Second International Workshop on Granulation

22 – 25 June 2004, University of Sheffield

ABSTRACTS
SESSION I: HIGH SHEAR PROCESSES
HIGH SHEAR GRANULATION IN THE PHARMACEUTICAL INDUSTRY;
INHOMOGENEITY PHENOMENA.
# #&
Kaspar
#
van den Dries , Herman Vromans
N.V. Organon, AKZO-NOBEL, P.O. Box 20, 5340 BH Oss, The Netherlands
&
Faculty of Pharmacy, University of Utrecht, P.O. Box 80082, 3508 TB Utrecht, The Netherlands
In the pharmaceutical industry high hear granulation is a commonly used process step in the
production of a solid formulation. Because of regulatory guidelines and product quality
assurance, an important granule requirement is homogeneity. However, often granule
inhomogeneity is observed, expressed as a granule size-dependent variation in composition of the
granules [1-4]. As a consequence of segregation during subsequent process steps this might
introduce content uniformity problems. In order to solve these quality problems the mechanisms
involved in the formation of non-homogeneous granulates have to be elucidated.
A research project focused on this subject revealed that at least three different mechanism are of
importance;
1. The mechanism of granule formation (nucleation); the granule inhomogeneity already
exists in the first minute of the granulation process, indicating that nucleation plays an
eminent role. To investigate the early second s of high shear granulation the process was
frozen with liquid nitrogen. From this approach the following insights have been
achieved: The mechanical agitation results in binder dispersion. Parallel to the dispersion
process penetration of the binder liquid leads to the formation of granules. This prevents
complete dispersion. The dynamic strength of the granules determines whether freshly
formed granules survive the shear forces.
2. Preferential growth during layering; Layering growth experiments with single granules
showed that fine primary particles have a higher affinity for granule growth than coarse
particles. This results in an accumulation of the fine particles in the granules. It was
argued that fine particles can enter the surface pores, while coarse particles cannot. This
provides “a headstart” for the fine particles in the powder mix with respect to growth. A
model based on this hypothesis was able to describe the experimental results
quantitatively.
3. Granule breakage behaviour; The breakage behavio ur in the high shear mixer was
investigated with tracer experiments [2]. Comparison of these results with the granule
homogeneity showed that there is a interrelationship between granule breakage and the
drug distribution over the granules. Granule breakage results in a continuous exchange of
(primary) particles between granules. This yielded a homogeneous distribution. On the
other hand, when there is minimal breakage, intact granules function as kernels for
preferential growth. Consequently, the development of a poor distribution depends on the
constituents’ particle size differences.

1
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

Unravelling the mechanism is the first step towards a high quality product. In addition,
experiments were performed to investigate the influence of process and formulation conditions
on the above described mechanisms. This level of understanding can prevent granule
inhomogeneity in the future. Eventually, this scientific approach to process development must
results in proper control of high shear granulation and adequate granule properties.

1. Hapgood, K.P., Hartman, H.E., Kaur, C., Plank, R., Harmon, P., Zega, J.A., 2002. A case study of drug
distribution in wet granulation. World Congress of Particle Technology 4, Sydney, Australia (21-25 July).
2. van den Dries, K., de Vegt, O., Girard, V., Vromans, H., 2003. Granule breakage phenomena in a high shear
mixer; influence of process and formulation variables and consequences on granule homogeneity. Powder
Technol. 133, 228-236.
3. van den Dries, K., Vromans, H., 2002. Relationship between inhomogeneity phenomena and granule
growth mechanisms in a high shear mixer. Int. J. Pharm. 247, 167-177.
4. Vromans, H., Poels -Janssen, H.G.M., Egermann, H., 1999. Effects of high shear granulation on granulate
homogeneity. Pharm. Dev. Technol. 4, 297 -303.

EMPIRICAL TO MECHANISTIC MODELLING IN HIGH SHEAR GRANULATION –

EXAMPLES FROM THE PHARMACEUTICAL INDUSTR Y
I. Niklasson Björn, A. Jansson, M. Karlsson, S. Folestad and A. Rasmuson
AstraZeneca Centre of Excellence for Process Analytical Technology P A R &D Mölndal, Sweden
Wet granulation in high shear mixers is an important unit operation step often used in
development and manufacturing of pharmaceutical oral solid dosage forms. It is a complex
process step and the knowledge about the fundamental mechanisms behind it is limited. To
improve the understanding, it is important to find different ways to model and monitor the
process to enable predictive scale-up and control of the process. Different types of models have
been used to describe the dynamic behaviour of the impeller torque and the granulate growth.
A change in the impeller torque is a direct measure of changes in the flow pattern and/or changes
in material properties within the mixer. However, the interpretation of the impeller torque
measurements is not trivial. The impeller torque has been modelled for both the dry and wet
mixing phases. In the dry mixing phase a mechanistic model found in the literature was used. In
the wet mixing phase a multivariate process model was used.
The model used in the dry mixing phase was slightly modified and a new term was added.
Experiments where made with microcrystalline cellulose. The results showed good agreement
between the modified model and the experimental data. The model was also used for predictions.
For the wet mixing phase the multivariate process model was constructed from timely
measurements of impeller torque, moisture content, particle size, density and NIR spectra. The
multivariate model was able to successfully explain the variation in the impeller torque for
different operating conditions and material properties.
To describe the granulate size distribution growth a population balance model was developed.
The theory of population balance modelling requires knowledge of how and to what extent two
particles are likely to coalesce or break. In the population balance equation, the probability
function for coalescence is known as the coalescence kernel. A coalescence kernel has been
constructed, where theories from the macro- and micro-level have been combined. The effect of
different process parameters like impeller speed, initial particle size, bowl size, liquid viscosity,

2
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

addition time and wet massing time have been included. A simple model of the shear forces was
found to be related to the impeller to rque exerted by the in the bowl.

WET GRANULATION IN A BATCH HIGH SHEAR MIXER

Khashayar Saleh, Lilian Vialatte & Pierre Guigon
Dépt. Génie Chimique, Université de Technologie de Compiègne, France
This study deals with the wet granulation in a high shear mixer. The experimental apparatus is a
laboratory scale "Lödige" granulator, with a maximum volume of 20 L and equipped with a
chopper and an pneumatic atomising system.
The main objectives of the study are, firstly, to point out the effect of binder-solid interactions on
mechanisms of granulation and, secondly, to establish the technological signature of the
granulation device.
Two kinds of alumina with different particle size distributions (alumina SH100 and alumina
SH30) were granulated using various liquids having different surface tensions, viscosities, binder
concentrations, densities, ... (water, aqueous solutions of polyethyleneglycol or polyvinylic
alcohol).
Experimental results showed that the granulation process proceeds generally through a three
distinct growth regimes independent of the nature of the powder, the binder liquid or the
operating conditions. However, the transition between different regimes depends on the physico
chemical properties of solids and liquids, on operating conditions and on the experimental
procedure.
Analysis of samples removed from the granulator at different time intervals permitted to point out
growth mechanisms involved in each regime as follows:
After a first stage in which the granulation mechanism is the nucleation of primary particles
(regime 0), it appears a first regime in which there is no creation of new agglomerate species.
During this regime, there is a balance between attrition phenomena which create fine particles or
agglomerates in the one hand and fine consuming growth mechanisms (agglomeration, layering)
on the other hand. In the second regime, granules grow by non-preferential layering of the fine
agglomerates onto the surface of the other species. The phenomena that governs the transition
between the first and the second regime are the densification of the granules and the binder
transport to the granule surface. Finally, when the fine agglomerates are entirely picked up, a
third regime appears in which the granulation mechanism is a preferential coalescence
mechanism of small and coarse granules.
A physico-chemical approach based on the individual surface free energy of the powder and the
binder liquids was used to assess the substrate-binder interactions. A unifying correlation to
predict the three stage granulation profile and transition points was also developed.
Finally, scale-up experiments, carried out on an industrial scale granulator, confirmed the three
stage granulation profile revealed in the laboratory scale granulator.

3
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

MECHANISMS IN HIGH VISCOSITY IMMERSION-GRANULATION

1 1 2 2
S.L.
1
Rough , D.I. Wilson , A.E. Bayly and D.W. York
Department of Chemical Engineering, University of Cambridge, UK
2
Procter and Gamble Ltd, Newcastle Technical Centre, Longbenton, UK
An atypical high shear granulation process is investigated in which a fine inert powder is bound
with a highly viscous surfactant paste. The mechanism comprises adsorption of powder particles
onto paste fragments, breakage of powder-coated paste granules, micro-mixing of the granules
with absorption of the powder, granule growth via coalescence, and finally granule consolidation.
These stages are supported by micrographic and granule size distribution data. The agglomeration
process features two main mechanisms, namely binder distribution follo wed by granule
consolidation and coalescence, with their key transitions exhibiting a different dependency upon
operating parameters.
A number of time-dependent consistency regimes can be identified and quantitatively described
by parameters obtained from bulk tapping compaction tests. Various dynamic granule
characteristics are inferred from these pseudo-steady state bulk parameters, and are used to track
the agglomeration process, the results of which are consistent with the Iveson et al.1 steady state
agglomeration regime maps. Of particular interest is the trend in Hausner ratio, which provides
information on the inter- granular friction and cohesivity.
The effects of paste/powder composition, paste rheology and mixing speed upon the transition
rate can be explained physically in terms of adsorption, viscous and mechanical energy
dissipation mechanisms. In summary, the work introduces a preliminary analysis of an
immersion- granulation mechanism in which a number of key features are identified. The results
suggest that a fundamental investigation of steady state and dynamic system properties at both
the macro- and micro-scale would be highly advantageous.

1. Iveson S.M., Wauters P.A.L., Forrest S., Litster J.D., Meesters G.M.H. and Scarlett B. (2001) Growth
regime map for liquid-bound granules: further development and experimental validation, Powder Tech.,
117, p.83-97, 2001.

AGGLOMERATION PROPER TIES FOR GLUTEN AND GLUTEN FREE CEREALS

FLOURS BY WATER ADDITION AND SHEARING - MIXING RHEOLOGY COUPLED
WITH MICROSCOPIC OBSERVATION
Hebrard A., Oulahna D.
Laboratoire de Génie des Procédés des Solides Divisés, Ecole des Mines d'Albi-Carmaux, France
The agglomeration properties of gluten content cereals such as durum wheat semolina and gluten
free cereals such as amaranth, corn, manioc, quinoa and buck wheat flour, have been
investigated. The different states of agglomeration have been identified by the evolution of
mixing rheological properties under binder addition and shearing conditions determined by a
mixer torque rheometer (MTR) [Hancock, 1994]. Surface state alterations during agglomeration
have been observed by environmental microscopic observations (ESEM) which do not impose
sample pre- treatment.
All these samples present the same rheological footprint under specific hydration and shear rate,
three agglomeration states can be identified. Progressively, particles agglomerate and form
aggregates which coalesce, their size increases up to the development of a paste, then the

4
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

maximum viscosity is reached. Further addition of water results in paste dilution and the
viscosity decreases.
The agglomeration properties of all these samples and microscopic observations, in this study
allows comparison of the hydration rate, the mixing viscosity to reaching a specific state of
agglomeration, the surface state of the agglomerates and the structure of the links formed
between particles composing the agglomerates.

GRANULAR FLOW IN A PLANETARY MIXER

B.F.C. Laurent1 and D.I. Wilson2
1
Institute for Manufacturing, Cambridge University, Department of Engineering, Cambridge CB2 1RX, UK
2
Department of Chemical Engineering, University of Cambridge, Cambridge, CB2 3RA, UK
Granulation is an important step in drugs manufacturing processes. Here, flow patterns in dry
systems and during wet-mass granulation were characterised using Positron Emission Particle
Tracking. Powder flows were studied first in a planetary mixer where two different regimes were
identified. In a dry state, the material flows along a torus formed by the beater of the agitator. The
angular velocity and total speed scaled with the agitator speed, showing that powder flow is
controlled by the number of blade passes. When granulation proceeds, the flow fields in the core
showed the presence of another torus situated above the one observed with the dry system,
creating a region of high shear where both tori meet. Related work using a simplified mixer
showed how granular flows may be related to torque rheometry. This opens the prospect of
designing new continuous processes for the pharmaceutical industry in order to implement leaner
manufacturing methods and just- in-time techniques of production.

5
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

SESSION II: FLUIDIZED BED PROCESSES

STUDIES OF FLUID BED GRANULATION IN AN INDUSTRIAL R&D CONTEX T
Renee Boerefijn
Following a brief overview of recent developments in laundry powder processing
and current drivers and trends, a review will be presented of collaborative
research between Unilever R&D and Sheffield University on fluidised bed
granulation. These fall apart into 4 areas:
1. Contact zone studies
2. Growth similarity for non- ideal systems
3. Growth and breakage on balance
4. Scale-up rules based on similarity
Together, this lays the broad ground-work for fluidised bed granulation. Some
concluding remarks provide pointers for the future of research in this area.

SPRAY GRANULATION, AGGLOMERATION AND COATING PROCESSES USING

CONTINUOUS FLUIDIZED BED TECHNOLOGY
Michael Jacob
Glatt Ingenieurtechnik GmbH / R&D Nordstrasse 12, 99427 Weimar, Germany
Fluidized bed techniques are widely used in all fields of industry. The range of application
includes both heat and mass transfer processes (for instance cooling and drying calcination) and
complex multi-phase processes like agglomeration, spray granulation and coating.
In many field of the industry products created using spray granulation or agglomeration processes
obtain growing importance. The fields of application include products like instant food, instant
drinks, mixtures and single components in the food, pharmaceutical or chemical industry. The
products range from dustfree spray-granulated fine chemicals with high bulk density and high
resistance against abrasion to instant drinks for health care which are homogeneous in
composition, free flowing, easily wettable and fast to dissolve in liquids.
To manufacture these kinds of products continuous fluidized bed processes are more than
suitable.
Based on basic principles of fluidized bed processes two main types of continuous fluidized bed
plants were developed. These types will be describ ed in the following cha pters.
The first version ( fig. 1 ) is a fluidized bed unit with rectangular design of bottom screen. The
bottom screen is used to equalize the flow of process air and is the lower border of the fluidized
bed. In the normal case the product discharge is realized using a rotary valve. The discharge flow
rate can be adjusted by changing the speed of this volumetric dosing device. Normally an internal
classifying of the product flow in the unit is not rea lized. In standard units the dedusting of the

6
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

exhaust is done using an integrated bag filter system. The filter bags are cleaned by pulse back
air. By using internal filter systems all solid streams are kept inside the process room and
additional conveying systems to recycle solids are not necessary. If needed this version of
fluidized bed units can be delivered alternatively with an external dedusting system or wet
scrubber. The main advantage of this unit is the possibility to carry out multi-step processes in
one apparatus. In fig. 1 a combination of an agglomeration or spray granulation process with a
cooling process is shown. The section below the bottom screen can be devided into several inlet
air chambers to distribute the air flow to the different process steps and feed different air qualities
to the fluidized bed.
Fluidized bed granulator / coater for continuous processing - type GFG

clean gas-
chamber

solid feed
seeds / powder / wet solids filter system
integrated

expansion
chamber

spray system fluidized bed

top-spray chamber
bottom-spray

bottom screen inlet air

chambers

product discharge
Process air Process air
for fluidization for fluidization
and drying and cooling

Figure 1
In contrast with the unit of fig. 1 in fig. 2 a fluidized bed unit with circular design of the bottom
screen is shown. Typical for this special version of fluidized bed unit is the central classifying
discharge pipe for continuous product discharge. The upper end of this vertical pipe is fitted
direct in the bottom screen. The bottom screen is used to equa lize the flow of process air and is
the lower border of the fluidized bed. A second air flow is fed to the discharge pipe. The flow
velocity is adjusted in such a way that small particles are blown back to fluidized bed. Particles
with a falling velocity higher than the flow velocity in the discharge is will be discharged.
Because of that mechanism the particles size of discharged solids can be adjusted and the product
is free of dust. These version of fluidized bed units are normally equipped with external
dedusting systems. The recovered dust of external dedusters can be recirculated to the
agglomeration or spray granulation process.

7
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

Fluidized bed granulator for continuous processing - type AGT

segregation chamber

expansion chamber
spray system
top-spray
bottom-spray
dust
fluidized bed chamber
starting material bottom screen
seeds
inlet air chamber process air
for fluidization
and drying

discharge pipe classifying air

product

Figure 2

In all agglomeration processes or in the special case of spray granulation liquids are sprayed into
the fluidized bed. In order to do that spray nozzles are installed in the process room to distribute
the liquid. Different spray systems can be used to spray either on top of the fluidized bed ( top-
spray ) or direct upwards into the fluidized bed ( bottom-spray ) . Both spray systems can be used
in both versions of fluidized bed units.
This article focuses on continuous processing options where liquids are sprayed onto fluidized
products.
The article will focus on both general principles and case studies of large-scale industrial
applications. For this reason the audience can enter the field of continuous fluidized bed
technology and recognize its potential for industrial application.

ON THE PREDICTABILITY OF FLUIDIZED BED SPRAY GRANULATION: AN ALYSIS

OF PARTICLE POPULATIONS AND TRANSPORT PHENOMENA
Stefan Heinrich
Otto-von-Guericke-University Magdeburg
Institute of Process Equipment and Environmental Technology
Universitätsplatz 2, D-39106 Magdeburg
e-mail: [email protected]
The fluidized bed spray granulation (FBSG) is a process used for the production of granular high-
quality, free- flowing, low-dust and low-attrition solids originating from liquid products, e. g.
solutions, suspensions, melts and emulsions. The advantage is the coupling of the wetting,
drying, particle enlarging, shaping, homogenisation and separation processes and the the
production in a single processing step. Especially for large production units a continuous
operation of the FBSG is desirable. The continuous granulation process presents, unlike to the
batch-operation, the advantage to operate the plant under stationary condition at high
throughputs. The stationary operation point is reached, provided constant granulate spectrum beside

8
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

constant mass flows and constant thermal conditions, whereby initially fed granulates have to be removed at
all. Sometimes this unsteady phase lasting up to a few hours.
This contribution presents calculations of the transient behaviour of the particle size distributions
in the fluidized bed and in the product flow with regard to the number of particles, surface, mass
and volume of particle, of the air temperature, of the air humidity, of the wetting efficiency, of
the temperature of the liquid film around the particles, of the particle temperature and of the
pneumatics (porosity, Reynolds number, velocities) during the unsteady start-up period,
including all input and output particle (disperse phase) and mass (continuous phase) flows. Seeds
are produced in several ways: (1.) Particle attrition, (2.) Overspray (nucleation) are non
deposited dried drops, (3.) Agglomeration of wet and dry particles, (4.) by a external crusher.
Therefore, the population balance model takes into account the seeds formation (fines
generation). An dynamic oscillating behaviour, which may occurre during the unsteady start-up
phase, caused by classifying and milling processes, is discussed. The aim of the examinations is
to study the stability behaviour of the FB (Fig. 1).
Furthermore, the complex correlations of a number of microprocesses, spraying, wetting, drop
deposition, heat transfer, drying / evaporation and mass transfer considering that the solid
dispersion were studied, and transient three-dimensional distributions of the air humidity, the air
temperature, the particle wetting efficiency, the liquid film temperature, the particle temperature,
the local liquid loading and liquid deposition (spraying zone) and the local evaporation rate were
calculated (Fig. 2).
The population balance model (Fig. 3) and the heat and mass transfer model (Fig. 4) was
evaluated in a semi- industrial fluidized bed pilot plant of the institute.

stable behaviour (coarse milling) unstable behaviour (fine milling)

Figure 1. Calculated particle size distributions of a continuous external separation
process.

9
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

Figure 2a: Unsteady distributions at a top-spray injection onto a fluidized bed at time t =
100 sec with low solid dispersion (D vertical = 0.01 m²/s, Dhorizontal = 0.001 m²/s).

Figure 2b: Unsteady distributions at a top-spray injection onto a fluidized bed at time t =
100 sec with high solid dispersion (Dvertical = 0.1 m²/s, Dhorizontal = 0.01 m²/s).

10
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

Bed - Measurement Bed - Simulation

Product - Measurement Product - Simulation

Figure 3. Comparison of measured and simulated particle size distributions of a continuous

internal separation process.

view on 120 - 300 degree plane

height above distribution plate [m]
height above distribution plate [m]

radius [m]

Figure 4. Measured stationary spatial air temperature distributions of a water sprayed FB

of plastic beads.

11
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

PARTICLE DESIGN AND EVALUATION OF DRY CO2 RECOVERY SORBENT WITH

A LIQUID HOLDING CAP ABILITY
Satoshi Kimura, Masamitsu Adachi, Reiji Noda and Masayuki Horio
Tokyo University of Agriculture and Technology, Graduate School of BASE, 2-24-16, Nakamachi, Tokyo, 184-
8588, Japan
e-mail: [email protected]
For CO2 capture at high temperature/high CO2 partial pressure conditions, new dry CO2 removal
sorbents were developed based on the recently developed lithium silicate with additives to
accelerate CO 2 sorption. To avoid adhesion and defluidization associated with the liquid
carbonate phase formation in the temperature range of 500-600 C, sorbent particles were
designed so that the liquid lithium carbonate cannot really wet the sorbent surface. To achieve
this the lithium silicate fine particles granulated forming the core of a sorbent particle were
coated with a layer of coarse alumina particles. Different types of sorbents were manufactured by
granulation and coating with different ceramic sol binders and different prescriptions in a
tumbling fluidized bed. Minimum fluidization velocity, fluidizing behavior and
adsorption/desorption properties of the sorbent particles were determined in a laboratory scale
fluidized bed reactor (i.d.,0.03m) in the temperature range up to 1073K. In cyclic
adsorption/desorption conditions, gradual decrease both in over all conversion and maximum
adsorption capacity took place. Nevertheless, the performance obtained for the sorbent
manufactured with alumina sol was found successful with no defluidization and with high
adsorption capacity of 50% in conversion (10wt% including the coat layer) over 60cycles.

PRELIMINARY DEM SIMU LATIONS OF FLUIDISED BED GRANULATION

K. D. Kafui & C. Thornton
School of Engineering, The University of Birmingham, UK
A fully 3D granular dynamics fluidised bed model based on the Distinct Element Method (DEM)
has been developed and adapted for a preliminary ‘proof-of-concept’ simulation of fluidised bed
granulation (Thornton & Kafui, 2004). In the model (Kafui et. al, 2002), the discrete particle-
particle or particle-wall interactions are based on theoretical contact mechanics in which the
actual material properties are utilised, ensuring that these properties govern the evolution of
contact forces and the resultant kinematics. The gas phase is modelled as a continuum and the
compressible Navier-Stokes equations of motion are solved using an adapted form of Patanker’s
SIMPLE methodology (Kuipers et. al, 1993).
The concept explored here is that of a spray zone notionally represented as a conical region of
fine mist of ‘adhesive’ or liq uid where the apex of the zone is the spray source. Discrete particles
which enter this region pick up ‘adhesive’ or liquid film, the amount (volume) and/or strength
(surface energy or viscosity) of which depends on the distance from the source and the time spent
in the zone.
For the ‘adhesive’ spray zone concept, surface energy based on the JKR theory is the notional
adhesive. A ‘drying’ adhesive bond between two particles is mimicked by increasing the
interface energy up to an asymptotic limit determined by the initial bond interface energy. The
duration of the activity of the surface energy of a particle outside the spray zone is limited, with
the surface energy ‘drying’ during this period and becoming inactive if no bonding occurs before
activity ceases.

12
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

Liquid spray is employed as the active binding medium in the spray zone for the second
modelling concept. Here, particles in the spray zone get a dosage of liquid on the surface, the
volume of dosage being a function of the time spent in the spray zone and the distance from the
spray source. Pendular liquid bridges are formed when a particle with ‘wet’ surface liquid makes
contact with another particle. The liquid bridge forces are evaluated using the formulations of
Lian et. al, 1993, and the viscosity of the liquid in a bridge or at a particle surface is increased
functionally from a ‘wet’ value to a ‘dry’ value according to its age. For viscosities below a set
proportion of dry viscosity dubbed the migration limit, it is assumed that liquid migration can
occur from a bridge to a bridgeless contact on the same particle, the driving force being the
difference between the migration limit and the viscosity in the bridge.
The adhesive conceptual formulation of the spray zone has been employed in a preliminary
simulation of the fluidised bed granulation of a bed of 100,000 Geldart group A particles (five
sizes, 45 mm – 55 mm). Video sequences of the evolution of granules and bonds will be
presented along with evolving granule size distributions. Work using the liquid spray formulation
is still in progress.
1. Thornton, C. & Kafui K.D. (2004). Fully 3D DEM simulations of fluidised beds including granulation. To
be presented at Fluidization XI, Ischia, 9-13 May, 2004.
2. Kafui, K.D., Thornton, C. & Adams, M.J. (2002). Discrete particle -continuum fluid mo delling of gas-solid
fluidised beds. Chemical Engineering Science, 57, 2395-2410.
3. Thornton, C. & Yin, K.K. (1991). Impact of elastic spheres with and without adhesion. Powder Technology,
65(1-3), 113-123.
4. Kuipers, J.A.M., van Duin, K.J., van Beckum, F.P.H. & van Swaaij, W.P.M. (1993). Computer simulation
of the hydrodynamics of a two -dimensional gas-fluidised bed. Computers in Chemical Engineering 17(8),
839-858.
5. Lian, G., Thornton, C. & Adams, M.J. (1993). A theoretical study of liquid bridge forces and stability
between two rigid spherical bodies. Journal of Colloid and Interface Science, 16, 138-147.

CONCEPTS TO IMPROVE THE EFFICIENCY OF FOOD AGGLOMERATION

S. Palzer
Nestlé Product Technology Centre, Lange Strasse 21, D-78224 Singen
In the food manufacturing industry several agglomeration technologies are applied: fluidised bed
agglomeration, steam jet agglomeration, agglomeration during drying and pressure agglomeration
like extrusion, roller compaction or tabletting. There are some concepts, which can be applied
across the length scale to improve the efficiency of agglomeration processes.
During various agglomeration processes moisture is added to the powder particles to increase the
adhesion forces by approaching or exceeding the glass transition temperature Tg of amorphous
substances. Calculating Tg for various moisture contents enables to identify for different
processes suitable moisture/temperature combinations.
Another approach is to describe pressure agglomeration by applying the laws of bulk solid
handling. The results obtained by trials and calculations allow characterising powdered foodstuffs
including their yielding and compaction behaviour. Thus suitable axial and radial stress
combinations and equipment design for the different stages of pressure agglomeration processes
can be defined.

13
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

BREAKAGE KINETICS DURING FLUIDISED BED GRANULATION

H.S. Tan, A.D. Salman and M.J. Hounslow,
Particle Products Group, Department of Chemical and Process Engineering, The University of Sheffield
It is widely acknowledged that in the great majority of granulation processes there are size
enlargement (agglomeration) processes and size reduction (breakage processes) active
simultaneously. Yet most modelling studies account only for agglomeration. The reasons for
this defect are two fold: first agglomeration-only models seem to give a good description of the
process and secondly even quite high- fidelity measurements of particle size distribution do not
allow model discrimination between agglomeration-only and agglomeration-plus-breakage
models. This defect poses a considerable problem in seeking to extrapolate or scale- up by means
of the model since in some processes the breakage rate might be quite high and so the actual
agglomeration rate is much greater tha n that predicted by the model.
In this paper we user specially made tracer particles to show the extent of breakage and to
quantify its kinetics. We address the issue of discriminating between attrition and fragmentation
and are able to draw conclusions about the effects of granule age and size.

14
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

SESSION III: COMPACTION PROCESSES

DRY GRANULATION OF ORGANIC POWDERS : DEPENDENCE OF PRESSURE 2D-

DISTRIBUTION ON DIFFERENT PROCESS PARAMETERS.
Thibaut Lecompte1, Pierre Doremus1, Jean-Claude Le Thiesse2, Jean-Claude
Masteau2, Laurent Perier-Camby3 and Gérard Thomas 3
1- Laboratoire 3S, INPG, 1025 rue de la piscine, F38041 St Martin d’Hères , France
2- Rhodia, Centre de recherche de Lyon, 55 av. des frères Perret, 69190 St Fons, France
3- Laboratoire Spin, LPMG-UMR CNRS 5148, Ecole Nationale Supérieure des Mines de St Etienne,158 Cours
Fauriel 42023 St Etienne, France
Correspondence: [email protected]
Nowadays dry granulation of powders
has become a very important research
topic because dry granulation is the most
economic way of granulation, making
easy the particle handling, and avoiding
the loss of matter during particle
processing, or particle transfers. This
kind of process has been deeply studied
but a better knowledge appears
necessary to control the great number of
parameters of the process. This is
particularly important in cases where the
nature of the material considered may
lead to compaction during which very
complex phenomena can occur.
Figure 1 - Instrumented roll compactor
In order to try and optimise dry
granulation process for organic compounds, a roll press has been realised with a series of
instruments allowing the control as high as possible of the compaction process. The apparatus
(figure 1) is constituted of three parts : first, a vertical container (50 kg maximum weight) with
rotating steel blades avoiding arches into which the powder is poured. Then a feeder transferring
the powder towards the rolls ; the feeder is equipped with an horizontal helical screw in a
cylindrical draft tube (10 mm in internal diameter, 500 mm long). In the end of the feeder, a
junction allows the change from the cylindrical symmetry of the feeder to the prismatic symmetry
existing in the roll gap. The roll press (0 to 50000 daN, load per unit length 0 to 104 daN.cm-1)
has been developed to record different major classical parameters : the roll speed, the roll gap, the
press strength, the rotation angle, and the feeding rate (from 0 to 80 L min-1 ).
In comparison with different kinds of roll press described in literature, original instrumentation
systems have been developed to catch specific data. The first one is the pressure distribution
pressure profiles (radial and axial pressures) at the interface between powder and the roll wall on
the roll width. Another one is the drive torque applied to the rolls. In addition a large size of
smooth steel- made roll (240 mm in diameter, 50 mm width) has been chosen to make easier the
extrapolation of the results to the industrial scale.

15
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

The results obtained with an organic compound exhibit the dependence –sometimes unexpected-
of the rotation angle, the feeding and the rotation speed on the pressure distribution, the roll
width, and the drive torque.
The France RhôneAlpes Region is kindly acknowledged for granting funds to this research
program.

COMPRESSION BEHAVIOU R OF AGGLOMERATES AN D THE EVOLUTION IN

TABLET STRUCTURE
Göran Alderborn
Department of Pharmacy, Uppsala University, Box 580, SE-751 23 Uppsala, Sweden
An insight into the factors that control the manufacturability of agglomerates is vital for a rational
approach to the problem of agglomerate engineering. In the presentation, the question of the
relationship between agglomerate properties and their compression and compaction behaviour is
addressed.
The dominating compression mechanisms for agglomerates seem to be permanent deformation,
densification and attrition (e.g. [1]). The degree of deformation and densification that
agglomerates undergo during the compression process are related to the composition of the
agglomerates (e.g. [2]) but also to their physical structure, such as their size, shape and porosity
(e.g. [3]). In addition, during compression of agglomerate mixtures, the relative strength of
agglomerates will affect their mode of deformation [4].
By the use of relationships between the applied stress during compression and the consequent
degree of compression of the agglomerate bed, an indicator of the compression mechanics of
agglomerates in terms of a failure strength can be derived [5]. The failure strength can be
modulated by the composition and the porosity of the agglomerates.

The response of agglomerates to the applied compressive stress will control the evolution in
physical structure of the compact (e.g. [6]). The structure will possibly control a number of
important pharmaceutical tablet properties, such as mechanical strength, disintegration and drug
release.
In conclusion, the compression behaviour and the evolution in tablet structure during confined
compression of agglomerates can be modulated and controlled by formulation factors and by
agglomeration method.

1. Johansson, B., Wikberg, M., Ek, R. and Alderborn, G. Compression behaviour and compactability of
microcrystalline cellulose pellets in relationship to their pore structure and mechanical properties. Int. J.
Pharm., 117, 57 (1995)
2. Nicklasson, F., Johansson, B. and Alderborn, G. Tabletting behaviour of pellets of a series of porosities - a
comparison between pellets of two different compositions. Eur. J. Pharm. Sci., 8, 11 (1999)
3. Johansson, B. and Alderborn, G. The effect of shape and porosity on the compression behaviour and tablet
forming ability of granular materials formed from microcrystalline cellulose. Eur. J. Pharm. Biopharm. 52,
347 (2001)
4. Tunón, Å. and Alderborn, G. Granule deformation and densification during compression of binary mixtures
of granules. Int. J. Pharm. 222, 65 (2001)

16
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

5. Adams, M.J., Mullier, M.A. and Seville, J.P.K. Agglomerate strength measurement using a uniaxial
confined compression test, Powder Technol. 78, 5 (1994)
6. Johansson, B., Nicklasson, F. and Alderborn, G. Effect of pellet size on degree of deformation and
densification during compression and on compactability of microcrystalline cellulose pellets. Int. J. Pharm.,
163, 35 (1998)

EFFECT OF PROCESS PARAMETERS ON MELT GRANULATION AND TABLET

PRESSING OF PHARMACEUTICAL MATERIALS
Gavin Walker
Queen’s University Belfast, School of Chemical Engineering, Belfast, UK
e-mail [email protected]
In this work, lactose was granulated in a fluidised bed dryer with polyethylene glycol 6000 used
as the binder. Granulation experiments were performed to assess the effect of granulation time
and binder content of the feed, on the resulting granule properties such as average granule size,
size distribution, granule yield stress, and granule porosity. These data were correlated using the
granule growth regime model. It was found that the dominant granule growth mechanisms in the
melt granulation of pharmaceutical powders were nucleation followed by steady growth.
However, with high binder content the granulation mechanism moved to the "over-wet massing"
regime in which discrete granule formation could not be obtained. Furthermore, the granules
from the granulation process were tablet pressed using commercial equipment. The physical
characteristics of the tablets were correlated with particle size, mechanical strength and PEG
content of the feed granules.

GRANULATION OF PHARMACEUTICAL EXCIPIENTS BY ROLL COMPACTION

G. Bindhumadhavan1, J. P. K. Seville1, M. J. Adams1, R. W. Greenwood1, S.
Fitzpatrick2
1- Centre for Formulation Engineering, The University of Birmingham, B15 2TT, U.K.
2- Merck Sharp & Dohme Ltd., Hertford Road, Hoddesdon, EN11 9BU, U.K.
e-mail [email protected]
Roll compaction is widely used in industry to produce free flowing agglomerates from a fine
particulate feed. Two of the main advantages of this process are that it is dry and continuous.
Despite being superficially a simp le process, a quantitative understanding has proved difficult to
develop because of the complex behaviour of particulate materials. Sub-optimal design and
operation of the equipment can lead to unsatisfactory products. Johanson (1965) developed a
theoretical model that enables the surface pressure, torque and separating force of the rolls to be
predicted from the physical characteristics of the powder and the dimensions of the rolls.
However, a detailed experimental validation of the theory has yet to be accomplished. The
current paper describes such a study using an instrumented roll press and a microcrystalline
cellulose powder. The measured pressure profiles in the nip region of the roll press were
comparable to the calculated values. The theory was also found to predict the effect of material
properties on the nip angle and the peak pressure but it was unable to account for the influence of
roll speed.

17
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

SESSION IV: GRANULE PROPERTIES

THE RELATIONSHIP BETWEEN SURFACE PROPERTIES AND BINDER
PERFORMANCE IN GRANULATION
S.J.R. Simons1, D. Rossetti1, P. Pagliai1, R. Ward2 and S. Fitzpatrick2
1- Colloid & Surface Engineering Group, Dept. Chemical Engineering, University College London, WC1 7JE, U.K.
2- Formulation Process & Design, Merck Sharp & Dohme, Hertford Road, Hoddesdon, EN11 9BU, U.K.
e-mail: [email protected]
The choice of the correct binder and wetting agent in a granulation process can critically
determine the performance and stability of the resultant granules. To date, a coherent basis for
making this choice has not been established, at least partly because the links between the surface
properties of the solids and the interaction with the binder is not fully understood. Clearly, the
surface energy of the solid particles will have an impact on the ability of binder solution to spread
across the surface of the particles. The same surface energy will also influence the strength of the
adhesion of the solids to the dry binder. Both factors will determine how well the particles are
bound into the granule – a key determinant of solids segregation and content uniformity
behaviour.

In this paper, the ability of a micromanipulator technique to differentiate between the interactions
of different binders (namely, HPMC, HPC and PVP) with drug particles is reported. Differences
between the powder-wetting characteristics of these binder solutions were apparent in spite of the
diversity of particle sizes and shapes employed. These differences are in line with those predicted
on the basis of surface energy calculations and granule properties.

BEHAVIOUR OF DRY BINDERLESS GANULES UNDE R IMPACT AND DIAMETRIC

COMPRESSION
Y.S. Cheong1, A.D. Salman1, A.F. Routh1, M.J. Adams2, C. Thornton2, D. K.
Kafui2 and M.J. Hounslow1
1
Particle Products Group, Department of Chemical and Process Engineering, University of Sheffield, Sheffield, UK.
2
School of Engineering, University of Birmingham, Birmingham, UK.
Wet high shear granulation is a size enlargement process with liquid addition to convert fine
powder to larger entities known as granules. However, disintegration and dispersion of these
granules during application may not be effective due to the extra strength caused by the presence
of binder at constituent particle contacts. The current investigation suggests that dry binderless
granules may be a better option as the interparticle bonding is due to the autoadhesive effect and
mechanical interlocking of the primary particles. Despite the weaker bonding, dry binderless
granules can withstand considerable stress before fracture, thus enabling them to survive
transportation. This work presents the mechanical behaviour of dry binderless granules,
composed of monodispersed polystyrene microspheres, under different loading rates, i.e. impact
and diametric compression. The method of producing dry granules with no binding agent using a
high shear granulator and their strength characterisation are highlighted. Besides that, the effect
of humidity on the surface interaction among the constituent particles of the granules is
investigated and the influence on the macroscopic response of the granules is examined. Discrete
element simulations are carried out to illustrate the evolution of damage mechanism during
impact and slow compression.

18
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

PARTICLE PROPERTIES AND ADHESIVE FORCES AS PARAMETER OF

AGGLOMERATION
Karl Sommer
Lehrstuhl für Maschinen- und Apparatekunde , TU München 85350, Freising Weihenstephan
e-mail: [email protected]
The properties of solids are determined by their chemical composition, the disperse state, and the
forces acting on the boundary surfaces. They are decisive for the behaviour of products in the
manufacturing process
With decreasing particle size the forces between the particles gain in importance as compared to
the particle‘s weight. Changing the adhesive forces can influence the product properties in the
process and the agglomeration qualities.
Of the adhesive forces, it is the capillary force that deserves particular attention in the
manufacture and application of solids. Capillary forces determine the pore structure of an
agglomerated solid and the external forces brought to play during the production process. For
example, if the interaction of these two effects on the pore structure is known, it is possible to
influence the strength of structural agglomerates and the dissolution rate of instant products.
A constantly acting adhesive force is the van der Waals force. The name ‘van der Waals ’ is well
known from our school or student days in the form of the so-called van der Waals equation of
gases. The reason of this is an intermolecular attractive force due to fluctuating dipoles. This
force is universal and is not identical to chemical bonds or electron bonds. The van der Waals
force acts not only in gases, but also between solids, for example between the graphite layers of
carbon
We all know from experience the phenomenon of electrostatic charge, which often leads to
undesirable adhesive forces especially in non-conductive materials. When particles of different
materials are brought into contact with one another, or are rubbed, electrons pass from one
material into the other. Because the partners become oppositely charged, they also attract one
another. In the case of very fine particles, electrostatic forces are an order of magnitude smaller
than the van der Waals forces, and are therefore generally negligible.
The significance of the electrostatic forces is not in their magnitude, but in the fact that they have
a much longe r range than van der Waals forces, and can thus influence the movement of
suspended particles. A classical example is the electrostatic separator, in which dust particles
made over a distance of several centimetres toward the precipitation electrode. In practice, in the
case of coarse particles, the distances between the particles are larger awing to surface
irregularities, and the adhesive forces are thus limited. Liquid changes the system in two ways: an
the one hand, "external“ forces are additionally generated by liquid menisci, while an the other
hand, the van der Waals interaction is directly influenced by the presence of another substance.
If a liquid is present at the contact points, capillary forces act as a result of surface tension.
Reduced pressure due to the capillary effect leads to adhesion between the two particles. The
result is adhesive forces that are larger than the van der Waals forces by a factor of 5. On the
other hand, the actual van der Waals interaction decreases in the presence of water by a factor of
up to 100, and even by a factor of up to 5,000 with surfactants. The knowledge of the physical
understanding of the adhesion forces as the bas of the agglomeration leads to practical
consequences.

19
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

AN EXPERIMENTAL STUDY OF THE IMPACT BREAKAGE OF SINGLE WET

GRANULES
J.S. Fu1, A.D. Salman1, G.K. Reynolds1, M.J. Adams2 and M.J. Hounslow1
1
Department of Chemical and Process Engineering, University of Sheffield, Sheffield, UK
2
Centre for Formulation Engineering, Chemical Engineering, University of Birmingham, UK
The behaviour of wet single granules under impact loading has been investigated. They were
prepared in a high-shear mixer granulator using calcium carbonate particles having different
mean sizes and a range of phase volumes of a liquid binder, which was either polyethylene glycol
or mixtures of glycerol and water to obtain a range of viscosities. It was found that the porosity
decreased with increasing granulation time for values less than some critical time; at greater
times the porosity then remained constant. A critical impact velocity was defined as the value at
which a wet granule appeared visibly cracked. This quantity increased monotonically with
increasing granulation time even at times greater than that corresponding to the asymptotic
porosity. The critical impact velocity also decreased with increasing granule and primary particle
size, exhibited a minimum value with increasing binder content and increased with increasing
binder viscosity. The extent of the damage increased with increasing impact velocity following
the general sequence: plastic deformation, multiple crack propagation from the plastically
deformed contact region, conical-shaped debris on the target with the rebound of a cracked but
coherent cap and, at greater velocities, the cap was fragmented. This behaviour corresponded to
a decrease in the survival ratio and the mean fragment size with increasing impact velocity.
Moreover, the size of the conical debris increased with increasing impact velocity and increasing
primary particle size. In summary, an extremely detailed characterisation of the impact damage
of wet granules and the dependence on their composition, porosity and loading conditions has
been obtained. This should provide an improved basis for understanding the factors that control
size evolution and product quality in granulation processes.

SINGLE AND BULK COMPRESSIONS OF SOFT GRANULES: DEM SIMULATION

AND EXPERIMENTAL EVALUATION
A. Samimi, A. Hassanpour and M. Ghadiri
Institute for Particle Science and Engineering, University of Leeds, Leeds, UK
In a number of powder processing operations such as compaction and tabletting, the behaviour of
bulk powders subjected to compression is of great interest. Bulk compression of particles has
been considered as a useful experimental method to characterise some properties of individual
particulate solids being compressed in the bed. A large number of correlations have been
developed mainly based on the experimental fittings of data to relate the applied pressure on the
bed to some state of consolidation measures of powders in the bed such as bed volume, strain and
relative density. Heckel’s correlation (Heckel 1961) is one of these models and is used to infer
the yield stress of individual particles based on the density-pressure relationship of the bed. In
spite of the extent of publications using Heckel’s model, the level of understanding of the
compression mechanism as well as the identity of the parameter that is essentially characterised
by the model si poor. Furthermore, there is not much detailed work investigating the bulk
compression of soft granules. The most appropriate approach to evaluate this model is the
combined experimental and computer simulation techniques.

20
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

This paper addresses the deformation and failure behaviour of detergent based soft granules using
uni-axial single and bulk compression methods. In particular the mechanical properties of single
granule such as yield strength and Young’s modulus obtained based on single particle
compression tests are used in Distinct Element Method (DEM) to simulate the bulk compression
behaviour. The DEM simulations are then compared with experimental results of the bulk
compression. Furthermore the identity of the Heckel parameter obtained based on the
experimental and simulation is discussed.

TOWARDS A DESKTOP ATTRITION TESTER FOR SIMULATING PNEUMATIC

CONVEYING
G.M.H. Meesters, B. van Laarhoven and S. Schaafsma
Delft University of Technology and DSM-Research
In industry solid products are sold more and more in the form of granules. Agglomerated
particles in a granule form eliminate transport problems caused by sticky, hygroscopic, dusty and
explosive powders. During transport of the granules they are subjected to both static and dynamic
loads. The challenge is to design granules resistant to damage induced during the transport.
Beekman [1] designed and developed a new test method to asses attrition tendency of small
samples of granules. Pitchumani [2], improved the design to the current Crank Slider Repeated
Impact Tester (henceforth CS-RIT).
In this work a comparison is made between attrition of sodium benzoate granules by pneumatic
transport and attrition by CS-RIT. Sodium benzoate granules in the size range of 710 µm –1000
µm are brought in the particle chamber of the CS-RIT. The particle chamber is subjected to a
linear oscillatory motion resulting in two collisions each cycle. By fixing the frequency and the
amplitude of the linear oscillatory motion a fixed impact velocity is obtained. Attrition on sodium
benzoate granules as function of the number of collisions is correlated to attrition induced by a
pneumatic transport setup. First results will be presented.

2. W.J. Beekman, G.M.H. Meesters, B. Scarlett and T. becker. Measurement of granule attrition and fatigue in
a vibrating box. Particle and Particle Systems Characterisation, 19 5-11, 2001.
3. R. Pitchumani. Breakage characteristics of particles and granules. Thesis, Delft University of Technology.
2003.

COMPUTER-AIDED DESIGN OF GRAN ULE MICROSTRUCTURE

Frantisek Stepanek
Department of Chemical Engineering and Chemical Technology, Imperial College London, UK
e-mail: [email protected]
The properties of multi-component granules depend not only on granule size and composition,
but also on the spatial distribution of components within the granule. Some properties – such as
the rate of release of an active component during dissolution – can be effectively controlled by
the granule structure. Granule structure follows from the granulation process, but it also depends
on the formulation and on the specification of the raw materials (e.g., primary particle size
distribution). Microstructure design thus brings together formulation design and process design.
In this contribution we will present computational methods for “virtual granulation” which enable

21
 Second International Workshop on Granulation
22 – 25 June 2004, University of Sheffield

the construction of realistic three-dimensional models of granules formed by the agglomeration

of primary particles of defined size, shape and surface properties, with binder droplets of defined
spreading and solidification rate. As a post-processing step, a software module for “virtual
dissolution” enables the release profiles of individual components from each designed granule to
be determined. The overall design methodology, which involves finding the inverse of
computationally determined structure and property functions, will also be presented.

22