Supercharging the silicon solar cell

Researchers at the Massachusetts Institute of Technology have shown that single-layer silicon solar cells could convert up to 35% of the energy in sunlight into useful electricity, a boost from the current efficiency ceiling of 29% (Nature 2019, DOI: 10.1038/s41586-019-1339-4). The trick, they say, involves finding the right combination of materials to allow them to exploit a phenomenon first predicted about 40 years ago.

Semiconductors like silicon excel at converting some, but not all, wavelengths of light into electricity. The range of wavelengths a semiconductor can use to generate electricity depends on an inherent electrical property of the material, called the bandgap. When these materials absorb wavelengths of light that pack more energy than those in that specific range, the energy gets lost as heat.

Solar cells could run more efficiently and put out more electricity if researchers could expand that range of wavelengths. There are tricks to do so, but none of them have gotten a good footing in the market, says Marc Baldo, an electrical engineer at MIT. One approach is the tandem solar cell, which stacks up multiple cells made of different semiconductors that each specialize in absorbing a particular band of the solar spectrum. These are expensive, and have complex wiring and internal inefficiencies, Baldo says. Like many other engineers, Baldo has been trying to find a way to make a regular solar cell work better.

To do so, they’ve turned to the fundamentals of what goes on in a solar cell. When a photon strikes a silicon solar cell, it excites a pair of negative and positive charges, an electron and a “hole.” This packet of charges is a quasiparticle called an exciton. In silicon, excitons rapidly separate as electrons join the flow of current. But in other materials, including the four-ringed organic semiconductor tetracene, excitons are capable of more exotic moves. Through a process called exciton fission, an exciton with a particular quantum spin can split into two lower energy excitons with different spins. Each of the daughter excitons carries half the energy of the parent.

This exciton fission could help expand the range of wavelengths a silicon solar cell can use to make electricity because the way tetracene splits excitons is a perfect match for silicon. The organic molecule’s daughter excitons have energy levels compatible with silicon, meaning tetracene could absorb light that silicon doesn’t know what to do with, get excited, and then donate that energy to silicon in the form of its daughter excitons. That donation would then excite silicon in a way to produce electricity.

Turning to tetracene for help with exciting silicon was first proposed in 1979 by chemist David Dexter (J. Luminescence 1979, DOI: 10.1016/0022-2313(79)90235-7). At the time, “it wasn’t clear what to do with this,” Baldo says. “The technology didn’t exist to build a silicon cell engineered for this process.”

Baldo’s lab has been looking for the right fix since 2009. Tetracene and silicon can only work together with a mediator; the challenge has been to find the right one. Silicon’s surface is a forest of bonds that must be coated with a protective layer. This protection, called the passivation layer, cannot be eliminated. But these protective layers typically disrupt the flow of charges between silicon and tetracene. Baldo says his group couldn’t find much guidance from theory, so they had to try many, many combinations of materials before they arrived at the right one.

After years of work, they found that a film of hafnium oxynitride about 8 Å thick (just under a nanometer) passivates the silicon but lets charges tunnel through from a tetracene layer. Baldo’s group has now shown that these three materials can couple together into a functional solar cell. Based on calculations, the team estimates that a tetracene-containing solar cell could have a 35% theoretical efficiency limit.

“There is nothing else like this in solar,” Baldo says. The way that tetracene splits excitons in half, making two new ones better suited for silicon solar cells—and the possibility of applying these basic physics in a real commercial product—is what has kept Baldo motivated to work on this problem for a decade.

Christopher Bardeen, a photochemist at the University of California, Riverside, is also excited about the team’s results. He points out that Baldo’s group has previously demonstrated the phenomenon in organic solar cells, and others have been working on it in other emerging solar technologies, such as dye-sensitized solar cells. But the recent research is the first demonstration in silicon, Bardeen says.

So far, the efficiency of the cells the MIT team has made is nothing to brag about, Baldo says. The devices perform worse than the silicon technology on the market today. The biggest challenge is that the tetracene daughter excitons tend to fuse back together faster than they travel into the silicon.

Columbia University chemist Luis Campos says his group and others are working on engineering tetracene and other organic molecules to extend the lifetime of the split excitons. Baldo’s paper, he says, “is a huge green light for organic chemists to say, yes, we should investigate this.” The exciton recombination problem is difficult one. “It’s a bit of a barrier,” he says. “But we just need to tunnel through it.”