3D Petrography - Serendipitous Discovery of Magmatic Vapor Deposition of Anhydrite at Mount Pinatubo by SEM Imaging of Outer Crystal Surfaces

A standard petrographic technique focuses upon examination of surfaces or planes cut through rock samples, with one approach studying chemical variations in a core to rim traverse using various microprobes, and more recently, another determining the distribution of crystal sizes to obtain information about nucleation and growth. We show that another mineral domain deserves petrographic attention: the outer surfaces of crystals, which are normally relegated to nearly invisible thin lines in a cut section. In studying anhydrite phenocrysts from the 1991 climactic eruption of Mt. Pinatubo, SEM examination of "raw" pumice fragments showed the existence of a Ca-sulfur-rich phase with hexagonal morphology residing upon plagioclase phenocryst surfaces in vesicles (Fournelle et al,1996, Fig 9). In 1992, Terry Gerlach suggested that the Pinatubo anhydrite phenocrysts should be evaluated with XRD to determine if they were indeed orthorhombic anhydrite (β -CaSO4), and not a lower temperature polymorph (i.e., α or γ ). In 1998, we recommenced this project, mounting several dozen 100-200 micron-size phenocrysts of the proper density fraction on tape (minerals had been separated from the pumices using standard techniques). They were examined by low resolution SEM with EDS to distinguish the anhydrite from apatite, prior to single-crystal XRD. We were surprised to find that many of the anhydrite surfaces were decorated with small mounds, which upon examination by high resolution SEM turned out to be micron and smaller pyramids, with some surfaces bearing hundreds. Single-crystal XRD verified that the phenocrysts were orthorhombic anhydrite, and EBSD verified that the small pyramids were the same. Eventually we found that these surface pyramids are common phenomena in experimental or industrial chemical vapor deposition processes when nucleation overwhelms growth. Textural relations were consistent with these pyramids being deposited in situ, within the Pinatubo magma chamber, significantly prior to eruption, with geochemical modeling supporting this hypothesis (Jakubowski et al, 2002, Am. Min 87, 1029; download from www.geology.wisc.edu/ ∼~johnf/Ryan.pdf) As demonstrated here, a polished thin section can entirely miss critical petrographic information present upon the outer crystal surface. Consequently, additional sample preparation may be necessary, including careful separation of minerals or clumps of minerals and matrix, followed by imaging by SEM. We suggest that one impact may be in the study of volcanic materials, where there may have been a vapor present at depth prior to eruption, and where magmatic vapor deposition processes may have left evidence on the surfaces of crystals.