Polar crystals are characterized by an axis that has a nonzero dipole due to the nature of the molecular packing. For these crystals, the growth rates of the faces delineating the polar axis are generally expected to be equal. Recent experiments, however, have revealed a few exceptions where the growth of these faces from the vapor phase is asymmetric, a notable case being crystals of resorcinol. Here, we present the mechanics of resorcinol crystal growth from the melt for the hemihedral faces (011) and (01̅ 1̅ ) delineating the polar axis as revealed by molecular dynamics simulations. The simulations reveal asymmetric growth consistent with experiment. The asymmetry is attributed to the slow-growing (011) face being less able to direct the correct alignment of the oncoming molecules and the presence of an alternate resorcinol conformation that readily incorporates into the lattice at this surface, serving to poison and retard subsequent growth. Putting the issue of the rogue conformation aside, the identified factors that influence molecular recognition are considered to be applicable to other polar crystals, which suggest asymmetric growth along the polar axis to be a common feature. ■ SIGNIFICANCE STATEMENTPolar crystals form the basis for important functional materials. For these crystals, we do not understand the nature and mechanism of crystal growth along the polar axis, which is considered to be identical at both ends. Recent experiments have revealed a few exceptions where the growth of these faces from the vapor phase is asymmetric. We have carried out molecular dynamics simulations on one of these exceptional cases, resorcinol. The simulations reveal the molecular processes involved and explain the asymmetric growth. The molecular recognition processes at play and the associated dynamics of molecule alignment during crystal growth suggest asymmetric growth along the polar axis to be a common feature.

No abstract

von Liebig discovered the first case of molecular polymorphism when crystallizing benzamide in 1832. 1 So far only microcrystalline mixtures of metastable phases with the thermodynamically stable form I were available. Optimizing the conditions for crystal growth now gave access to single crystals of the metastable form III which allowed for a detailed comparison of form I and form III by Raman and IR spectroscopy and thermoanalysis. Surprisingly, in the course of detailed DSC measurements for form I an endothermic peak prior to melting was observed. As proven by temperature dependent powder diffraction, thermal microscopy, MDsimulations, and solid-state NMR this pre-melting thermosignal is not related to an enantiotropic phase transition of form I into form III with a transition temperature close to the melting point.

The identification of two step nucleation mechanisms considerably extended our understanding of crystal nucleation. Here, we report an analogous observation of a two-step mechanism but in 2-D for deposition of molecules to a growing crystal face. Using molecular dynamics simulations connected with the Kawska-Zahn approach, α-resorcinol precipitation from the vapor is treated at the low driving force regime. Growth at the faster growing (01̅1̅) face reveals the deposition of molecules to form a disordered liquid-like layer. Strikingly, this apparently divergent (nonepitaxial) molecular arrangement does not represent self-poisoning which would lower the growth rate of the (01̅1̅) face. On the contrary, more favorable attachment energy along with a disorder–order transition, akin to a two-step nucleation observed in 3-D systems, leads to growth rates that are about 20 times faster than the more standard mode association of molecules at the (011) face where the molecules readily align according to the crystal lattice.

No abstract