ZIRCON ZONATION STYLES IN DATE-EU SPACE: EXPLAINING FORWARD MODEL DEVIATION FROM OBSERVED DATA

Natural zircon crystals frequently display strong internal zonation as a result of grain growth, and metamorphism. These heterogeneous concentrations of radioactive actinides, and resultant radiation damage accumulation, change the diffusion behavior of He in zircon and therefore affect radiation damage based modeling systems. To understand how zonation impacts He diffusion and forward model predicted grain dates, we characterized suites of zircon crystals from three sample sites, the northern Madison Range of Montana, the Mecca hills gneiss, and Punchbowl formation. These samples were previously investigated using (U-Th)/He thermochronology and represent varying relationships between date and U-Th concentration with notable outliers. Detailed grain characterization using scanning electron microscopy, Raman spectral mapping and laser-ablation ICP-MS revealed a range of zonation styles throughout our sample sets. Complex zonation behavior (oscillatory, chaotic, rim-core zonation) is observed in the majority of representative grains from these basement ranges, indicating that many zircon grains that could be used for (U-Th)/ He dating do not follow basic assumptions of homogenous alpha particle production used in radiation damage based models. These models assume homogenous eU distribution and alpha particle ejection correction proportional to eU concentration; zonation results in either higher or lower alpha particle retention than expected, depending on how eU concentration is distributed throughout the grain, by producing either He “traps” (i.e. baddeleyite subgrains) or rapid diffusion pathways. For example, the forward modeling program ZRDAAM overpredicts the ages of zoned Northern Madison range grains at 413 ppm eU, 443 ppm eU, and 997 ppm eU (with eU profile ages of 109.0, 86.1, and 78.2 Ma respectively) by between 45 and 440 Myr. A thorough understanding of zircon zonation types observed in these rocks is necessary to understand the disparity between model predictions and observed dates, particularly for those grains with high concentrations of radioactive actinides.