Facile Tool for Prediction of Catalytic Activity Cu-Lanthanoid Catalyst for Methanol Synthesis

Methanol (MeOH) and dimethyl ether (DME), which can be easily obtained from MeOH, are superior candidates for clean transportation fuel. A compact and simple process with good economy has been proposed to produce these fuels from dispersed unused carbon resources (Yamada, 2003). The key point of this process is development of a noble catalyst which is active for MeOH synthesis under mild reaction conditions. A number of new catalysts for MeOH synthesis from syngas were discovered and some of them are more active than the conventional Cu/Zn/Al catalyst. Cu-lanthanoid catalyst was reported as one of the alternative catalyst (Andriamasinoro et al., 1993; Nix et al., 1987; Walker et al., 1992). Under mild conditions Cu-Yb showed higher activity than Cu-Zn-Al (Sakata et al., 1998). Prediction of the characteristics of new catalysts or catalyst additives from the physicochemical properties of catalyst components would accelerate catalyst development. In the present study such prediction methodology was developed and applied for MeOH synthesis by Cu-Lanthanoid catalyst. As reviewed in the introduction of reference(Valero et al., 2009), modeling methodologies were suggested in the research field of catalysis (Baumes et al., 2004; 2007; Farrusseng et al., 2005; Grubert et al., 2003; Holeňa & Baerns, 2003; Serra et al., 2003; Serra, Corma, Valero, Argente & Botti, 2007; Wolf et al., 2000). It was also reported that an artificial neural network, especially a radial basis function network(RBFN),(Omata et al., 2004; Umegaki et al., 2003) was successfully applied for the regression of catalytic phenomena instead of the conventional polynomial equation. Such methodology is effective for integrating the observation(Serna et al., 2008) and the characterization(Barr et al., 2004; Baumes et al., 2009; 2008; Gilmore et al., 2004; Takeuchi et al., 2005). Successful prediction of catalytic properties from the physicochemical properties of the catalyst elements was, however, reported only in few cases (Kito et al., 1992; 1994). We recently succeeded in identifying an effective additive for Ni/active carbon (AC) catalyst for the carbonylation of methanol based on previous experimental results and the physicochemical properties of the elements (Omata & Yamada, 2004). The physicochemical properties of element X were correlated by means of RBFN to the catalytic selectivity Kohji Omata1, Tetuo Umegaki2 and Muneyoshi Yamada3 1Tohoku University 2Nihon University 3Akita National College of Technology Japan Facile Tool for Prediction of Catalytic Activity Cu-Lanthanoid Catalyst for Methanol Synthesis 8