Detection of Fentanyl and Fentanyl Analogs in Multi-Component Drug Mixtures by Combining Surface-Enhanced Raman Spectroscopy and Chemometric Approaches

The increasing prevalence of fentanyl analogs in illicit drug mixtures poses significant challenges for forensic laboratories and public safety. Trace levels of highly potent compounds, often combined with other drugs and common adulterants such as caffeine, acetaminophen, and metamizole, create complex matrices that are difficult to analyze. Rapid, sensitive, and reliable detection methods are essential to ensure accurate identification while minimizing laboratory hazards.
This study integrates surface-enhanced Raman spectroscopy (SERS) with chemometrics to detect fentanyl analogs in binary, tertiary, and more complex mixtures. Samples including fentanyl, 4-ANPP, acetyl fentanyl, para-fluoro fentanyl, metamizole, acetaminophen, and caffeine were mixed with optimized Au/Ag nanostar colloids and analyzed using portable Raman spectrometers. A spectral library was developed for individual components and their mixtures. Chemometric and machine learning techniques, including principal component analysis and partial least squares discriminant analysis, were examined to deconvolute overlapping spectra, isolate fentanyl signals, and predict relative concentrations of the drugs. DFT based models were developed to assist in the detection of both known and previously uncharacterized fentanyl analogs in multi-component samples.
The goal was to develop rapid, and sensitive, presumptive detection of fentanyl analogs in complex drug mixtures. By combining nanotechnology-enhanced spectroscopy with artificial intelligence, the study contributes to strategies for combating the opioid crisis.