Computational modelling and simulations of materials for photocatalytic hydrogen production

Abstract

Photocatalytic hydrogen production has been of interest since Fujishima and Honda first demonstrated it in 1972 with the help of a titanium-oxide electrode. Thanks to the rapid development of nanotechnology, this method has received much attention in recent years due to the need for cost-effective green hydrogen production. At the same time, solar cell technology has seen swift improvement in efficiency and reduced production costs due to increased research and political goodwill. However, there are still some significant issues to be addressed regarding energy storage and availability during dark or cloudy periods.

A solution would be to store the solar energy in hydrogen through electrolysis or photocatalytic water splitting. Although electrolysis is the most developed technology, it is still quite expensive, requiring a connection of the electrolysers to the electrical grid or a direct electrical connection to a solar farm. This makes it primarily suitable for developed countries with an extensive electrical grid and high-standard infrastructure. Developing countries, on the other hand, which are often located in regions with a lot of sun, lack an electrical grid of high quality to transport electricity from solar farms to the electrolysers. In these cases, photocatalytic hydrogen production could be the solution, as it would make it possible to produce hydrogen directly from the sunlight and transport it using existing infrastructure. Photocatalytic hydrogen production effectively combines the solar absorption abilities of photovoltaics with electrolysers’ ability to split water molecules into oxygen and hydrogen. Some challenges exist before the technology is viable for large-scale production facilities, such as low solar-tohydrogen efficiency and technology maturity.

In this project, we have used computational models and simulations to investigate materials and their structural, optical, electronic, and photocatalytic properties to address these issues. This was done by approximating solutions to the Schrödinger equation with the help of the Vienna Ab Initio Simulation Package (VASP) and our own developed tools. These results were analysed using post-processing tools in combination with physics, material science and chemistry. The applied computational models and simulations are incredibly reliant on computing power, which necessitates high-performance computing, parallelisation, and the development of efficient numerical methods.

The scientific contribution of this project is threefold:

First, we thoroughly reviewed state-of-the-art experimental and theoretical research on T iO2 based photocatalysts. In this work, we identified four major challenges that must be overcome for the field to advance further:

1. No standardised measurement setting for hydrogen production rates.

2. The intrinsic properties of T iO2 are not good enough, other materials can easily outperform T iO2 based photocatalysts.

3. Lack of cooperation between theoretical and experimental work. Not playing on the strength of the two approaches to complement each other.

4. There does not exist scalable photocatalytic hydrogen production facilities.

We identified other materials that could outperform T iO2 based photocatalysts on cost, efficiency, and lifetime through computer simulation studies and our review article. The most interesting candidates were perovskites and transition metal dichalcogenides such as MoS2.

Secondly, we investigated the structural and electronic of perovskite materials, CsPbI3 and CsSnI3, to determine their suitability for photocatalytic applications. In order to tackle the challenges such as toxicity and long-term stability issues faced by well-known lead-based organic perovskites, we carried out an in-depth analysis of the properties of lead-free double perovskites, Cs2AgBiBr6. To enhance the properties of Cs2AgBiBr6, we also conducted another numerical study by substituting Br with other halide atoms in Cs2AgBiX6 (X = Br, Cl, F, and I). These studies show that Cs2AgBiX6 (X = Br, Cl, F, and I) can be prominent candidates for photocatalytic and photovoltaic applications through clever substitutions of the halide component.

Thirdly, we investigated MoS2 as a potential photocatalyst. Fourteen different polymorphs were proposed and analysed for the very first time using first-principle calculations based on density functional theory (DFT). Seven of the polymorphs (1H, 2T, 2H, 2R1, 3Ha, 3Hb, and 4T) were both mechanically and dynamically stable with indirect bandgaps ranging from 1.87 eV to 2.12 eV. This work was extended as we looked into how dopants would influence the photocatalytic properties of MoS2. We used the stable polymorphs from our earlier work, and all seven showed promising results regarding their d-band model and Gibbs free energy. We chose to substitutional dope 3Hb with Al, Co, I, N and Ni. Replacing one Mo atom with either Al, Co, I, N and Ni lowered the Gibbs free energy by a factor of ten. However, only 3Hb with one Mo atom replaced with Al or Ni were stable structures. This shows that MoS2 through doping is a photocatalyst that, with further optimisation, could be used in large-scale photocatalytic hydrogen production.

We expect that the results from the studies in this project will result in new and efficient low-cost materials that will help push the field of photocatalytic water splitting further.