International Journal of Pharmaceutics 245 (2002) 9 /24

www.elsevier.com/locate/ijpharm

Ibuprofen crystals with optimized properties

Norbert Rasenack, Bernd W. Müller *
Department of Pharmaceutics and Biopharmaceutics, Christian Albrecht University Kiel, Gutenbergstr. 76, 24118 Kiel, Germany

Received 12 February 2002; received in revised form 27 May 2002; accepted 29 May 2002

Abstract

The common analgesic drug ibuprofen shows bad dissolution and tableting behavior due to its hydrophobic
structure. Additionally its high cohaesivity results in low flowability. Because of the bad compaction behavior
ibuprofen has to be granulated usually before tableting. Another problem in manufacturing is the high tendency for
sticking to the punches. A crystal form with optimized properties of ibuprofen was prepared and characterized in this
study. Crystallization was carried out using the solvent change technique in the presence of different water-soluble
additives. These additives were only present during the crystallization process and removed after precipitation by a
washing process. A nearly pure ibuprofen powder was received, as GC-analysis showed. Plate-shaped crystals with
increased powder dissolution, increased flowability and good tableting behavior were obtained. All crystals were
determined as isomorphic by DSC and X-ray analysis. Thus the improvement of the substance characteristics of
ibuprofen is reached by changes in the outer appearance of the crystals and in surface modifications. Due to the fact
that ibuprofen molecules can form hydrogen bonds, additives that can interact with these hydrogen bonds during the
crystallization process can modify the properties of the resulting crystals. # 2002 Elsevier Science B.V. All rights
reserved.

Keywords: Ibuprofen; Dissolution; Flowability; Tableting behavior; Crystallization techniques; Crystallization in the presence of
additives

1. Introduction flow is bad because of a high cohesivity and

adhesivity. Because of the bad compaction beha-
Ibuprofen is widely used as an analgesic and vior ibuprofen has to be granulated in most cases
antirheumatic drug. The common crystal form before tableting. Another problem in manufactur-
shows disadvantages concerning the properties ing is the high tendency for sticking to the
affecting the manufacturing properties. Powder punches. Beside these disadvantageous properties
ibuprofen shows bad dissolution behavior because
of its hydrophobic structure. However, a rapid
* Corresponding author. Tel.:
/49-431-880-1333; fax:
/49-
431-880-1352
drug release is preferable, especially for analgesic
E-mail address: bwmueller@pharmazie.uni-kiel.de (B.W. drugs.
Müller).

0378-5173/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 5 1 7 3 ( 0 2 ) 0 0 2 9 4 - 6
10 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

The properties of a drug can be affected by structures are mutually bonded which results in a
choosing a suitable polymorphic form, by choos- hydrophobic crystal structure. The most impor-
ing a suitable crystal habit, by using special tant faces in the ibuprofen crystal are the (001) and
crystallization techniques, and by using a suitable (100) faces. While the (001) face is dominant in
preparation with excipients. In the literature crystals grown from solvents other than alcohols,
several methods for improving the properties of the (100) face is dominant in ibuprofen grown
ibuprofen are described. Usually a preparation from ethanol or methanol (Bunyan et al., 1991).
with excipients is used to optimize the substance The (100) face is the slowest growing one and,
properties. A coprecipitation with Eudragit† S100 therefore, dominates the morphology (Shankland
is described by Khan et al. (1994). Kachrimanis et et al., 1996). This surface can either be polar or
al. (1998) also describe spherical crystal agglom- non-polar. Due to the fact that at the polar surface
erates obtained by crystallization by the solvent the carboxylic groups are projected out, they can
change technique in the presence of Eudragit† interact freely with polar molecules as no periodic
S100. A powder with a drug-load of 90% (m/m) is bond chains are disrupted by adsorption of polar
obtained. Flowability and compressibility were molecules. Thus, the growth rate is controlled by
improved because of the Eudragit† S100 the the non-polar surface variant. However, strong
drug release is sustained. Another way to obtain hydrogen bonding and difficult displacement of
a drug powder with improved flowability is the the solvent can change the growth rate as the
spherical crystallization technique as described by growth kinetics is transferred to the polar surface
Kawashima et al. (1982). Spherical crystallization (Bunyan et al., 1991). Nikolakakis et al. (2000)
is defined as an agglomeration technique that describe the effect of different solvents and the
transforms crystals directly into a compacted presence of a methacrylic polymer on the habit of
spherical form during the crystallization process. ibuprofen prepared by a temperature change
Gordon and Chowhan (1990) describe crystals of method. Crystallization was carried out with
naproxen that were modified by the spherical different cooling rates using different alcohols
crystallization technique. An improvement of the and acetone in the presence of different
flow characteristics and the compressibility of drug methacrylic polymers. Especially the solvent and
raw material was observed. However, an increase the cooling rate affect the micromeritic of the
in drug dissolution cannot be achieved by this growing crystals. The polymer has a smaller effect.
method. Another agglomeration technique is the Garekani et al. (2001) describe differences in
non-solvent-shock-agglomeration (Möller and crystal habit of ibuprofen that is crystallized by a
Korsatko, 1999). Melted ibuprofen is agglomer- cooling process from different alcohols and hex-
ated in a non-solvent. Beside differences in crystal ane. During the cooling process ibuprofen powder
agglomeration, differences concerning the indivi- is added as nuclei. While ibuprofen crystallized
dual crystal can also occur. Ibuprofen is able to from methanol and ethanol show a polyhedral
form different crystal forms with different proper- crystal habit, crystallization from hexane results in
ties. The shape of a crystal affects its tabletability. needlelike crystals. The crystal habit influences the
As the shape of ibuprofen crystals depends on the mechanical properties of the drug. If ibuprofen is
solvent, powder flow and compactability can be crystallized from ethanol and methanol the pow-
improved by the choice of a suitable solvent der flow is increased. If hexane is used, the crystals
(Gordon and Amin, 1984). Also hydrogen bond- show the worst powder flow and produced the
ing affects the crystal shape (Shankland et al., softest tablets. The importance of the crystal
1996). In the ibuprofen single crystal unit cell four structure for processing properties is described by
ibuprofen molecules are attached with two hydro- Romero et al. (1991). Five ibuprofen raw materials
gen bonds as described by Romero et al. (1993). obtained from different suppliers were compared.
The two remaining carboxylic-groups are available All crystals were isomorphic. However, differences
for binding to neighboring unit cells. Thus in the concerning the crystal habit, the crystal size, and
crystal structure of ibuprofen the hydrophilic the surface area that affect the granulation process
 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24 11

(different liquid requirements in wet granulation) 2. Materials and methods

and the coating process are described. In a screen-
ing of several crystallization methods (solvent 2.1. Materials
change, temperature change, solvent evaporation)
using different solvents Rasenack and Müller Ibuprofen 50 was supplied by BASF AG
(2002a) describe different crystal habits for ibu- (Ludwigshafen, Germany). These common ibu-
profen. Differences concerning the flowability profen crystals are employed as ‘control’ in this
were observed, but no improvement of dissolution study. Isopropyl alcohol (Merck KG, Darmstadt,
rate was reached. Einig et al. (2000) describe a Germany) was of analytical grade. Water was used
pharmaceutical mixture containing the common in double-distilled quality. All employed additives
crystal form of ibuprofen [90% (m/m)], a nonionic are of pharmaceutical quality. A list of additives
surfactant (0.5%) and other excipients such as used in this study is given in Table 1. The additives
diluents, disintegrants and binding agents nor- are divided in three groups referring to the
characteristics of the obtained ibuprofen crystals.
mally used in tablet production. The drug release
Excipients for tablet production were micro-
in powder dissolution according to the USP 25
crystalline cellulose (Avicel† PH 102, FMC Corp.
method was 100% after 5 min because of the
Philadelphia, USA), sodiumcarboxymethylcellu-
dissolution enhancing effect of the surfactants.
lose (AcDiSol† , FMC Corp. Philadelphia, USA),
Another way of improving drug dissolution is the colloidal silicium dioxide (Aerosil† , Degussa,
forming of an inclusion complex with a cyclodex- Frankfurt, Germany), and magnesium stearate
trin as described by Mura et al. (1998). A (Merck KG).
disadvantage of formulations with cyclodextrins
is the low drug-load. 2.2. Methods
The aim of this study was to find a crystal form
of ibuprofen that is isomorphic with the common 2.2.1. Crystallization procedure
crystal form but shows optimized properties as Crystallization was carried out using the solvent
pure drug powder. Crystallization was carried change method as described by Rasenack et al.
out by the solvent change technique in the (2001) in the presence of water-soluble additives
presence of different additives. These additives using a double-walled glass vessel with thermostat.
were removed from the product, so a pure drug In the first step ibuprofen was dissolved in
powder was obtained. Modified substance proper- isopropyl alcohol at 40 8C (80 g/100 ml). The
ties, which were affected by the additives, were concentration was below the saturation concentra-
investigated. tion to avoid any crystals remaining that would

Table 1
Surfactants and hydrophilic substances used as additives in crystallization process

Surfactants with PEG chain Surfactants without PEG chain Hydrophilic additives

Polyoxyethylene-sorbitane-fatty acid-ester Sucrosemonolaurate (Sisterna B.V., Trehalose (Sigma, Deisenhofen, Ger-

(polysorbates, Tween† ; ICI, Frankfurt, Roosendaal, Netherlands) many)
Germany)
Polyoxyethylene alcohols (Brij† ; ICI) Sucrosemonopalmitate (Sisterna B.V.) Dextran 200 (Sigma)
Polyoxyethylene-glycerylmonoisostearate Sucrosemonostearate (Sisterna B.V.) Hydroxypropyl cellulose (Klucel† ; Her-
(Tagat† I; Goldschmidt, Essen, Germany) cules Inc., Wilmington, USA)
Polyoxyethylen 40 hydrogenated castor oil Sodiumlaurate (Merck KG) Hydroxyethyl starch (Fresenius, Bad
(Cremophor† RH 40; BASF) Homburg, Germany)
Poloxamer 188 (Pluronic† F68; BASF) Sodiumstearate (Merck KG) Polyvinyl alcohol (Sigma), Polyvinylpyr-
rolidone (Kollidon† 25; BASF)
12 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

affect the crystallization process. If possible, the improvement of dissolution rate effected by sur-
additive (8 g) was dissolved in the isopropyl factants remaining on the crystals.
alcohol. Due to their tendency to form a gel at
higher concentrations only 3.2 g of the sodium 2.2.2.3. Dissolution studies: intrinsic dissolution. In
salts of fatty acids were used. Additives that are powder dissolution the specific surface area, crys-
not soluble in organic solvents were dissolved in tal habit, wettability, and differences in agglom-
the water. Precipitation was reached by adding 410 eration tendency all affect the dissolved amount.
ml of water (5 8C) continuously over 70 min In intrinsic dissolution the surface is equal for all
under stirring conditions. Temperature was low- samples, so differences in crystal lattice und
ered from 40 to 20 8C continuously using a wettability can be detected. Intrinsic dissolution
thermostat (Lauda Compact-Kältethermostat was carried out using the ‘rotating disk’ method
RKP20-D, Messgerätewerk Lauda, Lauda-König- (USP 25). Drug powder (200 mg) were compressed
shofen, Germany). The obtained crystals were at 30 kN for 90 s. Changes in polymorphic
collected by filtration under vacuum conditions. modifications were excluded by X-ray-analysis.
They were washed with ice-cold water (5 /50 ml) During dissolution the compact was fixed in a
and dried in a desiccator under vacuum condi- sample holder, so that only one side of the
tions. compact was in contact with the dissolution
medium. Dissolution was carried out in 900 ml
2.2.2. Characterizing methods of pH 7.4 phosphate buffer using an Erweka DT6
(Erweka, Heusenstamm, Germany). The dissolu-
tion medium was vacuum-degassed. The stirring
2.2.2.1. Dissolution studies: powder dissolution in speed was 100 rpm, and the temperature was
phosphate buffer. Powder dissolution and dissolu- maintained at 259/0.5 8C. The dissolved amount
tion from tablets were performed according to the of ibuprofen was quantified at 221 nm (Lambda40
USP 25 rotating paddle method in 900 ml of pH UV VIS Spectrometer, Perkin Elmer).
7.4 phosphate buffer using an Erweka DT6
dissolution apparatus (Erweka, Heusenstamm, 2.2.2.4. Scanning electron microscopy (SEM).
Germany). The dissolution medium was vacuum- Scanning electron-micrographs of crystals were
degassed. The stirring speed employed was 100 obtained using a Philips XL 20 (Philips, Eindho-
rpm, and the temperature was maintained at 259/ ven, Netherlands). Samples were fixed on an
0.5 8C. Quantification of the dissolved amount of aluminium stub with conductive double sided
ibuprofen was carried out spectrophotometrically adhesive tape (Leit-Tabs, Plannet GmbH, Wetzlar,
at 221 nm (Lambda40 UV VIS Spectrometer, Germany) and coated with gold in an argon
Perkin Elmer, Connecticut, USA). All samples atmosphere (50 Pa) at 50 mA for 50 s (Sputter
were analyzed in triplicate. Coater, Bal-Tec AG, Lichtenstein).

2.2.2.2. Dissolution studies: powder dissolution in 2.2.2.5. Differential scanning calorimetry (DSC).
simulated gastric fluids. Dissolution studies of A differential scanning calorimeter (DSC7, Perkin
selected ibuprofen samples were carried out in Elmer) was used. The equipment was calibrated
simulated gastric fluid as a second dissolution using indium and zinc. Samples were heated at
medium using the same technical parameters as 10 8C/min in aluminium pans under nitrogen
described above. This consists of 0.25% sodium atmosphere. The onsets of the melting points and
dodecyl sulfate and 0.2% sodium chloride and is enthalpies of fusion were calculated by the soft-
adjusted with hydrochloric acid to pH 1.29/0.1. ware (PYRIS, Perkin Elmer).
Surface tension is lowered by SDS in an attempt to
mimic in vivo conditions as described by Pedersen 2.2.2.6. X-ray diffractometry. Powder X-ray dif-
et al. (2000). Another aim of using a dissolution fraction (PXRD) patterns were collected in trans-
medium containing a surfactant is to exclude the mission using an X-ray diffractometer (Stoe, PSD
 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24 13

supply unit, Darmstadt, Germany) with Cu Ka 1 Table 2

radiation (monochromator: Germanium) gener- Readings used for the calculation
ated at 30 mA and 40 kV. Powder was packed Factor Unit
into the rotating sample holder between two films
(PETP). Plastic deformation % Percentage of the plastic energy
Total energy N m Plastic and elastic energy
Crushing strength N Stability of the comprimate
2.2.2.7. Relative crystallinity. To quantify possible Fc
disorder in crystal structure, the relative crystal- Fmax kN Upper punch force (maximum)
linity was determined. Crystalline substances show SF (up) mm Displacement of upper punch in
max
Fmax
sharp peaks, whereas amorphous substances only e cm3 Volume (true) tableted
show a ‘halo’. Partially amorphous substances
show both. So by comparing the intensity of the
PXRD patterns the relative crystallinity can be used. To compare different tableting behavior a
determined. By mixing the drug powder with an punch force /displacement-profile of the pure drug
internal standard as described for example by and powder mixtures was recorded. By mathema-
Saleki-Gerhardt et al. (1994), Cordes (1997) a tical calculation based on characteristic data
quantification can be carried out eliminating the (Table 2) of the compression process and the
effects caused by differences in sample density or resulting compact a factor (‘T -factor’) for compar-
sample preparation. In this study magnesium ing the tableting properties was obtained (Rase-
oxide (Merck KG; approximately 20%, accurately nack and Müller, 2002b). The T -factor (J cm4) is
weighted and mixed in an IKA-test tube shaker calculated using Eq. (2). A high value shows good
VF2, IKA Labortechnik, Staufen, Germany) is tabletability. The equation is suitable for compar-
used. PXRD patterns of ibuprofen and magne- ing different powders at the same machine para-
sium oxide are distinctly different: the main peaks meters: the values of the most important
of ibuprofen are below 30 2u; in this area parameters do not seem to be transferable when
magnesium oxide does not show any peaks. The using different adjustments. For comparability the
main peak of magnesium oxide (42.8 2u) appears same experimental set-ups (same adjustment of
in an area where ibuprofen shows no peaks. upper and lower punch, same true volume of each
Intensity of the ibuprofen peaks at 16.5 2u, 20.1 substance compressed) are, therefore, required.
2u, and 22.2 2u and in the case of magnesium oxide
at 42.8 2u were calculated using the computer Fc
T plastic def energytotal  sFmax (up) e (2)
program WINXPOW (Stoe). Relative crystallinity Fmax
was determined using Eq. (1). Eq. (2): Calculation of the T -factor.
P
weight(MgO)  intensityibuprofen
relcryst: (1) To obtain comparable data, constant true
weight(ibuprofen)  intensityMgO
volumes were poured manually into the previously
Eq. (1): Calculation of the relative crystallinity. cleaned die. The amount of powder required was
calculated from the true density. Each powder was
2.2.2.8. Method for analyzing tableting behavior. tableted ten times. The relative standard deviation
The compressions were carried out using a Fette of the calculated T -factors was lower than 5%.
Exacta 11 (Wilhelm Fette Inc., Schwarzenbek,
Germany) single punch press, equipped with flat 2.2.2.9. Specific surface area. The specific surface
punches of 12 mm diameter. Data (punch force, area was determined using the gas adsorption
displacement, time) were recorded (piezoelectric, method. Calculation is based on the BET equa-
resp. inductive). For measuring the mechanical tion. A Surface Area Analyzer Gemini-2360 (Mi-
properties (crushing strength) of the compacts a cromeritics Instrument Corporation, Norcross,
Pharma Test PTB300 (Frankfurt, Germany) was USA) was employed.
14 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

2.2.2.10. Particle size measurement. Particle size 2.2.2.15. Quantification of remaining additives.
measurement was carried out using a laserdiffract- Ibuprofen that was crystallized in the presence of
ometer (Helos KFS Rodos, Sympatec GmbH, sucrosemonolaurate and dextran was analyzed for
Clausthal Zellerfeld, Germany). remaining additives. Sucrosemonolaurate was
quantified after saponification to lauric acid and
derivatization with N -methyl-N -trimethylsilyl-tri-
2.2.2.11. True density. True density was deter-
fluoracetamid by GC (HP 5890, Hewlett Packard
mined using the helium gaspycnometer AccuPyc
Comp., Palo Alto, California, USA; carrier gas:
1330 (Micromeritics Instrument Corporation).
helium; capillar-column: DB5, 30 m/0.25 cm /
0.25 mm, Fisons Instruments, Manchester, UK;
2.2.2.12. Flowability. Flowability was quantified detector: FID; internal standard methyl heptade-
using avalanche analysis to quantify powder flow- canoate). Quantification limit was 50 mg/kg (/
ability (AeroFlow */TSI Modell 3250, TSI, Aa- 0.05 ppm). Dextran was quantified enzymatically
chen, Germany). The powder sample is put in a (testkit D-Glucose, Roche Diagnostics, Man-
cylindrical drum that slowly rotates about its nheim, Germany) as glucose after saponification
horizontal axis at a constant rate. When the incline (in view of the glucose from sucrosemonolaurate).
angle of the powder’s surface becomes too great
for its molecular structure to support, the powder 2.2.2.16. Tablet production. The tablet production
collapses down toward the bottom. This event is was carried out using a Fette Exacta 1 (Wilhelm
referred to as an ‘avalanche’. The time interval Fette Inc.) single punch press, equipped with flat
between avalanches and the amplitude of the punches of 9 mm diameter. Tablets containing 200
avalanche is recorded. Before measurement the mg of ibuprofen were produced. The punch force
powder was disagglomerated through a sieve was adjusted in order to receive tablets with a
(710). Sixty mililitre of each powder were used; crushing strength of approximately 50 N. The
measurement was carried out over 300 s with 1 tablets produced for analyzing the dissolution
UpM. Factors characterizing the flowability are were prepared using 85% ibuprofen (control/crys-
the mean time between avalanches, the scatter and tallized with sucrosemonolaurate/crystallized with
the maximum time. A high mean time and a high sucrosemonolaurate and dextran 200), 7% micro-
maximum time show cohesivity; irregular flow crystalline cellulose (Avicel† PH102), 7% sodium-
characteristics result in a high scatter. carboxymethylcellulose (AcDiSol† ), 0.5%
colloidal silicium dioxide (Aerosil† ) and 0.5%
magnesium stearate. This powder mixture was
2.2.2.13. Contact angle. The contact angle was
tableted directly. However, the composition was
measured by the sessile drop technique using a
not optimal for all crystal forms, but for evaluat-
goniometer (G1, Krüss GmbH, Hamburg, Ger-
ing the dissolution behavior the same powder
many). A compressed disc of the powder (200 mg)
mixture had to be tableted for all ibuprofen
was made at 30 kN for 90 s. The contact angle
crystals.
between the disc and water (saturated with ibu-
profen) was determined 10 and 180 s after the
droplet was put onto the disk.
3. Results and discussion

2.2.2.14. Moisture sorption behavior. The moisture The properties of ibuprofen crystals that were
sorption behavior (adsorption and desorption) obtained by the crystallization technique employed
was analyzed using a VTI MB 300G (VTI in this study differ significantly from the common
Corporation, Hialeah, Florida, USA) at 25 8C crystal form. A macroscopically visible difference
over the humidity range 0 /95% RH (steps of 5% is the optimized flowability. Ibuprofen crystallized
RH). Sample sizes of approximately 75 mg were in the presence of additives shows an increased
used. flowability. Especially crystals grown in the pre-
 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24 15

sence of hydrophilic additives, sucrose esters or (same melting point, same heat of fusion) and X-
sodium salts of fatty acids are free flowing drug ray analysis show. Also the relative crystallinity is
powders that only show little cohesivity or adhe- the same for all crystals (rel. cryst. /0.62, S.D. /
sivity. 0.02), and the true density is 1.11 g/cm3 (S.D./
The outer appearance of the crystals differs (Fig. 0.01) for all samples. The common crystal form of
1). However, all crystals are isomorphic as DSC ibuprofen is a needle-shaped habit with a rough

Fig. 1. SEM photographs of the ibuprofen crystals. (a) Control; (b) crystallized without additives; (c) crystallized in the presence of
polysorbate 80; (d) crystallized in the presence of sucrosemonolaurate; (e) crystallized in the presence of hydroxypropyl cellulose; (f)
crystallized in the presence of sucrosemonolaurate and dextran 200; (g, h) crystallized in the presence of sucrosemonolaurate and
hydroxypropyl cellulose.
16 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

Table 3
Flowability of ibuprofen

Tablettose † Control Without additives Tween† 80 Sucrose-monolaurate Dextran Sucrose-monolaurate
/dextran

Mean (s) 2.3 4.9 4.4 3.0 3.0 3.0 3.1

Scatter (s) 0.8 2.3 1.9 1.3 0.9 1.1 1.0
Max (s) 4.2 12.0 8.4 6.6 5.4 5.2 5.0

surface (Fig. 1a). Crystals that were prepared by tion of these are used. Each group of the additives
the solvent change technique are plate-shaped. is represented in Table 3 with data for polysorbate
However, within ibuprofen that was crystallized 80, sucrosemonolaurate, dextran and a combina-
in the presence of different additives differences tion of sucrosemonolaurate with dextran. Espe-
can be observed. The crystals grown without cially when comparing the data with the data
additives (Fig. 1b) are formed irregularly. Differ- obtained for preagglomerated lactose, the impor-
ences in crystal morphology of the crystals grown tance for processing parameters becomes evident.
in the presence of additives are discernible (Fig. 1): The flowability correlates with the observations by
All crystals are elongated thin plates, but the form SEM: The well-formed crystals that are mainly
of the edges varies: well-formed crystals with a agglomerated at the top have the best flowing
triangular top are obtained in the presence of properties; ibuprofen crystallized without addi-
surfactants without a PEG chain (Fig. 1d) and tives the worst. Polysorbate-modified powder
hydrophilic additives (Fig. 1e). The combination takes an intermediate position.
of both of them results in geometrically exact Also in the tabletability of the pure drug
crystals having symmetrical tops (Fig. 1f /h). powders, there exist differences: the ibuprofen
Crystals are either single or adjacented at their powders with an optimized flowability show a
tops forming hedgehog-like structures. In contrast better tabletability, represented by a higher T -
to this, crystals that are formed in the presence of factor (Table 4). If the powder is filled more
surfactants with a PEG chain (Fig. 1c) are homogenously into the die, the punch force is
agglomerated with their whole surfaces to each more effectively used; the resulting compacts show
other. The crystal edges are not symmetrically a higher crushing strength. The prepared ibupro-
shaped. The observed crystal morphology indi- fen samples can be divided into four groups: In the
cates that additives with hydroxyl groups affect the presence of surfactants with a PEG chain no
growing crystal in a special manner. increase in tabletability is observed. Samples
In correlation to the crystal structures observed crystallized in the presence of hydrophilic additives
by SEM, there exist differences in the powder flow or surfactants without a PEG chain show an
as shown for characteristic samples in Table 3. The increased tabletability which can be further im-
common crystal form is cohesive: the mean proved by their combination. In Table 5 charac-
between two avalanches is 5 s, with a high scatter teristic data for ibuprofen compacts (common
(2.3 s) and a maximum time of 12 s. This shows an crystal form and ibuprofen crystallized in the
irregular flow characteristic. When the crystal presence of sucrosemonolaurate and dextran) are
habit is changed from the needle-shaped crystals compared. At nearly the same maximal punch
to the thin plates, flowability increases. Ibuprofen force the total energy (/area under the punch
that is crystallized without any additives is not force /displacement curve) is significantly higher
such cohesive as the common crystal form. If for the plate-shaped crystals. The elastic recovery
crystallization is carried out in the presence of is lower, thus more energy is put into the drug
additives, the flowability increases significantly, powder during the tableting process, resulting in
especially if surfactants without a PEG chain or compacts that are much more stable. Depending
hydrophilic additives like dextran or a combina- on the alignment of the crystals in the die, the
Table 4

N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

Tabletability of ibuprofen

Ibuprofen crystallized in the presence of surfactants with PEG chain

Control Without additives Tween† 80 Brij† 98 Tagat† I Cremophor† RH40 Poloxamer 188

T -Factor 0.33 0.39 0.37 0.34 0.30 0.31 0.30

(J cm4)

Ibuprofen crystallized in the presence of surfactants without PEG chain

Sucrosemonolaurate Sucrosemonopalmitate Sucrosemonostearate Sodiumlaurate Sodiumstearate

T -Factor 0.46 0.44 0.41 0.45 0.43

(J cm4)

Ibuprofen crystallized in the presence of hydrophilic additives

Trehalose Dextran Hydroxypropyl-cellulose Hydroxyethylstarch Polyvinylalcohol Polyvinylpyrrolidone

T -Factor 0.49 0.45 0.43 0.41 0.50 0.49

(J cm4)

Ibuprofen crystallized in the presence of sucrosemonolaurate and hydrophilic additives

/Trehalose
/Dextran
/Hydroxypropyl-
/Hydroxyethyl-starch
/PVA
/PVP

cellulose

T -Factor 0.51 0.53 0.54 0.52 0.50 0.50

(J cm4)

Results: mean of ten measurements; S.D.B/5%.

17
18 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

Table 5 or a combination of these additives (Table 5). This

Comparison of ibuprofen compacts: characteristic data (pure indicates that differences concerning the surface
drug powder)
structure of the crystals occur during the employed
Control Crystallized in the presence crystallization process. The common ibuprofen
of sucrosemonolaurate and crystals have a rough surface, which may cause a
dextran higher tendency for sticking to the punches. Fig.
Flowability Cohesive Free flowing, not cohesive 2a shows the irregular surface of the ibuprofen
Sticking to the Adhesive Not adhesive tablet (control). An irregular structure with holes
punches caused by pulled out substances that are sticking
Elastic recovery 24 19
to the punch can be observed. In contrast, if
(%)
Fmax (kN) 16 15 tablets are prepared with ibuprofen that is crystal-
Total energy (J) 4.2 4.9 lized in the presence of sucrosemonolaurate and
Crushing 32 55 dextran, the tablet surface is much more regular
strength (N)
and smooth (Fig. 2b). These ibuprofen crystals can
Same true volumes of pure ibuprofen were tableted ten times be directly tableted to tablets with a drug-load of
with same machine adjustments. 85%.
All these differences described above concern
contact area between the particles can vary. the handling and processability of the drug pow-
Another difference concerns the affinity of the der. Beside these there exist differences concerning
ibuprofen powder to the punches. The common the drug release: the powder dissolution shows
crystals stick to the punches. A sticking to the differences depending on the crystallization pro-
punches was not observed for ibuprofen (not even cess. By changing the crystal habit, the drug
for pure drug powder without any lubricants) that powder dissolves only a little bit faster than the
was crystallized in the presence of sucrose esters, control (Fig. 3a). A further increase in dissolution
sodium salts of fatty acids, hydrophilic polymers behavior is observed if ibuprofen is crystallized in

Fig. 2. SEM-photograph of ibuprofen tablet surface. (a) Ibuprofen control; (b) ibuprofen crystallized in the presence of
sucrosemonolaurate and dextran.
 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24 19

Fig. 3. Dissolution profiles of ibuprofen. (a) Common crystal form and crystallized without additives; (b) crystallized in the presence of
surfactants with PEG chain; (c) crystallized in the presence of surfactants without PEG chain; (d) crystallized in the presence of
hydrophilic additives; (e) crystallized in the presence of different types of surfactants with PEG chain; (f) crystallized in the presence of
sucrosemonolaurate and different hydrophilic additives.

the presence of surfactants with a PEG chain (Fig. oleyl chain. A dramatic improvement in drug
3b) or in the presence of hydrophilic additives dissolution occurs if ibuprofen is crystallized in
(Fig. 3d). However, all these ibuprofen crystals the presence of surfactants without a PEG chain
show only a little improvement of the dissolution (sucrose esters and sodium salts of fatty acids)
rate. The type of the hydrocarbon chain or the (Fig. 3c). As a common property of surfactants
length of the PEG chain do not influence the these additives have a hydrophobic and a hydro-
dissolution profile in a significant way (Fig. 3e), philic part of the molecule. The hydrophilic part
but in tendency surfactants with a saturated has high hydrophilicity located in a small area,
hydrocarbon chain lead to an ibuprofen powder either in the form of a sugar with eight hydroxyl-
with a higher dissolution profile than those with an groups or an anionic carboxylic-group. In contrast
20 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

to these structures, the surfactants with a PEG

chain have a long hydrophilic chain where the
hydrophilicity is not located in such a concen-
trated manner. Surfactants with PEG chains can
form hydrogen bonds, but these are weaker than
those of hydroxyl-groups. This results in different
interactions with the growing ibuprofen crystals as
additives with a high tendency to form hydrogen
bonds can interact with the (100) face of the
growing crystal. By strong hydrogen bonding the
crystal growth rate is transferred to the polar (100)
face variant (Bunyan et al., 1991). The geometric
exactness and different types of adjacenting show
that different additives affect the resulting crystals
Fig. 4. Statistical evaluation (powder dissolution, 2 min).
in different ways. A further improvement in
dissolution rate can be achieved if ibuprofen is different (Table 7), but there is no correlation
crystallized in the presence of a sucrose ester and between the dissolution profile and the total sur-
hydrophilic additives, as shown in Fig. 3f. Differ- face. Because of the hydrophobic crystal surface of
ences in dissolution kinetics are made especially
ibuprofen, a distinction has to be drawn between
evident when comparing the slope in the sigma-
the surface area and the wettable surface that is
minus-plot [ln(m0/m )/t] as shown in Table 6 for
available to the dissolution medium. The crystals
the initial dissolution. To evaluate the effect on
formed in the presence of surfactants with a PEG
dissolution, the amount of drug released in minute
chain are bigger (X50 /50 mm) and more agglom-
2 was analyzed statistically (Fig. 4): The main
erated than in the presence of most of the
effect is attributable to sucrosemonolaurate, but
hydrophilic additives (X50/45 mm). This corre-
also dextran and the combination have significant
lates with the specific surface areas. If crystal-
effects. Beside the hydrophilization caused by
lization is carried out in the presence of
interactions, slight amounts of surfactant that
sucrosemonolaurate and a hydrophilic additive,
remain in the product can be anchored in the
the formed specific surface is similar as in the case
crystal surface. The effect of dextran, for example,
of crystallization with only the hydrophilic addi-
is lower; it can form hydrogen bonds, but due to its
tive. Different supersaturation profiles during
purely hydrophilic nature it cannot interact with
precipitation process can explain different crystal
ibuprofen in that way. If a combination is used,
growth rates. The specific surface area of samples
the effects increase.
prepared in the presence of sucrosemonolaurate
Different dissolution behavior can be attributed
takes a mean position, but the dissolution rate is
to differences in the wettable surface. The specific
high. This indicates that other effects influencing
surface area of the prepared ibuprofen crystals is
the drug release rate exist.
Agglomeration tendency, flowability and disso-
Table 6
Slope in sigma-minus-plot ln(m0/m) vs. time (initial) lution behavior can be affected by crystal mod-
ification. Agglomeration and surface properties
Crystallization in the presence of Slope (ln mg/min) affect the distribution in the dissolution medium
Control 0.09 and thus the dissolution rate. The common
Without additives 0.13 ibuprofen crystals remain agglomerated on the
Polysorbate 80 0.25 surface of the dissolution medium because of a
Dextran 0.26 high flotation. In contrast to this, ibuprofen that
Sucrosemonolaurate 0.88
Sucrosemonolaurate
/dextran 3.65
was crystallized in the presence of sucrosemono-
laurate and dextran 200 is spread finely over the
Table 7
Specific surface area of ibuprofen

N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

Ibuprofen crystallized in the presence of surfactants with PEG chain

Control Without additives Tween† 80 Brij† 98 Tagat† I Cremophor† RH40 Poloxamer

188

Spec. surface area 0.1129/0.013 0.2979/0.039 0.1819/0.019 0.1899/0.017 0.1829/0.012 0.1799/0.019 0.1959/
(m2/g) (9/S.D.) 0.019

Ibuprofen crystallized in the presence of surfactants without PEG chain

Sucrosemonolaurate Sucrosemonopalmitate Sucrosemonostearate Sodiumlaurate Sodiumstearate

Spec. surface area 0.2349/0.015 0.2589/0.017 0.2749/0.018 0.2119/0.015 0.2399/0.014

(m2/g) (9/S.D.)

Ibuprofen crystallized in the presence of hydrophilic additives

Trehalose Dextran Hydroxypropyl-cellu- Hydroxyethylstarch Polyvinylalcohol Polyvinylpyrrolidone

lose

Spec. surface area 0.3319/0.012 0.2819/0.013 0.2459/0.017 0.2249/0.015 0.1519/0.014 0.1939/0.009

(m2/g) (9/S.D.)

Ibuprofen crystallized in the presence of sucrosemonolaurate and hydrophilic additives

/Trehalose
/Dextran
/Hydroxypropyl-cel-
/Hydroxyethyl-
/PVA
/PVP

lulose starch

Spec. surface area 0.3119/0.019 0.3149/0.017 0.2799/0.009 0.3149/0.019 0.1729/0.011 0.2079/0.011

(m2/g) (9/S.D.)

Results: mean of three measurements; S.D., standard deviation.

21
22 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

Table 8
Vapor sorption measurement

Control Sucrosemonolaurate Sucrosemonolaurate
/dextran

Begin of water adsorption at RH (%) /95 85 60

Mass difference 0 /95% RH (%) 0.00 0.06 0.20
Hysteresis at 75% RH (%) 0.00 0.00 0.04

dissolution medium and suspended finely in it. crystals. This is an indication for a hydrophiliza-
However, ibuprofen that is crystallized with sur- tion of the crystal surface. Differences in water
factants of the PEG-type is agglomerated and uptake can also influence the tabletability.
dissolves more slowly. Thus a better wettability Fig. 5 shows that differences in powder dissolu-
caused by the residue of surfactants in the final tion are also important for the drug release from
product does not cause the effect. This is empha- tablets. Tablets, which are prepared using ibupro-
sized by the analysis on purity: the amount of fen that was crystallized in the presence of
sucrosemonolaurate was 0.08% (m/m) (S.D. / sucrosemonolaurate and dextran, show a drug
0.02; four separately prepared samples analyzed) release of over 90% after 2 min, including the
the content of dextran was below the quantifica- disintegration time (of approximately 40 s). Ta-
tion limit ( B/0.01%). Therefore, according to the blets (85% drug content) containing the common
pharmacopoeias, a pure drug powder was ob- ibuprofen crystals are relatively faster in release
tained. The most important effect influencing the than it could be expected from the results of
dissolution profile is the distribution behavior. powder dissolution. Because of the mass of a
Intrinsic dissolution shows nearly the same results tablet, flotation effects are lower. To estimate the
for all crystals, as agglomerating effects are relevance of detected differences in drug dissolu-
excluded by carrying out the dissolution from a tion, dissolution studies were carried out in
standardized compact. The contact angle of all simulated gastric fluid. This testing medium con-
samples does not differ (approximately 648, 10 s tains a surfactant (SDS), which results in a
after putting the droplet onto the compact; ap- relatively faster dissolution of the common crys-
proximately 638 after 180 s). These results confirm tals. The effect of flotation is lower because of the
that there are no high amounts of surfactants higher wettability, and agglomerates are disag-
remaining in the drug powder. However, these glomerated rapidly. No significant difference be-
methods are not very sensitive for detecting slight tween the common crystals, the ibuprofen
differences. Beside the determination of the con- crystallized without additives, the ibuprofen crys-
tact angle a more sensitive method to analyze the tallized in the presence of a surfactant with a PEG
affinity to water is the vapor sorption measure-
ment (Table 8): the common ibuprofen crystals do
not show any water adsorption, even at 95% RH
no increase in mass can be observed. Ibuprofen
crystallized in the presence of sucrosemonolaurate
shows a slight water uptake, beginning at 85% RH.
If ibuprofen is prepared in the presence of
sucrosemonolaurate and dextran, the water uptake
starts at 60% RH. Of course, the water uptake (
/
0.2%) is very low, but in comparison to other
samples there is a significant difference. A hyster-
esis (difference between adsorption and deso- Fig. 5. Dissolution from tablets (including destintegration
rption) is only observed at these ibuprofen time).
 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24 23

crystals are formed, which shows the influence on

the crystallization process. The environment af-
fects the external shape of the ibuprofen crystal,
without changing the internal structure; all crystals
were isomorphic. By the use of additives during
the crystallization process, an improvement of
handling properties and of dissolution properties
can be reached.

Fig. 6. Dissolution profile in simulated gastric fluid. Acknowledgements

chain or a hydrophilic additive can be detected. We would like to thank BASF AG for financial
Ibuprofen crystals precipitated in the presence of a support.
sucrose ester and especially in the presence of
sucrosemonolaurate and a hydrophilic additive
show a significant increase in powder dissolution
References
(see Fig. 6). In this dissolution study all hydro-
phobic surfaces can interact with the SDS in the Bunyan, J.M.E., Shankland, N., Sheen, D.B., 1991. Solvent
medium, thus it can be excluded that little effects on the morphology of ibuprofen. AIChE Symp. Ser.
amounts of remaining additives are responsible 87, 44 /57.
for the increase in drug dissolution. Thus the Cordes, D., Ph.D. Thesis, Christian-Albrecht-University Kiel,
modified crystals cause different dissolution ki- 1997.
Einig, H., Hach, H., Eason, R., Müller, B.W., Thompson, R.C.,
netics. This effect occurs especially in the presence 2000. Pharmaceutical Mixture Comprising a Combination
of sucrose esters or sodium salts of fatty acids and of a Profen and other active Compounds. Patent WO 00/
is intensified by addition of hydrophilic additives. 41680.
Garekani, H.A., Sadeghi, F., Badiee, A., Mostafa, S.A., Rajabi-
Siahboomi, A.R., 2001. Crystal habit modifications of
ibuprofen and their physikomechanical characteristics.
4. Conclusions Drug Dev. Ind. Pharm. 27, 803 /809.
Gordon, R.E., Amin, S.I., 1984. Crystallization of Ibuprofen.
US patent 4476248.
If crystallization is carried out in the presence of Gordon, M.S., Chowhan, Z.T., 1990. Manipulation of na-
additives, special effects can occur that influence proxen particle morphology via the spherical crystallization
the crystal habit and the crystal surface. Thus the technique to achieve a directly compressible raw material.
properties of the resulting drug powder can be Drug Dev. Ind. Pharm. 16, 1279 /1290.
Kachrimanis, K., Ktistis, G., Malamataris, S., 1998. Crystal-
affected, even if a pure drug results. Effects that lization conditions and physicochemical properties of ibu-
are not obtainable by different crystallization profen-Eudragit† S100 spherical crystal agglomerates
processes without additives can be achieved. In prepared by the solvent-change technique. Int. J. Pharm.
ibuprofen crystals the hydrogen bonds play an 173, 61 /74.
Kawashima, Y., Okumura, M., Takenaka, H., 1982. Spherical
important role. Additives that are able to interact
crystallization: direct spherical agglomeration of salicylic
strongly with the hydrogen bonds of ibuprofen are acid crystals during crystallization. Science 216, 1127 /1128.
able to affect the growth rate of the dominant Khan, M.A., Bolton, S., Kislalioglu, M.S., 1994. Optimization
(100) surface transferring the growth rate to the of process variables for the preparation of ibuprofen
polar variant of the surface as described by coprecipitates with Eudragit S100. Int. J. Pharm. 102,
Bunyan et al. (1991). Especially the carboxylic- 185 /192.
Möller, T., Korsatko, W., 1999. Die ‘Non-solvent-shock-
group of sodium salts of fatty acids or the sucrose agglomeration’ */Technologie als neuartige alternative
esters (many hydroxyl-groups) are suitable for this Methode zur Aufarbeitung von Ibuprofen. Pharmazie 54,
strong interaction. Geometrically exactly shaped 524 /530.
24 N. Rasenack, B.W. Müller / International Journal of Pharmaceutics 245 (2002) 9 /24

Mura, P., Bettinetti, G.P., Manderioli, A., Faucci, M.T., Rasenack, N., Müller, B.W., Einig, H., 2001. Verfahren zur
Bramanti, G., Sorrenti, M., 1998. Interactions of ketoprofen Kristallisation von Profenen. German Patent Application
and ibuprofen with b-cyclodextrins in solution and in the No. 0050/53009.
solid state. Int. J. Pharm. 166, 189 /203. Romero, A.J., Lukas, G., Rhodes, C.T., 1991. Influence of
Nikolakakis, I., Kachrimanis, K., Malamataris, S., 2000. different sources on the processing and biopharmaceutical
Relations between crystallisation conditions and micro- properties of high-dose ibuprofen formulations. Pharm.
meritic properties of ibuprofen. Int. J. Pharm. 201, 79 / Acta Helv. 66, 34 /43.
88. Romero, A.J., Savastano, L., Rhodes, C.T., 1993. Monitoring
Pedersen, B.L., Müllertz, A., Brondsted, H., Kristensen, H.G., crystal modifications in systems containing ibuprofen. Int. J.
2000. A comparison of the solubility of danazol in human Pharm. 99, 125 /134.
and simulated gastrointestinal fluids. Pharm. Res. 17, 891 / Saleki-Gerhardt, A., Ahlneck, C., Zografi, G., 1994. Assess-
ment of disorder in crystalline solids. Int. J. Pharm. 101,
894.
237 /247.
Rasenack, N., Müller, B.W., 2002a, Properties of ibuprofen
Shankland, N., Florence, A.J., Cox, P.J., Sheen, D.B., Love,
crystallized under various conditions */a comparative
S.W., Stewart, N.S., Wilson, C.C., 1996. Crystal morphol-
study. Drug Dev. Ind. Pharm, in press.
ogy of ibuprofen predicted from single-crystal pulsed
Rasenack, N., Müller, B.W., 2002b. Crystal habit and tableting
neutron diffraction data. J. Chem. Soc., Chem. Commun.
behavior. Int. J. Pharm 244, 45 /57. 7, 855 /856.