Earth-based dark matter searches

38  Download (0)

Full text


Dark matter

Detection methods



Ionisation (Charge)

Scintillation (Light) Phonons




dark energy (73%)

CDM (23%)

Atoms (4%)

Does dark matter matter?

WMAP best fit


'Matter not visible to us because it emits no radiation that we can observe, but it is detectable gravitationally.'

What is dark matter?


Dark matter Not dark matter


Particlephysicists say:

Most probable dark matter candidates are

W eakly I nteracting M assive P articles

in particular the lightest supersymmetric particle

the neutralino χ




is ...produced in the Big Bang ...very long-lived or stable a massive neutrino (~ 100 mProton)

...very weakly interacting

1015 through a human body each day only < 10 interact

the rest is passing through unaffected

...highly abundant (~ 10 - 100 WIMPs per liter) ...moving at ~ 230 km/s w.r.t us

χ has exactly the right properties to be the dark matter


Detection methods:


Indirect searches

Direct searches



It is hoped/expected that the LHC (planned start-up in 2007) will be able to produce neutralinos,

however, it will not be able to constrain its cross section and identify it as the cosmologically observed dark matter.

Will deliver important information required to fully identify and

understand the particle.

A section of the LHC-accelerator in the testlab


Indirect searches

WIMPs are expected to be gravitationally captured in high density regions as galactic centers, the center of the Sun or the Earth.

Indirect dark matter searches assume that the WIMP is its own antiparticle (as predicted by SUSY models) or that equal

numbers of WIMPs and anti-WIMPs are present.

Therefore one should be able to detect a larger WIMP

annihilations signal from such regions, which can be manifested as a flux of gamma-rays, neutrinos or antimatter.

See other talks for information on the detection methods.

The Milky Way in gamma-rays


M ajor A tmospheric G amma I maging C herenkov


Direct searches


WIMPs scatter elastically with target nuclei in the detector.


Expected counts per recoil energy:

with and

Detector physics F2 form factor

mn target nucleus mass NT number of target nuclei

ER recoil energy minor uncertainties


n00 /mw local halo density

f(v) WIMP velocity distr.

vmax escape velocity


Particle physics mw WIMP mass

σ0 WIMP elastic scattering cross section



Ionisation (Charge)

≈ 20% detected energy

Scintillation (Light)

≈ few % detected energy

Phonons (Heat)

≈ 100% detected

energy Recoil

How to detect a nuclear recoil:


Difficulties with direct detection experiments:

The recoil energy is small (tens of keV)

Since the WIMP signal is expected to originate from elastic scattering, a featureless, quasi exponentially decreasing energy spectrum will result.

Detector needs to have an as low energy threshold as possible

The interaction rate is low

Need high target masses in order to gain a sufficient statistic in a reasonable life-time of the experiment.

Extremely high background





β γ γ

γ γ γ

γ γ


γ γ

γ γ


γ γ β

β β

β β

β β

β β ββ


β β

β β β


β γ γ

γ γ

γ γ


γ γ

γ γ γ

n n

n n n

Backgrounds for dark matter experiments:

electrons (β) photons (γ)

neutrons (n) nuclear recoil


electron recoil


Ionisation (Charge)

Scintillation (Light) Phonons


Combining two detection methods one can distinguish between nuclear and electron recoil

Can reduce the background to neutrons


Neutron sources:

Neutrons from detector components ultra-pure materials, veto

Neutrons from rock

passive shielding Neutrons from radon

Radon removal, gas-tight sealing and ventilation with radon-free air, careful treatment at the surface

Neutrons from cosmic-ray muons large depth, veto


Cosmic-ray muons

Image of the shadow of the Moon in muons as produced by the 700 m subterranean Soudan 2 detector in the Soudan mine in Minnesota.


Need an internal shielding close to the detector to protect it from the background (from rock, ...).

Because of its high atomic number and very low intrinsic radioactivity lead provides good shielding. Lead extracted from mines is however polluted by naturally radioactive

elements such as uranium and thorium which decay into Pb210. Can clean the lead from everything but Pb210.

Roman lead

has been extracted from the grounds several centuries ago.

Pb210 has decayed (thalf ~ 22 yrs).

This lead shows extremely low activity.

Ingots of 'less archeological interest' of an ancient wreck that run agound off the Sept Iles (France).




Ionisation (Charge)

Scintillation (Light) Phonons



CRESST, Rosebud




C ryogenic D ark M atter S earch

at the Soudan lab in Minnesota


C ryogenic D ark M atter S earch


250 g Ge or 100 g Si crystal 1 cm thick x 7.5 cm diameter

ZIP detector in its mount

Qinner Qouter A


C Rbias Ibias

SQUID array Phonon D



Z-sensitive Ionisation and Phonon mediated detector


WIMP hits a Ge/Si nucleus the nucleus recoils and vibrates the whole crystal.

This breaks Cooper pairs in the thin aluminum layer which build quasiparticles that diffuse through the Al-fins and are trapped in the tungsten TES where they release their binding energy to the W electrons T raises R increases

Phonon detection

ZIP with 4144 QETs

(quasiparticle-trap-assisted electrothermal-feedback transition-edge sensors)

380 μm x 55 μm Al fins (300 nm thick) 250 μm x 1 μm W

(35 nm thick)

passive tungsten grid


WIMP hits a Ge/Si nucleus breaks up e--hole pairs in the crystal; these get separated by E-field charge is collected by electrodes on the surface of the crystal

Events within a few μm of the surface experience a deficit in the charge collection (due to the detectors 'dead layer'). More on that soon...

Ionization signal

holes electrons

TES side




Discrimination capability I

gamma & neutron source (252Cf) only gamma source (60Co) no missidentification above the detector threshold (10.4 keV)


Discrimination capability II

Particles (electrons) that interact in the surface 'dead layer' of the detector result in reduced ioniza- tion yield and could be missiden- tified as nuclear recoils.

Use phonon rise time for discrimination of surface electrons

This eliminates almost all electron recoils while keeping most of the nuclear recoils.

Combined with ionization yield rejects

> 99.9999% of gammas

> 99% of 'betas'

neutrons gammas

surface electrons


ZIP 1 (Ge) ZIP 2 (Ge) ZIP 3 (Ge) ZIP 4 (Si) ZIP 5 (Ge) ZIP 6 (Si)

SQUID cards FET cards

4 K 0.6 K 0.06 K 0.02 K

Tower 1 Arrange 6 detectors in a tower check for multiple scattering (characteristic for background) Use both Ge and Si detectors

WIMPs cross-section ~ 5 times higher in Ge than Si but neutron cross-section similiar

CDMS II project

manager Dan Bauer of Fermilab holding one of the towers


Use several towers simultaniously

increase interaction probability/time

1st tower took data Oct 03 – Jan 04 March – Aug 04: 2 towers

Starting summer 2005:

tower 1-5

19 Ge (4.75 kg) & 11 Si (1.1 kg) Tower 6 & 7 expected to be

ready in 2008

State-of-the-art technology!


The cryostat uses six nested layers to cool the detectors to <50 mK.

This reduces the background vibrations of the detector's atoms and makes them more sensitive to individual particle collisions.



"3Helium – 4Helium" dilution refrigerator

Energy is required to transport helium-3 atoms from the 3He-rich phase into the 3He-poor phase. Continuous crosssing of this boundary

Cooling of the mixture


outer polyethylene shield

Different layers of shielding around the icebox

muon veto paddles

inner lead shield

inner polyethylene shield

outer lead shield





The future: SuperCDMS

Phase A: 25 kg (2008/ $50M) Phase B: 150 kg (2010/ $70M) Phase C: 1000 kg (2015/ $500M) (compare to < 6 kg today)

SuperCDMS is approved to be installed in SNOLab. A space is approved and ready for occupancy in 2007.

A factor of 1000 reduction in muon flux

= 1000 x less neutrons


particle DA rk MA tter searches

with highly radiopure scintillators at Gran Sasso


particle DA rk MA tter searches

with highly radiopure scintillators at Gran Sasso

9 scintillating thallium-doped sodium iodide (NaI) crystals of 9.7 kg each electrons and nuclei recoiling after a

collision cause emissions of photons that are detected using photomultiplier


electromagnetic background rejection by pulse shape discrimination (only)



60º 230 km/s 30 km/s

Milky Way Sun


Annual modulation

larger WIMP flux in June & smaller WIMP flux in December


Requirements of the annual modulation:

Modulated rate according cosine In a definite low energy range With a proper period (1 year)

With a proper phase (about June 2nd) For single hit in a mulit-detector set-up With modulated amplitude in the region

of maximal sensitivity < 7%

DAMA observed annual modulation over 7 years

P(A=0) = 7 x 10-4

Bernabei et al. astro-ph/0307403


From the DAMA homepage:


If you can bear to hear the truth you've spoken twisted by knaves to make a trap for fools,

...'ll be a Man my son!

(from "IF" by R. Kipling)


results seem to be ruled out by all other experiments



Upper limits starting to get into the interesting regions in mWIMP-σ-space In the near future one will be able to rule out certain SUSY-theories.

High demands on instrumentation!




Related subjects :