r
X
i
v
:
g
r

q
c
/
9
9
0
4
0
3
5
v
2
1
6
A
p
r
1
9
9
9
NDAFP59
April 1999
Can primordial wormholes be induced by GUTs at
the early Universe?
S. Nojiri
1
, O. Obregon
2
, S.D. Odintsov
,3
and K.E. Osetrin
4
Department of Mathematics and Physics
National Defence Academy, Hashirimizu Yokosuka 239, JAPAN
Instituto de Fisica de la Universidad de Guanajuato
P.O. Box E143, 37150 Leon Gto., MEXICO
Tomsk State Pedagogical University, 634041 Tomsk, RUSSIA
NTZ, University of Leipzig, Augustusplatz 10/11
04109 Leipzig, GERMANY
abstract
Using large N, 4d anomaly induced oneloop eective action for conformally invariant
matter (typical GUT multiplet) we study the possibility to induce the primordial spheri
cally symmetric wormholes at the early Universe. The corresponding eective equations
are obtained in two dierent coordinate frames. The numerical investigation of these
equations is done for matter content corresponding to N = 4 SU(N) super YangMills
theory. For some choice of initial conditions, the induced wormhole solution with increas
ing throat radius and increasing redshift function is found.
1
email: nojiri@cc.nda.ac.jp
2
email: octavio@ifug3.ugto.mx
3
email: odintsov@mail.tomsknet.ru, odintsov@itp.unileipzig.de
4
email: osetrin@tspu.edu.ru
1
The wormholes are puzzling topological objects (kind of handles of topological origin)
which attract much attention in General Relativity for years. It is not strange as they
may be considered as bridges joining two dierent Universes or two separate regions of
the same Universe (for an introduction, see refs.[1]). There are many speculations related
with hypotetical eects which may be expected near wormholes.
Moreover, it is known that wormholes usually cannot occur as classical solutions of
gravity (with matter) due to violations of energy conditions [1]. Nevertheless, one can
expect that primordial wormholes may present at the very early Universe where quantum
eects play an essential role (or yet in fundamental Mtheory, see,for example, [2]).
Indeed as it was shown in ref.[3] (see also [4]) using quantum stress tensor for confor
mal scalar on spherically symmetric space and in ref.[5] using oneloop eective action in
large N and swave approximation for minimal scalar there may exist semiclassical quan
tum solution corresponding to a Lorentzian wormhole connecting two asymptotically at
regions of the Universe. The corresponding spherically symmetric wormhole solution has
been found numerically in both approximations [3, 5] as well as analytically [5]. That
shows the principal possibility of inducing primordial wormholes at the early Universe (in
its quantum regime). However above discussion has been limited strictly to scalar matter.
And what happens if other types of matter are included? Hence, the very natural question
is: Can primordial wormholes be induced from GUTs at the early Universe?
In the present Letter we try to answer this question for spherically symmetric worm
holes. We use 4d conformal anomaly induced eective action in oneloop and large N
approximation so that one can neglect quantum gravitational contributions. The eec
tive equations for an arbitrary massless GUT containing conformal scalars, spinors and
vectors are explicitly obtained for two forms of spherically symmetric background. Nu
merical solution of these equations is presented for N = 4 super YangMills theory. It
is observed that it depends on the choice of the initial conditions. The initial conditions
admitting inducing of primordial wormholes are also presented.
We rst derive the eective action for conformally invariant matter (for a general
review of eective action in curved space, see[6]). Let us start from Einstein gravity with
N
0
conformal scalars
i
, N
1
vectors A
and N
1/2
Dirac spinors
i
S =
1
16G
_
d
4
x
_
g
(4)
_
R
(4)
2
_
+
_
d
4
x
_
g
(4)
_
_
_
1
2
N
0
i=1
_
g
(4)
i
+
1
6
R
(4)
2
i
_
1
4
N
1
j=1
F
j
F
j
+
N
1/2
k=1
k
D/
k
_
_
_
. (1)
The above matter content is typical for asymptotically free GUT at high energies [6]
as interaction terms and masses at strong curvature are negligible due to asymptotic
freedom. For nite GUTs or asymptotically non free GUTs one should not include
masses if consider only conformally invariant theories and interaction plays no role as
anyway we consider purely gravitational background.
The convenient choice for the spherically symmetric spacetime is the following:
ds
2
= f()
_
f
1
()g
dx
dx
+ r
2
0
d
_
. (2)
2
where , = 0, 1, g
] where
ind
= [f, g
(4)
] [1, g
(4)
] is conformal anomaly
induced action which is quite wellknown [8], g
(4)
corresponds to
ds
2
=
_
g
dx
dx
+ r
2
0
d
_
, g
f
1
()g
. (3)
The conformal anomaly for above matter is wellknown
T = b
_
F +
2
3
2R
_
+ b
R
GB
+ b
2R (4)
where b =
(N
0
+6N
1/2
+12N
1
)
120(4)
2
, b
=
(N
0
+11N
1/2
+62N
1
)
360(4)
2
, b
= 0 but in principle, b
may be
changed by the nite renormalization of local counterterm in gravitational eective action,
F is the square of Weyl tensor, R
GB
is GaussBonnet invariant.
Conformal anomaly induced eective action
ind
may be written as follows [8]:
W = b
_
d
4
x
gF + b
_
d
4
x
g
_
_
22
2
+ 4R
4
3
R2 +
2
3
(
R)
_
+
_
R
GB
2
3
2R
_
1
12
_
b
+
2
3
(b + b
)
__
d
4
x
g
_
{R 62 6()()}
2
R
2
_
(5)
where =
1
2
ln f(), and independent terms are dropped. All 4dimensional quantities
(curvatures, covariant derivatives) in Eq.(5) should be calculated on the metric (3). The
calculation of [1, g
(4)
] =
_
d
4
x
g
__
bF + b
R
GB
+
2b
3
2R
_
ln
R
2
_
+O(R
3
) (6)
where is massdimensional constant parameter, all the quantities are calculated on the
background (3). The condition of application of above expansion is R < R
2
(curvature
is nearly constant), in this case we may limit by only few rst terms.
Let us solve the equations of motion obtained from the above eective Lagrangians
S + . In the following, we use g
as g
if there is no confusion.
ind
(W in Eq.(5)) is rewritten after the reduction to 2 dimensions as follows:
ind
4
=
br
2
0
3
_
d
2
x
g
_
_
R
(2)
+ R
_
2
+
2
3
R
R
(2)
+
1
3
R
2
+b
r
2
0
_
d
2
x
g
_
_
22
2
+ 4R
(2)
4
3
(R
(2)
+ R
)2
+
2
3
(
R
(2)
)
_
+
_
2R
R
(2)
2
3
2R
(2)
_
1
12
_
b
+
2
3
(b + b
)
_
r
2
0
_
d
2
x
_
_
R
(2)
+ R
62 6
_
2
_
R
(2)
+ R
_
2
_
(7)
3
Here R
=
2
r
2
0
is scalar curvature of S
2
with the unit radius.
Let us derive the eective equations of motion. In the following, we work in the
conformal gauge : g
=
1
2
e
2
, g
under consideration is
the product of the original metric tensor and the function e
2
, the equations given by
the variations over g
] with
respect to g
2
r
+ 2
r
_
+
b
r
2
0
16
_
8e
2
r
_
e
2
2
r
_
+8
2
r
2
r
+
8
3
e
2
r
{R
4
} +
32
3
e
2
r
R
4
_
b
+
2
3
(b + b
)
_
r
2
0
16
_
16e
2
r
R
4
+ 4(
r
)
2
2
r
12e
2
r
_
e
2
_
2
r
+ (
r
)
2
__
+ 12
_
2
r
+ (
r
)
2
_
(
r
)
2
_
r
2
0
16
_
1
2
2
r
2(
r
)
2
1
4
2
r
+
3
2
r
_
_
16
3
b
2
r
+ (
r
)
2
_
_
b
+
2
3
(b + b
)
_
_
2
r
+ (
r
)
2
_
_
+r
2
0
_
1
12
be
2
r
_
ln
_
R
4
2
__
r
R
4
_
1
2
2
r
2(
r
)
2
1
4
2
r
+
3
2
r
_
_
2
3
b
2
r
ln
_
R
4
2
_
+
b
4
2
r
_
ln
_
R
4
2
__
(8)
+
_
b
_
2
3
e
2
(
2
r
)
2
+
2e
2
3r
2
0
+
1
3
2
r
R
4
_
+
4
r
2
0
_
b
3
+ b
2
r
_
1
R
4
__
,
R
4
2e
2
2
r
+
2
r
2
0
, (9)
0 =
r
2
0
16G
_
2
r
e
2
+
4
r
2
0
e
2+2
_
+ b
r
2
0
_
2(
2
r
)
2
e
2
8
3
e
2
2
r
(
2
r
) +
4
3
2
r
_
e
2
2
r
1
3
_
2 (
r
)
2
e
2
2
r
+
2
r
_
(
r
)
2
e
2
__
_
_
b
+
2
3
(b + b
)
_
r
2
0
_
2
r
_
e
2
_
2
r
+ (
r
)
2
__
3e
2
_
2
r
+ (
r
)
2
_
2
_
4
+r
2
0
_
4
3
be
2
_
2
r
_
2
ln
_
R
4
2
_
+
4
3
b
2
r
_
e
2
2
r
ln
_
R
4
2
__
+
4be
2
3r
2
0
ln
_
R
4
2
_
4
r
2
0
_
b
3
+ b
2
r
ln
_
R
4
2
_
(10)
+
4
3
be
2
2
r
2
r
_
ln
_
R
4
2
__
4
3
b
2
r
_
e
2
2
r
_
ln
_
R
4
2
___
+
4e
2
2
r
R
4
_
2
3
be
2
_
2
r
_
2
+
2be
2
3r
2
0
+
1
3
b
2
r
R
4
4
r
2
0
_
b
3
+ b
2
r
_
+
2
r
_
2e
2
R
4
_
2
3
be
2
(
2
r
)
2
+
2be
2
3r
2
0
+
1
3
b
2
r
R
4
4
r
2
0
_
b
3
+ b
2
r
___
,
0 =
r
2
0
16G
_
2e
2
_
3
2
r
+ 3 (
r
)
2
+
2
r
_
+
4
r
2
0
e
2+2
_
+b
r
2
0
_
2
2
r
(e
2
2
r
) +
4
3
_
e
2
2
r
2
r
+
2
r
(e
2
2
r
)
_
8
3r
2
0
2
r
1
6
{
r
R
4
r
+
r
R
4
r
}
_
_
b
+
2
3
(b + b
)
_
r
2
0
_
1
2
2
r
R
4
+
1
2
{
r
R
4
r
+
r
R
4
r
}
+3
2
r
_
e
2
_
2
r
+ (
r
)
2
__
6
r
_
e
2
2
r
+ (
r
)
2
___
(11)
Note that now the real 4 dimensional metric is given by
ds
2
= e
2+2
dx
+
dx
+ r
2
0
e
2
d
2
. (12)
Let us make the following change:
r
etc. as
r
= e
+
l
,
2
r
= e
2+2
_
2
l
+
l
l
+
l
l
_
. (13)
Using the new radial coordinate l, the metric (12) is rewritten as follows:
ds
2
= f(l)dt
2
+ dl
2
+ r(l)
2
d
2
f(l) = e
2+2
r(l) = r
0
e
. (14)
Here f(l) is called a redshift function and r(l) is a shape function. If f(l) and r(l) are
smooth positivedenite functions, which satisfy the conditions:
f(l) 1, r(l) l when l
f(l), r(l) nite when l 0 , (15)
the metric expresses the wormhole which connects two asymptotically at universes.
For the choice of the metric (14), Eqs.(8), (9), (10) and (11) are rewritten as follows:
0 =
r
2
0
48G
e
4+4
_
2
l
+
l
_
5
+
b
r
2
0
e
4+4
16
_
8e
2
l
_
e
2
_
2
l
+
l
l
+ (
l
)
2
__
+8
_
2
l
+
l
l
+ (
l
)
2
_ _
2
l
+ (
l
)
2
+
l
_
+
8
3
e
2
l
{R
4
} +
32
3
e
2
l
R
4
_
b
+
2
3
(b + b
)
_
r
2
0
e
4+4
16
_
16e
2
l
R
4
+ 4(
l
)
2
_
2
l
+ (
l
)
2
+
l
_
12e
2
l
_
e
2
_
2
l
+
l
l
+ 2 (
l
)
2
__
+12
_
2
l
+
l
l
+ 2 (
l
)
2
_
(
l
)
2
_
r
2
0
e
2+2
16
_
1
2
2
l
3
2
(
l
)
2
+
1
2
l
1
4
2
l
+
5
4
1
4
l
_
e
2+2
_
16
3
b
2
l
+
l
l
+ (
l
)
2
_
+ (
l
)
2
_
_
b
+
2
3
(b + b
)
_
_
2
l
+
l
l
+ 2 (
l
)
2
_
_
+r
2
0
_
1
12
be
4+2
l
_
ln
_
R
4
2
__
l
R
4
(16)
e
2+2
_
1
2
2
l
3
2
(
l
)
2
+
1
2
l
1
4
2
l
+
5
4
1
4
l
_
e
2+2
_
2
3
b
_
2
l
+ (
l
)
2
+
l
_
ln
_
R
4
2
_
+
b
4
_
2
l
+
l
l
+
l
l
_
_
ln
_
R
4
2
__
+
_
b
_
2
3
e
2
_
2
l
+ (
l
)
2
+
l
_
2
+
2e
2
3r
2
0
+
1
3
_
2
l
+
l
l
+
l
l
_
R
4
_
+
4
r
2
0
_
b
3
+ b
_
_
2
l
+ (
l
)
2
+
l
_
_
1
R
4
__
,
R
4
2e
2
_
2
l
+ (
l
)
2
+
l
_
+
2
r
2
0
, (17)
0 =
r
2
0
e
2
16G
_
e
2
_
2
l
+
l
l
+
l
l
_
e
2
+
4
r
2
0
e
2
_
+b
r
2
0
e
2+4
_
2
_
2
l
+
l
l
+ (
l
)
2
_
2
8
3
_
2
l
+ (
l
)
2
+
l
_ _
2
l
+
l
l
+ (
l
)
2
__
+
4
3
e
2
_
2
l
+
l
l
+
l
l
_ _
e
2
_
2
l
+
l
l
+ (
l
)
2
__
1
3
_
2 (
l
)
2
_
2
l
+ (
l
)
2
+
l
_
6
+e
2
_
2
l
+
l
l
+
l
l
_ _
e
2
(
l
)
2
___
_
b
+
2
3
(b + b
)
_
r
2
0
e
2+4
_
e
2
_
2
l
+
l
l
+
l
l
_ _
e
2
_
2
l
+
l
l
+ 2 (
l
)
2
__
3
_
2
l
+
l
l
+ 2 (
l
)
2
_
2
_
(18)
+r
2
0
e
2+4
_
4
3
b
_
2
l
+ (
l
)
2
+
l
_
2
ln
_
R
4
2
_
+
4
3
be
2
_
2
l
+
l
l
+
l
l
_
_
e
2
_
2
l
+ (
l
)
2
+
l
_
ln
_
R
4
2
__
+
4be
4
3r
2
0
ln
_
R
4
2
_
4
r
2
0
_
b
3
+ b
_
e
2
_
2
l
+
l
l
+
l
l
_
ln
_
R
4
2
_
+
4
3
b
_
2
l
+ (
l
)
2
+
l
_ _
2
l
+
l
l
+
l
l
_
_
ln
_
R
4
2
__
4
3
be
2
_
2
l
+
l
l
+
l
l
_
_
e
2
_
2
l
+
l
l
+
l
l
_
_
ln
_
R
4
2
___
+
4e
2
_
2
l
+ (
l
)
2
+
l
_
R
4
_
2
3
be
2
_
2
l
+ (
l
)
2
+
l
_
2
+
2be
4
3r
2
0
+
e
2
3
b
_
2
l
+
l
l
+
l
l
_
R
4
4
r
2
0
_
b
3
+ b
_
e
2
_
2
l
+ (
l
)
2
+
l
_
+e
2
_
2
l
+
l
l
+
l
l
_
_
2
R
4
_
2
3
be
4
_
2
l
+ (
l
)
2
+
l
_
2
+
2b
3r
2
0
+
1
3
be
2
_
2
l
+
l
l
+
l
l
_
R
4
4e
2
r
2
0
_
b
3
+ b
_
_
2
l
+ (
l
)
2
+
l
_
___
,
0 =
r
2
0
e
2
16G
_
2e
4
_
3
2
l
+ 4
l
l
+ 6 (
l
)
2
+
2
l
+ (
l
)
2
_
+
4
r
2
0
e
2
_
+b
r
2
0
e
2
_
2e
2
_
2
l
+
l
l
+
l
l
_ _
e
2
_
2
l
+
l
l
+ (
l
)
2
__
+
4
3
_
e
4
_
2
l
+
l
l
+ (
l
)
2
_ _
2
l
+ (
l
)
2
+
l
_
+e
2
_
2
l
+
l
l
+
l
l
_ _
e
2
_
2
l
+ (
l
)
2
+
l
__
8e
2
3r
2
0
_
2
l
+
l
l
+ (
l
)
2
_
e
2
3
l
R
4
_
b
+
2
3
(b + b
)
_
r
2
0
e
2+2
_
1
2
_
2
l
+
l
l
+
l
l
_
R
4
+
l
R
4
l
(19)
+3
_
2
l
+
l
l
+
l
l
_ _
e
2
_
2
l
+
l
l
+ 2 (
l
)
2
__
7
6e
l
_
e
+3
2
l
+
l
l
+ 2 (
l
)
2
___
.
The equations (16) and (18) contain the 6th order derivatives of ,
6
l
and the equations
(16), (18) and (19) contain the 4th order derivatives of . Therefore we need to impose
the initial conditions including 5th order derivative of and 3rd order derivative of :
l

l=0
=
2
l

l=0
=
3
l

l=0
=
4
l

l=0
=
5
l

l=0
= 0
l

l=0
=
2
l

l=0
=
3
l

l=0
= 0 . (20)
The equations (16) and (18) contain the 5th order derivatives of from
4
l
R
4
but
5
l
always appears in the form
5
l
. Therefore these terms vanish at l = 0. This tells
that we cannot impose the initial condition
4
l
= 0 since we cannot solve the dierential
equations with respect to
5
l
at l = 0.
We should also note that all the equations (16), (18) and (19) are not the dynamical
equations of motion but the combinations with the constraint. The constraint can be
found by putting l = 0 under the initial conditions (20), when the equations (16) and
(18) have the following form:
0 =
_
b + b
+
3
2
b
+ 8b
0
_
(4)
0
+ 14be
2
0
r
0
2
(6)
0
, (21)
0 =
3e
2
0
8 G
2be
4
0
ln
_
2
2
r
0
2
_
+r
0
2
_
b + b
+
3
2
b
2b
0
_
(4)
0
be
2
0
r
0
4
(6)
0
, (22)
0 =
e
2
0
G r
0
2
+ 4 (2 b + 3 b
)
(4)
0
. (23)
Equations (21), (22) and (23) will be compatible if
0
satises to the following equation:
0 = 224 b (2 b + 3 b
) G ln
_
2
2
r
0
2
_
+e
2
0
(54 b 30 b
+ 81 b
+ 40 b
0
) . (24)
Then we can choose the initial conditions as (20) and (24) with arbitrary 
l=0
. Then
under these initial conditions, we can use any two of equations (16), (18) and (19) as
dynamical equations of motion. Equation (24) admits the real solution for
0
if
b (112 b + 168 b
) e
1
2
+
27 (2 b+3 b
)
20 b
G ln
_
2
2
r
0
2
_
< 5 b
(25)
and from (24) we have the restriction
0
1
4
+
54 b + 81 b
40 b

(26)
From equations (16), (18), (19) and initial conditions (20) we get the following behavior
of and near l = 0
(l) =
0
+
2b + 2b
+ 3b
+ 16b
0
80640 G r
0
4
b (2b + 3b
)
l
6
+ O(l
8
) (27)
(l) =
0
e
2
0
96 G r
0
2
(2b + 3b
)
l
4
+ O(l
6
) (28)
8
Then for initial conditions (20) (l) and (l) are decreasing functions near l = 0.
For the numerical calculation as an example we will consider N = 4 SU(N) superYM
theory where (see, for example, [10])
b = b
=
N
2
1
(8 )
2
, b
= 0. (29)
Such theory became popular recently in relation with AdS/CFT correspondence. We take
it as typical example of GUT. Note also that as it was mentioned in ref.[10] the explicit
choice of b
e
2
0
198 G r
0
2
b
l
4
+ O(l
6
) (31)
and from (2425) we obtain
(r
0
)
2
> 2 b > 0,
0
> 21/10 (32)
(r
0
)
2
= 2 b > 0,
0
= 21/10 (33)
(r
0
)
2
< 2 0 < b <
5 e
16/5
112 G ln(
2
2
r
0
2
)
,
8
5
<
0
<
21
10
. (34)
Let us choose for numeric investigation, that
G = 1,
2
= 2, r
0
2
= 2/
2
= 1.
The results of numerical calculation are given by Figures 1 and 2 for redshift and shape
functions where N = 5.
50 100 150 200
l
25
50
75
100
125
150
175
f
Fig.1 n=5
100 200 300 400 500 600
l
2
4
6
8
r
Fig.2 n=5
Hence we showed that for above choice of initial conditions the inducing of primordial
wormholes at the early Universe is rather unrealistic. The throat radius of wormhole
quickly shrinks to zero.
Let us now consider another choice of initial conditions
l

l=0
=
2
l

l=0
=
3
l

l=0
= 0,
4
l

l=0
= 0
l

l=0
=
2
l

l=0
=
3
l

l=0
= 0 . (35)
9
From equations (16), (18), (19) and initial conditions (35) we get the following behavior
of and near l = 0
4
l

l=0
=
1
792b
2
G
_
99be
2
0
+ 4b
2
G
0
0
_
51 + 36e
4
0
+ 20e
4
0
0
2
_
4
0
_
b
3
G
_
99e
2
0
(21 10
0
) + bG(51 + 36e
4
0
+ 20e
4
0
0
2
)
2
_
_
(36)
4
l

l=0
=
1
132 b
2
G
_
b
2
G
_
51 36 e
4
0
20 e
4
0
0
2
_
+
_
b
3
G
_
99 e
2
0
(21 10
0
) + b G (51 + 36 e
4
0
+ 20 e
4
0
0
2
)
2
_
_
(37)
where = 1.
Moreover we obtain the conditions
5
l

l=0
=
5
l

l=0
= 0, (38)
or
891 e
2
0
(10
0
21) = 8 b G
_
51 + 36 e
4
0
+ 20 e
4
0
0
2
_
2
,
0
>
21
10
. (39)
For case
0
= 21/10 we have
4
l

l=0
=
1
2200
_
275 e
21/5
b G
+ (1680 e
42/5
2380)
0
+ 4116 e
42/5
0
2
_
, (40)
4
l

l=0
=
1
110
_
85 60 e
42/5
147 e
42/5
0
_
(41)
Then functions (l) and (l) will increase near l = 0 if
0
<
5
294 b G
_
b G
_
12 + 17 e
42/5
_
e
42/5
b G
_
231 e
21
5
+ b
_
17 12 e
42
5
_
2
G
_
_
for G = 1 and b = (N
2
1)/(8)
2
we have
0
< 0.4081. The explicit numerical study
(see Figures 3 and 4) shows that shape and redshift functions are increasing giving the
window for inducing of primordial wormholes.
0.2 0.4 0.6 0.8
l
8.2
8.3
8.4
r
Fig.3 n=10
0.2 0.4 0.6 0.8
l
30
31
32
33
f
Fig.4 n=10
On these gures
0
= 21/10 and
0
= 0.41.
10
Hence we numerically proved that at least for some initial conditions GUTs at the early
Universe may help in producing of primordial wormholes. It is of course the open question
which initial conditions from the two classes presented above lead to more stable cong
uration. Moreover, due to complicated structure of eld equations and initial conditions
itself we cannot classify from the very beginning the initial conditions as supporting (or
not) wormholes production. Nevertheless for any specic GUT under discussion above
study may be easily repeated and principal possibility of wormholes inducing may be
shown at least numerically.
Acknowledgments. The research by SDO was partially supported by a RFBR Grant
N990216617, by Saxonian Min. of Science and Arts and by Graduate College Quantum
Field Theory at Leipzig University. The research by KEO was partially supported by
a RFBR Grant N990100912 and the research by OO has been partially supported by
CONACyT grants 3898PE9608 and 28454E.
References
[1] M.S. Morris and K.S. Thorne, Am.J.Phys. 56 (1988) 395; M.S. Thorne, K.S. Thorne
and U. Yurtsever, Phys.Rev.Lett. 61 (1988) 1446
[2] G. Gibbons, preprint hepth/9801106.
[3] D.Hochberg, A. Popov and S.N. Syshkov, Phys.Rev.Lett. 78 (1997) 2050.
[4] V. Khatsymovsky, In proceedings of II Int.Conf. Quantum Field Theory and Gravity,
Eds.I.L. Buchbinder and K.E. Osetrin, TGPU Publishing, Tomsk, 1997.
[5] S. Nojiri, O. Obregon, S.D. Odintsov and K.E. Osetrin, Phys.Lett. B449 (1999) 173.
[6] I.L. Buchbinder, S.D. Odintsov and I.L. Shapiro, Eective Action in Quantum Grav
ity, IOP Publishing, Bristol and Philadelphia, 1992.
[7] S. Nojiri and S.D. Odintsov, Phys.Rev. D59 (1999) 044026.
[8] R.J. Reigert, Phys.Lett. B134 (1984)56; E.S. Fradkin and A. Tseytlin, Phys.Lett.
B134 (1984) 187; I.L. Buchbinder, S.D. Odintsov and I.L. Shapiro, Phys.Lett. B162
(1985) 92, for a recent review see D. Anselmi, hepth/9903059.
[9] S. Nojiri and S.D. Odintsov, hepth/9802160, Int.J.Mod.Phys. A, to appear.
[10] I. Brevik and S.D. Odintsov, hepth/9902184, Phys.Lett. B, to appear.
[11] I. Antoniadis and E. Mottola, Phys.Rev. D45 (1992) 2013; S.D. Odintsov, Z.Phys.
C54 (1992) 531.
11