Graphic: Each cobalt metal site (purple) in CoMe-MFU-4l can grab and hold two carbon monoxide molecules (red and gray atoms). When the gas is removed, the site returns to its original state and can capture CO again.
Most materials designed to capture gases work like a crowded parking lot: one car per space. But chemists at UC Berkeley have now created a porous material that breaks this rule, fitting two gas molecules into a single binding site—a feat that could revolutionize the separation or purification of critically important industrial gases.
The research, published in Science, describes a metal–organic framework (MOF)—a class of porous crystalline materials—that can selectively grab carbon monoxide (CO) from gas mixtures with record-breaking efficiency and capacity. What makes this science groundbreaking isn't just how much CO the MOF captures, but how the MOF goes about the capture: after the first CO molecule attaches, the binding site undergoes chemical changes that open the door for the second CO to bind even more strongly than the first does.
"We've been exploring different mechanisms for cooperative gas binding in metal–organic frameworks for years," said Jeffrey Long, the C. Judson King Professor of Chemistry and Chemical Engineering at UC Berkeley and the corresponding author on the study. "What's remarkable here is achieving cooperativity through chemical reactivity at individual metal sites, rather than requiring the entire framework to reorganize."
Metal–organic frameworks recently gained widespread recognition when their pioneers, including Omar Yaghi of UC Berkeley, won the 2025 Nobel Prize in Chemistry for developing these highly porous materials. While that work focused on the fundamental discovery and synthesis of MOFs with record-breaking surface areas, Long's research demonstrates that the materials can be used for highly efficient gas purification.
Breaking the one-molecule rule
Traditional gas-capture materials work through simple adhesion; individual gas molecules stick to individual binding sites in a material. Among thousands of MOFs studied for gas capture, only a few have ever been shown to bind more than one molecule per metal site, and those required extremely low temperatures or have other limitations.
The material investigated by the Berkeley team, called CoMe-MFU-4l and initially discovered by Dirk Volkmer of the University of Augsburg in Germany, takes a fundamentally different approach. The MOF contains cobalt metal centers that can undergo dramatic chemical changes. When the first CO molecule binds, the cobalt ion undergoes a chemical change known as a "spin transition," flipping from a highly-magnetic state to a less-magnetic state.
"There's essentially a hidden second site that is only accessible once the first site is occupied," explained Kurtis Carsch, who led the research as an Arnold O. Beckman Postdoctoral Fellow in Long's lab and is now an assistant professor of chemistry at The University of Texas at Austin, where his research program addresses the exploration of porous materials for catalysis and separations.
This electronic change opens the second binding site. Then, when a second CO molecule arrives, the CO not only attaches, but triggers a chemical reaction that binds more tightly than the first molecule. Both CO molecules end up locked in place through actual chemical bonds, not just physical attraction. The new binding paradigm, coined by Carsch and Long as "single-site cooperativity," proceeds where each binding site operates independently, but the first CO molecule makes it easier for the second to bind, similar to how hemoglobin cooperatively takes up oxygen in blood.
"The moment we saw the infrared spectrum showing that acetyl peak, we knew we had something special," said Carsch. "This wasn't just physical adsorption; real chemistry was happening. The cobalt–methyl bond was reversibly breaking and reforming to accommodate both CO molecules."
In tests at room temperature, CoMe-MFU-4l captured 5.2 millimoles of CO per gram at just 10 millibar pressure—the highest capacity reported for any MOF under these conditions. At atmospheric pressure, it plateaued near 6.1 mmol/g, close to the theoretical maximum for binding two CO molecules per cobalt site.
The material was also highly selective, choosing CO over other gases including ethylene, a similar-sized molecule that many materials would struggle to distinguish in the presence of CO. In experiments simulating real-world conditions known as breakthrough measurements, the material captured nearly two CO molecules per site even from a mixture of 10% CO in ethylene. Importantly, the process is fully reversible, so the material can release captured CO and be used again.
A new design strategy
The work opens the door to developing other MOFs that can separate and purify chemicals more efficiently. Gas separations currently consume about 15% of global industrial energy and MOFs, with their extremely high surface areas, could offer more energy-efficient ways of separating gases—especially if other MOFs can break the one-molecule-per-site rule like CoMe-MFU-4l.
"We're thinking about ways that we can capitalize on using materials to be able to do these purifications with minimum energy expenditure," Carsch said.
However, challenges remain. The research team said the material is sensitive to trace oxygen and water, which could limit its use under real-world conditions.
"I would put our finding at a very low technology readiness level," Carsch said. "Upfront, our interest was doing the very fundamental science of just: can we merge these two disparate fields of absorption science and organometallic chemistry?"
Looking forward, technologies are already under development for other MOFs that help work around oxygen and water sensitivities, and the Long group at Berkeley has a collaboration with Baker Hughes aimed at bridging the gap between academic discovery and industrial application. Eventually, they said, CoMe-MFU-4l could find applications in purifying industrial gas streams, removing trace CO from hydrogen for fuel cells, or as a final purification step in its production.
In addition to other research groups at UC Berkeley, this research was conducted in collaboration with individuals at the National Institute of Standards and Technology (NIST), Montana State University, Lawrence Berkeley National Laboratory, University of Delaware, and the Miller Institute for Basic Research in Science. The authors thank the Department of Energy (DoE) for financial support of this work.