DISPERSION HAPPENS: EVALUATING THE UBIQUITOUS SCATTER IN ULTRA-PRECISE 40AR/39AR SANIDINE DISTRIBUTIONS

⁴⁰Ar/³⁹Ar sanidine geochronology is a leading method for determining absolute timing and rates of important geological processes. However, ⁴⁰Ar/³⁹Ar analyses of sanidine consistently yield over-dispersed (MSWD >> 1) single crystal populations when individual crystals have precision of ~±0.3 per mil or better. This dispersion creates ambiguity for assigning accurate eruption ages. We explore several possible mechanisms that may lead to dispersion that include 1) over estimation of analytical precision, 2) variable degrees of melt inclusion hosted excess argon, and 3) crystal-to-crystal variation in neutron dose during irradiation. For (1), a bulk sample of Fish Canyon sanidine (FCS) was ground to less than 100 μm following irradiation to homogenize any potential grain-to-grain variation in ⁴⁰Ar*/³⁹Ar. Individual ~0.3 mg aliquots of crushed FCS were analyzed in the same manner as typical single-crystal laser-fusion experiments. Crushed FCS aliquots yield analytical precision of ±3-5 ka and a normal distribution demonstrating that sanidine age dispersion is not caused by inaccurate error estimation. For (2), mid-Tertiary sanidine samples were handpicked into populations with and without melt inclusions. Some inclusion-bearing populations show more dispersion and skew to slightly older ages, however both FCS populations yield indistinguishable ages indicating that melt inclusions are not a primary source of dispersion. For (3), the spatial distribution of sanidine grains was carefully controlled by placing only a few crystals in each irradiation position and neutron flux was monitored using an irradiation disk with a dense placement of positions (152 0.08” holes in a 1” diameter disk). Dispersion was substantially reduced indicating that grain-to-grain variation of neutron flux plays a prominent role in apparent age dispersion and is consistent with our observations that dispersion is proportional to sample age. Reduction of fluence gradients during irradiation will play a key role towards reducing sanidine age dispersion. Temporally similar Tertiary volcanic rocks with known stratigraphic relationships from southern CO were dated. Using well-controlled irradiation geometries, we show that eruptions ages that vary by as little as 10-15 ka can be accurately distinguished from one another.