Primordial black hole dark matter from axion inflation

This paper demonstrates that primordial black holes formed via axion inflation coupled to a U(1) gauge field can constitute all dark matter in the asteroidal mass range even with subdominant inflaton gradient energy, while simultaneously predicting a detectable stochastic gravitational wave background at LISA that can distinguish between different density fluctuation statistics.

The Gist

Imagine the early universe not as a smooth, calm ocean, but as a turbulent sea where tiny ripples could suddenly grow into massive, invisible whirlpools. These whirlpools are Primordial Black Holes (PBHs), and this paper explores how they might have formed to become the "Dark Matter" that holds our universe together.

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

1. The Setup: The Inflaton and the Gauge Field

Think of the early universe as a giant balloon being blown up rapidly (this is Inflation). Inside this balloon, there is a rolling ball called the Axion (the inflaton). Usually, this ball rolls down a gentle hill, creating a smooth expansion.

However, in this model, the Axion is connected to a "magnetic field" (a U(1) gauge field) via a special rope. As the Axion rolls, it pulls on this rope, causing the magnetic field to vibrate and amplify, much like plucking a guitar string that gets louder and louder the faster you move your hand.

2. The Problem: The "Backreaction"

In previous studies, scientists tried to predict how loud these vibrations would get using a simple shortcut. They assumed the Axion kept rolling at a steady speed, ignoring the fact that the vibrating magnetic field pushes back on the Axion.

The Analogy: Imagine a child on a swing. If you push the swing, it goes higher. But if the swing is attached to a heavy spring, the spring pulls back. Previous models ignored the spring's pull. This paper says, "We can't ignore the spring!" If the magnetic field gets too strong, it slows down the Axion. This is called backreaction.

3. The Solution: A Better Simulation

The authors didn't just use a shortcut. They built a sophisticated computer simulation (a "homogeneous backreaction" model) to track exactly how the Axion and the magnetic field interact in real-time.

• They checked to make sure their simulation was valid by ensuring the "messiness" of the Axion's movement (gradient energy) stayed very small compared to its forward motion (kinetic energy). It was like checking that the child on the swing wasn't wobbling so much they fell off.
• Result: They found that even with the magnetic field pushing back hard, the Axion still rolls fast enough to create the conditions needed for black holes.

4. The Big Question: How "Lumpy" is the Universe?

To form a black hole, the universe needs to be "lumpy" enough in certain spots. The paper tackles a major uncertainty: What do these lumps look like?

• Scenario A (Gaussian): Imagine the lumps are like a standard bell curve. Most are average, and extreme lumps are very rare.
• Scenario B (Chi-Squared): Imagine the lumps are more "spiky." Extreme lumps are much more common than in the bell curve.

The authors ran the simulation for both scenarios.

• The Finding: In both cases, they found a "sweet spot" (specifically for black holes the size of asteroids) where the lumps are just right to collapse into black holes that could make up 100% of the Dark Matter in the universe.

5. The "Smoking Gun": Gravitational Waves

Here is the most exciting part. The same process that creates these black hole "whirlpools" also creates ripples in space-time called Gravitational Waves.

• The Analogy: If the Axion is a drumstick hitting a drum (the magnetic field), the black holes are the sound of the drum hitting the floor, but the Gravitational Waves are the hum of the drum itself.
• The Prediction: The paper predicts that this "hum" will be loud enough to be heard by LISA, a future space-based observatory (like a giant ear in space).
• The Twist: The "hum" sounds different depending on whether the universe was "Gaussian" or "Chi-Squared."
  • If the universe was Gaussian, the hum is louder.
  • If it was Chi-Squared, the hum is quieter.
  • Why this matters: By listening to LISA, we won't just know if these black holes exist; we will know which statistical rule (Gaussian or Chi-Squared) the universe followed to make them.

Summary of Claims

1. Dark Matter: It is possible that all the dark matter in the universe is made of tiny, asteroid-sized black holes formed in the first fraction of a second after the Big Bang.
2. Method: This is possible even when the magnetic fields push back hard on the inflaton, provided we use a more accurate simulation method than previous studies.
3. Verification: We don't need to guess if this is true. The process creates a specific gravitational wave signal that LISA will be able to detect.
4. Discrimination: The strength of that signal will tell us exactly how the density fluctuations (the "lumps") were distributed in the early universe, solving a long-standing mystery about the statistics of these events.

In short, the paper says: "We have a better way to calculate how these black holes form, and if they exist, the universe will sing a song that LISA can hear, proving exactly how they were made."

Technical Summary

1. Problem Statement

The paper addresses the production of Primordial Black Holes (PBHs) as a candidate for Dark Matter (DM), specifically in the "asteroidal" mass window ($M \sim 10^{-12} M_\odot$), where PBHs could constitute 100% of the DM. The authors focus on axion inflation coupled to a $U(1)$ gauge field via a pseudo-scalar interaction ($\phi F\tilde{F}$).

The core problem addressed is the reliability of previous calculations regarding PBH abundance in this regime. Earlier studies (Refs. [18–24]) relied on a "local backreaction" approximation. This method assumes the gauge field amplification parameter $\xi$ is constant and uses an analytic solution for gauge modes. However, this neglects the memory effect: the growth of gauge modes depends on the full history of $\xi(t)$, not just its instantaneous value. In the strong amplification regime required to produce significant PBHs, this local approximation becomes unreliable.

Furthermore, there is significant uncertainty regarding the statistics of the primordial density perturbations ($\delta\rho$). Previous works often assumed a $\chi^2$ distribution (arising from the square of Gaussian gauge modes), while lattice simulations suggest the statistics may be closer to Gaussian or intermediate. Since PBH formation depends on the extreme tail of the probability distribution function (PDF), the choice of statistics drastically alters the predicted abundance.

2. Methodology

The authors employ a rigorous, state-of-the-art approach to resolve these issues:

• Homogeneous Backreaction Regime: Instead of the local approximation, they solve the equations of motion for the gauge field modes numerically on a grid of Fourier modes. They treat the axion field as a homogeneous condensate (mean-field approximation). This "homogeneous backreaction" scheme accounts for the non-local memory effects of the gauge field evolution.
• Model Construction: They construct a specific axion potential $V(\phi)$ with five distinct branches:
  1. A flat region consistent with CMB observations (Planck pivot scale).
  2. Intermediate regions with steeper slopes to induce significant gauge field amplification.
  3. A quadratic minimum to end inflation.
     The parameters are tuned to ensure the gradient energy density of the axion remains subdominant to its kinetic energy (validating the homogeneous approximation).
• Perturbation Calculations:
  • They compute scalar ($\zeta$) and tensor (GW) power spectra by summing vacuum contributions and sourced contributions (from gauge fields).
  • They analyze two statistical scenarios for the curvature perturbation $\zeta$: Gaussian and $\chi^2$.
• PBH Abundance Calculation:
  • They use a modern method to calculate the collapse threshold ($C_{th}$) based on the specific shape of the power spectrum (using the compaction function).
  • They incorporate the non-linear relation between curvature and density contrast on super-Hubble scales.
  • For the $\chi^2$ case, they apply threshold statistics assuming a local quadratic relation between the curvature and an underlying Gaussian variable.

3. Key Contributions

1. Implementation of Homogeneous Backreaction: The paper replaces the analytic "local" gauge mode functions with numerically computed functions that include the full time-evolution history (memory effect). This is the first application of this improved scheme to PBH production in axion inflation.
2. Validation of the Approximation: They explicitly monitor the ratio of axion gradient energy to kinetic energy. They demonstrate that for the parameters yielding PBH DM, this ratio remains well below the $10^{-2}$ threshold (specifically $< 6 \times 10^{-3}$ for Gaussian and $< 10^{-4}$ for $\chi^2$), confirming that the homogeneous backreaction scheme is valid and full lattice simulations are not strictly necessary for these results.
3. Statistical Uncertainty Analysis: Rather than assuming a single statistics, they provide results for both Gaussian and $\chi^2$ limits. They quantify how the required power spectrum amplitude differs between these two cases to achieve the same PBH abundance.
4. Gravitational Wave (GW) Prediction: They calculate the Stochastic Gravitational Wave Background (SGWB) generated by the same mechanism, noting its detectability by LISA.

4. Key Results

• PBH Dark Matter: The mechanism successfully produces PBHs that can account for 100% of the Dark Matter in the asteroidal mass range ($M \sim 10^{-12} M_\odot$).
• Impact of Backreaction: Neglecting backreaction leads to a massive overprediction of PBH abundance (by factors of $\sim 12$ for $\chi^2$ and $\sim 6 \times 10^6$ for Gaussian). Properly accounting for backreaction is crucial for accurate predictions.
• Power Spectrum Requirements:
  • To produce the same PBH abundance, the Gaussian case requires a significantly larger power spectrum amplitude (variance) than the $\chi^2$ case because the Gaussian tail is less populated.
  • Consequently, the associated GW signal is stronger in the Gaussian case.
• GW Signal: The model predicts a SGWB with an amplitude detectable by the LISA space interferometer. The signal peaks in the LISA frequency band.
• Discrimination: The amplitude of the GW signal at LISA can experimentally discriminate between the Gaussian and $\chi^2$ statistical assumptions. If the GW signal is observed at the amplitude predicted by the Gaussian case, it implies a specific underlying physics; if it matches the $\chi^2$ prediction, it implies different statistics.

5. Significance

This work provides a robust theoretical framework for axion inflation models producing PBH dark matter. By moving beyond the simplified local backreaction approximation, the authors demonstrate that such models are viable and computationally controllable without needing prohibitively expensive full lattice simulations.

The paper establishes a clear link between PBH formation, axion dynamics, and gravitational wave astronomy. It suggests that the detection (or non-detection) of a specific GW signal by LISA could not only confirm the existence of asteroidal-mass PBHs as dark matter but also reveal fundamental properties of the primordial perturbations (specifically their statistical nature) that are otherwise inaccessible. This transforms the PBH production mechanism from a purely theoretical construct into a testable hypothesis with distinct observational signatures.