Endo-Cavity complexation and through-the-annulus threading of large calixarenes induced by very loose alkylammonium ion pairs

Article abstract of DOI:10.1021/ol100578z

Figure presented A general mode to obtain endo-complexation and through-the-annulus threading of scarcely efficient macrocyclic hosts was obtained by exploiting the inducing effect of the weakly coordinating tetrakis[3,5-bis(trifluoromethyl)phenyl]borate

Full text of DOI:10.1021/ol100578z

ORGANIC
LETTERS
2010
Vol. 12, No. 9
2092-2095
endo-Cavity Complexation and
Through-the-Annulus Threading of Large
Calixarenes Induced by Very Loose
Alkylammonium Ion Pairs
Carmine Gaeta,* Francesco Troisi, and Placido Neri*
Dipartimento di Chimica, UniVersita` di Salerno, Via Ponte don Melillo,
I-84084 Fisciano, Salerno, Italy
cgaeta@unisa.it; neri@unisa.it
Received March 10, 2010
ABSTRACT
A general mode to obtain endo-complexation and through-the-annulus threading of scarcely efficient macrocyclic hosts was obtained by
exploiting the inducing effect of the weakly coordinating tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (TFPB) anion that gives free “naked”
cations. This method works very well with simple, conformationally mobile calix[6-7]arene ethers, which give conformationally blocked endo-
calix or [2]pseudorotaxane complexes.
Over the past two decades calix[n]arenes have gained a
prominent position in the field of supramolecular chemistry
thanks to their wide range of different applications.1 In
contrast with other macrocyclic hosts such as cyclodextrins,2
crown ethers,3 and cucurbiturils,4 which readily give endo-
complexation with native unsubstituted derivatives,5 calix-
arenes often require an extensive chemical modification to
give similar results,6 and more frequently they act as simple
scaffolds to construct podand-like receptors where the calix
cavity very often remains unexploited.1
From another point of view, simple ethers of calixarenes
can be seen as modified crown ethers,3 and in principle they
should be able to give endo-complexation of organic cations
in a similar way. Unfortunately, with the notable exception
of some calix[5]arene derivatives,7 simple ethers of calix-
arenes do not give such kind of complexation because they
are too small in the case of calix[4]arenes or because they are
not well preorganized in the case of larger calix[6-8]arenes.
Thus, for example, it is well-known that hexamethoxy-p-
tert-butylcalix[6]arene 1a (Figure 1) is conformationally
mobile and does not complex halide, hexafluorophosphate,
or tetraphenylborate salts of organic ammonium cations.8
(1) Gutsche, C. D. Calixarenes, An Introduction; Royal Society of
Chemistry: Cambridge, U.K., 2008.
(2) Szejtli, J. Chem. ReV. 1998, 98, 1743.
On the other hand, the through-the-annulus threading of
dialkylammonium or bipyridium cations to give pseudoro-
taxane structures, which is very popular with large crown
(3) Gokel, G. W. Crown Ethers and Cryptands; Royal Society of
Chemistry: Cambridge, U.K., 1991.
(4) Lagona, J.; Mukhopadhyay, P.; Chakrabarti, S.; Isaacs, L. Angew.
Chem., Int. Ed. 2005, 44, 4844.
(5) Ogoshi, T.; Kanai, S.; Fujinami, S.; Yamagishi, T.; Nakamoto, Y.
J. Am. Chem. Soc. 2008, 130, 5022.
(7) Garozzo, D.; Gattuso, G.; Notti, A.; Pappalardo, A.; Pappalardo, S.;
Parisi, M. F.; Perez, M.; Pisagatti, I. Angew. Chem., Int. Ed. 2005, 44, 4892,
and references therein.
(6) Coquie`re, D.; de la Lande, A.; Mart`ı, S.; Parisel, O.; Prange´, T.;
Reinaud, O. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 10449, and references
therein.
(8) See Supporting Information for further details.
10.1021/ol100578z  2010 American Chemical Society
Published on Web 04/12/2010

ethers such as [24]crown-8 or [34]crown-10,9 respectively,
to obtain interlocked rotaxane and catenane topologies,10 is
very rare in the calix[n]arene series. In fact, the very few
examples regard the threading of viologen derivatives
through the annulus of calix[6]arene hosts in which the anion
coordinating ability of their ureido groups is exploited to
favor the ion-pair dissociation.11 Instead, to the best of our
knowledge, no examples of threading through conforma-
tionally mobile, larger calix[6-7]arene annuli by dialky-
lammonium cations are currently known.12
Therefore, to minimize the latter effect we decided to use
a weakly coordinating hydrophobic anion able to give a very
loose ion pair. An useful anion in this respect is tetrakis[3,5-
bis(trifluoromethyl)phenyl]borate (TFPB-) (Figure 1), which
is known to be noncoordinating (or very weakly coordinat-
ing) to metal ions and therefore is referred to as a
“superweak” anion.16
The addition of TFPB salt of n-butylammonium cation
+
3a (n-BuNH3 ·TFPB-) to a CDCl3 solution of hexahexyloxy
derivative 1b gave dramatic changes in its 1H NMR spectrum
(Figure 2a,b). The most evident ones were the sharpening
of all signals, the appearing of n-butyl resonances in the
upfield negative region of the spectrum (-0.87 and -1.07
ppm), and the formation of a well-defined AX system (4.48
and 3.48 ppm, J ) 14.4 Hz) for ArCH2Ar groups (Figure
2b). This is a clear indication that the conformationally
mobile calix[6]arene host 1b became blocked in the NMR
time scale in a cone conformation and that the n-butylam-
monium cation 3a gave an endo-calix complexation with the
alkyl chain shielded by the aromatic rings.
Here, we wish to report on how both problems of endo-
cavity complexation and through-the-annulus threading of
large calix[6-7]arenes can be solved in a general mode by
exploiting the inducing effect of a weakly coordinating anion.
Figure 1
As part of our ongoing program concerning the su-
pramolecular chemistry of large calixarene macrocycles,13
we decided to investigate the complexing ability toward
organic ammonium cations of some simple ether deriva-
tives. Not surprisingly, we found that in addition to
hexamethoxy-1a, both hexahexyloxy-1b and heptamethoxy-
p-tert-butylcalix[7]arene 2a (Figure 1) also gave no ap-
preciable interactions upon titration with chloride, hexaflu-
orophosphate, or tetraphenylborate salts of organic ammonium
cations.8 This can be ascribed to the sum of two unfavorable
effects, namely, the scarce preorganization of conformation-
ally mobile derivatives 1a,b and 2a (Figure 1) and the free
energy necessary to separate a tight ion pair.11,14,15
Figure 2.
1H NMR spectra (CDCl3, 400 MHz, 298 K) of (a) 1b, (b)
equimolar solution (3 mM) of 1b and n-butylammonium TFPB-, (c)
equimolar solution (3 mM) of 1b and di-n-pentylammonium TFPB-.
Complex formation was also confirmed by the ESI(+)
mass spectrum, which gave as the base peak a value of 1553
8
+
(9) Raymo, F. M.; Stoddart, J. F. Chem. ReV. 1999, 99, 1643.
(10) Molecular Catenanes, Rotaxanes and Knots: A Journey Through
the World of Molecular Topology; Sauvage, J. P., Dietrich-Buchecker, C.,
Eds.; Wiley-VCH: Weinheim, 1999.
m/z corresponding to supermolecular ion n-BuNH3 ⊂1b.
The presence of n-butylammonium cation inside the calix
cavity of 1b was proved by a 2D NOESY spectrum.8
This binding mode was confirmed by the lowest Amber-
(11) Arduini, A.; Ferdani, R.; Pochini, A.; Secchi, A.; Ugozzoli, F.
Angew. Chem., Int. Ed. 2000, 39, 3453.
+
energy structure of the complex n-BuNH3 ⊂1b (Figure 3a),
(12) Very recently, an example of threading through the conformationally
blocked pentakis(tert-butoxycarbonylmethoxy)-p-tert-butylcalix[5]arene deriva-
tive has been reported: Gattuso, G.; Notti, A.; Parisi, M. F.; Pisagatti, I.;
Amato, M. E.; Pappalardo, A.; Pappalardo, S. Chem.sEur. J. 2010, 16,
2381.
obtained with the program MacroModel-9.0,17 where it is
possible to observe the position of the nitrogen atom sitting
(13) (a) Neri, P.; Consoli, G. M. L.; Cunsolo, F.; Geraci, C.; Piattelli,
M. In Calixarene 2001; Asfari, Z., Bo¨hmer, V., Harrowfield, J., Vicens, J.,
Eds.; Kluwer: Dordrecht, 2001; Chapter 5, p 89. (b) For a review on the
chemistry of calix[7]arenes, see: Martino, M.; Neri, P. Mini-ReV. Org. Chem.
2004, 1, 219.
(16) (a) Strauss, S. H. Chem. ReV. 1993, 93, 927. (b) Hirochika, N.;
Suzuki, H.; Norob, J.; Kimura, T. Chem. Commun. 2005, 2963. (c) Nishida,
H.; Takada, N.; Yoshimura, M.; Sonoda, T.; Kobayashi, H. Bull. Chem.
Soc. Jpn. 1984, 57, 2600. (d) Blight, B. A.; Camara-Campos, A.; Djurdjevic,
S.; Kaller, M.; Leigh, D. A.; McMillan, F. M.; McNab, H.; Slawin, A. M.
J. Am. Chem. Soc. 2009, 131, 14116.
(14) Credi, A.; Dumas, S.; Silvi, S.; Venturi, M.; Arduini, A.; Pochini,
A.; Secchi, A. J. Org. Chem. 2004, 69, 5881
.
(17) Mohamadi, F.; Richards, N. G.; Guida, W. C.; Liskamp, R.; Lipton,
M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. MacroModel-
9.0/Maestro-4.1 program. J. Comput. Chem. 1990, 11, 440.
(15) Cafeo, G.; Gattuso, G.; Kohnke, F. H.; Notti, A.; Occhipinti, S.;
Pappalardo, S.; Parisi, M. F. Angew. Chem., Int. Ed. 2002, 41, 2122
Org. Lett., Vol. 12, No. 9, 2010
.
2093

(at a distance of 0.76 Å) above the mean plane of the
calixarene oxygens.
+
Figure 3
.
Lowest AMBER-energy structures of (a) n-BuNH3 ⊂1b,
+
+
(b) (n-Pent)2NH2 ⊂1b, and (c) Bn2NH2 ⊂1b complexes obtained
by Monte Carlo conformational searches. Hydrogen atoms of
calix[6]arene hosts 1b are omitted for clarity.
Three hydrogen bonds are present between +NH3 protons
and ethereal calixarene O-atoms. The lowest energy structure
is asymmetrical, and therefore time-averaging is observed
in the NMR time scale to give the observed C6V symmetry.
It is noteworthy that no hint of complexation of
3a·TFPB- was observed for hexamethoxycalix[6]arene 1a
and heptamethoxycalix[7]arene 2a. This implies that upon
complexation n-butylammonium cation induces the fitting
of the host that became effective only when sufficiently
bulky groups are present at the lower rim to give a minimal
partial preorganization.
Figure 4.
1H NMR spectra (CDCl3, 400 MHz, 298 K) of (a) 1a, b)
equimolar solution (3 mM) of 1a and di-n-pentylammonium TFPB-,
3b, (c) equimolar solution (3 mM) of 1a and dibenzylammonium
TFPB-, 3c.
+
endo-complexation with n-BuNH3 , both gave through-the-
annulus threading with di-n-pentylammonium cation 3b in
CDCl3 (Figure 4b). This demonstrates that the through-the-
annulus threading tolerates a larger cavity and that the exo-
+
oriented pentyl chains in pseudorotaxane (n-Pent)2NH2 ⊂1a
+
and (n-Pent)2NH2 ⊂2a positively contributed to the stabi-
To extend the validity of this “weakly-coordinating anion”
approach, we tested the complexation of TFPB salts of di-
n-alkylammonium cations. Thus, the addition of di-n-
lization of calixarene cone conformation, which would be
faster interconverting because of the small OMe groups.
+
+
pentylammonium 3b salt [(n-Pent)2NH2 ·TFPB-] to a CDCl3
Clearly, with n-BuNH3 cation this extra stabilization is
missing because only a single alkyl chain would be present.
Additional interesting through-the-annulus threadings were
obtained with dibenzylammonium salt 3c (Bn2NH2+·TFPB-)
and calix[6]arenes 1a (Figure 4c), 1b,8 or calix[7]arene 2a in
CDCl3.8 In the 1H NMR spectra of these pseudorotaxanes two
sets of unshielded and shielded benzyl signals can be always
observed corresponding to their exo- or endo-cavity disposition.
The most shielded signals were in the order benzylic-CH2 (∆δ
) 2.77-2.78), ortho-BnH (∆δ ) 2.60-2.61), meta-BnH (∆δ
) 2.10-2.34), and para-BnH (∆δ ) 1.49-1.64).8 The lowest
AMBER-energy geometry of complex Bn2NH2+⊂1b, (Figure
3c) evidenced a pseudorotaxane structure with the formation
of two NH· · ·O hydrogen bonds. Clearly, the position of
dibenzylammonium cation 3c is averaged on the NMR time
scale over the six equivalent O-atoms.
The apparent association constants for the above-mentioned
complexes (Table 1) were determined by integration of 1H NMR
spectra of their 1:1 titration mixture in CDCl3, which showed
slowly exchanging signals for both free and complexed guests.19
In the case of complexes n-BuNH3+⊂1b, Bn2NH2+⊂1b, and
Bn2NH2+⊂2a the signal of the free guest was below the
acceptable limit to allow a reliable direct determination of their
association constant.19
solution of calix[6]arene host 1b again caused dramatic
8
1
changes in its H NMR spectrum (Figure 2c), similar to
those discussed above, with the typical signature at highfield
negative values and the emergence of the ArCH2Ar AX
system (4.40 and 3.51 ppm, J ) 13.6 Hz).8,18 Complex
formation was also confirmed by a ESI(+) mass spectrum,8
which gave as the base peak a value of 1637 m/z corre-
8
+
sponding to supermolecular ion (n-Pent)2NH2 ⊂1b. A 2D
ROESY8 spectrum confirmed the presence of the di-n-
pentylammonium guest 3b inside the calix[6]arene cavity.
Monte Carlo conformational searches8 confirmed a pseu-
dorotaxane structure with the formation of two NH· · ·O
hydrogen bonds. In particular, the lowest Amber-energy struc-
ture of (n-Pent)2NH2+⊂1b shows (Figure 3b) that the secondary
ammonium nitrogen is slightly above the mean plane of the
six calixarene O-atoms at a distance of 0.34 Å, significantly
lower than that observed for n-BuNH3+⊂1b complex (0.76 Å).
Interestingly, hexamethoxycalix[6]arene 1a (Figure 4) and
heptamethoxycalix[7]arene 2a,8 which previously gave no
(18) Interestingly, a scalar J-coupling was observed in the COSY-458
spectrum between the +NH2 ammonium protons (5.42 ppm) and their linked
R (0.53 ppm) and R′ (2.75 ppm) methylene groups, making straightforward
the entire assignment.8 Thus, two different pentyl chains were observed in
+
the complex (n-Pent)2NH2 ⊂1b: one strongly shielded and inside the cavity
To overcome this latter problem and to compare the different
stability of the complex formed by a given host with different
(with ε, δ, γ, ꢀ, and R protons resonating at -0.42, -0.87, -1.20, -0.88,
and 0.53 ppm, respectively) and the other, at more normal values of chemical
shift (with ε′, δ′, γ′, ꢀ′, and R′ protons resonating at 0.94, 1.15, 1.63, 1.88,
and 2.75 ppm, respectively), outside the cavity.
(19) Hirose, K. J. Incl. Phenom. Macrocycl. Chem. 2001, 39, 193.
Org. Lett., Vol. 12, No. 9, 2010
2094

to two stereoisomeric directional pseudorotaxane com-
plexes.23 In particular, the addition of n-hexylbenzylammo-
nium 3f TFPB- salt [(n-Hex)NH2+Bn·TFPB-] to calix[6]arene
host 1b in CDCl3 clearly evidenced the formation of two
Table 1. Association Constants (Ka, M-1) and Percentages of
Formation of Butylammonium 3a and Dialkylammonium 3b,c
Complexes with Calixarene Hosts 1a,b and 2a Determined by
1H NMR Spectroscopy (400 MHz, CDCl3, 25 °C) upon
Addition of 1 equiv of Their Corresponding TFPB Salts
+
distinct diastereoisomeric (n-Hex)NH2 Bn⊂1b complexes
(Figure 5): one with a shielded endo-cavity n-hexyl chain
resonating at negative values and the other with a shielded
endo-cavity benzyl moiety resonating in the 3-5 ppm range.
Signal integration indicated that the endo-hexyl complex is
preferentially formed over the endo-benzyl stereoisomer in
a ratio of 4:1. Interestingly, the same experiment performed
with shorter alkyl chains such as pentyl (3e) [(n-
3a
3b
3c
1a
1b
2a
a
4.4 × 102 (42%)8
3.5 × 104 (84%)8
1.5 × 103 (59%)8
2.5 × 103 (67%)b
b
b
8.3 × 106
1.2 × 105
b
a
4.5 × 103
a No complexation takes place. b Obtained by competition experiments.8
+
+
Pent)NH2 Bn·TFPB-] and butyl (3d) (n-BuNH2 Bn·TFPB-)
groups led to higher endo-alkyl/endo-benzyl ratios of 10:1
and 30:1,8 respectively. Thus, the preference for endo-alkyl
stereoisomer increases with shorter alkyl chains. All these
results can be explained by assuming the above-mentioned
better conformational templation of a benzyl group with
respect to an alkyl one, in particular when this latter become
shorter. An alternative explanation can also be related to
different solvation patterns and/or entropic contributions of
benzyl versus alkyl group upon endo-calix inclusion.
guests, competition experiments20 were performed by treating
1 equiv of hexahexyloxycalix[6]arene 1b in CDCl3 with a
mixture of two different ammonium TFPB salts (1 equiv each).
+
Thus, it was evidenced that Bn2NH2 ⊂1b is preferentially
+
formed over (n-Pent)2NH2 ⊂1b in a ratio of 3:2, probably
because the less-flexible benzyl group may give a better
conformational templation at the lower rim with respect to
the pentyl one. A similar comparison indicated that
+
+
n-BuNH3 ⊂1b is preferentially formed over Bn2NH2 ⊂1b
in a 9:1 ratio probably because an higher number of hydrogen
bonds (3) can be formed in the first complex with respect to
the second one (2 H-bonds).21 From these results the relative
stability constants (Krel, Table S1 in Supporting Information)
can be calculated, which allowed the indirect determination
+
of association constants for complexes n-BuNH3 ⊂1b,
22
+
+
Bn2NH2 ⊂1b, and Bn2NH2 ⊂2a (Table 1).
As a final point, we decided to study the threading of an
unsymmetrical dialkylammonium cation that would give rise
(20) Tanaka, Y.; Kato, Y.; Aoyama, Y. J. Am. Chem. Soc. 1990, 112,
2807.
Figure 5.
1H NMR spectra (CDCl3, 400 MHz, 298 K) of equimolar
(21) The formation of the above endo-cavity and pseudorotaxane
complexes is induced by the very weakly coordinating anion TFPB-, and
it can be expected that these complexes can be broken by addition of any
specie able to interact competitively with the cation. In fact, we found that
the addition of 30 equiv of CD3CN leads to complete dethreading of the
di-n-pentylammonium guest from the cavity of calix[6]arene 1a [(n-
solution (3 mM) of 1b and n-hexylbenzylammonium TFPB- 3f.
In conclusion, we have reported here a general mode to obtain
endo-cavity complexation and through-the-annulus threading
of large calixarene by exploiting the inducing effect of the
weakly coordinating TFPB anion. This method works very well
with simple calixarene ethers, and no extensive or sophisticated
chemical modifications of the parent macrocycles are required.
In particular, the obtained pseudorotaxane complexes represent
the first examples of threading through larger calixarene annuli
by dialkylammonium cations.
+
Pent)2NH2 ⊂1a, 15 mM in CDCl3], whereas 62 equiv of CD3CN is
+
necessary for breaking the Bn2NH2 ⊂1a complex (15 mM in CDCl3). This
+
is in accordance with the higher stability (see Table 1) of Bn2NH2 ⊂1a
+
(Ka ) 2.5 × 103 M-1) with respect to (n-Pent)2NH2 ⊂1a complex (Ka )
+
4.4 × 102 M-1). Interestingly, the most stable complex n-BuNH3 ⊂1b (Ka
) 8.3 × 106 M-1) retained its structure even after the addition of 1300
equiv of CD3CN to its 5 mM solution in CDCl3. On the other hand, under
the same conditions the addition of 1.3 equiv of n-Bu4N+Cl- is sufficient
to give a complete decomplexation of the n-butylammonium guest of
+
n-BuNH3 ⊂1b, thus confirming the impressive effect produced by the ion-
pairing tendency of chloride anion. As expected, similar results were also
obtained by addition of appropriate amounts of bases.
(22) An alternative way to assess the stability of each complex can be based
on the reduction of the host conformational mobility upon complexation.
Therefore, VT 1H NMR studies (C2D2Cl4, 400 MHz) were conducted on both
1a and (n-Pent)2NH2+⊂1a, and a coalescence of ArCH2Ar signals was
ascertained at 278 and 358 K (see Table S2 in Supporting Information),
respectively. From these data an energy barrier for conformational intercon-
version of 12.4 and 16.3 kcal/mol, respectively, can be estimated. Comparison
of the two barrier values leads to a difference of about 3.9 kcal/mol, which
can be mainly attributed to the host conformational stabilization brought by
the ammonium templation. Indeed, this value compares very well with the ∆G
of complexation of 3.6 kcal/mol calculated by the experimental association
constant (4.4 × 102 M-1, Table 1). A similar favorable result was also obtained
in the case of Bn2NH2+⊂2a complex, whose ∆G of complexation of 5.0 kcal/
mol calculated by the experimental association constant (4.5 × 103 M-1, Table
1) was in accordance with a difference of energy barriers of 5.8 kcal/mol
between complexed and free 2a estimated by VT 1H NMR studies (see Table
S3 in Supporting Information).
Acknowledgment. Thanks are due to Dr. Patrizia Iannece
(Dipartimento di Chimica, Universita` di Salerno) for ESI-MS
measurements.
Supporting Information Available: Synthetic details, 1D
and 2D NMR spectra, details on molecular modeling, and
relative stability constants (Krel). This material is available
free of charge via the Internet at http://pubs.acs.org.
OL100578Z
(23) For recent examples diastereoisomeric pseudorotaxanes see: Arduini,
A.; Bussolati, R.; Credi, A.; Faimani, G.; Garaude´e, S; Pochini, A.; Secchi,
A.; Semeraro, M.; Silvi, S.; Venturi, M. Chem.sEur. J. 2009, 15, 3230.
Org. Lett., Vol. 12, No. 9, 2010
2095