Application of Fourier transform-mid infrared reflectance spectroscopy for monitoring Korean traditional rice wine ‘Makgeolli’ fermentation

Abstract

This study was performed to examine the application of Fourier transform-mid infrared (FT-MIR) spectroscopy to rapidly and non-destructively measure the quality of Makgeolli during fermentation process. A model using the entire ranges of the spectra and a Partial Least Squares Regression (PLSR) model using specific wavelength ranges closely associated with Makgeolli quality were compared to investigate the appropriate non-destructive monitoring method for Makgeolli production. In the global PLSR model using the entire range of the spectra and the local PLSR model using two separate regions (functional group spectral region and fingerprint spectral region), the optimal prediction model for alcohol concentration was from 2nd derivative of Savitzky-Golay pretreatment. The coefficient of determination of the developed model was 0.984, and standard error of prediction (SEP) was 0.595%. Reducing sugar was detected best with 1st derivative of Norris-Gap pretreatment, showing the determination coefficient of 0.983 and SEP of 0.579%. Titratable acidity was accurately predicted with 1st derivative of Savitzky-Golay pretreatment, showing the determination coefficient of 0.936 and SEP of 0.026%. This study showed that FT-MIR spectroscopy can be utilized to monitor the quality changes of Makgeolli.