Prospects for sub-EW-scale ALP searches via $γ+b\bar{b}$ signatures at the LHC using jet substructure techniques

TL;DR

The paper investigates the LHC's ability to probe light axion-like particles in the $m_a\in[10,100]$ GeV range that dominantly decay to $b\bar{b}$, by exploiting a boosted topology where $a$ recoils against a high-$p_T$ photon and is reconstructed within a large-radius jet using jet-substructure observables. The analysis combines soft-drop grooming, N$_2^{\mathrm{DDT}}$ tagging, and a novel variable-radius subjets flavour tagging to discriminate signal from multijet background, with a Crystal Ball fit to extract the $\mathrm{m}_{\mathrm{SD}}$ peak and a data-driven background estimate. Projected HL-LHC limits show sensitivity to ALP-fermion couplings of order $g_{\mathrm{aff}}\sim 0.04$–$0.1\ \mathrm{GeV}^{-1}$ for $m_a$ between $20$ and $70$ GeV, improving further if the photon threshold is lowered to $p_{T,\gamma}>100$ GeV (notably at $m_a\approx12$ GeV). The study demonstrates that a photon-tagged, jet-substructure-based search can access a previously unconstrained region of parameter space and complements existing astrophysical and collider bounds, with implications for trigger strategies and future explorations at HL-LHC and beyond.

Abstract

The current Large Hadron Collider (LHC) data show no clear indication of new physics and only incremental improvements are anticipated at the energy frontier in the near future. However, while the focus of the LHC has been on constraining TeV scale physics, new particles could still be hiding below the electroweak scale. In order to obtain sensitivity to a new light boson with couplings to SM fermions, a potentially promising decay channel, for resonances with mass $\gtrsim {\cal O}(10)$ GeV, would be the decay to $b\bar b$ pairs. The measurement of such signatures is challenging due to the trigger requirements at the LHC. In this work, we explore the LHC sensitivity to a light pseudoscalar, or axion-like particle (ALP), in the $b\bar{b}$ final state with an associated photon, using jet substructure techniques, in the mass range between 10 GeV and 100 GeV. We obtain projected exclusions on the ALP-fermion coupling in a region of phase space which has not so far been probed by direct searches. We further discuss the impact that lower trigger thresholds may have on the LHC reach.

Figures (11)

• Figure 1: (a) The Feynman diagram for the signal process, $pp\to a(\to b\bar{b})\gamma$, where the ALP-fermion couplings, $g_{\rm aff}$, are shown, and one representative Feynman diagram for the (b) non-resonant and (c) resonant background, $pp\to jj\gamma$.
• Figure 2: The soft drop mass distributions of AK1.0 jet by changing the soft drop parameter $\beta$, with fixed $z_{cut}=0.1$, (a) for the multijet background, and (b) ALP signal of mass $m_a=50$ GeV.
• Figure 3: The $10\%$ quantile of the $\mathrm{N}_2$ distribution in simulated multijet events, $\mathrm{N}_2^{10\%}$, used to define the $\mathrm{N}_2^{\mathrm{DDT}}$ variable. The distribution is shown in the $\rho-p_{\rm T}$ plane. The white dashed lines correspond to constant values of the jet mass, $\mathrm{m}_{\mathrm{SD}}$. We select the region $-7.5<\rho<-2.0$ for our analysis, as shown with vertical gray lines.
• Figure 4: The soft drop mass distributions of AK1.0 jet for the multijet background and ALP signal. The ALP signals are shown as $\gamma$ a$_{m_a}$, for example, $\gamma$ a$_{\rm 12~GeV}$ corresponds to signal process with $m_a=$ 12 GeV. The $\gamma+$multijet background includes contributions from $W+\gamma$ and $Z+\gamma$ processes. The event yield on the y-axis is evaluated after all the selection cuts (see Table \ref{['tab:cuts']}). The background distribution is corrected with an overall factor of $0.74$CMS:2017dcz.
• Figure 5: Comparison of double subjet $b$-labelling efficiency for different choices of the $\rho$$_{\mathrm{VR}}$ parameter while constructing the VR subjets. The efficiencies are shown for the simulated ALP signal with masses, (a) $m_a=$ 12 GeV, (b) $m_a=$ 40 GeV (c) $m_a=$ 60 GeV and (d) $m_a=$ 100 GeV. The $R_{\mathrm{min}}$ and $R_{\mathrm{max}}$ values are set to 0.02 and 0.4 ATL-PHYS-PUB-2017-010, respectively.