Complex short-range order in Si-Ge-Sn alloys revealed by machine-learning potential and modeling-assisted experimental characterization

Short-range order (SRO) has been recently demonstrated to play a crucial role in modulating a wide range of physical properties in medium-entropy alloys (MEAs) and high-entropy alloys. In this regard, our recent theoretical studies demonstrate that two distinct types of SROs, which carry different energies, distinct degrees of local ordering, and dissimilar electronic structures, can coexist across a wide range of compositions in Si-Ge-Sn MEAs, thus opening the opportunity for bandgap engineering through tuning SRO^1,2. However, experimental characterization of SRO has been demonstrated highly non-trivial. With our recently developed highly efficient neuroevolution machine-learning potential with a DFT-level accuracy³, we can extend the spatial scale of DFT modeling to directly match that of advanced characterization techniques including atom-probe tomography (APT) and 4D scanning transmission electron microscopy (4D-STEM). We further develop a novel statistical approach to retrieve Warren-Cowley SRO parameters directly from raw APT data. The developments enable cross-validation among theoretical predictions and different experimental characterizations techniques including APT, 4D-STEM and EXAFS. The developed approach is generally applicable for quantitatively characterizing SRO in a wide range of alloys.

1. Jin, X., Chen, S. & Li, T. Coexistence of two types of short-range order in Si–Ge–Sn medium-entropy alloys. Commun Mater 3, 66 (2022).

2. Jin, X., Chen, S. & Li, T. Enabling Type I Lattice-Matched Heterostructures in SiGeSn Alloys Through Engineering Composition and Short-Range Order: A First-Principles Perspective. IEEE J. Sel. Top. Quantum Electron. PP, 1–11 (2024).

3. Chen, S., Jin, X., Zhao, W. & Li, T. Intricate short-range order in GeSn alloys revealed by atomistic simulations with highly accurate and efficient machine-learning potentials. Phys. Rev. Mater. 8, 043805 (2024).