Journal of Controlled Release 67 (2000) 211–221

www.elsevier.com / locate / jconrel

Reduction in skin permeation of N,N-diethyl-m-toluamide

(DEET) by altering the skin / vehicle partition coefficient q
a ,1 a,
Julie S. Ross , Jaymin C. Shah *
a
Department of Pharmaceutical Sciences, Medical University of South Carolina, 171 Ashley Avenue, Charleston, SC 29425, USA

Received 2 November 1999; accepted 19 January 2000

Abstract

Reported adverse side effects after using N,N-diethyl-m-toluamide (DEET)-containing mosquito repellent products appear
to be the result of significant absorption of DEET through human skin. The overall objective was to develop formulations of
DEET with significantly reduced permeation using the basic principles and model of skin permeation based on Fick’s laws of
diffusion at steady state. Ternary phase diagrams of DEET with water and semipolar solvents, ethanol, PG and PEG 400,
showed an increase in the aqueous solubility of DEET. This resulted in a linear decline in octanol / water PC with an increase
in the concentration of the solvent. Permeation of DEET across human skin was studied from vehicles containing various
amounts of PG and PEG 400 using an infinite dose technique and Franz diffusion cell. DEET’s flux reduced with increasing
PG concentration and the flux from 90% PG was 9.962.1 mg / cm 2 h, 6-fold lower than flux of pure DEET control,
63.2624.5 mg / cm 2 h. Flux was reduced 6-fold from 60% PEG 400 solution, and permeation of DEET was totally prevented
from 90% PEG 400 which was very viscous. However, a combination of 60% PEG 400 with 30% PG not only reduced
permeation 9-fold but was suitable as a vehicle for formulation. The decrease in flux and permeability of DEET with
increasing concentration of solvent appeared to be a direct result of decrease in skin / vehicle PC and octanol / water PC. This
study clearly demonstrates that alternative formulations can be developed for DEET aimed at reduced permeation and
toxicity unlike the current formulations some of which contain ethanol which has been shown to enhance permeation of
DEET. A similar approach can be used for developing formulations of other industrial and occupational agents to prevent
their skin permeation when a user may be exposed to them.  2000 Elsevier Science B.V. All rights reserved.

Keywords: DEET; Mosquito repellent; Solvent; Skin permeation; Partition coefficient; Propylene glycol; Polyethylene glycol; Ethanol

q
This research was conducted as dissertation research by J.
Stinecipher towards partial fulfillment of the requirements of
doctor of philosophy degree at the Medical University of South
Carolina.
*Corresponding author. Present address: Pharmaceutical
RandD, Pfizer Central Research, Ms 8156-39, Eastern Point Road,
Groton, CT 06340, USA. Tel.: 11-860-715-2332; fax: 11-860-
441-0467.
1
Present address: IBAH Pharmaceutics Services, 525 Virginia
Drive, Fort Washington, PA 19034, USA.
1. Introduction
Use of N,N-diethyl-m-toluamide (DEET) contain-
ing mosquito repellent products have been associated
with many adverse reactions [1–4]. Recently, DEET
has also been implicated as a factor in neurotoxicity
resulting from the Gulf War chemical exposure [5].
A review of the biodistribution and toxicity of DEET
by Robbins and Cherniack suggested that there were
several areas of toxicity that needed further evalua-

0168-3659 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved.
PII: S0168-3659( 00 )00210-8
212 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
tion [6]. DEET is absorbed through the skin very
rapidly as shown by several investigators [7–10].
Our previous studies have shown not only significant
absorption of DEET from all commercial products
but enhanced permeation of DEET across human
skin from 15 to 60% ethanol, a solvent for DEET in
many products [11]. Therefore, it is very evident that
dermal absorption of DEET may be responsible for
the potential toxicity problems associated with
DEET-containing products, and therefore, new or
alternative formulations of DEET specifically aimed
at decreased permeation and absorption will be safe
and useful [12,13].
In order to develop these alternative formulations
with decreased absorption of DEET, the basic princi-
ples of skin permeation should be applied to DEET.
The simplest model to describe dermal absorption
based on Fick’s laws of diffusion at steady state is as
follows [14]:
dc / dt ss 5 Jss 5 DAK /h (Cd 2 Cr )
where Jss 5steady state flux or the permeation rate;
A5surface area of the skin exposed or of the
application site; h5the thickness of the rate-limiting
barrier, usually the stratum corneum; D5the diffu-
sion coefficient of the diffusant, K5the skin / vehicle
partition coefficient; Cd 5the thermodynamic activity
of the diffusant in the vehicle and Cr 5the con-
centration of the diffusant in receptor, close to zero
under sink conditions.
Therefore, only D, K, and Cd may be altered to
reduce the permeation of diffusant. By selecting an
appropriate vehicle, the thermodynamic activity and
the skin / vehicle partition coefficient can be reduced
to minimize skin permeation, although, high thermo-
dynamic activity of DEET in vehicle will be required
to maintain its ability to repell mosquitos. The
diffusion coefficient is relatively difficult to alter
since it is dependent on a number of factors such as
the skin characteristics, effect of vehicle on skin,
skin hydration and interaction of diffusant with skin.
Therefore, using an appropriate vehicle, either K, D,
and / or the skin can be influenced to reduce permea-
tion of DEET using this basic principle in order to
develop improved formulations of DEET.
The specific aim of the present study was to
reduce and minimize skin permeation of DEET by
reducing its skin / vehicle PC (K) using semipolar
solvents which may decrease DEET’s n-octanol /
water PC. The permeation of DEET from each
vehicle was evaluated by performing in vitro per-
meation studies across human skin with a Franz
diffusion cell using an infinite dose technique. Per-
meation profiles were constructed in each case and
the steady state flux, diffusion coefficient, lag time,
permeability, and skin / vehicle partition coefficient
were calculated to investigate the mechanism of
reduction in flux. The permeation of DEET from
each vehicle was compared to that of pure DEET in
an effort to determine the reduction in flux and
overall permeation.
2. Materials
DEET (97%) was purchased from Aldrich Chemi-
cal Company. Propylene glycol, and methanol
(HPLC grade) were purchased from Fisher (Fair
(1)
Lawn, NJ, USA). Carbowax  Sentry  Polyethylene
glycol 400 NF, FCC Grade was a gift from Union
Carbide (Danbury, CT, USA). Phosphate buffer
solution (0.1 M, pH 7.4) was prepared with sodium
phosphate monobasic monohydrate (Mallinckrodt,
St. Louis, MO, USA) and sodium phosphate dibasic
anhydrous (Curtin Matheson Scientific,Houston, TX,
USA).
3. Methods
3.1. Phase solubility diagram; effect of solvent on
aqueous solubility of DEET
Since DEET is a liquid at room temperature, a
ternary phase diagram was constructed for DEET,
water, and each solvent: ethanol, propylene glycol
(PG) or polyethylene glycol 400 (PEG 400), to
determine the miscibility of the three liquids. A pair
of liquids such as DEET and water that are miscible
with each other only to a slight extent would usually
exist as two phases. The addition of a third com-
ponent to the system that is miscible in all pro-
portions with one of the solvents would be expected
to produce a single phase in which all three com-
ponents are miscible with each other. Solutions were
prepared consisting of PG, PEG 400 or ethanol, and
 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
varying percentages of DEET from 1 to 99%. Water
was added to these solutions in 100 ml increments
until the solutions appeared turbid after vortexing for
30 s. A ternary plot was constructed in order to
determine the binodal curve, or the boundary of the
two-phase area as seen in Fig. 1. Regions above the
binodal curve represent homogeneous solutions of
the three liquids. Homogeneous or single phase
solutions containing 10% DEET and varying per-
centages of PG and PEG 400 were selected for in
vitro permeation experiments.
3.2. Effect of solvent on the n-octanol /water
partition coefficient of DEET
n-Octanol and water with various concentrations
of each solvent were presaturated with each other in
an amber colored bottle for 48 h. The two phases
were separated using a separating funnel. DEET was
added in small amounts to a mixture of equal
volumes of the octanol and water phases in a
scintillation vial, and allowed to agitate for 48 h to
reach equilibrium. At the end of 48 h, the octanol
and water phases were separated and individually
analyzed for DEET concentration using UV spectro-
photometric assay. The o / w partition coefficient was
calculated from the ratio of concentrations in the two
phases, octanol and water, and plotted as a function
of solvent concentration (Fig. 2). However, the
Fig. 1. Ternary phase diagrams of solvent, water and DEET
showing the homogeneous (single phase) and heterogeneous
regions.
213
Fig. 2. Effect of solvent on the apparent n-octanol / water partition
coefficient of DEET.
experimentally determined PC is an apparent o / w
PC since each solvent will itself partition between
octanol and water and thus change composition of
each phase and influence the PC of DEET directly.
3.3. Skin preparation
Human skin samples were obtained from elective
plastic surgery, and the fat and other visceral debris
were removed from the underside of the freshly
excised skin. The skin was then washed with 0.1 M,
pH 7.4 phosphate buffer solution before freezing at
2208C. The full thickness skin samples were cut
into 2 cm 2 pieces and allowed to thaw overnight at
room temperature in phosphate buffer solution prior
to the in vitro percutaneous permeation experiments.
3.4. In vitro percutaneous permeation
The permeation experiments were conducted with
vertical Franz diffusion cells (Crown Glass Com-
pany, Somerville, NJ, USA) each having a receptor
volume of 4.9 ml and a diameter of 0.9 cm. Pieces of
full thickness human skin (2 cm 2 ), which had been
previously thawed were mounted on the receptor
compartment of the diffusion cells. The receptor
compartment was filled with 0.1 M, pH 7.4 PBS
which was stirred continuously to ensure uniform
distribution and maintain sink conditions, and the
temperature of the entire diffusion cell assembly was
maintained at 378C using a recirculating water
jacket.
214 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221

The permeation studies were conducted using the

infinite dose technique in which a large excess of
permeant (46.88 mg / cm 2 ) is applied to the skin in
comparison to the amount permeating so that an
approximately constant concentration of the per-
meant is maintained in the donor compartment
during the entire course of the experiment. Homoge-
neous solutions consisting of 10% (100 mg / ml)
DEET and 45, 60, 75, and 90% PG in water were
evaluated for DEET permeation to study the effect of
PG on the permeation of DEET. An aliquot of 300
ml (100 mg / ml DEET) of the above homogeneous
solutions was applied to 0.64 cm 2 of skin in the
donor compartment (46.88 mg / cm 2 ) and then cov-
ered with a glass slip to prevent evaporation.
Aliquots of 300 ml of the receptor fluid were
withdrawn and replaced periodically with fresh
phosphate buffer for 36 h. All samples were refriger- Fig. 3. Average permeation profiles of DEET from 45% PG (m),
60% PG (3), 75% PG (♦), and 90% PG (j) solutions, where
ated at 48C until analysis for DEET using HPLC.
n56, as compared to DEET control (*), where n58. All in vitro
The permeation study for DEET from PEG 400 permeation studies were performed at 378C through full thickness
was conducted in a similar manner by applying human skin, and 46.88 mg / cm 2 of DEET was applied to the skin.
homogeneous solutions consisting of 10% DEET and
60, 75, and 90% PEG 400 in water. In addition to
vehicles containing PG or PEG, a homogeneous
solution consisting of 10% DEET, 30% PG, and 60%
PEG 400 was also studied for the permeation of
DEET across human skin in a similar manner at the
same exposure level (46.88 mg / cm 2 ). In all permea-
tion studies, exposure of skin to DEET was constant
at a level of 46.88 mg / cm 2 , and in no case did the
total amount permeated exceed 10% of the applied
DEET. The permeation profiles of DEET from PG,
PEG 400 and the combination were constructed
(Figs. 3–5). In an earlier study, the effect of ethanol
on permeation of DEET from aqueous solutions was
studied at the same exposure level [11]. The permea-
tion parameters obtained from the previous ethanol
study were compared to the PG and PEG 400 results
to compare the effects of solvents on DEET’s skin
permeation and understand the mechanism for reduc-
tion in flux (Figs. 6).

3.5. HPLC analysis

DEET was analyzed by reverse phase HPLC on an
Alltech C8 column (5 ml, 25 cm34.6 mm) and
eluted with the mobile phase at a flow-rate of 0.7
ml / min. The mobile phase consisting of methanol–
Fig. 4. Average permeation profiles for DEET from DEET control
(♦), 60% PEG 400 (j), 75% PEG 400 (m), and 90% PEG 400
(3), where (n53) except for the control where (n58). All in
vitro permeation studies were performed at 378C through full
thickness human skin, and 46.88 mg / cm 2 of DEET was applied to
the skin.
 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
Fig. 5. Average permeation profiles of DEET from PEG 400 / PG
mixture (j) as compared to DEET control (♦), where n56 for
PEG 400 / PG and n58 for the control. All in vitro permeation
studies were performed at 378C through full thickness human skin,
and 46.88 mg / cm 2 of DEET was applied to the skin.
water (80:20, v / v) was filtered through a Versapor
0.8 mm filter (Millipore) and degassed under vac-
uum. DEET was detected using UV absorbance at
240 nm at an average retention time of 8 min, and
the minimum detectable level (MDL) for DEET was
24 ng / injection.
3.6. Data analysis
The cumulative amount (A) in mg / cm 2 of DEET
permeating into the receptor compartment was
plotted against time (t) to obtain the permeation
profile. The steady state flux (Jss ) was estimated
from the slope of the linear portion of the profile.
The J-shaped curve can be described by the follow-
ing equation:
A 5 Jss (t 2 t L ).
The lag time (t L ) was estimated from the x-
intercept of the linear portion of the profile and was
used to calculate the apparent diffusion coefficient
(D) as follows:
215
D 5 h 2 / 6t L (3)
where h is the thickness of stratum corneum, the rate
limiting barrier, and assumed to be 0.001 cm [15].
From the values of Jss , D, h, and Cd , permeability
(P), and skin / vehicle partition coefficient (K) were
calculated using the following equations:
P 5 Jss /Cd (4)
K 5 Ph /D (5)
The estimated parameters are presented as
mean6standard deviation (S.D.) and were evaluated
for differences using an ANOVA test at P,0.05.
The flux, P and K were plotted as a function of
solvent concentration to understand the mechanism
for reduction in flux (Fig. 6a,b,c).
4. Results
The ternary phase diagrams for DEET, water, and
PG or PEG 400 are shown in Fig. 1. The binodal
curve denoted by symbols separates the area where
the three liquids are miscible with each other and the
area where a heterogeneous mixture exists. The
ternary phase diagrams for PG and PEG 400 are very
similar and follow the same trend as that for ethanol
obtained earlier (Fig. 1). As seen from the ternary
diagrams, DEET and water are almost immiscible,
however with addition of solvents, homogeneous
solutions containing DEET and water can be pre-
pared. Although all three solvents, ethanol, PG and
PEG 400, behave similarly, from the relative size of
the homogeneous area in the phase diagram, ethanol
is a better solvent than PG, which is slightly better
than PEG 400 for DEET. Thus, it is not surprising
that many DEET products contain ethanol as the
solvent [11]. The phase diagrams in Fig. 1 show that
to obtain a homogeneous solution containing 10%
(2) DEET, it must contain at least 40% PG or at least
50% PEG 400. Therefore, several homogeneous
solutions containing 10% DEET were chosen for the
permeation studies to evaluate the effect of PG
(.40%) and PEG 400 (.50%) on the permeation
of DEET.
216 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221

Fig. 6. Effect of solvent on (a) steady state flux, (b) permeability, and (c) skin / vehicle partition coefficient (K 310 21 ) of DEET in
permeation through whole thickness human skin at 378C from 10% DEET solutions in solvent–water mixtures.

4.1. Effect of solvent on n-octanol /water partition
coefficient of DEET
The octanol / water partition coefficients (o / w PC)
of DEET are plotted as a function of solvent
concentration in Fig. 2. Although, o / w PC were
calculated from the ratio of concentrations in the two
phases, the experimentally determined PC is an
apparent o / w PC since each solvent partitioned
between octanol and water and thus altered the
composition of each phase and influence the PC of
DEET directly. As expected, DEET is moderately
lipophilic with an o / w PC of approximately 50,
which decreases as the concentration of semipolar
solvent increases for all three solvents; ethanol, PG
and PEG 400. At 60% level of solvent, o / w PC
reduced to 6–10, and with further addition of
solvent, no DEET was detected in the octanol phase
suggesting solubilization of DEET in the water–
solvent mixture. No significant differences were
observed between the different cosolvents with re-
spect to their effect on o / w PC of DEET.
4.2. Effect of propylene glycol ( PG) on the
percutaneous permeation of DEET
The permeation profiles of DEET from various PG
solutions are shown in Fig. 3 as the cumulative
amount of DEET permeated across the skin versus
time. The profiles represent an average of six experi-
 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
ments for each solution and are compared to the
permeation of pure DEET (n58). DEET was found
to continuously permeate throughout the 36 h time
period from each solution, however permeation
profiles from all PG solutions were lower than that of
DEET control. The permeation of DEET across
human skin from the PG solutions decreased with an
increase in the percentage of PG. At 36 h the
cumulative amount of DEET permeated can be
ranked as: pure DEET.45% PG.60% PG.75%
PG.90% PG. The cumulative amounts permeated
from the 75 and 90% PG solutions were significantly
lower (P,0.05) than that from the 45, 60% PG
solutions and DEET control.
The permeation parameters of DEET were ob-
tained from the typically J-shaped profiles using Eqs.
(2)–(5), and listed in Table 1. DEET’s flux ranged
from 59.5669.73 mg / cm 2 h at 60% PG to
9.9062.06 mg / cm 2 h at 90% PG, depending on the
concentration of PG and was lower than the flux of
DEET control, 63.20624.52 mg / cm 2 h. Flux of
DEET from 75 and 90% PG were significantly lower
than the flux from the 45, 60% PG solutions and the
control (Table 1, Fig. 3). PG had no effect on the lag
times and diffusion coefficients of DEET since they
were similar for all of the solutions. The permeabili-
ty of DEET through full thickness human skin
decreased with increasing PG concentrations with
highest permeability from 60% PG at 5.983
10 24 60.98310 24 cm / h and lowest permeability at
0.99310 24 60.21310 24 cm / h from 90% PG solu-
tion. Again, the permeability of DEET from the 75
and 90% PG solutions were significantly lower (P,
0.05) than that from the 45 and 60% PG solutions.
The skin / vehicle partition coefficient (K) of DEET
Table 1
Permeation parameters of DEET from propylene glycol (PG) solutions through full thickness human skin (46.88 mg / cm 2 ) at 378C a
% PG Jss tL
(mg / cm 2 h) (h)
45 57.85626.85 12.6961.62
60 59.5669.73 16.9861.21
75 14.5963.33 b 14.9661.85
90 9.9062.06 b 9.9062.06
DEET (n58) 63.20624.52 12.6264.05
a
Values are presented as mean6S.D. (n56); 100 mg / ml DEET was applied to 0.64 cm 2 of skin.
b
Jss for DEET from 75 and 90% PG is significantly lower (P,0.05) than Jss for DEET from 45 and 60% PG and the DEET control.
c
P and K for DEET from 75 and 90% PG are significantly lower (P,0.05) than P and K for DEET from 45 and 60% PG.
217
was significantly lower (Fig. 6c) at 75 and 90% PG
compared to that from 45 and 60% PG (P,0.05). K
of DEET was found to be 61.2612.8 from a 10%
DEET solution containing 60% PG, however, K
appeared to decrease as the percentage of PG in
solution increased above 60%, and K for 90% PG
solution was 8.362.7. Therefore, the relative affinity
of DEET for skin as compared to that for vehicle
decreases with an increase in PG concentration
which results in parallel decreases in permeability
and flux (Fig. 6a,b,c).
4.3. Effect of PEG 400 on the percutaneous
permeation of DEET
Fig. 4 shows the permeation profiles of DEET
across human skin from various solutions of PEG
400. The average (n53) permeation profiles from
PEG solutions when compared to pure DEET control
(n58) show that the total amount of DEET permeat-
ing is significantly lower from 60 and 75% PEG 400.
Furthermore, no permeation of DEET was detected
from the 90% PEG 400 during the entire 36 h time
period. Effectively pure PEG 400 (90%) prevents
permeation of DEET. Thus, the total amount of
DEET permeated at the end of 36 h decreased with
an increase in PEG 400 concentration, and it can be
ranked as: pure DEET . . .60% PEG 400.75%
PEG 400.90% PEG 400¯0. The cumulative
amounts permeated from the 60 and 75% PEG 400
solutions were significantly lower by 7-fold (P,
0.05) than that from DEET control.
The permeation parameters, steady state flux (Jss ),
lag time (t L ), diffusion coefficient (D), permeability
(P), and skin / vehicle partition coefficient (K), ob-
D 310 8 P310 4 K 310 21
(cm 2 / h) (cm / h)
1.3360.16 5.8162.70 4.3762.12
0.9960.070 5.9860.98 6.1261.28
1.1360.13 1.4660.33 c 1.2960.21 c
1.2760.27 0.9960.21 c 0.8360.2 7 c
1.7360.95
218 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221

Table 2
Permeation parameters of DEET from PEG 400 solutions through full thickness human skin (46.88 mg / cm 2 ) at 378C a
% PEG 400 Jss TL D 310 8 P310 4 K
(mg / cm 2 h) (h) (cm 2 / h) (cm / h)
60 11.9561.42 b 3.2061.66 b 6.9164.94 1.2060.14 2.3961.38
75 10.4566.12 b 3.9260.80 b 4.3760.90 1.0560.62 2.2860.88
90 No permeation
DEET (n58) 63.20624.52 12.6264.05 1.7360.95
a
Values are presented as mean6S.D. (n53); 100 mg / ml DEET was applied to 0.64 cm 2 of skin.
b
Jss and t L for DEET from 60 and 75% PEG 400 are significantly lower (P,0.05) than Jss and t L from the DEET control.

tained from the permeation profiles are listed in
Table 2. No significant difference (P.0.05) in the
parameters exists between 60% PEG 400 and 75%
PEG 400, however, the flux values for these solu-
tions are significantly lower (6-fold) than the flux
value for pure DEET. Since no permeation was
observed from 90% PEG 400, the flux and skin /
vehicle partition coefficient values for 90% PEG 400
must be significantly lower than those from the 60
and 75% PEG 400 (Fig. 4).
4.4. PEG 400 /PG combination (60:30, v /v)
Although 90% PEG 400 prevented any permeation
of DEET across human skin, the solution was too
viscous. Therefore, a combination of 60% PEG 400
with 30% PG was studied to evaluate if the combina-
tion retained the synergistic effect of reducing
DEET’s permeation since it will be practical for use
due to its lower viscosity. The permeation profile for
DEET across human skin from the PEG 400 / PG
(60:30, v / v) solution, as compared to the pure DEET
control, is shown in Fig. 5. Although some permea-
tion of DEET was observed in contrast to 90% PEG
400, the permeation from the PEG 400 / PG combina-
tion was significantly lower (P,0.05) than that from
pure DEET control. Table 3 lists the permeation
parameters obtained from the profiles according to
the methods presented in Section 3.6. The steady
state flux (Jss ) from the PEG 400 / PG solution was
significantly lower (P,0.05) at 7.2065.54 mg / cm 2
h than the flux of the DEET control (63.20624.52
mg / cm 2 h), and slightly lower than the flux from
90% PG solution also. The lag time and diffusion
coefficients were similar for both, DEET control and
PEG 400 / PG solution. The permeability of DEET
from the PEG 400 / PG solution was lower at
7.2065.50310 25 cm / h than permeability for DEET
from the 90% PG and 60% PEG 400 solutions
previously evaluated suggesting a synergistic effect
on reduction in permeability.
Fig. 6a,b,c compares the average steady state flux,
permeability and skin / vehicle partition coefficient of
DEET, respectively, as a function of vehicle com-
position in terms of solvent levels for ethanol, PG
and PEG 400. For ethanol and PG, the flux and
permeability are at a higher plateau from 30 to 45%
and 45–60% solvent levels, respectively, and then
begin to decline significantly with further increase in
solvent concentration. For PEG 400, the flux is lower
to begin with, however with further increase in PEG
level, flux decreases and no permeation was detected
from the 90% PEG 400 for 36 h. A comparison of
trends in Jss (Fig. 6a) and P (Fig. 6b) with K (Fig.
6c) clearly show that changes in flux and permeabili-
ty are primarily due to changes in the skin / vehicle
PC as expected. Also, this is consistent with the
findings that no significant changes in D were
observed from the three solvent systems studied.
Table 3
Average permeation parameters for DEET across human skin
(46.88 mg / cm 2 ) at 378C from PEG 400 / PG (60:30 v / v) solution a
DEET control PEG 400 / PG (60:30, v / v)
2
Jss (mg / cm h) 63.20624.52 b
7.2065.54
t L (h) 12.6264.05 16.4466.31
D 310 8 (cm 2 / h) 1.7360.95 1.1260.34
P310 5 (cm / h) 7.2065.50
K 8.6861.79
a
The values were obtained from the permeation profile pre-
sented in Fig. 5. All values are presented as mean6S.D., where
n56 for the PEG 400 / PG solutions and n58 for the DEET
control; 100 mg / ml DEET was applied to 0.64 cm 2 of skin.
b
Jss for the DEET is significantly lower (P,0.05) from PEG
400 / PG solution than Jss for DEET control.
 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
The PEG 400 / PG combination exhibits the next
lowest flux value (9-fold lower than the DEET
control) and may be a better vehicle because of its
lower viscosity. As previously observed, the reduc-
tion in permeation of DEET from the PEG 400 / PG
combination was due to the lower skin / vehicle PC
(K) of DEET.
5. Discussion
Suggestions of an incomplete toxicological assess-
ment of N,N-diethyl-m-toluamide [6,16] and the
several reported cases of toxicity and side effects in
humans after using DEET-containing products [1–4],
appear to warrant the development of alternative
formulations of DEET designed to decrease permea-
tion and absorption of DEET [12,13]. Although the
actual mechanism for transport of DEET through
human skin is not known, DEET being a moderately
lipophilic small molecule will permeate through skin
very rapidly. DEET’s facile permeation is confirmed
by a relative large flux of 63.2 mg / cm 2 h which is in
agreement with previous findings [7–11,17]. There-
fore, an attempt was made to reduce its permeation
by formulating DEET in a vehicle which results in
reduced skin / vehicle partition coefficient with little
or no direct effect of the vehicle on skin.
Higuchi was the first one to develop equations
relating thermodynamic activity and skin absorption
from creams and ointments [18]. Barry has described
in detail the drug–vehicle–skin interactions based on
the theoretical model developed by Poulsen [19].
The flux of a permeant through skin should be
constant at a maximum level from saturated binary
solvent system (maximum thermodynamic activity)
if it does not have a direct vehicle effect on the skin.
However, decrease in flux should be observed with
an increase in solubilizer concentration from vehicles
of constant permeant concentration when the de-
crease in skin / vehicle PC is greater than the increase
in solubility due to solubilizer in the vehicle. The
above prediction assumes that the solvent acting as
the solubilizer in the vehicle does not alter the D or
solubility of permeant in skin. If the solvent in-
creases the solubility of permeant in the skin with
increasing concentration in vehicle, flux should
increase initially as solubility increase is not offset
219
by an equivalent decline in skin / vehicle PC. Increas-
ing the solvent concentration beyond that point
results in a decrease in flux due to the skin / vehicle
PC declining more rapidly combined with below
saturation concentration in the vehicle. Therefore,
irrespective of the effect of solvent on the permeant’s
solubility in skin, at high concentrations of solvent,
flux should decline due to reduction in skin / vehicle
PC. Ternary phase diagrams (Fig. 1) of DEET /
water / solvent (ethanol, PG, PEG 400) do indeed
show increasing solubility of DEET with addition of
the solvent. This is also confirmed by an almost
linear to exponential decline in octanol / water PC of
DEET with an increasing concentration of solvent
(Fig. 2). Thus, the decrease in flux (Fig. 6a) and
permeability (Fig. 6b) of DEET with increasing
concentrations of solvent appears to be a direct result
of decrease in skin / vehicle PC (Fig. 6c). Although it
is difficult to know the direct effect of solvents on
DEET’s solubility in skin, from Fig. 6, it appears
that solvent’s effect on DEET’s solubility in water is
greater, particularly at higher concentrations.
Ethanol is a known permeation enhancer and was
shown to increase permeation of DEET compared to
control except at high concentrations and is thus
unsuitable as a vehicle solvent for DEET [11,17].
Propylene glycol (PG) is a semipolar cosolvent and
although permeation of various drugs was enhanced
by PG concentrations up to 50%, reduced permeation
was seen at higher concentrations similar to the
permeation changes observed for DEET [20–23].
Propylene glycol may be classified as a humectant,
and the resultant dehydration effect on skin is
especially noted at high concentrations of PG [19].
The dehydration effect of PG on the skin and the
higher affinity of DEET for PG-water vehicle than
for the skin may contribute to the reduced DEET
permeation. The results of the in vitro percutaneous
permeation studies with high PG concentrations
suggest that PG may be a suitable substitution for
ethanol as a solvent for DEET in commercial
mosquito repellents aimed at reducing permeation.
Polyethylene glycol 400 (PEG 400) was another
potential vehicle that was evaluated for the permea-
tion of DEET. PEG 400 is a fairly viscous, hygro-
scopic liquid that is often found in topical prepara-
tions. PEG 400 has been shown to reduce the
permeation of drugs through human skin [24,25].
220 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
The ternary phase diagram of DEET / water / PEG 400
(Fig. 1) although it showed increase in the aqueous
solubility of DEET, it was not greater than that for
the DEET / water / PG and DEET / water / ethanol sys-
tems, and therefore, DEET solutions containing only
60, 75, and 90% PEG 400 could be evaluated. The
effect of concentration of PEG 400 on the steady
state flux appears to follow a similar parabolic
pattern to that of ethanol and PG [20–22,26–28].
Most strikingly, no permeation of DEET was de-
tected from the 90% PEG 400 solution throughout
the 36 h (Fig. 4). Similar results of negligible flux at
the 90% PEG 400 level have been reported for
oxaprozin and guanabenz solutions [25]. This de-
creased permeation was proposed to be due to the
formation of PEG 400 association complex with
oxaprozin and guanabenz, but this mechanism has
not been proven.
Since 90% PEG 400 eliminates the permeation of
DEET through skin, it may be used as a solvent in
new repellent formulations with reduced potential for
toxicity. However, PEG 400 is a viscous liquid and
may not have the customer appeal of the current
ethanolic mosquito repellents. The high viscosity
may also preclude the use of a spray pump, a
common method of application. In order to reduce
the viscosity, PG (30%) was added to a PEG 400
solution (60%) containing 10% DEET. Both ve-
hicles, PG and PEG 400, effectively reduced the
permeation of DEET, and similar results were ob-
served with the combination (Fig. 5). The two
solvents appear to exhibit a synergistic effect as the
flux from the combination is lower than that from
60% PEG 400 alone or 30% PG alone (Tables 2 and
3). Thus, the 60:30 PEG 400 / PG solvent system
appears to be a better alternative to ethanol than pure
PEG 400, due to its lower viscosity and overall
lower (9-fold reduction) permeation of DEET.
6. Summary
Using a simple model of skin permeation based on
Fick’s laws of diffusion at steady state, an approach
of formulating DEET with significantly reduced
permeation was successfully demonstrated. Semipo-
lar solvents, ethanol, PG and PEG 400, significantly
increased the aqueous solubility of DEET along with
a decreasing o / w PC with increase in concentration
of the solvent. The permeation of DEET across
human skin was reduced significantly from solutions
containing 60% or greater amounts of PG and PEG
400. Permeation of DEET was totally prevented
from 90% PEG 400 solution which was very visc-
ous, however a combination of 60% PEG 400 with
30% PG not only reduced permeation 9-fold but was
suitable as a solvent for vehicle. This study clearly
demonstrates that alternative formulations can be
developed for DEET aimed at reduced permeation
and toxicity unlike the current formulations some of
which contain ethanol which has been shown to
enhance permeation of DEET. A similar approach
can be used for developing formulations of other
industrial and occupational agents to prevent their
skin permeation and penetration when a user may be
exposed.
Acknowledgements
We would like to thank the American Foundation
of Pharmaceutical Education for a predoctoral fel-
lowship to J. Stinecipher, and Jayabharati Vai-
dyanathan and Kimberley Hill for technical help in
the PC experiments.
References
[1] J.C. Veltri, T.G. Osmitiz, D.C. Bradford, B.C. Page, Re-
trospective analysis of calls to poison control centers resulting
from exposure to the insect repellent N,N-diethyl-m-
toluamide (DEET) from 1985 to 1989, Clin. Toxicol. 32
(1994) 1–16.
[2] D.L. Edwards, C.E. Johnson, Insect-repellent-induced toxic
encephalopathy in a child, Clin. Pharmacol. 6 (1987) 496–
498.
[3] J.W. Lipscomb, J.E. Kramer, J.B. Leiken, Seizure following
brief exposure to the insect repellent N,N-diethyl-m-
toluamide, Ann. Emerg. Med. 21 (1992) 315–317.
[4] J.P. deGarbino, A. Laborde, Toxicity of an insect repellent:
N,N-diethyltoluamide, Vet. Hum. Toxicol. 25 (1983) 422–
423.
[5] M.B. Abou-Donia, K.R. Wilmarth, K.F. Jensen, F.W. Oehme,
T.L. Kurt, Neurotoxicity resulting from coexposure to
pyridostigmine bromide, DEET, and permethrin: Implica-
tions of gulf war chemical exposures, J. Toxicol. Environ.
Health 48 (1996) 35–56.
[6] P.J. Robbins, M.G. Cherniack, Review of the biodistribution
 J.S. Ross, J.C. Shah / Journal of Controlled Release 67 (2000) 211 – 221
and toxicity of the insect repellent N,N-diethyl-m-toluamide
(DEET), J. Toxicol. Environ. Health 18 (1986) 503–525.
[7] R.P. Moody, B. Nadeau, I. Chu, In vitro dermal absorption of
N,N-diethyl-m-toluamide (DEET) in rat, guinea pig, and
human skin, In Vitro Toxicol. 8 (1995) 263–275.
[8] R.P. Moody, F. Benoit, D. Riedel, L. Ritter, Dermal absorp-
tion of the insect repellent DEET (N,N-diethyl-m-toluamide)
in rats and monkeys: Effect of anatomical site and multiple
exposure, J. Toxicol. Environ. Health 26 (1989) 137–147.
[9] G.P. Schoenig, R.E. Hartnagel Jr., T.G. Osimitz, S. Llanso,
Absorption, distribution, metabolism, and excretion of N,N-
diethyl-m-toluamide in the rat, Drug Metab. Dispos. 24
(1996) 156–163.
[10] S. Selim, R.E. Hartnagel Jr., T.G. Osimitz, K.L. Gabriel, G.P.
Schoenig, Absorption, metabolism, and excretion of DEET
following dermal application to human volunteers, Fundam.
Appl. Toxicol. 25 (1995) 95–100.
[11] J. Stinecipher, J. Shah, Percutaneous permeation of N,N-
diethyl-m-toluamide (DEET) from commercial mosquito
repellents and the effect of solvent, J. Toxicol. Environ.
Health 52 (1997) 119–136.
[12] W.G. Taylor, T.J. Danielson, R.W. Spooner, L.R. Golsteyn,
Pharmacokinetic assessment of the dermal absorption of
N,N-diethyl-m-toluamide (DEET) in cattle, Drug Metab.
Dispos. 22 (1994) 106–112.
[13] Z.A. Mehr, L.C. Rutledge, E.L. Morales, V.E. Meixsell, D.W.
Korte, Laboratory evaluation of controlled-release insect
repellent formulations, J. Am. Mosq. Control Assoc. 1
(1985) 143–146.
[14] M. Katz, B.J. Poulsen, Absorption of drugs through the skin,
in: B.B. Brodie, J.R. Gillette (Eds.), Handbook of Ex-
perimental Pharmacology, Vol. 28, Springer–Verlag, New
York, 1971, pp. 103–174.
[15] S. Scheuplein, Skin permeation, in: A. Jarrett (Ed.), The
Physiology and Pathophysiology of the Skin, Vol. 5, Aca-
demic Press, New York, 1978, pp. 1669–1730.
[16] United States Environmental Protection Agency, N,N-dieth-
yl-m-toluamide (DEET): Pesticide registration standard, US
Environmental Protection Agency, Office of Pesticide and
Toxic Substances, Washington, DC, December 1980.
[17] J. Stinecipher, J. Shah, Percutaneous permeation of the meta,
221
ortho, and para isomers of N,N-diethyltoluamide (DEET),
Int. J. Pharmaceut. 160 (1) (1998) 31–41.
[18] T. Higuchi, Thermodynamic analysis of percutaneous ab-
sorption process from creams and ointments, J. Cos. Chem.
11 (1960) 85–87.
[19] B.W. Barry, in: Dermatological Formulations – Percutaneous
Absorption, Marcel Dekker, Inc, New York, 1983, pp. 49–
233.
[20] J.W. Hadgraft, Recent progress in the formulation of vehicles
for topical applications, Br. J. Derm. 87 (1972) 386–389.
[21] J. Ostrenga, C. Steinmitz, B. Poulsen, Significance of vehicle
composition. I: Relationship between topical vehicle com-
position, skin penetrability, and clinical efficacy, J. Pharm.
Sci. 60 (1971) 1175–1179.
[22] J. Ostrenga, C. Steinmitz, B. Poulsen, S. Yett, Significance of
vehicle composition. II: Prediction of optimal vehicle com-
position, J. Pharm. Sci. 60 (1971) 1180–1183.
[23] G.L. Flynn, R.W. Smith, Membrane diffusion. III: Influence
of solvent composition and permeant solubility on membrane
transport, J. Pharm. Sci. 61 (1972) 61–66.
[24] S. Kumar, A.W. Malick, N.M. Meltzer, J.D. Mouskountakis,
C.R. Behl, In vitro skin permeation and retention of a
leukotriene antagonist from topical vehicles using hairless
guinea pig model, J. Pharm. Sci. 69 (1992) 631–638.
[25] P.P. Sarpotdar, J.L. Gaskill, R.P. Giannini, Effect of poly-
ethylene glycol 400 on the penetration of drugs through
human cadaver skin in vitro, J. Pharm. Sci. 75 (1986) 26–28.
[26] B. Berner, J.H. Otte, G.C. Mazzenga, R.J. Steffens, C.D.
Ebert, Ethanol:water mutually enhanced transdermal thera-
peutic system II: Skin permeation of ethanol and nitro-
glycerin, J. Pharm. Sci. 78 (1989) 402–407.
[27] T. Kurihara-Bergstrom, K. Knutson, L.J. DeNoble, C.Y.
Goates, Percutaneous absorption enhancement of an ionic
molecule by ethanol–water systems in human skin, Pharm.
Res. 7 (1990) 762–766.
[28] Y. Wada, K. Nakajima, J. Yamazaki, T. Seki, K. Sugibayashi,
Y. Morimota, Influence of composition of l-menthol–etha-
nol–water ternary solvent system on the transdermal delivery
of morphine hydrochloride, Biol. Pharm. Bull. 16 (1993)
600–603.