THE EXTREME RANGE OF ARGON ISOTOPE RATIOS IN OBSIDIAN

In the course of ⁴⁰Ar/³⁹Ar dating of various archeological obsidians, a number of sub-atmospheric ⁴⁰Ar/³⁶Ar ratios have been inferred from 3-isotope correlation plots (i.e., isochron diagrams). To further investigate this phenomenon, ³⁸Ar/³⁶Ar ratios of unirradiated aliquots of these samples have been measured, along with directly determined ⁴⁰Ar/³⁶Ar and ³⁸Ar/³⁶Ar ratios of very young obsidians having negligible radiogenic ⁴⁰Ar (⁴⁰Ar*) ingrowth. Initial results from 15 “old” samples, with initial ⁴⁰Ar/³⁶Ar determined relatively imprecisely by isochron intercepts, show crudely correlated ⁴⁰Ar/³⁶Ar and ³⁸Ar/³⁶Ar ratios. Analysis of “young” samples with negligible ⁴⁰Ar* show an extremely well-defined positive linear correlation between these ratios, whose slope is slightly steeper than expected for kinetic fractionation (e.g. Young et al., 2002). These results are consistent with trends determined from some basalts and andesites (e.g., Ozawa et al., 2006; Matsumoto et al., 1989) except that the obsidians display a more extreme range, including substantially supra-atmospheric as well as sub-atmospheric values (δ⁴⁰Ar = -89 to + 45 permil; δ³⁸Ar = -34 to +24 permil with respect to air). Kinetic fractionation during incomplete equilibration between cooling magma and air, with net flux into or out of the magma causing sub- or supra-atmospheric values (respectively), is supported by a general correlation between ³⁸Ar/³⁶Ar and ³⁶Ar concentrations. Imminent measurements of multiple samples from a single flow at Glass Mtn., near Medicine Lake in NE California, will be used to examine the expected variability in the extent of fractionation arising from variable diffusive lengthscales and cooling rates. Obsidians with low ⁴⁰Ar/³⁶Ar ratios may yield significantly erroneous ⁴⁰Ar/³⁹Ar plateau ages, and in cases where isochrons are indeterminate, ³⁸Ar/³⁶Ar ratios of unirradiated aliquots should be measured to test for fractionation.