Hyperspectral Quantitative Mapping and Microanalysis of Minerals at Submicron Spatial Resolution

Improvements in electron optics, field emission gun design, and X-ray detector technology have led to the development of a new class of microanalysis tools. By combining multiple wavelength dispersive (WD) spectrometers, spectral energy dispersive (ED) spectrometers and cathodoluminescence (CL) data sets, a wealth of information can be collected during routine electron beam mapping and analysis. In this paper, data collection on a field emission electron microprobe equipped with five wavelength dispersive, two silicon drift energy dispersive X-ray spectrometers and a wavelength dispersive linear array optical spectrometer has been automated with the optical, electron and X-ray signals being collected in parallel. To analyse the large hyperspectral data sets we have implemented an autoclustering strategy that can be used to produce a quantitative chemical phase map. The cathodoluminescence spectra can be deconvoluted to produce peak intensity maps, which are then converted into elemental speciation images with quantitative sensitivity down to parts per million levels. Both methods are illustrated using typical mineral processing examples, firstly a Ni-laterite sample is investigated with the fine-scale distribution of Ni examined in detail. Secondly, a scheelite sample with rare earth zoning in mapped and rare earth levels quantified.