Silicon Detector Dark Matter Results from the Final Exposure of CDMS II

We report results of a search for weakly interacting massive particles (WIMPS) with the silicon detectors of the CDMS II experiment. This blind analysis of 140.2 kg day of data taken between July 2007 and September 2008 revealed three WIMP-candidate events with a surface-event background estimate of ${0.41}_{-0.08}^{+0.20}(\mathrm{stat}{)}_{-0.24}^{+0.28}(\mathrm{syst})$. Other known backgrounds from neutrons and ${}^{206}\mathrm{Pb}$ are limited to $<0.13$ and $<0.08$ events at the 90% confidence level, respectively. The exposure of this analysis is equivalent to 23.4 kg day for a recoil energy range of 7–100 keV for a WIMP of mass $10\mathrm{GeV}/c^2$. The probability that the known backgrounds would produce three or more events in the signal region is 5.4%. A profile likelihood ratio test of the three events that includes the measured recoil energies gives a 0.19% probability for the known-background-only hypothesis when tested against the alternative $\mathrm{\text{WIMP}}+\mathrm{\text{background}}$ hypothesis. The highest likelihood occurs for a WIMP mass of $8.6\mathrm{GeV}/c^2$ and WIMP-nucleon cross section of $1.9\times {10}^{-41}{\mathrm{cm}}^2$.

Figure 1

Nuclear-recoil efficiency as a function of recoil energy after application of each selection criterion is shown. Each curve from top to bottom shows the cumulative effect of successive cuts on the data, with the second curve from the top (gray dashed) encompassing all data-quality cuts, trigger, and ionization threshold efficiencies, and the nuclear-recoil yield band efficiency as measured on ${}^{252}\mathrm{Cf}$ neutrons. The third curve (black dashed) adds the ionization radial cut to the above. The bottom curve (solid blue) adds both the phonon timing criteria and the recoil-energy thresholds, and hence shows the overall efficiency of this analysis. The abrupt drops in acceptance at low recoil energies reflect the elevated energy thresholds chosen for some detectors.

Figure 2

Ionization yield versus recoil energy in all detectors included in this analysis for events passing all signal criteria except (top) and including (bottom) the phonon timing criterion. The curved black lines indicate the signal region ($-1.8\sigma$ and $+1.2\sigma$ from the mean nuclear-recoil yield) between 7 and 100 keV recoil energies for detector 3 in tower 4, while the gray band shows the range of charge thresholds across detectors. Electron recoils in the detector bulk have yield near unity. The data are colored (dark to light gray) to indicate recoil energy ranges of 7–20, 20–30, and 30–100 keV to aid the interpretation of Fig. 3.

Figure 3

Normalized ionization yield (standard deviations from the nuclear-recoil band centroid) versus normalized phonon timing parameter (normalized such that the median of the surface-event calibration sample is at $-1$ and the cut position is at 0) for events in all detectors from the WIMP-search data set passing all other selection criteria. The black box indicates the WIMP-candidate selection region. The data are colored (dark to light gray) to indicate recoil energy ranges of 7–20, 20–30, and 30–100 keV. The thin red curves on the bottom and right axes are the histograms of the data, while the thicker green curves are the histograms of nuclear recoils from ${}^{252}\mathrm{Cf}$ calibration data; both are normalized to have the same arbitrary peak value.

Figure 4

Experimental upper limits (90% confidence level) for the WIMP-nucleon spin-independent cross section as a function of WIMP mass. We show the limit obtained from the exposure analyzed in this work alone (blue dotted line), and combined with the CDMS II Si data set reported in Refs. [21, 25] (blue solid line). Also shown are limits from the CDMS II Ge standard [17] and low-threshold [20] analysis (dark and light dashed red), EDELWEISS low-threshold [30] (long-dashed orange), XENON10 S2 only [31] (dash-dotted green), and XENON100 [32] (long-dash-dotted green). The filled regions identify possible signal regions associated with data from CoGeNT [33] (dashed yellow, 90% C.L.), DAMA/LIBRA [10, 34] (dotted tan, 99.7% C.L.), and CRESST [12, 35] (dash-dotted pink, 95.45% C.L.) experiments. The 68% and 90% C.L. solid contours for a possible signal from these data alone are shown in light blue. The blue dot shows the maximum likelihood point at ($8.6\mathrm{GeV}/c^2$, $1.9\times {10}^{-41}{\mathrm{cm}}^2$).