Home /GBD technology
Grain boundary diffusion for high-temperature NdFeB magnets.
GBD puts dysprosium and terbium exactly where coercivity is decided: along the grain boundaries. That means SH and UH class thermal performance with far less heavy rare earth content and no unnecessary loss of magnetic strength.
GBD decision panel
Same Hcj target. Less Dy/Tb exposure.
Heavy rare earth mass index
Conventional route vs GBD route
lower is better
Diffusion path
50-70%
less Dy/Tb
Typical heavy rare earth reduction versus conventional alloying.
No Br loss
same flux target
Dy/Tb stays at the grain boundary instead of diluting the core phase.
2023
radial rings
GBD extended to multipole ring geometries for robotics motors.
H / SH / UH
high-temp grades
Best fit for motor magnets that need coercivity margin at heat.
The buyer problem
High-temperature magnets should not waste heavy rare earths.
Conventional high-Dy alloying works, but it spends expensive material throughout the whole magnet. GBD targets the boundary where demagnetization starts, so purchasing gets lower exposure and engineering keeps the thermal margin.
Where Dy/Tb goes
Conventional alloying
Mixed through the entire magnet blank.
Mainrich GBD route
Diffused along grain boundaries only.
Material use
Conventional alloying
Higher heavy rare earth content to reach SH / UH coercivity.
Mainrich GBD route
50-70% less Dy/Tb for the same thermal target.
Magnetic strength
Conventional alloying
Br can drop because the main magnetic phase is diluted.
Mainrich GBD route
Grain interiors stay cleaner, preserving Br and BHmax.
Supply exposure
Conventional alloying
More exposed to Dy/Tb price and export volatility.
Mainrich GBD route
Lower HREE mass per part, with the same licensing discipline.
| Decision point | Conventional alloying | Mainrich GBD route |
|---|---|---|
| Where Dy/Tb goes | Mixed through the entire magnet blank. | Diffused along grain boundaries only. |
| Material use | Higher heavy rare earth content to reach SH / UH coercivity. | 50-70% less Dy/Tb for the same thermal target. |
| Magnetic strength | Br can drop because the main magnetic phase is diluted. | Grain interiors stay cleaner, preserving Br and BHmax. |
| Supply exposure | More exposed to Dy/Tb price and export volatility. | Lower HREE mass per part, with the same licensing discipline. |
How it works
A surface process that changes the grain boundary, not the whole magnet.
Dy or Tb compound is applied to the surface and pulled inward through vacuum heat treatment. The atoms migrate along grain boundaries, strengthening the coercivity path while leaving the Nd2Fe14B grain interiors largely untouched.
01
Base NdFeB blank
Start with the target grade, geometry, and tolerance plan before diffusion.
02
Surface preparation
Clean and prepare the magnet surface so the Dy/Tb compound can bond evenly.
03
Dy/Tb application
Apply a controlled surface layer rather than alloying the full magnet body.
04
Vacuum diffusion
Heat treatment drives atoms inward along the grain boundary network.
05
Finish and verify
Final coating, BH curve confirmation, dimensional checks, and batch records.
Why it matters
The commercial argument is as important as the physics.
GBD is useful when the program already needs high-coercivity material. It reduces the cost and volatility burden without asking the motor team to accept a weaker magnet.
Thermal margin without overbuying Dy
For SH and UH grades, GBD gives the coercivity reserve motor programs need while reducing exposure to volatile heavy rare earth inputs.
Complex motor geometry support
The process now covers blocks, arcs, tiles, and radial multipole rings, so robotics and servo designs are not forced back to simple shapes.
Cleaner buyer documentation
Quotes can separate grade, diffusion route, coating, export licensing, and QC evidence so engineering and purchasing see the same risk picture.
Production fit
Best for hot motors, compact packages, and qualification-heavy programs.
GBD does not matter for every magnet. It matters when the design needs H, SH, UH, or EH class coercivity and the project cannot absorb avoidable Dy/Tb exposure.
2015
diffusion R&D began
3+ shapes
blocks, arcs, rings
Development timeline
From diffusion trials to radial multipole rings.
The important milestone is not just making GBD work on a block. It is applying it to motor geometries that buyers actually need to qualify.
2015
R&D started
Initial grain boundary diffusion development.
2018
Blocks in production
GBD-treated square and rectangular magnets.
2020
Arcs and tiles
Extended diffusion route to motor magnet geometries.
2023
Radial rings
Multipole ring production for compact robotics motors.
Request for quote
Ask for the GBD and conventional comparison.
Send the drawing, grade target, operating temperature, coating requirement, and destination. We can quote both routes with the cost, lead-time, QC, and export documentation separated clearly.