Neutron background simulation for the CRESST-II experiment

Abstract

The CRESST collaboration reported an excess signal in their latest physics run. After removal of expected electromagnetic background, low energy αs and degraded lead recoils from the signal region, a mean value of 35.4 excess events remain. In this work, we investigate if neutron induced nuclear recoils can explain the reported excess. Three results are found: The total event rate in the simulation of 2.04+0.23-0.15 x 10-3 cts /kgd is much lower than the one inferred from the remaining excess events of 4.85 x 10-2 cts /kgd in the experimental data. Additionally, the dominance of oxygen recoils (∼ 90 %) predicted by the simulation is in disagreement with the favored experimental lightyield distribution of the excess signal. Furthermore, the experimentally observed fraction of higher multiplicities is only half of the ∼ 20 % obtained by the simulation. Especially the experimental absence of double detector hits in the presence of higher multiplicities is in clear contradiction to the results of the simulation. In combination, these three discrepancies between the experimental and the simulated background allows the conclusion that the observed excess signal cannot be explained by neutron induced nuclear recoils alone.