Exploring the Higgs potential at ATLAS

Immediately after the Big Bang, all the particles we know about today were massless and moving at the speed of light. About 10^–12 seconds later, the scalar Higgs field spontaneously broke the symmetry of the electroweak force, separating it into the electromagnetic and weak forces, and giving mass to fundamental particles. Without this process, the universe as we know it would not exist.

Since its discovery in 2012, measurements of the Higgs boson – the particle associated with the new field – have refined our understanding of its properties, but it remains unknown how closely the field’s energy potential resembles the predicted Mexican hat shape. Studying the Higgs potential can provide insights into the dynamics of the early universe, and the stability of the vacuum with respect to potential future changes.

The Higgs boson’s self-coupling strength λ governs the cubic and quartic terms in the equation describing the potential. It can be probed using the pair production of Higgs bosons (HH), though this is experimentally challenging as this process is more than 1000 times less likely than the production of a single Higgs boson. This is partly due to destructive interference between the two leading order diagrams in the dominant gluon–gluon fusion production mode.

The ATLAS collaboration recently compiled a series of results targeting HH decays to bbγγ, bbττ, bbbb, bbll plus missing transverse energy (E_T^miss), and multilepton final states. Each analysis uses the full LHC Run 2 data set. A key parameter is the HH signal strength, μ_HH, which divides the measured HH production rate by the Standard Model (SM) prediction. This combination yields the strongest expected constraints to date on μ_HH, and an observed upper limit of 2.9 times the SM prediction (figure 1). The combination also sets the most stringent constraints to date on the strength of the Higgs boson’s self-coupling of –1.2 < κ_λ < 7.2, where κ_λ = λ/λ_SM, its value relative to the SM prediction.

Each analysis contributes in a complementary way to the global picture of HH interactions and faces its own set of unique challenges.

Despite its tiny branching fraction of just 0.26% of all HH decays, HH → bbγγ provides very good sensitivity to μ_HH thanks to the ATLAS detector’s excellent di-photon mass resolution. It also sets the best constraints on λ due to its sensitivity to HH events with low invariant mass.

The HH → bbττ analysis (7.3% of HH decays) exploits state-of-the-art hadronic–tau identification to control the complex mix of electroweak, multijet and top-quark backgrounds. It yields the strongest limits on μ_HH and the second tightest constraints on λ.

HH → bbbb (34%) has good sensitivity to μ_HH thanks to ATLAS’s excellent b-jet identification, but controlling the multijet background presents a formidable challenge, which is tackled in a fully data-driven fashion.

Studying the Higgs potential can provide insights into the dynamics of the early universe

The decays HH → bbWW and HH → bbττ in fully leptonic final states have very similar characteristics and are thus targeted in a single HH → bbll+E_T^miss analysis. Contributions from the bbZZ decay mode, where one Z decays to charged light leptons and the other to neutrinos, are also considered.

Finally, the HH → multilepton analy­sis is designed to catch decay modes where the HH system cannot be fully reconstructed due to ambiguity in how the decay products should be assigned to the two Higgs bosons. The analysis uses nine signal regions with different multiplicities of light charged leptons, hadronic taus and photons. It is complementary to all the exclusive channels discussed above.

For the ongoing LHC Run 3, ATLAS designed new triggers to enhance sensitivity to the hadronic HH → bbττ and HH → bbbb channels. Improved b-jet identification algorithms will increase the efficiency in selecting HH signals and distinguishing them from background processes. With these and other improvements, our prospects have never looked brighter for homing in on the Higgs self-coupling.