In DepthSolar Energy

Cesium fortifies next-generation solar cells

See allHide authors and affiliations

Science  08 Jan 2016:
Vol. 351, Issue 6269, pp. 113-114
DOI: 10.1126/science.351.6269.113

In a world looking for better, cheaper alternative energy, the solar cell materials called perovskites are a bright hope. Their efficiency at converting sunlight into electricity is climbing faster than that of any solar technology before them. They're cheap and easy to make, can be manufactured roll-to-roll like newsprint, and can even be layered atop conventional silicon solar cells to boost their output. But they are fragile stars: Moisture, air, heat, or even prolonged sunlight makes them fall apart.

Now, these materials are toughening up. Over the past few months, three separate teams have reported that adding a dash of cesium to their perovskite recipes produces efficient solar cells that are far more stable when exposed to the elements. It's still too early to say whether cesium-spiked perovskites will withstand years or decades on a rooftop. Even so, “this is really a breakthrough for the field,” says Michael Graetzel, a chemist at the Swiss Federal Insitute of Technology in Lausanne, who leads one of the groups.

Perovskites are rapidly overcoming other shortcomings. The first perovskite-based solar cells, made 6 years ago by Japanese researchers, turned just 3.8% of the energy in sunlight into electricity, an efficiency well below that of silicon and other commercial technologies (Science, 15 November 2013, p. 794). But last month, at a meeting of the Materials Research Society here, researchers from South Korea reported evidence that their latest cells rival silicon, reaching a record 21.7% efficiency. Researchers are growing ever more hopeful that perovskite solar cells will soon approach 30% efficiency, rarefied territory now occupied only by costly gallium arsenide cells. “There seems to be no fundamental reason these materials won't achieve the efficiencies gallium arsenide has achieved,” says Henry Snaith, a physicist at the University of Oxford in the United Kingdom.

Yet unlike gallium arsenide, perovskites are made from cheap components—typically including the inorganic elements lead and iodine, together with one of two simple organic compounds, methyl ammonium (MA) or formamidinium (FA)—in a layered crystalline arrangement. All the chemistry needed to make them can be done without the expensive high-temperature setups or clean room facilities needed for many other solar cell materials. “This is really quite remarkable for the perovskites,” says David Mitzi, a physicist at Duke University in Durham, North Carolina.

Tandem solar cells, such as this 10-centimeter disk, combine the benefits of perovskite and silicon.


Perovskites are also exceptionally good at absorbing photons. As a result, cells can be made very thin, which further lowers costs. A thinner cell can also be more efficient, because electrons energized by sunlight are less likely to get hung up on imperfections in its crystalline lattice as they travel to an electrode, which feeds them to an external circuit.

Yet “high efficiency is meaningless if the cells aren't stable,” says Giles Eperon, a physicist at Oxford. So researchers around the globe are searching for more stable perovskite recipes. They've swapped out lead for tin, antimony, and bismuth: other metals nearby in the periodic table. And they've replaced iodine with bromine and chlorine. Most such changes reduce the materials' efficiency.

Perovskite solar cells reflect different colors depending on their composition; darker ones absorb more light.


But one change—replacing MA with FA, a slightly larger organic molecule—actually increases it. In the 12 June 2015 issue of Science, for example, Sang Il Seok, a chemist at the Korea Research Institute of Chemical Technology in Daejeon, South Korea, and his colleagues reported achieving more than 20% efficiency with FA-lead iodide perovskite solar cells. Cells with FA alone or a mixture of FA and MA also appear to be somewhat more stable than pure MA-lead iodide cells. When MA cells are taken out of a protective glove box, they degrade almost immediately, turning from black to yellow—a change that shows they are absorbing a narrower band of visible light. Mixed FA-MA materials also degrade, but more slowly: in minutes rather than seconds, Snaith says.

Now, several groups have found that adding cesium to their recipe seems to further stabilize the other components and helps the mixture retain its perovskite structure and black appearance. Nam-Gyu Park, a chemical engineer at Sungkyunkwan University, Suwon, in South Korea, first described the approach. In a 21 October 2015 online paper in Advanced Energy Materials, he and colleagues reported that replacing 10% of the MA with cesium yielded solar cells that held up “significantly” better to humidity and sunlight, although they did not give specific numbers.

These cells had a top efficiency of 16.5%, a step behind the best MA-only cells. But progress has continued. In a paper published online 3 December 2015 in Energy & Environmental Science, Graetzel and colleagues reported perovskite cells with a mix of MA, FA, and cesium that had an efficiency of just over 21%, a result verified by an independent lab. It seems clear that cesium is a key to making cells more stable and powerful. “I'm sure this is where the field is going to go,” Graetzel says.

Cesium-containing perovskites can also be coaxed to work well with silicon cells, as Snaith and colleagues report online this week in Science. Such combinations typically layer a perovskite cell on top of a silicon cell. The materials absorb different wavelengths of light, because they have different band gaps—the amount of extra energy they must absorb to shake electrons loose from their atoms so they travel through the material. Silicon, with a band gap of 1.1 electron volts (eV), is good at absorbing photons on the red end of the visible spectrum. Typical MA-based perovskites have a bandgap of 1.5 eV and thus absorb slightly shorter, or bluer, photons. Combining the two materials captures more of the solar spectrum—and thus more energy—than either could harvest alone.

To make tandem cells perform better, researchers want to widen their net by pushing the band gap of perovskites higher so they will absorb even bluer light. Groups led by Snaith and others have done that by replacing some or all of the iodine with bromine. But the changes made the cells more vulnerable to heat and light.

In their current study, Snaith's team replaced 17% of the FA with cesium. The resulting bromine-based perovskites could withstand exposure to prolonged light and high temperatures. The cells were also 17% efficient, and they had the wider band gap needed to work well with silicon. The researchers calculate that by layering the material atop a 19% efficient silicon cell, they could create a tandem cell with an efficiency of 25%. Snaith says silicon-perovskite tandems eventually ought to exceed 30% efficiencies.

So far, numbers like that have been the sole territory of gallium arsenide cells—devices so expensive they are used primarily in space, where it's worth paying a premium for higher efficiency. If cesium-spiked perovskites can bring high-efficiency solar cells down to Earth, the temperamental challengers' brightest days may still lie ahead.

View Abstract

Stay Connected to Science

Navigate This Article