Determination of cocaine contamination on banknotes using tandem mass spectrometry and pattern recognition

Abstract

Tandem mass spectrometry is used to monitor the contamination of banknotes by cocaine. By introducing a series of banknotes into an instrument a distribution of contamination can be obtained. The distribution of samples arising from defendants where the banknotes have been in close proximity to cocaine should differ from the distribution from the general background population. Peak picking and integration is used to produce a series of intensity readings for a batch of banknotes. By visually inspecting these distribution, and applying a variety of chemometric methods (principal components analysis, cluster analysis and class modelling via Mahalanobis distance) it is possible to discriminate effectively between the two classes of distribution (7157 background notes and 4826 case notes alleged to be from drug dealers). By calculating the Mahalonobis distance over 100 bootstrap iterations, background samples were correctly classified 96.48% of the time, while case samples were correctly classified 89.37% of the time.