isotope exchange during cooling. Consequently, it has presumably preserved original G18O values acquired before or during the eclogite facies metamorphism. The large range of G18O in garnet shown in Figure 14 roughly correspond to the range of whole-rock values, and it can be explained by three possible processes (Zheng et al., 1998): (1) inhomogeneous water–rock interaction prior to the UHP metamorphism; (2) isotope exchange with crustal fluids during subduction

394

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 13. Plots of G18O values of quartz vs. coexisting minerals in eclogites from Shuanghe in the Dabie Mountains (solid dots) and Donghai of the Su-Lu terrane (solid triangles). Equilibrium isotope fractionation lines at different temperatures are shown for reference (after Zheng et al., 1998).

Geochemistry and isotope tracer study of UHP metamorphic rocks

395

Fig. 14. Oxygen isotope fractionation between omphacite and garnet ('18O) vs. G18O values of garnet from eclogites of the Dabie and Su-Lu terranes. The reference lines of '18O = 0 and 2‰ represent the limit of equilibrium fractionation between these two minerals. Data outside of this range are out of equilibrium. Data sources: Yui et al. (1997), Zhang et al. (1998), Zheng et al. (1998, 1999).

(prograde metamorphism) and/or exhumation (retrograde metamorphism); and (3) heterogenous crust–mantle interaction during the residence of subducted rocks at mantle depths. Of the three listed possibilities, water–rock interaction prior to or during subduction is the most likely explanation for the heterogeneity of garnet G18O values. Possibility (3) may be excluded because there is no evidence of oxygen isotope exchange between mantle peridotites and their UHP wall rocks (Zhang et al., 2000). As for possibility (2), retrograde metamorphism could exert a significant isotope exchange (both stable and radiogenic) in clinopyroxenes (Xie et al., 2003), but the maintenance of high-temperature garnet–quartz 18O/16O fractionation indicates that the range in oxygen isotope composition in garnet has little connection with retrograde metamorphism. Conclusions from oxygen isotope tracer studies Oxygen isotope tracer studies have provided the following important conclusions. (1) Garnet, and hence bulk rock, of UHP eclogites from the Dabie orogen shows a very large variation in oxygen isotope composition (G18O = –10 to +8‰), with the majority lower than +6‰, i.e. the presumed value of pristine mantle-derived basaltic rocks. Such variation is considered to have been inherited from the protoliths as a result of uneven water–rock interaction, in which the hydrothermal system was charged with meteoric

396

B.M. Jahn, D. Rumble & J.G. Liou

water from a cold climate. Rumble et al. (2002) suggested that the Qinglongshan’s cold climate could be a manifestation of Neoproterozoic “snowball” Earth. (2) Despite the large variation in oxygen isotope composition of the protoliths, most rocks show equilibrium isotope fractionation between constituent minerals. The preservation of protolith isotope signatures implies little hydrous fluid activity during subduction. Thus, the lowering of G18O values must have taken place before the subduction. On the other hand, the safeguarding of equilibrium isotope fractionation of eclogite facies temperatures equally indicate the absence of pervasive fluid activity during exhumation. However, channelised fluid flows were present locally as witnessed by the formation of kyanite-bearing quartz vein at Huangzhen and epidote-rich veins at Qinglongshan. (3) The presence of disequilibrium isotope fractionations in some eclogites indicates the effect of retrograde metamorphism (re-hydration reactions). The responsible fluids were probably equilibrated with low-G18O country gneisses (Yui et al., 1997) or came from structural hydroxyls originally bound to anhydrous minerals (Zheng et al., 1999). Coupled Nd and O isotopic disequilibrium One of the most interesting results in isotopic studies of UHP eclogites is the demonstration of a direct correspondence in equilibrium (or disequilibrium) state between the Sm–Nd and O isotopic systems in eclogitic minerals (Zheng et al., 2002). The state of O isotope equilibrium may provide a critical test for the validity of the Sm–Nd mineral isochron ages. Zheng et al. (2002) analysed O and Sm–Nd isotopic compositions for four eclogites from southern Su-Lu in China. In the first two cases where Omp–Grt oxygen isotope fractionation is normal (' = 1.4 to 0.8‰ at 600 to 900 °C; Zheng, 1993) and produces oxygen isotope temperatures compatible with those estimated from petrological geothermometry, their Sm–Nd isotope data also yield reasonable isochron ages of about 220 Ma. By contrast, the third eclogite from Yakou shows an Omp–Grt oxygen isotope temperature of 545 °C, much lower than the petrological temperature of 700–800 °C. The low temperature is likely due to isotope re-equilibration during retrograde metamorphism at amphibolite facies conditions. Its Grt and Omp Sm–Nd systems produced an aberrant age of 280 Ma. The fourth eclogite from Yangkou, near Qingdao city, shows a reversed Omp–Grt isotope fractionation (' = –0.5‰), clearly indicating oxygen isotope disequilibrium. In this case, the Sm–Nd systematics of Grt, Omp, Phg and WR fail to yield any isochron relationship. The correspondence of equilibrium state between the two isotope systems (Sm–Nd and O) is illustrated by Figure 15 (Fig. 5 of Zheng et al., 2002). Interpretation of such a correspondence is similar to that for the correlation of equilibrium states between the chemical (trace element) and Nd isotope compositions as discussed earlier. As explained by Zheng et al. (2002), the transformation of a basaltic rock to an eclogite may be expressed by a simplified reaction: Pl + Diop = Grt + Q + Omp. In this reaction, the Si2O6 structural unit in diopside is conveyed to omphacite, whereas the Si3O8 and Si2O8 units in plagioclase are reorganised to form SiO4 and SiO2 in garnet and quartz, respectively. In an incomplete isotope exchange as observed in the Yangkou eclogite, Grt inherits the oxygen isotope signature of the plagioclase structural units and

Geochemistry and isotope tracer study of UHP metamorphic rocks

397

Fig. 15. Relationships between the Nd and O isotope ratios of garnet and omphacite from eclogites of the SuLu terrane (data and diagram from Zheng et al., 2002).

thus has higher G18O than coexisting Omp, whereas Omp inherits the oxygen isotope composition of the Si2O6 structural unit in its precursor diopside with little change in G18O value. This work and subsequent studies (Zheng et al., 2002) provide an additional insight into the kinetics of isotopic disequilibrium. Aberrant Sm–Nd isochron ages are shown to be accompanied by oxygen isotope disequilibrium. Based on the literature data (see summary by Zheng et al., 2002), the rates of O diffusion in Grt and Omp are lower than, or close to, those of Nd diffusion at the same temperatures. Consequently, an attainment of O isotopic equilibrium in the Grt–Omp pair suggests that Nd isotopic compositions are homogenised at metamorphic temperatures. Aberrant Sm–Nd isochron ages, or no ages at all, are expected to show disequilibrated oxygen isotope fractionation. This is illustrated by the Hong’an eclogites (Jahn & Liu, 2002).

V. Application of isotope constraints to tectonic evolution – example of the Dabie orogen This section is here to show how radiogenic isotopes can be used to constrain the tectonic evolution of the Dabie orogen. Some of the opinions, conclusions and implications reached in the followings are not necessarily shared by co-authors Rumble and Liou.

398

B.M. Jahn, D. Rumble & J.G. Liou

It is generally agreed that the Qinling–Dabie Orogen was produced by a Triassic collision between two Precambrian cratons in China. The traditional and unchallenged concept about the pre-collisional plate movement is that the Yangtze craton was subducted northward underneath the Sino-Korean craton, and the exhumed UHP metamorphic rocks and subjacent “basement” gneisses and derivative Cretaceous granitoids represent part of the Yangtze craton. At present, the overwhelming arguments (geological, structural, and oxygen isotopes) are in favour of northward subduction of the Yangtze craton. However, our Sr–Nd–Pb isotopic tracer analysis indicates that the Cretaceous granitoids and mafic–ultramafic rocks of Dabieshan have a very close affinity with the Sino-Korean craton, and are quite distinguished from the Yangtze craton. This may suggest an opposite polarity of the Triassic subduction, and has a drastic consequence for all tectonic models. Despite some consensus, the tectonic evolution of the Dabie orogen has been controversial. While most agree that the UHP terrane (= Southern Dabie Complex or SDC) represents a “thin” slice exhumed from great depths (= 100 km), the Northern Dabie Complex (NDC) has been variously considered as a migmatite terrane originally presented as the upper part of subducted plate (Maruyama et al., 1994; Ernst & Liou, 1995), as a high-temperature part of a vertical extrusion complex which includes UHP and HP rocks (Hacker et al., 1996), as a high-temperature metamorphic terrane formed in the hanging wall of the Sino-Korean craton, intruded by crustal melts and mafic–ultramafic rocks (Zhang et al., 1996a), as a magmatic complex created during the Cretaceous N–S extension (Hacker et al., 1998), as a migmatitic dome (= core complex) formed by decompressional partial melting of an exhuming slice, soon after the plate collision but not in the Cretaceous (Faure et al., 1999), or as an asymmetric Cordillerantype extensional complex (= core complex) formed in the Cretaceous (Hacker & Peacock, 1995; Ratschbacher et al., 2000). Because the present orogenic architecture of Dabieshan is dominated by late Mesozoic structures, it is important to determine the origin of the gneiss terranes of the NDC as well as the Cretaceous mafic and granitic intrusions. Mafic magmas provide information about the characteristics of their upper mantle source(s), whereas granitic magmas reveal their source characteristics at lower to middle crustal levels. Here we employ a geochemical and Sr–Nd–Pb isotopic tracer technique to characterise the “source rocks” beneath Dabieshan and unravel the possible genetic relationship between different terranes and lithologic units. The result of the isotope tracer analysis reveals a severe problem to the traditionally held northward subduction and tectonic evolution of the Dabie orogen (e.g. Jahn et al., 2001b). Lithological and geochemical characteristics of the NDC and SDC gneisses and Cretaceous intrusions The Northern Dabie Complex (NDC) consists dominantly of granitic gneisses and subordinate migmatite, amphibolite, garnet granulite, marble and some conspicuous trains of mafic–ultramafic rocks. The presence of mafic granulites at several localities suggests that the NDC may have reached the granulite facies metamorphism and was strongly overprinted by the amphibolite facies and a later thermal event when massive

Geochemistry and isotope tracer study of UHP metamorphic rocks

399

Cretaceous granites were emplaced. The granitic gneisses of the NDC cover a wide range of chemical compositions (SiO2 = 53–80%). Grey gneisses with TTG composition appear to dominate over gneisses of granitic composition. In the TAS diagram of Middlemost (1994), the NDC gneisses plot in the fields following the sub-alkaline trend of intrusive rocks. Quartzofeldspathic gneisses of the ultrahigh-pressure SDC have been separated into paragneiss and orthogneiss based on their mineral assemblages (e.g. Zhang et al., 1996b; Carswell et al., 1997). The paragneisses (Pl + Qtz + Phe + Bt + Ep ± Gt ± Ttn ± Rt) are intimately associated with eclogites, often interlayered with marble, phengitic schists and jadeitic quartzite, and have undergone UHP metamorphism. The orthogneisses (Pl + Kf + Qtz + Mica + Ep ± Grt ± Ttn ± Am) are granitic and trondhjemitic; they are distinguished from paragneisses by the absence of rutile and presence of K-feldspar. Zircon dating revealed that most protoliths of the gneisses were formed in the Late Proterozoic (ca. 700–800 Ma; Ames et al., 1996; Rowley et al., 1997; Hacker et al., 1998; Zheng et al., 2003; Li, 2003), but a few, such as those occurring at Shuanghe, have definitely much older Late Archaean ages (Chavagnac et al., 2001; Ayers et al., 2002). Together with granulites from the Luotian gneiss dome and trondhjemitic gneisses and metapelites of the Kongling group (zircon ages up to 3.2 Ga; Chen et al., 1996; Wu et al., 2002; Qiu et al., 2000), these rocks represent the only known relics of Archaean rocks in the Yangtze craton. The Cretaceous granitic intrusions encompass a wide spectrum of rock types including monzonite, quartz monzonite, syenite and granite. They plot in the midalkaline field in the TAS diagram, and form a roughly continuous trend with alkali gabbros of Dabieshan (Chen et al., 2002). The granitic intrusions, like mafic–ultramafic rocks, were emplaced indiscriminately in all tectonic subunits of Dabieshan. In the N Huaiyang belt, the intrusions are dominated by Sr, Ba and REE-rich syenitic magmas (Zhou et al., 1995a). Overall, the Cretaceous granitoids have geochemical compositions indicative of an origin from enriched sources. They are not crustal melts produced in a continental collision zone, but are liquids formed in an extensional setting with significant input of metasomatised mantle. In addition to granitic and syenitic intrusions, numerous small mafic/ultramafic bodies were also emplaced in Dabieshan during the Cretaceous, practically contemporaneously with the granitoids. These mantle-derived rocks (gabbro and pyroxenite) show an astonishing isotopic signature of the continental crust, which led Jahn et al. (1999) to propose a model of crust–mantle interaction and production of the mafic magmas by melting of a metasomatised mantle source. Isotope test of the existing tectonic models In all the published tectonic models, the Yangtze craton is assumed to subduct northwards beneath the Sino-Korean craton (e.g. Maruyama et al., 1994; Ernst & Liou, 1995; Hacker et al., 1996, 1998; Faure et al., 1999). The exhumed blocks of UHP and HP units are thin slices underlain by unknown “basement” rocks. Cretaceous doming in the NDC (Hacker et al., 1998, 2000; Faure et al., 1999; Ratschbacher et al., 2000)

400

B.M. Jahn, D. Rumble & J.G. Liou

stretched and separated the thin slice of the UHP unit and resulted in two disproportional entities, now represented in majority by the UHP terrane of the SDC and a tiny part in the north of the NDC (Xu et al., 1999). Recently, microdiamond inclusions in garnet have been identified in eclogites from the NDC, thus concluding the UHP metamorphic history of the eclogite-bearing unit in the NDC (Xu et al., 2003). The Cretaceous doming in the NDC was accompanied by intense magmatism, resulting in voluminous granitoid intrusions and ubiquitous mafic/ultramafic stocks and dikes. Note that the magmatism took place within all the tectonic units of Dabieshan. The northward subduction models predict that the Cretaceous magmatism would show isotopic signature of the lower crust and upper mantle (= lithosphere) of the Yangtze craton. The thin slice of UHP/HP metamorphic rocks could have been part of a microcontinent (= Neoproterozoic island arc) lying between the two cratons, so whether it belonged to the Yangtze craton is debatable. In any case, the northward subduction models can be tested by the isotope tracers of the post-collisional Cretaceous magmatic rocks that were formed by melting of the subducted-and-exhumed lithosphere. The Sr–Nd–Pb isotopic data used in the present analysis are from our own work and the literature (references too many to cite individually; they are available upon solicitation). The key points are given as follows. (1) Nd–Sr isotopic compositions of granitic gneisses (Figs. 16a, b). At 220 Ma (Fig. 16a), all granitic gneisses have evolved into negative HNd values. However, the data sets for the UHP (SDC) and high-T terranes (NDC) of Dabieshan are easily distinguished, with HNd(T) values of –2 to –13 (except one) for SDC and –11 to –23 (except Huangtuling) for NDC. The SDC gneisses have more dispersed 87Sr/86Sr ratios (= ISr), up to 0.740, whereas the NDC gneisses are more restricted in the range of 0.706 to 0.715. The gneisses at Shuanghe locality are evidently extraordinary in all aspects with regard to other gneisses of the SDC. They have the lowest HNd(T) values, the highest TDM model ages, and the oldest zircon ages (Late Archaean) among the SDC gneisses. A granulite from Huangtuling village, near Yinshan, has a very low HNd(T) value of –30 and very high 87 Sr/86Sr of 0.742. Together with some known Archaean zircon ages for granulites from the same region, it indicates the presence of late Archaean relics in Dabieshan. The granitic gneisses of the Su-Lu UHP terrane seem to differ from those of the Dabie UHP terrane (SDC) in terms of their lower range of HNd(T) values, but are similar in the wide range of Sr isotope compositions. At 120 Ma (Fig. 16b) Cretaceous mafic/ultramafic rocks were derived from the upper mantle. Their restricted HNd(T) of –15 to –20 and ISr (0.706–0.710) suggest a metasomatised mantle presumably produced by interaction of subducted lower crust with mantle peridotites (Jahn et al., 1999). However, this interpretation is not unique as we will discuss later. The coeval but much more voluminous granitic intrusions are shown to have isotopic compositions remarkably similar to both the mafic rocks and NDC granitic gneisses. The Cretaceous granitoids were emplaced in all the tectonic subunits of Dabieshan (North Huaiyang, NDC, SDC, and Susong Blueschist terrane), yet their isotopic compositions are conspicuously uniform. (2) Sm–Nd isotopic and chemical characteristics (Figs. 17a, b). While the HNd(T) value represents time-integrated isotopic evolution of a rock, 147Sm/144Nd ratio is a

Geochemistry and isotope tracer study of UHP metamorphic rocks

401

Fig. 16. (a) HNd(T) vs. 87Sr/86Sr ratios, recalculated at 220 Ma, for the granitic gneisses from the Southern and Northern Dabie Complex (SDC, NDC), and the Su-Lu terrane. Shuanghe gneisses and Huangtuling granulite are exceptions. (b) HNd(T) vs. 87Sr/86Sr ratios, recalculated at 120 Ma, for the granitic gneisses and Cretaceous mafic/ultramafic and granitic rocks. Note the similarity between the Cretaceous mafic/ultramafic rocks, Cretaceous granites/syenites, and the Neoproterozoic NDC gneisses.

402

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 17. (a) HNd(T) vs. 147Sm/144Nd, recalculated at 220 Ma, for the granitic gneisses from the Southern and Northern Dabie Complexes (SDC, NDC) and the Su-Lu terrane. The NDC and SDC gneisses are separated at HNd(T) = –13, except for Shuanghe and Huangtuling. (b) HNd(T) vs. 147Sm/144Nd, recalculated at 120 Ma, for the granitic gneisses and Cretaceous mafic/ultramafic and granitic rocks. Note the similarity between the Cretaceous mafic/ultramafic rocks, Cretaceous granites/syenites, and the Neoproterozoic NDC gneisses.

Geochemistry and isotope tracer study of UHP metamorphic rocks

403

parameter characteristic of its chemical composition. The majority of the SDC granitic gneisses have Sm/Nd ratios higher than the average upper continental crust (UCC) value, whereas the opposite is true for the NDC gneisses. This indicates that NDC gneisses have more fractionated REE and greater enrichment in LREE. The Su-Lu gneisses have intermediate characteristics, showing data points straddling the SDC and NDC fields (Fig. 17a). At 120 Ma, the Cretaceous granitic magmas are shown to have both Nd isotopic and REE patterns similar to the NDC gneisses, and clearly distinguished from the SDC gneisses. (3) Depleted mantle-based model ages (TDM, Figs. 18a, b). The SDC gneisses have TDM ranging from 1.1 to 2.4 Ga, except two with high fSm/Nd values and those from the Shuanghe locality (Fig. 18a). On the other hand, the NDC gneisses have older TDM from 1.5 to 3.0 Ga, including two granulites from Huangtuling. The Cretaceous granitoids and mafic intrusions also have a range of TDM similar to the NDC gneisses. Fig. 18b shows the Nd isotopic distinction between the Sino-Korean Craton (SKC) and the South China Block (Yangtze Craton and Cathaysia included) based on Sm–Nd isotope data of Mesozoic granitoids and felsic volcanic rocks. The two blocks are roughly separated at HNd = –13. The field of Dabieshan rocks overlap much of the granitoids from the SKC, but have little in common with those from the South China Block. (4) Pb isotopic compositions (Figs. 19a, b). Based on data from the literature (e.g. compilations of Zhang, 1995; Chen & Jahn, 1999), the Pb isotope fields of Mesozoic granitoids from the SKC are contrasted with that from the Yangtze craton. The data of Mesozoic granitoids from Dabieshan and Su-Lu have Pb isotopic compositions similar to those of the SKC, but very different from the Yangtze craton. Granitic gneisses of Dabieshan at 120 Ma show the same isotopic characteristics. Discussion and tectonic implications The geochemical and isotopic constraints to the tectonic evolution of the entire Dabie orogen may be summarised below: (1) The Cretaceous mafic–ultramafic magmas in Dabieshan were derived from a metasomatised mantle source with EM1 signature. Their isotopic characteristics are similar to the coeval mafic rocks from the SKC (e.g. Qiu et al., 1997). (2) The Cretaceous granitoids (felsic volcanics included) have Sr–Nd isotopic and elemental characteristics similar to the mafic–ultramafic rocks, suggesting their possible genetic relationship. The isotopic signature is “typical” of the lower continental crust of moderately depleted granulite facies rocks. This is also supported by Pb isotopic data (Zhang, 1995; Zhou et al., 1995b; Chen & Jahn, 1999; Zhang et al., 2002). (3) Recent isotopic investigations on Cretaceous mafic and felsic volcanic rocks from western Shandong, eastern Shandong (Jiaodong peninsula), Xishan and Nankou (near Beijing) reveal that these rocks also have very similar isotopic signature as the contemporaneous intrusions of Dabieshan. The similar isotopic signature is also found in Cretaceous granites of the Taihang Mountains (Chen et al., 2003). (4) The quartzofeldspathic gneisses of the NDC and SDC are distinguished both in isotopic and trace element compositions. Sr–Nd isotopic data suggest that the SDC

404

B.M. Jahn, D. Rumble & J.G. Liou

Fig. 18. (a) HNd(T) vs. model age TDM for the granitic gneisses from the Southern and Northern Dabie Complexes (SDC, NDC), and the Su-Lu terrane. (b) HNd(T) vs. model age TDM for Mesozoic granitic rocks from the SKC and South China Block (Yangtze craton & Cathaysia). The two blocks can be separated at HNd = +13.

Geochemistry and isotope tracer study of UHP metamorphic rocks

405

(a)

(b)

Fig. 19. (a) 207Pb/204Pb vs. 206Pb/204Pb plot for Cretaceous granitoids and granitic gneisses of Dabieshan, and Mesozoic granitoids of the Su-Lu terrane. The distinction of isotope fields for the Yangtze and Sino-Korean cratons is clear. Most Cretaceous granitoids and gneisses (calculated at 120 Ma) show their close affinity with the Sino-Korean craton, and completely distinguished from the Yangtze craton. (b) 208Pb/204Pb vs. 206Pb/204Pb plot for Cretaceous granitoids and granitic gneisses of Dabieshan, and Mesozoic granitoids of the Su-Lu terrane.

406

B.M. Jahn, D. Rumble & J.G. Liou

gneisses have no role in the genesis of post-orogenic granitoids, but the data allow the NDC gneisses to be a likely source for Cretaceous granitoids. (5) A core complex interpretation for the NDC (e.g. Faure et al., 1999; Ratschbacher et al., 2000) is supported by the geochemical and isotopic data, as well as by the occurrence of Triassic eclogites in the northern margin of the NDC (Xu et al., 1999). The grossly similar Nd–Sr–Pb isotopic signature between Cretaceous mafic–ultramafic rocks, Cretaceous granitic rocks and the NDC gneisses is intriguing. The derivation of mafic and ultramafic rocks from a metasomatised mantle is beyond any doubt. However, it is not clear how the metasomatism was effected. Jahn et al. (1999) proposed that the process occurred in the post-collisional epoch, and between the hot asthenosphere and “trapped” lower continental crust after most subducted continental slices were uplifted. Implicitly, the lower crust dominated the Sr–Nd isotope budget for both the mafic liquids produced from the mantle and the granitic rocks formed by the melting of the lower crust. While this interpretation may be suitable for a terrane with deep subduction of continental blocks, it becomes difficult to explain the similar isotopic signature observed for other contemporaneous rocks emplaced in other parts of the Sino-Korean craton (SKC). Consequently, the mantle metasomatism cannot be regarded as a local Dabieshan phenomenon; it is of a cratonic scale. Moreover, a comparison of chemical compositions between the NDC gneisses and Cretaceous granitoids poses another problem. Petrological experiments constrain that melting of felsic gneisses at lower crustal conditions would produce liquids of silica-rich granitic composition, and in no way syenitic liquids could be formed (Litvinovsky et al., 2000). The occurrence of syenitic rocks in the North Huaiyang belt probably indicate an important additional input of the mantle component (Litvinovsky et al., 2001). Besides, high Sr–Ba concentrations in Cretaceous granitoids cannot be explained by melting of ordinary felsic gneisses. Consequently, in addition to the NDC gneisses, input from metasomatised mantle is required for the production of Cretaceous granitoids. If the lower crust and the upper mantle sources of Cretaceous magmatic rocks in Dabieshan are identical to that of the SKC, then it implies that the continental lithosphere underneath Dabieshan is the same as that underneath the SKC. Consequently, the traditional view that the Yangtze craton subducted northwards beneath the SKC may no longer be acceptable. The subduction polarity could have been just opposite. At present, we are working on an appropriate tectonic model to explain the isotopic features. Nevertheless, a recent detailed structural and deformation analyses on UHP metamorphic rocks at Yangkou in the Su-Lu terrane lend a strong argument for such a reversed subduction polarity (Zhao et al., 2002).

Acknowledgements Constructive comments and suggestions on an earlier version by Yong-Fei Zheng (Hefei, China) have greatly improved the final version of this article. The editorial assistance of Tamás Váczi is much appreciated. We thank the European Mineralogical Union for invitation to present this work in the 5th EMU School and Symposium on Ultrahigh Pressure

Geochemistry and isotope tracer study of UHP metamorphic rocks

407

Metamorphism held in Budapest, Hungary (21–25 July 2003). The preparation of this review was supported in part by the National Science Council (NSC) of Taiwan (Jahn), and the U.S. National Science Foundation, EAR Continental Dynamics Program, EAR0003276 (Rumble) and EAR-0003355 (Liou).

References Ames, L., Zhou, G. & Xiong, B. (1996): Geochronology and isotopic character of ultrahigh-pressure metamorphism with implications for collision of the Sino-Korean and Yangtze cratons, central China. Tectonics, 15:472–489. Ayers, J.C., Dunkle, S., Gao, S. & Miller, C.F. (2002): Constraints on timing of peak and retrograde metamorphism in the Dabie Shan ultrahigh-pressure metamorphic belt, east-central China, using U-ThPb dating of zircon and monazite. Chem. Geol., 186:315–331. Baker, J., Matthews, A., Mattey, D., Rowley, D.B. & Xue, F. (1997): Fluid-rock interaction during ultra-high pressure metamorphism, Dabie Shan, China. Geochim. Cosmochim. Acta, 61:1685–1696. Barnicoat, A.C. & Cartwright, I. (1997): Focused fluid flow during subduction: oxygen isotope data from high pressure ophiolites of the western Alps. Earth Planet. Sci. Lett., 132:53–61. Beard, B.L., Medaris, L.G., Johnson, C.M., Brueckner, H.K. & Misar, Z. (1992): Petrogenesis of Variscan high-temperature Group A eclogites from the Moldanubian zone of the Bohemian Massif, Czechoslovakia. Contrib. Mineral. Petrol., 111:468–483. Beard, B.L., Medaris, L.G., Johnson, C.M., Jelinek, E., Tonika, J. & Riciputi, L.R. (1995): Geochronology and geochemistry of eclogites from the Mariánské Lázně Complex, Czech Republic: Implications for Variscan orogenesis. Geol. Rundsch., 84:552–567. Becker, H. (1993): Garnet peridotite and eclogite Sm–Nd mineral ages from the Lepontine dome (Swiss Alps): New evidence for Eocene high-pressure metamorphism in the central Alps. Geology, 21:599–602. Bernard-Griffiths, J. & Cornichet, J. (1985): Origin of eclogites from South Brittany, France: A Sm–Nd isotopic and REE study. Chem. Geol., 52:185–201. Bernard-Griffiths, J., Peucat, J.J., Cornichet, J., Ponce de Leon, M.I. & Ibarguchi, J.I.G. (1985): U–Pb, Nd isotopic and REE geochemistry in eclogites from the Cabo Ortegal Complex, Galicia, Spain: an example of REE immobility conserving MORB-like patterns during high-grade metamorphism. Chem. Geol., 52:217–225. Blattner, P., Grindley, G.W. & Adams, C.J. (1997): Low 18O terranes tracking Mesozoic polar climates in the South Pacific. Geochim. Cosmochim. Acta, 61:569–576. Blichert-Toft, J., Chauvel, C. & Albarede, F. (1997): Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multicollector ICP-MS. Contrib. Mineral. Petrol., 127:248–260. Bocchio, R., de Capitan, L., Ottolini, L. & Cella, F. (2000): Trace element distribution in eclogites and their clinopyroxene/garnet pair: A case study from Soazza (Switzerland). Eur. J. Mineral., 12:147–161. Burton, K.W., Kohn, M.G., Cohen, A.S. & O’Nions, R.K. (1995): The relative diffusion of Pb, Nd, Sr and O in garnet. Earth Planet. Sci. Lett., 133:199–211. Carswell, D.A. & Cuthbert, S.J. (2003): Ultrahigh pressure metamorphism in the Western Gneiss Region of Norway. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 51–73. Carswell, D.A., O’Brien, P.J., Wilson, R.N. & Zhai, M. (1997): Thermobarometry of phengite-bearing eclogites in the Dabie Mountains of central China. J. Metamorph. Geol., 15:239–252. Chavagnac, V. & Jahn, B.M. (1996): Coesite-bearing eclogites from the Bixiling Complex, Dabie Mtns., China: Sm–Nd ages, geochemical characteristics, and tectonic implications. Chem. Geol., 133:29–51. Chavagnac, V., Jahn, B.M., Villa, I., Whitehouse, M.J. & Liu, D. (2001): Multichronometric evidence for an in-situ origin of the ultrahigh-pressure metamorphic terrane of Dabieshan, China. J. Geol., 109:633–646. Chen, J.F. & Jahn, B.M. (1999): Nd-Sr-Pb isotope tracer study and crustal evolution of southeastern China. In Zheng, Y.F. (ed.): Chemical geodynamics. Beijing: Science Publisher, 262–287 (in Chinese). Chen, N.-S., You, Z.-D., Suo, S.-T., Yang, Y. & Li, H.-M. (1996): Zircon U–Pb ages of intermediate granulite and deformed granites in the Dabie Mountains, central China. Chinese Sci. Bull., 41:1886–1890.

408

B.M. Jahn, D. Rumble & J.G. Liou

Chen, B., Jahn, B.M. & Wei, C. (2002): Petrogenesis of Mesozoic granitoids in the Dabie UHP complex, Central China: Trace element and Nd-Sr isotope evidence. Lithos, 60:67–88. Chen, B., Jahn, B.M. & Zhai, M.G. (2003): Sr-Nd isotopic characteristics of the Mesozoic magmatism in the Taihang-Yanshan orogen, north China craton, and implications for Archean lithosphere thinning. J. Geol. Soc. London, in review. Cheng, Y.Q., Liu, D.Y., Williams, I.S., Jian, P., Zhuang, Y.X. & Gao, T.S. (2000): SHRIMP U–Pb dating of zircons of a dark-coloured eclogite and a garnet-bearing gneissic-granitic rock from Bixiling, eastern Dabie area. Acta Geol. Sin., 74:193–205. (in Chinese with English abstract). Cong, B. (ed.) (1996): Ultrahigh-pressure metamorphic rocks in the Dabieshan-Sulu region of China. Dordrecht: Kluwer, 223 p. Cong, B., Zhai, M., Carswell, D.A., Wilson, R.N., Wang, Q., Zhao, Z. & Windley, B.F. (1995): Petrogenesis of UHP rocks and their country rocks at Shuanghe in Dabie Shan, central China. Eur. J. Mineral., 7:119–138. Degeling, H., Eggins, S. & Ellis, D.J. (2001): Zr budgets for metamorphic reactions and the formation of zircon from garnet breakdown. Mineral. Mag., 65:749–758. de Sigoyer, J., Chavagnac, V., Blichert-Toft, J., Villa, I., Luais, B., Guillot, S., Cosca, M. & Mascle, G. (2000): Dating the Indian continental subduction and collisional thickening in the northwest Himalaya: multichronology of the Tso Morari eclogites. Geology, 28:487–490. Dobrzhinetskaya, L.F., Green, H.W. & Wang, S. (1996): Alpe Arami: a peridotite massif from depths more than 300 km. Science, 271:1841–1845. Dodson, M.H. (1973): Closure temperature in cooling geochronological and petrological systems. Contrib. Mineral. Petrol., 40:259–274. Dong, S.W., Oberhänsli, R., Schmid, R., Liu, X.C., Tang, J.F. & Xue, H.M., (2002): Occurrence of metastable rocks in deeply subducted continental crust from the Dabie Mountains, central China. Episodes, 25:84–89. Duchêne, S., Blichert-Toft, J., Luais, B., Télouk, P., Lardeaux, J.M. & Albarède, F. (1997): The Lu-Hf dating of garnets and the ages of the Alpine high-pressure metamorphism. Nature, 387:586–589. Eide E.A. & Liou, J.G. (2000): High-pressure blueschists and eclogites in Hong’an: a framework for addressing the evolution of high- and ultrahigh-pressure rocks in central China. Lithos, 52:1–22. Eide, E., Liou, J.G. & McWilliams, M.O. (1994): 40Ar/39Ar constraints on high and ultrahigh pressure metamorphism, Hubei Province, China. Geology, 22:601–604. Ernst, W.G. & Liou, J.G. (1995): Contrasting plate-tectonic styles of the Qinling-Dabie-Sulu and Franciscan metamorphic belts. Geology, 23:353–356. Fan, Q., Liu, R., Ma, B., Zhao, D. & Zhang, Q. (1996): Protolith and ultrahigh-pressure metamorphism of Maowu mafic – ultramafic rocks, Dabieshan. Acta Petrol. Sin., 12:29–47. Faure, M., Lin, W., Shu, L., Sun, Y. & Schärer, U. (1999): Tectonics of the Dabieshan (eastern China) and possible exhumation mechanism of ultrahigh-pressure rocks. Terra Nova, 11:251–258. Fruh-Green, G.L. (1994): Interdependence of deformation, fluid infiltration, and reaction progress recorded in eclogitic metagranitoids (Sesia Zone, western Alps). J. Metamorph. Geol., 12:327–343. Fu, B., Zheng, Y.-F., Wang, Z., Xiao, Y., Gong, B. & Li, S. (1999): Oxygen and hydrogen isotope geochemistry of gneisses associated with ultrahigh pressure eclogites at Shuanghe in the Dabie Mountains. Contrib. Mineral. Petrol., 134:52–66. Fu, B., Touret, J.L.R. & Zheng, Y.-F. (2001): Fluid inclusions in coesite-bearing eclogites and jadeite quartzite at Shuanghe, Dabie Shan, China. J. Metamorph. Geol., 19:529–545. Fu, B., Zheng, Y.-F. & Touret, J.L.R. (2002): Petrological, isotopic and fluid inclusion studies of eclogites from Sujiahe, NW Dabie Shan (China). Chem. Geol., 187:107–128. Gebauer, D. (1996): A P-T-t path for an (ultra?-)high pressure ultramafic/mafic rock association and its felsic country rocks based on SHRIMP dating of magmatic and metamorphic zircon domains. Example: Alpe Arami (Central Swiss Alps). In Basu, A. & Hart, S. (eds.): Earth processes: Reading the isotopic code. /Geophys. Monogr., 95/. Washington (D.C.): Am. Geophys. Union, 307–329. Getty, S.R. & Selverstone, J. (1994): Stable isotope and trace element evidence for restricted fluid migration in 2 GPa eclogites. J. Metamorph. Geol., 12:747–760. Griffin, W.L. & Brueckner, H.K. (1980): Caledonian Sm–Nd ages and a crustal origin for Norwegian eclogites. Nature, 285:319–321.

Geochemistry and isotope tracer study of UHP metamorphic rocks

409

Griffin, W.L. & Brueckner, H.K. (1985): REE, Rb–Sr and Sm–Nd studies of Norwegian eclogites. Chem. Geol., 52:249–271. Günther, M. & Jagoutz, E. (1997): Isotope disequilibria (Sm/Nd, Rb/Sr) between minerals of coarse grained, low temperature garnet peridotites from Kimberley floors, Southern Africa. In Meyer, H.O.A. & Leonardos, O.H. (eds.): Proc. 5th Int. Kimberlite Conf. /CPRM Spec. Publ., 1A/. Brasilia: Serv. Geol. Bras., 354–365. Hacker, B. R. & Peacock, S. M. (1995): Creation, preservation, and exhumation of UHPM rocks. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 159–181. Hacker, B.R., Wang, X., Eide, E.A. & Ratschbacher, L. (1996): The Qinling-Dabie ultrahigh-pressure collisional orogen. In Yin, A. & Harrison, T.M. (eds.): Tectonic evolution of Asia. Cambridge: Cambridge Univ. Press, 345–370. Hacker, B.R., Ratschbacher, L., Webb, L., Ireland, T., Walker, D. & Shuwen, D. (1998): U/Pb zircon ages constrain the architecture of the ultrahigh-pressure Qinling-Dabie Orogen, China. Earth Planet. Sci. Lett., 161:215–230. Hacker, B.R., Ratschbacher, L., Webb, L., McWilliams, M.O., Ireland, T., Calvert, A., Shuwen, D., Wenk, H.R. & Chateigner, D. (2000): Exhumation of ultrahigh-pressure continental crust in East-central China: Late Triassic–Early Jurassic tectonic unroofing. J. Geophys. Res., 105:13,339–13,364. Harte, B. & Kirkley, M.B. (1997): Partitioning of trace elements between clinopyroxene and garnet: data from mantle eclogites. Chem. Geol., 136:1–24. Hirajima, T. & Nakamura, D. (2003): The Dabie Shan-Sulu orogen. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 105–144. Hirajima, T., Ishwatari, A., Cong, B., Zhang, R.Y., Banno, S. & Nozaka, T. (1990): Coesite from Mengzhong eclogite at Donghai county, northeastern Jiansu Province, China. Mineral. Mag., 54:579–583. Hirajima, T., Zhang, R., Li, J. & Cong, B. (1992): Petrology of the nyboite bearing eclogite in the Donghai area, Jiangsu province, E. China. Mineral. Mag., 56:37–46. Hiramatsu, N. & Hirajima, T. (1995): Petrology of the Hujialin garnet clinopyroxenite in the Su-Lu ultrahigh pressure province, eastern China. Isl. Arc, 4:310–323. Hoskin, P.W.O. & Black, L.P. (2000): Metamorphic zircon formation by solid-state recrystallization of protolith igneous zircon. J. Metamorph. Geol., 18:423–439. Humphries, F.J. & Cliff, R.A. (1982): Sm–Nd dating and cooling history of Scourian granulites, Sutherland. Nature, 295:515–517. Jagoutz, E. (1988): Nd and Sr systematics in an eclogite xenolith from Tanzania: Evidence for frozen mineral equilibria in the continental lithosphere. Geochim. Cosmochim. Acta, 52:1285–1293. Jahn, B.M. (1998): Geochemical and isotopic characteristics of UHP eclogites and ultramafic rocks of the Dabie orogen: Implications for continental subduction and collisional tectonics. In Hacker, B. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 203–239. Jahn, B.M. (1999): Sm–Nd isotope tracer study of UHP metamorphic rocks: implications for continental subduction and collisional tectonics. Int. Geol. Rev., 41:859–885. Jahn, B.M. & Liu, X.C. (2002): Age(s) of the Hong’an Block and the question of 400 Ma UHP metamorphic event. In 3rd UHPM Workshop, Beijing, Abstr. Vol., 68–71. Jahn, B.M., Cornichet, J., Cong, B. & Yui, T.F. (1996): Ultrahigh HNd eclogites from an UHP metamorphic terrane of China. Chem. Geol., 127:61–79. Jahn, B.M., Wu, F., Lo, C.H. & Tsai, C.H. (1999): Crust-mantle interaction induced by deep subduction of the continental crust: geochemical and Sr-Nd isotopic evidence from post-collisional mafic-ultramafic intrusions of the northern Dabie complex, central China. Chem. Geol., 157:119–146. Jahn, B.M., Caby, R. & Monie, P. (2001a) The oldest UHP eclogites of the World: age of UHP metamorphism, nature of protoliths and tectonic implications. Chem. Geol., 178:143–158. Jahn, B.M., Chen, B., Li, H.Y. & Potel, S. (2001b): Continental subduction and mantle metasomatism: consequence on the Cretaceous magmatism and implications for the architecture of the Dabie Orogen. In Ogasawara, Y., Maruyama, S. & Liou, J.G. (eds.): UHPM Workshop 2001 at Waseda University. Tokyo: Waseda Univ., 147–154. Jahn, B.M., Fan, Q.C., Yang, J.J. & Henin, O. (2003): Petrogenesis of the Maowu pyroxenite-eclogite body from the UHP metamorphic terrane of Dabieshan: chemical and isotopic constraints. Lithos, in press.

410

B.M. Jahn, D. Rumble & J.G. Liou

Jamtveit, B., Carswell, D.A. & Mearns, E.W., (1991): Chronology of the high-pressure metamorphism of Norwegian garnet peridotites/pyroxenes. J. Metamorph. Geol., 9:125–139. Jerde, E.A., Taylor, L.A., Crozaz, G., Sobolev, N.V. & Sobolev, V.N. (1993): Diamondiferous eclogites from Yakutia, Siberia: evidence for a diversity of protoliths. Contrib. Mineral. Petrol., 114:189–202. Johnson, K.T.M. (1994): Experimental cpx/and garnet/melt partitioning of REE and other trace elements at high pressures: petrogenetic implications. Mineral. Mag., 58A:454–455. Kalt, A., Hanel, H., Schleicher, H. & Kramm, U. (1994): Petrology and geochronology of eclogites from the Variscan Schwarzwald (FRG). Contrib. Mineral. Petrol., 115:287–302. Li, H. (2003): Crustal evolution of the UHP metamorphic terranes in the Dabie Orogen. Thesis, Université de Rennes 1, France. Li, S., Xiao, Y., Liou, D., Chen, Y., Ge, N., Zhang, Z., Sun, S., Cong, B., Zhang, R., Hart, S.R. & Wang, S. (1993): Collision of the N China and Yangtze Blocks and formation of coesite-bearing eclogite: Timing and processes. Chem. Geol., 109:89–111. Li, S., Wang, S., Chen, Y., Liu, D., Qiu, J., Zhou, H. & Zhang, Z. (1994): Excess Ar in phengite from eclogite: Evidence from dating of eclogite minerals by Sm–Nd, Rb–Sr, and 40Ar/39Ar methods. Chem. Geol., 112:343–350. Li, S.-G., Jagoutz, E., Chen, Y.-Z. & Li, Q.-L. (2000): Sm–Nd and Rb–Sr isotopic chronology and cooling history of ultrahigh pressure metamorphic rocks and their country rocks at Shuanghe in the Dabie Mountains, Central China. Geochim. Cosmochim. Acta, 64:1077–1093. Li, Y.-L., Zheng, Y.-F., Fu, B., Zhou, J.-B. & Wei, C.-S. (2001): Oxygen isotope composition of quartz-vein in ultrahigh-pressure eclogite from Dabieshan and implications for transport of high-pressure metamorphic fluid. Phys. Chem. Earth (A), 26:695–704. Liati, A., Gebauer, D. & Fanning, M. (2000): U–Pb SHRIMP dating of zircon from the Novate granite (Bergell, Central Alps): evidence for Oligocene-Miocene magmatism, Jurassic/Cretaceous continental rifting and opening of the Valais trough. Schweiz. Mineral. Petrogr. Mitt., 80:305–316. Liou, J.G. & Zhang, R.Y. (1996): Occurrences of intergranular coesite in ultrahigh-P rocks from the SuLu region, eastern China: Implications for lack of fluid during exhumation. Am. Mineral., 81:1217–1221. Liou, J.G. & Zhang, R.Y. (1998): Petrogenesis of an ultrahigh-pressure garnet-bearing ultramafic body from Maowu, Dabie Mountains, east-central China. Isl. Arc, 7:115–134. Liou, J.G., Zhang, R.Y., Wang, X., Eide, E.A., Ernst, W.G. & Maruyama, S. (1996): Metamorphism and tectonics of high-pressure and ultra-high-pressure belts in the Dabie–Sulu region, China. In Yin, A. & Harrison, T.M. (eds.): Tectonic evolution of Asia. Cambridge: Cambridge Univ. Press, 300–344. Liou, J.G., Zhang, R.Y. & Jahn, B.M. (1997): Petrology, geochemistry and isotope data on a ultrahigh-pressure jadeite quartzite from Shuanghe, Dabie Mountains, east-central China. Lithos, 41:59–78. Liou, J.G., Zhang, R.Y., Ernst, W.G., Rumble, D. & Maruyama, S. (1998): High-pressure minerals from deeply subducted metamorphic rocks. In Hemley, R. & Mao, D. (eds.): Ultrahigh pressure mineralogy /Rev. Mineral., 37/. Washington (D.C.): Mineral. Soc. Am., 33–96. Litvinovsky, B.A., Steele, I.M. & Wickham, S.M. (2000): Silicic magma formation in overthickened crust: melting of charnockite and leucogranite at 15, 20 and 25 kbar. J. Petrol., 41:717–737. Litvinovsky, B.A., Jahn, B.M., Zanvilevich, A.N., Saunders, A., Poulain, S., Kuzmin, D.V. & Titov, A.V. (2001): Petrogenesis of syenite-granite suites from the Bryansky complex (Transbaikalia, Russia): implications for the origin of A-type granitoids magmas. Chem. Geol., 189:105–133. Luais, B., Duchêne, S. & de Sigoyer, J. (2001): Sm–Nd disequilibrium in high-pressure, low-temperature Himalayan and Alpine rocks. Tectonophysics, 342:1–22. Masago, H., Rumble, D., Ernst, W.G., Parkinson, C.D. & Maruyama, S. (2003): Low G18O eclogites from the Kokchetav massif, northern Kazakhstan. J. Metamorph. Geol., 21:579–587. Maruyama, S., Liou, J.G. & Zhang, R. (1994): Tectonic evolution of the ultrahigh-pressure (UHP) and highpressure (HP) metamorphic belts from central China. Isl. Arc, 3:112–121. Mearns, E.W. (1986): Sm–Nd ages for Norwegian garnet peridotite. Lithos. 19:269–278. Medaris, L.G., Jr., Beard, B.L., Johnson, C.M., Valley, J.W., Spicuzza, M.J., Jelinek, E. & Misar, Z. (1995): Garnet pyroxenite and eclogite in the Bohemian Massif: geochemical evidence for Variscan recycling of subducted lithosphere. Geol. Rundsch., 84:489–505.

Geochemistry and isotope tracer study of UHP metamorphic rocks

411

Mezger, K., Essene, E.J. & Halliday, A.N. (1992): Closure temperatures of the Sm–Nd system in metamorphic garnets. Earth Planet. Sci. Lett., 113:397–409. Middlemost, E.A.K. (1994): Naming materials in the magma/igneous rock system. Earth Sci. Rev., 37:215–224. Miller, C. & Thöni, M. (1995): Origin of eclogites from the Austroalpine Ötztal basement (Tirol, Austria): geochemistry and Sm-Nd vs. Rb-Sr isotope systematics. Chem. Geol., 122:199–225. Miller, J.A., Cartwright, I., Buick, I.S. & Barnicoat, A.C. (2001): An O-isotope profile through the HP-LT Corsican ophiolite, France, and its implications for fluid flow during subduction. Chem. Geol., 178:43–69. Mørk, M.B.E. & Mearns, E.W. (1986): Sm–Nd isotopic systematics of a gabbro-eclogite transition. Lithos, 19:255–267. Nadeau, S., Philippot, P. & Pineau, F. (1993): Fluid inclusion and mineral isotopic compositions (H-C-O) in eclogite rocks as tracers of local fluid migration during high pressure metamorphism. Earth Planet. Sci. Lett., 114:431–448. Oberhänsli, R., Martinotti, G., Schmid, R. & Liu, X. (2002): Preservation of primary volcanic textures in the ultrahigh-pressure terrain of Dabie Shan. Geology, 30:699–702. Okay, A.I. (1993): Petrology of a diamond and coesite-bearing metamorphic terrain, Dabie Shan, China. Eur. J. Mineral., 5:659–675. Okay, A.I. (1994): Sapphirine and Ti-clinohumite in ultra-high-pressure garnet-pyroxenite and eclogite from Dabie Shan, China. Contrib. Mineral. Petrol., 110:1–13. O’Reilly, S.Y. & Griffin, W.L. (1995): Trace-element partitioning between garnet and clinopyroxene in mantlederived pyroxenites and eclogites: P-T-X controls. Chem. Geol., 121:105–130. Paquette, J.L., Menot, R.P. & Peucat, J.J. (1989): REE, Sm–Nd and U–Pb zircon study of eclogites from the Alpine External Massif (western Alps): evidence for crustal contamination. Earth Planet. Sci. Lett., 96:181–198. Philippot, P. & Rumble, D. (2000): Fluid-rock interactions during high-pressure and ultrahigh-pressure metamorphism. Int. Geol. Rev., 42:312–327. Philippot, P. & Selverstone, J. (1991): Trace-element brines in eclogitic veins: implications for fluid composition and transport during subduction. Contrib. Mineral. Petrol., 106:417–430. Philippot, P., Agrinier, P. & Scambelluri, M. (1998): Chlorine cycling during subduction of altered oceanic crust. Earth Planet. Sci. Lett., 161:33–44. Pidgeon, R.T. (1992): Recrystallisation of oscillatory zoned zircons: some geochronological and petrological implications. Contrib. Mineral. Petrol., 110:463–472. Pidgeon, R.T., Nemchin, A.A. & Hitchen, G.J. (1998): Internal structures of zircons from Archaean granites from the Darling Range batholith: implications for zircon stability and the interpretation of zircon U–Pb ages. Contrib. Mineral. Petrol. 132:288–299. Qiu, J.S., Wang, D.Z., Zeng, J.H. & McInnes, B.I.A. (1997): Study on trace element and Nd-Sr isotopic geochemistry of the Mesozoic K-rich volcanic rocks and lamprophyres in western Shandong. Geol. J. China Univ., 3:384–395. Qiu, Y.M., Gao, S., McNaughton, N.J., Groves, D.I. & Ling, W. (2000): First evidence of >3.2 Ga continental crust in the Yangtze craton of south China and its implications for Archean crustal evolution and Phanerozoic tectonics. Geology, 28:11–14. Ratschbacher, L., Hacker, B.R., Webb, L.E., McWilliams, M., Ireland, T., Dong, S.W., Calvert, A., Chateigner, D. & Wenk, H.R. (2000): Exhumation of the ultrahigh-pressure continental crust in east central China: Cretaceous and Cenozoic unroofing and the Tan-Lu fault. J. Geophys. Res., 105:13,303–13,338. Rowley, D.B., Xue, F., Tucker, R.D., Peng, Z.X., Baker, J. & Davis, A. (1997): Ages of ultrahigh pressure metamorphism and protolith orthogneisses from the eastern Dabie Shan: U/Pb zircon geochronology. Earth Planet. Sci. Lett., 151:191–203. Rubatto, D., Gebauer, D. & Compagnoni, R. (1999): Dating of eclogite-facies zircons: the age of Alpine metamorphism in the Sesia-Lanzo Zone (Western Alps). Earth Planet. Sci. Lett., 167:141–158. Rubatto, D., Liati, A. & Gebauer, D. (2003): Dating UHP metamorphism. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 341–363. Rumble, D. (1998): Stable isotope geochemistry of ultrahigh-pressure rocks. In Hacker, B. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 241–259.

412

B.M. Jahn, D. Rumble & J.G. Liou

Rumble, D. & Yui, T.F. (1998): The Qinglongshan oxygen and hydrogen isotope anomaly near Donghai in Jiangsu province, China. Geochim. Cosmochim. Acta, 62:3307–3321. Rumble, D., Wang, Q.C. & Zhang, R.Y. (2000): Stable isotope geochemistry of marbles from the coesite UHP terrains of Dabieshan and Sulu, China. Lithos, 52:79–95. Rumble, D., Giorgis, D., Ireland, T., Zhang, Z.M., Xu, H.F., Yui, T.F., Yang, J.S., Xu, Z.Q. & Liou, J.G. (2002): Low G18O zircons, U–Pb dating, and the age of the Qinglongshan oxygen and hydrogen isotope anomaly near Donghai in Jiangsu Province, China. Geochim. Cosmochim. Acta, 66:2299–2306. Rumble, D., Liou, J.G. & Jahn, B.M. (2003): Continental crust subduction and UHP metamorphism. In Rudnick, R. (ed.).: Treatise on geochemistry, Vol. 3. (in press) Sanders, I.S., van Calsteren, P.W.C. & Hawkesworth, C.J. (1984): A Grenville Sm–Nd age for the Glenelg eclogite in northwest Scotland. Nature, 312:439–440. Sassi, R., Harte, B., Carswell, D.A. & Han, Y. (2000): Trace element distribution in central Dabie eclogites. Contrib. Mineral. Petrol., 139:298–315. Scambelluri, M. & Philippot, P. (2001): Deep fluids in subduction zones. Lithos, 55:213–227. Scherer, E.E., Cameron, K.L. & Blichert-Toft, J. (2000): Lu–Hf garnet geochronology: closure temperature relative to the Sm–Nd system and the effects of trace mineral inclusions. Geochim. Cosmochim. Acta, 64:3413–3432. Selverstone, J., Franz, G., Thomas, S. & Getty, S.R. (1992): Fluid variability in 2 GPa eclogites as an indicator of fluid behavour during subduction. Contrib. Mineral. Petrol., 112:341–357. Sharp, Z.D., Essene, E.J. & Hunziger, J.C. (1993): Stable isotope geochemistry and phase equilibria of coesitebearing whiteschists, Dora Maira massif, western Alps. Contrib. Mineral. Petrol., 114:1–12. Stosch, H.G. & Lugmair, G.W. (1990): Geochemistry and evolution of MORB-type eclogites from the Münchberg Massif, southern Germany. Earth Planet. Sci. Lett., 99:230–249. Tatsumi, Y. & Eggins, S. (1995): Subduction zone magmatism. Oxford: Blackwell, 211 p. Taylor, L.A., Snyder, G.A., Crozaz, G., Sobolev, V.N., Yefimova, E.S. & Sobolev, N.V. (1996): Eclogitic inclusions in diamonds: evidence of complex mantle processes over time. Earth Planet. Sci. Lett., 142:535–551. Tilton, G.R., Schreyer, W. & Schertl, H.P. (1989): Pb-Sr-Nd behavior of deeply subducted crustal rocks from the Dora Maira Massif, western Alps, Italy. Geochim. Cosmochim. Acta, 53:1391–1400. Tilton, G.R., Schreyer, W. & Schertl, H.P. (1991): Pb-Sr-Nd isotopic behavior of deeply subducted crustal rocks from the Dora Maira Massif, western Alps, Italy-II: what is the age of the ultrahigh-pressure metamorphism? Contrib. Mineral. Petrol., 108:22–33. Thöni, M. (2002): Sm–Nd isotope systematics in garnet from different lithologies (Eastern Alps): age results, and an evaluation of potential problems for garnet Sm–Nd chronometry. Chem. Geol., 185:255–281. Thöni, M. & Jagoutz, E. (1992): Some new aspects of dating eclogites in orogenic belts: Sm-Nd, Rb-Sr and Pb-Pb isotopic results from the Austroalpine Saualpe and Koralpe type-locality (Carinthia/Styria, southeastern Austria). Geochim. Cosmochim. Acta, 56:347–368. Vance, D. & O’Nions, R.K. (1990): Isotope chronometry of zoned garnets: growth kinetics and metamorphic histories. Earth Planet. Sci. Lett., 97:227–240. Vance, D. & O’Nions, R.K. (1992): Prograde and retrograde thermal histories from the central Swiss Alps. Earth Planet. Sci. Lett., 114:113–129. Von Quadt, A., Günther, D., Frischknecht, R., Zimmermann, R. & Franz, G. (1997): The evolution of preVariscan eclogites of the Tauern Window (eastern Alps): A Sm/Nd, conventional and laser ICP-MS zircon U–Pb study. Schweiz. Mineral. Petrogr. Mitt., 77:265–279. Wang, X.M., Jing, Y., Liou, J.G., Pan, G., Liang, W., Xia, M. & Maruyama, S. (1990): Field occurrences and petrology of eclogites from the Dabie Mountains, Anhui, central China. Lithos, 25:119–131. Wang, X.M., Zhang, R.Y. & Liou, J.G. (1995): UHPM terrane in east central China. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 356–390. Webb, L.E., Hacker, B.R., Ratschbacher, L., McWilliams, M.O. & Dong, S.W. (1999): Thermochronological constraints on deformation and cooling history of high- and ultrahigh-pressure rocks in the Qinling-Dabie orogen, eastern China. Tectonics, 18:621–638. Wu, Y.-B., Chen, D.-G., Xia, Q.-K, Deloule, E. & Cheng, H. (2002): SIMS U–Pb dating of zircons in granulite of Huangtuling from Northern Dabieshan. Acta Petrol. Sin., 18:378–382 (in Chinese with English abstr.).

Geochemistry and isotope tracer study of UHP metamorphic rocks

413

Xie, Z., Zheng, Y.-F., Jahn, B.-M., Ballevre, M., Chen, J.-F., Gautier, P., Gao, T.-S., Gong, B. & Zhou, J.-B. (2003): Sm–Nd and Rb-Sr dating for pyroxene-garnetite from North Dabie in east-central China: Problem of isotope disequilibrium due to retrograde metamorphism. Chem. Geol., submitted. Xu, S.T., Shu, W., Liu, Y.C., Wang, R.C., Jiang, L.L. & Wu, W.P. (1999): Discovery of eclogites in northern Dabie Mountains and their petrofabric characteristics. Chinese Sci. Bull., 44:1452–1456 (in Chinese). Xu, S.-T., Liu, Y.-C., Chen, G.-B., Compagnoni, R., Rolfo, F., He, M.-C. & Liu, H.-F. (2003): New findings of microdiamonds in eclogites from the Dabie-Sulu region in central-eastern China. Chinese Sci. Bull., 48:988–994. Yang, J.J. & Jahn, B.M. (2000): Deep subduction of mantle-derived garnet peridotites from the Su-Lu UHP metamorphic terrane in China. J. Metamorph. Geol., 18:167–180. Yui, T.F., Rumble, D. & Lo, C.H. (1995): Unusually low G18O ultrahigh-pressure metamorphic rocks from the Sulu terrain, eastern China. Geochim. Cosmochim. Acta, 59:2859–2864. Yui, T.F., Rumble, D., Chen, C.H. & Lo, C.H. (1997): Stable isotope characteristics of eclogites from the ultrahigh-pressure metamorphic terrains, east-central China. Chem. Geol., 137:135–147. Zhang, L.G. (1995): Geology of East Asian Lithosphere – Isotope geochemistry and dynamics of the upper mantle, basement and granitic rocks. Beijing: Science Publ., 252 p. (in Chinese, with English abstr.). Zhang, R.Y. & Liou, J.G. (1994): Coesite-bearing eclogite in Henan Province, central China: detailed petrology, glaucophane stability and PT path. Eur. J. Mineral., 6:217–233. Zhang, R.Y. & Liou, J.G. (1997): Partial transformation of gabbro to coesite-bearing eclogite from Yangkou, the Sulu terrane, eastern China. J. Metamorph. Geol., 15:183–202. Zhang, R.Y., Liou, J.G. & Cong, B. (1995a): Talc-, magnesite- and Ti-clinohumite-bearing ultrahigh-pressure meta-mafic and ultramafic complex in the Dabie Mountains, China. J. Petrol., 36:1011–1037. Zhang, R.Y., Hirajima, T., Banno, S., Cong, B. & Liou, J.G. (1995b): Petrology of ultrahigh-pressure rocks from the southern Su-Lu region, eastern China. J. Metamorph. Geol., 13:659–675. Zhang, R.Y., Liou, J.G. & Tsai, C.H. (1996a): Petrogenesis of a high-temperature metamorphic terrane: a new tectonic interpretation for the north Dabieshan, central China. J. Metamorph. Geol., 14:319–333. Zhang, R.Y., Liou, J.G. & Ye, K. (1996b): Petrography of UHPM rocks and their country rock gneisses. Chapter 4. In Cong, B. (ed.): Ultrahigh-pressure metamorphic rocks in the Dabieshan-Sulu region of China. Beijing: Science Press & Dordrecht: Kluwer, 49–68. Zhang, R.Y., Rumble, D., Liou, J.G., & Wang, Q.C. (1998): Low GO, ultrahigh-P garnet-bearing mafic and ultramafic rocks from Dabie Shan, China. Chem. Geol., 150:161–170. Zhang, R.Y., Liou, J.G., Yang, J.S. & Yui, T.F. (2000): Petrochemical constraints for dual origin of garnet peridotites from the Dabie-Sulu UHP terrane, eastern-central China. J. Metamorph. Geol., 18:149–166. Zhang, H., Gao, S., Zhong, Z., Zhang, B., Zhang, L. & Hu, S. (2002): Geochemical and Sr-Nd-Pb isotopic compositions of Cretaceous granitoids: constraints on tectonic framework and crustal structure of the Dabieshan ultrahigh-pressure metamorphic belt, China. Chem. Geol., 186:281–299. Zhang, R.Y., Liou, J.G., Zheng, Y. F. & Fu, B. (2003): Transition of UHP eclogites to gneissic rocks of low-grade amphibolite facies during exhumation: Evidence from the Dabie terrane, central China. Lithos, in press. Zhao, Z.Y., Fang, A.M. & Yu, L.J. (2002): UHP ductile shear complex at Yangkou, Sulu UHP metamorphic belt, central China: implication for continent subduction and exhumation. In Proc. 3rd UHPM Workshop, Beijing, 46–49. Zheng, Y.F. (1993): Calculation of oxygen isotope fractionation in anhydrous silicate minerals. Geochim. Cosmochim. Acta, 57:1079–1091. Zheng, Y.F., Fu, B., Gong, B. & Li, S. (1996): Extreme 18O depletion in eclogite from the Su-Lu terrane in east China. Eur. J. Mineral., 8:317–323. Zheng, Y.F., Fu, B., Li, Y., Xiao, Y. & Li, S. (1998): Oxygen and hydrogen isotope geochemistry of ultrahighpressure eclogites from the Dabie Mountains and the Sulu terrane. Earth Planet. Sci. Lett., 155:113–129. Zheng, Y.F., Fu, B., Xiao, Y., Li, Y. & Gong, B. (1999): Hydrogen and oxygen isotope evidence for fluid-rock interactions in the stages of pre- and post-UHP metamorphism in the Dabie Mountains. Lithos, 46:677–693. Zheng, Y.-F., Gong, B., Li, Y.-L., Wang, Z.-R. & Fu, B. (2000): Carbon concentrations and isotopic ratios of eclogites from the Dabie and Sulu terranes in China. Chem. Geol., 168:291–305. Zheng Y.-F., Wang, Z.-R., Li, S.-G. & Zhao, Z.-F. (2002): Oxygen isotope equilibrium between eclogite minerals and its constraints on mineral Sm–Nd chronometer. Geochim. Cosmochim. Acta, 66:625–634.

414

B.M. Jahn, D. Rumble & J.G. Liou

Zheng Y.-F., Fu B., Gong, B. & Li, L. (2003): Stable isotope geochemistry of ultrahigh pressure metamorphic rocks from the Dabie-Sulu orogen in China: Implications for geodynamics and fluid regime. Earth Sci. Rev., 61:in press. Zhou, B. & Hensen, B.J. (1995): Inherited Sm/Nd isotope components preserved in monazite inclusions within garnets in leucogneiss from East Antarctica and implications for closure temperature studies. Chem. Geol., 121:317–326. Zhou, T.X., Chen, J.F. & Zhang, X. (1995a): Geochemical study of the granite-syenite belt in the N. Huaiyang region and its tectonic significance. Geol. Rev., 41:144–151. Zhou, T.X., Li, X., Zhang, X., Wei, G., Pan, G. & Chen, J. (1995b): A lead isotopic study of alkali intrusion in the N. Huaiyang region. J. China Univ. Sci. Tech., 25:467–473.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 13, 415–441

Three-dimensional mechanics of UHPM terrains and resultant P–T–t paths PETER O. KOONS1*, PHAEDRA UPTON1 and MICHAEL P. TERRY2 1

Department of Earth Sciences, Bryand Global Sciences Center, University of Maine, Orono, Maine 04469 USA 2 Bayerisches Geoinstitut, Universität Bayreuth, D-95440 Bayreuth, Germany; * e-mail: [email protected]

Introduction In this paper we examine the mechanical and petrological interactions during formation and exhumation of ultrahigh pressure (UHP) terrains with the intention of producing a general three-dimensional dynamic model of continental subduction. We consider aspects of ancient and modern continental subduction to provide boundary conditions, rheological constraints and characteristic scales of time and space for the dynamic model in the generation and evolution of UHP during continental collision. The UHP and HP assemblages of the Western Gneiss of Norway provide rheological, geometric, and geochronological information for the modelling, while the active obliquely convergent plate boundary of central New Zealand serves as a modern analogue of the collision–subduction transition. The geodynamic models of oblique convergence, conditioned by these observations, provide orogen-wide velocity and strain rates and identify the characteristic length scales of strain partitioning within oblique subduction. The petro-structural evolution of individual packets within the oblique orogen are examined within a Lagrangian mixing model currently being developed that allows us to relate the large scale dynamic model to observations made at the level of an outcrop. Analytical and analogue models of continental subduction and ultrahigh pressure terrains identify a rather delicate force balance between buoyancy-generated body forces and viscous drag forces within a thickening crust and a downgoing slab. This balance is controlled by the strength of lithosphere materials that glue the heavy slab to the light crust and also produce a strong lid on top of the convergent zone (e.g. Chemenda et al., 1996; Ernst & Peacock, 1996; Davies & von Blackenburg, 1995). Factors affecting buoyancy are composition, pressure, temperature, the degree of metamorphic equilibration (Austrheim, 1998; Ryan, 2001) and the degree of asthenosphere involvement in the convergent zone. Material strength within the orogen depends upon the same factors and, in addition, is a function of local fluid pressure and erosion along the upper surface (Koons et al., 2002). Given a steady plate tectonic setting, i.e. a convergent or obliquely convergent plate boundary with a roughly constant ridge–orogen separation, the mechanical, thermal, structural, and petrological evolution of that margin is dominated by the time-dependent material behaviour of the rocks

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

416

P.O. Koons, Ph. Upton & M.P. Terry

transferred into the margin. Another source of time-dependent velocity patterns arises from interaction of the lithospheric slab with upper and lower mantle flow (Funiciello et al., 2003a,b). Consequently, any of the questions that arise regarding the geodynamics of UHP terrains in convergent margins, relate to the interaction of metamorphic evolution of orogen material with mantle convection. Information on chemical and mechanical interaction of UHP evolution is therefore contained in the petrological and structural signatures of ancient orogens and the kinematics of modern orogens. In this respect, the inter-relationship of deformation and metamorphism is similar to that identified in subduction/accretionary complexes where links between intermediate depth seismicity and dehydration are well-established (e.g. Abers, 2000; Hacker et al., 2003a,b).

General mechanical considerations Recognition of the widespread, if infrequent, occurrence of ultrahigh pressure assemblages indicating subduction of crustal material to depths of more than 100km has forced a paradigm shift in geodynamic understanding of convergent zones (e.g. Chopin, 1984; Smith, 1984; Hacker & Peacock, 1994; Carswell & Compagnoni, 2003). In order to model the particle paths of crustal material that exhibited UHP assemblages in dynamic models, it is necessary not only to include a large component of downward velocity associated with the downgoing slab, but also to consider a slab trajectory that is not necessarily parallel to the slab (e.g. Funiciello et al., 2003a,b). These two conditions introduce a dominant vertical velocity component to the orogen as opposed to a dominantly horizontal component of earlier models, and, more importantly, a component of divergence within the velocity field associated with volume increase. The requirement of continuity, or mass balance, in the mechanics of an orogen: 


 (
v ) t

(1)

(in which
= mass density and v = the three-dimensional velocity field with components u, v, w in the Cartesian directions x, y, z, respectively) is therefore difficult to satisfy in the absence of synchronous contraction and expansion within the orogen. Consequently, both divergent and contractional velocity fields will coexist in an orogen cross-section with positive and negative dilatation rates:  u w  Dilatation  0.5   x z 

(positive = expansion)

(2)

in the absence of changes in the plate tectonic setting. In this paper, we consider the implications of this divergence to the petro-structural evolution of orogens, and we concentrate on those rock-related features that either influence or record the geodynamic evolution of continental subduction. The gross mechanical problems associated with formation and preservation of UHP terrains are relatively well-defined (e.g. Molnar & Gray, 1979) and can be arbitrarily separated into the conditions controlling descent and those controlling exhumation of the buoyant crustal material.

3D mechanics of UHPM terrains and resultant P–T–t paths

417

The descent phase is capable of transporting relatively deep crustal material to depths of > 100km material within time frames on the order of 2 Ma. During this descent phase, the crustal material remains cool and strong and appropriate compositions can record the attainment of UHP conditions. This phase is driven by the sinking of the heavy lithosphere, dragging with it the crust that adheres to the slab. The amount and depth attained of crustal material is a function of the plate velocity, buoyancy and crustal strength, all of which are variable (Ryan, 2001). During the exhumation phase UHP terrains detach from the heavy slab and are displaced upwards to regions within high-pressure (HP) metamorphic conditions. The cause of detachment can be related to crust–mantle delamination, to slab breakoff (Davies & von Blackenburg, 1995), or steady state channel flow (Cloos, 1982). The thermal signals of these three mechanisms are potentially quite different as discussed below, but the mechanical effect is similar. The detached block moves up to the base of the crust with a rate of vertical displacement dependent upon relative buoyancy, strength of the crustal lid and viscous forces applied to the base. Kinetics provides an important, potentially rate-limiting, step in detachment and some observations exist on relative reaction rates, fluid evolution, and deformation in HP and UHP terrains (e.g. Rubie, 1983; Austrheim, 1998; Ernst et al., 1998). Further vertical displacement and tectonic imbrication carries the previously subducted material into the upper crust. This stage of exhumation is generally associated with kinematic evidence for crustal extension. Above, in the simplified treatment of the mechanical setting, we have arbitrarily separated descent and exhumation, implying a discreet break between the two regimes. However, under certain rheological conditions, these two regimes can be part of a continuous channel flow process (Cloos, 1982). The balance between the plate tectonic forces is therefore maintained by the metamorphic behaviour of the rocks caught within the convergent zone. The determining variables in convergent dynamics thus include those related to relative, threedimensional plate velocities as well as to the processes of metamorphic evolution. A measure of the tendency for delamination for very viscous systems is provided by the dimensionless Grashof number relating the ratio of buoyancy forces to viscous forces: Gr = L3g
/2

(3)

(L = characteristic length, g = gravitational acceleration, 
density difference between lighter crust and heavier mantle and deep crust,  = effective viscosity for a given strain rate). For Gr less than a critical value, Grcr, viscous drag dominates and crustal material can be subducted. If Gr > Grcr then buoyant crust will tend to separate from the heavy lithospheric slab. An approximation of Grcr = 1 may be produced by comparing buoyancy stresses generated in continental subduction with effective viscosity predictions based upon experimentally constrained behaviour of crustal material: Gref  buoyancy/viscous .

(4)

Where buoyancy =
gh; viscous = flow stress at strain rates of 10 exhibiting power law behaviour of the form:

–14

–1

sec for material

P.O. Koons, Ph. Upton & M.P. Terry

418

Fig. 1. Diagram of the effective Grashof number representing buoyancy stress versus resisting stress for viscous crustal material. Critical Grashof (Grcr = 1) separates regions that tend to descend (subcritical) from those where buoyancy forces dominate (supercritical). The labelled trajectories indicate the evolution predicted for thermally activated materials (governed, for example by Eqn. 5) undergoing conductive heating after subduction (Eqn. 9; Fig. 7.). The plagioclase and wet granite rheologies are calculated for strain rates of 10–14. The eclogite trajectory assumes only a buoyancy effect due to relevant phase changes.



 Q   RT 

 A n  exp

(5)

(Brace & Kohlstedt, 1980; Ranalli, 1995). Figure 1 illustrates the division between material likely to descend with the slab (Gr < 1) and the material tending to rise under buoyancy (Gr > 1). The specific evolutionary paths are related to thermal equilibration discussed in a later section. We emphasise that this dimensionless number is offered as a guide to the qualitative response of a buoyant/viscous system and does not consider the coupled dynamics of an orogen, which will certainly influence the local material velocities.

Mechanical model: Constraints from natural analogues In order to reduce some of the variance in our model conditions, we have used specific ancient and modern natural analogues to provide initial velocities, geometries and scales. The Western Gneiss Region is an erosional window through a stack of overlying thrust nappes that were transported hundreds of kilometres onto the former craton of Baltica during the Late Silurian–Early Devonian Scandian continental collision. The Baltica basement and its tectonic cover were then ductilely deformed together in the later phases of this collision. The distinct rock types associated with these nappes make excellent stratigraphic and structural markers in an area where Baltica basement has experienced high and ultrahigh pressures during Scandian continental collision. Following evidence expanded on below, the Scandian orogen appears to have been deformed with distinct vertical partitioning of strain into dominantly contractional at UHP depth, synchronous with dominantly extensional strain in the mid and upper crust (Austrheim et al., 2003). Although the general shape and conditions of the Laurentia–Baltica plate boundary during the Early Devonian are approximately known, additional kinematic and rheological observations available for modern convergent boundaries are unavailable for

3D mechanics of UHPM terrains and resultant P–T–t paths

419

ancient orogens. The relevant observations include seismicity patterns, seismic velocities, relative plate vectors, surface strain field and gravity field that all provide potential constraints on a dynamic model. We have, therefore, chosen the accessible and relatively well-characterised Pacific–Australian plate boundary in the transition from oblique subduction to oblique collision of central New Zealand. Associated with the transition of convergent mode is also a transition from a high-temperature convergent zone in the north near Taupo, to a low-temperature, high-pressure continental collision in the south. This tectonic and thermal pattern association is analogous to the coeval Acadian/Baltica convergent margin of the Devonian. Natural occurrences: Western Gneiss Region, Norway The ancient analogue is located in the Western Gneiss Region (Fig. 2), which represents a window through tectonic cover nappes and exposes high-pressure (Griffin et al., 1985) and ultrahigh pressure metamorphic rocks (Smith, 1984; Dobrzhinetskaya et al., 1995; Wain, 1997; Wain et al., 2000; Terry et al., 2000b). The Western Gneiss Region is predominantly composed of Proterozoic allochthonous and para-autochthonous granitoid gneisses that contain eclogite (Gee et al., 1985). Temperature estimates of

Fig. 2. (a) Generalised geological map showing narrow refolded synclines of the tectonic cover in Baltica crust and the location of the study area. (b) Generalised tectonostratigraphic map modified from Gee et al. (1985). Large arrow shows the orientation of a time averaged (450–425 Ma) relative motion vector for Baltica with respect to Laurentia, approximated from reconstructions of Torsvik (1998). Small arrows show movement vectors of allochthons. From Terry et al. (2000b).

420

P.O. Koons, Ph. Upton & M.P. Terry

metamorphism from isolated crustal (in situ) eclogites in Proterozoic basement by Krogh & Carswell (1995) show a systematic increase in temperature from ~ 600 °C to ~ 800 °C (Fig. 3). Detailed mapping by Robinson (1995, 1997), and Terry & Robinson, (1998, 2003) indicates that numerous tectonic cover sequences have been tightly folded into the Baltica basement and that the basement is imbricated (Fig. 2). Near the area of the transition to orogen parallel structures there are four regionally extensive thrust nappes overlying Baltica basement (Robinson, 1995; Lutro et al., 1997). The nappes, from bottom to top, include: 1) The Risberget Nappe, dominated by 1190 Ma rapakivi granite and subordinate metamorphosed gabbro, representing a slice of an unknown part of the Baltica basement. 2) The Sætra Nappe, deformed and metamorphosed Late Proterozoic feldspathic quartzite and amphibolite representing sandstone cut by diabase dikes. This represents a sedimentary assemblage with related

Fig. 3. Pressure–temperature–time–deformation histories for high- and ultrahigh-pressure rocks of the northern segment (Terry et al., 2000b).

3D mechanics of UHPM terrains and resultant P–T–t paths

421

dikes formed on the margin of Baltica during rifting that formed the Iapetus Ocean. 3) The Blåhø Nappe, metamorphosed, garnet-mica schist and amphibolite representing a volcanic arc assemblage formed and metamorphosed proximal to Baltica during the Early Ordovician or earlier. 4) The Støren Nappe dominated by low amphibolite facies metamorphosed mafic volcanic rocks with subordinate volcanogenic sedimentary rocks. The Støren Nappe and equivalents in the Trondheim region contain early Ordovician fossils of Laurentian affinity and are interpreted as an oceanic and island-arc assemblage produced within Iapetus or on its Laurentian margin, and thrust over Baltica during the Late Silurian. These nappes correlate directly with the Tånnås Augen Gneiss Nappe and the Särv Nappe of the middle allochthon of the Swedish Caledonides, and the Seve Nappe and the Köli Nappe of the upper allochthon of the Swedish Caledonides and the upper allochthon of the Trondheim basin, Norway. The underlying basement is dominated by 1680–1650 Ma granitoid gneisses cut by ~ 1500 Ma rapakivi granite (now augen gneiss) and by a variety of 1450–950 Ma gabbros, now variably eclogitised. Tectonic framework The Scandian orogeny, which resulted from collision between Baltica and Laurentia, includes emplacement of thrust nappes followed by and synchronous with subduction of Baltica beneath Laurentia, that resulted in production of HP and UHP eclogites and extensional orogenic collapse. Available paleomagnetic and paleogeographic data indicate that Baltica and Laurentia probably began to collide about 425 Ma with a latitudinal velocity vector of 8–10 cm per year, that resulted in oblique collision (Torsvik, 1998). Production of UHP rocks continued at least until 402 ± 2 Ma (R.D. Tucker in Lutro et al., 1997; Carswell et al., 2003). This was accompanied by rapid extensional exhumation that is divided into two phases that occurred between ca. 400 and 390 Ma. The first phase, associated with deformed ductile detachment faults in the hinterland, is interpreted to have been active during continued convergence of Baltica and Laurentia (Lutro et al., 1997) and to have continued until 395 Ma. The second phase was associated with the development of low-angle detachments that ultimately brought Devonian sedimentary rocks and underlying nappes into contact with the high-pressure Baltica basement (Andersen & Jamtveit, 1990; Andersen et al., 1991; Andersen et al., 1994). Kinematic framework Structural analysis by Terry (2000) and Terry & Robinson (2003) integrated with wellconstrained P–T–t paths from UHP and HP rocks (Terry et al., 2000a,b) on Nordøyane allow development of a kinematic framework for the Baltica margin during the Scandian orogeny from eclogite to low amphibolite facies. The structural features identified are divided into two categories, 1) those that formed at depths greater than 65 km associated with top-SE shearing (contraction) and 2) those that formed at less than 45 km to less than 20 km associated with top-W or left-lateral shearing (extension). The HP and UHP structures are seen where partitioning of late strain around metamorphosed gabbro and diorite gneiss in Baltica crust allows direct inferences to be made regarding the geometry of folds, kinematics and original structural orientations related to production and exhumation of HP rocks. In the diorite gneiss, the geometry of

422

P.O. Koons, Ph. Upton & M.P. Terry

folds associated with eclogite facies fabrics is isoclinal to tubular with axes parallel to the trend of a stretching lineation. Results from strain estimates and the presence of L > S or L >> S fabrics indicate that these structures were formed in a constrictional strain field. Eclogite facies mylonite zones that locally have a minimum thickness of 40 m cut Proterozoic gabbro and adjacent gneiss. Removal of late folding by small-circle rotation indicates that these structures were formed during top-SE shearing (140º) during prograde metamorphism in a direction that was consistent with thrusting parallel to plate motion. These structures are interpreted to be segment(s) responsible for juxtaposing HP and UHP rocks at a minimum depth of 65 km. The transition to late structural features is constrained to occur between 65 and 45 km depth by PT estimates of 780 °C and 1.3 GPa of well equilibrated mafic augen gneiss. The progression of these structures is also well preserved by strain partitioning. The earliest of these were extensional detachments juxtaposing eclogite facies rocks against overlying amphibolite facies rocks that show no evidence for eclogite facies metamorphism. These detachments are strongly overprinted and complexly folded, and they represent a phase of upper crustal extension that was active during continued convergence at deeper levels. Younger, more localised mylonite zones formed synchronously, by tubular, sheath, isoclinal, tight and open folding that shows a progression from WNW to ENE trends. The earliest mylonite zones, interpreted as originally subhorizontal, range in strike through a 20° angle from 110° to 90°. Later steeply dipping mylonite zones formed under lower amphibolite facies conditions, strike 75° and locally truncate earlier structures. The youngest mylonite zones formed at lowest amphibolite conditions, strike 50° and truncate all earlier structures. Folds developed during this progression show the range in orientation from WNW to ENE reflected in the orientations of the mylonite zones that is interpreted to represent progressive evolution during top-west shearing. These changes in orientation of the late structural features are from orogen-normal to orogen-parallel. There are two major difficulties that must be overcome in order to apply the above interpretation to the scale of the orogen. The first deals with extrapolation of these structual features to the scale of the orogen. The HP structures show a transport direction that is normal to the strike of the metamorphic field gradient (Griffin et al., 1985), identical to the transport direction at the base of the Jotun Nappe (Fossen, 2000), and consistent with plate motion (Torsvik, 1998). The extensional structures dominate the Western Gneiss Region and the extensional fabrics change from WNW to EW in the area of the Jotun nappe and the Devonian sedimentary basins (Fossen, 2000; Fig. 2). Identical changes have been observed in syn-depositional deformational structures in the Devonian basins, which provide a connection to very shallow levels in the crust (Osmundsen et al., 1998). In the north part of the Western Gneiss region narrow synclines of the nappes can be seen in the basement, which show a progressive change from E–W to NE. Together the changes mimic the detailed changes in the extensional structures seen on Nordøyane. The structural trends and changes observed on Nordøyane that rocks experienced during exhumation are in agreement with structure seen on an orogen scale. Thus, we believe it is reasonable to extend our interpretation to provide a kinematic framework for the orogen.

3D mechanics of UHPM terrains and resultant P–T–t paths

423

The second problem is time. Obviously, applying the structures that developed along P–T–t paths relates structures that formed at different times. Therefore, it is necessary to show that both compression at depth and extension of the upper crust were operating synchronously. Recent geochronologic and structural data (Williams et al., 1999; Terry, 2000; Terry et al., 2000a,b; Carswell et al., 2003; Austrheim et al., 2003) indicate overlap between the deposition of clastic sediments in extensional basins (403–394 Ma) at high structural levels and prograde eclogite facies metamorphism (407 ± 2–402 ± 2 Ma) at greater depth. Austrheim et al. (2003) recently reported a U–Pb zircon age of 401 ± 2 Ma for pegmatites that are approximately normal to the late nearly orogenparallel stretching direction. This age is indistinguishable from the age of the UHP metamorphism for the Ulsteinvik eclogite (402 ± 2 Ma) and indicates that syn-collisional exhumation involving upper crustal extension was probably an important mechanism for exhuming HP and UHP rocks. With these problems considered, it is possible to assemble a generalised kinematic model at the time of UHP metamorphism which involves orogen-normal thrusting at depths below 65 km and orogen-normal extension at 45 km depth that evolve to orogenparallel at < 20 depth km with an intervening transition zone of 20 km. This transition is well exposed in the area of Hustad but is poorly understood in terms of its detailed P–T–t–d evolution and the nature of strain partitioning. This kinematic framework provides a starting place to test a variety of different exhumation models for UHP rocks. Observations of the modern analogue: Central New Zealand Oblique continental subduction is currently occurring between the Pacific and Australian plates as a result of relative motion of ca. 39 mm yr–1 at an azimuth of ca. 255° (Fig. 4) (de Mets et al., 1994). The geometry and plate velocities along this boundary are used to constrain our dynamic model. Oblique convergence and a change in the composition of the Pacific plate from oceanic plateau in the north to continental in the south has led to the subduction of continental material beneath the northern South Island over at least the last 10 Ma (Fig. 4). South of the Hikurangi subduction margin, around Kaikoura, dramatic changes in the mechanics occur. The crust of the Hikurangi Plateau subducting beneath the North Island thickens from ca. 10 km in the north to ca. 15 km adjacent to the Chatham Rise from northeast to southwest (Davy & Wood, 1994). The Chatham Rise itself has a crustal thickness of 23–26 km. The Benioff zone of the Tonga–Kermadec–Hikurangi subduction zone ends abruptly at a line trending northwest through Kaikoura (Fig. 4). The southernmost edge of the subducted plate is identified by an abrupt step in the depth of the deepest earthquakes from 250 km to about 100 km (Fig. 4) (Anderson & Webb, 1994). This step is aligned with the 2000 m bathymetric contour at the transition from oceanic lithosphere of the Pacific plate to continental lithosphere of the Chatham Rise. On the basis of microearthquakes recorded during a detailed survey of the North Canterbury region, Reyners & Cowan (1993) suggest that the subducted slab continues farther to the southwest. They find that the dip of the seismic lower crust changes near 43°S from being subhorizontal in the southwest to a dip

424

P.O. Koons, Ph. Upton & M.P. Terry

of ca. 10°NW in the northeast. It is unclear whether the lower crust near 43°S is flexed or torn (Reyners & Cowan, 1993). Determination of the three-dimensional velocity structure to a depth of 100 km images the subducted slab between 40 and 100 km as a relatively low-velocity feature in the upper mantle (Eberhard-Phillips & Reyners, 1998) (Fig. 4). This is interpreted to reflect the continental nature of the subducting crust in this region. The amplitude of the low-velocity feature increases from northeast to southwest as the crust increases in thickness. Using existing calculations of thermal regime in a downgoing slab (Peacock, 1996; Hacker et al., 2003a,b), we calculate the predicted positions of coesite and jadeite stability beneath the northern part of the South Island (Fig. 4). Above the zone of continental subduction is the Marlborough Region, a broad zone of diffuse deformation in the upper crust extending on the order of 200 km to either side of the Wairau Fault, the terrane boundary between Western province rocks of the Australian plate and the Torlesse terrane of the Pacific Plate. The style of late Cenozoic deformation differs across the Wairau Fault; in the west, Miocene to recent reverse faulting has resulted in crustal thickening and uplift; to the east, the dominantly strikeslip Marlborough Fault System overlies the southernmost edge of the Hikurangi subduction zone (Walcott, 1998).

Fig. 4. (a) Digital Elevation Model (GEOgraphix) of the northern South Island showing the Marlborough Fault System, the Alpine Fault and the Kaikoura Ranges. Dashed grey line shows the edge of the Hikurangi subduction zone as recorded by an abrupt shallowing of the deepest earthquakes from 250 km to 100 km in depth (Anderson & Webb 1994). Inset shows the plate tectonic setting of central New Zealand, the 1000 and 2000 m bathymetric contours defining the transition from the oceanic Hikurangi Plateau to the continental Chatham Rise. Plate vector is from DeMets et al. (1994). (b) Schematic interpretation of the tectonics based on the 3D velocity model at the cross-section shown in (a). Modified from Eberhart-Phillips & Reyners (1998).

3D mechanics of UHPM terrains and resultant P–T–t paths

425

Mechanical framework I: Orogen-scale dynamics Our mechanical discussion follows both numerical and analytical approaches. The numerical approach consists of a three-dimensional dynamic model of continental subduction for a region with dimensions of 1000 km normal to the plate boundary, 600 km parallel to the plate boundary, and with a vertical extent of 200 km (Fig. 5). The resultant model provides three-dimensional velocity, strain rate, vorticity and dilatation rates, and stress fields for a variety of boundary and rheological conditions (e.g. Koons et al., 2002). Our solutions also track the thermal evolution of the problem domain, discussed below, to provide information on equilibrium and kinetic states in an externally fixed reference frame. In general, these preliminary numerical models serve as the vehicle for testing the sensitivity of continental subduction to variation in boundary and internal variables. Initial boundary conditions for the example presented here are characterised by a set of oblique far field velocities of lateral velocity: normal velocity (Vy : Vx) of 20 mm yr–1 : 10 mm yr–1 with differing vertical slab velocities imposed upon the base illustrated in Figure 5. These initial models have no along-strike variation in the applied boundary or rheological conditions. Numerical solution of the motion and stress equations is based upon algorithms from ITASCA (FLAC3D; Cundall & Board, 1988), a three-dimensional finite difference code, which we have modified to accommodate large strains and local erosion. Materials are represented by polyhedral elements within a three-dimensional grid that uses an explicit, time-marching solution scheme and a form of dynamic relaxation. Each element behaves according to a prescribed linear or non-linear stress/strain law in response to applied forces or boundary restraints. The inertial terms in the equations of motion,  ij xi


bi


dvi dt

(6)

(Cauchy’s equations of motion where ij is the stress tensor, xi, vi are the vector components of position and velocity, respectively,
is the density of the material, [b] is the body force) are used as numerical means to reach the equilibrium state of the system under consideration. The resulting system of ordinary differential equations is then solved numerically using an explicit finite difference approach in time. The drawbacks of the explicit formulation (i.e., small timestep limitation and the question of required damping) are overcome by automatic inertia scaling and automatic damping that does not influence the mode of failure. The governing differential equations are solved alternately, with the output for the solutions of the equations of motion used as input to the constitutive equations for a progressive calculation. Solution is achieved by approximating first-order space and time derivatives of a variable using finite differences, assuming linear variations of the variable over finite space and time intervals, respectively. The continuous medium is replaced by a discrete equivalent with all forces involved, concentrated at the nodes of a three-dimensional mesh used in the medium representation. Initially, we employ standard steady-state rheological models with pressure-dependent upper crustal rheology on top of temperature-dependent lower rheology for various compositions (e.g. Brace & Kohlstedt, 1980; Ranalli, 1995) (Fig. 5).

426

P.O. Koons, Ph. Upton & M.P. Terry

Fig. 5. Mechanical model of continental subduction. The block consists of three layers: 0–15 km depth consists of a pressure-sensitive Mohr-Coulomb material representing the upper crust; the lower crust, from 15–40 km, consists of a pressure-insensitive, temperature-sensitive rheology based on wet quartzite; the upper mantle, from 40–200 km, consists of a pressure-insensitive, temperature-sensitive rheology based on diabase. (a) Contours of velocity parallel to the plate margin (Vy). Dashed line shows position of cross sections below. (b) Contours of rotation. (c) Contours of shear strain rate. (d) Contours of lateral strain rate.

Numerical results Application of the conditions discussed above produces the velocity and deformation rate fields of Figure 5. The incoming and sinking slab generates deformation in the orogen along the top of the slab and within the slab itself. The vertical velocity

3D mechanics of UHPM terrains and resultant P–T–t paths

427

field in the orogen separates into material moving toward the surface and that entrained with the downgoing slab. Maximum downward vertical velocities within the slab are ~ 2.5 mm yr–1, while maximum upward mass velocities attained at the upper free surface are ~ 1.5 mm yr–1. The petrological implications of this velocity structure are considered below. Strain in the numerical models is certainly not homogeneous with clockwise vorticity about the y axis ( = –0.5(Vx/z – Vz/x), concentrated in the lower and descending crust along the top of the dipping slab down to depths of ~ 200km. Another region of clockwise vorticity exists in the eastern (x  850 km), upper crust representing the oblique thrust belts of the upper, outboard side of the orogen (Koons, 1990, 1994; Willett et al., 1993). Anti-clockwise vorticity is concentrated further to the west (x  600 km) in the upper crust where material is rapidly exhumed. The shape of this upper crustal vorticity concentration is strongly influenced by the surface erosion condition (Koons et al., 2003). Material passing from the east to the west through the orogen is subjected to both senses of vorticity and, consequently, should record an opposite sense of shear before it is exposed in the west. It should be noted that this shear sense reversal requires no change in the net plate vectors. Shear strain rates in the x–z plane (xz/t = Vz/x + Vx/z; Fig. 5b) are highest along a mid-crustal detachment and along the slab. Lateral strain rates (yz/t = Vy/x + Vy/z; Fig. 5d) provide an indication of the location of strike-slip deformation within the orogen. For the rheological model that we have chosen in this initial model, in addition to a concentration of lateral deformation within the upper crust and along the slab, there is also strike-slip deformation along a vertical zone that slices through the slab at ~ 850 km. Partitioning of lateral velocity along this cross-cutting vertical structure reduces the degree of obliquity in the down going slab. The dynamic model presented in Figure 5 captures some of the general features of continental subduction with UHP descent and exhumation. However, it fails to incorporate numerous, important features recognised in natural occurrences. The model’s utility lies primarily in being a vehicle for testing its response to perturbations and sensitivity to basic rheological and boundary assumptions including variation in surface erosional regimes. Most of these potential perturbations involve coupling among the processes of reaction and deformation related to both equilibrium and disequilibrium effects of reaction weakening/hardening behaviour, deformation enhanced reaction and the complex interplay of fluid production, migration and deformation. Role of disequilibrium Evidence for large departures from this steady-state rheological model, derived from HP and UHP models (e.g. Rubie, 1983; Koons et al., 1987; Austrheim, 1998), suggest a strong degree of strain–rheological coupling. Rubie (1983) suggested that strain-related reaction has led to super-plastic flow in HP terrains, thereby providing a positive feedback for rapid and large strains and is considered in later models. Rate of metamorphic equilibration influences crustal strength (Rubie, 1983), crustal buoyancy (Ryan & Dewey, 1997), and the record of equilibration (Austrheim, 1998). To provide useful information on the dynamics of convergent margins, it is

P.O. Koons, Ph. Upton & M.P. Terry

428

insufficient for assemblages to travel through PT space, they must also be capable of recording at least some locations on the path. It has long been recognised that the kinetic barriers to equilibration for dehydration reactions are relatively low and that those for relatively anhydrous rocks are high (e.g. Fyfe et al., 1958; Austrheim, 1998; Koons et al., 1987; Rubie, 1983). Seismic investigations of devolatilisation in active subduction zones reinforces this empirical kinetic model of near-equilibrium conditions for hydrated assemblages (e.g. Peacock, 1996; Abers, 2000; Hacker et al., 2003a,b). It is also clear that disequilibrium persists in polymetamorphic assemblages from mid- to deep crustal levels and that metamorphic reaction in these rocks is spatially and causatively linked to deformation and transient fluid pulses (e.g. Austrheim, 1998; Rubie, 1983). Influence of surface processes Although the paleoclimate of Baltica in the Devonian is not immediately obvious, the influence of erosion on orogen dynamics, as identified in both numerical models (e.g. Koons, 1990; Willett et al., 1993; Beaumont et al., 1992; Avouac & Burov, 1996; Blythe, 1998) and analogue models (Chemenda et al., 1996), is sufficiently important that we consider several different erosional schemes in our geodynamic models. Processes at the earth’s surface influence the large-scale orogen dynamics in several ways related to load development and rheological modification (Koons et al., 2002; 2003). Local relief similar to that of the Himalaya can produce stress perturbations within the upper crust that strengthen the regions beneath mountain loads relative to the valleys where significant shear stress concentrations occur (Koons et al., 2002). Focussed exhumation due to either fluvial erosion (Zeitler et al., 2001a,b; Koons et al., 2002) or from an orographically generated erosional pattern significantly alter the integrated strength of an orogen. Over very long time frames, erosion can deplete the upper crust in radiogenic components, lower the geothermal gradient, and consequently strengthen the upper crust (e.g. Dewey et al., 1993). However, on the time scales of most orogenic processes (< 50 Ma) the effect of concentrated exhumation is to significantly weaken rather than strengthen the crust. Exhumation drives the reduction of the integrated strength of the crust (= Fc; Sonder & England, 1986) by thermal thinning due to advection of isotherms within a concentrated region (e.g. Allis et al., 1979; Koons 1987). The amount of advection-driven weakening for a thermally activated lower crust and Mohr-Coulomb upper crust can be approximated by: Fc  ( N 1)

2
ghBD
ghBD h1 ( N 1) 2 2

(7)

(Koons et al., 2002) in which N is Terzaghi’s (1943) flow value     1 sin  N tan 2  

 4 2  1 sin 

for a Mohr-Coulomb material with angle of friction, .

(8)

3D mechanics of UHPM terrains and resultant P–T–t paths

429

This approximation emphasises the sensitivity of Fc to the square of the thickness of the frictional upper crust. Upward advection rapidly thins this upper frictional layer and is consequently very efficient at reducing the total strength of the crust. In regions of active convergence, this thermal weakening leads to strain concentration and can lead to the generation of a tectonic aneurysm with an associated high temperature metamorphic signal (Zeitler et al., 2001a,b). In extensional regions, thermal weakening can produce necking behaviour and further concentrate strain within the region of rapid erosion (Fletcher & Hallet, 1983). Strain partitioning within the orogen Along oblique plate boundaries, thermal thinning associated with exhumation can result in the concentration of lateral and convergent strain onto a single plate boundary, as along the Alpine Fault in southern New Zealand (Koons et al., 2003). The resulting kinematic evidence recorded in the structural fabric could suggest a change in plate convergence vectors in the opposite sense of the actual vector change. Further north along the same plate boundary, contraction, strike-slip and extension displacements along the Hikurangi margin coincide with strong lateral gradients in material properties (Fig. 6). A recent modelling study (Upton et al., in press) investigated the influence of material properties on velocity partitioning within oblique subduction zones, using a geometry based on the Hikurangi margin to the north (Fig. 6a). Rheological variation in the oblique models was constrained by seismic velocity and attenuation information available from the Hikurangi margin. Extension and velocity partitioning occur if the subduction interface is weak, but neither develops if the subduction interface is strong (Fig. 6b). The simple mechanical model incorporating rheological variation based on seismic observations produced kinematics that closely match those published from the Hikurangi margin, including: extension within the Taupo Volcanic Zone, uplift of ponded sediments (Eberhart-Phillips & Reyners 1998) and dextral contraction in the south of the North Island.

Mechanical framework II: Viscous mixing, local kinematics and nappe formation One of the primary goals in our mechanical treatment is to relate petro-structural observations of natural UHP terrains to the dynamics of convergent margins. Required for this is an estimate of global flow pattern within an orogen-fixed reference frame in an Eulerian sense as a function of bounding velocities and rheological parameters as produced in our dynamic model. By itself, however, the Eulerian velocity field is insufficient for describing the generation of folds with their associated thermal and compositional mixing, which constitute the petro-structural architecture of the orogen (Escher & Beaumont, 1997). The natural petro-structural architecture contains the kinematic and petrological information necessary to reconstruct position and rates of processes that formed the orogen. This architecture evolves as a function of local stretching and vorticity fields and is best defined within a local Lagrangian reference frame where displacements relative to individual particles can be tracked for at least part

430

P.O. Koons, Ph. Upton & M.P. Terry

Fig. 6. The effect of coupling strength on extension and partitioning behind a subduction zone (Upton et al., in press). (a) Northern New Zealand and the Hikurangi Trough. The interpreted degree of plate coupling along the Hikurangi subduction margin is shown in italics (Reyners, 1998). (b) Model results showing margin normal velocity. In the north, where the interface is weak, margin normal motion is accommodated largely on the plate interface. Extension within the upper plate occurs above the downgoing slab. In the south, where the interface is strong, margin normal motion is accommodated within the upper plate and no extension occurs. (c) Model results showing margin-parallel velocity. Across the whole margin, this component is accommodated within the upper plate. (d) Margin-normal (no dash) and margin-parallel (dashed) components across the weak interface and strong interface (bold lines). Partitioning and extension occur above a weak interface but not a strong interface. Modified from Upton et al. (in press).

3D mechanics of UHPM terrains and resultant P–T–t paths

431

of their history. The evolution of the material derivatives of a Lagrangian system can lead quickly to a complex pattern of tendrils and whorls similar to that found along the route to chaos characterised by the “baker transformation” (Ottino, 1989). This Lagrangian complexity in the midst of Eulerian simplicity is responsible for the large and small scale viscous folds and geochemical mixing of geophysical fluids (e.g. Allegre & Turcotte, 1986; Passchier, 1997), and natural and analogue examples are provided in Figure 7. In these examples, non-coaxial, complex refolding is associated with vortex flow as opposed to dominantly shear flow (Ishii, 1992). In the following discussion, we assume that nappe formation results from a Lagrangian mixing process similar to that demonstrated for other geophysical fluids (e.g. Fig. 7), and that this process is facilitated in regions with a large component of vortex flow relative to shear flow, and we examine the thermal implications of nappe formation. High vortex flow is predicted within regions where the Eulerian velocity field contains a component of divergent shear similar to that illustrated in our dynamic model (Fig. 5) and in natural and analogue examples (Fig. 7). We predict that divergent shear will concentrate vortex flow, with associated nappe formation, immediately beneath the upper plate and, further, that the orientation and wavelength is a function of the relative amounts of vortex vs. shear flow and, consequently, that these folds will contain information on relative slab displacement. Thermal and petrological evolution Since at least Albarede (1976), there has been a recognition that the mass velocity in an active region strongly influences the thermal state of that region and, consequently, that

Fig. 7. Large scale folding in natural (Malaspina Glacier) and analogue materials illustrate local complexity in Lagrangian mixing for a simple Eulerian velocity field (vectors on the right image indicate instantaneous surface velocity field). The transition from simple shear flow to the complex folded nappe-type structure occurs where the velocity field is characterised by divergent shear and vortex flow becomes significant. (Image of Malaspina Glacier from http://earthobservatory.nasa.gov/Newsroom/NewImages/images.php3?img_id=10307)

P.O. Koons, Ph. Upton & M.P. Terry

432

the thermal state of an orogen can be modelled with the three-dimensional transient partial differential equation for an advecting medium: T k T T T

( 2T ) (u v w ) A( x , y , z ) , t x y z
C p( x,y,z )

(9)

in which k = thermal conductivity as a spatial function, Cp = heat capacity,
= density, u, v, w = velocity in an externally-fixed, Cartesian reference frame, A = spatially varying radioactive heat production term discussed below. In our numerical treatment, two-dimensional solutions employ a standard CrankNicholson approximation for the mildly non-linear advective conductive system (Koons, 1987). Three-dimensional solutions for the non-linear advection–conduction equations employ approximations derived from ITASCA (Cundall & Board, 1988). We address questions related to petrological and mechanical evolution with static, kinematic and dynamic models. Many of the difficulties and strategies involved in tracking the thermal history are thoroughly discussed by Roselle et al. (2002). The standard model geometry used in the static models is 400 km on a side with a depth of 200 km. The block is discretised at variable resolutions depending upon the model, with grid spacing in the standard model set at 20 × 20 × 50 grid elements that are further divided into tetrahedral elements with linear thermal distribution across subelements (Fig. 8). In the unperturbed model with 40 km of crust, 110 km of mantle lithosphere, and 50 km of asthenosphere, radioactive heat production is concentrated in the upper 20 km of the crust and reduced to 10% in the lower crust and further reduced to 1% in the mantle (Table 1). The asthenosphere/lithosphere boundary is defined as the 1380 °C isotherm at which the basal boundary is maintained. Model sides are no netflow boundaries and generally are sufficiently removed from thermal perturbations that they exert no thermal influence on the solutions. The steady state geotherm for these conditions is shown in Figure 8 outside the region of perturbation and is similar to those predicted from older, continental crust. We recognise that natural parameter ranges will result in a spread of temperatures and that our calculations therefore represent reasonable approximations to thermal trends but not absolute thermal values.

Table 1. Properties of thermal models

Crust Crustal Root Lithosphere Asthenosphere




A

2700 2700 3300 3400

3.5  10 3.5  10–7 1.2  10–8 1.2  10–8 –6



Cp

3.0 3.0 3.75 3.75

1.0  10 1.0  103 1.0  103 1.0  103

T (ºC) 3

Top fixed at 10 ºC --------------------1380 ºC

Where
is density (kg m–3), A is radioactive heat production term (W kg–1),  is the thermal conductivity (W m–1K–1) and Cp is specific heat (J kg–1K–1).

Fig. 8. (a) P–T deformation histories for the UHP Upper Plate (path 1) and HP Lower Plate (path 2) of northern Nordøyane. Grey areas indicate P–T estimates for the Upper Plate based on thermobarometry (Terry et al., 2000a,b; Ravna, 2000). (b) Thermal model of an instantaneously thickened crustal root, showing thermal profile at 3 and 6 Ma after emplacement. Position through time on the P–T diagram of the base of the crustal root (shown by x) is indicated by “model 1”. (c) Thermal model of slab break-off leading to instantaneous emplacement of asthenosphere to within 10 km of the crustal root, showing thermal profile at 3 and 6 Ma after break-off. Position through time on the P–T diagram of the base of the crustal root (shown by x) is indicated by “model 2”.

3D mechanics of UHPM terrains and resultant P–T–t paths 433

434

P.O. Koons, Ph. Upton & M.P. Terry

Results of static solutions: Crustal thickening Perturbation of the lithosphere due to crustal thickening to the 120 km depths indicated by UHP assemblages is represented in the models of Figure 8. Thickening was accomplished in the model by instantaneous thickening of material with the same properties as the lower crust by 80 km over a horizontal distance of 100 km in the x direction and 400 km in the y direction. Consequently, while the solution is fully threedimensional, T/y is minimised in the static models and the solutions are effectively two-dimensional. The initial temperature of this crustal root was homogenised to 600 °C. This assumption and the assumption of instantaneous emplacement are related to viscous mixing and are discussed below. The thermal solutions behave in a manner predictable from the initial solutions of England & Thompson (1984) and Grasemann et al. (1998). Initial cooling of the lithosphere by a vertical flow of cooler crustal material leads to UHP conditions in the underplated wedge. Thermal response of the wedge is very slow due to insulation from the asthenosphere by the thickened mantle lithosphere and results from radioactive heating with characteristic times of > 40 Ma. Eventually, the increased radioactive source distribution in the overthickened welt warms the welt and the steadystate solution, relevant for t > 100 Ma, contains a thermal hotspot in the overthickened crustal root (Fig. 8b). The time required to approach steady state is so great that it is largely irrelevant for our investigation of the processes at active convergent margins with time frames of < 20 Ma. On these short time frames, underplating has a cooling effect and thermal equilibration is very limited (Liou et al., 1996). Asthenospheric involvement The thermal effect on the crustal root of lithospheric slab breakoff (Davies & von Blackenburg, 1995) is represented by instantaneous upward displacement of the asthenosphere to within 10 km of the crustal root (Fig. 8c). For reference in the discussion, the evolution of a particle at 110 km below the surface, 10 km above the Moho, is chosen and plotted together with the position of the 800 °C isotherm (Fig. 8). The 10 km thick remaining lithospheric lid insulates the crustal root for more than 1 Ma. By 2 Ma, the thermal state of the crustal root has been significantly perturbed to a level that should be visible with present geothermometrics, assuming metamorphic reequilibration of the assemblage. The 800 °C isotherm has reached 10 km into the root by 3 Ma and may be representative of the thermal evolution of the Western Gneiss UHP assemblages (Krogh & Carswell, 1995) (Fig. 8a). Removal of more of the insulating lithospheric lid, bringing hot asthenosphere closer to the crust, will reduce the response time in the root in a manner well approximated by the tx = x2c/ expression. Evolution of the thermal structure following asthenospheric involvement at longer time scales is similar to that discussed extensively by Ryan & Dewey (1997). Thermal modelling of actively deforming regions has generally proceeded through solutions of transient non-linear equations for conductive and advective heat transport in two or three spatial dimensions (e.g. Koons, 1987; Peacock, 1995; Grasemann et al., 1998; Lin & Roecker, 1998; Roselle et al., 2002; Hacker et al., 2003a,b). While the thermal field in these solutions is transient, the models are kinematic and assumed

3D mechanics of UHPM terrains and resultant P–T–t paths

435

velocity fields have been steady and smooth. As indicated by Sleep (1979) the process of internal deformation in nappe formation can have a significant influence upon the thermal and therefore petrological evolution of an orogen. This influence is in part approximated by the characteristic length scale of conduction ( xt t ) and is consequently sensitive to fold wavelength (Sleep, 1979). The generation and evolution of nappes and folded isotherms is, as discussed above, spatially linked to regions of high vorticity (Fig. 9). The thermal evolution of these nappes is examined for various dominant wavelengths in the models presented in Figure 9 to see if petrological details of divergent P–T–t paths as a function of folding are likely to persist. For wavelengths of less than 5 km, thermal homogenisation occurs within the folded structures during subduction of crustal material (Fig. 9a). For wavelengths ~ 10 km, the thermal structure remains folded for periods of > 5 Ma (Fig. 9b) and should be visible within the assemblage record.

Fig. 9. Thermal relaxation of instantaneously placed folds of varying thickness. (a) A series of nappes of thickness 5 km are placed into the upper mantle as a crustal root. After 2 Ma, a pod of colder material exists within the crustal root. (b) Thermal relaxation for nappes of thickness 10 km.

436

P.O. Koons, Ph. Upton & M.P. Terry

Dynamic model with thermal results In regions of steep velocity gradients, the advective terms of the heat flow equation (Eqn. 9) dominate and the resultant thermal field carries the signature of the velocity field in a manner represented by the dimensionless Peclet number (Pe = VL/; where V = velocity normal to the thermal gradient,  = thermal diffusivity, and L = thermal length scale). The relevant length scale (L) for Peclet varies, but is generally on the order of either slab thickness or nappe thickness. As demonstrated in the initial static models, the former leads to thermal insulation represented by high Peclet, while the latter may allow local thermal homogenisation if nappe thickness is less than 5 km (Fig. 9). The velocity distribution of our dynamic model produces regions of high Peclet near the surface and within the downgoing slab where thermal recovery significantly lags behind pressure changes leading to near isothermal exhumation or burial. High Peclet along the slab results in formation of lenses of coesite and diamond stability within the subducted material (Fig. 10b). Particle paths of UHP must pass through those stability regions at rates that are, in the Western Gneiss region and the Pacific–Australian plate boundary, at least within the time frame of several million years. Depending upon the surface erosion conditions, a thermal hotspot can form in the near surface where exhumation is rapid, concentrating strain and leading to local high temperature metamorphism (Koons et al., 2002, 2003). In addition to the regions of high Peclet, there are regions in the mid to lower crust of low Peclet where there is little advective influence on the thermal structure. These regions of greenschist to amphibolite grade metamorphism are associated with high strain rates (Fig. 10b). Thermal–mechanical coupling In addition to the petrological evolution of crustal assemblages, thermal transience also influences the mechanical behaviour of the orogen for materials with thermally activated rheology (Eqn. 5). The degree of heating related to thermal influence from the asthenosphere reduces the flow resistance of material in the wedge. This thermally induced weakening can shift subducted continental material from subcritical to supercritical Grashof and permit this buoyant material to ascend (Fig. 1). The composition and rate of heating determines the trajectory of subducted material within Grashof space.

Discussion Our three-dimensional dynamic model is successful in reproducing some of the observations from ancient and modern collisional orogens, but fails in several important aspects. In general, those thermal, petrological and structural features associated with the exhumation phase of UHP formation are at least in part captured in the dynamic model. These include depression of isotherms to produce UHP conditions, length and time scales similar those of the modern Pacific–Australian plate boundary, and significant vertical and horizontal strain partitioning. All of these features result from a standard, time-independent rheological and boundary model in which the crustal material is separated from high temperature asthenosphere by a lithospheric lid.

3D mechanics of UHPM terrains and resultant P–T–t paths

437

Fig. 10. Mechanical and thermal model of continental subduction. (a) As for Figure 5a. (b) Cross-section through the model illustrating the thermal regime associated with the subducting slab. Stability fields of coesite and diamond are shown within the slab.

The exhumation phase is poorly represented with time-independent rheologies and constant boundary conditions. In the absence of the strong non-linear feedbacks that arise through mechanical and rheological coupling, the detachment of deeply subducted blocks of mafic material from the down going slab is unlikely. Rheological timedependence sufficient to permit exhumation can result from heating of thermally activated crustal material, causing the subducted crust to become supercritical. In addition, more rapid rheological transitions can be produced through metamorphic transitions such as fluid production or reaction-enhanced ductility (e.g. Rubie, 1983). If these processes are widespread, then rapid fluctuations in crustal viscosity are possible and the conduction time scale of crustal heating no longer constrains the rates of UHP separation and exhumation. In the dynamic model, we have not considered any shift in the far field velocities of the lithospheric slab. Given the recent observations from natural and analogue models of trench rollback (Funiciello et al., 2003b) it is clear that interaction with mantle flows, both in the upper and lower mantle, can profoundly affect the boundary conditions that influence continental subduction. Similarly, the high pressure heating part of the Western Gneiss P–T–t history (at 405 Ma) suggests an influence of high temperature asthenosphere on at least a limited scale prior to exhumation, in a manner not compatible with the steady state boundary models. Incorporation of these boundary variations in the dynamic model would initiate linking of mantle convection models with crustal dynamics and appears to be a rich direction for future investigation.

438

P.O. Koons, Ph. Upton & M.P. Terry

Acknowledgements We would like to thank D.A. Carswell and R. Compagnoni for the opportunity to present this study. Kirsten Jones is thanked for her contribution to the thermal modelling presented here and her assistance with many of the figures. Leigh Sterns and Erich Osterberg are thanked for Figure 7.

References Abers, G.A. (2000): Hydrated subducted crust at 100–250 km depth. Earth Planet. Sci. Lett., 176:323–330. Albarede, F. (1976): Thermal models of post-tectonic decompression as exemplified by the Haut Allier granulites (Massif Central, France). Bull. Soc. Geol. Fr., 18:1023–1032. Allegre, C.J. & Turcotte, D.L. (1986): Implications of a two-component marble-cake mantle. Nature, 323:123–127. Allis, R.G., Henley, R.W. & Carman, A.F. (1979): The thermal regime beneath the Southern Alps. In Walcott, R.I. & Creswell, M.M. (eds.): The origin of the Southern Alps /Bull. - R. Soc. N. Z., 18/, 79–85. Andersen, T.B. & Jamtveit, B. (1990): Uplift of deep crust during orogenic extensional collapse: A model based on field studies in the Sogn-Sunnfjord region, W. Norway. Tectonics, 9:1097–1111. Andersen, T.B., Jamtveit, B., Dewey, J.F. & Swensson, E. (1991): Subduction and eduction of continental crust: major mechanisms during continent-continent collision and orogenic extensional collapse, a model based on the south Norwegian Caledonides. Terra Nova, 3:303–310. Andersen, T.B., Osmundsen, P.-T. & Jolivet, L. (1994): Deep crustal fabrics and a model for extensional collapse of the southwest Norwegian Caledonides. J. Struct. Geol., 16:1191–1203. Anderson, H. & Webb, T. (1994): New Zealand seismicity: patterns revealed by the upgraded National Seismograph Network. N. Z. J. Geol. Geophys., 37:477–493. Austrheim, H. (1998): Influence of fluid and deformation on metamorphism of the deep crust and consequences for the geodynamics of collision zones. In Hacker, B.R. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 297–323. Austrheim, H., Corfu, F., Bryhni, I. & Andersen, T.B. (2003): The Proterozoic Hustad igneous complex: a low strain enclave with a key to the history of the Western Gneiss Region of Norway. Precambrian Res., 120:149–175. Avouac, J.P. & Burov, E.B. (1996): Erosion as a driving mechanism of intracontinental growth. J. Geophys. Res., 101:17,747–17,769. Beaumont, C., Fullsack, P. & Hamilton, J. (1992): Erosional control of active compressional orogens. In McClay, K.R. (ed.): Thrust tectonics. New York (N.Y.): Chapman & Hall, 1–18. Blythe, A.E. (1998): Active tectonics and ultrahigh-pressure rocks. In Hacker, B.R. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks, Dordrecht: Kluwer, 141–160. Brace, W.F. & Kohlstedt, D. (1980): Limits on lithospheric stress imposed by laboratory experiments. J. Geophys. Res., 89:6248–6252. Carswell, D.A. & Compagnoni, R. (2003): Introduction with review of the definiton, distribution and geotectonic significance of ultrahigh pressure metamorphism. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 3–9. Carswell, D.A., Tucker, R.D., O’Brien, P.J. & Krogh, T.E. (2003): Coesite micro-inclusions and the U/Pb age of zircons from the Hareidland Eclogite in the Western Gneiss Region of Norway. Lithos, 67:181–190. Chemenda, A., Mattauer, M. & Bokun, A.N. (1996): Continental subduction and a mechanism for the exhumation of high-pressure metamorphic rocks: New modeling and field data from Oman. Earth Planet. Sci. Lett., 143:173–182. Chopin, C. (1984): Coesite and pure pyrope in high-grade blueschists of the western Alps: a first record and some consequences. Contrib. Mineral. Petrol., 86:107–118. Cloos, M. (1982): Flow mélanges: numerical modeling and geologic constraints on their origin in the Franciscan subduction complex. Geol. Soc. Am. Bull., 93:330–345.

3D mechanics of UHPM terrains and resultant P–T–t paths

439

Cundall, P., & Board, M. (1988): A microcomputer program for modeling of large-strain plasticity problems. In Swoboda, C. (ed.): Numerical methods in geomechanics. Proc. 6th Int. Conf. on Numerical Methods in Geomechanics, Innsbruck, Austria. Rotterdam: Balkema, 2:101–108. Davies, J.H. & von Blackenburg, F. (1995): Slab breakoff: A model of lithosphere detachment and its test in the magmatism and deformation of collisional orogens. Earth Planet. Sci. Lett., 129:85–102. Davy, B. & Wood, R. (1994): Gravity and magnetic modelling of the Hikurangi Plateau. Mar. Geol., 118:139–151. De Mets, C., Gorden, R.G., Argus, D.F. & Stein, S. (1994): Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophys. Res. Lett., 21:2191–2194. Dewey, J.F., Ryan, P.D. & Andersen, T.B. (1993): Orogenic uplift and collapse, crustal thickness, fabrics and metamorphic phase changes; the role of eclogites. J. Geol. Soc. London, Spec. Publ., 76:325–343. Dobrzhinetskaya, L.F., Eide, E.A., Larsen, R.B., Sturt, B.A., Trønnes, R.G., Smith, D.C., Taylor, W.R. & Posukhova, T.V. (1995): Microdiamond in high-grade metamorphic rocks of the Western Gneiss Region, Norway. Geology, 23:597600. Eberhart-Phillips, D. & Reyners, M. (1998): Continental subduction and three-dimensional crustal structure: The northern South Island, New Zealand. J. Geophys. Res., 102:11,843–11,861. England, P.C. & Thompson, A.B. (1984): Pressure-temperature-time paths of regional metamorphism, Part I: Heat transfer during the evolution of regions of thickened continental crust. J. Petrol., 25:894–928. Ernst, W.G. & Peacock, S.M. (1996): A thermotectonic model for the preservation of ultrahigh-pressure phases in metamorphosed continental crust. In Bebout, G.E., Scholl, D.W., Kirby, S.H. & Platt, J.P. (eds.): Subduction: Top bottom /Geophys. Monogr., 96/. Washington (D.C.): Am. Geophys. Union, 171–178. Ernst, W.G., Mosenfelder, J.L., Leech, M.L. & Liu, J. (1998): H2O recycling during continental collision: phase-equilibrium and kinetic considerations. In Hacker, B.R. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 275–295. Escher, A. & Beaumont, C. (1997): Formation, burial and exhumation of basement nappes at crustal scale: a model based on the Western Swiss-Italian Alps. J. Struct. Geol., 19:955–974. Fletcher, R. C. & Hallet, B. (1983): Unstable extension of the lithosphere: a mechanical model for Basin-andRange structure. J. Geophys. Res., 88:7457–7466. Fossen, H. (2000): Extensional tectonics in the Caledonides: Synorogenic or postorogenic? Tectonics, 19:213–224. Funiciello, F., Morra, G., Regenauer-Leib, K. & Giardini, D. (2003a): Dynamics of retreating slabs: 1. Insights from two-dimensional numerical experiments. J. Geophys. Res., 108(B4):2206, doi:10.1029/2001JB000898. Funiciello, F., Faccenna, C., Giardini, D. & Regenauer-Leib, K. (2003b): Dynamics of retreating slabs: 1. Insights from three-dimensional laboratory experiments. J. Geophys. Res., 108(B4):2207, doi:10.1029/2001JB000896. Fyfe, W.S., Turner, F.J. & Verhoogen, J. (1958): Metamorphic reactions and metamorphic facies /Geol. Soc. Am. Mem., 73/. Gee, D.G., Kumpulainen, R., Roberts D., Stephens, M.B., Thon, A. & Zachrisson, E. (1985): Scandinavian Caledonides – tectonostratigraphic map. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen – Scandinavia and related areas, Chichester: Wiley. Grasemann, B., Ratschbacher, L. & Hacker, B.R. (1998): Exhumation of ultrahigh-pressure rocks: Thermal boundary conditions and cooling history. In Hacker, B.R. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks. Dordrecht: Kluwer, 117–139. Griffin, W.L., Austrheim, H., Brastad, K., Brynhi, I., Krill, A.G., Krogh, E.J., Mørk, M.B.E., Qvale, H. & Tørudbakken, B. (1985): High-pressure metamorphism in the Scandinavian Caledonides. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide orogen – Scandinavia and related areas. Chichester: Wiley, 783–801. Hacker, B.R. & Peacock, S.M. (1994): Creation, preservation, and exhumation of ultrahigh pressure metamorphic rocks. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 159–181. Hacker, B.R., Abers, G.A. & Peacock, S.M. (2003a): Subduction factory I. Theoretical mineralogy, densities and seismic wave speeds, and H2O contents. J. Geophys. Res., 108(B1):2029, doi:10.1029/2001JB001127 Hacker, B.R., Peacock, S.M., Abers, G.A. & Holloway, S.D. (2003b): Subduction factory 2. Are intermediatedepth earthquakes in subducting slabs linked to metamorphic dehydration reactions? J. Geophys. Res., 108(B1):2030, doi:10.1029/2001JB001129

440

P.O. Koons, Ph. Upton & M.P. Terry

Ishii, K. (1992): Partitioning of non-coaxiality in deforming layered rock masses. Tectonophysics, 210:33–43. Koons, P.O. (1987): Some thermal and mechanical consequences of rapid uplift; An example from the Southern Alps, New Zealand. Earth Planet. Sci. Lett., 86:307–319. Koons, P.O. (1990): The two sided wedge in orogeny; Erosion and collision from the sand box to the Southern Alps, New Zealand. Geology, 18:679–682. Koons, P.O. (1994): Three-dimensional critical wedges: Tectonics and topography in oblique collisional orogens. J. Geophys. Res., 99:12,301–12315. Koons, P.O., Rubie, D.C. & Frueh-Green, G. (1987): The effect of disequilibrium and deformation on the mineralogical evolution of an orthogneiss during high-pressure metamorphism. J. Petrol., 28:679–700. Koons, P.O., Zeitler, P.K., Chamberlain, C.P., Craw, D. & Melzer, A.S. (2002): Mechanical Links Between Erosion and Metamorphism in Nanga Parbat, Pakistan Himalaya. Am. J. Sci., 302:749–773. Koons, P.O., Norris, R.J., Craw, D. & Cooper, A.F. (2003): Influence of exhumation on the structural evolution of transpressional plate boundaries: An example from the Southern Alps, New Zealand. Geology. 31:3–6. Krogh, E.J., & Carswell, D.A. (1995): UP and UHP eclogites and garnet peridotites in the Scandinavian Caledonides. In Coleman, R.G. & Wang, X. (eds.): Ultrahigh pressure metamorphism. Cambridge: Cambridge Univ. Press, 244–298. Larsen, R.B., Eide, E.A. & Burke, E.A.J. (1998): Evolution of microdiamond-bearing granulites in the Western Gneiss Region, Norway. Contrib. Mineral. Petrol., 133:106–121. Lin, C.H. & Roecker, S.W. (1998): Active crustal subduction and exhumation in Taiwan. In Hacker, B.R. & Liou, J.G. (eds.): When continents collide: Geodynamics and geochemistry of ultrahigh-pressure rocks, Dordrecht: Kluwer, 1–25. Liou, J.G., Zhang, R.Y., Eide, E.A., Maruyama, S., Wang, X. & Ernst, W.G. (1996): Metamorphism and tectonics of high-P and ultrahigh-P belts in Dabie-Sulu Regions, eastern central China. In Yin, A. & Harrison, T.M. (eds.): The tectonic evolution of Asia. Cambridge: Cambridge Univ. Press, 300–344. Lutro, O., Robinson, P. & Solli, A. (1997): Proterozoic geology and Scandian overprinting in the Western Gneiss Region. Nor. Geol. Unders., Rep. 97.132, 86 p. Molnar, P. & Gray, D. (1979): Subduction of continental lithosphere: some constraints and uncertainties. Geology, 7:58–62. Mørk, M.B.E. (1985): A grabbro to eclogite transition on Flemsøy, Sunmøre, western Norway. Chem. Geol., 50:283–310. Mørk, M.B.E. & Mearns, E.W. (1986): Sm-Nd isotopic systematics of a gabbro-eclogite transition. Lithos, 19:255–267. Osmundsen, P.T., Andersen, T.B., Markussen, S. & Svendby, A.K. (1998): Tectonics and sedimentation in a supra-detachment environment: The Kvamshesten Devonian basin, western Norway. Basin Res., 10:213–234. Ottino, J.M. (1989): The kinematics of mixing: stretching, chaos and transport. Cambridge: Cambridge Univ. Press. Passchier, C.W. (1997): The fabric attractor. J. Struct. Geol., 19:113–127. Peacock, S.M. (1995): Ultrahigh-pressure metamorphic rocks and the thermal evolution of collisional belts. Isl. Arc, 4:376–383. Peacock, S.M. (1996): Thermal and petrologic structure of subduction zones. In Bebout, G.E., Scholl, D.W., Kirby, S.H. & Platt, J.P. (eds.): Subduction: Top to bottom /Geophys. Monogr., 96/. Washington (D.C.): Am. Geophys. Union, 119–133. Ranalli, G. (1995): Rheology of the Earth. 2nd ed. London: Chapman & Hall. Ravna, E.J.K. (2000): The garnet-clinopyroxene Fe2+-Mg geothermometer: an updated calibration. J. Metamorph. Geol., 18:211–219. Reyners, M. (1998): Plate coupling and the hazard of large subduction thrust earthquakes at the Hikurangi subduction zone, New Zealand. N. Z. J. Geol. Geophys., 41:343–354. Reyners, M. & Cowan, H. (1993): The transition from subduction to continental collision: crustal structure in the North Canterbury region, New Zealand. Geophys. J. Int., 115:1124–1136. Robinson, P. (1995): Extension of Trollheimen tectono-stratigraphic sequence in deep synclines near Molde and Brattvåg, Western Gneiss Region, southern Norway. Nor. Geol. Tidsskr., 75:181–198. Robinson, P. (1997): Quartzite of Reksdalshesten, Moldefjorden, Western Gneiss Region: Another parallel with the tectonostratigraphy of Trollheimen. Nor. Geol. Unders., 433:14–15.

3D mechanics of UHPM terrains and resultant P–T–t paths

441

Roselle, G.T., Thuring, M. & Engi, M. (2002): MELONPIT: A finite element code for simulating tectonic mass movement and heat flow within subduction zones. Am. J. Sci., 302:381–409. Rubie, D.C. (1983): Reaction-enhanced ductility: the role of solid-solid univariant reactions in deformation of the crust and mantle. Tectonophysics. 96:331–352. Ryan, P.D. (2001): The role of deep basement during continent-continent collision: a review. In Miller, J.A., Holdsworth, R.E., Buick, I.S. & Hand, M. (eds.): Continental reactivation and reworking /Geol. Soc. Spec. Publ./, 1–55. Ryan, P.D. & Dewey, J.F. (1997): Continental eclogites and the Wilson Cycle. J. Geol. Soc. London, 154:437–442. Sleep, N.H. (1979): A thermal constraint on the duration of folding with reference to Acadian geology, New England (USA). J. Geol., 87:583–589. Smith, D.C. (1984): Coesite in clinopyroxene in the Caledonides and its implications for geodynamics. Nature, 310:641–644. Sonder, L.J. & England, P. (1986): Vertical averages of rheology of the continental lithosphere: Relation to thin sheet parameters. Earth Planet. Sci. Lett., 77:81–90. Terry, M.P. (2000): Structural and thermal evolution of Baltica basement and infolded cover nappes on Nordøyane and their bearing on mechanisms for production and exhumation of high-pressure rocks, Western Gneiss Region, Norway. Ph.D. Thesis, Univ. of Massachusetts, Boston, 198 p. Terry, M.P. & Robinson, P. (1998): Eclogite-facies structural features and their bearing on mechanisms for production and exhumation of high-pressure rocks, Western Gneiss Region, Norway, including the microdiamond-bearing gneiss at Fjørtoft. In Int. Workshop on Ultra-High-pressure Metamorphism and Exhumation; Stanford University, 74–78. Terry, M.P. & Robinson, P. (2003): Evolution of amphibolite facies structural features and boundary conditions for deformation during exhumation of high- and ultrahigh-pressure rocks, Nordøyane, Western Gneiss Region, Norway. Tectonics, 22:1036. Terry, M.P., Robinson, P. & Hamilton, M.A. (2000a): Monazite geochronology of UHP and HP metamorphism, deformation, and exhumation, Nordøyane, Western Gneiss Region, Norway. Am. Mineral., 85:1651–1680. Terry, M.P., Robinson, P. & Ravna, E.K. (2000b): Kyanite eclogite thermobarometry and evidence for thrusting of UHP over HP metamorphic rocks, Nordøyane, Western Gneiss Region, Norway. Am. Mineral., 85:1637–1650. Terzaghi, K. (1943): Theoretical soil mechanics. New York (N.Y.): Wiley, 510 p. Torsvik, T.H. (1998): Palaeozoic palaeogeography: A North Atlantic viewpoint. Geol. Fören. Stockh. Förh., 120:109–118. Upton, P., Koons, P.O. & Eberhart-Phillips, D.: Extension and strain-partitioning in an oblique subduction zone, New Zealand: Constraints from three-dimensional numerical modeling. Tectonics, in press. Wain, A. (1997): New evidence for coesite in eclogites and gneisses: Defining an ultrahigh pressure province in the western gneiss region, Norway. Geology, 25:927–930. Wain, A., Waters, D., Jephcoat, A. & Olijynk, H. (2000): The high-pressure to ultrahigh-pressure eclogite transition in the Western Gneiss Region, Norway. Eur. J. Mineral., 12:667–687. Walcott, R.I. (1998): Modes of oblique compression: Late Cenozoic tectonics of the South Island of New Zealand. Rev. Geophys., 36:1–26. Willett, S. D., Beaumont, C. & Fullsack, P. (1993): A mechanical model for the tectonics of doubly vergent compressional orogens. Geology, 21:371–374. Williams, M.L., Jercinovic, M.J. & Terry, M.P. (1999): Age mapping and dating of monazite on the electronmicroprobe: Deconvoluting multistage tectonic histories. Geology, 27:1023–1026. Zeitler, P.K., Koons, P.O., Bishop, M.P., Chanberlain, C.P., Craw, D., Edwards, M.A., Hamidullah, S., Jan, M.Q., Khan, M.A., Khattak, M.U.K., Kidd, W.S.F., Mackie, R.L., Meltzer, A.S., Park, S.K., Pecher, A., Poage, M.A., Sarker, G., Schneider, D.A., Seeber, L. & Shroder, J.F. (2001a): Crustal reworking at Nanga Parbat. Evidence for erosional focusing of crustal strain. Tectonics, 20:712–728. Zeitler, P.K., Melzer, A., Koons, P.O., Craw, D., Hallet, B., Chamberlain, C.P., Kidd, W.S.F., Park, S., Seeber, L., Bishop, M. & Schroder, J. (2001b): Erosion, Himalayan geodynamics and the geomorphology of metamorphism. GSA Today, 11:4–8.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 14, 443–466

Metamorphism and textures of dry and hydrous garnet peridotites LAURO MORTEN1* and VOLKMAR TROMMSDORFF2 1

Dipartimento di Scienze della Terra e Geologico-Ambientali, Università di Bologna, Piazza di Porta S. Donato 1, 40126 Bologna, Italy 2 Institut für Mineralogie und Petrographie, ETH-Zentrum, 8092 Zürich, Switzerland; * e-mail: [email protected] Garnet peridotite forms a large portion of the upper mantle but also occurs not uncommonly in association with crustal rocks in collisional mountain belts. The examples described here are all of the latest type, with the emphasis placed upon texture and metamorphism. All these examples are from the Central Alps and from the Eastern Alps, i.e., areas that have not been dealt with in previous chapters.

Central Alpine domain The known outcrops of garnet lherzolite in the Central Alps all occur in the same tectonic zone, namely the upper part of the Adula–Cima Lunga unit (Fig. 1). Palaeogeographically, this unit has been considered part of the former European continental margin (Schmid et al., 1990) and thus it forms the uppermost nappe complex of the lower Penninic system. Heinrich (1982, 1986) mapped mineral assemblages in mafic and pelitic lithologies of the Adula nappe and established isograds of a regional high-pressure metamorphism predating the classic, Lepontine isograd belt (Trommsdorff, 1966; Jäger et al., 1967) of the Central Alps. The peak conditions of the high-pressure metamorphism increase from north to south from 1 GPa and 500 °C to over 2–5 GPa and 800 °C, based on thermobarometry of eclogites, metapelites and meta-ophicalcite rocks (Heinrich, 1982, 1986; Partzsch, 1996; Meyre et al., 1997, 1999; Pfiffner, 1999). Garnet lherzolites were found in the highest P–T region of the Adula–Cima Lunga nappe at three localities. These are, from east to west (Fig.1): Monte Duria (MD; Fumasoli, 1974), Alpe Arami (AA; Grubenmann, 1908; Möckel, 1969), and Cima di Gagnone (CdG; Evans & Trommsdorff, 1978). At all three localities, garnet lherzolites show a more or less pronounced layering determined by variations in olivine and pyroxene contents. At MD, garnet lherzolites occur in over 20 individual ultramafic bodies, which form 10–100 m boudins within migmatitic gneisses. Garnet lherzolites and gneisses are folded around a steeply plunging megafold having an amplitude of at least 4 km. At AA, garnet lherzolites form the core of a 1 km × 400 m chlorite peridotite boudin surrounded by steeply south-dipping migmatitic gneisses. A discontinuous layer of eclogitic rocks separates the peridotite body from the country gneisses. At CdG, garnet

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

444

L. Morten & V. Trommsdorff Fig. 1. Eocene high pressure metamorphism in the Adula–Cima Lunga Nappe (modified after Frese et al., 2003). Quartz eclogites (shaded dashed isograds and values in squares) indicate an increase in P–T from north to south (Heinrich, 1982). Eclogites and garnet peridotites show corresponding values around 800 °C and 3 GPa; some values for garnet peridotites (dots) are given after Nimis & Trommsdorff (2001). The isograds of the later amphibolite facies regional metamorphism (Oligocene decompression) crosscut the nappe boundaries.

lherzolite occurs in one of numerous ultramafic lenses of tens to hundreds of metres in size, surrounded by pelitic and semipelitic, migmatitic, gneisses, kyanite eclogites, marbles, and metaophicalcite rocks (Pfiffner & Trommsdorff, 1998). The peridotites show transitions to eclogite. Within the ultramafic rocks, the presence of metarodingite boudins, interpreted as former midocean ridge basalt (MORB) dykes, testifies to an early serpentinite stage of the metaperidotites (Evans et al., 1979, 1981). The ultramafic–mafic–carbonate suite at CdG has been interpreted as derived from an ocean basin near a continental margin (Pfiffner & Trommsdorff, 1998), with the ultramafic rocks representing former subcontinental mantle that had been exhumed during oceanic rifting. The chemical compositions of the garnet lherzolites from the Central Alps (O’Hara & Mercy, 1966; Fumasoli, 1974; Rost et al., 1974; Ernst, 1978; Evans & Trommsdorff, 1978; Pfiffner, 1999) are all remarkably similar and close to that of fertile mantle, which is typical for the subcontinental lithosphere in the Alpine realm and Liguria (Nicolas & Jackson, 1972; Piccardo et al., 1990; Menzies & Dupuy, 1991; Müntener, 1997). Isotope geochemical investigations of minerals from AA and CdG garnet lherzolites yielded consistent, Eocene ages (~ 40Ma), based on garnet–clinopyroxene–whole rock Sm–Nd isochrons (Becker, 1993) and of 43–35 Ma, based on the U–Pb method and SHRIMP analysis of zircons (Gebauer et al., 1992; Gebauer, 1996). These ages are consistent with the recent dating of Late Eocene high-pressure metamorphism in the Western Alps

Metamorphism and textures of dry and hydrous garnet peridotites

445

(Froitzheim et al., 1996; Gebauer et al., 1997). In agreement with the prograde metamorphism of the eclogite sequence mapped in the Adula–Cima Lunga unit by Heinrich (1982, 1986), most of the garnet peridotites from the Central Alps show evidence of prograde metamorphism, as documented by Evans & Trommsdorff (1978) for CdG. The garnet peridotite at CDG was interpreted by these authors, for the first time, as subduction zone garnet peridotite. Poikiloblastic and porphyroclastic garnet peridotites in the Central Alps Detailed petrographic descriptions of garnet lherzolites from MD, AA, and CdG have been given by Fumasoli (1974), Möckel (1969) and Evans & Trommsdorff (1978), respectively. Additional information regarding the textures is given here. The garnet lherzolites in the Central Alps exhibit two main textural types: poikiloblastic and porphyroclastic. The poikiloblastic garnet peridotites (Fig. 2b) occur at CdG and, locally, at MD. The poikiloblastic garnet peridotite is weakly foliated, sometimes folded, with completely anhedral, often elongated garnet crystals concentrated along pyroxene-rich layers. Poikiloblastic garnet is often stuffed with abundant, sometimes folded inclusions of orthopyroxene and, less commonly, olivine, clinopyroxene and rare, pale brownish to greenish, rounded, magnesian Ca-amphibole. Olivine occurs in two generations: the older generation (olivine I) is by far major in quantity and was formed during highpressure (HP) dehydration of a hydrous protolith, whereas the younger generation (olivine II) is minor and was formed during post-HP partial recrystallisation. Olivine I consists of up to 2 mm large, elongated grains containing numerous rod-shaped microinclusions of ilmenite (Fig. 2c). The rods are elongated parallel to [010] of olivine and form palisades aligned within the (001) plane of olivine (Risold et al., 2001). Especially along more pyroxenitic layers, pseudomorphs of olivine with inclusions of wormy ilmenite after titanian clinohumite are commonly found, with occasional relics of (OH)titanian clinohumite (Evans & Trommsdorff, 1978). The breakdown of titanian clinohumite and the presence of planar OH defects with ilmenite palisades in olivine I indicate prograde metamorphism at high H2O activity (Trommsdorff et al., 2001). In addition to olivine, the matrix around the garnets is formed by an idioblastic mosaic of orthopyroxene, clinopyroxene and Ca-amphibole with a typical grain size of 0.5–1 mm. At CdG, garnet has overgrown pre-existing, isoclinal folds involving all the matrix minerals, including Ca-amphibole and olivine + ilmenite pseudomorphs after titanian clinohumite (Pfiffner & Trommsdorff, 1998) (Fig. 3). Ca-amphibole grains enclosed in garnet have a distinctly higher K2O content (0.7 wt%) than those in the matrix (0.15 wt%). Thus there is a large body of evidence that, in the poikiloblastic type peridotites, garnet formed during prograde metamorphism. Evans & Trommsdorff (1978) interpreted the poikiloblastic peridotites from CdG as “subduction zone garnet peridotites” derived from a partially serpentinised, hydrous protolith. At CdG, prograde chlorite-amphibole peridotites are commonly found to be isofacial with the garnet peridotites. The simultaneous occurrence of both rock types can be explained in terms of variable bulk composition and/or H2O activity (Trommsdorff, 1990; Pfiffner, 1999).

446

L. Morten & V. Trommsdorff

Fig. 2. Photomicrographs of HP metaperidotites from the Central Alps, after Frese et al. (2003). Lineations and trace foliation are parallel to the long edge of the images. Crossed polarisers. All mineral abbreviations after Kretz (1983). (a) Cima di Gagnone garnet peridotite (sample Mg160Tro). The foliation is defined by elongated and slightly flattened grains of olivine, enstatite and garnet. (b) Prograde poikiloblastic pyrope-rich garnet (dark shading) with inclusions of enstatite, olivine and amphibole from Cima di Gagnone (sample Mg160-48) overgrowing an older foliation. (c) To the left: oriented inclusions of ilmenite rods elongated parallel to [010] olivine and forming palisades parallel to (001) of olivine (Mg160-98-1). To the right: schematic representation of the crystallographic orientation of the rods and palisades after Risold et al. (2001). (d) Porphyroclastic texture with two generations of olivine from Alpe Arami. The older olivine I forms large, mm-sized grains, surrounded by a mortar textured matrix of recrystallised olivine II. (e) Poikiloblastic pyrope-rich garnet (dark shading) from Alpe Arami (sample AA4; courtesy of P. Nimis).

Metamorphism and textures of dry and hydrous garnet peridotites

447

Fig. 3. Photomicrograph of garnet peridotite from CdG. Pokiloblastic garnet (grt) overgrows a preexisting microfold with diopside (dio), enstatite (en) and pargasitic amphibole (amph).

The second textural type, a porphyroclastic garnet lherzolite, dominates at AA. In comparison with CdG, the AA garnet peridotite is composed of a four-phase assemblage (garnet, olivine, enstatite, diopside) which has been described in detail by Möckel (1969). Its microtexture is characterised by more or less equant, anhedral porphyroblasts of pyrope-rich garnet in a porphyroclastic matrix with a mortar texture (Fig. 2d). Garnet grew, most probably, poikiloblastically (Fig. 2e): its present, often porphyroclastic texture was induced by a late deformation. Garnet is mostly rounded, with a grain size up to 10 mm. Olivine, enstatite, clinopyroxene and, rarely, chromian spinel inclusions in garnet (Möckel, 1969; Nimis & Trommsdorff, 2001) indicate that garnet growth was prograde. Chromian diopside, up to 5 mm in diameter, is frequently concentrated near garnet. The matrix around garnet is composed of two generations. A porphyroclastic generation of olivine (olivine I), consists of large grains up to 2 mm in size, showing deformation-induced undulating extinction, kinks and subgrains, and containing rodshaped ilmenite inclusions. A second generation (olivine II) displays small, subhedral grains forming the mortar texture between the porphyroclasts. The latter generation recrystallised during deformation along the Southern Steep Belt of the Central Alps (Trommsdorff et al., 2000). Similarly, enstatite and diopside form two generations (Paquin & Altherr, 2001). A poikiloblastic microtexture (Fig. 2e), analogous to CdG, has been reported by Möckel (1969, sample G01-A) in one titanian clinohumite-bearing garnet peridotite block near AA: this texture also rarely occurs in the main AA body (Nimis, pers. comm.) (Fig. 2e). Recent detailed mapping (Bay, 1999) has confirmed that

448

L. Morten & V. Trommsdorff

the locality of Alpe Stuello, where this block was sampled, represents remnants of a moraine deposit derived from the AA ultramafic body (see also Geological Map of Switzerland 1:25000, sheet 1313, Bellinzona; Bächlin et al., 1974). Oriented ilmenite rods in olivine The principal argument for an ultradeep origin of the AA rocks (Dobrzhinetskaya et al., 1996) was based upon the occurrence of topotactic FeTiO3 rods in olivine, originally described by Möckel (1969) as brown rutile needles, oriented parallel to [010]ol. On the basis of TEM electron diffraction patterns, Dobrzhinetskaya et al. (1996) re-determined the needles as ilmenite and three additional, previously unknown, polymorphs of FeTiO3 which they interpreted as intermediate between the high-pressure phase FeTi-perovskite (Leinenweber et al., 1991) and ilmenite. By means of optical measurements Dobrzhinetskaya et al. (1996) determined the bulk quantities of 1–3 vol% FeTiO3 rods in olivine corresponding to about 0.7 to over 2.0 wt% TiO2. These numbers have later been revised to lower values ranging from 0.1–0.9 vol% FeTiO3 (0.07 to 0.6 wt% TiO2 in olivine) with a mean of 0.53 vol% (Green et al., 1997). By comparison, the highest reported TiO2 content in natural olivine (Hervig et al., 1986) does not exceed 0.05 wt% (500 ppm). Maximum TiO2 solubility in olivine from experiments at oxygen fugacities equivalent to mantle conditions at 10 GPa/1400 °C (Ulmer et al., 1998) and 14 GPa/1600 °C (Gudfinnsson & Wood, 1998) yield 0.13 wt% and 0.11 wt% TiO2 in olivine, respectively. By contrast, Dobrzhinetskaya et al. (1996) determined, at uncontrolled oxygen fugacity, considerably higher solubility values, i.e. > 1 wt%, at 12 GPa/1700 K. Based on the very high Ti content and the determination of unknown polymorphs, Dobrzhinetskaya et al. (1996) inferred that the titanate rods in AA olivine exsolved as FeTiO3 perovskite during the polymorphic transformation of wadsleyite, which has a higher Ti solubility (Gudfinnsson & Wood, 1998), to olivine. The existence of the three new FeTiO3 polymorphs has been questioned by Risold et al. (1997) on the basis of theoretical considerations and by Hacker et al. (1997) and Risold et al. (1997) because the published diffraction patterns can be explained by dynamical diffraction between ilmenite and the olivine matrix. The bulk quantities of TiO2 in Arami olivine published by Dobrzhinetskaya et al. (1996) have been questioned by Hacker et al. (1997) on the basis of EPMA measurements by Risold et al. (1996) and Reusser et al. (1998) on the basis of UV-laser ablation ICP-MS: all these authors agree on a TiO2 content in olivine of approximately 350 ppm (0.035 wt%). These results are still 2 to 20 times lower than the revised values determined optically by Green et al. (1997). Ilmenite rods in olivine, identical in size, composition and quantity to those of Alpe Arami, have been detected in over ten peridotite localities of the Adula–Cima Lunga nappe (Fig. 1); among these are CdG and MD. Additional occurrences of ilmenite rods in olivine have been reported from the Nonsberg region in the Eastern Alps (Risold et al., 1999) and in the Sulu terrane, China (Hacker et al., 1997). If the ilmenite exsolutions are indicative of an ultradeep origin, the geological evidence for prograde metamorphism mentioned above (see Nimis & Trommsdorff,

Metamorphism and textures of dry and hydrous garnet peridotites

449

2001) would require that Central Alpine peridotites had to be subducted to more than 300 km and then exhumed again within a single collision cycle. However, based on new optical microscope and TEM observations, a mechanism has been proposed for the formation of the FeTiO3 rods in olivine that does not require mantle transition zone pressure, nor ultrahigh pressure (Risold et al., 2001): TEM investigation of the rods at AA, CdG and MD demonstrates that the titanate inclusions correspond to a Mg-bearing ilmenite with exclusively a trigonal crystal structure. No other FeTiO3 polymorphs have been identified. The shape of the ilmenite inclusions concurs with minimum topotactic misfits between the lattices of olivine and Mg-bearing ilmenite (Risold et al., 2003). Optical microscopy reveals that the ilmenite rods are not randomly distributed within olivine but have, at all localities, a preferred alignment parallel to (001)ol; Figure 2c, (Risold et al., 2001). TEM images of the ilmenite palisades at AA and CdG show that they are controlled by the presence of (001) planar defects in olivine along which the rods nucleated. The defects have been characterised by high resolution TEM and identified as 4.4 Å wide humite-type layers, i.e., (Mg2SiO4)·Mg(OH,F)2, intergrown within olivine. In such layers substitution up to 50 mol% MgH2 by Ti4+ is possible. Assuming Ti-saturated humite slabs, the maximum density of (001) faults observed at AA, approximately 1900 mm–1, can accommodate about 170 ppm TiO2 in olivine. This represents about half the values obtained by UV-laser ablation ICP-MS and EPMA analysis. Evidence for the annealing of the defects, however, suggests that only parts of the defects are still preserved. The formation of ilmenite rods is considered to be a consequence of the breakdown of isolated humite layers in olivine. The apparent TiO2 content in olivine at AA, CdG and MD is regarded as a function of the Ti-humite defect density. Lattice preferred orientation (LPO) of olivine from AA and CdG Orogenic peridotites (Den Tex, 1969) play a fundamental role in the understanding of the composition, geophysical properties and geodynamic processes of the Earth’s upper mantle. Forsterite-rich olivine dominates mantle composition and, being a relatively weak mineral, controls mantle rheology (e.g. Drury & Fitzgerald, 1998). During convection in the asthenosphere, deformation by intracrystalline slip, accompanied by dynamic recrystallisation, produces lattice preferred orientation (LPO) of olivine. Crystal LPO, with slip planes and slip directions parallel to the shear plane and shear direction, respectively, leads to an optimisation of the resolved shear stress on active slip systems, and thus to geometric softening of the aggregate. Because of a large elastic anisotropy of olivine crystals, the olivine LPO is the primary cause of seismic anisotropy in peridotites. The relationships between flow geometry, LPO and seismic anisotropy in olivine polycrystals have been repeatedly discussed for naturally and experimentally deformed aggregates (e.g. Nicolas & Christensen, 1987; Tommasi et al., 2000; Bystricky et al., 2000). Most published LPOs of olivine were obtained from ophiolites or xenoliths. They generally show strong preferred orientations of (010) or (0kl) parallel to foliation, and [100] or [u0w] parallel to lineation (Ben Ismail & Mainprice, 1998). This is in good

450

L. Morten & V. Trommsdorff

agreement with slip systems active at high temperature and low strain rates with the primary Burger vector [100] (e.g. Nicolas & Poirier, 1976). These LPOs have been considered to be inherited mantle fabrics, although the rocks may have suffered structural and chemical changes during tectonic and metamorphic activity. In the following, textures of the type (010)[100] will be referred to as mantle LPO. Numerous occurrences of peridotites are known in the Alpine orogen. They contain different structural and metamorphic records due to their different palaeogeographic settings and thus different geodynamic histories. Many of these rocks in the Western, Central and Ligurian Alps are not ophiolite sensu stricto, because they are lherzolites of subcontinental origin (Trommsdorff et al., 1993). Petrological and structural work was carried out on several of these occurrences with LPO data available (e.g. Boudier, 1978; Hoogerduijn Strating, 1991). Some of the investigated peridotites match olivine “mantle LPOs”, as they are ex-mantle fragments without significant Alpine overprint. Other peridotites in the Alps underwent subduction and a metamorphic overprint to high pressure and temperature. They now form part of the Alpine nappe system. These peridotites with an Alpine metamorphic signature display LPOs different from the mantle LPO outlined above (Frese et al., 2003). The perhaps best known peridotite with an Alpine HP metamorphism is the garnet peridotite of Alpe Arami. Its olivine LPO is known since over 30 years (Möckel, 1969) and is exceptional, with olivine a axes, i.e. [100], oriented subnormal to foliation and c axes, [001], parallel to lineation within the foliation plane. This fact has led Dobrzhinetskaya et al. (1996) to conclude that the Arami LPO is yet another characteristic of extraordinary high pressure as had been inferred from the occurrence of ilmenite rods in olivine. The LPOs of olivine from the unaltered garnet lherzolite at CdG and AA, however, show consistent patterns (Fig. 4a–c). The olivine [100] axes are distributed along a partial girdle normal to the lineation, with the highest concentrations subnormal to the foliation. The olivine [010] pole figures have a maximum near the foliation plane away from the lineation, and olivine [001] are concentrated in a point maximum close to the lineation. The first eigenvector for [100] plots nearly normal to the foliation plane and the first eigenvector for [001] plots nearly parallel to the lineation. The first eigenvalues for [100], [010] and [001] are generally twice as high as the second and third, which are about equal, indicating point-like distributions with a major random component. The orientation distribution of enstatite (Fig. 4d) at CdG (Mg160Tro) shows strong maxima of [100] normal to the foliation, of [010] normal to lineation within the foliation plane, and of [001] parallel to lineation. The first eigenvector for enstatite [100] plots nearly normal to the foliation plane, and the one for [001] lies parallel to lineation. The first eigenvalues for all three pole figures are three times higher than the second and third, which are similar. All this indicates strong point maximum distributions, and a (100)[001] preferred orientation for enstatite as well as for olivine. The enstatite LPO is very sharp whereas the LPO of olivine is less distinct, but all distributions possess approximately orthorhombic sample symmetry. The LPO of the first olivine generation from Alpe Arami (Fig. 4a) determined in this study confirms the results of Möckel (1969), with preferred orientation of olivine

Metamorphism and textures of dry and hydrous garnet peridotites

451

Fig. 4. Lattice preferred orientation for (a) olivine porphyroclasts from the AA garnet lherzolite, (b, c) olivine in two samples from the CdG garnet lherzolite, and (d) enstatite porphyroblasts from the CdG garnet lherzolite. Lower hemisphere equal area projection. Normal to foliation plane is vertical, and lineation is horizontal. All measuraments done by EBSD. The texture index pfJ, the minimum and maximum density is in the contoured pole figures, the eigenvalues (E1, E2, E3) and eigenvectors (5,„,z) of the corresponding orientation ellipsoid, and the number of data points per sample are given. After Frese et al. (2003).

452

L. Morten & V. Trommsdorff

[100] subnormal to the foliation. The significant new result is that CdG olivine I grains and AA olivine porphyroclasts have a similar LPO. Because the garnet peridotite at CdG formed its LPO at pressures around 3 GPa, the LPO at AA cannot be taken as argument for higher pressures. Criteria determining the deformation conditions The LPO of olivine in mantle peridotites is typically characterised by maxima of the [010] and [001] axes perpendicular to foliation, and of the [100] axes parallel to lineation. These fabrics originate from glide and dynamic recrystallisation with the dominant slip systems (010)[100] and (0kl)[100]. The (010)[100] system was experimentally determined for dry deformation at high temperatures (> 1200 °C; Nicolas et al., 1973). Recently Jung & Karato (2001) produced several new olivine fabrics in experiments under wet conditions in nearly simple shear deformation (ca. 2 GPa/1200 °C). Their “type C” fabric is characterised by an oblique pattern with concentrations of [001] 20° off the shear direction, and of the (100) plane 20° off the shear plane. This fabric was produced at high H2O activity and modest shear stress (~ 250 MPa). None of the investigated fresh garnet lherzolites of the Central Alps from CdG and AA show LPOs for olivine with [100] maxima parallel to the lineation, but mostly with [100] subnormal to foliation and [001] towards lineation. These LPOs are similar to the experimental “type C” fabric presented by Jung & Karato (2001). Their LPO suggests “easy slip” on the (100)[001] slip system in simple shear deformation. This system was observed to be dominant under high stress at temperatures below 900 °C in dry experiments (Carter & Avé Lallemant, 1970; Phakey et al., 1972). Jung & Karato (2001) argued that slip along [001] may be similarly enhanced by water at higher temperature as by high stress under dry conditions. A qualitative comparison of the LPOs measured at CdG and AA with the experimentally produced LPOs suggests deformation for AA and CdG with similar mechanisms being activated. The LPO of enstatite in the CdG peridotite has similar maxima and is much stronger than that of olivine. This is in good agreement with LPO work done on the chlorite-enstatite-olivine schists from Val Cama, Central Alps (Trommsdorff & Evans, 1969). This LPO pattern is consistent with crystal plastic slip on (100)[001], the dominant slip system of enstatite (Muegge, 1898; Raleigh, 1965; Mackwell, 1991). The LPO is dominated by an “easy slip” orientation for this slip system with respect to foliation and lineation in a simple shear regime. Based upon the distribution of the ilmenite rods within olivine I at CdG the following may also be considered: ilmenite palisades occur within olivine (001) planes (Risold et al. 2001), and the preferred orientation of their poles is therefore identical to the LPO of [001] of olivine. [001] of olivine is oriented preferentially parallel to the lineation within the foliation plane, with the rods aligned within a plane normal to the lineation. The olivine grains hosting the ilmenite palisades display shape and lattice preferred orientation along [001], with a mineral elongation which marks a macroscopically visible lineation. Normal to lineation, late extensional cracks normal to

Metamorphism and textures of dry and hydrous garnet peridotites

453

[001] of olivine are observed. These cracks have given access to fluid migration and to the formation of fluid inclusion trails parallel to the (001) plane of olivine. These extensional cracks are interpreted as having formed from the release of residual stress parallel to the lineation. The mineral growth must have occurred under glide active on (100)[001] of olivine, which is considered the reason for the observed fabric.

Eastern Alpine domain The Ulten zone (Andreatta, 1935; 1948; Hoinkes & Thöni, 1993) is a high-grade basement unit of the Italian Eastern Alps which is composed of migmatites, Grt + Ky ± Sil (fibrolitic) gneisses, lenses of amphibolitised eclogites and slices of Spl/Grt peridotites (mineral abbreviations after Kretz, 1983). It has been classically attributed to the Upper Austroalpine system of the Central Eastern Alps (Fig. 5), a nappe pile of Cretaceous age (Thöni, 1981). The Upper Austroalpine system consists of metasedimentary cover and upper-to-

Fig. 5. Geological sketch and main tectonic lineaments of the Nonsberg area (after Obata & Morten, 1987, and Nimis & Morten, 2000). (1) Bt-Ms paragneiss; (2) Grt-Ky paragneiss; (3) migmatite and orthogneiss; (4) siltstone, sandstone and conglomerate of the Insubric Flysch (Upper Cretaceous); (5) Triassic dolomite; (6) ultramafic rocks; (7) amphibolite; (8) tonalite; (9) marble.

454

L. Morten & V. Trommsdorff

lower crust slices derived from the Mesozoic passive margin of the Adria microplate (Dal Piaz, 1993). According to Flügel (1990) and Neubauer & von Raumer (1993), the Austroalpine system of the Eastern Alps is a piece of the Variscan belt. Its hypothetical evolution history, as pointed out by various authors, can be briefly summarised, according to Godard et al. (1996), as follows. (i) Ocean closure before Devonian and collision of continental and arc elements during the Early Carboniferous (Neubauer & von Raumer, 1993) produced the Central European part of the Variscan fold belt (Neugebauer, 1990), possibly enclosing minor elements of pre-Variscan age (Ziegler, 1984, 1993; Maggetti & Flisch, 1993; von Raumer & Neubauer, 1993; Neubauer & von Raumer, 1993). (ii) This belt was dismembered and uplifted during Late Carboniferous/Early Permian times along transtensive fault systems that transected the orogen (Ziegler, 1984). The Variscan basement was further unroofed and dismembered by Permian and Mesozoic lithospheric thinning and rifting along normal faults (Eberli, 1988; Dal Piaz, 1993). (iii) The Late Triassic–Late Jurassic rifting and drifting stages created the Adria microplate (Africa), in relation to the opening of the Tethys basin between Africa and Europe. The inversion of these normal faults during the Cretaceous (Eo-Alpine) orogeny led to the thrusting of the Austroalpine system (i.e., the Adria microplate) northward and northwestward, over the colliding Penninic domain of the European plate margin (Froitzheim & Eberli, 1990). In northwestern Trentino, the Austroalpine system is separated from the Southern Alps domain by the major Tertiary Periadriatic (Insubric) lineament, which is composed of the strike-slip to transpressive Tonale and Giudicarie faults (Fig. 5). North of these faults, the Austroalpine system is divided into the Tonale and Ortler nappes (Thöni, 1981) by the Peio line (Andreatta, 1948), a Cretaceous thrust which was folded and reactivated during the Tertiary (Martin et al., 1991). The northern cover-bearing Ortler nappe is overlain along the Peio thrust by the cover-free Tonale nappe, which includes the Tonale and Ulten zones (“series”, Andreatta, 1948). The pre-Alpine basement of the Ortler nappe consists largely of Grt + St ± Sil micaschists with minor amphibolites of tholeiitic composition, calc-alkaline metagranitoid, marbles and serpentinite slivers. The Grt + Sil paragneisses in the Tonale zone of the Tonale nappe include similar marbles and mafic–ultramafic intercalations, although they display a higher-grade pre-Alpine metamorphic overprint. This lithological assemblage partly represents a disrupted ophiolitic suite within the Variscan basement (Martin & Prosser, 1993). A back-arc basin environment may be envisaged in comparison to the Speik Complex (Eastern Austroalpine system; Neubauer, 1988; Neubauer & Frisch, 1993). The basement was retrogressed during the late- or postVariscan exhumation and was locally deformed under greenschist facies conditions during the Alpine orogeny (Thöni, 1981). By contrast, the Ulten zone of the Tonale nappe is mainly composed of kyanitebearing micaschists, migmatites and gneisses with well-preserved pre-Alpine highgrade metamorphic signatures, weakly overprinted by the Alpine metamorphism. It is bordered by two belts of sillimanite-bearing paragneisses, which represent the eastern continuation of the Tonale zone (Fig. 5). The Ulten zone is divided from the Bt - Ms ± Sil metasedimentary rocks of the Tonale zone by the Rumo line (Morten et al., 1976);

Metamorphism and textures of dry and hydrous garnet peridotites

455

the higher-grade rocks of the Ulten zone are located north of the Rumo line (Nonsberg area) and consist of migmatites, Grt-Ky gneisses, with boudins of amphibolitised eclogites and Spl/Grt peridotites (Fig. 5). Nonsberg ultramafic rocks The Nonsberg ultramafic rocks mainly consist of barrel-shaped bodies tens to hundreds of metres in size, which are generally located along the boundary between the underlying strongly-foliated gneisses and the overlying migmatites and orthogneisses (Fig. 5). They are predominantly peridotites of harzburgitic to lherzolitic composition (Bondi et al., 1992). Compositional banding, primarily due to modal variation of pyroxene is conspicuous in some outcrops. Layers or bands of pyroxenite, some of which are complexly folded, also occur (Morten & Obata, 1983). On a textural and grain size basis two petrographic types have been distinguished (Obata & Morten, 1987): coarse type and fine type. The coarse type rocks are coarsegrained (up to few centimetres in grain size) and are relatively undeformed. The fine type rocks are of much finer grain size (0.2–1 mm) and show various metamorphic textures. In the Nonsberg area, the fine type prevails in abundance over the coarse type. In some ultramafic bodies, the coarse type sporadically occurs in the more typical fine-grained matrix. The field and petrographic observations of many typical and transitional rock types have indicated that the fine types have been derived from the coarse types by syntectonic recrystallisation and that the Grt-bearing rocks derived from previous Splbearing peridotites (Obata & Morten, 1987). Various hypotheses have been put forth to explain the transformation from spinel peridotites to garnet peridotites and to depict the evolution of the Nonsberg ultramafic rocks. Proposed scenarios include: (i) upwelling of deep mantle material and partial melting in the spinel lherzolite field followed by isobaric cooling and then tectonic emplacement into the crust (Herzberg et al., 1977); (ii) high-pressure metamorphism after emplacement of spinel lherzolite into the crust (Rost & Brenneis, 1978); (iii) tectonic emplacement of spinel lherzolite into the crust followed by “burial” into deeper levels (Rost et al., 1979); (iv) emplacement of high-temperature spinel lherzolite into the lower crust followed by conductive cooling of the body essentially at the same depth (Obata & Morten, 1987); (v) emplacement of spinel lherzolite bodies into the crust and then sinking through the less dense, ductile, sialic lower crust (Obata & Morten, 1985); (vi) emplacement into the crust of sliced fragments of a mantle wedge overlying a subducting slab in a collisional environment (Godard et al., 1996). More recently, on the basis of P–T estimates, the inferred P–T path and physical models for subduction zones, Nimis & Morten (2000) proposed the following scenario. High temperatures, exceeding 1400 °C, were attained in the innermost portion of a mantle wedge overlying a subducting continental slab. Hydrous melts were produced within these hot mantle portions and rose into the overlying, lower-T spinel lherzolites (~ 1200 °C, 1.31–1.58 GPa), where the rising melts produced high-T (> 1400 °C) pyroxenitic segregates. Convection induced by the movement of the subducting plate caused wedge peridotites to cool while flowing towards the slab at essentially constant depth (isobaric cooling path). The mantle material was then driven to greater depth by

456

L. Morten & V. Trommsdorff

the downward flow near the wedge–slab interface (T decrease and P increase path) and was eventually incorporated in the crust as tectonic slices or as sinking blobs. Mantle flow and entrainment in the crust caused the peridotites to cool down to ~ 850 °C before, or while, being subducted together with the slab to a depth of about 90 km. The latest subduction stage, which led to the formation of garnet, is represented by a nearly isothermal path. The transformation from spinel peridotite to garnet peridotite (from coarse type to transitional and fine type) is accompanied by significant input of crustal metasomatic agents (Morten & Obata, 1990) as indicated by the crystallisation of abundant, up to 23 vol%, amphibole with crustal geochemical signatures (Rampone & Morten, 2001). During the subsequent exhumation history, peridotites and continental crust shared the same decompressional and cooling path, which is testified by mineral zoning patterns in ultramafic rocks and is best recorded by mafic and pelitic country rocks, as well as by kelyphite minerals after peridotitic garnets (Godard et al., 1996). Textures of the Nonsberg peridotites Coarse type Most of the coarse type peridotites are spinel lherzolite with protogranular texture (Fig. 6a,b) similar to that observed in many ultramafic xenoliths in alkali basalts (Mercier & Nicolas, 1975). The constituent minerals are olivine (up to 5 mm grain size), enstatite (2–4 mm), diopside (0.7–2 mm), and Cr-Al spinel (0.6–1 mm). In some samples Ca-amphibole (actinolitic) occurs along pyroxene grain boundaries and locally in finegrained recrystallised parts. Large grains of olivine and enstatite show undulating extinction and kink band textures. Enstatite contains exsolution lamellae of diopside and spinel, and diopside contains lamellae of enstatite and spinel. Spinel occurs as large independent grains, as well as exsolutions in pyroxenes, and they are typically intergrown with enstatite (Fig. 6c). In a few coarse type samples, garnet occurs in addition to spinel. The texture is transitional from protogranular to porphyroclastic, where large grains of olivine and pyroxene appear to be more strained than those in normal spinel lherzolites and the smaller, unstrained grains (neoblasts, approximately 0.3 mm grain size) are more abundant. Garnet occurs surrounding spinel (Fig. 7a) and also forms 10–50 µm thick exsolution lamellae, parallel to (100) of the host (Fig. 7b), in both enstatite and diopside. In addition to the garnet and pyroxene lamellae, the presence of very thin (less than 1 µm thick) lamellae of Cr-spinel was confirmed in enstatite parallel to (100) of the host by transmission electron microscopy (Obata & Morten, 1987). Some large diopside grains contain Ca-amphibole lamellae, whose composition is identical to the neoblasts in the same sample (Fig. 7c). The large spinel grains are always mantled by garnet; they are brown to reddish brown and amoeboid to holly leaf shaped, with a maximum grain size of 2 mm. The smaller the size, the darker the color and more chromian the composition. In finegrained, recrystallised parts of the thin section, the garnet lherzolite assemblage, i.e., olivine + enstatite + diopside + garnet, occurs together with hornblende, but without spinel. It is evident from petrographic observations reported above that the spinel-garnet

Metamorphism and textures of dry and hydrous garnet peridotites

457

Fig. 6. (a) Protogranular texture of coarse type Spl-lherzolite. Both pyroxenes exsolve lamellae of Ca-poor and Ca-rich pyroxene, respectively. Crossed polarisers. The field of view is 3 mm. (b) Protogranular texture of coarse type Spl-lherzolite. The clinopyroxene (left bottom corner) carries exsolution lamellae of Ca-poor pyroxene and spinel. Crossed polarisers. Width of view is 2 mm. (c) Coarse type Spl-lherzolite. Spinel intergrowths with enstatite. Plane polarised light. Field of view is 1.5 mm.

458

L. Morten & V. Trommsdorff

Fig. 7. (a) Coarse type Spl-Grt peridotite. Garnet porphyroblast with an inclusion of corroded Cr-spinel. Partly crossed polarisers. Field of view is 2 mm. (b) Coarse type Spl-Grt peridotite. Both clinopyroxene (left) and orthopyroxene (right) contain exsolution lamellae of garnet. Crossed polarisers. Field of view is 2 mm. (c) Coarse type Spl-Grt peridotite. Clinopyroxene with exsolution lamellae of orthopyroxene, garnet and amphibole. Crossed polarisers. Field of view is 2 mm.

Metamorphism and textures of dry and hydrous garnet peridotites

459

lherzolite was originally a protogranular spinel lherzolite that has been partly transformed to garnet lherzolite. Fine type As mentioned above, the fine type is more abundant in this region. It is finer grained and the texture is porphyroclastic to tabular or mosaic equigranular. It contains more varieties of mineral assemblages than the coarse type. Observed mineral assemblages are summarised as follows (Obata & Morten, 1987; Morten, 1993): Ol + Opx + Cpx + Grt + Ca-amph ± Spl, Ol + Opx + Grt + Ca-amph + Spl, Ol + Opx + Ca-amph + Spl, Ol + Opx + Ca-amph + Chl. Pyrope-rich garnet is the stable phase in the first two groups. Clinopyroxene is minor or absent. Ca-amphibole, containing up to 1 wt% K2O (Obata & Morten, 1987; Rampone & Morten, 2001) is an important hydrous, Ca-bearing phase. It was considered (Obata & Morten, 1987) to grow according to the clinopyroxene and garnet consuming reactions (Obata & Thompson, 1981): Cpx + Opx + Grt +H2O l Ca-amph + Ol and, under even higher partial pressure of H2O, Grt + Ol + H2O l Ca-amph + Opx + Spl. No recent data exist of LPO of olivine from the Nonsberg peridotites. However Andreatta (1934) measured with a U-stage the LPO of olivine from both granulartextured spinel and garnet peridotites. In both cases the LPO resulted in [010] normal to foliation, [001] and [100] parallel to foliation, i.e., the mantle LPO (see above). In porphyroclastic rocks, porphyroclasts of olivine, enstatite, and diopside are up to 5 mm in size and are more or less elongated and show signs of deformation such as undulating extinction, kink band boundaries and subgrains (Fig. 8a). Kinked orthopyroxene porphyroclasts contain numerous exsolution lamellae of garnet and clinopyroxene (Fig. 8b and c), and the clinopyroxene porphyroclasts contain those of orthopyroxene. Olivine neoblasts show granular polygonal texture (Fig. 8a). Pyroxene neoblasts are unstrained and contain no exsolution lamellae. Spinel inclusions are commonly observed in olivine neoblasts, while they are very rare in the olivine porphyroclasts. The amphibole is colorless or pale green to brownish green. It is typically zoned, being more tremolite-rich at the grain margins, where compositional change is so abrupt that sometimes Becke lines are visible within the grains. The amphibole grains typically include small, irregular shaped spinel (Fig. 9a) and, in few cases, thin lamellae of phlogopite occur in some amphibole grains. Garnet occurs as large porphyroclastic crystals (up to 2–3 mm size) and also as small grains which are in textural equilibrium with olivine, pyroxenes, amphibole, and perhaps with spinel. Large garnet crystals typically include vermicular grains of Crspinel, which are partly armoured by olivine (Fig. 9b). Spinel inclusions are less common in small garnet but olivine, pyroxene and, sometimes, carbonate (dolomite) inclusions are commonly observed.

460

L. Morten & V. Trommsdorff Fig. 8. (a) Porphyroclastic texture of Grt-Spl peridotite. Large kinked olivine porphyroclast within neoblastic, granular polygonal matrix. Crossed nicols. The field of view is 2.5 mm. (b) Porphyroclastic texture of Spl-Grt peridotite. The highly strained orthopyroxene porphyroclast contains garnet lamellae. Crossed polarisers. The field of view is 3 mm. (c) Porphyroclastic texture of Spl-Grt peridotite. The kinked orthopyroxene porphyroclast contains garnet and clinopyroxene lamellae. Small clinopyroxene grains crystallised along the kink planes. Crossed polarisers. Field of view is 0.6 mm.

Chrome spinel is a common accessory phase and it is dispersed as small (less than 0.2 mm in diameter) translucent brown grains. Although the spinel included in the garnet may be a relic of earlier assemblages, these small grains dispersed among the olivine and pyroxene matrix are considered to be in equilibrium with garnet (Obata & Morten, 1987). In the chlorite-bearing assemblage, chlorite appears to be in textural equilibrium with olivine, orthopyroxene and amphibole, while spinel is nearly opaque and rare: this indicates that in this assemblage chlorite, instead of spinel, is the stable aluminous phase (Fig. 9c). Other phases such as phlogopite, dolomite, and apatite were recognised in some samples. In a garnet-bearing sample, several apatite grains occur around a large grain of dolomite (Fig. 9d). Locally, the peridotites have undergone retrogression to an olivine + tremolite + cummingtonite + chlorite + talc assemblage.

Metamorphism and textures of dry and hydrous garnet peridotites

461

Fig. 9. (a) Fine type amphibole-bearing peridotite. Amphibole includes small, irregular shaped spinels. Parallel polarisers. Field of view is 2 mm. (b) Fine type Grt-Spl-Amph peridotite. Relict of coarse garnet porphyroblast with inclusion of corroded spinel set up in fine-grained matrix. Partly crossed polarisers. Field of view is 4 mm. (c) Fine type chlorite-bearing peridotite. Chlorite grains mark the rock foliation. Partly crossed polarisers. Field of view is 1.25 mm. (d) Granoblastic polygonal texture of a fine type peridotite. Dolomite grain surrounded by small apatite grains. Plane polarised light. Field of view is 1.3 mm.

Acknowledgements The authors would like to acknowledge scientific support by Katrine Frese and Anne Chantal Risold, technical support by Roberto Braga and financial support to L. M. by MIUR and CNR. A careful review by G. B. Piccardo helped to improve the manuscript and is also acknowledged.

References Andreatta, C. (1934): Analisi strutturali di rocce metamorfiche,V. Oliviniti. Period. Mineral., 5:237–253. Andreatta, C. (1935): La formazione gneissico-kinzigitica e le oliviniti della Val d’Ultimo (Alto Adige). Mem. Mus. Stor. Nat. Venezia Tridentina, 3:1–160. Andreatta, C. (1948): La linea di Peio nel massiccio dell’Ortles e le sue miloniti. Acta Geol. Alp., 1:1–63. Bächlin, R., Bianconi, F., Codoni, A., Dal Vesco, E., Knoblauch, F., Kündig, E., Reinhard, M., Spaenhauer, F., Spicher, A., Trommsdorff, V. & Wenk, E. (1974): Geologischer Atlas der Schweiz. Geologische Karte 1:25 000 Bellinzona (LK 1313), mit Anhängsel von Blatt 1312 Locarno.

462

L. Morten & V. Trommsdorff

Bay, L. (1999): Strukturgeologie und Petrographie im Gorduno Tal (Alpe Arami). Diploma thesis, ETH, Zürich, Switzerland. Becker, H. (1993): Garnet peridotite and eclogite Sm-Nd mineral ages from the Lepontine dome (Swiss Alps): new evidence for Eocene high-pressure metamorphism in the central Alps. Geology, 21:599–602. Ben Ismail, W. & Mainprice, D. (1998): An olivine fabric database: an overview of upper mantle fabrics and seismic anisotropy. Tectonophysics, 269:145–157. Bondi, M., De Francesco, A.M. & Morten, L. (1992): Major elements, 3d transition elements, Cu and Sr geochemistry of peridotitic rocks within the Austridic crystalline basement, Nonsberg area, Northern Italy. In Carmignani, L. & Sassi, F.P. (eds.): Contributions to the geology of Italy with special regard to Palaeozoic basements /IGCP No. 276 Newslett., 5/, 229–235. Boudier, F. (1978): Structure and petrology of the Lanzo peridotite massif (Piemont Alps). Geol. Soc. Am. Bull., 89:1574–1591. Bystricky, M., Kunze, K., Burlini, L. & Burg, J.P. (2000): High shear strain of olivine aggregates; rheological and seismic consequences. Science, 290:1564–1567. Carter, N.L. & Avé Lallemant, H.G. (1970): High temperature flow of dunite and peridotite. Geol. Soc. Am. Bull., 81:2118–2202. Dal Piaz, G.V. (1993): Evolution of Austroalpine and Upper Penninic basement in the Northwestern Alps from Variscan convergence to post Variscan Extension. In von Raumer, J.F. & Neubauer, F. (eds.): The preMesozoic geology in the Alps. Berlin: Springer-Verlag, 327–343. Den Tex, E. (1969): Origin of ultramafic rocks, their tectonic setting and history: a contribution to the discussion of the paper ‘The origin of ultramafic and ultrabasic rocks’ by Wyllie, P.J. Tectonophysics, 7:457–488. Dobrzhinetskaya, L., Green, H.V. & Wang, S. (1996): Alpe Arami: a peridotite massif from depths of more than 300 kilometers. Science, 271:1841–1845. Drury, M.R. & Fitzgerald, J.D. (1998): Mantle rheology: insights from laboratory studies of deformation and phase transition. In Jackson, I. (ed.): The Earth’s mantle: composition, structure, evolution. Cambridge: Cambridge Univ. Press, 503–559. Eberli, G.P. (1988): The evolution of the southern continental margin of the Jurassic Tethys Ocean as recorded in the Allgäu Formation of the Austroalpine Nappes of Graubünden. Eclogae Geol. Helv., 81:175–214. Ernst, W. G.(1978): Petrochemical study of the lherzolitic rocks from Western Alps. J. Petrol., 19:341–392. Evans, B.W. & Trommsdorff, V. (1978): Petrogenesis of garnet lherzolite, Cima di Gagnone, Lepontine Alps. Earth Planet. Sci. Lett., 40:333–348. Evans, B.W., Trommsdorff, V. & Richter, W. (1979): Petrology of an eclogite-metarodingite suite at Cima di Gagnone, Ticino, Switzerland. Am. Mineral., 64:15–31. Evans, B.W., Trommsdorff, V. & Goles, G.G. (1981): Geochemistry of high grade eclogites and metarodingites from the Central Alps. Contrib. Mineral. Petrol., 76:301–311. Flügel, H.V. (1990): Das vor Alpine Basement im Alpin-Mediterranen Belt – Überblick und Problematik. Jahrb. Geol. Bundesanst., 133:181–221. Frese, K., Trommsdorff, V. & Kunze, K. (2003): Olivine [100] normal to foliation: lattice preferred orientation in prograde garnet peridotite formed at high H2O activity, Cima di Gagnone (Central Alps). Contrib. Mineral. Petrol., 145:75–86. Froitzheim, N. & Eberli, G.P. (1990): Extensional detachment faulting in the evolution of a Tethys passive continental margin, Eastern Alps, Switzerland. Geol. Soc. Am. Bull., 102:1297–1308. Froitzheim, N., Schmid, S.M. & Frey, M. (1996): Mesozoic palaeogeography and the timing of eclogite facies metamorphism in the Alps: a working hypothesis. Eclogae Geol. Helv., 89:51–110. Fumasoli, M.W. (1974): Geologie des Gebietes nördlich und südlich der Jorio-Tonale-Linie im Western von Gravedona (Como, Italia). PhD thesis. Univ. Zürich. Gebauer, D. (1996): A P–T–t path for an (ultra?-) high-pressure ultramafic/mafic rock-association and its felsic country-rocks based on SHRIMP-dating of magmatic and metamorphic zircon domains. Example: Alpe Arami (Central Swiss Alps). In Basu, A. & Hart, S. (eds.): Earth processes: Reading the isotopic code. /Geophys. Monogr., 95/. Washington (D.C.): Am. Geophys. Union, 307–329. Gebauer, D., Grünenfelder, M., Tilton, G., Trommsdorff, V. & Schmid, S. (1992): The geodynamic evolution of garnet-peridotites, garnet-pyroxenites and eclogites of Alpe Arami and Cima di Gagnone (Central Alps) from Early Proterozoic to Oligocene. Schweiz. Mineral. Petrogr. Mitt., 72:107–111.

Metamorphism and textures of dry and hydrous garnet peridotites

463

Gebauer, D., Schertl, H.P., Brix, M. & Schreyer, W. (1997): 35 Ma old ultrahigh-pressure metamorphism and evidence for very rapid exhumation in the Dora Maira Massif, Western Alps. Lithos, 41:5–24. Godard, G., Martin, S., Prosser, G., Kienast, J.R. & Morten, L. (1996): Variscan migmatites, eclogites and garnet-peridotites of the Ulten zone, Eastern Austroalpine system. Tectonophysics, 259:313–341. Green, H.V., Dobrzhinetskaya, L., Riggs, E.M. & Zhen-Ming., J. (1997): Alpe Arami: a peridotite massif from the Mantle Transition Zone? Tectonophysics, 279:1–21. Grubenmann, U. (1908): Der Granatolivinfels des Gordunotales und seine Begleitgesteine. Vierteljahrsschr. Naturforsch. Ges. Zürich, 53:129–156. Gudfinnsson, G.H. & Wood, B.J. (1998): The effect of trace elements on the olivine-wadsleyite transformation. Am. Mineral., 83:1037–1044. Hacker, B.R., Sharp, T., Zhang, R.Y., Liou, J.G. & Hervig, R.L. (1997): Determining the origin of ultrahighpressure lherzolites. Science, 278:702–704. Heinrich, C.A. (1982): Kyanite-eclogite to amphibolite facies evolution of hydrous mafic and pelitic rocks, Adula Nappe, Central Alps. Contrib. Mineral. Petrol., 81:30–38. Heinrich, C.A. (1986): Eclogite facies regional metamorphism of hydrous mafic rocks in the central Alpine Adula nappe. J. Petrol., 27:123–154. Hervig, R.L., Smith, J.V. & Dawson, J.B. (1986): Lherzolite xenoliths in kimberlite and basalts: petrogenetic and crystallographic significance of some minor and trace elements in olivine, pyroxenes, garnet and spinel. Trans. R. Soc. Edinburgh, Earth Sci., 77:181–201. Herzberg, C., Riccio, L., Chiesa, A., Fornoni, A., Gatto, G.O., Gregnanin, A., Piccirillo, E.M. & Scolari, A. (1977): Petrogenetic evolution of a spinel-garnet-lherzolite in the Austridic crystalline basement from Val Clapa (Alto Adige, northeastern Italy). Mem. Ist. Geol. Mineral. Univ. Padova, 30:1–28. Hoinkes, G. & Thöni, M. (1993): Evolution of the Ötztal-Stubai, Scarl-Campo and Ulten Basement Units. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 485–494. Hoogerduijn Strating, E.H. (1991): The evolution of the Piemonte-Ligurian ocean. A structural study of ophiolite complexes in Liguria (NW Italy) /Geol. Ultraiectina, 74/. Jäger, E., Niggli, E. & Wenk, E. (1967): Rb-Sr Altersbestimmungen an Glimmern der Zentralalpen. /Beitr. Geol. Karte Schweiz, 134/, 67 p. Jung, H. & Karato, S.I. (2001): Water-induced fabric transitions in olivine. Science, 293:1460–1462. Kretz, R.(1983): Symbols for rock-forming minerals. Am. Mineral., 68:277–279. Leinenweber, K., Utsumi,W., Tsuchida, Y., Yagi, T. & Kurita, K. (1991): Unquenchable high-pressure perovskite polymorphs of MnSnO3 and FeTiO3. Phys. Chem. Miner., 18:244–250. Mackwell, S.J. (1991): High temperature rheology of enstatite: implications for creep in the mantle. Geophys. Res. Lett., 18:2027–2030. Maggetti, M. & Flisch, M. (1993): Evolution of the Silvretta nappe. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 469–484. Martin, S. & Prosser, G. (1993): Pre-Alpine evolution of Upper Austroalpine Units from Northwestern Trentino, Italy. Terra Nova, 5(4):16. Martin, S., Prosser, G. & Santini, L. (1991): Alpine deformation along Periadriatic lineament in the Italian Eastern Alps. Ann. Tecton., 5(2): 118–140. Menzies, M.A. & Dupuy, C. (1991): Orogenic massifs: protolith, process and provenance. In Menzies, M.A., Dupuy, C. & Nicolas, A. (eds.): Orogenic lherzolites and mantle processes /J. Petrol., Spec. Issue., 32/, 1–16. Mercier, J.-C.C. & Nicolas, A. (1975): Textures and fabrics of upper-mantle peridotites as illustrated by xenoliths from basalts. J. Petrol., 16:454–487. Meyre, C., De Capitani, C. & Partzsch, J.H. (1997): A ternary solid solution model for omphacite and its application to geothermobarometry of eclogites from the middle Adula Nappe (Central Alps, Switzerland). J. Metamorph. Geol., 15:687–700. Meyre, C., De Capitani, C., Zack, T. & Frey, M. (1999): Petrology of high-pressure metapelites from the Adula nappe (Central Alps, Switzerland). J. Petrol., 40:199–213. Möckel, J.R. (1969): Structural petrology of the garnet-peridotite of Alpe Arami (Ticino, Switzerland). Leidse Geol. Meded. 42:61–130.

464

L. Morten & V. Trommsdorff

Morten, L. (1993): Ultramafic rocks in the eastern Alps. In Morten, L. (ed.): Italian eclogites and related rocks /Accad. Naz. Sci. detta Accad. XL, 13/, 111–119. Morten, L. & Obata, M. (1983): Possible high-temperature origin of pyroxenite lenses within garnet peridotite, northern Italy. Bull. Minéral., 106:775–780. Morten, L. & Obata, M. (1990): Rare earth abundances in the Eastern Alpine peridotites, Nonsberg area, Northern Italy. Eur. J. Mineral., 2:643–653. Morten, L., Bargossi, G.M. & Landini Bargossi, F. (1976): Notizie preliminari sulle metamorfiti della Val di Rumo, Val di Non, Trento. Mineral. Petrogr. Acta, 21:137–144. Muegge, O. (1898): Ueber Translationen und verwandte Erscheinungen in Krystallen. Neues Jahrb. Mineral. Geol. Paläontol, (1):71–158. Müntener, O. (1997): The Malenco peridotites (Alps): petrology and geochemistry of subcontinental mantle and Jurassic exhumation during rifting. Ph.D. thesis, ETH Zürich, Switzerland. Neubauer, F. (1988): The Variscan orogeny in the Austro-Alpine and South-Alpine domains of the Eastern Alps. Schweiz. Mineral. Petrogr. Mitt., 98:339–349. Neubauer, F. & Frisch, W. (1993): The Austro-Alpine metamorphic basement East of the Tauern Window. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 515–536. Neubauer, F. & von Raumer, J.F. (1993): The Alpine basement – Linkage between the Variscides and EastMediterranean mountain belts. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 641–661. Neugebauer, L. (1990): The Japetus model: a plate tectonic concept for the Variscan belt of Europe. Tectonophysics, 169:229–256. Nicolas, A. & Christensen, N.I. (1987): Formation of anisotropy in upper mantle peridotites: a review. Rev. Geophys., 25:111–123. Nicolas, A. & Jackson, E.D. (1972): Répartition en deux provinces des peridotites des chaînes alpines logeant la Méditerranée; implications géotectoniques. Schweiz. Mineral. Petrogr. Mitt., 52:479–495. Nicolas, A. & Poirier, J.P. (1976): Crystalline plasticity and solid state flow in metamorphic rocks. London: Wiley. Nicolas, A., Boudier, F. & Bouiller, A.M. (1973): Mechanism of flow in naturally and experimentally deformed peridotites. Am. J. Sci., 273:853–876. Nimis, P. & Morten, L. (2000): P–T evolution of ‘crustal’ garnet peridotites and included pyroxenites from Nonsberg area (upper Austroalpine), NE Italy: from the wedge to the slab. J. Geodyn., 30:93–115. Nimis, P. & Trommsdorff, V. (2001): Revised thermobarometry of Alpe Arami and other garnet peridotites from the Central Alps. J. Petrol., 42:103–115. Obata, M. & Morten, L. (1985): The spinel lherzolite to garnet lherzolite transformation and the gravity sinking of ultramafic bodies in the lower crust. Terra Cognita, 5:440. Obata, M. & Morten, L. (1987): Transformation of spinel lherzolite to garnet lherzolite in ultramafic lenses of the Austridic crystalline Complex, Northern Italy. J. Petrol., 28:599–623. Obata, M. & Thompson, A.B. (1981): Amphibole and chlorite in mafic and ultramafic rocks in the lower crust and upper mantle. Contr. Mineral. Petrol., 77:74–81. O’Hara, M.J. & Mercy, E.L. (1966): Garnet peridotite and eclogite from Bellinzona, Switzerland. Earth Planet. Sci. Lett., 1:61–130. Paquin,J. & Altherr, R. (2001): New constraints on the P–T evolution of the Alpe Arami garnet peridotite body (Central Alps, Switzerland). J.Petrol., 42:1119–1140. Partzsch, J.H. (1996): The tectono-metamorphic evolution of the middle Adula nappe, Central Alps, Switzerland. Ph.D. thesis, Univ. of Basel, Switzerland. Pfiffner, M. (1999): Genese der Hochdruckmetamorphen ozeanischen Abfolge der Cima-Lunga Einheit (Zentralalpen). Ph.D. thesis, ETH Zürich, Switzerland. Pfiffner, M. & Trommsdorff, V. (1998): The high-pressure ultramafic-mafic-carbonate suite of Cima LungaAdula, Central Alps: excursions to Cima di Gagnone and Alpe Arami. Schweiz. Mineral. Petrogr. Mitt., 78:337–354. Phakey, P., Dollinger, G. & Christie, J. (1972): Transmission electron microscopy of experimentally deformed olivine crystals. Geophys. Monogr., 16:117–138.

Metamorphism and textures of dry and hydrous garnet peridotites

465

Piccardo, G. B., Rampone, E. & Vannucci, R. (1990): Upper mantle evolution during continental rifting and ocean formation: evidences from peridotite bodies of the Western Alpine-Northern Appennine system. In Roure, F., Heitzmann, P. & Polino, R. (eds.): Deep structure of the Alps /Mém. Soc. Géol. Fr., 156/, 323–333. Raleigh, C.B. (1965): Glide mechanisms in experimentally deformed minerals. Science, 150:739–741. Rampone, E. & Morten, L. (2001). Records of crustal metasomatism in the garnet peridotites of the Ulten Zone (Upper Austroalpine, Eastern Alps. J. Petrol., 42:207–219. Reusser, E., Risold, A-C., Günther, D. & Trommsdorff, V. (1998): Bulk Ti-content of ilmenite-bearing olivine from garnet lherzolites. EOS, Trans. Am. Geophys. Union, 79:F953. Riecke, E. (1895): Ueber das Gleichgewicht zwischen einem festen homogen deformierten Körper und einer flüssigen Phase. Ann. Phys., 54:731–738. Risold, A-C., Trommsdorff, V., Reusser, E. & Ulmer, P. (1996): Alpe Arami and Cima di Gagnone garnet peridotites: observations contradicting the hypothesis of ultra deep origin. EOS, Trans. Am. Geophys. Union, 77:F761. Risold, A-C., Trommsdorff, V., Reusser, E. & Ulmer, P. (1997): Genesis of FeTiO3 inclusions in garnet peridotites from the Central Alps. Terra Abstr., 9:28–29. Risold, A-C., Morten, L. & Trommsdorff, V. (1999): Ilmenite palisades in olivine of spinel- and garnetperidotites from the Nonsberg area, upper Austroalpine, northern Italy. In 4th Int. Eclogite Field Symp., August 14–18, 1999, Kazakhstan, Abstr., 20. Risold, A-C., Trommsdorff, V. & Grobety, B. (2001): Genesis of ilmenite rods and palisades along humite-type defects in olivine from Alpe Arami. Contrib. Mineral. Petrol., 140:619–628. Risold, A-C., Trommsdorff, V. & Grobety, B. (2003): Morphology of oriented ilmenite inclusions in olivine from garnet peridotites (Central Alps, Switzerland). Eur. J. Mineral., 15:289–294. Rost, F. & Brenneis, P. (1978): Die Ultramafitite im Bergzug südlich des Ultentales, Provinz Alto Adige (Oberitalien). Tschermaks Mineral. Petrogr. Mitt., 25:257–286. Rost, F., Wannemaker, J. & Grigel, W. (1974): Die Ultramafitite der Alpe Arami und Croveggio Tessin (Schweiz). Schweiz. Mineral. Petrogr. Mitt., 54:353–369. Rost, F., Wannemaker, J. & Anton, J. (1979): Garnet ultramafitites in the Ivrea zone and other crystalline zones along the Periadriatic Lineament. Mem. Sci. Geol., 33:183–191. Schmid, S.M., Rück, P. & Schreurs, G. (1990): The significance of the Schams nappes for the reconstruction of the palaeotectonic evolution of the Penninic zone along the NRP20-East traverse (Grisons, Eastern Switzerland). Mém. Soc. Géol. Fr., 156:263–287. Thöni, M. (1981): Degree and evaluation of the Alpine metamorphism in the Austroalpine Unit W of the Hohe Tauern in the light of K/Ar and Rb/Sr age determinations in micas. Jahrb. Geol. Bundesanst., 124:111–174. Tommasi, A., Mainprice, D., Canova, G. & Chastel, Y. (2000): Viscoplastic self-consistent and equilibriumbased modeling of olivine lattice preferred orientations: implications for the upper mantle seismic anisotropy. J. Geophys. Res., 105:7893–7908. Trommsdorff, V. (1966): Progressive Metamorphose kieseliger Karbonatgesteine in den Zentralalpen zwischen Bernina und Simplon. Schweiz. Mineral. Petrogr. Mitt., 46:431–460. Trommsdorff, V. (1990): Metamorphism and tectonics in the Central Alps: the Alpine lithospheric mélange of Cima Lunga and Adula. Mem. Soc. Geol. Ital., 45:39–49. Trommsdorff, V. & Evans, B.W. (1969): The stable association enstatite-forsterite-chlorite in amphibolite facies ultramafics of the Lepontine Alps. Schweiz. Mineral. Petrogr. Mitt., 49:325–332. Trommsdorff, V., Piccardo, G.B. & Montrasio, A. (1993): From magmatism through metamorphism to sea floor emplacement of subcontinental Adria lithosphere during pre-Alpine rifting (Malenco, Italy). Schweiz. Mineral. Petrogr. Mitt., 73:191–203. Trommsdorff, V., Hermann, J., Müntener, O., Pfiffner, M. & Risold, A.C. (2000): Geodynamic cycles of subcontinental lithosphere in the Central Alps and the Arami enigma. J. Geodyn. 30:77–92. Trommsdorff, V., Risold, A.C., Reusser, E., Connolly, J. & Ulmer, P. (2001): Titanian clinohumite: ilmenite rod inclusions and phase relations, Central Alps. In UHPM workshop 2001, Fluid/slab/mantle interactions and ultrahigh-P minerals, Waseda Univ., Tokyo, Abstr. Vol., 84.

466

L. Morten & V. Trommsdorff

Ulmer, P., Risold, A.C. & Trommsdorff, V. (1998): TiO2 solubility in mantle olivine as a function of pressure, temperature, aSiO2 and fH2. EOS, Trans. Am. Geophys. Union, 79:164. von Raumer, J.F. & Neubauer, F. (1993): Late Pre-Cambrian and Palaeozoic evolution of the Alpine basement – An overview. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 625–639. Wenk, H.R. (1985): Preferred orientation in deformed metals and rocks; an introduction to modern texture analysis. Orlando (Fl.): Academic Press. Ziegler, P.A. (1984): Caledonian and Hercynian crustal consolidation of Western and Central Europe – a working hypothesis. Geol. Mijnb., 63(1): 93–108. Ziegler, P.A. (1993): Late Palaeozoic-Early Mesozoic plate reorganisation: evolution and demise of the Variscan fold belt. In von Raumer, J.F. & Neubauer, F. (eds.): The pre-Mesozoic geology in the Alps. Berlin: Springer-Verlag, 203–216.

EMU Notes in Mineralogy, Vol. 5 (2003), Chapter 15, 467–487

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks JACQUES L.R. TOURET1* and MARIA-LUCE FREZZOTTI2 1

Musée de Minéralogie, École des Mines, 60, Bd. Saint-Michel, 75006 Paris, France 2 Dipartimento di Scienze della Terra, Università di Siena, Via Laterina 8, 53100 Siena, Italy; * e-mail: [email protected]

Introduction Fluid–rock interaction occurs at all steps of metamorphic evolution. Most metamorphic reactions, in fact, involve volatiles – aqueous for hydroxyl-bearing minerals, carbonic for carbonates. The determination of the P–T conditions of trapping and of the precise chemical composition of these fluids has always been a major problem of metamorphic petrology. Many methods are possible, mainly based on experimental or indirect approaches (e.g. thermodynamic modelling, experimental mineral reactions, stable isotopes etc.). Fluid inclusion studies (e.g. Roedder, 1984) represent a possible direct way to study fluids trapped at high P and T. High pressure (HP) and particularly ultrahigh pressure (UHP) metamorphic rocks have been involved in a highly complicated metamorphic history, being buried at great depth, 100 km or more, then exhumed back to the surface in a relatively short time, typically a few tens of millions years. These extreme values, together with the fact that the most abundant minerals, garnet and pyroxene, are not listed among the most favourable mineral species for fluid inclusion research, would at first sight indicate that fluid inclusion studies are of little interest in these important rock types. Yet in recent years a great amount of publications have appeared, notably on eclogites and related rocks, showing not only that workable fluid inclusions are commonly present, but also that they show a remarkable variety of observed inclusion types, aqueous (i.e. Ca and Na dominated brines) as well as gaseous (notably N2 and/or CO2) (see e.g. Andersen et al., 1990; Philippot & Selverstone, 1991; Klemd, 1989; Gao & Klemd, 2001; Fu, 2002). Interestingly, some of these inclusions may ultimately contain fluid remnants trapped at different stages of the rock evolution: not only during final uplift (e.g. Agard et al., 2000), as anyone might intuitively guess, but also during peak metamorphic conditions, despite the extreme depths at which this metamorphism occurs, and even, but only in exceptional cases (Dabie Shan ultrahigh pressure metamorphic rocks), possibly at the pre-metamorphic stage (Liou et al., 1997; Zheng et al., 1999, 2001; Fu et al., 2002, 2003a,b). Fluid inclusions alone are not sufficient to reach this provocative conclusion. Their data must be confronted to other pieces of evidence, notably the stable isotope signature (e.g. Rumble & Yui, 1998), which appears also to have been preserved from pre-metamorphic surface conditions, before burial and exhumation.

USE "PAGE UP" AND "PAGE DOWN" TO SEE THE PAPER

VOLUME INDEX

468

J.L.R. Touret & M.-L. Frezzotti

An up-to-date review of fluids in subduction zones, essentially based on fluid inclusion data, has been recently done by Scambelluri & Philippot (2001). The present paper does not attempt to duplicate it. It extends its scope to continental collision settings (UHP rocks) and, above all, aims at considering some of the special problems involved in the study of fluid inclusions in HP–UHP eclogitic rocks. From examples mainly issued from the Dabie Shan terrane, it describes the rather specific working method developed in our group to address correctly this type of research. It involves a detailed discussion both on the organisation of a correct fluid inclusion investigation and on the petrological significance of data so far obtained from fluid inclusions. Indeed, the study of fluid inclusions in such rocks is not easy, but a number of well-documented cases show that it is undoubtedly worth the required time and effort.

The fluid inclusion approach to HP and UHP rocks: The principles A major problem in fluid inclusion studies, notably in metamorphic rocks, is the great number (at least many thousands) of fluid inclusions which may occur within any grain of some rock-forming minerals, as well as the fact that inclusions may have been formed at different stages of the rock history. In high-pressure metamorphic rocks, an additional problem arises from the fact that some inclusions may have suffered the same large temperature and pressure variations as those applied to the rock system, with obvious possibilities of extensive “post-trapping” changes (Roedder, 1984). As a consequence, the study of fluid inclusions in this environment should follow the procedures that apply for any other metamorphic rock in a very precise and rigorous way, as described in some detail in a recent publication (Touret, 2001). The correct investigation procedure involves four successive steps, which should precede any attempt of overall interpretation (fluid origin, fluid movements through the rock system etc.). Investigation steps should occur rigorously in the proposed order, with partial conclusions precisely stated at the end of each step: 1) identification of the different fluid types, 2) chronology of the successive inclusion generations, 3) selection of few representative inclusions for each fluid type, 4) comparison between independent fluid inclusion and mineral P–T data. Step 1: Identification of the fluid types Fluid inclusions provide some unique information not only on the composition, but also on the density (i.e. molar volume) of the fluid system trapped in the mineral cavity. As discussed later, the density parameter plays a most important role in the interpretation of fluid inclusion data. With a few exceptions – which are not rare in high pressure rocks – fluid inclusions occur only in minerals of simple, not variable composition (e.g. quartz), in which the chemical interaction between fluid and host minerals is minimal. In the case of quartz, for instance, the fluid system will only be saturated in silica. Solubility changes at variable P and T will only introduce minor volume changes (by quartz dissolution/ deposition on the wall of the cavities), that are irrelevant compared to the volume variations caused by the constant adaptation of the fluid system to changing P and T conditions.

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 469 This important conclusion is illustrated in the diagram of Figure 1, the basic figure for the interpretation of inclusions in high grade metamorphic rocks. After initial formation of an inclusion at high P and T, the two variables will no longer be independent, being related within the inclusion by the equation of state of the enclosed fluid. Any change in the inclusion volume will be determined by the pressure difference in and out of the inclusion cavity: a decrease if the metamorphic P–T path is more or less parallel to the pressure axis (isobaric path, Fig. 1c), or an increase in the case of strong decompression (Fig. 1b, d). When the strength of the host mineral is no longer able to accommodate to the pressure difference, inclusions will be ruptured, either by implosion (trajectory 1c), or by explosion (decrepitation; Fig. 1b). New generations of inclusions can then be formed, with characteristics (smaller size, typical textures) extensively discussed in the fluid inclusion literature (see e.g. Touret, 2001 and references therein). Note that if the metamorphic P–T

Fig. 1. The fundamental principle of inclusion interpretation in high grade metamorphic rocks. Each inclusion formed at high P–T conditions (P–T box) defines its fluid isochore (solid lines), which has to be compared with the metamorphic P–T path (heavy broken lines) followed during retrogression (a: isochoric path; b and d: decompression path; c: isobaric cooling path). Depending upon pressure difference (P), the inclusion volume will change, until rupture either by implosion (c), or by explosion (decrepitation, b and d)

470

J.L.R. Touret & M.-L. Frezzotti

path remains parallel to the fluid constant density line (isochore), no pressure difference is generated, and the most fragile inclusion can then be preserved from the greatest depths up to the surface (pseudo-isochoric path, Fig. 1a). As discussed later, the notion of pseudoisochoric P–T path is very important to understand some apparently surprising facts in HP–UHP rocks (i.e. the preservation of very deep fluid remnants). In a very general way, however, the ubiquity of post-trapping changes, combined with the fact that inclusions may have been initially formed in a rather wide P–T range has obvious consequences: densities are much more variable than chemical compositions in fluids present at a given stage of the rock history. This important conclusion can be directly used for the definition of a fluid type, which can be defined as a fluid system of (roughly) constant composition, existing at a given stage of rock evolution (relatively constant P–T conditions), but eventually showing large density variations. Figure 2 gives some examples of fluid types, represented by a set of inclusions within a single host mineral grain. The fluid composition is expressed by a number of microthermometric parameters: final melting temperatures (Tm) for aqueous systems or non-permanent gases (CO2), homogenisation temperatures (Th) for low critical T gases (CH4, N2). These microthermometric measurements must be supplemented, when appropriate, by micro-Raman analyses, which have proven to be extremely important for the study of HP–UHP rocks. The identification of some essential UHP mineral phases has to be done first with Raman (Chopin, 1984; Chopin & Ferraris, 2003), and this technique represents an extremely powerful tool to identify mineral phases (including daughter minerals) in fluid inclusions. For gases, distinction between N2, CH4 and CO2 and quantification of relative mole percentage is also possible (cf. Burke, 2001). Most rock samples contain only one or two fluid types, and cases involving more than five fluid types are really exceptional. With some experience, their identification is a rapid and easy task, done by direct microscopic observation on the double polished sections commonly used for fluid inclusion studies. It is, however, essential to represent exactly the fluid inclusion distribution, as done in Figure 2. The drawings represent exactly all inclusions occurring in a given field of view of the microscope, in a volume precisely identified by the exact measurement of its upper and lower surfaces (in the present case, upper and lower surfaces of the double polished plate). This procedure is critical, not only to analyse the time relations between eventual successive generations of fluid inclusions (see following sections), but also to precisely locate any inclusion which, for a reason or another, is the object of special attention. For instance, analysing an inclusion by Raman without knowing the microthermometric parameters loses a great part of its potential interest. A correct investigation requires first measuring melting and/or homogenisation temperatures, and then doing the Raman analyses on the same inclusion. An impossible task, if the inclusion has not been precisely identified and located, by means of annotated drawings or microphotographs. Step 2: Chronology of the different fluid types Once the different fluid types have been identified, then comes the most important part of the study: are fluids synchronous, occurring at the same time within the rock system, or have they circulated through the rock system at very different times, notably during

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 471

Fig. 2. Fluid types identified in UHP eclogite from Shuanghe, Dabie Shan, China (Fu, 2002). All inclusions occur in quartz, itself often included in other minerals. (a) G1 (pure N2), (b) associated to W1 (high salinity aqueous), (c) G2 (pure CO2), (d) W1 (high salinity aqueous, Tm < –10 °C), (e) W2 (low salinity aqueous), (f) GW: mixed aqueous-gaseous. Numbers near inclusions: underlined – temperature of homogenisation (all to liquid), italic – final melting temperatures. Same identification of fluid types as in Figures 4, 5 and 7.

the final stages of post-metamorphic uplift. The principles for fluid inclusion chronology in metamorphic rocks have been extensively discussed in a number of recent publications (notably Touret, 2001 and references herein), and they do not need to be discussed extensively here. It will simply be recalled that the basic notion of primary versus secondary inclusions is better replaced by a three-fold division between isolated (= primary; type 1 and 2 in Fig. 3), trail-bound (= secondary; types 5, 6 and 7 in Fig. 3) and clustered (group of neighbouring inclusions occurring in the same mineral host, but

472

J.L.R. Touret & M.-L. Frezzotti

Fig. 3. Principles of fluid inclusion chronology. 1, 2, 3, and 4: primary or early inclusions. 5, 6 and 7: secondary or late inclusions; see text.

not bound to a former plane of a healed microfracture; type 4 in Fig. 3). The clustered inclusions are especially important, as their interpretation can be completely different: either they are neighbouring isolated inclusions, and in this case they have exactly the same significance, or they represent former exploded (transposed) inclusions, and then they are more comparable to secondary inclusions. The distinction between the two options is only possible at a more elaborate stage of the investigation (steps 2 and 3), but the different categories must first be identified with precision. Further elaboration of the fluid chronology is mainly done on the successive generations of secondary inclusions, using the timing of the microfractures (“intragrain” versus “intergrain”; types 5 and 7, respectively, in Fig. 3) and the intersection criteria of the different trails. Further examples can be found in the extensive literature concerning high temperature metamorphic rocks (granulites; Touret, 1981, 1987). At this stage, it is more appropriate to underline some typical features of high pressure metamorphic rocks of direct relevance for fluid inclusion studies: 1) The mineral assemblage may record a long part of the whole P–T evolution: a. Many high-grade metamorphic rocks show textures typical for incomplete (or telescoped) mineral reactions: zoned compositions, especially in resistant

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 473 minerals like garnet, which are the best P–T indicators, coronas or decompression reactions etc. Primary inclusions trapped at different steps of this mineral evolution may record the evolving fluid system if they have survived the changing P–T conditions. b. High pressure and especially ultrahigh pressure metamorphic rocks have been buried to extreme depths, than rapidly uplifted. It may be expected that tension fractures will occur during this uplift, allowing the incoming of late, externally derived fluids. Also, when high pressure metamorphism is superimposed on a dry, lower crustal protolith (e.g. eclogitised Precambrian granulites of the Norwegian Caledonian orogen; Austrheim, 1987), it has been demonstrated (incidentally, by fluid inclusion studies; Austrheim, 1987; Andersen et al., 1989; 1991) that the eclogite-forming mineral reactions could only be triggered by the influx of late aqueous fluids along discrete shear zones. In other words, late fluids are to be expected in these metamorphic rocks, but these fluids must not be too abundant, otherwise they would reset all mineral compositions and eradicated all traces of high pressure metamorphism. These remarks are in line with the incredible “memory” of some of these rocks, which in the case of Dabie Shan, China, may have preserved not only the peak metamorphic mineral assemblage (i.e. coesite, diamond), but also the stable isotopic signature (Rumble & Yui, 1998; Zheng et al., 1999, 2001; Xiao et al., 2000; Fu et al., 2001; Jahn et al., 2003). 2) A further indication of the common occurrence of free fluids in high-pressure metamorphic rocks during peak metamorphism and limited fluid movement during retrogression is given by the fact that, in strong opposition with other metamorphic rock types, most abundant fluid inclusions do not occur in quartz (incidentally rare or absent in most eclogites) but in other rock-forming minerals: garnet, pyroxene (omphacite), epidote etc. Application of Steps 1 and 2: Micro-mapping at the scale of the thin section on the example of the Shuanghe coesite-bearing eclogite (Fu, 2002) In high and ultrahigh pressure metamorphic rocks, the variety and complexity of host minerals, the relatively small abundance of late fluids and the abundance of isolated or clustered inclusions impose a working technique illustrated by the drawings of Figure 4 (case of Dabie Shan, work done by Bin Fu at the occasion of his PhD thesis; Fu, 2002). First of all, some representative rock sections are selected, including a number of rock-forming minerals in their typical assemblage, as well as relevant textural features. In the present case some discrete shear zones and open microfractures were visible on many eclogite samples (Fig. 4). Fluid inclusions were searched in all minerals from the figure, and it was found that, in this case, they only occurred in quartz mostly present as isolated grains included in other minerals. Aqueous, gaseous and mixed aqueousgaseous inclusions were observed, either dominant in a single quartz grain (general case), or present together. Additional microthermometric parameters (final melting temperatures for aqueous, homogenisation temperatures for gaseous inclusions) allowed to identify five fluid types, illustrated in Figure 2, labelled G1 (N2-rich), G2 (CO2-rich), W1 (high salinity aqueous), W2 (low-salinity aqueous), WG (mixed gaseous-aqueous).

474

J.L.R. Touret & M.-L. Frezzotti

Fig. 4. Shuanghe eclogite, Dabie Shan, China. Distribution of the different fluid types at the scale of the thin section (Fu, 2002). 12 quartz crystals (heavy lines) have been investigated in details, recording all inclusions occurring in each crystal. Aqueous fluid types (W1 and W2) are spread in the whole section, W2 inclusions being later than W1 (Proof: Fig. 2e, corresponding on this drawing to the crystal labelled in the upper right corner of the figure). Gaseous inclusions (G2 and GW, in this sample G1 inclusions are missing) are related to a mm-size shear zone (6), well visible on the hand specimen. 1: garnet, 2: omphacite, 3: epidote, 4: amphibole (barroisite), 5: quartz, 6: shear zone, 7: microfissures (open cracks). 8: identification of the different fluid types: see also Figure 2.

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 475 Note that the clear definition of some types, notably W1 and W2, required a rather advanced knowledge of the melting data (Tm < –10 °C for W1, Tm > –10 °C for W2), thus some information inherent to step 3 (Fig. 5). However, in the present case both groups are also clearly defined by the inclusion setting: isolated or clustered for W1, secondary (trail-bound) for W2 (Fig. 5 d and e, respectively). Some important conclusions can immediately be derived from a close observation of the different figures: – G2 (CO2), as well as W2 (low-salinity aqueous) are late fluids, related either to discrete shear zones or parallel to open microfracture. The same yields also probably for WG, also occurring within or at close distance of the shear zone. – G1 (N2) and W1 (high-salinity aqueous), on the other hand, are more primary (isolated or clustered) in quartz, which is included in other minerals (garnet, zoisite etc., discussed in more detail below at Step 3); these fluids are compositionally equilibrated with the immediately enclosing mineral assemblage at peak metamorphic conditions, eventually reset in density during post-metamorphic uplift (Step 4).

Fig. 5. Melting temperatures (Tm) vs. homogenisation temperatures (Th) diagram of inclusions occurring in the sample shown in Figures 2 and 4 (94-M-55), together with another comparable sample (92 HS4): identification of W1 and W2 fluid types for Tm = –10 °C. Modified from Fu (2002). Identification of the different fluid types: see Figure 2.

476

J.L.R. Touret & M.-L. Frezzotti

Step 3: Selection of few representative inclusions for each fluid type Each inclusion provides its own isochore, to be compared with independent P–T mineral data (Step 4). This corresponds to the basic principle of fluid inclusion data interpretation, but it needs to be applied with logic and rigour. In most cases, the variability of inclusions occurring in any hand specimen is so great that potential isochores may cover the whole P–T field of geological interest. Increasing the number of measurements, in the form of composite frequency histograms, often makes nothing else than complicate the picture. Vague maxima (peaks) may appear, but without any precise meaning and possible interpretation. The worst solution would be to select one isochore at random, especially if it matches some mineral P–T box (see discussion in Touret, 2001). The rational procedure is to minimise the variability and, above all, to indicate precisely on which arguments the selection has been done: 1) Identify a sufficiently great number of inclusions (typically between 20 and about 50) formed at the same time (GSI = Group of Synchronous Inclusions, Touret, 2001). The GSI exactly represents the spatial inclusion distribution, in such a way that any inclusion can be precisely (and rapidly) relocated in the inclusion section at any time (Fig. 2). 2) In most cases (the exception would be boiling fluids or immiscible fluid systems, extremely rare in high pressure metamorphic rocks), the fluid content of the inclusions is homogeneous. Then the fluid density is directly indicated by the homogenisation temperature (Th) and it can be indicated on the drawing (Fig. 2), as well as combined in a frequency histogram (Fig. 6). Then several situations may be encountered: – General (and ideal case). If the GSI has not suffered many perturbations, all Th should be almost identical, resulting in a well-defined histogram maximum (singlepeak histogram, Touret, 1977). The group can be characterised by a relatively small number of measurements, and the peak value taken for the representative isochore (Fig. 6a). – In the case of more severe perturbation, transposition, Th values will be more scattered. It can be interesting to contour iso-Th lines, in order to visualise the direction of density evolution. In high pressure rocks, inclusion perturbations tend mainly to occur during decompression, resulting in a progressive decrease of the fluid density (Fig. 6b). Contouring often reveals “islands” of relatively high density, surrounded by wider areas of lower density fluid. The transition may be progressive or more abrupt, when a set of inclusion had decrepitated and was refilled with a fluid equilibrated at the news P–T conditions. This is often the case for clustered inclusions, in which undecrepitated and decrepitated inclusions may coexist within the same mineral host. This situation will notably result in a twopeak histogram (in most cases, the high-density peak much less prominent than the low density one), both of which can be taken for the definition of representative isochores.

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 477

Fig. 6. Frequency histograms illustrating different scenarios for post-trapping modifications. (a) single-peak histogram resulting from preserved fluid inclusions (isochoric path a in Fig. 1). (b) Scattered homogenisation temperature distribution resulting from a decompression P–T path (path b in Fig. 1). Decompression causes decrepitation of most fluid inclusions, which reset at lower density (i.e. higher homogenisation temperatures). (c) Scattered homogenisation temperature distribution resulting from isobaric cooling (path c in Fig. 1). Most fluid inclusions implode and density is shifted to higher values (i.e. lower homogenisation temperatures).

In conclusion, the selection of these isochores will depend on the investigated case, following these simple (and obvious) arguments. But in any case, the inclusion that is used for the definition should be precisely indicated, either by a special symbol on the inclusion drawing, or (and) by an arrow on the frequency histogram (Fig. 6). Step 4: Comparison between fluid inclusion and independent P–T mineral data At this stage of the investigation, we have a limited (typically less than five) isochores, in a given chronological order, to be compared to the P–T “box” defined by the mineral assemblage. In high grade, high temperature metamorphic rocks (granulite), the sense of density variation gives a first indication on the possible shape of the whole P–T path: clockwise if later isochores correspond to a density decrease, anticlockwise for the opposite trend (Touret, 2001). In the present case, the extremely strong decompression imposed by high to ultrahigh-pressure metamorphism indicates that only the first eventuality needs to be considered (actually verified by direct observation). It can also be observed that, in a very general way, typical P–T paths for high and especially ultrahigh pressure rocks (e.g. China; see Hirajima & Nakamura, 2003) are roughly parallel to the steep aqueous isochores, but cross systematically the less steep gaseous isochores (Fig. 7). Consequently, at striking difference with granulites, aqueous inclusions are better preserved than gaseous inclusions, systematically overpressured (relative to the enclosing mineral) during post-metamorphic evolution. This potential preservation of aqueous inclusions may be extreme. For instance, typical peak metamorphic conditions for UHP rocks of about 3 GPa, 600 °C are roughly on the trajectory of the isochore for liquid water (d = 1 g/cm3; Fig. 8). The above implies that if surface rocks are buried to a depth of 100 km, then brought back at the surface along this isochore, fluid inclusions will be preserved (Fig. 8). Incidentally, this situation might be less exceptional that it seems. It has been shown (Fu, 2002) that 18O-depleted eclogites at

478

J.L.R. Touret & M.-L. Frezzotti

Fig. 7. P–T interpretation of fluid inclusion data for Shuanghe eclogite (Figs. 2, 4 and 5), modified from Fu (2002). M1: peak UHP metamorphism. M2: recrystallisation; M3: amphibolite facies metamorphism (Cong et al., 1995; Liou et al., 1997). Jd100: jadeite + quartz = albite (Holland, 1980, 1983). Jd50: jadeite + quartz = albite with jadeite content of 50 mol% in clinopyroxene. Coesite–quartz from Bohlen & Boettcher (1982); Al2SiO5 triple point from Holdaway (1971). Fluid isochores were calculated using the empirical equation from Zhang & Franz (1987) and the modified Redlich–Kwong equation of state of Holloway (1981). Identification of the different fluid types: see Figure 2.

Dabie Shan are characterised by low salinity isolated (primary) inclusions in quartz, itself included in garnet (W2, Figs. 2 and 3), whereas identical inclusions in 18O-undepleted eclogites contain more saline fluids, considered to be far remnants of seawater in the premetamorphic protolith. Low salinity inclusions, on the other hand, could be remnants of the meteoric water, which has given to the protolith its 18O-depleted signature. Note, however, that these inclusions are compositionally identical to very late W2 fluids, trapped when the rocks were back near the surface, at the end of the complete metamorphic cycle.

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 479

Fig. 8. Typical P–T paths for HP–UHP rocks (eclogites and granulites, arrows). H2O: maximum density for pure aqueous inclusion, see text. Coesite–quartz from Bohlen & Boettcher (1982); Al2SiO5 triple point from Holdaway (1971).

480

J.L.R. Touret & M.-L. Frezzotti

Only the inclusion location (setting) allows distinguishing between the two cases, evidently completely different for the interpretation of fluid inclusion data. In general, inclusion isochores cross the P–T path well below peak metamorphic conditions, indicating extensive fluid density resetting during uplift (Fig. 7). It has been shown in metamorphic rocks from Crete by Kuster & Stockhert (1997) that, during decompression, inclusion density constantly re-equilibrates by dislocation creep until a temperature of about 300 °C. This temperature corresponds to a pressure of about 0.2–0.3 GPa, a range of values very commonly observed in most high pressure rocks studied so far. However, at least for the host mineral quartz, it does not seem that the fluid composition has changed, especially in syn- and even pre-metamorphic inclusions; the inclusion content has preserved the memory of the initial, pre-metamorphic fluid. The situation may be far more complicated for the spectacular gaseous and multiphase solid inclusions repeatedly found in many eclogite-forming minerals, notably garnet, pyroxene and epidote (Philippot & Selverstone, 1991; Scambelluri et al., 1997, 1998). In this case, it is obvious that extensive fluid–mineral interaction has occurred during a large part of the rock history, and that the actual inclusion composition may be quite different from the synmetamorphic fluid. Despite these complications, however, these fluid inclusions may still indicate recycling of components which were taken up in rocks well before the main eclogite facies metamorphism (e.g. Cl, Sr, B, Li etc.; Scambelluri & Philippot, 2001). Frequently, microdomains are observed at the size of a crystal, with variable and sometimes contrasting fluid composition. The interpretation of these fluids is far more complicated than that of those contained in a non-reactive mineral such as quartz; it requires a precise knowledge of the complete chemical composition of the fluid system, only marginally possible with the most advanced analytical apparatus (ion microprobe, synchrotron radiation). Few reliable data are yet available, and it is obvious that this line of research will be considerably developed in coming years. At this stage, we can only emphasise the absolute necessity to select the analysed inclusions with the care and rigor, which have been highlighted throughout this paper. It will be of little help to have the best possible analytical data on a given inclusion, if the time of formation and the representativity of this inclusion are not known.

Applications: Some general considerations based on recent studies The principles outlined above must be set within the framework of some general considerations on HP–UHP rocks, dealing notably with the type of fluids which can be expected, as well as those actually found in recent investigations. We will finally conclude this paper by some comparative remarks between high pressure and high temperature metamorphic rocks, which both show opposite and very characteristic fluid inclusion patterns. Which fluids can be expected in HP–UHP rocks? After having defined the specific methods of investigation that should be used for fluid inclusion studies in HP–UHP metamorphic rocks, it may be of interest to speculate on the nature of the fluids that might be encountered, as well as on those which have been

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 481 actually found in a number of recent studies. Pre-metamorphic protoliths, typically altered oceanic crust, have been generated in a water-rich environment (sea floor), including widespread hydrothermal systems. Such an origin results in fluids with variable salinities, incorporated in the rock system as hydrous minerals (mostly low salinity aqueous fluids) or pore sediments (often saline brines). Carbon-bearing fluids, either resulting from the degradation of organic matter (methane) or from part of the hydrothermal system (CO2, CH4), may be abundant, but as a rule they are expelled from the rock system at an early stage of the burial, before the onset of metamorphism (mud volcanoes etc.; see Henry et al., 1996). Only some carbonates, often in restricted amounts, will participate to the metamorphic evolution. During burial, progressive mineral reactions will generate internally buffered fluids, like in any other metamorphic sequence. Most initial minerals are water bearing (clay minerals, chlorites etc.), thus metamorphic fluids will be dominantly aqueous. However, in contrast with high temperature settings (amphibolite–granulite), crustal melting (anatexis) is not reached, and some hydrous minerals (zoisite, lawsonite, chloritoid, phengite etc.) will remain stable until the highest P–T conditions: aqueous fluids will be progressively liberated during prograde metamorphism, but in lesser amounts than in high temperature metamorphic regimes. When the protolith is not altered oceanic crust, but a dry lower crustal granulite (e.g. the Norwegian Caledonian Nappe system, Bergen Arcs), the influx of externally derived fluid is necessary to trigger some key high pressure mineral reactions, notably the transition granulite/eclogite. Limited quantities of fluids are introduced along discrete shear zones, as exemplified by the spectacular outcrops of Western Norway (Jamtveit et al., 2000), where a Caledonian eclogite, equilibrated at about 2 GPa and 600 °C grades into a metastable Precambrian granulite, equilibrated at about 0.8 GPa, 800 °C. Note that, contrarily to some statements (Kostenko et al., 2002), the nature of this fluid (low salinity aqueous; Andersen et al., 1991) has been clearly defined by fluid inclusion studies. Other externally derived fluids may be introduced into the rock system during post-metamorphic uplift, as in any other metamorphic setting. The amount of these fluids, however, should be limited, otherwise all minerals would be re-equilibrated. The preservation of extreme pressure conditions (up to about 3 GPa, corresponding to more than 100 km burial) is not only an indication that surface rocks have reached these extreme depths, but also that they have escaped complete retrogression during uplift. Which fluids are actually found in HP–UHP rocks? The results of a number of recent studies show indeed that the actual observations match relatively well the expectations. Peak pressure fluid inclusions are quite rare in eclogites and are always found in peak mineral phases (e.g. garnet, omphacite and kyanite) or in non-recrystallised mineral inclusions (e.g. quartz, kyanite and apatite) enclosed within high pressure minerals. Right from the first studies in the early 1980’s–1990’s on eclogites from a number of Caledonian, Hercynian, and Alpine high pressure terranes, it was clear the presence of different fluid types: pure or nearly pure CO2, H2O, N2, CH4 or binary CO2–N2 or N2–CH4 fluids, highly saline brines (Fig. 9) or complex aqueous-

482

J.L.R. Touret & M.-L. Frezzotti

Fig. 9. Microphotographs showing the distribution of fluid inclusions in eclogitic rocks (M. Scambelluri collection). (a) Primary brine inclusions (liquid + vapour + halite) in omphacite from eclogites. (b) Brine inclusions in omphacite from high pressure veins in eclogites.

gaseous mixtures dominate during eclogite facies metamorphism (e.g. Andersen et al., 1989, 1990; Klemd et al., 1992; Philippot et al., 1995; Scambelluri & Philippot, 2001). Some of these fluids are CO2 dominated and comparable to the fluids encountered in granulites (Touret, 1992), but contrarily to what one might expect, they clearly postdate aqueous and/or N2 fluids. Sharp et al. (1993) and Philippot et al. (1995) made a very important contribution to the understanding of the evolution of fluids in the ultrahigh pressure coesite- and pyrope-bearing rocks from the Dora-Maira massif in the Western Alps. These authors recognised that large quantities of dehydration fluids were released from hydrous phases during prograde metamorphism and went to in situ formation of melts. Remnants of prograde dehydration fluids, predating the formation of pyrope and coesite-bearing ultrahigh pressure assemblage, are preserved as low salinity aqueous fluid inclusions containing minor CO2 in relic kyanite within pyrope. Further evidence for the preservation of relics of dehydration fluids attending the growth of garnet is given by polyphase decrepitated inclusions in pyrope, containing Mg-phosphate, phyllosilicates, chlorides and opaque phases but no preserved fluid. Supercritical dense COH-silicate fluids rich in alkalis (P > 4.5 GPa at T = 1000 °C) related to eclogitisation were also reported in garnet of metamorphic gneiss lenses within migmatites of the gneiss-eclogite unit from Erzgebirge at UHP conditions in diamondbearing garnet gneisses (Stockhert et al., 2001). These fluids are preserved in polyphase inclusions containing diamond, phlogopite, quartz, paragonite, phengite, kyanite, Na-plagioclase, apatite and rutile. At very high pressure conditions (> 3–4 GPa) there is no immiscibility between an aqueous fluid and a silicate melt and these fluids prove the interaction of eclogite garnet with COH-silicate fluids. Aqueous fluids have also been recognised in Dabie Shan eclogites although these have lower salinities and may contain

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 483 N2 (Fu et al., 2003a,b; Xiao et al., 2000, 2001). The fact that aqueous fluids trapped in some preserved primary inclusions represent peak metamorphic fluids has reached general consensus nowadays. The presence of hydrous mineral phases at peak conditions in many HP–UHP terrains argue for a crucial importance of aqueous fluids during eclogite facies metamorphism. Aqueous eclogitic fluid inclusions, however, may show both variable salinity and ion species in solution. In the jadeite quartzites and eclogites from the Dabie Shan, primary aqueous fluid inclusions have different salinities and, most importantly, different cation species at the scale of the crystal (Fu, 2002). Cation abundances could be related to the host mineral phases (e.g. Ca-dominated brines in epidote, low salinity fluids in garnet, and Na, K-bearing aqueous fluids in amphibole). The role of nitrogen in eclogitic fluids has been extensively discussed by Andersen et al. (1989, 1990, 1991) in western Norway, where eclogitisation is triggered by channelled fluids circulating along the shear zones (Austrheim, 1987). By careful observation of the successive fluid inclusion generations, these authors were able to identify remnants of the eclogite fluids preserved in primary fluid inclusions within nonrecrystallised quartz grains included in garnet and in clinopyroxene. In line with other studies, these fluids are water-rich (XH2O > 0.8). In the gaseous part of the fluid system, N2 predominates (XN2 = 0.7 to 0.8), a striking difference with the granulite fluids which are pure CO2 (no visible H2O and N2 < 1 mol%). N2 in eclogitic fluids has also been reported in the Dabie Shan eclogites (Fu et al., 2001) and in the Münchberg gneiss complex (Klemd et al., 1992). It is noteworthy that the isochores determined from the densities of these inclusions pass through the peak metamorphic conditions calculated from the corresponding mineral assemblages. In conclusion, we now have a few guidelines for the interpretation of the complex fluid assemblages in eclogites. The typical fluid assemblage is a homogeneous mixture of H2O (with variable solute contents) and N2, contaminated by CO2 to various extents. If the peak metamorphic fluids are preserved in rare fluid inclusions, they correspond to homogeneous aqueous-gaseous mixtures, where nitrogen represents the dominant apolar gaseous species. Pure gases (CO2, N2; CH4) or their mixtures (CO2 + N2 or CH4 + N2) are related to the successive post-peak metamorphic evolution, formed by repeated episodes of fluid immiscibility, as to be expected during post-metamorphism uplift. In most cases, liberated water will be immediately consumed by partial melting and/or mineral reactions.

High grade metamorphic rocks: Some fundamental differences between high pressure (eclogites and related rocks) and high temperature (granulites) metamorphic rocks In conclusion, we hope to have convinced the reader that fluid inclusions are worth to be studied in HP and UHP rocks, and that they provide an important contribution to the understanding of these most important rocks. Many problems remain to be clarified, notably the modalities of fluid–rock interaction at all stages of the rock evolution, but they will not be attempted to be further discussed here.

484

J.L.R. Touret & M.-L. Frezzotti

Some considerations on the fluid source are rather straightforward. For instance, micas and feldspars may contain a significant amount of nitrogen in their structure in the form of NH4+ ions, which are released when these minerals break down (Andersen et al., 1991). Most N2 in eclogites is found at the stage where these minerals become unstable and such a link cannot be coincidental. The problem of aqueous fluid salinities is much more complicated, and it cannot be fully answered at present, notably for the high salinity brines commonly found in low T eclogites. Philippot et al. (1998) proposed that in Alpine eclogites chlorine-rich inclusions are derived from hydrothermal alteration of the oceanic lithosphere. Scambelluri et al. (1997) suggested the involvement of recycled seawater, Cl and alkalies in HP fluids, with no relation to hydrothermal alteration prior to subduction. However, in other eclogites (e.g. the Dabie Shan terrane) most saline inclusions are Ca-dominated and not Na-rich as it would be expected for a former seawater (Xiao et al., 2001; Fu et al., 2001). This feature may be taken as possible evidence for a non-marine source of the initial fluid. Probably these are originally pore fluids, salt-enriched by metamorphic dehydration. To be really effective, this process must act on a small volume of fluids in local systems, yielding a local trend towards increasing salinity and a heterogeneous distribution of fluid composition even at the sample scale, as observed by Fu et al. (2001). More work is clearly needed, with many more data than presently available. Much effort is presently made for the complete chemical analysis of single inclusions, by non-destructive (PIXE, synchrotron radiation) and/or destructive methods (laser ablation ICP-MS). Scambelluri et al. (2001) have recently published the first trace element laser ablation ICP-MS analyses of fluid + mineral inclusions produced during breakdown of antigorite serpentine, a major dehydration reaction occurring at depth within subducting oceanic plates. The fluid in the inclusions is enriched in incompatible elements, indicating that serpentine breakdown can produce a metasomatic fluid phase which can play a relevant role in arc magma genesis. We will conclude this presentation with some comments on high grade metamorphic rocks in general, stressing the fact that despite extreme P–T conditions, preserved fluid remnants add an essential contribution to the knowledge of these most important rocks. The contrast is striking between pressure-dominated (eclogites and related rocks) and temperature-dominated environments (granulites), which can be opposed for almost all aspects (Fig. 9). In eclogites, a wide range of possible fluid composition is found, but always in small and relatively rare inclusions. Well-preserved aqueous inclusions are exceedingly rare, most are transposed by explosion (decrepitation). Gaseous inclusions are also not very abundant, but rather common. The typical gas is N2, released from the breakdown of feldspars. CO2, when present, is locally derived from neighbouring rocks, mostly granulites. Fluid movements are extremely limited, both in the fluid quantities and transport scale (probably not exceeding a few hundreds of meters). Rare and strongly transposed gaseous inclusions (including externally derived CO2) mostly derive from neighbouring granulites. All other fluids appear to have been internally buffered during the prograde evolution of the metamorphic pile, and most aqueous fluids are far remnants of supracrustal pore waters (saline aqueous fluids in the case of sea floor or hydrothermally altered rocks, and pure water from meteoric fluids).

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 485 In granulites, the amount of inclusions (both in size and number) is variable, but it can be extremely high in rocks that have escaped solid state recrystallisation (annealing = granulitic texture). Mostly gaseous inclusions are found, in the form of high density pure CO2. These fluids are to a large extent syn-metamorphic and externally derived. They are introduced in the lower crust from a distant source (transport distance at kilometre scale) by mantle-derived intrusions, which not only supply the fluids but also the heat responsible for the special metamorphic character (high to ultrahigh temperature fluid influx at peak metamorphism). Aqueous fluids might also have been present, but they are not preserved in inclusions, and their occurrence can only be guessed from indirect features (K-feldspar veining, high chlorine content of some minerals etc.). The origin of these fluids is highly hypothetical but, as well as for the CO2, it seems to be related to fluid–magma interaction (i.e. fluids dissolved in magmas, gradually expelled during progressive magma crystallisation). It is also interesting to note that the main characteristics of the fluid inclusion population are unambiguously associated to each metamorphic facies: eclogitised granulites (Western Norway) show typical eclogite inclusions, with rare remnants of former granulite inclusions, whereas granulitised eclogites (Dabie Shan) contain typical granulite inclusions, with almost no remnants of former eclogite inclusions. In this respect, the dominant fluid inclusion population appears to be almost as typical as the mineral assemblage for the characterisation of a given metamorphic facies.

Acknowledgements We thank B. de Vivo and M. Scambelluri for perspective reviews, and R. Compagnoni for invitation to the EMU School in Budapest. Research on fluid inclusions in UHP rocks are funded by the Siena University PAR project 2003.

References Agard, C., Goffé, B., Touret, J.L.R. & Vidal, O. (2000): Retrograde mineral and fluid evolution in high-pressure metapelites (Schistes Lustrés Unit, Western Alps). Contrib. Mineral. Petrol., 140:296–315. Andersen, T., Burke, E.A.J. & Austrheim, H. (1989): Nitrogen bearing aqueous fluid inclusions in some eclogites from the western gneiss region of the Norwegian Caledonides. Contrib. Mineral. Petrol., 103:153–165. Andersen, T., Austrheim, H. & Burke, E.A.J. (1990): Fluid inclusions in granulites and eclogites from the Bergen arcs, Caledonides of West Norway. Mineral. Mag., 54:145–158. Andersen, T.B., Jamtveit, B., Dewey, J.F. & Swensson, E. (1991): Subduction and eduction of continental crust: major mechanisms during continent-continent collision and orogenic extensional collapse. A model based on south Norwegian Caledonides. Terra Nova, 3:303–310. Austrheim, H. (1987): Eclogitization of lower crustal granulites by fluid migration through shear zones. Earth Planet. Sci. Lett., 81:221–232. Bohlen, S.R. & Boettcher, A.L. (1982): The quartz-coesite transformation: a precise determination of the effects of other components. J. Geophys. Res., 87:7073–7078. Burke, E.A.J. (2001): Raman microspectrometry of fluid inclusions. Lithos, 55:139–159. Chopin, C. (1984): Coesite and pure pyrope in high-grade blueschists of the western Alps: a first record and some consequences. Contrib. Mineral. Petrol., 86:107–118. Chopin, C & Ferraris, G. (2003): Minerali chemistry and mineral reactions in UHPM rocks. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Minerals., 5/. Budapest: Eötvös Univ. Press, 191–227.

486

J.L.R. Touret & M.-L. Frezzotti

Cong, Y., Zhai, M., Carswell, D.A., Wilson, R.N., Wang, Q., Zhao, Z. & Windsley, B.F. (1995): Petrogenesis of ultrahigh-pressure rocks and their country rocks at Shuanghe in Dabieshan, central China. Eur. J. Mineral., 7:119–138. Fu, B. (2002): Fluid regime during high- and ultrahigh-pressure metamorphism in the Dabie-Sulu terranes, Eastern China: a stable-isotope and fluid-inclusion study. Ph.D. thesis, Vrije Universiteit, Amsterdam, The Netherlands, 144 p. Fu, B., Touret, J.L.R. & Zheng, Y.-F. (2001): Fluid inclusions in coesite-bearing eclogites and jadeite quartzite at Shuanghe, Dabie Shan (China). J. Metamorph. Geol., 19:529–545. Fu, B., Zheng, Y.-F. & Touret, J.L.R. (2002): Petrological, isotopic and fluid inclusion studies of eclogites from Sujiahe, NW Dabie Shan (China). Chem. Geol., 187:107–128. Fu, B., Touret, J.L.R. & Zheng, Y.-F. (2003a): Remnants of pre-metamorphic fluid and oxygen-isotopic signatures in eclogites and garnet clinopyroxenite from the Dabie-Sulu terranes, eastern China. J. Metamorph. Geol., in press. Fu, B., Touret, J.L.R., Zheng, Y.-F. & Jahn, B.-M. (2003b): Fluid inclusions in granulites, granulitized eclogites and garnet clinopyroxenites from the Dabie-Sulu terranes, eastern China. Lithos, in press. Gao, J. & Klemd, R. (2001): Primary fluids entrapped at blueschist to eclogite transition: evidence from the Tianshan meta-subduction complex in northwestern China. Contrib. Mineral. Petrol., 142:1–14. Henry, P., Le Pichon, X., Lallemant, S., Lance, S., Martin, J., Foucher, J-P., Fiala-Médioni, A., Rostek, F., Guilhaumou, N., Pranal, V. & Castrec, M. (1996): Fluid flow in and around of a mud volcano field seaward of the Barbados accretionary wedge: results from Manon cruise. J. Geophys. Res., 101:20,297-20,323. Hirajima, T. & Nakamura, D. (2003): The Dabie ShanSulu orogen. In Carswell, D.A. & compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Mineral., 5/. Budapest: Eötvös Univ. Press, 365–414 Holdaway, M.J. (1971): Stability of andalusite and the aluminium silicate phase diagram. Am. J. Sci., 271:97–131. Holland, T.J.B. (1980): The reaction albite = jadeite + quartz determined experimentally in the range 600–1200 °C. Am. Mineral., 65:129–134. Holland, T.J.B. (1983): The experimental determination of activities in disordered and short-range ordered jadeite pyroxene. Contrib. Mineral. Petrol., 82:214–220. Holloway, R.J. (1981): Compositions and volumes of supercritical fluids in the Earth’s crust. In Hollister, L.S. & Crawford, M.L. (eds.): Fluid inclusions: Applications to petrology /MAC Short Course Handb., 6/. Ottawa (Ont.): Mineral. Assoc. Can., 13–18. Jahn, B.M., Rumble, D. & Liou, J.G. (2003): Geochemistry and isotope tracer study of UHP metamorphic rocks. In Carswell, D.A. & Compagnoni, R. (eds.): Ultrahigh pressure metamorphism /EMU Notes Minerals., 5/. Budapest: Eötvös Univ. Press, 105–144. Jamtveit, B., Austrheim, H. & Malthe-Sørenssen, A. (2000): Accelerated hydration of the Earth’s deep crust induced by stress perturbations, Nature, 408:75–78. Klemd, R. (1989): P-T evolution and fluid inclusion characteristics of retrograded eclogites, Münchberg gneiss Complex, Germany. Contrib. Mineral. Petrol., 102:221–229. Klemd, R., van den Kerkhof, A.M. & Horn, E.E. (1992): High-density CO2-N2 inclusions in eclogite facies metasediments of the Münchberg gneiss complex. Contrib. Mineral. Petrol., 111:409–419. Kostenko, O., Jamtveit, B., Austrheim, H., Polok, K. & Putnis, C. (2002): The mechanism of fluid infiltration in peridotites at Almklovdalen, Western Norway. Geofluids, 2:203–215. Kuster, M. & Stockhert, B. (1997): Density changes of fluid inclusions in high-pressure low-temperature metamorphic rocks from Crete. A thermobarometric approach based on the creep strength of the host minerals. Lithos, 41:151–167. Liou, J.G., Zhang, R.Y. & Jahn, B.M. (1997): Petrology, geochemistry and isotope data on a ultra high pressure jadeite quartzite from Shuanghe, Dabie Mountains, East-central China. Lithos, 41:59–78. Philippot, P. & Selverstone, J. (1991): Trace element rich brines in eclogitic veins: implications for fluid composition and transport during subduction. Contrib. Mineral. Petrol., 106:417–430. Philippot, P., Chevallier, P., Chopin C. & Dubessy, J. (1995): Fluid composition and evolution in the coesitebearing rocks of the Dora Maira Massif, western Alps. Contrib. Mineral. Petrol., 121:29–44. Philippot, P., Agrinier, P. & Scambelluri, M. (1998): Chlorine cycling in the subducted oceanic lithosphere. Earth Planet. Sci Lett., 161:33–44.

Fluid inclusions in high pressure and ultrahigh pressure metamorphic rocks 487 Roedder, E. (1984): Fluid inclusions /Rev. Mineral., 12/. Washington (D.C): Mineral. Soc. Am., 651 p. Rumble, D. & Yui, T.F. (1998): The Qinglongshan oxygen and hydrogen isotope anomaly near Donghai in Jiangsu Province, China. Geochim. Cosmochim. Acta, 62:3307–3321. Scambelluri, M. & Philippot, P. (2001): Fluids in subduction zones. Lithos, 55:213–227. Scambelluri, M., Piccardo, G.B., Philippot, P., Robbiano, A. & Negretti, L. (1997): High salinity fluid inclusions formed from recycled seawater in deeply subducted alpine serpentinites. Earth Planet. Sci. Lett., 148:485–500. Scambelluri, M., Philippot, P. & Pennacchioni, G. (1998): Salt-rich aqueous fluids formed during eclogitization of metabasalts in the alpine continental crust (Austroalpine Mount Emilius unit, Italian western Alps). Lithos, 43:151–161. Scambelluri, M., Bottazzi, P., Trommsdorff, V., Vannucci, R., Hermann, J., Gómez-Pugnaire, M.T. & LópezSánchez Vizcaíno, V. (2001): Incompatible element-rich fluids released by antigorite breakdown in deeply subducted mantle. Earth Planet. Sci. Lett., 192:457–470. Sharp, Z.D., Essene, E.J. & Hunziker, J.C. (1993): Stable isotope and phase equilibria of coesite bearing whiteschists. Dora Maira massif, western Alps. Contrib. Mineral. Petrol., 114:1–12. Stockerth, B., Duyster, G., Trepman, C. & Massonne, H.G. (2001): Microdiamond daughter crystals precipitated from supercritical COH + silicate fluids included in garnet, Erzgebirge, Germany. Geology, 29:391–394. Touret, J.L.R. (1977): The significance of fluid inclusions in metamorphic rocks. In Fraser, D.G. (ed.): Thermodynamics in geology /NATO ASI Ser. C, 28/. Dordrecht: Reidel, 203–227. Touret, J.L.R. (1981): Fluid inclusions in high grade metamorphic rocks. In Hollister, L.S. & Crawford, M.L. (eds.): Fluid inclusions: Applications to petrology /MAC Short Course Handb., 6/. Ottawa (Ont.): Mineral. Assoc. Can., 182–208. Touret, J.L.R. (1987): Fluid inclusions and pressure-temperature estimates in deep seated rocks. In Helgeson, H.C. (ed.): Chemical transport in metasomatic processes /NATO ASI Ser. C, 218/. Dordrecht: Reidel, 99–122. Touret, J.L.R. (1992): Fluid inclusions in subducted rocks. Proc. K. Ned. Acad. Wet., 95:385–403. Touret, J.L.R. (2001): Fluids in metamorphic rocks. Lithos, 55:1–27. Xiao, Y., Hoefs, J., van den Kerkhof, A. M., Fiebig, J. & Zheng, Y. (2000): Fluid history of UHP metamorphism in the Dabie Shan, China: a fluid inclusion and oxygen isotope study on the coesite-bearing eclogite from Bixiling. Contrib. Mineral. Petrol., 139:1–16. Xiao, Y., Hoefs, J. & van den Kerkhof, A. M. (2001): The Raobazhai complex in North Dabie Shan, China: eclogite- to granulite-facies metamorphism of subducted oceanic crust. J. Metamorph. Geol., 19:3–19. Zhang, Y. & Franz, J.D. (1987): determination of the homogenisation temperature and densities of supercritical fluids in the system NaCl-KCl-CaCl2-H2O using synthetic fluid inclusions. Chem. Geol., 64:335–350. Zheng, Y.-F., Fu, B., Xiao, Y.L., Li, Y.-L. & Cong, B. (1999): Hydrogen and oxygen isotope evidence for fluidrock interactions in the stages of pre-and post-UHP metamorphism in the Dabie Mountains. Lithos, 46:677–693. Zheng, Y.-F., Fu, B., Li, Y.-L., Wei, C.-S. & Zhou, J.-B. (2001): Oxygen isotope composition of granulites from Dabieshan in eastern China and its implications for geodynamics of Yangtze plate subduction. Phys. Chem. Earth (A), 26:673–684.