《基础化学》课程教学资源（文献资料）合成氨进展——科学家开发出氨合成节能技术.pdf

科学家开发出氨合成节能技术东京工业大学教授细野秀雄领导的研究小组10月22日在新一期英国期刊《自然一化学》网络版上报告说，他们开发出了一种高效合成氨的新技术，使用这种技术所消耗的能源只有传统方法的十分之一。氨对于地球上的生物相当重要，它是所有食物和肥料的重要成分，还会直接或间接参与药物合成，并且有望在燃料电池领域得到应用。氨是目前世界上产量最多的无机化合物之一，全球每年生产约1.7亿吨氨。氨由氮和氢反应合成，但是破坏氮分子之间强有力的结合，使其与氢发生反应需消耗大量能源。细野秀雄等研究者向其开发的超导物质C12A7中加入现在合成氨时常用的钉微粒，制成催化剂。C12A7是钙铝酸盐化合物，是高铝水泥的主要成分。研究人员发现，在这种催化剂作用下，氮和氢能高效合成氨。他们认为，这是由于在化合时相关电子变得容易移动，从而使氮分子容易成为原子。东京工业大学的研究人员准备今后进一步提高上述催化剂的性能，争取在5年至10年后使这项新技术达到实用水平。原文英文题目如下（原文见文献资料） Ammonia synthesis using a stable electride as an electron donor and reversible hydrogen store Industrially,the artificial fixation of atmospheric nitrogen to ammonia is carried out using the Haber-Bosch process,but this process requires high temperatures and pressures,and consumes more than 1%of the world' power production.Therefore the search is on for a more environmentally benign process that occurs under milder conditions.Here,we report that a Ru-loaded electride [Ca24Al28064]4+(e-)4 (Ru/C12A7:e-),which has high electron-donating power and chemical stability,works as an efficien catalyst for ammonia synthesis.Highly efficient ammonia synthesis is achieved with a catalytic activity that is an order of magnitude greater than those of other previously reported Ru-loaded catalysts and with almost half the reaction activation energy.Kinetic analysis with infrared spectroscopy reveals that C12A7:e-markedly enhances N2 dissociation on Ru by the back donation of electrons and that the poisoning of ruthenium surfaces by hydrogen adatoms can be suppressed effectively because of the ability of C12A7:e-to store hydrogen reversibly

科学家开发出氨合成节能技术东京工业大学教授细野秀雄领导的研究小组 10 月 22 日在新一期英国期刊《自然—化学》网络版上报告说，他们开发出了一种高效合成氨的新技术，使用这种技术所消耗的能源只有传统方法的十分之一。氨对于地球上的生物相当重要，它是所有食物和肥料的重要成分，还会直接或间接参与药物合成，并且有望在燃料电池领域得到应用。氨是目前世界上产量最多的无机化合物之一，全球每年生产约 1.7 亿吨氨。氨由氮和氢反应合成，但是破坏氮分子之间强有力的结合，使其与氢发生反应需消耗大量能源。细野秀雄等研究者向其开发的超导物质 C12A7 中加入现在合成氨时常用的钌微粒，制成催化剂。C12A7 是钙铝酸盐化合物，是高铝水泥的主要成分。研究人员发现，在这种催化剂作用下，氮和氢能高效合成氨。他们认为，这是由于在化合时相关电子变得容易移动，从而使氮分子容易成为原子。东京工业大学的研究人员准备今后进一步提高上述催化剂的性能，争取在 5 年至 10 年后使这项新技术达到实用水平。原文英文题目如下（原文见文献资料）： Ammonia synthesis using a stable electride as an electron donor and reversible hydrogen store Industrially, the artificial fixation of atmospheric nitrogen to ammonia is carried out using the Haber–Bosch process, but this process requires high temperatures and pressures, and consumes more than 1% of the world's power production. Therefore the search is on for a more environmentally benign process that occurs under milder conditions. Here, we report that a Ru-loaded electride [Ca24Al28O64]4+(e−)4 (Ru/C12A7:e−), which has high electron-donating power and chemical stability, works as an efficient catalyst for ammonia synthesis. Highly efficient ammonia synthesis is achieved with a catalytic activity that is an order of magnitude greater than those of other previously reported Ru-loaded catalysts and with almost half the reaction activation energy. Kinetic analysis with infrared spectroscopy reveals that C12A7:e− markedly enhances N2 dissociation on Ru by the back donation of electrons and that the poisoning of ruthenium surfaces by hydrogen adatoms can be suppressed effectively because of the ability of C12A7:e− to store hydrogen reversibly

Ammonia synthesis using a stable electride as an electron donor and reversible hydrogen store Masaaki Kitano1 , Yasunori Inoue1 , Youhei Yamazaki1 , Fumitaka Hayashi2, Shinji Kanbara1 , Satoru Matsuishi2, Toshiharu Yokoyama2, Sung-Wng Kim2†, Michikazu Hara1 * and Hideo Hosono1,2* Industrially, the artificial fixation of atmospheric nitrogen to ammonia is carried out using the Haber–Bosch process, but this process requires high temperatures and pressures, and consumes more than 1% of the world’s power production. Therefore the search is on for a more environmentally benign process that occurs under milder conditions. Here, we report that a Ru-loaded electride [Ca24Al28O64] 41(e2)4 (Ru/C12A7:e2), which has high electron-donating power and chemical stability, works as an efficient catalyst for ammonia synthesis. Highly efficient ammonia synthesis is achieved with a catalytic activity that is an order of magnitude greater than those of other previously reported Ru-loaded catalysts and with almost half the reaction activation energy. Kinetic analysis with infrared spectroscopy reveals that C12A7:e2 markedly enhances N2 dissociation on Ru by the back donation of electrons and that the poisoning of ruthenium surfaces by hydrogen adatoms can be suppressed effectively because of the ability of C12A7:e2 to store hydrogen reversibly. The commercial production of ammonia is greater than that of any other chemical, reaching 160 million tons per year. Most ammonia is consumed as ammonium sulfate, which is used as an essential fertilizer in crop production. Although industrial ammonia synthesis is conducted using the Haber–Bosch process with iron-based catalysts1,2 at 400–600 8C and 20–40 MPa, such high reaction temperatures are detrimental given that the reaction is exothermic (46.1 kJ mol21 ) (ref. 3). The rate-determining step of ammonia synthesis is cleavage of the N;N bond, because the bond energy is extremely large (945 kJ mol21 ) (refs 4,5). Transition metals, such as Fe or Ru, are indispensable for the promotion of N;N bond cleavage6–8, as are electron donors that provide electrons to the transition metals9–12. A N2 molecule is fixed to form a bond with a transition metal by donating electrons from its bonding orbitals and accepting electrons to its antibonding p-orbitals (back donation)13. Effectively, this back donation is enhanced by electron donors, which further weakens the N;N bond and results in the cleavage of N2 (refs 14–16). Electron donation from appropriate promoters is therefore key to enhancing the efficiency of ammonia synthesis using Fe or Ru catalysts10,14. However, in general it is extremely difficult to produce a material with a low work function as well as chemical and thermal stability. Although the catalytic activity of Ru is enhanced drastically by adding alkali or alkaline earth metals with small work functions17, these metals are unstable for ammonia synthesis because they are so chemically active that reaction with the produced ammonia forms metal amides18. Alkali and alkaline earth oxides are used exclusively as promoters; however, the activation of metal catalysts is still not efficient and the microscopic mechanism remains unclear19. If an efficient promoter with a distinct electrondonating ability could be found, then the performance of these catalysts would be increased significantly. Another obstacle for industrial ammonia synthesis with Ru-loaded catalysts is hydrogen poisoning under high hydrogen pressures. Siporin and Davis reported that promotion of Ru catalysts with basic compounds cannot be attributed solely to an effect on N2 dissociation20. Base promotion is a trade-off between sufficiently lowering the activation barrier for N2 dissociation without detrimentally increasing the competitive adsorption of H2. As the ammonia synthesis activity of Ru catalysts generally decays because of hydrogen adatoms formed on the Ru surface, the reaction order for H2 on Ru catalysts often approaches 21 (refs 11,20,21). Such H2 poisoning is a serious obstacle for industrial ammonia production that requires high-pressure conditions. The chemical industry is therefore currently searching for a supported Ru catalyst that promotes N2 dissociation, but suppresses H2 poisoning. Here we report a stable electride, C12A7:e2, that acts as an efficient electron donor for a Ru catalyst. The electride has a high electron-donating efficiency and chemical stability, and does not exhibit H2-poisoning because of its reversible hydrogen absorption/ desorption capability, which results from its crystal structure (Fig. 1a). Electrides are crystals with cavity-trapped electrons that serve as anions and were first synthesized by J. L. Dye in 1983 using crown ethers22. Although such materials are expected to have unique properties, no practical applications were reported because they were chemically and thermally unstable and decompose in an inert atmosphere or in air above approximately 230 8C. In 2003, an inorganic electride C12A7:e2 was synthesized by utilizing a stable inorganic oxide, 12CaO.7Al2O3 (C12A7), which is a constituent of commercial alumina cement. This material is able to form a complex with electrons and the resulting material became the first stable electride in air at and above room temperature23. The unit cell of C12A7 has a positively charged framework structure composed of 12 subnanometre-sized cages that connect to each other by sharing a mono-oxide layer to embrace four O22 ions in four cages as counteranions to achieve electroneutrality. Chemical reduction processes are used to inject four electrons into four of the 12 cages by extracting two of the O22 ions accommodated in the cavities as counteranions to compensate for the positive charge on the cage wall. The resultant chemical formula is represented by [Ca24Al28O64] 4þ(e2)4 (ref. 23). 1 Materials and Structures Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan, 2 Frontier Research Center, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan, † Present address: Department of Energy Science, SungKyunKwan University, Suwon, Korea. *e-mail: hosono@msl.titech.ac.jp; mhara@msl.titech.ac.jp ARTICLES PUBLISHED ONLINE: 21 OCTOBER 2012 | DOI: 10.1038/NCHEM.1476 NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry 1 © 2012 Macmillan Publishers Limited. All rights reserved

The injected electrons occupy a unique conduction band called the ‘cage conduction band’ (CCB)24, which is derived from the threedimensionally connected cages by sharing one oxide monolayer, and can migrate through the thin cage wall by tunnelling, which leads to metallic conduction (about 1,500 S cm21 at room temperature). This electron-trapped cage structure of the bulk is retained up to the top surface if the sample is heated appropriately25. The CCB in C12A7 was verified by photoemission spectroscopy and ab initio calculations and is derived from the very unique crystal structure of C12A7—threedimensionally connected subnanometre-sized cages with a positive charge (zeolite has a similar crystal structure, but the cage is negatively charged). No other stable electride has been realized since the first synthesis in 1983 by J. L. Dye, despite much interest in achieving an electride that is stable at room temperature. In addition, the electrons encapsulated in the cages of C12A7:e2 can be replaced readily with hydride ions (H2) by heating in H2 gas. The incorporated H2 ions desorb as H2 molecules at about 400 8C to leave electrons in the positively charged framework of C12A7 (ref. 26); the incorporation and release of hydrogen on C12A7:e2 is entirely reversible. The electride formation and reversible storage ability of hydrogen originates from the very unique crystal structure of C12A7 described above. Such a formation is, of course, impossible for other oxides, including Al2O3 and CaO (ref. 27). Results and discussion Structure and performance of Ru-loaded C12A7:e2. The electronic structure of C12A7:e2 is similar to that of an Fþ centre (see Fig. 1b), an electron trapped at the site of an O22 vacancy, in a CaO crystal. Basic oxides loaded with a noble metal, such as CaO and MgO with Fþ centres, are known to have catalytic activity for reactions that involve electron-transfer processes, because of electron transfer from the F centre on the substrate Valence band Valence band Cage conduction band 2.4 eV Conduction band 5.5 eV 7.4 eV CaO Evac Framework conduction band 4.7 eV Ru F+ centre Ef C12A7:e– 5 nm Electron H2 C12A7:e– Ca Al O Ru H– ion a b c d O Ca Ru CaO Electron 1.6 eV 4.1 eV Figure 1 | Ru-loaded C12A7 electride catalyst for ammonia synthesis. a, Schematic model of Ru-loaded C12A7:e2. High-density electrons (2.0 × 1021 cm23 ) are distributed statistically in the subnanometre-sized cages of C12A7 as counteranions and electrons encaged in C12A7 can be exchanged by H2 ions under an H2 atmosphere23,26. b, Fþ centres in the CaO crystal. Electrons are trapped at the oxygen-vacancy sites and octahedrally coordinated with six Ca2þ ions. The energy level of the Fþ centre in CaO is rather varied depending on the environment around the electron-trapping site30. c, Comparison of the energy levels for an Fþ centre in CaO, CCB in C12A7:e2 and the Fermi level (Ef ) in Ru. Evac denotes vacuum level. Here, CCB is the additional conduction band that originates from three-dimensionally connected nanocages in the fundamental band gap. Data for the relevant levels in CaO and C12A7:e2 are taken from previous reports30,31. The much higher energy level (that is, CCB) of the electrons trapped in the connected cages relative to that for the Fþ centre in CaO primarily results from a weaker Madelung potential caused by the larger separation between the electron and the nearest neighbour Ca2þ. d, TEM image of 0.3 wt% Ru-loaded C12A7:e2. ARTICLES NATURE CHEMISTRY DOI: 10.1038/NCHEM.1476 2 NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry © 2012 Macmillan Publishers Limited. All rights reserved

surface to the deposited metal cluster28. However, there are three distinct differences between the Fþ centres and C12A7:e2. Both the electron-donating efficiency and the stability of the Fþ centres are low because of a high potential barrier that originates from a strong electron confinement, low surface concentration and ease of structural alteration29. C12A7:e2 resolves these two drawbacks, as explained in the following. Figure 1c shows the electronic structure of C12A7:e2, along with that of Fþ centres in CaO (ref. 30) and the Fermi level of Ru. Although the levels of valence maximum and conduction-band minimum for C12A7:e2 are close to those of the Fþ centres in CaO, the location of the CCB level of C12A7:e2 is higher than those of the Fþ centres in CaO by 1.6–4.1 eV (refs 30,31). Therefore, the electrons encaged in C12A7:e2 can be donated effectively to the metal Ru (work function 4.7 eV) because of its intrinsic low work function (2.4 eV), which is comparable to that of potassium metal32. The third difference is a much higher encaged electron concentration at the surface of C12A7:e2, 1014–1015 cm22 , several orders of magnitude greater than that of the conventionally produced Fþ centres25. It is therefore anticipated that C12A7:e2 functions as an efficient electron donor for Ru nanoparticles with respect to electrondonating ability and the number of active sites for electron transfer to deposited Ru nanoparticles. In addition, we expect that the reversible hydrogen absorption/desorption capability of C12A7:e2 may prevent the poisoning of the Ru surface by hydrogen adatoms. Hydrogen adatoms on Ru would readily spill over onto C12A7:e2 and the excess hydrogen adatoms would be entrapped as H2 ions in the subnanometre-sized cages of C12A7, to reserve sufficient Ru sites to decompose N2. C12A7:e2 powders were prepared by a solid-state reaction according to a previous report33. The surface area of prepared C12A7:e2 was only 1 m2 g21 . Figure 1d shows a transmission electron microscopy (TEM) image of 0.3 wt% Ru/C12A7:e2 with Ru nanoparticles deposited on the C12A7:e2 surface. The mean particle size and the dispersion of Ru on C12A7:e2 were determined by CO chemisorption measurements (Supplementary Table S1) to be 30–40 nm and ,5%, respectively. Figure 2a shows the catalytic activities for ammonia synthesis over various 1 wt% Ru-loaded catalysts. Results for Ru-Cs/MgO and Ru-Ba/activated carbon (Ru-Ba/AC) are also shown for comparison; the former is one of the most active Ru catalysts for ammonia synthesis and the latter has been used for commercial ammonia production11,34. All the tested catalysts were reduced under reaction conditions before the reaction; nevertheless, Ru/C12A7:e2 exhibited a much higher catalytic activity than Ru/Al2O3, Ru/CaO, Ru/CaO−Al2O3 (Ru/CA), Ru/C12A7:O2 and Ru-Ba/AC and the ammonia effluent mole fractions for Ru/C12A7:e2 and Ru-Cs/MgO reached near-thermodynamic equilibrium (about 0.5%). Deposition of a large amount of Ru is required to achieve a high catalytic performance on Ru-based catalysts11,12,34–36. The relationship between the catalytic activity and the amount of deposited Ru was examined for each catalyst (Supplementary Table S1). The ammonia synthesis rate of Ru/C12A7:e2 increased with the Ru loading amount and reached a maximum at 1.2 wt% Ru deposition. Under optimal conditions, the catalytic activity of Ru/C12A7:e2 is comparable to those of Ru(6.0 wt%)-Cs/MgO and Ru(9.1 wt%)-Ba/AC, despite the lower surface area and lower amount of loaded Ru. Ru/C12A7:e2 has a higher turnover frequency (TOF), at least more than 10 times those of the conventional catalysts (Fig. 2b and Supplementary Table S1). For Ru-Ba/AC and Ru-Cs/MgO, these TOFs do not depend significantly on the amount of Ru loading, but the TOF value of the Ru/C12A7:e2 catalyst increases with a decrease in the amount of loaded Ru, to reach a maximum at 0.3 wt%. Supplementary Table S1 also shows that Ru on C12A7:e2 exhibits a much larger TOF than Ru-Ba/AC and Ru-Cs/MgO for Ru particles of similar size. There is no significant difference in Ru particle size (7–9 nm) between Ru(0.1 wt%)/C12A7:e2, Ru(9.1 wt%)-Ba/AC and Ru(6.0 wt%)-Cs/MgO, but nevertheless the TOF of Ru/C12A7:e2 is 20–50 times larger than those of Ru-Ba/AC and Ru-Cs/MgO. In addition, the catalytic activity of Ru/C12A7:e2 is superior to those of other catalysts at a reaction temperature of 320 8C, especially in terms of the TOF (Supplementary Table S2). As a consequence, the Ru/C12A7:e2 catalyst exhibits the smallest activation energy among the tested catalysts (Fig. 2b and Supplementary Table S1). The extremely high TOF and small activation energy indicate that C12A7:e2 imparts a high catalytic activity for ammonia synthesis onto the Ru surface. 0 0.15 0.05 0.10 × 10 a b 0 0.3 0.1 0.2 0.4 0 20 40 60 80 100 120 140 Ru/C12A7:e– Ru/C12A7:O2– Ru-Cs/MgO Ru-Ba/AC Ru/CaO 0 1,000 2,000 3,000 Ru/CA Ru/γ-Al2O3 Ru/γ-Al2O3 Ammonia synthesis rate (μmol g–1 h–1) Effluent NH3 mole fraction (vol%) Ru/C12A7:e– Ru/C12A7:O2– Ru-Cs/MgO Ru-Ba/AC Ru/CaO 0.20 0.25 TOF (molecule site–1 s–1) Apparent activation energy (kJ mol–1) Figure 2 | Catalytic performance of Ru/C12A7:e2 at atmospheric pressure. a, Ammonia synthesis rate and ammonia effluent mole fraction at 400 8C over various 1 wt% Ru-loaded catalysts. b, Turnover frequencies (TOFs; black bars) and apparent activation energies (red bars) for ammonia synthesis at 400 8C over various 1 wt% Ru-loaded catalysts. AC, activated carbon. Reaction conditions: catalyst, 0.2 g; synthesis gas, H2/N2 ¼ 3 with a flow rate of 60 ml min21 ; pressure, 0.1 MPa; temperature, 400 8C. An error analysis was performed for these data and the error range found was +5%. NATURE CHEMISTRY DOI: 10.1038/NCHEM.1476 ARTICLES NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry 3 © 2012 Macmillan Publishers Limited. All rights reserved.

Catalytic performance under high pressure. Figure 3a shows the variation of the TOF as a function of the total pressure on Ru/C12A7:e2 and Ru-Cs/MgO at 360 8C. Industrial catalysts for ammonia synthesis are required to have a high catalytic performance under high pressure, because high pressure is advantageous for the equilibrium of ammonia synthesis and the liquefied state of the resulting ammonia is much more convenient for utilization11. The TOF of Ru/C12A7:e2 is enhanced significantly by an increase in the total pressure and reaches a maximum (0.25 s21 ) at 1.0 MPa, whereas the TOF of Ru-Cs/MgO is almost independent of pressure. Table 1 summarizes the catalytic performance of various Ru-loaded catalysts under high-pressure conditions (1.0 MPa) at 400 8C. The amount of formed NH3 (see Table 1) is much smaller than that at thermodynamic equilibrium (about 4 vol%). Under these conditions, Ru/C12A7:e2 is substantially superior to Ru-Cs/MgO and Ru-Ba/AC with respect to TOF/surface Ru atoms and TOF/total Ru atoms. The maximum TOF of Ru/C12A7:e2 at 1.0 MPa reaches 0.98 s21 , which is still larger than those (0.05–0.28 s21 ) reported for Ru-Ba/AC and Ru-Cs/MgO at higher pressures (2.0–6.3 MPa)11,34–36. Such an enhancement of catalytic activity with pressure is not observed for conventional oxide-supported Ru catalysts, but instead the catalytic performance is independent of total pressure or decreases with increasing total pressure, because the dissociative adsorption of H2 on Ru prevents the adsorption and cleavage of N2 at high pressure20,37. Figure 3b demonstrates that the Ru/C12A7:e2 catalyst continuously produces ammonia without any decrease in activity for 75 hours, even under high pressure. The total amount of ammonia produced during the reaction reached 35 mmol. If all the electrons entrapped in the crystallographic cages of the whole C12A7:e2 sample were consumed in the ammonia synthesis reaction (that is, one electron used for the cleavage of one N2 molecule), then the ammonia yield would be only 452 mmol (per 0.2 g catalyst), which is smaller by two orders of magnitude than the amount of ammonia produced. This result indicates that Ru/C12A7:e2 functions as a stable catalyst for the ammonia synthesis reaction and electrons encapsulated in the Ru/C12A7:e2 catalyst can be used repeatedly during the reaction. Reaction mechanism. To understand the high catalytic performance of Ru/C12A7:e2, the adsorbed state of N2 on various Ru-loaded catalysts was examined using Fourier transform infrared (FT-IR) spectroscopy. Figure 4a shows that each catalyst exhibits broad peaks around 2,300–2,100 cm21 , caused by the stretching vibrational mode of chemisorbed N2 on the Ru surface with an end-on orientation10,38–40. The peak frequency (2,194 cm21 ) of N2 adsorbed on Ru/C12A7:O22 is lower than that (2,245 cm21 ) on Ru/g-Al2O3, which suggests that C12A7:O2 has alkaline characteristics similar to those of MgO (refs 38,39). In contrast, the spectrum for Ru/C12A7:e2 has three peaks at 2,234, Pressure (MPa) 0 0.2 0.4 0.6 0.8 1.0 0 0.30 0.25 0.20 0.15 0.10 0.05 a TOF (molecule site–1 s–1) Time (h) Ammonia production (mmol) 40 30 20 10 0 0 20 40 60 80 100 b Ru/C12A7:e– Ru-Cs/MgO Figure 3 | Catalytic performance of Ru/C12A7:e2 under high-pressure conditions. a, TOFs for high-pressure ammonia synthesis over 1 wt% Ru/C12A7:e2 and 6 wt% Ru-Cs/MgO. Reaction conditions: catalyst, 0.2 g; synthesis gas, H2/N2 ¼ 3 with a flow rate of 60 ml min21 ; temperature, 360 8C. The error range defined as the standard deviation for a set of experimental runs was+5% for the red and+7% for the black data points. b, Time course of ammonia formation over 1 wt% Ru/C12A7:e2. Reaction conditions: catalyst, 0.2 g; synthesis gas, H2/N2 ¼ 3 with a flow rate of 60 ml min21 ; pressure, 1.0 MPa, reaction temperature (360 8C). The error range was+5%. Table 1 | Catalytic performance of Ru catalysts on various supports at 400 8C under a pressure of 1.0 MPa. Catalyst Surface area (m2 g21 ) Ru loading (wt%)* Dispersion (%)† Particle size (nm)† NH3 synthesis rate (mmol g21 h21 ) ‡ TOF (s21 ) § TOF (3103 Ru atom21 s 21 ) NH3 (vol%)} Ru-Ba/AC 310 9.1 14.3 9.3 8,285 0.02 2.6 1.14 Ru-Cs/MgO 12 6.0 18.6 7.2 12,117 0.03 5.7 1.67 Ru/C12A7:O22 1–2 1.2 3.4 39.2 888 0.06 2.1 0.12 Ru/C12A7:e2 1–2 0.1 15.6 8.5 994 0.22 34.9 0.13 1–2 0.3 4.1 32.9 3,686 0.98 39.8 0.50 1–2 1.2 3.2 41.3 8,245 0.59 18.8 1.13 1–2 4.0 2.0 68.5 6,089 0.22 4.3 0.83 *Ru content was determined by inductively coupled plasma atomic emission spectroscopy. † Dispersion and particle size were calculated on the basis of CO chemisorption values, assuming spherical metal particles and that the stoichiometry of Ru/CO ¼ 1 (ref. 19). ‡ NH3 synthesis conditions: catalyst, 0.2 g; synthesis gas, H2/N2 ¼ 3 with a flow rate of 60 ml min21 ; temperature, 400 8C; pressure, 1.0 MPa. § TOF was calculated from the rate of ammonia synthesis divided by the number of CO atoms chemisorbed on the Ru surfaces. TOF was also calculated from the rate of ammonia synthesis divided by the number of Ru atoms deposited on the catalysts. } NH3 mole fraction in the reactor effluent. ARTICLES NATURE CHEMISTRY DOI: 10.1038/NCHEM.1476 4 NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry © 2012 Macmillan Publishers Limited. All rights reserved.