Skip to content

Purcell effect of nitrogen-vacancy centers in nanodiamond coupled to propagating and localized surface plasmons revealed by photon-correlation cathodoluminescence

MetadataDetails
Publication Date2021-05-14
JournalPhysical review. B./Physical review. B
AuthorsSotatsu Yanagimoto, Naoki Yamamoto, Takumi Sannomiya, K. Akiba
InstitutionsTokyo Institute of Technology, National Institutes for Quantum Science and Technology
Citations28
AnalysisFull AI Review Included
  • Core Achievement: Experimental confirmation of the Purcell effect in nitrogen-vacancy (NV) centers within nanodiamonds (NDs) coupled to both propagating surface plasmon polaritons (SPPs) and localized surface plasmons (LSPs).
  • Methodological Breakthrough: The study successfully combined cathodoluminescence (CL) in a scanning (transmission) electron microscope (S(T)EM) with Hanbury Brown-Twiss (HBT) interferometry, enabling lifetime measurements with simultaneous nanometer spatial and nanosecond temporal resolution.
  • Purcell Enhancement: Significant shortening of the NV center lifetime was observed in samples containing silver (Ag) structures, confirming enhanced decay rates due to increased electromagnetic local density of state (EMLDOS).
  • Structure Dependence: The strongest Purcell enhancement (Fp = 1.73) was achieved when NDs were embedded in an Ag nanohole structure (Sample C), compared to NDs placed on a flat Ag film (Fp = 1.15).
  • Plasmon Coupling Confirmation: Analytical calculations and Finite Element Method (FEM) simulations elucidated that the enhancement in Sample B is primarily due to coupling with propagating SPPs, while the enhancement in Sample C is dominated by coupling to localized LSP resonance modes.
  • Statistical Validation: Wilcoxon-Mann-Whitney statistical u-tests confirmed that the observed shifts in lifetime distributions were statistically significant and attributable to the Purcell effect, not statistical fluctuation.
ParameterValueUnitContext
Quantum EmitterNV0 centersN/AContained within nanodiamonds (NDs).
ND Average Diameter100nmUsed in all samples and FEM models.
ND Graphitized Layer2.5nmThickness of amorphous layer surrounding the diamond core (used in FEM).
NV Zero-Phonon Line (ZPL)575nmEmission wavelength used for lifetime analysis.
Reference Lifetime (Sample A)16.3nsMean lifetime of NDs on 30 nm SiO2 membrane.
Lifetime (Sample B)14.2nsMean lifetime of NDs on flat 300 nm Ag film (SPP coupling).
Lifetime (Sample C)9.4nsMean lifetime of NDs embedded in 300 nm Ag film (LSP/SPP coupling).
Experimental Purcell Factor (Fp, B)1.15N/ACalculated from mean lifetimes (τA / τB).
Experimental Purcell Factor (Fp, C)1.73N/ACalculated from mean lifetimes (τA / τC).
FEM Purcell Factor (Model B, Total)1.66N/AWavelength-averaged simulation result.
FEM Purcell Factor (Model C, Total)7.46N/AWavelength-averaged simulation result (LSP mode).
Electron Accelerating Voltage80kVUsed for CL-HBT measurements.
Electron Beam Current Range17 to 60pAUsed for excitation; confirmed not to affect lifetime measurement.
Spatial Resolution (Probe Size)10 or lessnmAchieved using STEM-CL setup.
Ag Film Thickness300nmUsed for Samples B and C.
SiO2 Membrane Thickness30nmUsed for Sample A and as a protective layer in Sample C fabrication.
  1. Sample Fabrication: Three distinct sample types (A, B, C) were prepared using drop-casting and sputter-deposition techniques to control the nanodiamond environment:
    • Sample A (Reference): NDs dispersed on a 30 nm thick free-standing SiO2 membrane.
    • Sample B (Flat SPP): NDs dispersed on a 300 nm thick flat Ag film (thermally deposited on InP).
    • Sample C (Embedded LSP/SPP): NDs dispersed on a 30 nm SiO2 membrane, followed by sputter-deposition of a 300 nm Ag film to embed the NDs, creating Ag nanohole structures around the NDs.
  2. Cathodoluminescence (CL) Excitation: A Scanning (Transmission) Electron Microscope (STEM, JEM-2000FX) operating at 80 kV was used to excite the NV centers with a highly localized electron beam (probe size < 10 nm).
  3. Lifetime Measurement (CL-HBT): The second-order correlation function, g(2)(Ļ„), of the CL intensity was measured using a Hanbury Brown-Twiss (HBT) interferometer setup integrated into the STEM column.
    • The HBT system splits the CL signal into two paths, measured by single-photon counting modules (SPCMs).
    • Lifetime (Ļ„0) was extracted by fitting the photon bunching feature in the g(2)(Ļ„) curve using the expression g(2)(Ļ„) = 1 + A exp(-|Ļ„|/Ļ„0Exp).
  4. Statistical Analysis: Lifetime distributions from 53 to 58 individual NDs per sample were compiled into histograms. The Wilcoxon-Mann-Whitney u-test was applied to confirm the statistical separation of the lifetime distributions due to the Purcell effect.
  5. Electromagnetic Simulation: The Purcell factors were numerically evaluated using the Finite Element Method (FEM) via COMSOL Multiphysics, modeling the ND sphere (100 nm diameter, 2.5 nm graphite layer) in the respective environments (Models A, B, C).
    • The Purcell factor was calculated by integrating the energy dissipation (P0(ω)) from the electric dipole, providing insights into EMLDOS enhancement by SPP and LSP modes.

The ability to precisely control the spontaneous emission rate of quantum emitters at the nanoscale using plasmonic structures has direct implications for several advanced technologies:

  • Quantum Communications and Computing: Enhanced transition probability (Purcell effect) speeds up optical signal processing, crucial for high-speed quantum communication links and integrated photonic circuits.
  • High-Efficiency Light Sources: The findings provide a basis for manipulating quantum emitters (QEs) to enhance the efficiency of classical light sources, such as Light Emitting Diodes (LEDs) and Laser Diodes (LDs), by increasing the radiative decay rate.
  • On-Chip Photonics: Coupling QEs to propagating SPPs (as demonstrated in Sample B) allows for the transfer of quantum information as a surface wave, enabling the realization of integrated on-chip devices where QEs communicate with distant optical elements.
  • Plasmonic Sensing: NV centers coupled to LSPs (Sample C) create highly sensitive nanoscale probes. The strong EMLDOS enhancement in nanoholes can be leveraged for advanced quantum sensing applications.
  • Nanodiamond-Based Devices: The use of stable NV centers in NDs makes this technology compatible with robust, room-temperature quantum device fabrication.
View Original Abstract

We measured the second-order correlation function of the cathodoluminescence intensity and investigated the Purcell effect by comparing the lifetimes of quantum emitters with and without metal structure. The increase in the electromagnetic local density of state due to the coupling of a quantum emitter with a plasmonic structure causes a shortening of the emitter lifetime, which is called the Purcell effect. Since the plasmon-enhanced electric field is confined well below the wavelength of light, the quantum emitter lifetime is changed in the nanoscale range. In this study, we combined cathodoluminescence in scanning (transmission) electron microscopy with Hanbury Brown-Twiss interferometry to measure the Purcell effect with nanometer and nanosecond resolutions. We used nitrogen-vacancy centers contained in nanodiamonds as quantum emitters and compared their lifetime in different environments: on a thin SiO2 membrane, on a thick flat silver film, and embedded in a silver film. The lifetime reductions of nitrogen-vacancy centers were clearly observed in the samples with silver. We evaluated the lifetime by analytical calculation and numerical simulations and revealed the Purcell effects of emitters coupled to propagating and localized surface plasmons. This is the first experimental result showing the Purcell effect due to the coupling between nitrogen-vacancy centers in nanodiamonds and surface plasmon polaritons with nanometer resolution.