Journal of Petrology | Volume 43 | Number 8 | Pages 1595-1616 | 2002
© Oxford University Press 2002
Experimental Melting of Cordierite Gneiss and the Petrogenesis of Syntranscurrent Peraluminous Granites in Southern Brazil
1PROGRAMA DE PÓS-GRADUAÇÃO EM GEOCIÊNCIAS, UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL, PORTO ALEGRE, 91509-900, RS, BRAZIL
2DEPARTMENT OF EARTH SCIENCES, UNIVERSITY OF MANCHESTER, MANCHESTER M13 9PL, UK
3CENTRO DE PETROLOGIA E GEOQUÍMICA, INSTITUTO DE GEOCIÊNCIAS, UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL, PORTO ALEGRE, 91509-900, RS, BRAZIL
Received August 5, 2000; Revised typescript accepted February 27, 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
To understand the petrogenesis of peraluminous granites syntectonic to the Dorsal de Canguçu Transcurrent Shear Zone in the Sul-rio-grandense Shield, Brazil, melting experiments were performed on one of the potential protoliths, a cordierite-bearing semi-pelitic metasedimentary gneiss (PE-1). Experiments were conducted at pressures of 5, 10 and 15 kbar, at temperatures of 700–900°C, and under fluid-absent and 5% H2O-present conditions. The experiments show that fluid-absent melting begins at near-solidus conditions, around 700°C, promoted by participation of retrogressive phengitic muscovite in the reaction Mus + Kf ± Qz = melt ± Fe–Ti oxide ± Als, producing a very small amount of melt (<9%) with widely ranging composition. All hypersolidus experiments (>800°C) produced S-type granitic melts promoted by participation of biotite or cordierite in the reactions Bio + Pl + Crd + Qz = Px + Fe–Ti oxide + melt at 5 kbar, and Bio + Pl + Crd ± Qz = Grt + Als ± Kf + melt at 10 and 15 kbar, both producing a high amount of melt (10–63% by volume). The melt compositions obtained at 900°C and 15 kbar under fluid-absent conditions, promoted by biotite or cordierite breakdown, are similar to the syntectonic granites. However, it is unlikely that the granites were formed at this pressure (corresponding to a depth of melting of
54 km). KEY WORDS: cordierite gneiss; fluid-absent and H2O-present melting experiments; peraluminous granites; shear zone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
The last two decades have seen an increasing interest in processes of melt production from mantle or crustal sources, extraction, transport, accumulation and storage in the crust (Clemens & Vielzeuf, 1987; Brown et al., 1995b). As well as the interest in magmatic processes, the role played by structures, such as shear zones, in collecting and transporting melts from their original sources to mid- and upper-crustal environments, has been an important field of study (Clemens & Mawer, 1992; D’lemos et al., 1992). In addition, melt production, transport and evolution in high-grade metamorphic terrains are important issues in the study of migmatites (Brown et al., 1995a; Thompson & Connolly, 1995), as is the possibility of segregation of this melt to produce plutonic bodies (Van der Molen & Paterson, 1979; Brown, 1994; Vigneresse et al., 1996).
The experimental petrological study of reactions to elucidate the processes that generate melts has contributed to the understanding of the petrogenesis of granitoids, and volcanic and metamorphic rocks, and has allowed the testing of different geological models, and the delineation of physical and chemical constraints. Tuttle & Bowen (1958) were among the first to perform experiments to understand the evolution of natural granitic rocks. Since then, a substantial number of papers have been published on the melting behaviour of the crust and mantle, using experimental petrology as the main tool. Until the mid-1980s the main problems investigated were the behaviour of minerals under different conditions of P and T (e.g. Storre, 1972; Johannes, 1984), and the location of solidus curves for synthetic and natural systems (e.g. Wyllie, 1977). Following this period, experiments were carried out to test different models of magma generation in specific geological areas (e.g. Le Gardien et al., 1995; Patiño-Douce & Beard, 1995), trying to explain melt productivity in terms of the effects of variables such as oxygen fugacity, water activity [i.e. fluid-absent vs (H2O, CO2)-fluid-present melting experiments], chemical composition of key meltable minerals (e.g. TiO2 content of biotite) and the protolith mineral assemblage (e.g. presence or absence of tourmaline in the starting material).
The main lithologies used as starting material in experiments on crustal melting are rocks with high proportions of hydrated minerals, such as muscovite, biotite and hornblende, which can produce high proportions of melt through fluid-absent melting reactions (e.g. Clemens & Vielzeuf, 1987). Such reactions are accepted as the most likely way to melt rocks in the crust, as the low porosity and permeability of rocks in the crust generally prevents the storage and/or influx of large volumes of aqueous fluid. However, melting reactions involving H2O as a reactant have been tested and sometimes accepted as a possible mechanism for producing melts, in general related to initial melting of sedimentary rocks in collisional tectonic settings (e.g. Patiño-Douce & Harris, 1998). Thus, model or natural metapelites, quartzo-feldspathic rocks (metagreywackes and tonalites) and amphibolites have been the rocks most commonly tested under various fluid-absent melting conditions, mainly because of the high modal abundance of muscovite, biotite or amphibole in these three rock types, respectively.
Reactions in which muscovite in large proportions is the main hydrated phase were investigated by Patiño-Douce & Harris (1998), using a muscovite schist. These experiments produced a peraluminous granitic melt compositionally similar to Himalayan leucogranites. Muscovite- and biotite-bearing rocks have been used in a number of experiments in which biotite was the main hydrated mineral (e.g. Patiño-Douce & Johnston, 1991; Pickering & Johnston, 1998); in general, the muscovite provided a small volume of melt, whereas the main melt-producing reaction (quantitatively speaking) involved the breakdown of biotite and the production of peraluminous granitic melts. There have been a large number of experiments that used only biotite as the hydrated phase. For example, experiments using greywacke compositions were performed by Le Breton & Thompson (1988), Le Gardien et al. (1994), Vielzeuf & Montel (1994), Patiño-Douce & Beard (1995, 1996), Stevens et al. (1997) and Montel & Vielzeuf (1997); experiments using tonalitic compositions were performed by Singh & Johannes (1996a, 1996b) and Soares et al. (1998). These studies show that the breakdown of biotite occurs around 830°C, producing high proportions of melt of peraluminous granitic composition. Amphiboles have been used mainly in experiments involving basaltic or amphibolitic rocks, for example those described by Wyllie & Wolf (1993), Patiño-Douce & Beard (1995) and Rapp & Watson (1995), and in a few experiments involving greywackes or tonalitic rocks (e.g. Carrol & Wyllie, 1990; Wolf & Wyllie, 1994). The melts produced have trondhjemitic to tonalitic compositions, with metaluminous to mildly peraluminous affinities and even liquids with mild peraluminosity (Wolf & Wyllie, 1994). A few experiments have been carried out using rocks containing other hydrous minerals, such as chlorite or staurolite (Vielzeuf & Holloway, 1988; Patiño-Douce & Harris, 1998).
Cordierite, a nominally hydrous phase, is a common mineral in restitic assemblages of migmatitic rocks produced by high-grade metamorphic events (e.g. Stevens et al., 1995). Fluid-absent melting experiments on metapelites and metagreywackes have also confirmed that cordierite is an important restitic mineral (Vielzeuf & Montel, 1994; Stevens et al., 1997). However, there has been no experimental study in which cordierite was one of the main phases present in the starting material.
In the Sul-rio-grandense Shield, cordierite is present in high-grade metasedimentary rocks belonging to the Varzea do Capivarita Metamorphic Suite (VCMS). The latter consists of high-grade forsterite marbles, calc-silicate gneisses, pyroxene–biotite gneisses and migmatitic garnet–cordierite gneisses, which were metamorphosed to granulite-facies conditions during the Proterozoic. Garnet–biotite and garnet–cordierite thermometry yield peak-metamorphic temperatures between 770 and 840°C, consistent with the formation of garnet- and cordierite-bearing leucosomes by the incongruent fluid-absent melting of biotite + sillimanite + quartz (Silva, 1999). These melts were segregated as centimetre-scale quartzo-feldspathic layers, rather than being extracted as large bodies. This first melting event (2·0 Ga) pre-dates generation of the Dorsal de Canguçu Transcurrent Shear Zone (650 Ma), a transcontinental strike-slip shear zone that crosscuts the cordierite-bearing metasedimentary rocks. A second melting event led to the emplacement of syntectonic peraluminous granitoids. These provide an opportunity to investigate the relationships between melt productivity and crustal evolution.
The principal aim of this study was to test the possibility that the high-grade semi-pelitic migmatites, which crop out in the region, and are retrometamorphosed to low grade, are the protolith of the peraluminous granites. We have therefore performed experiments in which cordierite is a major phase in the starting material, and have compared our results with previous experimental data from the literature.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
The continental crust of SW Gondwana is characterized by cratonic areas surrounded by Neoproterozoic deformation belts where the most impressive structures observed are continental-scale transcurrent shear zones (Fig. 1). The Dorsal de Canguçu Transcurrent Shear Zone (DCTSZ) is part of a post-collisional fault system that trends NE–SW, parallel to the Dom Feliciano Belt in Southern Brazil, along which tectonites of middle and upper crust were developed and magmas of several ages emplaced (Fig. 1c).
|
The early syntectonic magmatism in the DCTSZ is represented by the porphyritic Quiteria metagranite (mgQ) dated at 631 ± 6 Ma by conventional U–Pb age determination on zircons (Koester et al., 2001a). The mgQ is characterized by compositions ranging from granodiorite to monzogranite, with subordinate syenogranite. It displays a high-K calc-alkaline affinity and SiO2 contents varying from 64 to 71 wt %, low TiO2 + FeO* + MgO (5·8 wt %), high Al2O3 (A/CNK 1·1), low Ba and Sr contents and strongly fractionated rare earth element (REE) patterns (Lan/Lun = 20·87 and Lun = 6·93 on average). Major and trace element compositions generally show linear correlations on Harker diagrams. The mineralogy, enclaves and chemistry give the granitoid an I-type signature, whereas the initial Sr isotope composition (87Sr/86Sr = 0·716) and low Zr and V contents suggest a mixed origin involving crustal and mantle sources. Field, petrographic and geochemical evidence indicates that assimilation plus fractional crystallization (AFC) was the dominant mechanism of granitoid evolution. Microdioritic enclaves represent the parental magma, but crustal assimilation of orthogneiss contributed to the hybrid calc-alkaline high-K granitoid genesis (Fernandes & Koester, 1999). Sm–Nd model ages of 2·0 Ga and 
Nd - 7 suggest an involvement of Palaeoproterozoic continental crust in the generation of the granitoids (Frantz et al., 1999).
The late magmatism, which was syntectonic to the DCTSZ, corresponds to the emplacement of leucogranites of the Cordilheira Granitic Suite (CGS). The Arroio Francisquinho metagranite (mgAF) is the older and the Cordilheira metagranite (mgC) the younger, according to field structural relations and mineralogical and chemical characteristics. Rb–Sr ages of 629 ± 23 and 617 ± 48 Ma (Koester et al., 1997) have been obtained for these two metagranites, respectively. The CGS granitoids have compositions ranging from monzogranite to syenogranite, with subordinate granodiorite. They are strongly peraluminous (two-mica granites), characterized by SiO2 contents varying from 70 to 75 wt %, low contents of TiO2 + FeO* + MgO (<2 wt %) and high Al2O3 (A/CNK >1·2, <4·5). The mgAF has high contents of Sr and Ba, and low Rb in relation to the mgC. Both granitoids display weakly fractionated REE patterns; mgAF has Lan/Lun = 11·24 and Lun = 4·58, and mgC has Lan/Lun = 9·86 and Lun = 5·05, on average. In most cases, the samples of peraluminous granitoids show random trends in major and trace element variation diagrams and normalized trace element plots. The CGS granites have a consistent S-type signature as demonstrated by their mineralogical and chemical compositions, with high SiO2, normative corundum (>2%) and initial 87Sr/86Sr isotope ratios of 0·732 (mgAF) and 0·740 (mgC). These facts suggest crustal melting processes as the dominant mechanism operating in the evolution of the suite. On the other hand, field relations, enclave typology and petrological and chemical characteristics do not indicate an S-type source for either mgAF or mgC. The generation of these rocks may be explained by melting of crustal rocks such as paragneiss or orthogneiss, or even a mixed source such as orthogneiss containing xenoliths of metasedimentary rocks (Fernandes & Koester, 1999). Values of Nd between -5 and -19 and TDM of 2·0 Ga, obtained by Frantz et al. (1999), indicate the influence of Palaeoproterozoic sialic crust in the evolution of this suite.
The host rocks of the DCTSZ are orthogneisses and paragneisses, which record upper-amphibolite facies metamorphism and partial melting. The partial melting episode pre-dates the nucleation of the transcurrent shear zones, as indicated by the location of melt pockets along pressure shadows and boudin-necks and other extensional structures observed in these rocks. The paragneisses are roof pendants with outcrop in areas of tens of km2, in orthogneisses of tonalitic to trondhjemitic composition, which display the same upper-amphibolite-facies metamorphism and similar extensional structures but without any melt segregation (Fernandes & Porcher, 2000). The rocks record an east–west direction of tectonic transport as marked by sillimanite lineations and garnet porphyroblasts in the paragneisses, and feldspar lineations in the orthogneisses. Also present is a subhorizontal tectonic foliation, registered by alternating millimetre-scale mafic and felsic bands. These fabrics are clearly older than the nucleation of the DCTSZ, which contains magmatic structures and tectonic lineations and subvertical planar structures with NE–SW trends.
Radiometric ages of the host rock units are older than the ages of the DCTSZ syntectonic granitoids. The paragneisses, which are probably related to a marginal basin (Fernandes et al., 1995), are older than 2·0 Ga. The orthogneisses, which have been ascribed to an active margin magmatic arc environment (Fernandes et al., 1995), have a U–Pb SHRIMP age of 2·078 ± 13 Ga (Leite et al., 2000). This age places the arc and the most pervasive fabrics of the paragneisses into the Palaeoproterozoic. The older ages of the porphyritic Quiteria metagranite in relation to the peraluminous granites, and their close spatial distribution along the entire length of the shear zone, suggest a genetic relationship between these rocks. Tommasi et al. (1994) suggested a mechanism of advective heat transfer as a result of emplacement of the porphyritic granodiorites, triggering partial melting of the host rocks.
|
EXPERIMENTAL AND ANALYTICAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
Starting material
Experiments were performed using a natural metasedimentary migmatite from the Sul-rio-grandense Shield. The rock, a cordierite gneiss, is a potential protolith for the peraluminous granite because it occurs as mega-enclaves (
2 km) in the orthogneiss that surrounds the peraluminous granites. The sample (PE-1) was collected from a mega-enclave of this migmatite, which crops out 3 km west of the DCTSZ. The mineral assemblage comprises 59% feldspars and quartz, plus cordierite, garnet, biotite, retrogressive minerals (phengitic muscovite) and accessory minerals (apatite, zircon, pyrite and spinel). Bulk-chemical, modal and mineral compositions are given in Table 1.
|
The composition of PE-1, plotted on an ACF diagram (Fig. 2, Table 2), is closer to the metagreywackes used in previous experiments than to the metapelites. The significant proportion of cordierite, without the presence of aluminosilicate, argues in favour of this, but the substantial proportions of feldspars and micas suggest a semi-pelitic composition. No previous study has used a starting material containing cordierite together with biotite and muscovite, or one with a similar composition. In PE-1, biotite has an mg-number of 51 and a high TiO2 content (3·08%), muscovite is phengitic, with mg-number of 65 and without TiO2, and the cordierite is classified as magnesian, with an mg-number of 65. The plagioclase is sodic (An40) and unzoned; the garnet is almandine, and is zoned, with rims that show higher Ca and Mg contents and lower Mn contents than the cores.
|
|
Around 4 kg of the rock was crushed to an average grain size of 1 cm, then quartered (1 kg), ground to 115 mesh, then quartered to 10 g of material. This was ground under acetone in an agate mortar to 20 µm. Some large grains of biotite of 30 µm remained. Aliquots of 15 mg of sample were loaded into 7 mm long, 3 mm diameter gold tubes of 0·15 mm wall thickness, which were used as capsules in all experiments. In the fluid-absent melting experiments, capsules with one end already welded were placed in the drying oven at 120°C for at least 1 h to eliminate adsorbed water. After loading the starting material, the capsule was again dried for 1 h, and then crimped and arc-welded shut. In the experiments performed with added water, the appropriate amount (5%) of deionized H2O was added with a microsyringe to the empty dried capsule, then the solid starting material was added, and immediately the capsule was crimped and arc-welded shut. Weight loss was monitored during welding for both kinds of capsule, and when water was added, the capsules were also checked for leaks by immersing them in silicone oil at 100°C.
Apparatus
Experiments were performed at temperatures from 700 to 900°C in an internally heated gas pressure vessel (IHGPV) at 5 kbar, and in solid-media non-end-loaded piston-cylinder (PC) apparatus at 10 and 15 kbar (Table 3), at the Department of Earth Sciences, University of Manchester.
|
The PC experiments used assemblies of 1*27 cm (1/2 inch) diameter comprising an outer cylinder of pressed NaCl, plus a glass sleeve in experiments at T
850°C, surrounding the graphite furnace. The capsule was placed on a solid ceramic rod in the lower part of the furnace surrounded by powdered ceramic to avoid empty spaces around the capsule, and then covered with a ruby disc (0·6 mm) to prevent puncturing of the capsule by the thermocouple during the experiment. The upper part of the furnace was filled with a ceramic tube through which the Pt–PtRh thermocouple passes. Temperature was controlled by a Eurotherm 815 temperature controller, which maintained a temperature stability throughout all runs of better than ±3°C. Pressure was calibrated against the jadeite–albite–quartz and quartz–coesite transitions, and controlled by a hydraulic pressure controller. The overall uncertainty in pressure is probably less than ±0·5 kbar. No correction was made for the effect of pressure on thermocouple e.m.f. Run duration varied from 61 to 290 h and experiments were quenched by turning off the power, which caused temperatures to drop to close to 3°C in 30 s.
The IHGPV experiments used argon as the pressure medium. Capsules were packed together with silica wool inside the sample holder between the two chromel–alumel thermocouples, which were spaced 8 mm apart. Temperature was controlled using a Eurotherm controller with variation less than ±2°C. Thermal gradient was minimized by tilting the pressure vessel. The pressure gauge was calibrated against the melting point of Hg, and variability was ±0·4 kbar. Experiment durations varied from 168 h at the highest temperature to 428 h at the lowest temperature. Runs were quenched by turning off the power to the apparatus, whereupon temperatures dropped to 300°C in around 90 s.
After the experiments, the presence of excess H2O in the H2O-present experiments was confirmed by puncturing the capsule and measuring the weight loss after oven drying.
Observation and analysis of run products
Successful experimental products were sliced, mounted in epoxy resin, polished and carbon coated for use in electron-microprobe and scanning electron microscopy (SEM) studies. Analyses were obtained using a JEOL-JSM 6400 electron microprobe fitted with an SEM system, at the Department of Earth Sciences, University of Manchester. Analyses of minerals and glasses used an acceleration voltage of 15 kV, probe current of 15 nA, beam size of 1 µm, and a beryllium-window detector. Analyses of glasses were made using an Oxford Instrument Cryo-stage, which uses liquid nitrogen to hold the temperatures around -192°C, minimizing counting losses related to diffusion of some elements (e.g. Na, K). This approach has been used before with good results (e.g. Stevens et al., 1997), although other strategies for avoiding alkali loss in glasses, such as defocused beam or scanning over a large area, have also been used with good results (e.g. Le Gardien et al., 1995; Patiño-Douce, 1996).
Mineral composition data for the experimental run products are given in Electronic Appendices 1–10, which may be downloaded from the Journal of Petrology Web site at http://www.petrology.oupjournals.org.
Oxygen fugacities
Variations in oxygen fugacity may be expected to have an effect on our results, and so it is important to measure the oxygen fugacities imposed by the experimental apparatus. The graphite-based cell assemblies used in the piston-cylinder apparatus restrict the log fO2 in samples to between QFM (quartz–fayalite–magnetite) and QFM- 2 (Patiño-Douce, 1996). In the IHGPV, conditions are more oxidized (Patiño-Douce, 1996). Using an empirical calibration, Patiño-Douce (1993) estimated the oxygen fugacities for experiments that contained the mineral pair biotite–ilmenite or biotite–pyroxene. However, orthopyroxene and biotite or ilmenite and biotite crystallized in only a few of our experiments at 5 kbar, preventing either of these calibrations from being generally useful. Most of our runs at 10 and 15 kbar contain the product assemblage garnet + sillimanite + quartz. Following Pickering & Johnston (1998), we can write the following oxygen buffer:
 |
Application of this equilibrium is hampered by the difficulty of obtaining accurate estimates of the Fe3+ contents of garnets. Stoichiometric criteria (Droop, 1987) suggest that less than 2% of the octahedral sites contain Fe3+; for the most Fe3+-rich garnets (garnet PE1-26, rim; Electronic Appendix 7) this yields an andradite activity of 5·6 x 10-6, assuming ideal ionic mixing. Employing ideal mixing for other garnet components, calculation of equilibrium (A) using THERMOCALC (Powell & Holland, 1988, 1990, 1998) yields an oxygen activity of 10-14 for the piston-cylinder run PE1-26 (800°C and 15 kbar), which is comparable, within error, with the magnetite–wüstite buffer.
Mode calculation
Modal abundances in the starting material and experimental products were calculated using back-scattered electron (BSE) modal analysis of SEM images, and integrating these data with mass balance equations. Use of mass balance equations was necessary because the difficulty of distinguishing some phases that show very little contrast in BSE images (e.g. cordierite and feldspars) precludes an accurate modal estimate, whereas other phases (e.g. garnet and biotite) can be reasonably estimated. An average of three representative images were obtained in each experiment, and modal abundances of the well-constrained phases were used as an additional constraint to solve the mass balance equations (Patiño-Douce & Johnston, 1991; Stevens et al., 1997). Table 4 presents the results, using the maximum possible number of mineral phases in equations, so as to minimize the error.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
Modal abundances as a function of temperature at each studied pressure are shown for fluid-absent (Fig. 3a–c) and fluid-present (Fig. 3d and e) melting conditions. Quartz, plagioclase and glass are present in all experimental run products. Muscovite occurs in experiments at low temperature. Biotite appears at all pressure and temperature conditions, except 900°C at 5 kbar. Cordierite occurs only at 5 and 10 kbar at temperatures below 800°C under fluid-absent conditions, and also at 900°C at 10 kbar under H2O-present conditions. Amphibole occurs only at 15 kbar and orthopyroxene only at 5 kbar and under fluid-absent conditions. Garnet and aluminosilicate occur only at 10 and 15 kbar. K-feldspar occurs only at 900°C at 15 kbar, and Fe–Ti oxides are present only in experiments at 5 kbar. BSE images of selected run products are shown in Fig. 4. The presence of Fe–Ti oxides and a larger proportion of melt at low pressure (5 kbar, Fig. 4a) than high pressure (10 and 15 kbar, Fig. 4b–f) should be noted. Figure 4b and c shows the low melt proportion at low temperature (700°C and 850°C, respectively), and the increasing modal abundance of garnet with rising temperature. Figure 4d shows the interconnected melt and the presence of zoned garnet, amphibole and a small amount of biotite. Figure 4e and f shows results of experiments performed under fluid-present conditions, at 10 and 15 kbar, respectively; there is more melt in these experiments than in those without added water at the same P and T (Table 4). Results of all experiments have been combined to produce pressure–temperature diagrams (Fig. 5), in which ‘in’ and ‘out’ curves show the mineral stability limits, and fields of interstitial melt, pockets of melt and interconnected melt are also shown.
|
|
|
In near-solidus (
750°C) experiments glass (quenched melt) occurs between minerals, typically as a very thin film (<2 µm) around minerals or at triple junctions. In hypersolidus (800°C) experiments pools of glass up to 20 µm across are produced. The melt composition is variable in experiments at low temperature (750°C) but becomes uniform with rising temperature, suggesting interconnectivity of the melt pockets with rising temperature and an increasing approach to equilibrium on a millimetre scale.
Quartz is present in all run products under fluid-absent and fluid-present conditions, as globular grains (<20 µm). Its abundance and grain size decrease with rising temperature.
Biotite is a very common product phase, in both fluid-absent and fluid-present runs, and forms small euhedral crystals, around 5 µm long and 3 µm wide. Its modal abundance decreases with rising temperature. It is present in all experiments at 10 and 15 kbar; however, at 5 kbar, under fluid-absent conditions, it occurs only at 700–800°C. This suggests that the biotite-out curve is positioned above 900°C at 10 and 15 kbar, but below 900°C at 5 kbar. In the fluid-present experiments, the biotite-out reaction is located above 900°C for all pressures. Some large anhedral biotite flakes 20 µm long observed in low-temperature experiments are remnants of the starting material. The mg-number and TiO2 of biotite increase with rise in temperature at constant pressure, and in general are lower under fluid-present conditions than fluid-absent conditions (see Electronic Appendix 1).
Plagioclase is abundant in all experimental run products and can be observed as subhedral (<10 µm) crystals, sometimes with relict cores preserved. It is zoned and sometimes has a sponge-like texture, with melt inclusions. With rising temperature at all pressures, the Or content increases, and the Ab and An contents slightly decrease, although remaining within the andesine range. At 5 kbar, under fluid-absent conditions, plagioclase rims at 800 and 900°C are labradorite, i.e. more An rich than the cores (see Electronic Appendix 2).
Cordierite occurs under fluid-absent conditions only at 5 kbar, 700°C and 10 kbar, 800°C, and under H2O-present conditions at 5 kbar, 900°C and 10 kbar, 800°C. The abundance of cordierite decreases with rising temperature and pressure. It forms large (10 µm) generally anhedral, relict grains. Cordierite compositions from experiments, when compared with starting material, show a small variation in mg-number with rising temperature (see Electronic Appendix 3).
Amphibole is present only in experiments at 15 kbar, under fluid-absent and fluid-present conditions. It forms unzoned prismatic crystals (locally containing small inclusions of glass), which measure approximately 10 µm by 5 µm in low-temperature experiments, and decrease in size with rising temperature. It has a high Al2O3 content that increases with temperature rise, and high Na2O and CaO contents. In the highest-temperature experiments (850°C) the amphiboles have high FeO and MgO contents and are classified as gedrites, whereas at low temperatures (800°C) they are calcic (Na + K < 0·5; Ti < 0·5) and can be classified as Fe-tschermakitic hornblende (see Electronic Appendix 4).
Orthopyroxene is present only at 5 kbar and temperatures of 800 and 900°C under fluid-absent conditions. It forms prismatic crystals, approximately 5 µm long by 2 µm wide. Its mg-number increases with temperature rise (56 at 800°C and 66 at 900°C). It has a high Al2O3 content (8·8%) and the Wo–En–Fs contents suggest that it is enstatite (see Electronic Appendix 5).
Muscovite occurs as a relict mineral at 10 kbar, 700°C and at 15 kbar, 800°C under fluid-absent conditions, and at 15 kbar, 800°C under fluid-present conditions. It forms small (2 µm) anhedral grains. The muscovite in the starting material contains no significant TiO2, but apparent TiO2 contents in muscovite from experimental run products are as high as 1·8%. Other oxide contents also change from the starting material, mainly K2O, MgO and FeO, with lower mg-number at low-temperature conditions (see Electronic Appendix 6).
Garnet is the main phase produced in all experiments at 10 and 15 kbar, under fluid-absent and fluid-present conditions. It forms rounded crystals (<15 µm) showing a sponge-like texture, with inclusions of glass and quartz. Its modal proportion increases with temperature rise. Common features of the garnets are new rims formed around seed crystals, which occur in the highest-temperature experiments (800°C) at 10 kbar and at all temperatures at 15 kbar. The cores are relicts of the starting material, whereas the rims are products of crystallization. The rims in general have higher pyrope and grossular and lower almandine and spessartine contents than the cores. Compositions of rims from fluid-absent and fluid-present runs are similar (see Electronic Appendix 7).
Aluminosilicate is produced in the experiments only at 10 and 15 kbar. It occurs as small acicular to prismatic crystals (2 µm long and 1 µm wide), and appears at 10 kbar, 750°C and 15 kbar, 800°C under fluid-absent conditions, and at 800°C under H2O-present conditions. It invariably has a composition close to Al2SiO5 (see Electronic Appendix 8). The needles all appear to have straight extinction and some square basal sections, suggesting that they are sillimanite rather than kyanite. However, the stable polymorph at the pressure–temperature conditions is kyanite, and so the sillimanite must be metastable.
Despite the presence of K-feldspar in small proportions in the starting material, it was found in run products only at 900°C and 15 kbar, under fluid-absent and fluid-present conditions, suggesting that it participates in the reactions. It forms small subhedral crystals (2 µm) with low An content (see Electronic Appendix 9). Stevens et al. (1997), whose experiments used metapelites without K-feldspar in the starting material, found this mineral in experiments only above 900°C at 10 kbar.
Three kinds of Fe–Ti oxides occur in run products, but only at 5 kbar: ilmenite, ulvöspinel and magnetite. They form small (<10 µm) euhedral neoblasts. Under fluid-absent conditions magnetite coexists with either ilmenite or ulvöspinel. Under fluid-present conditions magnetite was not identified, and a coexistence of ilmenite and ulvöspinel was observed (see Electronic Appendix 10).
Accessory phases include pyrite, apatite and zircon. They are present in proportions lower than 0·01%, probably representing relict phases from the starting material.
|
MELTING REACTIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
Melt production in experiments without added water is related to breakdown of hydrated minerals, and in fluid-present runs to the influence of H2O and/or breakdown of hydrated minerals. Melt production in all of the fluid-absent and fluid-present experiments can be subdivided into the following stages: (1) near-solidus conditions; (2) hypersolidus conditions.
The first melt production stage, at near-solidus conditions, is responsible for low proportions of melts, produced by the breakdown of muscovite in fluid-absent conditions. This involves reaction of phengitic muscovite, with high water content, that was present in the starting material as a retrogressive phase and the influx of H2O in fluid-present conditions. The reaction occurs at low temperatures (750°C), and produces a small amount of melt (Table 4), <4% at 10 and 15 kbar (750°C) and <9% at 5 kbar (700°C), through the reaction
 |
Reaction (1) is suggested by the presence of very thin films of melt around the minerals or at triple junctions at low temperatures (<800°C) and the presence of aluminosilicate at 750°C at 10 kbar, and oxides at 700°C at 5 kbar. Reaction (1) is similar to that suggested by Patiño-Douce & Johnston (1991) for the breakdown of muscovite at 10 kbar, 825°C (Mus + Pl + Qz = melt + Als).
The second melt production stage, at hypersolidus conditions, which is responsible for high proportions of melts, occurs by the breakdown of biotite. To understand the melting reactions that occur in hypersolidus conditions, experiments were subdivided into fluid-present and fluid-absent sets, and up-temperature isobaric sequences at 5, 10 and 15 kbar considered in each set.
At 5 kbar, under fluid-absent conditions, the breakdown of biotite begins between 700 and 800°C. Biotite is progressively consumed, until it is exhausted at 900°C, producing pyroxene and Fe–Ti oxides in the residual assemblage. The presence of cordierite in the experiment only at 700°C indicates that it was consumed by 800°C. Between 800 and 900°C the reaction involving biotite or cordierite fluid-absent breakdown is Bio + Pl + Qz = Px + Fe–Ti oxide + melt. This reaction involves mineral compositional change, as demonstrated in Electronic Appendix 5, where orthopyroxene shows an increase of MgO content with rising temperature.
The H2O-present melting experiment at 900°C produces an assemblage marked by cordierite + biotite + plagioclase + quartz + Fe–Ti oxide + melt. The absence of cordierite at this temperature in fluid-absent melting experiments shows that the H2O added to the experiment prevents the breakdown of cordierite up to at least this temperature. Also, the absence of orthopyroxene, an anhydrous phase, is probably related to the high amount of water in the experiment.
Fluid-absent melting experiments at 10 kbar show that an important reaction begins at around 800°C: Bio + Pl + Crd + Qz = Grt + Als + melt. This is one of the most important fluid-absent biotite-melting incongruent reactions for pelites. The presence of cordierite in the experiment only at temperatures lower than 800°C indicates that it was consumed by 850°C. At this pressure, the main variations in chemical composition of the phases are represented by the increasing MgO content of biotite, and the decreasing Al2O3 content of garnet (Electronic Appendices 1 and 7, respectively).
The stable assemblages in H2O-present experiments are, at 800°C, Bio + Grt + Qz + Pl + Crd + Als + melt, and at 900°C, the same assemblage without the presence of cordierite, showing that it was consumed between these temperatures. The reaction is the same as at 10 kbar in the fluid-absent experiments, showing that the added H2O does not change the main reaction. As demonstrated in Electronic Appendix 1, biotite has a chemical composition similar to that under fluid-absent conditions, with increasing MgO content at higher temperature, and the bulk composition shows a lower Al2O3 content at higher temperature.
The reaction occurring under fluid-absent conditions at 15 kbar is similar to that at 10 kbar, represented by Bio + Pl + Crd ± Qz = Grt + Als ± Kf + melt. The presence of hornblende in all experiments at 15 kbar might be explained by instability of cordierite at this pressure and production of amphibole, which decreases in modal abundance from 700 to 900°C. A cordierite-out curve is plotted in Fig. 5 between 10 and 15 kbar, in agreement with the stability field for this mineral proposed by Green & Vernon (1974) and Stevens et al. (1995) where Mg-cordierite is stable at medium pressure and temperatures, respectively, around 10 kbar and 800°C. The chemical variation of biotites and garnets does not show any correlation with temperature (Electronic Appendices 1 and 7, respectively).
The melting experiments under H2O-present conditions at both 800 and 900°C are represented by the stable assemblage Bio + Grt + Qz + Pl + Als + Amp + melt, which is similar to the assemblage under fluid-absent conditions.
Thus, at 5 kbar, the main melt production occurs at 800°C, promoted by the reaction
 |
At 10 and 15 kbar, the main melt production begins at 800°C, when new rims of garnet appear around garnet seeds from the starting material. The proposed reaction is
 |
Reactions (2) and (3) are similar to those suggested, respectively, by Patiño-Douce & Beard (1995) at pressures from 3 to 10 kbar and temperatures around 850°C (Bio + Pl + Qz = melt + Px + oxide), and Le Gardien et al. (1995), who suggested for the breakdown of biotite, at 10 kbar and temperatures between 825 and 900°C, the reaction Bio + Pl + Qz = Grt ± Als + melt. Melting reactions in fluid-present conditions seem to be similar to those in the fluid-absent conditions.
|
EXPERIMENTAL MELT AND PERALUMINOUS GRANITOIDS SYNTECTONIC TO THE DCTSZ |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
Melt generated from sample PE-1 contains, in almost all hypersolidus experiments, >68 and <75 wt % SiO2 (on a normalized H2O-free basis). Lower SiO2 contents occur in only two experiments (PE1-1 and PE1-22). The heterogeneous chemical compositions of melt in near-solidus experiments (<800°C, Table 5) probably reflect the lack of interconnectivity between melt volumes, and as result, a lack of chemical equilibrium. In terms of the Ab–An–Or contents, using the classification of Barker (1979), experimental melts are granitic in composition (Fig. 6). They are potassic and have a K2O/Na2O ratio varying between 1·2 and 5·8, and their mafic contents (TiO2 + MgO + FeO) span a wide range, from 1·5 to 12·3 wt %. All melts are peraluminous, i.e. 1·0 < A/CNK < 2·0 (Fig. 7), and are corundum normative (>2·4 and <5·9). In addition, the relationship between SiO2 and aluminium saturation index (ASI) suggests an S-type granitic affinity for hypersolidus melts (Fig. 8).
|
|
|
|
The production of S-type granitic melts from a semi-pelitic source rock containing cordierite as an important phase demonstrates that sample PE-1 is a potential source rock for production of peraluminous granites syntectonic to the Dorsal de Canguçu Transcurrent Shear Zone. Detailed examination of run products shows that the melt produced in experiment PE1-8, at 900°C, 15 kbar and under fluid-absent conditions is the one that best fits the bulk composition of natural peraluminous granitic rocks from this shear zone. Comparison of experimental and natural compositions plotted on SiO2 and CaO vs TiO2 + FeO + MgO diagrams (Figs 9 and 10) shows that sample PE1-8 is a reasonable fit, and this is also demonstrated in other diagrams such as the Ab–An–Or diagram (Fig. 6). Other experiments show some small discrepancies, such as the SiO2 content of PE1-22, or the CaO content of PE1-27, PE1-11 and PE1-5, whereas all other experiments show larger discrepancies when compared with the compositions of natural peraluminous granitic rocks.
|
|
The melts produced in this study from PE-1 display trends of increasing proportion of melt with rising temperature, and a slightly higher melt fraction in fluid-present experiments than in fluid-absent experiments at comparable P and T. An important point of discussion is the possibility that PE-1 represents the source rock for the peraluminous granites of the shear zone, with the composition being related to the critical melt fraction. In others words, is it possible that segregation of this melt material has produced granitic bodies? Some papers have discussed the problem of melt extraction (e.g. Van der Molen & Paterson, 1979; Brown, 1994). Vigneresse et al. (1996) have suggested that the minimum melt fraction necessary to promote percolation of melt from its source is around 8%, the liquid percolation threshold (LPT), whereas a plausible melt fraction that allows segregation and transport toward the upper crust, the melt escape threshold (MET), is around 20–25%. By these criteria, melts produced from PE-1 in fluid-absent experiments at 750°C at 5 and 15 kbar, and 800°C at 10 kbar, should not have been able to coalesce to produce granitic bodies, if the experimentally produced melt proportions are a true reflection of the natural system. It is also likely that in nature melt coalescence would have occurred well before the high melt fraction at 5 kbar and 900°C were reached, where >50% of melt was produced. However, all other fluid-absent pressure–temperature conditions produced melt fractions between 16 and 32%. Under H2O-present conditions, all experiments yielded higher melt fractions than the LPT of 8%, and only the experiment at 900°C and 5 kbar has a much higher melt content (63%) than the MET suggested by Vigneresse et al. (1996). Melt fractions in the other experiments range from 10 to 36%. Thus, in terms of melt fraction, experiment PE1-8 is a good fit (30%), as it is in terms of chemical composition as demonstrated above.
Melt productivity of PE-1 at various pressures, temperatures and water contents is shown in Fig. 11, in which it is compared with other melting experiments using different starting materials. A pressure control on melt production is not suggested in this study, although it was reported by Singh & Johannes (1996b) and Soares et al. (1998) from their experiments. In our experiments at 10 kbar the proportion of melt is lower than at 15 kbar, but at 5 kbar it is higher than at both 10 and 15 kbar. PE-1 appears to be more fertile (at 10 kbar and fluid-absent conditions) than the two-mica metapelite (HP-60-1) used by Pickering & Johnston (1998) and the biotite–plagioclase–quartz (BPQI) assemblage used by Le Gardien et al. (1995). Its fertility is similar to that of the metapelite (HQ-36) used by Patiño-Douce & Johnston (1991), and less than that of the model two-mica pelite (BPQM) used by Le Gardien et al. (1995). When compared with fluid-present experiments (Fig. 11) PE-1 appears to yield reasonable melt fractions; for example, Holtz & Johannes (1991) suggested a melt productivity around 40–60% at 5 kbar and 800°C, exactly the interval of melt fractions produced by PE-1 at fluid-absent conditions at 800°C (32%) and at fluid-present conditions at 900°C (63%).
|
The similarities in percentages of melting between the two sets of runs—generally <10% separates dry from wet melting—can be attributed to the high proportion of minerals containing water and/or carbon dioxide in the starting material. Biotite and muscovite contribute, respectively, 1·35% and 0·11% water to the starting material (these values of H2O are estimates determined by difference from 100%), which, added to the 0·20% of water or carbon dioxide from cordierite, represents around 1·66% of potentially important components bound up in the hydrous minerals in the experiments.
In HP-60-1 starting material (Pickering & Johnston, 1998) the water present bound in minerals was around 0·90%, and as shown in Fig. 11, the melt productivity compared with PE-1 was 15% lower. On the other hand, in the HQ-36 starting material from Patiño-Douce & Johnston (1991) the amount of water was around 1·70, and the melt productivity was similar to that of the PE-1 (Fig. 11). Thus, the high amount of hydrated minerals in PE-1 increases the melt fraction under fluid-absent conditions, and does not enhance melting much more under fluid-present conditions.
H2O loss from hydrous melts during fluid-absent experiments has been documented by Patiño-Douce & Beard (1994), where the manifestation of this loss is a decrease in melt fraction and increase of plagioclase with rise in temperature. In this case, however, the similarities between fluid-absent and fluid-present conditions, which could be interpreted as a loss of H2O in the fluid-present experiments, is not associated with an increase of plagioclase (Table 4) and, together with the systematic good agreement between fluid-absent and fluid-present conditions in PE-1 experiments, suggests this hypothesis is not responsible for the melt fraction similarities. Other experiments such as SBG from Patiño-Douce & Beard, at 925°C and 7 kbar and fluid-absent conditions, produce around 43·5% of melt fraction, only 10·5% lower than the same starting material at the same pressure–temperature but in fluid-present conditions (Patiño-Douce, 1996).
The question of fluid-absent vs fluid-present conditions in the generation of peraluminous granites syntectonic to the Dorsal de Canguçu Transcurrent Shear Zone should also be discussed. As suggested by Harris et al. (1993) for metasedimentary protoliths, melts formed during H2O-present melting will have lower Rb/Sr and higher Ba contents than those formed under fluid-absent conditions. In the peraluminous rocks from DCTSZ, two kinds of granites have been distinguished on the basis of the trace elements Rb, Sr and Ba (see section on Geological Setting), which on the basis of major elements have similar compositions. The Arroio Francisquinho metagranite has an Rb/Sr ratio of 1·86 and Ba content of 337 ppm (on average), and the Cordilheira metagranite has an Rb/Sr ratio of 4·00 and a Ba content of 181 ppm (on average). Using these values, it is possible to argue that the Arroio Francisquinho metagranite was formed in the presence of fluid, whereas the Cordilheira metagranite was formed in the absence of fluid.
In this study, the conditions that have come the closest to reproducing, in terms of major element compositions, the peraluminous granites of the DCTSZ by melting of PE-1, were at 15 kbar, 900°C and under fluid-absent conditions. On the basis of their trace element compositions it could be suggested that the PE-1 melts resemble the Cordilheira metagranite rather than the Arroio Francisquinho metagranite. However, the high pressure (15 kbar) of the melting experiments is unusual for metapelites, which in general are metamorphosed and deformed at lower-pressure conditions. Thermobarometric estimates of the Várzea do Capivarite Metamorphic Suite suggest that the metamorphic peak occurred at 770°C and moderate pressures, <5 kbar (Silva, 1999). If 15 kbar does represent the depth of melting, and if the overlying rocks in this region at the time had an average density of 2·85 g/cm3, corresponding to the normal density of continental crust, the melt would be formed at a depth in the crust of 54 km. The Cordilheira metagranite was emplaced at a mesozonal depth (>7 and <20 km; Hughes, 1982), and today it can be traced to a depth of 25 km in the shear zone, which geophysical studies show to be filled with peraluminous granites (Costa, 1998; Fernandes & Koester, 1999). Thus, to be compatible with our results the peraluminous granites would have to have been generated at depth of 33 km and therefore could not represent in situ magmatism. All of these arguments militate against the semipelites of the VCMS (of which PE-1 is an example) as having been the protolith of peraluminous granites (even the Cordilheira metagranite) syntectonic to the Dorsal de Canguçu Transcurrent Shear Zone.
If the semipelites of the VCMS do not represent realistic source rocks for the production of these peraluminous granites, a different mechanism and/or source rock must be inferred; for example: (1) anatectic melts represented by different melts from the same source; (2) anatectic melts produced from different sources; (3) anatectic melts produced under different pressure–temperature–fluid conditions.
Structural data suggest a waning deformation history between the emplacement of the Arroio Francisquinho and Cordilheira metagranites (Fernandes & Koester, 1999), and the absolute ages are compatible with a genetic link between them. The similarities between Rb–Sr whole-rock ages, 629 ± 23 Ma and 617 ± 48 Ma, for the Arroio Francisquinho metagranite and the Cordilheira metagranite, respectively (Koester et al., 1997; Fernandes & Koester, 1999), together with the ages of 631 ± 6 Ma (U–Pb conventional, Koester et al., 2001a) for the oldest syntranscurrent porphyritic granodiorite and 600 ± 10 Ma (U–Pb SHRIMP age, Koester et al., 2001b) for post-transcurrent alkaline magmatism, allow the timing of melting to be constrained to have occurred between 630 and 600 Ma.
Migmatitic metasedimentary rocks of VCMS, containing centimetre-scale segregations of melt, represent a regional melting event in the region (Silva, 1999). However, this melting event cannot be responsible for the production of peraluminous syntectonic granites, as these rocks are xenoliths in orthogneiss with magmatic ages of 2·0 Ga (Leite et al., 2000). Therefore a second melting event must be proposed for the melting and production of the large bodies of peraluminous granitic rocks.
U–Pb ages have demonstrated that, in this region, the orthogneissic rocks give a ‘resetting’ age around 630 Ma (631 ± 8 Ma, U–Pb SHRIMP, da Silva et al., 1999; 627 ± 6 Ma, U–Pb conventional, Koester et al., 2001a), interpreted as a resetting of zircons during the emplacement of the first magmas syntectonic to the DCTSZ. All these data indicate that there are at least two different melt production episodes in this region, the first at 2·0 Ga, and the second between 630 and 600 Ma.
If the peraluminous granites represent anatetic melts from different sources, then there is still the question of which rock in the region could represent a plausible protolith to the Arroio Francisquinho metagranite. Results of previous experimental studies may help to answer this question. These studies have demonstrated that peraluminous granitoid compositions can be formed by partial melting of rocks of a wide range of compositions, from metapelites through quartzo-feldspathic rocks (meta-granitoids and greywackes) to mafic rocks (basalts and amphibolites). Metasedimentary rocks, in particular metapelites, have been among the favourite starting materials in previous studies because of their high mica content, which promotes melting under fluid-absent conditions. Thus Vielzeuf & Holloway (1988) melted a metapelite containing staurolite and kyanite under fluid-absent conditions, and obtained 40 vol. % of S-type granitic melt between 850 and 875°C at 10 kbar, with a Qz + Grt + Sill + Pl residual assemblage. Le Breton & Thompson (1988) used a series of three metapelites containing kyanite as starting material. Under fluid-absent conditions at 10 kbar they found that melting began at 760°C and was extensive at 850°C, producing peraluminous liquids with heterogeneous compositions through breakdown of biotite. Le Gardien et al. (1995) used a starting material containing quartz, plagioclase, biotite and muscovite. They found that at 10 kbar, 40–60% of peraluminous granitic melt was formed between 800 and 900°C. Comparison of these experiments with those that did not involve the participation of muscovite in the starting material show that melt volume is increased by 15%, suggesting high fertility for rocks containing muscovite. However, despite the fact that melts produced in all of these experiments show some chemical similarities to the peraluminous granites of the DCTSZ (e.g. peraluminous and high SiO2), their mafic contents (TiO2 + FeO + MgO) are too high.
Using a muscovite schist and a muscovite–biotite schist in which muscovite was the main hydrous mineral, Patiño-Douce & Harris (1998) produced peraluminous granitic melt compositions under fluid-absent conditions. Despite a good chemical agreement with the natural peraluminous granites of the DCTSZ, these muscovite-bearing starting materials are not appropriate for the VCMS, in which the main hydrous mineral is biotite.
Granitic compositions produced in previous experiments on metapelites, which used biotite-bearing starting materials compatible with the VCMS, have been presented by Patiño-Douce & Johnston (1991), Stevens et al. (1997) and Pickering & Johnston (1998). All of these experiments demonstrated that peraluminous granitic compositions, with similar SiO2 and mafic contents to the syntectonic granites to the DCTSZ, can be generated from a starting material in which quartz + plagioclase + biotite are the main minerals, with or without garnet, muscovite, K-feldspar or aluminosilicate.
Greywackes and meta-granitoids in which biotite, plagioclase and quartz are the dominant minerals may also be possible protoliths for peraluminous granitic rocks. However, the compositions of experimentally produced melts are not always peraluminous, as demonstrated by the experiments of Le Gardien et al. (1995), in which a small amount of granitic melt of metaluminous affinity was obtained at 900°C and 10 kbar using a biotite gneiss (biotite + plagioclase + quartz) as starting material. Granitic melts with metaluminous to slightly peraluminous affinity were obtained by Finger & Clemens (1995), using a biotite gneiss starting material under H2O-present conditions. Experiments have also been conducted using starting materials containing quartzo-feldspathic assemblages with biotite as the main hydrous mineral, and granitic melts with peraluminous affinity have been produced (Holtz & Johannes, 1991; Patiño-Douce & Beard, 1995; Singh & Johannes, 1996a, 1996b). In all of these experiments, the melts have high SiO2 contents. However, their mafic contents are higher than those of the peraluminous granites syntectonic to the DCTSZ.
Previous studies on the melting of mafic rocks, mostly basalts and amphibolites, have yielded tonalitic, trondhjemitic or granodioritic and subordinately granitic compositions (Rapp et al., 1991; Wyllie & Wolf, 1993; Wolf & Wyllie, 1994; Winther, 1996). Only mildly peraluminous affinities have been described (Patiño-Douce & Beard, 1995), showing that mafic rocks are unlikely to be the protoliths of the granites syntectonic to the DCTSZ.
Metapelites and quartzo-feldspathic rocks are volumetrically abundant as host rocks of the DCTSZ peraluminous syntectonic granites, and they have also been the subject of previous experimental studies that have yielded similar melt compositions to these granites; this suggests that production of the DCTSZ peraluminous granites probably involved melting of a pelitic and/or a quartzo-feldspathic protolith with variable amounts of water.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
The Neoproterozoic Dorsal de Canguçu Transcurrent Shear Zone (DCTSZ) of the Sul-rio-grandense Shield in Brazil is intruded by a suite of syntectonic, peraluminous two-mica granites. This shear zone transects the orthogneisses of the Arroio dos Ratos Gneiss Complex, which contain mega-enclaves of metasedimentary gneisses belonging to the Varzea do Capivarita Metamorphic Suite (VCMS). Rocks of the paragneiss suite may be much more voluminous at depth than their outcrop suggests, and are possible candidates for the source rocks of the peraluminous granitic magmas.
Melting experiments designed to test the possible role of the high-grade VCMS metasediments as source of the granitic magmas indicate that fluid-absent and H2O-present partial melting can generate a wide range of granitic compositions. Results show that at 5, 10 and 15 kbar, melting begins at 700°C, involving participation of the retrogressive muscovite and producing a very small amount of glass (quenched melt). Melt productivity greatly increases with the breakdown of biotite or cordierite at temperatures 800°C. Orthopyroxene is the main residual mineral at 5 kbar, garnet at 10 and 15 kbar.
The similarity in percentage of melting between the fluid-absent and fluid-present runs shows that the addition of H2O does not greatly enhance melting for any given temperature and pressure. This is attributed to the large amount of H2O bound up in the hydrous minerals, mainly biotite and cordierite, in the PE-1 starting material, demonstrating the influence of cordierite as a hydrous phase in melting previously migmatized protoliths. The melts generated under both conditions have a wide range of granitic compositions (61 < SiO2 < 75 wt %) and peraluminous affinities (1·1 < A/CNK < 2). At 900°C, 15 kbar and under fluid-absent conditions, melts have similar compositions to the syntectonic peraluminous granites associated with the DCTSZ and a reasonable 30% melt fraction.
The experimental results demonstrate that the peraluminous granites syntectonic to the DCTSZ could have been produced by fluid-absent melting of cordierite gneisses of the VCMS during Neoproterozoic transcurrent faulting. Rb/Sr ratios and Ba contents have been used to suggest that the Cordilheira metagranite could have been formed by melting VCMS semipelite but that the Arroio Francisquinho metagranite could not. The depth of melting (54 km) is, however, unlikely, considering the available geological knowledge of the area. Further experiments are currently in progress to assess whether or not such melt compositions can be generated by partial melting of other potential protoliths in the region, such as the orthogneisses or a combination of orthogneisses plus xenoliths of metasediments.
|
ACKNOWLEDGEMENTS |
|---|
The authors would like to thank M. B. Wolf, J. Kaszuba and R. J. Tracy for providing constructive reviews that considerably helped to improve the manuscript, as well as the editor, G. Bergantz. We thank L. D. de Oliveira and L. F. G. Morales for help with the preparation of PE-1 sample (starting material), G. Bromiley for help in the high-pressure–temperature laboratory, D. Johnston for thin-section preparation, and D. Plant and S. Caldwell for help with the microprobe analysis and SEM–BSE images. G. T. R. Droop, L. A. Hartmann, L. V. S. Nardi and M. R. Soares are thanked for discussions, and R. Hoff is thanked for helping with modal analysis using geoprocess software. This work was supported by CAPES (No. BEX0089/98-2) and CAPES–British Council Project (No. 081/98).
|
FOOTNOTES |
|---|
*Corresponding author. Telephone: 00 55 51 33167140. Fax: 00 55 51 33167270.
E-mail: koester{at}ufrgs.br and akscomazzon{at}yahoo.com.br
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGEXPERIMENTAL AND ANALYTICAL...RESULTSMELTING REACTIONSEXPERIMENTAL MELT AND...CONCLUSIONSREFERENCES |
|---|
Barker, F. (1979). Trondhjemite: definition, environment and hypotheses of origin. In: Barker, F. (ed.) Trondhjemites, Dacites and Related Rocks. Amsterdam: Elsevier, pp. 1–12.
Brown, M. (1994). The generation, segregation, ascent and emplacement of granite magma: the migmatite-to-crustally-derived granite connection in thickened orogens. Earth-Science Reviews 36, 83–130.[Web of Science]
Brown, M., Averkin, Y. A., McLellan, E. L. & Sawyer, E. W. (1995a). Melt segregation on migmatites. Journal of Geophysical Research 100, 15655–15679.[Web of Science]
Brown, M., Rushmer, T. & Sawyer, E. W. (1995b). Introduction to special section: mechanisms and consequences of melt segregation from crustal protoliths. Journal of Geophysical Research 100, 15551–15563.
Carrol, M. R. & Wyllie, P. J. (1990). The system tonalite–H2O at 15 kbar and genesis of calc-alkaline magmas. American Mineralogist 75, 345–357.[Abstract]
Chappell, B. W. & White, A. J. R. (1974). Two contrasting granite types. Pacific Geology 8, 173–177.
Clemens, J. D. & Mawer, C. K. (1992). Granitic magma transport by fracture propagation. Tectonophysics 204, 339–360.[Web of Science]
Clemens, J. D. & Vielzeuf, D. (1987). Constraints on melting and magma production in the crust. Earth and Planetary Science Letters 86, 287–306.[Web of Science]
Costa, A. F. U. (1998). Teste e modelagem geofísica da estruturação das associações litotectônicas Pré-Cambrianas no Escudo Sul-rio-grandense. Ph.D. thesis, Instituto de Geociências, Universidade Federal do Rio Grande do Sul, Porto Alegre, 291 pp.
da Silva, L. C., Hartmann, L. A., McNaughton, N. J. & Fletcher, I. R. (1999). SHRIMP U–Pb zircon dating of Neoproterozoic granite magmatism and collision in the Pelotas Batholith, southernmost Brazil. International Geology Review 41, 531–551.[Web of Science]
D’lemos, R. S., Brown, M. & Strachan, R. A. (1992). Granite magma generation, ascent and emplacement within a transpressional environment. Journal of the Geological Society, London 149, 487–490.
Droop, G. T. R. (1987). A general equation for estimating Fe3+ concentrations in ferromagnesian silicates and oxides from microprobe analysis, using stoichiometric criteria. Mineralogical Magazine 51, 431–435.[Web of Science]
Fernandes, L. A. D. & Koester, E. (1999). An overview of the Neoproterozoic Dorsal de Canguçu strike-slip shear zone and its role in the tectonic evolution of the continental crust in southern Brazil. Journal of African Earth Sciences 29, 3–24.
Fernandes, L. A. D. & Porcher, C. C. (2000). Geologia estrutural. In: Porcher, C. A., Cunha, R. C. & Camozzato, E. (eds) Mapa Geológico da Folha de Cachoeira do Sul. Brasília: PLGB CPRM-Serviço Geológico do Brasil.
Fernandes, L. A. D., Menegat, R., Costa, A. F. U., Koester, E., Porcher, C. C., Tommasi, A., Kraemer, G., Ramgrab, G. E. & Camozzato, E. (1995). Evolução tectônica do Cinturão Dom Feliciano no Escudo Sul-rio-grandense: parte II—uma contribuição a partir das assinaturas geofísicas. Revista Brasileira de Geociências 25, 375–384.
Finger, F. & Clemens, J. D. (1995). Migmatization and ‘secondary’ granitic magmas: effects of emplacement and crystallization of ‘primary’ granitoids in Southern Bohemia, Austria. Contributions to Mineralogy and Petrology 120, 311–326.[Web of Science]
Frantz, J. C., Botelho, N. F., Pimentel, M. M., Potrel, A., Koester, E. & Teixeira, R. S. (1999). Relações isotópicas Rb–Sr e Sm–Nd e idades do magmatismo granítico brasiliano da região leste do Cinturão Dom Feliciano no Rio Grande do Sul: evidências de retrabalhamento de crosta continental paleoproterozóica. Revista Brasileira de Geociências 29, 227–232.
Green, T. L. & Vernon, R. H. (1974). Cordierite breakdown under high-pressure, hydrous conditions. Contributions to Mineralogy and Petrology 46, 215–226.[Web of Science]
Harris, N., Inger, S. & Massey, J. (1993). The role of fluids in the formation of high Himalayan leucogranites. In: Treloar, P. J. & Searle, M. P. (eds) Himalayan Tectonics. Geological Society, London, Special Publications 74, 391–400.
Holtz, F. & Johannes, W. (1991). Genesis of peraluminous granites I. Experimental investigation of melt composition at 3 and 5 kbar and various H2O activities. Journal of Petrology 32, 935–958.
Hughes, C. J. (1982). Igneous Petrology. Amsterdam: Elsevier, 551 pp.
Johannes, W. (1984). Beginning of melting in the granite system Qz–Ab–Or–An–H2O. Contributions to Mineralogy and Petrology 84, 264–273.
Koester, E., Soliani, E., Jr, Fernandes, L. A. D., Kraemer, G. & Tommasi, A. (1997). Geocronologia Rb/Sr e K/Ar dos granitóides sintectônicos à Zona de Cisalhamento Transcorrente Dorsal de Canguçu na região de Encruzilhada do Sul, RS. Pesquisas 24, 67–77.
Koester, E., Fernandes, L. A. D., Oliveira, L. D., Cliff, R. A., Soliani, E. Jr., & Morales, L. F. G. (2001a). Cronologia de eventos magmáticos da região leste do Cinturão Dom Feliciano, RS. In: XI Congresso Latino-Americano de Geologia, Montevideo, Uruguay, 12–16/11/2001. Extended Abstract, CD-ROM, 6 pp.
Koester, E., Soliani, E., Jr, Leite, J. A. D., Hartmann, L. A., Fernandes, L. A. D., NcNaughton, N. J., Santos, J. O. S. & Oliveira, L. D. (2001b). SHRIMP U–Pb age for the emplacement of Santana Granite and reactivation of the Porto Alegre Suture, southern Brazil. Journal of South American Earth Sciences 14, 91–99.[Web of Science]
Le Breton, N. & Thompson, A. B. (1988). Fluid absent (dehydration) melting of biotite in metapelites in the early stages of crustal anatexis. Contributions to Mineralogy and Petrology 99, 226–237.[Web of Science]
Le Gardien, V., Thompson, A. B., Grujic, D. & Ulmer, P. (1994). The role of source composition on melt fractions generations during crustal anatexis. EOS Transactions, American Geophysical Union 75(16, Spring Meeting Supplement), 354 pp.
Le Gardien, V., Thompson, A. B., Grujic, D. & Ulmer, P. (1995). Experimental melting of biotite + plagioclase + quartz ± muscovite assemblages and implications for crustal melting. Journal of Geophysical Research 100, 15581–15591.[Web of Science]
Leite, J. A. D., Hartmann, L. A., Fernandes, L. A. D., NcNaughton, N. J., Soliani, E., Jr, Koester, E., Santos, J. O. S. & Vasconcellos, M. A. Z. (2000). Zircon U–Pb SHRIMP dating basement of Dom Feliciano, southernmost Brazil. Journal of South American Earth Sciences 13, 739–750.[Web of Science]
Montel, J. M. & Vielzeuf, D. (1997). Partial melting of metagreywackes, Part II. Compositions of minerals and melts. Contributions to Mineralogy and Petrology 128, 176–196.[Web of Science]
Patiño-Douce, A. E. (1993). Titanium substitution in biotite: an empirical model with applications to thermometry, O2 and H2O barometries, and consequences for biotite stability. Chemical Geology 108, 133–162.[Web of Science]
Patiño-Douce, A. E. (1996). Effects of pressure and H2O content on the compositions of primary crustal melts. Earth Sciences 87, 1–21.
Patiño-Douce, A. E. & Beard, J. S. (1994). H2O loss from hydrous melts during fluid-absent piston cylinder experiments. American Mineralogist 79, 585–588.[Web of Science]
Patiño-Douce, A. E. & Beard, J. S. (1995). Dehydration-melting of biotite gneiss and quartz amphibolite from 3 to 15 kbar. Journal of Petrology 36, 707–738.
Patiño-Douce, A. E. & Beard, J. S. (1996). Effects of P, f(O2) and Mg/Fe ratio on dehydration-melting of model metagreywackes. Journal of Petrology 37, 999–1024.
Patiño-Douce, A. E. & Harris, N. (1998). Experimental constraints on Himalayan anatexis. Journal of Petrology 39, 689–710.[Web of Science]
Patiño-Douce, A. E. & Johnston, A. D. (1991). Phase equilibria and melt productivity in the pelitic system: implications for the origin of peraluminous granitoids and aluminous granulites. Contributions to Mineralogy and Petrology 107, 202–218.[Web of Science]
Pettijohn, S. J. (1963). Sedimentary Rocks. New York: Harper, 718 pp.
Pickering, J. M. & Johnston, D. J. (1998). Fluid-absent melting behaviour of a two-mica metapelite: experimental constraints on the origin of Black Hills Granite. Journal of Petrology 39, 1787–1804.[Web of Science]
Powell, R. & Holland, T. J. B. (1988). An internally consistent thermodynamic dataset with uncertainties and correlations: 3. Applications, methods, worked examples and a computer program. Journal of Metamorphic Geology 6, 173–204.[Web of Science]
Powell, R. & Holland, T. J. B. (1990). Calculated mineral equilibria in the pelite system. KFMASH (K2O–FeO–MgO–Al2O3–SiO2–H2O). American Mineralogist 75, 367–380.[Abstract]
Powell, R. & Holland, T. J. B. (1998). An internally consistent thermodynamic data set for phases of petrological interest. Journal of Metamorphic Geology 16, 309–343.[Web of Science]
Rapp, R. P. & Watson, E. B. (1995). Dehydration melting of metabasalt at 8–32 kbar: implications for continental growth and crust–mantle recycling. Journal of Petrology 36, 891–931.
Rapp, R. P., Watson, E. B. & Miller, C. F. (1991). Partial melting of amphibolite/eclogite and the origin of Archean trondhjemites and tonalites. Precambrian Research 51, 1–25.[Web of Science]
Silva, A. M. (1999). Caracterização das condições de metamorfismo: Suite Metamórfica Várzea do Capivarita, Encruzilhada do Sul, RS. M.Sc. Dissertation, Instituto de Geociências, Universidade Federal do Rio Grande do Sul, Porto Alegre, 240 pp.
Singh, J. & Johannes, W. (1996a). Dehydration melting of tonalites. Part I. Beginning of melting. Contributions to Mineralogy and Petrology 125, 16–25.[Web of Science]
Singh, J. & Johannes, W. (1996b). Dehydration melting of tonalites. Part II. Composition of melts and solids. Contributions to Mineralogy and Petrology 125, 25–44.
Soares, M. R., Hartmann, L. A., Steele, I. M. & Newton, R. C. (1998). Dehydration melting of a natural metatonalitic rock: experiments at high pressures and temperatures. Anais da Academia Brasileira de Ciências 75, 79–87.
Stevens, G., Clemens, J. D. & Droop, G. T. R. (1995). Hydrous cordierite in granulites and crustal magma production. Geology 23, 925–928.
Stevens, G., Clemens, J. D. & Droop, G. T. R. (1997). Melt production during granulite-facies anatexis: experimental data from ‘primitive’ metasedimentary protoliths. Contributions to Mineralogy and Petrology 128, 352–370.[Web of Science]
Storre, B. (1972). Dry melting of muscovite + quartz in the range Ps = 7 kb to Ps = 20 kb. Contributions to Mineralogy and Petrology 37, 87–89.[Web of Science]
Thompson, A. B. & Connolly, J. A. D. (1995). Melting of the continental crust: some thermal and petrological constraints on anatexis in continental collision zones and other tectonic settings. Journal of Geophysical Research 100, 15565–15579.
Tommasi, A., Vauchez, A., Fernandes, L. A. D. & Porcher, C. C. (1994). Magma-assisted strain localisation in an orogen-parallel transcurrent shear zone of southern Brazil. Tectonics 13, 421–437.[Web of Science]
Tuttle, O. F. & Bowen, N. L. (1958). Origin of granite in the light of experimental studies in the system NaAlSi3O8–KAlSi3O8–SiO2–H2O. Geological Society of America Memoir 74, 153 pp.
Van der Molen, I. & Paterson, M. S. (1979). Experimental deformation of partially melted granite. Contributions to Mineralogy and Petrology 70, 218–229.
Vielzeuf, D. & Holloway, J. R. (1988). Experimental determination of the fluid-absent melting relations in the pelitic system. Contributions to Mineralogy and Petrology 98, 257–276.[Web of Science]
Vielzeuf, D. & Montel, J. M. (1994). Partial melting of metagreywackes. Part I. Fluid-absent experiments and phase relationships. Contributions to Mineralogy and Petrology 117, 375–393.[Web of Science]
Vigneresse, J. L., Barbey, P. & Cuney, M. (1996). Rheological transitions during partial melting and crystallisation with application to felsic magma segregation and transfer. Journal of Petrology 37, 1579–1600.
Winther, K. T. (1996). An experimentally based model for the origin of tonalitic and trondhjemitic melts. Chemical Geology 127, 43–59.[Web of Science]
Wolf, M. B. & Wyllie, P. J. (1994). Dehydration-melting of amphibolite at 10 kbar; the effects of temperature and time. Contributions to Mineralogy and Petrology 115, 369–383.[Web of Science]
Wyllie, P. J. (1977). Crustal anatexis: an experimental review. Tectonophysics 43, 41–71.[Web of Science]
Wyllie, P. J. & Wolf, M. B. (1993). Amphibolite dehydration melting: sorting out the solidus. In: Prichard, H. M., Alabaster, T., Harris, N. & Neary, C. R. (eds) Magmatic Process and Plate Tectonics. Geological Society, London, Special Publications 76, 405–416.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F. Ferri, S. Poli, and D. Vielzeuf An Experimental Determination of the Effect of Bulk Composition on Phase Relationships in Metasediments at Near-solidus Conditions J. Petrology, May 8, 2009; (2009) egp026v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Siebel, C. K. Shang, E. Reitter, J. Rohrmuller, and K. Breiter Two Distinctive Granite Suites in the SW Bohemian Massif and their Record of Emplacement: Constraints from Geochemistry and Zircon 207Pb/206Pb Chronology J. Petrology, November 11, 2008; (2008) egn049v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. D. Putirka Thermometers and Barometers for Volcanic Systems Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 61 - 120. [Full Text] [PDF]
|
|
|
|
 |
|
 B. Charoy and P. Barbey Ferromagnesian silicate association in S-type granites: the Darongshan granitic complex (Guangxi, South China) Bulletin de la Societe Geologique de France, January 1, 2008; 179(1): 13 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. HARRIS, A. McMILLAN, M. HOLNESS, R. UKEN, M. WATKEYS, N. ROGERS, and A. FALLICK Melt Generation and Fluid Flow in the Thermal Aureole of the Bushveld Complex J. Petrology, June 1, 2003; 44(6): 1031 - 1054. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||