Searching for Dark Matter with Bubble Chambers › sites › default › files › files ›...

Hugh Lippincott, Fermilab

Yale University, WIDGApril 29, 2014

Searching for Dark Matter with Bubble Chambers

Wednesday, May 7, 14

Weak Interaction Discussion Group History

• I was around for the founding WIDG• I think I gave the second WIDG ever on the

Klapdor 0vBB claim in January 2005

• 8 out of 11 slides were all text• This is my 9th


� �

PICOC. Amole, M. Besnier, G.⌘Caria, A.⌘Kamaha, A.⌘Noble, T.⌘Xie

M. Ardid,M. Bou-Cabo

D. Asner, J. Hall

D. Baxter, C.E. Dahl, M. Jin

E. Behnke, H. Borsodi, C.⌘Harnish, O. Harris, C.⌘Holdeman, I. Levine,

E.⌘Mann, J. Wells

P. Bhattacharjee, M. Das, S.⌘Seth

S.J. Brice, D. Broemmelsiek, P.S. Cooper, M. Crisler, W.H.⌘Lippincott, E. Ramberg, M.K. Ruschman, A.⌘Sonnenschein

J.I. Collar, R. Neilson, A.E. Robinson

F. Debris, M. Fines-Neuschild, C.M. Jackson, M.⌘Lafrenière, M.⌘Laurin, L. Lessard,J.-P.⌘Martin, M.-C.⌘Piro, A.⌘Plante, O. Scallon, N.⌘Starinski, V.⌘Zacek N. Dhungana, J. Farine, R.⌘Podviyanuk, U. Wichoski

R. Filgas, S. Pospisil, I. Stekl

S. Gagnebin, C. Krauss, D.⌘Marlisov, P. Mitra I. Lawson,

E. Vázquez Jáuregui

Collaboration

D. Maurya, S. Priya


We think dark matter exists


So what is it?The WIMP “Miracle”

(WIMP = Weakly Interacting Massive Particle)

A sampling of available dark matter candidates

Particles with mass and couplings at the weak scale

yield cross sections that correspond to ~correct relic density of cold dark matter

@94#$AH$,>=$ E$

,*&31)#$,";?(1?$

!"#$I*&);$J4(K#&?#$

It’s probablyWIMPs, right?


WIMPs not necessarily related to supersymmetry

• Dark sector could be as complicated as standard model

• Searches not limited by expectations from SUSY models

LM6,N$*$4-2$4#1#??*&();$/&-'$NJNO$

Dark Sector!$$784214(%*+%,#-(*%)$#$()%40+(45%/0$"%

#$%4(#)$%)',(%'9%$"(,%:(0;

How do we find it?

Fermi bubbles, courtesy of NASA

• Indirect - detect annihilation products from regions of high density like the sun or the center of the galaxy

• Accelerators - create a WIMP at the LHC

• Missing ET and monojet searches

• Direct detection - WIMPs can scatter elastically with nuclei and the recoil can be detected


How do we find it?





� �

��"��,��"��#��"��$

5�61��""��,��,�5,�DDD

��

��

��86��:& +;

��3�#��"��6P$��6��

��8��5�61�

� ��

� ��

�� 22 $��"�� 2 2 $��"��"��


How do we find it?





Rate calculationI The differential cross section (for spin-independent interactions)

per kilogram of target mass per unit recoil energy is

dRdQ

=⇢0m�

⇥ �0A2

2µ2p⇥ F 2(Q)⇥

Z

vm

f (v)v

dv (1)

I Dark matter density component, from local and galacticobservations with historically a factor of 2 uncertainty

I The unknown particle physics component, hopefully determinedby experiment

I Proportional to A2 for most models

I The nuclear part, approximately given by F 2(Q) / e�Q/Q0 whereQ0 ⇠ 80A5/3 MeV

I The velocity distribution of dark matter in the galaxy - of order30% uncertainty, and vm =

pQ/2m2r

3/2




dRdQ

=⇢0m�

⇥ �0A2

2µ2p⇥ F 2(Q)⇥

Z

vm

f (v)v

dv (1)





I The velocity distribution of dark matter in the galaxy - of order30% uncertainty, and vm =

pQ/2m2r

3/6Wednesday, May 7, 14



dRdQ

=⇢0m�

⇥ �0A2

2µ2p⇥ F 2(Q)⇥

Z

vm

f (v)v

dv (1)





I The velocity distribution of dark matter in the galaxy - of order30% uncertainty (not-statistical), and vm =

pQ/2m2r


The energy scale• Energy of recoils is tens of keV

• Entirely driven by kinematics, elastic scattering of things with approximately similar masses (100 GeV) and v ~ 0.001c

The energy scale

I Energy of recoils - ⇤ 10 � 100 keVI Entirely driven by kinematics - elastic scattering of particles with

approximately similar masses (100 GeV) and v ⇤ 0.001c (270km/s)

12

mNv2N =12⇥ 100 GeV ⇥ 10�6 = 50 keV (2)


How do we find it?

• Very low rate process (~events/year)

• Rate depends crucially on WIMP mass and thresholdEnectali Figueroa-Feliciano / Fermilab Seminar / 20130 10 20 30 40

Ethresh �keV⇥0.05

0.10

0.50

1.00

R⇤Ethresh⌅ �counts⇧10kg⇧year⇥Total Rate for different thresholds, m⇤ ⇥ 100 GeV⇧c2, ⌅ ⇥ 1.⇧10�45cm2

Ne

Si

Ar

Ge

Xe

Xe

Ge

Si

Ar

Ne

Integrated Rate as a function of low-energy threshold of experiment

R(cts/10kg/yr) for 10-45 cm2, 100 GeV


How do we find it?

• Very low rate process (~events/year)

• Rate depends crucially on WIMP mass and thresholdEnectali Figueroa-Feliciano / Fermilab Seminar / 20130 10 20 30 40

Ethresh �keV⇥0.05

0.10

0.50

1.00


Ne

Si

Ar

Ge

Xe

Xe

Ge

Si

Ar

Ne


Enectali Figueroa-Feliciano / Fermilab Seminar / 2013

0 10 20 30 40Ethresh �keV⇥

0.05

0.10

0.50

1.00


Ne

Si

Ar

Ge

Xe


Xe

Ge

Si

Ar

Ne

Enectali Figueroa-Feliciano / Fermilab Seminar / 2013

0 10 20 30 40Ethresh �keV⇥

0.05

0.10

0.50

1.00


Ne

Si

Ar

Ge

Xe


Knowing your energy scale and efficiency at threshold are crucial!

Xe

Ge

Si

Ar

Ne

R(cts/10kg/yr) for 10-45 cm2, 10 GeV


The canonical plot

• Limited at low mass by detector threshold

• Limited at high mass by density

Phys. Rev. Lett. 109, 181301 (2012)

5

]2WIMP Mass [GeV/c6 7 8 910 20 30 40 50 100 200 300 400 1000

]2W

IMP-

Nuc

leon

Cro

ss S

ectio

n [c

m

-4510

-4410

-4310

-4210

-4110

-4010

-3910

]2WIMP Mass [GeV/c6 7 8 910 20 30 40 50 100 200 300 400 1000

]2W

IMP-

Nuc

leon

Cro

ss S

ectio

n [c

m

-4510

-4410

-4310

-4210

-4110

-4010

-3910

]2WIMP Mass [GeV/c6 7 8 910 20 30 40 50 100 200 300 400 1000

]2W

IMP-

Nuc

leon

Cro

ss S

ectio

n [c

m

-4510

-4410

-4310

-4210

-4110

-4010

-3910

DAMA/I

DAMA/Na

CoGeNT

CDMS (2010/11)

EDELWEISS (2011/12)

XENON10 (2011)

XENON100 (

2011)

COUPP (2012)

SIMPLE (2012)

ZEPLIN-III (20

12)

CRESST-II (2012)

XENON100 (2012)observed limit (90% CL)

Expected limit of this run:

expectedσ 2 ± expectedσ 1 ±

FIG. 3: Result on spin-independent WIMP-nucleon scatter-ing from XENON100: The expected sensitivity of this run isshown by the green/yellow band (1�/2�) and the resultingexclusion limit (90% CL) in blue. For comparison, other ex-perimental limits (90% CL) and detection claims (2�) are alsoshown [19–22], together with the regions (1�/2�) preferred bysupersymmetric (CMSSM) models [18].

3 PE. The PL analysis yields a p-value of ⇥ 5% for allWIMP masses for the background-only hypothesis indi-cating that there is no excess due to a dark matter sig-nal. The probability that the expected background inthe benchmark region fluctuates to 2 events is 26.4% andconfirms this conclusion.A 90% confidence level exclusion limit for spin-

independent WIMP-nucleon cross sections ⇥� is calcu-lated, assuming an isothermal WIMP halo with a lo-cal density of �� = 0.3GeV/cm3, a local circular veloc-ity of v0 = 220 km/s, and a Galactic escape velocity ofvesc = 544 km/s [17]. Systematic uncertainties in the en-ergy scale as described by the Le� parametrization of [6]and in the background expectation are profiled out andrepresented in the limit. Poisson fluctuations in the num-ber of PEs dominate the S1 energy resolution and arealso taken into account along with the single PE resolu-tion. The expected sensitivity of this dataset in absenceof any signal is shown by the green/yellow (1⇥/2⇥) bandin Fig. 3. The new limit is represented by the thick blueline. It excludes a large fraction of previously unexploredparameter space, including regions preferred by scans ofthe constrained supersymmetric parameter space [18].The new XENON100 data provide the most strin-

gent limit for m� > 8GeV/c2 with a minimum of⇥ = 2.0 � 10�45 cm2 at m� = 55GeV/c2. The max-imum gap analysis uses an acceptance-corrected expo-sure of 2323.7 kg�days (weighted with the spectrum of a100GeV/c2 WIMP) and yields a result which agrees withthe result of Fig. 3 within the known systematic di�er-ences. The new XENON100 result continues to challengethe interpretation of the DAMA [19], CoGeNT [20], andCRESST-II [21] results as being due to scalar WIMP-nucleon interactions.

We acknowledge support from NSF, DOE, SNF, UZH,Volkswagen Foundation, FCT, Région des Pays de laLoire, STCSM, NSFC, DFG, Stichting FOM, WeizmannInstitute of Science, and the friends of Weizmann Insti-tute in memory of Richard Kronstein. We are grateful toLNGS for hosting and supporting XENON.

� Electronic address: [email protected]† Electronic address: [email protected]

[1] N. Jarosik et al., Astrophys. J. Suppl. 192, 14 (2011);K. Nakamura et al. (PDG), J. Phys. G37, 075021 (2010).

[2] G. Steigman and M. S. Turner, Nucl. Phys. B253, 375(1985); G. Jungman, M. Kamionkowski, and K. Griest,Phys. Rept. 267, 195 (1996).

[3] G. Bertone, D. Hooper, and J. Silk, Phys. Rept. 405, 279(2005).

[4] M. W. Goodman and E. Witten, Phys. Rev. D31, 3059(1985).

[5] E. Aprile et al. (XENON100), Astropart. Phys. 35, 573(2012).

[6] E. Aprile et al. (XENON100), Phys. Rev. Lett. 107,131302 (2011).

[7] M. Aglietta et al. (LVD), Phys. Rev.D58, 092005 (1998).[8] E. Aprile et al. (XENON100) Astropart. Phys. 35, 43

(2011).[9] E. Aprile et al., Phys. Rev. C79, 045807 (2009).

[10] X. Du et al., Rev. Sci. Instr. 75, 3224 (2004).[11] E. Aprile et al. (XENON100), (2012), arXiv:1207.3458.[12] E. Aprile et al. (XENON100), Phys. Rev. D84, 052003

(2011).[13] S. Yellin, Phys. Rev. D66, 032005 (2002).[14] G. Plante et al., Phys. Rev. C84, 045805 (2011).[15] E. Aprile et al., Phys. Rev. Lett. 97, 081302 (2006).[16] M. M. Szydagis et al., JINST 6, P10002 (2011).[17] M. C. Smith et al., Mon. Not. R. Astron. Soc. 379, 755

(2007).[18] Combined region using C. Strege et al., JCAP 1203,

030 (2012); A. Fowlie et al. (2012), arXiv:1206.0264;O. Buchmueller et al. (2011), arXiv:1112.3564.

[19] C. Savage et al., JCAP 0904, 010 (2009).[20] C. E. Aalseth et al. (CoGeNT), Phys. Rev. Lett. 106,

131301 (2011).[21] G. Angloher et al. (CRESST-II), Eur. Phys. J. C72, 1971

(2012).[22] Z. Ahmed et al. (CDMS), Science 327, 1619 (2010);

Z. Ahmed et al. (CDMS), Phys. Rev. Lett. 106,131302 (2011); E. Armengaud et al. (EDEL-WEISS), Phys. Lett. B 702, 329 (2011); J. Angleet al. (XENON10), Phys. Rev. Lett. 107, 051301 (2011);M. Felizardo et al. (SIMPLE), Phys. Rev. Lett. 108,201302 (2012); E. Behnke et al. (COUPP) (2012),arXiv:1204.3094; D. Y. Akimov et al. (ZEPLIN-III), Phys. Lett. B 709, 14 (2012); E. Armengaudet al. (EDELWEISS) (2012), arXiv:1207.1815.

LUX (2013)


The canonical plot

• What happened to “weakly” interacting? • Mediation via Z was excluded long ago (~10-39 cm2), but only

now are we probing Higgs exchange

5

]2WIMP Mass [GeV/c6 7 8 910 20 30 40 50 100 200 300 400 1000

]2W

IMP-

Nuc

leon

Cro

ss S

ectio

n [c

m

-4510

-4410

-4310

-4210

-4110

-4010

-3910

]2WIMP Mass [GeV/c6 7 8 910 20 30 40 50 100 200 300 400 1000

]2W

IMP-

Nuc

leon

Cro

ss S

ectio

n [c

m

-4510

-4410

-4310

-4210

-4110

-4010

-3910

]2WIMP Mass [GeV/c6 7 8 910 20 30 40 50 100 200 300 400 1000

]2W

IMP-

Nuc

leon

Cro

ss S

ectio

n [c

m

-4510

-4410

-4310

-4210

-4110

-4010

-3910

DAMA/I

DAMA/Na

CoGeNT

CDMS (2010/11)

EDELWEISS (2011/12)

XENON10 (2011)

XENON100 (

2011)

COUPP (2012)

SIMPLE (2012)

ZEPLIN-III (20

12)

CRESST-II (2012)

XENON100 (2012)observed limit (90% CL)

Expected limit of this run:

expectedσ 2 ± expectedσ 1 ±

FIG. 3: Result on spin-independent WIMP-nucleon scatter-ing from XENON100: The expected sensitivity of this run isshown by the green/yellow band (1�/2�) and the resultingexclusion limit (90% CL) in blue. For comparison, other ex-perimental limits (90% CL) and detection claims (2�) are alsoshown [19–22], together with the regions (1�/2�) preferred bysupersymmetric (CMSSM) models [18].

3 PE. The PL analysis yields a p-value of ⇥ 5% for allWIMP masses for the background-only hypothesis indi-cating that there is no excess due to a dark matter sig-nal. The probability that the expected background inthe benchmark region fluctuates to 2 events is 26.4% andconfirms this conclusion.A 90% confidence level exclusion limit for spin-

independent WIMP-nucleon cross sections ⇥� is calcu-lated, assuming an isothermal WIMP halo with a lo-cal density of �� = 0.3GeV/cm3, a local circular veloc-ity of v0 = 220 km/s, and a Galactic escape velocity ofvesc = 544 km/s [17]. Systematic uncertainties in the en-ergy scale as described by the Le� parametrization of [6]and in the background expectation are profiled out andrepresented in the limit. Poisson fluctuations in the num-ber of PEs dominate the S1 energy resolution and arealso taken into account along with the single PE resolu-tion. The expected sensitivity of this dataset in absenceof any signal is shown by the green/yellow (1⇥/2⇥) bandin Fig. 3. The new limit is represented by the thick blueline. It excludes a large fraction of previously unexploredparameter space, including regions preferred by scans ofthe constrained supersymmetric parameter space [18].The new XENON100 data provide the most strin-

gent limit for m� > 8GeV/c2 with a minimum of⇥ = 2.0 � 10�45 cm2 at m� = 55GeV/c2. The max-imum gap analysis uses an acceptance-corrected expo-sure of 2323.7 kg�days (weighted with the spectrum of a100GeV/c2 WIMP) and yields a result which agrees withthe result of Fig. 3 within the known systematic di�er-ences. The new XENON100 result continues to challengethe interpretation of the DAMA [19], CoGeNT [20], andCRESST-II [21] results as being due to scalar WIMP-nucleon interactions.

We acknowledge support from NSF, DOE, SNF, UZH,Volkswagen Foundation, FCT, Région des Pays de laLoire, STCSM, NSFC, DFG, Stichting FOM, WeizmannInstitute of Science, and the friends of Weizmann Insti-tute in memory of Richard Kronstein. We are grateful toLNGS for hosting and supporting XENON.

� Electronic address: [email protected]† Electronic address: [email protected]

[1] N. Jarosik et al., Astrophys. J. Suppl. 192, 14 (2011);K. Nakamura et al. (PDG), J. Phys. G37, 075021 (2010).

[2] G. Steigman and M. S. Turner, Nucl. Phys. B253, 375(1985); G. Jungman, M. Kamionkowski, and K. Griest,Phys. Rept. 267, 195 (1996).

[3] G. Bertone, D. Hooper, and J. Silk, Phys. Rept. 405, 279(2005).

[4] M. W. Goodman and E. Witten, Phys. Rev. D31, 3059(1985).

[5] E. Aprile et al. (XENON100), Astropart. Phys. 35, 573(2012).

[6] E. Aprile et al. (XENON100), Phys. Rev. Lett. 107,131302 (2011).

[7] M. Aglietta et al. (LVD), Phys. Rev.D58, 092005 (1998).[8] E. Aprile et al. (XENON100) Astropart. Phys. 35, 43

(2011).[9] E. Aprile et al., Phys. Rev. C79, 045807 (2009).

[10] X. Du et al., Rev. Sci. Instr. 75, 3224 (2004).[11] E. Aprile et al. (XENON100), (2012), arXiv:1207.3458.[12] E. Aprile et al. (XENON100), Phys. Rev. D84, 052003

(2011).[13] S. Yellin, Phys. Rev. D66, 032005 (2002).[14] G. Plante et al., Phys. Rev. C84, 045805 (2011).[15] E. Aprile et al., Phys. Rev. Lett. 97, 081302 (2006).[16] M. M. Szydagis et al., JINST 6, P10002 (2011).[17] M. C. Smith et al., Mon. Not. R. Astron. Soc. 379, 755

(2007).[18] Combined region using C. Strege et al., JCAP 1203,

030 (2012); A. Fowlie et al. (2012), arXiv:1206.0264;O. Buchmueller et al. (2011), arXiv:1112.3564.

[19] C. Savage et al., JCAP 0904, 010 (2009).[20] C. E. Aalseth et al. (CoGeNT), Phys. Rev. Lett. 106,

131301 (2011).[21] G. Angloher et al. (CRESST-II), Eur. Phys. J. C72, 1971

(2012).[22] Z. Ahmed et al. (CDMS), Science 327, 1619 (2010);

Z. Ahmed et al. (CDMS), Phys. Rev. Lett. 106,131302 (2011); E. Armengaud et al. (EDEL-WEISS), Phys. Lett. B 702, 329 (2011); J. Angleet al. (XENON10), Phys. Rev. Lett. 107, 051301 (2011);M. Felizardo et al. (SIMPLE), Phys. Rev. Lett. 108,201302 (2012); E. Behnke et al. (COUPP) (2012),arXiv:1204.3094; D. Y. Akimov et al. (ZEPLIN-III), Phys. Lett. B 709, 14 (2012); E. Armengaudet al. (EDELWEISS) (2012), arXiv:1207.1815.

Phys. Rev. Lett. 109, 181301 (2012)

LUX (2013)


So we look for WIMPs

• A few hundred just passed through us, and we might expect a handful of counts in a detector per year

• The problem is that background radioactivity is everywhere!


So we look for WIMPs

• A few hundred just passed through us, and we might expect a handful of counts in a detector per year

• The problem is that background radioactivity is everywhere!

100 events/second/kg =3,000,000,000,000 events/year

in a ton-scale experiment


Backgrounds!


Background sources

• Cosmic rays are constantly streaming through

• All experiments have to go underground to get away from cosmic rays

• Radioactive contaminants - rock, radon in air, impurities

• Emphasis on purification, everything must be clean

• The detector itself - steel, glass, detector components

• Discrimination - can you tell signal from background via some tag in the event itself?


Background sources




• Emphasis on purification and shielding

• The detector itself - steel, glass, detector components

• Discrimination - can you tell signal from background via some tag in the event itself?


Background sources



• The detector itself - steel, glass, detector components• Self-shielding to leave a clean inner region• Discrimination - can you tell signal from background (gamma

rays, alphas, neutrons, etc)?


• Emphasis on purification and shielding


5 10 15 20 25 300

0.2

0.4

0.6

0.8

1

Timing Parameter (µs)

Ioni

zatio

n Y

ield

Bulk electron recoilsSurface electron recoilsNuclear recoilsrecoil energy 10 100 keV

CDMS - Charge to heat

Xenon TPCs - Charge to light

Argon - Pulse shape discrimination

Electronic recoils (gammas)

vs. nuclear recoils (WIMPs)


Background sources



Bubble Chambers!


• Emphasis on purification and shielding• The detector itself - steel, glass, detector components• Self-shielding to leave a clean inner region• Discrimination - can you tell signal from background (gamma

rays, alphas, neutrons, etc)?


Chicagoland Observatory for Underground Particle Physics (COUPP)

[Some debate over the pronunciation (should the Ps be silent?)]


Chicagoland Observatory for Underground Particle Physics (COUPP)

[Some debate over the pronunciation (should the Ps be silent?)]

PICO


Bubble Chamber Operation Cycle

21 August 2013 15 M.B. Crisler SNOLAB Future Projects

Workshop

PICO/COUPP fast compression bubble chamber

• Pressure expansion creates superheated fluid, CF3I or C3F8

• I for spin-independent • F for spin-dependent• Particle interactions nucleate

bubbles

• Cameras see bubbles• Recompress chamber to reset

Bubble&chamber&operation&

Water&(buffer)&

Propylene&Glycol&(hydraulic&fluid)&

CF3I&(target)&

to&piston&/&pump&

!Expand!the!chamber!to!the!superheated!state!(10sec).!!Cameras!see!the!bubble!

Trigger!Stereoscopic!position!information!!

Recompress!the!chamber!(100msec)!and!wait!30sec!after!every!bubble.!

Synthetic&silica&jar&

February&2nd,&2013& 9&Russell&Neilson&


• By choosing superheat parameters appropriately (temperature and pressure), bubble chambers are blind to electronic recoils (10-10 or better)

• The probability for gamma interaction to produce a bubble:• To form a bubble requires two things• Enough energy• Enough energy density - length scale must be comparable

to the critical bubble size

2 4 6 8 10 1210−11

10−10

10−9

10−8

10−7

10−6

10−5

Threshold (keV)

Prob

abilit

y

CF3I (various)CF3I fitC3F8 (various)C3F8 fit

Preliminary

Why bubble chambers?



• By choosing superheat parameters appropriately (temperature and pressure), bubble chambers are blind to electronic recoils (10-10 or better)

• To form a bubble requires two things• Enough energy• Enough energy density - length scale must be

comparable to the critical bubble size


• Easy to identify multiple scattering events Neutron backgrounds

• Easy DAQ and analysis chain• Cameras• Piezos• No PMTs, no cryogenics



We take pictures


Why not bubble chambers?

• Threshold detectors - no energy resolution• Harder to distinguish some backgrounds, less information

about any potential signal

• Alphas (several MeV) were a big concern• Energy threshold calibrations are hard and important• Bubble chambers are slow - about 30 s of deadtime for every

event

• Overall rate must be low


About those alphas

• Discovery of acoustic discrimination against alphas by PICASSO (Aubin et al, New J. Phys 10:103017, 2008)

• Alphas deposit energy over tens of microns• Nuclear recoils deposit theirs in tens of nanometers• In COUPP bubble chambers, alphas are several times louder


The PICO program• COUPP4: A 2-liter chamber operated at SNOLAB from 2010-2012

• COUPP60: Up to 40 liters, running at SNOLAB now

• PICO-2L: Refurbished COUPP4 with C3F8, filled in October, 2013

• PICO-250L: Ton scale detector in G2 DM competition, at SNOLAB in 2016?


!  150 mm diameter fused silica jar !  Closed by a flexible stainless steel

pressure balancing bellows !  Instrumented with

"  Temperature, Pressure Transducers "  Fast Transient Pressure Transducer "  Piezo Electric Acoustic Transducers

!  Immersed in hydraulic fluid within a stainless steel pressure vessel

!  Hydraulic pressure controls the superheated fluid pressure

!  Viewed by machine vision cameras

The Bubble Chamber

26 July 2012 8 M.B. Crisler IDM 2012

COUPP4 at SNOLAB

• First run at SNOLAB: About 140 live days of data, with 79% acceptance after all cuts


• Better than 99.3% rejection against alphas at 16 keV threshold

• Limited by statistics, and backgrounds

COUPP4: Acoustic discrimination

−1 0 1 2 30

20

40

60

80

100

120

ln(AP)

Cou

nts/

0.05

AP

AmBe data sampleWIMP search data

20 recoil− like events

2474alphas

Acousticcut, 96%acceptance


This is what dark matter would sound like


This is what an alpha sounds like


Both together, just to hear the difference


• 20 WIMP candidates (6 at 8 keV, 6 at 11 keV, 8 at 16 keV)

• 3 multiple bubble events imply neutrons

• U, Th in the piezo-acoustic sensors and the viewports

• Remaining excess of singles at low threshold

• Time clustering

• Correlated with activity at water-CF3I interface

COUPP4: Results and sensitivity


• Given uncertainties on backgrounds, no background subtraction: PRD 86:052001 (2012)


WIMP Mass (GeV)

Spin−d

epen

dent

pro

ton

cros

s−se

ctio

n (c

m2 )

COUP

P (Jan.

2011)

KIMS (2

007)

PICASS

O (2009

)

(χχ→

W+W

− )

(χχ→

bb)

SIMPLE

(2010)

SIMPLE (2011)

COUP

P (this

result)

IceCu

be

Super−K

(hard)

COUPP

(Jan. 2011)PICASSO

(2012)

cMSSM

(χχ→

W+W

− )

(χχ→

bb)

SIMPLE

(2010)

SIMPLE (2011)

COUP

P (this

result)

Super−K

(soft)

CMS

CDF (A−V Eff.

theory)

CDF (A−V100 GeV med.)

101 102 103 10410−40

10−39

10−38

10−37

10−36


• Given uncertainties on backgrounds, no background subtraction: PRD 86:052001 (2012)


http://dmtools.brown.edu/ Gaitskell,Mandic,Filippini

WIMP Mass (GeV)Spi

n−in

depe

nden

t nuc

leon

cro

ss−s

ectio

n (c

m2 )

XENON10

CDMS

CDMS (SUF)

XENON100

COUPP (Jan

. 2011)

COUPP (thi

s result)

cMSSM

101 102 10310−45

10−44

10−43

10−42

10−41

10−40


• Engineering run at shallow site in 2010 • Low backgrounds and acoustic

discrimination

• Fluid darkening due to photodissociation of iodine

• Excessive surface rate• Improved purification and chemistry tested

in second run November, 2011

• Moved to SNOLAB beginning summer of 2012

COUPP60

Preliminary



discrimination


• Excessive surface rate• Improved purification and chemistry tested

in second run November, 2011


COUPP60


COUPP60


discrimination


• Excessive surface rate• Improved purification and chemistry,

tested in November, 2011


Preliminary


• Filled with 36.8 kg of CF3I at end of April, 2013• First bubble observed on May 1(radon decay)• Physics data started June 13

COUPP60


• Collected >2700 kg-days of dark matter search data between 9 and 25 keV threshold

• Good live fraction > 80%, no darkening• >1500 neutron source events from calibration runs

COUPP60


• Position reconstruction working well

• Clear set of events on surface and hemisphere

• Not a background, and rate is under control

COUPP60 - the dataZ&=(),%49=%';9@@*&'[(=$>(),'

49)?%+"&'AA'3'K\6]1',&$4=:'#$%$N'

^O_'

Z&=(),%49=%';9@@*&'[(=$>(),'49)?%+"&'AA'3'K\6]1',&$4=:'#$%$N'

^O`' _O`'

Preliminary

Preliminary


• Acoustics are working well

• Calibration data show a narrower distribution than COUPP4

• Alpha discrimination still very strong

−1 0 1 2 30

50

100

150

200

ln(AP)

Cou

nts/

0.05

AP

AmBe neutron source

−1 0 1 2 30

20

40

60

80

100

120

ln(AP)

Cou

nts/

0.05

AP

AmBe data sampleWIMP search data

20 recoil− like events

2474alphas

Acousticcut, 96%acceptance

103 104 10510 2

10 1

100

101

102

103

Pow

er ti

mes

freq

uenc

y2

Frequency (Hz)

Piezo 2

Alpha, 12000 < R2 < 16000

Neutron, 12000 < R2 < 16000Timeouts

COUPP60 - the dataPreliminary


• Analysis still under development

• Good news: Zero multiple bubbles, no neutrons. Limit is factor 3 below observed rate in COUPP4

• Bad news: Population of events that sound like nuclear recoils but are clearly not WIMPs

• Silver lining: statistics - we can actually study them in detail

• Early indications confirm a slightly different acoustic distribution and similar timing and spatial correlations for at least some fraction of events

COUPP60 - the data



• Good news: Zero multiple bubbles, no neutrons. Limit on neutron rate is factor 6 below observed rate in COUPP4



• Early indications confirm a slightly different acoustic distribution and similar timing and spatial correlations for at least some fraction of events

COUPP60 - the data



• Good news: Zero multiple bubbles, no neutrons. Limit on neutron rate is factor 6 below observed rate in COUPP4



• Indications confirm a slightly different acoustic distribution and similar timing and spatial correlations to COUPP4 background for at least some fraction of events

COUPP60 - the data


• Temperature dependence - more specifically, a dT/dt dependence

COUPP60 - the data

Singles'Recoil6like'rate'in'bulk'vs'Fme'during'first'period'at'low'pressure'

2'

loose'fiducial'

T'in'glycol'

T'in'water'

Rate'(not'to'scale)'

(not'to'scale)'

(not'to'scale)' Both'WIMP'and'gamma''runs'are''included'here,'since'no'gamma''turn6on'was''seen'

• Spatial dependence - looks temperature-related to me


• Clear shift in acoustics relative to neutron calibration data

COUPP60 - the data

Looking(at(26(psi(data(compared(to(neutron(data(at(28(psi(and(at(same(

temperature(

8(

34(deg(C(28(psi(

34(deg(C(26(psi(

Pressure@dependence(is(pre\y(well(calibrated(so(I(think(this(is(a(fair(comparison(

• Developing more advanced acoustic analysis that shows much better discrimination on a subset of data


PICO-2L (COUPP4 redux)

• Alternate fluid - remove the iodine - C3F8• Lower threshold (down to 3 keV in test stand)• Improved sensitivity at low WIMP mass• ~1 event per day from recent CDMS result• Improved SD sensitivity• First effort in concert with the PICASSO

collaboration

• Chamber filled in October, 2013Wednesday, May 7, 14


• Why C3F8 again? Excellent gamma rejection at a lower threshold

2 4 6 8 10 1210−11

10−10

10−9

10−8

10−7

10−6

10−5

Threshold (keV)

Prob

abilit

y

CF3I (various)CF3I fitC3F8 (various)C3F8 fit

Preliminary



• Alternate fluid - remove the iodine - C3F8• Lower threshold (down to 3 keV in test stand)• Improved sensitivity at low WIMP mass• ~1 event per day from recent CDMS result• Improved SD sensitivity• First effort in concert with the PICASSO

collaboration

• Chamber filled in October, 2013Wednesday, May 7, 14

PICO-2L

0.08 0.06 0.04 0.02 0

0.05

0

0.05

Time (s)

Vol

tage

(V)

•Now taking data!


PICO-2L

• 26 bubble event from AmBe neutron source! Wednesday, May 7, 14

PICO-2L

0 1 2 3 4 5 6 70

5

10

15

20

25

30

35

40

45

50

(AP)

Cou

nts/

0.05

AP

AmBe neutron source

• Acoustics from commissioning data (with neutron source)• Radon injection at start of run provides some alphas in

our neutron data set

Alphas from radon

Neutron events


ProjectionsN9:#&=06N$N-9Q*4$:&-h#13-4?$

=06N$MM$[#$

N9:#&=06N$N-9Q*4$$,&-h#13-4?$

0>6>$

@94#$AH$,>=$ BD$

Xenon10/100

PICO-2l

LUX


PICO-250L

• Begin with C3F8 to maximize leadership potential (Spin-dependent and low masses)

• Retain flexibility to respond to developments in the field

PICO-250L DOE G2 Proposal

21

superheated is accomplished by small proportional valves that can open the chamber to the reservoir tanks to tweak the pressure up or down as needed under servo control. Schematically, the system is very similar to the hydraulic control system of a forklift truck or a lift gate, and it can be entirely constructed from commercial hydraulic components.

Figure 13: Pressure Vessel shown with integral fast compression units (left). Site rendering showing the

Pressure Vessel in the water tank viewed from the top by the PMTs of the veto system (right).

In our earlier chambers, the hydraulic control unit was an integrated package connected to the bubble chamber by piping. For a large chamber, this piping connection would present a significant seismic risk because the pipe could be sheared in a seismic incident. A significant design innovation planned from the PICO-250L hydraulic controls is to mount the fast compression units directly on the pressure vessel body, as shown in Figure 13. Because the fast compression connection dominates the fluid flow requirements, this geometry allows us to eliminate the large piping connection to the chamber. The fluid flow during expansion is 0.1% of the flow during fast compression. Therefore, a thin, flexible pressure-rated hose can accommodate the flow required for expansion and for recharging the fast compression accumulators. This hose would connect to a manifold on the chamber body protected by check valves so that a failure would not result in a loss of chamber pressure.

It is possible to engineer the volumes of the pressure reservoirs so that the bubble chamber will be safe against any failure or incorrect configuration of valves or failure of the pump.

The PICO-250L pressure controller will be managed using a National Instruments Compact RIO controller. This system provides for significant instrument readout, data processing, and vetting of process control data at the FPGA level within the backplane of the device. The FPGA passes processed data to a real-time controller that provides higher-level control functions and passes the data on to the LabVIEW based Data Acquisition System. This system has been successfully implemented in COUPP/PICO calibration and prototype chambers.


5

environment called for it (e.g., to follow up a signal in a large xenon detector). In addition, the use of multiple mass targets will be needed to characterize the WIMP mass and velocity distribution in the event of a discovery [21][22].

Figure 2: Comparison of SD-proton vs. SI cross sections for a set of dark matter models, showing the complementary and necessary reach of both channels to explore possible parameter space (from [17]).

b)(Description(of(Experimental(Method;(Performance(Requirements(

The(Case(for(Bubble(Chambers(as(Dark(Matter(Detectors(Throughout this section, we refer to several existing bubble chambers. COUPP-4 is a 2-liter chamber that was filled with CF3I at Fermilab and then at SNOLAB, producing excellent physics results [9][10]. COUPP-60 has been running with 37 kg of CF3I at SNOLAB since June 2013. PICO-2L is a 2-liter chamber that has replaced COUPP-4 at SNOLAB; commissioned in October 2013, it represents the first large-scale bubble chamber to be filled with C3F8 and also the first joint effort of the combined PICO collaboration.

The strengths of the bubble chamber technology for dark matter searches can be summarized as follows:

Electron)recoil)insensitivity)A principal strength of the bubble chamber technology for a dark matter search application is the extraordinary insensitivity (~10-10) to electron recoil backgrounds. The ability of the bubble chamber to attain this large background rejection factor while maintaining high efficiency for detecting nuclear recoils arises naturally from the physics of bubble nucleation in a superheated liquid, which requires a critical energy deposition within a small volume to nucleate the bubble. The bubble formation scale and energy threshold are classical thermodynamic properties determined by the temperature and pressure of the superheated fluid [23][24][25]. These can be adjusted to cleanly discriminate between the nuclear

oughly scan the parameter spaces, we adopted theBayesian method that is the foundation of the Markovchain Monte Carlo approach. The DM models can havedistinct phenomenological predictions. We showed that theDM model possibilities can be narrowed by measurementsof both SI and SD elastic scattering. The direct signals forDM in recoil and neutrino telescope experiments are com-plementary to LHC experiments in distinguishing the be-yond the standard model physics scenarios [135].

We summarize below the model predictions for the DMcross sections; the posterior distributions are summarizedin Fig. 12.

(i) In mSUGRA, the FP region provides the largest SIand SD cross sections. This is due to the mixedHiggsino nature of the lightest neutralino; the neu-tralino couplings to the Higgs and Z bosons are large.The Bino nature of the lightest neutralino in the CAand AF regions causes these scenarios to have sub-stantially smaller cross sections.

(ii) The tadpole nMSSMmodel has large SD scattering,of order 10!3 pb, and a wide range of SI cross

section. This is a consequence of the DM annihila-tion occurring through the Z boson. To counter thesmall annihilation rate in the early universe (due tothe small neutralino mass in the model), the neu-tralino pair is required to have a larger Z bosoncoupling, resulting in a large SD cross section.

(iii) In the singlet extended SM, the DM candidate isspin-0, which gives a vanishing SD cross section.The SI cross section is generally small, below"10!8 pb, and occurs through Higgs boson ex-change. If SD scattering of DM is observed, theclass of models with spin-0 DM would be imme-diately excluded as being the sole origin of the DMin the Universe.

(iv) For stable Dirac neutrino DM, the Z boson domi-nates in the calculation of both the relic density andelastic scattering cross section and makes the SI andSD cross sections tightly correlated and large.

(v) In mUED, a sweet spot of !SD "Oð10!6Þ pb existsfor which the DM relic density is reproduced. Therelic density is strongly dependent on the curvatureparameter and fixes its value. The KK quarks haveapproximately the same mass as the inverse curva-ture and the SD cross section is thus closely tied tothe relic density. On the other hand, the !SI crosssection is more dispersed due to the larger variationof the Higgs boson mass.

(vi) In the LHT model, the SD interaction occursthrough T-odd quarks which have a small hyper-charge. Therefore, the SD cross section in thismodel is typically very small. In contrast, the SIscattering proceeds through the Higgs and T-oddquarks, giving experimentally accessible SI crosssection values.

We provide a summary of the SI and SD cross sectionsby the box and whisker plots in Fig. 13. The boxes repre-sents the coverage of the middle 50-percentile. We sum-marize the forecast for observing a signal in neutrino

IceCube muon rate

MDM

log 1

0Nµ

0 200 400 600 800 1000

−4−3

−2−1

01

2 mSUGRAUEDLHTxSMRHN

IceCube signal significance

MDM

log 1

0σ

0 200 400 600 800 1000

−4−3

−2−1

01

2 mSUGRAUEDLHTxSMRHNDiscovery

Evidence

FIG. 11 (color online). The expected numbers of events and the statistical significance of DM signals for 3 and 5 years running of theIceCube neutrino telescope. Only the FP region in mSUGRA and a portion of the mUED parameter space give a significance>5! for5 years of running.

SI vs. SD

log10σSI

log 1

0σS

D

−12 −11 −10 −9 −8 −7 −6

−10

−8−6

−4−2

mSUGRAUEDLHTnMSSMRHN

FIG. 12 (color online). The posterior distributions of !DM!pSIand !DM!pSD for six models.

SPIN DEPENDENCE OF DARK MATTER SCATTERING PHYSICAL REVIEW D 78, 056007 (2008)

056007-13

• Straightforward scale up of existing PICO-2L and COUPP60 detectors


PICO-250L• Funded by NSF and DOE as part of G2

Dark Matter (big showdown in February)

• Engineering well underway• Construction 2015-2016?


20

Figure 12: The PICO-250L outer vessel (left) with the inner vessel and bellows inside (right).

The procurement of a suitable pressure vessel does not present any major challenges. We have identified vendors familiar with screened materials and controlled welding with non-thoriated electrodes. We have identified a vendor who also has significant expertise in the design of pressure vessel sight glass windows for bubble chambers. The procurement process requires the creation of a technical specification that identifies the gross dimensions, the nozzle positions and orientations, and the pressure ratings. The specification will state that the vessel must be designed, fabricated, and tested in compliance with both US and Ontario Canadian codes. The entire process of designing and fabricating the vessel is entirely the responsibility of the vendor. As was done for COUPP-60, to ensure radio-purity we will procure the raw stainless steel from a Swedish supplier that is reliably low background. We will hold the steel while it is sampled and measured for radioactivity. The successfully screened material will then be released to the vendor for fabrication.

WBS)2.3:)Hydraulic)Pressure)Controls)The pressure control unit is responsible for expanding and recompressing the chamber, regulating chamber pressure in the expanded and compressed states, and maintaining the chamber in a safe state in the face of subsystem failures. The system developed for our most recent set of test chambers has simplified the design and led to better performance and a more robust and fault-tolerant system.

The basic design concept is simple. A high-pressure reservoir in the form of an accumulator tank connects to the bubble chamber through a fast, large-bore solenoid valve. Opening this valve accomplishes fast compression and closing it prepares the chamber for expansion. A second, smaller solenoid connects the chamber to a low-pressure accumulator tank. Opening the smaller valve accomplishes the expansion of the chamber. Regulation of the pressure when the chamber is active and


21

superheated is accomplished by small proportional valves that can open the chamber to the reservoir tanks to tweak the pressure up or down as needed under servo control. Schematically, the system is very similar to the hydraulic control system of a forklift truck or a lift gate, and it can be entirely constructed from commercial hydraulic components.

Figure 13: Pressure Vessel shown with integral fast compression units (left). Site rendering showing the

Pressure Vessel in the water tank viewed from the top by the PMTs of the veto system (right).

In our earlier chambers, the hydraulic control unit was an integrated package connected to the bubble chamber by piping. For a large chamber, this piping connection would present a significant seismic risk because the pipe could be sheared in a seismic incident. A significant design innovation planned from the PICO-250L hydraulic controls is to mount the fast compression units directly on the pressure vessel body, as shown in Figure 13. Because the fast compression connection dominates the fluid flow requirements, this geometry allows us to eliminate the large piping connection to the chamber. The fluid flow during expansion is 0.1% of the flow during fast compression. Therefore, a thin, flexible pressure-rated hose can accommodate the flow required for expansion and for recharging the fast compression accumulators. This hose would connect to a manifold on the chamber body protected by check valves so that a failure would not result in a loss of chamber pressure.

It is possible to engineer the volumes of the pressure reservoirs so that the bubble chamber will be safe against any failure or incorrect configuration of valves or failure of the pump.

The PICO-250L pressure controller will be managed using a National Instruments Compact RIO controller. This system provides for significant instrument readout, data processing, and vetting of process control data at the FPGA level within the backplane of the device. The FPGA passes processed data to a real-time controller that provides higher-level control functions and passes the data on to the LabVIEW based Data Acquisition System. This system has been successfully implemented in COUPP/PICO calibration and prototype chambers.



4

Figure 1: Spin-dependent and -independent expected sensitivities for PICO-250L, corresponding to exposures of one live-year at 10 keV threshold with an expected 0.26 events from neutrons, 80 live-days at 5 keV with one solar neutrino event, and 40 live-days at 3 keV with one solar neutrino event. The supersymmetric parameter space probed through SD and SI couplings can be significantly decoupled (see for instance, Fig. 1 in [4]), resulting in a true chance at discovery through SD-sensitive searches. Adopting the probabilistic treatment in [15], PICO-250L has the potential to explore the vast majority of supersymmetric WIMP parameter space.

Historically, the field of direct detection of dark matter has focused on SI channels, primarily motivated by the idea that experiments could explore smaller cross sections due to the assumed A2 enhancement of most SI dark matter models. However, many WIMP models have orders of magnitude larger SD couplings than SI (see Figure 2), giving PICO-250L sensitivity to a complementary set of WIMP models that will be invisible to the next generation of SI detectors (e.g. [16][17] for SUSY, [4] for Kaluza-Klein). An extensive exploration of the SD parameter space in both SD proton and SD neutron couplings is therefore a critical component to any global dark matter strategy. As available supersymmetric parameter space becomes more constrained by direct detection experiments and the LHC, other theoretical models that can explain dark matter become more interesting. For example, recent developments in Asymmetric Dark Matter or hidden sector models have generated a great deal of theoretical interest ([18][19][20] and references therein). The prediction of a light dark matter particle is a common feature of many of these models, increasing the desirability for a complete exploration of all potential dark matter masses. With this in mind, the physics reach of PICO-250L is an excellent complement to that of large xenon detectors like LUX/LZ and Xenon1T. Xenon provides sensitivity to traditional heavy mass SI dark matter in addition to SD neutron couplings. PICO-250L has unmatched sensitivity to SD proton couplings and can more effectively probe light dark matter models than xenon by taking advantage of the more favorable kinematics of scattering on a light target like fluorine. Furthermore, PICO-250L will have a very different set of systematics than xenon detectors, without any of the concerns regarding quenching factors for example (particularly relevant for light dark matter).

In the search for very light dark matter, PICO-250L will come up against the 8B solar neutrino coherent scattering signal relatively soon in its initial run (~1 event per 40 live-days at 3 keV threshold) [36][37]. Once observed, the solar neutrino signal can be avoided by raising the operating threshold as reflected by the operations plan in Section IV subsection d) Span of Operations.

A final advantage to the bubble chamber technology, one that is again complementary to other dark matter searches, is the potential for target exchange in the same detector; a target such as CF3I with improved sensitivity in the SI channel to heavier WIMPs could be used after the C3F8 run if the physics

101 102 103 104

10−42

10−41

10−40

10−39

10−38

10−37

10−36

PICASS

O

cMSSM

COUPP-

60

PICO-2

50L(CF3

I)

PICO-250

L (C3F8)

PICO-2L

COUPP−

4

W+W

bb

IceCub

e

W+W

bbSupe

r−K

CMS

WIMP mass [GeV/c2]

SD p

roto

n cr

oss

sect

ion

per n

ucle

on[c

m2 ]

101 102 10310−46

10−45

10−44

10−43

10−42

10−41

10−40 PICASSO

cMSSM

DAMA

CDMS−Si

CoGent

CRESST

COUPP-60

PICO-250L (C3F8)

PICO-250L (CF3I)LUX

PICO−2L

XENON100

XENON10

CDMS−II

CDMS−II LT

COUPP−4

WIMP mass [GeV/c2]

SI c

ross

sec

tion

per n

ucle

on[c

m2 ]

Projections


Dark Matter, Sept 2007 Rick Gaitskell, Brown University, DOE

DM Direct Search Progress Over Time (2012)

~1 event kg-1 day-1

~1 event 1 tonne-1 yr-1

(Gross Masses kg)

LZ 7t!=2 10-48

Plot does not track low mass WIMPs 10 GeV

Many of current projections omitted from this plot

5Friday, May 18, 12


• Dark matter searches are making fast progress (indirect, accelerator and direct)

• PICO is producing the best direct detection limits on spin-dependent dark matter

• PICO bubble chambers are competitive for spin-independent searches (particularly for light dark matter)

Conclusion


1 10 100 1000 10410!5010!4910!4810!4710!4610!4510!4410!4310!4210!4110!4010!39

10!1410!1310!1210!1110!1010!910!810!710!610!510!410!3

WIMP Mass !GeV"c2#

WIM

P!nu

cleo

ncr

osss

ectio

n!cm2 #

WIM

P!nu

cleo

ncr

osss

ectio

n!pb#

8BNeutrinos

Atmospheric an

d DSNB Neutrin

os

CDMS II Ge

(2009)

Xenon100 (2

012)

CRESST

CoGeNT(2012)

CDMS Si(2013)

EDELWEISS (2011)

DAMA SIMPLE (2012)

ZEPLIN-III (201

2)COUPP (201

2)

SuperCDMS Soudan Low ThresholdXENON 10 S2 (2013)

CDMS-II Ge Low Threshold (2011)

SuperCDMS S

oudan

Xenon1T

LZ

LUX (2013)

DarkSide G2

DarkSide 50

DEAP3600

PICO250-CF3

I

PICO250-C3F8

7BeNeutrinos

NEUTRINO C OHER ENT SCATTERING

NEUTRINO COH

ERENT SCATT

ERING

CDMSlite

(2013)

SuperCDMS S

NOLABLUX 300-d

ay

SuperCDMS SNOLAB


Searching for Dark Matter with Bubble Chambers › sites › default › files › files ›...

Documents

Transcript of Searching for Dark Matter with Bubble Chambers › sites › default › files › files ›...