Constraints on inelastic dark matter from XENON10

It has been suggested that dark matter particles which scatter inelastically from detector target nuclei could explain the apparent incompatibility of the DAMA modulation signal (interpreted as evidence for particle dark matter) with the null results from CDMS-II and XENON10. Among the predictions of inelastically interacting dark matter are a suppression of low-energy events, and a population of nuclear recoil events at higher nuclear recoil equivalent energies. This is in stark contrast to the well-known expectation of a falling exponential spectrum for the case of elastic interactions. We present a new analysis of XENON10 dark matter search data extending to $E_{nr}=75\mathrm{keV}$ nuclear recoil equivalent energy. Our results exclude a significant region of previously allowed parameter space in the model of inelastically interacting dark matter. In particular, it is found that dark matter particle masses $m_{\chi }\gtrsim 150\mathrm{GeV}$ are disfavored.

Figure 1

Events remaining after reanalysis of 58.6 live days of dark matter search data. The dark matter acceptance box is bounded by the stair-step lines indicating the centroid and $-3\sigma$ bounds obtained from fits to the neutron calibration data. The fitted electron recoil centroid is shown as a solid blue curve, and the $-3\sigma$ contour is shown dashed. The 13 events remaining in the acceptance box are circled. An $S2$ software threshold of 12 electrons was imposed in the analysis.

Figure 2

(top) The distribution of $f_p$ for electron recoil calibration data. The black dashed line indicates the discrimination cut bound on this parameter. (bottom) The distribution of $f_p$ for nuclear recoil calibration data. The acceptance ${ϵ}_f$ of the cut is listed in Table 1.

Figure 3

The fraction of leakage events remaining among single gamma ray scatters in the electron recoil calibration data set (black stars). Also shown is the Monte Carlo prediction for false single scatters, with an additional scatter below the cathode grid (red triangles), or in the outer 8.7 kg of xenon (blue circles). The sum of the two Monte Carlo components is indicated by pink squares. Note that the $x$ axis is electron (not nuclear) recoil equivalent energy; $20\mathrm{keVee}=76\mathrm{keVr}$.

Figure 4

The solid black curves indicate the 90% C.L. exclusion limits obtained from XENON10 data, with ${\mathcal{L}}_{\mathrm{eff}}$ given by [24, 25] (solid gray curves take ${\mathcal{L}}_{\mathrm{eff}}$ from [22]). The filled regions indicate the 90% (magenta, darker) and 99% (cyan, lighter) C.L. DAMA-allowed regions.