A graph theory approach to structure solution of network materials from two-dimensional solid-state NMR data

Abstract

An NMR crystallography strategy is presented for solving the structures of materials such as zeolites and related network materials from a combination of the unit cell and space group information derived from a diffraction experiment and a single two-dimensional NMR correlation spectrum that probes nearest-neighbour interactions. By requiring only a single 2D NMR spectrum, this strategy overcomes two limitations of previous approaches. First, the structures of materials having poor signal-to-noise in solid-state NMR experiments can be investigated using this approach since a series of 2D spectra is not required. Secondly, the structures of aluminophosphate materials can potentially be determined from ²⁷Al/³¹P solid-state NMR experiments since this approach does not require the isolated spin pairs which have been important for determining structures of silicate materials by ²⁹Si solid-state NMR. Using concepts from graph theory, the structure solution strategy is described in detail using a hypothetical two-dimensional network structure. A collection of two-dimensional network structures generated by the algorithm under various initial conditions is presented. The algorithm was tested on a series of 27 zeolite framework types found in the International Zeolite Association’s zeolite structure database. Finally, the structure of the zeolite ITQ-4 was solved from powder X-ray diffraction data and a single ²⁹Si double quantum NMR correlation spectrum. The limitations of the strategy are discussed and new directions for this approach are outlined.