hep-th/0410158

Born-Infeld Black Holes in (A)dS Spaces

Rong-Gen Cai∗
CASPER, Department of Physics, Baylor University, Waco, TX76798-7316, USA
Institute of Theoretical Physics, Chinese Academy of Sciences, P.O. Box 2735, Beijing 100080, China

Da-Wei Pang†
Institute of Theoretical Physics, Chinese Academy of Sciences, P.O. Box 2735, Beijing 100080, China
and Graduate School of the Chinese Academy of Sciences

Anzhong Wang‡
CASPER, Department of Physics, Baylor University, Waco, TX76798-7316, USA

We study some exact solutions in a D(≥ 4)-dimensional Einstein-Born-Infeld theory with a cos-
mological constant. These solutions are asymptotically de Sitter or anti-de Sitter, depending on the
arXiv:hep-th/0410158v2 8 Dec 2004

sign of the cosmological constant. Black hole horizon and cosmological horizon in these spacetimes
can be a positive, zero or negative constant curvature hypersurface. We discuss the thermodynam-
ics associated with black hole horizon and cosmological horizon. In particular we find that for the
Born-Infeld black holes with Ricci flat or hyperbolic horizon in AdS space, they are always ther-
modynamically stable, and that for the case with a positive constant curvature, there is a critical
value for the Born-Infeld parameter, above which the black hole is also always thermodynamically
stable, and below which a unstable black hole phase appears. In addition, we show that although the
Born-Infeld electrodynamics is non-linear, both black hole horizon entropy and cosmological horizon
entropy can be expressed in terms of the Cardy-Verlinde formula. We also find a factorized solution
in the Einstein-Born-Infeld theory, which is a direct product of two constant curvature spaces: one
is a two-dimensional de Sitter or anti-de Sitter space, the other is a (D − 2)-dimensional positive,
zero or negative constant curvature space.

I. INTRODUCTION

Black holes in anti-de Sitter (AdS) and de Sitter (dS) spaces are quite different from their counterparts in asymp-
totically flat spaces. In AdS spaces, there are so-called topological black holes whose event horizon could be a positive,
zero or negative constant curvature surface [1]. In asymptotically dS spaces, there are no globally time-like Killing
vector and spatial infinity, there is still not a well defined approach to define conserved charges like mass and angular
momentum of asymptotically dS gravitational configurations [2, 3, 4]. Due to the AdS/CFT correspondence [5], it
was argued by Witten [6] that the thermodynamics of black holes in AdS spaces can be identified with that of dual
conformal field theory (CFT) residing on the boundary of the AdS space. In the sense of the dS/CFT correspon-
dence [7], thermodynamics of black hole horizon and of cosmological horizon in the asymptotically de Sitter spaces
might be related to that of dual CFTs. These motivate a surge of study of black holes in AdS and dS spaces.
For a (1 + 1)-dimensional conformal field theory, there is a well-known entropy formula, namely, the Cardy for-
mula [8]. In an elegant paper [9], in the spirit of AdS/CFT correspondence, Verlinde argued that there is a similar
entropy formula for CFTs in higher dimensions. The formula “derived” by Verlinde is called Cardy-Verlinde formula in
the literature. Indeed this formula has been checked to hold for CFTs with AdS gravity duals, such as Schwarzschild-
AdS black holes [9], Kerr-AdS black holes [10], Hyperbolic and charged black holes [11], Taub-Bolt-AdS instanton [12],
Kerr-Newmann-AdS black holes [13] and so on (see also [14]). In addition, it has been found that entropies of black
hole horizons [15] and cosmological horizons [16] in asymptotically dS spaces can also be expressed in terms of the
Cardy-Verlinde formula.
In 1934 Born and Infeld proposed a non-linear electrodynamics with the aim of obtaining a finite value for the
self-energy of a point-like charge [17]. Although it became less popular with the introduction of QED, in recent years
the Born-Infeld action has been occurring repeatedly with the development of superstring theory, where the dynamics
of D-branes, some soliton solutions of supergravity, is governed by the Born-Infeld action. For various motivations,
extending the Reissner-Nordström black hole solutions in Einstein-Maxwell theory to the charged black hole solutions

∗ e-mail address: cairg@[Link]

† e-mail address: pangdw@[Link]
‡ e-mail address: Anzhong Wang@[Link]
 2

in Einstein-Born-Infeld theory with/without a cosmological constant has attracted some attention in recent years, for
example, see [18, 19].
In this paper, we first generalize the exact solutions of spherically symmetric Born-Infeld black holes with a cosmo-
logical horizon in arbitrary dimensions recently given in [19] to the case where black hole horizon and/or cosmological
horizon is a positive, zero or negative constant curvature surface. We then in Secs. III and IV study thermodynamics
associated with black hole horizon and cosmological horizon and show that although the electrodynamics is non-linear,
entropies of black holes and cosmological horizon in the Einstein-Born-Infeld theory can still be expressed in terms
of the Cardy-Verlinde formula, which shows the latter is of some universality. In Sec. V we present a factorized
solution in the Einstein-Born-Infeld theory. The solution is a direct product of two constant curvature spaces: one is
a two-dimensional dS or AdS space and the other is a positive, zero or negative constant curvature space. The result
is summarized in Sec. VI.

II. BORN-INFELD BLACK HOLES IN (A)DS SPACES

Consider an (n + 1)-dimensional (n ≥ 3) Einstein-Born-Infeld action with a cosmological constant Λ

√
 
R − 2Λ
Z
S = dn+1 x −g + L(F ) , (2.1)
16πG
where R is scalar curvature and L(F ) is given by
s !
2 F µν Fµν
L(F ) = 4β 1− 1+ . (2.2)
2β 2

Here the constant β is called Born-Infeld parameter with dimension of mass. In the limit β → ∞, L(F ) reduces to
the standard Maxwell form
L(F ) = −F µν Fµν + O(F 4 ), (2.3)
while one has L(F ) → 0 as β → 0. In what follows we set 16πG = 1 for simplicity, where G is the Newton constant
in (n + 1)-dimensions.
The equations of motion of the electromagnetic field and the Einstein equations can be obtained by varying the
action with respect to the gauge field Aµ and the metric gµν , which yields
√
 
µν
−gF
∂µ  q  = 0, (2.4)
F µν F
1 + 2β 2µν

1 1 2Fµα Fν α
Rµν − Rgµν + Λgµν = gµν L(F ) + q , (2.5)
2 2 F µν F
1 + 2β 2µν

respectively. Here Rµν stands for Ricci tensor.

Suppose the spacetime metric is of the form
1
ds2 = −V (r)dt2 + dr2 + R2 (r)hij dxi dxj , (2.6)
V (r)
where V (r) and R(r) are two functions of the coordinate r only, and hij is a function of coordinates xi which
span an (n − 1)-dimensional hypersurface with constant scalar curvature (n − 1)(n − 2)k. Here k is a constant and
characterizes the hypersurface. Without loss of generality, one can take k to be ±1 and 0 such that the black hole
horizon or cosmological horizon in (2.6) can be a positive (elliptic), zero (flat) or negative (hyperbolic) constant
curvature hypersurface. For the metric (2.6), we have non-vanishing components of Ricci tensor
V ′′ V ′ R′
Rtt = − − (n − 1) , (2.7)
2 2R
V ′′ V ′ R′ V R′′
Rrr = − − (n − 1) − (n − 1) , (2.8)
 2 2R R 
n−2 1
Rij = k− [V (Rn−1 )′ ]′ δji , (2.9)
R2 (n − 1)Rn−1
 3

where a prime stands for the derivative with respect to the coordinate r.
In the static and symmetric background (2.6), the equation (2.4) can be satisfied by setting F µν to zero, except for
rt
F , which gives
p
(n − 1)(n − 2)βq
F rt = p , (2.10)
2
2β R 2n−2 + (n − 1)(n − 2)q 2
where qRis an integration constant relating to the electric charge of the solution. Defining the electric charge via
1 ∗
Q = 4π F dΩ, we have
p
(n − 1)(n − 2)ωn−1
Q= √ q, (2.11)
4π 2
where ωn−1 represents the volume of constant curvature hypersurface described by hij dxi dxj .
From Eqs. (2.7) and (2.8), we have R′′ (r) = 0, which has two solutions. One is R = r; the other is R = a, here a
is a constant. We first consider the case of R = r. The case of R = a will be discussed later. With R = r, solving the
equation (2.9) yields
4β 2
 
m 1
V (r) = k − n−2 + + r2
r n(n − 1) l2
√
2 2β
Z p
− 2β 2 r2n−2 + (n − 1)(n − 2)q 2 dr. (2.12)
(n − 1)rn−2
Here we have redefined the cosmological constant as Λ = −n(n − 1)/2l2 . It turns out that the integration in (2.12)
can be worked out and can be expressed by using a hypergeometric function,
4β 2
 
m 1
V (r) = k − n−2 + + r2
r n(n − 1) l2
√
2 2β p
− n−3
2β 2 r2n−2 + (n − 1)(n − 2)q 2
n(n − 1)r
2(n − 1)q 2 n − 2 1 3n − 4 (n − 1)(n − 2)q 2
+ 2 F1 [ , , , − ]. (2.13)
nr2n−4 2(n − 1) 2 2(n − 1) 2β 2 r2n−2
Here m is an integration constant, relating to the mass of the gravitational configuration. Since the metric is asymp-
totically de Sitter (l2 < 0) or anti-de Sitter (l2 > 0), according to the definition of mass in asymptotically dS and AdS
spaces due to Abbott and Deser [2], we have the mass of the solution
M = (n − 1)ωn−1 m. (2.14)
Note that throughout this paper, the convention 16πG = 1 has been used. When k = 1 and hij dxi dxj denotes the
line element of an (n − 1)-dimensional unit round sphere, the solution (2.13) reduces to the one found in [19]. Note
that for a positive constant curvature space, the metric is not necessary to be a round sphere. Therefore the metric
(2.6) is more general than a spherically symmetric metric.

III. CARDY-VERLINDE FORMULA FOR THE BORN-INFELD BLACK HOLES IN ADS SPACES

In this section we first consider the case of l2 > 0, namely for a negative cosmological constant. In this case, the
spacetime asymptotically approaches to an AdS space. The solution (2.13) describes a Born-Infeld black hole in AdS
space. The black hole horizon is determined by V (r)|r=r+ = 0. The behavior of metric function V in the small r
region and large r region and thermodynamics of the black hole for the case of k = 1 have been analyzed and discussed
in [19]. We will not therefore repeat them here, instead we will mainly focus on the cases of k = 0 and k = −1 and
will show that the black hole entropy can be expressed in terms of the Cardy-Verlinde formula.
The temperature of the black hole can be obtained by continuing the metric (2.6) to its Euclidean sector via t = −iτ
and requiring the absence of conical singularity at the horizon. This results in a periodic Euclidean time τ with period
1/T , which is just the inverse Hawking temperature of the black hole. Calculation gives
√ !
4β 2
 
1 n 2 2β
q
2 2n−2
T = (n − 2)k + + r+ − n−3 2β 2 r+ + (n − 1)(n − 2)q 2 . (3.1)
4πr+ n − 1 l2 (n − 1)r+
 4

When T = 0, the black hole is an extremal black hole, there the black hole horizon and the internal horizon (Cauchy
horizon) coincide with each other. For the extremal black hole, the charge has a relation to the horizon radius
2n−6  2
(n − 1)2 r+ 4β 2
 
2n−2 n
(n − 1)(n − 2)q 2 = −2β 2 r+ + (n − 2)k + 2
+ 2 r+ . (3.2)
8β 2 n−1 l
The entropy of the black hole still obeys the so-called horizon area formula
n−1
S = 4πωn−1 r+ . (3.3)

It is easy to show that these thermodynamic quantities such as charge (2.11), mass (2.14), temperature (3.1) and
entropy (3.3) satisfy the first law of black hole thermodynamics

dM = T dS + ΦdQ, (3.4)

where Φ is the electrostatic potential at the black hole horizon, which is conjugate to the electric charge Q,
s
n − 1 16πq n − 2 1 3n − 4 (n − 1)(n − 2)q 2
Φ= n−2 2 F1 [ , , ,− 2n−2 ]. (3.5)
2(n − 2) r+ 2(n − 1) 2 2(n − 1) 2β 2 r+

Note that the mass of black hole can be expressed in terms of the horizon radius r+

4β 2
  
n−2 1 2
M = (n − 1)ωn−1 r+ k+ + r+
n(n − 1) l2
√
2 2β
q
2n−2
− n−3 2β 2 r+ + (n − 1)(n − 2)q 2
n(n − 1)r+
!
2(n − 1)q 2 n − 2 1 3n − 4 (n − 1)(n − 2)q 2
+ 2n−4 2 F1 [ , , ,− 2n−2 ] . (3.6)
nr+ 2(n − 1) 2 2(n − 1) 2β 2 r+

To see the thermodynamical stability of the black hole, let us calculate the heat capacity with a fixed charge, which
implies that we are discussing the stability in a canonical ensemble.
 
∂M
Cq =
∂T q
16π 2 r+
n
T
=   √ 2n−2 . (3.7)
4β 2 n 2 + 2 2β −2β 2 r+ +(n−1)(n−2)2 q2
−(n − 2)k + n−1 + l2 r+ n−3
(n−1)r+
q
2n−2
2β 2 r+ +(n−1)(n−2)q2

Note that from the metric (2.13), Hawking temperature (3.1) or the heat capacity (3.7), one can see that the case
of β = 0 or q = 0 reduces to the case of Schwarzschild black hole in AdS space. In other words, when β = 0, the
charge parameter q will disappear automatically, and vice versa. p When q = 0, we can see p that the heat capacity is
always positive
p if k = 0 or −1, while it is negative for r+ < l (n − 2)/n, positive for r+ > l (n − 2)/n, and diverges
at r+ = l (n − 2)/n, when k = 1. When q 6= 0, since the numerator in (3.7) is always positive, the sign of Cq is
therefore completely determined by that of the denominator, from which we are certainly able to get a relation of
the horizon radius to other parameters k, β, spacetime dimension n and charge q. Numerical check indicates that
when k = 0 or −1, the heat capacity is always positive. This can be seen from the behavior of Hawking temperature
(3.1). In Fig. 1 and Fig. 2, we respectively plot the inverse Hawking temperature for a given charge when k = 0 and
k = −1. We see that the inverse Hawking temperature always monotonically decreases from infinity (corresponding
to extremal black holes) to zero as the black hole horizon radius increases. Note that the region with a negative
Hawking temperature should be excluded from the physical phase space.
When k = 1, the heat capacity can be positive or negative. We plot the inverse Hawking temperature of the
Born-Infeld black hole when k = 1 in Fig. 3. In the case of k = 1, we notice that between two stable black hole phases
(small black holes near the extremal limit and large black holes where the effect of charge is negligible) there is a
unstable black hole phase. However, this unstable black hole phase will disappear when the Born-Infeld parameter
β increases. To clearly see this point, in Fig. 4 we plot the inverse temperatures of black holes with a fixed charge,
but with different β, from which we can see clearly how the unstable phase disappears. This implies that for the
Born-Infeld black holes with a given charge, there is a critical value of the Born-Infeld parameter β, above which the
 5

8
6
Inverse T
4
2
1 0.2
0.4
1.5 0.6
0.8
radius 1
2 beta

FIG. 1: The inverse Hawking temperature of a five-dimensional Born-Infeld black hole in AdS space for a fixed charge q/l2 = 0.5
for the case of k = 0.

10
Inverse T
5
0
11.25
1.5 0.40.2
0.6
1.75 0.8
radius 1
2 beta

FIG. 2: The inverse Hawking temperature of a five-dimensional Born-Infeld black hole in AdS space for a fixed charge q/l2 = 0.5
for the case of k = −1.

black holes are always stable thermodynamically. This critical point can be calculated from the Hawking temperature
(3.1). The expression, however, is a little bit complicated, we do not therefore present it here.
According to the AdS/CFT correspondence, the thermodynamics of Born-Infeld black holes in AdS spaces should
also be identified with that of a some dual CFT. Since the entropy of CFTs with AdS duals can be described by the
Cardy-Verlinde formula [9], so it is of interest to see whether or not the black hole entropy (3.3) can be recast to a
form of the Cardy-Verlinde formula.
Suppose there is a CFT residing on an (n − 1)-dimensional sphere with radius R,
ds2 = −dt2 + R2 dΩ2n−1 . (3.8)
The Cardy-Verlinde formula can be written down as [9]
2πR p
S= Ec (2E − Ec ), (3.9)
n−1
where E and Ec respectively represent the total energy and the Casimir energy, non-extensive part of energy, of the
CFT.
For the Born-Infeld black holes, the total energy E is the AD mass M given in (2.14). The corresponding Casimir
energy Ec is [9, 11]
Ec = nE − (n − 1)T S − (n − 1)ΦQ. (3.10)
Substituting those thermodynamic quantities into (3.10), we get
n−2
Ec = 2k(n − 1)r+ ωn−1 . (3.11)
We see that for Born-Infeld black holes with Ricci flat horizon (k = 0), the Casimir energy vanishes, while it is
negative for the hyperbolic horizon.
With the energy E and the Casimir energy Ec , we find that the black hole entropy (3.3) can indeed be cast to a
form of the Cardy-Verlinde formula as follows,
2πl
q
S= (2(E − Eq ) − Ec )Ec /k. (3.12)
n−1
 6

4
2
Inverse T
2 10
1.5 0.8
0.6
1
radius 0.4 beta
0.5 0.2

FIG. 3: The inverse Hawking temperature of a five-dimensional Born-Infeld black hole in AdS space for a fixed charge q/l2 = 0.5
for the case of k = 1.

Inverse T

radius
0.5 1 1.5 2

FIG. 4: The inverse Hawking temperature of a five-dimensional Born-Infeld black hole in AdS space for a fixed charge q/l2 = 0.5
for the case of k = 1. Three curves from right to left correspond to the case βl = 0.40, 0.22 and 0.15.

The dual CFT to the Born-Infeld black holes resides on the boundary of the bulk metric (2.6). Up to a conformal
factor, the boundary metric is

ds2 = −dt2 + l2 hij dxi dxj , (3.13)

it is just the boundary spacetime, in which the dual CFT resides and its entropy is described by the formula (3.12).
Eq in (3.12) is nothing but the energy of electromagnetic field outside the black hole horizon, which can be calculated
as
Z ∞
Eq ≡ ωn−1 T̄00 rn−1 dr
r+
√
4β 2 n 2 2βr+
q
2n−2
= ωn−1 r+ − 2β 2 r+ + (n − 1)(n − 2)q 2
n n
!
2(n − 1)2 q 2 n − 2 1 3n − 4 (n − 1)(n − 2)q 2
+ n−2 2 F1 [ , , ,− 2n−2 ] , (3.14)
nr+ 2(n − 1) 2 2(n − 1) 2β 2 r+

where T̄00 is the 0 − 0 component of the energy-momentum tensor of non-linear electromagnetic field,

1 2Fµα Fν α
T̄µν = gµν L(F ) + q . (3.15)
2 F µν F
1 + 2β 2µν

Thus we show that although the electrodynamics in the Einstein-Born-Infeld theory is non-linear, the entropy of the
Born-Infeld black holes in AdS space can still be expressed in terms of the Cardy-Verlinde formula. Of course, the
energy of electromagnetic field outside the black hole horizon should be subtracted from the total energy.
 7

IV. CARDY-VERLINDE FORMULA FOR THE BORN-INFELD BLACK HOLES IN DS SPACES

In this section we consider the Born-Infeld black holes in de Sitter space (l2 < 0 in (2.12)). In this case, a
cosmological horizon appears, in addition to black hole horizons. Both cosmological horizon and black hole horizon
have thermodynamic properties such as Hawking temperature, entropy and etc. As we mentioned in Sec. I, it is not
an easy matter to compute conserved charges associated with an asymptotically dS space because of the absence of
spatial infinity and a globally timelike Killing vector in such a spacetime. It is found that if one uses the definition of
mass due to Abbott and Deser in dS space [2], the entropy of Schwarzschild black holes in dS space can be expressed
by the Cardy-Verlinde formula, but not for the entropy of cosmological horizon [15]. On the other hand, if one
uses the definition of mass coming from the surface counterterm prescription in asymptotically dS spaces [3, 4], the
entropy of cosmological horizon can be described by the Cardy-Verlinde formula, but not for the entropy of black hole
horizon [16].
For convenience, we change l2 in (2.13) to −l2 so that in what follows one has l2 > 0. In this case, the metric
function V becomes
4β 2
 
m 1
V (r) = k − n−2 + − 2 r2
r n(n − 1) l
√
2 2β p
− 2β 2 r2n−2 + (n − 1)(n − 2)q 2
n(n − 1)rn−3
2(n − 1)q 2 n − 2 1 3n − 4 (n − 1)(n − 2)q 2
+ 2 F1 [ , , , − ]. (4.1)
nr2n−4 2(n − 1) 2 2(n − 1) 2β 2 r2n−2

The solution (4.1) is asymptotically de Sitter. When k 6= 1, the solution (4.1) generalizes the so-called topological
de Sitter solutions proposed in [4] to the Einstein-Born-Infeld theory with a positive cosmological constant. When
k = 1 and m > 0, in an appropriate parameter space, both cosmological horizon rc and black hole horizon r+ occur;
they satisfy V (r)|r=r+ ,rc = 0. When k = 0 or k = −1, black hole horizon disappears, but the cosmological horizon is
still there if m < 0. As the case of asymptotic AdS, the black hole horizon or cosmological horizon can be a positive,
zero or negative constant curvature hypersurface. Therefore these solutions can be dubbed topologically asymptotic
de Sitter solutions.
According to the definition of mass due to Abbott and Deser [2], the mass of the solution (4.1) is

M = (n − 1)ωn−1 m, (4.2)

where m can be expressed in terms of black hole horizon radius r+ via V (r+ ) = 0,

4β 2
  
n−2 1 2
m = r+ k+ − r+
n(n − 1) l2
√
2 2β
q
2n−2
− n−3 2β 2 r+ + (n − 1)(n − 2)q 2
n(n − 1)r+
!
2(n − 1)q 2 n − 2 1 3n − 4 (n − 1)(n − 2)q 2
+ 2(n−2) 2 F1 [ , , ,− 2n−2 ] . (4.3)
nr+ 2(n − 1) 2 2(n − 1) 2β 2 r+

The Hawking temperature of black hole horizon is found to be

√ !
4β 2
 
1 n 2 2β
q
2 2 2n−2 2
T = (n − 2)k + − 2 r+ − n−3 2β r+ + (n − 1)(n − 2)q . (4.4)
4πr+ n−1 l (n − 1)r+

The entropy of black hole horizon and the chemical potential conjugate to the charge Q are still given by (3.3) and
(3.5), respectively. Namely they keep the same forms as in the case of asymptotically AdS space. In this way it is
easy to check that the first law of thermodynamics holds for the black hole horizon, dM = T dS + ΦdQ.
For the black hole horizon, we find that the Casimir energy Ec is given b[15]
n−2
Ec = nE − (n − 1)T S − (n − 1)ΦQ = 2k(n − 1)r+ ωn−1 , (4.5)

and that the energy Eq of the electromagnetic field outside the black hole horizon remains the same form as the one
given in (3.14). Here the total energy E = M .
 8

Then the entropy of black hole horizon given by (3.3) can be recast to
2πl
q
S= Ec /k|(2(E − Eq ) − Ec )|. (4.6)
n−1
It is interesting to note that for the black hole in dS space, the extensive part of total energy
n 2
2(E − Eq ) − Ec = −2(n − 1)ωn−1 r+ /l , (4.7)
is negative. Its implication is not clear to us at the moment.
Next we discuss the thermodynamics of cosmological horizon rc . We relate the gravitational mass M̃ to the
cosmological horizon rc by using the surface counterterm method [3, 4]. In this method, the gravitational mass is just
the AD mass, but with an opposite sign,
M̃ = −(n − 1)ωn−1 m, (4.8)
where m can be expressed in terms of the cosmological horizon radius rc
4β 2
  
1
m = rc n−2
k+ − rc2
n(n − 1) l2
√
2 2β
q
− 2β 2 rc2n−2 + (n − 1)(n − 2)q 2
n(n − 1)rcn−3
2(n − 1)q 2 (n − 1)(n − 2)q 2

n − 2 1 3n − 4
+ F [ , ,
2(n−2) 2 1 2(n − 1) 2 2(n − 1)
, − ] . (4.9)
nrc 2β 2 rc2n−2
Note that the gravitational mass M̃ (4.8) is measured on the boundary of the asymptotically dS spacetime, namely at
the future infinity (I + ), which is outside the cosmological horizon. Similar to the black hole horizon, we can obtain
the Hawking temperature of the cosmological horizon
√ !
4β 2
 
1 n 2 2β
q
T̃ = −(n − 2)k − − rc2 + 2β 2 rc2n−2 + (n − 1)(n − 2)q 2 . (4.10)
4πrc n − 1 l2 (n − 1)rcn−3
In general, the Hawking temperature of cosmological horizon is not equal to that of black hole horizon. Therefore
the spacetime describes a black hole in de Sitter space is unstable quantum mechanically. The chemical potential
conjugate to the charge Q is defined as
s
n − 1 16πq n − 2 1 3n − 4 (n − 1)(n − 2)q 2
Φ̃ = − 2 F1 [ , , ,− ]. (4.11)
2(n − 2) rcn−2 2n − 2 2 2n − 2 2β 2 rc2n−2
And the entropy of the cosmological horizon obeys the area formula
S̃ = 4πωn−1 rcn−1 . (4.12)
For such defined thermodynamic quantities associated with the cosmological horizon, we can see that they indeed
satisfy the first law of thermodynamics
dM̃ = T̃ dS̃ + Φ̃dQ. (4.13)
Therefore it is seemingly reasonable to consider black hole horizon and cosmological horizon as two separate thermo-
dynamic systems [20].
For the cosmological horizon, the Casimir energy turns out to be [15, 16]
Ẽc = nẼ − (n − 1)T̃ S̃ − (n − 1)Φ̃Q = −2k(n − 1)rcn−2 ωn−1 , (4.14)
where Ẽ = M̃ . Similarly we can calculate the energy of Born-Infeld electromagnetic field outside the cosmological
horizon
Z ∞
Ẽq ≡ −ωn−1 T̄00 rn−1 dr
rc
√
4β 2 n 2 2βrc
q
= −ωn−1 rc − 2β 2 rc2n−2 + (n − 1)(n − 2)q 2
n n
2(n − 1)2 q 2 (n − 1)(n − 2)q 2

n − 2 1 3n − 4
+ F
2 1 [ , , , − ] . (4.15)
nrcn−2 2(n − 1) 2 2(n − 1) 2β 2 rc2n−2
 9

Then we find that as the case of black hole horizon, the entropy of cosmological horizon can be reexpressed in terms
of the Cardy-Verlinde formula
2πl
q
S̃ = |Ẽc /k|(2(Ẽ − Ẽq ) − Ẽc ). (4.16)
n−1
Here the extensive part of the energy

2(Ẽ − Ẽq ) − Ẽc = 2(n − 1)ωn−1 rc2 /l2 . (4.17)

Thus we show that both the entropies associated with black hole horizon and cosmological horizon of the Born-Infeld
black holes in dS space can be formally recast to a form of the Cardy-Verlinde formula, although the Casimir energy is
negative in some cases, or the extensive part of energy is negative in others. These results imply that thermodynamics
associated with black hole (cosmological) horizon is indeed related to that of a some CFT.

V. FACTORIZED SOLUTION

In Sec. II we mentioned that for the static electromagnetic field solution (2.10), we have R′′ (r) = 0 from the Einstein
equations of motion, and only the solution of R = r was considered there, which gives the Born-Infeld black hole
solutions in (A)dS spaces (2.13). In this section we discuss the case R = a with a being a positive constant. In this
case, the metric (2.6) is a direct product of two subspaces, in which ds22 = a2 hij dxi dxj is an (n − 1)-dimensional
constant curvature space.
From the equation (2.4) of the Born-Infeld electromagnetic field, we have
p
rt (n − 1)(n − 2)βq
F = p ≡ f, (5.1)
2
2β a 2n−2 + (n − 1)(n − 2)q 2

where q is still an integration constant, which can be related to the electric charge of the solution via (2.11). Note
that here the electric field (5.1) is a constant, independent of the radial coordinate r.
Solving (2.7) or (2.8) gives

V = −Λ0 r2 + c1 r + c0 , (5.2)

where c1 and c0 are two integration constants. Without loss of generality, we can take c1 = 0 through redefining the
radial coordinate r. And c0 can be normalized to be one by rescaling the coordinates t and r. Then we have

V = 1 − Λ0 r 2 . (5.3)

The constant Λ0 is

8 16β 2 p 8(n − 3)f 2

Λ0 = Λ− (1 − 1 − f 2 /β 2 ) − p . (5.4)
n−1 n−1 (n − 1) 1 − f 2 /β 2

From (2.9) we obtain a constraint

n−2 2 4β 2 4β 2
2
k− Λ+ − p = 0. (5.5)
a n−1 n − 1 (n − 1) 1 − f 2 /β 2

Therefore we see that the coordinates t and r in (2.6) now span a two-dimensional constant curvature spacetime.
The curvature is determined by Λ0 (dS if Λ0 > 0, AdS if Λ0 < 0). In addition, once the parameters, the curvature
parameter k, charge q, the cosmological constant Λ, and the Born-Infeld parameter β are given, then the curvature
radius a of the constant curvature space ds22 = a2 hij dxi dxj can be determined via the relation (5.5). Finally we
obtain an exact solution (2.6), which is a direct product of a two-dimensional dS (if Λ0 > 0) or AdS (if Λ0 < 0) space
and an (n − 1)-dimensional constant curvature space with a curvature radius a.
 10

VI. CONCLUSION

We have found black holes solutions in the Einstein-Born-Infeld theory with a cosmological constant in higher
dimensions. These black hole solutions are asymptotically de Sitter or anti-de Sitter depending on the cosmological
constant. Black hole horizon or cosmological horizon in these solutions can be a positive, zero or negative constant
curvature hypersurface. Therefore these solutions generalize some exact solutions of the Einstein-Born-Infeld theory
existing in the literature. We have studied thermodynamics of black hole horizon and of cosmological horizon. For the
Born-Infeld black holes in AdS spaces, when the horizon is a zero or negative constant curvature hypersurface, they
are always thermodynamical stable with positive heat capacity; when the horizon is a positive constant curvature
hypersurface, there is a critical value on the Born-Infeld parameter for a fixed charge, above which the black hole is
also thermodynamical stable. However, the Born-Infeld parameter is less than the critical value, a unstable black hole
phase will appear. In addition, it has been found that although the electromagnetics of the Born-Infeld theory is non-
linear, the entropies of black hole horizon and cosmological horizon can all be recast to a form of the Cardy-Verlinde
formula, which indicates that the Cardy-Verlinde formula is of universality in some sense. Note that if there are some
higher derivative terms of curvature, black hole entropy in these gravitational theories can not be expressed in terms
of the Cardy-Verlinde formula [11]. Therefore our results implies that higher order derivative terms of electromagnetic
field do not destroy the application of the Cardy-Verlinde formula. In other words, the spacetime of the Born-Infeld
black hole at least in AdS space is still dual to a some CFT residing on the boundary of the AdS space. The conformal
invariance of the dual CFT is not broken.
In the Einstein-Born-Infeld theory with a cosmological constant in D(≥ 4) dimensions, we have also found a
factorized solution, whose spacetime is a direct product of two constant curvature spaces: one is a two-dimensional
dS or AdS space and the other is a (D − 2)-dimensional positive, zero or negative constant curvature space with a
constant curvature radius.

Acknowledgements

DWP thanks Hong-Sheng Zhang, Zong-Kuan Guo, Wei He, Jian-Huang She, Qi Guo and Hao Wei for useful
discussions and kind help. RGC would like to express his gratitude to the Physics Department, Baylor University
for its hospitality. This work was supported by Baylor University, a grant from Chinese Academy of Sciences, a
grant from NSFC, China (No. 13325525), and a grant from the Ministry of Science and Technology of China (No.
TG1999075401).

[1] J. P. S. Lemos, Phys. Lett. B 353, 46 (1995) [arXiv:gr-qc/9404041]; J. P. Lemos, Class. Quant. Grav. 12, 1081 (1995)
[gr-qc/9407024]; J. P. S. Lemos and V. T. Zanchin, Phys. Rev. D 54, 3840 (1996) [arXiv:hep-th/9511188]; C. G. Huang and
C. B. Liang, Phys. Lett. A 201 (1995) 27; R. G. Cai and Y. Z. Zhang, Phys. Rev. D 54, 4891 (1996) [arXiv:gr-qc/9609065];
R. G. Cai, J. Y. Ji and K. S. Soh, Phys. Rev. D 57, 6547 (1998) [arXiv:gr-qc/9708063]; R. G. Cai, Nucl. Phys. B 524,
639 (1998) [arXiv:gr-qc/9801098]; S. Aminneborg, I. Bengtsson, S. Holst and P. Peldan, Class. Quant. Grav. 13, 2707
(1996) [gr-qc/9604005]; R. B. Mann, Class. Quant. Grav. 14, L109 (1997) [gr-qc/9607071]; R. B. Mann, Nucl. Phys.
B 516, 357 (1998) [hep-th/9705223]; L. Vanzo, Phys. Rev. D 56, 6475 (1997) [gr-qc/9705004]; D. R. Brill, J. Louko
and P. Peldan, Phys. Rev. D 56, 3600 (1997) [arXiv:gr-qc/9705012]; D. Klemm, Class. Quant. Grav. 15, 3195 (1998)
[gr-qc/9808051]; D. Klemm, V. Moretti and L. Vanzo, Phys. Rev. D 57, 6127 (1998) [Erratum-ibid. D 60, 109902 (1998)]
[gr-qc/9710123]; M. Banados, A. Gomberoff and C. Martinez, Class. Quant. Grav. 15, 3575 (1998) [hep-th/9805087];
D. Birmingham, Class. Quant. Grav. 16, 1197 (1999) [hep-th/9808032]; R. Cai and K. Soh, Phys. Rev. D 59, 044013
(1999) [gr-qc/9808067]; K. Behrndt, M. Cvetic and W. A. Sabra, Nucl. Phys. B 553, 317 (1999) [arXiv:hep-th/9810227];
M. Cvetic et al., Nucl. Phys. B 558, 96 (1999) [arXiv:hep-th/9903214]; M. J. Duff and J. T. Liu, Nucl. Phys. B 554, 237
(1999) [arXiv:hep-th/9901149]; M. F. A. da Silva, A. Wang, F. M. Paiva and N. O. Santos, Phys. Rev. D 61, 044003 (2000)
[arXiv:gr-qc/9911013]; R. G. Cai, Phys. Lett. B 572, 75 (2003) [arXiv:hep-th/0306140]; R. Aros, R. Troncoso and J. Zanelli,
Phys. Rev. D 63, 084015 (2001) [hep-th/0011097]; R. G. Cai, Phys. Rev. D 65, 084014 (2002) [arXiv:hep-th/0109133];
R. G. Cai, Phys. Lett. B 582, 237 (2004) [arXiv:hep-th/0311240]; W. L. Smith and R. B. Mann, Phys. Rev. D 56,
4942 (1997) [gr-qc/9703007]; Y. Wu, M. F. A. da Silva, N. O. Santos and A. Wang, Phys. Rev. D 68, 084012 (2003)
[arXiv:gr-qc/0309002]; S. Nojiri and S. D. Odintsov, Phys. Lett. B 521, 87 (2001) [Erratum-ibid. B 542, 301 (2002)]
[arXiv:hep-th/0109122]; M. Cvetic, S. Nojiri and S. D. Odintsov, Nucl. Phys. B 628, 295 (2002) [arXiv:hep-th/0112045];
S. Nojiri and S. D. Odintsov, Phys. Rev. D 66, 044012 (2002) [arXiv:hep-th/0204112]; Y. M. Cho and I. P. Neupane, Phys.
Rev. D 66, 024044 (2002) [arXiv:hep-th/0202140]; I. P. Neupane, Phys. Rev. D 67, 061501 (2003) [arXiv:hep-th/0212092];
I. P. Neupane, arXiv:hep-th/0302132; R. G. Cai and A. Wang, arXiv:hep-th/0406040; R. G. Cai and A. z. Wang, Phys.
Rev. D 70, 064013 (2004) [arXiv:hep-th/0406057]; and references therein.
 11

[2] L. F. Abbott and S. Deser, Nucl. Phys. B 195, 76 (1982).

[3] V. Balasubramanian, J. de Boer and D. Minic, Phys. Rev. D 65, 123508 (2002) [arXiv:hep-th/0110108]; A. M. Ghezel-
bash and R. B. Mann, JHEP 0201, 005 (2002) [arXiv:hep-th/0111217]; D. Klemm, Nucl. Phys. B 625, 295 (2002)
[arXiv:hep-th/0106247].
[4] R. G. Cai, Y. S. Myung and Y. Z. Zhang, Phys. Rev. D 65, 084019 (2002) [arXiv:hep-th/0110234].
[5] J. Maldacena, Adv. Theor. Math. Phys. 2, 231 (1998) [Int. J. Theor. Phys. 38, 1113 (1998)] [hep-th/9711200]; S. S. Gubser,
I. R. Klebanov and A. M. Polyakov, Phys. Lett. B 428, 105 (1998) [hep-th/9802109]; E. Witten, Adv. Theor. Math. Phys.
2, 253 (1998) [hep-th/9802150].
[6] E. Witten, Adv. Theor. Math. Phys. 2, 505 (1998) [arXiv:hep-th/9803131].
[7] A. Strominger, JHEP 0110, 034 (2001) [arXiv:hep-th/0106113].
[8] J. L. Cardy, Nucl. Phys. B 270, 186 (1986).
[9] E. Verlinde, arXiv:hep-th/0008140.
[10] D. Klemm, A. C. Petkou and G. Siopsis, Nucl. Phys. B 601, 380 (2001) [arXiv:hep-th/0101076].
[11] R. G. Cai, Phys. Rev. D 63, 124018 (2001) [arXiv:hep-th/0102113]; R. G. Cai, Y. S. Myung and N. Ohta, Class.
Quant. Grav. 18, 5429 (2001) [arXiv:hep-th/0105070]; R. G. Cai and A. z. Wang, Phys. Rev. D 70, 064013 (2004)
[arXiv:hep-th/0406057].
[12] D. Birmingham and S. Mokhtari, Phys. Lett. B 508, 365 (2001) [arXiv:hep-th/0103108].
[13] J. l. Jing, Phys. Rev. D 66, 024002 (2002) [arXiv:hep-th/0201247].
[14] L. Cappiello and W. Muck, Phys. Lett. B 522, 139 (2001) [arXiv:hep-th/0107238]; S. Nojiri, S. D. Odintsov and S. Ogushi,
Int. J. Mod. Phys. A 16, 5085 (2001) [arXiv:hep-th/0105117]; S. Nojiri, S. D. Odintsov and S. Ogushi, Phys. Rev. D 65,
023521 (2002) [arXiv:hep-th/0108172]; S. Nojiri, S. D. Odintsov and S. Ogushi, Int. J. Mod. Phys. A 17, 4809 (2002)
[arXiv:hep-th/0205187]; D. Astefanesei, R. Mann and E. Radu, JHEP 0401, 029 (2004) [arXiv:hep-th/0310273].
[15] R. G. Cai, Nucl. Phys. B 628, 375 (2002) [arXiv:hep-th/0112253].
[16] R. G. Cai, Phys. Lett. B 525, 331 (2002) [arXiv:hep-th/0111093].
[17] M. Born and L. Infeld, Proc. Roy. Soc. Lond. A 144, 425 (1934).
[18] D. L. Wiltshire, Phys. Rev. D 38, 2445 (1988); D. A. Rasheed, arXiv:hep-th/9702087. T. Tamaki and T. Torii, Phys.
Rev. D 62, 061501 (2000) [arXiv:gr-qc/0004071]; N. Breton, Phys. Rev. D 67, 124004 (2003) [arXiv:hep-th/0301254];
M. Aiello, R. Ferraro and G. Giribet, arXiv:gr-qc/0408078. M. Cataldo and A. Garcia, Phys. Lett. B 456, 28 (1999)
[arXiv:hep-th/9903257]; S. Fernando and D. Krug, Gen. Rel. Grav. 35, 129 (2003) [arXiv:hep-th/0306120].
[19] T. K. Dey, Phys. Lett. B 595, 484 (2004) [arXiv:hep-th/0406169].
[20] C. Teitelboim, arXiv:hep-th/0203258; A. Gomberoff and C. Teitelboim, Phys. Rev. D 67, 104024 (2003)
[arXiv:hep-th/0302204]; R. G. Cai and Q. Guo, Phys. Rev. D 69, 104025 (2004) [arXiv:hep-th/0311020].