A light scalar WIMP through the Higgs portal and CoGeNT

TL;DR

This work examines a light dark matter candidate in the form of a real scalar singlet $S$ coupled to the Higgs via the portal, noting that for $m_S \ll m_h$ the annihilation and spin-independent scattering cross sections are linked primarily through $m_S$ and the Higgs coupling. By confronting the model with CoGeNT and DAMA signals, along with CDMS-II and Xenon10/100 constraints that depend on the scintillation efficiency, the authors identify regions in $(m_S, \sigma_{SI})$ where the relic density $\Omega_{DM} h^2$ is consistent with observations and the direct-detection rate aligns with the data. They also discuss indirect constraints from Fermi-LAT dwarf galaxies and predictions for the Higgs invisible width, which differ between DAMA- and CoGeNT-favored regions, potentially offering collider-discriminating power. The results reveal a tension between light-WIMP explanations of direct-detection anomalies and Xenon100 limits, highly sensitive to the Leff assumptions, and highlight the role of the Higgs invisible decay channel as a complementary probe.

Abstract

If dark matter (DM) simply consists in a scalar particle interacting dominantly with the Higgs boson, the ratio of its annihilation cross section ---which is relevant both for the relic abundance and indirect detection--- and its spin-independent scattering cross section on nuclei depends only on the DM mass. It is an intriguing result that, fixing the mass and direct detection rate to fit the annual modulation observed by the DAMA experiment, one obtains a relic density in perfect agreement with its observed value. In this short letter we update this result and confront the model to the recent CoGeNT data, tentatively interpreting the excess of events in the recoil energy spectrum as being due to DM. CoGeNT, as DAMA, points toward a light DM candidate, with somewhat different (but not necessarily incompatible) masses and cross sections. For the CoGeNT region too, we find an intriguing agreement between the scalar DM relic density and direct detection constraints. We give the one $σ$ region favoured by the CDMS-II events, and our exclusion limits for the Xenon10 (2009) and Xenon100 data, which, depending on the scintillation efficiency, may exclude CoGeNT and DAMA. Assuming CoGeNT and/or DAMA to be due to scalar singlet DM leads to definite predictions regarding indirect detection and at colliders. We specifically emphasize the limit on the model that might be set by the current {\it Fermi}-LAT data on dwarf galaxies, and the implications for the search for the Higgs at the LHC.

Figures (5)

• Figure 1: SI cross section ($\sigma^0_{n}$) vs scalar singlet mass ($m_S$), for $\rho_{DM} = 0.3$ GeV/cm$^3$ and a standard Maxwellian velocity distribution (with mean velocity $220$ km/s and escape velocity $v_{esc}=650$ km/s, see our conventions in Arina:2009um). The green region corresponds to CoGeNT (minimum $\chi^2$, with contours at 90 and 99.9% C.L.), for which we have assumed that the excess at low recoil energies is entirely due to DM (assuming a constant background contamination). The DAMA regions (goodness-of-fit, also at 90 and 99.9% C.L.) are given both with (purple/orange) and without (purple, no fill) channelling. The blue region corresponds to the CDMS-II two events, at $1 \sigma$, which we obtained following the procedure of Kopp:2009qt. The blue (short-dashed) line is the 90% C.L. exclusion limit from CDMS-Si Akerib:2005kh. The black dotted line is the $90 \%$ C.L. exclusion limit from the Xenon10 2009 data set, using their scintillation efficiency Angle:2009xb, as also considered in Kopp:2009qt. The long-dashed line is based on the same data but using instead the smaller scintillation efficiency advocated in Manzur:2009hp (central value, at $1\sigma$ the corresponding exclusion can be found in Fitzpatrick:2010em). Finally, the brown lines (continuous) encompass the region predicted by the singlet scalar DM model corresponding to the WMAP range $0.094 \leq \Omega_{DM} h^2 \leq 0.129$, for $0.2 \leq f \leq 0.4$.
• Figure 2: $E dN/dE$ spectra produced by Pythia8.1 for a scalar singlet of mass $6$ GeV (blue, long dashed), $8$ GeV (dark blue, short dashed) and $10$ GeV (black, continuous line).
• Figure 3: Higgs invisible decay branching ratio for $m_h=120$ GeV (left panel) and $m_h=180$ GeV (right panel).
• Figure 4: Left panel: Effect of Poisson fluctuations on the expected signal (continuous curve) from a candidate with mass $= 10$ GeV and $\sigma_{SI} = 10^{-41}$ cm$^2$, assuming LeffMin (in green the Xenon100 bins, in black our bins -- see text). The two vertical dashed lines correspond to the 3PE and 4PE thresholds. Right panel: Effects of changing the cutoff in the recoil energy (see text) on the number of events above 3PE (in black) and 4PE (in red) for LeffMed (short dashed), LeffMin (continuous), for a candidate of $6.5$ GeV for LeffMed and of $7.5$ GeV for LeffMin, both for $\sigma_{SI} = 10^{-40}$ cm$^2$. The horizontal dashed line corresponds to the 90% C.L. exclusion limit, which corresponds to 2.3 events according to Poisson statistics.
• Figure 5: Xenon100 exclusion limits with 90% C.L., with threshold at 3 PE (in black, left panel) or 4 PE (in red, right panel). The curves correspond respectively to the LeffMed (short dashed), LeffMin (continuous) and LeffZep (long dashed) scintillation efficiency --- see text. For the sake of comparison, we have taken $v_{esc} = 544$ km$\cdot$s$^{-1}$ like the Xenon100 collaboration. The blue (dot-dashed) lines correspond to our predicted exclusion limit for Xenon100, using LeffMin and for an exposure of about 1 ton-days, assuming zero event.