About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
Chemistry and Physics of Interfaces
|
| Presentation Title |
Mechanics of the Bonded LiNbO3/SiO2 Interphase from a Machine-Learned Interatomic Potential |
| Author(s) |
Nuwan Dewapriya, John Gillespie Jr. |
| On-Site Speaker (Planned) |
Nuwan Dewapriya |
| Abstract Scope |
The bonded LiNbO3/SiO2 interphase governs the reliability of lithium-niobate-on-insulator (LNOI) device platforms, yet its reactive, mixed ionic-covalent character lies beyond classical interatomic potentials. We train an on-the-fly machine-learned interatomic potential from density functional theory configurations spanning non-bonded, transition, and bonded regimes, then drive 18,243-atom molecular dynamics of the interphase under Mode I and Mode II loading at 300, 450, and 600 K. The interphase resolves compositionally and mechanically, with operational thickness near 2.4 nm matching experiment. At 300 K the normal traction peaks at 8.5 GPa, while the shear response reveals a Coulomb friction shelf and a non-cohesive normal residual isolated from the cohesive law. Between 300 and 600 K the interphase loses 20% of its peak traction and 31% of its work of separation, softening several times faster than bulk LiNbO3. We also characterize mixed-mode loading at 300 K and provide cohesive-zone parameters for device-scale modeling. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, Mechanical Properties, Thin Films and Interfaces |