Raman and reflectance spectroscopy of serpentinites and related hydrated silicates: Effects of physical properties and observational parameters, and implications for detection and characterization on Mars

Abstract

Raman and high-resolution reflectance spectrometers will play an increasingly important role in future surface exploration of Mars. We undertook a study of the 532 nm Raman and 0.35–2.5 μm reflectance spectral properties of a suite of terrestrial serpentinites and related rocks from the Southern Quebec Ophiolite Belt, Canada in order to: (1) determine what factors control the appearance of their Raman and reflectance spectra; and (2) enable more robust analysis of Raman and reflectance observational data for Mars to be provided by future Mars landed missions. We examined the effects of surface texture, presence of weathered surfaces, sample heterogeneity, grain size (whole rock versus powder) on Raman and reflectance spectra, and of integration time on the Raman spectra. The Raman spectra are characterized by strong induced fluorescence, however they usually still allow the strongest peaks of the major silicate phase (serpentine, tremolite-actinolite, or talc) to be identified. The different serpentine polymorphs (antigorite, chrysotile, lizardite) can also normally be identified by differences in some Raman peak positions. Only a few of the accessory phases are recognizable in the Raman spectra. There is no systematic difference relating the presence or absence of Raman peaks between whole rock and powder spectra. Increasing integration time can allow weak Raman peaks to be more confidently recognized. Reflectance spectra are dominated by the numerous absorption bands of serpentine or talc, making these phases easily identifiable, but also making accessory phases difficult to identify. Factors such as tetrahedrally-coordinated Fe3+, mixed valence Fe (Fe2+/Fe3+), and the presence of octahedrally coordinated Fe2+ in serpentine can also be recognized. As in the Raman spectra, accessory phases are generally not identifiable in the reflectance spectra due to low abundance, weak absorption bands, or because their absorption bands are overlapped by the stronger hydrated silicate absorption bands. In the reflectance spectra the presence of magnetite can be recognized by a lowering of reflectance and a negatively-sloped spectrum. The presence of other opaque phases may be inferred from a lowering of reflectance. Reflectance spectroscopy is generally superior to Raman spectroscopy for identifying the presence of serpentine because of the aforementioned induced fluorescence seen in some of the Raman spectra. Raman and reflectance spectra do not always allow for identification of the same accessory phases, demonstrating their complementarity for analysis of hydrated silicate-dominated rocks.