The Cryogenic Dark Matter Search (CDMS) : Status and future Kipac SLAC April 2009 Department of Physics, Stanford University KIPAC associate member CDMS
The CDMS Collaboration Caltech Z. Ahmed, S. Golwala, D. Moore Case Western Reserve University D.S. Akerib, C.N. Bailey, D.R. Grant, R. Hennings- Yeomans, M.R. Dragowsky Fermilab D.A. Bauer, M.B. Crisler, J. Hall, D. Holmgren, E. Ramberg, J. Yoo MIT E. Figueroa-Feliciano, S. Hertel, K. McCarthy NIST K. Irwin Queens University W. Rau Santa Clara University B.A. Young Stanford University P.L. Brink, B. Cabrera, J. Cooley, W. Ogburn, M. Pyle, S. Yellin Syracuse University R.W. Schnee, M. Kos and M. Kiveni University of California, Berkeley M. Daal, J. Filippini, N. Mirabolfathi, B. Sadoulet, D. Seitz, B. Serfass, K. Sundqvist University of California, Santa Barbara R. Bunker, D. O. Caldwell, H. Nelson, J. Sander University of Colorado at Denver M. E. Huber, B. Hines University of Florida T. Saab, J. Hoskins, D. Balakishiyeva University of Minnesota P. Cushman, L. Duong, M. Fritts, V. Mandic, X. Qiu, A. Reisetter, Oleg Kamaev University of Texas A&M R. Mahapatra University of Zurich S. Arrenberg, T. Bruch, L. Baudis, M. Tarka
Composition of the Cosmos WMAP best fit WIMPs
Galaxy formation CDM simulations (U. Durham) Start 99% of way back to big bang Z Time 100 14 Gy 10 13 Gy 1 3 Gy 0.1 1 Gy 0.01 100 My 0.001 10 My 0 today
Galaxy formation CDM simulations (U. Durham) Start 99% of way back to big bang Z Time 100 14 Gy 10 13 Gy 1 3 Gy 0.1 1 Gy 0.01 100 My 0.001 10 My 0 today
Intergalactic Dark Matter Structure (U. Durham) Dark Matter filaments between galaxies contains Warm-Hot ionized Intergalactic (dark baryonic Matter (WHIM) detectable by (e.g.) O-VI absorption along line of sights to distant X-ray sources such as Quasers.
Intergalactic Dark Matter Structure (U. Durham) Dark Matter filaments between galaxies contains Warm-Hot ionized Intergalactic (dark baryonic Matter (WHIM) detectable by (e.g.) O-VI absorption along line of sights to distant X-ray sources such as Quasers.
Galaxy Rotation Curves Measurements Doppler shifts in stellar absorption lines Radio emission line of neutral hydrogen Satellite galaxies considered as test particles Conclusions Velocity curves do not roll over : mass density stays constant out to a radius x10 that of luminous matter. Mass density/critical density, Ω m > 0.07. Value of local mass density: 0.3 GeV/c 2 cm 3.
Unmasking Weakly Interacting Massive Particles (WIMPs) CDMS Tevatron WIMP production on Earth WIMP scattering on Earth HESS GLAST WIMP annihilation in the cosmos
SuperCDMS and the LHC, ILC
Direct Detection Astrophysics of WIMPs Energy spectrum & rate depend on WIMP distribution in Dark Matter Halo Spherical-cow assumptions: isothermal and spherical, Maxwell-Boltzmann velocity distribution V 0 = 230 km/s, v esc = 650 km/s, ρ = 0.3 GeV / cm 3 Energy spectrum of recoils is featureless exponential with E ~ 50 kev Rate (based on σ nχ and ρ) is fewer than 1 event per kg material per month moo At any instant may contain ten 60-GeV WIMPs on average. 20 billion WIMPs may pass through each second.
Direct Detection - Nuclear recoil Direct detection scattering experiment Few kev recoil energy < 1 event/kg/day Background suppression/rejection Low energy threshold Signal modulation WIMP detector ~10 kev energy nuclear recoil WIMP-Nucleus Scattering Importance of threshold and high quenching factor I/Xe a 50 kev true nuclear recoil threshold is equivalent to about 5 kev electron equivalent recoil
Discriminate against radio-active backgrounds Statistical inference Annual modulation (due to earth s motion around the sun) [DAMA], or even diurnal modulation (earth s rotation) [Drift] Event-by-event basis Each candidate (Nuclear recoil) event generates a unique signature: CDMS EDELWEISS DRIFT Phonons 0.3 kev Ionization 3 kev Bubbles 10 (?) kev CRESST COUPP PICASSO Scintillation 30 kev XENON, WArP ZEPLIN II & III DAMA, DEAP, XMASS
WIMP-detection Experiments Worldwide SNOLAB Picasso DEAP CDMS II COUPP DUSEL? CLEAN GEODM Boulby ZEPLIN II/III DRIFT 1/2 CanFranc ROSEBUD ANAIS ArDM EDELWEISS II ORPHEUS Gran Sasso DAMA/LIBRA CRESST I/II CUORE XENON WArP XMASS KIMS CsI LiF Elegant V&VI
Spin-Independent Exclusion Limit (46 zb) 4.75 kg Ge, 1.1 kg Si First 5-Tower results - published PRL 102 p. 011301 (Jan 2009)
Spin-Independent Exclusion Limit (46 zb) 4.75 kg Ge, 1.1 kg Si Projected CDMS II First 5-Tower results - published PRL 102 p. 011301 (Jan 2009)
CDMS II at Soudan - shielding Muon veto paddles can discriminate between muons and high energy gammas. Outermost polyethylene neutron moderator is 41 cm thick. Outer lead gamma shield is 23 cm thick. Secondary polyethylene neutron moderator is 14 cm thick.
Anatomy of Penetrating Neutron Event
Anatomy of Penetrating Neutron Event (i) ~100 GeV µ interacts in rock of tunnel generating neutron
Anatomy of Penetrating Neutron Event
Anatomy of Penetrating Neutron Event (ii) 330 MeV neutron from rock
Anatomy of Penetrating Neutron Event
Anatomy of Penetrating Neutron Event (iii)pb nucleus shattered 9 n (T 0.1-50 MeV) 9 g (E 0.1-2.5 MeV)
Anatomy of Penetrating Neutron Event
Anatomy of Penetrating Neutron Event (iv)lower energy neutrons moderate in polyethylene m.f.p ~ 3 cm@1 MeV
Anatomy of Penetrating Neutron Event
Anatomy of Penetrating Neutron Event (v)higher energy (30 MeV) neutron traverses poly m.f.p ~ 100 cm
Anatomy of Penetrating Neutron Event
Anatomy of Penetrating Neutron Event (vi)following ~12 scatters in Cu/poly neutron (now T~100 kev) (vi) scatters in two Ge detectors (Er~5 kev), and then (vii) ultimately captures on H in poly.
Photon and Neutron Calibration The Ionization response yield of Y the = detectors ratio of ionization is best demonstrated to phonon with - The calibration black lines demonstrate photon and the neutron +/- 2 σ bounds sources of the electron/nuclear recoil band Si ZIP Ge ZIP Bulk Electron Recoils (Gammas, X-rays and high-energy electrons) Bulk nuclear recoils (neutrons and WIMPS) Si ZIP Ge ZIP
Photon and Neutron Calibration Ionization yield Y = ratio of ionization to phonon - The black lines demonstrate the +/- 2 σ bounds of the electron/nuclear recoil band Bulk Electron Recoils (Gammas, X-rays and high-energy electrons) Bulk nuclear recoils (neutrons and WIMPS) Si ZIP Ge ZIP
CDMS Detector : Phonons + Ionization Ionization charge bias only 3 Volts Do not want fast Luke-phonons to contaminate phonon signal. 3 Volts sufficient to collect bulk events at sub-kelvin temperatures. Sufficient to collect bulk events at sub-kelvin temperaturesneed to discriminate against natural backgrounds Phonons Transition-edge sensors, transition temperature 100 mk. Athermal phonons collected fast enough to identify surface events. 3V bias Qinner Qouter + + - + + - - - + - + + + - - - Bulk Event Surface Bulk Surface Event Phonon Sensor side
SQUID array CDMS II Experimental Setup Phonon D SQUET card R bias R feedback I bias D C A B Q outer V qbias Q inner Tower Detectors: 250 g Ge or 100 g Si crystals 10 mm thick x 76 mm diameter. Two ionization electrodes, (Qinner and Qouter). Four phonon channels (A, B, C, D), reconstruct x, y and z coordinate. Detector Stack contains 6 ZIP detectors within 10 mk can of the icebox
ZIP detector phonon sensor technology TES s patterned on the surface measure the full recoil energy of the interaction Phonon pulse shape allows for rejection of surface recoils (with suppressed charge) 4 phonon channels allow for event position reconstruction 60 µm wide 380 µm Al fins ~25% QP collection eff.
Athermal phonon sensors High-energy phonons (~400 GHz) from particle recoil break Cooper pairs in superconducting Al (Tc = 1 K). The Al film acts as a phonon filter against other heating mechanisms. Resultant quasiparticles diffuse towards the tungsten trap where electron scattering heats up the W tungsten transition edge sensors (Tc ~ 70 mk). Al collector fins W - Al overlap Si or Ge surface 2 µm wide W transition edge senor Al fin Phonon Absorption Diffusion Losses ~30% ( D qp, τ qp, L diff ) ~75% Trapping Region Heated W electrons E ph 10Δ Al Cooper pairs sub 2Δ Al Cascade Phonons >50% loss Quasiparticles transport the energy to the W TES W TES
Transition Edge Sensors (TES) Steep Resistive Supeconducting Transition R T Voltage bias is intrinsically stable W T c ~ 70 mk T c width (10%-90%) <1 mk Unit-less measure of transition width α = dr dt R T The Joule heating produced by bias P J = V 2 B R P when R J is stable whereas for current bias P J = I B 2 R P J when R which is intrinsically unstable SQUID Array W ETF-TES I bias R shunt
Fabrication in Stanford Nanofabrication Facility ZIPs: Z-dependent Ionization and Phonon detectors. 1 cm thick Ge and Si substrates for CDMS II. Detectors utilize low temperature superconductivity of the Al and W thin film pattern layers. Gamma event rejection capability is 1 part in 5e5 for CDMS II Beta (surface) event rejection is 1 part in 200 for CDMS II. Phonon Sensors protected by Photoresist Inner and outer Ionization electrodes defined
WIMP Candidate: Blind Analysis All cuts set blind, without looking at signal In good Fiducial Volume Good Inner Fiducial Vol
WIMP Candidate: Blind Analysis All cuts set blind, without looking at signal In good Fiducial Volume In the Nuclear Recoil Band Good Inner Fiducial Vol
WIMP Candidate: Blind Analysis All cuts set blind, without looking at signal In good Fiducial Volume In the Nuclear Recoil Band Not surface event: phonon timing cut Good Inner Fiducial Vol
WIMP Candidate: Blind Analysis All cuts set blind, without looking at signal In good Fiducial Volume In the Nuclear Recoil Band Not surface event: phonon timing cut Not a Multiple Scatter Good Inner Fiducial Vol
The WIMP Search Data: Ge All cuts set and frozen! Predict 77 ± 15 single scatters in NR
The WIMP Search Data 97 Singles in Signal region, now try Surface Event Cut...
Open The Box: Surface Event Cut Expected Background: 0.6 ± 0.5 surface events and < 0.2 neutrons
Open The Box: Surface Event Cut Expected Background: 0.6 ± 0.5 surface events and < 0.2 neutrons
Open The Box: Surface Event Cut Zero observed Events in Nuclear recoil band Expected Background: 0.6 ± 0.5 surface events and < 0.2 neutrons
SuperCDMS phases - Moore s Law if zero bkgd 4 kg 15 kg 150 kg
SuperCDMS: 1-inch thick substrates 1-inch thick Ge and Si substrates processed at SNF (Stanford Nanofabrication Facility) aka CIS on Stanford campus. 1.4 micron wide TES
Improved phonon sensor layout
First calibration results from Ge 1-inch detectors
SuperCDMS phonon sensor layout Steep Resistive Supeconducting Transition W Tcc ~ 70 mk Tcc width (10%-90%) <1 mk R T Voltage bias is intrinsically stable
CDMS II tower vs SuperTower at Soudan SQUET card Tower
Path to Large Ge Crystals - purity Charge collection in dislocation-free Ge (Berkeley, February 2008) 3cm x 1cm sample of H-grown dislocation-free Ge (E.E. Haller) Unusable @77K, but can be neutralized at <100mK Excellent charge collection @ low voltages Available in >6 diameters (standard detector grade Ge limited to 3-4 ) 241 Am source 3 x1cm: 250g (CDMS II) 6 x2 : 5kg?
Path to Large Ge Crystals - photolithography 6-inch diameter mass-simulator spunup on SuperCDMS photoresist coater in SNF. Photoresist thickness in correct range for usual rpm ranges. Other fabrication equipment needed for 6-inch diameter detectors: Thin-film sputtering machine. Modified/New plasma etcher. Modifications to our present EV align U.V. Exposure machine.
Future detector ideas izip izip/double-sided detectors with outer phonon channel (A) to reject perimeter events. In izip charge electrodes interleaved with narrow strips occupied by phonon sensors. Less phonon timing information for surface events But now charge channels can veto surface events
Ice-box for 1 tonne of Ge (sub zeptobarn, < 10-45 cm 2 ) Icebox for 1 tonne of Ge ~ inner can 1 m on a side ~160 kg of Ge shown deployed, either 3 inch or 6 inch diameter detectors. Corresponding to SuperCDMS Phase B cross-section of 0.3 zeptobarns.
Ionization and Phonon read-out with SQUIDs Ionization: detector transformer-coupled to first stage of two-stage SQUID configuration Eliminate potential microphonics associated with FET readout Eliminate IR photon leakage Eliminate heated FET load on 4 K NIST transformer chip, ~12 mm x 6 mm Critically damped circuit, ~1 MHz sampling required. Simulations predict 0.4 kevee FWHM Phonons: standard two-stage SQUIDs Allows move to Al-Mn TESs to overcome W Tc variability. Commensurate with NIST-style time-domain multiplexing. Also allows improved phonon information.
Conclusions Cryogenic Dark Matter Search at Soudan Final (5-tower) run of CDMS II completed Analysis of last 1000 kg-days underway, Expect completion and announcement during summer 2009. First SuperTower (3kg) installed at Soudan Detector fabrication rate x5 faster (per unit mass) than CDMS II. Establish new detector design rejection capability. Measure reduced surface contamination levels. Construct more SuperTowers for SuperCDMS Soudan Detectors for second SuperTower already fabricated and undergoing cryogenic testing. Review this summer to allow further SuperTower deployment at Soudan. Sensitivity goal is WIMP-nucleon cross-section 5 x 10-45 cm 2 (5 zepto-barns) Beyond Soudan Background-free 100 kg Ge at >4000 mwe depth, looks straightforward Sensitivity goal is WIMP-nucleon cross-section 3 x 10-46 cm 2 (0.3 zepto-barns) Exploring new technologies and approaches to reach 1 tonne Ge under for
CDMS
WIMP-nucleon Spin-Dependent Limits CRESST I Phys. Rev. D73, 011102 (2006) CDMS II Si DAMA/NaI PICASSO CRESST I CDMS II Ge PICASSO CDMS II Si ZEPLIN I DAMA/NaI NAIAD Super-K different nuclear form factors CDMS II Ge Majorana ν 73 Ge is odd-n, even-p, nucleus (as is 27 Si) CDMS sensitive primarily to WIMP-neutron spin-dependent coupling CDMS rules out WIMP-neutron spin-dependent coupling to explain DAMA For WIMP-proton spin-dependent coupling, DAMA's allowed region (due to Iodine being odd-p) is ruled out at very low WIMP mass by CRESST-I, for standard halo model; and above 18 GeV/c 2 by SuperK (an indirect search).
Systematic Effects Systematic effects on the β rejection are now better incorporated in the background estimate Phonon and Charge sides have Different Rejection ability for the 2 primary cuts: Yield and Timing Yield-based Rejection Phonon Charge Counts More phonon side β in signal Ionization Yield
Systematic Effects Systematic effects on the β rejection are now better incorporated in the background estimate Phonon and Charge sides have Different Rejection ability for the 2 primary cuts: Yield and Timing Yield-based Rejection Phonon Charge Counts More phonon side β in signal Ionization Yield
Systematic Effects Systematic effects on the β rejection are now better incorporated in the background estimate Phonon and Charge sides have Different Rejection ability for the 2 primary cuts: Yield and Timing Yield-based Rejection Phonon Charge Timing-based Rejection Counts More phonon side β in signal Counts Phonon side better Ionization Yield Phonon Delay in µs
Systematic Effects Systematic effects on the β rejection are now better incorporated in the background estimate Phonon and Charge sides have Different Rejection ability for the 2 primary cuts: Yield and Timing Yield-based Rejection Phonon Charge Timing-based Rejection Counts More phonon side β in signal Counts Phonon side better Ionization Yield Phonon Delay in µs Phonon side βs dominate in signal region but have much better timing rejection
Scientific Reach of SuperCDMS Direct Detection complementary to collider searches for SUSY neutralinos Projected upper limits on WIMP cross section and mass for SuperCDMS experiments Gray background shape shows allowed regions of msugra parameter space for spinindependent WIMP-nucleon cross sections. Colored regions show representative models in several specific models considered by collider studies. Colored circles indicate the Linear Collider Cosmology benchmark points. Region A is the overlap region between LHC & SuperCDMS 25kg. Region B is accessible only to SuperCDMS 25 kg and Region C only to LHC.
SuperCDMS Phase A (1.4 zb) Backgrounds Background Events Before Veto After Veto Rejection Inefficiency Not Rejected Rate # Evts Rate # Evts (singles) Rate # Evts CDMS II T 1-5 at Soudan 4.0 kg 485 d (raw 1,300 kg-d) gammas n/a n/a 147 130,000 2.0E-06 4.2E-04 0.25 betas n/a n/a 0.4 370 2.1E-03 1.2E-03 0.75 neutrons (radio-nuclides) n/a n/a 2.0E-05 0.01 1 2.0E-05 0.01 neutrons (muon-induced) 1.1E-02 7 1.5E-04 0.09 1 1.5E-04 0.09 SuperCDMS ST6 1-2 at Soudan 7.5 kg 550 d (raw 2,800 kg-d) gammas n/a n/a 147 290,000 1.0E-07 2.1E-05 0.03 betas n/a n/a 0.16 320 2.5E-04 5.8E-05 0.08 neutrons (radio-nuclides) n/a n/a 2.0E-05 0.03 1 2.0E-05 0.03 neutrons (muon-induced) 1.1E-02 15 1.5E-04 0.2 1 1.5E-04 0.2 SuperCDMS ST6 1-5 at Soudan 19 kg 1200 d (raw 15000 kg-d) gammas n/a n/a 147 1,550,000 1.0E-07 2.1E-05 0.2 betas n/a n/a 0.16 1,714 2.5E-04 5.8E-05 0.4 neutrons (radio-nuclides) n/a n/a 2.0E-05 0.2 1 2.0E-05 0.2 neutrons (muon-induced) 1.1E-02 80 8.0E-05 0.5 1 8.0E-05 0.5 SuperCDMS ST6 1-7 at SNOLAB 27 kg 1000 d (raw 18000 kg-d) gammas n/a n/a 68 860,000 1.0E-07 1.0E-05 0.09 betas n/a n/a 0.16 2,000 2.5E-04 5.8E-05 0.51 neutrons (radio-nuclides) n/a n/a 1.5E-05 0.13 1 1.5E-05 0.13 neutrons (muon-induced) 6.8E-05 0.60 4.5E-07 0.004 1 5.0E-07 0.004
GLAST and Indirect Detection LAT instrument: wide field of view, fine angular/energy resolution at ~0.3-300 GeV Ideal for WIMP searches: Near galactic center Satellite halos Diffuse extragalactic... and for much non-wimp astronomy Detection very favorable in focus point models (Roszkowski et al., arxiv:0707.0622) Stringent background rejection and resolution demands => interplay with detector characterization
CDMS and Higgs-search at Tevatron M. Carena, D. Hooper and P. Skands, hep-ph/0603180 Tevatron detection of neutral Higgs with enhanced couplings to down-type fermions strongly affected by (non)-discovery of WIMPs by CDMS II. By the end of 2007, CDMS can rule out all of Tevatron future reach - or else they discover SUSY. Tevatron 2005 CDMS 2005 CDMS 2007 Tevatron 2007
What do we learn if we see a signal? Suppose we see 8 events at the rate of 1 evt per 50 kg-d of Ge Then mass & cross section determined as shown and SI vs SD determined from different targets Suggest properties to look for at LHC and future ILC actual signal CDMS Soudan 1-Tower A convincing signal would motivate large TPC such as DRIFT for velocity distribution If SUSY seen first at LHC would still want to determine if LSP is the dark matter, SO NEED TO PUSH DIRECT DETECTION EITHER WAY
Reduction of Backgrounds Reduce beta contamination via active screening/cleaning Observed alpha rate indicates dominated by 210 Pb on detectors Improved radon mitigation already in place -- will determine if it has effect Materials surface analysis (PIXE/RBS/SIMS/Auger) (in progress) Limits detector 14 C, 40 K beta contamination to ~10% of total Developing multiwire proportional counter or cloud chamber as dedicated alpha/beta screener (prototypes in progress) Necessary for 17 beta emitters that have no screenable gammas/alphas Reduce photon, beta backgrounds via improved shielding Active (inexpensive) ionization endcap detectors to shield betas, identify multiple-scatters Add inner Pb shielding (like we had at shallow Stanford site)