Desarrollo de un modelo de predicción para la identificación y cuantificación de adulterantes en drogas ilícitas incautadas en la ciudad de Cúcuta - Norte de Santander.

Abstract

In this study, a predictive model was developed to identify and quantify the substances present in illicit drug samples seized by the Office of the Attorney General’s Office. The validation of the method was carried out through the analysis of cocaine samples seized in the city of Cúcuta and his metropolitan area between July 2017 and June 2018 using Gas Chromatography with selective mass detector (GC / MS) for the analysis of the samples. Additional analysis was made using Infrared Spectroscopy technique with Fourier Transform Analysis and the infrared spectra obtained were analyzed using chemometrics in the next steps: Regression of Partial Least Squares (PLSR) in order to predict the composition and percentage of substances present in the samples, Analysis of Principal Components (PCA) to look for similarities between the samples and their possible relation to the location of seizure, and Linear Discriminant Analysis to differentiate between cocaine hydrochloride and crack cocaine. The adulterants studied were the most commonly used in the samples were: levamisole, caffeine, lidocaine, phenacetin, lactose, and sodium bicarbonate. The method developed in this study is fast, inexpensive, non-destructive, and environmentally friendly; it allows for a reduction in cost and analysis time, in addition to reducing the rate of false positives or false negatives in preliminary tests of illicit drugs as are currently given by using colorimetric methods. On the other hand, knowing the adulterants and their percentages in the samples seized will serve as tool for generating criminal policy in the Attorney General's Office, in addition to characterization of places of origin or adulteration and generation of public health policies by the Municipal Health Secretary.