Characterizing the Turbostratic Disorder in COF-5 with NMR Crystallography

Abstract

Since its initial synthesis in 2005, COF-5 has been known to have intrinsic disorder in the placement of the 2D layers relative to one another (i.e. turbostratic disorder). Prior studies of have demonstrated that the eclipsed layering found in the space group originally assigned to COF-5 (<i>P</i>6<i>/mmm</i>) is inconsistent with energy considerations. Herein it is demonstrated that eclipsed layers are also inconsistent with<sup> 13</sup>C solid-state NMR data. Crystal structure predictions are made in five alternative space groups and good agreement is obtained in <i>P</i>21<i>/m</i>, <i>Cmcm</i>, and <i>C</i>2<i>/m</i>. We posit that all three space groups are present within the stacked 2D layers and show that this conclusion is consistent with evidence from <sup>13</sup>C solid-state NMR linewidths and chemical shifts, powder x-ray diffraction data and energy considerations. An alternative explanation involving a mixture of multiple pure phases is rejected because the observed NMR spectra don’t exhibit the characteristic features of such mixed phase materials.

Computational Protocol for the Spectral Assignment of NMR Resonances in Covalent Organic Frameworks

“…In that case, some carbon classes that by default in the g S model are treated to be equivalent are better kept as separate classes. In a case such as COF-5, it is possible that considering the different existing symmetries of the (different regions of the) framework is preferable over defining the carbon classes based on the symmetry of the building block.…”