College of Food Science and Engineering, Jiangxi Agricultural University, Nanchang 330045, People's Republic of China.
Jiangxi Sericulture and Tea Research Institute, Nanchang 330043, China; Jiangxi Key Laboratory of Tea Quality and Safety Control, Nanchang 330043, China.
Spectrochim Acta A Mol Biomol Spectrosc. 2021 Apr 5;250:119366. doi: 10.1016/j.saa.2020.119366. Epub 2021 Jan 2.
Surface enhanced Raman spectroscopy based on rapid pretreatment combined with Chemometrics was used to determine chlorpyrifos residue in tea. Au nanoparticles were used to as enhance substrate. Different dosages of PSA and NBC were investigated to eliminate the tea substrate influence. Competitive adaptive reweighted sampling (CARS) was used to optimize the characteristic peaks, and compared to full spectra variables and the experiment selected variables. The results showed that PSA of 80 mg and NBC of 20 mg was an excellent approach for rapid detecting. CARS - PLS had better accuracy and stability using only 1.7% of full spectra variables. SVM model achieved better performance with R = 0.981, RMSEP = 1.42 and RPD = 6.78. Recoveries for five unknown concentration samples were 98.47 ~ 105.18% with RSD - 1.53% ~ 5.18%. T-test results showed that t value was 0.720, less than t0.05,4 = 2.776, demonstrating that no clear difference between the real value and predicted value. The detection time of a single sample is completed within 15 min. This study demonstrated that SERS coupled with Chemometrics and QuEChERS may be employed to rapidly examine the chlorpyrifos residue in tea towards its quality and safety monitoring.
基于快速预处理和化学计量学的表面增强拉曼光谱法用于测定茶叶中的毒死蜱残留。使用金纳米粒子作为增强基底。研究了不同剂量的 PSA 和 NBC 以消除茶基质的影响。采用竞争自适应重加权采样（CARS）对特征峰进行优化，并与全光谱变量和实验选择变量进行比较。结果表明，PSA 为 80mg 和 NBC 为 20mg 是快速检测的理想方法。仅使用全光谱变量的 1.7%，CARS-PLS 具有更好的准确性和稳定性。SVM 模型具有更好的性能，R=0.981，RMSEP=1.42，RPD=6.78。对于五个未知浓度样品的回收率为 98.47%105.18%，RSD 为 1.53%5.18%。t 检验结果表明，t 值为 0.720，小于 t0.05,4=2.776，表明真实值和预测值之间没有明显差异。单个样品的检测时间在 15 分钟内完成。本研究表明，SERS 结合化学计量学和 QuEChERS 可用于快速检测茶叶中的毒死蜱残留，以监测其质量和安全性。
