Artificial neural network-aided catalyst research for low-pressure DME synthesis from syngas

Artificial neural network (ANN) was applied to screen additives of Cu/Zn based catalyst which is effective for high activity of methanol synthesis from syngas in one-step DME synthesis at low pressure (1-3 MPa) and low temperature (500-550K). Besides the conventional additives such as Al, Ti and V were discovered to be effective additives for high activity of methanol synthesis when Cu/Zn based catalyst were prepared by oxalate-ethanol co-precipitation method and mixed with γ-alumina. They were concluded to mitigate the inhibiting effect of H 2 O formed at high CO conversion. ANN was also applied for optimization of catalyst composition. The methodology comprising of DOE, ANN and grid search, was confirmed to be useful and can cope with the catalyst optimization tailored for each process situation.