Recent advances in nano/microfabricated substrate platforms and artificial intelligence for practical surface-enhanced Raman scattering-based bioanalysis

Abstract

Surface-enhanced Raman scattering (SERS) can be a valuable analytical tool for bioanalysis. However, SERS is a complex optical phenomenon resulting from light-matter interaction. The SERS intensity and spectral feature strongly rely on the status of target molecules on the surface and structure of plasmonic nanomaterials. Among many prerequisites of SERS for practical applications, this review focuses on the fabrication method for sub-10 nm nanogap structures, the method for actively capturing strategies, and the method for data analysis. The recent advance of two different fabrication strategies with plasmonic nanostructures such as top-down and bottom-up have mainly been introduced and compared the advantages and disadvantages of each method, including basic principles. The capturing strategies for specific target analytes are discussed. There has been a significant challenge in data analysis of SERS spectra, because of the complexity of Raman spectrum and fluctuating Raman response. In this regard, recent developments in data science have resulted in the advancement of analytical technology. Beyond typical chemometrics, emerging technologies employing artificial intelligence have shown remarkable progress for practical SERS-based applications. In this review, we focus on the recent advances in the strategies for sensitive and reproducible SERS and the method for post-data processing.