Advances in High Energy Physics
By varying action () with respect to the gravitational eld
𝜇]
,thedilatoneldΦ,andthegaugeeld
𝜇
,wend
R
𝜇]
=
4
−2
𝜇
Φ
]
Φ+
1
4
𝜇]
(
Φ
)
+2
−4𝛼Φ/(𝑛−2)
𝑌
L
(
)
𝜇𝜂
]
𝜂
−
8
2
−2
4𝛼Φ/(𝑛−2)
2
𝑌
L
(
)
−L
(
)
𝜇]
,
()
∇
2
Φ
=
−2
8
Φ
−4
2
4𝛼Φ/(𝑛−2)
2
𝑌
L
(
)
−L
(
)
,
()
∇
𝜇
−4𝛼Φ/(𝑛−2)
𝑌
L
(
)
𝜇]
=0.
()
In the limiting case where →∞,wehaveL() = .In
this case the system of eld equations ()–() restore the well-
knownequationsofEMdgravity[–],asexpected.
2.1. Finite Action in Canonical and Grand-Canonical Ensem-
bles. In general, the total action given in () is divergent
when evaluated on a solution. One way of dealing with the
divergences of the action is adding some counterterms to
action (). e counterterms should contain a part which
removes the divergence of the gravity part of the action and
a part for dealing with the divergence of the matter action.
Since the horizon of our solution is at, the counterterm
which removes the divergence of the gravity part should be
proportional to
−. e counterterm for the matter part of
theactioninthepresenceofthedilatonisgivenby
ct
=−
1
8
𝜕M
𝑛−1
−
−2
e
+
deriv
,
()
where
e
is given by () and
deriv
is a collection of terms
involving derivatives of the boundary elds that could involve
the curvature tensor constructed from the boundary metric.
Sinceinourcasetheboundaryisatso
deriv
is zero on the
boundary. e variation of the total action (
tot
=+
ct
)about
the solutions of the equations of motion is
tot
=
𝑛−1
𝑎𝑏
𝑎𝑏
−
1
16
𝑛−1
−
−4𝛼Φ/(𝑛−2)
𝑌
L
(
)
𝑎
𝑎𝑏
𝑏
,
()
where
𝑎𝑏
=
−
16
Θ
𝑎𝑏
−Θ
𝑎𝑏
+
−2
e
𝑎𝑏
.
()
Equation()showsthatthevariationofthetotalaction
with respect to
𝑎
will only give the equation of motion of
the nonlinear massless eld
𝑎
provided the variation is at
xed nonlinear massless gauge potential on the boundary.
us, the total action,
tot
=+
ct
, given in () is appropriate
for the grand-canonical ensemble, where
𝑎
=0on the
boundary. But in the canonical ensemble, where the electric
charge [−
−4𝛼Φ/(𝑛−2)
𝑌
L()
𝑎
𝑎𝑏
]is xed on the boundary,
the appropriate action is
tot
=+
ct
+
1
16
⋅
𝜕M‘
𝑛−1
−
−4𝛼Φ/(𝑛−2)
𝑌
L
(
)
𝑎
𝑎𝑏
𝑏
.
()
e last term in () is the generalization of the boundary
term introduced by Hawking and Ross for linear electromag-
netic eld [] and the results of [, ] for the nonlinear
Lifshitz black holes to the exponential nonlinear gauge eld
coupledtothedilatoneld.us,bothincanonicaland
grand-canonical ensemble, the variation of total action about
the solutions of the eld equations is
tot
=
𝑛−1
𝑎𝑏
𝑎𝑏
. ()
at is, the nonlinear gauge eld is absent in the variation
of the total action both in canonical and grand-canonical
ensembles.
In order to obtain the conserved charges of the space-
time,weusethecountertermmethod[–]inspiredby
(A)dS/CFT correspondence. For asymptotically AdS solu-
tions this method works very well [–]. However, in
our paper we have the scalar dilaton eld with a Liouville
potential. It was argued that the presence of Liouville-type
dilaton potential, which is regarded as the generalization of
the cosmological constant, changes the asymptotic behavior
of the solutions to be neither asymptotically at nor (A)dS. It
has been shown that no dilaton dS or AdS black hole solution
exists with the presence of only one Liouville-type dilaton
potential [–]. But, as in the case of asymptotically AdS
spacetimes, according to the domain-wall/QFT (quantum
eld theory) correspondence [–], there may be a suitable
counterterm for the stress-energy tensor which removes the
divergences. In this paper, we deal with the spacetimes with
zero curvature boundary [
𝑎𝑏𝑐𝑑
() = 0], and therefore the
counterterm for the stress-energy tensor should be propor-
tional to
𝑎𝑏
. We nd the nite stress-energy tensor in -
dimensional Einstein-dilaton gravity with Liouville-type in
the form [, ]
𝑎𝑏
=
1
8
Θ
𝑎𝑏
−Θ
𝑎𝑏
+
−2
e
𝑎𝑏
,
()
where
e
is given by
2
e
=
(
−2
)
2
−+1
2Λ
−4𝛼Φ/(𝑛−2)
.
()
In the particular case =0,theeective
2
e
of ()
reduces to
2
= −(−1)(−2)/2Λof the AdS spacetimes.