the search for weakly interacting massive particle dark...

B.SadouletAPS 2013 W.K.H Panofsky Prize 1

The Search for Weakly InteractingMassive Particle Dark MatterScience motivations and strategies

Is dark matter made of particles?What physics? Complementarity of experimental approaches

Two paradigmsWeak scale WIMP (≈ 100GeV): how this business started!A Dark Sector (low mass?, structure?): emerging ideas!

Experimental strategy for Direct DetectionVery low rate, small energy deposition, need for nuclear recoilComplementarity of approaches:

Noble Liquids may be more natural at high massesLow temperature detectors are essential if we want to explore low mass WIMP

Bernard SadouletDept. of Physics /LBNL UC BerkeleyUC Institute for Nuclear and ParticleAstrophysics and Cosmology (INPAC)UC Dark Matter Initiative

B.SadouletAPS 2013 W.K.H Panofsky Prize

W.K.H. Panofsky

Thanks to APSParticularly honored because Pief was one of the

important mentors of my youth!Very happy to share this prize with Blas Cabrera

Coordinated talks

2


Dark Matter: A Central Cosmology Question

The nature of dark matter is a central problem of cosmology!

≠ baryons ≠ light neutrinosIs it made of particles produced in

the early universe? If yes: evidence for physics beyond

standard model! Tev scale or totally different

origin?


4 Complementary Approaches

Dark MatterGalactic Halo (simulation)

WIMP annihilation in the cosmos

GLAST

Fermi/GLAST

VERITAS, also HESS, Magic + IceCube (v)

WIMP production on Earth

LHC

CDMS

WIMP scattering on Earth:e.g. CDMS, Xenon 100,etc.

Cosmological Observations

Planck

Keck telescopes


Two Paradigms

Weak scale WIMPs <= hierarchy problem

Dark Matter Hidden Sector: no necessarily weak scale

5


Weak Scale WIMP paradigm

Fantastic success of Standard Model but unstableWhy is h, W and Z at ≈100 Mp? Need for new physics at that scale supersymmetry additional dimensions, global symmetries In order to prevent the proton to decay, a new quantum number => Stable particles: Neutralino Lowest Kaluza Klein excitation, little Higgs

Particles in thermal equilibrium + decoupling when nonrelativistic

Cosmology points to W&Z scaleInversely standard particle model requires new physics at this scale => significant amount of dark matter

Weakly Interacting Massive ParticlesDark Matter could be due to TeV scale physics

Freeze out when annihilation rate ≈ expansion rate

⇒Ωxh2 =

3 ⋅10-27cm3 / sσ Av

⇒σ A ≈α 2

MEW

2

A remarkable concidence


Weak scale WIMP: How It All Started!My subjective impression of key steps

Dark MatterB Lee and S. Weinberg 1977 annihilation rate-> densityArticle of Silk and Srednicki, 1984 annihilation in gamma rays

Low temperature detectors -> neutrinosDrukier and Stodolsky Dec 1984Cabrera, Krauss, Wilzcek, Dec 1984

Goodman and Witten Jan 1985Direct detection is possible!

Ionization and Nuclear recoil recognitionNuclear recoils will ionize! (Marv Cohen-> Lindhardt) ≈1986Germanium (Avignone, Caldwell) 1987-88

excluded Z0 Importance of nuclear recoil (CDMS) 1989 Low pressure TPC (Tao, ≈1990)DAMA (Bernabei) 1990-2000Liquid Xenon (Elena Aprile)1998-2002

7


An Active Field

8

Credit: Joerg Jaeckel

Direct DetectionAn expanding community US≈ 270 physicists ≈70% FTE ≈ 40% of world


Xe 100 (2012)

Direct Detection

9

CMSSM≈mSUGRA Focal point regionNo threshold for Direct Detection

LHC Monojetsχγ µγ 5χ( ) qγ µγ 5q( )


No Missing Energy at the LHC

10

mSUGRA≈CMSSM 4 parameter + signUseful simplification but easy to eliminate

Natural supersymmetryDo only what you really need to solve

the hierarchy problem : light s-top Much more open

But other problems (mh, Bs->𝛾)

Simplified models ->600 GeV

mSUGRA ≈out except FPbut SUSY parameter space still very much open although -> larger mass


Dwarfs

Dark Matter Annihilation (Fermi)

11

Galactic Halo

Extragalactic Background

cf Alex Drlica-Wagner

No clear signalPresentation in terms of bb, 𝝉𝝉 ,WW channels is simpleBut realistic model is a mix -> less restrictive limits:

No strong limit on low mass

- - -


Different Paradigm: Dark Sector

12

WIMP-less DM (e.g., Feng Kumar)MSSM breaking through gauge mediation in SUSY breaking sector+ Hidden sectors

Thermal relic with WIMP-less miracle

⇒ Naturally σv ≈gx

4

mx2 ≈

gEW4

mZ2 but mX ≠ mZ

What if there were a hidden dark sector?In contact with ordinary sector at high temperature.Now shielded from ordinary sector

e.g., preventing production of low mass WIMp at accelerators

Three examples

Dark Photon (e.g., Arkani-Ahmed, Finkbeiner, Weiner)Light mediator -> interesting phenomenology: excited stated,

Sommerfeld enhancement etc..

Asymmetric Dark Matter (e.g., K. Zurek)What if dark sector was also asymmetric betweendark matter and anti-dark matter?Reasonable to have ≈ same asymmetryif MDM=6*Mp we get ΩDM≈6*Ωb as observedbut no idea about cross section, no indirect


Fundamental Motivations?Weak: Make sense of a confused experimental situation

Just having more parameters to play with, to fit experiments which are probably plagued by misunderstood backgrounds, systematics, calibration and astrophysics problems

Stronger: Natural in many theoretical modelse.g., String theory inspiredNon necessarily related to weak scale

Hopefully: establish by solid observationsCurrent candidates

Serious failure of LCDM in dwarf galaxiesAstrophysics Or fundamental physics? e.g. self interacting

Direct detection signals at low energy????

Fermi-Glast GeV excess: unlikely that can distinguish from astrophysicsPamela,Fermi,AMS positron excess: difficult to distinguish from astrophysicsFermi-Glast 135 GeV gamma line in Fermi Glast if confirmed (no continuum -> internl structure of dark sector )

13


Dwarf Spheroidals

14

2) The density profile: NFW or core?Basic degeneracy between velocity

anisotropy and density profileWalker and Penarrubia: break the

degeneracy for Fornax and Sculptor with two populations of stars -> Core!

2 distinct but related problems1) The number of satellitesbut we keep discovering small ones

Not enough large masssatellites: “Too big to fail problem”

Frenk et al.Bullock et al.

σm

≈ 0.1 g / cm2( )−1 ≈ 0.18 barns/ GeV/c2( ) (Bullock's group)

Is it astrophysics or particle physics?Extremely efficient ejection of baryons by repetition of SNsOr strongly interacting dark matter => would point to dark sector


Low WIMP Mass Region (DD)

15

Claims for Very large modulationsCDMS does not see modulation (but needs to go

below 5 keV if possible )Statistical significance of CoGeNT marginal (2.8

sigma) => more statistics. Malbeck does not see a modulationControl of systematics is essential

CoGeNTCDMS

Trying to make sense of the various claims

CoGeNT evolving (surface events): no more claimCRESST likely 206Pb CDMS Ge interpreted by Collar/Fields: Our only comment so far similar effect in multiples

4

FIG. 3: Best-fit components of the null and alternative mod-els, overlapped on summed data, projected on Er and Ei.Red: ER and SE combined through the Crystal Ball PDF[15]. Green: ZC. Blue: NR. Black: sum of components. Thenull model requires a large deviation of the ZC centroid toEi ! 0.5 keVee, hard to reconcile with adequate Ei calibra-tions [12] and with the mean Ei of ZC events above ! 5 keVnr,a region where their true centroid can be assessed (Fig. 2).The separation between ZC and NR populations is noticeablefor data in the 5-11.9 keVnr analysis region used in [1].

stance, the addition of non-overlapping time periods [26]from detectors spanning an order of magnitude in back-ground rate within the signal box [14, 16], the negligibleoverlap with the CoGeNT spectral region containing aclear excess of events (Fig. 4), or unresolved issues relatedto CDMS’s energy scales [22]. However, we emphasizethat the choice of signal box boundaries, one that resultsin a poor signal-to-background ratio, is already su!cientto cripple its sensitivity.

In conclusion, we find that recently released 2007-2008CDMS data strongly support (5.7! C.L.) the presence ofa family of low-energy events concentrated in the nuclear

FIG. 4: Blue: best-fit NR component for CDMS summeddetector data, translated to ionization scale and overlappedon histogrammed CoGeNT data [4] after normalization tothe vertical scale. Neither is corrected for e!ciency nextto threshold. Dotted blue lines represent the 1! uncertaintyin the parameter A1 for NR. A dashed black line representsknown CoGeNT backgrounds (flat+cosmogenic, [4]), whichprovide an adequate fit to the data down to ! 1.2 keVee,the lower boundary of the CDMS annual modulation searchregion. Inset: 90% and 99% C.L. CDMS ROI in WIMP cou-pling vs. mass (see text), including all present uncertaintiesexcept for those related to CDMS’s energy scales [22]. ROI’sfor CoGeNT [4], CRESST [23] and DAMA [3] are from [6],and include the e"ect of a residual surface event contamina-tion in CoGeNT described in [4]. The DAMA ROI assumesa Maxwellian dark matter halo: deviations from it can dis-place it to lower WIMP couplings [5, 6, 24, 25]. Additionaluncertainties for DAMA exist [25].

recoil band. An origin in neutron scattering is highly un-likely. Their rate and spectral shape provides a matchto the majority of low-energy events unaccounted for byCoGeNT (Fig. 4). Both searches employ the same tar-get material, germanium, but perform orthogonal back-ground cuts [8], enhancing the possible meaning of thiscoincidence. If this excess is interpreted as a WIMP sig-nal, it falls in close proximity to other anomalies reportedby recent dark matter experiments (Fig. 4). The favoredregion of WIMP mass is also of present interest in indi-rect searches for dark matter [27]. We also determine thatthe recent search for an annual modulation signal by theCDMS collaboration is insu!ciently sensitive to excludea dark matter origin for this excess, due to an inadequateselection of analysis region. Unsupported quantitativestatements made in [7, 11] about background composi-tion in CDMS detectors are not compatible with ourfindings.


New result for CDMS on Si K. McCarthy yesterday

16

Analysis favors a WIMP region of interest ≈3σ

Most likely value at 8.6 GeV WIMP mass with 1.9x10-41 cm2 cross section

Consistent with earlier CDMS Ge and Si limits

Also consistent with a WIMP interpretation of the COGENT experiment

In tension with limits from Xenon 10, Xenon 100 experiments

DAMA

CRESST

CoGeNT-KelsoXENON100

CDMS-II Ge

DAMA

CRESST

XENON10

CDMS-II Ge

CDMS-II Si

CDMS-II Si

Data are insufficient to claim discovery of a WIMP signal, but does warrant further investigation


How do we check?

17

Still to be done with CDMS IIInvestigation of possible systematics, unsuspected backgroundGe+Si consistent likelihood analysisLow energy Si analysis in same style as Ge

SuperCDMS SNOLABLower noise in phonons and Ionization <-R&DSi towers?

SuperCDMS SoudanCDMS Lite 23eV rmsLow threshold

Si towers?

SuperCDMS Soudan low threshold

CDMS II limit


Compatibility of Low Mass WIMP with LHC

LHC Monojets in effective contact operatorProbably means that if indeed there is light dark matter it does not couple to gluons (in most models it does not).

18

q,g

q,g

X

X

jet


Experimental Strategy

19

log σ

log M50 GeV/c2

2 directions1. Improve

sensitivity at large mass

5 GeV/c2

2. Improve sensitivity at small mass


2 frontiers: 1.High Mass FrontierGet as many events as possible

with no background=> convincing signal, enough statistics to do physicsWeak scale WIMP: natural scale if Higgs exchange 10-45 -10-46 cm2/nucleon+ push as down as possible if no signal

The smaller, the more fine tuned...≈ 10-48 cm2/nucleon: irreducible background from atmospheric coherent

neutrino scattering

What technology?Noble liquids most natural if they can control their background

=> several tons (e.g., LZ 7 tons)However Xe requires a factor 3000 improvement to get 10-48 cm2/

nucleon

But low temperature detectors can play a useful role for Weak scale WIMP:Current CDMS proposal 200kg -> 10-46 cm2/nucleonNeeded background rejection demonstrated within a factor of 2.Complementary: very different backgrounds and detection mechanism

Need a diversity of target e.g., isospin dependence (J. Kumar)

20


What technology?Low temperature detectors are most natural because of their large S/N Best: improved phonons and ionization (see next slide) Giving up rejection, you can go to very low threshold with luke phonons

but then background limited

2 frontiers: 2.Low Mass Frontier

21

Need excellent signal to noiseEnergy deposition goes as square of the WIMP mass

Ed =2M χ

2MN

M χ + MN( )2 v2 1− cosθ*( ) ≈ 2M χ

2

MN

v2 1− cosθ*( ) for M χ << MN

Xenon with S2 only (Sorensen) could also be useful, but No rejection: will have to rely on self shielding Fiducialization difficult (no t0-> can we use diffusion?) Few 100eV/ electron? Single electron background from surface would have to be corrected

Both would need careful calibration of energy scale


Importance of Discrimination

22

How far can we go?Gamma level 1/210 CDMS, beta 200kg Phonon energy: ≈Tc3 σ=180 eV -> σ=10 eV More optimistic!Ionization: FET->HEMT σ=300 eV -> σ=150 eV+ noise environment not baseline!!

No electron recoil rejection

Electron recoil rejection

100 101 10210−47

10−46

10−45

10−44

10−43

10−42

10−41

10−40

WIMP Mass (GeV)

σ o (cm

2 )

WIMP Sensitivity

Likelihood sensitivityEstimate by M. PylePreliminary!

SuperCDMS SoudanMajorana Demonstrator

CDMS II 2013

DAMA

CRESSTSi

CoGeNT -Kelso

We may be even able to detect coherent scattering of 8B solar neutrino! L. Strigari

Important physics signal (30kg/yr) + sensitivity demonstration (if no Si)

8B


WIMPs: ≥Two paradigms•Weak scale•Dark SectorPhenomenology (e.g., dependence on the target nucleus) may be more complex

than for the “vanilla” WIMP scenarios.

Conclusions

23

Fascinating time4 prong approach=>

complementary coverageconstrain theory speculations

No convincing result so farBut both paradigms still in good shape

Other good DM candidates: axions and possibly sterile neutrinos

Strategy for Direct DetectionSeveral technologies with complementary capabilities 2 frontiers• high WIMP mass: natural region of noble liquids (background control)

• low WIMP mass: S/N of low temperature (target mass cost )Different susceptibility to background, energy scale uncertaintiesEnough sensitivity overlap to have at least a second experiment able to cross-check any claim.Several targets => physics, large target masses-> astrophysics


Last but not least

24


Additional Material

25


DAMA

26

Can it be instrumental?Unstability in threshold: modulation appears smaller in LIBRA than in DAMA? Delayed pulses from muons (not neutrons, but defects): problems single rate

+phaseAwkward to the community: What to do?

Lower threshold: LIBRA has changed Phototubes to high QE + background modelExperiment by other groups: DM-Ice,ANAIS,KIMS, Princeton

Clearly modulationalthough not blind

Is it a signal?incompatible with other

experiments (New: KIMs)Saturates single rate=>Unphysicallylarge modulation

If true important!


LHC<-> Direct and Indirect Detection

27

Monojets/MonophotonsTim Tait et al, Paddy Fox et al:Effective theory with various

operatorsBut: Physical interpretation? limits of formalism(e.g., E<<M)


Complementarity Direct-Indirect

28

pMSSM model scan (19-dimensional scan of the MSSM) shown in grayred = models which the LAT may be sensitive to over a 10 year mission


Sterile Neutrinos?

29

AnomaliesLSND antineutrinoMiniBoone

antineutrinos now similar to neutrinostwice as much statisticless fluctuation up in high energy

Now compatible with LSND3.6 sigma

Note: Karmen excludes large DM2

not keV neutrino! Best fit ≈ 1 eV Compatibility with cosmology???

Ignarra Oct 12 3+1

Deficit of reactor antineutrinosSage-GallexTension with µ disappearanceWe need probably ≥2 sterile neutrinos


Axions

3 directionsCosmological axions

with RF cavitiesSolar axionsLight-through the wall

experiments

cavity: next year

cavity: 4-year

cavity: very challenging

helioscope: current

helioscope: 10-year

Laser: current

Laser: locked FP

Rosenberg


What about the Galactic Center?

31

2 independent analysis Hooper,LindenAbazadjian,Klapinghat

More or less agree on conclusionsCould be dark matter annihilation: spatial shape, cross section, spectrumBut could be Millisecond Pulsars or Cosmic rays interaction

How will we ever know whether astrophysics or Particle Physics?

Better spectrumHigher spatial resolution? Gamma 400?

Same type of remarks about TeV positron excess


Fermi 135 GeV Gamma Line?

Very few events ≈15+ Fundamental difficulty: how do we determine the number of trials

32

What could it be if it is real?Not ordinary WIMP (large cross section/absence of continuum)Dark Sector with light mediator or more generally structure (Weiner et al.)Decay of heavy right handed neutrino (Bergstrom)

Can it be instrumental?A priori smooth efficiency (MC) but

Use the limb as efficiency calibrator : 30 events; large error bars“Signal” appears toward the limb and the sun (Daniel Whiteson)Can it be due to change of reconstruction algorithm in this energy region?

the search for weakly interacting massive particle dark...

Documents