Titanium dioxide (TiO2)

Desired Variable Speed Rotor Mill PULVERISETTE 14 or desired Cross Beater Mill PULVERISETTE 16 require smaller start sizes than 25 mm. Also results of 100 µm < 30 µm cannot be achieved with both types of mills. Almost all samples use to start sticking, when a d50 < 20 µm was reached. Even pre-crushed sample would clog fine sieve rings (with P-14s), or would jam the ribs of a P-16 relatively fast.

Desired level of fineness could be reached relatively fast with our planetary mills. Equipment of steel would be possible to use, but white samples would be discolored rapidly with metal abrasion.

For our demonstration, we used zirconium oxide which shows a bright (almost whitish) abrasion. We used a 500 ml bowl made of zirconium oxide which was equipped with 8x 30 mm Ø grinding balls to pre-crush the largest sample agglomerates. With PULVERISETTE 5 with 4 bowl fasteners, up to ~ 1 kg of sample could be ground in one batch.

We added 255 grams of TiO2 sample into the bowl and ground the sample for 60 seconds at maximum speed (400 rpm). On the opposite grinding position, a properly tared bowl, filled with sand was placed. After one minute of dry grinding, the bulk volume of ground sample almost filled the 500 ml bowl to its maximum. We did not find any larger fragments of sample. The sample started sticking lightly (see lid on separate photo #2).

Because a fineness of < 30 µm is required, an exchange of grinding balls would be required (5 mm Ø grinding balls should be capable to reach the desired fineness).

For this – see result 02.

After pre-crushing in result 01, we exchanged the 30 mm Ø grinding balls against ~ 2000x 5 mm Ø ZrO2 grinding balls and proceeded dry grinding for 2 more minutes of dry grinding (a total grinding time of 3 minutes). The sample was sticking stronger to bowl and grinding balls. Within compressed clusters of ground sample, surely also larger particles could be embedded which would not be ground any further.

We added 135 grams of water and proceeded grinding for 2 more minutes in suspension (5 minutes of total grinding time). By optical microscopy, still a few fragments > 30 µm were spotted, while the majority of sample was already ground (or deagglomerated) into small micrometer range (< 3 µm).

Just to be on the safe side, we added further 60 grams of water and proceeded grinding for a total grinding time of 8 minutes. The sample was checked by optical microscopy again; no more particles > 30 µm were spotted. To be on the safe side, we proceeded grinding for a total grinding time of 15 minutes. Surely even longer grinding times could be applied. By static light scattering (using our Fritsch Analysette 22 NeXT), we determined a d90 < 960 nm already.

We separated the grinding balls by sieving and packed the sample best possible (flushed with fresh water). To minimize agglomeration for drying, we recommend using a rotary speed evaporator to remove the water. Also, other solvents (e.g. like isopropyl alcohol) could have been used for comminution as well.

For an intensive cleaning of the used equipment. We ground a proper portion of sand and water for 2 minutes at maximum speed. The equipment was flushed with fresh water and dried off afterwards.