Every day, water utilities across the United States monitor our drinking water for harmful pollutants. Haoran Wei, a University of Wisconsin-Madison associate professor of civil and environmental engineering, wants to make that process more efficient.
With the support of a prestigious National Science Foundation CAREER Award, Wei is launching a research project to develop a sensor capable of detecting organic pollutants in water at trace levels—a few parts per billion, which is roughly equivalent to a few drops of water in an Olympic-sized swimming pool.
“Every water utility or public waterworks has to monitor for regulated organic contaminants before it can distribute drinking water to millions of people across the country. That is the law,” Wei says. “This takes time and can be expensive because the maximum allowable contaminant level for these pollutants is extremely low.”
To do this, most water utilities rely on sophisticated equipment that can perform tests, such as liquid chromatography and tandem mass spectrometry. The equipment is capable, Wei says, but it is expensive and requires specially trained operators. Costs can vary depending on the targeted contaminant, but Wei says some can run up to several hundred dollars per test.
Those costs can add up quickly. According to the U.S. Centers for Disease Control, under the Safe Drinking Water Act, testing schedules vary by contaminant and water system size, with systems that serve larger populations generally required to test more frequently.
Wei wants to create a test that’s less than $1 per use.
Through his CAREER Award, he plans to develop a test that uses surface-enhanced Raman spectroscopy, a highly sensitive optical measurement technique capable of detecting contaminants at extremely low concentrations. Researchers or testing technicians shine a laser on a water sample and can measure shifts in scattered light as photons (light particles that carry energy and information) interact with the contaminant’s molecular bonds. Every molecule scatters light differently, so that unique signature can be used to identify materials.
“Each chemical bond generates a unique Raman band,” Wei says. “It’s like a fingerprint. So if we capture the fingerprints of these organic molecules, we can know they’re there. With the enhanced signal, we can detect pollutants at the parts-per-billion or sub parts-per-billion level.”
But there are challenges, Wei says. The biggest is that water often contains other substances, such as natural organic material. That can make trace pollutants harder to detect, especially when their Raman bands overlap.
To address this, Wei is using machine learning algorithms in a two-step process. The first step identifies pollutants in “noisy” spectroscopy readings. The second step can help quantify how much of a pollutant is in a sample. Wei says he can infer pollutant density from how molecules arrange themselves on particle surfaces. Those arrangements change as concentration increases, which affects the relative intensities of a group of Raman bands.
The Safe Drinking Water Act regulates 53 organic compounds, many of which get into water through industrial or agricultural runoff. Wei says his research group plans to evaluate its method against the full list.
If the project proves successful, Wei says he sees his scanning method fitting in as an early-use option for water utilities, rather than replacing existing laboratory tests.
“This isn’t targeting replacing the standard, because the sensitivity isn’t comparable to a liquid chromatography-mass spectrometry,” Wei says. “Those can get down to the low parts-per-trillion. This would fit in as a pre-screening method that can flag violation events so you could then use the standard method to determine exactly how much of a contaminant is present.”
CAREER Awards also feature public outreach and education components. Wei will develop a machine learning or spectral analysis module for one of his graduate-level classes. That will use a case study based on drinking water in central Wisconsin. He will also work with undergrad research programs like Freshwater@UW and provide high school teachers with training in areas like data analytics and water contamination.
Featured image caption: Haoran Wei, a University of Wisconsin-Madison associate professor of civil and environmental engineering, is developing a low-cost test for drinking water. The test could function as a first step for identifying contamination in water systems. Photo: Joel Hallberg.