TOURMALINE PROVENANCE DETERMINATION IN MODERN SANDS, BLACK HILLS, SOUTH DAKOTA: MULTIVARIATE ANALYSIS OF LASER-INDUCED BREAKDOWN SPECTROSCOPY (LIBS) SPECTRA
A total of 162 tourmaline grains were separated and analyzed from the coarse (0.5-1.0 mm) sand fraction of drainages: French Creek (n=44), Spring Creek (n=99), Cheyenne River (n=19). Five LIBS spectra were averaged from each grain. These spectra were then analyzed by a matching algorithm developed using 60 spectra from six lithologic categories: calcareous metamorphic rocks, hydrothermal deposits, Li-rich pegmatites, Li-poor pegmatites, pelitic metamorphic rocks, and silicic igneous rocks. The matching algorithm is a sequence of PLSR models, each of which classifies spectra into two groups: the host lithology being defined and a group of all other lithologies. After a model defines a lithology, all associated spectra are removed from subsequent models.
This model showed that Li-rich pegmatites were the dominant source (84-85%), silicic igneous rocks (12-16%) and Li-poor pegmatites (0-3%) were secondary sources, and metasedimentary grains are absent in the French and Spring Creek drainages. The Cheyenne River sample was similar, with 79% from Li-rich pegmatites and 21% from silicic igneous rocks. These findings generally agree with those based on other, more involved methods (Viator 2003) that granitic and pegmatitic lithologies are the dominant source of tourmaline in the coarse sands from the Black Hills.