Solar water splitting, in which water is separated into oxygen and hydrogen using sunlight, could be a clean and renewable way to produce energy, and researchers are busy looking for efficient photoelectrode materials for use in this process. One such material is haematite, or iron oxide (rust). Iron is cheap and abundant and haematite has a high theoretical solar-to-hydrogen conversion efficiency of 14–17%, which corresponds to a photocurrent of 11–14 mA/cm2. However, it is still unclear as to how defects in this material affect its ability to convert solar energy into hydrogen fuel and why some haematite electrodes appear to be more efficient than others.

In an effort to answer these questions, a team led by Michael Graetzel of the École Polytechnique Fédérale de Lausanne and Avner Rothschild of the Israel Institute of Technology decided to try and find out how structure is related to performance at the single nanostructure level in haematite. In an individual centimetre-sized water-splitting electrode, there can be billions of individual nanostructures, explains team member Scott Warren. Until now, however, researchers have typically studied the structure and properties of these nanostructures in aggregate and neglected individual structures. "We have shown, on the other hand, that there are important differences among the individual nanostructures in a single electrode – differences that determine whether a nanostructure is active for water splitting or entirely inactive."

New performance record for water splitting

By understanding these differences, the team then developed synthesis techniques that allowed them to make the most efficient haematite nanostructures for water splitting. The researchers used these "champion" nanostructures to fabricate photoelectrodes capable of generating a photocurrent of around 4mA/cm2. This is the highest photocurrent ever achieved for haematite and indeed any metal oxide photoanode, and a new performance record for water splitting, says Warren.

"Our approach involves looking at a single nanostructure and determining how its structure is unique," he told "We then measure how current moves through that single nanostructure and sometimes find that no current passes through, because of structural defects – that we identified in a transmission electron microscopy technique (TEM) developed in our lab (see images). We can then begin to understand what aspects of a structure impact current transport. It is this sort of information that was inaccessible to researchers in the past."

In analogy to the best performing champion solar cells, the Switzerland-Israel team has shown that some nanostructures are very good at water splitting and other less so. The distinguishing characteristic of the champion nanostructures identified in this study is that all the haematite crystals within the nanostructure are oriented in the same direction – something that allows electrons to travel rapidly through the material.

The research could help make better batteries, solar and fuel cells, says Warren, although there is still much work to be done. "While the haematite we looked at performs better than any other equivalent cheap and stable material, we still need to improve its performance," he said. "Continuing to characterise the nanostructure of haematite in this way will help us identify defects and other bottlenecks hampering its solar-water splitting efficiency."

The team presented its results in Nature Materials.