Applications Of Trilinear Decomposition Algorithm To HPLC-DAD For Simultaneous Multicomponent Quantitative Analysis

High performance liquid chromatography (HPLC) is one of the most useful analytical techniques. Its chief limitation lies in the need to avoid the overlap of chromatographic peaks as much as possible. However, chemometrics can solve this problem. In this thesis, the developments of multivariate resolution and calibration in chemometrics, especially the alternating trilinear decomposition algorithms for the three-way data analysis and their applications have been reviewed. The methods for three-way trilinear decomposition have been applied to the second-calibration using high performance liquid chromatography-diode array detection (HPLC-DAD). The problems arising from the serious overlapping of spectra and chromatograms are successfully resolved during the analysis of environment contamination and drugs and so on. The assume replacing or improving the "chemical separation" with "mathematical separation" and the charming of three-way data analysis from "second-order advantage" is exhibited in second-order calibration. In Chapter 1, at first, the developments and applications of HPLC have been summarized, and from its shortcoming we introduced the chemometrics. Afterward, the developments of multivariate resolution and calibration in chemometrics, especially the alternating trilinear decomposition algorithms for the three-way data analysis and their applications have been reviewed. The overlapped chromatograms and spectra system of nitrobenzenes (1-chloro-2,4-dinitrobenzene and 3,5-dinitrobenzoic acid) in the presence of 2,4-dinitrobenzoic acid were resolved by ATLD combined with HPLC-DAD in Chapter 2. It successfully solved such problem caused by overlapped chromatograms and spectra even in the presence of unknown interfering components. It uses replacing or improving the "chemical separation" with "mathematical separation", and the analytical performances is improved. It also revealed the ATLD algorithm is one of the approaches to solve the second-order calibration problem in HPLC-DAD. In Chapter 3, we adopted simply eluting solutions and short separate time as the separate condition of HPLC-DAD and combined with ATLD algorithm to determine simultaneously and fast the two glucocorticoid drugs. Because of similarity of their physical and chemical characteristics, they have similarity spectra and adjacent elute time. In general, it is difficult to reach complete separation of them. However, ATLD method combined with HPLC-DAD can be used to determine simultaneously two glucocorticoid drugs in the presence of unknown interferences which exist in the urine. The prediction results are consistent with the actual one. It is indicated that chemometrics is a promising method to apply in analytical chemistry, biology and medicine analysis. In Chapter 4, two local anaesthetic drugs have been analyzed quickly. In the condition of overlapped chromatogram peaks, usualize a new algorithm: alternating fitting residue (AFR) method for resolving partly separated peaks into the pure chromatographic, spectral and concentration profiles even with very low levels of separation by analyzing HPLC-DAD data. The overlapped chromatograms and spectra arising from low separation level that is difficult to resolve have been settled by HPLC-DAD combining with AFR algorithm and simultaneous qualitative and quantitative analysis in the presence of unknown interfering components. It has been replace or improve the "chemical separation" with "mathematical separation". By our treating real HPLC-DAD data sets, the results demonstrated that the proposed algorithm performs well as long as the model dimensionality chosen is not less than the number of actual components and avoid the problem of so-called two-factor degeneracy and have lower convergence rate compared with PARAFAC method. The study proved that AFR algorithm is a powder chemometric tool for solving partly separate peaks. In Chapter 5, two anti-tuberculosis drugs (isoniazid and pyrazinamide) have been determined simultaneously by using alternating fitting residue (AFR) method combined with HPLC-DAD. Comparing to alternating trilinear decomposition (ATLD) and self-weighted alternating decomposition algorithm (SWATLD), the predicted results AFR resolved is closer to actual one. The experiment showed that the proposed method can successfully solved such problem that is difficult to resolved, caused by chromatograms and spectra overlapping and quantitative analysis fast and exactly. In Chapter 6, four vitamin drugs (VB1, VB2, VB6 and VB3) have been determined simultaneously by using alternating penalty trilinear decomposition (APTLD) method combined with HPLC-DAD. The results APTLD method predicted is closer to actual one. Moreover, APTLD method has the properties of fast convergence rate and being insensitive to overestimation of the component number compared with traditional PARAFAC. The results showed this experiment can be easily performed without complicated procedures and paying out time-consumingand quantified accurately. The experiment can simultaneously determine interesting compounds in complex system and achieve direct, fast, quanlitative and quantitative analysis.