As global electricity use soars, the need for clean, efficient energy sources has never been greater.
According to the International Energy Agency, electricity will account for more than half of the world’s energy use within 25 years, up from around 20% today.
To meet this demand sustainably, researchers are racing to develop new solar cell materials that are not only more efficient but also thin, flexible, and versatile—capable of powering everything from smartphones to skyscrapers.
One of the most promising options is a family of materials called halide perovskites.
These compounds can absorb and emit light with remarkable efficiency, making them ideal for solar cells and even LED lighting.
But they come with a major drawback: they are unstable and can degrade quickly. To unlock their full potential, scientists need to understand how these materials behave at a fundamental level.
A team at Chalmers University of Technology in Sweden has now made a breakthrough in this quest.
Using a combination of advanced computer simulations and machine learning, they have gained new insights into a compound called formamidinium lead iodide, one of the most exciting yet puzzling halide perovskites.
This material has excellent electronic properties but is limited by its instability. Researchers have tried mixing it with other perovskites to improve its durability, but until now, the details of its behavior were unclear.
The Chalmers group focused on the compound’s low-temperature phase, a structural state that has long baffled scientists because experiments alone could not fully explain it.
Their work has now clarified this missing piece of the puzzle, providing a deeper understanding of how the material changes as it cools.
“The low-temperature phase of this material has long been a missing piece of the research puzzle and we’ve now settled a fundamental question about the structure of this phase,” explained Chalmers researcher Sangita Dutta.
To achieve this, the team pushed the limits of material modeling. Traditional simulations could only handle hundreds of atoms for short timescales, but by adding machine learning to their methods, they were able to simulate millions of atoms over much longer periods.
This leap in scale allowed them to replicate more realistic scenarios and match their results with lab experiments.
Working with collaborators at the University of Birmingham, the team cooled samples of formamidinium lead iodide down to –200°C, confirming that their simulations reflected what happens in real materials. They found that as the compound cools, its molecules become stuck in a semi-stable arrangement—a key detail that helps explain its quirks.
These insights, published in the Journal of the American Chemical Society, could guide the design of more stable perovskite mixtures, bringing us closer to practical, long-lasting solar cells.
As researcher Erik Fransson from Chalmers put it, “We hope the insights we’ve gained from the simulations can contribute to how to model and analyze complex halide perovskite materials in the future.”
By piecing together this long-missing part of the perovskite puzzle, scientists are edging closer to solar materials that could transform how we power the world.
