Application of near-infrared spectroscopy to sensor based sorting of a porphyry copper ore

Test work was conducted to define the possibilities and limitations of a near-infrared (NIR) sensor to distinguish between Cu grades of 150 porphyry Cu ore samples. NIR spectroscopy allowed determination of NIR active minerals. At porphyry systems, NIR active minerals are usually alteration minerals that are produced by a specific type of hydrothermal alteration. f the Cu-bearing minerals that were determined from the sample set by petrography produced diagnostic absorptions in the NIR. Because of this, no spectral features were present in the measured NIR spectra of the ore samples that relate directly to the Cu grade. irect relation was present between the Cu grade and the NIR active mineralogy that was determined from the NIR response of the samples. This indirect relation was based on sample groups with different NIR active mineral assemblages that represented different zones of hydrothermal alteration and constituted different ranges of Cu grades. logistic regression resulted that based on the NIR response, it was possible to reliably estimate a probability that the Cu grade of a sample is below the cut-off grade for economic processing of the ore. The predictors used in this regression were several characteristic spectral features of the NIR active mineralogy.