Measurement of the Higgs boson total decay width using the H $\to$ WW $\to$ e$\nu\mu\nu$ decay channel in proton-proton collisions at $\sqrt{s}$ = 13 TeV

Using 138 fb$^{-1}$ of 13 TeV proton-proton collision data, the CMS experiment measured the Higgs boson total decay width to be 3.9$^{+2.7}_{-2.2}$ MeV via the H $\to$ WW $\to$ e$\nu\mu\nu$ channel, achieving a threefold improvement in uncertainty over previous results while confirming consistency with the Standard Model.

The Gist

The Big Picture: Weighing a Ghost

Imagine the Higgs boson as a very shy, incredibly fast ghost that appears for a split second in a massive particle collider (the Large Hadron Collider, or LHC) and then vanishes. Scientists want to know exactly how "heavy" this ghost is in terms of its energy, which physicists call its decay width.

Think of the decay width like the loudness of a bell.

• A bell that rings for a long time (a wide decay width) is loud and easy to hear.
• A bell that rings for a tiny fraction of a second (a narrow decay width) is a quiet "ping" that is very hard to catch.

The Standard Model (the rulebook of physics) predicts this Higgs ghost should be a very quiet "ping"—so quiet that our detectors are too fuzzy to hear it directly. It's like trying to measure the exact weight of a feather using a bathroom scale; the scale isn't sensitive enough.

The Trick: Listening to the Echo

Since they can't weigh the ghost directly, the CMS team at CERN used a clever trick. They looked at two different ways the ghost appears:

1. The "On-Shell" Ghost (The Main Event): This is the ghost appearing at its normal, expected energy (125 GeV). It's like the ghost showing up at a party exactly when invited.
2. The "Off-Shell" Ghost (The Rare Guest): This is the ghost appearing at much higher energies (over 160 GeV). It's like the ghost crashing the party at a much higher energy level. This happens very rarely.

The Analogy:
Imagine you are trying to figure out how fast a car engine runs, but you can't look at the engine. Instead, you look at how much fuel the car uses at a slow speed (on-shell) versus how much it uses when it's revving its engine at a redline (off-shell).

The paper explains that the ratio between these two "fuel usages" tells you the secret speed of the engine (the decay width). If the ghost is very "narrow" (quiet), the high-energy "off-shell" version is much harder to produce compared to the normal one. By measuring how often the high-energy version appears compared to the normal one, they can calculate the width.

The Experiment: The Great Filter

The scientists looked at 138 "femtobarns" of data. To put that in perspective, that's like watching 138 trillion proton collisions happen in the LHC between 2016 and 2018.

They were looking for a specific signal: A Higgs boson turning into two W particles, which then turn into an electron and a muon (and some invisible neutrinos).

• The Challenge: The background noise is huge. It's like trying to hear a whisper in a stadium full of cheering fans. The "fans" are other particle collisions that look similar but aren't the Higgs.
• The Solution: They used a Deep Neural Network (DNN). Think of this as a super-smart AI referee. It looked at every single collision and asked: "Does this look like the Higgs ghost, or is it just background noise?" It sorted the events into different categories based on how many other particles (jets) were flying around.

The Results: A Perfect Match

After sorting through the noise and using their AI referee, the team found:

• The Off-Shell Signal: They measured how often the high-energy ghost appeared. The result was 1.2 (with some uncertainty). In the rulebook, a value of 1.0 is perfect. So, 1.2 is very close to what was expected.
• The Total Width: Using the ratio of the high-energy ghost to the normal ghost, they calculated the total decay width.
  • Their Result: 3.9 MeV (plus or minus a bit).
  • The Prediction: 4.1 MeV.

The Verdict: The measurement is a perfect match for the Standard Model. The "ghost" is exactly as quiet and elusive as the rulebook said it would be.

Why This Matters

This isn't just a "we found it" paper; it's a "we measured it precisely" paper.

• Improvement: This result is 3 times more precise than the previous attempt by the same team using older data.
• New Channel: This is the first time the CMS team has measured this specific width using the H → WW (Higgs to W particles) channel at the high energy of 13 TeV. Previously, they had to use a different channel (H → ZZ).
• Consistency: The fact that the measurement matches the prediction so well means there are no "weird" new physics hiding in the shadows right now. The Higgs boson is behaving exactly as the Standard Model predicts.

Summary

The CMS team acted like detectives trying to weigh a ghost. They couldn't weigh it directly, so they compared how often the ghost appeared in a "normal" state versus a "high-energy" state. Using a massive amount of data and a smart AI to filter out the noise, they calculated the ghost's "width" to be 3.9 MeV. This matches the theoretical prediction of 4.1 MeV almost perfectly, confirming that our current understanding of the universe's building blocks is still holding up strong.

Technical Summary

Technical Summary: Measurement of the Higgs boson total decay width using the H →WW →eνµν decay channel

Problem and Motivation
The total decay width of the Higgs boson ($\Gamma_H$) is a fundamental parameter of the Standard Model (SM). For a SM Higgs boson with a mass of 125 GeV, the theoretical prediction is 4.1 MeV. However, this value is significantly smaller than the invariant mass resolution of the Large Hadron Collider (LHC) detectors, rendering direct measurement of the decay width infeasible. Consequently, $\Gamma_H$ must be constrained indirectly by comparing the production rates of the Higgs boson in "on-shell" (near the resonance mass) and "off-shell" (at high invariant masses) regimes.

Previous measurements by the CMS Collaboration in the $H \to ZZ$ channel and the $H \to WW \to 2\ell 2\nu$ channel (using 7 and 8 TeV data) provided constraints, but with limited precision. This paper addresses the need for a more precise determination of $\Gamma_H$ using the full Run 2 dataset (2016–2018) at $\sqrt{s} = 13$ TeV, specifically targeting the $H \to WW \to e\nu\mu\nu$ decay channel. The study aims to improve upon the previous CMS limit in this channel ($\Gamma_H < 26$ MeV) by leveraging increased center-of-mass energy, higher integrated luminosity, and refined theoretical and experimental modeling.

Methodology
The analysis utilizes proton-proton collision data corresponding to an integrated luminosity of 138 fb$^{-1}$ collected by the CMS experiment. The measurement relies on the ratio of off-shell to on-shell production cross-sections, which is proportional to the total decay width ($\Gamma_H \propto \mu_{\text{off-shell}} / \mu_{\text{on-shell}}$).

• Event Selection and Reconstruction: Events are selected requiring an opposite-sign electron-muon pair ($e\mu$) to suppress Drell-Yan backgrounds. Kinematic requirements include missing transverse momentum ($p^{\text{miss}}_T > 20$ GeV) to account for neutrinos, a dilepton invariant mass $m_{e\mu} > 20$ GeV, and a reconstructed transverse mass of the Higgs boson $m^H_T > 60$ GeV. Jets are clustered using the anti-$k_T$ algorithm, and events with $b$-tagged jets are vetoed to reduce top-quark backgrounds.
• Categorization: Events are categorized based on the number of jets (0, 1, and $\ge 2$) to distinguish between gluon-gluon fusion (ggF) and vector boson fusion (VBF) production modes.
• Machine Learning Discrimination: A multiclass deep neural network (DNN) is employed to separate signal from background. The DNN is trained to classify events into four classes for the 0- and 1-jet categories (off-shell ggF, on-shell ggF, top quark, nonresonant WW) and six classes for the $\ge 2$-jet category (adding off-shell/on-shell VBF). Input variables include kinematic properties of the dilepton system, dijet system, missing transverse momentum, and a proxy invariant mass of the WW system.
• Signal and Background Modeling:
  • Signal: Off-shell contributions are defined by a generator-level requirement of $m_{WW} > 160$ GeV. Simulations for ggF and VBF use POWHEG and JHUGEN (for VBF) or MCFM (for ggF), incorporating complex pole schemes and interference effects between signal and continuum backgrounds.
  • Backgrounds: Major backgrounds include nonresonant WW production, top quark ($t\bar{t}$), and Drell-Yan processes. Normalizations for nonresonant WW and $t\bar{t}$ are constrained simultaneously in signal regions (SR) and control regions (CR) using data. Nonprompt lepton backgrounds are estimated using a data-driven "matrix method" approach.
• Statistical Analysis: A profile likelihood fit is performed using the COMBINE framework. The likelihood function incorporates the DNN output distributions for off-shell signal regions and on-shell control regions. Systematic uncertainties (experimental and theoretical) are treated as nuisance parameters with Gaussian constraints.

Key Contributions

• First 13 TeV Measurement in $H \to WW$: This work presents the first CMS determination of the Higgs boson decay width using the $H \to WW \to e\nu\mu\nu$ channel at $\sqrt{s} = 13$ TeV.
• Improved Precision: The analysis achieves a factor of 3 improvement in uncertainty over the previous CMS result in this specific decay channel (derived from 7 and 8 TeV data).
• Advanced Modeling: The study implements sophisticated modeling of the interference between the off-shell Higgs signal and the continuum $WW$ background, which is critical for accurate extraction of the signal strength in the high-mass tail.
• DNN-Based Categorization: The application of deep neural networks allows for efficient discrimination between signal and background across different jet multiplicities, optimizing sensitivity in both ggF and VBF production modes.

Results

• Off-Shell Signal Strength: The off-shell signal strength is measured as $\mu_{\text{off-shell}} = 1.2^{+0.8}_{-0.7}$. This value is consistent with the SM prediction of unity.
• Total Decay Width: The Higgs boson total decay width is determined to be $\Gamma_H = 3.9^{+2.7}_{-2.2}$ MeV.
• Consistency with SM: The measured width is in agreement with the Standard Model prediction of 4.1 MeV.
• Exclusion Limits: The scenario of no off-shell Higgs boson production is excluded at 3.4 standard deviations (with an expected exclusion of 1.3 standard deviations). The 95% confidence level upper limit on the width is $\Gamma_H < 22.00$ MeV.
• Uncertainty Breakdown: The total uncertainty in the ratio $r = \mu_{\text{off-shell}}/\mu_{\text{on-shell}}$ is dominated by statistical uncertainties (56%), followed by theoretical uncertainties (27%).

Significance
The paper claims that this result represents a significant advancement in constraining the Higgs boson properties. By complementing previous measurements in the $H \to ZZ$ channel, this analysis provides a robust, independent constraint on $\Gamma_H$ using the $WW$ channel. The result confirms the SM prediction within uncertainties and demonstrates that deviations from the SM, which could indicate anomalous couplings or new physics, are currently constrained within the measured range. The improvement in precision over previous limits in this channel highlights the effectiveness of utilizing the full Run 2 dataset and advanced analysis techniques like deep learning and precise interference modeling.