⚛️ high-energy experiments

Search for heavy neutral leptons in decays of W bosons produced in 13 TeV pp collisions using prompt signatures in the ATLAS detector

Using 140 fb $^{-1}$ of 13 TeV proton-proton collision data from the ATLAS detector, this study searches for heavy neutral leptons in W boson decays via prompt same-charge dilepton or trilepton signatures, finding no significant excess over Standard Model backgrounds and thereby setting stringent upper limits on their mixing with electron and muon neutrinos across a mass range of 8–65 GeV.

Original authors: ATLAS Collaboration

Published 2026-07-24

📖 3 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

Imagine the universe as a giant, bustling party where every particle has a role to play. For decades, physicists have been trying to figure out why the "neutrinos"—the ghostly, invisible guests that zip through everything without stopping—have any mass at all. It's a bit like discovering that a ghost has weight; it shouldn't be possible according to the old rulebook. To solve this mystery, scientists proposed a new character: the Heavy Neutral Lepton (HNL). Think of this HNL as a "big brother" to the neutrino, a heavy, shy relative that might explain why the tiny neutrinos have mass, why the universe is made of matter instead of just antimatter, and even what dark matter is made of. The big question is: Do these heavy cousins actually exist, or are they just a clever idea on paper?

This paper is the story of a massive search party sent out to find these elusive HNLs. Using the ATLAS detector—a gigantic, multi-layered digital camera the size of a cathedral built inside the Large Hadron Collider (LHC)—scientists smashed protons together at incredible speeds. They were looking for a very specific "smoking gun": a heavy neutrino that pops into existence, lives for a split second, and then decays into three charged particles (like electrons or muons) that all show up at the same time. Because these heavy neutrinos are their own antiparticles (a weird quantum trick), they can produce pairs of particles with the same electric charge, a signature that is almost impossible for the standard "background noise" of the universe to fake.

The team analyzed a mountain of data equivalent to 140 inverse femtobarns of collisions (a fancy way of saying they watched 140 trillion trillion proton collisions). They set up a series of digital traps, or "signal regions," designed to catch these three-particle events while filtering out the billions of boring, standard events. They looked for the heavy neutrinos to appear in a mass range between 8 and 65 GeV (a unit of energy that tells us how heavy the particle is).

The result? The traps were mostly empty, though there was a tiny, intriguing blip. The scientists found no heavy neutral leptons hiding in the data, but they did observe a small fluctuation: the number of events they saw was slightly higher than expected, showing a 1.7 sigma excess. While this is a hint of something interesting, it isn't strong enough to claim a discovery; it's consistent with random statistical noise rather than a new particle. The paper explicitly rules out the existence of these heavy neutrinos if they mix with electron or muon neutrinos with a strength greater than certain tiny numbers. Specifically, for heavy neutrinos weighing between 15 and 30 GeV, the mixing strength must be less than 1.1 × 10⁻⁵ for electrons and 5 × 10⁻⁶ for muons. In the full range of 8 to 65 GeV, the mixing must be even weaker, below 8 × 10⁻⁵ and 5.0 × 10⁻⁵ respectively.

So, while the heavy neutrinos remain a theoretical possibility, this search has pushed the "Where are you?" sign much further out. If they do exist, they are either much heavier than this search could catch, or they are so shy and weakly connected to the rest of the universe that they are even harder to find than we thought. The search continues, but the map of the universe just got a little more precise.

Technical Summary: Search for Heavy Neutral Leptons in $W$ Boson Decays using Prompt Signatures in the ATLAS Detector

Problem and Motivation
The discovery of neutrino flavor oscillations confirms that neutrinos possess mass, a phenomenon not explained by the Standard Model (SM). This observation suggests the existence of right-handed heavy Majorana neutrinos, often referred to as Heavy Neutral Leptons (HNLs), which could arise from the Type-1 Seesaw mechanism. HNLs are theoretically motivated as potential explanations for the smallness of neutrino masses, the baryon asymmetry of the universe via leptogenesis, and as dark matter candidates. While previous searches have constrained HNLs in various mass ranges, including long-lived particle signatures, this paper addresses the search for prompt HNL decays produced in $W$ boson decays using the full Run 2 dataset of the Large Hadron Collider (LHC).

Methodology
The analysis utilizes $140 \text{ fb}^{-1}$ of proton-proton collision data at a center-of-mass energy of $\sqrt{s} = 13 \text{ TeV}$ collected by the ATLAS detector. The search targets the production of a single HNL ( $N$ ) via mixing with SM neutrinos in the decay of an on-shell $W$ boson ( $W \to \ell^\pm N$ ), followed by the prompt decay of the HNL into a charged lepton and a neutrino ( $N \to \ell^\mp \nu$ ). Due to the Majorana nature of the HNL, this process results in a final state with three charged leptons and a neutrino, where two leptons share the same flavor and electric charge (Same-Flavor Same-Charge, SFSC), and the third lepton has a different flavor and opposite charge.

Event Selection: The analysis defines signal regions (SRs) requiring either three prompt signal leptons or two signal leptons plus one baseline lepton. To suppress dominant SM backgrounds (such as $WZ$, $ZZ$, and $VVV$ production), events with three same-flavor leptons are vetoed, and kinematic requirements are applied to variables including transverse mass ( $m_T$ ), effective mass ( $m_{\text{eff}}$ ), and the invariant mass of the SFSC lepton pair ( $m_{\ell\ell}^{\text{SC}}$ ). A $b$ -jet veto is applied to reduce top-quark backgrounds.
Background Estimation: Backgrounds are categorized into three groups:
1. Prompt SM processes: Estimated using Monte Carlo (MC) simulations normalized to theoretical cross-sections (dominated by $WZ$).
2. Charge-flip electrons: Estimated using data-driven methods based on the probability of electron charge misidentification, though found to be negligible in the signal regions.
3. Fake and non-prompt leptons: Estimated using the data-driven matrix method, which leverages the differing efficiencies of loose and tight lepton identification criteria. This category is the dominant background source.
Systematic Uncertainties: Uncertainties include integrated luminosity (0.83%), lepton identification and reconstruction efficiencies, jet energy scale and resolution, and modeling uncertainties for both signal and background processes. The dominant uncertainties arise from the statistical and systematic errors in the fake/non-prompt lepton background estimation.

Key Contributions

Extended Mass Range: This analysis covers a heavy neutral lepton mass range of $8 $–$ 65 \text{ GeV}$, extending beyond previous ATLAS prompt decay searches performed on partial Run 2 datasets.
Comprehensive Signal Regions: The study defines and optimizes multiple signal regions for both $e^\pm e^\pm \mu^\mp$ and $\mu^\pm \mu^\pm e^\mp$ final states, as well as two-lepton signal regions to recover events where the third lepton fails strict signal criteria but passes baseline requirements.
Robust Background Validation: The background estimation methods are validated in control regions and validation regions (VRs) defined with relaxed kinematic cuts, showing good agreement between observed data and predicted SM backgrounds within $2\sigma$ .

Results
No significant excess of events over the expected Standard Model background was observed. The observed event yields in all signal regions are consistent with background predictions, with the largest local deviation being $1.7$ standard deviations in the SRE3 signal region.

Consequently, the analysis sets model-dependent exclusion limits on the mixing parameters $|U_e|^2$ and $|U_\mu|^2$ as a function of the HNL mass ( $m_{\text{HNL}}$ ):

Full Mass Range ($8 $–$ 65 \text{ GeV}$): The analysis excludes $|U_e|^2$ values above $8 \times 10^{-5}$ and $|U_\mu|^2$ values above $5.0 \times 10^{-5}$ .
Strongest Constraints: The most stringent limits are achieved in the mass range of $15 $–$ 30 \text{ GeV}$, excluding $|U_e|^2 < 1.1 \times 10^{-5}$ and $|U_\mu|^2 < 5 \times 10^{-6}$ .
Wider Ranges: Upper limits of $1.8 \times 10^{-5}$ and $7.0 \times 10^{-6}$ are set for $|U_e|^2$ and $|U_\mu|^2$ respectively over broader mass intervals ($15 $–$ 50 \text{ GeV}$ and $15 $–$ 40 \text{ GeV}$).

Significance
The paper claims that these results supersede previous ATLAS searches based on partial Run 2 data and complement existing searches for long-lived HNLs and other HNL production mechanisms (such as vector-boson fusion). By focusing on prompt signatures in the $8 $–$ 65 \text{ GeV}$ mass range, this work provides the most stringent constraints to date on the mixing of heavy neutral leptons with electron and muon neutrinos in this specific kinematic regime, effectively probing the parameter space where HNLs could explain neutrino mass origins and matter-antimatter asymmetry.

More like this