IDENTIFICATION OF DETRITAL HEAVY MINERALS USING LASER-INDUCED BREAKDOWN SPECTROSCOPY (LIBS) AND MULTIVARIATE ANALYSIS

Prior to the development of detrital zircon geochronology techniques, the primary method for understanding provenance trends was to visually identify the constituent detrital heavy minerals in a sediment or sedimentary rock. The development of Laser-Induced Breakdown Spectroscopy (LIBS), a rapid laser ablation optical emission analytical technique, makes possible rapid and accurate identification of the heavy minerals in a clastic sediment and measurement of their proportions, providing valuable information to enhance detrital zircon U-Pb studies. Multivariate analysis of LIBS spectra taken from heavy grains can both identify and separate minerals into clusters of compositional subgroups within a specific mineral. To accomplish this, LIBS spectra from 10+ samples of 17 mineral groups were collected. A multivariate decision tree was calibrated with 75% of these spectra. The remaining 25% were used as a validation set testing the accuracy of the algorithm. This mineral identification decision tree consists of a series of binary Partial Least Squares Regression (PLSR) models, with each model targeting a detrital heavy mineral. A previous iteration of this PLSR decision tree was able to separate and identify with 100% accuracy seven of the targeted mineral groups. LIBS spectra acquired from ca. 500 heavy mineral grains from each of three outcrops of the of the Early Permian (Wolfcampian) non-marine Abo Formation in New Mexico were classified by mineral species using the mineral identification decision tree. In addition to the counts of the 17 targeted mineral groups, each mineral group will be analyzed using PCA cluster analysis to test for compositional subgroups. The mineral counts and subgroups will supplement the findings of a previous detrital zircon analysis of the Abo Formation done by Bonar (2018) showing sediment dispersal pathways from active Ancestral Rocky Mountain uplifts.