Bogomolny Bounds and First-Order Equations

An energy admits a Bogomolny bound when, for a stated coupling and boundary sector, it can be written as a sum of nonnegative squares plus a boundary or topological term. The boundary term gives a lower bound, and a configuration saturates it only if every square vanishes. Those first-order equations imply the second-order static field equations under the same regularity and boundary assumptions. Saturation proves classical energetic minimality in that sector; it does not prove that a saturated solution exists, that every solution is saturated, or that the mass is quantum exact.

Required background. Finite-energy boundary data and charge supplies the sector and orientation. Helpful background. The concrete kink, vortex, and monopole models provide three normalizations of the same method.

What a square completion requires

Suppose the static energy in sector $Q$ can be rearranged as

$$E[\Phi] =\int\mathrm d^D x\, \sum_A c_A\,|\mathcal F_A[\Phi]|^2 +\mathcal Z(Q), \qquad c_A>0 .$$

The term $\mathcal Z$ must depend only on the declared boundary data or conserved charge, not on local deformations within the sector. Then

$$E\geq\mathcal Z(Q)$$

for the chosen orientation. If $\mathcal Z$ can have either sign, the invariant statement is usually $E\geq|\mathcal Z|$. Saturation requires

$$\mathcal F_A[\Phi]=0 \quad\text{for every }A .$$

Three checks are indispensable:

1. expanding the squares reproduces every coefficient and sign in the original energy;
2. integration by parts produces only the stated boundary term and no discarded interior or singular contribution;
3. the first-order system is compatible with the global boundary data and has a regular solution.

The method originated in the general analysis of Bogomolny 1976, pp. 449–454 and is developed across the canonical models in Manton and Sutcliffe 2004, chs. 5, 7, and 8.

Kink: one square and a smooth completion function

For one real scalar in one spatial dimension,

$$E=\int\mathrm dx\, \left[\frac12(\phi')^2+V(\phi)\right].$$

Let $W$ be a real function on the relevant interval satisfying

$$V(\phi)=\frac12[W'(\phi)]^2,$$

Here $W$ is simply a convenient real function for the square completion. Unless the model is embedded in a supersymmetric theory, calling it a superpotential does not by itself supply a supersymmetry algebra or a quantum BPS statement. The energy can then be written as

$$E =\frac12\int\mathrm dx\, \big(\phi'\mp W'(\phi)\big)^2 \pm\big[W(\phi(+\infty))-W(\phi(-\infty))\big].$$

Choosing the sign that makes the boundary term positive gives

$$E\geq|\Delta W|, \qquad \phi'=\pm W'(\phi).$$

Differentiating the first-order equation yields

$$\phi''=W''W'=V'(\phi),$$

so the second-order equation follows. This proof assumes a single smooth branch of $W$ along the profile. A formal square root of $2V$ that changes sign or is nonsmooth at an interior point does not automatically define a global first-order problem.

Critical Abelian Higgs vortex

Use the model and normalization of Vortices, Flux Quantization, and Core Scales in the transverse plane. At critical coupling $\lambda=e^2$,

$$E =\int\mathrm d^2x\, \left[ |D_i\phi|^2 +\frac12B^2 +\frac{e^2}{2}\left(|\phi|^2-v^2\right)^2 \right],$$

where $B=F_{12}$ and the plane has orientation $\mathrm dx^1\wedge\mathrm dx^2>0$. With $D_i=\partial_i-ieA_i$ and therefore $[D_1,D_2]=-ieB$, the exact derivative identity is

$$\begin{aligned} |D_1\phi|^2+|D_2\phi|^2 ={}&\left|(D_1+iD_2)\phi\right|^2+eB|\phi|^2\\ &-i\partial_1(\phi^*D_2\phi) +i\partial_2(\phi^*D_1\phi). \end{aligned}$$

The last line integrates to the boundary and vanishes for the stated regular finite-energy data. For positive flux,

$$\begin{aligned} E =\int\mathrm d^2x\, \Bigg[ &\left|(D_1+iD_2)\phi\right|^2\\ &+\frac12\left(B-e(v^2-|\phi|^2)\right)^2 \Bigg] +ev^2\Phi_B . \end{aligned}$$

Since $\Phi_B=2\pi n/e$,

$$E\geq2\pi v^2 n \quad (n>0).$$

The saturated equations are

$$(D_1+iD_2)\phi=0, \qquad B=e(v^2-|\phi|^2).$$

For negative flux, reverse both signs and obtain $E\geq2\pi v^2|n|$.

The first equation also provides an algebraic check on the scalar field equation. Acting with $D_1-iD_2$ gives

$$D_iD_i\phi+eB\phi=0,$$

and the second first-order equation turns this into

$$D_iD_i\phi =e^2(|\phi|^2-v^2)\phi,$$

the static scalar Euler–Lagrange equation. The gauge equation is just as explicit. Define

$$j_i =ie\left[\phi^*D_i\phi-(D_i\phi)^*\phi\right].$$

The first BPS equation is equivalent to $D_i\phi=-i\epsilon_{ij}D_j\phi$, so

$$j_i=e\epsilon_{ij}\partial_j|\phi|^2.$$

The second BPS equation gives $\partial_jB=-e\partial_j|\phi|^2$. Since $\partial_jF_{ji}=-\epsilon_{ij}\partial_jB$, the gauge Euler–Lagrange equation

$$\partial_jF_{ji}=j_i$$

follows with the same orientation and signs.

At critical coupling, the spatial stress of a saturated static vortex vanishes, so members of a smooth separated BPS family exert no static force on one another. The existence and dimension of that family are a separate result: for winding $n$, the regular critical equations have a $2|n|$-dimensional moduli space interpreted as vortex positions Weinberg 1979, pp. 3008–3012. Away from $\lambda=e^2$, the leftover potential coefficient cannot be absorbed into the same squares and forces generally lift the positional moduli.

Monopole: square completion in three dimensions

In the Prasad–Sommerfield limit of the adjoint $SU(2)$ gauge–Higgs model,

$$E =\frac12\int\mathrm d^3x\, \left[(B_i^a)^2+(D_i\Phi^a)^2\right].$$

Take $|\Phi|\to v$ at spatial infinity and define the magnetic orientation and normalization by

$$n_{\mathrm m} =\frac{g}{4\pi v} \lim_{R\to\infty} \int_{S_R^2}\mathrm dS_i\,\Phi^aB_i^a.$$

For the standard globally regular $SU(2)$ monopole this is an integer. More generally, the global form of the gauge group and the allowed matter representations determine the magnetic charge lattice, so that assumption must accompany the formula.

For positive magnetic orientation,

$$\begin{aligned} E &=\frac12\int\mathrm d^3x\, \left(B_i^a-D_i\Phi^a\right)^2 +\int\mathrm d^3x\,B_i^aD_i\Phi^a\\ &=\frac12\int\mathrm d^3x\, \left(B_i^a-D_i\Phi^a\right)^2 +\frac{4\pi v}{g}n_{\mathrm m}. \end{aligned}$$

The Bianchi identity $D_iB_i=0$ turns the cross term into the surface magnetic charge. Thus

$$E\geq\frac{4\pi v}{g}|n_{\mathrm m}|, \qquad B_i^a=\pm D_i\Phi^a .$$

Prasad and Sommerfield exhibited the regular unit-charge saturated solution Prasad and Sommerfield 1975, pp. 760–762. At nonzero Higgs self-coupling, the same square completion leaves an additional nonnegative potential. The lower bound survives, but a nontrivial smooth monopole cannot saturate the full energy and its classical mass exceeds the bound.

What saturation proves

For variations that preserve the sector and boundary conditions, a saturated configuration is an absolute classical energy minimum because every competing configuration has the same topological term and nonnegative squares. The Hessian is therefore nonnegative on admissible perturbations. Exact symmetries can still give zero modes, producing a moduli space rather than an isolated minimum.

The shared stability taxonomy places this conclusion among weaker and different stability statements: square completion proves a classical bound but does not construct a global solution or supply quantum protection. The soliton boundary and stability comparison records where the kink, vortex, and monopole bounds exist and where additional fluctuation or quantum tests begin.

Saturation does not establish:

• existence for arbitrary charge, geometry, or boundary conditions;
• uniqueness, since moduli or disconnected solution branches may remain;
• stability against variations that change the charge or boundary sector;
• absence of continuum radiation in time-dependent motion;
• exact equality after quantum corrections.

Supersymmetry can place the same charge in a central extension and protect a BPS mass, but that conclusion requires the supersymmetry algebra, representation theory, anomalies, and quantum state. It does not follow from the nonsupersymmetric square completion alone.

Stress and force as independent checks

For a static solution, varying the spatial metric gives the stress tensor $T_{ij}$. A saturated kink has equal gradient and potential energy densities in the transverse direction. Critical vortices have vanishing integrated and, for the BPS equations, local planar stresses. BPS monopoles exhibit cancellation of long-range magnetic and scalar forces.

These observations are checks, not alternative derivations of the bound. Vanishing net force between well-separated objects can occur approximately for other reasons, and a zero integrated pressure does not prove that all local squares vanish.

Common pitfalls

Completing the wrong coefficients. The saturable decomposition into the displayed BPS squares plus flux exists only at the critical coupling. For $\lambda>e^2$, a positive residual preserves the lower bound but prevents saturation; for $\lambda<e^2$, its sign prevents this argument from establishing the bound. In either case, importing the critical first-order equations fails when the full energy is expanded.

Dropping a boundary term without checking patches or singularities. Gauge potentials may require patches, and a singular core can contribute an interior boundary. Use gauge-invariant charge data and verify regularity.

Calling every first-order equation BPS-exact. Here “Bogomolny” means a classical energy bound. Quantum exactness is an additional supersymmetric statement.

Exercises

1. Expand the critical vortex squares and recover the original energy plus the flux term.

Solution

The magnetic square contributes

$$\frac12B^2+\frac{e^2}{2}(v^2-|\phi|^2)^2 -eB(v^2-|\phi|^2).$$

The exact identity can be rearranged as

$$\left|(D_1+iD_2)\phi\right|^2 =|D_i\phi|^2-eB|\phi|^2 +\text{total derivative}.$$

Combining this with the magnetic square leaves the original energy minus $ev^2B$ plus that boundary derivative. Its integral vanishes for the regular finite-energy fields used here. Moving the flux term to the right yields the displayed completion and $E\geq ev^2\Phi_B$.

2. Add a nonnegative scalar potential $V(\Phi)$ to the monopole energy and explain why $B_i^a=D_i\Phi^a$ no longer saturates the full energy unless $V$ vanishes on the entire profile.

Solution

The gauge and gradient terms still complete into a square plus magnetic charge, but $\int V(\Phi)\,\mathrm d^3x$ remains as an additional positive term. Saturation of the old square leaves this term nonzero for a profile that crosses the Higgs core. Hence the mass is strictly above $4\pi v|n_{\mathrm m}|/g$ in the generic theory.

Continue

Moduli-Space Dynamics and Collective Quantization explains how normalizable zero modes of saturated families become low-energy coordinates. Kinks, vortices, and monopoles retain the complete model data and failure boundaries.

References

• Bogomolny, Evgeny B. “Stability of Classical Solutions.” Soviet Journal of Nuclear Physics 24 (1976): 449–454. INSPIRE record.
• Manton, Nicholas, and Paul Sutcliffe. Topological Solitons. Cambridge Monographs on Mathematical Physics. Cambridge University Press, 2004, chs. 5, 7, and 8. DOI.
• Prasad, M. K., and Charles M. Sommerfield. “Exact Classical Solution for the ’t Hooft Monopole and the Julia–Zee Dyon.” Physical Review Letters 35 (1975): 760–762. DOI.
• Weinberg, Erick J. “Multivortex Solutions of the Ginzburg–Landau Equations.” Physical Review D 19 (1979): 3008–3012. DOI.