Purcell-enhanced emission from nitrogen-vacancy centers in nanodiamonds in a photonic integrated silicon nitride slot waveguide
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Journal of Physics Photonics |
| Published | 2026-05-07 |
| Authors | Jenny Paola Romero-Castro, Mauricio Gómez, Liudmila Trotsiuk, Anne Laure Baudrion, Karina Garay |
| DOI | 10.1088/2515-7647/ae6a55 |
Nanodiamond NV centers in a silicon nitride slot waveguide emitted at up to 3.125 ns⁻¹, producing an experimental Purcell factor of 8.928 against nanodiamonds on glass. The gap center shortened the amplitude-weighted lifetime from 2.82 ns to 0.32 ns. Nearby waveguide regions reached about 2.00 ns⁻¹ and a Purcell factor of 5.714. These results establish a large local rate increase, but the study did not yet verify single-photon emission through photon antibunching.
Slot waveguide design
Section titled “Slot waveguide design”Two parallel silicon nitride ridge waveguides form the slot. Each ridge was designed at 250 nm width and 4 µm length, with a 100 nm gap. An output waveguide lies on one side, while a 20-segment Bragg reflector occupies the other. The grating used a 240 nm period and had a simulated reflectance of 94% at 637 nm.
Finite-difference time-domain simulations represented the NV center as a Y-oriented electric dipole radiating at 637 nm. This orientation matched the dominant field of the gap and output waveguide modes. The gap mode had an effective index of 1.550, and the output TE mode had 1.786.
A 40 nm gap produced a simulated Purcell factor of 8.94. The selected 100 nm gap gave 3.22. Lateral emitter position changed the factor from about 2.80 to 3.05 through interference with reflected fields. The waveguide design differs from the photonic crystal cavity and diamond nanostructure approaches cited by the authors.
Device fabrication and emitter placement
Section titled “Device fabrication and emitter placement”The structure was fabricated on 250 nm silicon nitride on glass. Electron-beam lithography used AR-P 6200.09 resist at 20 kV, a dose of 60 µC cm⁻², and a 15 µm aperture. Reactive ion etching used CHF₃ gas at 100 W to transfer the pattern into the silicon nitride.
Measured ridge widths were 226 nm and 284 nm, and the measured gap was 96 nm. The Bragg segments were 116 nm wide with 84 nm spaces. Calculations using these measured features gave a reflector reflectance of 0.929.
Commercial fluorescent nanodiamonds containing NV centers were dispersed in deionized water at 0.1 mg ml⁻¹. Electron microscopy showed particle diameters of about 35 nm. A second lithography step defined rectangular deposition areas around the slot. Drop-casting and drying placed particles on the resist, and acetone treatment for 10 min removed excess particles. The remaining nanodiamonds were concentrated around the gap while avoiding the reflector and output waveguide.
Lifetime measurement and rate enhancement
Section titled “Lifetime measurement and rate enhancement”Time-resolved fluorescence used fluorescence lifetime imaging microscopy with time-correlated single-photon counting. The system was a Becker & Hickl DCS 120 FLIM platform. NV centers were excited with a 473 nm pulsed diode operating at 20 MHz. Excitation was linearly polarized perpendicular to the guided propagation axis and focused through a 50×, 0.85 NA objective. Decay histograms were fitted with multi-exponential models convolved with the instrument response function. The synthetic response had an effective width of 107.5 ps.
| Sample location | Fitted amplitude and lifetime components | Amplitude-weighted lifetime | Emission rate |
|---|---|---|---|
| Glass | 0.552, 1.22 ns; 0.447, 4.78 ns | 2.82 ns | 0.350 ns⁻¹ |
| Slot gap center | 0.895, 0.200 ns; 0.096, 1.03 ns; 0.007, 6.31 ns | 0.32 ns | 3.125 ns⁻¹ |
The glass decay was described by two components. The gap decay required three components. The authors attribute multi-exponential behavior to NV spin sublevels with different deactivation rates and coexisting negative and neutral NV centers. They did not separate these contributions quantitatively. The slower gap components were assigned to less efficiently coupled emitters, dipole orientation, position within each particle, and signal mixing from nearby regions.
The emission rate was calculated as the inverse amplitude-weighted lifetime. The gap center reached 3.125 ns⁻¹, against 0.350 ns⁻¹ on glass. Nearby zones reached about 2.00 ns⁻¹, giving experimental Purcell factors of 8.928 and 5.714, respectively.
Agreement with the fabricated device
Section titled “Agreement with the fabricated device”Simulations used the measured dimensions and normalized the radiated power of a Y-polarized dipole to the same dipole on glass. The maximum central-dipole factor was 5.75 at the edge of the wider lower waveguide. Nearby gap values were near 4, while positions above the waveguides fell near 1.
The experimental factor was 8.93, compared with 5.75 for the simulation, a difference of 3.17. The authors suggest that the experimental estimate includes radiative and non-radiative processes, whereas FDTD accounts only for radiative power. The difference therefore suggests that radiative processes dominate the measured interaction.
To test unknown defect placement, the authors kept the sphere fixed and moved the dipole through 60 random positions inside it. The resulting factors ranged from 4.79 to 12.04 and averaged 6.46. About 53.3% of values exceeded the centered-dipole result. Some positions produced values near 8.98, which matches the experimental factor.
Guided mode coupling
Section titled “Guided mode coupling”Modal coupling was evaluated only by FDTD. The beta factor is the emission rate into a guided mode divided by the total emission rate. At the brightest central-dipole position, the first four output modes received a total beta factor of 0.2020. The TE00 mode contributed 0.1374, and TE01 contributed 0.0613. The TM00 and TM01 modes each had values below 0.001.
The TE00 mode received the largest guided coupling because its electric field matched the Y-oriented dipole. The authors used these results to support coupling into the desired fundamental TE mode.
Limits and unresolved questions
Section titled “Limits and unresolved questions”The experimental Purcell factor was a ratio of fitted decay rates, not a direct measurement of radiative power. FLIM spatial resolution also mixed nearby emitters into some pixels. Exact vertical or lateral placement of each NV center remained unknown, and particles could contain one or more centers.
The decay analysis did not distinguish quantitatively between the effects of spin sublevels and multiple charge states. The beta calculation used one emitter position, so its mode coupling remains spatially dependent. Second-order correlation measurements are still required to establish photon antibunching and single-photon operation.
Controlled and homogeneous nanodiamond deposition remains an active goal. The authors also identify narrower gaps and more precise subwavelength fabrication as routes to higher performance.
Device engineering implications
Section titled “Device engineering implications”The 96 nm fabricated gap retained the enhanced decay and field confinement. Device design should therefore place nanodiamonds in the gap and use polarization perpendicular to guided propagation. The TE00 mode should remain the target for collection, while the Bragg reflector redirects otherwise uncoupled emission.
For material selection, the result applies to nanodiamond particles about 35 nm in diameter containing NV centers. It does not establish a result for bulk diamond windows, plates, or wafers. The next validation is photon antibunching, followed by integration into more complex photonic circuits.
View Original Abstract
Abstract We report a silicon nitride photonic structure that enhances and directs emission from nitrogen-vacancy centers in nanodiamonds (NDs), aiming at scalable integration into photonic integrated circuits. The design consists of two parallel ridge waveguides separated by a subwavelength air gap, supporting a confined optical mode that increases the local density of optical states, thereby enhancing spontaneous emission via the Purcell effect. A Bragg reflector is included to recover emissions that would otherwise radiate out of plane, directing it toward the output waveguide. Controlled localization of NDs within lithographically defined regions is achieved through lithographic masks. Fluorescence lifetime measurements reveal a ninefold emission rate enhancement, with Purcell factors up to 8.9 and simulated β -factors of up to 13.7%. This approach provides a compact platform that combines slot-waveguide emission enhancement with lithography-assisted ND localization for integration with photonic integrated circuits and compatibility with other solid-state quantum emitters.
Single Crystal Cvd Diamond Plate
View Specs →
Advanced Thermal Simulation Service Diamond Heat Spreader Performance Analysis
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →Original source
Section titled “Original source”References
Section titled “References”- 2005 - Single-photon sources [Crossref]
- 2017 - High-performance semiconductor quantum-dot single-photon sources [Crossref]
- 2020 - Chip-to-chip quantum teleportation and multi-photon entanglement in silicon [Crossref]
- 2025 - Single-photon generation and manipulation in quantum nanophotonics [Crossref]
- 2013 - The nitrogen-vacancy colour centre in diamond [Crossref]
- 1997 - Scanning confocal optical microscopy and magnetic resonance on single defect centers [Crossref]
- 2009 - Fluorescence and spin properties of defects in single digit nanodiamonds [Crossref]
- 2019 - Nanodiamonds: a powerful platform for quantum optics and sensing [Crossref]
- 2025 - Recent progress in hybrid diamond photonics for quantum information processing and sensing [Crossref]
- 2020 - Integration of diamond-based quantum emitters with nanophotonic circuits [Crossref]