The XENONnT dark matter experiment

Abstract

The multi-staged XENON program at INFN Laboratori Nazionali del Gran Sasso aims to detect dark matter with two-phase liquid xenon time projection chambers of increasing size and sensitivity. The XENONnT experiment is the latest detector in the program, planned to be an upgrade of its predecessor XENON1T. It features an active target of 5.9 tonnes of cryogenic liquid xenon (8.5 tonnes total mass in cryostat). The experiment is expected to extend the sensitivity to WIMP dark matter by more than an order of magnitude compared to XENON1T, thanks to the larger active mass and the significantly reduced background, improved by novel systems such as a radon removal plant and a neutron veto. This article describes the XENONnT experiment and its sub-systems in detail and reports on the detector performance during the first science run.

Fig. 1

Schematic of the XENONnT experiment highlighting the detectors (TPC, neutron veto, muon veto) as well as the sub-systems for cryogenics and purification. Orange and blue arrows are used to show the path of gaseous and liquid xenon, respectively. The solid lines indicate the default flow path during normal operation, while the dashed lines indicate redundant flow paths and the optional flow path to allow for online krypton distillation. Details on the cryogenic system (CRY), the gaseous and liquid purification systems (GXe-PUR and LXe-PUR), and the krypton and radon removal plants (Kr-DST and Rn-DST) are reported in Sect. 2.2. The TPC and the veto systems are described in Sects. 2.1 and 2.3, respectively

Fig. 2

CAD rendering of the XENONnT TPC within the cryostat. The zoomed insets show details about the field cage, the electrode stacks at the top and bottom of the TPC, and the implementation of the PMT arrays

Fig. 3

The XENONnT TPC suspended from the top flange of the inner cryostat vessel shortly before the assembly was completed within the cleanroom inside the water tank. One panel of the PTFE sheet surrounding the lower part of the TPC was removed to show how the cathode high voltage feedthrough connects to the cathode electrode

Fig. 4

PMT arrays during TPC assembly: (top) The top array installed inside the diving bell. (bottom) The cathode and bottom screening electrodes are installed above the bottom PMT array

Fig. 5

Illustration (not to scale) of setting LXe level of the TPC with a diving bell: The lower end of the exhaust pipe defines the height of the liquid–gas interface

Fig. 6

The Hamamatsu R11410-21 PMTs in the top (253 units) and bottom (241) arrays were distributed according to their room-temperature quantum efficiency (QE, color code). PMT 383 was added just before assembly to replace one of the PMTs which performed worst during testing, i.e., it was not positioned based on its QE

Fig. 7

(top) The electrode wires are individually fixed by copper pins to the electrode frames (here: anode). Two transverse wires, installed to reduce sagging, are shown as well. (bottom) Detailed outside view of the TPC field-shaping elements during installation: the massive guard rings (15 mm height) and the thinner wire field electrodes (2 mm diameter) are installed at half pitch at different distances from the 3 mm thick PTFE reflectors. The guard electrodes consist of two pieces fixed together by two M3 SS bolts at each junction. The other two bolts are used to fix the guard-ring resistor chain (installed within the guard rings). Also visible are the 5 G$\Omega$ resistors on PTFE supports connecting the wire field electrodes

Fig. 8

CAD rendering of the xenon handling system in XENONnT. It includes the cryostat hosting the LXe TPC, the cryogenic system used to cool the xenon (CRY), the liquid (LXe-PUR) and gaseous purification (GXe-PUR) systems for electronegative impurity removal, the cryogenic distillation columns for krypton (Kr-DST) and radon removal (Rn-DST), ReStoX1 and ReStoX2 for LXe storage, filling and recovery, and the gas bottle rack for adding xenon gas into the system

Fig. 9

3D CAD rendering of the radon distillation column. Its total height is 3.8 m and it is continuously operated to reduce the $_{}^{222}$Rn level to the design goal

Fig. 10

CAD rendering of the neutron veto structure (white reflectors, 120 PMTs) and the calibration system. The I-belt (blue) is used to move a tungsten collimator containing a source close to the cryostat. The L-shaped beam pipe (pink) for neutron calibrations consists of the neutron guide and beam arms, pointing towards the water tank top and the cryostat, respectively. Unshielded calibration sources can be inserted into the two U-tubes (red, green) surrounding the cryostat

Fig. 11

A picture of the neutron veto taken from below during its assembly (the bottom octagon of light reflectors was not yet in place). In the center the cryostat, covered with ePTFE sheets, and the LXe recirculation pipe are visible

Fig. 12

Overview of the XENONnT slow control system: the information from the various sub-systems underground is acquired by Programmable Automation Controllers (PACs) and eventually stored in a database. The remote accessibility of the system is crucial for safe detector operation and continuously monitored by heartbeat servers

Fig. 13

Overview of the XENONnT data acquisition system [90]. The PMT signals from the neutron and muon veto systems and the amplified TPC PMT signals (top and bottom array) are digitized and time-stamped before being read out. The attenuated signals (0.5$\times$) of the bottom array PMTs are summed using a cascade of linear fan-in/fan-out modules. Logic modules inhibit data readout in case of high-energy events or if digitizers are busy

Fig. 14

Signals detected during a calibration with an Am-Be source. They are associated to a neutron interaction in the TPC and a coincident 4.4 MeV $\gamma$-photon interacting in the neutron veto (filled with demineralized water), followed by a 2.2 MeV $\gamma$ from subsequent neutron capture in the water. The bottom combined display summarizes the time correlation between these events in the two detectors: at $t=0$ a neutron is emitted; the associated $\gamma$-photon interacts through Compton scattering in the water, inducing Cherenkov light, which is detected by neutron veto PMTs within few tens of ns. At the same time, the neutron scatters in the xenon target producing a nuclear recoil and inducing the emission of prompt light (S1). The back-scattered neutron leaves the TPC and, upon reaching the NV, is moderated and captured by a proton, inducing the emission of a 2.2 MeV photon detected by the neutron veto PMTs after $\sim$200 $\mathrm{\mu }$s. Within the TPC two S2-type signals are detected. Properties of the S1 and of the S2-like signals are used to define the primary S2 that, in this case, is the one recorded around 550 $\mathrm{\mu }$s. The alternative S2 is compatible with a single electron signal. The S1 and S2 signals are shown with higher time resolution in the top left two panels. The top-right plot shows a top view of TPC and neutron veto, with the hit pattern for the S2 signal as seen by the TPC top PMT array (see central circular panel; the magenta dot indicates the reconstructed S2 position) as well as the hit pattern from the detection of the 4.4 MeV $\gamma$-photon in the neutron veto PMTs. The staggered circles indicate the PMTs of the neutron veto from the bottom (inner) to the top (outer circles). The circle size is proportional to the detected signal and the color code shows the time of detection

Fig. 15

Event rate in the neutron veto as a function of time for 4 (red), 6 (green), and 10-fold (blue) PMT coincidence after filling the water shield with demineralized water. The initial high rate due to contamination with natural $_{}^{222}$Rn decreases consistent with its half-life, after which the rates stabilize to below 100 Hz even for the 4-fold coincidence requirement

Fig. 16

Electron lifetime evolution as measured with the purity monitor. A clear improvement occurs after the start of cryogenic LXe purification operated at 2 LPM with the high-efficiency O$_2^{}$ purifier (Q-5). The model assumes a purifier efficiency of 100% and an O$_2^{}$ outgassing rate of 0.11 mg/day

Fig. 17

Light yield variation as a function of the reconstructed interaction depth and radius derived from the 41.6 keV $_{}^{83m}$Kr summation line. The reference frame is such that the liquid/gas interface and cathode planes are located at z = 0 cm and $z\sim -$149 cm, respectively

Fig. 18

Relative S2 correction maps for the top array (top) and bottom array (bottom) as a function of the reconstructed coordinates $X_{\text{obs}}$ and $Y_{\text{obs}}$ derived from $_{}^{83m}$Kr data. The two evident narrow bands with enhanced S2 response are produced by the transverse wires located on the gate and anode electrodes to mitigate their sagging

Fig. 19

2D histogram of the S2 area (with XY spatial corrections applied) as a function of drift time for $_{}^{83m}$Kr decays. In red, overlaid to the 2D histogram, a negative exponential fit of the average S2 area values (black points) used to extract the electron lifetime ${\tau }_e=(15.0\pm 0.4)$ ms for this run

Fig. 20

Light yield $L_y$ and charge yield $Q_y$ (in PE/keV) for several monoenergetic ER lines between 2.8 keV ($_{}^{37}$Ar [85]) and 4.4 MeV ($_{}^{12}$C). The values from low-energy lines $_{}^{37}$Ar, $_{}^{83m}$Kr, $_{}^{131m}$Xe and $_{}^{129m}$Xe (the latter two activated during neutron calibration) are fitted with the function $Q_y=-\frac{g_2}{g_1}{L}_y+\frac{g_2}{W}$ derived from Eq. (2). The best-fit result is shown in red. Higher-energy lines were excluded from the fit since they are more affected by potential reconstruction biases