Electronic and electrocatalytic properties of nickel oxide thin films and interfacing on silicon for water splitting devices

The thesis entitled "Electronic and electrocatalytic properties of nickel oxide thin films and interfacing on silicon for water splitting devices" deals with the implementation of nickel oxide (NiO) at the anode of a photo-water splitting device for the oxygen evolution reaction (OER). The thesis can be tackled through three main parts. The first part consists in studying the surface electronic properties of NiO and its electrical behaviour, the second part deals with the catalytic properties of NiO towards adsorbates and the OER, finally in a third part, the Si/SiO2 interface has been studied as well as the deposition of NiO on top for assembling a functional photo-anode. Regarding the first part, the surface properties of nickel oxide thin films have been investigated by in-situ X-ray photoelectron spectroscopy (XPS) and ultra-violet photoelectron spectroscopy (UPS). It has been found that, according to the condition of preparation, which defines the concentration of doping in the nickel oxide thin film, the Fermi level can be varied from 1.1 eV to 0.6 eV while the workfunction can be varied from 4.5 eV to 5.2 eV. Eventually, a charge compensation mechanism of the defects is proposed. In collaboration with the EMAT department (Electron microscopy for Materials science) of the university of Antwerp, thin films prepared at room temperature have been studied by high resolution transmission microscopy and by high resolution electron energy loss spectroscopy. The study concluded the presence of a secondary oxygen-rich phase accumulating at the grain boundaries, which is unstable above 200 °C. This phase would be responsible for the high electrical conductivity reported for room temperature nickel oxide thin films. The instability of the secondary phase would be the origin of the electrical ageing process observed for such nickel oxide thin films. Then, in the second part, oriented nickel oxide thin films have been prepared at high temperature along the (100), (110) and the (111) direction and were subsequently fundamentally studied for in-depth understanding of the nickel oxide/electrolyte interface. The nickel oxide/electrolyte interface has been studied in-situ by XPS/UPS by exposing oriented surfaces to water in vacuum and also by carrying out electrochemical measurements in an electrolyte. In vacuum, it has been found that water adsorbs in a bi-layer fashion. The first layer in contact with the surface contains hydroxides and protons (originating from the water dissociation reaction), while the second layer contains undissociated water molecules. Supported by the electrochemical study on oriented surfaces in an electrolyte, it has been assumed that the (100) oriented nickel oxide thin film offers an equal number of adsorption sites for protons and hydroxides. On the contrary, the (110) and the (111) oriented thin films would offer primarily adsorption sites for hydroxides. Eventually, the electrochemical study of nickel oxide oriented thin films towards the oxygen evolution reaction shows that the (110) oriented thin film is the most active electrode followed by the (111) oriented thin film and then the (100) surface. The results suggest that a non-negligible nickel hydroxide layer grows on top of the nickel oxide surface during the oxygen evolution reaction and that the nickel hydroxide layer would sustain the electrochemical reaction. The interpretation of the results lead to the assumption that the (110) oriented nickel oxide thin film would stabilize the nickel hydroxide in a form, which is catalytically more active towards the oxygen evolution reaction than the nickel hydroxide growing on top of the (100) and the (111) oriented nickel oxide thin films. The nickel hydroxide growing on top of the (100) oriented surface might be less homogeneous and thinner than the nickel hydroxide growing on top of the (111) oriented thin film. However, the optimization of the catalytic properties of a nickel oxide based catalyst would be much more affected by the temperature of preparation. Thus, as a rule of thumb, it can be retained that, whatsoever the dominant orientation, best electrochemical performances are attained when nickel oxide thin films are prepared at room temperature and at relatively high oxygen concentration during sputtering. Finally in the third part, to interface nickel oxide by cathodic magnetron sputtering on silicon/silicon dioxide, it has been demonstrated that nickel oxide has to be prepared in such a way that it avoids the implantation of oxygen in the silicon dioxide, as for reactive sputtering depositions. A specific method to deposit nickel oxide by sputtering, referred to as metal layer oxidation (MLO), has been proposed and is basically split into two steps. The first step consists in the deposition of a metallic layer by sputtering in argon (oxygen free atmosphere), whereas the second step consists in oxidizing the metallic layer in an oxygen rich atmosphere while the cathode is off. The MLO method enables the elimination of the bombardment of the silicon dioxide by negatively charged oxygen ions when the sputtering is realized in the presence of oxygen in the chamber. Then, the silicon/silicon dioxide interface has been studied in the aim to realize a metal-insulator-semiconductor tunnelling junction with nickel oxide. The study of the silicon/silicon dioxide interface shows that the interface contains donor state, located in the top 2 nm of the silicon in the vicinity of Si/SiO2 interface, which is responsible for the pinning of the Fermi energy in silicon, especially when platinum is interfaced. When nickel oxide is deposited, by the MLO method, it is proposed that the donor state is ionized in totality. In consequence the band-deviation with nickel oxide when prepared by MLO is larger than with platinum. Moreover, the ionization of the donor state can lead to the formation of an intense electric field throughout the Si/SiO2 interface in the 100-500 MV/m range. At the end of the thesis, photo-anode structures based on silicon and nickel oxide have been fabricated by MLO and tested in a photo-water splitting cell. Although the devices provided positive response to light excitation, the experiments might suggest that the transfer of the charges from the silicon towards the catalytic site and the catalytic layer itself have to be improved. These last barriers should be taken into account in future works to achieve the realization of an efficient water-splitting device.