In pursuit of the 40K branching ratios:: K-Ca and 39Ar-40Ar dating of gem silicates

This paper presents combined Ar-39-Ar-40 and K-Ca analyses on gem quality, K-rich silicates: a muscovite and a sanidine, both from pegmatites. We attempted to directly determine the K-40 branching ratio by comparing the radiogenic ingrowth of the daughter isotopes Ca-40* and Ar-40*. Sub-microgram amounts of Ca were analysed for isotopic composition with an external reproducibility of 0.03%. Ca analytics involved a Ca-43/Ca-48 double spike, miniaturised ion-exchange columns with a 1 ng Ca blank, and a Ta2O5 based loading reagent to allow for sub-microgram analyses. The sanidine gave clearly conflicting Ar-39-Ar-40 (461 +/- 6 Ma, 2 sigma, including uncertainty on the age monitor) and K-Ca (477 +/- 2 Ma) ages. The reason of the age discordance may be due to diffusive Ar losses; a low-temperature recrystallization is ruled out by the disordered Si-Al distribution. The muscovite sample gave nearly coincident Ar-39-Ar-40 and K-Ca ages applying currently recommended decay constants. Thus, analytically, this sample fits the requirements for a direct determination of the K-40 branching ratio, which we calculate as B-Ar=0.1067, B-Ca=0.8933. These values are different from those generally used in geochronology, and closer to the values used in the physics community. It is clear, however, that the branching ratio proposed here needs to be verified by further investigation on supposedly ideal samples. The precision with which we are able to resolve the individual ages also reveals complications of the geological history even in gem quality silicates. (C) 2000 Elsevier Science B.V. All rights reserved.