目的 构建自动化柱色谱系统，探究低成本的物理量传感器电导率、pH值和氧化还原电位（oxidation-reduction potential，ORP）在色谱过程检测中的应用。方法 系统集成蠕动泵、液位反馈控制模块及在线检测模块，通过LabVIEW软件实现实时数据采集与过程调控。以银杏叶水沉液为色谱处理对象，分析物理量与乙醇体积分数及成分含量的相关性。结果 银杏内酯C与电导率呈强正相关（r＝0.92），乙醇体积分数与电导率（r＝−0.85）、ORP（r＝−0.83）显著负相关。采用多元回归模型关联了色谱出口液体中乙醇体积分数、银杏内酯C含量和物理量。模型用于验证实验，乙醇体积分数所得平均绝对偏差（average absolute deviation，AAD）为4.45%，银杏内酯含量所得AAD为0.65%。自动化系统显著提升数据采集效率，每小时可以获得1 800组物理量数据，为后续挖掘色谱过程规律提供基础。结论 研究证明在色谱柱后接物理量传感器采集数据有利于精准收集色谱液部分指标成分及乙醇，具有较好应用前景。

Objective To construct an automated chromatography system and investigate the application of low-cost physical sensors electrical conductivity, pH, and oxidation-reduction potential (ORP) for process monitoring. Methods The system was composed with a peristaltic pump, liquid level feedback control module, and online detection module, with real-time data acquisition and process control facilitated by LabVIEW software. Ginkgo biloba leaf water suspension was employed as the target for chromatography processing. The correlation between physical parameters and ethanol concentration, as well as component content, was investigated. Results Ginkgolide C exhibits a strong positive correlation with electrical conductivity (r = 0.92), whereas ethanol concentration displays significant negative correlations with electrical conductivity (r = −0.85) and ORP (r = −0.83). A multiple regression model was utilized to establish the relationship between ethanol concentration, ginkgolide C content, and physical parameters in the effluent from the chromatography column. The model was validated in an additional experiment, with the average absolute deviation (AAD) for ethanol concentration being 4.45%, and for ginkgolide C content, the AAD was 0.65%. The automated system markedly enhanced data collection efficiency, enabling the acquisition of 1 800 sets of physical parameter data per hour, thus laying a solid foundation for further exploration of chromatography process patterns. Conclusion The study demonstrates that employing physical quantity sensors after the chromatography column to collect data enables the accurate measurement of key components in the effluent, such as specific indicators and ethanol, which holds great potential for practical applications.

R283.6

浙江省科技计划项目（2022C03142）；金华市公益性技术应用研究项目（2023-4-188）