Abstract
In acidic media, many transition-metal phosphides are reported to be stable catalysts for the hydrogen evolution reaction (HER) but typically exhibit poor stability toward the corresponding oxygen evolution reaction (OER). A notable exception appears to be Rh(2)P/C nanoparticles, reported to be active and stable toward both the HER and OER. Previously, we investigated base-metal-substituted Rh(2)P, specifically Co(2-x)Rh(x)P and Ni(2-x)Rh(x)P, for HER and OER as a means to reduce the noble-metal content and tune the reactivity for these disparate reactions. In alkaline media, the Rh-rich phases were found to be most active for the HER, while base-metal-rich phases were found to be the most active for the OER. However, Co(2-x)Rh(x)P was not stable in acidic media due to the dissolution of Co. In this study, the activity and stability of our previously synthesized Ni(2-x)Rh(x)P nanoparticle catalysts (x = 0, 0.25, 0.50, 1.75) toward the HER and OER in acidic electrolyte are probed. For the HER, the Ni(0.25)Rh(1.75)P phase was found to have comparable geometric activity (overpotential at 10 mA/cm(geo)(2)) and stability to Rh(2)P. In contrast, for OER, all of the tested Ni(2-x)Rh(x)P phases had similar overpotential values at 10 mA/cm(geo)(2), but these were >2x the initial value for Rh(2)P. However, the activity of Rh(2)P fades rapidly, as does Ni(2)P and Ni-rich Ni(2-x)Rh(x)P phases, whereas Ni(0.25)Rh(1.75)P shows only modest declines. Overall water splitting (OWS) conducted using Ni(0.25)Rh(1.75)P as a catalyst relative to the state-of-the-art (RuO(2)||20% Pt/C) revealed comparable stabilities, with the Ni(0.25)Rh(1.75)P system demanding an additional 200 mV to achieve 10 mA/cm(geo)(2). In contrast, a Rh(2)P||Rh(2)P OWS cell had a similar initial overpotential to RuO(2)||20% Pt/C, but is unstable, completely deactivating over 140 min. Thus, Rh(2)P is not a stable anode for the OER in acidic media, but can be stabilized, albeit with a loss of activity, by incorporation of nominally modest amounts of Ni.