August 2011

Volume 7, Number 4

ISSN 1811-5209

When the Continental

Crust Melts
EDWARD W. SAWYER, BERNARDO CESARE, and MICHAEL BROWN, Guest Editors

How the Crust Gets Really Hot

Melting Experiments and
Thermodynamic Calculations
Interpreting Microstructures
Crustal Melting and the Flow
of Mountains
Melt Flow through
the Crust
Elements is published jointly by the Mineralogical
Society of America, the Mineralogical Society
of Great Britain and Ireland, the Mineralogical
Association of Canada, the Geochemical Society,
The Clay Minerals Society, the European
Association of Geochemistry, the International
Association of GeoChemistry, the Socit
Franaise de Minralogie et de Cristallographie,
the Association of Applied Geochemists,
the Deutsche Mineralogische Gesellschaft,
the Societ Italiana di Mineralogia e Petrologia,
the International Association of Geoanalysts,
the Polskie Towarzystwo Mineralogiczne
(Mineralogical Society of Poland), the Sociedad
Espaola de Mineraloga, the Swiss Society of
Mineralogy and Petrology, and the Meteoritical
Society. It is provided as a benefit to members of
these societies.
Elements is published six times a year. Individuals
are encouraged to join any one of the partici-
pating societies to receive Elements. Institutional
subscribers to any of the following journals
American Mineralogist, Clay Minerals, Clays and
Clay Minerals, Mineralogical Magazine, and The
Canadian Mineralogistalso receive one copy
of Elements as part of their 2011 subscription.
Institutional subscriptions are available for
US$160 (US$175 non-US addresses) a year in
2011. Contact the managing editor (tremblpi@
ete.inrs.ca) for information.
Copyright 2011 by the Mineralogical Society
of America
All rights reserved. Reproduction in any form,
including translation to other languages, or by
any meansgraphic, electronic or mechanical,
including photocopying or information storage
and retrieval systemswithout written permission
from the copyright holder is strictly prohibited.
Publications mail agreement no. 40037944
Printed in USA
ISSN 1811-5209 (print)
ISSN 1811-5217 (online)
www.elementsmagazine.org
www.elements.
geoscienceworld.org
Volume 7, Number 4 August 2011 ABOUT THE COVER:
Spider Wall on the south face
of Nuptse (the summit ridge
is at ~7650 m, and the wall is
When the Continental Crust Melts ~1700 m in height), showing a
network of leucogranite dykes
Guest Editors: Edward W. Sawyer, Bernardo Cesare, and Michael Brown
in metasedimentary rocks of
the Everest Series (centre)
When the Continental Crust Melts above the Nuptse leucogranite,
229 Edward W. Sawyer, Bernardo Cesare, visible at the bottom left and
right. Leucogranites emplaced
and Michael Brown in the shallow crust are the end
product of melting of the deep
crust in orogenic belts. The
view is from Pokalde Peak in
the Khumbu Himalaya, Nepal.
How Does the Continental Crust
235 Get Really Hot?
IMAGE COURTESY OF MICAH JESSUP,
UNIVERSITY OF TENNESSEE, USA
Chris Clark, Ian C. W. Fitzsimons, David
Healy, and Simon L. Harley
p q
kfs pl liq
pl liq
Is the Crucible Reproducible?
241
liq
q
grt bt kfs pl
grt opx bt
l li
Reconciling Melting Experiments
xp
op
grt bt pl liq grt opx kfs pl liq
t
gr
with Thermodynamic Calculations
grt opx crd kfs pl liq opx pl liq Richard W. White, Gary Stevens, and Timothy E. Johnson
Melted Rocks under the Microscope:
247 Microstructures and Their Interpretation
Marian B. Holness, Bernardo Cesare, and Edward W. Sawyer
Crustal Melting and the Flow of Mountains
253 Rebecca A. Jamieson, Martyn J. Unsworth, Nigel B. W. Harris,
Claudio L. Rosenberg, and Karel Schulmann
Organizing Melt Flow through the Crust
261 Michael Brown, Fawna J. Korhonen, and Christine S. Siddoway
D E PA R T M E N T S
Editorial Is Science a Contact Sport? . . . . . . . . . . . . . . . . 219
From the Editors John Valley, Principal Editor 20122014. . 220
The Elements Toolkit Smashing Up Stones . . . . . . . . . . . 221
People in the News Williams-Jones, Ferry, Stolper. . . . . . . 222
Meet the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Society News
Swiss Society of Mineralogy and Petrology . . . . . . . . . . . . . . 267
Association of Applied Geochemists . . . . . . . . . . . . . . . . . . .268
Mineralogical Society of Great Britain and Ireland . . . . . . . 269
International Association of GeoChemistry . . . . . . . . . . . . . .270
European Association of Geochemistry . . . . . . . . . . . . . . . .271
The Clay Minerals Society . . . . . . . . . . . . . . . . . . . . . . . . . . .272
The Meteoritical Society . . . . . . . . . . . . . . . . . . . . . . . . . . . .273
Mineralogical Society of America . . . . . . . . . . . . . . . . . . . . 274
International Association of Geoanalysts . . . . . . . . . . . . . . . . 276
Socit Franaise de Minralogie et de Cristallographie . . .277
Geochemical Society . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .278
Mineralogical Association of Canada . . . . . . . . . . . . . . . . . 280
Book Review Timescales of Magmatic Processes . . . . . . . . . . 282
Meeting Report Making Science Matter . . . . . . . . . . . . . 284
Calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Advertisers in This Issue . . . . . . . . . . . . . . . . . . . . . . . . . 286
Parting Shots Standing Stones . . . . . . . . . . . . . . . . . . . . 288
217

The Mineralogical
Society of America is
composed of individuals
interested in mineralogy,
crystallography, petrology,
and geochemistry. Founded
in 1919, the Society promotes,
through education and research, the under-
standing and application of mineralogy by
industry, universities, government, and the
public. Membership benefits include special
subscription rates for American Mineralogist
as well as other journals, a 25% discount on
Reviews in Mineralogy & Geochemistry
series and Monographs, Elements, reduced
registration fees for MSA meetings and short
courses, and participation in a society that
supports the many facets of mineralogy.
SOCIETY NEWS EDITOR: Andrea Koziol
(Andrea.Koziol@notes.udayton.edu)
Mineralogical Society of America
3635 Concorde Pkwy Ste 500
Chantilly, VA 20151-1110, USA
Tel.: 703-652-9950; fax: 703-652-9951
business@minsocam.org
www.minsocam.org
The Mineralogical
Society of Great Britain
and Ireland is an inter-
national society for all
those working in the
mineral sciences. The
Society aims to advance the
knowledge of the science of mineralogy and
its application to other subjects, including
crystallography, geochemistry, petrology,
environmental science and economic
geology. The Society furthers its aims
through scientific meetings and the publica-
tion of scientific journals, books and mono-
graphs. The Society publishes Mineralogical
Magazine and Clay Minerals. Students receive
the first year of membership free of charge.
All members receive Elements.
SOCIETY NEWS EDITOR: Kevin Murphy
(kevin@minersoc.org)
The Mineralogical Society
12 Baylis Mews, Amyand Park Road
Twickenham, Middlesex TW1 3HQ, UK
Tel.: +44 (0)20 8891 6600
Fax: +44 (0)20 8891 6599
info@minersoc.org
www.minersoc.org
The Mineralogical
Association of Canada
was incorporated in 1955
to promote and advance
the knowledge of miner-
alogy and the related disci-
plines of crystallography,
petrology, geochemistry, and economic
geology. Any person engaged or interested
in these fields may become a member of the
Association. Membership benefits include a
subscription to Elements, reduced cost for
subscribing to The Canadian Mineralogist, a
20% discount on short course volumes and
special publications, and a discount on the
registration fee for annual meetings.
SOCIETY NEWS EDITOR: Pierrette Tremblay
(ptremblay@mineralogicalassociation.ca)
Mineralogical Association of Canada
490, de la Couronne
Qubec, QC G1K 9A9, Canada
Tel.: 418-653-0333; fax: 418-653-0777
office@mineralogicalassociation.ca
www.mineralogicalassociation.ca
The Clay Minerals
Society (CMS) began as the
Clay Minerals Committee
of the US National Academy
of Sciences National
Research Council in 1952.
In 1962, the CMS was
incorporated with the primary purpose of
stimulating research and disseminating
information relating to all aspects of clay
science and technology. The CMS holds an
annual meeting, workshop, and field trips,
and publishes Clays and Clay Minerals and
the CMS Workshop Lectures series. Member-
ship benefits include reduced registration fees
to the annual meeting, discounts on the
CMS Workshop Lectures, and Elements.
SOCIETY NEWS EDITOR: Jeffery Greathouse
(jagreat@sandia.gov)
E LEMENTS
PARTICIPATING SOCIETIES
The Clay Minerals Society
3635 Concorde Pkwy Ste 500
Chantilly, VA 20151-1110, USA
Tel.: 703-652-9960; fax: 703-652-9951
cms@clays.org
www.clays.org
The Geochemical Society
(GS) is an international
organization founded in
1955 for students and
scientists involved in the
practice, study and teaching
of geochemistry. Our
programs include co-hosting the annual
Goldschmidt ConferenceTM, editorial over-
sight of Geochimica et Cosmochimica Acta
(GCA), supporting geochemical symposia
through our Meeting Assistance Program,
and supporting student development
through our Student Travel Grant Program.
GS annually recognizes excellence in
geochemistry through its medals, lectures
and awards. Members receive a subscription
to Elements, special member rates for GCA
and G-cubed, and publication and confer-
ence discounts.
SOCIETY NEWS EDITOR: Seth Davis
(seth.davis@geochemsoc.org)
Geochemical Society
Washington University
Earth & Planetary Sciences
One Brookings Drive, Campus Box #1169
St. Louis, MO 63130-4899, USA
Tel.: 314-935-4131; fax: 314-935-4121
gsoffice@geochemsoc.org
Explore GS online at www.geochemsoc.org
The European Association
of Geochemistry was
founded in 1985 to promote
geochemical research and
study in Europe. It is now
recognized as the premiere
geochemical organization
in Europe encouraging interaction between
geochemists and researchers in associated
fields, and promoting research and teaching
in the public and private sectors.
SOCIETY NEWS EDITOR: Liane G. Benning
(L.G.Benning@leeds.ac.uk)
Membership information:
www.eag.eu.com/membership
The International
Association of
GeoChemistry (IAGC) has
been a pre-eminent interna-
tional geochemical organi-
zation for over 40 years. Its
principal objectives are to
foster cooperation in the advancement of
applied geochemistry by sponsoring specialist
scientific symposia and the activities organized
by its working groups and by supporting its
journal, Applied Geochemistry. The administra-
tion and activities of IAGC are conducted by
its Council, comprising an Executive and ten
ordinary members. Day-to-day administration
is performed through the IAGC business office.
SOCIETY NEWS EDITOR: Chris Gardner
(iageochemistry@gmail.com)
IAGC Business Office
275 Mendenhall Laboratory
125 South Oval Mall
Columbus, OH 43210, USA
Tel.: 614-688-7400; fax: 614-292-7688
www.iagc-society.org
The Socit Franaise
de Minralogie et de
Cristallographie, the
French Mineralogy and
Crystallography Society,
was founded on March 21,
1878. The purpose of the
Society is to promote mineralogy and
crystallography. Membership benefits include
the bulletin de liaison (in French), the
European Journal of Mineralogy, Elements, and
reduced registration fees for SFMC meetings.
SOCIETY NEWS EDITOR: Anne-Line Auzende
(Anne-Line.Auzende@impmc.upmc.fr)
SFMC
Campus Boucicaut, Btiment 7
140 rue de Lourmel
75015 Paris, France
www.sfmc-fr.org
218
The Association of
Applied Geochemists is
an international organiza-
tion founded in 1970 that
specializes in the field of
applied geochemistry. It
aims to advance the science
of geochemistry as it relates to exploration
and the environment, further the common
interests of exploration geochemists, facili-
tate the acquisition and distribution of
scientific knowledge, promote the exchange
of information, and encourage research and
development. AAG membership includes
the AAG journal, Geochemistry: Exploration,
Environment, Analysis; the AAG newsletter,
EXPLORE; and Elements.
SOCIETY NEWS EDITOR: Patrice de Caritat
(Patrice.deCaritat@ga.gov.au)
Association of Applied Geochemists
P.O. Box 26099
Nepean, ON K2H 9R0, Canada
Tel.: 613-828-0199; fax: 613-828-9288
office@appliedgeochemists.org
www.appliedgeochemists.org
The Deutsche
Mineralogische
Gesellschaft (German
Mineralogical Society)
was founded in 1908 to
promote mineralogy and
all its subdisciplines in
teaching and research as well as the personal
relationships among all members. Its great
tradition is reflected in the list of honorary
fellows, who include M. v. Laue, G. v.
Tschermak, P. Eskola, C.W. Correns, P.
Ramdohr, and H. Strunz. Today, the Society
especially tries to support young researchers,
e.g. to attend conferences and short courses.
Membership benefits include the European
Journal of Mineralogy, the DMG Forum, GMit,
and Elements.
SOCIETY NEWS EDITOR: Michael Burchard
(michael.burchard@geow.uni-heidelberg.de)
Deutsche Mineralogische Gesellschaft
dmg@dmg-home.de
www.dmg-home.de
The Societ Italiana
di Mineralogia e
Petrologia (Italian Society
of Mineralogy and Petro-
logy), established in 1940,
is the national body repre-
senting all researchers deal-
ing with mineralogy, petrology, and related
disciplines. Membership benefits include
receiving the European Journal of Mineralogy,
Plinius, and Elements, and a reduced registra-
tion fee for the annual meeting.
SOCIETY NEWS EDITOR: Marco Pasero
(pasero@dst.unipi.it)
Societ Italiana di Mineralogia e Petrologia
Dip. di Scienze della Terra
Universit di Pisa, Via S. Maria 53
I-56126 Pisa, Italy
Tel.: +39 050 2215704
Fax: +39 050 2215830
segreteria@socminpet.it
www.socminpet.it
The International Asso-
ciation of Geoanalysts is
a worldwide organization
supporting the professional
interests of those involved
in the analysis of geological
and environmental mate-
rials. Activities include the management of
proficiency testing programmes for bulk rock
and micro-analytical methods, the production
and certification of reference materials and
the publication of the Associations journal,
Geostandards and Geoanalytical Research.
SOCIETY NEWS EDITOR: Michael Wiedenbeck
(michawi@gfz-potsdam.de)
International Association of Geoanalysts
Ms. Jennifer Cook, Hon. Sec.
British Geological Survey
Keyworth, Nottingham, NG12 5GC, UK
http://geoanalyst.org
Affi liated Societies
The International Mineralogical Association,
the European Mineralogical Union, and the
International Association for the Study of Clays are
affi liated societies of Elements. The affi liated status
is reserved for those organizations that serve as an umbrella for other groups in the
fields of mineralogy, geochemistry, and petrology, but that do not themselves have
a membership base.
A UGUS T 2011
The Polskie
Towarzystwo Mineral-
ogiczne (Mineralogical
Society of Poland), founded
in 1969, draws together
professionals and amateurs
interested in mineralogy,
crystallography, petrology, geochemistry,
and economic geology. The Society promotes
links between mineralogical science and
education and technology through annual
conferences, field trips, invited lectures, and
publishing. Membership benefits include
subscriptions to Mineralogia and Elements.
SOCIETY NEWS EDITOR: Zbigniew Sawowicz
(zbigniew.sawlowicz@uj.edu.pl)
Mineralogical Society of Poland
Al. Mickiewicza 30,
30-059 Krakw, Poland
Tel./fax: +48 12 6334330
ptmin@ptmin.pl
www.ptmin.agh.edu.pl
The Sociedad Espaola
de Mineraloga (Spanish
Mineralogical Society) was
founded in 1975 to promote
research in mineralogy,
petrology, and geochem-
istry. The Society organizes
annual conferences and furthers the training
of young researchers via seminars and
special publications. The SEM Bulletin
published scientific papers from 1978 to
2003, the year the Society joined the Euro-
pean Journal of Mineralogy and launched
Macla, a new journal containing scientific
news, abstracts, and reviews. Membership
benefits include receiving the European
Journal of Mineralogy, Macla, and Elements.
SOCIETY NEWS EDITOR: Juan Jimenez Millan
(jmillan@ujaen.es)
Sociedad Espaola de Mineraloga
npvsem@lg.ehu.es
www.ehu.es/sem
The Swiss Society of
Mineralogy and
Petrology was founded in
1924 by professionals from
academia and industry and
by amateurs to promote
knowledge in the fields of
mineralogy, petrology and geochemistry and
to disseminate it to the scientific and public
communities. The Society coorganizes the
annual Swiss Geoscience Meeting and
publishes the Swiss Journal of Geosciences
jointly with the national geological and
paleontological societies.
SOCIETY NEWS EDITOR: Urs Schaltegger
(urs.schaltegger@unige.ch)
Swiss Society of Mineralogy and Petrology
Universit de Fribourg, Dpartement des
Gosciences
Chemin du Muse 6, Prolles 1700
Fribourg, Switzerland
Tel. +41 26 300 89 36;
fax: +41 26 300 97 65
http://ssmp.scnatweb.ch
The Meteoritical Society
is an international organi-
zation founded in 1933 for
scientists, collectors, and
educators to advance the
study of meteorites and
other extraterrestrial mate-
rials and their parent asteroids, comets, and
planets. Members receive our journal, Mete-
oritics and Planetary Science, reduced rates for
Geochimica et Cosmochimica Acta, which we
cosponsor, the Meteoritical Bulletin, and
Elements. We organize annual meetings,
workshops, and field trips, and support
young planetary scientists worldwide.
Through our medals and awards, we recog-
nize excellence in meteoritics and allied
fields.
SOCIETY NEWS EDITOR: Cari Corrigan
(corriganc@si.edu)
MEMBERSHIP INFORMATION:
http://meteoriticalsociety.org

PRINCIPAL EDITORS
HARRY Y. (Hap) McSWEEN, University of
Tennessee, USA (mcsween@utk.edu)
JAMES I. DREVER, University of Wyoming, USA
(drever@uwy.edu)
GEORGES CALAS, IMPMC, France
(Georges.Calas@impmc.jussieu.fr)
ADVISORY BOARD 2011
JOHN BRODHOLT, University College London, UK
NORBERT CLAUER, CNRS/UdS, Universit de
Strasbourg, France
WILL P. GATES, SmecTech Research
Consulting, Australia
GEORGE E. HARLOW, American Museum
of Natural History, USA
JANUSZ JANECZEK, University of Silesia, Poland
HANS KEPPLER, Bayerisches Geoinstitut,
Germany
DAVID R. LENTZ, University of New Brunswick,
Canada
ANHUAI LU, Peking University, China
ROBERT W. LUTH, University of Alberta, Canada
DAVID W. MOGK, Montana State University, USA
TAKASHI MURAKAMI, University of Tokyo, Japan
ROBERTA OBERTI, CNR Istituto di Geoscienze
e Georisorse, Pavia, Italy
TERRY PLANK, Lamont-Doherty Earth
Observatory, USA
XAVIER QUEROL, Spanish Research Council, Spain
MAURO ROSI, University of Pisa, Italy
BARBARA SHERWOOD LOLLAR, University of
Toronto, Canada
TORSTEN VENNEMANN, Universit de
Lausanne, Switzerland
OLIVIER VIDAL, Universit J. Fourier, France
MEENAKSHI WADHWA, Arizona State
University, USA
BERNARD WOOD, University of Oxford, UK
JON WOODHEAD, University of Melbourne,
Australia
EXECUTIVE COMMITTEE
CARLOS AYORA IBEZ, Sociedad Espaola
di Mineraloga
LIANE G. BENNING, European Association
of Geochemistry
THOMAS D. BULLEN, International Association
of GeoChemistry
PETER C. BURNS, Mineralogical Association
of Canada
GIUSEPPE CRUCIANI, Societ Italiana di
Mineralogia e Petrologia
BARBARA L. DUTROW, Mineralogical
Society of America, Chair
W. CRAWFORD ELLIOTT, The Clay Minerals Society
MONICA M. GRADY, The Meteoritical Society
BERNARD GROBTY, Swiss Society of
Mineralogy and Petrology
GUY LIBOUREL, Socit Franaise
de Minralogie et de Cristallographie
MAREK MICHALIK, Mineralogical Society
of Poland
EDWIN A. SCHAUBLE, Geochemical Society
CLIFFORD R. STANLEY, Association
of Applied Geochemists
PETER TRELOAR, Mineralogical Society
of Great Britain and Ireland
FRIEDHELM VON BLANCKENBURG,
Deutsche Mineralogische Gesellschaft
MICHAEL WIEDENBECK, International
Association of Geoanalysts
MANAGING EDITOR
PIERRETTE TREMBLAY, tremblpi@ete.inrs.ca
EDITORIAL OFFICE
490, rue de la Couronne
Qubec (Qubec) G1K 9A9, Canada
Tel.: 418-654-2606 Fax: 418-653-0777
Layout: POULIOT GUAY GRAPHISTES
Copy editor: THOMAS CLARK
Proofreaders: THOMAS CLARK
and DOLORES DURANT
Printer: ALLEN PRESS
The publishers assume no responsibility for
any statement of fact or opinion expressed
in the published material. The appearance of
advertising in this magazine does not constitute
endorsement or approval of the quality or value
of the products or of claims made for them.
www.elementsmagazine.org
E LEMENTS
EDITORIAL
IS SCIENCE A CONTACT SPORT? new offerings, panel members who fairly decide
which research proposals are most deserving of
funding, members of advisory boards that thought-
fully set science priorities when everything cant
be supported. Our referees ensure that we play by
the rules, which for science means that truth and
accuracy should ultimately win.
For the most part, being a referee is a difficult
and thankless job, an uncompensated duty that
we assume for the betterment of our shared sci-
entific community. In science, virtually all of us,
Hap McSween sooner or later, get to be referees. Few of us are
actually trained for these responsibilities, though,
My university plays (American) footballthese
and that is probably unfortunate. Those of us who
are big-time contests, held in a stadium that seats
teach need to share with our students the rules
more than a hundred thousand spectators and
and techniques by which we referee, and instill
televised more often than not. Watching a game
in them an appreciation for the trust we place in
not long ago, I was taken aback by the crowds
referees and a sense of respect for its paramount
vociferous reaction to a referees ruling against
importance to science.
the home team. Based on an instant
replay projected onto a gigantic It is human nature to be loyal to the
The loyal crowd just
screen at the top of the stadium, home team, to our colleagues and
the ruling seemed fair to me. The wanted their team to friends, and even to our scientific
referee was close to the action and win, and any rulings passions. Loyalty colors the way we
obviously had some experience in respond to our pastimes, our poli-
by the guy in the tics, and our professions. It seems
such matters. But the crowd around
me, none of whom I assume actu- striped shirt that did to me that in the past few decades,
ally had any experience refereeing, not further that cause various factions of society have
was furious at the call and roared learned to express their loyalty in
its displeasure. Later in the game,
were roundly booed. some angry and less-than-produc-
a similar ruling against the oppo- tive ways, such as rudely booing the
nent resulted in the crowds jubilation. I doubt opposing team and polarizing our political discus-
that this was their vocal affirmation of excellence sions. I dont sense that this societal hardening
and accuracy in officiating. A good call was, by has spilled over into science yet. But we should
definition, one that favored the home team. The guard against it. My hope, and my expectation, is
loyal crowd just wanted their team to win, and that the community of scientists will continue to
any rulings by the guy in the striped shirt that did conduct their sparring contests with respect, and
not further that cause were roundly booed. This to value our referees and spare them the jeers that
is American football, but the same fan behavior greet the guys in the zebra shirts on game day.
(or worse) is a part of the other football played Hap McSween, University of Tennessee
around the world. mcsween@utk.edu
We all understand this, of course. Sports are all
about winning, for the players and for the specta-
tors, and it is easy to get caught up in the game
and lose track of the ideal of sportsmanship. The
reason I bring this up is that science is some-
times described as a contact sport. In science,
competing ideas often collide. We are obligated
to champion our hypotheses, at least until they
are proven wrong, and to marshal evidence against
hypotheses with which we disagree. This is the
way science works. Our contests are based on ideas
rather than brawn, but sometimes they can get
confrontational or even nasty.
Science, too, thankfully has its referees: reviewers
who provide insightful criticisms of manuscripts,
editors who adjudicate when reviewers dont agree, Football referee signaling a touchdown. PHOTO YOBRO10 |
book reviewers who provide valuable insights into DREAMSTIME.COM
219 A UGUS T 2011
 FROM THE EDITORS
THIS ISSUE
What started as a proposal on the traditional aspects of migmatites
evolved into When the Continental Crust Melts after the proposers
were challenged by the editors to think big and show the relevance of
their work to other disciplines. The focus became the impact of partial
melting on processes ranging from grain scale to crustal scale. As for
all issues, the guest editors worked hard with their international cast of
authors to bring you six stimulating papers.
JOHN VALLEY, PRINCIPAL EDITOR 20122014
John Valley has accepted our invitation to join the editorial team,
starting officially in January 2012. He will replace Hap McSween, whose
term ends at the end of 2011. We will welcome John formally in the
first issue of 2012. In the meantime, he is being integrated into the team
and participates in all discussions.
IMPACT FACTOR 2010
Elements 2010 impact factor was 3.105. Interestingly, Elements five-year
impact factor is 3.561. This probably reflects the fact that articles are
cited over several years.
The most cited articles from the time of publication to July 2011 are:
Geisler T, Schaltegger U, Tomaschek F (2007) Re-equilibration of
zircon in aqueous fluids and melts. Elements 3: 43-50 (70 citations)
Harley SL, Kelly NM, Moller A (2007) Zircon behaviour and the
thermal histories of mountain chains. Elements 3: 25-30 (64)
Charlet L, Polya DA (2006) Arsenic in shallow, reducing ground-
waters in southern Asia: An environmental health disaster.
Elements 2: 91-96 (59)
Cartigny P (2005) Stable isotopes and the origin of diamond.
Elements 1: 79-84 (51)
E LEMENTS
Morin G, Calas G (2006) Arsenic in soils, mine tailings, and former
industrial sites. Elements 2: 97-101 (42)
Ohtani E (2005) Water in the mantle. Elements 1: 25-30 (36)
Self S, Thordarson T, Widdowson M (2005) Gas fluxes from flood
basalt eruptions. Elements 1: 283-287 (32)
Lumpkin GR (2006) Ceramic waste forms for actinides. Elements 2:
365-372 (32)
ODay PA (2006) Chemistry and mineralogy of arsenic. Elements 2:
77-83 (31)
Rubatto D, Hermann J (2007) Zircon behaviour in deeply subducted
rocks. Elements 3: 31-36 (30)
Bruno J, Ewing RC (2006) Spent nuclear fuel. Elements 2: 343-349 (30)
NUCLEAR FUEL CYCLE ISSUE
I mentioned in the April issue that, after the Fukushima nuclear acci-
dent, we made the Nuclear Fuel Cycle issue freely available on our
GeoScienceWorld site (www.elements.geoscienceworld.org) and on
Elements website at www.elementsmagazine.org. This was advertised
as widely as possible across our network, thanks to the efforts of Barb
Dutrow and the members of the Executive Committee. Did it work?
Yes, there was a spike of at least one order of magnitude in downloads
from GeoScienceWorld for all articles in that issue. The article Spent
Nuclear Fuel was downloaded 35 times more than in previous months.
FACEBOOK
At the time of writing, we had gained 182 followers on Facebook in less
than two months. If you have a Facebook account, do like us. We will
keep you posted on when issues are taken to press and mailed, and we
will share timely news. Go to www.facebook.com/elementsmagazine.
Pierrette Tremblay, Managing Editor
220 A UGUS T 2011
SMASHING UP STONES
A critical, but often neglected, aspect of the entire geoanalytical pro-
cess involves how one actually gets ones sample from the field and
back to the laboratory for analysis. Clearly collecting material that is
representative of the process being studied is the fi rst critical step. But
what does one do once the specimen arrives back home? I can well
remember the many hours I spent as a graduate student some decades
ago in the crushing lab reducing kilograms of sample down to grams
of representative powder; this powder would be the starting mate-
rial for my assigned tasks as a budding geochemist. Jaw mill to reduce
to centimetre-size puck and ring mill to reduce to coarse powder
agate ball mill to reduce to fi ne powder hours of processing keep
everything clean dont contaminate, dont fractionate Tedious, to
say the least.
So what technological progress has this aspect of mineralogy/geochem-
istry seen of late? Over roughly the past decade interest has grown in
the use of electrodynamic disaggregation. In this procedure, a high-
voltage electrical impulse creates a shockwave either within the material
itself or within the fluid medium typically water that surrounds
it. This method of using lightning strikes to reduce walnut-size rock
chips down to individual mineral grains fi rst came to my attention
at the Goldschmidt 2009 meeting in Davos, Switzerland, where hard-
ware from the Swiss company Selfrag AG (www.selfrag.com) was on
display. Subsequently a paper by Giese et al. (2010) described in detail
the physical process involved in the various forms of electrodynamic
disaggregation and also demon-
strated that the high temperatures
that briefly affect the sample do not
bias apatite fission-track ages.
Here I would like to describe briefly
an alternative technology which
I learned about a year or so ago.
It is being developed by Zybek
Advanced Products (www.zapmate-
rials.com), a small company located
in Boulder, Colorado, USA. Zybeks
Aerody namic Impact Reactor
(FIG. 1) employs a high-pressure air
FIGURE 1
Model of the stream created by a series of impel-
Aerodynamic Impact lers. Within the reactor chamber,
Reactor (patent pending). The reactors
height is approximately 2 metres. the airflow is directed into a vortex
geometry with a high-speed pneu-
Diagram of the individual components within the Aerodynamic Impact
FIGURE 2 Reactor processor integrated with cyclone separators
E LEMENTS
An example of a phosphate starting material and end product from the
FIGURE 3 Aerodynamic Impact Reactor
matic flow established along the margin of the vessel. Within seconds
centimetre-size rock chips introduced into this environment undergo
a grain-size reduction through collisions within the reactor. The pro-
cessed material is ultimately ejected through a port in the base of the
reaction chamber (FIG. 2). The device has been integrated with multiple
cyclone separators, which allow the processed material to be binned
by grain size and which also remove particulates down to roughly
1 m grain size from the exhaust air. A number of parameters can be
adjusted on this apparatus, but it is commonly set to produce grain
size fractions smaller than 100 or even 50 microns. By reprocessing
the coarser-grained materials separated by the cyclone, it is possible to
produce ultimately a very fi ne-grained end product (FIG. 3).
So what are the advantages of this new approach to sample processing?
Though I have yet to see any concrete data, the method is supposed
to be relatively contamination free. High processing rates of up to sev-
eral metric tons per hour could be of interest to the mining industry.
Compared to some of the other competing methods, the Aerodynamic
Impact Reactor is energy efficient, meaning lower operating costs. It is
flexible in terms of the grain-size distribution it can produce and the
nature of the feed stock. In fact, the method has been applied to the
processing of coal and even switch grass (FIG. 4). It tends to liberate
material along grain boundaries, but it also has been found to favour
high surface-to-volume ratios for the end product a benefit if subse-
quent chemical treatment is planned.
Scanning electron microscope images of two of the more unusual
FIGURE 4 material types on which the device has been used
What is the future of this technology in either the mining industry or
basic research? This is hard to say as the technology has not yet been
widely disseminated. Time will tell.
Michael Wiedenbeck, (michawi@gfz-potsdam.de)
Helmholtz-Zentrum Potsdam
REFERENCE
Giese J, Seward D, Finlay MS, Wthrich E, Gnos E, Kurz D, Eggenberger U,
Schreurs G (2010) Electrodynamic disaggregation: Does it affect apatite
fission-track and (U-Th)/He analyses? Geostandards and Geoanalytical
Research 34: 39-48
221 A UGUS T 2011
 PEOPLE IN THE NEWS

ANTHONY E. WILLIAMS- MGPV-sponsored session entitled Turning Up the Heat: Metamorphic

JONES, LOGAN MEDALIST Perspectives on Mineral Equilibria, Heat Transport, Tectonics, and
Thermochronology (T35).
During last Mays joint annual
meeting of the Geological Association Dr. Ferry is cited for his contributions to the theory of fluidrock interac-
of Canada and the Mineralogical tions. His science is fundamentally field-based and his tools predomi-
Association of Canada in Ottawa, nantly chemical and petrological, but the insights gained have signifi-
Anthony E. Willy Williams- cant physical implications. His systematic evaluation of the role of fluid
Jones of McGill University received migration during regional metamorphism in many field areas worldwide
the Logan Medal, the Geological forms the basis for understanding the long-term permeability of the
Association of Canadas highest award middle to lower crust. He conducts careful, systematic field studies to test
and presented to an individual for his and others models. These studies have demonstrated that different
sustained distinguished achievement models best explain observations from individual field sites: nature is
in Canadian Earth science. We reproduce excerpts of the citation below. not as simple as any single model. On Earth, crustal devolatilization
during metamorphism influences the mechanical strength and thermal
For thirty years, Willys influence on Canadian economic geology, and
structure of the continents and contributes to element cycling between
Earth science in general, has been profound. Willy is an extremely cre-
the Earth, the atmosphere, and the oceans. Until the pioneering field
ative and innovative researcher. The scope and breadth of his research
studies of John Ferry and his students and colleagues, these impacts
into the genesis of mineral deposits is unparalleled, not only in the array
could not be realistically quantified.
of types of mineral deposits that he and his group have tackled but also
in the methodologies applied and the approaches taken. His studies of
ED STOLPER ELECTED
mineralizing systems have included sediment-hosted base metal min- FOREIGN MEMBER OF
eralization, uranium, porphyry CuMo, granitoid-related WSnMo, THE ROYAL SOCIETY
pegmatite and hydrothermal rare-element mineralization, epithermal
Edward Stolper is the William E.
precious metals, asbestos, fluorite, Archean gold, and modern geothermal
Leonhard Professor of Geology and
systems. He and his group made a major research breakthrough recently
Provost, Division of Geological and
by showing that both petroleum and high-temperature vapor have the
Planetary Sciences at Caltech. He was
ability to transport important quantities of ore-forming metals.
recently elected as a Foreign Member
Although Willys contributions to economic geology are remarkable of the Royal Society. Each year 44
indeed, they are even more so when considered in the context of his fun- Fellows, 8 Foreign Members and up
damental contributions to other fields, including environmental geology, to 1 Honorary Fellow are elected from a group of over 700 candidates
biomineralization, igneous petrology, and volcanology. In many varied who are proposed by the existing Fellowship.
ways, Willy is a truly exceptional teacher and mentor. He has produced
Ed Stolper is renowned for his experimental and theoretical work on
legions of outstanding graduate and undergraduate students, inspiring
melting and igneous processes on the Earth, Mars and asteroids. He
them to follow careers in Earth science.
was the first to propose that the SNC meteorites came from Mars. He
developed the sandwich method of multiple saturation, which enabled
JOHN FERRY, 2011
quantification of melting in the mantles of Earth and other planets. He
DISTINGUISHED GEOLOGIC
CAREER AWARDEE developed the first quantitative model of water speciation in glasses
and silicate melts and showed that water dissolves both as OH and as
John M. Ferry, Johns Hopkins
molecular H2O. He was the first to show that silicate crystals float in
University, is the 2011 Distinguished
their melts at very high pressures, with implications for the differentia-
Geologic Career Awardee of the
tion of large silicate planets. He was the first to demonstrate a linear
Mineralogy, Geochemisty, Petrology,
relationship between the extent of melting and source water contents
and Volcanology (MGPV) Division
in the back-arc environment.
of the Geological Society of America.
The award will be presented during
the 2011 GSA Annual Meeting,
Minneapolis, Minnesota, USA. The
presentation will take place at the MGPV Reception (held jointly with
the Mineralogical Society of America and the Geochemical Society).
Dr. Ferry will give the Distinguished Geologic Career Award Lecture,
When the Heat Is Turned Up, Look Out for the Hot Water, at the

E LEMENTS 222 A UGUS T 2011

www.wiley.com
 Michael Brown held academic appointments in
the UK until 1990 when he moved to the University
of Maryland as Professor of Geology and Chair. His
research interests are in high-temperature metamor-
phic petrology, crustal melting, and tectonic
geology. His work on migmatites and associated
granites has furthered our understanding of how
heat and mass are transferred within continents,
and in particular the relationships between sources and sinks of melt
and the importance of melt in the tectonic evolution of orogenic belts.
With time his research has broadened into ultrahigh-temperature and
high-pressure metamorphism, the origin of paired metamorphic belts,
and secular change.
Bernardo Cesare is a professor of petrology at the
University of Padova (Italy), where he earned a PhD
in geology in 1992 studying the Vedrette di Ries
contact aureole. His research interests include low-
pressure metamorphism and anatexis of pelitic
rocks, fluidmeltrock interactions in graphitic sys-
tems, fluid and melt inclusions, and the crystal
chemistry of high-temperature minerals. His
approach to crustal melting involves primarily the multidisciplinary
study of anatectic enclaves from SE Spain and of nanogranite inclu-
sions in migmatites and granulites. During the last decade he was deeply
involved in coordinating the international training of early-stage
researchers in metamorphic petrology.
Chris Clark is a senior research fellow in metamor-
phic geology and geochronology at Curtin
University in Western Australia. His principal
research interests are in the linking of geochro-
nology, specifically the UPb method using zircon
and monazite, with the development of metamor-
phic assemblages in order to constrain the dura-
tions of mountain-building events in high-grade
metamorphic terranes.
Ian C. W. Fitzsimons is a professor of metamorphic
geology at Curtin University in Perth, Western
Australia. After an undergraduate degree at the
University of Cambridge, he completed a PhD on
granulite facies metamorphism at the University of
Edinburgh, followed by research positions at Royal
Holloway University of London, the University of
Edinburgh, and Monash University. He moved to
Perth in 1998, where he focuses his research on the field geology, min-
eralogy, petrology, and geochronology of metamorphic rocks, particu-
larly the Precambrian granulites of Antarctica, India, and Madagascar.
Simon L. Harley is Professor of Lower Crustal
Processes at the University of Edinburgh, Scotland.
He has over 30 years experience in metamorphic
and experimental geology and geochemistry, which
he has applied to understanding the high-temper-
ature processes that take place during the evolution
of continents. His approach emphasizes their chem-
ical, isotopic, and petrographic records as preserved
in minerals and mineral assemblages. He is internationally recognized
as a leading authority on ultrahigh-temperature metamorphism and
granulites and their implications for continental evolution.
E LEMENTS 226
Nigel B. W. Harris, a graduate of the University of
Cambridge, is a petrologist and geochemist who
has studied the causes and consequences of melt
production in tectonically thickened crust since his
first Tibetan field campaign in 1985. He was
appointed Professor of Tectonics at the Open
University (UK) in 2001. In recent years his work
has focused on chemical proxies for global weath-
ering fluxes and the linkage between tectonics, orography, and climate
in the Himalaya.
David Healy is Lecturer in Geomechanics at the
University of Aberdeen. He has research interests
in structural geology, rock mechanics, tectonics,
and metamorphic geology. He has a keen interest
in the theory of natural rock deformation and uses
quantitative models to explore the consequences
of theoretical predictions in terms of field and labo-
ratory observations.
Marian B. Holness studied for both her degrees at
the University of Cambridge. After periods at the
University of Chicago and the University of
Edinburgh, she returned to Cambridge in 1997 to
take up a teaching post. Her interests are primarily
in decoding the record of rock history left behind
in grain-scale fabrics. She has progressed from
working on volatile fluids in metamorphic rocks,
through the partial melting of high-grade metamorphic rocks, and is
now investigating the complex problem of solidification, in particular
of gabbros.
Rebecca A. Jamieson is a graduate of Memorial
University of Newfoundland (PhD 1979) and is cur-
rently Carnegie Professor and Chair of Earth
Sciences at Dalhousie University in Halifax, Canada.
She studies interactions between metamorphic and
tectonic processes at all scales using a variety of
approaches, including field work, petrology, geo-
chronology, and geodynamic modeling. She has
worked on parts of the AppalachianCaledonian, Grenvillian, and
HimalayanTibetan orogenic belts. Her recent work has focused on the
causes and consequences of melting and ductile flow in orogens and on
the exhumation of ultrahigh-pressure metamorphic rocks.
Timothy E. Johnson is a postdoctoral research sci-
entist at the University of Mainz. He received his
BSc (1992) and PhD (1999) from the University of
Derby and held postdoctoral positions at the uni-
versities of Graz and Maryland before moving to
Germany. His expertise is in metamorphic geology
and mineral equilibria modeling of subsolidus and
suprasolidus rocks of varying compositions and
from a variety of tectonothermal environments. He has a particular
interest in the generation and segregation of melt and its consequences
for the compositional, thermal, and rheological evolution of the crust.
Contd on page 227
A UGUS T 2011
Contd from page 226

Fawna J. Korhonen is a research fellow at Curtin
University of Technology, Australia. She moved to
this position in 2009 after a three-year postdoctoral
appointment at the University of Maryland. Fawna
received her BA from Carleton College and her PhD
from the University of Minnesota. Her research
interests include the study of polyphase high-grade
metamorphism and anatexis, and the processes of
mass transfer within the crust that lead to intracrustal differentiation
during orogenesis. She is particularly interested in the chemical and
physical effects of melting and melt loss during protracted crustal evolu-
tion, and the genetic link between residual granulites, migmatites, and
granites.
Claudio L. Rosenberg graduated from the State
University of Milano (Italy) in 1989, where he com-
pleted an MSc thesis on the growth of K-feldspar
megacrysts in granites. After a PhD in Basel
(Switzerland) on the emplacement and solid-state
flow of the Bergell pluton (Central Alps), he moved
to Giessen (Germany) where he studied the fabrics
of synkinematic, partially melted rock using experi-
mental analogues. Based on these experiments and studies of natural
migmatitic fabrics, he described the modes of localization and melt
segregation and the rheological changes during deformation of melting
crust. He is now at the Freie Universitt Berlin (Germany), where he
works on the syncollisional exhumation of the Alpine chain.
Edward W. Sawyer received his first degree from
the University of Southampton. He then worked
for the Geological Survey of South Africa in Namibia
for six years and obtained a masters degree from
the University of Cape Town. He then moved to
Canada and received a PhD from the University of
Toronto, followed by postdoctoral research at the
Geological Survey of Norway. He returned to
Canada in 1986 to take a post at the Universit du Qubec Chicoutimi,
where he is now a professor. His principal research interest is in mig-
matites and the segregation and migration of anatectic melt in the con-
tinental crust.
Gary Stevens graduated with BSc and MSc degrees
from Rand Afrikaans University in Johannesburg
and received his PhD degree from the University
of Manchester in 1995. Following this, he spent 5
years at the University of the Witwatersrand, where
the Economic Geology Research Unit kindly
indulged his research interests in petrology. For the
past 10 years he has been employed at Stellenbosch
University, where he holds the position of South African Research Chair
in Experimental Petrology. His main research interests are the origins
of the continental crust, partial melting of the crust, and the processes
that shape granite chemistry.
Martyn J. Unsworth has been a professor of geo-
physics at the University of Alberta in Edmonton,
Canada, since 2000. He holds a BA in natural sci-
ences (1986) and a PhD in marine geophysics
(1991), both from the University of Cambridge. His
research focuses on the use of electromagnetic geo-
physics to study continental dynamics. His recent
studies have been in the Tibetan Plateau, eastern
Anatolia, Taiwan, and the Canadian Cordillera. He makes use of the
magnetotelluric method to study the composition of the crust and
mantle in regions undergoing deformation. He has also worked on the
use of these geophysical techniques in environmental and geothermal
applications.
Richard W. White is a professor of metamorphic
geology at the University of Mainz, Germany. He
received his BSc and MSc from the University of
Sydney and his PhD from Macquarie University
(Sydney) in 1998. He spent nine years at the
University of Melbourne in several postdoctoral
positions, undertaking mineral equilibria modeling
studies, focusing on partial melting. He then moved
to Germany, where his main interests center on high-temperature meta-
morphic processes, the development of mineral and melt activitycom-
position models, and their application to natural examples. He is cur-
rently an editor of the Journal of Metamorphic Geology.

Karel Schulmann started his career in 1987 at the

Czech Geological Survey in Prague. From 1991 to
2004 he was chair of the Department of Structural
Geology and Petrology of Charles University in
Prague. Since 2004 he has held the position of pro-
fessor of geology and tectonics at the University of
Strasbourg. His research interests include the struc-
tural geology and tectonics of orogenic collisional
systems, metamorphic petrology, metamorphic microstructures and tex-
tures, rock fabrics, orogenic processes such as the exhumation of orogenic
lower crust, the mechanisms of lower crustal flow, and the accretion of
juvenile crust in the Central Asian Orogenic Belt.

Christine S. Siddoway received her BA from

Carleton College and MSc from the University of
Arizona. Following her PhD from the University of
California, Santa Barbara, she completed a Fulbright
postdoctoral research fellowship at the University
of Siena (Italy). She has been an investigator in the
U.S. Antarctic research program since 1990, during
which time she has examined the history of breakup
between West Antarctica and New Zealand, the evolution of the active
margin of East Gondwana, and the deformation and metamorphism of
mid-crustal rocks in transcurrent settings. She is on the faculty of the
Geology Department at Colorado College and served as chair during
20072010.

E LEMENTS 227 A UGUS T 2011

 When the Continental
Crust Melts
Edward W. Sawyer1, Bernardo Cesare2 and Michael Brown3

1811-5209/11/0007-0229$2.50 DOI: 10.2113/gselements.7.4.229

P
artial melting of the continental crust has long been of interest to Ga) continental crust appears to be
petrologists as a small-scale phenomenon. Mineral assemblages in the sl ig ht ly more felsic t ha n
Proterozoic (2.5 0.5 Ga) or
cores of old, eroded mountain chains that formed where continents Phanerozoic (< 0.5 Ga) crust
collided show that the continental crust was buried deeply enough to have (Rudnick and Gao 2003). Thus,
melted extensively. Geochemical, experimental, petrological and geodynamic juvenile material added to the
crust must be modified in order to
modelling now show that when the continental crust melts the consequences b e come cont i nent a l c r ust.
are crustal-scale. The combination of melting and regional deformation is Evidence from modern arcs indi-
critical: the presence of melt on grain boundaries weakens rocks, and weak cates that more felsic compositions
arise because the mafic magmas
rocks deform faster, influencing the way mountain belts grow and how rifts fractionate and because they cause
propagate. Tectonic forces also drive the movement of melt out of the lower the crust to partially melt.
continental crust, resulting in an irreversible chemical differentiation of Consequently, a layer of mafic
cumulate and residual material
the crust. develops at the base of arc crust.
KEYWORDS : continental crust, partial melting, microstructures, As the arc crust thickens, this
metamorphic petrology cumulate and residual part at the
base converts to denser material,
detaches (a process called delami-
INTRODUCTION
nation) and sinks into the mantle. Thus, the bulk composi-
The continental crust is 41.4 km thick on average and tion of the remaining continental crust becomes more
covers 39% of the Earths surface. Information from the felsic. The residual and cumulate material that returns to
isotopic and trace element composition of >4-billion-year- the mantle contains, and hence is enriched by, a small
old (Ga) zircon grains and the evolution of mantle isotopic proportion of felsic melt and becomes the Enriched Mantle
reservoirs indicates that 75%, and possibly more, of the I (EMI) isotopic reservoir (Tatsumi 2005).
continental crust was created before 2.5 Ga (Harrison 2009;
Belousova et al. 2010). Thus, the continental crust is much
longer-lived than oceanic crust and, consequently, has
EVIDENCE THAT THE CONTINENTAL CRUST
acquired considerable complexity. This is reflected in the PARTIALLY MELTED
petrological and structural characteristics of the rocks At the beginning of the last century, extensive mapping
within it. was done in the shield areas of Scandinavia, Canada and
elsewhere. This pioneering work revealed that large parts
The continental crust began to form in the Hadean, more of the continental crust have been metamorphosed to a
than 4.0 billion years ago, first as the mantle differentiated, higher degree and more strongly deformed than adjacent
then from thickened oceanic crust above hotspots and areas. We now know that the structures in these highly
at shallow levels (~15 km) above convergent margins deformed regions are similar to those in modern orogens
(Harrison 2009). Since the late Archean (from ca 2.8 Ga), where continents have collided and that the metamorphic
most new, or juvenile, continental crust has formed in temperature in these regions was high enough (> 700 oC)
magmatic arcs above subduction zones, but about 10% was for large areas to partially melt. Some continental crust
formed where mantle magmas were added to existing crust has experienced repeated episodes of modification by
by hotspots or plumes. If new, juvenile continental crust intense deformation, high-temperature metamorphism and
is formed from mantle magma in magmatic arcs and at partial melting: examples occur in the Grenville Province
hotspots or plumes, then its average composition should of Canada, in southern West Greenland, in the Western
be mafic. It is not. The average composition of the conti- Gneisses of Norway and in East Africa. Different terms are
nental crust is broadly andesitic, although Archean (>2.5 used to describe this modification. It is simply called
reworking by petrologists and structural geologists, but from
a geochemists perspective, it is intracrustal differentiation.
1 Dpartement des Sciences Appliques,
Universit du Qubec Chicoutimi The largest and most intensely reworked regions of conti-
Chicoutimi, Qubec G7H 2B1, Canada nental crust are located where continents collided and
E-mail: ewsawyer@uqac.ca major mountain chains were formed, for example, the East
2 Dipartimento di Geoscienze, Universit di Padova African Orogen. Reworking is not restricted to thickened
Via Gradenigo 6, I-35131 Padova, Italy orogens. Mantle melts emplaced into the continental crust
E-mail: bernardo.cesare@unipd.it at rifts or in large igneous provinces associated with
3 Department of Geology, University of Maryland hotspots can result in high-temperature metamorphism.
College Park, MD 20742-4211, USA Partial melting in such settings can lead to intense, local
E-mail: mbrown@umd.edu

E LEMENTS , V OL . 7, PP. 229234 229 A UGUS T 2011

reworking of the continental crust, but such thermal
reworking is not generally accompanied by intense
deformation.
The deformed and metamorphosed continental crust is not
uniform. The upper part is approximately granodioritic in
composition and is richer in SiO2 and K 2O relative to the
lower part, which is more mafic and richer in Al2O3, FeO,
MgO and CaO (Rudnick and Gao 2003). These differences
as well as the considerable enrichment in light rare earth
elements and the large negative Eu anomaly in the upper
crust relative to the lower crust are best explained by partial
melting, a process that is also called anatexis. Thus intra-
crustal differentiation occurs by partial melting of the
lower part of the continental crust and migration of the
melt to the upper part, leaving the lower crust with a more
mafic and residual bulk composition (FIG. 1 AND 2). In addi-
tion to these geochemical differences, this process imparts
a layered structure to the continental crust, which is
revealed by an increase in seismic P- and S-wave velocities
with depth. Seismic profi les across young continental crust
affected by late Paleozoic collision and mountain building
in western Europe show the same sub-horizontal Moho
and internal velocity structure as old crust in northern
Europe that was reworked by mountain building events in
the Proterozoic and Archean. Thus, the acquisition of a
sub-horizontal layered structured must happen soon after
mountains stop growing. This same basic pattern of modi-
fication to continental crust has been going on since the
late Archean, at least.
The geochemical approach has revealed that the large-scale
process of intracrustal differentiation occurs by partial
melting, but it does not address other concerns, such as
Sill and dike network in stromatic metatexite migma-
FIGURE 1 tite at Maigetter Peak (height 480m) in the Fosdick
Mountains of West Antarctica (762638S, 1463000W). The grain boundaries but segregates from the residual solids along a
image is looking to the SE and was taken from the air (Twin Otter
wing tip in upper right). From the aerial perspective and also upon
close examination in outcrops, intersecting dikes do not appear to
truncate or displace each other; the sills and dikes of granite
crosscut foliation but may be continuous with or discordant to
leucosomes in the migmatite. The leucosomes contain peritectic
garnet and cordierite (see Figure 1 in Brown et al. this issue).
E LEMENTS
the source of the heat for melting, what happens at the
grain scale during anatexis, or how felsic melt moves from
the lower to the upper crust. Nor is it concerned with the
broader consequences of partial melting, such as its effect
on the rheology of the continental crust and how this
affects the way mountain chains are built when continents
collide. These and other questions are the subject of this
issue of Elements on the theme When the Continental
Crust Melts.
TYPES OF MELTING IN
THE CONTINENTAL CRUST
Rock types such as metapelite, metagreywacke and granite
may begin to partially melt when the metamorphic temper-
ature exceeds 650 oC (FIG. 3), and the melt they produce is
granitic in composition. Whether they melt or not and the
quantity of melt produced depend on the availability of
H2O. Melting may occur if H 2O is present as a free fluid in
the pores and grain boundaries of the rock; this is called
H 2O fluid-present melting and takes place at the lowest
temperatures. Melting may also occur when hydrous
minerals (hydrates), such as muscovite, biotite and amphi-
bole, melt incongruently (see glossary); other minerals,
most commonly quartz and feldspar, may also participate
in these melting reactions. Incongruent melting may be
either H 2O fluid-present or, at higher temperature, H 2O
fluid-absent. Crystalline rocks have very low porosity and
so contain very little fluid H 2O; thus the amount of melt
produced from H 2O in the pores is too small to be easily
detected. Consequently, the production of large volumes
of granitic melt in continental crust is widely thought to
occur by fluid-absent incongruent melting, except for
instances where large volumes of aqueous fluid were intro-
duced into rocks already at high temperature, as discussed
below.
Schematic representation of the reworking of conti-
FIGURE 2 nental crust by partial melting. Partial melting occurs
in the lower part of the crust where temperatures exceed the
solidus and migmatites are formed (brown). Melt is formed on
progressively more focussed pathway (shown in red), first through
leucosomes then dykes. The melt collects to form plutons, typically
at the transition from ductile middle crust (yellow) to brittle upper
crust (green); some felsic lavas may be erupted. It is not yet clear
whether melt ascent is uninterrupted or whether melt ponds at
intermediate levels, shown by the question marks. The ascent of
some melt ends in the middle crust as dyke complexes, without
forming plutons.
230 A UGUS T 2011
Pelitic rocks contain a large amount of muscovite and
biotite 30 to 50 vol% is not unusual and will produce
melt progressively as the temperature rises above the
temperatures of the incongruent melting reactions
involving these minerals, typically ~720 oC and ~820 oC,
respectively. Other rock types also undergo fluid-absent
incongruent melting. Metagreywackes and meta-andesites
begin to melt between 750 oC and 800 oC. Amphibolites
follow at about 850 oC, but they produce melt of tonalitic
composition. Fluid-absent incongruent melting of micas
in metapelites and metagreywackes can produce as much
as 50 vol% melt. After all the mica is consumed at about
925 o C, the rate of melt production decreases, and the
composition of the melt is no longer granitic.
Fluid-absent incongruent melting of micas and amphibole
describes the melting of metapelite, metagreywacke and
mafic rocks quite well. It explains both the volumes of melt
generated and the granulite facies, residual mineral assem-
blages found deep in the crust that are left behind after
melt has been extracted. However, it is not a good descrip-
tion of melting in hydrate-poor quartzofeldspathic rocks,
such as leucocratic granites, trondhjemites and tonalites.
Recent studies in metamorphic terranes, ranging in age
from Archean to Phanerozoic, show far higher degrees of
partial melting in granitic rocks than can be accounted for
by H2O in pores or by the breakdown of their mica and
amphibole. Melting in these rocks occurred because an
aqueous fluid infi ltrated them and led to what is called
water-fluxed melting at low temperature, around 700 oC.
Such an influx of H2O is now recognised as being respon-
sible for melting of metapelitic, metapsammitic and
metamafic rocks in some anatectic terranes (Ward et al.
2008; Berger et al. 2008). Oxygen stable isotope studies
reveal diverse sources for this H2O. In some terranes it came
from dehydration reactions in nearby metapelites or from
crystallizing plutons, whereas in others it originated as
deeply penetrating seawater or meteoric water, and in yet
others it came from the mantle. It is not surprising, there-
fore, that many of the places where water-fluxed melting
has occurred in the continental crust are adjacent to major
crustal-scale shear zones that provided the pathways for
the H2O to infi ltrate the continental crust (Sawyer 2010).
THE HEAT PROBLEM
The temperature required for H2O fluidpresent or water-
fluxed melting (700 oC) might be reached as a result of
mantle heat entering the base of the crust and radiogenic
heat generated in a continental crust thickened by orogen-
esis (FIG. 3). However, large granulite terranes that under-
went melting at temperatures well above 850 oC and appear
to have lost substantial volumes (>600,000 km3 for the
Ashuanipi subprovince in Quebec; Guernina and Sawyer
2003) of granitic melt as determined from the composition
of their residual rocks are problematic in that they required
a great deal of heat. The average continental crust does not
contain enough K, Th and U to produce sufficient radio-
genic heat to sustain this degree of melting on the required
timescale. Other sources of heat are required. The mantle
is an obvious source, and strain heating may be significant
in some circumstances. New measurements (Whittington
et al. 2009) indicate that the thermal diffusivity of rocks
at high temperature is low; consequently, the middle and
lower crust may retain heat better than previously thought.
Identifying the source of heat and the combination of
parameters or circumstances required to focus the heat
into thickening crust and produce a high degree of partial
melting remains a major problem. Hence, the article by
Clark et al. (2011 this issue) is the starting point for When
the Continental Crust Melts.
E LEMENTS 231
Types of melting in PT space for continental crust
FIGURE 3 thickened to 71 km. The base of average (41.4 km)
crust is shown by the blue dashed line. The red curve is the
H2O-present solidus in the haplogranite system; subsolidus condi-
tions occur in the yellow field to its left, and partial melting can
occur in the pink field. Fields for melting by hydrate breakdown are
shown: blue for muscovite (Ms), brown for biotite (Bt) and green
for amphibole (Amp). The purple line marks the start of ultrahigh-
temperature (UHT) metamorphism. Two equilibrium geotherms for
crust of normal thickness are shown as dotted black lines. Crustal
radiogenic heat production (0.61 Wm -3) and a mantle heat flux at
the Moho (30 mWm -2) are the same for both, but thermal conduc-
tivity is 3.0 Wm -1K-1 for geotherm A and 2.0 for B; hence
geotherm B is hotter but still does not reach UT conditions.
PETROLOGICAL ASPECTS OF MELTING
THE CONTINENTAL CRUST
The rocks in the continental crust that have partially
melted are called migmatites; the nomenclature specific to
these rocks and the means by which they are identified in
the field are outlined by Sawyer (2008a, b). Migmatites are
basically simple rocks with two components. One, which
is partially melted, is called neosome, and consists of the
crystallized products from the melt and the complementary
residual material. The second, called paleosome, consists of
rock that did not melt. In most cases, however, the melt
and residual solid have segregated from each other,
although not completely. The neosome then consists of
two petrologically different parts, one derived from the
melt and called leucosome, and the other derived from the
residual solid material and, if dark coloured, called melano-
some, otherwise simply residue. In most cases this simple
petrological framework is made morphologically complex
by deformation during the melting process. Deformation
results in the translation, rotation and distortion of the
constituents parts. If the strain is high enough, the migma-
tite becomes attenuated, resulting in a banded or layered
appearance (FIG. 4) typically seen in the deep parts of orogens.
EXPERIMENTS AND PETROGENETIC
MODELLING
The pressure and temperature conditions retrieved from
granulites and migmatites tell us how deep in the conti-
nental crust melting occurred and provide minima that
must be achieved by any proposed mechanism of heating.
Basic information for determining the pressure and temper-
ature (PT) history comes from well-controlled experi-
ments on the partial melting of rocks such as pelite,
greywacke and amphibolite. Phase equilibria modelling
using internally consistent thermodynamic datasets
derived from experiments has now been added to the set
of tools available for understanding the PT conditions for
partial melting in the continental crust. The article by White
et al. (2011 this issue) compares the results from both
A UGUS T 2011
approaches to better understand the conditions and petro-
logical processes that occur when the continental crust melts.
Dating the time of formation of metamorphic minerals
and adding this time constraint to PT information results
in a PTt trajectory, which charts the movement of rocks
through the continental crust. These trajectories provide
a powerful tool for testing numerical models that investi-
gate the combination of parameters governing the develop-
ment of orogens.
MELTED ROCKS UNDER THE MICROSCOPE
The microstructure in rocks continually readjusts to
changes in conditions. Minerals disappear, new ones grow,
and grain boundaries move, driven by the need to reduce
energy (e.g. Holness 2008), whether that is lattice, inter-
facial or surface energy. The extent to which microstructure
reaches the equilibrium state, often thought of as uniform
grain size and polygonal grain shapes, contains informa-
tion on driving forces and the kinetics of grain-boundary
migration. These factors could be related to such diverse
and interesting parameters as the cooling and deformation
histories of the rocks. The type of microstructural informa-
tion sought must be matched to the rock sampled. It is
fruitless, for example, to attempt to understand the melting
reactions or mineralmelt equilibration microstructures
by examining the paleosome, since it did not melt.
Similarly, the microstructure of a leucosome contains infor-
mation about the crystallization of anatectic melt rather
than the melt-producing reactions. The correct identifica-
tion of each petrological part of a migmatite is necessary
because each contains information about processes specific
to its origin.
A results in the formation of many different types of dilatant
B C
Examples of partially melted rocks. (A) Migmatite
FIGURE 4 derived from pelite and psammite protoliths,
Nemiscau subprovince, Quebec. The lightest-coloured parts are
leucosome and the darkest parts, rich in biotite and conspicuous
red garnet, are residual material; together these are the neosome.
The medium-grey-coloured part is a psammite that did not partially
melt; it is paleosome. Scale is 15 cm long. (B) Highly strained
E LEMENTS
Since leucosome cannot be considered as representative of
the initial melt composition, because of crystal fraction-
ation and contamination for example, the chemical compo-
sition of quenched glass from melting experiments has
been the principal source of information on the composi-
tion of anatectic melts. This situation is changing: micron-
sized inclusions of glass and nanogranite (FIG. 5), believed
to be respectively quenched anatectic melt and its crystal-
lization products, have been found in minerals from
migmatite terranes (Cesare et al. 2011). These inclusions
could provide the major, trace and isotopic compositions
of natural anatectic melts; such starting-point composi-
tions are required to understand what changes occur to
anatectic melts in the crust. How can anatectic melt remain
as glass in slowly cooled rocks from deep in the continental
crust? This and other questions are addressed in the contri-
bution by Holness et al. (2011 this issue), which outlines
what recent studies of the microstructure in partially
melted rocks tell us about the processes that occur when
the continental crust melts and subsequently cools.
TECTONIC AND GEODYNAMIC
IMPLICATIONS OF PARTIAL MELTING
The onset of partial melting has a profound effect on the
continental crust. The types of structures that form change
and strain rates increase when the temperature of the conti-
nental crust passes the solidus temperature. Because
anatectic melt is less dense and less viscous than either the
protolith or the solid residue, it is more mobile than the
solid fraction and will separate from it. Buoyancy is a
driving force, but differential stress acting on an inevitably
anisotropic crust induces pressure gradients, and these
constitute another, locally stronger, driving force for the
movement of melt. Differential stress in anisotropic rocks
structures, the space between boudins being one well-known
example. Melt migrates to and collects in these structures.
The transfer of heat in the continental crust is largely by
the slow process of conduction, so the deep parts of the
crust are slow to heat up and slow to cool. Consequently,
metamorphic temperatures can remain above the solidus
(650 oC) for times as long as 30 million years, e.g. in the
HimalayanTibetan system. In that period melt can move
from one set of dilatant structures to the next as the crust
progressively deforms, crystallizing partially in each and
creating a complex network of leucosomes.
migmatite derived from metatonalite partially melted under granu-
lite facies conditions in the Limpopo Mobile Belt, a deeply eroded
orogen. Penknife is 11 cm long. (C) Migmatite in which the garnet-
bearing neosomes have been highly strained, creating a banded or
layered structure typical of shear zones developed in melt-bearing
rocks. Scale is 15 cm long.
232 A UGUS T 2011
 produced when and where rocks become hot and melt.
Strain and advected heat may be focussed into a narrow
zone between a reverse-sense shear zone at the bottom and
a normal-sense one at the top, in a phenomenon called
channel flow. Over the past two decades, advances in under-
standing these topics and other tectonic and geodynamic
consequences of When the Continental Crust Melts have
occurred through the use of highly sophisticated numerical
models, and the article by Jamieson et al. (2011 this issue)
presents the state of the art in this critical field.

MOVING THE MELT TO DIFFERENTIATE

Backscattered electron image of a nanogranite
FIGURE 5 derived from a small (6 m) inclusion of granitic melt
trapped in a garnet (Grt) crystal from a migmatite at Ronda
(Spain). The melt inclusion has a typical polyhedric shape (nega-
tive crystal; see Cesare et al. 2011) and crystallized into a fine-
grained aggregate of quartz (Qtz), biotite (Bt), K-feldspar (Kfs),
apatite (Ap) and plagioclase (not visible in this image).
IMAGE COURTESY OF O MAR BARTOLI, U NIVERSITY OF PARMA , ITALY
Approximately 80% of grain boundaries have melt on them
when the melt reaches ~7 vol%, and this results in a loss
of about 80% of the pre-melting strength of the protolith
(Rosenberg and Handy 2005). Rocks become very weak
long before melting advances enough (~26 vol%) to turn
them into magma, i.e. a suspension of crystals in melt. The
onset of melting and the weakening it causes have a
profound effect on the rheology of the continental crust,
on the way it deforms and on how orogens develop. The
location of weak rocks is controlled by where the heat
source is and by the rate at which hot rocks and cold rocks
are moved to advect heat and mass. These factors are
controlled in part by isostasy, by the development of a
ductile root at the bottom of the continental crust and by
erosion at the top of it. A weak region in the crust is
THE CONTINENTAL CRUST
Granites are accumulations of anatectic melt, albeit melt
that has had its composition changed through contamina-
tion by residuum (peritectic phases), wall rocks, or mixing
with different magmas and through fractional crystalliza-
tion. Melting takes place deep (>25 km) in the continental
crust. However, most plutons of granite are emplaced in
its upper part, mostly at depths of 12 to 15 km where the
transition from ductile to brittle rheology occurs (FIG. 2).
To accomplish the differentiation of the continental crust,
anatectic melt must migrate from the grain boundaries
where it was formed and become progressively concen-
trated into a more focussed flow pattern. Thus, the melt is
able to traverse rocks that are at subsolidus temperatures
in the middle crust without freezing as dykes. In other
words the flow of granite melt must become organised.
How this happens When the Continental Crust Melts is
discussed by Brown et al. (2011 this issue) in the fi nal
article.
ACKNOWLEDGMENTS
Constructive reviews and comments by principal editor
Hap McSween and reviewers Tracy Rushmer, Nick Petford
and Gary Stevens have greatly improved this contribution.
On behalf of all the contributors we would like to express
our collective thanks to Pierrette Tremblay for her encour-
agement and help at all stages in the development of this
issue.

REFERENCES
Belousova EA, Kostitsyn YA, Griffi n WL,
Begg GC, OReilly SY, Pearson NJ (2010)
The growth of the continental crust:
Constraints from zircon Hf-isotope
data. Lithos 119: 457-466
Berger A, Burri, T, Alt-Epping P, Engi M
(2008) Tectonically controlled fluid flow
and water-assisted melting in the
middle crust: An example from the
Central Alps. Lithos 102: 598-615
Brown M, Korhonen FJ, Siddoway CS
(2011) Organizing melt flow through
the crust. Elements 7: 261-266
Cesare B, Acosta-Vigil A, Ferrero S, Bartoli
O (2011) Melt inclusions in migmatites
and granulites. Journal of the Virtual
Explorer 40: paper 2
Clark C, Fitzsimons ICW, Healy D, Harley
SL (2011) How does the continental
crust get really hot? Elements 7:
235-240
Guernina S, Sawyer EW (2003) Large-
scale melt-depletion in granulite
terranes: an example from the Archean
Ashuanipi Subprovince of Quebec.
Journal of Metamorphic Geology 21:
181-201
Harrison TM (2009) The Hadean crust:
Evidence from >4 Ga zircons. Annual
Review of Earth and Planetary Sciences
37: 479-505
Holness MB (2008) Decoding migmatite
microstructures. In: Sawyer EW, Brown
M (eds) Working with Migmatites.
Mineralogical Association of Canada,
Short Course Volume 38, pp 57-76
Holness MB, Cesare B, Sawyer EW (2011)
Melted rocks under the microscope:
Microstructures and their interpreta-
tion. Elements 7: 247-252
Jamieson RA, Unsworth MJ, Harris NBW,
Rosenberg CL, Schulmann K (2011)
Crustal melting and the flow of moun-
tains. Elements 7: 253-260
Jessup MJ, Cottle JM, Searle MP, Law RD,
Newell DL, Tracy RJ, Waters DJ (2008)
P-T-t-D paths of Everest Series schist,
Nepal. Journal of Metamorphic Geology
26: 717-739
Rosenberg CL, Handy MR (2005)
Experimental deformation of partially
melted granite revisited: implications
for the continental crust. Journal of
Metamorphic Geology 23: 19-28
Rudnick RL, Gao S (2003) The composi-
tion of the continental crust. In:
Rudnick RL (ed) The Crust. Treatise on
Geochemistry 3, Elsevier-Pergamon,
Oxford, pp 1-64
Sawyer EW (2008a) Atlas of Migmatites.
The Canadian Mineralogist Special
Publication 9, NRC Research Press,
Ottawa, Ontario, Canada, 371 pp
Sawyer EW (2008b) Identifying the parts
of migmatites in the field. In: Sawyer
EW, Brown M (eds) Working with
Migmatites. Mineralogical Association
of Canada, Short Course Volume 38, pp
29-36
Sawyer EW (2010) Migmatites formed by
water-fluxed partial melting of a leuco-
granodiorite protolith: Microstructures
in the residual rocks and source of the
fluid. Lithos 116: 273-286.
Tatsumi Y (2005) The subduction factory:
How it operates in the evolving Earth.
GSA Today 15: 4-10
Ward R, Stevens G, Kisters A (2008) Fluid
and deformation induced partial
melting and melt volumes in low-
temperature granulite-facies metasedi-
ments, Damara Belt, Namibia. Lithos
105: 253-271
White RW, Stevens G, Johnson TE (2011)
Is the crucible reproducible?
Reconciling melting experiments with
thermodynamic calculations. Elements
7: 241-246
Whittington AG, Hofmeister AM, Nabelek
PI (2009) Temperature-dependent
thermal diffusivity of the Earths crust
and implications for magmatism.
Nature 458: 319-321

E LEMENTS 233 A UGUS T 2011

GLOSSARY
Anatectic front The surface marking the beginning
of partial melting in the continental crust. It corre-
sponds to the fi rst occurrence of neosome in the
direction of increasing metamorphic grade.
Anatectic melt A melt, generally granitic in composi-
tion, produced by anatexis
Anatexis Partial melting of the continental crust, irre-
spective of the degree of partial melting
Brittleelastic fracturing Open-mode fracturing by
crack propagation normal to the direction of
minimum compression. It occurs when stresses at the
crack tips equal fracture toughness, or when reduced
stresses lead to subcritical crack growth.
Constrictional strain Deformation resulting in
prolate fabrics in which linear structures dominate
over planar structures
Diatexite A migmatite in which neosome dominates
and prepartial melting structures (bedding, folia-
tion, folds) have been destroyed and commonly
replaced by syn-anatectic flow structures
Ductile fracturing Fracturing due to creep and
growth of microscale voidsfi lled with either fluid
or melt in rockthat become interconnected leading
to rupture.
Ductile-to-brittle transition zone The depth in the
Earths crust where the brittle strength equals the
ductile strength. It occurs in the range of 12 to 18 km.
Flattening strain A deformation resulting in oblate
fabrics in which planar structures dominate over
linear structures
Haplogranite system A simplification of the composi-
tion of granite to just albite + orthoclase + quartz +
H2O components (the AbOrQz system). Adding an
anorthite component creates the haplogranodiorite
system.
Incongruent melting The process by which partial
melting of a rock, mineral or mineral assemblage
produces one or more new (peritectic) minerals, in
addition to melt
Leucosome The part of a migmatite derived from segre-
gated partial melt. Leucosome does not necessarily
have the composition of an anatectic melt because
E LEMENTS
processes such as fractional crystallization and
contamination may have modified its composition.
Melanosome A type of residuum composed predomi-
nantly of dark-colored minerals, such as biotite,
garnet, cordierite, amphibole or pyroxene
Metatexite A type of migmatite in which coherent
prepartial melting structures, such as bedding, folia-
tion and folds, are preserved
Migmatite A metamorphic rock formed by partial
melting. At the outcrop scale migmatites are hetero-
geneous. In addition to two petrogenetically related
parts called leucosome and residuum, migmatites can
also contain rocks, called paleosome, which did not
melt.
Neosome The part of a migmatite formed by partial
melting and consisting of melt-derived and residual
fractions. The neosome may, or may not, have under-
gone segregation.
Orogenesis The process of forming a mountain chain
in the Earths continental crust due to the conver-
gence and collision of tectonic plates
Paleosome The non-neosome part of a migmatite that
was not affected by partial melting because of its bulk
composition
Peritectic mineral(s) A new mineral (or minerals)
produced in addition to melt during incongruent partial
melting of a rock, mineral or mineral assemblage
Protolith or parent rock The rock from which the
neosome in a migmatite was derived
Pseudosection A map of phase assemblages for two
specified intensive and or/extensive variables (for
example, pressure and temperature) and a specified
bulk composition
Residuum The solid fraction left in a migmatite after
partial melting and the extraction of some or all of
the melt
Segregation The overall process in which anatectic
melt is separated from the residuum in a migmatite
Solidus The boundary separating the solid ( fluid)
phase assemblage fields (generally at lower tempera-
ture) from the melt-bearing phase fields (generally at
higher temperature) in a PT phase diagram
Stromatic migmatite A type of metatexite migmatite
in which the leucosome and melanosome, or just the
leucosome, occur as laterally continuous, parallel
layers called stroma, which are commonly oriented
along the compositional layering or the foliation
Supercontinent A large continental landmass created
from the collision of several continental cores or
cratons
Ultrahigh-temperature (UHT) metamorphism
Metamorphism that occurred at temperatures
above 900 oC and pressures compatible with the
stability of sillimanite
234 A UGUS T 2011
 How Does the Continental
Crust Get Really Hot?
Chris Clark1, Ian C. W. Fitzsimons1, David Healy2
and Simon L. Harley3
1811-5209/11/0007-0235$2.50 DOI: 10.2113/gselements.7.4.235
T
here is widespread evidence that ultrahigh temperatures of 9001000 C
have been generated in the Earths crust repeatedly in time and space.
These temperatures were associated with thickened crust in collisional
mountain belts and the production of large volumes of magma. Numerical
modelling indicates that a long-lived mountain plateau with high internal
concentrations of heat-producing elements and low erosion rates is the most
likely setting for such extreme conditions. Preferential thickening of already-
hot back-arc basins and mechanical heating by deformation in ductile shear
zones might also contribute to elevated temperatures.
KEYWORDS : metamorphism, ultrahigh temperature, heat production, mountain
belt, thermal modelling
INTRODUCTION
Evidence for the pressuretemperature (PT) conditions
under which Earths crust has generated large volumes of
magma is provided by metamorphic rocks that represent
the solid residue of partial melting. Many of these rocks
preserve minerals formed at moderate pressures and very
high temperatures, conditions consistent with substantial
partial melting of continental crust. Although originally
regarded as isolated anomalies, there is increasing evidence
that these ultrahigh-temperature (UHT) conditions were
attained repeatedly in time and space. Our ability to quan-
tify this record has increased dramatically in recent years
with improved thermodynamic constraints on mineral PT
stability that allow us to derive robust PT data for meta-
morphic rocks. These data can then be compared with
geothermal gradients predicted using mathematical models
that describe the thermal behaviour of continental crust
in simple tectonic settings. While standard numerical
models for mountain building can reproduce the condi-
tions recorded by most metamorphic rocks, UHT metamor-
phism is difficult to replicate. In this article we examine a
number of heat sources that might account for these
extreme temperatures.
RECOGNITION OF UHT METAMORPHISM
Metamorphic conditions are classified using metamorphic
facies, which are PT fields defi ned by distinctive mineral
assemblages (FIG. 1A). UHT conditions lie at the high-
temperature extreme of the granulite facies and were
defined by Harley (1998) as temperatures in excess of 900 C
1 The Institute for Geoscience Research (TIGeR), Department of
Applied Geology, Curtin University, GPO Box U1987
Perth WA 6845, Australia
E-mail: c.clark@curtin.edu.au
2 School of Geosciences, Kings College, University of Aberdeen
Aberdeen, AB24 3UE, UK
3 Grant Institute of Earth Science, The University of Edinburgh
Edinburgh, EH9 3JW, UK
E LEMENTS , V OL . 7, PP. 235240
and pressures of 0.7 to 1.3 GPa.
Brown (2006) proposed a revised
upper pressure limit equivalent to
a P/T gradient of 750 C GPa 1,
close to the kyanitesillimanite
reaction boundary (FIG. 1A). The
lower temperature limit of 900 C
is somewhat arbitrary, but it places
the onset of UHT metamorphism
beyond the conditions at which
many crustal rocks start to melt, a
process that represents a signifi-
cant barrier to the attainment of
higher temperatures.
Recognition of UHT metamor-
phism is problematic because few rocks develop diagnostic
minerals at these conditions and widespread chemical
equilibration during cooling makes temperature estimates
based on mineral composition unreliable. Although rare
in metamorphic belts, Mg-rich mudstone does develop
diagnostic mineral assemblages at UHT conditions, most
notably sapphirine + quartz (FIG. 1B), but also orthopy-
roxene + sillimanite + quartz, spinel + quartz, and osumilite
+ garnet (Harley 2008). However, the stability of these
assemblages is highly sensitive to minor chemical compo-
nents and the redox state of the rock, making them unreli-
able indicators of UHT conditions. Thus sapphirine +
quartz is stable down to 850 C in highly oxidised systems
(Taylor-Jones and Powell 2010), and components such as
Fe3+, Cr, Zn and Ti can extend spinel + quartz stability to
below 900 C (Harley 2008). Other evidence for UHT condi-
tions includes high Zr contents in rutile; aluminous ortho-
pyroxene coexisting with garnet, although the presence of
Fe3+ can again lead to temperature overestimates; and
extensive solid solution in feldspar and pyroxene resulting
in mesoperthite and pigeonite exsolution, although care
must be taken to ensure these are not igneous relics (Harley
2008).
Temperature estimates based on the distribution of Fe and
Mg between different minerals are reset on cooling from
UHT conditions, but calculations that correct for this effect
reveal a continuum in estimated peak temperature from
the lower granulite facies into the UHT field (FIG. 1A ;
Pattison et al. 2003). This suggests that UHT metamor-
phism occurs in similar tectonic settings to lower-temper-
ature granulite metamorphism and is not a result of
anomalous processes, a conclusion supported by the
discovery of UHT metamorphism at more than 40 localities
worldwide (Kelsey 2008) with ages spanning the last 3000
million years (Brown 2006). Inferred geothermal gradients
beneath the Himalaya are consistent with UHT conditions
at depth (Hacker et al. 2000), implying that the apparent
scarcity of Phanerozoic UHT metamorphism reflects the
time taken for deep crustal rocks to reach the surface and
235 A UGUS T 2011
 A B
(A) PT conditions of UHT and other styles of meta-
FIGURE 1 morphism, from Brown (2007). Red circles are PT
estimates for granulite facies rocks (Harley 1998; Pattison et al.
2003); their distribution shows that UHT metamorphism is contin-
uous with the granulite facies. Field abbreviations: A, amphibolite
facies; BS, blueschist facies; E-HPG, medium-T eclogite high-P
granulite facies; G, granulite facies; GS, greenschist facies; UHP,
ultrahigh-pressure metamorphism. Mineral abbreviations: Ab,
not an absence of UHT conditions. The ages of UHT meta-
morphism show some correlation with periods of super-
continent assembly (Brown 2006), suggesting that UHT
metamorphism occurs during continental collision or that
UHT terranes associated with collision are more likely to
be preserved than those formed in other settings.
WHAT DRIVES UHT METAMORPHISM?
While there is widespread agreement that UHT rocks occur
in many metamorphic belts, there is no consensus on the
heat source for such extreme temperatures. Pervasive defor-
mation and widespread chemical and textural re-equili-
bration at high temperature have destroyed much of the
field and petrological evidence for how UHT conditions
are achieved (FIG. 1C). Some constraints are provided by
the exposure of ancient UHT terranes at the surface of crust
that is now of normal thickness and by mineral reactions
in UHT rocks indicating that peak conditions are typically
followed by decompression. These relationships suggest
that UHT rocks form in the mid levels of thickened crust,
consistent with metamorphism during continental colli-
sion. However, some terranes, including the Napier
Complex of Antarctica, record prolonged cooling from UHT
conditions at near constant pressure, implying that these
areas were in isostatic equilibrium during and after meta-
morphism. Another important observation is that UHT
metamorphism is typically not associated with the intru-
sion of substantial mafic or ultramafic rock, ruling out
mantle-derived magma as a major heat source.
Given the limited geological evidence, the best quantitative
constraints on the cause of UHT metamorphism come from
numerical predictions of temperature variations in simple
tectonic settings. Two-dimensional numerical models are
increasingly used to reproduce the evolution of mountain
belts (Jamieson and Beaumont 2010), but these require an
understanding of regional-scale structure and rock distribu-
tion that is lacking for deeply eroded UHT terranes. For
this reason we investigate the factors that promote, or limit,
UHT metamorphism in simple models of crustal thickening
using the one-dimensional heat flow equation:
, (1)
E LEMENTS
C
albite; Coe, coesite; Jd, jadeite; Ky, kyanite; Qtz, quartz; Sil, silli-
manite. (B AND C) Mineralogical indicators and field relationships
of UHT metamorphism in the Napier Complex, Antarctica.
(B) Sapphirine (Spr) + orthopyroxene (Opx) + quartz (Qtz) assem-
blage. Opx contains up to 10 wt% Al2O3. (C) Interlayered sequence
of UHT metamorphic rocks including quartzofeldspathic gneiss (a),
garnet-sillimanite metapelite with Spr-bearing layers (b),
metatonalite (c) and granodioritic gneiss (d)
where T is temperature, t is time, z is depth, is thermal
diffusivity, is density, cp is specific heat capacity, u is
vertical transport velocity relative to Earths surface (rate
of burial, or exhumation if negative), and Arad, Amech and
Achem are rates of heat production per unit volume by radio-
active decay, mechanical deformation, and chemical reac-
tion, respectively. The fi rst term on the right-hand side of
equation 1 describes conductive heat flow between rocks
of different temperature, the second quantifies vertical heat
transport by advection (heat carried with rocks moving
relative to Earths surface), and the third describes three
mechanisms that create heat (and also consume heat in
the case of Achem). One-dimensional models cannot account
for lateral movement of heat or rock, which will be signifi-
cant close to plate boundaries and other large-scale dipping
structures in mountain belts, but they do provide a fi rst-
order assessment of potential heat sources for UHT meta-
morphism, particularly for rocks located some distance
from plate boundaries. The use of one-dimensional models
also maximizes the likelihood of replicating UHT condi-
tions, given that lateral heat flow will move heat away from
high-temperature rocks.
Important parameters in our models include the thick-
nesses of the crust and lithosphere before and after thick-
ening, the temperaturedepth profi le before thickening,
the geometry of thickening (e.g. homogenous deformation
or thrust stacking), the erosion rate, the values of , and
cp, the magnitude of heat flow from the mantle into the
lithosphere, the magnitudes and spatial and/or temporal
distributions of Arad, Amech and Achem, and the magnitude
of heat advection by magma into or within the crust. These
are constrained to varying degrees by geologic and experi-
mental data, and there has been considerable uniformity
in values used over the last 30 years (England and
Thompson 1984); however, recent studies have questioned
some of these assumptions. In particular, new experiments
show that has a much stronger temperature dependence
than thought previously, with values at lower crustal
temperatures being about 50% of those used in most
models (Whittington et al. 2009). This reduces the rates
of conductive heat flow, allowing regions of high radioac-
tive, mechanical or chemical heat production to attain
higher temperatures, and we adopt temperature-dependent
values of in our models. Unlike many studies that assume
Achem to be negligible, we incorporate a term for the heat
236 A UGUS T 2011
consumed by melting reactions, which could be significant Radioactive Heat Production
under UHT conditions given the potential for extensive
Radioactive decay of U, Th and K has long been recognised
melt generation.
as an important heat source in continental crust, with
We investigate three heat sources that have been proposed typical heat-production values of 0.13.0 W m-3 (Vil et
to account for UHT conditions during continental al. 2010). The influence of radioactive heating during
collision: mountain building depends on the initial distribution of
heat-producing elements (generally assumed to be greater
Elevated radioactive heat production in thickened crust
in the upper crust due to its more felsic composition) and
Increased mantle heat input to back-arc basins how this distribution is modified during collision, including
Mechanical heating in ductile shear zones the addition of radioactive material by thickening and its
loss by erosion (Jamieson et al. 1998; Sandiford and
Another possible heat source is the addition of mantle- McLaren 2002). We investigate these parameters using a
derived magma to the crust, but we ignore this given the one-dimensional model in which crust, comprising a
lack of evidence for significant mafic magmatism in UHT 20 km thick upper radioactive layer (Arad = 2.0 W m3)
terranes. We also ignore the effects of magma movement and a 15 km thick lower non-radioactive layer, is instan-
within the crust because this does not add extra heat to taneously doubled in thickness by homogenous deforma-
the system and cannot, on its own, drive UHT metamor- tion (FIG. 2). There is no thickening of mantle lithosphere,
phism. Partial melting could, however, play an important consistent with its partial detachment or subduction
role in limiting crustal heat production and enabling the during collision, and, following modelling studies of the
crust to attain higher temperatures in later metamorphic European Alps (England and Thompson 1984), erosion
events, and we discuss this at the end of the article.

pre-thickening post-thickening post-erosion

Top of lithosphere: T = 0 C Top of lithosphere: T = 0 C Top of lithosphere: T = 0 C

AA Erosion rate = 0.7 mm y-1 upper crust

5 km
upper crust
20 km

Arad = 2 W m-3
Arad = 2 W m-3
upper crust
Arad = 2 W m-3
40 km

30 km
lower crust
15 km

lower crust Arad = 0 W m-3

Arad = 0 W m-3
lithosphere = 150 km

MOHO at 35 km

lithosphere = 150 km
MOHO at 35 km
30 km

lower crust
Arad = 0 W m-3

mantle
lithosphere = 185 km

mantle MOHO at 70 km
Arad = 0 W m-3
Arad = 0 W m-3

mantle
Arad = 0 W m-3
Base of lithosphere: Base of lithosphere:
T = 1300 C T = 1300 C

Base of lithosphere:
T = 1300 C

Temperature (C) Temperature (C)

0 300 600 900 1200 0 300 600 900 1200
0 My (before thickening) PTt for initial depth 30 km
0 My (after thickening) PTt for initial depth 50 km
10 20 My 80 My
0.2
PTt for initial depth 70 km
40 My 100 My 0.4
Pressure (GPa)

60 My 120 My
Depth (km)

20
0.6
30 0.8
40 UHT 1.0 UHT

1.2
50
1.4
60 1.6
70
B
B 1.8 C
C

1-D thermal model for instantaneous doubling of
FIGURE 2 crustal thickness by homogenous deformation, with
erosion at 0.7 mm y-1 starting 20 My after thickening. (A) Model
geometry and A rad immediately before thickening, immediately
after thickening and 120 My after thickening when erosion has
returned crust to its original thickness. (B) Evolution of the
geothermal gradient with time, showing gradients immediately
before and after thickening and then at 20 My intervals. (C) PTt
particle paths for rocks buried to 30, 50 and 70 km depths on
thickening. All models solve equation 1 by finite difference with
fixed T at the surface (0 C) and the base of the lithosphere
(1300 C). The latent heat of melting is 320 kJ kg -1 (see FIG. 4C), and
the T-dependent expressions for and cp are from McKenzie et al.
(2005) for the mantle and from Mottaghy et al. (2008) for the
crust. Boxes in (B) and (C) mark UHT conditions.

E LEMENTS 237 A UGUS T 2011