Machine learning-integrated SERS platform based on AuSe nanowires for indigo carmine detection in real samples

Surface-enhanced Raman scattering (SERS) substrate composed of gold nanoparticles-decorated selenium nanowires (AuSe NWs) was developed for trace detection of the food dye, indigo carmine (IC). The Se NWs were synthesized via co-precipitation and functionalized with Au nanoparticles through photoreduction. The resulting substrate exhibited strong SERS activity due to synergistic electromagnetic and chemical enhancement within AuSe NWs. Detection of IC was achieved at concentrations as low as 10^-9 M, with an enhancement factor of 10⁸. The substrate demonstrated excellent uniformity, reproducibility, and stability for up to 30 days. Practical applicability was confirmed in real food samples such as chips and grape juice, reaching detection limits of 10^-8 M. Furthermore, machine learning algorithms were integrated with SERS spectra to accurately classify IC in various foods including cookies, chewing gum, cheese, and bread. This approach offers a reliable, sensitive, and rapid tool for real-time monitoring of synthetic dyes in food industries.