Researchers 3D-Print Vanilla Cookies From Plastic Bottles. No One Has Tasted Them.

Southern Illinois University Carbondale researchers presented 3D-printed vanilla cookies made from plastic bottles and corn waste at the American Chemical Society fall meeting on August 24. The snacks, called µBites, are produced by breaking polyethylene terephthalate and agricultural residue into carbon fragments, feeding those fragments to genetically engineered yeast, then extruding the resulting slurry through a printer. Lead researcher Lahiru Jayakody described a “pleasant, appealing aroma.” No one has eaten the cookies. The team is still waiting for university approval to conduct a taste test.

The process begins with oxidative hydrothermal dissolution: heat, water, pressure, and oxygen applied in a multi-step sequence that cracks PET bottles and discarded corn stalks into molecules microbes can digest. CRISPR-modified yeast strains then convert those fragments into proteins, fats, and flavor compounds. One engineered baker’s yeast produces vanillin. Another strain generates beta-carotene for color and vitamin A. Fiber, starch, and sweetener are added before 3D printing.

The project started inside NASA’s Deep Space Food Challenge as a way to turn waste carbon into calories on long missions. Jayakody has framed the same logic for Earth: plastic is carbon, food is carbon. Production currently costs about $60 per kilogram. The team hopes efficiency gains will bring that down. Jayakody has also said it could take two decades of additional research before plastic pulled from oceans or landfills becomes safe, palatable food.

Claire Robinson of GMWatch objected to the absence of rigorous safety testing. Genetically engineered yeast and bacteria, she noted, can unexpectedly produce toxins or allergens. She cited the 1989 Showa Denko L-tryptophan case, in which a genetically engineered production process was linked to 37 deaths and more than 1,500 illnesses in the United States. Robinson said she would not “dignify the product by expecting lab animals to eat it, let alone humans.”

The researchers say laboratory data indicate the cookies are safe to eat. That claim has not been tested in a public human trial. A patent application is pending on the production system. Until independent toxicology and allergen work is complete, the product remains a conference demonstration, not a food.

The team argues µBites could address plastic pollution — estimated at 3.4 percent of global emissions — and rising food demand projected to increase 35 to 56 percent by 2050. Robinson called those environmental claims highly questionable. Industrial bioreactors consume energy, water, and refined inputs. They do not replace the need to cut plastic production or improve deposit-return recycling. A $60-per-kilogram cookie printed from supermarket bottles does not drain the ocean of PET.

ACS reporting said the group hopes public consumption within a few years and imagines use in disaster zones, submarines, or off-world colonies. Those scenarios assume the safety file closes and the energy ledger works. Neither has been demonstrated at commercial scale.

Microbes can convert waste molecules into edible compounds. That is a laboratory fact. Whether the result belongs on a plate is a different question. Whole-food nutrition does not require a 32-step hydrothermal circuit or a CRISPR yeast consortium. Plastic pollution requires less plastic, not a new snack category. Until someone is allowed to eat the cookie and independent labs publish what is in it, µBites remain a proof of concept — and a reminder that “edible” and “food” are not the same word.