Skip to content

Synergistic Control of Crystal Planes and Defects in CVD Single-Crystal Diamond - A Review of Growth Mechanisms and Frontier Applications

FieldValue
JournalCoatings
Published2026-02-08
AuthorsXu Li, Jiaying Wei, Jie Gao, Yan Wang, Yongqiang Ma
DOI10.3390/coatings16020218
Citations1

CVD single-crystal diamond performance depends strongly on crystal plane, defect population, and doping. The review identifies (100) as the mature route for large-area homoepitaxy, with a reported growth rate of 165 µm/h at 300 Torr under nitrogen-assisted processing. Its application rule is direct: use (100) for large homoepitaxial films and (111) for aligned color centers. Use (110) for fast N-type layers or termination-controlled quantum surfaces.

The review focuses on microwave plasma chemical vapor deposition because its electrodeless plasma offers stable control of crystallographic orientation and intrinsic defects. It covers (100), (111), (110), (113), and hexagonal diamond, but excludes fusion-reactor windows and radiation damage. Published evidence is linked through DFT, first-principles calculations, AFM, TEM/SAED, STM, Raman spectroscopy, EBSD, photoluminescence, SIMS, and electrical measurements. The authors organize these results around growth kinetics, defect control, doping, surface termination, and device function. No new reactor dataset or systematic search protocol is reported, so the article should be read as a comparative synthesis.

(100): process maturity and defect tradeoffs

Section titled “(100): process maturity and defect tradeoffs”

Calculations reported in the review lower the (100) surface energy from 9.08 J/m² to 4.83 J/m² through 2 × 1 dimer reconstruction. This stability supports step-flow homoepitaxy, although early CH₂ insertion kinetics limited growth to below 1 µm/h. At 300 Torr, nitrogen-assisted MPCVD reached 165 µm/h, and growth at 350 Torr produced single-crystal diamond up to 18 mm thick. A 3° to 4° miscut increased growth from 3 µm/h to 9 µm/h and promoted step flow over hillock formation. AFM found rough surfaces, pyramidal hillocks, and 400 to 1000 nm penetration twins at 810 °C, while 1000 °C enabled atomically flat growth. Nitrogen from 3.1 ppm to 560 ppm increased lattice constant and Vickers hardness, but fracture toughness and strength fell 68.87% and 49.84%. Boron near 2500 ppm was proposed for conductivity without loss of single crystallinity. At 5000 ppm, the growth rate increased but surface defects appeared, while 10000 ppm produced polycrystalline structure. For NV sensing, (100) distributes centers across four equivalent <111> axes at about 25% each, compared with more than 90% preferred alignment on (111).

(111): slower growth and aligned color centers

Section titled “(111): slower growth and aligned color centers”

The (111) face is the lowest surface energy diamond plane, but dense atomic packing slows hydrogen abstraction and carbon attachment. Under conditions that suppress twins, its growth rate is typically about one tenth of the (100) rate. The process parameter α is the ratio of [100] growth to [111] growth. Keeping α below 1.5 suppresses twin survival. The review’s process rules use temperatures above 900 °C and methane below 4% to limit twin propagation. Growth at 850 °C and 200 mbar with a methane/oxygen ratio of 3, corresponding to 0.3% methane, gave 0.26 nm roughness on 100 nm films. The same process produced 200 nm roughness after 300 µm, showing that thickness can break the narrow quality window. Mn at 1 × 10^16 cm⁻³ changed step edges from linear to sawtooth. Above that concentration, dislocations and {111} micro-twins appeared. Wafer-scale (111) heteroepitaxy produced an 11.3 mm diameter freestanding diamond with uniform orientation by EBSD. A dual-probe gradiometer detected signals near a 10 mA electric-vehicle busbar without magnetic shielding, using preferentially aligned NV centers. P1 spin-bath decoherence and residual strain remain unresolved, according to the review.

(110): fast growth and surface termination

Section titled “(110): fast growth and surface termination”

The (110) surface retains a flat 1 × 1 termination, while anisotropic bonds create groove sites for hydrocarbon adsorption. Under 900 °C, 140 Torr, and 4% to 6% methane, <110> growth reached 17.6 to 22 µm/h, about twice the (100) rate. Adding a 0.5% O2/H2 ratio lowered that rate to 16.6 µm/h while suppressing defect luminescence. Phosphorus-doped homoepitaxy reached 66 µm at 33 µm/h with Rq near 15 nm. SIMS and electrical data showed more than 50% of phosphorus on active substitutional sites. They also showed a donor activation energy of 0.56 eV and NA/ND below 20%. The main bulk defects are Σ3 twins, a/2<110> perfect dislocations, a/6<211> partial dislocations, and stacking faults. Heat treatment can convert twin structures into more stable perfect-dislocation networks, although their bulk effects remain insufficiently quantified. For shallow NV centers, calculations found no in-gap states or surface spin after fluorine termination. Defect levels and the zero-phonon line remained close to bulk values, but this remains a theoretical route. The review gives conflicting electron-affinity signs for fluorine, so device teams should verify the relevant surface model.

High-index (113) diamond retained a smooth epitaxial surface at 1050 °C with 5% methane, while growing at about 4 to 5 µm/h. About 73% of its NV centers aligned along <111> at 29.5° to the surface normal. The remaining 27% occupied two <111> directions at 58.5°. The review contrasts this with 97% preferred alignment on (111) and an even split between two <111> directions on (110). Boron incorporation and electrochemical kinetics were better on (113) than on (118) or (115). Hexagonal diamond is not a conventional CVD epitaxial substrate. Millimeter-scale bulk crystals were produced by high-pressure compression of graphite. Synchrotron XRD confirmed the (1010) reflection without cubic (111), and HRTEM confirmed ABAB stacking. The phase had a Vickers hardness near 110 GPa, but simulations show that cubic diamond remains thermodynamically stable. Shock intensity and loading direction determine whether AB stacking or cubic diamond forms.

Application selection and unresolved design work

Section titled “Application selection and unresolved design work”

Device design should couple orientation, doping, termination, and defect control before growth.

PlaneRepresentative measured resultMain control issueReview-supported application
(100)165 µm/h at 300 Torr with nitrogen assistance; 18 mm at 350 TorrDislocations, twins, and doping strainLarge homoepitaxial films, drift layers, and heat spreaders
(111)0.26 nm roughness on a 100 nm film at 850 °C and 200 mbarTwins, stacking faults, and interface hillocksAligned NV and SiV layers and efficient doping
(110)17.6 to 22 µm/h at 900 °C and 140 TorrΣ3 twins, grooved morphology, and terminationN-type epitaxy and shallow NV surfaces
(113)4 to 5 µm/h at 1050 °C with 5% methaneMacrosteps and rough morphologyDirectional NV alignment and photonic coupling

Application-oriented plane selection is not yet routinely embedded in power-device layouts, quantum-chip layouts, or photonic architectures. The proposed response combines atomic calculations, mesoscale strain and thermal models, reactor-scale plasma kinetics, and high-resolution characterization. Device teams should treat growth rate, defect density, and application performance as one selection problem.

For homoepitaxial and quantum-device work, 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal plates and wafers up to 125 mm, with single-crystal thickness from 0.1 to 500 µm and substrates up to 10 mm. Boron-doped diamond and polishing to Ra below 1 nm provide catalog options for the review’s doped layers and low-roughness substrates. The engineering support link can help define crystal orientation and fabrication requirements for a specific request, while global shipping is available.

View Original Abstract

Single-crystal diamond (SCD) demonstrates immense potential in high-power electronics, quantum information, and precision sensing due to its exceptional hardness, high thermal conductivity, wide bandgap, and superior chemical stability. Focusing on the crystallographic dependence of chemical vapor deposition (CVD) synthesis, this review systematically examines the growth mechanisms, defect characteristics, and application progress of typical low-index planes, specifically (100), (111), and (110). The (100) plane, leveraging stable step-flow growth modes and a mature process window, has established itself as the primary orientation for large-size, high-quality homoepitaxy. Conversely, while the (111) plane faces challenges regarding growth rate and the suppression of twins and stacking faults, it offers unique advantages for high-efficiency doping and the preferential alignment of quantum color centers, such as NV and SiV centers. The (110) plane, characterized by its anisotropic surface structure and high effective growth rate, shows significant potential for textured film preparation, N-type doping epitaxy, and quantum sensing based on surface termination control. Furthermore, this article outlines progress in high-index planes (e.g., (113)) and hexagonal diamonds (HDs), highlighting their possibilities for rapid thick-film deposition, directional color center regulation, and novel superhard/quantum material design. Finally, from an integrated “Material-Defect-Device” perspective, we identify current critical scientific and engineering challenges, providing a roadmap for the synergistic optimization of crystal plane selection, defect engineering, and device structure.

  1. 2021 - Quantum Information Processing and Precision Measurement Using a Levitated Nanodiamond [Crossref]
  2. 2022 - Precision fabrication of diamond micro-optic elements with hybrid SiO2/Cr etching masks [Crossref]
  3. 2022 - Recent advances of boron-doped diamond electrochemical sensors toward environmental applications [Crossref]
  4. 2019 - Comparative study on stability of boron doped diamond coated titanium and niobium electrodes [Crossref]
  5. 2025 - High-Pressure, High-Temperature Heteroepitaxial Growth of Large Single-Crystal Diamonds [Crossref]
  6. 2020 - Chemical vapour deposition diamond single crystals with nitrogen-vacancy centres: A review of material synthesis and technology for quantum sensing applications [Crossref]
  7. 2019 - Influence of substrate holder configurations on bias enhanced nucleation area for diamond heteroepitaxy: Toward wafer-scale single-crystalline diamond synthesis [Crossref]
  8. 2021 - High quality CVD single crystal diamonds grown on nanorods patterned diamond seed [Crossref]
  9. 1993 - Nucleation of diamond during hot filament chemical vapor deposition [Crossref]