intemational

! i,'
journal of
ELSEVIER
pharmaceutics
International Journal of Pharmaceutics 137 (1996) 243-248

A comparison of electrical and rheological techniques for the

characterisation of creams
S. T a m b u r i c a, D . Q . M . C r a i g a'*, G . V u l e t a b, J. M i l i c b
"Centre Jbr Materials Science, School of Pharmacy, University of London, 29 39 Brunswick Square, London WCIN IAX UK
blnstitute for Pharmaceutical Technology, Faculty of Pharmacy, Vojvode Stepe 450, 11221 Belgrade, Yugoslavia
Received 22 January 1996; accepted 26 March 1996

Abstract

Two complex semi-solid emulsion systems, one ionic and one non-ionic, have been evaluated using both
conventional (conductivity measurements and continuous flow rheology) and oscillatory (dielectric spectroscopy and
oscillatory rheology) methods. The results of specific conductivity measurements indicated differences in charge
mobility, with the ionic cream showing a considerable higher conductivity (171.0/~S/cm compared to 9.3/~S/cm for
the non-ionic system). Dielectric spectroscopy allowed a more sophisticated electrical analysis to be obtained, and a
discussion is given of how the data relates to the current model for interpreting the low frequency response Flow and
oscillatory results indicated that greater internal structuring, leading to higher elasticity, was achieved for the ionic
system. It is demonstrated that both techniques yield different yet complimentary information on the cream structure.
The study has indicated that the use of dielectric spectroscopy and oscillatory rheometry, in addition to conventional
methods, may lead to a better understanding of the emulsion microstructure.

Keywords: Conductivity; Creams; Dielectric; Rheology

1. Introduction (e.g. Niirnberg and Muckenschnabel, 1982; Ca-

jkovac and Stivic, 1983; Kallioinen et al., 1994)
Semi-solid emulsion systems (creams) are have been widely used to characterise cream
widely used as a means of both altering the hydra- bases, although the interpretation of data may not
tion state of the skin and delivering drugs via the always be straightforward.
topical route. A considerable problem associated An alternative, but related electrical approach,
with the formulation and manufacture of creams is dielectric analysis. This technique involves the
has been the establishment of reliable techniques application of an alternating electric field to a
for their characterisation, largely because of the sample over a range of frequencies and the mea-
complexity of their physical structure. Continuous surement of the sample response. This response
flow and single-point conductivity measurements will comprise an ifi-phase and out-of-phase com-
ponent with respect to the fluctuations in field
* Corresponding author. direction, hence two parameters m a y be measured

0378-5173/96/$1500 1996 Elsevier Science B.V. All rights reserved

PH S0378-5173 (96)04 5 28 -0
244 S. Tamburic et al. / International Journal o f Pharmaceutics 137 (1996) 243 248

at any single frequency (Craig, 1995). These two of the product. Hence, the objective of the present
components are the capacitance (C) and dielectric study was to compare two electrical and two
loss (G/og, where G is the conductance and ~o is rheological characterisation techniques as meth-
the angular frequency). Previous studies (Rowe et ods for evaluating the microstructure of the
al., 1988; Goggin et al., 1994) have indicated that creams.
the technique may be highly useful in the charac-
terisation of creams, as not only may all the
information given by conductivity measurements 2. Materials and methods
be obtained, but in addition information on the
microstructure of the cream is given from both All the components used in the formulations
the absolute values of capacitance and dielectric (Table 1) were of pharmacopoeial quality. The
loss and also from the shape of the spectra. It creams were prepared by heating the water and oil
should be emphasised that while conductivity phases in separate containers up to 70-75C,
measurements are usually single point determina- mixing them using a high-speed agitation, and
tions, the equipment used in fact operates at a set cooling the resulting emulsion down to room tem-
frequency, hence these values are essentially single perature using a slow-speed agitation.
frequency dielectric loss measurements. The as- Conductivity measurements were performed
sumption is made that the conductivity is inde-
with a CDM 2 meter (Radiometer, Denmark),
pendent of frequency under the conditions of
fitted with a CDC 104 electrode, using a fre-
measurement.
quency of 50 Hz at room temperature (25C).
In addition to dielectric measurements, oscilla-
Dielectric measurements were performed using a
tory rheology may also be used to characterise
low frequency Dielectric Spectrometer (Dielectric
creams. This technique works on a similar princi-
Instrumentation Ltd., UK) with parallel platinum
ple to dielectric analysis in that an oscillating
electrodes (area approximately 0.5 cm 2, separation
mechanical (as opposed to electrical) signal is
applied to a sample over a range of frequencies. distance 1 ram) and a voltage of 0.1 Vrms. Fre-
Consequently, the two components obtained yield quency sweeps between 10 -2 and 104 Hz at 25C
information on the solid-like (G', storage mod- were obtained for each emulsion. At least three
ulus) and liquid-like (G", loss modulus) behaviour measurements were performed for each cream
of the sample. using both techniques, with a coefficient of varia-
To test the concept of using the two dynamic tion of < 5% found.
techniques in conjunction for the characterisation All rheological measurements were performed
of creams, we have chosen two complex semi- using a Carrimed controlled-stress rheometer (TA
solid emulsions (Table 1). Both contain a multi-
component emulsifier, consisting of: a lipophilic Table 1
The formulation of two amphiphilic creams
surfactant (glycerol monostearate), a hydrophilic
surfactant (either ionic sodium lauryl sulphate or Component Ionic cream Non-ionic
non-ionic Polysorbate 60) and an auxiliary (% w/w) cream
emulsifier (a homologue admixture of two fatty (% w/w)
alcohols, cetyl and stearyl). The two creams exem-
Cetostearyl alcohol 9.0 9.0
plify a special group of semi-solid emulsions, Glycerol monostearate 3.0 3.0
namely amphiphilic creams, whose structure com- Sodium lauryl sulphate 3.0
prises a crystalline gel network of complex Polysorbate 60 --- 8.0
emulsifiers, with oil and water phases distributed Liquid paraffin 12.0 15.0
through that network. It is important to develop Isopropyl myristate 3.0 3.0
White soft paraffin 25.0 12.0
means of characterising the level of liquid crystal Glycerol 5.0 5.0
structuring within the cream, as this will in turn Purified water 40.0 45.0
have a profound effect on the physical properties
 S. Tamburic et al. / International Journal of Pharmaceutics 137 (1996) 243-248 245

Table 2
Dielectric parameters (capacitance C, dielectric loss G/co, low frequency logarithm capacitance slope) and specificconductivity(k)
for the two creams

Cream C at 10 - t Hz C at 103 Hz G/09 at 10 -1 Hz G/~o at 103 Hz Log capacitance k

(10 5 F) 0 0 -8 F) (10 5 F) 0 0 -7 F) slope (0 to - 1 . 5 (#S/cm)
log Hz)

Ionic 2.360 1.420 1.028 2.228 0.1553 171.0

Non-ionic 1.585 0.067 1.574 0.177 0.4056 9.3

Instruments Ltd., UK) at 25C. Flow experiments and

involved stress sweeps from 0-100 Pa over 1 min,
(72(0. (R 2 . Cl)
followed by decreasing the stress to 0 Pa over 1 Cco"= G(co)/~o = 1 + oJ 2 . ( R 2 . C1)2 (2)
min. Oscillatory sweeps were performed from 1-
10 Hz using a torque of 489 /tNm over 15 min. where C',o and C~ are the real and imaginary
Repeat measurements showed a coefficient of components of the response at a frequency ~o and
variation of < 10%. R is the resistance (l/G) of the slab, while the
subscripts 1 and 2 refer to the bulk and electrode
layers respectively. These equations predict that at
3. Results and discussion low and high frequencies, two responses will be
seen corresponding to the electrode and bulk lay-
Conductivity measurements gave values of 9.3 ers respectively. These separate layers yield infor-
pS/cm for the non-ionic cream and 171.0 pS/cm mation on the structure of the system and may be
for the ionic system (Table 2), hence the ionic used as a diagnostic tool. In particular, the pres-
cream yielded a far higher specific conductivity ence of an electrical 'blocking' layer on the elec-
(k). While these k values may be due to the trodes is indicated by the slope of the low
differences in the degree of structuring and/or the frequency (below 1 Hz) capacitance, with an effec-
type of emulsion, they may also simply reflect the tive blocking layer resulting in a flat capacitance
higher level of free ions within the ionic system, slope. Similarly, the absolute values of the capaci-
hence it is not possible to unequivocally interpret
such a single point determination. -3.0
The dielectric responses of the two creams are
oo~ -4.0
shown in Fig. 1. The ionic system again yielded a
higher response, but in this case information is -50
o
given over a range of frequencies, providing spec- -6.0
tra which may be interpreted in terms of the
-7.0
emulsion microstructure (Rowe et al., 1988; Gog-
gin et al., 1994). -8.0
The model currently used is based on the work o
o~ -9.0
of Hill and Pickup (1985), which proposes that a
_d
number of systems may be considered to comprise -10.0
two layers, a bulk layer and an electrode layer 05 1, 2, ,5
which is composed of components of the materi- Log frequency (Log Hz)
als under study which have been adsorbed onto
Fig. l. Dielectric spectra of ionic and non-ionic creams.
the electrodes. The response may be described by Capacitance of ionic cream, capacitance of non-ionic
C1 cream, dielectric loss of ionic cream, dielectric loss of
Co)' = 1 -F o92 .(R 2. C1)2 (1) non-ionic cream. Dotted line corresponds to 50 Hz, solid line
shows the linear conductivity region.
246 S, Tamburic et al. / International Journal o f Pharmaceutic~ 137 (1996) 243-248

tance indicate the thickness and permittivity of 120

this layer, according to
_ 100
C = Ae/d (3)
where A is the area of the electrodes, is the
permittivity of the layer and d is the layer thick-
ness. At high frequencies, the response will be
~/ /j *j'>"
dominated by the bulk conductivity G, hence
when plotted logarithmically against frequency,
2o
G/~o will have a linear slope of - 1 in this region,
indicating that G is frequency independent (as is 0
assumed when using conductivity meters). -,0 '0 ,'o 2'0 3; ,o do 6; 70
Examination of Table 2 shows that the conduc- Shear rate (l/s)
tivity and dielectric measurements are in agree-
ment in that higher values for the high frequency Fig. 2. Flow curves for ionic and non-ionic creams. ~ Ionic
cream, non-ionic cream.
dielectric loss and conductivity ~c were found for
the ionic systems. As both of these parameters
reflect charge movement through the system, such the emulsion, since previous studies (Goggin et
corroboration is to be expected. However, exami- al., 1994) have demonstrated that for the same
nation of Fig. l indicates that at 50 Hz (the formulation, greater structuring leads to a lower
frequency used for the conductivity measure- conductivity.
ments) the dielectric loss is in the 'crossover' Further information on the structure of the two
region between the responses of the two layers, semisolid emulsion systems was obtained using
hence the conductivity is not independent of fre- rheological studies. Linear rheological tests (stress
quency and is a reflection of both the bulk and sweeps) revealed significantly different flow curves
electrode layers. This does not necessarily mean for the two creams (Fig. 2). The ionic system
that the conductivity values are unreliable, as the showed a much higher yield value than the non-
cell systems used for the dielectric and conductiv- ionic one (Table 3), indicating a larger resistance
ity measurements are different (which will alter to an external force before the system starts flow-
the response), but these results do indicate that ing, and thus a greater degree of structuring
care must be taken in assuming that the conduc- (Barry, 1974). The overall resistance to flow, ex-
tivity is frequency independent. A similar phe- pressed by viscosity, was also higher for the ionic
nomenon has been observed for surfactant system, accompanied by a 20-fold smaller hys-
solutions (Craig and McDonald, 1995) where it teresis loop, which is commonly used as a mea-
was demonstrated that the linear conductivity re- sure of the level of thixotropy and/or structural
gion may not, in some cases, be seen below the breakdown within the system (Table 3, Fig. 2).
kHz region. Oscillatory studies provide information on both
Table 2 also gives values for the slope of the elastic and viscous components of complex vis-
low frequency capacitance. The value for the ionic coelastic rheological behaviour of semi-solids
cream is lower than for the non-ionic, indicating a (Ferry, 1970). The overall level of viscoelasticity is
more rigid 'blocking' layer on the electrodes. given by the tan ~ value, which is the ratio
Rowe et al. (1988) have suggested that the re- between the loss and storage moduli (G"/G'). The
sponse in this region may be a reflection of the values of phase angle 6 (ranging from 0 for an
degree of structuring within gel and emulsion ideal elastic solid to 90 for an ideal viscous
systems, hence the results presented here indicate liquid) are given over the whole frequency range
that the ionic system is more structured. If so, this in the Fig. 3, with representative tan ~ values
in turn indicates that the higher conductivity val- given in Table 3. The values indicate that the
ues are simply a reflection of the ionic nature of non-ionic creams show greater viscous, as op-
 S. Tamburic et al. / International Journal of Pharmaceutics 137 (1996) 243-248 247

Table 3
Flow parameters (viscosity r/, hysteresis area H and yield value Zv) and oscillatory parameters at 1.8 Hz (storage modulus G', loss
modulus G" and tangent of the phase angle 3) for the two creams

Cream q at 100 Pa (Pa s) H (Pa/s) r v (Pa) G' (Pa) G" (Pa) Tan 6

Ionic 13.85 93.0 41.24 2961 845 0.2842

Non-ionic 1.71 2114.0 12.7 f 4929 2547 0.5154

posed to elastic behaviour compared to the ionic properties of creams, oscillatory measurements
systems, which in turn suggest that the ionic allow direct quantitative assessment of the vis-
creams have greater internal structuring (Eccle- coelastic properties of these systems. Overall, both
ston, 1984). the dielectric and rheological studies indicate that,
in this particular case, the ionic cream shows a
greater level of structuring.
4. Conclusions

This study has focused on the assessment of the Acknowledgements

methods available for the characterisation of
semi-solid emulsions. While conductivity and vis- This work was partially supported by the Well-
cosity measurements are undoubtedly of use, con- come Trust (grant number 045086/Z/93/B to Dr.
siderably more sophisticated information may be Tamburic).
obtained using oscillatory dielectric and rheologi-
cal techniques. In particular, the use of dielectric
analysis allows an insight into the degree of struc- References
turing within the cream and also indicates the
Barry, B.W., Rheology of pharmaceutical and cosmetic
frequency range over which d.c. conductivity semisolids. In Bean, H.S., Beckett, A.H. and Carless, J.E.
dominates the charge transport process, which (Eds.), Advances in Pharmaceutical Science, Academic
may not necessarily coincide with the frequency at Press, London 1974, pp. 1-72.
which commonly used conductivity equipment Cajkovac, M. and Stivic, I., Inversion properties of ambiphilic
emulsions, Acta Pharm. Jugosl. 33 (1983) 29-42.
takes measurements. Similarly, while flow mea- Craig, D.Q.M., Dielectric analysis of pharmaceutical systems,
surements allow useful insights into rheological Taylor and Francis, Basingstoke, 1995.
Craig, D.Q.M. and McDonald, C., Low frequency dielectric
40 effects in the measurement of fl-cyclodextrin interactions
with sodium dodecyl sulphate, J. Phys. Chem. 99 (1995)
35 \ 5413 5417.
Eccleston, M.E., Properties of fatty alcohol mixed emulsifiers
E *
and emulsifying waxes. In Florence, A.T. (Ed.), Materials
used in Pharmaceutical Formulation, Critical Reports on
Applied Chemistry, Vol.6, Blakewell Scientific Publica-
tions, Oxford, 1984, pp. 125-156.
~ eo Ferry, J.D., Viscoelastic properties of polymers, 2 nJ edn., Wi-
ley, New York, 1970.
Goggin, P., Craig, D.Q.M. and Gregory, D., Emulsion pro-
15
duction monitoring by dielectric spectroscopy, Pharm.
Res., 11 (1994) S-244.
10 i i i i i i
Hill, R.M. and Pickup, C., Barrier effects in dispersive media,
-0.2 -0.0 0.2 0.4 0.6 0.8 1.0 1.2
J. Mater. Sci., 20 (1985) 4431-4444.
Log frequency [log Hz) Kallioinen, S., Helenius, K. and Yliruusi, J., Physical charac-
teristics of some emulsion creams in which the length of
Fig. 3. Variation of phase angle with frequency for two ethylene oxide chain of emulsifier varies, Pharmazie, 49
creams. ~ Ionic cream, non-ionic cream. (1994) 750 756.
248 S. Tamburic et al. / International Journal o f Pharmaceutics 137 (I996) 243 248

Niirnberg, E. and Muckenschnabel, R., Hydrophile und am- Rowe, R.C., Dissado, L.A., Zaidi, S.H. and Hill, R.M., The
biphile Cremesysteme, Dtsch. Apoth. Ztg. 41 (1982) 2093- characterisation of heterogeneous gels by a dielectric tech-
2106. nique: 2. Formulation and structural considerations, J.
Coll. Int. Sci., 122 (1988) 354-366.