Diffusion-Selective Modulation Strategy for Enhanced Propane Dehydrogenation via Mixing Inert CHA Zeolite

Abstract Achieving a high propane conversion in propane dehydrogenation to propylene (PDH) carries tremendous scientific and economic value, particularly if achievable with established industrial units. Here we propose a diffusion-selective modulation strategy driven by large-scale machine-learning computation, where the key to success lies in finding auxiliary materials to drive the PDH reaction forward by enlarging the diffusion-rate difference between H2 and C3 species. This strategy is demonstrated theoretically via high-throughput computation and then validated experimentally through the introduction of inert siliceous CHA zeolites as a diffusion-selective modulator within the catalyst bed. By simply physically mixing CHA with various PDH catalysts, we show that the single-pass propane conversion can increase by 1–5% from industrial space velocities (10 h–1) to significantly higher throughputs (256 h–1), consistently improving performance across multiple reported catalysts without compromising stability. This “plug-and-play” methodology can be directly implemented in existing industrial PDH units without altering either the catalyst formulation or the reactor configuration, providing a readily implementable and scalable route for upgrading current PDH infrastructure.