Why Are Some Crystals Straight?

Abstract

Copyright © 2020 American Chemical Society. More than one-quarter of molecular crystals that are able to be melted can be made to grow in the form of twisted lamellae or fibers. The mechanisms leading to such unusual crystal morphologies lacking long-range translational symmetry on the mesoscale are poorly understood. Benzil (C6H5C(O)-C(O)C6H5) is one such crystal. Here, we calculate the morphology of rod-shaped benzil nanocrystals and other related structures. The ground states of these ensembles were twisted by 0.05-0.75°/Å for rods with cross sections of 50-10 nm2, respectively; the degree of twisting decreased inversely proportional to the crystal cross-sectional area. In the aggregate, our computational studies, combined with earlier observations by light microscopy, suggest that in some cases very small crystals acquire 3D translational periodicity only after reaching a certain size. Twisting is accompanied by conformational changes of molecules on the {101¯ 0} surfaces of the six-sided rods, although it is not easily answered from our data whether such changes are causes of the twisting, consequences of surface stress where symmetry is broken, or consequences of intrinsic dissymmetry when two or more geometric tendencies are in conflict. Nevertheless, it has become clear that, in some cases, the development of a crystal with a lattice having long-range translational symmetry is not foretold in the thermodynamics of aggregates of molecules. Rather, a lattice is sometimes a device for allowing a growing crystal to take advantage of the thermodynamic driving force of growth, the best compromise for a large number of molecules, which on a smaller scale would be dissymmetric (have a point symmetry only). The relationship between these calculations and the ubiquity of crystal twisting on the mesoscale are discussed.