Is apatite U-Th zonation information necessary for accurate interpretation of apatite (U-Th)/He thermochronometry data?

New U, Th and Ce concentration maps were acquired by LA-ICPMS for 70 apatites from 18 cratonic basement samples from the Canadian shield to characterize the nature and variability in apatite U–Th zonation. All apatites are zoned in effective U concentration (eU), with 80% exhibiting modest zoning that varies from a factor of ∼1.2 to ∼2.4. Zonation patterns include those with a general eU decrease from core to rim (∼25%), a core to rim eU increase (∼35%), and patchy or irregularly shaped eU distributions (∼40%). Most samples consist of individual apatites with markedly different and opposing eU profiles. Cathodoluminescence (CL) images were obtained for 258 apatites in 25 samples. Comparison of eU and CL patterns reveals no consistent relationships, suggesting that CL is not a reliable proxy for apatite eU zonation. We explore the implications of eU zonation for the interpretation of apatite (U–Th)/He (AHe) data by comparing the age predictions for representative and endmember apatite eU profiles with those for unzoned apatites. We focus on thermal histories that should magnify eU zonation effects, including slow monotonic cooling, extended He partial retention zone (HePRZ) residence, and protracted reheating and cooling. Application of incorrect α-ejection correction (FT) factors, different He concentration gradients, and variable intracrystalline retentivity due to heterogeneous radiation damage may cause the AHe dates for zoned and unzoned apatites to differ. In our dataset, the magnitude of the FT effect is \u3c1.5% for most samples. The He diffusion gradient and heterogeneous radiation damage effects cause apatites with eU enriched cores to yield AHe dates equal to or older than unzoned grains, with the opposite true for apatites with eU enriched rims. For monotonic cooling rates as slow as 0.1 °C/Myr, apatites of typical eU zonation exhibit a \u3c1 °C difference in effective closure temperature from equivalent unzoned grains. The difference in HePRZ temperature at 50% of an 150 Myr isothermal holding time is \u3c1.4 °C for typically zoned apatites. For most reheating simulations considered here, age deviations between zoned and unzoned apatites are not significant, and reach maximum differences of ∼10% for peak temperatures that induce partial He loss from the apatites. The age dispersion caused by zonation is comparable to that induced by grain size variations, but is considerably less than what can be caused by differences in mean eU and associated radiation damage for certain histories. By simulating zoned apatite suites from representative and endmember samples we find that eU zoning adds modest age dispersion to most samples. The commonly opposing eU profiles for individual apatites in most samples have different effects on the predicted age, and partially cancel each other out to reduce the predicted age difference at the sample level. Higher predicted age differences between zoned and unzoned samples generally correlate with greater sample scatter. The age deviations generally fall within the ∼15% uncertainty range that we commonly apply for sample AHe dates. Except in unusual circumstances, it appears unlikely that the unzoned apatite assumption will lead to misinterpretation of AHe datasets, thus precluding the need to routinely acquire apatite eU zonation information as part of AHe dating studies