StatusThe thesis was presented on the 27 February, 2008
Approved by NCAA on the 17 April, 2008
Abstract– 0.43 Mb / in romanian
3.03 Mb /
The work is dedicated to the development of optimisation's algorithms of all stages of biopharmaceutical analysis, which uses HPLC method, including the choice of the method of sample preparation, optimisation of chromatographic separation's selectivity and a choice of the optimal algorithms for information processing.
An original method for biomaterial sample preparation, based on “homogeneous” extraction, is offered for the first time. The comparative analysis of some non-selective methods for sample preparation has been carried out, being tested and optimised previously. The ranges of their applicability has been determined in dependence of analyte's polarity and other properties, then a general scheme of choosing the optimal method for sample preparation has been offered. The opportunity of biomaterial volume's minimisation has been studied, including analysis of blood spots on a paper.
Some recommendations on the choice of filling material for columns and precolumns and chromatographic equipment's module’s rational placement have been given. The possibility of compensation of the precolumn volume contribution in the blurring of chromatographic zones has been shown.
Recommendations on the choosing of the optimal parameters for the analog chromatographic signals digitisation and algorithms for chromatograms processing have been offered. Usage of the inverted second derivative of the chromatographic signal at chromatograms handling, complicated by considerable base-line drift has been offered for the first time.
Two algebraic functions most adequate for real peaks have been selected for mathematical modelling of chromatograms: the exponentially modified Gauss function (EMG) and the asymmetric double sigmoidal function. An algorithm for marking the chromatograms that have overlapped peaks, and evaluation of parameters of above termed algebraic functions has been designed. For the fitting of the model functions, both algebraic, and apparatus, the "Two step" algorithm has been offered to be utilised.
It was discovered, that in the most cases the linear relation between the logarithms of parameters of algebraic functions, that describes chromatographic peaks, and the logarithm of substitute's concentration in the mobile phase are respected, that has allowed to develop a simple and effective model method for the chromatographic system’s selectivity optimisation. The product of partial overlapping coefficients of all analytical peaks has been offered as an objective test for the chromatogram’s quality evaluation