A Filtered View of Time: Improving the performance of the 36Cl / 10Be chronometer in Greenland ice cores by separation of the 10Be budget in ice and dust

Absolute dating methods are required to provide accurate age estimates of extremely thinned and folded layers in the deepest sections of ice cores. The ³⁶Cl / ¹⁰Be chronometer works on the principle that the ratio of ³⁶Cl (half-life; 301 kyr) will decrease relative to ¹⁰Be (half-life; 1.4 Myr) with increasing age of ice. This method is especially desirable because it generally requires less ice for analysis than other absolute dating techniques such as ⁸¹Kr. However, both ³⁶Cl and ¹⁰Be are affected by processes that complicate their reliability as geochronological tools. For instance, at low accumulation sites, the ³⁶Cl inventory can be depleted through hydrogen chloride outgassing, however ³⁶Cl is largely preserved in glacial periods due to increased buffering from alkaline species associated with increased dust content. Conversely, increased dust content during glacial periods can complicate the ¹⁰Be inventory due to adsorption of  ¹⁰Be onto dust. In deep ice, such ¹⁰Be migration has resulted in observations of the ¹⁰Be concentration decreasing faster than expected from physical decay alone (Kappelt et al., 2025), which can lead to potential age underestimates when using the ³⁶Cl / ¹⁰Be chronometer.

 

Here we focus on the issue of ¹⁰Be migration in deep ice by using a 0.45μm filter to separate the ¹⁰Be inventory attached to dust particles and in ice prior to ¹⁰Be measurement. We present preliminary results using our filtration method on Holocene, LGM and MIS-4 samples from the Dye-3 ice core from southern Greenland. Our results confirm that the impact of dust on the ¹⁰Be budget is more pronounced during glacial periods. Additionally, we use our filtration technique to test its potential to resolve the depletion of ¹⁰Be observed in the deeper sections of ice cores, by working on deep core sections that are independently dated by ⁸¹Kr. To make progress on better constraining the ¹⁰Be signal, we also present a modified sequential leaching technique, previously applied to ocean and river sediments. By performing sequential leaching on the filtered dust, we aim to separate the labile meteoric ¹⁰Be fraction (adsorbed from the ice) from the meteoric ¹⁰Be fraction that was already present at the dust surface prior to the incorporation of the dust into the ice. In better constraining the impact of ¹⁰Be migration onto dust on the total ¹⁰Be inventory in deep ice cores we hope to improve the accuracy of the paired ³⁶Cl / ¹⁰Be chronometer for small-sample (< 1 kg) ice core analyses.