In thermal work diamond is a filler, not an abrasive. Nothing gets cut, nothing wears. The job is to move heat across an interface, and that makes coating and particle packing more important than crystal sharpness.
A Different Specification Entirely
An abrasive buyer asks about cutting behaviour, toughness and size distribution in the cut. A thermal buyer asks about thermal conductivity, interfacial resistance and how much filler can be loaded before the composite becomes unprocessable.
Diamond has the highest thermal conductivity of any bulk material, which is why it appears in heat spreaders, thermal interface materials and metal matrix composites. The difficulty is never the diamond itself; it is the boundary between diamond and matrix. Phonons do not cross a bad interface.
Coating Is The Whole Game
Bare diamond does not wet copper or aluminium, and the resulting gap acts as a thermal barrier that cancels most of the benefit. A titanium coating solves this: it bonds to the diamond on one side and reacts with the metal matrix on the other, forming a continuous thermal path.
At 0.1% to 0.3% by weight the coating is thin enough not to dominate the composite, which matters because titanium conducts heat far worse than diamond. The goal is a complete, thin, continuous layer, not a thick one.
Copper-coated and tungsten-copper-coated diamond exist for higher-end applications. In tungsten-copper coating the weight gain is substantial, roughly 1 carat becoming 1.3 carats at standard thickness, because the coating is a structural layer that also adjusts thermal expansion.
Size Grading And Packing
Thermal composites rarely use a single size. A graded mix packs more efficiently, and higher packing means more conductive paths per unit volume.
| Application | Typical size | Reason |
|---|---|---|
| Thermal interface materials | 0.5 – 3 um, plus a coarser fraction | Thin bond line, high loading |
| Heat spreaders, diamond-Cu | 4 – 40 um | Balance of conductivity and processability |
| Diamond-SiC composites | 4 – 54 um | Expansion matching with the SiC matrix |
| 3D printing / MMC feedstock | 30/40 – 40/60 mesh | Flowability of the powder bed |
Sizes commonly specified in this area run 4 to 8, 8 to 16, 10 to 20, 30 to 40 and 36 to 54 micrometres, plus 80/100 and 200/230 mesh for coarser fills. Multi-modal blends such as 1, 10 and 40 micrometres together are common in TIM work.
Which Base Grade
Crystal type matters less than it does in abrasives, but it still matters. HDY monocrystalline is the default because it is available in tight PSD at reasonable cost and plates well. Detonation nano material such as HDPN appears in nano-enhanced composites where the filler must stay below a micron.
Ash content becomes significant here. Measured ash across high-strength grit ranges from 0.0002% to 0.0093%, and in a thermal composite the residue sits exactly at the interface you are trying to engineer. Specify the low end and ask for the analytical method.
The Interface Problems That Actually Limit Performance
- Incomplete coating. Patchy titanium means part of the surface is still a barrier. Average weight gain will not reveal this; a cross-section will.
- Coating too thick. A heavy layer improves bonding but adds a low-conductivity shell. Thin and complete beats thick and partial.
- Poor packing. A single-size filler leaves voids. Graded blends raise conductivity without changing the diamond.
- Processing damage. High-shear mixing can break the coating. Check the composite, not just the incoming powder.
Loading, Voids And Why More Is Not Always Better
Thermal composites improve with filler loading up to a point, then degrade. The reason is voids: at high loading the matrix cannot wet every particle, and an air gap conducts heat far worse than the polymer or metal it replaced. The practical ceiling depends on particle shape, coating and mixing method rather than on the diamond itself.
Three levers raise the useful ceiling:
- Graded blends. Mixing coarse and fine fractions lets small particles fill the gaps between large ones, which raises packing density before viscosity becomes unworkable.
- Coating compatibility. A coating the matrix can wet reduces void formation at the interface and allows higher loading at the same viscosity.
- Mixing energy. High-shear mixing helps dispersion but can break coating. The correct energy is the one that disperses without stripping.
A typical TIM route uses two or three fractions, for example around 1, 10 and 40 micrometres, and adjusts the ratio until viscosity hits the process limit. The conductivity gain from grading is often larger than the gain from switching to a more expensive diamond grade.
What To Measure On The Powder Versus The Part
| Property | Measured on powder | Measured on composite |
|---|---|---|
| Coating continuity | Cross-section, SEM | Interfacial resistance |
| Particle size | PSD per fraction | Void fraction, packing |
| Purity | Ash, ICP-OES | Not directly measurable |
| Thermal performance | Single-crystal conductivity | Bulk conductivity, expansion |
Buying on powder data alone is how projects end up with a composite that looks correct on paper and underperforms on the bench. The interface is the product.
Common Mistakes
- Buying abrasive-grade powder for a thermal application. The size may be right and the coating may be wrong, and the result is an interface that does not conduct.
- Specifying a single size for maximum loading. Single-size fillers peak at lower packing density than graded blends.
- Judging by ash alone. Low ash helps, but coating continuity matters more for conductivity.
How To Verify
Ask for coating weight gain, a cross-section image showing a continuous layer, PSD for each fraction in the blend, and an impurity report with the method stated. Then measure the composite, not the powder: thermal conductivity, interfacial resistance and, for heat spreaders, expansion behaviour.
Next Step
Products: HDY, T titanium-coated diamond, HDPN nano.
Applications: diamond composite thermal management, thermal interface materials, 3D printing and metal matrix composites.