Hydrogen has long promised a clean energy future, but separating it from mixed gases with precision and efficiency has remained one of chemistry's stubborn frontiers. Researchers at KAIST in South Korea have now engineered a polymer membrane that filters hydrogen at the angstrom scale, threading a path between the fragility of crystalline materials and the imprecision of conventional polymers. Published in Nature Communications in July 2026, the work introduces not only a new material but a new way of measuring what makes such materials work — a quiet reminder that progress in science often be
KAIST develops hydrogen-selective membrane with molecular network design for clean energy
Stitching polymer chains together like Lego blocks
Mark
Why does hydrogen separation matter so much right now?
Mimi
Because hydrogen is finally being taken seriously as an energy carrier. We can make it from water using renewable electricity. But the hydrogen we produce in industrial processes comes mixed with other gases—nitrogen, carbon dioxide, methane. You need high purity hydrogen for fuel cells and other applications. Right now, separating it is expensive and energy-intensive. A better membrane changes that economics.
Mark
What makes this membrane different from what existed before?
Mimi
The real insight is the Bridge Connectivity Degree. Before, researchers could measure how much crosslinking they had, but not whether that crosslinking actually created useful pores. This team quantified whether the network was truly complete—whether the bridges connected end-to-end to form actual pathways. That number, 73 percent, directly predicts separation performance. It's a way to bring the precision of crystalline materials to the flexibility of polymers.
Mark
The membrane lasted 100 hours without degrading. Is that long enough for industry?
Mimi
It's a proof of concept. Industrial membranes need to run for months or years. But the fact that it showed zero performance loss over 100 hours is encouraging. Combined with the tensile strength being twice that of previous materials, it suggests the design is robust enough to handle real operating conditions—pressure, temperature changes, chemical exposure.
Mark
What's the next hurdle?
Mimi
Scale and cost. Making a laboratory membrane that works is one thing. Manufacturing it reliably in large sheets, at competitive prices, is another. They also need to test it in actual hydrogen production facilities, not just in controlled lab conditions. But the fundamental science is sound, and the metrics they developed give other researchers a clear target to aim for.
El Pulso
- Hydrogen's promise as a zero-emission fuel has been bottlenecked for decades by the sheer difficulty of extracting it cleanly and economically from mixed gas streams.
- Existing separation materials force an uncomfortable trade-off: crystalline porous materials offer precision but shatter at scale, while polymer membranes are durable but nearly impossible to tune at the molecular level.
- The KAIST team broke this impasse by crosslinking polymer chains with chemical bridges and introducing a new metric — Bridge Connectivity Degree — that measures whether those bridges actually form complete, functional transport pathways.
- Their best membrane achieved 73% bridge connectivity, outperforming predecessors in both hydrogen permeability and selectivity, with ultramicropores confirmed small enough to exclude carbon dioxide molecules entirely.
- After 100 hours of continuous operation without performance loss and tensile strength double that of comparable polymers, the material is signaling it may be ready for the long, demanding road toward industrial deployment.
Hydrogen has long promised a clean energy future, but separating it from mixed gases with precision and efficiency has remained one of chemistry's stubborn frontiers. Researchers at KAIST in South Korea have now engineered a polymer membrane that filters hydrogen at the angstrom scale, threading a path between the fragility of crystalline materials and the imprecision of conventional polymers. Published in Nature Communications in July 2026, the work introduces not only a new material but a new way of measuring what makes such materials work — a quiet reminder that progress in science often begins with asking better questions.
Hydrogen burns to produce only water — no carbon dioxide, no particulates. For decades, scientists have recognized it as an ideal fuel for a decarbonized world. The obstacle has always been extraction: hydrogen rarely exists alone in nature, and separating it from mixed gases with high purity has proven stubbornly difficult. A team led by Professor Tae-Hyun Bae at KAIST, South Korea's premier science and engineering university, believes they have found a meaningful way forward.
The researchers engineered a polymer membrane that filters hydrogen from mixed gases at the angstrom scale — a unit so small that one angstrom is roughly one-millionth the thickness of a human hair. The challenge was familiar: crystalline porous materials like metal-organic frameworks offer exquisite control over pore size but are fragile and difficult to manufacture without defects. Polymer membranes are robust and scalable, but controlling exactly where pores form inside them has remained nearly impossible. The KAIST team set out to combine the strengths of both.
They did so by stitching polymer chains together with crosslinkers — chemical bridges connecting one chain to another. The deeper innovation was measurement. Previous researchers described crosslinking density, but that metric said nothing about whether connections actually created useful separation pathways. The team introduced the Bridge Connectivity Degree, or BCD, which measures the proportion of crosslinkers connected at both ends, forming complete pathways through the network. Their best membrane achieved a BCD of 73 percent, yielding significantly higher hydrogen permeability and selectivity than its predecessors, with ultramicropores confirmed small enough to exclude carbon dioxide entirely.
Durability proved equally encouraging. The membrane ran continuously for 100 hours without performance degradation and demonstrated roughly twice the tensile strength of other high-performance polymer membranes — suggesting it could survive the stresses of real industrial environments. The research was published in Nature Communications in July 2026. What remains is the long journey from laboratory to deployment: scaling the membrane, testing it in actual hydrogen production facilities, and proving it can compete economically with existing technologies.
Hydrogen burns clean. When it combusts, it produces only water—no carbon dioxide, no particulates, no greenhouse gases. For decades, scientists have known this makes it an ideal fuel for a decarbonized world. The catch has always been extraction. Hydrogen rarely exists alone in nature. It comes mixed with other gases, and pulling it out with high purity has remained stubbornly difficult. A team at KAIST, South Korea's premier science and engineering university, believes they have found a way forward.
Professor Tae-Hyun Bae and his colleagues in the Department of Chemical and Biomolecular Engineering announced in August that they had engineered a polymer membrane capable of filtering hydrogen from mixed gases with unprecedented efficiency. The membrane works at the scale of angstroms—units so small that one angstrom is roughly one-millionth the thickness of a human hair. At that scale, the researchers designed a network of transport pathways that allow hydrogen molecules through while blocking larger molecules like nitrogen and carbon dioxide.
The challenge the team faced was not new. Scientists have long known that crystalline porous materials like metal-organic frameworks offer precise control over pore size and shape. But these materials are fragile and difficult to manufacture at scale without defects. Polymer membranes, by contrast, can be produced in large sheets and are mechanically robust. The problem is that controlling exactly where and how pores form inside a polymer has proven nearly impossible. The KAIST researchers set out to combine the best of both worlds.
They did this by stitching polymer chains together using crosslinkers—chemical bridges that link one chain to another, much like connecting Lego blocks. The innovation was not just the structure itself but how they measured it. Previous researchers had used metrics like the degree of crosslinking to describe how thoroughly the polymer network was connected. But these measurements did not tell you whether the connections actually created useful pores for separation. The KAIST team introduced a new metric called the Bridge Connectivity Degree, or BCD. It measures the proportion of crosslinkers that are connected at both ends, forming complete pathways through the network. In their best membrane, called ms-oDMB-DB50, they achieved a BCD of 73 percent.
The results were striking. The new membrane showed significantly higher hydrogen permeability than its predecessor while also becoming more selective—better at letting hydrogen through while blocking other gases. Analysis revealed the membrane contained numerous ultramicropores, pores smaller than three angstroms, too small for carbon dioxide molecules to enter. To verify these pores actually existed, the researchers used a clever technique: they introduced helium atoms, which are smaller than hydrogen, and tracked whether they could penetrate the membrane. They could, confirming the pores were real.
Durability matters in industrial applications. The team ran the membrane continuously for 100 hours without any degradation in performance. They also measured its tensile strength—the force it could withstand before tearing—and found it was roughly twice as strong as other high-performance polymer membranes reported in the literature. This suggests the material could survive the stresses of actual industrial hydrogen separation plants.
Dr. Hongju Lee, the first author of the study, noted that while others had tried to merge the advantages of crystalline and polymer materials, this work was different because it quantified network completeness and directly connected that number to separation performance. Professor Bae described the approach more simply: by stitching polymer chains together with crosslinkers that fit like Lego blocks, they created a selective filter for hydrogen. The research was published in Nature Communications in July 2026, funded by South Korea's Ministry of Science and ICT. What comes next is the long road from laboratory success to industrial deployment—scaling the membrane, testing it in real hydrogen production facilities, and proving it can compete economically with existing separation technologies.
Citas Notables
By stitching polymer chains together with crosslinkers that fit together like Lego blocks, we formed a network inside the membrane that selectively allows only small hydrogen gas molecules to pass through.— Professor Tae-Hyun Bae, KAIST
This study is different in that it defines 'how completely the network is connected' as a quantitative value and directly links that value to separation performance.— Dr. Hongju Lee, Korea Institute of Science and Technology
Quieres la nota completa?
Lee el original en eurekalert.org ↗
Politics & Governance