A Methodology to Estimate Concentration Profiles from Two-Dimensional Covariance Spectroscopy Applied to Kinetic Data

Abstract

Two-dimensional (2D) covariance analyses applied to spectroscopy data obtained while monitoring chemical processes allow exploring the chemical reactions involved. Some slices of the sample–sample covariance map are an approach to the concentration profiles of the reactive species and we present a novel methodology to identify them. The method overcomes problems habitually referenced in the application of this technique and it is based on the selection of spectral zones with similar standard deviation of the variables and row centering the spectra data in each zone. The slices are identified according to the correlation coefficient value. The method is illustrated using simulated spectra data set representatives of two model reactions A → B and A → I → B. It has been applied to analyze the effect of rare earth metal triflate initiators in the cationic curing process of diglycidyl ether of bisphenol A with γ-valerolactone. The number of significant slices found is equal to the number of reactive species. This is interesting information that can be used as an initial estimation to find profile concentrations using other methods such as multivariate curve resolution–alternating least squares (MCR-ALS).