A fresh approach to synthesizing ammonia from air and water

Ammonia is vital to society, but its manufacture is energy intensive, has a large carbon footprint and requires high initial capital outlays. An intriguing reaction now suggests that energy-efficient alternatives are possible.
Máté J. Bezdek is in the Department of Chemistry, Princeton University, Princeton, New Jersey 08544, USA.

Search for this author in:

Paul J. Chirik is in the Department of Chemistry, Princeton University, Princeton, New Jersey 08544, USA.

Search for this author in:

Global food production requires ammonia-based fertilizers. The industrial transformation of atmospheric nitrogen gas (N2, also known as dinitrogen) into ammonia (NH3) is therefore essential for human life. Despite the simplicity of the molecules involved, the cleavage of the strong nitrogen–nitrogen triple bond (the N≡N bond) in dinitrogen and the concomitant formation of nitrogen–hydrogen (N–H) bonds poses a difficult challenge for catalytic chemistry, and typically involves conditions that are costly in terms of energy requirements: high reaction temperatures, high pressures or combinations of reactive reagents that are difficult to handle and energy-intensive to make. Writing in Nature, Ashida et al.1 demonstrate that a samarium compound mixed with water and combined with a molybdenum catalyst can promote ammonia synthesis from dinitrogen under ambient conditions. The work opens up avenues of research in the hunt for ammonia-making processes that operate under ambient conditions, and raises the question of what an ideal process should be.

Motivated by a looming global fertilizer shortage at the turn of the twentieth century, and later by munitions shortages (ammonia can be used to make explosives), the chemists Fritz Haber and Carl Bosch were the first to demonstrate2 that dinitrogen could be “pulled from air” and converted to ammonia. In the modern version of the Haber–Bosch process, dinitrogen and hydrogen gas are combined over a catalyst typically based on iron to produce ammonia (Fig. 1a). Today, global ammonia production occurs at a rate of about 250–300 tonnes per minute, and provides fertilizers that support nearly 60% of the planet’s population3,4.

Figure 1 | Comparison of approaches for making ammonia. a, The industrial Haber–Bosch synthesis of ammonia (NH3) reacts nitrogen gas (N2, also known as dinitrogen) with hydrogen molecules (H2), typically in the presence of an iron catalyst. The process requires high temperatures and pressures, but is thermodynamically ideal — minimal energy is wasted on side processes. b, Nitrogenase enzymes catalyse the reaction of dinitrogen with six electrons (e) and six protons (hydrogen ions; H+) under ambient conditions to make ammonia. However, two extra electrons and protons form a molecule of H2, and the conversion of ATP (the cell’s fuel molecules) to ADP drives the reaction. The process therefore has a high chemical overpotential — it uses much more energy than is needed simply to drive the ammonia-forming reaction. c, Ashida et al.1 report that a mixture of water and samarium diiodide (SmI2) converts nitrogen to ammonia under ambient conditions in the presence of a molybdenum catalyst; the SmI2 weakens the oxygen–hydrogen bonds in water, effectively producing hydrogen atoms (red) that react with dinitrogen. This approach might allow the development of reactions that have low overpotentials.

The modern conditions for ammonia synthesis involve temperatures greater than 400 °C and pressures of approximately 400 atmospheres, and are therefore often said to be ‘harsh’. This common misconception has motivated chemists to find ‘milder’ alternatives that use new catalysts to lower the operating temperatures and pressures. In reality, the search for new catalysts should be inspired by the need to reduce the capital expenditure associated with building ammonia plants, and by the requirement to reduce carbon emissions — not only from ammonia synthesis itself, but also from production of the hydrogen used in the process5.

Chemists have turned to nature for inspiration, as they often do. The nitrogenase family of enzymes is largely responsible for the biological conversion of dinitrogen to ammonia (a process called nitrogen fixation), and is the source of nitrogen atoms in amino acids and nucleotides, the building blocks of life. Unlike the Haber–Bosch process, however, nitrogenases do not use hydrogen gas as a source of hydrogen atoms. Instead, they transfer protons (hydrogen ions; H+) and electrons to each nitrogen atom to form N–H bonds (Fig. 1b). But although nitrogenases fix nitrogen at ambient temperatures, they use eight equivalents of protons and electrons per dinitrogen molecule (rather than six, the number needed according to the stoichiometry of the reaction) to provide the necessary thermodynamic driving force for fixation and for other coupled processes6. This use of excess hydrogen equivalents means that nitrogenases operate with a large chemical overpotential — they use much more energy than is actually needed to drive fixation7.

Chemists have mimicked the nitrogenase reaction by adding sources of protons and electrons to metal-containing complexes that contain bound dinitrogen. For example, workers from the same group as Ashida et al. previously reported8 molybdenum complexes that catalyse fixation in this way, producing up to 230 molecules of ammonia per molybdenum complex. However, the associated overpotentials are substantial (reaching nearly 300 kilocalories per mole of dinitrogen, in some cases)9. Viewed through this lens, the Haber–Bosch process is close to being a thermodynamically ideal process for ammonia synthesis, and is not as energetically harsh as it is sometimes claimed to be.

A challenge for catalysis researchers is to combine the best of the biological and industrial approaches to nitrogen fixation — that is, to find a process that operates near ambient temperature and pressure, has minimal chemical overpotential, and does not require a capital-intensive plant to make ammonia on a large scale. This is a big challenge, because no combination of acids (which are proton sources) and reducing agents (electron sources) has been found that provides a thermodynamic driving force for fixation on a par with that of hydrogen gas, and which is reactive enough to form N–H bonds from dinitrogen at, or near, ambient temperature.

But what would happen if, instead of functioning separately, proton and electron sources can be coaxed into working together? Ashida et al. have adopted this strategy, and thereby report what could be a fundamentally new approach to catalytic ammonia synthesis. They make use of a phenomenon known as coordination-induced bond weakening10, which arises from the interplay of samarium diiodide (SmI2) and water (Fig. 1c).

Water that is not in a chemical complex contains strong oxygen–hydrogen (O–H) bonds that are difficult to cleave. But when the oxygen atom in water coordinates (donates its lone pair of electrons) to SmI2, the O–H bonds are weakened and the resulting mixture becomes a potent source of hydrogen atoms — effectively, an excellent source of both protons and electrons. Ashida et al. use this source of hydrogen atoms with a molybdenum catalyst to fix nitrogen. Considerable coordination-induced bond weakening has previously been measured in SmI2–water mixtures, and used to make carbon–hydrogen bonds11,12.

The extension of this idea to catalytic ammonia synthesis is noteworthy for two main reasons. First, it is remarkable that the molybdenum catalyst facilitates ammonia synthesis in aqueous solution, because molybdenum complexes often degrade in water. Second, the use of coordination-induced bond weakening provides a new way of fixing nitrogen under ambient conditions that avoids the use of potentially dangerous combinations of proton and electron sources — such combinations can spontaneously ignite. The authors’ approach also works when ethylene glycol (HOCH2CH2OH) is used instead of water, expanding the range of hydrogen-atom sources for making ammonia by this method.

Ashida and co-workers propose a catalytic cycle for their process in which the molybdenum catalyst first coordinates to dinitrogen and cleaves the N≡N bond to form a molybdenum nitrido complex (which contains a molybdenum–nitrogen triple bond). The SmI2–water mixture then delivers hydrogen-atom equivalents to this complex, ultimately producing ammonia. Forming N–H bonds with molybdenum nitrido complexes poses a considerable thermodynamic challenge, because N–H bonds are also weakened when bound to molybdenum, as noted by our group10; this effect is a source of chemical overpotential. The SmI2 not only facilitates hydrogen-atom transfer, but also keeps the metal in a reduced form and prevents the deleterious formation of molybdenum oxide in aqueous solution.

The method reported has considerable operational challenges that currently make it impractical for synthesizing ammonia: SmI2 is used in large quantities, which generates a lot of waste; separating ammonia from aqueous solutions is energetically costly; and a chemical overpotential of about 140 kcal mol–1 remains. Nevertheless, Ashida and colleagues’ work creates a playground in which chemists can explore methods for ammonia synthesis. Future research should focus on finding alternatives to SmI2, based on metals that are more abundant than samarium, to promote coordination-induced bond weakening, enable N–H bond formation and lower the energetic costs of making ammonia from air and water.

Nature 568, 464-466 (2019)

doi: 10.1038/d41586-019-01213-7


  1. 1.

    Ashida, Y., Arashiba, K., Nakajima, K. & Nishibayashi, Y. Nature 568, 536–540 (2019).

  2. 2.

    Hager, T. The Alchemy of Air (Random House, 2009).

  3. 3.

    Schlögl, R. Angew. Chem. Int. Edn 42, 2004–2008 (2003).

  4. 4.

    Smil, V. Enriching the Earth: Fritz Haber, Carl Bosch, and the Transformation of World Food Production (MIT Press, 2001).

  5. 5.

    Appl, M. in Ullmann’s Encyclopedia of Industrial Chemistry 2012 Vol. 3, 139−225 (Wiley, 2012).

  6. 6.

    Hoffman, B. M., Dean, D. R. & Seefeldt, L. C. Acc. Chem. Res. 42, 609–619 (2009).

  7. 7.

    Pappas, I. & Chirik, P. J. J. Am. Chem. Soc. 138, 13379−13389 (2016).

  8. 8.

    Eizawa, A. et al. Nature Commun. 8, 14874 (2017).

  9. 9.

    Bezdek, M. J., Pappas, I. & Chirik, P. J. Top. Organometal. Chem. 60, 1−21 (2017).

  10. 10.

    Bezdek, M. J., Guo, S. & Chirik, P. J. Science 354, 730–733 (2016).

  11. 11.

    Chciuk, T. V. & Flowers, R. A. II J. Am. Chem. Soc. 137, 11526−11531 (2015).

  12. 12.

    Kolmar, S. S. & Mayer, J. M. J. Am. Chem. Soc. 139, 10687−10692 (2017).

Download references

Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up