Simultaneous determination of pyruvate and acetate levels in xanthan biopolymer by infrared spectroscopy: effect of spectral pre-processing for solid-state analysis

A rapid, direct, and reagent-free procedure based on solid-state Fourier transform infrared spectroscopy (FT-IR) coupled with partial least squares (PLS) data analysis has been developed for simultaneous determination of pyruvate and acetate levels in a microbial xanthan biopolymer. The influences of various spectral pre-processing procedures were studied in order to eliminate effects caused by sample preparation. It was determined that the combination of first derivative and orthogonal signal correction pre-processing contributes to a significant increase in the predictive performance of PLS-1 regression models. By employing the wavenumber region 1320–1350 cm⁻¹ for pyruvate determination and 1500–1600 cm⁻¹ for acetate determination, the root mean square error of cross-validation (RMSECV) for pyruvate and acetate contents were obtained 0.13% and 0.29% w/w, respectively. Results of the proposed procedure for different real samples and those obtained by their reference methods were compared.