A Practical Nanoplasmonic SERS Substrate for Differential Diagnosis of Lung Normal and Cancer Cells through Multivariate Statistical Analysis

Abstract Lung cancer ranks first for cancer‐related mortalities primarily due to late diagnosis. Though Surface‐Enhanced Raman spectroscopy (SERS) is a popular bioanalytical technique, its direct application to diagnosis is impeded by low data reproducibility. Colloidal nanoparticles suffer from SERS intensity fluctuations due to unavoidable aggregation, and Brownian and diffusion motions in biological samples. The processes for solid‐state SERS substrates are either sophisticated or difficult to reproduce. Herein, we revisit the well‐established thermal evaporation process for the easy and reproducible preparation of silver nanoparticles loaded SERS glass substrates. The static mode of thermal evaporation yielded closely packed and uniformly distributed silver nanoparticles. The properties of these nanoparticles are tuned for the best performance by controlling the thermal evaporation process. And SERS substrate exhibited a reasonably good enhancement factor of ~10⁵ with uniformity and reproducibility <6 % RSD over a large area. It was utilized for label‐free SERS fingerprinting of lung adenocarcinoma cells A549 and normal lung fibroblast cells, WI‐38. The obtained data shows a slight distinction of Raman fingerprints in terms of certain biomolecules like nucleic acids, proteins, and lipids. Further multivariate statistical tools have been utilized which ensures a clear divergence between the cancerous cells and normal cells.