Selective Hydrogenation of Benzene to Cyclohexene over Monometallic Ru Catalysts: Investigation of ZnO and ZnSO4 as Reaction Additives as Well as Particle Size Effect

Monometallic Ru catalysts with different particle size were prepared via a precipitation method and reduced at different temperatures. In addition, their catalytic activity towards cyclohexene formation from selective hydrogenation of benzene was investigated. With the utilization of ZnO and ZnSO4 as reaction additives, (Zn(OH)2)3(ZnSO4)(H2O)3 could be generated and chemisorbed on the Ru surface, which played a crucial role on increasing the selectivity to cyclohexene and retarding the catalytic activity towards benzene conversion. Interestingly, without addition of ZnO and ZnSO4, no cyclohexene was observed over all tested Ru catalysts with different particle sizes. This suggested that particle size plays no role in cyclohexene synthesis from selective hydrogenation of benzene over the pure monometallic Ru catalysts in the absence of ZnO and ZnSO4. On the other hand, when both ZnO and ZnSO4 were applied, surface n(Zn2+)/n(Ru) molar ratio increased with increasing particle size of the monometallic Ru catalysts after catalytic experiments, demonstrating that the content of chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 on Ru catalysts surface is enhanced under such a circumstance. More importantly, the maximum cyclohexene yield obtained over monometallic Ru catalysts showed a volcanic-type variation with increasing particle size of Ru from 3.6 nm to 5.6 nm. When the particle size of the monometallic Ru catalyst was 4.7 nm, the highest cyclohexene yield of 60.4% was achieved with an optimum n(ZnO)/n(Ru) ratio of 0.19:1 in the presence of 0.62 mol·dm−3 ZnSO4 within 25 min of catalytic experiments at 423 K under 5.0 MPa of H2. In addition, no decrease of catalytic activity towards cyclohexene generation was observed over this catalyst after 10 catalytic experiments without any regeneration.


Introduction
Caprolactam and adipic acid are of great importance for the modern chemical industry, since they play significant roles in producing nylon 6, nylon 66, polyamide, polyester. Both are widely utilized in the textiles, transportation and electronic industries, among others. Therefore, the selective hydrogenation of benzene to cyclohexene has drawn great industrial and academic interest, because cyclohexene can be easily converted to caprolactam and adipic acid by typical olefin reactions, Figure 1a illustrates the XRD patterns of Ru-4.7 catalyst with addition of different content of ZnO in the presence of ZnSO4 (0.62 mol·L -1 ) after catalytic experiments. Characteristic diffractions corresponding to metallic Ru were observed over all measured samples, demonstrating that Ru exists mainly as a metallic state (PDF: 01-070-0247). Furthermore, when the molar ratio of ZnO to Ru rises to 0.17, the reflections related to (Zn(OH)2)3(ZnSO4)(H2O)3 starts to be shown (PDF: 01-078-0247). In addition, the intensity of its reflections enhances with increasing the molar ratio of ZnO to Ru from 0.17 to 0.34, indicating that (Zn(OH)2)3(ZnSO4)(H2O)3 was generated during the catalytic experiments via the reaction between ZnO and ZnSO4 and was then chemisorbed on the Ru surface. Moreover, the content of (Zn(OH)2)3(ZnSO4)(H2O)3 increases with rising the molar ration of ZnO to Ru. On the other hand, Figure 1b shows the XRD patterns of Ru-4.7 catalyst after catalytic experiments with addition of only ZnO (n(ZnO)/n(Ru) = 0. 19), only ZnSO4 (0.62 mol·L -1 ), both ZnO (n(ZnO)/n(Ru) = 0.19) and ZnSO4 (0.62 mol·L -1 ) or without any additives, respectively. Reflections attributed to metallic Ru were detected over all analyzed catalysts, revealing the existence of metallic Ru as well. In addition, characteristic diffractions of ZnO were observed when only ZnO was added as an additive (PDF:01-070-2551), indicating that ZnO could not be reduced during the reaction. In comparison, (Zn(OH)2)3(ZnSO4)(H2O)3 was only pronounced over Ru-4.7 after catalytic experiment with addition of both ZnO (n(ZnO)/n(Ru) = 0.19) and ZnSO4 (0.62 mol·L -1 ). These observations indicate that the presence of both ZnSO4 and ZnO is crucial for the formation of (Zn(OH)2)3(ZnSO4)(H2O)3. H2-TPR profile of a non-reduced Ru sample is shown in Figure 2. A reduction peak with a small shoulder peak can be observed at a range between 300-520 K, which is corresponding to the reduction of Ru 3+ →Ru 2+ →Ru 0 [23]. In addition, the hydrogen consumption of 11.9 μmol H2/mg catalyst is obtained from a standard CuO calibration, which gives 80.2% of Ru content. This is consistent with that obtained from AAS (e.g., 80.4%). These results indicate that Ru could be completely reduced via the reduction procedure and exist as metallic state through the whole time, which are consistent to the XRD results. H 2 -TPR profile of a non-reduced Ru sample is shown in Figure 2. A reduction peak with a small shoulder peak can be observed at a range between 300-520 K, which is corresponding to the reduction of Ru 3+ →Ru 2+ →Ru 0 [23]. In addition, the hydrogen consumption of 11.9 µmol H 2 /mg catalyst is obtained from a standard CuO calibration, which gives 80.2% of Ru content. This is consistent with that obtained from AAS (e.g., 80.4%). These results indicate that Ru could be completely reduced via the reduction procedure and exist as metallic state through the whole time, which are consistent to the XRD results.  Table 1 shows the concentration of ZnSO4 (cZnSO4), molar ratio of Zn to Ru as well as S to Ru, particle size of the Ru-4.7 catalyst and pH values of the slurry after catalytic experiments. Neither Zn nor S species can be detected after reaction without any additives (1st row). By adding only ZnSO4 as an additive, slight amount of Zn (e.g., n(Zn)/n(Ru) = 0.0313) and S (e.g., n(S)/n(Ru) = 0.0026) was observed after hydrogenation (2nd row). It was reported by Sun et al. [24] that slender quantities of (Zn(OH)2)3(ZnSO4)(H2O)3 can be generated via the hydrolysis of ZnSO4. The same situation can be applied in this case. Furthermore, with only addition of ZnO, only Zn was detected after catalytic experiment (9th row). When both ZnSO4 and ZnO are utilized as additives, the molar ratio of Zn to Ru as well as S to Ru increase with increasing the ZnO content after catalytic experiments, enlightening the enhancement of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 on the Ru surface. This is consistent with the XRD results. It is worth noting that the particle size of Ru after reaction with all measured samples are comparable, i.e., ≈4.7 nm, indicating that the particle size of Ru does not change very much during the reaction. Additionally, the pH value of the slurry considerably decreases after the catalytic experiments with the presence of ZnSO4 (e.g., 5.53 of pH value with only addition of ZnSO4 vs. 7.10 of pH value with only addition of ZnO). This can be rationalized in terms of the hydrolysis of ZnSO4, resulting in the acidic environment. The pH value of the slurry continuously increases when raising the ZnO content, which is because ZnO can react with ZnSO4 to form the (Zn(OH)2)3(ZnSO4)(H2O)3 salt. The concentration of ZnSO4 decreases, hence the pH value rises.    Table 1 shows the concentration of ZnSO 4 (c ZnSO 4 ), molar ratio of Zn to Ru as well as S to Ru, particle size of the Ru-4.7 catalyst and pH values of the slurry after catalytic experiments. Neither Zn nor S species can be detected after reaction without any additives (1st row). By adding only ZnSO 4 as an additive, slight amount of Zn (e.g., n(Zn)/n(Ru) = 0.0313) and S (e.g., n(S)/n(Ru) = 0.0026) was observed after hydrogenation (2nd row). It was reported by Sun et al. [24] that slender quantities of (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O )3 can be generated via the hydrolysis of ZnSO 4 . The same situation can be applied in this case. Furthermore, with only addition of ZnO, only Zn was detected after catalytic experiment (9th row). When both ZnSO 4 and ZnO are utilized as additives, the molar ratio of Zn to Ru as well as S to Ru increase with increasing the ZnO content after catalytic experiments, enlightening the enhancement of the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 on the Ru surface. This is consistent with the XRD results. It is worth noting that the particle size of Ru after reaction with all measured samples are comparable, i.e., ≈4.7 nm, indicating that the particle size of Ru does not change very much during the reaction. Additionally, the pH value of the slurry considerably decreases after the catalytic experiments with the presence of ZnSO4 (e.g., 5.53 of pH value with only addition of ZnSO 4 vs. 7.10 of pH value with only addition of ZnO). This can be rationalized in terms of the hydrolysis of ZnSO 4 , resulting in the acidic environment. The pH value of the slurry continuously increases when raising the ZnO content, which is because ZnO can react with ZnSO 4 to form the (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt. The concentration of ZnSO 4 decreases, hence the pH value rises.   (Figure 3e,f). As can be seen, Ru particles over all analyzed samples display a circular or elliptical shape after reaction. In addition, the particle size of Ru is around 5 nm, which is in good agreement with XRD results (Figure 3a-d). Moreover, the stick shape of ZnO was observed when only ZnO was added as an additive (Figure 3b), suggesting that ZnO can be hardly dispersed on the Ru surface. In addition, it can be observed from Figure 3e,f that Zn and S species homogeneously dispersed on the Ru surface of Ru-4.7 AH. This indicates that the generated (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt is uniformly dispersed on the Ru surface as well.  (Figure 3e,f). As can be seen, Ru particles over all analyzed samples display a circular or elliptical shape after reaction. In addition, the particle size of Ru is around 5 nm, which is in good agreement with XRD results (Figure 3a-d). Moreover, the stick shape of ZnO was observed when only ZnO was added as an additive (Figure 3b), suggesting that ZnO can be hardly dispersed on the Ru surface. In addition, it can be observed from Figure 3e,f that Zn and S species homogeneously dispersed on the Ru surface of Ru-4.7 AH. This indicates that the generated (Zn(OH)2)3(ZnSO4)(H2O)3 salt is uniformly dispersed on the Ru surface as well. The wettability of Ru catalyst surface was further investigated both in the absence and in the presence of ZnO and ZnSO4. Water droplets on pure Ru as well as Ru with ZnO and ZnSO4 are shown in Figure 4. As demonstrated, when ZnO and ZnSO4 were utilized as reaction additives, the contact angle significantly decreased in comparison to that observed from Ru without adding ZnO and ZnSO4. This suggests that hydrophilicity of Ru surface was highly improved by the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt. XPS spectra of the Ru catalyst with a particle size of 4.7 nm before and after catalytic experiments were demonstrated in Figure 5. Figure 5a gives a peak at 461.2 eV that is attributed to Ru 3p3/2. This shows that Ru exists in a metallic state [25], which is consistent with XRD results. As The wettability of Ru catalyst surface was further investigated both in the absence and in the presence of ZnO and ZnSO 4 . Water droplets on pure Ru as well as Ru with ZnO and ZnSO 4 are shown in Figure 4. As demonstrated, when ZnO and ZnSO 4 were utilized as reaction additives, the contact angle significantly decreased in comparison to that observed from Ru without adding ZnO and ZnSO 4 . This suggests that hydrophilicity of Ru surface was highly improved by the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt.  (Figure 3e,f). As can be seen, Ru particles over all analyzed samples display a circular or elliptical shape after reaction. In addition, the particle size of Ru is around 5 nm, which is in good agreement with XRD results (Figure 3a-d). Moreover, the stick shape of ZnO was observed when only ZnO was added as an additive (Figure 3b), suggesting that ZnO can be hardly dispersed on the Ru surface. In addition, it can be observed from Figure 3e,f that Zn and S species homogeneously dispersed on the Ru surface of Ru-4.7 AH. This indicates that the generated (Zn(OH)2)3(ZnSO4)(H2O)3 salt is uniformly dispersed on the Ru surface as well. The wettability of Ru catalyst surface was further investigated both in the absence and in the presence of ZnO and ZnSO4. Water droplets on pure Ru as well as Ru with ZnO and ZnSO4 are shown in Figure 4. As demonstrated, when ZnO and ZnSO4 were utilized as reaction additives, the contact angle significantly decreased in comparison to that observed from Ru without adding ZnO and ZnSO4. This suggests that hydrophilicity of Ru surface was highly improved by the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt. XPS spectra of the Ru catalyst with a particle size of 4.7 nm before and after catalytic experiments were demonstrated in Figure 5. Figure 5a gives a peak at 461.2 eV that is attributed to Ru 3p3/2. This shows that Ru exists in a metallic state [25], which is consistent with XRD results. As  XPS spectra of the Ru catalyst with a particle size of 4.7 nm before and after catalytic experiments were demonstrated in Figure 5. Figure 5a gives a peak at 461.2 eV that is attributed to Ru 3p 3/2 . This shows that Ru exists in a metallic state [25], which is consistent with XRD results. As can be observed from Figure 5b, the electron-binding energy of Ru 3p 3/2 over Ru-4.7 AH is 462.5 eV, indicating the existence of metallic Ru as well [26]. It is worth noting that, since the electron-binding energy of Ru-4.7 AH is 1.3 eV higher than that observed over Ru-4.7, it can be concluded that Ru δ+ species of electron deficiency is generated during the catalytic experiment. Additionally, the deconvoluted electron-binding energy of Zn 2p 2/3 and the Auger electron kinetic energy of Zn·LMM was observed at 1022.0 eV and 988.6 eV over Ru-4.7 AH, respectively (Figure 5c,d), suggesting that Zn species mainly exist as Zn 2+ after reaction [27,28]. In addition, it is observed that the electron-binding energy of Zn 2p 2/3 (1022.0 eV) is lower than that obtained over Zn(OH) 2 (1022.7 eV) [29] as well as ZnSO 4 (1023.0 eV) [30]. This implies that the lost electrons were transferred from Ru to Zn 2+ . Furthermore, Figure 5e shows that the electron-binding energy of S 2p is 169.6 eV, indicating that S species mainly exist as SO 4 2− [31]. Struijk et al. [32] also found that the electron-binding energy of S 2p on the surface of Ru catalyst after hydrogenation in ZnSO 4 was 169.9 eV, and S species were in the form of SO 4 2− . can be observed from Figure 5b, the electron-binding energy of Ru 3p3/2 over Ru-4.7 AH is 462.5 eV, indicating the existence of metallic Ru as well [26]. It is worth noting that, since the electron-binding energy of Ru-4.7 AH is 1.3 eV higher than that observed over Ru-4.7, it can be concluded that Ru δ+ species of electron deficiency is generated during the catalytic experiment. Additionally, the deconvoluted electron-binding energy of Zn 2p2/3 and the Auger electron kinetic energy of Zn·LMM was observed at 1022.0 eV and 988.6 eV over Ru-4.7 AH, respectively (Figure 5c,d), suggesting that Zn species mainly exist as Zn 2+ after reaction [27,28]. In addition, it is observed that the electron-binding energy of Zn 2p2/3 (1022.0 eV) is lower than that obtained over Zn(OH)2 (1022.7 eV) [29] as well as ZnSO4 (1023.0 eV) [30]. This implies that the lost electrons were transferred from Ru to Zn 2+ . Furthermore, Figure 5e shows that the electron-binding energy of S 2p is 169.6 eV, indicating that S species mainly exist as SO4 2- [31]. Struijk et al. [32] also found that the electron-binding energy of S 2p on the surface of Ru catalyst after hydrogenation in ZnSO4 was 169.9 eV, and S species were in the form of SO4 2− .  Figure 6 exhibits the catalytic activity towards cyclohexene formation from selective hydrogenation of benzene over Ru-4.7 catalyst with the different molar ratio of ZnO to Ru in the presence of ZnSO4. It is obvious that the catalytic activity towards benzene conversion declines when increasing the added ZnO content. Meanwhile, in contrast, the selectivity towards cyclohexene formation grows. In addition, cyclohexene yield of 60.4% was obtained when the molar ratio of ZnO to Ru is 0.19, which is one of the highest yields of cyclohexene reported by far [33][34][35]. Furthermore, as can be observed in Figure 7, no cyclohexene was detected without any additives. In comparison, 2.1% and 33% of cyclohexene yield was achieved when only ZnO or only ZnSO4 was added, respectively.  Figure 6 exhibits the catalytic activity towards cyclohexene formation from selective hydrogenation of benzene over Ru-4.7 catalyst with the different molar ratio of ZnO to Ru in the presence of ZnSO 4 . It is obvious that the catalytic activity towards benzene conversion declines when increasing the added ZnO content. Meanwhile, in contrast, the selectivity towards cyclohexene formation grows. In addition, cyclohexene yield of 60.4% was obtained when the molar ratio of ZnO to Ru is 0.19, which is one of the highest yields of cyclohexene reported by far [33][34][35]. Furthermore, as can  Figure 7, no cyclohexene was detected without any additives. In comparison, 2.1% and 33% of cyclohexene yield was achieved when only ZnO or only ZnSO 4 was added, respectively.  Combined with catalytic experimental results as well as the characterization outcomes, it can be concluded that the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt plays a key role in the synthesis of cyclohexene from selective hydrogenation of benzene. The significance of (Zn(OH)2)3(ZnSO4)(H2O)3 salt can be ascribed as follows: (1) Zn 2+ of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt on the Ru surface could transfer the electrons from Ru to Zn, which leads to generating the Ru δ+ species of electron deficiency. It has been reported that the generated Ru δ+ species of electron deficiency has a weaker ability to absorb cyclohexene, resulting in the faster desorption of the formed cyclohexene to avoid the further hydrogenation of cyclohexene [36]; (2) A stable complex could be formed between Zn 2+ of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt on the Ru surface and cyclohexene via hydrogen bonds, achieving the similar effect of the electron modification to increase the selectivity to cyclohexene [37]; (3) Zn 2+ of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt might selectively cover some strong active sites of Ru which are not suitable for the synthesis of cyclohexene, thus more benzene and cyclohexene could be adsorbed on the weaker active sites. These bring about the increase of selectivity to cyclohexene as well as the decline of the catalytic activity towards benzene conversion [32]; (4) Hydrophilicity of Ru surface could be significantly improved by the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt, resulting in the formation of a stagnant water layer. Since the solubility of cyclohexene is much lower than that of benzene, desorption of cyclohexene  Combined with catalytic experimental results as well as the characterization outcomes, it can be concluded that the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt plays a key role in the synthesis of cyclohexene from selective hydrogenation of benzene. The significance of (Zn(OH)2)3(ZnSO4)(H2O)3 salt can be ascribed as follows: (1) Zn 2+ of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt on the Ru surface could transfer the electrons from Ru to Zn, which leads to generating the Ru δ+ species of electron deficiency. It has been reported that the generated Ru δ+ species of electron deficiency has a weaker ability to absorb cyclohexene, resulting in the faster desorption of the formed cyclohexene to avoid the further hydrogenation of cyclohexene [36]; (2) A stable complex could be formed between Zn 2+ of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt on the Ru surface and cyclohexene via hydrogen bonds, achieving the similar effect of the electron modification to increase the selectivity to cyclohexene [37]; (3) Zn 2+ of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt might selectively cover some strong active sites of Ru which are not suitable for the synthesis of cyclohexene, thus more benzene and cyclohexene could be adsorbed on the weaker active sites. These bring about the increase of selectivity to cyclohexene as well as the decline of the catalytic activity towards benzene conversion [32]; (4) Hydrophilicity of Ru surface could be significantly improved by the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt, resulting in the formation of a stagnant water layer. Since the solubility of cyclohexene is much lower than that of benzene, desorption of cyclohexene Combined with catalytic experimental results as well as the characterization outcomes, it can be concluded that the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt plays a key role in the synthesis of cyclohexene from selective hydrogenation of benzene. The significance of (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt can be ascribed as follows: (1) Zn 2+ of the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt on the Ru surface could transfer the electrons from Ru to Zn, which leads to generating the Ru δ+ species of electron deficiency. It has been reported that the generated Ru δ+ species of electron deficiency has a weaker ability to absorb cyclohexene, resulting in the faster desorption of the formed cyclohexene to avoid the further hydrogenation of cyclohexene [36]; (2) A stable complex could be formed between Zn 2+ of the chemisorbed (Zn(OH) 2  more benzene and cyclohexene could be adsorbed on the weaker active sites. These bring about the increase of selectivity to cyclohexene as well as the decline of the catalytic activity towards benzene conversion [32]; (4) Hydrophilicity of Ru surface could be significantly improved by the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt, resulting in the formation of a stagnant water layer. Since the solubility of cyclohexene is much lower than that of benzene, desorption of cyclohexene from the catalyst surface would be obviously increased [32]. Therefore, a decrease of the benzene conversion as well as an increase of cyclohexene selectivity were observed when a higher amount of the (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt was adsorbed on the Ru surface. Based on the observations, a plausible sketch of surface structure of Ru-4.7 catalysts with the different additives was demonstrated in Figure 8. Without any additives ( Figure 8A), no Zn 2+ was chemisorbed on the Ru surface, resulting in the complete hydrogenation of benzene to cyclohexane. By applying ZnO as an additive ( Figure 8B), the adsorbed Zn was mainly stick shape of ZnO which cannot be dispersed on the Ru surface and no electrons could be transferred from Ru to Zn. Only 2.1% of maximum cyclohexene was, hence, obtained when only ZnO was added. However, when ZnSO 4 was utilized as the unique additive, a small quantity of (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt was generated due to the hydrolysis of ZnSO 4 under the reaction condition ( Figure 8C). In addition, the (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt could be chemisorbed and dispersed on the Ru surface, leading to a 33% of maximum cyclohexene yield. In addition, the amount of the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt rises when increasing the ZnO content when both ZnO and ZnSO 4 are present ( Figure 8D). Notably, even (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt is the essence of the synthesis of cyclohexene, and there is an optimum content for that since the enhancement of the cyclohexene selectivity is accompanied by the drop of the catalytic activity towards benzene conversion. Therefore, the highest yield of cyclohexene (60.4%) was achieved when the molar ratio of ZnO to Ru reached 0.19. from the catalyst surface would be obviously increased [32]. Therefore, a decrease of the benzene conversion as well as an increase of cyclohexene selectivity were observed when a higher amount of the (Zn(OH)2)3(ZnSO4)(H2O)3 salt was adsorbed on the Ru surface. Based on the observations, a plausible sketch of surface structure of Ru-4.7 catalysts with the different additives was demonstrated in Figure 8. Without any additives ( Figure 8A), no Zn 2+ was chemisorbed on the Ru surface, resulting in the complete hydrogenation of benzene to cyclohexane. By applying ZnO as an additive ( Figure 8B), the adsorbed Zn was mainly stick shape of ZnO which cannot be dispersed on the Ru surface and no electrons could be transferred from Ru to Zn. Only 2.1% of maximum cyclohexene was, hence, obtained when only ZnO was added. However, when ZnSO4 was utilized as the unique additive, a small quantity of (Zn(OH)2)3(ZnSO4)(H2O)3 salt was generated due to the hydrolysis of ZnSO4 under the reaction condition ( Figure 8C). In addition, the (Zn(OH)2)3(ZnSO4)(H2O)3 salt could be chemisorbed and dispersed on the Ru surface, leading to a 33% of maximum cyclohexene yield. In addition, the amount of the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt rises when increasing the ZnO content when both ZnO and ZnSO4 are present ( Figure 8D). Notably, even (Zn(OH)2)3(ZnSO4)(H2O)3 salt is the essence of the synthesis of cyclohexene, and there is an optimum content for that since the enhancement of the cyclohexene selectivity is accompanied by the drop of the catalytic activity towards benzene conversion. Therefore, the highest yield of cyclohexene (60.4%) was achieved when the molar ratio of ZnO to Ru reached 0.19.

Effects of Particle Size of Ru Catalysts
Monometallic Ru catalysts with different particle size before and after catalytic experiments were analyzed by XRD (Figure 9a,b). As can be seen from Figure 9a, characteristic diffractions related to metallic Ru were observed over all measured samples, indicating that Ru exists mainly in a metallic state. Additionally, the particle size of Ru from top to bottom was calculated to be 5.6, 4.7, 4.1 and 3.5 nm via Scherrer equation, respectively. Furthermore, the intensity of diffractions at 44° decreases from top to bottom, indicating the decline of the particle size as well. On the other hand, Figure 9b exhibits the XRD patterns of Ru catalysts with different particle size after the reaction. As

Effects of Particle Size of Ru Catalysts
Monometallic Ru catalysts with different particle size before and after catalytic experiments were analyzed by XRD (Figure 9a,b). As can be seen from Figure 9a, characteristic diffractions related to metallic Ru were observed over all measured samples, indicating that Ru exists mainly in a metallic state. Additionally, the particle size of Ru from top to bottom was calculated to be 5.6, 4.7, 4.1 and 3.5 nm via Scherrer equation, respectively. Furthermore, the intensity of diffractions at 44 • decreases from top to bottom, indicating the decline of the particle size as well. On the other hand, Figure 9b exhibits the XRD patterns of Ru catalysts with different particle size after the reaction. As can be observed, characteristic diffractions related to metallic Ru were also shown over all samples, revealing that Ru exists in a metallic state as well after catalytic experiments. In addition, the particle size of Ru from top to bottom was calculated to be 5.9, 4.7, 4.2 and 3.6 nm respectively, according to the Scherrer equation. These demonstrated that the Ru particle size did not change very much during the reaction process. However, several new diffractions were noticed in Figure 9b, which correspond to (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt. This can be rationalized by observing that (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 was generated from the reaction between ZnO and ZnSO 4 and was then chemisorbed on the Ru catalysts surface. In addition, it was also displayed that the intensity of (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 gradually increased with increasing particle size of Ru catalysts, demonstrating that (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 started to accumulate on the catalysts surface when Ru particle size increased from 3.5 to 5.6 nm. revealing that Ru exists in a metallic state as well after catalytic experiments. In addition, the particle size of Ru from top to bottom was calculated to be 5.9, 4.7, 4.2 and 3.6 nm respectively, according to the Scherrer equation. These demonstrated that the Ru particle size did not change very much during the reaction process. However, several new diffractions were noticed in Figure 9b, which correspond to (Zn(OH)2)3(ZnSO4)(H2O)3 salt. This can be rationalized by observing that (Zn(OH)2)3(ZnSO4)(H2O)3 was generated from the reaction between ZnO and ZnSO4 and was then chemisorbed on the Ru catalysts surface. In addition, it was also displayed that the intensity of (Zn(OH)2)3(ZnSO4)(H2O)3 gradually increased with increasing particle size of Ru catalysts, demonstrating that (Zn(OH)2)3(ZnSO4)(H2O)3 started to accumulate on the catalysts surface when Ru particle size increased from 3.5 to 5.6 nm.  Figure 10 gives the HTEM images, EDS, and particle size distribution of the Ru catalysts with the different particle sizes after catalytic experiments. All Ru catalysts display circular or elliptical shape. Furthermore, particle size of Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH is mainly distributed at 3.4, 4.0, 4.9 and 5.4 nm, respectively, which is consistent with the results obtained from XRD. In addition, as can be seen from the EDS results, the molar ration of Zn to Ru on the catalysts surface increased from 0.3 (1.44/44.46) over Ru-3.5 AH to 0.7 (3.23/47.02) over Ru-3.5 AH. This is attributed to the fact that (Zn(OH)2)3(ZnSO4)(H2O)3 salt starts to accumulate, which is in a good agreement with XRD results.
Textural properties of the Ru catalysts with different particle sizes before and after catalytic experiments are listed in Table 2. Decrease of the specific surface area and pore volume as well as an increase of pore diameter was observed with increasing particle size of Ru catalysts for both before and after catalytic experiments. Moreover, it is noticed that the specific surface area (e.g., 70 m 2 g -1 over Ru-3.5 AH) and the pore volume (e.g., 0.15 cm 3 g -1 over Ru-3.5 AH) after reaction is lower in comparison to that obtained over fresh catalysts (e.g., 78 m 2 g -1 of specific surface area and 0.17 cm 3 g -1 of pore volume over Ru-3.5). The pore diameter after catalytic experiments, however, is higher than that observed before catalytic experiments (e.g., 6.16 nm over Ru-3.5 AH vs. 5.48 nm over Ru-3.5). This might be because parts of the pores were blocked by the chemisorbed (Zn(OH)2)3(ZnSO4)(H2O)3 salt.
The catalytic activity towards benzene conversion and selectivity for cyclohexene formation over Ru catalysts with different particle size in the presence of ZnO and ZnSO4 was demonstrated in Figure 11a-c. An obvious increase of catalytic activity towards benzene conversion as well as a decrease of cyclohexene selectivity was observed with increasing the particle size of Ru catalysts from 3.5 nm to 5.6 nm. Furthermore, the maximum yield (Maxyield) of cyclohexene achieved over Ru  Figure 10 gives the HTEM images, EDS, and particle size distribution of the Ru catalysts with the different particle sizes after catalytic experiments. All Ru catalysts display circular or elliptical shape. Furthermore, particle size of Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH is mainly distributed at 3.4, 4.0, 4.9 and 5.4 nm, respectively, which is consistent with the results obtained from XRD. In addition, as can be seen from the EDS results, the molar ration of Zn to Ru on the catalysts surface increased from 0.3 (1.44/44.46) over Ru-3.5 AH to 0.7 (3.23/47.02) over Ru-3.5 AH. This is attributed to the fact that (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt starts to accumulate, which is in a good agreement with XRD results.
Textural properties of the Ru catalysts with different particle sizes before and after catalytic experiments are listed in Table 2. Decrease of the specific surface area and pore volume as well as an increase of pore diameter was observed with increasing particle size of Ru catalysts for both before and after catalytic experiments. Moreover, it is noticed that the specific surface area (e.g., 70 m 2 g −1 over Ru-3.5 AH) and the pore volume (e.g., 0.15 cm 3 g −1 over Ru-3.5 AH) after reaction is lower in comparison to that obtained over fresh catalysts (e.g., 78 m 2 g −1 of specific surface area and 0.17 cm 3 g −1 of pore volume over Ru-3.5). The pore diameter after catalytic experiments, however, is higher than that observed before catalytic experiments (e.g., 6.16 nm over Ru-3.5 AH vs. 5.48 nm over Ru-3.5). This might be because parts of the pores were blocked by the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 salt. The catalytic activity towards benzene conversion and selectivity for cyclohexene formation over Ru catalysts with different particle size in the presence of ZnO and ZnSO 4 was demonstrated in Figure 11a-c. An obvious increase of catalytic activity towards benzene conversion as well as a decrease of cyclohexene selectivity was observed with increasing the particle size of Ru catalysts from 3.5 nm to 5.6 nm. Furthermore, the maximum yield (Max yield ) of cyclohexene achieved over Ru catalysts with different particle size was displayed in Figure 11d. It can be seen that the highest maximum yield of cyclohexene of 60.4% was given over Ru-4.7 catalyst, and the maximum cyclohexene yield obtained over Ru catalysts shows a volcanic-type variation when increasing the particle size of Ru from 3.6 to 5.6 nm. Therefore, in this work, it can be concluded that there is an obvious particle size effect for the catalytic selective hydrogenation of benzene to cyclohexene over the Ru catalysts in the presence of ZnO and ZnSO 4 . Similar results were reported by Bu et al. [20] and Sun et al. [38] over Ru-Ba catalysts and Ru-Zn catalysts, respectively. particle size of Ru from 3.6 to 5.6 nm. Therefore, in this work, it can be concluded that there is an obvious particle size effect for the catalytic selective hydrogenation of benzene to cyclohexene over the Ru catalysts in the presence of ZnO and ZnSO4. Similar results were reported by Bu et al. [20] and Sun et al. [38] over Ru-Ba catalysts and Ru-Zn catalysts, respectively. Figure 10. HRTEM images, EDS, and particle size distribution of the Ru catalysts with the different particle size after hydrogenation. First row (a-c), second row (d-f), third row (g-i) and last row (j-l) belongs to Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH, respectively. Table 2. Textural properties of the Ru catalysts with the different particle size before and after catalytic experiments, including specific surface area (SBET), pore volume (Vp) and pore diameter (dp).

Catalyst
SBET/m 2 g −1 V p/cm 3 g −1 d p/nm Ru-3.5 78 0.17 5.48 Figure 10. HRTEM images, EDS, and particle size distribution of the Ru catalysts with the different particle size after hydrogenation. First row (a-c), second row (d-f), third row (g-i) and last row (j-l) belongs to Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH, respectively.  The reaction course of selective hydrogenation of benzene over the Ru-4.7 catalyst under the optimized condition is illustrated in Figure 12. As can be observed, the yield of cyclohexene increased in the first 25 min and the highest cyclohexene yield of 60.4% was achieved at 25 min. Then the yield of cyclohexene starts to decrease while cyclohexane yield keeps increasing during the whole reaction, suggesting that hydrogenation of benzene is a typical continuous reaction. However, it is worth mentioning that the rate of cyclohexene yield declining is rather slow (49.5% of cyclohexene yield was obtained when 98.1% of benzene conversion was observed at 60 min), indicating that the further hydrogenation of cyclohexene to cyclohexane is effectively retarded. The reaction course of selective hydrogenation of benzene over the Ru-4.7 catalyst under the optimized condition is illustrated in Figure 12. As can be observed, the yield of cyclohexene increased in the first 25 min and the highest cyclohexene yield of 60.4% was achieved at 25 min. Then the yield of cyclohexene starts to decrease while cyclohexane yield keeps increasing during the whole reaction, suggesting that hydrogenation of benzene is a typical continuous reaction. However, it is worth mentioning that the rate of cyclohexene yield declining is rather slow (49.5% of cyclohexene yield was obtained when 98.1% of benzene conversion was observed at 60 min), indicating that the further hydrogenation of cyclohexene to cyclohexane is effectively retarded.
To obtain a deeper understanding of the mechanism regarding how the particle size of Ru affects the cyclohexene formation, Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH were tested for selective hydrogenation of benzene to cyclohexene without utilizing ZnO and ZnSO 4 ·7H 2 O as well ( Figure 13). Complete conversion of benzene towards cyclohexane generation was observed over all tested Ru catalysts (2 g) within 5 min of reaction time, and no cyclohexene was detected. Furthermore, the hydrogenation of benzene was conducted over 1 g of the catalysts. Although the decline of the catalytic activity towards benzene conversion emerged, cyclohexene was still not detected. This demonstrates that there is no particle size effect of the Ru catalysts for selective hydrogenation of benzene to cyclohexene without any additives, which is confirmed by Milone et al. as well [19]. Therefore, based on aforementioned observations, how the particle size of Ru catalysts affects the catalytic activity towards cyclohexene formation can be deemed as a result of the following two reasons: firstly, the content of (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 on the Ru catalysts surface increased with increasing the particle size of Ru, leading to a raise of the surface Zn/Ru molar ratio. This results in more electrons being transferred from Ru to Zn 2+ , then more Ru δ+ species of electron deficiency being generated during the catalytic experiments. It has been reported that Ru δ+ species of electron deficiency could enhance the desorption of cyclohexene from Ru catalysts surface to retard the further hydrogenation of cyclohexane to cyclohexene, resulting in an increase of the selectivity to cyclohexene [36]. Another reason is that the pore width increases when raising the particle size of Ru catalysts, benefitting the desorption of cyclohexene from the pores of Ru catalysts. This could drastically suppress the further hydrogenation of cyclohexane to cyclohexene [39]. However, a decrease of catalytic activity towards benzene conversion was observed when increasing the particle size of Ru catalysts as well as the selectivity to cyclohexene. Thus, the highest cyclohexene yield was obtained over Ru-4.7 with considerably high benzene conversion (78.3%) and selectivity to cyclohexene (77.1%) after 25 min of reaction time. The reaction course of selective hydrogenation of benzene over the Ru-4.7 catalyst under the optimized condition is illustrated in Figure 12. As can be observed, the yield of cyclohexene increased in the first 25 min and the highest cyclohexene yield of 60.4% was achieved at 25 min. Then the yield of cyclohexene starts to decrease while cyclohexane yield keeps increasing during the whole reaction, suggesting that hydrogenation of benzene is a typical continuous reaction. However, it is worth mentioning that the rate of cyclohexene yield declining is rather slow (49.5% of cyclohexene yield was obtained when 98.1% of benzene conversion was observed at 60 min), indicating that the further hydrogenation of cyclohexene to cyclohexane is effectively retarded. To obtain a deeper understanding of the mechanism regarding how the particle size of Ru affects the cyclohexene formation, Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH were tested for selective hydrogenation of benzene to cyclohexene without utilizing ZnO and ZnSO4·7H2O as well ( Figure 13). Complete conversion of benzene towards cyclohexane generation was observed over all tested Ru catalysts (2 g) within 5 min of reaction time, and no cyclohexene was detected. Furthermore, the hydrogenation of benzene was conducted over 1 g of the catalysts. Although the decline of the catalytic activity towards benzene conversion emerged, cyclohexene was still not detected. This demonstrates that there is no particle size effect of the Ru catalysts for selective hydrogenation of benzene to cyclohexene without any additives, which is confirmed by Milone et al. as well [19]. Therefore, based on aforementioned observations, how the particle size of Ru catalysts affects the catalytic activity towards cyclohexene formation can be deemed as a result of the following two reasons: firstly, the content of (Zn(OH)2)3(ZnSO4)(H2O)3 on the Ru catalysts surface increased with increasing the particle size of Ru, leading to a raise of the surface Zn/Ru molar ratio. This results in more electrons being transferred from Ru to Zn 2+ , then more Ru δ+ species of electron deficiency being generated during the catalytic experiments. It has been reported that Ru δ+ species of electron deficiency could enhance the desorption of cyclohexene from Ru catalysts surface to retard the further hydrogenation of cyclohexane to cyclohexene, resulting in an increase of the selectivity to cyclohexene [36]. Another reason is that the pore width increases when raising the particle size of Ru catalysts, benefitting the desorption of cyclohexene from the pores of Ru catalysts. This could drastically suppress the further hydrogenation of cyclohexane to cyclohexene [39]. However, a decrease of catalytic activity towards benzene conversion was observed when increasing the particle size of Ru catalysts as well as the selectivity to cyclohexene. Thus, the highest cyclohexene yield was obtained over Ru-4.7 with considerably high benzene conversion (78.3%) and selectivity to cyclohexene (77.1%) after 25 min of reaction time.

Reusability of Ru-4.7 Catalyst
The reusability of Ru-4.7 catalyst was tested under the same reaction conditions without further regeneration ( Figure 14). No obvious deactivation was found after 10 catalytic experiments, and the Ru-4.7 catalyst gives at least 76.8% of benzene conversion and 75.4% selectivity to cyclohexene. The maximum yield of cyclohexene remains above 59.6% after 25 min of reaction time. This indicates

Reusability of Ru-4.7 Catalyst
The reusability of Ru-4.7 catalyst was tested under the same reaction conditions without further regeneration ( Figure 14). No obvious deactivation was found after 10 catalytic experiments, and the Ru-4.7 catalyst gives at least 76.8% of benzene conversion and 75.4% selectivity to cyclohexene. The maximum yield of cyclohexene remains above 59.6% after 25 min of reaction time. This indicates that the Ru-4.7 catalyst possesses a great potential for industrial application on selective hydrogenation of benzene towards cyclohexene production.

Chemicals
RuCl3·3H2O was commercially obtained from Sino-Platinum Co. Ltd. (Kunming, China). ZnSO4·7H2O was purchased from Fuchen Chemical Reagent Factory (Tianjin, China). NaOH and benzene were delivered from Kemiou Chemical Reagent Co. Ltd. (Tianjin, China). ZnO was bought from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). All chemicals were of analytical grade without further purification and distilled water was applied in all experiments.

Preparation of Catalysts
Monometallic Ru catalysts were prepared as follows: 10.0 g RuCl3·3H2O precursor was dissolved in 200 cm 3 of distilled water. Then 200 cm 3 of 5 wt % NaOH aqueous solution was added at 353 K with continuous stirring for 2 h, followed by a reduction procedure in a 1000 cm 3 Hastelloy autoclave under 5.0 MPa of H2 and a stirring speed of 800 min -1 at 373 K, 393 K, 413 K and 433 K for 5 h, respectively. Neither drying nor calcination procedure were conducted before the reduction of catalysts. In addition, it should be noticed that the particle size of Ru catalysts was obtained via X-ray diffraction method, which was calculated to be 3.5 nm, 4.1 nm, 4.7 nm and 5.6 nm, respectively. To be clarified, the Ru catalysts synthesized at different reduction temperature were denoted as Ru-3.5, Ru-4.1, Ru-4.7 and Ru-5.6, respectively.

Catalytic Experimental Procedure
All catalytic experiments were carried out in a GS-1 type Hastelloy autoclave of 1000 cm 3 . In a typical selective hydrogenation of benzene to cyclohexene reaction, 2.0 g monometallic Ru catalyst, 50.0 g ZnSO4·7H2O as well as 280 cm 3 distilled water were added in the autoclave. Then the reactor was purified using N2 for 4 times and followed by purification with H2 for another 4 times as well. After that, the reactor was heated up to 423 K under 5.0 MPa of H2 with a stirring speed of 800 min -1 , followed by adding 140 cm 3 benzene and adjusting the stirring speed to 1400 min -1 to eliminate the mass transfer limitation [18]. Subsequently, the liquid samples were taken periodically from the reactor every 5 min. All withdrawn samples were analyzed by gas chromatography from Hangzhou

Preparation of Catalysts
Monometallic Ru catalysts were prepared as follows: 10.0 g RuCl 3 ·3H 2 O precursor was dissolved in 200 cm 3 of distilled water. Then 200 cm 3 of 5 wt % NaOH aqueous solution was added at 353 K with continuous stirring for 2 h, followed by a reduction procedure in a 1000 cm 3 Hastelloy autoclave under 5.0 MPa of H 2 and a stirring speed of 800 min −1 at 373 K, 393 K, 413 K and 433 K for 5 h, respectively. Neither drying nor calcination procedure were conducted before the reduction of catalysts. In addition, it should be noticed that the particle size of Ru catalysts was obtained via X-ray diffraction method, which was calculated to be 3.5 nm, 4.1 nm, 4.7 nm and 5.6 nm, respectively. To be clarified, the Ru catalysts synthesized at different reduction temperature were denoted as Ru-3.5, Ru-4.1, Ru-4.7 and Ru-5.6, respectively.

Catalytic Experimental Procedure
All catalytic experiments were carried out in a GS-1 type Hastelloy autoclave of 1000 cm 3 . In a typical selective hydrogenation of benzene to cyclohexene reaction, 2.0 g monometallic Ru catalyst, 50.0 g ZnSO 4 ·7H 2 O as well as 280 cm 3 distilled water were added in the autoclave. Then the reactor was purified using N 2 for 4 times and followed by purification with H 2 for another 4 times as well. After that, the reactor was heated up to 423 K under 5.0 MPa of H 2 with a stirring speed of 800 min −1 , followed by adding 140 cm 3 benzene and adjusting the stirring speed to 1400 min −1 to eliminate the mass transfer limitation [18]. Subsequently, the liquid samples were taken periodically from the reactor every 5 min. All withdrawn samples were analyzed by gas chromatography from Hangzhou Kexiao Chemical Instrument and Equipment Co. Ltd. (Hangzhou, China) equipped with a flame ionization detector (FID). It is worth mentioning that all reagents and products (e.g., benzene, cyclohexene and cyclohexane) were calibrated, and correlation coefficient (R 2 ) of all compounds is no less than 0.99. In addition, the benzene conversion and selectivity towards cyclohexene were calculated with the obtained peak area with the area normalization method. After each reaction, the organic phase was removed via a separating funnel and the solid Ru catalyst together with the aqueous solution were added in the autoclave again in order to investigate the reusability of the catalysts via aforementioned experimental procedure. The Ru-3.5, Ru-4.1, Ru-4.7 and Ru-5.6 catalysts after hydrogenation reaction were denoted as Ru-3.5 AH, Ru-4.1 AH, Ru-4.7 AH and Ru-5.6 AH, respectively. To avoid Zn 2+ impacting the characterization results, catalysts after the reaction were filtered and washed until the filtrate become neutralization and no Zn 2+ was detected. Then solid samples were dried in Ar flow at 373 K and stored in the ethanol for further characterization.

Catalysts Characterization
X-ray diffraction (XRD) patterns for the fresh and spent catalysts were recorded at room temperature using an X'Pert Pro instrument from Philips (Almelo, The Netherlands). The diffracted intensity of Cu-Kα radiation (λ = 0.154 nm) was measured at the range of 2θ between 5 • and 90 • , with a step size of 0.03 • . Non-reduced solid Ru samples were pretreated under Ar atmosphere at 373 K for 6 h, which was further used for Ru content analysis and reducibility degree analysis, respectively. The Ru content was obtained by atomic absorption spectrometer (AAS) using a Perkin Elmer AAnalyst 300 from America (PerkinElmer, Waltham, MA, USA). In addition, temperature programmed reduction (TPR) was conducted with an Autosorb-IQ from Quantachrome (Boynton Beach, FL, USA). Typically, 10 mg of dried sample was firstly treated under an Ar flow for 2 h at room temperature. Then an Ar stream containing 10 Vol % H 2 was introduced instead (30 cm 3 min −1 ) while being heated till 573 K (10 K min −1 ) and held for 1 h. The hydrogen consumption was recorded and determined by using a standard CuO calibration. In addition, elemental analysis was obtained via X ray Fluorescence (XRF) using a S4 Pioneer instrument from Bruker AXS (Karlsruhe, Germany). Moreover, the valence state of Ru and Zn on the catalyst surface was analyzed by X-ray photoelectron spectroscopy (XPS) using a PHI Quantera SXM instrument from Ulvac-Phi (Chigasaki, Japan). Al K α (E b = 1486.6 eV) was chosen as the source of radiation and vacuum degree was adjusted to be 6.7 × 10 −8 Pa. The energy scale was calibrated and corrected for charging using the C1s (E b = 284.8 eV) line as the binding energy reference. The sample in the ethanol was supersonically pretreated for 30 min and dried in Ar. Then the powder was placed on the conducting resin for SEM analysis. The catalyst surface element scanning was tested by a ∑IGMA scanning electron microscope (SEM) from Carl Zeiss AG (Jena, Germany). The sample in the ethanol was pretreated using supersonic for 30 min as well, and then placed on a Cu board for TEM analysis. Furthermore, JEOL JEM 2100 transmission electron microscope (TEM) combined with energy dispersive spectrometer (EDS) was utilized to identify the surface composition of the selected samples. Textural properties were analyzed by the Nova 1000 e-Physisorption Analyzer (Boynton Beach, FL, USA). All the samples were evacuated at 523 K under the vacuum pressure for 2 h prior to the measurements and the isotherms were taken at 77 K. The specific surface area (S BET ) was determined by the Brunauer-Emmett-Teller (BET) model. To investigate the wettability of the catalyst surface, water contact angle values (CAs) were measured by a contact angle meter (JC2000 C1, Powereach, Shanghai, China) at ambient temperature for each sample.

Conclusions
(Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 could be generated from ZnO and ZnSO 4 during catalytic experiments, which was then chemisorbed and uniformly dispersed on the surface of Ru-4.7 catalyst. It was demonstrated that the chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 plays the crucial role on improving the catalytic activity towards cyclohexene synthesis over Ru-4.7 catalyst, i.e., 60.4% of cyclohexene yield was obtained over Ru-4.7 catalyst when applying ZnO and ZnSO 4 as additives in comparison to that achieved over Ru-4.7 catalyst when only utilizing ZnO (2.1%) or ZnSO 4 (33.0%). More importantly, without adding ZnO and ZnSO 4 , complete conversion of benzene was observed within 5 min and no cyclohexene was detected over all tested Ru catalysts with different particle sizes. This suggested that there is no particle size effect on cyclohexene synthesis from selective hydrogenation of benzene over the pure monometallic Ru catalysts in the absence of ZnO and ZnSO 4 . On the other hand, when both ZnO and ZnSO 4 were utilized as the reaction additives, surface Zn 2+ /Ru molar ratio increased with increasing the particle size of the monometallic Ru catalysts after catalytic experiments, indicating that the content of chemisorbed (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 on Ru catalysts raised under such a circumstance, which was confirmed from XRD results as well. In addition, the maximum cyclohexene yield obtained over monometallic Ru catalysts showed a volcanic-type variation when increasing the particle size of Ru from 3.6 to 5.6 nm. Therefore, the generation of (Zn(OH) 2 ) 3 (ZnSO 4 )(H 2 O) 3 during the catalytic experiment was considered as the essence of the particle size effect over Ru catalysts on the selective hydrogenation of benzene to cyclohexene. The highest cyclohexene yield of 60.4% was achieved over Ru-4.7 with an optimum ZnO/Ru ratio of 0.19:1 in the presence of 0.62 mol·L −1 ZnSO 4 within 25 min of catalytic experiment at 423 K under 5.0 MPa of H 2 . In addition, no decrease of catalytic activity towards cyclohexene generation was observed over this catalyst after 10 catalytic experiments without any regeneration.
Author Contributions: Haijie Sun, Zhihao Chen and Zhikun Peng conceived and designed the experiments; Chenggang Li and Yan Li performed the experiments; Zhongyi Liu and Shouchang Liu analyzed the data; Lingxia Chen contributed reagents/materials/analysis tools; Haijie Sun wrote the paper.