New insights into the support morphology-dependent ammonia synthesis activity of Ru/CeO₂ catalysts

Abstract

Three ceria nanocrystals with different morphologies (nanocubes, nanorods, and nanoparticles) were synthesized and used as supports to evaluate the support morphology-dependent ammonia synthesis activity in the structure-sensitive ammonia synthesis reaction. The Ru/r-CeO₂ catalyst shows a higher catalytic reaction rate (3830 μmol g⁻¹ h) than that of Ru/c-CeO₂ (1289 μmol g⁻¹ h) and Ru/p-CeO₂ (529 μmol g⁻¹ h) under the reaction conditions of 400 °C, 1 MPa, and 60 mL min⁻¹. In addition, promoters (Cs, K, and Ba) can effectively enhance the catalytic activity of the catalysts such as 2Cs–Ru/r-CeO₂ (14 266 μmol g⁻¹ h) and 2K–Ru/r-CeO₂ (11 227 μmol g⁻¹ h). The effect of the morphology of the CeO₂ support on the metal–support interaction and catalytic performance of the Ru/CeO₂ catalysts can be correlated to the exposed crystal planes and surface composition/electronic structure of the CeO₂ support with diverse morphologies.

---

1. Introduction

Supported ruthenium catalysts have much higher activity at lower temperatures and pressures than that of the conventional iron catalysts in the synthesis reaction of ammonia.^1,2 Therefore, extensive efforts have been focused on the development of effective Ru-based catalysts, including 1) Ru particle size and shape, with the most active sites (B₅-type sites) ranging from 1.8 to 3.5 nm, 2) basic promoters such as alkali or alkaline earth compounds (Cs⁺, K⁺, Ba²⁺, etc.), and 3) the search for a proper support for Ru particles, for the synthesis of ammonia. In particular, the focal point is the design of a support that possesses the ability to change the electronic properties of Ru and affect its morphology and structure. To date, many supports, such as carbon materials, nitrides, zeolites, and oxides, have been well-developed and successfully utilized for the synthesis of ammonia.^3–6 However, these supports, which consist of particles, are complex solid systems with ill-defined surface species and assorted polycrystals.

Ceria, as a typical polycrystalline metal oxide, is more effective for ammonia synthesis when used as a support rather than a promoter.⁷ It has been reported in the previous studies^8,9 that the reversible Ce⁴⁺/Ce³⁺ redox pair and oxygen vacancies cause the electron density of the Ru metal surface to increase and thereby enhances its activity for ammonia synthesis. Recently, ceria nanocrystals with uniform and well-defined planes, such as nanorods ({110} and {100}), nanocubes ({100}), and nanopolyhedrons ({111} and {100}), have been controllably synthesized.^10,11 Numerous theoretical calculations indicate that the different crystal planes of ceria could strongly affect the oxygen vacancy formation energy and the interaction with surface molecules¹² and metals such as Au,¹³ Pd,¹⁴ Ag,¹⁵ and Pt.¹⁶ Moreover, different exposed planes have various types of surface geometries or electronic structures and compositions, which could affect the chemical state of the active metal when being loaded on the different crystal planes. For instance, a recent study¹⁷ based on the oxidation of formaldehyde on Pd supported on ceria with different morphologies indicates that different facets have different concentrations and structures of oxygen vacancies, which can affect the chemical state of the active Pd component and further enhance its catalytic activity. Similar phenomena were observed when silver,¹⁸ aurum,¹⁹ platinum,²⁰ nickel,²¹ or ruthenium²² was supported on ceria with different morphologies. Ammonia synthesis over ruthenium catalysts is well known as a structure-sensitive reaction. Several groups^23,24 have reported the strong effect of Ru particle size and shape on its catalytic activity for ammonia synthesis. However, the morphological effect of the support on the ammonia synthesis activity over supported Ru catalysts, to the best of our knowledge, has never been reported.

Herein, we employed the different morphologies of ceria nanorods (r-CeO₂), nanocubes (c-CeO₂), and nanoparticles (p-CeO₂) as supports to prepare the Ru/CeO₂ catalysts (denoted as Ru/r-CeO₂, Ru/c-CeO₂, and Ru/p-CeO₂, respectively). Their physical and chemical properties were investigated by TEM, XRD, UV-vis, Raman, XPS, and TPR analyses. The catalysts exhibit significant support-morphology-dependent catalytic performances for the ammonia synthesis reaction, which could be correlated to the exposed crystal planes and surface composition of the CeO₂ nanocrystals. The promoters (Cs, K, and Ba) introduced to the catalysts obviously enhanced their ammonia synthesis activity.

2. Experimental section

2.1 Chemicals

All chemicals were of analytical grade and used directly without any further purification. They were purchased from Aladdin Industrial Corporation in China. The purity of nitrogen and hydrogen was 99.999%. Deionized water with a resistivity of 18.25 MΩ cm was used in all the reactions.

2.2 Preparation of supports

CeO₂ supports with various morphologies were prepared via an improved hydrothermal method, as previously reported.¹⁰ To obtain nanorods and nanocubes, 6.96 g of Ce(NO₃)₃·6H₂O and 19.6 g of NaOH were first dissolved in 5 mL and 35 mL of deionized water, respectively. Subsequently, these two solutions were mixed and directly transferred to a 50 mL Teflon-lined stainless steel autoclave without stirring, and then heated at 100 °C, 120 °C, and 180 °C for 24 h to obtain r-CeO₂, p-CeO₂, and c-CeO₂, respectively. After the hydrothermal treatment, the precipitates were washed with deionized water until the pH = 7, and then dried at 60 °C for 12 h, followed by calcination at 450 °C for 4 h in air at a heating rate of 2.5 °C min⁻¹.

2.3 Preparation of Ru/CeO₂ catalysts

Ru/CeO₂ catalysts were prepared using the incipient impregnation method. The CeO₂ powders with different morphologies were stirred with a solution of ruthenium carbonyl (Ru₃(CO)₁₂) in THF at room temperature for 12 h in a rotary evaporator. Then the solvent was removed under reduced pressure at 30 °C. The gray powder was calcined at 300 °C for 3 h in a hydrogen atmosphere at a heating rate of 3 °C min⁻¹, followed by cooling down to room temperature in a hydrogen atmosphere.

The promoters were introduced using the wet impregnation method with an aqueous solution of nitrates (KNO₃, Ba(NO₃)₂ or CsNO₃) at 30 °C. The resulting samples were dried at 110 °C for 12 h. In addition, 2Cs–Ru/MgO and 6Ba–Ru/AC catalysts were synthesized according to the previous reports.^25,26 The molar ratio of M/Ru (M = Cs, K, and Ba) in the 2M–Ru/CeO₂ and 2Cs–Ru/MgO catalysts is 2, whereas in the 6Ba–Ru/AC catalyst, it is 6.

2.4 Measurements of catalytic activity

In the ammonia synthesis reaction, 0.2 g catalyst was used for each experiment. The catalyst was diluted with quartz sand to 0.6 mL (stainless steel microreactor (φ = 6 mm)). A stainless steel reactor with an inner diameter of 6 mm was used. The catalysts were stabilized under different temperatures for 2 hours (reaction conditions: 1 MPa, 60 mL min⁻¹, and H₂/N₂ = 3). When the ammonia synthesis reaction reached a steady state, data were obtained. The ammonia synthesis rates in the effluent were determined via a chemical titration method with a known amount of H₂SO₄ (Congo red as the indicator).

2.5 Characterization of the catalysts

The crystalline structures of the samples were analyzed via X-ray powder diffraction (XRD) (X'pert, PANalytical, Dutch) using Cu Kα radiation (λ = 1.54050 Å). The crystallite size of the support was calculated using the Scherrer equation. The BET surface area (S_BET) and pore volume of all the samples were determined using N₂ physisorption at 77 K with a Micromeritics ASAP 2020 apparatus. Scanning electron microscope (SEM) images were obtained using a JEOL JSM-6701F scanning electron microscope. Transmission electron microscope (TEM) experiments were conducted using a JEM-2010 with an accelerating voltage of 200 kV. The element compositions were detected using an X-ray photoelectron spectrometer (XPS, ESCALAB 250Xi). The electron binding energy scale of all the spectra was calibrated using C 1s at 284.8 eV. Temperature programmed reduction (TPR) profiles of the samples were generated using a TP-5080 catalyst characterization instrument with a TCD detector. Before the H₂-TPR test, the samples (50 mg) were pretreated with an Ar flow (27 mL min⁻¹) at 573 K for one hour and were then cooled down to 288 K. The test was performed by heating the samples (50 mg) in a H₂/Ar (H₂, 10%) mixture flow (30 mL min⁻¹) at a linear heating ramp of 10 K min⁻¹ from 288 to 1173 K. CO chemisorption was determined using the pulse method by introducing a CO flow over the sample (150 mg) maintained at 293 K. Ru dispersion was calculated from the cumulative volume of CO adsorbed during the pulse, assuming a chemisorption stoichiometry CO/Ru = 1 : 1. UV-visible diffuse reflectance spectra were obtained using a UV-2550 (Shimadzu) spectrometer with BaSO₄ as the reference. Raman spectra were obtained using a Renishaw RM1000 (λ  =  514 nm and 325 nm). The percentage composition of Ru was determined by XRF using a MagixPW2403.

3. Results and discussion

First, three CeO₂ samples with different morphologies were fabricated (Fig. 1). Fig. 1a–c show the uniform cube morphology of c-CeO₂ with a side size of approximately 300 nm. Note that a few smaller cubes grew on the larger cubes (Fig. S1†). The high-resolution TEM (HRTEM) image, as shown in Fig. 1b, displays clear (100) lattice fringes with an interplanar spacing of 0.27 nm, which indicates that c-CeO₂ selectively exposes the (100) planes. The rod morphology of r-CeO₂ has a length in the range of 200–400 nm (Fig. 1d and f). An interplanar spacing of 0.19 nm, which is attributed to the (110) plane, is observed (Fig. 1e). This indicates that r-CeO₂ shows a preferred 1D growth direction along (110) and is thus enclosed by (110) and (100) planes. Moreover, some “dark pits” are observed on r-CeO₂ (Fig. S2†). These “dark pits” indicate that there are many defects in r-CeO₂. Fig. 1i shows irregular nanoparticles with a size of 100–300 nm, which are often referred as “classical nanoparticles”,²⁷ with the most stable (111) planes predominantly exposed. However, as previously reported^21,22 and revealed by the HRTEM images, only the (100) facet exists in c-CeO₂, whereas the (111) and (110) facets are the main exposed planes for p-CeO₂ and r-CeO₂, respectively.

Fig. 1 TEM and SEM images of the as-prepared CeO₂ nanocrystals and the corresponding catalysts: (a, b, and c) nanocubes (c-CeO₂), (d, e, and f) nanorods (r-CeO₂), (g and h) Ru/r-CeO₂ and Ru/c-CeO₂, and (i) nanoparticles (p-CeO₂).

Fig. 2 shows the XRD patterns of the three CeO₂ samples with different morphologies and the corresponding supported Ru samples. The XRD patterns of the three CeO₂ samples can be indexed to that of the fluorite CeO₂ cubic structure (JCPDS: 03-065-2975), with the lattice parameters of 5.422, 5.398, and 5.373 Å. After ruthenium loading on the CeO₂ samples, the resulting Ru/CeO₂ samples show Ru loadings close to the theoretical value of 4% (determined by XRF measurements, Table S1†), despite some weight loss. Moreover, the TEM images (Fig. 1g, h and S3†) and the elemental mapping of the Ru/CeO₂ catalysts (Fig. S4 and S5†) also disclose that the Ru nanoparticles are dispersed on the supports. The lattice spacing is 0.234 nm, which corresponds to the (100) plane of Ru (Hexagonal, JCPDS: 03-065-7645). However, no diffraction peaks related to Ru species are observed in the XRD patterns of the three Ru/CeO₂ catalysts, which implies that the Ru species are highly dispersed in the catalysts. CO chemisorption and XPS were further performed to obtain the Ru dispersity (Table S1†). The results show that Ru/r-CeO₂ and Ru/p-CeO₂ display much higher Ru dispersion (40.4% and 36.3%, respectively) than that of Ru/c-CeO₂ (23.1%), which is consistent with the trend observed by XPS. It is well-known that the strain is a type of metrology of lattice stress due to the surface effects and crystal imperfections.²⁸ Since oxygen vacancies are the main defects in the samples, the difference in microstrain (Table S1†) indicates that the density of oxygen defects in the r-CeO₂ sample is larger than that in the c-CeO₂ and p-CeO₂ samples. In addition, the larger microstrain of r-CeO₂ could be related to the “dark pits” observed in its TEM image (Fig. S2†). Loading Ru on the CeO₂ samples caused a decrease in the microstrain of the r-CeO₂ and p-CeO₂ samples, and contrarily, the microstrain of the c-CeO₂ sample increased. These results suggest that the surface of CeO₂ is enriched with oxygen vacancies, which can strongly bind to the Ru nanoparticles, and simultaneously, the Ru nanoparticles can induce the generation of oxygen vacancies in CeO₂. Since the introduction of strain within the lattice could further increase the charge carrier mobility,²⁹ we can deduce that the order of charge carrier mobility is Ru/r-CeO₂ > Ru/c-CeO₂ > Ru/p-CeO₂.

Fig. 2 XRD patterns for the different morphologies of CeO₂ and the corresponding supported Ru catalysts.

Raman spectroscopy was further employed to provide the bulk and surface information using 514 nm and 325 nm excitation laser lines (Fig. S6 and S7†). Fig. S6† presents the visible (514 nm) Raman spectra of the three CeO₂ supports and the corresponding catalysts. The Raman peak at around 462 cm⁻¹ is ascribed to the F_2g vibration mode of the cubic CeO₂, and the other peaks at around 593 cm⁻¹ and 1172 cm⁻¹ are associated with the defect-induced (D) modes and second-order longitudinal modes, respectively, of the cubic CeO₂ fluoride structure.²² Moreover, the Raman peaks at 690 and 964 cm⁻¹ can be assigned to the Ru–O–Ce bond resulting from the metal–support interaction.³⁰ The Raman spectra of Ru/CeO₂ are similar to that of the supports. However, the F_2g peaks shift to lower wavenumbers at around 452 cm⁻¹, which suggests that the deposited Ru lowers the symmetry of the Ce–O bond, and this results in an increased optical absorption.²¹ Moreover, the peaks of Ru species were not detected, which indicates that the Ru species were highly dispersed in the catalysts. Since UV Raman spectra are more sensitive to defect sites in ceria than the visible Raman spectra, the concentration of oxygen vacancies in the supports were detected with 325 nm excitation laser lines (Fig. S7†). The relative intensity ratio of I_D/I_F2g reflects the concentration of oxygen vacancies in the CeO₂ samples. This ratio for r-CeO₂ is much higher than that for the c-CeO₂ and p-CeO₂ samples, which suggests that r-CeO₂ has the most abundant oxygen vacancies. Theoretical calculations¹³ indicate that the oxygen vacancy formation energy on the different crystal planes of CeO₂ follow the order of (110) < (100) < (111). This phenomenon is also in accordance with the other results reported in the literature.^20–22

The ideal CeO₂ (100) surface is a polar plane and its surface is terminated by an O layer. The CeO₂ (110) and (111) surfaces are nonpolar planes (charge neutral) and their surfaces have both exposed Ce⁴⁺ and O atoms.³¹ However, the defects induced different surface compositions and electron structures on the planes. The zeta potential is the electrostatic potential that exists at the shear plane of a particle, which is related to both the surface charge and the local environment of the particle. The zeta potentials of r-CeO₂, c-CeO₂, and p-CeO₂ are shown in Fig. 3. For c-CeO₂ sample, the zeta potential is −22.9 ± 2.2 mV, which is due to the fact that the (100) plane was terminated by an O layer. The zeta potential of p-CeO₂ is −6.3 ± 2 mV, which is close to that of the ideal (111) plane. This negative potential is ascribed to the reduction of Ce⁴⁺ to Ce³⁺ and oxygen species absorbed on Ce⁴⁺. However, the zeta potential of r-CeO₂ is +9.4 ± 0.6 mV, which could be assigned to the “dark pits” observed on its HRTEM images, indicating the presence of abundant pit-surrounding Ce sites.³²

Fig. 3 Zeta potential of the three CeO₂ supports.

The surface electronic states were obtained from UV/vis diffuse reflectance measurements. The support and the catalyst exhibit two intense absorption bands at about 271 and 343 nm (Fig. 4). These bands could be assigned to the O²⁻ → Ce⁴⁺ charge transfer and the interband transitions,³³ respectively. Moreover, a weak peak at 219 nm is attributed to the O²⁻ → Ce³⁺ charge transfer. Compared with the absorption edge of p-CeO₂, which in c-CeO₂ and r-CeO₂ shifts to a larger wavelength, this redshift in fact is due to the presence of Ce³⁺ and F color centers.³⁴ These results suggest that the surface electron structures of CeO₂ with different planes are different, which is consistent with the zeta potential results.

Fig. 4 UV-vis diffuse reflectance spectra for the different morphologies of CeO₂ and the supported Ru catalysts.

XPS analysis was used to investigate the surface chemical state of Ce, O, and Ru elements in the three Ru/CeO₂ catalysts. The XPS survey spectra of the three Ru/CeO₂ catalysts (Fig. S8†) indicate that the as-prepared composites comprised Ce, O, and Ru elements. Fig. 5 shows the high-resolution XPS spectra of Ce 3d, O 1s, and Ru 3d for the three Ru/CeO₂ catalysts. The XPS spectra of Ce 3d are fitted into ten components. Both valences (3+ and 4+) are present in the three Ru/CeO₂ catalysts. The relative percentages of Ce, Ru, and O species are obtained from the area ratio of the peaks. As shown in Table S2,† Ru/r-CeO₂ has the highest Ce³⁺/Ce⁴⁺ ratio (0.59) among the three Ru/CeO₂ catalysts, followed by Ru/c-CeO₂ (0.46) and Ru/p-CeO₂ (0.34). Since the presence of Ce³⁺ is associated with the generation of oxygen vacancies, the concentration of surface oxygen vacancies should follow the same order: Ru/r-CeO₂ > Ru/c-CeO₂ > Ru/p-CeO₂, which agrees well with the Raman and microstrain results. The Ru/r-CeO₂ and Ru/c-CeO₂ catalysts exhibit larger Ce³⁺/Ce⁴⁺ ratios than their corresponding supports. This could be explained by the formation of more oxygen vacancies. In addition, O 1s could be fitted with the three peaks³⁵ at 529.4 eV (O_L), 531 eV (Ov), and 531.9 eV (Oc), which are assigned to the lattice oxygen bound to the metal cations, O²⁻ in oxygen-deficient regions, and chemisorbed and dissociated oxygen species or OH, respectively. As a consequence, the number of surface oxygen vacancies could be roughly estimated based on the ratio of O_V/O_L.³⁶ As shown in Table S2,† r-CeO₂ has more oxygen vacancies than c-CeO₂, whereas the loading of Ru increases the oxygen vacancies of all the catalysts, which is in accordance with the results obtained from the Ce 3d spectra.

Fig. 5 High resolution XPS spectra: Ce 3d, O 1s, and Ru 3d.

The Ru X-ray photoelectron spectra (XPS) of the catalysts present a doublet corresponding to Ru 3d5/2 and Ru 3d3/2 (Fig. 5). The binding energy at 280.2 eV is characteristic of metallic ruthenium (Ru⁰). The shift in the peaks for Ru 3d5/2 to higher values (281.2 eV and 282.4 eV) can be attributed to Ru_n⁺, which most likely exists at the interface between Ru and CeO₂.³⁷ The presence of Ru_n⁺ indicates that the Ce–O bond is bound to the Ru nanoparticles, and this feature is also observed in the Raman spectra at 964 and 690 cm⁻¹. The relative percentages of Ru_n⁺ follow the order of Ru/r-CeO₂ (84.2%) > Ru/c-CeO₂ (79.8%) > Ru/p-CeO₂ (43.7%).

H₂-TPR was employed to investigate the reducibility of the CeO₂ supports and Ru/CeO₂ catalysts. The peaks in the high temperature region (>600 °C) could be attributed to the reduction of bulk CeO₂. The p-CeO₂ and r-CeO₂ supports exhibit strong reduction peaks at 462 and 415–498 °C (Fig. 6a), respectively, which could be attributed to the reduction of ceria surface oxygen. However, c-CeO₂ shows a weak reduction peak for ceria surface oxygen at 443 °C. These results imply that the reduction of pure ceria is influenced by the morphology. After Ru loading, the reduction peaks of ceria surface oxygen shift to lower temperatures (Fig. 6b). Given that the samples were pretreated with Ar for 1 h, the reduction peaks at 30, 43 and 69 °C should be caused by the weakening of the Ce–O bond by the strongly bound Ru species.¹⁷ Campbell³⁸ proposed the electronic metal–support interaction (EMSI) related to ceria formed by strong chemical bonds to Ru nanoparticles. As such, this strong Ce–O–Ru bond causes electron transfer from Ru to the support, thus forming Ru_n⁺. As a consequence, the relative number of Ru_n⁺ and Ce–O–Ru bonds can be roughly estimated by the area ratio of the reduction peaks. The area ratio of Ce–O–Ru in Ru/r-CeO₂ is 7 and 15 times more than that of Ru/c-CeO₂ and Ru/p-CeO₂, respectively. The relative number of Ru–O–Ce and Ru_n⁺ follows the order of Ru/r-CeO₂ > Ru/c-CeO₂ > Ru/p-CeO₂, which is in accordance with the results obtained from the Ru XPS spectra. Moreover, the surface oxygen species reduction occurred together with the reduction of Ru species at 65 and 93–116 °C in the Ru/c-CeO₂ and Ru/r-CeO₂ samples, respectively. However, in the case of Ru/p-CeO₂, the lower temperatures at 161 and 276 °C could be ascribed to the abstraction of O²⁻ by hydrogen from the surface ceria facilitated by the presence of the Ru nanoparticles. The reduction peak at 422 °C appears because the reduced ceria is far from the Ru nanoparticles. It is interesting to compare the surface oxygen reduction peak (<600 °C) area of the Ru/CeO₂ catalysts with that of the CeO₂ supports. The surface oxygen reduction peak area of Ru/r-CeO₂ is 1.8 times larger than that of the r-CeO₂ support. For the Ru/c-CeO₂ and Ru/p-CeO₂ catalysts, this peak area is 3.1 and 2.8 times larger than that of the corresponding supports, respectively. Specifically, the H₂ consumption amount of surface oxygen in the Ru/CeO₂ catalysts is far more than that in the CeO₂ supports. Generally, after the surface oxygen reduction in the Ru/CeO₂ catalysts, a Ce film was formed on the CeO₂ surface and Ru was deposited onto the Ce film. According to a previous report,³⁹ the Ce/Ru metal system can uptake hydrogen to form CeH_2+x hydride. Thus, it is reasonable to deduce that the excess H₂ consumption should be a result of the formation of the CeH_2+x hydride. As such, the amount of CeH_2+x hydride on the different planes could be similar to the number of Ru–O–Ce and Ru_n⁺ following the order: Ru/r-CeO₂ > Ru/c-CeO₂ > Ru/p-CeO₂. The earlier studies^40,41 have reported that CeH_2+x hydride can transfer electrons to Ru, as detected by the X-ray absorption spectroscopy (XAS), which suggests that the ammonia synthesis activity is related to the formation of CeH_2+x hydride on different planes.

Fig. 6 H₂-TPR for the different morphologies of CeO₂ (a) and the supported Ru catalysts (b).

Ammonia synthesis is a structure-sensitive reaction, which determines the optimal loading amount and size of Ru. For the loading amount of Ru, a previous study⁴² revealed that the dispersion of Ru decreases in the range of 1–5% Ru and the optimal loading is around 3.5%. In addition, Li et al.⁴³ found that the catalytic activity does not apparently increase when the Ru loading exceeds 4%. Therefore, three different ceria morphologies with a Ru loading of 4% were employed to establish the structure–activity relationship of the support. The ammonia synthesis rates and apparent activation energy (E_a) were evaluated on the three Ru/CeO₂ catalysts (Fig. 7a). The reaction rate of the Ru/r-CeO₂ catalyst was 3830 μmol g⁻¹ h, which was much higher than that of the Ru/c-CeO₂ (1289 μmol g⁻¹ h) and Ru/p-CeO₂ (529 μmol g⁻¹ h) catalysts at a reaction temperature of 400 °C. The E_a values of the three Ru/CeO₂ catalysts can be ranked as follows: Ru/r-CeO₂ (108 kJ mol⁻¹) < Ru/c-CeO₂ (124 kJ mol⁻¹) < Ru/p-CeO₂ (132 kJ mol⁻¹). These results indicate that the morphology of the CeO₂ supports has a significant effect on the rate of ammonia synthesis, which agrees well with the amount of CeH_2+x hydride on the different planes. In addition, it is widely accepted that the basicity of the support would lead to electron donation to Ru, and the acidity would withdraw electrons from Ru. Therefore, the activity of Ru/γ-Al₂O₃ for the ammonia synthesis is lower than that of Ru/MgO (Fig. S9†). For the Ru/p-CeO₂ catalyst, its basicity is weaker than that of MgO, whereas the activity of Ru/p-CeO₂ is higher than that of Ru/MgO. This could be ascribed to the formation of CeH_2+x hydride, which could transfer electrons to Ru and reversibly store and release hydrogen.⁴⁴Fig. 7b shows the correlation of the ammonia synthesis rate with the reaction time on stream. The catalysts were used to continuously produce NH₃ for 70 hours without deactivation.

Fig. 7 a) Arrhenius plots of Ru/CeO₂ catalysts in the temperature range of 613–693 K. b) Time dependence of the catalytic activities of the Ru/CeO₂ catalysts at 673 K. Reaction conditions: N₂ : H₂ = 1 : 3, 60 mL min⁻¹, and 1 MPa.

Alkali and alkaline earth metals (Cs, K, and Ba) are exclusively used as promoters for ammonia synthesis. These promoters can effectively lower the N₂ dissociation barrier and facilitate the destabilization of NH_x species. Fig. 8 shows the effect of K, Ba, and Cs promoters introduced into the three Ru/CeO₂ catalysts on the ammonia synthesis rate. Note that the catalytic activities increase by the introduction of Cs, K, and Ba promoters. Especially, the Cs promoter is the most effective to improve the ammonia synthesis rate with different morphologies of the Ru/CeO₂ catalysts. For the same morphology, such as the r-CeO₂ support, the effect of the promoters on the catalytic performance of 2M–Ru/r-CeO₂ (M = K, Cs, and Ba) catalysts follows the order of 2Cs–Ru/r-CeO₂ (14 266 μmol g⁻¹ h) > 2K–Ru/r-CeO₂ (11 227 μmol g⁻¹ h) > 2Ba–Ru/r-CeO₂ (6638 μmol g⁻¹ h). Therefore, the XPS of the 2M–Ru/r-CeO₂ catalysts was further performed to investigate the effect of the promoters on the catalytic activity (Fig. S10†). It can be observed that the Cs 3d^5/2 binding energy at 724.2 eV is much lower than that of Cs₂O (725.2 eV). According to the previous reports,^45–47 ruthenium can facilitate the decomposition of CsNO₃, KNO₃, and Ba(NO₃)₂ at around 120 °C and partial Cs reduction has been observed by XAS. Thus, the Cs 3d5/2 binding energy at 724.2 eV can be attributed to the partially reduced cesium. The Ba 3d^5/2 binding energy at 780 eV is ascribed to the formation of BaO, which is in accordance with the previous report.⁴⁸ Interestingly, the promoters obviously changed the electronic structure of the ruthenium and the support. The relative percentage of Ru⁰ was 15.8% in Ru/r-CeO₂, whereas that of Ru⁰ increased to 47.6%, 20.3%, and 48.9% in 2Cs–Ru/r-CeO₂, 2Ba–Ru/r-CeO₂, and 2K–Ru/r-CeO₂, respectively. Simultaneously, the ratio of Ce³⁺/Ce⁴⁺ decreased from 0.58 to 0.29, 0.31, and 0.23 (Table S3†). This suggests that the introduction of promoters causes the electron transfer from the support to Ru. For the promoters, the electron effect may be the main promoting action.^49,50 However, there is no direct correlation between the activity of Ru and the relative percentage of Ru⁰ and the values for ratio of Ce³⁺/Ce⁴⁺. Comparing the three Ru/CeO₂ catalysts involving the same promoter, the catalytic activities follow the order of 2M–Ru/r-CeO₂ > 2M–Ru/c-CeO₂ > 2M–Ru/p-CeO₂. These results indicate that the different planes of CeO₂ and the promoters play crucial roles to enhance the ammonia synthesis activity. Table S4† compares the rates of ammonia synthesis over Ru catalysts supported on various supports with different Ru precursors. The 2Cs–Ru/r-CeO₂ catalyst has much higher activity (33.5 mmol g⁻¹ h) at 3 MPa and 400 °C than that of the other catalysts. The 2Cs–Ru/MgO (10 530 μmol g⁻¹ h⁻¹) and 6Ba–Ru/AC (6543 μmol g⁻¹ h) catalysts were also used for comparison (Fig. S11†) at 1 MPa and 400 °C. The former is one of the most active Ru catalysts for ammonia synthesis and the latter is used for commercial ammonia production. Under these reaction conditions, the ammonia synthesis rate of the 2Cs–Ru/p-CeO₂ catalyst was 5103 μmol g⁻¹ h, which is much smaller than that of the 2Cs–Ru/MgO catalyst. With a change in the exposed planes of ceria, the ammonia synthesis rate significantly increased, for example, the ammonia synthesis rate of 6061 μmol g⁻¹ h for 2Cs–Ru/c-CeO₂ and 14 266 μmol g⁻¹ h for 2Cs–Ru/r-CeO₂. These results suggest that the effect of the morphology of CeO₂ support on the catalytic performance of Ru/CeO₂ and 2M–Ru/CeO₂ catalysts could be correlated to the exposed crystal planes and surface composition/electronic structure of the CeO₂ support with different morphologies.

Fig. 8 Activity comparison of the Ru/CeO₂ catalysts (a–c) with the promoters (K, Ba, and Cs). Reaction conditions: 673 K, N₂ : H₂ = 1 : 3, 60 mL min⁻¹, and 1 MPa.

4. Conclusions

We have synthesized three different morphologies of CeO₂ supports (nanocubes, nanorods, and nanoparticles), which dominantly exposed the (100), (110), and (111) facets, respectively. Raman, UV-vis, zeta potential, XPS, and H₂-TPR analyses verify that the surface composition and electronic structure of the various facets are different. The three Ru/CeO₂ catalysts and catalysts involving promoters (Cs, K, and Ba) exhibit significant support-morphology-dependent catalytic activity towards ammonia synthesis. The Ru/r-CeO₂ catalyst shows a higher catalytic reaction rate (3830 μmol g⁻¹ h) than that of Ru/c-CeO₂ (1289 μmol g⁻¹ h) and Ru/p-CeO₂ (529 μmol g⁻¹ h) at the reaction temperature (400 °C) and pressure (1 MPa). The promoters (Cs, K, and Ba) effectively enhance the catalytic activity, for example, the catalytic activity of 14 266 μmol g⁻¹ h for 2Cs–Ru/r-CeO₂ and 11 227 μmol g⁻¹ h for 2K–Ru/r-CeO₂. These results not only demonstrate that the catalytic performance of the catalysts can be tuned by controlling the morphology of the support but also provide a novel strategy to enhance the ammonia synthesis activity via increasing the number of highly efficient planes of the support.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of China (Grant No. 21403257).

References

1. H. Bielawa, O. Hinrichsen, A. Birkner and M. Muhler, Angew. Chem., Int. Ed., 2001, 40, 1061–1063.
2. R. Schlögl, Angew. Chem., Int. Ed., 2003, 42, 2004–2008.
3. B. C. McClaine, S. E. Siporin and R. J. Davis, J. Phys. Chem. B, 2001, 105, 7525–7532.
4. D. Szmigiel, H. Bielawa, M. Kurtz, O. Hinrichsen, M. Muhler, W. Raróg, S. Jodzis, Z. Kowalczyk, L. Znak and J. Zieliński, J. Catal., 2002, 205, 205–212.
5. R. Kojima and K. I. Aika, Chem. Lett., 2000, 29, 514–515.
6. Z. W. Ma, S. L. Zhao, X. M. Xiong, B. Hu and C. L. Song, Catal. Lett., 2016, 146, 2324–2329.
7. Y. Niwa and K. Aika, Chem. Lett., 1996, 3, 3–4.
8. Z. W. Ma, X. M. Xiong, C. L. Song, B. Hu and W. Q. Zhang, RSC Adv., 2016, 6, 51106–51110.
9. X. Luo, R. Wang, J. Ni, J. Lin, B. Lin, X. Xu and K. Wei, Catal. Lett., 2009, 133, 382–387.
10. C. Pan, D. Zhang, L. Shi and J. Fang, Eur. J. Inorg. Chem., 2008, 2008, 2429–2436.
11. H. X. Mai, L. D. Sun, Y. W. Zhang, R. Si, W. Feng, H. P. Zhang, H. C. Liu and C. H. Yan, J. Phys. Chem. B, 2005, 109, 24380–24385.
12. M. Nolan and G. W. Watson, J. Phys. Chem. B, 2006, 110, 16600–16606.
13. R. Si and M. Flytzani Stephanopoulos, Angew. Chem., Int. Ed., 2008, 47, 2884–2887.
14. M. Nolan, J. Mater. Chem., 2011, 21, 9160–9168.
15. Y. Xu, J. Greeley and M. Mavrikakis, J. Am. Chem. Soc., 2005, 127, 12823–12827.
16. A. Bruix, J. A. Rodriguez, P. J. Ramirez, S. D. Senanayake, J. Evans, J. B. Park, D. Stacchiola, P. Liu, J. Hrbek and F. Illas, J. Am. Chem. Soc., 2012, 134, 8968–8974.
17. H. Tan, J. Wang, S. Yu and K. Zhou, Environ. Sci. Technol., 2015, 49, 8675–8682.
18. S. Liu, X. Wu, W. Liu, W. Chen, R. Ran, M. Li and D. Weng, J. Catal., 2016, 337, 188–198.
19. M. F. Camellone and S. Fabris, J. Am. Chem. Soc., 2009, 131, 10473–10483.
20. Y. Gao, W. Wang, S. Chang and W. Huang, ChemCatChem, 2013, 5, 3610–3620.
21. N. Wang, W. Qian, W. Chu and F. Wei, Catal. Sci. Technol., 2016, 6, 3594–3605.
22. F. Wang, C. Li, X. Zhang, M. Wei, D. G. Evans and X. Duan, J. Catal., 2015, 329, 177–186.
23. C. Fernández, C. Sassoye, D. P. Debecker, C. Sanchez and P. Ruiz, Appl. Catal., A, 2014, 474, 194–202.
24. Z. Song, T. Cai, J. C. Hanson, J. A. Rodriguez and J. Hrbek, J. Am. Chem. Soc., 2004, 126, 8576–8584.
25. F. Rosowski, A. Hornung, O. Hinrichsen, D. Herein, M. Muhler and G. Ertl, Appl. Catal., A, 1997, 151, 443–460.
26. C. Liang, Z. Wei, Q. Xin and C. Li, Appl. Catal., A, 2001, 208, 193–201.
27. K. Zhou and Y. Li, Angew. Chem., Int. Ed., 2012, 51, 602–613.
28. R. Si, Y.-W. Zhang, S. J. Li, B. X. Lin and C. H. Yan, J. Phys. Chem. B, 2004, 108, 12481–12488.
29. G. Giri, E. Verploegen, S. C. B. Mannsfeld, S. Atahan-Evrenk, D. H. Kim, S. Y. Lee, H. A. Becerril, A. Aspuru-Guzik, M. F. Toney and Z. Bao, Nature, 2011, 480, 504–508.
30. A. Satsuma, M. Yanagihara, J. Ohyama and K. Shimizu, Catal. Today, 2013, 201, 62–67.
31. H. F. Wardenga and A. Klein, Appl. Surf. Sci., 2016, 377, 1–8.
32. Y. Sun, Q. Liu, S. Gao, H. Cheng, F. Lei, Z. Sun, Y. Jiang, H. Su, S. Wei and Y. Xie, Nat. Commun., 2013, 4, 2899.
33. L. Liu, Z. Yao, Y. Deng, F. Gao, B. Liu and L. Dong, ChemCatChem, 2011, 3, 978–989.
34. J. Wang, D. N. Tafen, J. P. Lewis, Z. L. Hong, A. Manivannan, M. J. Zhi, M. Li and N. Q. Wu, J. Am. Chem. Soc., 2009, 131, 12290–12297.
35. J. C. Dupin, D. Gonbeau, P. Vinatier and A. Levasseur, Phys. Chem. Chem. Phys., 2000, 2, 1319–1324.
36. Z. Hu, X. Liu, D. Meng, Y. Guo, Y. Guo and G. Lu, ACS Catal., 2016, 6, 2265–2279.
37. H. Y. Kim, H. M. Lee and G. Henkelman, J. Am. Chem. Soc., 2012, 134, 1560–1570.
38. C. T. Campbell, Nat. Chem., 2012, 4, 597–598.
39. A. P. Walker, T. Rayment and R. M. Lambert, J. Catal., 1989, 117, 102–120.
40. A. P. Walker, T. Rayment, R. M. Lambert and R. J. Oldman, J. Catal., 1990, 125, 67–76.
41. A. P. Walker and R. M. Lambert, J. Phys. Chem., 1992, 96, 2265–2271.
42. I. Rossetti and L. Forni, Appl. Catal., A, 2005, 282, 315–320.
43. Z. Li, C. Liang, Z. Feng, P. Ying, D. Wang and C. Li, J. Mol. Catal. A: Chem., 2004, 211, 103–109.
44. H. J. Lin, J. J. Tang, Q. Yu, H. Wang, L. Z. Ouyang, Y. J. Zhao, J. W. Liu, W. H. Wang and M. Zhu, Nano Energy, 2014, 9, 80–87.
45. W. Raróg-Pilecka, E. Miśkiewicz, S. Jodzis, J. Petryk, D. Lomot, Z. Kaszkur, Z. Karpiński and Z. Kowalczyk, J. Catal., 2006, 239, 313–325.
46. I. Rossetti, L. Sordelli, P. Ghigna, S. Pin, M. Scavini and L. Forni, Inorg. Chem., 2011, 50, 3757–3765.
47. I. Rossetti, F. Mangiarini and L. Forni, Appl. Catal., A, 2007, 323, 219–225.
48. E. Truszkiewicz, W. Raróg-Pilecka, K. Schmidt-Szałowski, S. Jodzis, E. Wilczkowska, D. Łomot, Z. Kaszkur, Z. Karpiński and Z. Kowalczyk, J. Catal., 2009, 265, 181–190.
49. Z. Kowalczyk, S. Jodzis, W. Raróg, J. Zieliński and J. Pielaszek, Appl. Catal., A, 1998, 173, 153–160.
50. I. Rossetti, N. Pernicone and L. Forni, Appl. Catal., A, 2001, 208, 271–278.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6cy02089e

---