Calibration of Raman Bandwidths on the Scanning Habitable Environments with Raman and Luminescence for Organics and Chemicals (SHERLOC) Deep Ultraviolet Raman and Fluorescence Instrument Aboard the Perseverance Rover

Abstract

In this work, we derive a simple method for calibrating Raman bandwidths for the Scanning Habitable Environments with Raman and Luminescence for Organics and Chemicals (SHERLOC) instrument onboard NASA's Perseverance rover. Raman bandwidths and shapes reported by an instrument contain contributions from both the intrinsic Raman band (IRB) and instrumental artifacts. To directly correlate bandwidth to sample properties and to compare bandwidths across instruments, the IRB width needs to be separated from instrumental effects. Here, we use the ubiquitous bandwidth calibration method of modeling the observed Raman bands as a convolution of a Lorentzian IRB and a Gaussian instrument slit function. Using calibration target data, we calculate that SHERLOC has a slit function width of 34.1 cm–1. With a measure of the instrument slit function, we can deconvolve the IRB from the observed band, providing the width of the Raman band unobscured by instrumental artifact. We present the correlation between observed Raman bandwidth and intrinsic Raman bandwidth in table form for the quick estimation of SHERLOC Raman intrinsic bandwidths. We discuss the limitations of using this model to calibrate Raman bandwidth and derive a quantitative method for calculating the errors associated with the calibration. We demonstrate the utility of this method of bandwidth calibration by examining the intrinsic bandwidths of SHERLOC sulfate spectra and by modeling the SHERLOC spectrum of olivine.

Characterizing Hydrated Sulfates and Altered Phases in Jezero Crater Fan and Floor Geologic Units With SHERLOC on Mars 2020

“…The intrinsic Raman bands of well‐ordered crystalline solids are generally homogeneously broadened (Califano & Schettino, 1988; Mukamel, 1985), are likely Lorentzian and have full width at half‐maximum (FWHM) < 50 cm −1 . In the SHERLOC Raman spectra, the observed band shapes of peaks of FWHM up to ∼50 cm −1 are largely dominated by the SHERLOC instrument slit function, which is well‐approximated as a Gaussian (Jakubek et al., 2023). On the other hand, for amorphous solids and hydration bands in the ∼3,000–4,000 cm −1 region, the intrinsic Raman bands are generally inhomogeneously broadened due to the diversity of microenvironments (Leikin et al., 1997; Walrafen et al., 1996), giving rise to Gaussian band shapes.…”

“…In the mineral region, a lower bound to the FWHM was set at 20 cm −1 (∼2 spectral detection elements) to separate mineral signals from potential cosmic ray spikes. A SHERLOC Raman peak cannot be narrower than the 34.1 cm −1 FWHM of the slit function, and a feature with a FWHM <34.1 cm −1 is likely a cosmic ray or anomalous pixel feature (Jakubek et al., 2023). However, there is some error associated with band width value measurements that can report a FWHM <34.1 cm −1 for a Raman peak, thus a FWHM lower limit of 20 cm −1 was used to avoid the removal of these Raman peaks from consideration.…”