Effects of Annealing Temperature on Interfacial Structure and Thermal Conductivity of Hot-Pressed Copper/Cr-Coated Diamond Composites
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Materials |
| Published | 2026-04-11 |
| Authors | Yajing Liu, Xi Chen, Yong Liu, Wei Tian, Fanfan Zhou |
| DOI | 10.3390/ma19081534 |
Annealing a sputtered chromium layer on diamond for 2 h at 900 °C produced the paper’s highest measured thermal conductivity, 605.27 ± 18.16 W·m⁻¹·K⁻¹. The value was 111% above the uncoated composite and equaled 211% of its thermal conductivity. The 900 °C condition produced a 77.1 wt.% Cr3C2 interlayer with more than 98% carbide coverage, a total coating thickness of 368 ± 51 nm, and Ra of 33.3 nm. The authors attribute the result to a balance between chemical bonding, phonon matching, and interlayer resistance.
Processing route
Section titled “Processing route”MBD6-type diamond particles, 60 to 70 mesh, were ultrasonically cleaned in acetone and ethanol for 30 min each and rinsed to neutral pH. Chromium was deposited by magnetron sputtering for 10 min under 99.99% Ar at a chamber vacuum of 1 × 10⁻³ Pa. The process used 450 W Cr power, a 70 V substrate bias, dual targets, and substrate rotation at 4 rpm. Coated particles were annealed for 2 h under Ar/H2 at 700, 900, or 1100 °C. The furnace ramp was 10 °C/min, followed by furnace cooling to room temperature.
Annealed particles received an electroless copper coating at pH 13 to 14. CuSO4·5H2O supplied copper and HCHO was the reducing agent. NaOH adjusted pH, Na2EDTA·2H2O complexed copper ions, and C10H8N2 stabilized the bath. The plated particles were mixed with 70 μm and 20 μm copper powders in ethanol for 3 h. The copper powders had a 2:1 mass ratio, and the diamond content was 50 vol.%.
The mixtures were hot pressed under vacuum at 850 °C and 25 kN for 30 min. The heating rate was 10 °C/min, and the final samples were cylinders measuring Ø12.7 mm × 3 mm. XRD and SEM examined phases and morphology, XPS measured chemical bonding, and AFM measured roughness over a 5 × 5 μm area at 0.9 Hz.
Thermal diffusivity was measured at 25 °C by laser flash analysis with an accuracy of ±3%. Density was measured by Archimedes water immersion. Specific heat was calculated by the rule of mixtures, and thermal conductivity was derived from λ = αρCₚ. Each sample was measured three times.
Phase and coating structure
Section titled “Phase and coating structure”EDS detected mainly Cr before annealing, with almost no surface carbon signal. As annealing temperature increased, the detected carbon content rose while the chromium content fell. The authors interpret this as progressive conversion of metallic Cr into Cr3C2 through interfacial diffusion and chemical reaction.
XRD confirmed metallic Cr and Cr3C2 in every annealed coating. Weak Cr3C2 peaks appeared after 700 °C treatment. Increasing the temperature weakened the metallic Cr peaks and strengthened the carbide peaks. At 1100 °C, the coating was carbide dominated, although some metallic Cr remained.
The 900 °C coating had a total thickness of 368 ± 51 nm. It comprised a 288 ± 56 nm Cr3C2 layer and an 80 ± 5 nm Cr layer. The 700 °C coating was thinner at 194 ± 32 nm, while the 1100 °C coating reached 746 ± 94 nm through continued carbide growth.
Crystal orientation also changed. The texture coefficient for Cr (200) was 1.22 after annealing at 900 °C and fell to 0.70 after 1100 °C. The Cr3C2 (150) coefficient increased from 2.10 to 2.26 over the same temperature change. This indicates progressive loss of the deposited Cr texture and stronger preferential orientation in the carbide.
Interface morphology and bonding
Section titled “Interface morphology and bonding”The 700 °C coating formed thin, non-uniform carbide layers on the diamond {100} and {111} facets. Its interface remained discontinuous, with residual porosity after hot pressing. At 900 °C, carbide coverage exceeded 98% on both facets and formed a continuous structure. The {100} surface developed fine-grained features, while the {111} surface contained larger, irregular plate-like carbides.
The 1100 °C coating developed coarse carbide grains and a much thicker layer. Gaps appeared at the Cu/diamond interface, and less copper adhered to the diamond surface. The authors associate this structure with accelerated grain growth at high temperature.
Surface roughness also varied non-monotonically. The pristine diamond had Ra of 1.58 nm. Annealed coatings had Ra values of 45.6 nm at 700 °C, 33.3 nm at 900 °C, and 61.6 nm at 1100 °C. The 900 °C condition therefore combined the lowest coated roughness with continuous carbide coverage.
Electroless copper plating improved the subsequent hot-press bond. The authors report copper fragments attached to the diamond, carbon and chromium infiltration into the copper matrix, and a thin Cu/C solid solution layer. Cr3C2 formation bridged the copper and diamond phases.
XPS detected Cr-O bonds and copper oxides on the 900 °C sample. The authors state that residual oxidation and incomplete reduction left the interface chemically heterogeneous.
Measured thermal performance
Section titled “Measured thermal performance”The thermal conductivity increased from the uncoated baseline at 700 °C, reached its maximum at 900 °C, and declined after annealing at 1100 °C.
| Sample | Coating thickness, nm | Phase content, wt.% | Ra, nm | Thermal conductivity, W·m⁻¹·K⁻¹ |
|---|---|---|---|---|
| Uncoated | Not coated | Not applicable | 1.58 | 286.57 ± 8.59 |
| 700 °C for 2 h | 194 ± 32 | 57.5 Cr3C2 and 42.5 Cr | 45.6 | 532.93 ± 15.99 |
| 900 °C for 2 h | 368 ± 51 | 77.1 Cr3C2 and 23.9 Cr | 33.3 | 605.27 ± 18.16 |
| 1100 °C for 2 h | 746 ± 94 | 87.2 Cr3C2 and 12.8 Cr | 61.6 | 437.98 ± 13.14 |
For the 900 °C sample, thermal diffusivity was 240.37 ± 7.21 mm²·s⁻¹ and density was 5.51 ± 0.13 g·cm⁻³. Its calculated specific heat was 0.457 ± 0.05 J·g⁻¹·K⁻¹.
Infrared imaging on a platform maintained at 100 °C showed that the 900 °C composite reached 100 °C within 12 s. It heated 12 s faster than the unmodified composite and 8 s faster than pure copper.
The authors compare the maximum conductivity with three recent coating results: 431 W·m⁻¹·K⁻¹ for a molten-salt Cr-carbide layer, 469 W·m⁻¹·K⁻¹ for a V-based coating, and 550 W·m⁻¹·K⁻¹ for a W-based coating. They report 605.27 W·m⁻¹·K⁻¹ as the highest value among Cr-based Cu/diamond composites under comparable hot-pressing conditions.
Model-based transport interpretation
Section titled “Model-based transport interpretation”The diffusion mismatch model calculated an interfacial thermal conductance of 689.72 MW·m⁻²·K⁻¹ for diamond/Cr3C2 and 399.51 MW·m⁻²·K⁻¹ for diamond/Cr. The copper-side comparison was reversed. Cu/Cr had a calculated conductance of 645.77 MW·m⁻²·K⁻¹, while Cu/Cr3C2 had 461.07 MW·m⁻²·K⁻¹.
These results describe two competing effects. Partial carbide formation improves phonon matching at the diamond side. Complete conversion would add a layer with an intrinsic thermal conductivity of 19 W·m⁻¹·K⁻¹, compared with 90 W·m⁻¹·K⁻¹ for Cr. A thick carbide layer therefore raises resistance on the copper side.
The model also places acoustic impedance between the copper and diamond values. The listed impedances rise from 26.61 for Cu to 33.34 for Cr, 36.69 for Cr3C2, and 44.72 for diamond, in units of 10⁶ kg·m⁻²·s⁻¹. This gradient supports phonon transmission through the mixed interlayer.
The Hasselman-Johnson analysis inferred a full-interface conductance of 13.56 MW·m⁻²·K⁻¹ for the 900 °C structure. The values were 6.73 MW·m⁻²·K⁻¹ at 700 °C and 3.27 MW·m⁻²·K⁻¹ at 1100 °C. The uncoated interface had an inferred conductance of 0.95 MW·m⁻²·K⁻¹. The 900 °C value was therefore 1327% above the uncoated value. These interface values are model outputs rather than direct measurements.
Limits and engineering implications
Section titled “Limits and engineering implications”The annealing study covered only three temperatures, each with a fixed holding time of 2 h. The 900 °C optimum therefore applies to the tested process window. It does not define a continuous optimum or establish how shorter annealing times affect the same structure.
Thermal conductivity was measured at 25 °C, while infrared testing used a 100 °C platform. The paper does not report performance across a wider temperature range or after thermal cycling. Specific heat was calculated rather than measured directly, and the diffusion mismatch model assumes diffuse phonon scattering.
The test specimens were 50 vol.% particle composites in small cylindrical geometry. The paper does not evaluate monolithic heat spreaders, bonded joints, or complete devices. Surface oxidation detected by XPS may also affect reproducibility.
For device thermal management, the result supports partial carbiding rather than complete carbide conversion. Carbide coverage, roughness, phase fraction, and layer thickness must be controlled together. For the tested hot-press route, annealing at 900 °C for 2 h produced the best measured balance, with a 368 ± 51 nm Cr/Cr3C2 coating and conductivity of 605.27 ± 18.16 W·m⁻¹·K⁻¹.
View Original Abstract
Efficient heat dissipation is crucial for semiconductor devices; however, conventional thermal management materials often cannot meet practical demands because of inadequate thermal conductivity and mismatched coefficients of thermal expansion with semiconductor materials. In this study, we develop a synergistic process integrating magnetron sputtering and annealing to fabricate a composition-controllable Cr/Cr3C2 composite interlayer on diamond surfaces. By regulating the annealing temperature from 700 to 1100 °C, three key parameters of the Cr/Cr3C2 composite interlayer can be tailored: the thickness varies from ~200 to 800 nm, the Cr/Cr3C2 fraction is adjustable, and the surface roughness ranges from 33.3 to 61.6 nm. In the current research, the sample that was annealed at 900 °C for 2 h exhibited the highest coating uniformity, with carbide coverage exceeding 98% and no discernible porosity. This optimized annealing process produces an interlayer with robust coverage, moderate thickness (~300 nm), and low surface roughness (Ra = 33.3 nm), thereby markedly enhancing interfacial bonding and thermal-transport performance. The resulting composite achieves a maximum thermal conductivity of 605.27 W·m−1·K−1, corresponding to 211% of the experimentally measured value for the uncoated sample. Analyses combining the diffusion mismatch model and experimentation indicate that the enhancement originates from improved phonon spectral matching and increased interfacial adhesion energy. This work provides processing guidance for precise interface engineering in high-thermal-conductivity diamond/copper composites.
Electronic Grade Intrinsic Diamond Epitaxial Wafer H Terminated Superior 100 Mosfet Ready Substrate
View Specs →
Diamond Heat Spreader With Copper Coated Metalized Surface
View Specs →
Advanced Thermal Simulation Service Diamond Heat Spreader Performance Analysis
View Specs →References we cover
Section titled “References we cover”- An Investigation on the Spark Plasma Sintering Diffusion Bonding of Diamond/Cu Composites with a Cr Interlayer
- Fabrication of high thermal conductivity Cu/diamond composites at ambient temperature and pressure
Original source
Section titled “Original source”References
Section titled “References”- 2018 - A Theoretical Review on Interfacial Thermal Transport at the Nanoscale [Crossref]
- 2023 - Thermal Management of Electronics from −40 °C to 150 °C Enabled by Hydrate Salt Composite with Efficient Thermal Energy Storage [Crossref]
- 2018 - Largely Enhanced Thermal Conductivity of Graphene/Copper Composites with Highly Aligned Graphene Network [Crossref]
- 2020 - Research Progress of Diamond/Copper Composites with High Thermal Conductivity [Crossref]
- 2010 - Ceramics and Ceramic Matrix Composites for Heat Exchangers in Advanced Thermal Systems—A Review [Crossref]
- 2018 - Optimized Thermal Conductivity of Diamond/Cu Composite Prepared with Tungsten-Copper-Coated Diamond Particles by Vacuum Sintering Technique [Crossref]
- 2018 - Regulating Interface Adhesion and Enhancing Thermal Conductivity of Diamond/Copper Composites by Ion Beam Bombardment and Following Surface Metallization Pretreatment [Crossref]
- 2013 - Alloy Development for Highly Conductive Thermal Management Materials Using Copper-Diamond Composites Fabricated by Field Assisted Sintering Technology [Crossref]