Exploring N2 capturing in Liquid Argon using Li-FAU Mol Sieve in the Iceberg Cryostat

Authors:
Publisher website
Google Scholar
Abstract
Liquefied ultrapure noble gases are typically the chosen target for neutrino and dark matter experiments. Commonly, the required grade of purity of such cryogenic liquids in terms of O2 contamination is less than 100 ppt, which makes it necessary for the argon to circulate through adsorption columns filled, typically, with BASF Cu-0226S and Mol Sieve 4A for capturing O2 and H2O. Another requirement that emerged during the design of LBNF-DUNE is the purity of the Liquid Argon (LAr) in terms of N2 concentration, as it was observed that both N2 and O2 contaminations can cause the reduction of the LAr scintillation light emission. In particular, a decreasing behavior in the lifetime and the relative amplitude of the slow component is relevant from concentrations of 1 ppm of N2 and 0.1 ppm of O2. Recently, exploratory experiments performed in the Liquid Argon Purification Cryostat (PuLArC — ∼90 L of LAr) had been successfully used to purify LAr from N2 using an innovative solid adsorbent based on a Li-FAU Molecular Sieve. Here, we report experiments of N2 capturing in LAr carried out at the ICEBERG Cryostat (∼3000 L of LAr) at Fermilab. For the experiments, several controlled injections of N2 were made into the cryostat, and the impurity concentrations in the LAr (both for N2 and O2) were monitored using a calibrated gas analyzer during the purification process. The experimental data confirmed that the Li-FAU is capable of capturing the N2 from the LAr. A numerical methodology based on the mass and energy balances of the system was applied to better understand the behavior of the adsorption process and to help with the scale-up of the process. A Pore Diffusion Model (PDM) coupled with parameter optimization was proposed to predict the N2 concentration in the ICEBERG cryostat over time. The model parameters applied in this work were determined based on experimental data acquired in the PuLArC. The model was able to accurately reproduce the data from the ICEBERG, indicating that the mathematical model could be a valuable tool to design and operate the GAr/LAr system of LBNF-DUNE. It was possible to predict from simulations the necessary time to reach the required purity from an initial contamination concentration and also the number of purification cycles until solid saturation.

This publication has 7 references indexed in Scilit: