The valorization of glycerol wastes through its conversion into high-added value products using biotechnological processes is very promising due to the high amount of glycerol generated from the bio-diesel industry. In this sense, it is important to develop rapid, reliable and, if possible, environmentally friendly analytical techniques for the monitoring of high-added value compounds. With this in mind, Fourier transform near infrared (FT-NIR) spectroscopy was explored for the assessment of biomass, glycerol, l-phenylalanine and 2-phenylethanol (2-PE) parameters during the biotechnological production of 2-PE. A total of 80 samples were collected and analyzed through FT-NIR spectroscopy. Partial least squares (PLS) regression method was selected for establishing quantification models. The obtained figures of merit, in terms of, coefficients of determination of prediction (R2P), range error-ratios (RER) and residual predictive deviation (RPD) yielded values higher than 0.85, 7.7 and 2.6, respectively, for all the parameters assessed. The obtained results reveal that all the developed PLS models can be considered as models with a good predictive accuracy and therefore attest the suitability of FT-NIR spectroscopy for the quantification of these parameters. Concluding, a new rapid, reliable, cost-effective, accurate, multi-parametric and environmentally friendly analytical technique exploring the potential of FT-NIR spectroscopy was successfully developed.

The authors thank the Portuguese Foundation for Science and Technology (FCT) under the scope of the strategic fundings of UIDB/ BIO/04469/2020 and UIDB/QUI/50006/2020 units. Ricardo N.M.J. Páscoa and Cristina Quintelas thanks FCT for funding through program DL 57/2016 – Norma transitória. The authors also want to express their gratitude to Prio Energy – Prio Biocombustíveis, SA for providing the glycerol used in this experimental work.

info:eu-repo/semantics/publishedVersion