Insight into Catalysts for Methanol Synthesis

Methanol Synthesis over Copper Catalysts

The production of methanol from CO and H₂ is a widely used process in industry since the 1920s. The use of CO₂ is also feasible, either by a prior reverse water-gas shift reaction (CO₂ + H₂ ⇌ CO + H₂O) or via the direct hydrogenation of CO₂. This offers a sustainable production pathway to methanol—an important base chemical and fuel—by the valorization of waste CO₂. It could also allow the use of methanol as a storage molecule for hydrogen produced from renewable energy sources via water electrolysis [1]. However, in this scenario, the process and especially the catalyst would have to deal with a fluctuating gas supply, constantly changing the redox potential of the synthesis gas mixture.

Since the implementation of the low-pressure methanol process by Imperial Chemical Industries (ICI), UK, in the 1960s, the typical industrial catalyst for methanol synthesis is a Cu/ZnO catalyst with additional low-content promotion by Al or Mg. In this material, ZnO serves both as the support and an electronic promoter. This promotional effect is derived from a dynamic accumulation of ZnO_x species (0 ≤ x < 1) on the copper particle surface and is described as strong metal–support interactions (SMSI) [2]. Al₂O₃ is seen as a structural promoter for the ZnO support, but an additional effect of electronic promotion is discussed as well [3].

Figure 1. Schematic drawing of the dynamics in the Cu/ZnO system as a function of the chemical potential of the feed gas (depending on the CO:CO₂ ratio in the CO_x/H₂ synthesis gas). The actual eligible configurations of ZnO, i.e., oxygen-deficient ZnO_x (x<1) or surface alloy islands on Cu, can be monitored and discriminated spectroscopically, whereas the fully oxidized or reduced states are theoretical reference points.

The actual nature of the ZnO_x species and the catalytically active phases are subject to discussion. They are highly dependent on the redox potential of the gas feed and can influence the reaction to methanol at the catalyst surface (see Fig. 1).

Having said that, it becomes clear that it is very important to study the catalyst’s structure and performance under dynamic conditions. In the framework of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) priority program SPP2080, we adress these dynamic interactions using synthetic, in situ– and operando-analytical as well as computational methods for different promoter oxides [4–6].

Using MgO Instead of ZnO

The synthesis and catalytic performance of Cu/ZnO/Al₂O₃catalysts is well documented in the literature [7,8]. However, to study the promotional effect of ZnO, we need to separate this from the support function that ZnO also has in the classical system. Therefore, we developed a model system with the irreducible oxide MgO instead of ZnO.

During the synthesis of these materials, the co-precipitation yields mixed Cu,Mg hydroxycarbonates of the malachite or mcguinnessite structure type. Calcination and subsequent reduction lead to Cu/MgO catalysts. These materials are active in CO hydrogenation, but it is possible to add ZnO in small amounts by impregnation, re-enabling the hydrogenation of CO₂ and CO/CO₂ mixtures [9]. This offers the opportunity to investigate the state and dynamic changes of the ZnO promoter with element-specific X-ray absorption spectroscopy under dynamic reaction conditions, which we currently do together with our collaborators from the Karlsruhe Institute of Technology (KIT).

Precursor Chemistry of Cu/MgO Catalysts

Apart from the spectroscopic and computational investigations of the ZnO promotion, we are also interested in the precursor chemistry of these Cu/MgO catalysts from a solid-state chemical point of view: Crystallographically, the mineral mcguinnessite, (Cu_0.5Mg_0.5)₂CO₃(OH)₂, belongs to the malachite structure group. However, as its Cu,Zn equivalent rosasite, it is not structurally identical to the pure Cu malachite, Cu₂CO₃(OH)₂, but shows a slightly different crystal structure without a smooth transition from one to the other [10].

This leads to interesting findings when we gradually substitute Cu by Mg in a series of synthesis experiments: Typically, the hydroxycarbonates are synthesized by pH-controlled co-precipitation at 65 °C in an automated lab reactor. The precipitates are aged in the mother liquor, filtered, washed, and dried. When using small amounts of Mg (up to 10 at-%), we obtain moderately crystalline (partly Mg-substituted) malachites of the typical green color. In these experiments, a characteristic pH minimum appears in the synthesis protocol after a given time and is used as an indicator for successful crystallization—this aging time scales nearly linearly with the amount of Mg used. However, for larger ratios (e.g., 15 or 30 at-% Mg), another rather step-like pH drop is observed much earlier and only amorphous products of blue color could be isolated. These products are prone to Mg leaching during the washing step.

In a different approach, we applied a hydrothermal aging step at pH 8.0 (102 °C, 20 h) and were able to obtain turquoise-colored crystalline mcguinnessites—where Cu is substituting Mg in deviation from the regular 1:1 ratio (see Fig. 2). Hence, we expect a changeover between the formation of Mg-substituted malachite and Cu-substituted mcguinnessite at a point between 10 and 15 at-% Mg.

Schematic synthesis protocol, ageing times, PXRD patterns

Figure 2. Left: Schematic synthesis protocol with different shapes of pH drops during aging. Center: Ageing times increase with rising Mg content; for high amounts, amorphous products are formed after a few hours; so far, longer aging times led to the oxolation of Cu to CuO which is to be avoided during the wet-chemical synthesis. Right: PXRD patterns of the synthesized Cu,Mg hydroxycarbonates and comparison of the crystal structures.

Further experiments in the respective composition range as well as a thorough investigation of the hydrothermal treatment will help elucidate the crystal chemistry of these mixed Cu,Mg hydroxycarbonates. This will improve our knowledge about the possibilities to substitute Cu and Mg in the malachite and mcguinnessite crystal structures, and should allow us to optimize the synthesis and preparation of Cu/MgO catalysts from crystalline precursors.

Acknowledgment

We gratefully acknowledge the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) for funding through the priority program SPP2080 “Catalysts and reactors under dynamic conditions for energy storage and conversion” (project number 406903668) and especially Jan-Dierk Grunwaldt and Felix Studt at KIT for fruitful discussions and an interactive and lively collaboration.