Computational study of Raney nickel type catalysts for the alkaline oxygen evolution reaction

Abstract

The growth of the global population, together with an increase in pollution caused by the use of fossil fuels for energy production, has led to an environmental and energy crisis. This crisis necessitates the search for alternative sustainable and environmentally accommodating energy sources. One such alternative form of energy is hydrogen gas. Although hydrogen gas is a highly energy-dense and clean energy carrier, it does not occur naturally and needs to be produced. One of the cleanest hydrogen production methods is water electrolysis. However, much research is being conducted on the electrocatalysts that drive the two water electrolysis half-reactions, i.e. the hydrogen evolution reaction (HER) and oxygen evolution reaction (OER). The OER is more energy-intensive and requires an effective electrocatalyst. The material of these catalysts are dictated by the pH environment of the electrolysis. In an acidic environment, costly rare noble metals such as Ir and Ru are used. However, in an alkaline environment, base metals such as Ni, Co and Fe can be used. Literature is abundant with studies showing Ni and Ni-based oxides as effective catalytic materials for the OER in an alkaline environment. One Ni-based electrocatalyst of interest to this study is Raney nickel. Raney nickel, a porous nickel structure, is prepared from a bimetallic precursor, typically with Al as a secondary metal. This secondary metal is selectively leached out using an alkaline solution leaving a porous structure. However, the composition and morphology of the precursor can influence the final structure and composition of Raney nickel. Previous studies (experimental and computational) of Raney nickel focused mainly on three compositions (phases), namely Ni2Al3, NiAl and NiAl3. To gain a better understanding of the influence of the composition and the morphology of the precursor on the final structure of Raney nickel and, therefore, the catalytic activity of Raney nickel, there is a need to investigate a wider range of phases. That is what this study attempted to do by studying several Ni1-XAlX bimetallic phase surface structures. Before the Ni1-XAlX bimetallic surfaces could be constructed and investigated, an inclusive computational model needed to be developed. A Density Functional Theory (DFT) based model was developed using pure Ni and Al bulk structures and (111) surfaces. The model made use of the CAmbridge Serial Total Energy Package (CASTEP) included in the BIOVIA Software, Materials Studio 2020 (MS2020). The bulk model was developed using 2×2×2 supercells and validated by comparing the structural and electronic properties of these bulk structures with literature values. Surface slabs in the (111)-Miller plane were created for Ni and Al. Although most of the setting in the bulk model was used when surface slabs were optimised, the k-points were changed for the surface slab model. The surface slabs were validated by comparing the structural and electronic properties of the Ni (111)- and Al (111)-surface slabs with literature values. Next, the Ni1-XAlX bimetallic precursors were constructed using the Site Occupancy Disorder (SOD) program, which built all possible configurations with the given ratio of Ni and Al atoms (phases) and identified the unique configurations. The unique configurations were optimised using the General Utility Lattice Program (GULP). Further, SOD also allowed for the calculation of thermodynamic properties, such as probability distribution, configurational entropy, and the enthalpy of mixing. These properties were used to identify the most stable Ni1-XAlX bimetallic ratio and atomic configurations for further study. The bulk structures of the most thermodynamically stable unique configurations were optimised using the abovementioned DFT model. Surfaces slabs were generated by cleaving the optimised bulk structures along the (111)-Miller. These (111)-surface slabs were optimised and used to identify configurations for manually simulated leaching. The surface slabs for the manual leaching of Al were selected using calculated properties such as the work function, the surface energy, and the Ni and Al content. The manual leaching was done by removing a set of four Al atoms. The initial Al atom was selected by considering the nearest neighbour (NN) atoms on the surface and in the sub-surface layers. The subsequent Al atoms were chosen by considering NN atoms in the layers surrounding the cavity created by the previous removal. The leached surface slabs were optimised after each removal. The final leached surface slabs were those with the smallest average removal energy. Lastly, the simulated adsorption of the 𝑂𝑂𝑂𝑂− species was studied on the pure Ni (111)-surface slab and the final leached bimetallic (111)-surface slabs. The adsorption data for the Ni (111)-surface slab showed that Ni exhibited optimal adsorption energies for an OER catalyst. Comparatively, the adsorption data of five manually selected adsorption positions showed that the bimetallic (111)-surface slabs of Ni8Al8, Ni9Al7, and Ni11Al5 could be active catalysts for the OER. Of these Ni1-XAlX bimetallic phases, Ni8Al8 (NiAl) has been studied previously. However, Ni9Al7 and Ni11Al5 are newly identified phases that could be used for future studies of precursor composition and morphology influences on Raney nickel.