Solar energy is currently dominated by photovoltaic devices, which have ridden massive economies of scale to price dominance. But these devices are not necessarily the best choice in all circumstances. Unless battery technology improves, it’s quite expensive to add significant storage to solar production. And there are types of transportation—long-distance rail, air—where batteries aren’t a great solution. These limitations have made researchers maintain interest in alternate ways of using solar energy.
One alternative option is to use the energy to produce a portable fuel, like a hydrocarbon or hydrogen itself. This is possible to do with the electrons produced by photovoltaic systems, but the added steps can reduce efficiency. However, systems that convert sunlight more directly to fuel have suffered from even worse efficiencies.
But a Japanese group has decided to tackle this efficiency problem. The team started with a material that’s not great—it only absorbs in the UV—but is well understood. And the researchers figured out how to optimize it so that its efficiency at splitting water to release hydrogen runs right up against the theoretical maximum. While it’s not going to be useful on its own, it may point the way toward how to develop better materials.
Why are materials terrible at using the energy in sunlight to split water? Consider everything they have to do. For starters, they need to be decent photovoltaic materials, efficiently converting photons into free electrons and the corresponding hole with a positive charge. The materials have to somehow keep those two charges from recombining as they make their way to the surface, where they can interact with water. Once the charges are at the surface, the material has to also act as a catalyst, breaking open water and releasing hydrogen and oxygen. This last piece isn’t simple, as the formation of oxygen is driven by holes while hydrogen production requires electrons, meaning the two processes have to be physically separated.
Finding a single material that fits all of these requirements is not a simple task. The basic material that’s being used here, strontium-titanium oxide, has been used for this process for decades and has never reached much more than 60 percent of its theoretical maximum.
The Japanese team’s approach was to tackle each of these inefficiencies, although it’s not entirely clear from the paper whether each of their solutions was entirely intentional.
To start with, their choice of material—SrTiO3—handles the efficiency of converting photons to electrons and holes. It’s extremely good at it, seemingly capable of doing so at nearly the maximum efficiency predicted by theoretical calculations.
And due to its history, people had identified ways of improving the transport of charges within the material. For this work, the researchers doped the material with aluminum. The aluminum atoms tend to settle into the defects that slow down the transport of charges, allowing electrons and holes to recombine. When aluminum atoms are present, they sort of paper over these defects, allowing the charges to move freely throughout the material.
At the surface
Where the SrTiO3 tends to fall short is in the catalysis. Critically, the paper shows how the authors managed to make significant improvements. A number of developed catalysts are good at driving the splitting of water. But the researchers still had to keep the electrons and holes separated as they made their way to the catalysts.
It turns out that the material did the work for the researchers. Through the course of their work, the researchers discovered that the electrons and holes show up on different areas of the surface of the SrTiO3. While the surface of the material looks even and smooth to the eye, different areas will expose different faces of the underlying crystal structure at the atomic level. And as it turns out, the electrons and holes go to different surfaces because of these differences.
Amazingly, the researchers seemed to figure this out by depositing an additional catalyst on top of the SrTiO3. They used a process called photodeposition, in which high-energy photons are used to help chemically link a substance to an underlying surface. In this case, the underlying surface is the SrTiO3 material, and the wavelengths used were the same ones that produce electrons and holes. As a result, the appropriate catalysts ended up linked to the same areas where the charges they needed were delivered.
For the hydrogen-producing portion of the reaction, the researchers used a rhodium-based catalyst that will work for either oxygen or hydrogen production. But it was combined with a chromium oxide that physically blocked oxygen from interacting with the catalyst. This ensured that the electrons ended up where the catalyst for the hydrogen reacted. These chemicals were deposited through a reduction reaction, ensuring they ended up where there was a supply of electrons.
Meanwhile, a cobalt-oxygen catalyst was deposited through an oxidation reaction, ensuring it was linked to the areas supplied with holes. As a result, this catalyst for oxygen production ended up deposited only where the holes it needed were supplied.
Summing the process up, the structure of the underlying materials delivers electrons and holes to different areas of the SrTiO3 material. The researchers figured out how to use that fact to link the appropriate catalysts specifically to those sites.
At the edge of theory
It’s impossible to tell how efficient each individual step is in terms of converting incoming photons to the end products, hydrogen and oxygen. The system can only be examined as a whole, and from this perspective, it’s extremely impressive: the overall efficiency is 96 percent of the maximum possible efficiency derived from theoretical calculations. Thus, each individual step of the process is likely to be operating nearly at the theoretical efficiency.
This news is fantastic—other than the part where UV photons are required for the process to work. The Sun produces much of its energy at non-UV wavelengths, and a lot of the UV light is filtered out by our atmosphere. So this particular material isn’t going to drive the hydrogen revolution.
The key thing about this research is that it has identified the principle by which we might create the catalyst that could drive such a revolution. There is a wide variety of materials that can use light, including at visible wavelengths, to catalyze hydrogen production poorly. There is a much larger collection of photovoltaic materials that might do the same if combined with the right catalysts. The work described here provides a recipe that might convert some of them to useful materials.
Get rid of the defects. Find a material where electrons and holes take different routes through the material. use the presence of electrons and holes to link the right catalysts to where they are supplied with charges. If that works with a better starting material, we could be producing hydrogen through an extremely simple system.