Application Of Second-order Calibration Methods For The Quantitative Analysis In Complex Systems Mainly Focused On Vegetable Oil And Body Fluid

As society develops, a wide range of analytical instrumentation is available thatenables high dimensionality data to be obtained. In this situation, second-ordercalibration methods has gained widespread acceptance by the analytical community,because it exploit the intrinsic data structure of ‘hyphenated’ analytical methods, suchas excitation-emission fluorescence matrix spectroscopy and chromatographicseparation coupled to diode array detection, to enable one to quantify the analyte, evenin the presence of interferences not modelled in the calibration stage. This property ofsecond-order calibration methods has been termed the ‘second-order advantage’ byBooksh and Kowalski. This advantage makes it possible to directly quantify theinteresting analytes from complex system and provides the main motivation for thepresent parper.In this paper, second-order calibration methods are employed for the quantitativeanalysis in complex systems mainly focused on vegetable oil and body fluid. Thesecomplex systems are made of complex mixtures of compounds of a wide range ofchemical classes. Characterization of compounds in vegetable oils is of greatimportance in establishing the oil quality, and it is important to determine the levels ofmedicines or markers in the body fluid for understanding how healthy people are. Themost commonly used analytical techniques for analysis of these compounds are liqu idchromatography. These techniques are however expensive, time-consuming andrequire a pre-separation step before the analysis. Therefore, the aim of this paper is tocouple the second-order calibration methods with high-performance liquidchromatography-diode array detection (HPLC-DAD), high-performance liquidchromatography-fluorescence detection (HPLC-FLD) or excitation-emissionfluorescence matrix (EEM) for the development and optimization of a simple, rapid,efficient, fast and no-extraction methodology for quantitative analysis in complexsystems, such as vegetable oil and body fluid. The contents of our studies can besummarized as:Chapter2:A new chromatographic methodology is presented for fast analysis often synthetic phenolic antioxidants in five kinds of oil samples: propyl gallate (PG),2,4,5-trihydroxybutyrophenone (THBP), tert-butylhydroquinone (TBHQ),nordihydroguaiaretic acid (NDGA), ethoxyquin (EQ),3-tert-butyl-4-hydroxyanisole (BHA), octyl gallate (OG),2,6-di-tert-butyl-4-hydroxymethyphenol (Ionox-100),dodecyl gallate (DG),3,5-di-tert-butyl-4-hydroxytoluene (BHT). The alternatingpenalty trilinear decomposition (APTLD) algorithm has shown to be an excellent toolfor modelling the second-order data, where overlapping peak and baseline drift wereexisted, making the fast determination and resolution of the phenolic antioxidants inoils possible. The extraction is unnecessary and the ten antioxidants can be elutedwithin6mins. For the validation of the method, linearity, root-mean-square error ofprediction (RMSEP) and limit of detection (LOD) have been performed. The results ofoverall predictions are desirable and the LOD of our method is enough to detect SPAsin edible vegetable oils in most cases.Chapter3: A liquid chromatographic method has been developed, in combinationwith the second-order calibration method based on the alternating trilineardecomposition (ATLD) algorithm with fluorescence detection (FLD), for thesimultaneous determination of seven phenolic antioxidants in six kinds of oil samples.The extraction is unnecessary and a column washing is also not required. After asimple dilution step, oil samples can be directly injected into the detecting system andthe elution is accomplished in a short time (6min). Overlapped peaks have beensuccessfully resolved and the quantitative results of the analytes of interest have beenaccurately estimated. The limits of this method are much lower than the maximumresidue level established for the antioxidants. It is about one-tenth of the valuepreviously proposed by Chapter2.Chapter4: A novel algorithm named as self-weighted total least-squaresalternating trilinear decomposition (SWTT) has been proposed for quantitativeanalysis of EEM fluorescence data. Based on different input parameters includingcollinearity, interference intensity and noise level in three simulated data arrays, theperformances of the self-weighted alternating trilinear decomposition (SWATLD), theparallel factor analysis-alternating least squares (PARAFAC-ALS) and the SWTTalgorithms were evaluated in terms of predicting ability, error times and consistency ofresolved and real profiles. When the collinearity, interference intensity and noise levelof data are low, the results obtained by the three algorithms are nearly the same.However, the SWATLD can not get satisfactory results and the PARAFAC-ALS easilytrap into swamps many times when the leve is high. By utilizing the alternating totalleast-squares principle and weight factor, the SWTT algorithm can improve thetwo-factor degeneracy problem, which is difficult to handle for the PARAFAC-ALSalgorithm, and obtain more stable and more precise results. In addition, the SWTT and PARAFAC-ALS method was successfully applied to the simultaneous analysis forestriol and estrone in liquid cosmetic samples.Chapter5: Several studies have shown that an increase in amounts ofisoxanthopterin can be associated with the presence of cancer, thus its level isimportance parameter in clinical diagnosis. A simple, rapid, sensitive and selectivemethod for determining isoxanthopterin content in human urine samples has beendeveloped using secon-order calibration mehtod coupled with EEM. With theapplication of a second–order advantage, the proposed strategy could be utilized for adirect concentration determination of isoxanthopterin without any pretreatment step,even the spectra of isoxanthopterin is highly overlapping with other pterin compounds(neopterin, biopterin, pterin and pterin-6-carboxylic acid). The values of theisoxanthopterin from healthy volunteers, which is botained by our method, were ingood agreement with those obtained by HPLC-FLD method.Chapter6: A simple and rapid method is proposed for quantitative analysis ofadriamycin in cell culture media without any pretreatment procedure. It combinedsecond-order calibration methods based on both an alternating trilinear decomposition(ATLD) and an alternating normalization-weighter error (ANWE) algorithmscombined with excitation-emission fluorescent matrix spectra. The satisfactory resultsshowed that the second-order calibration method can be applied to directly quantifyadriamycin in cell culture media. Moreover, in the analysis of cell culture mediaincluding four interferences which are the fluorescent materials in cell, the adriamycinalso can be accurately determinated. It is found that second-order calibration method isappropriate for quantitative analysis of drug content in cell culture media even in thepresence of natural fluorescent interferences of cell.Chapter7: This work presents a novel approach for simultaneous determination ofbutylated hydroxyanisole (BHA) and propyl gallate (PG) in a very interferingenvironment by combining the sensitivity of molecular fluorescence and the selectivi tyof second-order calibration method based on the self-weighted alternating normalizedresidue fitting (SWANRF) algorithm. This method uses very small amounts of organicreagents and the quantification of PG and BHA were accomplished usinggreen-chemistry principles. In addition, the proposed method avoids preconcentrationand elution procedures, so it considerably decreases the analytical time and theexperimental expenses. Because the instrument involved in the measurement isnonsophisticated, the experiments could be carried out in routine laboratories.Compared to the HPLC method (the AOAC official method983.15,1994), the EEM-SWANRF is a more sensitive, simpler, faster, cleaner and more powerful method.