Probing the Chern-Simons Portal at the HL-LHC through Displaced Vertices from W Boson Associated Production

TL;DR

The paper investigates the HL-LHC sensitivity to the Chern-Simons portal, an SM extension introducing a massive neutral vector boson $X$ tied to a new $U_X(1)$ and anomaly-induced topological interactions. Focusing on the associated production $pp \to W^\pm j X$ with $W^\pm \to e^\pm \nu_e$ and $X \to \mu^+\mu^-$, the study employs LO matrix-element generation, Pythia showering, and Delphes-3 phase-II detector simulation under $\langle PU \rangle = 200$, incorporating a displaced-vertex selection to target LLP decays. A multivariate TMVA-based Boosted Decision Tree discriminates signal from backgrounds including diboson/triboson and heavy-flavor processes; baseline cuts are augmented by a DV cut $l_{DV} > 2$ cm and tracker geometry constraints. With $\mathcal{L}=3000\,\mathrm{fb}^{-1}$, the analysis yields 95% CL exclusions in the $(g_{Xll}^2, c_W^2)$ plane for $m_X=5$, 10, 15 GeV, reaching $c_W^2$ down to $\mathcal{O}(10^{-6}-10^{-7})$ and $g_{Xll}^2$ down to $\mathcal{O}(10^{-7})$, demonstrating that HL-LHC displaced-vertex searches can meaningfully probe gauge-anomaly-induced interactions and set robust constraints on the Chern-Simons portal.

Abstract

This study explores the Chern-Simons portal model, an extension of the Standard Model that introduces a massive neutral vector boson $X$ associated with a $U_X(1)$ gauge symmetry. Motivated by gauge anomaly cancellation, the model incorporates heavy chiral fermions that induce observable effects through topological Chern-Simons interactions, despite being inaccessible at Large Hadron Collider energies. We investigate the associated production of the $X$ boson with a $W$ boson and jets at the High-Luminosity LHC with a center-of-mass energy of 14 TeV, considering different $X$ masses benchmarks. A multivariate analysis using Boosted Decision Trees is employed to separate signal from background processes. Detector effects are modeled using a fast simulation tuned to the HL-LHC environment, including realistic pile-up conditions with an average of 200 interactions per bunch crossing. We derive expected $95\%$ confidence level exclusion limits in two-dimensional parameter spaces involving the $X$ boson couplings. Our results demonstrate that the HL-LHC can achieve high sensitivity to gauge-anomaly-induced interactions, setting robust constraints on the $X$ boson coupling to the $W$ boson down to $\mathcal{O}(10^{-4})$, depending on $m_X$.

Figures (7)

• Figure 1: Representative Feynman diagram displaying the $WWX$ and $X\ell\ell$ vertices, corresponding to the couplings $c_W$ and $g_{Xll}$.
• Figure 2: Representative Feynman diagrams contributing to the leading-order associated production of the $X$ gauge boson with a jet and a $W$ boson in proton--proton collisions.
• Figure 3: The selection efficiencies for signal processes as a function of lifetime. The dashed lines indicate our imposed limits for the X boson decay length.
• Figure 4: Normalized (unit-area) distributions of the input observables used in the BDT for the signal hypotheses $m_X=5,10,15~\mathrm{GeV}$ and the dominant SM backgrounds, after the baseline selection. Variables span global kinematics, angular correlations, and displaced-track measurements.
• Figure 5: BDT score distributions for signal and background. Training and test samples agree within statistical uncertainties, indicating no overtraining.