How Plastic-Eating Enzymes Crossed the Threshold to Industrial Viability
Recent breakthroughs in machine learning and chemical engineering have slashed the cost of enzymatic plastic recycling, making it cheaper than producing virgin fossil-fuel plastics for the first time.
By Ishani Patel
- Biochemical Engineers
- Focus on using machine learning and directed evolution to increase the catalytic speed and thermal stability of natural enzymes.
- Industrial Process Designers
- Prioritize techno-economic viability, focusing on bioreactor scaling, energy efficiency, and reducing expensive chemical inputs.
- Petrochemical & Packaging Industries
- Evaluate the technology based on its ability to achieve cost parity with fossil fuels and deliver virgin-quality clear plastic from mixed waste.
Humanity produces over 322 million metric tons of plastic annually, and the average person disposes of roughly 115 pounds of it each year. A significant portion of this burden is polyethylene terephthalate (PET), a highly versatile polymer used in single-use packaging, soda bottles, and synthetic textiles, which alone accounts for 12% of all solid waste on Earth.[2]
For decades, the primary strategy for managing this waste has been mechanical recycling. However, mechanical processes are highly sensitive to contamination and struggle to process mixed-color plastics, thermoforms, and textile fibers. The resulting recycled plastic is typically of lower quality than virgin material, meaning it is often downcycled into carpets or park benches before inevitably ending up in a landfill.
The search for a biological alternative gained global momentum following the discovery of Ideonella sakaiensis, a bacterium found in a Japanese recycling center that had naturally evolved the ability to use PET plastic as its primary carbon and energy source. This organism provided the first biological blueprint for deconstructing synthetic polymers.[3]
Researchers soon mapped the bacterium's mechanism, revealing a synergistic two-enzyme system. The first enzyme, PETase, attacks the polymer surface and liberates soluble intermediate compounds. A second enzyme, MHETase, then cleaves those intermediates into the plastic's foundational monomers: terephthalic acid (TPA) and ethylene glycol (EG).[3]
While groundbreaking, these natural enzymes were poorly suited for industrial applications. Because they evolved to operate in ambient environmental conditions, they worked far too slowly and lost their structural integrity when exposed to the elevated temperatures and varying pH levels required for commercial-scale chemical processing.[2]
The biological bottleneck was broken by researchers at the University of Texas at Austin, who utilized a structure-based machine learning algorithm to engineer a highly resilient variant known as FAST-PETase (Functional, Active, Stable, and Tolerant PETase). By predicting which amino acid mutations would stabilize the protein, the AI-guided design yielded an enzyme capable of surviving harsh industrial conditions.[1][2]
The performance of FAST-PETase represented a paradigm shift. In laboratory testing, the engineered enzyme completely degraded 51 different post-consumer PET waste samples—including mixed-color packaging and polyester fabrics—in as little as 24 hours at a sustained temperature of 50 degrees Celsius.[1][2]
The performance of FAST-PETase represented a paradigm shift.
Crucially, this enzymatic depolymerization allows for a true closed-loop lifecycle. The recovered TPA and EG monomers can be chemically repolymerized into new PET with a very high yield and purity. Because the enzymes selectively target the polymer backbone, colorants and impurities are left behind, allowing mixed-color waste to be recycled into perfectly clear, virgin-quality plastic without intensive pre-sorting.[2]
Despite these biological advances, scaling the technology from laboratory bioreactors to industrial facilities faced a severe economic hurdle. The depolymerization process traditionally required massive quantities of sodium hydroxide to manage the pH of the reaction, driving up operating costs and generating chemical waste that negated many of the environmental benefits.
A recent breakthrough in process engineering has now cleared this final economic barrier. A consortium of researchers from the National Renewable Energy Laboratory (NREL), the University of Massachusetts Lowell, and the University of Portsmouth published a comprehensive blueprint in Nature Chemical Engineering that fundamentally redesigned the chemical environment of the bioreactor.
By replacing the traditional sodium hydroxide base with ammonium hydroxide, the engineering team unlocked cascading efficiencies. This subtle chemical switch reduced the need for expensive acid and base additions by over 99%, while simultaneously streamlining the recovery of the purified monomers.
The resulting techno-economic analysis demonstrates that this optimized process cuts greenhouse gas emissions by nearly half, reduces energy use by 65%, and slashes annual operating costs by 74% compared to previous enzymatic techniques.
For the first time, the modeled cost of enzyme-recycled PET has fallen below the cost of domestic virgin PET produced from fossil fuels. The NREL study projects the cost of the enzymatically recovered plastic at $1.51 per kilogram, undercutting the $1.87 per kilogram market rate for virgin materials, creating a powerful financial incentive for industry adoption.
However, transparent uncertainties remain as the technology transitions from pilot plants to full commercial deployment. Highly crystalline PET, such as the rigid plastic used in standard water bottles, remains stubbornly resistant to enzymatic attack and must be thermally melted before the enzymes can efficiently degrade it, adding an energy penalty to the process.[2][4]
If these industrial scaling efforts succeed, the implications for global energy and waste are profound. According to NREL, the plastics currently landfilled in the United States contain enough embodied energy to supply 5% of the power needs of the entire U.S. transportation sector—energy that biocatalysis can now economically recapture.
Key points
- PET plastic accounts for 12% of all global solid waste, and traditional recycling degrades its quality.
- Machine learning was used to engineer FAST-PETase, an enzyme that degrades plastic in 24 hours at 50°C.
- A recent process innovation replaced sodium hydroxide with ammonium hydroxide, cutting operating costs by 74%.
- The modeled cost of enzyme-recycled PET is now $1.51/kg, cheaper than the $1.87/kg cost of virgin plastic.
- The recovered molecular building blocks can be used to create perfectly clear, virgin-quality plastic from mixed waste.
What we don’t know
- How efficiently the enzymes can be recovered and reused across multiple depolymerization cycles in a massive industrial vat.
- Whether the logistics of collecting, washing, and transporting municipal plastic waste will offset the economic gains of the bioreactor process.
- How quickly the technology can be adapted to degrade other prevalent plastics, such as polyolefins or polystyrene.
Key terms
- PET (Polyethylene terephthalate)
- The most common thermoplastic polymer, widely used in clothing fibers, liquid containers, and single-use packaging.
- PETase
- An enzyme that catalyzes the hydrolysis of PET plastic, acting as molecular scissors to cut the long polymer chains into smaller segments.
- Monomer
- A basic molecular building block that can bind chemically to other molecules to form a long chain, or polymer.
- Biocatalysis
- The use of natural substances, such as enzymes or microbes, to speed up chemical reactions in industrial processes.
- Repolymerization
- The chemical process of taking recovered monomers and stitching them back together to create a new, virgin-quality plastic polymer.
Sources
[1]NatureBiochemical EngineersMachine learning-aided engineering of hydrolases for PET depolymerization
Read on Nature →
[2]Argonne National LaboratoryPetrochemical & Packaging IndustriesPlastic-eating enzyme could eliminate billions of tons of landfill waste
Read on Argonne National Laboratory →
[3]Proceedings of the National Academy of Sciences (PNAS)Biochemical EngineersCharacterization and engineering of a two-enzyme system for plastics depolymerization
Read on Proceedings of the National Academy of Sciences (PNAS) →
[4]Factlen Editorial TeamPetrochemical & Packaging IndustriesSynthesis by Factlen editorial team
Read on Factlen Editorial Team →
Comments
Every angle. Every day.
Get science stories with full source coverage and perspective breakdowns delivered to your inbox.

