BPS Particles and Central Charges

A central charge can turn the supersymmetry algebra into a quantitative mass inequality. Saturation shortens the representation and protects an index, but it does not by itself prove that a particle of the chosen charge exists or remains stable throughout parameter space. Those are dynamical and chamber-dependent questions.

Required background. BPS shortening and central-charge bounds supplies the representation theory. Helpful background. Supercurrent multiplets explains how extended charges enter current algebra, and electric and magnetic one-form symmetries clarifies charge lattices and global form.

The four-dimensional N=2 particle bound

Fix the normalization

\begin{aligned} \{Q^I_\alpha,\bar Q_{\dot\beta J}\} &=2\delta^I{}_J\sigma^\mu_{\alpha\dot\beta}P_\mu,\\ \{Q^I_\alpha,Q^J_\beta\} &=2\varepsilon_{\alpha\beta}\varepsilon^{IJ}Z, \end{aligned}

where $I,J=1,2$ . For a massive state in its rest frame, $P^\mu=(M,\mathbf0)$ . Write $Z=|Z|e^{i\varphi}$ and take phase-rotated linear combinations of $Q^1$ and the Hermitian conjugate of $Q^2$ . Their anticommutators diagonalize to eigenvalues proportional to

2(M+|Z|), \qquad 2(M-|Z|).

The norm of any state obtained by acting with a supercharge is nonnegative. Therefore

M\ge |Z|.

This is the central-charge positivity argument of Witten and Olive 1978, pp. 97–101. The numerical coefficient belongs to the displayed algebra convention. Rescaling $Z$ changes the written bound, not its physics.

If $M>|Z|$ , all fermionic oscillators act nontrivially and the particle belongs to a long massive multiplet. If $M=|Z|$ , the oscillators associated with $M-|Z|$ annihilate the state. A half-BPS particle preserves four of the eight real supercharges and occupies a shorter multiplet. The preserved combination depends on the phase $\varphi$ ; two BPS states preserve the same supercharges only when their central charges have aligned phases.

Central charge as a function of vacuum and charge

In an abelian low-energy theory of rank $r$ , electric and magnetic charges form an integral lattice $\Gamma$ . In a symplectic basis,

\gamma=(p^I,q_I), \qquad \langle\gamma,\gamma'\rangle=p^Iq'_I-q_Ip'^I.

The central charge is linear in $\gamma$ :

Z_\gamma(u)=q_Ia^I(u)+p^Ia_{D,I}(u)+s_am^a,

where $u$ labels the vacuum, $(a^I,a_{D,I})$ are special coordinates, and flavor charges $s_a$ couple to complex masses $m^a$ . Factors such as $\sqrt2$ depend on the chosen normalization of the supersymmetry algebra and special coordinates.

Electric–magnetic monodromy changes the charge basis and the period vector together, leaving $Z_\gamma$ invariant. Thus a statement like “the particle is electric” is frame-dependent, whereas its charge-lattice pairing and mass are invariant.

For a one-particle BPS state,

M_\gamma(u)=|Z_\gamma(u)|.

The right-hand side is exact once the exact periods and masses are known. It predicts the mass conditional on the state’s existence. It does not populate the charge lattice automatically.

Existence, stability, and protection are different claims

For each proposed charge $\gamma$ , ask four questions in order.

Is the charge allowed? The global gauge group, line-operator spectrum, flavor group, and Dirac quantization determine $\Gamma$ .
Does a normalizable one-particle state exist? This can require semiclassical quantization, an exact construction, or another dynamical argument.
Is it stable in the chosen chamber? A decay $\gamma\to\gamma_1+\gamma_2$ is kinematically marginal when
$|Z_\gamma|=|Z_{\gamma_1}|+|Z_{\gamma_2}|,$
which for $Z_\gamma=Z_{\gamma_1}+Z_{\gamma_2}$ occurs when the two constituent phases align.
Which protected quantity is known? An index can remain invariant under continuous deformations that preserve the gap in the relevant sector even while unprotected degeneracies rearrange.

This order prevents a common logical reversal: algebra gives a lower bound for every state, not an existence theorem for every lattice vector.

Protected spin information

In four-dimensional $\mathcal N=2$ theories, a useful protected quantity is the second helicity supertrace. One common convention is

\Omega(\gamma;u) =-\frac12\operatorname{Tr}_{\mathcal H^{\mathrm{1p}}_{\gamma,u}} (-1)^{2J_3}(2J_3)^2.

The trace is over the one-particle Hilbert space of charge $\gamma$ in the vacuum $u$ , with the center-of-mass multiplet treated in the stated convention. Long multiplets cancel; short multiplets contribute. Other authors absorb signs or universal multiplet factors into $\Omega$ , so a wall-crossing formula is meaningful only after its convention is declared.

The index can jump on walls of marginal stability because a normalizable bound state merges with a multiparticle continuum. This does not contradict local constancy away from walls: the spectral condition required for index invariance fails exactly at the wall.

Example: weakly coupled SU(2)

On the Coulomb branch of pure $SU(2)$ $\mathcal N=2$ Yang–Mills theory, the gauge group is generically broken to $U(1)$ . In a conventional basis,

Z_{(p,q)}=q\,a+p\,a_D.

At weak coupling the charged $W$ boson is an electric BPS state, while semiclassical monopoles and dyons carry magnetic charge. Near a monopole singularity, $a_D\to0$ and the monopole becomes massless. The electric-variable effective action then appears singular; in a magnetic frame the light monopole is included explicitly Seiberg and Witten 1994, pp. 19–52, arXiv:hep-th/9407087.

This example illustrates all four layers: the charge lattice allows many $(p,q)$ ; the algebra fixes $M=|qa+pa_D|$ for BPS states; semiclassics or exact arguments establish existence in a chamber; and wall crossing controls how the protected spectrum changes between chambers.

From the algebra to a projector

For a BPS particle of phase $e^{i\varphi}=Z/|Z|$ , the null supercharge combination can be written schematically as

\left(Q^1_\alpha -e^{i\varphi}\varepsilon_{\alpha\beta} (Q^2_\beta)^\dagger\right)|\mathrm{BPS}\rangle=0,

up to the spinor and phase convention used to diagonalize the algebra. In a classical soliton background the same condition appears by setting the corresponding fermion variation to zero. This is why completing the energy into squares and solving a preserved-supercharge projector lead to the same first-order equations on the BPS solitons page.

Common pitfalls

Treating the bound as an existence theorem. $M\ge|Z|$ constrains states already present in the Hilbert space. It does not assert a BPS particle for every $\gamma\in\Gamma$ .

Suppressing normalization. The coefficient relating $M$ and $|Z|$ depends on how $Z$ enters the algebra. Display the algebra before quoting a numerical bound.

Ignoring the chamber. A BPS index is locally constant only away from walls. A charge and central charge without a vacuum and chamber do not specify a spectrum.

Exercises

Let $Z_1=2$ and $Z_2=e^{i\theta}$ in fixed mass units, with $Z=Z_1+Z_2$ .

Show that $|Z|\le |Z_1|+|Z_2|$ and determine when equality holds.
Interpret the equality as a possible marginal decay condition.
Explain why it does not prove that a bound state exists on either side.

Solution

Directly, $|Z|^2=5+4\cos\theta\le9$ , hence $|Z|\le3=|Z_1|+|Z_2|$ . Equality holds at $\theta=0$ modulo $2\pi$ , when the phases align. At that locus a BPS particle of charge $\gamma_1+\gamma_2$ can have the same mass as two separated BPS constituents, so decay is marginal. Existence and which side supports a normalizable bound state depend on interactions and the charge pairing; the triangle inequality alone supplies neither fact.

References

Seiberg, Nathan, and Edward Witten. “Electric–Magnetic Duality, Monopole Condensation, and Confinement in $\mathcal N=2$ Supersymmetric Yang–Mills Theory.” Nuclear Physics B 426 (1994): 19–52; erratum 430 (1994): 485–486. arXiv:hep-th/9407087.
Witten, Edward, and David Olive. “Supersymmetry Algebras That Include Topological Charges.” Physics Letters B 78 (1978): 97–101. doi:10.1016/0370-2693(78)90357-X.