Skip to content

Chemical bonding evolution of near-surface nitrogen and defects in ultra-low-energy nitrogen implanted (100) single crystal diamond probed by synchrotron x-ray spectroscopies

FieldValue
JournalJournal of Physics D Applied Physics
Published2026-04-17
AuthorsAmaresh Das, Mohan Kumar Kuntumalla, Sayantan Maity, Arsène Frédérique Michel Chemin, Maria Brzhezinskaya
DOI10.1088/1361-6463/ae6122

A 700 °C in situ anneal changed the implanted nitrogen bonding balance. The N K-edge peak-area share moved from 64% C-N and 36% C=N to 74% and 26%, respectively. The same anneal removed about 36% of the near-surface nitrogen. XPS and NEXAFS identify C=N as the less thermally stable configuration under these conditions. The remaining concern is an sp2-carbon π* state 1.8 eV below the NV− ground state.

Nitrogen bonding favors C-N after annealing

Section titled “Nitrogen bonding favors C-N after annealing”

Surface nitrogen measured by XPS fell from 4.7 at% in the as-prepared sample to 3.0 at% after annealing. The shift is consistent with partial desorption or transformation of nitrogen in C=N sites. C-N bonds remain comparatively stable.

This result applies to 100 eV N2+ implanted (100) single-crystal diamond treated in situ at 700 °C. The study did not test other anneal temperatures or implant doses.

Controlled implantation and depth-sensitive measurement

Section titled “Controlled implantation and depth-sensitive measurement”

Commercially available (100)-oriented single-crystal diamond was treated with microwave hydrogen plasma. The treatment removed surface graphitic and non-diamond carbon impurities and produced hydrogen termination.

Before implantation, the sample was annealed at 800 °C for 5 min in ultrahigh vacuum. The authors used this step to remove contaminants introduced during handling.

Implantation used 100 eV N2+ ions at room temperature. The beam struck normal to the surface at a dose of 1 × 10^15 ions cm−2, with a current of approximately 20 nA. Each N2+ ion dissociated on impact and distributed its 100 eV equally between two nitrogen atoms. The effective nitrogen energy was therefore 50 eV per atom.

A brief 300 °C anneal defined the as-prepared state. Synchrotron XPS and NEXAFS were measured at a base pressure of 1 × 10−9 mbar.

C 1s XPS used photon energies of 335, 435, 535, and 700 eV. They produced photoelectron kinetic energies of approximately 50, 150, 250, and 415 eV. N 1s XPS used 450, 550, and 650 eV photons, yielding kinetic energies of about 50, 150, and 250 eV.

NEXAFS spectra were collected in total electron yield mode with a −30 V retarding potential. The XPS resolution was better than 0.1 eV. The C K-edge monochromator resolution was approximately 100 meV.

The pristine hydrogen-terminated surface contained 2.1 at% oxygen. After implantation, oxygen rose to 5.2 at%, a roughly 2.5-fold increase. The authors suggest that new defect sites increased adsorption of oxygen and water from air.

After the 700 °C anneal, oxygen fell to 0.5 at%. This corresponds to nearly 90% oxygen desorption during annealing. The O 1s binding energy indicates that oxygen mainly occupied weakly bound C-Ox configurations rather than nitrogen-oxygen species.

XPS survey spectra detected no other impurity features within their detection limits.

Lower photoelectron kinetic energy gives greater surface sensitivity in XPS. The 283.7 ± 0.2 eV C 1s component decayed rapidly as photon energy increased. This locates the associated bonding near the surface.

The component cannot be assigned to one carbon state. It may contain sp2 carbon, carbon interstitials, vacancies, and implantation-induced stress. The 285.0 eV sp3 C peak remained strongest at every photon energy. This supports a largely preserved diamond framework after implantation.

Annealing brought the relative sp3 C signal close to its pristine value. About 20% of the combined sp2 C/Cdef/Cstress component remained near the surface.

The authors attribute partial sp3 recovery to mobile point defects and nitrogen substitution. Carbon interstitials become mobile near 300 °C, while vacancies gain substantial mobility near 700 °C. Interstitial nitrogen can migrate through vacancies and form substitutional nitrogen. Nitrogen and hydrogen desorption also reduce stress and permit local rearrangement.

At 335 eV, annealing reduced the CHx/C=N component by about 28%. The oxygen-bearing/C-N component fell by nearly 72%.

The N 1s peak also narrowed after annealing. Its FWHM changed from 2.1, 2.0, and 2.0 eV to 2.0, 1.8, and 1.8 eV. These values apply at photon energies of 450, 550, and 650 eV, respectively. The narrowing is consistent with reduced disorder and more uniform bonding. XPS alone cannot distinguish C-N from C=N.

The hydrogen-terminated diamond retained its core C K-edge features after implantation and annealing. These were the 289.1 eV core exciton and the 302.5 eV second bandgap minimum. Implantation caused only a moderate reduction in exciton intensity. The annealed spectrum showed no significant change in that intensity.

Implanted HOPG served as a reference for nitrogen-related resonances. It placed π*(C=N) near 286.8 eV and σ*(C-N) near 288.7 eV. The pristine diamond C-H σ* feature at 287.6 eV was no longer clearly resolved after implantation and annealing.

Pre-edge broadening prevented reliable separation of individual C-H and nitrogen contributions at the carbon edge. The extent of hydrogen desorption could not be quantified. The absence of the N3 resonance at the N K-edge argues against long-range near-surface graphitization. Short-range graphitization in the extreme surface could not be excluded.

XPS emphasizes the topmost surface, while NEXAFS samples a slightly deeper near-surface volume. The authors use this difference to explain the opposing defect signals after annealing.

Earlier 100 eV N2+ work on polycrystalline diamond assigned a C 1s feature near 284 eV to stress and interstitial nitrogen. Higher-energy implantations at 200 eV to 1 keV often assigned the same energy range to sp2 carbon and C=C bonds.

The present 283.7 ± 0.2 eV signal in (100) single-crystal diamond cannot separate these causes. Its rapid surface decay agrees with the detectable surface defects reported for this orientation.

Earlier Mg-K-alpha XPS used 1253.6 eV photons. The photoelectron inelastic mean free path was about 25 Å, exceeding the 5 to 6 Å projected nitrogen range. Synchrotron XPS here varied kinetic energy from 50 to 415 eV, giving stronger depth discrimination.

An earlier polycrystalline diamond study annealed to 1000 °C and reported a stable subsurface sp2-rich carbon layer. At 700 °C here, sp3 C partly recovered, but about 20% of the combined surface defect component remained.

Charge trapping implications for shallow NV centers

Section titled “Charge trapping implications for shallow NV centers”

The authors placed measured unoccupied states relative to the valence-band maximum and the NV− 3A2 ground state. They adopted a 5.5 eV diamond bandgap. The CBM was placed 0.19 eV above the 289.1 eV exciton, and the pre-edge scale was aligned to the VBM.

Surface stateEnergy above VBMPosition relative to NV− 3A2Charge-transfer assessment
C-H σ*3.8 eV0.8 eV aboveDoes not efficiently accept charge
C-N σ*4.9 eVWell aboveUnlikely to trap charge
C=N π*3.0 eVClose to or slightly aboveLimited charge trapping
sp2 C π*1.2 eV1.8 eV belowEfficient electron trap

The hydrogenated surface also has negative electron affinity, which complicates NV− charge stability despite the high C-H state.

The sp2 carbon-derived π* states sit well below the NV− ground state. They can accept charge from shallow NV− centers and may contribute to surface noise. Annealing reduces the relative C=N population. The authors suggest that this may lower charge transfer into that state and improve NV− charge stability.

The study did not measure NV− spin coherence or charge-state lifetime. Improved shallow-center stability is therefore a mechanistic expectation, not a measured device outcome. For device engineering, implantation, annealing, and sp2 suppression must be evaluated together. The paper identifies localized sp2 carbon as the main surface-state risk to NV− charge.

For shallow NV− quantum-device research, the material requirement is (100)-oriented single-crystal diamond with a hydrogen-terminated surface before nitrogen implantation and annealing. 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal diamond plates and wafers up to 125 mm. Any custom build would need a separate specification for orientation and surface treatment through engineering support.

View Original Abstract

Abstract Controlled modification of diamond surfaces is essential for achieving stable, high coherence shallow negatively charged nitrogen-vacancy (NV − ) centers for quantum technologies. Here, hydrogen-terminated (100) single crystal diamond surfaces were modified by ultra-low-energy (100 eV) N 2 + implantation, corresponding to an effective nitrogen energy of 50 eV per atom. The implantation-induced bonding configurations and their thermal evolution were investigated using synchrotron-based, energy-dependent x-ray photoelectron spectroscopy (XPS) and near-edge x-ray absorption fine structure (NEXAFS) spectroscopy. XPS reveals the formation of C-N and C=N bonding states together with implantation induced surface carbon defects associated with sp 2 -like carbon, structural disorder, and stress. NEXAFS identifies distinct unoccupied σ * and π * surface states while preserving the characteristic excitonic and second bandgap features of diamond, with no evidence of long-range graphitic domains in near-surface region. In situ annealing at 700 °C leads to preferential nitrogen desorption from C=N sites, stabilization of thermally robust C-N bonding, recovery of sp 3 carbon order through defect annihilation and nitrogen substitution, and relaxation of implantation induced stress, leaving a small stable fraction of near-surface sp 2 carbon. A mechanistic energy level analysis indicates that π *(C=N) and σ *(C-N) states lie above the NV − ground state and are less detrimental to shallow NV − charge stability, whereas sp 2 carbon π * states located approximately 1.85 eV below the NV − ground state may act as charge traps. These findings highlight the importance of minimizing surface sp 2 carbon. Overall, this study elucidates the bonding and electronic structure of ultra-low-energy nitrogen-modified diamond surfaces and provides guidance for engineering diamond terminations through N 2 + implantation for stabilization of shallow NV − centers.

  1. 2013 - The nitrogen-vacancy colour centre in diamond [Crossref]
  2. 2025 - Nitrogen-vacancy centers in epitaxial laterally overgrown diamond: towards up-scaling of color center-based quantum technologies [Crossref]
  3. 2022 - Diamond—the ultimate material for exploring physics of spin-defects for quantum technologies and diamondtronics [Crossref]
  4. 2025 - Detection of electron spin resonance of a single nitrogen-vacancy center in diamond using the excited-state lifetime [Crossref]
  5. 2016 - Production of bulk NV centre arrays by shallow implantation and diamond CVD overgrowth [Crossref]
  6. 2024 - Enhanced quantum properties of shallow diamond atomic defects through nitrogen surface termination [Crossref]
  7. 2020 - Chemical vapour deposition diamond single crystals with nitrogen-vacancy centres: a review of material synthesis and technology for quantum sensing applications [Crossref]
  8. 2019 - Origins of diamond surface noise probed by correlating single-spin measurements with surface spectroscopy [Crossref]
  9. 2011 - Chemical control of the charge state of nitrogen-vacancy centers in diamond [Crossref]
  10. 2021 - Magnetometry based on the excited-state lifetimes of a single nitrogen-vacancy center in diamond [Crossref]