Improving the quality of constituent materials may be the most pressing technological advancement necessary to achieve fault-tolerant superconducting quantum processors. It is currently believed that the robustness of the qubit state (relaxation time, or T₁) is limited by two-level system defects (TLS) hosted in the material. TLS can affect T₁ through both classical processes (e.g., dielectric loss) and quantum processes. 

Despite the postulated importance of TLS, their precise chemical and physical form has remained elusive. Where do these TLS reside? Perhaps in the bulk of the superconductor, or maybe its surface oxides? Or within the Josephson tunneling junction, which is formed with aluminum oxide? Do the TLS result from the intrinsic chemistry of our materials, or from residual effects of fabrication? The answers to these questions will provide key insights that enable the development of new fabrication processes. However, understanding the microscopic nature of TLS requires sophisticated tools and a breadth of expertise that are not typically found in a superconducting qubit lab.

In the Houck lab at Princeton University, we use nanofabrication to create superconducting qubits and characterize them in a dilution refrigerator (at 10 mK!) with microwave electronics. These processes have been optimized over two decades of superconducting qubit research, leading to improvements in T₁ by several orders of magnitude. However, our field has largely used the same superconductors-niobium and aluminum-until only a couple years ago. Our lab replaced niobium with tantalum, leading to the highest published T₁’s for 2D transmon qubits at the time [1]. The idea came out of a collaboration between our group, Nathalie de Leon in Princeton’s Electrical Engineering department, and Robert Cava in Princeton’s Chemistry department, showing the promise of an under-utilized collaboration between qubit researchers and materials experts.

In our study [2], we examined the properties of niobium films in a collaboration with Brookhaven National Lab (BNL), where scientists have access to state-of-the-art facilities like the National Synchrotron Light Source (NSLS) II. We used a wide array of spectroscopy and microscopy techniques to search for clues about the microscopic sources of qubit loss. Specifically, we deposited niobium with a varied set of sputtering parameters, yielding a wide range of qubit performance and materials properties. By comparing the deposition methods, we have pinpointed grain boundaries and suboxides near the surface as plausible TLS hosts. 

For me, a highlight of this study was visiting BNL and participating in the resonant inelastic x-ray scattering (RIXS) measurements, led by Dr. Ignace Jarrige. I was impressed by the sheer size of NSLS II and the SIX (soft inelastic x-ray scattering) beamline, and how a multitude of technologies come together to create record energy resolution [3]. I feel fortunate that, through the mentorship of Ignace and his colleagues, I had the opportunity to learn about a new field and a set of measurement techniques which deepened my understanding of my own field.

Post-measurement discussions with Ignace and Dr. Mark Hybertsen, a nanomaterials theorist at BNL, were as illuminating as the measurements themselves. As we learned to speak each other’s languages, we gained insight into the physical mechanisms that connect the materials properties observed at BNL with the T₁ values I had measured. Our proposed connections suggest several promising avenues for future exploration. In particular, it would be interesting to see whether a TLS-based mechanism can explain why tantalum qubits perform better than niobium qubits. We hope that our community sees our paper for much more than its scientific conclusions-but as a blueprint for how collaborations like ours have the power to propel the field of superconducting qubits forward.  Indeed, this partnership between BNL and academic scientists to understand qubit coherence will continue as part of the Co-Design Center for Quantum Advantage (C2QA), a new DOE QIS Research Center led by BNL.

[1] Place, A.P.M., Rodgers, L.V.H., et al. New material platform for superconducting transmon qubits with coherence times exceeding 0.3 milliseconds. Nat Commun 12, 1779 (2021).

[2] Premkumar, A., Weiland, C., Hwang, S. et al. Microscopic relaxation channels in materials for superconducting qubits. Commun Mater 2, 72 (2021).

[3] Jarrige, I., Bisogni, V., Zhu, Y., Leonhardt, W., & Dvorak, J.. Paving the Way to Ultra-High-Resolution Resonant Inelastic X-ray Scattering with the SIX Beamline at NSLS-II. Synchrotron Radiation News, 31, 2, 7-13 (2018).