A research team from multiple institutions including the Ningbo Institute of Materials Technology and Engineering (CAS), Beijing Computational Science Research Center, and Hangzhou Dianzi University has developed a new strategy to design metallic glasses (MGs) that are both kinetic ally ultra stable and mechanically ductile. The study shows that by engineering specific spatial patterns of oxygen atoms within a zirconium-copper metallic glass, the material can resist structural change at high temperatures (kinetic stability) while retaining the ability to deform plastically without turning brittle. This discovery effectively decouples two properties that were previously thought to be inextricably linked, opening new avenues for creating high-performance amorphous materials.
For decades, the development of metallic glasses (materials with a disordered, glassy atomic structure instead of a crystalline one) has faced a fundamental dilemma. To make a glass resist heat and avoid transforming into a crystal (a property known as kinetic stability), scientists have traditionally had to cool it extremely slowly or use other methods that force its atoms into a very low-energy, tightly packed configuration. This thermodynamic stability makes the material rigid and strong, but it also makes it extremely brittle, causing it to crack and fail under stress. Conversely, more ductile metallic glasses, which can bend and deform, tend to be less stable and will crystallize more easily when heated. This trade-off has limited the practical applications of these otherwise promising materials. While some experimental results hinted at the possibility of separating these properties, the underlying mechanisms remained a mystery, preventing the rational design of better materials.
The Solution: Turing Pattern-Driven Oxygen Patterning
In this new work, an interdisciplinary team from Chinese researcher centres, found a solution not in changing the glass's overall energy, but in changing its internal architecture at the nanoscale. Using advanced computer simulations powered by machine learning (ML), the researchers introduced oxygen atoms into a copper-zirconium metallic glass. They discovered that oxygen, which has a strong affinity for zirconium, naturally self-organizes into specific patterns through a process similar to reaction-diffusion, a mechanism famously proposed by Alan Turing. These patterns form what the team calls "oxygen-centred pinned structures" (OPSs).
These OPSs act like tiny, localized anchors throughout the material. They drastically slow down the movement of atoms, granting the entire glass exceptional kinetic stability, raising its "onset" glass transition temperature by about 200 Kelvin. Crucially, the rest of the material, the metallic regions between these anchored points, retains the high-energy, loosely packed structure of a rapidly cooled glass. This preserves the "soft spots" necessary for plasticity. When stress is applied, these regions activate uniformly, allowing the material to deform and preventing the formation of catastrophic, brittle fractures. In essence, the team created a material that is a composite at the nanoscale: a stable, rigid scaffold interwoven with a deformable, ductile matrix.
The team also validated their computer models against real-world experiments. Their simulations successfully reproduced key experimental observations reported in the literature, including the clustering of oxygen atoms seen under electron microscopes, the characteristic rise in glass transition temperature with oxygen content, and the transition from brittle to ductile behavior observed in mechanically tested samples. This cross-validation confirms that the pattern-engineering phenomenon predicted by their models is not just a computational approach but a realistic and achievable materials design strategy.
The Future: Intelligent scaling-up
This research establishes a new paradigm for materials design. By treating dopants not as a uniform addition, but as a tool for creating specific nanoscale patterns, scientists can now tailor the properties of metallic glasses with unprecedented precision. The team successfully demonstrated that they could use the reaction-diffusion model as a "blueprint" to directly build atomistic models with the desired OPS patterns, bypassing the need for lengthy simulation searches.
This pattern-engineering approach is not limited to the zirconium-copper-oxygen system. It can be extended to other chemically complex glasses containing small, reactive elements like carbon or nitrogen. The ability to digitally design and simulate these patterns opens the door to additive manufacturing techniques, where materials could be built layer-by-layer with prescribed heterogeneity for optimal performance.
Potential applications are vast and include nanostructured alloys for aerospace, high-strength composites for automotive engineering, and radiation-resistant coatings for nuclear applications—anywhere materials are needed that can withstand extreme conditions without sacrificing mechanical integrity. This work transforms a long-standing materials science limitation into a tuneable design feature, paving the way for a new generation of amorphous materials with on-demand stability and plasticity.
The research has been recently published in the online edition of Materials Futures , a prominent international journal in the field of interdisciplinary materials science research.
Reference: Huanrong Liu, Qingan Li, Shan Zhang, Rui Su, Yunjiang Wang, Pengfei Guan. Turing pattern engineering enables kinetically ultrastable yet ductile metallic glasses[J]. Materials Futures , 2026, 5(2): 025603. DOI: 10.1088/2752-5724/ae3d4f
Materials Futures
Turing pattern engineering enables kinetically ultrastable yet ductile metallic glasses
12-Feb-2026