E-coherent crystalline interfaces: Coherency enhanced by discohesion arrays

Coherent crystalline interfaces form when a pair of joined crystals share lattice sites. Such interfaces are ubiquitous in materials, minerals, and compounds, with examples including grain boundaries in polycrystals and phase boundaries in multiphase systems. Existing methodologies such as the topological model [R. C. Pond, X. Ma, Y. W. Chai, and J. P. Hirth, Topological modelling of martensitic transformations, in Dislocations in Solids Vol. 13, edited by F. R. N. Nabarro and J. P. Hirth (Elsevier, Amsterdam, 2007), Chap. 74, pp. 225--261] provide a framework for understanding the nature of coherency between two crystals and the line defect content within an interface. However, these methods only consider states of coherency achieved via affine transformations. Here we show that in some interfaces, local relaxations in the form of nonaffine transformations lead to the introduction of additional coincidence sites within the interface; we term this class of interfaces e-coherent. These nonaffine relaxations are topologically equivalent to inserting disconnection (or disclination) dipoles or loops into the interface. Unlike traditional interfacial line defects, the defects associated with e-coherency cannot have long-range stress fields, and their motion alters the state of coherency between the crystals. Given these unique properties, we differentiate them from other defects by referring to them as discohesions. Through atomistic simulations and transmission electron microscopy, we show that the energetics and kinetics of e-coherent interfaces are strongly affected by the discohesion content in the interface, leading to fundamentally different behavior compared with non-e-coherent interfaces. We demonstrate that e-coherency occurs in grain, twin, and phase boundaries and that a given interface can have multiple possible e-coherent states. These results suggest that e-coherency is likely to be pervasive in crystalline solids.