| Solid-phase microextraction (SPME) was introduced in 1990. SPME integrated with GC or GC/MS had been applied successfully in analysis of chemical compounds in environment, food and biological material. The advantage of SPME is revealded in the rapid analysis of food. There are six chapters in this paper. Our research work has been emphasized on the application of SPME in quantitative and qualitative rapid analysis in food.Chapter One reviewed and summarized application of SPME in analysis of food, and it was pointed out that SPME is especially adapted to the analysis of volatile and semi-volatile compounds. Some techlogies were reviewed including fast gas chromatography, experiment design and pattern recognition.Chapter Two studied the chromatographic behavior of organophoxphorus pesticides (OPs) under different chromatographic conditions. A microbore capillary column (i.d.0.10 mm) was used for fast analyzing of GC, compared with a narrow bore capillary column (i.d.0.25 mm) for conventional GC. The temperature program of fast chromatography can be translated from the condition in convential chromatography by GC method translation software. The analytical time is shortened from 38.0 to 8.75 min. While column head pressure was increased to 550 kPa. Thus, the analytical time was shortened further to 5.17 min. The response of peak is enhanced while the noise is not increased accordingly. It is interesting to see that height of peak is inversely proportional to width of half height of peak, although they are not exactly matched mathematically. The chromatographic condition which column head pressure is 385.1 kPa is chosen in this paper because it can meet the demand of qualification. Extraction efficiencies of OPs with three different fibers were compared, and PDMS fiber was chosen. The factors influencing extraction efficiency were optimized, the final chosen conditions were:extraction time was 80 min; extraction temperature was 70℃; the agitation speed was 750 rpm.30% of sodium chloride and 5% of acetonitrile were selected; the desorption temperature was 250℃. In order to eliminate the matrix effect, calibration curves were constructed by spiking appropriate amounts of OPs in each homogenized blank matrix of fruit samples (apples, pears, potatos and bananas). The results of method validation were satisfied. The real fruit samples were analyzed with the proposed method, Residues of OPs were found in most samples. OPs found in samples were chlorpyrifos, ethion, bromophos methyl, fenchlorphos, carbophenothion, cyanophenphos and parathion in which contents were even lower than LOD regulated in national standard of the People's Republic of China. It means that residues of OPs detected with this method may not be detected with other conventional technologies. Apples and pears were cut into two parts of husked or not. These samples were analyzed with the proposed method. Tomato samples were cleaned with flowing water. Some were detected directly, whereas the others were analyzed after 60 min of rinse. The analytical results showed that residues of OPs generally exist on the surface of fruits. It is understood that the risk of OPs intake can be reduced if fruit is husked or immersed in water for some time before eating.Chapter three studied the method which was developed for the determination of clenbuterol in meat product using stable-isotope-dilution gas chromatography coupled with Selected Ion Monitoring Mass Spectrometry. SPME was used to analyze clenbuterol in water and urine which matrix was simple. There is no report that clenbuterol in meat was detected with SPME because of complex matric. In our study, the protein was deposited by adding precipitator in homogenized meat sample. Then sodium hydroxide was added into the supernatant until the solution turn turbid and then turns clear. SPME and on-fiber derivatization could be optimized respectively. The condition of on-fiber derivatization was optimized first, and the optimized derivatization temperature was 80℃, derivatization time was 20 min, agitation speed was 500 rpm. The condition of SPME was optimized by three factors central composite design including 20 experiments. The three factors were extraction time, extraction temperature and agitation speed. Once the design is performed it is then possible to calculate the values of the response and coded factors using regression. Analysis of Variance was performed after the quadratic model was set up. The model was estimated with F-test, and the significance each term of model was assessed by t-test. A new quadratic model, which removed some terms of major factors and interactions according to the significane, was calculated. The optimum coded values for each of the three factors was calculated by genetic algorithm and hard search algorithm. The best extraction conditions were:extraction time was 100 min, extraction temperature was 90℃, and agitation speed was 504 rpm. After solid-phase microextraction and on-fiber derivatization, the content of clenbuterol was measured with the aid of the stable-isotope-dilution under optimal conditions. The limit of detection for the clenbuterol is 0.48μg/kg, the recovery is in the range of 96%～104%, and relative standard deviation is 7.1%. The real samples were analyzed with this method and the Chinese standard method. The results were satisfied through the compare of two methods.Chapter four studied the volatile compounds in beer and the discrimination of different band of beers. The samples, which were produced in six beer manufacturers, were collected. Three beer manufacturers were from China, and others were from Germany. The condition of fast gas chromatography was obtained from that of convential gas chromatography in literature by GC Method Translation Freeware. Total time analyzing a sample is only 10 min.43 compounds, which was searched with NIST library and similarity was higher than 80%, were found in Nanchang beer. 8 co-exist peaks were selected. PCA was carried out after the data were autoscaled. Samples were grouped according to six manufactures on the biplot. Loading of acetic acid and decanoic acid were almost overlapped. The good correlation revealed that the two compounds may be produced in same production process. Several architectures of the art neural network (ANN) were investigated to predict the origin of beers. ANNs used in this paper were back-propagation (BP), radial basis (RB) and probabilistic (PNN) neural networks. Leave-one-out cross-validation approach was used to evaluate the discriminating power of four architectures of the network. The prediction accuracy of BP-ANN adopting gradient descent algorithm or resilient backpropagation was less than 70%. The predication accuracy of RB-ANN was 84.1%. PNN-ANN was the best; its predication accuracy was 98.4%.Chapter five studied the volatile compounds in Nanfeng oranges and the discrimination of origin. 74 samples from representive geographical origin were collected. These samples were analyzed with data obtained from SPME-GC/MS. The SPME condition was optimized, and 57 volatile compounds were determined. These compounds include 13 alkanes,20 olefins,2 alkyl benzenes,14 alcohols,3 aldehydes,3 esters,1 morpholine and 1 hydroxybenzene. Areas of 44 co-existed chromatographic peaks were autoscaled for discriminating the origin of Nanfeng orange. These variables were selected by genetic algorithm because most of them were irrelevant to origin.7 selected variables were analyzed by PCA. Samples from 5 different origins were separated on PCA plot, and it was found that the distribution of samples on the PCA plot was almost as same as their geographical position in Nanfeng. It was showen that the volatile compounds were relevant with geographical origin.
|
PDF Full Text Request