Skip to content

High ODMR contrast and alignment of NV centers in microstructures grown on heteroepitaxial diamonds

MetadataDetails
Publication Date2024-04-15
JournalApplied Physics Letters
AuthorsJan Engels, Jürgen Weippert, Tingpeng Luo, Jan Kustermann, Patricia Quellmalz
InstitutionsFraunhofer Institute for Applied Solid State Physics
Citations4
AnalysisFull AI Review Included
  • Breakthrough in Wafer-Scale Quantum Sensing: Demonstrates pyramidal diamond microstructures grown on heteroepitaxial single crystal diamond (SCD) substrates (Ir/YSZ/Si) for highly aligned Nitrogen-Vacancy (NV) center ensembles.
  • Engineered Buffer Layer Architecture: Introduction of a nominally undoped buffer layer between the nitrogen-doped substrate and the active NV-rich pyramid significantly suppresses background spin noise, improving signal isolation.
  • Record Heteroepitaxial ODMR Contrast: Achieves an Optically Detected Magnetic Resonance (ODMR) contrast of 19% (no magnet) and 13% (single resonance with magnet) on heteroepitaxial structures—more than doubling previous benchmarks and approaching the theoretical weak-excitation limit of 22%.
  • Near-Perfect NV Alignment: Pyramidal {111} facets grown on the buffer layer show >90% alignment to a single NV orientation, enabling four spatially isolated vector channels on a single wafer.
  • Scalable Fabrication: Pyramid base lengths scalable from 6-50 µm, unlocking wafer-scale production of integrated quantum optical and vector magnetometry devices.
  • Application Target: Designed for AFM-tip vector magnetometers and chip-scale quantum navigation, requiring high NV density, uniform alignment, and homogeneous large-area facets.
ParameterValueUnitContext
Substrate TemplateIr / YSZ / Si (001)—Heteroepitaxial CVD single crystal diamond
Pyramid Base Length6 - 50µmScalable lateral facet dimension
Pyramid Side Facets{111}—Crystallographic orientation of NV-aligned facets
Pyramid Growth Height (Pyramid II)~1.6µmHeight difference at the base
Buffer Layer TypeNominally undoped SCD—Decouples active NVs from substrate defects
CVD Series I — C/H Ratio0.75%N-doped substrate growth
CVD Series I — Temperature785°CHeteroepitaxial substrate
CVD Series I — Pressure300mbarReactor process parameter
CVD Series II — C/H Ratio0.75%Buffer layer + pyramid growth
CVD Series II — Temperature825°CBuffer layer + pyramid growth
CVD Series II — Pressure300mbarReactor process parameter
CVD Series III — C/H Ratio0.5%Homoepitaxial (HPHT) reference
CVD Series III — Temperature825°CHomoepitaxial reference
P1 (Substitutional N) Density1.8 - 5.0ppmDetermined via UV-vis / DEER
NV Density (Wafer IV)3.2ppbCalibrated PL spectroscopy
NV Density (Series I/II)0.3ppbCalibrated PL spectroscopy
NV⁻ / NV Ratio (Pyramid II Facets)96%Negative charge state yield
N/C Ratio (Substrate I/II)1143ppmGas phase doping ratio
N/C Ratio (Wafer IV)3333ppmGas phase doping ratio
ODMR Contrast — Hetero Series I (no magnet)15%Substrate-coupled pyramids
ODMR Contrast — Hetero Series II (no magnet)19%Buffer-layer pyramids (this work)
ODMR Contrast — Homo Series III (no magnet)28%Reference homoepitaxial pyramids
ODMR Contrast — Hetero Series II (with magnet, single spin)13%Vector-resolved measurement
ODMR Contrast — Hetero Series I (with magnet)6%Previous benchmark
ODMR Contrast — Homo Series III (with magnet)16%Reference value
T2 Dephasing Time — Hetero Series I48 ± 7nsNV spin coherence in substrate-grown pyramid
T2 Dephasing Time — Hetero Series II25 ± 3nsNV spin coherence in buffer-layer pyramid
T2 Dephasing Time — Wafer IV160 ± 19nsBulk wafer measurement
T2 Dephasing Time — Homo Series III406 ± 22nsReference homoepitaxial pyramid
Theoretical Weak-Excitation Contrast Limit22%Maximum NV ODMR (low power)
Theoretical Strong-Excitation Contrast Limit~30%Maximum NV ODMR (high power, Rabi regime)
  1. Heteroepitaxial Diamond Growth: CVD diamond deposition on 4-inch Ir/YSZ/Si(001) templates to produce wafer-scale monocrystalline diamond.
  2. Selective Nitrogen Doping: Series I used an N/C ratio of 1143 ppm in the gas phase to generate a 1.8 ppm P1 background in the bulk substrate.
  3. Buffer Layer Deposition (Series II): A nominally undoped CVD layer grown over the N-doped substrate at 825 °C, 300 mbar, C/H = 0.75% to isolate the pyramid volume from substrate spin noise.
  4. Pyramid Microfabrication: Reactive ion etching and anisotropic selective CVD growth to produce inverted pyramids exposing four discrete {111} side facets per mesa, with base dimensions of 6-50 µm.
  5. Secondary Pyramid Overgrowth: {111} facets selectively overgrown with NV-friendly diamond (Series II: C/H = 0.75%; Series III: C/H = 0.5%) at 825 °C to embed uniformly aligned NV ensembles.
  6. Confocal PL Characterization: Single-NV and ensemble PL mapping to determine NV density and NV⁻/NV charge ratio.
  7. CW ODMR with and without External Magnet: Frequency-swept continuous-wave ODMR (2.7-2.9 GHz range) with aligned bias magnet to resolve all four NV orientations per facet and quantify vector alignment.
  8. T2 Coherence Measurement: Hahn-echo or equivalent dephasing sequences on NV ensembles across all sample series.
  9. Defect Density Quantification: UV-vis spectroscopy for bulk P1 concentration, DEER for facet-localized P1 density.
  • Applicable Materials: This research requires Optical Grade Single Crystal Diamond (SCD) heteroepitaxial wafers grown on Ir/YSZ/Si templates, optionally with engineered N-doped and undoped buffer layers for NV ensemble integration. 6CCVD supplies custom heteroepitaxial SCD substrates (up to 125 mm equivalent) and Polycrystalline Diamond (PCD) reference samples suitable for NV-host characterization or sacrificial buffer development.
  • NV Engineering Compatibility: For groups seeking to replicate or extend these quantum sensing architectures, 6CCVD provides nominally undoped SCD films (thickness 0.1-500 µm) with controlled nitrogen incorporation profiles, enabling the same buffer-layer-plus-active-volume layer stack demonstrated here.
  • Customization Potential: The pyramid geometries (6-50 µm) and {111} facet orientation require precision laser micromachining and orientation-controlled polishing. 6CCVD’s in-house laser cutting and shaping services deliver micrometer-accurate diamond structures, with optional custom metalization (Ti, Pt, Au, W, Pd, Cu) for backside contacts, microwave striplines, and AFM-tip integration.
  • Surface Quality for Facet Uniformity: The {111} facets must minimize scattering and strain inhomogeneities to preserve T2. 6CCVD’s SCD polishing reaches Ra < 1 nm, and PCD surfaces achieve Ra < 5 nm — directly addressing the lattice-strain gradients that limit coherence in this study.
  • Wafer-Scale Scalability: The authors explicitly call out wafer-scale manufacturing as the key advantage of heteroepitaxy. 6CCVD’s capability to deliver plates/wafers up to 125 mm (PCD) and equivalent large-area SCD substrates positions our materials as a direct enabler for transitioning this research from single-pyramid prototypes to full-wafer quantum device batches.
  • Engineering Support: 6CCVD’s in-house PhD engineering team can advise on material selection, doping profiles, buffer layer thickness optimization, and metallization stack design for NV-based vector magnetometers, AFM-tip quantum sensors, and integrated quantum photonic devices.
  • Global Logistics: All custom orders ship worldwide (DDU default, DDP available), with engineering consultation available throughout the design-to-delivery cycle.

For custom specifications, heteroepitaxial SCD wafer inquiries, or material consultation, visit 6ccvd.com or contact our engineering team directly.

View Original Abstract

Heteroepitaxial chemical vapor deposition is the most promising option to fabricate wafer-scale monocrystalline diamonds for quantum applications. Previously, we demonstrated the feasibility to manufacture functional micrometer-sized pyramids on as-grown heteroepitaxial diamond as well as their quantum optical characteristics. Due to high background signals and microfabrication challenges, these pyramids could not compete with homoepitaxially grown structures. In this study, we overcame these problems with a nominally undoped buffer layer between the heteroepitaxial substrate and the pyramidal microstructure to reduce the signal-to-noise ratio from the substrate on the spin measurements of the nitrogen-vacancy (NV) center. Moreover, the microfabrication was improved to reach a higher angle of the pyramidal side plane, corresponding to the {111} facets. These improvements lead to pyramids on which each facet contains almost purely only one of the four possible NV orientations as shown by optically detected magnetic resonance (ODMR). ODMR shows a very high contrast of 19% without an external magnet and of 13% for a single spin resonance in the presence of a magnetic field. The contrast is more than doubled compared to our previous study. The T2* dephasing time of the NV centers of the samples ranges from 0.02 to 0.16 μs. The P1 center is a single substitutional nitrogen center, and the P1 densities range from 1.8 to 5 ppm.