Feasibility of a Liquid Xenon Detector for Reactor Neutrino Detection Via Cevns

Feasibility of a Liquid Xenon Detector for Reactor Neutrino Detection via CEvNS

Kaixuan Ni University of California San Diego

Magniﬁcent CEvNS 2020 Low-threshold Two-phase Xenon Time Project Chambers are Well Developed for Dark Matter Search

• Liquid xenon (LXe) is the leading technology for WIMP dark matter search • LXe detectors from 10-kg to ton-scale (XENON1T) were operated stably over years. • Detector technology is well developed and calibration studied down to sub-keV NR

With both S1 and S2 signals (TPC mode):

• Ultra-low background • 3D ﬁducialization • ER/NR discrimination with S2/S1 • ER threshold: ~1 keV • NR threshold: 3~5 keV

K. Ni (UCSD) 2 Two-phase Xenon Detector in Electron Counting mode

• Using the ionization(S2)-only approach, LXe detector is sensitive to sub-keV nuclear recoils from sub-GeV DM or CEvNS.

With ionization(S2)-only signal (EC - Electron Counting mode):

• ER threshold: ~20 eV • NR threshold: ~300 eV

X Single-electron signals from the XeNeu detector (LLNL) Single electrons detection with 100% efficiency J. Xu et al., arXiv:1904.02885 3 K. Ni (UCSD) Noble Elements for Reactor Neutrino Detection

• based on reactor antineutrino spectrum from Mueller et al (1191.2553) and Hayes, Vogel (1605.02047) with 7.6% 238U, 25% 235U, 14.8% 241Pu, 51% 239Pu, normalized to a ﬂux of 6x1012 cm-2s-1 (~25 m from a 3 GWth reactor)

• ~10 events/kg/day expected at nuclear recoil energy threshold of 0.4~1.0 keV (target dependent)

CEvNS rate • A low-cost, compact neutrino detector with 10~100 kg above a threshold target: >100 neutrino events/day • LXe is a very promising target. LNe and LAr are also promising (require wavelength shift and 39Ar removal)

Liquid Density Threshold (keV) to Scintillation Intrinsic Elements Boiling Point (K) (kg/liter) get ~10/kg/day wavelength (nm) Radioactivity LHe 0.1 ~1.0 80 4 none

LNe 1.2 ~1.0 78 27 none

LAr 1.4 ~0.9 127 87 39Ar (1 Bq/kg)

LKr 2.4 ~0.5 148 120 85Kr (1 MBq/kg)

LXe 3.0 ~0.4 178 169 none pros cons 4 K. Ni (UCSD) LXe’s Response to Low Energy Nuclear Recoils well studied

LXe’s response to low energy recoils are well studied experimentally and fully modeled in the NEST software package down to 300-eV (single ionization electron)

Latest measurement Lenardo et al. 1908.00518

5 K. Ni (UCSD) Liquid Xenon detectors from 10-kg to ton-scale are built with single-electron sensitivity and operated for long-term

• 10-kg scale: XENON10, ZEPLIN-II/III • 100-kg scale: XENON100, RED-100, LUX, PandaX-I/II • 1~10 ton scale: XENON1T, PandaX-4T, XENONnT, LZ (commissioning)

E. Aprile et al. / Astroparticle Physics 35 (2012) 573–590 575

Fig. 3. (Top) The Hamamatsu R8520-06-Al PMTs on the top of XENON100 are arranged in concentric circles in order to improve the reconstruction of the radial event position. (Bottom) On the bottom, the PMTs are arranged as closely as possible in order to achieve high light collection, as required for a low detector threshold. Fig. 2. DrawingXENON100 of the XENON100 dark matter detector: detector the inner TPC contains (61-kg active target) single electron detection in XENON10 62 kg of liquid xenon as target and is surrounded on all sides by an active liquid xenon veto of 99 kg. The diving bell assembly allows for keeping the liquid–gas Two arrays of Hamamatsu R8520-06-Al 1’’ square photomulti- interface at a precise level, while enabling to fill LXe in the vessel to a height above plier tubes (PMTs), specially selected for low radioactivity [28], the bell. detect the light in the TPC: 98 PMTs are located above the target in the gas phase, arranged in concentric circles in order to improve The two-phase (liquid–gas) operation requires a precisely con- the resolution of radial event position reconstruction, see Fig. 3 K. Ni (UCSD) trolled liquid level just covering the gate grid. To minimize the im- (top). The outmost ring extends beyond the TPC radius to improve 6 pact of liquid density variations due to temperature changes as position reconstruction at the edges. The remaining photocathode well as fluctuations in the gas recirculation rate, a diving bell coverage is 43.9% of the TPC cross section area. The energy thresh- design was chosen to keep the liquid at a precise level. Outside old and hence the sensitivity of the detector is determined by the S1 the bell, the liquid in the detector vessel can be at an arbitrarily signal. Because of the large refractive index of LXe of (1.69 ± 0.02) high level. This made it possible to fill the vessel to a height of [36], and the consequent total internal reflection at the liquid–gas about 4 cm above the bell, enabling a 4p coverage of the TPC with interface, about 80% of the S1 signal is seen by the second PMT ar- a LXe veto. ray, which is located below the cathode, immersed in the LXe. Here, The bell keeps the liquid level at the desired height when a con- 80 PMTs provide optimal area coverage (in average 52% useful PMT stant stream of gas pressurizes it. This is accomplished by feeding photocathode coverage with 61% in the central part) for efficient S1 the xenon gas returning from the gas recirculation system (see light collection, see Fig. 3 (bottom). The bottom PMTs have a higher Section 3.5) into the bell. The pressure is released through a small quantum efficiency compared to the top PMTs. This is shown in pipe that reaches out into the veto LXe volume. The height of the Fig. 4, together with the distribution of the PMT quantum efficiency LXe level inside the bell is adjusted by vertically moving the open in the detector. The photoelectron collection efficiency from the end of the pipe which is connected to a motion feedthrough. photocathode to the first dynode for this type of PMT is about In order to minimize the dependence of the charge signal on the 70%, according to Hamamatsu. xy-position, the liquid–gas interface has to be parallel to the anode. A LXe layer of about 4 cm thickness surrounds the TPC on all To facilitate leveling, the detector can be tilted with two set screws sides and is observed by 64 PMTs, of the same type as used for from the outside of the radiation shield. Four level meters, measur- the TPC readout. In total, this volume contains 99 kg of LXe. The ing the capacitance between partially LXe filled stainless steel presence of this LXe veto, operated in anti-coincidence mode, is tubes and a Cu rod placed in their center, as well as the measured very effective for background reduction [27] and is one major dif- S2 signal width at different locations, are used to level the detector ference in design compared to XENON10. The LXe veto is optically (see Section 5.2). separated from the TPC by the interlocking PTFE panels. Optical CEvNS event rate in liquid xenon expected neutrino events vs. measured background

Using the liquid xenon ionization yield from the latest measurement by Lenardo, Xu et al., 1908.00518 and assumed a poisson distribution for expected # of e-

1206.2644 5 e-

1605.06262 1910.06190

1907.11485

7 K. Ni (UCSD) Detection sensitivity to CEvNS from reactor neutrinos in LXe

Assumptions: • reactor neutrino ﬂux of 6x1012 cm-2s-1 • S2-only background similar to those measured in XENON10/100 • Background uncertainty: 2% (2-5 e-), 8% (5-10 e-)

preliminary

Even with 5 e- threshold, a 5-sigma detection can be achieved in < 400 kg-day exposure. ➡ O(10) kg active target detector is suﬃcient for a ﬁrst detection ➡ further reducing the single-and-few electron background will improve the sensitivity for precision measurement 8 K. Ni (UCSD) Understanding the single-and-few electrons background

Investigation of background electron emission in the LUX detector (2004.07791)

● Show up several hundreds of milliseconds after an S2

● Single electrons are related to impurities in liquid xenon

● Electrons trapped at the liquid-gas interface emitted at a later time

9 K. Ni (UCSD) SE background from impurities in LXe

XENON100 (1311.1088) LUX (2004.07791)

Single e- rate in the bulk LXe: proportional Single e- rate decreases with higher to the concentration of impurities electron lifetime (LXe purity)

Solutions: significantly improve the LXe purity (1 ms -> 10 ms electron lifetime) to reduce SE background by a factor of 10 ➡ faster purification (technologies developed for the DM experiments) ➡ reduce outgassing materials e.g. PTFE ➡ improve purification efficiency (isolate active LXe target from other material)

10 K. Ni (UCSD) Sealed TPC to improve puriﬁcation efﬁciency (UCSD)

0.5-kg active LXe target

arXiv: 2007.16194

~0.5 ms electron lifetime achieved within 12 hr of circulation at 5 SLPM

11 K. Ni (UCSD) SE background from electrons trapped at the liquid-gas interface

● the Schottky barrier model: electrons approach the liquid-gas dielectric boundary are “trapped” (Sorensen, 1702.04805) ● a much stronger extraction field (>7 kV/cm in liquid) than previously thought is needed to extract all the electrons (Xu et al., 1904.02885) ● even a tiny fraction of trapped electrons will contribute significantly to SE bkg

Attempts/solutions to reduce the electrons trapped at liquid-gas interface:

● Apply very strong extraction field: > 7 kV/cm in the liquid (Xu et al., 1904.02885) ● Implement an electron shutter grid to block high-E events (Red-100, 1910.06190) ● Stimulate the electron emission using infrared light (Purdue) ● Eliminate the liquid-gas interface (UCSD)

12 K. Ni (UCSD) Detecting Electroluminescence (S2) in Liquid Xenon

● Electroluminescence in LXe was observed using a thin (10-um) wire at a threshold of 400 kV/cm (1408.6206), producing ~300 photons/electron ● Recently, we detected LXe electroluminescence from electrons emitted from a thin (25-μm) wire.

e- e- e- e- e- e-

S2 in Gas

Anode: +2 kV Cathode: -3 kV S2 in LXe wire

13 K. Ni (UCSD) Eliminating the liquid-gas interface

PMT Liquid surface S2 in LXe Anode

Gate

Anode: 0 kV Cathode: -3 kV Cathode wire

S2 in LXe electrons emitted from the wire

14 K. Ni (UCSD) NUXE: Reactor Neutrino Detection with Liquid Xenon

● O(10)-kg active LXe target ● ionization-only: single-electron sensitive ● detect ~100 neutrinos/day near a reactor ● Ultra-clean LXe target: ● ~10 ms electron lifetime ● Reduce/remove electron trapping: ○ S2 from LXe or very high extraction field ● Funded by DARPA YFA program 15 K. Ni (UCSD) Summary

● Liquid xenon is a very promising target to detect CEvNS from reactor neutrinos ● Special techniques developed recently to further suppress the single electron background will improve the sensitivity for precision measurement ● A low-cost O(10)-kg NUXE detector is currently being developed at UCSD

Collaborators are welcome!

Acknowledgement: the research reported in this presentation is sponsored by the Defense Advanced Research Project Agency, the content of the information does not necessarily reﬂect the position or the policy of the Government, and no oﬃcial endorsement should be inferred.

16 K. Ni (UCSD)