Deciphering chemical order/disorder and material properties at the single-atom level

Perfect crystals are rare in nature. Real materials are usually composed of crystal defects and chemical order/disorder such as grain boundaries, dislocations, interfaces, surface reconstructions and point defects that disrupt the periodicity of the atomic arrangement and determine their properties and functionality1-3. Although recent years have witnessed rapid development of quantitative material characterization methods1,4-18, correlating 3D atomic arrangements of chemical order/disorder and crystal defects with material properties remains a major challenge. On a parallel front, quantum mechanics calculations such as density functional theory (DFT) have made significant progress from modelling ideal bulk systems to “real” materials with dopants, dislocations, grain boundaries and interfaces19,20. Presently, these calculations rely heavily on average atomic models extracted from crystallography. To improve the predictive power of first-principle calculations, there is a pressing need to use atomic coordinates of real systems beyond average crystallographic measurements. Here, we determined the 3D coordinates of 6,569 iron and 16,627 platinum atoms in an iron-platinum nanoparticle to correlate 3D atomic arrangements and chemical order/disorder with material properties at the single-atom level. We identified rich structural variety and chemical order/disorder including 3D atomic composition, grain boundaries, anti-phase boundaries, anti-site point defects and swap defects. We show for the first time that experimentally measured 3D atomic coordinates and chemical species with 22 pm precision can be used as direct input for DFT calculations of material properties such as spin and orbital magnetic moments and local magnetocrystalline anisotropy. This work merges the forefront of 3D atomic structure determination of crystal defects and chemical order/disorder with DFT calculations, which is expected to transform our understanding of structure- property relationships at the most fundamental level.