Collar Group Website non-accelerator particle physics Kavli Institute for Cosmological Physics
WIMP Detectors


Home
Research Projects
Axions
Neutrino Detectors
WIMP Detectors
Group Members
Our Laboratory
News & Recent Publications
Photo Galleries & Multimedia
Links
Site Map & Search

Project Team
William 'Joe' Bolte, Undergraduate Student

Juan I. Collar, Faculty Member

Dante Nakazawa, Graduate Student

Kevin O'Sullivan, Undergraduate Student

Brian C. Odom, Post-Doctoral Member

Aza Raskin, Undergraduate Student

Andrew Sonnenschein, Post-Doctoral Member


News
A COUPP is in the making: A new experimental effort to search for particle dark matter, the Chicagoland Observatory for Underground Particle Physics (COUPP) will debut at the end of summer 2004. A prototype bubble chamber sensitive to Weakly Interacting Massive Particles (WIMPs), developed by our group, is to be installed in the MINOS near detector gallery at Fermilab. [more]


Related Websites

FermiLab (FNAL)

SOUDAN Underground Laboratory (MN)



 
Our Research
Articles about WIMPs and Superheated Liquids
Photo Gallery and Multimedia
 

Our Research

Go to top

COUPP, a Heavy-Liquid Bubble Chamber for WIMP Detection

Go to top
 

Can a bubble chamber be made stable enough to search for rare events?

Superheated Liquids present a number of extraordinary advantages as detectors for Weakly Interacting Massive Particles (WIMPs are arguably the best particle candidates for the cold dark matter in galaxies like our own). First and foremost, the degree of superheat can be tuned so as to have complete insensitivity to the minimum-ionizing backgrounds that plague these searches, while still being responsive to low-energy nuclear recoils like those expected from WIMPs (Figure 1).

 

Figure 1. Figure 1.
Left: Bubble chamber compression/decompression cycle in P,T parameter space. The energy deposition from a particle interaction, if local enough (BC 2, 2005), produces a protobubble of gas in a superheated liquid that in turn evolves to destabilize the whole (i.e., induces a first order phase transition, in this case violent vaporization).
Right: Evolution of the protobubble (initially just a few tens of nm) produced by a neutron recoil in superheated CF3Br. The vial contains ~20 ml of liquid and full boiling takes place after ~0.1s (Movie).

 

To this one can add the advantages of safety, low-cost, optimal target constitution, and room-temperature operation. Two ongoing experiments, SIMPLE and PICASSO exploit the technique using superheated droplet detectors (a.k.a. "bubble detectors". In this implementation of the method the delicate metastable state of the target liquid is preserved by dispersing it into an immiscible gel, effectively resulting in a collection of mini-bubble chambers for which the possibility of nucleation on microscopic surface cavities (vapor embryos, the main cause of instability in bubble chambers) is absent. Unfortunately, some of the most promising target liquids are too dense to permit a homogeneous dispersion in water-based gels, making their use in bubble detectors a rather complicated endeavor (saturation of the sol with density-matching salts introduces a number of problems). This limitation and some encouraging first attempts within the framework of SIMPLE (Figure 2) prompted us to re-examine the original proposal by Hahn [1] to use bubble chambers in WIMP searches. The challenge, as discussed below, is in finding ways to match the indefinite (meta)stability seen in bubble detectors with a device where the superheated state can normally be maintained for short periods of time only.

 

Figure 2. Figure 2.
Response of a rudimentary unshielded bubble chamber (30g of Freon-115 entirely surrounded by a viscoelastic liquid sheath -"Aquasonic"-) in the Rustrel (LSBB) underground laboratory (J. Puibasset, Ph.D. Dissertation, 2000, Université Paris VI). The degree of superheat was equivalent to an energy threshold for chlorine recoils of ~15 keV. The sheath degraded rapidly, but the experiment demonstrated that long (up to 12 h in that case) superheat times can be achieved in relatively well-isolated bulk volumes of superheated liquid.

 

Recently we have been able to demonstrate [2] that after taking some precautions it may indeed be possible to keep bulk volumes of heavy refrigerants in a radiation-sensitive superheated metastable state for long enough to perform a WIMP search. The prototypes developed (Figure 3) are operated at near room temperature and the industrial refrigerants used are inexpensive, non-flammable and non-toxic, with a chemical composition that maximizes sensitivity to neutralino interactions through both the spin-dependent and independent channels [3] (Figure 4). For these reasons the technique seems to be ideally fitted for the goal of building tonne or even multi-tonne detectors, devices able to prove most of the supersymmetric phase space where the neutralino dark matter may abide. Monitoring of large detector volumes in bubble chambers is trivially simple when compared to other devices.

 

Figure 3. Figure 3.
Different stages of development of small bubble chamber prototypes: rudimentary, improved version of it, and a 2 kg device with all the bells and whistles. How soon the first multi-ton WIMP detector?

 

Figure 4. Figure 4.
Left: Each point in the plot represents a possible combination of spin-dependent and spin-independent neutralino-nucleus scattering cross sections arising from a possible choice of supersymmetric parameters [3]. The lack of strong correlation between both channels is evident, emphasizing the need for detectors able to explore both fronts.
Right: Fluorine displays the largest spin-dependent cross section of any target, whereas the spin-independent channel favors heavy-mass nuclei: superheated liquids such as CF3I make ideal neutralino detectors.

 

Several techniques have been identified and exploited to maximize the stability of small bubble chamber prototypes containing CF3Br, CF3I and C3F8. Namely, avoidance of contact with rough metallic surfaces, use of an immiscible liquid "lid" above the active volume, outgassing of surfaces in the presence of a buffer liquid, surface cleaning techniques and improvement of surface wetting via vapor deposition of buffer liquid [2]. Interestingly, several of these techniques have been developed only recently and in fields remote from that of radiation detection. Their net result is to bring the bubble chamber one step closer to a bubble detector, i.e., to insure that the superheated liquid is effectively isolated from any rough surfaces, or more specifically, that it effectively wets only another liquid (Figure 5). Small prototypes (~20 g) remain superheated for periods of 15 minutes on the average at the shallow 6 m.w.e. depth of the EFI underground laboratory, a nucleation rate compatible with the measured neutron flux and energy spectrum in the site (Figure 6). The insensitivity (rejection factor) to minimum ionizing particles (MIPs) in operating conditions at which the liquids are nevertheless fully responsive to low energy nuclear recoils has been measured to be > 109 (Figure 6) using a strong (3 mCi) 88Y gamma source. This guarantees the ability to build much larger prototypes, of tonne or multi-tonne size, essentially without any concern for MIPs, including large concentrations of C-14.

 

Figure 5. Figure 5.
An example of deactivation of microscopic inhomogeneous nucleation sites on apparently smooth glass surfaces by use of a buffer liquid. A similar situation can be achieved by filling up a degassed chamber with buffer liquid vapor.

 

Calibrations using neutron sources with a well-defined endpoint energy (11.1 MeV for Am/Be and 152 KeV for 88Y/Be) have allowed testing of the response of the liquids to nuclear recoils down to 4 keV in the case of CF3I and to establish excellent agreement with theoretical models of this response (Figure 6). Data points in the figure represent the appearance of the first bubble nucleation upon decompression in the presence of each source (i.e., as the energy threshold for nucleation is reduced), each point corresponding to a compression/decompression cycle. For sufficiently high source intensities and/or slow decompression rates this bubble (and subsequent violent boiling) is the result of a recoil with an energy close to the highest that the source can produce. These maximum recoil energies are indicated in the labels. The lines represent the theoretical expectations for the onset of sensitivity to recoils of these energies using the classic Seitz "Hot Spike" model [4]. An excellent agreement with the data is observed by best-fitting the single free parameter in this model [2] (the best value obtained being compatible with previous bubble chamber data [4]). The predicted sensitivity for each species not being exactly the same, the lines represent the first one expected to react to the source (Br and F, closely matched, in this case). Further calibrations are planned to test the contribution of each individual component to the response of the chamber. The photonuclear 88Y/Be source emits a mixed field of ~108 high-energy gammas and just 3 x 103 monochromatic neutrons per second, allowing for a dramatic demonstration of total insensitivity to photoelectrons in operating conditions that nevertheless ensure optimal response to WIMP interactions (Figure 6). A recently procured 124Sb/Be source will be used to perform calibrations with recoil energies as low as ~1 keV, an unprecedented test of a WIMP detector.

 

Figure 6. Figure 6.
Left: Distribution of duration of the superheated state in a 12 ml CF3Br bubble chamber at -10oC and atmospheric pressure. A strong gamma source inducing ~1E6 interactions per second produces no measurable boiling in conditions where the chamber is sensitive to low-energy recoils.
Right: Response of the same chamber to neutron sources and comparison with theoretical models. Labels indicate the maximum recoil energy that can be produced by the source in each species. Lines indicate the pressure at each temperature below which full sensitivity to this energy is expected according to Seitz's "hot spike" model (the experimental points represent the appearance of the first bubble upon decompression -i.e., their upper boundary is expected to coincide with the lines-).

 

The construction of a 1-liter (~2 kg) active volume chamber is well-advanced (Figure 7). The purpose of this prototype and experiment (the Chicagoland Observatory for Underground Particle Physics, COUPP) is to study the ultimate limits to the stability of the superheated liquid in a deeper location, with much reduced neutron backgrounds. However a device of this mass can already be an extremely competitive WIMP detector, given the optimal choice of target nuclei and intrinsic insensitivity to most backgrounds (Figure 8). Several aspects in the design of this prototype, including safety issues, have been studied by our collaborators at Fermilab (FNAL). While it is our ultimate goal to deploy a large bubble chamber dark matter search in the Soudan Underground Laboratory (MN), initial testing of a prototype device will take place in an unused "muon alcove" in the MINOS near detector gallery, on Fermilab grounds. In this Fermilab site the experiment will profit from ~300 ft of rock overburden: a preliminary estimate of the cosmic-ray associated backgrounds at this depth reveals that the nucleation rate assumption used to generate the excellent prospects in (Figure 8) could in principle be met there. In the presence of ~30 cm of polyethylene shielding, (cosmic) muon-induced energetic neutrons producing a few nucleations per kg/day are expected to be dominant.

 

Figure 7. Figure 7.
Left: Conceptual design of the 2 kg CF3I chamber to be used in the preliminary phase of COUPP.
Center: The recompression vessel in this prototype.
Right: Inner quartz vessel and pressure-compensation bellows for the same.

 

Figure 8. Figure 8.
Sensitivity limits in the spin-independent (Left) and
spin-dependent (Right) neutralino parameter space achievable with COUPP at the 2 kg prototype stage, compared to other experiments. The neutron background rate used for these estimates is representative of what can be expected in the Minos near detector gallery and, according to Monte carlo calculations, a factor of ~100 too conservative for the Soudan depth. The limits are plotted for two different energy thresholds, one already demonstrated with Y-88/Be neutron calibrations and a second one (close to the best that can be expected before gamma background rejection is lost) soon to be tested with a Sb-124/Be source.
Note: recently released CDMS limits improve those in the figure by a factor 3-4.

 

MCNP-Polimi [5] Monte Carlo simulations of the response to a typical underground neutron flux indicate that large enough bubble chambers (few hundred liters) would have ideal features as WIMP detectors (Figure 9). For instance, a sizeable inner fiducial volume would be shielded against events produced by neutrons able to penetrate a reasonable thickness of neutron moderator [2].

 

Figure 9. Figure 9.
Large bubble chambers are "self-shielding" against neutron-induced recoils. The probability that a neutron (mfp=few cm) reaches deep into a sizeable central fiducial volume within the chamber, produces just one bubble and exits without further bubble production is extremely small in O(1) ton devices. This advantageous property is illustrated in the MCNP-Polimi simulations in the figure.

 

These represent the ultimate challenge for next-generation WIMP detectors. Their interactions would nevertheless be revealed in these chambers by multi-bubble events which WIMPs cannot produce (Figure 9 & Figure 10). To this unique feature one can add the ability to easily exchange liquids from those containing fluorine as the heaviest atom to those containing iodine or bromine instead: for targets like these the expected WIMP and neutron induced bubble-nucleation rates can be radically different, a signature that could be exploited for WIMP discovery Figure 11. The 2 kg prototype under construction will also serve the purpose of studying the feasibility of building much larger chambers. New challenges will certainly arise during its operation, besides those already envisioned. For instance, Radon emanation from metallic parts in the inner vessel can give rise to a recoiling-daughter background, but prospects based on BOREXINO measurements are reassuring at ~5 microBq/m2 of steel. The inner vessel can be sealed off from the external world by Rn-impermeable materials (steel, quartz, metallic gaskets), and hence the only concern is internal surface emanation and secondarily, keeping the initial concentration low.

 

Figure 10. Figure 10.
Left: A single bubble like those expected to be produced by WIMPs.
Center: Several simultaneous disconnected bubbles are a tell-tale signature of neutron recoils.
Right: Automatic pattern recognition of bubble sites in 3-D is achieved with two perpendicularly-placed cameras.

 

Figure 11. Figure 11.
Target liquids can be easily replaced in bubble chambers. Cycling available refrigerants containing I, F, Br, or combinations of these, would in many cases allow to clearly distinguish a neutron background from a WIMP signal, and might enable one to pinpoint the favored supersymmetric particle that best explains the combined measurements (i.e. would greatly reduce the number of supersymmetric models able to fit the observations, casting light on the nature of the neutralino).

 

Figure 12. Figure 12.
Expected rate of nucleation from underground neutrons as a function of hydrogenated moderator (polyethylene) thickness.

 

Figure 13. NEW: preliminary CF3I data (response to Am/Be neutron recoils).

 

References:

Go to top

  • [1] B. Hahn , Nucl. Phys. B (Proc. Suppl.) 36 (1994) 459.
  • [2] W.J Bolte et al. Development of Bubble Chambers With Enhanced Stability and Sensitivity to Low-Energy Nuclear Recoils, submitted to Phys. Rev. Lett.(2005)
  • [3] J.I. Collar et al., Phys. Rev. Lett. 85, 3083 (2000); J.I. Collar et al., New J. Phys. 2, 14.1 (2000).
  • [4] M. Harper, PhD Diss. (U. of Maryland, 1991), Nucl. Sci. Eng. 114, 118 (93), Nucl. Instr. Meth. A336, 220 (93); Ch. Peyrou in "Bubble and Spark Chambers", R.P. Shutt ed., (Academic Press, NY, 1967); F. Seitz, Phys. Fluids 1, 1 (1958); S.C. Roy et al., Nucl. Instr. and Meth. A255, 199 (1987).
  • [5] S.A. Pozzi et al., Nucl. Instr. Meth. A513 (2003) 550.

 

Articles about WIMPs and Superheated Liquids

PDF format

Go to top

 

Photo Gallery and Multimedia

Go to top

 
KICP Links: About KICPResearchEducation&OutreachNewsPeopleSeminarsWorkshops

Telephone:

   

(773) 702-4253

Fax:

   

(773) 834-8279

E-mail:

   

collaruchicago.edu

Postal Address:

   

5640 S. Ellis Ave., LASR 214, Chicago, IL 60637

Webmaster:

   

egaltsevakicp.uchicago.edu

Last update:

   

June 25, 2006