Multicomponent diffusion in silicate melts using eigen-component approach

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

This project will investigate diffusion of all major oxides (defined to be more than one weight percent) in dry natural silicate melts using a new approach. In the past, researchers could only successfully and quantitatively study the diffusion of oxides in natural silicate melts with monotonic, or flat, concentration profiles. But what about oxides with non-monotonic profiles? The goal of this project is to develop an easier way to quantify diffusion of all major oxides in dry silicate melts. One application of the new method will be the modeling of the mixing of two magmas. Preliminary results show that such mixing may result in unexpected high concentrations of some metals. Such high concentrations may lead to ore formation. Hence, one of the practical applications is to explain previously unexplained ore formation, and to predict possible locations of ore deposits. Other applications of the new method include more accurate quantification of crystal growth and dissolution rates. This research will also solve a major theoretical problem in high-temperature geochemical kinetics. This project will investigate multicomponent diffusion in dry natural silicate melts using eigen-component approach. Multicomponent diffusion in silicate melts is fundamental to mass transport in igneous processes and is also one of the most difficult experimental and theoretical problems facing geochemical kineticists. Multicomponent diffusion is often encountered in natural silicate melts but not treated due to the level of difficulties in terms of both theory and especially experiments. This team has extracted a diffusion matrix in basaltic melts at three different temperatures, and roughly verified that the diffusion eigenvector matrix in basaltic melts is roughly independent of temperature. They now hypothesize that the diffusion eigenvector matrix is also roughly independent of melt composition in dry melts. They are now further improving the eigenvector matrix and hope to verify that the eigenvector matrix is invariant with composition. In the eigen-component approach, plots of “concentrations” of eigen-components (rather than oxide wt% or mole fractions) versus distance will be made. If their hypothesis is correct, then using the eigenvector matrix, all eigen-component plots would be monotonic, and diffusion eigenvalues (i.e., diffusivity for the eigen-components) will be obtained from fitting the eigen-component profiles. If some eigen-component plots in some experiments show nonmonotonic profiles, within the framework of our hypothesis, it would mean inaccuracy of the eigenvector matrix, and hence the need to improve the eigenvector matrix. To complete the investigation of multicomponent diffusion in a melt, a couple of well-designed experiments at each temperature and pressure will be enough to determine all eigenvalues, and hence the full diffusion matrix. This team's hypothesis and new approach will contribute to the fundamental understanding of diffusion in natural silicate melts. The success of the proposed work will remove one difficult problem from the list of geochemical kineticists and allow future geochemists to realistically predict diffusion-controlled processes, such as magma mixing, mineral growth and dissolution, and post-entrapment interaction between a melt inclusion and its host mineral. They will build and publish an online resource to compute multicomponent diffusion profiles. This would enable interested students and scientists to use multicomponent diffusion in their research of geological problems. This planned online calculator will help facilitate the use of their results by the petrological and geochemical community. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.