Electrochemical ammonia synthesis via nitrate reduction on Fe single atom catalyst

If you care about water and you care about energy, nitrate and ammonia sit on opposite sides of the story. Nitrate is one of the most common pollutants found in rivers and groundwater. Ammonia, on the other hand, is a cornerstone chemical and a potential energy carrier, but we mostly make it today with the Haber-Bosch process, which is centralized and carbon intensive. The idea that Wu, Siahrostami, Wang, and colleagues explore is simple and bold: take the nitrate already dissolved in water and electrochemically convert it into ammonia. They aim to do this at room temperature, in a way that could be distributed, and do it selectively, which is the challenging part. Selectivity is where most catalysts struggle. Nitrate can take many pathways. On many metals, nitrogen atoms can meet and couple, resulting in nitrogen gas instead of ammonia. Wu and colleagues chose to use single atoms to address this—iron atoms, each positioned one at a time, anchored in a nitrogen-doped carbon scaffold. This setup lacks neighboring metals to facilitate the joining of two nitrogens. With one active site, only one nitrogen can be processed at a time. The results are impressive. They observe ammonia Faradaic efficiency hovering around seventy-five percent. They measure a partial current for ammonia close to one hundred milliamperes per square centimeter at more negative bias, and mass-normalized ammonia production rates that exceed twenty thousand micrograms per hour per milligram of catalyst. That’s not just a curiosity; it’s production-level throughput in a lab cell. How do you construct such a sparse catalyst while keeping it accessible? They employ a clever template-assisted carbonization method. Iron chloride and an organic precursor are mixed with silica nanoparticles, baked under argon, and then the silica is etched away. What remains is a porous carbon sponge with iron atoms nestled into nitrogen sites. The loading isn’t heavy—about one and a half weight percent iron—but the structure is optimized for transport. The surface area is around two hundred eighty-six square meters per gram, the pore volume is nearly 0.80 cubic centimeters per gram, and the mesopores are centered around eighteen nanometers. In simpler terms, there are many open channels for nitrate to locate those isolated iron atoms. Do we truly achieve single atoms? The microscopy confirms this. In high-angle annular dark-field images, individual bright dots are seen scattered across carbon—no clusters, no particles. Spectroscopy provides additional information. Extended X-ray absorption shows a prominent first-shell feature at about 1.6 angstroms, which matches iron-nitrogen bonds, and the usual iron-iron signal near 2.2 angstroms is absent. When fitting the data, they find roughly four nitrogens surrounding each iron and an iron-nitrogen distance of about 1.92 angstroms. That represents the Fe-N4 motif that is often sketched on whiteboards. X-ray absorption at the iron L-edge indicates delocalized three-d electrons consistent with a coordinated atomic site. X-ray photoelectron spectra on the support show familiar nitrogen types—pyridinic, pyrrolic, graphitic—that can anchor those single iron atoms. This consistent picture emerges from multiple perspectives. So, what occurs when it is placed in the cell? In a neutral electrolyte with potassium sulfate and a generous amount of potassium nitrate, the first detection of ammonia appears as you sweep past about minus half a volt. That is referenced against the reversible hydrogen electrode. At that point, ammonia accounts for approximately thirty-nine percent of the charge, and the mass-normalized yield is a few hundred micrograms per hour per milligram of catalyst. If you push a bit harder, you hit the ideal spot around minus 0.66 volts, where ammonia claims roughly seventy-five percent of the current. The total current density hovers in the mid-thirties of milliamperes per square centimeter. You are producing just over five thousand micrograms of ammonia per hour per milligram of catalyst. That’s an excellent balance point—high selectivity without depriving the reaction. If you are willing to apply a more negative bias, the catalyst becomes even more productive. At about minus 0.85 volts, it channels nearly one hundred milliamperes per square centimeter into ammonia formation, reaching that headline production rate of roughly twenty thousand micrograms per hour per milligram. On a per-area basis, they report about 0.46 millimoles per hour per square centimeter of ammonia. The Faradaic efficiency does not remain consistent at every potential—selectivity and activity fluctuate as the driving force changes. However, over a two-hour period at the seventy-five percent point, the selectivity remains remarkably stable. Durability often poses a challenge for high-activity catalysts, but here, the single-atom sites prove resilient. In a simple H-cell, they conducted twenty consecutive cycles and observed performance fluctuating within a narrow range. They then transitioned to a flow cell with a bipolar membrane and an iridium oxide anode for oxygen evolution. They maintained the system for thirty-five hours at tens of milliamperes per square centimeter without any drift in potential or ammonia selectivity. The metal remains stable as well. Under conditions that typically cause iron nanoparticles to lose about twenty percent of their iron into solution, the single-atom catalyst shows no detectable iron dissolution by inductively coupled plasma analysis. This isolated coordination seems to enhance both selectivity and stability. Controls are important, as carbon can sometimes contribute partially. In this case, bare nitrogen-doped carbon barely reduces nitrate to ammonia. The iron nanoparticle version does produce ammonia, showing some selectivity in certain cycles. However, when normalized by iron content, the single-atom catalyst generates approximately twenty times more ammonia per mole of iron. This strengthens the argument that it is the Fe-N4 sites that are doing the heavy lifting. To determine the origin of the nitrogen in ammonia, they utilize isotope labeling. By using nitrate containing the heavier nitrogen-15, and after electrolysis, proton nuclear magnetic resonance indicates only the signature of ammonium carrying that same heavy isotope, with no mysterious contaminants disguising themselves as product. There is also an interesting story regarding the intermediates. Early in the reaction, nitrite emerges, contributing significantly to the current—approximately two-thirds at onset. As you drive the potential more negatively, that nitrite signal diminishes to under ten percent, while ammonia increases. When they feed nitrite directly to the same catalyst, ammonia dominates with more than ninety percent efficiency. This strongly suggests that nitrite serves as a pathway—a stepping stone on the way to fully hydrogenated ammonia. Delving deeper, the mechanism appears clean and intuitive—once you work through the calculations. The team utilizes density functional theory and a computational hydrogen electrode framework. In layman's terms, this allows them to calculate how the free energies of reaction steps change with applied potential. The entire journey from nitrate to ammonia involves nine protons and eight electrons. The first step is to protonate nitrate in solution to form nitric acid, requiring no electrons. On the surface, the crucial intermediate is adsorbed nitric oxide, denoted as NO*. The two most challenging steps involve reducing NO* to HNO* and then HNO* to atomic nitrogen on the site. These serve as the thermodynamic bottlenecks. Examining the entire energy landscape, every step becomes more favorable by about minus 0.30 volts compared to the reversible hydrogen electrode. Experimentally, the process begins to accelerate closely around minus 0.40 volts, indicating a gap of about a tenth of a volt. This can be reasonably attributed to kinetic barriers beyond the basic thermodynamics. Now, why are single atoms so crucial for selectivity? On extensive metal surfaces, two nitrogen-containing species can coexist and couple, lowering the barrier for creating dinitrogen or nitrous oxide. An isolated iron atom does not allow for such neighboring interactions. The calculations support this: N-N coupling intermediates are energetically unfavored on a lone Fe-N4 site. The laboratory results confirm this—no nitrogen gas products appear under their operating conditions. When comparing models, the differences become sharper. On a flat iron surface, referred to as Fe(110), the limiting step occurs further along in the sequence—transitioning from NH* to NH2*. The potential needed to ensure every step is downhill is actually positive, around plus 0.50 volts. This translates to considerably lower activity for nitrate conversion to ammonia under reductive bias. Substituting iron with cobalt or nickel in the same single-atom motif shows that the computed limiting potentials for the transition from NO* to HNO* increase to around 0.42 and 0.39 volts, respectively. This is again less favorable than iron. In this family of materials, iron occupies the sweet spot. The environment in which you operate also influences the outcome in sometimes subtle ways. Altering the nitrate concentration from dilute to concentrated barely impacts the ammonia efficiency, which remains in the mid-seventies across a range, while the production rate increases as more nitrate is added. It plateaus beyond approximately half a molar concentration. If you switch to an alkaline environment, you can achieve similar ammonia selectivity at less negative potentials, which is beneficial for energy efficiency. However, moving to strongly acidic conditions harms the entire process—both activity and selectivity suffer. In a mixed alkaline electrolyte, they even demonstrate a scenario where ammonia efficiency reaches eighty-six percent at about minus 0.21 volts. This highlights that electrolyte design is an essential factor, not just a background consideration. All these measurements rely on the cell design, and the group pays close attention to this aspect as they scale up. In the flow configuration, they separate the cathode and anode using a bipolar membrane and utilize a one molar potassium hydroxide anolyte to facilitate oxygen evolution. They also circulate the nitrate-containing catholyte across a gas-diffusion electrode. The outcome is a stable platform where they can maintain approximately minus thirty-five milliamperes per square centimeter for a day and a half, with consistent ammonia production at steady efficiency. It may not be a pilot plant, but it goes far beyond a simple benchtop experiment. When you take a step back, there's a satisfying coherence throughout this work. The synthesis achieves a structural target—Fe-N4 single atoms embedded in a porous carbon—with multiple, independent measurements confirming the same motif. This structure accomplishes two conceptual goals simultaneously. It creates a site where forming and hydrogenating NO* is thermodynamically easier than on adjacent transition metals, while also eliminating the possibility for two nitrogens to couple. The electrochemistry validates both aspects of that design. You observe early nitrite available for further reduction, ammonia dominating as you drive the process, and the absence of nitrogen gas. Performance drops when you replace it with nanoparticles or different metals, aligning with the predictions of the models. Moreover, when the system is stressed—through long durations or numerous cycles—the sites remain intact, and the iron does not dissolve. Where does this leave us? If you are managing agricultural runoff or nitrate-rich industrial streams, the potential to extract value from what is typically a liability is enticing. Considering decentralized ammonia production—made where you need it, when you need it—having a catalyst that simultaneously delivers high selectivity, high throughput, and robustness is rare. However, there are still system-level challenges to address. Hydrogen evolution continuously competes for current at negative bias. Collecting and concentrating ammonia from dilute solutions requires intelligent reactor design. Additionally, real wastewater is messier than potassium nitrate in the lab. The authors recognize this as they point to the need for testing in complex mixtures and developing cells that enhance ammonia capture. Compared to where the field stood just a few years ago, this represents a significant advancement. In the capable hands of Wu and colleagues, single-atom iron not only exhibits elegance under a microscope; it also facilitates charge movement effectively, at the right potentials, while sustaining that performance over time. It transforms a widespread pollutant into a valuable molecule, doing so with a simplicity that can be captured in one line: one atom, one site, one nitrogen at a time.