Discrimination of legal and illegal Cannabis spp. according to European legislation using near infrared spectroscopy and chemometrics. Issue 9 (6th July 2020)

• Record Type: Journal Article
• Authors: Duchateau, Céline; Kauffmann, Jean‐Michel; Canfyn, Michaël; Stévigny, Caroline; De Braekeleer, Kris; Deconinck, Eric
• Is Part Of: Drug testing and analysis. Volume 12:Issue 9(2020)
• Page Start: 1309
• Page End: 1319
• Publication Date: 2020-07-06
• Subjects: Cannabis -- Chemometrics -- European legislation -- Handheld -- Near infrared spectroscopy; Drugs -- Analysis -- Periodicals; Drug testing -- Periodicals; Chemistry, Forensic -- Periodicals; 615.1901
• DOI: 10.1002/dta.2865
• Languages: English
• ISSNs: 1942-7603

Abstract

Abstract: Aerial parts containing cannabidiol can be purchased in a legal way but cannabis used as recreational drug is illegal in most European countries. Δ9‐tetrahydrocannabinol is one of the main cannabinoids responsible for the psychotropic effect. European Union countries and Switzerland authorize a concentration of THC of 0.2 % and 1.0 % w/w, respectively, for smoking products and industrial hemp. Public health inspectors and law enforcement officers need to check the legality of samples. Therefore there is a need for innovative approaches, allowing quality control of these products in an easy way and preferably on site. In many countries, cultivation of industrial hemp is permitted if the THC content does not exceed 0.2 % w/w. A portable equipment could be a useful measuring tool for farmers to check for the THC content at regular time. In this work, 189 samples were analysed with a benchtop and a handheld NIR device in order to create two classification methods according to European and Swiss laws. All samples were also analysed by GC‐FID to determine their THC concentration. Supervised analysis was applied in order to establish the best model. For the first classification, the accuracy was 91% for the test set with the benchtop data and 93 % for the test set with the handheld data. For the second classification, the accuracies were respectively 91 % and 95 %. The obtained models, hyphenating spectroscopic techniques and chemometrics, enable to discriminate legal and illegal cannabis samples according to European and Swiss laws. Abstract : (The key findings) Supervised analysis were applied in order to establish the best model. For the binary classification, the accuracy was 91% for the test set with the benchtop data and 93 % for the test set with the handheld data. For the tertiary classification, the accuracy was 91 % for the benchtop and 95 % for the handheld device.