PatrickScottMay152007KTH5A5461ExperimentalTechniquesinParticleAstrophysics DirectDetectionofDarkMatterFollow-up

(1)

Ionisation Signal in Liquid Noble Gases Velocity-dependent WIMP-nucleus cross-sections The Ultimate Detector Halo Models

Direct Detection of Dark Matter Follow-up

Patrick Scott

May 15 2007

KTH 5A5461

Experimental Techniques in Particle Astrophysics

May ’07 KTH 5A5461 Direct Detection of Dark Matter: Follow-up

(2)

Follow-up Topics

1

Ionisation Signal in Liquid Noble Gases

2

Velocity-dependent WIMP-nucleus cross-sections

3

The Ultimate Detector

4

Halo Models

(3)

Q. Does the ionisation signal in a liquid noble gas detector arise from fluorescence in the gas phase?

(4)

Q. Does the ionisation signal in a liquid noble gas detector arise from fluorescence in the gas phase?

A. YES

(5)

Q. Does the ionisation signal in a liquid noble gas detector arise from fluorescence in the gas phase?

A. YES

. . . more or less. . .

(6)

Q. Does the ionisation signal in a liquid noble gas detector arise from fluorescence in the gas phase?

so electrons are accelerated to produce light, but by colliding with and exciting/ionising gas atoms, not via bremsstrahlung

May ’07 KTH 5A5461 Direct Detection of Dark Matter: Follow-up Refer to Aprile et al. 2002, astro-ph/0207670 and Bolozdynya 1999, NIM A 422:314

electroluminescence =

production of light by passing a current through something proportional scintillation =

fluorescence caused by collisions with electrons, amplified by accelerating them using a strong current = one way of producing electroluminescence

Figure 3. The LXeTPC module for XENON: Schematic Design of the 100 Kg detector and its components.

Xe target is formed by a sandwich of Teflon spacers as UV diffuse reflector and copper rings for electric field shaping. The structure is closed at the bottom by a copper plate. The inside of this plate is coated with CsI as photocathode to convert Xe scintillation photons into free electric charges. The structure forms a 30 cm high cylinder with 38 cm inner diameter, holding about 100 kg of ultra pure liquid xenon.

On the top, the structure is hermetically sealed to a cylindrical copper vessel of larger diameter, housing the PMTs and the wire structure for the proportional scintillation process in the gas phase.

The LXeTPC structure is enclosed in a copper vessel containing the liquid xenon for active shielding. With both detectors at the same temperature and similar pressure, the amount of material for the inner detector walls is minimized. The scintillation light from the shield section is viewed by two rings of 16 PMTs each.

Xenon01: submitted to World Scientific on August 1, 2002 5

(7)

Q. What is the current theoretical status of

velocity-dependent WIMP-nucleus cross-sections?

A. In general SUSY, the neutralino does indeed have a velocity-dependent coupling to quarks:

L

eff

= ¯ χ ˜

⁰1

γ

^µ

γ

5

χ ˜

⁰1

¯ q

i

γ

µ

(α

1i

+ α

2i

γ

5

) q

i

+ α

3i

χ ¯ ˜

⁰1

χ ˜

⁰1

¯ q

i

q

i

+ α

4i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

γ

5

q

i

+ α

5i

χ ¯ ˜

⁰1

χ ˜

⁰1

q ¯

i

γ

5

q

i

+ α

6i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

q

i

,

spin-dependent, spin-independent, CP-violating and velocity-dependent terms

In the non-relativistic limit (manifestly true of CDM), velocity-dependent terms go to zero

(Jungman et al 1996 Phys Rep 267:195)

.

Generally not considered - though in self-interacting DM models a velocity-dependent DM-DM scattering cross-section is put in by hand (i.e. not from SUSY, nor any real theoretical model).

(8)

Q. What is the current theoretical status of

velocity-dependent WIMP-nucleus cross-sections?

A. In general SUSY, the neutralino does indeed have a velocity-dependent coupling to quarks:

L

eff

= ¯ χ ˜

⁰1

γ

^µ

γ

5

χ ˜

⁰1

¯ q

i

γ

µ

(α

1i

+ α

2i

γ

5

) q

i

+ α

3i

χ ¯ ˜

⁰1

χ ˜

⁰1

¯ q

i

q

i

+ α

4i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

γ

5

q

i

+ α

5i

χ ¯ ˜

⁰1

χ ˜

⁰1

q ¯

i

γ

5

q

i

+ α

6i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

q

i

,

spin-dependent, spin-independent, CP-violating and velocity-dependent terms

In the non-relativistic limit (manifestly true of CDM), velocity-dependent terms go to zero

.

Generally not considered - though in self-interacting DM models a velocity-dependent DM-DM scattering cross-section is put in by hand (i.e. not from SUSY, nor any real theoretical model).

May ’07 KTH 5A5461 Direct Detection of Dark Matter: Follow-up (Cerdeño et al. 2004 JHEP 12:048)

(9)

Q. What is the current theoretical status of

velocity-dependent WIMP-nucleus cross-sections?

A. In general SUSY, the neutralino does indeed have a velocity-dependent coupling to quarks:

L

eff

=

χ¯˜⁰1γ^µγ5χ˜⁰1¯qiγµ

(α

1i

+

α2iγ5

)

qi

+ α

3i

χ ¯ ˜

⁰1

χ ˜

⁰1

¯ q

i

q

i

+ α

4i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

γ

5

q

i

+ α

5i

χ ¯ ˜

⁰1

χ ˜

⁰1

q ¯

i

γ

5

q

i

+ α

6i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

q

i

,

spin-dependent, spin-independent, CP-violating and velocity-dependent

terms

In the non-relativistic limit (manifestly true of CDM), velocity-dependent terms go to zero

.

Generally not considered - though in self-interacting DM models a velocity-dependent DM-DM scattering cross-section is put in by hand (i.e. not from SUSY, nor any real theoretical model).

(10)

Q. What is the current theoretical status of

velocity-dependent WIMP-nucleus cross-sections?

A. In general SUSY, the neutralino does indeed have a velocity-dependent coupling to quarks:

L

eff

=

(α

1i

+

α2iγ5

)

qi+ α3iχ¯˜⁰1χ˜⁰1¯qiqi

+ α

4i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

γ

5

q

i

+ α

5i

χ ¯ ˜

⁰1

χ ˜

⁰1

q ¯

i

γ

5

q

i

+ α

6i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

q

i

,

spin-dependent,spin-independent, CP-violating and velocity-dependent

terms

In the non-relativistic limit (manifestly true of CDM), velocity-dependent terms go to zero

.

Generally not considered - though in self-interacting DM models a velocity-dependent DM-DM scattering cross-section is put in by hand (i.e. not from SUSY, nor any real theoretical model).

(11)

Q. What is the current theoretical status of

velocity-dependent WIMP-nucleus cross-sections?

A. In general SUSY, the neutralino does indeed have a velocity-dependent coupling to quarks:

L

eff

=

(α

1i

+

α2iγ5

)

+ α

4i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

γ

5

q

i

+

α5iχ¯˜⁰1χ˜⁰1q¯iγ5qi+ α6iχ¯˜⁰1γ5χ˜⁰1¯qiqi,

spin-dependent,spin-independent,CP-violating

and velocity-dependent terms

In the non-relativistic limit (manifestly true of CDM), velocity-dependent terms go to zero

.

Generally not considered - though in self-interacting DM models a velocity-dependent DM-DM scattering cross-section is put in by hand (i.e. not from SUSY, nor any real theoretical model).

(12)

Q. What is the current theoretical status of

velocity-dependent WIMP-nucleus cross-sections?

A. In general SUSY, the neutralino does indeed have a velocity-dependent coupling to quarks:

L

eff

=

(α

1i

+

α2iγ5

)

+ α

4i

χ ¯ ˜

⁰1

γ

5

χ ˜

⁰1

¯ q

i

γ

5

q

i

+

α5iχ¯˜⁰1χ˜⁰1q¯iγ5qi+ α6iχ¯˜⁰1γ5χ˜⁰1¯qiqi,

spin-dependent,spin-independent,CP-violating

and velocity-dependent terms

In the non-relativistic limit (manifestly true of CDM), velocity-dependent terms go to zero

.

Generally not considered - though in self-interacting DM models a velocity-dependent DM-DM scattering cross-section is put in by hand (i.e. not from SUSY, nor any real theoretical model).

(13)

Q. Could one build an SDMMD (Super-Duper Mega-Mega Detector)?

A. Maybe, but it would be tough.

Ge/Si (ionisation detectors) don’t scintillate NaI/CsI (scintillation detectors) don’t semiconduct

=⇒ no ionisation signal

Phonons cannot exist in liquid detectors

Solid Xe is still a scintillator

(Aprile et al. 1994 NIM 353:55)

It can also be made to semiconduct under the right T/P conditions

(Eremets 2000, PRL 85:2797)

The tricky bit would be getting it to solidify without crystal defects, in usable quantities, all under high pressure - and keeping it that way

(14)

Q. Could one build an SDMMD (Super-Duper Mega-Mega Detector)?

A. Maybe, but it would be tough.

Ge/Si (ionisation detectors) don’t scintillate NaI/CsI (scintillation detectors) don’t semiconduct

=⇒ no ionisation signal

Phonons cannot exist in liquid detectors

conditions

(Eremets 2000, PRL 85:2797)

The tricky bit would be getting it to solidify without crystal defects, in usable quantities, all under high pressure - and keeping it that way

(15)

Q. Could one build an SDMMD (Super-Duper Mega-Mega Detector)?

A. Maybe, but it would be tough.

Ge/Si (ionisation detectors) don’t scintillate NaI/CsI (scintillation detectors) don’t semiconduct

=⇒ no ionisation signal

Phonons cannot exist in liquid detectors Solid Xe is still a scintillator

(Aprile et al. 1994 NIM 353:55)

It can also be made to semiconduct under the right T/P conditions

(Eremets 2000, PRL 85:2797)

The tricky bit would be getting it to solidify without crystal defects, in usable quantities, all under high pressure - and keeping it that way

(16)

My halo model’s better than yours. . . Q1. Really? Q2. So what?

Typical approximation is of an isothermal dark matter halo

=⇒ spherical, isotropic, non-rotating

with clumps and gradients in ρ and hv i

rotating. . . but much slower than the baryonic disc

A. Direct detection exclusion limits have sensitivities of “a few tens of percent” to the chosen halo model, BUT annual

modulation signals (like DAMA’s. . . ) are super-sensitive to halo assumptions

(17)

My halo model’s better than yours. . . Q1. Really? Q2. So what?

Typical approximation is of an isothermal dark matter halo

=⇒ spherical, isotropic, non-rotating

‘Real’ halos are thought to be triaxial spheroids,

with clumps and gradients in ρ and hv i

rotating. . . but much slower than the baryonic disc

A. Direct detection exclusion limits have sensitivities of “a few tens of percent” to the chosen halo model, BUT annual

modulation signals (like DAMA’s. . . ) are super-sensitive to halo assumptions

(18)

My halo model’s better than yours. . . Q1. Really? Q2. So what?

Typical approximation is of an isothermal dark matter halo

=⇒ spherical, isotropic, non-rotating

‘Real’ halos are thought to be triaxial spheroids,

with clumps and gradients in ρ and hv i

tens of percent” to the chosen halo model, BUT annual

modulation signals (like DAMA’s. . . ) are super-sensitive to halo assumptions

(19)

My halo model’s better than yours. . . Q1. Really? Q2. So what?

Typical approximation is of an isothermal dark matter halo

=⇒ spherical, isotropic, non-rotating

‘Real’ halos are thought to be triaxial spheroids,

with clumps and gradients in ρ and hv i

rotating. . . but much slower than the baryonic disc

A. Direct detection exclusion limits have sensitivities of “a few tens of percent” to the chosen halo model, BUT annual

modulation signals (like DAMA’s. . . ) are super-sensitive to halo assumptions

(20)

My halo model’s better than yours. . . Q1. Really? Q2. So what?

Typical approximation is of an isothermal dark matter halo

=⇒ spherical, isotropic, non-rotating

‘Real’ halos are thought to be triaxial spheroids,

with clumps and gradients in ρ and hv i

rotating. . . but much slower than the baryonic disc

A. Direct detection exclusion limits have sensitivities of “a few tens of percent” to the chosen halo model, BUT annual

modulation signals (like DAMA’s. . . ) are super-sensitive to halo assumptions

May ’07 KTH 5A5461 Direct Detection of Dark Matter: Follow-up Refer to Green 2004, IAUS 220:483, Stiff & Widrow 2003, PRL 90:211301, Belli et al 2002, Phys Rev D 66:043503 and Green 2001, Phys Rev D 63:043005