An investigation of artificial biasing in detrital zircon U-Pb geochronology due to magnetic separation in sample preparation

The application of detrital zircon geochronology for provenance analysis is complicated by the presence of biases induced by natural processes and sample preparation. The biasing of age distributions as a result of magnetic susceptibility is illustrated using sensitive high resolution ion microprobe dating of detrital zircon from a metaquartzite sample partitioned using a Frantz magnetic barrier separator. The relationship of paramagnetism with U content, a-dose, and discordance is demonstrated, but no relationship between grain size and discordance or age is found. The data also demonstrate that previous limits of zircon survival in sedimentary processes based on U content alone are too simplistic. Two age modes at similar to3150 and similar to2960 Ma are present in all the paramagnetic fractions; there is a bias toward the similar to3150 Ma mode being more prominent in the least-paramagnetic fractions. While the similar to2960 Ma is present in the least-paramagnetic fraction, it is argued that such fortuitous representation cannot be assumed before analysis. Such there or not provenance interpretations are considered simplistic, and at the very least there is no harm in broadening the range of paramagnetic fractions sampled for analysis. The results indicate a compromise between broad representation and analytical efficiency (avoiding discordant and thus unreliable results) can be made with a Frantz setting of similar to1.8 Angstrom and 10degrees side-slope. Copyright (C) 2002 Elsevier Science Ltd.