FLUORESCENCE INTERFERENCE IN RAMAN SPECTROSCOPY: WILL THE GEOSCIENCE COMMUNITY DRIVE THE INSTRUMENT INDUSTRY?

Raman Spectroscopy (RS) has the potential to transform petrology and applied mineralogy due to its ability to perform in-situ mineral identification without the sample volume and sample destruction required by powder XRD and without the greater effort and expense of EPMA. Recently it has been shown that RS would be as effective as powder XRD for routine mineral identification if only the interference from specimen fluorescence (SF) could be minimized or eliminated (Bartholomew, 2011). Driven by the biomedical, pharmaceutical, and homeland security industries, the commercial RS instrument offerings have been moving to longer (NIR to IR) laser wavelengths which rarely excite SF in organic materials. However the decrease in Raman intensities accompanying longer wavelengths becomes unacceptably low for minerals when using an IR laser (e.g. 1064nm).

This study evaluates the efficacy for the geoscience industry of the two laser wavelengths most commonly offered in commercial RS instruments (780nm and 532nm) as well as one of the custom offerings (514nm) by reviewing spectra stored in the rruff.info database of Raman spectra. Selecting spectra from samples of rock forming minerals for which spectra from all 3 lasers are in the database resulted in 789 spectra representing 124 mineral species. Due to the variable spectral characteristics of SF these spectra had to be visually reviewed and, when SF was visible, grouped by type and intensity of spectral interference. While SF was most common for the 780nm spectra the frequency of SF interference was significant for both the 532nm and 514nm lasers. When SF interference was present it rarely interfered in the spectra from all 3 lasers, but interference affecting 2 of the three wavelengths was not uncommon. This study does not suggest that a 3 laser Raman instrument (prohibitively expensive to many potential users) is required for dependable mineral identification, but it does indicate that the most commercially available laser wavelengths are not the best choice for geoscience materials. Follow up work is under way to identify the spectral gaps in mineral fluorescence which would be the target laser wavelength regions for a geoscience-optimized Raman instrument. However, it will require demand for such a machine for the instrument industry to produce a commercial offering.