Identification and quantification of minerals within Apollo samples using LWIR spectroscopy

The mineral composition of the lunar surface is a critical scientific objective, as it provides key insights into the evolution of the Moon. For instance, the variations in Mg-Fe composition of pyroxene solid solution could reveal the Moon’s crustal evolution and magmatic history. Reflectance/emission spectroscopy provides a rapid and nondestructive technique for remote mineralogical detection. Visible and near-infrared (VNIR) reflectance spectroscopy has been widely used for lunar mineral identification in both remote sensing and proximal analyses. However, some silicate minerals, such as quartz and feldspar, are spectrally featureless in the VNIR range due to the absence of transition metals. The long-wave infrared (LWIR) spectroscopy hosts fundamental molecular vibrations of silicate minerals and can provide spectral, mineralogical, and chemical information of the samples. In this study, we investigated the potential of LWIR spectroscopy for identifying the mineral components of lunar samples and remotely determining the modal mineralogy of the lunar surface. We analyzed 26 laboratory LWIR spectra measured from eight representative samples collected during three Apollo missions. Spectral analysis revealed distinct mineralogical features between lunar mare and highland regions. Mare samples exhibit multiple features indicating plagioclase, pyroxene, olivine, while highland samples exhibit pronounced spectral features of plagioclase. We identified the transparency feature (TF) at approximately 12.2 μm as an indicator of fine-fraction (0–45 μm) samples. Furthermore, we introduced an approach to retrieval modal mineralogy from LWIR spectra without input endmember spectral libraries. Partial least squares regression (PLSR) models were developed to correlate the laboratory measured modal mineralogy with LWIR spectra. The PLSR predicted results are strongly correlated with laboratory measurements, with Spearman correlation coefficients exceeding 0.80 for plagioclase, pyroxene component (all the subtypes of pyroxene in the bulk sample) and olivine. The results of this study provide a spectral-compositional framework for both laboratory analyses and orbital remote sensing investigations.