Assessing the Sensitivity of Niobium- and Tantalum-Based Superconducting Qubits to Infrared Radiation

The use of tantalum films for superconducting qubits has recently extended qubit coherence times significantly, primarily due to reduced dielectric losses at the metal-air interface. However, the choice of base material also influences the sensitivity to quasiparticle-induced decoherence. In this study, we investigate quasiparticle tunneling rates in niobium and tantalum-based offset-charge-sensitive qubits. Using a source of thermal radiation, we characterize the sensitivity of either material to infrared radiation and explore the impact of the infrared background through the targeted use of in-line filters in the wiring and ambient infrared absorbers. We identify both radiation channels as significant contributions to decoherence for tantalum but not for niobium qubits and achieve tunneling rates of 100 Hz and 300 Hz for niobium and tantalum respectively upon installation of infrared filters. Additionally, we find a time-dependence in the observed tunneling rates on the scale of days, which we interpret as evidence of slowly cooling, thermally radiating components in the experimental setup. Our findings indicate that continued improvements in coherence times may require renewed attention to radiative backgrounds and experimental setup design, especially when introducing new material platforms.

💡 Research Summary

This paper investigates how the choice of base superconducting material—niobium (Nb) versus tantalum (Ta)—affects the sensitivity of offset‑charge‑sensitive transmon qubits to infrared (IR) radiation, which generates non‑thermal quasiparticles that can tunnel across the Al/AlOₓ/Al Josephson junctions and degrade qubit performance. The authors fabricate two sets of devices that are identical in geometry, junction fabrication, and packaging, differing only in the metal layer underlying the qubit capacitor (Nb or Ta). Both devices are operated at their first‑order flux‑insensitive point around 5 GHz, with an EJ/EC ratio of ≈20, giving a charge‑dispersion of ≈5 MHz. This design makes the qubit frequency depend on the parity of the offset charge (even vs. odd), allowing a Ramsey‑type pulse sequence to map the parity onto the computational basis (|0⟩/|1⟩) and to monitor parity switches in real time at repetition rates of 0.5–1 MHz.

The measurement protocol consists of an X90 pulse, free evolution for a time set by the parity splitting Δf, a Y90 pulse, and a single‑shot readout. By recording ≈500 k single‑shot I/Q points per trace (≈1 s) and repeating the experiment ten times, the authors obtain long binary time series indicating the parity state. The power spectral density (PSD) of these series shows a Lorentzian plateau at low frequencies, from which the quasiparticle tunneling rate Γ₀ is extracted. In the baseline configuration—where all control and readout lines already contain in‑line IR filters and the sample is enclosed in a mu‑metal/aluminium/copper shield—the average tunneling rates are 461 ± 16 Hz for Nb devices and 1542 ± 176 Hz for Ta devices, i.e., Ta exhibits roughly three times higher quasiparticle activity.

To probe the effect of external IR radiation, the authors introduce a resistive Manganin wire inside the shield and drive it with a variable current, thereby dissipating electrical power P_W up to a few microwatts. The wire heats to ≈5 K at the highest power, emitting thermal radiation estimated at ≤5 nW. Measured Γ₀ versus P_W follows a power law Γ₀ = Γ_base + α P_Wⁿ, with exponents n ≈ 1.4 for Nb and n ≈ 2.6 for Ta. This indicates that Nb’s quasiparticle generation scales roughly linearly with incident power (consistent with direct photon absorption), whereas Ta’s response is more nonlinear, suggesting a recombination‑limited regime where quasiparticle density grows faster with power.

Correlation analysis between Γ₀ and the energy‑relaxation time T₁ reveals that, even at the lowest applied powers, Ta qubits show a clear dependence of T₁ on Γ₀, implying that quasiparticle tunneling is the dominant relaxation channel for Ta. In contrast, Nb qubits display no systematic T₁‑Γ₀ correlation; their T₁ is limited by other loss mechanisms such as surface dielectric loss or material defects. This distinction underscores that the base‑metal material fundamentally influences quasiparticle dynamics: literature reports longer quasiparticle lifetimes and higher mobility in Ta than in Nb, leading to a larger flux of quasiparticles reaching the junction.

The authors then evaluate mitigation strategies. Adding additional in‑line IR filters on the readout output line (after the parametric amplifier) and surrounding the sample space with foam IR absorbers reduces the tunneling rates to ≈100 Hz for Nb and ≈300 Hz for Ta. While both materials benefit, Ta remains limited by quasiparticle‑induced relaxation, as evidenced by a residual slope in the T₁‑Γ₀ plot. This demonstrates that, for Ta‑based qubits, merely filtering external IR is insufficient; intrinsic quasiparticle diffusion from the base layer must also be addressed, e.g., via engineered normal‑metal traps or gap‑engineered regions.

A further observation concerns long‑term drift. After cooldown, the tunneling rates slowly decrease over several days, a behavior the authors attribute to slowly cooling components inside the cryostat that initially act as weak IR emitters. This temporal dependence suggests that background radiative environments evolve on timescales comparable to experimental runs, and that careful thermal anchoring and pre‑conditioning of cryogenic hardware are necessary for reproducible low‑quasiparticle operation.

In summary, the paper provides (1) a quantitative comparison of Nb and Ta qubits’ susceptibility to IR‑induced quasiparticles, (2) experimental validation that in‑line IR filtering and ambient absorbers can suppress but not fully eliminate quasiparticle tunneling in Ta devices, and (3) evidence that the base‑metal’s quasiparticle lifetime and diffusion properties dominate the residual decoherence. The findings imply that future high‑coherence superconducting qubits—especially those employing Ta for its low dielectric loss—must incorporate both aggressive radiative shielding and material‑specific quasiparticle mitigation (e.g., normal‑metal traps, gap engineering) to fully exploit the intrinsic advantages of the tantalum platform.