The weakest of those interfaces is the one between the perovskite film used to absorb light and the electron transport layer, which keeps current flowing through the cell. 

“A chain is only as strong as its weakest link, and we identified this interface as the weakest part of the whole stack, where failure is most likely,” said Padture, who directs the Institute for Molecular and Nanoscale Innovation at Brown. “If we can strengthen that, then we can start making real improvements in reliability.”

To do that, Padture drew on his experience as a material scientist, developing advanced ceramic coatings used in aircraft engines and other high-performance applications. He and his colleagues began experimenting with compounds known as self-assembled monolayers or SAMs.

“This is a large class of compounds,” Padture said. “When you deposit these on a surface, the molecules assemble themselves in a single layer and stand up like short hairs. By using the right formulation, you can form strong bonds between these compounds and all kinds of different surfaces.”

Padture and his team found that a formulation of SAM with silicon atom on one side, and iodine atom on the other, could form strong bonds with both the election transport layer (which is usually made of tin oxide) and the perovskite light-absorbing layer. The team hoped that the bonds formed by these molecules might fortify the layer interface. And they were right.

“When we introduced the SAMs to the interface, we found that it increases the fracture toughness of the interface by about 50%, meaning that any cracks that form at the interface tend not to propagate very far,” Padture said. “So in effect, the SAMs become a kind of molecular glue that holds the two layers together.”

Testing of solar cell function showed that the SAMs dramatically increased the functional life of the perovskite cells. Non-SAM cells prepared for the study retained 80% of their peak efficiency for around 700 hours of lab testing. Meanwhile, the SAM cells were still going strong after 1,300 hours of testing. Based on those experiments, the researchers project the 80%-efficiency-life of the SAM cells to be about 4,000 hours.

“One of the other things we did, which people don’t normally do, is we broke open the cells after testing,” said Zhenghong Dai, a Brown doctoral student and first author of the research. “In the control cells without the SAMs, we saw all kinds of damage such as voids and cracks. But with the SAMs, the toughened interfaces looked really good. It was a dramatic improvement that really kind of shocked us.”

Importantly, Padture said, the improvement in toughness did not come at the cost of power-conversion efficiency. In fact, the SAMs actually improved the cell’s efficiency by a small amount. That occurred because the SAMs eliminated tiny molecular defects that form when the two layers bond in the absence of SAMs.

“The first rule in improving the mechanical integrity of functional devices is ‘do no harm,’” Padture said. “So that we could improve reliability without losing efficiency — and even improving efficiency — was a nice surprise.” 

The SAMs themselves are made from readily available compounds and are easily applied with a dip-coating process at room temperature. So the addition of SAMs would potentially add little to the production cost, Padture said.

The researchers plan to build on this success. Now that they’ve fortified the weakest link in the perovskite solar cell stack, they’d like to move onto the next weakest, then the next and so on until they’ve fortified the entire stack. That work will involve strengthening not only the interfaces, but also the material layers themselves. Recently, Padture’s research group won a $1.5 million grant from the U.S. Department of Energy to expand on their research.

“This is the kind of research that’s required in order to make cells that are inexpensive, efficient and perform well for decades,” Padture said.

Srinivas K. Yadavalli, Min Chen, Ali Abbaspourtamijani and Yue Qi were co-authors of the research, which was funded by the Office of Naval Research (N00014-17-1-2232 and N00014-20-1-2574) and the National Science Foundation (1538893 and 2002158)