Higgs boson pair production in the D=6 extension of the SM

TL;DR

This paper investigates the dimension-6 EFT extension of the Standard Model to constrain Higgs sector modifications using Higgs pair production via gluon fusion at the LHC. By implementing the D=6 operators in a Monte Carlo framework and analyzing the hh → (bb)(ττ) channel with realistic backgrounds and theoretical uncertainties, the authors derive projected constraints on the c6 coefficient and, more broadly, on the EFT parameter space under several modeling assumptions. They demonstrate that HL-LHC data can substantially tighten bounds on c6 (to about |c6| ≲ 0.6 in the full model with current single-Higgs constraints) and illustrate how including multi-Higgs processes complements single-Higgs fits in a global EFT analysis. This work represents a first step toward incorporating multi-Higgs production into EFT fits and motivates exploring additional final states and differential observables for improved sensitivity.

Abstract

We derive the constraints that can be imposed on the dimension-6 effective theory extension of the Standard Model, using gluon fusion-initiated Higgs boson pair production at the LHC. We use a realistic analysis focussing on the $hh \rightarrow (b\bar{b}) ( τ^+ τ^- )$ final state, including initial-state radiation and non-perturbative effects. We include the statistical uncertainties on the signal rates as well as conservative estimates of the theoretical uncertainties. We first consider a theory containing only modifications of the trilinear coupling, through a $c_6 λ\, H^6/ v^2$ Lagrangian term, and then examine the full parameter space of the effective theory, incorporating current bounds obtained through single Higgs boson measurements. We also consider an alternative scenario, where we vary a smaller sub-set of parameters. Allowing, finally, the values of the other coefficients to vary within \textit{projected} experimental ranges, we find that the currently unbounded parameter, $c_6$, could be constrained to lie within $|c_6| \lesssim 0.6$ at 1$σ$ confidence, at the end of the high-luminosity run of the LHC (14~TeV) in the full model, and to $-0.6 \lesssim c_6 \lesssim 0.5$ in the alternative model. This study constitutes a first step towards the inclusion of multi-Higgs boson production in a full fit to the dimension-6 effective field theory framework.

Figures (11)

• Figure 1: The Feynman diagrams contributing to $gg \rightarrow hh$, including those induced by higher-dimensional operators. The grey blobs indicate the points of insertion of $D=6$ EFT vertices. At the order that we are considering in the present article, no two EFT insertions can occur in a single diagram. Diagrams with only one grey blob only appear in the effective theory.
• Figure 2: The effect of the variation of individual operators on the total cross section divided by the SM value. In the right panel we focus on a narrower region, showing in the grey-shaded area the $\pm 10\%$ variation with respect to the SM value. The solid portions of the curves represent the region which is compatible at 95% C.L. or more with the current Higgs boson data, obtained using HiggsBounds and HiggsSignals (see section \ref{['sec:full']} for details).
• Figure 3: The efficiency of the analysis is shown on the left panel. The right panel shows the efficiency times the cross section of the EFT point, divided by the SM cross section times the SM point efficiency.
• Figure 4: The $p$-value obtained for a given value of $c_6$, for the process $hh \rightarrow (b\bar{b}) (\tau^+ \tau^-)$ at 600 fb$^{-1}$ and 3000 fb$^{-1}$ of integrated luminosity. On the left figure we show the result without any theoretical uncertainty included ($f_\mathrm{th} = 0$) and on the right figure with theoretical uncertainty on the signal cross section prediction of 30% ($f_\mathrm{th} = 0.3$).
• Figure 5: The $p$-values obtained after marginalization over the directions orthogonal to the $(c_H,c_6)$-plane, for the process $hh \rightarrow (b\bar{b}) (\tau^+ \tau^-)$. On the top plots we show the results at 600 fb$^{-1}$ of integrated luminosity, without ($f_\mathrm{th} = 0.0$) and with ($f_\mathrm{th} = 0.3$) theoretical uncertainty included and on the bottom we show the corresponding plots at 3000 fb$^{-1}$. We also present the 1-sigma contours as black dashed lines.