Multi-way Calibration Coupled With HPLC-DAD For Quantitative Analysis Of Drugs In Complex Systems

Chemometrics is a developing composite discipline, which is an important branch of the analytical chemistry and widely used to the qualitative and quantitative analysis for the analytes in complex system. Its unique characteristics, "mathematical separation" partially or completely replace "chemical separation", which is increasingly causing the concern of chemists. Now chromatographic techniques play an important role in simultaneous determination of multiple components in samples. In this paper, the research work focuses on multi-way calibration combined with chromatographic techniques for quantitative analysis of interest in complex chemical systems.1. Second-order calibration coupled with HPLC-DAD for quantitative analysis of drugs in complex systems (Chapter 2 to Chapter 5)Accurate, rapid and quantitative determination of ingredient is an important thing in life science and plays an important role in drug quality control. Quantitative analysis of components of interesting even in the presence of unknown interferences is the most concern of analysts. Two components are difficult to be separated by simple chromatographic condition as they have similar structure and characteristic, accordingly, methods of chemometrics are used to helpfully resolve this problem. Second-order calibration methods were presented to allow accurate and reliable quantitative analysis of metronidazole and tindazole, metoprotol and atenolol in plasma samples, which fully reflects the"second-order advantage".Isomeric compounds have the same molecular weight and many also have similar chemical properties, so it is unnecessary to surprise that they have similar spectral profiles and close retention times in chromatographic analysis. How to effectively separate and quantitatively analyze them in drugs is also the interesting question of analysts. Recently there are many papers to aim at these problems, but most methods often suffer from being time-consuming and needing tedious pretreatments in chromatography or too expensive cost. This study described that HPLC-DAD coupled with second-order calibration methods was employed to determine isomeric compounds such as psoralen and isopsoralen, alantolactone and isoalantolactone in Chinese tradional medicine. With the second-order calibration method, one can simplify the sample preparation procedure only by using a simple and convenient solvent extraction with methanol. Thought isomeric compounds have not only similar spectral profiles and close retention times in chromatographic analysis, but also their profiles overlapping with those of unknown interference from complex system. Nevertheless second-order calibration methods could provide accurate concentration prediction and good resolution of elution and spectral profiles for the analytes of interests under this circumstance. Also LC/MS and GC/MS were performed to validate the accuracy of results which were obtained by second-order calibration methods. The results showed that second-order calibration methods is a powerful chemometric tool to decompose heavily overlapped peaks into their pure chromatographic, spectral and concentration profiles even in complicated systems such as pharmaceutical preparations. The second-order advantage was fully exploited in this paper, which made concentration prediction feasible even in the presence of unknown interferences. In the present work, a simple sample disposal and short analysis time has been achieved. It has been proved that"mathematical separation"instead of partial"physical or chemical separation"was useful in chromatographic analysis. It provides a new method of quantitative analysis for ingredient in complex drug samples.2. Quantitative analysis of overlapping chromatographic peaks using trilinear second-order methods (Chapter 6)Base on the application in practice of trilinear second-order calibration methods, resolution of overlapping chromatograms using these decomposition algorithms are also detailly studied in this paper. Chromatographic and spectral peaks were normalized to calculate resolution or overlapping degree of components in overlapping peaks, which made the difference of components more obvious in overlapping peaks especially in embedded peaks. The results obtained by simulated and real data have shown that trilinear decomposition algorithms have the superior performance to resolve overlapping peaks of elution and spectral profiles and quantify the analytes as there is discrepancy among overlapping peaks or embedded peaks and overlap degree of chromatographic and spectra peaks are not simultaneous severely approximate 100%. With the same noise level and collinearity, the greater of separation of chromatographic and spectral peaks the better results are attained from the algorithms.3. Third-order calibration coupled with LC×LC-DAD for quantitative analysis of drugs in complex systems (Chapter 7)With the development of analytical instruments and methods, it becomes increasing easier to obtain a large number of high-dimensional data such as four-dimensional data which involve more information than second-dimensional data. It requires more effective higher-order calibration methods of chemometrics to derive useful information from it, accordingly, third-order calibration methods that came into being developed. In this paper, LC×LC-DAD coupled with AQLD (alternating quadrilinear decomposition), APQLD (alternating penalty quadrilinear decomposition) and four-way PARAFAC (parallel factor analysis) was applied to quantitatively determine ferulic acid in chuanxiong samples, acceptable results can be supplied. It come to a conclusion, three third-order calibration methods are adapted to analyze four-way data attained from two-dimensional HPLC-DAD. Comparing with AQLD, APQLD and four-way PARAFAC algorithms, obviously, AQLD and APQLD are generally insensitive to the overestimates of the component number chosen and have fast convergence rate, which is difficult for four-way PARAFAC to surpass. The accuracy of AQLD and APQLD is better than that of four-way PARAFAC illustrated by the EJCR test. The performances of both AQLD and APQLD were slightly better than that of four-way PARAFAC for decomposing quadrilinear data in this paper. Since an additional temporal dimension was introduced and more information of samples was considered, theoretically, third-order calibration could give a higher sensitivity and selectivity than second-order calibration, which can be considered as"third-order advantage". Admittedly, the advantage of four-dimensional data analysis is still at an early stage of exploration, and complex or higher order advantage is still to be found and studied for us.