基于紫外可见光谱的红葡萄酒关键质量指标快速监测

对葡萄酒发酵和陈酿过程中质量进行有效监控，是葡萄酒品质的重要保证。通过监测葡萄酒发酵和陈酿过程中酚类物质和颜色的演变规律，以及紫外-可见吸收光谱特征，建立基于紫外-可见吸收光谱的酚类物质和颜色参数的偏最小二乘回归(Partial least squares regression, PLSR)模型。结果显示总酚、总单宁和总黄烷醇质量浓度在发酵过程中不断增加，在陈酿过程中逐渐减小。而总黄酮醇在发酵过程中先增加而后减小，在陈酿过程中逐渐减小。浸渍过程和发酵初期是颜色形成的关键时期，在陈酿过程中颜色逐渐老化。基于紫外-可见光谱的酚类物质和颜色参数预测模型的校正集决定系数R_cal²及验证集决定系数R_val²均不小于0.84,残余预测偏差(Residual predictive deviation, RPD)均不小于2.54,模型可以达到预测的目的。因此，紫外-可见光谱结合化学计量学是快速监测红葡萄酒发酵和陈酿过程中酚类化合物及色泽演变的一种简单经济且高效可行的手段。

Rapid Monitoring of Key Quality Indicators of Red Wines Based on UV-Vis Spectroscopy

Consumer demand for high-quality and safely produced wines requires high standards in terms of quality assurance and process control methods, which in turn requires proper analysis of wines during and after production. The role of phenolics and color characteristics in ensuring wine quality at all stages of the production process is gradually gaining recognition and importance. However, traditional chemical analysis methods required cumbersome pre-treatment, expensive instrumentation requirements, time-consuming determination procedures, and harsh operating conditions. The evolution of phenolics and color was monitored, as well as UV-Vis absorption spectral characteristics, and a partial least squares regression (PLSR) model was developed for phenolics and color parameters based on UV-Vis spectroscopy during wine fermentation and aging. The results showed that the content of total phenols, total tannins and total flavanols were increased continuously during fermentation and decreased gradually during aging. In contrast, total flavonoids were increased and then decreased during fermentation and gradually decreased during aging. In addition, the maceration process and the early stage of fermentation were the critical periods for color formation, while the color was gradually aged during the aging. The calibration set correlation coefficients R2cal and validation set correlation coefficients R2val of the UV-Vis spectroscopy-based phenolic and color parameter prediction models were both no less than 0.84, and the residual predictive deviation (RPD) was no less than 2.54, thus the models could achieve the prediction purpose. Therefore, UV-Vis spectroscopy combined with chemometrics was a simple, economical and efficient way to monitor phenolic compounds and color evolution during red wine fermentation and aging.