Research On Rapid Analysis And Detection Of Organophosphorus Pesticide Residues Based On Spectral Absorption/ Surface-Enhanced Raman Spectroscopy

As the poisoning and trade disputes, which are caused by organophosphorus pesticide residues, occur frequently in recent years, the rapid and batch detection of organophosphorus pesticide residues is so urgent for protecting people’s health and national economy from damage. Traditional detection methods such as chromatography, mass spectrometry, chromatography-mass and capillary electrophoresis are unable to achieve the above goal, but the combination of the emerging enzyme inhibition-spectral absorption and surface-enhanced Raman spectroscopy (SERS) can realize the rapid screening, qualitative and quantitative analysis of organophosphorus pesticide residues and achieve the goal. In the actual detection, some problems still exist. The enzyme inhibition-spectral absorption is for the rapid screening of organophosphorus pesticides, the stability and function of the existing instruments need further improvement. And SERS spectroscopy is for the qualitative and quantitative analysis of pesticides, but some problems such as poor stability and large noise need to solved. Additionally, the spectral analysis must depend on the professional personnel and knowledge, so it is not conducive for popularization and application of the technology.Focusing on the above problems, the dissertation discusses the related theory of the enzyme inhibition-spectrum absorption and SERS spectroscopy for organophosphorus pesticide residues detection and carry out the two following research work. Firstly, basing on the theory of the enzyme inhibition-spectral absorption, a novel and rapid screening device of organophosphorus pesticide residues is designed through the improvement of the optical structure and optimization of the weak signal processing and software design. Secondly, the chemometrics methods are applied for the SERS spectroscopy of organophosphorus pesticides to promote stability analysis, remove noise and simplify spectral analysis process, and then the qualitative and quantitative analysis of pesticides are achieved.The work includes the following specific content:1. The spectroscopic structure and switch modulation of the single laser are adopted to eliminate the interference of multiple light sources and improve the detection stability. Then, the embedded system and QT architecture are used to design the user software of friendly-interface and fully-function. The large capacity SD card and network module are also integrated into the instrument, so the local storage and network transmission of the detection data are easily performed.2. Basing on the designed detection device, the experiments of the performance assessment, and detection of organophosphorus pesticide residues in the vegetables and soil are carried out. The performance assessment experiment shows that the repeatability, stability and detection limit of the device are both well. And in the experiments of vegetables and soil, the related detection results of the device have good agreements with the results of the laboratory analysis equipment: spectrophotometer and gas chromatograph, and the stability is better than the commercial rapid detection instruments. In conclusion, the designed device can realize the screening of organophosphorus pesticides effectively.3. For the purpose of the qualitative and quantitative analysis of the pesticides further, the chemometrics methods are used for the processing of the SERS spectroscopy of organophosphorus pesticides. Through discussing the issues on the baseline deduction, clutter filtration, spectrum correction, noise smooth, feature extraction and variable selection, classification and regression model development systematically, a complete system is developed which is for the rapid qualitative and quantitative analysis of organophosphorus pesticide based on the SERS spectroscopy.4. For the practical problems in the SERS spectral analysis, some superior and novel algorithms are proposed. In clutter filtration, many impure peaks are well eliminated in the pesticides spectrum by comparing with the overall trend and average value of spectra. And the novel fusion classification algorithm which combines boosting algorithms with SVM is adopted for the qualitative detection of the structural analogues (methyl parathion, ethyl paraoxon), the classification accuracy is further prior to the original SVM. In variable selection, the adaptive genetic algorithm is introduced for the wavelength selection on the basis of PLSR, and the quantitative detection error is significantly reduced.