Figure

Description

Silicon-carbide defect qubits use optically addressable electron spins associated with point defects in a technologically mature wide-band-gap semiconductor. Two major branches must not be conflated: the neutral divacancy has an ground state analogous in control language to diamond NV centers, while the negatively charged silicon vacancy has an ground state and different level structure.

SiC combines optical initialization/readout, microwave spin control, nearby nuclear-spin registers, and wafer-scale growth and nanofabrication. Its near-infrared optical transitions are attractive for spin-photon interfaces, though spectral stability and photon collection remain less mature than the best diamond nanophotonics.

Hamiltonian

For an axial neutral divacancy, a useful ground-state spin Hamiltonian is

with , zero-field splitting , transverse strain/electric-field parameter , and hyperfine tensor . Both and the sign and magnitude of the optical spin contrast depend on polytype, lattice site, and defect configuration, so values such as GHz apply to a specified center (for example PL6), not to every SiC divacancy. Silicon-vacancy centers require an crystal-field Hamiltonian; they are a related defect branch, not the same qubit.

Motivation

  • Retain optically addressable defect-spin physics in a semiconductor with established wafer processing.
  • Integrate spin memories with nanophotonic cavities and electrical devices on the same material platform.
  • Exploit multiple defect species and nuclear-spin registers while maintaining a clear physical taxonomy.

Experimental Status

Room-temperature coherent control of SiC defect spins was demonstrated in 2011. Christle et al. isolated and coherently controlled individual neutral divacancies and measured ensemble Hahn-echo coherence exceeding 1 ms. Nagy et al. demonstrated high-fidelity spin and optical control of single silicon-vacancy centers in 4H-SiC.

The platform has since advanced from isolated-spin control to network-node primitives. Bourassa et al. entangled a divacancy electron spin with a single Si nuclear spin while reporting 99.984(1)% electron-spin control fidelity and ms. Fang et al. generated V1 silicon-vacancy spin–photon entanglement with fidelity %. In 2024, Hu et al. prepared an ambient-condition PL6 electron–C Bell state with fidelity 0.89 and retained 0.88 after integration into a silicon-carbide-on-insulator waveguide. Nishikawa et al. then demonstrated coherent room-temperature photoelectrical readout of an individual V2 silicon-vacancy spin with 1.7–2 times the signal-to-noise ratio of their optical readout. Morioka et al. broadened the electrical-readout picture in 2026 by demonstrating coherent photoelectrical detection and spectroscopy of PL3 and PL5–PL7 ensembles.

Key Metrics

MetricValueNotesFidelity reference
Hahn-echo >1 msNeutral divacancy ensemble in high-purity 4H-SiCChristle et al. 2015
Electron-spin control99.984(1)%Isotopically engineered divacancy; randomized benchmarkingBourassa et al. 2020
Dynamically decoupled >14.5 msIsotopically engineered divacancy; Hahn echo msBourassa et al. 2020
Spin–photon entanglement fidelity%V1 silicon vacancy, time-bin photonic qubitFang et al. 2024
Electron–nuclear Bell-state fidelity0.89 (0.88 after waveguide integration)PL6 divacancy–C register at ambient conditionsHu et al. 2024
Operating regimeRoom temperature demonstratedCoherent defect-spin controlKoehl et al. 2011
Single-spin photoelectrical readout1.7–2× optical-readout SNRIndividual V2 silicon vacancy at room temperatureNishikawa et al. 2025
Photoelectrical readoutRoom-temperature coherent PDMRPL3 and PL5–PL7 defectsMorioka et al. 2026
Defect branches and Divacancy and silicon-vacancy encodingsNagy et al. 2019

Scaling Considerations

  • Different polytypes, lattice sites, and charge states create a rich but fragmented device landscape.
  • Transform-limited optical lines and stable charge-state control are required for remote entanglement.
  • Isotopic purification and nuclear-spin engineering can extend coherence but add materials complexity.
  • Foundry compatibility is promising, but reproducible cavity-coupled single-defect yield remains a bottleneck.
  • Photoelectrical magnetic-resonance readout offers a path around inefficient near-infrared photon collection, but defect-dependent ionization and electrical contrast still require device-level optimization.

References

Linked Papers

Evergreen context