Identification Of Chemical Contaminants In Animal Production By Fourier Transform Infrared Spectroscopy

Screening chemical contaminants in animal production is the premise of controlling them from the source.The previous detection techniques focus on the quantitative analysis.However,new contaminants has been a challenge for them.Infrared spectrum is a powerful tool to analysis the material structure for its rich structural information.The application combining it with chemometrics in analyzing these comtaminants in complex matrixes made it possible to screening illegal chemicals in the animal production.The research set up the infrared spectrum data designed for chemical contaminants in animal production to improving target recognition and explored the effect of factors related to analyzing the target based on the data to identify the chemical contaminants.The main contents and results are as follows:1.The establishment of the IR spectrum database containing chemical contaminants probably used in animal production.In the Know It All software platform,333 substances were added on the basis of several databases purchased.In addition,the infrared spectra of 21 kinds of feed premix supporters and diluting agents were collected.For building functional database,take aniline agonists for example,the method to look for a reference material which is the one of these,or whose structure is the public is not available.2.Analysis of complex samples.The research discussed the significance of the matrix database,and the effect of the type of matrixes,target concentrations(9.09%,16.67%,33.33%,50.00%),the amount of the target in the tablet(0.10%,0.50%,1.00%,2.00%)and searching methods on the matching results of the target.It was found that the interference of organic matrixes was lower than that of inorganic ones,the results would be remarkably reduced when the concentration of the target was lower than 16.67%,the content of the target in tablets over 0.50% was suitable,the result was depressed with the order of First Derivative Euclidean Distance,Correlation,Second Derivative Euclidean Distance and Euclidean Distance.It was also found that increasing the content of the interfering substance blended with the target,to a certain extent,had no significant effect on the matching results,and the content corresponding to different substance was different when the matching results remarkably.3.To classify the complex samples with different matrixes,the principal component analysis and cluster analysis were used.Different bands containing the characteristic peaks respectively belong to organic matrixes and inorganic ones and data processing methods(square root transformation,logarithm transformation and logarithmic centralization transformation)were selected and implied for the original data.It turned out thatthe clustering results according to the type of the matrix will be improved when the information related to matrixes in the bands or the loading diagrams of principal components was more abundant,or there were marked difference in the absorption intensity between two kinds of matrixes.