Diamond processing for quantum defects
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Frontiers in Quantum Science and Technology |
| Published | 2026-05-29 |
| Authors | Ian Hammock, Tanvi Deshmukh, Midrel Wilfried Ngandeu Ngambou, Jonathan C. Marcks, Jessica C. Jones |
| DOI | 10.3389/frqst.2026.1745287 |
| Citations | 1 |
Quantum-grade diamond defects are limited by surface and processing damage as much as by bulk crystal quality. Isotopically engineered bulk NV centers have reached T₂ ≈ 1.8 ms at room temperature. Very shallow centers at depths ≤ 20 nm often show Hahn-echo T₂ in the tens of microseconds.
The review connects that gap to preparation, growth, activation, and interface control. It also reports a hydrogen-degassing experiment that increased ¹⁵NV⁻ yield in delta-doped diamond.
Processing quality sets the defect limit
Section titled “Processing quality sets the defect limit”The review collects reference points that show why no single processing metric predicts every type of defect.
| Metric | Material and condition | Reported value |
|---|---|---|
| NV T₂ | Isotopically engineered CVD diamond at room temperature | ≈ 1.8 ms |
| NV T₂ and T₂* | NV⁻ in a 100 nm deep isotopically purified CVD layer | T₂ ≈ 1.7 ms and T₂* ≈ 90 μs |
| Hahn-echo T₂ for shallow NV | Depth ≤ 20 nm | Tens of μs |
| SnV T₂ | Strain-tuned SnV in diamond at 4 K | ≈ 223 μs |
| SnV zero-phonon-line linewidth | Strain-tuned SnV at 4 K | 48 MHz |
| NV zero-phonon-line linewidth | Certain growth-incorporated near-surface NV centers | 1.2(5) GHz |
Depth, isotopic purity, growth conditions, strain, and post-growth processing all affect these results. Quantum communication favors buried, low-strain emitters with narrow, stable optical lines. Sensing often favors near-surface ensembles with high density and uniform properties. Scalable qubit architectures require deterministic placement and high activation yield.
Surface preparation removes damage before growth
Section titled “Surface preparation removes damage before growth”Mechanical polishing removes carbon through cleavage, abrasion, and thermal wear. It also creates roughness, morphological anisotropy, and a subsurface damage layer. Chemical-mechanical polishing adds chemical action and gentle abrasion, with reported roughness ≤ 10 nm Rq. One combined CMP and plasma process produced 0.30(15) nm RMS roughness over 400 μm². That value is near the 0.357 nm diamond lattice constant.
ICP-RIE removes damaged material before homoepitaxy. A typical guideline sets the etch depth at least three to five times the smallest polishing particle diameter. Argon causes micromasking, chlorine can impair near-surface NV centers, and oxygen etching is anisotropic. Intermixed Ar/Cl₂ and O₂ plasmas with regular pump/purge steps gave the best reproducibility in the cited work.
Before polishing, RCA1 uses H₂O₂:NH₄OH:H₂O at 1:1:5 with 30% peroxide, followed by ultrasonic deionized water. Post-CMP HF dissolves silica residues.
Before plasma ignition, the reactor is evacuated to about 2 × 10⁻⁶ Torr, and the substrate is heated to ≥ 500 °C for a few hours. This step desorbs interfacial nitrogen and oxygen-bearing contaminants. Hydrogen plasma passivates dangling bonds and removes sp² carbon, with bulk etch rates often ≤ 100 nm h⁻¹ under multi-kW power.
Growth and placement control defect depth
Section titled “Growth and placement control defect depth”Microwave plasma enhanced chemical vapor deposition is the main route to homoepitaxial quantum-grade diamond. A hydrogen-rich, low-pressure plasma converts CH₄ into reactive CHx radicals and supplies atomic hydrogen. Atomic hydrogen stabilizes sp³ bonding and preferentially removes nondiamond carbon.
Monocrystalline step-flow growth typically uses CH₄:H₂ ratios of 0.005% to 0.025%, with reactor-specific temperature and pressure. The (001) surface supports fourfold step propagation along ⟨110⟩. The (111) surface favors triangular terraces and can align NV axes perpendicular to the surface. Nitrogen incorporation is often higher on (111), which can help NV ensemble alignment. Commercial miscut tolerances of ±3° can exceed the narrower windows reported for defect-quality growth.
Delta-doping confines a dopant to a narrow epitaxial layer without implantation damage, but it still requires later vacancy creation. Shallow implantation at ≤ 20 keV confines damage to the upper few nanometers. Epitaxial overgrowth then sets the final emitter depth and repairs some implantation damage. FIB beam diameters of 20 nm to 100 nm support site-selective implantation, including single-ion counting.
Annealing and hydrogen removal control activation
Section titled “Annealing and hydrogen removal control activation”The authors’ own experiment isolates hydrogen degassing before electron irradiation. Two otherwise identical samples contained a 2 nm ¹⁵N delta-doped layer at 10 nm depth. One sample was degassed and the other was not. Both then received 850 °C activation for 2 h under similar vacuum at ≤ 1 × 10⁻⁶ Torr.
Residual-gas analysis detected H₂ desorption beginning near 500 °C during a 400 °C to 850 °C vacuum ramp. Pulsed optically detected magnetic resonance identified ¹⁵NV⁻ separately from substrate ¹⁴NV⁻. The degassed sample showed a clear increase in ¹⁵NV⁻ yield. Extracted hyperfine splitting changed from 2.85(5) MHz before degassing to 3.09(1) MHz afterward. The authors attribute the yield increase to reduced formation of NVH after hydrogen removal.
Annealing controls both repair and loss. Interstitial carbon becomes mobile near 400 °C, with an activation energy of about 1.68 eV. Neutral vacancies mobilize near 600 °C, with 2.2(2) eV. A 1,000 °C anneal reduces the vacancy amount by about 90%.
Small vacancy clusters dissociate in the 1,000 °C to 1,200 °C range, while some deep aggregates require ≥ 1,600 °C. NV dissociation is expected at ≥ 1,400 °C and SnV dissociation at ≥ 2,200 °C. Extreme budgets can etch the surface, contaminate tools, and require dopant depths ≥ 1 μm.
Interfaces and doping determine charge stability
Section titled “Interfaces and doping determine charge stability”Surface termination changes electron availability near the defect. Hydrogen termination has negative electron affinity and tends to favor neutral defect states in ambient conditions. Oxygen termination has positive electron affinity, suppresses surface conductivity, and helps preserve negative NV⁻ and G4V states. Some oxygen recipes leave subnanometer sp² regions that act as electrical traps, so termination must be verified for each workflow.
Boron is the most mature p-type dopant, but acceptor levels can drive NV⁻ toward NV⁰. Raising donor concentration can increase the NV⁻ fraction, yet excessive nitrogen also increases spin-bath noise. Planar gate structures have switched a single NV center through NV⁺, NV⁻, and NV⁰ on 10 ns to 100 ns timescales.
Contact design depends on termination. Pt/Au is nearly ohmic on hydrogen-terminated diamond, while Au, Zr, and Al Schottky contacts are common on oxidized surfaces. ALD caps can passivate dangling bonds and supply fixed charge, but a thick cap reduces near-field sensing sensitivity.
Open process questions guide device design
Section titled “Open process questions guide device design”The main unresolved issue is formation pathway control. Local vacancy creation beside substitutional dopants and long-range vacancy migration may both form XV centers. Their relative importance needs direct tracking before and after annealing. Surface reconstruction, subnanometer disorder, and electrically active states remain difficult to quantify with routine tools. This makes recipe transfer between laboratories difficult.
Templated growth adds mask-edge anisotropy and contamination risk from mask residues and redeposited species. Smart-cut membranes need temperatures about 50 °C to 100 °C lower than bulk homoepitaxy and have a narrower step-flow window. Surface reflectivity also prevents reliable comparison of in situ pyrometry between substrate types. Freshly formed NV, SiV, and GeV centers can recombine with mobile carbon interstitials, limiting practical yields to a few percent.
Material selection therefore couples orientation, defect depth, surface finish, thermal budget, and charge control. Sensing designs should minimize shallow-surface noise while preserving uniform ensembles. Photonic designs should favor buried, low-strain centers with narrow lines and reproducible orientation. Shared specifications for miscut, surface finish, and spin and optical benchmarks remain necessary for scale-up.
Materials and support
Section titled “Materials and support”6CCVD’s diamond materials catalog includes MPCVD-grown single-crystal plates and wafers up to 125 mm, with thickness from 0.1 to 500 micrometers. Single-crystal material can be polished to Ra below 1 nm, and the catalog lists Au, Pt, Pd, Ti, W, and Cu metallization. These are substrate and metal options for the paper’s (001) or (111) NV and G4V fabrication, where a custom specification can be requested through engineering support.
View Original Abstract
Advances in material processing are rapidly improving the quality and scalability of nitrogen-vacancy (NV) and group-IV vacancy (G4V) color centers in diamond—key building blocks for quantum sensing and photonic networks. Central challenges remain: precise control of defect-formation pathways and the mitigation of nearby charge traps and parasitic states, which is especially problematic for near-surface emitters. Recent progress in sample preparation and in-situ thermal treatments (before, during and after growth, and during implantation) have reduced nonradiative dark defects and suppressed interface doping. Optimizing thermal strategies have illustrated an increased conversion yield for NV and G4V centers, while limiting unwanted photoluminescence. Similarly, surface treatments play an imperative role in stabilizing near-surface charge states for sensing applications. Complementary ex-situ protocols, such as high-temperature vacuum anneals, and hybrid incorporation methods that combine shallow implantation with epitaxial overgrowth continue to improve yields and coherence for shallow NV and G4V centers. Together, these integrated strategies are enabling deterministic, high-fidelity quantum emitters embedded in scalable diamond nanostructures.
Electronic Grade Intrinsic Diamond Epitaxial Wafer H Terminated Superior 100 Mosfet Ready Substrate
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →
High Concentration Quantum Grade Single Crystal Diamond
View Specs →References we cover
Section titled “References we cover”- A coherence-protection scheme for quantum sensors based on ultra-shallow single nitrogen-vacancy centers in diamond
- Fabrication of oriented NV center arrays in diamond via femtosecond laser writing and reorientation
- Quantum biosensing on a multiplexed functionalized diamond microarray
- Enhancement of quantum coherence in solid-state qubits via interface engineering
- Sulfur in diamond and its effect on the creation of nitrogen-vacancy defect from ab initio simulations
Citing papers we cover
Section titled “Citing papers we cover”Original source
Section titled “Original source”References
Section titled “References”- 2005 - The control of growth parameters in the synthesis of high-quality single crystalline diamond by CVD [Crossref]
- 2019 - Ab initio theory of the nitrogen-vacancy center in diamond [Crossref]
- 2024 - Photonic-cavity-enhanced laser writing of color centers in diamond [Crossref]
- 2025 - Deterministic detection of single ion implantation [Crossref]
- 2019 - Multivariable study on growth of diamond on diamond substrates by microwave plasma chemical vapour deposition [Crossref]
- 1998 - The annealing of interstitial-related optical centres in type II natural and CVD diamond [Crossref]
- 2019 - Photoluminescence decomposition analysis: a technique to characterize n - v creation in diamond [Crossref]
- 2016 - Fabrication of all diamond scanning probes for nanoscale magnetometry [Crossref]
- 2021 - Diamond for electronics: materials, processing and devices [Crossref]