Saccharides are the major constituents of many herbs, and they are often utilized as quality indicators of many botanical drugs, such as Chinese medicines. A method for the rapid determination of saccharides in the in-process extract solutions is beneficial for process monitoring and ensuring consistency in the quality of the end-products during the manufacturing of Chinese medicines. In this work, a method based on Raman spectroscopy and a competitive adaptive reweighted sampling-partial least squares (CARS-PLS) model was established for the rapid quantification of saccharides. The accuracy and precision of this method were confirmed by employing one monosaccharide (glucose), one oligosaccharide (maltotriose), and two polysaccharides (Codonopsis radix polysaccharides and Polygonati rhizome polysaccharides) as reference substances. The determined results correlated well with the reference values of the four substances with the coefficient of determination of prediction (R_p²) ≥ 0.9939 and the root-mean-square error of prediction (RMSEP) ≤ 1.1052 mg/mL. Then, the method was applied to monitoring the simulated extraction process for Wenxin granule manufacture using total saccharides as a quality indicator. The CARS-PLS model exhibited satisfactory fitting and predictive capability, with R_p² and RMSEP values of 0.9743 and 1.4931 mg/mL, respectively. Our work demonstrated that Raman spectroscopy coupled with chemometrics can offer a reliable and nondestructive alternative for the determination of different types of saccharides, in addition to being useful for real-time monitoring of the extraction process during the manufacturing of Wenxin granules. The presented approach is expected to be applicable to other Chinese medicines.

糖是许多草药的主要成分，经常被用作许多植物药（例如中药）的质量指标。一种快速测定过程中提取液中糖类含量的方法，有利于过程监控和确保中药生产过程中最终产品质量的一致性。在这项工作中，建立了一种基于拉曼光谱和竞争性自适应加权加权采样最小二乘（CARS-PLS）模型的方法，用于糖的快速定量。通过使用一种单糖（葡萄糖），一种寡糖（麦芽三糖）和两种多糖（党参（Codonopsis radix）多糖和Poly子（Polygonati）的根茎）证实了该方法的准确性和精密度。多糖）作为参考物质。测定结果与四种物质的参考值具有很好的相关性，预测系数（R _p ²）≥0.9939，预测均方根误差（RMSEP）≤1.1052 mg / mL。然后，将该方法用于监测以总糖为质量指标的温心颗粒制造的模拟提取过程。CARS-PLS模型具有令人满意的拟合和预测能力，R _p ²和RMSEP值分别为0.9743和1.4931 mg / mL。我们的工作表明，拉曼光谱法与化学计量学结合可以为测定不同类型的糖类提供可靠且无损的替代方法，此外还可以用于实时监测文心颗粒生产过程中的提取过程。预期所提出的方法将适用于其他中药。