Resolution And Modeling Of The Complex Analytical Signals

Resolution of complex overlapping signals is always one of the challenging tasks in analytical chemistry studies. With the development of analytical instruments and complexity of analytical systems, it is necessary to build effective methods to resolve complex overlapping analytical signals. There are two ways to carry out qualitative and quantitative analysis of overlapping signals including resolution methods and multivariate calibration. In this dissertation, a series of resolution of overlapping signals methods were proposed for qualitative and quantitative analysis of overlapping signals. The proposed methods were applied in resolution of real analytical signals, including gas chromatography - mass spectroscopic (GC - MS) and near infrared (NIR) spectra. Furthermore, the feasibility of high-throughput analysis was also validated. This dissertation includes six main aspects:(1) Based on the concept of 'window' in factor analysis, a window independent component analysis (WICA) was proposed for resolution of complex overlapping GC - MS signals. Two experimental GC - MS data including the pyrolysis products of isoprocarb and flue-cured tobacco were analyzed with the proposed WICA. Results show that both mass spectra and chromatographic profiles of the components in the overlapping GC - MS signal are accurately extracted.(2) A non-negative ICA method was proposed by means of a post rotation of the independent components (ICs) and applied to the extraction of the chemcial information of the components from the signals of complex samples. Raman spectra of pharmaceutical tablets and gas chromatography - mass spectrometry (GC - MS) data of cigarette smoke were qualitatively analyzed. The results show that the two types of analytical signals can be effectively and accurately extracted by using the proposed method.(3) Based on a modification of adaptive immune algorithm (AIA), a non-negative immune algorithm (NNIA) approach is proposed to carry out high-throughput analysis of complex overlapping GC - MS signals. With the virtue of the non-negative correction, the chromatographic profile of each component in an overlapping signal can be extracted independently and sequentially along the retention time. It may be, therefore, an applicable tool for high-throughput analysis of complex GC - MS signals, because, with the method, multicomponent samples may be analyzed in a very fast elution way without considering the separation effect of the components. In order to investigate the effect of the proposed method, a simulated GC - MS data of six components and an experimental GC - MS data of 40 pesticides were employed. Both the two data are satisfactorily resolved.(4) An outlier detection method is proposed for near-infrared spectral analysis. The underlying philosophy of the method is that, in random test (Monte Carlo) cross-validation, the probability of outliers presenting in good models with smaller prediction residual error sum of squares (PRESS) or in bad models with larger PRESS should be obviously different from normal samples. The method builds a large number of PLS models by using random test cross-validation at first, then the models are sorted by the PRESS, and at last the outliers are recognized according to the accumulative probability of each sample in the sorted models. For validation of the proposed method, four data sets, including three published data sets and a large data set of tobacco lamina, are investigated. The proposed method is proved to be high efficient and veracious compared with the conventional leave-one-out (LOO) cross validation method.(5) A method named as independent factor diagnostics (IFD) is proposed for investigation of the contribution of each LV to the predicted results on the basis of discussions about the determination of LV number in PLS modeling for NIR spectra of complex samples. It is shown that several high order LVs constitute main contributions to the predicted results, albeit the contribution of the low order LVs should not be neglected, in the PLS models for NIR spectra of three data sets of complex samples, including a public data set and two tobacco lamina ones. Therefore, in practical uses of PLS for analysis of complex samples, it may be better to use a slightly large LV number for NIR spectral analysis of complex samples.(6) A weighted multiscale regression for building a combined model in multivariate calibration of near infrared spectra is proposed. In the approach, the spectra are decomposed into different scale blocks (or frequency components) by wavelet transform (WT) at first, then partial least squares (PLS) models are built with the decomposed components, and at last a combined model is built by a weighted averaging. The weight of each model is determined by the prediction residual error sum of squares (PRESS) value obtained with Monte Carlo cross validation (MCCV). The underlying philosophy of the strategy is that useful information may be embedded in all the components obtained by WT, although the higher and lower frequency components mainly represent noise and background, respectively. To validate the effectiveness and universality of the proposed method, it was applied to two different sets of near-infrared (NIR) spectra of tobacco lamina. Compared with the results obtained with commonly used PLS method, the proposed method is proved to be a high-performance tool for multivariate calibration of complex NIR spectra.