Search for low-mass resonances decaying to $\tau\tau$ and measurement of the $\Upsilon$ $\to$ $\tau\tau$ decay in proton-proton collisions at $\sqrt{s}$ = 13.6 TeV

Using 61.9 fb$^{-1}$ of proton-proton collision data at $\sqrt{s}$ = 13.6 TeV, the CMS collaboration performed an inclusive search for low-mass spin-zero resonances decaying to $\tau\tau$ in the 20–60 GeV range and achieved a 5.8$\sigma$ observation of $\Upsilon \to \tau\tau$ decays, while setting 95% confidence level upper limits on the production cross section times branching fraction for any new resonances.

The Gist

Imagine the Large Hadron Collider (LHC) as the world's most powerful particle smasher. It fires two beams of protons at each other at nearly the speed of light, creating a chaotic explosion of debris. Usually, scientists look for the "big" new particles, like the Higgs boson, which are heavy and rare.

This paper is about a different kind of hunt: looking for lightweight, invisible ghosts that might be hiding in plain sight.

Here is the story of the search, broken down into simple concepts:

1. The Mystery: Looking for "Tiny" New Particles

Scientists know the Standard Model (the rulebook of particle physics) works well, but it doesn't explain everything. Some theories suggest there are other, lighter particles (called $\phi$ bosons) that are much smaller than the Higgs boson.

Think of the Higgs boson as a heavy boulder. These new particles would be like feathers. The problem is, in the noisy, crowded environment of the LHC, feathers are incredibly hard to spot because they get lost in the sea of heavier debris.

2. The Challenge: The "Noise" Problem

When these light particles decay, they turn into tau leptons (a type of heavy electron). But because the original particle is so light, the resulting taus are "lazy"—they don't move very fast or far.

In a normal experiment, the computer system (the trigger) acts like a bouncer at a club. It only lets in events where particles are moving fast and have high energy. Because these "feather" particles are slow, the bouncer usually kicks them out before they can even be recorded. It's like trying to hear a whisper in a rock concert; the volume is turned up so high that the quiet sounds are filtered out.

3. The Solution: The "Scouting" Camera

To solve this, the CMS team used a special technique called Data Scouting.

Imagine the LHC is a busy highway. The standard cameras only take photos of speeding race cars (high-energy events). The Scouting system is like a high-speed, low-resolution security camera that takes pictures of everything, even the slow-moving bicycles.

• The Trick: Instead of saving every single detail of the crash (which takes up too much space), the scouting system saves just the "essence" of the event. This allows them to record four times more events than usual.
• The New Algorithm: They also built a new "flashlight" (a reconstruction algorithm) specifically designed to spot these slow, low-energy taus that the old flashlight missed.

4. The Discovery: Finding the "Upsilon"

Before hunting for the new "feather" particles, the team needed to prove their new flashlight worked. They looked for something they already knew existed: the Upsilon ($\Upsilon$) meson.

Think of the Upsilon as a known, heavy family of particles that also decays into slow taus. It's like testing a new metal detector in a park where you already know there are buried coins.

• The Result: They successfully found the Upsilon mesons decaying into tau pairs.
• The Significance: They found them with a statistical certainty of 5.8 sigma. In the world of physics, this is like flipping a coin and getting heads 5.8 times in a row in a row where getting heads is supposed to be impossible. It's a definitive "Yes, we found it!"

They measured how often this happens (the production cross-section) and found it matched their expectations perfectly. This proved their new "low-energy" tools work in the chaotic environment of a hadron collider.

5. The Search for New Physics: The "Feather" Hunt

Now that they knew their tools worked, they looked for the unknown $\phi$ boson in the mass range between 20 and 60 GeV.

• The Method: They scanned the data for a "bump" in the mass distribution—a sudden spike where more events happened than the background noise predicted.
• The Result: No new particles were found. The data looked exactly like what the Standard Model predicted. There were no mysterious "feathers" hiding in the noise.

6. The Conclusion: Setting the Boundaries

Even though they didn't find the new particle, the paper is a success.

• Firsts: This is the first time anyone has looked for these specific low-mass particles decaying into taus at a hadron collider.
• Limits: They set a "fence" around the possible existence of these particles. They can now say with 95% confidence that if these particles exist, they are rarer than a certain limit (between 40 and 400 pb).
• Legacy: They proved that by using "scouting" data and new algorithms, we can now see parts of the particle world that were previously invisible.

In short: The team built a new, sensitive net to catch slow-moving particles. They tested the net by catching a known fish (the Upsilon), and it worked perfectly. They then cast the net into the deep ocean looking for a mythical fish (the $\phi$ boson). They didn't find the mythical fish, but they proved the net works and mapped out exactly where the fish cannot be hiding.

Technical Summary

Technical Summary: Search for Low-Mass Resonances Decaying to $\tau\tau$ and Measurement of $\Upsilon \to \tau\tau$ in Proton-Proton Collisions at $\sqrt{s} = 13.6$ TeV

Problem and Motivation
While the discovery of the Higgs boson completed the Standard Model (SM), numerous phenomena remain unexplained, prompting theories that predict additional spin-zero bosons ($\phi$). These particles may be lighter than the Higgs and possess enhanced couplings to third-generation fermions, specifically tau leptons. Previous searches for inclusive spin-zero particles decaying to $\tau\tau$ at hadron colliders have been limited to invariant masses above 60 GeV. Below this threshold, the transverse momentum ($p_T$) of the resulting tau leptons is low, leading to substantial backgrounds from Quantum Chromodynamic (QCD) multijet events. Traditional data acquisition methods, which rely on high trigger thresholds to manage data rates, suffer from very low efficiency for these low-mass $\tau\tau$ events. Consequently, the mass range between 20 and 60 GeV has remained unexplored in the context of hadronic tau decays at hadron colliders.

Methodology
This analysis utilizes proton-proton collision data collected by the CMS detector at a center-of-mass energy of $\sqrt{s} = 13.6$ TeV during 2022–2023, corresponding to an integrated luminosity of 61.9 fb$^{-1}$. The final state consists of one tau lepton decaying to a muon and neutrinos ($\tau_\mu$) and the other decaying hadronically ($\tau_h$).

Key methodological innovations include:

• Data Scouting: The analysis employs the CMS scouting data stream, which retains only event data reconstructed by the high-level trigger (HLT). This reduces event size, allowing for significantly higher trigger rates (up to 25 kHz) compared to standard streams.
• Novel Reconstruction Algorithm: A new "scouting HPS" (hadrons-plus-strips) algorithm was developed to reconstruct $\tau_h$ candidates with visible transverse momentum ($p_T^{\text{vis}}$) as low as 5 GeV. Unlike the standard HPS algorithm, which constrains the $\tau$ cone opening angle to $\Delta R < 0.1$, the scouting HPS implements a dynamic maximum opening angle (up to $\Delta R = 0.4$ at 5 GeV) to capture low-momentum decays.
• Machine Learning Discriminants:
  • TAUNET: A "deep sets" neural network model is used to discriminate genuine $\tau_h$ decays from QCD backgrounds based on the constituents of the $\tau_h$ candidate and surrounding particles.
  • HLV$\mu$: A neural network isolation discriminant identifies $\tau_\mu$ candidates with low neighboring activity.
• Trigger Strategy: The search utilizes a logical OR of triggers targeting electromagnetic activity (ECAL deposit $\ge$ 30 GeV), hadronic activity (scalar $p_T$ sum of jets $>$ 360 GeV), or HCAL deposits ($p_T >$ 180 GeV).
• Analysis Regions: Three distinct selection regions are defined:
  1. $\Upsilon$ Region: To observe $\Upsilon(1S, 2S, 3S) \to \tau\tau$ decays, requiring low $p_T^{\text{vis}}$ ($>5$ GeV) and specific isolation criteria.
  2. $Z$ Region: To validate high-$p_T^{\text{vis}}$ reconstruction ($>20$ GeV) using $Z \to \tau\tau$ events.
  3. $\phi$ Search Region: To search for resonances in the 20–60 GeV mass range, applying tighter isolation and kinematic cuts.

Backgrounds are modeled using simulated SM processes (including $Z/\gamma^*$, $W$, $t\bar{t}$, dibosons, QCD, and $\Upsilon$ production) and validated against data. Continuum backgrounds in the search region are estimated using specific functional forms (exponential polynomials or UA2/ATLAS functions) determined by the Akaike Information Criterion.

Key Contributions and Results

1. Observation of $\Upsilon \to \tau\tau$:
   The analysis successfully measures the $\Upsilon(1S, 2S, 3S) \to \tau^+\tau^-$ process in a hadron collider environment. The combined maximum-likelihood fit across prong-only and prong-plus-strips channels yields a significance of 5.8$\sigma$ above the background-only hypothesis.

   • Measured Cross Section: $\sigma = 3.5 \pm 0.7 \text{ (stat)} \pm 0.7 \text{ (syst)}$ nb for the fiducial region $|y_{\text{vis}}| < 1.2$ and $p_T^{\text{vis}} > 15$ GeV.
   • This result validates the efficacy of the novel low-momentum $\tau_h$ reconstruction algorithm in a high-rate data stream.

2. Search for Low-Mass Spin-Zero Resonances ($\phi$):
   An inclusive search for spin-zero bosons decaying to $\tau\tau$ was performed in the mass range of 20 to 60 GeV.

   • Observation: No significant excess above the Standard Model background was observed in the $m_{\text{vis}}(\mu, \tau_h)$ distributions.
   • Limits: Upper limits at 95% confidence level (CL) on the product of the production cross section and branching fraction, $\sigma(pp \to \phi) \times \mathcal{B}(\phi \to \tau\tau)$, were set. These limits range between 40 and 400 pb across the mass range.
   • The results extend the previously analyzed mass region by CMS and ATLAS (which started at 60 GeV) down to 20 GeV for inclusive production with a hadronic tau signature.

Significance
The paper claims that these results constitute the first ever inclusive search for a spin-zero resonance produced via gluon fusion and decaying to $\tau\tau$ with a hadronic tau signature in the 20–60 GeV mass range at a hadron collider. The measurement of the $\Upsilon \to \tau\tau$ decay with 5.8$\sigma$ significance demonstrates that high-rate data scouting streams, combined with novel low-momentum reconstruction algorithms, can achieve unprecedented sensitivity to new physics involving tau leptons in challenging kinematic regimes previously inaccessible due to trigger and reconstruction limitations. The absence of a signal in the search region provides stringent constraints on BSM models predicting light spin-zero bosons with enhanced couplings to third-generation fermions.