Element transport along grain boundaries of synthetic silicate bicrystals: the effect of tilt orientations on transport properties (TP1)

点击翻译按钮获取中文摘要

Diffusional mass transport in a synthetic or natural rock matrix may occur via volume or grain boundary diffusion, or a combination of both. A huge amount of experimental data on volume diffusion exists for many elements within the respective mineral structures; data on grain boundary diffusion are lacking. This is particularly true for mass transport within single grain boundaries with well characterised grain boundary structures. Single grain boundaries with constant characteristics at the mm-scale are a prerequisite for such studies. These experimental data are urgently needed. Rim growth experiments under compressive stress performed by members of the FOR 741 group have shown that grain boundary diffusion in polycrystalline spinel depends on misorientation (Keller et al submitted). An increasing density of "faster" grain boundaries produced by reaction under stress may enhance rim growth even if the overall grain boundary density decreases with increasing stress. The relative orientation of crystals defining the grain boundary has an important effect on the transport rates. Rim growth experiments, however, can only record the overall effect. Here, simpler experiments in geologically relevant systems will be performed making use of pre-fabricated grain boundaries with well-defined, varying misorientations. Oriented garnet and olivine bicrystals are produced, sputtered or pulsed-laser coated with various components and subsequently heated at different temperatures and durations. Resulting concentration profiles along/across the specific grain boundaries will be measured by Analytical Transmission Electron Microscopy (ATEM) and Synchrotron X-Ray Fluosrescence (SXRF) methods and compared with volume diffusion profiles from the same experiment. Applying continuum models for diffusional transport, and estimating grain boundary widths from High Resolution TEM (HRTEM) observations, grain boundary diffusion coefficients are derived and directly compared with volume diffusion coefficients. Experiments will be complemented by atomistic simulations using advanced ionic interaction models that are derived from first-principles DFT calculations. The relaxed structural models are applied to interpret the transmission electron micrographs with atomic scale resolution. Furthermore, the calculations will provide interfacial energies needed to formulate larger scale continuum models derived from the experimentally determined profiles. Eventually, molecular dynamics simulations may reveal pathways for fast diffusion.