Chemical diffusivity measurements have been made on anhydrous metaluminous diffusion couples of dacite and rhyolite at 1 atm, 1200°–1400° C, and 10 kbar, 1300°–1600° C, and on anhydrous peraluminous and peralkaline dacite-rhyolite diffusion couples at 10 kbar, 1300°–1600° C. Chemical diffusivities for Si, Al, Fe, Mg, and Ca were measured in all experiments on the metaluminous diffusion couples using Boltzmann-Matano analysis, and Si diffusivities were measured on the other diffusion couples. Two 10 kbar metaluminous experiments were analyzed with the X-ray microprobe and diffusivities of Sr, Y, Zr and Nb were measured. Si diffusivity displays a weak negative correlation with SiO2 content over the range of 65%–75% SiO2. At a given SiO2 content chemical diffusivities of all non-alkali elements are usually within less than an order of magnitude of Si chemical diffusivity and are controlled by partitioning along the diffusion profile so as to maintain local equilibrium at each point along the profile. Alkali chemical diffusivities were not measured but can be estimated from the experiments to be orders of magnitude higher than non-alkali chemical diffusivities. Data were fit to Arrhenius equations for diffusivities measured at 65, 70 and 75% SiO2. At 1 atm the Arrhenius equation for non-alkalies at 70% SiO2 in the metaluminous system is: $$D_T = 72.5\exp \left( { - 82.5 kcal/RT} \right)$$ an at 10 kbar: $$D_T = 0.0673\exp \left( { - 56.5 kcal/RT} \right)$$ whereDT is the diffusivity in cm2/s,R is in calories, andT is in Kelvin. At 65 and 75% SiO2 the pre-exponential factors and activation energies are similar to the values determined at 70% SiO2. Results on the metaluminous system demonstrate that the effect of increasing pressure is to increase the diffusivity at constant temperature, by about a factor of 4 at 1300° C, less at higher temperatures. Ten kbar activation energies and pre-exponential factors for peraluminous and peralkaline systems are slightly smaller than for the metaluminous system and reflect the slightly higher diffusivities in the peraluminous and peralkaline systems consistent with their lower calculated viscosities when compared to the metaluminous system. 1-atm diffusivities can be calculated from melt viscosities using the Eyring equation to within a factor of 5, except for 75% SiO2 diffusivities which consistently display calculated diffusivities approximately an order of magnitude below measured diffusivities. Using fundamental equations of transition state theory the 1-atm chemical diffusivities of non-alkalies, and alkalies too, can be calculated from thermodynamic data and melt structure models. There are, however, discrepancies in the calculated and measured activation energies and pre-exponential factors. Application of diffusivity measurements to magma chamber processes demonstrates that diffusion is not an effective process for compositional modification and can only begin to have a significant effect on melt compositions if the dacitic and rhyolitic melts are convecting separately and separated by a thin, static zone where diffusive transport is occurring; even in this case diffusion is likely to modify alkali concentrations only, and perhaps isotopic ratios in small magma chambers, or chambers with large aspect ratios (width/height). If the dacitic melt forms enclaves which are mixed into the rhyolitic melt, then diffusion coupled with the physical mixing of enclaves has the potential to rapidly affect alkali and isotopic ratios of the rhyolite melt and dacitic enclaves. Non-alkali concentrations in both dacite enclaves and rhyolite are, however, unlikely to be significantly affected. Because of the ineffectiveness of diffusion, once a magma chamber becomes zoned in major and trace elments it will remain zoned, with the exception of alkalies and possibly isotopic ratios, unless physical mixing between the different compositions occurs.
Read full abstract
Jul 4, 2013
Willie Nicol + 1
Open Access
