⚛️ phenomenology

SHiP as a (post-)discovery machine: identifying the diphoton signals' origin

This paper demonstrates that the SHiP experiment can distinguish the underlying production mechanisms of a feebly coupled axion-like particle decaying into diphotons by analyzing the kinematics of just a few reconstructed events, even in the presence of realistic detector effects.

Original authors: Matei Climescu, Malte Fogde Mikkelsen, Maksym Ovchynnikov

Published 2026-09-18

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Matei Climescu, Malte Fogde Mikkelsen, Maksym Ovchynnikov

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

Deep within the subatomic world, physicists hunt for particles that barely interact with the rest of the universe. These "feebly interacting" particles are ghostly; they can pass through mountains of lead without leaving a trace, making them incredibly difficult to catch. One promising candidate for such a particle is the axion-like particle, a hypothetical object that could help explain why the universe has more matter than antimatter. If these particles exist, they might be created in high-energy collisions at facilities like CERN, but they would travel far away from the collision point before decaying into something we can see, such as two flashes of light. The challenge for scientists is not just finding these rare events, but understanding exactly how the particles were born. Since the moment of creation happens inside a thick, impenetrable target and cannot be watched directly, researchers must deduce the particle's origin by studying the path and energy of the light it emits later.

A new study by Matei Climescu, Malte Fogde Mikkelsen, and Maksym Ovchynnikov tackles this puzzle using the upcoming SHiP experiment, a specialized detector designed to catch these elusive decays. The team asked a fundamental question: if the experiment finds a new particle that breaks apart into two photons, can the pattern of that break-up tell us which force created it? There are two main ways such a particle could be made. In one scenario, it is born from the interaction of light itself, a process driven by electromagnetic forces. In the other, it emerges from the decay of heavier, unstable particles like B mesons, driven by the weak nuclear force. While both pathways lead to the same final result—two photons flying apart—the journey the particle takes before it decays is different. The researchers wanted to know if the energy of those photons and the specific location where they appear could act as a fingerprint to distinguish between these two birth stories.

To answer this, the team built a detailed computer simulation of the SHiP detector, modeling how a new particle would travel from the target, through a long vacuum chamber, and finally decay into two photons. They tracked how the particle's speed and the angle at which it was launched affected where it would eventually break apart. They found that the energy of the photons alone is not enough to solve the mystery. A particle created by light forces but living for a very short time can mimic the energy signature of a particle created by weak forces that lives much longer. The two scenarios can look nearly identical if one only looks at how fast the photons are moving. However, the location where the decay happens tells a different story. Because the two types of particles are launched at different angles and travel at different speeds, they decay in different spots within the detector. A particle born from light tends to decay closer to the start of the chamber if it is short-lived, while one born from heavy mesons tends to drift further down the line before breaking up.

By combining the energy of the photons with the precise coordinates of their decay, the researchers discovered that the two origins become easy to tell apart. In their simulations, they found that if the detector sees just two to four clear events of these decaying particles, it is enough to confidently say whether the particle was made by light or by the weak force, provided the particle's mass is between 0.2 and 1 GeV. If the particle is a mix of both forces, the task becomes harder, requiring perhaps a few dozen to a hundred events to untangle the contributions. The study also accounted for the real-world limitations of the detector, such as how the instrument measures the position of the photons and how the data is processed. Even with these realistic imperfections, the method holds up, showing that a very small number of events can reveal the hidden nature of the particle's creation.

This work transforms the SHiP experiment from a simple search engine into a powerful diagnostic tool. If the experiment discovers a new particle, the team's analysis shows that scientists will not have to wait for a massive amount of data to understand what it is. Instead, the very first few sightings could reveal the fundamental forces at play, distinguishing between a particle born of light and one born of heavy matter. This ability to identify the production mechanism so quickly would be a major step forward in understanding the hidden sector of the universe, turning a rare discovery into a clear window on the laws of physics.

Technical Summary: SHiP as a (post-)discovery machine: identifying the diphoton signals' origin

Problem Statement
The upcoming SHiP (Search for Hidden Particles) experiment aims to probe feebly interacting particles (FIPs) produced by 400 GeV protons striking a thick target. While SHiP has the potential to discover new particles, a critical post-discovery challenge remains: identifying the specific production mechanism of a discovered particle. In the baseline SHiP setup, production occurs within the thick target and is unobservable directly. If different theoretical models (e.g., an Axion-Like Particle (ALP) coupled to $U(1)_Y$ versus one coupled to $SU(2)_L$ ) produce the same particle with the same visible decay (e.g., a diphoton final state), the underlying physics cannot be distinguished solely by the particle's existence. Furthermore, the unknown lifetime of the new particle creates a degeneracy; a change in lifetime can mimic a change in the production interaction by altering the kinematic distributions of the decays that reach the detector. This paper investigates whether the kinematics of visible decays—specifically energy and decay position—can resolve this ambiguity and determine the production interaction.

Methodology
The authors analyze a benchmark scenario involving an ALP with a dominant diphoton decay, comparing two distinct production hypotheses:

Photophilic Model: Production via Primakoff scattering from primary and cascade photons in the target.
Electroweak ( $SU(2)_L$ ) Model: Production via flavor-changing neutral current decays of $B$ mesons and charged kaons ( $B \to X_s a$ , $K^\pm \to \pi^\pm a$ ), alongside photon production induced by the same operator.

The analysis proceeds in two complementary stages:

Semianalytic Calculation: Using the SensCalc and EventCalc-SHiP frameworks, the authors generate production distributions, propagate particles through the SHiP geometry (including the 50 m decay volume and the electromagnetic calorimeter, ECAL), and calculate acceptance. They assume perfect reconstruction of ALP energy ( $E_a$ ), polar angle ( $\theta_a$ ), and decay position ( $z, r_\perp$ ).
Statistical Inference: The study employs a likelihood-based approach to compare the two hypotheses. For a given number of observed events ( $N$ ), the unknown lifetime ( $c\tau$ ) is profiled independently under each hypothesis. The test statistic $T$ compares the maximum likelihoods of the pure photophilic and pure $SU(2)_L$ models. The required number of events ( $N_{90\%}$ ) is defined as the minimum count needed to achieve $\ge 90\%$ correct model identification across the allowed lifetime domains.

The analysis accounts for geometric acceptance (requiring both photons to intersect the ECAL), propagation effects, and the specific kinematic signatures of the parent particles (e.g., the softer spectrum of kaon decays vs. the harder spectrum of $B$ decays).

Key Contributions

Kinematic Discrimination: The paper demonstrates that while the ALP energy distribution alone suffers from a degeneracy between production mechanism and lifetime (e.g., a short-lived photophilic ALP can mimic the energy spectrum of a long-lived $SU(2)_L$ ALP), the inclusion of decay position breaks this degeneracy.
Role of Transverse Position: The transverse decay coordinate ( $r_\perp$ ) is identified as the most powerful discriminator. Since $r_\perp$ is linked to the production angle via the decay length, and different production sources (B-mesons vs. photons) have distinct angular distributions, $r_\perp$ effectively encodes the production origin.
Event Count Requirements: The study quantifies the statistical power of SHiP for model discrimination. It establishes that only a small number of reconstructed events are sufficient to distinguish between pure interactions, provided the lifetime is profiled correctly.
Extension to Mixtures: The methodology is extended to scenarios where both interactions contribute (mixed couplings), defining a statistical test to reject pure limits in favor of a mixture.

Results

Pure Interactions: For ALP masses between 0.2 and 1 GeV, distinguishing between pure $U(1)_Y$ $U (1)_{Y}$ and $SU(2)_L$ $S U (2)_{L}$ interactions requires remarkably few events.
- Using only energy ( $E_a$ ), over 130 events are required at $m_a = 0.3$ GeV due to the lifetime-energy degeneracy.
- Including the longitudinal decay position ( $z$ ) reduces the requirement to 28 events.
- Including the transverse decay position ( $r_\perp$ ) further reduces the requirement to 8 events.
- Using the full set of observables ( $E_a, z, r_\perp$ ), the requirement drops to 7 events at 0.3 GeV and 3 events at 1 GeV.
Mixed Interactions: For coupling admixtures, the study indicates that distinguishing the mixture from pure limits requires "several tens to $O(10^2)$ events" in favorable regions, depending on the mixing parameter $\xi$ and the interference sign.
Detector Effects: A full simulation of the SHiP electromagnetic calorimeter (including photon showers, digitization, and clustering) confirms that these discrimination capabilities hold even when realistic detector response is included, though the primary results presented in the summary above assume perfect reconstruction to establish the fundamental kinematic limits.

Significance
The paper claims that SHiP is not merely a discovery machine but a "post-discovery" instrument capable of characterizing the nature of new physics. By demonstrating that the production mechanism can be inferred from the kinematics of the decay products alone, the work establishes that SHiP can differentiate between photophilic and electroweak ALP scenarios with high confidence using a minimal dataset. This capability is crucial for understanding the underlying physics of any potential discovery, as it allows the experiment to probe the specific operators (e.g., $U(1)_Y$ vs. $SU(2)_L$ ) responsible for the particle's existence without needing to observe the production vertex directly. The results suggest that even with a small number of observed decays, SHiP can resolve the origin of diphoton signals, provided the analysis properly accounts for the interplay between lifetime, production kinematics, and geometric acceptance.

More like this