Plastic-derived cookies explained:
- Engineered yeasts convert carbon from PET and agricultural waste into proteins, flavourings and nutrients for 3D-printed cookies.
- Nobody has yet tasted a µBite, and extensive safety and regulatory testing will be required before commercial use.
- The 33-step process currently takes up to two days and costs around $60 per kilogramme, making mainstream production highly unlikely.
We’ve had lab-grown meat, precision-fermented proteins and insect-based snacks, but is an edible plastic-derived cookie a step too far? Microplastics have already infiltrated water, seafood, meat, fruit, vegetables and packaged products, and evidence that these particles are making their way into the human body is mounting.
A 2025 study by University of New Mexico researchers detected micro- and nanoplastics in every liver, kidney and brain sample examined, with markedly higher concentrations in the brain. Other studies have reported plastic particles in human blood, lungs, gut, placenta and testes. Scientists still don’t know exactly what this accumulation means for human health, but plastic is clearly travelling far beyond the food and drink through which much of it enters the body.
Food manufacturers are therefore under growing pressure to understand how plastic particles enter products and how to keep them out. Now, in a development closer to dystopian fiction than conventional bakery innovation, scientists want people to eat something derived from plastic deliberately – raising the question of whether food science has finally lost the plot.
Researchers at Southern Illinois University (SIU) Carbondale have created a protein-rich, vanilla-scented cookie partly derived from a discarded plastic bottle. Called µBites, pronounced ‘microbites’, it uses genetically engineered yeast to transform polyethylene terephthalate (PET), crop residues and other waste into proteins, fats, nutrients and flavouring compounds. These are combined with starch, fibre and sweeteners, passed through a 3D printer and microwaved. The researchers insist the resulting cookie doesn’t contain plastic particles, but nobody has yet been authorised to eat one.
Turning bottles into biscuits
The project emerged from NASA’s Deep Space Food Challenge, which sought technologies capable of producing food during long missions where conventional agriculture, refrigeration and resupply would be severely restricted.
A round trip to Mars could take approximately three years, making it impractical to carry every meal astronauts would need. Materials aboard the spacecraft would consequently have to be reused wherever possible, creating an incentive to turn discarded packaging and inedible biomass back into nutrients.
“We were trying to develop technologies for plastic upcycling to make more valuable products,” says Dr Lahiru Jayakody, associate professor of microbiology at SIU Carbondale and leader of the research. “We thought, why not focus on making food? Because plastic is carbon and food is carbon.”
There are, however, several substantial steps between an empty drinks bottle and printable cookie dough.
The process begins with oxidative hydrothermal dissolution, a treatment developed at SIU Carbondale that uses water, oxygen, high temperatures and pressure to break PET and agricultural residues such as discarded corn stalks and leaves into carbon-rich compounds that microbes can access.
Engineered yeast strains are then fed those compounds and programmed to convert them into food components. “Microbes are very clever, so we are using their traits to solve the problems we created,” says Dr Jayakody.
The latest phase of the research, presented at the American Chemical Society’s Fall 2026 meeting in Chicago, focuses on making µBites more appetising and nutritionally useful. The team engineered baker’s yeast, Saccharomyces cerevisiae, to produce vanilla-flavoured vanillin from plant-derived ferulic acid, while an adapted strain of Rhodosporidium toruloides used ethylene glycol derived from PET to produce beta-carotene, a precursor to vitamin A.
These ingredients were combined with yeast-derived protein, fibre, starch and sweeteners before being extruded through a 3D printer and cooked in a microwave. It’s far removed from grinding up plastic and tipping the fragments into a bakery mixer, but it’s equally far removed from a commercially viable biscuit line. The system currently involves 33 steps, takes between one and two days and costs around $60 to produce one kilogramme of cookies.
Microplastics out, plastic-derived ingredients in?
The food industry is scrambling to keep microplastics out of its products, yet µBites asks consumers to embrace plastic waste as the starting point for their next snack.
While microplastic contamination occurs when persistent particles enter food through water, soil, air, packaging, ingredients or processing equipment, the SIU Carbondale system is designed to dismantle PET and feed its carbon-rich components to microbes, which use them to construct new biological molecules.
Origin and final composition aren’t necessarily the same thing. Sugar fermented into alcohol is no longer sugar, while proteins produced through precision fermentation don’t contain the microorganisms used to make them. The principle behind µBites is that PET should similarly disappear during processing, leaving proteins, fats, vitamins and flavour compounds rather than plastic particles.
The chemistry may stack up, but plastic carries more emotional baggage than most food ingredients. Consumers associate it with ocean pollution, landfill, chemical migration and invisible particles entering their bodies. Convincing them that accidentally ingesting microplastics is fundamentally different from intentionally eating protein made from plastic-derived carbon will take considerably more than a clever name.
Safety evidence will determine whether µBites remain a provocative experiment or become a legitimate ingredient platform. Dr Jayakody says the products have been analysed internally and by accredited third-party laboratories for toxic chemicals, heavy metals, allergens and food pathogens. Simulated digestion studies and further testing are also being conducted ahead of human trials.
Those trials haven’t yet been approved, meaning the researchers themselves haven’t tasted the cookies. Blind sensory assessments reportedly found the aroma appealing and the texture pleasant to touch, while Jayakody says the product smells like a real cookie.
Smell and texture can’t establish safety, digestibility or nutritional value. The latest findings were presented as a conference poster rather than a completed human safety study, so considerable testing remains before the cookies can be considered ready for consumers. Regulators would need evidence covering purity, nutritional composition, toxicology, allergenicity and process consistency before any bakery or snack manufacturer could contemplate using the ingredients.
Consumer acceptance presents another obstacle because the proposition combines several technologies that already generate suspicion: genetically engineered microorganisms, precision fermentation, 3D-printed food and intensive processing. Deriving the ingredients from plastic waste makes that challenge even greater.
A solution to waste or a $60 distraction?
The SIU Carbondale system currently converts more than 50% of the carbon in its waste feedstock into food. Researchers hope to recirculate the remainder and eventually approach 100% conversion, although around 10% could still be released as gas or remain unused.
Even dramatic efficiency gains wouldn’t make µBites a realistic solution to the plastic crisis. More than 400 million tonnes of plastic waste are generated annually, and no plausible market for cookies, milk alternatives, meat substitutes or animal feed could consume it all.
“You’re not going to turn it all into cookies,” Professor Jason Hallett of Imperial College London told New Scientist. “It’s not a solution to the plastic-waste crisis. There’s no way you could do this commercially. We’re not gonna be eating plastic cookies.”
His criticism exposes the danger of presenting a niche space technology as an answer to two global problems. Plastic pollution won’t be solved by finding increasingly elaborate uses for waste while production continues to rise, and food insecurity won’t disappear because scientists can print a vanilla-scented protein disc from a drinks bottle.
The concept becomes more credible in the extreme environments for which it was developed. Spacecraft, submarines, polar research stations and disaster-response operations all face constraints around storage, resupply and waste disposal. A portable system capable of turning carefully controlled waste streams into protein, vitamins and flavourings could provide an emergency or supplementary food source.
Cookies are also only the proof of concept. Dr Jayakody says the ingredients could be used in milk alternatives, meat alternatives and other products requiring protein, vitamins and flavour compounds, while the researchers are exploring their potential in animal feed.
Food production already depends on transformation, with microbes turning milk into yoghurt, sugars into alcohol and dough into bread. µBites take that logic into more contentious territory by asking consumers to judge an ingredient entirely by what it becomes rather than where it came from.
The SIU Carbondale researchers may eventually prove that their cookies are safe, nutritious and useful where conventional food production is impossible. The harder challenge, however, will be persuading people who are already alarmed about accidentally consuming microplastics that a cookie derived from a plastic bottle isn’t plastic at all.
Studies:
Jayasekara S, Hou I, Jayakody L, et al. Engineered yeast consortia for converting plastic and biomass-derived compounds into valuable food additives. ACS Fall 2026 presentation.
Nihart AJ, Garcia MA, El Hayek E, et al. Bioaccumulation of microplastics in decedent human brains. Nat Med31, 1114–1119 (2025). https://doi.org/10.1038/s41591-024-03453-1
