Identification and Forensic Classification of Amphetamine

Sammanfattning: Amphetamine is one of the most abused drugs worldwide. In the fight against the manufacture and illegal distribution of amphetamine, the contribution from forensic laboratories is the identification of the drug and the establishment of links between seizures. These links contribute to the disclosure of distribution chains and illicit laboratories.For the identification of amphetamine and other drugs, a dual-column capillary gas chromatographic method is described. The separation was enhanced by the use of two columns in the same gas chromatograph. The phases in the two columns were of different polarities and were connected to a flame ionisation detector and a nitrogen phosphorous detector, respectively. This configuration generated more information on the identity of compounds as compared to ordinary one-dimensional gas chromatography.Links between amphetamine seizures can be established on two different levels, one level aiming at seizures originating from the same manufacturing batch and the other aiming at seizures manufactured by the same recipe. Such links were established by the study of impurities that emanated from the manufacturing process. By gas chromatographic analysis, impurity profiles were achieved from the seizures and compared. The extraction conditions of the impurities, the stability of the extract and the gas chromatographic performance were studied. The extraction method used proved to be robust and repeatable. Most of the impurities studied were extracted quantitatively from the amphetamine.Computerised methods were used to facilitate the comparison of the profiles and to find links between them. Selected peak areas were used to represent the profiles and were shown to be good descriptors. A quotient method that was able to find profiles from seizures originating from the same batch was presented. It also proved to be less sensitive to accidental errors in peak areas than other computerised methods studied.Principal Component Analysis (PCA) and Soft Independent Modelling of Class Analogies (SIMCA) proved to be good tools for finding similarities on the 'recipe' level. Classes were found among profiles and new profiles could be put into the classes previously found. Outliers were easily identified and were either the first members of new classes or had erroneous peak areas.The batch level statements could be given with high accuracy and the results used in court.The class level statements had less accuracy and could be used for intelligence purposes to get an overview of the illicit market and direct police work. With the help of these methods new seizures could be compared to a large number of previously analysed seizures, for the establishment of new forensic links.