Efficient creation of shallow NV$^-$ ensembles by high-angle ion implantation

AI summary

Overview

Research area: Quantum sensing and materials engineering — specifically the fabrication of shallow nitrogen-vacancy (NV⁻) centers in diamond for nanoscale spin sensing.

Technical level: Intermediate. The paper is readable for someone with basic familiarity with NV centers and ion implantation, but it assumes comfort with concepts such as implantation dose, vacancy generation, Ramsey/Hahn-echo sequences, and dynamical decoupling.

Scope (one sentence): The paper demonstrates that implanting ¹⁵N ions at high angles (exceeding 60°) into high-purity diamond produces shallow NV⁻ ensembles with yields approaching 10% and maximum effective depths below 10 nm, characterized optically and through spin-coherence and nuclear-spin sensing measurements.

What This Paper Is About

NV⁻ centers sitting only a few nanometers below a diamond surface are powerful sensors for detecting nuclear spins in materials placed on that surface, but making them shallow and numerous at the same time is a longstanding materials problem. Typical implantation-based methods yield only about 0.1–1% conversion of implanted nitrogen into NV⁻, largely because too few vacancies are created per nitrogen atom and because the charge state is unstable near the surface. The authors show that simply tilting the diamond relative to the ion beam during implantation — a geometric change requiring no new hardware — improves both the shallowness and the conversion yield.

Key Contributions

1. A mask-free, high-angle implantation recipe for shallow NV⁻ ensembles. Implanting ¹⁵N⁺ at 5 keV with effective angles exceeding 60° (achieved by tilting the diamond) followed by forming-gas annealing and oxygen termination produces charge-stable NV⁻ ensembles.

2. A physical concept backed by SRIM simulations. High-angle incidence reduces the projected dose by cos θ, causes ions to stop shallower through oblique scattering, and enhances backscattering — which suppresses retained nitrogen near the surface while vacancies are still generated, raising the vacancy-to-nitrogen ratio where NV formation is most vulnerable to vacancy loss.

3. Quantified yield enhancement. Yields of 7.8%, 8.3%, 4.5%, and 2.4% at 60°, 70°, 80°, and 85° respectively, cross-checked with an independent Poissonian-statistics estimate from photoluminescence fluctuations (6.5%, 7.9%, 4.5%, 2.5%).

4. Depth and coherence benchmarking. Direct depth estimation via ¹¹B nuclear spins in attached hexagonal boron nitride (hBN) flakes gives maximum effective depths of 8.2, 6.9, 6.2, and 5.9 nm, and spin measurements show coherence comparable to single shallow NV⁻ centers at similar depths.

Main Findings

• PL uniformity: Photoluminescence maps over a (100 μm)² area with 250 nm sampling resolution are spatially uniform except for a small number of localized spots attributed to surface topography. For the 60° sample, the intensity histogram is normal with I_PL = 7.04 Mcps and σ_PL = 0.21 Mcps.

• Charge-state purity: Emission is dominated by NV⁻ (zero-phonon line at 637 nm); NV⁰ emission (zero-phonon line at 575 nm) is weak and almost negligible within the detection window.

• NV⁻ areal densities: 3.56 × 10¹¹, 2.29 × 10¹¹, 0.48 × 10¹¹, and 0.10 × 10¹¹ cm⁻² for 60°, 70°, 80°, and 85°, obtained by calibrating PL intensity against single NV⁻ centers.

• Corresponding nitrogen densities (SRIM): 4.57 × 10¹², 2.76 × 10¹², 1.06 × 10¹², and 0.40 × 10¹² cm⁻², giving the yields listed above — well above the 0.1–1% typical of shallow NV⁻ creation, with 5% shown as a reference line.

• Yield falls at the highest angles: The drop is attributed to reduced effective nitrogen dose and enhanced backscattering, including surface-potential-induced reflection, which increases nitrogen–vacancy separation.

• Implanted nitrogen is the source: Ramsey fringes show a pair of peaks separated by 3.1 MHz, consistent with the hyperfine interaction of ¹⁵NV⁻, indicating negligible contribution from residual nitrogen in the CVD layer or substrate.

• Dephasing times: T₂* = 139 ns with p = 0.6 at 85°; T₂* = 184–272 ns with p = 0.5–0.7 at the other angles. The sub-exponential stretch is attributed to a distribution of T₂* across NV⁻ centers at different depths.

• Hahn-echo coherence improves with angle: (T₂, p) = (2.83 μs, 1.1), (3.97 μs, 1.0), (6.32 μs, 1.0), and (7.20 μs, 0.9) for 60°, 70°, 80°, and 85°.

• Decoherence tracks nitrogen density, with an offset: Relative nitrogen areal densities are 100%, 60%, 23%, and 8.8%; relative decoherence rates 1/T₂ are 100%, 71%, 45%, and 39%. The deviation at 85° suggests additional surface-related noise acting as an effective offset.

• No ¹³C envelope modulation was observed in any sample, confirming NV⁻ formation exclusively within the ¹²C-enriched CVD layer.

• Dynamical decoupling performance: With XY16-N128, (T₂, p) = (19.2 μs, 0.8), (28.4 μs, 0.8), (43.9 μs, 0.7), and (54.8 μs, 0.6) for 60° through 85°. T₂ scales approximately as N^0.4, weaker than the ~N^0.7 expected for bulk-nitrogen-dominated noise and consistent with surface-noise-dominated shallow single NV⁻ reports (N^0.2–N^0.7).

• Proton signals appear without attached samples: NMR signals from protons and from the ¹⁵N of NV⁻ centers are resolved around 1 MHz, with protons attributed to surface termination or hydrogen incorporated in the lattice, since no oil-immersion objective was used.

• Direct depth measurement: ¹¹B spins in hBN flakes were detected at 176 mT using XY16-N256, with signals at the expected ¹¹B resonance of 2.21–2.22 MHz and correlation spectroscopy confirming the assignment. Extracted nuclear-spin coherence times T₂,ₙ* = 35.4–66.7 μs.

• Depths below 10 nm: d_nv = 8.2, 6.9, 6.2, and 5.9 nm for 60°, 70°, 80°, and 85°, with spatial variation of about 0.5 nm. These are shallower than the 10–15 nm range reported for conventional 7° implantation at 4–6 keV.

• A residual offset: SRIM predicts mean depths and straggles of (4.8 nm, ±2.5 nm), (4.0 nm, ±2.2 nm), (3.3 nm, ±2.0 nm), and (3.2 nm, ±1.8 nm); measured d_nv exceeds these by about 3 nm, more than a typical hydrocarbon interlayer (1–2 nm) would explain, possibly indicating residual channeling.

• Prior high-angle work for context: One earlier study used θ = 85° at 85 keV, giving ~20 nm depth and about 1% yield; another combined θ = 45° with a SiO₂ mask and stayed below 1% yield.

Methodology in Plain English

The team started with high-purity diamond consisting of a ¹²C-enriched CVD layer grown on (100)-oriented type-IIa substrates. They cleaned the surface and implanted ¹⁵N⁺ ions at 5 keV with a nominal dose of 1 × 10¹³ cm⁻² — a dose defined by the implanter for normal incidence. Rather than using any special optics, they simply tilted the diamond relative to the beam so the effective angle exceeded 60°. After implantation they annealed in forming gas to let vacancies pair up with nitrogen, then annealed in air to oxygen-terminate the surface and stabilize the NV⁻ charge state.

They first used SRIM simulations to predict how angle changes the depth profiles of nitrogen and vacancies, then checked those predictions experimentally with a confocal microscope: photoluminescence maps gave NV⁻ density after calibration against known single centers, and spectra confirmed the charge state. Spin properties were measured with time-resolved PL, Ramsey interferometry, Hahn echo, and XY16 dynamical decoupling at roughly 29.5 mT. Finally, they estimated depth by attaching hBN flakes to the diamond and fitting the ¹¹B nuclear-spin signal to a theoretical model, treating the bias field, NV–hBN separation, and nuclear coherence time as free parameters with the flake assumed 100 nm thick.

Why This Matters

The work offers a practical route to the shallow, dense NV⁻ ensembles that nanoscale magnetic resonance needs, using a knob — sample tilt — that any implantation facility already has. It also reframes an old constraint: instead of fighting the low vacancy-per-nitrogen ratio near the surface, the geometry is used to change it.

Real-world applications:

• Nanoscale NMR and NQR spectroscopy of thin materials and interfaces, exploiting the demonstrated detection of ¹¹B spins in hBN.
• Sensing of nuclear spins in van der Waals materials and other samples simply placed on the diamond surface.
• Creation of vacancy-related quantum defects in other wide-bandgap materials such as hBN and SiC.
• Formation of NV⁻ centers in high-nitrogen, low-NV-density HPHT type-Ib diamond by implanting species such as phosphorus.

Industry relevance: The method requires no specialized implantation systems, no implantation masks, no post-implantation etching, and no delta-doped CVD growth layers. That simplicity lowers the fabrication barrier for producing shallow quantum sensors and could make wafer-scale processing more straightforward for quantum-sensing device developers.

Future Directions

• Raising the density further: Increasing the nitrogen dose could boost NV⁻ numbers, but the authors warn it may promote vacancy clustering or amorphize the diamond; high-temperature implantation is proposed to encourage vacancy diffusion during the process itself.

• Explaining the enhanced yield: The authors state that the origin of the markedly enhanced yield remains an open question, since the total vacancy density is not expected to be dramatically higher than in conventional methods.

• Modeling near-surface damage: More detailed molecular dynamics simulations are called for to understand how damage is distributed near the surface.

• Closing the depth gap: The roughly 3 nm discrepancy between measured d_nv and SRIM predictions, possibly from residual ion channeling or interfacial effects, needs investigation.

• Extending to other materials and emitters: Applying the geometric approach to vacancy defects in hBN, SiC, and to phosphorus implantation in type-Ib diamond.

Target Audience

Researchers and graduate students in quantum sensing and diamond defect engineering; materials scientists working on ion implantation and near-surface modification of wide-bandgap semiconductors; and engineers in quantum-device fabrication who need a simple, mask-free process for shallow spin sensors. Readers new to the field will need some background on NV centers and spin-coherence measurements to follow the characterization sections.

Authors’ abstract

Negatively charged nitrogen-vacancy (NV$^-$) centers located a few nanometers below the diamond surface are key quantum defects for nanoscale sensing of external spins. However, the creation of shallow NV$^-$ centers with high yield remains a materials challenge. Here, we demonstrate that high-angle ion implantation enhances the creation efficiency of shallow NV$^-$ centers. By implanting $^{15}$N ions at angles exceeding 60$^\circ$, we achieve high NV$^-$ yields approaching 10% with effective NV$^-$ depths below 10 nm. These yields are significantly higher than those typically reported for shallow NV$^-$ creation. The enhanced NV$^-$ yield is consistent with an increased vacancy-to-nitrogen ratio in the near-surface region, which is expected to promote NV formation during annealing. The created NV$^-$ ensembles show coherence properties comparable to those of single shallow NV$^-$ centers at similar depths, and allow detection of nuclear spins in van der Waals materials attached to the diamond surface. Our results establish geometric control of ion implantation as a simple and broadly applicable approach to engineer shallow vacancy-related quantum defects.