Rapid determination of ergosterol in grape mashes for grape rot indication and further quality assessment by means of an industrial near infrared/visible (NIR/VIS) spectrometer - A feasibility study

Perceptual quality and food safety of grape products like wine is generally impaired by grape rot. Commonly applied visual inspection for quality assessment remains subjective. However, conventional analytical determinations are tedious and time-consuming. In contrast, near infrared/visible (NIR/VIS) spectroscopy is an objective and rapid alternative. An industrial spectrometer especially designed for online process integration was applied in laboratory configuration to evaluate the feasibility of the prediction of ergosterol and 12 further quality-associated parameters in grape mashes. Calibration models were established, validated, and evaluated. Spectral ranges used were 450-850 nm (VIS) and 1050 1650 nm (NIR). The regression method applied was partial least squares (PLS) regression. Calibrations at best suitable for discriminating low and high values were obtained for gluconic, acetic, tartaric, and titratable acids contents, pH value, ethanol content, and laccase activity. Therefore, for these parameters limits of NIR/VIS spectroscopy may be revealed. Calibrations for semi-quantitative predictions or rough estimations were obtained for fructose, glucose, and malic acid contents. Best performance suitable for quantitative predictions was achieved for relative density, glycerol, and ergosterol contents. The industrial spectrometer proved to perform similar to a research laboratory spectrometer. Consequently, a future on-line implementation of NIR/VIS spectrometers in wineries may be a valuable and reliable tool for rapid grape quality assessment at reception. Particularly, ergosterol content prediction proved to be a promising measure for efficient and immediate quality control and management, respectively. Wine quality and food safety may be increased. Fair and incentive pricing of delivered grapes may be enabled. (C) 2014 Elsevier Ltd. All rights reserved.