Dark Matter Detection
with Liquid Xenon
Masahiro Morii
Harvard University
Laboratory for Particle
Physics and Cosmology
21 August 2009
1
Dark Matter
Existence of Dark Matter is well established
from its gravitational effects


Coma cluster [Zwicky], Galaxy rotation curve [Rubin]
Weak gravitational lensing, Bullet cluster
Amount of Dark Matter is inferred from cosmological data


~22% of the energy of the Universe
Local density 0.3 GeV/cm3
Identity of Dark Matter is unknown


Majority must be cold and non-baryonic
i.e. made of particles that are not a part of the SM
Dark Matter is a particle physics problem
as much as a cosmology problem
21 August 2009
Dark Matter
2
WIMP Dark Matter
No shortage of candidates, but…
WIMPs are the front runners



~100 GeV new particles with weak
(and gravitational) interactions
Such a particle would naturally have
the right thermal relic density
Predicted in many BSM theories
(e.g. the LSP)
Since the annihilation cross section
σ(χχ → ff ) is constrained by the
relic density, we can predict:
σ(χf → χf )
Direct detection
σ(ff → χχ)
Production at colliders
21 August 2009
Dark Matter
3
Direct WIMP Detection
Best limits on the WIMP-nucleon
cross section are ~5x10-44 cm2


CDMS : Ge and Si crystals at 10 mK, 121 kg-day exposure
XENON10 : liquid Xe, 136 kg-day
For LSPs, the interesting region is
around 10-44 cm2

Smaller cross sections possible, but
increasingly difficult to reconcile with
the flavor problem
Next generation of experiments
aim for <10-45 cm2
21 August 2009
Dark Matter
4
Liquid Xenon
A2
WIMP-nucleus cross section

Xe (A = 131.3) gives high signal rate
100 kg-year exposure can probe
σ(WIMP-p) < 10-45 cm2
Key liquid Xe properties



High density: 3 g/cm3
High boiling point: 165K
Good scintillator: 42 photons/keV


High ionization yield: W = 15.6 eV


λ = 175 nm easy to detect with PMTs
High electron mobility, low diffusion
No long-lived radioactive isotopes besides double-beta decays
  85Kr
must be removed by charcoal chromatography
21 August 2009
Dark Matter
5
Two-Phase Xe Detector
PMTs collect prompt (S1) and
proportional (S2) light signals


S1-S2 delay  Drift length
S2 light pattern  Horizontal location
S2/S1 ratio differs markedly between
electron and nuclear recoil

>98.5% rejection of EM backgrounds
Good scaling to larger masses



1 m3 holds 3 tonnes
Instrumentation (mass)2/3
Backgrounds improve with size due to
self shielding
21 August 2009
Dark Matter
6
LUX Experiment
LUX is a 350 kg (100 kg fiducial) liquid Xe experiment

Located in the Davis cavern, Sanford Underground Lab in Homestake, SD
XENON10 technology has been
improved to achieve <1 bkgd. in
100 kg-year




Xe purification system has 300 kg/day throughput using a heat exchanger
Ultra-low activity Ti vacuum vessel replaces SS + Cu
PMTs have low activity (9/3 mBq of U/Th per tube) and high QE (27%)
183 m3 purified water tank shields the detector from neutrons
Recoil energy threshold <5 keV  σ(WIMP-p) = 5x10-46 cm2
21 August 2009
Dark Matter
7
LUX Collaboration
Brown, Case Western, LBNL, Harvard,
LLNL, Maryland, Texas A&M, Rochester,
South Dakota, Yale
Funded by DOE & NSF
21 August 2009
Dark Matter
8
Harvard Group
Harvard joined LUX in June 2009

Morii (50%) is the PI
Took up a critical-path item: post-amplifier


preamp
PMT
120 channels of receiver-amplifier-shaper for the PMT signals
Full system is needed in November
postamp
Harvard took over production from UC Davis


Recruiting a postdoc and 1–2 graduate students


Will take part in detector integration, commissioning
Develop analysis software framework
21 August 2009
Dark Matter
Analog Trigger
Digital Trigger

FADC

Oliver and Morii improved the LLNL design
New LPPC engineer, Meghna Kundoor, working on testing
Components in hand. PC boards in fabrication
On track for November delivery
9
LUX Status and Schedule
Prototype LUX0.1 is operating at Case



1 liter of liquid Xe viewed by 4 PMTs
Test cryogenics and Xe purification system
>1 m electron drift achieved in 3 days
Assembly of LUX in Sanford surface
building will start in November


All major components are in hand
Building is being fitted out
Fully-assembled LUX lowered to
Davis cavern (4,850 ft) in Spring
2010

21 August 2009
Dark Matter
Dark Matter search will start!
10
LZ Proposal
LZ = LUX scaled up to 1500 kg (1200 kg fiducial)


Joint collaboration of LUX and ZEPLIN-III
LUX infrastructure designed to accommodate LZ
σ(WIMP-p) = 2x10-47 cm2 in 2 years

2000-fold improvement over current limits
Cost of liquid Xe ~$1000/kg

Maximize the fiducial/total mass ratio by rejecting single-scatter γ-ray background with liquid scintillator
Harvard will assume larger responsibilities


Development of low radioactivity, high-QE PMT
Complete analog electronics chain (pre + postamp)
MRI-R2 proposal submitted this month

3-year construction  Data taking in 2013
21 August 2009
Dark Matter
11
Summary and Prospect
Exciting time for Dark Matter detection
Cosmology points us to compelling particle physics
  Liquid Xe technology has the potential for first observation

Harvard is entering DM hunting with strong commitment
Producing critical component for the LUX experiment
  PMTs and analog electronics for the proposed LZ experiment

Discovery potential of LUX is excellent
σ(WIMP-p) = 5x10-46 cm2 covers the SUSY-favored region
  Dark Matter search run will start in 2010

LZ will push the sensitivity to 2x10-47 cm2 by 2015
21 August 2009
Dark Matter
12