Higgs pair production via gluon fusion in the Two-Higgs-Doublet Model

TL;DR

This work delivers NLO QCD predictions for Higgs pair production via gluon fusion in the Two-Higgs-Doublet Model (2HDM), including a loop-improved reweighting approach to extend beyond the infinite top-mass limit and matching to parton showers. It analyzes all seven di-Higgs final states within a set of phenomenologically viable 2HDM benchmarks, revealing that heavy-Higgs resonances and Yukawa-coupling patterns can dramatically modify total rates and differential distributions, while QCD corrections substantially reduce theoretical uncertainties. The results show large K-factors for gluon-fusion channels and detailed interference structures between triangle and box amplitudes, with resonant scenarios potentially enhancing light-Higgs production by orders of magnitude relative to the SM. The study provides fully differential, NLO+PS event samples accessible through public tools, enabling realistic simulations and informing strategies to identify extended Higgs-sector signatures at the LHC.

Abstract

We study the production of Higgs boson pairs via gluon fusion at the LHC in the Two-Higgs-Doublet Model. We present predictions at NLO accuracy in QCD, matched to parton showers through the MC@NLO method. A dedicated reweighting technique is used to improve the NLO calculation upon the infinite top-mass limit. We perform our calculation within the MadGraph5_aMC@NLO framework, along with the 2HDM implementation based on the NLOCT package. The inclusion of the NLO corrections leads to large K-factors and significantly reduced theoretical uncertainties. We examine the seven 2HDM Higgs pair combinations using a number of representative 2HDM scenarios. We show how the model-specific features modify the Higgs pair total rates and distribution shapes, leading to trademark signatures of an extended Higgs sector.

Figures (10)

• Figure 1: Generic Feynman diagrams describing the production of neutral Higgs boson pairs ($h = \PHiggslightzero, \PHiggsheavyzero, \PHiggspszero$) in the 2HDM through gluon fusion at leading order. The Feynman diagrams have been generated using FeynArts.styHahn:2000kx.
• Figure 2: Generic Feynman diagrams describing the production of mixed $\mathcal{C}\mathcal{P}$--even / $\mathcal{C}\mathcal{P}$--odd neutral Higgs boson pairs [$\PHiggslightzero\PHiggspszero, \PHiggsheavyzero\PHiggspszero$] in the 2HDM at leading order. We separately show the two possible partonic initial--states i) gluon fusion (left, center); ii) $q\bar{q}$ annihilation (right--most).
• Figure 3: Sample Feynman diagrams describing the production of neutral Higgs pairs ($h = \PHiggslightzero, \PHiggsheavyzero$) via gluon fusion at next--to--leading order in the 2HDM. The shaded blobs denote the effective Higgs couplings to the gluons in the HEFT approach.
• Figure 4: Light Higgs pair differential rates as a function of a) the di--Higgs invariant mass $m_{\PHiggslightzero\PHiggslightzero}$ (left panels); and b) the hardest Higgs transverse momentum $p_T^{\PHiggslightzero}$ (right panels). We separately show the results at LO+PS and NLO+PS accuracy in QCD, where the latter correspond to the "loop--improved" approach. The NLO+PS prediction for the SM is overlayed for comparison. In the lower subpannels we display the bin--by--bin ratio of the 2HDM prediction at NLO+PS over the corresponding SM result. The LHC center--of--mass energy is $\sqrt{S} = 14$ TeV. The 2HDM parameters are fixed to benchmark B1.
• Figure 5: Light Higgs pair differential rates as a function of a) the di--Higgs invariant mass $m_{\PHiggslightzero\PHiggslightzero}$ (left panels); and b) the hardest Higgs transverse momentum $p^{\PHiggslightzero}_T$ (right panels), in the same setup as for Figure 4. The 2HDM parameters are fixed as in benchmark B2.