Research
Two programmes, held to one standard of evidence.
This page separates what is established from what is reported, hypothesised or unknown. We have no Colour samples or measurements yet, and nothing here claims otherwise.
Programme 1
Copper-diamond thermal composites
Diamond conducts heat far better than copper, and copper is easier to shape. A composite aims to combine both, with thermal expansion tuned by the diamond fraction. The interface between the two decides the result.
| Label | Statement | Source |
|---|---|---|
| Established | Copper does not wet diamond, so the interface limits heat transfer. | Textbooks; interface reviews |
| Established | Carbide-forming elements (Ti, Cr, W, Mo, Zr, B) can bond the two, but a thick layer blocks heat. | Reviews, 2020 and 2025 |
| Reported | About 690 W/mK with an interface conductance of 45 MW/m2K, using W-coated 200 micron diamond at 1:1 volume. | Abyzov et al., Materials and Design, 2015 |
| Reported | 554 W/mK with only 93 MPa tensile strength in Ti-doped copper, showing the conductivity and strength trade-off. | University of Waikato study |
| Reported | About 800 W/mK and low thermal expansion in a commercial composite. | Element Six (manufacturer statement) |
| Hypothesis | A controlled interface and tuned expansion could reduce thermal strain in power-module baseplates. | Colour. Untested. |
| Unknown | Machining cost and behaviour at plate scale. | No data found |
Planned characterisation
| Property | Method | Reference standard | Status |
|---|---|---|---|
| Thermal diffusivity | Laser flash analysis | ASTM E1461 | Planned |
| Specific heat | Differential scanning calorimetry | ASTM E1269 | Planned |
| Density and porosity | Archimedes immersion | To be confirmed | Planned |
| Thermal expansion | Dilatometry | To be confirmed | Planned |
| Microstructure and interface | SEM with EDS, Raman spectroscopy | Not applicable | Planned |
Thermal conductivity equals thermal diffusivity times density times specific heat. Diffusivity alone does not establish conductivity.
Programme 2
High-temperature magnets with less heavy rare earth
Dysprosium and terbium keep sintered NdFeB magnets from weakening in hot motors, but they are scarce. Two techniques reduce the need: grain boundary diffusion and hydrogen processing of magnet scrap.
| Label | Statement | Source |
|---|---|---|
| Established | Grain boundary diffusion places heavy rare earths near grain surfaces, raising coercivity with a small loss of remanence. | Reviews, 2021 and 2025 |
| Reported | Hydrogen processing can recover about 90% of magnetic properties under low temperature and pressure with rare earth hydride additions. | Habibzadeh et al., ACS Omega, 2023 |
| Reported | The hydrogen scrap route is patented by the University of Birmingham, with HyProMag as exclusive licensee. | HyProMag regulatory filing, 2024 |
| Hypothesis | Recycled feed combined with diffusion could meet a high-temperature target with less heavy rare earth, repeatably. | Colour. Untested. |
| Unknown | Feedstock consistency across scrap sources, and freedom to operate. | No data found |
Publications
Open documents.
| Document | Type | Version |
|---|---|---|
| NdFeB recycling and heavy rare earth reduction | Literature review | 0.2.0 |
| Supply-secure high-temperature magnets | White paper | 0.2.0 |
| Primers: heat transport, interfaces, processing, measurement, qualification | Educational series | 0.2.0 |
| Research integrity policy | Policy | 0.2.0 |
| Specification-sheet templates | Templates | 0.2.0 |
Open questions
- What interface layer thickness maximises conductivity without hurting strength, and how repeatable is it?
- How well do coupon-scale measurements predict behaviour in a real power-module baseplate?
- Can grain boundary diffusion work on magnets made partly from recycled powder?
- How consistent is magnet scrap across sources, and where do patents constrain these routes?