11434 Langmuir 2007, 23, 11434-11442

Articles
Studies on an Ester-Modified Cationic Amphiphile in Aqueous
Systems: Behavior of Binary Solutions and Ternary Mixtures with
Conventional Surfactants
Dan Lundberg,*,†,‡ Johan Unga,§ Ashley L. Galloway,† and Fredric M. Menger*,†
Department of Chemistry, Emory UniVersity, Atlanta, Georgia 30322, and Department of Chemical and
Biological Engineering, Chalmers UniVersity of Technology, 412 96 Göteborg, Sweden

ReceiVed February 13, 2007. In Final Form: August 15, 2007

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

The aqueous behavior of an ester-modified cationic amphiphile with the molecular structure CH3CH2O(CdO)-
(CH2)6(CdO)O(CH2)8N+(CH3)3Br-, in the following referred to as A, has been investigated. Systems with A as the
only solute, as well as different aqueous mixtures with conventional cationic surfactants, primarily dodecyltrim-
Downloaded via ESS INFLIBNET PCA 1 on October 9, 2023 at 09:56:31 (UTC).

ethylammonium bromide (DTAB), were included in the study. Isotropic solution samples were characterized using
1H NMR, 13C NMR, NMR diffusometry, and conductivity measurements, whereas liquid crystalline samples were

investigated by optical polarization microscopy and small-angle X-ray diffraction. The results are compared to the
behavior of the binary system of DTAB and water. A does not exhibit a typical surfactant behavior. When it is present
as the only solute in a binary aqueous system, it forms neither conventional micelles nor liquid crystalline phases.
However, there is clear evidence that it assembles with lower cooperativity into loose clusters at concentrations above
25-30 mM. When A is mixed with DTAB in solution, the two amphiphiles form mixed assemblies, the structure
of which varies with the total amphiphile concentration. In concentrated mixtures with alkyltrimethylammonium
surfactants, A can participate in hexagonal liquid crystalline phases even when it constitutes a significant fraction of
the total amphiphile content.

Introduction of hydration, but it generally occurs at a well-defined concentra-

Most water-soluble organic substances show at least some
degree of amphiphilicity and, hence, have an inherent driving
force to minimize unfavorable exposure of their hydrophobic
moieties to water.1 The amphiphilic character of a compound is
manifested, for instance, by a decrease in the surface tension of
its aqueous solution. Even a small molecule, such as methanol,
which has an unlimited miscibility with water, causes a significant
decrease in surface tension by exposing its methyl group to the
air.1 Another common consequence of the amphiphilicity is
aggregation of the molecules in aqueous solutions. Depending
on the molecular structure of the solute, particularly the size,
type, and distribution of its hydrophobic portions, the aggregation
can show a varying degree of cooperativity, and the resulting
aggregates can have very different morphologies and sizes,
everything from small oligomers such as dimers or trimers to
large assemblies consisting of thousands of molecules is possible.1
Amphiphiles with distinctly separated hydrophilic and hy-
drophobic parts, such as synthetic surfactants and certain natural
polar lipids, commonly show self-assembly with a high degree
of cooperativity, resulting in, for instance, micelles or bilayer
structures. Micelle formation is governed by a delicate balance
between several factors, e.g., the relative sizes of the hydrophilic
and hydrophobic parts, the presence of charges, and the degree
* Corresponding authors. E-mail: delundb@emory.edu (D.L.), menger@
emory.edu (F.M.M.).
† Emory University.
‡ Present address: Department of Chemistry, University of Coimbra, 3004-
535 Coimbra, Portugal.
§ Chalmers University of Technology.
(1) Mukerjee, P. J. Pharm. Sci. 1974, 63, 972.
tion, the critical micelle concentration (cmc). It is critically
important for the ability of an amphiphile to form micelles that
its hydrophobic moieties can pack closely in order to effectively
exclude water from the aggregates. Hence, the hydrophobic parts
must be flexible. Surfactant-like molecules with rigid structures,
such as bile salts or substances carrying a polycyclic aromatic
group as the hydrophobic part do in general not form conventional
micelles, even if their hydrophilic and hydrophobic moieties are
well-separated.1
In recent years, Menger and co-workers have studied several
amphiphiles, whose structures are based on that of the simplest
type of surfactants, single-tailed with a small ionic headgroup,
but are modified by the inclusion of nonhydrocarbon groups in
their hydrophobic parts. One of the main incentives for studying
these substances has been to investigate how the presence of
different functional groups affect the self-assembly behavior of
amphiphiles. Thus, substances having one or two ester,2,3 ether,4
or thioether groups5 inserted at different positions along the
hydrocarbon chain have been synthesized and characterized with
regard to their basic physicochemical properties in aqueous
solution. All of the inserted functional groups can be classified
(to use the terminology of Laughlin6) as nonoperative functional
groups, i.e., they are by themselves not sufficiently hydrophilic
to act as a surfactant head group. On the other hand, the inserted
(2) Menger, F. M.; Galloway, A. L. J. Am. Chem. Soc. 2004, 126, 15883.
(3) Menger, F. M.; Galloway, A. L.; Chlebowski, M. E. Langmuir 2005, 21,
9010.
(4) Menger, F. M.; Chlebowski, M. E. Langmuir 2005, 21, 2689.
(5) Menger, F. M.; Shi, L. J. Am. Chem. Soc. 2006, 128, 9338.
(6) Laughlin, R. G. The Aqueous Phase BehaVior of Surfactants; Academic
Press: London, 1994.

10.1021/la700430u CCC: $37.00 © 2007 American Chemical Society

Published on Web 10/09/2007
An Ester-Modified Amphiphile in Aqueous Systems
Figure 1. The molecular structure of A.
functionalities can all be expected to show differences in their
interactions with water, both in magnitude and type, as compared
to the ordinary hydrocarbon segment that they replace.
Since hydrophobic and hydrophilic interactions are interde-
pendent, and hence not additive,7 it is difficult to predict the
degree of tail modification that can be sustained before an
amphiphile loses its typical surfactant behavior, e.g., the ability
to form micelles at a distinct concentration. It is also difficult
to predict the character of nonmicellar assemblies that may
possibly be formed by “overly modified” surfactants. If an inserted
functional group shows a significant attraction to water, this may
severely impede close packing of the hydrophobic parts, and
hence decrease the propensity for micelle formation. On the
other hand, if the interactions with the solvent are weaker, the
chain modification may have a minor, or even negligible, effect
on the qualitative behavior of the amphiphile.
Not surprisingly, the behavior of the modified surfactants varies
widely depending on the type, number, and positions of the
inserted functional groups. However, some general trends are
observed. For instance, among the studied compounds that are
modified with ester groups, which can act as hydrogen-bond
donors and thus show notable interactions with water, only
amphiphiles having a terminal segment of C10H21 or larger form
normal micelles.2,8 For related compounds that lack a sufficiently
large terminal hydrocarbon segment, the aggregation seems to
occur with lower cooperativity, resulting in smaller assemblies
with a looser structure than in a conventional micelle. Conse-
quently, it is in a sense possible to view the hydrocarbon portion
between the ester group and the polar head group effectively as
an extension of the latter. This observation is consistent with
previous reports on modified surfactants.6
The objective of this work was to make a more detailed
investigation of the aqueous behavior of one of the ester-modified
amphiphiles that were previously synthesized and studied in our
laboratories, namely the compound A depicted in Figure 1.2,3,9
One of the main aims was to obtain a better picture of its aggregates
in dilute solutions, but we also extended the composition range
from that in previous work to include much higher concentrations
(up to 90 wt % of amphiphile). Furthermore, in addition to studies
on solutions of A by itself, we present data on the aggregation
and phase behavior of mixtures of A and conventional cationic
surfactants of the alkyltrimethylammonium bromide type,
primarily dodecyltrimethylammonium bromide (DTAB). A
multimethod approach was adopted: isotropic solution samples
were characterized using 1H NMR, 13C NMR, NMR diffusometry,
and conductivity measurements, whereas liquid crystalline
samples were investigated by optical polarization microscopy
and small-angle X-ray diffraction. Throughout the text, the results
are discussed in relation to the binary system of DTAB and
water, which is taken to represent “typical” surfactant behavior.
Experimental Procedures
Materials. A was synthesized using previously described
procedures.2 DTAB (99%) and tetradecyltrimethylammonium bro-
mide (TTAB, 99%) were purchased from Sigma, decyltrimethy-
(7) Israelachvili, J. N. Intermolecular and Surface Forces, 2nd ed.; Academic
Press: London, 1992.
(8) As judged from surface tension vs concentration plots and solubilization
studies.
(9) A was designated E in ref 2 and A in ref 3.
Langmuir, Vol. 23, No. 23, 2007 11435
Table 1. The Used Values for the Densities of the Complete
Amphiphiles (G) and Their Apolar Parts (Gap), Respectively
F Fap F Fap
amphiphile (g/cm3) (g/cm3) amphiphile (g/cm3) (g/cm3)
A 1.12 0.98 DTAB 1.04 0.81
DeTAB - 0.80 TTAB - 0.82
lammonum bromide (DeTAB, >98%) and hexadecyltrimethylam-
monium bromide (CTAB, >99%) were from Fluka, and hexamethyl-
disilane (HMDS, 98%) was from Aldrich. Deuterium oxide (99.9%
D) was obtained from Cambridge Isotope Labs. All commercial
chemicals were used as received. The aqueous solutions used in the
conductivity experiments were prepared using water that had been
purified with a Millipore Milli-Q system.
Samples. Isotropic Solutions. Solution samples were prepared by
mixing and/or diluting concentrated solutions of A or DTAB prepared
with either D2O or H2O. The dilution was made by volume, but the
aliquots were also weighed for verification.
In samples containing HMDS, a small amount of this substance
(on the order of one molecule of HMDS per 1000 surfactant
molecules, verified by 1H NMR) was added to the sample solutions
using a microliter syringe. To ensure that the added HMDS did not
cause any significant change in the solution behavior, the self-
diffusion coefficient of the amphiphile was measured both before
and after the addition.
When the same samples were used in multiple experiments and
more than 2 days passed between these, the samples were stored at
5 °C in order to avoid hydrolysis of the ester groups. In some of the
most concentrated solutions of A, a detectable degradation of the
substance was observed (on the order of a few percent). However,
repeated experiments on these samples showed that the hydrolysis
products did not cause any appreciable changes in the behavior of
the samples.
Liquid Crystalline Phases. Liquid crystalline samples were
prepared by weighing the appropriate amounts of surfactants and
D2O into screw-capped vials. The surfactant components (which are
all dry powders) were mixed by shaking the vials before the solvent
was added. After preparation, the samples were allowed to rest for
a few days, centrifuged, and finally left for equilibration for at least
1 week before characterization. To monitor sample stability, 1H
NMR spectra were run on samples prepared by diluting small portions
of the liquid crystals. No appreciable degradation was detected during
the relevant storage times.
Only samples that by visual observation were deemed to be single-
phase were examined by optical microscopy and SAXD.
Calculation of the Volume Fraction of Aggregates. The volume
fractions of aggregates, Φ, used in the calculations involving eq 3
and in Figure 6, and the volume fractions of the apolar domains, Φap,
shown in Table 2, were estimated using the densities presented in
Table 1. These densities were calculated from the molecular weights
(or fractional molecular weights) and the following values of group
volumes: 26.99 Å3 for -CH2-, 54.19 Å3 for -CH3, 35.00 Å3 for
-COO-, and 141.0 Å3 for -N(CH3)3Br.10 For A, the whole chain,
including the ester groups, is assumed to contribute to the apolar
volume.
When calculating the volume fraction of aggregates in the isotropic
solution samples, we assumed that the concentrations of aggregated
amphiphile are c - 30 mM for A and c - 16 mM for DTAB, where
c denotes the total concentration of amphiphile. In the hexagonal
phases, the head group region is considered as part of the polar
domains.
NMR. All NMR experiments were performed at 25 °C on a Varian
INOVA 600 spectrometer equipped with a pulsed field gradient
(PFG) generator and a PFG amplifier. The samples were inserted
to the probe at least 20 min prior to the experiments to allow for
thermal equilibrium to be attained.
The 1D experiments were performed at 599.7 and 150.8 MHz for
the 1H and 13C experiments, respectively. The 13C NMR spectra
(10) Hagslätt, H.; Söderman, O.; Jönsson, B. Liq. Cryst. 1992, 12, 667.
11436 Langmuir, Vol. 23, No. 23, 2007
were recorded in a 1H-decoupled mode. It was assumed that the
chemical shift of the 2H lock signal was independent of the amphiphile
concentration (this has been shown to be a good approximation for
other ionic amphiphiles11), and since the frequency offset was kept
constant, the changes in the 13C NMR chemical shifts with
concentration could be calculated directly from the “absolute”
frequencies of the respective resonances.
In the 13C NMR spectra, one peak is observed for each
nonequivalent carbon. For A it was possible to assign the following
resonances with good certainty (as counted from the headgroup):
the methylene carbon neighboring the head group (C1), the methylene
carbons next to the ester oxygens (C8 and C17), the methylene
carbons next to the carbonyl groups (C10 and C15, as a pair), the
carbonyl carbons (C9 and C16, as a pair), the terminal methyl carbon
(C18), and the head group methyl carbons. For DTAB, C1, C3, C10,
C12, and the headgroup methyl carbons were assigned. The carbons
that are not specifically assigned are referred to as midchain methylene
carbons.
The diffusion experiments were run using a Hahn-echo sequence
with intervening pulsed field gradients (PG), resulting in a complete
pulse sequence of 90°-PG-180°-PG. The delays between the
gradient pulses (∆) and the width of the pulsed gradients (δ) was
kept constant at 140 and 7 ms, respectively, while the strength of
the pulsed gradient (G) was linearly incremented from 0.01 up to
0.6 T/m (maximum varied among experiments and samples) in 16
steps. The gradient strength was calibrated by making a measurement
on a trace amount of H2O in D2O (D ) 1.902 × 10-9 m2 s-1), while
linearity of the gradient amplifier in the applied gradient strength
interval was verified by measurements on poly(ethylene glycols)
with known D.12
The self-diffusion coefficients (D) of A and water were obtained
from the attenuation of the relevant echo peaks by linear least-
square fits to the Stejskal-Tanner equation (eq 1):13
1n(I/I0) ) -(γGδ)2 D(∆ - δ/3)
where I is the measured signal intensity, I0 the signal intensity in
the absence of gradient pulses, γ the magnetogyric ratio of protons,
and the rest of the parameters are as defined above. In all experiments
the observed echo-decays gave good fits to eq 1, which shows that
they represent single self-diffusion coefficients.
Conductivity Measurements. Electrical conductivity measure-
ments were performed using a Fisher Scientific Traceable conduc-
timeter, which was calibrated to standard solutions with known
conductivities (Fisher Scientific Traceable One-Shot). The measure-
ments were carried out at 25 °C, and the electrode was immersed
in stirred sample solutions until a stable reading was achieved.
Optical Microscopy. The liquid crystalline samples were
examined using a Nikon Diaphot-TMD microscope, with the samples
placed between crossed polarizers, in order to identify characteristic
birefringence patterns.
Small-Angle X-ray Diffraction (SAXD). The SAXD experiments
were performed on a Hecus X-ray Systems Kratky camera equipped
with a MBraun linear position-sensitive detector. CuKR X-rays with
a wavelength of 0.1542 nm were produced by a Philips PW 1830/40
X-ray generator operated at 50 kV and 40 mA, and a tungsten beam
stop was used to protect the detector from the primary beam. In each
experiment, data were collected for 30 min in a vacuum.
Results and Discussion
Overview of Phase Behavior. The modified surfactant A
exhibits a significant aqueous solubility. It forms an isotropic
solution when present at up to approximately 75 wt % (1850
mM) in a binary mixture with water (D2O), while at higher
(11) Ödberg, L.; Svens, B.; Danielsson, I. J. Colloid Interface Sci. 1972, 41,
298.
(12) Nydén, M. Measuring Micelle Size and Shape. In Handbook of Applied
Surface and Colloid Chemistry; Holmberg, K., Ed.; John Wiley & Sons: New
York, 2001; Vol. 2, p 281.
(13) Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 288.
Lundberg et al.
Figure 2. Phase diagram of the system A/DTAB/D2O at 25 °C. The
accuracy of the phase boundaries is within 5 wt %. The shaded
region represents compositions not characterized. L ) isotropic liquid
phase, Hex ) Hexagonal liquid crystalline phase.
concentrations the solution coexists with a solid, presumably
hydrated amphiphile. When the concentration of A is higher
than about 1000 mM (approximately 40 wt %), there is a visually
detectable increase in the viscosity of the solutions.
Contrary to what is common for conventional surfactants, the
A system shows no liquid crystalline phases within the whole
investigated composition range (up to 90 wt %). This departure
(1) from typical surfactant behavior does in itself indicate that
aggregates of A probably do not have well-defined hydrophobic
domains. If such assemblies are formed, these are generally
expected to be forced, by geometrical constraints, into an ordered
packing at higher concentrations. For comparison, the maximum
volume fraction of randomly close-packed hard spheres is about
0.64.14 However, in the case of spherical surfactant micelles, the
presence of aggregate-aggregate repulsions (due to electrostatic
interactions and/or hydration) generally causes the formation of
liquid crystalline phases at significantly lower volume fractions.15
The aqueous phase behavior of A can be compared to that of
our reference system, i.e., the aqueous system of the conventional
cationic surfactant DTAB, which shows aggregate growth at
surfactant concentrations above approximately 450 mM (13 wt
%)16 and a hexagonal liquid crystalline phase above about 50 wt
%.17 It can also be valuable to relate the behavior of the
concentrated, viscous solutions of A to that of solutions of CTAB,
a surfactant that is known to form large, wormlike micelles at
concentrations above 10 wt %. A visual comparison of a 60 wt
% (1500 mM) solution of A and a 16 wt % solution of CTAB
reveals that the viscosity of the latter is dramatically higher. This
finding suggests that if any aggregate growth occurs in solutions
of A, this growth is much less pronounced than that in solutions
of CTAB.
When A is present in concentrated aqueous mixtures with
surfactants of the XTAB type, it can participate in hexagonal
liquid crystalline phases even when it is present in quite high
proportions. The phase diagram of the system A/DTAB/D2O is
shown in Figure 2. Hexagonal phases are also formed in
(14) Scott, G. D.; Kilgour, D. M. J. Phys. D: Appl. Phys. 1969, 2, 863.
(15) Holmberg, K.; Jönsson, B.; Kronberg, B.; Lindman, B. Surfactants and
Polymers in Aqueous Solution, 2nd ed.; John Wiley & Sons: Chichester, 2002.
(16) Minardi, R. M.; Schulz, P. C.; Vuano, B. Colloids Surf. A 2002, 197, 167.
(17) McGrath, K. M. Langmuir 1995, 11, 1835.
An Ester-Modified Amphiphile in Aqueous Systems Langmuir, Vol. 23, No. 23, 2007 11437

Figure 3. Partial 1H NMR spectra of A in D2O at three different concentrations. 0.1 ppm ) 60 Hz.

concentrated mixtures of A and DeTAB or TTAB (see the Small-

Angle X-ray Diffraction section). The extension of the liquid
crystalline region toward the binary D2O-A axis increases with
an increasing alkyl chain length of the XTAB molecule.
Characterization of Isotropic Solution Samples. 1H NMR.
Figure 3 shows the 1H NMR spectra of three solutions of A at
different concentrations. When comparing the three spectra, one
can clearly see that the peaks become gradually broader with
increasing concentration. A pronounced “smearing” of the
resonances occurs for concentrations of A above 750 mM, which
means that the change in peak width probably reflects the increase
in solution viscosity that was mentioned above. Both an increase
in solution viscosity and a broadening of NMR peaks can be
signs of aggregate growth. However, in comparison to the peak
broadening that is observed when an amphiphile of similar size
as A forms large micelles, the change in width for the resonances
in Figure 3 is almost negligible. For instance, the half-height
bandwidths of the 1H NMR peaks from CTAB in threadlike
micelles are on the order of hundreds of hertz.18 Furthermore, Figure 4. The observed self-diffusion coefficients of A (O) or DTAB
(×) when each substance is present as single amphiphiles, of A (4)
the broad resonances do not show the typical band shape, or DTAB (+) when present in an equimolar mixture of the two, and
characterized by a broad base combined with a sharp apex, that HMDS present in trace amounts in solutions of A (0) or DTAB (])
is expected in the 1H NMR spectrum of an amphiphile as a function of the total surfactant concentration. For the surfactants,
participating in large micelles.18,19 These findings are consistent Dobs values were measured for A alone, DTAB alone, as well as for
with the difference in viscosity of solutions of A or CTAB and both compounds in the mixture, at all studied concentrations. DHMDS
give further support to the notion that A does not form typical values were determined in solutions of 50-500 mM A or 50-400
micelle-like aggregates. mM DTAB.
In Figure 3 it can also be seen that the resonances shift downfield to determine the self-diffusion coefficients of the amphiphiles,
with increasing concentration. This phenomenon will be further the water, and the highly hydrophobic probe molecule HMDS
discussed in the 13C NMR section. in samples with different concentrations of A and/or DTAB.
For the mixtures of A and DTAB, there is no qualitative When HMDS is added to a micellar solution, it is heavily
difference in the appearance of the 1H NMR spectra as compared partitioned to the hydrophobic regions of the aggregates. Thus,
to those of samples containing A as the lone solute. under the assumption that HMDS resides only in the aggregates,
NMR Diffusometry. The self-diffusion coefficients of the its self-diffusion coefficient can be taken to represent the diffusion
different species present in an aqueous surfactant solution depend coefficient of the aggregates.
on the formation of self-assemblies and on other interactions Figure 4 shows the self-diffusion coefficients for A and DTAB
between the components. Thus, in order to obtain information in aqueous solutions where the compounds are present either as
on the aggregate sizes and the degrees of hydration in the presently the sole amphiphiles or in equimolar aqueous mixtures of both,
investigated systems, the PGSE-NMR technique20-23 was used as well as the diffusion coefficients for HMDS that is present
in trace amounts in solutions of A or DTAB. In addition to the
(18) Ulmius, J.; Wennerström, H. J. Magn. Reson. 1977, 28, 309.
(19) Olsson, U.; Söderman, O.; Guering, P. J. Phys. Chem. 1986, 90, 5223.
data shown in Figure 4, the self-diffusion coefficient of A was
(20) Furo, I. J. Mol. Liq. 2005, 117, 117. also measured in samples with concentrations between 500 and
(21) Söderman, O.; Stilbs, P.; Price, W. S. Concepts Magn. Reson. A 2004, 1500 mM. In this interval, the observed diffusion coefficient
23A, 121.
(22) Price, W. S. Concepts Magn. Reson. 1997, 9, 299. decreases essentially linearly down to 1.4 × 10-11 m2 s-1 at
(23) Price, W. S. Concepts Magn. Reson. 1998, 10, 197. 1500 mM. The observed decrease in the amphiphile self-diffusion
11438 Langmuir, Vol. 23, No. 23, 2007
Figure 5. The observed self-diffusion coefficients of A (O) or DTAB
(×) as a function of the reciprocal of the total surfactant concentration.
The dashed line is a prediction of the observed diffusion coefficients
obtained using eq 2 with a Dmono of 4.7 × 10-10 m2 s-1, a Dagg of
8.7 × 10-11 m2 s-1, and a cmc of 15.6 mM.
coefficients with increasing concentration is consistent with
assembly into some sort of aggregate. A comparison between
the diffusion coefficients for the two amphiphiles show that the
aggregates of A have an apparent size on the same order as that
of DTAB micelles.
Under the assumptions that only monomers are present below
the cmc and that the micelles can be approximated as mono-
disperse, discrete entities, the observed self-diffusion coefficient,
Dobs, is the weighted average of the diffusion coefficients for the
monomers, Dmono, and the aggregates, Dagg, according to eq 2
cmono cagg
Dobs ) Dmono + D (2)
c c agg
where cmono and cagg are the concentrations of surfactant molecules
at the respective sites, and c is the total surfactant concentration.15
For a conventional, micelle-forming surfactant, eq 2 gives a
good approximation of reality. It follows from eq 2 that, for a
conventional surfactant, a plot of Dobs versus the reciprocal of
the total surfactant concentration should show two approximately
straight lines intersecting at the cmc, a horizontal line Dobs )
Dmono for concentrations up to the cmc and a line with a slope
of approximately (Dmono - Dagg) × cmc for higher concentrations.
Plots of Dobs of A and DTAB versus c-1, see Figure 5, show
approximately linear sections at both high and low c-1 for both
the amphiphiles. However, the plots also indicate a higher
curvature around the intersection of the two lines in the case of
A than in the case of DTAB. A smoother change in Dobs with
c-1 indicates a less cooperative self-assembly and/or a gradual
growth in aggregate size with increasing concentration.24 Note
that Dobs of A is lower than Dmono at values of c-1 higher than
0.033 mM-1.3 This finding indicates that A begins to assemble
at concentration lower than 30 mM, i.e., the concentration where
a break in a surface tension curve for A has been previously
observed.3
To make a more detailed analysis of the aggregation behavior
of A, we took a closer look at the observed self-diffusion
coefficients of HMDS, DHMDS. One can see in Figure 4 that
DHMDS decreases slowly with increasing concentration. A decrease
with increasing volume fraction of aggregates in a solution is
expected, due to obstruction effects, but it is also possible that
there is a contribution from an increase in aggregate size. In the
case of monodisperse spherical aggregates of constant size, the
decrease in the diffusion coefficient, Dagg, as a function of the
volume fraction of aggregates, Φ, can, at low volume fractions,
(24) Evans, D. F.; Wennerström, H. The Colloidal Domain: Where Physics,
Chemistry, Biology, and Technology Meet, 2nd ed.; Wiley-VCH: New York,
1999.
Lundberg et al.
generally be described by a first-order expression (eq 3)12,21
Dagg ) D0,agg(1 - kΦ) (3)
where D0,agg is the diffusion coefficient of the micelles at infinite
dilution and k is a constant that usually varies between 1.7 and
2.5, depending on the type and degree of interactions between
the aggregates.12,25,26
DHMDS for DTAB shows an essentially linear decrease with
concentration, and a fit of these diffusion coefficients versus Φ
(see Experimental Procedures for calculation of Φ) to eq 3 gives
a value of k of 2.2, which is consistent with spherical aggregates,
and a D0,agg of 8.7 × 10-11 m2 s-1. The latter is related to the
hydrodynamic radius of the aggregates, RH, via the Stokes-
Einstein equation (eq 4)
kBT
D0,agg ) (4)
6πηRH
where kB is the Boltzmann constant and η the viscosity of the
solvent at the experimental temperature T. Setting η ) 1.132
mPa s for D2O in eq 4 gives a hydrodynamic radius of the DTAB
micelles of 2.2 nm. This value approximates the extended length
of a DTAB molecule, i.e., 2.1 nm, and is hence close to the
expected micellar radius.7
In the A system, DHMDS show a nonlinear decrease with
increasing amphiphile concentration that cannot be fit directly
to eq 3. This nonlinearity supports the idea that the assemblies
of A grow continuously with increasing concentration. It is
important to keep in mind, however, that since the aggregates
may have a loose structure, e.g., contain water due to the presence
of the more polar ester groups, it is likely that partitioning of
HMDS into these assemblies is weaker than it is into the more
well-defined hydrophobic domains of conventional micelles, e.g.,
those formed by DTAB. In this case, the measured DHMDS can,
particularly at lower concentrations, be expected to be higher
than the true Dagg and may hence give an underestimation in the
effective size of the A aggregates when this is calculated using
eq 4. If, however, the discrepancy is assumed to be negligible
at 500 mM A (where Φ ) 0.19) and obstruction as an
approximation is accounted for via eq 3 with a k of 2 (which is
roughly the average of the k values generally observed for
spherical aggregates), one obtains an RH of 3.1 nm. This value
approximates the extended length of a molecule of A. Hence,
the diffusometry results for HMDS suggest that the assemblies
of A at higher amphiphile concentrations have an apparent size
similar to that expected had it formed conventional micelles.
The estimates of Dagg for A and DTAB together with the
experimental values of Dmono and the cmc values15 were inserted
into eq 2 to calculate the corresponding Dobs vs c-1. As can be
seen in Figure 5, the experimental values of Dobs for DTAB show
a good fit to the calculated concentration dependence, except at
the lowest c-1. This discrepancy is probably caused by obstruction
effects and/or minor micellar growth. However, regardless of
the cmc value used, the measured Dobs values of A will not fit
eq 2, which is consistent with our interpretation of the diffusion
data.
Returning to Figure 4, one can see that the values of Dobs for
A and DTAB in the equimolar mixtures are very close to the
values obtained for the individual amphiphiles at equivalent total
concentrations. This observation gives clear evidence for the
formation of mixed assemblies. If the onset of aggregation of
(25) Evans, G. T.; James, C. P. J. Chem. Phys. 1983, 79, 5553.
(26) Ohtsuki, T.; Okano, K. J. Chem. Phys. 1982, 77, 1443.
An Ester-Modified Amphiphile in Aqueous Systems
Figure 6. Normalized self-diffusion coefficients of water as a
function of the volume fraction of aggregated A (O) or DTAB (b).
D0,water, i.e., the self-diffusion coefficient for water in bulk, is 1.902
× 10-9 m2 s-1. In terms of molarity, the concentrations range from
50 to 1000 mM A and from 25 to 1400 mM DTAB. The calculation
of Φ is detailed in the Experimental Procedures. The dashed line
represents a prediction of the contribution from obstruction to the
decrease in the water diffusion (see the text).
one amphiphile was unaffected by the presence of the other,
signs of aggregation would not be observed until each of the
components reached their respective aggregation concentrations
(i.e., the cmc in the case of DTAB). Another important observation
is that Dobs of A is notably higher than that of DTAB in a significant
concentration range above the onset of assembly. Since Dmono
of DTAB is higher than Dmono of A, this observation directly
shows that the first formed micelles (with increasing concentra-
tion) consist mainly of DTAB and that the fraction of A in the
micelles increases with increasing total concentration (refer to
eq 2).
Figure 6 shows the concentration dependence of the self-
diffusion coefficients of water, Dwater, in solutions of either A
or DTAB. The data are normalized to the self-diffusion coefficient
of water in bulk (i.e., D of a trace amount of HDO in pure D2O),
D0,water, and plotted versus the volume fraction of aggregated
amphiphile, Φ.
The self-diffusion coefficient of water in a micellar solution
is reduced due to the combination of two effects. First, the
aggregates exclude a fraction of the total volume for the diffusing
water and thereby obstruct the diffusion pathway. Second, a
fraction of the water hydrates the surfactant molecules. If the
aggregates in a solution can be approximated as hard spheres,
the decrease in the diffusion coefficient of the solvent due to
obstruction can be predicted by eq 527
D0,water
Dwater ) (5)
1 + Φ/2
where all parameters are defined above. Such an approximation
is relevant for spherical micelles and hence applicable at least
for the DTAB system. A plot of a prediction from eq 5 is included
in Figure 6. It is evident from Figure 6 that hydration of the
solutes provides a major contribution to the decrease in water
mobility in both of the studied systems. One can also see that
the decrease in Dwater with amphiphile concentration is signifi-
cantly larger in solutions of A than in solutions of DTAB at a
given concentration.
As can be understood from eq 5, the obstruction of solvent
diffusion exerted by spherical aggregates depends on the volume
(27) Jönsson, B.; Wennerström, H.; Nilsson, P. G.; Linse, P. Colloid Polym.
Sci. 1986, 264, 77.
Langmuir, Vol. 23, No. 23, 2007 11439
fraction of dispersed matter but not on the radius of the spheres.28
Thus, if we for a moment assume the hypothetical situation where
A forms conventional micelles, a difference in Dwater at a given
Φ between solutions of A or DTAB can be attributed to a
difference in the degree of hydration of the respective compounds.
Since the two amphiphiles both carry trimethylammonium groups
as their hydrophilic head groups and since it hence is reasonable
to assume that the hydration number per head group is similar
in the two cases, one would then expect the decrease in Dwater
to be larger in the sample with the higher molar concentration,
i.e., the DTAB solution (due to its lower molecular weight).
However, the opposite is observed. This finding gives further
strong support to the proposal that A forms loosely assembled
clusters. For instance, the degree of hydrophobic hydration per
amphiphile molecule can be expected to increase with an increased
area of exposure to the surrounding water,29 which is one of the
consequences of “looseness”. In addition, confinement of water
within the clusters, as well as attractive interactions between the
ester groups and the solvent, may contribute to the decrease in
Dwater in the A system relative to that in the DTAB system.
13C NMR. It could be seen in Figure 3 that the peaks in the
1H NMR spectra of A in D O shifted downfield with increasing
2
concentration. A concentration dependence of NMR chemical
shifts is commonly observed for amphiphiles in aqueous solution
and can in general be attributed to a combination of differences
in the direct effects of the environment (“medium effects”) and
differences in the average conformation of the molecules
(“conformation effects”) between the monomeric and aggregated
states. The phenomenon has been used to monitor micellization
of surfactants for several decades.30-32
It has been shown that the changes in 13C NMR shifts for an
alkyl chain show only a very weak dependence on medium effects,
and thus to a good approximation can be ascribed exclusively
to changes in chain conformation,33-35 and that a downfield shift
can be related to an increase in the average ratio of trans to
gauche conformations in the chains.34,35 In contrast to the case
of alkyl carbons, the chemical shift of a carbonyl carbon can
show a significant dependence on direct interactions with the
surrounding solvent.33 This is particularly true in a protic solvent,
as water, where the degree of solvent-solute hydrogen bonding
have a pronounced influence on the observed chemical shift.
Hence, the change in shift for the carbonyl carbons of A can to
a significant extent be taken to reflect the degree of exposure of
the ester bonds to water.
When only one peak for each nonequivalent carbon in a
molecule appears in the spectra, the observed chemical shift,
δobs, for each atom is the population-weighted average of the
values at the different sites where the molecules reside.
Accordingly, in the case of a conventional, micelle-forming
surfactant, one obtains an expression for δobs of the same form
as eq 2, given in eq 6
(28) This is true as long as the aggregate dimensions are much larger than the
size of a water molecule, a condition that is fulfilled for conventional surfactant
micelles.
(29) Lindman, B.; Wennerström, H.; Gustavsson, H.; Kamenka, N.; Brun, B.
Pure Appl. Chem. 1980, 52, 1307.
(30) Clifford, J.; Pethica, B. A. Trans. Faraday Soc. 1964, 60, 1483.
(31) Muller, N.; Birkhahn, R. H. J. Phys. Chem. 1967, 71, 957.
(32) Drakenberg, T.; Lindman, B. J. Colloid Interface Sci. 1973, 44, 184.
(33) Stothers, J. B. Carbon-13 NMR Spectroscopy; Academic Press: New
York, 1972.
(34) Batchelor, J. G.; Prestegard, J. H.; Cushley, R. J.; Lipsky, S. R. Biochem.
Biophys. Res. Commun. 1972, 48, 70.
(35) Persson, B. O.; Drakenberg, T.; Lindman, B. J. Phys. Chem. 1976, 80,
2124.
11440 Langmuir, Vol. 23, No. 23, 2007 Lundberg et al.

cmono cagg
δobs ) δmono + δ (6)
c c agg
where δmonoand δmic are the chemical shifts of amphiphile
molecules present as monomers or residing in micelles, respec-
tively, while cmono, cmic, and c are as defined above.
Figure 7 shows the change in δobs with c-1 for selected carbons
in A and DTAB. The data for DTAB show the expected
dependence on c-1, i.e. a distinct kink at cmc-1, a downfield
shift on micelle formation, and a larger change in shift for the
midchain carbons than for those in either end of the alkyl chain.32
For A, the change in chemical shift for all alkyl carbons, except
those neighboring the ester oxygens, is also downfield, but the
change is significantly smaller than that observed for DTAB.
Furthermore, the shifts change with c-1 in a more gradual fashion.
If the chemical shifts are assumed to exclusively reflect the average
chain conformation, these observations strongly support an Figure 7. Change in 13C NMR shifts relative to the shift for
assembly structure for A that differs significantly from that of monomeric amphiphile versus the reciprocal of the concentration
for selected peaks from A or DTAB. The peaks are from the terminal
conventional micelles and that the structure of the assemblies methyl (O), two different “midchain” methylene groups (0, ]), the
changes gradually with concentration. Consistent with the results methylene groups neighboring the ester oxygens (+), and the carbonyl
from the NMR diffusion experiments, the 13C NMR data also carbons (4) of A, or the terminal methyl (b), the midchain methylene
indicate that self-assembly of A begins at a concentration well groups (9), or the C10 methylene group (2) of DTAB. ∆δ ) δobs
below 30 mM, i.e., the previously suggested aggregation - δlow, where δlow is the observed chemical shift at the lowest
concentration (30 mM corresponds to a c-1 of 0.033 mM-1).3 investigated concentration, i.e., 5 mM (not shown). The ∆δ values
of the carbonyl carbons (4) of A for the four most concentrated
The signals from the carbonyl carbons, as well as those from samples (i.e., at low c-1) have been omitted from the plot. These
the carbons adjacent to the ester groups, show upfield shifts with decrease monotonically with increasing concentration to a value of
increasing concentration. An upfield shift can be related to an -1.9 for the 100 mM sample (i.e., at c-1 ) 0.01 mM-1).
average decrease in hydrogen bonding to the oxygen atoms,33
indicating at least a partial loss of ester-water interactions when
A form assemblies. However, the change in shift from the lowest
to the highest concentration of A, i.e. approximately -2 ppm,
is significantly smaller than, for instance, the difference of more
than 11 ppm in the carbonyl carbon shift of acetone in water and
cyclohexane.33 Hence, these results do not suggest that the ester
bonds are transferred to a “dry” hydrocarbon environment when
A assembles and thus support the interpretation of the diffusion
data. Interestingly, the carbonyl peaks (4 in Figure 7) begin to
shift from δlow at a lower concentration (i.e., a higher c-1) than
the other signals. This observation suggests that there is a loss
of hydrogen bonding at a concentration lower than where the
average molecular conformation is affected and further supports Figure 8. Change in 13C NMR shifts relative to the shift for
that the character of the aggregates changes with concentration. monomeric amphiphile versus the reciprocal of the total concentration
Figure 8 compares the changes in δobs for selected peaks of for selected peaks from A or DTAB when present as either the only
amphiphile (solid symbols) or in an equimolar mixture of A and
A or DTAB when the compounds are present as either the only DTAB (open symbols). For A, the peaks are from a midchain
solute or in an equimolar mixture of the two. The observed changes methylene group (9, 0) and the carbonyl carbons (2, 4), and for
in δobs with c-1 are consistent with the results from the NMR DTAB, the peaks are from the midchain methylene groups (b, O).
diffusion measurements in that they support that mixed aggregates ∆δ values for the carbonyl carbons for A (2, 4) for the most
are formed (by equivalent arguments). One can also see that at concentrated samples (i.e., at low c-1) have been omitted from the
higher concentrations (i.e., lower c-1) the average chain plot. These decrease monotonically with increasing concentration
to values of -1.9 and -2.0 for the 100 mM samples (i.e., at c-1 )
conformations for the two compounds seem to be similar to 0.01 mM-1) where A is present as the lone solute and where it is
those experienced in the respective single-amphiphile aggregates. admixed with DTAB, respectively.
The most interesting observation is that for some of the midchain
carbons in A (9 and 0 in the plot), at intermediate total amphiphile ConductiVity Measurements. Determining the concentration
concentrations, the shifts deviate from δlow to upfield shifts when dependence of the specific conductivity, κ, of an aqueous solution
the substance is present in the mixture. This finding suggests a of an ionic surfactant gives values of both the cmc and the degree
decrease in the ratio of trans to gauche conformations at certain of counterion dissociation, R, for the micelles. The cmc is
carbon positions and can be explained by the tails of A being, generally obtained from the intersection of two essentially straight
on average, more bent in an intermediate concentration range lines and a good estimate of R can often be obtained by simply
but adopting a more extended conformation at higher concentra- calculating the ratio of the slopes above and below the cmc. The
tions. In terms of aggregate structure, the observed data are results from a conductivity study on A and DTAB are shown in
consistent with a picture in which the first molecules of A that Figure 9.
participate in assemblies are oriented with their ester groups The plot for DTAB shows the expected appearance with a
close to the surface of the clusters, but are forced into the core distinct break at about 16 mM and a change in slope that
of the assemblies with an increasing total concentration, when corresponds to an R of about 0.3. Both the cmc and the counterion
also the fraction of A in the aggregates increases. dissociation are in good agreement with values found in the
An Ester-Modified Amphiphile in Aqueous Systems Langmuir, Vol. 23, No. 23, 2007 11441

Table 2. Values of the Lattice Parameter (a) Obtained from the

SAXD Experiments Together with Values of the Apolar Volume
Fraction (Φap) and Cylinder Radii (R) Calculated for the
Hypothetical Cases Where All (ΦXTAB+A, RXTAB+A) or None
(ΦXTAB, RXTAB) of the Fraction of A Is Included in the
Cylindrical Aggregates That Make Up the Hexagonal Phase (see
Text for Further Details)a
a RXTAB+A RXTAB
sample (nm) ΦXTAB+A (nm) ΦXTAB (nm)
DTAB 60/0 4.16 0.44 1.45 0.44 1.45
DTAB 60/10 4.11 0.44 1.43 0.40 1.36
DTAB 60/20 4.14 0.45 1.45 0.35 1.29
DTAB 60/30 4.19 0.45 1.47 0.31 1.23
DTAB 60/40 4.21 0.45 1.49 0.27 1.14
DTAB 70/0 3.96 0.51 1.48 0.51 1.48
DTAB 70/10 3.95 0.51 1.48 0.46 1.40
DTAB 70/20 3.98 0.52 1.50 0.41 1.34
DTAB 70/30 4.00 0.52 1.51 0.36 1.26
DTAB 70/40 4.04 0.53 1.54 0.31 1.18
DTAB 70/50 4.13 0.53 1.58 0.26 1.11
DeTAB 60/0 3.53 0.41 1.19 0.41 1.19
DeTAB 60/10 3.60 0.42 1.22 0.37 1.15
DeTAB 60/20 3.64 0.43 1.25 0.33 1.10
TTAB 60/0 4.74 0.46 1.68 0.46 1.68
TTAB 60/10 4.62 0.46 1.64 0.42 1.56
TTAB 60/20 4.61 0.46 1.64 0.37 1.47
TTAB 60/30 4.65 0.46 1.66 0.33 1.40
TTAB 60/40 4.66 0.46 1.67 0.28 1.30
Figure 9. Specific conductivity as a function of concentration of
TTAB 60/50 4.61 0.47 1.65 0.24 1.17
A (a) or DTAB (b). TTAB 60/60 4.64 0.47 1.66 0.19 1.06
a
literature.15,17,36 For A, the dependence of κ on concentration is The samples are named according to the following system: in a
sample XTAB x/y, the total concentration of amphiphiles, i.e., of A +
much weaker. The slight change in slope at about 25 mM
XTAB, is x wt%, while A constitutes y wt% of the total amount of
corresponds to an R of about 0.9, which indicates a very limited amphiphiles.
condensation of ions when the assemblies are formed. The plot
for A can be compared to the results from conductivity studies concentration of A suggests that whatever the actual structure
on surfactant aggregation in situations where the cooperativity of the assemblies, these are not well-defined enough to be forced
is impaired, for instance in aqueous solutions containing ethylene into an ordered structure, even at very high volume fractions.
glycol37 or urea.38 Under such conditions one obtain plots that Clusters of A probably begin to overlap at about 40 wt % of
are similar to that obtained for A. In conclusion, the findings amphiphile, the concentration above which an increase in viscosity
from the conductivity measurements on A are consistent with is observed. At the highest concentrations, the assemblies may
those from the previously discussed experiments, in that they merge into an essentially homogeneous solution, and it is possible
suggest that loosely assembled aggregates are formed at a that hydrophobic interactions and/or water-mediated hydrogen
concentration around 25-30 mM. bonds between the monomers contribute to the increase in
Structure in Highly Concentrated Solutions of A. Several of viscosity.
the results presented above suggest that the assemblies of A in Characterization of Liquid Crystalline Samples. Optical
dilute solutions have a “looser” structure than conventional Microscopy. Anisotropic liquid crystalline phases (e.g., lamellar
micelles. At higher concentrations (500 mM, Φ ) 0.19), DHMDS and hexagonal phases) are optically birefringent and show
data correspond to an apparent aggregate size similar to that characteristic textures when placed between crossed polarizers
expected for conventional spherical micelles. When the am- and viewed in an optical microscope.6 Hence, this technique
phiphile concentrations is increased above 500 mM, the decrease provides a convenient method for an initial characterization of
in D of A is quite limited and can be most probably accounted the liquid crystalline phases in an aqueous surfactant system. A
for by obstruction effects; any notable growth of the aggregates screening of the concentrated aqueous mixtures of A with DeTAB,
should show a larger effect on the observed self-diffusion DTAB, or TTAB shows that all liquid crystalline samples give
coefficient.39 Hence, it is reasonable to assume that A remains birefringence patterns typical of hexagonal structures.
in roughly spherical clusters up to the point (with regard to Small-Angle X-ray Diffraction (SAXD). In order to obtain
increasing concentration) where these are forced into close contact. quantitative information on the liquid crystalline samples, the
At very high amphiphile volume fractions, the self-diffusion SAXD technique was employed. The diffractograms of all liquid
data are difficult to evaluate in terms of aggregate size and shape, crystalline samples indicate hexagonal structures, consistent with
because when the aggregates are in close contact, there is a the optical microscopy results. For hexagonal structures, the
significant contribution from molecular diffusion between the relative positions of the diffraction peaks obey the relationship
assemblies to the observed diffusion coefficient.39 However, the 1:x3:2:x7:3..., and the lattice constant a, i.e., the distance
simple fact that no liquid crystalline phases are formed at any between the centers of adjacent cylinders, can be obtained from
the position of the first diffraction peak, q*, via eq 7:
(36) Carpena, P.; Aguiar, J.; Bernaola-Galvan, P.; Ruiz, C. C. Langmuir 2002,
18, 6054.
4π
(37) Carnero Ruiz, C. Colloid Polym. Sci. 1999, 277, 701. q* ) (7)
(38) Carnero Ruiz, C. Colloids Surf. A 1999, 147, 349.
(39) Jonströmer, M.; Jönsson, B.; Lindman, B. J. Phys. Chem. 1991, 95, 3293.
ax3
11442 Langmuir, Vol. 23, No. 23, 2007
It can be seen in Table 2 that when an increasing amount of
DTAB is exchanged for A, at a constant total amphiphile
concentration, a shows a shallow minimum for both the DTAB
60/y and the DTAB 70/y series. For the DeTAB 60/y series, a
increases with an increasing fraction of A, whereas it is essentially
constant or decreases slightly for TTAB 60/y.
Given the differences in the aqueous behavior of A as compared
to that of a conventional surfactant, it is difficult to predict the
degree of inclusion of A in the cylindrical aggregates that
constitute the hexagonal phase, as well as its average location
and conformation in these. In order to obtain some clues on the
role of A in the hexagonal structures, one can consider the two
hypothetical cases where (1) all of the present A is included in
the cylinders and (2) all A resides in the continuous aqueous
domains, i.e., that the cylinders are made up solely of DTAB.
For the two cases, the volume fractions of apolar domains, Φap,
are denoted ΦXTAB+A and ΦXTAB, respectively. The procedure
for estimating the Φap values is given in the Experimental
Procedures, and the obtained values are presented in Table 2.
The hypothetical values of Φap can then, together with the
experimental values of a, be related to their corresponding
aggregate radii, R, via eq 8, which can be derived by geometric
reasoning:
( )
x3 1/2
R)a Φ (8)
2π ap
The calculated values of R are presented in Table 2.
In samples containing DTAB as the only amphiphile, R is
1.45 and 1.48 nm, respectively, for amphiphile concentrations
of 60 and 70 wt %. These values are, as expected, slightly smaller
than the extended length of a C12H25 chain, which is 1.67 nm.7
When a fraction of DTAB is exchanged for A, and all of this
is included in Φap, the change in R essentially reflects the change
in a, since ΦDTAB+A is practically constant within each sample
series. If A is excluded from Φap, on the other hand, there is, as
expected, a considerable decrease in R with an increasing fraction
of A. For instance, when 40% of the DTAB is exchanged for
A, cylinder radii of 1.49 and 1.54 are obtained for 60 and 70 wt
% total amphiphile concentration, respectively, if ΦDTAB+A is
used, whereas the corresponding values with ΦDTAB are 1.14 and
1.18. It is highly unlikely that DTAB would arrange itself into
cylinders with such small radii as the two latter, since this would,
due to geometrical constraints, require substantial exposure of
its hydrophobic tails to water. For this reason we conclude that,
most likely, a significant fraction of the A participates in the
cylinders that form the hexagonal liquid crystalline phase. The
same conclusions hold for the DeTAB and TTAB systems.
With the above presented calculations on hand it is fair to
exclude the possibility that a major part of A is molecularly
dissolved in the aqueous domains of the liquid crystals. This
means that the differences in the variation in a with composition
Lundberg et al.
among the sample sets mainly reflect a difference in the average
conformations of A and the XTAB compounds when they
participate in the cylinders. If a conventional surfactant of the
same size as A was used instead of this, it should swell the
cylinders in hexagonal phases of all the XTAB surfactants that
are used in this study. As can be seen in Table 2, A do swell the
cylinders in a hexagonal phase of DeTAB, but have a very limited
effect on the cylinder radii for DTAB or TTAB, both of which
have significantly shorter tails than A. This observation can be
explained by A having, on average, a larger surface exposed to
the polar surroundings per unit volume when present in the
hexagonal phase than that expected for a conventional surfactant
of similar size. Such a difference in exposed surface per volume
can be rationalized by a larger degree of chain bending in A,
most probably in order to allow for better contact between the
ester groups and the solvent. This means that the findings from
the SAXD study are in line with the above-discussed results on
the mixed aggregates that form in isotropic solution and hence
give further evidence for the strong influence on the aggregation
behavior of A that is exerted by the driving force for ester bond-
water interactions.
Conclusions
The ester-containing amphiphile A does not exhibit typical
surfactant behavior. When it is present as the only solute in a
binary aqueous system, it forms neither conventional micelles
nor liquid crystalline phases. However, there is clear evidence
that A assembles with lower cooperativity into some kind of
loose clusters at concentrations above 25-30 mM that cannot
be characterized in greater molecular detail at this time. The
apparent hydrodynamic radius of the assemblies, as determined
from NMR diffusometry, increases with concentration up to the
order of what would be expected had they been conventional
spherical micelles, but due to the uncertainty in their structure
and the indications of strong hydration, the apparent size may
not directly reflect the physical size. At concentrations of A
above approximately 1000 mM (40 wt %), there is a notable
increase in the viscosity of the solutions. The increased viscosity
might be explained by the formation of a network-like structure.
When A is mixed with DTAB in solution, the two amphiphiles
form mixed assemblies, the structure of which varies with the
total amphiphile concentration. In concentrated mixtures with
alkyltrimethylammonium surfactants, A participates in the
formation of hexagonal liquid crystalline phases, even when it
constitutes a significant fraction of the total amphiphile content.
Acknowledgment. This work was financially supported by
an NIH grant to F.M.M. (D.L., A.L.G., F.M.M.) and the Chalmers
Foundation (J.U.). The authors are grateful to Dr. Shaoxiong Wu
for technical assistance with the setup of the NMR spectrometer.
Prof. Björn Lindman is acknowledged for valuable discussions.
LA700430U