U^Pb LA-ICP-MS Geochronology and Pseudosecti

Sveconorwegian Mid-crustal Ultrahigh-temperature Metamorphism in Rogaland,Norway: U^Pb LA-ICP-MS Geochronologyand Pseudosections of Sapphirine Granulitesand Associated Paragneisses

KIRSTEN DRU« PPEL1,2*, LIZ ELSA« �ER1, SO« NKE BRANDT3 ANDAXEL GERDES4

1TECHNICAL UNIVERSITY BERLIN, INSTITUTE FOR APPLIED GEOSCIENCES, ACKERSTR. 71^76, D-13355 BERLIN,

GERMANY2KIT KARLSRUHE, INSTITUTE FOR APPLIED GEOSCIENCES, ADENAUERRING 20B, D-76131 KARLSRUHE, GERMANY3UNIVERSITY OF KIEL, INSTITUTE FOR GEOSCIENCES, LUDEWIG-MEYN-STR. 10, D-24118 KIEL, GERMANY4GOETHE-UNIVERSITY FRANKFURT, INSTITUTE FOR GEOSCIENCES, ALTENHO« FERALLEE 1, D-60438 FRANKFURT AM

MAIN, GERMANY

RECEIVEDJANUARY 13, 2011; ACCEPTED SEPTEMBER 12, 2012ADVANCE ACCESS PUBLICATION NOVEMBER 5, 2012

MgAl-rich sapphirine granulites (bulk XMg 0·71^0·75) occur as

boudinaged layers in migmatitic garnet^orthopyroxene^cordierite^

spinel gneisses and migmatitic garnet^sillimanite metapelites in the

vicinity of the c. 930^920 Ma Rogaland anorthosite^mangerite^

charnockite complex, SW Norway. Investigation of the mineral

reaction history of the sapphirine granulites and the surrounding

paragneisses, combined with geothermobarometric calculations and

constraints from pseudosections calculated in the Na2O^CaO^

K2O^FeO^MgO^Al2O3^SiO2^H2O^TiO2 (NCKFMASHT)

system, indicates a clockwise P^T path that reached peak-

metamorphic ultrahigh-temperature (UHT) conditions of c.10008C at c. 7·5 kbar by prograde heating. UHT peak metamorph-

ism is followed by near-isothermal (ultra)high-temperature decom-

pression to P55·5 kbar at 900^10008C and subsequent

near-isobaric cooling to5750^8008C at c. 5 kbar. In situ U^Pb

laser ablation inductively coupled plasma mass spectrometry dating

of metamorphic zircon within the sapphirine granulites yields con-

cordant ages of 1010� 7Ma and 1006� 4Ma for zircon

presumably formed during prograde breakdown of garnet at

T4850^9408C as estimated fromTi-in-zircon thermometry, sug-

gesting that UHT metamorphism and the deduced clockwise P^Tevolution is linked to regional Sveconorwegian metamorphism at c.1010 Ma. Most of the metamorphic zircon surrounds largely resorbed

inherited oscillatory zoned zircon cores (207Pb/206Pb apparent ages

1220^1841 Ma), testifying to the sedimentary origin of the sapphir-

ine granulites. Epitactic growth of xenotime on metamorphic zircon

at 933� 5Ma is suggested to be related to crystallization of anatec-

tic melt during post-decompressional cooling. The clockwise P^Tpath culminating at mid-crustal UHTconditions at c. 1010Ma fol-

lowed by (U)HTdecompression is interpreted to result from colli-

sional tectonics during the early stages of the Sveconorwegian

Orogeny, followed by gravitational collapse of the mountain plateau.

KEY WORDS: anorthosite; isograds; pseudosection; sapphirine; U^Pb

dating; UHT metamorphism

*Corresponding author. E-mail: [email protected]

� The Author 2012. Published by Oxford University Press. Allrights reserved. For Permissions, please e-mail: [email protected]

JOURNALOFPETROLOGY VOLUME 54 NUMBER 2 PAGES 305^350 2013 doi:10.1093/petrology/egs070

Downloaded from https://academic.oup.com/petrology/article-abstract/54/2/305/1485547by gueston 13 February 2018

I NTRODUCTIONRocks of highly magnesian and aluminous bulk compos-ition are highly sensitive to P^Tchanges and often preservewell-developed reaction textures that provide informationabout P^T conditions. Such MgAl-granulites are increas-ingly being used to elucidate the P^T paths of granulite-facies terranes (e.g. Bertrand et al., 1992; Harley, 1998;Kriegsman & Schumacher, 1999; Baba, 2003; Martignole& Martelat, 2003; Goncalves et al., 2004; Sajeev et al.,2004; Brandt et al., 2007, 2011).We report new petrological and geochronological data

for sapphirine-bearing MgAl-granulites from Rogaland(SW Norway). The sapphirine granulites of Rogalandoccur as boudinaged lenses in migmatitic metasedimen-tary gneisses (GrtÔpx^Crd^Spl gneisses and Grt^Silmetapelites; mineral abbreviations after Kretz, 1983)enclosed by migmatitic charnockite. They are exposed inthe vicinity of the 930^920Ma anorthosite^mangerite^charnockite (AMC) plutonic suite of Rogaland, SWNorway, some 8 km from the actual contact. This region isaffected by ultrahigh-temperature (UHT) granulite-faciesmetamorphism and an interference exists between regionalSveconorwegian metamorphism (M1 between c. 1035 and970 Ma) and contact metamorphism related to intrusionof the anorthosite plutons (M2 between 930 and 920 Ma;Tobi et al., 1985). Considering alignment of part of the sap-phirine crystals, Jansen & Tobi (reported by Maijer &Padget,1987) inferred that their formation might be relatedto the regional metamorphism. Detailed petrological andgeochronological investigations are, however, thus far lack-ing for these rocks.In this study, we combine petrographic investigations of

reaction textures, mineral chemistry, bulk geochemistryand geothermobarometric data with calculated P^T pseu-dosections of sapphirine granulites and the surroundingmigmatitic GrtÔpx^Crd^Spl paragneisses and Grt^Silmetapelites to constrain both their peak-metamorphicconditions and retrograde P^Tevolution. In situ laser abla-tion inductively coupled plasma mass spectrometry(LA-ICP-MS) dating of zircon and xenotime in thin sec-tions of the sapphirine granulites links their metamorphicevolution to the main metamorphic events reportedfrom the Rogaland area; that is, regional Sveconorwegianmetamorphism (M1) versus contact metamorphism relatedto intrusion of the Rogaland Complex during thepost-collisional stage of the Sveconorwegian Orogeny(M2). Results are discussed in a regional context, especiallyconsidering the close spatial association of the UHTgranu-lites with the anorthositic^noritic Rogaland Complex.Our results place new constraints on the tectonothermalevolution of the Rogaland Sector during Late Proterozoiccrust-forming processes and on the generation of contactmetamorphic envelopes surrounding large Proterozoicanorthosite massifs.

GEOLOGICAL SETT INGThe study area forms part of the SveconorwegianRogaland^Vest Agder terrane of southwestern Norway.The high-grade metamorphic basement exposed in thispart of the Baltic Shield mainly comprises large bodies ofmigmatitic charnockites and charnockitic gneisses. Attheir margins, the large charnockitic plutons are inter-layered with mafic to ultramafic, and metasedimentarylenses (Fig. 1). The latter are mainly composed of migmati-tic metagreywackes and metapelites referred to as ‘garneti-ferous migmatites’ (Huijsman et al., 1981), as well assubordinate quartzites, calcsilicates, and marbles of theso-called ‘Faurefjell formation’ (Hermans et al., 1975).Deposition ages of c. 1·5 Ga have been suggested for theserocks (e.g. Verschure, 1985). The whole area underwenthigh-grade regional metamorphism during the Sveconor-wegian Orogeny (M1) at c. 1035^970Ma (U^Pb monazite:Bingen & van Breemen, 1998; Bingen et al., 2008a; U^Pbzircon:Wielens et al.,1981; Mo« ller et al., 2002, 2003; Tomkinset al., 2005; ReÔs molybdenite: Bingen & Stein, 2003;Bingen et al., 2006, 2008b; Table 1). This metamorphicevent was associated with strong deformation of the rocksand is suggested to record crustal thickening of the Roga-land^Vest Agder terrane (Bingen et al., 2008c). Corre-sponding amphibolite-facies peak-metamorphic conditionsof 600^7008C and 6^8 kbar, increasing from NE to SW(Tobi et al., 1985), have been calculated for garnet-bearingmigmatites (Jansen et al., 1985; Tobi et al., 1985; Tomkinset al., 2005). Zircon included in cordierite coronas formedby garnet breakdown of metasediments of the Oltedalarea (Fig. 1), near the orthopyroxene isograd, dates re-gional decompression following the M1 event at955�8Ma (Tomkins et al., 2005).The high-grade metamorphic basement was subse-

quently intruded by the Rogaland Complex (Fig. 1), a1200 km2 anorthosite^mangerite^charnockite suite at930^920Ma (Scha« rer et al., 1996), presumably duringpostorogenic exhumation of the Rogaland^Vest Agder ter-rane (Versteeve, 1975). The Rogaland Complex comprisesthree main massif-type anorthosite bodies that are them-selves intruded by the layered mafic series of theBjerkreim^Sokndal lopolith, comprising anorthosite,norite, gabbronorite, mangerite, and charnockite layers(Michot, 1960). Thermal ionization mass spectrometry(TIMS) U^Pb-dating of zircon and baddeleyite inclusionsin orthopyroxene megacrysts in the anorthosites hasyielded ages of 931^926Ma (Scha« rer et al., 1996). Theend of the igneous activity is dated at 920�3Ma byzircon from ilmenite-rich norite from Tellness (Scha« reret al., 1996). Low pressures of c. 4 kbar are postulatedfor the emplacement depth of the Rogaland Complexbased on barometric estimates for garnetôrthopyroxene^plagioclase^sillimanite equilibria in osumilite-bearinggneisses (Jansen et al., 1985;Westphal, 1998;Westphal et al.,

JOURNAL OF PETROLOGY VOLUME 54 NUMBER 2 FEBRUARY 2013

306Downloaded from https://academic.oup.com/petrology/article-abstract/54/2/305/1485547by gueston 13 February 2018

2003) in the surrounding metamorphic basement andexperiments on the stability of osumilite in metapelites(i.e. 5·5�0·2 kbar for the stability of an osumilite^garnet^orthopyroxene assemblage at 800^8508C; Hollandet al., 1996). Melting experiments on jotunite from thelate-stage Bjerkreim^Sokndal layered intrusion supportthese relatively low M2 pressures of 55 kbar (VanderAuwera & Longhi, 1994).The intrusion of the Rogaland Complex is suggested

to be responsible for a multi-stage contact-metamorphicoverprint of the surrounding basement, comprising anearly, high-temperature to UHT granulite-facies stage(M2; 49008C, c. 5 kbar), followed by a retrogradeamphibolite-facies overprint (M3; 550^7008C, 3^5 kbar)attributed to isobaric cooling of the intrusive bodies(Hermans et al., 1975; Maijer et al., 1981; Maijer, 1987).According to Jansen et al. (1985) and Maijer (1987), thecontact-thermal M2 metamorphism was a very intenseevent, erasing most of the regional metamorphic M1 para-geneses during static recrystallization, with undeformed,high-grade contact-metamorphic M2 assemblages beingpreserved in a 10^30 km wide aureole around theRogaland Complex. More recently, Bingen et al. (2006)were able to show that the post-collisional anorthositeplutonism is associated with significant ductile deform-ation in the surrounding basement at least 10 km awayfrom the plutons.

Following Jansen et al. (1985) and Maijer (1987), contact-metamorphism is reflected by several mineral isograds;that is (with increasing distance from the contact): (1) the(inverted) pigeonite-in isograd in metabasites c. 5 km fromthe contact; (2) the osumilite-in isograd in metapelites c.10^13 km from the intrusive contact; (3) the orthopy-roxene-in isograd in leucocratic ortho- and paragneisses, c.20 km from the contact; (4) the clinopyroxene-in isograd ingranitic gneisses further east (Fig. 1). Whereas the first twoisograds wrap around the western lobe of the RogalandComplex, the last two run subparallel to the southeasternigneous contact, but diverge from it in the north.According to Jansen et al. (1985) and Westphal et al. (2003)contact-metamorphic temperatures during the M2 event de-crease from 9008C close to the contact, judging from the oc-currence of pigeonite (now inverted) and estimates for theosumilite-bearing assemblage, to c. 7508C at 15 km distancefrom the igneous complex at the orthopyroxene isograd. Incontrast to the pigeonite-in and osumilite-in isograds, forwhich a genetic relation with the emplacement of theRogaland Complex is generally accepted, the origin ofthe Opx-in and Cpx-isograds is still debated. They areeither discussed in terms of the anorthosite emplacement(M2; Tobi et al., 1985; Maijer, 1987; Westphal et al., 2003) orsuggested to represent transitions in the regional meta-morphic grade during the Sveconorwegian Orogeny(M1; e.g. Bingen & van Breemen,1998; Bingen et al., 2008a).

Fig. 1. Geology of the Rogaland^Vest Agder metamorphic terrane, modified after Falkum (1982). Isograds after Tobi et al. (1985). Location ofstudy area (Fig. 2) is indicated by the rectangle.

DRU« PPEL et al. UHT METAMORPHISM, ROGALAND, NORWAY


Table 1: Compilation of published geochronological data on igneous and metamorphic events in the Rogaland area

Igneous/metamorphic stage Reference Lithology Mineral and method Age (Ma)

Igneous (pre- to syn-M1) Pasteels & Michot (1975) Granitic gneiss U–Pb zircon c. 1486

Zhou et al. (1995) Charnockitic gneiss U–Pb zircon 1159� 5

Bingen & van Bremen (1998) Augen gneiss U–Pb zircon 1049 þ2/–8, 1051 þ2/–8, 1051 þ2/–4

Andersen et al. (2002) Granite U–Pb zircon 1036 þ23/–22

Moller et al. (2002) Pigeonite charnockite U–Pb zircon 1588� 10 (inherited), 1033� 20, 1056� 10

Pigeonite charnockite U–Pb zircon 1520� 7 (inherited), 1035� 6

Augen gneiss U–Pb zircon 1034� 7

Moller et al. (2003) Pegmatitic leucosome U–Pb zircon 1039� 11

Migmatitic garnet gneiss U–Pb zircon 1046� 12

Tomkins et al. (2005) Metapelitic migmatite U–Pb zircon 1233� 67 to 3053� 67

Metamorphic (syn-M1) Pasteels & Michot (1975) Garnet gneiss U–Pb zircon c. 1018–951

Augen gneiss U–Pb zircon 1001–1041

Bingen & van Bremen (1998) Augen gneiss U–Pb monazite 1006� 3, 975� 2, 1010� 2, 1000� 1,

1012� 1, 1008� 1, 990� 1, 974� 2

Charnockitic gneiss U–Pb monazite 1019� 1, 1005� 1, 1004� 1, 1001� 1

Augen gneiss U–Pb monazite 1024� 1, 1009� 1, 987� 2, 986� 2,

970� 5, 971� 2, 951� 6

Charnockitic gneiss U–Pb monazite 997� 1, 986� 2, 985� 1, 972� 1,

1004� 1, 950� 1, 943� 1

Augen gneiss U–Pb monazite 1007� 2, 979� 1

Moller et al. (2002) Augen gneiss U–Pb zircon 1020� 7 to 980� 7

Pigeonite charnockite U–Pb zircon 1013� 8, 1014� 11, 1017� 12 to

992� 14, 972� 20

Tomkins et al. (2005) Metapelitic migmatite (peak) U–Pb zircon 1035� 9, 989� 11

Metapelitic migmatite

(decompression)

U–Pb zircon 955� 8

Bingen et al. (2008b) Granitic gneiss U–Pb monazite 1002� 7

Augen gneiss Th–Pb monazite 999� 5, 997� 5

Charnockitic gneiss Th–Pb monazite 1013� 5, 980� 5

Felsic granulite U–Pb monazite 1032� 5, 990� 8

Metamorphic

(post-M1 decompression)

Bingen & Stein (2003) Qtz–Pl–Kfs-Leucosome

in Grt–Opx-gneiss

Re–Os molybdenite 974� 3 to 969� 3

Tomkins et al. (2005) Metapelitic migmatite

(decompression)

U–Pb zircon 955� 8

Bingen et al. (2006) Granitic gneiss Re–Os molybdenite 982� 4 to 974� 3; 959� 3, 956� 3

to 947� 3, 946� 3 to 939� 3

Augen gneiss Re–Os molybdenite 953� 3 to 931� 3

Igneous (syn-M2) Pasteels et al. (1979) Pegmatite U–Pb zircon 914� 6

Charnockite U–Pb zircon 931� 10

Scharer et al. (1996) Anorthosite U–Pb baddeleyite 915� 4, 929� 2, 932� 3, 932� 3

Norite U–Pb zircon 920� 3

Quartz mangerite U–Pb zircon 931� 5

Metamorphic (M2 to M3) Pasteels & Michot (1975) Garnet gneiss U–Pb zircon 910� 30

Bingen & van Bremen (1998) Augen gneiss U–Pb monazite 928� 3, 927� 1, 925� 2, 912� 3,

907� 5, 930� 1, 928� 1, 904� 5, 904� 8

Charnockitic gneiss U–Pb monazite 932� 1

Moller et al. (2002) Pigeonite charnockite U–Pb zircon 931� 22, 920� 5, 911� 6

Bingen et al. (2008b) Augen gneiss Th–Pb monazite 927� 5, 924� 5, 922� 5, 914� 4, 910� 9

Bingen et al. (2006) Qtz vein Re–Os molybdenite 918� 3 to 917� 3



F I ELD RELATIONS OF THEV IKESDALEN AREAThe occurrence of sapphirine-bearing rocks in the vicinityof the Rogaland Complex was first decribed by Hermanset al. (1976). The exposure is situated c. 7 km north of theintrusive contact with the anorthosite massif adjacent tothe osumilite-in isograd and north of Vikes� (Fig. 1). Thearea is dominated by banded to massive, partially garnet-bearing, migmatitic charnockite, interlayered withmigmatitic metasedimentary units and metabasites. Meta-sediments mainly comprise GrtÔpx^Crd^Spl gneissesintercalated with subordinate metapelites (Fig. 2). Garnet-rich granite forms single, concordant layers, which weinterpret as leucosomes of the metasedimentary units. Thesapphirine granulites are exposed as concordant, discon-tinuous and folded layers within the GrtÔpx^Crd^Spl-gneisses (Fig. 3a^c). They form either sharplybounded, decimetre- to metre-thick boudins (Fig. 3a) ordiffuse schlieren (Fig. 3b), which grade into the surround-ing gneisses. Prismatic sapphirine crystals of up to 2 cm inlength display a weak alignment, oriented subparallel to

the regional foliation (Fig. 3c). The surrounding stronglymigmatitic GrtÔpx^Crd^Spl gneisses are banded on acentimetre to metre scale and display a foliation that wassubsequently folded (Fig. 3d). Leucosomes form irregularpatches or concordant layers, mainly composed of quartz,alkali feldspar, plagioclase and garnet. Restitic domainsare rich in orthopyroxene, spinel, plagioclase and garnet,the last of which forms large crystals of up to 5 cm in diam-eter largely replaced by symplectitic orthopyroxene,spinel, cordierite, and plagioclase (Fig. 3d). Rare migmati-tic Grt^Sil metapelites form narrow, concordant layers inthe GrtÔpx^Crd^Spl gneisses and contain broad con-cordant garnet-rich leucosomes of granitic composition.

ANALYT ICAL METHODSMajor and trace element contents of bulk-rock sampleswere analysed with a PHILIPS PW 1404 wavelength-dispersive X-ray fluorescence (XRF) spectrometer at theInstitute of Applied Geosciences, Technical UniversityBerlin, Germany, using fused glass discs.

Fig. 2. Geological map of the Ivesdalen area (see Fig.1 for location), showing the sample locations of the sapphirine granulite, migmatitic meta-pelite and migmatitic GrtÔpx^Crd^Spl gneiss.



Electron microprobe (EMP) analyses of minerals wereperformed on a Cameca Camebax instrument at theZentraleinrichtung fu« r Elektronenmikroskopie (ZELMI),Technical University Berlin, Germany. Natural and syn-thetic standards were used for instrument calibration.Mineral analyses were performed with an accelerating volt-age of 15 kV, a beam current of 20 nA and an electronbeam of 2·5 mm in diameter. Exsolved feldspar wasadditionally analysed with a defocused beam of 20 mm indiameter. EMP profiles across minerals were analysedusing a Cameca SX100 instrument at the GeoForschungs-Zentrum (Helmholtz Centre) Potsdam, Germany. Analyseswere performed with an accelerating voltage of 15 kV, abeam current of 20 nA and a beam diameter of 2mm.In situ trace element analyses of the major phases in both

the sapphirine granulites and the surrounding gneisseswere performed on polished �100 mm thick sectionsby LA-ICP-MS at the Institute for Geodynamicsand Geomaterials, University of Wu« rzburg, Germany.

A Merchantek 266 LUV laser coupled with an Agilent7500i ICP-MS device was used [plasma power 1250W, car-rier gas (Ar) 1·28 lmin^1, plasma gas (Ar) 14·9 lmin^1,auxiliary gas (Ar) 0·9 lmin^1]. The diameter of the laserbeam was 50 mm and the laser repetition rate was 10Hzfor all analyses. We measured background and sample at20 s each. Si in silicates and Mn in ilmenite and magnetitewere used as internal standards. The glass reference mater-ials NIST 612 [values given by Pearce et al. (1997)] andNIST 614 [values given by Horn et al. (1997)] were usedfor external instrument calibration and for control of theresults. Data processing was conducted using GLITTER3.0 software (On-line Interactive Data Reduction for theLA-ICP-MS, Macquarie Research Ltd., 2000). The preci-sion is approximately �7% for element concentrations410 ppm, �10% for concentrations 45 ppm, �15% forconcentrations 41ppm and �20% for concentrations51ppm. The minimum detection limits were in the rangeof 0·02^0·5 ppm.

Fig. 3. Outcrop photographs. (a) Sharply bounded and weakly foliated sapphirine granulite boudin surrounded by foliated GrtÔpx^Crd^Splgneiss. (b) Folded schlieren of sapphirine granulite with diffuse contacts to the bordering GrtÔpx^Crd^Spl gneiss. (c) Weakly aligned, pris-matic crystals of deep blue sapphirine in the sapphirine granulite, visible on a macroscopic scale. (d) Porphyroblastic garnet of the GrtÔpx^Crd^Spl gneiss largely replaced by an orthopyroxene^spinel^plagioclase intergrowth and set in a foliated Kfs^Pl^Qtz^Bt matrix.



For the in situ U^Pb LA-ICP-MS analyses, two selectedsapphirine granulite samples were prepared as polishedthick sections (c. 80^100 mm thick). The internal structuresof zircon and xenotime were characterized via back-scattered electron (BSE) images on a high-resolution scan-ning electron microscope at ZELMI, Technical UniversityBerlin, Germany, and under cathodoluminescence (CL)using a JEOL JSM-6400 electron microprobe at theGoethe-University Frankfurt (GUF). U^Pb analyses werecarried out by LA-ICP-MS at GUF using a Thermo-Finnigan Element II sector field ICP-MS system coupledto a New Wave UP213 ultraviolet laser with a teardroplow-volume cell following the method described byGerdes & Zeh (2006, 2009) and Frei & Gerdes (2009). Thelaser was fired with 5Hz repetition rate and an energydensity of about 3 J cm^2 over 18 s of ablation. The spotsize was adjusted to the grain size and varied between 8and 20 mm for zircon and between 8 and 18 mm for xeno-time. The estimated depth penetration was about10^20 mm. Signal was tuned for maximum sensitivitywhile keeping the oxide formation rate below 0·2%(UO/U). All data were corrected for common-Pb basedon the background and interference-corrected 204Pbsignal (daily 204Hg¼170�20 c.p.s) and a model Pb com-position (Stacey & Kramers, 1975). Within-run Pb/Ufractionation was corrected for each analysis using alinear regression through all ratios. Subsequently instru-mental mass bias and drift were corrected by normaliza-tion to reference zircon GJ-1. Previous studies at GUFhave shown that LA-ICP-MS with non-matrix matchedstandardization can yield precise and accurate results(e.g. Meier et al., 2006; Dill et al., 2007; Millonig et al.,2008). Reported uncertainties (�2s) were propagated byquadratic addition of the external reproducibility (SD,standard deviation) of the GJ-1 reference standard (1·3%for 206Pb/238U; n¼19) of the day and the within-run statis-tics of each analysis (2 SE, standard error). Accuracy andprecision of the non-matrix matched standardization waschecked by repeated analyses (n¼ 5) of Moacir monaziteand Ples› ocive zircon. Moacir analyses yielded a weightedmean 207Pb/235U age of 502·4�5·5Ma (MSWD¼ 0·7;SD¼ 2·2%) and Ples› ocive analyses a concordia age of336·4�2·8Ma (MSWD¼1·5). This is in perfect agree-ment with recent isotope dilution (ID)-TIMS analyses ofthese reference minerals (i.e. 511·2�0·4Ma and337·13�0·37, respectively; Slama et al., 2008; Gasquetet al., 2010).

RESULTSPetrographySapphirine granulites

The sapphirine granulites preserve peak-metamorphicassemblages of coarse porphyroblastic sapphirine

(15^25 vol. %) and orthopyroxene (15^20 vol. %) that areset in a fine-grained matrix of alkali feldspar (10^15vol. %), plagioclase (10^15 vol. %) and biotite (15^25vol. %; Fig. 4a) and are partly replaced by fine-grainedspinel^cordierite(^biotite) intergrowths (21^28 vol. %).The minerals mostly display a homogeneous distributionwith a weak foliation being defined by slightly aligned sap-phirine and biotite (Fig. 4a).Sapphirine forms subhedral prismatic crystals (mostly

1mm to 2 cm in length) with a strong colourless to laven-der blue pleochroism (Fig. 4b). They preserve rare inclu-sions of biotite. Porphyroblastic, subhedral orthopyroxene(mostly 1mm to 1·5 cm in diameter) with fine-grained,exsolved ilmenite platelets, displays strong greenish tobrownish pleochroism (Fig. 4b) and also contains rarebiotite inclusions. Granoblastic alkali feldspar and plagio-clase of the equigranular, fine-grained matrix (grain sizesof 0·2^0·5mm) are generally anhedral and display strongmicroperthitic and antiperthitic exsolution, respectively,indicating their formation at high temperatures. Alongthe albite exsolution lamellae, alkali feldspar is stronglysericitized, whereas plagioclase displays only minoralteration.Commonly, porphyroblastic sapphirine is replaced by a

fine-grained symplectite composed of cordierite and greenspinel (Fig. 4a, b and d) whereby its original shape is pre-served. A granoblastic cordierite^spinel reaction rim, onthe other hand, occurs between porphyroblastic sapphirineand orthopyroxene (Fig. 4b and c). Sapphirine (includingthe pseudomorphous cordierite^spinel symplectites) is sub-sequently mantled by an almost monomineralic rim ofgranoblastic spinel, followed by cordierite towards theneighbouring orthopyroxene (Fig. 4b). Spinel of both reac-tion textures displays exsolution of magnetite and platyreddish brown ilmenite. In places, spinel also containsexsolved corundum. Brownish biotite mainly replaces cor-dierite of the cordierite^spinel intergrowths (Fig. 4b^d)but also occurs as corona around porphyroblastic ortho-pyroxene. In addition, biotite is an abundant replacementproduct of matrix alkali feldspar.Accessory phases include zircon, xenotime, magnetite,

and ilmenite. Subhedral zircon (50^100 mm in diameter)occurs in the feldspar^biotite matrix, as inclusions in sap-phirine and orthopyroxene and as inclusions in cordieritewithin the cordierite^spinel reaction textures (Fig. 4d).

Migmatitic Grt^Sil metapelites

Migmatitic metapelites are made up of greyish^bluishlayers rich in garnet (up to 10 vol. %), cordierite (8^30vol. %), sillimanite (up to 5 vol. %), and spinel (up to 10vol. %); they contain minor plagioclase, quartz, ilmenite,sulphides, and alkali feldspar, as well as accessory zircon,alternating with felsic leucosomes mainly consisting ofmedium-grained, granoblastic quartz, plagioclase and



minor microperthitic orthoclase, as well as accessoryspinel.Anhedral, medium-grained garnet 1^3mm in diameter

within the melanocratic layers contains abundant inclu-sions of sillimanite, quartz and minor biotite (Fig. 5a).Garnet is associated with anhedral, porphyroblastic spineland ilmenite (both 1^2mm in diameter) and porphyro-blastic quartz (0·5^3mm in diameter; Fig. 5a). Porphyro-blastic spinel and quartz occur in direct contact or areseparated from each other by retrograde coronas (Fig. 5b,d and e; see below). Spinel preserves inclusions of silliman-ite and quartz whereas porphyroblastic quartz enclosessillimanite, biotite and spinel (Fig. 5b). Matrix sillimaniteoccurs as euhedral, fine-grained prisms of up to 0·2mmin length that are associated with garnet-rich, foliation-parallel stringers (Fig. 5a). Minor plagioclase, alkali feld-spar and quartz in the matrix form anhedral, granoblasticgrains 0·1^0·2mm in diameter.

Garnet is separated from prismatic sillimanite and/orquartz by narrow plagioclase coronas, which are followedby broad cordierite moats (Fig. 5a and c). Porphyroblasticspinel and quartz are locally separated by a fine-grainedintergrowth of quartz and hercynitic spinel (Fig. 5a).More often porphyroblastic spinel is separated from por-phyroblastic quartz by a broad moat of cordierite (Fig. 5dand e). Locally, this cordierite forms symplectites withquartz (Fig. 5d and e), which occasionally containK-feldspar. A second generation of fine-grained (50·1mm)garnet forms a corona around porphyroblastic spinel(Fig. 5a, b, e and f). Coronitic garnet also forms rimsaround sulfides, ilmenite (Fig. 5e) and spinel of thespinel^quartz intergrowths (Fig. 5a). Locally, coroniticgarnet replaces cordierite of the cordierite^quartz sym-plectites around spinel, as evident from vermicularquartz, which is preserved as inclusions in the newlyformed garnet (Fig. 5e and f). Rarely cordierite is partly

Fig. 4. Mineral assemblages and reaction textures of the sapphirine granulites (a, thin-section scan; b^d, thin-section photomicrographs). (a)Sapphirine and orthopyroxene porphyroblasts are set in a medium-grained Kfs^Pl^Bt matrix. (b) At the contact with orthopyroxene, por-phyroblastic sapphirine is subsequently mantled by granoblastic spinel and cordierite. (c) Sapphirine itself is replaced by spinel^cordierite sym-plectites, the latter of which is preferentially replaced by biotite. (d) Zircon is most abundant in granoblastic cordierite of the cordierite^spinelreaction rims.



Fig. 5. Mineral assemblages and reaction textures of the Grt^Sil metapelites (thin-section photomicrographs). (a) The porphyroblastic assem-blage of spinel, quartz, and garnet, intergrown with sillimanite, constitutes the melanocratic layers of the migmatitic Grt^Sil metapelite.Garnet is surrounded by a broad cordierite rim. A spinel^quartz intergrowth occurs between porphyroblastic spinel and quartz. Spinel of theintergrowth is surrounded by a narrow garnet corona. (b) Porphyroblastic spinel is in direct contact with quartz, which also occurs as an inclu-sion in spinel. Quartz preserves a small biotite inclusion. Coronas of garnet and cordierite locally separate porphyroblastic spinel and quartz.(c) A narrow plagioclase rim followed by a broad cordierite corona separates porphyroblastic garnet from sillimanite. (d) Porphyroblasticquartz and spinel are separated by a broad cordierite corona, which also occurs around sillimanite and ilmenite. Locally, cordierite forms asymplectitic intergrowth with quartz. Spinel is partly replaced by corundum, retrogressed to Al-hydroxides. (e) Spinel and ilmenite are sur-rounded by a cordierite^quartz symplectite, partly overgrown by garnet, with the vermicular quartz being preserved as inclusions in thenewly formed garnet. (f) Biotite formed at the expense of cordierite and porphyroblastic spinel is rimmed by a garnet corona. In addition, bio-tite overgrows the margins of the garnet (upper left).



replaced by biotite, which, in turn, is surrounded by a rimof garnet, comprising abundant quartz inclusions (Fig. 5f).Locally, euhedral garnet, which forms around spinel, coex-ists with biotite^quartz intergrowths that have grownalong cordierite margins.

Migmatitic GrtÔpx^Crd^Spl gneisses

The GrtÔpx^Crd^Spl gneisses are inhomogeneous intexture and mainly comprise equigranular, medium-grained leucocratic zones composed of alkali feldspar(30^50 vol. %), plagioclase (30^50 vol. %), and quartz(10^20 vol. %), together with minor biotite (0^10 vol. %),orthopyroxene (0^5 vol. %), spinel (0^5 vol. %), and ac-cessory cordierite, zircon, and ilmenite (Fig. 6a). Maficschlieren of up to 10 cm in width and up to several metresin length are irregularly distributed in the outcrops, butmay also alternate with quartz^feldspar-rich zones ona thin-section scale defining a compositional layering(Fig. 6a). These melanocratic zones are mainly composedof relics of porphyroblastic orthopyroxene and garnet, thelatter of which is largely replaced by complex intergrowthsof orthopyroxene (50^70 vol. %), hercynite spinel/magnet-ite (10^20 vol. %), plagioclase (�10 vol. %) biotite (5^10vol. %), cordierite (5^15 vol. %), and minor alkali feldspar(510 vol. %; Fig. 6a^d).Garnet forms coarse porphyroblastic grains of up to

6 cm in diameter, locally associated with porphyroblasticorthopyroxene (1^3 cm in diameter) intergrown with fine-grained, greenish spinel (1^2mm in diameter; Fig. 6a, band e). The orthopyroxene porphyroblasts frequently dis-play exsolution of greenish spinel (Fig. 6c and e). Porphyro-blastic garnet and orthopyroxene are surrounded by afine- to medium-grained (grain sizes of 0·1^1·2mm), leuco-cratic matrix composed of anhedral, micro- to mesoperthi-tic alkali feldspar, subhedral, antiperthitic plagioclase,weakly aligned biotite, quartz, and minor fine-grained, an-hedral ilmenite (Fig. 6a and b).Porphyroblastic garnet (1^6 cm in diameter) is largely to

completely replaced by a fine-grained intergrowth ofmedium-grained, anhedral, antiperthitic plagioclase,brownish, anhedral orthopyroxene with vermicular spinelexsolution, and isometric greenish spinel (Fig. 6b and c)associated with magnetite. Plagioclase mainly occurs indirect contact with the garnet relicts, separating it fromthe replacing orthopyroxene (Fig. 6c). Fine-grained anhe-dral cordierite, in places together with biotite, occurs as anarrow reaction rim between orthopyroxene and exsolvedspinel in the Opx^Spl^Pl intergrowth replacing porphyro-blastic garnet (Fig. 6d and e). Alkali feldspar and biotiteare preserved only in the outermost zones of the pseudo-morphs after garnet. Biotite also forms coronas aroundporphyroblastic orthopyroxene and garnet (Fig. 6e). Asecond generation of garnet overgrows coronitic biotite assub- to euhedral grains (Fig. 6f). In addition, fine-grainedsecondary garnet of up to 0·1mm in diameter is intergrown

with orthopyroxene, replacing orthopyroxene of the Opx^Spl^Pl intergrowth as garnetôrthopyroxene symplectites,or forms monomineralic rims around spinel.

Mineral chemistrySelected samples of both the sapphirine granulites and thesurrounding migmatitic GrtÔpx^Crd^Spl gneisses andmigmatitic Grt^Sil metapelites were analysed for themajor and trace element composition of their majorphases by EMP and LA-ICP-MS (only sapphirine granu-lites), respectively.The XMg of Fe^Mg silicates is calculatedas molar Mg/(MgþFe2þ). Representative analyses arelisted inTables 2^9.

Garnet

Garnet formulae were calculated on a 24-oxygen basis. Forthe estimation of the Fe3þ contents of garnet a calculationmodus on the basis of 16 cations was chosen. End-membercalculation followed the sequence andradite, grossular, al-mandine, spessartine and pyrope. Resorbed porphyro-blastic garnet of the GrtÔpx^Crd^Spl gneisses is analmandine^pyrope solid solution with minor proportionsof spessartine (�10mol %) and grossular (�3mol %)and an average composition of Prp41Alm46Sps10Grs3 (XMg

0·46^0·48; Table 2). Core^rim zoning is not developed(Fig. 7). Rims of secondary garnet around spinel, late sym-plectitic garnet intergrown with orthopyroxene, andregrown euhedral garnet on biotite (average compositionPrp35Alm53Sps11Grs1; XMg 0·38^0·40) are always less mag-nesian than porphyroblastic garnet (Fig. 7).Porphyroblastic garnet of the migmatitic metapelite is a

Ca- and Mn-poor pyropeâlmandine solid solution(Prp38^35Alm58^61Sps1Grs2^0; XMg 0·36^0·40; Table 2). Thepyrope content slightly increases towards the rim (Fig. 7).Secondary garnet (Prp33^29Alm67^63Sps2Grs2; XMg 0·30^0·34; Table 2, Fig. 7) rimming spinel is also Ca- andMn-poor but less magnesian than porphyroblastic garnet.

Orthopyroxene

Orthopyroxenes of the sapphirine granulite and GrtÔpx^Crd^Spl gneiss show only minor differences in their com-positions (Table 3; Fig. 7). Porphyroblastic orthopyroxeneof the sapphirine granulites is characterized by higherXMg values (0·69^0·75), when compared with porphyro-blastic matrix orthopyroxene of the GrtÔpx^Crd^Splgneisses (XMg 0·63^0·67).The Al2O3 contents of porphyro-blastic orthopyroxene from the sapphirine granulite andthe GrtÔpx^Crd^Spl gneiss, on the other hand, are simi-lar (8·1^8·9wt %), with the highest values being preservedin the cores of large grains. Orthopyroxene replacing por-phyroblastic garnet in the GrtÔpx^Crd^Spl gneiss hassimilar XMg (0·63^0·68) and Al2O3 contents (7·3^9·3wt%). Symplectitic orthopyroxene intergrown with lategarnet is characterized by higher XMg values (0·70^0·77)and the lowest Al2O3 contents (5·5^6·3wt %).



Regarding its trace element composition, orthopyroxeneof the sapphirine granulite is characterized by high andrimward increasing Ti (860^2500 ppm) and Mn (1250^2150 ppm), and minor contents of Sc (40^90 ppm),V (60^105 ppm),Y (10^80 ppm), and Zr (5^15 ppm).

Spinel

Spinel of the Spl^Crd symplectites replacing sapphirine inthe sapphirine granulites is essentially a hercynite^spinelsolid solution and always less magnesian (XMg 0·48^0·56)than the coexisting cordierite and associated sapphirine

Fig. 6. Mineral assemblages and reaction textures of the GrtÔpx^Crd^Spl gneisses (a, b, thin-section scans; c^f, thin-section photomicro-graphs). (a) The GrtÔpx^Crd^Spl gneisses are characterized by a heterogeneous texture with melanocratic GrtÔpx^Spl^Pl^Crd-richzones alternating with Akfs^Pl^Qtz leucosomes. (b) Porphyroblastic garnet is partially to completely replaced by a pseudomorphic,fine-grained Opx^Spl^Pl^Crd intergrowth. (c) Relicts of garnet are preserved and subsequently mantled by plagioclase and spinel/orthopyrox-ene. (d) Texturally late cordierite, in places associated with biotite, separates orthopyroxene from spinel. (e) Late biotite also replaces garnetand orthopyroxene. Orthopyroxene may also preserve early biotite inclusions. (f) Late garnet overgrows biotite and spinel.



(Table 4). Contents of Fe2O3, as calculated from ideal stoi-chiometry, are low (3^6wt %), and ZnO and Cr2O3 con-tents are always 50·2wt %. Regarding the traceelements, the spinel contains minor Ti (17^140 ppm), V(60^230 ppm), and Mn (290^1050 ppm), whereas theamounts of the REE, Y, and Zr (50·1ppm) are mostlybelow the detection limit.Spinel intergrown with orthopyroxene in Opx^Spl^Pl

symplectites replacing garnet in the Grt^Opx^Crd^Splgneisses is also an unzoned hercynite^spinel solid solution,but significantly less magnesian (XMg 0·33^0·43) than

that of the sapphirine granulites. It contains only minorCr2O3, NiO, and ZnO of50·1wt % and shows low valuesof Fe2O3 (3^5wt %) and MnO (0·6^0·9wt %).Magnetite exsolution lamellae of spinel have an almostpure end-member composition.Green spinel intergrown with quartz, as well as the

matrix spinel of the metapelites is an almost pure, unzoned,hercynite^spinel solid solution with negligible Cr2O3

(50·4wt %) and ZnO (50·3wt %) contents and very lowFe2O3 (1·4^2·0wt %) as calculated from ideal stoichiom-etry. The XMg of symplectitic spinel (XMg 0·39^0·40)

Table 2: Representative electron microprobe analyses of garnet in the migmatitic Grt^Opx^Crd^Spl gneiss and the migma-

titic metapelite

Rock type: Grt–Opx–Crd–Spl-gneiss Metapelite

Sample: Ro-IV-05-07 Ro-IV-05-07 Ro-IV-07-03

Texture: porph. porph. porph. porph. porph. porph. porph. secondary garnet porphyroblastic secondary

around Spl

Grt Grt Grt Grt Grt Grt Grt around Spl with Opx

SiO2 39·2 39·4 39·2 39·2 39·2 39·3 39·2 39·5 39·7 38·3 38·3

TiO2 0·02 0·03 0·01 0·02 0·02 0·01 0·03 0·02 0·00 0·04 0·07

Al2O3 21·8 21·8 21·8 21·8 21·7 21·8 21·8 22·2 22·4 22·4 22·0

Cr2O3 0·03 0·00 0·00 0·00 0·01 0·03 0·03 0·00 0·00 0·00 0·00

Fe2O3 1·70 0·05 0·08 0·00 0·00 0·00 0·00 0·54 0·32 0·61 0·68

FeO 22·7 22·1 22·0 21·9 21·6 21·7 21·6 25·1 24·7 26·7 29·8

MnO 4·60 4·31 4·36 4·37 4·46 4·33 4·49 5·00 4·81 0·94 0·81

MgO 11·1 10·9 10·8 10·8 10·9 11·0 10·9 8·94 9·39 9·66 7·84

CaO 0·88 0·95 0·99 0·94 1·00 1·01 1·03 0·87 0·92 0·69 0·77

Sum 100 99·5 99·2 99·1 99·0 99·1 99·1 102 102 99·4 100

Formula (O¼ 24)

Si 5·95 6·03 6·02 6·03 6·03 6·03 6·03 5·98 5·99 5·93 5·95

Ti 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01

Al 3·90 3·94 3·95 3·95 3·94 3·94 3·95 3·97 3·99 4·07 4·03

Cr 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Fe3þ 0·20 0·01 0·01 0·00 0·00 0·00 0·00 0·06 0·04 0·07 0·08

Fe2þ 2·69 2·83 2·83 2·82 2·78 2·78 2·78 3·18 3·11 3·45 3·88

Mn 0·59 0·56 0·57 0·57 0·58 0·56 0·58 0·64 0·61 0·12 0·11

Mg 2·51 2·49 2·47 2·48 2·50 2·51 2·49 2·02 2·11 2·23 1·82

Ca 0·14 0·16 0·16 0·15 0·17 0·17 0·17 0·14 0·15 0·11 0·13

Sum 16·0 16·0 16·0 16·0 16·0 16·0 16·0 16·0 16·0 16·0 16·0

XMg 0·48 0·47 0·47 0·47 0·47 0·47 0·47 0·39 0·40 0·39 0·32

andradite 0·00 0·13 0·22 0·00 0·00 0·00 0·00 1·54 0·92 1·8 2·0

grossular 2·41 2·45 2·48 2·56 2·74 2·76 2·82 0·81 1·56 0·1 0·2

almandine 45·4 46·9 46·9 46·8 46·2 46·2 46·1 53·2 52·0 58·4 65·4

spessartine 9·91 9·25 9·42 9·45 9·63 9·34 9·71 10·7 10·3 2·1 1·8

pyrope 42·3 41·3 41·0 41·2 41·5 41·7 41·4 33·8 35·3 37·6 30·6

porph., porphyroblastic.



resembles that of the resorbed porphyroblastic garnet (XMg

0·36^0·40), whereas matrix spinel is less magnesian (XMg

0·28^0·29; Table 4).

Sapphirine

Regarding its major element composition, the por-phyroblastic sapphirine of the sapphirine granulites isunzoned and does not vary in composition between sam-ples. With high XMg values of 0·81^0·78, sapphirine ismore magnesian than the coexisting orthopyroxene. Allanalyses are mixtures between the ideal 7:9:3 and 2:2:1

end-members (Fig. 8). Significant amounts of ferric iron(0·21^0·33 a.p.f.u.) are calculated from ideal stoichiometry(Table 5).In terms of its trace element composition, sapphirine

contains minor Ti (350^1450 ppm), Mn (500^900 ppm),Be (3^15 ppm), Sc (5^13 ppm), Y (1^13 ppm), and Zr(0·1^0·2 ppm), which strongly decrease towards the grainmargins and cracks (especially Ti and Mn), whereasLi (30^110 ppm) increases in the same direction. Likeorthopyroxene, sapphirine displays enrichment of theheavy REE (HREE)4Gd (Yb 1^4 ppm) with respect to

Fig. 7. Zoning profiles of garnet in the Grt^Opx^Crd^Spl gneisses and migmatitic metapelite and of orthopyroxene in the Grt^Opx^Crd^Splgneiss, and sapphirine granulite. Profiles always extend from rim to rim, through the core.



the light REE (LREE), which are always below the detec-tion limit.

Cordierite

Late cordierite separating orthopyroxene from spinel ofthe Grt^Opx^Crd^Spl gneiss is Mg-rich (XMg 0·88^0·92).High totals of c. 100wt % indicate the absence of H2O orCO2 fluids in the structural channels.Cordierite of the Crd^Spl reaction rims around sapphir-

ine in the sapphirine granulite shows no systematic com-positional variation depending on its textural position;

that is, as granoblastic reaction product of sapphirine andorthopxyroxene or as a symplectitic phase formed duringthe breakdown of sapphirine alone (Table 6). With XMg

values of 0·86^0·88 cordierite is always more magnesianthan both the associated spinel and sapphirine. It showshigh totals of around 100wt %, K2O contents at or belowthe detection limit and low Na2O contents (50·9wt %).Regarding its major element composition, cordierite ofthe sapphirine granulites is unzoned. Minor concentriczoning is revealed by the trace elementsTi (15^3000 ppm),Mn (270^1100 ppm), V (0·5^340 ppm), Rb (3^100 ppm),

Table 3: Representative electron microprobe analyses of orthopyroxene in the migmatitic Grt^Opx^Crd^Spl gneiss and the

sapphirine granulite

Rock type: Sapphirine granulite Grt–Opx–Crd–Spl gneiss

Sample: Ro-05-16b Ro-05-15 Ro-05-33B Ro-IV-05-07

Texture: porph. porph. porph. porph. porph. porph. Matrix Opx repl. Opx repl. Opx repl. Opx repl. Opx repl. Opx repl. Opx

Opx Opx Opx Opx Opx Opx Opx Grt Grt Grt Grt Grt Grt smpl.

SiO2 48·7 49·4 48·7 49·9 49·8 50·0 47·1 47·8 48·1 47·9 48·6 48·8 48·0 50·8

TiO2 0·19 0·19 0·13 0·14 0·13 0·14 0·08 0·11 0·06 0·06 0·05 0·05 0·04 0·02

Al2O3 9·10 8·91 9·15 8·11 8·23 8·80 8·79 8·78 8·76 8·82 9·27 8·89 8·48 6·34

Cr2O3 0·04 0·01 0·01 0·01 0·01 0·01 0·02 0·00 0·00 0·02 0·00 0·00 0·01 0·00

Fe2O3 1·60 0·72 3·14 1·78 1·99 1·41 4·09 3·27 1·75 2·59 3·73 2·85 2·52 2·59

FeO 16·4 16·6 14·7 15·8 15·4 16·5 17·8 18·5 19·3 19·1 17·4 18·3 19·2 18·0

MnO 0·34 0·28 0·21 0·35 0·35 0·28 1·43 1·26 1·28 1·30 1·04 0·95 1·28 0·93

MgO 22·6 23·7 23·9 24·4 24·5 24·1 20·7 21·0 20·7 20·6 22·1 21·8 20·7 23·3

CaO 0·58 0·11 0·28 0·05 0·09 0·05 0·06 0·06 0·07 0·09 0·06 0·07 0·05 0·05

Na2O 0·13 0·02 0·10 0·01 0·02 0·02 0·02 0·00 0·00 0·02 0·04 0·02 0·01 0·03

K2O 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01 0·00

Sum 99·8 99·9 100 101 101 101 100 101 100 100 102 102 100 102

Formula (O¼ 6)

Si 1·78 1·80 1·76 1·80 1·80 1·79 1·75 1·76 1·78 1·77 1·75 1·77 1·78 1·83

Al(4) 0·22 0·20 0·24 0·20 0·20 0·21 0·25 0·24 0·22 0·23 0·25 0·23 0·22 0·17

Al(6) 0·17 0·18 0·15 0·14 0·15 0·16 0·13 0·14 0·17 0·16 0·15 0·15 0·15 0·10

Ti 0·01 0·01 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Cr 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Fe3þ 0·04 0·02 0·09 0·05 0·05 0·04 0·11 0·09 0·05 0·07 0·10 0·08 0·07 0·07

Fe2þ 0·50 0·50 0·44 0·48 0·47 0·49 0·55 0·57 0·60 0·59 0·52 0·55 0·59 0·54

Mn 0·01 0·01 0·01 0·01 0·01 0·01 0·04 0·04 0·04 0·04 0·03 0·03 0·04 0·03

Mg 1·23 1·28 1·29 1·31 1·31 1·29 1·15 1·15 1·14 1·14 1·19 1·18 1·14 1·25

Ca 0·02 0·00 0·01 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Na 0·01 0·00 0·01 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

K 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Sum 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00 4·00

XMg 0·71 0·72 0·74 0·73 0·74 0·72 0·67 0·67 0·66 0·66 0·69 0·68 0·66 0·70

Al p.f.u. 0·39 0·38 0·39 0·35 0·35 0·37 0·38 0·38 0·38 0·38 0·39 0·38 0·37 0·27

p.f.u., per formula unit; porph., porphyroblastic; smpl., symplectitic; repl., replacing.



Be (4·5^12 ppm), Sc (0·5^4 ppm), and Nb (0·1^5 ppm),which decrease towards the margins of the larger cordier-ite grains forming the reaction rims between sapphirineand spinel. The amounts of Zr (50·1 to 0·5 ppm) and Y(50·1 to 8 ppm) are very low. Symplectitic cordieritegrown exclusively at the expense of sapphirine is unzonedand has the lowest overall trace element contents.Cordierite separating garnet and sillimanite in the

metapelite has the lowest Mg/Fe ratios in all the rocktypes investigated (XMg 0·77^0·81; Table 6). According to

high totals of up to 99·1wt %, cordierite contains onlyminor structurally bound H2O or CO2.

Feldspar

Alkali feldspar is a major constituent of the leucocraticmatrix of both the sapphirine granulite and theGrtÔpx^Crd^Spl gneiss. The original composition of themicroperthitic alkali feldspar of the GrtÔpx^Crd^Splgneiss, composed of orthoclase (average Or89Ab10An1)with plagioclase exsolution lamellae (average Or2Ab75

Table 4: Representative electron microprobe analyses of spinel in the migmatitic GrtÔpx^Crd^Spl gneiss, the migmatitic

metapelite and the sapphirine granulite

Rock type: Sapphirine granulite Grt–Opx–Crd–Spl gneiss Metapelite

Sample: Ro-05-16b Ro-05-15 Ro-05-14 Ro-Iv-05-07 Ro-Iv-07-03

Texture: Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl intergr. Spl–Qz matrix

w Crd w Crd w Crd w Crd w Crd w Crd w Opx w Opx w Opx w Opx sympl

SiO2 0·03 0·03 0·01 0·04 0·14 0·01 0·01 0·05 0·03 0·03 0·00 0·00

TiO2 0·00 0·02 0·02 0·02 0·01 0·01 0·03 0·00 0·00 0·01 0·00 0·01

Al2O3 60·8 61·5 62·3 60·6 59·9 60·6 58·7 60·6 59·9 60·3 61·7 59·7

Cr2O3 0·06 0·06 0·04 0·13 0·08 0·05 0·22 0·04 0·04 0·06 0·10 0·46

MgO 12·0 12·7 13·5 12·9 13·4 14·0 7·80 10·1 10·4 10·2 9·94 7·10

MnO 0·18 0·10 0·16 0·15 0·17 0·23 0·87 0·72 0·72 0·72 0·05 0·15

FeO 22·9 22·1 20·9 21·4 20·2 19·4 28·2 25·4 24·8 25·1 26·4 30·4

Fe2O3 4·11 3·54 3·11 5·32 6·02 5·69 3·63 2·64 4·22 3·23 1·47 2·04

NiO 0·03 0·02 0·02 0·00 0·02 0·02 0·00 0·00 0·03 0·01 0·00 0·00

ZnO 0·16 0·16 0·19 0·19 0·14 0·05 0·17 0·09 0·04 0·02 0·26 0·31

Sum 99·8 99·9 100 100 99·5 99·4 99·3 99·3 99·6 99·4 99·9 100

Formula (O¼ 4)

Si 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Ti 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Al 1·92 1·93 1·94 1·90 1·88 1·90 1·92 1·94 1·91 1·93 1·97 1·95

Cr 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01

Mg 0·48 0·50 0·53 0·51 0·53 0·55 0·32 0·41 0·42 0·41 0·40 0·29

Mn 0·00 0·00 0·00 0·00 0·00 0·01 0·02 0·02 0·02 0·02 0·00 0·00

Fe3þ 0·08 0·07 0·06 0·10 0·11 0·10 0·08 0·05 0·08 0·07 0·03 0·04

Fe2þ 0·52 0·49 0·46 0·48 0·46 0·44 0·65 0·58 0·56 0·57 0·60 0·70

Ni 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Zn 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01 0·01

Sum 3·00 3·00 3·00 3·00 3·00 3·00 3·00 3·00 3·00 3·00 3·00 3·00

XMg 0·48 0·50 0·53 0·51 0·54 0·56 0·33 0·41 0·43 0·42 0·40 0·29

hercynite 47·7 45·9 43·4 43·6 40·9 39·1 61·9 54·8 52·3 53·8 58·0 67·9

spinel 47·9 50·4 53·2 51·1 53·3 55·3 32·3 40·8 41·9 41·3 40·2 29·4

magnetite 3·99 3·46 3·04 4·92 5·43 5·13 3·78 2·74 4·17 3·28 1·67 2·34

galaxite 0·40 0·23 0·36 0·34 0·38 0·51 2·04 1·65 1·65 1·66 0·11 0·35

intergr. w, intergrown with.



An23), was measured with a broadened beam of 25 mm andranges between Or35Ab49An16 and Or46Ab40An16 (Table7; Fig. 9). The composition of micro- to mesoperthiticalkali feldspar of the sapphirine granulite, on the otherhand, was both measured with a broadened beam (c.25 mm) and recalculated from the volumetric proportionsof the alkali feldspar host (average Or78Ab20An2) andplagioclase exsolution lamellae (average Or3Ab70An27) asOr52Ab36An11.Matrix andesine (Ab67^70) of the sapphirine granulites

shows minor orthoclase contents of 53·5mol % and ismore Ca-rich than the oligoclase (Ab70^75; Or 52·5mol%) of the Grt^Opx^Crd^Spl gneisses (Table 7; Fig. 9).The latter shows only minor compositional variations,which are unrelated to its textural position; that is, asmatrix feldspar and as a replacement product of garnet,associated with orthopyroxene and spinel.Matrix plagioclase (Ab65^68) and coronitic plagioclase

resorbing garnet (Ab67^70) of the migmatitic metapeliteare unzoned and characterized by low orthoclase contentsof51mol % (Table 7; Fig. 9).

Biotite

Biotite of the sapphirine granulite was observed indifferent textural positions. Biotite replacing cordierite of

the Spl^Crd symplectites displays the highest XMg of0·77^0·83, whereas biotite replacing orthopyroxene andalkali feldspar (XMg 0·73^0·79) is always less magnesian(Table 8). Biotite inclusions in the peak-metamorphicphases were too small to gain reliable results. Highamounts of F (up to 1·9wt %) and TiO2 (4·3^6·0wt %)in biotite from all textural settings suggest its growth athigh temperatures.The lowest XMg values of 0·66^0·67 are recorded by

matrix biotite and biotite of the Opx^Pl^Spl intergrowthsreplacing porphyroblastic garnet in the Grt^Opx^Crd^Splgneisses, corresponding to lower values of F (up to 1·2wt%) but similar, highTiO2 contents (5·6^5·9wt %). Biotiteof single samples from both rock types is compositionallyuniform and shows no significant chemical zoning.Biotite inclusions in peak-metamorphic garnet of the

migmatitic metapelite display XMg values of 0·69^0·75and variable but high TiO2 contents (4·6^6·3wt %;Table 8). Rare late biotite replacing garnet and cordieriteshows similar XMg values of 0·68^0·74 but trends to lowerTiO2 (3·7^5·1wt %).

Fe^Ti oxides

Matrix ilmenite in the sapphirine granulites is charac-terized by high hematite contents of up to 27mol %,

Table 5: Representative electron microprobe analyses of sapphirine in the sapphirine granulite

Sample: Ro-05-16b Ro-05-16b Ro-05-16b Ro-05-16b Ro-05-15 Ro-05-15 Ro-05-15 Ro-05-15 Ro-05-33B Ro-05-33B

SiO2 14·5 14·2 14·4 14·3 14·3 14·4 15·0 14·2 14·7 14·4

MgO 16·3 16·5 16·1 16·1 16·3 16·6 16·4 16·4 16·5 16·5

Al2O3 58·6 58·8 58·4 58·7 59·2 58·9 58·8 59·2 57·8 59·4

Cr2O3 0·04 0·05 0·03 0·02 0·03 0·03 0·05 0·02 0·04 0·05

FeOtot 10·9 10·9 11·0 10·7 10·1 10·2 10·0 10·0 10·3 10·1

NiO 0·00 0·01 0·00 0·01 0·00 0·00 0·03 0·00 0·00 0·01

ZnO 0·00 0·00 0·00 0·01 0·00 0·04 0·00 0·02 0·00 0·01

MnO 0·05 0·06 0·10 0·08 0·14 0·14 0·14 0·07 0·08 0·09

Sum 100 100 100 99·9 100 100 100 99·9 99·5 101

Formula (O¼ 20)

Si 1·73 1·70 1·73 1·72 1·71 1·72 1·79 1·71 1·77 1·72

Al(4) 4·27 4·30 4·27 4·28 4·29 4·28 4·21 4·29 4·23 4·28

Al(6) 3·99 3·98 4·00 4·03 4·06 4·01 4·07 4·06 3·99 4·06

Fe3þ 0·27 0·32 0·27 0·24 0·22 0·27 0·13 0·23 0·24 0·21

Cr 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Fe2þ 0·82 0·76 0·83 0·83 0·79 0·75 0·86 0·77 0·80 0·79

Mn 0·00 0·01 0·01 0·01 0·01 0·01 0·01 0·01 0·01 0·01

Mg 2·91 2·93 2·89 2·88 2·91 2·95 2·91 2·93 2·96 2·92

Zn 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Ni 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Sum 14·0 14·0 14·0 14·0 14·0 14·0 14·0 14·0 14·0 14·0

XMg 0·78 0·79 0·78 0·78 0·79 0·80 0·77 0·79 0·79 0·79



suggesting late re-equilibration during retrogression of therocks. Magnetite is a common constituent of the leuco-cratic domains of the GrtÔpx^Crd^Spl gneisses, whereit forms subhedral grains or exsolution lamellae in hercyni-tic spinel, displaying an almost pure end-membercomposition.

Whole-rock geochemistrySapphirine granulites and the hosting GrtÔpx^Crd^Splgneisses and migmatitic Grt^Sil metapelites were analyzedfor their bulk composition, using the rock chips left overfrom thin-section preparation. The sapphirine granulitesare characterized by high MgO (10^12wt %; XMg 0·70^0^75) and Al2O3 (23^25wt %), whereas SiO2 (45^50wt

%), Na2O, K2O and CaO are low (Fig. 10, Table 10). Thecomposition of the hosting GrtÔpx^Crd^Spl gneisses ismore variable owing to their heterogeneous texture butalways less magnesian (XMg¼ 0·51^0·69) and slightly lessAl-rich (Al2O3 21^22wt %) than that of the sapphirinegranulite (Table 10). Melanocratic layers of the surround-ing GrtÔpx^Crd^Spl gneisses, composed mainly oforthopyroxene and spinel, are characterized by very lowSiO2 contents (43wt %), moderate XMg values (0·51) andhigh Al2O3 (22wt %), whereas the GrtÔpx-rich melano-cratic zones display strong compositional similarities tothe sapphirine granulites, with almost identical values forXMg (0·69) and Al2O3 (22wt %), but higher SiO2 (54wt%) and CaO. When compared with the hosting Grt^

Table 6: Representative electron microprobe analyses of cordierite in the migmatitic GrtÔpx^Crd^Spl gneiss, the migmati-

tic metapelite and the sapphirine granulite


Sample: Ro-05-16b Ro-05-15 Ro-05-33 Ro-IV-07-03

Texture: Crd Crd Crd Crd Crd sep. Crd sep. Crd sep. rim on sympl.

intergr. intergr. intergr. intergr. Opx and Opx and Opx and Grt–Sill with Kfs

with Spl with Spl with Spl with Spl Spl Spl Spl

Na2O 0·13 0·16 0·26 0·93 0·07 0·07 0·03 0·22 0·36

SiO2 50·7 50·6 50·4 50·7 50·9 51·1 50·6 49·1 49·9

MgO 12·2 12·1 11·8 11·4 12·2 12·3 12·1 10·6 10·5

Al2O3 34·0 34·0 34·1 34·0 34·3 34·2 34·0 33·0 33·5

K2O 0·03 0·02 0·01 0·01 0·03 0·01 0·00 0·01 0·01

CaO 0·02 0·02 0·04 0·02 0·03 0·01 0·02 0·01 0·01

FeO 2·84 3·07 3·11 2·22 2·90 2·01 2·98 4·68 4·71

TiO2 0·00 0·00 0·00 0·00 0·02 0·05 0·00 0·02 0·05

Cr2O3 0·01 0·01 0·00 0·00 0·00 0·00 0·00 0·04 0·00

MnO 0·05 0·03 0·03 0·02 0·05 0·06 0·05 0·05 0·08

Sum 100 100 99·8 99·3 100 99·8 99·7 97·7 99·1

Formula (O¼ 18)

Na 0·03 0·03 0·05 0·18 0·01 0·01 0·01 0·04 0·07

Si 5·00 5·00 4·99 5·04 5·00 5·03 5·00 5·01 5·02

Mg 1·80 1·78 1·74 1·68 1·78 1·81 1·78 1·60 1·57

Al 3·96 3·96 3·98 3·98 3·98 3·97 3·98 3·96 3·97

K 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Ca 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Fetot 0·23 0·25 0·26 0·18 0·24 0·17 0·24 0·40 0·40

Ti 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Cr 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Mn 0·00 0·00 0·00 0·00 0·00 0·01 0·00 0·00 0·01

Sum 11·0 11·0 11·0 11·1 11·0 11·0 11·0 11·0 11·0

XMg 0·88 0·88 0·87 0·90 0·88 0·92 0·88 0·80 0·80

intergr., intergrown; sep., separating.



Opx^Crd^Spl gneisses, the sapphirine granulites alwaysshow higher K2O and LOI (loss of ignition) values, reflect-ing the abundance of texturally late biotite.The metapeliteshows high SiO2, Al2O3 (19wt %) and Fe2O3 (10wt %),minor amounts of CaO, K2O, Na2O, and TiO2 of52wt%, as well as a comparably low XMg value of 0·36.Relationships between the bulk composition and the

observed mineral assemblages of the sapphirine granulitesand the associated gneisses and metapelites are illustratedin a schematic Al2O3^FeO^MgO (AFM) diagram pro-jected from K-feldspar and plagioclase (Fig. 10). The pres-ence of peak-metamorphic Opx^Spr assemblages is inagreement with the high XMg of the sapphirine granulites.In contrast, the hosting and less magnesian sapphirine-freegneisses preserve the peak-metamorphic assemblage oforthopyroxene^spinel^plagioclase, replacing early garnet,therefore plotting in the Opx^Spl stability field. TheFeÂl-rich migmatitic metapelite is situated in the stability

field of Grt^Sil, consistent with the observed mineralassemblage.

P^T^X EVOLUTIONGeothermobarometryGeothermobarometric calculations were performed for thesapphirine granulites, the GrtÔpx^Crd^Spl gneisses andthe migmatitic metapelites to constrain the peak andretrograde P^T conditions. Results and applied geother-mobarometer calibrations are summarized inTable 11.


The P^T conditions during the formation of the peak-metamorphic GrtÔpx^Pl^Qtz assemblage preserved inrare leucocratic domains of the GrtÔpx^Crd^Splgneisses were calculated by combining analyses fromporphyroblastic garnet cores (XMg 0·48; XGrs 0·02) with

Table 7: Representative electron microprobe analyses of plagioclase and alkali feldspar in the migmatitic GrtÔpx^Crd^Spl

gneiss, the migmatitic metapelite and the sapphirine granulite

Sample: Grt–Opx–Spl–Crd gneiss Ro-IV-05-07 Sapphirine granulite Ro-05-16b Meta-pelite

Ro-IV-07-03

Texture: matrix matrix matrix Kfs Kfs Pl Pl matrix Pl repl Pl repl Pl repl Kfs Kfs Pl Pl matrix matrix matrix matrix

Akfs Akfs Akfs host host lam. lam. Pl Grt Grt Grt host host lam. lam. Pl Pl Pl Pl

BB BB BB

SiO2 62·8 63·3 62·8 64·5 64·4 62·4 62·6 61·3 61·2 60·9 62·4 65·7 65·3 61·3 61·1 60·8 60·8 60·5 60·0

MgO 0·01 0·00 0·04 0·00 0·01 0·00 0·00 0·00 0·02 0·00 0·00 0·00 0·00 0·11 0·00 0·10 0·00 0·26 0·00

Al2O3 21·8 22·0 21·6 18·5 18·7 23·9 23·7 24·4 24·4 24·4 23·8 19·2 19·2 25·4 25·3 25·3 25·4 25·3 24·8

K2O 6·91 6·67 7·05 14·90 14·69 0·31 0·40 0·40 0·39 0·45 0·39 13·23 13·75 0·89 0·45 0·62 0·13 0·19 0·14

CaO 3·31 3·27 3·12 0·18 0·19 5·26 5·00 5·79 6·09 5·91 5·06 0·48 0·39 5·61 5·98 5·77 6·82 6·61 6·35

Na2O 5·35 5·49 5·26 1·04 1·10 9·11 9·09 8·27 8·42 8·54 9·16 2·24 1·86 8·30 8·41 8·42 8·26 8·06 7·85

FeO 0·03 0·04 0·09 0·03 0·08 0·04 0·02 0·03 0·06 0·08 0·04 0·07 0·03 0·19 0·05 0·20 0·05 0·34 0·19

TiO2 0·03 0·03 0·01 0·00 0·00 0·00 0·03 0·02 0·00 0·00 0·04 0·04 0·03 0·01 0·03 0·01 0·00 0·00 0·00

BaO 0·07 0·02 0·00 0·06 0·05 0·00 0·02 0·00 0·02 0·00 0·02 0·06 0·06 0·00 0·00 0·00 0·02 0·01 0·00

SrO 0·05 0·04 0·07 0·12 0·15 0·00 0·00 0·00 0·00 0·00 0·00 0·03 0·07 0·01 0·00 0·00 0·00 0·00 0·00

Sum 100 101 100 99·3 99·4 101 101 100 101 100 101 101 101 102 101 101 101 101 99·3

Formula (O¼ 8)

Si 2·83 2·83 2·84 2·99 2·98 2·75 2·76 2·72 2·71 2·71 2·75 2·98 2·97 2·69 2·69 2·68 2·67 2·67 2·69

Mg 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01 0·00 0·01 0·00 0·02 0·00

Al 1·16 1·16 1·15 1·01 1·02 1·24 1·23 1·28 1·28 1·28 1·24 1·02 1·03 1·31 1·31 1·31 1·32 1·31 1·31

K 0·40 0·38 0·41 0·88 0·87 0·02 0·02 0·02 0·02 0·03 0·02 0·76 0·80 0·05 0·02 0·03 0·01 0·01 0·01

Ca 0·16 0·16 0·15 0·01 0·01 0·25 0·24 0·28 0·29 0·28 0·24 0·02 0·02 0·26 0·28 0·27 0·32 0·31 0·31

Na 0·47 0·48 0·46 0·09 0·10 0·78 0·78 0·71 0·72 0·74 0·78 0·20 0·16 0·71 0·72 0·72 0·70 0·69 0·68

Fe 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01 0·00 0·01 0·00 0·01 0·01

Ti 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Ba 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Sr 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Sum 5·02 5·01 5·02 4·99 4·99 5·03 5·02 5·01 5·02 5·03 5·03 4·99 4·99 5·03 5·03 5·04 5·02 5·02 5·00

orthoclase 38·7 37·6 39·8 89·6 88·8 1·66 2·17 2·24 2·15 2·46 2·10 77·7 81·3 4·87 2·44 3·35 0·73 1·02 0·80

albite 45·6 47·0 45·1 9·51 10·12 74·5 75·0 70·5 69·8 70·6 75·0 20·0 16·8 68·7 70·0 69·6 68·2 67·0 68·5

anorthite 15·6 15·5 15·1 0·90 1·03 23·8 22·8 27·3 28·0 27·0 22·9 2·35 1·97 26·4 27·5 27·0 31·1 32·0 30·7

BB, measured with broadened beam; Akfs, alkali feldspar; lam., lamellae.



those of unzoned porphyroblastic matrix orthopyroxene(XMg 0·63; Alpfu/2¼0·195), and associated matrix plagio-clase (An27). Very high temperatures of 1000^10608C arecalculated for a reference pressure of 7·5 kbar, using theGrtÔpx Fe^Mg exchange thermometer calibrations ofLee & Ganguly (1988), Carswell & Harley (1990) and Bhat-tacharya et al. (1991). Similar temperatures of c. 10008C areconstrained with the Al-in-Opx thermometer of Aranovich& Berman (1997), indicating that the porphyroblastic as-semblage in leucocratic domains formed at UHTconditionsof 1020�308C. Uniform, high pressures of 7·5�1 kbar arecalculated for a reference temperature of 10008C, usingthe GrtÔpx^Pl^Qtz barometer calibrations of Newton &

Perkins (1982), Powell & Holland (1988) and Bhattacharyaet al (1991), which are, within error, consistent with pressuresestimated by Al-in-Opx barometry (i.e. 8·2 kbar, calibra-tion of Harley & Green, 1982). Feldspar thermometry ofternary matrix feldspar of the GrtÔpx^Crd^Spl gneisses,using the thermodynamic dataset of Elkins & Grove(1990) and the computer software SolvCalc by Wen &Nekvasil (1994), yields high temperatures of 1060�708C(Fig. 9) for a reference pressure of 7 kbar and thus corrobor-ates the extreme GrtÔpx temperatures. For the calcula-tion of the P^T conditions during garnet replacementthrough Opx^Spl^Pl intergrowths in the melanocratic do-mains of the same samples, analyses from margins of

Table 8: Representative electron microprobe analyses of biotite in the migmatitic GrtÔpx^Crd^Spl gneiss, the migmatitic

metapelite and the sapphirine granulite

Rock type: Grt–Opx–Crd–Spl gneiss Sapphirine granulite Metapelite

Sample: Ro-IV-05-07 Ro-05-14 Ro-05-15 Ro-05-16 Ro-IV-07-03

Texture: matrix matrix matrix repl. repl. Bt repl. Bt repl. Bt repl. Bt repl. repl. Bt repl. Bt repl. Bt repl. Bt repl. Bt repl. Bt repl. incl. in repl.

Bt Bt Bt Grt Grt Kfs Kfs Crd Crd Opx Kfs Kfs Crd Crd Opx Opx Grt Crd

SiO2 37·1 37·4 37·1 37·2 37·2 38·1 38·0 38·2 38·9 38·7 37·9 37·1 38·5 38·9 38·6 38·5 38·51 37·87

Al2O3 15·7 15·3 15·4 15·6 15·4 14·8 14·8 15·6 15·9 15·2 14·9 14·7 15·7 15·6 15·8 15·7 13·29 13·58

TiO2 5·62 5·93 5·89 5·81 5·77 4·93 4·67 4·42 3·94 4·92 5·88 5·24 4·26 4·52 5·88 5·59 6·25 3·74

FeO 12·6 13·2 13·2 13·2 13·0 9·80 9·17 8·61 7·50 8·99 10·43 10·79 8·82 8·41 9·31 10·2 12·21 11·34

MnO 0·18 0·13 0·12 0·14 0·16 0·03 0·00 0·01 0·01 0·01 0·01 0·00 0·05 0·04 0·02 0·05 0·08 0·02

MgO 14·6 14·4 14·6 14·4 14·4 17·7 18·0 19·4 20·5 18·6 16·7 16·6 19·6 19·6 17·6 17·0 15·12 17·38

BaO 0·01 0·05 0·05 0·05 0·11 0·04 0·02 0·06 0·00 0·07 0·03 0·02 0·08 0·01 0·04 0·08 0·02 0·01

CaO 0·03 0·07 0·00 0·01 0·03 0·02 0·02 0·02 0·01 0·00 0·00 0·01 0·10 0·05 0·08 0·06 0·00 0·00

Na2O 0·12 0·12 0·10 0·10 0·11 0·14 0·21 0·07 0·15 0·16 0·15 0·12 0·19 0·22 0·21 0·18 0·07 0·13

K2O 9·68 9·54 9·66 9·74 9·68 9·88 9·80 9·60 9·73 9·98 9·83 9·59 9·23 9·79 9·76 9·35 10·30 10·33

F 1·13 1·18 1·19 1·14 1·13 1·77 1·75 1·52 1·71 1·74 1·22 1·25 1·30 1·51 1·18 1·08 n.d. n.d.

Cl 0·02 0·01 0·02 0·00 0·02 0·02 0·02 0·01 0·00 0·00 0·02 0·03 0·03 0·01 0·01 0·01 n.d. n.d.

Sum 95·6 96·2 96·1 96·3 95·9 95·4 94·7 96·0 96·6 96·5 95·7 94·1 96·6 97·2 97·3 96·6 95·9 94·4

Formula (O¼ 22)

Si 5·45 5·48 5·44 5·45 5·47 5·54 5·55 5·48 5·51 5·53 5·50 5·49 5·49 5·51 5·48 5·50 5·67 5·65

Al(4) 2·55 2·52 2·56 2·55 2·53 2·46 2·45 2·52 2·49 2·47 2·50 2·51 2·51 2·49 2·52 2·50 2·31 2·35

Al(6) 0·17 0·11 0·11 0·13 0·14 0·07 0·09 0·11 0·16 0·09 0·04 0·06 0·12 0·12 0·12 0·15 0·00 0·04

Ti 0·66 0·70 0·69 0·68 0·68 0·58 0·55 0·51 0·45 0·57 0·69 0·62 0·49 0·51 0·67 0·64 0·69 0·42

Fe 1·55 1·62 1·62 1·62 1·60 1·19 1·12 1·03 0·89 1·08 1·27 1·34 1·05 1·00 1·10 1·22 1·50 1·41

Mn 0·02 0·02 0·01 0·02 0·02 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·01 0·00 0·00 0·01 0·01 0·00

Mg 3·19 3·14 3·18 3·14 3·15 3·84 3·92 4·15 4·32 3·95 3·61 3·66 4·17 4·12 3·71 3·63 3·32 3·86

Ba 0·00 0·00 0·00 0·00 0·01 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00

Ca 0·00 0·01 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·00 0·02 0·01 0·01 0·01 0·00 0·00

Na 0·03 0·03 0·03 0·03 0·03 0·04 0·06 0·02 0·04 0·04 0·04 0·03 0·05 0·06 0·06 0·05 0·02 0·04

K 1·81 1·78 1·81 1·82 1·81 1·83 1·83 1·76 1·76 1·82 1·82 1·81 1·67 1·77 1·76 1·71 1·93 1·97

Sum 15·5 15·4 15·4 15·4 15·4 15·6 15·6 15·6 15·6 15·6 15·5 15·5 15·6 15·6 15·4 15·4 15·5 15·7

OH 3·47 3·45 3·45 3·47 3·47 3·18 3·19 3·31 3·24 3·21 3·44 3·41 3·41 3·32 3·47 3·51 n.d. n.d.

F 0·53 0·55 0·55 0·53 0·53 0·82 0·81 0·69 0·76 0·79 0·56 0·59 0·58 0·68 0·53 0·49 n.d. n.d.

Cl 0·00 0·00 0·00 0·00 0·01 0·00 0·00 0·00 0·00 0·00 0·00 0·01 0·01 0·00 0·00 0·00 n.d. n.d.

XMg 0·67 0·66 0·66 0·66 0·66 0·76 0·78 0·80 0·83 0·79 0·74 0·73 0·80 0·81 0·77 0·75 0·69 0·73

n.d., not detected.



Table9:

RepresentativeLA-ICP-M

Straceelementdataofselected

mineralsinsapphirinegranulite

Sam

ple:

Ro-IV-05-16

Ro-IV-05-15

Ro-IV-05-16

Ro-IV-05-15

Ro-IV-05-16

Ro-IV-05-15

Ro-IV-05-16

Ro-IV-05-15

Ro-IV-05-15

Mineral:

Orthopyroxene

Orthopyroxene

Sap

phirine

Sap

phirine

Spinel

Spinel

Cordierite

Cordierite

Zirco

n

Point:

2324

815

31

22

68

1415

1016

1221

core

3co

re13

rim

16rim

9rim

15

Analysis

(ppm)

Li

15·5

18·8

18·7

22·5

35·1

36·6

31·5

50·9

53·8

14·9

125

72·6

539

344

485

414

14·8

17·0

22·1

36·8

34·7

Be

50·33

0·444

50·11

50·32

12·9

7·80

8·22

6·45

50·25

50·20

0·988

0·827

12·1

6·09

4·60

5·64

51·9

51·6

51·4

1·95

51·3

Sc

68·1

57·2

51·7

77·1

11·3

13·3

11·2

13·0

50·28

50·30

0·421

0·376

1·26

0·783

0·631

0·437

n.a.

n.a.

n.a.

n.a.

n.a.

Ti

878

1223

917

1679

1190

1135

1111

956

56·7

69·9

139

59·2

35·3

377

38·1

69·5

510·0

14·3

10·4

10·3

55·4

V68·1

73·1

105

103

151

190

185

208

168

184

230

222

77·9

5·66

23·4

2·71

1·84

50·23

50·27

0·188

2·61

Mn

1550

1632

2148

1868

688

747

795

842

852

921

1029

775

724

358

519

492

7·32

29·1

2·94

9·66

53·8

Rb

3·94

15·3

0·100

17·4

50·06

50·09

50·07

0·384

1·03

0·301

6·19

4·92

13·1

20·1

34·2

21·0

50·12

0·390

0·512

0·134

2·37

Sr

0·391

0·762

0·184

0·490

0·275

0·274

50·08

0·642

0·439

0·053

0·477

0·393

1·79

1·73

6·30

3·64

0·258

3·41

0·457

0·525

2·22

Y14·6

14·4

57·6

38·5

6 ·90

9·86

9·37

12·4

0·050

50·03

0·088

0·101

0·306

0·099

0·176

50·10

744

913

1558

2767

1008

Zr

5·92

5·01

8·91

10·7

0·221

0·199

0·231

0·232

50·03

50·06

50·05

50·05

50·10

50·05

50·07

50·13

n.a.

n.a.

n.a.

n.a.

n.a.

Nb

0·218

0·643

50·02

1·40

50·05

50·04

50·05

50·05

50·03

0·066

50·03

50·03

50·06

0·584

50·03

0·256

1·52

1·24

1·51

1·50

1·37

Ba

0·844

4·24

50·18

4·87

0·347

50·27

50·29

50·31

1·24

0·388

0·534

0·745

0·704

4·03

11·8

2·99

50·51

1·93

1·32

50·32

18·0

La

50·05

0·021

50·02

50·08

50·03

50·04

50·05

50·04

50·02

50·04

50·02

50·03

50·02

50·03

50·03

50·06

50·07

50·08

50·07

50·05

0·09

Ce

50·06

50·03

50·02

50·03

50·02

50·03

50 ·03

50·03

50·02

50·03

50·01

50·02

50·04

50·02

50·02

50·04

4·40

4·74

3·62

1·55

5·12

Pr

50·06

50·02

50·02

50·05

50·01

50·03

50·05

50·02

50·02

50·02

50·02

50·01

50·02

50·01

50·01

50·05

0·071

0·090

50·05

0·077

0·088

Nd

50·15

50·12

50·07

50·38

50·11

50·19

50·20

50·16

50·10

50·06

50·09

50·16

50·17

50·09

50·09

50·24

0·945

1·69

0·393

0·583

0·781

Sm

50·21

50·10

50·14

50·38

50·22

50·25

50·33

50·23

50·12

50·18

50·06

50·08

50·20

50·16

50·17

50·45

1·68

2·16

3·17

2·12

3·22

Eu

50·08

50·04

50·04

50·13

50·05

50·06

50·04

50·06

50·02

50·04

50·03

50·03

50·08

50·03

50·05

50·07

50·12

0·234

0·150

0·126

0·778

Gd

50·13

50·13

0·948

1·03

50·20

0·275

0·415

0·219

50·16

50·17

50·11

50·15

50·12

50·05

50·13

50·30

9·67

12·0

17·7

19·0

16·3

Tb

0·069

0·090

0·379

0·282

0·031

0·072

0·060

0·043

50·02

50·03

50·02

50·02

50·03

50·01

50·03

50·08

3·78

5·41

8·41

10·8

5·99

Dy

1·43

1·25

5·90

4·58

0·468

0·998

0·907

0·869

50·07

50·09

50·09

50·06

50·10

50·08

50·05

50·25

52·1

69·4

124

158

81·7

Ho

0·568

0·540

1·72

1·36

0·221

0·372

0·280

0·368

50·02

50·03

50·02

50·02

50·03

50·06

50·02

50·07

23·3

29·1

48·3

66·1

33·5

Er

3·09

2·87

8·38

8·69

1·28

1·32

1·25

2·04

50·05

50·07

50·08

50·07

50·10

50·06

50·06

50·13

118

141

235

300

171

Tm

0·802

0·832

1·59

1·90

0·231

0·385

0·303

0·391

50·02

50·02

50·02

50·03

50·05

50·01

50·02

50·06

27·0

30·8

54·6

59·9

35·7

Yb

7·03

6·37

13·4

15·5

2 ·22

3·19

2·93

4·08

50·11

50·11

50·09

50·12

50·18

50·09

50·11

50·25

279

319

541

539

359

Lu

1·16

1·17

1·99

2·99

0·356

0·490

0·552

0·713

50·03

50·03

50·03

50·02

50·04

50·02

50·02

50·06

n.a.

n.a.

n.a.

n.a.

n.a.

Hf

0·862

0·601

0·886

1·75

50·11

50·13

50·13

0·105

50·07

50·09

50·08

50·08

50·09

50·05

50·06

50·28

12017

14237

12986

16233

9288

Ta

50·04

0·036

0·045

50·08

50·04

50·05

50·06

50·04

50·03

0·036

50·02

50·02

50·05

0·075

50·04

50·09

0·816

1·41

1·27

1·50

0·912

Ca

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

5178

5177

5165

5106

5195

Cr

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

54·3

50·42

53·6

52·7

50·6

Co

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

0·509

0·543

50·19

0·248

0·964

Ni

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

1·21

51·2

50·89

50·69

5·47

Cu

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

50·92

50·89

4·28

50·57

93·3

Zn

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

56·6

56·7

56·1

54·4

58·1

Ga

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

0·589

50·23

50·19

0·744

1·89

Sn

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

50·58

50·61

0·856

0·390

6·02

Cs

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

50·07

50·07

50·06

50·04

0·159

n.a.notan

alysed

.



resorbed garnet (XMg 0·47; XGrs 0·03^0·02) are combinedwith those of the replacing, adjacent Al-rich orthopyroxene(XMg 0·63; Al

pfu/2¼0·20), and associated unzoned plagio-clase (An28). Pressures of 7·5�1 kbar are calculated usingthe GrtÔpx^Pl^Qtz barometer (calibrations as above),which are identical to the pressures calculated for the por-phyroblastic assemblage but represent maximum values, asquartz is absent in the reaction texture. Remarkably, how-ever, the GrtÔpx^Pl pressures are close to the Al-in-Opx

pressures (i.e. 8·1 kbar, calibration of Harley & Green,1982). Corresponding temperatures, calculated from GrtÔpx Fe^Mg and Al-in-Opx thermometry (calibrations asabove and for a reference pressure of 7·5 kbar) range be-tween 980 and 10308C (1000�308C), and hence are similarto those calculated for the peak-metamorphic GrtÔpxpairs of the leucocratic domains.Symplectitic late-stage garnet (XMg 0·40^0·43) inter-

grown with low-Al-orthopyroxene (XMg 0·67^0·68; Alpfu/2

Fig. 8. Ternary SiO2^(Mg,Fe2þ)O^(Al,Fe3þ,Cr)2O3 diagram illustrating the compositional variation of sapphirine in the sapphirine granu-lites. Positions of the 2:2:1 and 7:9:3 endmembers are marked. The compositional trend runs almost parallel to theTschermaks substitution line.



up to 0·14), on the other hand, formed at lower P^Tcondi-tions of 760�308C (both GrtÔpx Fe^Mg exchange andAl-in-Opx thermometry with calibrations as above) and 3^5 kbar (Al-in-Opx barometry of Harley & Green,1982).

Migmatitic Grt^Sil metapelites

Peak pressures experienced by the garnet^sillimanite-bearing migmatitic metapelite were estimated for plagio-clase (An31) and coexisting garnet (XMg 0·38; Grs0-2)from the Grt^Sil^Pl^Qz (GASP) equilibrium (calibrationsof Newton & Haselton, 1981; Koziol & Newton, 1988;Powell & Holland, 1988) yielding a pressure of 7·5�0·5kbar (for a reference temperature of 10008C), which is in

agreement with the pressure estimates for the associatedGrtÔpx^Crd^Spl gneisses. Pressure calculations for theformation of coronitic plagioclase resorbing garnetthrough the GASP equilibrium have been performed bycombining garnet rim compositions (XMg 0·38; Grs0-2)with coronitic plagioclase (An30). The results (for a refer-ence temperature of 10008C) are about 1 kbar lower(6·5�0·7 kbar) than the peak pressures.

Sapphirine granulites

Peak temperatures of 1075�1108C were estimated fromthe Fe^Mg exchange thermometer of Kawasaki & Sato(2002) using the compositions of coexisting unzoned

Fig. 9. Compositions of alkali feldspar and plagioclase in the sapphirine granulites and GrtÔpx^Crd^Spl gneisses, plotted in the ternary feld-spar diagram. Solvi at 7 kbar were calculated with the software SOLVCALC (Wen & Nekvasil, 1994) using the feldspar activity model ofElkins & Grove (1990).



porphyroblastic Al-rich orthopyroxene (XMg 0·69^0·75;Alpfu/2 up to 0·20) and coexisting sapphirine (XMg 0·81^0·78). Similar extreme temperatures of 1090�508C8Cwere estimated for a reference pressure of 7 kbar for theformation of ternary feldspar from feldspar solvus therm-ometry (Elkins & Grove, 1990; Fig. 9).

P^T pseudosectionsTo constrain detailed P^T paths of the migmatitic metape-lites, the sapphirine granulites, and hosting migmatiticGrtÔpx^Crd^Spl gneisses pseudosections were calcu-lated in the NCKFMASHT system, using theTHERIAK-DOMINO software (v. 20/03/07) of DeCapitani & Brown (1987), with the internally consistentthermodynamic dataset of Holland & Powell (1998;THERIAK-DOMINO filename tcdb55c2d.txt) and themineral activity models of Baldwin et al. (2005) for feld-spar,White et al. (2007) for garnet, biotite, spinel, ilmenite,orthopyroxene and liquid, Holland & Powell (1998) forcordierite, and Kelsey et al. (2004) for sapphirine.To ensure a close correspondence between the observed

mineral assemblages and the calculated pseudosections,sample compositions were determined on the rock chipsleft over from the thin-section preparation. The chemistryof the sapphirine granulites and the migmatitic Grt^Silmetapelites is adequately expressed in the systemNCKFMASHT and that of the hosting, partly Ti-freeGrtÔpx^Crd^Spl gneisses in the NCKFMASH system,as additional components (e.g. Zn, Cr in spinel) are

present at only trace element level and will thus not signifi-cantly affect the topology of the pseudosections. Absence(in garnet and orthopyroxene) or low amounts (in sapphir-ine and spinel) of (calculated) Fe3þ suggests reducingconditions of metamorphism. For the dry and weaklyretrogressed GrtÔpx^Crd^Spl gneisses, water contentsare directly taken from the determined LOI. Sapphirinegranulites and migmatitic Grt^Sil metapelites, on theother hand, always show variably elevated water contents,owing to the formation of biotite during retrogression indi-cating post-peak H2O influx. As the present study focuseson the reconstruction of the peak-metamorphic conditionswe have calculated the pseudosections for the least retro-graded sapphirine granulite sample and the migmatitcGrt^Sil metapelite sample for a reduced water content.Based on a comparison with the water contents of theleast retrogressed samples of the associated GrtÔpx-Spl^Crd granulites, containing �1mol % H, we performedcalculations assuming a low water content of 1mol %H. These low H2O values are realistic for the formation ofthe ‘dry’ peak-metamorphic mineral assemblages of boththe sapphirine granulite and the Grt^Sil metapelite andmoreover in accordance with the generally assumed ordetermined low H2O content of such UHT rocks(e.g. Kelsey et al., 2004; Brandt et al., 2011). As a conse-quence the retrograde segment of the P^T evolutioninferred for these samples is only semi-quantitative.Temperatures constrained for the crystallization of the par-tial melt at the solidus and for the regrowth of biotite

Fig. 10. Relationship between bulk-rock composition and mineral assemblages of the sapphirine granulite, migmatitic metapelite, and migma-titic GrtÔpx^Crd^Spl gneiss, illustrated in a schematic Al2O3^FeO^MgO diagram projected from H2O, plagioclase, and K-feldspar.



represent maximum and minimum values, respectively, ashigher H2O contents would shift the solidus towardslower temperatures and would expand the biotite stabilityfield towards higher temperatures. The studied samplesare migmatites or restitic domains of migmatites thatunderwent significant melt and H2O loss and hence adramatic change of the bulk-rock composition near or atpeak-metamorphic conditions. Therefore, the prograde

segment of the inferred P^T path, as inferred from min-eral inclusions in the porphyroblastic minerals combinedwith calculated pseudosections, must be regarded as onlysemi-quantitative. To refine the estimation of the peak con-ditions we have calculated isopleths for the aluminiumcontents (calculated as XMg-Tschermaks component¼Alpfu/2)of orthopyroxene, which is robust against post-peakFe^Mg exchange, and the XMg [calculated as molar

Table 10: Bulk-rock geochemistry of representative samples of the sapphirine granulites and associated migmatitic metape-

lites and Grt^Opx^Crd^Spl gneisses


Sample: RO-IV-05-16 RO-IV-05-15 RO-IV-05-14 RO-05-33B RO-IV-05-09 RO-IV-05-07 RO-05-33B2 RO-IV-07-03

Opx–Spl rich Grt–Opx rich

wt %

SiO2 48·9 45·2 47·5 49·7 52·6 42·5 53·6 61·1

Al2O3 24·0 24·8 23·3 24·3 21·9 21·5 21·2 19·1

Fe2O3 6·86 8·17 8·71 8·77 7·23 19·5 8·83 9·79

MnO 0·06 0·08 0·08 b.d.l. 0·1 0·91 b.d.l. 0·05

MgO 9·00 12·1 10·8 10·2 5·76 10·1 9·71 2·83

CaO 1·31 0·85 1·09 2·07 3·17 1·46 2·05 0·94

Na2O 2·32 1·21 1·68 2·13 3·75 2·27 2·49 1·29

K2O 4·39 4·25 4·16 3·28 3·95 1·22 2·82 1·88

TiO2 0·74 0·98 0·82 0·89 0·71 50·10 0·60 1·14

P2O5 0·07 0·05 0·06 0·06 0·07 0·08 0·07 0·03

LOI 3·13 3·41 2·44 0·80 0·95 0·04 0·72 2·06

Sum 101 101 101 102 100 99·6 102 100

XMg 0·72 0·75 0·71 0·70 0·61 0·51 0·69 0·36

ppm

Ba 269 208 251 212 256 176 238 457

Ce n.a. n.a. n.a. 530 n.a. n.a. 530 n.a.

Co n.a. n.a. n.a. 67 n.a. n.a. 78 n.a.

Cr 49·8 66·5 54·2 45 52·6 49·3 36 122

Ga 23·2 29·1 23·0 24 22·0 515 27 23

Hf n.a. n.a. n.a. 10·7 n.a. n.a. 9·1 n.a.

La n.a. n.a. n.a. 540 n.a. n.a. 50 n.a.

Nb 13·1 16·2 12·9 18 15·1 50·0 14 17

Nd n.a. n.a. n.a. 22 n.a. n.a. 510 n.a.

Pb 510 510 510 b.d.l. 10·2 510 510 13

Rb 140 153 149 110 218 20·5 111 64

Sc n.a. n.a. n.a. 21 n.a. n.a. 19 n.a.

Sr 90·1 62·6 71·2 115 284 99·3 130 127

Th 13·5 510 510 b.d.l. 24·8 510 520 520

V n.a. n.a. n.a. 90 n.a. n.a. 67 n.a.

Y 38·5 41·5 31·7 137 23·1 771 40 43

Zn 76·6 72·1 80·3 114 87·7 43·8 108 145

Zr 251 268 234 270 219 394 217 245

b.d.l., below detection limit; n.a., not analysed.



Mg/(MgþFe)] and XGrs [calculated as molar Ca/(CaþMgþFeþMn)] of garnet. Peak P^Tconditions were con-strained by comparing the contours with the analysedXMg and XGrs of garnet (calculated as above) and the Alcontent of orthopyroxene (calculated as Alpfu/2).

Migmatitic Grt^Sil metapelite

The pseudosection for the Si- and Fe-rich migmatitic meta-pelite (bulk rock XMg 0·36) is characterized by the stabilityof Grt^Sil assemblages (þ ilmenite, quartz, plagioclase andK-feldspar) over a very large P^T range (Fig. 11a). Owingto the low H2O content used for the calculations the solidusis situated at high temperatures of 800^9108C. At subsolidusconditions,58008C and pressures45·5 kbar, Grt^Sil coex-ists with biotite. At temperatures above the solidus and

pressures46·5^7 kbar biotite is replaced by melt. At lowerpressures Grt^Sil coexists with cordierite. Towards tempera-tures 41020^10508C Grt^Sil is progressively replaced bySpr^Qtz assemblages at pressures 46·5^7 kbar and bySpl^Qtz assemblages at pressures56·5^7 kbar. A clockwiseP^T path is constrained for the migmatitic Grt^Sil metape-lite, based on the following criteria (Fig. 11a).

(1) Prograde inclusions of biotite and sillimanite in coarsegarnet and quartz indicate that the early mineral as-semblage Grt^Bt^Sil^Kfs^Pl^Qtz^Ilm was partiallyconsumed by melting reactions during prograde heat-ing to T48008C (maximum temperature owing toprograde H2O and melt loss), thereby crossing the sol-idus at P47 kbar.

Table 11: Results of geothermobarometric calculations for the sapphirine granulites and the migmatitic metapelites and Grt^

Opx^Crd^Spl gneisses

Sample Texture Temperature (8C) Pressure (kbar)

Grt–Opx Fe–Mg Al in Spr–Opx Feldspar Grt–Opx–Pl–Qtz Al in GASP

Opx Opx

Pref LG CH B AB KS EG Tref NP PH B (Mg) HG PH NH KN

Migmatitic Grt–Sil metapelite

Peak

Ro-IV-07-03 Grt(core)–Sil–

Pl(matrix)–Qtz

950 — — — — 6·4 6·8 7·5

1000 — — — — 7·1 7·5 8·1

Garnet replacement

Ro-IV-07-03 Grt(rim)–Sil–

Pl(corona)–Qtz

950 — — — — 5·5 5·9 6·6

1000 — — — — 6·1 6·5 7·2

Sapphirine granulite

Peak

Ro-IV-05-14 Spr–Opx–Fsp 7·0 — — — — 1075� 110 1090� 50

Grt–Opx–Crd–Spl gneiss

Porphyroblastic

Ro-IV-05-07 Grt(core)-Opx(porph.)–

Pl(matrix)–Qtz

7·5 1054 999 1009 994 — 1060� 70 980 7·7 7·1 6·7 7·6 — — —

8·0 1058 1003 1016 1004 — — 1000 7·8 7·2 6·8 8·2 — — —

Garnet replacement

Ro-IV-05-07 Grt(margin)–

Opx(repl.)–Pl(repl.)

7·5 1033 979 989 990 — — 980 7·8 7·1 6·8 7·5 — — —

(maximum pressure!) 8·0 1037 983 996 1001 — — 1000 7·9 7·2 6·9 8·1 — — —

Symplectitic

Ro-IV-05-07 Grt(sympl.)–

Opx(sympl.)

3·5 789 744 751 753 — — 760 — — — 3·4 — — —

4·0 792 747 757 762 — — 800 — — — 4·8 — — —

AB, Aranovich & Berman (1997); B, Bhattacharya et al. (1991); B (Mg), Bhattacharya et al. (1991), Mg-exchange; CH,Carswell & Harley (1990); EG, Elkins & Grove (1990); HG, Harley & Green (1982); KN, Koziol & Newton (1988); KS,Kawasaki & Sato (2002); LG, Lee & Ganguly (1988); NH, Newton & Haselton (1981); NP, Newton & Perkins (1982); PH,Powell & Holland (1988).



(2) The absence of early cordierite suggests that the fur-ther evolution proceeded through the large trivariantfield Grt^Sil^Kfs^PlÎlm^Qtz^melt at P46·5^7·2kbar andT48008C.

(3) The studied Grt^Sil metapelite displays the peak-metamorphic assemblage of Grt^Spl^Sil^Kfs^PlÎlm^Qtz^melt, which is stable in a narrow divariant

field at 6·6^7·2 kbar and 990^10408C.The upper tem-perature limit for the peak conditions is given by theabsence of sapphirine, which is stable at T410208C.Based on the composition of porphyroblastic garnet(XMg up to 0·40; XGrs 0·015^0·020) P^Tconditions of10008C and 6·6 kbar are estimated for the formationof the peak assemblage, broadly consistent with the

Fig. 11. P^T pseudosections for (a) the migmatitic metapelite, (b) the migmatitic GrtÔpx^Crd^Spl gneiss, and (c) the sapphirine granulite,contoured for the XAlpfu/2 of orthopyroxene (migmatitic GrtÔpx^Crd^Spl gneiss and sapphirine granulite) and the XMg and XGrs of garnet(migmatitic GrtÔpx^Crd^Spl gneiss and metapelite). The bulk compositions are given as normalized mole proportions of theNCKFMASH(T) components. P^T paths are derived from mineral reaction textures and compositions. Deduced peak P^Tconditions overlapwith results from the thermobarometric calculations (c. 10008C, 7·5 kbar; Table 11). Bold dashed line marks the solidus. (See text for discussionof proposed P^T paths.)

(continued)



results of the GASP barometry and indicating slightlydecreasing pressures during heating to UHT condi-tions. The formation of spinel^quartz intergrowthsbetween porphyroblastic spinel and quartz probablyoccurred in the same stability field.

(4) The growth of cordierite between spinel and quartz isconsistent with entry into the divariant field Grt^Spl^Sil^Kfs^PlÎlm^Qtz^melt through decompres-sion still at UHTconditions.

(5) Subsequent entry into the divariant field Grt^Crd^Sil^Kfs^PlÎlm^Qtz^melt through decompression^cooling is evident from the growth of cordierite

coronas between garnet, sillimanite and quartz,which was accompanied by the formation of plagio-clase coronas around garnet that formed through theGASP equilibrium.

(6) Regrowth of garnet (XMg 0·30^0·35), which forms ascoronas around ilmenite and spinel, thereby replacingpost-peak cordierite, is not clearly evident from thepseudosection. However, calculated isopleths for themode of garnet abundance have a positive slope inthe trivariant Grt^Crd^Sil^Kfs^PlÎlm^Qtz field(not shown) and indicate garnet growth with decreas-ing temperature, suggesting that late garnet forms in

Fig. 11. Continued



response to post-decompressional near-isobaric cool-ing at sub-solidus conditions.

(7) The limited growth of late biotite, which forms at theexpense of retrograde cordierite, suggests cooling toT57008C. Biotite regrowth was accompanied bycontinued garnet regrowth, as indicated by garnetrims around late biotite, consistent with cooling.Biotite formation is most probably related to reactionwith crystallizing melt.This is apparently inconsistentwith the position of the solidus at 8008C. However,the high solidus temperature is an artefact of the low-ered water content used for the pseudosection

calculations. Applying the analysed water content ofc. 11mol % H (taken from the LOI), representativefor the retrograde evolution, the solidus is shifted toT57008C, consistent with inferred biotite formationthrough interaction with melt.


The calculated pseudosections for several samples of theGrt^Opx^Crd^Spl gneiss are rather similar and thereforeare illustrated for only the least altered sampleRo-IV-05-07 (Fig. 11b). Because sample Ro-IV-05-07

Fig. 11. Continued



contains Ti below the detection limit, consistent with theabsence of Ti phases, the pseudosection was calculated inthe NCKFMASH system. The topology of the pseudosec-tion is characterized by the large stability field of orthopyr-oxene and the absence of quartz over the wholeinvestigated P^T range (700^11008C, 4^8 kbar). Quartz ispresent in the thin section, but only in rare leucocratic sub-domains, which are not represented by the overall Si-poorbulk chemistry. Owing to the low H2O content and thehigh bulk XMg, the solidus is located at high temperaturesof 850^9508C. At sub-solidus conditions, orthopyroxenecoexists with spinel^K-feldspar^plagioclase^biotite.Additional cordierite is stable at low pressures (5 c. 4·5^5·5 kbar) and is replaced by garnet at higher pressures (4c. 5^5·5 kbar). At elevated pressures 46·5 kbar garnet isprogressively replaced with increasing temperature byorthopyroxene and spinel (þ feldspar, liquid), which arestable in a large pentavariant stability field. Garnet finallydisappears at T4900^10008C. Contours of XMg indicatethat garnet becomes less magnesian with deceasing P andT, whereas Al-in-Opx isopleths demonstrate that theamount of Al in orthopyroxene increases withtemperature.We have constrained a clockwise P^T path for the mela-

nocratic domains of the GrtÔpx^Crd^Spl gneiss(bulk-rock XMg 0·51), which is based on the following ob-servations (Fig. 11b).

(1) Biotite and spinel inclusions in porphyroblastic garnetand orthopyroxene, respectively, suggest an early stagesubsolidus assemblage of GrtÔpx^Bt^Spl^Kfs^Pl,which is stable in a large trivariant field at P45·5kbar and T58508C. Owing to prograde H2O andmelt loss experienced by the GrtÔpx^Crd^Spl gneis-ses these temperatures must be regarded as maximumvalues.

(2) During heating to T49008C biotite was completelyconsumed via melt-producing reactions. The resultingassemblage garnetôrthopyroxene^spinel^plagioclasecoexisting with melt, representing the observed por-phyroblastic assemblage of the melanocratic sample,is stable at P46·7 kbar and 910^9908C.

(3) At higher temperatures garnet becomes unstable andthe observed decomposition of porphyroblasticgarnet (XMg up to 0·47) to form orthopyroxene^spinel^plagioclase intergrowths coexisting with meltis in agreement with continued heating toT4910^9908C and entry into the pentavariant Opx^Spl^Fsp^Liq field. Based on the composition of ortho-pyroxene from the Opx^Spl^Pl domains (Alpfu/2 upto 0·20), combined with a peak pressure of 7·5�1kbar, as calculated by barometry for the rare quartz-bearing subdomains, UHTconditions of c. 9908C areconstrained for the formation of the Opx^Spl^Plintergrowths (Fig. 11b), which are consistent with the

maximum thermal stability of garnet at a pressure of7·5 kbar in the pseudosection (c. 9608C) and are ingood agreement with results of GrtÔpx thermom-etry for the garnet replacement (1000�308C).

(4) The occurrence of symplectitic alkali-feldspar andbiotite in the outermost zones of the Opx^Pl^Splpseudomorphs after garnet testifies to mineral^meltreactions during cooling into the trivariant Opx^Spl^Bt^Kfs^Pl^Liq field atT58758C.

(5) The reappearance of biotite is locally accompanied bythe growth of cordierite as a narrow reaction rim be-tween orthopyroxene and spinel. Cordierite forma-tion, plus continued biotite growth, is consistent withentry into the trivariant Opx^Spl^Crd^Bt^Kfs^Plfield, which occurs at subsolidus conditions and lowpressures of 55·5 kbar. Hence cordierite formationproves marked decompression of the order of c. 2kbar from peak pressures at still high temperatures ofc. 800^8508C (Fig. 11b).

(6) The formation of euhedral garnet (XMg 0·39^0·40;XGrs 0·02) in association with Al-poor orthopyroxene(XMg 0·67^0·68; Al pfu/2 up to 0·14) overgrowing boththe Opx^Spl^Pl symplectites and Al-rich porphyro-blastic orthopyroxene is consistent with re-entry intothe GrtÔpx^Bt^Spl^Kfs^Pl trivariant field throughcontinued cooling to �800^7508C at a pressure of c.5·5^5 kbar (Fig. 11b), broadly in agreement with thethermobarometric data for this stage (760�308C, 3^5kbar). Regrowth of euhedral garnet formed at theexpense of late biotite also indicates cooling.

Sapphirine granulites

The topology of all pseudosections for the sapphirinegranulites is rather similar and therefore is illustrated foronly sample Ro-05-16 (Fig. 11c). Orthopyroxene is stableover the entire P^Twindow (700^11008C, 4^8 kbar). Thesolidus is situated at a high temperature of 9008C. At sub-solidus conditions orthopyroxene coexists with biotite(plus other phases). Coexisting sapphirine and orthopyrox-ene are stable at temperatures both above and below thesolidus. The following P^T path was established for thesapphirine granulites based on the observed mineral reac-tion sequence (Fig. 11c).

(1) Biotite inclusions in orthopyroxene and sapphirineindicate initial temperature conditions 59008C.The lack of spinel, corundum, garnet or cordieriteinclusions, which are stable together with biotite atsubsolidus conditions in various assemblages,might be explained by either their complete con-sumption during the melting reaction or by a differ-ent, less aluminous bulk composition prior to partialmelting.

(2) During heating biotite was consumed through ortho-pyroxene- and sapphirine-producing melt reactions,



which initiate at temperatures49008C. The inferredhighest-grade assemblage SprÔpx^Pl^Kfs^melt isstable together with ilmenite and rutile in a large sta-bility field at UHT conditions of 900^9908C andP46 kbar and is limited towards higher temperaturesby the incoming of spinel coexisting with sapphirine.Based on the composition of Al-rich porphyroblasticorthopyroxene (Alpfu/2 up to 0·20) coexisting withsapphirine, peak-metamorphic P^T conditions of950^10008C at 7^8 kbar are constrained from theAl-in-Opx isopleths, which are slightly lower than thethermometric data for the sample (Opx^Spr therm-ometry and Fsp thermometry: c. 1050^11008C), butare consistent with the P^T data estimated from theGrt^Sil metapelites and the GrtÔpx^Crd^Splgneisses. Both rutile and ilmenite are inferred phasesin the melt-present Pl^KfsÔpx^Spr peak assemblageof the pseudosections, but rutile is not observed in thesamples. The lack of rutile can be explained by the in-corporation of high amounts of Ti in orthopyroxeneat peak conditions, which is evidenced by ilmeniteplatelets exsolved from orthopyroxene during cooling.However, Ti is not included in the available orthopyr-oxene mixing models. This explains the mismatch be-tween the calculated and observed occurrence ofrutile.

(3) Peak metamorphism is followed by decompression of1·5^2 kbar, as documented by the formation of cor-dierite^spinel symplectites between porphyroblasticsapphirine and orthopyroxene; this indicates entryinto the narrow 2FspÔpx^Crd^Spl^RtÎlm^Liqfield at a pressure of55·7 kbar and still UHT tem-peratures of 910^9508C.

(4) Subsequent regrowth of biotite at the expense oforthopyroxene, sapphirine, and symplectitic cordier-ite reflects interactions between peak-metamorphic aswell as early retrograde mineral assemblages withcrystallizing melt during cooling to subsolidus tem-peratures59008C. This comparably high solidus tem-perature must be regarded as a maximum value aspseudosection modelling for this sample was per-formed with a reduced H2O content.

(5) The late formation of corundum in the sapphirinegranulites is in agreement with continued cooling toT57508C at c. 5 kbar.

U^PB LA- ICP-MS DAT ING OFZ IRCON AND XENOTIME ANDTI- I N-Z IRCON THERMOMETRYU^Pb zircon datingTwo sapphirine granulite samples (i.e. Ro-IV-05-14 andRo-05-33B) were selected for U^Pb LA-ICP-MS analysis.

In both samples, large, well-rounded zircon grains of50^100 mm diameter are randomly distributed throughoutthe feldspar^biotite matrix (Fig. 12a) or occur as raretiny inclusions in sapphirine and orthopyroxene. In add-ition, zircon is especially abundant within the decompres-sional cordierite^spinel reaction textures replacingporphyroblastic sapphirine and orthopyroxene (Fig. 12b).CL and BSE images of zircons from all textural settingsreveal a complex internal structure for most grains(Fig. 12c^f). They exhibit corroded cores with an internalfine-scale oscillatory zoning (Fig. 12c and d) that is inter-preted to reflect magmatic zircon growth. In most grains,the oscillatory zoned core is surrounded by a thinrim (525 mm), which is not zoned and occurs apparentlyrandomly relative to the core. These rims, which we inter-pret as metamorphic overgrowths, are characterized bya slightly higher BSE brightness than the oscillatoryzoned cores and display a very weak, homogeneous CLwith sharp boundaries against the magmatic zones(Fig. 12c^f). The textural relationships indicate that theoscillatory zoned zircon cores are older than the rims.Rare internally featureless and anhedral zircon occursin the cordierite^spinel reaction rims between orthopyr-oxene and sapphirine as very small (515 mm) grainsthat display a similar homogeneous luminescence tothe overgrowth rims on corroded zircon cores. Basedon the lack of any zoning and similarity to the meta-morphic rims they are interpreted as completely newlygrown.Oscillatory zoned zircon cores have variable contents of

Y (740^2600 ppm) and Ti (up to 25 ppm; Table 9) as wellas variable U (130^1800 ppm) and Th/U ratios (0·11^0·90;Fig. 13; Table 12). The U, Y and Ti contents of the meta-morphic zircon rims and small single grains are alsovariable and partly even higher (U 280^2600 ppm, Y960^2800 ppm, Ti up to 55 ppm; Tables 9 and 12), whereastheir Th/U ratios are, with only three exceptions, very low(0·02^0·19; Table 12), consistent with their inferred meta-morphic growth (Fig. 13). When compared with the cor-roded zircon cores, they display stronger enrichment ofLREE over HREE (Table 9).In sample Ro-05-33B, a total of 45 U^Pb analyses were

performed in 25 large oscillatory zoned cores with a beam8^20 mm in diameter. Forty-two of these analyses have adiscordance lower than 5%. Their 207Pb/206Pb ages rangefrom 1841�26 to 1220� 40Ma with a clear cluster be-tween 1495 and 1470Ma (21 analyses, Fig. 14a). This distri-bution strongly suggests a detrital origin for the zirconcores, representing various igneous source rocks. In add-ition, 20 analyses were acquired with a 8^12 mm beam onnarrow metamorphic rims in 12 zircon crystals. Fourteenconcordant analyses (discordance 52%) collected onhomogeneous metamorphic rims wider than 15 mm, to-gether with four concordant analyses of xenotime



Fig. 12. Photomicrographs of representative zircon grains in sapphirine granulite Ro-IV-05-14 (a, b, e, f, BSE images; c, d, CL images). (a)Zircon occurs as inclusions in biotite and perthitic alkali feldspar of the leucocratic matrix. (b) Zircon is abundant in granoblastic cordieriteof the Crd^Spl reaction rims. (c, d) Zircon inclusions in cordierite of the Crd^Spl intergrowth showing oscillatory zoned cores and broad meta-morphic rims of up to 20 mm with a very weak and homogeneous CL. (e, f) Broad, homogeneous BSE-bright metamorphic rims surroundingoscillatory zoned cores are partly overgrown by anhedral BSE-bright xenotime. Images are taken after LA-ICP-MS analysis (note small diam-eter of LA-ICP-MS craters in zircon cores and rims of c. 20 mm and c. 10 mm, respectively). LA-ICP-MS U^Pb results are indicated as207Pb/206Pb ages with 2s errors.



(discordance 52%) displaying a similar age range (seebelow), define a concordia age of 1006�4Ma (Fig. 14a).In sample Ro-IV-05-14, 35 analyses were performed on

21 crystals, comprising 24 analyses of oscillatory zonedcores, eight analyses of metamorphic rims and three ana-lyses of metamorphic zircon grains in cordierite^spinelsymplectites that are large enough to be measured. Thebeam diameter was changed between 8 and 16 mm toremain smaller than the target. The 207Pb/206Pb ages ofthe oscillatory zoned cores vary between 1501�21 and1265�54Ma with a cluster of the concordant data at1501^1455Ma (Fig. 14b and Table 12). Although youngerand/or older core ages are not detected in sampleRo-IV-05-14, a detrital origin of the magmatic coresseems very likely, based on data for the associated sapphir-ine granulite Ro-05-33B. Metamorphic zircon rims andtwo newly formed metamorphic zircon grains in Crd^Splreactions rims around sapphirine yield very similar207Pb/206Pb ages of 989� 49 to 1064�38Ma and of993�23 to 1029�32 Ma, respectively (Table 12). Ten con-cordant analyses (discordance 54%) of metamorphiczircon grains and rims define a concordia age of1010�7Ma (Fig. 14b), very similar to that of sampleRo-05-33B (1006�4 Ma) and pointing to metamorphicgrowth of zircon during regional Sveconorwegian meta-morphism M1 in both samples.

U^Pb xenotime datingSingle grains of xenotime, as well as xenotime intergrownwith the zircon margins (U 1210^5950 ppm, Th/U 0·9^6·3), occur in sample Ro-05-33B. They show a strong,homogeneous BSE response and, like the metamorphiczircon rims, define similar Sveconorwegian concordant207Pb/206Pb ages of 1001^979Ma (five analyses).

In contrast, small, BSE-bright anhedral xenotime(520 mm; U 970^3490 ppm, Th/U 0·8^4·2), epitacticallygrown on the margins of zircon in samples Ro-05-33Band Ro-IV-05-14 (Fig. 12f), yields younger concordia agesof 933�5Ma (eight spots) and 928�10Ma (three spots),respectively (Fig. 14a and b).

Ti-in-zircon thermometryTemperatures during metamorphic zircon formation in thesapphirine granulites at c. 1000Ma were calculated fromthe titanium concentration in zircon, using the revisedTi-in-zircon thermometer calibration of Ferry & Watson(2007), based on the original version of the thermometerby Watson et al. (2006). In their modified Ti-in-zirconthermometer, Ferry & Watson (2007) took into accountthatTi could take the place of either Zr or Si in zircon, fol-lowing reaction (1) ZrSiO4þTiO2¼ZrTiO4þ SiO2 or(2) SiO2þTiO2¼TiSiO4. Experimental data demon-strate that the Ti content in zircon increases with decreas-ing aSiO2 and increasing aTiO2. Following this, the revisedthermometer of Ferry & Watson (2007) involves theactivities of both TiO2 and SiO2 in the rock system. Attemperatures of 400^10008C the thermometer returnstemperatures with an estimated uncertainty of �108 orbetter. Using unconstrained aSiO2 and aTiO2 additionalmaximum uncertainties are estimated as 60^708C at7508C.In the case of the sapphirine granulites of Rogaland,

rutile was not identified but minor ilmenite is present in-stead as Ti-rich phase, suggesting that the activity of titan-ium is almost equal to unity or only slightly lower in theserocks. Quartz, on the other hand, is absent from oursamples, providing evidence for a certain amount ofsilica undersaturation in the rock system and indicatingsilica activities �0·5 and51. Following this, temperatures

Fig. 13. Th/U vs 207Pb/206Pb age diagram for the analysed zircon of the sapphirine granulites. Magmatic cores generally show higher Th/Uratios than metamorphic zircon rims (core^rim mixture analyses are excluded from the diagram).



Table12:U^PbLA-ICP-M

SanalysesforsapphirinegranulitesamplesRo-05-33B

andRo-IV-05-14

Analysis

Spot

Calc.

Texture

Position

207Pb*

Uy

Pby

Thy/

206Pb/

206Pbz/

�2s

207Pbz/

�2s

207Pbz/

�2s

rho§

Age(M

a).

Disc

(mm)

(c.p.s.)

(ppm)

(ppm)

U204Pb

238U

(%)

235U

(%)

206Pb

(%)

207Pb/

�2s

206Pb/

�2s

207Pb/

�2s

(%)

235U

238U

206Pb

Zirco

nin

matrixmineralsofRo-05-33b

16-1_1

20detritalco

rein

Bt

30771

481

121

0·27

34536

0·2424

2·0

2·998

2·3

0·08968

1·2

0·85

1399

251407

181419

231

16-1_2

20detritalco

rein

Bt

55745

705

188

0·26

18393

0·2572

2·6

3·268

2·8

0·09215

1·0

0·93

1475

351473

221471

190

16-1_3

20detritalco

rein

Bt

56120

746

192

0·38

10587

0·2437

2·0

2·996

2·2

0·08917

1·0

0·90

1406

251407

171408

180

13-2_1

16co

re–rim

mixture

inOpx

17469

338

710·23

1042

0·2032

2·5

2·243

3·2

0·08006

2·1

0·76

1192

271195

231198

410

13-2_2

16detritalco

rein

Opx

16825

462

133

0·57

6041

0·2628

2·3

3·348

2·9

0·09240

1·8

0·78

1504

301492

231476

341

13-3_1

16detritalco

rein

Opx

22446

515

143

0·41

11563

0·2590

1·9

3·321

2·5

0·09302

1·6

0·76

1485

251486

201488

300

13-3_2

16co

re–rim

mixture

inOpx

29806

1000

207

0·16

11080

0·2064

3·0

2·443

3·2

0·08587

1·2

0·93

1209

331256

231335

236

14-1_1

16co

re–rim

mixture

inSpr

13079

355

820·22

14964

0·2281

2·4

2·732

2·9

0·08685

1·7

0·81

1325

281337

221357

331

14-1_2

16co

re–rim

mixture

inSpr

15250

373

930·21

3786

0·2470

2·2

3·004

2·4

0·08822

1·0

0·91

1423

281409

181387

192

14-1_3

16detritalco

rein

Spr

11962

280

690·25

13184

0·2383

2·0

2·990

2·4

0·09097

1·4

0·81

1378

251405

191446

273

11c-1_

18

xmetam

orphic

rim

inKfs

7721

837

141

0·05

2456

0·1715

2·5

1·722

3·1

0·07283

1·8

0·81

1020

241017

201009

371

11c-1_

216

detritalco

rein

Kfs

25398

607

164

0·18

27691

0·2645

2·6

3·363

2·9

0·09221

1·3

0·90

1513

351496

231472

252

11c-1_

316

detritalco

rein

Kfs

31605

778

207

0·19

20370

0·2587

3·6

3·290

4·0

0·09225

1·6

0·92

1483

481479

311473

300

11c-1_

416

detritalco

rein

Kfs

39950

882

240

0·18

42541

0·2640

2·5

3·397

2·6

0·09332

0·8

0·95

1510

341504

211494

151

11a-1_

18

xmetam

orphic

rim

inOpx

4617

634

105

0·08

4708

0·1676

2·0

1·685

2·7

0·07292

1·9

0·72

999

181003

181012

381

11a-1_

28

xmetam

orphic

rim

inOpx

3105

561

920·02

3130

0·1715

2·6

1·733

3·5

0·07331

2·4

0·75

1020

251021

231022

480

11a-1_

312

detritalco

rein

Opx

25688

1160

307

0·27

18723

0·2565

2·7

3·246

3·1

0·09177

1·5

0·87

1472

361468

251463

290

11a-1_

412

xmetam

orphic

rim

inOpx

19122

1548

300

0·45

26494

0·1717

2·6

1·719

2·9

0·07261

1·3

0·90

1022

251016

191003

261

2-1_

18

detritalco

rein

Kfs

3968

339

880·14

1006

0·2479

3·3

3·121

4·2

0·09132

2·6

0·79

1427

431438

331453

491

2-1_

216

detritalco

rein

Kfs

45318

1083

311

0·25

47754

0·2721

2·5

3·566

2·6

0·09505

0·7

0·96

1552

341542

211529

131

6-1_

216

detritalco

rein

Kfs

16784

424

112

0·14

2501

0·2582

2·5

3·322

2·8

0·09333

1·2

0·91

1481

331486

221495

221

6-1_

316

detritalco

rein

Kfs

49021

1320

357

0·18

13932

0·2638

2·1

3·383

2·4

0·09302

1·0

0·90

1509

291501

191488

201

8-1_

416

detritalco

rein

Opx

22652

531

146

0·21

12599

0·2638

2·4

3·442

2·8

0·09464

1·3

0·88

1509

331514

221521

250

8-1_

516

detritalco

rein

Opx

36270

1239

273

0·13

16648

0·2194

2·4

2·666

2·7

0 ·08813

1·1

0·91

1279

281319

201385

215

Zirco

nin

Crd–S

pl(–B

t)zones

ofRo-05-33B

17-1_1

20detritalco

rein

Crd

32086

461

121

0·13

7619

0·2653

2·0

3·410

2·6

0·09320

1·6

0·77

1517

271507

211492

311

17-1_2

12detritalco

rein

Crd

14286

493

129

0·19

2678

0·2584

2·5

3·311

3·0

0·09294

1·6

0·84

1482

331484

231487

300

18-2_1

16detritalco

rein

Crd

16914

362

970·24

18191

0·2626

1·9

3·342

2·6

0·09232

1·7

0·75

1503

261491

201474

321

18-2_2

16detritalco

rein

Crd

15451

363

880·20

17153

0·2393

2·7

3·010

2·9

0·09123

1·2

0·91

1383

331410

231451

233

18-1_1

16detritalco

rein

Crd

10034

240

610·26

3755

0·2558

2·5

3·242

2·9

0·09191

1·6

0·84

1468

321467

231466

310

18-1_2

16detritalco

rein

Crd

16882

358

101

0·45

18194

0·2589

2·4

3·317

2·8

0·09294

1·4

0·87

1484

321485

221487

270

18-1_3

16detritalco

rein

Crd

18262

383

105

0·34

19771

0·2555

2·5

3·253

2·8

0·09235

1·2

0·91

1467

331470

221475

220

18-4_1

16detritalco

rein

Crd

14953

348

900·28

8971

0·2472

2·6

3·052

3·2

0·08955

1·8

0·83

1424

341421

241416

340

18-4_2

16detritalco

rein

Crd

16624

380

101

0·33

3618

0·2542

2·2

3·233

2·5

0·09225

1·2

0·87

1460

281465

191472

230

18-6_1

16detritalco

rein

Crd

20441

446

117

0·25

21906

0·2553

2·4

3·260

2·8

0·09260

1·5

0·85

1466

321472

221480

291

(continued

)



Table12:Continued

Analysis

Spot

Calc.

Texture

Position

207Pb*

Uy

Pby

Thy/

206Pb/

206Pbz/

�2s

207Pbz/

�2s

207Pbz/

�2s

rho§

Age(M

a).

Disc

(mm)

(c.p.s.)

(ppm)

(ppm)

U204Pb

238U

(%)

235U

(%)

206Pb

(%)

207Pb/

�2s

206Pb/

�2s

207Pb/

�2s

(%)

235U

238U

206Pb

18-6_2

12x

metam

orphic

rim

inCrd

5747

400

670·11

4508

0·1708

2·1

1·701

2·9

0·07224

1·9

0·74

1016

201009

19993

392

18-6_3

16detritalco

rein

Crd

26967

552

154

0·36

8241

0·2614

2·3

3·351

2·5

0·09297

1·0

0·91

1497

301493

201487

190

15-1_1

16detritalco

rein

Crd

15684

470

101

0·16

5864

0·2137

2·9

2·386

3·5

0·08095

2·0

0·82

1249

321238

251220

401

14-2_1

16detritalco

rein

Crd

9635

134

540·82

4104

0·3280

2·4

5·089

2·8

0·11253

1·4

0·86

1829

391834

241841

260

14–2_2

16detritalco

rein

Crd

10224

158

560·56

9384

0·3124

2·4

4·766

3·2

0·11064

2·2

0·74

1753

361779

271810

402

14-2_3

12co

re–rim

mixture

inCrd

6533

241

590·30

6123

0·2297

2·3

2·882

3·7

0·09099

2·9

0·61

1333

271377

281446

565

14-3_1

12x

metam

orphic

rim

inCrd

6887

538

880·10

9669

0·1690

2·1

1·675

2·7

0·07190

1·7

0·78

1007

19999

17983

342

14-3_2

16detritalco

rein

Crd

21813

718

180

0·59

2913

0·2344

2·0

2·769

2·4

0·08569

1·5

0·80

1357

241347

181331

281

14-3_3

12detritalco

rein

Crd

15480

904

192

0·21

5965

0·2078

2·9

2·454

3·7

0·08565

2·3

0·78

1217

321259

271330

455

11b-3_1

12x

metam

orphic

rim

inBt

12888

921

155

0·10

2098

0·1711

2·4

1·704

2·7

0·07223

1·3

0·88

1018

231010

18993

272

Analysis

Spot

Calc.

Texture

Position

207Pb*

Uy

Pby

Thy/

206Pb/

206Pbz/�2s

207Pbz/�2s

207Pbz/�2s

rho§

Age(M

a)Disc.

(mm)

(c.p.s.)

(ppm)

(ppm)

U204Pb

238U

(%)

235U

(%)

206Pb

(%)

207Pb/�2s

206Pb/�2s

207Pb/�2s

(%)

235U

238U

206Pb

Zirco

nin

Crd–S

pl(–B

t)zones

ofRo-05-33B

11b-3_2

8x

metam

orphic

rim

inBt

4714

569

960·14

2071

0·1671

1·8

1·679

2·6

0·07287

1·8

0·70

996

171001

171010

371

11b-3_3

12detritalco

rein

Bt

32477

1096

242

0·18

4562

0·2164

2·1

2·519

2·5

0·08443

1·3

0·84

1263

241278

181302

262

11b-3_4

12detritalco

rein

Bt

23811

1774

372

0·23

2978

0·2036

1·8

2·321

2·1

0·08267

1·0

0·87

1195

201219

151261

203

11b-5_1

16detritalco

rein

Crd

19007

665

160

0·13

7668

0·2409

1·5

2·901

2·3

0·08734

1·7

0·65

1392

191382

181368

341

11b-1_1

12detritalco

rein

Crd

8807

389

105

0·15

2622

0·2630

2·9

3·355

3·5

0·09252

2·0

0·83

1505

401494

281478

371

11b-1_2

8detritalco

rein

Crd

4251

409

109

0·13

1455

0·2592

4·4

3·248

4·8

0·09090

2·0

0·91

1486

591469

381445

382

11b-1_3

12detritalco

rein

Crd

16130

651

179

0·20

17130

0·2611

3·8

3·354

4·1

0·09316

1·6

0·93

1495

511494

331491

300

11b-1_4

12detritalco

rein

Crd

7019

292

790·26

7502

0·2677

2·1

3·433

2·9

0·09300

1·9

0·75

1529

291512

231488

362

11b-3_1

8x

metam

orphic

rim

inBt

5387

946

156

0·02

956

0·1698

2·5

1·717

2·9

0·07333

1·6

0·85

1011

231015

191023

311

11b-3_2

8x

metam

orphic

rim

inBt

7655

1442

236

0·03

4986

0·1668

2·0

1·676

3·1

0·07290

2·4

0·63

994

181000

201011

491

11b-3_3

16detritalco

rein

Bt

28594

770

203

0·17

4322

0·2559

2·9

3·195

3·2

0·09054

1·2

0·92

1469

381456

251437

241

11b-3_4

16detritalco

rein

Bt

66076

938

309

0·63

13218

0·2644

1·7

3·422

1·9

0·09385

1·0

0·87

1513

231509

151505

180

9-4_

18

xmetam

orphic

rim

inBt

1962

403

670·10

1069

0·1683

2·4

1·695

3·4

0·07304

2·4

0·71

1003

221007

221015

481

9-4_

212

xmetam

orphic

rim

inBt

28408

2082

441

0·83

12061

0·1687

2·3

1·674

2·8

0·07194

1·5

0·84

1005

22999

18984

301

9-4_

38

xmetam

orphic

rim

inBt

2306

518

850·05

3307

0·1700

2·9

1·709

3·6

0·07290

2·1

0·81

1012

281012

231011

420

9-4_

412

detritalco

rein

Bt

7367

490

117

0·43

1704

0·2182

2·3

2·712

3·4

0·09016

2·5

0·68

1272

261332

251429

477

9-2_

28

xmetam

orphic

rim

inCrd

7151

1362

224

0·06

9167

0·1703

2·5

1·712

3·1

0·07288

1·7

0·83

1014

241013

201011

340

9-2_

38

core–rim

mixture

inCrd

12082

1934

369

0·11

5374

0·1925

2·9

2·227

3·2

0·08389

1·3

0·91

1135

301189

221290

268

9-2_

412

detritalco

rein

Crd

24571

922

257

0·22

26272

0·2631

2·6

3·381

3·1

0·09318

1·7

0·84

1506

351500

251492

321

(continued

)



Table12:Continued

Analysis

Spot

Calc.

Texture

Position

207Pb*

Uy

Pby

Thy/

206Pb/

206Pbz/�2s

207Pbz/�2s

207Pbz/�2s

rho§

Age(M

a)Disc.

(mm)

(c.p.s.)

(ppm)

(ppm)

U204Pb

238U

(%)

235U

(%)

206Pb

(%)

207Pb/�2s

206Pb/�2s

207Pb/�2s

(%)

235U

238U

206Pb

1-1_

112

detritalgrain

inSpl

28021

3424

603

0·10

33097

0·1791

2·4

1·892

4·1

0·07662

3·2

0·60

1062

241078

271111

653

5-2_

312

detritalco

rein

Crd

62536

1809

522

0·90

17376

0·2130

2·2

2·605

2·6

0·08868

1·5

0·82

1245

251302

191397

297

5-1_

112

detritalco

rein

Crd

17519

885

217

0·28

20043

0·2293

2·3

2·758

2·6

0·08726

1·3

0·88

1331

281344

201366

242

3-2_

116

detritalco

rein

Crd

24324

555

149

0·19

12564

0·2584

2·4

3·337

2·5

0·09365

0·8

0·95

1482

321490

201501

151

Xen

otimein

matrixmineralsofRo-05-33b

11-c

xt_1

12x

metam

orphic

grain

inBt

56863

2776

1001

3·15

7690

0·1703

2·1

1·702

2·3

0·07250

0·9

0·91

1014

191009

151000

191

11-c

xt_2

8x

metam

orphic

grain

inBt

45472

5946

1436

0·88

7444

0·1709

2·2

1·708

2·5

0·07246

1·3

0·85

1017

201011

16999

271

11-c

xt_3

8x

metam

orphic

grain

inBt

8056

1839

636

3·29

2771

0·1682

1·8

1·664

2·5

0·07176

1·7

0·72

1002

17995

16979

352

11-c

xt_4

8metam

orphic

grain

inBt

6892

1564

465

2·39

9490

0·1641

1·8

1·637

2·7

0·07235

2·0

0·68

980

17985

17996

401

6-1_

18

xmetam

orphic

grain

inKfs

10804

2795

847

2·95

4375

0·1550

1·9

1·487

2·6

0·06959

1·8

0·73

929

17925

16916

371

8-1_

18

xmetam

orphic

grain

inOpx

6184

1583

579

4·20

3948

0·1562

2·3

1·512

3·2

0·07020

2·3

0·70

936

20935

20934

470

8-1_

212

xmetam

orphic

grain

inOpx

91362

2190

506

1·61

45551

0·1564

1·7

1·531

2·1

0·07099

1·3

0·80

937

15943

13957

261

8-1_

38

xmetam

orphic

grain

inOpx

13263

3487

754

1·35

6666

0·1548

2·1

1·491

2·8

0·06985

1·8

0·77

928

19927

17924

360

Xen

otimein

Crd–S

pl(–B

t)zones

ofRo-05-33B

18-6_0

12x

metam

orphic

grain

inCrd

18480

1836

449

2·21

9794

0·1559

1·8

1·493

2·0

0·06947

0·9

0·89

934

15928

12913

182

9-2_

112

xmetam

orphic

grain

inCrd

72042

2700

876

3·13

53940

0·1558

1·6

1·512

1·9

0·07042

1·1

0·81

933

14935

12941

231

5-2_

112

xmetam

orphic

grain

inCrd

8992

972

254

2·06

3321

0·1572

1·8

1·522

2·3

0·07024

1·5

0·76

941

16939

14935

310

5-2_

218

xmetam

orphic

grain

inCrd

7572

1846

512

2·46

2873

0·1545

2·2

1·493

2·7

0·07005

1·5

0·83

926

19927

16930

310

3-1_

116

xmetam

orphic

grain

inCrd

116364

1212

610

6·29

31104

0·1677

1·8

1·677

1·9

0·07254

0·7

0·92

999

161000

121001

150

Analysis

Spot

Calc.

Texture

Position

207Pby

Uy

Pby

Thy/

206Pb/

206Pbz/�2s

207Pbz/�2s

207Pbz/�2s

rho§

Age(M

a)Disc.

(mm)

(c.p.s.)

(ppm)

(ppm)

U204Pb

238U

(%)

235U

(%)

206Pb

(%)

207Pb/�2s

206Pb/�2s

207Pb/�2s

(%)

235U

238U

206Pb

Zirco

nin

matrixmineralsofRo-IV-05-14

14-9-2_1

12x

meta.

rim

matrixBt

21233

836

139

0·10

52276

0·1695

2·1

1·725

2·6

0·07378

1·6

0·78

1010

191018

171035

332

14-9-2_2

16detritalco

rematrixBt

17567

257

630·33

5339

0·2353

2·4

2·928

2·8

0·09028

1·4

0·86

1362

301389

221431

273

14-9-3_1

12meta.

rim

matrixKfs

31474

1259

220

0·19

23065

0·1780

1·9

1·813

2·3

0·07385

1·3

0·83

1056

181050

151037

261

14-9-3_2

12detritalco

rematrixKfs

19441

489

137

0·42

21544

0·2598

1·9

3·298

2·5

0·09208

1·6

0·77

1489

261481

201469

311

14-9-1_1

8detritalco

rematrixKfs

8605

493

121

0·87

7035

0·2410

4·3

2·970

4·8

0·08939

2·0

0·91

1392

541400

371412

381

14-9-1_2

8x

meta.

rim

matrixKfs

2474

284

480·14

5619

0·1674

1·9

1·665

3·1

0·07212

2·4

0·61

998

19995

20989

491

14-9-1_3

16detritalco

rematrixKfs

28785

394

102

0·21

61655

0·2530

2·0

3·240

2·4

0·09289

1·2

0·85

1454

261467

181486

241

14-10-1_

18

detritalco

rematrixKfs

6424

303

800·23

7758

0·2559

2·9

3·298

3·9

0·09345

2·7

0·73

1469

381480

311497

501

14-10-1_

216

detritalco

rematrixKfs

16615

216

500·34

36114

0·2287

2·3

2·854

3·0

0·09053

1·8

0·79

1327

281370

231437

355

14-7-1_1

8detritalco

rematrixKfs

4471

289

750·48

9737

0·2580

4·0

3·325

4·6

0·09349

2·3

0·87

1479

531487

371498

441

(continued

)



Table12:Continued

Analysis

Spot

Calc.

Texture

Position

207Pby

Uy

Pby

Thy/

206Pb/

206Pbz/�2s

207Pbz/�2s

207Pbz/�2s

rho§

Age(M

a)Disc.

(mm)

(c.p.s.)

(ppm)

(ppm)

U204Pb

238U

(%)

235U

(%)

206Pb

(%)

207Pb/�2s

206Pb/�2s

207Pb/�2s

(%)

235U

238U

206Pb

14-7-1_2

8x

meta.

rim

matrixKfs

4689

464

740·02

12686

0·1670

2·2

1·669

3·0

0·07247

2·0

0·74

996

21997

19999

410

14-7-1_3

16detritalco

rematrixKfs

38322

516

146

0·49

7492

0·2519

2·5

3·210

2·7

0·09242

1·0

0·93

1448

321460

211476

191

14-7-2_3

16detritalco

rematrixOpx

16583

225

590·23

4457

0·2534

3·3

3·229

3·7

0·09243

1·7

0·89

1456

431464

291476

321

14-1-1_2

16grain

matrixKfs

30813

431

116

0·29

67883

0·2595

3·9

3·350

4·4

0·09363

2·0

0·89

1487

521493

351501

381

Zirco

nin

Crd–S

pl(–B

t)reactionzones

ofRo-IV-05-14

14-8-1_3

8x

meta.

rim

Crd–S

pl

4930

568

990·19

13731

0·1711

2·0

1·711

2·8

0·07254

2·0

0·70

1018

191013

181001

411

14-8-1_4

16detritalco

reCrd–S

pl

40410

594

152

0·33

11724

0·2452

2·1

3·087

2·6

0·09131

1·5

0·81

1414

261429

201453

292

14-6-1_1

16detritalgrain

Crd–S

pl

64919

806

234

0·50

10375

0·2622

2·8

3·385

3·0

0·09363

1·1

0·93

1501

381501

241501

210

14-4-7_1

16detritalgrain

Crd–S

pl

29108

476

109

0·33

67953

0·2263

3·2

2·716

3·5

0·08705

1·4

0·91

1315

381333

261362

272

14-4-2_1

12x

meta.

grain

Crd–S

pl

19062

893

150

0·12

8499

0·1724

2·1

1·737

2·7

0·07306

1·7

0·77

1025

201022

171016

351

14-4-2_2

16x

meta.

grain

Crd–S

pl

12525

309

540·15

33860

0·1708

2·0

1·732

2·5

0·07354

1·6

0·78

1017

181021

161029

321

14-4-1_1

16x

meta.

grain

Crd–S

pl

22052

572

980·16

59983

0·1717

1·6

1·710

1·9

0·07225

1·1

0·82

1021

151012

13993

232

14-4-3_1

12detritalco

reCrd–S

pl

16791

531

108

0·11

4118

0·2075

4·5

2·369

5·3

0·08281

2·8

0·85

1216

501233

381265

542

14-4-3_2

12detritalco

reCrd–S

pl

17361

525

116

0·12

40978

0·2234

2·3

2·650

3·1

0·08601

2·0

0·75

1300

271315

231338

392

14-4-3_3

16detritalco

reCrd–S

pl

19299

314

720·15

44996

0·2283

2·8

2·704

4·1

0·08591

3·0

0·68

1325

331330

301336

570

14-4-5_1

8detritalco

reCrd–S

pl

4538

238

560·16

4956

0·2285

6·5

2·812

7·8

0·08927

4·4

0·83

1326

781359

601410

844

14-4-5_2

12rim–core

mixture

Crd–S

pl

13591

533

990·31

6790

0·1791

4·8

1·961

5·3

0·07941

2·3

0·90

1062

471102

371182

467

14-4-5_3

8x

meta.

rim

Crd–S

pl

29148

2594

433

0·18

7773

0·1678

2·0

1·731

2·7

0·07484

1·9

0·73

1000

181020

181064

384

14-4-6_1

16detritalco

reCrd–S

pl

37885

492

132

0·31

82178

0·2548

1·8

3·230

2·2

0·09196

1·2

0·85

1463

241464

171467

220

14-4-8_1

16rim–core

mixture

Crd–S

pl

40651

765

172

0·31

26020

0·2165

2·0

2·471

2·3

0·08280

1·0

0·89

1263

231264

171264

200

14-4-4_1

8x

meta.

rim

Crd–S

pl

5092

361

610·08

3530

0·1693

2·1

1·722

3·3

0·07377

2·6

0·64

1008

201017

221035

522

14-4-4_2

16detritalco

reCrd–S

pl

17495

239

630·21

10064

0·2545

2·1

3·217

2·5

0·09169

1·3

0·85

1462

281461

191461

250

14-4-5_1

8x

meta.

rim

Crd–S

pl

3121

310

550·30

3896

0·1668

2·3

1·663

3·3

0·07231

2·4

0·69

994

21994

21995

490

14-4-5_2

8rim–core

mixture

Crd–S

pl

6093

299

760·17

13107

0·2479

6·8

3·123

7·4

0·09138

2·9

0·92

1428

881438

591455

561

14-4-5_3

12detritalco

reCrd–S

pl

36202

804

217

0·27

14625

0·2583

2·5

3·304

2·8

0·09278

1·3

0·88

1481

331482

221483

250

14-4-5_4

12detritalco

reCrd–S

pl

29510

770

202

0·24

66557

0·2558

2·6

3·238

3·0

0·09181

1·5

0·87

1468

351466

241464

280

Xen

otimein

matrixmineralsofRo-IV-05-14

14-7-2_1

8x

meta.

grain

inOpx

18766

2163

456

1·69

52225

0·1544

1·9

1·507

2·4

0·07076

1·5

0·78

926

16933

15950

302

14-7-2_2

8x

meta.

grain

inOpx

10981

1455

246

1·08

31882

0·1555

1·8

1·488

2·9

0·06955

2·2

0·64

930

16925

18915

461

Xen

otimein

Crd–S

pl(–B

t)reactionzones

ofRo-IV-05-14

14-8-1_2

8x

meta.

grain

inCrd

2942

1216

181

0·76

5982

0·1539

2·2

1·499

3·5

0·07064

2·7

0·64

923

19930

21947

542

Spotsize

8,12,16

or20

mm;dep

thofcrater�20

mm.206Pb/2

38U

erroristhequad

raticad

ditionofthewithin-runprecision(2

SE)an

dtheexternal

reproducibility

(2SD)ofthereference

zircon.207Pb/2

06Pberrorpropag

ation(207Pbsignal

dep

enden

t)followingGerdes

&Zeh

(2009).207Pb/2

35U

erroristhequad

raticad

ditionofthe

207Pb/2

06Pban

d206Pb/2

38U

uncertainty.Calc.,datausedforco

nco

rdia

agecalculation;Disc.,disco

rdan

ce;meta.,metam

orphic.

*Within-runbackg

round-correctedmean

207Pbsignal

inc.p.s.(counts

per

seco

nd).

yU

andPbco

ntentan

dTh/U

ratiowerecalculatedrelative

toGJ-1

reference

zircon.

zTim

e-resolved

datawereco

rrectedforbackg

round,within-runPb/U

fractionationan

dco

mmonPbusingStacey&

Kramers(1975)

model

Pbco

mpositionan

dsubsequen

tlynorm

alized

toGJ-1

(ID-TIM

Svalue/measuredvalue);207Pb/2

35U

calculatedusing

207Pb/2

06Pb/(

238U/2

06Pb�1/137·88).

§Rhois

the

206Pb/2

38U/2

07Pb/2

35U

errorco

rrelationco

efficien

t.



Fig. 14. U^Pb concordia diagrams for LA-ICP-MS analysis of zircon and xenotime from two sapphirine granulite samples. (a) Sapphirinegranulite sample Ro-05-33B; (b) sapphirine granulite sample Ro-IV-05-14. Insets illustrate concordia diagrams for the growth of metamorphiczircon and xenotime during the regional Sveconorwegian metamorphism (M1) and of xenotime during post-collisional decompression and in-trusion of the anorthosite^mangerite^charnockite suite (M2). Xenotime analyses are marked as bold grey lines in the upper inset of (a). conc,concordance; MSWD (CþE), concordance and equivalence.



were calculated using determining factors of (1)aSiO2¼ aTiO2¼1 and (2) aSiO2¼0·5 and aTiO2¼1.Titanium concentrations in the oscillatory zoned zirconcores (11 spots) range from 11·6 to 25·1ppm Ti (Table 9),yielding temperatures of 760^840�508C, assuming silicaand titanium saturation in the rock system, and minimumtemperatures of 760^840�508C using a reduced Siactivity of 0·5. The Ti contents of the homogeneousmetamorphic zircon rims (six spots) are 10·3^55·4 ppm(Table 9), yielding temperatures of 750^940�508C and690^850�508C, for aSiO2¼1 and 0·5, respectively.Following this, maximum temperatures of c. 850^9408Care constrained for metamorphic zircon growth at1006�4Ma during M1, consistent with (U)HTconditions.

DISCUSS IONOrigin of the sapphirine granulitesMagnesiumâluminium-rich sapphirine granulites havebeen regarded as the product of several possible processes,including metamorphism of hydrothermally altered maficto ultramafic rocks or high-Mg clays (e.g. Sheraton, 1980;Harley et al., 1990), syn-metamorphic metasomatism (e.g.Herd et al., 1969; Vry & Cartwright, 1994; Dunkley et al.,1999) or formation of restitic MgAl-rich domains in meta-sedimentary rocks via partial melting (e.g. Clifford et al.,1981; Baba, 2003; Brandt et al., 2007). For the sapphirinegranulites of Ivesdalen (XMg 0·70^0·75) a metasedimen-tary origin is evidenced by their apparent structural andcompositional resemblance to the GrtÔpx-rich melano-cratic layers of the surrounding banded GrtÔpx^Spl^Crd gneisses and their partly diffuse contact relationshipsto the latter. A metasedimentary origin is also supportedby the abundance of inherited detrital igneous zircon inthe sapphirine granulites. The major part of the oscillatoryzoned zircon cores displays nearly concordant 207Pb/206Pbapparent ages ranging from 1841 to 1220 Ma, consistentwith a sedimentary origin derived from different igneoussources and deposition at51220 Ma. The large populationof 1501^1455Ma igneous zircon suggests emplacement ofthe igneous protoliths during the Telemarkian igneousevent at 1520^1480 Ma, defined by Bingen et al. (2008a,2008b). Magmatic zircon with preserved oscillatoryzoning has also been identified by Mo« ller et al. (2002,2003) in a number of ortho- and paragneiss samples fromthe same area, yielding significantly younger ages of c.1035Ma and c. 1050 Ma. The lack of such detrital zirconin our samples is consistent with sedimentation of theirprotoliths prior to these igneous events. An alternative pro-tolith could be metasomatically altered S-type granite, aspostulated for the formation of GrtÔpx-bearing alumin-ous migmatitic gneisses from Rogaland by Mo« ller et al.(2003). Regarding the close association of the sapphirinegranulites with GrtÔpx-rich metasedimentary units andthe large range of almost concordant ages of the oscillatory

zoned zircon in our samples, this scenario is rather un-likely. Most probably, these rocks represent Grt(Ôpx)-rich metasedimentary layers that were metamorphosed tosapphirineôrthopyroxene granulites and GrtÔpx^Crd^Spl gneisses during heating to UHTconditions.

P^T^t evolutionBased on the observation of sequential reaction texturesand stable mineral assemblages, combined with constraintsfrom pseudosections in the NCKFMASH(T) system andthermobarometric calculations, a clockwise P^T path isinferred for the sapphirine granulites, the host GrtÔpx^Spl^Crd gneisses and associated Grt^Sil metapelites ofIvesdalen, SW Norway (Fig. 15). This P^T path comprisesheating to MPÛHTconditions (c. 7·5 kbar, 10008C) fol-lowed by near-isothermal (U)HT decompression toP55·5 kbar at 900^10008C and subsequent near-isobariccooling toT5750^8008C at c. 5 kbar. Observed reactiontextures and results of the thermobarometric calculationsmostly correlate well with predictions from pseudosectionsin the NCKFMASH(T) system.The P^T path can be sub-divided into the following stages (Fig. 15).

Peak metamorphism

Relics of porphyroblastic prograde garnet and orthopyrox-ene associated with matrix plagioclase and quartz in leu-cocratic domains of the GrtÔpx^Spl^Crd gneisses testifyto peak metamorphism at temperatures of c. 10208Cand pressures of c. 7·5 kbar. In the melanocraticdomains, porphyroblastic garnet was replaced by a peak-metamorphic orthopyroxene^spinel^plagioclase inter-growth during prograde heating to temperatures of c.10008C. Similar UHTconditions of c. 10508C are calculatedby feldspar thermometry and Opx^Spr thermometry forthe peak-metamorphic sapphirineôrthopyroxene^feld-spar assemblage of the associated sapphirine granulites.These inferred ultrahigh temperatures are consistent withthe high Al2O3 contents of orthopyroxene in both rocktypes (up to 9·7wt %). Corresponding mid-crustalpressures of c. 7·5 kbar, determined by GrtÔpx^Pl^Qtzbarometry of the GrtÔpx^Spl^Crd gneisses, are in agree-ment with GASP pressures of 7·5�0·5 kbar, calculatedfor a spatially associated migmatitic Grt^Sil metapelite,and generally correlate well with the stability of thepeak-metamorphic assemblages as predicted by thepseudosections.

UHTdecompression

Post-peak decompression to pressures56 kbar at still ultra-high temperatures of c. 900-10008C is documented by vari-ous decompression textures, which frequently occur in allrock types investigated: (1) formation of cordierite^spinelreaction rims between porphyroblastic sapphirine andorthopyroxene of the sapphirine granulites; (2) growth ofcordierite as narrow reaction rims between orthopyroxene



and spinel in the migmatitic Grt^Opx-Spl^Crd gneisses;(3) formation of spinel^quartz symplectites at the expenseof garnet^sillimanite in the migmatitic metapelite; (4) for-mation of a cordierite corona between spinel and quartzin the migmatitic metapelite; (5) formation of a cordieritecorona between garnet, sillimanite, and quartz in the mig-matitic metapelite; (6) formation of a plagioclase coronaaround garnet (GASP reaction) in the migmatitic metape-lite.The interpretation of the post-peak path of the samplesheavily relies on cordierite-bearing symplectites and

corona textures. These textures are traditionally inter-preted in terms of decompression (e.g. Harley, 1998), asalso evidenced by the pseudosections of this study. Inmore recent years, however, it has been demonstrated thatcordierite corona structures may also form owing to an in-crease of the water activity, whereby cordierite becomesstable towards higher pressures (e.g. Harley & Thompson,2004; Kelsey et al., 2004; Baldwin et al., 2005). The gener-ally anhydrous peak-metamorphic and early retrogradeassemblages of our samples indicate low water activities

Fig. 15. Suggested P^T^t paths (grey, literature; black, this study) of metamorphism in high-grade metamorphic basement rocks intruded bythe Rogaland Complex at 920^930 Ma. Regional amphibolite-facies conditions of 6^8 kbar, 600^7008C are suggested for M1 metamorphismduring regional Sveconorwegian metamorphism at c. 1035^970 Ma (Jansen et al., 1985; Tomkins et al., 2005). Estimates for the maximum tem-peratures reached during later contact-related metamorphism M2 within the proposed contact aureole range from 7608C (Jansen et al., 1985)to49008C at the contact at low pressures of c. 5 kbar (Westphal et al., 2003). Ages of the metamorphic stages were determined by U^Pbzircon dating (i.e. Mo« ller et al., 2003; Tomkins et al., 2005; see text for discussion). Metamorphic zircon growth in sapphirine granulites of thisstudy occurred at 1006�4 Ma following prograde garnet breakdown during heating toT49008C. Sapphirine granulites and related rocks ofthis study are characterized by subsequent mid-crustal UHT metamorphism (c. 10008C, 7·5 kbar) followed by decompression to P55·5 kbarat 900^10008C. Growth of texturally late xenotime at 933�5 Ma is presumably related to cooling below the solidus and crystallization of leuco-some melt during M3.



during these stages of the metamorphic evolution. In add-ition, decompression to P55·5 kbar of these samples isalso evidenced by a number of cordierite-free,‘dry’ reactiontextures (i.e. quartz^spinel- and plagioclase-producing re-action textures in the metapelites and garnetôrthopyrox-ene regrowth in GrtÔpx^Crd^Spl gneisses), whoseformation is independent of the water activity.

Cooling

Subsequent cooling toT5750^8008C at c. 5 kbar is docu-mented by (1) the regrowth of biotite at the expense of thecordierite, spinel and orthopyroxene in the sapphirinegranulites and the GrtÔpx^Crd^Spl gneisses, reflectinginteractions between high-grade and early retrograde min-eral assemblages with crystallizing melt, (2) the late forma-tion of corundum at the expense of symplectitic spinel inthe sapphirine granulites, (3) the decomposition ofpeak-metamorphic Al-rich orthopyroxene into Al-poororthopyroxene^garnet symplectites in the GrtÔpx^Crd^Spl gneisses, (4) regrowth of garnet at the expense of cor-dierite in the migmatitic metapelite, and (5) regrowth ofgarnet as coronas around spinel in the migmatitic metape-lite. These predictions from the pseudosections are inagreement with P^T conditions of 760�308C and 3^5kbar as calculated by GrtÔpx Fe^Mg exchange andAl-in-Opx barometry of texturally late symplectitic assem-blages in the GrtÔpx^Spl^Crd gneisses. As a note of cau-tion, the cooling history of the sapphirine granulites andmigmatitic metapelites is not perfectly estimated by thepseudosections in Fig. 11, as these were calculated for areduced water content. Low water contents lead to a shiftof the biotite stability and solidus towards higher tempera-tures. Pseudosections for the comparably ‘dry’ GrtÔpx^Crd^Spl gneisses, on the other hand, were calculated forthe original water content but still document cooling toT58508C at pressures of c. 5 kbar. The late formation of agarnet corona around late biotite in the metapelite andregrowth of euhedral garnet at the expense of biotite inthe GrtÔpx^Crd^Spl gneisses provides possible evidencefor reheating of the samples. However, because of the lackof any significant zoning pattern in the newly formedgarnet a definite discrimination between prograde andretrograde garnet regrowth is difficult.

Timing of UHT metamorphismIn the sapphirine granulites, both oscillatory zoned,rounded zircon cores as well as broad, homogeneous,metamorphic zircon rims and single grains of meta-morphic zircon are observed. Concordia ages of1010�7Ma and 1006�4Ma demonstrate that meta-morphic zircon in the sapphirine granulites (both as over-growth rims on detrital zircon cores and as small singlegrains) formed during the regional Sveconorwegian meta-morphism between 1035 and 970Ma (Pasteels & Michot,1975; Bingen et al., 1993, 2008a, 2008b; Mo« ller et al., 2002,

2003; Tomkins et al., 2005). In contrast to Tomkins et al.(2005) and Mo« ller et al. (2002, 2003), we record no zircongrowth related to regional decompression after 970Ma orcontact metamorphism at around 930Ma (M2).Remarkably, metamorphic zircon in the sapphirine granu-lites, which is locally intergrown with xenotime, is charac-terized by high Ycontents of up to 2800 ppm and usuallylow Th/U ratios of 0·02^0·19, a feature also observed byMo« ller et al. (2003) for their metamorphic zircon grownduring M2 and M3, which also co-precipitated with xeno-time. According to Mo« ller et al. (2003) these highYcontentsobserved in their zircon are related to breakdown ofgarnet during M2. Even though garnet relicts are not pre-served in the sapphirine granulites, its previous existenceduring the prograde evolution of these samples is predictedby the calculated pseudosections (Fig. 11c). Therefore, for-mation of zircon and co-precipitating xenotime might berelated to prograde garnet breakdown during M1.Maximum temperatures of c. 840^9508C, attained duringmetamorphic zircon growth in the sapphirine granulitesat c. 1010 Ma, as calculated fromTi-in-zircon thermometry(Ferry & Watson, 2007), support the interpretation ofzircon formation during prograde garnet breakdown atHT to UHT conditions and are furthermore consistentwith constraints from the P^T pseudosections for the sap-phirine granulites (Fig. 11c), indicating garnet breakdownat temperatures 49008C. The fact that rims of meta-morphic zircon grown during M2 are not observed in oursamples does not necessarily mean that the sapphirinegranulites were not affected by this event. The generallack of zircon rims related to both post-M1 decompressionand contact metamorphism M2 in our samples may berelated to (1) the absence of peak-metamorphic garnet inthe sapphirine granulites releasing Zr during retrogradebreakdown, (2) the comparably low Zr contents of re-sorbed peak-metamorphic and early retrograde phases(510 ppm), which cannot provide enough Zr to producezircon during retrogression, and (3) the ‘dry’ mineralogyof the samples owing to previous extensive melt lossduring M1, hampering prograde dissolution^regrowth ofzircon as well as retrograde growth of zircon related tomelt crystallization during M2 (Kelsey et al., 2008).Evidence for M2 metamorphism of the sapphirine

granulites is restricted to the presence of late anhedralxenotime, epitactically grown on the metamorphic zirconand defining concordia ages of 928�10Ma and 933�5Ma. In garnet-absent samples such as the sapphirinegranulites, however, xenotime has a wide P^T stabilityrange and may persist throughout all metamorphicgrades (e.g. Bea & Montero, 1999; Pyle & Spear, 1999;Spear & Pyle, 2002). Nevertheless, we were unable to iden-tify an unambiguous source for the Y and P needed forxenotime crystallization, as the main phases of the sap-phirine granulites contain only minor Y,



generally515 ppm (except Opx with up to 75 ppm;Table 9). Most probably, xenotime formation at c. 930Main these samples is related to leucosome crystallizationduring cooling below the solidus. Late biotite crystalliza-tion in the sapphirine granulites implies introduction offluid, possibly leading to recrystallization of xenotime bya dissolution^reprecipitation process.In summary, our data indicate that garnet breakdown in

the sapphirine granulites during prograde heating toUHTconditions of49008C was responsible for the forma-tion of metamorphic zircon at 1006�4Ma during theSveconorwegian Orogeny M1 (Fig. 15). Consequently, theUHT metamorphism at c. 10008C,7·5 kbar and subsequentdecompression to pressures55·5 kbar at still high tempera-tures documented by our samples is probably related tothe same metamorphic event. This interpretation is sup-ported by the coincidence in the clockwise P evolution ofour samples and those of Degeling et al. (2001); the latterare unaffected by M2 metamorphism and yield post-M1 re-gional decompression ages of 955�8Ma (Degeling et al.,2001; Tomkins et al., 2005). Alternative formation of theUHTassemblages in the sapphirine granulites, GrtÔpx^Crd^Spl gneisses and Grt^Sil metapelites duringcontact-thermal M2 metamorphism cannot be definitelyexcluded, even though we find no petrological evidencefor a second UHT overprint of our samples. Subsequentpost-decompressional cooling into the amphibolite facies,on the other hand, most probably occurred during M2 toM3, judging from (1) xenotime formation at c. 930 Ma,probably related to leucosome crystallization during cool-ing, (2) similarities between the cooling-related reactiontextures of our samples and those observed by Westphalet al. (2003) in samples from the same area, which havebeen dated at 908�9Ma (Mo« ller et al., 2003), and (3) thecoincidence of pressures estimated for cooling of oursamples (i.e. c. 5 kbar) and those proposed for the anortho-site^norite-related contact metamorphism M2 and subse-quent cooling M3 (c. 3^5 kbar; Jansen et al., 1985; VanderAuwera & Longhi, 1994; Westphal, 1998; Westphal et al.,2003).

Links with regional metamorphic phasesOne of the main goals of this study was to link the variousstages of the metamorphic evolution of our samples to oneor more of the main metamorphic events established forthis area; that is: (1) the regional Sveconorwegian meta-morphism M1 bracketed between c. 1035 and 970 Ma;(2) contact metamorphism M2 related to intrusion of theanorthosite^mangerite^charnockite suite at c. 930 Ma;(3) retrograde metamorphism M3 associated with progres-sive re-equilibration of the basement rocks during coolingof the anorthosite^ mangerite^charnockite massif. To datethere is only limited information on the P^T paths experi-enced by the basement rocks from both inside and outsidethe osumilite-in isograd, although P^T estimates have

been made for the subsequent metamorphic stages M1,M2, and M3 (Jansen et al., 1985; Degeling et al., 2001),linked with growth episodes of zircon as determined fromin situ U^Pb dating (Degeling et al., 2001; Mo« ller et al.,2002, 2003; Tomkins et al., 2005).Based on phase equilibria for texturally early,

garnet-bearing, M1 assemblages in the migmatitic gneissessurrounding the Rogaland Complex, Jansen et al. (1985)concluded that the highest-grade conditions during theregional Sveconorwegian metamorphism M1 peaked atamphibolite-facies conditions of 600^7008C and 6^8 kbar.Similar conditions were proposed by Degeling et al. (2001)for peak metamorphism and prograde wet partial meltingof a migmatitic Grt^Sil metapelite exposed at the Opx-inisograd some c. 25 km from the Rogaland Complex andapparently unaffected by a later prograde overprintduring M2. In situ U^Pb dating combined with investiga-tion of the trace element composition of metamorphiczircon from this sample link metamorphic zircon growthat 1035�9Ma with incipient prograde migmatization ofthe metapelite (Tomkins et al., 2005). This interpretation issupported by U^Pb dating of metamorphic monazite in afelsic granulite from the same area, yielding ages of1032�5 and 990�8Ma (Bingen et al., 2008b). Recent de-tailed investigation of the peak-metamorphic conditionsand P^T paths experienced by the migmatitic GrtÔpxgneisses and Grt^Sil metapelites from the same area dem-onstrate that regional metamorphism M1 culminated atsignificantly higher overall temperatures of 900�1008C(GrtÔpx Fe^Mg equilibria, Al-in-Opx thermometry)than those previously proposed, whereas calculated pres-sures of 6�1 kbar (GrtÔpx^Pl^Qtz and GASP equili-bria) are comparable (Franke & Dru« ppel, 2007).Post-peak decompression following the M1 event andpre-dating anorthosite emplacement is evident fromcommon decompression textures (Degeling et al., 2001;Tomkins et al., 2005), with conditions of 5·6 kbar,7108C being estimated for the replacement of the peak-metamorphic garnet^sillimanite assemblage by cordierite(Tomkins et al., 2005). Zircon formed during this reactiondates the regional decompression at 955�8Ma (Degelinget al., 2001; Tomkins et al., 2005), hence pre-datinganorthosite^norite emplacement. Remarkably, similarpeak-metamorphic mid-crustal pressures of 7·5�0·5 kbarfollowed by post-peak decompression by about 2 kbar alsocharacterize the P evolution of the sapphirine granulitesand related rocks of this study, whereas the peak-metamorphic temperatures are at least 1308 higher thanthe granulite-faciesTcalculated for M1.For the subsequent contact metamorphism M2 induced

by the emplacement of the Rogaland Complex thermom-etry supports UHT peak-metamorphic conditions of c.9008C, calculated for osumilite-bearing gneisses in thevicinity of the igneous massif and our study area



(Westphal, 1998; Westphal et al., 2003). Low pressures of c.4^5 kbar are postulated for the emplacement depth ofthe Rogaland Complex and associated contactmetamorphism M2 based on barometric estimates forgarnetôrthopyroxene^plagioclase^sillimanite equilibriain osumilite-bearing gneisses (Jansen et al., 1985;Westphal,1998;Westphal et al., 2003) and experiments on the stabilityof osumilite in metapelites (Holland et al., 1996). Meltingexperiments on jotunite from the late-stage Bjerkreim^Sokndal layered intrusion support these relatively low M2

pressures of 55 kbar (Vander Auwera & Longhi, 1994),which are moreover consistent with the regional decom-pression to c. 5·6 kbar at 7108C postdating M1 andpre-dating anorthosite emplacement (Degeling et al., 2001;Tomkins et al., 2005). In situ U^Pb zircon dating of meta-morphic zircon grains or rims on inherited cores, over-grown by or intergrown with M2 minerals such asmagnetite, spinel and orthopyroxene, in migmatiticgneisses surrounding the Rogaland Complex yields agesof 927�7 Ma, hence directly linking UHT metamorph-ism with the intrusion of the anorthosite^norite massif(Mo« ller et al., 2002, 2003).Even though the UHT temperatures of c. 9008C previ-

ously proposed for M2 contact metamorphism are broadlyconsistent with those constrained for the sapphirine granu-lites and associated rocks of this study (i.e. c. 10008C), ourobserved P evolution with mid-crustal UHT metamorph-ism at c. 7·5 kbar being followed by decompression toP� 5·5 kbar is incompatible with the previously proposedupper crustal simple heating^cooling path at c. 5 kbar.The observed discrepancy between our pressure estimates(c. 7·5 kbar) and those proposed byWestphal et al. (2003)for GrtÔpx-bearing gneisses inside the osumilite-in iso-grad (c. 5 kbar) is probably related to the selection of refer-ence temperatures of only 700^8508C for the pressurecalculations of Westphal et al. (2003). Resetting of theFe^Mg system of garnetôrthopyroxene pairs used for thetemperature calculation of Westphal et al. (2003) is evidentfrom the partially extreme Al2O3 contents of orthopyrox-ene (up to 9·7wt %), which is more robust against retro-grade exchange and indicates temperatures in excess of9508C (using Al-in-Opx thermometry for P� 5 kbar), con-sistent with our data. Given these higherTestimates as ref-erence temperature, the resulting pressures are shifted tosignificantly higher, mid-crustal pressure values. Followingthis, the osumilite isograd could be an M1 isograd deflectedduring intrusion of the anorthosite plutons and associateddoming of the crust.A number of subsequent, retrograde, low-temperature,

low-pressure reaction textures are observed in the base-ment rocks in the vicinity of the igneous contacts and areinterpreted to overprint the high-grade M2 assemblages atP^Tconditions of c. 550^7008C, 3^5 kbar during slow cool-ing of the anorthosite^norite pluton from M2 to M3

(Maijer, 1987; Westphal, 1998; Wesphal et al., 2003). In situ

U^Pb dating of zircon overgrown by retrograde mineralassemblages in a migmatitic orthogneiss (i.e. zircon inclu-sions in coronitic garnet and in garnetôrthopyroxenesymplectites replacing orthopyroxene) yields an age of908�9Ma (Mo« ller et al., 2003), supporting this interpret-ation. The latter reaction textures strongly resemble thoseobserved in our migmatitic Grt^Sil metapelite and GrtÔpx^Crd^Spl gneisses (garnet regrowth around spineland cordierite, and decomposition of peak-metamorphichigh-Al orthopyroxene to low-Al orthopyroxene^garnetsymplectites), which occur in our samples at c. 3^5 kbar,700^8008C and are likewise interpreted to result from con-tinued cooling from UHTconditions into the amphibolitefacies.

Geodynamic implicationsThe sapphirine granulites and associated rocks of thisstudy are exposed at c. 7 km distance from the igneouscontact of the Rogaland anorthosite complex, within theproposed contact thermal aureole and next to theosumilite-in isograd. A typical regional metamorphicclockwise P^T path is deduced for the sapphirine granu-lites and the surrounding GrtÔpx^Crd^Spl gneisses andGrt^Sil metapelites studied here, with peak-UHT meta-morphism at 10008C and 7·5 kbar being followed bynear-isothermal decompression to55·5 kbar at still UHTconditions of 900^10008C and subsequent, near-isobariccooling to T5750^8508C. Syn-metamorphic ductiledeformation is common in the sapphirine granulites,marked by an alignment of the peak-metamorphic sap-phirine and retrograde biotite in the sapphirine granulites;this runs subparallel to the regional foliation and bandingdefined by the migmatitic GrtÔpx^Crd^Spl gneisses andGrt^Sil metapelites, as well as the surrounding migmatiticcharnockites. Our data demonstrate that temperatures ofat least 850^9408C were already reached during theSveconorwegian Orogeny M1 at c. 1010 Ma, recorded byzircon grown during prograde garnet breakdown in thesapphirine granulites.It remains uncertain whether the UHT peak meta-

morphism at c. 10008C, 7·5 kbar and early UHT decom-pression to P55·5 kbar at 900^10008C recorded by thesamples of the present study document continuous heatingto even higher temperatures during M1 and subsequentpre-M2 decompression or heating and decompressionduring contact-thermal M2 metamorphism.We prefer thefirst scenario as calculated regional P conditions attainedduring M1 (6^8 kbar) and the proposed post-peak decom-pression are almost identical to those of the sapphirinegranulites and associated rocks of this study. After decom-pression to P55·5 kbar at still UHT conditions of 900^10008C the sapphirine granulites and associated rockscould have formed the country-rocks for the intrudingRogaland Complex. In this case, elevated temperatures



still in excess of 9008C would have hindered the rocks fromattaining new low-pressure UHTassemblages during M2.Cooling of the samples below the solidus and crystalliza-tion of the leucosome, on the other hand, presumablyoccurred during emplacement and cooling of the anortho-site^norite massif at 930^908Ma (M2^M3) at low pres-sures of c. 5 kbar. Alternatively, the mid-crustal peakassemblages of the sapphirine granulites and their hostrocks and the post-peak decompression might haveformed during M2 metamorphism, with metamorphiczircon remaining completely unaffected. In that case, thelow pressures of c. 5 kbar postulated for the emplacementof the Rogaland Complex and associated contact-metamorphism (Jansen et al., 1985; Vander Auwera &Longhi, 1994; Westphal et al., 2003) require further atten-tion. Our interpretation that the UHT metamorphism re-corded by the studied rocks from the vicinity to theRogaland Complex is related to regional M1 metamorph-ism is consistent with the recent finding that ortho- andparagneisses exposed at 425 km distance from the an-orthosite complex at the Opx-in isograd, which are henceclearly unaffected by the contact^metamorphic M2 eventat 930^920 Ma, experienced medium-pressure near-UHTconditions at 900�1008C, 6�1 kbar (Franke & Dru« ppel,2007), indicating the widespread occurrence of granulite-facies rocks formed during regional Sveconorwegian meta-morphism. Our data suggest that UHT metamorphism inRogaland is not exclusive to M2 metamorphism during an-orthosite^norite emplacement but already occurredduring the regional Sveconorwegian metamorphism M1 atc. 1010 Ma, reaching temperatures4850^9408C (presum-ably c. 10008C). According to Bingen et al. (2006, 2008c)high-grade metamorphism during M1 is related to the col-lision between Fennoscandia and an unknown craton, pos-sibly Amazonia, at the end of the Mesoproterozoic,leading to formation of the Sveconorwegian orogenexposed over a total width of c. 500 km. This event wasmainly characterized by crustal thickening associatedwith tectonic imbrication and voluminous syn-collisionalgranite magmatism at c. 1050Ma (Bingen et al., 2008a),consistent with our constrained clockwise P^T paths,documenting granulite-facies metamorphism at mid-crustal levels, followed by decompression and cooling. Theextreme temperatures in excess of c. 10008C at mid-crustalpressures attained by our samples, however, imply a zoneof very high heat flow, which is difficult to explain by aconventional collisional model. Based on numerical model-ing, Clark et al. (2011) suggested that crustal thickening toform a wide and long-lived mountain plateau, which atthe same time displays high internal concentrations ofheat-producing elements to substantially raise the crustaltemperatures and low erosion rates, is the most likely geo-dynamic scenario to reach UHTconditions at mid-crustallevels. Preferential thickening of crust that has previously

undergone pre-heating in an extensional back-arc settingor mechanical heating in shear zones, on the other hand,contributes to elevated temperatures, but will not usuallylead to ultrahigh temperatures (Clark et al., 2011). In theRogaland Sector, a direct heat source, for example volu-minous, coeval, mafic plutons, is not observed.Accordingly, a collisional scenario is appealing, consistentwith our reconstructed clockwise P^T paths. The wholeRogaland area underwent high-grade regional meta-morphism from c. 1035 to 970Ma (Wielens et al., 1981;Bingen & van Breemen, 1998; Mo« ller et al., 2002, 2003;Bingen & Stein, 2003; Tomkins et al., 2005; Bingen et al.,2006, 2008a, 2008b) that peaked in granulite-facies condi-tions at c. 1010Ma (Degeling et al., 2001; Mo« ller et al., 2002,2003; Tomkins et al., 2005; Bingen et al., 2006, 2008b).Investigation of reaction textures combined with geochron-ology demonstrate that the high-grade lithologies of theRogaland^Vest Agder Sector resided at high-grade condi-tions for at least 60 Myr until c. 970 Ma, when decompres-sion commenced (Tomkins et al., 2005; Bingen et al., 2006,2008b). Following this, both the P^T conditions and dur-ation of high-grade metamorphism recorded by the crustalbasement of the Rogaland Sector, including our samples,are in agreement with its formation as a collision-related,long-lived, high-grade metamorphic mountain plateau. Inthis context, the retrograde decompression still at veryhigh temperatures, as recorded by our samples, is finallyassociated with melting of mafic lower crust giving rise tovoluminous anorthosite plutonism at 930^920 Ma; thiscan be interpreted as gravitational collapse of the moun-tain plateau.

CONCLUSIONSWe have documented a single-phase clockwise P^Tevolu-tion for regional MP^UHT granulite-facies MgAl-richsapphirine granulites and associated paragneisses of theRogaland Sector (South Norway), which are exposedclose to the 930^920Ma anorthosite^mangerite^charnock-ite suite of the Rogaland Complex. Our data, in combin-ation with U^Pb ages of zircon and xenotime measuredin thin sections, provide new insights into the crustal evolu-tion of the Rogaland Sector during latest Mesoproterozoicto early Neoproterozoic times, as follows.

(1) The metasedimentary protoliths of the sapphirinegranulites were deposited between 1220 and 1050Maand mainly incorporate zircon from early Mesopro-terozoic igneous rocks.

(2) The rocks record a clockwise P^Tevolution culminat-ing at UHT conditions of c. 10008C at a mid-crustallevel (c. 7·5 kbar). Retrograde decompression to pres-sures of c. 5 kbar, initially at prevailing UHTcondi-tions, was followed by near-isobaric cooling.



(3) In situ U^Pb zircon and xenotime age data combinedwith Ti-in-zircon thermometry indicate that regionalMPÛHT metamorphism occurred during regionalSveconorwegian metamorphism M1 at c. 1010 Ma,prior to the emplacement of the Rogaland Complex(930^920 Ma). In the sapphirine granulites, theformation of texturally late xenotime is related to theemplacement and cooling of the anorthosite^manger-ite^charnockite suite (M2^M3) at low pressures of c. 5kbar.

(4) The clockwise P^T path of the regional, c. 1010MaMPÛHT metamorphism is interpreted to be relatedto collisional tectonics during the early stages of theSveconorwegian Orogeny, followed by gravitationalcollapse of the mountain plateau.

ACKNOWLEDGEMENTSWe wish to thank Christa Zecha (TU Berlin) for carefulsample preparation, and Petra Marsiske (TU Berlin) forhelp with the XRFmeasurements. The assistance of OonaAppelt (GFZ Potsdam) and Francois Galbert (TU Berlin)during EMP analysis is greatly appreciated.We are grate-ful to Helene Bra« tz and Reiner Klemd (University ofErlangen) for their help and advice with the LA-ICP-MSmeasurements. We also wish to thank Astrid Kowitz(Humboldt University Berlin) for her companionshipduring the field investigations. The manuscript benefitedfrom discussions with Bernard Bingen, Mogens Marker,and Trond Slagstad (NGU Trondheim). Thoughtful com-ments on an earlier version of the paper by Chris Clarkand Andreas Mo« ller, and on the final version by BernardBingen, Chris Clark and Jacqueline Halpin helped toimprove the paper. The logistical support by theGeological Survey of Norway (NGU Trondheim), espe-cially the introduction to the regional geology of theRogaland area by Mogens Marker and Peter Ihlen, isgratefully acknowledged.

REFERENCESAndersen, T., Andresen, A. & Sylvester, A. G. (2002). Timing of late-

to post-tectonic Sveconorwegian granitic magmatism in SouthNorway. Norges geologiske unders�kelse Bulletin 440, 5^18.

Aranovich, L. Y. & Berman, R. G. (1997). A new garnetôrthopyrox-ene thermometer based on reversed Al2O3 solubility in FeOÂl2O3^SiO2 orthopyroxene. American Mineralogist 82, 345^353.

Baba, S. (2003). Two stages of sapphirine formation during progradeand retrograde metamorphism in the Palaeoproterozoic LewisianComplex in South Harris, NW Scotland. Journal of Petrology 44,329^354.

Baldwin, J. A., Powell, R., Brown, M., Moraes, R. & Fuck, R. A.(2005). Modelling of mineral equilibria in ultrahigh-temperaturemetamorphic rocks from the Ana¤ polisÎtauc� u Complex, centralBrazil. Journal of Metamorphic Geology 23, 511^532.

Bea, F. & Montero, P. (1999). Behavior of accessory phases and redis-tribution of Zr, REE, Y, Th, and U during metamorphism and

partial melting of metapelites in the lower crust. Studies in theKinzigite Formation of Ivrea Verbano, NW Italy. Geochimica et

Cosmochimica Acta 63, 1133^1154.Bertrand, P., Ouzegane, K. & Kienast, J. R. (1992). P^T^X relation-

ships in the Precambrian Al^Mg-rich granulites from In Ouzzal,Hoggar, Algeria. Journal of Metamorphic Geology 10, 17^31.

Bhattacharya, A., Krishnapur, K. R., Raith, M. & Sen, S. K. (1991).An improved set of a^X parameters for Fe^Mg^Ca garnets and re-finements of the orthopyroxene^garnet thermometer and theorthopyroxene^garnet^plagioclase^quartz barometer. Journal of

Petrology 32, 629^656.Bingen, B. & Stein, H. J. (2003). Molybdenite ReÔs dating of biotite

dehydration melting in the Rogaland high-temperature granulites,S Norway. Earth and Planetary Science Letters 208, 181^195.

Bingen, B. & Van Bremen, O. (1998). U^Pb monazite ages in am-phibolite- to granulite-facies orthogneiss reflect hydrous mineralbreakdown reactions: Sveconorwegian Province of SW Norway.Contributions to Mineralogy and Petrology 132, 336^353.

Bingen, B., Demaiffe, D., Hertogen, J., Weis, D. & Michot, J. (1993).K-rich calc-alkaline augen gneisses of Grenvillian age in SWNorway; mingling of mantle-derived and crustal components.Journal of Geology 101, 763^778.

Bingen, B., Stein, H. J., Bogaerts, M., Bolle, O. & Mansfeld, J. (2006).Molybdenite ReÔs dating constrains gravitational collapse of theSveconorwegian orogen, SW Scandinavia. Lithos 87, 328^356.

Bingen, B., Andersson, J., So« derlund, U. & Mo« ller, C. (2008a). TheMesoproterozoic in the Nordic countries. Episodes 31, 29^34.

Bingen, B., Davis, W. J., Hamilton, M. A., Engvik, A., Stein, H. J.,Sk�r, �. & Nordgulen, �. (2008b). Geochronology of high-grademetamorphism in the Sveconorwegian belt, S. Norway: U^Pb,Th^Pb and ReÔs data. NorwegianJournal of Geology 88, 13^42.

Bingen, B., Nordgulen, O. & Viola, G. (2008c). A four-phase model forthe Sveconorwegian orogeny, SW Scandinavia. NorwegianJournal ofGeology 88, 43^72.

Brandt, S.,Will, T. M. & Klemd, R. (2007). Magmatic loading in theProterozoic Epupa Complex, NW Namibia, as evidenced byultrahigh-temperature sapphirine-bearing orthopyroxene^silli-manite^quartz granulites. Precambrian Research 153, 143^178.

Brandt, S., Schenk, V., Raith, M., Appel, P., Gerdes, A. &Srikantappa, C. (2011). Late Neoproterozoic P^Tevolution of HPÛHT granulites from the Palni Hills (South India): new con-straints from phase diagram modelling, LA-ICPMS zircon datingand in-situ EMP monazite dating. Journal of Petrology 52, 1813^1856.

Carswell, D. A. & Harley, S. L. (1990). Mineral thermometry and bar-ometry. In: Carswell, D. A. (ed.) Eclogite Facies Rocks. Glasgow:Blackie, pp. 83^110.

Clark, C., Fitzsimons, I. C.W., Healy, D. & Harley, S. L. (2011). Howdoes the continental crust get really hot? Elements 7, 235^240.

Clifford, T. N., Stumpfl, E. F., Burger, A. J., McCarthy, T. S. &Rex, D. C. (1981). Mineral-chemical and isotopic studies ofNamaqualand granulites, South Africa: A Grenville analogue.Contributions to Mineralogy and Petrology 77, 225^250.

De Capitani, C. & Brown, T. H. (1987). The computation of chemicalequilibrium in complex systems containing non-ideal solution.Geochimica et Cosmochimica Acta 51, 2639^2652.

Degeling, H., Eggins, S. & Ellis, D. (2001). Zr budgets for meta-morphic reactions and the formation of zircon from garnet break-down. Mineralogical Magazine 65, 761^770.

Dill, H., Gerdes, A. & Weber, B. (2007). Cu^FeÛphosphate mineral-ization of the Hagendorf^Pleystein pegmatite province, Germany:With special reference to laser-ablation-inductive-coupled-plasmamass spectrometry (LA-ICP-MS) of iron-cored torbernite.Mineralogical Magazine 71, 423^439.



Dunkley, D. J., Clarke, G. L. & Harley, S. L. (1999). Diffusion meta-somatism in silica-undersaturated sapphirine-bearing granulitefrom Rumdoodle Peak, Framnes Mountains, east Antarctica.Contributions to Mineralogy and Petrology 134, 264^276.

Elkins, L. T. & Grove, T. L. (1990). Ternary feldspar experiments andthermodynamic models. American Mineralogist 75(5^6), 544^559.

Falkum, T. (1982). Geologisk kart over Norge, berggrunnskartMandal, 1: 250,000. Norges Geologiske Unders�kelse.

Ferry, J. M. & Watson, E. B. (2007). New thermodynamic models andrevised calibrations for the Ti-in-zircon and Zr-in-rutile thermo-meters. Contributions to Mineralogy and Petrology 154, 429^437.

Franke, H. & Dru« ppel, K. (2007). Geochemistry and petrology ofgranulite-facies rocks from Rogaland, SW Norway. Geochimica et

Cosmochimica Acta 71, Supplement Goldschmidt Meeting 2007, A293.Frei, D. & Gerdes, A. (2009). Accurate and precise in-situ zircon

U^Pb age dating with high spatial resolution and high samplethroughput by automated LA-SF-ICP-MS. Chemical Geology 261,261^270.

Gasquet, D., Bertrand, J. M., Paquette, J. L., Lehmann, J.,Ratzov, G., De Ascenc� a‹ o Guedes, R.,Tiepolo, M., Boullier, A. M.,Scaillet, S. & Nomade, S. (2010). Miocene to Messinian deform-ation and hydrothermalism in the Lauzie' re massif (FrenchWestern Alps): new U^Th^Pb and Ar ages. Bulletin de la Socie¤ te¤

Ge¤ ologique de France 181, 227^241.Gerdes, A. & Zeh, A. (2006). Combined U^Pb and Hf isotope LA-

(MC-)ICP-MS analyses of detrital zircons: Comparison withSHRIMP and new constraints for the provenance and age of anArmorican metasediment in Central Germany. Earth and PlanetaryScience Letters 249, 47^62.

Gerdes, A. & Zeh, A. (2009). Zircon formation versus zircon alter-ationçNew insights from combined U^Pb and Lu^Hf in-situ

LA-ICP-MS analyses of Archean zircons from the Limpopo Belt.Chemical Geology 261, 230^243.

Goncalves, P., Nicollet, C. & Montel, J. M. (2004). Petrology andin situ U^Th^Pb monazite geochronology of ultrahigh-temperature metamorphism from the Andriamena Mafic Unit,north^central Madagascar. Significance of a petrographical P^Tpath in a polymetamorphic context. Journal of Petrology 45,1923^1957.

Harley, S. L. (1998). On the occurrence and characterizationof ultrahigh-temperature crustal metamorphism. In: Treloar, P. J.& O’Brien, P. J. (eds) What Drives Metamorphism and

Metamorphic Reactions? Geological Society, London, Special Publications

138, 81^107.Harley, S. L. & Green, D. H. (1982). Garnetôrthopyroxene barom-

etry for granulites and peridotites. Nature 300, 697^701.Harley, S. L. & Thompson, P. (2004). The influence of cordierite on

melting and mineral^melt equilibria in ultra-high-temperaturemetamorphism. Transactions of the Royal Society of Edinburgh, Earth

Sciences 95, 87^98.Harley, S. L., Hensen, B. J. & Sheraton, J.W. (1990).Two-stage decom-

pression in orthopyroxene^sillimanite granulites from ForefingerPoint, Enderby Land, Antarctica: implications for the evolution ofthe Archean Napier Complex. Journal of Metamorphic Geology 8,591^613.

Herd, R. K.,Windley, B. F. & Ghisler, M. (1969). The mode of occur-rence and petrogenesis of the sapphirine-bearing and associatedrocks of west Greenland. Rapport Gr�nlands Geologiske Unders�gelse

24, 1^44.Hermans, G. A. E. M., Tobi, A. C., Poorter, R. P. E. & Maijer, C.

(1975). The high-grade metamorphic Precambrian of the SirdalÔrsdal area, Rogaland/Vest-Agder, S.W. Norway. Norges GeologiskeUnders�kelse, Bulletin 318, 51^74.

Hermans, G. A. E. M., Hagstege, A. L., Jansen, J. B. H. & Poorter, R.P. E. (1976). Sapphirine occurrence near Vikes� in Rogaland, SWNorway. Norsk GeologiskTidsskrift, Supplement 56(4), 397^412.

Holland,T. J. B. & Powell, R. (1998). An internally consistent data setfor phases of petrological interest. Journal of Metamorphic Geology 16,309^343.

Holland, T. J., Babu, E. V. & Waters, D. J. (1996). Phase relations ofosumilite and dehydration melting in pelitic rocks: a simplethermodynamic model for the KFMASH system. Contributions toMineralogy and Petrology 124, 383^394.

Horn, I., Hinton, R., Jackson, S. & Longerich, H. (1997). Ultra-traceelement analysis of NIST SRM 616 and 614 using laser ablationmicroprobe inductively coupled plasma mass spectrometry(LAM-ICP-MS): a comparison with secondary ion mass spectrom-etry (SIMS). Geostandards Newsletter 21, 191^203.

Huijsmans, J. P. P., Kabel, A. B. E.T. & Steenstra, S. E. (1981). On thestructure of a high-grade metamorphic Precambrian terrain inRogaland, south Norway. Norsk GeologiskTidsskrift 61, 183^192.

Jansen, J. B. H., Blok, R. J. P., Bos, A. & Scheelings, M. (1985).Geothermometry and geobarometry in Rogaland and preliminaryresults from the Bamble area, S Norway. In: Tobi, A. C. &Touret, J. L. R. (eds) The Deep Proterozoic Crust in the North Atlantic

Provinces. Dordrecht: D. Reidel, pp. 499^516.Kawasaki, T. & Sato, K. (2002). Experimental study of Fe^Mg ex-

change reaction between orthopyroxene and sapphirine and itscalibration as a geothermometer. Gondwana Research 5, 741^747.

Kelsey, D. E., White, R. W., Holland, T. J. B. & Powell, R. (2004).Calculated phase equilibria in K2O^FeO^MgOÂl2O3^SiO2^H2O for sapphirine^quartz-bearing mineral assemblages. Journalof Metamorphic Geology 22, 559^578.

Kelsey, D. E., Clark, C. & Hand, M. (2008). Thermobarometric mod-elling of zircon and monazite growth in melt-bearing systems:examples using model metapelitic and metapsammitic granulites.Journal of Metamorphic Geology 26, 199^212.

Koziol, A. M. & Newton, R. (1988). Redetermination of the anorth-ite breakdown reaction and improvement of the plagioclase^garnetÂl2SiO5^quartz geobarometer. American Mineralogist 73,216^233.

Kretz, R. (1983). Symbols for rock-forming minerals. American

Mineralogist 68, 277^279.Kriegsman, L. M. & Schumacher, J. C. (1999). Petrology of

sapphirine-bearing and associated granulites from central SriLanka. Journal of Petrology 40, 1211^1239.

Lee, H.Y. & Ganguly, J. (1988). Equilibrium compositions of coexist-ing garnet and orthopyroxene: experimental determinations in thesystem FeO^MgOÂl2O3^SiO2, and applications. Journal of

Petrology 29, 93^113.Maijer, C. (1987). The metamorphic envelope of the Rogaland intru-

sive complex. In: Maijer, C. & Padget, P. (eds) The Geology of

Southernmost Norway: a Geological Excursion Guide. Norges Geologiske

Unders�kelse, Special Publication 1, 68^72.Maijer, C. & Padget, P. (1987). The Geology of Southernmost Norway: a

Geological Excursion Guide. Norges Geologiske Unders�kelse, SpecialPublication 1.

Maijer, C., Andriessen, P. A. M., Hebeda, E. H., Jansen, J. B. H. &Verschure, R. H. (1981). Osumilite, an approximately 970Ma oldhigh-temperature index mineral of the granulite-facies meta-morphism in Rogaland, SW Norway. Geologie en Mijnbouw 60,262^272.

Martignole, J. & Martelat, J. E. (2003). Regional-scale Grenvillian-age UHT metamorphism in the Mollendo^Camana block (base-ment of the Peruvian Andes.). Journal of Metamorphic Geology 21,99^120.



Meier, F. M., Kolb, J., Skallaris, G. A. & Gerdes, A. (2006). New agesfrom the Mauritanides: Recognition of Archean IOCG mineraliza-tion at Guelb Moghrein, Mauritania.Terra Nova 18, 345^352.

Michot, P. (1960). La ge¤ ologie de la catazone: le proble' me desanorthosites, la palingene' se basique et la tectonique catazonaledans le Rogaland me¤ ridionale (Norve' ge me¤ ridionale). Norske

Geologiske Unders�kelse 212, 1^54.Millonig, L., Zeh, A., Gerdes, A. & Klemd, R. (2008). Neoarchean

high-grade metamorphism in the Central Zone of the LimpopoBelt (South Africa): Combined petrological and geochronologicalevidence from the Bulai Pluton. Lithos 103, 333^351.

Mo« ller, A., O’Brien, P. J., Kennedy, A. & Kro« ner, A. (2002).Polyphase zircon in ultrahigh-temperature granulites (Rogaland,SW Norway): constraints for Pb diffusion in zircon. Journal of

Metamorphic Geology 20, 727^740.Mo« ller, A., O’Brien, P. J. & Kro« ner, A. (2003). Linking growth epi-

sodes of zircon and metamorphic textures to zircon chemistry: Anexample for the ultrahigh-temperature granulites of Rogaland(SW Norway). In: Vance, D., Mu« ller, W. & Villa, I. M. (eds)Geochronology: Linking the Isotopic Record with Petrology and Textures.

Geological Society, London, Special Publications 220, 65^81.Newton, R. & Haselton, H. T. (1981). Thermodynamics of the garnet^

plagioclaseÂl2SiO5^quartz geobarometer. In: Newton, R.,Navrotsky, A. & Wood, B. J. (eds) Thermodynamics of Minerals and

Melts. NewYork: Springer, pp. 131^147.Newton, R. C. & Perkins, D. (1982). Thermodynamic calibration of

geobarometers based on the assemblages garnet^plagioclaseôrthopyroxene (clinopyroxene)^quartz. American Mineralogist 67,203^222.

Pasteels, P. & Michot, J. (1975). Geochronologic investigation of themetamorphic terrain of southwestern Norway. Norsk Geologisk

Tidsskrift 55, 111^134.Pasteels, P., Demaiffe, D. & Michot, J. (1979). U^Pb and Rb^Sr geo-

chronology of the eastern part of the south Rogaland igneous com-plex, southern Norway. Lithos 12, 199^208.

Pearce, N. J. G., Perkins, W. T., Westgate, J. A., Gorton, M. P.,Jackson, S. E., Neal, C. R. & Cherery, S. P. (1997). A compilationof new and published major and trace element data for NISTSRM 610 and NIST SRM 612 glass reference materials.Geostandards Newsletter 21, 115^144.

Powell, R. & Holland, T. J. B. (1988). An internally consistent datasetwith uncertainties and correlations: 3. Applications to geobarome-try, worked examples and a computer program. Journal of

Metamorphic Geology 6, 173^204.Pyle, J. M. & Spear, F. S. (1999).Yttrium zoning in garnet: coupling of

major and accessory phases during metamorphic reactions.Geological Materials Research 1, 1^49.

Sajeev, K., Osanai, Y. & Santosh, M. (2004). Ultrahigh-temperaturemetamorphism followed by two-stage decompression of garnetôrthopyroxene^sillimanite granulites from Ganguvarpatti,Madurai block, southern India. Contributions to Mineralogy and

Petrology 148, 29^46.Scha« rer, U., Wilmart, E. & Duschesne, J. (1996). The short duration

and anorogenic character of anorthosite magmatism: U^Pbdating of the Rogaland complex, Norway. Earth and Planetary

Science Letters 139, 335^350.Sheraton, J. W. (1980). Geochemistry of Precambrian metapelites

from East Antarctica: secular and metamorphic variations. Bureauof Mineral Resources Journal of Australian Geology and Geophysics 5,279^288.

Slama, J., Kos› ler, J., Condon, D. J., Crowley, J. L., Gerdes, A.,Hanchar, J. M., Horstwood, M. S. A., Morris, G. A., Nasdala, L.,Norberg, N., Schaltegger, U., Schoene, B., Tubrett, M. N. &

Whitehouse, M. J. (2008). Ples› ovice zirconça new natural refer-ence material for U^Pb and Hf isotopic microanalysis. ChemicalGeology 249, 1^35.

Spear, F. S. & Pyle, J. M. (2002). Apatite, monazite, and xenotime inmetamorphic rocks. In: Kohn, M. J., Rakovan, J. & Hughes, J. M.(eds) Phosphates: Geochemical, Geobiological, and Materials Importance.

Mineralogical Society of America and Geochemical Society, Reviews in

Mineralogy and Geochemistry 48, 293^335.Stacey, J. S. & Kramers, J. D. (1975). Approximation of terrestrial lead

isotope evolution by a two-stage model. Earth and Planetary Science

Letters 26, 207^221.Tobi, A. C., Hermans, G. A. E. M., Maijer, C. & Jansen, J. B. H.

(1985). Metamorphic zoning in the high-grade Proterozoic ofRogaland^Vest Agder, SW Norway. In: Tobi, A. C. & Touret, J. L.R. (eds) The Deep Proterozoic Crust in the North Atlantic Provinces.Dordrecht: D. Reidel, pp. 477^497.

Tomkins, H. S., Williams, I. S. & Ellis, D. J. (2005). In situ U^Pbdating of zircon formed from retrograde garnet breakdown duringdecompression in Rogaland, SW Norway. Journal of Metamorphic

Geology 23, 201^215.Vander Auwera, J. & Longhi, J. (1994). Experimental study of a jotu-

nite (hypersthene monzodiorite): constraints on the parentmagma composition and crystallization conditions (P, T, fO2) ofthe Bjerkreim^Sokndal layered intrusion (Norway). Contributions toMineralogy and Petrology 118, 60^78.

Verschure, R. H. (1985). Geochronological framework for thelate-Proterozoic evolution of the Baltic shield in SouthScandinavia. In: Tobi, A. C. & Touret, J. L. R. (eds) The Deep

Proterozoic Crust in the North Atlantic Provinces. Dordrecht: D. Reidel,pp. 381^410.

Versteeve, A. J. (1975). Isotope geochronology in the high-grade meta-morphic Precambrian of southwestern Norway. Norges Geologiske

Unders�kelse, Bulletin 318, 1^50.Vry, J. K. & Cartwright, I. (1994). Sapphirine^kornerupine rocks

from the Reynolds Range, central Australia: constraints on theuplift history of a Proterozoic low pressure terrain. Contributions toMineralogy and Petrology 116, 78^91.

Watson, E. B., Wark, D. A. & Thomas, J. B. (2006). Crystallizationthermometers for zircon and rutile. Contributions to Mineralogy and

Petrology 151, 413^433.Wen, S. & Nekvasil, H. (1994). SOLVCALC: An interactive graphics

program package for calculating the ternary feldspar solvus andfor two-feldspar geothermometry. Computers and Geosciences 20(6),1025^1040.

Westphal, M. (1998). Phasenpetrologie, Bildungsbedingungen undDruck^Temperatur Geschichte von Osumilith-fu« hrenden und ver-gesellschafteten Gesteinen in Rogaland, SW Norwegen. PhDthesis, University of Freiburg.

Westphal, M., Schumacher, J. & Boschert, S. (2003). High-temperature metamorphism and the role of magmatic heat sourcesat the Rogaland Anorthosite Complex in southwestern Norway.Journal of Petrology 44(6), 1145^1162.

White, R.W., Powell, R. & Holland, T. J. B. (2007). Progress relatingto calculation of partial melting equilibria for metapelites. Journalof Metamorphic Geology 25, 511^527.

Wielens, J. B., Andriessen, P. A. M., Boelrijk, N. A. M. & Hebeda, E.H. (1981). Isotope geochronology in the high-grade metamorphicPrecambrian of southwestern Norway: new data and reinterpret-ations. Norges Geologiske Unders�kelse, Bulletin 359, 1^30.

Zhou, X. Q., Bingen, B., Demaiffe, D., Lie¤ geois, J.-P., Hertogen, J.,Weis, D. & Michot, J. (1995). The 1160 Ma old Hidderskog meta-charnockite: implications of this A-type pluton for theSveconorwegian belt inVest Agder (SW Norway). Lithos 36, 51^66.



U^Pb LA-ICP-MS Geochronology and Pseudosecti

Documents

Transcript of U^Pb LA-ICP-MS Geochronology and Pseudosecti