Fingerprinting and Differentiation of Landfill Leachate and Domestic Sewage in Surface Water with Machine Learning Based Surface Enhanced Raman Spectroscopy

The major sources of chemical pollutants in drinking water springs are believed to be from landfill leachate and municipal wastewater. Whenever a water related pollution incident happens, the rapid identification of the pollution sources could greatly help in its management. Various techniques and methods have been proposed for this purpose; here, we propose a spectral fingerprinting and differentiation method for water samples using machine learning (ML) based on surface-enhanced Raman spectroscopy (SERS). To simulate the discharge of wastewater into waterbodies, surface water (lake water) was spiked with landfill leachate and domestic sewage (as pollution sources) at different COD gradients, respectively. Two SERS protocols were evaluated for extracting Raman features from the wastewater samples. The collected SERS spectra were subjected to different ML methods for best differentiation results. These results demonstrated for the first time that by tuning the surface adsorption on the SERS substrate, SERS combined with ML can extract spectral fingerprint features to identify and distinguish water pollution sources.