Monomeric, dimeric and hexameric resorcin[4]arene assemblies with alcohols in apolar solvents

Article abstract of DOI:10.1039/b803890b

Resorcin[4]arenes in an apolar solvent containing alcohols exist in three forms of self-assembled aggregates which have been characterised by the technique of diffusion NMR spectroscopy. The Royal Society of Chemistry.

Full text of DOI:10.1039/b803890b

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Monomeric, dimeric and hexameric resorcin[4]arene assemblies with
alcohols in apolar solventswzy
Bjoern Schnatwinkel,a Ion Stoll,a Andreas Mix,b Mikhail V. Rekharsky,c
Victor V. Borovkov,c Yoshihisa Inouec and Jochen Mattay*a
Received (in Cambridge, UK) 6th March 2008, Accepted 12th May 2008
First published as an Advance Article on the web 27th June 2008
DOI: 10.1039/b803890b
Resorcin[4]arenes in an apolar solvent containing alcohols exist
in three forms of self-assembled aggregates which have been
characterised by the technique of diffusion NMR spectroscopy.
independently and almost simultaneously the first X-ray crys-
tal structures of cubic hexameric nano-capsules of resorci-
n[4]arene 1b and pyrogallo[4]arene 2a. Studies on these
hexameric species in solution have almost primarily been made
by the groups of Cohen,6 Rebek7 and Atwood8 using NMR
spectroscopy techniques including NOE and diffusion coeffi-
cient determination. Later, Atwood9 suggested the formation
of the smaller self-assembled resorcin[4]arene 1c dimer facili-
tated by hydrogen-bond bridges involving eight 2-propanol
molecules, which was characterised by X-ray diffraction ana-
lysis. More recently Ugono and Holman10 discovered an
achiral hexameric capsule related to the chiral form reported
by Atwood. In contrast to the water bridged chiral assembly,
this capsule is held together by a hydrogen-bonding network
with six 2-ethylhexanol molecules and additionally two water
molecules on the edges of the formed cube.
The aggregation phenomenon controlled by one compound
(e.g. alcohol) in a three-component system (e.g. solvent,
resorcin[4]arene and alcohol) is not very well studied. The
works of Holman, Aoyama and Atwood prompted us to
elucidate the controlled molecular assembling processes of
1 : 1 complexes, dimers and hexamers in apolar solvents by
addition of selected alcohols. In order to visualize such inter-
actions, at first we attempted to detect and characterise such
species by mass spectrometry methods such as MALDI-TOF,
ESI and CSI. None showed evidence supporting the formation
of aggregates that survived in the gas phase. Also, Williams et
al.11 reported that neither variable temperature 1H NMR
spectroscopy nor GPC measurements yielded coherent infor-
mation concerning the aggregation of similar systems in solu-
tion. However, diffusion NMR spectroscopy provided useful
data concerning the solution states of aggregated species.
Based on the diffusion coefficients, it is possible to conclude
the particle size in solution. In our NMR experiments,z we
used a solution of undecylresorcin[4]arene 1a8 in dry CDCl3,
to which alcohols 3–5 were added.
Since Hogberg developed a practical synthesis of the resor-
¨
cin[4]arenes in 1980,1 numerous articles have been published
on these macrocyclic hosts and their molecular recognition
and self-assembling behavior. The bowl shaped structure of
resorcinarenes was one of the origins of the host–guest study
made by Aoyama et al.2 on the complexation of non-ionic
polar compounds in 1988, which was followed by Kobayashi
et al.3 Regarding the self-assembly of resorcin[4]arene 1a,
Aoyama also suggested in the same work that higher aggre-
gates such as pentamers or hexamers could possibly be
formed. The first evidence of higher supramolecular structures
in the solid state was discovered by Atwood and MacGilliv-
ray4, and Mattay et al.5 in the late 1990s. Both reported
a Organic Chemistry I, Department of Chemistry, Bielefeld University,
Bielefeld, Germany. E-mail: oc1jm@uni-bielefeld.de; Fax: +49 521
106 6417; Tel: +49 521 106 2072
We first determined the diffusion coefficients of species
existing in the NMR sample (0.7 mL) of 1a (25 mM) and
racemic 3 (1a : 3 E 1 : 10) in dry CDCl3. Because of the high
acidity of the alcoholic proton that should support the neces-
sary hydrogen-bonding network of such an aggregate, we
expected the formation of a hexameric capsule. However, the
diffusion coefficient (Table 1) of 0.42 ꢀ 0.01 ꢁ 10ꢂ5 cm2 sꢂ1
indicated a monomeric species in accordance with the reports
of Cohen et al.6 Considering the studies of Kobayashi et al.3 we
propose a 1 : 1 complexation of 2,2,2-trifluoro-1-phenylethanol
b Inorganic Chemistry 3, Department of Chemistry, Bielefeld
University, Bielefeld, Germany
c Department of Applied Chemistry, Osaka University, 2-1 Yamada-
oka, Suita 565-0871, Japan
w This work was supported by the Deutsche Forschungsgemeinschaft
(DFG).
z Electronic supplementary information (ESI) available: 1H NMR
spectra of 1aꢃ3 1 : 1 complex, 1aꢃ4 dimer, and 1aꢃ5 hexamer. See
DOI: 10.1039/b803890b
y Dedicated to Professor Ernst-Ulrich Wurthwein (University of
¨
Munster) on the occasion of his 60th birthday.
¨
ꢄc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 3873–3875 | 3873

View Article Online
Table 1 Effect of the addition of alcohols 3–5 on the diffusion
coefficient D (ꢁ 10ꢂ5 in cm2
s
ꢂ1) of 1a at 6.12 ppm in dry CDCl3 at
296 K
System
D/10ꢂ5 cm2 sꢂ1
)
Aggregation state
1aꢃ3
0.42 ꢀ 0.01
0.32 ꢀ 0.01
0.25 ꢀ 0.01
2.56 ꢀ 0.02
1 : 1 Complex
Dimer
Hexamer
—
1aꢃ4
1aꢃ5
Free CHCl3
Fig. 3 Stick model of the hexameric nano-capsule of 1a6ꢃ(5)6ꢃ(H2O)2.
compared to our previous experiments since the formation of
the hexameric nano-capsule has been reported by Holman and
Ugono10 (Fig. 3) and formulated as 1a6ꢃ(5)6ꢃ(H2O)2. In fact,
we determined a diffusion coefficient of 0.25 ꢀ 0.01 ꢁ 10ꢂ5 cm2
sꢂ1 respective to the signal at 6.12 ppm. This result is con-
sistent with the observed diffusion coefficients of several
experiments in related systems reported by Cohen et al.6 In
contrast to the solid state, in our experiments under absolutely
dry conditions we assume the complete ‘‘substitution’’ of all
water molecules by racemic 2-ethylhexanol resulting in the
hexameric complex 1a6ꢃ58 in solution.
Fig. 1 Stick model of the 1aꢃ3 complex. The alkyl chains are omitted
for clarity. The hydrogen-bonding network is described by the green
dotted lines.
in the cavity of resorcin[4]arene (Fig. 1). The driving force for
this complexation is apparently the p–p interaction between the
aromatic residue of the alcohol and the cavity. Moreover the
association may be assisted by the hydrogen-bonding network
formed by the alcohol and two resorcinol units. In addition,
solvophobic forces may also play a minor role in strengthening
the complexation.
Moreover under the solid state conditions it is likely that
three additional molecules of 5 are essential to stabilize the
aggregate’s inner volume (1290 A3) by encapsulation.
Depicted in Fig. 4 are the signal decays as a function of the
gradient strength combined for each experiment. It is apparent
that the system 1aꢃ5 has the strongest effect of all, whereas the
differences in the signal decay between the systems 1aꢃ3 and 1aꢃ4
are not so strong. This may be due to the moderate increase in
Next we treated the NMR sample of the stock solution of 1a
with racemic 4 (1a : 4 E 1 : 10) and determined the diffusion
coefficient. In good agreement with the results of Atwood et al.
in the solid state, we found a diffusion coefficient of 0.32 ꢀ
0.01 ꢁ 10ꢂ5 cm2 sꢂ1 which is typical for a dimer.6 This
aggregate is formed by 24 delocalized hydrogen bonds be-
tween two molecules of 1a and eight molecules of 4 (Fig. 2).
The effective van der Waals volume of the resorcin[4]arene’s
cavity was calculated by Barbour12 to be 230 A3. In order to
maintain structural stability, we assume that the capsule
contains an undefined number of alcohol molecules.
the particle size from monomeric to dimeric species (Vinner =
230 A3) and then dramatic increase in the case of hexameric
species (Vinner = 1290 A3). Filling the inner volume of these
Finally we treated the NMR sample of 1a with racemic 5
(1a : 5 E 1 : 10). We expected a lower diffusion coefficient
Fig. 4 1H NMR signal decays as a function of the gradient strength
(G) (600 MHz, 296 K) of 1aꢃ5 (’), 1aꢃ4 (K) and 1aꢃ3 (m) at 6.12 ppm
in dry CDCl3.
Fig. 2 Tube model of the butan-2-ol bridged resorcin[4]arene dimer.
3874 | Chem. Commun., 2008, 3873–3875
ꢄc
This journal is The Royal Society of Chemistry 2008

View Article Online
aggregates is of main interest in all of these experiments. We
suggest that the formation of these complexes is strongly
dependent on the shape and structure of the alcohol employed
and the resulting van der Waals interactions in the inner cavity.
In summary, we have elucidated the origin of the puzzling
aggregation behavior of resorcin[4]arenes with various alco-
hols by revealing the dramatic influence of bridging molecules
on the multicomponent aggregation in solution under fixed
conditions (concentration, pH, etc.). Further studies with
enantiomerically pure alcohols are currently in progress and
the results will be reported in due course.
3 K. Kobayashi, Y. Asakawa, Y. Kikuchi, H. Toi and Y. Aoyama,
J. Am. Chem. Soc., 1993, 115, 2648.
4 L. R. MacGillivray and J. L. Atwood, Nature, 1997, 389, 469.
5 (a) T. Gerkensmeier, W. Iwanek, C. Agena, R. Frohlich, S. Kotila,
¨
Iwanek, R. Kotila, R. Frohlich and J. Mattay, 5th International
¨
C. Nather and J. Mattay, Eur. J. Org. Chem., 1999, 2257; (b) W.
¨
Summer School on Supramolecular Chemistry, Ustron (Poland),
1996, Abstract P-31; (c) T. Gerkensmeier, W. Iwanek, C. Agena,
R. Frohlich, C. Nather and J. Mattay, 4th International Conference
¨
on Calixarenes, Parma (Italy), 1997, Abstract P-88.
¨
6 (a) L. Avram and Y. Cohen, J. Am. Chem. Soc., 2002, 124, 15148;
(b) L. Avram and Y. Cohen, Org. Lett., 2002, 4, 4365; (c) L. Avram
and Y. Cohen, Org. Lett., 2003, 5, 1099; (d) L. Avram and Y.
Cohen, Org. Lett., 2003, 5, 3329; (e) L. Avram and Y. Cohen, J.
Am. Chem. Soc., 2003, 125, 16180; (f) L. Avram and Y. Cohen, J.
Am. Chem. Soc., 2004, 126, 11556; (g) L. Avram and Y. Cohen, J.
Am. Chem. Soc., 2005, 127, 5714; (h) T. Evan-Salem and Y. Cohen,
Chem.–Eur. J., 2007, 13, 7659.
Notes and references
z NMR diffusion measurements were performed on a Bruker Avance
600 MHz FT-NMR spectrometer. The diffusion experiments were
performed using an LED pulse sequence with bipolar gradients as
delivered by the manufacturer. The gradient length (d) was set to 4000
ms and was varied linearly from 5 to 95%. The diffusion delay (D) was
set to 85 ms. 16 data points per diffusion experiment were collected
and 32 scans were applied per increment. Prior to the measurements
the heater and the airflow were switched off. and the sample was left in
the magnet for an additional 4 h in order to have constant temperature
conditions (296 K). In all measurements, we used the signal of 1a at
6.12 ppm to determine the diffusion coefficients. All spectral data
showed only averaged resonances of the alcohol signals on the NMR
timescale, because the formation of the different species is subjected to
a fast thermodynamic equilibria.
7 (a) A. Shivanyuk and J. Rebek, Jr, Proc. Natl. Acad. Sci. U. S. A.,
2001, 98, 7662; (b) A. Shivanyuk and J. Rebek, Jr, Chem. Com-
mun., 2001, 2424; (c) A. Shivanyuk and J. Rebek, Jr, J. Am. Chem.
Soc., 2003, 125, 3432; (d) M. Yamanaka, A. Shivanyuk and J.
Rebek, Jr, J. Am. Chem. Soc., 2004, 126, 2939; (e) T. Evan-Salem,
I. Baruch, L. Avram, Y. Cohen, L. C. Palmer and J. Rebek, Jr,
Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 12296; (f) E. S. Barrett,
T. J. Dale and J. Rebek, Jr, J. Am. Chem. Soc., 2007, 129, 3818.
8 (a) J. L. Atwood, L. J. Barbour and A. Jerga, Chem. Commun.,
2001, 2376; (b) J. Antesberger, G. W. V. Cave, M. C. Ferrarelli, M.
W. Heaven, C. L. Raston and J. L. Atwood, Chem. Commun.,
2005, 892.
9 K. N. Rose, L. J. Barbour, G. W. Orr and J. L. Atwood, Chem.
Commun., 1998, 407.
10 O. Ugono and K. T. Holman, Chem. Commun., 2006, 2144.
11 C. J. M. Stirling, L. J. Fundin and N. H. Williams, Chem.
Commun., 2007, 1748.
12 L. J. Barbour, CAVITY—computer program. A program to
compute the volume available to a sphere of given radius within
a molecular cavity. The program requires the van der Waals radii
and Cartesian coordinates of all the atoms surrounding the cavity.
A cube of length j is systematically stepped through the Euclidean
space containing all the atoms. If the cube can reside within any
sphere of radius r, which can in turn reside completely within the
cavity, the volume (initially set to zero) is incremented by i3. Values
of j = 0.2 A and r = 1.52 A were used. Van der Waals radii were
obtained from A. Bondi, J. Phys. Chem., 1964, 68, 441.
8 Resorcinol and n-dodecanal were obtained from Aldrich and used as
supplied.
Synthesis of undecylresorcin[4]arene, 1a: concentrated hydrochloric
acid (25 ml) was added to an ethanolic solution (125 ml) containing
n-dodecanal (33.2 g, 180 mmol) and resorcinol (19.8 g, 180 mmol). The
mixture was heated to reflux for 3 h. Upon cooling, the product
precipitated as a fine colourless solid. The crude product was collected
and washed with methanol and recrystallized from ethanol. Yield
75%.
¨
1 A. G. S. Hogberg, J. Org. Chem., 1980, 45, 4498.
2 Y. Aoyama, Y. Tanaka, H. Toi and H. Ogoshi, J. Am. Chem. Soc.,
1988, 110, 634.
ꢄc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 3873–3875 | 3875