Analytical strategy for representative subsampling of Raman-based robotic planetary exploration missions: The case study of solid dispersions of β-carotene and L-cysteine in gypsum

Raman spectroscopy is a commonly applied technique for identifying themolecular nature of the constituents of solid samples. However, the quantification of the intensity associated with a particular molecular solute in solid dispersions requires analytical strategies to integrate the possible microheterogeneity in terms of composition. Many quantitative applications of solid dispersions rely on large spectral datasets. However, analytical procedures with limited data acquisition must be developed in preparation for future robotic planetary exploration missions or for field applications on earth, both of which involve miniaturised instrumentation. Based on statistical models, we evaluate the minimum number of spectra required to deliver robust data for quantitative measurements associated with differing concentrations of β‐carotene or L‐cysteine dispersed in gypsum. In this respect, reference materials were prepared, and we investigated an analytical methodology based on a large, automated multipoint scanning approach enabling the representative interrogation of the samples. We demonstrate from a comprehensive and reproducible model that the intensity associated with different organic contents within a mineral matrix can be estimated from a reduced number of spectra. Calibration curves were established for our reference dispersions of L‐cysteine in gypsum in the range 1–10 wt%, using both confocal benchtop and nonconfocal miniaturised portable instruments. We propose a reference system and a Raman analytical strategy, that is, adaptable for planetary exploration constraints, and which allows quantitative comparison of performance of instrumentation in terms of signal detection.