Non-BPS ’t Hooft-Polyakov Monopoles

• The paper presents non-BPS ’t Hooft-Polyakov monopoles as finite-energy solutions beyond the BPS limit, solved by full second-order Yang-Mills-Higgs equations.
• It details far-field asymptotics, transseries constructions, and a resurgent structure highlighting universal gauge profiles and spectral resonances.
• It compares non-BPS deformations and generalizations, emphasizing their nontrivial linearized spectrum and implications for soliton dynamics.

Non-BPS ’t Hooft-Polyakov monopoles are finite-energy $SU(2)$ Yang-Mills-Higgs solitons with an adjoint Higgs field outside the Bogomol'nyi-Prasad-Sommerfield limit, typically characterized by nonzero Higgs self-coupling. In this regime the monopole no longer obeys first-order self-dual equations, but instead solves the full second-order radial system; the resulting background is more generic than the BPS monopole, and its asymptotics, fluctuation spectrum, and analytic structure are correspondingly different (Constantinidis et al., 2017, Dunne et al., 19 Feb 2026, Malinský, 16 Feb 2026).

1. Field-theoretic definition and radial system

The standard setting is $SU(2)$ Yang-Mills-Higgs theory with an adjoint scalar. A convenient static, purely magnetic, spherically symmetric ansatz is

$$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$

In this sector the monopole mass functional can be written as

$$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$

and the profile functions satisfy the coupled second-order equations

$W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$

with finite mass requiring

$W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$

A closely related normalization, used in other analyses, writes the radial equations in terms of the Higgs self-coupling $\lambda$ and emphasizes that the non-BPS monopole solves the full second-order Yang-Mills-Higgs equations rather than the Bogomolny equations (Dunne et al., 19 Feb 2026, Constantinidis et al., 2017).

The BPS limit is special. When the Higgs self-coupling vanishes, the energy admits a Bogomolny completion and the second-order equations reduce to

$$W'(r) = -H(r)\,W(r), \qquad H'(r) = \frac{1}{r^2}\left(1-W(r)^2\right),$$

with explicit solution

$$W_{\rm BPS}(r) = \frac{r}{\sinh r}, \qquad H_{\rm BPS}(r) = \coth r-\frac{1}{r}.$$

By contrast, for nonvanishing Higgs self-coupling the background no longer satisfies the BPS equation, the Higgs is massive, and the profiles must be found numerically from the second-order radial equations (Dunne et al., 19 Feb 2026, Russell et al., 2010).

2. Far-field asymptotics and resurgent structure

Recent work has analyzed non-BPS monopoles through resurgence theory. In one formulation the gauge profile is denoted $SU(2)$0, the scalar profile $SU(2)$1, and the non-BPS regime is $SU(2)$2, with equations

$SU(2)$3

A key result is that for any $SU(2)$4, $SU(2)$5 exponentially fast and the gauge asymptotics become universal: $SU(2)$6 The associated Borel kernel is a hypergeometric function with a single primary logarithmic branch point at $SU(2)$7, and the nonlinear Volterra formulation propagates this seed into a discrete real-axis singularity set

$SU(2)$8

with a resonance structure at the origin of the Borel plane. In the same analysis, the BPS case is singled out as qualitatively different: its Borel transform has simple poles on the imaginary axis and $SU(2)$9 is a regular point rather than part of a logarithmic branch-point tower (Malinský, 16 Feb 2026).

A complementary transseries construction treats radially symmetric non-BPS monopoles directly in the far-field region. For the equal-mass case $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$0, the large-$$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$1 solution is written as

$$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$2

Here the exponential sectors are powers of $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$3, while the fluctuation factors $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$4 and $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$5 are generally factorially divergent asymptotic series in $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$6. The leading decaying gauge fluctuation is controlled by a modified Bessel function of imaginary order, whereas the leading scalar fluctuation is simply $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$7. All higher exponential sectors are generated recursively from the leading homogeneous modes, and near-origin matching fixes the transseries parameters. For $$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$8, the reported matching values are

$$A_0^a = 0, \qquad A_i^a = \epsilon_{iak} \frac{x_k}{r^2} (W(r)-1), \qquad \Phi^a = H(r)\frac{x_a}{r}.$$9

with rescaled monopole mass

$$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$0

In the BPS limit, by contrast, the fluctuation factors truncate and the transseries becomes a convergent sum (Dunne et al., 19 Feb 2026).

3. Linearized spectrum, resonances, and global integral constraints

The non-BPS monopole remains a rich background for linearized bosonic dynamics. For $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$1, direct hedgehog perturbation analysis shows that the effective long-range attraction in the massive channel is no longer Coulombic, as in the BPS case, but has inverse-square asymptotics: $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$2 This still supports infinitely many bound states, now with threshold accumulation governed by an attractive $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$3 tail. In the coupled scattering problem the same mechanism produces Feshbach resonances. For $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$4, the decoupled massive channel yields bound-state frequencies

$$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$5

and the phase shift in the coupled system exhibits a clear resonance near

$$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$6

Thus switching on the Higgs self-coupling does not destroy the resonance phenomenon; it deforms the threshold structure and makes the near-threshold spectrum denser (Russell et al., 2010).

The non-BPS monopole has also served as a test background for integral formulations of Yang-Mills theory. In the generalized integral Bianchi identity for static monopoles, only the parameter $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$7 survives, and the ’t Hooft-Polyakov background provides a nontrivial test beyond the BPS sector. Numerical evaluation shows that the non-BPS monopole satisfies the first- and second-order integral equations obtained from the $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$8-expansion, with coefficient matching typically about $$\mathcal M_\beta = \frac{4\pi v}{e} \int_0^{\infty} dr \ \left[W'^2+\frac{r^2}{2}H'^2+\frac{(1-W^2)^2}{2r^2}+\frac{\beta^2r^2}{8}(H^2-1)^2+W^2H^2 \right],$$9. This supports the claim that the integral Yang-Mills equations are not merely BPS artifacts but encode nonlocal constraints compatible with ordinary finite-energy non-BPS solitons (Constantinidis et al., 2017).

4. Explicit non-BPS generalizations and deformations

Several papers construct monopole configurations that retain the topological and gauge-theoretic character of the ’t Hooft-Polyakov solution while departing from the BPS regime in distinct ways. One class consists of axially symmetric Jacobi-elliptic one-monopoles. These are constructed by replacing the large-distance angular structure of the usual one-monopole with Jacobi elliptic functions and then solving the full second-order equations numerically. For $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$0 they are regular finite-energy unit-charge solutions, but unlike the standard BPS monopole they are only cylindrically symmetric. They are explicitly described as non-BPS for both vanishing and nonvanishing Higgs potential, with total energies around $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$1 in units of $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$2 for $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$3 and $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$4 for $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$5. A recurring source of confusion is removed directly in that work: even when $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$6, these solutions are still non-BPS because they are obtained from the second-order field equations rather than the first-order Bogomol’nyi equations (Teh et al., 2010).

A different deformation is the spiked monopole, defined in an $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$7 model with two adjoint Higgs fields, one free and one self-interacting. The self-interacting Higgs produces a sharply localized non-BPS-like core, while the free Higgs generates a BPS-like $W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$8 tail,

$W''(r)-W(r)H(r)^2-\frac{W(r)(W(r)^2-1)}{r^2} = 0,$9

Numerically one finds

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$0

with equality only in the BPS case. For the representative choice $$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$1, the paper reports

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$2

about half the BPS value, and the resulting monopoles remain repulsive in all simulations. The construction therefore realizes a hybrid object with a non-BPS core and only partial long-range force cancellation (Evslin, 2018).

A more radical extension arises from arbitrary-gauge spherical parametrizations of the Georgi-Glashow model. In the real parametrization, the usual profile function acquires a gauge-invariant interpretation through

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$3

while the gauge angle is pure gauge. In the complex sector, exact monopole and dyon solutions are constructed by imposing $$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$4 with $$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$5. The resulting monopole profile is

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$6

and its longitudinal magnetic field coincides with the Wu-Yang monopole. These solutions are explicitly non-BPS, and their energy density is real,

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$7

but singular at the origin, so they are not regular finite-energy monopoles in the ordinary ’t Hooft-Polyakov sense (Korenblit et al., 2021).

5. Alternative realizations and broader monopole landscapes

Non-BPS ’t Hooft-Polyakov monopoles also appear in contexts where the monopole profile is realized indirectly rather than as a classical soliton of the Georgi-Glashow equations. In a quantum-mechanical model of a free spinless particle on a circle with two point interactions, the parameter space of boundary conditions contains an $$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$8 subfamily on which all energy levels are doubly degenerate. The non-Abelian Berry connection in the excited-state sector takes the hedgehog form

$$H''(r) - \frac{2}{r^2}H(r)W(r)^2 - \frac{\beta^2}{2}H(r)(H(r)^2-1)+\frac{2H'(r)}{r}=0,$$9

The asymptotics are

$W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$0

so the connection is regular at the origin and pure gauge asymptotically. The paper identifies these excited-state Berry connections as non-BPS ’t Hooft-Polyakov monopoles, in contrast with the ground-state sector, which reproduces the BPS profile (Ohya, 2015).

A broader effective-field-theory viewpoint places the ’t Hooft-Polyakov monopole inside a large landscape of finite-mass monopoles. In a general $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$1 theory with scalar-dependent gauge couplings and, more generally, with dipole-coupled charged vector fields, the familiar classical monopoles are described as special points of a common parameter space. In this classification the ’t Hooft-Polyakov monopole corresponds to

$W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$2

where the effective $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$3-plus-scalar description becomes the unitary gauge of an $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$4 theory with adjoint scalar. This broader landscape is not itself a construction of the classical non-BPS monopole profile, but it situates the non-BPS ’t Hooft-Polyakov solution within a wider family of dressed finite-mass monopoles (Blaschke et al., 2022).

6. BPS-specific constructions and the limits of direct extension

A substantial neighboring literature concerns the ’t Hooft-Polyakov monopole but remains explicitly BPS or self-dual. This matters because non-BPS monopoles are often informally imported into frameworks where only the BPS solution is actually under analytic control. In the geometric theory of defects, for example, the monopole is reinterpreted as a Riemann-Cartan configuration in a solid with Euclidean metric, nontrivial $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$5 connection, torsion, and curvature. The explicit Burgers- and Frank-vector densities are computed for the BPS profile $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$6, and while the same geometric interpretation is said to apply formally to arbitrary $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$7, the non-BPS defect densities are not worked out explicitly (Katanaev, 2020).

A similar limitation holds in higher-dimensional embeddings. The construction of six-dimensional non-abelian self-dual strings from four-dimensional monopoles uses BPS monopole seeds and treats the ’t Hooft-Polyakov monopole only in the BPS limit; the paper states directly that it does not address non-BPS ’t Hooft-Polyakov monopoles. Likewise, in five-dimensional gauge-Higgs unification the classical absence of a Higgs potential makes the BPS monopole natural, and the monopole equation becomes equivalent to a five-dimensional self-duality condition. The discussion of non-BPS configurations there is only indirect: they would arise once a quantum-generated Higgs potential is included, but the paper stays within the self-dual regime (Chu, 2013, Hasegawa et al., 2019).

The same caution applies to several generalized monopole programs. Analytical self-dual solutions in nonstandard Yang-Mills-Higgs models require $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$8, internal-structure and compact monopoles in $W(r)\sim e^{-r}+\cdots, \qquad H(r)\sim 1+\cdots .$9 models satisfy $\lambda$0, and shell-like monopoles in effective $\lambda$1 theories are constructed in the BPS limit with the standard ’t Hooft-Polyakov monopole appearing as a special case. These works are relevant for monopole morphology and model-building, but they are not analyses of non-BPS ’t Hooft-Polyakov monopoles themselves (Casana et al., 2013, Bazeia et al., 2018, Beneš et al., 2023).

Taken together, these distinctions delimit the subject sharply. Non-BPS ’t Hooft-Polyakov monopoles are ordinary finite-energy Yang-Mills-Higgs solitons beyond self-duality; they are governed by second-order equations, possess nontrivial resurgent and spectral structures, admit non-spherical and multi-field generalizations, and appear in several indirect realizations. At the same time, many of the most explicit analytic constructions in adjacent monopole literature remain BPS-specific, so extending those results to the non-BPS regime typically remains a formal suggestion rather than a completed calculation.