Two-Level-System Defects
Microscopic two-level systems in amorphous materials and interfaces absorb energy from qubits and are a leading limit on superconducting coherence.
What They Are
A two-level system, or TLS, is a microscopic defect that can sit in one of two nearly degenerate configurations, for example an atom tunneling between two positions or a dangling bond flipping state. They live in amorphous oxides, at material interfaces, and in the native oxide on superconducting films. Individually tiny, they are numerous, and because many carry an electric dipole moment, they couple to the electric fields of a qubit or resonator.
How They Hurt Qubits
A TLS whose transition frequency happens to coincide with a qubit's frequency can resonantly absorb the qubit's energy, opening a decay channel and shortening the qubit lifetime. A near-resonant TLS causes dispersive frequency shifts and dephasing. Because TLS frequencies wander over time and with temperature, a qubit's coherence can fluctuate from hour to hour as defects drift into and out of resonance, producing the characteristic instability seen in coherence measurements.
- Reside in amorphous oxides, interfaces, and film surfaces.
- Couple to qubit electric fields through their dipole moments.
- Cause both energy relaxation and time-dependent dephasing.
- Can be tuned by strain and electric field, revealing their nature.
Evidence and Diagnosis
TLS reveal themselves as sharp dips in a qubit's lifetime at particular frequencies, as avoided crossings in spectroscopy when the qubit is tuned across a defect, and as loss in resonators that depends on drive power and temperature in a signature way. Applying mechanical strain or a DC electric field shifts TLS frequencies, which is used both to study them and, in tunable qubits, to steer a qubit away from a harmful defect.
Mitigation
The main defenses are materials and geometry. Cleaner interfaces, removal or replacement of lossy native oxides, substitution of crystalline for amorphous dielectrics, and circuit designs that keep electric fields out of lossy regions all reduce the TLS density a qubit sees. Tunable qubits can also be parked away from known defects during operation. TLS remain one of the dominant coherence limits, and progress against them is largely a materials-science effort.