Solar Fuels

I was first introduced to the idea of solar fuels after I finished my undergraduate degree. The funding for my PhD was yet to be confirmed and so I was persuaded to work on an unpaid summer project in the meantime. The project instantly caught my attention – the idea that you could use sunlight to make fuels was fascinating. It didn’t take much persuasion in the end. 

So, what fuels can be made using sunlight? To name a few examples, it's possible to make hydrogen from sunlight and water. Fuels such as methane, methanol and ethanol can be produced from sunlight, water, and carbon dioxide. Ammonia can be made from sunlight, water, and nitrogen from the air. This particular project was on focused producing hydrogen. The scientific phrase for this field of research is photoelectrochemical water splitting – so basically, using sunlight to break down water molecules to produce hydrogen and oxygen gas. Theoretically speaking, it takes 1.23 eV of energy to split water, although in reality, it can take over 1.8 eV due to energy losses and kinetic issues.  

Hydrogen is thought of as the fuel of the future because it can be burned to produce energy without the production of carbon dioxide (a greenhouse gas); the only biproduct of burning hydrogen is water. There are many other ways to produce hydrogen, most of which use a lot of natural resources and have a large carbon footprint, but if fuels like hydrogen can be made in a green way using solar energy, I think it would truly solve the worlds energy crisis. Unfortunately, as I discovered whilst working on this project, achieving this dream is far from easy and it is not as simple as I once thought! 

To utilise the energy from the sun, we need to use a type of material called a semiconductor. Semiconductors are types of material that have a conductivity in between that of a conductor (usually metals) and an insulator (e.g. glass). To understand why we need semiconductors to harvest the energy from the sun, its useful to become familiar with the “Band Theory” of solids. For a free atom, the electrons have discrete energy levels depending on the atomic orbitals they are in. In a solid that contains a large number of atoms, the orbitals overlap to form bands. The highest occupied molecular orbitals (HOMO) form the valance band, and the lowest unoccupied molecular orbitals (LUMO) form the conduction band. It is the gap between these two bands that determines if the material is an insulator (very large bandgap), semiconductor (medium sized bandgap) and a metal (no gap – continuous band). Please see the diagram below for a visual representation. 

If the bandgap of a semiconductor is between 1.6 – 3.1 eV, visible light provides enough energy to promote electrons from the valence band to the conduction band. During this process, an electron-hole pair is generated. The electrons and holes are separated and flow in separate directions due to the electric field generated by the semiconductor/electrolyte interface. These promoted electrons can be used to drive reactions. If the band gap is 3.2 eV or higher, then UV light (which has higher energy) will be required to promote the electron. This is not desirable as UV light only makes up around 5% of the solar irradiance on Earth, compared to >40% with visible light.  

For photoelectrochemical water splitting, it’s not only the bandgap that is important; the valence and conduction bands must be positioned appropriately in order to drive hydrogen evolution reactions (HER) and the oxygen evolution reactions (OER): 

H2O + 2(h+) --> 1/2 O2 + 2H+ (OER)

2H+ + 2e- --> H2 (HER)

As you can see from the diagram above, the conduction band of the semiconductor must be positioned above the HER potential to allow electrons to flow into the electrolyte to reduce the H+ ions. Also, the valance band must be positioned below the OER potential to allow electrons to fill the generated holes in the semiconductor. One example of a semiconductor which fulfils these criteria is titanium dioxide (TiO2) and it was indeed this material that was used to split water using sunlight for the first time by Japanese scientists Fujishima and Honda in 1972 in a paper called "Electrochemical Photolysis of Water at a Semiconductor Electrode".

TiO2 also had the advantages of being an inexpensive material with good chemical stability. However, one of the drawbacks of this material is that it has low electron and hole mobility, meaning that it has a high level of charge recombination. Once light is absorbed and an electron-hole pair is generated, these travel in separate directions to reach the electrode surface so they can participate in the desired reactions. But the slower these electrons and holes move through the semiconductor, the higher the chance an electron (or hole) will encounter another hole (or electron) and cancel out. Charge recombination is a major issue in photoelectrochemical water splitting, however, there are many strategies for overcoming this. Another major drawback of TiO2 is that it has a wide band gap of 3.2 eV, which means it can only absorb UV light. This led scientists to start searching for materials with smaller band gaps whilst retaining suitable band energy positions relative to HER and OER potentials. This search also led to the possibility of using a cell with two semiconductors rather than one, as shown in the diagram below:

These two semiconductors would have complementary light absorption characteristics to absorb a greater proportion of visible light. Also, using two semiconductors also simplifies the requirements for suitable band energetics. One semiconductor could have a suitable conduction band position for the HER, and the second semiconductor can have a suitable valance position for the OER.

My old research group (the Energy Research Laboratory (ERL) at Loughborough University) were heavily involved in this field of research and published a number of papers on a material called haematite, which was essentially iron oxide (Fe2O3) or rust. Fe2O3 was widely investigated for photoelectrochemical water splitting because it was cheap, easy to make, had a band gap of around 2.1 eV and had good stability. Our research group (in collaboration with others) wanted to understand this material better and try and improve it’s performance.

Here are a few of the key papers that the group published:

I joined the research group at a time when most of this research was concluded and the group wanted to explore new potential materials. I was involved in investigating a material called bismuth vanadate (BiVO4), which quickly became the next most widely investigated material in the literature after haematite for photoelectrochemical water splitting. This material had slightly better characteristics than haematite, however, it was not easy to make pure thin films of this material, at least by the technique I was using: a variant of chemical vapour deposition (CVD). It was challenging to synthesise this material in a pure form, due to formation of impurities such as V2O5. However, by optimising the synthesis conditions, we finally managed to make it. Several talented researchers from our group continued this research until we could get reliable and reproducible results:

I also collaborated with other research groups to investigate ways of optimising the synthesis of this material by different methods:

I hope with this simplified explanation that you, the reader, have obtained a basic understanding of this field of research and why time and public resources have been spent on it. I must admit that I haven’t kept up with this field of research in recent years, however the major challenges still remain:

  • Increasing the efficiency of the reactions, which involves maximising the conversion of sunlight to hydrogen and minimising energy losses due to recombination of electrons and holes, and improving the overall stability of the system.
  • Developing the ideal materials, which have suitable band gaps and band positions, good electron/hole conductivity and mobility and good chemical stability.
  • Interfacial engineering – optimising the morphology of the photoelectrode and development of surface catalysts to improve efficiency of the reactions.
  • Lowing cost and scaling up the technology.

To me, the future of this field of research is uncertain as other strategies are becoming more favourable. For example, having an external solar cell connected to a water electrolyser. This could overcome stability issues of having a semiconductor in an electrolyte, which leads to poor stability of photoelectrochemical systems.

If you wish to read in more detail about this field of research, I refer you to this review paper: