Reaction-diffusion coupling across scales during the induction period of methanol-to-olefin conversion over ZSM-5 catalysts

The coupling of reaction and diffusion across the catalyst pore, grain, pellet and reactor bed has been studied using a combination of novel theoretical concepts for incorporating dynamics into kinetics, and a particle-resolved transient microkinetic model applied to temperature-programmed desorption (TPD), and step response studies of methanol, and dimethyl ether over ZSM-5 catalysts, respectively. A resolution of TPD profiles showcases the coupling between adsorption, desorption, and surface diffusion across scales. Six dynamic models are investigated to describe the 44-min induction period in the first step-response cycle and the 95% reduction in induction period in subsequent step-response cycles of dimethyl ether conversion at constant temperature. These include coverage, fixed site-interconversion, dynamic site-interconversion, mass transport of gas, and adsorbed species, and anomalous surface diffusion. None of these dynamic models could explain the overall step-response studies. A combination of adsorption, desorption, and surface reaction of dimethyl ether occurs alongside with competitive adsorption of water and methanol leading to a series of stable intermediates that produce propylene after a 44-min induction period. These intermediates are best described within the methoxymethyl mechanism. The reduction in induction period is due to high binding energies of dimethyl ether such that lower competitive adsorption occurs, fewer stable intermediates exist and a more direct pathway towards propylene formation is followed. The optimum surface coverage of the dissociative products of dimethyl ether (surface methoxy groups, methanol) is the key descriptor that governs the changing induction period behaviour at constant temperature. This site-distribution has a major influence on the dynamic reactor models which predict catalyst activity and product selectivity during methanol-to-olefin conversion. First-principles particle-resolved transient microkinetic models bridge the gap between transient molecular models for the pores and microkinetic models for the grain, pellet, and reactor bed scales.