Character Of Signal And Noise Sources In Dispersive And Static Fourier Transform Remote-sensing Raman Spectrometers

Abstract

Raman scattering, the inelastic scattering of photons by polarizable vibrational modes, was first reported by Raman and Krishnan in Nature (Raman 1928) as combination scattering in liquids. Landsberg and Mandelstam reported observing the same effect in quartz simultaneously in Naturwissenschaften (Landsberg 1928). The discovery of the Raman effect was pivotal in support of the quantum theory of light and matter as the quantity of energy imparted to a scattered photon is always equal to a harmonic vibrational energy, regardless of the intensity of the excitation light source. Raman spectroscopy probes stereochemistry in a given target analyte and is therefore sensitive to the quantity of electron density shared between elements in a molecular vibrational mode as well as its reduced mass. Raman spectral features occur at well-defined energies as sharp, narrow peaks in contrast to IR spectra that occur as broad features caused by the multitude of symmetry-allowed ro-vibrational transitions for a given vibrational harmonic. The position of Raman peak centers is so selective that polymorphs, hydration states and compositional variation can be quantitatively characterized by peak centers alone. Applying remote-sensing Raman spectrometers to probe the natural world was proposed as early as 1965 (Cooney 1965; Leonard 1970), but the successful implementation of such an instrument was said to be “impossible” by early authors (Hirschfeld 1972) because of the low value of Raman scattering cross-sections, the diminution of Raman signal with distance, the low number of Raman-active molecules illuminated in a remote-sensing Raman scattering experiment, and competition for a detector’s limited dynamic range with more luminous radiation sources. However, the advent of pulsed laser excitation sources and time-gated detectors made possible remote Raman measurements. Since that time, a number of research groups have proposed utilizing remote-sensing Raman spectrometers for planetary science missions. Recent work has demonstrated the ability of remote-sensing Raman spectrometers to measure planetary science materials at hundreds to thousands of meters distance (Acosta-Maeda 2016; Misra 2020). Extensive calibrations have already been performed on planetary science materials, including modeling compositional variation as a function of Raman peak shifts, and crystallinity as a function of Raman linewidth. At the time of this publication, the first two Raman spectrometers are en route to Mars aboard the Perseverance rover (Beegle 2021; Wiens 2021). Two classes of remote-sensing Raman spectrometers exist: dispersive grating spectrometers (Sharma 2003) and static Fourier Transform spectrometers (Gomer 2011). Differentiation between the two categories arises from how each spectrometer separates wavelength. The former separating wavelengths by use of an entrance slit and increasing with focal length of the spectrometer, and the latter separating wavelengths by phase differences between crossed wavefronts. Dispersive grating spectrometers attain high sensitivities as a result of localized detection of Raman modes on a detector and correlated photon-noise, while constrained by light throughput and narrow fields-of-view. Static Fourier Transform spectrometers attain medium sensitivities and large fields-of-view, while being limited by multiplicative photon-noise and diluted photon detection. In this work, the merits of both classes of remote-sensing Raman spectrometers are evaluated and their limits of sensitivity defined.