Atomic-Level Exploration of Local Structural Heterogeneity in Liquid Ga-In Alloys Using a Machine Learning Potential

Liquid gallium-indium (Ga-In) alloys have emerged as an intriguing class of materials that combine the advantages of both metals and liquids. Their unique properties enable a range of potential applications in chemistry, physics, materials, and engineering. Thus, elucidating the local structural properties of liquid Ga-In alloys at the nanoscale is essential for understanding the processes occurring within them and for advancing new applications. Here, we performed an atomic-level investigation of local coordination behavior and structural ordering around Ga and In atoms in five liquid Ga-In alloys (i.e., Ga_79.3In_20.7, Ga_83.2In_16.8, Ga_85.8In_14.2, Ga_91.8In_8.2, and Ga_96.9In_3.1) using machine learning force field molecular dynamics simulations. Our findings reveal that the Ga_79.3In_20.7, Ga_83.2In_16.8, and Ga_85.8In_14.2 alloys exhibit a continuous and predominant Ga-In phase, which is characterized by a spatial distribution of clustered In atoms. In stark contrast, the Ga_91.8In_8.2 and Ga_96.9In_3.1 alloys demonstrate the coexistence of pure Ga and Ga-In alloy phases, with In atoms forming dimers in the former and being atomically dispersed in the latter. More importantly, it is shown that there is a prominent difference in the local structural ordering surrounding Ga and In atoms in the first three Ga-In alloys, following a specific order of In-In > Ga-In > Ga-Ga, which is ascribed to the differences in interaction strengths. Overall, this study presents the first theoretical evidence of local structural heterogeneity in liquid Ga-In alloys and offers new insights into the nanoscale structural details in liquid Ga-based alloys.