Among the technical inquiries spanning topics like cyclodextrin polymers and mass spectrometry that were stacked in Damian Helbling's inbox was a question about four golden retrievers.
A local woman had contacted Helbling, a professor in Cornell's School of Civil and Environmental Engineering and a leading researcher of trace organic chemicals in water, wanting to know if it was safe for her dogs to swim in a nearby pond.
Damian Helbling, professor of civil and environmental engineering.
"I also have a golden retriever that loves to swim and I absolutely sympathize with this question," Helbling said, "but I hesitate to give a concrete answer when people ask me such specific questions about the safety of their water. As a researcher, I tend to look at the situation through a totally different lens."
For Helbling, the question isn't simply whether a pond is safe. It's what science can do to better understand the problem, and what engineering can do to solve it.
Few contaminants illustrate that challenge better than PFAS, or per- and polyfluoroalkyl substances, commonly known as "forever chemicals." The family of synthetic chemicals includes thousands - and by some definitions, millions - of compounds used in products ranging from non-stick cookware and firefighting foams to medical devices and food packaging. Their defining characteristic is the incredibly strong bond between carbon and fluorine atoms, which makes them exceptionally durable but also remarkably difficult to break down in the environment.
That persistence has made PFAS a growing public health concern, with studies showing that exposure to some PFAS may be linked to a variety of harmful health effects. As policy makers reconsider drinking water standards and monitoring of industrial PFAS discharges, environmental engineers like Helbling are racing to better understand, detect and remove the chemicals from water supplies.
Faster solutions for a forever problem
Industrial production of PFAS began to ramp up in the 1940s and continues today. As a result, "forever chemicals" are found nearly everywhere - in water, air, soil and even in the blood of people and animals throughout the world, according to the U.S. Environmental Protection Agency.
One of the best ways to solve the problem, Helbling said, is to limit the production and use of PFAS in the first place. He and his research group are helping advance that effort through "Faster Solutions for Forever Chemicals," an IBM-led project funded by the National Science Foundation that is developing computational tools to help manufacturers identify PFAS in their supply chains and replace them with safer alternatives.
Doctoral student Fanshu Geng in the Helbling Research Group's Hollister Hall laboratory.
"Companies don't necessarily know where or how they are using PFAS in the manufacturing processes," Helbling said. "They just know that there is a product they can acquire from some supplier that helps them make their widget faster or better or more resilient. We are trying to help companies track down PFAS in their supply chains and discover new chemistries that can provide the same benefits while being less persistent in the environment and less toxic."
That challenge is easier said than done. PFAS are widely used because they can resist heat, repel water and withstand harsh manufacturing conditions, among other properties. Replacing them requires understanding exactly what makes each PFAS useful for a particular application before designing a new molecule that delivers the same benefit.
Rather than synthesizing and testing thousands of potential replacements in the lab, the "Faster Solutions for Forever Chemicals" project is using IBM's expertise in AI to screen candidate molecules and predict which candidates are most likely to deliver the desired performance and have the smallest impact on the environment.
The team also wants to avoid "regrettable substitutions" that replace one harmful chemical with another.
"That's why we're using AI to answer questions like: Is this replacement molecule going to be persistent in the environment? Is it going to accumulate in organisms? Is it going to be a public health threat?" Helbling said. "Eliminating hazardous chemicals from manufacturing processes overnight isn't realistic, but understanding where or how they are used and pursuing safer alternatives is essential."
But what happens to the PFAS that still can't be replaced?
Engineering molecular nano sponges
When safer alternatives aren't yet possible, the problem changes. It's no longer how to eliminate PFAS from manufacturing processes, it's how to capture and destroy them before they enter the environment.
To do this, Helbling has spent the past decade designing specialized polymer materials that selectively capture PFAS from water. He describes them as molecular nano sponges.
"Imagine dropping a tiny sponge into a glass of contaminated water," Helbling said. "Instead of soaking up the water itself, the sponge selectively attracts PFAS molecules. Remove the sponge, and the water is cleaner. That's the idea that motivates this work."
Samples to be studied in the laboratory of Damian Helbling, professor of civil and environmental engineering.
Behind that simple analogy lies sophisticated chemistry. The polymers contain molecular receptors engineered to recognize the unique features of fluorinated compounds. Because different industries use different kinds of PFAS, the materials must also be customizable.
Helbling compares the process to building a molecule with a set of Legos.
"We've learned how to remove certain functional groups from the polymers and put new ones on to tailor these materials for specific applications," Helbling said. "We can look at a new PFAS and say, 'I bet we need the Lego set to look like this for this situation.'"
The technology grew out of a long-running collaboration with polymer chemist Will Dichtel, formerly of Cornell and now at Northwestern University. Helbling and Dichtel are now commercializing the technology through a startup company and partnerships with other stakeholders.
Once concentrated inside the nano sponges, the chemicals can be destroyed using emerging chemical treatment methods. That approach is far more practical than trying to destroy PFAS dispersed throughout millions of gallons of water, Helbling said.
Of course, before PFAS can be removed they first have to be found. While most researchers focus on a relatively small number of well-studied compounds, many other PFAS have never been fully characterized - and in some cases, scientists don't even know they exist.
The hunt for hidden PFAS
Every mystery starts with a clue.
For Helbling, the first clue to discovering a hidden PFAS often appears as an unexplained signal inside an acquisition from his high-resolution mass spectrometer. His lab uses the technology to analyze water samples without knowing beforehand which chemicals they contain. The resulting data becomes a molecular fingerprint of everything in the sample. From there, Helbling's team uses computational tools to search for subtle patterns that point to previously unidentified PFAS.
Doctoral student Vie Villafuerte in the Cornell NanoScale Facility.
"This is a really valuable technology, especially when you have new industries thinking about this problem and becoming more concerned about PFAS in their wastewater," Helbling said. "We have helped industries quantify known PFAS in their wastewater emissions, and we have also helped them discover more exotic and unknown PFAS that are unique to their industry or even their company."
The work has proven especially valuable for the semiconductor industry, where manufacturers often use proprietary fluorinated chemistries that aren't included on standard monitoring lists. Through collaborations with the industry and Cornell's NanoScale Science and Technology Facility, Helbling's lab has characterized PFAS that are uniquely associated with semiconductor manufacturing processes such as photolithography and plasma etching. Their research was among the first to identify these compounds in semiconductor wastewater, and that work led Helbling to being selected as the 2023 recipient of the Sustainable Future Award from the Semiconductor Research Corporation.
While semiconductor manufacturing is not considered the primary source of PFAS contamination, the research is helping manufacturers identify opportunities to develop safer alternatives at a time when the United States is aggressively expanding domestic chip manufacturing.
Helbling said the work illustrates how environmental engineering has become critical to helping tech industries grow responsibly.
"The tech industry is aware of emerging concern and regulations on trace chemicals such as PFAS in their emissions," Helbling said. "My industry partners tell me that they need environmental engineers to help them innovate in a responsible way. If a Cornell student wants to be in the tech industry but their passion is in the protection of fragile ecosystems or public health, environmental engineering is absolutely a way to do both."
Following forever chemicals
If PFAS have a future, Helbling wants artificial intelligence to predict it.
AI has become one of his lab's most powerful research tools, helping to forecast how forever chemicals will transform, travel and accumulate long before those processes unfold in the real world. By analyzing decades of scientific literature, machine learning models can predict how newly designed PFAS are likely to behave once they enter the environment.
Postdoctoral associate Liqing Yan and doctoral student Fanshu Geng collect water samples from a creek in Ithaca.
"There are an uncountable number of PFAS that may exist in the environment, and many of them transform over time through natural environmental processes to end up as the relatively few PFAS that the Environmental Protection Agency is focused on regulating," Helbling said. "We have developed tools to predict those transformations to forecast how the chemical composition of a contaminated site might evolve over years or decades."
Another project uses AI to answer a different question: Whose PFAS are these anyway?
Supported by the U.S. Department of Defense, the research examines the chemical fingerprints left behind by different sources, including firefighting foams, landfills, wastewater treatment plants and industrial facilities. After training machine learning models on well-characterized samples, the models can predict where contamination originated.
"That information could help industries and regulators prioritize cleanup efforts and identify responsible sources," Helbling said. "And these methods are transferable. They're applicable to lots of environmental problems beyond PFAS."
The lab is also using machine learning for the FLX PFAS Project, founded by Helbling and his students to collect and analyze water samples throughout New York's Finger Lakes region. The results are then shared online with the public. As the project's dataset grows, Helbling hopes researchers will soon be able to predict the likelihood that PFAS are present in any body of water in the Finger Lakes region.
Helbling knows people want certainty. They want to know whether the water is safe, whether their families are at risk and, sometimes, whether their dogs can go for a swim. Science can't always provide simple answers to people. Environmental engineering, however, offers the promise of better solutions.
