The goal of my undergraduate honors research project was to quantitatively derive intracrystalline diffusivities for lithium in partially resorbed garnets, using LA-ICPMS and numerical modeling. This was the first time diffusivities have been derived for Li in garnet. Understanding and quantifying diffusivity is vital to proper interpretation of geochemical features retained in minerals during growth. Garnet is a very important mineral in metamorphic rocks because it is largely resistant to internal diffusion and often retains a wealth of information about changes in growing conditions in its chemical zonation, similar to rings of a tree.
The garnet samples I analyzed are from the Makhavinek Lake Pluton Aureole, Labrador, Canada. These garnets are unique because they initially had nearly flat concentration profiles (a homogeneous distribution of the chemical constituents) resulting from formation at granulite facies conditions (high pressures and temperatures) over a long period of time (~60 Ma). Later, a pluton intruded into the pre-existing rock, heating the area and destabilizing the garnet. This resulted in a garnet consuming reaction from the outer edges of the crystal inward, forming a reaction corona. As the reaction progressed, material from the garnet either entered the new reaction phases or diffused into the garnet developing a stranded diffusion profile. These stranded diffusion profiles can be numerically simulated in order to extract diffusivities from them. I was able to numerically simulate Li, Y, and Yb SDPs for nine garnet samples, producing 27 diffusivity values over a range of temperatures. In order to do so, I had to measure the profiles using LA-ICPMS, develop an acid leaching technique to remove a grain boundary contaminant that was distorting the measured SDPs, correct the SDP for spatial averaging effects caused by the laser, and learn to use the Resorb model (cf. Carlson (2012)) to simulate the profiles and extract diffusivities. Through this study I found that Li diffuses much slower in garnet than measured in most other minerals and that Li and Y diffusion are likely related by a coupled diffusion mechanism where Li and Y are substituting into the VIII-fold coordination site. Therefor Li can diffuse through garnet no more rapidly than Y (+HREEs). For a more detailed look into the background, methods, and conclusions, check out my Senior Thesis in the library.
Advisor: Bill Carlson