Experimental evidence for a surface distribution of two-level systems in superconducting lithographed microwave resonators (0802.4457v2)

Abstract: We present measurements of the temperature-dependent frequency shift of five niobium superconducting coplanar waveguide microresonators with center strip widths ranging from 3 $\mu$m to 50 $\mu$m, taken at temperatures in the range 100-800 mK, far below the 9.2 K transition temperature of niobium. These data agree well with the two-level system (TLS) theory. Fits to this theory provide information on the number of TLS that interact with each resonator geometry. The geometrical scaling indicates a surface distribution of TLS, and the data are consistent with a TLS surface layer thickness of order a few nm, as might be expected for a native oxide layer.

• The paper demonstrates that frequency shifts in niobium CPW resonators are primarily due to surface-localized TLS, consistent with theoretical predictions.
• It employs temperature-dependent measurements across five resonators with varying center strip widths to quantify TLS contributions.
• The findings suggest that mitigating surface TLS via passivation or material modifications could reduce noise in superconducting devices.

Experimental Evidence for a Surface Distribution of Two-Level Systems in Superconducting Lithographed Microwave Resonators

This paper explores the contributions of two-level systems (TLS) to the frequency shifts observed in niobium superconducting coplanar waveguide (CPW) microresonators. Specifically, the research seeks to discern the geometrical distribution of TLS, which are posited to be responsible for excess frequency noise in superconducting devices. The authors provide experimental evidence supporting the hypothesis that TLS are primarily distributed within a surface layer rather than the bulk substrate of the resonators.

Overview

The authors conducted temperature-dependent measurements of frequency shifts across five niobium CPW resonators with varying geometries, at temperatures significantly lower than the critical temperature of niobium. The resonator geometries incorporated center strip widths of 3 to 50 micrometers. The experimental data closely adhered to TLS theory, suggesting a surface-based scattering of TLS, likely within a native oxide layer only a few nanometers thick. These results shed light on the source of the noise, which universally affects resonator behavior regardless of materials used or device design.

Methodology

The approach capitalized on systematic variation of the CPW resonator geometry, enabling an exploration of filling factors $F \delta$ corresponding to dielectric loss tangents over the studied temperature range. The geometric scaling pattern discovered indicates that TLS are not evenly distributed throughout the material but are located at a surface level. This finding is notably consistent with the thin layer thickness of native oxides.

Using detailed measurements and fits to TLS theory, the paper achieved valuable quantifications of the interaction between electric fields and TLS. The theoretical basis relied heavily on microwave resonance interactions, extending the interpretive model of TLS to encompass surface distributions validated through variations in center strip width geometry.

Key Findings

• The frequency shifts were significantly affected by resonator geometry, consistent with a distribution of TLS on the surface.
• Corrected filling factor values evidenced strong correlation with theoretical predictions for surface-distributed TLS.
• TLS effects exhibited scaling patterns indicative of their localization at interfaces, such as metal surfaces or exposed substrates.
• Absolute measured values suggest a thin layer of oxidized or adsorptive material serves as the TLS host.

The work brings critical understanding to the field of superconducting quantum systems by confirming the presence of TLS at interfaces, contributing to excess frequency noise. This is crucial for ongoing efforts in advancing the sensitivity and efficiency of quantum computing and photon detection technologies.

Implications and Future Directions

The implications of this research are profound for the future design and material selection in superconducting devices. By identifying the surface as the host region for TLS, efforts to mitigate their effects can now be methodically targeted towards surface chemistry modifications, potentially through the use of alternative materials or passivation methods. The success of these methodologies could significantly reduce noise and achieve increased coherence times in qubits, improving the fidelity of quantum operations.

Future research could explore varied material compositions, specifically focusing on non-oxidizing substrates and coatings compatible with niobium-based superconductors. Additionally, experimentation with geometry variations beyond strip width could further isolate and clarify the precise contributions of different surface interfaces to the observed phenomena. The environment's role, including adsorptive effects driven by ambient conditions, is another promising area of inquiry.

This investigation bridges an important gap in our understanding of TLS behavior in superconducting microresonators, establishing a strong foundation for enhancing the robustness and capabilities of these devices in advanced technological applications.