Journal of Petrology 45(7) © Oxford University Press 2004; all rights reserved
Orthopyroxene–Corundum in Mg–Al-rich Granulites from the Oygarden Islands, East Antarctica
1 DIVISION OF GEOLOGY AND GEOPHYSICS, SCHOOL OF GEOSCIENCES, F05, UNIVERSITY OF SYDNEY, SYDNEY, N.S.W. 2006, AUSTRALIA
2 SCHOOL OF GEOSCIENCES, UNIVERSITY OF EDINBURGH, GRANT INSTITUTE, KINGS BUILDINGS, EDINBURGH EH9 3JW, UK
RECEIVED MARCH 4, 2003; ACCEPTED FEBRUARY 2, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
High-Mg–Al, silica-undersaturated metapelites from the Oygarden Group of islands, East Antarctica, preserve clear evidence for the stable coexistence of the assemblage orthopyroxene + corundum in natural rocks. The quartz-absent metapelite occurs as pods and isolated layers within a high-strain zone related to deformation during the c. 0·93 Ga Rayner Structural Episode. Assemblages that include orthopyroxene, corundum, sapphirine, sillimanite, cordierite, garnet and kornerupine are developed across a pre-existing compositional zoning, leading to contrasting mineral Fe–Mg ratios. The assemblage orthopyroxene–corundum is shown to exist in only a very restricted range of bulk compositions and P–T histories. Simplified qualitative FMAS grids have been constructed for kornerupine-absent and -present systems, illustrating MAS terminations and divariant equilibria that help to describe the mineral assemblage and reaction history. Reaction textures that include coronas of sapphirine and sillimanite separating orthopyroxene and corundum, and symplectites of orthopyroxene + sapphirine ± cordierite/plagioclase, orthopyroxene + sillimanite ± cordierite/plagioclase and orthopyroxene + sapphirine + sillimanite embaying garnet, imply a clockwise P–T–t evolution. Conditions of P > 9–10 kbar and T
800–850°C were attained prior to an initial phase of decompression that was accompanied by heating of up to 100°C. Peak temperatures of T 850–900°C were achieved at P 9 kbar followed by near-isothermal decompression to pressures of P 5 kbar. This clockwise isothermal decompression path contrasts markedly with anticlockwise isobaric cooling paths recorded elsewhere in the Rayner Complex, and reflects a second phase of orogenesis within the Rayner Structural Episode. KEY WORDS: decompression; kornerupine; orthopyroxene–corundum; reaction textures; sapphirine; symplectite
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Aluminous magnesian gneisses are commonly reported from high-grade metamorphic terranes. Rocks with such compositions are useful for the reconstruction of pressure–temperature–time (P–T–t) paths as the mineral assemblages they contain commonly preserve spectacular reaction textures that document segments of P–T–t paths (Droop & Bucher-Nurminen, 1984
; Windley et al., 1984; Waters, 1986; Droop, 1989). Because of their unusual composition relative to country rocks, these gneisses are also commonly subjected to partial alteration of their margins, thereby creating variations in bulk composition across individual layers. As the position of a particular multi-variant equilibrium in P–T space generally depends upon the compositions of the phases involved, an assemblage with given phase compositions may pass through a reaction earlier or later than the same assemblage containing phases with different compositions (e.g. Hensen, 1971). Hence, segments of a P–T path can be ascertained by comparing rocks with subtle differences in their bulk compositions, especially variations in Fe–Mg ratios, dependent only on their spatial position with respect to layer boundaries.
Silica-undersaturated metapelites with highly magnesian compositions have been reported from a number of localities in East Antarctica, including MacRobertson Land (Fig. 1; Sheraton et al., 1982; Dunkley et al., 1999), the Napier Complex (Motoyoshi et al., 1995), the Vestfold Hills (Harley, 1993), and the Rauer Group (Harley, 1998a). This paper describes a new locality, in the Oygarden Group of islands, East Antarctica, where silica-undersaturated metapelitic gneiss preserves assemblages that include orthopyroxene, corundum, sapphirine, kornerupine and garnet, developed within distinct zones across inherited compositional domains. Contrasting mineral compositions and subtly different reaction textures are preserved in the compositionally different zones that are inferred to have formed as a result of a prior metasomatic event. Coexisting orthopyroxene and corundum, which are predicted by theoretical petrogenetic grids to form a stable assemblage under restricted P–T conditions (e.g. Hensen, 1987), but have been loosely inferred to exist in nature (e.g. Windley et al., 1984; Bertrand et al., 1992; Goscombe, 1992; Kihle & Bucher-Nurminen, 1992), or isolated within extensively developed reaction textures (Ouzegane et al., 2003), are shown more clearly here to have reached textural equilibrium at near-peak P–T conditions. These rocks illustrate the importance of subtle variations in bulk composition on resulting reaction products and assemblages, and the preservation of reaction textures and phases. Garnet, sapphirine, orthopyroxene–corundum, and kornerupine-bearing parageneses provide ‘snapshots’ of a clockwise P–T–t evolution that is interpreted to have resulted from prograde thickening of the crust in Kemp Land during the Rayner Structural Episode and decompression from high pressures and temperatures. This P–T path contrasts markedly with the dominantly anticlockwise P–T–t paths dominated by apparent isobaric cooling, reported elsewhere in the Rayner Complex (Clarke et al., 1989; Fitzsimons & Thost, 1992; Thost & Hensen, 1992; Hand et al., 1994; Nichols, 1995; Boger & White, 2003).
|
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
The Oygarden Group of islands forms some of the most westerly exposures of the Proterozoic Rayner Complex in Kemp Land (Fig. 1a), which is interpreted to represent Archaean crust that was reworked by the Rayner Structural Episode during the earliest Neoproterozoic (Sheraton & Black, 1983
; Clarke, 1987; Sheraton et al., 1987; Grew et al., 1988; Kelly et al., 2002). Lying near the western extremity of this zone of reworking (Sheraton & Black, 1983; Sheraton et al., 1987), the Oygarden Group is considered to have formed part of the Napier Complex from at least c. 2·5 Ga, and was tectonically reworked at c. 0·93 Ga (Kelly et al., 2002). The Napier Complex, an Archaean ultrahigh-temperature metamorphic terrane, experienced multiple episodes of deformation and metamorphism between c. 2·98 Ga and c. 2·45 Ga (Sheraton et al., 1987; Harley & Black, 1997).
The Rayner Complex is a composite granulite-facies metamorphic terrane that comprises both reworked Archaean and Proterozoic crust. Orogeny between 1·0 and 0·9 Ga occurred as a result of convergence between an Indo-Napier craton (comprising the Napier Complex and peninsular India) and part of what now constitutes East Antarctica, during the Rayner Structural Episode (RSE: Sandiford & Wilson, 1984; Sheraton et al., 1987). The RSE is characterized by two main phases of deformation that are separated in time by the intrusion of voluminous charnockite in MacRobertson Land and the northern Prince Charles Mountains at 0·98–0·96 Ga (Young & Black, 1991; Kinny et al., 1997; Young et al., 1997; Zhao et al., 1997). The location of the Oygarden Group, which records only the second phase of this deformation (Kelly et al., 2000, 2002), makes this area critical to the understanding of the evolution of this region.
The Oygarden Group of islands was affected by at least five episodes of deformation (D1–D5) over a 1·8 Ga period during the Archaean and Proterozoic. The structures in the islands are dominated by those that formed during the RSE: D3 and D4. At least two deformation events or sets of structures are recognized to pre-date D3/D4. An S1 gneissosity, defined in part by a migmatitic layering in layered felsic orthogneiss, is identified with confidence only where cut by a homogeneous felsic orthogneiss that intruded during D1 at c. 2·75 Ga (Kelly et al., in preparation). Layered composite orthogneiss, the most abundant lithology, is composed of layers of felsic, intermediate and mafic gneiss that alternate on a centimetre to metre scale. Subordinate lithologies include mafic dykes, calc-silicate gneiss, layered quartz-rich metasediments, quartz-bearing metapelitic gneiss, and silica-undersaturated metapelitic gneiss. D2 resulted in transposition of S1 and the homogeneous felsic orthogneiss into S2 at c. 2·45 Ga (Kelly et al., in preparation).
The RSE D3 deformation is characterized by the effects of east-directed thrusting that occurred at medium- to high-P granulite-facies conditions at c. 0·93 Ga (Kelly et al., 2002). S3 assemblages in mafic granulite, including garnet, orthopyroxene, clinopyroxene, plagioclase and quartz, suggest that D3 occurred at P 9–10 kbar and T 800–850°C (Kelly et al., 2000). D4 resulted in a 2–3 km wide extensional shear zone that recrystallized rocks in the south of the island group. D4 occurred at granulite-facies conditions, similar to those that accompanied D3 (Kelly et al., 2000). The last event recognized, D5, is characterized by mylonites and ultramylonites that formed at amphibolite-facies conditions and cut structures that formed during the RSE. Mylonites show a broadly northward sense of transport.
|
OUTCROP DESCRIPTION AND ASSEMBLAGES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
High Mg–Al silica-undersaturated metapelite occurs as more than 15 distinguishable pods and lenses in an outcrop approximately 20 m x 10 m in size within a D4 high-strain zone (Figs 1b and 2a). The pods form part of a layered sequence that includes garnet-rich felsic and intermediate gneisses, quartz-rich gneiss, and also lenses of quartz-bearing pelitic and semi-pelitic gneiss. S4 layering envelopes the pods of silica-undersaturated metapelite, but S4 is not penetratively developed within the pods. Most pods preserve similar assemblage relationships, summarized using four pods (1, 2, 4 and 10) as examples that span the common compositional and textural variations observed (Fig. 1c–f; Table 1). Three main composition/assemblage types occur within the pods, with other subordinate compositions and assemblages occurring as layers bounding or enveloping the pods. These main types are: (1) garnet-poor, where garnet occurs only as fine-grained and volumetrically insignificant inclusions in sapphirine or sillimanite; (2) garnet-rich, where garnet is abundant in outer zones of pods; (3) kornerupine-bearing, which may or may not contain abundant garnet. Nearly all pods preserve an assemblage zoning pattern characterized by anhydrous cores dominated by coarse-grained orthopyroxene and sapphirine with a transition zone into 10–15 cm wide, phlogopite-rich schistose rinds. The anhydrous cores commonly preserve a foliation (S3) that is oblique to the enveloping S4 layering and foliation developed in schistose rinds (e.g. Fig. 1c). The orientation of S3 is not consistent between pods, which is interpreted to reflect variable rotation of pods during D4 deformation.
|
|
Garnet-poor (quartz-absent) assemblages
The core assemblages in garnet-poor pods (Pods 1 and 4; Table 1) are composed of S3 sapphirine, orthopyroxene and corundum, with or without minor sillimanite, cordierite and phlogopite. Rutile also occurs as part of S3, and is the only Fe–Ti oxide present in any zone. Garnet is low in abundance and occurs only as fine-grained inclusions (less than 0·5 mm; Fig. 2b) in elongate sapphirine grains (Pod 1), and is absent from the core of Pod 4. Sapphirine commonly occurs as elongate grains and clusters of grains up to 2 cm in length, and may also form elongate ‘cuneiform’ grains in orthopyroxene. In Pod 4, orthopyroxene may occur as large (up to 5–6 mm), elongate grains with lobate and eroded grain boundaries, and abundant fine-grained rutile inclusions. Sapphirine may form large xenoblasts that contain inclusions of rounded orthopyroxene. Corundum is scattered in this pod and is not confined to the core region. Where corundum occurs it is always separated from orthopyroxene by a thin (up to 0·3 mm), but complete, corona of sapphirine (Fig. 3a). Importantly, the grain size of sapphirine in the coronas is appreciably smaller than that of the orthopyroxene and corundum grains that they separate. Sillimanite may rarely form an outer, continuous or discontinuous rim between sapphirine and orthopyroxene (similar to Fig. 3b), and can appear to embay these minerals. Cordierite is rare, forming narrow rims between sillimanite and orthopyroxene and inclusions in orthopyroxene. Rare phlogopite occurs as inclusions in, or along grain boundaries of orthopyroxene.
|
A two-stage transition zone occurs between the core and schistose rind. The first stage involves a marked increase in sillimanite (Fig. 2c and d), which interfingers with orthopyroxene and sapphirine, and a minor increase in phlogopite content (Fig. 4). In some local domains orthopyroxene and sapphirine form rounded inclusions in larger sillimanite grains. The second stage is characterized by a marked increase in phlogopite and cordierite. Corundum is present throughout the transition zone, and is commonly separated from orthopyroxene by double coronas of sapphirine and sillimanite (Fig. 3b). A schistose rind up to 15 cm in width occurs on all pods (Fig. 1c–f), and in garnet-poor types is characterized by a smaller grain size and lower abundance of rutile compared with the core. Sillimanite-rich domains occur; however, cordierite-rich domains are predominant. Sapphirine and orthopyroxene both appear to be in overall textural equilibrium with other phases, but locally may be partially embayed by sillimanite or cordierite. Rarely, sapphirine may contain inclusions of phlogopite, orthopyroxene, corundum and garnet. Corundum is surrounded by coronas of sapphirine and either sillimanite or cordierite (Fig. 3c). Proximal to the margin of the schistose rind, phlogopite and sapphirine are foliated parallel to the pod margin and to S4 (Fig. 1b and c).
|
Garnet-rich (quartz-absent) assemblages
These assemblages (summarized here using Pod 10 as an example; Fig. 1f; Table 1) show similar assemblage and zoning patterns to other pods but are comparatively garnet-rich, contain plagioclase and have no corundum. Core assemblages have higher abundances of cordierite, sillimanite and phlogopite compared with other pods, which is interpreted to reflect more intense metasomatic alteration prior to the development of the assemblages. The core assemblages may preserve cordierite-poor domains that contain sub-idioblastic, tabular sapphirine grains that appear needle-like in cross-section. Cordierite-rich domains that surround cordierite-poor areas contain large poikiloblastic sapphirine (>2 mm) and sillimanite intergrown with, or containing inclusions of, orthopyroxene, cordierite and garnet. Cordierite occurs as rims on and interfingering with orthopyroxene, sapphirine and sillimanite. An outer core, with higher abundance of phlogopite and sillimanite, may envelop the main core, and commonly contains interstitial patches of plagioclase. The transition zone differs from other pods in its preservation of garnet porphyroblasts (up to 1 mm) and higher modal abundance of cordierite. Intergrowths of tabular sapphirine with orthopyroxene in this zone may be partially replaced by coarse-grained cordierite, which may also embay coarse-grained orthopyroxene, sapphirine and sillimanite. Garnet is present as inclusions in sapphirine, orthopyroxene and cordierite.
The schistose rind is relatively garnet-rich compared with other pods. However, garnet-poor compositional sub-domains do occur that are similar to schistose rinds in garnet-poor rocks. Garnet-rich domains contain coarse-grained garnet, orthopyroxene and sillimanite, with or without sapphirine. In one sample (OG581) a quartz vein (approximately 2 mm in width) cuts the quartz-absent rock, armoured from the latter by a 0·25–0·5 mm wide moat of orthopyroxene. A zone of coarse-grained plagioclase up to 3 mm in width separates this moat from coarse-grained garnet in the quartz-absent host, and is commonly rimmed by cordierite. Rare symplectites of orthopyroxene and cordierite may also occur adjacent to garnet grains. Approximately >5 mm from the quartz vein, garnet is surrounded and embayed by symplectites of orthopyroxene + sillimanite, with or without plagioclase and cordierite (Fig. 3d). In sub-domains 2–3 cm away from the large garnet grains described above, the assemblage is composed of a coarse-grained intergrowth of sillimanite and orthopyroxene, and medium-grained garnet.
Kornerupine-bearing (quartz-absent) assemblages
Kornerupine has been identified in a single pod (Pod 2) within the suite (Fig. 1d). This pod preserves core and transition assemblages similar to those in garnet-poor rocks. However, the schistose rind has higher garnet abundance compared with garnet-poor assemblages described above. All zones are kornerupine-bearing. Kornerupine occurs in the core assemblage as elongate grains, generally <1 mm in length, that define S3 (Fig. 2e). Kornerupine may contain inclusions of garnet and corundum, be overgrown by orthopyroxene, and occur as inclusions in S3 sapphirine. Corundum inclusions are separated from kornerupine and orthopyroxene by coronas of sapphirine. Garnet occurs as fine-grained inclusions in sapphirine, whereas phlogopite may contain inclusions of sapphirine. In the schistose rind sapphirine, sillimanite, phlogopite and kornerupine surround coarse-grained (up to 2 mm) garnet, which may be embayed by kornerupine and orthopyroxene intergrowths (Fig. 2f). Garnet may also occur as coarse-grained inclusions in sapphirine, sillimanite and orthopyroxene. Sapphirine may contain inclusions of kornerupine. Adjacent to the core, the rind assemblage is aligned parallel to S3, but closer to the pod margin, sapphirine, sillimanite, phlogopite and kornerupine are parallel to S4.
Other quartz-absent assemblages
Present in the outcrop are compositionally distinct layers that may occur adjacent to the schistose rind on a single side of a pod (e.g. Pods 4 and 10), on both sides (Pod 1) or completely surrounding the pod (Pod 2; Fig. 1). In Pod 1 (Fig. 1c), the straight boundary between the schistose rind and a sillimanite-rich gneiss is parallel to S3 preserved in the core of the pod, and oblique to the enveloping S4. It is possible that these compositional layers reflect a primary compositional layering or are the result of metasomatic alteration of a more homogeneous protolith prior to D4. Sillimanite-rich gneiss (Pods 1 and 4) contains an assemblage composed of randomly oriented, coarse-grained orthopyroxene, sillimanite, phlogopite, cordierite and plagioclase, with minor garnet. Patchy phlogopite and sillimanite may define S4 close to the pod margin, but away from the pod margin sillimanite may occur as radiating aggregates. Sillimanite may be separated from orthopyroxene by medium- to coarse-grained rims of cordierite, which contain inclusions of phlogopite, sillimanite and orthopyroxene. In Pod 4 the sillimanite-rich gneiss preserves spectacular ‘cuneiform’ intergrowths of sillimanite and sapphirine in orthopyroxene (Fig. 3e). Garnet occurs as inclusions in sapphirine and sillimanite. Sillimanite–sapphirine gneiss, occurring as a distinct layer or outer rind (Pods 1, 2, 10; Table 1), is composed of coarse-grained sillimanite, sapphirine, orthopyroxene, phlogopite and garnet. In Pod 10, sillimanite-absent domains occur where large garnet porphyroblasts (2–4 mm) are embayed by symplectites of orthopyroxene and random, ‘spinifex-like’ arrays of tabular sapphirine, with or without plagioclase (Fig. 3f). Smaller garnet grains may also occur with coarse-grained orthopyroxene, sapphirine and plagioclase, and as inclusions in elongate sapphirine grains. Phlogopite is rare.
Quartz-bearing metapelite
Quartz-bearing metapelitic gneiss associated with quartz-absent rocks commonly preserves similar assemblages and reaction textures. The rocks predominantly preserve a coarse-grained S4 assemblage of orthopyroxene, garnet and sillimanite, with medium-grained biotite, plagioclase, quartz and rutile, with or without K-feldspar. In more leucocratic samples, S4 is defined by alternating garnet–orthopyroxene-rich and quartzo-feldspathic domains. In some samples coarse-grained plagioclase (2–3 mm) occurs in leucosomes and contains fine-grained vermicular inclusions of quartz and K-feldspar. In leucocratic domains, cordierite occurs as large grains, but also forms part of symplectites with orthopyroxene that separate garnet and biotite, or biotite and quartz.
|
MINERAL CHEMISTRY OF QUARTZ-ABSENT METAPELITE |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Electron microprobe point analyses and X-ray intensity maps were collected using the Cameca Camebax SX-50 electron microprobe housed at the Electron Microscope Unit, University of New South Wales, operating with an accelerating voltage of 15 kV, beam current of 20 nA, beam width of 1–5 µm and PAP data reduction software supplied by the manufacturer and described by Pouchou & Pichoir (1984)
. X-ray intensity maps were collected with a 1–3 µm beam size and count times of 300 ms at each point. X-ray intensity data (Figs 3 and 6) were reprocessed to display weight percent oxide values and molecular proportions following the method described by Clarke et al. (2001), which uses the computer software Matlab incorporating a matrix correction algorithm based on the empirical 
-factor approach of Bence & Albee (1968). Additional samples were analysed at the Department of Geology and Geophysics, University of Edinburgh, using a Cameca Camebax Microbeam electron microprobe operating at an accelerating voltage of 20 kV and beam current of 25 nA. No systematic differences were detected in compositions between minerals analysed in Edinburgh and Sydney. Microprobe analyses representing the range of mineral compositions are presented in Table 2. Back-scattered electron images (Fig. 2) were collected at the University of Edinburgh, using a Phillips XL30 scanning electron microscope operating at an accelerating voltage of 20 kV. In addition to analysis by electron microprobe, kornerupine was analysed by secondary ion mass spectrometry (SIMS) using the Cameca IMS-4f at the Department of Geology and Geophysics, University of Edinburgh, following methods outlined in detail by Hinton (1995). A primary 16O– beam at 8 nA was accelerated through 10 kV and positive secondary ions drawn at a 75 V offset. 1H, 7Li, 9Be and 11B, and a suite of other major and trace elements (Table 3), were analysed as isotopic ratios referenced against 28Si, and ppm concentrations of elements calculated by reference to ion yields determined using the NIST (SRM610) glass standard.
|
|
|
Across each pod the XMg values [= Mg/(Mg + Fe)] preserved by all ferromagnesian minerals gradually decrease from core domains to the schistose rinds (Fig. 5; Table 1), but are variable in sillimanite-rich gneiss and sillimanite–sapphirine gneiss. This decrease in mineral XMg correlates with an increase in phlogopite abundance in the rinds. Relative XMg values [Mg/(Mg + Fetot)] between minerals occur in the same order in nearly all pods and zones, with cordierite > sapphirine > phlogopite > orthopyroxene > garnet. When considering Fe3+,
of sapphirine may in some cases be equal to that of cordierite, and in rare circumstances phlogopite may be slightly more magnesian than sapphirine. Where present, kornerupine XMg may be less than sapphirine (Pod 2 core zone) or intermediate between sapphirine and cordierite values. However, when compositions are recalculated to consider Fe3+, kornerupine may become more magnesian than sapphirine (Pod 2 core zone) or marginally less magnesian (Pod 2 schistose rind).
|
Sapphirine
Compositional zoning of sapphirine is characterized by a systematic decrease in XMg [= Mg/(Mg + Fetot)] in grains with proximity to the edge of pods. Pod 1 core compositions have XMg
0·93–0·94, and decrease to XMg 0·90–0·92 in the schistose rinds (Table 1). Pod 2 sapphirine has slightly lower XMg values (0·91–0·89). Sapphirine from the sillimanite-rich and sillimanite–sapphirine gneiss also commonly has lower XMg compared with the schistose rinds. Sapphirine in all pods and zones has between 4·5 and 4·3 Al cations (per 10 oxygens) and shows a rimward increase in Al and decrease in Si. Sapphirine rims adjacent to corundum or sillimanite are commonly more enriched in Al when compared with rims adjacent to orthopyroxene, garnet or biotite. Compositions approach a stoichiometric ratio of 7:9:3 (Mg:Al:Si; where Al = 4·5, Si = 0·75 cations p.f.u.), with cores (higher Si) trending approximately towards 33% of a 2:2:1 composition (Al = 4, Si = 1 cations p.f.u.; Fig. 7a). Analysed compositions lie off the ideal Tschermak's substitution line (AlAlSi–1Mg–1) indicating a proportion of Fe3+ in total Fe. Fe3+ concentration, calculated by assuming 14 cations per 20 oxygens and using the structural formula M7(M)O2[T6O18] (after Moore, 1969), where Fe3+ = AlVI – AlIV [AlVI = 6 – (Si + Ti); AlIV = Altot – AlVI], ranges between 0·05 and 0·18 cations p.f.u. (average 0·10 ± 0·3). These values show no consistent trend within or between pods, or in relation to the Tschermak's substitution. Fe3+/Fetotal ranges between 0·14 and 0·48, but averages 0·35 ± 0·04. By taking Fe3+ into account XMg values increase by between 0·02 and 0·05 units (Tables 1 and 2a).
Orthopyroxene
Individual grains are rarely zoned with respect to XMg and values within domains are relatively uniform. However, rare grains preserve weak increases in XMg at rims adjacent to garnet and sapphirine. Orthopyroxene XMg compositions typically decrease from core domains (0·89–0·91) to schistose rinds (0·84–0·87; Pod 10 0·81). Sillimanite-rich gneiss and sapphirine–sillimanite gneiss orthopyroxenes have XMg 0·84–0·86, which is similar to, or lower than the schistose rinds, with the exception of Pod 10 (Table 1). Individual orthopyroxene porphyroblasts show a consistent rimward decrease in Al2O3 from a broad core plateau that is high in Al2O3 (up to 9·5 wt %) to narrow (<35 µm), comparatively Al2O3-poor rims (>3 wt %; Fig. 6a). Zoning in Al2O3 is most marked when orthopyroxene is adjacent to cordierite or plagioclase, and in some cases significant rimward depletions in Al2O3 are observed in orthopyroxene adjacent to sapphirine (Fig. 6b) and phlogopite. Fe3+ has been calculated for orthopyroxene from charge balance based on four cations and six oxygens; contents are minor (0·030 ± 0·02 cations p.f.u.) and show no systematic variation or zoning features. Fe3+ in Fetotal averages 0·11 ± 0·08, increasing XMg by generally less than 0·03 units, but is higher in some cases (Tables 1 and 2a).
|
|
) vs higher-Si cores (
). (b) Compositional traverse across a garnet porphyroblast from the rind of Pod 2. Garnets typically show a decrease in XPyp mirrored by an increase in XGrs and XAlm + XSps adjacent to rims. The garnet is approximately 0·5 mm in diameter. (c) (MgO + FeO)–SiO2–Al2O3 ternary plot for Pod 2 (Krn-bearing); it should be noted that kornerupine in this plot is uncorrected for Fe3+ in total Fe, and therefore plots just below the Opx–Crn tie-line. (d) (MgO + FeO)–SiO2–Al2O3 ternary plot showing chemographic relationships between phases from Pod 1 (Krn-absent). (e) Partial ternary diagram illustrating the position of the minerals Crd, Sil, Crn, Krn, Spr and Grt projected from Opx onto the SiO2–FeAl2O4–MgAl2O4 (quartz–hercynite–spinel) plane, using mineral compositions corrected for Fe3+. The minor deviation of Krn to the magnesian side of the Opx–Spr–Crd plane should be noted.Biotite
All biotite is close to phlogopite in composition [classified after Deer et al. (1992, p. 284)
]. Consistent with other ferromagnesian minerals, XMg of phlogopite decreases systematically from core zones (0·93–0·90) to schistose rinds (0·85–0·88), whereas phlogopites from sillimanite-rich and sapphirine–sillimanite gneiss are similar in composition or slightly less magnesian (XMg 0·85–0·89). Total F in phlogopite is commonly between 1·2 and 1·5 wt %, with no systematic variation between zones or pods. As F rarely accounts for more than 2% of the total percentage of elements in the phase, up to 6 wt % of total oxides is unaccounted for in the analysis. This is inferred to be H2O or Cl in the hydroxyl site, suggestive of a relatively hydrous phlogopite.
Garnet
General grain compositions are variable within and between pods, although predominantly following the Mg zoning patterns already described for other minerals. The most magnesian garnet occurs as inclusions in sapphirine in the core zone of Pod 10 [XPyp = 0·74–0·77, where XPyp = Mg/(Mg + Fetot + Ca + Mn)] and Pod 1 (XPyp = 0·67–0·70). In general, XGrs [Ca/(Mg + Fetot + Ca + Mn)] = 0·02–0·03. Most garnets display a systematic decrease in XPyp between the core and schistose rind (Table 1), with minor increases in XGrs and XAlm [Fetot/(Mg + Fetot + Ca + Mn)]. Sillimanite-rich and sillimanite–sapphirine gneiss typically have garnet compositions similar to, or slightly more magnesian than schistose rind garnet. Garnet porphyroblasts, which are preserved only in Pods 2 and 10, commonly preserve a rimward decrease in XPyp, with a broad core compositional plateau and narrow zoned rim (e.g. Fig. 7b). The core composition of some garnet grains in the schistose rinds may be as high as XPyp = 0·67 (Pod 2), similar to the composition of garnet inclusions in sapphirine from core domains. The composition of the rims of these grains is equivalent to that of smaller garnet porphyroblasts and inclusions in sapphirine in the schistose rind domains.
Kornerupine
Electron microprobe analyses indicate XMg 0·89–0·91 (where Fe2+ = Fetot) and a (Mg,Fe)O:Al2O3:SiO2 ratio that is close to 11:10:11, such that kornerupine lies to the Si-poor side of the Crn–Opx tie-line in MAS (Fig. 7c). However, considering the Fe3+ content in Fetotal [after Grew et al. (1999): Fe3+ in kornerupine is inferred to be equal to that estimated for associated sapphirine], approximately 35% in the Oygarden samples, increases XMg to 0·93–0·94. These values are similar to XMg values previously recorded for kornerupine (Droop, 1989; Goscombe, 1992; Vry, 1994; Carson et al., 1995; Friend, 1995; Grew, 1996). In addition to a change in XMg, removing Fe3+ from Fetotal shifts the position of kornerupine to being effectively collinear with orthopyroxene–corundum in MAS. Total Al cations p.f.u. (for 21·5 oxygens) are approximately 6·9 and AlIV [=5 – (Si + B)] is calculated to lie between 1·06 and 1·16. The Oygarden kornerupines are ‘boron-poor’ examples (after Friend, 1995), with nine SIMS analyses yielding an average of 0·94 wt % B2O3, 0·48 wt % F, 0·85 wt % H2O, 0·05 wt % Na2O, 91 ppm Be and 314 ppm Cl (see Table 3 for full analyses).
Other phases
In cordierite (where present), XMg varies from cores (0·96–0·97) to schistose rinds (0·94–0·95), consistent with the XMg trends recorded by the other minerals (Table 1). Cordierite in sillimanite-rich gneiss commonly has a similar composition to that in the schistose rind (XMg = 0·94–0·95). Plagioclase occurs in a number of zones in Pod 10, and only in sillimanite-rich gneiss in Pod 1, and has a variable composition with XAn [=Ca/[Ca + Na + K)] between 0·30 and 0·55. Calcium contents are higher in samples from Pod 10 (average XAn 0·50) and individual grains commonly preserve a rimward increase in XAn. Zoning is especially pronounced where plagioclase is adjacent to garnet. Sillimanite is near pure Al2SiO5, with Fe being the only impurity (<0·3 wt % Fe2O3, measured as FeO). Fe values show no systematic variation when in association with sapphirine. Corundum contains less than 0·15 wt % Fe2O3 and less than 0·3 wt % Cr2O3.
|
MINERAL CHEMISTRY OF QUARTZ-PRESENT METAPELITE |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Garnet from quartz-bearing metapelitic gneiss is a pyrope–almandine mix (XAlm = 0·35–0·46, and XPyp = 0·51–0·62) with low grossular contents (XGrs = 0·02–0·03). Large garnet grains (>2 mm in diameter) show a subtle rimward increase in Fe and a decrease in Mg content, commonly by approximately XAlm = 0·03. Orthopyroxene has XMg = 0·75–0·83 [Mg/(Mg + Fetot)] with alumina contents varying between 0·37 and 0·11 cations per six oxygens, and shows a consistent rimward decrease in porphyroblastic orthopyroxene of up to 0·17 cations. XMg of cordierite ranges between 0·90 and 0·94 and rarely preserves zoning. However, some symplectic cordierite has higher Mg contents than grains adjacent to biotite or garnet. Plagioclase is oligoclase in composition, with XAn = 0·18–0·32 and XAb = 0·80–0·66. Where plagioclase occurs as exsolution lamellae in perthite, it has the same composition as larger matrix grains. Alkali feldspar is sanidine to orthoclase in composition, with XOr = 0·73–0·92 and XAb = 0·27–0·08. Rare analyses taken from exsolution lamellae in K-feldspar have higher albite components (XAb
0·87). Biotite compositions range from phlogopite to eastonite, with XMg of between 0·79 and 0·91. Sillimanite generally has less than 0·5 wt % Fe2O3 (analysed as FeO). |
BULK-ROCK GEOCHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Bulk-rock composition data were obtained using standard X-ray fluorescence (XRF) techniques at the University of New South Wales and are available in Supplementary Data Table 1, which may be downloaded from the Journal of Petrology website at http://www.petrology.oupjournals.org/. The components MgO–FeO–SiO2–Al2O3 make up 97–98% of core samples and 94–97% of schistose rind samples, with the remainder being mostly K2O. Compositional relationships are therefore easily compared with analysed mineral compositions using a MAS ternary diagram (Fig. 8). The whole-rock analyses confirm that the rocks are undersaturated with respect to silica (38–47%), are highly magnesian (XMg = 0·85–0·93), and have core compositions characteristically low in CaO and Na2O. Data indicate systematic K enrichment in the transition zones and schistose rinds of the pods, an increase that is reflected in the increased modal abundance of phlogopite. In addition, most pods also show a minor, although systematic, enrichment in Fe and Si in the rinds. No major elements exhibit consistent depletion patterns; however, an apparent decrease in Mg was observed in data from some pods. Ti is commonly depleted in transition zones and enriched in the schistose rind correlating with a decrease in rutile abundance and grain size with increasing distance from the core. This may also indicate mobility of Ti from the core into the schistose rind where it has been incorporated in phlogopite. The presence of plagioclase with a significant albite component (XAb = 0·7 in Pod 1 outer zones; XAb = 0·45–0·69 in Pod 10; Table 1) suggests that a proportion of Na was present. There are no consistent zoning patterns with respect to Na, although some pods do have a rimward increase (Pods 1 and 10), therefore Na metasomatism accompanying silica cannot be confirmed.
|
On a MAS compatibility diagram, most bulk-rock compositions plot within the orthopyroxene–sillimanite–sapphirine tie-triangle and commonly on the silica-poor side of the orthopyroxene–corundum tie-line (Fig. 8). The compositions plot between garnet and phlogopite, and close to orthopyroxene, which correlates with the high proportion of orthopyroxene within core assemblages and abundance of phlogopite in rind assemblages. Exceptions to this include the quartz-bearing country rock from Pods 1 and 12, and samples OG577 (Pod 4) and OG581 (Pod 10). Sample OG577 is from sillimanite-rich gneiss that envelopes Pod 4 and its composition plots close to the orthopyroxene–sillimanite tie-line. Sample OG581 (schistose rind) plots above the orthopyroxene–sillimanite tie-line within the stability field orthopyroxene–sillimanite–cordierite, which corresponds to the observed textures and the absence of sapphirine in the assemblage.
|
THERMOBAROMETRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
The majority of quartz-absent assemblages described here are not appropriate for conventional thermobarometry. As a result, P–T estimates were made from garnet–orthopyroxene-bearing S3 and S4 assemblages from quartz-bearing metapelite that formed layers within the same outcrop as the silica-undersaturated metapelite. Average P and T calculations using the coarse-grained assemblage garnet, orthopyroxene, plagioclase, sillimanite, biotite, quartz and cordierite (sample OG562) using THERMOCALC v.3.0 [Powell & Holland (1988)
, based on the internally consistent dataset of Holland & Powell (1998)], gave an average P = 9·9 ± 0·3 kbar (for T = 850°C; aH2O = 1·0; 1
) and an average T = 844 ± 27°C (for P = 10 kbar; 1). Using aH2O = 0·5 reduced pressures by 1 kbar but raised temperatures by 50–60°C, whereas using aH2O = 0·1 reduced pressures by only 0·5 kbar and raised temperatures by 140°C. Using the ‘average P–T’ function on the same assemblage gave P = 10·1 ± 0·6 kbar and T = 875 ± 65°C (1). Excluding cordierite from this assemblage gave slightly lower pressure and similar temperature estimates: average P–T of P = 8·7 ± 1·3 kbar and T = 876 ± 66°C (1). A second sample (OG589) gave similar, although poorly constrained, results for the S4 assemblage garnet, orthopyroxene, plagioclase, biotite, cordierite and quartz, with an average P–T estimate of P = 8·4 ± 2·1 kbar and T = 819 ± 156°C (1).
Temperature estimates for suitable quartz-absent samples where garnet and orthopyroxene are in textural contact and interpreted to be in chemical equilibrium, or inferred to have originally been in textural contact, have also been obtained using garnet–orthopyroxene equilibria (Fe–Mg: Ganguly et al., 1996; Mg–Al: Harley & Green, 1982; Fe–Al: Aranovitz & Berman, 1997). A 10 kbar pressure estimate, obtained above, was used. Adjacent rim compositions of garnet and orthopyroxene produced estimates of T 810–910°C (for P = 10 kbar; Fe–Mg calibration) and T 850–960°C (for P = 10 kbar; Mg–Al calibration). The Fe–Al calibration was consistently lower than the Fe–Mg and Mg–Al calibrations and gave estimates of T 740–845°C (for P = 10 kbar). To investigate the effects of post-peak re-equilibration of Fe–Mg between orthopyroxene and garnet, the approach of Fitzsimons & Harley (1994) was used to ‘back-calculate’ the peak composition of garnet–orthopyroxene pairs. This technique produced internally consistent (convergent) estimates generally in the range T 910–960°C. Results obtained using garnet–orthopyroxene thermobarometers alone are only indicative of the general metamorphic conditions, and do not provide tight constraints on the P–T evolution. However, these results are consistent with estimates from THERMOCALC and those obtained using conventional garnet–orthopyroxene–plagioclase–quartz thermobarometry (see Supplementary Data Tables 2 and 3, which may be downloaded from the Journal of Petrology website). In all, the application of thermobarometry suggests peak or near-peak P–T conditions of P 9–10 kbar and T 850–950°C.
|
REACTION HISTORY AND PHASE RELATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Reaction textures observed in the orthopyroxene–sapphirine granulites occur in all compositional zones. This implies that the reaction textures post-date the development of the compositional zones and hence were not a direct result of interaction with a metasomatic fluid. Instead, the textures are interpreted to have been formed as a consequence of changes in ambient pressure–temperature conditions. For the purposes of describing the phase relations, the minerals orthopyroxene–corundum–sapphirine–sillimanite–cordierite–garnet–spinel have been investigated in the simplified FeO–MgO–Al2O3–SiO2 system (FMAS). By comparing the assemblages and reaction textures observed with existing petrogenetic grids (Waters, 1986
; Hensen, 1987), a qualitative partial P–T grid involving FMAS univariant reactions has been constructed (Fig. 9), and relevant FMAS divariant equilibria within this system have been modelled in terms of their MAS end-member analogues. The new grid observes known 
Vr contraints and restricts sapphirine-bearing assemblages to high temperatures by its terminal reaction relations. In addition, the grid restricts orthopyroxene–corundum to low temperatures relative to sapphirine, cordierite to low pressures and garnet to high pressures. The new reaction grid differs from previously published grids (e.g. Windley et al., 1984; Droop, 1989; Goscombe, 1992) in ignoring gedrite and spinel, and in being focused on the high-pressure reactions involving assemblages including orthopyroxene and corundum. In contrast to the grids of Droop (1989) and Goscombe (1992), the grid presents FMAS univariant reactions linked to MAS terminations and divariant equilibria. The new grid incorporates and satisfies the topology of the [Krn Spl Qtz] univariant reaction (Hensen & Green, 1973; Hensen, 1986, 1987) and the kornerupine-present experimental results of Wegge & Schreyer (1994), and is consistent with observations made for similar rocks from the Limpopo Belt (Windley et al., 1984; Droop, 1989).
|
Kornerupine-absent equilibria
Early Opx–Crn and pre- to syn-S3 assemblages
The S3 minerals sapphirine, orthopyroxene and sillimanite each contain relics of Mg-Grt. Although from the textures it is difficult to uniquely deduce reactions leading to the early formation of this garnet, its decrease in modal abundance suggests the traversal of the divariant FMAS [Crn Spl Crd] reaction
 | (1a) |
 | (1b) |
Reaction (1b) lies at high pressure in Fig. 9 and has a slope that is constrained using the thermodynamic dataset of Holland & Powell (1998), and is consistent with the same reaction as calculated by Ouzegane et al. (2003). This Mg end-member reaction lies at P = 16·25 ± 0·25 kbar (1) for T = 850°C and XMgTs = 0·12 (Fig. 10; Gasparik & Newton, 1984). However, the reaction is displaced to lower pressures with increasing Fe content in garnet. For example, using the highest recorded magnesium content for relic garnet from a core zone (XPyp = 0·77), with the corresponding orthopyroxene for that zone (XMg = 0·88), displaces this curve to P 11·7 kbar for T = 850°C (calculated using THERMOCALC, v.3.1, Powell & Holland, 1988). This pressure may be considered an upper limit to conditions reached during D3, and agrees with textural evidence for the absence of kyanite either as inclusions in garnet or pseudomorphed by sillimanite. The position of the kyanite–sillimanite boundary between 800 and 850°C is P = 10·2–11·3 ± 0·11 kbar (1, Fig. 10; calculated using THERMOCALC, v.3.1, Powell & Holland, 1988). These pressure estimates are also consistent with estimates made using S3 assemblages from mafic granulite (P 9–10 kbar; Kelly et al., 2000) and quartz-bearing metapelitic gneiss described above.
|
Post-S3/S4 reaction textures—corundum breakdown
Reaction textures that are preserved in most pods are random in habit. They overgrow S3 in core and transition zones, and S4 in schistose rinds, and therefore formed late- to post-D4. The dominant reaction texture observed in corundum-bearing samples involves single- or double-layer coronas of sapphirine and sillimanite separating corundum from orthopyroxene. These textures may be explained by the MAS model reaction
 | (2a) |
, based on the experimental data of Ackermand (1970), proposed a steeply positive slope (dP/dT 
50 bar/°C), whereas others have suggested moderate, positive slopes (Bertrand et al., 1992: dP/dT = 10–20 bar/°C), or steep negative slopes (Kihle & Bucher-Nurminen, 1992: dP/dT = –50 to –70 bar/°C, based on the thermodynamic data of Berman et al., 1985). A moderately steep, positive slope consistent with Ouzegane et al. (2003; dP/dT 25 bar/°C) has been used here to allow this reaction to link FMAS divariant equilibria generated from adjacent FMAS univariant reactions (Fig. 9).
In zones with higher XSiO2 (outer transition zone and schistose rinds), cordierite becomes increasingly abundant and appears as a replacement phase. Rather than by direct replacement of corundum (+orthopyroxene), the cordierite selectively replaces those textural domains in which sapphirine, with or without sillimanite, already appears to have formed. That is, the development of sapphirine and cordierite symplectites separating orthopyroxene and corundum (Fig. 3c) was not a situation of crossing reaction (2b) (Fig. 9), which would imply initial MAS tie-triangles sapphirine–orthopyroxene–corundum and orthopyroxene–corundum–cordierite. In addition, crossing this reaction is inconsistent with the formation of orthopyroxene and sillimanite in lower XSiO2 bulk compositions. As an alternative, we suggest that, following reaction (2a), the MAS analogue
 | (2c) |
; Kihle & Bucher-Nurminen, 1992; Ouzegane et al., 2003). Reactions (2a) and (2c) intersect at an MAS [Grt Spl] invariant point (Fig. 9), which has been constrained indirectly, using experiments on reactions (2a) and (2c), to lie at P = 11 ± 1·5 kbar and T = 820 ± 50°C (Fig. 10; Ackermand, 1970). With increasing Fe in the system, this MAS invariant point will be displaced as an FMAS univariant to lower pressures and higher temperatures, and will define the FMAS univariant reaction [Grt, Spl] (Figs 9 and 10).
|
Post-S3/S4 reaction textures—garnet breakdown textures
Where garnet occurs as medium- to coarse-grained porphyroblasts it is commonly surrounded by symplectites that are interpreted to have developed during or after D4. Typically, garnet is surrounded and embayed by intergrowths involving orthopyroxene and sapphirine, with plagioclase and/or cordierite. These symplectites may be ‘spinifex-like’ in cross-section (e.g. OG582A; Fig. 3f) or comprise more coarse-grained intergrowths of elongate sapphirine with orthopyroxene (OG580). Small, relic garnet grains commonly occur within the symplectites, and may occur as inclusions within sapphirine. These textures are consistent with breakdown of garnet by the simultaneous progression from left to right, with decreasing pressure, of two component reactions in the CMAS and FMAS sub-systems:
 | (3a) |
 | (3b) |
). However, plagioclase is widespread in some Oygarden samples and found intergrown throughout the sapphirine and orthopyroxene symplectites. Remnant garnet grains also occur in many of these zones, suggesting that garnet was present as relatively large grains and in appreciable abundances (possibly >30%) in the peak metamorphic assemblage, prior to garnet breakdown. A further, localized garnet breakdown texture occurs in OG577. Here, cuneiform intergrowths of sapphirine and sillimanite with orthopyroxene (Fig. 3e) occur around large garnet grains, and sapphirine and sillimanite contain inclusions of garnet. These textures are interpreted to reflect the effects of decompression across a reaction similar to reaction (1a).
Kornerupine-present equilibria
The presence of kornerupine in the S3 and S4 assemblages in Pod 2 further constrains the P–T evolution of these rocks. The reaction and assemblage system being considered has to be expanded to include two new MAS invariant points, [Grt Spl Crd] and [Grt Spl Crn] (Fig. 12), generated from the [Grt Spl] MAS invariant from Fig. 9. This produces a kornerupine field, the size of which will be dependent, in the first instance, on Fe–Mg ratios in the ferromagnesian phases involved. More crucially, the kornerupine field and stability relations will depend on the B and H2O (and F) content of the kornerupine involved. Increasing boron content in kornerupine will act to stabilize the kornerupine-present field, by shifting Krn-present invariant points and along Krn-absent univariants to higher pressures and temperatures (e.g. [Grt Spl Crn]), and to lower pressures (eg. [Grt Spl Crd]; Werding & Schreyer, 1996), with respect to its usual breakdown products that incorporate only minor or negligible B (Grew, 1996).
|
A qualitative, partial petrogenetic grid has been constructed for the minerals including garnet–orthopyroxene–sapphirine–corundum–kornerupine–sillimanite–cordierite in the simplified FMAS(B) system, centred on the [Crn Spl] and [Crd Spl] invariant points (Fig. 12a). The grid includes MAS terminations for selected reactions that are pertinent to the Oygarden assemblages and are discussed below. The topology of the FMAS univariant curves is determined by the [Krn Spl Crn] univariant reaction Opx + Sil = Grt + Spr + Crd. This reaction is equivalent to the [Krn Spl Qtz] univariant reaction and is constrained from the experiments and topological analysis of Hensen & Green (1973)
, Hensen (1986, 1987) and other workers (Bertrand et al., 1991; Audibert et al, 1995; Carrington & Harley, 1995; Harley, 1998b) to have moderate, positive dP/dT slope, with orthopyroxene and sillimanite on the high-P side. The stability of B-free kornerupine in the grid is constrained by the experimental data of Wegge & Schreyer (1994). Reactions (k3), (k7) and (k8) are identical to reactions (2a), (2c) and (1a), respectively, from the [Krn] grid (Fig. 9) and have been given similar nominal slopes. The topology of univariant reactions around the [Crn Spl] FMAS invariant point is largely controlled by the Fe–Mg composition of kornerupine and its position relative to the Opx–Spr–Crd plane in FMAS (Fig. 12a). The deviation of kornerupine from coplanarity with Opx–Spr–Crd as a result of differences in XMg is very small, and based on ambiguities surrounding the calculation of Fe3+ for the phases involved, it is reasonable to assume coplanarity to be the case. Figure 12a and b has therefore been calculated assuming this situation. The result of coplanarity between Opx–Spr–Crd–Krn is the generation of the degenerate reaction Krn = Spr + Crd [Sil Opx Grt] emanating from the [Crn Spl] FMAS invariant point. An alternative case, where kornerupine lies to the magnesian side of the Opx–Spr–Crd plane in FMAS, such as observed using average compositions for minerals from samples OG566 and OG567A when incorporating Fe3+ calculated using the cation/charge balance approach (Fig. 7e), is presented in Fig. 12c. This shift in composition results in changes to the [Grt Crn Spl], [Spr Crn Spl] and [Sil Crn Spl] FMAS univariants. No changes occur to reactions around the [Crd Spl] invariant point.
Based on textural evidence alone it is not clear exactly what sequence of reactions led to the formation of S3 and S4 kornerupine. However, indirect evidence can be taken from inclusion assemblages within or including kornerupine, and from equilibrium mineral associations. A number of potential kornerupine-forming reactions in this system can be considered, associated with the compositional sub-domains:
- core zone: garnet-poor or -absent assemblages, with inferred low XSi and high XMg;
- rind zone: garnet-present assemblages, with inferred lower XMg compared with (1) above.
In garnet-poor assemblages, garnet, corundum and sapphirine inclusions remain enclosed within S3 kornerupine and orthopyroxene. Corundum inclusions are commonly separated from the kornerupine host by sapphirine. These textures are suggestive of an early period of garnet breakdown and also the reaction of corundum, possibly with orthopyroxene, early in D3. A number of possible reaction scenarios at different P–T conditions can be considered. At low T ([Grt Spr Spl] < T < [Grt Spl Crd]), sequentially crossing reactions
 | (k1) |
 | (k2) |
 | (k3) |
 | (k4) |
The dominant S3/S4 assemblage of kornerupine, sapphirine and orthopyroxene may be stable over a broad P–T range (Fig. 12a), although is largely restricted to low XSiO2 bulk compositions (dark shaded areas in Fig. 11b and c). At low T (Figs 11b and 12a) this assemblage is stable at pressures below reaction (k2). But at higher temperatures (T > [Grt Spl Crd]; Fig. 11c), kornerupine begins to break down with increasing pressure to the assemblage orthopyroxene–sapphirine–sillimanite at reaction (k4). Figure 11b and c suggests that for the restricted, relatively low XSiO2 bulk composition of the Oygarden rocks, equilibration in the field bounded by reactions (k2), (k4) and (k6) will produce the observed S3 assemblage kornerupine–sapphirine–orthopyroxene. From the observed textures it is probable that decompression through this field, and/or equilibration in it, was responsible for the S3 assemblages.
Garnet-present assemblages are inferred to have lower bulk XMg compositions compared with garnet-poor or -absent assemblages. S3/S4 kornerupine, sapphirine and orthopyroxene are inferred to have formed in a similar reaction sequence to those described above, but because of the increased stability of garnet with lower XMg, this probably occurred at lower pressures. The preservation of garnet embayed by kornerupine and orthopyroxene suggests decompression late in, or after, the development of S4, possibly sequentially crossing a reaction similar to reaction (k8) and (k4) at low pressure. Alternatively, similar breakdown assemblages may be formed through a number of garnet-breakdown reactions described above, terminating in the orthopyroxene–sapphirine–kornerupine stability field. However, textural evidence for prior reactions is lacking.
Silica metasomatism and the effects of bulk-rock composition on assemblages
The combination of mineral and bulk-rock chemistry with assemblage and textural information illustrates the control that bulk composition exerts on assemblages and reaction products. For example, the modal abundance of sillimanite increases with increasing silica in the bulk composition of transition and rind zones. Similarly, although the appearance of cordierite in some zones is ultimately controlled by changes in pressure and temperature [reaction (2c); Figs 9 and 11a], cordierite is rare in the silica-poorer core zones. In addition, the percentage of progress of reaction (2a) is controlled by the XSiO2 of each zone. A progressive increase in the extent of reaction occurs between the core, transition and rind zones, which is correlated with increasing XSiO2. This increase can be illustrated by comparing the size of reaction coronas in Fig. 2a–c. However, corundum is still preserved in the rind areas. This preservation also indicates that the process that led to the variation in bulk composition between zones is not likely to have occurred during the decompression phase of the RSE metamorphism, and that reaction textures developed in the rocks are not the result of metasomatic reactions. The slight enrichment of Fe and Si, and an apparent depletion in Mg, coupled with the increased abundance of phlogopite in the schistose rinds, suggests that the pods underwent fluid-assisted alteration at some stage in their history. However, if silica-metasomatism did occur during the development of the reaction textures, the orthopyroxene–corundum reactions would be expected to have gone to completion, leaving no relic corundum. In addition, the addition of a free silica phase in this outer zone may also be expected to produce aluminosilicate coronas on, or pseudomorphs after, corundum through a reaction such as Crn + SiO2 = Al2SiO5. No evidence for this exists, therefore it is more likely that metasomatism occurred at least prior to the onset of decompression, and possibly prior to D3 during the earlier Proterozoic or Archaean history of the rock.
The appearance of the stable high-pressure assemblage involving orthopyroxene and corundum is also, to a degree, dependent on the bulk composition. Figure 11a illustrates the restricted bulk compositions that can ‘see’ orthopyroxene and corundum. For the P–T conditions that the Oygarden rocks experienced, only bulk compositions with XSiO2 0·35–0·43 will produce this assemblage. This figure also illustrates why spinel, a common constituent of high-Mg–Al metapelites, is absent from these assemblages. For the high-pressure conditions that accompanied metamorphism in the Oygarden Group, only bulk compositions with XSiO2 < 0·35 will produce spinel. However, spinel may form at lower pressure conditions than those inferred for the Oygarden Group. In addition to the variations discussed above, XSiO2 is inferred to influence the products of garnet-breakdown reactions. For example, the bulk compositions of samples that preserve sillimanite-absent symplectites that formed after garnet plot between the garnet and orthopyroxene composition in MAS (Fig. 8d). Sample OG577, which preserves symplectites that contain both sillimanite and sapphirine, is more silica-rich in comparison with samples with no sillimanite, and plots closer to sillimanite in the MAS ternary diagram (Fig. 8c). Sample OG581 is the most silica-rich sample analysed, plotting above the orthopyroxene–sillimanite tie-line, and hence preserves no sapphirine. It is apparent that silica availability has to an extent controlled the product assemblages of these garnet-breakdown reactions.
The presence of boron
The occurrence of kornerupine in Pod 2 assemblage, and absence from all other pods investigated, is most certainly related to elevated boron in the bulk composition of Pod 2. Three models for the growth of kornerupine in these samples can be considered: (1) isochemical breakdown of a B-bearing phase; (2) addition to the system of a B-bearing fluid; or (3) primary bulk-rock controls (Waters & Moore, 1985; Carson et al., 1995). There is no evidence in any samples for other B-bearing phases such as tourmaline as either inclusions or mineral relics from which the B may be sourced. This suggests that the breakdown of a B-phase was probably not responsible for the production of kornerupine. The pods show ample evidence for metasomatism during an earlier phase of metamorphism, so it could be argued that kornerupine may form through reaction involving a B-bearing fluid (Waters & Moore, 1985). There are no bulk-rock data for Pod 2. However, kornerupine is evenly distributed throughout Pod 2, and not restricted to the schistose rinds and transition zones. It is therefore likely that boron was not introduced during the metasomatic event that produced the dominant zoning that is described in detail in this paper. It is therefore likely that boron was present in the pre-D3/D4 bulk composition (in low abundance in typically B-free minerals) either as a primary compositional variation from the other rocks from which the pods were sourced or as the result of a previous alteration event.
Prograde and retrograde garnet stability
The increase in size and abundance of garnet in the outer zones and rinds of the pods can be correlated directly with differences in garnet chemistry. Garnet XPyp values are lower in rind areas than in cores, and can differ by as much as XPyp = 0·16. In addition, lower XPyp values correlate with higher garnet abundances and grain size. Pod 10, which is garnet-rich in comparison with other pods, also preserves garnet grains that are higher in XGrs (up to 0·06). Moreover, the bulk-rock compositions of samples from Pod 10 commonly have higher CaO contents. The textural context of the metasomatic zones is interpreted to reflect a bulk compositional zoning profile that existed prior to D4, and most probably before D3. Therefore variations in garnet compositions between zones reflect an initial bulk-rock composition and not changes caused by Fe–Mg exchange or metasomatism during metamorphism.
Garnet inclusions in S3 sapphirine and kornerupine suggest that D3 garnet growth during a prograde pressure increase was followed by a period of garnet breakdown. Garnet breakdown is interpreted to have been the result of decompression from peak pressures. This decompression during the D3 event would have allowed growth of S3 sapphirine and kornerupine with inclusions of garnet. Symplectites that replace garnet and overgrow S4 in garnet-rich rocks imply that further garnet breakdown occurred during continued decompression late in D4. These two apparent stages of garnet breakdown can be correlated with the Mg and, to a lesser extent, Ca contents of the garnet grains from the different zones. Garnet grains with high XPyp values mainly occur as inclusions in S3 sapphirine and kornerupine, and broke down early in the decompression path. Garnet grains with less magnesian compositions, for example those located in the transition zones and rinds, are more commonly preserved in the S3 assemblage as small garnet relics. This garnet probably began to break down late in D3 or early in D4. Breakdown of garnet grains that are now preserved as larger, embayed porphyroblasts produced symplectites that overprint S4 and therefore are interpreted to have occurred late in D4. Preservation of larger garnet grains in the rind zones is therefore a result of higher Fe contents displacing the garnet-present equilibria [e.g. reaction (1a); Fig. 9] to lower pressures. This may also explain the relative abundances of garnet in different samples, as prograde garnet would probably have formed at lower pressures in samples with higher Fe contents than in samples with higher Mg contents. This garnet would in turn be consumed at lower pressures on the decompression path.
Equilibrium assemblage: orthopyroxene and corundum
In corundum-bearing samples, corundum is commonly separated from orthopyroxene by reaction rims of sapphirine, sapphirine + sillimanite, or sapphirine + cordierite (Fig. 3). In the case of mono-mineralic sapphirine coronas, these rims can be narrow and mimic the grain boundaries of the corundum (Fig. 3a). The widths of these coronas are much less than the size of the grains that they separate. Moreover, several orthopyroxene grains may be adjacent to any one corundum grain, suggesting that corundum does not occur as metastable inclusions in orthopyroxene. This texture implies that textural, if not chemical, equilibrium was attained between orthopyroxene and corundum in the Oygarden rocks prior to reaction. This assemblage has been previously proposed to exist from chemographic and topological requirements by Hensen (1986, 1987) and Waters (1986), and also from experimental evidence (Ackermand, 1970). Morse & Talley (1971) reported orthopyroxene + corundum (+ spinel) amongst the breakdown products of sapphirine + quartz (Wilson Lake, Labrador), and interpreted the association to reflect a high-P instability of sapphirine + magnetite. Grew (1983) also reported from Enderby Land (East Antarctica) the occurrence of corundum within sillimanite coronae around sapphirine, where orthopyroxene formed further coronae around sillimanite, but did not infer stability of the orthopyroxene–corundum association in these rocks. In addition, Motoyoshi et al. (1990) inferred orthopyroxene + corundum as reaction products after sapphirine + quartz breakdown during apparent isobaric cooling. Other workers have inferred prior equilibrium between orthopyroxene and corundum (e.g. Windley et al., 1984; Bertrand et al., 1992; Goscombe, 1992; Kihle & Bucher-Nurminen, 1992), usually on the basis of small grains that are not always in proximity to orthopyroxene, or from inclusion assemblages (Kriegsman & Schumacher, 1999). Recently, more convincing evidence was presented by Ouzegane et al. (2003), who reported the extensive development of sapphirine + sillimanite coronae armouring corundum, with sapphirine + cordierite coronae separating sillimanite from orthopyroxene. Ouzegane et al. (2003) inferred orthopyroxene–corundum breakdown through prograde heating at P 10 kbar. These textures form a similar sequence to that reported here for the Oygarden Group metapelites. However, the Oygarden rocks are one of the clearest examples of orthopyroxene–corundum stability that we are aware of, proving that this assemblage can occur in natural systems.
This rare association is predicted to occur only in rocks with a restricted bulk-rock composition metamorphosed under restricted P–T conditions. Grew (1983) suggested that the reaction orthopyroxene + corundum = garnet + sapphirine + sillimanite reflected the upper temperature limit of the assemblage (as indicated in Fig. 9 of this study). In addition, conditions of P = 9–11·5 kbar and T = 750–850°C (above the kyanite–sillimanite univariant curve) with low water activities were inferred to be required to stabilize the assemblage (Grew, 1983). Currently available thermodynamic data also indicate that the rocks must pass through, or equilibrate at, moderately high pressures of P > 11 kbar (for the MAS system) or slightly lower for rocks with lower XMg. The Oygarden rocks suggest that this assemblage formed at conditions of P > 10 kbar and T = 800–850°C. The restricted P–T range within which orthopyroxene and corundum will coexist means that this assemblage is unlikely to be preserved in most circumstances. During the later stages of typical clockwise P–T paths, characterized by decompression after peak metamorphic conditions, this assemblage easily reacts to produce assemblages involving orthopyroxene, sapphirine, and sillimanite or cordierite, or to form kornerupine-bearing assemblages. Anti-clockwise P–T paths, preserved in a number of granulite terranes, are unlikely to intersect conditions appropriate for orthopyroxene–corundum stability. However, in this situation this assemblage may form through reaction between sapphirine and sillimanite if the sapphirine involved has high Mg content, and the cooling path passed through T 800–850°C at P > 8–10 kbar. This circumstance may be rare.
|
P–T–t PATH FOR KEMP LAND |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
The sequence of kornerupine-absent and -present equilibria described above can be coupled to establish a clockwise P–T–t evolution for the Oygarden Group of islands during the Rayner Structural Episode (Fig. 10).
Prograde history and initial decompression
The S3 assemblages, including kornerupine, orthopyroxene and sapphirine, record increasing pressure through the kornerupine stability field (Figs 9 and 11) to reach peak pressures of up to 10–11 kbar, and possibly higher. This pressure increase reflects tectonic thickening of the terrane during D3 thrusting that occurred in an overall transpressional setting (Kelly et al., 2000). It is unclear at what crustal level this part of Kemp Land resided prior to the RSE. However, coarse-grained garnet coronas and ‘necklaces’ in mafic granulites that are also embayed by post-D3 reaction textures (Kelly, 2000) suggest that thickening did in fact occur, and the S3 assemblages described here do not simply reflect thrust-related activation of the middle to lower continental crust.
Garnet inclusions in S3 sapphirine and kornerupine are consistent with garnet breakdown subsequent to peak pressures being attained. These textures are interpreted to result from crossing one or more reactions that emanate from the [Crd Sil] or the [Crd Spl] MAS invariants (Fig. 9), and reactions (k1), (k2) or (k8) in kornerupine-bearing rocks (Fig. 12). These reactions reflect a phase of decompression that commenced during D3, with peak pressures reached prior to the thermal maximum. Corundum with sapphirine coronas included in kornerupine also suggests that reactions (k3) and (k4) may have been crossed with decreasing pressure. This P–T path is on the high-T side of the [Grt Spl Crd] invariant point in the Krn-present grid, implying that early decompression from peak D3 pressures above the [Grt Spl Crd] point was accompanied by a rise in temperature late in D3 or during D4, followed by minor cooling across reaction (k4).
The continued heating of the Oygarden terrane, or at least residence at high temperatures during D3 and D4, is reflected in the grain size and texture of thrust zones. D3 thrust zones are typically not mylonitic in texture but instead are coarse-grained, suggesting that deformation was characterized by a slow strain rate or the thrust zones were annealed following deformation (Kelly et al., 2000). The latter interpretation is consistent with a period of higher temperatures following D3. Based on the experimentally constrained KFMASH grid (Hensen & Green, 1973; Carrington & Harley, 1995), fine-grained intergrowths of orthopyroxene, sillimanite and quartz after garnet and biotite in quartz-present metapelites (Kelly, 2000) indicate that peak temperatures may have exceeded 900°C for a limited period of time. The absence of ultrahigh-temperature assemblages, for example the association sapphirine + quartz, provides a limit to peak temperatures achieved, such that the temperature rise during decompression was probably not greater than 100°C.
Post-peak decompression
Reaction textures between orthopyroxene and corundum in kornerupine-absent rocks are interpreted to represent decompression across reactions (2a) and (2c) (Fig. 9) during D4. The slope of reaction (2a) is not well constrained experimentally, and it is possible that this reaction is crossed through heating (e.g. Bertrand et al., 1992) or decompression with either rising or falling temperature. Reaction (2c), in corundum-bearing rocks, has a shallow, positive slope. The products of this reaction are commonly fine-grained with respect to the reactant phases, so it is realistic to assume that these reactions were crossed as a result of decompression. Assemblages from kornerupine-bearing rocks suggest that this may have followed, or was accompanied by, minor heating to conditions above those appropriate to [Grt Spl Crd]. Furthermore, garnet breakdown textures that overgrew S4 reflect decompression across reaction (1a) late in, or following, D4. Reactions (1a) and (2a), which occur at different relative pressures and temperatures (Fig. 9), appear to have occurred in different rocks during the same period of decompression. This contradiction can be explained by differences in bulk-rock compositions and the resulting phase compositions, which are inferred to displace these equilibria a considerable distance in P–T space.
Although no quantitative pressure estimates could be made for the reaction textures developed in these rocks, estimates from mafic gneiss (Kelly, 2000; Kelly et al., 2000), and estimates made for reaction textures in mafic gneiss from elsewhere in Kemp Land (Else Nunataks, c. 80 km SW of the Oygarden Group; Pieters & Wyborn, 1977; Ellis, 1983), suggest that decompression from P 10 kbar to P 5 kbar occurred during and after D4. This exhumation, which was initially triggered during D3 thrusting, is inferred to have continued during the shift from thrusting to extension during D4. This latter event is characterized in the Oygarden Group by a 2–3 km crustal-scale shear zone. The coarse-grained nature of rocks reflects a degree of annealing of mylonitic textures following cessation of movement on this shear zone, and suggests the possibility of as yet unidentified structures controlling the continued exhumation of the terrane.
Mineral assemblages that characterize S3 and S4 in the silica-undersaturated rocks are commonly similar and of similar grain size. Moreover, the grain sizes of minerals that are involved in the reaction textures that developed during D3 are similar to, or larger than, those that developed later in D4. The post-D4 textures are commonly appreciably smaller in grain size than the peak mineral assemblages that they overprint. This suggests that the textures did not form purely by heating, which would produce later textures that were coarser in grain size, or completely overprint the earlier textures. Therefore the reaction textures described here did not form as a result of a second metamorphic event at lower pressures but are consistent with a single metamorphic/orogenic cycle. This continuous clockwise P–T path was dominated by decompression and accompanied by a minor component of heating. Age estimates for D3 and D4 overlap (c. 930–900 Ma; Kelly et al., 2002), supporting this interpretation.
Kemp Land within the Rayner Complex
The characterization in Kemp Land of a clockwise P–T trajectory dominated by post-peak decompression (Pieters & Wyborn, 1977; Ellis; 1983; this study) contrasts markedly with P–T paths recorded elsewhere in the Rayner Complex. Reaction textures in rocks from the Mawson Coast are interpreted to be consistent with an anticlockwise P–T path with isobaric cooling after the peak of metamorphism (Clarke et al., 1989; Dunkley et al., 2003). In the Stillwell Hills, an early phase of apparent cooling may have preceded decompression (Clarke, 1987; Harley & Hensen, 1990) and is similar to the P–T path recorded by rocks in the northern Prince Charles Mountains (Fitzsimons & Thost, 1992; Thost & Hensen, 1992; Hand et al., 1994; Nichols, 1995). More recent work (Boger & White, 2003) has suggested an anticlockwise P–T path dominated by retrograde cooling for the northern Prince Charles Mountains.
These differences may be linked to the relative timing of metamorphism and/or the location of these areas with respect to contraction across the orogen. Sheraton et al. (1987) suggested that areas that preserve lower-pressure assemblages, such as MacRobertson Land and the northern Prince Charles Mountains, record the later stages of the metamorphic event, and the areas that preserve higher-pressure assemblages, such as Kemp Land, record only the earlier, high-pressure part of the event. The higher-pressure rocks were interpreted to have been exhumed by thrusting over the Napier Complex, so were not recrystallized at lower pressures (Sheraton et al., 1987; Harley & Hensen, 1990). However, isotopic data from the Oygarden area indicate that thrusting and high-pressure granulite-facies metamorphism affected Kemp Land between c. 0·93 and 0·90 Ga (Kelly et al., 2002), after the low-P, high-T event in MacRobertson Land that is interpreted to have occurred between c. 0·99 and 0·97 Ga (Manton et al., 1992; Kinny et al., 1997; Boger et al., 2000; Dunkley et al., 2003). Therefore, the earliest deformation in MacRobertson Land does not correlate with that seen in Kemp Land, and the isobaric cooling event in MacRobertson Land cannot be directly compared with the isothermal decompression event in Kemp Land. Interestingly, a second phase of metamorphism and deformation, correlated with a complex-wide shift from broadly contractional to transpressional tectonism, affected the Rayner Complex of MacRobertson Land between c. 0·94 and 0·90 Ma (Young & Black, 1991; Manton et al., 1992; Boger et al., 2000; Carson et al., 2001; Dunkley et al., 2003). Apparent decompression inferred from rocks of the northern Prince Charles Mountains (Hand et al., 1994; Nichols, 1995) is probably related to this later event, the age of which correlates with the dominant phase of metamorphism and deformation in Kemp Land.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
The high-Al–Mg rocks described here contain textural evidence for the existence in nature of the stable assemblage involving orthopyroxene and corundum. This assemblage has been previously proposed to exist from chemographic and topological requirements, from experimental evidence and on the basis of reaction textures and inclusion assemblages. Textural evidence in the Oygarden rocks shows that this mineral assemblage was stable, and will occur as a stable assemblage only in rocks with a restricted bulk-rock composition (XSiO2 = 0·35–0·43 for high-magnesian rocks) metamorphosed under restricted P–T and low water activity conditions. For these compositions, moderately high pressures of P > 11 kbar are required (for the MAS system) or slightly lower for rocks with higher XFe. In the Oygarden rocks, this assemblage formed at conditions of P > 9 kbar and T = 800–850°C. At pressures below this, assemblages will develop minerals including sapphirine, sillimanite and cordierite with orthopyroxene; at much higher pressures (P > 13 kbar) orthopyroxene and corundum will react to form garnet; and at lower temperatures, or if boron is present, late kornerupine will develop. The restricted P–T range within which orthopyroxene and corundum can coexist means that this assemblage is unlikely to be preserved in most circumstances.
The reaction grid presented to describe textures from the Oygarden Group silica-undersaturated metapelites differs from previously published grids by ignoring gedrite and spinel, instead focusing on the high-pressure reactions. The grid presents FMAS univariant reactions linked to MAS terminations and divariant equilibria, and incorporates and satisfies the topology of the [Krn Spl Crn] univariant reaction and the kornerupine-present experimental results of Wegge & Schreyer (1994). The integration of assemblage and reaction texture information with the reaction grid allows the definition of a clockwise P–T path that reached peak pressures of P 9–10 kbar prior to a thermal maximum. Initial decompression was accompanied by thermal relaxation of the crust toward peak temperatures of up to T 850–900°C and continued decompression. The clockwise isothermal decompression path recorded by Kemp Land contrasts with elsewhere in the Rayner Complex, and is interpreted to correlate with the second phase of tectonism to affect MacRobertson Land.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Supplementary data for this paper are available at Journal of Petrology online.
|
ACKNOWLEDGEMENTS |
|---|
This work was completed with funding from the Antarctic Science Advisory Committee (Geoff L. Clarke and S.L.H.; ASAC Project No. 2214). Fieldwork was conducted during the 1996–1997 and 1997–1998 Australian National Antarctic Research Expeditions. The Australian Antarctic Division and the personnel of Mawson Base are thanked for logistic support. Part of this work was completed while N.M.K. was a postgraduate student visitor at the University of Edinburgh, with the assistance of an Edgeworth David Travel Grant-In-Aid (University of Sydney). N.M.K. also acknowledges the support of a University of Sydney Postgraduate Research Award (1996–2000) and a Royal Society of Edinburgh SEELLD Personal Fellowship (2001–). XRF analyses were funded by a Royal Society Grant to S.L.H. Drs Peter Hill and Nicola Cayzer are thanked for assistance with the electron probe and scanning electron microscope (University of Edinburgh); Barry Searle (UNSW) is thanked for his assistance with electron probe and element map analysis. Nathan Daczko is thanked for help with element map processing. SIMS analyses of kornerupine were performed with the assistance of Richard Hinton at the University of Edinburgh. Chris Carson, Julie Hollis, Richard White and Venessa Bennett are thanked for assistance in the field, and Geoff Clarke is thanked for editing an early version of this manuscript. The authors would also like to thank E. S. Grew, K. K. Podlesskii and V. N. Balashov for thorough and constructive reviews, and L. L. Perchuk for editorial assistance.
|
FOOTNOTES |
|---|
* Corresponding author. Present address: School of GeoSciences, University of Edinburgh, Grant Institute, Kings Buildings, Edinburgh EH9 3JW, UK. Telephone: +44-(0)131-650 5885. Fax: +44-(0)131-668 3184. E-mail: Nigel.Kelly{at}glg.ed.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Ackermand, D. (1970). Stabilitat und Synthese des Sapphirins im System MgO–Al2O3–SiO2–(H2O). Ph.D. thesis, Bochum University, 97 pp.
Aranovitz, L. Y. & Berman, R. G. (1997). A new garnet–orthopyroxene barometer based on reversed Al2O3 solubility in FeO–Al2O3–SiO2 orthopyroxene. American Mineralogist 82, 345–353.[Abstract]
Audibert, N., Hensen, B. J. & Bertrand, P. (1995). Experimental study of phase relations involving osumulite in the system K2O–FeO–MgO–Al2O3–SiO2–H2O at high pressure and temperature. Journal of Metamorphic Geology 13, 331–344.[Web of Science]
Bence, A. E. & Albee, A. L. (1968). Empirical correction factors for the electron microanalysis of silicates and oxides. Journal of Geology 76, 382–403.[Web of Science]
Berman, R. G., Brown, T. H. & Greenwood, H. J. (1985). An internally consistent thermodynamic data base for minerals in the system Na2O–K2O–CaO–MgO–FeO–SiO2–Al2O3–Fe2O3–TiO2–H2O–CO2. Atomic Energy Canada Ltd, Technical Report TR-377.
Bertrand, P., Ellis, D. J. & Green, D. H. (1991). The stability of sapphirine–quartz and hypersthene–sillimanite–quartz assemblages: an experimental investigation in the system FeO–MgO–Al2O3–SiO2 under H2O and CO2. Contributions to Mineralogy and Petrology 108, 55–71.[CrossRef][Web of Science]
Bertrand, P., Ouzegane, K. & Kienast, J. R. (1992). P–T–X relationships in the Precambrian Al–Mg-rich granulites from In Ouzzal, Hoggar, Algeria. Journal of Metamorphic Geology 10, 17–31.[Web of Science]
Boger, S. D. & White, R. W. (2003). The metamorphic evolution of metapelitic granulites from Radok Lake, northern Prince Charles Mountains, east Antarctica; evidence for an anticlockwise P–T path. Journal of Metamorphic Geology 21, 285–298.[Web of Science]
Boger, S. D., Carson, C. J., Wilson, C. J. L. & Fanning, C. M. (2000). Neoproterozoic deformation in the Radok Lake region of the northern Prince Charles Mountains, east Antarctica; evidence for a single protracted orogenic event. Precambrian Research 104, 1–24.[CrossRef][Web of Science]
Carrington, D. P. & Harley, S. L. (1995). Partial melting and phase relations in high-grade metapelites: an experimental petrogenetic grid in the KFMASH system. Contributions to Mineralogy and Petrology 120, 270–291.[Web of Science]
Carson, C. J., Hand, M. & Dirks, P. H. G. M. (1995). Stable coexistence of grandidierite and kornerupine during medium pressure granulite facies metamorphism. Mineralogical Magazine 59, 327–339.[Abstract]
Carson, C. J., Boger, S. D., Fanning, C. M., Wilson, C. J. L. & Thost, D. E. (2001). SHRIMP U–Pb geochronology from Mt Kirkby, northern Prince Charles Mountains, east Antarctica. Antarctic Science 12, 429–442.
Clarke, G. L. (1987). A comparative study of the structural and metamorphic evolution of the Olary (South Australia) and Stillwell Hills (Antarctica) Precambrian terrains. Ph.D. thesis, University of Melbourne, 254 pp.
Clarke, G. L., Powell, R. & Guiraud, M. (1989). Low-pressure granulite facies metapelitic assemblages and corona textures from MacRobertson Land, east Antarctica: the importance of Fe2O3 and TiO2 in accounting for spinel-bearing assemblages. Journal of Metamorphic Geology 7, 323–335.[Web of Science]
Clarke, G. L., Daczko, N. & Nockolds, C. (2001). A method for applying matrix corrections to X-ray intensity maps using the Bence–Albee algorithm and Matlab. Journal of Metamorphic Geology 19, 635–644.[Web of Science]
Deer, W. A., Howie, R. A. & Zussman, J. (1992). An Introduction to the Rock Forming Minerals. Harlow: Longman, 696 pp.
Droop, G. T. R. (1989). Reaction history of garnet–sapphirine granulites and conditions of Archaean high-pressure granulite-facies metamorphism in the Central Limpopo Mobile Belt, Zimbabwe. Journal of Metamorphic Geology 7, 383–403.[Web of Science]
Droop, G. T. R. & Bucher-Nurminen, K. (1984). Reaction textures and metamorphic evolution of sapphirine-bearing granulites from the Gruf Complex, Italian Central Alps. Journal of Petrology 25, 766–803.
Dunkley, D. J., Clarke, G. L. & Harley, S. L. (1999). Diffusion metasomatism in silica-undersaturated sapphirine-bearing granulite from Rumdoodle Peak, Framnes Mountains, east Antarctica. Contributions to Mineralogy and Petrology 134, 264–276.[CrossRef][Web of Science]
Dunkley, D. J., Clarke, G. L. & White R. W. (2003). The c. 1020–900 Ma Rayner Structural Episode in MacRobertson Land, east Antarctica: a case of oblique continental collision? Proceedings of the 8th International Symposium on Antarctic Earth Sciences, Wellington, New Zealand, 1999. Royal Society of New Zealand Bulletin 35, 31–42.
Ellis, D. J. (1983). The Napier and Rayner Complexes of Enderby Land, Antarctica—contrasting styles of metamorphism and tectonism. In: Oliver, R. L., James, P. R. & Jago, J. B. (eds) Antarctic Earth Science. Canberra, A.C.T.: Australian Academy of Science, pp. 20–24.
Fitzsimons, I. C. W. & Harley, S. L. (1994). The influence of retrograde cation exchange on granulite P–T estimates and a convergence technique for the recovery of peak metamorphic conditions. Journal of Metamorphic Geology 12, 411–428.[Web of Science]
Fitzsimons, I. C. W. & Thost, D. E. (1992). Geological relationships in high-grade basement gneiss of the northern Prince Charles Mountains, east Antarctica. Australian Journal of Earth Sciences 39, 173–193.[Web of Science]
Friend, C. R. L. (1995). Occurrences of boron-free and boron-poor kornerupine. Mineralogical Magazine 59, 163–166.
Ganguly, J., Cheng, W. & Tirone, M. (1996). Thermodynamics of aluminosilicate garnet solid solution: new experimental data, an optimized model, and thermometric applications. Contributions to Mineralogy and Petrology 126, 137–151.[CrossRef][Web of Science]
Gasparik, T. & Newton, R. C. (1984). The reversed alumina contents of orthopyroxene in equilibrium with spinel and forsterite in the system MgO–Al2O3–SiO2. Contributions to Mineralogy and Petrology 85, 186–196.[CrossRef][Web of Science]
Goscombe, B. (1992). Silica-undersaturated sapphirine, spinel and kornerupine granulite facies rocks, NE Strangeways Range, central Australia. Journal of Metamorphic Geology 10, 181–201.[Web of Science]
Grew, E. S. (1983). Sapphirine–garnet and associated parageneses in Antarctica. In: Oliver, R. L., James, P. R. & Jago, J. B. (eds) Antarctic Geoscience. Canberra, A.C.T.: Australian Academy of Science, pp. 40–43.
Grew, E. S. (1996). Borosilicates (exclusive of tourmaline) and boron in rock-forming minerals in metamorphic environments. In: Grew, E. S. & Anovitz, L. M. (eds) Boron: Mineralogy, Petrology and Geochemistry. Mineralogical Society of America, Reviews in Mineralogy 33, 387–480.
Grew, E. S., Manton, W. I. & James, P. R. (1988). U–Pb data on granulite facies rocks from Fold Island, Kemp Coast, east Antarctica. Precambrian Research 42, 63–75.[CrossRef][Web of Science]
Grew, E. S., Redhammer, G. J., Amthauer, G., Cooper, M. A., Hawthorne, F. C. & Schmetzer, K. (1999). Iron in kornerupine: a 57Fe Mössbauer spectroscopic study and comparison with single-crystal refinement. American Mineralogist 84, 536–549.[Abstract]
Hand, M., Scrimgeour, I., Powell, R., Stuwe, K. & Wilson, C. J. L. (1994). Metapelitic granulites from Jetty Peninsula, east Antarctica: formation during a single event or by polymetamorphism? Journal of Metamorphic Geology 12, 557–573.[Web of Science]
Harley, S. L. (1993). Sapphirine granulites from the Vestfold Hills, East Antarctica: geochemical and metamorphic evolution. Antarctic Science 5, 389–402.[Web of Science]
Harley, S. L. (1998a). Ultrahigh temperature granulite facies metamorphism (1050°C, 12 kbar) and decompression in garnet (Mg70)–orthopyroxene–sillimanite gneisses from the Rauer Goup, East Antarctica. Journal of Metamorphic Geology 16, 541–562.[CrossRef][Web of Science]
Harley, S. L. (1998b). On the occurrence and characterisation of ultrahigh-temperature crustal metamorphism. In: Treloard, P. J. & O'Brien, P. J. (eds) What Drives Metamorphism and Metamorphic Reactions? Geological Society, London, Special Publications 138, 81–107.
Harley, S. L. & Black, L. P. (1997). A revised Archaean chronology for the Napier Complex, Enderby Land, from SHRIMP ion-microprobe studies. Antarctic Science 9, 74–91.[Web of Science]
Harley, S. L. & Green, D. H. (1982). Garnet–orthopyroxene barometry for granulites and peridotites. Nature 300, 697–701.[CrossRef]
Harley, S. L. & Hensen, B. J. (1990). Archaean and Proterozoic high-grade terranes of east Antarctica (40–80°): a case study of diversity in granulite facies metamorphism. In: Ashworth, J. R. & Brown, M. (eds) High-temperature Metamorphism and Crustal Anatexis. Mineralogical Society Special Publications 2, 320–370.
Harley, S. L. & Motoyoshi, S. L. (2000). Al zoning in orthopyroxene in a sapphirine quartzite: evidence for >1120°C UHT metamorphism in the Napier Complex, Antarctica, and implications for the entropy of sapphirine. Contributions to Mineralogy and Petrology 138, 293–307.[CrossRef][Web of Science]
Hensen, B. J. (1971). Theoretical phase relations involving cordierite and garnet in the system, MgO–FeO–Al2O3–SiO2. Contributions to Mineralogy and Petrology 3, 191–214.[Medline]
Hensen, B. J. (1986). Theoretical phase relations involving garnet and cordierite revisited: the influence of oxygen fugacity on the stability of sapphirine and spinel in the system Mg–Fe–Al–Si–O. Contributions to Mineralogy and Petrology 92, 191–214.
Hensen, B. J. (1987). P–T grids for silica-undersaturated granulites in the systems MAS (n + 4) and FMAS (n + 3)—tools for the derivation of P–T paths of metamorphism. Journal of Metamorphic Geology 5, 255–271.[Web of Science]
Hensen, B. J. & Green, D. H. (1973). Experimental study of the stability of cordierite and garnet in pelitic compositions at high pressures and temperatures. III. Synthesis of experimental data and geological applications. Contributions to Mineralogy and Petrology 38, 151–166.[CrossRef][Web of Science]
Hinton, R. W. (1995). Ion microprobe analysis in geology. In: Reed, S. J. B. & Potts, P. J. (eds) Microanalytical Techniques. London: Chapman & Hall, pp. 235–289.
Holland, T. J. B. & Powell, R. (1998). An internally consistent thermodynamic dataset for phases of petrological interest. Journal of Metamorphic Geology 16, 309–343.[CrossRef][Web of Science]
Kelly, N. M. (2000). Deformation and metamorphism in the lower continental crust: Oygarden Group of Islands, east Antarctica. Ph.D. thesis, University of Sydney, 287 pp.
Kelly, N. M., Clarke, G. L., Carson, C. J. & White, R. W. (2000). Thrusting in the lower crust: evidence from the Oygarden Islands, Kemp Land, East Antarctica. Geological Magazine 137, 219–234.
Kelly, N. M., Clarke, G. L. & Fanning, C. M. (2002). A two-stage evolution of the Neoproterozoic Rayner Structural Episode: new U–Pb sensitive high resolution ion microprobe constraints from the Oygarden Group, Kemp Land, East Antarctica. Precambrian Research 116, 307–330.[CrossRef][Web of Science]
Kihle, J. & Bucher-Nurminen, K. (1992). Orthopyroxene–sillimanite–sapphirine granulites from the Bamble granulite terrane, southern Norway. Journal of Metamorphic Geology 10, 671–683.[Web of Science]
Kinny, P. D., Black, L. P. & Sheraton, J. W. (1997). Zircon U–Pb ages and geochemistry of igneous and metamorphic rocks from the northern Prince Charles Mountains, Antarctica. AGSO Journal of Australian Geology and Geophysics 16, 637–654.
Kretz, R. (1983). Symbols for rock-forming minerals. American Mineralogist 68, 277–279.[Abstract]
Kriegsman, L. M. & Schumacher, J. C. (1999). Petrology of sapphirine-bearing and associated granulites from central Sri Lanka. Journal of Petrology 40, 1211–1239.[CrossRef][Web of Science]
Manton, W. I., Grew, E. S., Hofmann, J. & Sheraton, J. W. (1992). Granitic rocks of the Jetty Peninsula, Amery Ice Shelf area, east Antarctica. In: Yoshida, Y., Kaminuma, K. & Shiraishi, K. (eds) Recent Progress in Antarctic Earth Science. Tokyo: Terra, pp. 179–189.
Moore, P. B. (1969). The crystal structure of sapphirine. American Mineralogist 54, 21–49.
Morse, S. A. & Talley, J. H. (1971). Sapphirine reactions in deep-seated granulites near Wilson Lake, central Labrador, Canada. Earth and Planetary Science Letters 10, 325–328.[CrossRef]
Motoyoshi, Y., Hensen, B. J. & Matsueda, H. (1990). Metastable growth of corundum adjacent to quartz in a spinel-bearing quartzite from the Napier Complex, Antarctica. Journal of Metamorphic Geology 8, 125–130.[Web of Science]
Motoyoshi, Y., Ishikawa, M. & Fraser, G. L. (1995). Sapphirine-bearing silica-undersaturated granulites from Forefinger Point, Enderby Land, East Antarctica: evidence for a clockwise P–T path? Proceedings of the NIPR Symposium on Antarctic Geosciences 8, 121–129.
Newton, R. C. (1972). An experimental determination of the high pressure stability limits of magnesian cordierite under wet and dry conditions. Journal of Geology 80, 398–420.[Web of Science]
Nichols, G. T. (1995). The role of mylonites in the uplift of an oblique lower crustal section, east Antarctica. Journal of Metamorphic Geology 13, 223–238.[Web of Science]
Ouzegane, K., Guiraud, M. & Kienast, J. R. (2003). Prograde and retrograde evolution in high-temperature corundum granulites (FMAS and KFMASH systems) from In Ouzzal Terrane (NW Hoggar, Algeria). Journal of Petrology 44, 517–545.
Pieters, P. E. & Wyborn, D. (1977). Geological work in Antarctica—1974/75. Bureau of Mineral Resources, Australia, Record 1977/16 (unpublished).
Pouchou, L. & Pichoir, F. (1984). A new model for quantitative X-ray microanalysis. Recherche Aerospatiale 3, 167–192.
Powell, R. & Holland, T. J. B. (1988). An internally consistent dataset with uncertainties and correlations: 3. Applications to geobarometry, worked examples and a computer program. Journal of Metamorphic Geology 6, 173–204.[Web of Science]
Sandiford, M. & Wilson, C. J. L. (1984). The structural evolution of the Fyfe Hills–Khmara Bay region, Enderby Land, east Antarctica. Australian Journal of Earth Sciences 31, 403–426.[Web of Science]
Schreyer, W. & Seifert, F. (1969). Compatibility relations of the aluminium silicates in the systems MgO–Al2O3–SiO2–H2O and K2O–AlgO–Al2O3–SiO2–H2O at high pressures. American Journal of Science 267, 371–388.[Abstract]
Sheraton, J. W. & Black, L. P. (1983). Geochemistry of Precambrian gneisses: relevance for the evolution of the East Antarctic Shield. Lithos 16, 273–296.[CrossRef][Web of Science]
Sheraton, J. W., England, R. N. & Ellis, D. J. (1982). Metasomatic zoning in sapphirine-bearing granulites from Antarctica. BMR Journal of Geology & Geophysics 7, 269–273.
Sheraton, J. W., Tingey, R. J., Black, L. P., Offe, L. A. & Ellis, D. J. (1987). Geology of an unusual Precambrian high-grade metamorphic terrane—Enderby Land and western Kemp Land, Antarctica. Australian Bureau of Mineral Resources Bulletin 223, 51 pp.
Spear, F. S. (1993). Metamorphic Phase Equilibria and Pressure–Temperature–Time Paths. Mineralogical Society of America, Monograph, pp. 608–625.
Thost, D. E. & Hensen, B. J. (1992). Gneisses of the Porthos and Athos Ranges, northern Prince Charles Mountains, constraints on the prograde and retrograde P–T path. In: Yoshida, Y., Kaminuma, K. & Shiraishi, K. (eds) Recent Progress in Antarctic Earth Science. Tokyo: Terra, pp. 93–102.
Vry, J. K. (1994). Boron-free kornerupine from the Reynolds Range, Arunta Block, central Australia. Mineralogical Magazine 58, 27–37.[Abstract]
Waters, D. J. (1986). Metamorphic history of sapphirine-bearing and related magnesian gneisses from Namaqualand, South Africa. Journal of Petrology 27, 541–565.
Waters, D. J. & Moore, P. B. (1985). Kornerupine in Mg–Al-rich gneisses from Namaqualand, South Africa: mineralogy and evidence for late-metamorphic fluid activity. Contributions to Mineralogy and Petrology 91, 247–259.
Wegge, S. & Schreyer, W. (1994). Boron-free kornerupine: its upper pressure stability limit in the system MgO–Al2O3–SiO2–H2O (MASH). European Journal of Mineralogy 6, 67–75.
Werding & Schreyer, W. (1996). Experimental studies on borosilicates and selected borates. In: Grew, E. S. & Anovitz, L. M. (eds) Boron: Mineralogy, Petrology and Geochemistry. Mineralogical Society of America, Reviews in Mineralogy 33, 117–159.
Windley, B. F., Ackermand, D. & Herd, R. K. (1984). Sapphirine/kornerupine-bearing rocks and crustal uplift history of the Limpopo belt, Southern Africa. Contributions to Mineralogy and Petrology 86, 342–358.[CrossRef][Web of Science]
Young, D. N. & Black, L. P. (1991). U–Pb zircon dating of Proterozoic igneous charnockites from the Mawson coast, east Antarctica. Antarctic Science 3, 205–216.[Web of Science]
Young, D. N., Zhao, J.-X., Ellis, D. J. & McColloch, M. T. (1997). Geochemical and Sr–Nd isotopic mapping of source provinces for the Mawson charnockites, east Antarctica: implications for Proterozoic tectonics and Gondwana reconstruction. Precambrian Research 86, 1–19.[CrossRef][Web of Science]
Zhao, J.-x., Ellis, D. J., Kilpatrick, A. & McColloch, M. T. (1997). Geochemical and Sr–Nd isotopic study of charnockites and related rocks in the northern Prince Charles Mountains, East Antarctica: implications for charnockite petrogenesis and Proterozoic crustal evolution. Precambrian Research 81, 37–66.[CrossRef][Web of Science]
|
REFERENCES IN SUPPLEMENTARY TABLES NOT CITED IN THE TEXT |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGOUTCROP DESCRIPTION AND...MINERAL CHEMISTRY OF QUARTZ...MINERAL CHEMISTRY OF QUARTZ...BULK-ROCK GEOCHEMISTRYTHERMOBAROMETRYREACTION HISTORY AND PHASE...P-T-t PATH FOR KEMP...CONCLUSIONSSUPPLEMENTARY DATAREFERENCESREFERENCES IN SUPPLEMENTARY... |
|---|
Bhattacharya, A., Krishnakumar, K. R., Raith, M. & Sen, S. K. (1991). An improved set of a–X parameters for Fe–Mg–Ca garnets and refinements of the orthopyroxene–garnet thermometer and the orthopyroxene–garnet–plagioclase–quartz barometer. Journal of Petrology 32, 629–656.
Bohlen, S. R., Wall, V. J. & Boettcher, A. L. (1983a). Experimental investigation and application of garnet granulite equilibria. Contributions to Mineralogy and Petrology 83, 52–61.[CrossRef][Web of Science]
Bohlen, S. R., Wall, V. J. & Boettcher, A. L. (1983b). Experimental investigations and geological applications of equilibria in the system FeO–TiO2–Al2O3–SiO2–H2O. American Mineralogist 68, 1049–1058.[Abstract]
Carswell, D. A. & Harley, S. L. (1989). Mineral thermometry and barometry. In: Carswell, D. A. (ed.) Eclogite Facies Rocks. Glasgow: Blackie, pp. 83–110.
Eckert, J. O., Newton, R. C. & Kleppa, O. J. (1991). The H of reaction and recalibration of garnet–pyroxene–plagioclase–quartz geobarometers in the CMAS system by solution calorimetry. American Mineralogist 76, 148–160.[Abstract]
Harley, S. L. (1984). An experimental study of the partitioning of Fe and Mg between garnet and orthopyroxene. Contributions to Mineralogy and Petrology 86, 359–373.[CrossRef][Web of Science]
Lee, H. Y. & Ganguly, J. (1988). Equilibrium compositions of coexisting garnet and orthopyroxene—experimental determinations in the system FeO–MgO–Al2O3–SiO2, and applications. Journal of Petrology 29, 93–113.
Moecher, D. P., Essene, E. J. & Anovitz, L. M. (1988). Calculation and application of clinopyroxene–garnet–plagioclase–quartz geobarometers. Contributions to Mineralogy and Petrology 100, 92–106.[CrossRef][Web of Science]
Newton, R. C. & Perkins, D. (1982). Thermodynamic calibration of geobarometers based on the assemblages garnet–plagioclase–orthopyroxene (clinopyroxene)–quartz. American Mineralogist 67, 203–222.[Abstract]
Wood, B. J. (1974). The solubility of alumina in orthopyroxene coexisting with garnet. Contributions to Mineralogy and Petrology 46, 1–15.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G.A. Belyanin, H.M. Rajesh, D.D. Van Reenen, and H. Mouri Corundum + orthopyroxene {+/-} spinel intergrowths in an ultrahigh-temperature Al-Mg granulite from the Southern Marginal Zone, Limpopo Belt, South Africa American Mineralogist, January 1, 2010; 95(1): 196 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Phillips, D. E. Kelsey, A. F. Corvino, and R. A. Dutch Continental Reworking during Overprinting Orogenic Events, Southern Prince Charles Mountains, East Antarctica J. Petrology, November 1, 2009; 50(11): 2017 - 2041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kawakami, E. S. Grew, Y. Motoyoshi, C. K. Shearer, T. Ikeda, P. V. Burger, and I. Kusachi Kornerupine sensu stricto associated with mafic and ultramafic rocks in the Lutzow-Holm Complex at Akarui Point, East Antarctica: what is the source of boron? Geological Society, London, Special Publications, January 1, 2008; 308(1): 351 - 375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Halpin, R. W. White, G. L. Clarke, and D. E. Kelsey The Proterozoic P-T-t Evolution of the Kemp Land Coast, East Antarctica; Constraints from Si-saturated and Si-undersaturated Metapelites J. Petrology, July 1, 2007; 48(7): 1321 - 1349. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. ULIANOV and A. KALT Mg-Al Sapphirine- and Ca-Al Hibonite-bearing Granulite Xenoliths from the Chyulu Hills Volcanic Field, Kenya J. Petrology, May 1, 2006; 47(5): 901 - 927. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||