The XENON1T experiment is the most recent stage of the XENON Dark Matter Search , aiming for the direct detection of dark matter candidates , such as the Weakly Interacting Massive Particles ( WIMPs ) .
The projected sensitivity for the spin-independent WIMP-nucleon elastic scattering cross-section is \sigma \approx 2 \times 10 ^ { -47 } \mathrm { cm } ^ { 2 } for a WIMP mass of m _ { \chi } = 50 \mathrm { GeV / c ^ { 2 } } .
To reach its projected sensitivity , the background has to be reduced by two orders of magnitude compared to its predecessor XENON100 .
This requires a water Cherenkov muon veto surrounding the XENON1T TPC , both to shield external backgrounds and to tag muon-induced energetic neutrons through detection of a passing muon or the secondary shower induced by a muon interacting in the surrounding rock .
The muon veto is instrumented with 84 8 " PMTs with high quantum efficiency ( QE ) in the Cherenkov regime and the walls of the watertank are clad with the highly reflective DF2000MA foil by 3M .
Here , we present a study of the reflective properties of this foil , as well as the measurement of its wavelength shifting ( WLS ) properties .
Furthermore , we present the impact of reflectance and WLS on the detection efficiency of the muon veto , through the use of a Monte Carlo simulation carried out with the Geant4 toolkit .
The measurements yield a specular reflectance of \approx 100 \% for wavelengths larger than 400 nm , while \approx 90 \% of the incoming light below 370 nm is absorbed by the foil .
Approximately 3 - 7.5 \% of the light hitting the foil within the wavelength range 250 \mathrm { nm } \leq \lambda \leq 390 \mathrm { nm } is used for the WLS process .
The intensity of the emission spectrum of the WLS light is slightly dependent on the absorbed wavelength and shows the shape of a rotational-vibrational fluorescence spectrum , peaking at around \lambda \approx 420 nm .
Adjusting the reflectance values to the measured ones in the Monte Carlo simulation originally used for the muon veto design , the veto detection efficiency remains unchanged .
Including the wavelength shifting in the Monte Carlo simulation leads to an increase of the efficiency of approximately 0.5 \% .