Impact of CVD-synthesis parameters and film thickness on growth rate of single-crystal diamond
At a Glance
Section titled “At a Glance”| Metadata | Details |
|---|---|
| Publication Date | 2021-02-01 |
| Journal | IOP Conference Series Materials Science and Engineering |
| Authors | A L Maslov, N. I. Polushin, А. I. Laptev, T. V. Martynova, A M Grechikhina |
| Institutions | National University of Science and Technology |
| Analysis | Full AI Review Included |
Executive Summary
Section titled “Executive Summary”- Objective: Systematic investigation of MPCVD single-crystal diamond (SCD) growth rate dependence on chamber pressure, microwave power, deposition time, and accumulated film thickness.
- Reactor Platform: Ardis-300 (2.45 GHz-class industrial MPCVD system); methane-to-hydrogen ratio held constant at 3:97.
- Key Finding 1 — Pressure Window: Growth rate peaked at ~34.7 kPa with 4000 W microwave power, reaching 26-27 µm/h, but produced defective polycrystalline-like surfaces.
- Key Finding 2 — Quality-Regime Optimum: Lower pressure (~29.3-30.7 kPa) yielded defect-free (100) epitaxial layers at 7.6-24.7 µm/h, demonstrating a clear quality-throughput tradeoff.
- Key Finding 3 — Self-Accelerating Growth: Across five sequential regrowth cycles, growth rate monotonically increased from 15.9 µm/h → 24.7 µm/h (+55%) due to progressive reduction in (100) misorientation and surface roughening.
- Substrate Quality Threshold: Starting substrates with roughness 1.0-2.5 nm and misorientation 1-2° were sufficient for high-quality homoepitaxy.
- Engineering Implication: Cycle-based regrowth is a viable strategy to scale SCD thickness beyond the conventional single-run ceiling while maintaining crystalline quality.
Technical Specifications
Section titled “Technical Specifications”| Parameter | Value | Unit | Context |
|---|---|---|---|
| Substrate Type | Single-crystal CVD diamond (Bhapsali Diamond Tools) | — | Homoepitaxial (100) growth |
| Initial Substrate Roughness | 1.0 - 2.5 | nm | Measured via AFM (MFP 3D) |
| Initial Misorientation (100)/[100] | 1 - 2 | ° | High-resolution XRD (D8 DISCOVER) |
| CH4 : H2 Ratio | 3 : 97 | — | Constant across all runs |
| CH4 Purity | 99.9999 | % | 6N gas |
| H2 Purity | 99.99995 | % | 6N5 gas |
| Substrate Temperature | 1100 | °C | Optical/IR pyrometer |
| Microwave Power Range | 3800 - 4000 | W | 2.45 GHz cavity |
| Pressure Range Studied | 29.3 - 40.0 | kPa | Full optimization sweep |
| Optimal Pressure (Quality) | 29.3 - 30.7 | kPa | Defect-free epitaxy |
| Optimal Pressure (Rate) | 34.7 | kPa | Max throughput, defective |
| Peak Growth Rate (Single Run) | 27 | µm/h | Sample 6, 34.7 kPa / 4000 W |
| Stable Quality Growth Rate | 7.1 - 10.0 | µm/h | 38.7-40.0 kPa range |
| Cyclic Growth Rate Range | 15.9 - 24.7 | µm/h | 5-cycle experiment |
| Max Film Thickness Achieved | 234 | µm | Single-cycle run |
| Cyclic Cumulative Thickness | ~702 | µm | Sum across 5 cycles |
| Film Thickness Measurement | Dial indicator (1 µm resolution) | — | Destructive cross-check |
| Surface Characterization | SEM (JEOL JSM-IT500) | — | Morphology / step analysis |
Key Methodologies
Section titled “Key Methodologies”- Substrate Pre-Qualification
- AFM roughness screening (Ra 1.0-2.5 nm acceptance window).
- High-resolution XRD misorientation mapping (θ ≤ 2°).
- Reactor Conditioning (Ardis-300)
- Bake-out and hydrogen plasma etch prior to deposition.
- Process gas: CH4/H2 = 3/97; ultra-high purity feed.
- Stage 1 — Pressure Optimization (Table 1)
- Fixed T = 1100 °C, target thickness ≈ 50 µm.
- Sweep at 40.0 / 38.7 kPa with 3800-4000 W.
- Stage 2 — Reduced Pressure Sweep (Table 2)
- 34.7 kPa and 29.3 kPa at 4000 W to map rate-quality boundary.
- Stage 3 — Cyclic Regrowth (Table 3)
- 5 sequential homoepitaxial cycles.
- Cycle 1: 34.7 kPa (8.5 h); Cycles 2-5: 30.7 kPa (5.75-7.5 h each).
- Continuous rate monitoring per cycle to track self-acceleration.
- Ex Situ Characterization
- Thickness: dial indicator (1 µm graduation).
- Morphology: SEM (JEOL JSM-IT500).
- Crystallinity: implicit from XRD substrate baseline + SEM step-feature imaging.
- Data Interpretation Framework
- Correlation of rate vs. accumulated thickness → attributed to misorientation reduction + surface roughening increasing active carbon-radical adsorption sites.
6CCVD Solutions & Capabilities
Section titled “6CCVD Solutions & Capabilities”The Ardis-300 dataset in this paper maps directly to the engineering decisions our customers face when specifying homoepitaxial SCD for quantum, photonic, and high-power RF applications.
Applicable Materials
- Optical Grade SCD (Type IIa): Required for any replication of this work. The (100)/1-2° misorientation window used by the authors is precisely the substrate class 6CCVD qualifies and ships, with FTIR and Raman pre-certification available upon request.
- Heavy Boron-Doped PCD (BDD): For applications where the 27 µm/h rate regime is acceptable and electrode/electrochemical functionality is required.
- Free-Standing SCD Membranes: For groups targeting the 100-700 µm cumulative thickness range demonstrated here, particularly for X-ray, photonics, or MEMS platforms.
Customization Potential
- Substrate Geometry: 6CCVD supplies Type IIa SCD substrates in custom cuts aligned to within <1° of (100) upon request, eliminating the rate-limiting misorientation term identified in this paper.
- Metalization: Although this study focused on bare growth, downstream devices (Schottky electrodes, coplanar waveguides, BDD electrochemical cells) require Ti/Pt/Au, Ti/W, or Pt ohmic stacks. 6CCVD offers in-house metallization (Au, Pt, Pd, Ti, W, Cu) with sub-micron photolithographic patterning and laser cutting for device-ready SCD/PCD wafers.
- Thickness Range: 6CCVD routinely delivers SCD from 0.1 µm to 500 µm and free-standing substrates up to 10 mm, covering both the thin-film (50 µm) and thick-plate (700 µm+) regimes explored here.
- Wafer Scale: Custom plates and wafers up to 125 mm diameter for PCD; inch-class polished SCD with Ra < 1 nm available — a finishing quality that mirrors or exceeds the 1.0-2.5 nm starting substrate spec in this work.
Engineering Support
- 6CCVD’s in-house PhD team can advise on:
- Selection of starting substrate misorientation grade for target growth rates.
- Pressure/power recipes translated between reactor platforms (Ardis-300, Seki, IPLAS).
- Tradeoff modeling between throughput (27 µm/h peak) and surface quality for downstream polishing budgets.
- Multi-cycle regrowth campaign planning for thick SCD plates where cumulative thickness >500 µm is required.
Call to Action
For custom specifications, multi-cycle regrowth planning, or material consultation on translating these results to your reactor platform, visit 6ccvd.com or contact our engineering team directly.
View Original Abstract
Abstract The widespread use of CVD-reactors of various constructions and the influence of many factors on the formation of energy-saturated reaction zone has led to the fact of significant difference of technological parameters optimal for each reactor even of one brand. An economically important parameter, in addition to the quality of the obtained single-crystal diamond layer, is the growth rate, which depends on many factors: the flow rate and chemical composition of gases, pressure in the reaction chamber, substrate temperature, the shape of the formed plasma cloud, dimensions and shape of molybdenum substrate holders and et al. This work is devoted to the experimental determination of the single-crystal CVD-diamonds growth rate across various synthesis conditions on Ardis 300. In research, we shown the effect of the thickness of the synthesized diamond layer on the rate of formation of subsequent diamond layers, which is associated with a change in the degree of misorientation between the crystallographic plane (100) and direction [100]. Also, this can be associated with an increase in the roughness of the growing surface, leading to an increase in the specific number of active centers thermodynamically favorable for carbon deposition during the formation of the crystal structure of diamond.