Artificial Photosynthesis Moves on From Water Splitting to CO2 Reduction

Commercial ventures in artificial photosynthesis have struggled, but the science is marching on

5 min read

Water splitting device in which the solar collecting element can be angled
Photo: Dexter Johnson

The road toward commercial artificial photosynthesis has been a bumpy one. Stories like the so-called artificial leaf generated a lot of hype in 2011, but the company initially behind the technology—Sun Catalytix—soon abandoned their commercial efforts in 2012 when it became clear the economics simply did not add up.

While other companies that launched around this time, Hypersolar for example, have continued to try to make their technology work commercially,  the scientific community seemingly has had far better luck advancing the fundamental science of photoelectrochemical reduction.

This scientific effort largely has been organized in the United States under the Department of Energy’s Joint Center for Artificial Photosynthesis (JCAP).  IEEE Spectrum had the opportunity to sit down with scientists at the northern branch of JCAP, located at Berkeley National Laboratory. (The southern arm is at the California Institute of Technology in Pasadena.) Our discussion covered where the technology is at this point, what’s next, and how nanomaterials are helping to shape its development.

JCAP was founded in 2010 under an initial five-year U.S. Department of Energy contract with the broad aim of developing methods for producing hydrogen and carbon-based fuels using only sunlight, water, and carbon dioxide as inputs. That initial agreement was renewed in 2015 for another five years.

“In the first five years, our mission was pretty broad in terms of what kind of artificial photosynthesis we would actually work on,” explains Frances A. Houle, deputy director for science and research integration at JCAP North. “Initially, we worked primarily on water splitting, and we actually were pretty successful in meeting our goals there.”

Houle says that JCAP research was able to demonstrate technologies that were at least 10 percent hydrogen efficient, which is a ratio of power from chemical energy to power from sunlight. The devices also displayed high levels of stability and durability and could be made in a way that was actually scalable. While Houle notes that some experimental subsystems have demonstrated 18 percent hydrogen efficiency, these devices are quite different from what JCAP has been producing.

“Our systems not only provide water splitting but also have full product separation,” explains Houle. “Normally in water splitting, you would generate hydrogen and oxygen at the same time, and that can be an explosive mixture. So our devices have membranes embedded in the system and designed in such a way that you get good efficiency while providing complete separation of the products.”

After those first five years, JCAP applied a developmental guideline known as the Technology Readiness Levels (TRL) to the research developed thus far. The TRL scale runs from one to 10 with one representing fundamental research and 10 meaning it was ready for production and deployment.

The water-splitting research that had been conducted by JCAP was deemed to be at a TRL three where a lab device had been demonstrated and it was no longer fundamental science. At TRL three it was ready to transition to a developmental stage and will continue to be developed through funding from the Office of Energy Efficiency & Renewable Energy (EERE), according to Houle.

The charge for JCAP since 2015 has been to investigate carbon dioxide reduction through photoelectric chemical methods, not through electrolysis. “This is a very aggressive long-term project, very different in character than before,” says Houle.

The challenges in carbon dioxide reduction are far more complex, and the developmental path is much steeper than it had been with water splitting. The main issue with carbon dioxide reduction is that it usually produces a soup of products when what you really want is a specific fuel, like ethanol. The project is so recent within JCAP that no papers have yet been published.

Meanwhile, the most recent research in water splitting was reported last November in Nature MaterialsA water-splitting catalyst was engineered onto a semiconductor for artificial photosynthesis.

As a bit of background into this line of research, semiconductor materials have a bandgap that generates an electron-and-hole pair when it is struck by a photon with an energy level that is higher than the bandgap of the semiconductor. This is how semiconductor materials convert solar energy into electrical energy in photovoltaic devices, and it forms the basis of photoelectrochemical reduction.

Francesca Maria Toma, a principal investigator at JCAP, and her colleague Jason Cooper, a research scientist at JCAP, contributed to this paper and other work by focusing on semiconductor growth and characterization. Specifically, they have been looking into how silicon, III-V semiconductors, and metal-oxide semiconductors behave in artificial photosynthesis schemes.

“In essence we’re trying to lower the cost of the overall package by utilizing metal-oxide semiconductors. And [we’re] trying to improve their performance and understand what limitations may be involved in this new class of materials that are less understood and more prone to defects than classical systems that are well understood, like silicon,” says Cooper.

Toma explains that they are using the electrons to split water or reduce the CO2, but with the aim of doing it inside of the cell, where the holes can be used in the oxidation side of the reaction. “The way they do this is by looking at nanostructured materials in confined environments. Specifically what we’re trying to do is to utilize light absorbers such as copper oxide and make it nanostructured so that you can have higher photocurrent,” says Toma. “In this way, the reaction can be higher just because you have more photocurrent. We also have looked into making these confined environments in which you have either nanoparticles or interfaces [that] allow you to look at how your reaction is proceeding.”

Toma, Cooper, and their colleagues at JCAP have developed multiple device architectures that are in the pipeline. They have fabricated one device that’s totally assembled and on its own can be brought outside into the sunlight to generate hydrogen and oxygen. This device uses commercially available high-efficiency semiconductor materials combined in a way to stabilize and protect them from the caustic solutions that are present during water oxidation or the water reduction chemistry.

Echoing what Houle had explained earlier, Cooper acknowledged that a large portion of their work in producing a device like this is separating the semiconductor from that solution. “That was one aim in this project: how to take these high-efficiency devices and get them into a package so that they can be stable for many hours, generating lots of bubbles,” says Cooper.

In this configuration, the researchers used an epoxy separator with the catalyst mesh embedded in the epoxy. Thus the electrons that are generated in the semiconductor can be extracted into the catalyst layer and transferred to a solution.

The use of an epoxy represents a kind of bulk method. So to address that issue the researchers also employed atomic layer deposition (ALD) to thin that bulk epoxy down to 4 nanometers. These extremely thin layers in combination with silicon make it possible to have the catalyst itself serve as the protection layer.

All of this work, and more that is still to be published, was produced with the aim of water splitting. Now the research has shifted to carbon dioxide reduction and as a result, so too has the research direction. But the expertise generated thus far will still be applicable to carbon dioxide reduction, even though the challenges may be a bit more daunting.

Toma adds: “It’s not that we need to learn a totally new field that resides outside our expertise, because the backgrounds that we have are still important to understanding carbon dioxide reduction. The processes are very similar in the end.”

The Conversation (0)