A Consistent Dark Matter Interpretation For CoGeNT and DAMA/LIBRA

TL;DR

Reconciles DAMA/LIBRA modulation with CoGeNT excess under a single elastically scattering DM particle with $m_{\rm DM}\sim 7$ GeV and $\sigma_{\rm DM-N}\sim 2\times 10^{-4}$ pb ($2\times 10^{-40}$ cm$^2$). Uses a standard halo model, the Helm form factor, and quenching-factor uncertainties to map compatible regions, assuming negligible channeling. Finds an overlapping parameter region and good fits to both signals, with preliminary CRESST oxygen-band events supporting the interpretation; shows these results remain consistent with XENON10/100 and CDMS constraints once $L_{\rm eff}$ and energy-scale uncertainties are considered. Highlights the role of experimental systematics in reconciling direct-detection results and proposes a falsifiable CoGeNT modulation test via an estimated ~40 kg-days per season exposure to reach ~3σ significance.

Abstract

In this paper, we study the recent excess of low energy events observed by the CoGeNT collaboration and the annual modulation reported by the DAMA/LIBRA collaboration, and discuss whether these signals could both be the result of the same elastically scattering dark matter particle. We find that, without channeling but when taking into account uncertainties in the relevant quenching factors, a dark matter candidate with a mass of approximately ~7.0 GeV and a cross section with nucleons of sigma_{DM-N} ~2x10^-4 pb (2x10^-40 cm^2) could account for both of these observations. We also comment on the events recently observed in the oxygen band of the CRESST experiment and point out that these could potentially be explained by such a particle. Lastly, we compare the region of parameter space favored by DAMA/LIBRA and CoGeNT to the constraints from XENON 10, XENON 100, and CDMS (Si) and find that these experiments cannot at this time rule out a dark matter interpretation of these signals.

Figures (6)

• Figure 1: Measurements of the germanium quenching factor ($Q_{\rm Ge} \equiv E_{\rm ionization}/E_{\rm Recoil}$) over the energy range of the excess events observed by CoGeNT. The solid line denotes the best fit normalization to these measurements, assuming the slope predicted by Lindhard theory ($k=0.20$). The dashed lines represent the upper and lower $2 \sigma$ normalizations, accounting only for statistical errors. For the measurements used, see Ref. Ge. Additional measurements by the CoGeNT collaboration span down to $E_{\rm Recoil}=0.7$ keV Ge2.
• Figure 2: The regions in the elastic scattering cross section (per nucleon), mass plane in which dark matter provides a good fit to the excess CoGeNT events and to the annual modulation reported by DAMA/LIBRA (upper frame), as well as the region in which the combination of CoGeNT+DAMA/LIBRA is well fit (lower frame). We have assumed that any effects of channeling are negligible and have adopted $v_0=230$ km/s and $v_{\rm esc}=600$ km/s. No errors associated with uncertainties in the form factors have been taken into account. If these and other systematics were fully included, the allowed region would be expected to increase considerably. See text for more details.
• Figure 3: The spectrum of events in CoGeNT (upper frame) and the spectrum of the annual modulation in DAMA/LIBRA (lower frame) for overall best fit dark matter parameters of $m_{\rm DM}=6.8$ GeV and $\sigma_{\rm DM-N}=1.58\times 10^{-4}$ pb. In the upper frame, the solid black line is the predicted result for signal plus background (with triggering and signal acceptance efficiency built into the model), whereas the dashed line is the background alone and points denote the measured values. In the lower frame, the solid line is the predicted signal and the points denote the measurements reported by DAMA/LIBRA. We have assumed that any effects of channeling are negligible and have adopted $v_0=230$ km/s and $v_{\rm esc}=600$ km/s. See text for more details.
• Figure 4: The preliminary spectrum of events in the oxygen band of the CRESST experiment, compared to the spectral shape predicted for the case of $m_{\rm DM}=6.8$ GeV and $\sigma_{\rm DM-N}=1.58\times 10^{-4}$ pb (which provides good fit to both CoGeNT and DAMA/LIBRA). The solid line is the predicted signal and the error bars denote the preliminary spectrum of events reported by the CRESST collaboration. We have adopted $v_0=230$ km/s and $v_{\rm esc}=600$ km/s. See text for more details.
• Figure 5: Constraints from the XENON 10 experiment Savage:2010tg. In each frame, the dashed line denotes the limit when using the central values of the scintillation efficiency, $L_{\rm eff}$, as measured by Manzur et al.manzur, whereas the dotted lines are derived using $\pm 1 \sigma$ values of $L_{\rm eff}$. In the upper frame, no assumptions are made regarding the values of $L_{\rm eff}$ at energies below 4 keV (for which no measurements exist). In the lower frame, $L_{\rm eff}$ is assumed to fall linearly below 4 keV. Considerably more relaxed constraints are obtained from other existing measurements of $L_{\rm eff}$leffuncertain. See text for more details.