Complementarity of dark matter detectors in light of the neutrino background

Direct detection dark matter experiments looking for WIMP-nucleus elastic scattering will soon be sensitive to an irreducible background from neutrinos which will drastically affect their discovery potential. Here we explore how the neutrino background will affect future ton-scale experiments considering both spin-dependent and spin-independent interactions. We show that combining data from experiments using different targets can improve the dark matter discovery potential due to target complementarity. We find that in the context of spin-dependent interactions, combining results from several targets can greatly enhance the subtraction of the neutrino background for WIMP masses below $10\mathrm{GeV}/c^2$ and therefore probe dark matter models to lower cross sections. In the context of target complementarity, we also explore how one can tune the relative exposures of different target materials to optimize the WIMP discovery potential.

Figure 1

Left: Relevant neutrino fluxes to the background of direct dark matter detection experiments: Solar, atmospheric, and diffuse supernovae [22, 23, 24]. Note that for the DSNB fluxes we show the three different contributions corresponding to the temperatures of each neutrino flavor: $T_{{\nu }_e}=3\mathrm{MeV}$, $T_{{\overline{\nu }}_e}=5\mathrm{MeV}$ and $T_{{\nu }_x}=8\mathrm{MeV}$ where ${\nu }_x$ represents all remaining flavors. Right: Neutrino background event rates for a germanium based detector. The black dashed line corresponds to the sum of the neutrino induced nuclear recoil event rates. Also shown is the similarity between the event rate from a $6\mathrm{GeV}/c^2$ WIMP with a SI cross section on the nucleon of $4.4\times 10^{-45}{\mathrm{cm}}^2$ (black solid line) and the ${}^8\mathrm{B}$ neutrino event rate.

Figure 2

Evolution of the discovery limit for a SI interaction as a function of the exposure for idealized Xe experiments with perfect efficiency and a 3 eV (4 keV) threshold. The discovery limit is shown for a $6\mathrm{GeV}/c^2$ ($100\mathrm{GeV}/c^2$) WIMP mass for different values of the systematic uncertainty on the ${}^8\mathrm{B}$ (atmospheric) flux in the left (right) panel. The second and third regions (background subtraction and saturation regime) are well described by Eq. (8) as shown by the dashed lines corresponding to the different systematic uncertainties.

Figure 3

Evolution of the discovery limit as a function of exposure on the WIMP mass vs SI cross section plane. These limits are computed for an idealized Xe-based experiment with no other backgrounds, 100% efficiency, and an energy threshold of 3 eV to fully map the low WIMP mass discovery limit. Features appearing on the discovery limits with increasing exposures are due to the different components of the total neutrino background, see Table 2.

Figure 4

Discovery limits for single (left) and multitarget (right) based experiments considering the SI interaction. Also shown are the current sensitivity limits of several experiments: DAMIC [27], SIMPLE [28], SuperCDMS-LT [29], COUPP [30], ZEPLIN-III [31], EDELWEISS [32], CDMS II Ge [33], CDMSLite [34], XENON100 [35], LUX [36], CDMS-II Si [37], DAMA/LIBRA [38] and CRESST [39].

Figure 5

Same as Fig. 4 but considering the SD interaction on the proton.

Figure 6

Same as Fig. 4 but considering the SD interaction on the neutron.

Figure 7

Left: Reconstructed WIMP masses when considering the ${}^8\mathrm{B}$ neutrino background under the WIMP only hypothesis. Right: Figure of merit of target complementarity showing the mean values of the reconstructed WIMP-nucleon cross sections computed by considering the distributions of the maximum likelihood of the ${}^8\mathrm{B}$ neutrino under the WIMP only hypothesis for different target nuclei and 1000 Monte Carlo pseudoexperiments.

Figure 8

Left: Discovery limits for a Xe only, and combinations of Xe/Ge, Xe/Ge/Si and $\mathrm{Xe}/\mathrm{Ge}/\mathrm{Si}/{\mathrm{C}}_3{\mathrm{F}}_8$ based experiments for the SI interaction with thresholds set such that we do not have pp neutrino events in the data: 3, 5.3, 14, and 33 eV for Xe, Ge, Si, and C respectively. The exposure has been set such as an expected number of 1000 ${}^8\mathrm{B}$ neutrino events equally distributed between the targets was expected in the four cases. For the ${\mathrm{C}}_3{\mathrm{F}}_8$ based experiment we computed two limits with and without energy sensitivity. Also shown are the background-free limits obtained without considering the neutrino background (sensitivity limits). Right: Ratio between the discovery limits of a Xe based experiment, combinations of Xe/Ge, Xe/Ge/Si and $\mathrm{Xe}/\mathrm{Ge}/\mathrm{Si}/{\mathrm{C}}_3{\mathrm{F}}_8$ based experiments with their respective sensitivity limits. The gap between this ratio and 1 gives the effect of the background on the discovery potential. Also shown as dashed vertical lines are the WIMP masses that can be mimicked by different neutrino components.

Figure 9

Same as Fig. 8 but considering the SD interaction on the proton. For the ${\mathrm{C}}_3{\mathrm{F}}_8$ based experiment we computed two limits with and without energy sensitivity. The right panel shows clearly that combining data from ${\mathrm{C}}_3{\mathrm{F}}_8$ with other targets results in a significant decrease of the impact of the neutrino background on the discovery potential.

Figure 10

Same as Fig. 8 but considering the SD interaction on the neutron.

Figure 11

Evolution of the discovery limit as a function of the exposure for a Xe based experiment and combinations of Xe/Ge, Xe/Ge/Si based experiment for a $6\mathrm{GeV}/c^2$ WIMP mass. The considered threshold for Xe, Ge, and Si are respectively 1.0, 1.8, and 4.7 keV. We considered both the SI interaction (left panel) and the SD interaction on the proton (right panel). For an expected number of 100 ${}^8\mathrm{B}$ events, the corresponding exposures for Xe, Xe/Ge, and Xe/Ge/Si are respectively 1.005, 1.989, and 3.154 ton-year.

Figure 12

Evolution of the discovery limit of a combination of Ge and Si based experiments as a function of the considered Si fraction for several different exposures considering a WIMP mass of $6\mathrm{GeV}/c^2$. We considered both the SI and the SD interaction on the proton. For the SI interaction there is no optimal Si fraction because there is no visible effect from target complementarity. However, for the SD interaction on the proton, the discovery limit at $6\mathrm{GeV}/c^2$ has a minimum value for a certain Si fraction which depends on the considered exposure.