Living Plastic Can Self-Destruct on Command

The film looks like ordinary polycaprolactone, the same polymer used in 3D-printing filaments and some dissolvable surgical sutures. Flexible, translucent, unremarkable. But sealed inside it, dormant and waiting, are two populations of engineered bacterial spores with a preprogrammed task: given the right signal, they will wake up, work together, and consume their own home entirely. Zhuojun Dai, a synthetic biologist at the Shenzhen Institutes of Advanced Technology, put it directly. “The realization that traditional plastics persist for centuries, while many applications, like packaging, are short-lived, led us to ask: Could we build degradation directly into the material’s life cycle?” His team’s answer, published in ACS Applied Polymer Materials, is a living plastic that can reduce itself to small molecules within six days, leaving behind no microplastic fragments at all.

The notion of living plastics has been circulating in synthetic biology for a few years now. The broad idea is to embed microbial spores into a polymer matrix during fabrication, so that the material functions normally during use but can be triggered to biodegrade at end of life. Pioneer work showed it was at least technically possible: spores are formidably tough little structures, capable of surviving the heat and organic solvents typically involved in plastic processing. The problem was efficiency. Existing living plastic systems relied on a single bacterial strain secreting a single enzyme, and that single enzyme, it turned out, was not quite up to the job on its own.

The Problem with Processive Enzymes

The fundamental difficulty has to do with plastic’s physical architecture. Solid polymers are semicrystalline, meaning their long molecular chains fold into ordered regions interspersed with disordered ones. Processive enzymes, the kind that grab one end of a polymer chain and walk steadily along it, snipping off monomers as they go, are well suited to complete degradation in principle. In practice, though, those chain ends are often buried or inaccessible, pinned at the interface between crystalline and disordered regions. The enzyme has nowhere to start. It’s a bit like trying to unroll a tightly wound ball of string when you can’t find the loose end.

Dai and his colleagues, including Jin Geng and Dianpeng Qi, solved this by looking to natural microbial ecosystems for inspiration. In the wild, microbes rarely work in isolation; complex communities partition labour among specialised members, gaining efficiency no single organism could manage. The team built an artificial version of this logic into their plastic.

They engineered two separate strains of Bacillus subtilis, a thoroughly well-studied and safe soil bacterium, each programmed to secrete a different enzyme after germination. The first produces a lipase from Candida antarctica that acts by random scission: it hits polymer chains wherever it happens to encounter them, snipping them into shorter fragments and, crucially, generating new free chain ends throughout the material. The second strain produces a lipase from Burkholderia cepacia that works processively, latching onto those newly exposed ends and threading the polymer chains through its active site in a kind of ratcheting motion, cleaving off monomers one by one until nothing remains. Alone, each enzyme is limited. Together, they are rather efficient. The random-scission enzyme effectively unlocks the doors; the processive enzyme walks through them.

To fabricate the living plastic, both spore populations were dissolved into a toluene solution of polycaprolactone (PCL) pellets and cast into films. The spores, having survived heat treatment at 100 degrees Celsius and a 24-hour soak in toluene without losing viability, were dispersed throughout the polymer matrix. Mechanical testing found no meaningful difference between living and ordinary PCL films; tensile strength and melting temperature were comparable. The spores, it seems, are passenger enough not to disturb the host material’s structural character.

Triggering degradation required warming the living plastic in a nutrient broth heated to roughly 50 degrees Celsius. At that temperature, the PCL matrix softens enough to release the spores. They germinate, activate their programmed enzyme circuits on exposure to a sugar inducer, and begin their cooperative work. In experiments tracking plastic weight loss, the material had dropped to roughly 2 percent of its original mass within four days. Gel permeation chromatography showed both the main polymer peaks and the smaller fragment peaks disappearing by day six; liquid chromatography and mass spectrometry confirmed that what remained were small molecules below 500 daltons in molecular mass. “By embedding these microbes,” Dai has said, “plastics could effectively ‘come alive’ and self-destruct on command, turning durability from a problem into a programmable feature.”

A Wristband That Eats Itself

Perhaps the more striking demonstration involved wearable electronics. The team fabricated a four-channel electrode by laminating copper circuitry between two sheets of living PCL, using a hot press and laser cutting to expose the monitoring contacts. Worn against the forearm, the device recorded electromyography signals from arm muscles during hand movements with a signal-to-noise ratio comparable to electrodes made from conventional polyimide, even after a month of storage. At end of life, the same heating and nutrient treatment degraded the polymer encapsulation in about 12 days. The copper circuitry remained, recoverable and recyclable; only the plastic had vanished.

That copper recovery point is worth holding onto. Living plastics don’t dissolve electronics into an undifferentiated sludge; they selectively remove the polymer component, potentially simplifying the sorting problem that makes e-waste recycling so difficult today. Whether that proof of concept survives the journey from laboratory flask to industrial reality is a different question, and one the team acknowledges openly. Polycaprolactone was chosen precisely because it is among the most enzyme-susceptible polyesters, so how far the cooperating-enzyme strategy extends to tougher polymers like polyethylene or polypropylene remains to be seen. The trigger mechanism, too, depends on controlled heating and a liquid nutrient supply, conditions that are rather more tractable in a laboratory than in the ocean where most plastic pollution ends up. Future work, the researchers say, aims to develop a water-based activation route.

Still, the core design principle is modular. The enzyme pair, the bacterial consortium, and the polymer matrix were each optimised independently before integration, which means, at least in theory, each component can be swapped out. Different plastics could, in principle, host different enzyme pairs; different triggers could replace heat. Synthetic biology has become very good at programming bacteria to do specific molecular things at specific times, and the idea of embedding that programmability into the material itself rather than in a separate disposal process is, arguably, a more honest confrontation with the problem than designing for recyclability after the fact. Plastics are durable by design. Living plastics would be durable by design too, only with an expiry date written into the material at the molecular level.

For now, a wristband electrode that eats itself on command is the technology at its most evocative. Whether it ever becomes a bin liner is the longer experiment.

Frequently Asked Questions

How does the “living plastic” actually work?

The material contains dormant bacterial spores from two engineered strains of Bacillus subtilis, suspended within the polymer during fabrication. While dormant, the spores don’t affect the plastic’s performance. When the material is warmed in a nutrient broth (around 50 degrees Celsius), the spores are released, germinate, and begin secreting two complementary enzymes that cooperatively break the polymer chains down to their molecular building blocks within days.

Why do you need two bacteria instead of one?

Solid plastic is semicrystalline, and polymer chain ends are often buried and inaccessible. A processive enzyme that chews through chains from the end has nowhere to start. The first strain’s enzyme uses random scission, cutting polymer chains at random points to expose new chain ends throughout the material. The second strain’s processive enzyme then grabs those exposed ends and dismantles the fragments completely. Each enzyme is limited working alone; together, they achieve near-complete breakdown in days rather than weeks.

Does it leave behind microplastics?

No, and this is one of the key advantages of the two-enzyme approach. Because the random-scission and processive enzymes work cooperatively, fragments don’t accumulate. Mass spectrometry analysis of the degradation products showed the end result was small molecules below 500 daltons, not plastic fragments. Earlier single-enzyme attempts left persistent debris after weeks.

Could this be used in everyday plastics like packaging?

The current work used polycaprolactone, a relatively enzyme-susceptible polyester used in 3D printing and medical sutures, not the polyethylene or polypropylene found in most packaging. The researchers designed the system to be modular, with the enzyme pair and bacterial strains potentially swappable for different polymers, but extending the approach to tougher commodity plastics remains a significant challenge for future research.

What triggers the degradation, and could it happen accidentally?

Degradation requires two things simultaneously: a temperature around 50 degrees Celsius to soften the plastic enough to release the spores, and a liquid nutrient medium to support germination and enzyme production. Normal use conditions don’t combine those triggers, so accidental degradation is unlikely. However, developing a trigger that works in water at ambient temperatures, which would be necessary for addressing ocean plastic pollution, remains an open research problem.

Source: ACS Applied Polymer Materials, Tang et al., 2026. DOI: 10.1021/acsapm.5c04611