Prediction of Sugar Content by Near-Infrared Spectroscopy during the Fermentation of Apple Juice with and without the Addition of Pomace
摘要
Cider is produced by fermenting apple juice, generating a by-product apple pomace (AP). In this work, apple juice fermentation was performed in both laboratory and bench scales, in systems with and without 10% AP. During fermentation, daily samples were taken and measured sugar content and NIR spectra. All systems followed a typical sugar consumption, and significant differences on sugar content were observed only in systems with AP. Partial least squares (PLS) calibration was performed considering all systems as one group to obtain robust models. The models and linearity improved by selecting the range from 1640 to 1710 nm using the interval PLS, obtaining R2 values between 0.81 and 0.86 in the prediction of total majority sugar, fructose, and glucose content. The study concludes that total majority sugars, fructose, and glucose can be accurately predicted using the developed linear models by measuring the NIRS on the two evaluated scales, both with and without the addition of AP. This study presents a novel application of NIRS combined with chemometric techniques to quantify fermentable sugars in cider fermentation under varying conditions, addressing a gap in literature by simultaneously exploring scaling, fermentation, and pomace addition in apple juice. While the current model is constrained to tracking this single fermentation control parameter, this work establishes an industrially relevant, scalable platform for apple waste valorization through fermentation, demonstrating both technical feasibility and production-ready operation. The adaptable framework enables future expansion to multi-parameter monitoring, creating opportunities for sustainable process optimization in beverage manufacturing.