Simultaneous production of methanol and dimethylether from synthesis gas

Abstract

Dimethylether is a non-toxic liquefied gas, which is projected to become one of the fundamental chemical feedstock in the future. Dimethylether can be produced from syngas via a two-step (indirect) process that involves synthesis of methanol by hydrogenation of CO/CO2 over a copper based catalyst and subsequent dehydration of methanol to DME over an acidic catalyst. Alternatively, DME can be produced in an one-step (direct) process using a hybrid (bifunctional) catalyst system that permits both methanol synthesis and dehydration in a single process unit. In the present research work the production of DME has been studied by applying both the indirect and direct processes. Firstly, the methanol synthesis and methanol dehydration reactions involved in the indirect process have been studied separately. Afterwards, these two reactions have been combined in the direct DME production process by using a hybrid catalyst comprising a methanol synthesis and a methanol dehydration component. The methanol synthesis by CO2 hydrogenation has been investigated over commercial and home-made CuO/ZnO/Al2O3 catalysts with the aim to identify optimal experimental conditions (CO2:H2 ratio, flow rate, temperature) that could be then used in the direct conversion of CO2/H2 mixtures into methanol/DME. Obtained results reveal that the conversion of CO2 and the yields of reaction products (CH3OH and CO) increase when the concentration of H2 in the feed and the reaction contact time are increased. It was found that both Cu+/Cu0 species are important for the conversion of CO2/H2, although the presence of Cuo seems to be more important for selectivity/yield of methanol. The stability of the CuO/ZnO/Al2O3 catalyst has been also investigated. It was observed that the main reason for the deactivation of catalyst is the water produced via the methanol synthesis and reverse water gas shift reactions. However, the catalytic activity and products selectivity were recovered slowly to their original levels after applying a regeneration procedure, indicating that deactivation by water is reversible.The dehydration of methanol to dimethylether (DME) has been investigated over a range of catalysts including alumina, silica-alumina and zeolites with different physicochemical characteristics. The effects of temperature and the presence of water vapour in the feed on catalytic performance have been studied in detail. The reactivity of catalysts has been evaluated by determining the reaction rates per gram of catalyst per acid site (total: Brönsted+Lewis) and per Brönsted/Lewis mole ratio. In addition, the reaction mechanism has been investigated over a selected catalyst, with the use of transient-MS and in situ DRIFTS techniques. Results obtained for alumina catalysts show that the catalytic activity and selectivity are determined to a large extent by the textural properties, degree of crystallinity and total amount of acid sites of catalysts. In particular, the methanol conversion curve shifts toward lower reaction temperatures with an increase of specific surface area. However, the enhanced catalytic activity of high-SSA samples cannot be attributed solely to the higher amount of surface acid sites, implying that the reaction rate is determined to a large extent from other parameters, such as textural properties and degree of crystallinity. Results of mechanistic studies indicate that interaction of methanol with the Al2O3 surface results in the formation of two kinds of methoxy groups of different adsorption strength. Evidence is provided that DME evolution is associated with methoxy species that are weakly adsorbed on the Al2O3 surface, whereas more strongly held species decompose to yield surface formate and, eventually, CH4 and CO in the gas phase. Results obtained over zeolite catalysts show that catalytic performance depends on the topology of zeolites due to differences in micropore structure and Si/Al ratio as well as on the number, strength and nature of active acid sites. The activity of zeolite catalysts for the methanol dehydration to DME follows the order ZSM-5 > Ferrierite > Mordenite ~ Beta ~ USY > H-Y. The strong Brönsted acid sites of ZSM-5 zeolites with relatively high Si/Al ratio represent the most active sites in methanol dehydration to DME reaction. However, the overall reactivity of the ZSM-5 zeolites is also affected by the balance of the Brönsted to Lewis acid sites. The activity of Beta and USY zeolites is determined by both Lewis and Brönsted acid sites. The moderate/low reactivity of Ferrierite, Mordenite and H-Y zeolite are determined by the abundant Brönsted acid sites of relatively weak/moderate strength.The direct CO2 hydrogenation to methanol/DME has been investigated using admixed catalysts comprising a methanol synthesis (commercial copper based catalyst: CZA1) and a methanol dehydration component (different alumia/zeolite catalysts: γ-Al2O3, ZSM-5, W/γ-Al2O3, USY(6), Ferrierite(10)). It has been revealed that the conversion of CO2 is always lower than the corresponding equilibrium values predicted by thermodynamics, indicating operation in the kinetic regime. The nature of the methanol dehydration component of the admixed catalysts was found to be important for both CO2 conversion and methanol dehydration. In particular, DME selectivity/yield, depends strongly on the nature of acid sites (both Lewis and Brönsted) as well as the textural (meso/macro porosity) and topological properties of methanol dehydration component of the admixed catalysts. The yield of DME obtained at a temperature of 250oC decreases following the order CZA1/ZSM-5, CZA1/USY(6) > CZA1/Ferrierite(10) > CZA1/ W/γ-Al2O3 >> CZA1/γ-Al2O3. The long-term stability experiments conducted over selected bifunctional catalytic systems revealed that the catalysts deactivate with time-on-stream, mainly due to water produced via methanol synthesis, methanol dehydration and reverse water gas shift reactions. In case of the CZA1/ZSM-5 admixed catalyst the catalytic activity and products selectivity were almost recovered after regeneration indicating that deactivation by water is reversible.