Two-Higgs simulations reveal unusually durable cosmic domain walls

Numerical simulations of a symmetry-restricted two-Higgs-doublet model have produced an unexpected type of cosmic defect: closed domain walls that slow their collapse sharply and can, in an idealized classical limit, survive for cosmological timescales. The researchers identify the remnants as high-amplitude oscillons, localized configurations in which the fields oscillate inside one vacuum while the exterior remains static in a second, degenerate vacuum. Under a special parameter choice, radiation from the structures is strongly suppressed, making their lifetime effectively unbounded in the ideal limit.

The result comes from a theoretical study of the ℤ₂-symmetric two-Higgs-doublet model, an extension of the Standard Model containing two complex Higgs doublets rather than one. The paper is a classical field-theory calculation, not a laboratory measurement or a prediction of a detected cosmic signal. Its simulations use natural units, with masses, energies, lengths and times expressed relative to a 125 GeV Higgs mass, and examine both two- and three-dimensional versions of the model. The authors say further analysis in an expanding Friedmann–Robertson–Walker background, as well as a treatment of quantum effects, remains necessary.

Domain walls can form when a symmetry-breaking phase develops and different regions of the universe settle into distinct, physically distinguishable vacuum states. In the model studied here, the sign of a bilinear combination of the two Higgs fields separates the two vacuum manifolds, with zero marking the wall between them. Such walls normally contract because of their curvature and collapse as the surrounding regions become uniform. In the new simulations, however, the total wall length stops following that straightforward decay pattern once the network breaks into isolated closed loops. The authors report that this departure from the expected contraction rate signals a transition to stable, high-amplitude oscillon remnants.

The simulations began with uniformly distributed random values assigned to the field components on a spatial grid. A damping term was applied during the initial iterations to simulate the cooling of an initially hot, dense and approximately homogeneous universe. The researchers then evolved the equations of motion in flat spacetime with a fourth-order Runge–Kutta method. In the first large-scale survey, periodic boundaries supplied a dense background of radiation particles and stochastic perturbations, as might occur in the early universe. The resulting wall structures remained at nonzero length for extended periods, with the decay rate falling markedly after roughly 50 time units in two dimensions and 250 in three dimensions. Because three-dimensional runs are computationally costly, most subsequent quantitative comparisons used two-dimensional calculations.

Separate trials defined a bubble lifetime beginning when the wall network became too short to wrap around the periodic simulation box and ended when its length first reached zero. Trials with no defined bubble lifetime were excluded so that stripe modes would not be counted. Of 300 independent runs for each system, only trials that ultimately reached a single vacuum were included, amounting to roughly 60 percent of the runs. Long-lived bubbles occurred frequently in the full eight-field two-Higgs-doublet model and less often in a four-field system containing two complex scalars. They were almost absent in a two-field reference model containing two real scalars, suggesting that two complex fields, and the unitary symmetry they carry, are required for the phenomenon seen in these simulations.

To isolate the mechanism, the researchers turned to a four-real-field theory with two complex scalars. This was not presented as a truncated version of the two-Higgs-doublet model, but as an independent toy model designed to retain the essential nonlinear structure without gauge degrees of freedom. They then found a parameter regime in which third and higher harmonics of the oscillation cancel exactly. The equations for the static amplitude and the oscillating amplitude consequently close on the fundamental frequency, suppressing the higher-order terms that would otherwise radiate energy away. In this idealized limit, the solution has an effectively infinite lifetime, and the researchers report that the computer modeling is consistent with that prediction.

The resulting oscillon exists as a continuous one-parameter family rather than a configuration with a uniquely favored radius. Its frequency can range from zero to a maximum value determined by the vacuum mass, and each frequency produces a different radial profile. At low frequency, the solution resembles a broad domain wall separating extended regions in opposite vacua. As the frequency rises, the oscillating region narrows and its amplitude changes, eventually approaching a small-amplitude oscillation around a single vacuum. As the frequency approaches zero, the interior of the opposite vacuum becomes arbitrarily large, linking the oscillon continuously to a large closed wall. The researchers therefore interpret the long-lived remnants of wall decay as the high-amplitude end of this continuous family.

The configuration also has the properties of a Q-ball. By forming linear combinations of the two complex fields, the static part can be separated from a component that rotates uniformly in phase. That rotating component carries a conserved global U(1) charge, with a local density proportional to the oscillation frequency and the squared oscillating amplitude. The charge is distributed through the bubble’s interior rather than concentrated on its surface. This distinguishes the time-dependent oscillons from static two-Higgs-model domain walls that contain charge-breaking condensates, although the authors argue that both involve localized charge accumulation within the model’s vacuum structure. The conserved charge provides an analytical explanation for why the oscillating configuration can remain localized rather than dispersing.

Stability was tested with Floquet analysis, which follows small perturbations over a complete oscillation period and examines whether their amplitudes grow or decay. For a representative frequency of 0.45 in the dimensionless units used by the study, no radial perturbation mode had a Floquet multiplier above one within a stated uncertainty of 10⁻⁵. The same stability was reported across the frequency family. Angular perturbations were also assessed using effective potentials that include centrifugal barriers: for angular momentum greater than zero, these barriers confine the modes more strongly than radial motion, so the demonstrated radial stability is taken to support stability in the other angular sectors. Repeating the analysis for the complete eight-field model produced the same linear-stability conclusion.

The special suppression condition corresponds to full mixing, expressed in the paper as tan a = tan b = 1, while the vacuum masses and tan β remain free along a continuous locus in parameter space. For the full-mixing example, the authors use μ² = 0.5, λ = 1.78 and λ₄ = −2.56, with Mₕ set to one. They embed the oscillating profile into the two complex doublets and again find that the radiation-producing higher harmonics cancel. Although the mathematical structure is stable in the classical simulations, its precise topological protection remains unresolved. The authors also caution that quantum behavior may differ from classical wave motion and that the investigated parameter choices need not be favored by collider constraints.

Because no cosmological evolution, production history or observational signature was calculated, the findings do not establish that such walls formed in the real universe or survive today. The authors nevertheless identify dark matter, stochastic gravitational-wave backgrounds and possible laboratory analogues in cold-atom or condensed-matter systems as subjects for future investigation. The immediate advance is a mechanism by which ordinary domain-wall collapse can leave a charge-carrying, radiation-suppressed remnant rather than disappear completely. In the model’s ideal classical limit, that remnant can outlast cosmic times; determining whether it can form in a realistic early universe and survive quantum effects requires further work.

Subject of Research: Classical simulations of radiation-suppressed oscillon domain walls in a ℤ₂-symmetric two-Higgs-doublet model

Article Title: Long-lived oscillons and domain walls in the \(\mathbb {Z}_{2}\)-symmetric two-Higgs-doublet model

Article References: Meng, Z. (2026). Long-lived oscillons and domain walls in the $$\mathbb {Z}_{2}$$-symmetric two-Higgs-doublet model. The European Physical Journal C, 86(10), Article 1191. https://doi.org/10.1140/epjc/s10052-026-16467-5

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16467-5

Keywords: two-Higgs-doublet model, domain walls, oscillons, Q-balls, topological defects, early universe, Higgs fields, radiation suppression, Floquet analysis, cosmology, Long-lived, domain