(190b) Revised Activity Descriptor for Recombinative Hydrogen Desorption

The widespread adoption of water electrolysis using renewable energy for the sustainable production of H₂ is partially hindered by the scarcity and high price of Pt, the catalyst of choice for the hydrogen evolution reaction (HER). The high catalytic activity of Pt for HER has been attributed to its optimal binding strength of hydrogen. The empirical correlation between the metal-hydrogen binding strength and HER activity was first reported by Trasatti in 1972 and later popularized by Nørskov et al by providing a theoretical basis within the framework of the volcano curve.

Despite its widespread adoption, however, the optimal activity requirement of ΔG=0 has shortcomings. Zero coverage values are often unrealistic and attempts to estimate ΔG near saturation coverage are futile, because thermodynamics dictate ΔG=0 at equilibrium. Figure 1a confirms that the activation barriers (E_a) for recombinative H₂ desorption from a near equilibrium surface plotted against ΔG collapse to a vertical line. The small deviations from ΔG=0 originate from the finite size of the unit cell used in the density functional theory (DFT) calculations.

We observed that the two H* atoms involved in the Tafel step on equilibrated surfaces typically bind in two different types of sites, e.g. top and fcc on fcc(111) surfaces. When using the binding energy difference (ΔΔG) between the weaker and stronger binding sites as descriptor instead, we show in Figure 1b that it linearly correlates with E_a for pure transition metals, near surface and some single atom alloys (SAA). Notably, we identified several deviating SAAs which have significantly lower E_a and could be interesting candidates for further investigations.

Overall, the kinetics of the Tafel step are more rigorously captured by ΔΔG as alternative descriptor, as it conforms to thermodynamic principles and removes ambiguity in choosing surface coverages for DFT calculations.