Here is a fact that surprises many first-time buyers: our "8 micron" grade has a median particle size of 2.10 µm. Most of its particles are far finer than 8 µm. So what does the name mean — and which number should you actually care about?
A grade is a distribution, not a size
Ground calcium carbonate is never one size. Every batch is a population of particles from sub-micron fines to the coarsest survivors of classification. A laser-diffraction analyzer measures that whole population and reports it as a cumulative curve: for any diameter, what percentage of the material is finer.
The D-values are simply points on that curve:
- D50 — the median. Half the material is finer. This drives average behavior: opacity contribution, oil demand, packing.
- D90 / D97 / D98 — the top end. Ninety, ninety-seven or ninety-eight percent finer. These describe the coarse tail — the particles that scratch films, block screen packs and show up as surface defects.
- Span — the width. A tighter span means a more uniform powder and more predictable rheology.
Why the top cut usually matters more
Two powders can share an identical D50 and behave completely differently, because problems in most applications come from the coarse tail, not the median. A thin PE film doesn't fail because of the average particle; it fails at the one 20 µm particle that spans a quarter of the film's thickness. Paint gloss doesn't drop because of the median; it drops where oversized particles roughen the film surface.
That's why our laboratory reports the full curve — D3 through D98 — for every grade, and why our specification tables always show D50, D90, D97 and D98 together:
| Grade | D50 | D90 | D97 | D98 |
|---|---|---|---|---|
| 5 µm | 1.52 | 3.46 | 4.10 | 5.25 |
| 8 µm | 2.10 | 5.94 | 8.05 | 8.71 |
| 10 µm | 2.56 | 7.12 | 9.61 | 10.18 |
| 12 µm | 2.78 | 8.62 | 11.69 | 12.58 |
| 15 µm | 3.05 | 9.99 | 13.94 | 15.14 |
| 20 µm | 3.38 | 12.50 | 18.26 | 20.06 |
Three questions to ask any supplier
- Is the fineness on your TDS a measured value from laser diffraction, or a nominal figure?
- Do you report the top cut (D97/D98), or only the flattering median?
- Is every batch tested, or only the occasional reference sample?
Our answers: measured, both, and every batch — the full comparison lives on our Technical Data page.