Disequilibrium for Ca during growth of pelitic garnet

Compositional zoning in hundreds of almandine-rich garnets in amphibolite by facies micaceous quartzites from the Picuris Range, north-central New Mexico USA, indicates that although Mn, Mg and Fe achieve chemical equilibrium at hand-sample scale during garnet growth, Ca does not. Instead, Ca concentrations at the surface of growing garnets appear to depend strongly on kinetic factors that govern the local chemical environment, yielding disequilibrium for Ca at scales larger than the region immediately surrounding an individual porphyroblast. Detailed zoning profiles were obtained for 371 garnet crystals in a small volume of a single sample of garnetiferous quartzite, and core analyses were made of 97 additional crystals. Each analysis was made on a section that passed precisely through the morphological centre of the crystal, located by means of 3-D imagery from computed X-ray tomography. The data reveal strong correlations between crystal size and concentrations of Mn, Mg and Fe (but not Ca) in garnet cores; a relationship between crystal size and isolation; rigorous cross-correlations among concentrations of Mn, Mg and Fe (but not Ca); and systematic variations in Ca concentrations as a function of crystal size and core composition that are anomalous in comparison to the behaviour of the other divalent cations. We interpret these observations as the result of thermally accelerated diffusion-controlled garnet growth, in circumstances that promoted rapid intergranular diffusion and thus rock-wide equilibration of Mn, Mg and Fe, but that prevented equilibration at similar scale for Ca because of its more sluggish intergranular diffusion. The anomalous behaviour of Ca is made evident in these garnets by the presence of sharp spikes in Ca concentration, which are demonstrably not a consequence of any simultaneous rock-wide event, such as a change in pressure, temperature, or some other intensive parameter. Instead, Ca concentrations probably reflect the local extent of reaction in the immediate vicinity of each porphyroblast. To the degree that such kinetic factors introduce departures from chemical equilibrium for Ca, thermobarometric estimates that involve grossular contents of pelitic garnet will be in error.