Title of article :
Application of chemometric methods in searching for illicit Leuckart amphetamine sources Original Research Article
Author/Authors :
Waldemar Krawczyk، نويسنده , , Andrzej Parczewski، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2001
Pages :
8
From page :
107
To page :
114
Abstract :
A chromatogram of contaminants of illicit amphetamine is referred to as contamination profile. The profile depends on the method and conditions of the drug synthesis. Therefore, it allows to link a drug sample sized by the police to the source of illicit production as well as enables to find out the links between dealers and users. In our study, statistical and chemometric methods were applied. The areas of 15 more informative chromatographic peaks were selected and used as original variables describing the drug samples. The selected peaks formed a profile. As a rule, samples submitted to the police laboratory are classified using correlation coefficient between the corresponding profiles as a measure of similarity. A preliminary classification and the affiliation of a given profile to a class is verified by means of principal component analysis and cluster analysis. Different distances between profiles are used including Euclidean and Pearson ones. The data structure is visualized in two- or three-dimensional spaces of two or three most important principal components. For the purpose of this study, a total of nearly 1000 drug samples were tested. It has been concluded that amphetamine samples may be attributed to one source if the Euclidean distance between the corresponding profiles is less than 1, and the Pearson distance is less than 0.04. This approach proved very effective in identification of drug traffic routes and clandestine laboratories by the police.
Keywords :
Forensic research , Illicit amphetamine , Impurities profile , Correlation coefficient , cluster analysis , Principal component analysis
Journal title :
Analytica Chimica Acta
Serial Year :
2001
Journal title :
Analytica Chimica Acta
Record number :
1029803
Link To Document :
بازگشت