Synthesis and solution properties of gemini surfactants containing oleyl chains

Article abstract of DOI:10.1039/b507910a

Gemini surfactants 18:1-s-18:1, where s = 2, 3, and 6 methylene groups and 18:1 refers to oleyl carbon chains, have been synthesized, characterized and a number of micelle solution properties measured by using electrical conductance, fluorescence probe em

Full text of DOI:10.1039/b507910a

View Online / Journal Homepage / Table of Contents for this issue
R E S E A R C H P A P E R
Synthesis and solution properties of gemini surfactants containing
oleyl chainsw
Xingfu Li,a Shawn D. Wettig,b Chuanzhong Wang,c Marianna Foldvarib and
Ronald E. Verrall*c
a Department of Chemistry, McGill University, 801 Sherbrooke St. W., Montreal,
Canada QC H3A 2K6
b College of Pharmacy and Nutrition, University of Saskatchewan, 110 Science Place,
Saskatoon, Canada SK S7N 5C9
c Department of Chemistry, University of Saskatchewan, 110 Science Place, Saskatoon,
Canada SK S7N 5C9. E-mail: verrallr@sask.usask.ca
Received 6th June 2005, Accepted 19th July 2005
First published as an Advance Article on the web 26th July 2005
Gemini surfactants 18:1-s-18:1, where s ¼ 2, 3, and 6 methylene groups and 18:1 refers to oleyl carbon chains,
have been synthesized, characterized and a number of micelle solution properties measured by using electrical
conductance, fluorescence probe emission, light scattering (DLS), surface tension and isothermal titration
calorimetry (ITC) methods at 25 1C. The cmc values of 18:1-2-18:1, 18:1-3-18:1, and 18:1-6-18:1 were found to be
26.9, 23.4, and 18.0 mM, respectively, using the electrical conductance method. Surface tension results suggest
that in 0.01 N NaCl solutions, the s ¼ 2 and 3 members of the series form multilayer rather than monolayer
structures, while the s ¼ 6 homologue adopts a close-packed arrangement. This is consistent with DLS and EM
measurements which show vesicle formation for the s ¼ 2 and 3 compounds, and micelle formation for the s ¼ 6
compound. The enthalpies of micellization (DH1M) are more exothermic for the 18:1-s-18:1 series of surfactants,
than for the 12-s-12 series. The differences are rationalized in terms of steric and configurational contributions to
DH1M arising from difficulties associated with packing of the bulky cis-9-octadecene tails.
monoquaternary ammonium surfactants with one or two alkyl
chains.9 The Krafft temperature for gemini surfactants posses-
Introduction
Surfactants containing two hydrophilic and two hydrophobic
sing long, saturated carbon chains is expected to be high,
groups (so-called ‘‘gemini’’ surfactants) have been the focus of
considerable research interest since the early 1990s. Several
chains is a means of reducing the Krafft temperature and, also,
439 1C for m Z 16. The introduction of unsaturated carbon
comprehensive reviews have documented the solution beha-
provides an opportunity to explore the morphological nature
of the aggregates formed using gemini compounds having
viour of gemini compounds.1–3 Their unique structure, con-
sisting of two typical single-head, single-tail surfactant
unsaturated chains.
molecules linked chemically at or near the head group, show
The use of cationic amphiphiles for mediating DNA trans-
a rich array of aggregate morphologies and solution properties
fection has become widespread since a formulation consisting
that are dependent upon the nature and size of the linking
of dimethyldioctadecylammonium bromide and dioleoyl-phos-
phatidylethanolamine (DOPE) was shown to be an effective
spacer.1,4–6
DNA transfection agent.10 The use of transfection agents is
The aggregation properties of the gemini surfactant series
CsH2s-a,o-bis(CmH2m11N1(CH3)2Brꢀ), referred to as m-s-m,
have been measured as a function of the number of s and m
necessary to overcome difficulties associated with transferring
high molecular weight, negatively charged species across the
(predominantly) negatively charged cell membrane. This is
generally accomplished through encapsulation of DNA, for
example, in cationic liposomes. The gemini surfactants are
good candidates to act as transfection agents because they
carbon atoms of the alkanediyl spacer and of the alkyl chain,
respectively.6,7 The cmc of the surfactant series with m ¼ 12 or
16 has been found to go through a maximum of small
amplitude for s ¼ 5 ꢁ 1, irrespective of the length of the alkyl
side chain. The cmc behaviour has been interpreted in terms of
conformational changes of the surfactant spacer group.6 The
effect, and they possess structural characteristics suitable for
are effective at low concentrations, thus diminishing any toxic
minimum head group area at the air–water interface, ao, shows
a maximum for s ¼ 10–12.8 This behaviour was interpreted as
transfection of DNA. In particular, through simple synthetic
modifications, they afford opportunities to tailor their struc-
ture to improve transfection efficiencies that are not available
with traditional surfactant or lipid systems.
reflecting a relocation of the spacer at the interface, from being
in contact with water to the air-side of the interface at s Z 10,
when the spacer becomes sufficiently hydrophobic. For con-
stant s, log cmc decreases linearly with increasing m with a
Gemini surfactants with oleyl chains were reported to med-
iate transfection of DNA in mammalian cell lines in the
absence of a colipid.11 In more recent studies, dermal delivery
slope similar to that for monoquaternary ammonium surfac-
tants. The changes of the degree of counterion dissociation, a,
with s and m are qualitatively similar to those found for
of plasmid DNA has been achieved using gemini surfactants
with a colipid.12 Also, the elasticity of lipid bilayers has been
shown to depend on the degree of unsaturation of lipid-like
materials.13 For these reasons, our recent transfection studies
using synthetic, nonviral transfection agents have focused on
w Electronic supplementary information (ESI) available: NMR and
elemental analysis data. See http://dx.doi.org/10.1039/b507910a
3172
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Online
water under a nitrogen atmosphere, allowing for the solution
to equilibrate.
Surface tension was determined using a Kruss (Model
¨
K10T) tensiometer, applying the du Nuoy ring technique.
Surface tension values (g) were corrected by using the method
of Harkins and Jordan.15 Experimental temperatures were
maintained at 25 ꢁ 0.05 1C using a Haake (Model F3)
circulating water bath. Solutions of 18:1-s-18:1 were prepared
in a 0.01 N NaCl solution, stirred, and equilibrated overnight.
Some samples took long periods for equilibrium to be reached.
This was a random occurrence.
Measurements of pyrene fluorescence intensities were carried
out using a Hitachi Fꢀ2500 fluorescence spectrophotometer
with slit widths of 1 mm at an excitation wavelength of 340
nm.16,17 The wavelength increment was 0.5 nm and the inte-
gration time was 0.5 s. Solutions were prepared by mixing a
concentrated surfactant with a known volume of solvent
containing pyrene such that the concentration of pyrene was
less than 1 ꢂ 10ꢀ6 M in the final solution. Spectra were
collected for each solution and there was no evidence of
exciplex formation.
Scheme 1 General structure of the 18:1-s-18:1 gemini surfactants.
the use of gemini compounds possessing long, unsaturated
carbon chains. In order to better understand their efficiency in
transfection, and because there has not been a comprehensive
study of the physicochemical, thermodynamic, and self-aggre-
gating properties of gemini surfactants having oleyl chains, we
undertook the study reported herein. It is shown that the s ¼ 2
and 3 members of the 18:1-s-18:1 series (Scheme 1) show
aggregation features that are different from gemini series
possessing saturated carbon chains.
Materials and methods
Materials
The hydrodynamic radii of micelles of gemini surfactants
were determined by using a Protein Solution Inc. Dyna-Pro
800 dynamic light scattering instrument.18 The data acquisition
time was set at 5 s and the minimum S/N ratio (ratio of
intensity of light scattered from the sample to the light scat-
tered from the solvent, alone) was set between 2–2.5 : 1.
Solutions were filtered prior to measurement by using a 0.02
mm filter (Whatman). The autocorrelation functions were
analyzed using the Protein Solutions Inc. Dynamics (v. 5.26)
software, which analyzes the correlation function for the
scattering data using a modified regularization method.19,20
The DLS output provides an estimate of the translational
diffusion coefficient, Dt, of the particles in solution. Assuming
Brownian motion and a spherical model, this coefficient is
converted to the hydrodynamic radius, Rh, of the particles
from the Stokes–Einstein equation:
The alkanediyl-a,o-bis(dimethyloleylammonium bromide),
CsH2s–a,o-(CmH2m11N1(CH3)2Brꢀ))2 (s ¼ 2, 3 and 6, abbre-
viated 18:1-2-18:1, 18:1-3-18:1, and 18:1-6-18:1, respectively),
surfactants were prepared via a bis-Menshutkin reaction of the
appropriate tetramethylalkylenediamine with a long-chain,
alkene bromide.
Oleyl alcohol was converted into the corresponding bromide
according to a procedure previously described.14 To a solution
of 5 g (18.62 mmol) of oleyl alcohol (cis-9-octadecen-1-ol,
Sigma, 99% purity) and 2.33 g (29.78 mmol) of pyridine
dissolved in acetonitrile (36 mL) was added, at 0 1C over the
course of 10 min, 10.22 g, (24.18 mmol) of solid Ph3PBr2. After
stirring at room temperature for 1 h (disappearance of alcohol
was checked by TLC), the reaction mixture was filtered
through a short pad of silica gel and rinsed with an ether–
pentane mixture (1 : 10, 200 mL). Upon further work up, (5.0 g,
15.09 mmol, 81% yield) of pure oleyl bromide was obtained.
The 18:1-s-18:1 surfactants were synthesized by reaction of 1
equivalent of N,N,N0,N0-tetramethylalkylenediamine (Aldrich)
(3.02 mmol) with 2.1 equivalents of 1-bromo-cis-9-octadecene
(6.34 mmol) in anhydrous acetonitrile under reflux for 72 h.
Upon cooling, a white solid was recovered by filtration and it
was recrystallized several times from a mixture of acetonitrile
and ethyl acetate. The purified gemini salts were normally
stored at ꢀ20 1C. The structures of all synthesized compounds
Rh ¼ kbT/6pZDt
(1)
where kb is the Boltzmann constant, T is the absolute tem-
perature, and Z is the solvent viscosity. The reported hydro-
dynamic radii were calculated as previously reported.21
Electron micrographs were obtained with a Philips Model
C10 transmission electron microscope operating at 60 keV. A
small drop of about 10 mL of a 3.00 mM surfactant solution
was placed on a carbon coated and glow discharged 200 mesh
copper grid. Negative staining was performed using 1% w/v
phosphotungstic acid. Enthalpies were measured in duplicate,
1
were confirmed by H NMR spectroscopy (CDCl3, 500 MHz
Bruker) and the purity verified by surface tension measure-
ments and elemental analysis (see supplementary informa-
tion).w The gemini compounds and oleyl bromide were found
to be pure, there being no evidence of starting materials or
monoquaternary intermediate in the proton NMR spectra or
analytical TLC of the purified gemini products. Stock solutions
containing 0.50 wt% or less of surfactant were prepared by
weight using analytical grade water obtained from a Millipore
Milli-Q filtration system.
Methods
Critical micelle concentrations were determined using specific
conductivity, surface tension, light scattering, fluorescence
emission, and isothermal titration calorimetry (ITC) techni-
ques. Specific conductivities were measured using a Wayne–
Kerr precision component analyzer (Model 6425) operating at
1.5 kHz, with a Tacussel electrode having a cell constant of
1.23 cmꢀ1 7
. Experimental temperatures were maintained at
25 ꢁ 0.05 1C using a Haake (Model F3) circulating water
bath. The specific conductance was measured after each addi-
tion of an aqueous solution of concentrated surfactant into
Fig. 1 Variation of the cmc with alkyl chain carbon number (m) at
25 1C for the gemini surfactants. Data for the 18-8-18 gemini surfactant
is for saturated C18 tails (ref. 22; data for 18:1-2-18:1, 18:1-3-18:1, and
18:1-6-18:1 gemini surfactants correspond to unsaturated C18 tails.
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
3173

View Online
Table 1 Critical micelle concentrations for the 18:1-s-18:1 gemini surfactants
Surfactant
STa/mM
SCb/mM
a
FLc/mM
LSd/mM
ITCe/mM
18:1-2-18:1
18:1-3-18:1
18:1-6-18:1
17.3 ꢁ 0.4
15.1 ꢁ 0.4
11.0 ꢁ 0.4
26.9 ꢁ 0.4
23.4 ꢁ 0.4
18.0 ꢁ 0.4
a
0.61 ꢁ 0.08
0.42 ꢁ 0.06
0.66 ꢁ 0.08
b
24.7 ꢁ 0.8
22.0 ꢁ 0.8
20.5 ꢁ 0.8
c
19 ꢁ 1
18 ꢁ 1
19 ꢁ 1
21.0 ꢁ 0.4
27.0 ꢁ 0.7
29 ꢁ 2
d
e
Critical micelle concentrations determined using Surface tension. Specific conductance. Fluorescence. Light scattering. Calorimetry.
as previously described,18 using a Calorimetry Sciences Cor-
poration model 4200 isothermal titration calorimeter (ITC).
with s, and a tendency to level off at s Z 12. The values of a
obtained for the m ¼ 18 : 1 series are consistent with the
increasing trend in a previously found,6,7,27 as m increased
from 12 to 16 (for constant s). A high degree of micelle
ionization predicts that head group repulsion would play an
important role in determining the structure of the aggregates
formed in the self-assembly process and favor traditional
micelles. However, light scattering results do not appear to
support this expectation for the s ¼ 2 and s ¼ 3 homologues of
the m ¼ 18 : 1 series and reasons for this will be discussed later.
Additional techniques were used to determine the cmc of the
18:1-s-18:1 series to compare with the values obtained from
conductance measurements (see Table 1). Both light scatter-
ing18 and spectrofluorimetry16,17 with pyrene as a probe, have
been used successfully to determine the cmc of gemini surfac-
tants. Fig. 3 shows the variation of the I1/I3 vibronic intensity
ratio of pyrene as a function of concentration of 18:1-s-18:1 in
Results and discussion
Fig. 1 shows the variation of the cmc for gemini surfactants
(m-s-m) with shorter spacers (s) as a function of m (number of
carbon atoms in saturated carbon chains). A linear decrease in
the logarithm of the cmc with an increase in m is observed for
m r 16, as found by Zana for the m-2-m surfactants with
8 r m r 16, and for which premicelle association is very weak
or absent.17 However, for m ¼ 18 (saturated, Fig. 1), the cmc
value is greater than would be predicted by the linear trend for
8 r m r 16, and the conductivity behaviour has been shown
to be consistent with premicelle aggregates.22 The cmc values
(Table 1) obtained from specific conductance for the 18:1 series
(s ¼ 2, 3, and 6) show similar behaviour to that of m ¼ 18
(saturated). While a plot similar to that shown for saturated
compounds can not be presented for gemini series having
unsaturated carbon chains because of lack of data, the fact
that the cmc values for the 18:1-s-18:1 series (Table 1) lie above
the linear correlation in Fig. 1 suggests the possibility that
premicelle aggregates may exist in these systems, as well.
Some support for this interpretation can be observed (Fig. 2)
in the variation of the specific conductivity, k, vs. surfactant
concentration, C, and of the molar conductivity, L ¼ (k ꢀ k0)/
C, vs. C0.5 (k0 ¼ specific conductivity of water) of these
systems. It was shown that curvature towards the k axis22 in
k vs. C plots indicates the presence of premicelle aggregates.
Also, a L vs. C0.5 plot will increase with concentration and
reach a maximum near the cmc, a behaviour contrary to that of
non-associating systems. The presence of a maximum has been
reported in the L vs. C0.5 plots for some dialkyldimethylam-
monium chlorides23 and 12-s-12,22 with s ¼ 12, 14, 20, and
attributed to premicelle association. Also, studies of gemini
surfactants having rigid hydrophobic spacers, including the
18-s-18 series, have shown evidence of premicelle aggregates
when m Z 16.24–26
Contrary to results for gemini compounds in which m ¼ 12
or 16 and the spacers are short, the gemini surfactant with m ¼
18 : 1 and s ¼ 6 shows a strong upward curvature at C o cmc
in the plot of k vs. C (Fig. 2a, inset). In addition, the plot of L
versus C0.5 (Fig. 2b) shows a maximum at a concentration close
to that corresponding to the inflection in the plots of k vs. C.
This is behaviour consistent with formation of premicelle
aggregates. While the upward curvature in the k vs. C plots
for s ¼ 2 and 3 (Fig. 2a) are less evident, the maxima in the
plots of L versus C0.5 plots (Fig. 2b) for these compounds
suggest the presence of premicelle aggregates in these systems.
In this work the apparent cmc values of the surfactants have
been taken as the concentration corresponding to the intercept
of the two linear segments of the k vs. C plots. The cmc values
of 18:1-2-18:1, 18:1-3-18:1 and 18:1-6-18:1 were found to be
26.9, 23.4, and 18.0 mM, respectively (Table 1), decreasing with
an increase in the spacer chain length. This is different from the
behaviour for the 12-s-12 series where the plot of cmc vs. spacer
s shows a maximum at s E 4–6 and decreases beyond s 46.6,7
Also, the degree of micelle ionization (a) decreases from 0.61
(s ¼ 2) to 0.42 (s ¼ 3), and then increases to 0.66 (s ¼ 6)
whereas for the 12-s-12 series, a shows a sigmoidal increase
Fig. 2 (a) Specific conductance as a function of surfactant concentra-
tion for the 18:1-2-18:1 (&), 18:1-3-18:1 (J), and 18:1-6-18:1 (n). The
apparent cmc’s were determined from the intercept of the two linear
regions of the curve. (b) Equivalent conductivity curves for the 18:1-2-
18:1 (&), 18:1-3-18:1 (J), and 18:1-6-18:1 (n). The apparent cmc’s
(from Fig. 2a) are indicated by arrows.
3174
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Online
Fig. 3 Pyrene vibronic intensity ratios as a function of surfactant
concentration at 25 1C for 18:1-2-18:1 (&), 18:1-3-18:1 (J), and
18:1-6-18:1 (n).
Fig. 4 Apparent hydrodynamic radii, RH, as a function of 18:1-s-18:1
concentration at 25 1C; 18:1-2-18:1 (&), 18:1-3-18:1 (J), and 18:1-6-
18:1 (n). cmc values were obtained from a Boltzmann fit of the data.
aqueous solutions, at 25 1C. In the I1/I3 plots the cmc is usually
taken as the concentration corresponding to the intercept of
the lines drawn by extrapolation of the rapidly varying part of
the plot and of the nearly horizontal part of the plot at high
concentration.28 Using this procedure, cmc values of 24.7, 22.0,
and 20.5 mM were obtained for 18:1-2-18:1, 18:1-3-18:1, and
18:1-6-18:1, respectively, in good agreement with values deter-
mined using conductance methods.
The I1/I3 vs. surfactant concentration plots (Fig. 3) for the
18:1-s-18:1 series show features that are not observed in the
case of the 12-s-12 series. Although the values of the I1/I3 ratios
in the plateau regions for the s ¼ 3 and 6 members are similar
in magnitude to those reported for the 12-s-12 series,27 there is
a less abrupt decrease in the magnitude of the I1/I3 ratios in the
cmc region. This indicates a less cooperative self-aggregation
process and that the probe is experiencing a variety of envir-
onments between totally aqueous, at low surfactant concentra-
tion, to micelle-like environment at high concentrations.29 In
the case of the s ¼ 2 homologue, the lower magnitude of the I1/
I3 ratio in the premicelle concentration region could be due to
pyrene either inducing small aggregates to form around it or
locating within existing, premicelle aggregates. While the re-
sults of these two techniques do not provide unambiguous
evidence for the existence of premicelle aggregates in all three
compounds, they are suggestive of their presence.
Dynamic light scattering studies were carried out to provide
additional estimates of the cmc and, especially, to obtain an
estimate of their hydrodynamic radius. Data analyses show
(Fig. 4) the presence of a scattering species below the cmc with
average value of Rh E 1–1.5 nm for all systems studied. Within
the estimated error, this value would be consistent with the
presence of submicelle aggregates. However, above the cmc the
s ¼ 6 homologue differs markedly from the other two members
of the series. Whereas for s ¼ 2 and 3 a large scattering species
having Rh E 35 nm is observed, for s ¼ 6 a mass weighted
aggregate having Rh ¼ 2.5 nm is observed. Within the esti-
mated error, it is difficult to determine whether any micelle
growth occurs with increasing concentration for these systems.
Using a curve fitting function, the cmc values were estimated to
be 19 mM, 18 mM, and 19 mM for the 18:1-2-18:1, 18:1-3-18:1,
and 18:1-6-18:1, respectively.
viously.18 This value is consistent with mean aggregation
numbers determined for the m ¼ 12 series.30,31 It appears from
this study that the s ¼ 2 and 3 members may form vesicle-like
aggregates spontaneously when self-aggregation occurs,
whereas the s ¼ 6 homologue forms traditional sphere-like
aggregates. Hydrocarbon chain unsaturation is known to have
a profound effect on the properties of phospholipids (i.e.
saturated vs. unsaturated).32 It effectively decouples parts of
the hydrocarbon chain on either side of the double bond.33
This reduces the length of parallel, strongly interacting chain
segments and mimics the consequences of chain shortening,
i.e., the packing parameter v/ao increases toward a value of
unity leading to lamellar or vesicle type structures. A similar
situation may prevail for gemini compounds possessing short
spacers and long, unsaturated hydrocarbon tails. Further
studies are needed to resolve this speculation.
Transmission electron microscope studies of the three gemini
surfactant systems (all at 3 mM concentration) indicated the
presence of an elongated network of particles made up of
stacks of flat structures (Fig. 5). These interconnected worm-
like micelles are typical in the 18:1-2-18:1 and 18:1-3-18:1
surfactant solutions (Fig. 5a,b), whereas the 18:1-6-18:1 solu-
tion contains fewer wormlike structures and more of individual
particles as small as 5 nm (Fig. 5c). Larger particles in the range
of 10–25 nm are also observable and these could represent
micellar clusters. The wormlike structures in the 18:1-2-18:1
and 18:1-3-18:1 surfactant solutions have dimensions of about
20–40 nm wide and several hundred nm long. These results are
in good agreement with those obtained from dynamic light
scattering (above) and implicate a preference for the formation
of higher order structures for the 18:1-s-18:1 surfactants,
particularly for small s.
Fig. 6 shows a plot of surface tension (g) vs. log surfactant
concentration for the 18:1-s-18:1 series at 25 1C with 0.01 N
NaCl in the subphase. Table 2 lists the cmc, gcmc (surface
tension at cmc), and a0 (minimum area per surfactant molecule
at the aqueous solution/air interface in nm2 moleculeꢀ1) prop-
erties of the gemini compounds obtained from these plots. The
minimum head group areas were determined from the surface
excess concentration, G, according to eqn. (2)
a0 ¼ (NAG)ꢀ1
(2)
The average particle sizes for s ¼ 2 and 3 at concentrations
above the cmc are much greater than the extended length of the
surfactant alkene chain, estimated to be ca. 2.3 nm. This
suggests that traditional micelles do not occur in these systems.
However, for s ¼ 6 a mean aggregation number of 40 was
determined by using a method similar to that described pre-
where, G is calculated from the Gibbs adsorption isotherm
relation
ꢀ
ꢁ
1
dg
2:30nRT d log C
G ¼ ꢀ
ð3Þ
T
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
3175

View Online
Fig. 5 Electron micrographs of (a) 18:1-2-18:1; (b) 18:1-3-18:1; (c) 18:1-6-18:1, bar represents 50 nm. Concentration of all surfactants was
3.00 mM.
The parameters R and T have their usual meaning, NA is the
Avogadro number, and n is the number of species at the
interface whose concentration changes with surfactant concen-
tration. It has been set at 2 or 3 in aqueous media by
investigators. In the presence of a swamping electrolyte it has
a value of one and there is no change in the ionic strength of
the aqueous solution with increasing surfactant concentration,
thus minimizing any change in the effective charge on the head
groups.34
The cmc (as determined from surface tension) is observed to
decrease as the spacer length increases in the series (Table 2).
Compared with the cmc data obtained from electrical conduc-
tance, light scattering and fluorescence probing for aqueous
surfactant solutions without salt, the presence of the electro-
lyte, as expected, decreases the cmc values for the18:1-s-18:1
series because of the reduction of electrostatic repulsion be-
tween the head groups. The surface tension at the cmc, gcmc, is
ca. 45 mN mꢀ1. The variation in the values of a0 with spacer
length for the 18:1-s-18:1 series is different from that observed
for the 12-s-12 and 16-s-16 series with s o 10. Rather than an
approximately linear increase with s, a minimum in ao is
observed for the 18:1-3-18:1 surfactant with a very large
increase for 18:1-6-18:1 occurring relative to s ¼ 3. Values of
a0 were calculated to be 0.29, 0.24, and 0.74 nm2 moleculeꢀ1 for
18:1-2-18:1, 18:1-3-18:1, and 18:1-6-18:1, respectively. The ao
values for s ¼ 2 and 3 are abnormally low. Similar results were
obtained for a four carbon ether spacer with a C18 saturated
carbon chain in the presence of a swamping electrolyte.29
Estimates of cross-sectional areas of an aliphatic chain oriented
perpendicular to the interface and of a –CH2– group lying flat
in the interface are about 0.2 and 0.07 nm2, respectively. The
cross-sectional area of the cis-alkene chain is likely to be
slightly larger than the aliphatic isomer. Therefore, the ao
values for the s ¼ 2 and 3 members do not appear to indicate
formation of a monolayer. Rather, these longer alkene chains
may be forming some type of multilayer structure at higher
solution ionic strength.11,35 This behaviour may have signifi-
cance with regard to the transfection efficiency of DNA shown
by the s ¼ 3 member with or without the presence of a helper
lipid.11 The s ¼ 6 member of the series shows a close-packed
arrangement at the air–water interface that is more consistent
with cross-sectional area projected from the alkene-chains and
the aliphatic spacer.
The enthalpies of micellization obtained from ITC measure-
ments are listed in Table 3. The DH1M for the 18:1-s-18:1
surfactants are exothermic. However, the quantitative beha-
viour of DH1M as a function of s for the 18:1-s-18:1 series
differs from that observed for the 12-s-12 series of gemini
surfactants that have been studied by
a number of
authors.18,31,36,37 The DH1M for s ¼ 2 in the 18:1-s-18:1 series
is less exothermic than for the 12-s-12 series and there is a
substantial exothermic increase in DH1M upon increasing s
from 2 to 3, which is not observed for the 12-s-12 series where
DH1M becomes progressively less exothermic with s up to s ¼
6–8. Also, DH1M for s ¼ 6 becomes less exothermic and
approximately the same as for the s ¼ 2 member of the series.
However, it is still more exothermic than the s ¼ 6 homologue
of the 12-s-12 series. The DH1M values for the s ¼ 3 and 6
members of the 18:1-s-18:1 series are ca. 5 kJ molꢀ1 more
exothermic than the corresponding members of the 12-s-12
series. While a more exothermic value is expected because of
the longer hydrocarbon chain length in the former series, the
difference turns out to be far less negative per –CH2– group
than the ꢀ1.6 kJ molꢀ1 of –CH2– found for many single-head,
single-tail surfactants.38 However, since a carbon–carbon dou-
ble bond is more hydrophilic than a carbon–carbon single
bond,39 this could give rise to a less exothermic DH1M for the
unsaturated compared to the saturated hydrocarbon chain
systems. Conformational differences between the unsaturated
and saturated carbon chains are also likely to contribute to the
reduced exothermic nature of the self-aggregation process. This
situation appears to be even more acute when one compares
the DH1M values of the compounds s ¼ 2 and m ¼ 12 or 18 : 1.
Both light scattering and surface tension results suggest self-
aggregation of s ¼ 2 and m ¼ 18 : 1 in the bulk or at the air–
water interface leads to lamellar or vesicle structures. Con-
sidering the bulkier nature of the cis-9-octadecene hydrocarbon
tails, the difficulty of packing these chains in aggregates must
place severe constraints on rotation about the C–C bonds in
the spacer, particularly when it is short. This factor decreases
the overall exothermicity of DH1M and it would be expected to
be greatest for the case of s ¼ 2. Studies of phospholipids show
that corresponding effects of a double bond in a trans config-
uration are considerably weaker than a cis double bond in the
Fig. 6 Surface tension plots for the 18:1-2-18:1 (&), 18:1-3-18:1 (J),
and 18:1-6-18:1 (n) gemini surfactants. Linear fits are represented by
lines for 18:1-2-18:1 (solid), 18:1-3-18:1 (large dash), and 18:1-6-18:1
(short dash).
3176
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Online
Table 2 Surface tension properties for the 18:1-s-18:1 gemini surfactants
Surfactant
cmc/mM
g
cmc/mN mꢀ1
a0a/nm2 moleculeꢀ1
a0b/nm2 moleculeꢀ1
18:1-2-18:1
18:1-3-18:1
18:1-6-18:1
a
17.3 ꢁ 0.4
15.1 ꢁ 0.4
11.0 ꢁ 0.4
43.5
45.0
45.0
0.88 ꢁ 0.05
0.72 ꢁ 0.05
2.22 ꢁ 0.08
0.29
0.24
0.74
b
Calculated assuming n ¼ 3 in eqn. (3). Calculated assuming n ¼ 1 in eqn. (3).
methods were in reasonable agreement. Analysis of the con-
ductance curves indicate premicelle aggregate formation for all
homologues of the 18:1-s-18:1 series, and determinations of the
vibronic ratio of pyrene are also consistent with this. The
Gibbs energy for the premicelle aggregation process shows
that it is less favored as the length of the spacer group
increases.
Table 3 Thermodynamic properties of micellization for a series of
18:1-s-18:1 gemini surfactants at 25 1C
a
Surfactant
DG1M
/
DG1premic
/
DH1M
/
TDSM1/
kJ molꢀ1
kJ molꢀ1
kJ molꢀ1
kJ molꢀ1
18:1-2-18:1
18:1-3-18:1
18:1-6-18:1
a
ꢀ46.4
ꢀ57.1
ꢀ45.5
ꢀ4.1
ꢀ4.5
ꢀ3.0
ꢀ15
ꢀ22
ꢀ14
31.4
35.1
31.5
Surface tension results show that in 0.01 N NaCl solutions,
the s ¼ 2 and 3 surfactants may be forming multilayer rather
than monolayer structures, consistent with dynamic light
scattering measurements, where large vesicular aggregates
(RH E 35 nm) are observed. The s ¼ 6 compound appears
to form spherical micelles, with an aggregation number of 40.
Electron microscopy confirms the presence of large, higher
order structures for the s ¼ 2, and 3 surfactants.
Calculated assuming DG1 ¼ (3 ꢀ 2a)RT ln cmc.
same location.33 Studies of 18:1-s-18:1 gemini compounds
having hydrocarbon chains in a trans configuration may be
able to shed more light on the interdependency of the hydro-
carbon tail configuration and the spacer conformation as it
effects the thermodynamics of self-aggregation.
The Gibbs energies, determined from cmc values obtained
by specific conductance and entropies of micellization, were
calculated in the usual manner and the results are shown in
Table 3. The micellization process is entropy driven for all
surfactants in this series, similar to the 12-s-12 series.36 In
addition to the thermodynamic micelle properties, values of the
premicelle Gibbs energy of micellization were estimated from
the relation22
As anticipated, the enthalpies of micellization are more
exothermic for the 18:1-s-18:1 surfactants as compared to the
12-s-12 surfactants, due in part to the longer hydrocarbon tails.
Differences in DH1M between the two series are also consistent
with significant steric and configurational contributions, aris-
ing from difficulties associated with packing of the bulky cis-9-
octadecene tails.
Acknowledgements
The authors thank the Saskatchewan Structural Sciences Cen-
tre (SSSC) for the use of NMR, ITC, and DLS instruments,
and Dr Jeremy Lee (Department of Biochemistry) for use of
the fluorescence spectrometer. Financial assistance provided by
the Natural Science and Engineering Research Council of
Canada (NSERC) and the Saskatchewan Synchrotron Insti-
tute is gratefully acknowledged.
DG1premic ¼ ꢀRT(1.5 ꢀ a) ln (cmc/cI)
(4)
where cI represents the value of the cmc if it had fallen on the
straight line shown in Fig. 1, i.e., in the absence of premicelle
association. They reflect mainly the loss of alkyl chain/water
contacts during formation of premicelle aggregates, and are
only a small fraction of the total decrease of contacts between
the alkyl chains and water in the formation of the regular
micelles. The DG1premic values are shown in Table 3 and are
based on cmc and a values derived from the specific conduc-
tance measurements (Table 1). Although the DG1premic data are
referenced to the projected value of cI for m ¼ 18, saturated,
since we are unaware of any studies of gemini compounds
possessing unsaturated alkyl chains, nevertheless this exercise
offers an opportunity to make a qualitative comparison of the
Gibbs energy change upon formation of premicelle aggregates
for 18:1-s-18:1 with 18-8-18. The DG1premic values for the three
members of the 18:1-s-18:1 series (s ¼ 2, 3, or 6) are less
negative (ꢀ2.25 to ꢀ1.5 kJ molꢀ1 of alkyl chain) than that of
18-8-18 (ꢀ6.2 kJ molꢀ1 of alkyl chain).22 These results indicate
that unsaturated alkyl chains are more hydrophilic than satu-
rated, in keeping with the lower Krafft points of unsaturated
compared to saturated compounds. Furthermore, the
DG1premic data in Table 3 show that premicelle aggregates are
less favored as s increases in the 18:1-s-18:1 series, contrary
to behaviour shown by the saturated gemini series.22
References
1
2
R. Zana, J. Colloid Interface Sci., 2002, 248, 203.
F. M. Menger and J. S. Keiper, Angew. Chem. Int. Ed. Engl., 2000,
39, 1907.
3
A. J. Kirby, P. Camilleri, J. B. F. N. Engberts, M. C. Feiters, R. J.
M. Nolte, O. Soderman, M. Bergsma, P. C. Bell, M. L. Fielden, C.
L. G. Rodriguez, P. Guedat, A. Kremer, C. McGregor, C. Perrin,
G. Ronsin and M. C. P. van Eijk, Angew. Chem. Int. Ed. Engl.,
2003, 42, 1448.
4
5
6
7
R. Zana and Y. Talmon, Nature, 1993, 362, 228.
D. Danino, Y. Talmon and R. Zana, Langmuir, 1995, 11, 1448.
R. Zana, M. Benrraou and R. Rueff, Langmuir, 1991, 7, 1072.
S. D. Wettig and R. E. Verrall, J. Colloid Interface Sci., 2001, 235,
310.
8
9
E. Alami, G. Beinert, P. Marie and R. Zana, Langmuir, 1993, 9,
1465.
R. Zana, J. Colloid Interface Sci., 1980, 78, 330.
10 J. K. Rose, L. Buonocore and M. A. Whitt, Biotechniques, 1991,
10, 520.
11 H. S. Rosenzweig, V. A. Rakhmanova and R. C. MacDonald,
Bioconjugate Chem., 2001, 12, 258.
12 I. Badea, R. E. Verrall, M. Baca-Estrada, S. Tikoo, A. Rosenberg,
P. Kumar and M. Foldvari, J. Gene Med., 2005, DOI: 10.1002/
jgm.763.
Conclusions
13 W. Rawicz, K. C. Olbrich, T. McIntosh, D. Needham and E.
Evans, Biophys. J., 2000, 79, 328.
14 J. Sandri and J. Viala, Synth. Commun., 1992, 22, 2945.
15 W. D. Harkins and H. F. Jordan, J. Am. Chem. Soc., 1930, 52,
1751.
A series of gemini surfactants, 18:1-s-18:1, with oleyl alkyl
chains and s ¼ 2, 3, and 6, were synthesized, characterized, and
their thermodynamic properties determined. The cmc values of
18:1-2-18:1, 18:1-3-18:1, and 18:1-6-18:1 were found to be 26.9,
23.4, and 18.0 mM, respectively, by using the electrical con-
ductance method. Values of the cmc obtained by using other
16 K. M. Jenkins, S. D. Wettig and R. E. Verrall, J. Colloid Interface
Sci., 2002, 247, 456.
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
3177

View Online
29 M. J. Rosen, J. H. Mathias and L. Davenport, Langmuir, 1999, 15,
7340.
30 D. Danino, Y. Talmon and R. Zana, Langmuir, 1995, 11, 1448.
31 S. D. Wettig, X. F. Li and R. E. Verrall, Langmuir, 2003, 19, 3666.
32 D. Marsh, Biophys. J., 1989, 55, 1093.
33 J. H. Seddon and G. Cevc, in Lipid polymorphism: Structure and
stability of lyotropic mesophase of phospholipids, ed. G. Cevc,
New York, 1993.
17 R. Zana and H. Levy, Colloids Surf. A, 1997, 127, 229.
18 S. D. Wettig, P. Nowak and R. E. Verrall, Langmuir, 2002, 18,
5354.
19 K. W. Mattison, in PSI Books (DynaPro Dynamic Light Scatter-
ing Instrument Control Software Manual), Protein Solutions,
Lakewood, NJ, 2001.
20 Y. L. Wang, K. Kimura, P. L. Dubin and W. Jaeger, Macro-
molecules, 2000, 33, 3324.
21 X. F. Li, S. D. Wettig and R. E. Verrall, J. Colloid Interface Sci.,
2005, 282, 466.
22 R. Zana, J. Colloid Interface Sci., 2002, 246, 182.
23 A. W. Ralston, D. N. Eggenberger and P. L. Du Brow, J. Am.
Chem. Soc., 1948, 70, 977.
24 L. D. Song and M. J. Rosen, Langmuir, 1996, 12, 1149.
25 F. M. Menger and C. A. Littau, J. Am. Chem. Soc., 1991, 113, 1451.
26 F. M. Menger and C. A. Littau, J. Am. Chem. Soc., 1993, 115, 10083.
27 S. D. Wettig, Ph.D. Thesis, University of Saskatchewan, Saska-
toon SK, 2000.
34 M. J. Rosen, Surfactants and Interfacial Phenomena, Wiley, 1989.
35 R. Nagarajan, in Structure–Performance Relationships in Surfac-
tants, ed. K. Esumi and M. Ueno, Marcel Dekker, Inc., New
York, 1997.
36 G. Y. Bai, H. K. Yan and R. K. Thomas, Langmuir, 2001, 17,
4501.
37 L. Grosmaire, M. Chorro, C. Chorro, S. Partyka and R. Zana,
J. Colloid Interface Sci., 2002, 246, 175.
38 M. T. Bashford and E. M. Woolley, J. Phys. Chem., 1985, 89,
3173.
28 M. Frindi, B. Michels and R. Zana, J. Phys. Chem., 1994, 98,
6607.
39 P. P. S. Saluja, T. M. Young, R. F. Rodewald, F. H. Fuchs,
D. Kohli and R. Fuchs, J. Am. Chem. Soc., 1977, 99, 2949.
3178
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 1 7 2 – 3 1 7 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5