Experimental Determination of Zn Isotope Fractionation During Evaporative Loss at Extreme Temperatures

Abstract Zinc isotopes fractionate during evaporation, and thus can potentially be used to calculate the proportion of volatile elemental loss from objects such as tektites, nuclear fallout melt glasses formed from silicate soils, and rocks from the Moon. The utility of the Zn isotope system in constraining the magnitude of volatile loss depends on accurate knowledge of its fractionation behavior during evaporation, i.e., the Zn isotope fractionation factor. Here, we present new results of Zn isotope analyses of experimentally heated soil samples, together with analyses of environmentally derived fallout melt glasses, and australite tektites, to better constrain how and to what extent Zn isotopes fractionate during thermally-driven evaporation. Major and trace element analyses of melt glass derived from the experimental heating of rhyolitic and arkosic soils demonstrates that elemental fractionation progressively becomes more important with increasing temperature and time. Initial Zn isotope ratios in the rhyolitic and arkosic soils are similar to upper continental crustal (UCC) values but become increasingly isotopically heavy as more Zn is lost from the heated glass, with a maximum δ 66 Zn value of +1.49 ± 0.05 ‰. The progressive loss of Zn from the rhyolitic soil is consistent with a fractionation factor (α) of 0.99879 ± 13. After loss of a labile component from the arkosic soil further evaporative loss of Zn from the silicate rich residue is also consistent with an α of ∼0.9988. This fractionation factor is not sensitive to either the heating temperature or duration. The fallout melt glass, derived from near surface nuclear weapons tests, and natural tektite samples are both relatively enriched in isotopically heavy Zn compared to the UCC. Although the fallout melt glass is largely comprised of the same rhyolitic soil used in our experiments the extent of Zn isotope fractionation is lower; consistent with an α of 0.9997. This is similar to estimated values of α associated with loss of Zn from Trinitite and lunar rocks. It is more difficult to constrain a value for α from australite tektites because the location of the impact site is not well constrained. Based on our analyses, we estimate a range of α values between 0.9988-0.9997, which is broadly consistent with experimental data and analyses of fallout melt glass. In all cases the measured Zn isotope fractionation factors are much lower than would be expected from evaporative loss of Zn in a high vacuum. We hypothesize that, for Zn, the value of α is dependent on the localized gas pressure and composition, with the degree of isotopic fractionation decreasing from high vacuum ( -9 bar) to atmospheric pressure. This would have consequences for the interpretation of Zn systematics on the Moon. Recent studies suggest that Zn loss from the Moon was dominantly caused by the Giant Impact and proceeded with an α of ∼0.9997. If correct, and taken together with our experimental data, this suggests that evaporative loss of Zn from the Moon is unlikely to have occurred in high vacuum and is more likely to have taken place at pressures similar to or higher than the current terrestrial atmosphere.