Taking a Razor to Dark Matter Parameter Space at the LHC

TL;DR

This paper extends collider searches for dark matter by applying the CMS razor analysis to multi-jet plus missing-energy final states, offering a more inclusive and information-rich alternative to monojet searches. By employing simplified DM models with vector/axial-vector mediators or gluon-coupled scalars, and by considering both EFT and UV-complete mediator scenarios, it derives competitive bounds on the DM–quark/gluon couplings and translates them into direct-detection and annihilation cross-section limits. The study also assesses EFT validity via unitarity arguments and explores the impact of light mediators, showing that mediator mass and width can either strengthen or weaken the bounds depending on the mass regime. Overall, razor-based analyses provide a complementary and increasingly powerful tool for constraining DM at the LHC, with clear prospects for improvement as more data become available.

Abstract

Dark matter (DM) has been searched for at colliders in a largely model independent fashion by looking for an excess number of events involving a single jet, or photon, and missing energy. We investigate the possibility of looking for excesses in more inclusive jet channels. Events with multiple jets contain more information and thus more handles to increase the signal to background ratio. In particular, we adapt the recent CMS "razor" analysis from a search for supersymmetry to a search for DM. We consider simplified models where DM is a Dirac fermion that couples to the quarks of the Standard Model (SM) through exchange of vector or axial-vector mediators or to gluons through scalar exchange. We consider both light and heavy (leading to effective contact interactions) mediators. Since the razor analysis requires multiple jets in the final state, the data set is complementary to that used for the monojet search and thus the bounds can be combined.

Figures (10)

• Figure 1: $R^2$ vs. $M_R$ distribution for SM backgrounds (a) $(Z\to \bar{\nu}\nu)+$jets, (b) $W+$jets (including decays to both $\ell^{inv}$ and $\tau^h$, (c) $t\bar{t}$, and (d) DM signal with $M_{\chi}=100$ GeV and $\Lambda= 644$ GeV. In all cases the number of events are what is expected after an integrated luminosity of 800 pb$^{-1}$. The cuts applied in $M_R$ and $R^2$ are shown by the dashed lines and the "signal" region is the upper right rectangle.
• Figure 2: $R^2$ vs. $M_R$ for various DM masses with $u$-only vectorial couplings with arbitrary normalization.
• Figure 3: $R^2$ vs. $M_R$ for DM coupling to (a) sea quarks (in this case the $s$-quark) and (b) gluons with arbitrary normalization.
• Figure 4: Cutoff scale $\Lambda$ bounds for vector, axial, and gluon couplings. The error band is determined by varying $\sigma_{SM}$ between $\sqrt{N_{SM}}$ and $\sigma_{SM} = 2\sqrt{N_{SM}}$. The dashed line is the bound determined by the monojet analysis Fox:2011pm.
• Figure 5: Combined razor and monojet $\Lambda$ bounds. The solid lines are the razor bounds and the dashed lines are the combined bounds.