From Atomic Defects to Integrated Photonics: A Perspective on Solid-State Quantum Light Sources

Single-photon emitters (SPEs) constitute a foundational resource for quantum technologies, including secure communication, photonic quantum computing, and emerging quantum network architectures. A wide range of quantum materials, from atom-like point defects in bulk crystals to excitonic states in low-dimensional semiconductors, now provide bright, coherent, and scalable sources of non-classical light. Meanwhile, advances in photonic integration have enabled efficient routing, filtering, and on-chip manipulation of these emitters. From this perspective, we survey and discuss the technological landscape in which solid-state emitters interface with quantum sensing, quantum communication, quantum computation, and emerging photonic AI platforms. Further, we discuss the materials landscape underpinning modern single-photon sources from the zero-dimensional, one-dimensional, two-dimensional and three-dimensional materials. Lastly, we highlight key integration pathways for these single-photon emitters into scalable quantum photonic systems.

💡 Research Summary

**
This perspective paper surveys the rapidly evolving landscape of solid‑state single‑photon emitters (SPEs) and their integration into scalable quantum photonic platforms. Beginning with an overview of quantum information science, the authors delineate three principal application domains—quantum sensing, quantum communication, and photonic quantum computing—and identify the photon‑level performance metrics required for each: high purity (g²(0) ≪ 0.5), high brightness, near‑transform‑limited linewidth, and, critically, photon indistinguishability for interference‑based protocols.

The review then examines the material families that host SPEs, organized by dimensionality. Bulk (3‑D) systems such as diamond host nitrogen‑vacancy (NV) and silicon‑vacancy (SiV) centers; NV offers room‑temperature spin coherence but suffers from relatively broad optical lines, whereas SiV/GeV provide narrower lines at cryogenic temperatures. Silicon carbide (SiC) is highlighted for its CMOS compatibility and favorable optical properties. Zero‑dimensional quantum dots (InAs/GaAs, InP, etc.) deliver high repetition rates and narrow linewidths but require low‑temperature operation; they can be Purcell‑enhanced via photonic crystal cavities or micro‑cavities. Two‑dimensional transition‑metal dichalcogenides (TMDCs) such as WSe₂ enable strain‑tunable emission across telecom windows and can operate near room temperature, though environmental stability remains a challenge. One‑dimensional nanowires and carbon nanotubes are mentioned briefly as emerging platforms.

Integration strategies form the core of the paper. The authors advocate hybrid platforms that combine deterministic SPEs with ultra‑low‑loss Si₃N₄ waveguides (propagation loss ≈10⁻² dB m⁻¹). A representative device places InAs quantum dots in a GaAs nanowaveguide onto a buried Si₃N₄ bus, followed by an adiabatic mode transformer and a 50:50 multimode‑interference coupler. This architecture achieves g²(0) ≈ 0.04, on‑chip insertion loss ≈1 dB m⁻¹, and maintains photon indistinguishability under resonant excitation, thereby meeting the stringent requirements of measurement‑based photonic quantum computing, Gaussian boson sampling, and quantum repeaters.

Application‑specific sections illustrate how SPEs are already being leveraged. In quantum sensing, spin‑based magnetometers (SERF, NV) achieve sub‑femtotesla sensitivity; when networked, they enable dark‑photon searches and biomedical magnetometry. In quantum communication, strain‑engineered WSe₂ monolayers serve as on‑chip BB84 transmitters, delivering g²(0) ≪ 0.1, quantum bit error rates below 1 % and supporting pulse rates up to 10 MHz. Purcell‑enhanced extraction and integration with low‑dark‑count superconducting nanowire detectors are projected to push secret‑key rates toward the rate‑loss frontier required for global quantum networks.

The photonic quantum computing discussion emphasizes the bottleneck of loss in increasingly complex linear‑optical circuits. Si₃N₄ waveguides provide a CMOS‑compatible, broadband, low‑loss backbone, while deterministic SPEs supply a scalable source of high‑quality photons. The authors argue that such hybrid systems enable long on‑chip delay lines, time‑multiplexed architectures, and integrated detectors, all essential for fault‑tolerant, measurement‑based computation.

Finally, the paper introduces quantum‑AI concepts based on boson‑sampling reservoirs (QORC). Multi‑photon interference in large interferometers creates high‑dimensional, nonlinear feature maps that, when combined with classical linear classifiers, outperform purely classical random‑Fourier‑feature approaches. Experiments show that true single‑photon inputs yield superior classification accuracy compared with coherent‑state inputs of equal photon number, underscoring the value of quantum interference as a computational resource.

In conclusion, the authors identify four critical pathways toward practical quantum photonic systems: (1) development of room‑temperature, high‑coherence SPEs; (2) efficient coupling of these emitters to ultra‑low‑loss waveguides and resonators; (3) adoption of heterogeneous, wafer‑scale fabrication compatible with existing semiconductor foundries; and (4) co‑integration of fast electronic control and low‑noise detection. By addressing these challenges, solid‑state SPEs can transition from laboratory curiosities to the backbone of future quantum sensing networks, secure communication links, large‑scale photonic processors, and quantum‑enhanced AI accelerators.