Dark Matter Production from Bubble Collisions during a First-Order Phase Transition at the End of Inflation
Zihong Cheng, Fa Peng Huang · 2026
Read Dark Matter Production from Bubble Collisions during a First-Order Phase Transition at the End of Inflation by Zihong Cheng, Fa Peng Huang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-03758.txt.
Introduction
Zihong Cheng and Fa Peng Huang study whether dark matter could have been produced by bubble collisions during a first-order phase transition at the end of inflation. The transition happens in a spectator scalar sector whose potential depends on the inflaton, so bubbles form faster as inflation ends. The authors first map where vacuum decay is led by the Coleman-De Luccia channel and the transition completes.
In the viable regime, elastic self-scatterings spread the spectator particles' momenta efficiently, while their decay into dark matter is the main channel moving the population into the dark sector. The final abundance combines direct production in bubble collisions with later decays of spectator particles. Competing number-changing processes stay too slow compared with cosmic expansion.
Catalog record
| CREATOR | Zihong Cheng, Fa Peng Huang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 104 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.03758
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Dark Matter Production from Bubble Collisions during a First-Order Phase Transition at the End of Inflation Zihong Cheng, Fa Peng Huang Abstract We study whether a first-order phase transition at the end of inflation can generate the observed dark matter abundance through bubble collisions. The transition occurs in a spectator scalar sector with an inflaton-dependent effective potential, so that the nucleation rate grows during inflation and becomes significant only near its end. We identify the region of parameter space in which vacuum decay is dominated by the Coleman–De Luccia channel, the Hawking–Moss transition remains subdominant, and the nucleated bubbles admit a consistent physical interpretation in an inflating background. Requiring also that the phase transition completes successfully, we then analyze particle production from bubble collisions. In the viable regime, elastic self-scatterings of the spectator particles can efficiently redistribute their momenta, while their decay into dark matter provides the dominant channel for transferring the spectator population to the dark sector. Other competing number-changing or sink processes remain inefficient compared with the Hubble expansion. The final relic abundance receives contributions from both direct production in bubble collisions and the subsequent decay of spectator field particles. We find that the observed dark matter abundance can be accommodated, within the order-of-magnitude accuracy of the collision-production treatment, in a restricted region of parameter space. 1 Introduction The particle nature and cosmological origin of dark matter (DM) remain among the central open questions in particle physics and cosmology 32 ; 25 ; 71 ; 21 ; 24 . While thermal freeze-out 48 ; 9 ; 77 and freeze-in 74 ; 52 ; 20 ; 49 provide the standard benchmarks for generating the DM relic abundance, the increasingly extensive experimental searches, together with the lack of positive signals in direct detection, indirect detection, and collider experiments, have motivated sustained interest in alternatives to the simplest thermal weakly interacting massive particle (WIMP) paradigm. In this context, cosmological first-order phase transitions (FOPTs) offer a particularly rich setting for DM genesis. FOPT can generate the DM abundance or substantially modify its final value through several distinct channels. Ultra-relativistic bubble walls provide a non-thermal mechanism for producing particles far heavier than the phase transition scale, whereas bubble–plasma collisions furnish an additional efficient mechanism for populating secluded dark sectors 14 ; 12 ; 11 . In addition, FOPTs may produce asymmetric DM or macroscopic non-thermal relics, such as (gauged) Q-ball and Fermi-ball DM 69 ; 57 ; 51 ; 55 ; 60 ; 62 ; 63 . Distinct from these scenarios, bubble wall filtering can also set the relic abundance of particle DM: if the DM mass changes across the wall, only sufficiently energetic particles can enter the broken phase, while the rest are reflected and depleted, realizing the filtered DM mechanism 13 ; 31 ; 61 . Against this broader backdrop, bubble collisions constitute another especially compelling channel, as they can directly convert the energy stored in the expanding bubble walls into DM particles and thereby provide a distinct non-thermal production mechanism 86 ; 40 ; 73 ; 47 ; 30 ; 42 . Among the possible realizations of this mechanism, a particularly intriguing one is that the FOPT takes place already during inflation rather than after reheating. In such a setup, an inflaton-induced FOPT in a spectator sector can produce superheavy DM far from equilibrium, as pointed out in Ref. 8 . Motivated by this possibility, in this work we study DM production from a FOPT during inflation in greater detail. A FOPT proceeds through the nucleation of true vacuum bubbles inside a metastable false vacuum. In the scenario considered here, however, the transition occurs during inflation, so the bubbles expand in an approximately empty vacuum rather than in a thermal plasma. As a consequence, plasma friction is absent and the bubble walls can continue to accelerate until they enter an ultrarelativistic runaway regime before collision 8 ; 30 ; 39 ; 67 ; 23 . The vacuum energy stored in these highly boosted walls is then released in bubble collisions, efficiently exciting fields coupled to the spectator sector 40 ; 73 ; 47 ; 30 . If the spectator field is coupled to a stable dark sector particle, these non-thermal processes can provide an efficient source of DM. Our focus is therefore on the regime of an inflationary vacuum with runaway bubbles, in which DM production is governed by vacuum bubble dynamics rather than by thermal scattering after reheating 8 . This focus immediately raises a first central issue: in an inflationary background, the vacuum decay problem is qualitatively different from its familiar flat-spacetime counterpart. In Minkowski space, false vacuum decay is described by the semiclassical bounce formalism of Refs. 34 ; 28 , with explicit quartic potential solutions given in Ref. 2 . Once gravitational effects are included, however, the relevant tunneling process must be analyzed in an approximately de Sitter background, where the structure of Euclidean saddle points can differ substantially from that in flat spacetime 33 ; 79 . Depending on the shape of the barrier and its relation to the Hubble scale, a Coleman–De Luccia (CDL) bounce may dominate, may coexist with the Hawking–Moss solution 87 ; 53 , or may even cease to exist 15 ; 66 ; 85 . This distinction is crucial for the present work, because the DM production mechanism of interest requires a genuine localized bubble nucleation event, followed by bubble expansion and subsequent collisions, rather than a homogeneous transition over the barrier. At the same time, identifying the physically relevant tunneling saddle is subtle in de Sitter space, since some Euclidean solutions possess additional negative modes and therefore cannot be straightforwardly interpreted as the tunneling saddle 70 ; 65 ; 19 ; 26 . We therefore analyze vacuum decay directly in de Sitter space, rather than relying on the flat-space approximation, and identify the relevant saddle in the parameter region of interest. In particular, we restrict attention to the parameter regime in which localized bubble nucleation is the relevant semiclassical description of the transition. A second central issue is whether the phase transition can occur during inflation without either disrupting the inflationary background too early or failing to complete near its end. If bubble nucleation becomes efficient too early, large bubbles may form while inflation is still ongoing and generate unacceptable inhomogeneities. Conversely, if the nucleation rate rises too slowly, then even if bubbles are eventually produced, accelerated expansion can separate them faster than they can grow and collide with one another, thereby preventing efficient percolation and completion of the phase transition 84 ; 38 . A viable scenario therefore requires the nucleation rate to remain sufficiently suppressed throughout most of the inflationary era, and then to increase rapidly near the end of inflation so that bubbles nucleate abundantly and collide efficiently within a Hubble time. In practice, this imposes nontrivial constraints on the time dependence of the tunneling exponent and on the dimensionless parameter \beta/H , which measures the timescale over which the nucleation rate rises relative to the Hubble expansion. A third essential aspect of the problem is the connection between particle production at the phase transition and the present-day DM abundance. It is not sufficient to show