Dry reforming of methane involves conversion of two of the most abundant greenhouse gases, namely methane and carbon dioxide, to a gas mixture composed mainly of carbon monoxide and hydrogen. This gas mixture may then be used as a resource for the synthesis of valuable chemicals and fuels, through Fischer-Tropsch and methanol/dimethyl ether synthesis. This process has also been considered as an attractive method for the production of hydrogen from biogas. Dry reforming reaction is generally accompanied by reverse water gas shift reaction (RWGS).

                          CH₄ + CO₂ ↔ 2CO + 2H₂       (Reforming)

                                CO₂ + H₂ ↔ CO + H₂O         (RWGS)

Coke minimization during dry reforming of methane is an important challenge^1,2 and our recent studies have shown that modification of Ni based catalysts with tungsten oxide caused significant improvement in minimization of coke formation during this reaction^3,4. Due to thermodynamic limitations, quite high reaction temperatures are generally preferred for this process. Elimination of RWGS reaction is another challenge to obtain high hydrogen yields.

            In this study, mesoporous alumina supported bimetallic Ni-W catalysts were synthesized by a one-pot hydrothermal route and tested in dry reforming of methane in a tubular membrane reactor  having a Pd wall. Product distributions obtained in the membrane reactor were compared with the results obtained in a conventional fixed bed catalytic reactor. The catalytic material, which was synthesized following a one-pot route by using P123 as the template, was shown to have an ordered pore structure. It contains 5% Ni and 10% W, and it has a surface area of 125 m²/g. Fixed bed reactor results of dry reforming of methane, which were obtained at 600^oC with a feed gas mixture containing equimolar CH₄/CO₂, and at a space time of 0.2 s.g.cm^-3, proved that activity of the catalysts synthesized in this study was quite high. However, due to the occurrence of RWGS reaction, H₂ selectivity values were always less than CO selectivity. In fact, the ratio of CO to H₂ in the product stream was always higher than 1.5 in the fixed bed reactor. Contribution of RWGS reaction also caused higher CO₂ fractional conversion values than the fractional conversion of methane. On the other hand, performance of the same catalyst in the membrane reactor, at similar conditions, proved elimination of RWGS reaction as a result of in-situ removal of produced hydrogen from the reaction zone. As a result of removal of hydrogen from the reaction zone through the Pd membrane wall of the reactor, significant enhancement of hydrogen yield was observed. In fact, hydrogen yield was even higher than CO yield in the membrane reactor. Our results obtained with other Ni based catalysts also supported the conclusion that RWGS reaction was eliminated in the membrane reactor. Membrane reactor did not only eliminate equilibrium limitations through removal of one of the products (H₂in this case) from the reaction zone, it also minimized occurrence of RWGS reaction during dry reforming of methane. Hence, much higher hydrogen yields were obtained in the membrane reactor, as compared to conventional fixed bed reactor.

Acknowledgement: TUBITAK project (111M449), Turkish Academy of Sciences and Collaboration with Dr. Pintar of Slovenian Institute of Chemistry.

References:

[1] Arbag H., Yasyerli S., Yasyerli N., Dogu G., Int. J. Hydrogen Energy 35 (2010) 2296.

[2] Djinovic P., Crnicev I.G.O, Erjavec B., Pintar A., Appl. Catal. B: Env. 125 (2012) 259.

[3] Arbag H., Yasyerli S., Yasyerli N., Dogu T., Dogu G., Topics in Catal. 56 (2013) 1695.

[4] Arbag H., Yasyerli S., Yasyerli N., Dogu G., Dogu T., Ind. Eng. Chem. Res. (2015) doi: 10.1021/ie504477t