The sun blankets the Earth with enough photons every hour to meet the entire world’s energy needs for a year. The question is how to efficiently convert them into electricity. Even under small-scale, laboratory conditions the world’s best single-junction solar cells, the kind found in most solar panels, still max out at capturing 29 percent of the sun’s energy. That puts them just shy of the hard limit of about one third that solar researchers calculated half a century ago. But scientists studying photovoltaics—the process by which sunlight is converted into electricity—have also long suspected that this limit is not as hard as it once seemed.
The ceiling on solar cell efficiency, known as the Shockley-Queisser limit, is between 29 and 33 percent, depending on how you measure it. It assumes a single-junction cell, meaning it’s made using only one type of semiconductor and is energized by direct sunlight. To nose past the limit, researchers have tried stacking multiple types of semiconductors or using lenses to concentrate light so that the cell receives a blast hundreds of times more powerful than the sun. Earlier this year, the National Renewable Energy Lab set a world record when it used a six-junction solar cell and a beam 143 times more concentrated than sunlight to achieve a whopping 47.1 percent energy efficiency.
But this technology will never be deployed at scale. The reason, says Marc Baldo, a professor of electrical engineering and computer science at MIT, is that these ultra-high-efficiency, multilayer solar cells are far too complex and expensive to produce as solar panels. To actually get more solar energy on the electric grid requires figuring out how to hit the Shockley-Queisser limit with single-junction, silicon-based solar cells, which are comparatively easy and cheap to produce. Better yet would be finding a way to bump the limit higher. And after a decade of work, Baldo and his colleagues may have finally figured out how.
As detailed in a paper published last week in Nature, Baldo’s team coated solar cells in a thin layer of tetracene, an organic molecule that effectively splits incoming photons in two. This process is known as exciton fission and means that the solar cell is able to use high energy photons from the blue-green part of the visible spectrum.
Here’s how it works. Silicon solar cells generate an electric current by using incoming photons to knock electrons from the silicon into a circuit. How much energy does that take? It depends on an attribute of the material known as its bandgap. Silicon’s bandgap corresponds to infrared photons, which carry less energy than photons in the visible part of the electromagnetic spectrum. Photons outside silicon’s bandgap essentially go to waste. But here’s where the tetracene comes in: It splits blue-green photons into two “packets” of energy that are each equivalent to an infrared photon. So rather than each infrared photon knocking free one electron, a single photon in the blue-green spectrum can knock free two electrons. It’s essentially getting two photons for the price of one.
This new cell represents a fundamentally new approach to a well-known truism in photovoltaics research: if you want to pass the Shockley-Queisser limit, you have to capture energy from a wider range of solar photons. Because this cell doesn’t rely on an expensive stack of materials with different bandgaps to broaden its range, it might ultimately be more practical too. Baldo says that using tetracene could bump the theoretical energy efficiency limit up to 35 percent—higher than was ever thought possible for single-junction cells.
Though the addition of tetracene is conceptually simple, implementing it was less so. The reason, says Baldo, is that if you put the tetracene directly onto the silicon, they interact in such a way that kills the electric charge. The challenge for Baldo and his colleagues was finding a material that could be sandwiched between the two materials to allow the energy packets to flow from the tetracene to the silicon. The theoretical literature gave them little guidance, so the team engaged in a lengthy process of trial and error to find the right interface material. This turned out to be a layer of hafnium oxynitride just eight atoms thick.
But this cell hasn’t bested any records yet. Its efficiency was about 6 percent in tests, so it has a long way to go before it can compete against existing silicon solar cells, let alone show up on a rooftop. But this work was only meant as a proof of concept of exciton fission in a solar cell. To bump the cell’s efficiency higher, Baldo says, will require some engineering work to optimize it for exciton fission.
In this sense, what the MIT team demonstrated wasn’t so much a competitive technology but a new tack for going beyond the limits of existing photovoltaics, says Joseph Berry, a senior scientist at the National Renewable Energy Laboratory. “What’s cool here is that this is a fundamentally different approach from traditional photovoltaics,” he says. “It’s an idea that’s been around for a long time, but hadn’t been translated into any kind of functional device.”
Berry and his colleagues at NREL are exploring other ways of advancing solar cell efficiency without the added complexity and cost of multi-junction cells. One of the most promising directions being explored by Berry are perovskite cells, which use synthetic materials that have structural properties similar to the naturally occurring mineral Perovskite. The first perovskite solar cells were only produced a decade ago, but since then they have witnessed the fastest efficiency gains of any type of solar cell to date.
Perovskite cells have a number of advantages over traditional silicon solar cells, says Berry, in particular their tolerance for material defects. Just a few unwanted particles on a silicon solar cell can render it useless, but perovskite materials still function well even if they’re not perfect. They also handle photonic energy more efficiently than silicon. Indeed, one of the main reasons silicon has dominated solar cell technology is not because it’s the best material for the job, but simply because scientists know so much about it due to its widespread use in digital technologies.
So far, none of these next-generation solar cells have found their way into commercial products. Almost all of the solar panels currently in operation are using traditional single-layer silicon cells, which have been proven to withstand the elements for decades. Getting perovskite-based solar panels into the field will require demonstrating that they’re stable and can last for 20 or more years. Berry says a number of companies have already deployed small-scale perovskite panels, which he hopes will pave the way for wider adoption down the road.
Looking to the future, Berry says it’s conceivable that the exciton fission technology under development at MIT could be combined with perovskite solar cells to increase their efficiency. “It’s not an either/or proposition,” Berry says, but first exciton fission must prove that it’s efficient enough for real-world applications. Ultimately, getting more sunlight on the grid will likely involve a suite of solar technologies, each with its own advantages.