As the commercial catalyst in the propane direct dehydrogenation (PDH) reaction, one of the biggest challenges of Pt catalysts is coke formation, which severely reduces activity and stability. In this work, a first-principles DFT-based kinetic Monte Carlo simulation (kMC) is performed to understand the origin of coke formation, and an effective method is proposed to curb coke. The conventional DFT calculations give a complete description of the reaction pathway of dehydrogenation to propylene, deep dehydrogenation, and C−C bond cracking. The rate-limiting step is identified as the dissociative adsorption of propane. Moreover, a comparison between different exchange-correlation functionals indicates the importance of van der Waals corrections for the adsorption of propane and propylene. The lateral interactions between the surface adsorbates are significant, which implies that mean field microkinetic modeling might not adequately describe the reaction process. There are two distinct stages in PDH, which are quick deactivation and steady state, respectively, as revealed from the kMC simulation. The precursor of coke mainly formed during the quick deactivation. The calculations indicate that the geometries of the active sites for the dehydrogenation and deep reactions are different. Therefore, the availability of surface sites is a crucial factor in the formation of propylene and side products. The active sites from quick deactivation are mainly occupied by C 2 /C 1 species, which are hard to remove. On the other hand, the surface sites that are left are mainly active toward dehydrogenation to propylene due to the geometry constraint. Therefore, a stable activity and selectivity is reached. Furthermore, the effect of hydrogen molecules in the input stream is also explored. The calculations indicate that the inclusion of hydrogen in PDH reactants not only enhances the forward reactions to the propylene formation but also reduces the consumption of the resulted propylene during the reaction. Therefore, hydrogen is very helpful to the selectivity increase in PDH in addition to other effects. Overall, the current study lays out a solid base for the future optimization of the Pt catalysts in PDH and we propose that the fine control of the surface sites on Pt has paramount importance in reducing coke formation.

Proton exchange membrane water electrolysis (PEM-WE) has emerged as a promising technology for hydrogen production and shows substantial advantages over conventional alkaline water electrolysis. To enable efficient PEM-WE in acidic media, iridium (Ir)- or ruthenium (Ru)-based catalysts are indispensable to drive the thermodynamically and kinetically demanding oxygen evolution reaction (OER). However, developing Ir/Ru catalysts with high efficiency and long-term durability still remains a formidable challenge. Herein, we report one-pot hydrothermal synthesis of ultrafine IrRu intermetallic nanoclusters loaded on conductive, acid-stable, amorphous tellurium nanoparticle support (IrRu@Te). Benefiting from the large exposed electrocatalytically active surface of ultrafine IrRu clusters and the strong electronic coupling between IrRu and Te support, the as-obtained IrRu@Te catalysts show good catalytic performance for the OER in strong acidic electrolyte (i.e., 0.5 M H2SO4), requiring overpotentials of only 220 and 303 mV to deliver 10 and 100 mA cm–2 and able to sustain continuous OER electrolysis up to 20 h at 10 mA cm–2 with minimal degradation. Moreover, IrRu@Te exhibits high specific activity, illustrating intrinsically better performance compared with that of unsupported IrRu and other commercial Ir- and Ru-based catalysts. It also demonstrates unprecedentedly high mass activity of 590 A gIrRu –1 at an overpotential of 270 mV, outperforming most Ir- and Ru-based OER catalysts reported in the literature. Furthermore, IrRu@Te catalysts reveal good OER performance in neutral electrolyte as well, holding great potential to be used for PEM-WE in environmentally friendly conditions. Density functional theory (DFT) calculations based on oxidized IrRu confirm that the catalyst/support coupling results in a lower energy barrier for the oxygen–oxygen bonding formation, offering a rational explanation to the experimentally observed OER performance.

Achieving an efficient and stable oxygen evolution reaction (OER) in an acidic or neutral medium is of paramount importance for hydrogen production via proton exchange membrane water electrolysis (PEM-WE). Supported iridium-based nanoparticles (NPs) are the state-of-the-art OER catalysts for PEM-WE, but the nonhomogeneous dispersion of these NPs on the support together with their nonuniform sizes usually leads to catalyst migration and agglomeration under strongly corrosive and oxidative OER conditions, eventually causing the loss of active surface area and/or catalytic species and thereby the degradation of OER performance. Here, we design a catalyst comprising surface atomic-step enriched ruthenium–iridium (RuIr) nanocrystals homogeneously dispersed on a metal organic framework (MOF) derived carbon support (RuIr@CoNC), which shows outstanding catalytic performance for OER with high mass activities of 2041, 970 and 205 A gRuIr –1 at an overpotential of 300 mV and can sustain continuous OER electrolysis up to 40, 45, and 90 h at 10 mA cm–2 with minimal degradation in 0.5 M H2SO4 (pH = 0.3), 0.05 M H2SO4 (pH = 1), and PBS (pH = 7.2) electrolytes, respectively. Comprehensive experimental studies and density functional theory (DFT) calculations reveal that the good performance of RuIr@CoNC can be attributed, on one hand, to the presence of abundant atomic steps that maximize the exposure of catalytically active sites and lower the limiting potential of the rate-determining step of OER and, on the other hand, to the strong interaction between RuIr nanocrystals and the CoNC support that endows homogeneous dispersion and firm immobilization of RuIr catalysts on CoNC. The RuIr@CoNC catalysts also show outstanding performance in a single-cell PEM electrolyzer, and their large-quantity synthesis is demonstrated.