Practical implementation of a Safe-by-Design framework: a proof of concept for early-phase hazard assessment of PFAS-free formulations
This paper demonstrates the practical implementation of a tiered Safe-by-Design framework within the EU-funded PROPLANET project, utilizing regulatory data mining, in silico screening, and in vitro testing to assess and guide the development of PFAS-free formulations for technical textiles, food-packaging machinery, and glass applications while highlighting methodological gaps and the need for further hazard characterization.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
For decades, the world of industrial coatings has relied on a family of synthetic chemicals known as per- and polyfluoroalkyl substances, or PFAS. These materials are prized for their ability to repel water and oil, resist heat, and withstand harsh chemicals, making them ideal for everything from non-stick cookware to protective gear. However, the very strength of the carbon-fluorine bonds that make PFAS so useful also makes them nearly impossible to break down in nature. They persist in the environment for centuries, accumulate in living organisms, and have been linked to serious health issues, including immune system damage and increased cancer risk. As governments move to ban these "forever chemicals," scientists and engineers face a difficult challenge: how to create new materials that offer the same high performance without the toxic legacy. The solution lies in a philosophy called Safe-by-Design, which asks researchers to consider safety and sustainability at the very beginning of the creation process, rather than trying to fix problems after a product is already made.
A team of researchers from across Europe recently put this philosophy to the test in a practical, real-world setting. Working within the EU-funded PROPLANET project, they set out to develop new, PFAS-free coatings for three specific uses: technical textiles, metal parts in food-packaging machinery, and low-maintenance glass. Instead of simply swapping one chemical for another, the team built a rigorous, step-by-step process to screen their new formulas for safety before they ever reached the market. Their approach combined a deep dive into existing safety records, computer-based predictions, and a series of biological tests using cells and tiny organisms. The goal was to catch any potential dangers early, allowing them to discard risky formulas and focus their resources on the safest, most effective options.
The process began with a thorough review of every single ingredient the team considered for their new coatings. They scoured safety data sheets and official European chemical databases to identify any substances known to cause cancer, damage genetic material, or harm reproduction. This initial screening acted as a strict filter. If a chemical was officially classified as dangerous, it was immediately removed from the list of candidates. The team also looked for gaps in the data. For many of the new, specialized chemicals they wanted to use, there was little to no existing safety information. To fill these holes, they turned to computer models that predict how a chemical might behave based on its molecular structure. However, they quickly discovered that standard computer models were not built to handle the specific types of silicon-based compounds they were working with. The predictions were often unreliable because the chemicals were too different from the ones the models had been trained on. To solve this, the team developed their own custom computer models specifically tuned to these new materials, ensuring their digital screening was accurate.
Once the ingredients were vetted, the researchers moved to the next stage: testing the actual mixtures. This was a crucial step because chemicals in a mixture can interact in ways that change their toxicity. A substance that is safe on its own might become harmful when mixed with others, or the mixture might create new byproducts. The team tested their liquid formulations on human lung cells and a tiny, transparent worm called C. elegans, which is a standard model for studying reproduction and development. They looked for signs of cell death, DNA damage, inflammation, and hormonal disruption. The results were immediate and decisive for some candidates. One formulation designed for textiles, which contained zinc oxide, caused significant damage to the cells and DNA. Despite the ingredient's potential benefits, the clear evidence of harm led the team to abandon that specific path entirely. Similarly, a formulation for glass that included a suspected carcinogen was dropped because the safety risks outweighed the performance gains.
For the formulations that passed the initial safety checks, the team conducted a second, more advanced round of testing. These experiments used more complex models, including a three-dimensional lung tissue model that mimics the airways more closely than simple cell cultures. They also tested for long-term risks like cancer development and reproductive harm. The results here were nuanced. One textile formulation, NIC 4.1, showed no signs of danger in any of the tests, marking it as a strong candidate for further development. However, the story was different for the coatings intended for glass and metal. While these mixtures performed exceptionally well in their intended jobs, they raised red flags in the safety tests. The metal coating showed signs of damaging DNA, and the glass coating showed potential issues with reproduction. The researchers did not rule these out immediately, but they flagged them as needing significant further investigation and reformulation before they could be considered safe enough for the next stage of development.
The study ultimately demonstrated that a structured, tiered approach to safety is not just a theoretical ideal but a practical tool for innovation. By integrating safety checks into the earliest stages of design, the team was able to identify and eliminate hazardous paths before investing time and money into full-scale production. They found that relying on a single method was insufficient; the combination of data mining, custom computer modeling, and biological testing provided a complete picture that no single tool could offer alone. The work also highlighted the challenges of working with complex mixtures, particularly when ingredients are difficult to dissolve or react unpredictably in a test environment. These difficulties can sometimes obscure the true risks, requiring careful interpretation of the data.
In the end, the project succeeded in proving that it is possible to design safer alternatives to PFAS without sacrificing performance, provided that safety is treated as a core design parameter from day one. The team identified a promising textile coating that is ready to move forward, while clearly identifying the specific safety hurdles that must be overcome for the glass and metal applications. This work serves as a blueprint for how the chemical industry can navigate the transition away from hazardous substances, using a blend of modern technology and biological testing to ensure that the products of tomorrow are both effective and safe for people and the planet.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.