V33A-4836:
Multicomponent Diffusion in Silicate Melts: SiO2-TiO2-Al2O3-MgO-CaO-Na2O-K2O System

Wednesday, 17 December 2014
Chenghuan Guo, University of Michigan, Earth and Environmental Sciences, Ann Arbor, MI, United States and Youxue Zhang, Univ of Mich- Earth & Env Sci, Ann Arbor, MI, United States
Abstract:
Multicomponent diffusion in silicate melts is a very important process for mass transport in natural systems. It plays a large role in mixing and contamination of magmas, double-diffusive convection of magmas, and growth or dissolution of minerals in magma. The often-observed uphill diffusion profiles in natural systems and experiments require multicomponent diffusion treatment, but that has not been done. Instead, effective binary diffusion approximation has been widely used to treat chemical diffusion of components without uphill diffusion, even though it doesn’t offer a consistent model for diffusion in a multicomponent system. In this report, we follow a long series of efforts in studying multicomponent diffusion in various melt systems (including SiO2-Al2O3-CaO, SiO2-Al2O3-MgO, SiO2-Al2O3-MgO-CaO, SiO2-Al2O3-Na2O-K2O-H2O and a basaltic melt).

Chemical diffusion experiments in a haplobasaltic melt SiO2-TiO2-Al2O3-MgO-CaO-Na2O-K2O (50% SiO2, 1.5% TiO2, 15% Al2O3, 10% MgO, 19% CaO, 3% Na2O, 1.5% K2O) were conducted at 1500°C and 1 GPa. Nine successful experiments were carried out, of which 8 had initial compositions differing in only one independent component by 3 wt% (and compensated by SiO2). The concentration profiles in all 9 experiments were simultaneously fit using the Levenberg-Marquardt algorithm to obtain the 6 × 6 diffusion matrix. All features in the 7 concentration profiles of 9 experiments are well reproduced by the obtained diffusion matrix. More experiments will be carried out to further understand multicomponent diffusion, and we will use the diffusion matrix to simulate diffusion in natural systems.