Field-enhancement calculations in plasmonic tagged nanoplastics improved by machine learning

Great scientific interest has recently been directed towards the development of sensitive methods for the detection, monitoring, and characterization of nanoplastics dispersed in the environment. A promising approach to increase sensitivity is based on the use of SERS (surface-enhanced Raman scattering), producing strong, easily recognized signals that allow sensitive and quantitative detection of nanoplastics thanks to the induced electromagnetic field enhancement. Computational approaches capable of simulating such enhanced optical response appear crucial for the development of SERS detection methods. Here, we model a system of plasmonic tagged nanoplastic particles using a rigorous electromagnetic description combined with neural networks to speed up the calculations and explore the field-enhancement response in a wider range of plastics and metal nanoparticle sizes and compositions. Our results show that this combined approach may serve as a rapid tool to explore the parameter space and optimize SERS detection.