Optimization of Cu oxide catalysts for methanol synthesis by combinatorial tools using 96 well microplates, artificial neural network and genetic algorithm

Compact and economic processes for methanol synthesis from syngas demand a new catalyst that is active under low-pressure and low temperature. Combinatorial approach comprising a high-pressure high-throughput screening (HTS) reactor system, an artificial neural network (NN), and a genetic algorithm (GA) was applied for the catalyst development. A variety of 96 microplates were used in the HTS reactor system for both preparation and activity testing to handle 96 catalyst samples simultaneously. Activity test results were used as training data for NN. After training, the NN can map catalyst activity as a function of catalyst composition and parameters for catalyst preparation. GA was used as an optimization tool to find maximum catalyst activity in the trained artificial neural network. Composition of methanol synthesis catalyst (Cu–Zn–Al–Sc–B–Zr), calcination temperature and the amount of precipitant were optimized simultaneously under pressure (1 MPa) because optimum catalyst composition is usually affected by both preparation and reaction conditions. The composition of the optimum catalyst was Cu/Zn/Al/Sc/B/Zr=43/17/23/11/0/6 prepared using 2.2 times the equivalent of oxalic acid and calcined at 605 K. The activity (427 g-MeOH/kg-cat./h) was much higher than that of industrial catalyst (250 g-MeOH/kg-cat./h) at 1 MPa, 498 K.