Gauss–Bonnet Black Branes in Lovelock Gravity

• Gauss–Bonnet black branes are extended black objects in Lovelock gravity, defined by the inclusion of the Gauss–Bonnet term that alters horizon geometry and thermodynamic properties.
• They manifest in various settings—such as asymptotically flat, AdS, and Lifshitz spacetimes—with distinct features like modified entropy, tension, and transport coefficients.
• Critical aspects including stability thresholds, existence bounds, and phase transitions emerge from the nontrivial interplay between higher-curvature corrections and black brane dynamics.

Gauss–Bonnet black branes are extended black objects in higher-curvature gravity theories whose action contains the Gauss–Bonnet density, most commonly in Einstein–Gauss–Bonnet gravity and, more generally, in Lovelock theories. In dimensions $D>4$, the Gauss–Bonnet combination contributes nontrivially to the field equations while preserving their second-order character, and black branes then appear in several distinct but related settings: asymptotically flat Kaluza–Klein black strings, planar anti-de Sitter black branes, Lifshitz branes, rotating and charged branes, and string-like bulk extensions of lower-dimensional black holes. Across these realizations, the higher-curvature coupling modifies horizon geometry, entropy, tension, transport coefficients, perturbative spectra, and, in some sectors, the very domain of existence of solutions (Kleihaus et al., 2012, Jacobson et al., 2011).

1. Field-theoretic framework and basic geometry

In the standard Einstein–Gauss–Bonnet formulation, the bulk action is

$$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$

For $D\ge 5$, the Gauss–Bonnet term is dynamical; in $D=4$ it is topological in the ordinary metric formulation, while several lower-dimensional constructions instead use scalar–tensor completions or dimensional regularizations (Jacobson et al., 2011, Hennigar et al., 2020).

A black brane is an extended black object with a translationally invariant worldvolume. The simplest case is the black string, with one extended spatial direction. In the asymptotically flat Kaluza–Klein setting, the spacetime approaches ${\cal M}_{d-1}\times S^1$, whereas in the anti-de Sitter setting one often considers planar horizons, $k=0$, with metric ansatz

$$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$

or its planar specialization $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$ (Jacobson et al., 2011). In the asymptotically flat spinning-string sector, the $D=5$ uniform black string solutions are stationary, axisymmetric, and uniform along $z$, with three commuting Killing vectors $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$0, $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$1, and $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$2, and horizon topology $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$3 (Kleihaus et al., 2012).

The geometric role of the Gauss–Bonnet term is not merely perturbative. In pure Einstein gravity, a trivial direct product of a lower-dimensional black hole with a flat extra direction is often sufficient to generate a black string. In Einstein–Gauss–Bonnet gravity, this simple factorization generally fails, and one must solve for a genuinely deformed extended geometry. This is already visible in the asymptotically flat uniform-string ansatz

$$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$4

where the metric function $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$5 along the extended direction becomes nontrivial once the Gauss–Bonnet term is present (Brihaye et al., 2010, Suzuki et al., 2022).

2. Asymptotically flat black strings and spinning extensions

The canonical asymptotically flat sector consists of uniform black strings in vacuum Einstein–Gauss–Bonnet gravity. Nonperturbative solutions were constructed for $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$6, together with perturbative small-$$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$7 expansions around the Einstein limit. These strings are asymptotically $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$8, possess mass $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$9, tension $D\ge 5$0, entropy $D\ge 5$1, and Hawking temperature $D\ge 5$2, and obey the first law

$D\ge 5$3

For the static uniform branch, the relative tension

$D\ge 5$4

ceases to be fixed at its Einstein value $D\ge 5$5; its variation with $D\ge 5$6 is one of the simplest signatures of Gauss–Bonnet backreaction on extended horizons (Brihaye et al., 2010).

A distinguished five-dimensional subset is the family of spinning uniform black strings. These generalize the $D\ge 5$7 Kerr black string by adding the Gauss–Bonnet density to the action. They share many qualitative properties with their Einstein counterparts and satisfy the generalized Smarr relation

$D\ge 5$8

At the same time, they exhibit a hard existence bound: solutions exist only up to a maximal Gauss–Bonnet coupling $D\ge 5$9, which depends on the horizon scale and the spin. In the static limit, the dimensionless parameter $D=4$0 reaches $D=4$1; at zero spin this implies $D=4$2. For fixed $D=4$3, increasing $D=4$4 decreases the relative tension $D=4$5 and the reduced area $D=4$6, while increasing reduced temperature $D=4$7, reduced entropy $D=4$8, and reduced spin $D=4$9. The metric component ${\cal M}_{d-1}\times S^1$0 also develops nontrivial angular dependence when ${\cal M}_{d-1}\times S^1$1 (Kleihaus et al., 2012).

Large-coupling asymptotics reveal an additional structural feature. In the regime of large dimensionless coupling ${\cal M}_{d-1}\times S^1$2, asymptotically flat Einstein–Gauss–Bonnet black strings separate into a Gauss–Bonnet region near the horizon, a general-relativistic region far away, and a transition region where the two contributions are comparable. For ${\cal M}_{d-1}\times S^1$3, the transition scale is

${\cal M}_{d-1}\times S^1$4

Within this large-${\cal M}_{d-1}\times S^1$5 approximation, one finds ${\cal M}_{d-1}\times S^1$6 for the relative tension, together with analytic leading-order formulas for ${\cal M}_{d-1}\times S^1$7, ${\cal M}_{d-1}\times S^1$8, ${\cal M}_{d-1}\times S^1$9, and $k=0$0, all confirmed numerically (Suzuki et al., 2022).

3. Planar AdS branes, holography, and transport

The best-studied AdS sector is the planar black brane in five-dimensional Gauss–Bonnet gravity with negative cosmological constant. A standard metric is

$k=0$1

with

$k=0$2

The horizon entropy density is

$k=0$3

and the Hawking temperature is

$k=0$4

This geometry underlies most holographic analyses of finite-coupling corrections in Gauss–Bonnet gravity (Buchel et al., 17 Jun 2026).

Jacobson, Mohd, and Sarkar developed a membrane-paradigm description for Einstein–Gauss–Bonnet black objects in $k=0$5, deriving a horizon-fluid stress tensor and the corresponding transport coefficients. For constant-curvature horizons, the horizon stress tensor takes the viscous-fluid form

$k=0$6

with a negative bulk viscosity and unchanged ratio

$k=0$7

For planar AdS black branes,

$k=0$8

which in $k=0$9 reduces to

$$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$0

This agrees with the Kubo-formula result from AdS/CFT and makes explicit how positive Gauss–Bonnet coupling lowers the shear-viscosity-to-entropy-density ratio below the Einstein value (Jacobson et al., 2011).

Perturbative dynamics of AdS Gauss–Bonnet black branes are likewise non-Einstein in two distinct ways. The quasinormal spectrum contains a perturbative branch, continuously connected to Schwarzschild–AdS as $$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$1, and a non-perturbative purely imaginary branch whose damping diverges in the Einstein limit. In the planar regime, the only instability found in the stable parameter sector is the eikonal one. For $$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$2, the brane is stable for

$$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$3

and becomes unstable outside that interval; near the threshold, the relaxation time becomes anomalously large (Konoplya et al., 2017).

A later five-dimensional analysis sharpened the relation between instability and boundary causality. It showed that Gauss–Bonnet black branes are unstable whenever the coupling lies outside the conformal collider bounds

$$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$4

with the helicity-2 sector controlling the upper bound and the helicity-0 sector the lower bound. Independently, hyperbolicity outside the horizon requires

$$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$5

The unstable modes at imaginary momentum and the causality-violating modes at real momentum are related by a phase rotation in complex momentum space and a Lorentz boost, making instability and superluminal propagation two aspects of the same mode structure in the Gauss–Bonnet truncation (Buchel et al., 17 Jun 2026).

For asymptotically AdS black strings and branes more generally, the boundary-counterterm program provides finite actions, conserved charges, and holographic stress tensors up to $$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$6. In this framework, planar $$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$7 horizons are special: the Gauss–Bonnet correction to the horizon entropy vanishes, so the entropy remains strictly area-proportional, even though the effective AdS radius and boundary stress tensor still carry explicit $$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$8-dependence (0806.1396).

4. Charged, Lifshitz, and other matter-coupled brane families

Several matter-coupled branches show that Gauss–Bonnet black branes are not restricted to neutral AdS or asymptotically flat sectors. One important class consists of rotating AdS black branes in Gauss–Bonnet gravity coupled to a power-Maxwell invariant source. In $$ds^2 = - f(r)\, dt^2 + \frac{dr^2}{f(r)} + r^2\, d\Sigma_{k,D-2}^2,$$9 dimensions with $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$0 independent rotation parameters $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$1, the solutions have planar horizons, a gauge potential aligned with the co-rotating Killing one-form, and conserved quantities

$d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$2

while the entropy obeys the area law,

$d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$3

The first law holds in the form

$d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$4

The power $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$5 in the nonlinear electromagnetic Lagrangian controls the asymptotics: for $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$6 or $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$7, including the conformal case $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$8, the solutions are asymptotically AdS; for $d\Sigma_{0,D-2}^2 = d\vec x^{\,2}$9, the gauge sector dominates asymptotically and the spacetime is not asymptotically AdS (Hendi et al., 2010).

A closely related but distinct family arises in $D=5$0 Gauss–Bonnet gravity coupled to a conformally invariant Maxwell source. The rotating planar AdS black branes admit up to $D=5$1 rotation parameters and obey

$D=5$2

together with

$D=5$3

These solutions satisfy the first law, possess an extremal mass $D=5$4, are locally stable in both canonical and grand-canonical ensembles, and exhibit no Hawking–Page transition for planar horizons (Hendi, 2010).

Non-AdS scaling geometries are realized by Gauss–Bonnet–dilaton Lifshitz black branes. In $D=5$5-dimensional spacetime, with metric

$D=5$6

the entropy density for planar horizons remains

$D=5$7

while the conserved quantity built from radial first integrals leads to the generalized Smarr formula

$D=5$8

Only one asymptotic integration constant contributes to the energy density, which the authors interpret as a no-hair statement for this sector, and the heat capacity is positive,

$D=5$9

so the solutions are thermally stable (Zangeneh et al., 2015).

There are also lower-dimensional and braneworld realizations. Codimension-2 braneworlds with a five-dimensional Gauss–Bonnet bulk and induced gravity on a 2-brane admit BTZ-like black holes on the brane that extend into the bulk as regular black strings, with the fine-tuning $z$0 playing a central role in horizon regularity (Cuadros-Melgar et al., 2010). In a different lower-dimensional scalar–tensor realization of Gauss–Bonnet gravity, rotating four-dimensional black strings with AdS asymptotics were also obtained; their thermodynamics uses an effective AdS radius $z$1, and the first law and Smarr relation remain valid, but rotation explicitly breaks the static thermodynamic universality of the corresponding charged BTZ sector (Hennigar et al., 2020).

5. Perturbations, instabilities, and phase structure

The stability theory of Gauss–Bonnet black branes is markedly dimension-dependent. For asymptotically flat Einstein–Gauss–Bonnet uniform black strings in $z$2, linearized Gregory–Laflamme analysis shows that $z$3 is special: for sufficiently large $z$4, the threshold eigenvalue $z$5 becomes negative, so the uniform string is dynamically stable. In the other dimensions studied, the strings remain dynamically unstable and retain negative specific heat. The $z$6 system therefore gives an explicit realization of the Gubser–Mitra conjecture, with the dynamically stable branch coinciding numerically with the thermodynamically stable branch of positive heat capacity (Brihaye et al., 2010).

The same work also constructed nonuniform black strings in six-dimensional Einstein–Gauss–Bonnet gravity. These emerge from the Gregory–Laflamme threshold, are more entropic and cooler than the corresponding critical uniform strings, and become less nonuniform as $z$7 increases at fixed temperature. This was interpreted as a smoothing effect of the Gauss–Bonnet term on the nonuniform branch (Brihaye et al., 2010).

Pure Gauss–Bonnet theory exhibits an analogous instability in a different guise. In seven dimensions, the exact uniform black string obtained by oxidizing a six-dimensional pure Gauss–Bonnet black hole admits unstable s-wave perturbations governed by a single master equation. The instability has the standard Gregory–Laflamme band structure: a discrete dispersion relation $z$8, a critical wavelength, and exponentially growing modes below threshold. Since the same background also has negative specific heat, this sector again supports the Gubser–Mitra correspondence, now in pure quadratic Lovelock gravity rather than Einstein–Gauss–Bonnet theory (Giacomini et al., 2015).

Codimension-2 braneworld strings reveal a different perturbative pattern. The Gauss–Bonnet term modifies the Lichnerowicz operator itself,

$z$9

so linearized perturbations are sourced by explicit Gauss–Bonnet curvature terms. For the five-dimensional Gauss–Bonnet black string studied in this framework, scalar s-wave modes are stable away from a Chern–Simons-type strong-coupling limit, while vector and tensor sectors become degenerate or inconclusive within the chosen transverse-traceless ansatz (Cuadros-Melgar et al., 2010).

At large dimension, the black-hole/black-string merger inherits nontrivial Gauss–Bonnet dependence. When $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$00, the merger geometry is governed by the same leading logarithmic-diffusion equation as in Einstein gravity and approaches the Einstein–Gauss–Bonnet black hole away from the neck. For the larger but still subordinate regime $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$01, the leading flow equation is modified and the merger geometry no longer asymptotes to the black-hole geometry away from the neck. This suggests that a topology-changing transition may still occur, but not obviously as a direct interpolation to a localized black hole in the same sense as the Einstein case (Nair et al., 2021).

6. Existence bounds, extremality, and global constraints

One of the most characteristic features of Gauss–Bonnet black branes is the appearance of existence bounds in coupling space. For the five-dimensional spinning uniform black strings, numerical continuation stops at a maximal $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$02, signaled not by an observed curvature singularity but by the loss of real roots in a quadratic equation for near-horizon coefficients. The extremal limit is also atypical: although the upper branch of rotating solutions approaches very low temperatures as $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$03, a near-horizon analysis with $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$04 symmetry produces logarithmic singularities in the Gauss–Bonnet corrections, and nonperturbative attempts to construct a regular extremal solution fail to converge. The available evidence therefore points to the absence of regular extremal spinning uniform black strings in this sector (Kleihaus et al., 2012).

Static higher-dimensional black strings show other dimension-specific bounds. In $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$05, regularity near the horizon implies

$$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$06

so for positive $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$07 one obtains a minimal horizon radius $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$08 and hence a mass gap. For $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$09, there is no analogous upper bound on positive $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$10, although a negative lower bound $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$11 appears in the near-horizon analysis (Brihaye et al., 2010).

Product-topology solutions make the sensitivity of Gauss–Bonnet gravity to horizon Weyl curvature especially explicit. For static black holes with $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$12 topology, which encompasses black strings, branes, and generalized Nariai geometries, the Gauss–Bonnet and Einstein–Gauss–Bonnet field equations reduce to algebraic square-root forms containing a negative constant induced by the product-sphere Weyl curvature. In pure Gauss–Bonnet gravity this generates a non-central naked singularity unless the mass lies in a $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$13-dependent window; in Einstein–Gauss–Bonnet gravity, a limited window of negative $$S = \frac{1}{16\pi G} \int d^D x \,\sqrt{-g}\,\Big[\,R - 2\Lambda + \alpha\,(R_{abcd}R^{abcd} - 4R_{ab}R^{ab} + R^2)\,\Big].$$14 is also allowed. This topological sector therefore provides a clean example in which horizon topology and higher-curvature terms jointly determine the global regularity of the spacetime (Pons et al., 2014).

Taken together, these results suggest a consistent pattern. Gauss–Bonnet corrections do not merely shift black-brane parameters quantitatively; they often introduce qualitatively new constraints: maximal couplings, minimal horizon radii, altered extremal behavior, modified Gregory–Laflamme thresholds, and holographic causality windows. At the same time, several open problems remain explicit in the literature: closed-form uniform-string solutions in full Einstein–Gauss–Bonnet theory are still unavailable in the generic case; the stability of spinning five-dimensional uniform strings was not analyzed; nonuniform branches beyond the known six-dimensional sector remain incomplete; and higher-dimensional, matter-coupled, or holographic generalizations continue to require either numerical construction or matched asymptotics rather than exact solutions (Kleihaus et al., 2012, Suzuki et al., 2022).