2 + 1−dimensional electrically charged black holes in Einstein - Power - Maxwell

Theory
O. Gurtug,∗ S. Habib Mazharimousavi,† and M. Halilsoy‡
Department of Physics, Eastern Mediterranean University, G. Magusa, north Cyprus, Mersin 10 - Turkey.

A large family of new black hole solutions in 2 + 1−dimensional Einstein-Power-Maxwell (EPM)

gravity with prescribed physical properties is derived. We show with particular examples that ac-
cording to the power parameter k of the Maxwell field, the obtained solutions may be asymptotically
flat for 1/2 < k < 1 or non-flat for k > 1 in the vanishing cosmological constant limit. We study
the thermodynamic properties of the solution with two different models and it is shown that ther-
modynamic quantities satisfy the first law. The behaviour of the heat capacity indicates that by
employing the 1 + 1−dimensional dilaton analogy the local thermodynamic stability is satisfied.
arXiv:1010.2340v3 [gr-qc] 11 Apr 2012

I. INTRODUCTION

It is well known that the main motivation to study nonlinear electrodynamics (NED) was to overcome some of the
difficulties that occur in the standard linear Maxwell theory. Divergence in self-energy due to point charges was one
such difficulty that kept physics community busy for decades. The model of Born - Infeld NED was developed with
the hope to remove these divergences; see e.g.[1–6], and by a similar trend NED was employed to eliminate black hole
singularities in general theory of relativity. A striking example of regular black hole solutions in (3 + 1) −dimensions
was given in [7] that considers Einstein field equations coupled with NED which satisfies the weak energy condition
and recovers the Maxwell theory in the weak field limit. The use of NED for eliminating singularities was also proved
successful in (2 + 1) − dimensions [8].
During the last decade, (2 + 1) - dimensional spacetimes admitting black hole solutions have attracted much
attention. As a matter of fact the d = 3 case singles out among other (d ≥ 4) spacetimes with special mass and
charge dependence. The first example in this regard is the BTZ black hole [9]. Later on, Einstein-Maxwell [10]
and Einstein-Maxwell-dilaton [11] extensions have also been found. The black hole solutions found in this context
include all typical characteristics that can be found in (3 + 1) or higher dimensional black holes such as; horizon(s),
black hole thermodynamics and Hawking radiation. The black hole solution derived in [12] is another example for
(2+1)−dimensions within the context of a restricted class of NED in which the Maxwell scalar has a power in the form
3/4
of (Fµν F µν ) . This particular power results from imposing traceless condition on the energy-momentum tensor.
The main objective of the present study is two-fold. First, we construct a large class of black hole solutions sourced
k
by the power Maxwell field in which the Maxwell scalar has the form (Fµν F µν ) . Here, the power k is a real rational
number which will be restricted to some intervals as a requirement of the energy conditions. In general, for d -
dimensional spacetimes, the specific choice of k = d4 yields a traceless Maxwell’s energy momentum tensor [13] which
is known to satisfy the conformal invariance condition. In recent years, the use of power Maxwell fields has attracted
considerable interest. It has been used for obtaining solutions in d− spacetime dimensions [14], Ricci flat rotating
black branes with a conformally Maxwell source [15], Lovelock black holes [16], Gauss-Bonnet gravity [17] and the
effect of power Maxwell field on the magnetic solutions in Gausss-Bonnet gravity [18]. Therefore in [12], the power 3/4
of the Maxwell scalar in (2 + 1) −dimensions is the unique case which results from this traceless condition. Our first
motivation in this study is to find the most general solution in (2 + 1)−dimensional Einstein-Power-Maxwell (EPM)
spacetime without imposing the traceless condition. Stated otherwise, choosing a traceful energy momentum tensor
amounts to treating k as a new parameter and we wish to investigate this freedom as much as we can. However, our
analysis on the obtained solutions has revealed that the power parameter k can not be arbitrary. For a physically
acceptable solution it must be a rational number. Hence, our general solution overlaps with the solution presented
in [12], if one takes the power parameter k = 34 . Depending on the value of k, however, the resulting metric displays
different characteristics near r = 0 which makes the present study more interesting. With the freedom of k we explore
a rich possibility in the structure of singularities. For values 1/2 < k < 1, for instance, the resulting spacetime
becomes asymptotically flat in the vanishing cosmological constant (Λ = 0), and for Λ > 0, it is the asymptotically

∗ Electronic address: ozay.gurtug@emu.edu.tr

† Electronic address: habib.mazhari@emu.edu.tr
‡ Electronic address: mustafa.halilsoy@emu.edu.tr
 2

de-Sitter spacetime. For k > 1 the resulting spacetime is non-asymptotically flat. Furthermore, the resulting metric
depends not only on the parameter k but also on the mass M , the charge Q and the cosmological constant Λ. When
Λ > 0, the solution describes a charged de-Sitter black hole spacetime with inner and outer horizons for 1/2 < k < 1,
and a cosmological horizon for k > 1. For specific values of these parameters the resulting spacetime singularity at
r = 0 is naked whose strength becomes k dependent. When the cosmological constant Λ < 0, the resulting spacetime
corresponds to charged anti de-Sitter with a cosmological horizon in the range of the power 1/2 < k < 1. For k > 1,
the resulting charged de-Sitter spacetime becomes naked singular at r = 0.
Another important issue in black hole physics is the concept and analysis of thermodynamic properties. This issue
has gained a significant momentum not just in the linear Maxwell theory but also in the NED. As an example in [19],
higher dimensional gravity coupled to NED sourced by power Maxwell field has been analyzed thermodynamically
for d > 3. The local and global thermodynamic stability is investigated by calculating the Euclidean action with
appropriate boundary term in the grand canonical ensemble. Our second objective in this study is to investigate the
local thermodynamic stability of the resulting black holes. This is achieved by employing the method presented in
[20], in which the local Hawking temperature is found from the Unruh effect. We calculate the heat capacities and
show that our solution conditionally displays local thermodynamic stability. For specific values of the parameters, the
calculated specific heat capacity at constant charge and electric potential both change sign at particular points. This
behaviour indicates that there may be a possible phase change in the black hole state. Alternatively, for a thorough
thermodynamical analysis we appeal to the dilatonic analogy established in 1 + 1−dimensions [25–27].
The organization of the paper is as follows. Section II, introduces the theory of EPM with solution and spacetime
structure. Thermodynamic properties of the solution is considered in Section III. We complete the paper with
Conclusion in Section IV.

II. EINSTEIN- POWER - MAXWELL (EPM) SOLUTIONS AND SPACETIME STRUCTURE

The 3−dimensional action for EPM theory with cosmological constant Λ is given by (c = kB = ℏ = 8G = 1)

√ 1
dx3 −g R − 23 Λ − L (F ) ,
R


I= 2π (1)
k
in which L (F ) = |F | and F is the Maxwell invariant

F = Fµν F µν ,

while the parameter k is arbitrary for the time being. Variation with respect to the gauge potential A yields the
Maxwell equations
 
k−1
d (⋆ FLF ) = 0 → d ⋆ F |F | = 0, (2)

⋆
where denotes duality. Variation of the action with respect to the spacetime metric gµν yields the field equations

1
Gνµ + Λδνµ = πTµν , (3)
3
where
1
T µν = µλ
LF − δ µν L ,



4 Fνλ F (4)
2
is the energy-momentum tensor of the power Maxwell field and explicitly reads
!
k
4k Fνλ F µλ


µ |F | µ
Tν= −δ ν . (5)
2 F

Our metric ansatz for (2 + 1)− dimensions, is chosen as

dr2
ds2 = −f (r) dt2 + + r2 dθ2 . (6)
f (r)
 3

Static, electrically charged potential ansatz is given by

A = A(r)dt,

which leads to

F = dA =E(r)dr ∧ dt, (7)

with its dual

⋆
F =E(r)rdθ (8)

and

F = Fµν F µν = −2E(r)2 . (9)

Accordingly, the Maxwell equation reads now

 k−1 
d E(r)rdθ 2E(r)2

= 0, (10)

which leads to the solution as

rE(r)2k−1 = constant, (11)

or equivalently
constant
E(r) = 1 . (12)
r 2k−1
By using the latter result in (7) and choosing the integration constant proportional to the electric charge Q, one
obtains the potential
(
Q ln r k=1
A(r) = Q(2k−1) 2(k−1) . (13)
2(k−1) r
(2k−1) k 6= 1, 12

The resulting energy - momentum tensor follows from (5) as,

1
T µν = |F |k diag (ξ, ξ, −1) (14)
2
where ξ = (2k − 1) and the explicit form of F is given by

Q2
F =− 2 (15)
r 2k−1
in which we recall that Q is a constant related to the charge of the black hole. One can show that the WEC and SEC
restrict us to the set k ∈ 21 , ∞ (see Appendix). The tt component of Einstein equations (3) reads

1 ′ 1 k
f (r) + Λ = 4πξ |F | . (16)
2r 3
whose integration gives,
2
r2 π (2k − 1) 2k 2(k−1)
f (r) = D + 2
− Q r 2k−1 , (17)
l 2 (k − 1)

in which D is an integration constant, Λ = −1/l2 and k 6= 1. The choice k = 1, which we exclude here gives the
2
known charged BTZ black hole solution in Einstein-Maxwell theory [10]. We note that for k = 1, F = − Q r 2 diverges
at r = 0 which is weaker than the cases for k < 1. This behaviour, however, turns opposite for the choice k > 1. In
order to illustrate this important effect of the power parameter k, we calculate the Kretschmann invariant for k = 1,
 4

k < 1 and k > 1. Since the resulting expressions for any k > 1 or k < 1 is too complicated, we prefer to calculate the
Kretschmann invariant for specific values of k;

12 4Q2 2 Q2
 
K = 4 − 2 − 2 for k = 1, (18)
l r l2 r
12 8Q2
K = 4 − 6 for k = 3/4,
l r
12 2πQ4 10 19πQ4
 
K = 4 − 4/3 − for k = 2.
l r l2 2r4/3

It is clear from these results that the rate of divergence of the Kretschmann invariant for k = 43 is faster than the
cases k ≥ 1.The case when the power of r, 2(k−1) 1
2k−1 < 0, bounds the value of k to 2 < k < 1 which is also consistent
with the energy conditions. Note that this case corresponds to asymptotically flat spacetime if one takes Λ = 0.
The integration constant D can be associated with the mass of the black hole i.e., D can be expressed in terms of
mass at infinity by employing the Brown-York [12, 21, 22] formalism. Following the quasilocal mass formalism it is
known that, a spherically symmetric 3−dimensional metric solution as
1
ds2 = −F (r)2 dt2 + dr2 + r2 dθ 2 (19)
G(r)2

admits a quasilocal mass MQL defined by

MQL = lim 2F (rb ) [Gr (rb ) − G(rb )] . (20)

rb →∞

Here Gr (rb ) is an arbitrary non-negative reference function, which yields the zero of the energy for the background
spacetime, and rb is the radius of the space-like hypersurface. According to our line element we get

r2 π (2k − 1)2 2k 2(k−1)

F (r)2 = G(r)2 = D + − Q r 2k−1 , (21)
l2 2 (k − 1)
2
r2 π (2k − 1) 2k 2(k−1)
Gr (r)2 = 2
− Q r 2k−1 (22)
l 2 (k − 1)

which yield
2(k−1)
   
2 2 2 2k−1 2
r l l ar r 2(k−1)

MQL = lim 2  2b 1 + 2 D − b  − (D + b − ar 2k−1 ) .

b (23)
rb →∞ l 2rb 2rb2 l2

2
Here we expanded the square roots and a = π(2k−1) 2k 2(k−1)
2(k−1) Q . We note that since 2k−1 < 1 for all values of 1/2 < k < ∞,
this limit results in −D, independent of the value of k. We would like to add that the same result may be found by
applying the method introduced in [23, 24] with a proper choice of the background metric.
Therefore the metric function, irrespective of the power of r, for M > 0 is given by

r2 π (2k − 1)2 2k 2(k−1)

f (r) = −M + − Q r 2k−1 . (24)
l2 2 (k − 1)

Finally in this section we give the Ricci scalar

6 πQ2k (4k − 3)
R=− + 2k . (25)
l2 r 2k−1
which indicates the occurrence of true curvature singularity for any k > 21 . Although the particular choice k = 34
shows R to be regular at r = 0 this is not supported by the Kretschmann scalar expression in (18). Nevertheless, the
energy conditions (i.e. at least WEC and SEC) - given in Appendix - always result in negative exponents in radial
coordinate r therefore r = 0 is a true curvature singularity.
 5

III. THERMODYNAMICS

A. Analysis with finite boundary model

In this section, we study the thermodynamical properties of the solution (24). A similar analysis for d−dimensional
charged black holes with a NED sourced by power Maxwell fields was considered in [11], by employing Euclidean
action with a suitable boundary term in the grand canonical ensemble. The analysis was carried out for spacetime
dimensions d > 3.
In this study, we follow an alternative method as demonstrated in [19] to find the local Hawking temperature by
using the Unruh effect in curved spacetime which is equivalent to finding the periodicity in the time coordinate in the
Euclidean version of the metric covering the outer region of the black hole. In the Unruh effect, an observer outside
the black hole experiences a thermal state with local temperature defined by
 
2f ′ (rh ) 32 r
h π (2k − 1) Q2k 
TH (r) = √ = p − 1 , (26)
π −χα χα π f (r)  l2 2r 2k−1

h

where χα is the Killing vector field generating the outer horizon and the location of the horizons are given by the
roots of f (rh ) = 0 which implies

2
rh2 π (2k − 1) 2k 2(k−1)
M= 2
− Q rh2k−1 . (27)
l 2 (k − 1)
It should be noted that the power parameter k in the analysis of thermodynamic properties is assumed to satisfy
1/2 < k < 1. It is remarkable to note that in the limits, TH (r) |r→rh → ∞ and TH (r) |r→∞ → 0. This is expected
because the solution given in (24) is non asymptotically flat, hence, we have a vanishing temperature (i.e. from
(26)) at infinity. Following the same procedure as demonstrated in [19], we define the reenergized temperature as
√ ′
T∞ = TH (r) −χα χα = f 4π (rh )
which gives
 
4  rh π (2k − 1) Q2k 
T∞ = − 1 . (28)
π  l2 2r 2k−1

h

The internal energy of the system on a constant t hypersurface can be defined from the Brown-York [12, 21, 22]
quasilocal energy formalism as

p rb 
E(rb ) = −2 f (rb ) − (29)
l
where r = rb is a finite boundary of the spacetime. One may vary the internal energy E(rb ) with respect to rh and
Q which leads to the first law of thermodynamics in the following form

dE(rb ) = TH (rb )dS + Φ (rb ) de, (30)

in which the entropy S, after our unit convention (c = kB = ℏ = 8G = 1) is given by

S = 4πrh , (31)
and e is the electric charge

√
2 2k−1
e= πk (2k − 1) 2k Q2k−1 . (32)
4
Here,
 
1 rh π − 1
TH (rb ) = p 2
− (2k − 1) Q2k rh 2k−1 (33)
2π f (rb ) l 2
 6

is the Hawking temperature at the boundary of the black hole spacetime and
1 √  2(k−1) 
2 (2k − 1) 2k 2Q 2k−1
2(k−1)
2k−1
Φ (rb ) = p rh − rb (34)
(1 − k) f (rb )

is the electric potential difference between the horizon and boundary rb . On the other hand the electric potential
difference between the boundary and infinity is given by
1 √  2(k−1) 
2 (2k − 1) 2k 2Q
Ψ (rb ) = p rb 2k−1 . (35)
(1 − k) f (rb )

We consider the black hole inside a box bounded by r = rb and calculate the heat capacity of the black hole at
constant Q, Φ and Ψ. The local thermodynamical stability conditions are determined by the sign of the heat capacities
calculated at constant quantities in the limit of large values of rb . which is defined by,
 
∂S
CX ≡ T ≥ 0, (36)
∂T X

in which T = TH (rb ) and X is the quantity to be held constant. Note that, we consider S = S(rh ) and T = T (rh , Q2k ),
therefore Eq. (36) can be written as
−1  −1
∂Q2k
       
∂S ∂T ∂T ∂rh ∂T
CX ≡ T =T =T + , (37)
∂T X ∂S X ∂rh ∂S ∂Q2k ∂S X

∂S
and with ∂rh = 4π, latter equation reduces to,

4πT
CX = n     o. (38)
∂T ∂T ∂Q2k
∂rh + ∂Q2k ∂rh X

The heat capacities for constant Q, Φ and Ψ are calculated by using Eq. (38). Because of the complexity of the
resulting expressions we prefer to give only the expressions for large values of rb , hence, the limiting heat capacities
as rb → ∞ are

 2k

2k−1 πl2 Q2k (2k−1)
 4πrh rh
 − 2
∂S
CQ = T ≃
 2k  , (39)
∂T 2 2k
Q rh2k−1 + πl 2Q
 2k 
2 2k
  4πrh rh2k−1 − πl Q 2(2k−1)
∂S
CΦ = T ≃  2k ,
∂T Φ
rh2k−1 + πl2 Q2k (2k − 1)2
 2k 
2 2k
  4πrh rh2k−1 − πl Q 2(2k−1)
∂S
CΨ = T ≃  2k  .
∂T Ψ 2k−1 πl2 Q2k
rh + 2

One observes that since, from (28), f ′ (rh ) > 0 i.e.

2k
πl2 Q2k (2k − 1)
rh2k−1 − > 0, (40)
2
thermodynamically our solution indicates a locally stable black hole.
 7

B. Analysis with 1+1-dimensional dilaton gravity model

In this section, we employ the method presented in [25–27] to study the thermodynamics of the EPM black hole
found above. In this method the solution given in Eq. (24) will be obtained from the dilaton and its potential of
two-dimensional dilaton gravity through dimensional reduction. Now we consider

ds2 = gab dxa dxb = g̃µν dx̃µ dx̃ν + φ2 (x̃) dθ2 (41)

where φ denotes the radius of the circle S 1 in M3 = M2 × S 1 . The Greek indices represent the two-dimensional
spacetime. After the Kaluza-Klein dimensional reduction, the action (1) reads as,
!
R̃ − 2 Λ̃
Z
k
p
S2D = 2π dx̃2 −g̃φ − L (F ) , L (F ) = |F | , (42)
2π

in which R̃ is the Ricci scalar of M2 and Λ̃ = Λ/3. Varying the above action leads to the following field equations

d (φLF ⋆ F) = 0, (43)
2
∇ φ + 2φΛ̃ = 2πφ (L (F ) − 2F LF ) , (44)
R̃ − 2Λ̃ = −2πL (F ) . (45)

Herein, the electric field 2−form is given by F =E (φ) dt ∧ dφ and its dual becomes 0−form ⋆ F =E (φ) . Note that, φ
is effectively one of our coordinates. The electric field invariant Fab F ab is
1
F = − E (φ)2 (46)
2
which implies from (43)

k−1
Eφ E 2 = constant. (47)

The latter equation yields the following electric field

q
E(φ) = 1 , (48)
φ 2k−1
where q is an integration constant. Then, the Lagrangian L (F ) can be written as,

1 q 2k
L (F ) = 2k , (49)
2k φ 2k−1

and
−k q 2(k−1)
LF = . (50)
2(k−1) φ
2(k−1)
2k−1

The rest of the field equations are given by

!
2 1 q 2k
∇ φ = V (φ) = −2φΛ̃ + 2πφ 2k (1 − 2k) (51)
2k φ 2k−1

and
!
′ 1 q 2k
R̃ = −V (φ) = 2Λ̃ − 2π 2k . (52)
2k φ 2k−1

It is remarkable to observe that, these equations correspond to the 2− dimensional field equations of dilaton gravity
with an action
Z p  
S2D = dx̃dt −g̃ φR̃ + V (φ) (53)
M2
 8

and the line element

dx̃2
ds2 = −f (x̃) dt2 + . (54)
f (x̃)
After manipulating Eq.s (51) and (52) one finds
∇2 φ = f φ′′ + f ′ φ′ = V (φ) (55)
and
R̃ = −f ′′ = −V ′ (φ) , (56)
in which a prime means derivative with respect to the argument. Our dilaton ansatz
φ = x̃ (57)
admits
f ′ = V (φ) , (58)
such that
f (φ) = J (φ) − C (59)
in which
 
φ2 2π (1 − 2k)2 q 2k  1 1 
Z
J (φ) = V (φ) dφ = 2 − − 2(1−k) . (60)
ℓ 2k+1 (k − 1) 2(1−k)
φ 2k−1 φ02k−1
Herein φ0 is a reference potential, Λ̃ = − ℓ12 and C represents the ADM mass of the EPM black hole [25–27]. Also the
line element (54) becomes
dφ2
ds2 = −f (φ) dt2 + . (61)
f (φ)


As it was introduced in [25–27], the extremal value of φ+ is obtained from V φ+ = φe = 0, which yields
2k
2πq 2k ℓ2
φe2k−1 = (2k − 1) . (62)
2k+1
This implies that the extremal mass (from (60)) is given by
Me = J (φe ) , (63)
in which for M ≥ Me the metric function admits at least one horizon φ+ that indicates the outer horizon. The
Hawking temperature at the outer horizon, the heat capacity and free energy are given respectively by

   
V φ+ φ+ 2k
TH = =  2 + 2π  1 q 2k  (1 − 2k) , (64)
4π 4π ℓ2 2k φ 2k−1
+

! !
2 1 q2k

! 4πφ+ ℓ2 + 2π 2k 2k (1 − 2k)

V φ+ 2k−1
φ+
Cq φ+ = 4π
=   (65)
V ′ φ+ 2 1 q2k
ℓ2 + 2π 2k 2k
φ 2k−1

and
     
2 2k
φ2+ 2π (1 − 2k) q 1 1 2k
 − J (φe ) − φ2+  2 + 2π  1 q 2k  (1 − 2k) .


F φ+ = −  − (66)
ℓ2 2k+1 (k − 1) 2(1−k)
2k−1
2(1−k)
2k−1 ℓ2 2k φ 2k−1
φ+ φ0 +

In summary, in this section the thermodynamic analysis of the EPM black hole is investigated by two entirely
different methods. Our analysis reveals that by rescaling the constant q we recover the results obtained in so (40)
that two different approaches for thermodynamic stability are in agreement.
 9

IV. CONCLUSION

In this study, the most general solution in (2 + 1)−dimensional spacetime in EPM theory, without imposing the
traceless condition on the energy - momentum tensor is derived. The obtained solutions describe black holes sourced
by the power Maxwell fields. From physics standpoint and in analogy with the self - interacting scalar fields, k can
be interpreted as the measure of self - interaction that electromagnetic field undergoes. As such, it alters much of
physics and, in particular, the black hole / singularity formations. We have shown with particular examples that the
power parameter k has a significant effect on the physical interpretation of the obtained solutions. For specific values
of parameter k, it is possible to obtain asymptotically flat ( with Λ = 0) or non - asymptotically flat solutions. As it
has been shown in the Appendix, for the choice 23 < k < 1, all energy conditions ( WEC, SEC and DEC) are satisfied,
as well asR the causality condition. The character of the singularity at r = 0 is timelike, since a new coordinate defined
by r∗ = dr f is finite as r → 0. In solutions admitting black holes, this timelike singularity is covered by horizon(s).
But, in some cases it remains naked and violates the cosmic censorship hypothesis. It becomes worthful therefore
to investigate the structure of this singularity in quantum mechanical point of view. For 2(k−1) 2k−1 > 0 the resulting
spacetime geometry is very similar to the BTZ black hole whose quantum singularity structure is investigated in
[28] by quantum test particles obeying the Klein-Gordon and Dirac equations. The results reported in [28] are; for
massive scalar fields the spacetime is quantum singular but for massless scalar bosons and for fermions, the spacetime
is quantum regular. On the other hand, naked singularity that occurs for 2(k−1) 2k−1 < 0 is structurally similar to the
solution given in [12]. The quantum nature of this singularity is recently investigated in [29] with the test particles
obeying the Klein-Gordon and Dirac equations. It was shown that the spacetime is quantum singular for massless
scalar particles obeying Klein-Gordon equation but quantum regular for fermions obeying Dirac equation. Therefore,
these results are also applicable to the solutions presented in this study. Thermodynamic quantities such as Hawking
temperature, entropy and specific heat capacity are also calculated by two different methods in which we obtain
the same result for stability. Magnetically charged non-black hole EPM solutions are considered in a recent study
[30] in which singularities, both classically and quantum mechanically are investigated thoroughly. The fact that by
employing NED in 2 + 1−dimensions one can construct regular black holes through cutting and pasting method has
also been shown in a separate study [31]. Finally, we note that by choosing k 6= 1 in the flat spacetime electrodynamics,
we avoid the logarithmic potential, once and for all.

[1] M. Born and L. Infeld, Proc. R. Soc. London A 144, 425 (1934).
[2] H. Salazar, A. Garcia and J. Plebanski, J. Math.Phys. 28, 2171 (1987).
[3] H. Salazar, A. Garcia and J. Plebanski, Nuovo Cimento B 84, 65 (1984).
[4] G. W. Gibbons and D. A. Rasheed, Nucl. Phys. B 454, 185 (1995).
[5] E. Fradkin and A. Tseytlin, Phys. Lett. B 163, 123 (1985).
[6] S. Deser and G. W. Gibbons, Class. Quant. Grav. 15, L35 (1998).
[7] E. Ayón-Beato and A. Garcia, Phys. Rev. Lett. 80, 5056 (1998).
[8] M. Cataldo and A. Garcia, Phys. Rev. D 61, 084003 (2000).
[9] M. Banados, C. Teitelboim and J. Zanelli, Phys. Rev. Lett. 69, 1849 (1992).
[10] C. Martinez, C. Teitelboim and J. Zanelli, Phys. Rev. D 61, 104013 (2000).
[11] K.C.K. Chan and R. B. Mann, Phys. Rev. D 50, 6385 (1994).
[12] M. Cataldo, N. Cruz, S. D. Campo, and A. Garcia, Phys. Lett. B 484, 154, (2000).
[13] M. Hassaı̈ne and C. Martı́nez, Phys. Rev. D 75, 027502 (2007).
[14] M. Hassaı̈ne and C. Martı́nez, Class. Quantum Grav. 25, 195023 (2008).
[15] S. H. Hendi, H. R. Rastegar-Sedehi, Gen. Rel. Grav. 41, 1355 (2009) .
[16] H. Maeda, M. Hassaı̈ne and C. Martı́nez, Phys. Rev. D 79 044012 (2009).
[17] S. H. Hendi and B. E. Panah, Phys. Lett. B 684, 77 (2010).
[18] S. H. Hendi, S. Kordestani and S. N. D. Motlagh, Prog. Theor. Phys. 124, 1067 (2010).
[19] H. A. González, M. Hassaı̈ne and C. Martı́nez, Phys. Rev. D 80, 104008 (2009).
[20] D. R. Brill, J. Louko and P. Peldan, Phys. Rev. D 56, 3600 (1997).
[21] J. D. Brown and J.W. York, Phys. Rev. D 47, 1407 (1993).
[22] J. D. Brown, J. Creighton and R.B. Mann, Phys. Rev. D 50, 6394 (1994).
[23] S. Deser and B. Tekin, Phys. Rev. Lett. 89, 101101 (2002).
[24] S. Deser and B. Tekin, Phys. Rev. D 67, 084009 (2003).
[25] D. Grumiller, R. McNees, JHEP 0704, 074 (2007);
[26] Y.S. Myung, Y.-W. Kim, Y.-J. Park, Phys. Rev. D 78, 044020 (2008);
[27] D. Grumiller, W. Kummer, D.V. Vassilevich, Phys. Rept. 369, 327 (2002).
[28] Joao Paulo M. Pitelli and P. S. Letelier, Phys. Rev. D 77, 124030 (2008).
 10

[29] O. Unver and O. Gurtug, Phys. Rev. D 82, 084016 (2010).

[30] S. H. Mazharimousavi, O. Gurtug, M. Halilsoy and O. Unver, Phys. Rev. D 84, 124021 (2011).
[31] S. Habib Mazharimousavi, M. Halilsoy and T. Tahamtan, Phys. Lett. A 376, 893 (2012).
[32] M. Salgado, Class. Quant. Grav. 20, 4551 (2003).
[33] S. H. Mazharimousavi, O. Gurtug and M. Halilsoy , Int. Jour. Mod. Phys. D 18, 2061 (2009).
APPENDIX: Energy Conditions
When a matter field couples to any system, energy conditions must be satisfied for physically acceptable solutions.
k
This is achieved by following the steps as given in [32, 33].T µν = 12 |F | diag(ξ, ξ, −1)

A. Weak Energy Condition (WEC)

The energy - momentum tensor is given in Eq.(14) implies

1 k 1 k 1 k
ρ = −Ttt = |F | ξ, pr = Trr = |F | ξ, pθ = − |F | ,
2 2 2
in which ρ is the energy density and pi are the principal pressures.
The WEC states that,

ρ≥0 and ρ + pi ≥ 0 (i = 1, 2) (A1)

which imposes k > 12 .

B. Strong Energy Condition (SEC)

This condition states that;

2
X
ρ+ pi ≥ 0 and ρ + pi ≥ 0, (A2)
i=1

which yields k ≥ 0. The SEC together with the WEC constraint the parameter k to k > 21 .

C. Dominant Energy Condition (DEC)

DEC states that pef f ≥ 0 in which

2
1X i
pef f = T . (A3)
2 i=1 i

1
and this gives the constraint k ≤ 1. One can show that DEC, together with SEC and WEC, imposes 2 < k ≤ 1.

D. Causality Condition (CC)

Beside the energy conditions, one can impose the causality condition (CC)

pef f
0≤ < 1, (A4)
ρ
2
which implies 3 < k ≤ 1. Therefore if CC is imposed, naturally all other conditions are satisfied.