The DMTPC project

The simplest models of the distribution of WIMPs in our Galaxy suggest that the orbital motion of the Sun about the Galactic center will result in an Earth-bound observer to experience a WIMP wind with speed 220 km/s (the galacto-centric velocity of the Sun) originating from the direction of the Sun’s motion. Because the Earth’s rotation axis is oriented at 48 • with respect to the direction of the WIMP wind, the average direction of DM particles recorded by an Earth-bound observer changes by about 90 o every 12 hours [2]. The ability to measure such a direction would provide a powerful suppression of insidious backgrounds such as cosmogenic neutrons and neutrinos from the Sun [3], as well as a unique instrument to test local DM halo models [4,5]. This capability makes directional detectors unique observatories for underground WIMP astronomy.

Dark Matter interactions in the detector generate nuclear recoils with typical energies of few tens of keV (Fig. 1, left). The direction of the recoiling nucleus encodes the direction of the incoming DM particle (Fig. 1, right). To observe the daily modulation in the direction of the DM wind, an angular resolution of ≈ 30 • in the reconstruction of the recoil nucleus is required.

Directional DM experiments use low-pressure (40-100 torr) gas as target material, in which typical DM induced nuclear recoils have a length of a few millimeters. A 3-D reconstruction of the recoil track with a spatial resolution of several hundred microns in all three coordinates is sufficient to achieve the desired angular resolution. A 2-D reconstruction is still valuable, although it slightly degrades the sensitivity [6]. “Head-tail” discrimination is also very important as it improves the sensitivity for detecting DM by about one order of magnitude [6].

Due to the low-density of the target material, directional detectors tend to be large in volume: a typical directional detector with a fiducial mass of one ton occupies O(10 3 ) cubic meters. Therefore, the success of a directional DM program depends very strongly on the developement of detectors with a low cost per unit volume. The largest expense for standard gaseous detectors is the cost of the readout electronics, making it essential to utilize low-cost readouts in order to make DM directional detectors financially viable.

The DMTPC detector is a low-pressure TPC with optical readout. The detection principle is illustrated in Fig. 2, left. The TPC is filled with CF 4 at a pressure of about 50 torr. At this pressure, a typical 50 keV nuclear recoil has a length of about 2 mm. The average ionization energy in CF 4 is 54 eV [7], which results in about 10 3 primary electrons from the nuclear recoil. Such electrons drift toward the amplification region, where photons are produced together with electrons in the avalanche process. A CCD camera mounted above the cathode mesh records the image of the nuclear recoil projected along the amplification plane. The sense of the recoiling nucleus can be determined (“head-tail” discrimination), because the energy loss (dE/dx) is not constant along the trajectory. An array of photomultipliers (PMTs) mounted above the cathode mesh determines the length of the recoil in the drift direction through time measurements. CF 4 was chosen as the target material primarily because of its high fluorine content. Fluorine is an excellent element to detect spin-dependent interactions on protons [8], due to its large spin factor and isotopic abundance (Table 1). Spin-dependent (SD) interactions are predicted to dominate over spin-independent ones in models where the lightest super-symmetric particle has a substantial Higgsino contribution [9,10].

CF 4 is also an excellent detector material [11,12]. Its good scintillation properties are very important for the optical readout. Our recent measurements [13] indicate that in CF 4 the number of photons produced between 200 and 800 nm wavelength is about 1 for every 3 avalanche electrons. Moreover, the low transverse diffusion characteristic of electrons in CF 4 allows for good spatial resolution in the reconstruction of the recoil track despite the long (25 cm) electron drift distance. Finally, CF 4 is non-flammable and non-toxic allowing for safe operation underground.

CCDs provide a true 2-D readout at a much lower cost-per-channel than any other readout technology used in particle physics. A modern low-noise CCD camera with 4 megapixels can be purchased today for a few thousand US dollars, which corresponds to 10 -3 dollars/channel. The cost of a directional DM detector is dominated by the readout electronics, making the optical readout a solution toward an econom

…(Full text truncated)…