Establishing Accurate High-Resolution Crystal Structures in the Absence of Diffraction Data and Single Crystals—An NMR Approach

Predicting accurate crystal structures from theoretical consideration has proven to be remarkably challenging. Although significant progress has been made and numerous approaches have now been investigated, selection of the correct structure as the first choice in blind studies is still rarely achieved. Here a process is described that consistently identifies the correct structure from the myriad candidates created from typical crystal structure prediction software. This approach relies on 13 C solid-state NMR data and a secondary refinement process that includes lattice fields. Four structures are considered, and in all cases the correct structure is selected as the first choice and the only statistically feasible candidate. Data from 13 C chemical shift tensor principal values are found to provide better selectivity, but 13 C isotropic shifts also consistently identify the correct structure. This process involves only experimental NMR data and computer-generated structures, yet the structures created appear to rival the accuracy of structures derived from single crystal diffraction methods including single crystal neutron diffraction.