200 keV energy electron irradiation of single-crystal diamond - Quantification of vacancy and nitrogen-vacancy production
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Physical review. B./Physical review. B |
| Published | 2026-06-15 |
| Authors | Anonymous, Ben L. Green, Mark E. Newton, Lillian B. Hughes Wyatt, Ania C. Bleszynski Jayich |
| DOI | 10.1103/hls1-92jc |
A 200 keV electron dose estimated at (10^{19}\ \mathrm{e^-/cm^2}) produced an exponential vacancy profile with a (12 \pm 1\ \mu\mathrm{m}) decay length and a surface concentration of ((1.1 \pm 0.2)\times10^{17}\ \mathrm{V/cm^3}) in (001) electronic-grade single-crystal diamond. The NV profile did not follow the vacancy profile. The authors’ kinetic model attributes this mismatch to vacancy clusters and nitrogen-divacancy (NV_2) formation.
Irradiated material and measurement method
Section titled “Irradiated material and measurement method”The specimen was a (2.09 \times 2.09 \times 0.52) mm CVD plate with (001) orientation, less than 1 ppb boron, and less than 5 ppb nitrogen. The beam was directed perpendicular to the 2.09 mm face. Its 32 mm path through air minimized energy loss. A comparison with 200 keV irradiation in a vacuum transmission electron microscope indicated that the remaining energy loss was negligible.
The 2.09 by 0.52 mm sides were laser cut and polished so photoluminescence could scan depth from the edge. Optical absorption ran through the irradiated face at 77 K. Confocal PL scans ran at room temperature and 77 K with 2 µm steps. Excitation used 660, 514, and 488 nm lasers. The 488 nm light partially converts (NV^-) to (NV^0), while 514 nm probes both charge states with less perturbation.
The initial sideband scan used 660 nm excitation, a 0.9 NA ×100 objective, a 1 µm spot, and a focus 10 µm below the polished side. Cryogenic anneal scans used a 0.5 NA ×50 long-working-distance objective, so those measurements fitted zero-phonon lines.
Quantified vacancy profile
Section titled “Quantified vacancy profile”Only neutral vacancies appeared in absorption through the 741 nm GR1 zero-phonon line. The 393.5 nm ND1 line for negative vacancies was absent. The mean concentration was ((2.5 \pm 0.5)\times10^{15}\ \mathrm{V/cm^3}) across the 0.52 mm thickness, equal to (14 \pm 3) ppb if the vacancies were uniformly distributed.
The authors fitted the GR1 sideband profile with an exponential convolved with a 2.1 µm Gaussian spectrometer response. Normalization linked the PL profile to bulk absorption and supplied the depth-dependent concentration. Significant vacancy signal extended to about 38 µm, consistent with the point at which 200 keV electrons lost enough energy for displacement.
Annealing response of vacancies
Section titled “Annealing response of vacancies”Each stage lasted 1 h at atmospheric pressure in nitrogen. A calibrated thermocouple checked the sample temperature, and diamond grit covered the sample to limit surface graphitization. Loading took about 15 min, which matters at high temperature because defect concentrations can change during loading.
No GR1 depth-profile change followed the 400 °C or 600 °C anneal. The 800 °C anneal reduced the signal by the same factor at every depth, suggesting uniform fractional vacancy loss. After 1000 °C, the profile changed shape and its maximum shifted to (49 \pm 2\ \mu\mathrm{m}). No GR1 remained after 1200 °C. The authors list complex formation, recombination with interstitials, and loss to extended defects or surfaces as possible losses.
NV and 733 nm defect profiles
Section titled “NV and 733 nm defect profiles”Neutral and negative NV emission became detectable after the 800 °C anneal. The neutral NV profile then peaked at (30 \pm 5\ \mu\mathrm{m}). After 1000 °C, 488 nm excitation placed the neutral NV peak at (48 \pm 5\ \mu\mathrm{m}). Under 514 nm excitation, neutral NV peaked at (54 \pm 5\ \mu\mathrm{m}), while negative NV reached its maximum at (64 \pm 5\ \mu\mathrm{m}).
After 1200 °C, both charge-state signals increased, and the neutral profile developed a shallow shoulder. The relative negative-to-neutral emission also rose, which the authors interpret as fewer competing electron traps.
The 733 nm zero-phonon line appeared after 1000 °C and followed the initial GR1 depth profile. Its intensity fell after 1200 °C, and the line was no longer detectable after that anneal. The authors place its detectable interval between 1000 and 1200 °C. Prior work associates this line with a multivacancy center. Its profile supports a vacancy-cluster assignment, but its structure remains unconfirmed. The authors speculate that it may be a trivacancy.
Kinetic model for cluster and NV formation
Section titled “Kinetic model for cluster and NV formation”The kinetic model represents isolated vacancies, substitutional nitrogen, NV, (NV_2), and vacancy clusters (V_2), (V_3), and (V_4). Vacancies diffuse, while the multivacancy clusters are immobile. The model omits charge states and interstitial reactions. The simulation started with 1000 ppb surface vacancies, a 12 µm decay length, and 5 ppb nitrogen. Formation used a 2.2 eV activation energy. Vacancy diffusion used 2.0 eV, (V_2) and (V_3) dissociation used 3.5 eV, and (NV_2) dissociation used 3.8 eV.
At 1000 °C, the model produced a surviving vacancy peak near 40 µm and an NV concentration maximum near 64 µm. It placed the vacancy-nitrogen crossover near 58 µm. Nearly all substitutional nitrogen became (NV_2) in the first about 40 µm. At 100 µm, nitrogen exceeded vacancies by about 20 to 1.
A second 1 h anneal at 1200 °C released vacancies from (NV_2) and increased modeled NV in the first 40 µm. The model placed 97% of vacancies in (V_4) clusters after this stage. That result conflicts with the proposed (V_3) assignment for the 733 nm line because the model loses too much (V_3). The authors claim only qualitative agreement and did not fit the rate constants to the measured profiles.
Comparison with theory and prior work
Section titled “Comparison with theory and prior work”Damage-threshold measurements for (100) incident electrons give (145 \pm 15) keV, consistent with a carbon displacement energy of (30 \pm 4) eV. The 200 keV beam was therefore above threshold.
An earlier simulation predicted a gradual vacancy decrease over 38 µm for 200 keV irradiation, while this experiment found an exponential decay with a (12 \pm 1) µm constant. Their maximum penetration depths agree, but the profiles differ. Earlier 300 keV transmission electron microscopy showed a gradual decrease to 140 µm.
The paper reports that measured vacancy production was about 50 times lower than a recent model predicted. The authors identify dose overestimation and incorrect production modeling as possible causes. They suggest including electronic-excitation effects on interstitial migration and the vacancy-interstitial recombination barrier. This uncertainty remains because the carbon interstitial concentration was not measured.
The paper’s implantation example uses (2 \times 10^9) 30 keV nitrogen ions per cm². SRIM calculations give about 5 ppb nitrogen and 400 ppb vacancies at 40 nm. The resulting vacancy excess supports the authors’ explanation for reduced NV yield after ion implantation.
Limits and engineering implications
Section titled “Limits and engineering implications”The model was not calibrated against enough data to constrain its many rate constants. The authors call for measurements with different vacancy and nitrogen concentrations, more annealing times and temperatures, and varied electron dose and energy. Direct detection of neutral or negative (NV_2) is required to test its role in charge trapping and NV yield.
The results change how NV production should be designed. In the first 20 µm, the data indicate about 200 vacancies for each substitutional nitrogen atom. Nitrogen availability therefore does not set the yield by itself. The authors identify (5 \pm 3) ppb (V^0) at (60 \pm 5\ \mu\mathrm{m}) as optimal for negative NV creation in material with 5 ppb nitrogen. For near-surface NV ensembles after the stated irradiation dose, they recommend removing the top 30 to 50 µm. Device planning must also account for the 733 nm cluster signal during the 1000 to 1200 °C interval and the loss of GR1 after 1200 °C.
Materials and support
Section titled “Materials and support”The study calls for a single-crystal (001) plate for electron irradiation and defect spectroscopy, with controlled low nitrogen and boron. 6CCVD’s diamond materials catalog lists single-crystal MPCVD material and plates up to 125 mm, but its listed single-crystal thickness ends at 500 µm. The paper’s 0.52 mm geometry and specified orientation should therefore be addressed through engineering support.
View Original Abstract
Electron irradiation and annealing treatments are a method of color center/defect creation in diamond. The depth profile of defects created by low-energy 200 keV electrons in single-crystal high-purity (type II) electronic grade diamond grown via chemical vapor deposition has been investigated. The depth profile of monovacancies created was found, using photoluminescence (PL), to decay exponentially, with a decay length <a:math xmlns:a=“http://www.w3.org/1998/Math/MathML”> <a:mrow> <a:mn>12</a:mn> <a:mo>±</a:mo> <a:mn>1</a:mn> <a:mspace width=“4pt”/> <a:mi>µ</a:mi> <a:mi mathvariant=“normal”>m</a:mi> </a:mrow> </a:math> and a production rate of <d:math xmlns:d=“http://www.w3.org/1998/Math/MathML”> <d:mrow> <d:mrow> <d:mo>(</d:mo> <d:mn>1.1</d:mn> <d:mo>±</d:mo> <d:mn>0.2</d:mn> <d:mo>)</d:mo> </d:mrow> <d:mo>×</d:mo> <d:msup> <d:mn>10</d:mn> <d:mrow> <d:mo>−</d:mo> <d:mn>1</d:mn> </d:mrow> </d:msup> <d:mspace width=“0.16em”/> <d:mi mathvariant=“normal”>V</d:mi> <d:mo>/</d:mo> <d:msup> <d:mrow> <d:mi mathvariant=“normal”>e</d:mi> </d:mrow> <d:mo>−</d:mo> </d:msup> <d:mo>/</d:mo> <d:mi>cm</d:mi> </d:mrow> </d:math> at the surface. The depth distribution of the neutral and negatively charged nitrogen-vacancy <h:math xmlns:h=“http://www.w3.org/1998/Math/MathML”> <h:mo>(</h:mo> <h:msup> <h:mrow> <h:mi>NV</h:mi> </h:mrow> <h:mrow> <h:mn>0</h:mn> <h:mo>/</h:mo> <h:mn>1</h:mn> <h:mo>−</h:mo> </h:mrow> </h:msup> <h:mo>)</h:mo> </h:math> center, and the 733 nm zero phonon line defect formed during an isochronal annealing study, have also been investigated by PL. The observed NV profiles, which do not match the vacancy profiles, can be qualitatively explained in terms of a simple model that includes the formation of vacancy clusters and the nitrogen-divacancy <i:math xmlns:i=“http://www.w3.org/1998/Math/MathML”> <i:mo>(</i:mo> <i:msub> <i:mi>NV</i:mi> <i:mn>2</i:mn> </i:msub> <i:mo>)</i:mo> </i:math> defect. The production of <j:math xmlns:j=“http://www.w3.org/1998/Math/MathML”> <j:mo>(</j:mo> <j:msup> <j:mrow> <j:mi>NV</j:mi> </j:mrow> <j:mrow> <j:mn>0</j:mn> <j:mo>/</j:mo> <j:mo>−</j:mo> </j:mrow> </j:msup> <j:mo>)</j:mo> </j:math> has been assumed to be nitrogen-limited, but this paper has shown that this is not the case, with NVs lost to the production of <k:math xmlns:k=“http://www.w3.org/1998/Math/MathML”> <k:msub> <k:mi>NV</k:mi> <k:mn>2</k:mn> </k:msub> </k:math> when the concentration of vacancies greatly exceeds that of substitutional nitrogen.
Boron Doped Single Crystal Diamond Plate
View Specs →
Single Crystal Cvd Diamond Plate
View Specs →
Customized Single Crystal Cvd Diamond Plate
View Specs →References we cover
Section titled “References we cover”- NV centres by vacancies trapping in irradiated diamond - experiments and modelling
- Creation of NV Centers in Diamond under 155 MeV Electron Irradiation
- Photoionization of negatively charged NV centers in diamond - Theory and ab initio calculations
- Window into NV center kinetics via repeated annealing and spatial tracking of thousands of individual NV centers
- Laser writing of individual nitrogen-vacancy defects in diamond with near-unity yield