← Latest papers
🔬 materials science

Defects and Structural Evolution in CrSBr: Machine-Learning Potentials for Atomistic Simulations

This study develops and validates charge-aware machine-learning interatomic potentials to enable large-scale atomistic simulations of CrSBr, revealing that pre-existing sulfur vacancies significantly lower chromium migration barriers and identifying bromine as the most susceptible species to electron-beam-induced displacement, thereby establishing a framework for defect engineering in this multifunctional material.

Original author: Arsalan Hashemi, Daniil Kruklinskii, Xiaoya Chang, Tapio Ala-Nissila, Mikko Karttunen, Arkady V. Krasheninnikov, Mahdi Ghorbani-Asl

Published 2026-11-10
📖 5 min read🧠 Deep dive

Original author: Arsalan Hashemi, Daniil Kruklinskii, Xiaoya Chang, Tapio Ala-Nissila, Mikko Karttunen, Arkady V. Krasheninnikov, Mahdi Ghorbani-Asl

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where the tiny, invisible building blocks of matter can be rearranged to create materials that think, sense, and react to light and magnetism all at once. This is the promise of a new generation of electronics, moving beyond the silicon chips that power our current devices toward materials that are flexible, efficient, and capable of handling information in entirely new ways. At the heart of this potential lies a specific class of materials known as layered semiconductors. These are substances made of atom-thin sheets stacked like a deck of cards, held together by weak forces that allow them to be peeled apart or manipulated with great precision. Among these, a material called chromium sulfur bromide has recently captured the attention of scientists. It is special because it naturally combines the ability to conduct electricity with a built-in magnetic order, making it a prime candidate for devices that could switch between magnetic and optical states. However, to use this material in real-world technology, researchers must understand how it behaves when it is imperfect. In the real world, no material is perfect; atoms are missing, others are in the wrong places, and external forces like beams of electrons can knock atoms out of their spots. Understanding how these tiny flaws change the material's behavior is essential, but it is also incredibly difficult to study with traditional methods because the events happen too fast and involve too many atoms to track with standard computer models.

To solve this puzzle, a team of researchers has developed a new computational tool that acts as a highly accurate, fast-forward simulator for these atomic interactions. Instead of trying to calculate the behavior of every single electron in the material—a task that is too slow for large systems—they trained a machine-learning model to predict how atoms push and pull on one another. This model, built on a vast library of data generated from precise quantum mechanical calculations, learned the rules of the atomic game so well that it could simulate millions of atoms moving together with nearly the same accuracy as the much slower, more rigorous methods. The researchers used this tool to explore the life of chromium sulfur bromide under various conditions, specifically looking at how the material responds when it is squeezed, heated, or bombarded with electrons. They wanted to know how the material's structure changes when atoms go missing or when they are forced to move into new positions, and whether these changes could be controlled to create useful devices.

The simulations revealed a clear picture of how this material behaves when it is stressed. When the researchers applied pressure to the material, the model showed that it compresses unevenly, shrinking much more in one direction than the others, a behavior that matched real-world experiments perfectly. This confirmed that the new tool could reliably predict how the material would react to physical forces. But the most significant discoveries came when the team looked at defects, or missing atoms. They found that the material naturally contains many missing bromine atoms, which are the most easily knocked out of place. However, the story changes when a missing sulfur atom is present. The simulations showed that if a sulfur atom is missing, it dramatically lowers the energy barrier required for a chromium atom to jump out of its spot and move into the empty space between the layers. In fact, the presence of a missing sulfur atom made it about 1.1 electron-volts easier for a chromium atom to migrate, a massive difference in the atomic world. This finding provides a concrete explanation for why electron beams, which are often used to image or manipulate these materials, cause chromium atoms to redistribute and the material to change its crystal structure. The beam knocks out sulfur atoms first, and these missing spots then act as open doors, allowing chromium atoms to slip into new positions and trigger a transformation of the entire material.

The researchers also calculated exactly how much energy is needed to knock an atom out of place. They found that bromine atoms are the most fragile, requiring the least amount of energy to be displaced, while chromium atoms are much more firmly held in place. This hierarchy of stability explains why the material first loses bromine atoms under radiation, creating a landscape of defects that eventually allows the more stable chromium atoms to move. The study did not just identify these mechanisms; it also established a reliable framework for predicting them. By proving that their machine-learning model could accurately reproduce the energy costs of creating these defects and the barriers for atoms to move, the team provided a blueprint for future engineers. They showed that by controlling the types of defects present, perhaps through specific growth conditions or targeted irradiation, it might be possible to steer the material toward a desired state. This work does not claim to have solved all the mysteries of chromium sulfur bromide, nor does it suggest that the material is ready for immediate commercial use. Instead, it offers a validated, high-speed method for exploring the atomic-scale disorder that defines the material's future potential. It suggests that the path to building better magneto-optical devices lies in understanding and managing these tiny, chaotic rearrangements, turning what was once a source of instability into a tool for precise control.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →