⚛️ phenomenology

Freezing-in the Axiverse

This paper employs an effective field theory approach to quantify how multiple light axions contribute to the relativistic degrees of freedom ( $N_{\rm eff}$ ), revealing that their detectability by current and future cosmic microwave background surveys depends sensitively on the flavor structure of their interactions with Standard Model fermions.

Original authors: Christopher Dessert, Soubhik Kumar, Joshua T. Ruderman

Published 2026-02-13

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Christopher Dessert, Soubhik Kumar, Joshua T. Ruderman

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, bustling party. For decades, physicists have been trying to figure out the guest list. We know the "Standard Model" is the VIP section, but we suspect there are many more guests hiding in the shadows—particles so light and shy they barely interact with anything.

This paper, titled "Freezing-in the Axiverse," is about a specific group of these shy guests called axions.

Here is the story of the paper, broken down into simple concepts and analogies.

1. The Mystery of the "Axiverse"

For a long time, physicists thought there might be just one axion. It was proposed to solve a specific puzzle about why the universe doesn't explode with magnetic imbalance (the "Strong CP problem").

But recent theories, especially those involving extra dimensions (like String Theory), suggest there isn't just one axion. There could be hundreds of them. Imagine a whole galaxy of these particles, all with different names and slightly different personalities. This collection is called the "Axiverse."

2. The Problem: Too Many Guests at the Party

The universe has a strict rule about how much "energy" (or radiation) it can hold during its early, hot days. This is measured by a number called $N_{eff}$ (the effective number of neutrino species). Think of $N_{eff}$ as the capacity of the dance floor.

If too many light particles (like axions) are produced in the early universe, they crowd the dance floor.
If the dance floor gets too crowded, the universe expands and cools differently than we observe today.
Current telescopes (like the Planck satellite) have measured the dance floor capacity very precisely. It's almost full, but not too full.

The Dilemma: If there are 100 axions, and they all get produced, the dance floor would be crushed. The universe would look very different. So, either:

There aren't that many axions.
The axions are so shy they never show up to the party.
The party started at a temperature too low to wake them up.

3. The Solution: "Freezing-In" and the "Interaction Menu"

The authors of this paper decided to stop guessing why the axions exist (the "UV" theory) and instead just look at how they might interact with the known particles (the "Effective Field Theory" or EFT approach).

They created a menu of possible interactions, categorized by how "strong" or "complex" they are:

The "Dimension-5" Menu (The VIPs)

These are the simplest ways axions can talk to the Standard Model.

The Analogy: Imagine the axions are trying to get into the VIP section.
The Twist: The authors realized that even if you have 100 axions, they might all be trying to talk to the same bouncer. If they all talk to the same bouncer (the gluon field), they effectively act as one group.
The Result: At this level, you can only have about 44 independent "conversations" happening at once. Even if you have 100 axions, only 44 of them can actually "freeze-in" (get produced) and crowd the dance floor. The rest are "sterile"—they sit in the corner and don't interact.

The "Dimension-6" Menu (The Crowd)

These are more complex interactions.

The Analogy: This is like a group hug. Instead of one axion talking to one particle, two axions have to bump into each other to interact with the Standard Model.
The Twist: Because this interaction requires two axions to be present, every single axion in the universe gets dragged into the party.
The Result: If these interactions exist, all 100 axions could potentially crowd the dance floor. This is much more dangerous for our "dance floor capacity" rules.

4. The Flavor Factor: Who is Talking to Whom?

The paper also looked at which axions talk to which particles (electrons, quarks, etc.). They tested three scenarios:

Anarchy (Chaos): Every axion talks to every particle with equal strength.
- Result: Disaster. The dance floor gets way too crowded. This scenario is mostly ruled out by current data.
Froggatt-Nielsen (The Hierarchy): Some axions talk to heavy particles (like the Top quark) and ignore light ones.
- Result: Better. The "heavy" axions are produced less often. This leaves a little more room on the dance floor, making it possible for the Axiverse to exist.
Minimal Flavor Violation (The Minimalist): Axions only talk to particles in a very specific, organized way.
- Result: Very safe. Only a few axions get produced. This is very hard to detect but fits the rules perfectly.

5. The "Freeze-In" Mechanism

Why call it "Freezing-in"?

Imagine the early universe is a hot soup.
As the universe expands, it cools down.
If the axions interact strongly, they "freeze-out" (stop interacting) early.
If they interact weakly, they never fully join the soup; they just "freeze-in" as a tiny, cold sprinkle of particles that never quite reach the temperature of the soup.
The paper calculates exactly how many axions "freeze-in" based on how hot the universe got (the Reheat Temperature) and how shy the axions are.

6. The Big Conclusion: What Can We Find?

The authors used this math to predict what future telescopes (like the Simons Observatory or CMB-HD) might see.

The Good News: If the universe reheated to a very high temperature, and if the axions have "Anarchic" couplings, we might see a signal soon! The extra radiation would be detectable.
The Bad News: If the axions are "Minimal Flavor Violation" types, or if the universe didn't get very hot, we might never see them. They are too shy.
The Warning: If we find too much extra radiation, it means the "Axiverse" is real, but we have to be very careful about how many axions there are. If there are too many, the universe breaks the rules.

Summary Analogy

Think of the universe as a concert hall.

The Audience: The Standard Model particles (electrons, quarks, etc.).
The Axions: A massive crowd of invisible people trying to sneak in.
The Bouncers (Dimension-5): They can only let in 44 people at a time, no matter how many are waiting outside.
The Group Hug (Dimension-6): If the axions hold hands in pairs, they can all sneak in at once.
The Ticket Counter ( $N_{eff}$ ): The hall has a fire code limit. If too many axions get in, the hall collapses (or rather, the universe looks different than it does).

This paper is the fire marshal's report. It says: "If you have 100 invisible guests, you can only let 44 in unless they hold hands. If they hold hands, you have to check the temperature of the party. If the party was too hot, the hall is full, and we would have seen it. If the party was cool, they stayed outside, and we might never know they were there."

The Takeaway: The existence of a "multiverse of axions" is possible, but it depends entirely on how they interact and how hot the early universe was. Future telescopes will be the ultimate bouncers, checking the guest list to see if the Axiverse is real.

1. Problem Statement

The paper addresses the cosmological implications of the Axiverse, a theoretical scenario arising from string theory and extra-dimensional models where a large number ( $N \sim 10 - 100+$ ) of light axion-like particles (ALPs) exist alongside the QCD axion.

The Conflict: While a single QCD axion typically contributes a negligible amount to the effective number of relativistic degrees of freedom ( $\Delta N_{\text{eff}} \approx 0.027$ if thermalized), the presence of $N$ axions poses a severe challenge. If all $N$ axions were produced thermally in the early universe, the resulting $\Delta N_{\text{eff}} \approx 0.027 N$ would likely exceed observational bounds from the Cosmic Microwave Background (CMB) and Big Bang Nucleosynthesis (BBN), which constrain $\Delta N_{\text{eff}} \lesssim 0.3$ .
The Gap: Previous literature largely focused on single-axion cosmology or specific production mechanisms (like moduli decay). There was no systematic Effective Field Theory (EFT) framework to calculate the "freeze-in" production of a spectrum of $N$ axions coupled to the Standard Model (SM) via generic operators, nor was there a clear understanding of how flavor structures and operator dimensions affect the count of active degrees of freedom.

2. Methodology

The authors adopt a bottom-up Effective Field Theory (EFT) approach to model the interactions between $N$ axions and the SM up to operator dimension six.

EFT Construction:
- Dimension-5 Operators: Linear in the axion field ( $\partial_\mu \phi_i J^\mu_{\text{SM}}$ ). The authors perform a systematic field redefinition (Gram-Schmidt orthogonalization) to identify the interaction basis. They demonstrate that the number of independent axion states coupled to the SM at dimension-5 is limited by the number of independent SM operators ( $N_{\text{ind}} \leq 44$ ), regardless of the total number of axions $N$ .
- Dimension-6 Operators: Quadratic in the axion field ( $\partial_\mu \phi_i \partial_\nu \phi_j O_{\text{SM}}$ ). Unlike dimension-5, these operators generically couple the entire spectrum of $N$ axions to the SM because the interaction matrix is rank- $N$ (or close to it). The authors identify a previously unconsidered charge-radius operator (antisymmetric coupling to hypercharge) and a Higgs-coupling operator.
Flavor Structures: The paper analyzes three distinct benchmarks for the flavor structure of axion-fermion couplings:
1. Anarchy: Random $O(1)$ couplings.
2. Froggatt-Nielsen (FN) Textures: Hierarchical couplings suppressed by powers of a small parameter $\epsilon \approx 0.23$ .
3. Minimal Flavor Violation (MFV): Couplings aligned with SM Yukawa matrices.
Cosmological Calculation:
- The authors calculate the freeze-in production rate of axions from the thermal SM bath.
- They solve the Boltzmann equation for the axion phase-space distribution, deriving analytic expressions for the decoupling temperature ( $T_d$ ) and the resulting energy density.
- They distinguish between regimes where the reheating temperature ( $T_{\text{RH}}$ ) is above or below $T_d$ .
- They verify that mass mixing between axions is negligible for the parameter space of interest ( $f_a \gtrsim 10^7$ GeV).

3. Key Contributions

Systematic EFT Derivation: The first complete derivation of the axiverse EFT up to dimension six, explicitly counting the independent degrees of freedom ( $N_{\text{ind}}$ ) that couple to the SM.
Discovery of a New Operator: Identification and inclusion of a dimension-6 charge-radius operator ( $\partial_\mu \phi_i \partial_\nu \phi_j B^{\mu\nu}$ ) which was previously unconsidered in the context of the axiverse.
Dimension-Dependent Scaling:
- Dimension-5: $\Delta N_{\text{eff}} \propto N_{\text{ind}}$ (capped at $\sim 44$ ). Even with $N=100$ , only $\sim 44$ axions can be produced via dimension-5 operators.
- Dimension-6: $\Delta N_{\text{eff}} \propto N$ . Since these operators are quadratic, they can thermalize the entire spectrum of $N$ axions, making the total number of axions a critical observable.
Flavor Sensitivity: Demonstration that the flavor structure (Anarchy vs. FN vs. MFV) drastically alters the decoupling temperatures and the resulting $\Delta N_{\text{eff}}$ , creating a rich interplay with terrestrial flavor constraints (e.g., $K \to \pi + \text{invisible}$ ).

4. Key Results

Hadronic Axiverse (Gauge Couplings Only):
- Only 3 axions (coupling to gluons, $W$ , and $B$ ) interact at dimension-5.
- $\Delta N_{\text{eff}}$ is small ( $\approx 0.08$ max) unless dimension-6 operators are active.
- Future CMB experiments (CMB-HD) are required to probe the QCD axion decoupling line.
Fermionic Axiverse (Flavor Couplings):
- Anarchy: Predicts the largest $\Delta N_{\text{eff}}$ . Current ACT data already excludes scenarios with high $T_{\text{RH}}$ and $N \gtrsim 10$ . Flavor-violating constraints (NA62) are often stronger than astrophysical bounds.
- Froggatt-Nielsen: Suppression of flavor-violating couplings relaxes constraints, but $\Delta N_{\text{eff}}$ remains significant for high $T_{\text{RH}}$ .
- MFV: Only $\sim 6$ axion states couple significantly (mostly to the top quark). Constraints are weaker, but future surveys (Simons Observatory, CMB-S4) can probe this region.
Dimension-6 Impact:
- Dimension-6 operators (Higgs and Charge Radius) allow all $N$ axions to be produced.
- For $N \sim 100$ , dimension-6 production can easily violate $\Delta N_{\text{eff}}$ bounds unless $T_{\text{RH}}$ is very low or $f_a$ is very high.
- The paper derives a new upper limit on $N$ based on $\Delta N_{\text{eff}}$ that can be more stringent than the "species scale" bound ( $N \lesssim M_{\text{Pl}}^2/f_a^2$ ).
Discovery Space:
- Current CMB data (Planck, ACT, SPT) already rules out high-scale reheating scenarios for models with many axions and unsuppressed couplings.
- Simons Observatory (SO) and CMB-S4 will probe a vast parameter space, potentially detecting the cosmological axion background.
- CMB-HD offers the sensitivity to probe the QCD axion decoupling line and lower reheating temperatures.

5. Significance

This work fundamentally shifts the perspective on axion cosmology from a single-particle problem to a multi-field statistical problem.

Observational Viability: It shows that the "Axiverse" is not automatically ruled out by $\Delta N_{\text{eff}}$ constraints; rather, it requires specific conditions (low reheating temperature, specific flavor textures, or suppression of dimension-6 operators).
New Probes: It establishes $\Delta N_{\text{eff}}$ as a powerful probe of the total number of axions ( $N$ ) and the UV flavor structure, complementing direct detection experiments.
Theoretical Guidance: The identification of the dimension-6 charge-radius operator and the quadratic nature of dimension-6 couplings provides a crucial target for UV completions (e.g., String Theory) to calculate specific Wilson coefficients.
Future Outlook: The paper provides a roadmap for upcoming CMB surveys to either discover a cosmological axion background or place stringent constraints on the existence of light axion spectra, effectively testing the "Axiverse" hypothesis.