231Pa EXCESS IN ZIRCON REVEALED BY MULTI-STEP CA-TIMS ANALYSES

Isotopic disequilibrium effects in the ²³⁸U->²⁰⁶Pb* system in zircon, such as partial exclusion of ²³⁰Th, are well understood and can be corrected for quite accurately. Inclusion of excess amounts of ²³¹Pa in the ²³⁵U->²⁰⁷Pb* system is less well understood and much more difficult to correct for, given the absence of any long-lived Pa isotopes. In a few spectacular cases very large amounts of excess ²³¹Pa, which in turn produce large excesses of ²⁰⁷Pb*, have been identified (i.e., Mortensen et al., 1997; Anczkiewicz et al., 2001), and there is little danger of isotopic age misinterpretation. However, smaller more subtle ²³¹Pa disequilibrium effects are less easily recognized. Such effects could easily be mistaken for discordance produced by Pb loss or inheritance of older zircon.

Detailed multi-step CA-TIMS analysis of a wide range of zircon samples reveals that a small number of them do indeed appear to show significant ²³¹Pa excess effects. Each of these samples has a plateau of statistically equivalent ²⁰⁶Pb*/²³⁸U ages, strongly indicative of the absence of either Pb loss or any older inheritance. However, all show a progression of ²⁰⁷Pb*/²⁰⁶Pb* ages: late-stage partial dissolution steps representing early crystallized cores of zircon have ²⁰⁷Pb*/²⁰⁶Pb* ages that are typically identical to, or only slightly older than, the corresponding ²⁰⁶Pb*/²³⁸U ages. This is suggestive of only very minor ²³¹Pa excess. Intermediate partial dissolution steps and early dissolution steps representing later-stage growth of zircon have progressively older ²⁰⁷Pb*/²⁰⁶Pb* ages, despite identical ²⁰⁶Pb*/²³⁸U ages. For the 69 Ma to 150 Ma samples used in this study, the ²⁰⁷Pb*/²⁰⁶Pb* ages within individual samples showed increases from ca. core zircon to ca. rim zircon of between ca. 5 and 16 m.y. These results suggest an increase in D_Pa/D_U from ca. 1-3 during the early stages of zircon crystallization to ca. 6-12 during the late stages of zircon crystallization.

These results emphasize the importance of reliance on robust ²⁰⁶Pb*/²³⁸U ages for high accuracy geochronology, especially for Phanerozoic samples.