Study On Matrix Effects In Plant Derived Food Applied In Fully Automated Pesticide Residue Analysis

In China,the workload of pesticide residue analysis is extremely high and rely on manual.With the development of making big data technology,intelligent analysis becomes the future development direction in multi-pesticide residue analysis.In the intelligent research of liquid chromatography-mass spectrometry(LCMS)method,it is necessary to solve the interference problem caused by matrix effects(MEs)first.MEs are the main cause of unreliable results.Matrix suppression effects affect the method sensitivity.ME values varied with different matrices,analytes,pre-treatment,chromatography,and mass spectrometry conditions.Therefore,for the intelligence of LCMS based analysis,it is essential of establish fully automated analytical methods with low matrix suppression levels and controlled ME variability to provide a basis for forming big data technology and establishing a fully intelligent pesticide multi-residue analysis method.In this work,the effects of matrix species,mass spectrometry methodology and mobile phase additives on the variability of MEs were systematically investigated based on plant foods and Qu ECh ERS pre-treatment,and the relationship between co-eluents and matrix suppression was analyzed to provide a basis for the optimal selection of controlled chromatographic mass spectrometry conditions.Then,by improving the existing robot pretreatment platform that is online coupled with UHPLC MS/MS system,establishing an automated standard solution preparation method,optimizing chromatographic conditions,segmented multiple reaction monitoring(s MRM)method,and micro solid-phase extraction(μSPE)clean-up method,a fully automated pesticide multiresidue analysis method was established and verified.This study provides a basis for subsequently establishing a large MEs database and comparative analysis,and also provides a basis for finally achieving intelligent pesticide multi-residue analysis.Firstly,the impact of matrix species and mass spectrometry methodology on MEs was studied under the Qu ECh ERS cleanup.Both full scan(TOF-MS scan)method using ultrahigh performance liquid chromatography-quadrupole time-of-flight mass spectrometry(UHPLC QTOF MS)in information-dependent acquisition(IDA)mode and multiple reaction monitoring(MRM)scan method using ultra-high performance liquid chromatography-tandem quadrupole mass spectrometry(UHPLC MS/MS)were validated.ME levels of 73 pesticides in 32 different matrix species were determined and compared under the two methods.The results showed that the significant MEs accounted for 59.5% and 31.6% in 32 matrix varieties under MRM and TOF-MS scan,respectively,showing mainly matrix suppression effects.19 and 33 pesticides were disturbed by significantly increased suppression compared to other matrices in bay leaf,ginger,rosemary,Amomum tsao-ko,Sichuan pepper,cilantro,Houttuynia cordata and garlic sprout respectively.The result indicated that the variability of MEs between matrix varieties was similar under two mass spectrometry methodologies.When the mass spectrometry methodology was changed from MRM to TOF-MS scanning,although the MEs were significantly attenuated for 24 pesticides,11 targets failed to reach the limit of quantification and there were nine cases where the signal of specific targets was completely suppressed by the matrix.With MRM scanning,MEs could be made less variable between matrix species by preferential ion pair selection,while avoiding the occurrence of strong matrix suppression.Secondly,the effect of mobile phase additives on matrix effects(MEs)was studied under the Qu ECh ERS cleanup.By comparing the levels of 244 ions of 68 pesticides in 16 types of matrix,it was found that changing the addition of 0.1% formic acid to the use of 5 m M ammonium formate alone or in combination significantly increased the proportion of significant MEs from 55.7% to 77.2% and 71.6% in the 16 matrices,and changed from matrix inhibition to promotion.The variability of MEs in pepper,bay leaf and Amomum tsao-ko mainly showed reduced matrix suppression,while the variability pairs of MEs in the other 13 substrates mainly showed a change from matrix suppression to matrix enhancement.At the same time,the types of differential ion pairs varied in the two groups of matrices.A total of10 pesticides were unable to avoid changes in MEs induced by mobile phase additives by optimizing ion pairs.These results suggest that the effect of mobile phase additives on MEs is matrix specific.In addition,the addition of 5 m M ammonium formate helped to attenuate matrix inhibition or enhance matrix promotion to achieve increased sensitivity.Thirdly,the effects of co-eluents on matrix suppression under the IDA method were further studied.Analysis of the total signal of co-extracted substances and matrix suppression levels revealed an increase in matrix suppression in a group of matrices where the total signal of co-extracted substances increased by 2.7 times.A total of 4007 qualitative components were obtained from 31 matrices,accounting for 3.5% to 7.9% of the total co-eluted signal,indicating that the co-eluted signal mainly originated from non-qualitatively identified substances.Through analyzing the relationship between the MEs of four pesticides(aphidicloprid,amitraz,dichlorvos and pyrimethanil)in soybean and the concentration and signal strength of four co-eluted isoflavones,it was found that the signal strength of coeluents had a stronger indication of matrix suppression.The sum of the peak areas of each ion at a time interval of 0.25 min was used as a variable,and the grouping results of the 31 substrates under TOF-MS scanning were used to establish an OPLS-DA model,which identified five differential co-eluted signal areas.The signal strength of four of these differential co-eluted signal areas showed negative correlation with the integral peak area of the differential pesticides between the tow matrix groups.These results suggested that the coeluted signal had some indication of substrate inhibition,and lower total signal strength of coeluted substances benefited the reduction of substrate inhibition.Research on the variability of MEs was conducted as the basis for the development and validation of a fully automated method for multi-residue analysis of pesticides.The traditional method was reconstructed into four independent experimental units executed by robots.By introducing two new functional modules,an automated working method of the robotic platform was established,consisting of a 37-step basic process and 4 solvent parameter calculation equations,enabling the robotic platform to independently perform automated standard solution preparation,micro solid phase extraction(μSPE)clean-up,and UPLCMS/MS detection.A method for the analysis of 325 pesticides under UPLC MS/MS with segmented multiple reaction monitoring(s MRM)scan was developed and optimized.The robot-made mixed standard solution with 664 ion pairs showed good linear correlation(R2 ≥0.98)and comparison with commercial mixed standards resulted in an RSD between 1.2%and 2.8%,indicating that the automated standard solution preparation method is accurate and reliable.The μSPE clean-up parameters were optimized to remove 71.8%-96.7% of the coeluent signal from the crude extract at 200 μL acetonitrile elution,while the recovery of90.7% of the pesticide was in the range of 70%-120% with RSD <15%,removing 71.8%-96.7% of the co-eluent signal from the crude extract.For red onion,pakchoi,black tea and king oyster mushroom,more than 317 pesticide targets had a good linear R2≥0.98 for the matrix-matched standard curves;the number of pesticide targets with LOQs matching the national standard MRL levels were 314,312,305 and 304,respectively;the number of pesticides with recoveries ranging from 70% to 120% and RSD≤15% at the four spiked levels were 252,247,238 and 263;the percentage of significant MEs were 19%,31%,64% and39%,respectively,with 172 of the target pesticides showing significant matrix enhancement in the king oyster mushroom-induced MEs.Compared to Qu ECh ERS,a 2.2-fold reduction in the level of significant matrix suppression effect was found in black tea under μSPE cleanup,while 66 pesticides were found to change from negligible MEs and moderate intensity of matrix inhibition to equal intensity of matrix promotion in four matrices under μSPE.These results indicate that the fully automated pesticide residue analysis method developed in the thesis has good performance and weak ME strength,and can achieve unattended calibrationautomatic clean-up-automatic injection-instrument analysis,which meets the basic requirements for a controlled chromatography-mass spectrometry system in establishing a database of MEs.