Automated analysis of carbon in powdered geological and environmental samples by Raman spectroscopy.

Sparkes, R, Hovius, N, Galy, A, Kumar, RV and Liu, JT (2013) Automated analysis of carbon in powdered geological and environmental samples by Raman spectroscopy. Applied Spectroscopy, 67.

Preview

Download (150kB) | Preview

Abstract

Raman spectroscopy can be used to assess the structure of naturally occurring carbonaceous materials (CM), which exist in a wide range of crystal structures. The sources of these geological and environmental materials include rocks, soils, river sediments, and marine sediment cores, all of which can contain carbonaceous material ranging from highly crystalline graphite to amorphous-like organic compounds. In order to fully characterize a geological sample and its intrinsic heterogeneity, several spectra must be collected and analyzed in a precise and repeatable manner. Here, we describe a suitable processing and analysis technique. We show that short-period ball-mill grinding does not introduce structural changes to semi-graphitized material and allows for easy collection of Raman spectra from the resulting powder. Two automated peak-fitting procedures are defined that allow for rapid processing of large datasets. For very disordered CM, Lorentzian profiles are fitted to five characteristic peaks, for highly graphitized material, three Voigt profiles are fitted. Peak area ratios and peak width measurements are used to classify each spectrum and allow easy comparison between samples. By applying this technique to samples collected in Taiwan after Typhoon Morakot, sources of carbon to offshore sediments have been identified. Carbon eroded from different areas of Taiwan can be seen mixed and deposited in the offshore flood sediments, and both graphite and amorphous-like carbon have been recycled from terrestrial to marine deposits. The practicality of this application illustrates the potential for this technique to be deployed to sediment-sourcing problems in a wide range of geological settings.