传统的农药残留检测为理化检验方法, 具有前处理过程繁琐、 耗时、 复杂等不足, 以目前农业上使用较广的毒死蜱作为研究对象, 提出了一种有机磷农药残留快速定量检测方法。 首先, 根据毒死蜱的化学特性, 综合考虑比色剂的显色效果以及二次污染问题, 确定在弱碱性环境下以间苯二酚对毒死蜱进行显色反应的预处理方案; 然后, 通过分析0.5 ～400 mg·kg-1之间毒死蜱样品的紫外-可见光光谱数据, 确定了显色反应后的特征信息主要集中在365～420 nm之间。 接着, 以偏最小二乘法构建全谱预测模型, 其校正相关系数达到0.999 6, 预测相关系数达到0.995 6, 校正标准差RMSEC为2.814 7 mg·kg-1, 验证标准差RMSEP为8.012 4 mg·kg-1; 提取400 nm为中心波段的特征区域构建预测模型, 其校正相关系数达到0.999 6, 预测相关系数达到0.999 3, 校正标准差RMSEC为2. 654 6 mg·kg-1, 验证标准差RMSEP为3.465 5 mg·kg-1。 最后, 通过分析0.5～16 mg·kg-1之间毒死蜱样品的近红外光谱数据, 发现其显色功能团的特征不是很明显, 但会引起间苯二酚本身5 200 cm-1处吸收峰的变化。 实验结果证明了所提出用于快速定量预测有机磷农药残留的方法是有效可行的, 该方法通过比色剂的显色加强有机磷农药在全光谱特别是紫外-可见光范围内的信息, 为实现农产品农药残留的快速检测提供了一条新途经。

The methods of physical-chemical inspection is adopted in the traditional pesticide residue detection, which require a lot of pretreatment processes, are time-consuming and complicated. In the present study, the authors take chlorpyrifos applied widely in the present agricultural field as the research object and propose a rapid and quantitative detection method for organophosphorus pesticide residues. At first, according to the chemical characteristics of chlorpyrifos and comprehensive chromogenic effect of several colorimetric reagents and secondary pollution, the pretreatment of the scheme of chromogenic reaction of chlorpyrifos with resorcin in a weak alkaline environment was determined. Secondly, by analyzing Uv-Vis spectrum data of chlorpyrifos samples whose content were between 0.5 and 400 mg·kg-1, it was confirmed that the characteristic information after the color reaction mainly was concentrated among 360～400 nm. Thirdly, the full spectrum forecasting model was established based on the partial least squares, whose correlation coefficient of calibration was 0.999 6, correlation coefficient of prediction reached 0.995 6, standard deviation of calibration (RMSEC) was 2.814 7 mg·kg-1, and standard deviation of verification (RMSEP) was 8.012 4 mg·kg-1. Fourthly, the wavelengths whose center wavelength is 400 nm was extracted as characteristic region to build a forecasting model, whose correlation coefficient of calibration was 0.999 6, correlation coefficient of prediction reached 0.999 3, standard deviation of calibration (RMSEC) was 2.566 7 mg·kg-1, standard deviation of verification (RMSEP) was 4.886 6 mg·kg-1, respectively. At last, by analyzing the near infrared spectrum data of chlorpyrifos samples with contents between 0.5 and 16 mg·kg-1, the authors found that although the characteristics of the chromogenic functional group are not obvious, the change of absorption peaks of resorcin itself in the neighborhood of 5 200 cm-1 happens. The above-mentioned experimental results show that the proposed method is effective and feasible for rapid and quantitative detection prediction for organophosphorus pesticide residues. In the method, the information in full spectrum especially UV-Vis spectrum is strengthened by chromogenic reaction of a colorimetric reagent, which provides a new way of rapid detection of pesticide residues for agricultural products in the future.