Deep learning potential-assisted surface engineering for HfO2/SiO2 interface and enhanced laser damage resistance

YN Qi and XG Guo and X Gao and P Zhou, APPLIED SURFACE SCIENCE, 706, 163540 (2025).

DOI: 10.1016/j.apsusc.2025.163540

Surface and interface defects are key limitations affecting the performance of optical films. In this study, the interface of HfO2/SiO2 films was enhanced by utilizing a substrate with high absorption capacity for deposited HfO2 molecules. To systematically map the relationship between absorption structures and absorption energy, a DeepMD potential for the HfO2-molecule/SiO2-surface system was developed. Based on this mapping, the surface structures was modified through plasma treatment at controlled temperature to regulate the treatment depth and further remove carbon-and fluorine-related contaminants. The prepared films exhibited an increase in HfO2 density at the interface, and the laser-induced damage threshold (LIDT) was improved after surface modification. These findings provide an atomistic understanding of the complex reaction dynamics controlled by surface defects on amorphous surfaces, offering insights into the potential of surface engineering to enhance optical performance in space applications or high-power laser systems.

Return to Publications page