Research On Application Of Multivariate Calibration Method In Analysis Of Multi-component Complex System

At present, the quantitative analysis of multi-component complex system is an importantproblem in the analytical chemistry. This thesis mainly studies combined with spectral analysismethod of multivariate calibration methods to establish the method for rapid quantitativeanalysis of multi-component complex system. In the analysis of complex compounds,traditional analysis methods with pre separation for mixture and have slowly analysis speed. Itis necessary to develop some new quickly, simply and high accuracy analysis methods to meetthe need of the practice an application. The spectral analysis method is a high throughputanalysis method. It can realize the rapid analysis of multi-component complex system. Formulti-component complex system, particularly those with similar components, an intenseoverlapping between the spectra and an inereased complexity of samples practically involvedpersist in a need of useful approaches to build robust and stable linear calibration models. Theproper use of multivariate calibration can solve the problem. In the first chapter, severalmultivariate calibration methods and spectral analysis methods are reviewed. The experimentincluded three parts.（1）A method which combines fluorescence spectroscopy, guest–hostchemistry and principal component regression (PCR) was proposed for determining theenantiomeric composition of tryptophan (Trp) and propranolol. Fluorescence spectroscopywas used to measure the chiral interaction between the enantiomers and the added bovineserum albumin (BSA), which plays the role of chiral selector. PCR was used to model thequantitative relationship between the obtained fluorescence spectral data and enantiomericcomposition. The result demonstrates that the method is practicable in quantitativelydetermining enantiomeric composition of Trp and propranolol.（2）A method which combinesUV spectroscopy, guest–host chemistry and principal component regression was proposed fordetermining the enantiomeric composition of DOPA samples. The calibration models weredeveloped from UV spectral data of a series of samples containing DOPA with differentknown enantiomeric compositions by using PCR. The obtained model was subsequentlyvalidated by determining the enantiomeric composition of a set of independently preparedsamples. This method shows high sensitivity for determining the enantiomeric composition of DOPA.（3）Raman spectroscopy and near-infrared (NIR) transmittance spectroscopy approachwas proposed to determine the active substance in Escitalopram tablets. The regression modelwas established by using stepwise regression method. The obtained model was validated bydetermining the active substance of a set of independently prepared samples. The resultdemonstrates that the method is effective and practicable to determine the active substance inEscitalopram tablets. Compared with the Raman spectroscopy method proposed in the fifthchapter, the NIR method proposed in this chapter shows higher accury for determining theactive substance in Escitalopram tablets.