Enantioselective helical folding inside a self-assembled, cylindrical capsule

Article abstract of DOI:10.1039/b925334c

The combination of experimental with theoretical CD spectroscopy allows the observation of stereoselective helical folding in a self-assembled capsule.

Full text of DOI:10.1039/b925334c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Enantioselective helical folding inside a self-assembled,
cylindrical capsulew
a
a
b
b
¨
Carsten Siering, Jakob Torang, Holger Kruse, Stefan Grimme* and
Siegfried R. Waldvogel*a
Received (in Cambridge, UK) 1st December 2009, Accepted 28th January 2010
First published as an Advance Article on the web 5th February 2010
DOI: 10.1039/b925334c
The combination of experimental with theoretical CD spectro-
scopy allows the observation of stereoselective helical folding in
a self-assembled capsule.
As only the asymmetric C14 chain revealed complete
folding, we focused on the encapsulation of 2-tetradecanol
(2). This compound is obtained enantiomerically enriched
by kinetic resolution of the corresponding racemic
acetate.10 The optical purity as determined by using Mosher’s
ester was 90% and 66% for the (R)- and (S)-enantiomer,
respectively.
Self-assembled capsules are well known in the field of supra-
molecular chemistry.1 The most prominent architecture was
introduced by Rebek et al. and consists of two concave,
self-complementary half-capsules (1; Fig. 1) which combine to
form a closed dimeric assembly.2,3 These capsules enclose a
volume of approx. 325 A3 and are able to accommodate various
guest molecules.4 Such containers have been successfully
employed as molecular flasks5 and for the binding of gases.6
A very fascinating part of Rebek’s work deals with the
encapsulation of linear alkanes inside a self-assembled capsule.7
It was found that linear hydrocarbons with chain lengths of
8 to 14 carbon atoms are bound as single molecules.8 For the
homologues containing more than 11 carbon centers, helical
folding of the chain inside the host was discovered by NMR
spectroscopy. The uncoiled alkyl chain would be too long to
be situated inside the cavity, therefore a helical folding is a
prerequisite for the formation of a closed capsule complex.
The highest affinity was found for n-undecane which exhibits
only slight coiling at the far ends of the molecule, whereas the
central region still remains extended. In the case of tetra-
decane, NOE experiments gave strong indications of a tightly
coiled conformation. There have been investigations on the
effect of chiral modifications at the periphery of the capsule on
the binding of chiral guests,9 but there is no report on a
stereocontrolled coiling of alkanes inside the cavity.
Initially, encapsulation of (rac)-2-tetradecanol was investi-
gated by 1H NMR spectroscopy. As typically found, the
protons of the encapsulated guest give rise to signals in a
range between 0.2 and À3.7 ppm. In contrast to Rebek’s case,
the lack of symmetry yields a large number of signals. Assign-
ment by COSY-NMR revealed a diastereotopic splitting of the
protons at C3, C4 and C5. The extent of this splitting
apparently depends strongly on the distance from the stereo-
genic center. While the signals for 3-H exhibit a difference in
chemical shift of 0.8 ppm, the proton resonance signals of C5
are only 0.1 ppm apart. It is known that the magnetic shielding
increases towards the ends of the capsule due to the
resorcinarene units.8 This anisotropy will be even stronger
when a nucleus is moved closer to the walls. Therefore, the
relative positions of the individual protons with respect to the
wall will strongly influence the shielding. Surprisingly, splitting
at the other terminus of the alcohol is not found. Apparently,
the chiral information induced at one end of the chain is not
promoted to the other in a way that allows detection on the
NMR time scale.
This interpretation is supported by a molecular modelling
using the AMBER force field. A Monte Carlo conformational
analysis revealed that a large majority of conformers exhibit a
high structural similarity to each other close to the alcohol
whereas the other end is rather disordered (see ESI for
detailsw). Fig. 2 depicts an average structure of the largest
Furthermore, there is evidence that interconversion of
P and M helices occurs rapidly on the NMR time scale since
diastereotopic protons were not observed. Rebek proposed a
‘‘crankshaft motion’’ for this particular process.8
In this report, we discuss the enantioselective helical folding
of an enantiomerically enriched alcohol inside an achiral,
self-assembled capsule.
a
´
Rheinische Friedrich-Wilhelms-Universitat Bonn, Kekule-Institut fur
Organische und Biochemie, Gerhard-Domagk-Str. 1, 53121 Bonn,
¨
¨
Germany. E-mail: waldvogel@uni-bonn.de; Fax: +49 228 739608;
Tel: +49 228 732653
b Westfalische Wilhelms-Universtitat Munster, Organisch-Chemisches
¨
¨
¨
Institut, Corrensstr. 40, 48149 Munster, Germany.
¨
E-mail: grimmes@uni-muenster.de; Fax: +49 251 8336515;
Tel: +49 251 833241
w Electronic supplementary information (ESI) available: Synthesis
and characterization of enantiomerically enriched 2-tetradecanol;
experimental details of CD experiments, further discussion of
theoretical results. See DOI: 10.1039/b925334c
Fig. 1 Structure of capsule monomer 1.
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 1625–1627 | 1625

View Article Online
involving simulation of a significant number of conformers
and averaging of individual CD spectra has already been
successfully applied to other all organic host–guest systems,12
molecular dynamics simulations with the semi empirical
SCC-DFTB-D13 method were performed for each possible
stereoisomer. Since the coiling of the alcohol induces
an additional chiral element (M/P helix) four possible
combinations of the chiral centers are conceivable (R-M,
S-M, R-P and S-P). The CD spectra resulting from trajectory
snapshots (about 70 for each diastereoisomer) were averaged
arithmetically (for details see ESIw).
Fig. 2 Complex of (R)-2-tetradecanol in 1Á1. Molecular Mechanics
model, irrelevant protons and substituents are omitted for clarity.
The experimental CD band at about 344 nm is reasonably
well reproduced by the applied electronic structure method
(TD-PPP) which yields CD bands at about 336–337 nm
(error of about 0.1 eV). The spectra shown in Fig. 4 typically
result from 20 individual electronic transitions in the
resorcinarene units that have been broadened by a standard
Gaussian shape function. The CD spectra clearly show that
the sign of the CD originates from the chirality of the coiled
alkane chain of 2-tetradecanol. The M helix results in a
positive CD while the P helix yields a negative CD sign. The
alcohol group itself seems to have no direct influence on the
observed CD effect. Since the TD-PPP method treats only
pi-electrons there is no electronic contribution from the guest
to the CD. The major effect seems to be the deformation of the
capsule induced by its helical guest. The deformation energy of
the capsule compared to its empty state is computed to be
9.0 kcal molÀ1 at the dispersion-corrected density functional
theory level (B97-D/TZVP14,15) for one example structure
of R-M configuration. The folding of the guest ‘‘costs’’ only
cluster of structurally similar conformers. In almost all
structures, the hydroxy group does not point towards the
tip of the cavity but to C2-H (this is in agreement with
the DFT-D structures discussed below). The axial
(Fig. 2, right) view gives an impression of the curvature of
the lower part.
Electronic CD spectroscopic investigations were performed
in order to demonstrate the stereoselective coiling of the chiral
alcohol. Mixing of (R)-2-tetradecanol with the achiral capsule
gives rise to
a positive CD at 344 nm, whereas the
(S)-enantiomer yields a negative signal. (Fig. 3) The varying
intensity of the CD effect is assigned to the differing enantio-
purity of the guest. As the alcohol exhibits no chromophore in
this specific wavelength range, the occurrence of a CD may be
attributed to a specific (‘chiral’) deformation of the capsule
induced by the alcohol. If the racemic alcohol is applied, no
CD can be observed.
Since the enantiomerically enriched alcohols were
applied in high excess relative to the concentration of the
capsule, it is necessary to prove that the CD does not arise
due to ‘‘solvent’’ effects, i.e. non-specific binding. Complete
loss of CD activity upon addition of n-undecane (which is
known to have a higher affinity to the capsule) proves that
this is not the case (see ESIw). Therefore, binding of the
enantioenriched alcohol inside the capsule is the origin of
the CD effect.
3.1 kcal molÀ1
.
The relative energies of the various diastereomers have been
computed at the same DFT level of theory. Using the
enantiomers R-M and S-P as reference the combinations
R-P and S-M (the corresponding diastereoisomers) are
2.8 kcal molÀ1 higher in energy, indicating that the combina-
tion R-M (S-P) is energetically favored over the combination
R-S (S-M). A partial cancellation of the relatively intense CD
bands for R-M and S-P would also explain the relatively small
signals observed experimentally.
In order to gain further insight into the host–guest system,
theoretical CD spectra were calculated using the semi
empirical TD-PPP11 method. As expected the system proves
to be highly flexible and small conformational changes have a
large effect on the computed CD spectra. Since an approach
Fig. 4 Computed TD-PPP CD spectra for 2-tetradecanol@1Á1. The
varying combinations of the alcohol head group (R/S configuration)
and the helical folding of the alkane chain (M/P configuration) lead to
4 isomers: (a) R-M, (b) R-P, (c) S-M and (d) S-P. The spectra result
from 20 individual electronic transitions.
Fig. 3 CD spectra of 1Á1 with enantiomerically enriched and racemic
2-tetradecanol.
ꢀc
This journal is The Royal Society of Chemistry 2010
1626 | Chem. Commun., 2010, 46, 1625–1627

View Article Online
The combination of experimental CD studies with theore-
tical investigations allows for the first time direct insight into
the helical arrangement of an alkyl chain in a self-assembled
Rebek-type capsule. By use of a chiral and optically enriched
secondary alcohol it was possible to induce a specific helical
folding of the guest. NMR investigations indicate a rather
fixed conformation close to the chiral center whereas the
remote part is as flexible as in normal alkanes. CD studies
further underline the binding and the subsequent formation of
a chiral inclusion complex. Theoretical investigation reveals
chiral induction in the resorcinarene chromophores and a
direct correlation between the helical chirality of the alkane
chain and the sign of the CD effect, independently of the
alcohol head group configuration.
Notes and references
1 J. W. A. Steed and J. L. Atwood, Supramolecular Chemistry,
Wiley, Chichester, 2000.
2 T. Heinz, D. M. Rudkevich and J. Rebek, Angew. Chem., Int. Ed.,
1999, 38, 1136–1139.
3 T. Heinz, D. Rudkevich and J. Rebek, Nature, 1998, 394, 764–766.
4 J. Rebek, Chem. Commun., 2007, 2777–2789.
5 J. Chen and J. Rebek, Org. Lett., 2002, 4, 327–329.
6 D. Ajami and J. Rebek, Angew. Chem., Int. Ed., 2008, 47, 6059–6061.
7 L. Trembleau and J. Rebek, Science, 2003, 301, 1219–1220.
8 A. Scarso, L. Trembleau and J. Rebek, J. Am. Chem. Soc., 2004,
126, 13512–13518.
9 T. Amaya and J. Rebek, J. Am. Chem. Soc., 2004, 126, 6216–6217.
10 T. Ohtani, H. Nakatsukasa, M. Kamezawa, H. Tachibana and
Y. Naoshima, J. Mol. Catal. B: Enzym., 1998, 4, 53–60.
11 F. Vogtle, A. Hunten and S. Grimme, Angew. Chem., Int. Ed.,
¨
2001, 40, 2468–2469.
¨
The support by providing samples of capsule monomer 1 by
J. Rebek, Jr. is gratefully acknowledged. This work was
supported by the Coordinated Research Collaboration SFB
624, DFG and by the IRTG 1444, DFG with a scholarship
to H.K.
12 C. Siering, S. Grimme and S. Waldvogel, Chem.–Eur. J., 2005, 11,
1877–1888.
13 M. Elstner, P. Hobza, T. Frauenheim, S. Suhai and E. Kaxiras,
J. Chem. Phys., 2001, 114, 5149–5155.
14 S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
15 A. Schafer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994, 100, 5829.
¨
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 1625–1627 | 1627