A crystal design strategy is described that produces a series of solid-state molecular host frameworks with prescribed lattice metrics and polar crystallographic symmetries. This represents a significant advance in crystal engineering, which is typically limited to manipulation of only gross structural features. The host frameworks, constructed by connecting flexible hydrogen-bonded sheets with banana-shaped pillars, sustain one-dimensional channels that are occupied by guest molecules during crystallization. The polar host frameworks enforce the alignment of these guests into polar arrays, with properly chosen guests affording inclusion compounds that exhibit second harmonic generation because of this alignment. This protocol exemplifies a principal goal of modern organic solid-state chemistry: the precise control of crystal symmetry and structure for the attainment of a specific bulk property.

The propensity of hydrogen-bonded guanidinium (G) organodisulfonates (S) to form crystalline inclusion compounds has been investigated in the context of separating isomeric mixtures of xylenes and dimethylnaphthalenes via selective inclusion. Pairwise competition experiments, in which inclusion compounds are grown from solutions containing an isomeric mixture of guests, map the inclusion selectivity of a particular host as a function of guest content in solution. Whereas the G 2 [4,4′-biphenyldisulfonate] host is minimally selective with respect to inclusion of o-, m-, or p-xylene, the homologous G 2 [2,6-naphthalenedisulfonate] is highly selective toward the inclusion of p-xylene, by a factor of 36 and 160 versus o-xylene and m-xylene, respectively. Similarly, the hosts of the homologous seriesand G 2 [4,4′azobenzenedisulfonate] display different selectivity for the 10 isomers of dimethylnaphthalene. The details of the selectivity behavior are highly dependent on the molecular structure of the GS host and the solid-state structures of the corresponding inclusion compounds. Single crystal structure determinations reveal that isomer selectivity is most pronounced when the structures of corresponding inclusion compounds are significantly different, i.e., when the isomeric guests template different architectural isomers of the host. Furthermore, selectivity appears to be a consequence of size and shape compatibility between the host and guest. The observation of selective inclusion demonstrates the feasibility of a crystallization-based separation process based on these host compounds.

Crystalline clathrates formed from two-dimensional guanidinium sulfonate hydrogen-bonded networks connected by 4,4‘-biphenyldisulfonate “pillars” in the third dimension exhibit a “brick-like” molecular framework that is a predictable architectural isomer of a previously observed bilayer architecture based on the same pillars. The amount of void space in the brick framework is nominally twice that of the bilayer form, with the framework occupying only 30% of the total volume. The formation of the brick architecture can be attributed to steric templating by the included molecular guests and host−guest interactions that favor assembly of this framework over its bilayer counterpart. The brick framework conforms to the different steric demands and occupancies of various aromatic guests (1,4-dibromobenzene, 1-nitronaphthalene, nitrobenzene, and 1,4-divinylbenzene) by puckering of the flexible, yet resilient, hydrogen-bonded network and by rotation of the pillars about their long axes, the latter also governing the width of the pores in the framework. These observations demonstrate that cystal engineering, and the ability to direct architectural isomerism in porous molecular lattices by the appropriate choice of molecular guest, is simplified by the use of robust 2-D networks.

No abstract