Niobium Bilayers: XPS Demonstrates 17 Capping Layers Resist Surface Oxidation

Why It Matters

By curbing surface oxidation, the findings enable lower dielectric loss and longer qubit coherence, accelerating quantum‑hardware performance.

Key Takeaways

• XPS rapidly screens 17 niobium capping materials for oxidation.
• Metallic zirconium and nitrides show strongest oxygen barrier.
• 5 nm noble metal caps fail to prevent surface oxide growth.
• Oxide thickness directly correlates with resonator loss tangent.
• Annealing can increase interfacial oxidation, reducing qubit coherence.

Pulse Analysis

Superconducting microwave resonators and qubits rely on niobium’s low‑loss properties, yet native oxides on the metal surface introduce two‑level‑system defects that degrade quality factors. Dielectric loss from these oxides limits coherence times, prompting researchers to explore protective capping layers. X‑ray photoelectron spectroscopy (XPS) offers depth‑sensitive, non‑destructive analysis, allowing rapid assessment of oxygen diffusion across thin films. By applying XPS to 17 candidate caps, the Cornell team established a practical workflow for material screening before costly device fabrication.

The study revealed that metal nitrides such as NbN, ZrN, HfN, TiN, and TaN, along with pure zirconium, provide the most effective oxygen barriers. These 5 nm caps limited niobium oxidation, resulting in thinner surface oxides and lower loss tangents when resonators operated between 4 GHz and 8 GHz at 500 mK. In contrast, noble‑metal caps failed to suppress oxide growth, and annealing unexpectedly promoted interfacial oxidation, underscoring the sensitivity of the process to thermal budgets. The direct correlation between XPS‑measured oxidation states and measured medium‑power loss validates XPS as a predictive tool for device performance.

For quantum‑hardware manufacturers, the ability to down‑select capping materials early shortens development cycles and improves qubit coherence without redesigning the entire stack. The open dataset encourages community verification and paves the way for alloy engineering, alternative cleaning chemistries, and controlled annealing to further mitigate oxidation. As superconducting processors scale, integrating robust, low‑loss capping layers will be essential for achieving the high‑fidelity operations demanded by error‑corrected quantum computing.