ACS PRF | ACS
All e-Annual Reports
43708-AC10
Crystal Growth, Polar Ordering and Domain Switching in Ferroelastoelectric Inclusion Compounds
Our work on crystal engineering and physical properties of ferroelastic and ferroelectric channel inclusion compounds has focused on both molecular recognition properties during crystal growth and optical and diffraction studies of domain switching events. This grant has supported one postdoctoral student, one research assistant and one graduate student (in part) during the preparation of his Ph.D. dissertation. The following two sections provide a brief overview of the work that we have done during the first part of this grant period.
Crystal growth mechanisms in channel inclusion compounds
As part of our efforts to understand polar ordering of guests in urea inclusion compounds (UICs) containing mixtures of guests, we have conducted extensive studies on the mechanism of crystal growth of these materials. Particularly relevant has been a series of mixed UICs of 2,9-decanedione and 2-decanone in urea; these crystals exhibit polar ordering of guests, but they also exhibit a dramatic and abrupt habit change from {001} plates to needles when the mole fraction of 2,9-decanedione drops below 50%. This solvent-dependent habit change is associated with a change in surface roughening on the {001} faces of these crystals, which, in turn, is correlated with the fraction of host-guest hydrogen bonding sites occupied in these crystals. Through X-ray diffraction studies, we have identified a surface roughening mechanism in which guest screws up the channel (by 60º rotation and1.84Å translation) to the next position along the urea helix. This screw-like twinning creates new ledges on the {001} surface, but it also reduces the size of lateral domains, which become observable as optical anomalies in crystals containing large fractions of 2,9-decanedione. With crystals containing smaller fractions of 2,9-decanedione, the domains are too small to scatter coherently, and the screw-like twinning is manifested as temperature dependent diffuse scattering. With ultrahigh resolution X-ray equipment at the University of Rennes, we have shown that crystals containing less than 40% 2,9-decanone are incommensurate with a slight misfit between host and guest, whereas those containing 40% or more are commensurate structures in which the host and guest repeats along the channel axis coincide. These studies conclusively demonstrate that the surface roughening that gives rise to needles can occur even when the inclusion crystal is commensurate, as long as there is a molecular mechanism that promotes ledge formation.
The screw-like twinning was initially identified through refinement of site occupancies of eight possible screw-related guest sites in the crystal structure of 2-decanone/urea. Now that our X-ray studies have shown this crystal to be incommensurate, we are currently trying to resolve the same kind of disorder in a commensurate system (60:40 2,9-decanedione:2-decanone/urea) in which the optical properties suggest intimate screw-like twinning. Initial refinements suggest that the disordered sites have low occupancies in this system, so other compositions need to be explored. Our discovery of a solid-solid transition in a related system (2,11-dodecanedione/urea) containing a high temperature phase with significant disorder (and presumably surface roughening) and an ordered, low temperature phase should provide an instructive comparison with the 2,9-decanedione:2-decanone/urea system.
Mechanistic studies of ferroelastic domain switching
Much of this work has focused on the role of nanoscopic twins in the ferroelastic domain switching of certain urea inclusion compounds. In crystals grown from solution, these twins are too narrow to diffract coherently, but they are visible through dynamical diffraction contrast in synchrotron white beam X-ray topography experiments. These twins, which are epitaxially matched with the surrounding mother domain in the original crystal, become mismatched with the daughter that is formed in the domain switching process. Curiously, these nanoscopic twins increase in size (sometimes so they are large enough to observe in the microscope) upon release of stresses large enough to give rise to ferroelastic domain switching. Through a series of in situ diffraction experiments (with the crystal under stress), we have formulated a model in which the daughters merge with the nanoscopic twins to minimize the interfacial contact between mismatched twins. It appears, therefore, that the memory effects and rubber-like behavior or pseudoelasticity are controlled by interfaces instead of defects distributed throughout the volume of the crystal. This contradicts the accepted mechanisms in other materials exhibiting "memory effects," including certain shape memory alloys and perovskite ferroelectrics.
By necessity, the plate-like crystals used in the in situ stress studies must be larger than the X-ray beam, so peak intensities cannot easily give exact populations of the different domains that are generated and destroyed during the domain switching process. However, deconvolution of the peak profiles of several reflections (and scaling for scattering power) suggests that the nanoscopic twins, which initially occupy less than 1-2% of the crystal, can occupy as much as 5-10% or more of the crystal after release of stress. Further in situ experiments, including those in which acoustomechanical vibrations are used to anneal metastable sites, should allow us to refine this mechanism even further.
