Interfacial Water Structure as a Descriptor for Its Electro-Reduction on Ni(OH)2-Modified Cu(111)

The hydrogen evolution reaction (HER) has been crucial for the development of fundamental knowledge on electrocatalysis and electrochemistry, in general. In alkaline media, many key questions concerning pH-dependent structure–activity relations and the underlying activity descriptors remain unclear. While the presence of Ni(OH)2 deposited on Pt(111) has been shown to highly improve the rate of the HER through the electrode’s bifunctionality, no studies exist on how low coverages of Ni(OH)2 influence the electrocatalytic behavior of Cu surfaces, which is a low-cost alternative to Pt. Here, we demonstrate that Cu(111) modified with 0.1 and 0.2 monolayers (ML) of Ni(OH)2 exhibits an unusual non-linear activity trend with increasing coverage. By combining in situ structural investigations with studies on the interfacial water orientation using electrochemical scanning tunneling microscopy and laser-induced temperature jump experiments, we find a correlation between a particular threshold of surface roughness and the decrease in the ordering of the water network at the interface. The highly disordered water ad-layer close to the onset of the HER, which is only present for 0.2 ML of Ni(OH)2, facilitates the reorganization of the interfacial water molecules to accommodate for charge transfer, thus enhancing the rate of the reaction. These findings strongly suggest a general validity of the interfacial water reorganization as an activity descriptor for the HER in alkaline media.

. Integration of the OH adsorption peak. The mean square roughness Sq was determined for three different positions on the surfaces (two are exemplarily given in Figure S2), using the corresponding tool in Gwyddion. 2

Determination of the thermal coefficients from the laser transients
Generally, in the absence of specific adsorption phenomena or when their responses are negligible, the potential change ∆E recorded after the laser pulse follows the relaxation of the temperature at the interface. Assuming that the non-reflected part of the irradiated laser beam is suddenly converted into heat, the change of temperature with time can be described as follows 3,4 : (1) Where the maximum temperature change at the surface (at t=t0=5ns) is: R is the reflectivity of the surface, I the laser intensity, κ,α and κ1,α1 are the thermal conductivity and the thermal diffusivity of the metal and the solution, respectively, where α = κ/ρC with ρ the material density and C the heat capacity.
The heat flux on the metal (q) can be described as: Using the thermal and optical constant for copper (Table S3), we can estimate the maximum change of temperature to be 28 K. Note that we cannot account for possible variations of these constants upon Ni(OH)2 deposition. A beam energy I of 3.54 MWcm -2 was used.
Since the change of the electrode potential (∆E) follows the change of temperature, it can be described as: S6 (5) Where is the thermal coefficient at any given potential and q is the charge density on the metal. We can easily extract these coefficients from the slopes obtained after linearization of the transients (∆E vs 1/ ), shown in Figure S5, using the following equation: ∆Sdl is defined as the difference in the entropy of the double layer components when they are present in the formed interface and in the bulk of the adjoining phases. 7 The pzr, i.e. the pme, is hence the potential at which water molecules distribute randomly and there is no net dipolar contribution to the electrode potential. Therefore, it is closely related to the potential of zero free charge, since water dipoles orient according to the electric field at the surface.

Correction for the thermodiffusion potential
The thermodiffusion potential arises from the temperature differences between the solution surrounding the reference and the working electrode, which causes a potential drop due to the motion of ions resulting from the thermal gradient. While in most cases such contributions can be neglected, in the case of highly alkaline solutions, due to the abnormally high mobility of OH − ions, uncorrected thermal coefficients are slightly overestimated (see Figure S6). To provide more accurate results, we therefore need to estimate the thermodiffusion potential from the Eastman entropies of transport and mobility of ions 9 . This results in an approximate value of −0.43 mV K -1 in 0.1 M NaOH, 10 which is then used to obtain the corrected thermal coefficients shown in Figure 5 in the main text. where . It has to be noted that in the case of 0.2 ML of Ni(OH)2 there are three potentials at which the thermal coefficient becomes zero, however, only the change of sign going from negative to positive can be considered as a second pme, according to the integration of equation (7).