Superconducting qubits free of fabrication residues

Nanofabrication techniques for superconducting qubits often suffer from residual contamination in the dielectric interfaces of these devices. This occurs because these techniques rely on resist-based electron-beam or optical lithography masks, which are placed and processed directly on the surface of the same wafer where the thin-film metallic qubit structures are deposited.

Tsioutsios et al. developed a technique for making superconducting qubits that minimizes residual contamination and is compatible with high temperature processes. Their technique uses free-standing silicon shadow masks fabricated from silicon-on-insulator wafers, which are separate from the wafers that the qubit metallic layers are deposited on.

“Several experimental results indicate that one of the main sources of energy loss in superconducting circuits comes from their dielectric interfaces, rather than the bulk dielectrics,” said author Ioannis Tsioutsios. “Therefore, eliminating those residues along with optimal preparation of the device-wafer surface can lead to better understanding of surface dielectric losses and potentially to improved superconducting qubit lifetimes.”

Their method involves a type of stencil lithography that naturally separates mask fabrication from device-wafer preparation. This not only minimizes cross-contamination between the mask and the device-wafer, but also makes it possible for researchers to optimize the surface of the device-wafer, without the limitations imposed by standard fabrication techniques.

“The next step in our research is to better understand and control surface dielectric losses in superconducting qubits,” said Tsioutsios. “We will rely on the specific advantages of our technique to perform controlled studies on the energy relaxation properties of aluminum superconducting qubits as a function of the surface preparation of their substrate.”

Source: “Free-standing silicon shadow masks for transmon qubit fabrication,” by I. Tsioutsios, K. Serniak, S. Diamond, V. V. Sivak, Z. Wang, S. Shankar, L. Frunzio, R. J. Schoelkopf, and M. H. Devoret, AIP Advances (2020). The article can be accessed at https://doi.org/10.1063/1.5138953 .