Solvent-based recycling and AI could be the key to solving plastic waste

Plastics are everywhere—in the packaging around your food, the insulation in your home, and even the parts of your phone. They make life easier and safer, but there’s a cost. Nearly 75% of plastic ends up in landfills or the environment, and just 9% is actually recycled. Scientists say that this needs to change. But how can recycling become more efficient, and can we really switch to better alternatives?

A group of researchers at the University at Buffalo believes that the answer lies in a smarter, more integrated way of thinking. Led by Dr. Aurora del Carmen Munguía-López, an assistant professor of chemical and biological engineering, their recent review explores how process systems engineering—an approach that uses design, analysis, and tools like machine learning—can reshape the future of plastics. Their findings take a deep dive into the options, trade-offs, and technical challenges of managing plastics sustainably.

Despite their bad reputation, plastics are essential. They help keep food fresh, lower the weight of cars and airplanes to save fuel, and make medical equipment safer and cheaper. "Eliminating the use of plastics is not currently a viable option," says Dr. Munguía-López.

The problem isn’t with the material itself—it’s with how we manage it after use. Without better systems, plastic ends up in oceans, landfills, and even in our own bodies. Exposure has been linked to cancer, fertility issues, and respiratory problems. So the goal isn’t to remove plastics from our lives entirely, but to manage them in smarter and cleaner ways.

Traditional recycling methods are limited. They work well for certain types of plastic, but not for more complex materials like multilayer packaging. That’s where solvent-based recycling comes in.

Solvents can selectively dissolve specific polymers from plastic waste. This allows them to remove valuable materials from contaminated or mixed plastic streams. One study, led by researchers at the University of Wisconsin-Madison and co-authored by Dr. Munguía-López, showed that this method was the most cost-effective way to recycle multilayer films used in packaging coffee grounds.

But it’s not a perfect solution. While it emits less greenhouse gas than some other advanced methods, it can still produce more emissions than traditional recycling if not managed carefully. Cooling methods, for instance, are better than heating methods when reforming polymers, according to several studies.

“The best approach is likely combining both solvent-based and traditional recycling,” says Munguía-López. This hybrid method could offer higher recovery rates while keeping emissions low.

Meanwhile, artificial intelligence is making sorting and planning much more precise. One model, PlasticNet, developed at the University of Wisconsin-Madison, reached over 87% accuracy—and even 100% for some plastics. Other tools use natural language processing to sift through scientific literature, pulling out important data to guide further development.

AI is also expected to help on a broader level. It could improve transportation planning, help coordinate different players in the plastics industry, and simulate the outcomes of new policies. “AI models will also be needed to address demands at the supply chain level,” explains Dr. Munguía-López.

Biobased plastics sound like a great idea. They’re made from renewable sources like corn and sugarcane and are sometimes compostable. But switching to them on a large scale comes with problems.

For one, growing the crops takes up land and water—and that land might otherwise be used to grow food. Also, composting bioplastics requires specialized facilities. Most places don’t have those yet, and separating biobased plastics from traditional ones isn’t easy for consumers.

Dr. Munguía-López points out that to truly understand their value, we have to look at the full life cycle of these materials. That means tracking everything from raw material extraction and manufacturing to disposal and emissions.

“We can’t validate biobased plastics until we consider the impact of their entire life cycle,” she says. So while biobased plastics offer promise, the switch won’t be simple—or quick.

Plastic pollution isn’t just an environmental problem. It’s also a social and economic one. That’s why researchers are calling for a systems-level approach. This means considering not just how to recycle or compost plastics, but how every step in the process connects.

That includes designing better products, improving waste collection, building the right facilities, and changing behavior at the consumer level. All of this requires coordination across many sectors—from scientists and engineers to policy makers and businesses. Process systems engineering can guide that effort. It offers tools for modeling and simulating complex systems, identifying weak points, and testing new solutions before they’re rolled out on a large scale.

In their review, the University at Buffalo team also highlights key gaps in current research. Many strategies are still too focused on single solutions. There’s a need to integrate technologies—like combining AI with solvent-based recycling—or to analyze how changes in one part of the system affect the others.

Making plastics sustainable is a huge task. But it’s not impossible. Solvent-based recycling offers an affordable and scalable solution for now. Artificial intelligence adds the precision and speed needed to handle growing waste streams. And biobased plastics, while not ready for full adoption, could be part of the solution down the line.

But none of these options will work on their own. It will take an entire shift in how we design, use, and manage plastics. As Dr. Munguía-López says, “We need holistic and comprehensive approaches, but to consider the pros and cons of those approaches throughout their entire life cycle.”

The future of plastics isn’t just about materials—it’s about making smarter systems to manage them all.

Research findings are available online in the journal Industrial & Engineering Chemistry Research.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.