You are here

The Chemical Detective

What’s in our drinking water? Duke professor of civil and environmental engineering Lee Ferguson uses non-targeted analysis to gather clues about chemical contaminants, making it possible to identify them and trace them back to their points of origin.

Listen

Transcript

[Bells/intro music]

Miranda Volborth: This is Rate of Change, a podcast from Duke University dedicated to the ingenious ways that engineers are solving society’s toughest problems. I’m Miranda Volborth, and today I’m talking to Lee Ferguson, an environmental chemist and professor of civil and environmental engineering at Duke University. But for the purpose of this podcast, let’s think of Lee as a detective– a chemical detective, who uncovers clues about unknown water contaminants that help them to be identified and traced back to a source. Let’s set the scene.

June 7, 2017: the Wilmington Star-News breaks the story of an emerging contaminant that has been found at significant levels in the Cape Fear River, which supplies Wilmington, North Carolina with about 80% of its drinking water.

The chemical was eventually identified as GenX, and traced to the Chemours Company’s Fayetteville Works plant, situated about 100 miles upstream of Wilmington. That plant was using GenX to produce Teflon, the nonstick material coating our pots and pans.

GenX is chemically similar to an entire class of compounds that are known to be problematic for human health and the environment—per and poly fluorinated alkyl substances, or PFAS. These compounds are really useful in consumer products because they’re highly resistant to degradation—but, that same quality also makes them difficult to remove from water.  In fact, this class of compounds is known as “forever chemicals.”

Lee Ferguson: The discovery of GenX in the Cape Fear River was actually quite an accident. Mark Strynar at EPA had just purchased a high-resolution mass spectrometer for the first time, and he was starting to test that system… During that process, they happened upon some compounds that nobody had seen before in environmental waters, and were certainly not being monitored anywhere in the country, and certainly not in the Cape Fear River. Alarmingly, they found that these molecules are very persistent, and they're making their way through drinking water treatment processes, and contaminating the drinking water of Wilmington, North Carolina, as well as a number of other communities down the Cape Fear River.

Volborth: But how was this mystery chemical eventually identified and traced to a source? It’s thanks to that specialized piece of equipment that Mark Stryner was testing out.

Ferguson: The equipment that we rely on is a high-resolution mass spectrometer. This is a device that is able to measure with very, very high precision and accuracy the molecular weight of many, many, many molecules in a sample at very, very low concentration. Okay? At its core, that's what the equipment does, and so there's really two aspects that are important here, the accuracy with which you can measure the mass, and therefore, get clues to the identity to the different molecules, and then also the sensitivity that we can measure those molecules to.

A relevant question that we might ask is, "What's in our water?" It's a very open question, and it seems like the obvious question. It's much more difficult to answer that question scientifically than we might expect. In most cases in environmental regulation and environmental assessment, those analyses are made on a molecule-by-molecule basis. We decide a priori which chemicals we are going to look for in a water sample, or a human sample, or a food sample. And we go, and we target those molecules, and it's almost like putting blinders on a horse. You see only what's in front of you.

Volborth: There are a few hundred substances on the priority pollutants list established by EPA. Those are substances that are regulated. Those regulations are enforceable. So it makes some kind of sense to look for them on a case by case basis. But— there are about 85,000 chemicals in use that are not yet regulated, including but certainly not limited to PFAS.

Ferguson: EPA doesn't necessarily have the ability to maintain focus on all potential emerging pollutants that might cause health effects in drinking water and in water supplies. So just because EPA isn't monitoring for it doesn't necessarily make a compound safe, and so it really has fallen to researchers, to academic researchers in many cases, to do this kind of work, where we are essentially performing environmental reconnaissance and forensics to try to understand what are the compounds that are slipping through the regulatory cracks.

We use what I like to term a chemical fingerprint approach to do this. Molecules, just like individual people, have specific characteristics. They have identities, and to be able to observe and to assign those identities, we need to measure the properties of those molecules as accurately as we can. So if you think about facial recognition to try to identify a single person in a concert venue, we can imagine the same corollary to being able to fingerprint, and identify, and isolate individual molecules in a chemical soup.

Volborth: What do those fingerprints look like?

Ferguson: Sure. This is the hard part of doing high-resolution mass spectrometry. Whereas our eyes as humans are really sort of tuned to understand a face or a fingerprint, an actual human face or a human fingerprint. We can understand and visually see the difference between individuals. We are not trained as humans to be able to recognize intrinsically the differences between chemical fingerprints.

Therefore, we have to rely on software approaches, things like machine learning and other types of data reduction techniques to be able to interpret the spectrum, which is a mass spectrum that is output by the mass spectrometer, and to take those spectra, and turn them into chemical structures that chemists like myself can understand and compare with each other. And the process of going from spectrum to structure is extremely complex, and it's also an active area of research currently on how we actually make those correlations.

And so there are specific chemical reactions that happen in these mass spectrometers, which result in the chemical spectra and the mass spectra that we actually measure. And so the data reduction is one of the hardest parts about this entire process, and it's really led me down quite a rabbit hole from being a wet chemist, bench chemist, to learning how to do complex data processing and data science routines, using things like machine learning and other kinds of approaches to take information and turn it into knowledge, and that's what our algorithms are designed to do.

Volborth: Lee really became involved with the GenX situation in 2018, because of his expertise in environmental chemistry.

Ferguson: The North Carolina Coastal Federation asked me to speak on their behalf in front of the North Carolina general assembly about the issues of unregulated contaminants in our water supplies. And this was initiated, of course, as a direct result of the GenX contamination in the Cape Fear River. However, the legislature was worried about whether this could be a bigger problem.

What if there were unrecognized and emerging contaminants that are in water supplies in other areas of North Carolina that simply had not been detected? So we set out to develop a monitoring network, and this monitoring network is termed the Per- and Polyfluorinated Alkyl Substances Testing Network, or PFAST Network, and it's a consortium of researchers from all the major universities in the state. And we have gone around to every public water supply in the state of North Carolina, and tested those waters for this class of per- and polyfluorinated alkyl substance, or PFAS.

Volborth: Lee’s team spent a year and a half driving around the state to collect samples from more than 400 sites, that they brought back to the lab for analysis.

Ferguson: Now, in many cases, we found exactly what we expected. So for example, in the Cape Fear River, unsurprisingly to us, we found elevated levels of PFAS compounds coming from the Chemours Fayetteville Works Plant. However, in some cases we were surprised. One of the biggest surprises that we encountered was the small town of Maysville North Carolina, which it's just a small community in Eastern North Carolina that relies on a single drinking water well, and that drinking water well pulls from an aquifer that is somewhere around 100 to 300 feet below the surface. That drinking water well at some point had been contaminated by what we believe is firefighting foams that contain PFAS.

So, of course, their water at Maysville, the water supply had never before been tested for PFAS chemicals, because as I mentioned, these compounds are not on any regulatory list. So there are no set water quality criteria for these compounds currently. When we measured the drinking water supply from Maysville, we found that several of these PFAS compounds were present at levels higher than the EPA's health advisory level that they had published for two specific PFAS compounds, perfluorooctanoic acid and perfluorooctane sulfonate. We immediately notified the town, and we notified their representatives in the general assembly, and we notified the utility that was responsible for this water supply.

Within a matter of days, we were able to verify that their neighboring community, Jones county, had water that was devoid of measurable PFAS compounds, and they were able to switch their water supply over to the Jones County supply.

Volborth: Water from the Haw River, sampled not too far downstream of the Greensboro’s textile mills, were covered in suspicious chemical fingerprints, leading Lee and his Duke colleague Heather Stapleton, to delve more deeply into what these chemicals were and where they had entered the river. At this point, their guess is that the mills manufacturing stain-resistant textiles is probably the culprit.

Of course, chemical contamination of drinking water is not limited to North Carolina—it’s just that we have the network in place to help detect it. In fact, Lee is putting this experience to work for communities halfway across the world.

Ferguson: The project that we've initiated in Sri Lanka with my colleague, Nishad Jayasundara, he's a new faculty in the Nicholas School of the environment… there's a phenomenon called a CKDu, chronic kidney disease of unknown etiology. And so this is a phenomenon that is happening around the world, including actually areas here in North Carolina, in southeastern North Carolina, but in Sri Lanka, there are a number of communities that have experienced chronic kidney disease in populations that don't seem to be associated with any particular genetic risk, or sort of there's very few other population factors that seem to explain the occurrence of this disease.

There is some indication that the incidents of the disease may be connected to water supplies, and specifically water supplies that are associated with agricultural areas in Sri Lanka. And so it turns out that in Sri Lanka, as well as many other places, pesticides and herbicides are used quite frequently in agricultural processes, and because of limited resources for water treatment, it's possible that many of those agrochemicals may be entering the water supplies of the people in the area. And so what we are working on is to try to examine and assess links between specific water contaminants as well as non-targeted analysis of water to understand what are the types of chemicals that might be present in these water supplies that are associated with high incidents of kidney disease.

So I'm sending one of my grad students over to Sri Lanka. He's leaving in about a week along with Nishad Jayasundara to collect water samples, and to bring them back here to do detailed chemical analysis to understand those chemical fingerprints. And so our hypothesis is that specific chemical fingerprints and specific chemical agrochemicals species are associated with this chronic kidney disease in these areas, and so that's what we're going to be testing over the next months to years.

Volborth: What’s the outcome you’re hoping for, Lee? What’s your grand vision?

Ferguson: In the end, with all of this water quality work that we're doing, and all the technology development that we've been participating in in our laboratory to develop these non-targeted analysis tools, what I want to see is a emerging contaminant reconnaissance program that is built in place, so that we can have early warning of chemical discharges that might pose risks downstream. And so this may sound like a pie in the sky idea, but in fact, this is exactly the sort of processes that have been implemented in Europe, for example. On the Rhine River there's an example of a monitoring station that does exactly this. It's an emerging contaminant reconnaissance station with early warning, so that we can understand discharges when they're happening, and try to intervene immediately to protect downstream water supplies. This is all part of holistic water quality protection, and this needs to happen not just here in North Carolina, but across the country and around the world.

Volborth: Thanks for listening. Follow Rate of Change for the latest research from Duke Engineering. And if you learned something from this podcast, please share it with others.

Music: "Deadly Roulette" byKevin McLeod via incompetech.com. Licensed under the Creative Commons 3.0: By Attribution license.