講演情報
[SMP31-P07]Non-destructive peak metamorphic temperature estimates of carbonaceous materials with visible and near-infrared micro-Raman spectroscopy
*中村 佳博1、金木 俊也1 (1.国立研究開発法人産業技術総合研究所 地質調査総合センター)
キーワード：
near-infrared micro-Raman spectroscopy、carbonaceous material、graphite、RSCM thermometry
We develop revised Raman spectroscopy of carbonaceous material (RSCM) thermometries based on both visible (532 nm) and near-infrared (785 nm) micro-Raman spectroscopy, with particular emphasis on non-destructive estimation of peak metamorphic temperatures using rock chips. Sixteen pelitic rock samples with independently constrained peak metamorphic temperatures of 200–650 °C were selected as reference materials for calibration. Raman spectral parameters obtained under different excitation wavelengths were systematically evaluated using both deconvolution and non-deconvolution approaches. Although excitation wavelength strongly affects Raman peak positions (e.g., the D1 band), key spectral parameters, including the D1 band full width at half maximum (FWHM) and the R2 ratio (area ratio of D1 to D1 + D2 + G bands), exhibit consistent temperature-dependent trends across both visible and near-infrared datasets. We demonstrate that peak metamorphic temperatures can be reliably estimated from rock chips using both 532 nm and 785 nm excitation, with precision comparable to that obtained from thin sections, particularly for low- to medium-grade metamorphic rocks (200–400 °C). At higher metamorphic grades (400–650 °C), the spectral evolution of carbonaceous material differs between polished thin sections and rock chips owing to the preferred orientation of graphite crystals. Nevertheless, by applying thermometric calibrations that explicitly account for these effects, peak metamorphic temperatures can be robustly estimated irrespective of excitation wavelength or sample condition. Furthermore, by integrating automated micro-Raman mapping, autofocus control, and a Python-based data-screening algorithm, we establish a fully automated workflow for non-destructive temperature estimation without manual selection of carbonaceous material grains, particularly in carbonaceous material–rich pelitic rocks. This integrated approach substantially reduces operator-dependent analytical bias and sampling heterogeneity in metamorphic rocks, enabling acquisition of large, high-quality temperature datasets using conventional visible and near-infrared Raman spectrometers. These datasets provide a robust framework for quantitative reconstruction of tectonic and thermal histories in collisional and subduction-zone settings.