NEW METHOD FOR HEAVY MINERAL ASSEMBLAGE ANALYSIS IN THE GEOLOGICAL SURVEY OF ISRAEL USING SCANNING ELECTRON MICROSCOPE - ENERGY DISPERSIVE SPECTROSCOPY PHASE MAPPING

Heavy mineral assemblage analysis is a fundamental tool in geological studies. Heavy mineral analysis is commonly performed manually under optical microscopy by identifying and counting transparent minerals only. This process is very time consuming and largely depends on the researcher’s level of expertise and experience. Adding to that are long-standing problems that cannot be solved confidently with optical microscopy alone. These include the identification of; opaque and intensely-altered minerals, grains smaller than a few microns, and colorless grains with uncertain orientation and rounded morphology. Here we present a relatively simple method for heavy mineral assemblage analysis routinely performed in the Geological Survey of Israel scanning electron microscope (SEM) laboratory. Analysis is performed using an FEI Quanta 450 SEM equipped with an Oxford X-max 20mm energy dispersive spectroscopy (EDS) detector. The acquired EDS data is processed using Oxford Instruments INCA^® software. Heavy mineral analysis is performed by EDS mapping of a 2×2 mm frame with an EDS resolution of ~10µm pixels. The resulting EDS element maps are then used to construct mineralogical phase maps by the following method. A ternary diagram scatter plot is generated from three selected element maps, where each pixel in the element maps is plotted based on its relative proportion of the chosen elements. Distinct mineralogical phases will appear as separate clusters on the ternary diagrams, which are then selected individually and defined as unique mineral phases. Successive series of ternary diagrams for different elements are used until all major mineralogical phases are identified. These mineral phases are finally plotted in different colors on top of the SEM electron image to create a complete mineralogical phase map. The relative abundances of the different mineralogical phases in the sample are then calculated from the total area of all phases. Although automation of the process is considered, at this time accurate application of the method still relies on the mineralogical knowledge and experience of the operator. The potential of the method is demonstrated here by highlighting the differences in heavy mineral assemblages of modern Israel coastal sand and Miocene sandstone formation from the Golan Heights in Northern Israel.