Evidence for stable grain boundary melt films in experimentally deformed olivine-orthopyroxene rocks

The microstructure of olivine-olivine grain boundaries has been studied in experimentally deformed (1200-1227 degrees C, 300 MPa) partially molten olivine and olivine-orthopyroxene rocks. In-situ melting produced similar to 1 vol% melt in all samples studied. Grain boundary analyses were carried out using a number of trammission electron microscopy techniques. The grain boundary chemistry in undeformed olivine-orthopyroxene starting material showed evidence for the presence of an intergranular phase along some, but not all, of the olivine-olivine boundaries. In the deformed samples, ultrathin Si-rich, Al- and Ca-bearing amorphous films have been observed along all investigated olivine-olivine grain boundaries. The chemistry of the grain boundaries, which is considered to be indicative for the presence of a thin film, was measured with energy-dispersive X-ray spectroscopy (EDX) and energy-filtering imaging. The amorphous nature of the films was confirmed with diffuse dark field imaging, Fresnel fringe imaging, and high-resolution electron microscopy. The films range in thickness from 0.6 to 3.0 nm, and EDX analyses show that the presence of Al and Ca is restricted to this ultrathin film along the grain boundaries. Because thin melt films have been observed in all the samples, they are thought to be stable features of the melt microstructure in deformed partially molten rocks. The transition from the occasional presence of films in the undeformed starting material to the general occurrence of the films in deformed materials suggests that deformation promotes the formation and distribution of the films. Alternatively, hot-pressing may be too short for films to develop along all grain boundaries. A difference in creep strength between the studied samples could not be attributed to grain boundary melt films, as these have been found in all deformed samples. However, a weakening effect of grain boundary melt films on olivine rheology could not be ruled out due to the lack of confirmed melt-film free experiments.