Assimilation at Shallow-Crust Phonolitic Chamber Controlled by Local Anomalous Heating

Abstract In a closed system differentiation, the alkali feldspar thermal divide prevents quartz-bearing magmatic rocks from occurring together with feldspathoidal rocks. Open system differentiation is required to find co-existing quartz and feldspathoids in the same pluton, as gave rise to silica-oversaturated trachytes, quartz alkali feldspar syenite and alkali feldspar syenites of the Passa Quatro Alkaline Complex (PQAC) in SE Brazil. These quartz-bearing rocks, first described here, were formed by phonolite crystallization concomitant to paragneiss assimilation in a shallow-crust magma chamber in a rifted lithosphere at the Late Cretaceous–Paleocene boundary. The proposal of such differentiation is supported by detailed field geology, petrography, mineral chemistry, lithogeochemical and Sr–Nd–Pb isotope data. The PQAC silica-oversaturated rocks are in contact with a paragneiss wall rock and host amphibolite and paragneiss xenoliths and quartz pockets. Our geochemical modeling based on Sr–Nd–Pb isotope data shows that at least 27% of phonolite crystallization concomitant to paragneiss wall rock assimilation would have been necessary to generate the contaminated silica-oversaturated trachyte. Furthermore, thermodynamic phase-equilibria modeling supports the assumption that the assimilation started before abundant alkali feldspar crystallization, allowing latent heat to improve crustal assimilation through magmatic enthalpy. This process led to an alkali feldspar mush zone equivalent to the quartz-bearing syenites that outcrop in the eastern portion of the PQAC. For modeled thermodynamic liquid line of descent approaches PQAC trachytes, a silica increment inversely proportional to alkalinity is needed. This feature resulted from the injection of a contaminant of subalkaline rhyolite into the phonolite uncontaminated magma. The contaminant melt is thermodynamically derived from a rock with a mineralogy similar to the PQAC paragneiss wall rock. The modeling indicates that the wall rock has a half mass fraction compared to the uncontaminated magma chamber. In addition, the wall rock initial temperature is 685°C, which is significantly higher than the typical geothermal gradient expected for shallow-depth continental crust. This hypothesis of heated crust is supported by the ratio of assimilation to crystallization mass fraction in the mass balance models, utilizing isotope and trace element data. We propose that the transposition of the thermal divide requires the presence of an anomalously heated crust. In the case of the PQAC, we suggest this heating could be due to the geographical proximity between Itatiaia and Passa Quatro alkaline complexes, which can represent the coalescence of magmatic chambers in this region of SE Brazil.