Shape retention and properties of diamond/Cu composites prepared by BJ3DP combined with different Cu infiltration contents
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Additive Manufacturing Frontiers |
| Published | 2026-02-28 |
| Authors | Pengfei Gao, Mengmeng Ding, Hao Fu, Jian Sun, Xue Yang |
| DOI | 10.1016/j.amf.2026.200314 |
Forty volume percent Cu infiltration produced W-coated diamond/Cu composites with linear shrinkage below 1% in all three principal directions. Infiltration at 1250 °C yielded 96.68% relative density and 392 W·m⁻¹·K⁻¹ thermal conductivity. The dimensional result came from supplying Cu close to the green part’s calculated pore volume.
Process route and green-part properties
Section titled “Process route and green-part properties”The authors printed two rectangular green parts from tetrahedral W-coated diamond powder averaging about 100 μm. Spherical Cu powder served as the second feedstock, with a particle size range of 15 μm to 53 μm. Plasma spheroidization gave the Cu particles their rounded shape.
Ten volume percent Cu was mechanically mixed with W-coated diamond to create the composite feedstock. The comparison feedstock contained only W-coated diamond. BJ3DP used a 100 μm layer thickness and 50% binder saturation in an ExOne Innovent+.
After printing, curing, and depowdering, the parts were infiltrated with Cu in a tube furnace. The furnace held each assembly at 1250 °C for 30 min. Rectangular Cu blocks with purity above 99.99% sat below the green parts during infiltration.
The 40 vol% condition used the green part containing 10 vol% Cu. The 50 vol% condition used the W-coated diamond green part without Cu. Both routes were designed for a nominal W-coated diamond/50 vol% Cu composite.
The composite green part had 58.31% green density and 1.01 MPa green strength. The W-coated diamond green part had 57.72% density and 0.86 MPa strength. The authors relate the density difference to Cu’s higher density and better flow. Finer Cu particles also filled voids between the larger diamond particles.
Matching copper to pore volume
Section titled “Matching copper to pore volume”The green-part geometry was measured in three dimensions after depowdering. Calculated pore volumes were 180.25 mm³ for W-coated diamond/10 vol% Cu and 183.41 mm³ for W-coated diamond. The selected Cu volumes were 168.07 mm³ and 250.39 mm³, respectively.
The 40 vol% condition supplied slightly less Cu than the calculated pore volume. The 50 vol% condition supplied substantially more Cu than its pore volume. The authors attribute infiltration of the porous network to capillary action.
When available Cu matched pore volume, dimensional change stayed below 1%. Excess Cu could not enter all pores and accumulated below the sample. Residual Cu and surface spillage increased the final dimensions, producing negative shrinkage.
The paper generalizes the matching rule to W-Cu and Mo-Cu through a volume-balance method. That method sets required infiltration volume approximately equal to green-part pore volume. The reported experiments do not validate those two composites.
Shape retention, density, and leakage
Section titled “Shape retention, density, and leakage”Shrinkage was measured along the X, Y, and Z axes after infiltration. The results separated the two infiltration conditions clearly.
| Measurement | 40 vol% infiltration | 50 vol% infiltration |
|---|---|---|
| X, Y, Z linear shrinkage | 0.52%, 0.48%, 0.67% | −1.99%, −1.75%, −22.59% |
| Relative density | 96.68% | 97.01% |
| Measured leak rate | 2.1 Pa·m³·s⁻¹ | 1.5 Pa·m³·s⁻¹ |
Negative values denote growth after infiltration, with the largest change along Z. The 40 vol% specimens retained flat upper and lower surfaces and good side surfaces. At 50 vol%, yellow Cu exudation appeared on top, and residual Cu roughened the lower surface.
Relative density was measured by Archimedes drainage. Leak rates were measured with a helium mass spectrometer. Thus, 50 vol% infiltration produced slightly higher density and a lower leak rate, but damaged dimensional control.
Microstructure and interface chemistry
Section titled “Microstructure and interface chemistry”SEM-EDS examined microstructure, fracture morphology, and elemental distribution. SEM imaging showed a nearly pore-free interior in both composites. W-coated diamond particles were uniformly distributed through the Cu matrix without detectable agglomeration.
At 1000× magnification, the authors observed continuous diamond-Cu contact without visible voids or gaps. These images support low interfacial porosity, but they do not measure interfacial thermal resistance.
XRD detected diamond, Cu, W, WC, and W₂C in both infiltrated samples. Confocal Raman spectra detected diamond but showed no graphite or amorphous carbon peaks. The authors therefore found no evidence of W-coated diamond graphitization during infiltration.
They suggest that W reacted with surface carbon to form WC and W₂C. These phases were proposed to improve Cu wettability and strengthen the diamond-Cu interface. Residual W peaks indicate that some coating remained after the reaction. That observation matters because a thick W coating can raise interfacial thermal resistance.
Thermal and electrical performance
Section titled “Thermal and electrical performance”Thermal diffusivity came from a laser thermal conductivity meter. Thermal conductivity was calculated from diffusivity, measured density, and specific heat capacity. Thermomechanical analysis measured thermal expansion, and a resistance tester supplied the resistivity used for IACS.
| Property | 40 vol% infiltration | 50 vol% infiltration |
|---|---|---|
| Thermal conductivity | 392 W·m⁻¹·K⁻¹ | 403 W·m⁻¹·K⁻¹ |
| Thermal expansion coefficient | 8.18 × 10⁻⁶ K⁻¹ | 7.36 × 10⁻⁶ K⁻¹ |
| Electrical conductivity | 32.3% IACS | 28.3% IACS |
The 40 vol% result was 49.78% of the 787.5 W·m⁻¹·K⁻¹ theoretical value. The measured conductivities were about 50% to 51% of theory. The authors link the shortfall to residual internal porosity and an excessively thick W coating.
Reference thermal expansion values were 17.5 × 10⁻⁶ K⁻¹ for Cu and 2.3 × 10⁻⁶ K⁻¹ for diamond. W has a thermal conductivity of 171 W·m⁻¹·K⁻¹. The authors note that this value is much lower than Cu and diamond.
The authors attribute the lower thermal expansion and electrical conductivity at 50 vol% to less Cu inside the sample. They state that actual Cu and diamond fractions differed from the nominal 50% and 50% values. Because diamond does not conduct electricity, the lower Cu fraction reduced relative electrical conductivity. Cu’s higher expansion coefficient also explains why reducing its fraction lowered the measured expansion.
Comparison with earlier binder jet results
Section titled “Comparison with earlier binder jet results”The authors’ earlier diamond/Cu study reported sintering shrinkage above 10% in length, width, and height. The combined BJ3DP and infiltration route reduced each directional shrinkage below 1% in the reported experiments.
The paper cites BJ3DP W-40 wt% Cu with 97.04% density, 228 W·m⁻¹·K⁻¹ thermal conductivity, and 46.5% IACS. That prior composite used a 50 μm layer and sintering at 1500 °C, rather than melt infiltration. Those results are not a direct composition or process match.
Here, the 50 vol% infiltration gave the highest thermal conductivity but failed shape retention. By contrast, pore-matched 40 vol% infiltration gave slightly lower thermal conductivity with much better dimensional control. The density difference was small, but the shrinkage difference was decisive. Process selection therefore depends on whether dimensional fidelity or maximum measured conductivity dominates.
Limits and device engineering implications
Section titled “Limits and device engineering implications”The 40 vol% composite retained its shape, but thermal conductivity remained below half the theoretical value. Residual W, coating thickness, and incomplete internal densification remain visible performance constraints. The experiments include W-coated diamond but no uncoated-diamond control.
The paper reports no post-infiltration hardness, fracture strength, or interfacial thermal resistance. The demonstration used rectangular green parts about 12 mm by 12 mm by 3 mm. It does not establish dimensional retention for larger parts or different geometries.
The route is useful when machining of the final geometry would be difficult. For an electronic heat-spreader design, the authors propose matching infiltration Cu volume to measured or calculated pore volume. Engineers must account for green density and binder saturation when setting that balance.
Reducing W coating thickness could raise thermal conductivity, although the paper did not test a thinner coating. For electrical and expansion control, nominal infiltration content should not be treated as the measured internal Cu fraction.
The demonstrated process combines binder jetting, 1250 °C melt infiltration, and a 30 min hold. A device requiring stable external dimensions can adopt pore-matched Cu infiltration, with the paper’s demonstrated shrinkage below 1%.
View Original Abstract
• High-shape-retention Dia/Cu composite was obtained via BJ3DP combined with infiltration. • The shrinkage of the sample obtained by infiltrating 40 vol.% copper was within 1%. • The TC of the sample obtained by infiltrating 40 vol.% copper was 392 W·m -1 ·K -1 . Diamond/Cu (Dia/Cu) composites have emerged as a focal point in research on thermally conductive materials owing to their exceptional thermal conductivity, low coefficient of thermal expansion, and outstanding corrosion resistance. Considering the difficult machinability of diamond, binder jet 3D printing (BJ3DP) overcomes the machining limitations of conventional processes and constitutes an emerging near-net-shaping technology. In this study, to further improve the shape accuracy of printed parts, W-coated Dia/Cu composites were prepared using BJ3DP combined with melt-infiltration technology to explore the feasibility of using BJ3DP for preparing highly dense and low-shrinkage samples. The effects of different Cu melt-infiltration contents on the shrinkage and properties were investigated, and the optimum Cu melt-infiltration content was determined. The results showed that the shrinkage of the samples obtained by melt infiltration with 40 vol% Cu was within 1% in all directions, with better shape retention and a higher accuracy of the external dimensions. The relative density, thermal conductivity, coefficient of thermal expansion, and electrical conductivity of the samples measured by infiltrating with 40 vol% Cu were 96.68%, 392 W·m -1 ·K -1 , 8.18 × 10 -6 K -1 , and 32.3% IACS, respectively. This study demonstrates a fabrication route for low-shrinkage and high-shape-retention Dia/Cu composites by combining BJ3DP and melt infiltration. This strategy provides a simplified and effective method for achieving high-dimensional accuracy in other composites.
Diamond Heat Spreader With Copper Coated Metalized Surface
View Specs →
Advanced Thermal Simulation Service Diamond Heat Spreader Performance Analysis
View Specs →
Electronic Grade Intrinsic Diamond Epitaxial Wafer H Terminated Superior 100 Mosfet Ready Substrate
View Specs →References we cover
Section titled “References we cover”- Progress in the Copper-Based Diamond Composites for Thermal Conductivity Applications
- Lightweight diamond/Cu interface tuning for outstanding heat conduction
Original source
Section titled “Original source”References
Section titled “References”- 2025 - Surface-metallized diamond/liquid metal composites through diamond size engineering as high-performance thermal interface materials
- 2024 - Research progress, potentials, and challenges of Cu composite for metal injection moulding feedstock [Crossref]
- 2025 - Enhanced thermal conductivity and thermal shock resistance in diamond/Cu composites through diamond surface etching [Crossref]
- 2021 - Binder jetting additive manufacturing of Cu/diamond composites: an experimental study [Crossref]
- 2012 - Selective interfacial bonding and thermal conductivity of diamond/Cu-alloy composites prepared by HPHT technique [Crossref]
- 2018 - High thermal conductivity of diamond/Cu composites produced with Cu-ZrC double-layer coated diamond particles [Crossref]
- 2015 - Effects of dual-layer coatings on microstructure and thermal conductivity of diamond/Cu composites prepared by vacuum hot pressing [Crossref]
- 2010 - Thermal conductivity and thermal expansion coefficient of diamond/5wt%Si-Cu composite by vacuum hot pressing [Crossref]
- 2015 - Diamond-W based coating-Cu composites with high thermal conductivity produced by pulse plasma sintering [Crossref]
- 2010 - Thermal conductivity of SPS consolidated Cu/diamond composites with Cr-coated diamond particles [Crossref]