Prospects for the Direct Detection of Dark Matter - BNL theory groups

The Direct Detection ofDark MatterProspects and TechniquesNeutrino Helicity at 50A Celebration of the Goldhaber-Grodzins-Sunyar ExperimentHank SobelBrookhaven, May 2, 2008

Background• In the past decade, breakthroughs in cosmologyhave transformed our understanding of theUniverse.• A wide variety of observations now support aunified picture in which the known particlesmake up only one-fifth of the matter in theUniverse, with the remaining four-fifthscomposed of dark matter.• The evidence for dark matter is nowoverwhelming, and the required amount of darkmatter is becoming precisely known.2

Despite this progress, the identity ofdark matter remains a mystery• Constraints on dark matter properties thebulk of dark matter cannot be any of theknown particles.– One of the strongest pieces of evidence that thecurrent theory of fundamental particles andforces, is incomplete.• Because dark matter is the dominant form ofmatter in the Universe, an understanding ofits properties is essential to attempts todetermine how galaxies formed and how theUniverse evolved.– Dark matter therefore plays a central role in bothparticle physics and cosmology, and the discoveryof the identity of dark matter is among the mostimportant goals in basic science today.3

Dark MatterWhat We Know• How much:Ω DM = 0.23 ± 0.04• What it’s not:Not short-lived: τ > 10 10 yearsNot baryonic: Ω B = 0.04 ± 0.004Not hot: “slow” DM is requiredto form structureWhat We Don’t know• Mass?• Spin?• Other quantum numbers andinteractions?• Absolutely stable?• One particle species or many?• How produced?• When produced?• Why does Ω DM have the observedvalue?• Role in structure formation?• How distributed now?4

The Properties of a Good DarkMatter Candidate• stable (protected by a conservedquantum number)• no charge, no color (weakly interacting)• cold, non-dissipative• relic abundance compatible toobservation• motivated by theory (vs. “ad hoc”)5

Dark Matter Candidates• The theoretical study of dark matter is verywell-developed, and has led to many concreteand attractive possibilities.• Two leading candidates for dark matter areAxions and weakly-interacting massiveparticles (WIMPs). These are well-motivated,not only because they resolve the darkmatter puzzle, but also because theysimultaneously solve longstanding problemsassociated with the standard model ofparticle physics.Independently of this, if we try to understand theweak scale in particle physics, new particles appear. Ifwe add these to the universe:6

Add New Particle to the Universe• Particle initially inthermal equilibrium.As universe cools:Freeze-outDecrease dueto expansion ofuniverseChange due toannihilation andcreationNumber density now by integratingfrom freeze-out to present:Ω DM ~ -17

The “WIMP Miracle”(Weakly Interacting Massive Particle)Mass of Dark Matter particle (TeV)0.1 1.0HEPAP LHC/ILC Subpanel (2006)Ω DM0.1The amount of dark matter leftover is inversely proportional tothe annihilation cross section:Ω DM ~ -1If we take : σ A = kα 2 /m 2 ,then Ω DM ~ m 2andFor Ω DM ~ 0.1M ~ 100 GeV – 1 TeV.“WIMP”Cosmology alone tells uswe should explore theweak scale.8

WIMPs• In many supersymmetric models, thelightest supersymmetric particle is,stable, neutral, weakly-interacting, mass~ 100 GeV. All the right properties forWIMP dark matter!• In addition:Ω DM = 23% ± 4% stringently constrains models9

Direct Detection of WIMPssunhalobulgediskDark matter responsible forgalaxy formation (includingours).We are moving through a darkmatter halo.Usually assume sphericaldistribution with Maxwell-Boltzmann velocitydistribution.V=230 km/s, ρ=0.3 GeV/cm 3WIMP nucleus scattering rate calculated from theory.Elastic nuclear scattering: interactions are either spindependentor spin-independent.For SI, Low velocity coherent interaction ~ A 2Largeuncertainty10

Experimental ChallengesThe WIMP “signal” is a low energy(~10-100 keV) nuclear recoil.• Overall expected rate is very small– (limit now σ< 10 -43 cm 2 gives about 0.1 event/kg/day , mSUGRAmodels go to ~ 10 -46 cm 2 ).• Need a large low-threshold detector which candiscriminate against various backgrounds.– Photons scatter off electrons.– WIMPs and neutrons scatter off nuclei.• Need to minimize internal radioactive contamination.• Need to minimize external incoming radiation.– Deep underground location11

Possible WIMP Signatures• Nuclear vs electronic recoil– (discrimination required)• No multiple interactions• Recoil energy spectrum shape– (exponential, rather similar to background…)• Consistency between targets ofdifferent nuclei– (essential once first signal is clearlyidentified)• Annual flux modulation– (Most events close to threshold, smalleffect ~few%)• Diurnal direction modulation– (nice signature, but very short tracksrequires low pressure gaseous target,XeGe12

Minimizing backgrounds• Critical aspect of any rare event search• Purity of materials– Copper, germanium, xenon, neon among the cleanest with nonaturally occurring long-lived isotopes– Ancient Lead, if free of Pb-210 (T1/2 = 22 years)• Shielding– External U/Th/K backgrounds• Radon mitigation• Material handling and assaying– surface preparation– cosmogenic activation• Underground siting and active veto– Avoid cosmic-induced neutrons• Detector-based discriminationlog(sensitivity)bkg free: ~tbkg: ~t 1/2systematicslog(exposure)13

Current State of Experiments• Rapid advances in detector technologyhave reached interesting sensitivitylimits and should be able to go further.• Broad spectrum of technologies.• New ideas, and new collaborations areappearing…14

Direct Detection TechniquesXe, Ar,NeNaI, Xe,Ar, NeZEPLIN II, IIIXENONWARPArDMLUXSIGNNAIADZEPLIN IDAMAXMASSDEAP/CLEANScintillationFew % of EnergyGe, CS 2, C 3F 8DRIFTIGEXCOUPP~20% of EnergyIonizationHeat -CRESST IIROSEBUDCaWO 4, BGOPhononsZnWO 4, Al 2O 3…~100% of EnergyCDMSGe, SiEDELWEISSCRESST IAl 2O 3, LiF15

CDMS•The Cryogenic Dark Matter Search (CDMS) Collaboration,currently operating in the Soudan mine in Minnesota haspioneered the use of low temperature phonon-mediated Ge orSi crystals.Al CollectorAlphononsquasiparticlediffusionquasiparticletrapWTransition-EdgeSensorSi or GeR TES (Ω)~ 10mK4321SuperconductingT c ~ 80mKnormalT (mK)• Nucleus is hit, it recoils, causing the whole germanium crystal to vibrate.• Vibrations (phonons) propagate to surface of crystal. They excite quasi-particlestates, which propagate to the tungsten and heat it up.• Temperature of the tungsten rises, and therefore so does the resistance of thecircuit.• Bias current decreases since the voltage across the tungsten held constant.16

CDMS II Active Background RejectionIonization Yield/Recoil EnergyRecoil Energy•Radioactive source data defines the signal (NR– neutrons from 252 Cf) and background (ER –gammas from 133 Ba) regions.•Ionization Yield (Charge/Heat)>10 4 Rejection of γ•Ionization collection incomplete on surface. Yield canbe sufficiently low to pollute the signal region•Faster down conversion of athermal phonons atsurface provides faster phonon signal for β’sZero backgroundmaintainedProjectedsensitivity17

New Technologies• The field has been energized by the emergence ofnoble liquids (argon, xenon, neon) in various detectorconfigurations, as well as new ideas for use of warmliquids and various gases under high or low pressure.• These offer several things, some are:– An increased reach in sensitivity by at least three orders ofmagnitude for WIMP’s .– The possibility of recoil particle direction measurement.– Detector sizes well beyond the ton scale.• The complementarity of detector capabilitiesprovides:– A range of target types suitable for establishing WIMPsignature– Diverse background control methods (e.g., single phase vs.two-phase in noble liquids; various combinations of multiplesignatures).18

Noble Liquids• Relatively inexpensive, easy to obtain, densetarget material.• Easily purified as contaminants freeze out atcryogenic temperatures.• Very small electron attachment probability.• Large electron mobility (Large drift velocityfor small E-field).• High scintillation efficiency• Possibility for large, homogenous detectors.• Problem - 39 Ar, 85 Kr19

Single-Phase TechniquesCLEAN, XMASSRelative probability10101-1• Pulse shape discriminationto discriminate electronsfrom nuclear recoils.• Argon and Neon especiallygoodPrompt/SingletLight (τ ~ 6 ns)electronsnuclear recoils+Events/0.01 wide bin81071061051041010 8 simulated e-’sGets betteras sizeincreases.10 keV eventse -nuclear recoils-210-310-410-110 1 10Late/TripletLight (τ ~ 1.6 μs LAr,20 ms LNe)210310T photon10(ns)M.G.Boulay and A.Hime, Astroparticle Physics 25, 179 (2006)431021010100 simulatedWIMPs10 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9F120prompt

360 kg Mini-CLEANLiquid Argon/Neon•Spherical vessel filled with purifiedLNe or LAr.•R&D with micro-CLEAN and DEAP-1measuring salient properties of LAr &LNe•Engineering design and constructionplan for Mini-CLEAN (360 kg) undercompletion.•Engineering design underway forDEAP/CLEAN (3600 kg)Installation planned in SNOLAB21

800 kg XMASSLiquid XenonHexagonal tubewith Quartzwindow•R&D with 100kg detectorsuccessful•Very low background PMT’sinside LXe•Purification of Xe, => 85 Krreduced•Installation in Kamioka thisSpring22

2-Phase Noble LiquidsLiquid-gas interface withenhanced electric fieldGASSecondaryscintillationIonizationelectronsLIQUIDPMT’sS2S1Primaryscintillation-HV (Cathode)WARP, ArDM, XENON, ZEPLIN, LUX(Argon, Xenon)Multiplying gridExperimental handles•Primary scintillationintensity•Primary scintillationpulse shape•Secondaryscintillation intensity•S2/S1•Multiple recoils•Fiducial volumeSome best inArgon, somebest in Xenon23

WARP (Italy, U.S., Poland)LArCurrent Detectors UnderConstruction~100 kg Fiducial SizeZEPLIN (UK, U.S.) LXeXENON-100 (U.S.,Germany, Italy,Portugal) LXeLUX (U.S.)LXe24

Dark Matter Sensitivity Reach:2008-12Current bestworld limits !Ton-scale25

Warm Liquids - COUPP• Based on room temperature bubble chamberof CF 3 I. Other targets possible• Fundamentally new idea is to operate thechamber with a threshold in specificionization (dE/dx) above the sensitivityneeded to detect minimum ionizing particles,so that it is triggered only by nuclear recoils.(~10 10 rejection of MIP’s.)• Already reached stable operation of a 1-liter(2kg) version at shallow depth.• Demonstrated excellent γ rejection.26

Control Nucleation Rate and TriggeringIn HEP applications, bubble chambers stable for ~ 10 ms.Surface and bulk spontaneous nucleation rates reduced.Decreasing superheat Intrinsic γ rejection >10 10at 10 KeV thresholdMuon tracks only visible with high superheats27Operate at low superheat MIP blind 1 nuclear recoil = 1 bubble

COUPP• First results, no attempt to controlRn during these engineering runs.Science,Feb. 15thWith Rnabatementand moredepth (300m.w.e.)project this.28

DAMA/LIBRA – 250 kg NaIAnnual Modulation Signal30 km/sDecember~ 232km/s60°June30 km/sExpect:Modulated sinusoidal rate in particular energy range.Proper period (1 year)Proper phase (v(⊕ is maximum ~2 ~ 2 June: : V~245, comparedto December V~215 km/s).Modulation amplitude expected to be ~5% for usuallyadopted halo distributions.29

Results – April 2008•“A clear modulation is present in all the energy intervals (2-6 keV) andthe periods and phases agree with those expected in the case of a DMparticle induced effect. The cumulative analysis favors the presence of amodulated cosinelike behaviour at 8.2 σ C.L.”•“The superimposed curve represents Acos ω(t − t 0 ) with a period T =2π/ω = 1 yr, with a phase t 0 = 152.5 day (June 2nd)”. Amplitude based onbest fit to all data.30

Questions…Different halo models predict different phase and amplitudes.Because many processes vary with the seasons on Earth, and because theexpected modulation is so slight, minor errors in the analysis could produce afalse signal.Modulation in PMT noise.Modulation in 1-2 keV region?Other experiments,both SI and SD ruleit out…need tounderstand why.31

WIMP Search Complementary toCollider Search for SUSYNowExcluded by DirectDetectionCDMS II limit (2006)(10 -7 pb = 10 -43 cm 2 )Assuming zero-backgroundSensitivity:25 kg of Ge (Xe, I, W)(100 kg Ar, 200 kg Ne)Excluded byAccelerators2008?LHCILC TeV10 -471000 kg of GeDirect detection is crosssectionlimited.But sensitive to >TeVmass WIMPsColliders are masslimited.And can’t determine ifWIMP is stable32

Conclusions• Cosmology and particle physics independently point tothe weak scale for dark matter• Past investments are now paying dividends as currentexperiments are beginning to be sensitive to the ratespredicted in well-motivated models.• Recent advances in detector technology imply that thesesensitivities may increase by orders of magnitude in thecoming few years. Such rapid progress will revolutionizethe field, and could lead to the discovery of dark matterfor many of the most well-motivated WIMP candidates.• Direct search experiments, in combination with collidersand indirect searches, may not only establish the identityof dark matter in the near future, but may also provide awealth of additional cosmological information.33

“One day, all of these will be supersymmetryphenomenology papers”34

Additional Material

Gaseous Detectors• Low Pressure gas– Major goal is to identify dark matter by observing diurnalperiodicity.– Direction of the recoil nucleus must be reliably measured.• Achieve a full 3-D reconstruction for very short tracks (

DRIFT-II(U.S.,G.B.)•TPC filled with low-pressure electro-negative gas (CS 2).•Recoil tracks are ~few mm long•Ion drift limits diffusion in all 3 dimensions•End planes allow determination of range, orientation & energy•Excellent discrimination based on range and ionisation-density•Important R&D efforts by DRIFT groups and others, include improvementsin readout sufficient for the achieving of full directionality…GEMs,Micromegas, combinations of wires and scintillation optics or isochronous37cells and time-resolved pads.

AXIONS AND OTHER CANDIDATES• The relic density argument could be anaccident, or just part of the picture.• Axions and other candidates notproduced through freeze-out could besome or all of the dark matter.38

AxionsPeccei, Quinn (1977); Wilczek (1978); Weinberg (1978)• The theory of the strong interactionsnaturally predicts large CP violating effectsthat have not been observed. Axions resolvethis problem by elegantly suppressing CPviolation to experimentally allowed levels.• Cosmology and astrophysics sets the allowedaxion mass range from 1 μeV to 1 meV, wherethe lower limit follows from the requirementthat axions not provide too much dark matter,and the upper limit is set by otherastrophysical constraints.39

ADMX(LLNL, U.W.)• In a static magnetic field, there is asmall probability for halo axions to beconverted by virtual photons to a realmicrowave photon by the Primakoffeffect. This would produce a faintmonochromatic signal with a line widthof ΔE/E of 10 -6 . The experimentconsists of a high-Q (Q=200,000)microwave cavity tunable over GHzfrequencies.40

Excluded Axion Coupling g Aγγ vs. Axion Mass m A(retracted)CAST (projected)ADMX Upgrades• Completing phase Iconstruction - SQUIDtechnology.– 1-2 years to cover~10 -6 -~10 -4 eV down toKSVZ• Phase II to cover samerange down to DFSZand extending the massrange of the search.– Requires dilutionrefrigerator to go from1.7 to 0.2 KRegion of mass whereaxions are a significantcomponent of dark matter41