Rotating Sen Black Hole Overview

• The rotating Sen black hole is a charged, rotating solution in low-energy heterotic string theory, uniquely supported by graviton, dilaton, axion, and U(1) gauge fields.
• Its metric formulations include the exact Kerr–Sen solution and Newman–Janis algorithm variants, revealing complex horizon, ergoregion, and shadow properties with quantifiable observational differences.
• Studies on geodesics, accretion disks, and holographic duality demonstrate how Sen charge parameters influence radiative efficiency, photon trajectories, and superradiant instabilities.

The rotating Sen black hole, usually identified with the Kerr–Sen solution, is a rotating charged black hole arising in the four-dimensional low-energy effective theory of heterotic string theory. In this setting the black hole carries external hair associated with the graviton, a $U(1)$ gauge field, the dilaton, and the axion obtained from the Kalb–Ramond sector. In recent phenomenological work, the same label has also been used for a rotating metric generated from a static Sen-inspired seed by the Newman–Janis algorithm (NJA), with an effective charge parameter $q_m$ and a radially varying mass function. The subject therefore includes both the exact Kerr–Sen geometry and a broader family of rotating Sen-like constructions used to study horizons, geodesics, thin accretion disks, black-hole shadows, polarization, lensing, superradiance, and holographic properties (Narang et al., 2020, Cai et al., 1 Sep 2025).

1. String-theoretic origin and field content

In the low-energy heterotic-string description, the rotating Sen black hole is not a vacuum solution of Einstein gravity. It is supported by coupled bosonic fields. In four dimensions these are the graviton $g_{\mu\nu}$, a $U(1)$ gauge field $A_\mu$, the Kalb–Ramond 3-form $B_{\mu\nu\lambda}$ or its pseudoscalar axion representative, and the dilaton. The effective action used in the literature appears in both heterotic-string and Einstein–Maxwell–dilaton–axion (EMDA) forms, depending on whether one emphasizes the original string-theory derivation or its four-dimensional effective description (Narang et al., 2020, Cayuso et al., 2019, Sakti, 2022).

A standard exact rotating realization is the Kerr–Sen black hole. Its dyonic extension in EMDA carries both electric and magnetic charges, with dilaton and axion charges fixed by those electromagnetic parameters. In the notation used for the dyonic family,

$d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$

and many shadow properties depend on $Z_c^2 \equiv P^2+Q^2$ (Jana et al., 2023). Gauged and AdS versions introduce an AdS length $l$ and preserve the same basic EMDA field content while changing the horizon function and thermodynamics (Sakti, 2022, Sakti et al., 2022).

This field-theoretic origin distinguishes the rotating Sen black hole from Kerr–Newman. In Kerr–Newman the charge enters pure Einstein–Maxwell gravity, whereas in Kerr–Sen the charge is inseparable from dilaton and axion structure. Several observational proposals exploit precisely that distinction, especially the coupling between shadow morphology and axion-induced polarization rotation (Narang et al., 2020).

2. Metric formulations and parameterizations

A convenient Boyer–Lindquist–like form of the exact Kerr–Sen metric introduces

$$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$

The corresponding line element is

$q_m$0

In alternative but equivalent notations one often uses $q_m$1, so that $q_m$2 and the combination $q_m$3 becomes $q_m$4 (Narang et al., 2020, Guo et al., 2019).

A second formulation, central to recent phenomenological analyses, starts from a static Sen-inspired seed metric

$q_m$5

with

$q_m$6

Applying the NJA yields a rotating spacetime with Kerr-like structure but with

$q_m$7

$q_m$8

$q_m$9

In this construction the metric components are identical in form to Kerr but with $g_{\mu\nu}$0. The Kerr limit is recovered continuously as $g_{\mu\nu}$1, for which $g_{\mu\nu}$2 and $g_{\mu\nu}$3 (Cai et al., 1 Sep 2025).

A central conceptual point is that these two constructions are not equivalent in status. The exact Kerr–Sen black hole is a bona fide heterotic-string solution. By contrast, the NJA-derived rotating Sen metric with $g_{\mu\nu}$4 is described as a phenomenological solution and is not guaranteed to solve the full low-energy heterotic string field equations with consistent dilaton and axion profiles for arbitrary rotation (Cai et al., 1 Sep 2025).

3. Horizons, ergoregions, extremality, and integrability

For the exact Kerr–Sen geometry, the horizons are

$g_{\mu\nu}$5

or, in the $g_{\mu\nu}$6 notation,

$g_{\mu\nu}$7

Regular horizons require the corresponding horizon-regularity bound, and extremality occurs when the two roots coincide (Narang et al., 2020, Uniyal et al., 2017). A characteristic feature emphasized in the literature is the increased charge capacity of Kerr–Sen relative to Kerr–Newman before the solution becomes naked (Narang et al., 2020).

The ergosurface is defined by $g_{\mu\nu}$8. In the exact Kerr–Sen family this gives the standard stationary-limit surface modified by the stringy charge. In the NJA-derived rotating Sen metric the horizon equation is

$g_{\mu\nu}$9

while the ergosurface is determined by

$U(1)$0

Numerically, increasing either $U(1)$1 or $U(1)$2 shrinks $U(1)$3; the effective charge also reduces the equatorial radius of the ergosurface (Cai et al., 1 Sep 2025).

Geodesic motion retains a high degree of integrability. For exact Kerr–Sen, null geodesics are completely integrable using the Hamilton–Jacobi method, with conserved energy $U(1)$4, azimuthal angular momentum $U(1)$5, and a Carter-like constant $U(1)$6. In one common notation,

$U(1)$7

$U(1)$8

where $U(1)$9 and $A_\mu$0 (Narang et al., 2020). For the NJA-derived rotating Sen metric, null geodesics also separate via the Hamilton–Jacobi method with a conserved Carter constant $A_\mu$1 (Cai et al., 1 Sep 2025).

This separability extends beyond null and timelike geodesics. In the string frame, Kerr–Sen admits a principal tensor with torsion, and the massive vector (Proca) equations can be fully separated by exploiting that hidden symmetry (Cayuso et al., 2019). That property places rotating Sen black holes alongside Kerr and Kerr–Newman in the class of rotating spacetimes with unusually rich hidden symmetry structure.

4. Circular orbits, ISCO structure, and photon trajectories

On the equatorial plane of the NJA-derived rotating Sen metric, circular timelike geodesics are described by the orbital frequency

$A_\mu$2

together with the specific energy and angular momentum

$A_\mu$3

The ISCO follows from the marginal-stability condition $A_\mu$4. For fixed $A_\mu$5, increasing $A_\mu$6 moves $A_\mu$7 inward and raises the Novikov–Thorne radiative efficiency

$A_\mu$8

The same trend is reported in the exact Kerr–Sen literature: charge deforms the radial sector strongly enough to change circular-orbit locations and associated observables (Cai et al., 1 Sep 2025, Uniyal et al., 2017).

For exact Kerr–Sen geodesics, the radial and latitudinal effective potentials are modified asymmetrically by charge. The Sen charge affects the radial structure through the linear-in-$A_\mu$9 term that replaces the Kerr–Newman constant-$B_{\mu\nu\lambda}$0 deformation, while the latitudinal potential has no explicit Sen-charge dependence. Increasing $B_{\mu\nu\lambda}$1 narrows the allowed radial regions for null geodesics, and the range of spherical photon orbits shrinks accordingly (Uniyal et al., 2017).

The shadow boundary in both exact and NJA-derived formulations is generated by unstable spherical photon orbits. For the NJA metric the critical impact parameters are expressed as $B_{\mu\nu\lambda}$2 and $B_{\mu\nu\lambda}$3, and the celestial coordinates for an observer at inclination $B_{\mu\nu\lambda}$4 are

$B_{\mu\nu\lambda}$5

At fixed $B_{\mu\nu\lambda}$6 and $B_{\mu\nu\lambda}$7, the shadow diameter decreases monotonically with increasing $B_{\mu\nu\lambda}$8, while at very high spin, such as $B_{\mu\nu\lambda}$9, a characteristic D-shaped distortion appears (Cai et al., 1 Sep 2025).

A further exact-Kerr–Sen result is the absence of terminating photon orbits for $d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$0. The charge-induced structure blocks null geodesics from reaching the ring singularity under the required simultaneous radial and latitudinal conditions (Uniyal et al., 2017).

5. Thin accretion disks, shadow observables, and EHT bounds

The NJA-derived rotating Sen black hole has been analyzed in the Novikov–Thorne thin-disk framework. For a steady, geometrically thin and radiatively efficient equatorial disk, the flux is written in Page–Thorne form as

$d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$1

or, in the paper’s spherical-coordinate implementation,

$d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$2

The effective temperature is then

$d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$3

and the observed spectral luminosity is computed from a multicolor blackbody integral with redshift factor

$d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$4

At fixed $d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$5, increasing $d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$6 raises $d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$7, shifts its peak inward, increases $d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$8, and boosts $d=\frac{(P^2-Q^2)}{2M}, \qquad k=\frac{PQ}{M},$9. At fixed $Z_c^2 \equiv P^2+Q^2$0, increasing $Z_c^2 \equiv P^2+Q^2$1 lowers the peak flux, temperature, and spectral luminosity (Cai et al., 1 Sep 2025).

The shadow analysis uses two observables: the shadow radius $Z_c^2 \equiv P^2+Q^2$2 and the distortion parameter $Z_c^2 \equiv P^2+Q^2$3. Increasing $Z_c^2 \equiv P^2+Q^2$4 reduces $Z_c^2 \equiv P^2+Q^2$5 and increases $Z_c^2 \equiv P^2+Q^2$6. Comparison with EHT diameter data gives explicit parameter windows. For M87$Z_c^2 \equiv P^2+Q^2$7, adopting $Z_c^2 \equiv P^2+Q^2$8 Mpc and $Z_c^2 \equiv P^2+Q^2$9, the reported bounds are:

• for $l$0: $l$1 and $l$2,
• for $l$3: $l$4 and $l$5.

For Sgr A$l$6, adopting $l$7, $l$8 kpc, and $l$9, the reported bounds are:

• for $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$0: $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$1 and $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$2,
• for $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$3: $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$4 and $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$5.

Combining both sources and the inclinations analyzed gives the shared range

$$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$6

The analysis also identifies a covariance among $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$7: within the allowed domain, increasing $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$8 can be partially compensated by decreasing $$\Delta \equiv r(r + r_{0}) - 2Mr + a^2, \qquad \rho^2 \equiv r(r + r_{0}) + a^2 \cos^2\theta, \qquad r_{0} \equiv \frac{Q^2}{M}.$$9 or varying $q_m$00 to match a given angular diameter (Cai et al., 1 Sep 2025).

6. Observational discriminants beyond size-only shadows

A distinctive exact-Kerr–Sen proposal combines shadow shape with axion-induced polarization rotation. In this framework the deviation from circularity constrains the charge through the shadow, while the same parameters determine a frequency-independent rotation of the polarization plane produced by axion hair. The rotation angle is proportional to the change in the axion field $q_m$01 along the photon trajectory, and the crucial claim is that in Kerr–Sen the shadow deformation and the polarization rotation are tied to the same underlying parameters $q_m$02. That correlation is absent in Kerr–Newman with independent axion hair (Narang et al., 2020).

Current EHT data are not yet decisive for that test. For M87$q_m$03 the bound $q_m$04 does not sharply constrain $q_m$05 in Kerr–Sen, whereas a refinement to $q_m$06 would enable a definitive inference of $q_m$07 and therefore the $q_m$08 charge. The same work emphasizes that frequency-independent polarization rotation would favor an axionic-hair interpretation over plasma Faraday rotation, which scales as $q_m$09 (Narang et al., 2020).

Time-dependent optical appearance has also been studied for near-extremal Kerr–Sen. For an isotropically emitting hot spot on a near-horizon circular orbit, the leading-order image locus is a vertical line at

$q_m$10

with a charge-dependent vertical extent $q_m$11. The orbital period remains

$q_m$12

while the typical redshift of secondary images increases with the reduced twist $q_m$13. Relative to Kerr–Newman, the Kerr–Sen case differs quantitatively because the image position $q_m$14 does not shift with charge at leading order, whereas in near-extremal Kerr–Newman it does (Guo et al., 2019).

For dyonic Kerr–Sen black holes, shadow analyses use the deviation from circularity $q_m$15 and the fractional deviation

$q_m$16

The static shadow radius satisfies

$q_m$17

and there is no stable photon orbit outside the horizon. The current M87$q_m$18 bound on $q_m$19 does not constrain the theoretically allowed parameter region, but the Sgr A$q_m$20 bound on $q_m$21 implies

$q_m$22

and therefore, if $q_m$23 is negligible or bounded independently,

$q_m$24

This is a bound on the combined dyonic scale rather than on electric or magnetic charge separately (Jana et al., 2023).

Weak-field lensing gives a complementary signature. In the material-medium approach to Kerr–Sen, prograde trajectories show larger refractive index and larger deflection than retrograde ones because of frame dragging, while increasing the charge/dilaton parameter suppresses the refractive index and the weak-field deflection for both prograde and retrograde motion (Roy et al., 16 Apr 2025).

7. Dynamical probes, AdS extensions, and holography

Rotating Sen black holes support several nontrivial dynamical phenomena beyond ray optics. One is superradiance of massive vector fields. In Kerr–Sen the Proca equations can be fully separated in the string frame using the torsionful hidden symmetry, and the resulting radial equation differs from Kerr and Kerr–Newman by stringy terms involving the parameter $q_m$25. Numerically, turning on charge decreases the superradiant growth rate relative to Kerr, but Kerr–Sen remains slightly more unstable than Kerr–Newman for the same $q_m$26 and $q_m$27 (Cayuso et al., 2019).

A second development concerns Kerr/CFT and hidden conformal symmetry. For dyonic Kerr–Sen and its gauged family, a neutral massless scalar in the low-frequency near-region exhibits a solution-space $q_m$28 symmetry. The associated left- and right-moving temperatures reproduce the Bekenstein–Hawking entropy through the Cardy formula, and absorption cross-sections match the dual CFT calculation. In the gauged case the same structure survives in the near-horizon near-extremal regime, with $q_m$29 replaced by a second near-horizon root $q_m$30 and AdS factors $q_m$31 and $q_m$32 entering the temperatures (Sakti, 2022).

The extremal Kerr/CFT formulation for dyonic Kerr–Sen gives two ungauged branches of the central charge,

$q_m$33

corresponding to the two branches of the extremal mass relation. The duality persists for the gauged AdS family and for the ultraspinning limit. Unlike ultraspinning dyonic Kerr–Newman–AdS, the ultraspinning dyonic Kerr–Sen–AdS black hole is not always superentropic; whether the reverse isoperimetric inequality is violated depends on the dyonic parameters through the dilaton and axion charges (Sakti et al., 2022).

Holographic chaos has also been analyzed for dyonic Kerr–Sen–AdS$q_m$34. In that setting the scrambling time for a large-entropy black hole has a leading logarithmic dependence on the entropy, supporting fast scrambling. The relevant bound for rotating shocks is

$q_m$35

and the scrambling delay produced by electrically and magnetically charged shocks depends on the chemical potentials $q_m$36 and $q_m$37. The same formalism extends to the ultra-spinning geometry (Prihadi et al., 2023).

These developments indicate that the rotating Sen black hole is not merely a charged deformation of Kerr. In exact, dyonic, gauged, and phenomenological NJA-derived realizations, it serves as a testing ground for how dilaton–axion structure modifies strong-field geometry, radiation, scattering, and holographic duality. At the same time, a methodological distinction remains essential: the exact Kerr–Sen family is a heterotic-string solution, whereas the NJA-derived rotating Sen metric with $q_m$38 is a useful but phenomenological construction whose field-equation status is limited (Cai et al., 1 Sep 2025).