Are Primordial Black Holes a Natural Dark Matter Candidate?

This paper challenges the prevailing view that Primordial Black Holes (PBHs) are generically fine-tuned dark matter candidates by demonstrating that, when evaluated across a broad landscape of production mechanisms and compared to particle dark matter benchmarks using uniform naturalness measures, PBHs span a full spectrum of naturalness tiers ranging from as natural as standard WIMPs to severely tuned, thereby proving that dismissing them as a whole conflates worst-case scenarios with a diverse landscape of viable models.

The Gist

The Big Question: Is "Dark Matter" Too Hard to Make?

Imagine the universe is a giant, complex machine. We know most of it is made of "Dark Matter," a mysterious substance we can't see but can feel through gravity. For decades, scientists have tried to build this machine using different blueprints.

Some blueprints use tiny, invisible particles (like WIMPs or Axions). Others suggest the dark matter is actually made of tiny, ancient black holes formed right after the Big Bang, called Primordial Black Holes (PBHs).

A common complaint against the "Black Hole" blueprint is that it is too fine-tuned. Critics say: "To make the universe look the way it does, you have to adjust the knobs on the black hole machine with such extreme precision that it's impossible by accident. It's like trying to balance a pencil on its tip during an earthquake."

This paper asks: Is that true? Or is the "Black Hole" blueprint actually just as natural as the particle blueprints?

The Tool: The "Sensitivity Meter"

To answer this, the author (Stefano Profumo) invented a universal "Sensitivity Meter." He didn't just look at black holes; he looked at 12 different ways to make dark matter (including black holes, particles, and weird mixtures).

He applied the same test to all of them: "How much do I have to tweak the settings to get the right amount of dark matter?"

• Low Sensitivity (Natural): If you turn a knob a little bit, the result changes a little bit. It's easy to hit the target.
• High Sensitivity (Fine-Tuned): If you turn a knob a tiny, tiny bit, the result explodes or disappears. You have to be incredibly precise to hit the target.

The Results: Three "Naturalness" Tiers

The paper found that all 12 methods fall into three distinct "tiers" of difficulty. Surprisingly, both black holes and particles appear in every single tier.

Tier 1: The "Easy Mode" (Natural)

These are the most forgiving blueprints. You can turn the knobs almost anywhere, and you still get the right amount of dark matter.

• The Winners:
  • Asymmetric Dark Matter: Like a scale where the weight is set by a simple ratio.
  • Post-Inflationary Axions: A specific type of particle that naturally settles into place.
  • Biased Domain Wall Black Holes: This is the paper's big surprise. Imagine a cosmic net (domain walls) that collapses. If the net is slightly "biased" (uneven), it naturally forms black holes in the perfect size range. The author found this method is just as "easy" and natural as the best particle theories. It requires no magical precision.

Tier 2: The "Medium Mode" (Mildly Tuned)

These require a bit more care. You need to aim for a specific spot, but it's not impossible.

• The Contenders:
  • Co-annihilating WIMPs: Particles that help each other disappear at just the right rate.
  • Early Matter Domination Black Holes: Black holes formed when the universe was filled with a heavy, slow-moving fluid.
  • First-Order Phase Transition Black Holes: Black holes formed when the universe "froze" like water turning to ice, creating bubbles that collapsed.
  • Note: These are all roughly equally difficult to tune, regardless of whether they are particles or black holes.

Tier 3: The "Hardcore Mode" (Highly Tuned)

These are the "pencil on a tip" scenarios. You need to adjust the settings to a fraction of a percent to work.

• The Strugglers:
  • Higgs-Funnel WIMPs: A particle that only works if it hits a very specific "resonance" (like a radio tuned to exactly one frequency). If you miss by a hair, it fails.
  • Single-Field Ultra-Slow-Roll Black Holes: This is the specific black hole model critics usually complain about. It requires a "double exponential" sensitivity. Imagine a machine where turning a knob changes the output by a factor of 10, but that knob itself is controlled by another knob that changes the output by another factor of 10. It's a "tuning nightmare."

The Big Takeaways

1. "Fine-Tuning" isn't about what the dark matter is; it's about how it's made.
The paper proves that the difficulty of the blueprint depends on the math of the formation process, not whether the result is a particle or a black hole.

• You can have natural black holes (Tier 1).
• You can have tuned particles (Tier 3).
• You can have tuned black holes (Tier 3).
• You can have natural particles (Tier 1).

2. The "Black Hole" reputation is unfair.
The claim that "Black Holes are always fine-tuned" is wrong. It conflates the worst-case black hole scenario (Tier 3) with the best-case scenario (Tier 1). The "Biased Domain Wall" black holes are actually some of the most natural candidates in the entire universe.

3. The "Two-Layer" Problem for Inflationary Black Holes.
For the black holes formed during "Inflation" (the rapid expansion of the early universe), the tuning is hard for two reasons stacked on top of each other:

• Layer 1: Getting the black holes to form at all.
• Layer 2: Getting the "Inflation" engine to produce the exact right conditions to trigger Layer 1.
  This double-layer problem makes these specific black holes very hard to tune, but it's a specific problem with that model, not all black holes.

Summary Analogy

Imagine you are trying to bake a cake that weighs exactly 1 pound.

• Tier 1 (Natural): You have a recipe where the ingredients are in a simple ratio. If you add a cup of flour or a cup of sugar, the weight changes predictably. It's easy to hit 1 pound. (This includes Biased Domain Wall Black Holes).
• Tier 2 (Medium): You have a recipe where the oven temperature matters a lot. If you are off by 10 degrees, the cake is too light or too heavy. You need to be careful, but it's doable. (This includes Early Matter Domination Black Holes).
• Tier 3 (Hard): You have a recipe where the cake rises only if you tap the table at exactly the right frequency while pouring the batter. If you miss by a millisecond, the cake is flat. (This includes Higgs-Funnel Particles and Single-Field Inflation Black Holes).

The paper's conclusion: Don't dismiss the "Black Hole" cake just because one specific recipe (Tier 3) is impossible to bake. There is another black hole recipe (Tier 1) that is just as easy to bake as the best particle recipes.

Technical Summary

Technical Summary: "Are Primordial Black Holes a Natural Dark Matter Candidate?"

Problem Statement
The identity of dark matter remains an open question, with candidates ranging from Weakly Interacting Massive Particles (WIMPs) and axions to Primordial Black Holes (PBHs). While PBHs in the asteroid-mass window ($10^{17} \lesssim M_{\text{PBH}}/\text{g} \lesssim 10^{22}$) can account for all dark matter without violating observational constraints, they are frequently dismissed as "fine-tuned." This objection arises because PBH production in inflationary models typically requires the primordial curvature power spectrum to be exponentially sensitive to inflaton potential parameters. However, this qualitative dismissal has rarely been subjected to the same quantitative scrutiny applied to particle dark matter candidates, where similar exponential sensitivities (e.g., in thermal freeze-out) are often accepted as part of the "WIMP miracle." Furthermore, existing literature lacks a unified framework comparing PBH and particle dark matter fine-tuning, and non-inflationary PBH mechanisms have not been analyzed quantitatively.

Methodology
The author applies the Barbieri–Giudice (BG) fine-tuning measure uniformly across a broad landscape of dark matter scenarios to test the claim that PBHs are generically fine-tuned. The BG measure, $\Delta \equiv \max_i |\partial \ln \Omega / \partial \ln x_i|$, quantifies the sensitivity of the dark matter relic abundance ($\Omega h^2$) to fractional variations in fundamental input parameters $x_i$.

The study evaluates:

1. Three non-inflationary PBH mechanisms: Biased domain walls, early matter domination, and first-order phase transitions (FOPT).
2. Six classes of inflationary PBH models: Including curvaton/spectator-field, multi-field stochastic, hybrid, single-field with features, and single-field ultra-slow-roll (USR)/inflection-point models.
3. Seven particle dark matter benchmarks: Covering thermal-relic WIMPs (off-resonance, Higgs-funnel, coannihilation), freeze-in particles (FIMPs), asymmetric dark matter (ADM), and axions (misalignment and post-inflationary).

All scenarios are evaluated against the same observable target: the Planck 2018 measurement $\Omega_{\text{DM}}h^2 = 0.120$. The analysis employs a "two-layer" decomposition for inflationary models to separate the sensitivity of the PBH abundance to the power spectrum amplitude ($\sigma$) from the sensitivity of $\sigma$ to the inflaton potential coefficients. Three complementary measures are used: the BG measure (worst-case direction), the Strumia–Rattazzi (SR) measure (quadrature sum), and the island half-width ($\epsilon$).

Key Contributions and Results

1. Three Universality Classes of Naturalness
The analysis reveals that the fine-tuning of any dark matter construction is determined by the analytic structure of its abundance map, not by whether the candidate is a particle or a gravitational relic. Three distinct tiers emerge:

• Class I (Natural, $\Delta \lesssim 5$): Characterized by pure power-law abundance maps with no exponential factors. This tier includes Asymmetric Dark Matter ($\Delta=1$), post-inflationary axions ($\Delta=1.19$), off-resonance WIMPs ($\Delta=2$), freeze-in particles ($\Delta=2$), and biased-domain-wall PBHs ($\Delta=2$ for gravity-induced bias, $\Delta=4.5$ for free bias).
• Class II (Intermediate, $\Delta \approx 15\text{--}50$): Characterized by a single exponential factor in the abundance map. This tier unifies PBH early matter domination ($\Delta \approx 14$), coannihilating WIMPs ($\Delta \approx 48$), and FOPT-PBHs (within a single-exponential approximation, $\Delta \approx 35$). The paper derives a structural identity for this class: $\Delta \approx \ln(T_{\text{form}}/T_{\text{eq}})$, where the tuning is set by the ratio of the formation scale to the matter-radiation equality scale.
• Class III and Beyond (Highly Tuned, $\Delta \gtrsim 10^3$): Characterized by resonance/cancellation mechanisms or double-exponential structures. This includes Higgs-funnel WIMPs ($\Delta \approx 6500$) and single-field inflationary PBH models.

2. The Status of Primordial Black Holes
The paper demonstrates that the PBH paradigm spans the entire naturalness landscape, contradicting the notion that PBHs are generically fine-tuned:

• Biased Domain Walls: These are shown to be as natural as the most natural particle dark matter candidates. The abundance map is a pure monomial ($f_{\text{PBH}} \propto \eta^{9/2} V_b^{-2}$), resulting in a constant, low fine-tuning measure ($\Delta = 4.5$ or $\Delta=2$ for gravity-induced bias).
• Inflationary Collapse: The fine-tuning depends critically on the reheating history and the model class.
  • For standard high-reheating scenarios ($T_{\text{reh}} \gg T_{\text{form}}$), all inflationary classes fall into Class III or beyond due to the radiation-dominated collapse threshold ($\Delta^{(\text{RD})}_\sigma \approx 57\text{--}70$) combined with the sensitivity of the power spectrum to potential coefficients.
  • For low-reheating scenarios ($T_{\text{reh}} \lesssim T_{\text{form}}$), curvaton and spectator-field models drop to Class II ($\Delta \approx 21\text{--}42$), comparable to coannihilating WIMPs.
  • Single-field USR/inflection-point models remain in Class III or beyond ($\Delta \gtrsim 10^3$) regardless of reheating, due to a "double-exponential" structure where the power spectrum amplitude is exponentially sensitive to potential coefficients, which are themselves exponentially sensitive to the collapse condition.

3. Resolution of Literature Tensions
The paper resolves the apparent tension between previous studies (e.g., Iovino and Riotto) that claimed inflationary PBHs are natural and those finding them highly tuned. The authors argue that the BG measure (evaluated on the $f_{\text{PBH}}=1$ contour) and Wilson's naturalness criterion answer complementary questions. The BG measure correctly captures the local sensitivity required to maintain the observed abundance, while the Wilson criterion addresses the existence of a technically natural UV completion. The two-layer decomposition clarifies that the "naturalness" of the UV completion does not negate the fine-tuning required to map those parameters to the specific observed abundance.

Significance
The paper's primary conclusion is that the claim "PBH dark matter is fine-tuned" conflates the worst-case scenario (single-field inflationary collapse) with a landscape that includes genuinely natural mechanisms (biased domain walls). The study establishes that fine-tuning is a property of the abundance map's analytic structure rather than the nature of the dark matter candidate.

Specifically, the work:

• Vindicates biased-domain-wall PBHs as a robustly natural candidate, occupying the same naturalness tier as off-resonance WIMPs and freeze-in particles.
• Quantifies the hierarchy within the PBH paradigm, showing a span of over seven orders of magnitude in fine-tuning ($\Delta \approx 2$ to $\Delta \gtrsim 10^{10}$), a range larger than that of all particle dark matter scenarios considered.
• Provides a unified framework for comparing disparate dark matter production mechanisms, demonstrating that the "WIMP miracle" and the "PBH miracle" (in specific mechanisms) are subject to the same quantitative naturalness criteria.
• Identifies multi-messenger testability for FOPT-PBHs, noting that the nucleation rate parameter required for the correct abundance ($\beta/H_* \sim 5\text{--}15$) falls within the sensitivity range of future gravitational-wave observatories like LISA, offering an independent test of the mechanism.

The analysis concludes that while single-field inflationary PBHs are indeed highly tuned, other PBH mechanisms are not, and the dismissal of the entire PBH paradigm based on fine-tuning arguments is not supported by a quantitative, cross-paradigm comparison.