A compact Fourier-transform near-infrared spectrophotometer and chemometrics for characterizing a comprehensive set of seized ecstasy samples.
Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy June 5, 2024 Jennifer A Cavalcante, Jamille C Souza, Jarbas J R Rohwedder et al. 3 citations
A compact, low-cost near-infrared (NIR) spectrophotometer, combined with chemometric models, can classify ecstasy tablets and estimate their active ingredient content with over 96% accuracy. Using seized samples from the Brazilian Federal Police, the study applied SIMCA, LDA, PLS-DA, and PLS regression to spectral data. The models estimated total active compounds, MDMA, and MDA with root mean square errors of 4.4%, 4.2%, and 2.7% (m/m), respectively. The approach offers a practical protocol for in-field forensic screening, though limitations such as false positives and negatives are discussed, and a new method to improve classification robustness is proposed.