Rice University spots soil contamination more rapidly

13 May 2025

Raman and machine learning are combined in a new strategy for identifying hazardous pollutants.

Down to earth: Rice University grounds

Described in PNAS, the technique could make it easier for farmers, communities and environmental agencies to test soil for hazardous compounds without needing to send samples to specialized laboratories.

Current testing usually involves labs comparing samples taken in the field with standard physical reference samples of the suspected contaminants. But for many environmental pollutants thought to pose a public health risk, there is no experimental data available that can be used to detect them. The process can also take several days.

Solving this problem would address a a critical gap in environmental monitoring and open the door to identifying a much broader range of hazardous compounds, including those that have changed over time, said the Rice team. This is especially important given that soil is a dynamic environment where chemicals are subject to transformations that can render them harder to detect.

The new method combines surface-enhanced Raman spectroscopy with a spectral reference library constructed using density functional theory (DFT), a computational modelling approach already employed to simulate the energetics of molecules.

As applied by the Rice team, DFT first isolates distinctive spectral features using a characteristic peak extraction (CaPE) algorithm. A second operation using a characteristic peak similarity (CaPSim) algorithm then identifies the analytes present, a determination made with high robustness to spectral shifts and amplitude variations.

This combination overcomes several limitations associated with traditional experimental libraries, according to the project, including spectral background interference, solvent effects, and commercially unavailable or challenging to synthesize compounds.

On-site field testing to monitor the environment

In trials, the project tested its method on soil from a restored watershed and natural area using both artificially contaminated samples and a control sample, looking for polycyclic aromatic hydrocarbons (PAHs) and their derivatives in soil. A common by-product of combustion, these molecules have been linked to cancer, developmental issues and other serious health problems.

Validation of the approach "showed strong similarity values (>0.6) between DFT-calculated and experimental surface-enhanced Raman spectra" for multiple PAHs including lesser-known and largely unstudied pollutant molecules, noted the project in its published paper. It also picked out even minute traces of PAHs, doing so more rapidly than conventional methods.

This method could offer a route to faster on-site field testing, potentially by integrating the machine learning algorithms and theoretical spectral library with portable Raman devices into a mobile system. Such a platform would make it easier for individuals or groups working with soil to test for hazardous compounds themselves, without needing laboratory resources.

"We are using PAHs in soil to illustrate this very important new strategy," commented Naomi Halas of Rice University.

"There are tens of thousands of PAH-derived chemicals, and this approach - calculating their spectra and using machine learning to connect the theoretically calculated spectra to those observed in a sample - allows us to identify chemicals that we may not, or do not, have any experimental data for."