14:20 〜 14:40
[SIT17-03] Partial melting and pre-melting in the upper mantle
★招待講演
キーワード：部分溶融、プレメルティング、粒界無秩序化
Melt phase plays an important role in the chemical evolution of the Earth. Therefore, many attempts have been performed to detect the melt phase from seismological observations. Seismic low-velocity and high-attenuation zones, which are frequently observed near the volcanic source regions in the upper mantle, have long been considered as partially molten zones. Also, low viscosity of the asthenosphere has been attributed to the presence of melt. Because melt exists in the relatively small spaces surrounded by the mineral grains, in order to clarify the effects of melt on the seismic wave velocity and viscosity, not only the melt properties but also the properties determining the interaction between solid and melt play important roles. One of such important properties is the dihedral angle determining the extent to which melt wets the mineral grain boundaries. So far, moderate values of dihedral angle have been reported for the melts in the Earth, indicating that melt has only a limited effect on the mechanical properties. In addition, geochemical studies have shown that only a small amount of melt (less than one per cent) can exist in the upper mantle (e.g., McKenzie, 2000, Chem. Geol.; Hirschmann, 2010, PEPI), which is not enough to explain the seismic low velocity regions and weak asthenosphere.
Our experimental studies by using rock analogue samples (Takei et al 2014, JGR; Yamauchi and Takei, 2016, JGR) have shown that a significant reduction of seismic velocity and viscosity starts from considerably below the solidus temperature in the absence of melt, and that the underlying mechanism of this subsolidus weakening can be explained by ``grain boundary disordering’’, which significantly enhances the sliding and matter diffusion along the grain boundaries. Significant increase in grain boundary disorder at near solidus temperatures (sometimes called pre-melting) has been well known in material sciences, but not in rock rheology. This physical process is important because it resolves the above-mentioned difficulty to explain the geophysical and geochemical observations in a consistent manner (Takei, 2017, Ann. Rev.; Yamauchi and Takei, 2020, G-cubed). Recently, significant reduction of viscosity at subsolidus temperatures quite similar to that reported by Yamauch and Takei (2016) for the rock analogue samples was observed in the olivine aggregates (Yabe et al, 2020, JGR), and the possible occurrence of pre-melting in the upper mantle has increased in realness.
Pre-melting, or grain boundary disordering, is driven by interfacial energy, which is usually neglected in the calculation of phase diagram. Tang et al (2006, Phys. Rev. Lett.) developed a thermodynamic model of a binary eutectic system by taking into account the interfacial energy and predicted the occurrence of both pre-melting and partial melting. Based on this model, I will discuss the physical mechanism of pre-melting and its important consequences on the geophysical and geological processes.
Our experimental studies by using rock analogue samples (Takei et al 2014, JGR; Yamauchi and Takei, 2016, JGR) have shown that a significant reduction of seismic velocity and viscosity starts from considerably below the solidus temperature in the absence of melt, and that the underlying mechanism of this subsolidus weakening can be explained by ``grain boundary disordering’’, which significantly enhances the sliding and matter diffusion along the grain boundaries. Significant increase in grain boundary disorder at near solidus temperatures (sometimes called pre-melting) has been well known in material sciences, but not in rock rheology. This physical process is important because it resolves the above-mentioned difficulty to explain the geophysical and geochemical observations in a consistent manner (Takei, 2017, Ann. Rev.; Yamauchi and Takei, 2020, G-cubed). Recently, significant reduction of viscosity at subsolidus temperatures quite similar to that reported by Yamauch and Takei (2016) for the rock analogue samples was observed in the olivine aggregates (Yabe et al, 2020, JGR), and the possible occurrence of pre-melting in the upper mantle has increased in realness.
Pre-melting, or grain boundary disordering, is driven by interfacial energy, which is usually neglected in the calculation of phase diagram. Tang et al (2006, Phys. Rev. Lett.) developed a thermodynamic model of a binary eutectic system by taking into account the interfacial energy and predicted the occurrence of both pre-melting and partial melting. Based on this model, I will discuss the physical mechanism of pre-melting and its important consequences on the geophysical and geological processes.