Can blocking sound waves make qubit-damaging defects live longer?
Atomic-scale defects inside superconducting qubit junctions lived about a hundred times longer when their frequency fell inside an engineered band where sound waves (phonons) cannot travel.
Source
Phonon engineering of atomic-scale defects in superconducting quantum circuits
Study at a glance
- Design
- Other — Cryogenic experiment: transmon qubits with junctions on suspended silicon platforms surrounded by a phononic bandgap shield, used as sensors to swap into and time the decay of individual defects.
- N
- N=56 · 56 two-level-system defects with unique frequencies, measured across seven transmon devices on two chips (including thermal cycling to redistribute frequencies).
- Population
- Two-level-system defects in the aluminium oxide barriers of Josephson junctions on silicon-on-insulator chips
- Outcome
- Defect energy relaxation time T1 versus frequency relative to the acoustic bandgap; defect coherence T2*; temperature dependence of T1
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The team built frequency-tunable transmon qubits whose Josephson junctions sit on suspended silicon platforms tethered by a patterned 'cross-shield' structure with an acoustic bandgap. In a dilution refrigerator, they used each qubit to find strongly coupled two-level-system (TLS) defects in the junctions, swapped excitations into them, and measured how long they stayed excited. They repeated this across devices and thermal cycles, then studied one long-lived defect's coherence and temperature dependence.
What they found
Across 56 defects, lifetimes split into two families, and the long-lived family sat mostly inside a frequency band that matched the simulated acoustic bandgap. The median lifetime rose from 4.4 μs outside the band to 505 μs inside, and one defect reached about 1100 μs. However, the defect's dephasing time stayed near 0.91 μs, limited by low-frequency noise, and its quality factor dropped sharply above about 75 mK.
The limits
What it doesn't show
Longer energy lifetime did not bring longer coherence, so the defects are not yet useful memories without techniques like dynamical decoupling, which were not tried. Several outliers remain, attributed to fabrication variation or dipole orientation but not directly tested. The low-temperature plateau in defect lifetime is only tentatively explained by quasiparticles, and the qubits themselves had short lifetimes of about 3 μs, so the paper does not show improved qubit performance.
Key terms
- Two-level system (TLS)
- An atomic-scale defect in an amorphous material that tunnels between two nearly equal configurations and can absorb energy from qubits.
- Transmon qubit
- A superconducting circuit qubit built from Josephson junctions and a capacitor, with microwave-frequency transitions.
- Acoustic (phononic) bandgap
- A range of frequencies in which a patterned structure supports no propagating sound waves.
- T1 relaxation time
- The characteristic time for an excited quantum state to decay to its ground state.
- SWAP spectroscopy
- Transferring an excitation between the qubit and a defect by bringing them into resonance for a set time.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
Why are phonons the dominant bath for TLS at microwave frequencies?
Common questions
Why target phonons rather than photons for the defects' decay?
Sound is about five orders of magnitude slower than light, so there are far more phonon states at microwave frequencies, making phonons the dominant bath.
How do they know the long lifetimes come from the bandgap?
The long-lived family of defects clusters in a frequency band that matches the simulated bandgap of the cross-shield structure.
Did the defects become better qubits?
Their energy lifetime grew greatly, but their dephasing time stayed around a microsecond due to low-frequency noise.
More on Qubits and quantum devices