Fuel Cells That Clean Themselves Could Finally Run on Dirty Gas

Sulfur is everywhere fuel isn’t supposed to be. Natural gas pipelines carry traces of hydrogen sulfide. Biogas fermented from organic waste contains it. Even carefully processed syngas, the mix of hydrogen and carbon monoxide produced by gasifying coal or biomass, arrives with sulfur along for the ride. For solid oxide fuel cells, which can in principle run on all of these fuels and have long been held up as a cornerstone of flexible clean energy, this is a persistent embarrassment: a device that promises to work on almost anything, stopped in its tracks by a contaminant measured in parts per million.

The problem sits at the atomic scale. Inside a solid oxide fuel cell, the anode is the workhorse, the electrode where fuel arrives and gives up its electrons. Conventional anodes are built around nickel, which is cheap, catalytically active, and excellent at the job. Right up until sulfur appears. Nickel and sulfur have an unfortunate affinity for one another. They bond readily, forming stable nickel-sulfur species that coat the active surface and effectively switch the anode off. Even 1 to 10 parts per million of hydrogen sulfide is enough to start the damage. And those Ni-S bonds, once formed, are stubborn. They don’t break apart easily under the cell’s normal operating conditions. The cell degrades; you either remove the sulfur upstream, which costs energy and money, or you watch performance fall.

Chuancheng Duan and his team at the University of Utah have been trying to think their way out of this problem differently. Rather than treating sulfur removal as a job for engineers upstream of the fuel cell, they wanted to know whether the anode itself could be made to deal with it. Their answer, published in the Journal of the American Chemical Society, turns on an unusual property of the metal rhodium, and on steam.

The key move was adding rhodium to the nickel anode, not as a surface coating but as a genuine component of the material. The synthesis method, called in situ exsolution, incorporates rhodium and nickel ions directly into a samarium-doped ceria lattice. Heat the material under reducing conditions and the metal ions migrate out of the oxide structure, nucleating at the surface as anchored bimetallic nanoparticles, part-nickel, part-rhodium, physically embedded in their support and thus far more resistant to the kind of sintering and migration that plague conventional catalyst surfaces at the temperatures solid oxide fuel cells operate.

What rhodium does to the anode’s relationship with sulfur is, in a word, electronic. The presence of rhodium alters the local chemical environment of nearby nickel atoms, weakening the bond that sulfur would otherwise form with them. Sulfur still arrives at the surface; it still adsorbs. But the Ni-S species that form are looser, more transient, less likely to accumulate into the irreversible blockage that kills a conventional anode. The difference shows up vividly in infrared spectra taken at 650°C while the catalyst is actually operating: the rhodium-nickel system shows weak, flickering Ni-S vibrational bands during hydrogen sulfide exposure, where the nickel-only anode shows strong, persistent signals, the spectroscopic signature of a surface being steadily poisoned.

That’s one half of the mechanism. The other half involves water. When steam is present, something more active happens. Rhodium, it turns out, is unusually good at dissociating water molecules, breaking them apart into reactive hydroxyl groups. Those hydroxyls then go to work on the sulfur that has adsorbed onto the nickel surface, oxidising it into sulphur dioxide. SO2 is volatile; it desorbs from the surface and escapes. The anode, in effect, cleans itself. “We show that catalysts can be engineered not just to tolerate sulfur, but to actively clean themselves during operation,” says Duan.

The numbers make the case. Under 100 parts per million of hydrogen sulfide at 600°C, the rhodium-nickel cell maintained more than three times the power output of a conventional nickel anode and showed substantially lower polarisation resistance. Thermogravimetric measurements at 650°C found the NiRh system took up more water than the Ni-only equivalent, and released it more exothermically, consistent with the more vigorous surface chemistry the rhodium is driving. Mass spectrometry confirmed the mechanism directly: switch the atmosphere to steam, and the rhodium-nickel catalyst produces a burst of SO2. The nickel-only catalyst produces rather less. Its sulfur, largely, stays put.

Solid oxide fuel cells have an awkward position in the clean energy landscape. They’re efficient, converting fuel to electricity directly, without combustion, at efficiencies that turbines can’t match. They’re flexible, able in principle to run on whatever the local energy economy produces. But that flexibility has always been compromised by sensitivity. Real fuels are messy. The hydrogen sulfide problem is one of the most persistent reasons why SOFCs that look promising in laboratory conditions, running on pure hydrogen, struggle when they encounter the dirtier gases of actual infrastructure.

Part of the appeal of Duan’s approach is that it doesn’t require the fuel to be cleaned. The cell cleans itself, continuously, as a byproduct of the steam already present during normal operation. Regeneration happens passively. There’s no separate cleaning cycle, no valve system, no ancillary energy cost. “This work establishes a new design strategy for sulfur-tolerant electrochemical materials,” Duan says.

Lead author Yue Bao, a graduate student in Duan’s Materials Research Laboratory for Sustainable Energy, notes the findings reach beyond solid oxide fuel cells. The same bimetallic exsolution approach, and the same steam-enabled self-cleaning chemistry, could in principle be applied to high-temperature catalysis more broadly, anywhere that a catalyst must hold its activity in the presence of sulfur-contaminated gas streams. Natural gas reformers, biogas processors, syngas reactors: all face versions of the same poisoning problem.

The broader question is whether a rhodium-containing anode can be made cheaply enough to compete. Rhodium is among the most expensive metals on earth, more so even than platinum, and the economics of adding it to a fuel cell material at industrial scale are not yet worked out. The research group’s task now is to understand how little rhodium is actually needed, whether the same self-cleaning effect can be achieved with smaller and smaller additions, until the material cost stops being an obstacle and solid oxide fuel cells can finally eat the dirty fuels they’ve always promised to handle.

Study link: https://pubs.acs.org/doi/10.1021/jacs.6c01484