European Journal of Pharmaceutical Sciences 115 (2018) 223–232

Contents lists available at ScienceDirect

European Journal of Pharmaceutical Sciences

journal homepage: www.elsevier.com/locate/ejps

Breakage and drying behaviour of granules in a continuous fluid bed dryer: T

Influence of process parameters and wet granule transfer
F. De Leersnydera,1, V. Vanhoorneb,1, H. Bekaertb, J. Vercruysseb, M. Ghijsc, N. Bostijna,
M. Verstraetena, P. Cappuynsd, I. Van Assched, Y. Vander Heydene, E. Ziemonsf, J.P. Remonb,
⁎
I. Nopensc, C. Vervaetb, T. De Beera,
a
Laboratory of Pharmaceutical Process Analytical Technology, Ghent University, Belgium
b
Laboratory of Pharmaceutical Technology, Ghent University, Belgium
c
BIOMATH, Department of Mathematical Modelling, Statistics and Bioinformatics, Ghent University, Belgium
d
Department of Pharmaceutical Development, Johnson & Johnson Pharmaceutical Research and Development, Janssen Pharmaceutica, Belgium
e
Department of Analytical Chemistry and Pharmaceutical Technology, VUB, Belgium
f
CIRM, Laboratory of Analytical Chemistry, University of Liege, Belgium

A R T I C L E I N F O A B S T R A C T

Keywords: Although twin screw granulation has already been widely studied in recent years, only few studies addressed the
Continuous manufacturing subsequent continuous drying which is required after wet granulation and still suffers from a lack of detailed
Fluid bed drying understanding. The latter is important for optimisation and control and, hence, a cost-effective practical im-
Critical quality attributes plementation. Therefore, the aim of the current study is to increase understanding of the drying kinetics and the
Twin screw granulation
breakage and attrition phenomena during fluid bed drying after continuous twin screw granulation. Experiments
Process understanding
Particle size distribution
were performed on a continuous manufacturing line consisting of a twin-screw granulator, a six-segmented fluid
Moisture content bed dryer, a mill, a lubricant blender and a tablet press. Granulation parameters were fixed in order to only
examine the effect of drying parameters (filling time, drying time, air flow, drying air temperature) on the size
distribution and moisture content of granules (both of the entire granulate and of size fractions). The wet
granules were transferred either gravimetrically or pneumatically from the granulator exit to the fluid bed dryer.
After a certain drying time, the moisture content reached an equilibrium. This drying time was found to depend
on the applied airflow, drying air temperature and filling time. The moisture content of the granules decreased
with an increasing drying time, airflow and drying temperature. Although smaller granules dried faster, the
multimodal particle size distribution of the granules did not compromise uniform drying of the granules when
the target moisture content was achieved. Extensive breakage of granules was observed during drying. Especially
wet granules were prone to breakage and attrition during pneumatic transport, either in the wet transfer line or
in the dry transfer line. Breakage and attrition of granules during transport and drying should be anticipated
early on during process and formulation development by performing integrated experiments on the granulator,
dryer and mill.

1. Introduction However, it is essential to gain fundamental knowledge about the me-

Traditionally, batch processes are used to produce pharmaceutical
products. However, the pharma industry is willing to invest in con-
tinuous processes since many benefits are associated with continuous
manufacturing (e.g. flexibility, smaller equipment, cost, time-to-market
reduction and improved product quality) (FDA, 2004; Plumb, 2005;
Teżyk et al., 2015; Vercruysse et al., 2013; Vervaet and Remon, 2005).
chanisms occurring during continuous processes to allow process opti-
misation and control for reliable real-time release of produced products.
Tablets are the most commonly used solid dosage forms in the
pharmaceutical industry (Gohel and Jogani, 2005). Direct compression
is the preferred continuous manufacturing method for production of
tablets, but often an intermediate agglomeration step is required to
improve the flowability, homogeneity and compressibility (Armstrong,

Abbreviations: L/S, liquid-to-solid; LOD, loss-on-drying; PBM, Population Balance Model; V0, bulk volume; V500, tapped volume after 500 taps; ρb, bulk density; ρt, tapped density after
500 taps
⁎
Corresponding author at: Ghent University, Laboratory of Pharmaceutical Process Analytical Technology, Ottergemsesteenweg 460, 9000 Ghent, Belgium.
E-mail address: Thomas.debeer@ugent.be (T. De Beer).
1
Shared first authorship.

https://doi.org/10.1016/j.ejps.2018.01.037
Received 28 September 2017; Received in revised form 12 December 2017; Accepted 20 January 2018
0928-0987/ © 2018 Elsevier B.V. All rights reserved.
F. De Leersnyder et al.
Fig. 1. Horizontal ConsiGma™-25 system (left) and vertical ConsiGma™-25 system (right) with 1. loss-in-weight feeder, 2. twin screw granulator, 3. wet transfer line (only in horizontal
set-up), 4. six-segmented fluid bed dryer, 5. dry transfer line and 6. product control unit with option for milling.
Fig. 2. Visual presentation of the filling cycle in the six-segmented fluid bed dryer of a ConsiGma™-25 line.
Table 1
Overview of the granulation parameters.
Screw speed (rpm) 900
Barrel temperature (°C) 25
Throughput (kg/h) 20
L/S ratio 0.15
2007; Gohel and Jogani, 2005; Vervaet and Remon, 2005).
Several equipment manufacturers currently offer integrated from-
powder-to-tablet lines for continuous production of tablets via con-
tinuous twin screw granulation (e.g. QbCon® by L.B. Böhle,
Granuformer® by Freund Vector, ConsiGma™ by GEA Pharma Systems,
MODCOS system by Glatt). Whereas in the Granuformer® line (Freund
Vector) a spiral dryer is implemented, the other continuous lines
pneumatically transfer the wet granules from the granulator to a seg-
mented fluid bed dryer (Kleinebudde, 2017; Byrn et al., 2014). How-
ever, on the ConsiGma™ system (GEA Pharma systems) gravimetric
instead of pneumatic granule transfer is also possible. Therefore two
different set-ups can be distinguished: (1) a horizontal set-up (granu-
lator is positioned next to the dryer) with pneumatic granule transfer
via a wet transfer line, (2) a vertical set-up (granulator is positioned
above the dryer) with gravimetric transfer of wet granules to the dryer
(Fig. 1).
Segmented fluid bed dryers are currently most often implemented in
continuous manufacturing lines. Although other designs of continuous
fluid bed dryers are available such as horizontal fluid bed dryers (e.g.
GF series by Glatt, Niro Contipharm granulator, Heinen drying tech-
nologies) or the AGT fluid bed dryer by Glatt, these systems are typi-
cally not applicable in the pharmaceutical industry due to the long
material residence time and lack of plug-flow (Gotthardt et al., 1999;
Vervaet and Remon, 2009; Teżyk et al., 2015). Short and controllable
residence times are of utmost importance during granulation of phar-
maceutical products for traceability and to avoid product degradation.
224
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
While twin screw granulation was widely studied in recent years,
only few studies addressed subsequent drying on a segmented fluid bed
dryer which is required after wet granulation. Fonteyne et al. and
Chablani et al. reported on the implementation of Raman and NIR
probes in the six-segmented fluid bed dryer unit of the ConsiGma™-25
for monitoring of moisture content and polymorphism while Vercruysse
et al. evaluated the stability and repeatability of the granulation and
drying units during long runs with constant process parameters
(Chablani et al., 2011; Fonteyne et al., 2014a; Vercruysse et al., 2013).
However, these studies did not focus on the influence of dryer para-
meters on critical quality attributes of granules and did not aim to ex-
plore and understand breakage and attrition phenomena taking place
during drying.
Although batch-wise fluid bed drying has been intensively studied
and the influence of most parameters (air flow, drying air temperature
and drying time) on the granule quality is expected to be similar during
continuous fluid bed drying, the set-up of a batch fluid bed dryer (with
a single drying cell) is fundamentally different in comparison to a
segmented fluid bed dryer (with multiple drying cells). Whereas an
entire batch of wet granules is introduced at the same moment into a
batch fluid bed dryer, granules are continuously transferred to a specific
cell of the segmented fluid bed dryer for a set time (filling time) after
which the next cell of the segmented fluid bed dryer is filled. Therefore,
the filling time of a dryer cell is an additional variable inherent to
drying in a segmented fluid bed dryer after continuous granulation.
However, the effect of the filling time has not yet been investigated.
Therefore, the effect of filling time, in addition to other drying para-
meters (air flow, drying air temperature, drying time), was evaluated in
the current study.
The impact of the transfer of wet granules after continuous granu-
lation was also neglected up to now. Using an in-line particle imaging
tool, Kumar et al. observed that tray dried granules showed wider
granule size distributions in comparison to fluid bed dried granules
which was attributed to breakage and attrition in the segmented fluid
F. De Leersnyder et al. European Journal of Pharmaceutical Sciences 115 (2018) 223–232

Table 2 suboptimal dryer settings or due to pneumatic granule transfer, in the

Overview of performed experiments (*: no granules could be collected on vertical set-up). current study the effect of drying parameters and granule transfer
(gravimetric or pneumatic) on the granule quality was evaluated.
Experimental Drying air Airflow Filling Drying
conditions temperature (°C) (m3/h) time (s) time (s) Efforts were made by Mortier et al. to gain fundamental under-
standing of the physics determining the drying behaviour of granules in
A 40 360 60 80 the six-segmented fluid bed dryer via application of mass-energy bal-
100
ances and mechanistic models (Mortier et al., 2012, 2013, 2014). Cal-
200
300 culating mass and energy balances, Mortier et al. could predict the
324 moisture content of granules, produced with different granulator and
B 60 360 60 80 dryer settings, based on the in-line data logged by the ConsiGma™-25
100 line (Mortier et al., 2014). Although mass and energy balances proved
200
to be powerful to monitor the overall granule moisture content in-line,
300
324 obtaining detailed insight of the drying process and kinetics at granule
C 40 440 60 80 level proved to be a tedious endeavor (Mortier et al., 2014). Therefore,
100 a mechanistic model was developed by Mortier et al. describing the
200
drying of a single granule in a dryer cell of the ConsiGma™-25 system
300
324 (Mortier et al., 2012). This model was calibrated and validated using
D 60 440 60 80 experimental data for a pharmaceutical formulation and allowed to
100 gain insight in the drying behaviour of a single granule. In a next step,
200 this model was rewritten as a Population Balance Model (PBM) to si-
300
mulate the moisture content of a population, or batch, of granules
324
E* and J* 50 400 120 140 (Mortier et al., 2013). In this PBM, the evolution of the moisture con-
200 tent of granules with different initial moisture contents during drying
300 can be described at the same time. This generates a distribution of
400
granules during the drying process in function of their current moisture
679
F 40 360 180 250*
content, which is the output of the model. Another feature of the PBM is
300 that the effect of the continuous filling of the dryer cell can be in-
400 corporated, which implies that different granules are not at the same
700 stage in the drying process (Mortier et al., 2012).The aforementioned
1034
PBM has finally been extended by Mortier et al. to describe the evo-
G 60 360 180 200*
250 lutions of particle size distribution and moisture content distribution of
300 the granules simultaneously during drying (Mortier et al., 2013). These
400 PBM models still need to be calibrated and validated to experimental
700
moisture content distribution data. However, moisture content dis-
1034
H 40 440 180 250*
tribution data is exhaustive to gather. Because of this, the model for
300 single granule drying by Mortier et al. is being focused on for calibra-
400 tion and validation to different formulations (Mortier et al., 2012).
700 Therefore, in the current study, experiments were setup to better
1034
understand the effect of the drying parameters (filling time, drying
I 60 440 180 200*
250 time, drying air temperature and air flow) and wet granule transfer on
300 the critical quality attributes of granules (particle size and moisture
400 content). Moisture content and granule size distribution are critical
700 quality attributes of granules as they determine the flowability, segre-
1034
gation tendency and stability of the granules (Crouter and Briens, 2014;
El Hagrasy et al., 2013; Emery et al., 2009). Additionally, the relation
bed dryer or during pneumatic granule transfer from the twin screw between particle size and moisture content was investigated in the
granulator to the dryer (Kumar et al., 2015). As the study of Kumar current study by determination of the moisture content of different
et al. could not differentiate between breakage and attrition due to sieve fractions in function of drying time at different drying conditions.

Fig. 3. Total moisture content of the granules produced on the horizontal ConsiGma™-25 set-up in function of drying for filling times of 60 s (left) and 180 s (right).

225
F. De Leersnyder et al.
Fig. 4. Total moisture content of the granules produced on the horizontal set-up in
function of drying time for the duplicate experiments at intermediate drying conditions.
The results of these experiments can in the future be used to apply the
mechanistic model for single granule drying of Mortier et al. for si-
mulation of the moisture content evolution for different granule sizes in
function of drying time and drying conditions (Mortier et al., 2012).
Fig. 5. Moisture content of the granules produced on horizontal set-up in function of drying time, drying air temperature and airflow with blue and red circles representing short (60 s)
and long (180 s) filling times for different granule size fractions respectively (smaller and larger size fractions are represented by smaller and larger circles, respectively). (For inter-
pretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
226
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
2. Materials and methods
2.1. Materials
The formulation consisted of two active ingredients, maize starch
(Cargill, Vilvoorde, Belgium) and powdered cellulose (Solka-Floc,
International Fiber Corporation, Ohio, USA) were used as fillers and
pregelatinised starch (National Starch & Chemical company,
Westchester, USA) and sodium starch glycolate (Roquette, Lestrem,
France) were used as binder and desintegrant, respectively.
2.2. Methods
2.2.1. Description of the ConsiGma™-25 system
Granulation and drying experiments were performed on a
ConsiGma™-25 system (GEA Pharma Systems, Wommelgem, Belgium).
This continuous line consists of a twin screw granulator, a six-seg-
mented fluid bed dryer, a granule conditioning unit with a mill, a
continuous blender for addition of for instance a lubricant such as
magnesium stearate and a tablet press (latter not shown in Fig. 1). As
fluid bed drying is not an inherently continuous process, the fluid bed
dryer unit of the ConsiGma™-25 system consists of six cells, which are
consecutively fed with wet granules produced by twin screw granula-
tion. After filling a dryer cell for a set filling time, granules are fed to the
next cell, while the previous cells complete their drying cycle. De-
pending on the applied filling and drying times, two to six cells can
contain granules simultaneously. This process is schematically pre-
sented in Fig. 2. It is clear that once filling and drying of the granules in
cell 4 starts, cell 1, 2 and 3 still contain granules in their drying phase.
F. De Leersnyder et al.
Fig. 6. Fraction of fines, yield and oversized for experiments A–D (filling time 60 s) performed on the horizontal set-up, in function of drying time. Drying time 0: granules directly after
wet granulation.
The design of the dryer unit implies that after the drying operation in
the ConsiGma™-25 system, mini-batches are created, which are suc-
cessively transported to the granule conditioning unit, lubricant blender
and the tablet press.
Depending on the position of the six-segmented fluid bed dryer and
granulator, two different set-ups of the ConsiGma™-25 can be dis-
tinguished: (1) a horizontal set-up (the granulator is positioned next to
the dryer) including pneumatic transport of wet granules from the exit
of the granulator to the dryer via a wet transfer line and (2) a vertical
set-up (the granulator is positioned above the dryer) which comprises
gravimetric addition of wet granules to the dryer.
2.2.2. Preparations of granules
A preblend of the formulation was prepared in a tumbling blender
(Inversina-Bioengineering, Wald, Switzerland) during 15 min at 25 rpm
and transferred to the loss-in-weight feeder (KT20, K-Tron Soder,
Niederlenz, Switzerland) of the ConsiGma™-25 system (GEA Pharma
Systems, Wommelgem, Belgium).
The barrel of the twin screw granulator (length-to-diameter ratio:
20:1) can be divided into a feed segment consisting of conveying ele-
ments and a work segment where the powder is intensively mixed with
granulation liquid by kneading elements. The granulation liquid was
pumped into the barrel just before the first kneading element via a
double liquid addition port, injecting granulation liquid on top of each
screw. The equipment has a built-in torque gauge which monitors the
torque with a frequency of 1 Hz. The applied granulation parameters for
current formulation were optimised by Vercruysse et al. and are listed
in Table 1 and were kept constant during all runs (Vercruysse et al.,
2013). The screw configuration was composed of 2 kneading zones each
227
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
consisting of 4 kneading elements (L = D/4 for each kneading element)
at an angle of 60 degrees. Both kneading zones were separated by a
conveying element (L = 1.5D). An extra conveying element (L = 1.5D)
was implemented after the second kneading block together with 2
narrow kneading elements (L = D/6 for each kneading element) in
order to reduce the amount of oversized agglomerates, as reported by
Van Melkebeke et al. (2008).
After granulation, the wet granules were transferred pneumatically
by a wet transfer line to the six-segmented fluid bed dryer unit of the
horizontal ConsiGma™-25 set-up (wet transfer line indicated in red in
Fig. 1) or transferred gravimetrically to the six-segmented fluid bed
dryer of the vertical ConsiGma™-25 set-up (Fig. 1).
After drying the granules were pneumatically transferred to the
product control unit via the dry transfer line (indicated in green in
Fig. 1) and were collected at the exit of the product control unit. In the
current study no mill was implemented in the product control unit.
Additionally, wet granules were collected at the exit of the granulator
and tray dried in an oven during 24 h at 40 °C prior to granule char-
acterization.
The filling time, drying time, air flow and drying air temperature
were varied during the experiments on the horizontal and vertical
ConsiGma™-25 set-up. An overview of the experiments is presented in
Table 2. Filling times of 60 (A–D) and 180 s (F–I) per cell were used. At
least four different drying times were applied, varying between the
shortest and longest drying time applicable at a set filling time. The
minimal drying time needed to discharge the wet granules from the
dryer unit, depended on the drying parameters and was set as shortest
drying time. For example in Table 2, it is indicated that experiments F
and H could not be performed with a drying time of 200 s, while
F. De Leersnyder et al. European Journal of Pharmaceutical Sciences 115 (2018) 223–232

Fig. 7. Fraction of fines, yield and oversized for experiments F–I (filling time 180 s) performed on the horizontal set-up, in function of drying time. Drying time 0: granules directly after
wet granulation (no granules could be collected after 200 s of drying time for experiments F and H).

Fig. 8. Fraction of fines, yield and oversized for experiment E and J (replicates) on the horizontal set-up, in function of drying time. Drying time 0: granules directly after wet granulation.

experiment G and I could be performed. This could be attributed to
application of a higher drying air temperature resulting in faster drying
and therefore less sticky granules. The longest drying time corre-
sponded to six times the applied filling time minus the time needed for
cell discharge. All experiments were performed at low (360 m3/h) or
high (440 m3/h) airflow in combination with a low (40 °C) or high
(60 °C) drying air temperature (Table 2). Additionally, experiments on
the horizontal ConsiGma™-25 set-up were performed in duplicate (E
and J) at intermediate conditions (120 s filling time, 50 °C drying air
temperature and 400 m3/h intermediate airflow) to estimate the
process repeatability. No granules could be collected on the vertical
ConsiGma™-25 set-up when applying drying times of 200 s in experi-
ments G and I and 250 s in experiments F and H because a blockage in
the dry transfer line (i.e. the transfer line transporting the granules from
the dryer to the product control unit) occurred. The granules were too
wet and sticky and larger than when they were discharged from the
dryer unit in the horizontal set-up (discussed in Sections 3.2.1 and
3.2.2).

228
F. De Leersnyder et al.
Fig. 9. Fraction of fines, yield and oversized for experiments A–D (filling time 60 s) on the vertical set-up, in function of drying time. Drying time 0: granules directly after wet
granulation.
2.2.3. Characterization granules
2.2.3.1. Determination of the overall moisture content and moisture content
of different size fractions. The residual moisture content of granules
collected after the product control unit was determined via loss-on-
drying (LOD) using a moisture analyser (Mettler LP16, Mettler-Toledo,
Zaventem, Belgium) including an infrared dryer and a balance. A
sample of 1 g was dried at 105 °C until the weight was constant for 30 s.
Additionally, the granules produced on both the horizontal and the
vertical set-up were sieved on a series of sieves (150, 300, 500, 850,
1000, 2000 μm) and the moisture content of each sieve fraction was
determined. The target moisture content for the formulation was 1.5%
measured by LOD.
2.2.3.2. Particle size analysis. The granule size of the tray dried
granules and of the granules collected after the product control unit
(after fluid bed drying) was analysed via dynamic image analysis using
the QICPIC™ system (Sympatec, Clausthal-Zellerfeld, Germany)
equipped with a vibrating feeder system (Vibri/L™) for gravimetrical
feeding of the granules. Samples of 20 g were measured in duplicate and
the averages of volume granule size distributions were reported,
WINDOX 5 software (Sympatec, Clausthal-Zellerfeld, Germany) was
used to calculate the granule size distributions. The amounts of fines
and oversized granules were defined as the fractions < 150 μm
and > 1000 μm, respectively. The yield of the process was defined as
the percentage of granules between 150 and 1000 μm.
2.2.3.3. Flowability testing. The flowability of the granules collected
after the product control unit was evaluated by the Hausner ratio. The
Hausner ratio was calculated from the bulk and tapped densities of the
229
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
granules. The bulk volume (V0) of 50 g granules was measured in a
250 ml graduated cylinder as well as the tapped volume after 500 taps
(V500) using a tapping machine (J. Englesman, Ludwigshafen,
Germany). Experiments were performed in duplicate. Bulk and
tapped densities were calculated as 50 g/V0 and 50 g/V500,
respectively. The Hausner ratio was calculated from the bulk (ρb) and
tapped (ρt) densities using the following equation:
ρt
HR =
ρb
3. Results and discussion
3.1. Moisture content
The total moisture content of the granules in function of drying time
was evaluated to gain better understanding of the drying kinetics. The
results were similar on the horizontal and vertical set-ups (data of
vertical set-up not shown). After a certain drying time, the moisture
content did not further decrease. Depending on the air flow and drying
air temperature, a constant moisture was reached after a specific drying
time. This is illustrated in Fig. 3 for the experiments with a filling time
of 60 (A–D) and 180 s (F–I). A constant moisture content was reached
after 200–300 and 400–700 s, respectively.
Intense drying conditions (D, I: high airflow and high drying air
temperature) resulted in faster drying of granules whereas milder
drying conditions (A, F: low airflow and low drying air temperature)
resulted in slower drying of the granules (Fig. 3). Consequently, the
moisture content after the longest possible drying time (324 and 1034 s
F. De Leersnyder et al.
Fig. 10. Fraction of fines, yield and oversized for experiments F–I (filling time 180 s) on the vertical set-up, in function of drying time. Drying time 0: granules directly after wet
granulation.
after filling times of 60 and 180 s, respectively) was slightly lower if
intense drying conditions were applied in comparison to milder drying
conditions. The moisture content results of replicate experiments E and
J (at intermediate airflow, drying air temperature and filling time on
horizontal set-up) are shown in Fig. 4 and demonstrate that the results
were highly repeatable.
Next, the evolution of the moisture content of the granules was
evaluated in function of granule size and drying time (Fig. 5). The re-
sults of the size fraction > 2000 μm were excluded as the moisture
content determination by LOD of 1 g of granules was not reproducible
since 1 g of granules often consisted of only 2–4 granules. Fig. 5 shows
that smaller size fractions of the granules, and especially the fines
fraction, dried faster than the larger size fractions.
During the initial drying phase, water from the surface of the
granules is evaporated and the surface/volume ratio of the granules in
the smaller size fractions is larger than of those in the larger size
fractions, resulting in faster drying of the former. Similar observations
were reported by Fonteyne et al. studying the end-point determination
of the fluid bed drying process (Fonteyne et al., 2014b). However, the
differences in moisture content between the different size fractions
disappeared at longer drying times. Additionally, the differences in
moisture content of the different size fractions disappeared faster when
applying intense drying conditions (D, I: high airflow and drying air
temperature) in comparison to milder drying conditions (A, F: low
airflow and drying air temperature). However, for all drying conditions,
there was no relevant difference in the moisture content of the different
size fractions once the target moisture content of 1.5% was approached.
This signifies that the multimodal particle size distribution of the
granules or the difference in drying time between granules filled into
230
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
the same cell (at a filling time of 180 s, the first granule entering a dryer
cell is dried 180 s longer in comparison to the last granule entering the
dryer cell) do not compromise uniform drying of the granules if suffi-
ciently large drying times are applied.
3.2. Particle size
In this section, the granule size distributions were evaluated in
function of the applied filling time, drying time, drying air temperature
and air flow to study breakage and attrition phenomena during fluid
bed drying. The granule size distributions of tray dried granules col-
lected at the outlet of the granulator were also determined and are
included in the graphs (Figs. 6–10) as the granule size distributions
after 0 s of drying time. It was decided to represent the granule size
distributions by the fines (< 150 μm) fraction, yield (150–1000 μm)
and oversized fraction (> 1000 μm) to make the interpretation less
complex and more relevant from a pharmaceutical point of view.
However, conclusions made were also valid when more size classes
were taken into account.
3.2.1. Horizontal set-up
Drying of the granules in the horizontal set-up resulted in extensive
breakage and attrition of the granules (i.e. loss in oversized and gain in
yield and fines, respectively). However, breakage and attrition of
granules exclusively occurred during the first drying period of 0–80 s
for filling times of 60 s (Fig. 6) and of 0–200 s for filling times of 180 s
(Fig. 7). When comparing the granule size distributions of the granules
obtained after longer drying times, no relevant differences in granule
size distributions could be detected.
F. De Leersnyder et al.
It could be concluded that granules are most prone to breakage
during the first seconds of drying. However, between the exit of the
granulator (0 s) and discharge from the dryer after the shortest drying
time (80 s drying time after a filling time of 60 s and 200 s drying time
after a filling time of 180 s) not only drying of the granules occurred but
also pneumatic transport of granules from the exit of the granulator to
the inlet of the dryer via a wet transfer line took place. During this
transport, wet granules could break up by air friction, interparticular
collisions and collisions with the wet transfer line. As the granules were
wet, solid bonds between the particles were not yet formed.
Additionally, exiting the wet transfer line, the granules entered the
dryer at high velocity and could consequently collide with the wall or
bottom of the dryer unit with high impact.
Comparison of the granule size distributions of experiments A and B
to C and D and comparison of experiments F and G to H and I for filling
times of 60 and 180 s, respectively, showed that application of higher
airflow resulted in slightly less oversized granules and a slightly higher
yield (Figs. 6 and 7). This could be attributed to more interparticle
collisions and more collisions between the particles and the stainless
steel dryer wall when a higher airflow was applied. In contrast, drying
air temperature did not affect the granule size distributions.
The granule size distribution of the granules produced in duplicate
(experiments E and J) were compared to ensure the repeatability of the
experimental set-up. The experiments were highly repeatable (Fig. 8).
3.2.2. Vertical set-up
The granulator settings applied on the horizontal and vertical set-up
of the continuous line were identical to allow direct comparison of both
set-ups and indeed tray dried granules produced on both set-ups
showed identical granule size distributions.
Similarly as on the horizontal set-up, breakage and attrition of the
granules was observed comparing the tray dried granules collected at
the granulator outlet and the fluid bed dried granules (Figs. 9 and 10).
However, the extent of breakage and attrition was much smaller for the
granules dried on the vertical set-up in comparison to the horizontal
set-up.
Whereas fluid bed drying of the granules on the horizontal set-up
resulted in a 40–65% decrease in oversized granules, a 50–60% increase
in yield and a 4–8% increase in fines, fluid bed drying of the granules on
the vertical set-up resulted in a 20–50% decrease in oversized granules,
a 20–40% increase in yield and 1–6% increase in fines. As clearly less
breakage and attrition of granules occurred during processing on the
vertical set-up, the more extensive breakage and attrition observed
during processing on the horizontal set-up was mainly attributed to the
pneumatic transfer of the wet granules in the wet transfer line.
However, as granule breakage and attrition also occurred on the ver-
tical set-up (although to a lesser extent), breakage and attrition was also
caused by the fluid bed drying process itself (i.e. collisions of the
granules with the dryer wall and interparticular collisions).
As a result of less breakage occurring during fluid bed drying on the
vertical set-up, the yield percentages obtained on the vertical set-up
were lower in comparison to the yield obtained on the horizontal set-up
as more oversized granules remained after drying. This signifies that the
granulation parameters on the vertical set-up should be optimised to
decrease the oversized fraction of granules formed during granulation
and consequently to increase the yield. Initially, the granulation para-
meters were set based on preliminary experiments on the horizontal set-
up anticipating extensive breakage and attrition during fluid bed
drying. However, as less breakage and attrition occurred on the vertical
set-up, the granulation parameters could be optimised or a mill could be
implemented after the drying step to increase the yield on the vertical
set-up.
Interestingly, drying time affected the obtained granule size dis-
tributions after fluid bed drying on the vertical set-up. Applying short
drying times (80–100 s for A–D and 300–400 s for F–I) resulted in more
breakage and attrition of the granules in comparison to longer drying
231
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
times. Although this seems contradictory at first instance, this can be
explained considering the moisture content of granules. Shortly dried
granules are still wet and therefore less resistant to pneumatic transport
in the dry transfer line (i.e. transfer line between the fluid bed dryer and
the product control unit) as less solid bonds can be formed by crystal-
lisation of solubilised material. Therefore, this phenomenon occurring
in the dry transfer line of the vertical set-up is comparable to the
breakage and attrition of wet granules in the wet transfer line of the
horizontal set-up. It can be concluded that a critical drying time on the
vertical set-up is necessary for the granules to acquire sufficient
strength by formation of strong solid bonds. Further drying after the
critical drying time, did not affect the granule size distributions (Figs. 9
and 10). The effect of drying time on breakage and attrition was less
pronounced for the experiments with long filling times (F–I). This could
be attributed to the dryer concept of the segmented dryer as the filling
time is included in the drying time. Application of 180 s filling time and
200 s drying time signifies that the first granule entering the cell was
dried during 200 s whereas the last granule added to the cell was dried
during only 20 s.
Similar to the horizontal set-up, the drying temperature did not
influence the granule size distributions. Additionally, the airflow did
not influence the granule size distributions although on the horizontal
set-up a minor, non-relevant, effect of the airflow was observed.
In conclusion, breakage and attrition mainly occurred when wet
granules were exposed to higher shear force caused by pneumatic
transport in the wet and/or dry transfer line. Integrated optimisation
experiments varying granulation, drying and milling parameters in the
same experimental design are recommended to obtain the highest yield
prior to tableting.
3.3. Flowability
The flowability of the granules was assessed by their Hausner ratio.
Granules produced on both horizontal and vertical set-ups of the con-
tinuous lines were classified as fairly to good flowing according to the
classification of the European Pharmacopeia (Council of Europe, 2011),
as Hausner ratios ranging from 1.15 to 1.24 and from 1.12 to 1.25 were
obtained after processing on the horizontal and vertical set-up, re-
spectively. No influence of the drying parameters on the flowability of
the granules could be detected even though coarser granules were ob-
tained at longer drying times on the vertical set-up.
3.4. General discussion
Possibly, breakage and attrition of the granules during drying in the
continuous line could be reduced if a binder was added via the gran-
ulation liquid as then no time for binder activation is required during
granulation. Alternatively, a higher concentration of binder or a more
potent binder could be included in the formulation to increase the
granule strength and to reduce breakage and attrition during dynamic
drying of the granules. The formulation used in the current study was
established for batch granulation. However, in contrast to traditional
batch granulation processes, the residence time and therefore also the
time available for bond formation during twin screw granulation is very
short, typically 5–20 s in twin-screw granulation versus minutes for
batch granulation (El Hagrasy et al., 2013; Kumar et al., 2014).
Therefore, more potent binders could be necessary for continuous
granulation in comparison to batch granulation.
The results of this experimental study will also be used to calibrate
the mechanistic model for drying of single pharmaceutical granules.
This will allow gaining more fundamental knowledge on the drying
behaviour and breakage and attrition phenomena occurring during
drying of granules on a continuous line. For the calibration of the PBM
on granule moisture content, as well as the two-dimensional PBM in-
volving granule size and moisture content distribution, both developed
by Mortier et al., moisture content distribution data is required,
F. De Leersnyder et al.
involving categorisation of granules of a specific sieve fraction ac-
cording to moisture content (Mortier et al., 2012, 2013).
4. Conclusions
The moisture content of the granules decreased with an increasing
drying time, airflow and drying temperature. Smaller granules dried
faster, resulting in a lower moisture content due to their relative larger
surface area. When drying time increased, differences between moisture
content of larger and smaller granules disappeared. Therefore the
multimodal particle size distribution of the granules or the difference in
drying time between granules filled into the same cell (which is in-
herent to the principle of the segmented fluid bed dryer) did not
compromise uniform drying of the granules if a sufficiently long drying
time is applied.
Extensive breakage and attrition of granules observed on the hor-
izontal set-up was mainly due to the pneumatic transfer of wet granules
between the granulator and the dryer. In the vertical set-up, less
breakage and attrition of granules was observed, since the wet granules
were not subjected to pneumatic transport during transport to the six-
segmented dryer. A higher degree of breakage and attrition of the
granules was observed when shorter drying times were applied on the
granules. When applying short drying times, granules were still wet
when entering the dry transfer line, causing breakage and attrition in
the dry transfer line due to pneumatic transport. Breakage and attrition
of granules during fluid bed drying should be taken into account early
on during process and formulation development on a continuous pro-
duction line for tablets by performing integrated experiments on all unit
operations of the line. Additionally, attention should be given to
granule transfer as it was demonstrated that less breakage and attrition
occurred after gravimetric granule transfer between the granulator and
the dryer.
Overall, the segmented dryer proved to be a robust system towards
granule size distribution and flowability as air flow and drying air
temperature only showed a negligible influence on these granule
characteristics. Therefore, small deviations in these settings during long
term manufacturing are not expected to influence the granule size
distribution or flowability of the granules, which are important critical
quality attributes for further downstream processing.
Acknowledgement
The authors would like to acknowledge GEA Pharma Systems for
offering the possibility to use the vertical ConsiGma™-25 set-up at their
facilities in Wommelgem.
References
Armstrong, N.A., 2007. Tablet manufacture by direct compression. In: Swarbrick, J. (Ed.),
Encyclopedia of Pharmaceutical Technology, 3rd ed. Informa Healthcare, New York,
pp. 3673–3683.
Byrn, S., Futran, M., Thomas, H., Jayjock, E., Maron, N., Meyer, R.F., Myerson, A.S.,
232
European Journal of Pharmaceutical Sciences 115 (2018) 223–232
Thien, M.P., Trout, L.T., 2014. Achieving continuous manufacturing for final dosage
formation: challenges and how to meet them. J. Pharm. Sci. 104, 792–802.
Chablani, L., Taylor, M.K., Methotra, A., Rameas, P., Stagner, W.C., 2011. Inline real-time
near-infrared granule moisture measurements of a continuous granulation–drying–-
milling process. AAPS PharmSciTech 12, 1050–1055.
Council of Europe, 2011. European Pharmacopoeia 7.0. pp. 308–311.
Crouter, A., Briens, L., 2014. The effect of moisture on the flowability of pharmaceutical
excipients. AAPS PharmSciTech 15, 65–74.
El Hagrasy, A.S., Hennenkamp, J.R., Burke, M.D., Cartwright, J.J., Litster, J.D., 2013.
Twin screw wet granulation: influence of formulation parameters on granule prop-
erties and growth behavior. Powder Technol. 238, 108–115.
Emery, E., Oliver, J., Pugsley, T., Sharma, J., Zhou, J., 2009. Flowability of moist phar-
maceutical powders. Powder Technol. 189, 409–415.
FDA, U.S.D. of H. and H.S, 2004. Guidance for Industry PAT — A Framework for
Innovative Pharmaceutical Development, Manufacuring, and Quality Assurance.
vol. 16.
Fonteyne, M., Arruabarrena, J., de Beer, J., Hellings, M., Van Den Kerkhof, T.,
Burggraeve, A., Vervaet, C., Remon, J.P., De Beer, T., 2014a. NIR spectroscopic
method for the in-line moisture assessment during drying in a six-segmented fluid bed
dryer of a continuous tablet production line: validation of quantifying abilities and
uncertainty assessment. J. Pharm. Biomed. Anal. 100, 21–27.
Fonteyne, M., Gildemyn, D., Peeters, E., Mortier, S.T.F.C., Vercruysse, J., Gernaey, K.V.,
Vervaet, C., Remon, J.P., Nopens, I., De Beer, T., 2014b. Moisture and drug solid-state
monitoring during a continuous drying process using empirical and mass balance
models. Eur. J. Pharm. Biopharm. 87, 616–628.
Gohel, M.C., Jogani, P.D., 2005. A review of co-processed directly compressible ex-
cipients. J. Pharm. Pharm. Sci. 8, 76–93.
Gotthardt, S., Knoch, A., Lee, G., 1999. Continuous wet granulation using fluidized-bed
techniques I. Examination of powder mixing kinetics and preliminary granulation
experiments. Eur. J. Pharm. Biopharm. 48, 189–197.
Kleinebudde, P., 2017. Continuous manufacturing of pharmaceutical products. In: 8th
International Granulation Workshop; Granulation Conference, Sheffield, UK, 28–30th
June 2017.
Kumar, A., Vercruysse, J., Toiviainen, M., Panouillot, P.E., Juuti, M., Vanhoorne, V.,
Vervaet, C., Remon, J.P., Gernaey, K.V., De Beer, T., Nopens, I., 2014. Mixing and
transport during pharmaceutical twin-screw wet granulation: experimental analysis
via chemical imaging. Eur. J. Pharm. Biopharm. 87, 279–289.
Kumar, A., Dhondt, J., De Leersnyder, F., Vercruysse, J., Vanhoorne, V., Vervaet, C.,
Remon, J.P., Gernaey, K.V., De Beer, T., Nopens, I., 2015. Evaluation of an in-line
particle imaging tool for monitoring twin-screw granulation performance. Powder
Technol. 285, 80–87.
Mortier, S.T.F.C., De Beer, T., Gernaey, K.V., Vercruysse, J., Fonteyne, M., Remon, J.P.,
Vervaet, C., Nopens, I., 2012. Mechanistic modelling of the drying behaviour of single
pharmaceutical granules. Eur. J. Pharm. Biopharm. 80, 682–689.
Mortier, S.T.F.C., Gernaey, K.V., De Beer, T., Nopens, I., 2013. Development of a popu-
lation balance model of a pharmaceutical drying process and testing of solution
methods. Comput. Chem. Eng. 50, 39–53.
Mortier, S.T.F.C., Gernaey, K.V., De Beer, T., Nopens, I., 2014. Analysing drying unit
performance in a continuous pharmaceutical manufacturing line by means of mass -
energy balances. Eur. J. Pharm. Biopharm. 86, 532–543.
Plumb, K., 2005. Continuous processing in the pharmaceutical industry: changing the
mind set. Chem. Eng. Res. Des. 730–738.
Teżyk, M., Milanowski, B., Ernst, A., Lulek, J., 2015. Recent progress in continuous and
semi-continuous processing of solid oral dosage forms: a review. Drug Dev. Ind.
Pharm. 9045, 1–20.
Van Melkebeke, B., Vervaet, C., Remon, J.P., 2008. Validation of a continuous granula-
tion process using a twin-screw extruder. Int. J. Pharm. 356, 224–230.
Vercruysse, J., Delaet, U., Van Assche, I., Cappuyns, P., Arata, F., Caporicci, G., De Beer,
T., Remon, J.P., Vervaet, C., 2013. Stability and repeatability of a continuous twin
screw granulation and drying system. Eur. J. Pharm. Biopharm. 85, 1031–1038.
Vervaet, C., Remon, J.P., 2005. Continuous granulation in the pharmaceutical industry.
Chem. Eng. Sci. 60, 3949–3957.
Vervaet, C., Remon, J.P., 2009. Continuous granulation. In: Swarbrick, J. (Ed.),
Handbook of Pharmaceutical Granulation Technology. 2009. Informa Healthcare,
New York, pp. 308–322.