(387f) Designing Coke-Resistant Dehydrogenation Catalysts with DFT and Grand Canonical Monte Carlo Simulations

Platinum-group metals are employed for non-oxidative alkane dehydrogenation because they offer a high degree of coke-resistance in carbon-rich environments. Cheaper iron-based catalysts are an attractive alternative, but they are more reactive and suffer from unselective cracking side reactions and rapid deactivation through coke formation. Iron-carbide catalysts can achieve selective propane dehydrogenation if surface modifications are introduced to suppress coke formation. In this work, we used grand canonical Monte Carlo (GCMC) with density functional theory (DFT) to identify effective surface modification strategies to enhance the coking resistance of Fe-based catalysts. Two strategies we investigate are (1) alloying Fe with other transition metals or main group elements and (2) introducing electronegative adsorbate groups to weaken carbon binding to the surface. GCMC simulations were performed to simulate carbon uptake and coke formation on the pristine iron surface, iron sulfide surfaces, iron-aluminum alloy surfaces, as well as surfaces with electron-withdrawing adsorbates. GCMC simulations were essential for sampling the large number of viable surface structures that can arise under realistic reaction conditions. These structures then were used to populate phase diagrams evaluated using the formalism of ab initio thermodynamics, which demonstrate that several viable surface modification strategies can effectively suppress coke formation. Analysis of the many resulting structures reveals that both electronic and geometric effects play an important role in different contexts. We find that iron carbide and iron sulfide have lower d-band center values, indicating carbon destabilization through an electronic effect. However, the d-band value of the iron-aluminum alloy is close to that of the pure iron, which suggests that a geometric effect is more important to coke resistance for this alloy. Overall, we show that GCMC is a robust approach for sampling coke formation structures and for finding strategies to suppress them under dehydrogenation conditions.