Advanced Materials Based on Plastic-Eating Enzymes (I): Recycling Challenges in Spain.

Does the yellow bin actually work?

There's a gesture most of us make almost every day, and it leaves us with a pleasant feeling: separating the packaging, walking to the yellow bin, tossing it in. You hear the dry thud of plastic hitting the bottom, and something inside you considers the matter closed. Done. I've done my part.

I made that gesture for years without thinking much about it. Then I spent four years on a thesis studying what exactly happens to that package, and now every time I open that yellow lid, I think about something different: what's the statistical probability that this piece of plastic and I will ever meet again?

If I'm being honest, it's not high.

Our municipal waste recycling rate was 43.2% in 2022, and Eurostat estimates it dropped to 41.4% in 2023. The European Union required us to reach 55% by 2025. You don't need to wait for the definitive data, which Eurostat won't validate until 2027, to know how that turned out: in July 2026, the European Commission sent Spain a reasoned opinion for failing to meet the 50% target that was due in 2020. A reasoned opinion is the last step before ending up in front of the EU Court of Justice. We're still arguing about the exam from six years ago while the next one has already come and gone.

In this article, I'm not here to tell you to stop recycling because it all ends up in the same place. Recycle. Keep doing it.

This is an article to explain why 43% is so hard to raise, what's being tried, and why in my thesis we decided to bet on a path that, until ten years ago, sounded like science fiction: teaching proteins to eat plastic.

The underlying problem: plastics are too good

Here's the irony that holds this entire story together. Plastics aren't a bad material. They're an extraordinary material. They're lightweight, cheap, moldable, waterproof, resistant to acids, bases, fungi, bacteria, and the passage of time. That's why we use them for absolutely everything. Since the mid-twentieth century, we've produced more than 8 billion tonnes.

The problem is exactly that list of virtues. What makes a package protect your food for months is the same thing that makes that package still be there a century from now. The chemical inertness we love so much on the supermarket shelf becomes an unsolvable problem at the bottom of the sea. A PET bottle can remain structurally intact for more than fifteen years submerged in the ocean, and its complete decomposition is projected to take centuries.

Part of that resistance has a beautiful explanation: crystallinity. The chains of a polymer like PET can organize themselves into ordered, tightly packed, nearly crystalline regions (between 30 and 50% of the material, depending on how it was manufactured). Think of a tangled skein of yarn versus a well-knitted sweater. You can push your fingers into the skein. Not into the sweater. Those crystalline regions are forbidden territory for any microorganism that might want to try.

And then there's the detail nobody talks about: modern plastic is almost never just plastic. It's a mixture. Multilayer, with additives, colorants, plasticizers, and inks. A juice carton is an engineering feat of three different materials laminated together, and that engineering feat is exactly what arrives at the recycling plant as a problem. To recycle it, you'd need to separate the aluminum, plastic, and cardboard to recycle all three separately without contamination.

The five pathways, and why none of them close the loop

When we talk about "recycling" in the singular, we're actually lumping together five technologies that look nothing alike.

Mechanical recycling is what almost everyone imagines and is the most widely used. You separate, wash, shred, melt, and remold. It's cheap, it's simple, and it's the backbone of the current system. Its problem is one of polymer physics: every time you melt a plastic, the chains break a little from thermal and mechanical stress. The material comes out worse than it went in. After a few cycles, your bottle can no longer be a bottle; it'll become a flower pot, then a filler fiber, then nothing. In the industry, they call it downcycling, which is a polite way of saying it's not a circle but a downward spiral. Add that it's extremely sensitive to contamination (a single wrong material in the batch ruins the whole run), and you understand why it needs clean waste streams that almost never exist in the real world.

Chemical recycling attacks the problem at its root: instead of melting the polymer, it breaks it down into its original monomers to rebuild from scratch. On paper, it's the perfect solution because the recovered material is virgin again. In practice, pyrolysis and gasification require temperatures between 400 and 800 °C. That's a lot of energy, a large carbon footprint, and complex equipment. Moreover, those temperatures produce mixtures of compounds that need to be separated and purified afterward, adding another layer of cost. Milder variants exist (hydrolysis, methanolysis, glycolysis), but they still demand aggressive chemical environments and significant amounts of solvent.

Thermal management (incineration with energy recovery) is honest about what it promises: it doesn't recycle anything; it burns. In return, it reduces volume by up to 90% and takes advantage of the fact that plastic has an extremely high calorific value — higher than almost any other waste. It's the only pathway that handles mixed and contaminated waste without complaint. The price is what you'd imagine: between 38 and 55 MJ per kilo processed, greenhouse gas emissions, and sophisticated flue gas treatment systems. And the material is lost forever. In the waste hierarchy, it sits just above the landfill, which is already saying very little. Nothing will ever be recovered.

Classical biological degradation (letting microorganisms do the work) is the most elegant from an environmental standpoint. An 80% volume reduction compared to landfill, no dioxins, no toxic emissions. The process occurs in four phases: biodeterioration, depolymerization, bioassimilation, and mineralization. It's beautiful. It's also useless for practical purposes because it takes decades or centuries. Microorganisms didn't evolve with PET in mind; they've been spending seventy years trying to catch up with a material that appeared in an evolutionary blink of an eye.

And this is where the fifth pathway comes in.

The idea: keep the tool and ditch the organism

In 2016, a discovery was published that changed everything. At a recycling plant in Japan, a bacterium, so-called Ideonella sakaiensis, was isolated. It had developed the ability to use PET as a carbon source. It wasn't eating it by accident: it had two specialized enzymes for the job. One, PETase, breaks the polymer. The other, MHETase, finishes the job and returns the two original building blocks of PET: terephthalic acid and ethylene glycol. And it did this at 30–37 ºC. Warm room temperature, not industrial furnace temperature.

The bacterium itself is slow. But the bacterium isn't the interesting part. The interesting part is its tools.

That's the conceptual leap of the entire field, and the one that explains why biological degradation and enzymatic degradation are different things, even though they sound similar. Instead of waiting for a living organism to go about its life at its own pace, you isolate the enzyme, produce it in quantity, optimize it, and put it to work in a reactor under whatever conditions you choose. You go from waiting on an organism to handling a catalyst.

And the advantages stack up all at once. You work under mild conditions, so energy costs plummet. You recover quality monomers, not degraded material, so you truly close the loop. Enzymes are specific, meaning you can target a particular polymer within a mixture without touching the rest (something mechanical recycling only achieves through expensive prior separation). You don't generate toxic byproducts. And you can process dirty, colored, or contaminated plastics that the mechanical pathway rejects outright.

If this sounds like a laboratory promise, I have a fact that brings it down to earth. Since September 2021, there has been an industrial demonstration plant operating in Clermont-Ferrand, at a Michelin Group site, run by the French company Carbios. Its 20 m³ reactor processes 2 tonnes of plastic (about 100,000 shredded bottles) and completely depolymerizes them within 10 to 16 hours. Not decades. In less than a day. And it's not alone: Samsara Eco in Australia operates at room temperature in an aqueous medium, Birch Biosciences is designing enzymes with generative artificial intelligence, and a consortium between NREL, the University of Massachusetts Lowell, and the University of Portsmouth has developed a process that reduces acid and base addition by more than 99 %, operating costs by 74 %, and energy consumption by 65 %.

Now the uncomfortable part

If you've read this far thinking "so it's solved," let me spoil it a little. Because honesty about the limitations is what separates science communication from a sales brochure.

First, the scope problem. The enzymes we know how to use attack ester bonds. That works with PET, with polyurethane, with polyesters in general. But look at the actual composition of European plastic waste: of the 57.2 million tonnes counted in 2021, polypropylene accounts for 16.6%, LDPE for 14.7 %, PVC for 11.4 %, HDPE for 9.3 %... and PET, our star substrate, barely 5.3 %. Polymers with a pure carbon-carbon backbone (PE, PP, PS, PVC) make up more than half of all waste, and they have no bond that a hydrolase can break. For those, you need prior oxidation, and the field is far less mature there.

Second, the enzyme problem itself. Proteins are exquisite and fragile machines. They work because they're folded in a very specific way, and that shape unravels with heat, with solvents, with time. Producing them costs money. Reusing them is difficult because, once you put them in the reactor with the plastic, recovering them from the medium is a problem in itself. And there's an uncomfortable design tension: the reaction goes faster at high temperatures, mainly because the plastic softens and the chains become accessible, but high temperature is precisely what destroys the enzyme. Protein engineering has gained ground here (variants like LCC-ICCG degrade 90% of crystalline PET working at 72 ºC), but the fundamental conflict remains.

Third, the access problem. An enzyme is a molecule floating in a liquid. Plastic is a solid. The entire reaction happens at the surface, at that boundary of contact, and the enzyme can only bite what it can reach. The crystalline regions we talked about at the beginning remain, to a large extent, impregnable.

And the bonus: bioplastics aren't what you think either

There's a twist in this story that deserves its own article, and it's worth noting here because it dismantles a very widespread intuition.

PLA, or polylactic acid (the material of "compostable" cups and much of 3D printing), is currently produced at a rate of about 2.47 million tonnes per year and is expected to reach 5.73 million by 2029. It's marketed as biodegradable. And it is, technically. The catch is in the fine print: it needs an industrial composting facility with controlled temperature, humidity, and oxygen. Outside those conditions, its degradation in the environment can take between 100 and 1,000 years.

Only 12 to 15 % of the PLA produced worldwide ends up in an adequate facility. The rest slips into the conventional plastic circuit, where it neither recycles well mechanically (it contaminates PET streams) nor degrades. Some 450,000 tonnes per year accumulate in landfills with the "biodegradable" label still on.

In Spain, the case is textbook: we have no large-scale domestic PLA production; we import it, and our industrial composting infrastructure for bioplastics is severely underdeveloped. In other words, we've imported the solution without importing the infrastructure that makes it work. A material designed to be part of the answer ends up fattening the problem, simply because the system that was supposed to receive it doesn't exist.

Why I ended up here

When I had to decide what to research for the next four years, this was the map I had in front of me. A massive problem. Five technologies, all with a structural flaw. And one of them (the enzymatic pathway) has the best ratio between what it promises and what stands in the way: it truly closes the loop, uses little energy, doesn't generate toxic waste, and has already proven it works at industrial scale.

Its three bottlenecks (the enzyme denatures, it can't be recovered, and it can't reach the substrate well) have something in common. None of them is a biology problem. All three are material problems. And if the problem is about materials, you can design a materials solution.

That was the bet of my thesis: to stop treating the enzyme as a simple biomolecule and start treating it as a component of a material. To combine it with polymers to build hybrids that protect it from heat, keep it anchored so it can be recovered, and place it right where it needs to be — even inside the plastic itself.

But that's chapter four. We still need to understand what exactly an enzyme is and how it can be capable of eating a bottle.

Back to the yellow bin: keep opening it. The gesture matters, and the system, with all its holes, works better with you than without you. We need to show that it works so that improving the system even becomes a conversation. It's just that now you know that on the other side of that lid there isn't a perfect circle waiting — there's a whole lot of people trying to close one.

In the next chapter, I'll introduce you to the culprits.