Switch to Low Carbon Version

Malaysian Researchers Develop AI-optimised Palm Ash for Scalable CO₂ Capture

March 20, 2026
by CSN Staff

Researchers in Penang have transformed oil palm ash into an efficient, AI-optimised material for CO₂ capture, offering a sustainable solution to palm industry waste and advancing climate mitigation strategies.

, Malaysian researchers have pioneered a method to transform oil palm ash, a byproduct of the palm oil industry, into a highly efficient carbon dioxide (CO₂) capture material. This innovation not only provides a sustainable solution to the disposal challenges posed by palm kernel shell ash but also presents a promising tool in the fight against climate change.

The team at Universiti Sains Malaysia employed a process involving acid treatment, heating at 700 degrees Celsius, and activation with potassium hydroxide, which produced a mesoporous carbon structure capable of adsorbing 2.9 millimoles of CO₂ per gram. Remarkably, this capacity matches that of advanced commercial adsorbents engineered with much larger surface areas, underscoring the significance of the material’s finely tuned pore architecture rather than sheer surface size.

The secret to this material’s efficiency lies in its molecular design: with an average pore diameter of 72.71 angstroms, the structure creates optimally sized channels that allow CO₂ molecules to enter and adhere effectively. This mesoporous framework facilitates faster diffusion and enhanced adsorption kinetics, crucial for industrial-scale applications. While surface area is often highlighted as paramount in adsorbents, this research highlights how pore size and arrangement, along with material functionalization, can be equally decisive in performance. The optimal activation ratio was found to be 1:2 for ash and potassium hydroxide; exceeding this ratio led to structural collapse, reducing efficacy.

Significantly, the Malaysian team integrated machine learning into their experimental approach, representing a novel fusion of artificial intelligence and materials science. Employing a bilayered neural network, the researchers achieved an R-squared value exceeding 0.99 in predicting CO₂ adsorption under varying conditions such as temperature, gas flow rate, and concentration. This computational model enables rapid simulation and optimisation, slashing the time and costs associated with traditional experimental testing. The thermodynamics revealed that CO₂ adsorption primarily occurs via physisorption—weak physical forces effective at lower temperatures around 30 degrees Celsius—but diminishes sharply at 60 degrees Celsius due to increased molecular energy allowing CO₂ to escape.

Malaysia’s palm oil industry produces more than 20 million tons annually, generating significant amounts of ash that are difficult to dispose of sustainably. By upcycling this waste into a carbon-capture material requiring less energy than synthesising commercial adsorbents, the research offers a closed-loop approach that combines waste valorisation with climate mitigation. Experimental tests showed that a gas flow rate of 30 millilitres per minute maximised contact time, enhancing CO₂ capture while higher flow rates led to premature breakthrough and loss of efficiency. Optimal CO₂ concentrations reached up to 15%, with unexpected adsorption dynamics observed at 12%, suggesting complex interactions between adsorption rates and gas flow patterns.

This pioneering work positions Universiti Sains Malaysia’s School of Chemical Engineering as a regional leader in sustainable materials development, particularly emphasizing AI-driven innovations. The researchers have ambitions to conduct pilot-scale experiments applying their material in industrial settings such as flue gas treatment, biogas upgrading, and direct air capture systems. Long-term durability studies will also determine if the material can be regenerated multiple times without significant loss of efficiency, a critical factor for commercial viability.

Although other biomass-derived materials like activated carbon from palm kernel shells may boast higher surface areas—sometimes exceeding 1000 m²/g—and different functionalizations such as magnesium oxide additions, the Malaysian oil palm ash material competes strongly in terms of carbon capture capacity. Comparative studies indicate surface area alone does not guarantee superior adsorption, reinforcing the importance of pore architecture and surface chemistry in adsorbent design.

Broader implications of this research connect to regional initiatives exploring bioenergy with carbon capture and storage (BECCS) using various palm oil waste streams, including fronds, trunks, and empty fruit bunches. These integrated systems offer sustainable pathways to produce electricity with reduced carbon footprints while valorising agricultural residues. The Malaysian innovation complements these approaches by focusing specifically on ash valorisation and direct CO₂ capture.

As global climate targets become more stringent, demand escalates for affordable, scalable carbon capture technologies. The Malaysian development addresses these needs by harnessing an abundant waste material to create a cost-effective, high-performance adsorbent, utilising machine learning to optimise and accelerate research progress. While challenges remain—such as proving long-term material stability and scaling production—the work signals a significant step towards circular economy solutions that merge agriculture, advanced materials science, and artificial intelligence in tackling climate change.

Next time palm oil products appear on shelves, it is worth remembering that from the industry’s residue streams, scientists are crafting innovative tools to actively remove carbon dioxide from the atmosphere, molecule by molecule.