A cheaper enzyme is quietly rewriting the plastics ledger

The most interesting molecule in plastics right now was found in a compost heap. A team of French scientists in Toulouse, working with a company called Carbios, spent years engineering a natural enzyme — a protein that eats through plant matter — until it could do something new: chew a plastic bottle back into its original chemical ingredients, in a few hours, in warm water.

That sounds like a magic trick. It matters because of what recycling has actually meant, until now.

The plastic bottle you drop into a recycling bin is usually not turned back into a bottle. It gets shredded, melted, and reused for something lower-quality — a park bench, a fleece jacket, a piece of packaging that will end up in a landfill one cycle later. Every trip through the melting process shortens the plastic's molecular chains a little, and the material gets weaker. Recycling, in its ordinary form, is really a slow route to the dump.

The Toulouse enzyme skips all of that. Instead of melting the plastic, it breaks it back down into the two chemicals it was originally built from. Those chemicals are identical to the ones an oil refinery would produce from crude. A bottle can, in principle, become a bottle again, and then a bottle again, without ever losing quality.

Carbios broke ground on a full-scale industrial plant in northeastern France in 2024, with L'Oréal, Nestlé, PepsiCo and the Japanese drinks giant Suntory lined up as its first customers. When it runs at full capacity, around 2027, it will handle roughly 50,000 tonnes of plastic a year. That is a huge number in laboratory terms and a tiny one against the 400 million tonnes of plastic the world produces annually. Nobody at the company pretends otherwise.

The economics are the honest problem. Making virgin plastic from oil is cheap, especially when oil is cheap. The enzyme-recycled version costs more. What is changing that math is regulation: the European Union agreed in 2024 that plastic drink bottles sold in the bloc must contain at least 30 percent recycled material by 2030. Suddenly, brands that spent a decade making vague sustainability promises need actual recycled molecules, and the enzyme is the only route that produces a molecule good enough to trust.

Parallel work is opening up the target list. A team at the University of Portsmouth in the UK, working with an American national lab, published an even faster enzyme variant in 2022. Danish and Dutch biotech companies are quietly building similar tools for other kinds of plastic — the tough foam in car seats, the nylon in your clothes — that ordinary recycling has never really been able to handle.

The scepticism is fair. One 50,000-tonne plant does not solve a 400-million-tonne problem. The enzymes that eat drink bottles do not touch the polyethylene bags that end up in the ocean. And an enzyme plant powered by coal is not really a climate story.

But the direction is legible for the first time. A pathway now exists in which the recycled material is not a downgrade. The next decade will decide whether policy, capital and the enzyme scale together fast enough to matter. The molecule, at least, is no longer the thing holding it up.