Pb diffusion in monazite: A combined RSB/SIMS study

Abstract

We report measurements of Pb diffusion in both synthetic (CePO4) and natural monazites run under dry, 1-atm conditions. Powdered mixtures of prereacted CePO4 and PbZrO3 were used as the source of Pb diffusant for "in-diffusion" experiments conducted in sealed Pt capsules for durations ranging from a few hours to several weeks. Following the diffusion anneals, Pb concentration profiles were measured with Rutherford Backscattering Spectroscopy (RBS) and supplemented by measurements with secondary ion mass spectrometry (SIMS). In order to evaluate potential compositional effects upon Pb diffusivity and simulate diffusional Pb loss that might occur in natural systems, we also conducted "out-diffusion" experiments on Pb-bearing natural monazites. In these experiments, monazite grains were surrounded by a synthetic zircon powder to act as a "sink." Monazites from these experiments were analyzed with SIMS. Over the temperature range 1100 to 1350°C, the Arrhenius relation determined for in-diffusion experiments on synthetic monazite is given by: D=0.94 exp (-592±39kJ mol-1/RT)m2s-1 Diffusivities for synthetic and natural monazites are similar, as are results of measurements made on the same samples using both RBS and SIMS. The activation energy for Pb diffusion we determined is more than three times that for natural monazite reported previously, with Pb diffusivities at <1200°C significantly lower than those earlier reported but similar to that of zircon. Our data indicate that a 10-μm-sized monazite grain would have a Pb closure temperature in excess of 900°C, given a cooling rate of 10°C/Ma. Clearly, other factors may affect U-Th-Pb systematics in monazite including possible nondiffusive effects such as recrystallization mediated by fluids or triggered by metamorphic reactions with other REE-bearing minerals. However, our data suggest that monazite may be quite resistant to alteration of Pb isotopes by the specific mechanism of solid-state diffusional exchange.