Search for low-mass WIMPs with EDELWEISS-II heat-and-ionization detectors

We report on a search for low-energy ($E<20\mathrm{keV}$) WIMP-induced nuclear recoils using data collected in 2009–2010 by EDELWEISS from four germanium detectors equipped with thermal sensors and an electrode design (ID) which allows to efficiently reject several sources of background. The data indicate no evidence for an exponential distribution of low-energy nuclear recoils that could be attributed to WIMP elastic scattering after an exposure of $113\mathrm{kg}·\mathrm{d}$. For weakly interacting massive particles (WIMPs) of mass 10 GeV, the observation of one event in the WIMP search region results in a 90% C.L. limit of $1.0\times {10}^{-5}\mathrm{pb}$ on the spin-independent WIMP-nucleon scattering cross-section, which constrains the parameter space associated with the findings reported by the CoGeNT, DAMA and CRESST experiments.

Figure 1

Top panel: Distribution of the fiducial ionization energy, $E_i$, in neutron calibration for the ID3 detector, before (black) and after (red) the ionization cut. The spectrum contains mostly nuclear recoils and a tail of ionizationless events. Bottom: cut efficiency determination. The fit function (continuous red line) is ${ϵ}_{\mathrm{ion}}=0.95[1-\exp (-1.87(E_{\mathrm{ion}}-1.25))]$ where $E_{\mathrm{ion}}$ is in keVee (electron recoil scale). The dashed lines represent the efficiency functions estimated in the same way for the three other detectors used in this search.

Figure 2

Scatter plot of heat vs ionization for all selected events in the $113\mathrm{kg}·\mathrm{d}$ WIMP search. Neutron calibration data are also displayed as grey dots. The continuous (dashed) blue line represents the 95% gamma-ray rejection cut for detector ID3 (ID401). The blue, circled points are events contained in the WIMP search region for at least some WIMP masses.

Figure 3

Color map of the WIMP signal density $\rho (E_r,E_i)$ together with the 90% WIMP search box (red line) for a 10 GeV WIMP in detector ID3. Overlaid are background (black dots) and neutron calibration data (grey dots) for the same detector. The average position of the fiducial gamma-ray band and the corresponding 95% rejection cut are shown as continuous and dashed blue lines, respectively.

Figure 4

90% C.L. Poisson limit on ${\sigma }_{\mathrm{SI}}$ as a function of WIMP mass derived from the analysis of the four bolometers (bold red line). We also show the location of potential WIMP signals from the CoGeNT [7], CRESST [8] and DAMA [21] experiments, as well as constraints from EDELWEISS-II [11], CDMS-II [17] and XENON100 [20], and the dedicated low-mass searches by CDMS-II [4] and XENON10 [5].