AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).
Evaluation of a New Coprocessed Compound Based on Lactose and Maize
Starch for Tablet Formulation
Submitted: July 24, 2003; Accepted: March 17, 2004; Published: May 26, 2004
Karsten Hauschild,1 and Katharina M. Picker1
1Facultyof Pharmacy, Institute of Pharmaceutical Technology and Biopharmacy, Martin-Luther-University Halle-
Wittenberg, Wolfgang-Langenbeck-Strasse 4, 06120 Halle/Saale, Germany
ABSTRACT
The development of new direct compression excipients
should include a comprehensive and rapid determination of
deformation properties. The aim of this study was to charac-
terize StarLac, a new coprocessed compound for direct com-
pression based on lactose and maize starch. For this purpose,
the effects of the base materials (maize starch and spray-
dried lactose) were considered and the influence of the spray-
drying process was investigated. This was performed by
comparing the physical mixture of starch and spray-dried
lactose at the same ratio as for StarLac. For analysis of the
deformation behavior, the 3-D model and the Walker equa-
tion were applied; for verification, the Heckel equation and
the pressure time function (a modified Weibull equation)
were used. The advantages of StarLac are its good flowabil-
ity depending on the spray-drying process, an acceptable
crushing force due to its lactose content, its rapid disintegra-
tion depending on starch, and a brilliant fast release of an
active ingredient, such as theophylline monohydrate. The
volume-pressure deformation properties of StarLac were
dependent on the lactose properties. Only at high maximum
relative density (ρ rel, max) did the influence of starch cause a
change in these properties. A network-like structure can be
observed using scanning electron microscopy pictures.
Overall, StarLac deformed plastically with a low portion of
elasticity. The physical mixture exhibited a more elastic
behavior than StarLac. However, the part of the powder that
was irreversibly compressed was much lower than was
observed for the single substances. This behavior is caused
by an interaction between the components, which in StarLac
is prevented by spray drying.
KEYWORDS: compression, tableting behavior, lactose,
starch, drug release
Corresponding Author: Katharina M. Picker-Freyer,
Martin-Luther-University Halle-Wittenberg, Institute of
Pharmaceutical Technology and Biopharmacy, Wolfgang-
Langenbeck-Strasse 4, 06120 Halle/Saale, Germany.
Tel: +49-345-552 5138. Fax: +49-345-552 7029.
Email: picker@pharmazie.uni-halle.de.
1
INTRODUCTION
Direct compression is a major formulation process in phar-
maceutical technology. With this formulation procedure,
granulation is no longer necessary before tableting. Thus,
direct compression is very efficient. In order to characterize
a new excipient, the investigation of powder technological
properties, including flowability, crystallinity, and water con-
tent, is necessary as is the study of the tableting properties.
The process of tableting consists of both compression and the
final formation of the tablet during storage.1 The whole
process has been described in terms of compressibility, elas-
tic recovery, compactibility, disintegration, and drug release.2
StarLac is a new direct compression excipient, produced by
the spray-drying of α-lactose monohydrate and maize
starch.3 The tableting properties of StarLac were compared
with those of FlowLac (spray-dried lactose), native maize
starch, and a physical mixture of FlowLac and maize starch.
The physical mixture contains FlowLac and native maize
starch in the ratio of 85:15, which is the same ratio as for
StarLac. Thus, the aim of the study was to characterize
StarLac, while considering what effects the base materials
and the spray-drying process may have. In order to investi-
gate the combination of both materials and not the influence
of spray drying, FlowLac, a spray-dried lactose was used.
What is the advantage of a new excipient for direct compres-
sion? The range of excipients for direct compression is small
compared with that of other pharmaceutical-technological
processes. At the beginning, untreated materials were applied
for direct compression, such as powder cellulose or α-lac-
tose-monohydrate. Then, the materials were improved by
using other manufacturing processes, such as spray drying
(spray-dried lactose),4 or by using a special sieve fraction
(lactose 100-mesh).5 The combination of 2 or more sub-
stances in one excipient (compound), for example,
Cellactose or MicroceLac,6 was the next technological inno-
vation advancing the positive excipient properties for direct
compression.
Why should lactose be spray dried in combination with
maize starch? Inferior flowability and insufficient crushing
force of native starch have always been a disadvantage.7
Flowability has to be within a certain limit in order for scale-
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).
up to occur. To minimize the interaction between the 2 pow-
ders and to improve flowability, the spray-drying process
was chosen. The new product should combine the good
flowability and plastic deformation of spray-dried lactose6
with the elastic deformation8 and rapid disintegration of
native maize starch.9
How can the compression properties of a new excipient be
characterized comprehensively and rapidly? The 3 parame-
ters that change during the compression process are time,
pressure, and volume. The volume can be expressed as
porosity. Most of the compression models describe a change
in volume with increasing pressure, as shown by the models
of Heckel,10 Walker,11 Cooper-Eaton,12 Kawakita,13
Sønnergaard,14 and others. Other models demonstrate the
relationship between time and pressure, for instance, the
pressure-time-function15,16 or the Fraser-Suzuki equation.17
The only compression model that combines the change of
porosity (similar to Heckel) with pressure and time is the 3-
D model according to Picker.18 This model characterizes the
compression process, while comprehensively and rapidly
taking into account pressure, normalized time, and porosity.
Furthermore, the part of the powder that is irreversibly com-
pressed should be determined using Walker’s compression
model.11 To verify the 3-D model, the Heckel equation and
the pressure-time function can be used. In additional, some
other parameters can be analyzed (eg, the compression pres-
sure of one particle fraction [Cooper-Eaton], the mean com-
pression pressure according to the linearized form of the
Cooper-Eaton equation,19 the coefficient of compression
[Kawakita]). Sønnergaard characterized the compression
properties by using the Cooper-Eaton and the Walker model
for his new log-exp-model.14
MATERIALS AND METHODS
The new excipient was StarLac, a spray-dried compound of
lactose and maize starch (lot no. L0007, Meggle AG,
Wasserburg, Germany). For comparison, FlowLac 100
(spray-dried lactose, lot no. L0043, Meggle AG, Wasserburg,
Germany) and maize starch (lot no. S8463, Roquette Freres,
Lestrem, France) were tableted. For drug release, theo-
phylline monohydrate (lot no. 21835894, Carl Roth GmbH,
Karlsruhe, Germany) was used as the active ingredient.
Magnesium stearate (lot no. 93810410, Caelo GmbH,
Hilden, Germany) was used as lubricant.
Test Conditions
The powders were equilibrated and the tablets were produced
and stored in a climate-controlled room, which has a constant
temperature (23°C ± 0.5°C) and relative humidity (43% ±
2
2% RH). The powder was mixed with 0.5% (vol/vol) mag-
nesium stearate in a cubic mixer at level 6 for 5 minutes
(ERWEKA, Heusenstamm, Germany) prior to tableting.
Tablets were produced on an instrumented eccentric tableting
machine (Korsch EK0, Korsch Maschinenfabrik, Berlin,
Germany) with 11-mm flat-faced punches (Ritter GmbH,
Berlin, Germany).Tablets were produced at different maxi-
mum relative densities (ρ rel, max). Equation 1 defines ρ rel, max:
(1)
where ρ max is the maximum density under load and ρ true is
the true density of the material
The tablet mass was calculated for each relative density used
(0.75, 0.80, 0.85, 0.90, 0.95). The minimum tablet height
under load was always 3 mm. Each tablet was produced to an
accuracy of ± 0.001 at ρ rel, max by manually weighing the
powder and hand filling the die. The tableting machine is
instrumented with strain gauges and an inductive displace-
ment transducer. Data were collected using a DMCplus
amplifier and an A/D converter (Hottinger Baldwin
Messtechnik GmbH, Darmstadt, Germany) and the software
BEAM 3.1 for MAC (AMS GmbH, Flöha, Germany).
Powder Properties
Water Content
The water content of the powders was determined by thermo-
gravimetry. A Netzsch TG 209 measuring device was used
(Netzsch Gerätebau GmbH, Selb, Germany). Measurement
was performed in the temperature interval in between 23°C
and 170°C with a heating rate of 10 K/min for each powder
equilibrated at 43% RH in triplicate.
Particle Shape and Size
The mean particle size (D50 [μm]) and the particle size distri-
bution were obtained by laser light diffraction using a dry
powder feeder (series 2600c, 63-mm focal length, beam length
10 mm, Malvern Instruments, Worcestershire, UK). Using the
software from Malvern, the mean particle size (D50) and the
specific surface area of the powder were calculated, and the
particle size distribution with parameters for 10% (D10) and
90% (D90) particle size was determined. The optical method
used for calculation of particle size was ISO 13320, a combi-
nation of Fraunhofer and Mie. Measurements were repeated 4
times. Both powder and tablet were additionally analyzed by
means of scanning electron microscopy (SEM) (JSM 6400,
JEOL Ltd, Tokyo, Japan) to determine the particle shape.
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).
Densities
The true density was determined by helium gaspycnometry
using a Micrometrics Accupyc 1330 device (Micrometrics,
Norcross, GA). The measurement was performed in tripli-
cate. Bulk and tap density were determined. Bulk density is
the ratio of the powder weight and the volume of this pow-
der. For tap density, the cylinder was tapped up to a constant
volume (Stampfvolumter SVM, ERWEKA), at least 2500
times.
Flowability
Two parameters that theoretically describe flowability, the
Hausner ratio and the Carr Index, were used. In addition, the
flow rate of the powder was determined as the ratio of mass
(g) to time (seconds); a steel funnel (orifice: 10, 15, and 25
mm diameter; DIN 53916) was used. Mass flow was given
for FlowLac and StarLac.
Crystallinity
The crystallinity of the powders was compared by using an
x-ray diffractometer (URD 63, Freiberger
Präzisionsmechanik, Freiberg, Germany). Source for the
radiation was copper, and a nickel filter was used. Bragg's
angle was analyzed between 3 and 50 2Θ.
Deformation Behavior
To analyze the deformation properties, the data from 5 tablets
of each substance and each ρ rel, max were recorded and dis-
played by BEAM software. The parameters of several defor-
mation models, which are described in the following sec-
tions, were calculated from all data above 1 MPa pressure.
Porosity-Pressure Profile According to Heckel
(2)
Heckel describes the decrease in porosity with pressure by
first-order kinetics.10
where ε indicates porosity; Drel, relative density; K, slope of
the Heckel equation; P, pressure; and A, point of intersection
with the y-axis.
The slope of the Heckel equation provides information on the
plastic deformation of the powder. Heckel found a relation-
ship between the slope of this function and the yield strength
of a powder.20 It has also been published that the slope of the
Heckel equation can be correlated with the elastic modulus
3
(Young’s modulus)21 and that the slope strongly depends on
the true density of the powder.22
Volume-Pressure Plot According to Walker
Walker observed a logarithmic relationship between the rel-
(3)
ative volume and the applied pressure.11 For practical rea-
sons, the relative volume has been multiplied by 100, and the
slope is now defined as the compressibility coefficient W.
where V indicates relative volume; W, compressibility coef-
ficient; P, pressure; and C, constant.
The compressibility coefficient W presents a measure of the
irreversible compressibility of the particles.11 In
Sønnergaard’s opinion, the practical use of the compressibil-
ity coefficient is better than using the slope of the Heckel
equation.21 Thus, he used the Walker equation as the base for
his log-exp-model.14
Pressure-Time Function
The pressure-time function23 was used to describe the influ-
(4)
ence of time. The pressure-time function is a Weibull equa-
tion, which is repeatedly modified.15,16 In the present form,
the equation is able to describe the normalized pressure-time
curve of the tableting process.24
where t indicates normalized time; tend, time at the lifting of
the upper punch (for normalized time, tend = 1); p(t), pressure;
pOmax, maximum pressure of the upper punch; γ, parameter of
asymmetry; and ß, time difference between the time at max-
imum pressure and the lifting of the upper punch.
The parameter γ indicates the asymmetry of the plot. Thus, it
is a measure for the elasticity of the powder.23 With increas-
ing γ, elastic deformation increases. The parameter ß repre-
sents the time difference between maximum pressure and the
lifting of the upper punch from the tablet. Information con-
cerning the fast elastic recovery is given.24
Porosity-Pressure-Time Plot: 3-D Model
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).
(5)
The new 3-D modeling technique is based on porosity, pres-
sure, and time. The only possibility to analyze these 3 impor-
tant variables during the tableting process (normalized time,
pressure, and relative density) in one deformation model is
the 3-D model.18
where Drel indicates relative density; t, normalized time; tmax,
time at maximum pressure; Pmax, maximum pressure; d, time
plasticity; e, pressure plasticity; f, intersection; P, pressure;
and ω, angle of torsion.
To describe the tableting process, the 3 variables were pre-
sented in a 3-D plot and a twisted plane was fitted to the data.
This presentation is called the 3-D data plot. The parameters
of the fitted plane (ie, time plasticity d, pressure plasticity e,
and the angle of torsion ω) were also exhibited in a 3-D plot,
and this plot is called the 3-D parameter plot.
Time plasticity d describes the plastic deformation of the
excipient according to time. It shows how fast the substance
is deformed. It can be influenced by tableting speed.25 With
increasing time plasticity d, the powder deforms faster dur-
ing tableting.2,25 Therefore, with increasing densification
time plasticity d increases. Pressure plasticity e describes the
relationship between density and pressure. The pressure plas-
ticity correlates with the slope of the Heckel equation.2 It
only takes plastic deformation due to pressure into account.
With increasing pressure plasticity e the necessary pressure
for deformation decreases. The angle of torsion ω is a meas-
ure of the material’s elasticity. When ω increases, elasticity
decreases.18 The angle of torsion ω can be interpreted as the
ratio between compression and decompression and thus indi-
rectly describes fast elastic decompression during the tablet-
ing process.18
Tablet Properties
Elastic recovery
The axial elastic recovery26 was determined after ejection of
the tablet by a micrometer screw (Mitutoyo, Tokyo, Japan) at
1, 3, and 6 hours and at 1 and 10 days after tableting. Radial
elastic recovery was only determined directly after ejection
of the tablet and 10 days later. All 20 tablets were analyzed.
The fast axial elastic recovery (fast ER) was calculated by
BEAM 3.1 software (AMS GmbH, Flöha, Germany). The
fast ER is the recovery between minimum tablet height and
lifting of the upper punch from the tablet.24 Axial elastic
recovery over time was additionally determined by thermo-
mechanical analysis. A Netzsch TMA 202 measurement
device was used. This method was only used for tablets made
4
of the pure substances at a ρ rel, max of 0.90.
Crushing Force
Usually the radial crushing force of tablets has to exceed a
certain limit to ensure tablet stability and robustness. The
radial crushing force was determined 5 times (TBH 30,
Erweka Apparatebau) with 2.3 mm/sec rate of compression
during the crushing force test.
Disintegration
To measure disintegration an Erweka ZT 2 coupled with a
water bath was used (Julabo 5, Julabo Labortechnik GmbH,
Seelbach, Germany). Determination was performed for 6
tablets.
Drug Release
For determination of drug release, 20% theophylline mono-
hydrate were blended with the excipients and then tableted.
The low ratio of theophylline monohydrate (20%) reduces its
effect on release properties. Therefore, the drug release main-
ly depends on the excipient. The method was performed
according to United States Pharmacopeia (USP) 24 (sink
conditions were achieved by using maximally 100 mg theo-
phylline monohydrate in one tablet for determination in 1000
mL water). A Pharmatest dissolution tester (PTW II,
Pharmatest Apparatebau GmbH, Hainburg, Germany) was
used for 6 tablets at each ρ rel, max. The concentration of theo-
phylline monohydrate was analyzed spectrophotometrically
(UV-spectrometer Spectronic 601, wavelength 272 nm,
Milton Roy Corp, Ivyland, PA). The wavelength was used
(6)
according to USP 24. A complete UV spectrum was record-
ed as control. The data were interpreted by using the
Schering plot (inverse drug release vs inverse time).27
Reason for using the Schering plot was the quick release of
theophylline-monohydrate from the tablets.
where M indicates the released mass; k, drug release con-
stant; and Mmax, maximally released mass (here 100).
The inverse slope of this plot correlates with the release rate
(inverse slope = product of the release constant and maximal
concentration).27
RESULTS AND DISCUSSION
Powder Properties
Thermogravimetric analysis showed a water content of
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).

Figure 1. Thermogravimetric analysis of StarLac, FlowLac, and maize starch between 20°C and 180°C.

Figure 2. Particle size distribution of StarLac, FlowLac, and maize starch determined by laser light diffractometry.
4.92% (w/w) for FlowLac. This result corresponds to litera-
ture values (4.5%-5.5% [w/w])28 of 5% (m/m) crystalline-
bound water (Figure 1). Maize starch had a water content of
10.94% (w/w). Maize starch29 lost water continuously. This
indicates that there is no tightly bound water. FlowLac lost
most of its water between 130°C and 160°C, indicating that
its crystalline water is tightly bound. The plot of StarLac
shows a combined behavior (Figure 1).
Maize starch showed mainly small particles and a narrow
particle size distribution (Figure 2). The results for StarLac
and FlowLac indicate a similar particle size and particle size
distribution. Both particle size distributions were located at
higher mesh sizes than for maize starch. The only difference
between StarLac and FlowLac was that the particle size dis-
tribution of StarLac was shifted to lower particle sizes and it
was a little bit more narrow (Figure 2). Most probably, these
results are due to the physical interaction between lactose
and maize starch particles during the spray-drying process.
The main parameters D10, D50 (mean particle size), and D90

5
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).
Table 1. Particle Size Distribution Parameters and Specific Surface Area of StarLac, FlowLac, and Maize Starch
FlowLac
D10 [μm] 31.5 ± 1.7
D50 [μm] 110.8 ± 3.6
D90 [μm] 194.5 ± 3.7
Specific surface area [m2/cm3] 0.10 ± 0.01
and the specific surface area are given in Table 1. The specif-
ic surface area of StarLac is higher than that of FlowLac. The
SEMs at 500 magnification show a similarity between
FlowLac and StarLac (Figure 3). Both photomicrographs
Figure 3. SEM of the particles of StarLac (A, C) and
FlowLac (B, D) at different magnifications (A and B at
original magnification ×500; C and D at ×15 000).
show crystals (most likely α-lactose monohydrate) and areas
with amorphous lactose. The particles are spherically shaped.
At a magnification of 15 000, slightly modified maize starch
particles can be detected inside StarLac. This finding may be
due to the spray-drying process.
The apparent particle (true), tap, and bulk density are shown
in Figure 4. Both spray-dried products, FlowLac and
StarLac, had excellent flow properties. This was indicated by
a high flow rate (for both, 16 g/sec at 15-mm diameter fun-
nel). Maize starch and the physical mixture never flowed
through the funnel. This result could be confirmed by the
Hausner ratio (1.15 for the spray-dried products and 1.3 for
starch) and the Carr-Index (15 for the spray-dried products
and 25 for starch). Flowability depends on the production
process.
The x-ray diffractogram of maize starch showed no crys-
talline parts when compared with FlowLac and StarLac. The
diffractograms of FlowLac and StarLac look similar in spite
of the 15% starch-fraction in StarLac. This phenomenon
could be the result of a better recrystallization in combination
with maize starch during spray drying of StarLac.
6
StarLac Maize Starch
13.5 ± 0.2 8.1 ± 1.2
54.8 ± 2.8 13.8 ± 0.5
151.4 ± 6.2 20.5 ± 1.4
0.20 ± 0.01 0.54 ± 0.07
Deformation Behavior
The slope of the Heckel equation indicates the plastic defor-
mation of a material. With increasing slope of the Heckel
equation, plastic deformation increases. This slope correlates
to the pressure plasticity gained by 3-D modeling.2 The
Heckel slopes of FlowLac, the physical mixture, and StarLac
decreased up to a ρ rel, max of 0.90. Thus, plastic deformation
decreased with increasing pressure (Figure 5). The Heckel
slope of maize starch increased mainly up to a ρ rel, max of
0.95. StarLac showed the same change at a ρ rel, max of 0.95,
but the change was not as strong. Due to this difference, the
change must be caused by maize starch. StarLac and the
physical mixture had a higher Heckel slope than FlowLac.
Thus, starch had a general effect on StarLac and on the phys-
ical mixture. However, why are the Heckel slopes of StarLac
and the physical mixture different? Only some other change
caused by the spray-drying process can explain this differ-
ence.
In order to characterize the change of the volume, the Walker
equation with the irreversible compressibility coefficient W
was used. Walker describes the changes relative to bulk den-
sity and determines the part of the powder that was irre-
versibly compressed at maximum pressure. The part of the
StarLac powder that was irreversibly compressed was
revealed as constant 40% (Figure 6). Maize starch and
FlowLac exhibited a higher irreversible compressibility W at
all times. At first, the irreversible compressibility W of
FlowLac decreased with increasing pressure up to a pressure
of 120 MPa, and then it increased. Maize starch exhibited a
continuous increasing irreversible compressibility W. The
behavior of the physical mixture showed an extreme lower
part of the irreversible compressibility W (from 5%-15%).
Thus, the interaction between starch and lactose particles
minimizes the part of the irreversibly compressed powder.
The spray-drying process increased the interaction between
different particles, and the part of the irreversibly com-
pressed powder decreased (compressibility coefficient W).
Elasticity can be determined by the ß- and γ-values of the
pressure time-function. The parameter ß describes the timing
relationship between compression (1-ß) and decompression
(ß). Thus, elasticity increases with higher values for the
parameter ß. The parameter γ is the symmetry factor of the
curve, which means that if the γ-value increases, the curve
starts to look more and more alike for compression and
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).

Figure 4. Bulk, tap, and apparent particle (true) densities of StarLac, FlowLac, physical mixture, and maize starch.

Figure 5. The slope of the Heckel equation of FlowLac, StarLac, physical mixture, and maize starch at different ρ rel, max
(0.75 to 0.95).
decompression (the same characteristics were arrived at a γ-
value of 3.3; the curve is then a normal distribution). When
looking at the ß-γ diagram, maize starch exhibited the high-
est elastic deformation, indicated by high values of the
parameters ß and γ (Figure 7). FlowLac and StarLac exhibit-
ed similar elastic deformation up to a ρ rel, max of 0.90.
However, the influence of maize starch in StarLac was high-
er. The deformation behavior of the physical mixture was
more elastic when compared with that of StarLac (Figure 7).
In Figure 7 it can be seen that starch had the highest elastic-
ity, followed by the physical mixture and both the spray-
dried products (FlowLac and StarLac). The figure also shows
that the change of the γ-value with higher compression
(arrows) was nearly the same for the physical mixture and
starch. The γ-value did not increase. At low compression, the
change of the γ-value was the same for FlowLac and StarLac.
The γ-value increased, and the ß-value did not change. At
high compression, the γ-value and the ß-value of StarLac
decreased as they did for starch. Thus, the elasticity of
StarLac is influenced at low compression by lactose and at
high compression by starch. Furthermore, the result indicates
that the elastic behavior is mainly decided by maize starch

7
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).

Figure 6. Compressibility coefficient W of FlowLac, StarLac, physical mixture, and maize starch at different ρ rel, max.

Figure 7. β-γ-plot result of the pressure-time function at ρ rel, max of 0.75 to 0.95 (arrows point to higher ρ rel, max).
despite its low volume part.
Characterized by 3-D modeling, the elasticity according to
the angle of torsion ω of the 3-D model is generally higher
for maize starch than for all other substances (Figure 8).
However, pressure plasticity e and time plasticity d for starch
are high as well. Maize starch showed a different plot form
when compared with FlowLac, the physical mixture, and
StarLac. Maize starch exhibited similar pressure plasticity at
all ρ rel, max used (Figure 8). However, time plasticity d
increased rapidly with increasing ρ rel, max. The angle of tor-
sion ω decreased with increasing ρ rel, max. This result means
that fast elastic decompression and thus elasticity increased
with increasing pressure. A change in elasticity was especial-
ly visible between a ρ rel, max of 0.90 and 0.95 (Figure 8). The
plots of FlowLac, the physical mixture, and StarLac
(between a ρ rel, max of 0.75 and 0.90) indicate similar com-
pression properties. In comparison to maize starch, they
exhibited lower pressure and time plasticity, and they exhib-

8
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).

Figure 8. The 3-D parameter plot of StarLac, FlowLac, maize starch, and physical mixture (ρ rel, max increased with increas-
ing time plasticity d from 0.75 to 0.95).
ited a strong decrease in ω-values with increasing ρ rel, max. help of the properties of FlowLac. Time plasticity d corre-
With increasing ρ rel, max, pressure plasticity e and the angle lates exponentially to ρ rel, max (Table 2). A special phenome-
of torsion ω decreases, and time plasticity d increases. This non was the change in deformation properties of StarLac
finding indicates that a part of the powder exhibits brittle between a ρ rel, max of 0.90 and 0.95. Time plasticity did not
deformation behavior.2 The pressure plasticity e increased in increase as strongly as before and pressure plasticity
the following order: physical mixture > StarLac > FlowLac. increased more strongly than usual. This finding could indi-
Time plasticity d was identical for all of them. Thus, for cate that the maize starch incorporated in StarLac only influ-
StarLac brittleness was slightly reduced in favor of some ences tableting behavior at ρ rel, max higher than 0.90.
plasticity resulting from the incorporated maize starch parti-
cles.
Tablet Properties
FlowLac and StarLac showed up to a ρ rel, max of 0.90 much
more brittle deformation than maize starch (Figure 8). This
finding corresponds to the Heckel-slope (Figure 5). The
change of the pressure plasticity e of StarLac and the physi-
cal mixture with increasing ρ rel, max can be interpreted with

Table 2. Exponential Regression of Time Plasticity (y = a *

e b x, a = 0.0005) of FlowLac, Maize Starch, and StarLac
(without 0.95 ρ rel, max) at Different ρ rel, max (0.75 to 0.95)
Exponent b
R2
(Time Plasticity d)
FlowLac 8.3161 0.9990 Figure 9. SEM of StarLac (B, D, and E) and FlowLac (A
StarLac 8.4229 0.9672 and C) before (A and B) and after tableting (C and E at ρ rel,
Physical mixture 8.5786 0.9729
max of 0.95 and D at ρ rel, max of 0.90) at original magnifica-
Starch 8.9403 0.9941 tion ×15 000 (1 in part E indicates the newly built net-
work).
9
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).

Table 3. Elastic Recovery at Different ρ rel, max of FlowLac, StarLac, and Maize Starch*
ρ rel, max Fast ER (%) ER After Ejection (%) ER After 1 Hour (%) ER After 10 Days (%)
Maize starch 0.75 5.20 ± 0.10 12.87 ± 0.22 14.09 ± 0.18 15.44 ± 0.09
0.80 5.42 ± 0.09 13.43 ± 0.21 14.75 ± 0.17 16.17 ± 0.19
0.85 5.47 ± 0.12 13.55 ± 0.22 14.85 ± 0.16 16.25 ± 0.24
0.90 5.65 ± 0.09 13.24 ± 0.24 14.66 ± 0.21 16.23 ± 0.33
0.95 5.80 ± 0.10 13.05 ± 0.24 14.37 ± 0.26 15.92 ± 0.32
StarLac 0.75 1.71 ± 0.26 4.99 ± 0.20 5.27 ± 0.16 5.62 ± 0.09
0.80 2.06 ± 0.08 4.25 ± 0.9 4.72 ± 0.10 4.89 ± 0.13
0.85 2.20 ± 0.20 4.30 ± 0.11 4.71 ± 0.11 4.73 ± 0.11
0.90 2.51 ± 0.09 4.30 ± 0.10 4.61 ± 0.10 4.63 ± 0.11
0.95 2.76 ± 0.09 4.37 ± 0.12 4.54 ± 0.09 4.54 ± 0.13
FlowLac 0.75 1.69 ± 0.05 1.80 ± 0.10 1.85 ± 0.10 1.93 ± 0.10
0.80 1.84 ± 0.06 2.06 ± 0.16 2.09 ± 0.15 2.13 ± 0.16
0.85 2.03 ± 0.08 2.35 ± 0.11 2.38 ± 0.13 2.37 ± 0.10
0.90 2.55 ± 0.07 1.94 ± 0.13 1.99 ± 0.10 2.00 ± 0.13
0.95 2.72 ± 0.24 1.69 ± 0.12 1.76 ± 0.15 1.80 ± 0.12
*ER indicates elastic recovery.

The SEMs of the interior of the maize starch tablets show
that most of the maize starch particles were still intact
(Figure 9), and the surfaces of the starch particles touch each
other. The SEM of maize starch indicates small particles that
can be the result of a broken starch particle. The interior of
StarLac and FlowLac tablets looks similar at a ρ rel, max of
0.90 (Figure 9); at a ρ rel, max of 0.95 the interior of StarLac
tablets exhibits a special structure (Figure 9), which could be
due to fused starch particles.
Axial and radial elastic recovery were determined at various
times. Maize starch tablets showed the highest axial elastic
recovery and tablets made of FlowLac, the lowest (Table 3).
StarLac tablets had a similar fast ER compared with FlowLac
tablets. At the other times, they recovered more than tablets
made of FlowLac. This result is because they are partly com-
posed of maize starch. At all ρ rel, max, FlowLac tablets recov-
ered elastically up to the lifting of the upper punch from the
tablet. Elastic recovery of StarLac tablet stopped after 1 day,
and elastic recovery of maize starch tablets stopped after 3
days. With increasing densification, fast recovery of all of the
substances increased continuously (Table 3), since the densi-
ty at maximum pressure increases. Therefore, in order to
reach the same density there must be an increase in fast elas-
tic recovery. And, this increase is visible in spite of there
being less time to recover as indicated by the parameter ß of
the time-pressure function (Figure 7). The elastic recovery of
all of the tablets had a maximum at ρ rel, max of 0.85. The rea-
son for this are 2 oppositely acting processes: in the first
process, the density at maximum pressure increased (fast ER
increased) and in the second one, a part of the powder was
irreversibly compressed as indicated by the compressibility
coefficient W (Figure 7). Thus, the ER decreased at high ρ rel,
max. Theradial elastic recovery of tablets made from all of the
substances was much lower (1%) than the axial recovery,
because of uniaxial compression in tablet production.
Maize starch tablets never exceeded an acceptable crushing
force. The compactability profiles showed different plots for
StarLac and FlowLac (Figure 10; compression pressure to
crushing force was used for a better understanding of com-
pactibility compared with other studies, instead of using ρ rel,
max). StarLac and FlowLac tablets achieved a minimal limit
of 60 N at a ρ rel, max of 0.90. Theoretically, StarLac tablets
exceeded the minimal limit with 112 MPa and FlowLac
tablets with 104 MPa compression pressure. The crushing
force behavior of tablets made of the physical mixture was
similar to that of FlowLac tablets, however at a lower level.
Tablets made from all of the substances disintegrated rapid-
ly. The limit of European Pharmacopoeia (15 minutes) was
never exceeded.
Tablets made with StarLac, maize starch, and FlowLac all
showed a fast release of theophylline monohydrate (active
ingredient). At least 80% of theophylline monohydrate were
released after 20 minutes. Based on this fast release, the
Schering plot was used for analysis of the release data. The
k-value of the Schering-plot is directly proportional to the
release rate. StarLac tablets showed a strongly higher release
rate than the other tablets (Figure 11). Concerning FlowLac,
up to a volume fraction of 85%, no difference could be seen
in drug release. Then the rate of release increased.
CONCLUSION
StarLac is well suited for direct compression; it demonstrat-

10
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).

Figure 10. Compactibility plot of StarLac, FlowLac, physical mixture, and maize starch and linear fittings (thin black lines)
at different ρ rel, max (0.75 to 0.95).

Figure 11. Inverse slope of the Schering plot of the mixtures (FlowLac, maize starch, and 20% theophylline monohydrate)
and StarLac with 20% theophylline monohydrate.
ed good compactibility and very fast release behavior. The mation properties, as indicated by the parameters of the 3-D
powder properties were similar when compared with those of model and the Walker equation, allowed for the following
FlowLac, and this was true in spite of the interaction between conclusion to be made: StarLac exhibited deformation
the 2 components during the spray-drying process as indicat- behavior with higher parts of plastic and elastic deformation
ed by particle size, particle size distribution, SEM pictures, than FlowLac, therefore StarLac is of interest for the manu-
and x-ray-diffractograms of the powders. facture of pressure-sensitive drugs.
The comprehensive and rapid characterization of the defor- Only at high ρ rel, max a change in the deformation properties

11
 AAPS PharmSci 2004; 6 (2) Article 16 (http://www.aapspharmsci.org).
of StarLac tended to result in those that take place in starch.
This finding can be confirmed by SEM pictures of StarLac
tablets. A network-like structure can be identified, which
could result from “fused” starch particles. At the part where
the starch fused, the plastic deformation part increased. This
result is exhibited in both deformation parameters, the time-
dependent ones and the volume-dependent ones. The tablet-
ing behavior of the physical mixture is dependent on starch
(elasticity) and on lactose (pressure plasticity). The crushing
force is lactose dependent, and the drug release is nearly the
same as for the starch tablets. The different tableting and
tablet behavior of StarLac and the physical mixture can be
explained as being caused by the spray-drying process.
The fast elastic recovery increased continuously with ρ rel,
max. However, during further time dependent elastic recovery,
the part of the powder that was irreversibly compressed was
more and more influenced by elastic recovery. The crushing
force, which was insufficient for maize starch tablets, was
enhanced for StarLac. However, this was not caused by the
spray-drying process (which decreased the crushing force
slightly) but rather by the higher lactose content. The release
rate will gain an advantage with fast disintegration. The
biggest advantage of StarLac tablets is its very fast release of
the active ingredient.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge Roquette Freres and
Meggle AG for materials and financial support.
REFERENCES
1. Picker KM. Soft tableting: A new concept to tablet pressure-sensitive
drugs. Pharm Dev Technol. 2004;9(1):107-121.
2. Picker KM. New Insights in the Process of Tablet Formation: Ways to
Explore Soft Tableting. [Habilitationsschrift]. Halle, Germany: Martin-
Luther-University Halle-Wittenberg and Görich und Weiershäuser
Verlag, Marburg; 2002.
3. Meggle AG. Technical information at www.meggle-pharma.de/en/
products/uebersicht/starlac.
4. Vromans H, Bolhuis GK, Lerk CF. Studies on tableting properties of
lactose. VI. Consolidation and compaction of spray-dried amorphous lac-
tose. Acta Pharm Suec. 1986;23:231-240.
5. Vromans H, de Boer AH, Bolhuis GK, Lerk CF, Kussendrager KD.
Studies on tableting properties of lactose. I. The effect of initial particle
size on binding properties and dehydration characteristics of lactose. Acta
Pharm Suec. 1985;22:163-172.
6. Goto K, Sunada H, Danjo K, Yonezawa Y. Pharmaceutical evaluation
of multipurpose excipients for direct compressed tablet manufacture:
comparisons of the capabilities of multipurpose excipients with those in
general use. Drug Dev Ind Pharm. 1999;25:869-878.
7. Bolhuis GK, Chowhan ZT. Lactose. In: Alderborn G, Nyström C, eds.
Pharmaceutical Powder Compaction Technology. New York, NY: Marcel
Dekker; 1996:459-478.
12
8. Paronen P, Juslin M. Compressional characteristics of four starches. J
Pharm Pharmacol. 1983;35:627-635.
9. Bos CE, Bolhuis GK, van Doorne H, Lerk CF. Native starch in tablet
formulations: properties on compaction. Pharm Weekbl Sc Ed.
1987;9:274-282.
10. Heckel RW. Density-pressure relationship in powder compaction.
Trans Metall Soc AIME. 1961;221:671-675.
11. Walker EE. The influence of the velocity of compression on the appar-
ent compressibility of powders. Trans Faraday Soc. 1923;19:614-622.
12. Cooper AR, Eaton LE. Compaction behavior of several ceramic pow-
ders. J Am Ceram Soc. 1962;45:97-101.
13. Kawakita K, Lüdde KH. Some considerations on powder compres-
sion equations. Powder Technol. 1971;4:61-68.
14. Sønnergaard JM. Investigation of a new mathematical model for com-
pression of pharmaceutical powders. Eur J Pharm Sci. 2001;14:149-157.
15. Dietrich R, Mielck JB. Parametrisierung des zeitlichen Verlaufs der
Verdichtung bei der Tablettierung mit Hilfe der modifizierten Weibull-
Funktion: 1.Mitt.: Gedanklicher und experimenteller Ansatz. Pharm Ind.
1984;46:863-869.
16. Dietrich R, Mielck JB. Parametrisierung des zeitlichen Verlaufs der
Verdichtung bei der Tablettierung mit Hilfe der modifizierten Weibull-
Funktion: 2.Mitt.: Auswertung und Interpretation der Ergebnisse. Pharm
Ind. 1985;47:216-220.
17. Fraser RDB, Suzuki E. Resolution of overlapping bands, functions
for simulating band shapes. Anal Chem. 1969;41:37-39.
18. Picker KM. The 3-D Model: Comparison of Parameters Obtained
From and by Simulating Different Tableting Machines. AAPS
PharmSciTech. 2003; 4(3):E35.
19. Kurup TRR, Pilpel N. Compression characteristics of pharmaceutical
powder mixtures. Powder Technol. 1978;19:147-155.
20. Heckel RW. An analysis of powder compaction phenomena. Trans
Metall Soc AIME. 1961;221:1001-1008.
21. Sønnergaard JM. A critical evaluation of the Heckel equation. Int J
Pharm. 1999;193:63-71.
22. Sønnergaard JM. Impact of particle density and initial volume on
mathematical compression models. Eur J Pharm Sci. 2000;11:307-315.
23. Dietrich R, Mielck JB. Eignung der Weibull-Funktion zur
Charakterisierung des zeitabhängigen Verformungsverhaltens von
Tablettierhilfsstoffen. Acta Pharm Technol. 1985;31:67-76.
24. Picker KM. Hydrophile Matrixtabletten: Tablettierung und
Freisetzung unter besonderer Berücksichtigung der relativen Feuchte
während der Herstellung [Dissertation]. Hamburg, Germany: University
of Hamburg; 1995.
25. Picker KM. The 3-D Model: Does Time Plasticity Represent the
Influence of Tableting Speed? AAPS PharmSciTech. 2003;4(4):E66.
26. Armstrong NA, Haines-Nutt RF. Elastic recovery and surface area
changes in compacted powders systems. J Pharm Pharmacol.
1972;24(suppl):135P-136P.
27. Koch HP. Die Technik der Dissolutionsbestimmung (2.Teil). Pharm
Acta Helv. 1984;59:130-139.
28. Kibbe AH, Weller PJ. Lactose. In: Rowe RC, Sheskey PJ, Weller PJ,
eds. Handbook of Pharmaceutical Excipients. London, UK:
Pharmaceutical Press; 2003:323-332.
29. Rowley G. Starch. In: Rowe RC, Sheskey PJ, Weller PJ, eds.
Handbook of Pharmaceutical Excipients. London, UK: Pharmaceutical
Press; 2003:601-608.