Full- or half-encapsulation of sulfate anion by a tris(3-pyridylurea) receptor: effect of the secondary coordination sphere.

Self-assembly of the [Fe(DABP)(3)]SO(4) (DABP = 5,5'-diamino-2,2'-bipyridine) or [Fe(bipy)(3)]SO(4) (bipy = 2,2'-bipyridine) complex with a tripodal tris(3-pyridylurea) ligand (L) results in a layered structure that includes a sulfate anion in the cleft of one L molecule. The two compounds, [Fe(DABP)(3)][SO(4) L] x 10 H(2)O (2) and [Fe(bipy)(3)][SO(4) L] x 9 H(2)O (3), show very similar sheets formed by the anionic units [SO(4) L](2-) and cationic building blocks ([Fe(DABP)(3)](2+) or [Fe(bipy)(3)](2+)). However, there are different water clusters that link the adjacent layers in the two products, that is, water parallelograms and quasi "water cubes" in 2 versus single water molecules, water dimers, and hexamers in 3. The half-encapsulation of sulfate by a single L molecule contrasts with the previously reported full-encapsulation of the sulfate ion by two L molecules in [M(H(2)O)(6)][SO(4) L(2)] (1). This different anion encapsulation is traced to the hydrogen-acceptor properties of the pyridyl groups of L together with the hydrogen-bonding properties of the cation secondary coordination sphere for a solid-state packing optimization. In 1 the direct hydrogen bonding from the secondary coordination sphere of octahedral [M(H(2)O)(6)](2+) to L-pyridyl helps in the formation of an octahedral cation-anion coordination in the NaCl-type structure. In 2 and 3, crystal water instead of the cations has to satisfy the hydrogen-accepting demands of L. Consequently, a non-spherical and only partly water-surrounded half-encapsulated [SO(4) L](2-) anion allows for a closer approach of the [Fe(DABP)(3)](2+) or [Fe(bipy)(3)](2+) cations than the [SO(4) L(2)](2-) anion. Then, the similar cation and anion size in 2 and 3 with the Coulomb attraction confined to a two-dimensional plane leads to the formation of a hexagonal BN (or graphite) lattice. Competition experiments with different anions for compound 2 reveal that SO(4)(2-) can be selectively crystallized against NO(3)(-), OAc(-), or ClO(4)(-).