Intergranular diffusion rates from the analysis of garnet surfaces: Implications for metamorphic equilibration

Novel approaches to garnet analysis have been used to assess rates of intergranular diffusion between different matrix phases and garnet porphyroblasts in a regionally metamorphosed staurolite‐mica‐schist from the Barrovian‐type area in Scotland. X‐ray maps and chemical traverses of planar porphyroblast surfaces reveal chemical heterogeneity of the garnet grain boundary linked to the nature of the adjacent matrix phase. The garnet preserves evidence of low temperature retrograde exchange with matrix minerals and diffusion profiles documenting cation movement along the garnet boundaries. Garnet–quartz and garnet–plagioclase boundaries preserve evidence of sluggish Mg, Mn and Fe diffusion at comparable rates to volume diffusion in garnet, whereas diffusion along garnet–biotite interfaces is much more effective. Evidence of particularly slow Al transport, probably coupled to Fe³⁺ exchange, is locally preserved on garnet surfaces adjacent to Fe‐oxide phases. The Ca distribution on the garnet surface shows the most complex behaviour, with long‐wavelength heterogeneities apparently unrelated to the matrix grain boundaries. This implies that the Ca content of garnet is controlled by local availability and is thought likely to reflect disequilibrium established during garnet growth. Geochemical anomalies on the garnet surfaces are also linked to the location of triple junctions between the porphyroblasts and the matrix phases, and imply enhanced transport along these channels. The slow rates of intergranular diffusion and the characteristics of different boundary types may explain many features associated with the prograde growth of garnet porphyroblasts. Thus, minerals such as quartz, Fe‐oxides and plagioclase whose boundaries with garnet are characterized by slow intergranular diffusion rates appear to be preferentially trapped as inclusions within porphyroblasts. As such grain boundary diffusion rates may be a significant kinetic impediment to metamorphic equilibrium and garnet may struggle to maintain chemical and textural equilibrium during growth in pelites.     

Diffusion models for corona formation in metagabbros from the Western Grenville Province,Canada

Metagabbro bodies in SW Grenville Province display a variety of disequilibrium corona textures between spinel-clouded plagioclase and primary olivine or opaque oxide. Textural evidence favours a single-stage, subsolidus origin for the olivine coronas and diffusive mass transfer is believed to have been the rate-controlling process. Irreversible thermodynamics have been used to model two different garnet symplectite-bearing corona sequences in terms of steady state diffusion. In the models the flux of each component is related to the chemical potential gradients of all diffusing species by the Onsager or L-coefficients for diffusion. These coefficients are analogous to experimentally determined diffusion coefficients (d), but relate the flux of components to chemical potential rather than concentration gradients.The major constraint on the relative values of Onsager coefficients comes from the observed mole fraction, X, of garnet in the symplectites; in (amph-gt) symplectites X _Gt ^Sym 0.80, compared with 0.75 in (cpx-gt) symplectites. Several models using simple oxide components, and two different modifications of the reactant plagioclase composition, give the following qualitative results: the very low mobility of aluminium appears to control the rate of corona formation. Mg and Fe have similar mobility, and Mg can be up to 6–8 times more mobile than sodium. Determination of calcium mobility is problematical because of a proposed interaction with cross-coefficient terms reflecting uphill Ca-diffusion, i.e., calcium diffusing up its own chemical potential gradient. If these terms are not introduced, it is difficult to generate the required proportions of garnet in the symplectite. However, at moderate values of the cross-coefficient ratios, Mg can be up to 4–6 times more mobile than calcium (L _MgMg/L_CaCa<4–6) and="" calcium="" must="" be="" 3–4="" times="" more="" mobile="" than="" aluminium="">L _CaCa/L_AlAl>3).     

Measurement and calculation of OH reactivity at a United Kingdom coastal site

Measurements of OH reactivity were made at the Weybourne Atmospheric Observatory on the North Norfolk coast, UK in May 2004. A wide range of supporting species was also measured concurrently as part of the TORCH-2 field campaign, allowing a detailed study of the OH oxidation chemistry to be carried out. Measurements were made in a variety of air masses, with the 3 most prevalent being air from the Atlantic that arrived at the site from over mainland UK in a South Westerly direction, and much cleaner Northerly air that originated over the far North Sea or Arctic, passed over the North Sea and arrived at the site from a North/North Easterly direction. Direct OH reactivity measurements were made on 6 days during the campaign and with influence of 2 of the 3 air masses prevalent during the study period. The average, minimum and maximum measured OH reactivity are: 4.9, 1.3 and 9.7 respectively. The measured OH reactivity was compared to key OH sinks such as NO₂ and CO and a general positive correlation was observed. OH reactivity (k′) was then calculated using the full range of OH sinks species that were measured (including >30 NMHCs) and their pseudo first order rate constants for reaction with OH. For much of the measurement period there is a significant difference between the measured and calculated k′, with an average value of k′_meas- k′_calc?=?1.9 s^-1, indicative of unmeasured OH sinks. A zero-dimensional box model containing a subset of the Master Chemical Mechanism was used to calculate the OH reactivity more accurately. The simultaneously measured trace species were used as inputs to the model and their oxidative degradation was described by a chemical mechanism containing ~5,000 species. The extra OH sinks species produced by the model, resulted in an improvement in the agreement between k′_meas and k′_calc, however the averaged missing OH reactivity across the entire measurement period remained at 1.4 s^-1. Speculation is made as to the source of this missing reactivity, including reference to studies showing that a potentially large number of high molecular weight aromatic species could be unmeasured by standard instrumentation.