Scientists may have finally found a solution to tackle the problem of radioactive iodine, one of nuclear energy’s most stubborn threats, thanks to AI.
A research team in South Korea has used artificial intelligence to discover a powerful new material that can trap radioactive iodine, specifically isotope I-129, one of the most persistent and dangerous byproducts of nuclear energy that poses serious environmental and health risks.
With a half-life of 15.7 million years, I-129 is highly mobile in the environment and notoriously difficult to remove from contaminated water.
Developed by researchers from Korea Advanced Institute of Science and Technology (KAIST) in collaboration with the Korea Research Institute of Chemical Technology (KRICT), the breakthrough material belongs to a class called Layered Double Hydroxides (LDHs).
These compounds are known for their structural flexibility and ability to trap negatively charged particles like iodate (IO₃⁻), the form radioactive iodine most often takes in aqueous environments.
AI narrows down the options
Instead of testing thousands of LDH combinations manually, which would be difficult to search through conventional trial-and-error experiments, the team turned to machine learning to identify optimal iodate adsorbents.
Starting with experimental data from 24 binary and 96 ternary compositions, they trained a model to predict the most promising candidates from a vast pool of metal combinations.
The team focused on the fact that LDHs, like high-entropy materials, can incorporate a wide range of metal compositions and possess structures favorable for anion adsorption.
The AI model guided the researchers to a quinary compound made of copper, chromium, iron, and aluminum, named dubbed Cu₃(CrFeAl).
This material showed over 90 percent efficiency in removing iodate from water, outperforming traditional silver-based absorbents, which often fail to trap iodate effectively.
Small sample, big leap
Remarkably, the team only needed to test about 16 percent of all possible material combinations to find the optimal one, demonstrating the power of AI in reducing both time and cost in nuclear environmental research.
“This study shows the potential of using artificial intelligence to efficiently identify radioactive decontamination materials from a vast pool of new material candidates,” said KAIST professor Ho Jin Ryu.
“It is expected to accelerate research for developing new materials for nuclear environmental cleanup.”
The research team has filed a domestic patent application for the developed powder technology and is currently proceeding with an international patent application. They are also working to improve the material’s stability under real-world conditions.
The team is now looking for academic and industrial partnerships to develop iodine-absorbing powders and water filters that can be used in contaminated nuclear sites to trap radioactive iodine.
The study was led by Professor Ho Jin Ryu from the Department of Nuclear and Quantum Engineering at KAIST, in collaboration with Dr. Juhwan Noh of the Digital Chemistry Research Center at KRICT.
Dr. Sujeong Lee, a graduate of KAIST’s Department of Materials Science and Engineering, and Dr. Noh were listed as co-first authors on the paper.
