Problems with argon: K-Ar ages in Gulf Coast shales

K-Ar ages of Gulf Coast shales do not correspond to diagenetic `events', but are a combination of inherited detrital ages and authigenic overprints. The key to understanding K-Ar ages of shales is the fate of the Ar-40 when the host minerals dissolve or react during the illitisation reaction: is Ar-40 incorporated within the illite, or is it lost to the porewaters? Ages are modelled as due to a single illitisation reaction, assuming a uniform precursor sediment and open-system import of K. Both whole-rock and clay mineral fraction data can be reproduced by modelling. The Ar-40 incorporated into illite is a surprisingly high proportion (75-100%) of that produced within detrital phases (smectite, K-feldspar, mica), perhaps implying that the illitisation reaction is a solid-state transformation, or that Ar-40 is imported into the shales along with incoming K. The model does not provide a good fit to the data below 4000 m, when the illitisation reaction is apparently complete; but shales lose Ar-40 by solid-state diffusion or recrystallisation. (C) 2002 Elsevier Science B.V. All rights reserved.