When hydrogen gas interacts with uranium metal, the combination creates a chemically reactive powder and a runaway reaction that is difficult to stop. The result can impact the safety and lifespan of technology critical for fusion energy, hydrogen storage and nuclear fuels.
In a recent study published in npj Materials Degradation, researchers from Lawrence Livermore National Laboratory (LLNL) observed and characterized the beginning stages of hydrogen-uranium corrosion for the first time. The result will lead to more predictive and physically grounded models for how uranium components degrade.
Imagine the hydrogen-uranium interaction like a geyser. Much like surface water seeping through cracks to make its way underground, hydrogen dissolves and diffuses into the uranium metal. This happens silently and invisibly until it becomes too much hydrogen for the uranium to hold. The two materials combine to form a new compound called uranium hydride, which takes up significantly more volume than the original uranium metal.
With that increase in volume comes an increase in pressure. The new hydride expands and is forced upward toward the surface, where it forms a tiny blister. That blister grows until the surface cannot take the strain. Much like a geyser shoots water into the air, the blister eventually bursts open and releases uranium hydride powder.
"Once that protective surface is breached, fresh metal is exposed, and the reaction accelerates," said LLNL scientist Jibril Shittu. "In short: adsorb, dissociate, diffuse, accumulate, blister, rupture, spall. That's the cycle, and once it starts, it's hard to stop."
To design fusion reactors that last, hydrogen storage that works, and nuclear fuels that can confidently remain in storage for decades, researchers need to understand the very beginnings of this hydrogen-uranium geyser. But historically, pinpointing the origin has been a challenge.
"The two workhorse techniques in this field both work beautifully once the reaction is well underway. But both are essentially blind to the very first events," said Shittu. "What was missing was a way to watch the surface itself, at the right length scale and the right time scale, continuously, without disturbing the experiment. That's the gap we set out to fill."
To do so, the team used white-light interferometry, which assembles a tiny topographic map by measuring how light reflects off the uranium surface compared to a reference beam. The method provides the sensitivity needed to watch hydride blisters - which are wide and shallow - without touching or destroying them.
"We can scan the same uranium surface repeatedly through the entire reaction, building a frame-by-frame record," said Shittu. "It's the difference between hearing about an event after the fact and having a security camera rolling the whole time."
From that footage, the team uncovered a few genuine surprises. The hydride blister didn't show up exactly where they predicted, and it spread out sideways as opposed to reaching deep into the uranium metal.
This work was done at a narrow temperature range and a single hydrogen pressure and material state. The next step is to extend it across a wider range of conditions.
"That's how we get from 'we can see it now' to 'we can predict it under any condition you give us,'" said Shittu.
White-light interferometry could also be used to understand hydrogen reactions with other metals, which has implications for the rapidly growing field of hydride superconductors, other corrosion problems, and other degradation processes.
Shittu emphasized that this first-of-its-kind study was made possible by the long-held institutional knowledge at LLNL.
"A lot of what made this study work wasn't in any textbook - it came from the senior scientists at the lab. Sitting down with the old-timers, asking dumb questions and listening hard to the answers is what kept us from re-learning things the field already knew and let us spend our effort on the parts that were genuinely new," said Shittu. "National labs are unusual in that they preserve that kind of institutional memory across generations of researchers, and this study is a direct beneficiary of it."
