Optimizing Raman Analyses of Quartz and Zircon Inclusions in Garnet

Raman peak shifts of mineral inclusions can be used to calculate pressures and temperatures (P's and T's) of metamorphic rocks. However, for accurate estimates of P-T conditions, precise measurements (typically better than 0.1 cm^-1) are needed. In this study, we investigated how to optimize data collection, and consequently, obtain more precise P-T estimates. Specifically, we sought to answer the following questions: (1) How does power density affect Raman peak positions of inclusions and standards (e.g. because of heating); (2) What is the best analysis location; (3) What is the most efficient and precise peak fitting tool; and lastly, (4) How much do peak positions drift over time? We explored these questions by analyzing standards and inclusions using a Horiba Scientific LabRAM HR Evolution with three different lasers [442 nm (blue), 532 nm (green), 633 nm (red)], a range of Neutral Density (ND) filters (0.01 to 100%) and acquisition times (30 to 210s), and a grating density of 1800 line/mm. For fitting the Raman peaks, we used four different tools: LabSpec 6, PeakFit, and "home-grown" software developed using MATLAB and RStudio. Our results show that: (1) power density does not affect Raman peak positions on quartz or zircon grains, and a 442 nm excitation wavelength yields the highest count rate; (2) Inclusions located in the middle of a 100 μm thick section of host mineral show better peak position reproducibility (<0.1 cm^-1 for the "464" peak in quartz) than shallower or deeper inclusions (20 μm of surface; 0.7 cm^-1). (3) All four software packages yield similar peak positions (to within 0.3 cm^-1) for all quartz and zircon Raman peaks, but the MATLAB code runs fastest; (4) Peak positions measured with the 532 nm laser drift by as much as 0.8 cm^-1 within the first hour of analysis, following by stabilization to very slow drift (0.1-0.2 cm^-1/10hr for the "464" peak) ; however, some peak position "jumps" of 1 cm^-1 can also occur over a period of 1 hour during a 24h session. The results imply that heating is not a concern for Raman analysis of quartz and zircon in garnet, so blue lasers that yield the highest count rates are preferred. Centrally-located inclusions yield more consistent results than shallower or deeper inclusions. Analysis should be avoided during the first hour of instrument warm-up and standards should be monitored closely.