Searches for extra-dimensional excitations in light-by-light scattering

TL;DR

This work analyzes the Randall-Sundrum Radion as a scalar from warped extra dimensions and its production in ultra-peripheral proton collisions via light-by-light scattering. It derives the Radion's couplings to SM fields, highlighting the trace anomaly–driven photon coupling and the non-minimal Higgs-curvature mixing that generates substantial interference with the Higgs sector. Through an EFT modeling of Radion and ALP production and a recasting of existing ALP limits, it shows that pure gravitational coupling is loop-suppressed while Higgs-Radion mixing can enhance signals by factors up to several hundred, enabling first exclusion contours in the $(\Lambda_r, \xi)$ plane for TeV-scale radions. The analysis emphasizes the necessity of incorporating full scalar mixing in interpreting LHC light-by-light data and provides a concrete methodology to translate ALP limits into RS Radion constraints. Overall, the paper demonstrates that current LHC data are already beginning to probe Radion scenarios with moderate mixing, while pure-gravity Radion remain largely unconstrained in this channel.

Abstract

We present a comprehensive phenomenological analysis of the Radion in the Randall-Sundrum model, focusing on its production via light-by-light scattering in ultra-peripheral proton-proton collisions at the LHC. We provide a consistent derivation of the effective couplings to Standard Model fields, clarifying the normalization of the trace anomaly-induced coupling to photons and the role of kinetic mixing with the Higgs boson. We demonstrate that while the pure gravitational coupling is loop-suppressed relative to Axion-Like Particles (ALPs), making the unmixed Radion elusive, the non-minimal mixing with the Higgs sector can induce constructive interference that enhances the signal by orders of magnitude. Using forward proton tagging to select exclusive high-mass events, we reinterpret recent experimental limits on ALPs to derive the first exclusion contours for the Radion in the $(Λ_r, ξ)$ plane, showing that mixing scenarios are beginning to be constrained by current LHC data.

Figures (5)

• Figure 1: The interference landscape for the Radion (Left: $m_\phi=200$ GeV, Right: $m_\phi=600$ GeV) in the $(\Lambda_r, \xi)$ plane. The greyscale indicates the enhancement factor relative to the pure curvature-induced prediction. Darker regions indicate constructive interference ($\xi > 0$), while lighter regions indicate destructive interference. The dotted line marks the "photophobic" region where the signal is suppressed.
• Figure 2: Simulated diphoton invariant mass distribution (normalized to unity) for a Radion resonance ($m_{\phi} = 150$ GeV, $\Gamma = 1$ GeV) at truth level. This validation plot confirms the correct generation of the Breit-Wigner line shape by the Monte Carlo framework described in Section 5.1, within the kinematic requirements region defined in Table \ref{['tab:selection_pp']}.
• Figure 3: ALP and Radion simulation distributions at truth level, both normalized to unity. The figure shows the comparison of the pseudorapidity separation $\Delta\eta$ between the two photons for ALP (grey) and Radion (red) signals at $m=150$ GeV. The close agreement supports the assumption that the selection efficiencies are comparable.
• Figure 4: Comparison of the light-by-light scattering cross-section $\sigma(\gamma\gamma \to X \to \gamma\gamma)$ for an ALP (grey solid line) and a Radion (colored dashed lines) as a function of mass, assuming a fixed scale $\Lambda = 1$ TeV for all models. The blue dotted line ($\xi=0$) shows the immense suppression of the pure trace anomaly coupling. The red dashed line ($\xi=0.5$) demonstrates the recovery of sensitivity due to Higgs-Radion mixing, though the rate remains suppressed relative to the ALP by a factor of $\sim 5 \times 10^{-4}$.
• Figure 5: Projected exclusion limits on the inverse symmetry breaking scale $1/\Lambda$ as a function of mass, assuming ${\cal B}(a, \phi \rightarrow \gamma \gamma)=1$. The grey area represents the region excluded for Axion-Like Particles (ALP) by current LHC light-by-light scattering searches. The red dashed line shows the translated limit for a Radion with mixing parameter $\xi=0.5$, and the blue dotted line shows the limit for a pure Radion ($\xi=0$). Higher values on the Y-axis correspond to stronger couplings (smaller $\Lambda$); thus, curves higher up indicate weaker experimental sensitivity to the scale $\Lambda$.