Chemical Disorder Suppresses Thermal Transport through Local Stress Redistribution in SrTiO3-Based High-Entropy Perovskites
By distilling a MACE teacher model into an efficient NEP for large-scale molecular dynamics, this study reveals that local stress disorder, rather than simple mass or size fluctuations, is the primary mechanism suppressing thermal transport in high-entropy SrTiO₃ perovskites through the loss of phonon spectral coherence.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Heat is a form of energy that moves through materials, often carrying unwanted warmth from one place to another. In the world of thermoelectrics, scientists try to stop this heat from flowing while keeping electricity moving freely. This is a delicate balancing act. If a material conducts heat too well, it cannot be used to convert waste heat into useful electricity. For decades, researchers have looked to a family of crystal structures called perovskites, specifically a compound known as strontium titanate, as a promising starting point. The challenge has been to disrupt the path of heat without breaking the path of electricity. One powerful strategy involves "high-entropy" engineering, a method where scientists mix several different types of atoms into the same spot within the crystal lattice. The idea is that this chemical chaos should scatter heat-carrying vibrations, known as phonons, and slow them down. However, while scientists knew this mixing worked, they did not understand exactly how the atoms were interacting to cause this slowdown, nor did they know which specific type of atomic disorder was doing the most work.
A team of researchers has now peeled back the layers of this mystery using a new approach to computer modeling. They focused on strontium titanate, a material where the atoms are arranged in a rigid, repeating framework. In this structure, a central titanium atom is surrounded by six oxygen atoms, forming a small octahedron. These octahedra share corners to build the entire crystal. The researchers replaced the strontium atoms on the outside with a mix of other elements, including lead, calcium, barium, and lanthanum, creating a highly disordered chemical environment. To see what was happening inside this chaotic mix, they could not rely on standard computer simulations, which are often too slow or inaccurate for such complex systems. Instead, they developed a sophisticated workflow. They first trained a highly accurate but slow artificial intelligence model on data from quantum physics calculations. They then used this slow, accurate model to teach a second, much faster model how to predict the behavior of atoms. This allowed them to run massive simulations involving thousands of atoms over long periods, capturing the subtle, fleeting movements that determine how heat travels.
The results revealed a surprising hierarchy in how different atoms affect heat flow. The researchers found that simply adding more types of atoms to the mix did not always lead to a continuous drop in heat conductivity. While the initial addition of disorder caused a sharp decline in heat transport, adding more components to an already messy mix yielded diminishing returns. The thermal conductivity dropped quickly at first but then leveled off, remaining well above the theoretical minimum limit for how poorly a crystal can conduct heat. This suggested that the mere presence of many different atoms was not the whole story. The team then investigated why some specific atoms were more effective than others at blocking heat. They discovered that the size of the atoms mattered more than their weight. When they substituted strontium with calcium, which has a different ionic size, the heat flow was suppressed much more effectively than when they used lead, which is much heavier but has a size closer to strontium. This finding challenged the long-held assumption that the random difference in atomic mass was the primary driver of heat scattering.
To understand why size was so critical, the researchers looked deeper into the local environment of the crystal. They found that the mismatch in atomic size caused the surrounding oxygen framework to twist and distort in complex ways. This distortion created a landscape of local stress that varied from atom to atom. The researchers developed a way to measure this stress and found a direct link between the disorder of these local stresses and the reduction in heat flow. The more chaotic the stress landscape became, the more the vibrations carrying heat lost their coherence and scattered. In the simulations, the calcium-rich mixtures showed a much broader and more chaotic distribution of these local stresses compared to the lead-rich ones. This stress disorder effectively shortened the lifetime of the heat-carrying vibrations, turning sharp, organized waves into diffuse, scattered noise. The study also showed that the heavy lead atoms did introduce a different effect, creating a soft, low-frequency vibration that interacted with the heat carriers, but this was less effective at blocking heat than the structural distortion caused by size mismatches.
The work provides a clear map of how chemical disorder translates into physical properties. It demonstrates that in these complex crystals, the way atoms push and pull on their neighbors, creating a jumbled landscape of local stress, is a more accurate predictor of thermal performance than simply counting the number of different elements or measuring their mass differences. By using a refined artificial intelligence method to simulate these interactions, the researchers were able to see the invisible mechanics of heat transport. They showed that the key to suppressing heat lies not just in mixing ingredients, but in how those ingredients reshape the local mechanical environment of the crystal. This insight offers a new guide for designing better materials. Instead of just aiming for maximum chemical variety, future designs can focus on creating specific types of local stress and structural distortion that are most effective at scattering heat. The study confirms that while high-entropy mixing is a powerful tool, its success depends on the specific mechanical consequences of the atomic substitutions, revealing a path toward materials that can more efficiently manage thermal energy.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.