Methanol Synthesis from CO/CO2/H2 and CO/H2 over Cu/ZnO/Al2O3 and Cu/CeO2 Catalysts

1. Introduction

Methanol is an industrial chemical of high significance with an annual production of 62 million tons in 2014. Commercial production of methanol from carbon monoxide and hydrogen dates back nearly 100 years,[1] during which period significant improvements to the catalyst and process have been achieved. Methanol is currently produced from CO/CO₂/H₂ feed gas mixtures using a Cu/ZnO/Al₂O₃ catalyst which was developed by ICI in the 1960s.[1] The reaction mechanism and nature of the active site as a function of reaction conditions have been studied intensively and remain subject of active industrial and academic research. Based on numerous spectroscopic, isotope labelling and kinetic studies dedicated to study the mechanism and active site of this system, it has been established that CO₂ is the carbon source of methanol produced from CO/CO₂/H₂ feeds, where a linear correlation exists between Cu metal surface area and methanol synthesis activity. Whilst the presence of Cu is essential for the production of methanol, rates of reaction are enhanced significantly by the ZnO component in the catalyst. The correlation between Cu surface area and activity does not apply to all types of Cu catalyst however, and CO₂ is not the only possible carbon source: methanol formation over Cu/CeO₂ catalysts occurs via CO hydrogenation and involves a different active site and reaction mechanism.[2,3] In this contribution the generation of methanol synthesis active sites on different Cu-based catalysts is explored as a function of the feed gas composition, focusing on the interplay between metal, support and gas phase. Catalyst performance is evaluated under steady and non-steady state conditions using spatially resolved experimental techniques to assess catalyst performance in different catalyst bed sectors. The effect of reaction conditions on the active sites and prevailing reaction mechanism is studied by combining kinetic data with results from spectroscopic studies on catalysts measured under relevant reaction conditions.

2. Experimental

A commercial CuO/ZnO/Al₂O₃ (60/30/10 weight ratio) catalyst precursor and a 10 wt% CuO/CeO₂ mixed oxide, prepared by deposition precipitation of CuO onto a CeO₂ support,[3] were used throughout the study. Reaction studies were performed using a set of 6 parallel fixed-bed, down-flow micro-reactors. The oxidic CuO/ZnO/Al₂O₃ and CuO/CeO₂ catalyst precursors were reduced in dilute H₂ and subsequently exposed to relevant methanol synthesis conditions in the temperature range of 180 – 225 °C and pressures ranging from 10 to 35 bar. Gas compositions used were 0 - 3% CO / 0 – 3% CO₂ / 67% H₂ / N₂ for Cu/ZnO/Al₂O₃ and 2 or 33% CO / H₂ for Cu/CeO₂. Parallel difference testing [4,5] was used to obtain spatially resolved activity profiles throughout the catalyst beds. In situ DRIFTS spectra were recorded on a Thermo Nicolet 750 Magna FTIR, equipped with an environmental DRIFTS cell, operated at atmospheric pressure. Samples for XPS analysis were studied using a Thermo Fisher Escalab 250. The high-pressure gas cell was operated by introducing gas at ~3 bar under flow (~50 ml/minute at the exhaust). Cooling down of the samples was done under reactive gas flow, followed by evacuation of the cell and transfer of the sample to the analysis chamber.

 

3. Results and discussion

The exit gas composition measured during the first 40 minutes of exposure of the H₂/N₂ reduced Cu/ZnO/Al₂O₃ catalyst to 3%CO/3%CO₂/67%H₂/N₂ at 25 bar, 200 °C, reveals that the reverse water-gas shift (RWGS) reaction occurs from t = 0, whilst methanol synthesis is not seen until t = 2.5 min. Analysis of kinetic data obtained under the entire range of conditions indicates that the occurrence of RWGS during the first minutes of operation results in a build-up of O_s surface species. The detailed kinetic modeling study [5] revealed that a model including generation of bidentate surface carbonate species (CO₃.2s), formed upon reaction of CO₂ with O_s, gives the best fit of the experimental data. The effect of CO:CO₂ ratio in the feed on steady-state performance of Cu/ZnO/Al₂O₃ in different catalyst bed sectors is illustrated in Figure 1. The extent of the RWGS reaction is dependent on the CO₂ concentration in the feed, and RWGS equilibrium is reached at different stages along the length of the reactor, depending on the CO:CO₂ ratio. The results presented in the Figure illustrate that the methanol production is not affected by a change in CO:CO₂ ratio from 10 to 1.

 

 

Figure 1. Steady state axial concentration plots for Cu/ZnO/Al₂O₃ under 3%CO/0.3%CO₂/67%H₂/N₂ (A) and 3%CO/3%CO₂/67%H₂/N₂ (B) conditions at 25 bar, 200 °C. Blue: CO; Red: CO₂; Green: CH₃OH; Purple: H₂O.

 

In the absence of CO₂ in the feed a different situation arises. Transient CO₂ and H₂O formation is observed during start-up, evenly distributed over the entire catalyst bed (Figure 2A). Methanol synthesis is observed after a short delay, at a 5 to 6 times lower rate than in the presence of CO₂. The methanol production rate decreases with time on stream, whilst the CO₂ and H₂O formation do not disappear entirely. These results suggest that the catalyst undergoes changes during exposure to CO/H₂. Subsequent exposure of the CO/H₂-treated Cu/ZnO/Al₂O₃ catalyst to 3%CO/3%CO₂/67%H₂/N₂ results in complete restoration of the activity as seen in the original activity test under CO/CO₂/H₂, highlighting the reversibility of the catalyst reduction process in CO/H₂.

Figure 2. Catalyst performance for Cu/ZnO/Al₂O₃ during start-up under 3%CO/67%H₂/N₂ conditions at 25 bar, 200 °C (A), and for Cu/CeO₂ during start-up under 2%CO/H₂, 25 bar, 195 °C (B).

Studying the Cu/CeO₂ reference catalyst in CO/H₂ feeds revealed a number of similarities and key differences between these systems: i) Methanol synthesis was observed from CO/H₂ feeds, with transient CO₂ and H₂O formation during start-up (Figure 2B), in a similar vein to that seen for Cu/ZnO/Al₂O₃ in CO/H₂; ii) Methanol synthesis starts only after this CO₂ and H₂O formation; iii) Addition of CO₂ to the CO/H₂ feeds leads to irreversible catalyst deactivation due to build-up of surface carbonate and formate species in case of Cu/CeO₂. XPS and DRIFTS spectroscopy confirmed the presence of cationic Cu species, Cu^d+, and partially reduced ceria, CeO_2-x, under CO/H₂ reaction conditions.

4. Conclusions

Micro-kinetic modelling tools have increased the understanding of the methanol synthesis reaction over Cu/ZnO/Al₂O₃ under a wide range of reaction conditions. Generation of surface carbonate (CO₃.2s) species was found to be key to generation of catalytic activity from CO/CO₂/H₂ feeds. Methanol formation in the absence of CO₂ (i.e., from CO/H₂) occurs at a significantly reduced rate and involves a different mechanism and active site, involving the reducible ZnO component. Similar observations during catalyst start-up in CO/H₂ were made for the Cu/CeO₂ system, where cationic Cu species, Cu^d+, were found to be present under reaction conditions, the active state of the catalyst being Cu^d+/CeO_2-x. Similarly, a Cu^d+/ZnO_1-x active phase is likely to exist for the Cu/ZnO/Al₂O₃ catalyst in CO/H₂. Addition of CO₂ to the feed has a detrimental effect on Cu/CeO₂-based methanol synthesis, whereas a change from Cu⁺/ZnO_1-x to a Cu⁰/ZnO active phase occurs for Cu/ZnO/Al₂O₃, with a change from CO hydrogenation to CO₂ hydrogenation.

 

1. Twigg, M.V. Methanol Synthesis. In Catalyst Handbook, 2nd Edition ed.; Manson Publishing Ltd: London, 1989; 441-468.

2. Nix, R.M.; Rayment, T.; Lambert, R.M.; Jennings, J.R.; Owen, G. J. Catal. 1987, 106, 216-234.

3. van de Water, L.G.A.; Wilkinson, S.K.; Smith, R.A.P.; Watson, M.J.. J. Catal. 2018, 364, 57-68.

4. Birtill, J.J. Catal. Today 2003, 81 (4), 531-545.

5. Wilkinson, S.K.; van de Water, L.G.A.; Miller, B.; Simmons, M.J.H.; Stitt, E.H.; Watson, M.J. J. Catal. 2016, 337, 208-220.