1

Direct Dark Matter Searches

Aldo Ianni, INFN-LNGS12 Mar., 2014

Cagliari

2

Evidence of Dark Matter• Spiral galaxies rotation curves: Whalo ~ 10Wstars (Wmatter>0.1)

• Clusters: galaxy motion, gravitational lensing: Wmatter~0.2-0.3

• CMB anisotropy: Wmatter~0.27, Wbaryons~0.04– ~84% of mass in the Universe dark and non-baryonic– <rDM> ~ 0.23rcrit~ 10-6 GeV/cm3

– around our Sun: rDM ~ 0.3-0.4 GeV/cm3

• Using early universe nucleosynthesis: Wbaryons~0.04 • Large Scale Structures:

– Formation of structures by gravitational clustering– Comparison of observations with non-relativistic (cold) dark matter

clustering agree well

3

The Cosmic Inventory

• CMB 1st peak: = 1W• Dark Energy ~ 0.72• Dark Matter ~ 0.23• Baryons ~ 0.04• Luminous matter ~ 0.01

4

Summary of DM properties

• DM reality based on strong observations• Unknown particle(s)– WIMPs, axions, keV-scale sterile neutrinos, …

• Make up ~80% of matter in the Universe• Neutral gravitationally interacting• Cold or not “too” warm (p/m <<1 at CMB

formation) • Stable or very long lived ( t >> 1017 years)

5

Dark Matter Searches

c

c

c c

c

c

q

q

qq

q

q_ _ _ _

INDIRECT COLLIDER

DIRECT

6

Direct DM Searches

•WIMPs

•Axion-like particles

7

WIMPs: Weakly Interacting Massive Particles

• A general class of weakly interacting massive (1GeV – 10 TeV) particles not from the Standard Model

• Assuming thermal equilibrium in the early Universe and non-relativistic decoupling, the energy density for these relic particles is predicted to be:

– W c ~ 0.2 for s ~ 10-36cm2

8

Few words on axions

9

DM as Axion-like Particle• ma ~ 6×10-6 eV (1012 GeV/fa) • Couplings:

• Axions are effectively collisionless• A non-relativistic population of axions can be present to match DM energy

density• For ma= 10-4 eV the number density of axions ~ 3 x 1014 m-3

• Mass range for Axion Cold DM: ~10-5 eV– Wa ~ (10-5 eV/ma)7/6 – v/c ~ 10-3

– Velocity dispersion ~ 3×10-17 c (10-5 eV/ma)5/6

a

f f

10

Direct Axions SearchesSearching for axions and ALPs• Haloscope: Axion DM eXperiment (ADMX)

– resonant conversion of axion to photons in a microwaves cavity permeated by a strong magnetic field (B = 8 Tesla)

– Enhancement when photon’s frequency corresponds to cavity’s resonant frequency’s, Q ~ 106 [Sikivie, PRL 51, 1415, 1983]

• Helioscope: CAST (2003 – 2014) -> IAXO

• Photon regenerations: OSQAR, ALPs-II

2 21

2)ah m c 1

11

12

Axions searches: sensitivities

13

WIMPs

14

Direct Search for WIMPs: nuclear recoil tagging

c c

N N

• 170 km/s < v0 < 270 km/s• 450 km/s < vesc < 650 km/s• rc ~ 0.3 Gev/cm3

• F(E) = nuclear form factor• f(v) = velocity distribution of WIMPs in the galaxy

Goodman and Witten, PRD31, 1985

15

“Standard” cross-section

Spin Independent interaction:

So for the “standard” fp = fn , ~ s A2

Spin Dependent interaction:

Deviations from the “standard” scenario are considered:Isospin Violating Interactions, fn/fp ~ -0.7Electromagnetic coupling …

Important to use different target detectors

16

Velocity Distribution

17

WIMPs Velocity relative to Earth

June 1st

Flux of 100 GeV/c2 WIMPs on Earth ~ 7×104 cm-2s-1

18

Form Factor (nuclear physics)

Na

Ar

GeXe

I

19

Vmin (Er, Mc, A)

Na, Er=6.7 keV

Ar, Er=31 keV

Xe, Er=2.5 keV

vesc

20

Recoil Spectrum

Ar

Xe

Na Ge

0.3 events/kg/year in Xe for 10-45 cm2 and 50 GeV/c2

21

Recoil Spectrum

2 events/ton/year in Xe for 10-47 cm2 and 100 GeV/c2

22

Annual Modulation: model independent signature

Galactic plane

Sola

r pla

ne

60°

Vsun ~ 235 km/s

Vearth ~ 15 km/sJUNE

Expected modulation (at % level) of1. rate2. spectral shape

23

Remarks on expected WIMP signal

• No specific feature• Low nuclear recoil energy• few events for 1 ton-year exposure at 10-47 cm2

– NEED background free experiment• GOAL of WIMPs search:– If Dark Matter is made of WIMPs we need:

• Strong background suppression (below WIMPs interaction rate)

• As low as possible sensitivity to WIMP-nucleon cross section

24

Low Mass WIMPs: threshold and target effect

Xe, 10 GeV/c2 0.06% above 7 keVr

Xe, 8 GeV/c2 0.06% above 7 keVr

Ge, 10 GeV/c2 36% above 2 keVr

s = 10-44 cm2

25

Solar Neutrinos as Background in e-recoil channel

Xe with data on light yield from LUX 2013

solar neutrinos with 99.6% discrimination power for e-recoil background

26

Solar Neutrinos Background in the n-recoil channel

• n-nucleus choerent scattering– Maximum recoil energy for 8B neutrinos = 4.3 keV– Flux of 8B ~ 6×106 cm-2s-1

Xe

27

Technologies• Cryogenic solid state

– Ionization spectrometer + bolometer operated at < 100mK– CDMS(Si and Ge); CRESST(Ca); EDELWEISS(Ge)

• LXe (XENON100, LUX) and LAr (DarkSide, MiniCLEAN, DEAP)– Scintillation + ionization

• Superheated liquid – Nuclear recoil induce bubble nucleation– Mainly with F for SD

• Scintillator crystal detectors– DAMA/LIBRA (NaI), CoGeNT (Ge), KIMS(CsI)

• Directional detectors– Gas TPC

28

WIMPs Direct Detection: Methods

Ionization

LightPhonons

DAMA

Xenon/LUX/DarkSideCDMS

CRESST

CoGeNT

CUORE

29

Background reduction

• Background reduction performed by means of discrimination to select ER vs NR

• Detector response different for ER and NR:– Ionization vs scintillation– Ionization vs heat – Pulse shape timing

30

DM searches worldwide• Gran Sasso

– DAMA/LIBRA– CRESST– XENON– DarkSide

• SNOlab– Picasso– COUPP– DEAP– CLEAN

• Homestake– LUX

• Soudan– SuperCDMS– CoGeNT

• Kamioka– XMASS– NewAge

• Canfranc– ArDM– ANAIS

• Modane– EDELWEISS

• YangYang– KIMS

• Boulby– Zeplin

• Jinping– PandaX– CDEX

31

Plot by R. Geitskell at DM2014

Background in DM, double beta decay and solar neutrino experiments

32

DAMA/LIBRA• DAMA– Low radioactivity 100 kg NaI array operated from 1996 to

2002 at Gran Sasso– Measures scintillation in crystal– No discrimination between ER and NR– 0.29 ton-year– Positive signal for annual modulation

• LIBRA– 250 kg NaI array operated since 2003– 1.04 ton-year– Positive signal for annual modulation

33

DAMA/LIBRA: results

• Modulation observed over 14 cycles• Cumulative exposure = 1.33 ton-year• Significance of modulation signal is 9.3s• Modulation amplitude in [2,6]keV = 0.0112±0.0012 cpd/kg/keV• Phase = 144±7 days• Period = 0.998 ± 0.002 year

34

DAMA/LIBRA: WIMPs fitTarget LY

[pe/keV]Threshold ER

[pe/keV]Threshold NR

[keVr]s/E

NaI 5.5-7.5 2 6.7(Na)22(I)

~7% at 60keV

Single-hit scintillation eventsEur. Phys. JC56, 2008

Mc = 12 GeV/c2 scp = 1.5×10-41 cm2 c2/Ndof = 1.02Mc = 8.6 GeV/c2 scp = 1.9×10-41 cm2 c2/Ndof = 1.69

Modulated amplitude, 2008

35

KIMS• Korea Underground Lab.• 12 crystals of CsI – 104.4 kg– Thereshold 1.5 keV– Background ~ 2 cpd/keV/kg– 2.5 years of data– PSD– Mean amplitude in 3-6 keV is 0.008 ± 0.068 cpd/keV/kg

• Future development (2015-2016) with low background NaI to test DAMA/LIBRA

36

CoGeNT

• Location: Soudan Underground Lab. USA• 440 g point contact HPGe at liquid nitrogen

temperature• Copper, lead and borated shielding + muon veto• Discrimination of surface events using rise time cut• Low threshold (400 eV) allows > 5 GeV/c2 WIMPs

search• After subtraction of known background it remains an

exponential signal excess• 2s hint for annual modulation

37

CDMS II

• 30 detectors (19 Ge, 11 Si) in Soudan Underground Lab (4.6 kg Ge + 1.2 kg Si)

• Data taking finished, moving to SuperCDMS (200 kg Ge array @ SNOlab)

• Measures ionization and phonons (TES read-out)• Discrimination– NR yield ~ 0.3– ER yield ~ 1– Surface events rejected by timing properties of phonon and

charge pulses

38

CDMS II Discrimination

39

CDMS II – Si results

• 140.2 kg-day• 3 events observed• Surf. backg. ~ 0.4• n backg < 0.13• 0.08 NR from 210Pb• p-value for null hypothesis is 0.19%• Best-fit:

– Mc = 8.6 GeV/c2

– Scp = 1.9×10-41 cm2

40

SuperCDMS: SI constraint at low mass

11 candidates seenExpected backg. = 6.6+1.1

-0.8

Feldman-Cousin gives a 90%CL upper limit at 11.2

577 kg-days at SOUDAN underground siteThreshold at 1.6 keVr

41

Liquified noble gases as WIMPs targetAr Xe Ne

Atomic number 18 54 10

Mean atomic mass 40 131.3 20.2

Melting point @ 1atm [K]

83.8 161.4 24.6

Density for liquid [g/cm3]

1.40 2.94 1.21

Volume fraction in atmosphere [ppm]

9340 0.09 18.2

Scintillation l [nm] 128 178 78

Scint. fast component [ns]

7 3

Scint. Slow component [ns]

1600 27

42

Argon and Xenon Complementarity

• Ar spin-independent interaction• Xe also spin-dependent– 129Xe (J=1/2, 26.2%) and 131Xe (J=3/2, 21.8%)

• Xe more sensitive at lower masses• Ar more sensitive at greater masses

43

Liquid Noble Gases Detectors

• Single phase (XMASS …)– 4p geometry for high light collection– Prompt, S1, signal + PSD

• Two-phase (Xenon1t, DarkSide … )– Cylindrical geometry with a TPC– Prompt, S1, and delayed, S2, signals + PSD

• Locate cryostat inside a Water Cherenkov muon veto with passive or active neutron shielding/veto

44

LXe• LUX @ Homestake

– 250 kg dual-phase LXe TPC with two 61-array PMTs– 85.3 days and 118.3 kg exposure– 2-30 phe window– 8.8 phe/keVee yield at null field for 122 keV gamma-ray– No events observed with 50% NR acceptance and 0.64±0.16 events of ER

background– Best limit on WIMP spin-independend search (~ 10-45 cm2)

• XENON1t @ Gran Sasso– Under construction at LNGS– 3.5ton LXe in total for projected sensitivity at ~ 10-47 cm2

– Based on success of XENON100 with 225days × 34kg• XMASS @ Kamioka

– 835 kg (100 kg fiducial) single-phase– First commissioning with unexpected background– Under refurbishment

45

LUX 2013 results

• 118 kg and 85.3 days

46

LAr strategy at Gran Sasso

• Two-phase TPC with LAr– Exploit background rejection power

• Low background technology– Careful selection of materials– Assembling in Rn-free clean room

• Active neutron veto– Boron-loaded scintillator

• Underground Argon– 39Ar activity (1 Bq/kg) reduced to ~0.6%

47

Expected WIMPs Signal in LAr

100 GeV/c2 WIMP mass10-45 cm2 WIMP-nucleon cross section~ 10-4 interactions/day/kg

Exposure of 1 ton-year gives about 40 events with these assumptionsExposure of 50 kg x 3 years gives about 5 events

48

More on Backgrounds to WIMPs

• / b g radioactivity• g radioactivity from surface close to sensitive

mass• Radiogenic neutrons: (a,n) and spontaneous

fissions• Cosmogenic neutrons from muons

Neutrons can mimic a nuclear recoil

49

Double-Phase TPC

Liquid

Gas Layer

Cathode

Drift FieldField Cage

Extraction Grid

Anode

Extraction Field

Photodetectors

Wavelength shifter

50

Two-phase TPC at Work: basic

gas

liquid

Upper PMTs array

Lower PMTs array

S1 prompt signalfrom LAr

Upper PMTs array

Lower PMTs array

Drift of free e-

and delayed signal in gas S2

S1 measures energy and time of eventS2 measures position of event in LAr and is proportional to the fraction of charge that escapesrecombination (this fraction depends on the drift field)

E field

51

LAr two-phase TPC at Work: signals~ 370 ms recoil-like

electron-likeS1 S2

52

Two-phase TPC at Work: figure-of-merit for background discrimination

Figure from WARP (Astropart. Phys. 28, 6

(2008) 495-507)

Pulse Shape Parameter

Log(

S2/S

1)

Background reduction performedby exploiting

a) Pulse shape of S1 through a parameter which measures thefraction of fast to slow componentin scintillation. F90 = Int_S1(<90ns) / Int_S1(all)

b) S2/S1: larger for e-like

Electromagnetic Events

Nuclear Recoils

Data from DS-10

53

The problem of 39Ar• Ar is naturally present in the atmosphere at 1% level• Atmospheric Ar contains 39Ar which is formed by

cosmic muon interactions– 40Ar(n,2n)39Ar

• 39Ar is a b decay emitter with Qb=565 keV and T1/2=269 years

• 39Ar is at the level of 1 Bq/kg– ~ 9×104 decays/kg/day– WIMPs(100GeV, 10-45 cm2) ~ 10-4 events/kg/day

PSD efficiency strongly depends on photoelectron statistics, so 39Ar rate asks for a higher detection threshold to obtain good background suppression39Ar must be removed to make a WIMPs experiment

54

DarkSide Program

• DarkSide-10– a prototype detector with 10 kg of LAr– operated underground at Gran Sasso for one year– test cryogenic technology with two-phase TPC– test Light Yield and background rejection power (F90 vs S2/S1)

• DarkSide-50– 50 kg of depleted LAr for 10-45 cm2 sensititivy for 100 GeV WIMP – currently under commissioning– Test active veto concept and low background procedures

• DarkSide-G2– 3 tons target mass of depleted LAr– WIMP sensitivity 10-47 cm2 for 100 GeV

55

Underground Argon

• 40Ar produced from 40K• The Earth is reach in 40K in underground• 40Ar moves into the atmosphere and makes 39Ar with

muons interactions• Underground 39Ar is expected to be much less, due to the

lower muon flux• However, 39Ar underground can be produced by radiogenic

neutrons interactions– 39K(n,p)39Ar

• Recipe: go deep underground where surrounding rocks are poor in U and Th

56

Neutrons from natural radioactivity• Radiogenic neutrons– from (a,n) and spontaneous fission of heavy

elements such as U and Th– energy ~ a few MeV (<10 MeV)

• Source in DarkSide:– PMTs (low background PMTs ~ few n/year/PMT)– Steel in cryostat and support structures

• Recipe: – Passive shielding effective for ~MeV energies but

asks for more surrounding materials– Active veto

57

Dark Side – 50 Detector

58

59

60

61

~ 10 mBq/m3 Rn clean room

62

63

64

65

Dark Side – 50 results: PDS

66

67

Dark Side – 50 results: summary

1. Ar purity: electron live time of 5m wrt drift time of 370ms

2. LAr LY at null field from 83Kr (41.5 keVee) is ~ 8 p.e./keVee

3. LAr LY at 200 V/cm from 83Kr (41.5 keVee) is ~ 7 p.e./keVee

4. PSD based on S1 well established

5. 278 kg-day exposure with AAr == 41851 kg-day UAr DS-50

6. LY in the neutron veto is ~ 0.5 p.e/keVee

7. 222Rn in LAr: <0.85mBq/kgAr

68

Dark Side – 50 Sensitivity

44.1 kg x 3 yr

35 keVr threshold

Based on present DS-50 results on ER rejectionpower

69

Dark Side – G2 Sensitivity

70

Conclusions• Dark Matter Direct search for Axions and WIMPs

(mainly)• Background free (or almost free) detectors needed• Low mass region (< 20 GeV/c2) with positive signals• Null experiments not in agreement with positive

observations• Upcoming efforts to test DAMA/LIBRA result• Dark Side – 50 at work at LNGS and already shows

the potentiality of LAr for Direct DM search• LAr is complementary to LXe

71

Cryogenic for DS-50 TPC

166K

30 l/min

15.6 psia

72

Underground Argon • In exhaust stream gas (CO2) of commercial mining facilities Ar at 400-

600ppm level– In DS extraction site: CO2 plant output, CO2(96%)+N2(2.4%)+He(0.4%)+Ar(0.06%)

• Make on-site preconcentration to ~40,000ppm, then cryogenic distillation at FNAL– After distillation: CO2(~0)+N2(<0.05%)+He(~0)+Ar(>99.95%)

• Purified depleted argon produced at 1 kg/day

• 39Ar depleted at <0.6% level wrt atmospheric level– ~500 decays/kg/day– For WIMPs ~ 10-4 interactions/kg/day

• Use LAr scintillation properties to perform background reduction– Prompt signal PSD shows: 90% n-recoil acceptance with <10-5 e-recoil leakage

73

Underground Argon

39Ar depletion factor >100

In progress study with countingdetector underground at KURF (1400 m.w.e.) Virginia, USA

74

Cosmogenic Neutrons

<En> ~ 90 MeV• Flux at Gran Sasso lab: 2.4 m-2 day-1

0.7 m-2 day-1 for > 10 MeV• Expected rate ~ 3×10-33 /s/atom• WIMPS rate ~ 10-34 /s/atom

• Neutrons from surrounding rocks reduced by shieldingIn DS-50 3m of water ~ 10-3 and 0.04from 1.5m of liquid scintillator: ~4 ×10-5

Water

LS

LAr/LXeTPC

m-induced neutrons

75

ULAr at low masses

Mass range: 10 GeV/c2

Cross-section: 10-41 – 10-40 cm2

39Ar

CDMS best-fit

5 keVr = 10 p.e.

S/B as large as 1.5

76

UAr and Supernova Neutrinos

Standard SN at 10 kpc

~ 60 events above 15 keVrfor 20 tons

Measure fluence and averageenergy of nx

77

SCENE:dependece of scintillation from nclear recoil in LAr on the Drift Field

arXiv: 1306.5675

Set-up installed and run at Univ. of Notre DameNuclear recoil response in Lar tested in 10.8 – 49.9 keVrThe drift field reduces the recombination probability affecting the light-yield

This result set the operation HV in DS-50

78

Rafael F. Lang, Purdue: The Direct Search for Dark Matter

~99.5% ER rejection @ 50% NR acceptanceenergy / keVnr

Discrimination using S2/S1 in XENON10060Co, 232Th and 241AmBe calibration

log 10

(S2 b/

S1) –

ER

mea

n

nuclear recoils(calibration)

electronic recoils(background)

S1 / PE

79

Rafael F. Lang, Purdue: The Direct Search for Dark Matter

1.0 ± 0.2 expected2 observedPoisson p=26%

likelihood analysis: background-only p-value energy / keVnr

XENON100: No Signal Observed225 live days, 34kg fiducial:

log 10

(S2 b/

S1) –

ER

mea

n

S1 / PE

arXi

v:12

07.5

988

electronic recoils(background)