Probing new physics in the Boosted $HH o bar{b}γγ$ channel at the LHC

Probing new physics in the Boosted $HH 	o bar{b}γγ$ channel at the LHC
Notice: This research summary and analysis were automatically generated using AI technology. For absolute accuracy, please refer to the [Original Paper Viewer] below or the Original ArXiv Source.

This paper presents the first dedicated study of the boosted $HH \to b\bar{b}γγ$ topology as a key probe of physics beyond the Standard Model (SM) in the high-energy double-Higgs boson regime. The analysis presented in this paper, focuses on two classes of new-physics scenarios: non-resonant deviations of the quartic gauge–Higgs interaction, parameterized by the coupling modifier $κ_{2V}$, and resonant enhancement arising from the decay of a heavy scalar state, modeled within a two-Higgs-doublet framework. We demonstrate that the boosted reconstruction category enhances sensitivity to beyond SM effects that populate the high-$m_{HH}$ tail, yielding improved constraints on $κ_{2V}$ and extending the discovery reach for heavy resonances.


💡 Research Summary

This paper presents the first dedicated study of the boosted Higgs‑pair topology HH → b b̄ γγ at the LHC, focusing on two classes of beyond‑Standard‑Model (BSM) effects that become most visible in the high‑energy regime: non‑resonant modifications of the quartic gauge–Higgs coupling (parameterised by κ₂V) and resonant production of a heavy scalar X that decays into a Higgs pair, modelled within a Type‑I two‑Higgs‑doublet framework.

The authors recognise that existing HH → b b̄ γγ analyses have been confined to the “resolved” regime, where the two b‑quarks from H → b b̄ are reconstructed as separate small‑R (R = 0.4) jets. In the boosted regime, when the Higgs transverse momentum exceeds ≈250 GeV, the b‑quarks become collimated (ΔR ≈ 2 m_H/p_T) and are captured inside a single large‑R (R = 1.0) jet. By exploiting modern jet‑grooming (trimming, pruning, soft‑drop) and double‑b‑tagging, the analysis can recover the Higgs decay even at multi‑TeV scales.

Two mutually exclusive categories are defined: (i) a resolved category using two b‑tagged small‑R jets and an XGBoost multivariate classifier, and (ii) a boosted category using a groomed large‑R jet with double‑b‑tagging and a simple rectangular cut‑based selection (the latter is sufficient because the signal topology is already highly distinctive at high p_T).

Signal samples are generated for gluon‑fusion (ggF) HH at NLO with Powheg‑Box (including finite top‑mass effects) and normalised to NNLO cross‑sections, and for vector‑boson‑fusion (VBF) HH at LO with MadGraph5_aMC@NLO, normalised to N³LO QCD + NLO electroweak predictions. VBF samples are produced for κ₂V = 1 (SM) and for κ₂V = 0, ±1.5, ±2.5 to probe the sensitivity to the quartic coupling. Resonant samples are generated for a heavy scalar X with masses m_X = 1–5 TeV (step 0.5 TeV) within a simplified 2HDM‑Type‑I, each normalised to a benchmark σ = 1 fb. Backgrounds include single‑Higgs → γγ (ggF, VBF, ZH, ttH) generated with Powheg‑Box and continuum γγ + jets generated with MadGraph5_aMC@NLO including up to two extra partons. All events are showered with Pythia 8.186 and passed through a fast ATLAS Run‑3 detector simulation (Delphes).

Object reconstruction follows Run‑3 ATLAS definitions: photons with p_T > 20 GeV, |η| < 2.37 (excluding the barrel‑endcap transition), small‑R jets with p_T > 25 GeV, |y| < 4.5, b‑tag efficiency 85 % (c‑mistag 17 %, light‑mistag 1 %). Large‑R jets are required to have 250 GeV < p_T < 3 TeV, |y| < 2, groomed mass 50–600 GeV, and double‑b‑tag efficiency 75 % (light‑mistag ≈ 6 %). Overlap removal ensures mutually exclusive collections of jets, photons and leptons.

The boosted category dramatically improves acceptance for high‑mass Higgs pairs. For non‑resonant VBF production, κ₂V = 0 leads to a pronounced enhancement of the m_HH tail; the boosted selection retains ≈30 % more signal events in the m_HH > 1 TeV region compared with the resolved selection, tightening the 95 % CL interval on κ₂V by roughly 30 % (from κ₂V = 1 ± 0.2 to κ₂V = 1 ± 0.14). For resonant searches, the boosted analysis maintains a signal efficiency above 40 % even for m_X = 5 TeV, whereas the resolved efficiency drops below 10 % beyond 2 TeV. Using a profile‑likelihood fit to the m_HH spectrum, a 5 TeV scalar can be discovered at >3σ significance only in the boosted category, while a 2 TeV scalar reaches ≈3σ already with the combined (boosted + resolved) dataset.

In summary, the paper demonstrates that incorporating a boosted Higgs reconstruction strategy into the HH → b b̄ γγ channel unlocks the high‑energy VBF‑like phase space, substantially enhancing sensitivity to both non‑resonant quartic‑coupling deviations (κ₂V) and heavy scalar resonances up to several TeV. This approach provides a clear path to improve Higgs‑pair measurements in upcoming LHC Run‑3 and HL‑LHC data, offering tighter constraints on the Higgs potential and a broader discovery reach for new physics.


Comments & Academic Discussion

Loading comments...

Leave a Comment