⚛️ high-energy experiments

Search for resonant and nonresonant production of pairs of dijet resonances in proton-proton collisions at $\sqrt{s}$ = 13 TeV

Using 138 fb $^{-1}$ of 13 TeV proton-proton collision data, the CMS collaboration conducted a search for resonant and nonresonant production of dijet resonance pairs, setting new limits on diquark and top squark models while observing local excesses with significances of 3.9 and 3.6 standard deviations that remain statistically insignificant globally.

Original authors: CMS Collaboration

Published 2026-10-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: CMS 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

The universe is built on a set of rules known as the Standard Model, a framework that successfully describes the fundamental particles and forces that make up everything we see. Yet, physicists know this map is incomplete. It does not explain gravity, the nature of dark matter, or why the universe is made of matter rather than antimatter. To find the missing pieces, scientists at the Large Hadron Collider smash protons together at nearly the speed of light, creating a chaotic spray of debris. They sift through this debris looking for patterns that do not fit the known rules, searching for new, heavy particles that might have been created in the collision and then immediately decayed into jets of ordinary matter.

In a recent study, the CMS collaboration at CERN turned its attention to a specific, complex scenario: the simultaneous creation of two new particles, each of which immediately breaks apart into a pair of jets. This is a more intricate search than looking for a single new particle because it involves four jets of energy appearing at once. The researchers analyzed data from 138 trillion proton-proton collisions, looking for two distinct possibilities. The first was a "resonant" process, where a massive, unseen parent particle decays into two identical new particles. The second was a "nonresonant" process, where the two new particles are created independently without a parent. By examining the masses of these four-jet clusters, the team set strict limits on what new particles could exist, effectively ruling out many theoretical ideas while keeping a few intriguing possibilities alive.

The search focused on two specific types of hypothetical particles. For the resonant search, the team looked for a heavy particle called a diquark, which would decay into two "vector-like quarks," each of which then splits into a jet pair. For the nonresonant search, they looked for pairs of "top squarks," which are predicted by a theory called supersymmetry. The data was processed by reconstructing the paths and energies of the particles, grouping them into pairs, and then calculating the total mass of the four jets. The team had to be extremely careful to distinguish a genuine signal from the overwhelming background of ordinary particle interactions, which naturally produce four jets but with a smooth, predictable distribution of masses.

The results were a mix of exclusion and tantalizing hints. In the search for the heavy diquark, the researchers found no evidence of a new particle, but they were able to rule out the existence of diquarks with masses below 7.6 trillion electronvolts for a specific model scenario. This is a significant achievement, as it pushes the boundary of where new physics could hide to much higher energy scales than previously tested. Similarly, in the search for top squarks, the team excluded masses between 0.50 and 0.77 trillion electronvolts, with only a tiny gap remaining. These findings significantly extend the limits set by previous experiments, narrowing the window for where these theoretical particles might be found.

Despite the lack of a definitive discovery, the data contained two rare events that stood out from the background. These events occurred in the high-mass tail of the distribution, where the total energy of the four jets was around 8 trillion electronvolts, and the mass of the two jet pairs was around 2 trillion electronvolts. If these events were interpreted as a signal of a new particle, they would represent a statistical fluctuation with a local significance of 3.9 standard deviations. However, when the researchers accounted for the fact that they were searching over a wide range of masses—a factor known as the "look-elsewhere effect"—the global significance dropped to 1.6 standard deviations. This means that while the events are interesting and worthy of further study, they are not yet strong enough to claim a discovery. The team noted that these two events, along with a third isolated event seen in a related analysis, could potentially be explained by a broad, heavy resonance, but more data is required to confirm this.

The study also examined a different kind of excess in the nonresonant search. At an average mass of 0.95 trillion electronvolts, the data showed a slight surplus of events compared to the predicted background, with a local significance of 3.6 standard deviations. Again, after correcting for the look-elsewhere effect, the global significance settled at 2.5 standard deviations. This is a modest excess that suggests a possible deviation from the standard model, but it falls short of the threshold required to be considered a discovery. The researchers emphasized that these fluctuations could simply be a statistical coincidence, and they urged caution in interpreting them as new physics without further confirmation.

Ultimately, this work serves as a rigorous filter for theoretical models. By setting precise upper limits on the production of these particle pairs, the CMS team has told theorists which versions of their ideas are no longer viable. The exclusion of diquarks below 7.6 trillion electronvolts and top squarks in the 0.50 to 0.77 trillion electronvolt range forces new models to either predict heavier particles or different interaction mechanisms. The few events that did appear in the high-mass tails serve as a reminder that the data still holds surprises, but they also highlight the difficulty of distinguishing a true signal from the noise of the universe. As the Large Hadron Collider continues to collect data, these rare events will be scrutinized further, and the limits will be pushed even higher, continuing the slow, steady march toward a deeper understanding of the fundamental laws of nature.

Technical Summary: Search for Resonant and Nonresonant Production of Pairs of Dijet Resonances

Problem and Motivation
This paper presents a search for physics beyond the Standard Model (BSM) in the form of pairs of dijet resonances produced in proton-proton collisions at $\sqrt{s} = 13$ TeV. While single production of dijet resonances ( $pp \to X \to jj$ ) has been extensively studied, the pair production of such resonances ( $pp \to XX \to (jj)(jj)$ ) offers a unique discovery channel, particularly for models involving heavy intermediate states or specific supersymmetric scenarios. The analysis targets final states with at least four jets, distinguishing between two production modes:

Resonant Production: A massive intermediate particle $Y$ decays into a pair of identical dijet resonances $X$ ( $pp \to Y \to XX$ ).
Nonresonant Production: The pair of dijet resonances $X$ is produced directly without an intermediate resonance ( $pp \to XX$ ).

The motivation for this specific search is twofold. First, it extends the reach to high-mass resonances ( $M_X > 0.5$ TeV, $M_Y > 2.0$ TeV) using resolved dijet techniques, complementing previous searches for boosted, low-mass pairs. Second, the search is motivated by a specific event observed in a prior CMS dijet search [1]—a candidate for resonant pair production with a four-jet mass near 8 TeV—which was not optimized for in that earlier analysis.

Methodology
The analysis utilizes a dataset corresponding to an integrated luminosity of 138 fb $^{-1}$ collected by the CMS detector during 2016–2018.

Event Reconstruction and Selection: Events are reconstructed using the particle-flow algorithm. Jets are clustered using the anti- $k_T$ algorithm with a distance parameter of 0.4. The selection requires at least four jets with $p_T > 80$ GeV and $|\eta| < 2.5$ .
Jet Pairing: To reconstruct the two dijet systems from the four leading jets, an algorithm minimizes a specific measure of the $\eta$ - $\phi$ space separation ( $\Delta R$ ) between jets in each pair, aiming to identify pairs that are closer together than uncorrelated QCD jets. The algorithm selects the pairing that minimizes $|\Delta R_1 - 0.8| + |\Delta R_2 - 0.8|$ .
Kinematic Variables: The analysis introduces a dimensionless variable $\alpha = m_{jj} / m_{4j}$ , where $m_{jj}$ is the average dijet mass and $m_{4j}$ is the four-jet mass. This variable is used to bin the data to create unbiased one-dimensional distributions for background estimation, avoiding the "sculpting" effects seen in direct $m_{4j}$ vs. $m_{jj}$ binning.
Background Estimation: The background is derived exclusively from data using a data-driven approach. Empirical functional forms (specifically a modified dijet function, ModDijet-3p) are fitted to the observed mass distributions in bins of $\alpha$ . The goodness-of-fit is validated using pseudo-experiments.
Signal Models:
- Resonant Search: Benchmarked against a diquark model where a scalar diquark $S$ decays to a pair of vector-like quarks $\chi$ , which subsequently decay to a quark and a gluon ( $S \to \chi\chi \to (ug)(ug)$ ).
- Nonresonant Search: Benchmarked against R-parity-violating (RPV) supersymmetry, specifically the pair production of top squarks ( $\tilde{t}$ ) decaying to down and strange quarks ( $\tilde{t}\tilde{t}^* \to (ds)(ds)$ ).
Statistical Analysis: A multibin counting experiment likelihood is used to set limits and estimate significances. Systematic uncertainties (jet energy scale, resolution, luminosity, and background modeling) are treated as nuisance parameters. The modified frequentist criterion ( $CL_s$ ) is used to set 95% confidence level (CL) upper limits.

Key Results

Resonant Search:
- Two events were observed in the tails of the distributions with four-jet masses of 8.0 TeV and 7.9 TeV, and average dijet masses of approximately 2 TeV.
- Interpreted as a signal of a four-jet resonance with mass 8.6 TeV decaying to dijet resonances of 2.1 TeV, these events yield a local significance of 3.9 standard deviations ( $\sigma$ ) and a global significance of 1.6 $\sigma$ (accounting for the look-elsewhere effect).
- Model-independent upper limits on the production cross section are set. For the benchmark diquark model with a mass ratio $\alpha_{true} = 0.25$ , diquark masses below 7.6 TeV (observed) and 7.8 TeV (expected) are excluded at 95% CL.
- An extension to broad mediators (widths up to 10%) was performed. The significance of the 8.6 TeV excess remains high (3.6–3.9 $\sigma$ ) regardless of the width, and the broad resonance interpretation remains compatible with the data.
Nonresonant Search:
- The most significant excess above the predicted background occurs at an average dijet mass of 0.95 TeV.
- This yields a local significance of 3.6 $\sigma$ and a global significance of 2.5 $\sigma$ .
- Top squarks with masses between 0.50 TeV and 0.52 TeV, and between 0.58 TeV and 0.77 TeV, are excluded at 95% CL, with the exception of a small interval between 0.52 and 0.58 TeV. This significantly extends previous limits (which excluded masses up to 0.52 TeV).

Significance and Claims
The paper claims to present the first LHC limits on the resonant pair production of dijet resonances via high-mass intermediate states. It represents the first time CMS has searched for this specific topology.

Regarding the observed excesses, the author maintains a modest tone. While the local significances (3.9 $\sigma$ for resonant, 3.6 $\sigma$ for nonresonant) are notable, the global significances (1.6 $\sigma$ and 2.5 $\sigma$ , respectively) are not sufficient to claim a discovery. The paper explicitly states that "more data are needed to establish if these events are the first hints of a signal." The results are presented as constraints on specific BSM models (diquarks and RPV SUSY) and as a demonstration of the sensitivity of the four-jet final state to high-mass resonances, rather than as a definitive discovery of new physics. The paper concludes that the observed events are interesting candidates that warrant further investigation with higher luminosity and collision energy.

More like this