Chemometrics-assisted High-order Instrument For Targeted And Non-targeted Analysis In Complex System

Modern analytical chemistry is characterized by two essential features,i.e.the instrumentation of analytical tools and the complexity of analytical objects,this brings the discipline to an unprecedented "data tsunami" era,chemometrics occurring at the right moment to provide a powerful tool for confronting these challenges.By decomposing the multi-way data collected from modern high-order instruments,the valuable information of targeted analytes can be exactly extracted out from heavily interferential but information-rich measurement data.By replacing and/or enhancing traditional "physical and/or chemical seperation”using smart "mathematical seperation",analysts can obtain relative high-efficient and green analytical strategies for accurate quantitative determination of targeted analytes in complex unknown systems.Acquiring prominent "second-or high-order advantage",which can avoid or significantly simplify laborious and time-consuming sample pretreatment processes and exclude the impacts of unknown background matrices interferences,makes direct,rapid,simultanous and accurate quantitative determination of targeted analytes in complex unknown system become possible.Recently,these strategies are increasely accepted and widely applied in the field of food safety analysis,environment minitoring,drug regulation,etc.Moreover,with the aid of the feature of relative uniqueness of decomposition of multi-way calibration methods with "second-or high-order advantage",an unprecedented analytical strategy can be developed and applied for exploring the laws of changes(e.g.changse of content and dynamic tendency)of unknown but interested factors and corresponding chemical stucture information in dynamic complex systems.Based on this,in this paper,two parts of work were performed and elaborated,i.e.rapid,simultanous and accurate determiantion of multi-targeted analytes in complex systems and non-targeted exploration of unknown but interested information in dynamic processes of complex systems by coupling chemometrics multi-way calibration methods with high-order instrument detection data.The main contents include the following 6 parts:Part I:Chemometric multi-way calibration methods coupled with high-order instrument measurement data for accurate quantitative analysis of targeted analytes in complex systemsIn Chapter 2,in the present work,a novel 'dilute-and-shoot' analytical strategy that combined self-weighted alternating normalized residue fitting(SWANRF)algorithm with excitation-emission matrix(EEM)fluorescence detection enhanced by photochemical derivatization(PD)was proposed for rapid,simultaneous and accurate quantitative analysis of aflatoxin B1(AFB1)and G1(AFG1)in complex foodstuff systems(including cereals,honey,and edible oil).By coupling the predominant"second-order advantage" of SWANRF algorithm with the ultra-sensitivity of EEM enhanced by off-line PD,specific quantitative information of both analytes could be successfully extracted out from heavily interferential matrices without complicated purification and chromatographic separation steps.Consequently,the whole analytical time and expense were significantly decreased,the accurate recoveries(with relative standard deviation,RSD)(93.5±6.6%?102.8±4.0%for AFB1,and 96.4±3.6%?107.2±6.0%for AFG1,respectively)and extremely low limits of detection(LODs)(0.12-0.21 ng mL-1 for AFB1,and 0.27-0.75 ng mL-1 for AFG1,respectively)were obtained for all four foodstuffs.In addition,quantitative results obtained from the proposed strategy were carefully compared with standard LC-MS method coupled with IAC clean-up for the further confirmation,which proved that SWANRF-assisted EEMs was excellent,and it is promising as an alternative analytical strategy for routine analysis of multiplex aflatoxins and the theoretical basis for developing portable detecting device for mycotoxin-analysis.In Chapter 3,a novel chemometrics-assisted high performance liquid chromatography-diode array detection(HPLC-DAD)method was proposed for rapid,accurate and simultaneous determination of five bio-active alkaloids,i.e.vincristine(VCR),vinblastine(VLB),vindoline(VDL),catharanthine(CAT)and yohimbine(YHB)in traditional Chinese medicine catharanthus roseus(C.roseus)and human serum samples.With the aid of prominent "second-order advantage" of alternating trilinear decomposition(ATLD)method,the resolution and rapid determination of five components of interest in complex matrices were performed even in the presence of heavy overlapped peaks and unknown interferences.Therefore,multi-step purification could be omitted and five components were fast eluted out within 7.5 min under a simple isocratic elution condition(acetonitrile/water containing 0.2%formic acid,37:63,v/v).Statistical parameters,e.g.linear correlation coefficient(R2),root-mean-square error of prediction(RMSEP),limit of detection(LOD)and limit of quantitation(LOQ)had been calculated to investigate the accuracy and reliability of the method.The average recoveries of five vinca alkaloids ranged from 97.1%to 101.9%and 98.8%to 103.0%in C.roseus and human serum samples,respectively.The five vinca alkaloids were adequately quantified with limits of detection(LODs)of 29.5-49.3 ng mL-1 in C.roseus and 12.4-27.2 ng mL-1 in human serum samples,respectively.The obtained results indicated that this analytical strategy provided a feasible alternative method for synchronously monitoring the quality of raw C.roseus herb and the concentration of blood drugs in clinic.In Chapter 4,in the present work,a novel chemometrics-assisted analytical strategy that combines three-way high performance liquid chromatography-diode array detection(HPLC-DAD)data with second-order calibration method based on alternating trilinear decomposition(ATLD)algorithm was developed for direct,accurate and simultaneous determination of thirteen phenolic compounds in complex red wine without intricate clean-up steps.All analytes were fast eluted out(7.5 min)under a simple gradient LC-separation and then detected in a multi-channel UV window.With the aid of prominent "second-order advantage" of ATLD algorithm,four common HPLC problems,i.e.solvent peaks,peak overlaps,unknown interferences and baseline drifts being mathematically calibrated,making "pure signals" of analytes could be extracted out from heavily interferential but information-rich HPLC-DAD profiles.The new strategy could avoid the loss of analytes of interest to significantly improve analytical accuracy.Validation parameters,i.e.recovery(97.7-104%),precision(RSD<7.1%),matrix effect,limits of detection(LODs,0.02-0.27?g mL-1)and limits of quantitation(LOQs,0.06-0.82?g mL-1)of thirteen analytes,were surveyed and further confirmed by standard LC-MS/MS method.Based on the indexes of phenolic compositions in wines,pattern recognition methods,i.e.principal component analysis and linear discriminant analysis(PCA-LDA)were applied for distinguishing wines of different vintage years,the discriminant accuracies were higher than 90%,which proved that this chemometrics-assisted HPLC-DAD strategy was an excellent method for direct,accurate and simultanous determination of phenolic compositions in complex wine samples as well as the authentication of vintage year.In Chapter 5,mycotoxins are a class of highly carcinogenic substances often naturally occurring in the moldy foods especially cereals.In the present work,a smart chemometrics-assisted analytical strategy that combines liquid chromatography-full scan-mass spectrometry(LC-MS)detection with second-order calibration method based on alternating trilinear decomposition(ATLD)algorithm was developed for direct,fast and interference-free determination of multi-class regulated mycotoxins in complex cereal samples with one-step ultrasound-assisted extraction.Ten mycotoxins with different property were fast eluted out(9.0 min)under a simple gradient LC-separation and detected by full scanning MS with a segmented fragment program.With the aid of prominent "second-order advantage" of the algorithm,the problems of co-eluted peaks,unknown interferences and baseline drifts occurring in the LC-MS profiles were mathematically resolved,making "pure signals" of targeted analytes can be extracted out from heavily interferential information.The new strategy avoided intricate physical/chemical clean-up steps to significantly improve the accuracy of the trace analysis of mycotoxins,average recoveries of ten analytes in both complex cereal samples(maize and rice)ranged from 93.8 to 109%with standard deviations(SD)lower than 9.8%,and the limits of detection(LOD)ranged from 0.01 to 1.17 p.g kg-1.In order to further confirm the reliability of the method,the same batch of samples was analyzed using LC-MS2 method with complicated immunoaffinity column(IAC)purification step,the elliptical joint confidence region(EJCR)tests of quantitative results of ten mycotoxins showed that higher precision was obtained using the proposed method.Therefore,this new strategy could be as an attractive alternative for simple,fast and accurate determination of multi-class mycotoxins in complex cereal samples.Part ?:Chemometric multi-way calibration methods coupled with the measurement data obtained from various high-order instruments for non-targeted analysis in dynamic complex systemsIn Chapter 6,a novel pattern recognition strategy based on two-dimensional(2-D)HPLC-DAD fingerprints resolved using chemometrics methods was developed and applied for the authentication of propolis of different varieties.By decomposing the three-way array of samples using alternating trilinear decomposition(ATLD)algorithm,"pure signals" of components of significant difference can be successfully extracted out from heavily interferential but information-rich 2-D fingerprints,including the normalized chromatographic matrix A,the normalized spectral matrix B and correspongding relative concentration matrix C,with the so-called "second-order advantage".Then,the resolved "pure signals”of components of significant difference were used for reconstructing the 1-D high-resolution fingerprints of samples,realizing not only the extention of low-resolution 2-D fingerprints at chromatographic dimension but also the compression of multi-wavelength spectral information at the other dimension.Invalid information,including the components of no significant difference,solvent peaks and baseline drift,were fitted into co-factors and removed away via the algorithm.Compared with traditional pattern recognition methods based on single-wavelength chromatographic fingerprints,the new method greatly reduced the difficulty of chromatographic separation and decreased anayltical time.With the aid of high-efficiency and green "mathematical separation" performed on the low-resolution 2-D fingerprints,the resolution of chromatographic separation was significantly improved and the multi-wavelength spectral information of samples was also comprehensivly enriched,thereby,the accuracy of pattern recognition can be significantly improved.Based on this strategy,2-D HPLC-DAD fingerprints of 94 propolis samples from different countries(including China,Brazil and Japan)were analyzed,resolved high-resolution 1-D fingerprints were used for unsupervised principal component analysis(PCA)and supervised linear discriminant analysis(LDA),the accuracy of LDA was up to 96.6%,and misdiscriminant rate was less than 3.2%.The accuracy was significantly higher than that of traditional methods based on single-wavelength fingerprint.Therefore,this strategy provided a high-efficiency and accurate identification tool for the authenticity of complex propolis system,and it was promising to be as an alternative method for the quality assessment of other complex food systems.In Chapter 7,the screening of bioactivity components in dynamic complex systems and their structural identification have been difficult to overcome in the field of analytical chemistry.In this paper,a smart chemometrics-assisted liquid chromatography-fullscan mass spectrometry(LC-MS)analytical strategy was developed and applied for non-targeted screening and structural identification of bioactivity components in the biodegradation process of flavonoids of propolis in serum.The real-time three-way LC-MS array data of biodegradation process of flavonoids of propolis were analyzed resolved using alternating trilinear decomposition(ATLD)algorithm,which could extract the "pure signals" of bioactive components out from heavily interferentical but information-rich LC-MS profile,i.e.the normalized chromatography,the normalized mass spectrometry and corresponding kinetic curves,non-bioactive components and baseline interfering signals were fitted into co-factors and removed away by various discriminant analysis,which was so-called "second-order advantage".Differring from traditional static screening methods of bioactivity components,the new non-targeted analysis strategy directly focused on the interaction process in dynamic complex systems,which used high-efficiency and eco-friendly "mathematical separation" to enhance traditional"physical/chemical separation" for deeply digging dynamic information of systems while shielding non-dynamic information.With the aid of "pure signals" resolved by ATLD algorithm,we can synchronously realize selectively chromatographic collection of unknown bio-active components,their molecular structural identification and kinetics study in dynamic complex systems.This strategy considerately improved the screening accuracy and efficiency of unknown bioactivity components in complex dynamic systems,and it provided a new way of thinking for exploration of complex unknown systems for analyst.