7
72
D. Menozzi et al.
typical columnar disposition along the a-axis of the unit
2
cell (see Figure 1, right). The distance between I and the
entropy driven. As shown in Table 1 (entries III and IV),
the thermodynamic parameters of Tiiii[C H , H, Ph]
3
7
˚
positive nitrogen of the guest is 7.679(3) A. All the above
complexation in DCE are not influenced by the
counterions, confirming the results reported for the
parent Tiiii[C H , CH , Ph].
values are in good agreement with similar complexes
reported in the literature (2, 14).
3
7
3
The first set of ITC measurements was made to assess
the solvent effect on the complexation. For this purpose,
dichloroethane (DCE) and methanol were chosen. The
former does not compete for the cavity due to its size and
its low propensity to form hydrogen bonds with the PvO
units of the cavitand. Vice versa, in the case of methanol,
crystal structures showed its inclusion in the cavity of Tiiii
receptors (10), driven by the synergistic action of
In order to probe the contribution of cation–dipole
interactions on the overall binding, tetrathiophosphonate
cavitand TSiiii[C H , H, Ph] was prepared (Scheme 1). It
3
7
is structurally identical to the corresponding tetrapho-
sphonate Tiiii[C H , H, Ph], except for the presence of
3
7
four inward-facing PvS instead of four PvO bridging
units. The PvS group is more polarisable than the PvO
counterpart, but it has a much smaller dipole moment (16).
Its larger size does not preclude the insertion of methyl
groups into the cavity (17); therefore, the introduction of
four PvS bridges has little influence on CH–p
interactions. As a corollary, this substitution should
decrease substantially the cation–dipole interactions
with the charged methylpyridinium guests. The ITC
results were beyond our expectations: the affinity of the
cavitand for guests 2 and 3 was completely shut off (Table
1, entries V and VI). Therefore, the cation–dipole
interactions are indispensable for binding methylpyridi-
nium guests. Without PvO, the complexation is
negligible, well below the detection limit of ITC.
H-bonding with the phosphonates and CH –p interactions
3
with the cavity. The different affinity of the two solvents
for the cavity is clearly reflected in the complexation of
guests 2 and 3 by Tiiii[C H , CH , Ph]. A decrease of the
3
7
3
association constant of two orders of magnitude was
determined for both guests by moving from DCE to MeOH
(
Table 1, entries I and II). In both cases, the reduced
affinity in MeOH is mainly enthalpic in origin, with the
DH reduction attributable to the solvent competition for
the cavity and the electrostatic stabilisation of the cationic
guest by the more polar solvent. Interestingly, the
counterion does not influence complexation, leading to
comparable values of DG and K . Such unexpected
ass
silence in impact may be attributed to the preferential
2
positioning of both iodide and PF within the alkyl chains,
6
3
.
Conclusions
as was proven in NMR studies in chloroform (14). Yet, the
absence of evidence of a role for the counterions in the
binding process can also arise from an adventitious
cancellation of enthalpic and entropic effects of com-
plexation at the less structured, well-solvated and solvent-
like site of the host compound.
The selective complexation of methylpyridinium salts by
tetraphosphonate cavitands relies on two interactions,
namely cation–dipole and CH –p, which operate
3
synergically. The crystal structure of the reported
complex clearly evidences this synergy in the solid
state. Using ITC, the thermodynamic profile in the
formation of several host–guest complexes was deter-
mined. They all are driven both by enthalpy and entropy
(18, 19), reflecting the importance and substantial
participation of solvent interactions for the binding
process. The relative weights of the two direct mutual
interaction modes have been dissected by changing two
structural parameters on the cavitands. The dominant role
of cation–dipole interactions was revealed by substituting
the PvO bridges with the PvS ones, which completely
suppressed complexation. The relevance of CH–p
interactions was highlighted by the removal of the four
methyl groups in the cavity apical positions, which led to
the decrease of one order of magnitude in the Kass of both
2 and 3. On the guest side, the counterion does not play a
significant role in complexation, which can be due to an
adventitious cancellation of effects, but may also emerge
from the basic lack of binding. The solvent instead affects
the binding significantly, particularly when the solvent
competes with the guest for the cavity space, like in the
case of methanol.
Next, the role of the cavity structure on complexation
was investigated. The presence of four methyl groups in
the apical positions of Tiiii[C H , CH , Ph] deepens the
3
7
3
cavity and enhances its p-basic character. Their removal
should reduce the complexation efficiency of the receptor.
Experimentally, a decrease of the Kass of one order of
magnitude was observed for the titration of methyl-
depleted cavitand Tiiii[C H , H, Ph] with guests 2 and 3
3
7
in DCE, as compared to the original ones (Table 1,
compare entry I with entry III, and entry II with entry IV).
A large enthalpic drop was determined, only partially
compensated by an entropic gain. This is ascribed to the
reduced contribution of CH –p interactions (15) to the
3
overall binding and to the enhanced solvation of the more
open cavity of the demethylated host Tiiii[C H , H, Ph].
3
7
The latter origin is also suggested by the entropic
outcome which testifies to more dramatic cavity
desolvation and guest mobility in the case of Tiiii[C3-
H , H, Ph] upon binding, while the guest desolvation is
7
comparable in the two cases. It is worth noting that the
methylpyridinium binding remains both enthalpy and