MATERIALS DISCOVERY: STRATEGIES TO DOWNSELECT RADIATION DETECTION MATERIALS USING PERFORMANCE-BASED METRICS

Materials informatic approaches offer new options for accelerating the development of semiconducting gamma detection materials. The ability to identify promising materials from large numbers of candidate material compositions and a priori curtail less fruitful inquiry is based upon the ability to construct robust models for property estimation. Informatic-defined approaches relate atomic-level composition and structure to performance-directed properties for property estimation. To downselect to a smaller number of more likely materials, the estimated potential of a candidate composition is assessed against a series of performance metrics for a semiconducting detection material. Design rules based on structure-property relationships and candidate analysis offer an assessment of property trade-offs and the validity of perceived design constraints. For the II-VI semiconductor class (e.g., CdTe + ZnTe ? CdZnTe2), the initial number of candidate compositions is over 700 which undergoes a 90% reduction upon application of density, band gap and mobility property screens. Many of these surviving materials lie outside the classical compositions associated with wide-band-gap semiconductors, and employ transition-metal or even group IA and IIA elements.