Boron Incorporation Efficiency in Diamond - Influence of Gas Composition, Crystal Orientation, and Substrate Temperature
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | C – Journal of Carbon Research |
| Published | 2026-08-27 |
| Authors | Vincent Mortet, Mahebub Alam, Patrik Straňák, K. D. Sung, Yan Busby |
| DOI | 10.3390/c12030067 |
Measured boron incorporation efficiency reached 11 in epitaxial diamond grown on a (111) substrate at 1% methane and a gas-phase boron-to-carbon ratio of 1000 ppm. The corresponding boron concentration was 1.7 × 10²¹ cm⁻³. Across the experiments, efficiency ranged from 0.01 at low methane concentration on (100) to values above unity on every tested orientation at high methane concentration. Methane concentration and crystal orientation therefore provided strong control over boron uptake. Efficiency, η, is the solid boron-to-carbon atomic ratio divided by the gas-phase ratio.
Growth and measurement method
Section titled “Growth and measurement method”Epitaxial layers were grown in a 1.5 kW resonance cavity microwave plasma-enhanced chemical vapor deposition system. HPHT Ib diamond substrates of about 10 mm² had (100), (115), (113), and (111) orientations. Reactor base pressure was below 2 × 10⁻⁶ mbar. Methane was diluted in hydrogen, and trimethylboron in hydrogen supplied boron. Substrate temperature during deposition was measured at about 1000 °C.
For the methane and orientation matrix, pressure was 100 mbar, microwave power was 580 W, and total flow was 500 sccm. Methane ranged from 0.05% to 1%, and gas-phase ratios were 200, 400, 700, and 1000 ppm. Each doped layer targeted about 200 nm, but measured thicknesses ranged from 51 to 975 nm. Undoped spacers separated doped layers. Nitrogen δ-doped layers marked changes in oxygen content or substrate position. Boron values were taken from each layer’s plateau away from the marker layers.
Oxygen and temperature studies used 100 mbar and 570 W. The (100) series used 0.4% methane and a 1000 ppm gas-phase ratio. The (111) series used 0.1% methane and 4000 ppm. These conditions compensated for the large orientation difference in incorporation.
Depth profiles came from magnetic sector SIMS and time-of-flight SIMS. The latter used a 2 keV O₂⁺ sputter beam and a 30 keV Bi⁺ primary beam for crater analysis. Boron was quantified from the ¹¹B⁺ to ¹²C⁺ signal ratio using a relative sensitivity factor from an implanted diamond standard and an isotope correction. The calculation assumed a carbon concentration of 1.76 × 10²¹ cm⁻³. Calibration uncertainty was below 3%, and the detection limit was approximately 10¹⁷ cm⁻³.
Methane concentration
Section titled “Methane concentration”Boron concentration rose quasi-linearly with methane concentration for all orientations and gas-phase ratios. The rise had a slight deviation at low methane concentration. Incorporation efficiency also increased with methane for every orientation, even though methane slightly raised the substrate temperature. The authors excluded temperature as the cause because efficiency decreased as temperature increased.
Efficiency fell as the gas-phase ratio increased at fixed methane concentration and orientation. This agrees with the authors’ prior review, which linked the decline at elevated ratios to boron saturation or shifts in incorporation kinetics. Reported efficiency-versus-methane slope values were approximately 130 to 170 for (100) and 790 to 960 for (111). The (111) response was therefore much stronger.
Crystal orientation
Section titled “Crystal orientation”At constant methane concentration and gas-phase ratio, efficiency increased from (100) through (115) and (113) to (111). Earlier studies showed parallel trends. Direct numerical comparison remains difficult because reactor design, gas composition, growth parameters, and substrate mounting differed. At comparable deposition rates, (100) layers had about one order of magnitude lower efficiency than (111) layers. The (115) data followed the (100) trend. The (113) data showed more scatter and lay between the two families.
The authors say tanθ/h can describe the rise from (100) to (113), where θ is the angle to the (100) normal and h is step height. Beyond (113), efficiency continued rising while nominal step density decreased. The highest efficiency on atomically flat (111) also conflicts with models based only on step or riser sites.
Efficiency increased with deposition rate for all orientations. The orientation difference remained at a given rate. The authors therefore argue that dominant incorporation mechanisms differ between (100)-like and (111)-like terraces or surfaces.
Oxygen addition
Section titled “Oxygen addition”Adding oxygen reduced boron incorporation on both (100) and (111) substrates. Across this work and the cited studies, efficiency fell by about one order of magnitude for each order-of-magnitude increase in gas-phase O/C. The present (100) and (111) trends were parallel despite different methane concentrations and gas-phase boron ratios.
Volpe et al. reported a decrease from about 2 × 10⁻² without oxygen to approximately 3.4 × 10⁻⁵ at O/C = 10. Sakaguchi et al. reported factors of 100 to 350 when comparing no oxygen with 0.25% added oxygen. Issaoui et al. observed an initial decline followed by an increase above 7% methane in hydrogen.
Oxygen is often assumed to reduce active boron species in plasma. The authors argue that plasma scavenging alone cannot explain the similar O/C response at different B/O ratios. They point to a surface contribution in which oxygen-containing species lower desorption energies and enable boron removal, similar to oxygen-assisted etching. This mechanism remains a proposed interpretation.
Deposition temperature
Section titled “Deposition temperature”Lower deposition temperature increased boron retention on both orientations in the controlled temperature series. The authors varied substrate position and calibrated temperature with a Williamson PRO 92-38 pyrometer. The (100) series used 0.4% methane and 1000 ppm gas-phase B/C. The (111) series used 0.1% methane and 4000 ppm. Both used 100 mbar and 570 W.
Arrhenius fitting for (100) gave R² = 0.983 and an apparent activation energy of 144 ± 11 meV. The high-temperature (111) fit gave R² = 0.926 and 355 ± 58 meV. The (111) point at the lowest temperature was excluded from the fit. Apparent saturation at low temperature remains unexplained.
The authors interpret these data as a thermally activated boron-loss process. The fitted energies agree with the 0.1 to 0.6 eV range reported for boron-loss mechanisms. Hydrogen abstraction under low-energy hydrogen irradiation has a similar energy scale. The boron trend contrasts with Demlow et al. and agrees with Kato et al.’s phosphorus trend on (100) diamond.
The authors suggest the (111) saturation may relate to its boron concentration of 10²⁰ cm⁻³, compared with no more than 3.3 × 10¹⁹ cm⁻³ for (100). This suggestion was not tested directly.
Process controls and open questions
Section titled “Process controls and open questions”Process control for epitaxial boron-doped diamond must include orientation, methane fraction, gas-phase boron ratio, oxygen content, and temperature. Within the tested series, higher methane increased uptake, (111) gave the highest efficiency, and oxygen reduced it. Lower temperature also retained more boron. These variables must be calibrated together for the selected substrate orientation and reactor conditions.
The study measured composition and growth behavior, but did not report carrier concentration, mobility, conductivity, or device performance. Those electrical properties also depend on impurities and defects, so incorporation efficiency cannot qualify a finished device.
Literature comparisons remain difficult when reactor and mounting conditions differ, and digitized values carry about 10% uncertainty. Each orientation and temperature condition therefore needs its own calibration.
Materials and support
Section titled “Materials and support”The relevant material family for the paper’s electronic, electrochemical, and quantum-device work is boron-doped single-crystal diamond. The 6CCVD diamond materials catalog lists boron-doped diamond and MPCVD-grown single-crystal and polycrystalline diamond, with plates and wafers up to 125 mm and single-crystal thickness from 0.1 to 500 micrometers. The paper’s HPHT Ib (100), (115), (113), and (111) substrates, target boron concentration, and layer thickness should be stated when requesting engineering support for device-specific discussion.
View Original Abstract
In this work, boron incorporation in diamond grown by microwave plasma-enhanced chemical vapor deposition is investigated as a function of methane concentration, boron precursor concentration, crystalline orientation, oxygen addition to the plasma, and deposition temperature. The boron incorporation efficiency, i.e., the ratio between the boron concentration in diamond and the gas-phase boron-to-carbon ratio, spans several orders of magnitude and can exceed unity for all investigated crystalline orientations. Overall results demonstrate that methane concentration and the substrate’s crystalline orientation are key factors governing boron incorporation. The temperature study reveals that boron incorporation is also governed by thermally activated boron-loss mechanisms. Finally, all results show that boron incorporation is not exclusively controlled by the plasma composition but also by surface growth mechanisms.
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”- Enhanced boron doping for diamond growth in cylindrical CVD reactors using a Faraday cage
- Experimental Insights into Impurity Incorporation in Chemical Vapor Deposition Doped Diamond