⚛️ phenomenology

Cosmological Collider Searches beyond the Hubble Scale with Planck Data

Using Planck data, this paper conducts the first cosmological collider searches for primordial non-Gaussianity mediated by heavy scalar particles, finding no evidence for masses near the Hubble scale but reporting a $1.7\sigma$ global hint of non-zero non-Gaussianity for parametrically heavier particles excited by a chemical potential.

Original authors: Soubhik Kumar, Qianshu Lu, Zhong-Zhi Xianyu, Yisong Zhang

Published 2026-03-18

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Soubhik Kumar, Qianshu Lu, Zhong-Zhi Xianyu, Yisong Zhang

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, expanding balloon. About 13.8 billion years ago, this balloon underwent a period of incredibly rapid expansion called Cosmic Inflation. During this split second, the universe grew from the size of a grain of sand to the size of a grapefruit almost instantly.

Scientists believe that if we look closely at the "frosting" left on this balloon today (the Cosmic Microwave Background, or the afterglow of the Big Bang), we might find tiny ripples. These ripples are the seeds of all the galaxies, stars, and planets we see today.

This paper is about a new way of listening to the "music" of these ripples to find particles so heavy and energetic that we could never create them in any particle accelerator on Earth.

Here is the breakdown of the paper's story, using simple analogies:

1. The "Cosmological Collider"

Usually, to find new particles (like the Higgs boson), we smash atoms together in giant machines like the Large Hadron Collider (LHC). But the LHC can only reach energies up to a certain limit.

The authors propose that the early universe was actually a much bigger, more powerful collider. During inflation, the energy was so high that it could create particles that are 100 times heavier than anything we can make on Earth.

The Analogy: Imagine the LHC is a small kitchen blender. It can make a smoothie, but it can't crush a diamond. The early universe, however, was a giant industrial crusher that could smash diamonds into dust. We can't go back in time to see the dust, but we can look at the "shrapnel" (the ripples in the sky) left behind to guess what kind of diamonds were crushed.

2. The Problem: The "Heavy" Particles are Silent

In the standard theory, if a particle is too heavy (much heavier than the energy of the inflation), it's like trying to push a boulder up a hill that's too steep. Quantum mechanics says these heavy particles just won't appear. They are "off-shell" and silent.

The Analogy: Imagine trying to hear a whisper from a person standing 10 miles away. If the wind is too weak (low energy), you can't hear them. The heavier the person (the heavier the particle), the quieter they are. For a long time, scientists thought we could only hear people standing close by (light particles).

3. The First Discovery: The "Triple-Exchange"

The authors looked at a specific type of interaction where three heavy particles swap energy back and forth before creating the ripples we see. They called this the "Triple-Exchange" diagram.

The Analogy: Imagine three friends passing a ball. If they just toss it once, it's a simple game. But if they pass it back and forth three times in a complex dance, the pattern of their movement creates a very specific, unique rhythm.
The Result: The authors calculated exactly what this "rhythm" (the shape of the signal) should look like and searched for it in data from the Planck satellite (a telescope that mapped the early universe).
The Verdict: They didn't find it. The search came up empty. This is actually good science! It tells us that if these specific heavy particles exist, they are either too rare or too heavy for us to see with current data. It sets a new limit on how heavy these particles can be.

4. The Second Discovery: The "Chemical Potential" (The Volume Knob)

This is the most exciting part. The authors realized that in the early universe, there might have been a "chemical potential."

The Analogy: Imagine the heavy particles are trying to whisper, but the wind is too weak. Now, imagine someone turns up the volume knob on the wind. Suddenly, even the heaviest, quietest particles can be heard clearly.
How it works: This "volume knob" is a specific energy field that "excites" the heavy particles, allowing them to pop into existence even if they are much heavier than the universe's expansion energy.
The Result: When the authors searched for this specific "loud" signal, they found something interesting.
- They found a hint of a signal (a "blip" in the data) that looks like a heavy particle with a mass about 6.7 times the energy scale of inflation.
- The Catch: In science, a "hint" isn't a discovery yet. This hint has a statistical significance of 1.7 sigma.
- The "Sigma" Scale: Think of it like a coin flip.
  - 1 Sigma: You flipped heads once. Could be luck.
  - 3 Sigma: You flipped heads 10 times in a row. Unlikely, but possible.
  - 5 Sigma: You flipped heads a million times. That's a discovery.
- Their result (1.7 sigma) is like flipping heads 3 or 4 times. It's interesting, but it could easily be random noise.

5. Why This Matters

Even though they didn't find a "smoking gun" (a 5-sigma discovery), this paper is a huge step forward for two reasons:

New Tools: They built a new "detector" (mathematical tools) that can listen for these specific heavy particles. Before this, we didn't know how to look for them properly.
Expanding the Horizon: They proved that we can look for particles that are much heavier than we thought possible, provided there is a "volume knob" (chemical potential) turning them on.

The Bottom Line

The authors took a map of the baby universe (Planck data) and used a new, sophisticated listening device to hunt for the echoes of super-heavy particles.

They found nothing for the standard "quiet" heavy particles.
They found a faint, intriguing whisper for the "loud" heavy particles, but it's not loud enough to be sure it's not just static.

The Future: The authors say, "Don't give up!" They are waiting for new telescopes (like SPHEREx and Spec-S5) that will take even sharper pictures of the universe. With better data, that faint whisper might turn into a clear shout, revealing the fundamental building blocks of reality at energies we can only dream of.

1. Problem Statement

The paper addresses the challenge of detecting primordial non-Gaussianity (NG) to probe the physics of cosmic inflation and fundamental interactions at energy scales far exceeding those accessible by terrestrial colliders (up to $10^{16}$ GeV). Specifically, the authors focus on the Cosmological Collider (CC) program, which aims to detect massive particles ( $M$ ) produced during inflation through their distinct oscillatory signatures in the bispectrum (three-point correlation function) of curvature fluctuations.

Two primary technical hurdles are identified:

Mass Suppression: In standard scenarios, the production of heavy particles ( $M \gg H$ , where $H$ is the Hubble scale) is exponentially suppressed by the Boltzmann factor $\exp(-\pi M/H)$ , limiting the reach of current searches to $M \lesssim 2H$ .
Incomplete Shape Modeling: Previous searches often relied on simplified templates or focused only on "single-exchange" diagrams. Realistic models involve complex interactions (e.g., "triple-exchange" diagrams) and full shape functions that include both oscillatory signals and smooth backgrounds. Ignoring the full shape reduces sensitivity.

2. Methodology

The authors employ an Effective Field Theory (EFT) approach within the framework of slow-roll inflation, extending it to include heavy scalar fields. Their methodology consists of three main components:

A. Theoretical Framework & Diagrams

Triple-Exchange Diagram: The authors identify that for heavy scalars with mass $M \sim H$ , the triple-exchange Feynman diagram (involving three massive scalar propagators) can yield a parametrically larger bispectrum compared to single- or double-exchange diagrams. This is due to specific mixing terms between the inflaton and the heavy scalar.
Scalar Chemical Potential Mechanism: To overcome the exponential suppression for $M \gg H$ , they utilize a mechanism where a chemical potential ( $\omega$ ) excites on-shell heavy scalars. This mechanism, derived from a complex scalar field with a time-dependent background, allows for the production of particles with masses up to $M \sim 10H$ without the $\exp(-\pi M/H)$ suppression, provided $\omega - M \approx 3H$ .

B. Full Shape Function Computation

Numerical Integration (Triple-Exchange): Calculating the triple-exchange bispectrum analytically is intractable due to four nested time integrals. The authors use the "coupled mode function" method within the Schwinger-Keldysh (in-in) formalism. They solve the coupled equations of motion for the inflaton and heavy scalar non-perturbatively to obtain mixed propagators, reducing the problem to a single non-nested integral.
Analytical Bootstrap (Chemical Potential): For the chemical potential scenario (single-exchange), they utilize cosmological bootstrap techniques to derive analytical expressions for the shape functions arising from temporal derivative and covariant derivative couplings.
Grid Generation: They generate high-resolution 3D shape functions ( $S$ ) on a grid of $216,000$ points in the $(k_1, k_2, k_3)$ momentum space to capture the full kinematic dependence, including the oscillatory features.

C. Data Analysis

Tool: They utilize the CMB-BEST package to perform the search on Planck 2018 CMB data.
Normalization: To avoid artificial fluctuations in the estimated $f_{NL}$ due to rapid oscillations in the equilateral limit, they normalize the shape function such that its maximum value over the physical volume is unity.
Statistical Treatment: They account for the look-elsewhere effect (LEE) using the correlation matrix from CMB-BEST to calculate global significance.

3. Key Contributions

First Search for Triple-Exchange Bispectrum: This is the first observational search specifically targeting the bispectrum mediated by the triple-exchange diagram, which is theoretically predicted to be the dominant contribution for $M \sim H$ in certain EFTs.
First Search Beyond $H$ with Chemical Potentials: They perform the first CC search for particles with masses parametrically larger than the Hubble scale ( $M \gg H$ ) using the scalar chemical potential mechanism.
Full-Shape Analysis: Unlike previous works that used simplified templates, this study uses the full shape function (including background and oscillatory components) derived from first principles, which is crucial for detecting signals that do not resemble standard local, equilateral, or orthogonal shapes.
Constraint on QSFI: They provide the first constraints on the Quasi-Single-Field Inflation (QSFI) model based on the triple-exchange contribution.

4. Results

The analysis of Planck 2018 data yields the following results:

Triple-Exchange Search ( $M \sim H$ ):
- No evidence for non-zero NG was found.
- The results are consistent with $f_{NL} = 0$ at the 95% Confidence Level (CL).
- This places the first constraints on the parameter space of the QSFI model for $1.5H \leq M \leq 2H$ .
Scalar Chemical Potential Search ( $M \gg H$ ):
- Local Significance: For a specific parameter point with chemical potential $\omega = 10H$ and mass $M = 6.7H$ (satisfying $\omega - M \approx 3H$ ), they found a 2.5 $\sigma$ local significance for non-zero NG.
- Global Significance: After correcting for the look-elsewhere effect, this corresponds to a 1.7 $\sigma$ global significance.
- Measured Amplitude: The estimated bispectrum strength is $f_{NL} = -203 \pm 82$ (68% CL) for the "covariant shape" coupling.
- Shape Distinctiveness: The signal shape ( $S_{cov}$ ) is distinct from standard templates (local, equilateral, orthogonal) and the triple-exchange shape, showing distinctive oscillations starting in the mildly equilateral regime ( $k_3/k_1 \sim 0.4$ ).
- Model Consistency Check: While the 1.7 $\sigma$ hint is intriguing, the authors note that in the specific parameter space tested ( $\omega=10H$ ), the required $f_{NL}$ conflicts with constraints from the power spectrum correction ( $\Delta P_\zeta < P_\zeta$ ). However, they argue that for higher chemical potentials (e.g., $\omega \sim 50H$ ), the power spectrum bounds relax, potentially allowing the observed significance to be explained by this model.

5. Significance

Probing Super-Hubble Physics: The paper demonstrates a viable pathway to probe physics at energy scales $M \gg H$ (up to $\sim 10H$ ) using CMB data, a regime previously thought inaccessible due to Boltzmann suppression.
Methodological Advancement: It establishes a robust pipeline for computing and searching for complex, non-factorizable bispectrum shapes, moving beyond simplified templates.
Future Prospects: The authors emphasize that while the current Planck data shows only a mild hint (1.7 $\sigma$ ), future Large Scale Structure (LSS) surveys like SPHEREx and Spec-S5 will significantly improve sensitivity. These future datasets could confirm or rule out the chemical potential mechanism and extend the mass reach of the Cosmological Collider.
Theoretical Insight: The work highlights the importance of "triple-exchange" processes and chemical potentials in maximizing the detectability of heavy states, guiding future model building and observational strategies.