Electrically tunable, two-photon interface from remote silicon-vacancy centers in industrial silicon carbide

Scalability is widely recognized as the central challenge for quantum computing, but it is equally critical for quantum communication. A quantum network requires a large number of quantum memory nodes. Each node must emit optically indistinguishable photons, host long-lived spin qubits for quantum information processing, and provide an efficient spin–photon interface. Few quantum systems can meet this combination of requirements, and fewer still can meet it reproducibly with mass-deployable fabrication. The silicon vacancy (VSi) center in 4H silicon carbide already combines favorable spin coherence and spin–optical properties with a mature semiconductor platform.
Here we address the remaining key requirement for scalable quantum memory nodes: stable, high-visibility two-photon interference between independently fabricated emitters. We integrate VSi centers into lateral p-i-n diodes made from commercially available 6-inch 4H-SiC wafers. The reverse bias provides both DC Stark tuning and charge-depletion-induced linewidth narrowing. Without pre-selection, all 19 characterized centers across two samples can be electrically tuned to a common target frequency. Their optical linewidths are below 60 MHz. Using two such centers in separate cryostats two meters apart, we observe two-photon interference with a raw Hong–Ou–Mandel visibility of 82.0(6)%, on par with the state of the art for remote solid-state emitters. This performance is sustained throughout a fully automated 628-hour (26-day) measurement campaign, with bias readjustments needed only every 8.4 h and 19.4 h on average. A quantitative model attributes the residual visibility loss to session-to-session spectral diffusion and non-transform-limited linewidths. These results establish VSi centers in industrial SiC as a scalable, spectrally stable building block for quantum memory nodes in distributed quantum networks.