1H NMR spectra modeling for predicting the acid, saponification, and iodine values of bio-waxes derived from vegetable oils

Abstract

Bio-waxes from vegetable oils are of recent interest due to the high availability of raw materials and their applications in the pharmaceutical, cosmetic, and food industries. The acid (AV), saponification (SV) and iodine (IV) values in bio-waxes are critical properties for strict quality control. In this work, the AV, SV and IV were predicted from ¹H NMR spectra and chemometric methods including Partial Least Squares Regression (PLSR) and Principal Component Regression (PCR). PLSR yielded the best results with R² above 0.95 for the three properties, and root-mean-square error of cross-validation (RMSECV) below 2 mg KOH/g, 5 mg KOH/g, and 3 g I₂/100 g for the AV, SV and IV, respectively. The developed models rely on both compositional and structural information, particularly on the alkyl chain length negative contribution and the RCOOH, RCOOR¹ and RCCR¹ groups positive contribution. The proposed strategy offers a rapid, low-cost, and greener alternative to conventional methods.