Chem Soc Rev

View Article Online

REVIEW ARTICLE View Journal | View Issue
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.

Process engineering with planetary ball mills†

Cite this: Chem. Soc. Rev., 2013,
42, 7660
Received 7th November 2012
DOI: 10.1039/c3cs35455e
www.rsc.org/csr
Christine Friederike Burmeister* and Arno Kwade
Planetary ball mills are well known and used for particle size reduction on laboratory and pilot scales
for decades while during the last few years the application of planetary ball mills has extended to
mechanochemical approaches. Processes inside planetary ball mills are complex and strongly depend on
the processed material and synthesis and, thus, the optimum milling conditions have to be assessed for
each individual system. The present review focuses on the insight into several parameters like properties
of grinding balls, the filling ratio or revolution speed. It gives examples of the aspects of grinding and
illustrates some general guidelines to follow for modelling processes in planetary ball mills in terms of
refinement, synthesis’ yield and contamination from wear. The amount of energy transferred from the
milling tools to the powder is significant and hardly measurable for processes in planetary ball mills.
Thus numerical simulations based on a discrete-element-method are used to describe the energy
transfer to give an adequate description of the process by correlation with experiments. The simulations
illustrate the effect of the geometry of planetary ball mills on the energy entry. In addition the imaging
of motion patterns inside a planetary ball mill from simulations and video recordings is shown.

Technische Universität Braunschweig, Institute for Particle Technology, Volkmaroder
Straße 5, 38104 Braunschweig, Germany. E-mail: c.burmeister@tu-braunschweig.de,
a.kwade@tu-braunschweig.de; Fax: +49 531391 9633; Tel: +49 531391 9643
† In dedication to the memory of Stefan Rosenkranz.
Introduction
Grinding is one of the most important unit operations due to
the need for size reduction in chemical, pharmaceutical and
food industries. Planetary ball mills are suitable for wet and
dry, fine and ultra-fine grinding of particles down to the
nanometre size range due to the high stress intensities reached

Christine Friederike Burmeister Prof. Dr-Ing. Arno Kwade studied

studied biological engineering at
the Technische Universität
Braunschweig and the Royal
Institute of Technology Stockholm
from 2005 to 2011 and completed
her diploma with her thesis ‘‘DEM
simulations of micromechanical
properties of modeled aggregates’’.
Her doctoral research under the
supervision of Prof. Arno Kwade is
focusing on simulations of stress
Christine Friederike conditions in ball mills and the
Burmeister correlation of simulation results
with mechanochemical processes
in ball mills.
mechanical engineering from
1986 to 1992 at the Technische
Universität Braunschweig and the
University of Waterloo. He
finished his PhD thesis on
‘‘Autogenous Grinding in Stirred
Media Mills’’ in 1996. Afterwards
he worked as a general manager
and an executive director before
he started his career as a full
professor for mechanical process
Arno Kwade engineering in 2005 at the
Institute for Particle Technology.
His main research focuses on experimental and numerical
investigations in the fields of grinding and dispersing, flow and
compression of powders as well as preparation of particulate
coatings. Main application fields of basic research are in battery
technology and pharmaceutics.

7660 Chem. Soc. Rev., 2013, 42, 7660--7667 This journal is c The Royal Society of Chemistry 2013
 View Article Online

Review Article Chem Soc Rev

from high accelerations.1 Other benefits in comparison to other

ball mill types like mixer mills or stirred media mills are for
instance the simple set up, the good cleanability and moderate
cost of the device.2 Due to their construction, especially the
complicated movement of the grinding chamber with high
centrifugal forces, today planetary ball mills are built only on
laboratory and pilot scales, which means with jar sizes up to
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.

1 litre. Typical applications of planetary ball mills are material

development and liberation of minerals and other materials for
analysis. The highly efficient mixing of materials in the grinding
Fig. 1 Scheme of planetary disks with movement in a normal and counter
chamber makes planetary ball mills also useful tools for direction; with pot height h; pot diameter dP; revolution radius R.
mechanochemical approaches. In a mechanical alloying (MA)
process a mixture of different powders is milled under material
transfer to a homogenous alloy. In mechanical milling uniform,
often stoichiometric composition powders are processed in ball
mills without material transfer. Thus no homogenization is
required.3,4
Mechanochemical processes in planetary ball mills have
been widely employed in industry to synthesize a variety of
commercially useful and scientifically interesting materials.
The advantage of using planetary ball mills instead of conventional Fig. 2 Scheme of a ball motion pattern in a single pot of a planetary ball mill –
chemical or processing routes is the reduction of the amount of (a) cascading, (b) cataracting, (c) rolling.

reagents for example through a highly efficient solvent-free route

to arylthioglycosides without the need for purification5–8 as well as
the reduction of reaction time and the gain of higher yield which
leads consequently to a more efficient use of materials and
energy.2,9 Also products can be processed that have improved
properties or are not available via conventional routes at all for
example new fluorite type f-Bi2VO2 or high quality sintered ceramic
bodies with average grain size in the nanometre range.10–13
Planetary ball mills
Planetary ball mills are used for sample processing, colloidal
grinding and material development due to their good reprodu-
cibility, safe handling and short processing times. Planetary ball
mills are known for more than a hundred years and normally
have 2 or 4 pots. The pots are attached to a disk that rotates
around a common central axis while the pots are simultaneously
rotating around their own axis. The high rotational speed of
both, the pots and the disk, leads to large impact energies of
milling balls inside the pots to achieve an effective grinding
performance. The materials under comminution, in MA processes
mainly powder particles, are located between surfaces and
crushed due to the impact and/or frictional forces that are caused
by collisions.2,4,9,14,15
The relationship between the velocity of the disk and the
pots is defined by the ratio k of the rotational speed nP of the
pot and the revolution speed nR of the disk:14,16,17
nP
k¼ (1)
nR
The ratio can be positive or negative depending on the moving
direction of the disk and pots. In Fig. 1 a planetary ball mill
set-up with a normal and counterdirection of the pots and the
disk is shown. As well as the rotation of the pot the powder
charge influences the ball motion inside. The powder particles
are trapped between the colliding balls and make the collision
less elastic so that the ball motion is slowed down and
the energy dissipation is reduced.18 The motion of the balls
influences the number of collisions, the ball velocity and
therefore the impact intensity of the balls which contributes
to the energy input to the powder charge.4 The balls show
complex behavior in the grinding chamber during processing
and the motion pattern varies depending on the operation
parameters. It changes from cascading to cataracting up to
centrifugation or rolling with increasing filling ratio and/or
increasing revolution speed.14,17,19–21
In the cascading regime (Fig. 2a) the milling balls are taken
along by the pot wall and unroll upon each other from the bulk
top to its base while in the cataracting regime (Fig. 2b) the balls
detach from the wall and impact with high intensity the bulk or
the opposite wall. Rolling or centrifuging balls (Fig. 2c) align to
the wall rotation with almost no relative velocity.20
Processing in planetary ball mills
Fine grinding, chemical reactions and mechanical alloying in
planetary ball mills have been investigated for a long time.
However, still challenges exist. For example several different
variables influence the success of each individual milling
process and so it is important to make a good parameter
assessment. The parameters to be considered regarding the
milling process are:15,22–24
 Revolution speed or rotational speed at a constant
speed ratio
 Milling time (=reaction time)
 Filling ratio of milling balls or the number of milling balls
at constant chamber size

This journal is c The Royal Society of Chemistry 2013 Chem. Soc. Rev., 2013, 42, 7660--7667 7661

Chem Soc Rev
 Filling ratio of grinding material or ball to powder ratio
These process parameters are apparently not completely
independent of each other and play an important role in the
refinement of particles and in the kinetics when optimized for
the best yield.6 On a local scale process parameter variation
leads to changes in the nature of impacts, the stress frequency
and the stress energy as well as the amount of powder trapped
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.
between colliding bodies.3
One of the most important parameters for synthesis and
refinement is the specific energy that is transferred to the
powder. Processing at a laboratory scale enables a relatively
tight control but for practical manufacturing methods large
scale processes and due to the limitations of planetary mills
other processing setups or milling devices are needed. The
specific energy is used to adjust the required milling parameters
for a certain material or synthesis under different operating
conditions. Also for comparison of different devices or set-ups
the specific energy has to be considered.25,26
The specific energy Ei can be obtained by relating the total
energy to the total mass of the powder as it is shown in the
following:27
Pn
j¼1 SEj DSFj SN  SE
Ei ¼ t¼ (2)
mP mP
The total amount of energy is determined by the frequency of
stress events DSFj at interval j, the average stress energy SEj at
interval j and the processing time t, which is equal to the
product of the total number of stress events SN and the average
stress energy SE. The stress energy SE describes the maximum
amount of energy which can be supplied to the product
particles at one collision event and can be calculated from
the relative impact velocity vj and the masses of the colliding
bodies m1 and m2 as given by eqn (3).20,27 The frequency and
energy are on their part affected by the milling parameters
mentioned above.
vj 2 m1 m2
SE ¼ (3)
2ðm1 þ m2 Þ
The energy is supplied to the particles due to collisions of the
balls which are caused by the movement of the disk and pots.
Thus the revolution speed is connected with the energy entry
and it influences yield as well as selectivity and can change the
formation of the powder as a result of temperature effects. It is
well understood in the case of particle refinement and inorganic
processing but deeper understanding is needed for mechano-
chemical approaches which are more complex and often differ
from each other.15,24
In most cases a minimum revolution speed is required to
reach a desired product yield or product quality. An increase in
revolution speed often leads to higher yields or better substrate
conversion at constant milling times as a result of the more
efficient mixing and the increasing number of collisions.2,22,28
For some reactions the speed has no influence above a particular
number of revolutions or no influence at all.24,29 The speed is
apparently limited by the design of the mill and also by
unwanted effects on the product. High speeds may induce
7662 Chem. Soc. Rev., 2013, 42, 7660--7667
View Article Online
Review Article
unwanted side reactions such as aldol reaction between ketones
and aldehydes15,30 or lead to the decomposition of supersaturated
solid solutions or metastable phases as it was found from
mechanochemical alloying of Al–Cu–Fe systems31,32 due to a
significant temperature increase.6 A high temperature increase
may be helpful when diffusion is required in powder alloying or
the reaction is promoted by high temperatures.6 On the other
hand a lower revolution speed leads to less energy consumption
and hence to lower cost and a more energy efficient process.24
The selection of the material of milling balls is determined
by the reaction itself in a way that it will not induce side
reactions or cause high abrasion and memory effects with
reconstitution of the original structure (9).33 Mostly the used
milling balls are made of tungsten carbide, steel, zirconia,
corundum or agate.6,8,22,30,34,35 Balls with large diameter or
high density can provide higher impact energy during the
collision event and, therefore, are applicable for reactions
where the effect of particle refinement will lead to higher
surface activity. Higher surface activity benefits for instance
in catalytic reactions or in the formation of thermodynamically
stable products.6,15 Higher densities do not automatically lead
to a higher energy transfer into the powder charge and thereby
a higher yield. High stress intensities also can have a damaging
effect on the product as the result of the energy dissipation in
heat.6,15
In some cases more ‘‘soft milling conditions’’ with low
ball-to-powder ratios and small ball sizes which intense the
frictional action are helpful to support amorphization and
metastable phase formation in alloying.6 Small balls are also
useful if no high intensity but more an efficient mixing and/or a
high number of collisions is required. In organic synthesis the
diameter of the milling balls is crucial, if the overall mass is not
constant for example for Suzuki–Miyaura reaction where the
yield significantly drops when the ball diameter is changed
from 15 mm to 10 mm at a constant ball number.22 However, if
a constant mass of milling balls and, thus, a constant filling
ratio is adjusted the product yield is unchanged independent
from the ball diameter. This independency is in contrast to
particle refinement processes where the ball diameter influences
the product particle size.24 The number of balls depends on the
volume of the grinding chamber and is limited by the filling
ratio r. It is known that for dry grinding a ball filling ratio of r =
0.3 results in the most effective feed grinding.20 Depending on
the size of the milling balls a certain amount of balls is
necessary to guarantee a qualitative yield.24 A higher number
of balls at constant density and diameter influence the reaction
due to the change in stress number. The correlation between
yield and number of balls seems to be almost linear for
chemical synthesis.15,22 The use of too many balls will result
in only a low yield due to the restricted movement of balls
inside the loaded pot.22 Otherwise in a charge with too few balls
strong abrasion can take place and may not be in line with the
recommendation of the mill manufacturer.24 A mixture of balls
with different size may randomize the motion of the ball charge
in a way that without well-defined trajectories tracks in the pot
are avoided.36–39
This journal is c The Royal Society of Chemistry 2013

Review Article
The yield increases with milling time22,30,40 due to the
consistent increase in the number of ball collisions and the
total amount of transferred energy. However, decomposition of
the product can be observed with prolonged reaction time for
some reactions.22
Wear in planetary ball mills
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.
The collision of balls against each other and the wall of the pot
during the grinding process will lead to wear of the container
and the balls. The dislodged material gets incorporated into the
powder where it contaminates or alters the powder chemistry.6
From an economic and synthetic point of view wear is crucial in
ball mills. When wear contaminates the final product complex
separation strategies for the remained particles are often neces-
sary and the total cost of the process will rise.6 Mostly the wear is
seen as a contamination of the final product and could be a
serious problem, for instance 1–5% Fe is normally present in
most powders (e.g. B5% Fe in quartz powder and B2% Fe in
silicon powder)41,42 milled with steel balls.3 With harder balls
the abrasive wear is less, but harder balls tend to spall.43
Only in some rare cases it is possible to use the wear of the
grinding chamber and grinding balls to introduce reagents or
catalysts in a process.44 Iron silicates, for instance, can be synthe-
sized from milling quartz powder with steel balls in the presence of
oxygen because nanoscale iron and iron oxide particles are formed
as a result of wear and react with the silica.41
The wear is highly affected by the adhesion of the powder
material to the ball, whether the material is able to form a layer
around the ball respectively.20,43 If the ball surface is insufficiently
coated due to low adhesion of a free flowing powder the energy is
wasted in wearing the balls.39 This effect is even stronger when
not a dry powder but a suspension is processed: the wear in wet
grinding is up to 7 times higher than in dry processes due to an
unlayered ball surface accompanied with corrosive wear.43–46
The amount of dislodged material from balls and the
grinding chamber which is incorporated into the powder also
depends on time and energy input of milling, stress intensity of
milling as well as on the milling atmosphere.3,43 Most important
for the relative ball wear is the overall energy input into the
grinding chamber.47 The total amount of wear increases with
milling time or energy input, respectively, so that the milling
process should be carried out just for the required duration.48 In
order to gather information about the influence of the process
parameters on the wear, studies were carried out with steel balls
in a planetary ball mill without using powder. It seems to be
clear that powder charge in the grinding chamber will reduce the
total amount of wear while without powder the wear progress
and dependencies remain in the same manner.48,49
The wear rate constant Kc of the dislodged material can
be measured experimentally and it quantifies the amount of
wear released into the powder per time step.49 The wear rate
constant Kc increases almost linearly with the revolution speed of
the mill and is mainly independent from milling time as it was
found from experiments with steel balls processed without powder.
With the Pulverisette 7 (Fritsch) it has been found that for high
speeds above 700 rpm the amount of wear is increasing rapidly.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
The rapid increase may be explained from the dislodged
material acting as an additional abrasive component and from
frictional heat which may change the material properties to a
more brittle material or the abrasion changes from the lower
shelf to the upper shelf.49
The wear rate constant Kc can reach a minimum for a certain
ball diameter while at inappropriate ball filling ratios r it will
become maximum.49 For example in a Pulverisette 7 (Fritsch,
VPot = 45 ml) the minimum Kc is achieved at a ball diameter of
dB = 12.7 mm.49 For the same device Kc increases with increasing
ball filling ratio r till the maximum is reached for r = 0.5. Above
this value the wear rate starts to decrease again. The reason is
that a higher number of balls lead to a higher number of
collisions and therefore a higher Kc. But with an excess number
of balls the motion of balls inside the grinding chamber is
interrupted and collisions with high enough impact energy to
cause wear are emerging less frequent.
In addition the contamination can be minimized by using
the same material for the milling chamber and milling balls as
the feed material so there will be no cross-contamination while
the changing chemistry of the product powder still has to be
considered. This alternative is only useful in cases where a
suitable medium for the grinding chamber and the balls is
available.3,50 Especially in reactive milling the atmosphere is
significant. However, often the container is not sealed properly
and the surrounding atmosphere which means mostly air
containing oxygen and nitrogen attain into the reaction and
the powder is contaminated as a result of side reactions with
the leaking atmosphere.3,50 As an inert atmosphere argon with high
purity of more than 99.99% would be suitable but is not economic-
ally viable and thus not feasible on an industrial scale.3
Determination of energy input using MSR
The processing of a highly exothermic powder mixture which
will react in a mechanically induced self-sustaining reaction
(MSR)37,51 can provide valuable information about the energy
supply from a milling device to powder charge. The advantage of
MSR is that the start of the self-sustaining reaction is conducted
by a high temperature increase52 hence these ideal systems are
useful to study the kinetics of mechanical alloying.53
The time period from activation to the reference point of
high temperature increase when the critical state is reached is
defined as ignition time tig.37 For the investigated reaction it is
assumed, that the net amount of energy necessary to generate
the self-sustaining reaction is constant51 and all mechanical
energy provided by the mill is consumed by the powder mixture
and used for reactant activation to reach the critical state
because the reaction always takes place after a certain milling
time independent from milling parameters. From the experi-
ments it can be seen that there is a direct relationship between
the reciprocal of ignition time and the revolution speed nR to
the third as it was proposed from theoretical computation
based on a collision model.51,54,55
1
/ nR 3 (4)
tig
Chem. Soc. Rev., 2013, 42, 7660--7667 7663
 View Article Online

Chem Soc Rev Review Article

The ignition time decreases almost linearly with increasing

ball-to-powder ratio at a constant ball-to-powder ratio and the
activation takes place uniformly over the whole sample which
gives evidence for the scalable character of experiments in
planetary ball mills.
Higher ball densities or ball diameters result in a shorter
ignition time due to the higher intensity of collision at higher
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.

ball masses.51,56 The big advantage of the discussed experiment

is that the ignition time is a clear indicator for a certain energy
supply and thus can be used for comparison of the efficiency of
different milling devices which are almost incomparable with
respect to milling time and loading mechanisms.51
Fig. 3 Influence of varied frictional coefficients on simulated ball motion
qualitatively fitted to different feeds.20

Simulation of ball motion in planetary ball

mills
The prediction of the outcome of mechanochemical synthesis
and mechanical alloying from simulation data is quite difficult
due to changes in composition, particle size, morphology,
structure and crystallinity.57 However, several studies have been
found to use simulations to get a better understanding of the
effect of process parameters on the energy entry and, hence, on
the grinding and mechanochemical progress. Discrete element
method-based simulations are used to describe the energy
transfer from milling tools to the powder, which are helpful
to get a better understanding of how milling process para-
meters or the design of the planetary ball mills affect the
grinding process by comparison of computational results and
experiments.58 The discrete-element-method (DEM) is a numer-
ical approach for the analysis of systems with discrete bodies.59
At each time step the acceleration and velocity of each grinding
ball as well as contact forces for colliding bodies (compressive
forces + shear force) are calculated according to Newton’s law of
motion and the trajectories of the balls during grinding are
predicted.16,58 A collision is identified when milling balls touch
or a ball hits the milling chamber wall and, thus, energy is
dissipated in this discrete event.26 A collision event also leads to
a change in motion of the ball in a normal and/or tangential
direction to the impact as well as a change in rotational motion
with a possible slip of the ball against the wall.60 The main
challenge that has to be met when utilizing DEM predictions is
the effect of the powder charge on the motion of the balls which
influences the energy exchange and energy fractions.26 The
energy is transferred and dissipated into the charge while only
a fraction of energy dissipated by collision or attrition is
utilized for grinding and the other fraction is only used for
the movement of the milled material without refining. It is
difficult to estimate the efficiency at which the dissipated
energy can be conversed to grinding of particles between the
surfaces of colliding and sliding bodies, so for simplification it
is assumed that the efficiency of energy conversion to grinding
is constant.60
The powder trapped between colliding bodies makes the
collisions of balls and wall less elastic.18 The friction caused by
the powder influences the motion pattern as well as the relative
impact velocity and stress energy distribution. With increasing
friction values the motion pattern changes from cascading to
cataracting and rolling (see Fig. 3).20
The results show the importance of the friction conditions
because with increased static or gliding friction when an
adhesive powder is processed e.g. marble relative velocities
are higher and more collisions take place per time step. With
increasing rolling friction the relative velocity as well as the
number of collisions per time step decrease.20
In DEM-simulation only the motion of each ball is observed
instead of dealing with powder particles to be milled. The
simplification reduces the difficulty of handling the powder
particles in the calculation which would require much more
efficient computers and induce inadmissible simulation times.
The presence of the milled material in the simulation is taken
into account by appropriate selection of a coefficient of restitu-
tion and a coefficient of friction.26,57,60 These coefficients have
significant influence on the process predictions, e.g. on the
energy consumption,61 and vary as a function of powder coat-
ing thickness, impact velocity and the type of impact and are
hence system dependent. For this reason the coefficients have
to be26 carefully chosen via parameter fitting methods for
instance from experiments62 determining the angle of repose63
or the behaviour of a single rolling ball.26
Influence of mill parameters on energy entry
DEM-simulations of planetary ball mills operated at different
parameter sets give better insight into the mechanisms inside
the grinding chamber like the motion, the collisions of balls
and balls with the wall as well as the energy input. The
frequency distributions of collision duration26 and collisions’
relative velocity20 calculated from ball motion are narrow which
indicates a very regular motion pattern. From relative velocity
distributions of stress energy fractions smaller than a certain
value can be derived. The fraction distributions can be used for
comparison of stress conditions inside the mill at different
operation parameters.20 The specific energy Ei transferred to
the powder is a key factor for identifying the most important
parameter setting to evaluate the milling conditions or mill
configuration.

7664 Chem. Soc. Rev., 2013, 42, 7660--7667 This journal is c The Royal Society of Chemistry 2013

Review Article
It is calculated as the total sum of impact energy of collision
events n as given by eqn (5) and it is equivalent to eqn (2) for a
constant grinding ball mass m.14,16,21,40
Xn
1
Ei ¼ mvj 2 (5)
j¼1
4m P
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.
However, the comparison between experimental results and
simulation data is challenging due to changes in composition,
particle size, morphology, structure and crystallinity.
The specific energy increases with increasing revolution
speed nR and increases linearly with the number of balls and,
thus, the number of collisions. Higher energy and/or the
collision number will lead to a better mechanochemical activation
and mechanical alloying which can also be seen from experimental
data. It can be seen that the rotation of the disk and pots in a
counterdirection (cf. Fig. 1) is more effective in grinding and
mechanochemical activation than the normal direction as
a result of higher energy and higher number of collisions.
These computational results are in very good agreement with
experimental results.14,17
The mill geometry has a significant effect on the number of
balls and the speed that can be used as well as the ball motion
and has to be carefully considered. At a constant pot height to
diameter ratio the pot diameter influences the energy Ei which
can be explained from the ball motion in the pot.17
Ei p d p 4 (6)
The kinetic energy Ekin of a single ball is equal to the product
of the ball mass m, the pot diameter dP and the acceleration a
acting on a ball moving from one point to another. The
acceleration can be expressed by the revolution radius R and
the angular velocity O.16
Ekin = 1/2mvj2 = madP (a = RO) (7)
Therefore, while the pot diameter is changed the specific
energy changes linearly. However, with the diameter of the pot
the number of balls nB will increase at a constant filling ratio
and a constant pot height to diameter ratio so that nB p dp3
and as a result the energy are a function of the pot diameter to
the fourth power. From eqn (6) it can also be derived that the
energy is dependent on the revolution radius R:
Ei p R (8)
The maximum Ei can be reached by optimal setting of
revolution radius R and speed ratio k.14 The specific energy
increases with rotational-to-revolution speed ratio k. The speed
ratio at the maximum specific energy shifts to higher values
when the revolution radius increases. The maximum specific
energy is defined as the critical energy at which just no rolling
takes place. It is possible to compare the particle refinement to
computational results over a wide range of conditions using the
relation between specific energy and the grinding rate KP.64,65
The grinding rate representing the rate of particle breakage is
well correlated with the specific energy in planetary ball mills
for several inorganic powders and the relationship can be
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
expressed by a linear function.16,64,66 Therefore the specific
energy can be used for prediction and estimation of the
grinding rate and optimization of conditions.
The particle size change during the experimental grinding
process can be described using an empirical equation:16
 
Dt Dl Dl
¼ 1 expðKP tÞ þ (9)
D0 D0 D0
with Dt/D0 as the normalized particle size and the grinding rate
KP. The correlation with experiments shows that the rate of size
reduction KP increases with decreasing ball diameter at high
rotational speeds while it is independent of ball size for low
speeds.66 KP increases with increasing revolution speed and
decreases with sample weight.16
Simulation of geometry scale-up
A geometrical scale up of fscale times of the mill size (pot
diameter dP, pot height h and revolution radius R) shall
investigate the effect on the energy Ei. The computational
predicted effect based on the dependencies in eqn (6) and (8)
on the energy should be:
Ei p fscale5 (10)
As expected the simulated energy Ei increases with the geometry
factor fscale. The simulation results diverge slightly from the
predicted value so within the change of the three parameters
pot diameter, height and revolution radius the energy depends on
Ei p fscale4.87 (11)
The error may be caused from the effect of other parameters
like ball diameter, filling ratio or the speed ratio which are not
completely independent and lead to different motion and
collision patterns in the modified mill geometry.16
Conclusions
Comminution and mechanochemical processes in planetary
ball mills are complex and mainly dependent on the milled
material or the conducted synthesis as well as the desired
product. There is no general rule for the choice of optimal
conditions. Several parameters influence the success of each
individual milling or mechanochemical process therefore it is
inevitable to make a good parameter assessment as it is
discussed in the section above.
The assessment depends on whether high stress energies for
size reduction are needed or the process depends on the total
energy input. High stress energies can be provided by large ball
diameters or high ball densities and a high revolution speed.
The counterdirection of rotational to revolution speed is more
effective in refinement than the motion in a normal direction.
Especially for chemical synthesis the specific energy is
crucial and, thus, an effective mixing with high stress frequencies
is advantageous and can be achieved by the use of small balls at a
constant filling ratio. As the specific energy can hardly be measured
directly, self-sustaining reactions are useful to determine the energy
input of a certain milling device. In addition simulations based on
Chem. Soc. Rev., 2013, 42, 7660--7667 7665

Chem Soc Rev
DEM show the influence of milling geometry and milling ball
properties on the specific energy that can be computed and
correlated with the grinding rate.
With regard to wear and cost from energy consumption it is
recommended to choose a moderate revolution speed and carry
out a process only for the desired time. Especially sensitive
materials require softer milling conditions at lower temperatures
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.
and smaller collision energies.
Notes and references
1 G. R. Karagedov and N. Z. Lyakhov, KONA, 2003, 21, 76–87.
2 A. Bruckmann, A. Krebs and C. Bolm, Green Chem., 2008,
10, 1131.
3 C. Suryanarayana, Non-Equilibrium Processing of Materials,
Elsevier, 1999.
4 C. Suryanarayana, Prog. Mater. Sci., 2001, 46, 1–184.
5 P. Ramrao Patil and K. P. Ravindranathan Kartha, Green
Chem., 2009, 11, 953.
6 V. P. Balema, J. W. Wiench, M. Pruski and V. K. Pecharsky,
Chem. Commun., 2002, 724–725.
7 A. Pichon, A. Lazuen-Garay and S. L. James, CrystEngComm,
2006, 8, 211.
8 P. R. Patil and K. P. R. Kartha, Green Chem., 2009, 11,
953–956.
9 E. L. Fokina, N. I. Budim, V. G. Kochnev and G. G. Chermik,
J. Mater. Sci., 2004, 39, 5217–5221.
10 P. P. Chattopadhyay, I. Manna, S. Talapatra and S. K. Pabi,
Mater. Chem. Phys., 2000, 68, 85–94.
11 F. Cavalieri and F. Pardella, Waste Manage., 2002, 22,
913–916.
12 C. Miclea, C. Tanasoiu, A. Gheorghiu, C. F. Miclea and
V. Tanasoiu, J. Mater. Sci., 2004, 39, 5431–5434.
13 A. Castro, P. Millán, J. Ricote and L. Pardo, J. Mater. Chem.,
1999, 10, 767–771.
14 H. Mio, J. Kano, F. Saito and K. Kaneko, Int. J. Miner.
Process., 2004, 74, S85–S92.
15 A. Stolle, T. Szuppa, S. E. Leonhardt and B. Ondruschka,
Chem. Soc. Rev., 2011, 40, 2317–2329.
16 H. Mio, J. Kano and F. Saito, Chem. Eng. Sci., 2004, 59,
5909–5916.
17 H. Mio, J. Kano, F. Saito and K. Kaneko, Mater. Sci. Eng., A,
2002, 332, 75–80.
18 M. Magini, Mater. Sci. Forum, 1992, 88–90, 121–128.
19 X. Jiang, M. A. Trunov, M. Schoenitz, R. N. Dave and
E. L. Dreizin, J. Alloys Compd., 2009, 478, 246–251.
20 S. Rosenkranz, S. Breitung-Faes and A. Kwade, Powder
Technol., 2011, 212, 224–230.
21 M. Matsuoka, J. Hirata and S. Yoshizawa, Chem. Eng. Res.
Des., 2010, 88, 1169–1173.
22 F. Schneider, A. Stolle, B. Ondruschka and H. Hopf, Org.
Process Res. Dev., 2009, 13, 44–48.
23 F. Schneider, T. Szuppa, A. Stolle, B. Ondruschka and
H. Hopf, Green Chem., 2009, 11, 1894.
24 T. Szuppa, A. Stolle, B. Ondruschka and W. Hopfe, Green
Chem., 2010, 12, 1288.
7666 Chem. Soc. Rev., 2013, 42, 7660--7667
View Article Online
Review Article
25 W. Peukert, Int. J. Miner. Process., 2004, 74, S3–S17.
26 P. R. Santhanam and E. L. Dreizin, Powder Technol., 2012,
221, 403–411.
27 A. Kwade, Int. J. Miner. Process., 2004, 74, S93–S101.
28 M. H. Enayati, Z. Sadeghian, M. Salehi and A. Saidi, Mater.
Sci. Eng., A, 2004, 375–377, 809–811.
29 V. Dereck, N. P. J. Martinez and F. Lamatz, Angew. Chem.,
Int. Ed., 2009, 121, 9482–9485.
30 B. Rodriguez, A. Bruckmann and C. Bolm, Chem.–Eur. J.,
2007, 13, 4710–4722.
31 S. D. Kaloshkin, V. V. Tcherdyntsev, Y. V. Baldokhin,
I. A. Tomilin and E. V. Shelekhov, J. Non-Cryst. Solids,
2001, 287, 329–333.
32 V. V. Tcherdyntsev, S. D. Kaloshkin, E. V. Shelekhov,
A. I. Salimon, S. Sartori and G. Principi, Intermetallics,
2005, 13, 841–847.
33 K. Takehira and T. Shishido, Catal. Surv. Asia, 2007, 11,
1–30.
34 K. Sasaki, T. Masuda, I. Hideki and T. Mitsuda, J. Am.
Ceram. Soc., 1996, 79, 1569–1574.
35 S. Tadier, N. Le Bolay, C. Rey and C. Combes, Acta Biomater.,
2011, 7, 1817–1826.
36 L. Takacs, Materials Week, Cleveland, 1996.
37 L. Takacs, Prog. Mater. Sci., 2002, 47, 355–414.
38 J. L. Harringa, B. A. Cook and B. J. Beaudry, J. Mater. Sci.,
1992, 27, 801–804.
39 L. Takacs and J. S. McHenry, J. Mater. Sci., 2006, 41,
5246–5249.
40 H. Mio, S. Saeki, J. Kano and F. Saito, Environ. Sci. Technol.,
2002, 36, 1344–1348.
41 F. K. Urakaev, Int. J. Miner. Process., 2009, 92, 58–66.
42 N. Stevulová, M. Bálintová and K. Tkácová, J. Mater. Synth.
Process, 2000, 8, 265–270.
43 I. Iwasaki, Int. J. Miner. Process., 1988, 22, 345–360.
44 D. A. Fulmer, W. C. Shearouse, S. T. Medonza and J. Mack,
Green Chem., 2009, 11, 1821.
45 C. S. Gundewar, K. A. Natarajan, U. B. Nayak and
K. Satyanarayana, Int. J. Miner. Process., 1990, 29, 121–139.
46 G. Pintaude, A. P. Tschiptschin, D. K. Tanaka and
A. Sinatora, Wear, 2001, 250, 66–70.
47 B. Joost, PhD thesis, TU Braunschweig, 1995.
48 C. Suryanarayana, Int. Mater. Rev., 1995, 40, 41–64.
49 A. Sato, J. Kano and F. Saito, Adv. Powder Technol., 2010, 21,
212–216.
50 E. Ivanov, J. Mater. Synth. Process., 2000, 8, 235–244.
51 F. J. Gotor, M. Achimovicova, C. Real and P. Balaz, Powder
Technol., 2012, 233, 1–7.
52 S.-B. Li and H.-X. Zhai, J. Am. Ceram. Soc., 2005, 88,
2092–2098.
53 G. B. Schaffer and P. G. McCormik, Metall. Trans. A, 1992,
23, 1285–1290.
54 Abdellaoui, 1995.
55 M. Magini, A. Iasonna and F. Padella, Scr. Mater., 1996, 34,
13–19.
56 H. Yang and P. G. McCormik, J. Solid State Chem., 1993, 107,
258–263.
This journal is c The Royal Society of Chemistry 2013
 View Article Online

Review Article Chem Soc Rev

57 T. S. Ward, W. Chen, M. Schoenitz, R. N. Dave and
E. L. Dreizin, Acta Mater., 2005, 53, 2909–2918.
58 P. A. Cundall and O. D. L. Strack, Geotechnique, 1979, 29, 47–65.
59 Y. T. Feng, K. Han and D. R. J. Owen, Mater. Sci. Eng., A,
2004, 375–377, 815–819.
60 T. Inoue and K. Okaya, Int. J. Miner. Process., 1996, 44–45,
425–435.
Published on 07 February 2013. Downloaded by National University of Singapore on 12/10/2023 7:39:36 AM.
62 P. L. Blau, Tribol. Int., 2001, 34, 585–591.
63 J. Kano, N. Chujo and F. Saito, Adv. Powder Technol., 1997, 8,
39–51.
64 J. Kano, H. Mio and F. Saito, AIChE J., 2000, 46,
1694–1697.
65 J. Kano, H. Mio, F. Saito and M. Miyazaki, Miner. Eng., 2001,
14, 1213–1223.

61 K. S. Liddell and M. H. Moys, Anglais, 1988, 88, 49–57. 66 J. Kano and F. Saito, Powder Technol., 1998, 98, 166–170.

This journal is c The Royal Society of Chemistry 2013 Chem. Soc. Rev., 2013, 42, 7660--7667 7667