Structure Refinement of Al-Si Interface at the Atomic Level for Identification of Two-level Systems in Superconducting Qubits

Superconducting qubits are promising platforms for quantum computing. However, they are susceptible to decoherence caused by two-level system (TLS) defects. The interface between the superconducting metal and the Si substrate is one of the possible origin sites for TLS defects. Possible candidate sites include interface steps, metal film grain boundaries, mixing across the interface, or strain fields associated with any of these microstructures. We have investigated Al-Si interfaces in Josephson junctions using aberration-corrected atomic-resolution scanning transmission electron microscopy (STEM) imaging for precise structural characterization. The STEM data were used as targets for the refinement of structural models using FANTASTX (Fully Automated Nanoscale To Atomistic Structure from Theory and eXperiments) software which also minimizes the formation energy calculated using density functional theory. Candidate TLS structures derived from FANTASTX models will be discussed.

This work was supported by DOE DE-SC0020313.