Meet the New Haven startup that wants to digest your plastic

Michael Magaraci, left, Protein Evolution enzyme platform scientist, explains his part of the enzyme development process with Brian Mansaku, center, chief process developer, and Connor Lynn, chief business officer, in their New Haven Lab in February. The company builds custom enzymes to digest plastic down to its component molecules for true recycling.

Tucked into an unassuming brick building in East Rock, a New Haven startup is fighting to replace the global recycling system. Protein Evolution, founded in 2021, has been quietly developing a new way to recycle plastic. It thinks it can eventually recycle polyester fabrics, rugs and other plastics that end up in overwhelmed landfills. The company says their primary competitor is the recycling system itself.

"Two-thirds of polyester goes into textiles, your polyester shirts, your running shorts, this rug," said chief business officer Connor Lynn, gesturing at the rug in the Protein Evolution offices. "Most people are familiar with recycling bottles, but we're not necessarily stopping there. ... We use biology to recycle."

The company builds and designs custom enzymes to break down polyethylene plastic (PET) and polyester into its component parts. These plastic components can then be filtered and synthesized into entirely new plastic. The products of the digestion are ethylene glycol and terephthalic acid, the building blocks of new polyester.

"Instead of seeing waste as this material that is basically end of life," Lynn said, "we can unlock the potential of it as a very cheap feedstock or source to create valuable chemicals, like those that go into producing polyester."

The process uses minimal energy and occurs at room temperature. If it scales, it would be an incredibly energy-efficient means of recycling.

Their lab is arrayed sort of like a funnel. At one end, powerful computers simulate possible enzymes, screening them for their ability to chew up plastic.

"We use computational methods, like AI, to predict the best enzymes," said Brian Mansaku, director of process development at Protein Evolution. "Computational methods are really good at identifying needles in the haystack."

After narrowing down possible enzymes from millions of candidates, a wet lab runs an array of tests to weed them. Toward the back of the lab, incubators with genetically modified bacteria grow the best enzymes at scale.

This is a carousel. Use Next and Previous buttons to navigate

The large test reactor in the lab space of Protein Evolution, a New Haven plastics recycling startup, is seen wrapped in black against the windows. The test reactor digests PET plastic into its component monomers.

Scrap textiles from fashion designer Stella McCartney await digestion testing in Protein Evolution's test lab. If successful, Protein Evolution will be able to recycle polyester fabrics in addition to more conventional recycled materials.

"We're really leveraging a lot of advances in synthetic biology," said Mansaku. "We're engineering biological units, the enzymes and the DNA that encodes them."

By a window overlooking an Industrial Revolution-era courtyard, a massive enzyme reactor stirs a solution of enzymes and PET powder, running the recycling reaction as a proof of concept. Michael Magaraci, director of platform engineering, said that eventually, the idea is to generate enough enzyme to build a plant to process 50,000 kilotons of polyesters annually.

"In the reactor over here, we actually combine the enzyme with the plastic to break it into a monomer," Margarci said, gesturing to a refrigerator-sized bottle with a mechanical stirring rod spinning away inside. "The rest of the equipment is just what we need to separate out the monomer to turn it back into plastic."

It's like digestion. When you eat a banana, your body's enzymes go to work breaking down the proteins and starches into amino acids and simple sugars. You take the big, complex molecules and break them down into parts. Protein Evolution's process takes the long chains of PET and breaks them into their components. Those monomers can be purified and recombined into plastic indistinguishable from plastic newly synthesized from fossil fuels.

"In theory, if you have a good source of plastic and a way to do this, you can basically use plastic like crude oil," said Benjamin Elling, a professor of chemistry and plastic researcher at Wesleyan University. He explained that the existing infrastructure of the chemical industry was already very good at taking small carbon chains and putting them to use.

Protein Evolution is a New Haven startup company founded in 2021. The company builds custom enzymes to digest plastic down to its component molecules for true recycling. "It's part of a circular economy," said Connor Lynn, chief business officer.

Most plastic doesn't get recycled. Lots of the stuff that enters single-stream recycling bins — including toys, plastic film, bags, and fabrics — ends up in a landfill. Much of that gets caught up in the recycling machinery, forcing shutdowns to disentangle equipment.

"Right now, plastic recycling is not circular," said Kim O’Rourke, recycling coordinator for Middletown. She explained that most recycling only happens once, if at all. "Your plastic container is mechanically broken down ... turned into a T-shirt or carpeting or filling for a sleeping bag, and after that, it goes in the trash."

Normal recycling is more accurately called downcycling. When recyclable plastic is shredded and melted down for reuse, it degrades the chains of molecules that make up plastic. To be a useful product, virgin plastic is often added to recycled plastic to strengthen it. This is why products often advertise the percentage of recycled material used.

The normal recycling process is often costly and imperfect. Very little plastic actually ends up being recycled from single-stream sources. In the past 40 years, less than 10 percent of plastic produced has been recycled. According to an analysis by Beyond Plastics, a research group based at Bennington College in Vermont, in recent years, this rate has declined to 5 percent.

"Plastics go through a sorting process, shipped to a plastic recycler, then they get chopped down, made into pellets and turned into a new product," O'Rourke said. She explained that this was only the process for bottles and large plastic food containers. Other plastic or mixed material containers typically don't get recycled due to sorting and size issues.

According to an NPR investigation, in the 1980s, the plastics industry had "serious doubts" that recycling would ever work but spent millions of dollars promoting recycling in a concerted greenwashing campaign.

"We're not in a great place, honestly," Elling said. "Globally, we produce 380 million tons of plastic every year, and 75 percent of that goes into a landfill in the United States. Lots of it gets incinerated."

A major part of the problem is that the plastics entering a recycling stream have to be of fairly high purity and quality. This is why containers, like bottles, are the most frequently recycled products. Bottles are made of high-purity, uniform plastic. Bottles are also easily broken down and shredded into meltable fragments.

"It's difficult to get clean material," O'Rourke said. "With single-stream recycling, all that recycling goes in together ... so people don't know what the rules are, and plastics are especially confusing, so people over-recycle."

Fabrics, fibers, films, packaging and rugs get caught in the machinery and can't be broken down. Lower-quality plastics, dirty plastics, or plastic that has been left out in the environment are often too low-quality to degrade. As they degrade, they become microplastics.

If Protein Evolution is successful in its effort to scale up enzyme-based plastic recycling, it can circumvent a lot of these issues. Since there's no melting process, different kinds of items can be added to the enzyme reactor for digestion. This means that polyester fabrics and films could become recyclable. Small containers could be digested — even some mixed materials.

"You really can't heat a new Patagonia fleece and turn it back into a bottle," Elling said. "The advantage of something like this is it could break it down into small molecules that could, in theory, be separated out from other, non-digestible components."

The company has partnered with the high-fashion brand Stella McCartney to develop a process to recycle polyester fabrics and rugs. Samples from a recent design line await testing, rolled up in barrels. Stella McCartney is one of Protein Evolution's early investors.

"These are actually stock materials from her production line," said Lynn. "Otherwise, they'd go to an incinerator or landfill."

An enzyme-based process could also accept lower-quality container plastic than conventional methods. Enzymes only react to specific targets so anything they can't degrade could be filtered out during the purification process. This would dramatically reduce sorting costs and make recycling more affordable generally.

Protein Evolution is only focused on PET and polyester plastics at the moment. PET is the easiest plastic to recycle due to its molecular structure and some of the commonest in circulation. Lynn explained that they would have to build new enzymes for each plastic, like nylon, elastane (Spandex), or other kinds of polyester.

"What this looks like now is a single-enzyme system," said Lynn. "As we go to more complex materials, you can envision a cocktail, a multi-enzyme system."

That may be more difficult for some of the higher-number plastics. Many plastics like Styrofoam, vinyl and acrylic plastics don't have as easy of a chemical backbone to degrade. Unlike polyesters, which are bound by a carbon-oxygen bond, other polymers are made of carbon-carbon chains, which are very strong.

"Our bodies make carbon bonds all the time. We have enzymes that do it, but it's an energy-intensive process," said Elling. "To go in the reverse direction, you need to put in more energy, or you need a really good catalyst that will lower the energy barrier."

But all of that is a distant problem. For now, Protein Evolution is focused on scaling up. Lynn hopes they can have a scale demonstration facility up before the end of 2025.

"We've made a tremendous amount of progress," said Lynn. "We're generating real monomer that's being sent out to be produced into new plastic. From there, it's all about scale."