Calcium Carbonate

With a start size up to ~ 500 µm, a pre crushing would be required for comminution. With 20 mm diameter, the largest grinding balls will have a factor of 20 (compared to largest particles), we would be on the safe side.

We filled 25x 20 mm Ø grinding balls into our 500 ml bowl made of zirconium oxide and added 163 grams of the calcium carbonate sample on top. The second bowl position was equipped with a properly balanced counter weight.

The mill was programmed with 450 rpm speed and the sample was ground for five minutes in total. After this first five minutes of grinding, the sample was compressed hard into the radius of the bowl. This kind of sticking occurs when majority of sample was comminuted < 20-30 µm.

With this level of fineness, interacting forces will get bigger as a particles own g-force. The particles do not fall back on the bottom of a grinding vessel; they attached to all available surfaces. By longer dry grinding, those fine particles will get compressed by the grinding tools to clusters which also can contain bigger particles (typically up to 150 µm) which will not be ground any further. A longer dry grinding is not recommended when such a fineness is present.

We added 75 grams of isopropyl alcohol (IPA) to procced grinding in suspension. To avoid over pressure, we set up the mill with one minute of grinding time and one minute of pausing time. After five cyclones (10 minutes of total grinding time), we checked the present particle size distribution by static light scattering. Our Laser Particle Sizer ANALYSETTE 22 NanoTec found a d90 < 18,4 µm and d50 < 3,6 µm.

We recommend to exchange the grinding balls at this point. See result 2 where the grinding process was proceeded with 0,5 mm Ø grinding balls.

After pre crushing with 20 mm Ø grinding balls in result 1 (10 minutes with effectively only five minutes of grinding), we exchanged the grinding balls.

The 20 mm Ø grinding balls were picked out with a tweezers and 800 grams of 0,5 mm Ø grinding balls were added into the bowl. The mill was programmed for one minute of grinding time, followed by a pausing time of 15 minutes to be on the safe side.

After 20 cycles and 30 minutes of total grinding time, we checked the particle size distribution again. A d50 < 910 nm was determined. We added further 25 grams of isopropyl alcohol to maintain a motor oil like slurry and proceeded grinding. After further 15 minutes of grinding (45 minutes of total grinding time and almost doubling of fine grinding time), we checked the particle size distribution again. We found a d50 value < 760 nm with d90 < 1,7 µm.

Surely a longer grinding time will improve the grinding result, but as shown with this last step, the grinding result will only improve slightly. It might take several hours (or days) until the desired level of fineness would be achieved.

Also important is, that nanometer fine particles have a tendency to reagglomerate, even in slurry. It is required to find out which solvent and which additive have to be combined to stabilize such a fine suspension. Even so, it can be difficult to determine the true particles size by static light scattering. This technic detects flocculated particles as a single, large particle when the sample was not dispersed correctly. More suitable methods might be DSL or EM where.

We aborted our trial at this point and packed the sample best possible.