Overview of Dark Matter Direct Detection Experiments

Ning Zhou Shanghai Jiao Tong University PIC2018, 2018-09-14, Bogota, Colombia Outline

• Dark matter candidates • Dark matter direct searches • DAMA/LIBRA anomaly • Spin-independent WIMP-nucleon interaction – High mass – Low mass • Spin-dependent WIMP-nucleon interaction • WIMP-electron interaction • Directional direct search • Summary

Ning Zhou, PIC2018 2 Dark Matter

• Strong evidences for the existence of dark matter • The nature of dark matter is unkown

Bullet Cluster

Chandra X-ray Observatory and Hubble Space Telescope

Galaxy rotation curve

Structure formation

CMB

Ning Zhou, PIC2018 3 Dark Matter Candidates

• Weakly Interacting Massive Particles (WIMPs) • Axions, ALPs • …

Ning Zhou, PIC2018 4 Dark Matter Detection

• DM searches – Direct detection – Indirection detection – Collider search Collider Indirect production search

Direct search

Ning Zhou, PIC2018 5 Dark Matter Direct Detection

• DM mass range: GeV – TeV • Local density: 0.3 GeV /cm3 • Isothermal velocity distribution, v0~220 km/s • Nuclear recoil (NR) vs Electron recoil (ER)

Ning Zhou, PIC2018 6 14

Standard Assumptions for Dark Mat ter Direct Detection

WIMPsχ incoming DM particles ! DM mass range: GeV~TeV F igur e•11. Background sources and shielding in a typical direct detection experiment χ • local WIMP density: 0.3 GeV/cm3 neutrinos form a fundamental lower bound on the cross section for background-free WIMP detection [43]. Next generat ion exp•eriIsothermalments will ha velocityve sensitiv distribution:ity within an or der of magnitude of the neutrino signal for most of the mass rangev, an~220d wi lkm/sl actual ly detect the 8B solar neutrino signal. 0 nuclear recoils Finally, anot her method to deal with backgrounds is to exploit the fact that the Earth is moving through the dark mat ter that sur•rouWIMPnds our escapegalaxy, y ivelocityelding a “W ~544IMP km/swind” that appears to come from the constellat ion DarkCygnus. T hMatteris should, in pri nSignalsciple, creat e a s mall “ annual modulat ion” in the detected WIMP rat es, as well as a somewhat large•r dStandaraily modudl atchannels:ion, as sho wSIn andin F iSDg. 12. However, if such e↵ ects were detected in an Z dR ⇢ vesc dσ = 0 χ N (v, E )vf (v)dv dE m m dE R June R χ N vmin R

dσχ N mN SI 2 SD 2 = 2 2 (σ0 FSI (ER )+ σ0 FSD (ER )) dE R 2µN v

Dec 利用一期的探测器寻找低质量的暗物质上海交通大学博士学位论文 F igur e 12. SKaixuanchema Niti c o f t h e p o s s i bRecentle so uResultsrces ofrfoma nDarknua Matterl mo Dirduectionlatio nDetection(left) a n d d a i l yCIPmANPod u2018,lati o5/29-6/3/2018,n (right) Palm Springs, CA 4 e↵ ects if WIMPs are detected in direct detection experiments

experiment, there would still have to be a convincing demonstrat ion that there are no such modulat ions in background sources. Recoil energy Electron yield spectrum in xenon

Communit y Pl anning St udy: Snowmass 2013

Ning Zhou, PIC2018 7

图假设自旋不相关散射截面， − 9 质量为 2 的粒子与不同靶标核素碰撞的积分能谱：，，和。实线上的标记是各实验的典型探测阈值。来源见。

子的平均自由程小于毫米，而的伽马光子的平均自由程大约为公分。换而言之，较大的原子系数使得液氙有较高的密度，从而使得液氙本身就可以对探测器外围的伽马本底产生很好的屏蔽效果，这种屏蔽效应也被称为“自屏蔽”效应（）。 µ l =(µ)− 1 =( × ρ)− 1 ρ

由于液氙非常好的自屏蔽效应，在液氙探测器中，来自探测器外围的本底事例被极大的降低。而实际上，在和实验中，外层的液氙将绝大多数的本底事例被阻挡在置信区域之外，从而在置信体积内营造一个更为干净的暗物质探测区域。图

显示了探测器内不同位置的本底分布，颜色代表事例率， 10 ，定义为“ ”（每能量沉淀每天在公斤液氙内产生的事例数）。图中的虚线显示的是最后选择的置信体积的边界。所以，基于低能区本底事例率的分布与事例能量不相关的假设，我们可以很明显得看到，本底事例从探测器边缘到内部急剧减少。相比较事例率最高的区域（暗红色），置信体积内的本底事例率减少了到个数量级。 Deep Underground Laboratories

• Shielding cosmic rays • Mine or tunnel

Ning Zhou, PIC2018 8 A world-wide effort

•

Ning Zhou, PIC2018 9 Detection Methods

• Charge, Light and Heat

K. Ni

Ning Zhou, PIC2018 10 Current Constraints

• Lower threshold and lower background

Ning Zhou, PIC2018 11

• DAMA/LIBRA Anomaly

Ning Zhou, PIC2018 12 DAMA/LIBRA

• DAMA/LIBRA-phase2 arXiv:1805.10486 – NaI (Tl) 250 kg, 6 annual cycles, 1.13 ton-year, light collection • The only experiment claiming signal with > 5 sigma – 50 GeV with 7x10-6 pb or 6-16 GeV with 2x10-4 pb • Not confirmed by experiments with other targets

8.0sigma 9.5sigma

12sigma

Ning Zhou, PIC2018 13 Other NaI Experiments

• SABRE, COSINE, DM-ICE, KIMS and ANAIS, etc • Stay tuned!

COSINE-100 5-year expectation

From Reina Maruyama’s talk at CIPANP 2018

Ning Zhou, PIC2018 14

• Spin Independent WIMP-Nucleon Interaction • High Mass WIMP

Ning Zhou, PIC2018 15 Noble liquid Experiment

signal to reject electronic backgrounds; • however, this achieved modest rejection efficiencies and only at relatively high recoil energies. When the first double-phase Xe detectors were deployed for dark-matter searches, in the ZEPLIN-II/III [9,10] and XENON10 [11] experiments, the increase in engineering complexity soon paid off in sensitivity, and this technique has been at the • Dense and homogenous target, self-shieldingforefront of the field ever sin ce. Comprehensive reviews on the application of the noble liquids to rare-event searches can be found in the literature [12,13]. • High light and charge yields The TPC configuration at the core of double- phase detectors, illustrated in Figure 3.1.2, has several notable advantages for WIMP searches, in that two signatures are detected • Dual-phase: scintillation light and ionizationfor every interaction: a p roelectronmpt scintillation signal (S1) and the delayed ionization response, detected via electroluminescence in a thin gaseous phase above the liquid (S2). • Single-phase: scintillation light These permit precise event localization in 利用利用一期的一探期测的器探寻测找器低寻质找量低的质暗量物的质暗物质上海交上通海th大re交e d学im通e博ns大io士ns学 (to学博with位士in论 a 学fe文w位 mm论 [14]文) and discrimination between electron and nuclear recoil events (potentially reaching 99.99% rejection [15]). Both channels are sensitive to very low NR energies. The S2 response enables detection of single ionization electrons extracted from the liquid surface due to the high photon yield that can be achieved with proportional scintillation in the gas [16-18]. In LUX we have demonstrated sufficient S1 light collection to achieve a NR energy threshold below 5 keV [4]. The comNuclearbination of accurate 3-D imaging capability within a monolithic volume of a readily purifiable, Nuclear highly self-shielding liquid is nearly an ideal architecture for minimizing backgrounds. It allows optimal exploitation of the powerful attenuation of external gamma rays and neutrons into LXe, distinguishes multiplyrecoil-scattered back ginroun dXes from single-site signals, and precisely tags events on the surrounding recoil in Xe surfaces. This latter feature is important, given the difficulty of achieving contamination-free surfaces. The low surface-to-volume ratio of the large, homogeneous TPC lowers surface backgrounds in comparison to signal, and stands in stark contrast to the high surface-to-volume ratio of segmented detectors. These concepts are illustrated in Figure 3.1.3, which shows neutron interactions occurring just a few millimeters apart in the ZEPLIN-III detector. The S1 signals are essentially time-coincident, but the S2 pulses have different time delays corresponding to different vertical coordinates, making the rejection of such multiple scatters extremely efficient. The figure shows also a pulse observed in delayed coincidence in the surrounding veto detector, indicating radiative capture of this neutron on the gadolinium-loaded plastic installed around the WIMP target. LZ will utilize a similar anticoincidence detection technique to Light yield Chargecharacter izeyield the radiatio n environment around the Xe detector and to further reduce backgrounds. Nevertheless, when the first tonne-scale Xe experiments were proposed just over a decade ago, it was unclear whether LXe technology could be monolithically scaled as now proposed for LZ, or if it would be necessary to replicate smaller devices with target masses of a few hundred kilograms each. The latter option, while conceptually simple, fails to fully exploit the power of self-shielding. Since then, several

Ning Zhou, PIC2018 16 3-2 图图预测的预不测同的电不场同强电度场下强，度不下同，能不量同的能核量反的冲核的反光冲产的额光（产额（，左图，左和图电产和额电产额（（，右图，）右，图来）源，于来源于。。

形成形漂成移漂电移场电（场（）；而）门；而电门极电和极阳和极之阳间极形之成间所形谓成的所萃谓取的电萃场取（电场（）。）。外来外的来粒的子粒轰子击轰探击测探器测并器与并探与测探器测内器的内液的氙液相氙互相作互用释作放用能释量放。能这量些。在这液些氙在中液沉氙淀中的沉淀的能量能，量据，前据文前介文绍介的绍反的应反机应理机，理首，先首产先生产初生始的初闪始烁的光闪信烁号光，信被号命，名被为命名为以及电以子及—电子— 离子离对子。对电。子电—子离—子离对子进对一进步一再步结再合结形合成形进成一步进的一步的信号。信由号于。这由两于者这的两时者间的间时隔间非间隔非常小常，小所，以所该以过该程过产程生产的生两的个两个信号信便号融便合在融一合起在成一为起一成个为一个信号。信在号漂。移在电漂场移的电作场的作用下用，下那，些那未些与未离与子离复子合复的合自的由自电由子电漂子移漂到气移液到交气界液面交，界然面后，萃然取后电萃场取将电这场些将漂这移些至漂液移至液面的面电的子电拽子出拽到出气到氙气中氙。中而。在而气在氙气中氙，在中更，强在的更电强场的作电用场下作，用电下子，被电加子速被与加气速氙与原气子氙碰原子碰撞产生所谓的电致发光（）或者比例发光。这个过程产生的信号被称撞产生所谓的电致发光（）或者比例发光。这个过程产生的信号被称作。如前文所述，液氙中，初级发光的信号非常快。实验中典型的信号的宽度为作。如前文所述，液氙中，初级发光的信号非常快。实验中典型的信号的宽度为 ∼ 。不同实验中的漂移电场和电子学系统都对信号的宽度有影响。而典型的 ∼ 。不同实验中的漂移电场和电子学系统都对信号的宽度有影响。而典型的的信号则要宽得多，取决于阳极和门电极之间气体层的厚度。为方便软件有效地区的信号则要宽得多，取决于阳极和门电极之间气体层的厚度。为方便软件有效地区分和信号，现有实验的信号宽度一般在 µ 的量级。分和信号，现有实验的信号宽度一般在 µ 的量级。实验中，我们在的周围会安装光敏型的传感器来探测内产生的和实验中，我们在的周围会安装光敏型的传感器来探测内产生的和信号。一般来说，选用最多的光敏器件是对紫外光（）具有高量子化效率的光电信号。一般来说，选用最多的光敏器件是对紫外光（）具有高量子化效率的光电倍增管。、以及本文要阐述的实验都是在的顶部和底部安装倍增管。、以及本文要阐述的实验都是在的顶部和底部安装了光电管阵列。而且，为了增强探测器的光采集效率，在的周围，一般会装配具有了光电管阵列。而且，为了增强探测器的光采集效率，在的周围，一般会装配具有高反射效率材料比如聚四氟乙烯，构成的光反射墙。基于两相技术研发的高反射效率材料比如聚四氟乙烯，构成的光反射墙。基于两相技术研发的探测器，其一大特点是可以三维地重建探测器内部事例的发生位置。具体来讲，我们假探测器，其一大特点是可以三维地重建探测器内部事例的发生位置。具体来讲，我们假设电子在漂移电场线的作用下竖直往上运动。那么，液面处产生的水平位置也可以设电子在漂移电场线的作用下竖直往上运动。那么，液面处产生的水平位置也可以 Dual-Phase Detector

• S1: prompt scintillation light in liquid • S2: ionizationDar electronk mat tdriftinger se ator cgash e=>s with dual-phase TPC delayed light signal • Good ER/NR discrimination

Drift time S1 S2 Drift time S1 S2

S1 S2 Dark matter: nuclear recoil (NR) Ning Zhou, PIC2018 17 Gamma background: electron recoil (ER)

(S2/ S1)NR<<(S2/ S1)ER 5 Xenon Detectors

•

48cm

60cm

100cm

LUX PandaX-II XENON1T Sensitive volume 250 kg Sensitive volume 580 kg Sensitive volume 2000 kg Completed in 2016 54 ton-day, ongoing 1 ton-year, ongoing

Ning Zhou, PIC2018 18 Xenon: LUX, PandaX-II, XENON1T

• Strongest constraints on the high mass WIMP • XENON1T sets 4.1x10-47 cm2 at 30 GeV WIMP

Ning Zhou, PIC2018 19 5

run, an exponential energy dependence to extend the

) t

ER Surface neutron applicability to high-energy (up to ⇠20 keVee), and an ´

h 10 energy-dependent Fermi-Dirac suppression of the recom-

t CEnNS AC WIMP d i 3 Total BG (1.3 t) Data (1.3 t) binat ion at low-energy (. 2keVee). The resulting light

w 10

n and charge yields after ﬁtting are consistent with mea- i Total BG (0.9 t) Data (0.9 t)

b 2 (

10 surements [33, 36–38]. The ﬁt posterior is used to pre-

t dict the ER and NR distributions in the analysis space of

n 10 e

v the DM search dat a, achieving an ER rejection of 99.7% E 1 in the signal reference region, as shown in Table I.ER uncertainties in (cS1, cS2 ) are propagat ed for stat isti- -1 b 10 cal inference via variat ion of the recombinat ion and its -2 fluctuat ion, as these show the most dominant e↵ect on 10 -4 -3 -2 -1 0 1 (cS2 - m ) / s sensitivity (here defined as the median of an ensemble b ER ER of confidence intervals derived under the background- only hypot hesis [39, 40]). For WIMP signals, the uncer- 2 FIG. 4: Background and 200 GeV/ c WIMP signal best-fit tainties from all modeled processes are propagat ed into − 47 2 predictions, assuming σSI =4.7⇥10 cm , compared to DM an uncertainty of 15% (3%) on the tot al efficiency for search data in the 0.9 t (solid lines and markers) and 1.3 t 6 (200) GeV/ c2 WIMPs. (dotted lines and hollow markers) masses. T he horizontal axis is the projection along the ER mean (µER), shown in Fig. 3, normalized to the ER 1σ quantile (σ ). Shaded bands ER TABLE I: Best-fit expected event rates with 278.8 days live- indicate the 68% Poisson probability region for the total BG time in the 1.3 t fiducial mass, 0.9 t reference mass, and 0.65 t expectations. core mass, for the full (cS1, cS2b ) ROI and, for illustration, in the NR signal reference region. T he table lists each back- XENON1T Latest Resultsgro und (BG) component separately and in total, the observed diogenic neutrons originat ing from detector mat erials, data, and the expectation for a 200 GeV/ c2 WIMP prediction − 47 2 coherent elastic neutrino nucleus scat tering (CE⌫NS) assuming the best-fit σSI =4.7 ⇥ 10 cm . 8 mai•nly frExposureom B solar neut ri1nos ton, and c-osyearmogen ic neutrons Mass 1.3 t 1.3 t 0.9 t 0.65 t from secondary particles produced by muon showers out- (cS1, cS2b ) Full Reference Reference Reference side the TPC (negligible due to the muon veto [11]). The ER 627± 18 1.62± 0.30 1.12± 0.21 0.60± 0.13 8 CE⌫•NS rFiducialat e is constrai nvolumeed by B solar neutrino flux [26] neutron 1.43± 0.66 0.77± 0.35 0.41± 0.19 0.14± 0.07 and cross-section [27] measurements. The rat e of radio- CE⌫NS 0.05± 0.01 0.03± 0.01 0.02 0.01 genic neutrons is modeled with Geant 4 MC [28, 29] AC 0.47+0.27 0.10+0.06 0.06+0.03 0.04+0.02 using the measured radioactivity of mat erials [30], as- − 0.00 − 0.00 − 0.00 − 0.00 Surface 106± 84.84± 0.40 0.02 0.01 suming a normalizat ion uncertainty of 50% based on the Total BG 735± 20 7.36± 0.61 1.62± 0.28 0.80± 0.14 uncertainty in the So ur ces 4A [31] code and the di↵er- ence bet ween the Geant 4 and MCNP particle propa- WIMPbest -fi t 3.56 1.70 1.16 0.83 gat ion simulat ion codes [32]. Fast neutrons have a mean Data 739 14 2 2 free pat h of ⇠15 cm in LXe and produce ⇠5 times more multiple-scat ter than single-scat ter events in the detector, allowing for background suppression. A dedicat ed Energy deposits in light- or charge-insensitive regions search for multiple-scat ter events finds 9 neutron candi- produce lone S1s or S2s, respectively, that may acciden- dat es, consistent with the expectat ion of (6.4 ± 3.2) de- tally coincide and mimic a real interaction. The lone-S1 rived from the Geant 4 and detector response simulat ion spectrum is derived from S1s occurring before the main described below, which is used to further constrain the S1 in high energy events and has a rat e of [0.7, 1.1] Hz. expected single-scat ter neutron event rat e in DM search The uncertainty range is determined from di↵ering rat es dat a. of single electron S2s and dark counts in the time win- The detector response to ERs and NRs is modeled sim- dow before the event. The lone-S2 sample is composed of ilarlarXiv:1805.12562y to the method describe d in Refs. [5, 33]. All 220Rn, all triggered low-energy events containing S2s without a 241AmBe, and neutron generat or calibrat ion dat a from validly paired S1 and has a rat e of (2.6± 0.1) mHz (with- bot h science runs are simultaneously fitted to account for out requiring the S2 threshold). The AC background rat e correlat ions of model parameters across di↵erent sources and distribution are estimat ed by randomly pairing lone- and runs. To fit the 220Rn dat a, the parameterizat ion S1s and -S2s and simulat ing the necessary quantities for of the ER recombinat ion model is improved from [5]by applying the event selection defined above. modifying the Thomas-Imel model [34]. These modifica- 222Rn progeny plat e-out on the inner surface of the tions include a power law field-dependence similar to [35] PTFE panels may decay and contaminat e the search re- to account for the di↵erent drift fields in each science gion if the reconstructed position falls within the fiducial Ning Zhou, PIC2018 20 Future Xenon Detectors

Experiment Sensitive Fiducial Expected Expected Status • Volume Volume exposure Sensitivity PandaX-4T 4 ton 2.8 ton 5 ton-year 10-47 cm2 Commissioning 2020 XENONnT 6 ton 5 ton 20 ton-year 2x10-48 cm2 Commissioning 2019 LZ 7 ton 5.6 ton 20 ton-year 2x10-48 cm2 operations start April 2020 Darwin 40 ton 30 ton 200+ ton-year Neutrino CDR in 2-3 years floor

PandaX-4T XENONnT LZ

Ning Zhou, PIC2018 21 Future Xenon Detectors

• Darwin, with 200+ ton-year, can cover most of the region above neutrino floor for high mass WIMPs

Ning Zhou, PIC2018 22 Argon Detectors

• Pulse shape of prompt scintillation signal – Singlet (6ns) and triplet (1.5ms) • Dual-phase: Ionized electron vs prompt scintillation light

DarkSide-10

F90: the fraction of light collected within the first 90ns

Ning Zhou, PIC2018 23 Current Running Argon Detectors arXiv:1802.071198 • DarkSide-50 @LNGS – 46 kg underground Argon – Dual-phase • DEAP-3600 @SNOLAB – 3600 kg natural Argon – Containing 39Ar (269 year) – Single-phase, R=85cm 6.8 ton-day

Ning Zhou, PIC2018 arXiv:1707.08042 24 GADMC

• Global Argon Dark Matter Collaboration • DarkSide-20k (2021 - ) – dual-phase, low radioactivity Ar – 50 tonne total mass, 30 tonnes fiducial mass – > 20 m2 of SiPM coverage • 300 tons fiducial mass detector (2026 - )

Ning Zhou, PIC2018 25

• Spin-independent WIMP-Nucleon Interaction • Low Mass WIMP

Ning Zhou, PIC2018 26 Toward Low DM Mass: S2-only Signature

• Dual-phase noble liquid detector • Usual signal region – PandaX: S1 [3PE, 45PE], S2 [100PE, 10000PE], Threshold ~ 1 keVee • Low mass signal region – S2-only, no ER/NR discrimination – Threshold ~ 0.1 keVee (DarkSide-50) – ~10-41 cm2 at 2 GeV

Phys. Rev. Lett. 121, 081307 (2018)Ning Zhou, PIC2018 27 Low Mass DM: Germanium detectors

• CDEX-10 @ CJPL • 10 kg Ge in LN2 • Analysis threshold 160 eVee • 102.8 kg-days exposure

PRL 120, 241301 (2018) Ning Zhou, PIC2018 28 Low Mass DM: Germanium detectors

• SuperCDMS experiment @Soudan • Standard iZIP mode: phonon and ionization, ER/NR discrimination • CDMSlite HV mode: phonon only – Ionization electrons generate a large number of NTL phonons – => Low threshold 56 eVee 50-80V

Phys. Rev. D 97, 022002 (2018) Ning Zhou, PIC2018 29 Low Mass DM: Crystal detector

• CRESST-III experiment

• CaWO4 crystal, 24g, @ ~15mK – Phonon signal: precise measurement of deposited energy – Scintillation light: particle-type dependent – Nuclear recoil threshold 30.1 eV => subGeV WIMP

F. Reindl, IDM2018

Ning Zhou, PIC2018 30

• Spin Dependent WIMP-Nucleon Interaction

WIMP

Ning Zhou, PIC2018 31 Spin-dependent WIMP-neutron Interaction

• Xenon odd-A isotope with unpaired neutron => High Mass – Xe129 (26.4%) Spin 1/2 – Xe131 (21.2%) Spin 3/2 • Germanium odd-A with unpaired neutron => Low Mass – Ge73 (7.73%) Spin 9/2

-35 )

2 10 m

c -36 (

n o

i -37 Xenon Germanium t

c 10

e s

-38 s

s 10 CDMSlite o

r -39 c

n o

r -40 t

u 10

e n - -41 PandaX-II full SM

P 10

PandaX-II chiral EFT M

I -42

10 LUX W

-43 XENON100 D

S 10 10-44 10 102 103 104 105 WIMP mass (GeV/c2) arXiv:1807.01936 Ning Zhou, PIC2018 Phys. Rev. D 97, 022002 (2018)32 Spin-dependent WIMP-proton Interaction

• PICO-60 experiment with C3F8 as the target – F19, unpaired proton, spin 1/2 – Bubble chamber with superheated liquid – 3.4 x10-41 cm2 at 30 GeV/c2 • PICO 40L, starting ~ December 2018 • PICO 500, under design

PRL 118 (2017) no.25, 251301 Ning Zhou, PIC2018 33

• WIMP-Electron Interaction

WIMP

Ning Zhou, PIC2018 34 WIMP-electron Interaction

• Light mass: Dark photon, ALP etc PandaX -II DarkSide-50

PRL119, 181806 (2018) Ning Zhou, PIC2018 arXiv:1802.06998 35 SuperCDMS HVeV

• Single-charge sensitive detector – Charge resolution: 0.1 electron-hole pairs • 0.93 g Si crystal (1 x 1 x 0.4 cm3) @ 33-36mK • 0.49 gram-days

WIMP-electron interaction, heavy mediator

Ning Zhou, PIC2018 PRL 121, 051301 (2018) 36 WIMP-Electron Annual Modulation

• 4-year exposure in XENON100 PRL 118, 101101 (2017) • Weak modulation signature at a period of 431 days in low energy SS events • Not compatible with DAMA modulation

Ning Zhou, PIC2018 37

• Directional Direct Search MIMAC MIcro-tpc MAtrix of Chambers A Large TPC for Directional Dark Matter detection

Daniel Santos Laboratoire de Physique Subatomique et de Cosmologie (LPSC-Grenoble) (UJF Grenoble 1 -CNRS/IN2P3-INPG)

Ning Zhou, PIC2018 38 Directional Direct Search

• Directionality: – Cygnus direction – Can help with neutrino floor! – 30deg angular resolution necessary to distinguish Cygnus from Sun • To reconstruct the recoil track • R&D work in progress

MIMAC DRIFT DMTPC NEWAGE Micromegas Negative ion MWPC CCD Micro pixel Ning Zhou, PIC2018 39 Summary

• World-wide efforts in the dark matter direct detection • A large variety of techniques and targets • No compelling positive results are obtained yet. DAMA’s anomaly is under further cross-check • Within ~10 years, we are able to reach the neutrino floor • Stay tuned!

THANK YOU!

Ning Zhou, PIC2018 40 Backup

•

Ning Zhou, PIC2018 41