MIT researchers use computer models to accelerate search for greener ammonia catalysts

Ammonia is everywhere in the world's food system, yet almost nobody thinks about it. The chemical ranks second only to sulfuric acid in total global production, and nearly all of it becomes fertilizer—the invisible foundation that allows us to feed eight billion people. But making ammonia the way we have for over a century comes at a steep cost. The Haber-Bosch process, which has dominated industrial ammonia production since the early 1900s, devours about 2% of the world's total energy consumption and generates roughly 1.5% of all greenhouse gas emissions. It relies on fossil fuels to generate the extreme heat and pressure needed to coax nitrogen from the air into ammonia molecules.

There is an alternative. Electrochemistry—using electricity instead of heat and pressure—could theoretically produce ammonia with far fewer emissions. The basic science works. But the process has never been efficient enough or cheap enough to compete with the entrenched Haber-Bosch method at industrial scale. The world currently consumes about 200 million metric tons of ammonia annually, and more than 90% of it still comes from the fossil fuel-dependent route. For the alternative to matter, it needs to work better and cost less.

A team at MIT believes they have found a way to speed up the search for the missing piece: the right catalyst. Catalysts are materials that accelerate chemical reactions without being consumed themselves, and their atomic structure determines how efficiently they work. Traditionally, finding better catalysts meant years of trial and error—synthesizing and testing millions of possible alloy combinations in the lab. Bilge Yildiz, a professor in MIT's departments of Nuclear Science and Engineering and Materials Science and Engineering, and doctoral students Constantine Athanitis and Filip Grajkowski developed a computational approach that predicts which materials are most likely to work before anyone builds them. Their findings, published in August in the journal EES Catalysis, use density functional theory—a quantum mechanical simulation tool—to model how different atomic arrangements behave. Machine learning then screens through the possibilities to identify the most promising candidates.

The researchers focused on transition metal nitrides as catalyst materials. These compounds have shown promise in electrochemical nitrogen reactions, and they work through an elegant mechanism: the nitrogen already present in the catalyst itself becomes part of the reaction pathway. This creates a cascade of chemical steps where each one provides energy to drive the next, reducing the total electrical input needed. One of the biggest obstacles in converting nitrogen gas to ammonia is the enormous energy required to break nitrogen's triple bond—one of the strongest bonds in chemistry. Metal nitrides help overcome this bottleneck by distributing the work across multiple reaction steps.

Athanitis explains the practical stakes clearly. "Even though a technology might be better for the world or for the climate, companies and capitalism won't really allow it unless it's cost competitive." The electrochemical method exists in principle, but it remains too slow and produces too little ammonia to make economic sense at scale. A catalyst that could reduce energy requirements while being more selective—producing ammonia while minimizing unwanted side reactions—would change the equation entirely. "If we can find a catalyst that reduces the energy needed and is more selective for ammonia production, we could essentially hit the jackpot," he says.

The MIT work is purely theoretical at this stage. The researchers have used computer models to identify promising alloys, but those materials have not yet been synthesized or tested in actual reaction cells. Dane Morgan, a materials engineering professor at the University of Wisconsin who was not involved in the study, calls the work "exciting" and notes that it clarifies how fundamental electronic properties of a material translate into catalytic performance. But he also sounds a note of caution: "Translating these calculations into practical catalysts will require many additional steps, so meaningful real-world impact is likely still some distance away."

The next phase requires moving from simulation to the laboratory. The researchers plan to build working reaction cells—devices that use the predicted catalysts to actually produce ammonia under real operating conditions. Only then will they know whether the computational predictions hold up in practice. Athanitis frames the work as part of a larger human endeavor to push at the boundaries of what is possible. "There have always been pushes at the frontiers," he says. "We like to think we've pushed the boundary of candidate materials here beyond what was thought of before." Whether this particular push leads to a breakthrough that transforms global ammonia production remains to be seen. But the direction is clear, and the stakes—feeding a growing world population while reducing emissions—could hardly be higher.