Quantum Computer Flaws Pinpointed Using Novel Energy Decay Spectroscopy Technique

Why It Matters

The method enables rapid, tunable‑free detection of decoherence sources, accelerating materials optimisation and improving coherence for scalable quantum computers.

Key Takeaways

• Fixed-frequency spectroscopy detects TLS without qubit tuning
• Dominant TLSs identified even >100 MHz detuned
• Anti‑correlation between T1e and T1f reveals single TLS
• Two‑TLS model required for device B dynamics
• Method accelerates materials diagnostics for scalable quantum processors

Pulse Analysis

Superconducting transmon qubits are the workhorse of most near‑term quantum processors, yet their performance is routinely limited by microscopic two‑level systems (TLS) that act as loss channels. Traditional TLS spectroscopy relies on sweeping the qubit frequency or the defect into resonance, a process that adds hardware complexity and is ill‑suited for fixed‑frequency architectures. The recent work from the SQMS Center and Fermilab introduces an energy‑decay spectroscopy technique that sidesteps frequency tuning entirely, leveraging correlated decay from the first and second excited states to expose the underlying defect landscape.

The authors repeatedly prepare the qubit in its second excited state and simultaneously monitor the relaxation rates of both the |1⟩→|0⟩ and |2⟩→|1⟩ transitions. By analysing the statistical correlation between the two decay channels, they can pinpoint dominant TLSs and reconstruct their frequency drift over multi‑day intervals. Experiments on two devices demonstrate that a single TLS can produce a strong anti‑correlation between T1e and T1f, while a second device requires a two‑TLS model. Remarkably, TLSs detuned by more than 100 MHz—up to 200 MHz in the study—still exert measurable influence on qubit relaxation, overturning the assumption that only near‑resonant defects matter.

This capability has immediate ramifications for quantum‑hardware manufacturers. Fixed‑frequency spectroscopy can be integrated into routine device testing, accelerating the identification of harmful defects without redesigning the chip or adding tunable elements. The reconstructed TLS drift data also feed back into materials‑science programs, guiding the selection of substrates, dielectrics, and fabrication processes that suppress the most disruptive defects. As quantum processors scale to thousands of qubits, the ability to diagnose and mitigate TLS‑induced decoherence at the wafer level will be a decisive factor in achieving reliable, error‑corrected computation.