Binding of acetylcholine and tetramethylammonium to flexible cyclophane receptors: Improving on binding ability by optimizing host's geometry

Article abstract of DOI:10.1021/jo049899j

The structure of a cyclophanic tetraester (1), previously employed for investigations on the cation-π interaction, has been optimized to better accommodate acetylcholine (ACh) and tetramethylammonium (TMA) guests. Following indications from molecular modeling calculations, a flexible cyclophane receptor of significantly improved binding properties has been obtained by removing the four carbonyl groups of the parent host. 2,11,20,29-Tetraoxa[3.3.3.3]paracyclophane (2) was prepared by an improved procedure, which was conveniently devised to avoid the formation of contiguous cyclooligomers that caused serious separation issues. Association of 2 with TMA picrate was measured in CDCl₃ at T = 296 K by ¹H NMR titrations and compared to binding data obtained for a set of reference hosts, including the parent tetraester 1, the corresponding cyclophanic tetraamine, the open-chain counterpart of 2, and its cyclooligomers from pentamer to octamer. Binding enhancements ranging from 15-fold (with respect to the tetraester and the tetraamine) to over 80-fold (with respect to the open-chain tetraether) were achieved by geometry optimization of the host. Binding of 2 to ACh and TMA was investigated for a variety of counterions. A constant binding free energy increment of nearly 8 kJ mol^-1 with respect to 1 was observed, independent from the anion and irrespective of the different structure of the cationic guests. Results showed that the electrostatic inhibiting contribution of the counterion to the cation's binding is a characteristic constant of each anion. The value of -ΔG° = 44.9 kJ mol^-1 extrapolated for TMA in the absence of a counterion indicates that 28-34 kJ mol^-1 of binding free energy are lost in ion pairing.

Full text of DOI:10.1021/jo049899j

Bin d in g of Acetylch olin e a n d Tetr a m eth yla m m on iu m to F lexible
Cyclop h a n e Recep tor s: Im p r ovin g on Bin d in g Ability by
Op tim izin g Host’s Geom etr y
Paolo Sarri, Francesca Venturi, Francesco Cuda, and Stefano Roelens*
CNR, Istituto di Chimica dei Composti Organometallici, Dipartimento di Chimica Organica,
Universita` di Firenze, Polo Scientifico, I-50019 Sesto Fiorentino, Firenze, Italy
stefano.roelens@unifi.it
Received J anuary 16, 2004
The structure of a cyclophanic tetraester (1), previously employed for investigations on the cation-π
interaction, has been optimized to better accommodate acetylcholine (ACh) and tetramethylam-
monium (TMA) guests. Following indications from molecular modeling calculations, a flexible
cyclophane receptor of significantly improved binding properties has been obtained by removing
the four carbonyl groups of the parent host. 2,11,20,29-Tetraoxa[3.3.3.3]paracyclophane (2) was
prepared by an improved procedure, which was conveniently devised to avoid the formation of
contiguous cyclooligomers that caused serious separation issues. Association of 2 with TMA picrate
was measured in CDCl3 at T ) 296 K by 1H NMR titrations and compared to binding data obtained
for a set of reference hosts, including the parent tetraester 1, the corresponding cyclophanic
tetraamine, the open-chain counterpart of 2, and its cyclooligomers from pentamer to octamer.
Binding enhancements ranging from 15-fold (with respect to the tetraester and the tetraamine) to
over 80-fold (with respect to the open-chain tetraether) were achieved by geometry optimization of
the host. Binding of 2 to ACh and TMA was investigated for a variety of counterions. A constant
binding free energy increment of nearly 8 kJ mol-1 with respect to 1 was observed, independent
from the anion and irrespective of the different structure of the cationic guests. Results showed
that the electrostatic inhibiting contribution of the counterion to the cation’s binding is a
characteristic constant of each anion. The value of -∆G° ) 44.9 kJ mol-1 extrapolated for TMA in
the absence of a counterion indicates that 28-34 kJ mol-1 of binding free energy are lost in ion
pairing.
In tr od u ction
On the other hand, flexibility is not always an enemy;
indeed, nonpreorganized receptors have been useful in
gaining an insight into the basic attractive force under-
lying the cation-π interaction: by making use of flexible
cyclophanic and open-chain phanic hosts, that is, recep-
tors possessing little or no preorganization at all, we were
able to gather a body of information on the intrinsic
affinity of quaternary ammonium cations for aromatics
The interaction of alkylammonium guests with syn-
thetic receptors possessing aromatic binding sites has
been central in the investigation of the cation-π inter-
action, mainly because of its biological implications.1
Since binding of quaternary ammonium cations to aro-
matics is weak, a substantial effort has been dedicated
to the design and the synthesis of rigid hosts possessing
preorganized cavities, which could exploit the advantage
of preorganization to increase their binding strength.
Typical examples are capsules2 and rigidified arenes,3
cyclophanes,4 and calixarenes,5 of which some cases
exhibiting outstanding binding properties have recently
been reported.6 Significant binding enhancements have
also been achieved with hosts capable of complexing the
whole ion pair,7 or by tayloring the host around the guest
through the innovative dynamic combinatorial approach.8
(2) (a) Shivanyuk, A.; Friese, J . C.; Doring, S.; Rebek, J . J . Org.
Chem. 2003, 68, 6489-6496. (b) Mansikkamaki, H.; Nissinen, M.;
Schalley, C. A.; Rissanen, K. New J . Chem. 2003, 27, 88-97. (c)
Corbellini, F.; Fiammengo, R.; Timmerman, P.; Crego, C. M.; Versluis,
K.; Heck, A. J . R.; Luyten, I.; Reinhoudt, D. N. J . Am. Chem. Soc. 2002,
124, 6569-6575. (d) Vysotsky, M. O.; Pop, A.; Broda, F.; Thondorf, I.;
Bohmer, V. Chem. Eur. J . 2001, 7, 4403-4410. (e) Shivanyuk, A.;
Rebek, J . Chem. Commun. 2001, 2374-2375. (f) Araki, K.; Hayashida,
H. Tetrahedron Lett. 2000, 41, 1209-1213. (g) Schalley, C. A.; Martin,
T.; Obst, U.; Rebek, J . J . Am. Chem. Soc. 1999, 121, 2133-2138. (h)
Zhong, Z. L.; Ikeda, A.; Shinkai, S. J . Am. Chem. Soc. 1999, 121,
11906-11907.
* Address correspondence to this author. Phone: +39-055-457-3546.
Fax: +39-055-457-3570.
(3) J asper, C.; Schrader, T.; Panitzky, J .; Klarner, F. G. Angew.
Chem., Int. Ed. 2002, 41, 1355-1358.
(1) (a) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem.,
Int. Ed. 2003, 42, 1210-1250. (b) Gokel, G. W.; DeWall, S. L.; Meadows,
E. S. Eur. J . Org. Chem. 2000, 2967-2978. (c) Lhotak, P.; Shinkai, S.
J . Phys. Org. Chem. 1997, 10, 273-285. (d) Ma, J . C.; Dougherty, D.
A. Chem. Rev. 1997, 97, 1303-1324. (e) Dougherty, D. A. Science 1996,
271, 163-168. (f) Scrutton, N. S.; Raine, A. R. C. Biochem. J . 1996,
319, 1-8.
(4) (a) Akabori, S.; Miura, M.; Takeda, M.; Yuzawa, S.; Habata, Y.;
Ishii, T. Supramol. Chem. 1996, 7, 187-193. (b) Meric, R.; Lehn, J .
M.; Vigneron, J . P. Bull. Soc. Chim. Fr. 1994, 131, 579-583. (c)
Kearney, P. C.; Mizoue, L. S.; Kumpf, R. A.; Forman, J . E.; Mccurdy,
A.; Dougherty, D. A. J . Am. Chem. Soc. 1993, 115, 9907-9919. (d)
Dhaenens, M.; Lehn, J . M.; Fernandez, M. J .; Vigneron, J . P. New J .
Chem. 1991, 15, 873-877.
10.1021/jo049899j CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/29/2004
3654
J . Org. Chem. 2004, 69, 3654-3661

Binding of Guests to Flexible Cyclophane Receptor
and on the energetics of the cation-π interaction.9 In
these studies, focused on tetramethylammonium (TMA)
and acetylcholine (ACh), we have extensively employed
the cyclophanic tetraester 1, mainly because of the
conveniently large upfield shifts induced upon binding
on the guests’ signals in the 1H NMR spectra, which
allowed for accurate measurements of binding constants.
paper we report that a significant improvement of
binding ability with respect to the parent cyclophane has
indeed been achieved just by appropriate modifications
of the geometry and the constitution of the cyclophane,
without resorting to a more rigid structure or to ad-
ditional attractive contributions.
Resu lts a n d Discu ssion
Design a n d Syn th esis. Molecular modeling calcula-
tions showed that 1 needs to wrap around TMA and the
cationic head of ACh to establish binding interactions;9e
such a conformational reorganization suggests that the
cyclophane cannot accommodate the guest inside the ring
cavity, even though the ring size is large enough for the
task. This feature is presumably due to the conforma-
tional restraints imposed by the four ester groups in the
backbone, which hamper the cyclophane from reaching
the proper binding conformation. From computational10
and crystallographic11 literature data we gathered that
optimal binding of TMA by a cyclophane receptor is
achieved when the cation is located in the center of a
region delimited by four aromatic rings, at a distance of
4.2-4.4 (computational) or 4.5-4.6 Å (crystallographic)
from the aromatics’ centroids. Such an arrangement
could be achieved by removing the ester carbonyl groups
from the structure of 1 to give a paracyclophanic tetra-
ether, in which four aromatic rings are bridged by three-
atom spacers; the central ether oxygen atom in the bridge
would avoid the steric hindrance of a full-carbon spacer,
caused by hydrogen atoms perching into the ring cavity;
removal of constraints and correct geometry could thus
be simultaneously achieved from 1 through this trans-
formation. Indeed, molecular modeling calculations in-
dicated that the tetraoxa[3.3.3.3]paracyclophane 2 can
assume a binding conformation capable of accommodat-
ing TMA in the center of the ring’s cavity at appropriate
distances from the aromatics’ centroids (4.2-4.7 Å), and
corresponding space-filling models showed that the cavity
is correctly filled by the guest (Figure 1).
Despite the useful shift variations, the observed as-
sociation constants were indeed very modest. Some
obvious questions are the following: Can the binding
ability of the cyclophanic tetraester 1 be improved
through structural modifications, still preserving the
flexibility of the host and the large induced shifts? How
much can be gained in binding energy with a flexible
cyclophane just by geometry optimization? Since the
interaction is electrostatic in nature,1d,9b,e how critical are
ring size and functional groups nature? In an effort to
answer these questions, we thought that even flexible
nonpreorganized cyclophanes might show significantly
enhanced binding properties based on pure cation-π
interactions, provided that the steric, geometric, and
coordinative requirements of the guest could be finely
matched by a correctly designed host. In the present
(5) (a) Arduini, A.; Pochini, A.; Secchi, A. Eur. J . Org. Chem. 2000,
2325-2334. (b) Sansone, F.; Barboso, S.; Casnati, A.; Sciotto, D.;
Ungaro, R. Tetrahedron Lett. 1999, 40, 4741-4744. (c) Arduini, A.;
Mcgregor, W. M.; Paganuzzi, D.; Pochini, A.; Secchi, A.; Ugozzoli, F.;
Ungaro, R. J . Chem. Soc., Perkin Trans. 2 1996, 839-846. (d) Casnati,
A.; J acopozzi, P.; Pochini, A.; Ugozzoli, F.; Cacciapaglia, R.; Mandolini,
L.; Ungaro, R. Tetrahedron 1995, 51, 591-598. (e) Takeshita, M.;
Nishio, S.; Shinkai, S. J . Org. Chem. 1994, 59, 4032-4034.
(6) (a) Hof, F.; Trembleau, L.; Ullrich, E. C.; Rebek, J . Angew. Chem.,
Int. Ed. 2003, 42, 3150-3153. (b) Atwood, J . L.; Szumna, A. Chem.
Commun. 2003, 940-941. (c) Ballester, P.; Shivanyuk, A.; Far, A. R.;
Rebek, J . J . Am. Chem. Soc. 2002, 124, 14014-14016. (d) Atwood, J .
L.; Szumna, A. J . Am. Chem. Soc. 2002, 124, 10646-10647. (e) Garel,
L.; Lozach, B.; Dutasta, J . P.; Collet, A. J . Am. Chem. Soc. 1993, 115,
11652-11653.
(7) (a) Shivanyuk, A.; Friese, J . C.; Rebek, J . Tetrahedron 2003, 59,
7067-7070. (b) Arduini, A.; Brindani, E.; Giorgi, G.; Pochini, A.; Secchi,
A. J . Org. Chem. 2002, 67, 6188-6194. (c) Arduini, A.; Giorgi, G.;
Pochini, A.; Secchi, A.; Ugozzoli, F. J . Org. Chem. 2001, 66, 8302-
8308. (d) Kubik, S.; Goddard, R. Eur. J . Org. Chem. 2001, 311-322.
(e) Kubik, S. J . Am. Chem. Soc. 1999, 121, 5846-5855. (f) Kubik, S.;
Goddard, R. J . Org. Chem. 1999, 64, 9475-9486.
(8) (a) Furlan, R. L. E.; Ng, Y. F.; Cousins, G. R. L.; Redman, J . E.;
Sanders, J . K. M. Tetrahedron 2002, 58, 771-778. (b) Otto, S.; Furlan,
R. L. E.; Sanders, J . K. M. Science 2002, 297, 590-593. (c) Roberts, S.
L.; Furlan, R. L. E.; Cousins, G. R. L.; Sanders, J . K. M. Chem.
Commun. 2002, 938-939. (d) Cousins, G. R. L.; Furlan, R. L. E.; Ng,
Y. F.; Redman, J . E.; Sanders, J . K. M. Angew. Chem., Int. Ed. 2001,
40, 423-428.
(9) (a) Bartoli, S.; De Nicola, G.; Roelens, S. J . Org. Chem. 2003,
68, 8149-8156. (b) Bartoli, S.; Roelens, S. J . Am. Chem. Soc. 2002,
124, 8307-8315. (c) Bartoli, S.; Roelens, S. J . Am. Chem. Soc. 1999,
121, 11908-11909. (d) Roelens, S.; Torriti, R. Supramol. Chem. 1999,
10, 225-232. (e) Roelens, S.; Torriti, R. J . Am. Chem. Soc. 1998, 120,
12443-12452.
The synthesis of 2 was accomplished by the method of
Komatsu,12 by reacting terephthalaldehyde and trieth-
(10) (a) Felder, C.; J iang, H. L.; Zhu, W. L.; Chen, K. X.; Silman, I.;
Botti, S. A.; Sussman, J . L. J . Phys. Chem. A 2001, 105, 1326-1333.
(b) Pullman, A.; Berthier, G.; Savinelli, R. J . Comput. Chem. 1997,
18, 2012-2022. (c) Caldwell, J . W.; Kollman, P. A. J . Am. Chem. Soc.
1995, 117, 4177-4178. (d) Duffy, E. M.; Kowalczyk, P. J .; J orgensen,
W. J . J . Am. Chem. Soc. 1993, 115, 9271-9275.
(11) From a search of the Cambridge Crystallographic database
(CSD), the distance between the quaternary nitrogen and the centroid
of the aromatic ring for all the available structures containing
quaternary ammonium-aromatic ring contacts below 5 Å shows a
maximum at 4.56 Å (116 structures), with 126 structures showing a
distance in the 4.2-4.4 Å range.
(12) Komatsu, N.; Chishiro, T. J . Chem. Soc., Perkin Trans. 1 2001,
1532-1537.
J . Org. Chem, Vol. 69, No. 11, 2004 3655

Sarri et al.
repeated chromatographic and crystallization steps and
eventually leading to very poor isolated yields of pure
compounds.13 To overcome these limitations, we devised
a modified procedure that conveniently avoids the forma-
tion of contiguous oligomeric products (Scheme 2). The
key feature of this procedure consists of reacting a
preformed “trimeric” dialdehyde instead of terephthala-
ldehyde with the bis-silyl ether of 1,4-benzenedimethanol
in a 3 + 1 cyclization reaction, which ties up the
cyclooligomerization to proceed in steps multiple of 4
units, thus excluding the formation of those cyclooligo-
mers that caused separation issues. Reaction mixtures
obtained this way could be efficiently separated by a
single chromatographic step to give 2 in pure form.
A second major improvement was obtained by changing
the order of addition of the reactants and taking advan-
tage of the influx technique in the cyclization step, that
is, by slow addition through a syringe pump of a 1:1
mixture of the two reactants to a solution of the reducing
agent and the catalyst. Analytical yields of 2 in the
reaction mixture were above 10%, giving 8-9% of isolated
pure product by flash column chromatography. The
longer synthesis compared to the original method, re-
quired for the preparation of the dialdehyde 9 from
commercial terephthalaldehyde mono-diethylacetal, was
compensated by the limited number of high-yielding steps
that did not require the isolation of intermediates 7 and
8.
The phanic tetraether 10, the open-chain counterpart
of 2, was prepared as a reference compound by a similar
procedure, i.e., by reacting the dialdehyde 9 with the
trimethylsilyl ether of benzyl alcohol in the presence of
catalytic amounts of trimethylsilyl triflate. For compari-
son reasons, the cyclophanic tetraamine 11 was also
prepared by literature methods,14 to ascertain whether
binding properties may be affected by the nature of the
heteroatom in the spacers or by slight variations in the
size of the ring.
F IGURE 1. PM3-computed model of the complex between 2
and TMA: (A) ball-and-stick representation and (B) space-
filling representation. PM3 optimization was performed on a
complex’s geometry obtained through a molecular mechanics
(MMFF94) conformational search.
SCHEME 1. Syn th esis of Oligom er ic
Oxa [3.n ]p a r a cyclop h a n es
ylsilane in the presence of catalytic amounts of trimeth-
ylsilyl triflate (Scheme 1).
Bin d in g Stu d ies. To evaluate the binding properties
of the described hosts, the association constants (Ka) for
the formation of the 1:1 complexes with tetramethylam-
monium picrate (TMAPic) were measured in CDCl3 at T
This reductive cyclooligomerization reaction afforded
a whole family of oxa[3.n]paracyclophanes, from which
the oligomers from the tetramer to the octamer (2-6)
could be isolated as pure compounds. Besides the target
cyclophanic tetraether, a set of cyclooligomeric homo-
logues were thus available to ascertain the effect of ring
size on binding properties. Although Komatsu’s reaction
is a practical entry to cyclophanic polyethers, it suffers
from two major drawbacks: (a) optimized (analytical)
yields of tetrameric 2 never exceeded 5% on a 10 mmol
scale at 0.02 M substrate concentration and (b) separa-
tion of cyclic oligomers is extremely tedious, requiring
1
) 296 K by H NMR titrations, in analogy to previous
studies,9 monitoring the time-averaged signal of the free
and complexed TMA with increasing host concentration.
The results are reported in Table 1, together with data
(13) In the original work, the authors stated that separation of
oligomers was accomplished by GPC chromatography. See ref 12, p
1532.
(14) Shinmyozu, T.; Shibakawa, N.; Sugimoto, K.; Sakane, H.;
Takemura, H.; Sako, K.; Inazu, T. Synthesis 1993, 1257-1260.
3656 J . Org. Chem., Vol. 69, No. 11, 2004

Binding of Guests to Flexible Cyclophane Receptor
SCHEME 2. Syn th esis of Tetr a oxa [3.3.3.3]p a r a cyclop h a n e 2
TABLE 1. Associa tion Con sta n ts Ka (M-1), Sta n d a r d
1) indicates to what extent the correct size of the
F r ee En er gies of Bin d in g -∆G° (k J m ol-1), a n d Gu est
Lim itin g Sh ift Va lu es ∆δ∞ (p p m ) for th e F or m a tion of 1:1
Com p lexes of Tetr a m eth yla m m on iu m P icr a te (TMAP ic)
w ith P h a n ic Hostsa
cyclophane is significant to its binding capabilities:
except for the heptamer 5, whose association is undetect-
able, all the other oligomers (3-6) show binding free
energies measurable but distinctly smaller than 1, sug-
gesting a reduced adaptivity to the guest with respect to
the latter, despite the larger number of binding sites and
the lack of conformational restraints.
host
Ka (SE)b
-∆δ∞
-∆G°
1c
1c,d
2
3
4
29.7 (0.4)
6.6 (0.1)
458 (6)
14.2 (0.1)
11.3 (0.1)
n.d.e
1.479
1.150
2.090
1.225
0.781
n.d.e
8.35(3)
4.64(4)
15.08(3)
6.53(2)
5.97(2)
n.d.e
Some degree of organization is however evident from
comparison with the open-chain tetraether 10, which can
be viewed as the nonpreorganized reference; 3, 4, and 6
5
bind TMAPic stronger than 10 by 1.8-2.7 kJ mol-1
,
6
16.4 (0.5)
5.5 (0.1)
29 (1)
0.798
0.909
0.755
0.910
6.88(8)
4.20(4)
8.29(8)
5.57(2)
10
11
11d
demonstrating little binding energy gain, with respect
to an open-chain counterpart, for cyclophanic hosts of
9.6 (0.1)
a
Measured by 1H NMR (200/300 MHz) at T ) 296 K in CDCl3
on 0.1 mM solutions of TMAP, using host concentrations up to 60
mg/mL. Standard error of the nonlinear least-squares fit. c Data
b
from ref 9b. Data for tetramethylammonium chloride. e Non
d
detectable; for Ka e 2 titration curves become indistinguishable
from each other in the investigated concentration range.
previously obtained for 1 under the same conditions. It
can immediately be noted that the cyclophanic tetraether
2 is by far the best receptor of the whole set: the observed
binding constant is over 15 times larger that that of the
parent cyclophane 1, with an increase in binding free
energy of nearly 7 kJ mol-1, while the larger induced shift
upon complexation indicates a closer contact of the guest
with the aromatic π faces. The improvement achieved
through the devised transformation supports the hypoth-
eses made about the optimal geometry of a cyclophanic
receptor for TMA. A further support is provided by the
structure of 2 in the solid state that, to avoid empty
space, adopts a flattened conformation in which opposite
pairs of aromatic rings are parallel but staggered (Figure
2); extensive reorganization is consequently required to
reach the correct binding conformation but, despite the
energetic cost of such a reorganization, the binding
improvement obtained with respect to 1 is in fact
substantial. Comparison with higher cyclooligomers (Table
F IGURE 2. Ball-and-stick representation of the X-ray struc-
ture of 2: (A) view down two of the four aromatic ring planes
and (B) view down the other two aromatic ring planes.
Hydrogen atoms are omitted for clarity in part B. Opposite
pairs of aromatic rings are parallel but staggered.
J . Org. Chem, Vol. 69, No. 11, 2004 3657

Sarri et al.
TABLE 2. Associa tion Con sta n ts Ka (M-1), Gu est Lim itin g Sh ift Va lu es ∆δ∞ (p p m ), Sta n d a r d F r ee En er gies of Bin d in g
-∆G° (k J m ol-1), In cr em en ts of Sta n d a r d F r ee En er gy of Bin d in g -∆∆G° (k J m ol-1) w ith Resp ect to 1, a n d Ca lcu la ted
Electr osta tic P oten tia l EP (k ca l m ol-1) for 1:1 Com p lexes of 2 w ith Acetylch olin e (ACh ) a n d Tetr a m eth yla m m on iu m
(TMA) Sa ltsa
guest
AChe
anion
Ka (SE)b
-∆δ∞
-∆G°
-∆∆G° c
EPd
Cl
92.5 (0.4)
2.012 (CH3N)
1.522 (CH2N)
2.702 (CH2O)
2.351 (CH3N)
1.647 (CH2N)
2.857 (CH2O)
2.078 (CH3N)
1.440 (CH2N)
2.531 (CH2O)
2.025
2.169
1.659
2.147
2.038
2.050
2.069
2.090
11.14(1)
7.93
36.715
f
Me2SnCl3
Pic
443 (6)
362 (2)
15.00(3)
14.50(1)
7.45
8.17
47.815
TMA
Cl
165 (2)
1004 (2)
75 (2)
1068 (3)
215 (2)
556 (4)
303 (5)
458 (6)
12.57(3)
17.01(1)
10.63(6)
17.16(1)
13.22(2)
15.56(1)
14.06(4)
15.08(3)
7.93
8.20
8.03
41.590
37.511
g
Me2SnCl3
AcO
h
Me2Sn(OAc)Cl2
TFA
TfO
PFF
Pic
7.17
7.88
7.48
6.73
44.708
47.920
45.253
49.362
a
Measured by 1H NMR (200/300 MHz) at T ) 296 K in CDCl3 on 0.1-1 mM solutions of salt, using host concentrations up to 0.04 M.
b
Pic: picrate; AcO: acetate; TFA: trifluoroacetate; TfO: triflate; PFF: pentafluorophenate. Standard error of the nonlinear least-squares
fit. c Increments calculated from data reported in ref 9b. Data from ref 9b. e Simultaneous fit of the three most shifting signals. f [AChCl]
d
) 1.1 × 10-3 M; [DMTC] ) 4.3 × 10-2 M. [TMACl] ) 5.0 × 10-4 M; [DMTC] ) 4.2 × 10-2 M. [TMAOAc] ) 7.5 × 10-4 M; [DMTC] )
g
h
4.2 × 10-2 M.
grossly mismatched size for the guest. On the contrary,
the association constant ratio between 2 and 10 sets to
over 80-fold the binding advantage shown by the cyclic
over the open-chain structure, i.e., the macrocyclic effect
exhibited by this phanic receptor. That a correct host-
guest fit is critical is demonstrated even more clearly by
the results for tetraamine 11. Although the ring size is
very similar to that of 2, the observed binding constant
is nearly identical with that of the parent cyclophane 1,
that is, 15-fold smaller than 2, a much too large drop to
be explained on the basis of pure size arguments. In
addition, it has been demonstrated that the contribution
from ether oxygen donors to the binding of TMA is
markedly smaller than that of aromatic rings;9a thus, a
significant contribution from ether oxygen donors in 2
can be ruled out and, likewise, it can hardly be expected
from amine nitrogen donors in 11. We believe that the
marked drop of binding ability, compared to 2, observed
for 11 should most reasonably be ascribed to the hydro-
gen atoms on nitrogen, likely pointing toward the cavity
and therefore preventing the guest from fitting in or,
alternatively, hampering the cyclophane from reaching
the correct binding conformation. Tetraamine 11 may
therefore suffer from a binding inhibition that may be
steric or conformational in origin, although possessing
the correct size. Eventually, in 11 the hydrogen atoms
on nitrogen could be involved in hydrogen bonding to the
anion, leading to enhanced association with the host due
to synergistic binding of the ion pair.7 Occurrence of
hydrogen bonding may not be evident with picrate, but
it would emerge with chloride, a counterion much more
effective than picrate as H-bonding acceptor. Comparison
of results for the association of TMA chloride with 1 and
11 clearly shows that the chloride anion exerts a very
similar inhibiting effect on binding for both hosts, exclud-
ing any synergistic contribution from hydrogen bonding.
In conclusion, the improved binding capabilities of 2
appear to rely essentially on correct geometry of the
binding site.
Con tr ibu tion of th e An ion . It has been shown that
the counterion exerts an inhibiting effect on the inter-
action of ACh and TMA with 1, which is due to the
electrostatic attraction exerted by the anion on the cation
and is stronger for anions of larger charge density.9b,c It
has also been shown that converting a strongly inhibiting
anion, such as chloride, into a more charge-dispersed
species, such as dimethyltrichlorostannate, induces a
marked binding enhancement of the corresponding TMA
and ACh salts to 1.9b,c Since we expected these effects to
be general, we tested the contribution of several coun-
terions to the association of ACh and TMA salts with 2.
The results obtained for a set of anions selected among
those employed in binding studies with 1 are reported
in Table 2.15 Significantly enhanced binding constants
with respect to 1 were obtained with both guests for all
the salts investigated; corresponding -∆G° values were
all larger than 10 kJ mol-1, spanning a range of 7 kJ
mol-1, whereas for 1 -∆G° values, although covering the
same range, did not exceed the limit of 9 kJ mol-1
.
Surprisingly, conversion of chloride into dimethyltrichlo-
rostannate induced for both guests the same binding
enhancement of 4 kJ mol-1 observed for 1, despite the
different structures of the cations and independently of
the markedly different binding abilities of the two hosts.
While confirming the beneficial effect of charge dispersion
on cation binding, which raised the association constant
to the order of 103 M-1, the unexpectedly constant
increment suggested that participation of anions with a
constant contribution may be a general feature. The effect
of charge dispersion was even more evident in the
conversion of acetate into the corresponding dimethyl-
acetoxydichlorostannate anion, which raised the binding
free energy of TMA to 2 by 6.5 kJ mol-1, with a 14-fold
increment of the association constant. In the latter case,
the combined contributions of the host and anion set to
(15) Stannate salts are generated in situ by addition of excess
dimethyltin dichloride to a solution of the chloride or acetate salt in
deuteriochloroform as previously described. See ref 9b,c.
3658 J . Org. Chem., Vol. 69, No. 11, 2004

Binding of Guests to Flexible Cyclophane Receptor
0.5 kJ mol-1 average increment value of -∆G° found
between 1 and 2. The increment 1 kJ mol-1 smaller than
average for picrate may be explained on the basis of steric
effects. The larger shifts induced on all guests upon
binding (Table 2) compared to 1 indicate a closer contact
between the guest and the aromatic cavity of the host
for 2, that is, a deeper fit into the host’s cavity; under
these conditions, binding of tight ion pairs18 may suffer
from steric hindrance for large anions possessing bulky
substituents, such as picrate, to a greater extent than in
the case of 1, which relies on wrapping around the guest
for binding.
Although based on a limited number of points, the
linear correlation running parallel to that obtained for 1
inequivocally confirmed (a) the extent of the binding
increase from the intercept difference, (b) that the
enhancement is independent of the counterion, and (c)
that the sensitivity to the electrostatic inhibition from
the anion is the same for both hosts. Eventually, as in
the case of 1, the free energy of binding for the “free”
TMA cation, i.e., in the absence of the counterion, may
be extrapolated for 2, giving a value of -∆G° ) 44.9 kJ
mol-1, which consistently is 6.3 kJ mol-1 larger than that
found for 1 and which shows that, for the anions
scrutinized, 28-34 kJ mol-1 of binding free energy are
lost in ion pairing because of the unavoidable presence
of a counterion.19 The most relevant conclusion is,
however, that a characteristic constant can be assigned
to each anion to describe the electrostatic inhibition
factor: calculated EP values may be used as a quantita-
tive measure of these constants, which can be generally
employed for predicting binding strengths as long as
factors other than pure electrostatics are not involved.
Con clu sion . We have redisigned the structure of a
cyclophanic tetraester (1) that has been previously used
to investigate the cation-π interaction to improve its
binding ability. Molecular modeling calculations indi-
cated that the cyclophanic tetraether resulting from
removal of the four carbonyl groups would better accom-
modate the tetramethylammonium (TMA) cation inside
the ring cavity. Cyclophanic tetraether 2 has thus been
prepared, together with a set of appropriate reference
compounds, and its binding properties toward TMA and
acetylcholine (ACh) were measured by using a variety
of counterions. From a comparative analysis of binding
data, the following conclusions have been drawn: (a) the
binding ability of the flexible cyclophane 1 was signifi-
cantly improved by a correct redesign of the host’s
structure, without resorting to a more preorganized
architecture; (b) a marked binding enhancement was
achieved with 2 not only with respect to the parent
tetraester, but also compared to its open-chain counter-
part, to its oligomeric higher homologues, and to the
corresponding cyclophanic teraamine of the same ring
F IGURE 3. Plot of ion pair electrostatic potential vs standard
free energy of binding to 2 for TMA salts. Data points are
experimental -∆G° values vs calculated EP values from Table
2; the solid line is the best fit line calculated by linear
regression: slope ) 0.465; intercept ) -6.822; r2 ) 0.997.
over 400-fold the binding increase achieved with respect
to TMA acetate and 1, corresponding to nearly 15 kJ
mol-1 in terms of binding free energy.
A general feature of data in Table 2 was that binding
constant values followed the same trend observed for host
1, that is, they were ranked according to the previously
observed anion’s inhibiting power. Even more interest-
ingly, the binding enhancement with respect to 1 ap-
peared to be constant throughout the series. This feature
can be better appreciated by looking at the increments
in binding free energy -∆∆G° with respect to 1 that, with
only one exception, cluster within 1 kJ mol-1 around the
value of 8 kJ mol-1 (7.7 ( 0.5 kJ mol-1). This remarkable
evidence shows that (a) 2 binds nearly 8 kJ mol-1
stronger than 1 to both guests regardless of the counter-
ion,16 (b) the anion’s inhibiting effect on cation’s binding
is a constant contribution for each anion independent
from the host, as anticipated on the basis of the results
for anion conversion, and (c) the observed trend appears
to be generally valid for different hosts. To validate these
conclusions, since for host 1 the inhibiting power of
anions was found to correlate with the electrostatic
potential EP of the ion pair as long as purely electrostatic
factors are involved,9b -∆G° values obtained for 2 were
plotted vs the calculated ion pair EP values reported in
Table 2.17 A good linear correlation was found for all data
(r2 ) 0.945) that became excellent (r2 ) 0.997) when
excluding ACh salts, which were shown to be affected
by factors other than pure electrostatics,9b with the value
slightly out of range for picrate (Figure 3).
The slope of the linear regression (slope ) 0.465) is
very close to the value obtained in the case of 1 (slope )
0.483), while the intercept (intercept ) - 6.82) indicates
that the correlation line is shifted up by 8.6 kJ mol-1 with
respect to that of 1, in close agreement with the 7.7 (
(18) The investigated salts have been shown to be present as tight
ion pairs in CDCl3 solution by spectroscopic evidence and conductivity
measurements. See ref 9b,c,e. Binding of quaternary ammonium salts
as tight ion pairs has also been demonstrated in the formation of
complexes between aromatic crown ethers and paraquats and vio-
lagens: Huang, F.; J ones, J . W.; Slebodnick, C.; Gibson, H. W. J . Am.
Chem. Soc. 2003, 125, 14458-14464.
(16) The independence of the intrinsic host’s binding ability from
the counterion in cation-π interactions has also been observed in other
systems: Hunter, C. A.; Low, C. M. R.; Rotger, C.; Vinter, J . G.; Zonta,
C. Chem. Commun. 2003, 834-835.
(17) For a discussion on calculated EP values and computational
methods, see ref 9b.
(19) The large amount of binding free energy lost in ion pairing,
corresponding to ion pair association constants Kip ) 105-106 M-1
,
confirmed that in chloroform ion pair dissociation is negligible and,
therefore, considering the whole ion pair as guest in the host-guest
association is a correct approximation.
J . Org. Chem, Vol. 69, No. 11, 2004 3659

Sarri et al.
2,11,20,29,38,47,56,65-Oct a oxa [3.3.3.3.3.3.3.3]p a r a cy-
size; (c) geometry is as crucial as ring size, but the
functional group’s nature is not, as long as it does not
impose conformational constraints; (d) the binding en-
hancement is independent from the counterion and from
the structure of the cationic guest, as may be expected
for electrostatic interactions; (e) the counterion’s inhibit-
ing contribution to binding appears to be a characteristic
constant of each anion, the value of which can be
expressed by the calculated electrostatic potential of the
ion pair; (f) with the appropriate choice of the anion, i.e.,
with a poorly inhibiting counterion, the association
constant for TMA is raised to the order of 103 M-1, with
a binding increase of over 400-fold with respect to 1,
corresponding to nearly 15 kJ mol-1 in terms of binding
free energy. In summary, a rational optimization of the
structure of a cyclophanic host has led to a synthetic
receptor of significantly improved binding properties and
to a deeper understanding of the factors and the contri-
butions involved in the cation-π interaction, such as the
predominant role of ion pairing.
clop h a n e, 6. White solid (35 mg, 2.4%). Mp 116.5-117 °C
1
(lit.12 mp 122-124 °C). H NMR (200 MHz, 0.1 F in CDCl3) δ
7.35 (s, 32 H), 4.54 (s, 32 H). 13C NMR (50 MHz, 0.1 F in CDCl3)
δ 137.7, 127.9, 71.8. Anal. Calcd for C64H64O8: C 79.97, H 6.71.
Found: C 79.75, H 6.72.
Syn t h esis of 2,11,20,29-Tet r a oxa [3.3.3.3]p a r a cyclo-
p h a n e, 2. To a solution of terephthalaldehyde mono-diethyl-
acetal (5.03 g, 24.16 mmol) in THF (50 mL) was added sodium
borohydride (1.41 g, 37.24 mmol) portionwise under stirring
at 0 °C. The mixture was stirred at room temperature for a
further 1.5 h, then poured into water and extracted with 4
portions of dichloromethane, washed with water, and dried
over anhydrous Na2SO4. The hydroxymethyl derivative 7 (5.07
g, quantitative) was obtained by removing the solvent under
reduced pressure as a colorless oil, which was used as such
1
for the next step. H NMR (200 MHz, CDCl3) δ 7.48-7.32 (m,
4 H), 5.49 (s, 1 H), 4.66 (br s, 2 H), 3.69-3.44 (m, 4 H), 2.07
(br s, 1 H), 1.23 (t, J ) 7.4 Hz, 6 H). 7 (4.73 g, 22.50 mmol)
was dissolved in DMSO (60 mL) and the solution was degassed
and saturated with Ar. NaH (60% dispersion in oil, 0.9 g, 22.50
mmol) was added and the suspension was heated under
stirring in an Ar atmosphere at 80 °C, until hydrogen evolution
had ceased leaving a clear solution. 1,4-Bis-bromomethylben-
zene (2.97 g, 11.25 mmol) was quickly added, stirring the
mixture for a further 0.5 h at 80 °C. The solution was then
poured into water (150 mL), extracted with dichloromethane
(4 × 50 mL), washed with water (4 × 50 mL), and dried over
anhydrous Na2SO4. By removing the solvent under reduced
pressure, the bis-benzyl ether 8 was obtained as a yellow oil
(5.37 g, 91%), which was used for the subsequent step without
Exp er im en ta l Section
Syn t h esis of Oligom er ic Oxa [3.n ]p a r a cyclop h a n es,
2-6. To a solution of terephthalaldehyde (1.6 g, 11.93 mmol)
in dry dichloromethane (480 mL) was added trimethylsilyl
triflate (0.22 mL, 1.2 mmol) at 0 °C under stirring. A solution
of triethylsilane (3.05 g, 26.23 mmol) in dry dichloromethane
(120 mL) was added dropwise in 30 min at 0 °C and the
mixture was then stirred at 0 °C for a further 3 h. The solution
was washed with NaHCO3 (2 × 200 mL) and water (2 × 200
mL) and dried over anhydrous Na2SO4, and the solvent was
removed under reduced pressure to give a white solid (4.76 g)
that was purified by flash column chromatography (silica gel
60, dichloromethane/diethyl ether 95:5) to give the cyclic
oligomers 2-6 (493 mg, 34% yield), of which 5 and 6 were
obtained as pure compounds. The residual mixture [analytical
yields (NMR): 4% (2), 14% (3), 7% (4)] was repeatedly
chromatographed (dichloromethane/diethyl ether, from 95:5 to
85:15) until 2 and 3 could be obtained as pure compounds. The
residual solid was then crystallized from acetonitrile to give
pure 4.
1
further purification. H NMR (200 MHz, CDCl3) δ 7.49-7.36
(m, 12 H), 5.50 (s, 2 H), 4.55 (s, 8 H), 3.70-3.46 (m, 8 H), 1.24
(t, J ) 7.1 Hz, 12 H). A solution of 8 (5.28 g, 10.10 mmol) in
acetone (150 mL) and water (1.5 mL) was left to stand
overnight at room temperature over Amberlyst 15 (1.2 g). The
resin was filtered off and the solvent was removed under
reduced pressure to give 3.97 g of a yellow oil that solidified
on standing. The crude dialdehyde 9 was purified by flash
column chromatography (silica gel 60, petroleum ether/ethyl
acetate 3:2) to give 9 as a pure compound. White solid (2.76 g,
1
66% overall yield). Mp 84-86 °C. H NMR (200 MHz, CDCl3)
δ 10.00 (s, 2 H), 7.89-7.51 (m, 8 H), 7.39 (s, 4 H), 4.64 (s, 4
H), 4.61 (s, 4 H). 13C NMR (50 MHz, CDCl3) δ 191.9, 145.4,
137.5, 129.9, 127.9, 127.7, 72.4, 71.4. Trimethylsilyl triflate
(0.035 mL, 0.2 mmol) was added at 0 °C to a solution of
triethylsilane (308 mg, 2.65 mmol) in dry dichloromethane (42
mL). A solution of 9 (296 mg, 0.79 mmol) and 1,4-bis-
trimethylsilyloxymethylbenzene (223 mg, 0.79 mmol) in dry
dichloromethane (21 mL) was added in 7 h at 0 °C by means
of an infusion pump. The clear solution was washed with water
(4 × 20 mL) and dried over anhydrous Na2SO4 to give, after
solvent removal, 531 mg of a white solid that was washed twice
with petroleum ether to remove siloxanes, affording a crude
oligomeric mixture (330 mg) as a white powder. The mixture
showed the presence of 2 (10.5% by NMR) and no evidence of
oligomers 3-5 by gel permeation chromatography (GPC).
Separation by flash column chromatography (silica gel 60,
petroleum ether/acetone 8:2) gave pure 2 (32 mg, 8.4%) as a
white powder. A sample of 2 was crystallized from acetonitrile
by slow evaporation of the solvent to obtain X-ray quality
crystals. The identity and the structure assignment of 2 was
confirmed by an X-ray crystallographic structure determina-
tion that, however, could not be refined to an R factor lower
than 14%. Details of structure determination and refinement,
together with an ORTEP plot with atom numbering, are
reported in the Supporting Information.
2,11,20,29-Tetr a oxa [3.3.3.3]p a r a cyclop h a n e, 2. White
solid (21 mg, 1.5%). Mp 139.5-140 °C (lit.12 mp 145-147 °C).
1H NMR (200 MHz, 0.1 F in CDCl3) δ 7.19 (s, 16 H), 4.50 (s,
16 H). 13C NMR (50 MHz, 0.1 F in CDCl3) δ 137.9, 128.1, 71.8.
ESI-M: 498.2 (M + NH4+), 503.2 (M + Na+), 519.3 (M + K+).
Anal. Calcd for C32H32O4: C 79.97, H 6.71. Found: C 79.71, H
6.62.
2,11,20,29,38-P en t a oxa [3.3.3.3.3]p a r a cyclop h a n e, 3.
White solid (145 mg, 10%). Mp 138-139 °C (lit.12 mp 143-
1
145 °C). H NMR (200 MHz, 0.1 F in CDCl3) δ 7.35 (s, 20 H),
4.56 (s, 20 H). 13C NMR (50 MHz, 0.1 F in CDCl3) δ 137.6,
127.8, 71.5. ESI-MS 618.3 (M + NH4+), 623.3 (M + Na+), 639.4
(M + K+), 646.3 (M + 2Na+). Anal. Calcd for C40H40O5: C
79.97, H 6.71. Found: C 80.21, H 6.60.
2,11,20,29,38,47-Hexa oxa [3.3.3.3.3.3]p a r a cyclop h a n e, 4.
White solid (65 mg, 4.5%). Mp 141-142 °C (lit.12 mp 155-156
1
°C). H NMR (200 MHz, 0.1 F in CDCl3) δ 7.36 (s, 24 H), 4.55
(s, 24 H). 13C NMR (50 MHz, 0.1 F in CDCl3) δ 137.6, 127.9,
71.7. ESI-MS 738.3 (M + NH4+), 743.3 (M + Na+). Anal. Calcd
for C48H48O6: C 79.97, H 6.71. Found: C 80.07, H 6.56.
2,11,20,29,38,47,56-H ep t a oxa [3.3.3.3.3.3.3]p a r a cyclo-
p h a n e, 5. White solid (26 mg, 1.8%). Mp 89-90 °C (lit.12 mp
1
Syn th esis of Op en -Ch a in P h a n e 10. Trimethylsilyl tri-
flate (0.030 mL, 0.17 mmol) and triethylsilane (362 mg, 3.11
mmol) in dry dichloromethane (15 mL) were reacted with 9
(500 mg, 1.33 mmol) and trimethylsilyloxymethylbenzene (480
mg, 2.66 mmol) in dry dichloromethane (15 mL) according to
97-100 °C). H NMR (200 MHz, 0.1 F in CDCl3) δ 7.36 (s, 28
H), 4.54 (s, 28 H). 13C NMR (50 MHz, 0.1 F in CDCl3) δ 137.6,
127.9, 71.7. ESI-MS 858.4 (M + NH4+), 863.4 (M + Na+), 879.4
(M + K+). Anal. Calcd for C56H56O7: C 79.97, H 6.71. Found:
C 79.76, H 6.52.
3660 J . Org. Chem., Vol. 69, No. 11, 2004

Binding of Guests to Flexible Cyclophane Receptor
the previous procedure, performing the addition in 15 min. A
white solid (562 mg) was obtained after workup and washing
with petroleum ether, which was chromatographed (silica gel
60, dichloromethane/diethyl ether 96:4) to give 183 mg (25%
yield) of pure 10 as a white solid. Mp 94-95 °C. 1H NMR (200
MHz, 0.025 M in CDCl3) δ 7.37-7.29 (m, 22 H), 4.56 (s, 16
H). 13C NMR (50 MHz, 0.025 M in CDCl3) δ 138.2, 137.7, 128.4,
127.9, 127.8, 127.6, 72.8, 72.1. Anal. Calcd for C38H38O4: C
81.69, H 6.86. Found: C 81.37, H 6.81.
Cristallografia Strutturale (CRIST) of the University of
Florence. We would like to thank Dr. S. Ciattini for the
structural analysis.
Su p p or tin g In for m a tion Ava ila ble: ORTEP plot of the
structure of 2 with atom numbering and X-ray crystallographic
data in CIF format. This material is available free of charge
via the Internet at http://pubs.acs.org.
Ackn owledgm en t. The X-ray crystallographic struc-
ture determination was performed at the Centro di
J O049899J
J . Org. Chem, Vol. 69, No. 11, 2004 3661