Synthesis and vesicle formation from hybrid bolaphile/ amphiphile ion-pairs. Evidence of membrane property modulation by molecular design

Article abstract of DOI:10.1021/jo980315l

Four new hybrid (bolaphile/amphiphile) ion-pairs were synthesized. Electron microscopy indicated that each of these forms bilayer membranes upon dispersion in aqueous media. Membrane properties have also been examined by differential scanning calorimetry, microcalorimetry, temperature-dependent fluorescence anisotropy measurements, and UV-vis spectroscopy. The T_m values for the vesicular 1,2,3,4, and 5 were 38,12,85,31.3, and 41.6 °C, respectively. Interestingly the T_m values for 1 and 3 were found to depend on their concentration. The entrapment of small solute and the release capability have also been examined to demonstrate that these bilayers form enclosed vesicles. X-ray diffraction of the cast films has been performed to understand the nature and the thickness of these membrane organizations. The membrane widths ranged from 33 to 47 A. Finally, the above observations have been analyzed in light of the results obtained from molecular modeling studies. Thus we have demonstrated that membrane properties can be modulated by simple structural changes at the amphiphile level. It was shown that by judicious incorporation of central, isomeric, disubstituted aromatic units as structural anchors into different bolaphiles, one can modulate the properties of the resulting vesicles.

Full text of DOI:10.1021/jo980315l

7640
J . Org. Chem. 1998, 63, 7640-7651
Syn th esis a n d Vesicle F or m a tion fr om Hybr id Bola p h ile/
Am p h ip h ile Ion -P a ir s. Evid en ce of Mem br a n e P r op er ty
Mod u la tion by Molecu la r Design
Santanu Bhattacharya,*,† Soma De, and M. Subramanian
Department of Organic Chemistry, Indian Institute of Science, Bangalore 560 012, India
Received February 19, 1998
Four new hybrid (bolaphile/amphiphile) ion-pairs were synthesized. Electron microscopy indicated
that each of these forms bilayer membranes upon dispersion in aqueous media. Membrane
properties have also been examined by differential scanning calorimetry, microcalorimetry,
temperature-dependent fluorescence anisotropy measurements, and UV-vis spectroscopy. The Tm
values for the vesicular 1, 2, 3, 4, and 5 were 38, 12, 85, 31.3, and 41.6 °C, respectively. Interestingly
the Tm values for 1 and 3 were found to depend on their concentration. The entrapment of small
solute and the release capability have also been examined to demonstrate that these bilayers form
enclosed vesicles. X-ray diffraction of the cast films has been performed to understand the nature
and the thickness of these membrane organizations. The membrane widths ranged from 33 to 47
Å. Finally, the above observations have been analyzed in light of the results obtained from molecular
modeling studies. Thus we have demonstrated that membrane properties can be modulated by
simple structural changes at the amphiphile level. It was shown that by judicious incorporation
of central, isomeric, disubstituted aromatic units as structural anchors into different bolaphiles,
one can modulate the properties of the resulting vesicles.
In tr od u ction
branes from a structurally defined lipid or its analogues
are currently receiving widespread interest. Early pio-
neering studies in this direction demonstrated the fea-
sibility of exploring various membrane-forming properties
from molecules of totally synthetic, nonbiological origin.9
Since then, investigations by different research groups10
including ourselves11 on synthetic lipids and amphiphiles
have enriched the understanding in this area consider-
ably.
It is known that by mixing single chain amphiphiles
of same charge in water, one can obtain mixed micelles.12
However, when equivalent amounts of single chain
amphiphiles of opposite charges are mixed, lamellar or
bilayer type aggregates are produced.13 Presumably the
electrostatic interactions at or near the headgroup region
Lipids and related amphiphiles are important building
blocks of biological membranes.1 Several naturally oc-
curring lipids have been isolated from sources as diverse
as plants, mammalian tissues, bacteria and even marine
organisms.2-5 However, due to the extreme difficulties
associated with their isolation, effective purification, and
characterization, most of the biophysical approaches to
understanding membrane behavior at the molecular level
have been largely devoted to synthetic analogues of lipids.
In addition, synthetic amphiphiles that form micellar or
vesicular aggregates in aqueous media are also attractive
for their potential applications in drug delivery,6 cataly-
sis,7 and the synthesis of nanomaterials.8 Consequently,
attempts to correlate the physical properties of mem-
† Fax: 91-080-334-1683. E-mail: sb@orgchem.iisc.ernet.in. Also at
the Chemical Biology Unit, J awaharlal Nehru Centre for Advanced
Scientific Research, J akkur, Bangalore 560 064, India.
(1) (a) Merz, K. M., J r.; Roux, B. Biological Membranes. A Molecular
Perspective from Computation and Experiment; Birkha¨user, Boston,
1996. (b) Slater, J . L.; Huang, C.-H. In The Structure of Biological
Membranes; Yeagle, P., Ed.; CRC Press: Boca Raton, FL, 1992; pp
175-210. (c) Marsh, D. CRC Handbook of Lipid Bilayers; CRC Press:
Boca Raton, FL, 1990.
(2) (a) Grunner, S. M.; J ain, M. K. Biochim. Biophys. Acta 1985,
818, 352. (b) Bittman, R.; Clejan, S.; J ain, M. K.; Deroo, P. W.;
Rosenthal, A. F. Biochemistry 1981, 20, 2790. (c) Friedberg, S. J .;
Holpert, M. J . Lipid. Res. 1978, 19, 57.
(3) Langworthy, T. A. In The Bacteria. The Treatise on Structure
and Function; Woese, C. R., Wolfe, R. S., Eds.; Academic: Orlando,
FL, 1985; pp 459-497. De Rosa, M.; Gambacorta, A.; Gliozzi, A.
Microbiol. Rev. 1986, 50, 70.
(4) Setter, K. O. Nature 1982, 300, 258.
(5) Brock, T. D.; Brock, K. M.; Belley, R. T.; Weiss, R. L. Arch.
Mikrobiol. 1972, 84, 54.
(9) (a) Kunitake, T. Angew. Chem., Int. Ed. Engl. 1992, 31, 2137.
(b) Fendler, J . H. Science 1984, 223, 888. (c) Fendler, J . H. Membrane
Mimetic Chemistry; Wiley: New York, 1982.
(10) (a) Sommerdijk, N. A. J . M.; Hoeks, T. H. L.; Synak, M.; Feiters,
M. C.; Nolte, R. J . M.; Zwanenburg, B. J . Am. Chem. Soc. 1997, 119,
4338. (b) Menger, F. M.; Ding, J . Angew. Chem., Int. Ed. Engl. 1996,
35, 2137. (c) Song, X.; Perlstein, J .; Whitten, D. G. J . Am. Chem. Soc.
1995, 117, 7816. (d) Murillo, O.; Watanabe, S.; Nakano, A.; Gokel, G.
W. J . Am. Chem. Soc. 1995, 117, 7665. (e) Lee, Y.-S.; Yang, J .; Sisson,
T. M.; Frankel, D. A.; Gleeson, J . T.; Aksay, E.; Keller, S. L.; Gruner,
S. M.; O’Brien, D. F. J . Am. Chem. Soc. 1995, 117, 5573. (f) Moss, R.
A.; Li, G.; Li, J .-M. J . Am. Chem. Soc. 1994, 116, 805. (g) Thompson,
D. H.; Wong, K. F.; Humphry-Baker, R.; Wheeler, J . J .; Kim, J .-M.;
Rananavare, S. B. J . Am. Chem. Soc. 1992, 114, 9035. (h) Wagenaar,
A.; Rupert, L. A. M.; Engberts, J . B. F. N.; Hoekstra, D. J . Org. Chem.
1989, 54, 2638.
(11) (a) Bhattacharya, S.; Snehalatha, K. J . Org. Chem. 1997, 62,
2198. (b) Bhattacharya, S.; De, S.; George, S. K. Chem. Commun. 1997,
2287. (c) Bhattacharya, S.; Mandal, S. S. Biochim. Biophys. Acta
(Biomembranes) 1997, 1323, 29. (d) Bhattacharya, S.; Haldar, S.
Biochim. Biophys. Acta (Biomembranes) 1996, 1283, 21. (e) Ghosh, S.;
Easwaran, K. R. K.; Bhattacharya, S. Tetrahedron Lett. 1996, 37, 5769.
(f) Bhattacharya, S.; Haldar, S. Langmuir 1995, 11, 4748. (g) Bhat-
tacharya, S.; Subramanian, M.; Hiremath, U. Chem. Phys. Lipids 1995,
78, 177. (h) Bhattacharya, S.; De, S. J . Chem. Soc., Chem. Commun.
1995, 651.
(6) Ringsdorf, H.; Schlarb, B.; Venzmer, J . Angew. Chem., Int. Ed.
Engl. 1988, 27, 113.
(7) (a) Bhattacharya, S.; Snehalatha, K.; George, S. K. J . Org. Chem.
1998, 63, 27. (b) Bhattacharya, S.; Snehalatha, K. Langmuir 1995,
11, 4653. (c) Scrimin, P.; Tecilla, P.; Tonellato, U. J . Am. Chem. Soc.
1992, 114, 5086.
(8) (a) Schnur, J . M. Science 1993, 262, 1669. (b) Whitesides, G. M.;
Mathias, J . P.; Seto, C. T. Science 1991, 254, 1312.
(12) Tanford, C. The Hydrophobic Effect: Formation of Micelles and
Biological Membranes; Wiley: New York, 1980.
10.1021/jo980315l CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/03/1998

Hybrid Bolaphile/Amphiphile Ion-Pairs
J . Org. Chem., Vol. 63, No. 22, 1998 7641
and hydrophobic association of the nonpolar chains drive
the formation of “tight” complexes which eventually
develop into lamellae or bilayers. However, so far these
studies have been confined to amphiphilic surfactants
only and the corresponding mixtures of monopolar am-
phiphile and bipolar bolaphiles14 have not been examined
for vesicle formation. Synthetic bolaphiles are mimics
of archaebacterial membranes which can sustain extreme
physiological conditions (85 °C, high pH) and form
ultrathin “monolayer” membranes.15 Due to these ex-
ceptional properties, much interest is growing toward the
development of their mimics to sort out their unusual
tolerance to external stimuli compared to the membrane
present in other organisms. We thought that it would
be interesting in this context to examine ion-pairs from
a mixture of oppositely charged amphiphilic and bo-
laphilic molecules.
properties of bilayer assemblies, we synthesized five
systems, as shown in Scheme 1. Three of these com-
pounds are composed of the following molecular modules.
First, the bolaphilic counterparts in these ion-pairs
contain central aromatic units from which two lipophilic
chains are connected and the ends of these chains contain
carboxylate residues. Second, the central aromatic units
are isomeric in relation so that disubstituted (1,2-, 1,3-,
1,4-) aromatic units can act as structural anchors in these
bolaphiles. Thus, depending on the position of the
covalent linkage of the undecanoate chains to the central
aromatic core, these lipophilic chains propagate at direc-
tions dictated by the manner in which they are connected
In a preliminary communication,16 we have demon-
strated that one can generate vesicles via ion-pairing of
cationic amphiphile with bisanionic bolaphiles. In this
paper we present in detail, the synthesis, the vesicle
formation, the thermotropic behavior, and the micropo-
larity of a series of bis(hexadecyltrimethylammonium)-
phenyl-1,2-, 1,3-, and 1,4-di(oxyundecanoate) amphiphiles
1-3 in which the alkyl chains of the bolaamphiphilic
counterions originate from the different isomeric posi-
tions of the central phenyl ring. We also present the dye
entrapment abilities of these vesicles and the results of
the permeability experiments with these vesicles. Cast
film X-ray diffraction studies on the multilayers were also
performed to investigate the possible nature of the bilayer
organization in these membranes. To put our results into
proper perspective, we also examined the vesicular
properties of the corresponding amphiphile (monopolar)
anion/cation pair cetyltrimethylammonium (CTA) 11-
phenoxyundecanoate 4 and CTA palmitate, 5. It is
evident that by judicious incorporation of central, iso-
meric, disubstituted aromatic units as structural anchors
into different bolaphiles one can modulate the properties
of the resulting vesicles. Remarkably the phase transi-
tion temperatures and the permeabilities of these ve-
sicular aggregates were found to be highly influenced by
the geometry of the bolaamphiphilic counterion. Finally,
we provide a rationale for the effects of such shape
variation of the bolaform counterion by considering
energy-minimized structures obtained by molecular mod-
eling studies for each amphiphile.
-
with the respective aromatic ring. Third, the end CO2
residues are finally ion-paired with amphiphilic cetyl-
trimethylammonium (CTA+) cations.
The synthesis of the ion-paired systems 1-4 began
with the preparation of three new aromatic diacids (8a ,
8b, and 8c) and one phenoxy acid (8d ) (Scheme 1). These
were accomplished by the conversion of catechol, resor-
cinol, quinol, and phenol to the corresponding ethers upon
treatment with ethyl 11-bromoundecanoate or the cor-
responding acids generally in high yields. The corre-
sponding esters were hydrolyzed in the presence of
ethanolic KOH. Then the ion-paired systems (1-4) were
synthesized in a key step as described below. First, by
the passage of a methanolic solution of freshly recrystal-
lized cetyltrimethylammonium bromide (CTAB) (1 equiv)
through ion-exchange resin (Amberlite IRA-900, OH-
form), CTA+OH- was generated. Then for the prepara-
tion of 1-3, to each of the reaction mixtures containing
1 equiv of CTA+OH-, 0.5 equiv of diacid 8a , 8b, or 8c,
respectively, in dry methanol was added. The reaction
mixtures were stirred at room temperature for 24 h under
nitrogen inlet, and the solvents were removed by rotary
evaporation to obtain solid residues from each of the
reaction mixtures. For the synthesis of 4, similar reac-
tion conditions were employed using 1 equiv of monoacid
8d in methanol. All the compounds isolated as solids
after evaporation of methanol were recrystallized several
times from EtOAc/MeOH (10:1), and each ion-pair gave
expected analytical and spectroscopic data consistent
with their structures, cf. Experimental Section. Due to
their strong hygroscopic nature, only freshly recrystal-
lized ion-pairs were employed for vesicle preparation and
subsequent characterization by different physical tech-
niques.
Resu lts a n d Discu ssion
Molecu la r Design a n d Syn th esis. To realize bilayer
like assemblies from the hybrid ion-paired bolaphile/
amphiphiles and to explore whether subtle structural
differences in the starting ion-pairs are reflected in the
(13) (a) Fukuda, H.; Kawata, K.; Okuda, H.; Regen, S. L. J . Am.
Chem. Soc. 1990, 112, 1635. (b) Kaler, E. W.; Murthy, A. K.; Rodrigueg,
B. E.; Zasadzinski, J . A. N. Science 1989, 245, 1371.
(14) Bolaphiles (bolaamphiphiles), in contrast to their monopolar
counterparts (single chain/single polar headgroup amphiphile), consist
of two hydrophilic headgroups linked to the hydrophobic core in R,ω-
position. See for reviews: (a) Escamilla, G. H.; Newkome, G. R. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1937. (b) Fuhrhop, J . H.; Bach, R. In
Advances in Supramolecular Chemistry: Gokel, G. W., Ed.; J AI
Press: Greenwich, CT, 1992; Vol. 2, pp 25-63.
(15) (a) De Rosa, M.; Gambacorta, A. J . Mol. Biol. 1985, 182, 131.
(b) Gliozzi, A.; Rolandi, R.; de Rosa M.; Gambacorta, A. J . Membr. Biol.
1983, 75, 45.
(16) Bhattacharya, S.; De, S. J . Chem. Soc., Chem. Commun. 1996,
1283.
P a ck in g P a r a m eter Ca lcu la tion s. It remains an
intriguing question why certain amphiphiles form vesicles
and others form micellar or other nonvesicular ag-
gregates. This question has been theoretically addressed
in the literature by Israelachvilli,17 in particular. An

7642 J . Org. Chem., Vol. 63, No. 22, 1998
Bhattacharya et al.
Sch em e 1a
were used in an effort to reduce ambiguity. The corre-
sponding values for the hydrophobic chain length (l) were
obtained from the equation l ) (1.5 + 1.265n) Å, where
n is the number of carbon atoms. The lengths of the
hydrocarbon segments of these ion-pairs in their fully
extended conformation were also obtained from the
calculation of energy-minimized conformations using the
DISCOVER package of INSIGHT II (version 2.3.5, Bio-
sym Technology Software). The values of hydrocarbon
lengths obtained by either approach were in good agree-
ment.
According to the above hypothesis, when the p value
(v/la) falls between the limiting values of 0.5 and 1.0,
vesicle formation is predicted. We obtained p values of
∼0.6 ( 0.02 for 4 and ∼0.56 ( 0.02 for 1, 2, and 3. These
orders of p values predict vesicle formation and agree
with the experimental observations (see below).
Vesicle F or m a tion . Encouraged by the results of the
calculations of packing parameters, we then decided to
examine the abilities of the ion-pairs toward vesicle
formation by dispersal of these compounds in water. The
aggregates were obtained by modified reverse phase
evaporation method (REV).24 A given ion-pair (2.5 mmol)
was dissolved in 1 mL of CHCl3. To this was added water
(pH 6.8), and the two phase mixture was briefly bath-
sonicated at ambient temperature to produce a stable
emulsion. The organic solvent from that emulsion was
reversibly evaporated under reduced pressure. The
resulting suspension was then freeze-thawed several
times. Then a brief bath-sonication above the phase
transition temperature afforded stable, translucent ve-
sicular aggregates.
a
Conditions. (a) Br(CH2)10CO2Et, anhydrous K2CO3, acetone,
reflux, 30 h, 85-90%; (b) aqueous KOH (5%), reflux, cool, aqueous
acid, 80-85%; (c) i. Br(CH2)10CO2H, NaOH (3.6 M), heat (∼70 °C);
ii. filter, acidification, 84%; (d) n-C16H33N+Me3, OH- (2 equiv for
1, 2, and 3 and 1 equiv for 4), MeOH, 25 °C, stir 24 h, 90-95%.
For the fluorescence, UV-vis spectroscopy, and the
X-ray diffraction studies, the vesicles were generated as
follows. A dry lipid film was hydrated by keeping it in
the presence of water for 12 h. Then the mixture was
dispersed by vortexing for a few minutes followed by 5-6
freeze-thaw treatments. For the trapping experiments,
the vesicles were formed by sonicating the screw-cap vials
containing the required amount of the lipid and water
having the fluorescence probe in a bath sonicator at ∼60
°C or higher for 30 min. Vesicles prepared by either
protocol remained optically stable for several days except
that of 4, as judged by their turbidity measurements at
450 nm, as a function of time.
Electr on Micr oscop y. To confirm the formation of
vesicular aggregates from this new series of synthetic,
ion-paired systems, all the aggregates were examined
under transmission electron microscopy (TEM) using a
J EOL-TEM 200 CX instrument. For the TEM studies,
the vesicles were generated by the REV method as
described in the preceding section except that uranyl
acetate was also included in the water in which the
vesicles were formed. TEM examination of the air-dried,
individual aqueous suspensions of 1-4 dried on carbon-
Formvar coated copper grids revealed the existence of
closed aggregate structures in all the cases (Figure 1).
While, 1, 2, and 3 formed predominantly multilamellar
spherical vesicles having diameters of 500-525 Å, 950-
1000 Å, and 1200-1300 Å, respectively, vesicles from 4,
in contrast, formed predominantly unilamellar spheres
empirical rationale correlating molecular architecture
with possible aggregate morphology has been extended
by Israelachvilli,18 Evans,19 and Ninham.20 According to
this model, the critical packing parameter (p) is related
to the geometric properties of the amphiphilic molecules
by the equation v/la, where v is the volume of the
hydrophobic chains of the amphiphile, l is the length of
the hydrocarbon chain in its fully extended conformation,
and a is the area of cross section occupied by the
headgroup of an amphiphile at the aqueous interface.
Limiting areas of the condensed phases of a number
of ion-paired systems have been experimentally deter-
mined21 from the pressure-area diagram involving mea-
surements in Langmuir balance. From the reported
information, one can obtain an estimate of this cross-
sectional area of the headgroup ion-pair Me3N+‚‚‚-O2C,
as 68 ( 3 Å2. We have used this area of cross section (a)
for calculation of packing parameters of all the ion-pairs
examined in the present study as all of them contain the
Me3N+‚‚‚-O2C ion-pair as headgroup. The amphiphile
tail volumes (v) were obtained from the published22
relationship, v ) (27.4 + 26.9n) Å3, where n is the number
of carbon atoms. The same was also calculated using the
method of adding volume increments.23 Both approaches
(17) Israelachvilli, J . N. Intermolecular and Surface Forces, 2nd ed.;
Academic Press: New York, 1992.
(18) Israelachvilli, J . N.; Mitchell, D. J .; Ninham, B. W. J . Chem.
Soc., Faraday Trans. 1976, 72, 1525.
(19) Mitchell, D. J .; Ninham, B. W. J . Chem. Soc., Faraday Trans.
2 1981, 77, 601.
(20) Evans, D. F.; Ninham, B. W. J . Phys. Chem. 1983, 87, 5025.
(21) (a) Shibata, O. J . Colloid Interface Sci. 1983, 96, 182. (b)
Kaneshina, S.; Nakamura, M.; Matuura, R. J . Colloid Interface Sci.
1983, 95, 87.
(23) Kitaigorodsky, A. I. Molecular Crystals and Molecules; Academic
Press: New York, 1973.
(24) DuZguenes, N.; Wilschut, J .; Hong, K.; Fraley, R.; Perry, C.;
Friend, D. S.; J ames, T. L.; Papahadjopoulos, D. Biochim. Biophys.
Acta 1983, 732, 289.
(22) Tanford, C. J . Phys. Chem. 1972, 76, 3020.

Hybrid Bolaphile/Amphiphile Ion-Pairs
J . Org. Chem., Vol. 63, No. 22, 1998 7643
F igu r e 1. Uranyl acetate stained transmission electron micrographs of vesicles prepared by REV of 1-4: (a) 1, (b) 2, (c) 3, and
(d) 4.
having diameters ranging from 750 to 800 Å. Closer
scrutiny of the micrographs shows that with vesicles of
1 the width of each lamellae ranged from ∼42 to 52 Å.
Similarly vesicles of 3 had a lamellar width of ∼34.5 Å
on average. These results are consistent with the XRD
measurements on the cast films from the aggregates of
the corresponding ion-pairs (see below). However, the
information on the lamellar widths of vesicular 2 and 4
fluorescence probe are widely used due to their simplicity
and convenience.25 A number of probes such as pyra-
nine,26 carboxyfluorescein,27 calcein,28 etc. have been used
to estimate membrane permeability. However, due to
their ionic nature it is possible that any of these could
compete for ion-pairing during such experiments. Due
to this concern we preferred to employ riboflavin, a
neutral water soluble molecule which is strongly fluo-
rescent in its neutral form but becomes nonfluorescent
upon deprotonation (pKa ∼ 10). It can also withstand
low ionic strength and, as a result, can be studied even
in water alone. The usefulness of riboflavin has already
been demonstrated independently by Kunitake,,29 Regen,13b
and Bhattacharya11b for the estimation of entrapment
capacity and for the determination of transmembrane pH
gradients with unrelated vesicle forming surfactants of
diverse surface charges.
were not obvious.
Dyn a m ic Ligh t Sca tter in g. Mean hydrodynamic
diameters were determined by laser light scattering.
Each suspension was generated by the reverse-phase
evaporation method and examined for its ability to
scatter. The mean hydrodynamic diameters were 600 (
50 Å, 1000 ( 70 Å (>97%), 1400 ( 50 Å (>95%), and 800
( 20 Å for 1, 2, 3, and 4, respectively. Minor populations
with suspensions of 2 and 3 (<3 and <5%, respectively)
consisting of larger aggregates (>2000 Å) were also seen.
En tr a p m en t of F lu or escen t Dye in Vesicles. After
having confirmed the formation of vesicles from these
newly developed ion-paired systems, we then decided to
examine whether these vesicles were closed and in that
situation whether these can sustain transmembrane ion
gradients. Such an expectation is based on the fact that
closed vesicles have the capacity to hold hydrophilic
molecules or ions inside their inner aqueous cavity. In
the entrapped situation, these solutes cannot permeate
across the lipidic bilayers freely.
We first generated vesicles by dispersal of 1-5 (con-
centration 1 mg/1 mL) in a bath sonicator at ∼60 °C over
a period of ∼30 min in water (pH 6.8) containing
riboflavin (2 × 10-5 M). Sonication up to 30 min
produced a translucent suspension, the turbidity (400
nm) of which did not change upon additional sonication.
(25) Brunner, J .; Graham, D. E.; Hauser, H.; Semenza, G. J . Membr.
Biol. 1980, 57, 133.
(26) Kano, K.; Fendler, J . H. Biochim. Biophys. Acta 1978, 509, 289.
(27) (a) Neumann, R.; Ringsdorf, H.; Patton, E. V.; O’Brien, D. F.
Biochim. Biophys. Acta 1987, 898, 338. (b) Weinstein, J . N.; Yoshikami,
S.; Henkart, P.; Blumenthal, R.; Hagins, W. A. Science 1977, 195, 489.
(28) Matzuzaki, K.; Harada, S.; Funakoshi, N.; Fujii, K.; Miyajima,
K. Biochim. Biophys. Acta 1991, 1063, 162.
The membrane permeability can be measured with
suitable water soluble markers. Although methods based
on radiochemical, redox, or enzymatic techniques are
well-documented in literature, method utilizing the
(29) Kunitake, T.; Okahata, Y.; Yasunami, S. Chem. Lett. 1981,
1397.

7644 J . Org. Chem., Vol. 63, No. 22, 1998
Bhattacharya et al.
Ta ble 1. Ribofla vin En tr a p m en t a n d Estim a ted OH-
P er m ea bility in Differ en t Bila yer Mem br a n es fr om
Ion -P a ir ed System s 1-4 a t 25 °Ca
lipid concn
(mM)
%
103kperm
108Pc,d
(cm/s)
lipid
entrapmentb
(s-1
)
1
2
3
4
0.96
0.96
0.96
1.78
e
2.28
0.82
e
f
3.6
1.9
a
The time-dependent decay of the fluorescence intensity at 514
nm was monitored as a function of time. The decay process
followed apparent pseudo-first-order kinetics. The rate constant
values represent averages of three independent experiments, and
the reproducibility was (4%. In each case, the extent of fluores-
cence decay was followed beyond 90% and the decay followed an
b
apparent pseudo-first-order rate constant. The entrapment of
riboflavin given in percentages has been corrected for an apparent
absorption of marker molecules in each case. c Calculated on the
basis of diameters obtained in transmission electron microscopy.
d
Bilayer thicknesses have been obtained from the X-ray diffraction
studies. We also interpret the time-dependent loss of fluorescence
intensity as a permeation-limited deprotonation of riboflavin
entrapped in the internal aqueous compartment of OH-. e Vesicles
did not entrap any dye. f The fluorescence loss in this case was
“instantaneous”.
assuming the time-dependent loss of riboflavin fluores-
cence as a rate-limiting permeation of exovesicular OH-
into the vesicle’s interior aqueous compartment, we
calculate P to be ∼1.9 × 10-8 cm/s for the permeation
constant of OH- toward vesicles at pH 10.2.
F igu r e 2. OH- permeability profile of riboflavin entrapped
vesicular 3; inset shows the gel filtration profile of the dye
entrapped vesicles of 3 using a Sephadex G-50 column.
The resulting vesicular suspension was loaded onto a
Sephadex G-50 column, and gel filtration was performed
to separate vesicle entrapped riboflavin from the free
riboflavin. A representative gel filtration profile of the
separation of free riboflavin from that bound at the
membrane surface and also the riboflavin that is en-
trapped in vesicles was shown in Figure 2 (inset). Except
with 1 and 4, riboflavin-entrapped vesicles could be
separated out. We suspect that the vesicles of 1 and 4
are so leaky that the encapsulated dye could not be
sustained in the aqueous cavities of these vesicles. This
leakiness could be due to the structural flaws and defects
associated with the bilayers of these ion-paired systems
(see below).
Kin etics of Tr a n sm em br a n e P er m ea tion . The
fluorescence emission due to the membrane-bound and
entrapped riboflavin was measured at 514 nm upon
excitation at 374 nm at 25 ( 0.5 °C. This includes the
dye molecules that were adsorbed on the outer membrane
surfaces and the ones that were entrapped within the
inner water pools. As the pH of the vesicle dispersion
was raised from 6.8 to 10.2, the fluorescence intensity
(514 nm) decreased initially “instantaneously” to about
78 ( 5% of the original value (recorded prior to pH
adjustment). We attribute this immediate ca. 22 ( 5%
loss in the fluorescence intensity to the deprotonation of
riboflavin molecules that are bound at the outer surfaces
of the vesicles. At the exposed exovesicular surfaces,
riboflavin molecules underwent instantaneous ionization
as the pH was increased. The residual fluorescence
intensity, however, disappeared as a function of time for
vesicular 3 (Figure 2). Since the residual loss in fluo-
rescence intensity was time-dependent, we calculated the
half-time of this process (t1/2 ∼3.2 min, kperm ∼3.6 × 10-3
s-1) (Table 1). The fraction of the entrapped riboflavin
was ∼0.82%. Time-dependent loss of the fluorescence
intensity due to the conversion of neutral riboflavin into
its deprotonated form upon pH adjustment (6.8 f 10.2)
demonstrated that it contained an internal aqueous
compartment. Taking a bilayer thickness of ∼35 Å for
vesicles of 3 based on TEM and XRD results and
In the case of vesicular 2, the initial fluorescence
intensity underwent a rapid decay upon pH adjustment
(6.8 f 10.2). However, the subsequent time-dependent
loss of fluorescence in this instance was not observed.
Presumably due to “loose” packing of the monomers of 2
in vesicles, the ionization of the riboflavin molecules that
are adhering at the membrane outer surfaces and the
transbilayer OH- permeation occur simultaneously and
could not be distinguished. Indeed the vesicles of 2 at
25 °C are in their fluid state as indicated by thermal
phase transition studies (see below), and therefore one
can further expect these bilayers to be rather disordered
(and hence leaky) under the conditions of this study.
In contrast to the above situations, vesicular 1 and 4
did not show any evidence of riboflavin entrapment.
Therefore these vesicles do not appear to have the
binding ability at their outer surfaces nor do they have
the capacity to sustain riboflavin molecules in their inner
aqueous pools. It is not, however, obvious why these
vesicles cannot bind or entrap the dye molecules espe-
cially when the TEM studies indicated the existence of
closed aggregates from 1 and 4.
Differ en tia l Sca n n in g Ca lor im etr y. To further
ascertain the presence of lamellar organizations in such
aggregates, the vesicles of the above ion-paired systems
1-5 were examined by differential scanning calorimetry
(DSC). Since the presence of a small amount of impurity
can affect the thermotropic properties and can also
broaden the transition, all the ion-paired systems were
recrystallized prior to calorimetric experiments. DSC
heating and cooling scans were obtained from individual
aqueous dispersions of the surfactants (120 mM, pH 6.8)
in a Perkin-Elmer Model DSC-2C differential scanning
calorimeter. Well-defined, sharp and single phase tran-
sitions were obtained for 3 (∼85 °C) (Figure 3), 4 (∼31
°C), and 5 (∼42 °C) during heating runs. 4 and 5 showed
almost reversible phase transition behavior during heat-
ing as well as cooling cycles. However, vesicles of 3
showed a lower transition (∼61.5 °C) in the cooling run
although a repeat heating run with the same sample of

Hybrid Bolaphile/Amphiphile Ion-Pairs
J . Org. Chem., Vol. 63, No. 22, 1998 7645
F igu r e 3. Gel to liquid-crystalline phase transition of the
vesicles of the ion-paired lipids 1-3: (a) 1, (b) 2, and (c) 3.
Ta ble 2. Gel to Liqu id -Cr ysta llin e Ma in P h a se
Tr a n sition P r op er ties for th e Ion -P a ir ed Am p h ip h iles
1-5 by Differ en tia l Sca n n in g Ca lor im etr ya
F igu r e 4. Fluorescence anisotropy (r) vs temperature (°C)
plots for different lipid vesicles.
Tm (°C)
Ta ble 3. P h a se Tr a n sition Tem p er a tu r es (Tm /°C) of
Vesicu la r 1-5 by Micr oca lor im etr y, F lu or escen ce
An isotr op y, a n d UV Meth od . F lu or escen ce An isotr op y
(r ) a t 15 °C a n d I1/I3 a t 25 a n d 50 °C Ar e Also Given for
Com p a r ison a
b
d
∆Hcal
∆Sc
∆HvH
CUe
lipid heat cool (kcal/mol) (cal/K‚mol) (kcal/mol) (molecules)
1
2
3
4
5
f
f
f
f
85.0 61.2
31.3 29.1
41.6 40.2
4.4
5.34
6.0
12.5
17.55
19.1
330.1
213.7
600.5
75
40
100
main transitionb (°C)
lipid microcal. fluor. ani. UV method (r) at 15 °C 25 °C 50 °C
I1/I3 at
anisotropy
a
b
Concentration of gel was 120 mM. The ∆Hcal values quoted
are the average of the total enthalpy for successive runs. We
1
2
3
4
5
39.1
c
84.9
35.3
41.8
37.0
d
e
30.5
41.0
38.0
12.0
f
g
42.0
0.29
0.17
0.24
0.26
0.3
0.98
0.97
0.93
1.24
1.26
0.93
0.93
0.89
0.93
0.93
estimate that transition enthalpies are accurate to (0.5 kcal mol-1
.
c ∆S values were calculated by dividing ∆Hcal/Tm assuming the
d
phase transition as
a
first-order process. ∆HvH, van’t Hoff
2
enthalpy values, were calculated from the relation ∆HvH) 6.9Tm
/
a
Concentration of vesicular solutions in all the cases was 2.5
∆T1/2, where ∆T1/2 is the spread of temperature at half-height.e The
CU values are the average for successive runs. f No detectable
transition at this concentration (see text for details).
b
mM. Average deviation is (0.5 °C for microcalorimetry, (1 °C
for fluorescence anisotropy method, and (2 °C for UV method.
c No detectable transition. No break in r vs T plot. e Broad
d
transition. f No maxima and break in Aenol/Aketo vs T plot. Not
g
vesicular 3 reproduced the thermogram with Tm at ∼85
°C. Notably the corresponding dispersions from 1 and 2
did not, however, show any detectable transition in the
temperature region 25-90 °C either in the heating or in
the cooling thermograms (Figure 3). The phase transi-
tion temperatures (Tm), calorimetric enthalpies (∆Hcal),
van’t Hoff enthalpies (∆HvH), and cooperativity units are
given in the Table 2.
Micr oca lor im etr y. Examination of the correspond-
ing vesicles at lower concentration (2.5 mM) using
microcalorimetry reproduced the peaks related to Tm for
4 and 5 as obtained under DSC at 120 mM concentration.
However, with 3, the cooperativity was drastically re-
duced at 2.5 mM concentration. Strikingly, the vesicles
of 1 at this concentration (2.5 mM) now gave a pro-
nounced peak at ∼39 °C under microcalorimetry in
contrast to what was observed under DSC at higher (120
mM) concentration. Vesicular 2, however, still did not
show any detectable peak in the temperature range 22-
90 °C, consistent with what was observed with suspen-
sions of 2 at higher concentration (120 mM) under DSC.
These results indicated that thermal meltings of some
of these bilayers do indeed depend on concentrations
which may find their origin in the formation of various
types of aggregates that include SUV, LUV, and MLV’s.
Flu or escen ce An isotr opy Measu r em en ts. To probe
the apparent concentration dependence observed with
determined.
vesicular 1 and 3, we then measured the fluorescence
anisotropy as a function of temperature. This technique
has been extensively used for probing the fluidity of the
interior of the vesicle bilayer and for measuring the
temperature (Tm) at which the bilayer undergoes a
transition from a gel-like to a liquid-crystalline state.30
We have measured the fluorescence anisotropy (r) as a
function of temperature (in the range 10-75 °C) using
1,6-diphenylhexa-1,3,5-triene (DPH), a hydrophobic fluo-
rescent probe which intercalates between the alkyl chains
in the core of the vesicle bilayer.
The anisotropy sensed by DPH in respective mem-
branes is a function of the membrane fluidity. Thus, the
relationship of steady-state fluorescence anisotropy (r)
of DPH and temperature was utilized to evaluate the
relative fluidity and phase transition behaviors of the
lipid membranes (Figure 4). Fluorescence anisotropy
values and phase transition temperatures for 2.5 mM
1-5 are included in Table 3.
As is evident from Figure 4, the vesicular solutions
showed rather well-defined and cooperative phase transi-
(30) (a) Andrich, M. P.; Vanderkooi, J . M. Biochemistry, 1976, 15,
1257. (b) Shinitzky, M.; Barenholz, Y. Biochim. Biophys. Acta 1978,
515, 367.

7646 J . Org. Chem., Vol. 63, No. 22, 1998
Bhattacharya et al.
Ta ble 4. Com p a r ison of Un it La yer Th ick n esses
Obta in ed fr om X-r a y Diffr a ction of Self-Su p p or tin g Ca st
F ilm s a n d Molecu la r Mod elin g Stu d ies
unit layer thickness (Å)
lipid
obsda
calcdb
1
2
3
4
5
46.8
39.1
36.1
33.4
37.9
44.0, 33.0
39.0, 33.7
34.5
40.5
41.0
An optically clear BAA-doped vesicular solution (1 ×
10-3 M) was first prepared. The absorbance maxima due
to the vesicle-doped enol and keto forms of BAA were
indicated at 315 and 250 nm, respectively. The ratio
[Aenol/Aketo] was found to increase sharply during the
phase transition of the bilayers, and the maximum was
seen at Tm.32 Then it decreased as the temperature
increased beyond Tm. The vesicular solutions of 1-3
were studied in the temperature range of 6-70 °C. The
temperature (Tm) related to the melting of “frozen” gel-
like state to “fluid” liquid-crystalline state, as measured
by using tautomerism of vesicle-doped BAA was found
to be comparable with corresponding values obtained by
the determination of fluorescence anisotropy as a function
of temperature. Sharp maxima obtained (Figure 5) from
the plot of [Aenol/Aketo] with temperature for 1 and 2 were
at 38 and 12 °C, respectively. Similarly, upon cooling
they showed a single break within (2 °C of the respective
Tm values observed during heating runs. BAA-doped
vesicles of 3 did not show any maxima in the temperature
range studied (10-60 °C). These findings further confirm
the concentration dependent modulation of the phase
transition properties of the membranes of 1-3.
Micr op ola r ity a n d P h a se Tr a n sition Tem p er a -
tu r e by P yr en e. To secure additional and independent
confirmation of the observed concentration dependence,
we also measured the monomer vibrational intensity
ratio (I3/ I1) of bilayer-doped pyrene as a function of
temperature. This method has been widely used to
determine the apparent Tm from the break of the I3/ I1
vs T plot for various lipid dispersions.33 The relevant
data are summarized in Table 3. In the cases of 4 and
5, I3/ I1 vs T gave sigmoidal plots as expected, i.e., I3/ I1
value sharply increased during the phase transition and
then became constant (1.06-1.07) (figure not shown). In
the cases of 1 and 2, the I3/ I1 vs T plot gave straight
lines with positive slopes in the temperature region 15-
50 °C. No break was obtained. In case of 3, I3/ I1 vs T
gave a broad, noncooperative transition. But strikingly,
the four vesicular solutions that contain an aromatic ring
in their ion-pairs showed different I3/ I1 values at room
temperature (25 °C) which suggested differences in the
micropolarities of the vesicular solutions. Probably the
emission due to pyrene probes interfered with the
aromatic units in lipids which might be responsible for
lack of any break in the I3/ I1 vs T plot related to apparent
Tm of these vesicles.
a
b
As obtained from reflection XRD of cast films. Molecular
lengths of unit layers of ion-pairs as estimated from energy-
minimized CPK models (INSIGHT).
tions in the cases of 1, 4, and 5. In contrast the
corresponding thermograms were broader in the cases
of 2 and 3 and did not show clear evidence of breaks
related to Tm. Further examination of Figure 4 reveals
that the fluorescence anisotropy value (r) at 15 °C was
quite different for 2 (r ) 0.17) than that for 1, 3, 4, and
5 (r ) 0.24-0.29) although there was little difference in
the polarization values after the phase transitions. Since
the anisotropy is directly related to microfluidity of
bilayers, the above observation suggests that the bilayers
of 2 are already in the fluid state at 15 °C. DPH is a
lipophilic molecule and locates only in the hydrophobic
center of the bilayer membrane. Therefore, the decrease
in the anisotropy value indicates the higher movement
of DPH in the bilayers of 2, i.e., the intramonomer van
der Waals interactions between the chains in 2 are
severely “loose”. The bilayers of 1, 3, 4, and 5 are
considerably more rigid at this temperature and should
therefore be in their gel-like solid states. Note, however,
that the thermogram in the case of 3 was really broad
and hence the transition is quite noncooperative at this
concentration (2.5 mM). This observation is consistent
with the results obtained at the same concentration in
microcalorimetry where although the transition was seen,
it was quite poor in cooperativity (not shown). It is
difficult to rationalize all of the above observations.
However, noncooperativity has been observed previ-
ously31 for bolaform amphiphiles having a membrane
spanning spacer, probably due to the presence of gaps
in the outer leaflets of the vesicles. We believe that this
could be one possible explanation why 3 shows poor
cooperativity in its thermotropic phase transition. In
addition, the increased rigidity offered by quinol-based
structural segments in 3 might also cause apparent
mitigation of cooperativity in the thermal phase transi-
tion of these bilayers.
Tem p er a tu r e-Dep en d en t UV Sp ectr a l Va r ia tion s
w ith Vesicle-Dop ed BAA. Benzoylacetanilide (BAA)
is a water soluble molecule that undergoes keto-enol
tautomerism. Ueno used this as a probe for the deter-
mination of apparent thermal phase transition temper-
ature (Tm) of various bilayer membranes.32 To further
examine the concentration dependent melting behavior
of these ion-paired systems, we used the same probe in
our studies. However, vesicular 4 and 5 were not
examined using this method since well-defined DSC
profiles were obtained for both of them irrespective of
their concentrations.
The value of I1/ I3 depends on the chemical composition
of the medium surrounding the probe and, thus, gives
an estimation of the micropolarity of the surrounding
medium. Table 3 showed the I1/ I3 values of 1-5 at 25
and 50 °C. It prompted us to believe that the local
composition at the vesicle solubilization site of pyrene is
different depending on the nature of the substitution of
alkyl chain on the aromatic unit. The value of I1/ I3 is
(31) (a) Duivenvoorde, F. L.; Feiters, M. C.; van der Gaast, S. J .;
Engberts, J . B. F. N. Langmuir, 1997, 13, 3737. (b) Fuhrhop, J .-H.;
Liman, U.; Koesling, V. J . Am. Chem. Soc. 1988, 110, 6840.
(32) Ueno, M.; Katoh, S.; Kobayashi, S.; Tomoyama, E,; Obata, R.;
Nakao, H.; Ohsama, S.; Koyama, N.; Morita, Y. Langmuir 1991, 7,
918.
(33) Kalyanasundaram, K. Photochemistry in Microheterogeneous
Systems; Academic Press: New York, 1987; p 177.

Hybrid Bolaphile/Amphiphile Ion-Pairs
J . Org. Chem., Vol. 63, No. 22, 1998 7647
F igu r e 5. Plots of Aenol/Aketo due to vesicle-doped BAA vs temperature (°C) for (a) 1 and (b) 2.
significantly different for 1-3 (0.97) compared to that of
4 and 5 (∼1.25) at 25 °C where all but 2 are present in
their gel states. Pyrene sensed more hydrophobic mi-
croenvironments in 1-3 compared to that in 4 at 25 °C.
This indicates that instead of residing near the interfacial
region of the vesicle, pyrene localizes itself near the
aromatic ring embedded in the hydrophobic membrane
core probably due to more favorable aromatic-aromatic
interactions.
attributable to a tilted chromophore orientation. There-
fore, these studies clearly show the existence of the
aromatic-aromatic stacking interaction between the
aromatic rings in 3 and to a limited extent in 2 while
they are in their aggregated vesicular form. No such
indications were, however, apparent from the vesicles of
1 or 4.
X-r a y Diffr a ction Stu d ies. To understand the ori-
entation of these ion-paired systems in the membranes
and also to gain knowledge about their lamellar packing,
X-ray diffraction studies were performed on supported,
cast films of 1-5. Reflections up to 20° were analyzed
and interpreted in terms of higher order reflections of
stacked bilayer structures.36
A series of reflections mostly up to the eighth order
reflections were obtained (not shown) for 1-4 and 5, the
highest intensity peak corresponding to the lowest 2θ
value being the long spacing. The long spacings were
46.8, 39.11, 36.06, and 33.41 Å for 1-4, respectively on
the basis of the higher order reflections (n ) 3, 4, 5, 6,
7). Closer examination of the scan profiles also showed
the existence of another series of very weak reflections
presumably indicating the presence of a second kind of
molecular packing in these aggregates. The X-ray dif-
fraction of 5 was run for comparison which gave the long
spacing value of 37.85 Å. When we extended the studies
in the low angle range, we identified a weak reflection
at 1.74°, corresponding to a d spacing of 50.85 Å. These
diverse long spacing values clearly indicate the formation
of different kinds of aggregates by these amphiphiles in
aqueous dispersions which is in agreement with observed
differences in the UV-visible spectra taken in CHCl3 and
in aqueous dispersions.
UV-Vis Stu d ies. To examine whether there is any
existence of interaromatic stacking interactions in such
aggregates, we then proceeded to compare their UV-vis
spectra in organic solvent with that of aggregates in
water. The absorption peaks for 1 and 4 in CHCl3
appeared at 241.8 and 278.6 nm and 242.5, 272.8, and
278.8 nm, respectively. The corresponding absorption
maxima for vesicular 1 and 4 were 228.0, 274.8, and
278.6 nm and 227.8, 272.5, and 278.7 nm, respectively
(Figures 6b and 6a). Thus, with 1 and 4, the UV-vis
spectra in CHCl3 and in vesicular solutions were quite
similar. On the other hand, the absorption peaks for 2
in CHCl3 occurred at 241.4, 276.6, and 282.5 nm (shoul-
der peak), respectively, which shifted to 226.0, 275.2, and
281.6 nm in aqueous dispersion (Figure 6c). Similarly,
the absorption peaks for a solution of 3 in CHCl3 occurred
at 241 and 292.8 nm, but for the corresponding vesicular
dispersion, the absorption maxima were located at 226.0,
294.0, and 302.0 nm (Figure 6d). Thus, in the case of 3,
a red shift as large as 10 nm and, for 2, a 3 nm red shift
were observed in vesicular solutions when compared to
the respective spectra taken in CHCl3.
According to the molecular exciton model,34 the red
shift observed for the bilayer aggregate should be derived
from the π-stacking of the chromophore, which has also
been defined as the J -like aggregation.35 It is also
Assuming the long hydrocarbon and the fatty acid
chains in these amphiphiles to have a s-trans conforma-
(34) Kasha, M. In Spectroscopy of Excited State; Plenum Press: New
York, 1976; p 337.
(35) (a) Tai, Z.; Qian, X.; Wu, L.; Zhu, C. J . Chem. Soc., Chem.
Commun. 1994, 1965. (b) Everaars, M. D.; Marcelis, A. T. M.;
Sudho¨lter, E. J . R. Langmuir 1993, 9, 1986.
(36) (a) Kimizuka, N.; Kawasaki, T.; Kunitake, T. J . Am. Chem. Soc.
1993, 115, 4387. (b) Kunitake, T.; Shimomura, M.; Kajiyama, T.;
Harada, A.; Okuyama, K.; Takayanagi, M. Thin Solid Films 1984, 121,
L89.

7648 J . Org. Chem., Vol. 63, No. 22, 1998
Bhattacharya et al.
F igu r e 6. Comparison of UV-vis spectra in aqueous dispersion of (a) 4, (b) 1, (c) 2, and (d) 3; inset shows the corresponding
UV-vis spectra in CHCl3.
tion, the thicknesses of the bilayers corresponding to 4
and 5 were estimated to be ∼ 41 Å from a CPK model of
two respective molecules oriented parallel to the bilayer
normal. The measured values of 33.41 and 37.85 Å,
respectively, suggest a tilted bilayer arrangement of the
alkyl tails. In the case of 1, i.e., with the counterion
anchoring through an ortho-substituted aromatic ring,
a folded arrangement with two hydrophobic chains
pointing to one direction is possible and thus gives rise
to a bilayer type orientation. In case of 2, the meta-
oriented isomer supports the interdigitated mode of
packing. In contrast, 3 with para directionality allows
the spanning of the undecanoate chains to opposite
directions in the membrane and thus the unit layer
thickness is very close to the length of the corresponding
bolaphile.
Molecu la r Mod elin g Stu d ies. To rationalize the
possible origin of such variations of the biophysical
properties of vesicular 1-5, depending on the geometric
features of the bolaphilic segments in these ion-pairs we
also carried out molecular mechanics studies. All the ion-
pairs were drawn in the computer using INSIGHT II,37
and their energies were minimized (see Supporting
Information). In 1, the two apolar undecanoate chains
in the bolaphilic core originate from a central catechol
unit and are at ∼60° in their extended conformations.
Even after ion-pairing, chain propagation in 1 in an
angular fashion leads to severe packing impairment. In
addition, as indicated from molecular modeling, other
conformational plans (tilted) are also possible to minimize
apolar chain/water contacts. It may be possible that a
number of such arrangements could coexist leading to
concentration dependent phase transition behavior. With
2, the undecanoate chains stem from a central resorcinol
unit and remain at an angle of ∼120°. In the conforma-
tional plan with a unit layer thickness of ∼39 Å (consis-
tent with the value obtained from the reflection X-ray
diffraction experiment on the cast film of 2), the intra-
monomer van der Waals interactions between the chains
become severely “loose”. Variation of its concentration
may lead to the formation of alternative structures.
Indeed XRD studies with cast films of aggregates of 2
also indicated the existence of additional organizations.
Despite the presence of structural “flaws” in the bo-
laphilic core of 2, the self-assembly of ion-pair persists
in such a way that a lower phase transition temperature
(<15 °C) is manifested. In contrast, the presence of an
aromatic quinol moiety at the bolaphilic core of 3 allows
the formation of tightly organized assemblies as the
undecanoate chains from the central quinol units span
along opposite directions (∼180°). This conformational
plan gives optimal van der Waals contacts upon ion-
pairing with CTA+ and also interaromatic π-stacking
association. The existence of a stacking type of interac-
tion is supported by a comparison of the UV-vis
(37) For details please consult BIOSYM programs, which are
available from Biosym Technologies, 9685 Scranton Road, San Diego,
CA 92121-3752.

Hybrid Bolaphile/Amphiphile Ion-Pairs
J . Org. Chem., Vol. 63, No. 22, 1998 7649
filtrate gave the corresponding diacid (8a ) as white crystals
(0.22 g, 80%): mp 88-90 °C; H NMR (DMSO-d6, 90 MHz) δ
1.25-1.80 (s + “br” m, 32H), 2.25 (t, 4H), 3.98 (t, 4H), 6.88 (s,
4H); IR (Nujol) 1685, 1245 cm-1; LRMS, EI, m/z (%) 478 (22).
Anal. Calcd for C28H46O6: C, 70.26; H, 9.68. Found: C, 70.31;
H, 9.61.
studies of the aqueous aggregates of 3 against its solution
in CHCl3. It is difficult to explain the observed effect on
dilution of 3 on the basis of molecular mechanics calcula-
tions alone. Presumably dilution to lower concentration
leads to the formation of other plans that may not
support bilayer type organization. Whatever may be the
exact reason, the present observations are novel and
unprecedented and also show for the first time that the
vesicle formation could be controlled by variation of
concentration.
In summary we have synthesized four new gemini
carboxylate surfactants and their corresponding ion-pairs
with CTA+ which upon dispersion in water afforded
vesicular aggregates. Detailed examination of their
morphological and biophysical properties by a variety of
physical methods indicates that it is possible to modulate
their finer bilayer organizations which in turn affect their
properties at the membranous levels. Present findings
illustrate a novel approach to fine-tuning vesicular
properties and suggest new recipes for stable vesicular
structures that may be of practical value. In particular,
mixed bilayers composed of bacterial lipid (bolaam-
phiphile) and fatty acids or mammalian lipids (monopolar
amphiphile) like cardiolipins could be tailored to have
useful biological properties. Some of these mixtures may
also produce transient membrane-like milieu. These
aspects are currently under examination.
1
Dieth yl P h en yl-1,3-bis(oxyu n d eca n oa te) (7b). A mix-
ture of resorcinol (0.27 g, 2.5 mmol), ethyl 11-bromounde-
canoate (1.46 g, 5 mmol), and fused potassium carbonate (1.1
g, 8.0 mmol) in 25 mL of dry acetone was refluxed for 20 h.
Then the reaction mixture was filtered, and upon evaporation
of acetone from the filtrate a solid was obtained. It was
chromatographed on silica gel with hexane/EtOAc (96:4) to
give 7b (1.14 g, 85%) as a white solid: mp 45-46 °C; TLC
(hexane/EtOAc 4:1), Rf 0.8; 1H NMR (CDCl3, 90 MHz) δ 1.16-
2.0 (m, 38H), 2.28 (t, 4H), 3.92 (t, 4H), 4.12 (q, 4H), 6.36-6.52
(m, 3H), 7.12 (t, 1H); 13C NMR (CDCl3, 22.5 MHz) δ 14.67,
24.44, 26.67, 34.67, 60.22, 68.0, 101.56, 106.67, 130.0, 160.67,
174.0; IR (Nujol) 1715, 1275 cm-1; LRMS, EI, m/z (%) 534 (22);
HRMS (EI) calcd for C32H54O6 534.4206, found 534.3921. Anal.
Calcd for C32H54O6: C, 71.87; H, 10.18. Found: C, 71.81; H,
10.13.
P h en yl-1,3-bis(oxyu n d eca n oic a cid ) (8b). 7b (0.3 g, 0.56
mmol) was heated with potassium hydroxide (0.4 g, 8.7 mmol)
in ethanol (10 mL) on a water bath for 5 h. Excess ethanol
was removed, the residue obtained was diluted with water (20
mL), and the mixture was filtered. Acidification of the clear
filtrate gave the corresponding diacid (8b) as white crystals
(0.23 g, 85%): mp 119-121 °C; 1H NMR (DMSO-d6, 90 MHz)
δ 1.25-1.80 (s + “br” m, 32H), 2.25 (t, 4H), 3.95 (t, 4H), 6.40-
6.52 (m, 3H), 7.15 (apparent t, 1H); 13C NMR (CDCl3, 22.5
MHz) δ 24.89, 26.0, 29.11, 34.22, 67.78, 101.33, 106.89, 130.22,
160.67, 175.1; IR (Nujol) 1685, 1275 cm-1; LRMS, EI, m/z (%)
478 (23). Anal. Calcd for C28H46O6: C, 70.26; H, 9.68.
Found: C, 70.21; H, 9.71.
Exp er im en ta l Section
Gen er a l. Melting points were recorded in open capillaries
and are uncorrected. NMR spectra of CDCl3 solutions were
obtained at 90, 200, or 270 MHz (1H). Descriptions of
instruments used for various characterizations have been
published.7a
Cetyltrimethylammonium bromide (CTAB), 11-bromounde-
canoic acid, 1,6-diphenyl-1,3,5-hexatriene (DPH), riboflavin,
pyrene, and Amberlite IRA-900 (OH- form) were from Aldrich
Chemical Co. Benzoylacetanilide (BAA) was prepared by the
dropwise addition of aniline to ethyl benzoylacetoacetate at
150 °C in xylene. Phenol was distilled, and catechol, resorci-
nol, and quinol were used after purification. Thin-layer
chromatography was performed on Merck silica gel-G plates.
Column chromatography was performed on silica gel (60-120
mesh, Merck). All the reagents and solvents were the highest
grade available commercially and used purified, dried, or
freshly distilled as required.
Dieth yl P h en yl-1,4-bis(oxyu n d eca n oa te) (7c). A mix-
ture of quinol (0.66 g, 6 mmol), ethyl bromoundecanoate (3.52
g, 12 mmol), and fused potassium carbonate (2.7 g, 20 mmol)
in 50 mL of dry acetone was refluxed for 24 h. The solvent
was removed from the reaction mixture. This afforded a solid
which upon chromatography on silica gel using CHCl3 pro-
duced 7c (2.76 g, 86%) as colorless crystals: mp 68-69 °C;
1
TLC (CHCl3), Rf 0.65; H NMR (CDCl3, 90 MHz) δ 1.0-1.92
(m, 38H), 2.28 (t, 4H), 3.88 (t, 4H), 4.12 (q, 4H), 6.8 (s, 4H);
13C NMR (CDCl3, 22.5 MHz) δ 14.44, 25.33, 26.44, 30.0, 34.44,
60.0, 68.44, 115.78, 153.56, 174.0; IR (Nujol) 1725, 1225 cm-1
;
LRMS, EI, m/z (%) 534 (100); HRMS (EI) calcd for C32H54O6
534.4206, found 534.3998. Anal. Calcd for C32H54O6: C, 71.87;
H, 10.18. Found: C, 71.83; H, 10.08.
P h en yl-1,4-bis(oxyu n d eca n oic a cid ) (8c). 7c (0.3 g, 0.56
mmol) was heated with potassium hydroxide (0.4 g, 8.7 mmol)
in ethanol (10 mL) on a water bath for 5 h. Excess ethanol
was removed, the residue obtained was diluted with water (20
mL), and the mixture was filtered. Acidification of the clear
filtrate gave the corresponding diacid (8c) as white crystals
(0.25 g, 90%): mp 128-130 °C; 1H NMR (DMSO-d6, 90 MHz)
δ 1.25-1.80 (s + “br” m, 32H), 2.25 (t, 4H), 3.92 (t, 4H), 6.82
(s, 4H); IR (Nujol) 1695, 1230 cm-1; LRMS, EI, m/z (%) 478
(20). Anal. Calcd for C28H46O6: C, 70.26; H, 9.68. Found: C,
70.23; H, 9.69.
11-P h en oxyu n d eca n oic a cid (8d ). To a mixture of
freshly distilled phenol (1.41 g, 15 mmol) and 11-bromounde-
canoic acid (3.95 g, 15 mmol) was added slowly dropwise with
stirring a solution of sodium hydroxide (1.0 g, 25 mmol) in 7.0
mL of water. Stirring was continued for an additional 10 min.
Then the reaction mixture was heated at 70 °C. This gave a
residue to which 50 mL of water was added, and the resulting
mixture was filtered. Acidification of the alkaline filtrate gave
8d as a white solid. Recrystallization of crude product using
hexane/methanol mixture gave colorless crystals (3.51 g,
84%): mp 75-77 °C (lit.11g mp 77-78 °C); TLC (CHCl3, Rf 0.5);
1H NMR (CDCl3, 90 MHz) δ 1.25-1.80 (s + “br” m, 16H), 2.40
(t, 2H), 3.95 (t, 2H), 6.85-7.00 (m, 3H), 7.20-7.40 (m, 2H); IR
(Nujol) 1690, 1235 cm-1; LRMS, EI, m/z (%) 278 (8).
Syn th esis. The syntheses of four ion-paired amphiphiles
are described below.
Dieth yl P h en yl-1,2-bis(oxyu n d eca n oa te) (7a ). A mix-
ture of catechol (0.66 g, 6.0 mmol), ethyl bromoundecanoate
(3.52 g, 12.0 mmol), and fused potassium carbonate (5.4 g, 40
mmol) in 50 mL of dry acetone was refluxed for 18 h until
TLC indicated the disappearance of the spot due to catechol.
The insoluble solid residue from the reaction mixture was
removed by filtration, and the filtrate was concentrated. The
residue obtained on evaporation of acetone was chromato-
graphed over silica gel with hexane/EtOAc (98:2) to give 7a
(2.88 g, 90%) as a white solid: mp 40-41 °C; TLC (hexane/
EtOAc 10:1); Rf 0.75; 1H NMR (CDCl3, 90 MHz) δ 1.12-2.0
(m, 38H), 2.25 (t, 4H), 4.0 (t, 4H), 4.12 (q, 4H), 6.88 (s, 4H);
13C NMR (CDCl3, 22.5 MHz) δ 14.30, 25.03, 26.00, 29.36, 34.35,
60.13, 69.24, 75.74, 77.14, 78.55, 114.2, 121.03, 149.3, 173.79;
IR (Nujol) 1730, 1250 cm-1; LRMS, EI, m/z (%) 534 (62); HRMS
(EI) calcd for C32H54O6 534.4206, found 534.3921. Anal. Calcd
for C32H54O6: C, 71.87; H, 10.18. Found: C, 71.78; H, 10.11.
P h en yl-1,2-bis(oxyu n d eca n oic a cid ) (8a ). 7a (0.3 g, 0.56
mmol) was heated with potassium hydroxide (0.4 g, 8.7 mmol)
in ethanol (10 mL) on a water bath for 5 h. Excess ethanol
was removed, the residue obtained was diluted with water (20
mL), and the mixture was filtered. Acidification of the clear

7650 J . Org. Chem., Vol. 63, No. 22, 1998
Bhattacharya et al.
Gen er a l P r oced u r e for th e Syn th esis of Ion -P a ir ed
Am p h ip h iles (1-4). The ion-paired amphiphiles (1-4) were
synthesized by passage of a methanolic solution of freshly
recrystallized CTAB (1 equiv) through a column (9 cm × 1.5
cm) filled with freshly washed ion-exchange resin (Amberlite
IRA-900, OH- form). The eluents were collected together, and
to this were added 0.5 equiv of 8a , 8b, and 8c for 1-3 or 1
equiv of 8d for 4, respectively. Upon completion of this step
the respective mixture was stirred at room temperature for
24 h under a nitrogen inlet. Then the solvent was evaporated
under vacuum, and the solid obtained was recrystallized
several times from a mixture of MeOH:EtOAc (1:10). This
procedure afforded yields of ∼90-95% for all the ion-paired
systems, 1-4. IR confirmed the absence of a carboxylic acid
moiety and the appearance of the ammonium carboxylate
(Me3N+‚‚‚-O2C) group (∼1560 cm-1).
were prepared in the presence of 0.5% (w/v) uranyl acetate
solution by the reverse-phase evaporation method. Opalescent
to optically clear aqueous dispersions were obtained in all the
cases. One drop of the above dispersion was placed into a
carbon-Formvar coated copper grid (400 mesh). Filter paper
was employed to wick away excess water. It was then placed
under a mechanical vacuum for approximately 0.5 h. A J EOL-
TEM 200 CX electron microscope with an accelerating voltage
of 120 keV was used for micrograph recording at magnifica-
tions of 57 000 and 96 000.
Ligh t Sca tter in g Exp er im en ts. Mean hydrodynamic
diameters were determined by laser light scattering using
Zetasizer 3000 (Malvern Instrument Ltd., Malvern, U.K).
Light scattering employed a He-Ne laser source at a wave-
length of 633 nm keeping the detector angle at 90°. Each
suspension was generated by reverse-phase evaporation as
described previously. The data were analyzed using internal
instrument software involving Malvern 7132 digital (16-bit)
autocorrelator. A 220 nm latex standard was used for calibra-
tion.
En tr a p m en t of Ribofla vin a n d Gel F iltr a tion . The
vesicles were prepared by bath sonication of the lipids (1 mg/1
mL) in aqueous riboflavin solution (concentration 2 × 10-5 M,
pH 6.8) for 30 min at ∼60 °C. Since the lamellar phases
formed from the ion-paired amphiphiles are extremely sensi-
tive to the presence of salt, the trapping experiments were
performed in pure water. After cooling to room temperature,
the vesicle suspension (3 mL) was loaded onto a sephadex G-50
M (Pharmacia) column (26 cm × 1 cm) and eluted with water
(pH 6.8). Fractions containing most of the vesicles were
pooled. These contained riboflavin molecules that were bound
to the vesicles at the outer surfaces and also the ones that got
entrapped in the vesicular inner pools.
All compounds (1-4) were found to be quite hygroscopic,
and all of them crystallized as hydrates despite prolonged
drying under vacuum. IR and repetitive elemental analyses
confirmed the presence of water molecules in them. High
1
resolution H NMR spectra of 1-4 also confirmed the presence
of an additional peak possibly due to hydration. However, this
additional peak could be vanished upon treatment with D2O
as they were found to be fully exchangeable with D2O. All
the compounds gave satisfactory 1H NMR, IR, and C,H,N
analysis. Pertinent spectroscopic (1H NMR, IR) and analytical
(elemental analyses) data are given below.
Bis(h exa d ecyltr im eth yla m m on iu m )p h en yl-1,2-d i(ox-
yu n d eca n oa te) (1): mp 110-130 °C; 1H NMR (200 MHz,
CDCl3) δ 0.88 (t, 6H), 1.25-1.76 (s + “br” m, 88H), 2.16 (t,
4H), 3.06 (“br” m, exchangeable with D2O), 3.32-3.40 (sharp
s + m, 22H), 3.98 (t, 4H), 6.88 (s, 4H); IR (Nujol) 1560 cm-1
.
Anal. Calcd for C66H128N2O6, H2O: C, 74.52; H, 12.32; N, 2.63.
Found: C, 74.49; H, 12.21; N, 2.38.
Kin etics of Tr a n sm em br a n e P er m ea tion . An aliquot
of 1 mL (pH 6.8) was taken in a fluorescence cuvette, and the
time-dependent decrease in the fluorescence intensity at 514
nm (excitation at 374 nm, excitation and emission slit widths
10 and 20 nm respectively) with time at 25 ( 0.5 °C was
followed using a Hitachi Model F-4500 fluorescence spectro-
photometer upon adjustment of pH from 6.8 to 10.2 by the
addition of an aliquot of 66.3 µL of a 0.034 M KOH solution.
Since riboflavin fluorescence decreases upon ionization alka-
line media (pKa ∼ 10.2), the time-dependent loss of the
fluorescence intensity at 514 nm due to riboflavin under an
imposed transmembrane pH gradient of 3.4 pH units was
taken as a measure of OH- permeation across these vesicles.
Rate constants were obtained from the monoexponential time
dependent portion of the loss of the fluorescence intensity.
Differ en tia l Sca n n in g Ca lor im etr y. (a ) P r ep a r a tion
of Su r fa cta n t Su sp en sion for Ca lor im etr y. Individual
lipid dispersions in pure water (Millipore) were prepared by
weighing an appropriate amount of the solid ion-paired sur-
factant into the DSC pan and adding 12 µL water. The pan
was immediately sealed, transferred to the calorimeter, and
kept at 80-90 °C to accomplish proper hydration. A 120 mmol
solution of each compound was used for DSC measurements.
(b) Ca lor im etr ic Mea su r em en ts. DSC measurements
were carried out with a Perkin-Elmer Model DSC-2C dif-
ferential scanning calorimeter at a sensitivity of 1 mcal/s and
a scanning rate of 5 K/min. A pan containing an identical
volume of pure water (Millipore) was used as reference in all
cases. Heating and cooling thermograms for each sample were
repeated several times to get reproducible thermograms. The
peak in the excess heat capacity vs temperature plot was taken
as the main transition temperature Tm. At different scanning
rates, identical values of enthalpy, transition temperature,
half-height width, and height of the transition profile were
obtained. The transition enthalpy, ∆H, was determined by
measuring the area under the transition profile and was the
average of at least three separate isolated runs. This integral
was used to calculate both the calorimetric (∆Hcal) and van’t
Hoff (∆HvH) enthalpies of the transition. These enthalpies
were used to determine the “cooperative unit” (CU) of the
transition, the number of monomers undergoing the phase
Bis(h exa d ecyltr im eth yla m m on iu m )p h en yl-1,3-d i(ox-
yu n d eca n oa te) (2): mp 50 °C; 1H NMR (200 MHz, CDCl3) δ
0.88 (t, 6H), 1.25-1.75 (s + “br” m, 88H), 2.17 (t, 4H), 2.34 (s,
exchangeable with D2O), 3.36-3.44 (sharp s + m, 22H), 3.93
(t, 4H), 6.45-6.50 (m, 3H), 7.14 (apparent t, 1H); IR (Nujol)
1560 cm-1. Anal. Calcd for C66H128N2O6, H2O: C, 74.52; H,
12.32; N, 2.63. Found: C, 74.45; H, 12.28; N, 2.43.
Bis(h exa d ecyltr im eth yla m m on iu m )p h en yl-1,4-d i(ox-
yu n d eca n oa te) (3): mp 105-110 °C; 1H NMR (270 MHz,
CDCl3) δ 0.88 (t, 6H), 1.15-1.80 (s + “br” m, 88H), 2.1 (“br” s,
exchangeable with D2O), 2.27 (t, 4H), 3.34-3.50 (s + m, 22H),
3.92 (t, 4H), 6.82 (s, 4H); IR (Nujol) 1550 cm-1. Anal. Calcd
for C66H128N2O6, H2O: C, 74.52; H, 12.32; N, 2.63. Found: C,
74.41; H, 12.26; N, 2.41.
Hexa d ecyltr im eth yla m m on iu m p h en oxyu n d eca n oa te
(4): mp 35 °C (soften), 80 °C (clear melt); 1H NMR (200 MHz,
CDCl3) δ 0.88 (t, 3H), 1.16-1.77 (s + “br” m, 44H), 2.19 (s,
exchangeable with D2O), 2.24 (t, 2H), 3.40-3.49 (sharp s +
m, 11H), 3.94 (t, 2H), 6.87-6.92 (m, 2H), 7.23-7.30 (m, 3H);
IR (Nujol) 1560 cm-1. Anal. Calcd for C36H67NO3, 0.5 H2O:
C, 75.74; H, 12.00; N, 2.45. Found: C, 75.61; H, 11.81; N, 2.27.
Vesicle P r epar ation . Reverse-phase evaporation24 method
was applied for the preparation of vesicles from these ion-
paired amphiphiles for transmission electron microscopy.
Typically, 2.5 × 10-3 M vesicular suspensions in pure, deion-
ized water were prepared in the presence of a 0.5% uranyl
acetate solution. For microcalorimetric studies, fluorescence
anisotropy measurements, UV-vis spectroscopy studies, and
X-ray diffraction experiments, the required amount of the ion-
paired surfactants was placed in a container and the solid
dissolved completely in chloroform. The solvent was evapo-
rated by keeping it continuously in high vacuum for 12 h to
produce a film of the surfactant. This was then hydrated with
deionized water, and the hydrated films were subjected to 4-5
freeze-thaw (above the phase transition temperature) cycles.
The resulting opalescent solution was then sonicated to give
stable aggregates. For entrapment experiments, the bath-
sonicated vesicles were used. In all the cases, clear and
optically translucent dispersions were obtained.
Tr a n sm ission Electr on Micr oscop y Stu d ies. In all the
cases, 2.5 × 10-3 M vesicular suspensions in deionized water
transition from the relationship CU ) ∆HvH/∆Hcal
. All the

Hybrid Bolaphile/Amphiphile Ion-Pairs
J . Org. Chem., Vol. 63, No. 22, 1998 7651
thermodynamic parameters are the averages of at least two
independent sample preparations, and each of the specimens
was scanned at least three times. Main phase transition
entropy, ∆S, in the units of cal mol-1 K-1, was determined on
the basis of the assumption that at the phase transition
temperature, the gel, and the liquid-crystal like phases were
in equilibrium. ∆S was measured from the equation ∆S )
∆Hcal/Tm, where ∆H is expressed in cal mol-1 and Tm in Kelvin.
Phase transition temperatures (Tm) are accurate up to (0.5
the wavelength range of 200-450 nm. Similarly, during the
cooling of the vesicular solutions, the spectra at different
temperatures were also recorded. The plots of the ratios of
the absorbances due to enol (315 nm) and to keto (250 nm),
[Aenol/Aketo], vs temperature (T) gave apparent Tm values for
the individual vesicles.
Micr op ola r ity a n d P h a se Tr a n sition Tem p er a tu r e by
P yr en e. A solution of pyrene in MeOH was introduced into
a solution of a given ion-pair in CHCl3. The solvent from this
mixture was removed under vacuum. Water was added to the
dry lipid film containing pyrene, and vesicles were obtained
by freeze-thaw method. The vesicle and pyrene concentra-
tions were 5 × 10-4 and 7.9 × 10-8 M, respectively. A Hitachi
650-60 spectrofluorimeter was used to determine the systemic
micropolarities at various temperatures and apparent phase
transition temperature (Tm). The temperature of the sample
cell holder was regulated by circulating water using a constant
temperature water bath (model J ulabo F10). Keeping the
excitation wavelength fixed at 337 nm, the emission spectra
of the region 360-400 nm were recorded using bandwidths of
5 nm. The samples were allowed to thermally equilibrate for
10 min prior to each measurement. At each temperature, the
ratio I3/ I1, where I3 and I1 represent the intensities of the third
and the first vibronic peaks in the fluorescence emission
spectrum due to pyrene was recorded and plotted against T,
the temperature. The measurements were done in the tem-
perature range of 25-70 °C. The value of I1/ I3 also gives an
idea about the micropolarity of the vesicular suspension.
UV-Visible Sp ectr oscop y. A 2.5 mmol vesicular suspen-
sion was prepared as described previously. UV-vis spectra
of ion-paired systems 1, 2, 3, and 4 were taken separately as
solutions in pure CHCl3 and in the form of vesicular aggregates
in aqueous media at 25 °C using Shimadzu model 2100 UV-
visible recording spectrophotometer. Spectra in the wave-
length range of 190-500 nm were recorded.
X-r a y Diffr a ction Stu d ies. Self-supported cast films for
the X-ray diffraction studies were prepared by dispersing the
amphiphile in water (10 mg/1.0 mL) as described above. Few
drops of this suspension at hot condition were dispersed on a
precleared glass plate and air-dried at room temperature.
Finally it was kept under vacuum for 15 min. X-ray diffraction
was carried out by the reflection method with a X-ray diffrac-
tometer (model XDS-2000 Scintag Inc., USA). The X-ray beam
was generated with a Cu anode and the wavelength of KR1
beam was 1.5406 Å. The X-ray beam was directed toward the
film edge, and the scanning was done up to the 2θ value of
20°.
°C and enthalpies up to (1.0 kcal mol-1
. The van’t Hoff
2
enthalpy ∆HvH was obtained from the relation ∆HvH ) 6.9 Tm
/
∆T1/2 where ∆T1/2 is the width at the half-maximum excess
specific heat and is accurate within (10%.
Micr oca lor im etr y. Lipid vesicles (2.5 mmol) were made
by the hydration of the dry lipid film. These hydrated films
were subjected to 4-5 freeze-thaw (above the phase transition
temperature) cycles and vortexed. Then individual samples
were degassed under vacuum prior to the transfer to the
sample cell. Identical volumes of vesicular suspension (1.2
mL) and water were loaded into the sample and reference cell,
respectively, of an MC-2 ultrasensitive scanning calorimeter
(Microcal Co., Amherst, MA). Samples were maintained for
at least an hour below the gel-to-liquid-crystalline phase
transition temperature before being scanned. The scan rate
was 90 °C/h. The calorimeter was calibrated electronically.
The transition enthalpy and van’t Hoff enthalpy were obtained
from the data analysis software (origin) coupled with it.
F lu or escen ce An isotr op y Stu d ies. Steady state fluo-
rescence measurements were carried out in a Hitachi F-4500
fluorescence spectrophotometer equipped with polarizers.
Temperature was regulated by a constant temperature water
circulating bath (model J ulabo F10) through the sample cell
holder. The samples were allowed to equilibrate for 10 min
prior to each run, and the temperature of the sample was
measured immediately prior to each measurement. Keeping
the excitation wavelength fixed at 360 nm, the emission
spectra in the wavelength region of 390-480 nm were studied
using a band-pass of 5 nm. The fluorescence anisotropies (r)
of DPH (8.6 × 10-8 M), as sensed by this probe doped in
individual vesicular solutions (2.5 mM), were measured. At
each temperature, the fluorescence emission spectra were
recorded by adjusting the polarizers in four different positions.
The anisotropy (r) at different temperatures for each vesicular
solution was calculated by employing Perrin’s equation, r )
(III - I G)/(III + 2I G) where, III and I are the observed
intensities measured with polarizers parallel and perpendicu-
lar to the vertically polarized exciting beam, respectively. G
is a factor used to correct for the inability of the instrument
to transmit differently polarized light equally. For each
aggregate, an r vs T plot gave the information about the gel
to liquid-crystalline phase change as function of temperature.
The values reported herein for r are the average values of three
independent measurements. The gel to liquid-crystalline
phase transition temperatures were calculated from the
midpoints of the breaks related to the temperature-dependent
anisotropy values. The temperature range for the phase
transition was calculated from the two temperature points for
each experiment which marked the beginning and the end of
the phase transition process.
Molecu la r Mod elin g Stu d ies. All five ion-paired amphi-
philes were drawn, and their energy-minimized, preferred con-
formations were calculated using the INSIGHT II 2.3.5. pack-
age (DISCOVER, Biosym. Technologies). DISCOVER is a mol-
ecular simulation program that performs energy minimization
to optimize initial geometries of different amphiphilic mol-
ecules constructed from appropriate fragments in INSIGHT.
Minimization path followed conjugate gradient optimization.
Consistent valence force field was selected for computations.
Calculated thickness and preferred drawings of different
amphiphilic monomers appear in Supporting Information.
Ack n ow led gm en t. This work was supported by a
TDM grant of Genetic Engineering and Biotechnology
and in part by a grant from HLRC, Bangalore. We are
grateful to Professor A. Surolia for giving access to
microcalorimeter, Dr. G. Subanna for TEM studies, and
SERC for the computational facilities.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR of 1, 2, 3
and 4, mass spectra of 7a , 7b, and 7c, and energy-minimized
drawings of different ion-paired systems 1-4 (Figure 1,
supplementary information) (14 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
UV-Absor ba n ce Meth od . In a typical experiment, a mix-
ture of BAA (2 × 10-5 M) and the ion-pair lipid (1 × 10-3 M)
was first dissolved in 100 µL of ethanol and then from this
ethanol was removed by keeping the sample in high vacuum
for 12 h to produce a BAA-doped lipid film. Then 2.5 mL of
water was added to this film, and the resulting mixture was
first vortexed and then subjected to 4-5 freeze-thaw cycles.
This gave a translucent BAA-doped vesicular solution. Then
an aliquot of this sample was transferred into the sample
cuvette, and to the reference compartment another vesicular
sample of identical lipid concentration that did not contain any
BAA was taken. The temperature of the cuvette chamber was
maintained by a thermoelectric temperature controller, TCC-
60. After attainment of a particular temperature, UV-vis
spectra were recorded in a Shimadzu model UV 2100 spec-
trophotometer at various temperatures from 6 °C onward in
J O980315L