Quantification of Silicate Minerals by SEM Based Image Analysis

An analytical program using a combination of SEM-based image analysis techniques is developed for automated modal analysis of ores containing silicate and other gangue minerals. Previous knowledge, gathered by means of either optical microscopy, X-ray diffraction, SEM, or electron microprobe analysis is used to establish discrimination criteria in the image analysis program. The program initially discriminates minerals on the basis of their relative grey values in a backscattered electron image. For those minerals which have overlapping grey levels the minerals are identified by the program using energy dispersive X-ray spectrometry with X-ray counting for twelve or more elements. This technique is useful to quantify ores in which the gangue minerals are numerous. Two cases of application are presented.