Hatano Y., Katsumura Y., Mozumder A. (Eds.) Charged Particle and Photon Interactions with Matter - Recent Advances, Applications, and Interfaces

Подождите немного. Документ загружается.

900 Charged Particle and Photon Interactions with Matter

Figure 31.19 Electronic image of a cosmic-ray shower as observed in the ICARUS 3 t prototype. The

drift direction is along the horizontal axis, spanning a distance of about 40 cm. The vertical axis corre-

sponds to 192 sense wires with a 2 mm pitch. (From Benetti, P. etal., Nucl. Instrum. Methods Phys. Res.

Sect. A., 332, 395, 1993b. With permission.)

Xe flash lamp

Quartz fiber

Field

ring

=50M

Grid

Faraday

cage

LAr

= 10M

= InF

–

RC >>

Waveform digitizer

Charge amplifier

HV filter

+HV

(Photo) Cathode

Anode

Figure 31.20 Layout of the purity monitor and readout scheme for the ICARUS T600 detector (Bettini

etal., 1991; Amerio etal., 2004). (From Amerio, S. et al., Nucl. Instrum. Methods Phys. Res. Sect. A, 527, 329,

2004.

With permission.)

Applications of Rare Gas Liquids to Radiation Detectors 901

are chosen to ensure complete transparency of the grids. The two grids are separated by a long

drift space. The attenuation due to electronegative impurities of an electron cloud drifting over this

distance is measured by comparing the charge leaving the cathode with that reaching the anode.

The currents from the cathode and the anode feed the same charge amplier. Free electrons are

produced by a short pulse Xe ash lamp (or UV laser) brought to a gold deposit on the cathode

through an optical ber. The electron cloud, photoproduced by the light pulse, drifts initially to

the rst grid producing a positive current in the amplier. This current vanishes at the instant the

electrons cross the cathode grid. The charge integrated by the amplier is then equal to the charge

produced at the cathode. While the charge drifts in the uniform eld between the two grids,

no current is experienced by the amplier. If electronegative impurities are present, the amount of

charge diminishes. When the charge nally crosses the anode grid, it produces a negative current

in the input of the amplier and a negative step in the output (charge) signal; the height of the step

is equal to the surviving charge (Q

) reaching the anode. The free-electron lifetime τ can simply be

expressed

as follows:

= −













exp

(31.17)

where T

is the electron drift time. Obviously, the sensitivity is higher for longer drift times, which

can

be obtained by working with low electric eld intensities.

The

performance of the LAr TPC has been studied through the analysis of a wide variety of

events

occurring in the detector (Cennini etal., 1994).

31.4.1.1

space

esolution

Cosmic muons crossing the detector have been used to evaluate the single-point space resolution

along the drift coordinate (σ

). Values of about 150μm are the norm and appear to be independent on

the eld. Instead, σ

strongly depends on the signal-to-noise ratio (≈10 in the case of ICARUS). The

space resolutions on the other two coordinates are strictly related to the wire pitch (p): σ

y ≈ p/

31.4.1.2 energy

esolution

By means of cosmic muons, the distribution of the charge deposited over 2 mm (wire pitch) was

measured. Its width is the convolution of a Landau function with the Gaussian electronic noise

and is directly related to the resolution of the energy deposited by ionization. A comparison with a

test pulse distribution shows that the noise is the main contribution to the width. Hence, an energy

resolution of ≈ 10% over a 2 mm track is the typical value in the 3 t prototype detector. An energy

resolution in the MeV range has also been evaluated by studying the Compton spectrum and the pair

production peak produced by a 4.43MeV monochromatic γ ray source. The best t gives a resolution

7% at 4

MeV

in agreement with the calorimetric measurement found in the literature.

31.4.1.3

particle

dentication

Particle identication, a vital feature, is directly related to the ability of measuring the dE/dx value

along the range of a stopping particle. A study of cosmic muons and protons stopping in the 3 t

prototype has shown that the charge deposited along the track is not proportional to the energy

deposited because of the electron–ion recombination effect that strongly depends on the ionization

density. This nonlinear response could degrade the particle identication capability. Based on the

fact that the electron–ion recombination has the result of producing UV scintillation photons, a

solution to recover linearity has been found and consists in dissolving in LAr a photosensitive dop-

ant able to convert back the scintillation light due to recombination into free electrons (see Section

31.2) (Cennini etal., 1995). Tetra-methyl-germanium (TMG) is one of the suitable dopants, because

has a large photoabsorption cross section (62

Mb)

and a large quantum yield (close to 100%); this

902 Charged Particle and Photon Interactions with Matter

implies that small quantities of TMG are enough to convert all the scintillation photons into elec-

trons in the vicinity of the ionizing track, that is, without spoiling the space resolution. Figure 31.21

illustrates

the dQ/dx versus dE/dx relation before and after the doping with TMG.

The

ICARUS project has demonstrated the feasibility of the technology by an extensive R&D

program, which included 10 years of study on small LAr volumes (proof of principle, LAr puri-

cation methods, readout schemes, and electronics) and 5 years of study with several prototypes

of increasing mass (purication technology, collection of physics events, pattern recognition, long

duration tests, and readout technology). The largest of these devices had a mass of 3 t of LAr (Benetti

etal., 1993a,b; Cennini etal., 1994) and had operated continuously for more than 4 years, collect-

ing a large sample of cosmic-ray and γ source events. Furthermore, a smaller device (50L) of LAr

(Arneodo etal., 2006) was exposed to the CERN neutrino beam, demonstrating the high recogni-

tion

capability of the technique for neutrino interaction events.

Thereafter,

the project has entered an industrial phase, a necessary path to follow toward the realiza-

tion of large volume detectors for astroparticle and neutrino physics. The rst step was the realization

of a large cryogenic prototype (14t of LAr) (Arneodo etal., 2003) to test the nal industrial solutions

that were adopted. The second step was represented by the construction of the T600 module: a detec-

tor employing about 600t of LAr to be operated at the National Laboratory of Gran Sasso (LNGS),

Italy (see Figure 31.22). This stepwise strategy allowed for the progressive development of the neces-

sary know-how to build a multi-kton LAr detector (Amerio etal., 2004). The T600 was transported to

the LNGS, where it will start taking data from atmospheric and solar neutrino interactions.

The LXe TPC is also developed to study γ ray astronomy in the energy range of MeV. The energy

and direction of an incident γ ray are reconstructed from the 3D locations and energy deposits of

individual interactions taking place in the homogeneous detector volume. While the charge signals

provide energy information and x–y positions, the fast xenon scintillation signal is used to trigger

the detector. The drift time measurement, referred to as the time of the trigger signal, gives the

z position with the known drift velocity (Aprile etal., 2008).

31.4.2 xMaSS project

An advantage of LXe is that Xe does not have an isotope with a lifetime longer than a couple of

months. The longest isotope is

127

Xe with a half-life of 36.4 days, which decays through electron cap-

ture with a Q-value of 0.664MeV. The second longest isotope is

131

Xe with the lifetime of 11.8 days.

Thus, experiments on rare phenomena (such as dark matter searches and double-beta decay) may

be carried out shortly after moving the xenon underground from the surface. Therefore, the cosmo-

genic production of Xe isotopes is not a serious problem. One of the main sources of the background

E = 200 V/cm

With TMG

(3.5 ppm)

Pure LAr

1 10

dE/dx (MeV/cm)

100

dQ/dx (el./2 mm)

Figure 31.21 dQ/dx versus dE/dx from stopping muons and protons. Charge saturation is evident in pure

LAr. Linearity is recovered with TMG. (From Cennini, P. etal., Nucl. Instrum. Methods Phys. Res. Sect. A.,

355,

660, 1995. With permission.)

Applications of Rare Gas Liquids to Radiation Detectors 903

is the presence of trace amounts of radioactive impurities, such as

Kr. The commercially available

xenon contains ∼1ppb ∼ppm of Kr, and such purity is sufcient for most of the applications of Xe.

Kr is a radioactive nucleus that decays into

Rb with a half-life of 10.76 years, and emits γrays

with a maximum energy of 687keV and a 99.57% branching ratio. Large quantities of

Kr are

produced articially in nuclear ssion. Xe is usually abstracted from air. The concentration of Xe

in air is ∼0.1 ppm, while the concentration of Kr is ∼1ppm. This constitutes the main source of

in air. The concentration of

Kr in air is measured to be ∼1 Bq/m

(Igarashi etal., 2000; Cauwels

etal., 2001), which corresponds to

Kr/Kr = ∼10

−11

. The possible methods for removing Kr from Xe

are distillation and adsorption. These are commonly used for industrial processes, but systems that

could reduce Kr sufciently to meet the requirements of experiments never existed. Further studies

for both adsorption-based chromatography and distillation followed the McCabe–Thiele method

that succeeded in reaching ∼ppt levels (McCabe and Smith, 1976; Bolozdynya etal., 2007).

The XMASS project aims at a multipurpose detection, which observes not only dark matter,

but also pp and

Be solar neutrinos, which the Super-Kamiokande cannot observe, as well as

136

neutrinoless double-beta decay to measure the neutrino mass, with a 20 t LXe scintillation detector

in the Kamioka observatory. The possibility of using ultrapure LXe as a low-energy solar neutrino

detector by means of ν + e scatterings has been evaluated. A detector possibly with 10t of ducial

volume* will give ∼14 events for pp neutrinos and ∼6 events for

Be neutrinos with the energy thresh-

old at 50keV. The high density and high Z of LXe would provide very good self-shields against the

incoming backgrounds originating from the container and outer environments (PMTs, cables, etc.).

A 30cm thick self-shield is equivalent to 4m of water shield for the electromagnetic component.

* Fiducial volume (or mass) is the effective volume (or mass) for particle detection in a 3D detector. The events that happen

the volume are considered to be reliable.

Wires of the TPC

VUV-sensitive PMTs

Drift length (1.5 m)

Cathode

Field-shaping electrodes

(during installation)

Figure 31.22 Picture of the internal detector layout inside the half-module: the cathode divides the volume

in two symmetric sectors. The picture refers to the left sector where wires and the mechanical structure of the

TPC and some PMTs are visible. (From Amerio, S. etal., Nucl. Instrum. Methods Phys. Res. Sect. A., 527,

329,

2004. With permission.)

904 Charged Particle and Photon Interactions with Matter

In the rst phase of the XMASS program, a 100 kg prototype for dark matter detection was built

and deployed at the Kamioka underground laboratory. The 30cm cubic detector was contained in

a low-activity copper vessel further shielded for the reduction of γ and neutron backgrounds. Nine

PMTs are attached on each face of the vessel, and scintillation photons are detected by a total of 54

low-background PMTs of 2in. (Hamamatsu R8778) through MgF

windows. The PMT photocath-

ode coverage for light detection was 17%. The background spectra were measured before and after

the distillation, as shown in Figure 31.23. The background rate around 100–300keV before distilla-

tion is higher than that after distillation by ∼5 × 10

−2

kg/keV/day. This background level corresponds

to a Kr concentration of 2–3ppb, assuming a

Kr/Kr ratio of 1.2 × 10

−11

. The dashed curve shows

the expected

Kr β spectrum, assuming a Kr concentration of 3ppb. The Kr concentration of puri-

ed Xe was measured to be 3.3 ± 1.1 ppt using a gas chromatography apparatus and an atmospheric

pressure ionization (API) mass spectrometer. The results conrmed that the cryogenic distillation

system can reduce the amount of Kr in Xe gas by up to three orders of magnitude (Abe etal., 2009).

The second phase, currently under construction (Suzuki, 2008), consists of an 856kg LXe detec-

tor with the main goal of physically detecting dark matter (Figure 31.24). The large liquid volume

is viewed by 642 hexagonal PMTs, arranged in an approximately spherical shape (radius ∼43cm)

with 67% photocathode coverage. A light yield of 4.4 photoelectrons/keV was predicted by simula-

tions. The ducial volume is a region inside the radius of 20 cm from the center, and the mass

of LXe is 100 kg. A new low-background PMT, Hamamatsu R10789, with a hexagonal photo-

cathode shape was developed, and the total radioactive contamination was reduced by a factor

of 10 from R8778 (used for the 100 kg prototype). The measured activities per PMT are 0.37 ×

−3

Bq for the U-chain, 1.0 × 10

−3

Bq for the Th-chain, 5.9 × 10

−3

Bq for

K, and 3.1× 10

−3

Bq for

Co. In order to reduce the environmental radioactive background, such as those due to γrays

and fast neutrons, the detector will have to be installed in the center of a water tank (10 m in

diameter and 10 m high). Twenty-inch PMTs will have to be arranged on the inner wall of the

tank to observe water Cherenkov signals for anti-coincidence counting. The sensitivity of the

–1

–2

200 400 600 800 1000 1200

Energy (keV)

Event rate (1/keV/kg/day)

1400 1600 1800 2000

Figure 31.23 The background spectra measured before distillation (solid histogram) and after distillation

(dotted histogram). The dashed curve shows the expected

Kr beta spectrum assuming a Kr concentration of

3ppt.

(From Abe, K. etal., Astropart. Phys., 31, 290, 2009. With permission.)

Applications of Rare Gas Liquids to Radiation Detectors 905

detector for 5 years of running is expected to reach ∼10

−45

for spin-independent interactions

and ∼10

−39

for spin-dependent interactions.

A similar experiment using liquid neon (LNe) is also proposed by the cryogenic low energy

astrophysics

with noble gases (CLEAN) project (Coakley and McKinsey, 2004).

31.4.3 dark Matter Search

Most of the constituent matter in the universe is dark with no emission, no absorption, and even no

scattering of light. It is, in fact, transparent. In 1933, Zwicky observed that the rotational velocity of

galaxies in the Coma cluster is greater than that expected from the virial theorem, and announced

that the universe contains more matter than is inferred from optical observation (Zwicky, 1933).

Further scientic evidence conrmed his view: the rotational velocity in a galaxy, gravitational

lensing, the microwave radiation background, and even the shape of thin galaxies require unseen

mass to prevent a galaxy from collapsing. The unseen dark matter accounts for a quarter of the uni-

verse.

Ordinary matter makes up only 4%. The rest, ∼73%, is supposed to be dark energy.

The

leading candidate is WIMPs (weakly interacting massive particles), cold massive elementary

particles with a typical mass of several tens of GeV. The supersymmetric theorem (SUSY) pre-

dicts a neutralino as a favorable candidate (Ellis and Flores, 1991). Review articles for dark matter

candidates and other dark matter detectors are found in Lewin and Smith (1996), Sumner (2002),

Freedman

and Turner (2003), Gaitskell (2004), and Spooner (2007).

Our

solar system travels around the galactic center at about 230 km/s (Spergel, 1988). WIMPs

are surrounding the Galaxy. Scientists are in search of recoil nuclei of a few tens of keV energy,

produced by elastic scattering with WIMPs. However, the event is quite rare (<10

−2

to 10

−4

/kg/day).

Almost all the signals produced even in an underground detector are due to the background,

mostly γ rays. The resulting kinetic energy spectrum for recoil ions will not be monochromatic,

but may be described as similar to the exponential. Both the exposure (mass × time) and the

ability to distinguish the WIMP signal from the background is essential. RGLs have an excellent

capability of discriminating nuclear recoil signals from the γ background using both the ioniza-

tion and the scintillation, and can provide a strong possibility of the direct observation of dark

matter in the Galaxy.

XMASS at Kamioka mine

Mt. Ikenoyama 1000 m deep

21 m

LXe

15 m

PMT

Figure 31.24 The XMASS detector in Kamioka. An LXe detector (bottom left) is set in the center of the

water tank. PMTs of 20 in. are installed on the wall of the water tank. LXe is surrounded by 642 PMTs. The

whole detector is installed 1000m below the mountain. (

Kamioka Observatory, Institute for Cosmic Ray

Research

(ICRR). The University of Tokyo, Tokyo, Japan.)

906 Charged Particle and Photon Interactions with Matter

31.4.3.1 interaction of slow ions with matter

The interaction mechanism for recoil ions of a few tens of keV energy with matter is considerably

different from that of fast ions described in Section 31.2. The nuclear stopping, S

, which can be

ignored in fast ion collisions, becomes comparable to or larger than the electronic stopping, S

. The

total stopping, S

, can be expressed as the sum of the two. Apart from the difculty in obtaining

values for S

and S

for slow collisions, a simple relation, S

= S

+ S

, produces another complex-

ity. The interaction of slow ions with matter has been extensively studied by Lindhard etal. (1963).

The nuclear process follows the usual procedure of a screened Rutherford scattering. The electronic

interaction is based on the Thomas–Fermi model. S

is expressed as (dε/dρ)

= kε

1/2

, where ε and ρ

are the dimensionless reduced energy and reduced range. When the projectile and the target are the

same element (Z

= Z

; indices 1 and 2 stand for the projectile and the target atom, respectively), k

given by the atomic number Z and mass A, as follows:

k Z A=

−

0 133

2 3

1 2

/ /

(31.18)

The secondary ions will enter into the collision processes again, and this will go on. After the

cascade processes for stopping collisions take place, most of the ion energy is spent in the atomic

motion, ν, and wasted as heat in ordinary detectors. Only the energy η is utilized in electronic exci-

tation and contributes to the ionization or scintillation. The ratio η/ε is called the nuclear quenching

factor, q

, or the Lindhard factor. The asymptotic equation for ν (=ε − η) for recoil ions in a homo-

nuclear

medium (Z

= Z

) is given as follows (Lindhard etal., 1963):

+ ⋅1 k g( )

(31.19)

approximate expression for g(ε) is given by (Lewin and Smith, 1996)

g( ) .

. .

ε ε ε ε= + +3 0 7

0 15 0 6

(31.20)

Once a value for q

is obtained, it can be used for the study of the track structure and quenching.

The electronic LET (LET

= −dη/dR), that is, the electronic energy deposited per unit length along

the ion track, can be an important parameter to consider the radiation effects due to slow recoil

ions (Hitachi, 2005, 2008). The value of LET

obtained for recoil ions is close to the LET value for

α-particles. This makes α-particles a good measure for decay times, the S/T ratio, etc., for recoil

ions,

as discussed in Section 31.2.

The

excitation density in the recoil ion track can be as high as that in an α-track; therefore,

scintillation quenching, q

, is expected, as discussed in Section 31.2. The total quenching factor,

, is expressed as q

× q

. The track structure can be considered by using LET

and ion velocity.

Then, the value of q

is estimated as discussed in Section 31.2. We have T = q

E in Equation 31.9.

Most of the δ rays produced by recoil ions do not have sufcient energy to escape from the core

and form an undifferentiated core, that is, T

/T = 1 for recoil ions. The quenching factors calculated

for recoil ions in LAr and LXe (Hitachi, 2005) are shown in Figure 31.25 together with reported

experimental values (Arneodo etal., 2000; Bernabei et al. 2002; Akimov etal., 2002; Aprile etal.,

2005; Chepel etal., 2006).

Practically, the recoil ion-to-γ ratio, RN/γ, is a good parameter for the design and development of

a detector for dark matter searches, and it can be expressed as (Hitachi, 2005)

RN q q

Lγ

nc el

⋅

(31.21)

Applications of Rare Gas Liquids to Radiation Detectors 907

where L

is the scintillation efciency for γ rays. It is less than 1 because of a reduced efciency

for the electron–ion recombination (see Figure 31.7). The value of q

is larger for light atoms as

shown in Table 31.2. However, the value of W

, the (electronic) energy required to produce one

photon, is larger for light atoms. These factors compensate each other. As a result, the number

of photons, N

, per keV is typically 10–13 at about 50 keV, and it does not change much for LAr

and LXe, as shown in Table 31.2. q

for LNe was estimated simply by taking into consideration

LET

and the scintillation efciency. The N

value estimated for LNe is small mainly because

of a large W

value due to a low luminescence yield and electronic quenching. The main contri-

bution to the error in RN/γ comes from an uncertainty in L

. Experimental values are shown in

Figure 31.25a for LXe. The values reported are 0.26 for 387keV recoil ions in LNe (Nikkel etal.,

2008) and 0.28 for 65 keV recoil ions in LAr (Brunetti etal., 2005), and compare well with the

estimated values.

Natural Ar and Xe are contaminated with

222

Rn. Daughter nuclei of Rn are produced in the ion-

ized state, and drift to the electrodes or attach onto the wall. Their decay produces αs, βs, and very

heavy recoil ions (A ∼ 200). The quenching factor estimated by using the numerical calculation and

the low-power approximation (Lindhard etal., 1963) for very heavy recoil ions in LAr is shown in

Figure 31.25b. Very heavy ions of 100–200keV recoiled in α-decay give a charge distribution quite

similar to that for ions recoiled by WIMPs in detector media (Hitachi, 2008).

0.5

0.4

0.3

0.2

0.1

0.0

0 20 40 60

Energy (keV)

Quenching factor q

Recoil/γ ratio

80 100 120

0.5

0.4

0.3

0.2

0.1

(k = 0.165)

TTL

Liquid Xe

Liquid Ar

(a)

0.5

0.6

0.4

0.3

0.2

0.1

0.0

0.5

0.6

0.7

0.4

0.3

0.2

0.1

Quenching factor

Energy (keV)

RN/γ ratio

TTL

(gas)

Ar ions

Numerical

Power low appr.

Recoil ion by α-decay

0 50 100

(b)

150 200

Figure 31.25 The nuclear, q

, and the total, q

TTL

= q

× q

, quenching factors as a function of recoil ion

energy in (a) LXe (Hitachi, 2005) and in (b) LAr. The values of RN/γ ratios are shown on the right axis. (a) The

points on the graph denote experimental values (Arneodo etal., 2000; Akimov etal., 2002; Bernabei etal.,

2002; Aprile etal., 2005; Chepel etal., 2006). (b) q values estimated for very heavy recoil ions (Pb) in α-decay

are

also shown for 80–180

keV,

•

denotes experimental q

values in gas.

908 Charged Particle and Photon Interactions with Matter

31.4.3.2 detectors for dark matter searches

Dark matter detectors are installed deep underground to avoid cosmic rays, as shown, for example,

in Figure 31.24 for the XMASS dark matter detector being constructed at Kamioka mine. High-

energy neutrons from muon spallation in the rock are shielded by a water tank. The main detec-

tor is an 80 cm diameter 800 kg LXe. XMASS relies on shielding for background rejection. The

outer layer of 20cm LXe is used as an active shield. XMASS is a multipurpose detector aiming

also at solar neutrinos and double-beta decay, as discussed in detail in Section 31.4.2. The nal

backgrounds are low-energy neutrinos or double-beta decay. Theoretically, γ’s and recoil ions can

be separated by observing the difference in the decay shape, as discussed in Section 31.2. A wave-

form analysis has been tested in LXe (ZEPLIN I, Alner, 2005). However, it is difcult to separate

these effectively by the scintillation decay in LXe. The reason may be the fact that the number of

photoelectrons observed is not enough. Also, the main difference in decay is in the recombination

process, which can change with particle energy in LXe. Other LXe dark matter detectors rely on the

scintillation

and the charge for particle identication.

The

discrimination by the decay shape is possible in LAr and LNe. The decay does not show

the recombination component but the singlet and triplet states. The decay times for these states dif-

fer by several orders of magnitude. This difference can be exploited for particle identication and

background rejection. Both, the differences in scintillation decay, and scintillation and charge yields

for γ’s and recoil ions can be simultaneously used in LAr. The principle for a dual-phase detector, a

TPC with photon detection capability, is described in the following text.

The WArP prototype detector (WARP, 2008) uses 2.3L (3.2kg) of LAr and is in operation in the

INFN LNGS. WArP uses a large difference in decay times in LAr, namely, 7ns and 1.6μs (Section

31.2.2), for the singlet and triplet states, respectively. The wavelength for LAr scintillation is quite short,

and so, a wavelength shifter, such as tetra-phenyl-butadiene (TPB), is needed. The presence of cosmic

Ar is considered to be a major source of background.

Ar is a 565keV β emitter.

Ar-depleted Ar

is available via centrifugation or thermal diffusion; however, the method is expensive, particularly for

a ton-scale detector. It has been found that Ar contained in underground wells shows a highly sup-

pressed

Ar contamination (WARP, 2008). The 140kg active target detector is under construction at

LNGS. The detector is surrounded by a neutron shield consisting of 70cm thick polyethylene and a 4π

active neutron veto (8t LAr with 303 PMTs). The inner dual-phase detector provides the x–y position

by 37 PMTs. The z-coordinate is given by the time delay between S1 and S2 signals, as discussed in

the following text. The particle discrimination capability using a waveform is shown in Figure 31.26.

ArDM (Rubbia, 2006; Otyugova, 2008) based on the dual-phase method and a system of large

electron multiplier (LEM) is implemented for electron multiplication in the gas phase. ArDM is in the

research and development stage. A ton-scale detector has a LAr target volume of 850kg with a vertical

drift length of 120cm. A voltage up to ∼400kV is supplied by a Greinacher (Cockroft–Walton) chain.

table 31.2

estimated

cintillation

for r

ecoil

ons

in rgl

liquid units lne lar lxe

Energy keV 20 40 60

Range

μm

0.1 0.13 0.07

= η/E

0.4 0.37 0.28

= q

× q

(0.3) 0.22 0.19

RN/γ

(0.3) 0.26 0.24

Photon/keV (2) 11 13

Note: Values

in parenthesis obtained by assuming L

= 1.

Hitachi (2005).

Applications of Rare Gas Liquids to Radiation Detectors 909

The DEAP/CLEAN (Boulay and Hime, 2006) detectors at SNOLAB, located in an active nickel

mine in Sudbury, Canada, observe only the scintillation light from LAr or LNe. DEAP-1 uses 7 kg

of LAr viewed by 2 PMTs of 5in. LNe contains no long-lived radio isotopes and is easily puried

using cold-traps. Mini-CLEAN is a planned WIMP detector containing 100 kg of LNe or LAr

viewed by 32 PMTs (McKinsey, 2007). The ability to exchange the liquids having different sensi-

tivities with WIMPs and fast neutrons will allow both of these event populations to be distinguished

and

characterized.

typical liquid–gas dual-phase detector, XMASS II, for WIMP searches is shown in Figure

31.27. It is a position-sensitive LXe TPC having scintillation-observing capability. The difference in

scintillation and charge yields due to γ’s and recoil ions (Figure 31.3) is used for particle identica-

tion (Benetti etal., 1993a,b; Aprile etal., 2006). The liquid phase is basically a 3D detector with an

allay of PMTs for scintillation detection. It uses the primary scintillation, S1, and the charge signal

2.5

1.5

0.5

–0.5

–1

0.1 0.2

(a) (b)

0.3

Neutron-induced ion recoils WIMP exposure of 96.5 kg day

0.4

Log (S2/S1)

2.5

1.5

0.5

–0.5

–1

0.5

40–60 keV

>2.8 × 10

triggers

0.6

Pulse shape discrimination parameter (F) Pulse shape discrimination parameter (F)

0.7 0.8 0.9 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Figure 31.26 Particle discrimination capability in a LAr dark matter detector using the scintillation pulse

shape (a) obtained with an Am–Be source and (b) obtained from WIMP exposure. The box shown is indicative

(WARP,

2008). (From Benetti, P. et al., Astroparticle Physics, 28, 495, 2008. With permission.)

Grid

Liquid surface

Anode

MgF

(cathode)

MgF

(cathode)

160 mm

220 mm

165 mm

LXe

PMT

(HAMAMTSU

R8778×7)

Gas Xe

15 kg dual-phase Xe detector

Grid

OFHC

PTFE

(Field-shaping ring)

Figure 31.27 Schematic view of XMASS II showing a dual-phase LXe detector.