236
H. Ohtsu et al.
12.1.7 Kinetic Assembly Using a Labile CuI Unit
The success of S 3 encapsulation prompted us to design materials that contain intrinsic
interactivity within their pores using the principles of kinetic assembly. To produce
coordination networks under kinetic control, we employed labile metal connectors.
If a metal can readily exchange ligands under the reaction conditions, usually in
solution, it may generate several species during network formation because of the
presence of several intermediate metal complexes. In particular, labile multinuclear
metal clusters could give rise to multiple species existing simultaneously in solution.
However, in most metal ions the intermediate state geometries are predictable, which
limits the number of easily accessible structures for the network metal connectors.
Therefore, to demonstrate the kinetic control for network formation, we decided to
utilize labile metal connectors instead of commonly used metal ions [41, 76]. These
were combined using a rigid, highly symmetric ligand to produce structurally robust
networks. For this study, [Cu 4 I 4 (PPh 3 ) 4 ] was selected as a suitable metal precursor,
because its central Cu 4 I 4 cubane core is kinetically labile in solution and can be
readily converted into other copper iodide clusters [77, 78]. For the ligand, tetra-4(4-pyridyl)phenylmethane (tppm) was used, which possesses all the required characteristics, namely structural rigidity, thermal stability, and tetrahedral symmetry. Upon
heating a suspension of [Cu 4 I 4 (PPh 3 ) 4 ] and tppm in DMSO in air at 453 K for 30 min,
a homogenous colorless solution was obtained. Depending on the cooling rate, two
kinds of network crystals could be obtained. Rapid cooling (ca. 20 K min
−1 , kinetic
assembly) exclusively produced yellow needles with the formula [(CuI) 2 (tppm)] in
99% yield. In contrast, slow cooling (ca. 3 K min
−1 , thermodynamic assembly)
resulted in the formation orange prisms with the formula [(Cu 2 I 2 )(tppm)] in 95%
yield. Although these two networks had the exact same chemical composition, their
connectivity and topology were markedly different. Therefore, these materials could
be considered as structural isomers (Fig. 12.12).
The single-crystal structure analysis of the kinetic isomer revealed that CuI helical
chains bridged by tppm ligands formed a non-interpenetrating porous network along
the c-axis, where each Cu(I) ion is coordinated by two nitrogen atoms of tppm ligand
and two bridging iodide groups. The helical chain network has 1D channels with
pore windows of 5.8 × 5.5 Å and a 35% void space (without solvent) in the unit cell
volume. The most important feature of this structure is that the bridging iodide in
the connecting cluster faces into the 1D channel, and therefore can act as interactive
pore site (Fig. 12.12a, c, e).
The crystal structure analysis of the thermodynamic isomer on the other hand,
revealed that it has a quadruply interpenetrating network consisting of Cu 2 I 2 dimer
units and tppm ligands. The structure is a 4,4-connected grid with the PtS topology
if the Cu 2 I 2 dimer units are considered as square-planar sites and the central carbon
atom of tppm as a tetrahedral node. In addition, this network contains 1D channels
with pore windows of 4.0 × 3.9 Å and a 22% void space in the unit cell volume. The
iodide groups in the Cu 2 I 2 dimer are obscured by tppm ligands from the interpenetrating nets, which is different from the kinetic structure. As a result, the channels
Précédent

- 236/532

Suivant