Defect-controlled [formula omitted] diffusion-domain structure of white micas from high-resolution [formula omitted] crystal-mapping in slowly-cooled muscovite

Abstract

Using target-matching techniques combining Ar40/Ar39 crystal-mapping with elemental mapping and high-resolution electron microscopy, this study investigates the Ar40 behavior in very-slowly cooled muscovite from the Harney Peak Granite (HPG, South Dakota, USA). Detailed age mapping along (001) in single crystals from different localities of the HPG documents age gradients in excess of ∼ 300–400 m.y., with conspicuous internal Ar40/Ar39 zoning. This suggests (001) layer-parallel Ar40 transport driven by diffusion, consistent with previous Ar40/Ar39 crystal-mapping studies. The age distribution pattern is complex, however, and defines a mosaic of sub-grain domains with more retentive core zones, broadly ∼ 250–300 μm across, separated by zones of high diffusivity varying in shape and extent. The maximum ages preserved in the core domains are independent of their size but vary linearly with the bulk areal extent of the peripheral (or surrounding) high-diffusivity zones. Spatial Ar40/Ar39 relationships inside each grain point to a mechanism of multipath continuum-diffusion interaction between subdomains across the whole crystal, rather than via discrete non-interracting domains such as in K-feldspars. A close spatial correlation exists between younger ages, Na-depleted (K-enriched) zones, and density of microstructural defects. These defects, identified as lenticular voids and basal partings (< 100 nm–long), developed in response to inward K ↔ Na interdiffusion during late-magmatic stages, in the absence of deformation. Coupled variations in density of microstructural defects and Na–K interchange are inferred to control the bulk diffusion-domain structure of HPG muscovite. Quantitative diffusion modeling of coupled compositional–defect–isotopic variations indicates that Ar40 diffusivity may be enhanced by up to six orders of magnitude in defect-controlled high-diffusivity zones relative to less defective (pristine) domains. On the other hand, empirical diffusivity estimates required to preserve the core ages are commensurate with diffusion estimates independently derived from recent atomistic simulations.