Evaluation of copper compounds using automated multivariate statistical phase analysis of EDX spectral imaging data

Abstract

In ferrous metal smelting operations, each element is refined from very complex compound material in which the major elements are Cu , Zn and Pb. Especially in smelting Cu, it is very important to understand the complex morphology of the various compounds in the raw material because of its very diversity. This detailed understanding of the complex morphology is necessary for the high purification process of each element and to improve the refining efficiency. Currently, ICP, XRD and EPMA are used for the evaluation of the raw materials. However, ICP and XRD can provide information only about average composition and EPMA requires highly experienced analysts to operate the instrument and evaluate a mixed compound material. In this study we analyzed Cu-compound raw material by the auto phase analysis method using the multivariate statistical analysis of EDX spectral imaging data (Thermo Scientific COMPASS). We analyzed the phases which were extracted objectively by present method mentioned above. As the results of the analysis of the EDX spectral imaging data acquired for 30 minutes, we could confirm the 15 phases including Cu-Sn alloy (as the major component), Cu sulfide and compounds with rare elements such as Sb and As, and we recognized that those phases are mixed with the typical distribution.