Tripod facial surfactants with benzene as the central core: Design, synthesis and self-assembly study

Article abstract of DOI:10.1039/c2nj20876h

The design, synthesis and self-assembled study of a new class of benzene-derived tripod facial amphiphiles are reported. The synthetic route chosen based on a central mesitylene as scaffold allows easy tuning of lipophilic and hydrophilic groups and thus control of the tensioactive properties of these new surfactants. This new class of surfactants exhibits three glucose moieties as the hydrophilic polar head and three hydrocarbon chains each having 3 to 7 carbons as the lipophilic part. These tripod facial amphiphiles exhibit well-defined tensioactive and aggregative properties. Their critical aggregation concentration, their particle size in water (less than 20 nm), and their aggregation behavior are closely linked to the nature of their lipophilic chains and can therefore be easily modulated. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012.

Full text of DOI:10.1039/c2nj20876h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
NJC
Cite this: New J. Chem., 2012, 36, 1170–1179
www.rsc.org/njc
PAPER
Tripod facial surfactants with benzene as the central core: design,
synthesis and self-assembly studyw
ab
Julien Dauvergne,ab Anissa Bendjeriou,ab Franc
¸
´
oise Bonnete,
Joachim Kohlbrecher,c Bernard Pucci,ab Laurie-Anne Barretab and Ange Polidori*ab
Received (in Montpellier, France) 11th October 2011, Accepted 15th February 2012
DOI: 10.1039/c2nj20876h
The design, synthesis and self-assembled study of a new class of benzene-derived tripod facial
amphiphiles are reported. The synthetic route chosen based on a central mesitylene as scaffold
allows easy tuning of lipophilic and hydrophilic groups and thus control of the tensioactive
properties of these new surfactants. This new class of surfactants exhibits three glucose moieties
as the hydrophilic polar head and three hydrocarbon chains each having 3 to 7 carbons as the
lipophilic part. These tripod facial amphiphiles exhibit well-defined tensioactive and aggregative
properties. Their critical aggregation concentration, their particle size in water (less than 20 nm),
and their aggregation behavior are closely linked to the nature of their lipophilic chains and can
therefore be easily modulated.
For example, clip-like glycoluril-based structures display an
Introduction
interesting counterion-dependent aggregation.12 Amphiphilic
calix[4]arene scaffolds possessing lipophilic side chains on
the lower rim and either dendritic hydrophilic residues or
carboxylic moieties on the upper rim have also been
obtained.13–15 Kellermann et al.14 showed that these amphiphiles
adopt a cone-shaped conformation involving small aggregates
of high curvature, forming uniform aggregates consisting of
seven units, non-deformable upon drying. Facially amphiphilic
dendrimers have also been designed, forming well-controlled and
stable self-assembled structures.16–18 Facial amphiphiles offer
particular promise in biology, where they can be used to over-
come difficulties in handling membrane proteins in aqueous
solution.10
In the last few decades, synthetic amphiphiles with unusual
architectures such as bolaamphiphiles or gemini surfactants
have been obtained.1 Amphiphilic surfactants too, sometimes
called facial or contrafacial surfactants, have attracted increasing
interest.2 They are defined as molecules with a rigid framework
(such as the steroid skeleton) with hydrophilic and lipophilic
parts located on two opposite faces.3,4 The controlled self-
assembly of facial amphiphiles offers an attractive opportunity
to construct defined aggregates and supramolecular architec-
tures, specifically designed for particular applications.
Currently, cholic acid is the most popular platform for the
design of facial amphiphiles, leading researchers to derivatize
its hydroxyl groups by grafting on sugars, carboxylates,
amines/ammoniums and sulfonates, thus enhancing its facial
amphiphilic character.5–10a Non-steroidal facial amphiphiles
have also been synthesized with a block pattern derived from
bridged annulene10b or [2 + 2] and [4 + 2] norbornadiene
cycloaddition cores.11 In recent years, other facial amphiphiles
with original scaffolds have appeared in the literature.
Here, we describe how we designed the synthesis of a
new class of benzene-based tripod facial amphiphiles and
determined their tensioactive and aggregative properties in
aqueous media. These new facial amphiphiles were prepared
following a chemical pathway based on a central benzene
group as scaffold (Fig. 1).
Simple chemical modifications of this aromatic core allow
lipophilic and hydrophilic moieties to be tuned easily. Regarding
the aromatic ring, persubstituted aromatic compounds
having a D3d symmetry structure have been shown to adopt
an alternating up–down conformational network in which
cooperative interactions direct the vicinal groups to point to
opposite sides of the benzene ring, resulting in a 1,3,5/2,4,6
facial segregation.19 In the present case, the lipophilic face of
these amphiphiles is made up of three short hydrocarbon
chains each having 3 to 7 carbons and the hydrophilic face
consists of 3 glucosyl groups introduced via click-chemistry
resources. The physico-chemical properties of these facial
a
´
´
Universite d’Avignon et des Pays de Vaucluse, Faculte des Sciences,
´
`
Equipe Chimie Bioorganique et Systemes Amphiphiles, 33,
rue Louis Pasteur, 84000 Avignon, France.
E-mail: ange.polidori@univ-avignon.fr
´
Institut des Biomolecules Max Mousseron, UMR 5247,
CNRS—Universites Montpellier I et II 15, avenue Charles Flahault,
b
´
34093 Montpellier cedex 05, France
c Laboratory for Neutron Scattering, Paul Scherrer Institut,
5232 Villigen PSI, Switzerland
w Electronic supplementary information (ESI) available: TEM micro-
graphs of a dispersion of C7Glu3 in water after sonication, all 1H and
13C NMR spectra scanned of tripod compounds CnGlu3 and some
data concerning SAXS experiments. See DOI: 10.1039/c2nj20876h
c
1170 New J. Chem., 2012, 36, 1170–1179
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 1 General structure of tripod amphiphiles with 1,3,5/2,4,6 facial
segregation.19
amphiphiles, notably tensioactive properties, shape and size of
aggregates, are closely linked to the length of the lipophilic
chains grafted onto the benzene ring.
Results and discussion
The key step in obtaining alternately substituted benzene was the
sequential functionalization of 1,3,5-tribromo-2,4,6-trihydroxy-
methyl-benzene 1, which can easily be achieved on a multiple-
gram scale. Its synthesis from mesitylene as starting material, as
described by Anthony et al.,20 involves a four-step synthetic route
giving a 59% overall yield. In the present case we obtained a 90 g
batch without chromatography, which we used to synthesize
tripod facial amphiphiles according to the chemical route depicted
in Scheme 1.
The first decisive step involved a threefold nucleophilic
aromatic substitution of 1 by thiolate anions derived from
either propyl (a), butyl (b), pentyl (c), hexyl (d) or heptyl
mercaptan (e). The reaction proceeded smoothly under mild
conditions to give thioether compounds in excellent yield, due
to the use of 1,3-dimethyl-2-imidazolidone (DMI) as solvent
rather than DMF, which requires harsher conditions.21 To
efficiently introduce the three hydrophilic groups onto the
aromatic core, copper(I)-catalyzed alkyne–azide cycloaddition
was used. The hydroxyl groups were subjected to mesylation
using a combination of methanesulfonyl chloride/TEA in cold
CH2Cl2. Surprisingly, the reaction afforded the trichlorinated
derivatives 2a–e instead of the desired products. The mesylate
groups produced in the reaction mixture appear to be substituted
again by the existing chloride ions. Conversion requires a long
reaction time but the product is formed in high yields with good
purity.
Scheme 1 Synthetic route leading to tripod facial amphiphiles
CnGlu3, reagents, temperature, time and yield: (a) R1SH, tBuOK,
DMI, rt, 16 h, 97%; (b) Et3N, MsCl, CH2Cl2, 5 1C - rt., 48 h,
83%; (c) propargyl alcohol, NaH 60%, THF/DMF (1 : 1), 0 1C - rt.,
18 h, 65%; (d) 1-azido-1-deoxy-b-D-glucopyranoside tetraacetate,
DIEA, CuI, THF, rt, 24 h, 71%; (e) MeONa cat., MeOH/THF
(4 : 1), pH 9–10, rt, 24 h, 93%.
The subsequent substitution of chlorine groups by propargyl
alcoholate in DMF readily provided the trisubstituted derivatives
3a–e in 60% yield. The addition of 1-azido-1-deoxy-b-D-gluco-
pyranoside tetraacetate on trisubstituted derivatives 3a–e in the
presence of active CuI as catalyst afforded triglucosylated
compounds in good yield after 24 h reaction time at room
temperature. Finally, deprotection of acetyl groups by trans-
esterification in the presence of sodium methoxide as catalyst
provided tripod facial amphiphiles (CnGlu3). These alkynyl
derivatives 3a–e were used to efficiently attach sugar moieties
by the usual Cu(I)-catalyzed azide-alkyne [1,3]-dipolar cyclo-
addition reaction, in high yield and with complete regioselectivity
according to the optimized procedure for glycolipid ana-
logues.22,23 The treatment of derivatives 3a–e and peracetylated
glucosylazide in THF by direct use of active CuI(I) as catalyst
combined with DIEA, after a 24 h reaction time to ensure
complete conversion, afforded stereoselective acetylated com-
pounds CnGlu3 in good yield. This was confirmed by
1H NMR spectra, which revealed a characteristic peak at around
8 ppm assigned to the triazole protons and disappearance of the
triplet at around 2.5 ppm assigned to the alkyne proton.
Deprotection of acetylated group sodium methoxide and
purification by Sephadex LH-20 size exclusion chromatographies
provided pure CnGlu3 facial amphiphiles as white powder which
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1170–1179 1171

View Article Online
Table 1 Physico-chemical properties of tripods
CACa/mM
gCACa/mN mꢀ1
Amin/A2 a
Pe
log K0W
Dh/nmc
%Vol
NAgg
Rg/nmd
b
d
Entry
C3Glu3
C4Glu3
C5Glu3
C6Glu3
C7Glu3
2.38 ꢁ 0.03
39.2 ꢁ 0.2
36.3 ꢁ 0.2
35.8 ꢁ 0.2
33.2 ꢁ 0.2
31.3 ꢁ 0.2
235 ꢁ 58
211 ꢁ 26
210 ꢁ 8
0.260
0.292
0.295
0.324
0.296
6
3.3
5.2
8.1
11.5
4200
99.9
99.8
99.9
99.8
52
3
5
13
n.d
n.d
1.4
1.8
3.6
n.d.
n.d.
0.84 ꢁ 0.05
5.9
7.4
8.5
9.6
0.13 ꢁ 0.01
0.01 ꢁ 0.001
192 ꢁ 24
211 ꢁ 13
1.2 ꢂ 10ꢀ3 ꢁ 0.1 ꢂ 10ꢀ3
a
b
Measured by tensiometry (Wilhelmy plate method). Each value is the mean of three different measurements ꢁ SD. Measured by reversed phase
c
HPLC. The hydrodynamic or Stokes diameter (Dh) of particles is measured at a concentration at least 15 times above the CAC. The values are
d
the average of 10 runs. Aggregation number (NAgg) and radius of gyration (Rg) measured by SAXS at a concentration about 15 times the CAC
e
Packing parameter.
to linearly decrease as tripod concentration increases, showing
breaks corresponding to the critical aggregation concentration
(CAC) depending on the hydrophobic chain length, as for a
conventional surfactant.24
Such behavior clearly indicates a facial segregation of polar
and apolar parts in water on both sides of the aromatic ring
(Fig. 3). Furthermore, it should be noted that the triglucosylated
compound without hydrocarbon chains (C0Glu3) prepared
from 1,3,5-tris(bromomethyl)benzene following synthetic routes,
summarized in Scheme 2, does not exhibit surfactant properties,
indicating that the aromatic ring alone was not enough to endow
the molecule with a sufficiently hydrophobic character.
For the tripod compounds (Fig. 4), log(CAC) linearly
decreases when the number of methylenes of each hydrophobic
chain increases: log(CAC) = ꢀ1.96nC + 7.26 with R2 = 0.9804.
However it should be pointed out that within a homologous
Fig. 2 Surface tension vs. log C for compounds C3Glu3 (&),
C4Glu3 (’), C5Glu3 (J), C6Glu3 (ꢂ) and C7Glu3 (m).
were fully characterized by 1H, 13C and HR mass spectrometry.
All these compounds have good water solubility except C7Glu3;
this lower solubility was associated with the longer hydrocarbon
tails of the lipophilic part.
Tensiometric measurements allowed us to determine the
main tensioactive properties of these tripods (Table 1 and
Fig. 2). For all compounds tested, the surface tension is observed
Scheme 2 Synthetic route of C0Glu3, reagents, temperature, time
and yield: (a) propargyl alcohol, NaH 60%, DMF, 0 1C, 48%;
(b) 1-azido-1-deoxy-b-D-glucopyranoside tetraacetate, DIEA, CuI,
Fig. 3 Representation of a stick molecular model of the tripod
C3Glu3 (Chem3D Pro 11.0).
THF, RT, 50%; (c) MeONa cat., MeOH/THF (4 : 1), pH 9–10, rt.,
24 h, 87%.
c
1172 New J. Chem., 2012, 36, 1170–1179
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 4 logCAC plotted vs. number of carbons of the hydrophobic chains.
Fig. 5 Log(K0w) of tripods CnGlu3 plotted against number of carbon
series of surfactants, the CAC usually decreases by a factor of
about 10 every time two methylene groups are added to the
hydrophobic tail.24 In the present case, this factor increases from
three to eight every time one methylene group is added to each
tail. This weaker contribution to the hydrophobic effect is
explained by the fact that the three methylenes are not added
linearly to one chain, but to three different chains. Similar
behavior can be observed with lipids. A second hydrocarbon
chain added to an amphiphilic molecule makes a smaller con-
tribution to the hydrophobic effect than the one already linked to
that molecule.24 By applying Gibbs equation G = ꢀ(1/RT)[dg/
dLog C], the relative excess of surfactant at the air–water inter-
face G was determined, leading to the evaluation of the area per
surfactant polar head from Amin = 1/GN. These tripod surfac-
tants exhibit a wide polar head (Amin = 210 ꢁ 60 A2), whatever
the nature of the hydrophobic tails (Table 1).
of the hydrophobic chains.
log(K0w) is a parameter closely related to the molecule’s
water/octanol partition coefficient, which can be obtained
from reverse-phase HPLC.25 This parameter is usually con-
sidered to reflect the hydrophobic character of the surfactant.
log(K0w) shows a linear evolution from C4Glu3 to C7Glu3
when the carbon number of hydrophobic chains increases
(Fig. 5). It can be seen that the log(K0w) of the C3Glu3
compound is equal to that of the C4Glu3 compound. Thus,
we can assume that the hydrophobic contribution of the 3
additional methylenes remained marginal for very short
chains. Above their CAC, all surfactants form monodispersed
self-assembled systems with a Stokes (or hydrodynamic) dia-
meter lower than 20 nm, except for C7Glu3 (Table 1 and
Fig. 6). The tripod hydrodynamic diameters increase with the
length of the hydrophobic tail. Particle diameters obtained
from DLS measurements suggest that C3Glu3 and C4Glu3
should form spherical micelles in water with a diameter of less
than 6 nm, but the increase in the particle size observed with
C5Glu3 and C6Glu3 indicates the formation of larger micelles.
Further information about the shape of the tripod assemblies
was obtained by Small Angle X-ray Scattering experiments
(SAXS). The last tripod C7Glu3, which has low water solubility,
leads to unstable large particles larger than 0.2 mm as observed by
TEM after negative staining (see ESIw, Fig. S1). Moreover, this
compound precipitates quickly after dispersion by sonication.
DLS measurements also reveal very large aggregates testifying to
the rapid fusion of vesicles formed from the tripod C7Glu3.
Fig. 6 Stokes (or hydrodynamic) diameter distribution from DLS
measurements for amphiphile solutions as a function of length of
linear alkyl chains: C3Glu3 at 15 CMC (—, blue), C4Glu3 at 15 CMC
(–ꢃ–, green), C5Glu3 at 120 CMC (ꢃ ꢃ ꢃ, red), C6Glu3 at 15 CMC (– –,
black), C7Glu3 at 120 CMC (–ꢃꢃ–, brown).
To better understand the whole process of evolution, the
self-aggregated properties of C3- to C5Glu3 tripods having
good water solubility were studied by SAXS at different
concentrations. Different behaviours were observed for
C3- to C5Glu3 tripods as the hydrophobic tail increased in
length. Comparisons between C3, C4 and C5Glu3 at the same
concentration (about 40 g Lꢀ1) are shown in Fig. 7A. Evolution
of scattering from factors of C3Glu3 to C5Glu3 as a function
of concentration is shown in the ESIw (Fig. S2). As the
concentration increases up to 42 g Lꢀ1 (15CMC), C3Glu3
and C4Glu3 form small aggregates (see ESIw, Table S2),
probably trimers for C3Glu3 and larger aggregates for
C4Glu3. Radii of gyration and molecular masses have been
determined via the Guinier approximation, the latter being
determined by comparing the forward scattering intensity
value I(0) at the highest concentration with the I(0) value
below the CAC (taken to be monomeric). Analysis of the pair
distribution function P(r) for C3Glu3 using the program
GNOM26 yielded a maximum particle dimension Dmax of
about 4 nm, consistent with DLS measurements. In the case
of C4Glu3, although the evolution is essentially the same, it
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1170–1179 1173

View Article Online
Fig. 7 (A) Experimental SAXS curves of C3- to C5Glu3 at their highest concentrations. (B) Scattering intensities of C5Glu3 at different
concentrations. The curves are not normalized for sake of clarity. The symbols represent the experimental data and the lines the data
fits.
was impossible to characterize the monomeric form below the
CAC due to a poor SAXS signal below 0.9 g Lꢀ1
preferred type of aggregation is through the surfactant pack-
ing parameter.
.
In the case of C5Glu3, the scattering curves clearly present a
different behaviour. The C5Glu3 pair distribution functions
(see ESIw, Fig. S3) and the Guinier analysis suggest the
formation of elongated objects, which lengthen as the tripod
concentration increases. The increase in intensity at very low
angles confirms the presence of larger objects, as compared
to C3Glu3 and C4Glu3. The SAXS curves were fitted using
least-square models, for C3- to C5Glu3 at the same high
concentration (Fig. 7A) and for C5Glu3 as a function of
concentration (Fig. 7B). By using a model of a generalized
Gaussian coil, C3Glu3 looks like an oblate ellipsoid. For
C4Glu3 and C5Glu3, a model of rod-like copolymer micelles
has been used, which shows that C4Glu3 is thinner and shorter
than C5Glu3 at their highest concentration, but thicker than
C3Glu3. Cross-sectional radius (1.52 ꢁ 0.04 nm), radius of
gyration and length were therefore determined for each
C5Glu3 concentration (see ESIw, Table S2). SAXS measure-
ments (Rg and Dmax) are in accordance with dynamic light
scattering measurements in terms of evolution of particle
dimensions as the hydrophobic tail increases in length. In
the case of C3- and C4Glu3, DLS is not able to distinguish
between the different species in solution (monomer and small
aggregates), and gives only an average hydrodynamic diameter
of tripods in solution. For C5Glu3, the DLS diameter is
slightly lower than the length determined by SAXS, since for
very long cylinders the diffusion and therefore the hydrodynamic
radius are mainly determined by the diameter of the rod, and not
by its length.27 Differences in behaviour among C3Glu3, C4Glu3
and C5Glu3 have yet to be clearly explained. SAXS studies on
C6Glu3 and C7Glu3 would probably have shed further light on
tripod self-assembly, but their solubility was too low for SAXS
experiments.
P = v/al
(1)
where v is the volume of the chain, l is the length of the chain
and a, the cross-sectional area of the head group. Cross-
sectional head group areas were determined from CMC
measurements as described in the Experimental section. For
single-chain hydrocarbon tails, both the length of the chain
and its volume can be estimated according to Tanford:
v = 3(27.4 + 26.9n)
l = 1.5 + 1.265n
(2)
(3)
where l is the length of the fully extended chain in A, v the
volume of the 3 alkyl chains in A3, and n the number of
hydrocarbon units in each chain.
Tanford et al. studied the factors determining the packing
patterns of amphiphilic molecules and the shape of aggregates.
Amphiphiles having P o 1/3 prefer to form spherical micelles
while those having either 1/3 o P o 12 or 21 o P o 1, prefer to
form either rod-like or cylindrical micelles, or lamellar phases,
respectively. We can observe for all compounds (Table 1) that
the calculated values for each compound show small changes
in the critical packing parameter. The general trend predicted
by the molecular packing parameter does not seem valid for
our new tripod facial surfactants. Actually, this predictive
theory is valid if the micelle core is packed with surfactant
tails leaving no empty space.29 The volume occupied by the
hydrophobic part is certainly greater than that calculated with
eqn (2) because there is an empty space between the 3 chains due
to the presence of the rigid aromatic central core of this new
family of tensioactives. Trappman et al. pointed out that the
interaction of space-demanding dendritic amphiphiles dominates
their self-assembly process.30 Kratzat and Finkelmann have also
reported that asymmetrically branched non-ionic surfactants do
not follow the packing parameter model of Tanford.31 Due to the
very low aggregation number of compounds C3Glu3, C4Glu3
and C5Glu3 measured by SAXS, the aggregation properties of
An interesting question is whether the results obtained by
SAXS studies confirm the concept of the molecular packing
parameter introduced by Tanford.28
According to this theory, a simple and common conceptual
way of relating the geometric shape of an amphiphile and the
c
1174 New J. Chem., 2012, 36, 1170–1179
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
these compounds appear to be similar to those of bile salts.32 In
contrast to classical detergents, where the hydrophilic head group
and the lipophilic flexible aliphatic chains are clearly separated,
bile salt molecules have a lipophilic surface, which is the convex
side of the rigid steroid ring system and a hydrophilic surface
which is the polyhydroxylated concave side of the molecule. Bile
salt micelles also have much smaller aggregation numbers than
classical detergent micelles. Moreover, it is interesting to note
that bile salts behave quite differently in water from classical
detergents.33 They form two types of aggregates, small aggregates
and much larger ones. Small aggregates with an aggregation
number of less than 10 lead to the formation of primary micelles
of almost spherical shapes which transform with increasing
tensioactive concentration into secondary micelles with rod-like
structures. The results obtained by SAXS measurement with
C5Glu3 suggest an association behaviour of this amphiphilic
tripods equivalent to that of bile salts with a face-to-face
arrangement of the hydrophobic parts and the formation of
rod-shaped micelles whose length increases as a function of the
surfactant concentration (see ESIw in Table S2).
grade and used as received unless stated. THF and MeOH
were dried over Na and CH2Cl2 over CaH2 under argon
atmosphere. Commercially anhydrous DMF was stored over
activated molecular sieves of 3 A. The progress of the reactions
was monitored by thin layer chromatography (TLC, Merck
254, silica plates) and the compounds were detected either by
exposure to ultraviolet light (254 nm) or by spraying with a
0.05% permanganate aqueous solution or with an acidic
solution of ceric molybdate following heating at 150 1C. Flash
chromatography purifications were carried out on silica gel
(40–63 mm granulometry). Size exclusion chromatography
1
purifications were carried out on a Sephadex LH-20 resin. H
and 13C spectra were recorded on a Brucker AC 250 spectro-
1
meter at 250 MHz for H and 62.86 MHz for 13C. Chemical
shifts (d values) were reported in ppm downfield from internal
residual solvent as a heteronuclear reference. HRMS(ESI) was
determined on a QStar Elite mass spectrometer.
Synthesis
1,3,5-Tris(chloromethyl)-2,4,6-tris(propylthio)benzene (2a).
To a solution of potassium tert-butoxide (2.27 g, 20.23 mmol)
in DMI (30 mL) was added under argon atmosphere
propanethiol (3.8 mL, 21.76 mmol). The solution was stirred
for 10 min then compound 1 (2.28 g, 5.63 mmol) was added
and the thick mixture was stirred for 16 h at room tempera-
ture. The reaction mixture was poured into water (300 mL)
and extracted with AcOEt (3 ꢂ 100 mL). The organic layer
was washed first with H2O (2 ꢂ 50 mL), then with brine (50
mL), dried with anhydrous MgSO4, filtered and concentrated.
The crude compound was subjected to flash chromatography
on a silica gel column with AcOEt/cyclohexane (2 : 8), giving
the tris(hydroxymethyl) benzenic product as colorless oil (1.43
g, 97%). Rf 0.24 in AcOEt/cyclohexane (3 : 7). 1H NMR
(CDCl3) d 1.03 (t, 3J = 7.34 Hz, 9H, 1.65 (six, 3J = 7.39
Hz, 6H), 2.89 (t, 3J = 7.55 Hz, 6H), 2.9 (br s, 3H), 5.31 (s, 6H).
13C NMR (CDCl3) d 13.5, 23.0, 40.7, 63.3, 137.2, 150.7.
Et3N (1.25 mL, 9 mmol) was added slowly to a solution of
the tris(hydroxymethyl)benzene derivative (1.2 g, 2 mmol) in
dry CH2Cl2 (30 mL) cooled with an ice bath at 5 1C under argon
atmosphere. Methane sulfonyl chloride (610 mL, 7.8 mM) was
added dropwise at 5 1C and the reaction mixture was stirred for
48 h at room temperature. The reaction mixture was monitored
by TLC with CH2Cl2/cyclohexane (1 : 9) as eluent until only one
spot was obtained, and then concentrated to give a crude
compound as an orange oil. Purification by flash chromato-
graphy on a silica gel column with CH2Cl2/cyclohexane (1 : 9)
gave the expected product as a white powder (1.09 g, 83%), Rf
Conclusion
The synthetic route chosen for the synthesis of these new
benzene-based facial amphiphiles allows the selective introduc-
tion in good yield of hydrophilic and hydrophobic parts of the
molecule. Thus, both overall electrical charges and/or water
solubility of the polar head and nature and length of hydro-
phobic parts can be tuned easily. We show that the increased
number of small lipophilic chains and their segregation on one
side of these amphiphilic compound tripods afford them
tensioactive properties. Whatever the length of the hydrocarbon
part, these amphiphilic compounds exhibit self-aggregation
properties in water. They lead to particles whose size and shape
are closely dependent on the length of the hydrophobic tail. Long
hydrocarbon chains (C7Glu3) decrease tripod solubility in water,
and thus their ability to form small aggregates. The attachment
of other types of hydrophilic groups on the polar face of these
tripods should easily offset hydrophobicity, modulate the hydro-
philic lipophilic balance and solve problems of water solubility.
We use SAXS experiments to characterize two different beha-
viours in the self-aggregation of CnGlu3 tripods. The tripods
form aggregates whose size and shape depend on the length of
the hydrophobic tail. Interestingly, although the aggregation
numbers, Nagg, of these facial amphiphiles are very low, this
aggregation behaviour in water is close to that of natural facial
surfactants such as bile salt derivatives. The aggregation process
of C5Glu3 seems to proceed stepwise over a broad concentration
range, as with bile salt detergents. The design of these tripod
facial surfactants and their ability to interact face-to-face with the
hydrophobic part suggest that they are suitable for use in the
solubilization of large hydrophobic nanoparticles. The study of
their ability to handle and pinch-off specifically hydrophobic
nanoparticles in aqueous solution is currently in progress.
1
0.53 in CH2Cl2/cyclohexane (1 : 9), Mp 88.7–89.1 1C. H NMR
3
3
(CDCl3) d 1.07 (t, J = 7.36 Hz, 9H), 1.72 (six, J = 7.36 Hz,
6H), 3.02 (t, 3J = 7.36 Hz, 6H), 5.51 (s, 6H). 13C NMR (CDCl3)
d 13.6, 23.1, 40.9, 46.2, 139.9, 149.5. HRMS (ESI+) calcd for
C18H28Cl3S3 ([M + H]+) 445.0436, found 445.0419.
Compounds 2b–2e were obtained as described previously
for 2a from 5.63 mmol of compound 1 by using the appropriate
thiol (butane-, pentane-, hexane- or heptanethiol) instead of
propane thiol.
General method
All starting materials were commercially available and were
used without further purification. All solvents were of reagent
1,3,5-Tris(chloromethyl)-2,4,6-tris(butylthio)benzene (2b). Yield
865 mg (88%), colorless oil, Rf 0.5 in CH2Cl2/cyclohexane (1 : 9).
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1170–1179 1175

View Article Online
1H NMR (CDCl3) d 0.96 (t, 3J = 7.32 Hz, 9H), 1.49 (six, 3J =
7.3 Hz, 6H), 1.69 (six, 3J = 6.73 Hz, 6H), 3.04 (t, 3J = 7.53 Hz,
6H), 5.51 (s, 6H). 13C NMR (CDCl3) d 13.7, 22.1, 31.6, 38.7,
46.2, 140.0, 149.5. HRMS (ESI+) calcd for C21H34Cl3S3 ([M +
H]+) 487.0936, found 487.0888.
1,3,5-Tris(prop-2-ynyloxymethyl)-2,4,6-tris(pentylthio) benzene
(3c). Yield 492 mg (84%), yellow oil, Rf 0.45 in CH2Cl2/
1
3
cyclohexane (6 : 4). H NMR (CDCl3) d 0.9 (t, J = 7.06 Hz,
4
9H), 1.29–1.42 (m, 12H), 1.58 (m, 6H), 2.49 (t, J = 2.32 Hz,
3H), 2.89 (t, 3J = 7.49 Hz, 6H), 4.33 (d, 4J = 2.36 Hz, 6H), 5.3
(s, 6H). 13C NMR (CDCl3) d 14.0, 22.3, 29.4, 31.2, 39.0, 58.1,
71.6, 74.4, 79.9, 141.4, 147.0.
1,3,5-Tris(chloromethyl)-2,4,6-tris(pentylthio)benzene (2c).
Yield 925 mg (87%), colorless oil, Rf 0.56 in CH2Cl2/cyclo-
hexane (1 : 9). 1H NMR (CDCl3) d 0.93 (t, 3J = 7.12 Hz, 9H),
1,3,5-Tris(prop-2-ynyloxymethyl)-2,4,6-tris(hexylthio)benzene
3
3
1.32–1.46 (m, 12H), 1.71 (p, J = 7.49 Hz, 6H), 3.03 (t, J =
7.63 Hz, 6H), 5.51 (s, 6H). 13C NMR (CDCl3) d 14.0, 22.3,
29.3, 31.1, 39.0, 46.2, 140.0, 149.5. HRMS (ESI+) calcd for
C24H40Cl3S3 ([M + H]+) 529.1328, found 529.1358.
(3d). Yield 488 mg (77%), yellow oil, Rf 0.45 in CH2Cl2/
1
3
cyclohexane (6 : 4). H NMR (CDCl3) d 0.9 (t, J = 6.81 Hz,
9H), 1.26–1.43 (m, 18H), 1.59 (p, 3J = 8.88 Hz, 6H), 2.49
(t, 4J = 2.37 Hz, 3H), 2.89 (t, 3J = 7.5 Hz, 6H), 4.33 (d, 4J =
2.39 Hz, 6H), 5.3 (s, 6H). 13C NMR (CDCl3) d 14.1, 22.6, 28.7,
29.6, 31.5, 39.1, 58.1, 71.6, 74.4, 79.9, 141.4, 147.0.
1,3,5-Tris(chloromethyl)-2,4,6-tris(hexylthio)benzene (2d).
Yield 1.03 g (91%), colorless oil, Rf 0.66 in CH2Cl2/cyclo-
hexane (1 : 9). 1H NMR (CDCl3) d 0.91 (t, 3J = 6.73 Hz, 9H),
1,3,5-Tris(prop-2-ynyloxymethyl)-2,4,6-tris(heptylthio) benzene
3
3
1.29–1.47 (m, 18H), 1.7 (p, J = 7.8 Hz, 6H), 3.02 (t, J =
7.4 Hz, 6H), 5.51 (s, 6H). 13C NMR (CDCl3) d 14.1, 22.5, 28.6,
29.5, 31.4, 39.0, 46.2, 140.0, 149.5. HRMS (ESI+) calcd for
C27H46Cl3S3 ([M + H]+) 571.1854, found 571.1827.
(3e). Yield 452 mg, (67%) yellow oil, Rf 0.45 in CH2Cl2/
1
3
cyclohexane (6:4). H NMR (CDCl3) d 0.9 (t, J = 6.88 Hz,
9H), 1.27–1.4 (m, 24H), 1.59 (p, 3J = 8.88 Hz, 6H), 2.49 (t, 4J =
2.37 Hz, 3H), 2.89 (t, 3J = 7.5 Hz, 6H), 4.33 (d, 4J = 2.39 Hz,
6H), 5.3 (s, 6H). 13C NMR (CDCl3) d 14.1, 22.6, 28.7, 29.6, 31.5,
39.1, 58.1, 71.6, 74.4, 79.9, 141.4, 147.0.
1,3,5-Tris(chloromethyl)-2,4,6-tris(heptylthio)benzene (2e).
Yield 993 mg (81%), colorless oil, Rf 0.38 in CH2Cl2/cyclo-
1
3
hexane (1 : 9). H NMR (CDCl3) d 0.91 (t, J = 6.73 Hz, 9H,
1,3,5-Tris[1-(b-D-glucopyranosyl)-1,2,3-triazole-4-yloxymethyl]-
2,4,6-tris(propylthio) benzene (C3Glu3). To a solution of 3a
(358 mg, 0.5 mmol) in THF (15 mL) at room temperature was
added 2,3,4,6-tetra-O-acetyl-b-D-glucopyranosyl azide (740 mg,
1.8 mmol), DIEA (520 mL, 3 mmol) and CuI (29 mg, 0.15 mmol).
The resulting green mixture was stirred for 24 h at room
temperature. AcOEt (60 mL) was then added and the organic
layer was washed first with saturated aqueous NH4Cl
(3 ꢂ 20 mL), then with brine (20 mL), dried with anhydrous
MgSO4, filtered and concentrated. The crude solid was subjected
to flash chromatography on a silica gel column with a gradient of
AcOEt/CH2Cl2 (1 : 9 to 4:6), giving the acetylated product as a
white powder (677 mg, 83%). Rf 0.26 in AcOEt/CH2Cl2 (4 : 6).
Mp 190.1–192.9 1C. [a]2D5 = ꢀ56.2 (c 1, CH2Cl2). 1H NMR
(CDCl3) d 0.92 (t, 3J = 7.41 Hz, 9H), 1.53 (six, 3J = 7.4 Hz, 6H),
3
CH3), 1.31–1.44 (m, 24H), 1.71 (p, J = 7.65 Hz, 6H), 3.03
(t, 3J = 7.63 Hz, 6H), 5.51 (s, 6H). 13C NMR (CDCl3) d 14.1,
22.7, 28.6, 28.9, 29.0, 29.3, 29.6, 31.8, 39.0, 46.2, 140.0, 149.5.
HRMS (ESI+) calcd for C30H52Cl3S3 ([M + H]+) 613.2203,
found 613.2234.
1,3,5-Tris(prop-2-ynyloxymethyl)-2,4,6-tris(propylthio) benzene
(3a). To a solution of propargyl alcohol (210 mL, 3.6 mmol)
in anhydrous DMF (5 mL) cooled to 0 1C was slowly added
NaH (60% w/w, oil dispersion) (144 mg, 3.6 mM) under argon.
The reaction mixture was stirred for 10 min and a solution of 2a
(656 mg, 1 mmol) in dry THF (5 mL) was slowly added at 0 1C.
The resulting mixture was stirred for 18 h at room temperature.
Then ice was added. The mixture was poured into water (50 mL)
and extracted with AcOEt (3 ꢂ 50 mL). The organic layers
were gathered, washed first with H2O (2 ꢂ 30 mL), then with
brine (30 mL), dried with anhydrous MgSO4, filtered and
concentrated. Purification by flash chromatography on a silica
gel column with CH2Cl2/cyclohexane (6 : 4) as eluent gave the
expected product as yellow oil (329 mg, 65%). Rf 0.36 in
3
1.88 (s, 9H), 2.05–2.09 (3s, 27H), 2.79 (t, J = 7.41 Hz, 6H),
3
3
3
4.02 (ddd, J = 2.82 Hz, J = 4.76 Hz, J = 9.98 Hz, 3H),
4.14 (dd, 3J = 2 Hz, 2J = 12.53 Hz, 3H), 4.3 (dd, 3J = 4.95 Hz,
2J = 12.56 Hz, 3H), 4.82 (s, 6H), 5.21–5.33 (m, 9H), 5.4–5.53
(m, 6H), 5.9 (d, 3J = 9.0 Hz, 3H), 7.83 (s, 3H). 13C NMR (CDCl3)
d 14.0, 20.2, 20.6, 20.7, 23.0, 40.8, 61.6, 64.2, 67.7, 70.2, 72.0, 72.8,
75.0, 85.7, 121.1, 141.1, 146.0, 147.1, 168.9, 169.4, 170.0, 170.6.
1
3
CH2Cl2/cyclohexane (8 : 2). H NMR (CDCl3) d 1.02 (t, J =
3
4
7.34 Hz, 9H), 1.6 (p, J = 7.34 Hz, 6H), 2.5 (t, J = 2.38 Hz,
3H), 2.88 (t, 3J = 7.24 Hz, 6H), 4.34 (d, 4J = 2.35 Hz, 3H), 5.33
(s, 6H). 13C NMR (CDCl3) d 13.6, 23.0, 41.0, 58.1, 71.6, 74.4,
79.9, 141.4, 147.0.
Deacetylated compound. An acetylated compound (550 mg,
0.3 mmol) was dissolved in 50 mL of a dry mixture of THF/
MeOH (1 : 4). A sufficient catalytic quantity of sodium meth-
oxide was added to reach a pH of 9–10 and the reaction
mixture was stirred for 24 h at room temperature. The solution
was neutralized with the resin acid IRC-50 (pH = 7), filtered
and concentrated. The solid was purified by Sephadex LH-20
size exclusion chromatography with methanol, giving the
expected product as a white powder (246 mg, 73%). Rf 0.18
Compounds 3b–3e were obtained as described previously
for 3a by using the appropriate 1,3,5-tris(chloromethyl)-2,4,6-
tris(alkylthio) benzene 2b–2e (0.5 mmol) instead of 2a.
1,3,5-Tris(prop-2-ynyloxymethyl)-2,4,6-tris(butylthio)benzene
(3b). Yield 339 mg (62%), yellow oil, Rf 0.43 in CH2Cl2/
1
3
cyclohexane (7: 3). H NMR (CDCl3) d 0.92 (t, J = 7.2 Hz,
3
3
9H), 1.44 (six, J = 8.0 Hz, 6H), 1.57 (p, J = 7.01 Hz, 3H),
2.49 (t, 4J = 2.37 Hz, 3H), 2.9 (t, 3J = 7.5 Hz, 6H), 4.33
in AcOEt/MeOH/H2O (7 : 2 : 1). Mp 182.8–184.1 1C. [a]D25
=
ꢀ8.0 (c 1, MeOH). HRMS (ESI+) calcd for C45H71N9O18S3
([M + 2H]2+): 560.7034, found 560.7033. 1H NMR (DMSO)
d 0.90 (t, 3J = 7.4 Hz, 9H), 1.45 (six, 3J = 7.23 Hz, 6H),
4
(d, J = 2.39 Hz, 6H), 5.3 (s, 6H). 13C NMR (CDCl3) d 14.1,
22.6, 29.6, 39.1, 58.1, 71.6, 74.4, 79.9, 141.4, 147.0.
c
1176 New J. Chem., 2012, 36, 1170–1179
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
3
3
3
2.79 (t, J = 7.28 Hz, 6H), 3.23 (m, 3H), 3.37–3.48 (m, 9H),
3H), 5.36 (d, J = 6.03 Hz, 3H), 5.53 (d, J = 9.23 Hz, 3H),
8.23 (s, 3H). 13C NMR (DMSO) d14.3, 22.2, 29.2, 30.9, 38.2,
61.2, 64.2, 70.0, 71.8, 72.6, 77.4, 80.4, 87.9, 123.5, 140.4,
144.1, 147.3.
3.66–3.77 (m, 6H), 4.63 (t, 3J = 7.5 Hz, 3H), 4.69 (s, 6H), 5.18
(d, 3J = 5.36 Hz, 3H), 5.19 (s, 6H), 5.3 (d, 3J = 4.81 Hz, 3H),
5.37 (d, 3J = 6.06 Hz, 3H), 5.53 (d, 3J = 9.23 Hz, 3H),
8.27 (s, 3H). 13C NMR (DMSO) d 13.8, 22.9, 39.0, 61.6, 64.2,
70.0, 71.8, 72.5, 77.4, 80.4, 87.9, 123.6, 140.4, 144.1, 147.3.
Compounds CnGlu3 were obtained as previously described
for the synthesis of C3Glu3 by using the appropriate 1,3,5-
tris(prop-2-ynyloxymethyl)-2,4,6-tris(alkylthio)benzene 3b–3e
(0.5 mmol) instead of 3a, then 0.3 mmol of acetylated
compound.
1,3,5-Tris[1-(b-D-glucopyranosyl)-1,2,3-triazole-4-yloxymethyl]-
2,4,6-tris(hexylthio) benzene (C6Glu3)
Acetylated compound. Yield 643 mg (73%), white powder,
Rf 0.47 in AcOEt/CH2Cl2 (4 : 6). Mp 123.7–125.4 1C. [a]D25
=
ꢀ54.8 (c 1, CH2Cl2). 1H NMR (CDCl3) d 0.87 (t, 3J = 7.0 Hz,
3
9H), 1.25–1.35 (m, 18H), (p, J = 7.38 Hz, 6H), 1.89 (s, 9H),
2.05–2.11 (3s, 27H), 2.79 (t, 3J = 7.28 Hz, 6H), 4.02 (m, 3H),
3
2
3
1,3,5-Tris[1-(b-D-glucopyranosyl)-1,2,3-triazole-4-yloxymethyl]-
2,4,6-tris(butylthio) benzene (C4Glu3)
4.14 (dd, J = 1.75 Hz, J = 12.53 Hz, 3H), 4.3 (dd, J =
4.96 Hz, 2J = 12.71 Hz, 3H), 4.81 (s, 6H), 5.22–5.28 (m, 9H),
5.39–5.53 (m, 6H), 5.9 (d, 3J = 8.85 Hz, 3H), 7.83 (s, 3H). 13
Acetylated compound. Yield 627 mg (75%), white powder,
Rf 0.32 in AcOEt/CH2Cl2 (4 : 6). Mp 194.3–195.9 1C. [a]D25
C
=
ꢀ54.9 (c 1, CH2Cl2). 1H NMR (CDCl3) d 0.86 (t, 3J = 7.12 Hz,
NMR (CDCl3) d 14.1, 20.2, 20.5, 20.6, 20.7, 22.5, 28.6, 29.6,
31.4, 38.9, 61.6, 64.2, 67.7, 70.2, 71.9, 72.8, 75.0, 85.7, 121.1,
141.1, 146.0, 147.0, 168.8, 169.4, 170.0, 170.5.
6H), 1.34 (six, 3J = 7.65 Hz, 6H), 1.45–1.68 (m, 6H), 1.9 (s, 9H),
3
2.04–2.09 (3s, 27H), 2.81 (t, J = 7.41 Hz, 6H), 4.02 (m, 3H),
3
4.14 (dd, J = 1.73 Hz, J = 12.55 Hz, 3H), 4.3 (dd, J =
2
3
Deacetylated compound. Yield 347 mg (88%), white powder,
Rf 0.46 in AcOEt/MeOH/H2O (7 : 2 : 1). Mp 186.8–188.1 1C.
[a]2D5 = ꢀ8.2 (c 1, MeOH). HRMS (ESI+) calcd for
C54H89N9O18S3 ([M + 2H]2+): 623.7738, found 623.7739.
2
4.85 Hz, J = 12.62 Hz, 3H), 4.83 (s, 6H), 5.22–5.33 (m, 9H),
5.4–5.53 (m, 6H), 5.9 (d, J = 8.75 Hz, 3H), 7.84 (s, 3H). 13C
3
NMR (CDCl3) d 13.7, 20.2, 20.6, 20.7, 22.1, 31.7, 38.6, 61.6,
64.2, 67.7, 70.3, 72.0, 72.8, 75.0, 85.7, 121.1, 141.1, 146.0, 147.1,
168.8, 169.4, 170.0, 170.5.
1H NMR (DMSO) d 0.82 (t, J = 6.96 Hz, 9H), 1.18–1.25
3
3
(m, 12H), 1.42 (m, 6H), 2.79 (t, J = 7.24 Hz, 6H), 3.19–3.26
3
(m, 3H), 3.34–3.45 (m, 9H), 3.66–3.78 (m, 6H), 4.62 (t, J =
Deacetylated compound. Yield 289 mg (87%), white powder,
Rf 0.4 in AcOEt/MeOH/H2O (7 : 2 : 1). Mp 184.3–186.0 1C.
[a]2D5 = ꢀ8.4 (c 1, MeOH). HRMS (ESI+) calcd for
5.56 Hz, 3H), 4.68 (s, 6H), 5.18 (2s, 9H), 5.31 (d, 3J = 4.67 Hz,
3
3
3H), 5.36 (d, J = 6.03 Hz, 3H), 5.53 (d, J = 9.23 Hz, 3H),
8.23 (s, 3H). 13C NMR (DMSO) d 14.3, 22.2, 29.2, 30.9, 38.2,
61.2, 64.2, 70.0, 71.8, 72.6, 77.4, 80.4, 87.9, 123.5, 140.4,
144.1, 147.3.
C48H77N9O18S3 ([M
+
2H]2+): 581.7269, found
581.7275.1H NMR (DMSO) d 0.82 (t, 3J = 7.03 Hz, 9H),
1.26–1.41 (m, 12H), 2.81 (t, 3J = 6.91 Hz, 6H), 3.22 (m, 3H),
3
3.37–3.46 (m, 9H), 3.67–3.79 (m, 6H), 4.63 (t, J = 5.09 Hz,
1,3,5-Tris[1-(b-D-glucopyranosyl)-1,2,3-triazole-4-yloxymethyl]-
2,4,6-tris(heptylthio) benzene (C7Glu3)
3H), 4.69 (s, 6H), 5.19 (s, 9H), 5.31 (d, 3J = 4.49 Hz, 3H), 5.36
(d, 3J = 5.96 Hz, 3H), 5.53 (d, 3J = 9.2 Hz, 3H), 8.25 (s, 3H).
13C NMR (DMSO) d 14.1, 21.9, 31.7, 38.0, 61.2, 64.2, 70.0,
71.8, 72.6, 77.4, 80.4, 87.9, 123.5, 140.4, 144.1, 147.3.
Acetylated compound. Yield 645 mg (72%), white powder,
Rf 0.54 in AcOEt/CH2Cl2 (4 : 6). Mp 134.2–136.7 1C. [a]D25
=
ꢀ50.8 (c 1, CH2Cl2). 1H NMR (CDCl3) d 0.88 (t, 3J = 6.9 Hz,
3
9H), 1.25 (m, 24H), 1.47 (p, J = 7.54 Hz, 3H), 1.89 (s, 9H),
1,3,5-Tris[1-(b-D-glucopyranosyl)-1,2,3-triazole-4-yloxymethyl]-
2,4,6-tris(pentylthio) benzene (C5Glu3)
2.05–2.11 (3s, 27H), 2.79 (t, 3J = 7.3 Hz, 6H), 4.02 (ddd,
3J = 1.99 Hz, J = 5.27 Hz, J = 10.17 Hz, 3H), 4.14 (dd,
3
3
Acetylated compound. Yield 685 mg (79%), white powder,
Rf 0.42 in AcOEt/CH2Cl2 (4 : 6). Mp 179.5–181.2 1C. [a]D25
3J = 1.95 Hz, J = 12.54 Hz, 3H), 4.3 (dd, J = 4.97 Hz,
2
3
=
ꢀ55.9 (c 1, CH2Cl2). 1H NMR (CDCl3) d 0.86 (t, 3J = 7.1 Hz,
2J = 12.63 Hz, 3H), 4.81 (s, 6H), 5.22–5.28 (m, 9H), 5.4–5.53
(m, 6H), 5.9 (d, J = 8.98 Hz, 3H), 7.83 (s, 3H). 13C NMR
3
9H), 1.25–1.31 (m, 12H), 1.5 (p, 3J = 7.26 Hz, 6H), 1.88
3
(s, 9H), 2.05–2.08 (3s, 27H), 2.8 (t, J = 7.41 Hz, 6H), 4.02
(CDCl3) d 14.1, 20.2, 20.5, 20.6, 20.7, 22.6, 28.9, 29.7, 31.7,
38.9, 61.6, 64.2, 67.7, 70.2, 71.9, 72.8, 75.0, 85.7, 121.2, 141.1,
146.0, 147.0, 168.8, 169.4, 170.0, 170.5.
(m, 3H), 4.14 (dd, 3J = 1.98 Hz, 2J = 12.55 Hz, 3H), 4.3
(dd, 3J = 4.83 Hz, 2J = 12.52 Hz, 3H), 4.81 (s, 6H), 5.21–5.32
3
(m, 9H), 5.39–5.52 (m, 6H), 5.9 (d, J = 8.93 Hz, 3H), 7.83
(s, 3H). 13C NMR (CDCl3) d 14.0, 20.2, 20.6, 20.7, 22.3, 29.3,
31.1, 38.9, 61.6, 64.2, 67.7, 70.0, 71.9, 72.7, 75.0, 85.7, 121.1,
141.1, 146.0, 147.1, 168.8, 169.4, 170.0, 170.6.
Deacetylated compound. Yield 341 mg (87%), white powder,
Rf 0.53 in AcOEt/MeOH/H2O (7 : 2 : 1). Mp 183.0–184.7 1C.
[a]2D5 = ꢀ7.4 (c 1, MeOH). HRMS (ESI+) calcd for
C57H95N9O18S3 ([M + 2H]2+): 644.7973, found 644.7978.
3
Deacetylated compound. Yield 346 mg (95%), white powder,
Rf 0.46 in AcOEt/MeOH/H2O (7 : 2 : 1). Mp 187.8–189.3 1C.
[a]2D5 = 17.7 (c 1, MeOH). HRMS (ESI+) calcd for
C51H83N9O18S3 ([M + 2H]2+): 602.7504, found 602.7506.
1H NMR (DMSO) d 0.85 (t, J = 6.95 Hz, 9H), 1.22–1.41
(m, 30H), 2.79 (t, 3J = 6.77 Hz, 6H), 3.17–3.27 (m, 3H),
3
3.36–3.45 (m, 9H), 3.67–3.76 (m, 6H), 4.62 (t, J = 5.54 Hz,
3H), 4.68 (s, 6H), 5.17 (2s, 9H), 5.3 (d, 3J = 4.73 Hz), 5.35 (d,
3
3
1H NMR (DMSO) d 0.82 (t, J = 6.96 Hz, 9H), 1.18–1.25
3J = 6.06 Hz, 3H), 5.53 (d, J = 9.21 Hz, 3H), 8.24 (s, 3H).
3
(m, 12H), 1.42 (m, 6H), 2.79 (t, J = 7.24 Hz, 6H), 3.19–3.26
13C NMR (DMSO) d 14.4, 22.5, 28.7, 28.8, 29.6, 31.6, 38.3,
61.3, 64.2 70.0, 71.7, 72.6, 77.4, 80.4, 87.9, 123.5, 140.3,
144.0, 147.2.
3
(m, 3H), 3.34–3.45 (m, 9H), 3.66–3.78 (m, 6H), 4.62 (t, J =
5.56 Hz, 3H), 4.68 (s, 6H), 5.18 (2s, 9H), 5.31 (d, 3J = 4.67 Hz,
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1170–1179 1177

View Article Online
1,3,5-Tris(prop-2-ynyloxymethyl) benzene (4). To a solution
of propargyl alcohol (311 mg, 5.5 mmol) in anhydrous DMF
(8 mL) cooled to 0 1C, was added, little by little, under argon
atmosphere a 60% NaH (131 mg, 5.5 mmol) oil dispersion.
The reaction mixture was stirred for 10 min and a solution of
tris(bromomethyl)benzene (505 mg, 1.4 mmol) in dry DMF
(1 mL) was slowly added at 0 1C. The resulting mixture was
stirred for 18 h at room temperature and ice was added to
neutralize the basic solution. The mixture was poured into water
(50 mL) and extracted with AcOEt (3 ꢂ 50 mL). The organic
layers were gathered, washed first with H2O (2 ꢂ 30 mL), then
with brine (30 mL), dried with anhydrous MgSO4, filtered and
concentrated. Purification by flash chromatography on a silica
gel column with CH2Cl2/cyclohexane (85 : 15) gave the expected
product as yellow oil (511 mg, 48%). Rf 0.51 in CH2Cl2/
Dynamic light scattering measurements
Hydrodynamic diffusion coefficients and polydispersity index
of CnGlu3 solutions at different concentrations were measured
at 25 1C using a Zetasizer Nano-S model 1600 (Malvern
Instruments Ltd., UK) equipped with a He–Ne laser (l =
633 nm, 4.0 mW). The solutions were prepared and stored
at room temperature overnight before measurements. The
samples were filtered through a 0.45 mm filter and placed in
a 45 mL-cuvette. The experimental run time was 10 s and
experimental data are reported as an average of 10 values from
10 scans. Scattered light intensity was measured at a scattering
angle of 1731 relative to the laser source (backscattering
detection). The time-dependent correlation function was
analyzed using an exponential decay model. When there was
a low polydispersity index (PDI o 20%), a Stokes radius (RS)
of particles was estimated from the diffusion coefficient (D)
using the Stokes–Einstein equation D = kBT/6pZRS, where
kB is the Boltzmann constant, T the absolute temperature and
Z the viscosity of the solvent. This enabled intensity size
distribution to be converted into volume size distribution via
the manufacturer’s software, according to the Mie theory
which gives a good estimation of the mean size of particles
in solution.
1
4
cyclohexane (9 : 1). H NMR (CDCl3) d 2.5 (t, J = 2.38 Hz,
3H), 4.22 (d, 4J = 2.38 Hz, 6H), 4.64 (s, 6H), 7.33 (s, 3H).
13C NMR (CDCl3) d 57.33, 71.30, 74.77, 79.55, 127.12, 137.93.
1,3,5-Tris[1-(b-D-glucopyranosyl)-1,2,3-triazole-4-yloxymethyl]
benzene (C0Glu3). Compound C0Glu3 was obtained as
described for the synthesis of C3Glu3 by using the appropriate
1,3,5-tris(prop-2-ynyloxymethyl)-benzene instead of 3a.
Acetylated compound. Yield 481 mg (50%), white powder,
Rf 0.48 in AcOEt/CH2Cl2 (8 : 2). 1H NMR (CDCl3)
d 2.00–2.06 (3s, 26H), 4.04–4.32 (m, 9H), 4.55 (s, 6H) 4.62
(s, 6H), 5.27–5.88 (m, 9H), 6 (d, 3J = 9.04 Hz, 3H), 7.32
(s, 3H), 8.04 (s, 3H). 13C NMR (CDCl3) d 20.1, 20.5, 20.6,
20.7, 61.5, 62.7, 67.7, 70.3, 71.6, 72.7, 74.9, 85.6, 121.7, 127.3
138.2, 145.6 168.9, 169.5, 170.0, 170.5.
Determination of log K0w values
We used reversed-phase high-performance liquid chromato-
graphy, a common and rapid technique for indirect determi-
nation of the octanol/water partition coefficient as a measure
of the lipophilicity of organic compounds. The facial amphi-
philes were dissolved in MeOH at a concentration of 1 g Lꢀ1
and injected into a reverse-phase column (C18, 5 mm granu-
lometry, 250 ꢂ 4.6 mm) at room temperature.25 Compounds
were eluted with various MeOH/H2O mixtures (from 95 : 5 to
65 : 35). The measurements were performed at a flow rate of
0.8 mL minꢀ1 and detected at 220 nm. The value of log k0w
was calculated as log k0w = log[(t–t0)/t0], where t is the
retention time of the compound and t0 is the elution time of
MeOH, which is not retained on the column. log k0w values
were obtained by extrapolation of the linear regression to
0% MeOH.
Deacetylated compound. Yield 308 mg (87%), white powder,
Rf 0.51 in AcOEt/MeOH/H2O (5 : 4 : 3). Mp 175.0–176 1C. [a]D25
= ꢀ7.4 (c 1, MeOH). 1H NMR (DMSO) d 2.51–3.47 (m, 24H),
3.68–3.82 (m, 6H), 4.15 (m, 3H), 4.57 (s, 6H), 4.60 (s, 6H), 4.68
(m, 3H), 5.24 (m, 3H), 5.39–5.46 (m, 9H), 5.54 (d, 3J = 9.2 Hz,
3H), 7.27 (s, 3H), 8.38 (s, 3H). 13C NMR (DMSO-d6) d 49.06,
61.20, 63.35, 69.99, 71.67, 72.51, 77.39, 80.40, 87.96, 123.82,
126.43, 138.77, 144.17.
Surface tension measurements
Small angle X-ray scattering (SAXS)
The surface activity of facial amphiphiles in solution at the air/
water interface was determined by the Wilhelmy plate technique
The structure of particles in solution (radius of gyration,
dimensions of self-assembled structures, molecular mass) was
characterized by SAXS using Synchrotron radiation on the
ID14-eh3 beamline at European Synchrotron Research Facility
(Grenoble, France). Scattering patterns were measured at several
solute concentrations ranging from 2 to 40 g Lꢀ1 in H2O. For
a sample–detector distance of 1.83 m and an X-ray wavelength
l = 0.0931 nm, a 0.01 o q o 3.5 nmꢀ1 range of momentum
transfer was covered. To avoid radiation damage during the
scattering experiments, the sample was kept circulating in the cell
while the data were collected in 10 successive 30 s frames.
Individual frames were averaged after normalization to the
intensity of the incident beam and corrected for detector response,
and the scattering of the buffer was subtracted. The difference
curves were scaled for the solute concentration. All data mani-
pulations were performed via the program package PRIMUS.34
using a Kruss K100 tensiometer controlled by Labdesk software
¨
(Kruss, Germany). All solutions were prepared at least 12 hours
¨
before the measurements with water purified using a Milli-Q
system (Millipore; resistivity = 18.2 MO cm; surface tension =
72.8 mN mꢀ1). A 20 mL initial volume of the facial amphiphile
solution was taken in a glass trough, and surface tensions were
determined by a dilution technique. The platinum plate was
cleaned by flaming before experiments. All measurements were
carried out at 25 1C and repeated 3 times unless otherwise noted.
The surface excess G at the air–water interface was calculated by
the Gibbs adsorption isotherm equation G = ꢀ(1/RT)(dg/dlog C),
where g is the surface tension (N mꢀ1) at the surfactant concen-
tration C (mol Lꢀ1). The occupied area (Amin) per surfactant
molecule was calculated from Amin = 1/NAG, where NA is
Avogadro’s number.
c
1178 New J. Chem., 2012, 36, 1170–1179
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
The forward scattering values I(q = 0) and the radii of gyration
Rg were evaluated using the Guinier approximation assuming
that, at very small angles (qRg o 1), the intensity is represented as
I(q) = I(0)exp(ꢀ(qRg)2/3). For rod-like particles, scattering
curves were analyzed by plotting qI(q) E exp(ꢀ(qRg)2/2). The
scattering data were separated into two parts: a low-angle range
where the length of the rod L = O12 Rg can be determined and a
higher-angle region where the diameter can be obtained from
data at angles where 2p/L o q o 1/Rc (Rc is the cross-sectional
radius of gyration). The molecular weight (MW) of the solutes
was estimated by comparing their forward scattering intensity
with that of a reference solution35 of bovine serum albumin
(MWBSA = 66 kDa), corrected for electron density and partial
specific volume of solutes. Partial specific volumes were estimated
from the chemical compositions of the molecules. Nagg is calcu-
lated by dividing the aggregate molecular weight obtained from
the forward intensity by the molecular weight of the monomer.
The SAXS data in Fig. 7 were further analyzed using the SASfit
software package.36 For C4Glu3 and C5Glu3, a model from Jan
Skov Pedersen of rod-like copolymer micelles was used. This
model worked reasonably well except for sample C3Glu3, which
needed to be treated with a model of a generalized Gaussian coil.
J. Am. Chem. Soc., 2010, 132, 16750–16752; (b) D. T. McQuade,
D. G. Barett, J. M. Desper, R. K. Hayashi and S. H. Gellman,
J. Am. Chem. Soc., 1995, 117, 4862–4869.
11 F. M. Menger and J. L. Sorrels, J. Am. Chem. Soc., 2006, 128,
4960–4961.
12 J. A. A. W. Hlemans, R. R. J. Slangen, A. E. Rowan and
R. J. M. Nolte, J. Org. Chem., 2003, 68, 9040–9049.
13 R. Roy and J. M. Kim, Angew. Chem., Int. Ed., 1999, 38,
369–372.
14 M. Kellermann, W. Bauer, A. Hirsch, B. Schade, K. Ludwig and
C. Bottcher, Angew. Chem., Int. Ed., 2004, 43, 2959–2962.
¨
15 (a) S. Jebors, F. Fache, S. Balme, F. Devoge, M. Monachino,
S. Cecillon and A. W. Coleman, Org. Biomol. Chem., 2008, 6,
319–329; (b) K. Suwinska, O. Shkurenko, C. Mbemba, A. Leydier,
S. Jebors, A. W. Coleman, R. Matar and P. Falson, New J. Chem.,
2008, 32, 1988–1998.
16 A. P. H. J. Schenning, C. Elissen-Roman, J.-W. Weener, M. W. P.
L. Baars, S. J. van der Gaasdt and E. W. Meijer, J. Am. Chem.
Soc., 1998, 120, 8199–8208.
17 K. Ariga, T. Urakawa, A. Michiue, Y. Sasaki and J.-I. Kikuchi,
Langmuir, 2000, 16, 9147–9150.
18 Y. Chen, A. V. Ambade, D. R. Vutukuri and S. Thayumanavan,
J. Am. Chem. Soc., 2006, 128, 14760–14761.
19 K. V. Kilway and J. S. Siegel, Tetrahedron, 2001, 57, 3615–3627.
20 J. E. Anthony, S. I. Khan and Y. Rubin, Tetrahedron Lett., 1997,
38, 3499–3627.
21 M. Gingras, J.-M. Raimundo and Y. M. Chabre, Angew. Chem.,
Int. Ed., 2006, 45, 1686–1712.
22 (a) H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew. Chem.,
Int. Ed., 2001, 40, 2004–2021; (b) V. V. Rostovtsev, L. G. Green,
V. V. Fokin and K. B. Sharpless, Angew. Chem., Int. Ed., 2002, 41,
2596–2599.
23 F. Fazio, M. C. Bryan, O. Blixt, J. C. Paulson and C. H. Wong,
J. Am. Chem. Soc., 2002, 124, 14397–14402.
24 C. Tanford, in The hydrophobic effect: formation of micelles and
biological membranes, Krieger publishing company, Malabar Florida,
2nd edn, 1991.
25 T. Braumann, J. Chromatogr., A, 2008, 373, 191–225.
26 D. Svergun, J. Appl. Crystallogr., 1992, 25, 495–503.
27 R. Pecora, Basic Concepts Scattering and Time Correlation
Functions, in Soft Matter Characterization, ed. Borsali and
Percora, Springer, 2007, vol. 1, pp. 1–38.
Acknowledgements
We are grateful to the European Synchrotron Radiation
Facility for provision of synchrotron radiation facilities and
to Petra Pernot for assistance in using beamline ID14-eh3. We
would like to thank Professor Marc Gingras for his helpful
discussion concerning the synthesis of tripods, and Marjorie
Sweetko for English language revision.
References
28 (a) C. Tanford, J. Phys. Chem., 1972, 76, 3020–3024;
(b) Israelachvili, D. J. Mitchell and B. W. Ninham, J. Chem.
Soc., Faraday Trans. 2, 1976, 72, 1525–1568; (c) K. Kratzat and
H. Finkelmann, Langmuir, 1996, 12, 1765–1770.
1 (a) J. H. Fuhrhop and T. Wang, Chem. Rev., 2004, 104, 2901–2938;
(b) F. M. Menger and J. S. Keiper, Angew. Chem., Int. Ed., 2000,
39, 1906–1920; (c) M. Wathier, A. Polidori, K. Ruiz, A.-S. Fabiano
and B. Pucci, New J. Chem., 2001, 25, 1588–1599.
29 R. Ngarajan, Langmuir, 2002, 18, 31–38.
30 B. Trappmann, K. Ludwig, M. R. Radowski, A. Shukla, A. Mohr,
2 Y. Zhao, Curr. Opin. Colloid Interface Sci., 2007, 12, 92–97.
3 Y. Cheng, D. M. Ho, C. R. Gottlieb, D. Kahne and M. A. Bruck,
J. Am. Chem. Soc., 1992, 114, 7319–7320.
4 P. Venkatesan, Y. Cheng and J. D. Kahne, J. Am. Chem. Soc.,
1994, 116, 6955–6956.
´ ´
5 Y. R. Vandenburg, B. D. Smith, M. N. Perez-Payan and
H. Rehage, C. Bottcher and R. Haag, J. Am. Chem. Soc., 2010,
¨
132, 11119–11124.
31 K. Kratzat and H. Finkelmann, J. Colloid Interface Sci., 1996, 181,
542–550.
32 P. Garidel, A. Hildebrand, R. Neubert and A. Blume, Langmuir,
2000, 16, 5267–5275.
33 A. Hidebrand, P. Garidel, R. Neubert and A. Blume, Langmuir,
2004, 20, 320–328.
A. P. Davis, J. Am. Chem. Soc., 2000, 122, 3252–3253.
6 U. Taotafa, D. B. McMullin, S. C. Lee, L. D. Hansen and
P. B. Savage, Org. Lett., 2000, 2, 4117–4120.
7 P. B. Savage, Eur. J. Org. Chem., 2002, 759–768.
34 P. V. V. Konarev, V. Vladimir, A. V. Sokolova, M. H. J. Koch and
D. Svergun, J. Appl. Crystallogr., 2003, 36, 1277–1282.
35 E. Mylonas and D. I. Svergun, J. Appl. Crystallogr., 2007, 40,
s245–s249.
36 J. Kohlbrecher, SASfit ver. 0.93.3; available at http://kur.web.psi.
ch/sans1/SANSSoft/sasfit.html.
8 H. M. Willemem, A. T. M. Marcelis and E. J. R. Sudholter,
¨
Langmuir, 2003, 19, 2588–2591.
9 Z. Zhang, J. Yan and Y. Zhao, Langmuir, 2005, 21, 6235–6239.
10 (a) P. S. Chae, K. Gotfryd, J. Pacyna, L. J. W. Miercke, S. G.
F. Rasmussen, R. A. Robbins, R. R. Rana, C. J. Loland,
B. Kobilka, R. Stroud, B. Byrne, U. Gether and S. H. Gellman,
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1170–1179 1179