Revisiting metastable cosmic string breaking

The paper investigates the decay (breaking) rate of metastable cosmic strings arising from hierarchical symmetry breaking, e.g., G ⊃ U(1) ⊃ nothing, with vacuum expectation values V ≫ v and π1(G)=0, π2(G/U(1))=π1(U(1))=Z. In such models, strings are classically stable in the low-energy effective theory but metastable in the full theory because monopoles from the first breaking can nucleate inside the string and sever it via a tunneling process. The standard Preskill–Vilenkin treatment models an infinitely thin string and pointlike monopoles, yielding a breaking rate per length Γ = (μ/2π) e^{-κ} with κ = m_Μ^2/μ. PTA observations of a nHz gravitational-wave background prefer metastable strings with √κ ≈ 8 and −7 ≤ log10(G_N μ) ≤ −4, but this lies in a regime where the assumed hierarchy is not extreme, casting doubt on the thin-string approximation. The authors revisit the breaking rate including finite sizes of the string and monopole, constructing an effective worldsheet theory that captures unwinding dynamics and quantifies corrections to the tunneling exponent S_B, aiming to provide robust limits relevant to interpreting PTA data.

The study situates itself within prior work on cosmic strings, topological defects, and their gravitational-wave signatures (e.g., Vilenkin & Shellard; Hindmarsh; Auclair et al.; Gouttenoire et al.). Recent PTA collaborations (NANOGrav, EPTA+InPTA, PPTA, CPTA) reported a stochastic background with Hellings–Downs correlations, prompting interest in metastable strings as a source (e.g., Buchmuller, Domcke & Schmitz; Leblond, Shlaer & Siemens). The conventional decay-rate estimate for metastable defects was formulated by Preskill & Vilenkin (thin-string, pointlike monopoles). Shifman & Yung proposed ansätze to model unwinding in non-Abelian embeddings, serving as the basis for including finite-size effects. The paper also references classic monopole (’t Hooft–Polyakov) and ANO string literature and notes related symmetry-breaking chains and bead-on-string scenarios relevant in GUT contexts.

Model: An SU(2) gauge theory with an adjoint scalar φ^a (a=1,2,3) and a doublet scalar h_i (i=1,2) is used. The Lagrangian (Minkowski metric) is L = −(1/2)(D_μφ^a)^2 − (D_μ h)†(D^μ h) − (1/4g^2)F^a_{μν}F^{aμν} − V_Higgs(φ,h), with covariant derivatives D_μ h = ∂μ h − i(g/2)A^a_μ τ^a h, D_μ φ^a = ∂μ φ^a + ε^{abc}A^b_μ φ^c. The potential is V_Higgs = λ[(1/2 φ^2 − v^2)^2 + λ_2(φ^aφ^a − V^2)^2 + γ(V/2 − φ^3/V)^2], with V ≫ v and positive couplings λ, λ_2, γ. Vacuum: ⟨φ^a⟩ = V δ^{a3} breaks SU(2)→U(1). With γ>0, ⟨h⟩=(v, 0)^T breaks the remaining U(1) (h has charge ±1/2). Monopoles: The ’t Hooft–Polyakov monopole arises at SU(2)→U(1). The effective U(1) field strength and monopole charge Q_m=−4π are defined; monopole mass m_M = (4π V/g) f_M(m_φ/m_W), with m_φ=√(8λ)V and m_W=gV; a numerical fit for f_M(x) is provided. Cosmic strings: At the second breaking, the theory reduces to U(1) with complex scalar h_1 (charge 1/2), permitting ANO strings. Profiles ξ(ρ), f(ρ) satisfy boundary conditions; string tension μ = 2π v^2 f_r(m{h1}/m_γ) with m{h1}=2√λ v and m_γ = (1/√2) g v; numerical fit for f_r(x) is provided, with asymptotic forms for μ. Monopole–string connection: Using a combed gauge with two charts (northern and southern hemispheres), the monopole’s gauge transition t_{NS}=e^{iφ} implies that a string with winding n=−1 must attach to a monopole; the magnetic flux through the string equals the monopole’s charge. Hence, strings can end on monopoles, allowing string breaking via monopole–antimonopole nucleation. Conventional thin-string estimate: In Euclidean worldsheet description, the monopole worldline forms a circle (bubble nucleation). The bounce action is S_B^{(thin)} = π m_M^2/μ = πκ, giving Γ ≈ (μ/2π) e^{-S_B}. Finite-size unwinding ansätze: Following Shifman & Yung, introduce an unwinding parameter β that interpolates between the ANO string (β=0) and the unwound state (β≈π/2). Primitive Ansatz (regular/singular gauge forms provided) specifies h, A_μ, and φ with profiles ξ_β(ρ), f_β(ρ), Φ_β(ρ) and boundary conditions. This yields a β-dependent string tension T(β) obtained by minimizing the radial profiles. Improved Ansatz allows separate radial profiles for the r^3 component (string) and the others (monopole) via f_β and f_β^W, to reflect different core sizes ~ v^{-1} vs V^{-1}. Effective 1+1D theory: Promote β→β(t,z). To avoid singularities, introduce an additional profile a_s(ρ) in A_n (n=t,z) with boundary conditions a_s(0)=1, a_s(∞)=0. After integrating over transverse coordinates, obtain an effective Euclidean action S_E = 2π ∫ ρ_E dρ_E [ (1/2) K_eff(β)(∂_{ρ_E}β)^2 + T(β) ], where K_eff(β) and T(β) are computed numerically from radial integrals (explicit formulae in Appendix A). The O(2)-symmetric bounce satisfies β(0)=π/2, β′(0)=0, β(∞)=0 and the EOM K_eff β″ = T′ − (1/2)K_eff′ (β′)^2 − (1/ρ_E)K_eff β′. Thin-wall approximation (for large hierarchy V ≫ v, small ΔT ≡ T(0)−T(π/2)): The action reduces to S_B ≈ π m_eff^2 / [T(0) − T(π/2)], with m_eff = ∫_0^{π/2} dβ √[2 K_eff(β)] [T(β) − T(0)], mirroring the thin-string formula with m_M → m_eff. Numerical procedure: (1) For each β, minimize T(β) over ξ_β, f_β (and f_β^W for improved), and Φ_β; (2) Determine a_β(ρ) by minimizing ∫ dρ H_β; (3) Build the 2D effective action by integrating over transverse coordinates; (4) Solve the bounce equation with SO(2) symmetry. Mass parameters are expressed via couplings and g; S_B scales as g^{-2}. Visualizations: Field configurations illustrate that the monopole core size remains ~ m_W^{-1} while the string width is ~ m_γ^{-1}.

• Agreement with thin-string in strong hierarchy: For m_W ≫ m_γ, the primitive Ansatz reproduces the asymptotic behavior of the thin-string bounce action S_B^{(thin)} = π m_M^2/μ up to O(1) factors; the thin-wall approximation accurately matches the full bounce solution in this regime, with m_eff close to m_M at the several-tens-of-percent level. - Finite-size effects and improved Ansatz: Allowing separate profiles (improved Ansatz) lowers T(β) for given β relative to the primitive Ansatz, but K_eff(β) can increase significantly, leading to a larger S_B for m_W > m_γ in the authors’ minimization scheme. The anticipated large logarithmic enhancement in T(β) from monopole spreading is absent; instead, the string core shrinks to maintain a small monopole core (~ m_W^{-1}). - Breakdown of thin-wall/thin-string for mild hierarchy: For m_W/m_γ ≲ O(10), the thin-wall approximation underestimates the bounce action, signaling that the infinitely thin string approximation is unreliable in this regime. This is particularly relevant since PTA-preferred values correspond to √κ ≈ 7.5–8.5, which imply only mild hierarchy. - Robust constraints beyond thin limit: The unwinding-ansatz-based bounce actions provide upper limits on S_B (and thus lower limits on the tunneling factor e^{−S_B}) even when m_W/m_γ = O(1), where S_B^{(thin)} is not reliable. - Parameter dependences: Increasing m_{h1}/m_W decreases S_B, while S_B depends only mildly on m_{h2}/m_W and m_φ/m_W. Contour plots indicate regions where the improved-ansatz S_B is smaller than S_B^{(thin)}, implying that the thin-string formula can underestimate the breaking rate in parts of parameter space. - PTA relevance: Bands with S_B = π κ for √κ ≈ 7.5–8.5 are identified; results show that conventional fits based solely on thin-string κ may misestimate the required microphysical parameters when hierarchy is mild.

The study directly addresses the reliability of conventional (infinitely thin) estimates of the cosmic string breaking rate in the parameter regime relevant to PTA observations. By explicitly modeling the unwinding with finite monopole and string cores and constructing an effective worldsheet theory, the authors quantify when S_B^{(thin)} remains valid (large hierarchy) and when it fails (mild hierarchy). For large V/v, the primitive ansatz’s S_B agrees with thin-string expectations up to O(1), validating use of κ-based parameter inferences in that limit. However, for m_W/m_γ ≲ O(10), thin-wall/thin-string approximations underestimate S_B, so naive extrapolations from κ can lead to incorrect mappings between the GW signal and underlying particle-physics parameters. The improved ansatz, while physically motivated to separate core scales, yields larger S_B due to enhanced kinetic terms in the adopted minimization scheme, emphasizing the importance of a consistent treatment of kinetic contributions. Overall, the finite-size framework provides robust upper limits on S_B (lower bounds on e^{−S_B}), enabling conservative constraints on models attempting to explain the PTA signal via metastable strings without requiring extreme hierarchies.

The paper develops and applies finite-core unwinding ansätze to compute the bounce action for the breaking of metastable cosmic strings in SU(2)→U(1)→nothing models. It shows that for strong hierarchies (m_W ≫ m_γ), the primitive ansatz reproduces the thin-string result for S_B up to O(1), affirming the conventional Preskill–Vilenkin estimate. Crucially, for mild hierarchies m_W/m_γ ≲ O(10), relevant to PTA-favored √κ ≈ 7.5–8.5, the thin-wall/thin-string approximations underestimate S_B, while the ansatz-based calculations provide reliable upper limits on S_B and thus robust lower limits on the string-breaking rate. The work refines interpretation of PTA data in terms of underlying symmetry-breaking scales and couplings and delineates parameter dependencies (notably the impact of m_{h1}/m_W). Future directions include: a full stability analysis of string configurations (beyond the β direction) especially for m_W/m_γ = O(1); extending to other symmetry-breaking chains (e.g., SU(2)×U(1)→U(1)×U(1)→U(1)) where unconfined flux appears; improving the minimization procedure for the improved ansatz and exploring mixed ansätze that can further reduce S_B; and cosmological lattice simulations to study network formation when phase transitions are not well separated.

• The classical stability of the string for m_W/m_γ = O(1) is not fully demonstrated; only the stability along the β (unwinding) direction is checked heuristically. - Results rely on specific unwinding ansätze (primitive and improved) and a particular minimization procedure; the improved ansatz may overestimate S_B due to separate minimizations of K_eff and T. - Analysis is restricted to the SU(2)→U(1)→nothing chain; scenarios with unconfined flux (e.g., SU(2)×U(1)→U(1)×U(1)→U(1)) are not treated. - Cosmological aspects (formation of metastable string networks, interplay of close-in-time phase transitions) are not simulated; thermal history effects are only discussed qualitatively. - Thin-wall approximations break down for mild hierarchy, and while the ansätze provide upper limits on S_B, the true minimal action path might yield smaller S_B than obtained here.