Metal-halide perovskite solar cells have seen tremendous growth and development in recent years mainly due to their impressive power conversion efficiency, now at a record 23.7%. However, the first expected commercial perovskite solar cells will be tandem devices, in which the perovskite rests on top of silicon. The materials absorb different parts of the solar spectrum, making tandem devices highly efficient. However, tandem devices that use perovskites for both layers would be easier to make and cost much less. The challenge is finding perovskites with silicon’s ability to capture low energy light.

Now by chemically tailoring a tin- and lead-based perovskite material, researchers have made an efficient low-energy absorber. They have combined this material with a conventional top perovskite layer to make tandem devices with a power conversion efficiency of 25%.

All-perovskite tandem solar cells could reach efficiencies over 30%. In tandem solar cells, the top wide-bandgap layer absorbs higher energy wavelengths and the bottom layer soaks up the remaining lower-energy infrared light. Lead-based perovskites have the wide bandgap needed for a top cell in a tandem device. For a lower perovskite layer, the most efficient way to narrow the bandgap is to replace some or all of the lead with tin. But solar cells made with narrow-bandgap tin-lead perovskites suffer from low efficiency and stability. This happens because the tin combines with oxygen and forms defects that trap charge carriers.

Joseph Berry and Kai Zhu at the National Renewable Energy Laboratory, Yanfa Yan of the University of Toledo, and their colleagues have found a way to overcome that problem. They add an organic compound guanidinium thiocyanate (GuaSCN) to thin films of (FASnI₃)_0.6(MAPbI₃)_0.4, where FA is formamidinium and MA is methylammonium.

The GuaSCN additive forms two-dimensional layers that coat the perovskite crystals. This protects the tin from oxidation, leading to 10 times less charge-trapping defects. This means charges flow longer—the charge carrier mobility goes up to more than 1 µs—and contribute to electric output. The efficiency of the Sn-Pb perovskite is 20.5%; best ones made so far have been 18% efficient. The researchers used this with a conventional wide-bandgap top cell to make a 4-terminal device that was 25% efficient. “We asked how good an optoelectronic material and absorber a tin-lead perovskite can be,” Berry says. “And we found we could manipulate them in a way that they’re spectacularly good materials.”

One challenge is that the devices are not very stable. Their initial efficiency drops down to 88% after about 100 hours of continuous operation. This is an order less than necessary for practical devices. Berry says that they plan to improve this stability and device performance by improving contacts, and by engineering the perovskite some more with additives that control chemistry.

The best silicon-perovskite tandem solar cells to date, made by startup Oxford Photovoltaics in the United Kingdom, have an efficiency of 28%. So these new cells have a lot of catching up to do still, says Pabitra Nayak, a physicist at the University of Oxford. But it is still respectable for a new material and additive. “Controlling the defects in lead-tin perovskites is a challenge,” he says. “The addition of guanidinium thiocyanate seems to control defect densities in these materials which is essential for all-perovskite tandem cells.”

Read the abstract in Science.