AI Uses Crystal Symmetry to Find New Antiferromagnets for Spintronics

A symmetry-guided AI model narrowed 110,000 generated crystals to four stable antiferromagnetic candidates, pointing to a faster computational route for discovering materials that could shape future spintronic memory technologies.

A recent study published in the journal npj Computational Materials introduced a generative deep learning framework for the inverse design of magnetic crystalline materials. Termed “Space Group Crystal Diffusion Variational Autoencoders (SG-CDVAE)”, this model incorporates crystallographic space-group information directly into its latent representation, improving the generation of high-symmetry crystal structures. By combining global crystal symmetries with local atomic arrangements, SG-CDVAE supports more targeted exploration of vast chemical design spaces and identifies materials with targeted magnetic properties. From an initial pool of approximately 110,000 generated crystal structures, the SG-CDVAE generation and high-throughput screening workflow identified 80 high-symmetry antiferromagnetic (AFM) candidates, demonstrating its potential to accelerate computational materials discovery.

Detailed molecular picture of tooth enamel reveals adaptations to diet

From chewing to chomping to grinding, teeth suffer from a lifetime of repeated mechanical stress. It makes sense, then, that enamel is one of the hardest natural materials. University of Wisconsin-Madison physics professor Pupa Gilbert and colleagues previously showed that the hydroxyapatite nanocrystals that make up enamel are arranged perfectly parallel to one another, like hairs in a ponytail, but their crystal lattices are not co-oriented—a structure that contributes to the biomaterial's resistance to fracture, also known as toughness. In a new study published in the journal Nature, Gilbert and her colleagues developed a technique to quantitatively measure enamel nanocrystal orientation angles across human and non-human primate enamel from different epochs, finding a strong correlation between how tough food is and the misorientation angle.

A spinel crystal structure exhibits unusual, pressure-induced superconductivity

Superconductors are materials that conduct electricity with an electrical resistance of zero. Superconductivity is generally observed when materials are cooled down to extremely low temperatures. In some cases, however, like in so-called high-temperature superconductors, this property emerges at higher temperatures. Researchers at the Center for High Pressure Science & Technology Advanced Research, Chinese Academy of Sciences and other institutes recently observed pressure-induced superconductivity in CuIr2S4, a spinel that typically becomes an insulator when cooled below about 230 K, meaning that electricity can no longer flow through it. Their paper, published in Physical Review Letters, shows that progressively tuning this material's crystal structure using pressure prompts the emergence of two distinct superconducting phases, dubbed SC-I and SC-II, with a maximum transition temperature of 18.2 K.

Chaos as a matter of direction: Researchers build layered material where order and disorder coexist

Some materials behave unexpectedly. They crack differently than expected, or react in ways that are hard to explain. The answer often lies in their atomic structure. Is it neatly arranged, as in a crystal, or disordered, as in glass? Researchers at the University of Twente have now created a material that is both simultaneously. In two directions it is disordered; in the third, perfectly ordered. Their findings have been published in Nature Communications. In materials science, the distinction between crystalline and amorphous has been fundamental for more than a century. Crystals such as salt or diamond have a strictly repeating atomic pattern. Amorphous materials such as glass lack that long-range order. "We use that distinction every day," says Mark Huijben, a researcher at the University of Twente. "But we often assume that order or disorder is a property of the whole material. Our work shows it can also be a matter of direction."

Silver nanoparticles enable assembly of a theorized, previously unobserved crystal metallic structure

Using finely tuned nanoscale building blocks, researchers from Brown University and the University of Michigan College of Engineering have stabilized a fleeting structural phase of matter that had been predicted theoretically but never before stabilized in a physical material. The new nanoparticle superlattice, described in the journal Science, freezes an elusive intermediate state between two of nature's most common crystal metallic arrangements. Beyond describing new details about how this transition works, the new structure exhibits extraordinary optical properties that could be useful in quantum computing or other quantum information systems. More broadly, the work provides a new recipe for using custom-shaped nanoparticles to engineer entirely new classes of materials with tailored properties.

How Disorder Regulates Heat Flow

A new theory linking atomic disorder to heat flow offers a strong foundation for predicting thermal conductivity in materials ranging from crystals to amorphous carbon. [...] Designing new materials for next-generation energy technologies requires a knowledge of the relationship between atomic structure and other thermomechanical properties such as heat transport. From solar cells to nuclear reactor moderators, many technologies depend on materials whose structures lie between perfectly ordered crystals and perfectly disordered glasses. Capturing this wide spectrum of disorder in a general theory of thermomechanical properties has proven difficult. Physicists have now devised a model of structural disorder based on the statistics of atomic arrangements, which reveals a clear relationship between disorder and the atomic vibrations that underpin thermal transport [1]. Using the relationship, the team, led by Michele Simoncelli at Columbia University, was able to explain the wide range of thermal conductivities of many structural forms of carbon. The researchers hope that this disorder model will open new pathways for assessing the macroscopic thermal conductivity of materials on the basis of their atomic structure.

Crystals in a new light: Research team proposes rethinking crystal structure analysis

Every crystal's shape is a mirror of the internal arrangement of its molecules, but the molecules in photoswitchable crystals can expand, twist and change properties—from their color to their electronic conductivity—with a simple flash of light. This has made them highly sought-after for applications like pharmaceuticals and data servers. But scientists have very little control over the shape that crystals take. After chemist Jason Benedict's team painstakingly grew photoactive molecules known as dithienylethenes into 19 distinct crystal structures, just two responded to light. "It was a real bummer," says Benedict, Ph.D., professor in the Department of Chemistry, within the University at Buffalo's College of Arts and Sciences. "I can assure you, there's not a lot of interest in non-photoactive crystals made from photoactive molecules." But, "after sitting back and licking our wounds a little bit," as Benedict puts it, they began to ask some questions.