Investigating the Utility of 40Ar/39Ar Data Acquired by Step-Heating of Alkali Feldspar for Thermochronology

Daniil Popov, Richard Alan Spikings
40 Ar/ 39 Ar data obtained from alkali feldspar by step-heating have been frequently interpreted using the multi-diffusion domain (MDD) theory to constrain cooling histories of rocks between ~150-350 ºC. This theory assumes that each alkali feldspar grain contains a population of variably-sized noninteracting diffusion domains, which form at high temperatures (e.g. during crystallisation from magma) and remain intact throughout the geologic history and step-heating experiment. It is further
more » ... . It is further assumed that the domains only loose Ar by volume diffusion over both geological and laboratory timescales. However, some researchers suggest that the redistribution and loss of 40 Ar over geological timescales is dominantly controlled by fluid interaction, and that laboratory step-heating of alkali feldspar significantly modifies its bulk diffusion properties. Their conclusions are incompatible with the assumptions behind MDD theory and thus dispute its ability to accurately reconstruct the thermal histories of rocks. This work aims to resolve the existing controversy and includes three studies that tackle different aspects of the problem. The first study revaluates the significance of high correlations between log(r/r0) and 40 Ar/ 39 Ar age spectra, which are frequently observed. Some researchers suggest that a close correlation between these spectra can only be explained by invoking diffusive loss of 40 Ar and thus validates the assumptions behind MDD theory. However, numerical modelling results presented here illustrate that highly correlated log(r/r0) and 40 Ar/ 39 Ar age spectra can also be obtained by fluid-assisted removal of 40 Ar, and that even volumetrically low degrees of alteration (3-6 vol.%) can result in gross inaccuracies in time-temperature paths obtained using MDD theory. The second study addresses the nature of the hypothetical intra-grain diffusion domains and their stability during laboratory step-heating. The existence of these domains was initially proposed to account for the frequent acquisition of non-linear Arrhenius trajectories of 39 Ar release from alkali feldspar. However, the experimental results provided here indicate that non-linearity of Arrhenius trajectories of 39 Ar release from alkali feldspar can be related to heating-induced fracturing, which in some cases progresses over the course of analysis. The third study re-examines the causes of intra-grain 40 Ar/ 39 Ar date variations in gem-quality Itrongay feldspar. Previous in situ 40 Ar/ 39 Ar work on this feldspar suggested that it has partially lost 40 Ar by volume diffusion and thus supported some of the assumptions behind MDD theory. However, the petrological and geochronological data reported here indicate that only a minor component of the variability of in situ 40 Ar/ 39 Ar dates of the studied Itrongay feldspar crystal was caused by diffusive loss of 40 Ar, while most variation relates to fluid-induced dissolution and overgrowth. Cumulatively, these results indicate that step-heating 40 Ar/ 39 Ar data derived from alkali feldspar should not be interpreted using MDD theory, unless the underlying assumptions of this theory can be validated for a given sample using petrological evidence. This work also shows that in situ 40 Ar/ 39 Ar dating of alkali feldspar combined with petrological characterisation is a promising tool for dating fluid flow events in non-sedimentary rocks. First of all, I would like to thank Richard Spikings and Urs Schaltegger for choosing me for this project.
doi:10.13097/archive-ouverte/unige:133126 fatcat:pjltcsvjrnga7hicjdvrn5y2xa