Minerals

We started our first trial with sample ATR-1 ( and desired Planetary Mill PULVERISETTE 5 classic line. The present agglomerates (up to ~1 cm diameter and harder) required a pre crushing with 20 mm Ø grinding balls. Therefore, we started our trial with 130 g of ATR-1 sample and 25x 20 mm Ø grinding balls as a first step of comminution. After the first minute, sample started sticking already.

Sticking of samples is happening by dry grinding, when interacting forces of fine ground particles will become bigger as a particles own g-force. The particles does not fall to the bottom of a bowl anymore; they remain slightly attached to other surfaces (grinding balls, wall of the bowl or other particles). When a d50 < 20 µm is reached, dry ground samples will start to become compressed by the flying grinding balls. Such compressed clusters of sample also contain bigger particles which will not be comminuted anymore.

Because of sticking of dry ground sample, we added 230 grams of isopropyl alcohol (IPA) in order to maintain motor oil like slurry. We proceeded grinding in suspension for further nine minutes, so a total grinding time of 10 minutes was provided.

To avoid over pressure, we ground the sample in steps of one minute, followed by a programmed pausing time of two minutes (factor 1:2). The outside temperature of the bowl should be checked on a regular basis and a temperature above the boiling point of the used solvent should be avoided. Else, longer pausing times should be adjusted.

After ten minutes of total grinding time, the bowls filling level reached the maximum capacity of the used 500 ml grinding bowl. Because sample became too pasty, it was required to take out sample slurry. We removed 150 grams of the slurry (- 55 g of sample and -95 g of isopropyl alcohol), so only 75 g of sample was left inside the bowl. The slurry was replaced by 150 g of fresh isopropyl alcohol and the used grinding balls were replaced by 5 mm Ø grinding balls as a further step in order to reduce the particle size (the smaller the grinding balls, the finer the sample can be ground). We proceeded grinding until a total grinding time of 30 minutes was reached. At this time, we checked the particle size distribution by static light scattering (using a Laser Particle Sizer ANALYSETTE 22 NanoTec). We determined a d10 < 770 nm, d50 < 2,7 µm and d90 < 11,9 µm. Such a level of fineness allows using grinding balls of only one millimeter.

We used a test sieve of 1,4 mm to separate the 5 mm Ø grinding balls from sample and added 800 grams of 1 mm Ø grinding balls for fine grinding.

Further 120 g of isopropyl alcohol were added in order to maintain a motor oil like slurry again. After 40 minutes of total grinding time (10 minutes with 1 mm Ø zirconium oxide grinding balls), we checked particle size distribution again and found a d10 < 420 nm, d50 < 1,07 µm and d90 < 2,7 µm. This use to be a typical result which can be achieved with classic line planetary mills. Smaller grinding balls of 0,5 mm or 0,1 mm Ø grinding balls are usually taken only for premium line models.

We proceeded grinding with 1 mm grinding balls until a total grinding time of one hour was achieved. Now, the prepared sample to check the particle size distribution started flocculating. This is happening often, when particles are ground into sub micrometer range. We checked a few different methods of dispersing, but always a second population of particles up to 30 µm was determined.

We aborted the grinding trial after one hour and packed the sample best possible. It is required to check if a different kind of solvent, or a grinding additive can improve the grinding result. If the sample should not be ground in water, it can be helpful drying the sample to > 100 °C in order to remove crystalline water inside the mineral sample. With water, next to surfactants as grinding agent, also a shift of pH-value or adding of a proper salt can stabilize sub micrometer particles.

As second sample, we picked the DIR-1 marked sample material. Usually up to ~ 200 grams of sample can be ground in a 500 ml bowl. Because in result one, sample was taken out of the bowl again (only 75 grams of sample remained inside the bowl), we started this trial with 70 grams of DIR-1 sample.

We also used the Planetary Mill PULVERISETTE 5 premium line which can speed up grinding balls up to > 50x g-force. The classic line model is only capable to speed up grinding balls to > 22x g-force. We started our trial, using 100x 10 mm Ø grinding balls made of zirconium oxide.

For the beginning of the trial, we added about 120 grams of isopropyl alcohol. To maintain a motor oil like slurry. After 10 minutes of total grinding time, particle size distribution was determined with Laser Particle Sizer ANALYSETTE 22 NanoTec. We found a d50 < 7,5 µm and d90 < 21,2 µm already.

The grinding balls were exchanged against 0,5 mm grinding balls (800 grams) and further 70 grams of isopropyl alcohol were added to keep the viscosity motor oil like.

Now also the speed could be set to maximum (800 rpm) and pausing times were increased to factor 10x. After 15 minutes of total grinding time (5 min. with 0,5 mm grinding balls), further 50 g of isopropyl alcohol were added to dilute the higher viscosity. We proceeded grinding until 20 minutes total grinding time passed and found a d50 < 670 nm and d90 < 2,7 µm already.

Because the bowl was filled to its maximum already, we proceeded grinding until 40 minutes of total grinding time was reached. Now we observed a flocculation of sample inside the bowl with present setup. Static light scattering is no proper tool to determine the primary particles anymore. We checked several pre treating methods of dispersion, but always found a new population from 10-30 µm range which is constantly rising inside the measuring circuit.

We packed the ground sample best possible and brushed of dried sample powder from the used equipment afterwards. As mentioned in result one, a proper solvent for grinding need to be found, eventually an additive is required to keep sub micrometer particles stabilized. Also a pre treating to get rid of crystalline water can be beneficial to grind the sample as desired (see hints of result one).