International Journal of Pharmaceutics 454 (2013) 4757

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

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

Hydrogel blends with adjustable properties as patches for

transdermal delivery
Stefania Mazzitelli a,1 , Cinzia Pagano b,1 , Danilo Giusepponi b ,
Claudio Nastruzzi a, , Luana Perioli b
a
b

Department of Life Sciences and Biotechnology, University of Ferrara, Ferrara, Italy

Department of Chemistry and Pharmaceutical Technology, University of Perugia, Perugia, Italy

a r t i c l e

i n f o

Article history:
Received 15 March 2013
Received in revised form 26 June 2013
Accepted 30 June 2013
Available online 12 July 2013
Keywords:
Gelatin
Pectin
Hydrogels
Transdermal delivery formulation

a b s t r a c t
The effect of different preparation parameters were analyzed with respect to the rheological and pharmaceutical characteristics of hydrogel blend patches, as transdermal delivery formulation. Mixtures of
pectin and gelatin were employed for the production of patches, with adjustable properties, following
a two-step gelation procedure. The rst gelation, a thermal one, is trigged by the presence of gelatin,
whereas, the second gelation, an ionic one, is due to the formation of the typical egg box structure
of pectin. In particular, the patch structural properties were assessed by oscillation stress sweep measurements which provided information concerning their viscolelastic properties. In addition, different
modalities for drug loading were analyzed with respect to drug homogeneous distribution; testosterone
was employed as model drug for transdermal administration. Finally, the performances of the produced
transdermal patches were studied, in term of reproducibility and reliability, by determination of in vitro
drug release proles.
2013 Elsevier B.V. All rights reserved.

1. Introduction
Many active pharmaceutical ingredients, including gene drugs,
biotech products and sex hormones are almost exclusively
delivered by parenteral administration, because they result ineffective when orally administered in reason of their pre-systemic
metabolism due to the acidic conditions of the stomach and degradation by gastrointestinal enzymes leading to poor bioavailability
(Lee and Knighton, 2008). As alternative route of administration,
the generally employed parenteral delivery is unfortunately invasive and requires supervision of health care professionals, causing
low patient compliance.
In order to overcome these drawbacks, new routes and formulations for the delivery of various labile drugs have been proposed,
such as buccal, nasal, vaginal and transdermal formulations, as
alternative to conventional administration strategies (Yoo and Lee,
2006).
In this respect, transdermal delivery system (TDS) can provide
some desirable advantages over conventional pharmaceutical

Corresponding author at: Department of Life Sciences and Biotechnology, University of Ferrara, via Fossato di Mortara 17/19, 44100 Ferrara,
Italy. Tel.: +39 0532 455255.
E-mail address: nas@unife.it (C. Nastruzzi).
1
These authors equally contributed to the study.
0378-5173/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ijpharm.2013.06.081

dosage forms, including: (i) the avoidance of gut and hepatic rstpass metabolism, (ii) the reduction of dose frequency trough multi
day therapy regimes by a single application, (iii) the maintenance of
stable drug delivery proles, resulting in a substantial decrease of
over or under dosing, (iv) the prolongation of the activity for drugs
characterized by a short half-life, (v) the possibility of an ease and
immediate cessation of drug administration by patch removal, (vi)
the rapid identication, in emergencies, of the medication regimes
in unconscious or comatose patients and nally (vii) the alternative
administration route for patients unable to take oral medications
(Berti and Lipsky, 1995; Prausnitz and Langer, 2008).
TDS formulations are usually represented by ointments, semisolid emulsions or polymeric patches (Subedi et al., 2010), in
particular, the latter, in spite of being largely preferred by
pharmaceutical industry, still requires substantial improvements.
Transdermal patches can indeed cause different adverse reactions
such as irritation due to skin occlusion and impairment of water
vapour permeation, intense pain when peeled off from skin and
they require complex procedures for the preparation. Generally, oil
based formulations are easy to use but they have a scarce patient
acceptability since they leave a greasy layer on the skin, after application (Brown et al., 2006).
Taking into account the above reported considerations, the main
aim of the current paper is the development of new transdermal
formulations based on the combined use of two natural hydrophilic
polymers resulting in biocompatible, composite patches allowing

48

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

the modulation of the rheological properties and the drug release

prole. Furthermore, the patches have been designed for possibly
decreasing the skin irritation and improving the patient acceptability. Gels based on hydrophilic polymers offer indeed important
advantages of being easily prepared without the need of organic
solvents, cross-linkers and catalysts and of being suitable for a variety of drugs (Hoffman, 2012). Notably, TDS are usually prepared by
synthetic polymers such as poly(2-hydroxy ethyl methacrylate),
poly(N-vinyl pyrrolidone), poly(methyl methacrylate), poly(vinyl
alcohol) and poly(ethylene glycol) (Kandavilli et al., 2002).
The choice of the most appropriate polymeric composition
is essential to guarantee good patch characteristics in terms of
mechanical properties and drug release kinetics. Both during the
manufacturing process and at the time of application, the patch
is subjected to mechanical solicitations that can compromise its
integrity and nal performance. For this reason the mechanical
properties play an important role in the patch nal performances
since they should possess an adequate exibility to avoid breaking
(Kandavilli et al., 2002).
The tailoring of TDS mechanical properties could be achieved by
polymer based hydrogels; their three-dimensional network structure is indeed capable of imbibing large amounts of water or
biological uids, resulting in a rheological behavior composed of
viscous and elastic components, depending on polymer composition/concentration (Peppas et al., 2000).
As model drug, testosterone was chosen since it represents
a typical example of drug possessing the general characteristics
suitable for transdermal delivery. For instance, testosterone is
largely used in the therapy for natural androgen level restoration
in patients with primary or secondary hypogonadal disorders and
to treat symptoms of partial androgen deciency in aging men
(andropause) (Geurts and Coelingh Bennink, 2000; Houwing et al.,
2003). After oral administration, the drug shows a low bioavailability due to liver pre-systemic metabolism, phenomenon that
could be avoided by formulation in gels, subcutaneous implants,
parenteral and transdermal formulations (Wokovich et al., 2006).
Summarizing, the current paper describes: (i) the preparation
and the rheological characterization of hydrogel patches based on
two natural polymers with different chemical and gelling characteristics (i.e. low-methoxyl pectin and gelatin), (ii) the loading of
testosterone (as model drug) into patches, (iii) the dissolution test
of patches for the determination of testosterone release prole and
(iv) the mathematical analysis of the experimental release kinetics.

of pectin and reaching the maximum clearness. Thereafter, the

pectin dispersion was subjected to 4 consecutive ltration steps
through pharmaceutical grade cellulose Durapore membrane
lters, with decreasing nominal pore size: 5, 1.2, 0.8 and 0.4 m,
using a stainless steel barrel, base, lter support screen, top cap
and tubing adapter. Before ltration, the lters were wetted with
EtOH in order to reduce pectin microbial charge and to prolong its
stability. The ltration rate was set at very low values (typically
10 mL/min) to allow a greater interaction between material and
lters; the ltrate was collected into polycarbonate plastic bottles.
Gelatin powder was then added, under magnetically stirring
(300 rpm), to the puried pectin dispersions maintained at 60 C.
After complete gelatin dissolution occurred (typically in 30 min),
the obtained polymeric mixture was immediately ltered through
disposable sterile syringe lter (0.22 m pore size), the ltrate
was then maintained under vacuum for 15 min at 60 C, in order
to remove the air bubbles generated during the stirring/ltration
processes; 2 mL of the degassed dispersion were poured into each
of the 2.4 cm diameter dish of a 24-multiwell culture plate (Coring
Incorporated, New York, USA). The hydrogel patches were obtained
by a two-step procedure, namely: a thermal and an ionic gelation
process, that were respectively trigged by the presence of gelatin
and pectin. The rst step was accomplished by maintaining the
multiwell culture plates containing the polymeric dispersion at 5 C
for 24 h. Once the gelation of gelatin resulted in a partial consolidation of the patches, the second gelation was achieved by layering
2 mL of a BaCl2 1.5% (w/v) solution onto the forming patches. After
complete gelation (3 h), barium chloride was removed by pipette
aspiration and patches were maintained in BaCl2 0.5% (w/v).
2.3. Rheological studies
Rheological measurements were carried out by means of a controlled stress rheometer Stresstech HR (Reologica Instruments, AB
Milano, Italy) equipped with a parallel plate geometry (40 mm
diameter; 25 mm gap). Hydrogel patches were analyzed by oscillatory stress sweep measurements. As preliminary experiment, the
linear viscoelastic region of the patches, was determined, using a
frequency of 1 Hz and at stress value comprised between 1 and
500 Pa. The frequency sweep measurements were performed using
the frequency range 0.110 Hz and the stress value previously
determined in the linear viscoelastic region.
2.4. Design of Experiments and statistics

2. Materials and methods

2.1. Materials
Testosterone was from SigmaAldrich (St. Louis, MO, USA);
pectin LM Unipectine type OF 805 and gelatin type B OC 200 were
kindly provided from Sano Bioindustries (Angouleme, France) and
Degussa (Puteaux, France), respectively. All other materials and
solvents were from SigmaAldrich and were of highest purity available.
2.2. Preparation of polymeric dispersions and hydrogel patches
For the preparation of the hydrogel patches, pectin/gelatin
dispersion mixtures were prepared and puried as follows. Pectin
dispersions (26%, w/v) in ultrapure water were initially prepared
by adding a precisely weighted amount of pectin powder into ultrapure water as dispersing medium. The powder was sifted into the
vortex of the rapidly mechanically stirred water (300 rpm), after
the powder adding completion, the stirring speed was reduced at
50 rpm and left overnight in order to obtain the complete hydration

The screening of a selected number of parameters controlling

the production of hydrogel patches was performed by: (i) classical
intuitive approach Changing One Separate factor a Time (COST) and
(ii) Design of Experiments (DoE). Table 1 reports the investigated
experimental parameters and their range of variation, namely: total
polymer content, pectin to gelatin ratio, concentration of gelling
ions and type of cations. Taking into consideration the results
obtained by COST, the process optimization was performed by DoE.
A randomized central composite face-centered design (CCF), consisting of 17 runs, was used to study the inuence of experimental
parameters on elastic (G ) and viscous (G ) moduli (Table 2). The
experimental design and the evaluation of the experiments were
performed by the PC software MODDE 8.0 (Umetrics AB, Sweden),
followed by multiple linear regression (MLR) algorithms.
2.5. Drug loading
Testosterone was loaded by different procedures into the
hydrogel patches at the concentration of 10 mg/patch, corresponding to a concentration of 5 mg/mL. Method #1, nely grinded
(by mortar and pestle) testosterone powder was suspended

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

49

Table 1
Experimental parameters for the hydrogel patch production and list of the prepared samples.
Parameter

Abbreviation

Meaning

Range

Type of cations
Concentration of gelling ions
Total polymer content
Pectin to gelatin ratio

Cations
Ions
Polymer
p/g ratio

The type of cations used to induce the ionic gelation of pectin

The weight/volume percentage of barium chloride present in the ionic gelling solution
The total concentration of polymers employed for the preparation of the hydrogel patches
The weight ratio between pectin and gelatin used for the preparation of the hydrogel patches

Ca2+ , Ba2+ , Sr2+

1.02.0%, w/v
8.012.0%, w/v
35/6550/50%, w/w

#M-P2G2-Ca1.5
#M-P2G2-Ba1.5
#M-P2G2-Sr1.5
#M-P2G2-Ba1.0
#M-P2G2-Ba2.0
#L-P2G2-Ba1.5
#H-P2G2-Ba1.5
#M-P1G3-Ba1.5
#M-P3G1-Ba1.5
#M-P2G2-Ba1.5/T1
#M-P2G2-Ba1.5/T2
#M-P2G2-Ba1.5/T3
#H-P2G2-Ba1.5/T1
#H-P2G2-Ba1.5/T2
#H-P2G2-Ba1.5/T3

Ions type

Ions concentration (%, w/v)

Total polymer (%, w/v)

Pectin/gelatin ratio (%, w/w)

Testosterone loading method

Ca2+
Ba2+
Sr2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+
Ba2+

1.5
1.5
1.5
1.0
2.0
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5

10.0
10.0
10.0
10.0
10.0
8.0
12.0
10.0
10.0
10.0
10.0
10.0
12.0
12.0
12.0

42/58
42/58
42/58
42/58
42/58
42/58
42/58
35/65
50/50
42/58
42/58
42/58
42/58
42/58
42/58

Method #1
Method #2
Method #3
Method #1
Method #2
Method #3

Table 2
Experimental design matrix and results of the DoE (design of experiments) approach for the preparation of hydrogel patches.
Run

#L-P1G3-Ba1.0
#H-P1G3-Ba1.0
#L-P3G1-Ba1.0
#H-P3G1-Ba1.0
#L-P1G3-Ba2.0
#H-P1G3-Ba2.0
#L-P3G1-Ba2.0
#H-P3G1-Ba2.0
#L-P2G2-Ba1.5
#H-P2G2-Ba1.5
#M-P1G3-Ba1.5
#M-P3G1-Ba1.5
#M-P2G2-Ba1.0
#M-P2G2-Ba2.0
#M-P2G2-Ba1.5
#M-P2G2-Ba1.5
#M-P2G2-Ba1.5

Responsesa

Factors
Ions (%, w/v)

Polymer (%, w/v)

p/g ratio (%, w/w)

1.0
1.0
1.0
1.0
2.0
2.0
2.0
2.0
1.5
1.5
1.5
1.5
1.0
2.0
1.5
1.5
1.5

8.0
12.0
8.0
12.0
8.0
12.0
8.0
12.0
8.0
12.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0

35/65
35/65
50/50
50/50
35/65
35/65
50/50
50/50
42/58
42/58
35/65
50/50
42/58
42/58
42/58
42/58
42/58

G (Pa)
6500
12,600
8020
16,000
8510
15,030
10,800
17,950
7800
12,100
8400
12,900
9800
12,050
9310
9720
10,900

G (Pa)
1100
1720
1360
2030
1390
2100
1520
2700
1200
2200
1360
1980
1430
2750
1430
1490
1570

a
Considered response: elastic modulus G (the elastic or storage modulus represents a measure of the elasticity of materials. The ability of the material to store energy):
viscous modulus G (the viscous or loss modulus represents the ability of the materials to dissipate energy, lost as heat).

into the gelatin/pectin polymeric dispersion at the concentration of 5 mg/mL, maintained at 60 C, by magnetically stirring at
300 rpm. Method #2, testosterone was previously solubilized in
ethanol at 50 mg/mL, thereafter the ethanolic solution was added
to the gelatin/pectin polymeric dispersion maintained at 60 C,
at 100 L/mL. Method #3, testosterone powder was previously
dispersed in the BaCl2 solution used as gelling agent, at a concentration of 5 mg/mL; thereafter the suspension was layered
onto the forming patches, as previously described in Section 2.2.
The testosterone loaded patches were microscopically analyzed
by stereomicroscopy, using a Nikon SMZ 1500 Stereomicroscopy
(Nikon, Japan) equipped with a digital camera.

20 mm obtaining a release surface of 3.14 cm2 . The cell is positioned

in the bottom of the vessel with the cover facing upwards and at
a distance of 25.0 2.0 mm from the paddle blade. The test was
carried out by working at 100 rpm at 32.0 0.5 C in sink conditions and using as dissolution medium a water/dry ethanol solution
(80/20 w/w). At predetermined intervals, 2 mL of sample were
withdraw from the vessel and replaced by same amount of fresh
dissolution medium. Drug concentration from samples was determined by a UVvis spectrophotometer (8453 Agilent, Germany).
Drug calibration curve in the dissolution medium was previously
prepared (max = 249.0 nm, r = 0.9991), using a water/ethanol solution (80/20) as blank. The results are expressed as average of three
measurements and the error is reported as standard deviation (SD).

2.6. In vitro release studies

3. Results and discussion
The in vitro release of testosterone from the hydrogel was performed by dissolution test for transdermal patches employing a
paddle apparatus for solid oral dosage forms with the addition of
an extraction cell (Ph. Eur. VII Ed.). The central part of the cell forms
a cavity (depth of 2.6 mm, diameter 27 mm) to hold the patch. On
the top of the cell is positioned a cover with a central opening of

3.1. Preparation of hydrogel patches: general aspects and

analysis of the experimental parameters
The most common drug products applied to the skin for systemic effects are referred to as self-adhering transdermal delivery

50

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

Fig. 1. Schematic representation of the two step preparation procedure for hydrogel patches based on pectin and gelatin.

systems (TDS) or transdermal patches. Notably, the development

and commercialization of TDS for drug delivery require a number of
specic requisites, such as general product quality, mechanical and
rheological characteristics to t handling and skin wearing, drug
content uniformity (uniformity of dosage units) and effectiveness
at release and during shelf life.
In this respect, we started a set of experiments aimed at the
preparation of precisely controlled hydrogel patches, specically
designed as TDS. The procedure, developed in our laboratory,
is reported in Fig. 1. The patches are constituted by blends of
two natural hydrophilic polymers of different nature: pectin and
gelatin.
Pectins are a family of complex polysaccharides that contain 1,4linked -d galactosyluronic, in which some of the carboxyl groups
are methyl esteried, and in lower amounts other saccharidic units,
including homogalacturonan, rhamnogalacturonan-I and substituted galacturonans. In pectins, the degree of methylation (DM) is a
very important structural characteristic and it is dened as the percentage of carbonyl groups esteried with methanol. If more than
50% of the carboxyl groups are methylated, the pectins are called
high-methoxy pectins (HM) and less than that degree of methylation are called low methoxy (LM) pectins. Notably, LM pectin water
dispersions can be conveniently converted in gels by the addition
of divalent cations, such as calcium, barium and strontium. The formation of ionic bridges between cations and the ionized carboxyl

groups of the galacturonic acid results in gels characterized by the

so-called egg box-model structure (Liu et al., 2003).
Conversely, gelatin is a protein polymer, being obtained by the
denaturation and structural degradation of collagen. Its physicochemical and structural characteristics are similar to those of high
molecular weight rigid-chain synthetic polymers, which is not the
case with native collagen. Therefore, in a similar way to synthetic
polymers, in aqueous solutions, gelatin molecules assume, under
specic conditions (high temperature, solvent, pH) a exibility sufcient to realize a wide variety of conformations. These properties
confer to gelatin the ability to form a large variety of supermolecular geometries, from globular to brillar structures and the capacity
to form specic thermoreversible triple-stranded helical structures
(formed in solutions at low temperatures) (Rathna, 2008). The rate
of the formation of helical structures depends on many factors
including the gelatin molecular weight and the polymer concentration in the solution.
Taking advantage of the different gelling properties of pectin and
gelatin (i.e. ionic and thermal gelation mechanisms), transdermal
patches were designed and produced, with a diameter of 2.4 cm and
a thickness of 2.0 0.2 cm, according to the procedure reported in
the method section and schematized in Fig. 1.
Special attention was paid to the rheology of patches since it can
affect, in different ways, the drug release. For instance, rheological
properties may directly inuence the diffusion rate of a drug at

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

the microstructural level. Semisolid formulations with a relatively

high viscosity can exhibit high drug diffusion rates if compared
to formulation of lower viscosity. Moreover, the rheology of TDS
affects the application to the skin and the reproducibility maintenance of a product drug release. The latter is obviously a crucial
product characteristic that rstly researchers and later manufacturers should use to maintain and demonstrate batch-to batch
consistency. Finally it is noteworthy that the rheological properties
of semisolid formulations can also inuence their stability during
the storage and therefore inuence the product shelf life. These
observations support the importance of performing rheological
studies of TDS formulations also considering that the viscosity of
a semisolid dosage form is highly inuenced by factors such as the
inherent physical structure of the product and the specic analysis
and methodology.
A variety of methods can be used to evaluate the consistency
of semisolid dosage forms, including rheometry employing different geometries such as concentric cylinders and coneplate or
plateplate. Concentric cylinders are usually employed for uid,
owable semisolid dosage forms, while plateplate geometries are
typically used, as in the current paper, when the sample size is
small or the test samples are quite viscous and less owable, as for
hydrogels.
Keeping in mind all these considerations, the rst part of the
study was devoted to the investigation of the experimental conditions inuencing the rheological characteristics of the produced
patches. The analysis was initially performed by a ChangingOne-Separate-factor-a-Time (COST) approach, which represents an
intuitive method of designing experiments, involving the testing of
factors, or causes, one at a time instead of all simultaneously. We
specically employed this approach since we believe COST analysis can give practical information, later used for the successive
DoE. DoE is indeed more appropriately designed when there is a
good knowledge of the investigated system, and in this respect
the preliminary COST approach permits the selection of the factors
affecting the process and process variability (Mazzitelli et al., 2008).
Notably, the experimental approach for hydrogel patches preparation is relatively simple and governed by a limited number of
experimental parameters, including: (i) type of cations (cations),
(ii) concentration of gelling ions (ions), (iii) total polymer content
(polymer) and (iv) pectin to gelatin weight ratio (p/g ratio) (see
Table 1).
3.2. Effects of type and concentration of cations on the
rheological properties of hydrogel patches
As rst set of experiment, we investigated the effects of a key
parameter known to affect the mechanical properties of ionically

51

cross-linked gels, namely the type of cation triggering the polysaccharide gelation. The effect of three different divalent cations was
analyzed on the rheological properties, by varying the constituent
of the gelling solution using calcium, barium or strontium ions. The
gelling solutions were prepared by dissolving CaCl2 , BaCl2 or SrCl2
in water at 1.5% (w/v).
The effect of type of ions was analyzed on samples prepared
maintaining constant both the total polymer content and p/g ratio.
After patch preparation, their rheological characteristics were analyzed by frequency stress sweep measurements to determine the
frequency dependence of both elastic (G ) and viscous (G ) moduli.
Viscoelastic materials exhibit indeed both solid- and liquid-like
behavior that are usually analyzed by the evaluation of G , that
describes the solid-like character of a material and G describing
the liquid-like character.
The data reported in Fig. 2 clear indicate that the formation of
the pectin/gelatin co-gel is very sensitive to the type of cations. As
already reported for alginate based gels, divalent cations bind to the
pectin in a highly cooperative manner, thereby forming a gel with
various afnity for different ions (Morch et al., 2006) This behavior
was conrmed by the values of both G and G found also in the
case of p/g blends. The moduli were indeed consistently higher in
the case of gels cross-linked with barium, followed by calcium and
strontium.
From the obtained data, it was therefore possible to withdraw
a number of general considerations (later conrmed by further
experiments): (i) the hydrogel patches showed good homogeneity
as attested by oscillation stress sweep measurements that gave very
reproducible rheological proles, (ii) the tested patches behaved as
solid-like materials, with G always greater than G , (iii) the samples showed a at rheological trend in small amplitude oscillatory
experiments (110 Hz) with G roughly one order of magnitude
greater than G .
In consideration of the above reported statements, the following
experiments were performed using barium ions (Ba2+ ) as gelling
agents for the pectin gelation. This choice was not only due to
the superior mechanical properties of the barium cross-linked gels,
but also taking into consideration that barium hydrogels were successfully employed for in vitro and in vivo experiments. The use
of barium as alternative to the generally used calcium has indeed
given promising results both in allo-and xenograft transplantation
protocols as well as in biocompatibility studies (Luca et al., 2007).
As previously described for pure alginate hydrogels, the here
presented results demonstrated that also in the case of the hydrogel
blends, the concentration of barium ions inuences the gelation
behavior of the polymer mixture.
The inuence of Ba2+ ion concentration on the mechanical
parameters of gel structures, at constant total polymer content and

Fig. 2. Frequency sweep tests showing the frequency dependence of the elastic modulus G (A) and viscous modulus G (B). Determinations were performed on patches
constituted of hydrogel blends of pectin and gelatin, with a total polymer concentration of 10% (w/v) and a p/g ratio equal to 42/58. The pectin fraction was gelled by CaCl2
(squares), BaCl2 (circles) or SrCl2 (triangles). Data represent the average of three independent measurements run in triplicate.

52

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

Fig. 3. Frequency sweep tests showing the frequency dependence of the elastic modulus G (A) and viscous modulus G (B). Determinations were performed on patches
constituted of hydrogel blends of pectin and gelatin, with a total polymer concentration of 10% (w/v) and a p/g ratio equal to 42/58. The pectin fraction was gelled by BaCl2
aqueous solution at the concentration of 1.0 (triangles), 1.5 (circles) or 2.0% (w/v) (squares). Data represent the average of three independent measurements run in triplicate.

p/g ratio, is shown in Fig. 3. The amount of BaCl2 , present in the

gelling solution, was progressively increased from 1.0 to 2.0% (w/v).
The choice of the concentration range was performed following the
indications obtained in previously investigations on alginate gels,
indicating that 1.5% was the optimal value for the alginate gelation
and biocompatibility (Mazzitelli et al., 2011).
The raising of the barium concentration, from 1.0 to 2.0%, caused
a progressive increase of G modulus that passed from an average value of about 9000 to 12,000 Pa; in agreement, G displayed a
similar trend passing from about 1400 to 2000 Pa.

3.3. Effects of concentration and ratio of polymers on the

rheological properties of hydrogel patches
Both total polymer concentration and pectin to gelatin ratio
were modied to investigate their effect on the rheological behavior of the produced patches. Firstly, the ratio between the polymer
and the concentration of Ba2+ were kept constant and the total
polymer was varied from 8% to 12% (Fig. 4).
The obtained results indicated that the total polymer concentration is one the key parameter inuencing the hydrogel mechanical

Fig. 4. Frequency sweep tests showing the frequency dependence of the elastic modulus G (A) and viscous modulus G (B). Determinations were performed on patches
constituted of hydrogel blends of pectin and gelatin, with a total polymer concentration of 8.0 (triangles), 10 (circles) or 12.0% (w/v) (squares) and a p/g ratio equal to 42/58.
The pectin fraction was gelled by BaCl2 aqueous solution at the concentration of 1.5% (w/v). Data represent the average of three independent measurements run in triplicate.

Fig. 5. Frequency sweep tests showing the frequency dependence of the elastic modulus G (A) and viscous modulus G (B). Determinations were performed on patches
constituted of hydrogel blends of pectin and gelatin, with a total polymer concentration of 10.0% (w/v) and a p/g ratio equal to 35/65 (triangles), 42/58 (circles) or 50/50
(squares). The pectin fraction was gelled by BaCl2 aqueous solution at the concentration of 1.5% (w/v). Data represent the average of three independent measurements run
in triplicate.

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

53

Fig. 6. Response surface plots obtained by DoE analysis of the preparation of patches constituted of hydrogel blends of pectin and gelatin. The following factors and responses
were respectively investigated: polymer and ions vs. G (A) and G (B); p/g ratio and ions vs. G (C) and G (D) and p/g ratio and polymer vs. G (E) and G (F). The meanings of
factors and responses are reported in Tables 1 and 2.

characteristics. In particular, G appears to be strongly affected

by the total polymer concentration, with a marked difference
between the values obtained at 8% (7000 Pa) to those at 12%
(12,000 Pa).
As further analysis, Fig. 5 reports the effect of the polymer ratio.
Interestingly, pectin appears to have a predominant effect on the
nal rheological characteristic of the hydrogel patches. When G
and G were measured at constant total polymer concentration
(10%, w/v), it was evident that patches they displayed the highest value in the case the larger proportion of pectin (50/50%, w/w),
whilst both moduli progressively decreased with the concomitant
decrease of pectin content.

3.4. Combined effects of experimental parameters on the

rheological properties of hydrogel patches: DoE analysis
As above discussed, the hydrogel composition and gelling conditions (i.e. factors: ions, polymer and p/g ratio) were found to
be critical experimental parameters strongly affecting the patches
rheological characteristics. In this respect, these factors were further investigated in a second set of experiments, following a
statistical approach able to evaluate their possible interactions.
Therefore, a factorial design was performed, consisting of 17 runs,
where ions, polymer and p/g ratio were the independent factors,
while G and G represented the dependent responses.

54

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

Fig. 7. Bar plot showing the results for the t and the predictivity of the model for
the three-level factorial design describing the combined effect of ions, polymer and
p/g ratio on elastic and viscous moduli. R2 and Q2 respectively represents the percent
of the variation of the response explained by the model and the fraction of variation
of the response predicted by the model. When the model validity bar is larger than
0.25, there is no lack of t of the model (the model error is in the same range as
the pure error). Finally, the reproducibility represents the variation of the response
under the same conditions (i.e. pure error), of the three center points, compared to
the total variation of the response.

Table 2 reports the matrix for the three-level factorial design

describing the combined effects of the above reported factors, while
Fig. 6 shows the response surface plots obtained by the DoE analysis.
The plots reported in panels A and B showed the direct relation
between both polymer and ions and G and G . In detail, the G
modulus increased from below 7000 to over 15,000 Pa, in response
to a polymer increase from 8% to 12% (w/v) (panel A); a similar
trend, even if in a lower extent, was observed for the viscous modulus G (panel B). The variation of the amount of barium (ions) had
a similar trend, but less pronounced; as expected, the increase of
this factor provoked a progressive increase of the G and G moduli;
this effect is particularly evident when the factor ions is set at the
higher value (2.0%, w/v) (panels C and D). Finally, the plots showing
the effect of the factor p/g ratio showed that both G and G had an
inverse relation with the amount of gelatin present in the polymer
mixture. At high level of gelatin (i.e. p/g ratio 34/66) elastic and
viscous moduli had indeed the lowest values, below 7000 and
1500 Pa, respectively. In addition, Fig. 7 reports the summary of
the t and the predictivity of the model for the three-level factorial
design, describing the combined effect of ions, polymer and p/g
ratio on G and G . In particular, the high values of all coefcients
indicate that the model has a good validity and a signicant
reproducibility.
3.5. Drug loading and its effect on hydrogel patch rheological
characteristics
Testosterone, used as model drug, was loaded into the hydrogel patches following different procedures that were alternatively
employed in order to asses which of them could outcome in a
homogeneous dispersion of the drug into the patch, resulting in
convenient and reliable release proles. Notably, testosterone is a
very hydrophobic compound; its solubility in the hydrogel matrix
is extremely low and therefore the drug is present in the patches
as suspended particles.
Drug loading was performed by three different procedures, as
described in the experimental section, namely: methods #1, #2 and
#3. In the rst case, the drug powder was previously grinded by

Fig. 8. Stereophotomicrographs of patches constituted of hydrogel blends of pectin

and gelatin after testosterone loading. The microscopic analysis was performed
on patches constituted of blends of pectin and gelatin, with a total polymer concentration of 10% (w/v) and a p/g ratio equal to 42/58. The pectin fraction was
gelled by BaCl2 aqueous solution at the concentration of 1.5% (w/v). Testosterone
was loaded following three different procedures as described in the experimental
section, namely: methods #1 (A), #2 (B) and #3 (C). Bar corresponds to 100 m.

mortar and pestle and thereafter suspended into the pectin/gelatin

polymeric dispersion at 60 C, under magnetic stirring. The second method relies on a preliminary solubilization of testosterone
in ethanol, followed by addition of the obtained solution to the
p/g dispersion at 60 C, causing the ethanol evaporation and the
consequent precipitation of testosterone as very ne particles.
Finally, the third procedure was based on the addition of testosterone as suspension in the BaCl2 solution, later used as gelling
agent.

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

55

Fig. 9. Frequency sweep tests showing the frequency dependence of the elastic modulus G (A) and viscous modulus G (B). Determinations were performed on patches
constituted of hydrogel blends of pectin and gelatin, with a total polymer concentration of 10% (w/v) and a p/g ratio equal to 42/58. The pectin fraction was gelled by BaCl2
aqueous solution at the concentration of 1.5% (w/v) in the presence (lled symbols) or in the absence (open circles) of testosterone. The drug was loaded following three
different procedures as describe in the experimental section, namely: methods #1 (lled circles), #2 (lled squares) and #3 (lled triangles). Data represent the average of
three independent measurements run in triplicate.

The comparative analysis of the loading methods was performed

by stereomicroscopic analysis, in order to determine the degree of
drug particle dispersion within the patches. The microphotographs,
reported in Fig. 8, indicate that patches prepared by method #2 displayed superior characteristics, as conrmed by the homogeneous
distribution of the powder particles throughout the entire eld;
more importantly the method resulted in the lower presence of
hard agglomerates, while the primary powder particles, that are
typically <10 m across, were predominant. On the contrary, samples prepared by methods #1 and #3 are characterized by a large
presence of irregularly scattered hard agglomerates, manly located
at the patch surface.
The homogenous distribution of testosterone obtained by
method #2 was also conrmed by the rheological analyses reported
in Fig. 9, showing the frequency sweep curves for empty and loaded
patches. As expected, all loaded patches displayed highest values
for G and G ; particularly, the highest level for both moduli was
reached by the patches loaded with method #2. This behavior was
attributed to the fact that the rheological properties of suspensions
are strongly affected by the nature and content of dispersed powder
particles, such as particle shape and size (Lenglov et al., 2003).
Noteworthy, different papers reported that generally there is an
inverse relation between the size of the dispersed particles and
viscosity of the dispersing phase (Patankar and Hu, 2001; Bao et al.,
2002). As shown in the microphotographs reported in Fig. 8, the
testosterone particles, present in the patches loaded with method
#2, are mainly single primary particles, therefore displaying lower
mean dimensions with respect to those present in patches loaded
with methods #1 and #3, accounting for the highest values of G
and G .
In addition, to evaluate the drug loading effect on the characteristics of the patches and the reproducibility of the process, a further
rheological parameter was considered, namely tan It represents
the tangent of the phase angle the ratio of viscous modulus to
elastic modulus and it can be usefully employed as quantier of
the presence and extent of elasticity in a uid.
In this respect, empty patches (constituted of blends of p/g, with
a total polymer concentration of 10% (w/v) and a p/g ratio equal to
42/58) displayed a tan value equal to 0.162, therefore reecting
the prevalence of elastic modulus toward viscous modulus.
The loaded patches presented very similar tan values equal to
0.172, 0.210 and 0.155 for methods #1, #2 and #3, respectively.
The slight increase in tan value for the loaded patches indicates
that the presence of drug powder resulted in a minimal change
in the balance between energy loss and storage. The rheological

characterization was run in triplicate on three independent samples and the analysis of the standard deviation revealed that the
samples prepared by the loading method #2 gave more reproducible results in comparison to the others (smaller SD), conrming
that this procedure is the most appropriate for testosterone loading. Taken together, the study aimed to evaluate the inuence of
the experimental parameters on the patch characteristics allowed
the selection of a limited number of samples for the following
in vitro release studies. Namely, samples #M-P2G2-Ba1.5/T2 and
#H-P2G2-Ba1.5/T2 were considered in reason of their mechanical, rheological and reproducible properties (see Table 1 for the
complete composition of the selected samples). Finally, the patches
loaded with methods #1, #2 and #3 were also compared in term of
in vitro release studies.
3.6. In vitro release studies
Transdermal patches are usually formulated to assure a sustained systemic drug release, from a few days up to a couple of
weeks. In this respect, the accurate rheological characterization
performed on patches with different formulations had the aim to
select those presenting the best viscoelastic properties, allowing an
easy administration (i.e. skin application) and duration of use.
As further objective aimed to evaluate the drug product performance, specic tests for determining the drug release from the
produced patches were performed. Determination of drug release
proles from transdermal medicines, although does not representing a measure of bioavailability, gives important information on
the drug release characteristics that have the potential to alter the
biological performance of the drug in the dosage form.
As general consideration for the setup of drug release experiments, it is mandatory to remember that an in vitro test must
be able to evaluate the rate and extent of release of drugs from a
transdermal patch. Although the test may not model in vivo performance, it is a critical quality attribute to be specied in the release
and shelf life nished product specication. Data coming from such
assays can help also in the evaluation of the effect that formulation
composition and shape may inuence drug permeation through the
skin (EMA, 2012).
Notably, transdermal patches are complex dosage forms, where
the rate and extent of drug release is inuenced by various parameters, therefore the chosen assay inuences strongly the quality and
reproducibility of the in vitro release test (Ueda et al., 2009).
In this respect, several apparatus and procedures have
been utilized and proposed to study in vitro release from

56

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

transdermal patches including Franz diffusion apparatus, rotating

cylinder method, ow through diffusion cell. The latest resulted
the most appropriate, being the ofcial method described in Ph.
Eur. monograph (VII Ed.) and strongly encouraged from FDA (Ueda
et al., 2009). The ow through diffusion represents a modication
of the paddle method used for oral solid dosage form dissolution
studies. This method does not inuence the release mechanism;
this means that the obtained results can be ascribed only to the
formulation composition and properties.
The apparatus consists of an iron steel cell, in which the patch
is placed, projected whit the aim to expose only one surface to the
dissolution medium. By this method it is possible to test only a
dened sample area of patch which represents the side that, during
the in vivo application, takes contact with the stratum corneum
where the drug is released and then absorbed.
In this respect, we employed the method described in Ph Eur
monograph for transdermal patches, following the dissolution
test and, as specied in the Guideline on quality of transdermal
patches, we assessed that the test itself and/or sample preparation
did not damage or otherwise alter the performance of the hydrogel
patches.
With the aim to evaluate the effect of the formulation on the
in vitro release prole of testosterone from hydrogel blends, different parameters were considered, namely: (a) the procedure
adopted for testosterone loading into the patches; (b) the length of
patch storage in BaCl2 0.5% (w/v) solution and (c) the total polymer
concentration, employed for hydrogel patch preparation (intended
as the sum of the two polymers).
The suitability of the most appropriate drug loading method,
able to guarantee a sustained testosterone release, was initially
investigated. Testosterone release determinations were performed
on the pectin/gelatin patch #H-P2G2-Ba1.5, in which the drug
was loaded following the described methods (see Section 3.5).
The obtained release proles, reported in Fig. 10A, conrmed
the already discussed superior quality of samples loaded with
the procedure #2. It is indeed evident that loading methods
#1 and #3, resulting in an uneven distribution of testosterone,
mainly located on the patch surface, caused a faster release of the
drug.
With respect to point (b), the hydrogel patches, after preparation, were maintained in BaCl2 0.5% (w/v) solutions for different
length of time (i.e. 18 or 42 h). The drug release proles reported in
Fig. 10B showing the possible effect of storage, display that for patch
#H-P2G2-Ba1.5/T2 not big differences could be observed from the
dissolution proles determined after 18 and 42 h of incubation in
BaCl2 solution.
Finally, regarding the effect of the total polymer content on
the patch release characteristics, as displayed in Fig. 10C, the
drug is released in a slightly faster fashion from hydrogel patches
containing a lower amount of polymers (#M-P2G2-Ba1.5/T2 ) in
comparison to the drug release kinetics of patches produced with
12% total polymer concentration (#H-P2G2-Ba1.5/T2 ) (Fig. 10C). For
instance, after 6 h 32.5% and 28.0% of testosterone are released from
#M-P2G2-Ba1.5/T2 and #H-P2G2-Ba1.5/T2 ), respectively.
The observed difference was attributed to the characteristics
of the matrix network. It could be hypothesized that in the case of
#H-P2G2-Ba1.5/T2 the higher total polymer concentration causes a
slower release of the drug in reason of the tighter hydrogel network
structure.
Taken together, the obtained release proles are in perfect
agreement with the aim to obtain a system in which the drug
release is completely controlled by the patch matrix. The polymeric
matrix creates a network where both water and drug molecules can
be entrapped inuencing the structure, viscosity, and drug release
mechanism. In vitro release data were further analyzed to possibly determine the mechanism governing drug release. As reported

Fig. 10. Testosterone release proles from hydrogel patches based on pectin and
gelatin. Panel A: patches loaded following procedure #1 (lled squares), #2 (lled
circles) and #3 (lled diamonds) (as described in the experimental section). Patches
were maintained (after preparation) in BaCl2 0.5% (w/v) solution for 42 h. Samples
were prepared with a total polymer concentration of 12% (w/v), a p/g ratio equal
to 42/58 and gelled by BaCl2 aqueous solution at the concentration of 1.5%. Panel
B: patches maintained (after preparation) in BaCl2 0.5% (w/v) solution for 18 (open
circles) or 42 h (lled circles); samples were prepared with a total polymer concentration of 12% (w/v), a p/g ratio equal to 42/58 and gelled by BaCl2 aqueous solution
at the concentration of 1.5%. Panel C: patches prepared with a total polymer concentration of 10% (gray circles) or 12% (w/v) (lled circles), a p/g ratio equal to 42/58 and
gelled by BaCl2 aqueous solution at the concentration of 1.5%; samples were maintained (after preparation) in BaCl2 0.5% (w/v) solution for 42 h. Testosterone was
loaded following method #2, determinations were performed by paddle method
for solid oral dosage forms and an extraction cell. Data represent the average of
three independent experiments (bars = SD).

in Table 3, the release proles were tted to the rst order kinetics Mt /M = 1 ekt , and Ritger and Peppass kinetics mathematical
model Mt /M = 1 ekt , where Mt /M is the drug fraction released
at time t, k is a constant, depending on structural and geometric
characteristics of the system, n is the diffusional coefcient related
to the release mechanism (Ritger and Peppas, 1987).
The analysis of r values found for sample #H-P2G2-Ba1.5/T2 suggests that drug release from the matrix follows a ckian diffusional

S. Mazzitelli et al. / International Journal of Pharmaceutics 454 (2013) 4757

57

Table 3
Ritger and Peppass kinetics mathematical model and rst order kinetic model tting.
Equation

18 h

42 h

#M-P2G2-Ba1.5/T2

Mt /M = kt1 (zero order)

Mt /M = kt0.9
Mt /M = kt0.8
Mt /M = kt0.7
Mt /M = kt0.6
Mt /M = kt0.5
Mt /M = 1 ekt (rst order)

#H-P2G2-Ba1.5/T2

#M-P2G2-Ba1.5/T2

#H-P2G2-Ba1.5/T2

0.067
0.126
0.240
0.458
0.886
1.744
0.001

0.294
1.370
2.688
4.352
6.535
9.551
0.005

0.998
0.998
0.997
0.994
0.989
0.982
0.998

0.068
0.130
0.249
0.479
0.934
1.857
0.001

0.107
3.901
2.421
0.550
1.904
5.295
0.019

0.973
0.981
0.987
0.992
0.996
0.998
0.985

0.079
0.150
0.285
0.549
2.882
2.119
0.001

4.099
2.734
1.059
1.056
2.244
7.663
0.013

0.980
0.986
0.991
0.995
0.996
0.998
0.991

0.063
0.120
0.230
0.443
0.863
1.758
0.001

5.407
4.286
2.909
1.168
1.116
4.271
0.021

0.969
0.977
0.984
0.990
0.995
0.998
0.981

mechanism as indicated by the slightly higher values of r calculated

for the tting with n = 0.5. On the contrary, for sample #M-P2G2Ba1.5/T2 the higher r values were for n = 0.9 and n = 1, meaning that
drug release in governed from polymeric chains mechanism and
from drug concentration. Analysing the composition, this sample
shows the lowest total polymeric content that probably promotes
change in chains disposition when kept in solution for long time.
4. Conclusion
The design and development of transdermal patches must take
into account the rheological characteristics that can inuence both
patient compliance and therapy efcacy. Transdermal patches are
indeed exposed to many mechanical stresses during the manufacturing process, packaging and application (e.g. force exerted to
promote its adhesion on skin surface and continuous mechanical stimuli connected to patient daily life style). For this reason
it is very important to know their rheological characteristics and
microstructure stability. The structural properties of the patches
prepared were assessed performing oscillation stress sweep measurements which provide information concerning the viscolelastic
properties of the polymer blends employed.
References
Bao, Y., Senos, A.M., Almeida, M., Gauckler, L.J., 2002. Rheological behaviour of aqueous suspendions of hydroxyapatite (HAP). J. Mater. Sci. 13, 639643.
Berti, J.J., Lipsky, J.J., 1995. Transcutaneous drug delivery: a practical review. Mayo
Clin. Proc. 70, 581586.
Brown, M.B., Martin, G.P., Jones, S.A., Akomeah, F.K., 2006. Dermal and transdermal
drug delivery systems: current and future prospects. Drug Deliv. 13, 175187.
EMA, 2012. Guideline on quality of transdermal patches.
Geurts, T.B.P., Coelingh Bennink, H.J.T., 2000. Testosterone replacement therapy:
testosterone undecanoate (Andriol). J. Urol. Uroge., 2435, Special Edition.
Hoffman, A.S., 2012. Hydrogels for biomedical applications. Adv. Drug Deliv. Rev. 54,
312.
Houwing, N.S., Maris, F., Schnabel, P.G., Bagchus, W.M., 2003. Pharmacokinetic study
in women of three different doses of a new formulation of oral testosterone
undecanoate. Andriol Testocaps. Pharmacother. 23, 12571265.

Kandavilli, S., Nair, V., Panchagnula, R., 2002. Polymers in transdermal drug delivery
systems. Pharm. Technol., 6280.
Lee, S.K., Knighton, S., 2008. Scenario 4: choice of hormone replacement therapy
(HRT). In: North-Lewis, P. (Ed.), Drugs and the Liver. Pharmaceutical Press, London, pp. 257274.
Lenglov, A., Pavlnek, V., Sha, P., Quadrat, O., Stejskal, J., 2003. The effect of
dispersedparticlesize and shape on the electrorheological behaviour of suspensions. Colloids Surf. A 227, 18.
Liu, L., Fishman, M.L., Kost, J., Hicks, K.B., 2003. Pectin-based systems for colonspecic drugdelivery via oral route. Biomaterials 24, 33333343.
Luca, G., Calvitti, M., Nastruzzi, C., Bilancetti, L., Becchetti, E., Angeletti, G., Mancuso,
F., Calaore, R., 2007. Encapsulation, in vitro characterization and in vivo biocompatibility of Sertolis cells in alginate based microcapsules. Tissue Eng. 13,
641648.
Mazzitelli, S., Capretto, L., Carugo, D., Zhang, X., Piva, R., Nastruzzi, C.,
2011. Optimised production of multifunctional microbres by microuidic chip technology for tissue engineering applications. Lab Chip 21,
17761785.
Mazzitelli, S., Tosi, A., Balestra, C., Nastruzzi, C., Luca, G., Mancuso, F., Calaore,
R., Calvitti, M., 2008. Production and characterization of alginate microcapsules produced by a vibrational encapsulation device. J. Biomater. Appl. 23,
123145.
Morch, Y.A., Donati, I., Strand, L.B., Skjak-Braek, G., 2006. Effect of Ca++ , Ba++ , and Sr++
on alginate microbeads. Biomacromolecules 7, 14711480.
Patankar, N.A., Hu, H.H., 2001. Rheology of a suspension of particles in viscoelastic
uids. J. Non-Newtonian Fluid Mech. 96, 427443.
Peppas, N.A., Bures, P., Leobandung, W., Ichikawa, H., 2000. Hydrogels in pharmaceutical formulations. Eur. J. Pharm. Biopharm. 50, 2746.
Prausnitz, M.R., Langer, R., 2008. Transdermal drug delivery. Nat. Biotechnol. 26,
12611268.
Rathna, G.V.N., 2008. Gelatin hydrogels: enhanced biocompatibility, drug release
and cell viability. J. Mater. Sci. Mater. Med. 19, 23512358.
Ritger, P., Peppas, N.A., 1987. A simple equation for description of solute release,
II: Fickian and anomalous release from swellable devices. J. Control. Release 5,
3742.
Subedi, R.K., Oh, S.Y., Chun, M.K., Choi, H.K., 2010. Recent advances in transdermal
drug delivery. Arch. Pharm. Res. 33, 339351.
Ueda, C.T., Shah, V.P., Derdzinski, K., Ewing, G., Flynn, G., Maibach, H., Marques, M.,
Rytting, H., Shaw, S., Thakker, K., Yacobi, A., 2009. Topical and transdermal drug
products. Pharmacopeial Forum 35, 750764.
Wokovich, A.M., Prodduturi, S., Doub, W.H., Hussain, A.S., Buhse, L.F., 2006. Transdermal drug delivery system (TDDS) adhesion as a critical safety, efcacy and
quality attribute. Eur. J. Pharm. Biopharm. 64, 18.
Yoo, J.W., Lee, C.H., 2006. Drug delivery systems for hormone therapy. J. Control.
Release 112, 114.