The energy of grain boundaries and interfaces in metamorphic rocks: comparison of natural materials with experimental charges

The relative magnitude of the energies of grain boundaries and heterophase interfaces in texturally equilibrated materials can be measured using the dihedral angle method, which is based on the control of the geometry of three-grain junctions involving two different phases by the relationship γ_AA = 2γ_AB cos (Θ/2), where Θ is the dihedral angle, γ_AA is the energy of the AA grain boundary and γ_AB is that of the heterophase interface. We have compiled a dataset of dihedral angles between 35 mineral pairs found in well-equilibrated granulite-facies rocks. Such a dataset permits the ranking of grain boundary energies, and we find that this ranking correlates with that of the crystalloblastic series. More significantly, we also find that for almost all mineral pairs, both dihedral angles (i.e. those at both AAB and BBA junctions) are <120˚, meaning that the energy of the heterophase interface is lower than that of either grain boundary. An analogous situation occurs in nearly all binary systems of metals. One notable exception is the Zn-Sn pair for which one angle is <120˚ and the other is >120˚, meaning that the energy of the heterophase interface is intermediate between that of the two grain boundaries. This intermediate situation is also observed in experimental mineral charges created by hot-pressing powders, and matches the predictions of a simple model of interfaces involving randomly oriented crystal lattices. That neither metamorphic rocks nor most metals fit this theoretical framework must be a consequence of the creation and preservation of low-energy heterophase interfaces during solidification, reaction, deformation and grain growth. Preliminary work shows that epitaxy governs the location and orientation of nucleation in most metamorphic reactions, resulting in low-energy heterophase interfaces. One corollary of our results is that textural maturation of metamorphic rocks generally results in phase mixing rather than separation and layer formation. Furthermore, that dislocations can cross epitaxial heterophase interfaces, and that the rate of migration of interfaces is dependent on their energy, means that an understanding of rock rheology and microstructural development derived from experimental studies using initially hot-pressed powders may not be directly applicable to natural systems.