Mineral identification by elemental composition: a new tool within PDF-4 databases

Abstract

The ICDD has developed a microanalysis tool to help scientists identify minerals from their elemental analyses, most typically micro-XRF or a microprobe analysis. Many minerals have characteristic elemental profiles that can often distinguish the mineral from others by their composition differences. In Release 2016 ICDD® PDF-4 databases 20 670 unique compositions have been identified out of 45 497 mineral and mineral-related entries. The application utilizes several common features of PDF® databases to enhance correct identification, most notably those formulas are expressed in weight and atomic percent, data sets are classified by mineral nomenclature and structural classifications, and most minerals have associated atomic and molecular structures. These crystal structures are very useful in determining compositional variants and solid solutions. The ICDD has developed algorithms that are analogous to the search/match processes used for powder diffraction identification. Data can be input as either the element or common oxide. To test the algorithm and graphics interfaces we compared results from the microanalysis module to published data from the Smithsonian Microbeam reference mineral collection. The software correctly identified 24/28 minerals by the highest merit score in the algorithm. In two cases, an isoelemental mineral was identified and in two other cases, the specimens had more elements than the reference standards hindering positive phase identification.