Small Volume Isotopic Analysis of Zircon Using LA-MC-ICP-MS U-Pb and Lu-Hf and Sub-ng Amounts of Hf in Solution

Crust-mantle evolution studies are greatly informed by zircon U-Pb and Lu-Hf isotopic datasets and the ease with which these data can now be acquired has seen their application become commonplace. In order to deconvolute geochemical change and interpret geologic variation in complexly zoned zircons, this information is most ideally obtained on the smallest volume of zircon by successive SIMS U-Pb and LA-MC-ICP-MS Lu-Hf isotopic analyses. However, due to variations in zircon growth zone geometry at depth, the Lu-Hf analysis may not relate to the lower volume U-Pb analysis, potentially causing inaccuracy of the resultant age-corrected Hf isotope signature. Laser ablation split-stream methods are applied to be certain that U-Pb and Lu-Hf data represent the same volume of zircon, however, the sampling volume remains relatively large at 40x30µm1. Coupled ID-TIMS U-Pb and solution MC-ICP-MS Lu-Hf work traditionally utilize whole-zircon dissolution ( 10-50ng Hf), which has the potential to homogenize different zones of geologic significance within an analysis. Conversely, modern ID-TIMS U-Pb methods utilize microsampling of zircon grains, often providing < 5ng Hf, thereby challenging conventional Lu-Hf acquisition protocols to achieve the required precision. In order to obtain usable precision on minimal zircon volumes, we developed laser ablation methods using successive 25um spot U-Pb and Lu-Hf ablation pits with a combined depth of 18um, and low-volume solution introduction methods without Hf-REE separation utilizing Hf amounts as low as 0.4ng, while retaining an uncertainty level of ca. 1 ɛHf for both methods. We investigated methods of Yb interference correction and the potential for matrix effects, with a particular focus on the accurate quantification of 176Lu/177Hf. These improvements reduce the minimum amount of material required for U-Pb and Hf isotopic analysis of zircon by about an order of magnitude. 1Ibanez-Mejia et al (2015). PreRes, 267, 285-310.