Skip to content

Thermal characterization of polycrystalline diamond thin film heat spreaders grown on GaN HEMTs

MetadataDetails
Publication Date2017-07-24
JournalApplied Physics Letters
AuthorsYan Zhou, Rajesh Ramaneti, J. Anaya, Svetlana Korneychuk, Joff Derluyn
InstitutionsUniversity of Bristol, Hasselt University
Citations150
AnalysisFull AI Review Included

Since the PDF text extraction returned no content, I cannot fabricate specific data points (e.g., exact temperatures, voltages, or coherence times) from a paper I cannot read. Fabricating those numbers would violate the integrity standards required for technical documentation.

What I can do reliably is produce a structured template scaffolded against your catalog, with the fields left as explicit placeholders (<...>) so a human reviewer can drop in the verified values from the PDF once it’s available.


  • Objective: <State the paper’s primary goal — e.g., improving spin coherence of NV centers, scaling BDD electrodes, demonstrating a novel resonator geometry, etc.>
  • Material System: <SCD / PCD / BDD, plus any dopants — e.g., nitrogen-vacancy ensemble, boron-doped polycrystalline, isotopically engineered 12C>
  • Key Achievement: <Quantitative headline result — e.g., T2 > 1 ms, growth rate > 10 µm/h, Rs < 100 Ω/sq>
  • Methodology: <MPCVD reactor type, substrate, gas chemistry, temperature/pressure window>
  • Relevance to Industrial CVD Diamond: <Why the result matters for procurement or process engineers>
  • Limitations Noted by Authors: <e.g., substrate size, defect density, cost-per-wafer>

ParameterValueUnitContext / Notes
Substrate<e.g., Type Ib HPHT (100)>—Seed crystal orientation
Growth Method<e.g., MPCVD — 2.45 GHz / 915 MHz>—Reactor class
CH4 / H2 Ratio<e.g., 1 : 99>%Gas chemistry
Pressure<e.g., 200-300>TorrProcess window
Substrate Temperature<e.g., 900-1100>°CPyrometer reading
Growth Rate<e.g., 2-8>µm/hVertical [100]
Thickness Range<e.g., 0.1-500>µmFree-standing after lift-off
Crystal Quality (FWHM)<e.g., < 5>arcsecX-ray rocking curve (400)
Dopant / Defect<e.g., [N] < 1 ppb, [B] 1018-1020 cm-3>—Target metric
Surface Roughness (Ra)<e.g., < 1 nm (SCD), < 5 nm (PCD)>nmPost-polish
Coherence / Mobility Metric<e.g., T2 = __ ms, µ = __ cm2/V·s>—Application-specific
Electrode Metal Stack<e.g., Ti / Pt / Au>nmMetallization scheme
Operating Voltage / Current<e.g., __ V, __ mA>—Device-level
Thermal Conductivity<e.g., > 1800>W/m·KRoom temp

  1. Substrate Preparation: <e.g., HPHT Ib seeding, laser-cut seed array, ultrasonic cleaning in acetone/methanol>
  2. Reactor Configuration: <e.g., 2.45 GHz ASTeX-style, 6 kW max, water-cooled base>
  3. Gas Recipe: <List CH4, H2, O2 (if any), N2 (if NV doping), B(CH3)3 (if BDD) with flow rates in sccm>
  4. Process Parameters: <Pressure (Torr), temperature (°C), microwave power (W)>
  5. In-Situ Monitoring: <e.g., pyrometer, optical emission, laser reflectance>
  6. Post-Processing: <e.g., lift-off, lapping, CMP polishing, laser dicing, annealing>
  7. Characterization Suite: <e.g., Raman, PL, XRD, SIMS, Hall, EPR, coherence measurements>

Paper Requirement6CCVD CapabilityMatch Status
<e.g., High-purity SCD with low NV ensemble>Optical Grade SCD, [N] < 1 ppb available✅ Aligns
<e.g., Boron-doped electrode > 1020 cm-3>Heavy Boron-Doped PCD (BDD)✅ Aligns
<e.g., 50 mm × 50 mm free-standing plate>PCD plates up to 125 mm diameter✅ Aligns
<e.g., 200 µm free-standing membrane>SCD / PCD 0.1 µm - 500 µm✅ Aligns
<e.g., Ti/Pt/Au contact stack>Internal Au, Pt, Pd, Ti, W, Cu metallization✅ Aligns
<e.g., Sub-nm RMS for cavity coupling>Ra < 1 nm (SCD), < 5 nm (PCD)✅ Aligns
<e.g., 4-inch wafer for fab integration>Substrate thickness up to 10 mm + laser cutting✅ Aligns

Based on the generic template, the most likely 6CCVD SKU this research would map to is <Optical-Grade Single Crystal Diamond> or <Heavy Boron-Doped Polycrystalline Diamond (BDD)>, depending on whether the paper targets quantum / photonic or electrochemical / power-device applications. For NV / spin-coherence work, isotopically engineered 12C-enriched SCD should be considered; specify enrichment level (e.g., 99.999%) when requesting a quote.

  • Geometry: Square or rectangular free-standing plates up to 125 mm (PCD); custom seed orientations ([100], [110], [111]) for SCD.
  • Metallization: In-house Ti / Pt / Au, Ti / W, Cu, or full-stack recipes with controlled interface chemistry (carbide-forming vs. non-carbide-forming).
  • Laser Cutting & Dicing: Edge geometry, kerf width, and taper angle can be specified for device-ready chips.
  • Surface Engineering: CMP polishing to sub-nm Ra, or controlled as-grown roughness for adhesion-dependent applications.

6CCVD’s in-house PhD team can assist with material selection, defect-density trade-offs, and metallization stack design for programs that extend the results reported in the source paper. Typical support engagements include:

  • Doping concentration vs. target resistivity curves.
  • Thermal management calculations for high-power stacks.
  • Coherence-time or device-yield optimization loops over multiple growth lots.

> To replicate or extend the findings in this paper, request a custom quote with the parameters above filled in. For material consultation, visit 6ccvd.com or contact our engineering team directly.


Action required: Please re-upload the PDF (or paste the abstract + key figures / tables) and I will populate every &lt;placeholder&gt; with verified numbers, units, and context drawn directly from the source — no fabrication.

View Original Abstract

Polycrystalline diamond (PCD) was grown onto high-k dielectric passivated AlGaN/GaN-on-Si high electron mobility transistor (HEMT) structures, with film thicknesses ranging from 155 to 1000 nm. Transient thermoreflectance results were combined with device thermal simulations to investigate the heat spreading benefit of the diamond layer. The observed thermal conductivity (κDia) of PCD films is one-to-two orders of magnitude lower than that of bulk PCD and exhibits a strong layer thickness dependence, which is attributed to the grain size evolution. The films exhibit a weak temperature dependence of κDia in the measured 25-225 °C range. Device simulation using the experimental κDia and thermal boundary resistance values predicts at best a 15% reduction in peak temperature when the source-drain opening of a passivated AlGaN/GaN-on-Si HEMT is overgrown with PCD.