Research On The Construction Of Three-dimensional Correction Model And Anti-interference Quantitative Application Of UV-visible Spectroscopy And Its Use As Chromatographic Detection Signa

Chemometrics is an interdisciplinary subject combining statistics and chemistry.Chemometrics has been a practical field of analytical chemistry since its early days.It is a discipline that solves problems neglected by mainstream statisticians.Chemometrics can be widely defined as the use of mathematical and statistical protocols to analyze chemical data,usually using multivariate analysis methods.These methods include many mathematical methods,models and algorithms.In the field of chemometrics,chemical multiway calibration is one of the research focuses and hotspots.This dissertation focuses on the new quantitative analysis application of ultraviolet-visible spectroscopy combined with multiway calibration and response surface modeling.The main contents are as follows:In Chapter 2,Aromatic amino acids play an extremely important role in life activities and participate in many biological processes.Their concentration levels are associated with a variety of diseases,such as phenylketonuria and colorectal cancer.Therefore,the quantification of aromatic amino acids is an important task.In the present work,a novel and rapid three-way analytical method was proposed to detect the levels of aromatic amino acids in prostate cancer cells（PC3 cells）and Dulbecco’s modified minimal essential medium（DMEM cell culture）,by using the affordable ultraviolet-visible spectroscopy.Firstly,spectrum-pH second-order data were designed per sample;Secondly,properties of the resulted spectrum-pH-sample three-way data were investigated by utilizing the parallel factor analysis（PARAFAC）,alternating trilinear decomposition（ATLD）,and constrained alternating trilinear decomposition（CATLD）algorithms,and a flexible scanning approach for determining the constraint parameters of CATLD was proposed;Thirdly,a three-way calibration method based on the CATLD algorithm with the proposed scanning approach was developed for interference-free quantification of aromatic amino acids in these systems.The average relative predictive errors of validation（ARPEV）for phenylalanine,tyrosine,and tryptophan were 1.4%,3.0%,and 0.7%in prostate cancer cells,and ARPEV for phenylalanine,tyrosine,and tryptophan were 4.1%,1.2%,and 0.7%in DMEM cell culture.The predicted contents of tyrosine and tryptophan in DMEM cell culture were 64.2±2.9μg m L^-1,5.6±0.3μg m L^-1,there are no significant differences in the concentrations between the developed analytical method and high performance liquid chromatography method.The proposed spectrum-pH-sample three-way calibration method based on CATLD algorithm can provide an interesting analytical strategy with high selectivity and accuracy for ultraviolet-visible spectroscopy.In chapter 3,a Box-Behnken design optimization method was used to establish an extraction method for carotenoids in Solanaceae plants.Firstly,based on the single factor method,the extraction solvent,liquid-solid ratio,extraction time,extraction temperature and other factors were studied.Then,based on this,the extraction temperature,extraction time and ratio of liquor to material were used to carry out the three-factor and three-level center combination design.The response surface method was used to optimize the extraction parameters of carotenoids in tobacco leaves,and then the mathematical model was constructed and the experimental test was carried out.The optimal extraction conditions given by the model are as follows:the ratio of liquor to material is 20:1（m L:g）,the extraction temperature is 45℃,and the extraction time is 25 min.The predicted value of the extraction amount under this condition is 8.36 mg/L.The actual extraction amount obtained by the experiment under this optimal process condition is 8.29±0.40 mg/L（at 95%confidence level）,and the relative error is only 0.84%.In Chapter 4,carotenoids are a class of natural pigments widely found in animals and plants.They are composed of a series of conjugated isoprene units,which are highly conjugated through double bonds,making them difficult to identify and quantify.Carotenoids have rich effects and are the main source of vitamin A in the body.At the same time,they also have anti-cancer,anti-oxidation,immune regulation,and anti-aging effects.Therefore,the quantification of carotenoids is an important task.Based on HPLC-DAD,a novel and rapid three-way calibration method was proposed for the rapid detection of six carotenoids（α-carotene,β-carotene,lutein,zeaxanthin,lycopene and astaxanthin）in the three solanaceae plants（tomato,eggplant and tobacco）.Even in the above-mentioned complex systems where interference coexists,there is serious chromatographic overlap,and there is a slight drift in retention time.The proposed analysis method can still achieve anti-interference quantitative analysis of six carotenoids.This method enhances chromatographic separation through three-way calibration of mathematical sub-functions,which can save time,simplify the requirements for complete chromatographic separation of samples,reduce the use of organic reagents,and achieve a new development concept of green environmental protection.The calibration set,validation set and prediction set ofα-carotene,β-carotene,lutein,zeaxanthin,lycopene and astaxanthin showed satisfactory results.The content of lutein in tobacco leaves was predicted to be 498.8±22.6μg m L^-1（Y28-Ⅰ）,267.1±12.8μg m L^-1（Y28-Ⅱ）,311.1±1.6μg m L^-1（Yunyan 87-Ⅰ）and 314.0±10.2μg m L^-1（Yunyan 87-Ⅱ）,and the content of lycopene in tomato was predicted to be 95.1±4.2μg m L^-1.The traditional HPLC method was used as a control.F test and t test showed that at a 95%confidence level.There was no significant difference in precision and accuracy between the proposed method and the traditional HPLC method.The above results show that the proposed method is an alternative rapid analysis method for rapid quantitative determination of carotenoids in complex systems.