An ongoing criticism of solar energy is that it only works during the day. This problem is being tackled in a plethora of different ways, from oversizing solar systems and connecting them to battery backups and thermal storage to creating syngas (synthetic gas) via algae. New research from CalTech and the Swiss Federal Institute of Technology are looking at a new way of producing hydrogen and syngasses in another way, by using solar funnels.

A BBC report citing a paper appearing in the journal Science describes how the prototype device uses a quartz window and cavity to focus sunlight into a cylinder lined with cerium oxide. Cerium oxide, also known as ceria, is hygroscopic (meaning that it attracts and holds water molecules from the surrounding environment) and will also absorb a small amount of carbon dioxide. As the sunlight heats the ceria, it thermochemically breaks down the water and carbon dioxide pumped into the cylinder to produce carbon monoxide and hydrogen that can be converted to a liquid fuel.

Through a two-step process, the ceria (cerium dioxide) converts carbon dioxide or water into its constituent elements. “Ceria is a metal oxide, what that material will do when heated is it will release oxygen.…It happens at high temperatures, when we cool it back down it wants to absorb oxygen,” Haile said. The ceria replaces the oxygen by stripping it from the supplied material, carbon dioxide or water, thereby creating carbon monoxide—used for syngas, or hydrogen—which can be used directly. Either resulting fuel could be used to store the sun’s energy for use in power generation.

The funnels can be small, but they’re not nano-sized. “It’s like a sponge it’s porous and the gases flow through it,” Haile said. But “it’s not nano because these temperatures are too high for nano-structures.”

At this point, the material isn’t efficient enough for commercial use. The prototype is inefficient, converting between 0.7 percent and 0.8 percent of the solar energy in the funnel into fuel. With advances that could change. “We calculated efficiency should be between 15 percent and 19 percent,” according to Haile. “We’re working with University of Minnesota on that. Right now it’s limited by the thermal design of the reactor. We need a better thermal design,” she said.