Discussion

measurements (Figure 8.7). This is a fingerprint of an electrochemical and non-limited reaction. As shown in the Appendix A.5, this has been assigned to the oxygen reduction reaction (ORR) following the reaction: ¹₂O₂ + H₂O+2 e⁻ → 2 OH⁻ where the oxygen comes from the remaining dissolved oxygen molecules in the electrolyte.

Effect of the surface orientation

Looking into details of the electrochemical results displayed in Figure 8.7, when the NiO is deposited on MgO(100)/Pt(100) substrates, the cyclic voltammetry highlights stronger adsorption features related to the adsorption of protons (H⁺) in the 0.1-0.3 V vs RHE potential range, in comparison to the sample prepared on MgO(110)/Pt(110) or on α-Sapphire(0001)/Pt(111) either the samples are prepared with 10 % of oxygen or with 20 % of oxygen.

For the NiO(100)-10% sample, it should be mentioned that the decrease in current down to -8 µA/cm², in the 0.1-0.3 V vs RHE potential range, as the current measured with quasi-static experiments and cyclic voltmetry are similar (Figure 8.7, left part). At such potential, the hydrogen reduction reaction is very unlikely. Thus, the continuous reduction of protons could be associated to a non limited reduction reaction: H⁺+e⁻ → ”H” but the final product cannot be assigned to a chemical compound with certitude. In addition, the partial reduction of surface Ni²⁺could also be another option to explain the current behaviour of the NiO 100-10% surface.

For the samples prepared with 10% of oxygen concentration, the EIS results show that seemingly hydrogen is adsorbed for the dominantly (100) oriented thin film but not for the other orientations (Figure 8.7, right upper part). The affinity of the dominantly (100) oriented sample towards hydrogen adsorption over the other orientations is also noticed when the sample is prepared with 20% of oxygen in the deposition chamber as the determined equivalent capacitive element is higher for the NiO(100)-20% sample than for the NiO(110)-20% and the NiO(111)-20% samples (Figure 8.7, right lower part). Moreover, the CV measurements display a more pronounced peak for hydrogen adsorption of the dominantly (100) oriented sample either prepared with 10 % of oxygen or with 20 % of oxygen (Figure 8.7, left part).

Therefore, it could be said that the (100) oriented sample has a sharper tendency to adsorb protons on its surface in comparison to the other orientations. This is in line with the in-situ XPS/UPS measurements (Section 8.2.3), which suggest that no charged layer is formed on the (100) surface and that hydrogen and hydroxide might be adsorbed equally on the (100) oriented surface. On the contrary, water adsorption experiments in vacuum on the (110) and (111) oriented NiO thin films could lead to the formation of a dipole where electrons are facing the positive hydrogen species (H⁺) within the first layer of adsorbed water (Section 8.2.3). This assumption is supported by the electrochemical measurements, which show that these two surfaces are not that much interacting towards hydrogen adsorption.

XPS measurements after electrochemical characterization

Looking back at Figure 8.4, it can be observed that the valence band position after the electrochemical measurements is not modified in comparison to the valence band maximum measured after surface exposure to water in vacuum.

Moreover, the difference of spectra of the Ni 2p region after electrochemical

measurement with as deposited thin film shows a marginal swelling of the density of state in the 855-860 eV region in comparison to the difference of spectra after surface exposure to water in vacuum. In the meantime, the shoulder at higher binding energy in the O 1s region is substantially increased.

The increase of the shoulder is more important for the NiO(110)-20% and the NiO(111)-20% sample. Thus, it could be assumed that the intensity of the shoulder at higher binding energy in the O 1s spectrum, after the electrochemical experiments might be closely related to the presence of defects and not necessarily to the intrinsic properties of a specific surface orientation.

The difference of the Ni 2p spectra displays a slight increase of the photoelectron emission between 854 and 860 eV in comparison to the water exposure experiments in vacuum. It can be assumed that the layer of adsorbates leading to the a nickel hydroxide layer on the oriented NiO surfaces does not vary much during the electrochemical experiments. In comparison, the shoulder in the O 1s region increases substantially. Thus, it can be assumed that the electrochemical experiments generates additional adsorbate species but bonded to the hydroxide already adsorbed on the NiO surface.

This additional layer of adsorbate would not directly interact with the NiO surface but could be mitigated in the presence of defects on the surface as this shoulder is larger if the defective state O 1s(Def.) is visible in XPS spectra of the as-deposited samples.

8.4 Conclusion

Differently oriented NiO films were prepared by reactive magnetron sputtering at 10 % and 20 % oxygen concentration onto heated Platinum (100), (110) and (111) oriented thin films grown onto MgO and sapphire substrates. The films are analyzed in-situ using photoelectron spectroscopy before and after surface water dissociation in vapour phase and by electrochemistry in an oxygen-poor and 0.1 M NaOH solution where experiments are carried in the window stability of water to study in-operando adsorption processes.

The similarity of the UPS results obtained on the most oriented thin films with what is provided in the literature suggests that an orientation could support a specific surface electronic fingerprints and even though our thin films are not perfect, their electronic surface properties could originate from the dominant surface orientation. In particular, the dominantly (100) oriented NiO thin films, displays a O 2p[z] electronic state not visible in dominantly (110) and (111) oriented NiO thin films.

The XP spectra reveal that water is absorbed in a bi-layer fashion on the as-prepared oriented NiO thin films. The first layer would be made of dissociated water molecules and the second layer, made of undissociated water molecules, is interacting with the first layer. The second layer forms a dipole structure (polarized water molecule), which would explain the decrease of the workfunction measured after water exposure. The dissociation of water within

the first layer leads to the formation of adsorbates. These adsorbates adsorb differently according to the dominant surface orientation. Thus, it has been assumed that hydroxides (OH⁻) and protons (H⁺) are adsorbed equally on the dominantly (100) oriented thin films but only the hydroxides might be adsorbed on the dominantly (110) and (111) oriented thin films.

Finally, the study of the adsorption reactions in the window stability of water evidences that the dominantly (100) oriented sample provide adsorption sites for hydrogen and hydroxide. On the contrary, the dominantly (110) and (111) oriented thin films show less reactivity towards the adsorption of hydrogen. Also, the presence of defects on the surface might be indispensable to enhance adsorption on the NiO surfaces as samples prepared at high oxygen concentration develop electrochemical features suggesting larger catalytic activity of the surfaces.

CHAPTER 9

Oriented NiO thin film activity towards the oxygen evolution reaction

Summary

Oriented NiO films (111), (100) & (110), reactively DC-sputtered on oriented platinum films on top of MgO substrates were studied electrochemically and were characterized on a rotating disk electrode for their performance towards the oxygen evolution reaction (OER). From the results, two electrochemical domains can be defined according to their catalytic activities: high overpotential (> 1.55 V vs RHE) and low overpotential (< 1.55 V vs RHE). We found that at low overpotential, electrochemical performances are rather similar for all orientations, but on the contrary, at higher overpotential, the (110) sample has the best electrochemical performance, followed by (111) and (100) samples. During OER, a thin hydroxide layer, about 1-2 nm thick growing on top of the thin film surfaces, has been identified as being the active material for the electrochemical reactions. The (111) and the (100) oriented thin film samples would produce a nickel hydroxide having the same catalytic properties, but the (100) oriented thin film would be less covered by the nickel hydroxide offering lower catalytic site density, and, on the other hand, (110) samples would form a more stable and more catalytically active nickel hydroxide material over both (111) and (100) thin films. The observed differences is ascribed to the formation of theβ-Ni(OH)2phase on top of the (110) oriented thin film being more crystalline and more stable nickel hydroxide phase than theα-Ni(OH)2, which is supposedly produced during the OER on the NiO(111) thin film and, to a lesser extent, on the NiO(100) oriented thin film.

The scientific material of this Chapter served to compose the article:

R. Poulain, A. Klein, and J. Proost. Electrocatalytic Properties of (100)-, (110)-, and (111)-Oriented NiO Thin Films toward the Oxygen Evolution Reaction. The Journal of Physical Chemistry C, 2018

Readable online: https://doi.org/10.1021/acs.jpcc.8b05790. The author of this manuscript, cited also as first author of the article above would like to point out he did not access the final version of the article.

9.1 Introduction

Although the hydrogen evolution reaction (HER) is efficient, the oxygen evolution reaction (OER) provides bottleneck issues and hinders greatly the performance of water splitting devices. Indeed, OER involves four electrical charges where HER only requires two, which could considerably decrease the kinetics of the reaction. However, among all the parameters that can be tuned to optimize the reaction, the material used as electrode can have a great influence on the OER mechanism and so on the performance of a water-splitting device.

The crystal orientation could be one aspect for optimizing the efficiency of an electrode towards an electrochemical reaction. For instance, it has been proven that the thermodynamically unstable (100) orientation of IrO2

and RuO2 structure is electrochemically more active towards OER than the (110) orientation [211]. Up to now, no experimental study has been carried out on the NiO surface orientation towards the OER activity. However, theoretical and experimental studies on adsorbates suggest a relatively great catalytic properties discrepancy between the three NiO dominant orientations [58, 61, 63, 65–67, 79, 212].

NiO is a stiff element, able to resist to relatively aggressive environment.

Structurally, it adopts a stacked rocksalt like structure with a lattice constant of 4.17 A. Dominant surface orientations for NiO are the (100), (110) and (111) in the (hkl) Miller index formatting. As seen before (see Section 2.2.1), the (100) and (110) surfaces are non-polar, which means the electric charge with the underlying plane is neutral, while the (111) surface is polar because of the formation of a dipole perpendicular to the surface [52].

Also, during electrochemical experiments on NiO, a thin nickel hydroxide layer might grow. Nickel hydroxide forms layers maintained by a weak OH-OH liaison and oxidizes to oxy-hydroxide above 1.1-1.4 V vs RHE. β/β refers to a hydroxide/oxy-hydroxide couple when no water or ions intercalate the structure. If water intercalates, hydroxide crystallinity decreases and the hydroxide is said to be turbostratic. In such case the hydroxide/oxy-hydroxide phase transformation involves theα/γcouple [171]. Oxidation state of cationic sites inγ-oxy-hydroxide phase is 3.66 and might be due to the intercalation of external compounds as potassium in the structure [170]. The whole picture between the four species is described by the so-called Bode scheme (Figure 9.1). The Nickel hydroxide properties towards the OER has been largely studied theoretically and experimentally, and could be one of the best base material as alternative to rare earth based catalyst [32–35, 213–216].

Although higher control on the material properties are achieved with sputtered thin films, only few reports have been found on the use of NiO thin films towards OER. The systematic study on NiO thin film carried out by Miller et al. [218] highlighted that OER activity is increased with NiO sputtered

Figure 9.1: Bode diagram representing the phase transformation and relationship between the four states of nickel hydroxide derived compounds. Taken from Lyons et al. [217].

films when they present a high degree of crystallinity and when there is an excess of oxygen in the film composition. However, the study did not point out any relationship with the surface orientation and the hydroxide properties developing on top of the film.

Instead of carrying out a systematic study, we propose here to introduce electrochemical results obtained on oriented samples in order to give hints for preferential path for tailoring NiO based electrode for OER. Moreover, we would like to emphasize that NiO electrochemical properties might be closely related to the nature of the nickel hydroxide growing on the surface during OER.