Quantification of PVC plasticizer mixtures by compact proton NMR spectroscopy and indirect hard modeling

Abstract

A novel approach is introduced for the fast, reliable, and low-cost recognition and quantification of plasticizers in plasticizers mixtures. It uses benchtop 1H NMR spectroscopy and indirect hard modeling, a mechanistic multivariate regression technique. The approach is demonstrated on five different PVC plasticizers having similar spectral signatures in proton NMR spectra. With only 16 scans per spectrum, i.e., 2 min 40 s measurement time, quantification limits down to 0.14 mg mL−1, or 0.35 wt% plasticizer in PVC, were achieved. Apart from the rapid data acquisition, the use of spectral hard modeling enabled the quantification of plasticizer mixtures while using only 4 to 6 training samples per component. Despite strongly overlapping signals in the NMR spectra, various plasticizers were differentiated and quantified, as exemplarily demonstrated for binary mixtures. A commercial PVC specimen with three different layers was also examined, confirming the applicability of benchtop NMR spectroscopy. Additionally, the use of the proposed method to validate official regulations concerning the plasticizer content in PVC is assessed. The presented results demonstrate that the combination of benchtop NMR and spectral hard modeling is a very promising analytical tool for rapid PVC plasticizer recognition and quantification with high analytical throughput. Moreover, the results indicate a high potential for benchtop NMR and spectral hard modeling for microchemical analysis, even for complex samples.