Attenuated Total Reflection Fourier Transform Infrared Spectroscopy and Chemometrics for the Forensic Investigation of Intentional Rodenticide Poisoning in Ground and Liquid Coffee
Year: 2026
Type: article
Abstract: Advances in forensic science and analytical technologies have significantly improved the collection, preservation, and interpretation of material evidence. Food and beverage matrices can serve as valuable forensic evidence when biological samples are absent or degraded. Among these matrices, coffee is particularly relevant to investigations of intentional poisoning because its sensory characteristics can mask toxic substances. Pesticides marketed as rodenticides, including carbamate and organophosphate compounds, together with the anticoagulant rodenticide brodifacoum, are of forensic interest because of their toxicity, widespread use, and occurrence in poisoning cases. Traditional chromatographic techniques, while robust and sensitive, are costly and labor-intensive, and many forensic laboratories face limited access to such equipment. Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIR) combined with chemometric tools such as Principal Component Analysis (PCA) and Soft Independent Modeling of Class Analogy (SIMCA) offers a rapid, nondestructive, and cost-effective alternative for screening rodenticide-related contamination in complex matrices. This study evaluated the potential of ATR-FTIR coupled with chemometric analysis for detecting rodenticide-related contamination in coffee and its applicability to forensic investigations of intentional poisoning. Commercial brodifacoum-based rodenticides and rodenticide formulations of unknown composition were investigated as representative toxic agents. A preliminary evaluation of ground coffee contaminated with one unknown formulation at 5.0 and 10.0% (w/w) was conducted to define the experimental conditions. The 5.0% level was selected for subsequent analyses because it was the lower of the two evaluated levels and already provided multivariate differentiation from uncontaminated samples. A total of 85 ground-coffee observations and 84 liquid-coffee observations were initially evaluated. Following Hotelling’s T2/Q-SPE assessment and inspection of the original spectral profiles, the refined data sets were used for exploratory and supervised multivariate analyses. Three SIMCA-based decision rules, data-driven SIMCA (DD-SIMCA), alternative SIMCA (Alt-SIMCA), and a Hotelling’s T2-only rule, were evaluated using the corrected spectral range for ground coffee or the full spectral range for liquid coffee, as well as the fingerprint region from 1800 to 900 cm–1, with external-validation sets comprising 26 to 28 samples depending on the matrix and spectral region. For ground coffee, both DD-SIMCA and Alt-SIMCA achieved 100% external-validation accuracy and 100% sensitivity for all classes in both spectral regions, with specificities of at least 75.00% and 80.00%, respectively. For liquid coffee, the best performance was obtained with Alt-SIMCA using the full spectral range, which achieved 100% external-validation accuracy, 100% sensitivity for all classes, and class specificities ranging from 89.47 to 100%. Within the tested formulations, coffee brands, and contamination level, these findings indicate the potential of ATR-FTIR combined with chemometrics as a rapid screening tool for the forensic analysis of coffee samples suspected of contamination with toxic agents.
Source: ACS Omega
Language: English
Subfield: Analytical Chemistry
Field: Chemistry
Domain: Physical Sciences
Keywords: rodenticide poisoning, coffee, brodifacoum, anticoagulant rodenticides, carbamate pesticides, organophosphorus pesticides, ATR-FTIR spectroscopy, SIMCA, principal component analysis, forensic toxicology, food contaminant detection, ground coffee, non-destructive screening, Hotelling's T2, external validation, decision rules
Open Access status: gold
Estimated APC paid: $1,935
Award: G22113365
OpenAlex ID: W7236445885
