Internal Calibration of the PandaX-II Detector with Radon Gaseous Sources
Wenbo Ma, Abdusalam Abdukerim, Zihao Bo, Wei Chen, Xun Chen, Yunhua Chen, Chen Cheng, Xiangyi Cui, Yingjie Fan, Deqing Fang, Changbo Fu, Mengting Fu, Lisheng Geng, Karl Giboni, Linhui Gu, Xuyuan Guo, Ke Han, Changda He, Shengming He, Di Huang, Yan Huang, Yanlin Huang, Zhou Huang, Xiangdong Ji, Yonglin Ju, Shuaijie Li, Huaxuan Liu, Jianglai Liu, Yugang Ma, Yajun Mao, Yue Meng, Kaixiang Ni, Jinhua Ning, Xuyang Ning, Xiangxiang Ren, Changsong Shang, Lin Si, Guofang Shen, Andi Tan, Anqing Wang, Hongwei Wang, Meng Wang, Qiuhong Wang, Siguang Wang, Wei Wang, Xiuli Wang, Zhou Wang, Mengmeng Wu, Shiyong Wu, Weihao Wu, Jingkai Xia, Mengjiao Xiao, Pengwei Xie, Binbin Yan, Jijun Yang, Yong Yang, Chunxu Yu, Jumin Yuan, Ying Yuan, Jianfeng Yue, Xinning Zeng, Dan Zhang, Tao Zhang, Li Zhao, Qibin Zheng, Jifang Zhou, Ning Zhou, Xiaopeng Zhou
PPrepared for submission to JINST
Internal Calibration of the PandaX-II Detector withRadon Gaseous Sources
PandaX-II Collaboration
Wenbo Ma, a Abdusalam Abdukerim, a Zihao Bo, a Wei Chen, a Xun Chen, a,b
YunhuaChen, c Chen Cheng, d Xiangyi Cui, e Yingjie Fan, f Deqing Fang, g Changbo Fu, g Mengting Fu, h Lisheng Geng, i,j
Karl Giboni, a Linhui Gu, a Xuyuan Guo, c Ke Han, a, Changda He, a Shengming He, c Di Huang, a Yan Huang, c Yanlin Huang, k ZhouHuang, a Xiangdong Ji, l Yonglin Ju, m Shuaijie Li, e Huaxuan Liu, m Jianglai Liu, a,b,e, Yugang Ma, g,n
Yajun Mao, h Yue Meng, a,b
Kaixiang Ni, a Jinhua Ning, c Xuyang Ning, a Xiangxiang Ren, o Changsong Shang, c Lin Si, a Guofang Shen, i Andi Tan, l AnqingWang, o Hongwei Wang, n,p
Meng Wang, o Qiuhong Wang, n,q
Siguang Wang, h WeiWang, d Xiuli Wang, m Zhou Wang, a,b
Mengmeng Wu, d Shiyong Wu, c Weihao Wu, a Jingkai Xia, a Mengjiao Xiao, l,r
Pengwei Xie, e Binbin Yan, a Jijun Yang, a Yong Yang, a Chunxu Yu, f Jumin Yuan, o Ying Yuan, a Jianfeng Yue, c Xinning Zeng, a Dan Zhang, l Tao Zhang, a,b
Li Zhao, a,b
Qibin Zheng, k Jifang Zhou, c Ning Zhou, a and XiaopengZhou i, a INPAC and School of Physics and Astronomy, Shanghai Jiao Tong University, MOE Key Labfor Particle Physics, Astrophysics and Cosmology, Shanghai Key Laboratory for Particle Physicsand Cosmology, Shanghai 200240, China b Shanghai Jiao Tong University Sichuan Research Institute, Chengdu 610213, China c Yalong River Hydropower Development Company, Ltd., 288 Shuanglin Road, Chengdu 610051,China d School of Physics, Sun Yat-Sen University, Guangzhou 510275, China e Tsung-Dao Lee Institute, Shanghai 200240, China f School of Physics, Nankai University, Tianjin 300071, China g Key Laboratory of Nuclear Physics and Ion-beam Application (MOE), Institute of Modern Physics,Fudan University, Shanghai 200433, China h School of Physics, Peking University, Beijing 100871, China i School of Physics, Beihang University, Beijing 100191, China j International Research Center for Nuclei and Particles in the Cosmos & Beijing Key Laboratoryof Advanced Nuclear Materials and Physics, Beihang University, Beijing 100191, China k School of Medical Instrument and Food Engineering, University of Shanghai for Science andTechnology, Shanghai 200093, China l Department of Physics, University of Maryland, College Park, Maryland 20742, USA Corresponding author Spokesperson Corresponding author a r X i v : . [ phy s i c s . i n s - d e t ] A ug School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China n Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201800, China o School of Physics and Key Laboratory of Particle Physics and Particle Irradiation (MOE), Shan-dong University, Jinan 250100, China p Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201210, China q University of Chinese Academy of Sciences, Beijing 100049, China r Center for High Energy Physics, Peking University, Beijing 100871, China
E-mail: [email protected]; zhou [email protected]
Abstract : We have developed a low-energy electron recoil (ER) calibration method with
Rn for the PandaX-II detector.
Rn, emanated from natural thorium compounds,was fed into the detector through the xenon purification system. From 2017 to 2019, weperformed three dedicated calibration campaigns with different radon sources. We studiedthe detector response to α , β , and γ particles with focus on low energy ER events. Duringthe runs in 2017 and 2018, the amount of radioactivity of Rn were on the order of 1%of that of
Rn and thorium particulate contamination was negligible, especially in 2018.We also measured the background contribution from
Pb for the first time in PandaX-IIwith the help from a
Rn injection. ontents α Spectrum 74.3 Low Energy β Spectrum 84.4 Delayed Coincidence Events 94.5
Pb measurement from
Rn Injection 10
While the existence of dark matter in the Universe is firmly established by cosmologicalobservations [1], the possible particle nature of dark matter is still unknown and widelystudied [2–4]. Direct detection experiments examine the possible scattering of galactic darkmatter particles on ordinary matter in detectors. So far the most sensitive direct detec-tion results, specifically for Weakly Interacting Massive Particles (WIMPs) at mass rangearound 100 GeV, are given by dual-phase xenon Time Projection Chambers (TPC) [5–7].Among them, the PandaX-II detector was operated from 2015 to 2019 in China Jin-Pingunderground Laboratory (CJPL-I) [8] to search for WIMPs as well as other dark mattercandidates and neutrinoless double β decay [9–14]. The PandaX-II detector had an activemass of 580 kg of xenon in a cylindrical field cage with a diameter of 646 mm and height of600 mm. The top and bottom ends of the TPC were instrumented with 110 Hamamatsu-R11410 3-inch photomultiplier tubes (PMTs). Detailed description of the detector andsupporting structures can be found in [15].A dual-phase xenon TPC excels in discriminating possible WIMP signals from back-ground events with detector response. A scattering event gives a prompt scintillation signal(usually called S
1) and a delayed electroluminescence signal ( S
2) when drifted ionizationelectron enters a stronger electric field in the gas phase. WIMP would most likely scatterwith xenon nuclei and trigger a nuclear recoil (NR) signal. On the contrary, electron recoil(ER) events are predominantly from β and γ particles interacting with electrons in xenonatoms. NR and ER events differ from the relative amplitude of their corresponding S S S /S
1) and therefore ER backgrounds can be effectively rejected.– 1 –alibration with an internal β source provides an uniformly distributed low-energyER events, which help determine a detector-specific ER band in the ( S S /S
1) phasespace. Multiple types of sources, including CH T [16],
Th-based
Rn [17], Ar [18],
Xe [19], and
Kr [20], have been developed by several collaborations for this purpose.With a few multi-ton scale dual phase xenon TPCs under construction [21–23], internalcalibrations with gaseous sources are becoming even more essential.In this paper, we report the results from three internal calibration campaigns withradon sources at PandaX-II. Three different sources were inserted into the xenon onlinepurification loop and emanated
Rn reached the TPC under the influence of gas flow. β decay of the daughter nucleus Pb gave low energy ER events that were uniformlydistributed in the detector volume. A large sample of α events from Rn and
Pooffered calibration at MeV scale, which is particularly important for future large scalemulti-purpose TPCs. We collected about ten thousand ER events in the dark matter searchregion of interest (ROI) from 0 to 20 keV ee . A data-driven ER distribution model in the( S S /S
1) phase space was constructed based on those events and used for dark matterphysics analysis [24]. The temporal event rate evolution and contaminations of radoninjection were studied with high energy α events, low energy β events, and β − α coincidenceevents. We also presented a new method to calculate the Pb activity from measuring
Po-
Pb-
Bi activity ratio, which is essential for the ER background estimation oflarge scale noble gas detectors.
Rn sources made from natural thorium compounds were used in PandaX-II.
Rn isa noble gas isotope with a half-life of 55 s and part of
Th decay chain.
Th has anatural isotopic abundance of 99.98% with a long half-life of 1.4 billion years. Possiblechemical processing would break the secular equilibrium of the
Th decay chain butkeep its fast-decaying daughter
Th, which decays to
Ra and then
Rn.
Rn gasemanates from the solid and may be injected into the detector for calibration.
Rn and itsdaughters decay to insignificant level within a couple of days after injection and thus a goodchoice for internal calibration. However, the main concerns for natural thorium sources arecontaminations of
Rn and particulate thorium compounds.
Rn originates from
Thand takes weeks to decay away, and the latter introduces permanent backgrounds to thedetector.The
Rn emanation rate varies for different thorium sources, depending on the me-chanical structure, gas flow rate, and arrangement inside the gas loop. The relative ac-tivities of emanated
Rn and
Th is measured as an indicator of emanation efficiency.We arranged the sources in a source chamber and flushed it with boil-off nitrogen gas ata flow rate of 2 standard liter per minute (SLPM) and activities of
Rn is measured bya commercial RAD7 radon detector [25] downstream. The radioactivity of
Th is mea-sured by its characteristic γ peaks with a High Purity Germanium (HPGe) detector [26].We measured activities of emanated Rn by RAD7 and also by a cold-trap system for– 2 – a) (b) (c)
Figure 1 : Thoriated tungsten electrodes (a) and lantern mantles (b) are shown in thesource chambers. Figure (c) shows the pipe male union where resin source were coatedinside.better accuracy. The three radon sources and corresponding measurements are describedas follows.Thoriated tungsten electrode with 1% of thorium was used in the calibration campaignof 2017. The diameter of the electrodes is about 1 mm and the length 150 mm. A
Thactivity of 48.0 ± Rn activity was less than 3 . Rnand
Rn activities were 1.22 ± . ± .
01 Bq/kg respectively.Lantern mantles treated with thorium nitrate (Th(NO ) ) were used as the radonsources in the 2018 campaign. Commercial lantern mantles made from ramie fibers wereused for their porous texture. They were soaked in Th(NO ) solution and thoroughlydried for further usage. The Th activity was measured to be 343 ± Rn less than 0 .
58 Bq/piece. Ten pieces of lantern mantles were put into a 0.62-literchamber for the emanation measurement. A 0.5 µ m particulate filter was added after thechamber to prevent any fiber from reaching RAD7. With a continuous flushing of 24 hours,we measured the emanated Rn and
Rn activities to be 1 . ± .
01 and 0 . ± .
01 Bqper piece, respectively.We conducted an additional
Rn emanation measurement with a cold-trap collectionsystem. Instead of a direct measurement with RAD7, the source chamber was connectedto a liquid nitrogen cold trap, where emanated
Rn and
Rn were freezed on the innersurface.
Rn would immediately decay while
Rn accumulated on the surface. Thecollection process in the cold-trap lasted for 6 hours and then freezed gas was transferredto a silicon photodiode detector setup for measurement. The freeze-in efficiency of the coldtrap and measurement efficiency of the silicon photodiode were determined by dedicatedcalibrations using a standard
Rn source. The silicon photodiode measured the temporalevolution of
Po and
Po activities during a 300-hour long measurement. The fitted
Rn activity was 3 . ± . Rn might introduce bias to the measurement of low rate of
Rn. It was also– 3 – igure 2 : The gaseous source delivery system (in red) is shown with a simplified schematicof the PandaX-II detector and the xenon circulation and purification system [15].worth noting that we packed 60 pieces of lantern mantles for the cold trap measurement,which might negatively impact the emanation rates.The source in the 2019 campaign emanated large amounts of
Rn and
Rn isotopesand were used to study event rates of
Pb and its parents. The source was fabricatedby diffusing thorium nitrate into a thin layer of resin coated in a commercial pipe fitting,as seen in Figure 1c. Two 3 nm-rated filters were connected at the ends of the fitting.Measured by the RAD7 detector, the
Rn (
Rn) activity was 300 ± ±
2) Bq. Thecalibration was done right before the end of PandaX-II operation and therefore the high
Rn activity was not a concern.
We inserted radon sources into the PandaX-II circulation gas loop with an added set ofdelivery pipeline. The central component of the system was a small chamber to containthe sources. The auxiliary structure included a mass flow controller (MFC), a vacuumpump, a couple of bypass pipelines, valves, and filters, as shown in Figure 2. The sourcechamber was a stainless steel cylinder with standard Conflat flanges on the two ends. Thenominal diameter was 63 mm and length 200 mm. The Conflat flanges were coupled to thepipeline with 1/4-inch VCR fittings. All chambers, flanges, and pipes were electrochemi-cally polished to reduce possible
Rn emanation. A 0.4 µ m-rated gasket-type filter and a3 nm-rated filter were mounted downstream of the source chamber to prevent radioactiveparticulates from entering the gas loop. A 0.4 µ m-rated filter upstream was also installedto avoid any negative impact when the gas flow might have been paused.– 4 –e mounted the setup in parallel with the existing circulation loops (Figure 2). ThePandaX-II detector was connected to two parallel circulation loops, in which xenon gaswas continuously purified through high temperature zirconium getters at a typical flowrate of 40 to 60 SLPM. Liquid xenon in the detector was extracted and warmed up togas phase before purification. Purified gas was then liquefied by a heat exchanger anda dedicated cooling module before fed into the TPC again. For calibration campaigns,the source module was placed upstream of the getter on loop 2 to minimize the impactof possible impurity gas emanated from the Th source. The amount of radon reachingthe TPC depended on the flow rate through the source as well as the overall flow rate ofcirculation xenon gas. During the three calibration campaigns, the gas flow rate throughthe source chamber was about 20, 4, and 1 SLPM respectively.
We performed three radon calibration campaigns from 2017 to 2019. Detector performancewas carefully monitored during the source injections. Initial trial injections in 2017 causeddrift electron lifetime to decrease to about 100 µ s in several runs but it quickly recov-ered during normal calibrations. The flow rate through the source chambers in the othertwo campaigns were much smaller than that of 2017 and therefore negligible impact wasintroduced to the TPC.Performance of data acquisition (DAQ) system was monitored and tuned for highertrigger rates during calibration. The DAQ system of PandaX-II saturated with a detectortrigger rate at about 50 Hz. During the calibration campaign in 2017, the average eventrate was only around 15 Hz with dark matter data taking trigger settings. In 2018, weimplemented a new set of calibration trigger to reject events with more than 70 firedphotomultipliers and thus suppressed the high energy events. The trigger rates were around32 Hz during our calibration runs. To measure the event rates of α s, we took data for 43minutes with a random trigger of about 120 Hz and each triggered events recorded 1 µ s-longwaveforms of all PMTs. The results will be discussed later.For the calibrations in 2017 and 2018, we characterized the event rate changes of Rnand its daughters in the detector over time, and examined possible permanent
Th con-taminations and
Rn contaminations during and after the injection period. The resultsare presented in subsection 4.1 and detailed event selection procedures described in sub-section 4.2 to 4.4. For the final radon calibration in May 2019, the emphasis was on
Rn.About 1 Bq of
Rn was accumulated in detector in two days before we stopped injection.We were able to collect a large sample of
Pb and
Bi-
Po coincidence events foranother 5 days after injection. The ratio of event rates of
Po and
Bi-
Po is reportedin subsection 4.5.
Event rates during calibration were used to characterize the accumulation, equilibration,and decay of the gaseous source in the TPC. We measured rates from high energy α eventsfrom Rn and
Po, low energy β events from Pb, and β and α coincidence events– 5 –ampaigns 2017 2018 2019Thoriated Source Tungsten electrodes Lantern mantles Coated resinDuration [days] 35 20 2.1Live Time [days] 18.9 11.9 1.3Injected Rn [Bq] 2 . ± .
001 31 . ± . . ± . Rn [Bq] (1 . ± . × − . ± .
04 6 . ± . Rn from thoriumparticulate [ µ Bq/kg] 0.20 ± ± Table 1 : The activity of injected
Rn and
Rn were calculated.
Rn from thoriumparticulate refers to the possible contaminations induced by thorium compound particulatereached the TPC active volume. The value was measured by the difference between
Rnactivity before and after the radon calibration campaigns. The activities after injectionswere measured after the injected
Rn had decayed to insignificant levels. - - - - -
10 110 E v e n t r a t e [ H z ] Rn Po Pb Po Bi -
Figure 3 : The event rate evolution of
Rn and daughters during the 2017 calibration.The vertical error bar refers to the statistical uncertainty and the horizontal bar representsthe run duration. Event rates of
Rn and
Po are from α s, Pb from events in theenergy range of [0, 600] keV ee , and Bi-
Po from the coincidence algorithm in 4.4.Spikes in the event rates in November are due to test injections.from
Bi and
Po respectively. In Figure 3, we showed the temporal evolution of therates during the 2017 calibration campaign. For 2017, the duration of source injectioncampaign lasted 35 days, during which we were able to accumulate 16.9 days of live datawith stable α event rates. We performed a few trial injections in late November 2017 andwe could see α event rate spikes in the figure. Rn rate increased by 2 . ± .
001 Hzduring stable source injection comparing with background data before injection. The event– 6 – .2 5.4 5.6 5.8 6 6.2 6.4 6.6 6.8 7 7.2Energy [MeV] C oun t s [ p e r k e V ] dataRn Po Bi Rn Po Flat backgroundTotal
Figure 4 : High energy region of the spectrum for calibration data in 2017.
Rn (5490keV),
Po (6002 keV),
Bi (6090 keV),
Rn (6288 keV), and
Po (6778 keV) peaksare fitted with Gaussian functions and drawn in blue, green, yellow, red, and orange,respectively. A flat background is added in the fitting, as shown in gray. The right-handsides of the
Rn and
Po peaks deviate from fitting due to response distortion for S Rn was also introduced into the TPC by the injection. Theratio between event rates of injected
Rn and
Rn for 2017 (2018) was 0 . ± .
01 %(0 . ± .
12 %). The small ratios here meant that introduced
Rn had only a manageablecontribution to dark matter data taking.The other major concern for the radon sources was permanent contamination by par-ticulate thorium compound infiltration. Due to the long half-life,
Th and daughterisotopes would slowly but steadily release α , β , and γ background from inside of the TPC.We quantified the effect by comparing Rn event rate differences before and after thecalibration campaigns. The measured activities, normalized by fiducial volume, are givenin Table 1. The result from the 2018 campaign was consistent with zero and showed thatno significant amount of thorium particulate was introduced to the detector.We stopped PandaX-II operation after the radon calibration campaign in 2019, andno measurement was done about particulate contaminations. α Spectrum
High energy α events from Rn,
Rn, and their progenies were selected with the z-coordinate cut to exclude events from top and bottom electrodes. The maximum drifttime ∆ t = 360 µs corresponded to the full height of the TPC, and the z-coordinate cutwas defined that ∆ t ∈ [10 , µs . A trigger cut was required to reject α events with a– 7 –
100 200 300 400 500 600] ee Energy [keV0246810121416 - · / s ] ee R a t e [ c oun t s / e V CalibrationPhysics run
Figure 5 : Energy spectra in the range of 0 to 600 keV ee during calibration (red) andphysics runs (black) in 2017. For the physics (calibration) runs, a total of 427 (406) hours’data were used.foregoing coincidence β . The signal size of S , β and other low energy events. We also compared the amount of charge collectedby the top and bottom PMT array, the ratio of which depended on the z-coordinate.Therefore we removed events with charge ratio mis-matched with z-coordinates. Bulk α events were selected with 3D mapping corrected (see below) S , S > α events close to the edge of the field cage (called surface events)were rejected by requiring S > α s was reconstructed with scintillation S S S S Rn-
Pocoincidence events. S S α peak energies and the resulting light yield was L y = 8 . ± .
10 PE/keV. We fitted thepeaks with Gaussian functions and a flat background in Figure 4. Fitted counts were usedto report event rates of radon isotopes during the calibration runs. Deviation from Gaus-sian were seen on the right-handed sides of the
Rn and
Po peaks. We confirmed thedeviations were caused by events nearby the bottom of the TPC, where detector responseto charges of large S β Spectrum
We monitored the event rate change of the low energy region in the range of [0 , ee During the calibration runs, this part of the spectrum was dominated by
Pb events,which were typical multi-scattering events of β and γ particles from the Pb decay.– 8 – ee Energy [keV020406080100120140160180 C oun t s [ p e r . k e V ]
20 40 60 80 100 120 140S1 [PE]11.522.53 ( S / S ) l og Figure 6 : Reconstructed energy spectrum (left) and ER band (right) obtained from thecalibrations run in 2018. Events were selected with identical fiducial volume cuts and pulsequality cuts as in Ref. [24].The electron equivalent recoil energy is reconstructed as the linear combination of the S S β spectra before and during the calibration runs of 2017 in the volume that ∆ t ∈ [20 , µs and x + y <
720 cm . The β and the subsequent 238 keV γ were captured by the TPCalmost simultaneously and the spectrum clearly shows the signature.Low energy electron recoil data from 2018 calibration were used for the ER model inthe PandaX-II full-exposure analysis [24]. Besides fiducial volume cuts and quality cuts,we introduced an additional quality cut to reject events with excessive noise in pulses,which was also adopted in the dark matter analysis. In the range of S ∈ [3 , ± Pb and
Rn and took intoaccount an energy acceptance of 0.8% in [0 ,
20] keV for
Pb events. The small
Pb ratecould be attributed partly to the mobility of charged ions drifting towards the cathode andout of the fiducial volume [27, 28].
Bi-
Po events in Figure 3 refer to the coincidence events of a β from Bi end point2252 keV) and an α from Po (8784 keV), with a typical delay time scale of 0 . µ s. Thecoincidence events had a typical double- S α sand thus could be easily identified.Selection cuts were implemented specifically for coincidence events and correspondingefficiencies were calculated. The S Bi β was firstly selected to have energy inthe range of 50 to 3000 keV ee , which introduced a 98.1% efficiency [29]. The events wereselected in the bulk volume of the detector. Electrode events were rejected by requiring– 9 – .5 1 1.5 2 2.5 3s] m [ S1 t D s ] m C oun t s [ p e r . s m – = 0.3045 t Figure 7 : The time difference between decay events of
Bi and
Po. Deviation fromexponential with smaller ∆ t S is due to the inefficiency of separating two S t ∈ [10 , µs . Fiducial cuts as well as α -specific surface event cuts were appliedto selected events in the bulk of the detector with the same cut values as mentionedpreviously. S β and α events are often overlapped since their width is on thescale of microseconds. Therefore, the position of the events was reconstructed from thetiming difference between S S β . Random coincidence events were rejected by thetop-bottom charge ratio cut, which checked the correlation between the ratio and eventz-coordinate. The timing difference between S β and α ∆ t S of events survivedall cuts is plotted in Figure 7. An exponential decay curve could fit data with ∆ t S largerthan 0 . µ s. The curve deviated from exponential at smaller ∆ t S when S α and S β couldnot be effectively separated. The efficiency of this effect was estimated to be 75%. Pb measurement from
Rn Injection
A common difficulty in the spectrum fitting in dark matter searching data was the contri-bution from
Pb in the low energy range. While the radioactivities of ancestors on thedecay chain, such as
Rn, can be estimated from high energy events, it cannot be used toestimate
Pb activity reliably. A decrease of activities when compare parent to daughterisotopes is often observed and we call this phenomena chain depletion. A quantitativeevaluation of chain depletion is desirable to better constrain the
Pb contribution in thelow energy spectrum.The thoriated resin calibration source used in the 2019 campaign injected a largeamount of
Rn in the TPC and provided us a rare opportunity to study the chain deple-tion. In the first week after calibration was stopped,
Rn and its daughters quickly de-cayed to insignificant levels and
Pb events dominated the energy region up to 1000 keV ee .The activity of Pb in the TPC was determined based on event rates in the energy range– 10 – - -
10 1 E v e n t r a t e [ H z ] Po Po Bi- Pb Figure 8 : Event rate evolution of
Pb and parent isotopes in 2019. The starting timein the horizontal axis is about 35 hours after the source injection stopped.
Po rates arecalculated from α s and Bi-
Po rates from the coincidence algorithm.
Pb rates arecalculated with events in the range of [10,200] keV ee . Background is subtracted for all ratecalculations.[10, 200] keV ee . After subtracting backgrounds, the event rate was then divided by 4.3%,the energy acceptance. Delayed coincidence events of Bi-
Po were selected by the cutsoutlined in the previous subsection. The event rates of
Po,
Pb, and
Bi-
Po areas shown in Figure 8 as a function of time. The chain depletion ratio between
Pb and
Po was determined to be 36 . ± . Bi-
Po coincidence eventsand
Pb was 43 . ± . Pb based on measured α rate of Po or
Bi-
Po coincidence rate in the dark matter data. This new procedure was applied inthe PandaX-II final dark matter analysis in Ref. [24].
Internal calibration sources are of increasing importance with the larger liquid xenon de-tectors being commissioned and planned. We investigated three
Rn emanation sourcesbased on natural thorium compounds. All the sources have been tested in the PandaX-IIdetector. By flushing thoriated tungsten electrodes and lantern mantles, we were able tointroduce a significant amount of
Rn without serious
Rn contaminations. We alsoobserved no thorium particulate infiltration in the TPC with lantern mantles, which isespecially important to avoid introducing any permanent contaminations. The two sourcesare made of commercial products and can be easily prepared. A large sample of low energyER calibration events, mainly from
Pb, were collected and used to construct a data-driven ER response model for the dark matter searches of the PandaX-II. The thoriated– 11 –esin source were used mainly as a
Rn source to measure the activity ratios between
Pb and its parent isotopes. This allows a robust estimate of
Rn-induced backgroundin the low energy dark matter search region.We are investigating the possibility to use similar natural thorium compound sourcesfor the PandaX-4T detector, which is currently under construction. Even by conservativelyassuming that the ratio between
Rn and
Rn for PandaX-4T calibration is the sameas what we observed in 2018, it would take around 25 days of waiting time before we couldresume dark matter data taking. However, we envision that calibrations for other purposes,such as neutron calibrations, can be performed during the
Rn cooling-off period. Withthis strategy, we can minimize the impact to overall physics run time.
Acknowledgement
We thank Shoukang Qiu and Quan Tang from University of South China for helping fab-ricate the thoriated resin source. This project is supported by grants from the Ministry ofScience and Technology of China (No. 2016YFA0400301 and 2016YFA0400302), a DoubleTop-class grant from Shanghai Jiao Tong University, grants from National Science Founda-tion of China (Nos. 11435008, 11505112, 11525522, 11775142 and 11755001), grants fromthe Office of Science and Technology, Shanghai Municipal Government (Nos. 11DZ2260700,16DZ2260200, and 18JC1410200), and the support from the Key Laboratory for ParticlePhysics, Astrophysics and Cosmology, Ministry of Education. This work is supported alsoby the Chinese Academy of Sciences Center for Excellence in Particle Physics (CCEPP)and Hongwen Foundation in Hong Kong. Finally, we thank the CJPL administration andthe Yalong River Hydropower Development Company Ltd for indispensable logistics andother supports.
References [1] P.A.R. Ade et al. ,, Planck 2015 results. XIII. Cosmological parameters.
Astron. Astrophys. ,594:A13, 2016.[2] M. Schumann, Direct Detection of WIMP Dark Matter: Concepts and Status.
J. Phys. G ,46(10):103003, 2019.[3] J. Liu, X. Chen, and X. Ji, Current status of direct dark matter detection experiments.
Nature Phys. , 13(3):212–216, 2017.[4] L. Baudis, Direct dark matter detection: the next decade.
Phys. Dark Univ. , 1:94–108, 2012.[5] E. Aprile et al. , Dark Matter Search Results from a One Ton-Year Exposure of XENON1T.
Phys. Rev. Lett. , 121(11):111302, 2018.[6] X. Cui et al. , Dark Matter Results From 54-Ton-Day Exposure of PandaX-II Experiment.
Phys. Rev. Lett. , 119(18):181302, 2017.[7] D.S. Akerib et al. , Results from a search for dark matter in the complete LUX exposure.
Phys. Rev. Lett. , 118(2):021303, 2017.[8] K.J. Kang et al. , Status and prospects of a deep underground laboratory in China.
J. Phys.Conf. Ser. , 203:012028, 2010. – 12 –
9] K. Ni et al. , Searching for neutrino-less double beta decay of
Xe with PandaX-II liquidxenon detector.
Chin. Phys. C , 43(11):113001, 2019.[10] C. Fu et al. , Spin-Dependent Weakly-Interacting-Massive-Particle–Nucleon Cross SectionLimits from First Data of PandaX-II Experiment.
Phys. Rev. Lett. , 118(7):071301, 2017.[Erratum: Phys.Rev.Lett. 120, 049902 (2018)].[11] C. Fu et al. , Limits on Axion Couplings from the First 80 Days of Data of the PandaX-IIExperiment.
Phys. Rev. Lett. , 119(18):181806, 2017.[12] X. Ren et al. , Constraining Dark Matter Models with a Light Mediator at the PandaX-IIExperiment.
Phys. Rev. Lett. , 121(2):021304, 2018.[13] J. Xia et al. , PandaX-II Constraints on Spin-Dependent WIMP-Nucleon EffectiveInteractions.
Phys. Lett. B , 792:193–198, 2019.[14] X. Chen et al. , Exploring the dark matter inelastic frontier with 79.6 days of PandaX-IIdata.
Phys. Rev. D , 96(10):102007, 2017.[15] A. Tan,
Pandax-II Dark Matter Detector and its First Results . PhD thesis, Maryland U.,2019.[16] D.S. Akerib et al. , Tritium calibration of the LUX dark matter experiment.
Phys. Rev. D ,93(7):072009, 2016.[17] E. Aprile et al. , Results from a Calibration of XENON100 Using a Source of DissolvedRadon-220.
Phys. Rev. D , 95(7):072008, 2017.[18] E.M. Boulton et al. , Calibration of a two-phase xenon time projection chamber with a Arsource.
JINST , 12(08):P08004, 2017.[19] D.S. Akerib et al. , Ultralow energy calibration of LUX detector using
Xe electron capture.
Phys. Rev. D , 96(11):112011, 2017.[20] D.S. Akerib et al. , Kr calibration of the 2013 LUX dark matter search.
Phys. Rev. D ,96(11):112009, 2017.[21] D.S. Akerib et al. , The LUX-ZEPLIN (LZ) Experiment.
Nucl. Instrum. Meth. A ,953:163047, 2020.[22] E. Aprile et al. , Projected WIMP Sensitivity of the XENONnT Dark Matter Experiment.ArXiv:2007.08796 [physics.ins-det], 2020.[23] H. Zhang et al. , Dark matter direct search sensitivity of the PandaX-4T experiment.
Sci.China Phys. Mech. Astron. , 62(3):31011, 2019.[24] Q. Wang et al. , Results of Dark Matter Search using the Full PandaX-II Exposure.ArXiv:2007.15469 [astro-ph], 2020.[25] DURRIDGE Company Inc,
RAD7 Electronic Radon Detector User Manual .[26] X. Wang, et al. , Material Screening with HPGe Counting Station for PandaX Experiment.
JINST , 11(12):T12002, 2016.[27] J.B. Albert et al. , Measurements of the ion fraction and mobility of α - and β -decay productsin liquid xenon using the EXO-200 detector. Phys. Rev. C , 92(4):045504, 2015.[28] E. Aprile et al. , Intrinsic backgrounds from Rn and Kr in the XENON100 experiment.
Eur.Phys. J. C , 78(2):132, 2018. – 13 –
29] S. Li et al. , Krypton and radon background in the PandaX-I dark matter experiment.
JINST , 12(02):T02002, 2017., 12(02):T02002, 2017.