Research Activities
CF3I development:
Given the initial performance of the C2ClF5
SDD, a natural question was whether or not SDDs might have a similar impact on
spin-independent measurements. Given the then-prevailing attitude that, since
the cross section scales with the squares of both the mass number and the
WIMP-nucleus reduced mass, exploring the spin-independent channel of WIMP
interactions would require target nuclei with a significantly higher mass
number. Although several readily available "heavy"
refrigerants exist (eg. CF3Br, CF3I, XeF6,...),
the problem of density-matching them (~ 2-3 g/cm3) with the
suspension gels (r
~1 g/cm3) in order to achieve a homogeneous dispersion of the
refrigerant without introducing additional radio-contaminants inhibited their
development. For this reason, some recent attention has focused on the
development of a gel-free bubble chamber approach of COUPP, headed by Juan
Collar (a former SIMPLE person).
Given that the principle background
contribution in SDDs arises from the gel, and that this should be the same as
in the SIMPLE C2ClF5 experiment, the impact of the new
SDD implementation was estimated. Fig. 1 shows projections in both the
spin-independent and –dependent sectors for a 34 kgd exposure (comparable to
that of CDMS) with 1 evt/kgd background rate, with the “no evt” contours
indicating an ultimate limit to be realized if discrimination techniques were
to be identified and implemented.
|
Fig. 1a: projected spin-independent CF3I SDD exclusions. |
Fig. 1b:
projected spin-dependent CF3I SDD exclusions. |
|
An
alternative approach, at least in principle, was to match in viscosity rather
than density. We have recently succeeded to produce a gel matrix, using the standard
SIMPLE ingredients with the addition of agarose to provide a viscosity of
0.17 kg/m/s, as well as shift upwards the sol-gel transition temperature.
This permitted production of a prototype SDD with CF3I in 1-3
times the concentration of the SIMPLE devices, shown in Fig. 2 (the R&D of
the device prototype continues in progress. .In the current fabrication
protocol, and unlike in the C2ClF5 fabrications, about
50\% of the refrigerant dissolves into the gel due to its high solubility in
the weak hydrogen bond gel matrix, consistent with the solubility of CF3I
in water (16% f the gel) and glycerin (78% f the gel). This leads to a
fracturing of the gel once a bubble nucleates, and performance degradation.
This fracturing is inhibited by overpressuring the devices, but not
eliminated. There is also a significant presence of clathrates hydrates at
low temperature, implying that the device cannot be stored at temperature
below 0o C because the use of clathrates
hydrates breaks down locally the metastability of the droplets. Various
techniques to include the use of gelifying agents not requiring water as a
solvent or the use of others techniques to inhibit the diffusion of the
dissolved gas, are being explored. Also under investigation are new
constructs in the detector fabrication which would eliminate the fracturing
entirely. |
Fig. 2: prototype CF3I
SDD |
|
New Device Development: An
extensive program of device R&D is being carried out. The focus is both
on existing device improvements (such as detector lifetime extension using
agarose), and on the development of new devices such as C4F10,
C3F8, and C4F8. |
C4F10 C3F8 |
|
Device Response & Calibrations: The SDD
response has been studied using “distributed” sources of 252Cf and
Am/Be neutrons. In 2001, a filtered neutron beam facility was constructed on the
thermal column of the Portuguese research reactor and the Institute
Tecnologico e Nuclear, producing quasi-monoenergetic beams of 25-160 keV. The
thermal column facility remains in use; a new filtered beam facility has been
commissioned on a higher energy port of the reactor. Besides device
calibrations, it will be used for response studies of the SDDs for use in
neutron spectroscopy and dosimetry. |
SDD calibration using
filtered neutron beams at the RPI. |
Instrumentation:
The first SIMPLE data acquisition
was based on a piezoelectric transducer (PKM 13EPY-4002-Bo) connected to a
low-noise pre-amplifier (SSM2019) which coupled through a wide bandwidth dual
JFET input operational amplifier (TL082) to the input of an acquisition
channel. Mechanically the piezoelectric transducer was located in a protective
glycerine layer within the emulsion containment vessel, enclosed within a
copper mesh which acts as a Faraday cage. The transducer signal was amplified
by a factor of 105 , and recorded in a LabView platform, together
with the signals from other detectors, a wide-band hydrophone (Benthos AQ 4),
and an acoustic monitor placed outside the bath/shielding. The device was
however unable to discriminate bubble nucleations from microleaks arising from
the escape of the overpressuring gas into the surrounding water bath, nor
fracturing of the gel with bubble growth during the device use.
New
microphone–based instrumentation has been developed, based on a new high-quality
Panasonic MCE-200 electret microphone cartridge with adaptive electronics
(PGA2500).
|
SDD microphone plus
electronics |
SDD bubble nucleation
ouput waveform |
SDD bubble nucleation
ouput Fast Fourier Transform |
The
team has also explored pulse shape and frequency identification techniques as a
means of discriminating microleaks and fractures, and begins to focus on the
spatial identification of the bubble nucleation site using additional
microphones and x-y reduction techniques.
Currently,
the discrimination capability of the SDD is limited to the rejection of
coincident SDD signals in the experiment matrix, which addresses only the
neutrons leaking through the surrounding water shield. The bubble nucleation
process is however a four-stage process, the last two of which can generate an
acoustic pulse and the last of which generally provides the recorded signal.
The formation of a high temperature, high pressure zone (stage 2) is followed
by its rapid expansion (10-9 s) to a size at which the pressure
inside the bubble almost equals the external pressure; if the bubble size is
above a critical dimension, the fourth stage then sets in. The bubble expansion
in the third stage is soley attributable to the transformed energy of the
incident particle, whereas the fourth stage is due to the energy stored in the
liquid. The full signal should therefore consist of both a
fast and slow pulse, the fast component of which depends on the nature of the
incident radiation. To what extent this is detectable remains in question, and
the feasibility of measuring this stage as a discrimination technique using
ultrasound technology is being explored.
|
Analyses & Interpretations: Following
on the work of Tovey et. al., a model-independent framework for the
interpretation of spin-dependent search results has been elaborated. This
expands the spin-dependent scattering cross section as σSD ~ [
ap<Sp>
+ an <Sn>
]2 , and
yields 3D exclusion (ap, an, MW) plots
without using the Odd Group Approximation: each experiment provides a
different set of an,ap for a given MW, and
“standard” exclusions are thus of contours in the (an,ap)
plane for a given slice in MW. |
Model-independent
spin-dependent exclusions For NAIAD (left) and
DAMA/Xe (fight) . |
|
model-independent isospin-dependent exclusions for
the spin-independent sector. |
This
model-independent formalism has been extended to the spin-independent sector
assuming isospin-dependence of the WIMP-nucleon interaction, σSI ~ [ gp Z+ gn N ]2 where gp,n
are the spin-independent WIMP couplings with the proton, neutron.
|
|
Light
Nuclei experiments The
conventional wisdom in such experiments is that large A target nuclei are
required in spin-independent searches. The general interaction rate however
includes a nuclear form factor, which loses the coherence advantage at large
momentum transfers, permitting light nuclei experiments to provide comparable
restrictions given sufficient exposures. |
Light nuclei experiment projections in the spin-independent sector. |
