Photon production in top quark events at ATLAS and CMS

Imagine the Large Hadron Collider (LHC) as the world's most powerful particle accelerator, essentially a giant cosmic collision course where scientists smash protons together to see what happens. In this chaotic environment, the top quark is the heavyweight champion—it's the heaviest known elementary particle, like a massive boulder in a stream of pebbles.

This paper is a report from two giant teams of scientists, ATLAS and CMS, who are like two different detective agencies working at the same crime scene. They are investigating a very rare and specific event: what happens when a top quark (or a pair of them) is created alongside a photon (a particle of light).

Here is a breakdown of their findings using simple analogies:

1. The Rare Event: Finding a "Spark" in the Storm

Usually, when top quarks are made, they come in pairs and don't carry a photon. Finding a top quark with a photon attached is like finding a specific, rare coin in a massive pile of sand. It's much harder to find than just finding the sand (standard top quark pairs), but because the LHC has been running for so long, they have collected enough "sand" to finally count these rare coins with high precision.

Why do they care? Because the way the top quark interacts with the photon is a direct test of the Standard Model (the rulebook of physics). If the interaction looks slightly different than the rulebook predicts, it could be a clue that there is "new physics" hiding in the shadows.

2. The Detective Work: Sorting the Clues

The scientists face a tricky problem: Where did the photon come from?
In the collision, a photon can be emitted by:

The initial particles crashing together (the "start" of the event).
The heavy top quark itself.
The debris left over after the top quark decays.

It's like trying to figure out who threw a ball in a crowded stadium. You can't see the thrower clearly, but you can guess based on how fast and where the ball is going. The scientists use complex computer models to simulate these different "throwing" scenarios. They have to be very careful because their computer models aren't perfect yet; they are trying to stitch together different pieces of a puzzle where some pieces are only half-finished.

3. The "Fake" Photons: Distinguishing Real from Imitation

A major challenge is that sometimes things look like photons but aren't.

The Imposter: An electron or a jet of particles can get misidentified as a photon.
The Background Noise: Sometimes light comes from other messy parts of the collision (like "pileup," where multiple collisions happen at once).

To solve this, the teams use data-driven methods.

CMS uses a strategy called the ABCD method. Imagine they have four rooms. Three rooms are filled with "fake" imposters. By counting how many imposters are in those rooms, they can mathematically predict how many imposters are hiding in the "Signal Room" (where the real photons are) and subtract them out.
ATLAS uses a similar trick, looking at how often electrons are mistaken for photons to estimate the error rate.

4. The Results: What Did They Find?

Counting the Coins: Both teams measured the total number of these events (the "inclusive cross-section"). Their numbers match the Standard Model predictions very closely (within about 5%). It's like weighing a bag of gold coins and finding it matches the expected weight perfectly.
Looking at the Details (Differential Measurements): They didn't just count the coins; they looked at how fast the photons were moving and where they were pointing. They found that while the overall numbers match, there are some small "trends" or wiggles in the data compared to the computer models. This suggests the models need to be tweaked to be more accurate.
The "Charge Asymmetry": They checked if top quarks and anti-top quarks behave differently when a photon is involved. The Standard Model predicts a tiny difference. The teams found a result that matches this prediction, though the data is still a bit fuzzy (statistically limited).

5. Searching for New Physics (The EFT)

The scientists used these measurements to test the Standard Model Effective Field Theory (EFT). Think of this as checking if the rulebook has any hidden footnotes or secret clauses.

They looked at the energy of the photons. If the photons were behaving in a way that suggested a "new force" or a "new particle" was influencing them, the data would have shown a big deviation.
The Verdict: So far, no new physics has been found. The data fits the existing rulebook. However, they have set very strict "speed limits" (limits on coefficients) for how much new physics could be hiding without being noticed yet.

6. The Single Top Mystery

There is another rare process where a single top quark is made with a photon.

CMS saw "evidence" of this in 2018.
ATLAS officially "observed" (confirmed) it in 2023.
Interestingly, they found about 30-40% more of these events than the theory predicted. This is a bit of a mystery that the teams are eager to solve with more data.

7. What's Next?

The paper concludes that while the current results are great, the job isn't done.

Run 3: The LHC is now collecting even more data (Run 3).
Better Tools: The teams have upgraded their "cameras" and "algorithms" to identify photons even better than before.
The Goal: With more data and sharper tools, they hope to measure these top-photon interactions with even higher precision, potentially catching that elusive "new physics" if it's there.

In summary: The ATLAS and CMS teams have successfully counted and analyzed rare top quark events involving light. They found that the universe is behaving mostly as predicted by current theories, but they are keeping a very sharp eye out for any tiny cracks in the rulebook that might reveal something new.

Technical Summary: Photon Production in Top Quark Events at ATLAS and CMS

Problem and Motivation
Top quark production in association with a photon ( $t\gamma$ and $t\bar{t}\gamma$ ) serves as a critical testing ground for Standard Model (SM) predictions, specifically regarding the top-photon coupling. While these processes are rarer than standard top pair production, they possess relatively high cross-sections compared to other top-boson processes and are significant backgrounds for various SM measurements and Beyond the Standard Model (BSM) searches. Crucially, these processes are sensitive to the top-photon coupling, particularly at high photon transverse momentum ( $p_T$ ), allowing for indirect searches for new physics via the SM Effective Field Theory (EFT). The availability of large datasets from LHC Run 2 (approx. 140 fb $^{-1}$ ) has enabled a transition from observation to precision measurements of these rare processes.

Methodology and Experimental Challenges
The analysis relies on data from the ATLAS and CMS collaborations collected during LHC Run 2. The primary experimental challenge involves modeling the various origins of photons within $t\bar{t}\gamma$ events: those emitted from initial state quarks, the top quarks themselves, or top decay products. While distinguishable in theory, they are experimentally indistinguishable.

Theoretical Modeling: At Leading Order (LO), modeling is straightforward. However, at Next-to-Leading Order (NLO), simulating the full $2 \to 7$ process is computationally infeasible. Current strategies involve either using an LO model corrected to NLO cross-sections or "stitching" an NLO model (including initial state/off-shell top photons) with an LO model (including on-shell top/decay photons). Additionally, the $tW\gamma$ process, which interferes with $t\bar{t}\gamma$ beyond LO, is difficult to model; current published results often rely on LO simulations where this interference is absent.
Background Estimation: The dominant background consists of $t\bar{t}$ events with non-prompt or "fake" photons (misreconstructed electrons, jets, or photons from hadronization/pileup). Since simulation often inadequately models these, data-driven methods are employed. CMS utilizes the ABCD method, inverting photon isolation and shower width selections to estimate fake rates. ATLAS employs a similar strategy for hadronic/pileup photons and estimates misreconstructed electrons using $Z \to e\gamma$ candidates from $Z \to ee$ decays.
Signal Selection: Events are selected via single- and dileptonic triggers, requiring specific lepton and jet multiplicities and exactly one high- $p_T$ isolated photon. Deep Neural Networks (DNNs) are trained to separate the $t\bar{t}\gamma$ signal from backgrounds.

Key Contributions and Results

Inclusive Cross-Section Measurements:
- ATLAS: Measured the inclusive cross-section for $t\bar{t}\gamma$ in both single-lepton and dilepton channels. For the first time, the $t\bar{t}\gamma$ production component was measured independently from the decay component. The total inclusive cross-section was measured as $\sigma_{t\bar{t}\gamma} = 319 \pm 4 \text{ (stat)} \pm 15 \text{ (syst)}$ fb (5% precision), consistent with the NLO prediction of $296 \pm 30$ fb. The measurement is dominated by systematic uncertainties related to signal modeling and jet/ $b$ -tagging.
- CMS: Measured fiducial cross-sections for the total $t\bar{t}\gamma$ process (production + decay) in lepton+jets and dilepton channels separately. Results are compatible with NLO QCD calculations.
- Combined $t\bar{t}\gamma + tW\gamma$ : A measurement of the sum of these processes was performed, showing good agreement with fixed-order calculations, though dedicated $tW\gamma$ measurements with improved modeling remain a future need.
Differential Cross-Section Measurements:
Both collaborations performed differential measurements of absolute and normalized cross-sections as a function of photon $p_T$ , pseudorapidity ( $\eta$ ), and angular variables (e.g., $\Delta R$ between photon and closest lepton or $b$ -jet). While data generally agree with theoretical predictions, clear trends indicate difficulties in accurately modeling the process across all kinematic regions.
EFT Interpretation:
The photon $p_T$ distribution was used to constrain Wilson coefficients ( $c_{tZ}$ and $c^I_{tZ}$ ) in the SM EFT.
- CMS: Derived limits using $t\bar{t}\gamma$ events.
- ATLAS: Performed a simultaneous fit to photon $p_T$ in $t\bar{t}\gamma$ and $Z$ boson $p_T$ in $t\bar{t}Z$ regions to exploit the relationship between $t\gamma$ and $tZ$ couplings. The results show that sensitivity is primarily driven by the $t\bar{t}\gamma$ channel, with the combination providing additional constraints.
Charge Asymmetry and Single Top ( $tq\gamma$ ):
- Charge Asymmetry ( $A_C$ ): ATLAS measured the top quark charge asymmetry in $t\bar{t}\gamma$ events, finding $A_C = -0.003 \pm 0.029$ . This is consistent with the SM prediction (which varies between -0.5% and -2% depending on phase space) but is currently limited by statistical uncertainty.
- Single Top ( $tq\gamma$ ): Evidence for $tq\gamma$ was reported by CMS (2018), and observation was reported by ATLAS (2023) using the full Run 2 dataset. Measured cross-sections in both analyses are 30–40% above SM predictions, motivating further differential studies.

Significance and Outlook
The paper highlights that $t\gamma$ and $t\bar{t}\gamma$ processes provide a unique interface between the electroweak and strong sectors, offering a stringent test of the SM and a probe for EFT operators. The primary achievement is the transition to precision measurements (down to ~4-5% for inclusive cross-sections) and the development of sophisticated modeling strategies to handle the complex photon origins and backgrounds.

The authors note that while Run 2 data has established a baseline, significant open tasks remain, particularly regarding the modeling of the $tW\gamma$ interference and differential measurements for $tq\gamma$ . Looking forward, the paper emphasizes that Run 3 data, combined with improved photon reconstruction and identification algorithms (specifically multivariate approaches in CMS), will enable even higher precision measurements of top-photon processes.