Turning palm waste into a machine-learning-optimized CO2 capture material
Led by Dr. Azam Taufik Mohd Din from USM's School of Chemical Engineering, the team developed a novel adsorbent named OPA-KOH(1:2) by treating palm ash with acid, then carbonizing and activating it using potassium hydroxide. The resulting mesoporous material achieved a CO2 adsorption capacity of 2.9 mmol/g despite its modest surface area of 30.95 m/g - performance comparable to, or better than, many commercial activated carbons.
"This isn't just recycling - it's upcycling at the molecular level," said Dr. Mohd Din. "We're taking a waste product that often ends up in landfills and turning it into a high-performance tool for carbon capture."
The secret lies in its optimized pore structure, with an average pore size of 72.71 A, enabling CO2 molecules to flow in and bind efficiently. Analysis showed that the adsorption process is exothermic and spontaneous, dominated by physical adsorption with a smaller chemisorption component - enhancing both effectiveness and durability under real-world operating conditions.
To further refine the process, the researchers applied machine learning algorithms to model and predict CO2 adsorption behavior. A bilayered neural network achieved an R value above 0.99, showing exceptional predictive power. "This shows ML isn't just a trend - it's becoming essential," noted Dr. Mohd Din. "It allows us to simulate, optimize, and understand adsorption processes faster and more deeply than ever before."
The findings demonstrate how Malaysia's vast palm oil residues - over 20 million tons produced annually - can be repurposed into a circular, value-added product for carbon management. The resulting adsorbent is both affordable and scalable, offering a practical route toward industrial CO2 capture systems such as flue gas treatment and biogas upgrading.
Future studies will explore regeneration cycles, material stability, and pilot-scale testing. As global climate goals demand rapid innovation, the Malaysian team's fusion of waste materials science and AI points toward a new era of green, intelligent engineering.
Research Report:Enhanced CO2 capture using KOH-functionalized oil palm ash adsorbent: experimental and applied machine learning approach
Related Links
Shenyang Agricultural University
Carbon Worlds - where graphite, diamond, amorphous, fullerenes meet
