Search for Light Dark Matter-Electron Scatterings in the PandaX-II Experiment
Chen Cheng, Pengwei Xie, Abdusalam Abdukerim, Wei Chen, Xun Chen, Yunhua Chen, Xiangyi Cui, Yingjie Fan, Deqing Fang, Changbo Fu, Mengting Fu, Lisheng Geng, Karl Giboni, Linhui Gu, Xuyuan Guo, Ke Han, Changda He, Di Huang, Yan Huang, Yanlin Huang, Zhou Huang, Xiangdong Ji, Yonglin Ju, Shuaijie Li, Qing Lin, Huaxuan Liu, Jianglai Liu, Liqiang Liu, Xiaoying Lu, Wenbo Ma, Yugang Ma, Yajun Mao, Yue Meng, Nasir Shaheed, Kaixiang Ni, Jinhua Ning, Xuyang Ning, Xiangxiang Ren, Changsong Shang, Guofang Shen, Lin Si, 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, Xiang Xiao, Binbin Yan, Jijun Yang, Yong Yang, Chunxu Yu, Jumin Yuan, Ying Yuan, Xinning Zeng, Dan Zhang, Tao Zhang, Li Zhao, Qibin Zheng, Jifang Zhou, Ning Zhou, Xiaopeng Zhou
SSearch for Light Dark Matter-Electron Scatterings in the PandaX-II Experiment
Chen Cheng, Pengwei Xie, Abdusalam Abdukerim, Wei Chen, Xun Chen,
1, 4
Yunhua Chen, XiangyiCui, Yingjie Fan, Deqing Fang, Changbo Fu, Mengting Fu, Lisheng Geng,
9, 10
Karl Giboni, LinhuiGu, Xuyuan Guo, Ke Han, Changda He, Shengming He, Di Huang, Yan Huang, Yanlin Huang, Zhou Huang, Xiangdong Ji, Yonglin Ju, Shuaijie Li, Qing Lin,
14, 15
Huaxuan Liu, Jianglai Liu,
1, 3, 4, ∗ Xiaoying Lu,
16, 17
Wenbo Ma, Yugang Ma, Yajun Mao, Yue Meng,
1, 4, † Nasir Shaheed,
16, 17
KaixiangNi, Jinhua Ning, Xuyang Ning, Xiangxiang Ren,
16, 17
Changsong Shang, Guofang Shen, Lin Si, AndiTan, Anqing Wang,
16, 17
Hongwei Wang, Meng Wang,
16, 17
Qiuhong Wang, Siguang Wang, Wei Wang, Xiuli Wang, Zhou Wang,
1, 4
Mengmeng Wu, Shiyong Wu, Weihao Wu, Jingkai Xia, Mengjiao Xiao, Xiang Xiao, Binbin Yan, Jijun Yang, Yong Yang, Chunxu Yu, Jumin Yuan,
16, 17
Ying Yuan, XinningZeng, Dan Zhang, Tao Zhang, Li Zhao, Qibin Zheng, Jifang Zhou, Ning Zhou, and Xiaopeng Zhou (PandaX-II Collaboration) School of Physics and Astronomy, Shanghai Jiao Tong University, MOE Key Laboratory for Particle Astrophysicsand Cosmology, Shanghai Key Laboratory for Particle Physics and Cosmology, Shanghai 200240, China School of Physics, Sun Yat-Sen University, Guangzhou 510275, China Tsung-Dao Lee Institute, Shanghai Jiao Tong University, Shanghai, 200240, China Shanghai Jiao Tong University Sichuan Research Institute, Chengdu 610213, China Yalong River Hydropower Development Company, Ltd., 288 Shuanglin Road, Chengdu 610051, China School of Physics, Nankai University, Tianjin 300071, China Key Laboratory of Nuclear Physics and Ion-beam Application (MOE),Institute of Modern Physics, Fudan University, Shanghai 200433, China School of Physics, Peking University, Beijing 100871, China School of Physics, Beihang University, Beijing 100191, China International Research Center for Nuclei and Particles in the Cosmos & Beijing Key Laboratoryof Advanced Nuclear Materials and Physics, Beihang University, Beijing 100191, China School of Medical Instrument and Food Engineering, Universityof Shanghai for Science and Technology, Shanghai 200093, China Department of Physics, University of Maryland, College Park, Maryland 20742, USA School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China State Key Laboratory of Particle Detection and Electronics,University of Science and Technology of China, Hefei 230026, China Department of Modern Physics, University of Science and Technology of China, Hefei 230026, China Research Center for Particle Science and Technology, Institute of Frontier andInterdisciplinary Scienc, Shandong University, Qingdao 266237, Shandong, China Key Laboratory of Particle Physics and Particle Irradiation of Ministryof Education, Shandong University, Qingdao 266237, Shandong, China Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201210, China (Dated: January 20, 2021)We report constraints on light dark matter through its interactions with shell electrons in thePandaX-II liquid xenon detector with a total 46.9 tonne · day exposure. To effectively search for thesevery low energy electron recoils, ionization-only signals are selected from the data. 1821 candidatesare identified within ionization signal range between 50 to 75 photoelectrons, corresponding to amean electronic recoil energy from 0.08 to 0.15 keV. The 90% C.L. exclusion limit on the scatteringcross section between the dark matter and electron is calculated based on Poisson statistics. Underthe assumption of point interaction, we provide the world’s most stringent limit within the darkmatter mass range from 15 to 30 MeV / c , with the corresponding cross section from 2 × − to3 × − cm . The existence of dark matter (DM) is established byoverwhelming evidence in cosmological and astronomi-cal observations [1]. Possible interactions between DMparticles and baryonic matter have been searched for inunderground laboratories using ultra-low background de-tectors by directly detecting recoil signals [2, 3]. DM ∗ Spokesperson: [email protected] † Corresponding author: [email protected] within a mass range between GeV / c to TeV / c havebeen mostly searched for via its elastic scattering offatomic nucleus [4–12]. The scatterings of DM within thismass range with electrons, while still possible, is difficultto be kinematically probed as the energy of electron re-coils (ERs) is suppressed by the smallness of electronmass. For sub-GeV light DM, on the other hand, thenuclear recoil energy becomes much more difficult to de-tect with conventional detection techniques. It was re-alized that these light DM can scatter with shell elec- a r X i v : . [ h e p - e x ] J a n R a t e [ e v en t s / da y ] Run 9 Run 10 Run 11
FIG. 1. Event rate of the US2 signals with charge range from 50 to 200 PE trons, which may subsequently produce sufficiently largeionization signals in the detector [13]. Such DM-electronscatterings open up a new experimental paradigm, whichhas since been pursued by numbers of groups [14–19].The PandaX-II experiment [4, 20–29], located in theChina Jinping Underground Laboratory (CJPL), utilizesa dual-phase Time Projection Chamber (TPC), whichcontains an active cylindrical target with 580 kg of liq-uid xenon (LXe). DM scattering produces prompt scin-tillation photons ( S
1) in the liquid region, and ionizedelectron signal ( S
2) through delayed electroluminescencein the gaseous region. Both S S S S S S Rn, and
Am-Be sources areused in this analysis. Most S S µ s window preceding an US2, anda charge ratio cut in the top and bottom PMTs. Twocuts are tightened, including the full-width-10%-maximaof US2 and the rising edge of US2 (defined as the ratioof the charge in the first 1 µ s to the total) based on thedistribution of the ER calibration data. Events are fur-ther selected within a reconstructed radius of 15 cm fromthe central axis for three DM search runs, leading to a117 kg fiducial mass of LXe. The time evolution of thesurviving candidates in the three DM search runs withcharge range from 50 to 200 photoelectron (PE) is shown in Fig. 1, and appears to be reasonably stable within atotal time span of two years.The combined event selection efficiency for US2 eventsis a product of the trigger efficiency and data quality cutefficiency. PandaX-II utilizes an FPGA-based realtimetrigger system, with its efficiency directly measured from
20 40 60 80 100 120 140 160 180Ionization Signal [PE]0.10.20.30.40.50.60.70.80.9 E ff i c i en cy Run 9TriggerQuality cut (tritium)Am-Be)
Quality cut (Combined event selection E ff i c i en cy Runs 10/11TriggerQuality cut (tritium)Am-Be)
Quality cut ( Rn)
Quality cut (Combined event selection
FIG. 2. Data quality cut efficiency (red), trigger efficiency(blue) and combined efficiency (black) vs. detected ion-ization signals in the Run 9 (top) and Runs 10/11 (bot-tom). The uncertainties are obtained from tritium calibra-tion data. The quality cut efficiencies (dashed) from
Am-Be and
Rn calibration runs are used to validate thatfrom tritium. The upper axes correspond to mean ER en-ergy computed using the constant model ( f e = 0 .
83 and E = 13 . × S / SEG / EEE /f e , see Fig. 3). events with multiple S
2s [30]. The data quality cut ef-ficiency is obtained from the tritium calibration run bythe product of each individual cut efficiency. Each cut ef-ficiency is estimated by the ratio of survived events withall cuts to those with all-but-this-cut. The trigger, dataquality, and the combined efficiencies are shown in Fig. 2for US2 events with charge range between 20 to 200 PE.For comparison, data quality efficiency curves obtainedusing
Am-Be (Runs 9/10/11) and
Rn (Runs 10/11)data are overlaid, which are agrees with that from tritiumdata within the uncertainty band. ER E1020304050607080 e n Run 9NEST 2.0 modelPandaX-II modelConstant model ER E010203040506070 e n Runs 10/11NEST 2.0 modelPandaX-II modelConstant model
FIG. 3. Number of primary ionized electrons vs. electronrecoil energy for three DM search runs for NEST 2.0 [31, 32],the constant model [33], and the PandaX-II model [34], withuncertainty bands obtained from original publications.
As mentioned in the introduction, the light DM-electron scatterings produce sub-keV recoil energy, there-fore knowledge of the photon and ionization productionsin LXe in this energy regime is required. Three indepen-dent signal response models are compared, all under thestandard W value of 13.7 eV to produce either a pho-ton or electron [35]: 1) the Noble Element SimulationTechnique (NEST 2.0) model [31, 32], 2) the constantmodel in which the fraction of ionization to total quantais f e = 0 .
83 [33] with no energy dependence and with-out recombination effects, and 3) the PandaX-II model,obtained by fitting the tritium calibration data with cor-related S S e vs. recoil energy is shown in Fig. 3for Run 9 and Runs 10/11 under a drift field of 400 and 318 V/cm, respectively. In general, the constant modelpredicts less n e in comparison with NEST 2.0. For Run9, the PandaX-II model agrees with other two modelswithin 1 σ at 0.9 keV. On the other hand, for Runs 10/11,the PandaX-II model agrees with the constant model,but has slight tension with NEST 2.0. Therefore, theconstant model is selected as the nominal model in thisanalysis to conservatively estimate the number of pri-mary ionized electrons.The spectrum of detected ionization signals, i.e. US2events in PE, can then be predicted based on the mea-sured detector parameters [28], listed in Table I for con-venience, and the efficiencies in Fig. 2. TABLE I. The PandaX-II detector parameters, including elec-tron extraction efficiency (EEE), single electron gain (SEG)and its measured resolution ( σ SE ) [28]. They are used to esti-mate the relation between ER energy and detected ionizationelectrons. Run 9 Run 10 Run 11EEE (%) 46 . ± . . ± . . ± . . ± . . ± . . ± . σ SE (PE) 8.3 7.8 8.1
20 40 60 80 100 120 140 160 180 200Ionization Signal [PE]110 da y ] (cid:215) t onne (cid:215) R a t e [ e v en t s / PE ROI = 1 DM F cm -39 · = 1.5 x s ) =20 MeV/c X DM (m ) =200 MeV/c X DM (mData da y ] (cid:215) t onne (cid:215) R a t e [ e v en t s / PE ROI /q m a = DM F cm -35 · = 2 x s ) =20 MeV/c X DM (m ) =200 MeV/c X DM (mData
FIG. 4. Detected ionization signals (US2, black his-tograms), and expected signals from DM-electron scatteringswith F DM =1 (upper) and α m e /q (lower), with blue (red)histogram corresponding to a DM mass of 20 MeV / c / c ). The gray shadow shows the ROI of this analysis.The excess in the data peaking at ∼
25 PE are single electronevents, likely due to stray electrons in LXe.
TABLE II. The number of US2 candidates, exposure, and known ER background events for the three DM search runs. Thespan 1 and span 2 of the Run 11 are listed separately due to the different background rates. ROI is chosen as from 50 PE and75 PE, corresponding to a mean ER energy between 0.08 to 0.15 keV. The flat ER background includes Kr,
Rn,
Rn,material ER, solar neutrino and
Xe [28].
Run 9 Run 10 Run 11 span 1 Run 11 span 2 TotalExposure [tonne · day] 9.3 9.0 28.6 46.9DM-electron candidates [events] 287 340 1194 1821Flat ER background [events] 0.8 0.2 0.3 0.6 1.8Tritium background [events] 0 0.1 0.2 0.3 0.6 The predicted rates and recoil energy spectra of DM-electron scatterings in LXe target are calculated followingthe procedures in Refs. [13, 36]. Two benchmark interac-tion models, the point-like interaction with form-factor F DM = 1, and light mediator with F DM = α m e /q , areconsidered, with their corresponding ionization cross sec-tions from xenon atoms computed. The US2 candidatesfrom the data are overlaid with the predicted ionizationspectra for the two models with DM masses of 20 and 200MeV / c , respectively, in Fig. 4. The region-of-interest(ROI) is chosen to be 50 to 75 PE to keep The lower cut isset to keep at least 50% trigger efficiency, and the highercut is set to enclose the most high energy tail for theDM benchmark DM mass of 20 MeV / c . The number ofcandidates are summarized in Table II for the three DMsearch runs, which are significantly higher than the dom-inating known ER background with approximately flatspectrum at low energy (flat ER) and the tritium contri-bution [28]. We note that in comparison to a standard S S S
1. To be conservative, we assume that alldetected candidates are DM-electron scattering eventsto derive the upper limits using Poisson statistics [37].The 90% C.L. exclusion curves of DM-electron scatteringcross section under the two benchmark interaction mod-els are shown in Fig. 5, assuming both the constant andNEST 2.0 signal response models. For comparison, theworld data are also overlaid in Fig. 5. Due to the achieved50 PE analysis threshold ( ∼ / c , with the corresponding cross section from2 × − to 3 × − cm . At 25 MeV / c , our result isa few times more constraining than that from XENON10and XENON1T which used the same xenon target andthe ionization-only channel [15, 38]. Alternative choice ofthe NEST 2.0 model would increase the ionization yieldrelative to the constant model at a given energy, lead-ing to a more constraining limit. In the near future, thePandaX-4T experiment will be under operation. Withlarger exposure, lower background and a lower thresholdusing triggerless readout [39], PandaX-4T will providemore sensitive search for DM-electron scatterings.This project is supported in part by Office of Scienceand Technology, Shanghai Municipal Government (grant No. 18JC1410200), a grant from the Ministry of Scienceand Technology of China (No. 2016YFA0400301), andgrants from National Science Foundation of China (Nos.12005131, 11905128, 12090061, 11775141). We thanksupports from Double First Class Plan of the ShanghaiJiao Tong University. We also thank the sponsorshipfrom the Chinese Academy of Sciences Center for Ex-cellence in Particle Physics (CCEPP), Hongwen Founda-tion in Hong Kong, and Tencent Foundation in China.Finally, we thank the CJPL administration and the Ya-long River Hydropower Development Company Ltd. forindispensable logistical support and other help. P a nd a X - II ( T h i s W o r k ) F DM = XE N O N T XE N O N S E N S E I D a r k S i d e - DA M I C E D E L W E I SS - - - - D M - E l e c t r on C r o ss S e c t i on [ c m ] P a nd a X - II ( T h i s W o r k ) F DM = α m e2 / q XE N O N SENSEI D a r k S i d e - D A M I C E D E L W E I S S - - - - - DM Mass [ MeV / c ] D M - E l e c t r on C r o ss S e c t i on [ c m ] FIG. 5. 90% C.L. upper limit (solid: the constant model,dashed: NEST 2.0) on DM-electron scattering cross sec-tion from PandaX-II data for F DM = 1 (upper) and F DM = α m e /q (lower). Results from XENON1T [15],XENON10 [38], DarkSide-50 [16], SENSEI [18], DAMIC [17],EDELWEISS [40] are also shown for comparison. [1] G. Bertone, D. Hooper, and J. Silk, Particle dark mat-ter: evidence, candidates and constraints, Physics Re-ports , 279 (2005).[2] J. Liu, X. Chen, and X. Ji, Current status of directdark matter detection experiments, Nature Phys. , 212(2017).[3] T. M. Undagoitia and L. Rauch, Dark matter direct-detection experiments, Journal of Physics G: Nuclear andParticle Physics , 013001 (2015).[4] A. Tan et al. (PandaX), Dark Matter Search Results fromthe Commissioning Run of PandaX-II, Phys. Rev. D ,122009 (2016).[5] E. Aprile et al. (XENON), Dark Matter Search Resultsfrom a One Ton-Year Exposure of XENON1T, Phys. Rev.Lett. , 111302 (2018).[6] D. Akerib et al. (LUX), Results from a search for darkmatter in the complete LUX exposure, Phys. Rev. Lett. , 021303 (2017).[7] P. Agnes et al. (DarkSide), DarkSide-50 532-day DarkMatter Search with Low-Radioactivity Argon, Phys. Rev.D , 102006 (2018).[8] R. Ajaj et al. (DEAP), Search for dark matter with a231-day exposure of liquid argon using DEAP-3600 atSNOLAB, Phys. Rev. D , 022004 (2019).[9] R. Agnese et al. (SuperCDMS), Search for Low-MassWeakly Interacting Massive Particles Using Voltage-Assisted Calorimetric Ionization Detection in the Super-CDMS Experiment, Phys. Rev. Lett. , 041302 (2014).[10] H. Jiang et al. (CDEX), Limits on Light Weakly Inter-acting Massive Particles from the First 102.8 kg × dayData of the CDEX-10 Experiment, Phys. Rev. Lett. ,241301 (2018).[11] A. Abdelhameed et al. (CRESST), First results from theCRESST-III low-mass dark matter program, Phys. Rev.D , 102002 (2019).[12] C. Amole et al. (PICO), Dark Matter Search Results fromthe PICO-60 C F Bubble Chamber, Phys. Rev. Lett. , 251301 (2017).[13] R. Essig, J. Mardon, and T. Volansky, Direct detection ofsub-GeV dark matter, Phys. Rev. D , 076007 (2012).[14] R. Essig, A. Manalaysay, J. Mardon, P. Sorensen, andT. Volansky, First Direct Detection Limits on Sub-GeVDark Matter from XENON10, Phys. Rev. Lett. ,021301 (2012).[15] E. Aprile et al. (XENON), Light Dark Matter Searchwith Ionization Signals in XENON1T, Phys. Rev. Lett. , 251801 (2019).[16] P. Agnes et al. (The DarkSide Collaboration), Con-straints on Sub-GeV Dark-Matter–Electron Scatteringfrom the DarkSide-50 Experiment, Phys. Rev. Lett. ,111303 (2018).[17] A. Aguilar-Arevalo et al. (DAMIC), Constraints on LightDark Matter Particles Interacting with Electrons fromDAMIC at SNOLAB, Phys. Rev. Lett. , 181802(2019).[18] L. Barak et al. (SENSEI), SENSEI: Direct-Detection Re-sults on sub-GeV Dark Matter from a New Skipper-CCD,Phys. Rev. Lett. , 171802 (2020).[19] T. Emken, R. Essig, C. Kouvaris, and M. Sholapurkar,Direct detection of strongly interacting sub-GeV darkmatter via electron recoils, Journal of Cosmology and Astroparticle Physics (09), 070–070.[20] C. Fu et al. (PandaX-II), Spin-Dependent Weakly-Interacting-Massive-Particle–Nucleon Cross Section Lim-its from First Data of PandaX-II Experiment, Phys. Rev.Lett. , 071301 (2017), [Erratum: Phys.Rev.Lett. 120,049902 (2018)], arXiv:1611.06553 [hep-ex].[21] C. Fu et al. (PandaX), Limits on Axion Couplings fromthe First 80 Days of Data of the PandaX-II Experiment,Phys. Rev. Lett. , 181806 (2017), arXiv:1707.07921[hep-ex].[22] X. Chen et al. (PandaX-II), Exploring the dark matterinelastic frontier with 79.6 days of PandaX-II data, Phys.Rev. D , 102007 (2017).[23] X. Cui et al. (PandaX-II), Dark Matter Results From 54-Ton-Day Exposure of PandaX-II Experiment, Phys. Rev.Lett. , 181302 (2017).[24] J. Xia et al. (PandaX-II), PandaX-II Constraints onSpin-Dependent WIMP-Nucleon Effective Interactions,Phys. Lett. B , 193 (2019).[25] X. Ren et al. (PandaX-II), Constraining Dark Mat-ter Models with a Light Mediator at the PandaX-II Experiment, Phys. Rev. Lett. , 021304 (2018),arXiv:1802.06912 [hep-ph].[26] Q. Wang et al. (PandaX-II), An Improved Evaluation ofthe Neutron Background in the PandaX-II Experiment,Sci. China Phys. Mech. Astron. , 231011 (2020).[27] K. Ni et al. (PandaX-II), Searching for neutrino-less dou-ble beta decay of Xe with PandaX-II liquid xenon de-tector, Chin. Phys. C , 113001 (2019).[28] Q. Wang et al. (PandaX-II), Results of dark mattersearch using the full PandaX-II exposure, Chin. Phys.C , 125001 (2020).[29] X. Zhou et al. (PandaX-II), A Search for Solar Axionsand Anomalous Neutrino Magnetic Moment with theComplete PandaX-II Data, Chinese Physics Letters ,011301 (2021).[30] Q. Wu et al. , Update of the trigger system of the PandaX-II experiment, JINST (08), T08004.[31] M. Szydagis, C. Levy, G. M. Blockinger, A. Kamaha,N. Parveen, and G. R. C. Rischbieter, Investigatingthe XENON1T low-energy electronic recoil excess usingNEST, Phys. Rev. D , 012002 (2021).[32] Noble Element Simulation Technique, http://nest.physics.ucdavis.edu/download/calculator .[33] E. Aprile, K. L. Giboni, P. Majewski, K. Ni, and M. Ya-mashita, Observation of anticorrelation between scintilla-tion and ionization for MeV gamma rays in liquid xenon,Phys. Rev. B , 014115 (2007).[34] B. Yan et al. (PandaX-II), Determination of responsesof the PandaX-II liquid xenon detector to low energyelectron and nuclear recoils (in preparation).[35] M. Szydagis, N. Barry, K. Kazkaz, J. Mock, D. Stolp,M. Sweany, M. Tripathi, S. Uvarov, N. Walsh, andM. Woods, NEST: A Comprehensive Model for Scintilla-tion Yield in Liquid Xenon, JINST , P10002, 1106.1613.[36] Direct Detection of sub-GeV Dark Matter, http://ddldm.physics.sunysb.edu/ddlDM/ .[37] F. James, Statistical Methods in Experimental Physics2nd Ediition (World Scientific, 2006).[38] R. Essig, T. Volansky, and T.-T. Yu, New constraints andprospects for sub-GeV dark matter scattering off elec- trons in xenon, Phys. Rev. D , 043017 (2017).[39] H. Zhang et al. (PandaX), Dark matter direct search sen-sitivity of the PandaX-4T experiment, Sci. China Phys.Mech. Astron. , 31011 (2019). [40] Q. Arnaud et al. (EDELWEISS), First germanium-basedconstraints on sub-MeV Dark Matter with the EDEL-WEISS experiment, Phys. Rev. Lett.125