Bubble Spacetimes: Interfaces in Curved Spacetime

• Bubble spacetimes are geometries in which a compact region exhibits a different metric, topology, or causal structure from its surrounding spacetime.
• They are used to model phenomena like false-vacuum decay, AdS solitons, and wormhole analogues, highlighting diverse cosmological and string-theoretic applications.
• Key studies use junction conditions, Raychaudhuri equations, and numerical simulations to explore the dynamics, stability, and thermodynamics of these localized spacetimes.

Bubble spacetimes are geometries in which a compact region differs significantly from the surrounding spacetime, typically via a different metric glued to an exterior background, and, in a common formulation, a composite spacetime

$$\mathcal M=\mathcal M^+ \cup \mathcal M^- \quad\text{with a common boundary}\quad \Sigma$$

(Tippett et al., 2013, Haque et al., 2017). In the literature surveyed here, this usage includes false-vacuum and Coleman–De Luccia bubbles, AdS shellworlds and AdS solitons, stringy cosmic-brane bubbles, topological baby-universe bubbles, wormhole-like bubble universes, and low-regularity causal bubbles (Berglund et al., 2021, El-Menoufi et al., 2013, Cai et al., 2024, Lemos et al., 2022, García-Heveling et al., 2022). The common theme is a localized interface, cap, or causal region whose intrinsic geometry, topology, or causal structure differs from that of the ambient spacetime.

1. Definitions, taxonomy, and geometric archetypes

A broad geometric definition appears in the TARDIS construction: bubble spacetimes are geometries in which a compact region differs significantly from the surrounding spacetime, typically via a different metric glued to an exterior background, often Minkowski (Tippett et al., 2013). Standard examples listed there include Alcubierre warp bubbles, false vacuum or Coleman–De Luccia bubbles, and Krasnikov tubes, while the specific construction considered is a compact region of nontrivial curvature embedded in an otherwise flat Minkowski background, with a Rindler interior and a flat exterior (Tippett et al., 2013).

A complementary cosmological definition treats a bubble spacetime as a spherically symmetric, homogeneous and isotropic region $\mathcal M^-$, typically modeled as an FLRW cosmology, embedded inside a spherically symmetric background spacetime $\mathcal M^+$, either FLRW or Schwarzschild, and joined across a timelike bubble wall $\Sigma$ (Haque et al., 2017). In that setting the bubble is not a single metric but a composite spacetime, and the main question is which such embeddings are consistent with general relativity under weak assumptions on matter.

Other papers extend the notion beyond thin-wall cosmology. In asymptotically planar AdS, a bubble may be a regular fixed-point set of a spacelike Killing field, with the AdS soliton described as a horizonless “bubble” where a compact circle caps off smoothly (El-Menoufi et al., 2013). In string theory, a bubble may be an exceptional four-dimensional sub-spacetime $W^{1,3}$ of positive curvature embedded in a ten-dimensional background, realized either as a $dS_4$ bubble wall in AdS decay or as a Lorentzian image of an exceptional Fano surface inside a Lorentzian Calabi–Yau 5-fold (Berglund et al., 2021). In low-regularity Lorentzian geometry, the term acquires a causal meaning: a causal bubble is the open set

$$\mathcal{B}^{\pm}(p):=J^{\pm}(p)\setminus\overline{I^{\pm}(p)},$$

non-empty when causal but not chronological accessibility extends beyond $\overline{I^\pm(p)}$ (García-Heveling et al., 2022).

This range of usages shows that “bubble spacetime” is a structural notion rather than a single ansatz. The bubble may be a thin shell, a regular cap, a topological throat, an exceptional cycle, or an open causal region. What remains invariant is the presence of a distinguished localized sector whose geometry is not that of the ambient spacetime.

2. Junctions, shell dynamics, and focusing constraints

The standard thin-wall formalism is based on continuity of the induced metric and a jump condition for the extrinsic curvature,

$\left[ K_{ab} \right]-h_{ab}[K]=-8\pi G\,S_{ab},$

where $S_{ab}$ is the surface stress tensor on the wall (Haque et al., 2017). In the AdS black-hole nucleation problem this is implemented explicitly for a spherical wall with stress-energy $\mathcal M^-$0, yielding the angular junction condition

$\mathcal M^-$1

for a wall separating two AdS black-hole geometries (Li et al., 2022). In the cosmological radiation-inside-dust problem, the same formalism reduces the shell motion to

$\mathcal M^-$2

with $\mathcal M^-$3 determined by the interior and exterior densities and by the shell energy density $\mathcal M^-$4 (Casadio et al., 2011).

A more geometric consistency condition comes from the null Raychaudhuri equation,

$\mathcal M^-$5

assuming the null energy condition (Haque et al., 2017). For an ingoing null congruence crossing the wall, one must have

$\mathcal M^-$6

In the spatially flat FLRW–FLRW case this becomes

$\mathcal M^-$7

so a flat or positively curved bubble in a non-positively curved cosmological background must expand no faster than the background, or must contract (Haque et al., 2017). For a Schwarzschild exterior and a flat or positively curved FLRW bubble, the constraint sharpens to

$\mathcal M^-$8

so the bubble must be contracting and bounded by its apparent horizon (Haque et al., 2017).

These focusing constraints are stronger than merely solving the Israel equations in a chosen matter model. They show that admissible bubble spacetimes are restricted by the monotonicity of null expansion, independently of many details of the microphysics. A recurring consequence is that rapidly inflating, connected, NEC-respecting bubbles are much more difficult to embed than the thin-wall intuition alone suggests.

3. Vacuum decay, cosmological evolution, and collisions

In false-vacuum decay, bubble spacetimes arise from quantum tunneling. A thin spherical wall separates a higher-energy exterior from a lower-energy interior, and the Euclidean bounce controls the tunneling exponent. In AdS with black holes, the exterior and interior can both be AdS black-hole spacetimes,

$\mathcal M^-$9

with different AdS radii $\mathcal M^+$0, masses $\mathcal M^+$1, and common charge $\mathcal M^+$2 in the RNAdS case (Li et al., 2022). The Euclidean action gives the decay exponent

$\mathcal M^+$3

and the numerical analysis shows that black holes catalyze vacuum decay; for RNAdS black holes, the tunneling rate to the final RNAdS black hole with the minimum critical mass is the highest among the possible channels (Li et al., 2022).

When the bubble interior is not vacuum but an inflating or radiative cosmology, the classical post-nucleation dynamics become sensitive to the surrounding matter. For a de Sitter bubble embedded in dust or radiation, the Israel equations give an effective proper-radius dynamics together with an evolution equation for the shell density $\mathcal M^+$4 (Pannia et al., 2016). Numerical evolution shows that inhomogeneities and equation of state matter significantly: bubbles nucleated in sub-density regions expand more slowly than in homogeneous backgrounds, and radiation environments slow the growth of the proper radius more strongly than the corresponding dust cases (Pannia et al., 2016). In the short-time analytic treatment of radiation inside dust, a bubble of radiation with vanishing initial expansion speed can be matched with an expanding dust exterior, but not with a collapsing dust exterior, regardless of the dust energy density (Casadio et al., 2011).

Bubble collisions add intrinsically non-linear structure. Full general-relativistic simulations of eternal-inflation collisions show that vacuum bubbles and bubbles with realistic inflationary cosmology can collide to produce false-vacuum pockets, oscillons, repulsive or normal post-collision walls, and classical transitions to vacua of lower or even higher energy than the colliding interiors (Johnson et al., 2011). In large-field inflationary bubbles, collisions typically leave a region with fewer e-folds but do not halt inflation; in small-field models, collisions can completely disrupt inflation in the future light cone of the collision unless the potential barriers are suitably asymmetric (Johnson et al., 2011). This suggests that bubble spacetimes are not only interfaces between vacua but also dynamical laboratories in which cosmological observables depend sensitively on wall dynamics, ambient geometry, and scalar potential structure.

4. Higher-dimensional, AdS, and modified-gravity realizations

A string-theoretic realization appears in the “stringy bubbles” framework. There the ten-dimensional metric is taken schematically as

$\mathcal M^+$5

with the induced $\mathcal M^+$6 slice at $\mathcal M^+$7 interpreted as the bubble wall or brane (Berglund et al., 2021). The warp factor depends on $\mathcal M^+$8, producing a $\mathcal M^+$9-contribution in the Ricci tensor and hence a brane-localized stress tensor at the wall. In this codimension-2 “axilaton” construction, a nontrivial axion–dilaton monodromy and localized brane sources generate warped localization of gravity and matter, an exponentially large four-dimensional Planck scale,

$\Sigma$0

and an exponentially suppressed positive cosmological constant on the $\Sigma$1 bubble (Berglund et al., 2021). The same paper argues that such $\Sigma$2 bubbles are naturally realized as exceptional Fano surfaces in a Lorentzian Calabi–Yau 5-fold, with the hyperbolic complement playing the role of an AdS-like exterior (Berglund et al., 2021).

A different higher-dimensional archetype is the AdS soliton bubble. In asymptotically planar AdS, the metric

$\Sigma$3

describes a regular horizonless spacetime in which the $\Sigma$4-circle shrinks smoothly at $\Sigma$5 (El-Menoufi et al., 2013). Regularity requires

$\Sigma$6

and the bubble enters a Smarr relation

$\Sigma$7

which is structurally symmetric to the black-hole Smarr formula under $\Sigma$8 and $\Sigma$9 (El-Menoufi et al., 2013). In AdS/CFT terms, the soliton bubble represents the confining phase, while the planar black hole represents the deconfined phase (El-Menoufi et al., 2013).

Quadratic $W^{1,3}$0 gravity adds another distinct realization. For

$W^{1,3}$1

a spherical bubble can separate two constant-curvature regions with $W^{1,3}$2, and the junction carries not only an ordinary thin shell but also a gravitational double layer of $W^{1,3}$3-type (Eiroa et al., 2017). In particular, pure double layers are possible: the matching hypersurface can have $W^{1,3}$4, $W^{1,3}$5, and $W^{1,3}$6, while the double-layer strength

$W^{1,3}$7

remains non-zero (Eiroa et al., 2017). The explicit pure double-layer bubbles require $W^{1,3}$8, and constitute the first example of a pure double layer in a gravitational theory (Eiroa et al., 2017).

5. Causality, topology change, and wormhole analogues

Some bubble spacetimes are designed precisely to alter causal structure. The TARDIS spacetime is a bubble of curvature whose interior is Rindler space with periodic timelike coordinate,

$W^{1,3}$9

after the transformation $dS_4$0 (Tippett et al., 2013). Identifying $dS_4$1 makes curves of constant $dS_4$2 into closed timelike curves. The bubble worldtube itself traces a closed loop in ambient Minkowski spacetime, so from the external viewpoint it appears as pair creation and annihilation of bubble segments, while internally it supports explicit non-geodesic CTCs (Tippett et al., 2013). The price is a wall stress tensor that violates the classical energy conditions and a non-compactly generated Cauchy horizon (Tippett et al., 2013).

A topological variant is the “topological drive,” where a compact FRW-like bubble universe forms inside Minkowski space, the throat pinches off, and the bubble later re-attaches at another spacetime point (Cai et al., 2024). The metric

$dS_4$3

interpolates between a closed FRW region and an exterior Minkowski region (Cai et al., 2024). In the pinch-off limit, the detachment point is a quasiregular singularity: curvature invariants remain finite, but causal geodesics become incomplete (Cai et al., 2024). Because the bubble may re-attach at a spacelike- or past-separated external event, the construction permits effective superluminal travel or backward-in-time travel, again at the cost of null-energy-condition violation in the throat (Cai et al., 2024).

The same cut-and-paste logic also links bubble universes to traversable wormholes. Joining two flat 3-balls along a thin-shell $dS_4$4 produces a Minkowski–Minkowski static closed universe, i.e. a bubble universe, while joining two complements of flat balls along a thin-shell $dS_4$5 produces a Minkowski–Minkowski static open universe, i.e. a traversable wormhole (Lemos et al., 2022). In the closed case the shell has

$dS_4$6

and satisfies the classical energy conditions; in the wormhole case

$dS_4$7

and violates them (Lemos et al., 2022). This complements a different topological-defect construction connecting two asymptotically Minkowski spacetimes: there the bubble acts as a nontraversable wormhole supported by matter satisfying all classical energy conditions, while scalar waves and quantum particles can tunnel through, with the wave dynamics determined by a choice of self-adjoint extension at the defect (Pitelli et al., 2015).

These examples show that bubble spacetimes lie close to wormhole physics but are not reducible to it. Some are compact universes bounded by a shell, some are traversable or nontraversable inter-universe bridges, and some produce chronology violation by modifying topology or light-cone structure rather than by supporting a stationary throat.

6. Energy conditions, stability, thermodynamics, and causal refinement

Energy conditions divide the subject sharply. TARDIS bubbles and topological drives require exotic matter violating the classical energy conditions (Tippett et al., 2013, Cai et al., 2024). Traversable Minkowski–Minkowski wormholes likewise require negative surface density, whereas the corresponding closed bubble universe satisfies the null, weak, strong, and dominant energy conditions (Lemos et al., 2022). By contrast, the topological-defect bubble connecting two asymptotically Minkowski spacetimes is supported by a spherical shell interpreted as two orthogonal families of Nambu strings, with positive surface energy density and negative pressure satisfying the classical energy conditions, even though the resulting wormhole-like geometry is nontraversable for classical bodies (Pitelli et al., 2015). AdS soliton bubbles are regular vacuum solutions with no horizon and no shell matter at all (El-Menoufi et al., 2013).

Stability is similarly model-dependent. In the AdS-decay shellworld cited by the stringy-bubble paper, the shellworld is stable against small perturbations, and in the axilaton model the de Sitter geometry resolves what would otherwise be a naked singularity (Berglund et al., 2021). The topological-defect wormhole admits both stable and unstable scalar-field evolutions depending on the chosen self-adjoint extension: one boundary condition yields a stable “delta-like” scattering pattern, while another yields unstable resonant modes (Pitelli et al., 2015). The Minkowski–Minkowski bubble universe and traversable wormhole are neutrally stable under radial perturbations, whereas the Einstein static universe is unstable and the Friedmann static hyperbolic universe is linearly stable (Lemos et al., 2022).

Bubble spacetimes also support thermodynamic and holographic interpretations. In planar AdS, the ADM mass and tensions satisfy the trace constraint

$dS_4$8

and the bubble Smarr relation mirrors the black-hole relation, exhibiting a black hole–bubble symmetry suggestive of a confining/deconfined phase correspondence in the dual gauge theory (El-Menoufi et al., 2013). In stringy cosmic-brane models, the same warp-factor integrals that localize the graviton also produce an exponentially suppressed positive cosmological constant, so the bubble simultaneously acts as a geometric cap, a localized four-dimensional world, and a source of mass hierarchy (Berglund et al., 2021).

Finally, low-regularity Lorentzian geometry introduces a distinct causal refinement. The globally hyperbolic continuous metric

$dS_4$9

provides an explicit example where $$\mathcal{B}^{\pm}(p):=J^{\pm}(p)\setminus\overline{I^{\pm}(p)},$$0: there is an open region in $$\mathcal{B}^{\pm}(p):=J^{\pm}(p)\setminus\overline{I^{\pm}(p)},$$1, foliated by branching null curves (García-Heveling et al., 2022). This shows that even global hyperbolicity and orthogonal splitting do not prevent bubble structure in the causal future when the metric is only continuous, and that the synthetic timelike curvature-dimension condition does not, by itself, exclude causal bubbling (García-Heveling et al., 2022). A plausible implication is that “bubble spacetime” names not only a class of shell or cap geometries, but also a broader mode of failure of the classical smooth causal picture.