α-Emitting mineral inclusions in apatite, their effect on (U-Th)/He ages, and how to reduce it

U-Th rich mineral inclusions in apatite are often held responsible for erroneously old (U-Th)/He ages, because they produce “parentless” He. Three aspects associated with this problem are discussed here. First, simple dimensional considerations indicate that for small mineral inclusions, the parentless helium problem might not be as serious as generally thought. For example, a mineral inclusion that is 10% the length, width and height of its host apatite needs to be a thousand times more concentrated in U and Th to produce an equal amount of He. Therefore, single isolated inclusions smaller than a few μm are unlikely to contribute significant helium. For larger or more abundant inclusions, the parentless helium problem can be solved by dissolution of the apatite and its inclusions in hot HF. Second, besides creating parentless helium, inclusions also complicate α-ejection corrections. Mathematical exploration of this latter problem for spherical geometries reveals that for randomly distributed inclusions, the probability distribution of single-grain ages is predicted to have a sharp mode at the mean age, with tails towards younger and older ages. Multiple-grain measurements will yield accurate and precise age estimates if 10 or more randomly distributed α-emitting mineral inclusions are present in a sample. Third, thermal modeling indicates that mineral inclusions have a non-trivial but minor (<5 °C) effect on the closure temperature. These predictions were tested on apatites from rapidly cooled migmatites of Naxos (Greece) which contain abundant U-rich zircon inclusions. Thirty-seven samples were subjected to two kinds of treatment. The “pooled” age (i.e., the synthetic multi-grain age computed from a number of single-grain analyses) of four inclusion-free samples (13 apatites), prepared in HNO₃ is 10.9 Ma, close to apatite and zircon fission-track ages from the same rock. (U-Th)/He ages of 14 inclusion-bearing samples dissolved in HNO₃ range between 9 and 45 Ma, with a pooled age of 22.6 Ma. The ages of 19 HF-treated samples range between 5 and 16 Ma, with 10 of 14 single-grain samples between 9 and 13 Ma and a pooled age of 10.9 Ma. These observations agree with the theoretical predictions and support the addition of HF-treated apatite (U-Th)/He dating to the thermochronological toolbox.