Diffusion kinetics of 3He and 21Ne in quartz and implications for cosmogenic noble gas paleothermometry

The simultaneous production and diffusion of cosmogenic noble gases offers the potential to constrain past temperatures on Earth and other planetary surfaces. Knowledge of both the production rate and diffusion kinetics of cosmogenic nuclide pairs is required to utilize this open-system behavior for paleothermometry. Here, we investigate the diffusion kinetics of spallogenic He-3 and Ne-21 in quartz through a series of step-degassing experiments on individual, proton-irradiated quartz grains. Quartz often, but not always, exhibits two stages of linear Arrhenius behavior, with He and Ne exhibiting similar release patterns. This two-stage behavior does not appear to correlate with heating-induced structural changes or anisotropy, nor is it an artifact of proton irradiation. The behavior may instead be associated with a sample-specific property such as radiation damage, mineral inclusions, fluid inclusions, or structural defects. We interpret these two Arrhenius arrays to represent multiple diffusion domain (MDD)-type behavior in quartz, as two-domain models closely reproduce the experimental data. However, we are currently unable to link this behavior with a clear physical mechanism; a different, more mechanistic model may be more appropriate in future studies. For both He and Ne, modeled Arrhenius diffusion parameters (activation energy, E-a, and pre-exponential factor, D-0) display a range of values in the quartz samples analyzed. For 3 He, E-a ranges from 73.0 to 99.8 kJ/mol and D-0 from 5.9 x 10(0) to 1.0 x 10(4) cm(2) s(-1) for the initial, low-temperature linear Arrhenius arrays; when observed, a second array at higher temperatures corresponds to Ea ranging from 85.2 to 106.4 kJ/mol and D-0 from 1.7 x 10(-1) to 3.5 x 10(0) cm(2) s(-1). For Ne-21, E-a ranges from 95.7 to 153.8 kJ/mol and D-0 from 6.6 x 10(-1) to 3.2 x 10(3) cm(2) s(-1) for the initial, low-temperature array; linearity at high temperatures is not well constrained, likely because the alpha- to beta-quartz transition occurs during the relevant temperature range. When extrapolated to Earth surface temperatures and geologically relevant timescales, these results suggest that 1 mm-radius quartz grains lose significant amounts of cosmogenic He-3 by diffusion at sub-zero temperatures from the low-retentivity domain over >10(3) yr timescales and from the high-retentivity domain over >10 4 yr, whereas quantitative retention of cosmogenic Ne-21 occurs over >10(6) yr at temperatures <= 40 degrees C in most cases. While these results are generally consistent with previously reported studies, they also reveal that sample-specific diffusion parameters are required for quantitative application of cosmogenic noble gas paleothermometry. The cosmogenic He-3 abundance in one quartz sample with a simple Holocene exposure history and the stepwise degassing pattern of cosmogenic He-3 and Ne-21 from another quartz sample with a similar to 1.2 Ma exposure history agree well with diffusion experiments on proton-irradiated aliquots of the same samples. For the sample with a simple Holocene exposure history, a production and diffusion model incorporating sample-specific diffusion parameters and the measured 3 He abundance predicts an effective diffusion temperature consistent with the effective modern temperature at the sample location. This internal consistency demonstrates that the empirically