(163k) Preparation of Highly Catalytically Active Cu/Zno and Cu/Zno/Al2O3 Materials by Metal-Organic Chemical Vapor Deposition

Copper-based catalysts (especially Cu/ZnO and Cu/ZnO/Al₂O₃) make the basis of the industrial methanol synthesis and methanol oxidation and they are important component of fuel cell technology. For the preparation of these catalysts, metal?organic chemical vapor deposition (MOCVD) technique has proven its potential. From the scientific point of view, the binary Cu/ZnO catalyst is also a prototype for the exploration of synergistic metal?support interactions for the understanding of catalytic reactions on the molecular level - the so-called strong metal support interactions (SMSI). Eventually, strategies can be developed to gain molecular control over the Cu/ZnO interface that would provide the basis for the knowledge-driven improvement of heterogeneous catalysts. There are numerous reports that suggest that the properties of the active sites are strongly influenced by the presence of alumina species. Moreover, the stability of the Cu/ZnO catalyst is improved by the addition of alumina. It is also believed that aluminium oxide has the ability to keep the copper particles apart and stabilise them against sintering, i.e. to behave as a so-called structural promoter in that catalyst system. Therefore, synthesis strategies should be developed that enable strict control over the properties of this material, which would lead to the establishment of the model system for the Cu/ZnO/Al₂O₃ system. There is in principle no direct limit in respect to the simultaneous maximization of the specific copper surface area and the Cu/ZnO interface area, and thus to increase of the catalytic activity beyond that which is currently possible as long as low density and high surface area mesoporous host matrices are used to anchor the catalytically active species. For further enhancement of the catalytic performance of these systems, but also for further scientific understanding of the mechanisms, chemical vapor deposition of the components is studied in more detail and the introduction of Al₂O₃ as the third component is considered. Since the activity of these catalysts depends on the numerous (structural) factors such as chemical composition, crystallite size, copper surface area, as well as bulk structural parameters like microstrain in the copper particles, further research in this area is needed in order to establish clear structure-activity relationships.

Approach and techniques

Nanochemistry in the pores and cavities can be tuned depending on the type of supporting mesoporous material i.e. its pore size, available surface area, surface properties, etc. The influence of the mesoporous matrix properties on the catalyst properties can be studied by using various periodic mesoporous silicate supports with different pore size, pore size distributions, pore shapes, etc. Owing to their high specific surface area and precisely controllable pore structures in the lower nanometer range, periodic mesoporous silica (PMS), such as MCM-41, MCM-48, SBA-15, MCF and related materials including inorganic-organic hybrid derivatives and similar, have been proven to be excellent supports for numerous catalytically active species, but also suitable matrix materials for the research of the in-pore chemistry.

In addition to conventional aqueous impregnation/calcination procedures, metal?organic chemical vapor deposition (MOCVD) techniques has proven its potential for the preparation of the highly active catalysts. The clear advantage of this technique comparing with the typical solution-phase method that uses salt type precursor is that unwanted effects of the solvent can be completely avoided in the gas phase. Moreover, the molecular organo-metallic (MOCVD) precursors tend to cleanly generate the desired materials with controlled growth rate at relatively low growth temperature. [Cu(OCHCH₃CH₂N(CH₂)₂] and Zn(CH₂CH₃)₂ are used as the MOCVD precursors. ZnEt2 acts as elegant reducing agent for the copper alkoxide and in parallel covers the silica surface with ZnEt-species which in turn are converted into the ZnO promoter component upon calcination/reduction cycles. The ternary Cu/ZnO/Al₂O₃ system can be made by the MOCVD loading of the PMS support using an appropriate Al₂O₃ precursors, for example aluminum alkoxides such as Al(OiPr)₃ or aluminum alkyls and hydrides such as Al(CH₃)₃ and H₃Al(N(CH₃)₃), all being typical choices and commercially available MOCVD precursors of aluminium oxide and aluminium, respectively. In order to maximize the specific copper surface area and the Cu/ZnO interface area, the loading of copper in the matrices can be increased by the repetition of the loading cycles. During the process, copper particle size and agglomeration should be kept as low as possible, while accessible surface area of the loaded porous matrix stays as high as possible. Simultaneously, the particle size and porosity as a function of the loading cycle number and the conditions are determined. Furthermore, fluidized bed chemical vapour deposition (FB-CVD) type reactors are investigated, since that is an industrially relevant catalyst preparation technique.

PXRD is the main technique used to determine the degree of crystallisation and particle size of copper. In principle, the copper XRD peak should be kept as broad as possible even under the conditions of repeated loading cycles. The influence of the Al₂O₃ component is also studied in this way. Additional information about particle aggregation and species can be obtained using XANES and EXAFS. Further, FTIR is applied to reveal the surface properties of the materials under consideration. For the determination of the pore size and internal surface area standard nitrogen absorption (BET/BJH methods) is used. As additional routine methods temperature-programmed reduction and copper area determination by N2O frontal chromatography is applied. Visual examination and the chemical composition of the final materials is carried out using TEM-EDX. Finally, detailed catalysis tests of selected and promising samples (those with high copper loading, broad PXRD copper peaks and high specific surface area) is carried out. Microstructure of the samples that show respectably high or superior activity would be again investigated in some more detail to elucidate the particular properties aiming to contribute to the understanding of the SMSI effect and related metal support interactions in the particular system.

Results

Between 8 and 12% of copper and 21-25% of zinc were successfully deposited by MOCVD using MCM48 and MCF materials as a matrix. The specific copper surface area of the obtained Cu/ZnO@PMS samples was in the range of 5-6 m2Cu/gcat, while the XRD and EXAFS studies revealed exceptionally low copper crystallite size and degree of aggregation. When the MCM48 was used as a matrix, the absence of characteristic copper peaks suggests particle size lower than 2 nm, while in the case of MCF average particle size of 5 nm was estimated using the Scherrer formula. However, repeated loading led to a significant decrease of the material internal surface area, This is assigned to the low pore size of this material, and for further studies of the repeated loading, silica based porous materials with the average pore size of 10 nm and above were used. The methanol synthesis performance is found to be 19-130 μmol/g_cath, using a special catalytic test set-up referenced to the industrial ternary catalyst having a value of about 450 μmol/g_cath. Catalysts of the type Cu/ZnO/A₂O₃@Matrix have also been prepared and their catalytic activity as well as synergetic effect between the components is under investigation.

Conclusions

MOCVD technique, both fixed and fluidized bed, has been demonstrated as a successful method for the synthesis of Cu/ZnO and Cu/ZnO/Al₂O₃ and catalysts using templated silica and porous alumina particles as a support. The molecular control of the Cu/ZnO interface is achieved by variation of dimension and framework of the support pores and by exploration of the precursor chemistry. Catalytic activity of the studied materials is higher then expected for the catalysts with the corresponding copper surface area, which proves their potential for the exceeding activity of industrially co-precipitated bench-mark catalyst.