Rb-Sr dating of sphalerite based on fluid inclusion-host mineral isochrons; a clarification of why it works

Abstract

Nakai et al. (1990, 1993), Brannon et al. (1992), and Christensen et al. (1995a, b) recently published precise Rb-Sr formation ages of sphalerites from various Mississippi Valleytype deposits. This important advance in direct radiometric dating of ore deposition is based on a crush-leach procedure in which the Rb-Sr isotope composition of the fluid inclusion fraction (termed leachate) is analyzed separately from the residual host crystal (termed residue). Isochron arrays were obtained by combining the fluid-host mineral data pairs from several samples. The resulting ages have yielded decisive chronological constraints on controversial genetic models (as summarized, for example, by Christensen et al., 1995a) where few alternative methods for dating are available. Although the method appears successful, the above authors were unable to explain why the separation of sphalerite samples into fluid inclusion and residual solid fractions should yield an isochron age at all. Without an explanation, the accuracy of the ages remains uncertain. The aims of the present communication are to clarifY the origin of the isochron patterns and to evaluate their age significance in the way they have been interpreted hitherto, that is, in the absence of independent geochronologic information. Our arguments are based first on analyses of synthetic fluid inclusions in quartz, which we use as an experimental analogue of fluid inclusion-bearing sphalerites. In building on the analytical approach of Shepherd and Darbyshire (1981) for fluid inclusion dating, Pettke and Diamond (1995a) synthesized RbSr-bearing, aqueous fluid inclusions in quartz under high temperatures and pressures. Subsequent analysis of the separate fluid inclusion and quartz-host fractions revealed Rb-Sr patterns which are very similar to those found in natural sphalerite by Nakai eta!. (1990). Although recognizing that the analogy between the synthetic and natural samples is limited by important differences in crystal-chemical behavior between sphalerite and quartz, we are confident that the similarities between the isotopic results of these studies justifY inferences being drawn regarding the dating of natural sphalerites. Second, by manipulating the data published by Nakai et a!. (1990, 1993), Brannon et a!. (1992), and Christensen et a!. (1995a, b), we examine the nature of partitioning of Rb and Sr between sphalerite and hydrothermal parent fluid, and thus arrive at conclusions on the validity of dating with fluid inclusion-host sphalerite isotope arrays.