Mild Ammonia Synthesis over Ba-Promoted Ru / MPC Catalysts: E ﬀ ects of the Ba / Ru Ratio and the Mesoporous Structure

: A series of novel mesoporous carbon-supported, Ba-promoted, Ru catalysts with Ba / Ru ratios of 0.1–1.6 and a Ru loading of 10 wt% (denoted as 0.1–1.6Ba-10Ru / MPC) were prepared via stepwise impregnation of Ru and Ba precursors on the mesoporous carbon materials. The catalysts were applied to mild ammonia synthesis and compared to reference materials, including an analog ofthepreparedcatalystwithaBa / Ruratioof 1.6andaRu loadingof10wt%(denoted as1.6Ba-10Ru / AC). Characterization by X-ray di ﬀ raction (XRD), nitrogen physisorption, and electronic microscopy revealed that the 0.1–1.6Ba-10Ru / MPC catalysts contained Ru particles (approximately 2 nm) that were well-dispersed on the mesoporous structure and nanostructured Ba(NO 3 ) 2 species. These species decomposed into amorphous BaO x species, acting as a promoter on the metallic Ru particles forming catalytically active sites for ammonia synthesis. All the 0.1–1.6Ba-10Ru / MPC catalysts showed a synergistic e ﬀ ect of the active Ba and Ru species, which were stabilized in the mesoporous carbon framework with fast molecular di ﬀ usion and could e ﬀ ectively catalyze mild ammonia synthesis (280–450 ◦ C and 0.99 MPa) even under intermittently variable conditions, particularly for those with Ba / Ru ratios of > 0.5. In contrast, the 1.6Ba-10Ru / AC analog showed poor activity and stability for ammonia synthesis due to the sintering of Ba and Ru particles on the outer surface of the microporous carbon framework, resulting in low molecular di ﬀ usion and weak synergistic e ﬀ ect of the catalytically active sites. MPC catalysts were exposed to air. The HRTEM image shows that the Ru size (2.1 ± 0.9 nm) and mesoporous carbon framework of the used 1.6Ba-10Ru / MPC catalyst resembled those of the fresh 1.6Ba-10Ru / MPC catalyst. The HRTEM-mapping shows that the distributions of Ru and Ba over the 0.1–1.6Ba-10Ru / MPC catalysts were largely unchanged after ammonia synthesis. In contrast, the XRD and HRTEM results clearly show that the used 1.6Ba-10Ru / AC catalyst contained large BaCO 3 and Ru 0 particles with crystallite and particle sizes of ca. 13 and 4.2 ± 2.0 nm, respectively. The used 1.6Ba / MPC sample exhibited strong di ﬀ raction peaks large BaCO 3 particles formed the decomposition Ba(NO to


Introduction
The atmospheric CO 2 concentration has rapidly increased over the past two decades due to the burning of fossil fuels, causing increases in global temperature, sea level, and extreme climate [1]. As part of the adoption of the Paris Agreement, the Japanese government has set a goal to cut 26% and 80% of national CO 2 emissions by 2030 and 2050, respectively, based on the data recorded in 2013 [2]. To achieve this goal, hydrogen is a promising energy source with clean emissions, particularly for hydrogen produced by water electrolysis using renewable energy [3]. However, hydrogen is flammable, expensive, and hard to liquify, rendering its storage, transportation, and utilization quite difficult. The conversion of hydrogen to various chemicals, so-called "hydrogen carriers", is a potential method to store, transport, and use hydrogen energy more safely [4]. For instance, ammonia (NH 3 ) is composed of one nitrogen and three hydrogen atoms, corresponding to 17.6 wt% hydrogen and its industry is experienced including well-developed infrastructure for production, transportation, storage, Figure 1 shows the wide-angle XRD pattern of the prepared 1.6Ba-10Ru/MPC compared to those of the reference samples. The 1.6Ba-10Ru/MPC sample showed two sets of X-ray diffraction peaks and no signal arising from Ru was observed, suggesting that Ru is too small to be examined by XRD. A set of different peaks at 18.9, 21.8, 31.1, 36.6, 38.4, and 50.1 • are associated with the (111), (200), (220), (311), (222), (331), and (420) planes of Ba(NO 3 ) 2 (PDF card number: 3424; precursor). The Ba(NO 3 ) 2 particle size was estimated using the Scherrer equation as 49 nm. The diffraction peaks of Ba(NO 3 ) 2 with crystallite sizes ranging from 29 to 76 nm were observed for the 0.5-1.6Ba-10Ru/MPC, 1.6Ba/MPC, and 1.6Ba-10Ru/AC samples (Table 1). This indicates that Ba easily forms large particles in the microporous 1.6Ba-10Ru/AC samples. In addition to the decreased Ba size as a function of Ba Catalysts 2019, 9,480 3 of 12 loading, the decreased Ba size in the 1.6Ba/MPC and 0.5-1.6Ba-10Ru/MPC samples indicates that the mesoporous structure of MPC is a suitable support for impregnation of Ba species. In addition, the Ba sizes of the 1Ba-10Ru/MPC and 1.6Ba-10Ru/MPC catalysts were smaller than that of 1.6Ba/MPC, indicating that improved Ba dispersion can be obtained due to strong interaction of the Ba and Ru species. The diffraction peaks of Ba were hardly observed for the 0.1Ba-10Ru/MPC sample, implying that Ba was present in the form of semi-crystalline species at low Ba loadings. It should be noted that 1.6Ba-10Ru/MPC and the reference samples showed no Ru diffraction peaks, suggesting that Ru was finely dispersed in the MPC support with low or no crystallinity. The other set of diffraction peaks observed at 26.4 and 42.6 • are associated with the (002) and (100) planes of the graphite originally present in the MPC support. However, the graphite feature of MPC was slightly weakened after impregnation of Ba and Ru.  1 Microporous pore volume (V Micro ) was calculated using the Dubinin-Astakhov (DA) plot and α s -plot method. 2 Mesoporous pore volume (V Meso ) was calculated as V Total -V Micro . 3 Pore sizes determined at the peak maxima of the non-linear density function theory (NLDFT) calculation. 4 Determined from the high-resolution transmission electron microscopy (HRTEM) images. 5 Determined from CO chemisorption. The data in the parentheses are the Ru dispersions calculated by CO chemisorption.  6Ba-10Ru/AC. The "asterisk" peaks denote those associated with the carbon materials. Figure 2 shows the N2 adsorption-desorption isotherms and NLDFT pore size distributions of the prepared 1.6Ba-10Ru/MPC sample compared to those of the reference samples. The N2 adsorption-desorption isotherms can be divided into two groups: the classical type IV isotherm with an H1 hysteresis loop for 1.6Ba/MPC, 10Ru/MPC, and 0.1-1.6Ba-10Ru/MPC, which is associated with characteristic features of the MPC with mesoporous structure. The H1 hysteresis loop is slightly shifted to the lower P/P0 region by impregnation of Ru and Ba, and its intensity decreased. This indicates that the Ru and Ba species are impregnated inside the mesopores of MPC. The other group 10

Figure 1.
Wide-angle XRD patterns of the prepared Ba-Ru catalysts and reference materials-(a) 1.6Ba-10Ru/MPC, (b) 1Ba-10Ru/MPC, (c) 0.5Ba-10Ru/MPC, (d) 0.1Ba-10Ru/MPC, (e) 10Ru/MPC, (f) 1.6Ba/MPC, and (g) 1.6Ba-10Ru/AC. The "asterisk" peaks denote those associated with the carbon materials. Figure 2 shows the N 2 adsorption-desorption isotherms and NLDFT pore size distributions of the prepared 1.6Ba-10Ru/MPC sample compared to those of the reference samples. The N 2 adsorption-desorption isotherms can be divided into two groups: the classical type IV isotherm with an H 1 hysteresis loop for 1.6Ba/MPC, 10Ru/MPC, and 0.1-1.6Ba-10Ru/MPC, which is associated with characteristic features of the MPC with mesoporous structure. The H 1 hysteresis loop is slightly shifted to the lower P/P 0 region by impregnation of Ru and Ba, and its intensity decreased. This indicates that the Ru and Ba species are impregnated inside the mesopores of MPC. The other group can be classified as a typical type I isotherm with no apparent hysteresis loop (1.6Ba-10Ru/AC). It is evident that the 1.6Ba-10Ru/AC sample only contained a microporous structure. Table 1 lists the structural properties of the prepared 1.6Ba-10Ru/MPC sample, in comparison with reference samples.
can be classified as a typical type I isotherm with no apparent hysteresis loop (1.6Ba-10Ru/AC). It is evident that the 1.6Ba-10Ru/AC sample only contained a microporous structure. Table 1 lists the structural properties of the prepared 1.6Ba-10Ru/MPC sample, in comparison with reference samples. The specific surface area (SBET) of 1.6Ba-10Ru/MPC was similar to that of 1.6Ba-10Ru/AC, whereas 1.6Ba-10Ru/MPC exhibited higher total pore (VTotal) and mesopore volumes (VMeso) than those of 1.6Ba-10Ru/AC, which contained a large micropore volume (VMicro). The pore size analysis was calculated using the desorption branch via the non-linear density function theory (NLDFT) method and slit-pore model. The pore size of 1.6Ba-10Ru/MPC was determined to be approximately 5-6 nm, which is much larger than that of 1.6Ba-10Ru/AC. Similar results were observed for 1.6Ba/MPC, 10Ru/MPC and 0.1-1Ba-10Ru/MPC. Thus, the MPC-supported Ru catalysts with and without Ba promoter are large-pore mesoporous materials and 1.6Ba-10Ru/AC is a microporous material.
The microstructure and particle size distributions of the 1.6Ba-10Ru/MPC and reference samples were carefully examined by high-resolution transmission electron microscopy (HRTEM) and highangle annular dark field scanning transmission microscopy (HAADF-STEM). The related HRTEM images are shown in Figure 3 and Figure S1. The Ru size distributions and HAADF-STEM images are shown in Figures S2 and S3, respectively. The Ru nanoparticles of the 10Ru/MPC and 0.1-1.6Ba-10Ru/MPC catalysts were clearly observed on the mesoporous carbon framework and their sizes were approximately 1.8 nm, regardless of Ba loading. The Ru nanoparticles of the 1.6Ba-10Ru/AC catalyst also contained small Ru particles approximately 1.2 nm in size and apparently larger than the micropores (ca. 0.8 nm) of the AC support. This suggests that the Ru nanoparticles in 1.6Ba-10Ru/AC should be impregnated on the pore mouths of the AC microporous material. The HRTEM images with EDX mapping show that large Ba particles in the nanometer scale (approximately several tens of nm) aggregated on the prepared catalysts ( Figure S3), which is presumably associated with the Ba(NO3)2 particles observed in the XRD pattern ( Figure 1). These Ba(NO3)2 particles were converted into active BaOx species on the Ru particles for ammonia synthesis, which will be discussed in detail in Section 2.2.
The influences of the Ba promoter and porous structure on the Ru size and chemical environment of the prepared Ba-Ru catalysts were further investigated by CO chemisorption. All samples were reduced at 450 °C for 2 h under a H2 flow (50 mL min −1 ) before CO chemisorption. Table  1 shows that the Ru particle sizes of 10Ru/MPC and 0.1-1.6Ba-10Ru/MPC were 2-11 nm, which  The specific surface area (S BET ) of 1.6Ba-10Ru/MPC was similar to that of 1.6Ba-10Ru/AC, whereas 1.6Ba-10Ru/MPC exhibited higher total pore (V Total ) and mesopore volumes (V Meso ) than those of 1.6Ba-10Ru/AC, which contained a large micropore volume (V Micro ). The pore size analysis was calculated using the desorption branch via the non-linear density function theory (NLDFT) method and slit-pore model. The pore size of 1.6Ba-10Ru/MPC was determined to be approximately 5-6 nm, which is much larger than that of 1.6Ba-10Ru/AC. Similar results were observed for 1.6Ba/MPC, 10Ru/MPC and 0.1-1Ba-10Ru/MPC. Thus, the MPC-supported Ru catalysts with and without Ba promoter are large-pore mesoporous materials and 1.6Ba-10Ru/AC is a microporous material.
The microstructure and particle size distributions of the 1.6Ba-10Ru/MPC and reference samples were carefully examined by high-resolution transmission electron microscopy (HRTEM) and high-angle annular dark field scanning transmission microscopy (HAADF-STEM). The related HRTEM images are shown in Figure 3 and Figure S1. The Ru size distributions and HAADF-STEM images are shown in Figures S2 and S3, respectively. The Ru nanoparticles of the 10Ru/MPC and 0.1-1.6Ba-10Ru/MPC catalysts were clearly observed on the mesoporous carbon framework and their sizes were approximately 1.8 nm, regardless of Ba loading. The Ru nanoparticles of the 1.6Ba-10Ru/AC catalyst also contained small Ru particles approximately 1.2 nm in size and apparently larger than the micropores (ca. 0.8 nm) of the AC support. This suggests that the Ru nanoparticles in 1.6Ba-10Ru/AC should be impregnated on the pore mouths of the AC microporous material. The HRTEM images with EDX mapping show that large Ba particles in the nanometer scale (approximately several tens of nm) aggregated on the prepared catalysts ( Figure S3), which is presumably associated with the Ba(NO 3 ) 2 particles observed in the XRD pattern ( Figure 1). These Ba(NO 3 ) 2 particles were converted into active BaO x species on the Ru particles for ammonia synthesis, which will be discussed in detail in Section 2.2.
The influences of the Ba promoter and porous structure on the Ru size and chemical environment of the prepared Ba-Ru catalysts were further investigated by CO chemisorption. All samples were reduced at 450 • C for 2 h under a H 2 flow (50 mL min −1 ) before CO chemisorption. Table 1 shows that the Ru particle sizes of 10Ru/MPC and 0.1-1.6Ba-10Ru/MPC were 2-11 nm, which increased with increasing Ba loading and were much bigger than those determined by the HRTEM images, especially for the 0.5-1.6Ba-10Ru/MPC catalysts. Similar results were observed for the 1.6Ba-10Ru/AC catalyst. The Ru particle sizes of the prepared Ba-Ru catalysts were overestimated by the CO chemisorption method, likely due to that the covering of Ba species on the Ru surfaces of the prepared Ba-Ru catalysts, especially for the 0.5-1.6Ba-10Ru/MPC catalysts. Therefore, the adsorption of CO on the Ru surfaces was hindered, leading to the overestimation of Ru sizes. images, especially for the 0.5-1.6Ba-10Ru/MPC catalysts. Similar results were observed for the 1.6Ba-10Ru/AC catalyst. The Ru particle sizes of the prepared Ba-Ru catalysts were overestimated by the CO chemisorption method, likely due to that the covering of Ba species on the Ru surfaces of the prepared Ba-Ru catalysts, especially for the 0.5-1.6Ba-10Ru/MPC catalysts. Therefore, the adsorption of CO on the Ru surfaces was hindered, leading to the overestimation of Ru sizes.

Mild Ammonia Synthesis
The catalytic performance of the 0.1-1.6Ba-10Ru/MPC samples as solid catalysts for ammonia synthesis was examined in a stainless-steel fixed-bed reactor with a quartz inlet under mild conditions (280-450 °C and 0.99 MPa) and compared to reference samples. After the reaction, the downstream gas was analyzed using an online gas chromatography instrument equipped a thermal conductivity detector (GC-TCD) using a Thermon-3000 + KOH (2 + 2)% Sunpak-N 60/100 mesh column. The space velocity (SV) was maintained at 9000 h −1 and the standard G1-grade gas of H2 and N2 was used as a feedstock and the H2/N2 ratio was kept at 3. Prior to ammonia synthesis, the samples were reduced at 450 °C for 2 h using a pure H2 flow with an SV value of 10000 h −1 . The ammonia synthesis activity was calculated by dividing the synthesized amount per unit time by the catalyst mass (mmol g −1 h −1 ). Figure 4 shows the ammonia synthesis activity as a function of reaction temperature over the 0.1-1.6Ba-10Ru/MPC catalysts in comparison to those of the 10Ru/MPC, 1.6Ba/MPC, and 1.6Ba-10Ru/AC catalysts. A volcano-shaped curve with a maximum activity of approximately 10 mmol g −1 h −1 at 380 °C was observed for the 0.5-1.6Ba-10Ru/MPC catalysts, corresponding to the equilibrium of ammonia formation and decomposition reactions ( Table 2).

Mild Ammonia Synthesis
The catalytic performance of the 0.1-1.6Ba-10Ru/MPC samples as solid catalysts for ammonia synthesis was examined in a stainless-steel fixed-bed reactor with a quartz inlet under mild conditions (280-450 • C and 0.99 MPa) and compared to reference samples. After the reaction, the downstream gas was analyzed using an online gas chromatography instrument equipped a thermal conductivity detector (GC-TCD) using a Thermon-3000 + KOH (2 + 2)% Sunpak-N 60/100 mesh column. The space velocity (SV) was maintained at 9000 h −1 and the standard G1-grade gas of H 2 and N 2 was used as a feedstock and the H 2 /N 2 ratio was kept at 3. Prior to ammonia synthesis, the samples were reduced at 450 • C for 2 h using a pure H 2 flow with an SV value of 10000 h −1 . The ammonia synthesis activity was calculated by dividing the synthesized amount per unit time by the catalyst mass (mmol g −1 h −1 ). Figure 4 shows the ammonia synthesis activity as a function of reaction temperature over the 0.1-1.6Ba-10Ru/MPC catalysts in comparison to those of the 10Ru/MPC, 1.6Ba/MPC, and 1.6Ba-10Ru/AC catalysts. A volcano-shaped curve with a maximum activity of approximately 10 mmol g −1 h −1 at 380 • C was observed for the 0.5-1.6Ba-10Ru/MPC catalysts, corresponding to the equilibrium of ammonia formation and decomposition reactions ( Table 2).
Similar results were observed for the 0.1Ba-10Ru/MPC catalyst, although ammonia synthesis activity decreased and its maximum number was observed in the higher temperature region. Compared to the 0.5-1.6Ba/10Ru/MPC catalysts, the 1.6Ba-10Ru/AC catalyst showed only 1/4 activity for ammonia synthesis and the 10Ru/MPC catalyst was only active when the reaction temperature was >470 • C. The 1.6Ba/MPC sample was completely inactive for ammonia synthesis under the mild reaction conditions. These observations indicate that the Ba/Ru molar ratio should be >0.5 for the Ba-10Ru/MPC catalysts and that the mesoporous carbon framework can significantly facilitate ammonia synthesis. The turnover frequency (TOF) was calculated by dividing the ammonia synthesis rate by the number of surface Ru atoms estimated from CO chemisorption. The TOF values of the 0.1-1.6Ba-10Ru/MPC catalysts were positively correlated with the Ba/Ru molar ratio due to the formation of B 5 sites on the Ru surfaces [30,31]. Among the prepared Ba-Ru catalysts, 1.6Ba-10Ru/MPC showed a high rate of ammonia synthesis per Ru species and higher TOF than its counterparts and reference catalysts. The Ru and Ba species were homogeneously dispersed in the nanospace of the MPC support, resulting in a synergistic effect and high ammonia synthesis activity. It should be noted that the 1.6Ba-10Ru/MPC catalyst, with proper amounts of Ba and Ru, is superior to the previously developed catalyst 2.5Cs-10Ru/MPC with similar Cs and Ru contents synthesized using the same method. Thus, it is clear that Ba assists Ru-catalyzed ammonia synthesis, probably due to the electronic and structural promotion effects that create more B 5 sites on the Ru surfaces.  The maximum ammonia synthesis rate was observed from the curves in Figure 4 and corresponding reaction temperature. 2 The Cs/Ru molar ratio was referred to our recent study [32].
Similar results were observed for the 0.1Ba-10Ru/MPC catalyst, although ammonia synthesis activity decreased and its maximum number was observed in the higher temperature region. Compared to the 0.5-1.6Ba/10Ru/MPC catalysts, the 1.6Ba-10Ru/AC catalyst showed only 1/4 activity for ammonia synthesis and the 10Ru/MPC catalyst was only active when the reaction temperature w ˃ as 470 °C. The 1.6Ba/MPC sample was completely inactive for ammonia synthesis under the mild reaction conditions. These observations indicate that the Ba/Ru molar ratio should be ˃0.5 for the Ba-10Ru/MPC catalysts and that the mesoporous carbon framework can significantly facilitate ammonia synthesis. The turnover frequency (TOF) was calculated by dividing the ammonia synthesis rate by the number of surface Ru atoms estimated from CO chemisorption. The TOF values of the 0.1-1.6Ba-10Ru/MPC catalysts were positively correlated with the Ba/Ru molar ratio due to the formation of B5 sites on the Ru surfaces [30,31]. Among the prepared Ba-Ru catalysts, 1.6Ba-10Ru/MPC showed a high rate of ammonia synthesis per Ru species and higher TOF than its counterparts and reference catalysts. The Ru and Ba species were homogeneously dispersed in the nanospace of the MPC support, resulting in a synergistic effect and high ammonia synthesis activity. It should be noted that the 1.6Ba-10Ru/MPC catalyst, with proper amounts of Ba and Ru, is superior to the previously developed catalyst 2.5Cs-10Ru/MPC with similar Cs and Ru contents synthesized using the same method. Thus, it is clear that Ba assists Ru-catalyzed ammonia synthesis, probably due to the electronic and structural promotion effects that create more B5 sites on the Ru surfaces.   The maximum ammonia synthesis rate was observed from the curves in Figure 4 and corresponding reaction temperature. 2 The Cs/Ru molar ratio was referred to our recent study [32].
The microstructure and particle size distribution of the used Ba-Ru catalysts were examined by XRD and HRTEM and compared to those of fresh catalysts (Figures 5 and 6). The used 0.1-1.6Ba-10Ru/MPC catalysts exhibited weak diffraction peaks at 2θ = 26.4 and 42.6 • , corresponding to the graphite structure of MPC, and several weak diffraction peaks arising from the around 6 nm BaCO 3 particles. The mesoporous carbon framework with small graphite character was unaffected by the ammonia synthesis conditions. BaCO 3 was presumably formed by the reaction of BaO or Ba(OH) 2 species, which are typically formed by decomposition of the Ba(NO 3 ) 2 precursor and reaction with atmospheric CO 2 molecules when the used 0.1-1.6Ba-10Ru/MPC catalysts were exposed to air. The HRTEM image shows that the Ru size (2.1 ± 0.9 nm) and mesoporous carbon framework of the used 1.6Ba-10Ru/MPC catalyst resembled those of the fresh 1.6Ba-10Ru/MPC catalyst. The HRTEM-mapping shows that the distributions of Ru and Ba over the 0.1-1.6Ba-10Ru/MPC catalysts were largely unchanged after ammonia synthesis. In contrast, the XRD and HRTEM results clearly show that the used 1.6Ba-10Ru/AC catalyst contained large BaCO 3 and Ru 0 particles with crystallite and particle sizes of ca. 13 and 4.2 ± 2.0 nm, respectively. The used 1.6Ba/MPC sample exhibited strong diffraction peaks associated with large BaCO 3 particles formed by the decomposition of Ba(NO 3 ) 2 to large BaO and Ba(OH) 2 species, and subsequent reaction of these BaO x species with atmospheric CO 2 . This observation indicates that highly-dispersed, small BaO x particle-promoted metallic Ru species can be formed in the 0.1-1.6Ba-10Ru/MPC catalysts during ammonia synthesis, whereas the reverse is true for the 1.6Ba/MPC and 1.6Ba-10Ru/AC catalysts. Combining the XRD and HRTEM results with the aforementioned characterization and catalytic studies, the Ba(NO 3 ) 2 precursor is transformed into amorphous BaO or Ba(OH) 2 species in the 0.1-1.6Ba-10Ru/MPC catalysts under the ammonia synthesis reaction conditions. These amorphous BaO or Ba(OH) 2 species act as efficient promoters for Ru-catalyzed ammonia synthesis, where the catalytically active sites are the well-dispersed Ru particles stacked with activated BaO x species at the nanoscale. These conclusions are particularly supported by the CO chemisorption data. In the 1.6Ba-10Ru/AC catalyst, Ba(NO 3 ) 2 , which is probably inhomogeneously impregnated on the AC support, should exhibit low interaction with the Ru species. Both Ba and Ru aggregate easily on the AC support during ammonia synthesis resulting in the low activity of the 1.6Ba-10Ru/AC catalyst for ammonia synthesis. The stability and durability of the 1.6Ba-10Ru/MPC catalyst were surveyed by intermittently variable ammonia synthesis, continuously operated in a fixed-bed reaction system under a H 2 pressure of 0.99 MPa for >70 h, where the reaction temperatures and SV values varied between 300-380 • C and 9000-18000 h −1 , respectively. Figure 7 shows that the ammonia synthesis rate over the 1.6Ba-10Ru/MPC catalyst decreased with decreasing reaction temperature and SV. However, 1.6Ba-10Ru/MPC catalyzed intermittently variable ammonia synthesis at each stage with high stability, indicating that the ammonia synthesis activity can be finely and reversibly adjusted by the reaction parameters. This also indicates that the 1.6Ba-10Ru/MPC catalyst has potential application for intermittently variable ammonia synthesis under mild conditions, where its activity can be quickly adjusted to meet the supply requirements of renewable hydrogen derived from water hydrolysis using renewable electricity.
1.6Ba-10Ru/MPC catalyst resembled those of the fresh 1.6Ba-10Ru/MPC catalyst. The HRTEMmapping shows that the distributions of Ru and Ba over the 0.1-1.6Ba-10Ru/MPC catalysts were largely unchanged after ammonia synthesis. In contrast, the XRD and HRTEM results clearly show that the used 1.6Ba-10Ru/AC catalyst contained large BaCO3 and Ru 0 particles with crystallite and particle sizes of ca. 13 and 4.2 ± 2.0 nm, respectively. The used 1.6Ba/MPC sample exhibited strong diffraction peaks associated with large BaCO3 particles formed by the decomposition of Ba(NO3)2 to large BaO and Ba(OH)2 species, and subsequent reaction of these BaOx species with atmospheric CO2. This observation indicates that highly-dispersed, small BaOx particle-promoted metallic Ru species can be formed in the 0.1-1.6Ba-10Ru/MPC catalysts during ammonia synthesis, whereas the reverse is true for the 1.6Ba/MPC and 1.6Ba-10Ru/AC catalysts. Combining the XRD and HRTEM results with the aforementioned characterization and catalytic studies, the Ba(NO3)2 precursor is transformed into amorphous BaO or Ba(OH)2 species in the 0.1-1.6Ba-10Ru/MPC catalysts under the ammonia synthesis reaction conditions. These amorphous BaO or Ba(OH)2 species act as efficient promoters for Ru-catalyzed ammonia synthesis, where the catalytically active sites are the well-dispersed Ru particles stacked with activated BaOx species at the nanoscale. These conclusions are particularly supported by the CO chemisorption data. In the 1.6Ba-10Ru/AC catalyst, Ba(NO3)2, which is probably inhomogeneously impregnated on the AC support, should exhibit low interaction with the Ru species. Both Ba and Ru aggregate easily on the AC support during ammonia synthesis resulting in the low activity of the 1.6Ba-10Ru/AC catalyst for ammonia synthesis. The stability and durability of the 1.6Ba-10Ru/MPC catalyst were surveyed by intermittently variable ammonia synthesis, continuously operated in a fixed-bed reaction system under a H2 ˃ pressure of 0.99 MPa for 70 h, where the reaction temperatures and SV values varied between 300-380 °C and 9000-18000 h −1 , respectively. Figure 7 shows that the ammonia synthesis rate over the 1.6Ba-10Ru/MPC catalyst decreased with decreasing reaction temperature and SV. However, 1.6Ba-10Ru/MPC catalyzed intermittently variable ammonia synthesis at each stage with high stability, indicating that the ammonia synthesis activity can be finely and reversibly adjusted by the reaction parameters. This also indicates that the 1.6Ba-10Ru/MPC catalyst has potential application for intermittently variable ammonia synthesis under mild conditions, where its activity can be quickly adjusted to meet the supply requirements of renewable hydrogen derived from water hydrolysis using renewable electricity.

Characterization
The specific surface area and porosity of the prepared catalysts were determined by N2 physisorption using a BELSORP-max instrument (MicrotracBEL Corp., Osaka, Japan) at 77 K. The surface areas of the prepared samples were determined using the Brunauer-Emmet-Taylor (BET) equation while the micropore and mesopore volumes were calculated using the Dubinin-Astakhov (DA) plot or αs-plot method. The pore size distribution was obtained using the non-linear density function theory (NLDFT) method from the desorption isotherm using the slit-pore model. The compositions of the prepared catalysts were analyzed by elemental analysis (hydrogen, carbon and oxygen) and X-ray fluorescence. The crystallinity of the prepared catalysts was determined using a Rigaku MiniFlex600 diffractometer with Cu Kα radiation (λ = 0.15418 nm) operating at 40 kV and 15 mA. The Ru particle size and related size distribution were statistically analyzed via high-resolution transmission electron microscopy (HRTEM) using a TOPCON EM002B instrument operating at 120 kV. The Ru size distributions were calculated by counting more than 100 particles using a digital micrograph GMS 3 software (GATAN Inc., Pleasanton, CA, USA, 1996). The pulse chemisorption of CO was determined using an Ohkura Riken R6015 instrument. Freshly prepared samples were

Characterization
The specific surface area and porosity of the prepared catalysts were determined by N 2 physisorption using a BELSORP-max instrument (MicrotracBEL Corp., Osaka, Japan) at 77 K. The surface areas of the prepared samples were determined using the Brunauer-Emmet-Taylor (BET) equation while the micropore and mesopore volumes were calculated using the Dubinin-Astakhov (DA) plot or α s -plot method. The pore size distribution was obtained using the non-linear density function theory (NLDFT) method from the desorption isotherm using the slit-pore model. The compositions of the prepared catalysts were analyzed by elemental analysis (hydrogen, carbon and oxygen) and X-ray fluorescence. The crystallinity of the prepared catalysts was determined using a Rigaku MiniFlex600 diffractometer with Cu Kα radiation (λ = 0.15418 nm) operating at 40 kV and 15 mA. The Ru particle size and related size distribution were statistically analyzed via high-resolution transmission electron microscopy (HRTEM) using a TOPCON EM002B instrument operating at 120 kV. The Ru size distributions were calculated by counting more than 100 particles using a digital micrograph GMS 3 software (GATAN Inc., Pleasanton, CA, USA, 1996). The pulse chemisorption of CO was determined using an Ohkura Riken R6015 instrument. Freshly prepared samples were reduced under a H 2 flow (50 mL min −1 ) at 450 • C for 2 h, followed by purging with a He flow (50 mL min −1 ) until the TCD signal was stable at 50 • C. For the CO chemisorption, a sequential pulse using a standard gas of 10% CO/He was introduced to the reduced samples at 50 • C until no CO was adsorbed.

Mild Ammonia Synthesis
Ammonia synthesis over the prepared Ba-Ru catalysts was studied in a stainless steel fixed-bed reactor with a quartz inlet (12 mm, internal diameter) under mild reaction conditions (280-450 • C, <1 MPa). It should be noted that the High Pressure Gas Safety Act of Japan has defined "high pressure gas" as equal to or higher than 1 MPa at 35 • C. We specifically performed mild ammonia synthesis at the reaction pressure of <1 MPa using G1 grade N 2 and H 2 standard gases as feedstocks. The H 2 /N 2 ratio in the feed gas was maintained at 3. The prepared Ba-Ru catalysts were finely packed in the quartz inlet and placed in a stainless-steel cylindrical reactor, which was controlled by an automatic reaction test system (Taiyo system Corp., Japan). The Ba-Ru catalysts were reduced under a H 2 flow (SV = 10000 h −1 ) at 450 • C for 2 h before ammonia synthesis. To start the reaction, hydrogen and nitrogen (H 2 /N 2 ratio = 3) was fed to the fixed-bed reactor at 280-450 • C under a pressurized atmosphere. The stability test of the 1.6Ba-10Ru/MPC catalyst was performed for >70 h, where the reaction temperatures and SV were repeatedly and quickly changed in the ranges of 300-380 • C and in 9000-18000 h −1 , respectively. The quantitative analysis of the downstream products was performed using a Shimadzu gas chromatograph (GC-2014) equipped with a TCD detector and a Thermon-3000 + KOH (2 + 2)% Sunpak-N 60/100 mesh column (2.1 m length and 3.2 mm internal diameter, Shinwa Chemical Industries Ltd.). The ammonia synthesis rate was calculated by dividing the synthesized amount per unit time by the catalyst mass (mmol g −1 h −1 ).

Conclusions
Ba-promoted Ru nanoparticles supported on the mesoporous carbon materials with various Ba/Ru ratios (0.1-1.6Ba-10Ru/MPC) were prepared by the impregnation method and tested under mild ammonia synthesis conditions. The influences of the Ba/Ru ratio and mesoporous structure on the ammonia synthesis activity of the prepared 0.1-1.6Ba-10Ru/MPC catalysts were studied and compared to those of reference catalysts. The as-made 0.1-1.6Ba-10Ru/MPC catalysts contained Ru nanoparticles with sizes of approximately 1-2 nm, independent of Ru loading, and small Ba(NO 3 ) 2 crystallites (29-49 nm) that increased in size with increasing Ba loading. In contrast, the 1.6Ba-10Ru/AC catalyst with similar Ba and Ru loadings contained large Ba(NO 3 ) 2 crystallites and small Ru nanoparticles on the pore mouths of microporous structure. The Ba size in 1.6Ba/MPC was larger than that of 1.6Ba-10Ru/MPC, whereas the Ru sizes in 10Ru/MPC and 1.6Ba-10Ru/MPC were similar. In the catalytic study, all prepared 0.1-1.6Ba-10Ru/MPC catalysts were active towards ammonia synthesis, and their activities were much higher than that of the reference catalyst 1.6Ba-10Ru/AC. The 10Ru/MPC showed low activity for ammonia synthesis and the 1.6Ba/MPC sample without Ru showed no activity. The XRD, CO chemisorption, and HRTEM studies of the fresh and used catalysts showed that the 0.1-1.6Ba-10Ru/MPC catalysts contained small BaO x species close to the surface of the metallic Ru particles as the catalytically active sites, which were able to catalyze mild ammonia synthesis efficiently, due to the synergistic effect of Ba and Ru. Moreover, these active sites were highly stable in the mesoporous structure and remained nearly unchanged after use. In contrast, the 1.6Ba-10Ru/AC catalyst with Ba and Ru on the outer surface of AC with a microporous structure was unstable for ammonia synthesis and serious aggregation of Ba and Ru was observed. The intermittently variable synthesis of ammonia using the 1.6Ba-10Ru/MPC catalyst was performed in a fixed-bed reaction system under a H 2 pressure of 0.99 MPa for >70 h by frequently varying the reaction temperatures and SV values. Although the ammonia synthesis activity varied depending on the reaction parameters, the 1.6Ba-10Ru/MPC catalyst showed high stability at all stages of intermittently variable synthesis of ammonia. Thus, it can be concluded that the 1.6Ba-10Ru/MPC catalyst has potential application for the synthesis of ammonia under mild and variable conditions, which can be supplied renewable hydrogen produced via water electrolysis and renewable energy as a sustainable production process.
Author Contributions: M.N. designed and performed the experiments including the preparation and characterization of the catalysts and their catalyst activity tests and wrote the original paper; S.-Y.C. conceived of the characterization and catalytic tests of the prepared catalysts as well as reviewed and edited the paper; H.T. proposed and supervised the project. All the authors discussed and commented on the paper.
Funding: This research was funded by Japan Science and Technology Agency (JST), the Council for Science, Technology, and Innovation (CSTI), the Cross-ministerial Strategic Innovation Promotion Program (SIP), and the Energy Carriers program.