⚛️ phenomenology

Spectator Dark Matter and the Higgs Portal

This paper proposes that an extremely feeble Higgs portal interaction can suppress the catastrophic overproduction of light scalar spectator dark matter generated during inflation by inducing early oscillations and reducing field variance, thereby allowing sub-GeV dark matter to constitute the observed relic abundance without thermalizing the spectator.

Original authors: Stephen E. Henrich, Yann Mambrini, Keith A. Olive

Published 2026-10-07

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Stephen E. Henrich, Yann Mambrini, Keith A. Olive

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

Dark matter is the invisible substance that holds galaxies together, yet we have never seen it directly. For decades, physicists have assumed that this mysterious material was created long after the universe began, born from the hot, dense soup of particles that filled the early cosmos. However, a new line of thinking suggests that if dark matter is made of a specific kind of light, invisible particle, it might have been created much earlier, during the violent birth of the universe itself. This idea, known as the "spectator" scenario, proposes that these particles were present from the very first moments of cosmic expansion. The problem is that if they were created then, the math suggests there should be far too much of them today—so much that they would have crushed the universe under their own weight. A team of physicists has now found a way to resolve this paradox, showing how a tiny, almost invisible connection to the known particles of our world could have naturally thinned out this excess, leaving behind just the right amount of dark matter to exist today.

The story begins with the theory of inflation, a period of incredibly rapid expansion that occurred a fraction of a second after the Big Bang. During this time, the universe stretched so fast that quantum fluctuations—tiny, random jitters in energy—were blown up to cosmic scales. If dark matter consists of a light, scalar particle (a type of field that has a value at every point in space but no direction), these fluctuations would have created a vast, uniform sea of the particle, essentially a "condensate," filling the universe before it even cooled down. In the simplest version of this story, where the particle has no special interactions, this sea would have remained dense and heavy, leading to a catastrophic overproduction of dark matter. The amount predicted would be billions of times greater than what we actually observe, making the existence of stars and galaxies impossible.

To solve this, the researchers, Stephen Henrich, Yann Mambrini, and Keith Olive, investigated a mechanism where this dark matter particle interacts very weakly with the Higgs field, the field responsible for giving other particles their mass. They call this the "Higgs portal." Even though the connection is incredibly faint—so weak that it is far below the threshold where it would normally create new particles—it has a profound effect on the history of the dark matter. The researchers found that this portal does two things. First, it creates a tiny self-interaction for the dark matter particle. Second, and more importantly, it gives the particle a "thermal mass" when the universe is hot. This thermal mass is not a permanent weight but a temporary one that depends on the temperature of the surrounding universe.

The key discovery is that this thermal mass can trigger the dark matter to start moving and oscillating much earlier than it would have on its own. Imagine a heavy object sitting in a thick fluid; if the fluid suddenly becomes more viscous, the object might start to sink or shift sooner than expected. In this cosmic scenario, the thermal mass acts like that fluid, causing the dark matter condensate to begin its rhythmic motion while the universe is still very young and hot. Once this motion starts, the energy of the dark matter begins to dilute, or spread out, much faster than it would if it were just sitting still. Instead of fading away slowly like ordinary matter, the energy density of this dark matter drops off rapidly as the universe expands. This rapid dilution happens because the thermal mass itself is shrinking as the universe cools, creating a feedback loop that drains the energy from the condensate efficiently.

The researchers calculated that this process can reduce the initial, overwhelming abundance of dark matter to the precise, tiny amount we see today. They explored different scenarios, including cases where the universe expanded for a very long time during inflation and cases where it was shorter. In the longer scenarios, the dark matter reaches a stable, balanced state of fluctuations before inflation ends, and the Higgs portal still manages to thin it out effectively. In the shorter scenarios, where the fluctuations are smaller to begin with, the mechanism still works, though it requires specific conditions to be met. A major constraint in their work is the "isocurvature" limit, which is a rule derived from observations of the cosmic microwave background. This rule says that if inflation was too short, the random fluctuations would leave a detectable imprint on the early universe that we simply do not see. The researchers found that for their mechanism to work, inflation must have lasted at least ten billion cycles of expansion, a number that fits comfortably within current observational limits.

The study also considered a modification involving gravity. By adding a slight, non-standard interaction between the dark matter and the curvature of spacetime itself, the researchers found that the initial fluctuations of the dark matter could be suppressed even further. This change allows the thermal mass to take over the role of starting the oscillations even more decisively, opening up a much wider range of possible masses and interaction strengths for the dark matter. This suggests that the dark matter could be much lighter than previously thought, potentially weighing as little as a fraction of an electron volt, or as heavy as ten billion electron volts, while still fitting the model.

Crucially, the researchers showed that this mechanism does not rely on the dark matter ever becoming part of the hot thermal bath of the early universe in the traditional sense. It does not "freeze out" like a gas turning to ice, nor does it "freeze in" by being slowly created from collisions. Instead, the Higgs portal acts as a drain, siphoning off the excess energy stored in the primordial condensate. The dark matter remains a coherent, classical wave-like field that never fully breaks apart into individual particles, yet its energy density is reduced to the correct level by the time the universe is old enough to form galaxies.

The findings suggest that the Higgs portal, often studied as a way to create dark matter, might actually be more important for destroying the overabundance of dark matter created during inflation. This offers a new perspective on how the invisible sector of the universe connects to the visible one. The work rules out the idea that a simple, non-interacting light scalar particle could be the dark matter, as it would inevitably lead to a universe that is too dense. It also argues against the notion that the dark matter must be heavy or interact strongly to be viable. Instead, the paper points to a delicate balance where a feeble, almost imperceptible link to the Higgs field is sufficient to save the universe from being overwhelmed by its own dark matter, leaving behind the quiet, stable cosmos we inhabit today.

Technical Summary: Spectator Dark Matter and the Higgs Portal

Problem Statement
Light scalar spectator fields generically acquire large stochastic fluctuations during cosmic inflation. On super-Hubble scales, these fluctuations behave as a nearly homogeneous classical condensate. If this primordial condensate survives until late times without significant suppression, it leads to a catastrophic overproduction of dark matter (DM). For a standard quadratic spectator with a weak-scale mass and a typical inflationary scale ( $H_{end} \sim 10^{13}$ GeV), the resulting relic abundance exceeds the observed value by approximately 30 orders of magnitude. While self-interactions (quartic terms) or specific inflationary durations (finite $N$ ) can mitigate this, they often require fine-tuned parameters or conflict with isocurvature constraints. The central question addressed is whether a minimal, gauge-invariant interaction connecting the spectator to the visible sector—specifically the Higgs portal—can sufficiently suppress this inflationary memory without thermalizing the spectator or destroying the condensate.

Methodology
The authors analyze a model where a light scalar singlet $S$ couples to the Standard Model Higgs doublet $H$ via the potential:
$V(S, H) \supset \frac{1}{2}m_S^2 S^2 + \frac{\lambda_S}{4}S^4 + \frac{\lambda_{HS}}{2}S^2 H^\dagger H$
The study focuses on "sub-freeze-in" couplings ( $10^{-28} \lesssim \lambda_{HS} \lesssim 10^{-9}$ ), where the spectator does not thermalize. The analysis incorporates:

Effective Mass Hierarchy: The effective mass squared, $m_{S,eff}^2 = m_S^2 + 3\lambda_S S^2 + c_S T^2 + \lambda_{HS}\langle H^\dagger H \rangle$ , determines the onset of oscillations. The authors consider contributions from the bare mass, radiatively induced quartic self-coupling ( $\lambda_S \approx \lambda_{rad} \propto \lambda_{HS}^2$ ), and thermal mass ( $c_S T^2 \propto \lambda_{HS} T^2$ ).
Inflationary Dynamics: The paper treats two regimes for the duration of inflation ( $N$ $N$ ):
- Stochastic Equilibrium: $N$ is sufficiently long for fluctuations to saturate at their equilibrium variance.
- Finite $N$ : $N$ is shorter than the equilibration time, leaving fluctuations in a random-walk regime.
Joint Stochastic Evolution: For the equilibrium case, the authors perform a two-field stochastic evolution of $S$ and the Higgs field $h$ during inflation, rather than treating $S$ as an isolated field. This accounts for the modification of the initial fluctuation amplitude by the portal coupling.
Non-Minimal Gravity: The model is extended to include a non-minimal gravitational coupling $\xi S^2 R$ , which alters the inflationary mass and fluctuation amplitude.

Key Contributions and Mechanisms
The paper identifies a novel mechanism where the Higgs portal suppresses the relic abundance not by producing dark matter, but by draining the energy density of a pre-existing inflationary condensate. This occurs through two distinct effects induced by the portal coupling:

Radiative Self-Interaction: Even if the bare quartic coupling is zero, Higgs loops generate a radiative quartic term ( $\lambda_{rad} \propto \lambda_{HS}^2$ ). If this term dominates the effective mass, it triggers early oscillations. In the quartic regime, the energy density redshifts as $\rho_S \propto a^{-4}$ (during radiation domination) or steeper, significantly faster than the standard matter-like $\rho_S \propto a^{-3}$ .
Thermal Mass: The coupling to the thermal bath generates a temperature-dependent mass term ( $m_{th} \propto T$ $m_{t h} \propto T$ ). This term can dominate the effective mass at high temperatures, triggering oscillations much earlier than the bare mass would. Crucially, because $m_{th}$ $m_{t h}$ decreases as the universe expands ( $T \propto a^{-1}$ $T \propto a^{- 1}$ or $a^{-3/8}$ $a^{- 3/8}$ ), the condensate undergoes enhanced dilution:
- During radiation domination: $\rho_S \propto a^{-4}$ .
- During reheating (matter-like): $\rho_S \propto a^{-27/8}$ .

The paper delineates four possible oscillation sequences based on the hierarchy of these mass terms:

Bare mass only.
Quartic $\to$ Bare mass.
Thermal $\to$ Bare mass.
Quartic $\to$ Thermal $\to$ Bare mass.

Results

Isolated Spectators: Without the Higgs portal, achieving the correct relic density for a quadratic spectator requires either extremely large self-couplings ( $\lambda_S \sim 1$ ) or very low reheating temperatures combined with specific finite- $N$ scenarios. However, finite- $N$ scenarios are severely constrained by isocurvature limits ( $N \gtrsim 10^{10}$ ), which often forces the system back into the overproduction regime.
Higgs Portal (Standard Gravity):
- Scenario 3 (Thermal $\to$ Bare): In the standard model, this scenario is generally ruled out. The isocurvature constraint requires large $N$ , which implies large initial fluctuations. The thermal mass alone cannot dilute this large condensate sufficiently to match the observed abundance without violating the mass bound imposed by the Higgs vacuum expectation value (VEV).
- Scenario 4 (Quartic $\to$ Thermal $\to$ Bare): This is the primary viable solution. The radiative quartic term triggers the initial oscillations, followed by a transition to thermal oscillations, and finally bare mass oscillations. This sequence allows for the correct relic abundance across a wide parameter space. The viable region favors sub-GeV dark matter ( $0.1 \text{ eV} \lesssim m_S \lesssim 10 \text{ GeV}$ ) and generally prefers low reheating temperatures ( $T_{RH} \lesssim 100 \text{ GeV}$ ) in the absence of non-minimal gravity.
- Scaling Laws: For finite $N$ , the relic density contour scales as $m_S \propto \lambda_{HS}$ . For stochastic equilibrium, it scales as $m_S \propto \lambda_{HS}^{3/2}$ .
Higgs Portal (Non-Minimal Gravity):
- Adding a non-minimal coupling $\xi S^2 R$ dramatically alters the results. The curvature term dominates the effective mass during inflation, suppressing the initial fluctuation amplitude ( $\langle S^2 \rangle \propto 1/\xi$ ).
- This suppression prevents the radiative quartic term from ever dominating the oscillation onset. Instead, the thermal mass directly triggers the oscillations (restoring the viability of Scenario 3).
- This modification expands the parameter space significantly, allowing for higher reheating temperatures (up to $10^{12}$ GeV) and larger portal couplings ( $\lambda_{HS}$ ) and masses ( $m_S$ ) compared to the minimal case. The isocurvature constraints are naturally satisfied for $\xi \gtrsim 0.026$ .

Significance
The paper claims that the Higgs portal plays a fundamentally different cosmological role than standard freeze-out or freeze-in mechanisms. Instead of generating the dark matter abundance, the portal acts as a "drain" that suppresses a pre-existing inflationary condensate. This mechanism:

Solves the catastrophic overproduction problem for light scalar spectators using extremely feeble couplings ( $\lambda_{HS} \lesssim 10^{-9}$ ).
Provides a natural initial condition for standard Higgs portal freeze-in, as the portal simultaneously dilutes the primordial condensate and eventually allows thermal production to dominate at large couplings.
Extends the viable mass range for spectator dark matter down to sub-eV scales, well below the typical GeV scale emphasized in previous literature.
Demonstrates that the interplay between radiative self-interactions and thermal masses can qualitatively alter the evolution of the spectator condensate, offering a robust solution to the inflationary memory problem even when the spectator remains non-thermalized.

More like this