Limits on Spin-Independent Interactions of Weakly Interacting Massive Particles with Nucleons from the Two-Tower Run of the Cryogenic Dark Matter Search

We report new results from the Cryogenic Dark Matter Search (CDMS II) at the Soudan Underground Laboratory. Two towers, each consisting of six detectors, were operated for 74.5 live days, giving spectrum-weighted exposures of 34 (12) kg d for the Ge (Si) targets after cuts, averaged over recoil energies 10–100 keV for a weakly interacting massive particle (WIMP) mass of $60\mathrm{GeV}/c^2$. A blind analysis was conducted, incorporating improved techniques for rejecting surface events. No WIMP signal exceeding expected backgrounds was observed. When combined with our previous results from Soudan, the 90% C.L. upper limit on the spin-independent WIMP-nucleon cross section is $1.6\times {10}^{-43}{\mathrm{cm}}^2$ from Ge and $3\times {10}^{-42}{\mathrm{cm}}^2$ from Si, for a WIMP mass of $60\mathrm{GeV}/c^2$. The combined limit from Ge (Si) is a factor of 2.5 (10) lower than our previous results and constrains predictions of supersymmetric models.

Figure 1

Ionization yield versus timing parameter (see text) for calibration data in T2Z3(Ge), with recoil energies in the range 10–100 keV. Typical bulk electron recoils from the ${}^{133}\mathrm{Ba}$ source of gamma rays (dots, light red) have yield near unity. Low-yield ${}^{133}\mathrm{Ba}$ events ($+$, black), attributed to surface electron recoils, have small values of timing parameter, allowing discrimination from neutron-induced nuclear recoils from ${}^{252}\mathrm{Cf}$ (○, medium blue), which show a wide range of timing parameter values. The vertical dashed line shows the minimum timing parameter allowed for WIMP candidates, and the box shows the approximate signal region, which is, in fact, weakly energy dependent.

Figure 2

Variables used to reject surface electron recoils (7–100 keV), for data from T2Z4 (Si, left) and T1Z2 (Ge, right), for the timing parameter (upper) and $\Delta {\chi }^2$ (lower). Light (red) lines show distributions of low-yield electron recoils from the ${}^{133}\mathrm{Ba}$ source; dark (blue) lines show distributions of nuclear recoils from the ${}^{252}\mathrm{Cf}$ source. Dashed lines indicate the minimum values for acceptable WIMP candidates. A cut on the timing parameter is used for the Ge detectors; a cut on $\Delta {\chi }^2$ is used for the Si detectors.

Figure 3

Ionization yield versus recoil energy for events in all Ge detectors (upper) and all Si detectors (lower) passing initial data selection cuts prior to applying the surface electron recoil rejection cut. The signal region consists of recoil energies exceeding 7 keV, shown with a vertical dashed line, and yields between the curved lines defined with recoils induced by the ${}^{252}\mathrm{Cf}$ source while WIMP-search data were still masked. Below 7 keV, separation between nuclear and electron recoils becomes poor. Events passing the surface electron recoil cut are a star (red) inside the signal region and dark filled circles (blue) outside. Bulk electron recoils with yield near unity are above the vertical scale limits.

Figure 4

WIMP-nucleon cross-section upper limits (90% C.L.) versus WIMP mass. The upper CDMS Ge curve uses data from only the current run; the lower Ge curve results from data from the current and previous runs [6]. Supersymmetric models allow the largest shaded region [17], and the smaller shaded (green) region [18]. The shaded region in the upper left (see text) is from DAMA [19], and experimental limits are from DAMA [20], EDELWEISS [21], and ZEPLIN [22].