Two ultrathin sheets of strontium titanate can form a square grid of crystal defects when they are stacked, rotated and strongly bonded, according to a study led by the Technical University of Denmark. The sheets are about 10 nanometers thick. A report by Tore Vind Jensen describes the finding, published in Nature, and its possible relevance to future electronics and energy storage.
The result resembles a familiar effect called a moiré pattern: overlapping two regular meshes at an angle creates a larger repeating pattern. But in this material, the atoms do more than produce an apparent pattern from two overlaid lattices. Strong bonding lets atoms at the sheets’ interface rearrange to fit together. The mismatch that remains gathers along narrow lines of distorted crystal, called dislocations, forming a square grid.
The grid’s spacing can be adjusted by changing the twist angle between the sheets. That gives researchers a way to tune the size of the pattern, although the report does not establish a practical device or a useful operating range. Strontium titanate is described as a hard, chemically stable material often used as a model oxide in electronics.
The study also points to an electrical effect. Modeling found that distortions around the grid could slightly separate positive and negative charges, a property called polarization. The modeled polarization forms swirling patterns. Because polarization control is relevant to memory and energy storage, the patterns may offer a starting point for future device designs. They are not evidence that a working memory or energy-storage device has been made. Whether the grid lines can conduct electricity remains an open question for future experiments.
The grid was difficult to identify. Initial microscope images appeared to show an ordinary moiré pattern, without a clear view of the predicted defects. The team used simulations to work out how the electron beam would pass through the material, then examined samples cut in different directions to reveal the buried interface in cross section. The report explains that the beam passed through material on both sides of the interface, complicating the image.
William Sandholt, then a Ph.D. student at DTU, first predicted the grid using a machine-learning model of how atoms would move as the sheets bonded and twisted. Electron microscopy by collaborators at the University of Antwerp and the SuperSTEM Laboratory in the United Kingdom helped identify and characterize the structure. The report says that further microscopy, interpreted alongside simulations, supported the connection between the atomic rearrangement and the electrical patterns.
The work’s significance lies in combining control over how the sheets are assembled with predictions that helped guide how researchers looked for the hidden structure. The team now wants to test whether the approach can produce useful electrical behavior in strontium titanate or other oxides. Whether that promise extends to devices—and whether the grid can conduct—will require further research.