Journal of Colloid and Interface Science 318 (2008) 421–429

www.elsevier.com/locate/jcis

Phosphatidylcholine embedded micellar systems:

Enhanced permeability through rat skin ✩
Aviram Spernath a , Abraham Aserin a , Amnon C. Sintov b,∗ , Nissim Garti a,∗
a Casali Institute of Applied Chemistry, The Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904, Israel
b Department of Pharmacology and School of Pharmacy, Ben-Gurion University of the Negev, P.O. Box 653, Beer Sheva 84105, Israel

Received 6 September 2007; accepted 18 October 2007

Available online 19 November 2007

Abstract
Micellar and microemulsion systems are excellent potential vehicles for delivery of drugs because of their high solubilization capacity and im-
proved transmembrane bioavailability. Mixtures of propylene glycol (PG) and nonionic surfactants with sodium diclofenac (DFC) were prepared
in the presence of phosphatidylcholine (PC) as transmembrane transport enhancers. Fully dilutable systems with maximum DFC solubilization
capacity (SC) at pH 7 are presented. It was demonstrated that the concentrates underwent phase transitions from reverse micelles to swollen
reverse micelles and, via the bicontinuous transitional mesophase, into inverted O/W microstructures. The SC decreases as a function of dilution.
DFC transdermal penetration using rat skin in vitro correlated with SC, water content, effect of phospholipid content, presence of an oil phase,
and ethanol. Skin penetration from the inverted bicontinuous mesophase and the skin penetration from the O/W-like microstructure were higher
than that measured from the W/O-like droplets, especially when the micellar system containing the nonionic surfactant, sugar ester L-1695, and
hexaglycerol laurate. PC embedded within the micelle interface significantly increased the penetration flux across the skin compared to micellar
systems without the embedded PC at their interface. Moreover, the combination of PC with HECO40 improved the permeation rate (P ) and
shortened the lag-time (TL ).
© 2007 Elsevier Inc. All rights reserved.

Keywords: Microemulsions; Phosphatidylcholine; Diclofenac; Permeability; Transdermal drug delivery

1. Introduction stances into and through the skin compared to conventional

In recent years, various colloidal systems have been investi-
gated as suitable vehicles for delivery of drugs and nutraceuti-
cals into and through the skin. Microemulsion systems, owing
to their thermodynamic stability, ease of preparation, trans-
parency, low viscosity, and considerable potential for solubi-
lization of a variety of active materials, have often been the
subject of various investigations related to drug delivery [1].
The results of numerous studies [2–7] have suggested that mi-
croemulsion vehicles have a significant potential to increase
penetration of hydrophilic, lipophilic, and amphiphilic sub-
✩
This paper is part of Aviram Spernath’s doctoral dissertation, The Hebrew
University of Jerusalem.
* Corresponding authors. Faxes: +972 (8) 647 2960 (A.C. Sintov), +972 (2)
652 0262 (N. Garti).
E-mail addresses: asintov@bgu.ac.il (A.C. Sintov), garti@vms.huji.ac.il
(N. Garti).
0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2007.10.036
vehicles. The release behavior depends on several properties,
such as physicochemical properties of the components, the
microemulsion’s internal structure (water-in-oil (W/O), oil-in-
water (O/W), or bicontinuous), and the interactions between
drug and vehicle [1,8].
SMEPS (self-microemulsifying pharmaceutical systems) are
isotropic mixtures of oil, surfactant, cosurfactant (or solubi-
lizer), and a drug. The basic principle of this system is its
ability to form fine oil-in-water (O/W) microemulsions under
gentle agitation following dilution to specific aqueous phase
content [9]. The O/W microemulsion region in these systems
is usually quite limited. In our recent publications [10–12], we
have introduced new SMEPSs capable of forming W/O, bicon-
tinuous, and O/W microemulsions progressively, upon water
dilution, without phase separation.
The ability of a drug in a topical formulation to permeate
the skin and to exert its effect is dependent on two consec-
422 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429

utive events. In the commonly used topical formulations, the

drug solute must first diffuse out of the vehicle to the skin sur-
face and then it must penetrate through the external cornifying
layer (stratum corneum) of the skin; it should then diffuse to
its target (local tissue or systemic circulation). Both steps are
dependent upon the physicochemical properties of the drug, ve-
hicle/carrier, and skin barrier [13].
Diclofenac sodium salt (diclofenac, DFC) is a nonsteroidal
anti-inflammatory drug that is widely used because of its robust
analgesic, antipyretic, and anti-inflammatory effects [14]. DFC
is not easily absorbed transdermally [15] and its delivery may
be improved by incorporation into microemulsion systems. Mi-
croemulsions are one of the colloidal systems used in topical
and transdermal delivery of particularly lipophilic compounds
[6,7,16]. It has also been demonstrated that skin permeation of
hydrophilic compounds in microemulsions may be significantly
enhanced compared to conventional vehicles, such as hydro- Fig. 1. Pseudo-ternary phase diagram (25 ◦ C) of a water/MCT/ethanol/
gels, emulsions, and liposomes [6,7,16]. Microemulsions also PC/HECO40/PG system with a constant weight ratio of PC/HECO40/PG
have excellent solubilization capacity [16], so these systems are (1:3:10) and a constant weight ratio of MCT/ethanol (1:3). Line N100 contains
an alternative for the solubilization of sparingly soluble com- only PC/HECO40/PG (1:3:10) with different water concentrations, while line
N82 contains surfactant mixture and oil phase in a 4:1 wt ratio with different
pounds such as diclofenac [16,17]. Microemulsions may also water concentrations.
act as penetration enhancers that modify the diffusion barrier of
the skin, depending on their oil/surfactant constituents [4,16]. tives (Hamburg, Germany). Polyoxyethylene-40 hydrogenated
It should be noted that in many of the reported studies the castor oil (Simulsol 1293, HECO40) was obtained from Sep-
drugs were solubilized in microemulsions of W/O or O/W, and pic (Paris, France). Hexaglycerol monolaurate, SY-Glyster
very little attention was paid to evaluate what happens to the
ML-500, was obtained from Sakamoto Yakuhin Kogyo Co.,
drug upon high dilution with water in our gastro-intestinal tract
Ltd. (Osaka, Japan). Sucrose monolaurate, L-1695, was ob-
(GIT) prior to approaching the membrane. Full dilution and no
tained from Ryoto Sugar Ester, Mitsubishi-Kagaku Foods Cor-
drug detachment are essential in order to confirm that the drug
poration (Tokyo, Japan). Brij 96v (polyoxyethylene (10) oleyl
will not leave the vehicle in the stomach or the GIT or precipi-
alcohol, C18:1 E10 ) was purchased from ICI Specialty Chemicals
tate or decompose.
(Essen, Germany). 1,2-Propanediol (PG) was purchased from
Penetration enhancers can also be used to enhance the per-
Merck KGaA (Darmstadt, Germany). Ethanol (EtOH) was pur-
meation of a drug from topical formulations [18,19]. Mild pene-
chased from Frutarom (Haifa, Israel) and diclofenac sodium
tration enhancers such as propylene glycol (PG) which are often
salt was purchased from Sigma Chemical Co. (St. Louis, MO,
used [20] are presumed to act by solvation of the α-keratin
USA). Medium chain triglyceride, MCT (Neobee M-5), was
structures in cells and facilitate transcellular diffusion [21].
Studies have shown that phosphatidylcholine (PC), if embed- obtained from Stepan (Northfield, IL, USA). Neobee M-5 con-
ded within the emulsion, liposome, or droplet interface, can tains 66% caprylic and 32% capric fatty acids. All chemicals
enhance transmembrane passage across the digestive tract and were used without further purification. The water was double
the skin, thus improving the bioavailability of the guest mole- distilled.
cules [22–24]. The aims of this study are: (1) to find the solu-
bilization profile of DFC along the water dilution line in fully 2.2. Microemulsion preparation
dilutable systems, (2) to evaluate the influence of phosphatidyl-
choline molecules on the permeation of DFC through rat skin,
Stock solutions of oil phase composed of MCT and ethanol
and (3) to evaluate the influence of microstructure changes on
in 1:3 weight ratios, and surfactants mixture composed of PC,
the DFC permeability.
hydrogenated PEG40 castor oil (HECO40), and propylene gly-
col (PG) in 1:3:10 weight ratios were prepared. The oil phase
2. Materials and methods and the surfactant mixture were mixed together in predeter-
mined concentrations to form a clear (transparent) concentrate.
2.1. Materials The concentrates can be further diluted with water, with no
phase separation, to form crystal clear, thermodynamically sta-
Epikuron 200 containing 92% phosphatidylcholine (PC ble microemulsions. The pseudo-ternary phase diagram of wa-
<92%, lysoPC 3%, other phospholipids 2%; fatty acid com- ter diluted concentrate (without drug) is given in Fig. 1. The
position: 13–17% saturated, 6–8% mono-unsaturated, 75–81% same systems were involved in the solubilization process of
poly-unsaturated fatty acid of which 68–72% is linoleic acid DFC in a clear oil-based concentrate prior to the water dilu-
and 7–9% is linolenic acid) was obtained from Degussa BioAc- tion. All the samples were kept at 25 ◦ C.
 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429
2.3. In vitro skin permeation of diclofenac
The permeability of DFC through rats’ skin was measured
in vitro with a Franz diffusion cell system (PermeGear, Inc.,
Bethlehem, PA, USA). The diffusion area was 1.767 cm2
(15 mm diameter orifice), and the receptor compartment vol-
umes varied from 11 to 12 ml. The solutions in the receiver
side were stirred by externally driven, Teflon-coated magnetic
bars. Sprague–Dawley rats (males, 200–300 g, Harlan Labo-
ratories, Ltd., Rehovot, Israel) were sacrificed by aspiration
of ethyl ether. The abdominal hair was clipped carefully and
sections of full-thickness skin were excised from the fresh car-
casses. After subcutaneous fat was removed with a scalpel, the
transepidermal water loss (TEWL) was measured before the
skin sections were mounted in the diffusion cells. TEWL ex-
aminations were performed on skin pieces using Dermalab®
Cortex Technology instrument (Hadsund, Denmark), and only
those pieces where the TEWL levels were less than 10 g/(m2 h)
were mounted in the diffusion cells, ready for testing. The
skin was placed on the receiver chambers with the stratum
corneum facing upwards, and then the donor chambers were
clamped in place. The excess skin was trimmed off and the re-
ceiver chamber, defined as the side facing the dermis, was filled
with phosphate buffered saline (PBS, pH 7.4). After 30 min
of skin washing at 37 ◦ C, the buffer was removed from the
cells. Infinite doses of liquid formulations containing 1 wt%
of DFC were applied on the skin, and the receiver chambers
were filled with phosphate buffered saline (pH 7.4). Samples
(2 ml) were withdrawn from the receiver solution at predeter-
mined time intervals, and the cells were replenished to their
marked volumes with fresh buffer solution. Addition of solution
to the receiver compartment was performed with great care to
avoid trapping air beneath the dermis. The samples were taken
into 1.5 ml amber vials and kept at −20 ◦ C until analyzed by
HPLC.
2.4. HPLC analysis of drug from receiver solutions
Aliquots of 20 µl from each sample were injected into the
HPLC system (Shimadzu VP series with an auto sampler and
a diode array detector), equipped with a prepacked C18 col-
umn (Lichrosphere 60 RP-select B, 5 µm, 125 × 4 mm). The
detection of diclofenac was carried out at 275 nm. The samples
were separated using an isocratic mobile phase consisting of
dibasic sodium phosphate (0.008 M, pH 2.5)—methanol (1:3)
at a flow rate of 1 ml/min. Calibration curves (peak area ver-
sus drug concentration) were linear over the range 1–20 µg/ml.
As a result of the sampling from the receiver solution—and the
replacement of these quantities with equal volumes of buffer—
the receiver solution was constantly being diluted. Taking this
process into account, the cumulative drug permeation (Qt ) was
calculated from the equation:

t−1
Qt = Vr Ct + Vs Ci ,
i=0
where Ct is the drug concentration of the receiver solution at
each sampling time, Ci is the drug concentration of the ith sam-
423
ple, and Vr and Vs are the volumes of the receiver solution and
the sample, respectively. In vitro data were expressed as the cu-
mulative diclofenac permeation per unit of skin surface area,
Qt /S (S = 1.767 cm2 ).
3. Results and discussion
3.1. Maximum solubilization of diclofenac sodium salt
Microemulsion systems have previously been shown to
increase the solubilization capacity of water-insoluble drugs
[25–28]. In the present study, we prepared a new system con-
taining a high concentration of phosphatidylcholine and ex-
amined its solubilization capacity for DFC. We measured the
solubilization capacity of the drug along dilution line N100
(dilution line N100: 100% surfactant mixture without an oil
phase, diluted with water). The calculated DFC concentrations
at various dilution compositions along dilution line N100 are
shown in Table 1; the result of the surfactants mixture loaded
with an increased measurable concentration of DFC is seen
in the first row in Table 1. The non-shaded area in Table 1
represents the formation of a clear isotropic system, while the
shaded area represents formulations that are beyond their exper-
imentally determined diclofenac solubilization capacity (i.e.,
containing precipitated drug). It can be seen (Table 1) that a
surfactants mixture which contains up to 25 wt% DFC can be
diluted with water without phase separation or drug precipita-
tion. When the surfactants mixture contains 26 wt% diclofenac,
sedimentation of the drug occurred over a specific range of wa-
ter concentrations (45–85% water). Above 85% water, the drug
is re-solubilized. As the initial DFC concentration (first row
in Table 1) in the surfactant mixture increases above 26 wt%,
a relatively wider range of water concentration is obtained in
which drug sedimentation occurs. The surfactants mixture that
contains more than 48 wt% diclofenac cannot be solubilized
at any water concentration. It can be said, therefore, that the
maximal diclofenac solubilization capacity of this microemul-
sion system is when the drug is dissolved in the surfactants
mixture at 48 wt% concentration. The maximum DFC solubi-
lization capacity of a formulation containing 96 wt% water is
1.92 wt% (19 200 ppm). The pH values all along the N100 di-
lution line were natural (pH 7.6 to 6.6) as we reported in our
previous paper [11]. The solubility of diclofenac under phys-
iological conditions was reported to range from 17.8 ppm at
neutral pH to less than 1 ppm at an acidic pH, while DFC solu-
bility in pure water was reported to be 1113 ppm [29–32]. Thus,
the solubilization capacity at 96 wt% water dilution within the
microemulsion was almost 20 times more than the solubility of
DFC in pure water.
The measured maximum solubilization capacity curve for
DFC (Fig. 2) can be divided into three isotropic regions based
on the deflection in the slope of the curve. The identified regions
are: (I) 0–20 wt% water, (II) 20–55 wt% water, and (III) above
(1) 55 wt% water. In our previous paper [12] we found, by using
several analytical techniques (viscosity, electrical conductivity,
and PGSE-NMR), that upon water dilution the microstruc-
ture of the system (along dilution line N100) is progressively
424 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429
Table 1
Calculated diclofenac concentration in any given water concentration upon the addition of measurable quantity of DFC to the surfactant mixture (0 wt% water),
along dilution line N100 (arrow)
Fig. 2. Measured maximum solubilization capacity curve for diclofenac, along
water dilution line N100 at 25 ◦ C (derived from Table 1).
transferred from reversed micelles (W/O-like structures) into
bicontinuous mesophase. Upon further dilution the bicontin-
uous structure transforms progressively into direct O/W mi-
croemulsions. The microstructural transitions were found to
occur at about 20 and 55 wt% of water, respectively. The pres-
ence of DFC at the interface (penetration into to surfactant
film) all along the dilution line has been proved by a PGSE-
NMR technique that was extensively discussed in our previ-
ous manuscript [12]. Other works used the same techniques
to define the location of different active molecules along the
dilution lines of different microemulsion systems [25–28,33].
Similarly, it has clearly been shown that the solubilization ca-
pacity of diclofenac in this microemulsion system is affected
by the microstructural changes, which occur along the water
dilution line. The diclofenac solubilization capacity decreases
as the water content increases up to 20 wt% water. This signifi-
cant decrease occurs because the reverse-micelles progressively
swell, resulting in a decrease in the interfacial area leading
to decreased diclofenac solubilization capacity. Upon further
water dilution (above 20 wt%), the formation of bicontinuous
microstructure occurs. These transformations were detected by
analytical techniques as described in previous papers [12,34].
Upon dilution of a system with bicontinuous microstructure,
the interfacial area is almost unchanged, thus the decrease in di-
clofenac solubilization capacity is less pronounced (the slope of
the solubilization curve is more shallow; Fig. 2). A sample that
contains more than 55 wt% water seems to possess an O/W-like
structure (direct micelles). Although the transformation from
bicontinuous to O/W-like structure causes a decrease in the in-
terfacial area, the dominant solubilization factor in this system
is the solubility of the drug in the water-continuous phase. The
decrease in diclofenac solubilization capacity along the water
dilution line (N100) can also be explained by the packing of
diclofenac into the interface. From the schematic illustration
(Fig. 3) of diclofenac packing capability along the water dilu-
tion line (N100), it can be seen that in a W/O-like structure
the diclofenac molecules can easily penetrate between the sur-
factant hydrophobic chains. Upon water dilution and formation
of bicontinuous microstructures, the ability of DFC to be em-
bedded between the surfactant hydrophobic chains decreases. It
should be noted that two different bicontinuous microstructures
are formed as the water dilution progresses. At first (Fig. 4a),
elongated worm-like reverse micelles are dispersed in continu-
 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429 425

Fig. 3. Schematic illustration (not to scale) of possible packing of diclofenac along water dilution line N100 (surfactants mixture and water) at four different dilution
regions. Upper right: W/O microemulsion; upper left: W/O-bicontinuous; bottom left: O/W-bicontinuous; bottom right: O/W microemulsion.

Fig. 4. Schematic illustration (not to scale) of bicontinuous microstructures created upon water dilution: (a) W/O-bicontinuous microstructure (oil continuous phase),
(b) O/W-bicontinuous microstructure (water continuous phase).

ous oil phase; upon further dilution the microstructure is con-
verted to elongated worm-like direct micelles dispersed in wa-
ter continuous phase (Fig. 4b). The point of inversion between
reverse-bicontinuous to direct-bicontinuous microstructure can
be detected by viscosity measurements along the dilution line
as we have shown in our previous publication [35]. The water
concentration at which the highest viscosity is detected is con-
sidered the point of inversion. The DFC solubilization seems
to be higher in the W/O-bicontinuous microstructure than in
the O/W-bicontinuous one because in the W/O-bicontinuous
microstructure the surfactant tails are dangling in the oil con-
tinuous phase and the DFC molecules can penetrate into the
surfactant film more easily. On the other hand, in the O/W-
bicontinuous microstructure the surfactant tails are confined to
the inner oil phase, which has less space to accommodate the
DFC molecules. Upon the inversion into O/W-like droplets, the
surfactant is tightly packed around the oil droplets (oil-in-water
droplets) leaving less free space for DFC. Since the packing pa-
426 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429

rameters of the Tween is ca. 1/3 while that of DFC is >1, the than the lag time of diclofenac dissolved in a PG vehicle alone.
packing of DFC into the interface is not favorable, thus its sol- When diclofenac was applied in PG vehicles containing both a
ubilization capacity is the lowest. nonionic surfactant (HECO40) and PC, the cumulative penetra-
tion through rat skin after 24 h was considerably higher (Q24 =
3.2. Diclofenac permeability 58.32 ± 8.15 µg/cm2 ) than the penetration of diclofenac dis-
solved in a PG vehicle alone or solubilized in a vehicle con-
3.2.1. Effect of PC on diclofenac penetration taining HECO40 only, as a surfactant. The lag time after ap-
Many materials such as surfactants, fatty acids, terpenes, plication of a vehicle containing PC is similar (6.04 ± 2.09 h)
dimethylsulfoxide (DMSO), Azone analogs, alcohols, and wa- to the lag time after application of a vehicle which contained
ter are known to enhance topical and transdermal penetration HECO40 only as a surfactant. By replacing the nonionic surfac-
of active materials [36,37]. In a recent publication [12] we have tant (HECO40) with Brij 96v, no differences in the diclofenac
shown that PC affects diclofenac penetration through Caco-2 cumulative penetration after 24 h or in the permeability coeffi-
monolayer cells. The permeability of Caco-2 cells can provide cient (Q24 = 52.71 ± 1.83 µg/cm2 and P = 3.48 ± 0.20 cm/h)
a good estimation of the permeability of an active drug through were noted, although the lag-time increased to 9.05 ± 0.48 h.
Thus, it seems that both nonionic surfactants had little influence
the intestinal wall. In the present study we have demonstrated
on the extent of diclofenac penetration; however, the combina-
the importance of PC to micellar and microemulsion systems
tion of PC with HECO40 improved the extent of penetration
and its contribution to the permeability enhancement of di-
(Q24 ), increased the permeation rate (P ) and shortened the lag-
clofenac through the skin.
time (TL ) (Table 2). From this set of experiments it can be pos-
The permeability of diclofenac through rat skin, using a
tulated that this system, which contains PC molecules, can serve
Franz cell, was evaluated after applications of the following for-
mulations: (1) DFC in PG as a vehicle, (2) DFC solubilized in
a formulation containing a nonionic surfactant, HECO40, with
and without PC, (3) DFC solubilized in a formulation contain-
ing PC and Brij 96v as another nonionic surfactant, and (4) DFC
solubilized in SMEPS formulation containing PG, HECO40,
PC molecules, MCT, and ethanol. The Franz diffusion cell has
been used to evaluate drug permeation through a skin barrier
[38]. The effect of the surfactants on the cumulative penetra-
tion of diclofenac is shown in Fig. 5. It can be seen that when
applied with PG alone, diclofenac has a relatively low cumu-
lative penetration after 24 h (Q24 = 33.16 ± 11.24 µg/cm2 ;
Table 2). It can also be seen from Fig. 5 and Table 2 that a
steady state permeation of the drugs starts after a lag-time of
10.39 h. When DFC was applied in a formulation containing
only HECO40 as a surfactant and PG as the solvent, the di-
clofenac cumulative penetration after 24 h and the permeability
coefficient were comparable (Q24 = 37.38 ± 2.66 µg/cm2 and
Fig. 5. Cumulative penetration of diclofenac through rat skin in different de-
P = 2.00 ± 0.05 cm/h) to DFC penetration from a PG vehicle livery systems. The formulation contains: PG (F); PG and HECO40 (2); PC,
alone. However, the lag-time of diclofenac applied in formula- Brij 96v, and PG (Q); PC, HECO40, and PG (×). All formulations contain: 1
tions containing PG and HECO40 were shorter (5.05 ± 1.98 h) wt% diclofenac and 50 wt% water, at 25 ◦ C. All formulations were clear.

Table 2
Summary of the permeability parameters [quantity of drug permeated after 24 h (Q24 ), permeability coefficient (P ), and lag time (TL )] obtained from the various
formulations of diclofenac sodium salt
Formulation/vehiclea Q24 (µg/cm2 ) P (cm/h) TL (h)
PG/50% water 33.16 (±11.24) 2.37 (±0.86) 10.39 (±0.47)
HECO40/PG/50% water 37.38 (±2.66) 2.00 (±0.05) 5.05 (±1.98)
PC/Brij 96v /PG/50% water 52.71 (±1.83) 3.48 (±0.20) 9.05 (±0.48)
PC/HECO40/PG/50% water 58.32 (±8.15) 2.95 (±1.44) 6.04 (±2.09)
PC/HECO40/PG/15% water 53.61 (±3.12) 3.52 (±0.12) 8.73 (±0.72)
PC/HECO40/PG/70% water 46.63 (±12.88) 2.82 (±1.26) 5.44 (±0.59)
PC/L-1695/HGL/PG/15% water 13.95 (±4.90) 0.89 (±0.18) 8.62 (±1.70)
PC/L-1695/HGL/PG/50% water 30.48 (±5.42) 1.80 (±0.35) 6.54 (±0.43)
PC/L-1695/HGL/PG/80% water 33.08 (±5.11) 1.82 (±0.29) 5.65 (±0.61)
PC/HECO40/PG/33% water 40.40 (±8.09) 2.73 (±0.43) 8.16 (±1.25)
PC/HECO40/PG/33% water/EtOH/MCT 48.01 (±9.05) 2.92 (±0.97) 6.19 (±0.98)
a Abbreviations: EtOH—ethyl alcohol, HGL—hexaglycerol laurate, L-1695—sugar ester, MCT—medium chain triglycerides, PC—phosphatidylcholine, PG—
propylene glycol.
 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429
Fig. 6. Cumulative penetration of diclofenac through rat skin delivered in for-
mulations containing different water concentrations. All formulations contain:
PC, HECO40, and PG at 1:3:10 wt ratio as surfactant, 1 wt% diclofenac and
water along dilution line N100, at 25 ◦ C. All formulations were clear and trans-
parent.
as a good vehicle for transdermal delivery of diclofenac, as the
PC molecules actually enhance the drug permeability. Several
modes of action were proposed for the improvement of active-
molecule absorption in the presence of PC molecules: (1) PC
that is embedded at the interface can act as an adhesion agent,
thus improving the active-molecule’s permeability [39]; (2) PC
molecules create a complex with the active-molecules forming
‘ion-pairs,’ thus altering the hydrophilic–lipophilic balance and
improving absorption kinetics [39]; (3) systems embedded with
PC can form vesicles (e.g., as in liposomes), providing some
protection to the active-molecules against chemical and enzy-
matic degradation [39]; (4) alteration of the epidermal lipid
barrier properties and alteration in the membrane fluidity lead
to increased active molecule diffusion and absorption through
the skin [39,40].
3.2.2. Effect of the microstructure on diclofenac penetration
In Section 3.1 and in our previous studies [11,12], we have
shown that the microemulsion’s microstructure has a vital role
in the ability of the system to solubilize and to deliver active
molecules such as diclofenac. We have demonstrated that the
solubilization capacity of diclofenac is greatly dependent on the
microstructure of the surfactants system while water is added.
The question is, therefore, whether the skin permeability for
the drug is also influenced by the microstructure of this for-
mulation. We examined the permeability of diclofenac applied
on skin in three formulations having different microstructures:
(1) containing 15 wt% water having a W/O-like microstruc-
ture, (2) containing 50 wt% water having a bicontinuous mi-
crostructure, and (3) containing 70 wt% water having a O/W-
like microstructure. The cumulative penetration of diclofenac
after 24 h and the permeability as shown in Fig. 6 seems to
be minimally affected by microstructure changes. Although not
basically different, it should be noted that formulations with bi-
continuous and O/W-like microstructures look slightly superior
427
Fig. 7. Cumulative penetration of diclofenac through rat skin in formulations
containing different water concentrations along dilution line N100, at 25 ◦ C.
The systems were stabilized by a mixture of PC and nonionic surfactant, sugar
ester L-1695, hexaglycerol laurate, and PG in 1:1.5:1.5:10 wt ratios. All formu-
lations contained 1 wt% diclofenac. All formulations were clear and transpar-
ent.
to the formulation having a W/O-like microstructure; however,
by replacing the HECO40 surfactant with a mixture of nonionic
surfactants (sugar ester L-1695 and hexaglycerol laurate in 1:1
wt ratio), a more pronounced effect (Fig. 7) resulted. It can
be clearly seen that the cumulative penetration and the perme-
ability coefficients of diclofenac applied in systems having bi-
continuous and O/W-like microstructures (Q24 = 30.48 ± 5.42
and 33.08 ± 5.11 µg/cm2 and P = 1.80 ± 0.35 and 1.82 ±
0.29 cm/h, respectively) are much higher than the cumulative
penetration and the permeability coefficient of diclofenac when
applied in a system with a W/O-like microstructure (Q24 =
13.95 ± 4.90 µg/cm2 and P = 0.89 ± 0.18 cm/h). Moreover,
diclofenac penetration after 24 h and its permeability are the
highest when administrated as an O/W-like microstructure.
These results seem to be in agreement with the results of
Djordjevic et al. [1] who studied the influence of the mi-
crostructure of microemulsions on the in vitro release rate of
the amphiphilic drug diclofenac diethylamine through a re-
generated cellulose membrane. The microemulsion was com-
posed of water, PEG-8 caprylic/capric glycerides, hexaglyc-
erol dioleate, and isopropyl myristate. It was found that the
release of diclofenac diethylamine from a formulation with
W/O structure, in which the drug is predominantly located in
the water droplets, was hampered by the external hydropho-
bic phase. In the case of bicontinuous and nonspherical water-
continuous microemulsions, the experimental data suggest that
those microstructures slightly hamper the amphiphilic drug re-
lease, probably due to drug/vehicle interactions. The maximum
diclofenac diethylamine permeation was obtained from a mi-
croemulsion containing the highest percentage of water having
an O/W microstructure. This observation was consistent with
the previously detected higher diclofenac diethylamine diffu-
sion coefficient from the lotion (containing a huge quantity of
428 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429
Fig. 8. Cumulative penetration of diclofenac through rat skin delivered in for-
mulations with and without oil phase. Concentrate contains 20 wt% oil phase,
dilution line N82 (2) contains: ethanol and MCT at 1:1 wt ratio as oil phase
and PC, HECO40, and PG at 1:3:10 wt ratio as surfactants mixture. Concen-
trate without oil phase (Q) contains only PC, HECO40, and PG at 1:3:10 wt
ratio as surfactant. All formulations contain 1 wt% diclofenac and 33 wt% wa-
ter at 25 ◦ C. All formulations were clear and transparent.
continuous water phase) compared to the lipogel and the mixed
micelle gel [41].
Adding 20 wt% oil phase (dilution line N82, Fig. 1),
composed of ethanol and MCT, to the surfactants mixture
(PC/HECO40/PG in 1:3:10 wt ratio) slightly increased the cu-
mulative penetration (Q24 = 48.01 ± 9.05 µg/cm2 in line N82
compared to 40.40 ± 8.09 µg/cm2 in line N100) of diclofenac
after 24 h, but somewhat more improvement was observed in
decreasing the lag-time from 8.16 (±1.25) h to 6.19 (±0.98) h
(Fig. 8). This result has generally been expected, since the in-
corporation of a lipophilic phase, especially ethanol, enhances
drug permeability through the skin [42]. A comparison be-
tween O/W-like formulation without ethanol and a W/O-like
formulation containing MCT and ethanol (Fig. 9) revealed that
incorporation of ethanol in the formulation is basically unnec-
essary, in particular since the use of ethanol should generally be
avoided due to its skin irritation.
4. Summary
The present study demonstrates that micellar systems em-
bedded with membrane transport enhancers such as phos-
phatidylcholine can enhance the transport of DFC across the
skin. To achieve a high transport efficacy it is essential to form
U-type phase diagrams composed of several ingredients so that
any concentrate will be progressively and fully dilutable with
aqueous phase. Such concentrate will transform from reverse
micellar vehicles into W/O microemulsions and via a bicontin-
uous mesophase into O/W microdroplets. The concentrate can
solubilize high levels of DFC. We examined the maximum solu-
bilization capacity at any dilution point along dilution line N100
(without oil/ethanol) and correlated it to the microstructural
transitions occurring along the dilution line. The solubilization
capacities decreased as a function of water dilution and of the
Fig. 9. The effect of microstructure and the presence of ethanol on the cu-
mulative penetration of diclofenac through rat skin delivered in formulations
containing: formulation with W/O microstructure containing ethanol (ethanol
and MCT at 1:1 wt ratio as oil phase and PC, HECO40, and PG at 1:3:10 wt
ratio as surfactants mixture and 33 wt% water) (2); formulation with O/W
microstructure without ethanol (PC, HECO40, and PG at 1:3:10 wt ratio as sur-
factant, and 70 wt% water) ("). All formulations were clear, transparent, and
contain 1 wt% diclofenac, at 25 ◦ C.
interfacial capacity at each microstructure. It was also found
that transport from the inverted bicontinuous and O/W-like mi-
crostructure was higher than that of the W/O-like structures. PC
was found to increase drug transport across rat skin compared
to similar micellar systems that did not contain PC embedded
at their interface.
References
[1] L. Djordjevic, M. Primorac, M. Stupar, Int. J. Pharm. 296 (2005) 73.
[2] M. Kreilgaard, E.J. Pedersen, J.W. Jaroszewski, J. Controlled Release 69
(2000) 421.
[3] M. Kreilgaard, M.J. Kemme, J. Burggraaf, R.C. Schoemaker, A.F. Cohen,
Pharm. Res. 18 (2001) 593.
[4] B. Baroli, M.A. Lopez-Quintela, M.B. Delgado-Charro, A.M. Fadda,
J. Blanco-Mendez, J. Controlled Release 69 (2000) 209.
[5] P. Spiclin, M. Homar, A. Zupancic-Valant, M. Gasperlin, Int. J. Pharm.
256 (2003) 65.
[6] A.C. Sintov, L. Shapiro, J. Controlled Release 95 (2) (2004) 173.
[7] A.C. Sintov, S. Botner, Int. J. Pharm. 311 (1–2) (2006) 55.
[8] A. Spernath, A. Aserin, Adv. Colloid Interface Sci. 128–130 (2006) 47.
[9] X. Sha, G. Yan, Y. Wu, J. Li, X. Fang, Eur. J. Pharm. Sci. 24 (2005) 477.
[10] A. Spernath, A. Aserin, N. Garti, J. Therm. Anal. Calorim. 83 (2) (2006)
297.
[11] A. Spernath, A. Aserin, N. Garti, J. Colloid Interface Sci. 299 (2) (2006)
900.
[12] A. Spernath, A. Aserin, L. Ziserman, D. Danino, N. Garti, J. Controlled
Release 119 (2007) 279.
[13] T. Nishihata, A. Kamada, K. Sakai, K. Takahahi, K. Matsumoto, K. Shi-
nozaki, Y. Tabat, M. Keigami, T. Miyagi, N. Tatsumi, Int. J. Pharm. 46
(1988) 1.
[14] S. Naito, H. Tominaga, Int. J. Pharm. 24 (1985) 115.
[15] T. Nishihata, K. Kotera, Y. Nakano, M. Yamazaki, Chem. Pharm. Bull. 35
(1987) 3807.
[16] M. Kreilgaard, Adv. Drug Deliv. Rev. 54 (Suppl.) (2002) S77.
[17] C. Corswant, P. Thoren, S. Engström, J. Pharm. Sci. 87 (1988) 200.
[18] P. Karande, A. Jain, S. Mitragotri, J. Controlled Release 110 (2006) 307.
 A. Spernath et al. / Journal of Colloid and Interface Science 318 (2008) 421–429
[19] P. Karande, A. Jain, S. Mitragotri, J. Controlled Release 115 (2006) 85.
[20] J. Hadgraft, Int. J. Pharm. 224 (2001) 1.
[21] M. Goodman, B.W. Barry, in: R.L. Bronaugh, H. Maibach (Eds.), Percu-
taneous Absorption: Mechanisms–Methodology–Drug Delivery, Marcel
Dekker, New York, 1989, p. 567.
[22] M. Kirjavainen, A. Urtti, J. Mönkkönen, Eur. J. Pharm. Sci. 10 (2) (2000)
97.
[23] D.Z. Liu, E.L. LeCluyse, D.R. Thakker, J. Pharm. Sci. 88 (11) (1999)
1161.
[24] D.Z. Liu, S.L. Morris-Natschke, L.S. Kucera, K.S. Ishaq, D.R. Thakker,
J. Pharm. Sci. 88 (11) (1999) 1169.
[25] A. Spernath, A. Yaghmur, A. Aserin, R.E. Hoffman, N. Garti, J. Agric.
Food Chem. 50 (23) (2002) 6917.
[26] A. Spernath, A. Yaghmur, A. Aserin, R.E. Hoffman, N. Garti, J. Agric.
Food Chem. 51 (8) (2003) 2359.
[27] N. Garti, M. Avrahami, A. Aserin, J. Colloid Interface Sci. 299 (1) (2006)
352.
[28] N. Garti, A. Yaghmur, A. Aserin, A. Spernath, R. Elfakess, S. Ezrahi, Col-
loids Surf. A 230 (1–3) (2004) 183.
[29] K. Tahereh, J. Fakhreddin, J. Pharm. Pharm. Sci. 6 (3) (2003) 352.
[30] R. Menasse, P.R. Hedwall, J. Kraetz, C. Pericin, L. Riesterer, A. Sallmann,
R. Ziel, R. Jaques, Scand. J. Rheumatol. 22 (Suppl.) (1978) 5.
429
[31] A. Chiarini, A. Tartarini, A. Fini, Arch. Pharm. 317 (3) (1984) 268.
[32] A. Fini, M. Laus, I. Orienti, V. Zecchi, J. Pharm. Sci. 75 (1) (1986) 23.
[33] N. Garti, I. Amar, A. Spernath, Phys. Chem. Chem. Phys. 6 (2004)
2968.
[34] A. Kogan, N. Garti, A. Aserin, J. Colloid Interface Sci. 315 (2) (2007) 637.
[35] D. Libster, A. Aserin, N. Garti, J. Colloid Interface Sci. 302 (1) (2006)
322.
[36] P. Karandea, A. Jaina, A.V. Aroraa, M.J. Hoa, S. Mitragotri, Eur. J. Pharm.
Sci. 31 (1) (2007) 1.
[37] S.T.K. Narishetty, R. Panchagnula, J. Controlled Release 102 (1) (2005)
59.
[38] D.Z. Liu, M.T. Sheu, C.H. Chen, Y.R. Yang, H.O. Ho, J. Controlled Re-
lease 118 (3) (2007) 333.
[39] L.M. Lichtenberger, Z.M. Wang, J.J. Romero, C. Perez, M.N. Giraud, J.C.
Barreto, Nat. Med. 1 (2) (1995) 154.
[40] L.M. Lichtenberger, J.J. Romero, W.M.J. De Ruijter, F. Behbod, R. Dar-
ling, A.Q. Ashraf, S.K. Sanduja, J. Pharmacol. Exp. Ther. 298 (1) (2001)
279.
[41] S. Parsaee, M.N. Sarbolouki, M. Parnianpour, Int. J. Pharm. 241 (2002)
185.
[42] Y.C. Chao, S.F. Chang, S.C. Lu, T.C. Hwang, W.H. Hsieh, J. Liaw, J. Con-
trolled Release 118 (1) (2007) 105.