Chemometric prediction of alginate monomer composition: A comparative spectroscopic study using IR, Raman, NIR and NMR

The potential of using infrared (I R), Raman and near infrared (NIR) spectroscopy combined with chemometrics for reliable and rapid determination of the ratio of mannuronic and guluronic acid (M/G ratio) in commercial sodium alginate powders has been investigated. The reference method for quantification of the M/G ratio was solution-state H-1 nuclear magnetic resonance (NMR) spectroscopy. For a set of 100 commercial alginate powders with a M/G ratio range of 0.5-2.1 quantitative calibrations using partial least squares regression (PLSR) were developed and compared for the three spectroscopic methods. All three spectroscopic methods yielded models with prediction errors (RMSEP) of 0.08 and correlation coefficients between 0.96 and 0.97. However, the model based on extended inverted signal corrected (EISC) Raman spectra stood out by only using one PLS component for the prediction. The results are comparable to that of the experimental error of the reference method estimated to be between 0.01 and 0.08. (C) 2007 Elsevier Ltd. All rights reserved.