The goal of this study was to ascertain the nature of the Ar release mechanism(s) for a single, well characterized muscovite heated incrementally in vacuo: (1) at temperatures and time intervals typical of 40Ar/ 39Ar step heating experiments and (2) isothermally at 500°C and 700°C for time intervals much longer than typical 40Ar/ 39Ar experiments. The rate of delamination was monitored by SEM, and the rate of dehydroxylation was determined by XRD analyses. The rate of in vacuo Ar loss was compared to that for muscovite heated at 700°C and 2 kbar H 2O pressure. For all experiments, the rate of Ar loss was found to be consistent with the rates of dehydroxylation and delamination. At 500°C, Ar loss behavior for a 3000 μm diameter muscovite grain was approximated by volume diffusion, but the Ar release pattern became increasingly non-Fickian with decreasing grain size. We propose that the cause of the non-Fickian behavior is delamination. The portions of the isothermal runs exhibiting diffusion-like profiles also yielded discordant 40Ar/ 39Ar ages and low atmospheric Ar contamination. We attribute this phenomenon to the preferential concentration of atmospheric Ar in interlayer extended defects and of 40Ar∗ in interlayer vacancies, whereas 39Ar is uniformly distributed throughout the muscovite structure by recoil. The 40Ar/ 39Ar age spectrum of the muscovite heated under hydrothermal conditions has a staircase shape, comparable to that usually interpreted as due to 40Ar∗ diffusive loss, whereas 40Ar/ 39Ar laser spot analyses did not allow this interpretation. The in vacuo Ar release rate for this sample, however, indicates that it had been altered during the experiment. Therefore, we propose that this 40Ar/ 39Ar age spectrum reflects the differential release of Ar from the alteration phases vs. muscovite. For our sample, dehydroxylation and delamination control the in vacuo Ar release at temperatures less than 900°C, whereas Ar diffusion from thin dehydroxylated slabs may dominate the Ar release mechanism for temperatures greater than 900°C. This model would predict that the recovery of fossil 40Ar∗ concentration gradients from muscovite by typical incremental heating schedules is not possible. The rates of delamination in vacuo and the rate of Ar diffusion for muscovite in nature, however, both depend upon its the chemical composition, structural integrity, and grain size. Consequently, for samples with either a naturally occurring broad range of grain sizes, such as detrital muscovite and authigenic clay from sedimentary rocks, or a compositionally or structurally heterogeneous population of grains, such as muscovite from shear zones, 40Ar/ 39Ar incremental heating may be able to detect intergranular variations in the K/Ar age.
Read full abstract