Vibrational spectroscopy for online monitoring of extraction solvent degradation products
In our research, we are exploring the potential of online monitoring of the organic solvents for the flowsheets relevant to the used nuclear fuel reprocessing and tributyl phosphate (TBP)- based extraction processes in particular. Utilization of vibrational spectroscopic techniques permits the discrimination of the degradation products from the primary constituents of the loaded extraction solvent. Multivariate analysis of the spectral data facilitates development of the regression models for their quantification in real time and potentially enables online implementation of a monitoring system. Raman and FTIR spectral databases were created and used to develop the regression partial least squares (PLS) chemometric models for the quantitative prediction of HDBP (dibutyl phosphoric acid) degradation product, TBP, and UO{sub 2}{sup 2+} extraction organic product phase. It was demonstrated that both these spectroscopic techniques are suitable for the quantification of the Purex solvent components in the presence of UO{sub 2}(NO{sub 3}){sub 2}. Developed PLS models successfully predicted HDBP and TBP organic concentrations in simulated Purex solutions.