IMPROVING PRESSURE ESTIMATION IN HIGH PRESSURE TERRANES WITH RAMAN SPECTROSCOPY: NEW CALIBRATION AND AN IMPLEMENTATION ON BLUESCHISTS FROM SIFNOS, GREECE

Schmidt and Ziemann (2000) measured the thermobarometric influence on Raman peak position (Δu) for α-quartz, reporting a systematic change in peak position as a function of P & T. Here we present an equation describing these relationships, which can be used to constrain P in high-P terranes with quartz inclusions in rigid phases (e.g. garnet). Published data for the positions of two α-quartz peaks (c. 464 and 206 cm^-1) were evaluated through a database of surface regression fits, with the equation selected based on both its simplicity and the resultant root mean squared error. Calibrations were developed for both Δu₄₆₄ and Δu₂₀₆, however the equation for Δu₂₀₆ was selected for this study owing to its larger pressure dependence. A Taylor A series fit was implemented and resulted in a high coefficient of determination (R²) of 0.9986 and a low total significance of probability (p-value) of 1.11x10^-16. Maximum uncertainty expected from analytical and statistical sources (excluding textural, geological uncertainty) is ± 0.4 kbar.

The relationship was tested on quartz- and garnet-bearing blueschists from Sifnos, Greece. Care was taken to select those inclusions completely encapsulated by garnet in 3D which have the highest chance of preserving the PT conditions at which they were included in garnet. Garnet growth was estimated to start c. 460°C and end at 560°C (Dragovic et al., 2012), with P between 20–22 kbar. We estimate P to be 18.1 kbar at 460°C and 19.7 kbar at 560°C (Δu₂₀₆ = 20.97 cm^-1). Differences in calculated P could result from partial relaxation of the garnet that surrounds the inclusion and cracks/surface-exposure. Thus we suggest that calculated pressures are minimum values. Further experimental work at high-P conditions could better refine the relationship between T, P and Δu at these high-P conditions.