Mantle-derived CO2-rich fluid entrapment in the lower crust and release via decompression: fluid inclusion systematics from metasedimentary granulite xenoliths in the Pannonian Basin

Carbon is mainly present as a CO₂-rich fluid in lithospheric upper mantle, while upon further upward transport, fluid-involved reactions can be expected in the lower crust, like promoting granulite facies metamorphism. Isotope composition of fluids, like CO₂-rich fluid inclusions in lower crustal xenoliths, serves as a direct tool to trace lithosphere-scale fluid processes and its potential effects on global carbon cycle. This study explores the significance and the fate of mantle-derived fluids, represented by primary and secondary fluid inclusion assemblages in metasedimentary granulites from the lower crust of the Pannonian Basin.

The studied xenoliths are made up mostly by garnet and sillimanite together with plagioclase, quartz, graphite, rutile and zircon in minor quantities. Garnet is mostly surrounded by fine-grained symplectitic aggregates of orthopyroxene, spinel ± plagioclase. The width of symplectitic corona and the extent of garnet breakdown vary at a thin section scale showing few micron sized rims and also the total replacement of garnet. Intact garnet, however hosts abundant primary graphite and fluid inclusions within its core, which were commonly co-entrapped with quartz, rutile and zircon. The smallest (2-3 µm) primary negative crystal-shaped fluid inclusions in garnet dominantly contain high-density (1.05-1.10 g/cm³) CO₂-rich fluid. He-Ne isotope analyses on primary garnet-hosted primary fluid inclusions showed R_c/R_a ratios of 6.3 ± 0.2, thus suggesting a subcontinental lithospheric mantle origin. Results on combined quartz-in-garnet and zircon-in-garnet elastic thermobarometry indicate entrapment at UHT conditions. The intersection of the entrapment isomekes is at a P-T of 1.3 ± 0.4 GPa and 1100 ± 70 ºC. Such P-T conditions are far not compatible with the recent MOHO depth in the Pannonian Basin and clearly indicate that entrapment took place in a crust much thicker than in present days. In the light of previous studies from the Pannonian Basin, garnet is unstable in the present-day lower crust, due to pronounced lithospheric thinning during the Miocene. We provide a calculation on the quantity of CO₂, which has been potential released by decompressional garnet breakdown. Such process serves as a newly discovered source of a delayed, secondary mantle degassing due to the residence of abundant mantle-derived CO₂ in fluid inclusions in garnet, – the most common mineral in the lower crust –, for millions of years. Potential release of such CO₂ may occur much later than leaving the stability field of garnet due to its metastable behavior, as evidenced by xenoliths.

In the light of our results, recent mantle degassing detected by surface-subsurface noble gas isotopic measurements does not always require a direct connection to the mantle or to a cooling magma chamber, but can be derived by destabilization of the lower crust. Accordingly, this mechanism should be taken into account when estimating geological carbon fluxes, as its contribution may be substantial and could potentially influence the overall carbon budget within the Earth system. Incorporating this factor into flux calculations is therefore essential, particularly in post-rift basins, characterized by significant crustal thinning for achieving a more accurate and comprehensive understanding of long-term carbon cycling and its geodynamic controls.