| In this thesis, chromatography technology, chemometrics and the identification of unknowns have been introduced. As a new technique in separation science, supercritical fluid chromatography has gained increasing attention in recent years. In this work, supercritical fluid chromatography-mass spectrometry combined with chemometrics was used in the following three researches.1. Direct detection of long-chain free fatty acids (FFAs) in edible oils using supercritical fluid chromatography-triple quadrupole mass spectrometry (SFC-TQMS). In this study, SFC followed by an electrospray ionization triple quadrupole mass spectrometry (ESI-MS), was shown to provide a novel method for the separation and detection of FFAs in edible oils without any pretreatment. Eight FFAs were separated on a HSS C18 SB column with gradient elution within 3 min. Effects of different columns, modifiers and column temperature changes were evaluated. Compared with ultra-performance liquid chromatography (UPLC), SFC showed a greater separation. All FFAs exhibited much more symmetrical peaks with relatively narrow peak shapes, especially the baseline separated α/γ-C18:3. The method validation results indicated the feasibility of this method for the high-throughput determination of individual FFAs with satisfactory correlation coefficients (R2 ≥0.994) and good reproducibility of RSD<13.5%(intraday) and ≤15.0% (interday). Combined with principal component analysis (PCA), different types of edible oil were successfully distinguished into several categories, showing a potential application for the determination of oil quality or authenticity.2. Characterization of medium-and long-chain free fatty acids contents of different beer types by the established SFC-TQMS method. In this part, medium- and long-chain FFAs which are related to beer taste were determined. N-hexane was selected as extracting agent, four factors that could potentially affect the extraction efficiency were evaluated, including standing time, vortex time, NaCl addition and extraction times. Response surface methodology (RSM) was used for the optimization of liquid-liquid extraction procedure, and results were compared with that of the single-factor experiments. RSM is more efficient as it allows the investigation of the effects of multiple variables at one time, which reduces the whole development time and the accidental error. Besides, interactions among the variables are considered. After the method validation,30 commercial beers were tested. PCA was used to discriminate them according to their FFAs contents. Beer samples from Europe were clearly separated from the Chinese samples. And beers made from different raw materials could be easily differentiated.3. In the present study, non-targeted analysis method of transformer oil based on supercritical fluid chromatography coupled with time-of-flight mass spectrometry (SFC-QTofMS) was established. Firstly, the representative waste transformer oil and 25# new oil samples were analyzed by MS, IR and GC-MS, then methanol was chosen as extracting agent, the profiling data of transformer oil was obtained by using SFC-QTofMS with full scan in both positive and negative mode. The separation efficiency of SFC was compared with that of UPLC, the result showed that SFC exhibited a better separation. Combined with PCA, transformer oil aging level evaluation system was built through waste transformer oil samples and verified by transformer oil in use and samples from fault simulation experiment. Furthermore, characteristic ions that played the most important roles for differentiating the transformer oil samples were screened out by loading plots, and identified by UNIFI software and SciFinder database. |
PDF Full Text Request