Thermobarometry and geochronology of metapelites and granitoids of the eastern Sierra de Velasco, Sierras Pampeanas, Argentina: P-T paths and exhumation rates during the Famatinian Cycle.

DE LOS HOYOS, CAMILO; WILLNER, ARNE; LARROVERE, MARIANO; ROSSI, JUANA; TOSELLI, ALEJANDRO

Göttingen, Alemania

Conferencia; 21 International Lateinamerika-Kolloquium; 2009

German Science Foundation

The Sierra de Velasco belongs to the Sierras Pampeanas Geological Province and is formed almost exclusively by Famatinian shear-deformed and undeformed I- and S-type granitoids corresponding to two geochronological groups: Lower/Middle Ordovician and Lower Carboniferous granitoids (e.g. Báez et al., 2005). The metasedimetary host-rock, the La Cébila Formation (LCF), crops out as small N-S belts along the eastern flank of the sierra. Most Ordovician granitoids are affected by NNW-SSE Devonian shear zones, while the Carboniferous ones are undeformed. The eastern flank of the sierra is composed of probably Ordovician S-type granitoids intruded in the LCF, this block being then intruded by Carboniferous granites. Field observations combined with petrographical and geochemical studies suggest that the LCF was first intruded by moderately to strongly peraluminous tonalites and granodiorites. Irregular to diffuse contacts and coeval migmatization of the host-rock with the assemblage Grt+Bi+Wm+Sil+Pl+Qtz indicate relatively deep emplacement levels for these bodies, while shear deformation affecting them suggests a probable Ordovician age. These granitoids and the LCF were then intruded by Crd-bearing granites at shallower levels, because of the sharp contacts and the development of Crd-Sil-Kfs-Bi-bearing hornfelses (Rossi et al., 2005). The relatively deep-emplaced granitoids and the shallower Crd-bearing granites are referred as DEG and CBG respectively. The above mentioned units were finally intruded by the Huaco and Sanagasta S-type plutons in Early Carboniferous times (Grosse et al., 2008) at emplacement levels probably shallower than those for the CBG, indicated by almost unaffected metasedimentary and granitic xenoliths. Samples representing the DEG, CBG and LCF were collected for thermobarometric and geochronological studies. The main goal of this contribution is to estimate exhumation rates during the Early/Middle Ordovician to the Early Carboniferous in order to constrain a geodynamic model for the Famatinian Orogen in this part of the Palaeozoic Gondwana margin. The chosen samples for this study are: 7992, representing a Grt-Bi-Ms-Pl-Sil migmatite ~2 Km away from the contact with the DEG; 7993, a Grt-Bi-Ms-Pl-Sil-bearing migmatite immediately at the contact; 7901, a Crd-Sil-Bi-Wm-Pl-bearing hornfelse septa within the CBG; 7833 and 7903 representing the DEG and CBG respectively. Monazite single-grain fractions from 7833 and 7903 were analyzed by the ID-TIMS method for U-Pb dating. Early/Middle Ordovician ages were obtained for both samples: 476±2 Ma (MSWD=0.5) for 7833 and 472±16 Ma (MSWD=1.5) for 7903. In addition, Pb, Th and U contents in monazite cores and rims within two polished sections of 7992 (leucosome and mesosome) were analized in situ using an electron microprobe. We found no important difference in monazite compositions from leucosomes and mesosomes, and a mean age of 477.2±4.7 (2σ) Ma was acquired from 21 grains, in good agreement with ID-TIMS for the DEG, suggesting that migmatization of the host-rock was coeval with the intrusion. Major elements for minerals within 7992, 7993 and 7901 were analyzed in situ using an electron microprobe for thermobarometry. Two whole-rock powders of 7993 and 7901 were analyzed for major elements to be used to calculate P-T pseudosections. In order to test the reliability of the P-T estimates, we compared independent thermobarometric methods such as conventional thermobarometers, multiphase equilibria and P-T pseudosections. The mineral assemblage within 7992 and 7993 permitted to use conventional and multiphase-equilibria thermobarometric methods. Spessartite profiles for garnets within 7992 are nearly flat in cores and inner rims with an increase in this component towards outer rims, indicating homogenization by high-temperature diffusion. Garnet core and rim compositions for 7992 combined with Wm and Bi inclusions yielded average P-T values of 610-577oC and 4.8-4.4 Kbar for cores and rims respectively. Some spessartite profiles for relatively large garnets in 7993 are “bell-shaped”, suggesting preservation of the prograde history. Garnet rim compositions and small “spessartite-flat” garnets with biotite and white mica from the matrix yielded average values of 728oC and 6.4 Kbar. Multiphase equilibria calculations (TWQ 2.32 software; Berman, 1991 with the corresponding database to different versions) yielded average temperatures of 598-565oC for garnet cores and rims in 7992, while 717oC was calculated for garnet rims in 7993. Pressures for garnet cores in 7992 and rims in 7993 were not determined because of the lack of plagioclase inclusions in 7992 and the low grossular content in 7993. An average pressure of 4.3 Kbar was obtained for garnet rims in 7992. P-T pseudosections for 7993 and 7901 were calculated using the Perplex software (Connolly & Petrini, 2002) to reconstruct stability fields of theoretical assemblages and compare them to observed mineral assemblages. Prograde garnets in 7993 were used to calculate garnet end-member isopleths. Isopleth intersections were always good and permitted to reconstruct a prograde P-T path defined by initial P-T conditions of 582oC and 6.2 Kbar for the core, and peak P-T conditions of 720oC and 9.1 Kbar for the rim. For sample 7901, a small stability field was calculated for the assemblage Bt+Wm+Crd+Pl+Sil+Qtz, in good agreement with the observed assemblage. The stability field is within the range of T=600-645oC and P=2.2-2.6 Kbar, and gives a good estimate for the emplacement conditions of the CBG. Combining results from the different methods, the LCF underwent during the Ordovician a clockwise P-T path in which loading during prograde garnet growth buried the block to a crustal level of ~33 Km. Subsequently, isothermal decompression during upift was accompanied by extensive granite intrusion at ~17 Km depth, followed by a final stage of isobaric cooling. Probably in the Middle Ordovician, the block was exhumed to depths of ~9 Km and intruded by the CBG. Despite the low resolution in ages for metamorphism and intrusion, exhumation rates can be roughly estimated. If prograde garnets within 7993 grew at the beginning of the Famatinian cycle (~490 Ma) and granitoids of the Sierra de Velasco intruded within a period of around 480-460 Ma, about 16 Km crust were exhumed in ~10 Ma, giving a fast initial exhumation rate of 1.6 mm/y during the Early Ordovician. During the Early to Middle Ordovician, about 8 Km crust would have been exhumed in around 20 Ma before the intrusion of the CBG, giving a four times slower velocity of 0.4 mm/y for this period. Between the Middle Ordovician and Early Carboniferous, the entire sequence was uplifted and intruded by the Huaco Pluton at ~7 Km depth (De los Hoyos, in prep), giving a twenty times slower exhumation rate of 0.02 mm/y. Thus, blocks uplift was first fast and then slowed down up to a steady-state stage between the Middle Ordovician and Early Carboniferous. This stage was accompanied by shear deformation in the Devonian and followed by S-type magmatism. If extensive shear deformation was related to a collisional event it is unlikely to be orthogonal, because uplift is virtually absent, but rather oblique or strike-slip. However, a detailed discussion of the exhumation and deformation mechanism cannot be given here, because structural data are not available.