A collaborative research effort between UNIST and the Korea Institute of Science and Technology (KIST) has successfully synthesized three novel porous materials using a data-driven structure prediction algorithm. These new materials, classified as zeolitic imidazolate frameworks (ZIFs), exhibit exceptional selectivity in greenhouse gas separation, particularly for carbon dioxide (CO₂). Professor Wonyoung Choe of the Department of Chemistry at UNIST, along with Professor Hyunchul Oh and Dr. Jung-Hoon Lee from KIST, spearheaded the project. The team's groundbreaking work, which includes the synthesis of UZIF-31, UZIF-32, and UZIF-33, was published in the March 2025 issue of JACS Au and even graced the journal's cover. Metal-organic frameworks (MOFs) are crystalline materials constructed from metal ions and organic ligands, forming highly porous structures. ZIFs, a specific type of MOF, are renowned for their chemical stability and flexible pore architectures, making them excellent candidates for applications in catalysis, gas storage, and separation. However, despite the theoretical potential for millions of ZIF structures, only around 50 have been synthesized since their discovery in 2006, a bottleneck known as the "zeolite conundrum." To overcome this limitation, the researchers developed an innovative algorithm that combines chemical intuition with detailed structural analysis. This algorithm evaluates critical parameters such as bond angles, ring connectivity, and network regularity. When applied to a virtual dataset of over 4.45 million candidate structures, the algorithm narrowed down the options to 420 promising structures, from which 90 were identified as top-tier based on their energy stability. Experimental verification of these top-tier candidates led to the successful synthesis of three ZIFs, each demonstrating remarkable gas separation capabilities. Most notably, UZIF-33 showed more than tenfold selectivity for CO₂ over methane, highlighting its significant potential for greenhouse gas separation and purification applications. "This study highlights the effectiveness of digital prediction in achieving real-world experimental success," Professor Choe stated. "By merging our algorithm with automated synthesis technologies, we can greatly expedite the development of advanced ZIF materials with precisely tailored properties." These findings not only advance the field of materials science but also offer a promising avenue for addressing environmental challenges related to greenhouse gas emissions. The ability to efficiently separate CO₂ from other gases could play a crucial role in industries aiming to reduce their carbon footprint.