Missing Energy Signatures of Dark Matter at the LHC

TL;DR

This work investigates missing-energy signatures at the LHC to constrain dark matter interactions with the Standard Model via an effective field theory (EFT) framework. By analyzing mono-jet and mono-photon channels from ATLAS and CMS, the authors set bounds on DM–quark and DM–gluon operators $O_V$, $O_A$, $O_t$, and $O_g$, translated into direct and indirect detection cross sections, and examine the impact of heavy vs. light mediators. They find that high-$p_T$ mono-jet analyses yield the strongest LHC limits for many operators up to $m_\chi \sim 1$ TeV, with the LHC providing particularly strong constraints on spin-dependent scattering. The study also highlights the Higgs portal as a powerful alternative: invisible Higgs decays can dominate sensitivity for light DM and light Higgs masses, and current Higgs limits can even set lower bounds on DM–nucleon interactions mediated by the Higgs. Overall, the paper demonstrates the complementarity of collider searches with direct and indirect detection, and discusses the limitations of EFT in scenarios with light mediators or Higgs-mediated couplings.

Abstract

We use ATLAS and CMS searches in the mono-jet + missing energy and mono-photon + missing energy final state to set limits on the couplings of dark matter to quarks and gluons. Working in an effective field theory framework we compare several existing mono-jet analyses and find that searches with high p_T cuts are more sensitive to dark matter. We constrain the suppression scale of the effective dark matter-Standard Model interactions, and convert these limits into bounds on the cross sections relevant to direct and indirect detection. We find that, for certain types of operators, in particular spin-independent dark matter-gluon couplings and spin-dependent dark matter-quark couplings, LHC constraints from the mono-jet channel are competitive with, or superior to, limits from direct searches up to dark matter masses of order 1 TeV. Comparing to indirect searches, we exclude, at 90% C.L., dark matter annihilating to quarks with the annihilation cross section of a thermal relic for masses below ~ 15-70 GeV, depending on the Lorentz structure of the effective couplings. Mono-photon limits are somewhat weaker than mono-jet bounds, but still provide an important cross check in the case of a discovery in mono-jets. We also discuss the possibility that dark matter--Standard Model interactions at LHC energies cannot be described by effective operators, in which case we find that constraints can become either significantly stronger, or considerably weaker, depending on the mass and width of the intermediate particle. We also discuss the special case of dark matter coupling to the Higgs boson, and we show that searches for invisible Higgs decays would provide superior sensitivity, particularly for a light Higgs mass and light dark matter.

Figures (11)

• Figure 1: Dark matter production in association with a single jet in a hadron collider.
• Figure 2: Measured missing energy spectra of $j + \slashed {E}_{T}$ for the three ATLAS analyses and the CMS analysis discussed in the text (black data points with error bars) compared to the collaborations' background predictions (yellow shaded histograms) and to our Monte Carlo prediction with (blue histograms) and without (black dotted lines) a dark matter signal. In all cases the DM signal comes from the vector operator, $\mathcal{O}_V$, and $m_\chi = 10\,{\rm GeV}$, $\Lambda=400\,{\rm GeV}$. Our simulations are rescaled to match the overall normalization of the collaborations' background predictions.
• Figure 3: Limits on the suppression scale $\Lambda$ for the vector operator, $\mathcal{O}_V$, where only the coupling to up quarks is considered, for the three ATLAS analyses and the analysis of CMS. In all cases the observed (expected) bound is represented by a solid (dashed) line.
• Figure 4: Limits on the suppression scale $\Lambda$ for various operators, where only the coupling to one quark flavor at a time is considered, for the veryHighPT ATLAS analysis. In all cases the observed (expected) bounds are shown as solid (dashed) lines.
• Figure 5: ATLAS limits on (a) spin-independent and (b) spin-dependent dark matter--nucleon scattering, compared to limits from the direct detection experiments. In particular, we show constraints on spin-independent scattering from CDMS Ahmed:2009zw, XENON-10 Angle:2007uj, XENON-100 Aprile:2011hi, DAMA Bernabei:2008yi, CoGeNT Aalseth:2011wpFox:2011px and CRESST Angloher:2011uu, and constraints on spin-dependent scattering from DAMA Bernabei:2008yi, PICASSO BarnabeHeider:2005pg, XENON-10 Angle:2008we, COUPP Behnke:2010xt and SIMPLE Girard:2011xc. DAMA and CoGeNT allowed regions are based on our own fits Kopp:2009qtFox:2011fxFox:2011px to the experimental data. Following Hooper:2010uy, we have conservatively assumed large systematic uncertainties on the DAMA quenching factors: $q_{\rm Na} = 0.3 \pm 0.1$ for sodium and $q_{\rm I} = 0.09 \pm 0.03$ for iodine, which leads to an enlargement of the DAMA allowed regions. All limits are shown at 90% confidence level, whereas for DAMA and CoGeNT we show 90% and $3\sigma$ contours. For CRESST, the contours are $1\sigma$ and $2\sigma$ as in Angloher:2011uu.