Time Series Photoelectric Detection Of Pesticide Residues Based On Paper-based Microfluidic Chip

In China,there are more than 100 thousand people are poisoned by accidental ingestion of fruits and vegetables containing pesticide residue every year,and the fatality rate is about 20%.The pesticide residues detection equipment at present cannot satisfy the demand for high precision and portability at the same time.Despite the variety of means of residue detection,there are still low-precision and high-cost technical bottlenecks.These problems are solved by the detection method of pesticide residues based on time series photoelectric detection using a paper-based microfluidic chip.The pesticide residues in the fruits and vegetables are hazardous for human health.Because of the severe threats,the government of China has made various rules to adjust the crop market-deal and food industry.According to those rules,the Maximum Residual Limit(MRL)for Imidacloprid is 0.5 mg/kg.And,the MRL value for both the highly toxic pesticides like phorate and avermectin is 0.01 mg/kg.This research is divided into following aspects: overview of pesticide residue,the basic principles of pesticide residue detection based on(μPADs),detail description of the seven-layer microfluidic paper chip,circuit design for pesticide residue detection system,experimental conditions and machine learning techniques,results and discussion as well as error analysis.Research is devoted on(μPADs)to get the optimum chip structure,examine the factors affect the internal structure of(μPADs)(such as sample diameter,micromixing mechanism,selection of paper materials,contact area,reaction time,temperature,and wavelength).Furthermore,the absorption of light and enzyme inhibition reaction is analyzed and modelled.Different techniques used for preprocessing the data,like,Multiplicative Scatter Correction(MSC)is used to purge the noise points from the data.Translational Processing(TP)is used to improve the performance of pesticide identification system.Then,Principal Component Analysis(PCA)is used for dimension reduction by focusing mostly on the variations.Then,Linear Discriminant Analysis(LDA)used to improve the separability of different pesticides.After pretreatment of figures,we mainly targeted on the reflected light variations aroused through the enzyme inhibition assay.So,the linear discriminant analysis showed satisfying discrimination of imidacloprid(Y = 1.6525 X 139.7500),phorate(Y= 3.9689 X 483.0526),and avermectin(Y = 2.3617 X 28.3082).The correlation coefficients for these linearity curves were 0.9635,0.8093,and 0.9094,correspondingly,with a 95 % limit of settlement.Then,the avermectin class chemicals and real-global samples(i.e.,lettuce and rice)examined,which showed possible graphic consequences to distinguish all the chemical substances.Therefore,it is far feasible for classification of three sorts of pesticides by using the modifications in the reflected light spectrum in time sequence analysis with an extraordinary level of automation and integration,good reliability,and contribution to the further development in the future.