DM-TPC: a new approach to directional detection of Dark Matter
G. Sciolla, S. Ahlen, D. Dujmic, V. Dutta, P. Fisher, S. Henderson, A. Kaboth, G. Kohse, R. Lanza, J. Monroe, A. Roccaro, N. Skvorodnev, H. Tomita, R. Vanderspek, H. Wellenstein, R. Yamamoto
aa r X i v : . [ a s t r o - ph ] M a y DM-TPC: a new approach to directional detection of Dark Matter
G. SCIOLLA a , S. AHLEN , D. DUJMIC , V. DUTTA , P. FISHER , S. HENDERSON ,A. KABOTH , G. KOHSE , R. LANZA , J. MONROE , A. ROCCARO , N. SKVORODNEV ,H. TOMITA , R. VANDERSPEK , H. WELLENSTEIN , R. YAMAMOTO (The DM-TPC Collaboration) Massachusetts Institute of Technology, Cambridge, MA 02139 (USA) Boston University, Boston, MA 02215 (USA) Brandeis University, Waltham, MA 02454 (USA)
Directional detection can provide unambiguous observation of Dark Matter interactions evenin presence of insidious backgrounds. The DM-TPC collaboration is developing a detectorwith the goal of measuring the direction and sense of nuclear recoils produced in Dark Matterinteractions. The detector consists of a Time Projection Chamber with optical readout filledwith CF gas at low pressure. A collision between a WIMP and a gas molecule results ina nuclear recoil of 1-2 mm. The measurement of the energy loss along the recoil allows usto determine the sense and the direction of the recoil. Results from a prototype detectoroperated in a low-energy neutron beam clearly demonstrate the suitability of this approach tomeasure directionality. A full-scale module with an active volume of about one cubic meteris now being designed. This detector, which will be operated underground in 2009, will allowus to set competitive limits on spin-dependent Dark Matter interactions using a directionaldetector. Searches for non-baryonic Dark Matter (DM) in the form of Weakly Interacting Massive Particles(WIMPs) rely on detection of nuclear recoils created by the elastic scattering between a WIMPand the detector material. In presence of backgrounds, an unambiguous positive observationcan be provided by detecting the direction of the incoming WIMP.1 As the Earth moves in thegalactic Dark Matter halo, the WIMPs appear to come at us with an average velocity of 220km/s from the direction of the constellation Cygnus. Due to the relative orientation betweenthe Earth’s rotation axis and the direction of the Dark Matter wind, the direction of the WIMPschanges with respect to our detector on Earth by about 90 degrees every twelve hours. Since nobackground is expected to correlate with the position of Cygnus in the sky, directional detectionwill improve the sensitivity to Dark Matter by orders of magnitude.2 If the detector is also ableto determine the sense of the direction, the sensitivity to DM is further enhanced.2 In additionto background rejection, the measurement of the direction of Dark Matter will also allow us todiscriminate between various Dark Matter models.Dark Matter particles can interact with ordinary matter via spin-independent (scalar) orspin-dependent (axial vector) interactions. Most of the current experiments concentrate onsearches for scalar couplings and are able to place very stringent limits on spin-independentinteractions excluding cross-sections above 10 − cm . a Corresponding author: [email protected] n contrast, the existing limits on axial-vector couplings are almost seven orders of magnitudeless stringent. Despite the modest experimental effort in this sector, these interactions areinteresting because they are expected to be enhanced with respect to the scalar interactions intheoretical models in which the Lightest Supersymmetric Particle has a substantial Higgsinocontribution.3 Therefore there is an urgent need for improving the searches for spin-dependentinteractions of Dark Matter. Materials rich in fluorine, with nuclear spin of 1/2, are the mostsuitable detector materials for such searches.4
The DM-TPC detector5 consists of a low pressure Time Projection Chamber (TPC) filled withtetra-fluoro-methane (CF ). For a pressure inside the vessel of 50 torr, the typical collisionof a WIMP with a gas molecule causes a nucleus to recoil by about 1 mm. The ionizationelectrons produced by the recoiling nucleus drift in the gas along a uniform electric field towardan amplification region created by two parallel woven meshes.6 The large electric fields presentin this region cause the avalanche process, during which a substantial amount of scintillationlight is produced.7 A CCD camera is used to detect such photons and to image the projectionof the nucleus parallel to the amplification plane. The total amount of light deposited in theCCD measures the total energy of the recoil. Because the energy loss is not uniform along thetrajectory, we can determine not only the direction of the incoming WIMP, but also its sense(“head-tail” measurement). A phototube provides a measurement of the recoil parallel to thedrift direction and serves as a trigger for reading out the CCD camera.The combination of the various measurements provided by this detector is very effective insuppressing backgrounds due to alpha particles, electrons, and photons. As an example, therejection factor measured for photons is better than 2 parts per million.The DM-TPC detector is designed with the goal of maturing into a large undergroundexperiment. The choice of an optical readout is motivated by the modest cost-per-channelobtainable with the use of a CCD camera, making it possible in the future to economically scalethe detector to large volumes. The choice of CF gas as active material is motivated by the lowtransverse diffusion and good scintillation properties of this gas. In addition, CF is non-toxicand non-flammable, making it easier to operate the detector underground. Finally, this gascontains four atoms of fluorine for each atom of carbon, making the DM-TPC detector ideal tostudy spin-dependent interactions. A small prototype of the DM-TPC detector, with an active volume of 10 × × . , hasbeen operational since Spring 2007. This chamber was built to demonstrate the validity of thedetector technology. In particular, this prototype was used to prove that our technology canindeed determine the sense of the direction of low-energy nuclear recoils (“head-tail” effect).The prototype was calibrated with 5.5 MeV alpha particles produced by a Am source.The energy loss (dE/dx) for 5.5 MeV alpha particles in CF was measured and compared withthe MC simulation. The agreement between data and MC was found to be excellent. The samealpha source was also used to measure the resolution as a function of the drift distance of theprimary electrons to study the effect of the diffusion. These measurements showed5 that thedrift distance should be limited to 25 cm.We have used this chamber to study the recoils of fluorine nuclei in interaction with lowenergy neutrons, which provide a signature very similar to that of a WIMP. For our tests weused a 14 MeV neutron source from a deuteron-triton tube as well as a Cf source. Theenergy of the reconstructed recoils was between 100 and 800 keV. Because of the small energy igure 1: Left: dE/dx distribution for a fluorine nucleus recoiling with an energy of about 250 keV after beingstruck by a neutron. The direction of the incident neutron is right to left. The larger energy loss at the right ofthe track and lower energy loss at the left of the track is characteristics of the “head-tail” effect. Right: skewnessof recoil events versus length of the recoil track, which is proportional to the recoil’s kinetic energy. Full (open)dots show data taken with the detector oriented parallel (anti-parallel) to the neutron flux. of the recoiling nucleus, we expect the energy deposition to uniformly decrease along the pathof the recoil, allowing us to identify the “head” (“tail”) of the event by a smaller (larger) energydeposition.Figure 1 (left) shows the energy-loss per unit length (dE/dx) of one recoiling nucleus. Thechange of dE/dx along the recoil track is clearly visible. To quantify the “head-tail” effect wedefine the skewness as γ ≡ µ /µ . , where µ and µ are the second and third moments of theenergy deposit along the track, respectively. Figure 1 (right) shows skewness as a function ofthe length of the recoil track. For our setup, we expect negative skewness to dominate, which isclearly observed in the data. Averaging over all energies we determine that (74 ± Cf source and a pressure of 75 torr provedthat our detector has good head-tail discrimination for recoils as low as 100 keV. The angularresolution for such recoils was measured to be better than 15 degrees. Figure 2 shows a typicalnuclear recoil reconstructed in our detector using the mesh-based amplification technique.
Given the promising results obtained by our R&D efforts, we decided to take this project tothe next level and build a larger detector with an active volume of about one cubic meter.When operated at 50 torr, this device will have an active mass of 250 g. In one year of operationunderground, such a detector will accumulate 90 kg-day of exposure. The successful operation ofthis detector will provide very competitive spin-dependent limits on Dark Matter cross-sections.If successful this effort will lay the foundation for a large (a few hundred kg) directional DarkMatter detector with substantial potential to directly observe Dark Matter and determine itsdirection. This experiment is an ideal fit for the Deep Underground Science and Engineering igure 2: Scintillation profile of a typical nuclear recoil candidate in a
Cf run at 75 torr. The neutron wasincident from the right (x axis) and the energy of the recoil shown was about 300 keV.
Laboratory (DUSEL) that is being planned in the Homestake mine in South Dakota.
Directional detection may hold the key to unambiguous observation of Dark Matter in presenceof backgrounds, and allows us to discriminate between models that predict Dark Matter to comefrom different directions in our galaxy.The DM-TPC collaboration is developing a detector to achieve this goal. The device consistsof a low-pressure TPC filled with CF gas and read out by an array of CCD cameras. Ourprototypes proved the detector concept and demonstrated its ability to reconstruct both thesense and direction of nuclear recoils above 100 keV. A larger detector is now being designed forunderground operations in 2009 with the goal of obtaining competitive results on spin-dependentinteractions using directional information. The success of this device will lay the foundationfor a large Dark Matter experiment that will be able to detect the direction of WIMPs anddiscriminate between DM models in our galaxy. Acknowledgments
This work was supported by the Advanced Detector Research Program of the U.S. Department ofEnergy (contract number 6916448), the National Science Foundation, the Reed Award Program,the Ferry Fund, the Pappalardo Fellowship program, the MIT Kavli Institute for Astrophysicsand Space Research, and the Physics Department at the Massachusetts Institute of Technology.
References
1. D. N. Spergel, Phys. Rev. D (1988) 1353.2. A. M. Green and B. Morgan, Astropart. Phys. , 142 (2007).3. U. Chattopadhyay and D. P. Roy, Phys. Rev. D 68 (2003) 033010; B. Murakami andJ. D. Wells, Phys. Rev. D 64 (2001) 015001; J. D. Vergados, J. Phys. G 30 (2004) 1127.4. J. R. Ellis and R. A. Flores, Phys. Lett. B 263 (1991) 259.5. D. Dujmic et al. , Nucl. Instrum. Meth. A 584 (2008) 327-333.6. D. Dujmic et al. , arXiv:0804.0497, submitted to Nucl. Instrum. Meth.7. A. Kaboth et al.et al.