Hydrogen From Sunlight

Smog-filled air, climate change, poisoned water: problems like these are all accelerating the search for green energy alternatives. One possible solution uses two of Earth’s abundant resources—sunlight and water—to produce hydrogen, a non-greenhouse gas that can deliver energy to fuel cells for powering vehicles or electronic devices without contributing to climate change.

Producing hydrogen using solar energy and water is known as solar water splitting, where the sun's energy is harnessed and stored in the hydrogen's chemical bonds. The process relies on the use of materials known as photocatalysts to split water into its two components, hydrogen and oxygen.  Although the technique is promising, the low efficiency of photocatalysts has hampered the large-scale use of solar water splitting. A group of ICTP scientists is hoping to change that.

In work published recently in The Journal of Chemical Physics, the researchers analyzed the characteristics of a particular photocatalyst, hematite, the iron oxide that makes rust. When coated with layers of a separate material, gallium(III) oxide, hematite reacts in a mysterious way: it seems to increase the production of hydrogen. To find out why and how gallium oxide provokes this reaction, ICTP physicists Kanchan Ulman, Manh-Thuong Nguyen, Nicola Seriani and Ralph Gebauer performed a series of atomistic computations to study the electronic and structural properties of hematite.

According to the researchers, it's a particularly odd effect because gallium oxide itself is not catalytic--it can't act as a photocatalyst by itself. "I find it fascinating that it makes the material more active by putting an inactive material on top," says Seriani, who along with his co-authors wanted to understand what was going on at the atomic level of this process. "Experimentalists can easily change macroscopic things, like pH or temperature, and see whether a lot of hydrogen is produced or not," explains Gebauer. "But this does not tell them what is really happening or why. In the end it is the mechanism at an atomic scale which defines whether the process works or not." He adds that thanks to their use of computational physics, he and his colleagues can very precisely model certain key steps that can help elucidate how the reaction works.

The theorists' calculations revealed the key role gallium oxide is playing for hematite in the water-splitting reaction: it renders inert the peculiar behavior of atoms at hematite's surface, behavior that was causing energy losses when hematite was used alone. When an electron is hit by sunlight and absorbs that energy, it pops up to a higher energy level, where it can then contribute to, in this case, the water-splitting reaction. But the surface of hematite is characteristically 'sticky'- it's harder for electrons to stay up at higher energy levels to have anything to do with splitting water. On the surface of hematite, the electrons easily and often fall back down to lower energy levels, in a process called recombination. When an atom recombines, all the energy it absorbed is lost, and can't be used to convert water to hydrogen. It's a different situation in the interior of hematite, considerably less sticky, which is why Ulman and his colleagues were excited that gallium oxide got rid of hematite's surface stickiness.

Hematite is inexpensive, naturally common, stable in water, and absorbs a good amount of the sun's energy to use to split water. If coating hematite with gallium oxide gets around one of hematite's main disadvantages as a photocatalyst, it's looking even better as a candidate for use at a large scale. These new calculations point at a way around hematite's surface state stickiness issue, but at the same time raise a whole new set of questions.

"In this study we showed that the surface states are not present when you add gallium oxide, but now the question is what happens when water actually comes and tries to separate out into oxygen and hydrogen," says Ulman. Gebauer explains: "This is very challenging because so far we have dealt only with the surface of hematite, which is a solid crystal. From our computation point of view, in a liquid everything's moving and changing and dancing around all the time, and you have to take this into account." The computational power needed to model the behavior of water in this photocatalytic reaction, he says, will be roughly 100-fold greater than the group used for their recent paper.

Which is why it's exciting that Seriani recently was awarded time on high-powered computers to investigate this. The grant includes 28 million hours of computation time from the Partnership for Advanced Computing in Europe, time that will be useful for modeling the complex movement of water and the full photocatalytic reaction using hematite. "This paper was 1 year of work, we do not want the next step to be 100 years of work," laughs Gebauer. With gallium oxide circumventing the recombination problem, and a host of computers ready to get to work on modeling the full reaction, the goal of converting the sun's energy to usable hydrogen is getting closer and closer. 

The title of the paper is Passivation of surface states of α-Fe2O3(0001) surface by deposition of Ga2O3 overlayers: A density functional theory study (http://dx.doi.org/10.1063/1.4942655)

--Kelsey Calhoun