Pharmaceutical Development and Technology, 2012; 17(3): 268276

2012 Informa Healthcare USA, Inc.

ISSN 1083-7450 print/ISSN 1097-9867 online
DOI: 10.3109/10837450.2010.531740

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

RESEARCH ARTICLE

Comparative study of the uniformity of coating thickness of

pellets coated with a conventional Wurster chamber and a
swirl generator-equipped Wurster chamber
Matev Lutrik1,2, Rok Dreu1, Rok ibanc1, and Stanko Sri1
Faculty of Pharmacy, University of Ljubljana, 1000 Ljubljana, Slovenia and 2Brinox d.o.o. Process systems, Sora 21, 1215
Medvode, Slovenia
1

Abstract
This study evaluated the performance of two bottom-spray coaters and the effect of pellet-size variability on coating
uniformity. A conventional Wurster chamber was used for the first group of trials, and a Wurster chamber with a
novel swirl-flow generator design was used for the second. The results confirmed that when using a conventional
Wurster coating chamber, pellets with a smaller diameter receive significantly less coating material compared to
those with larger diameters. The swirl generator-equipped Wurster chamber achieved close to uniform coating
thickness regardless of pellet size. The ratio (MS) of the mass of dye deposited in the coating layer to pellet surface
area indicates that coating was much more evenly distributed using the swirl-flow coater. Coating thickness was also
analyzed using SEM micrographs and the results were in close agreement with the MS factor values. Inter-particle
coating mass variation was also lower in case of swirl-flow coater. The results of this study show that a swirl-flow
coater is suitable for coating particles of variable size. They also showed an improvement in coating process yield
when using the swirl-flow coater.
Keywords: Wurster chamber, swirl-flow, pellets, coating thickness, fluid-bed

Introduction

The problem is how to manufacture particles with the

optimal amount of coating. When functional coatings are
applied, the coating must be of a certain thickness and
be evenly and homogenously distributed over the whole
particle surface. Intra- and inter-particle coating thickness variations should thus be reduced to a minimum.
In this study, two different bottom-spray coaters were
evaluated in terms of coating uniformity. The first coater
was a conventional Wurster (CW) process chamber[6]
(Figure 1) and the second was a Wurster-type process
chamber with a novel swirl generator design (swirl
Wurster; SW)[7,8] (Figure 2).
The primary difference between these two coating
devices lies in the design of the air distribution plate
located at the bottom of the process chamber. The air distribution plate in the CW process chamber is designed as
a flat perforated plate.[6] The bottom part of an SW process

Coating based on fluidized bed technology (i.e. air suspension coating) is commonly used in the pharmaceutical industry to coat small particles and tablets. For over
50 years the pharmaceutical industry has been using
fluidized bed technology for particle coating in order to
provide acid resistance,[1] modify release,[2,3] mask the
taste of unpalatable substances, protect APIs from light,
moisture, and oxygen, and improve the appearance.[4]
However, the question of how to achieve optimal and
homogenous particle coating is still an open issue.
Fluidized bed technology coating is usually a batch process, with particles continuously coming into contact
with droplets of coating material and then being dried
using heated air. After many passes through the spraying
and drying zones, a uniform coat is formed with a specific thickness and composition.[5]

Address for Correspondence: Dr. Rok Dreu, MPharm, Department of Pharmaceutical Technology, Faculty of Pharmacy, University of
Ljubljana, Akereva 7, SI-1000 Ljubljana, Slovenia. Tel: +386 1 476 9622. Fax: +386 1 425 8031. E-mail: rok.dreu@ffa.uni-lj.si
(Received 23 August 2010; revised 06 October 2010; accepted 08 October 2010)

268

Fluid-bed coating of particles with broad size distribution 269

4

10
5

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

1
4
7
5

3
2

Figure 1. Schematic diagram of the conventional Wurster coater:

(1) Air preparation unit; (2) fluidizing air intake; (3) air distribution
plate; (4) draft tube; (5) spray nozzle; (6) coating solution; (7)
product; (8) representation of particle motion; (9) outlet filters;
and (10) exhaust blower.

chamber consists of a curved, perforated plate with an

attached swirl generator, embodied as a series of inclined
grooves oriented in a circular fashion. The swirl generator is designed to impart a tangential component to the
air flow, which creates a swirling motion of the particles
inside the draft tube. In conventional fluidized bed coaters, perforations in the air distribution plate are arranged
such that a large proportion of the fluidizing air can flow
directly through the draft tube and only a small proportion around it, causing vertical pneumatic transport of
particles inside the tube.[9] Particles located in the partially
fluidized bed at the bottom of the chamber are allowed
to enter the central area with the spray nozzle by passing the gap either through gravity or the Venturi effect.
The Venturi effect is weaker when using a conventional
Wurster coater compared to the swirl generator-equipped
coater.[10] The Venturi effect is the result of a lower pressure
region inside and directly below the draft tube compared
to the region around the tube created by a higher central
air velocity. The pressure difference is the driving force
directing particles towards the central area.[11]
In addition to the swirling air motion and stronger
particle suction, the swirl generator enables particles to
achieve a screw-like (swirling) movement during transition through the draft tube. CW spray coating analysis
showed denser and slower pellet flow inside the draft
tube compared with swirl generator-equipped coating
machine.[10]
It was reported that the distribution plate design can
significantly affect the hydrodynamics of the coating
chamber and consequently increase coating uniformity
using particles of minimal size variation and decrease the
degree of unwanted agglomeration.[12] Swirl air flow was
2012 Informa Healthcare USA, Inc.

Figure 2. Schematic diagram of the SW chamber: (1) Swirl

generator; (2) curved air distribution plate; (3) fluidizing air
intake; (4) representation of swirling air/particle motion inside
the draft tube (5).

also found to improve drying efficiency, primarily due to

improved heat transfer resulting from specific particle
and air motion.[13,14]
All particles involved in the coating process, regardless of their size, should have the same coating thickness.
The mass of coating deposited on the particles during the
coating process should only be a function of the projected
particle surface area. This implies that the mass of the
coating (m) per pellet should increase with the square of
the particle diameter (d), m=kd2.00, where (k) represents
the proportionality factor.[15,16]
The particle coating thickness is an important parameter that can be assessed using various techniques, such as
analysis of images obtained using optical microscopy,[17]
fluorescence microscopy,[18] Raman spectroscopy,[19]
NIR (near-infrared) spectroscopy,[20] and the recently
developed terahertz pulsed imaging, which offers a
potential technique for examining coatings quickly and
non-destructively.[21,22] Pellet coating thickness can also be
estimated using scanning electron microscopy (SEM[23,24]
Prior to SEM imaging, a cross section of the particle must
be prepared, typically by slicing the particles into two
sections or by removing the upper half of a particle that is
embedded in a polymeric holder.
This study focused on comparing the conventional
bottom spray chamber with a Wurster insert and Wursterlike coating chamber with attached swirl generator. We
evaluated the mass of coating deposited and its uniformity in relation to particle size variations, inter-particle
coating thickness variation within narrow particle size
range and coating process yield.

Materials and methods

Materials
Pellets (Cellets, Syntapharm, Germany) of different size
fractions were obtained by sieving pellets of different

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

270 M. Lutrik etal.

size ranges (Cellets 200, 350, 500, and 700) using test
sieves of various sizes (300, 400, 500, 600, 710, 800, 900,
and 1000 m) and a sieve shaker (Retsch AS 200 basic,
Germany). In order to achieve an accurate size distribution, every portion of pellets was mechanically sieved
twice and also manually sieved.
Pellets were coated with an aqueous dispersion consisting of 8.08% hydroxypropyl methylcellulose (Pharmacoat
606, Shin-Etsu, Japan), 1.01% polyethylene glycol (PEG
6000, Fluka, Switzerland), and 1.09% Tartrazine coloring agent (Sigma-Aldrich, Germany). Coating polymer
dispersion was prepared by adding the hydroxypropyl
methylcellulose and polyethylene glycol to preheated
water at 70C and then stirred for 30min. After leaving
the solution to cool down to room temperature, coloring
agent was admixed and the portion of evaporated water
was replaced. In each experiment 980g of the coating
dispersion was applied to the starting cores.

Coating of pellets
The coating dispersion was sprayed using a binary
nozzle onto 1000g batches of pellet cores. A CW coater
(GPCG-1, Glatt GmbH, Germany) was used for the first
set of experiments and an SW process chamber was used
for the second set to ensure a swirling airflow inside the
draft tube. For both coating processes, the binary nozzle
had a 0.8mm opening diameter and a 2.50mm cap opening diameter.
The process parameters were chosen to ensure comparable coating conditions for all pellet-size fractions
involved in the coating process. The parameters were the
same in all experiments: Batch size 1000g; inlet airflow
rate 130 m3/h; inlet air temperature 55C; spray rate 10g/
min; atomizing air pressure 2.0 bars; gap between distribution plate/swirl generator and the Wurster insert bottom edge 20mm.
At the end of each coating process, the coated pellets
were dried for 5min at an air inlet temperature of 65C.
The coating process yield was determined as the ratio of
the pellet mass increase to the total mass of solid material in the coating solution sprayed onto the pellet cores.
A set of five coating experiments were conducted for
each process chamber design, varying the mass share of
individual pellet-size fractions each time. The first three
mixtures in each set were obtained by mixing four different pellet-size fractions (300400, 500600, 710800,
and 9001000 m) in different proportions, and the last
two sets were prepared by combining the two upper or
lower size fractions of pellets (300400 with 500600 and
710800 with 9001000 m) in equal proportions.

Particle size and shape

Particle size (Feret diameter) and shape (aspect ratio,
projection sphericity, and circularity) were determined
using an optical microscope with bottom illumination (Olympus BX50, Japan), a color camera (Sony
950P, Japan), and image analysis software (Analysis
PRO, Germany). Upper and lower limit color intensity


threshold values had to be set in order to obtain a proper

image and to distinguish the particles from the surface.
This also enabled the elimination of shadows cast by the
particles, which was already minimized through the use
of bottom illumination. Measurements were taken at
40 magnification. The accuracy of measurements was
4.12 m, which represents the length of a pixel at that
magnification.
The maximum Feret diameter of a pellet (dM) is defined
as the maximum caliper distance of 180 measurements
around the particle using 1 steps.[25]
Various shape factors can be used to characterize the
size, elongation, and roundness of a particle. The limit
values of the shape factors analyzed, except for aspect
ratio, vary from zero to one, where one represents an ideally spherical particle.
The aspect ratio (AR) allows for characterization of the
particle shape anisotropy in other words the lengthening of the particle and it is derived as the ratio between
the maximum caliper distance (dM) and the caliper distance perpendicular to the maximum.[25]
The projection sphericity () of a particle indicates
how spherical the object is. It is defined as the ratio of the
projected area of the two-dimensional particle outline (A)
to the surface area of the circle with the maximum caliper
distance (dM) and was calculated using Equation1:[26,27]
c=

4A
pd M 2

(1)

4 pA
P2

(2)

Circularity (C) indicates the roundness of the twodimensional picture of the particle. C is calculated from
the overall perimeter (P) and the projected area (A) of the
two-dimensional particle image (Equation 2):[27]
C=

Calculation of surface area and volume of the prolate

particle
Shape analysis data led us to classify the pellets as ellipsoid. Equation 3 calculates the surface area (Ae) of a prolate particle (ellipsoid), generated by an ellipse rotating
around its major axis (such that the equatorial radius b is
smaller than the polar radius a, which represents half the
length of the maximum Feret diameter):
a

b arc sin(e )
Ae = 2 p b 1+

where

b2
e = 1 2
a

and

b=

(3)

a
AR

The volume of an ellipsoid (Ve) can be calculated using

Equation 4:

Ve =

4
p a b2
3

(4)

Pharmaceutical Development and Technology

Fluid-bed coating of particles with broad size distribution 271

The equivalent to the spherical diameter (d) of an irregularly-shaped object (e.g. prolate particle) is the diameter
of a sphere of equivalent volume. It can be calculated
using Equation 5:
d=3

6Ve
p

(5)

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

Determination of coating uniformity

The evaluation of the amount of coating material deposited on the pellet cores is based upon spectrophotometric
measurements of the dye concentration that was added
to the coating dispersion. The particle-size fractions that
comprised the batch were separated after coating, using
the same sieve set as was used in the preparation of the
initial batch. The dye distribution analysis was conducted
using the population of particles from each size fraction
of every experiment. Thirty samples of 10 particles were
randomly sampled from each size fraction. Each set of
10 particles was put in a vial and 6.00mL of dihydrogen
phosphate buffer with pH 6.5 was added to enable dissolution of the coating layer. Prior to UV analysis, each test
solution was subjected to filtration to eliminate potential
impurities larger than 0.45 m (Minisart RC 25 filters,
Sartorius, Germany). A wavelength of 425nm was used
to assess the dye concentration (UV spectrophotometer
HP 8453, Hewlett-Packard, USA).
The concentration data derived from individual samples (Csample, j) enabled calculation of the arithmetic mean
value (Ci) of dye concentration for individual particle in
the sample and the standard deviation (SDi) of dye concentration for individual particle in the sample according to Equations 6 and 7, where N represents number of
samples and nsample is referred to number of particles in
the sample:[28]
N

Ci =

(C
j =1

sample, j

N nsample

SDi =

(C

sample, j

j =1

2 C sample )2

(6)

(7)

N nsample

The relative standard deviation (RSDi) of dye concentration, which represents inter-particle coating thickness
variation, was calculated using Equation 8:
RSDi =

SDi
100
Ci

(8)

Coating thickness measurement

The evaluation of the thickness of the coating layer deposited on the pellet surface was conducted using scanning
electron microscopy (Supra 32VP, Zeiss, Germany). Prior
to taking SEM micrographs, cross sections of the pellets
2012 Informa Healthcare USA, Inc.

were prepared. 1g of pellets of a specific size fraction was

placed into a round mold and then poured with polymer
(Demotec 95, Demotec Siegfried Demel, Germany). The
polymer was allowed to polymerize at 25C for 15min in
order to obtain round plates with embedded pellets. The
upper portion of the coated pellets was ground off using
silicon carbide abrasive grinding paper with a median
diameter 2.5 m grind particles (standard 1200 ANSI
grind paper).
This technique enables the preparation of cross sections of even the smallest pellets obtained from the coating process and also prevents damage to the much softer
coating layer that might result from techniques such as
the cutting or slicing of pellets.
Ten pellets were randomly chosen from each cross
section plate. The thickness of the coating was measured
at six points for each of the 10 pellets (Figure 3). SEM
micrographs of the pellet cross sections were taken using
an electron beam acceleration voltage of 0.9kV at 1500
magnification.

Results
Image analysis of pellet cores
An image shape analysis system was used to measure
aspect ratio, projection sphericity, circularity, and pellet diameter for each pellet-size fraction. 300 particles
were analyzed from each size fraction used in the coating
experiments in order to achieve a sufficient accuracy. The
acquired data were used to calculate pellet surface area,
volume, and equivalent spherical diameter (Table 1).
Using the same principle as the study performed by
Podczeck etal.,[26] in which spherical pellets had an aspect
ratio value equal to or below 1.02, a projection sphericity value above 0.94, and a circularity value above 0.93,
we determined that our pellets were ellipsoid and were
treated as such in the further representation of results.

Uniformity of coating thickness

The main focus of this work was to evaluate the performance of two different bottom spray coating chambers in
regard to the distribution of coating material. We analyzed
the mass and inter-particle uniformity of dye deposited
in the coating layer and the coating process yield.
The mean value of dye concentration and standard
deviation were calculated for each pellet-size fraction
from each experiment. The mass of dye deposited on the
pellet surface can be obtained from the dye concentration data. The ratio (MS) of the mass of dye deposited
on the pellet to the pellet surface (Ae) was calculated for
each pellet-size fraction. Due to the fact that the dye is
homogenously dispersed throughout the entire coating
layer volume, the value of this factor can be an indicator of coating thickness uniformity. The results indicate
that in a CW chamber (Table 2), the larger pellets receive
significantly more coating dispersion in comparison to
the smaller ones. The mass of dye deposited per square
millimeter of pellet surface varied in experiment 1C from

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

272 M. Lutrik etal.

Figure 3. SEM micrographs of pellet cross sections: (I) Cross section of the pellet taken at 300 magnification. (II) Part of the coating layer
of the pellet coated in the SW coater at 1500 magnification.
Table 1. Aspect ratio (AR), projection sphericity (), circularity (C), particle surface area (Ae), volume (Ve), and equivalent spherical
diameter (d) for pellet-size fractions used in the coating experiments.
Pellet-size fraction (m)
300400
500600
710800
9001000

AR
1.130.09
1.140.09
1.090.04
1.080.04

0.810.11
0.790.12
0.850.07
0.880.06

0.96 g/mm2 for the smallest size fraction (300400 m)

up to 2.10 g/mm2 for the largest pellets (9001000 m).
Using the SW chamber and the same pellet batch (experiment 1S), the MS factor values ranged from 1.12 g/mm2
for 300400 m pellets to 1.75 g/mm2 for 9001000
m pellets. The effect was even more evident in experiments 2C and 2S, in which the mass ratio of the larger
pellets was higher than that of the smaller pellets. The
CW chamber dye per particle surface values varied from
0.88 g/mm2 for the smallest pellets up to 2.12 g/mm2
for the largest pellets. The SW coater values had a much
narrower range of only 1.211.79 g/mm2. The reason
for this effect could be that larger particles entered the
CW chamber spray zone more often. The acceleration
of the particles inside the draft tube is proportional to
the drag force and inversely proportional to the particle
mass. This results in a greater acceleration of smaller
particles and, consequently, higher velocities compared
to larger pellets. Higher pellet velocities enable them to
travel higher, resulting in the smaller particles having
longer trajectories. This is connected with longer circulation times, fewer passes through the spray zone, and
thus with receiving a lower mass of coating. When using
SW process chamber, centrifugal force, imposed on particles by swirling airflow, orients the particles passing the
upper edge of the draft tube towards the wall of the process chamber (expansion part), presumably causing all
particles regardless of size to have comparable flow paths
and similar numbers of exposures to the spray.
Sudsakorn and Turton proposed an alternative explanation for these differences in the mass of coating deposited onto pellets of different diameters.[16] Their hypothesis


C
0.920.04
0.930.08
0.950.02
0.950.03

Ae (mm2)
0.3830.063
1.1640.136
2.0100.125
2.9200.180

Ve (mm3)
0.0230.006
0.1180.021
0.2680.025
0.4690.044

d (mm)
0.3497
0.6084
0.8000
0.9369

suggests that the particles in the CW process chamber

gain considerable radial momentum while crossing the
gap between the draft tube and distribution plate, driving them into the center of the tube during the vertical
acceleration. Large particles accelerate vertically more
slowly than smaller ones and this enables them to travel
closer to the center of the tube and consequently closer
to the spray area, due to longer residence times inside
the draft tube. The more pronounced vertical acceleration of the small particles causes them to move up much
closer to the wall of the tube and thus further away from
the spray zone. The small particles are therefore shielded
from droplets of the coating dispersion by larger particles
located in the midsection of the tube. When a swirling
motion of air and particles is used, all particles regardless of size move closer to the wall of the draft tube due to
centrifugal force exerted by airflow.
In our opinion both hypotheses are very plausible and
contribute to the difference in the coating thickness as a
function of particle size. It would be very hard to decouple
the contribution of both hypotheses experimentally.
An additional aim of this research was to demonstrate
how pellet diameter affects the mass of coating received
during the coating experiment. Assuming that all pellets regardless of size would obtain the same coating
thickness, the mass of the coating layer per unit of pellet surface would be identical for the whole batch. This
can only be achieved if the particles obtain a portion
of coating mass during the coating procedure that is in
proportion to the projected particle surface area, which
represents their potential to capture droplets of coating
solution. In such an ideal case the value of the n-factor,
Pharmaceutical Development and Technology

Fluid-bed coating of particles with broad size distribution 273

Table 2. Particle-size fraction percentages comprising the experiment batches, ratio of mass of coating deposited on the pellet to the
pellet surface area (MS), relative standard deviation of the dye concentration (RSD) for each particle-size fraction, and the total yield of
each coating experiment.
Experiment

Particle-size fraction

Fraction of batch (%)

RSD (%)

25.00
0.96
22.40
300400 m
25.00
1.12
12.70
500600 m
25.00
1.62
8.72
710800 m
25.00
2.10
10.00
9001000m
1S
25.00
1.12
18.10
300400 m
25.00
1.17
11.10
500600 m
25.00
1.51
13.10
710800 m
25.00
1.75
8.63
9001000m
2C
12.50
0.88
31.60
300400 m
12.50
1.20
11.60
500600 m
37.50
1.67
10.50
710800 m
37.50
2.12
6.75
9001000 m
2S
12.50
1.21
23.30
300400 m
12.50
1.32
8.87
500600 m
37.50
1.56
4.59
710800 m
37.50
1.79
4.74
9001000 m
3C
37.50
0.92
29.60
300400 m
37.50
1.08
14.50
500600 m
12.50
1.49
14.10
710800 m
12.50
1.85
9.24
9001000 m
3S
37.50
1.03
23.00
300400 m
37.50
1.05
10.10
500600 m
12.50
1.27
7.44
710800 m
12.50
1.49
8.35
9001000 m
4C
50.00
0.87
20.10
300400 m
50.00
1.07
12.20
500600 m
4S
50.00
0.97
18.10
300400 m
50.00
0.98
10.60
500600 m
5C
50.00
1.78
9.09
710800 m
50.00
2.07
8.55
9001000 m
5S
50.00
1.89
5.20
710800 m
50.00
1.95
4.97
9001000 m
C indicates use of the conventional Wurster process chamber; S indicates use of the swirl Wurster process chamber.
1C

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

MS (g/mm2)

representing the exponent n in the equation m=kdn,

would be 2.00, where m is the mass of coating per particle, k is the proportionality factor, and d is the diameter
of the particle. If the n-factor is more than 2 it means
that bigger particles obtain greater amount of coating
material per square millimeter than smaller ones. The
opposite is true if the n-factor is less than 2. The n-factor
values were calculated for each experiment set (Table 3).
The n-factor values represent the slope of the regression
line between the logarithmic value of the diameter of the
particle and logarithmic value of the average mass of dye
deposited per particle (Figure 4). In the case of coating
experiments 4 and 5, the R2 value of the regression line
was omitted, because the batch of particles was made
up of only two different sizes and the n-factor value was
equal to the line slope between those two experimental
points.
By comparing n-factor values, it can be clearly seen
that higher values were found when the CW process
2012 Informa Healthcare USA, Inc.

yield (%)
58.19

63.89

62.31

68.71

59.17

66.38

54.01
62.69
63.72
75.16

chamber was used. Using the SW process chamber, these

values are closer to the ideal value of 2.00.
We also made comparisons among those experiments
that were conducted using the same type of equipment.
When a greater portion of large particles was used for
CW coating (experiment 2C), a higher n-factor value was
noted, but when a greater portion of small particles was
used the result was a lower n-factor value (experiment
3C) (Table 3).
The difference is even more evident when comparing experiment 5C, with only two larger size fractions,
and experiment 4C, with only two smaller size fractions.
N-factor value dropped significantly in case of smaller
particles. The varying n-factor values for two different
batch mixtures might be the result of having the same
fluidizing air volume rates in all experiments, which may
have led to more uniform movement of smaller pellets.
Pellet diameter exponents (n) for experiments performed
in the SW process chamber show much lower values and

274 M. Lutrik etal.

Table 3. N-factor and R2 values for experiments conducted using a conventional Wurster chamber and a swirl Wurster chamber.
Conventional Wurster
Swirl Wurster
Experiment
1C
2C
3C
4C
5C
1S
2S
3S
4S
n-factor
2.75
2.86
2.68
2.38
2.81
2.43
2.37
2.33
2.02
0.988
0.996
0.993
0.992
0.997
0.992
R2

m2C=kd 2.86
R2=0.996

3.3

log (m)

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

3.1
2.9

m2S=kd 2.37
R2=0.997

2.7
2.5
2.3
2.1

0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 0,5
log (d)

Figure 4. The relationship between the logarithmic value of

mass of dye deposited in the coating layer and the logarithmic
value of pellet diameter d for experiments 2C () and 2S (); the
equations represent the mass of coating deposited with regard to
the exponent value n.

less dependence on the size of particles. However, a difference can still be noted in the case of experiments 4S
and 5S, where the same trend is present as when using
the CW process chamber. The n-factor value for experiment 4S was 2.02; very close to the ideal value of 2.0. This
represents an example of the coating process where the
amount of coating per square millimeter (MS values) was
the same for both size fractions used.

SEM analysis of pellet coating thickness

Coating thickness was additionally determined using
SEM micrograph image analysis of pellet cross sections
for each size fraction coated in experiments 2C and 2S
(Figure 5). A considerable difference was found between
pellets coated in the CW and SW process chambers. The
data confirmed that particles with smaller diameters
received significantly less coating material compared to
larger ones when coated in the CW chamber.[16] The coating thickness of pellets from experiment 2C increased
steadily in connection with the increase in pellet core
diameter (Figure 5). The thickness values ranged from
10 m for the smallest size fractions to 27 m for pellets of the largest size fraction. The SW process chamber
values only varied from 1820 m. This indicates that the
SW process chamber was able to ensure more uniform
contact between droplets of coating solution and pellets
inside the draft tube and therefore ensure a more homogenous coating thickness in experiment 2S.
The coating thickness values established by analyzing SEM micrographs were compared to the MS factor
values. There is a strong correlation between the thickness of the coating layer and the MS value. The R2 value of
the regression of the MS values versus SEM results of the
coating thickness was 0.998 experiment 2C and 0.979 for


Thickness of the coating layer (m)

3.5

5S
2.19

35
30
25
20
15
10
5
0

0,2

0,4

0,6

0,8

Diameter (mm)

Figure 5. SEM-assessed thickness of the coating layer deposited

on the pellets as a function of the equivalent spherical diameter of
pellet-size fractions for experiments 2C () and 2S ().

experiment 2S. This indicates that data representing the

average mass of dye deposited can also be used to estimate coating thickness in experiments in which a SEM
analysis of coating thickness was not performed.

Influence of the Wurster type and particle size on the

variance of coating mass
One of the factors that determine the uniformity of the
coating layer is the inter-particle variation in mass of coating received. This represents the distribution of coating
on particles within a particular pellet-size fraction. When
the deposition of the coloring agent in the coating layer
is uniform, the variance in dye concentration can serve
as a factor for evaluating the differences in the amount
of coating deposited for each size fraction of each experiment. As shown in Figure 6A and in Table 2, particle
diameter does influence the RSD values of dye amount,
which decrease when the particle diameter increases. In
comparing the RSD values of dye amount for pellet-size
fractions coated in the SW and CW process chambers, it
can be noted that the values are significantly lower for
experiments conducted using the SW coater. The only
deviation was found in the case of pellet fraction 700800
m for experiments 1C and 1S, in which the RSD value
was lower using the CW process chamber. Experiment 1S
was repeated but the deviation persisted.
The lower RSD values when using the SW process
chamber might be explained by the particle motion[16]
and the sheltering effect[28] inside the draft tube. Larger
particles tend to move closer to the center of the draft
tube in the CW process chamber, and consequently
closer to the spray zone. This may cause less uniform
coating of smaller particles and higher RSD values of dye
amount due to inter-particle sheltering, which is reduced
Pharmaceutical Development and Technology

for larger particles and increased for smaller particles

(Figure 6A). In the case of the SW chamber, it is assumed
that both small and large pellets are concentrated in the
region closer to the wall of the draft tube, eliminating
size-specific transition of particles into the vicinity of the
spray zone. As described by Cheng and Turton, interparticle RSDs of coating thickness are the result of the
sheltering effect.[28]
When coating two particle-size fractions (Figure 6B),
the RSD values of dye amount also indicate more uniform deposition of coating material in the experiments
conducted using the SW process chamber. The RSD values of dye amount were lower in experiments with larger
pellet batches (5C and 5S) compared to experiments with
smaller pellet batches (4C and 4S). A reasonable explanation would be that, when still wet, smaller pellets tend
to move in loose clusters that consequently increase the
coating thickness RSD values, as opposed to larger particles, which move in a more discrete manner due to having greater momentum. The intensified sheltering effect
and increase in RSD values of dye amount for smaller size
fractions may also result from their much higher numbers of particles and the particles specific surface area.

Coating process yield

Some differences in the coating process yields were also
observed. The loss of coating material, in general, due to
spray drying, deposition on the walls of the draft tube, or

Relative standard deviation (%)

B
Relative standard deviation (%)

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

Fluid-bed coating of particles with broad size distribution 275

35,00

Exp 1c
Exp 1s
Exp 2c
Exp 2s
Exp 3c
Exp 3s

30,00
25,00
30,00
15,00
10,00
5,00
0,00

300-400

500-600

710-800

900-1000

Fraction of pellets (um)

35,00

Exp 4c
Exp 4s
Exp 5c
Exp 5s

30,00
25,00
30,00
15,00

Conclusion
The results of our study clearly indicate that the bottom
spray coating of pellets with broad size distribution varying from 3001000 m can be adequately performed. The
bottom spray coater with a Wurster insert and a swirl generator deposited coating much more evenly on particles
of various sizes compared to the conventional Wurster
process chamber. The coating thickness was almost the
same in spite of the broad particle-size distribution.
When using the CW process chamber, larger pellets
received significantly more coating compared to particles
of smaller diameter. Particle diameter was found to have
a direct linear effect on the amount of coating per square
millimeter deposited. In other words, the thickness of
the coating layer increased linearly with increasing pellet size. The varying mass shares of pellet size fractions
seemed to have a greater impact on the RSD and average mass values of coating deposited when using the CW
coating process compared to the SW process. The RSD of
the amount of dye deposited in the coating layer per particle was significantly lower when using the SW process
chamber. It was also discovered that a greater share of
large particles resulted in a more uniform coating deposition. A noticeable improvement in the coating process
yield was present in all experiments conducted using the
SW process chamber.

Declaration of interest

10,00
5,00
0,00

attrition resulting from inter-particle collisions or collisions with the walls of the process chamber was lower in
all experiments conducted using the SW process chamber. A comparison of experiments with the same batch
composition shows a 9.80% relative increase in coating
process yield in experiment 1S compared to 1C, a trend
that increased in experiments 2, 3, and 4, with the highest
relative increase of 17.95% in experiment 5.
The increase in yield could be explained by the movement of particles in the SW process chamber. All particles
tend to move closer to the wall of the tube. The loss of
coating material due to deposition on the draft tube wall
is diminished in connection with an increased density of
particles in front of the tube wall. It is also assumed that
in the case of the SW process chamber, a lesser coating
layer attrition is present due to lower axial velocities at
the end of the draft tube, resulting in a lower intensity of
particle-to-wall collisions in the filter region.

300-400

500-600

710-800

900-1000

Fraction of pellets (um)

Figure 6. Relative standard deviation (RSD) as a function of

average particle diameter for experiments 1, 2, and 3 (A) and
experiments 4 and 5 (B) conducted using a CW process chamber
(open symbols with dotted lines) and an SW process chamber
(solid symbols with unbroken lines).
2012 Informa Healthcare USA, Inc.

This work was partly funded by the European Union,

European Social Fund (Grant number: P-MR07/06). The
authors report no conflicts of interest. The authors alone
are responsible for the content and writing of the paper.

References
1. Rahman MA, Ali J. Development and in vitro evaluation of enteric
coated multiparticulate system for resistant tuberculosis. Indian J
Pharm Sci 2008;70(4):477481.

Pharmaceutical Development and Technology Downloaded from informahealthcare.com by The University of Manchester on 10/26/14
For personal use only.

276 M. Lutrik etal.

2. Van Savage G, Rhodes CT. The sustained release coating of
solid dosage forms: A historical review. Drug Dev Ind Pharm
1995;21(1):93.
3. Dashevsky A, Wagner K, Kolter K, Bodmeier R. Physicochemical
and release properties of pellets coated with Kollicoat SR 30 D, a
new aqueous polyvinyl acetate dispersion for extended release. Int
J Pharm 2005;290(12):1523.
4. Pearnchob N, Siepmann J, Bodmeier R. Pharmaceutical
applications of Shellac: Moisture-protective and taste-masking
coatings and extended-release matrix tablets. Drug Dev Ind Pharm
2003;29(8):925938.
5. Teunou E, Poncelet D. Batch and continuous fluid bed coating
review and state of the art. J Food Eng 2002;53(4):325340.
6. Wurster DE. Method of applying coatings to edible tablets or the
like. US patent 2 648 609; 1953.
7. Savic S, Gregorka M, Dreu R, Srcic S, Sibanc R, Lustrik M, et al.
A process device for coating particles. WIPO Patent Application
WO/2010/065000, 2010.
8. Savic S, Gregorka M, Dreu R, Srcic S, Sibanc R, Lustrik M, et al.
Procesna naprava za oblaganje delcev. SI patent 22932 A, 2010.
9. Christensen FN, Bertelsen P. Qualitative description of the
Wurster-based fluid-bed coating process. Drug Dev Ind Pharm
1997;23(5):451.
10. Chan L, Tang E, Heng P. Comparative study of the fluid
dynamics of bottom spray fluid bed coaters. AAPS PharmSciTech
2006;7(2):E4553.
11. Porter SC, Bruno CH. Coating of pharmaceutical dosage
forms. In: Lieberman HA, Lachman L, Schwartz JB, editors.
Pharmaceutical dosage forms: Tablets. New York: McGraw-Hill,
1990:77159.
12. Heng P, Chan L, Tang E. Use of swirling airflow to enhance coating
performance of bottom spray fluid bed coaters. Int J Pharm.
2006;327(12):2635.
13. zbey M, Sylemez M. Effect of swirling flow on fluidized bed drying
of wheat grains. Energy Convers Manage 2005;46(910):14951512.
14. Tang E, Wang L, Liew C, Chan L, Heng P. Drying efficiency and
particle movement in coating impact on particle agglomeration
and yield. Int J Pharm 2008;350(12):172180.
15. Ragnarsson G, Johansson MO. Coated drug cores in multiple unit
preparations: Influence of particle size. Drug Dev Ind Pharm
1988;14(15):2285.

16. Sudsakorn K, Turton R. Nonuniformity of particle coating on a size

distribution of particles in a fluidized bed coater. Powder Technol
2000;110(12):3743.
17. Larsen CC, Sonnergaard JM, Bertelsen P, Holm P. Validation of an
image analysis method for estimating coating thickness on pellets.
Eur J Pharm Sci 2003;18(2):191196.
18. Andersson M, Holmquist B, Lindquist J, Nilsson O, Wahlund
K. Analysis of film coating thickness and surface area of
pharmaceutical pellets using fluorescence microscopy and image
analysis. J Pharm Biomed Anal 2000;22(2):325339.
19. Sovny T, Nikowitz K, Regdon Jr. G, Ksa Jr. P, Pintye-Hdi K.
Raman spectroscopic investigation of film thickness. Polym Test
2009;28(7):770772.
20. Kirsch JD, Drennen JK. Near-infrared spectroscopic monitoring
of the film coating process. Pharmaceut Res 1996;13(2):234237.
21. Fitzgerald AJ, Cole BE, Taday PF. Nondestructive analysis of tablet
coating thicknesses using terahertz pulsed imaging. J Pharm Sci
2005;94(1):177183.
22. Ho L, Cuppok Y, Muschert S, Gordon KC, Pepper M, Shen Y,
et al. Effects of film coating thickness and drug layer uniformity
on in vitro drug release from sustained-release coated pellets:
A case study using terahertz pulsed imaging. Int J Pharm
2009;382(12):151159.
23. Nitz M, Taranto OP. Film coating of theophylline pellets in a
pulsed fluid bed coater. Chem Eng Process: Process Intensificat
2008;47(8):14121419.
24. Gao C, Huang J, Jiao Y, Shan L, Liu Y, Li Y, etal. In vitro release and
in vivo absorption in beagle dogs of meloxicam from Eudragit FS
30 D-coated pellets. Int J Pharm 2006;322(12):104112.
25. AnalySIS Users Guide. Soft Imaging System GmbH. Germany,
1998.
26. Podczeck F, Rahman SR, Newton JM. Evaluation of a standardised
procedure to assess the shape of pellets using image analysis. Int J
Pharm 1999;192(2):123138.
27. Bouwman AM, Bosma JC, Vonk P, Wesselingh JA, Frijlink HW.
Which shape factor(s) best describe granules? Powder Technol
2004;146(12):6672.
28. Cheng XX, Turton R. The prediction of variability occurring in
fluidized bed coating equipment. II. The role of non uniform
particle coverage as particles pass through the spray zone. Pharm
Dev Technol 2000;5(3):323332.

Pharmaceutical Development and Technology