A new dual-spectroscopy approach promises real-time, molecular-level detection of hazardous chemicals in complex environments, according to researchers from the Hefei Institutes of Physical Science, Chinese Academy of Sciences.
The method – Surface Plasmon-Enhanced Dual Spectroscopy (SPEDS) – combines two complementary optical techniques: surface-enhanced Raman spectroscopy (SERS) and plasmon-mediated differential UV–Vis spectroscopy (P-DUS). Together, they enable rapid, high-sensitivity monitoring of target analytes, while maintaining the specificity required for accurate identification.
"SPEDS not only improves detection sensitivity but also significantly expands the scope of detectable substances," said Associate Professor Bao Haoming, who led the study, in a press release. “With the integration of machine learning algorithms, we achieved over 98 percent accuracy in both chemical quantification and identification – surpassing conventional single-mode detection technologies.”
The study addresses longstanding challenges in chemical sensing under real-world conditions, where interfering matrices and fluctuating analyte concentrations have limited the performance of traditional techniques. By synchronizing SERS and P-DUS data streams and applying support vector machine classifiers and multivariate regression models, the SPEDS platform demonstrated robust analyte recognition and quantification across a range of plasmonic substrates.
In field tests, the researchers used CuS-coated gold nanoarrays to detect mercury ions (Hg²⁺) in real water samples, successfully identifying and quantifying target species in real time. The technique showed strong reproducibility and adaptability across multiple plasmonic architectures, indicating broad utility in environmental and industrial applications.
According to the authors, SPEDS offers a promising pathway toward intelligent chemical sensing systems for use in environmental surveillance, public health, and industrial hazard mitigation. Future work will focus on optimizing the algorithmic framework and extending the system’s capability to monitor multi-analyte mixtures.
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
