Argon-loss by alkali feldspars

Abstract

Ar-loss is generally thought to preclude the use of alkali feldspars for dating old rocks by the K-Ar method, and has been thought to occur via structural features such as the boundaries of perthite lamellae or twins. We have done a 40Ar- 39Ar step-heating study of a suite of alkali feldspars from the Klokken syenite intrusion (U-Pb age 1166 ± 4 Ma) which have been fully characterized by transmission electron microscopy. All are low albite-low microcline intergrowths, but some are non-turbid, ‘strain-controlled’ cryptoperthites, while others are turbid, coarse microperthites. Geological and textural evidence suggests that the turbidity developed by fluid-rock interactions during the initial cooling of the pluton, in the stability field of low microcline (<500°C), 10 5 a after crystallization. The finest cryptoperthites have numerous fully coherent perthite lamellar boundaries (lamellar periodicity ~ 40 nm) and twin composition planes. The have retained radiogenic 40Ar and yield total degassing ages as high as 1125 ± 16 Ma and maximum ages on the age spectra of 1162 ± 16 Ma. The coarse turbid microperthites, with patchy intergrowths and twins on the mm-scale, gave total degassing ages as low as 650 Ma (maximum age 750 Ma). Argon leakage occurs least in feldspars with the most lamellar interfaces, but in proportion to the degree of turbidity, which is thought to be caused by large numbers of micropores. Strain-controlled intergrowths in perthites allow the recognition of volumes of alkali feldspar crystals which have not interacted with deuteric fluids. These retain argon and can give reliable K-Ar ages in old rocks. Saddle-shaped age spectra appear to result from the preferential degassing, at 1000°C, of part of the feldspar structure which has lost more of its argon in nature. The laboratory degassing does not occur in a manner analogous to that which has occurred naturally.