European Journal of Pharmaceutical Sciences 18 (2003) 3745

www.elsevier.com / locate / ejps

Floating matrix tablets based on low density foam powder: effects of

formulation and processing parameters on drug release
A. Streubel, J. Siepmann, R. Bodmeier*
Berlin, Kelchstr. 31, 12169 Berlin, Germany
College of Pharmacy, Freie Universitat
Received 18 June 2002; received in revised form 11 October 2002; accepted 22 October 2002

Abstract
The aim of this study was to develop and physicochemically characterize single unit, floating controlled drug delivery systems
consisting of (i) polypropylene foam powder, (ii) matrix-forming polymer(s), (iii) drug, and (iv) filler (optional). The highly porous foam
powder provided low density and, thus, excellent in vitro floating behavior of the tablets. All foam powder-containing tablets remained
floating for at least 8 h in 0.1 N HCl at 37 8C. Different types of matrix-forming polymers were studied: hydroxypropyl methylcellulose
(HPMC), polyacrylates, sodium alginate, corn starch, carrageenan, gum guar and gum arabic. The tablets eroded upon contact with the
release medium, and the relative importance of drug diffusion, polymer swelling and tablet erosion for the resulting release patterns varied
significantly with the type of matrix former. The release rate could effectively be modified by varying the matrix-forming polymer / foam
powder ratio, the initial drug loading, the tablet geometry (radius and height), the type of matrix-forming polymer, the use of polymer
blends and the addition of water-soluble or water-insoluble fillers (such as lactose or microcrystalline cellulose). The floating behavior of
the low density drug delivery systems could successfully be combined with accurate control of the drug release patterns.
2002 Elsevier Science B.V. All rights reserved.
Keywords: Extended drug release; Floating drug delivery system; Foam; HPMC; Matrix tablet

1. Introduction
Different studies reported in the literature indicate that
pharmaceutical dosage forms exhibiting good in vitro
floating behavior show prolonged gastric residence in vivo
(Ichikawa et al., 1991; Kawashima et al., 1991; Atyabi et
al., 1996; Iannuccelli et al., 1998). The physical properties
of the drug delivery system (e.g., density and size) as well
as the presence of food in the stomach have been identified
as the two most important parameters determining the in
vivo performance of the dosage form (Hwang et al., 1998).
Under fasted conditions the stomach is cleared of undigested material every 1.5 to 2 h by housekeeper waves. To
provide good floating behavior in the stomach, the density
of the device should be less than that of the gastric
contents (1.004 g / cm 3 ). However, it has to be pointed
out that good in vitro floating behavior alone is not
sufficient proof for efficient gastric retention in vivo. The
effects of the simultaneous presence of food and of the
*Corresponding author. Tel.: 149-30-8385-0643; fax: 149-30-83850692.
E-mail address: bodmeier@zedat.fu-berlin.de (R. Bodmeier).

complex motility of the stomach are difficult to estimate.

Obviously, only in vivo studies can provide definite proof
that prolonged gastric residence is obtained.
Extended-release dosage forms with prolonged residence
times in the stomach are highly desirable for drugs (i) that
are locally active in the stomach, (ii) that have an
absorption window in the stomach or in the upper small
intestine, (iii) that are unstable in the intestinal or colonic
environment, and / or (iv) have low solubility at high pH
values. In addition, as the total gastrointestinal transit time
of dosage forms is increased by prolonging the gastric
residence time, these systems can also be used as sustained
release devices with a reduced frequency of administration
and, therefore, improved patient compliance. Recent approaches to increase the gastric residence time of drug
delivery systems include (i) bioadhesive devices, (ii)
systems that rapidly increase in size upon swallowing, and
(iii) low density devices that float on the gastric contents
1993; Deshpande et al., 1996; Rouge et al., 1996;
(Moes,
Hwang et al., 1998; Singh and Kim, 2000).
A floating, single unit dosage form with extended
release consisting of a capsule containing a mixture of
drug and hydrocolloids has been described by Sheth and

0928-0987 / 02 / $ see front matter 2002 Elsevier Science B.V. All rights reserved.
PII: S0928-0987( 02 )00223-3

38

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

Tossounian (1978). Upon contact with gastric fluids, the

capsule shell dissolves and a gelled system with a bulk
density of less than one is formed. Based on this principle,
pharmaceutical products containing L-dopa combined with
a decarboxylase inhibitor and diazepam have been commercially developed (Sheth and Tossounian, 1984a,b; Erni
and Held, 1987). Floating tablets containing a mixture of
drug and hydrocolloids that remain in the stomach for an
extended period of time have been described (Sheth and
Tossounian, 1979a,b). Matrix tablets based on hydroxypropyl methylcellulose (HPMC K4M) have been developed by Baumgartner et al. (2000). Upon contact with
gastric fluid, the systems take up water and swell. As the
increase in volume is greater than the increase in mass
during swelling, the densities of these devices decrease and
the systems start to float after a short lag time. The
influence of different processing and formulation parameters on the floating properties of matrix tablets has been
studied (Colombo et al., 1989; Gerogiannis et al., 1993;
Rouge et al., 1997; Baumgartner et al., 1998). Reduced
floating lag times could be achieved by reducing the
compression forces (thus, increasing tablet porosities),
increasing polymer molecular weights and increasing the
particle sizes of the matrix-forming polymer.
An interesting approach to provide floating drug delivery systems is based on the formation of carbon dioxide
within the device upon contact with body fluids. Multilayer matrix tablets have been described containing an
effervescent layer with carbonate and, optionally, citric
acid (Ingani et al., 1987; Yang and Fassihi, 1996; Yang et
al., 1999). Upon contact with acidic aqueous media,
carbon dioxide is generated and entrapped within the
gelling hydrocolloid, causing the system to float. Timmer (1990) quantified the floating capabilities
mans and Moes
of matrix systems based on swellable polymers, including
gas-generating systems. However, long-term floating behavior was not observed with any of the investigated
dosage forms. Floating mini-tablets based on HPMC and
sodium bicarbonate as gas-generating agent have been
developed by Rouge et al. (1998). The floating properties
of these systems containing either piretanide or atenolol as
model drug could be improved by introducing a wet
granulation step. The generated carbon dioxide was entrapped for a longer period within the tablet matrix of the
granulated system compared with the non-granulated system. The floating lag times ranged from 1 to 27 min, and
the floating periods partially exceeded 6 h.

Krogel
and Bodmeier (1999) developed a floating
device consisting of two drug-loaded HPMC matrix tablets, which were placed within an impermeable, hollow
polypropylene cylinder (open at both ends). Each matrix
tablet closed one of the cylinders ends so that an air-filled
space was created in between, providing a low total system
density. The device remained floating until at least one of
the tablets was dissolved.
A floating drug delivery system, being less dense than

gastric juice due to the incorporation of at least one porous

structural element, such as foam or a hollow body, was

described by Muller
and Anders (1989). Recently, we
proposed a new type of multiparticulate floating drug
delivery system consisting of a highly porous carrier
material (foam powder), drug and polymer: low density
microparticles (Streubel et al., 2002). Verapamil HCl, a
drug with a strongly pH-dependent solubility (poorly
soluble at high pH values, highly soluble at low pH
values), was used as the model drug to demonstrate the
performance of this system in vitro. The microparticles
were prepared using a solvent evaporation method.
Different mass transport processes may occur during
drug release from polymer-based matrix tablets, including
(i) water imbibition into the system, (ii) polymer swelling,
(iii) drug dissolution, (iv) drug diffusion out of the tablet,
and (v) polymer dissolution. Depending on the type of
drug, polymer and release medium and on the tablet
composition, the respective processes are more or less
important (Siepmann and Peppas, 2001; Siepmann et al.,
2002). Velasco et al. (1999) reported that the rate and
mechanism of diclofenac sodium release from HPMC
K15M-based matrices were mainly controlled by the drug /
HPMC ratio, and that drug release was independent of the
compression force in the range between 3 and 12 kN. The
effects of the two formulation variables HPMC / lactose
ratio and HPMC viscosity grade on the release of
adinazolam mesylate from cylindrical tablets was studied
by Sung et al. (1996). The resulting drug release rate was
found to increase with decreasing HPMC / lactose ratio
and decreasing HPMC viscosity grade. Colombo et al.
(1990) varied the drug release rate from HPMC-based
matrix tablets by physically restricting the swelling of the
polymer. Different surface portions of HPMC tablets were
covered with impermeable coatings. The underlying drug
release mechanisms and the influence of the type of
coating on the resulting release rate were investigated. To
facilitate the industrial production of this type of drug
delivery system, the manual film-coating process can be
replaced by press-coating techniques (Conte et al., 1993).
The major objectives of the present study were (i) to
develop a single unit, floating drug delivery system
consisting of low density polypropylene foam powder,
matrix-forming polymer(s), drug, and filler (optional), and
(ii) to study the effect of important formulation and
processing parameters on the floating and drug release
behavior of these systems.

2. Materials and methods

2.1. Materials
Polypropylene foam powder (Accurel MP 1002 and
MP 1000, Membrana, Obernburg, Germany), chlorpheniramine maleate (CPM; ICN Biomedicals, Aurora, OH,

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

USA), diltiazem HCl (Godecke,

Freiburg, Germany),

theophylline anhydrous (Synopharm, Barsbuttel,

Germany), verapamil HCl (Knoll, Ludwigshafen, Germany),
hydroxypropyl methylcellulose (HPMC; Methocel E5,
E50 and K15M, Colorcon, Orpington, UK), carrageenan
(type CHP-2; Hercules, Lille Skensved, Denmark), corn
starch, gum guar (Sigma, St. Louis, MO, USA), gum
arabic (Caelo, Caesar and Loretz, Hilden, Germany),
polyacrylic acids [Carbopol 934P (polymerized in benzene, highly crosslinked with allyl sucrose), Carbopol
971P (polymerized in ethyl acetate, lightly crosslinked
with allyl pentaerythritol), Carbopol 974P (polymerized
in ethyl acetate, highly crosslinked with allyl pentaerythritol), and Noveon AA1 (polycarbophil; polymerized in
ethyl acetate, crosslinked with divinyl glycol), Noveon,
Cleveland, OH, USA], sodium alginate (Protanal LF 20 /
200; Pronova Biopolymer, Drammen, Norway), dibasic
calcium phosphate (Emcompress ; E. Mendell, Patterson,
NY, USA), lactose (a-lactose monohydrate, Flowlac 100;
Meggle, Wasserburg, Germany), microcrystalline cellulose
(MCC, Avicel PH-101; FMC Corporation, c / o Lehmann
und Voss, Hamburg, Germany), magnesium stearate
(Herwe Chemisch-technische Erzeugnisse, Sinsheim
Duhren,
Germany). The polypropylene foam powder was
sieved to obtain different size fractions, all other materials
were used as received.

39

2.4. In vitro drug release

In vitro drug release studies were conducted by placing
the tablets in 500 ml plastic containers filled with 300 ml
preheated release medium (0.1 N HCl, pH 1.2, 37 8C),
followed by horizontal shaking for 8 h (37 8C, 75 rpm,
Labortechnik, Burn53; GFL 3033, Gesellschaft fur
gwedel, Germany). The amount of drug released was
detected UV-spectrophotometrically at the following wavelengths: CPM, l 5 264 nm; diltiazem HCl, l 5 236 nm;
theophylline, l 5 270 nm; verapamil HCl, l 5 278 nm
(UV-2101 PC, Shimadzu Scientific Instruments, Columbia,
MD, USA).

2.5. Density measurements

The apparent densities of the tablets were calculated
from their volumes and masses (n56). The volumes V of
the cylindrical tablets were calculated from their heights h
and radii r (both determined with a micrometer gauge)
using the mathematical equation for a cylinder (V 5 p 3
r 2 3 h). The density of 0.1 N HCl at 37 8C was determined
with a pycnometer (n53).

3. Results and discussion

3.1. Floating behavior

2.2. Tablet preparation
Tablets containing 0.5% w / w magnesium stearate as
lubricant were prepared by direct compression. The respective powders [drug, foam powder, polymer(s) and optional
additives, compositions listed in Table1] were blended
thoroughly with a mortar and pestle. 8505 mg of the
mixture was weighed and fed manually into the die of an
instrumented single-punch tableting machine (EK0,
Korsch, Berlin, Germany) to produce tablets using flatfaced punches (2, 9, 12, or 16 mm in diameter). The
hardness of the tablets was kept constant (approximately
80 N, unless otherwise stated) and was measured with a
hardness tester (PTB 311, Pharma Test, Hainburg, Germany).

2.3. Floating behavior of the tablets

The in vitro floating behavior of the tablets was studied
by placing them in 500 ml plastic containers filled with
300 ml preheated 0.1 N HCl (pH 1.2, 37 8C), followed by
horizontal shaking for 8 h (37 8C, 75 rpm, n53; GFL
Labortechnik, Burgwedel, Ger3033, Gesellschaft fur
many). The floating lag times (time period between
placing the tablet in the medium and tablet floating) and
floating durations of the tablets were determined by visual
observation.

The structure of the floating tablets is shown schematically in Fig. 1. Incorporation of the highly porous foam
powder in the matrix tablets provides densities that are
lower than the density of the release medium [0.690.98
g / cm 3 (Table1), compared with 1.00 g / cm 3 for the release
medium]. 17% w / w foam powder (based on the mass of
the tablet) was sufficient to achieve proper in vitro floating
behavior for at least 8 h. In contrast to most conventional
floating systems (including gas-generating ones), these
tablets floated immediately upon contact with the release
medium, showing no lag times in floating behavior because the low density is provided from the beginning
(t 5 0). Extended floating times are achieved due to the air
entrapped within the foam powder particles, which is only
slowly removed from the system upon contact with the
release medium.
As expected, tablets without polypropylene foam powder (e.g., consisting of 240 mg HPMC K15M and 120 mg
verapamil HCl) first sank before floating, showing floating
lag times of between 9 and 33 min. Replacing only 8%
w / w (based on the mass of the tablet) of the HPMC with
the foam powder reduced the lag times to 2 min.

3.2. In vitro drug release

The drug release decreased when reducing the amount
of foam powder from 180 to 0 mg and simultaneously

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

40

Table 1
Compositions of the investigated tablets (all quantities are given in mg) and densities at a tablet hardness of 80 N
Formulation No.

Verapamil HCl
CPM
Diltiazem HCl
Theophylline
HPMC K15M
HPMC E50
HPMC E5
Carbopol 934P
Carbopol 974P
Carbopol 971P
Noveon AA1
Na-alginate
Corn starch
Carrageenan
Gum guar
Gum arabic
Lactose
MCC
Emcompress
Foam powder,
125160 mm
Density (g/cm 3 )

10

11

12

13

120

240

120

150

120

60

120

150

120

150

120
150

18

252

36

234

108

162

180

90

42

52

92

115

168

210

90

180

90

90

90

90

90

90

90

31

69

126

1.02

0.87

0.72

0.94

0.90

0.93

0.86

0.86

0.87

0.89

0.85

0.93

0.85

14
3

2
0.98

15

16

17

18

19

20

21

22

23

53

66

157

196

120

150

120

150

120

150

120

100

50

120

100

50

120

100

50

120

50

100

40

118

90

90

90

90

90

90

90

0.96

0.89

0.89

0.89

0.69

0.93

0.89

0.93

0.90

Formulation No.

Verapamil HCl
CPM
Diltiazem HCl
Theophylline
HPMC K15M
HPMC E50
HPMC E5
Carbopol 934P
Carbopol 974P
Carbopol 971P
Noveon AA1
Na-alginate
Corn starch
Carrageenan
Gum guar
Gum arabic
Lactose
MCC
Emcompress
Foam powder,
125160 mm
Density (g/cm 3 )

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

120

75

75

120

20

30

100

120

20

30

100

120

20

30

100

120

20

30

100

120

20

30

100

120

20

30

100

120

20

30

100

120

20

30

100

120

20

30
100

120

40

10

100

120

30

20

100

120

50

100

120

40

10

100

120

30

20

100

120

50

100

90

90

90

90

90

90

90

90

90

90

90

90

90

90

90

90

0.93

0.87

0.92

0.92

0.91

0.92

0.94

0.91

0.93

0.93

increasing the amount of HPMC K15M from 60 to 240 mg

(keeping the amount of drug constant: 120 mg) (Fig. 2).
This can probably be attributed to the different properties
of the polymer networks through which the drug must
diffuse. Polypropylene can be regarded as impermeable for
the drug (extremely low drug diffusion coefficients). Thus,
drug diffusion is restricted to water-filled pores within the

0.88

0.88

0.85

0.90

0.91

0.92

foam powder particles and to the swollen HPMC hydrogel.

With decreasing amounts of HPMC, the density of the
swollen hydrogel network decreases, presenting less hindrance for drug diffusion. Consequently, the drug release
rates increase. The shape of the observed drug release
curves from foam powder-containing and foam powderfree tablets is very similar, which indicates that the swollen

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

41

Fig. 1. Schematic presentation of the structure of the low density, floating

matrix tablets.

hydrogel network and not the foam powder predominantly

controls drug release.

3.2.1. Effect of the type of drug and initial drug loading

Drug release strongly depended on the type of drug,
decreasing in the rank order CPM.diltiazem HCl.
verapamil HCltheophylline (Fig. 3A). The reasons for
these differences are not straightforward and can probably
be related to various overlapping factors, such as (i) the
solubilities of the drugs within the bulk fluid [574, 588,
392 and 15.4 mg / ml in 0.1 N HCl at 37 8C for CPM
(Bodmeier and Paeratakul, 1991), diltiazem HCl (Bodmeier et al., 1996), verapamil HCl (Streubel et al., 2000)
and theophylline (Bodmeier and Chen, 1989), respectively], (ii) drug molecular weights [274.79, 414.53, 454.61

Fig. 3. Effect of (A) the type of drug (initial radius 6 mm, 360 mg tablet
weight / 120 mg drug, 150 mg polymer, 90 mg polypropylene foam
powder; compositions in Table1, formulation Nos. 2, 4, 5 and 6); and (B)
the initial verapamil HCl loading (initial radius 6 mm, 360 mg tablet
weight; compositions in Table1, formulation Nos. 7 to 10) on the release
patterns from HPMC K15M-based floating matrix tablets in 0.1 N HCl.

Fig. 2. Effect of the HPMC K15M / polypropylene foam powder ratio

on verapamil HCl release in 0.1 N HCl from matrix tablets (initial radius
6 mm, 360 mg tablet weight / 120 mg drug, polymer / foam powder ratio
given in the figure; compositions in Table1, formulation Nos. 1 to 3).

and 180.17 Da for CPM, diltiazem, verapamil and theophylline, respectively (The Merck Index, 1996)], which
affect the drug diffusivities within the swollen polymeric
networks, (iii) the drug dissolution rates, and (iv) polymerdrug interactions. A detailed analysis of these phenomena is beyond the scope of this study.
The initial drug loading of the low density matrix tablets
significantly affected the resulting drug release rate in 0.1
N HCl (Fig. 3B). The release rate, both in mg / time unit
and in % / time unit (data not shown), increased with
increasing drug content. This can again be explained by the
different properties of the swollen hydrogel networks. At
5% w / w drug loading, the tablet consists mainly of HPMC
(70% w / w). Thus, upon water imbibition a relatively tight
macromolecular network results, presenting a significant

42

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

hindrance for drug diffusion. With increasing drug loadings, the relative HPMC content decreases and, thus, the
tightness of the swollen hydrogel network decreases.
Consequently, the drug diffusivity and release rate increase.

3.2.2. Effect of the tablet geometry

A simple, but very effective tool for modifying the
release kinetics from matrix tablets is to vary their
geometry. Varying the initial radius and height of cylindrical tablets strongly affects the resulting drug release rate,
which can be predicted theoretically (Siepmann et al.,
2002). The release rate of verapamil HCl from HPMC
K15M-based devices with 25% w / w foam powder and
33% w / w drug loading as a function of initial tablet height
(1.35.2 mm at a constant tablet radius of 6 mm) is shown
in Fig. 4A. The absolute release rate of the drug increased
with increasing tablet height. This can be attributed to the
higher absolute verapamil HCl amounts incorporated within the system with increasing matrix volume. When
plotting the respective relative amounts of drug released in
percent versus time, it can be seen that the relative drug
release rate decreases with increasing tablet height (data
not shown). This can probably be explained by the
decreasing relative surface area of the matrices with
increasing tablet height (Siepmann et al., 2002).
The tablet radius (1.08.0 mm at a constant tablet height
of 2.6 mm) also had a very pronounced effect on the
absolute drug release rate (Fig. 4B). With increasing initial
tablet radius, the volume of the system and, thus, the
amount of drug available for diffusion increases, resulting
in increased absolute amounts of drug released. In contrast,
the relative surface area of the device decreases, and the
amount of drug released in % / time unit decreased (data
not shown).
3.2.3. Effect of type of matrix-forming polymer
The effect of the type of matrix polymer (HPMC E5,
HPMC E50, HPMC K15M, or Carbopol 934P) used for
the preparation of floating, low density tablets on the
resulting drug release kinetics is shown in Fig. 5. The three
HPMC types / grades differ in the type of substitution
and / or molecular weight (which can be correlated with the
polymer viscosity): Methocel E and K contain 2830 and
1924% methoxyl groups; the viscosities of 2% aqueous
solutions of HPMC E5, E50 and K15M at 20 8C are 5, 50,
and 15,000 cps, respectively. The drug release rate decreased in the rank order HPMC E5.HPMC E50.HPMC
K15M.Carbopol 934P. This can probably be attributed to
the different diffusion and swelling behavior in / of these
polymers. With increasing macromolecular weight, the
degree of entanglement of the polymer chains increases.
Thus, the mobility of the macromolecules in the fully
swollen systems decreases. According to the free volume
theory of diffusion, the probability for a diffusing molecule
to jump from one cavity into another, hence, decreases

Fig. 4. Effect of tablet geometry on verapamil HCl release from HPMC

K15M-containing floating matrix tablets in 0.1 N HCl. (A) Effect of tablet
height (initial tablet radius 6 mm, initial tablet height given in the figure,
33% w / w initial drug loading, 25% w / w polypropylene foam powder;
compositions in Table1, formulation Nos. 11 to 13). (B) Effect of tablet
radius (initial tablet height 2.6 mm, initial tablet radius given in the
figure, 33% w / w initial drug loading, 25% w / w polypropylene foam
powder; compositions in Table1, formulation Nos. 14 to 16).

(Fan and Singh, 1989). This leads to decreased drug

diffusion coefficients and decreased drug release rates with
increasing molecular weights. In addition to the effect on
the polymer swelling behavior, the increase in polymer
molecular weight also modifies the polymer dissolution
behavior. With increasing macromolecular weight the
dissolution rate decreases, thus the drug release decreases.
The different polymer dissolution behaviors could be
observed visually by the release of the low density
polypropylene foam powder. Clearly, the dissolution of
HPMC E5-containing tablets was faster than that of HPMC
E50-based systems. Polypropylene foam powder release
from HPMC K15M-containing tablets was very slow, and
practically no foam powder was released from the Car-

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

43

Fig. 5. Effect of the type of matrix-forming polymer on verapamil HCl

release in 0.1 N HCl from floating matrix tablets (initial radius 6 mm, 360
mg tablet weight / 120 mg drug, 90 mg foam powder, 150 mg polymer;
compositions in Table1, formulation Nos. 2, 17, 18 and 19).

bopol 934P-based system during the observation period.

This can be explained by the different chemical structures
of the polymers (substitution patterns, type of polymer
backbones).

3.2.4. Effect of the type and amount of filler

The effect of adding water-soluble (lactose) and waterinsoluble [microcrystalline cellulose (MCC) and dibasic
calcium phosphate (Emcompress)] fillers to low density
matrix tablets containing verapamil HCl, polypropylene
foam powder, and HPMC K15M on the resulting drug
release kinetics is shown in Fig. 6A. Clearly, the release
rate increased when adding the fillers, however no differences were seen between the three different fillers at the
investigated polymer / filler ratio of 2:1. The slight increase
in drug release can probably be explained by the decreasing relative HPMC amounts and, thus, the less tight
hydrogel structures upon swelling. Thus, the physicochemical properties of the filler do not seem to affect the
underlying drug release mechanisms and release rates in
the present case. Vargas and Ghaly (1999) did also not
observe any significant difference between lactose, MCC
and Emcompress when used as filler in theophyllinecontaining HPMC K4M-based matrix tablets with respect
to the resulting drug release kinetics (containing up to 59%
w / w filler). Thus, HPMC is clearly the dominating compound controlling the release rate of the drug in the
investigated low density matrix tablets.
When increasing the amount of added lactose, while
simultaneously reducing the amount of the matrix-forming
polymer (up to a ratio of 1:2 HPMC K15M / lactose,
keeping the total tablet weight constant), the drug release
rate in 0.1 N HCl significantly increased (Fig. 6B). As

Fig. 6. Effect of (A) adding different types of fillers (initial radius 6 mm,
360 mg tablet weight / 120 mg drug, 90 mg foam powder, 100 mg HPMC
K15M, 50 mg filler; compositions in Table1, formulation Nos. 2, 20, 21
and 22) and (B) the HPMC K15M / lactose ratio (initial radius 6 mm,
360 mg tablet weight / 120 mg drug, 90 mg foam powder, polymer /
lactose ratio given in the figure; compositions in Table1, formulation Nos.
2, 20, 23 and 24) on verapamil HCl release in 0.1 N HCl from floating
matrix tablets.

discussed above, this can probably be attributed to the

decreasing tightness of the swollen hydrogel.

3.2.5. Effect of using blends of matrix-forming polymers

Instead of using only a single polymer, blends of
different macromolecules can also serve as matrix formers.
A broad range of drug release behaviors was obtained from
matrix tablets prepared with blends of HPMC K15M and
various other hydrogel formers (Fig. 7A). Gum arabic,
gum guar, carrageenan and corn starch used as second
hydrogel former led to a rather rapid drug release (more
than 89% of the drug was released within the first 2 h).
Thus, these systems cannot provide extended drug delivery
over prolonged periods of time, probably due to rapid

44

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

partial tablet disintegration (visual observation) and / or

slower swelling of these polymers (resulting in a lack of
contribution to hydrogel formation). However, with sodium
alginate, NoveonAA1 and the different Carbopol types,
sustained drug release was achieved, similar to systems
based on only HPMC.
The effect of varying the blend ratio on drug release for
two hydrogel former combinations (HPMC K15M / Carbopol 934P and HPMC K15M / sodium alginate) is illustrated in Fig. 7B and C. The effect of the blend ratio was
much more pronounced in the case of HPMC K15M /
Carbopol (Fig. 7B) mixtures compared to the HPMC
K15M / sodium alginate blends (Fig. 7C). As discussed
above, drug release from tablets containing only Carbopol
934P was more sustained than from HPMC K15M-based
systems (Fig. 5). All blends of these polymers showed
intermediate behavior. Interestingly, drug release from 4:1
and 3:2 blends (HPMC K15M / Carbopol) was very similar
to the pure HPMC K15M systems. Thus, HPMC K15M is
the dominant compound, controlling drug release in these
blends. However, if more Carbopol than HPMC is present
in the system, this domination no longer holds. Intermediate release between the two pure polymers is obtained
in the case of 2:3 blends. Interestingly, the release of
verapamil HCl from pure HPMC K15M and pure sodium
alginate-based systems was faster than from tablets containing the respective polymer blends (Fig. 7C). This
might be explained by polymerpolymer interactions
between HPMC K15M and sodium alginate. An exact
analysis of this phenomenon is beyond the scope of this
study.

3.2.6. Effect of the foam powder particle size and the

tablet hardness
The effect of the particle size of the polypropylene foam
powder on the resulting densities of the investigated
systems and on drug release was not very pronounced in
the range between 125 and 1000 mm (data not shown).
Thus, the foam powder particle size is not a crucial
parameter affecting drug release.
The release rate slightly increased with decreasing tablet
hardness from 160 to 40 N (data not shown). This can
probably be attributed to an increase in the porosity of the
system and / or to faster disintegration / dissolution upon
water imbibition.

Fig. 7. Effect of (A) the use of blends of matrix-forming polymers (initial

radius 6 mm, 360 mg tablet weight / 120 mg drug, 90 mg foam powder, 20
mg HPMC K15M, 30 mg second hydrogel former as given in the figure,
100 mg lactose; compositions in Table1, formulation Nos. 25 to 33), and
the matrix-forming polymer blend ratio on verapamil HCl release in 0.1 N
HCl from floating tablets: (B) blends of HPMC K15M and Carbopol
934P; and (C) blends of HPMC K15M and sodium alginate (initial radius
6 mm, 360 mg tablet weight / 120 mg drug, 90 mg foam powder,
polymer / polymer blend ratio given in the figure, 100 mg lactose;
compositions in Table1, formulation Nos. 23, 25, 29, 34 to 39).

4. Conclusions
In conclusion, a single unit, floating drug delivery
system has been developed, which is based on low density
foam powder and matrix-forming polymer(s). Its in vitro
floating performance and the ability to control drug release
over prolonged periods of time have been demonstrated.
The drug release patterns can effectively be adjusted by
varying simple formulation parameters, such as the ma-

A. Streubel et al. / European Journal of Pharmaceutical Sciences 18 (2003) 3745

trix-forming polymer / foam powder ratio, initial drug

loading, tablet height and diameter, type of matrix-forming
polymer, addition of water-soluble and water-insoluble
fillers, and the use of polymer blends. Thus, desired release
profiles adapted to the pharmacokinetic / pharmacodynamic
properties of the incorporated drug can easily be provided.

References
Atyabi, F., Sharma, H.L., Mohammad, H.A.H., Fell, J.T., 1996. In vivo
evaluation of a novel gastric retentive formulation based on ion
exchange resins. J. Control. Release 42, 105113.
Baumgartner, S., Smid-Korbar, J., Vrecer, F., Kristl, J., 1998. Physical and
technological parameters influencing floating properties of matrix
tablets based on cellulose ethers. STP Pharma Sci. 8, 285290.
Baumgartner, S., Kristl, J., Vrecer, F., Vodopivec, P., Zorko, B., 2000.
Optimisation of floating matrix tablets and evaluation of their gastric
residence time. Int. J. Pharm. 195, 125135.
Bodmeier, R., Chen, H., 1989. Evaluation of biodegradable poly(lactide)
pellets prepared by direct compression. J. Pharm. Sci. 78, 819822.
Bodmeier, R., Paeratakul, O., 1991. Process and formulation variables
affecting the drug release from chlorpheniramine maleate-loaded beads
coated with commercial and self-prepared aqueous ethyl cellulose
pseudolatexes. Int. J. Pharm. 70, 5968.
Bodmeier, R., Guo, X., Sarabia, R.E., Skultety, P.F., 1996. The influence
of buffer species and strength on diltiazem HCl release from beads
coated with the aqueous cationic polymer dispersions, EudragitRS, RL
30D. Pharm. Res. 13, 5256.
Colombo, P., Provasi, D., Borazzo, M.G., Maggi, L., Catellani, P.L., 1989.
The role of compression force in floating tablet formula optimization.
Acta Pharm. Technol. 35, 168170.
Colombo, P., Conte, U., Gazzaniga, A., Maggi, L., Sangalli, M.E.,
Peppas, N.A., La Manna, A., 1990. Drug release modulation by
physical restrictions of matrix swelling. Int. J. Pharm. 63, 4348.
Conte, U., Maggi, L., Colombo, P., La Manna, A., 1993. Multi-layered
hydrophilic matrices as constant release devices (Geomatrix Systems). J. Control. Release 26, 3947.
Deshpande, A.A., Rhodes, C.T., Shah, N.H., Malick, A.W., 1996.
Controlled-release drug delivery systems for prolonged gastric residence: an overview. Drug Dev. Ind. Pharm. 22, 531539.
Erni, W., Held, K., 1987. The hydrodynamically balanced system: a novel
principle of controlled drug release. Eur. Neurol. 27 (Suppl. 1), 2127.
Fan, L.T., Singh, S.K., 1989. Controlled Release: A Quantitative Treatment. Springer, Berlin.
Gerogiannis, V.S., Rekkas, D.M., Dallas, P.P., Choulis, N.H., 1993.
Floating and swelling characteristics of various excipients used in
controlled release technology. Drug Dev. Ind. Pharm. 19, 10611081.
Hwang, S.J., Park, H., Park, K., 1998. Gastric retentive drug-delivery
systems. Crit. Rev. Ther. Drug Carrier Syst. 15, 243284.
Iannuccelli, V., Coppi, G., Sansone, R., Ferolla, G., 1998. Air compartment multiple-unit system for prolonged gastric residence. Part II. In
vivo evaluation. Int. J. Pharm. 174, 5562.
Ichikawa, M., Kato, T., Kawahara, M., Watanabe, S., Kayano, M., 1991.
A new multiple-unit oral floating dosage system. II: In vivo evaluation
of floating and sustained-release characteristics with p-aminobenzoic
acid and isosorbide dinitrate as model drugs. J. Pharm. Sci. 80,
11531156.
A.J., 1987. Conception and in vivo
Ingani, H.M., Timmermans, J., Moes,
investigation of peroral sustained release floating dosage forms with
enhanced gastrointestinal transit. Int. J. Pharm. 35, 157164.
Kawashima, Y., Niwa, T., Takeuchi, H., Hino, T., Ito, Y., 1991. Prepara-

45

tion of multiple unit hollow microspheres (microballoons) with acrylic

resin containing tranilast and their drug release characteristics (in
vitro) and floating behavior (in vivo). J. Control. Release 16, 279290.

Krogel,
I., Bodmeier, R., 1999. Development of a multifunctional matrix
drug delivery system surrounded by an impermeable cylinder. J.
Control. Release 61, 4350.
A.J., 1993. Gastroretentive dosage forms. Crit. Rev. Ther. Drug
Moes,
Carrier Syst. 10, 143195.

Muller,
W., Anders, E., 1989. Floating system for oral therapy. WO Patent
89 / 06956.
Rouge, N., Buri, P., Doelker, E., 1996. Drug absorption sites in the
gastrointestinal tract and dosage forms for site-specific delivery. Int. J.
Pharm. 136, 117139.
Rouge, N., Cole, E.T., Doelker, E., Buri, P., 1997. Screening of
potentially floating excipients for minitablets. STP Pharma Sci. 7,
386392.
Rouge, N., Cole, E.T., Doelker, E., Buri, P., 1998. Buoyancy and drug
release patterns of floating minitablets containing piretanide and
atenolol as model drugs. Pharm. Dev. Technol. 3, 7384.
Sheth, P.R., Tossounian, J.L., 1978. Sustained release pharmaceutical
capsules. U.S. Patent 4,126,672.
Sheth, P.R., Tossounian, J.L., 1979a. Sustained release tablet formulations. U.S. Patent 4,140,755.
Sheth, P.R., Tossounian, J.L., 1979b. Novel sustained release tablet
formulations. U.S. Patent 4,167,558.
Sheth, P.R., Tossounian, J.L., 1984a. Hydrodynamically balanced controlled release compositions containing L-dopa and a decarboxylase
inhibitor. U.S. Patent 4,424,235.
Sheth, P.R., Tossounian, J., 1984b. The hydrodynamically balanced
system (HBS): a novel drug delivery system for oral use. Drug Dev.
Ind. Pharm. 10, 313339.
Siepmann, J., Peppas, N.A., 2001. Modeling of drug release from delivery
systems based on hydroxypropyl methylcellulose (HPMC). Adv. Drug
Deliv. Rev. 48, 139157.
Siepmann, J., Streubel, A., Peppas, N.A., 2002. Understanding and
predicting drug delivery from hydrophilic matrix tablets using the
sequential layer model. Pharm. Res. 19, 306314.
Singh, B.N., Kim, K.H., 2000. Floating drug delivery systems: an
approach to oral controlled drug delivery via gastric retention. J.
Control. Release 63, 235259.
Streubel, A., Siepmann, J., Dashevsky, A., Bodmeier, R., 2000. pHindependent release of a weakly basic drug from water-insoluble and
-soluble matrix tablets. J. Control. Release 67, 101110.
Streubel, A., Siepmann, J., Bodmeier, R., 2002. Floating microparticles
based on low density foam powder. Int. J. Pharm. 241, 279292.
Sung, K.C., Nixon, P.R., Skoug, J.W., Ju, T.R., Gao, P., Topp, E.M., Patel,
M.V., 1996. Effect of formulation variables on drug and polymer
release from HPMC-based matrix tablets. Int. J. Pharm. 142, 5360.
The Merck Index, XII, 1996. Merck & Co., Inc., Whitehouse Station, NJ.
A.J., 1990. How well do floating dosage forms
Timmermans, J., Moes,
float? Int. J. Pharm. 62, 207216.
Vargas, C.I., Ghaly, E.S., 1999. Kinetic release of theophylline from
hydrophilic swellable matrices. Drug Dev. Ind. Pharm. 25, 1045
1050.
Velasco, M.V., Ford, J.L., Rowe, P., Rajabi-Siahboomi, A.R., 1999.
Influence of drug:hydroxypropylmethylcellulose ratio, drug and polymer particle size and compression force on the release of diclofenac
sodium from HPMC tablets. J. Control. Release 57, 7585.
Yang, L., Fassihi, R., 1996. Zero-order release kinetics from a selfcorrecting floatable asymmetric configuration drug delivery system. J.
Pharm. Sci. 85, 170173.
Yang, L., Eshraghi, J., Fassihi, R., 1999. A new intragastric delivery
system for the treatment of Helicobacter pylori associated gastric
ulcer: in vitro evaluation. J. Control. Release 57, 215222.