Search for a new heavy scalar resonance decaying to a pair of Z bosons in the four-lepton final state in proton-proton collisions at $\sqrt{s}$ = 13 TeV

The CMS collaboration analyzed 138 fb$^{-1}$ of proton-proton collision data at 13 TeV to search for a heavy scalar resonance decaying to two Z bosons in the four-lepton final state, finding no significant excess over the Standard Model background and setting 95% confidence level upper limits on the production cross section across a mass range of 130 GeV to 3 TeV.

The Gist

The Big Picture: Hunting for a Ghost in the Machine

Imagine the Large Hadron Collider (LHC) at CERN as the world's most powerful particle-smashing machine. It takes two beams of protons (tiny subatomic particles) and crashes them together at nearly the speed of light. When they collide, they create a chaotic explosion of energy that briefly turns into new particles.

For years, scientists have been looking for a specific "ghost": a new, heavy particle called a scalar resonance (let's call it "Particle X"). They suspect this particle might exist because our current rulebook for physics (the Standard Model) has some gaps, like not explaining gravity or dark matter. If "Particle X" exists, it would be a heavy cousin of the famous Higgs boson (discovered in 2012).

The Detective Work: How They Looked

The CMS team (the detectives) didn't just look for "Particle X" directly. Instead, they looked for its "footprints." They hypothesized that if "Particle X" exists, it would instantly fall apart into two Z bosons (another type of particle), which would then immediately fall apart into four leptons (electrons or muons).

Think of it like this: You are looking for a rare, invisible bird. You can't see the bird, but you know that if it lands, it will drop four specific, glowing feathers. Your job is to scan the forest for those four glowing feathers.

The Search Parameters:

• The Forest: They scanned a massive range of "masses" (how heavy the particle would be), from 130 GeV (a bit heavier than the Higgs) all the way up to 3,000 GeV (very heavy).
• The Data: They analyzed data from 2016 to 2018, which is like having a library containing 138 "petabytes" of collision records (138 inverse femtobarns).
• The Scenarios: They checked two ways the particle could be made:
  1. Gluon Fusion (ggF): Like two cars crashing head-on to create a new object.
  2. Vector Boson Fusion (VBF): Like two cars skimming past each other and exchanging a part to create a new object.

The Tools: Sorting the Noise

The problem is that the "forest" is full of other things that look like four glowing feathers. The background noise is huge.

• The Background: Most of the time, four leptons appear just by chance from other common processes (like two Z bosons being made naturally without a new heavy particle). This is the "static" on a radio.
• The Filter: To find the signal, the scientists used a sophisticated filter called a kinematic discriminant. Imagine you are trying to find a specific song in a noisy room. You don't just listen for any sound; you look for a specific rhythm and pitch. The scientists used math to calculate how "likely" a set of four particles is to be the new heavy particle versus just random background noise.

They also looked at the "shape" of the data. If "Particle X" exists, it should show up as a bump or a peak in the data graph, rising above the flat line of background noise.

The Results: The Silence of the Data

After running their complex statistical models and checking every possible mass and width (how "fuzzy" or spread out the particle might be), here is what they found:

1. No New Particle: They did not find a significant bump. The data looked almost exactly like what the Standard Model predicts (just the background noise).
2. A Tiny Fluke: There was one spot around 138 GeV where the data looked slightly higher than expected. It was a "blip" with a significance of about 3 standard deviations. However, when they accounted for the fact that they looked at many different spots (the "look-elsewhere effect"), this blip turned out to be just a random statistical fluctuation. It's like flipping a coin 1,000 times and getting a streak of heads once; it's surprising, but not proof of a magic coin.
3. Setting Limits: Even though they didn't find the particle, they didn't come up empty-handed. They set exclusion limits.
   • The Analogy: Imagine you are searching for a specific type of fish in a lake. You don't find it. But you can say, "If this fish exists, it must be smaller than 1 inch or rarer than 1 in a million."
   • The Paper's Claim: They can now say with 95% confidence that if this heavy particle exists, it cannot be produced more often than a certain rate. In the low-mass region, they ruled out production rates above 0.05–0.1 picobarns; in the high-mass region, they ruled out rates above 0.005 picobarns.

The Conclusion

The paper concludes that, based on the 138 fb⁻¹ of data collected, there is no evidence for a new heavy scalar resonance decaying into two Z bosons in the mass range of 130 GeV to 3 TeV.

The "ghost" remains invisible. The Standard Model continues to hold up, and the search for new physics must continue with even more data or different strategies. The scientists have effectively drawn a map of where the particle isn't, narrowing down the search for future experiments.

Technical Summary

Technical Summary: Search for a New Heavy Scalar Resonance Decaying to a Pair of Z Bosons in the Four-Lepton Final State

Problem and Motivation
While the 2012 discovery of the Higgs boson ($H$) with a mass near 125 GeV validated the Standard Model (SM), the SM remains incomplete, lacking explanations for gravity, dark matter, and matter-antimatter asymmetry. Many Beyond the Standard Model (BSM) theories predict the existence of additional scalar resonances, such as extra Higgs bosons in Two-Higgs-Doublet Models or radions in warped extra dimension models. This paper presents a search for a new heavy scalar resonance ($X$) decaying into a pair of $Z$ bosons ($X \to ZZ$), where each $Z$ boson subsequently decays into a pair of electrons or muons ($ZZ \to 4\ell$). The search covers a wide mass range ($130 \text{ GeV} \le m_X \le 3 \text{ TeV}$) and investigates both narrow-width and broad-width scenarios, including interference effects between the signal, the SM Higgs boson, and continuum backgrounds.

Methodology
The analysis utilizes proton-proton collision data collected by the CMS experiment at a center-of-mass energy of $\sqrt{s} = 13 \text{ TeV}$, corresponding to an integrated luminosity of $138 \text{ fb}^{-1}$ from the 2016–2018 data-taking periods.

• Event Selection: Events are selected requiring four isolated leptons (electrons or muons) originating from the primary vertex. Lepton candidates are reconstructed using the Particle-Flow algorithm. $Z$ boson candidates are formed from opposite-charge, same-flavor lepton pairs with invariant masses between 12 and 120 GeV. A $ZZ$ candidate is constructed from two such $Z$ candidates, with the one closest to the nominal $Z$ mass defined as $Z_1$.
• Categorization: To enhance sensitivity to Vector Boson Fusion (VBF) production, events are categorized into "VBF-tagged" and "untagged" (primarily Gluon Fusion, ggF) regions. The VBF category requires at least two jets with specific kinematic properties and a discriminant ($D^{VBF}_{2jet}$) based on matrix element likelihoods to distinguish VBF topology from ggF.
• Signal and Background Modeling:
  • Signal: Signal samples are generated at Next-to-Leading Order (NLO) for ggF and VBF processes. A parametric approach models the signal lineshape, efficiency, and resolution as functions of the resonance mass ($m_X$) and width ($\Gamma_X$). For broad-width scenarios, interference between the signal, the 125 GeV Higgs boson, and non-resonant backgrounds is explicitly modeled.
  • Backgrounds: Irreducible backgrounds ($qq \to ZZ$, $gg \to ZZ$, VBF $ZZ$, triboson, and top-associated production) are estimated using Monte Carlo (MC) simulations corrected with higher-order K-factors. Reducible backgrounds ($Z + \text{jets}$), where jets are misidentified as leptons, are estimated from data using a "misidentification rate" method in control regions.
• Statistical Analysis: A two-dimensional likelihood fit is performed using the reconstructed four-lepton mass ($m^{\text{reco}}_{4\ell}$) and a kinematic discriminant ($D^{\text{kin}}_{\text{bkg}}$) to distinguish signal from background. Systematic uncertainties (experimental and theoretical) are treated as nuisance parameters. Upper limits on the production cross section times branching fraction ($\sigma \times \mathcal{B}$) are set at the 95% confidence level (CL) using the $CL_s$ criterion.

Key Contributions

• Broad-Width Interference: The analysis incorporates a sophisticated treatment of interference effects for broad-width resonances, considering the interference between the new resonance, the SM Higgs boson, and continuum backgrounds in both ggF and VBF production modes.
• Parametric Signal Modeling: A flexible parametric model is employed to describe signal shapes across a continuous range of masses and widths, avoiding the need to simulate every specific hypothesis.
• Comprehensive Mass and Width Scan: The search extends previous results by covering a mass range up to 3 TeV and testing various width assumptions, including widths up to 30% of the resonance mass ($\Gamma_X/m_X = 0.3$).

Results

• Observation: No significant excess of events over the Standard Model background expectation is observed in the examined phase space. The largest local fluctuation occurs at $m_X \approx 137.8 \text{ GeV}$, with a local significance of 3.0 standard deviations ($\sigma$). However, after accounting for the look-elsewhere effect, the global significance is reduced to 1.8 $\sigma$.
• Exclusion Limits: Upper limits on $\sigma(pp \to X \to ZZ)$ are established. In the low-mass region, limits range from 0.05 to 0.1 pb, while in the high-mass region, they reach down to 0.005 pb.
  • For the narrow-width approximation (NWA), limits are generally stronger for the VBF production mechanism compared to ggF due to the higher signal-to-background ratio in the VBF-tagged category.
  • For broad-width scenarios, the limits vary with the assumed width. In some cases (e.g., $\Gamma_X = 100 \text{ GeV}$), the observed limits are more stringent than the median expected limits because the observed data shows a slight deficit relative to the background model, particularly in regions where the signal lineshape is broad.

Significance
The paper concludes that within the sensitivity of the current dataset and analysis strategy, there is no evidence for a new heavy scalar resonance decaying to $ZZ$ in the four-lepton final state. The results provide updated exclusion limits that constrain BSM models predicting additional scalar particles, particularly those with broad widths or significant interference effects. The analysis demonstrates the capability of the CMS experiment to probe complex signal scenarios involving interference and broad resonances, setting the stage for future searches with increased luminosity.