Determining chemical characteristics of plutonium solutions using visible spectrometry and multivariate chemometric methods

Abstract

Two chemometric analysis approaches for rapidly screening samples are presented. The first method is for determining Pu(III) and nitric acid concentrations by using the multivariate calibration technique of partial least squares (PLS) regression. Quantitation of plutonium using its visible spectrum is straightforward; however, the effects of nitric acid on the Pu(III) absorption spectra are subtle, and nitric acid quantitation from the absorbance spectrum is more difficult. In this study PLS regression is successfully applied to quantitate both plutonium and nitric acid by using the information contained in the absorption spectra of appropriate solutions. Evaluation of the calibration models, using test samples that span the range of the calibration concentrations, gave predictions consistent with the standard error of the calibration models. Secondly, pattern recognition methods are used to investigate the effects of various amounts of nitric acid, fluoride, or oxalate on visible spectra of Pu(IV) solutions. The methods enable qualitative estimates of the solution composition, which can potentially be used to adjust solution properties to desired specifications. The main pattern recognition methods employed are nearest neighbor classification and principal components analysis.