PHAs have different physical properties depending on the production process and the materials used. Courtesy of EuroPlas.
Plastic production reached a staggering 413.8 million metric tons in 2023, nearly 40% of which was used in packaging applications. The rest serves industries like construction, automotive, electronics, and healthcare. However, the environmental and health impacts of plastic pollution are dire. Microplastics contaminate ecosystems, and recycling efforts fall short, leaving many plastics to persist in the environment for centuries.
Bioplastics, particularly PHAs, offer a sustainable alternative to traditional plastics. Microorganisms produce these biodegradable polymers using CO₂ and organic materials, creating a closed-loop, eco-friendly lifecycle. PHAs have a wide range of chemical structures, but finding a proper replacements for conventional plastics has been slow and costly. While high-throughput testing could speed up the process, it requires significant resources.
To solve this problem, researchers are using artificial intelligence. They created multitask deep neural networks (DNNs). These AI tools predict bioplastic properties, study large datasets, and find PHA-based materials that work as well as regular plastics. This AI approach is changing how we design sustainable plastics, making the process faster and more efficient.
You can also read: Mango Materials Transforms Waste Methane into PHA.
To address these challenges, researchers used machine learning to screen a vast chemical space. The process involved:
Researchers trained DNN-based property predictors using almost 23,000 experimental values. These models focused on key properties such as thermal, mechanical, and gas permeability. They learned specific characteristics like glass transition, melting temperature, degradation temperature, young modulus, tensile strength, and elongation.
Three multitask neural networks analyzed the candidate set. Each focused on a specific property: thermal, mechanical, or gas permeability. The models predicted how PHAs and copolymers could replace plastics such as polyethylene, polypropylene, PVC, and polystyrene. They also predicted replacements for PET and polyamide-6.
To ensure that the AI-selected materials were viable for common applications, the researchers followed a two-step approach:
Researchers compared predicted properties of bioplastic candidates with those of common plastics, shortlisting the closest matches.
The team evaluated the synthesizability of candidates, focusing on those with known biosynthetic or chemical synthesis pathways.
The models identified 14 promising PHA-based materials that could replace petroleum-based plastics. All the top candidates had aromatic side-chain groups, which improves their mechanical properties. Studies show that microorganisms like Pseudomonas oleovorans can naturally produce PHAs with these features, making biosynthesis a practical production method.
Blending PHAs with conventional polymers showed great potential. Copolymers of PHAs and traditional plastics were more flexible, stronger, and thermally stable. These hybrid materials provide a smooth transition to biobased alternatives, helping industries adopt sustainable materials without major changes to their processing systems.
Many bacteria and archaea can produce PHAs. Courtesy of Microbiology Society.
PHAs are biodegradable and come from renewable sources, but producing them on a large scale is still difficult. While microbial fermentation is a proven method, it needs to have lower costs and higher scalability. Advances in genetic engineering and metabolic optimization could boost PHA production efficiency, helping it compete with petroleum-based plastics.
Chemical synthesis routes also offer promising solutions:
A synthesis approach similar to PET recycling could integrate lactone-based PHAs into polyesters.
Copolymerization techniques used for polyethylene and polypropylene could enable PHA integration, creating partially biobased materials.
By combining biosynthesis with chemical synthesis, researchers can expand the range of functional bioplastics while maintaining cost-effectiveness.
AI is transforming how we discover materials. It makes the process faster, cost-effective, and efficient. Consequently, with machine learning, scientists can study millions of bioplastics to find the right combination of strength, sustainability, and manufacturability.
This innovation is crucial as global plastic waste rules tighten. Countries are banning single-use plastics and introducing stricter recycling laws. This is driving the need for sustainable, bio-based materials. With AI-powered discovery, industries can adopt eco-friendly solutions without losing performance.
An open data-sharing platform provides information on PFAS alternatives for use in manufacturing.
The market for polymer-based chiller components is expanding rapidly as industries seek cost efficiency, corrosion…
A groundbreaking photosensitive PEEK ink is revolutionizing additive manufacturing and the polymer industry.
Despite current economic challenges, K 2025 remains the center of the plastics world, showcasing innovations…
The diesel effect can occur in molds with insufficient venting, leading to burned parts in…
Kevlar is the ultimate ballistic material, known for its exceptional strength, high tenacity, and lightweight…