Search for Low-Mass Weakly Interacting Massive Particles with SuperCDMS

We report a first search for weakly interacting massive particles (WIMPs) using the background rejection capabilities of SuperCDMS. An exposure of 577 kg days was analyzed for WIMPs with mass $<30\mathrm{GeV}/c^2$, with the signal region blinded. Eleven events were observed after unblinding. We set an upper limit on the spin-independent WIMP-nucleon cross section of $1.2\times 10^{-42}{\mathrm{cm}}^2$ at $8\mathrm{GeV}/c^2$. This result is in tension with WIMP interpretations of recent experiments and probes new parameter space for WIMP-nucleon scattering for WIMP masses $<6\mathrm{GeV}/c^2$.

Figure 1

Cumulative efficiencies after sequential application of each stage of event selection. From top to bottom, these are data-quality criteria, trigger and analysis thresholds, preselection criteria, and BDT discrimination with 68% C.L. ($\text{stat}+\text{syst}$) uncertainty band. The preselection and BDT selection efficiencies are interpolated from measurements in 1-keV bins. Steps are due to time-dependent analysis thresholds for individual detectors. For illustrative purposes, the Lindhard nuclear-recoil energy is provided as an approximate nuclear-recoil energy scale.

Figure 2

Top: Stacked histogram showing the components of the background model passing the preselection criteria summed over all detectors (neutron backgrounds are negligible and not included). For comparison, a $10\mathrm{GeV}/c^2$ WIMP with cross section $6\times {10}^{-42}{\mathrm{cm}}^2$ is shown on top of the total background. Events passing preselection criteria are overlaid (markers with statistical errors). A $p$-value statistic comparing the data to background model is 14% for this selection. Bottom: Difference between the data and the background expectation. Tan bars indicate the systematic uncertainty (68% C.L.) on the background estimate. Each component of the background model was computed prior to unblinding and was not fit or rescaled to match the data.

Figure 3

Small gray dots are all veto-anticoincident single-scatter events within the ionization-partition fiducial volume that pass the data-quality selection criteria. Large encircled shapes are the 11 candidate events. Overlapping shaded regions (from light to dark) are the 95% confidence contours expected for 5, 7, 10, and $15\mathrm{GeV}/c^2$ WIMPs, after application of all selection criteria. The three highest-energy events occur on detector T5Z3, which has a shorted ionization guard. The band of events above the expected signal contours corresponds to bulk electron recoils, including the 1.3-keV activation line at a total phonon energy of $\sim 3\mathrm{keV}$. High-radius events near the detector sidewalls form the wide band of events with near-zero ionization energy. For illustrative purposes, the Lindhard nuclear-recoil energy is provided as an approximate nuclear-recoil energy scale.

Figure 4

The 90% confidence upper limit (solid black) based on all observed events is shown with 95% C.L. systematic uncertainty band (gray). The preunblinding expected sensitivity in the absence of a signal is shown as 68% (dark green) and 95% (light green) C.L. bands. The disagreement between the limit and sensitivity at high WIMP mass is due to the events in T5Z3, whose occurrence weakens the limit in this range. Closed contours shown are CDMS II Si [3] (dotted blue, 90% C.L.), CoGeNT [4] (yellow, 90% C.L.), CRESST-II [5] (dashed pink, 95% C.L.), and DAMA/LIBRA [34] (dash-dotted tan, 90% C.L.). The 90% C.L. exclusion limits shown are CDMS II Ge [22] (dotted dark red), CDMS II Ge low threshold [17] (dashed-dotted red), CDMSlite [20] (solid dark red, extending to $\sim 3\mathrm{GeV}/c^2$), LUX [35] (solid green, extending to $\sim 5.5\mathrm{GeV}/c^2$), XENON10 S2 only [19, 36] (dashed dark green, extending to $\sim 4.5\mathrm{GeV}/c^2$), and EDELWEISS low threshold [18] (dashed orange, extending to $\sim 7\mathrm{GeV}/c^2$).