⚛️ high-energy experiments

Constraints on the Higgs boson total width from on-shell signal-background interference in the $H\to\gamma\gamma$ decay channel with $pp$ collisions at $\sqrt{s}=13$ TeV with the ATLAS detector

Using 140 fb $^{-1}$ of 13 TeV proton-proton collision data from the ATLAS detector, this study reports a direct, model-independent on-shell measurement of the Higgs boson total width via $H\to\gamma\gamma$ signal-background interference, setting an observed upper limit of 307 MeV at 95% confidence level.

Original authors: ATLAS Collaboration

Published 2026-09-18

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: ATLAS Collaboration

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the heart of the subatomic world, a particle known as the Higgs boson plays a unique role. Discovered in 2012, it is the physical manifestation of a field that gives mass to other fundamental particles. While scientists have spent over a decade measuring how this particle interacts with others, one of its most basic properties remains elusive: its total decay width. In the language of particle physics, this width represents the sum of all the ways the Higgs boson can break apart into other particles. According to the Standard Model, the prevailing theory of physics, the Higgs boson should decay in very specific, predictable ways, resulting in a total width of just 4.1 million electronvolts. However, if the Higgs boson is hiding a secret—perhaps decaying into invisible particles that our detectors cannot see—this total width would be larger. Measuring this width is therefore a direct test of whether the Higgs boson is a perfect match for the Standard Model or a gateway to new, undiscovered physics.

The challenge in measuring this width is that the Higgs boson is incredibly short-lived. It exists for such a fleeting moment that its natural width is far too narrow to be seen directly by even the most powerful detectors. It is like trying to measure the thickness of a human hair by looking at a blurry photograph of a speeding train; the blur of the measurement is thousands of times wider than the object itself. For years, physicists have tried to infer the width indirectly by comparing how often the Higgs boson appears at its normal mass versus how often it appears at much higher, rarer energies. This method works, but it relies on the assumption that the Higgs boson behaves exactly the same way at high energies as it does at normal energies. If that assumption is wrong, the measurement could be misleading.

A team of physicists working with the ATLAS detector at the Large Hadron Collider has now taken a different approach, one that does not require such assumptions. They focused on a specific, subtle effect that occurs when the Higgs boson decays into two photons. In the quantum world, particles can behave like waves. When the Higgs boson is created and immediately decays into two photons, it produces a wave that overlaps with the background waves of photons that are created by other, non-Higgs processes. Just as two sound waves can interfere to create a louder or softer sound, these particle waves interfere with each other. This interference distorts the shape of the mass distribution of the two photons, shifting the peak of the signal slightly to one side. The size of this shift depends directly on the total width of the Higgs boson. By studying this distortion, the researchers can measure the width without needing to guess how the particle behaves at different energies.

Using data from proton collisions recorded between 2015 and 2018, the team analyzed over 140 inverse femtobarns of collision events. This massive dataset allowed them to look at the distribution of photon pairs with extreme precision. They divided the data into two groups based on the momentum of the photon pairs, a technique that helps isolate the interference effect. The researchers then compared the actual data against computer simulations that modeled what the distribution would look like if the Higgs boson had different widths. They found that the data was consistent with the Standard Model prediction, showing no evidence of a hidden, invisible decay mode that would have broadened the width.

The study set a new limit on the total width of the Higgs boson. The researchers concluded that the width is less than 307 million electronvolts at a 95 percent confidence level. This means that if the width were any larger, the distortion in the photon data would have been too obvious to miss. The expected limit, based on what the experiment should have been able to see if the data were perfect, was 258 million electronvolts. While these numbers are still about 75 times larger than the Standard Model prediction of 4.1 million electronvolts, the measurement is significant because it is a direct, on-shell observation that does not rely on theoretical assumptions about the particle's couplings. It provides a complementary way to check the Higgs boson's properties, independent of the methods used in previous studies.

The precision of this result is currently limited by uncertainties in the theoretical calculations of the interference effect and the exact mass of the Higgs boson. The researchers noted that the largest source of uncertainty comes from the theory side, specifically the calculation of how the interference behaves. This suggests that future improvements in theoretical physics will be just as important as collecting more data. Despite the current limits, the method itself is a triumph. It proves that scientists can use the subtle interference of quantum waves to measure a property that is otherwise hidden by the limits of experimental resolution. This opens a new path for testing the Standard Model, offering a way to search for new physics that does not depend on the same assumptions as other techniques.

The findings confirm that, within the current limits of measurement, the Higgs boson behaves as the Standard Model predicts. There is no sign of it decaying into invisible particles or other exotic states that would have widened its total decay width. While the measurement is not yet precise enough to rule out all possible new physics, it establishes a robust, assumption-free constraint. As theoretical calculations improve and more data is collected, this technique will become even more powerful, potentially revealing the first cracks in the Standard Model or confirming its resilience with unprecedented clarity. For now, the Higgs boson remains a mystery in the sense that its width is still not measured with the precision of a ruler, but the shadow it casts on the quantum landscape has been mapped with a new kind of light.

Technical Summary: Constraints on the Higgs boson total width from on-shell signal–background interference in the $H \to \gamma\gamma$ decay channel

Problem Statement
The total decay width of the Higgs boson, $\Gamma_H$ , is a fundamental parameter representing the sum of all its partial decay widths. In the Standard Model (SM), $\Gamma_H$ is predicted to be approximately 4.1 MeV at a Higgs mass of 125 GeV. Any decay into particles beyond the SM (BSM), particularly invisible or undetected particles, would increase $\Gamma_H$ above this prediction. Directly measuring $\Gamma_H$ is challenging because the natural width is two to three orders of magnitude smaller than the experimental mass resolution ( $\mathcal{O}(1\text{--}3)$ GeV) in the diphoton ( $H \to \gamma\gamma$ ) and four-lepton channels. Consequently, the width cannot be resolved directly from the reconstructed lineshape.

Previous indirect constraints have relied on comparing on-shell (near the mass peak) and off-shell (far above the peak) production rates in $H \to ZZ$ and $H \to WW$ channels. However, these indirect measurements rely on the assumption that Higgs boson couplings to gluons and vector bosons are identical in on-shell and off-shell regimes. This paper addresses the need for a direct, on-shell measurement that does not depend on such coupling assumptions, thereby providing a complementary probe sensitive to potential BSM decays.

Methodology
This analysis utilizes the full Run 2 dataset of proton–proton collisions at $\sqrt{s} = 13$ TeV collected by the ATLAS detector, corresponding to an integrated luminosity of $140.1 \pm 1.2 \text{ fb}^{-1}$ . The measurement exploits the quantum mechanical interference between the resonant $gg \to H \to \gamma\gamma$ amplitude and the non-resonant continuum $gg \to \gamma\gamma$ background.

Interference Mechanism: The interference term distorts the diphoton invariant mass ( $m_{\gamma\gamma}$ ) distribution. It introduces an antisymmetric component that shifts the apparent peak position to lower masses (approximately $-70$ MeV at the SM width) and a symmetric component that reduces the overall signal yield (approximately $-1.7\%$ ). Both effects scale with $\sqrt{\Gamma_H}$ .
Event Selection and Categorization: Events are selected with two photon candidates satisfying tight identification and isolation criteria. The dataset is divided into two mutually exclusive categories based on the diphoton transverse momentum ( $p_T^{\gamma\gamma}$ $p_{T}^{γ γ}$ ):
- Low- $p_T^{\gamma\gamma}$ : $p_T^{\gamma\gamma} < 30$ GeV (where the interference contribution is largest).
- High- $p_T^{\gamma\gamma}$ : $p_T^{\gamma\gamma} \ge 30$ GeV.
  A finer categorization is avoided due to limitations in the theoretical modeling of the interference at higher $p_T$ and the LO accuracy of the $p_T$ spectrum in the simulation.
Statistical Model: A simultaneous maximum-likelihood fit is performed on the $m_{\gamma\gamma}$ $m_{γ γ}$ distributions in both categories.
- Signal: Modeled by a Double-Sided Crystal Ball (DSCB) function.
- Interference: Modeled using a dedicated analytic procedure to handle negative weights, constructed from asymmetric DSCB functions fitted to positive and negative weight subsets of simulated events.
- Background: Described by an exponentiated second-order polynomial, with parameters determined from data.
Systematic Uncertainties: Key uncertainties include the external constraint on the Higgs boson mass ( $m_H$ ) from the $H \to ZZ^* \to 4\ell$ channel, photon energy scale and resolution, and theoretical uncertainties in the normalization of the $gg$-interference (specifically the $K$ -factor). The signal strength parameter $\mu$ is profiled to ensure the measurement is independent of Higgs coupling assumptions.

Key Results
The analysis yields the following constraints on the Higgs boson total width at 95% confidence level (CL):

Observed Limit: $\Gamma_H < 307$ MeV.
Expected Limit: $\Gamma_H < 258$ MeV.
Comparison to SM: These limits correspond to approximately 75 (observed) and 63 (expected) times the SM prediction of 4.1 MeV.
Best-Fit Value: $\hat{\Gamma}_H = 13^{+75}_{-13}$ MeV, which is consistent with the SM prediction.
Statistical Sensitivity: When considering only statistical uncertainties, the observed (expected) upper limit improves to 105 (87) MeV, or roughly 26 (21) times the SM width.

The measurement is dominated by systematic uncertainties. The largest single contribution arises from the theoretical uncertainty in the normalization of the $gg$-interference ( $K_I$ ), contributing 49 MeV to the 68% CL uncertainty. The external constraint on the Higgs mass is the second-largest contributor (43 MeV).

Significance and Claims
The paper claims that this measurement provides a novel, direct, on-shell constraint on the Higgs boson total width. Its primary significance lies in its independence from theoretical assumptions regarding the equality of on-shell and off-shell Higgs couplings, which are required for indirect off-shell measurements. By relying solely on the interference-induced distortion of the $m_{\gamma\gamma}$ lineshape, this approach offers a complementary probe to off-shell determinations.

While the sensitivity is approximately 30 times weaker than the most precise indirect off-shell determinations, the paper emphasizes that the two approaches have different sensitivities to new physics. The direct on-shell method is particularly sensitive to decays into new particles that would increase the total width without necessarily altering the couplings assumed in off-shell extrapolations. The author notes that future improvements in the precision of this measurement will depend on higher-order theoretical calculations of the interference, specifically matching to parton showers, and reduced experimental uncertainties.

More like this