Abstract. A potential background for the SuperCDMS SNOLAB dark matter experiment is from radon daughters that have plated out onto detector surfaces. To reach desired backgrounds, understanding plate-out rates during detector fabrication as well as mitigating radon in surrounding air is critical. A radon mitigated cleanroom planned at SNOLAB builds upon a system commissioned at the South Dakota School of Mines & Technology (SD Mines). The ultra-low radon cleanroom at SD Mines has air supplied by a vacuum-swing-adsorption radon mitigation system that has achieved >1000× reduction for a cleanroom activity consistent with zero and <0.067 Bq m −3 at 90% confidence. Our simulation of this system, validated against calibration data, provides opportunity for increased understanding and optimization for this and future systems.
RADON BACKGROUNDS FOR SUPERCDMS SNOLABA potential source of dominant backgrounds for many rare-event searches or screening detectors is from radon daughter plate-out [1,2]. Backgrounds from 210 Pb and the recoiling 206 Pb nucleus from the α decay of 210 Po were the dominant low-energy backgrounds for XMASS [3,4], SuperCDMS Soudan [5], and EDELWEISS [6]. These backgrounds remain the dominant surface backgrounds for EDELWEISS-III [7]. Mitigation of radon daughters on surfaces has also been critical for SuperNEMO [8] Radon-daughter backgrounds are important to the expected low-mass sensitivity of the SuperCDMS SNOLAB experiment, which will use detectors of germanium and silicon to search for dark matter interactions [13]. Although the SuperCDMS Interleaved Z-sensitive Ionization and Phonon (iZIP) detectors provide excellent rejection of surface events above 8 keV [14], at lower energies the rejection is expected to worsen to the point that radon-daughter backgrounds may dominate [13]. The situation is similar for the SuperCDMS high-voltage (HV) detectors, which provide the experiment's lowest energy threshold by amplifying the ionization signal [5,15]. Below energies of ∼0.5 keV, rejection of events on the detector sidewall surface, based on the relative sizes of signals in the detectors' phonon sensors, becomes ineffective, so radon-daughter surface backgrounds are expected to dominate over bulk backgrounds for 210 Pb concentrations 50 nBq cm −2 . Figure 1 shows how the expected sensitivity of SuperCDMS SNOLAB varies for different surface concentrations of 210 Pb.Without mitigation, radon daughter plate-out during assembly underground at SNOLAB could dominate the expected background. The duration of the SNOLAB assembly is estimated at about 150 hours. At the average SNOLAB radon concentration of 130 Bq m −3 , the expected plate-out would be about 70 nBq cm −2 of 210 Pb. To make this plateout rate negligible, SuperCDMS SNOLAB has a goal of achieving a radon concentration underground of <0.1 Bq m −3 in its assembly cleanroom underground.