A novel heterojunction aerogel substrate for SERS machine learning-based explosive wastewater detection

Zihan Wang, Mengdan Ma, Man Yuan, Sisi Shang, Wei Liu, Sheng Cui, Xuan He,
A novel heterojunction aerogel substrate for SERS machine learning-based explosive wastewater detection,
Sensors and Actuators B: Chemical,
Volume 444, Part 1,
2025,
138337,
ISSN 0925-4005,
https://doi.org/10.1016/j.snb.2025.138337.
(https://www.sciencedirect.com/science/article/pii/S092540052501113X)
Abstract: Explosive wastewater contains organic pollutants with strong acidity and biological toxicity, posing severe environmental risks if not properly managed. However, the coexistence of multiple explosives and extreme acidity complicates pollutant signal identification. In this study, we integrated machine learning with an ultra-sensitive and stable SERS substrate for accurate detection of organic pollutant wastewater. The large specific surface area and porous structure of heterostructure aerogel enhanced the adsorption of molecules. And specific heterostructure promoted efficient charge separation, significantly improving charge transfer pathways and efficiency between substrate and molecules. This structural design amplified Raman signals through enhanced SERS effects and demonstrated exceptional chemical stability, maintaining detection performance even under prolonged exposure to highly acidic or alkaline conditions. This stable and efficient charge transfer made it highly sensitive to various organic pollutants wastewaters. Furthermore, integrating machine learning models with Raman spectroscopy not only improved efficiency of manual identification but also achieved an accuracy of up to 96 % in distinguishing wastewater mixtures. Therefore, the high sensitivity, stability and accuracy of the heterojunction aerogel SERS substrate make it have great potential in the detection of explosive wastewater in extreme environments.
Keywords: SERS; Heterojunction aerogel; Explosive wastewater; Machine learning; Charge transfer