Calixarene-based surfactants: Evidence of structural reorganization upon micellization

Article abstract of DOI:10.1021/la204004h

The self-aggregation of five amphiphilic p-sulfonatocalix[n]arenes bearing alkyl chains at the lower rim was investigated by NMR spectroscopy and electrical conductivity. The critical micelle concentration was determined, and the tendency of this special class of surfactants to self-aggregate in aqueous solution was analyzed as a function of the alkyl chain length and the number of aromatic units in the macrocyclic ring. The structure of the surfactants in the monomeric and micellized states was elucidated by means of ¹H NMR and, in the case of the calix[6]arene derivative, with 2D NMR experiments. While all amphiphilic calix[4]arenes studied here are blocked in the cone conformation, in the monomeric state the calix[6]arene adopts a pseudo-1,2,3-alternate conformation and the calix[8]arene is conformationally mobile. These calixarenes undergo an aggregation-induced conformational change, adopting the cone conformation in the micelles. The structure and size of the aggregates were studied by diffusion ordered spectroscopy (DOSY) experiments, and the results indicate that these surfactants self-assemble into ellipsoidal micelles.

Full text of DOI:10.1021/la204004h

Article
pubs.acs.org/Langmuir
Calixarene-Based Surfactants: Evidence of Structural Reorganization
upon Micellization
Nuno Basílio,† Luis Garcia-Rio,*,† and Manuel Martín-Pastor‡
†Departamento de Química Física y Centro Singular de Investigacion
‡Unidad de Resonancia Magnet
ica (RIAIDT), Universidad de Santiago, 15782 Santiago, Spain
́ ́
en Química Biologica y Materiales Moleculares (CIQUS) and
́
S
* Supporting Information
ABSTRACT: The self-aggregation of five amphiphilic p-sulfonatocalix[n]arenes bearing alkyl chains at the lower rim was
investigated by NMR spectroscopy and electrical conductivity. The critical micelle concentration was determined, and the
tendency of this special class of surfactants to self-aggregate in aqueous solution was analyzed as a function of the alkyl chain
length and the number of aromatic units in the macrocyclic ring. The structure of the surfactants in the monomeric and
micellized states was elucidated by means of 1H NMR and, in the case of the calix[6]arene derivative, with 2D NMR experiments.
While all amphiphilic calix[4]arenes studied here are blocked in the cone conformation, in the monomeric state the calix[6]arene
adopts a pseudo-1,2,3-alternate conformation and the calix[8]arene is conformationally mobile. These calixarenes undergo an
aggregation-induced conformational change, adopting the cone conformation in the micelles. The structure and size of the
aggregates were studied by diffusion ordered spectroscopy (DOSY) experiments, and the results indicate that these surfactants
self-assemble into ellipsoidal micelles.
the hydrophilic headgroup is also important to control the
aggregation behavior of this type of amphiphile, and it was
showed that calixarenes in the cone conformation can also form
vesicles.11,13,16
When compared with conventional surfactants, amphiphilic
calixarenes have lower critical micelle concentration (cmc) values
and tend to present exchange rates between monomers in bulk
solution and in the micelles several orders of magnitude slower
than that observed for the formers,14,15,17 though these prop-
erties are not yet well understood. Furthermore, calixarene-
based amphiphiles can display special properties such as self-
assembling into completely uniform and structurally persistent
micelles, as reported by Hirsch’s group.12
Among the several calixarene-based surfactants described in
the literature, p-sulfonatocalix[n]arenes bearing alkyl groups at
the lower rim were the first to be described and are probably
the most studied.7−9,17 As recognized by Shinkai et al.,9 the ag-
gregation properties can be significantly affected by the calixarene
architecture, being the length of the alkyl chain the most im-
portant factor contributing to the variations on the aggregation
INTRODUCTION
■
Calix[n]arenes1 are macrocyclic oligomers made of phenol
units linked by methylene bridges. They can be readily func-
tionalized at the phenolic hydroxyl groups (lower rim) and at
the para positions (upper rim) and give rise to wide range of
structural derivatives. Research on calixarene chemistry was
initially directed to the design of host compounds with receptor
abilities. Depending on the functional groups introduced on
the calixarene framework, these compounds can display high
affinity and selectivity for the inclusion of neutral, cationic, or
anionic guests. However, the versatility of the calixarenes is not
limited to host−guest chemistry, and when properly function-
alized, they can be applied in diverse areas such as self-assembly
on monolayers2 and nanoparticles,3 formation of rotaxanes and
molecular machines,4 dendrimers,5 and surfactants.6
Calixarene-based surfactants can be obtained by introducing
hydrophilic groups at the upper or lower rim and hydrophobic
groups (generally alkyl chains) at the opposite rim.7−16 Shinkai
et al.10 showed that the conformation adopted by the calixarene
is a crucial parameter in modulating their aggregation behavior
and suggested that the cone shaped conformation is ideal for
the formation of globular micelles. However, the conformation
is not the only factor that determines the type of self-assembled
aggregates formed from amphiphilic calixarenes. The nature of
Received: October 5, 2011
Revised: December 28, 2011
Published: December 29, 2011
© 2011 American Chemical Society
2404
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
1H NMR, COSY, ROESY, and DOSY spectra were recorded at
25 °C in a Varian Inova 400. The DOSY spectra were acquired with
the standard stimulated echo pulse sequence using LED and bipolar
gradients pulses.19 Rectangular-shaped pulse field gradients of duration
2 ms were applied with a power level linearly incremented from 2.1 to
64.3 G cm−1 in 20 steps. The gradient power was previously calibrated
with the same DOSY experiment with a reference sample of 99% D2O
at 25 °C (D = 1.872 × 10−9 m2 s−1). To obtain reliable results of the
diffusion coefficient, the diffusion time Δ of the experiment was opti-
mized for each sample to a value between 60 and 200 ms so as to
capture smoothly the attenuation of the signal intensity, while afford-
ing the maximum difference in intensity between the traces with the
maximum and minimum gradient power. The raw data were processed
using the MestreC program (Mestrelab Research Inc.).
Electrical conductivity was measured by using a Radiometer CDM3
conductivity meter with a cell constant of 0.968 cm−1. The conduc-
tivity meter was calibrated with two KCl conductivity standard solu-
tions (0.0100 M, with κ = 1413 μS cm−1 at 25.0 °C; and 0.100 M, with
κ = 11.28 mS cm−1 at 25.0 °C) supplied by Crison. The temperature
was kept constant to within 0.1 °C by passing thermostated water
through a jacketed vessel holding the solution.
characteristics of these amphiphiles. Accordingly, these
calixarenes (along with other derivatives) were classified9 into
three categories: (i) nonmicellar calixarenes: those calixarenes
that do not bear long alkyl chains (e.g., CH3) do not forms
micelles; (ii) micelle-forming calixarenes: those with moderate
alkyl chain length (e.g., hexyl) self-aggregate into micelles; and
(iii) unimolecular micelle calixarenes: calixarenes that possess
long alkyl chains (e.g., dodecyl) behave as a unimolecular micelle.
(The authors suggested that these calixarenes are stable in
water as a monomer or oligomer and proposed that a “single
molecule having six dodecyl chains can form a stable, micelle-
like closed shell by itself”.) This classification was suggested on
the basis of measurements of the critical micelle concentrations
(cmc) by surface tension, conductance, and spectroscopic
methods. The micellization of calixarenes in the categories ii
and iii was confirmed by a small-angle X-ray scattering study.18
Ise et al. found that p-sulfonatocalix[6]arene hexahexyl ether
(SC6HH) forms ellipsoidal micelles with aggregation numbers
that varies between 7 and 15 depending on the concentration.
However, in the same work it was also estimated that p-sulfo-
natocalix[6]arene hexadodecyl ether self-aggregates into micelles
with aggregation numbers higher than 80 which is definitively
not compatible with the existence of unimolecular micelles. It
was suggested by the authors that the different results could
be attributed to the higher concentration used in their work
in comparison with that used in the work developed by Shinkai
et al.7−9
All solutions used in the NMR experiments were prepared in D2O
(99.9%) while the solutions used for conductivity measurements were
prepared in deionized and doubly distilled water.
For the molecular modeling calculations an initial model of the
SC6HH calixarene in the pseudo-1,2,3-alternate conformation having
the alkyl chains completely extended and outside the calixarene cavity
was built with ChemOffice 2004 modeling software (CambrigeSoft
Corp.). This initial model was exported for refinement with the model-
ing software Vega-ZZ v.2.420 using molecular mechanics and dynamics
calculations. The module NAMD21 incorporated in the Vega-ZZ pro-
gram was used for molecular dynamics simulations. Both molecular
mechanics and molecular dynamics calculations were performed with
the CHARMM22_LIG force field and Gasteiger charges using a di-
electric constant ε = 80.
To get more in depth into the structure−aggregation
relationships of surfactant-based calixarenes, in this work, we
characterized the self-assembly process of five p-sulfonatocalix-
arenes (Scheme 1) with different alkyl chain length and number
The initial conformation of SC6HH was extensively optimized with
molecular mechanics minimization until convergence through a maxi-
mum of 60K cycles of conjugated gradient optimization. The resulting
conformation was used as starting conformation for a 2 ns molecular
dynamics simulation at 300 K. During the trajectory the calixarene
molecule and in particular the alkyl chains were able to optimize its
geometry overcoming small energy barriers. The final conformation
obtained in the MD trajectory was finally submitted to extensive mole-
cular mechanics minimization with a maximum of 60K cycles of con-
jugated gradient optimization. The final conformation obtained using
this molecular modeling protocol is discussed in the text.
Scheme 1. Calixarene-Based Surfactants Used in This Study
RESULTS AND DISCUSSION
■
Conductivity. The cmc’s of the different calixarene-based
surfactants studied in this work were determined by electrical
conductivity (κ) measurements. For all surfactants κ increases
as the concentration increases (see Supporting Information for
an example); however, the rate of increase in κ, relative to the
concentration, is different below and above the cmc. Each plot
of κ against the surfactant concentration furnishes two straight
lines that intersect at the concentration corresponding to the
micelle formation, allowing the determination of the cmc. From
the slopes of the lines it is possible to calculate the micelle
ionization degree (α). Assuming that the micelles do not con-
tribute significantly to the conductivity, α can be calculated
from the ratio of the slopes of the two straight lines above and
below the cmc.22 The degree of ionization can be replaced by
the degree of counterion binding (β) using the relationship
β = 1 − α. This parameter reflects the ability of the micelles to
bind counterions present in solution. The cmc’s and β values
obtained by electrical conductivity measurements are listed in
Table 1. The cmc values obtained for SC4TB, SC4TH, and
of phenol units in the macrocyclic ring. The cmc and the degree
of counterion binding (β) were determined by electrical con-
1
ductivity experiments, and H NMR spectroscopy was used
to investigate possible conformational changes induced by the
aggregation process. Finally, the determination of the diffusion
coefficients of the calixarenes both below and above the cmc,
using diffusion ordered NMR spectroscopy (DOSY) experi-
ments, provided information about the type and geometry of
the aggregates present in solution.
EXPERIMENTAL SECTION
■
All commercial reagents were of the highest purity available and used
as received. p-Sulfonatocalix[n]arenes (SCn) were prepared by ipso-
sulfonation of the correspondent p-tert-butylcalix[n]arene in H2SO4 at
80 °C. Amphiphilic calixarenes were synthesized from the reaction of
SCn with the corresponding alkyl bromide in basic media following
7
1
literature procedures. The products were characterized by H NMR
and MALDI-TOF mass spectrometry (see Supporting Information).
2405
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
amphiphilic moieties linked at the level, or as close as possible,
of the headgroups by spacer groups) to self-aggregate increases
as the number of alkyl chains increases.26 In addition, it was
observed that for p-sulfonatocalixarenes bearing butyl chains at
the lower rim the cmc decreases as the number of phenol units
increases.9
Table 1. Critical Micelle Concentration (cmc) and Degree
of Counterion Binding (β) Obtained from Electrical
Conductivity Data for the Five Amphiphilic
a
p-Sulfonatocalixarenes Investigated in This Work
surfactant
n
m
cmc (mM)
β
SC4TB
SC4TH
SC4TO
SC6HH
SC8OH
4
4
4
6
8
4
6
8
6
6
3.18
0.59
0.61
0.59
0.70
0.78
As in the past series, it is useful to compare the obtained
values for the cmc in mole of alkyl chain units (ncmc) instead
of mole of surfactant units. When the values of log ncmc are
represented against the number of phenolic units in the macro-
cyclic ring (see Supprting Information), it is observed that
the tendency to self-aggregate into micelles decreases with the
increment of phenolic units in the calixarene structure. Since
SC4TH is blocked in the cone conformation17 and, as we will
show later, the cone conformation is preferred in the micellar
state, the higher tendency of SC4TH to aggregate can be inter-
preted in terms of preorganization effects (there is no unfavor-
able energetic cost related to conformational changes). On the
other hand, SC6HH and SC8OH are more flexible (due to the
unhindered rotation of the phenol unit through the annulus)27
and are allowed to adopt conformations other than cone in
bulk solution (monomeric state). Thus, it is probable that the
free energy of micellization of these two calixarenes include an
extra and unfavorable term associated with the conformational
reorganization into cone in the aggregates. Apparently, in
the case of the most flexible SC8OH the entropic cost to fix
the cone conformation is higher, and consequently, it presents
a lower tendency to self-assemble into micelles. While this sup-
posed preorganization effect is effective in the case of p-sulfo-
natocalixarenes bearing hexyl chains, according to ref 9, it is not
observed for their analogues bearing butyl chains. The aggre-
gation of surfactants is a complex process and results from the
balance of several weak interactions. In the latter case increasing
the number of phenol units (and thus alkyl chains) seems to be
more important to the micellization process than the preorga-
nization effect.
0.491
0.0911
0.734
0.729
a
All data were obtained at 25 °C. n is the number of phenolic units in
the macrocyclic ring, and m is the number of carbons per alkyl chain.
SC6HH are in good agreement with those obtained in previous
works.7,9,17
For surfactants derived from calix[4]arene (SC4TB, SC4TH,
and SC4TO) the cmc decreases with increasing the number of
carbons atoms in the alkyl chains. In this series, all surfactants
have the same type and number of heads groups, the same
type and number of counterions, and the same number of alkyl
chains. The only factor that varies is the number of carbons
in the alkyl chains (m). In analogy with gemini surfactants,23
which have two alkyl chain and two head groups per molecule,
the cmc values obtained here are correlated with the number of
carbon atoms per alkyl chain (m) instead of the total number of
carbon in the alkyl chains. The correlation between log 4cmc
(4cmc is the cmc in units of mM of alkyl chain) against m pre-
sents a linear behavior with a negative slope as generally ob-
served for conventional surfactants (see Supporting Information).
It is worth noting that if log cmc is used instead of log 4cmc,
the same slope is obtained but the intercept will be more
negative (−0.97 0.08). From this representation a line of the
type log 4cmc = A + Bm was obtained with B = −0.39 0.02
and A = −0.37 0.08. Since log cmc is proportional to the free
energy of transfer, ΔGt, of a surfactant molecule from the
aqueous phase to the micellized state,23−25 the negative B value
reflects the fact that the hydrophobic interactions contributes
favorably to the micellization process making ΔGt more nega-
tive. On the other hand, the A value is generally positive for
conventional surfactants25 (e.g., in the case of anionic surfac-
tants the A values usually found varies between 1.3 and 1.9),
indicating that the transfer of a hydrophilic group from the
aqueous bulk to the micellar phase contributes unfavorably to
the micellization process. In the present case A is negative since
in this correlation it includes the contribution from the ionic
groups, the aromatic rings, the phenolic oxygen atoms, and the
CH2 groups of the methylene bridges. In analogy with conven-
tional surfactants the ionic and polar groups should contribute
unfavorably to the aggregation process while the CH2 groups
and the aromatic rings should contribute in a favorable way
to the formation of micelles due to hydrophobic, CH-π, and
π-stacking interactions.
Table 1 also shows the obtained values for β. As can be
observed, for the calix[4]arene derivatives, β is independent of
the alkyl chain length. This suggests that the hydration of the
surfactant headgroups does not change with chain length. The
β values are associated with the capacity of the micelles to bind
counterions, and in general, higher values of β arise from a
closer packing of the headgroups and a higher surface charge
density at the micelle−solution interface.28 So the results indi-
cate that all amphiphilic calix[4]arene derivatives studied here
present a similar surface charge density and suggest that the
aggregates should present similar geometries. On the other hand,
the β values increase with increasing the number of phenol units
in the calixarene structure. This increase in the surface charge
density indicates that amphiphilic calixarenes with larger macro-
cyclic ring and higher flexibility form more compact aggregates
than those with more rigid frameworks.
1
1H NMR study of Calix[4]arene Derivatives. The H
In this work we also studied the aggregation of three p-sulfo-
natocalix[n]arenes bearing hexyl chains at the lower rim with
different number of phenol units in the macrocyclic ring
(SC4TH, SC6HH, and SC8OH). In Table 1 it can be observed
that the calix[4]arene derivative (SC4TH) presents the lower
value for the cmc (higher tendency to self-aggregate), while the
calix[6]arene (SC6HH) and calix[8]arene (SC8OH) deriva-
tives present similar values for the cmc. This behavior is con-
trary to what is expected since, in a general way, the tendency of
oligomeric surfactants (i.e., molecules with two or more identical
NMR spectra of all amphiphilic calix[4]arene derivatives show a
pair of doublets for the methylene protons (Hexo and Hendo) at
concentrations below the cmc, indicating that those calixarenes
adopt the cone conformation.27 In addition, it had been shown
that when the alkyl groups present in the lower rim of tetra-O-
alkylated calix[4]arenes has 3 (i.e., propyl) or more carbons the
rotation of the benzene unit through the annulus is inhibited.27
So the calix[4]arene derivatives are immobilized in the cone
1
conformation. As can be observed in the H NMR spectra of
SC4TB (see Supporting Information), when the concentration
2406
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
increases above the cmc, the chemical shifts of the protons
Hendo, OCH2 are displaced upfield while the chemical shift of
Hexo remains constant. This behavior was previously observed,9
and it was suggested that the butyl groups form the micellar
core while the calixarene rings form stacks at micellar surface.
From the graphical representation of the chemical shifts of the
mentioned protons against the concentration of calixarene, a cmc
value of 2.3 mM was determined (see Supporting Information).
This value is in good agreement with that obtained in an early
work (2.5 mM)9 and slightly lower than the value obtained
by conductivity (3.18 mM, Table 1). This discrepancy can be
attributed to the fact that the first was measured in D2O while
the second was obtained in H2O. It is known that the cmc of
ionic and neutral surfactants is generally lower in D2O since
hydrophobic interactions are stronger in this solvent.29
chemical shifts variation upon micellization is approximately the
same for SC4TB as for SC4TH).
Analysis of the 1H NMR spectra of SC4TO below and above
the cmc (see Supporting Information) seems to suggests that
the chemical shifts are unaffected by the aggregation process,
and new signals (due to slow exchange) do not appear in
the spectra at concentrations above the cmc. However, careful
1
observation of the H NMR spectra of SC4TO shows that
below the cmc (0.06 mM) the signals of the Hendo and OCH2
protons present considerable broadening and appear displaced
to higher field (4.45 and 3.9 ppm) when compared with the
signals of the same protons in the spectra of SC4TB and SC4TH
monomers (4.6 and 4.1 ppm). In addition, the SC4TO signals
present approximately the same chemical shifts as those observed
for the micelles formed from SC4TB and SC4TH. These results,
considered together, strongly suggest the presence of aggregates
at premicellar concentrations.
As we showed in an earlier paper,17 when the concentration
of SC4TH is increased to values above the cmc, new signals for
1
the Hendo, OCH2 protons appear in the H NMR spectra (at
NMR Study of Calix[6]arene Derivative below the
1
25 °C) due to the slow exchange between monomers in the
bulk and in the micelles. This phenomenon is clearly observed
for the proton signals subjected higher chemical shift variations
upon micellization. To inspect if the slow exchange is also ob-
served in the case of SC4TB, we decided to study the influence
of the temperature on the Hendo, OCH2, and Hexo 1H NMR
signals of this compound above the cmc (4 mm, see Supporting
Information). The results show that when the temperature
decreases from 25 to 5 °C, the peaks broadened, indicating that
exchange rate becomes slower at that temperature. At 1 °C the
signals started to resolve, but the separation between the signals
of the micelle and those of free monomers is not sufficient to
obtain the exchange rate constant with 2D EXSY experiments.
It is worth noting that the signal corresponding to the Hexo
protons remains practically unaffected by the exchange process.
Though it was not possible to extract the exchange rate
constants for SC4TB, this study seems to suggest that the
monomer−micelle exchange rate become slower with increas-
ing the length of the alkyl chains (since the magnitude of the
CMC. The H NMR spectra of SC6HH (Figure 1) show a
more complex pattern than that observed for calix[4]arene
derivatives. Larger calix[6]arenes display a higher flexibility and
are more difficult to block into a given conformation because
of the facile “through the annulus” rotation of their aromatic
units.27 Calix[6]arenes derivatives can adopt eight theoretical
conformations, which differ in the syn or anti orientations of the
aromatic units with respect to one another.27 To our knowledge,
only the 1,2,4-conformation has not been reported to date.30
In addition to the eight “up−down” conformations, numerous
others in which one or more of the aryl groups project outward
from the average plane of the molecule are possible. In the solid
state, p-sulfonatocalix[6]arene (SC6) adopts an 1,2,3-alternate
conformation31 stabilized by intramolecular hydrogen bonding
between the three hydroxyl groups in each half of the molecule.
Because of the high degree of hydration in the solid state, it is
reasonable to think that SC6 adopts a similar conformation in
solution. In the case of SC6HH, hydrogen bonding is no longer
1
possible, but the H NMR spectra indicate that this molecule is
1
Figure 1. H NMR spectra of SC6HH 0.4 mM in D2O at 25 °C. The inset show the (triangles) ArH, (squares) ArCH2Ar, and (circles) OCH2
regions.
2407
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
stabilized in a specific conformation, probably by hydrophobic
interaction between the alkyl chains.
While the protons of the methylene bridges (ArCH2Ar)
connecting adjacent aromatic units with syn orientations appear
1
as a pair of doublets in the H NMR spectra (where the
low field doublet corresponds to Hendo protons and the upfield
doublet to Hexo protons), when the adjacent aromatic units are
anti oriented, those protons appear as a singlet.30 The 1H NMR
spectra of SC6HH were analyzed before,7,9 but it was not speci-
fied if the spectra were acquired at concentrations below or
above the cmc. According to the authors, the signals corre-
sponding to the ArCH2Ar protons afforded a broadened singlet
resonance assigned to a mobile conformation. As can be ob-
served in Figure 1, our results are quite different: the ArCH2Ar
signals of SC6HH present a pattern that consists of two pairs of
doublets signals (2:2:2:2) and two singlets (2:2), suggesting
that the conformation adopted by SC6HH presents four meth-
ylene bridges connecting adjacent aromatic units with syn
orientations (the doublet at 3.65 ppm appears overlapped with
one pair of OCH2 protons) and two methylene bridges
connecting adjacent aromatic units with anti orientations.
Between the eight possible conformations of SC6HH only
the 1,2,3-alternate, the partial cone, and the 1,2-alternate are
compatible with this combination of syn and anti methylene
bridges. The Hendo (low field) and Hexo (high field) protons (of
the syn methylene bridges) were assigned by COSY (Figure 2).
This experiment reveals one pair of methylene bridges with
higher syn character (Δδobs = 0.9 ppm) and another one with
lower syn character (Δδobs = 0.5 ppm).
The splitting pattern observed for the ArH (2:2:2:2:2:2)
signals is characteristic of a Cs symmetrical conformation. The
cross peaks in ArH region of the COSY spectrum (Figure 2)
exclusively reflect correlations between protons in the same
aromatic unit. The results indicate that there are three different
types of aromatic rings in SC6HH, and therefore each signal
represents two equivalent protons in different aromatic units.
This observation combined with presence of two nonequivalent
anti methylene bridges strongly suggests that the molecule
should have a symmetry C2 axis.
Figure 2. Part of the 2D 1H, 1H COSY spectra of SC6HH 0.4 mM in
D2O at 25 °C corresponding to the ArCH2Ar signals (up) and to the
ArH signals (down). Signals corresponding to the three different types
of aromatic rings are labeled in the horizontal inset corresponding to
The signals corresponding to the OCH2 protons (2:2:2:2:2:2)
(see Figure 1) indicate that each OCH2 group experiment a dif-
ferent chemical environment and that the conformation adopted
by the calixarene is somewhat distorted. Besides, the appearance
of a high-field and relatively broadened signals at 2.9 ppm sug-
gests that at least one of the alkyl chains is self-included in the
SC6HH cavity. Conformational stabilization by self-inclusion
of alkyl groups have been observed for other calix[6]arene
derivatives.32
1
the 1D H spectrum.
Scheme 2. Structure of SC6HH in the Proposed Pseudo-
1,2,3-Alternate Conformation and Simplified Scheme
a
Representing This Conformer to Facilitate the View
1
In order to account for the H NMR splitting patterns, we
propose that SC6HH adopts a pseudo-1,2,3-alternate confor-
mation where two adjacent aromatic units, oriented in anti
configuration relative to each other, are pointing outside of
the average plane of the molecule in such a way that one or two
alkyl chains can be (partially) self-included in the calixarene
cavity (alternatively, the dynamic motions of these units could
bend fast on the NMR time scale giving rise to equivalent peaks
in the spectra). This structure is characterized by a C2 axis that
passes thru the anti methylene bridges (Scheme 2).
The proposed conformation for SC6HH was further sup-
ported by NOESY and ROESY experiments. The NOE cross
peaks were all negative (i.e., same sign as the diagonal), and the
ROESY spectra (see Supporting Information) served to rule
out the possibility that any of the relevant cross peaks observed
a
The circles represent aromatic units and are code-colored to identify
NMR equivalent aromatic units.
in the NOESY spectra were due to chemical exchange. Figure 3
shows part of the 2D NOESY spectra of SC6HH correspond-
ing to the regions where the ArH−ArH and ArH−ArCH2Ar
2408
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
Figure 3. Selected regions of the NOESY spectra of SC6HH (0.4 mM in D2O, mixing time = 500 ms) showing NOE cross signals between the
ArH−ArH protons (up) and between ArH and ArCH2Ar protons (down). The labeling of the ArCH2Ar protons in the horizontal inset uses letters S
for singlet and D for doublet.
correlations appears. In Scheme 3, it is shown a model of SC6HH
in the pseudo-1,2,3-alternate conformation (see Scheme 2) that
was considered for the assignment of the H NMR spectra. In
this scheme circles with the same color represents equivalent
aromatic units. The NOE cross-peaks labeled in the NOESY
spectra of Figure 3 were all nicely compatible with the pro-
posed pseudo-1,2,3-alternate conformation.
Molecular modeling calculations based in unrestrained molec-
ular mechanics and molecular dynamics provided an opti-
mized conformation for the pseudo-1,2,3-alternate conformation
of SC6HH (see Supporting Information) (the details of this cal-
culation are given in the Experimental Section). The interproton
distances in this model were found to be fully compatible with
the qualitative analysis of the NOE intensities for those cross
peaks labeled in Figure 3. The model has one alkyl chain self-
included in the cavity, a characteristic that had been discussed
above. The structure is defined by two half-cavities, one of them
being highly symmetric, defined by the 1, 2, and 3 aromatic rings,
and the other, defined by rings 4, 5, and 6, distorted. It is worth
noting that the alkyl chain that is included in SC6HH cavity is
the one corresponding to the aromatic ring 4 instead of the
ring 5. If the alkyl chain attached to aromatic ring in position 5
(or 2) was self-included into the cavity, the structure of the mole-
cule should have a symmetry plane instead of a rotation axis C2,
resulting in different 1H NMR spectra. To account for the 2-fold
1
1
symmetry observed in the H NMR spectra of SC6HH, molec-
ular flexibility should permit the fast interconversion on the
NMR time scale between the half-undistorted and half-distorted
parts of the molecule, leading to the average structure depicted
on Scheme 3. Such a type of dynamic equilibrium is consistent
with the possibility of the aromatic unit undergo rotation
through the annulus and with the observation that will be
described in the next section that increasing the concentration of
SC6HH toward values above the cmc produces substantial
changes in the 1H NMR spectra. On overall, the single structural
model of SC6HH was found to provide a very consistent
2409
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
time scale) between the several possible conformations adopted
by SC8OH.
Scheme 3. Simplified Scheme, Equivalent to That Depicted
in Scheme 2, Representing the Pseudo-1,2,3-Alternate
Conformer
a
As was observed for SC4TH and SC6HH, increasing the
concentration of SC8OH above the cmc leads to the appear-
1
ance of new signals in the H NMR spectra of this molecule,
again compatible with slow exchange (on the NMR time scale)
between monomers in the bulk and in the aggregates. At con-
centrations well above the cmc (4.0 mM) it can be observed
two signals (1:1) for the methylene protons separated by
0.7 ppm, indicating that in the micelles SC8OH adopts the
cone conformation. Scheme 4 summarizes all the conforma-
tions adopted by the amphiphilic calixarenes studied herein
both in bulk solution and in the micellized state.
DOSY. Diffusion ordered NMR spectroscopy (DOSY)
experiments were used to determine the self-diffusion coeffi-
cients (D) of the five calixarene-based surfactants studied here.
The experiments were performed for several concentrations of
surfactant below and above the cmc. The signal attenuation
obtained during the influence of a pulsed field gradient in the
stimulated echo based DOSY experiment is a function of the
molecular motion. For single diffusing species and when square-
shaped pulsed field gradients are applied, the attenuation of the
signal intensities (I) is given by
a
This scheme is exposed to help the reader to visualize the correlations
observed in the NOESY spectra. Small circles represent protons or groups
or protons. Ar stands for aromatic, S for singlet, and D for doublet.
explanation for most of the characteristics observed in the NMR
spectra.
⎡
⎤
⎛
⎞
⎟
I
δ
3
2
2 2
⎜
ln
= −D γ G δ Δ −
⎢
⎥
⎦
⎝
⎠
⎣
1H NMR Study of Calix[6]arene Derivative above the
CMC. The effect of increasing the concentration of SC6HH
toward values above the cmc (see Supporting Information)
I
(1)
0
where γ is the magnetogyric ratio of the nucleus under obser-
vation (in this case H), G is the gradient strength, δ is the
1
1
produces substantial changes in the H NMR spectra. At inter-
duration of the gradient, and Δ is the diffusion time. If several
spectra with different gradient strengths are acquired and all
other parameters are kept constant (δ and Δ), D can be directly
obtained by representing the signal intensities against (γGδ)2-
(Δ − δ/3) and fitting data to eq 1. Above the cmc the surfactant
molecules are in dynamic equilibrium between monomeric and
micellar states. When the exchange kinetics is fast on the diffu-
sion time scale, the observed diffusion coefficient (Dobs) is the
molar fraction weighted average of the diffusion coefficients of
the monomer (Dm) and the micelle (DM):
mediate concentrations it is possible to observe the appearance
of new peaks in the spectra assigned to the aggregates and
the presence of the signals of the monomer, suggesting that for
SC6HH exchange between monomers in the bulk and in the
micelles is slow on the NMR time scale. At sufficiently high
concentrations (10 mM) only the signals corresponding to the
aggregates are observable. Judging by the simplicity of the spec-
tra, it is evident that in the aggregates SC6HH adopts a more
symmetrical conformation. The two signals that appear in the
ArCH2Ar and OCH2 regions were found to have relative areas
of 1:3. Considering this pattern, the first signal can be assigned
to the Hendo protons while the second is assigned to the Hexo
and OCH2 protons, indicating that SC6HH adopts the cone
conformation in the aggregates. This conformation is expect-
able since in the micelles the alkyl chains should point to the
hydrophobic interior while the sulfonate groups remain in con-
tact with the solvent. When SC6HH changes from the pseudo-
1,2,3-alternate conformation, in bulk solution, to cone, in the
micelles, the aromatic units should undergo a “through the
annulus” rotation. This conformational conversion demon-
strates that SC6HH is not blocked in a pseudo-1,2,3-alternate
conformation though this structure is stabilized, probably, by
hydrophobic interactions between the alkyl chains. Conforma-
tional changes induced by self-assembly have been observed for
other calix[6]arenes derivatives such as in the case of hydrogen-
bonded molecular cages.33
[S] − cmc
cmc
0
D
=
D +
D
M
obs
m
[S]
[S]
(2)
0
0
[S]0 being the total surfactant concentration. When the exchange
is slow, each signal shows the attenuation expected for appro-
priate diffusion coefficient, and the effects of the exchange may
be neglected. In the intermediate case, for two-site exchange, the
signal attenuation no longer shows a exponential dependence on
G2 but rather a sum of two exponentials.34−37 In this work we
used the signals of the ArH protons for the determination of the
D
obs since they are not subjected to variations on their chemical
shifts due to the aggregation process and thus not broadened
out due to slow exchange. In the case of SC6HH all signals are
subjected to significant broadening above the cmc, and in this case
we have used the ArH signals below the cmc and the OCH2 +
Hexo signals above the cmc. In all cases we observed good
monoexponential decays, suggesting that eq 2 can be applied
(i.e., fast exchange in the diffusion time scale). For further con-
firmation, the Dobs were obtained using distinct diffusion times,
and the same value (within the experimental error) was obtained
(see Supporting Information).
1H NMR Study of Calix[8]arene Derivative. The
1
influence of the concentration on the H NMR of SC8OH (see
Supporting Information) shows that below the cmc (0.30 mM)
the spectra of this highly flexible calix[8]arene derivative present
a pattern compatible with a conformational mobile struc-
ture. A broadened signal at 4.15 ppm is assigned to the ArCH2Ar
protons and is compatible with rapid exchange (on the NMR
Figure 4 shows the obtained values for Dobs plotted against
the concentration of surfactant. The data for SC4TH (Figure
4b) was already published in a early paper and is showed again
2410
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
Scheme 4. Conformations Adopted by the Calixarene-Based Surfactants in Bulk Solution and in the Micelles
for comparison.17 In the case of SC6HH, due to slow exchange
for all H NMR signals, it was not possible to obtain accurate
is in line with the observations made from the H NMR
experiments.
1
1
Using eq 2 along with the cmc and the Dm values (the
average of the Dobs obtained below cmc), it is possible to obtain
the DM value. In this work we only calculated DM for concen-
trations well above the cmc to avoid any complications that
could arise from the monomer−micelle exchange process. In
the case of SC4TO we assumed that for concentrations much
higher than the cmc Dobs ≈ DM. Once the DM values are known,
it is possible to calculate the “effective” hydrodynamic radius
of the aggregates (rh) using the Stokes−Einstein equation for
spherical particles
values of Dobs for intermediate concentrations (ca. 1−5 mM).
The variation of Dobs with the concentration of surfactant pres-
ents the expected tendency and (except in the case of SC4TO)
allows the determination of the cmc, Dm, and DM. Below the
cmc the solution contains only surfactant molecules in their
monomeric state, and thus Dobs is constant and equal to Dm.
The cmc is indentified in the point where the Dobs abruptly
drops due to the presence of aggregates in solution. The cmc’s
determined by this technique are presented in Table 2. As can
be observed, the values obtained in D2O are, in all cases, lower
than that obtained in H2O using conductivity measurements,
and this difference can explained by the fact that hydrophobic
interactions are stronger in D2O,29 as it was already mentioned
in this paper.
In the case of SC4TO it was not possible to determine the
cmc due to the lack of sensitivity of the technique at dilute con-
centrations, required to reach constant values of Dobs. However,
as can be observed, we have been able to measure the Dobs for a
concentration (0.03 mM) well below the cmc determined
by conductivity (cmc = 0.09 mM), and the value obtained for
Dobs = (1.5 0.3) × 10−6 cm2 s−1 is considerably lower than
what is expected for monomers (≈ 2.5 × 10−6 cm2 s−1) based
on the values obtained for SC4TB and SC4TH. This result
can be attributed to the formation premicellar aggregates and
kbT
6πηrh
D =
(3)
where kb is the Boltzmann constant, T the temperature, and
η the viscosity of the solution. The calculated rh values are
presented in Table 2. As can be observed, these values are
typical of micellar aggregates. Simple calculations of the length
of the monomeric unit of the amphiphilic calixarenes, using
ChemOffice Chem3D, with the alkyl chain fully extended give
the values (b) presented in the Table 2. As can be observed, the
calculated rh for the micellar aggregates are much higher than
the length of the surfactants. This observation suggests that
the micelles have geometries other than spherical. If we assume
that the micellar assemblies present ellipsoidal geometries, the
2411
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
Figure 4. Variation of the observed self-diffusion coefficient (Dobs) against the concentration of the amphiphilic p-sulfonatocalixarenes studied in this
work. (a) SC4TB, (b) SC4TH, (c) SC4TO, (d) SC6HH, and (e) SC8OH.
following equation applies38
Table 2. Critical Micelle Concentration (cmc) in D2O at
25 °C Obtained from DOSY Experiments, Hydrodynamic
Radius (rh) Considering Spherical Micelles, and Length of
the Minor (b) and Major (a) Semiaxis Considering
Ellipsoidal Micelles with Prolate and Oblate Geometries
k T
6πηb
b
D =
f(ρ)
(4)
where b is the minor semiaxis length of the micelle (and
assumed to be equal to the length of the surfactant); ρ = a/b is
the axial ratio, with a being the length of the major semiaxis;
and f(ρ) is a function representing the shape factor.
For a prolate ellipsoid (rodlike), f(ρ) is defined by eq 5 and
for an oblate ellipsoid (disklike) by eq 6.
a (nm)
surfactant
cmc (mM)
rh (nm)
b (nm)
prolate
oblate
SC4TB
SC4TH
SC4TO
SC6HH
SC8OH
2.3
2.7 0.2
3.5 0.2
3.4 0.1
2.9 0.2
3.1 0.1
1.15
1.40
1.65
1.40
1.40
6.6
8.9
7.8
6.6
7.3
3.5
4.6
4.3
3.7
4.0
0.32
0.56
0.50
2
ln(ρ + ρ − 1 )
f(ρ) =
By applying eqs 4−6, we have calculated the length of the
major semiaxis (a) of the micelles assuming prolate and oblate
ellipsoids (Table 2). However, the results obtained here do not
permit to distinguish between prolate and oblate geometries. As
commented in the Introduction, it was proposed,18 on the basis
of small-angle X-ray study, that SC6HH self-aggregates into
2
ρ − 1
(5)
2
arctan( ρ − 1 )
f(ρ) =
2
ρ − 1
(6)
2412
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
ellipsoidal micelles. These results are in agreement with those
obtained here, supporting the evidence that at concentrations
well above the cmc the micelles formed from amphiphilic
p-sulfonatocalixarenes adopt ellipsoidal geometries. We plan
to study this system in more detail using DOSY experiments
and a hydrophobic probe molecule, such hexamethyldisilane,39
in order to obtain detailed information about the structure of
these aggregates.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: luis.garcia@usc.es.
ACKNOWLEDGMENTS
■
This work was supported by Ministerio de Ciencia y Tecnologia
(Project CTQ2008-04420/BQU) and Xunta de Galicia
(PGIDIT07-PXIB209041PR, PGIDIT10-PXIB209113PR, and
2007/085). N.B. acknowledges FCT for a Ph.D. Grant (SFRH/
BD/29218/2006).
CONCLUSIONS
■
Amphiphilic p-sulfonatocalix[n]arenes self-assembled into
globular micelles probably with ellipsoidal geometry as revealed
by DOSY experiments. The cmc of p-sulfonatocalix[4]arenes
derivatives blocked into the cone conformation decreases when
length of the alkyl chains is increased, due to stronger hydro-
phobic interactions, as generally observed for conventional sur-
factants. Analysis of the 1H NMR spectra of SC6HH, supported
by 2D NMR experiments indicates that, in the monomeric
state, SC6HH adopts an asymmetric pseudo-1,2,3-alternate
conformation with a self-included alkyl chain. Since for this
calixarene the aromatic units can undergo “through the annulus”
rotation, it is likely that this conformation is stabilized by hydro-
phobic interactions between the alkyl chains. On the other hand,
the higher flexible calix[8]arene derivative (SC8OH) is confor-
mationaly mobile and exchange rapidly, on the NMR time scale,
between several possible conformation. However, in the micellized
state both SC6HH and SC8OH adopt the cone conformation.
This aggregation induced conformational change is expected due
to the globular structure of the micelles. This geometry favors
the cone conformation where all alkyl chains are directed in-
ward and the hydrophilic sulfonate groups are in contact with
the bulk aqueous solution.
For p-sulfonatocalixarenes bearing hexyl chains at the lower
rim, the cmc (in moles of alkyl chain units) increases when
the number of aromatic units in the macrocycle increases. This
behavior is unexpected and can be explained taking into account
the conformations adopted by these amphiphilic molecules.
SC4TH shows a higher tendency to micellize since it is pre-
organized in the cone conformation, and thus there are no con-
formational changes (and thus energetic costs associated with
it) when the monomers are transferred from the bulk aqueous
phase to the micelle. SC6HH and SC8OH change their confor-
mation when they are transferred from the monomeric to the
micellized state, and this presents an extra energetic cost that
results in a relative destabilization of the micelles and cmc in-
crease. When the aggregation tendency of SC6HH is compared
with that of SC8OH, the last is disfavored, probably, due to a
decrease in degrees of freedom, since SC8OH changes from
his highly flexible monomeric state to a micellar environment
where the molecule is conformationally constrained into the
cone conformation.
REFERENCES
■
(1) Calixarenes; Asfari, Z., Bohmer, V., Harrowfield, J., Vicens, J.,
̈
Eds.; Kluwer Academic Publishers: Dordrecht, 2001.
(2) Mulder, A.; Auletta, T.; Sartori, A.; Del Ciotto, S.; Casnati, A.;
Ungaro, R.; Huskens, J.; Reinhoudt, D. N. J. Am. Chem. Soc. 2004, 126,
6627−6636.
(3) Arduini, A.; Demuru, D.; Pochini, A.; Secchi, A. Chem. Commun.
2005, 645−647.
(4) Arduini, A.; Ciesa, F.; Fragassi, M.; Pochini, A.; Secchi, A. Angew.
Chem., Int. Ed. 2005, 44, 278−281.
(5) Rudkevich, D. M. Bull. Chem. Soc. Jpn. 2002, 75, 393−413.
(6) Helttunena, K.; Shahgaldian, P. New J. Chem. 2010, 34, 2704−
2714.
(7) Shinkai, S.; Mori, S.; Koreishi, H.; Tsubaki, T.; Manabe, O. J. Am.
Chem. Soc. 1986, 108, 2409−2416.
(8) Shinkai, S.; Kawabata, H.; Arimura, T.; Matsuda, T.; Satoh, H.;
Manabe, O. J. Chem. Soc., Perkin Trans. 1 1989, 1073−1074.
(9) Shinkai, S.; Arimura, T.; Araki, K.; Kawabata, H.; Satoh, H.;
Tsubaki, T.; Manabe, O.; Sunamoto, J. J. Chem. Soc., Perkin Trans. 1
1989, 2039−2045.
(10) Arimori, S.; Nagasaki, T.; Shinkai, S. J. Chem. Soc., Perkin Trans.
2 1995, 679−683.
(11) Strobel, M.; Kita-Tokarczyk, K.; Taubert, A.; Vebert, C.; Heiney,
P. A.; Chami, M.; Meier, W. Adv. Funct. Mater. 2006, 16, 252−259.
(12) Kellermann, M.; Bauer, W.; Hirsch, A.; Schade, B.; Ludwig, K.;
Bottcher, C. Angew. Chem., Int. Ed. 2004, 43, 2959−2962.
(13) Lee, M.; Lee, S. J.; Jiang, L. H. J. Am. Chem. Soc. 2004, 126,
12724−12725.
(14) Sansone, F.; Dudic, M.; Donofrio, G.; Rivelli, C.; Baldini, L.;
Casnati, A.; Cellai, S.; Ungaro, R. J. Am. Chem. Soc. 2006, 128, 14528−
14536.
(15) Consoli, G. M. L.; Granata, G.; Lo Nigro, R.; Malandrino, G.;
Geraci, C. Langmuir 2008, 24, 6194−6200.
(16) Micali, N.; Villari, V.; Consoli, G. M. L.; Cunsolo, F.; Geraci, C.
Phys. Rev. E 2006, 73, 0519041−0519048.
(17) Basílio, N.; García-Río, L.; Martín-Pastor, M. J. Phys. Chem. B
2010, 114, 4816−4820.
(18) Matsuoka, H.; Tsurumi, M.; Ise, N. Phys. Rev. B 1988, 38,
6279−6286.
(19) Wu, D.; Chen, A.; Johnson, C. S. J. Magn. Reson. 1995, 115,
260−264.
(20) Pedretti, A.; Villa, L.; Vistoli, G. J. Comput.-Aided Mol. Des. 2004,
18, 167−173.
(21) Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid,
E.; Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K. J. Comput.
Chem. 2005, 26, 1781−1802.
ASSOCIATED CONTENT
■
(22) Zana, R. J. Colloid Interface Sci. 1980, 78, 330−337.
(23) Zana, R.; Benrraou, M.; Rueff, R. Langmuir 1991, 7, 1072−1075.
(24) Mukerjee, P. Adv. Colloid Interface Sci. 1967, 1, 241−275.
(25) Rosen, M. J. Surfactants and Interfacial Phenomena, 3rd ed.; John
Wiley & Sons: Hoboken, NJ, 2004.
S
* Supporting Information
Characterization data, typical electrical conductivity against
surfactant concentration plot, graphs of the cmc plotted against
the alkyl chain length and the number of phenolic units, NMR
spectra, plot of the chemical shifts against surfactant concen-
tration, energy-minimized molecular models of SC6HH, and a
typical echo attenuation plot. This material is available free of
charge via the Internet at http://pubs.acs.org.
(26) Laschewsky, A.; Wattebled, L.; Arotcarena, M.; Habib-Jiwan, J.;
Rakotoaly, R. V. Langmuir 2005, 21, 7170−7179 and referencestherein.
(27) Ikeda, A.; Shinkai, S. Chem. Rev. 1997, 97, 1713−1734.
(28) Quirion, F.; Magid, L. J. J. Phys. Chem. 1986, 90, 5435−5441.
(29) Chang, N. J.; Kaler, E. W. J. Phys. Chem. 1985, 89, 2996−3000.
2413
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414

Langmuir
Article
́
(30) Menand, M.; Leroy, A.; Marrot, J.; Luhmer, M.; Jabin, I. Angew.
Chem., Int. Ed. 2009, 48, 5509−5512 and references therein.
(31) Atwood, J. L.; Clark, D. L.; Juneja, R. K.; Orr, G. W.; Robinson,
K. D.; Vincent, R. L. J. Am. Chem. Soc. 1992, 114, 7558−7559.
(32) van Duynhoven, J. P. M.; Janssen, R. G.; Verboom, W.; Franken,
S. M.; Casnati, A.; Pochini, A.; Ungaro, R.; de Mendoza, J.; Nieto,
P. M.; Prados, P.; Reinhoudt, D. N. J. Am. Chem. Soc. 1994, 116,
5814−5822.
(33) Arduini, A.; Domiano, L.; Ogliosi, L.; Pochini, A.; Secchi, A.;
Ungaro., R. J. Org. Chem. 1997, 62, 7866−7868.
(34) Morris, K. F.; Johnson, C. S. J. Am. Chem. Soc. 1992, 114, 3139−
3141.
(35) Tanner, J. E. J. Chem. Phys. 1970, 25, 2523−2526.
(36) Cabrita, E. J.; Berger, S.; Brauer, P.; Karger, J. J. Magn. Reson.
̈
̈
2002, 157, 124−131.
(37) Johnson, C. S. Jr. J. Magn. Reson., Ser. A 1993, 102, 214−218.
(38) Hiemenz, P. C. Principles of Colloid and Surface Chemistry;
Marcel Dekker: New York, 1986.
(39) Istjerndahl, M.; Lundberg, D.; Zhang, H.; Menger, F. M. J. Phys.
Chem. B 2007, 111, 2008−2014.
2414
dx.doi.org/10.1021/la204004h | Langmuir 2012, 28, 2404−2414