Self-Interacting Kalb-Ramond Field

• The self-interacting Kalb-Ramond field is a 2-form gauge potential featuring nonlinear self-interactions via potentials and nonminimal curvature couplings.
• It combines mechanisms like Higgs-generated masses, BF* mixing, and T-dual string dynamics to yield effects such as confinement, torsion, and Lorentz violation.
• Its versatile formulations, including dual scalar representations and emergent lattice models, offer practical insights into black-hole physics and modified gravity.

Self-interacting Kalb–Ramond field denotes a class of theories built around an antisymmetric rank-2 gauge field $B_{\mu\nu}=-B_{\nu\mu}$ in which the Kalb–Ramond sector acquires nonlinear structure beyond a purely free 2-form gauge theory. In the literature considered here, that nonlinear structure appears in several distinct ways: through an explicit self-interaction potential $V(B^{\mu\nu}B_{\mu\nu})$ combined with nonminimal curvature couplings, through Higgs-generated masses and $BF^\ast$ mixing that induce confining effective interactions, and through effective background dependence on T-dual coordinates in string-derived theories. The common gauge-invariant object is the 3-form field strength $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$, but the physical realizations range from confinement and torsional geometry to Lorentz-violating black-hole hair and modified conservative binary dynamics (Smailagic et al., 2020, Kumar et al., 2020, Fathi et al., 17 Jan 2025).

1. Field content, gauge structure, and source couplings

The Kalb–Ramond field is a 2-form gauge potential. In four-dimensional continuum formulations, its field strength is typically written as

$$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$

or equivalently as $\partial_{[\mu}B_{\nu\rho]}$. A standard free bulk term is quadratic in the field strength, as in

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$

or, in differential-form language,

$$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$

The gauge symmetry is the usual 2-form invariance,

$$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$

and in source-coupled formulations it requires an antisymmetric conserved current, such as

$\partial_\mu J^{\mu\nu}=0$

for a term $V(B^{\mu\nu}B_{\mu\nu})$0 (Barone et al., 2010).

A recurrent structural feature is that a 2-form does not couple to ordinary point-particle charge in the same way as a Maxwell field. The admissible conserved sources are instead extended distributions localized on branes or string-like objects. In the brane-interaction analysis, the source

$V(B^{\mu\nu}B_{\mu\nu})$1

is interpreted as a generalized dipole-like distribution, while

$V(B^{\mu\nu}B_{\mu\nu})$2

represents a Kalb–Ramond charge density on a brane (Barone et al., 2010).

The same field also admits dual descriptions. In four dimensions, one standard relation is

$V(B^{\mu\nu}B_{\mu\nu})$3

so the dual of the 3-form is a pseudoscalar. In the conservative gravity-plus-Kalb–Ramond binary problem, the 2-form is equivalently represented by a massless scalar $V(B^{\mu\nu}B_{\mu\nu})$4 through

$V(B^{\mu\nu}B_{\mu\nu})$5

which reduces the bulk action to a free scalar form (Bhattacharjee et al., 2010, Undheim et al., 2023).

2. Meanings of “self-interaction” in the literature

Taken together, the cited works show that “self-interacting Kalb–Ramond field” is not used in a single uniform sense. In some models it refers to an explicit self-interaction potential for $V(B^{\mu\nu}B_{\mu\nu})$6; in others it designates an induced or effective interaction structure generated after integrating out fields, solving constraints, or coupling the 2-form to gravity, matter, or lattice defects.

Usage	Realization	Representative papers
Explicit self-interaction	$V(B^{\mu\nu}B_{\mu\nu})$7 with nonminimal curvature couplings and a nonzero vacuum expectation value	(Kumar et al., 2020, Fathi et al., 17 Jan 2025)
Induced effective interaction	Local $V(B^{\mu\nu}B_{\mu\nu})$8–$V(B^{\mu\nu}B_{\mu\nu})$9 parent theory with $BF^\ast$0 and Higgs-generated masses	(Smailagic et al., 2020)
Effective nontrivial $BF^\ast$1-dynamics in string theory	$BF^\ast$2 depends on the T-dual coordinate $BF^\ast$3, so the antisymmetric term survives $BF^\ast$4-projection	(Davidović et al., 2011)
Non-intrinsic nonlinearities	Bulk KR sector remains quadratic; nonlinear effects arise from sources, gravity, compactness, or mixed sectors	(Barone et al., 2010, Undheim et al., 2023, Lozano et al., 2019)

This distinction is essential. Several papers explicitly state that they do not introduce intrinsic nonlinear self-interaction of the Kalb–Ramond field itself; instead, they study effective interactions mediated by the 2-form, or interactions induced after dualization, compactification, or coupling to other sectors (Barone et al., 2010, Undheim et al., 2023, Lozano et al., 2019).

3. Higgs-generated interaction sectors, $BF^\ast$5 mixing, and confinement

A concrete interacting Abelian realization contains both an ordinary gauge vector $BF^\ast$6 and a Kalb–Ramond tensor $BF^\ast$7, coupled through a $BF^\ast$8 term. The parent Euclidean Lagrangian is given schematically by

$BF^\ast$9

with the same parameter $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$0 acting both as a coupling and as the Kalb–Ramond mass. The formulation is local, but after integrating out one field it becomes a nonlocal effective theory for the other (Smailagic et al., 2020).

The decisive distinction is between a massive and a massless Kalb–Ramond sector. When $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$1 is massive, integrating it out produces a nonlocal effective electrodynamics containing the operator

$H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$2

and the static potential between two charges becomes

$H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$3

This is the Cornell potential: a Coulomb term plus a linear confining term. The linear rise is not added by hand; it follows from the propagator structure induced by a massive Kalb–Ramond field coupled through $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$4. In momentum-space language, the relevant operator behaves as

$H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$5

which in position space yields the Coulomb-plus-linear form (Smailagic et al., 2020).

The same model admits a Higgs interpretation with two scalar sectors: a charged complex Higgs field $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$6, coupled to the vector through

$H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$7

and a neutral scalar Higgs field $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$8, coupled to the Kalb–Ramond sector. The scalar potentials are chosen with opposite-sign quadratic terms so that only one field condenses at a time. This yields three vacua:

• Coulomb vacuum: no condensate; both fields remain massless.
• Yukawa vacuum: $H_{\mu\nu\rho}=\partial_{[\mu}B_{\nu\rho]}$9 condenses; $$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$0 becomes massive while $$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$1 remains massless.
• Cornell vacuum: $$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$2 condenses; $$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$3 becomes massive while $$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$4 remains massless.

The corresponding interaction energies are the unscreened Coulomb form,

$$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$5

the screened Coulomb or Yukawa form,

$$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$6

and the Cornell form

$$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$7

The model therefore ties confinement to the Kalb–Ramond mass, while photon mass produces screening. The paper emphasizes that the Higgs construction replaces an ad hoc mass parameter by a dynamically generated scale in the Kalb–Ramond sector (Smailagic et al., 2020).

4. Effective-string, T-dual, and torsional formulations

In weakly curved open-string backgrounds, the usual statement that the Kalb–Ramond field must vanish in an unoriented theory is shown to fail once one passes to the effective theory obtained after solving the boundary conditions. The starting background is

$$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$8

with infinitesimal linear field strength. Solving the boundary constraints yields an effective coordinate $$H_{\mu\nu\rho}\equiv \partial_\mu B_{\nu\rho}+\partial_\nu B_{\rho\mu}+\partial_\rho B_{\mu\nu},$$9, built from the $\partial_{[\mu}B_{\nu\rho]}$0-even part of the open-string coordinate, together with its T-dual partner

$\partial_{[\mu}B_{\nu\rho]}$1

To leading order,

$\partial_{[\mu}B_{\nu\rho]}$2

The effective antisymmetric background depends not on $\partial_{[\mu}B_{\nu\rho]}$3, but on the T-dual coordinate: $\partial_{[\mu}B_{\nu\rho]}$4 Because $\partial_{[\mu}B_{\nu\rho]}$5 is then $\partial_{[\mu}B_{\nu\rho]}$6-odd, the term $\partial_{[\mu}B_{\nu\rho]}$7 is overall $\partial_{[\mu}B_{\nu\rho]}$8-even and survives integration over the symmetric interval. The resulting effective action is

$\partial_{[\mu}B_{\nu\rho]}$9

From the effective world-sheet equations of motion, the Kalb–Ramond field is then identified with a torsion potential, through a generalized connection

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$0

This embeds the effective $\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$1-field in a non-Riemannian geometry with torsion (Davidović et al., 2011).

A related but distinct string-inspired line of work treats the Kalb–Ramond field as the source of spacetime torsion and augments the field strength by gauge terms. In the heterotic-string setting,

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$2

so the Bianchi identity becomes modified by Yang–Mills and gravitational Chern–Simons forms. For higher-form gauge couplings, the paper proposes a different class of augmentation,

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$3

with the corresponding gauge transformation

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$4

After dualization through $\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$5, the effective action contains the axion-like couplings

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$6

The parity-even term $\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$7 produces achromatic rotation of the plane of polarization of electromagnetic waves at high frequency, while the parity-violating term $\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$8 produces attenuation or amplification rather than simple rotation (Bhattacharjee et al., 2010).

5. Explicit self-interaction, Lorentz-breaking vacua, and black-hole hair

In gravity-coupled models, “self-interacting Kalb–Ramond field” is used in the direct sense of a 2-form subject to a self-interaction potential and nonminimal curvature couplings. One representative action is

$\frac{1}{12}H_{\mu\nu\rho}H^{\mu\nu\rho},$9

with a vacuum expectation value

$$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$0

The nonzero vacuum tensor spontaneously breaks local Lorentz symmetry and diffeomorphism symmetry. The resulting static hairy black hole has metric function

$$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$1

while the rotating generalization is Kerr-like with

$$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$2

The special limits are explicit: $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$3 gives Kerr, $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$4 gives a Kerr–Newman-like geometry with $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$5, and $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$6 returns the static power-law hairy solution. For fixed $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$7, there exists an extremal value $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$8; $$\mathcal L=-\frac{1}{6\kappa^2}\mathcal F^{\mu\nu\rho}\mathcal F_{\mu\nu\rho}.$$9 gives a nonextremal black hole, $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$0 an extremal one, and $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$1 no black hole (Kumar et al., 2020).

The same rotating solution modifies null geodesics and optical observables. The shadow becomes smaller and more distorted as either $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$2 or $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$3 increases. The paper defines the shadow area

$$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$4

and the oblateness

$$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$5

finding that $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$6 decreases and $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$7 increases with increasing Kalb–Ramond deformation. The M87* circularity deviation bound

$$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$8

mainly constrains $$B_{\mu\nu}\to B_{\mu\nu}+\partial_\mu \xi_\nu-\partial_\nu \xi_\mu,$$9, while the observed shadow angular diameter

$\partial_\mu J^{\mu\nu}=0$0

implies

$\partial_\mu J^{\mu\nu}=0$1

within the $\partial_\mu J^{\mu\nu}=0$2 region. The weak-field deflection angle is reduced relative to Kerr or Schwarzschild, with corrections depending on $\partial_\mu J^{\mu\nu}=0$3, $\partial_\mu J^{\mu\nu}=0$4, $\partial_\mu J^{\mu\nu}=0$5, and the impact parameter $\partial_\mu J^{\mu\nu}=0$6 (Kumar et al., 2020).

A more recent static model combines a self-interacting Kalb–Ramond background with a global monopole. The action contains

$\partial_\mu J^{\mu\nu}=0$7

with

$\partial_\mu J^{\mu\nu}=0$8

For the global monopole matter sector,

$\partial_\mu J^{\mu\nu}=0$9

the approximate lapse function becomes

$V(B^{\mu\nu}B_{\mu\nu})$00

The spacetime has a true central singularity, and the horizon structure depends on the sign of $V(B^{\mu\nu}B_{\mu\nu})$01: for $V(B^{\mu\nu}B_{\mu\nu})$02 there is one positive root, while for $V(B^{\mu\nu}B_{\mu\nu})$03 there are two positive roots. The thermodynamics departs from the standard area law; the entropy is

$V(B^{\mu\nu}B_{\mu\nu})$04

and the specific heat can change sign, so sufficiently large black holes become locally stable. Solar-system tests then require $V(B^{\mu\nu}B_{\mu\nu})$05 and $V(B^{\mu\nu}B_{\mu\nu})$06 to be very small, with individual bounds ranging from

$V(B^{\mu\nu}B_{\mu\nu})$07

depending on the observable channel (Fathi et al., 17 Jan 2025).

6. Extended sources, emergent realizations, and conceptual boundaries

The pure source-coupled Kalb–Ramond theory in flat spacetime is quadratic and does not by itself define an intrinsically self-interacting 2-form. Its novelty lies instead in the admissible source sector. For point-like branes in $V(B^{\mu\nu}B_{\mu\nu})$08, $V(B^{\mu\nu}B_{\mu\nu})$09, the dipole-like interaction energy is

$V(B^{\mu\nu}B_{\mu\nu})$10

which has the same tensorial form as an electromagnetic dipole-dipole interaction but with the opposite overall sign. Equal-sign Kalb–Ramond charge distributions therefore attract rather than repel. In the Cremmer–Scherk–Kalb–Ramond model,

$V(B^{\mu\nu}B_{\mu\nu})$11

the topological parameter $V(B^{\mu\nu}B_{\mu\nu})$12 generates Yukawa factors $V(B^{\mu\nu}B_{\mu\nu})$13, mixed tensor-vector source interactions, and a massive propagating mode, but the bulk theory remains quadratic (Barone et al., 2010).

An emergent condensed-matter realization places rotor variables on the faces of a $V(B^{\mu\nu}B_{\mu\nu})$14-dimensional cubic lattice. The antisymmetric tensor variables are

$V(B^{\mu\nu}B_{\mu\nu})$15

and the low-energy effective Hamiltonian is

$V(B^{\mu\nu}B_{\mu\nu})$16

This yields an emergent compact $V(B^{\mu\nu}B_{\mu\nu})$17 2-form gauge field, a string charge generated by Gauss-law violation, and a coupling to an electromagnetic sector localized on a D-brane-like lattice submanifold. The nonlinear-looking cosine terms are compactness-induced lattice interactions, not continuum self-couplings such as $V(B^{\mu\nu}B_{\mu\nu})$18 or $V(B^{\mu\nu}B_{\mu\nu})$19 (Lozano et al., 2019).

In the conservative binary problem with gravity, the Kalb–Ramond field again does not enter as a self-coupled two-form. After dualization to a massless scalar $V(B^{\mu\nu}B_{\mu\nu})$20, the worldline interaction is

$V(B^{\mu\nu}B_{\mu\nu})$21

The leading conservative potential is Coulomb-like,

$V(B^{\mu\nu}B_{\mu\nu})$22

and the 1PN effective binary Lagrangian contains velocity corrections, $V(B^{\mu\nu}B_{\mu\nu})$23 seagull terms, and mixed gravity–Kalb–Ramond contributions. Here the nonlinearities are induced by worldline couplings and the metric expansion rather than by a separate Kalb–Ramond self-potential (Undheim et al., 2023).

A persistent misconception is that any nontrivial Kalb–Ramond phenomenology automatically implies intrinsic self-interaction of the 2-form. The cited literature draws a sharper distinction. Explicit self-interaction appears in the gravity-coupled, Lorentz-breaking models with $V(B^{\mu\nu}B_{\mu\nu})$24 (Kumar et al., 2020, Fathi et al., 17 Jan 2025). By contrast, the Abelian confinement model, the string-effective T-dual construction, the brane-source analysis, the post-Newtonian binary EFT, and the lattice emergence program all show that substantial nonlinear or effective dynamics can arise even when the underlying Kalb–Ramond sector remains free or quadratic in the bulk (Smailagic et al., 2020, Davidović et al., 2011, Barone et al., 2010, Undheim et al., 2023, Lozano et al., 2019).