Growing global energy demand is pushing photovoltaic technologies to their limits. Researchers in Sweden have made a significant breakthrough in developing innovative solar materials with higher efficiency than those currently in use. These materials can be produced in extremely thin and flexible forms, making them ideal for use in both small devices, such as smartphones, and on the surfaces of large structures.
At Chalmers University of Technology, scientists are working on a promising type of solar material: halide perovskites. Although they show exceptional potential, they remain difficult to study and put to use.
Researchers are using artificial intelligence and machine learning models to study the behavior of perovskites more accurately. According to the International Energy Agency, electricity currently accounts for 20% of total consumption, but by 2050 it is expected to exceed 50%. This makes the development of sustainable and efficient technologies, such as new solar cells, essential. As lead researcher Julia Wiktor explains, simulations can now answer questions that remained unresolved until recently, helping to improve the materials for practical applications.
This category of materials can absorb and emit light with high efficiency. This gives them an advantage in both solar applications and optoelectronic products such as LEDs. However, their stability remains a challenge. One of the most promising materials is formamidinium lead iodide, but it degrades easily, limiting its commercial potential.
The solution may lie in mixing two different types of perovskites, but more needs to be learned about their properties to precisely control the resulting material.
The Chalmers team succeeded in describing in detail a critical phase of the material that previously could not be fully interpreted using laboratory methods. This phase, which is crucial to the material’s functionality and performance, was revealed through simulations and published in the scientific journal Journal of the American Chemical Society.
Sangeeta Dutta, a member of the team, said that the key was deciphering the low-temperature phase, which had remained unclear to scientists for years.
AI-Based Materials Modeling
The team specializes in creating computational models that simulate the behavior of complex materials. This allows them to analyze experimental scenarios under conditions that reflect reality. However, perovskites require enormous computational resources and precise simulation because of the complexity of their crystal structure.
By combining classical simulation with machine learning, the team was able to increase computational power, simulating materials with millions of atoms—instead of hundreds, as before. This brings the analyses much closer to real-world application conditions. To validate their models, the scientists collaborated with the University of Birmingham. They conducted experiments by cooling the material to –200°C to accurately replicate the conditions used in the models.
The scientists observed that formamidinium molecules become “trapped” in an intermediate state as the temperature drops, affecting the material’s stability. This information is crucial for the material’s future use and for optimizing its combinations with other elements.
Physicist Erik Fransson said that the knowledge gained from the study would make a significant contribution to better understanding and developing advanced solar materials.
At Solar Home, we use artificial intelligence models to calculate system output and the optimal placement of equipment on your property.
Request a quote now
At Solar Home, we use artificial intelligence models to calculate system output and the optimal placement of equipment on your property.
Request a quote now
