Abstract The retrieval of accurate thermal histories recorded by the (U-Th)/He system relies on empirical estimates of diffusion kinetics from natural minerals, although the difficulty in obtaining samples with homogeneous gas concentrations (required for accurate diffusivities) has limited the collection of such datasets. Whole-grain analyses are relatively time- and cost-effective, but natural 4He concentration heterogeneities caused by α-ejection, diffusional rounding, and/or parent nuclide zonation present an unquantified source of error. We employ a 1D spherical finite difference diffusion model to simulate effects of natural sources of He concentration heterogeneities, with a focus on the zircon (U-Th)/He (ZHe) system. Using He concentration profiles affected by various natural heterogeneities, models consistently predict that diffusion kinetics are strongly underestimated using such methods, especially at low gas release fractions. These underestimations are driven by He depletion near the grain boundary caused by α-ejection, and exacerbated by U and Th depletion. Conversely, models with parent nuclide (and therefore He) enrichment near the grain boundary compensate for ejection-based loss, yielding more accurate diffusion kinetics overall. A comparison of these results to real step-heating data yields few similarities, however, with He releases from real zircons either being linear in Arrhenius space, or having anomalously high diffusivities at low fractional releases. Radiation damage and its annealing are also significant factors affecting He diffusional mechanics. Using diffusivities and annealing as predicted by the current ZHe forward model, cycled step-heating of high-dose zircons (>1 × 1018 α/g) predict that diffusion decreases consistently with increased heating temperatures, similar to observations. The forward model predicts that annealing beyond this threshold dose should increase diffusivity, which is not observed, however. These results suggest that whole-grain step-heating data may provide a reasonable proxy for empirically measuring damage annealing kinetics in zircon. Major discrepancies between model results and real data reveal the need for more cycled-step heating data to understand the full complexities of He diffusion dynamics.

中文翻译：

---

非均质 4He 浓度分布和辐射损伤退火对全晶锆石扩散率分析的模拟影响

摘要 由 (U-Th)/He 系统记录的准确热历史的检索依赖于对天然矿物扩散动力学的经验估计，尽管获得具有均匀气体浓度的样品（准确扩散率所需）的困难限制了收集这样的数据集。全谷物分析相对具有时间和成本效益，但由 α 喷射、扩散舍入和/或母体核素分带引起的自然 4He 浓度异质性存在未量化的误差来源。我们采用一维球形有限差分扩散模型来模拟自然源 He 浓度异质性的影响，重点是锆石 (U-Th)/He (ZHe) 系统。使用受各种自然异质性影响的 He 浓度分布，模型一致预测，使用此类方法会严重低估扩散动力学，尤其是在气体释放率较低的情况下。这些低估是由 α 喷射引起的晶界附近的 He 耗尽驱动的，并因 U 和 Th 耗尽而加剧。相反，在晶界附近具有母核素（因此 He）富集的模型补偿基于喷射的损失，从而产生更准确的整体扩散动力学。将这些结果与真实的阶梯加热数据进行比较，几乎没有相似之处，然而，真实锆石的 He 释放要么在 Arrhenius 空间中是线性的，要么在低分数释放时具有异常高的扩散率。辐射损伤及其退火也是影响氦扩散力学的重要因素。使用当前 ZHe 正向模型预测的扩散率和退火，高剂量锆石 (>1 × 1018 α/g) 的循环步进加热预测扩散随加热温度升高而持续降低，与观察结果相似。正向模型预测，超过此阈值剂量的退火会增加扩散率，但并未观察到这一点。这些结果表明，全晶粒阶梯加热数据可以为经验测量锆石的损伤退火动力学提供合理的代理。模型结果与实际数据之间的主要差异表明，需要更多的循环步骤加热数据来了解 He 扩散动力学的全部复杂性。类似于观察。正向模型预测，超过此阈值剂量的退火会增加扩散率，但并未观察到这一点。这些结果表明，全晶粒阶梯加热数据可以为经验测量锆石的损伤退火动力学提供合理的代理。模型结果与实际数据之间的主要差异表明，需要更多的循环步骤加热数据来了解 He 扩散动力学的全部复杂性。类似于观察。正向模型预测，超过此阈值剂量的退火会增加扩散率，但并未观察到这一点。这些结果表明，全晶粒阶梯加热数据可以为经验测量锆石的损伤退火动力学提供合理的代理。模型结果与实际数据之间的主要差异表明，需要更多的循环步骤加热数据来了解 He 扩散动力学的全部复杂性。