Dynamic redox cycle of Cu⁰ and Cu⁺ over Cu/SiO₂ catalyst in ester hydrogenation

Abstract

Esters (methyl acetate, ethyl acetate, dimethyl oxalate) can oxidize Cu⁰ of the reduced Cu/SiO₂ catalyst to Cu⁺ and the obtained Cu⁺ can be reduced back to Cu⁰ by H₂ under reaction conditions. Cu⁰ and Cu⁺ are in a dynamic cycle during ester hydrogenation. The ratio of Cu⁺/Cu⁰ in the presence of esters under reaction conditions is probably different from that in the absence of esters under characterization conditions. The effect of the dynamic cycle on the Cu⁺/Cu⁰ ratio should be taken into consideration when establishing the relationship between the Cu⁺/Cu⁰ ratio and reaction performance.

---

Hydrogenation of esters to the corresponding alcohols is a fundamental reaction in organic chemistry and is employed in a large number of chemical processes. Of particular interest are the hydrogenation of diesters to diols, which are used in the production of polyesters, and the hydrogenation of acetic esters to the promising fuel ethanol. Cu/SiO₂ catalysts have been identified as active and selective catalysts for these reactions.¹ In order to gain better activity and stability, much work has been done to explore the nature of the active species for ester hydrogenation. The activity was frequently found to be proportional to the surface area of metallic copper and thus Cu⁰ was deemed as the sole active site.² Based on this, Cu⁰ activates both H₂ and the ester. However, it is reported that Cu⁺ can polarize the C=O bond of ester molecules and stabilize the intermediates formed during the reaction, implying that Cu⁺ could play a role in ester hydrogenation.³ Therefore, a synergetic effect of Cu⁰ and Cu⁺ for ester hydrogenation was advanced and proper ratio of Cu⁺/Cu⁰ is believed to benefit the catalytic performance of Cu/SiO₂ catalysts in addition to improved dispersion of copper species.⁴ Chen et al. reported that a Cu/SiO₂ catalyst prepared by ammonia evaporation method at 363 K with the highest Cu⁺/Cu⁰ ratio of 1.22 exhibited the best catalytic activity for the hydrogenation of dimethyl oxalate to ethylene glycol.⁵ Yin et al. also prepared Cu/SiO₂ catalysts by the same method and investigated the effect of the preparation temperature on the Cu⁺/Cu⁰ ratio.⁶ In contrast, they gained the best performance at their minimum Cu⁺/Cu⁰ ratio of 0.16 for the same reaction. In the published papers, the ratio of Cu⁺/Cu⁰ detected by XPS/XAES (X-ray Photoelectron Spectroscopy and X-ray Auger Electron Spectroscopy) before the reaction was adopted to establish the relationship with the reaction performance. However, the XPS/XAES analysis was exclusively carried out in high vacuum and without ester present. He et al. observed the Cu₂O phase in the spent Cu/SiO₂ catalyst, which did not appear in the fresh reduced catalyst.⁷ This hinted at the variation of the Cu⁺/Cu⁰ ratio during ester hydrogenation. Here we present proof that the esters (methyl acetate, ethyl acetate, dimethyl oxalate) can oxidize Cu⁰ to Cu⁺ and the obtained Cu⁺ can be reduced back to Cu⁰, implying a dynamic cycle of Cu⁰ and Cu⁺ during ester hydrogenation. The Cu⁺/Cu⁰ ratio in the presence of esters under reaction conditions is likely to be different from that in the absence of esters under characterization conditions. This may be the reason why different researchers observed diverse results about the effect of the Cu⁺/Cu⁰ ratio.^5,6

Fig. 1 shows TPR profiles of Cu/SiO₂ catalyst prepared by ammonia evaporation method (ESI†). With a heating rate of 10 K min⁻¹, the fresh calcined catalyst exhibited a broad reduction peak centered at 565 K. Ethyl acetate (EA) was introduced onto the catalyst after the reduction by a carrier gas of helium at 493 K, the temperature at which ester hydrogenation is usually carried out. After the treatment with EA, a second reduction demonstrated a smaller H₂-consumption peak at a relatively low temperature. This implies that EA oxidized Cu⁰ to its oxidation state, which showed similar reduction temperature to Cu₂O obtained from the oxidation of Cu⁰ in the reduced Cu/SiO₂ catalyst by N₂O (454 K vs. 458 K).

Fig. 1 TPR profiles of Cu/SiO₂ catalyst. (a) First reduction for the calcined catalyst, (b) second reduction after treatment with ethyl acetate.

XRD patterns of Cu/SiO₂ catalyst with different treatments are illustrated in Fig. 2. The broad diffuse peak at 2θ = 21.7° can be attributed to amorphous silica. No diffraction peak of copper species was detected in the calcined catalyst, indicating that the cupric compounds were in amorphous state. Fig. 2b implies that the characteristic diffraction peak of Cu⁰ at 2θ = 43.4° showed up after the reduction at 623 K (JCPDS 003-1015). Fig. 2c was for the catalyst suffering the reduction at 623 K and EA treatment at 493 K in series. Compared to Fig. 2b, the peak at 2θ = 43.4° abated drastically or even disappeared after the treatment with EA. Instead, two new peaks emerged at 2θ = 36.5° and 42.4°, respectively, characteristic of Cu₂O (JCPDS 065-3288). This suggests that EA oxidized Cu⁰ to Cu₂O, which is consistent with the TPR results. After EA oxidation, H₂ was introduced onto the catalyst to check the reducibility of the newly generated Cu₂O at 493 K. The peak of Cu₂O at 2θ = 36.5° decreased significantly after H₂-treatment while the peak of Cu⁰ at 2θ = 43.4° became obvious again (Fig. 2d). This demonstrated that Cu₂O derived from the oxidation of Cu⁰ by EA can be reduced back to Cu⁰ at 493 K. The highly dispersed metallic copper might be oxidized to CuO to some extent when exposure to air. The weak peak at about 2θ = 36.5° existing in the reduced catalyst in Fig. 2b should be ascribed to this type of CuO because it did not appear during the in situ XRD (Fig. S1, ESI†). This type of CuO may also make some contribution to the intensity of the peak at 2θ = 36.5° in Fig. 2c and d. However, we think its contribution was small. All in all, it can be drawn from the XRD and TPR results that the oxidation of Cu⁰ to Cu₂O by EA and the reduction of Cu₂O to Cu⁰ by H₂ occur simultaneously during EA hydrogenation. The existing formation of Cu⁺ during the reaction was determined by the catalyst system together with reactants and products. Cu⁺ in the copper chromite catalyst responsible for CO chemisorption in methanol synthesis from syngas exists as a crystalline CuCrO₂ phase.⁸ The active site Cu⁺ in ester hydrogenation seems to be in the form of Cu₂O.^6,7,9 The Cu₂O obtained from oxidation of Cu⁰ by EA in this work and the Cu⁺ species in the published papers should be same in essence. Therefore, the active sites Cu⁰ and Cu⁺ are in a dynamic redox cycle during EA hydrogenation. This dynamic cycle also exists during the hydrogenation of other esters, such as methyl acetate and dimethyl oxalate (Fig. S2, ESI†).

Fig. 2 XRD patterns of Cu/SiO₂ catalyst. Ethyl acetate (EA) was fed onto the catalyst by helium.

Fig. 3 shows the XPS/XAES spectra of the Cu/SiO₂ catalyst. The copper in the calcined catalyst existed in the oxidation state of Cu²⁺, as evidenced by the broad Cu 2p_3/2 peak at about 933 eV and the 2p to 3d satellite at 942–945 eV characteristic of Cu²⁺ with electron configuration of 3d⁹. For the catalyst reduced at 623 K, the binding energy of Cu 2p_3/2 shifted to 932.2 eV and the satellite disappeared, suggesting that the copper species became reduced (Cu⁰ or Cu⁺). Because the BE values of Cu⁰ and Cu⁺ are almost identical, the distinction between these two species has to be made on the basis of the XAES spectra. Two overlapping Cu LMM Auger kinetic energy peaks centered at about 918.8 eV and 914.2 eV can be ascribed to Cu⁰ and Cu⁺, respectively, impling the coexistence of Cu⁰ and Cu⁺ on the surface (2–10 nm) of the reduced catalyst. Cu⁰ and Cu⁺ of Cu/SiO₂ catalyst prepared by ammonia evaporation method were attributed to the reduction of highly dispersed CuO and copper phyllosilicate respectively in some papers and further reduction of this kind of Cu⁺ to Cu⁰ was believed to require a high temperature of 873 K.^5–7,9,10 However, the appointed reduction peak for the Cu⁺ was not observed in these papers although XPS/XAES gave a high Cu⁺ content of more than 50% sometimes. Copper phyllosilicate used as a reference compound in Van Der Grift's work did not show hydrogen consumption above 600 K in TPR process with a heating rate of 3.4 K min⁻¹.¹¹ Lin et al.¹² and Zheng et al.¹³ reported that the reduction of copper phyllosilicate yielded Cu nanocrystallites and increased amount of copper phyllosilicate boosted the dispersion of Cu⁰. It seems that the source of Cu⁺ is not clear. Deconvolution results of the XAES peak show that the proportion of surface Cu⁺ species was 21% for our 623 K reduced catalyst. If there was so much Cu⁺ in the bulk of the catalyst, there should be a reduction peak above 623 K for further reduction of Cu⁺. However, no other peak was obtained until 1200 K except the peak centered at about 565 K (Fig. S3, ESI†). In addition, the H₂ consumption implied that almost all the Cu²⁺ was reduced to Cu⁰ at 623 K. Therefore, it is reasonable to deduce that the freshly reduced catalyst contained only trace amount of Cu⁺, which may mainly exist on the surface of the catalyst. Its formation may be related to the special surface topographies such as terrace, kink or edge. Yet, the possibility of Cu⁺ coming from the reduction of copper phyllosilicate could not be excluded.

Fig. 3 XPS/XAES spectra of Cu/SiO₂ catalyst.

In general, the reaction rate is mainly determined by the surface active sites of the catalyst. Our results showed that Cu₂O derived from the oxidation of Cu⁰ by the esters yielded distinct XRD diffraction peaks and an obvious TPR H₂-consumption peak, suggesting that the Cu⁰ in the bulk of the catalyst can be reached and oxidized to Cu⁺ by the esters. Therefore, we tentatively proposed that the copper species in the bulk should be involved in ester hydrogenation in addition to the species on the surface. Herein, we attempted to combine the Cu⁺/Cu⁰ ratio of the entire catalyst to the reaction performance. Two methods were established by applying the TPO and TPR techniques to determine the Cu⁺/Cu⁰ ratio of the entire catalyst after the reaction (ESI†). The two methods gave identical results, which proved them feasible and reliable.

Fig. 4 shows the activity of Cu/SiO₂ catalyst for EA hydrogenation and the Cu⁺/Cu⁰ ratio of the spent catalyst. The Cu/SiO₂ catalyst gave more than 98% selectivity to ethanol (not shown here). The by-products were ethane, diethyl ether and acetaldehyde. Two experiments were carried out under same reaction conditions of 493 K, 0.1 MPa and H₂/EA ratio of 80, but terminated at different reaction times of 30 min and 9 h. The catalysts after 30 min and 9 h reaction possessed a Cu⁺/Cu⁰ ratio of 67/33 and 80/20 and exhibited 13% and 28% EA conversion, respectively. High pressure was beneficial for the reaction. Under reaction conditions of 493 K, 1.0 MPa and H₂/EA ratio of 80, the catalysts demonstrated 22% and 96% EA conversion at 20 min and 9 h, respectively, with a Cu⁺/Cu⁰ ratio of 49/51 and 60/40. Fig. 4 displays that EA conversion initially increased and then reached a stable level. Repeat experiments confirmed the rising period at the beginning but its duration varied with reaction conditions. The initial increase indicated an evolution of the active sites. At the start of the reaction, the catalyst possessed high Cu⁰ content, but exhibited a low EA conversion. After a few hours, EA conversion stabilized at a relatively high level, however the catalyst was in a low Cu⁰ content. We think that Cu⁺ and Cu⁰ attained a dynamic equilibrium at the moment. These results implied that EA conversion was not proportional to the content of Cu⁰ and a synergetic effect existed between Cu⁺ and Cu⁰ for EA hydrogenation. Therefore, the Cu⁺/Cu⁰ ratio is important for the catalytic performance of the catalyst. Promoters and preparation methods were reported to influence the reaction performance via regulating the Cu⁺/Cu⁰ ratio.¹⁴ In these papers, the Cu⁺/Cu⁰ ratio was detected by XPS/XAES without the presence of the reactants and it was just for the surface of the catalyst. Ester and H₂ are actors in the inter-conversion of Cu⁰ and Cu⁺. As excepted, high reaction pressure promoted the reduction of Cu⁺ to Cu⁰ and resulted in a lower ratio of Cu⁺/Cu⁰ of the entire catalyst (Fig. 4). It seems that the Cu⁺/Cu⁰ ratio of the entire catalyst was affected by the reaction conditions, which certainly can alter the surface Cu⁺/Cu⁰ ratio of the catalyst. We think that the relationship between the Cu⁺/Cu⁰ ratio and reaction performance should be estimated under operando conditions, no matter whether the active species in the bulk of the catalyst make some contribution to the reaction or not.

Fig. 4 Catalytic activity and Cu⁺/Cu⁰ ratio of Cu/SiO₂ catalyst in the hydrogenation of ethyl acetate (EA). Reaction conditions: (a) 493 K, 1.0 MPa, H₂/EA ratio = 80; (b) 493 K, 0.1 MPa, H₂/EA ratio = 80. The feed rate of EA was 3.3 μmol h⁻¹ g_cat⁻¹.

One ester molecule turns into two alcohol molecules according to the equation of ester hydrogenation. The two oxygen atoms in the ester molecule should be present in the alcohol molecules after the reaction. In this work, the transformation of Cu⁰ to Cu₂O should take some of the oxygen from the esters. If Cu₂O changed into water and Cu⁰ with the reduction of H₂, then significant amounts of water along with alkanes should be detected in the products. However, the amount of alkanes was very small. In addition, the stoichiometric yield of alcohol molecules per ester molecule was close to two. Therefore, the inter-conversion of Cu⁰ and Cu⁺ did not transform the oxygen of the esters into water. Although how esters and H₂ participate in the dynamic cycle of Cu⁰ and Cu⁺ is not clear, we think that esters and H₂ may be activated during the inter-conversion of Cu⁰ and Cu⁺ and then transformed into the target products.

In summary, Cu⁰ and Cu⁺ are in a dynamic redox cycle during the hydrogenation of esters such as methyl acetate, ethyl acetate and dimethyl oxalate. As the participators in inter-conversion of Cu⁰ and Cu⁺, ester and H₂ can shift the equilibrium of Cu⁰ and Cu⁺ and affect the Cu⁺/Cu⁰ ratio. Our results implied that the Cu⁺/Cu⁰ ratio is closely related to the reaction performance. Operando study using X-ray Absorption Fine Structure will be carried out to further explore the relationship between the Cu⁺/Cu⁰ ratio and reaction performance.

Acknowledgements

This work was financially supported by Energy Innovation Laboratory of BP company and National Natural Science Foundation of China (no. 21403215).

Notes and references

1. (a) J. Lin, X. Zhao, Y. Cui, H. Zhang and D. Liao, Chem. Commun., 2012, 48, 1177–1179; (b) C. Wen, Y. Cui, W. Dai, S. Xie and K. Fan, Chem. Commun., 2013, 49, 5195–5197; (c) D. S. Brands, E. K. Poels and A. Bliek, Appl. Catal., A, 1999, 184, 279–289.
2. (a) M. Mokhtar, C. Ohlinger, J. H. Schlander and T. Turek, Chem. Eng. Technol., 2001, 24, 423–426; (b) J. Gong, H. Yue, Y. Zhao, S. Zhao, L. Zhao, J. Lv, S. Wang and X. Ma, J. Am. Chem. Soc., 2012, 134, 13922–13925.
3. (a) E. K. Poels and D. S. Brands, Appl. Catal., A, 2000, 191, 83–96; (b) M. A. N. Santiago, M. A. Sanchez-Castillo, R. D. Cortright and J. A. Dumesic, J. Catal., 2000, 193, 16–28.
4. (a) A. Yin, C. Wen, X. Guo, W. Dai and K. Fan, J. Catal., 2011, 280, 77–88; (b) X. Ma, H. Chi, H. Yue, Y. Zhao, Y. Xu, J. Lv, S. Wang and J. Gong, AIChE J., 2013, 59, 2530–2539; (c) A. Yin, X. Guo, W. Dai, H. Li and K. Fan, Appl. Catal., A, 2008, 349, 91–99.
5. L. Chen, P. Guo, M. Qiao, S. Yan, H. Li, W. Shen, H. Xu and K. Fan, J. Catal., 2008, 257, 172–180.
6. A. Yin, X. Guo, W. Dai and K. Fan, Catal. Commun., 2011, 12, 412–416.
7. Z. He, H. Lin, P. He and Y. Yuan, J. Catal., 2011, 277, 54–63.
8. (a) G. Apai, J. R. Monnier and M. J. Hanrahan, Appl. Surf. Sci., 1984, 19, 307–314; (b) J. R. Monnier, M. J. Hanrahan and G. Apai, J. Catal., 1985, 92, 119–126.
9. A. Yin, X. Guo, W. Dai and K. Fan, J. Phys. Chem. C, 2009, 113, 11003–11013.
10. A. Yin, X. Guo, W. Dai and K. Fan, J. Phys. Chem. C, 2010, 114, 8523–8532.
11. C. J. G. Van Der Grift, A. Mulder and J. W. Geus, Appl. Catal., 1990, 60, 181–192.
12. H. Lin, X. Zheng, Z. He, J. Zheng, X. Duan and Y. Yuan, Appl. Catal., A, 2012, 445, 287–296.
13. X. Zheng, H. Lin, J. Zheng, X. Duan and Y. Yuan, ACS Catal., 2013, 3, 2738–2749.
14. (a) Z. Huang, F. Cui, H. Kang, J. Chen, X. Zhang and C. Xia, Chem. Mater., 2008, 20, 5090–5099; (b) L. Lin, P. Pan, Z. Zhou, Z. Li, J. Yang, M. Sun and Y. Yao, Chin. J. Catal., 2011, 32, 957–969; (c) C. Wen, Y. Cui, A. Yin, K. Fan and W. Dai, ChemCatChem, 2013, 5, 138–141.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and characterization data. See DOI: 10.1039/c5ra04389a

---