Morris -Thorne wormholes with a

cosmological constant
Jose P. S. Lemos
Department of Physics, Columbia University,
New York, NY 10027, &
Centro Multidisciplinar de Astrofsica - CENTRA,
Departamento de Fsica, Instituto Superior Tecnico,
Av. Rovisco Pais 1, 1049-001 Lisbon
e-mail: [email protected]
Francisco S. N. Lobo
Centro de Astronomia e Astrofsica da Universidade de Lisboa,
Campo Grande, Ed. C8 1749-016 Lisbon
e-mail: [email protected]
Sergio Quinet de Oliveira
Observat orio Nacional - MCT,
Rua General Jose Cristino 77, 20921-400 Rio de Janeiro
e-mail: [email protected]
Abstract
First, the ideas introduced in the wormhole research eld since
the work of Morris and Thorne are reviewed, namely, the issues of
energy conditions, wormhole construction, stability, time machines
and astrophysical signatures. Then, spherically symmetric and static
traversable Morris -Thorne wormholes in the presence of a generic cos-
mological constant are analyzed. A matching of an interior solution
to the unique exterior vacuum solution is done using directly the Ein-
stein equations. The structure as well as several physical properties
and characteristics of traversable wormholes due to the eects of the
cosmological term are studied. Interesting equations appear in the
process of matching. For instance, one nds that for asymptotically
at and anti-de Sitter spacetimes the surface tangential pressure P
of the thin shell, at the boundary of the interior and exterior solu-
tions, is always strictly positive, whereas for de Sitter spacetime it
can take either sign as one could expect, being negative (tension) for
relatively high and high wormhole radius, positive for relatively high
mass and small wormhole radius, and zero in-between. Finally, some
specic solutions with , based on the Morris -Thorne solutions, are
provided.
1
1 Review and Introduction
It is now 15 years that traversable wormhole theory started in earnest from
the work of Mike Morris and Kip Thorne published in 1988 [1]. It was rst
introduced as a tool for teaching general relativity, as well as an attempt to
allure young students into the eld, for instance those that had read Contact,
a novel of Carl Sagan that uses a wormhole to shortcut a large astronomical
distance, but it rapidly spread into several branches. These developments
culminated with the publication of the book Lorentzian Wormholes: From
Einstein to Hawking by Visser [2], where a review on the subject up to 1995,
as well as new ideas are developed and hinted at. It is our intention in this
introduction to do a brief review on the subject of wormholes. The subject
has grown substantially, and it is now almost out of control. We will focus on
the work developed after Vissers book was published (most of the references
prior to its releasing are in it), paying attention to the issues that branched
out of [1], like the issue of energy conditions, wormhole construction, stability,
time machines and astrophysical signatures.
1.1 The beginning
It is true that Wheeler [3], before the work of Morris and Thorne [1], had seen
wormholes, such as Reissner-Nordstr om or Kerr wormholes, as objects of the
quantum foam connecting dierent regions of spacetime and operating at
the Planck scale, which were transformed later into Euclidean wormholes by
Hawking [4] and others, but these Wheeler wormholes were not traversable,
one could not cross them from one side to the other and back, and further-
more would, in principle, develop some type of singularity [5]. Having been
a student of Wheeler, and having further learned through Wheelers interac-
tion with Zeldovich on the trace energy condition (which states 3p, with
being the energy density and p the pressure of the uid on its rest frame)
that energy conditions are on a shaky ground [6, 7], Thorne together with
his student Morris [1], understood that wormholes, with two mouths and a
throat, might be objects of nature, as stars and black holes are. Indeed, it
is a basic fact for the construction of traversable wormholes that the null
energy condition, the weakest of the conditions, has to be violated.
2
1.2 Energy conditions
The weak energy condition says that the energy density of any system at
any point of spacetime for any timelike observer is positive (in the frame of
the matter this amounts to > 0 and + p 0), and when the observer
moves at the speed of the light it has a well dened limit, called the null
energy condition ( + p 0). The weak and null energy conditions are the
weakest of the energy conditions (the null being even weaker than the weak),
their violation signals that the other energy conditions are also violated. In
Hawking and Ellis book [8] the weak energy condition is thought a physically
reasonable energy condition, that at least all classical systems should obey.
Afterwards it was found that it could be violated for quantum systems, such
as in the Casimir eect and Hawking evaporation (see [9] for a short review).
It was further found that for quantum systems in classical gravitational back-
grounds the weak or null energy conditions could only be violated in small
amounts, and a violation at a given time through the appearance of a nega-
tive energy state, would be overcompensated by the appearance of a positive
energy state soon after. This idea gave rise to the averaged energy condition
[10], and to the quantum inequalities which, being intermediate between the
pointwise energy conditions and the averaged energy conditions, limit the
magnitude of the negative energy violations and the time for which they are
allowed to exist, yielding information on the distribution of the negative en-
ergy density in a nite neighborhood [11, 12, 13]. It seems that the situation
has changed drastically, it has been now shown that even classical systems,
such as those built from scalar elds non-minimally coupled to gravity, vi-
olate all the energy conditions [14] (see also [15] for other violations of the
energy conditions). Thus, gradually the weak and null energy conditions,
and with it the other energy conditions, might be losing their status of a
kind of law.
1.3 Wormhole construction: a synthesis
Surely, this has had implications on the construction of wormholes. First, in
the original paper [1], Morris and Thorne constructed wormholes by hand,
that is, one gives the geometry rst, which was chosen as spherically sym-
metric, and then manufacture the exotic matter accordingly. The engineering
work was left to an absurdly advanced civilization, which could manufacture
3
such matter and construct these wormholes. Then, once it was understood
that quantum eects should enter in the stress-energy tensor, a self-consistent
wormhole solution of semiclassical gravity was found [16], presumably obey-
ing the quantum inequalities. These quantum inequalities when applied to
wormhole geometries imply that the exotic matter is conned to an extremely
thin band of size only slightly larger than the Planck length, in principle
preventing traversability [12]. Finally with the realization that nonminimal
scalar elds violate the weak energy condition a set of self-consistent classical
wormholes was found [17]. It is fair to say that, though outside this main-
stream, classical wormholes were found by Homer Ellis back in 1973 [18],
and related self-consistent solutions were found by Kirill Bronnikov in 1973
[19], Takeshi Kodama in 1978 [20], and Gerard Clement in 1981 [21], these
papers written much before the wormhole boom originated from Morris and
Thornes work [1] (see [22] for a short account of these previous solutions).
A self-consistent Ellis wormhole was found again by Harris [23] by solving,
through an exotic scalar eld, an exercise for students posed in [1].
1.4 Further wormhole construction
Traversable wormhole theory achieved the end of its rst stage after the writ-
ing of the monograph on the subject by Visser in 1995 [2]. This monograph
is fairly complete on citations, so we refer the reader to it for a bibliographic
search up to 1995. We refer here to some developments afterwards, quoting
older references when appropriate.
Further wormhole construction in general relativity:
Visser led the way through several works. Indeed, Visser [24] constructed
wormholes with polyhedrical symmetry in 1989, generalized a suggestion of
Roman for a conguration with two wormholes [1] into a Roman ring [25],
he started a study on generic dynamical traversable wormhole throats [26]
in 1997, found classically consistent solutions with scalar elds [14] in 1999,
and has also found self-dual solutions [27]. Other authors have made also
interesting studies. Before Vissers book we can quote the paper by Frolov
and Novikov, where they mix wormhole and black hole physics [28]. After
the book, particularly interesting wormholes with toroidal symmetry were
found by Gonzalez-Das [29], wormhole solutions inside cosmic strings were
found by Clement [30], and Aros and Zamorano [31], wormholes supported
4
by strings by Schein, Aichelburg and Israel, [32], rotating wormholes were
found by Teo [33], consistent solutions of the EinsteinYang-Mills theory
in connection with primordial wormhole formation were found in [34], theo-
rems for the impossibility of existence of wormholes in some Einstein-scalar
theories were discussed by Saa [35], wormholes with stress-energy tensor of
massless neutrinos and other massless elds by Krasnikov [36], wormholes
made of a crossow of dust null streams were discussed by Hayward [37] and
Gergely [38], and self consistent charged solutions were found by Bronnikov
and Grinyok [39].
Wormhole construction with arbitrarily small violations of the energy con-
ditions:
One of the main areas in wormhole research is to try to avoid as much
as possible the violation of the null energy condition. For static wormholes
the null energy condition is violated [1, 2]. Several attempts have been made
to overcome somehow this problem: Morris and Thorne already had tried
to minimize the violating region in the original paper [1], Visser [24] found
solutions where observers can pass the throat without interacting with the
exotic matter, which was pushed to the corners, and Kuhttig [40] has found
that the region made of exotic matter can be made arbitrarily small. For dy-
namic wormholes, the violation of the weak energy condition can be avoided,
but the null energy condition, more precisely the averaged null energy condi-
tion is not preserved [26, 41, 42], although in [43] it has been found that the
quantity of violating matter can be made arbitrarily small, a result in line
with [40] for static wormholes.
Wormhole construction with a cosmological constant :
Some papers have added a cosmological constant to the wormhole con-
struction. Kim [44] found thin shell solutions in the spirit of Visser [2],
Roman [45] found a wormhole solution inating in time to test whether one
could evade the violation of the energy conditions, Delgaty and Mann [46]
looked for new wormhole solutions with , and DeBenedectis and Das [47]
found a general class with a cosmological constant. Here, we will further
study wormholes in a spacetime with a cosmological constant, as will be
detailed below.
5
Wormhole construction in other theories of gravitation:
In alternative theories to general relativity wormhole solutions have been
worked out. In higher dimensions solutions have been found by Chodos and
Detweiler [48], Clement [49], and DeBenedictis and Das [50], in Brans-Dicke
theory by Nandi and collaborators [51], in Kaluza-Klein theory by Shen and
collaborators [52], in Einstein-Gauss-Bonnet by Kar [53], Anchordoqui and
Bergliaa found a wormhole solution in a brane world scenario [54] further
examined by Barcelo and Visser [55], and Koyama, Hayward and Kim [56]
examined a two-dimensional dilatonic theory.
1.5 Stability
To know the stability of an object against several types of perturbation is
always an important issue. Wormholes are not an exception. Not many
works though are dedicated to the stability theory of wormholes, although
the whole formalism developed for relativistic stars and black holes could be
readily used in wormholes. Visser [57], Poisson and Visser [58] and Ishak
and Lake [59] studied the stability of wormholes made of thin shells and
found, in the parameter space (P/)(throat radius/mass), where and P
are, respectively, the surface energy density and surface tangential pressure,
those wormholes for which there are stable solutions. For the Ellis drainhole
[18], Armendariz-Picon [60] nds that it is stable against linear perturbations,
whereas Shinkai and Hayward [61] nd this same class unstable to nonlinear
perturbations. Bronnikov and Grinyok [39, 62] found that the consistent
wormholes of Barcelo and Visser [17] are unstable.
1.6 Wormholes as time machines
An important side eect of wormholes is that they can be converted into
time machines, by performing a sucient delay to the time of one mouth in
relation to the other. This can be done either by the special relativistic twin
paradox method [63] or by the general relativistic redshift way [64]. The im-
portance of wormholes in the study of time machines is that they provide a
non-eternal time machine, where closed timelike curves appear to the future
of some hypersurface, the chronology horizon (a special case of a Cauchy hori-
zon) which is generated in a compact region in this case. Since time travel
to the past is in general unwelcome, it is possible to test whether classical or
6
semiclassical eects will destroy the time machine. It is found that classically
it can be easily stabilized [63, 2]. Semiclassically, there are calculations that
favor the destruction [65, 66], leading to chronology protection [66], others
that maintain the machine [67, 25]. Other simpler systems that simulate a
wormhole, such as Misner spacetime which is a species of two-dimensional
wormhole, have been studied more thoroughly, with no conclusive answer.
For Misner spacetime the debate still goes on, favoring chronology protec-
tion [68], disfavoring it [69], and back in favoring [70]. The upshot is that
semiclassical calculations will not settle the issue of chronology protection
[71], one needs a quantum gravity, as has been foreseen sometime before by
Thorne [72].
1.7 Towards a unied view: From stars to wormholes
There is now a growing consensus that wormholes are in the same chain of
stars and black holes. For instance, Gonzalez-Das [29] understood that an
enormous pressure on the center ultimately meant a negative energy density
to open up the tunnel, DeBenedectis and Das [47] mention that the stress-
energy supporting the structure consists of an anisotropic brown dwarf star,
and the wormhole joining one Friedmann-Robertson-Walker universe with
Minkowski spacetime or joining two Friedmann-Robertson-Walker universes
[26] could be interpreted, after further matchings, as a wormhole joining a
collapsing (or expanding) star to Minkowski spacetime or a wormhole joining
two dynamical stars, respectively. It has also been recognized, and empha-
sized by Hayward [73], that wormholes and black holes can be treated in a
unied way, the black hole being described by a null outer trapped surface,
and the wormhole by a timelike outer trapped surface, this surface being
the throat where incoming null rays start to diverge [42, 73]. Thus, it seems
there is a continuum of objects from stars to wormholes passing through black
holes, where stars are made of normal matter, black holes of vacuum, and
wormholes of exotic matter. Although not so appealing perhaps, wormholes
could be called exotic stars.
1.8 Astrophysical signatures
Stars are common for everyone to see, black holes also inhabit the universe
in billions, so one might tentatively guess that wormholes, formed or con-
7
structed from one way or another, can also appear in large amounts. If
they inhabit the cosmological space, they will produce microlensing eects
on point sources at non-cosmological distances [74], as well as at cosmologi-
cal distances, in this case gamma-ray bursts could be the objects microlensed
[75, 76]. If peculiarly large wormholes will produce macrolensing eects [77].
1.9 Aim of this paper
In this paper we extend the Morris -Thorne wormhole solutions [1] by in-
cluding a cosmological constant . Morris -Thorne wormholes, with = 0,
have two asymptotically at regions. By adding a positive cosmological con-
stant, > 0, the wormholes have two asymptotically de Sitter regions, and
by adding a negative cosmological constant, < 0, the wormholes have two
asymptotically anti-de Sitter regions. There are a number of reasons to study
wormholes with generic that a technologically absurdly advanced civiliza-
tion might construct. For > 0, we know that an inationary phase of
the ultra-early universe demands it, and moreover, from recent astronomical
observations, it seems that we live now in a world with > 0. On the other
hand, < 0 is the vacuum state for extended theories of gravitation such as
supergravity and superstring theories, and, in addition, even within general
relativity, a negative cosmological constant permits solutions of black holes
with horizons with topology dierent from the usual spherical [78, 79] (see
[80] for a review), which could be turned into wormhole solutions by adding
some exotic matter, although we do not attempt it here.
We follow the spirit of the Morris and Thorne paper [1], in that worm-
hole theory is a good tool to teach general relativity and a subject that
attracts students. We analyze distributions of matter similar to [1] but now
with generic , i.e, we analyze spherically symmetric and static traversable
Morris -Thorne wormholes in the presence of a cosmological constant. The
more complicated issue of the formalism of junction conditions, that Mor-
ris and Thorne so well evaded [1], is here treated also in a pedagogical way
through the direct use of Einstein eld equation, and the matter content of
the thin shell separating the wormhole from the exterior spacetime is found.
In this way, an equation connecting the radial tension at the mouth with
the tangential surface pressure of the thin shell is derived. The structure as
well as several physical properties and characteristics of traversable worm-
holes due to the eects of the cosmological term are studied. We nd that
8
for asymptotically at and anti-de Sitter spacetimes the surface tangential
pressure P of the thin shell is always strictly positive, whereas for de Sitter
spacetime it can take either sign as one could expect, being negative (tension)
for relatively high and high wormhole radius, positive for relatively high
mass and small wormhole radius, and zero in-between. Finally, some specic
solutions with , based on the Morris -Thorne solutions, are provided. In
presenting these solutions we dwell mostly on the case = 0, and > 0,
and comment briey on < 0. The plan of the paper is as follows: In section
2 we present the Einstein eld equation for a wormhole metric and perform
the junction to an external asymptotically Minkowski, de Sitter, or anti-de
Sitter spacetime. In section 3 we give some wormhole geometries, analogous
to [1] having = 0, > 0 and < 0, and study some of their properties.
In section 4 we conclude.
2 Einstein eld equation for wormholes with
a generic cosmological constant
2.1 The Einstein eld equation with generic , setting
the nomenclature
To set the nomenclature, the Einstein eld equation with a cosmological
constant is given, in a coordinate basis, by G

+ g

= 8Gc
4
T

, in
which G

is the Einstein tensor, given by G

= R

1
2
g

R, R

is the
Ricci tensor, which is dened as a contraction of the Riemann (or curvature)
tensor, R

= R

, and R is the scalar curvature dened as a contraction

of the Ricci tensor, R = R

. The Riemann tensor is a function of the second

order derivatives of the metric components g

. T

is the stress-energy tensor

of the matter, and the cosmological constant [81].
2.1.1 The spacetime metric
We will be interested in the spacetime metric, representing a spherically
symmetric and static wormhole, given by [1]
ds
2
= e
2(r)
c
2
dt
2
+
dr
2
1 b(r)/r
+ r
2
(d
2
+ sin
2
d
2
) , (1)
9
where (c t, r, , ) are the usual spacetime spherical coordinates, and (r)
and b(r) are arbitrary functions of the radial coordinate r. (r) is designated
the redshift function, for it is related to the gravitational redshift, and b(r)
is denominated the shape function, because as can be shown by embedding
diagrams, it determines the shape of the wormhole [1]. The radial coordinate
has a range that increases from a minimum value at r
o
, corresponding to the
wormhole throat, to a maximum a corresponding to the mouth. At r
o
one
has to join smoothly this spherical volume to another spherical volume copy
with r ranging again from r
o
to a, see Figure 1. In addition, one has then to
join each copy to the external spacetime from a to , as will be done.
r
o
r
o
smoothly
join
a
a
Figure 1 - The two copies of the spherical volume from r
o
to a have to be smoothly
joined at r
o
.
The mathematical analysis and the physical interpretation will be sim-
plied using a set of orthonormal basis vectors. These may be interpreted
as the proper reference frame of a set of observers who remain at rest in the
10
coordinate system (ct, r, , ), with (r, , ) xed. Denote the basis vectors
in the coordinate system as e
t
, e
r
, e

, and e

. Then, using the following

transformation, e

=

, with

= diag
_
e

, (1 b/r)
1/2
, r
1
, (r sin )
1
_
, (2)
where the notation means that the non-diagonal terms of the matrix are zero,
one nds
_

_
e

t
= e

e
t
e
r
= (1 b/r)
1/2
e
r
e

= r
1
e

= (r sin )
1
e

.
(3)
In this basis the metric components assume their Minkowskian form, given
by,
g

=

= diag(1, 1, 1, 1). (4)

In the orthonormal reference frame, the Einstein eld equation with a generic
cosmological constant, is given by
G

+

=
8G
c
4
T

. (5)
The Einstein tensor, given in the orthonormal reference frame by G

=
R

1
2
Rg

, yields for the metric (1) the following non-zero components
G

t
=
b

r
2
, (6)
G
r r
=
b
r
3
+ 2
_
1
b
r
_

r
, (7)
G

=
_
1
b
r
__

+ (

)
2

r b
2r(r b)

r b
2r
2
(r b)
+

r
_
, (8)
G

= G

, (9)
where a prime denotes a derivative with respect to the radial coordinate r.
11
2.1.2 The stress-energy tensor
The Einstein eld equation requires that the Einstein tensor be proportional
to the stress-energy tensor. In the orthonormal basis the stress-energy ten-
sor, T

, must have an identical algebraic structure as the Einstein tensor
components, G

, i.e., Equations (6)-(9). Therefore, the only non-zero com-
ponents of T

are T

t
, T
r r
, T

, and T

. These are given an immediate

physical interpretation,
T

t
= (r)c
2
, (10)
T
r r
=(r) , (11)
T

= p(r), (12)
T

= p(r) , (13)
in which (r) is the energy density, (r) is the radial tension, with (r) =
p
r
(r), i.e., it is the negative of the radial pressure, p(r) is the pressure
measured in the tangential directions, orthogonal to the radial direction. T

may include surface quantities as we will see.
2.1.3 The cosmological constant and the total stress-energy tensor
To obtain a physical interpretation of the cosmological constant, one may
write the Einstein eld equation in the following manner: G

= 8Gc
4
(T

+
T
(vac)

), in which T
(vac)

= g

(c
4
/(8G)) may be interpreted as the stress-
energy tensor associated with the vacuum, and in the orthonormal reference
frame is given by
T
(vac)

= diag
_
c
4
/(8G), c
4
/(8G), c
4
/(8G), c
4
/(8G)
_
. (14)
We see it is thus possible to adopt the viewpoint that the cosmological term
is an integral part of the stress-energy tensor, being considered as a uid.
Accordingly, we can dene the total stress-energy tensor, T

, as
T

= T

+ T
(vac)

(15)
such that G

= 8Gc
4
T

. Thus, the components of the total stress-
energy tensor of the wormhole, (r), (r) and p(r), are given by the following
12
functions
(r) = (r) +
c
2
8G
, (16)
(r) = (r) +
c
4
8G
, (17)
p(r) = p(r)
c
4
8G
. (18)
This viewpoint may be interesting to adopt in some cases.
2.1.4 The Einstein equations
We are interested in matching the interior solution, whose metric is given
by Equation (1), to an exterior vacuum solution, which will be considered
below. Using Equation (5) and equating Equations (6)-(8) with (10)-(12) we
obtain the following set of equations
(r) =
c
2
8G
_
b

r
2

_
, (19)
(r) =
c
4
8G
_
b
r
3
2
_
1
b
r
_

r

_
, (20)
p(r) =
c
4
8G
__
1
b
r
_ _

+ (

)
2

rb
2r
2
(1b/r)

rb
2r
3
(1b/r)
+

r
_
+
_
. (21)
By taking the derivative of Equation (20) with respect to the radial co-
ordinate r, and eliminating b

and

, given in Equation (19) and Equation

(21), respectively, we obtain the following equation

= (c
2
)

2
r
(p + ) . (22)
Equation (22) is the relativistic Euler equation, or the hydrostatic equation
for equilibrium for the material threading the wormhole, and can also be
obtained using the conservation of the stress-energy tensor, T

;
= 0, putting
= r. The conservation of the stress-energy tensor, in turn can be deduced
from the Bianchi identities, which are equivalent to G

;
= 0.
2.1.5 Method for solving the Einstein equations
The conventional approach to solving the Einstein equations would be to
assume a specic and plausible type of matter or elds for the source of the
13
stress-energy tensor. One would then derive equations of state for the radial
tension and the tangential pressure, as functions of the energy density. These
equations of state, together with the three eld equations would provide the
geometry of the spacetime given in terms of the metric, g

, as we would
have ve equations for ve unknown functions, i.e., b(r), (r), , and
p. Morris and Thornes approach [1], which will be followed in this paper,
diers as they rst xed a convenient geometry for a wormhole solution and
then derived the matter distribution for the respective solution (see [47] for
a careful analysis of the various approaches).
2.2 Construction of a wormhole with generic . I:
General comments
2.2.1 The mathematics of embedding
We can use embedding diagrams to represent a wormhole and extract some
useful information for the choice of the shape function, b(r), which will be
used in the specic solutions considered below. Due to the spherically sym-
metric nature of the problem, one may consider an equatorial slice, = /2,
without loss of generality. The respective line element, considering a xed
moment of time, t = const, is given by
ds
2
=
dr
2
1 b(r)/r
+ r
2
d
2
. (23)
To visualize this slice, one embeds this metric into three-dimensional Eu-
clidean space, whose metric can be written in cylindrical coordinates, (r, , z),
as
ds
2
= dz
2
+ dr
2
+ r
2
d
2
. (24)
Now, in the three-dimensional Euclidean space the embedded surface has
equation z = z(r), and thus the metric of the surface can be written as,
ds
2
=
_
_
1 +
_
dz
dr
_
2
_
_
dr
2
+ r
2
d
2
. (25)
Comparing Equation (25) with (23) we have the equation for the embedding
surface, given by
dz
dr
=
_
r
b(r)
1
_
1/2
. (26)
14
To be a solution of a wormhole, the geometry has a minimum radius, r =
b(r) = r
o
, denoted as the throat, at which the embedded surface is vertical,
i.e., dz/dr , see Figure 2. Outside the wormhole, far from the mouth,
space can be asymptotically at, de Sitter, or anti-de Sitter.
z(r)
r/r

r
0
0
1
2 3 4 5 6
1
2
3
(a)
r
o
a
a
(b)
Figure 2 - (a) The embedding diagram of a two-dimensional section (t =constant,
= /2) of the wormhole in three-dimensional Euclidean space, here a/r
o
= 6;
(b) For the full visualization of the surface sweep through a 2 rotation around
the zaxis. Figure 1 with one dimension less is equivalent to this gure.
One can dene the proper radial distance for the upper part of the worm-
hole z > 0 as
l(r) =
_
r
ro
dr
[1 b(r)/r]
1/2
, (27)
and for the lower part z < 0 as
l(r) =
_
r
ro
dr
[1 b(r)/r]
1/2
. (28)
15
The maximum upper limit of integration is a, the radius of the wormhole
mouth. The shape function b(r) should be positive and such that b/r < 1
in order to have
_
r/b 1 real. For generic cosmological constant this may
not be possible. For instance, in vacuum for suciently large radii Equation
(26) becomes imaginary, in which case the embedding process is no longer
valid. However, the importance of the embedding is near the throat where a
special condition, the are out condition, should be obeyed.
2.2.2 Exotic matter
Following [1] closely we will see that the wormhole needs exotic matter, i.e,
matter that does not obey the null energy condition, and thus does not obey
the weak or any other energy condition. The null energy condition applied
to the matter considered in (16)-(17) is c
2
> 0. Thus a good way to
dene exoticity is through the parameter dened as [1] =
c
2
| c
2
|
. This
parameter is dimensionless, and when positive signals exotic matter. Using
equations (19)-(20) one nds
=
c
2
| c
2
|
=
b/r b

2r(1 b/r)

|b

r
2
|
. (29)
To be a solution of a wormhole, one needs to impose that the throat ares
out, as in Figure 2. Mathematically, this aring-out condition entails that
the inverse of the embedding function r(z), must satisfy d
2
r/dz
2
> 0 near
the throat r
o
. Dierentiating dr/dz = (r/b(r) 1)
1/2
with respect to z, we
have
d
2
r
dz
2
=
b b

r
2b
2
> 0 . (30)
Combining Equation (29) with Equation (30), the exoticity function takes
the form
=
2b
2
r|b

r
2
|
d
2
r
dz
2
2r
_
1
b
r
_

|b

r
2
|
. (31)
Considering the nite character of , and therefore of b

, and the fact that

(1 b/r)

0 at the throat, we have the following relationship

(r
o
) =

0

0
c
2
|
0
c
2
|
> 0 . (32)
Thus matter at the throat is exotic (see [1, 2, 26] for a detailed discussion).
16
2.3 Construction of a wormhole with generic . II:
Interior and exterior solutions, and junction con-
ditions
We will distinguish the interior cosmological constant,
int
, from the exterior
one,
ext
. Equations (19)-(21) demonstrate that once the geometry is xed
by the redshift function (r), and the shape function b(r), the inclusion of
the cosmological constant
int
will shift the respective values of (r), (r)
and p(r) and might help in minimizing the violation of the energy conditions.
2.3.1 Interior solution of the Einstein equations with generic
int
To nd an interior solution of the Einstein equations with generic , we
combine the analysis developed in the previous sections, taking into account
the notation that
int
represents the cosmological constant associated with
the interior solution. The respective Einstein equations provide the following
relationships
(r) =
c
2
8G
_
b

r
2

int
_
, (33)
(r) =
c
4
8G
_
b
r
3
2
_
1
b
r
_

r

int
_
, (34)
p(r) =
c
4
8G
__
1
b
r
_ _

+ (

)
2

rb
2r
2
(1b/r)

rb
2r
3
(1b/r)
+

r
_
+
int
_
.(35)
The metric quantities should carry a subscript
int
, but we do not put it in
order to not overload the notation. It is of interest to nd an expression for
the radial tension at the throat. From equation (34) one nds that at the
throat (b(r
o
) = r
o
) the tension is
(r
o
) =
c
4
8G
_
1
r
2
o

int
_
. (36)
Thus the radial tension at the throat is positive for wormholes whose struc-
ture yields
int
<
1
r
2
o
, this includes wormholes with negative and zero cos-
mological constant. The radial tension is negative, i.e., it is a pressure, for
wormholes with the cosmological constant obeying
int
>
1
r
2
o
. The total radial
tension, (r
o
) = (r
o
) +
c
4
8G

int
, is always positive, of course.
17
2.3.2 Exterior vacuum solution of the Einstein equations with
generic
ext
The spacetime geometry for a vacuum exterior region is simply determined
considering a null stress-energy tensor, T

= 0, i.e., (r) = (r) = p(r) =
0. Note that
ext
represents the cosmological constant associated with the
exterior solution. In the most general case, the exterior radial coordinate r,
should be dierent from the interior one r. Here we put them equal, both
are denoted by r, since it simplies the junction and it gives interestingly
enough results. The Einstein equations then reduce to
0 =
b

r
2

ext
, (37)
0 =
b
r
3
2
_
1
b
r
_

r

ext
, (38)
0 =
_
1
b
r
_ _

+ (

)
2

rb
2r
2
(1b/r)

rb
2r
3
(1b/r)
+

r
_
+
ext
. (39)
The metric quantities should carry a subscript
ext
, but again we have not put
it as to not overload the notation. Solving the system of dierential equations
of Equations (37)-(39), the exterior vacuum solution with a cosmological
constant is given by
ds
2
=
_
1
2GM
c
2
r

ext
3
r
2
_
c
2
dt
2
+
dr
2
(1
2GM
c
2
r

ext
3
r
2
)
+r
2
(d
2
+ sin
2
d
2
) . (40)
This metric is the unique solution to the vacuum Einstein equations for
a static and spherically symmetric spacetime with a generic cosmological
constant. The denomination given to it depends on the sign of
ext
. The
Schwarzschild solution, which is a particular case, is obtained by setting

ext
= 0. In the presence of a positive cosmological constant,
ext
> 0, the
solution is designated by the Schwarzschild-de Sitter metric. For
ext
< 0,
we have the Schwarzschild-anti de Sitter metric. For = 0, note that this
metric is not asymptotically at as r , it is either asymptotically de
Sitter (
ext
> 0), or asymptotically anti-de Sitter (
ext
< 0). However,
if
ext
is extremely small, there is a range of the radial coordinate, i.e.,
1/

ext
r GM/c
2
, for which the metric is nearly at. For values of
r below this range, the eect of the mass M dominates, whereas for values
above this range, the eect of the cosmological term dominates, as for very
18
large values of r the large-scale curvature of the spacetime must be taken
into account.
(i) The Schwarzschild spacetime,
ext
= 0
Equation (40) with
ext
= 0 is the Schwarzschild solution. The full
vacuum solution represents a black hole in a asymptotically at spacetime.
The factor f(r) = (1
2GM
c
2
r
) is zero at
r
b
=
2 GM
c
2
, (41)
the black hole event horizon. Since the wormhole matter will ll the region
up to a radius a larger than r
b
this radius does not enter into the problem.
It is important to have it in mind, since if after construction one nds that
r
b
> a than the object constructed is a black hole rather than a wormhole.
(ii) The Schwarzschild-de Sitter spacetime,
ext
> 0
Equation (40) with
ext
> 0 represents a black hole in asymptotically de
Sitter space. If 0 < 9
ext
(GMc
2
)
2
< 1, the factor f(r) = (1
2GM
c
2
r

ext
3
r
2
)
is zero at two positive values of r, corresponding to two real positive roots.
Dening
A =
_
3c
4
8
ext
G
2
M
2
_
1/3
3

_
1 +

1
c
4
9
ext
G
2
M
2
, (42)
B =
_
3c
4
8
ext
G
2
M
2
_
1/3
3

_
1

1
c
4
9
ext
G
2
M
2
, (43)
the solutions are given by
r
b
=
2GM
c
2
_

A + B
2

A B
2

3
_
, (44)
r
c
=
2GM
c
2
(A + B) . (45)
When
ext
(GM/c
2
)
2
1 (see appendix A for details), one gets
r
b
=
2GM
c
2
_
1 +
4
3

ext
_
GM
c
2
_
2
_
, (46)
r
c
=

ext
_
_
1
GM
c
2

ext
3
_
_
. (47)
19
The smaller of the values, denoted by r = r
b
, can be considered as the event
horizon of the vacuum black hole solution, but since the wormhole matter
will ll the region up to a radius a superior than r
b
this radius does not enter
into the problem. The larger value, denoted by r = r
c
, can be regarded as
the position of the cosmological event horizon of the de Sitter spacetime.
Keeping
ext
constant, but increasing M, r = r
b
will increase and r = r
c
will decrease. If 9
ext
(GMc
2
)
2
= 1, both horizons coincide and are situated
at r = r
b
= r
c
= 3GM/c
2
. Thus we will consider 9
ext
(GMc
2
)
2
< 1.
Particular cases are,
ext
= 0 yielding the Schwarzschild solution, and M = 0
yielding the de Sitter solution. When r the metric tends to the de Sitter
spacetime
ds
2
=
_
1

ext
3
r
2
_
c
2
dt
2
+
dr
2
_
1
ext
3
r
2
_
+ r
2
(d
2
+ sin
2
d
2
) . (48)
For
ext
0, the de Sitter metric tends to the Minkowskian spacetime. In
the coordinates adopted above, the metric of the de Sitter spacetime will be
singular if r = (3/
ext
)
1/2
, but this is a mere coordinate singularity signaling
the presence of a cosmological event horizon.
(iii) The Schwarzschild-anti de Sitter spacetime,
ext
< 0
For the Schwarzschild-anti de Sitter metric, with
ext
< 0, the equation
f(r) = (1
2GM
c
2
r
+
|ext|
3
r
2
) = 0 will have only one real root, therefore implying
the existence of one horizon. If one denes
A =
_
3c
4
8|
ext
|G
2
M
2
_
1/3
3

_
1 +

_
1 +
c
4
9|
ext
|G
2
M
2
, (49)
B =
_
3c
4
8|
ext
|G
2
M
2
_
1/3
3

_
1

_
1 +
c
4
9|
ext
|G
2
M
2
, (50)
then the solution is (see appendix for details),
r
b
=
2GM
c
2
(A + B) . (51)
For |
ext
|(GM/c
2
)
2
1 one obtains
r
b
=
2GM
c
2
_
1
4
3
|
ext
|
_
GM
c
2
_
2
_
. (52)
20
Once again this event horizon is avoided by lling the space with exotic
matter from the throat at r
o
up to the mouth at a, where a > r
b
in order
that the wormhole is not a black hole. If
ext
= 0, the metric is reduced to
the Schwarzschild solution. If r the metric tends to the anti-de Sitter
solution
ds
2
=
_
1 +
|
ext
|
3
r
2
_
c
2
dt
2
+
dr
2
_
1 +
|ext|
3
r
2
_
+ r
2
(d
2
+ sin
2
d
2
) . (53)
For |
ext
| 0, the anti-de Sitter metric tends to the Minkowskian spacetime.
2.3.3 Junction conditions in wormholes with generic
ext
To match the interior to the exterior one needs to apply the junction condi-
tions that follow from the theory of general relativity. One of the conditions
imposes the continuity of the metric components, g

, across a surface, S,
i.e., g
(int)
|
S
= g
(ext)
|
S
. This condition is not sucient to join dierent
spacetimes. One formalism of matching, that leads to no errors in the calcu-
lation, uses the extrinsic curvature of S (or second fundamental form of the
surface S, the rst fundamental form being the metric on S) see, e.g, [82].
However, for spacetimes with a good deal of symmetry, such as spherical
symmetry, one can use directly the eld equations to make the match, see
e.g., [83] (see also Taub [84]). We follow this latter approach. Indeed, due
to the high symmetries of the solution, we can use the Einstein equations,
Equations (19)-(21), to determine the energy density and stresses of the sur-
face S necessary to have a match between the interior and exterior. If there
are no surface stress-energy terms at the surface S, the junction is called
a boundary surface. If, on the other hand, surface stress-energy terms are
present, the junction is called a thin shell.
(i) Matching of the metric
As was mentioned above the unique vacuum static and spherically sym-
metric solution, in the presence of a non-vanishing cosmological constant, is
given by Equation (40). A wormhole with nite dimensions, in which the
matter distribution extends from the throat, r = r
o
, to a nite distance
r = a, obeys the condition that the metric is continuous. Due to the spher-
ical symmetry the components g

and g

are already continuous, and so

21
one is left with imposing the continuity of g
tt
and g
rr
,
g
tt(int)
= g
tt(ext)
, (54)
g
rr(int)
= g
rr(ext)
, (55)
at r = a, with g
tt(int)
and g
rr(int)
being the metric components for the in-
terior region at r = a, and g
tt(ext)
and g
rr(ext)
the exterior metric com-
ponents for the vacuum solution at r = a. For the sake of consistency
in the notation this could have been done in the orthonormal frame with
the hat quantities g

, but in this case it is more direct to do with coordi-
nate frame quantities. One can start the analysis by considering two gen-
eral solutions of Equation (1), an interior solution and an exterior solution
matched at a surface, S. The continuity of the metric then gives generi-
cally
int
(a) =
ext
(a) and b
int
(a) = b
ext
(a). If one now uses, Equations (1),
(40), (54) and (55), one nds then e
2(a)
= [1 2GM/(c
2
a)
ext
a
2
/3] and
[1 b(a)/a] = [1 2GM/(c
2
a)
ext
a
2
/3] which can be simplied to
(a) =
1
2
ln
_
1
2GM
c
2
a


ext
3
a
2
_
, (56)
b(a) =
2GM
c
2
+

ext
3
a
3
. (57)
From Equation (57), one deduces that the mass of the wormhole is given by
M =
c
2
2G
_
b(a)

ext
3
a
3
_
. (58)
(ii) Matching of the equations I: The surface pressure
We are going to consider the case where static interior observers mea-
sure zero tidal forces, i.e.,
int
=constant,

int
= 0. This is, of course, a
particular choice which simplies the analysis. As we have seen, the met-
ric is continuous through the surface S. However, their rst and second
derivatives might not be. Since the metric is static and spherically sym-
metric the only derivatives that one needs to worry about are radial. Now,
second derivatives in the metric are related to the Einstein tensor G

, or
since we are working with hat quantities, to G

. But G

is proportional
to the stress-energy tensor T

. Thus, something in the stress-energy tensor
22
has to reect this discontinuity. Indeed, at the boundary S, T

is propor-
tional to a Dirac function, and we can write T

= t

( r a), where
r =

g
rr
r means the proper distance through the thin shell. To nd t

one then uses
_
+

G

d r = (8 G/c
4
)
_
+

t

( r a) d r, where
_
+

means an
innitesimal integral through the shell. Using the property of the function
(f(x)) = (1/|f

(x)|) (x), and

_
+

g(x)(x x
0
) = g(x
0
), one nds
t

=
c
4
8 G
_
+

G

d r . (59)
Since the shell is innitesimally thin in the radial direction there is no ra-
dial pressure, thus we are left with a surface energy term , and a surface
tangential pressure P.
First we calculate the surface energy density . From Equation (6) we
see that G

t
only depends on rst derivatives of the metric, so that when
integrated through the shell it will give metric functions only, that by deni-
tion are continuous. Thus, since the integral gives the value of the metric on
the exterior side (b
+
, say) minus the value of the metric on the interior side
(b

), it gives zero, and one nds

= 0 . (60)
Now we nd the surface tangential pressure P. From Equation (8) we
see that G

has an important term (1

b
r
)

. The other terms depend at

most on the rst derivative and as before do no contribute to the integral.
Thus, in this case Equation (59) gives 8 G/c
2
P =
_
1 b(a)/a
+

. Now,

= 0 by assumption, and
+
= [GM/(c
2
a
2
)
ext
a/3] / (1 b(a)/a).
Thus, P =
c
4
8 Ga
GM/(c
2
a
2
)ext a/3

1b(a)/a
, or more explicitly,
P =
c
4
8Ga
GM
c
2
a

ext
3
a
2
_
1
2GM
c
2
a

ext
3
a
2
. (61)
P is always positive for the Schwarzschild and the Schwarzschild-anti de
Sitter spacetime, i.e.,
ext
0. The Schwarzschild-de Sitter spacetime,

ext
> 0, has to be analyzed more carefully. In Figure 3, we plot in a
graph 9
ext
(GM/c
2
)
2
2GM/(c
2
a) the regions where P is negative, zero
or positive. For high
ext
(GMc
2
)
2
(either
ext
big or M big) and at low
23
2GM/(c
2
a) (either M small or a big), one needs a surface tension to sup-
port the structure. In the other case one needs a surface pressure. This is
expected in the sense that for a positive
ext
one has an expanding external
de Sitter spacetime. If a is big (and so 2GM/(c
2
a) small), the wormhole
boundary is participating somehow in the expansion, so one needs a tension
to hold it. For small a (
2GM
c
2
a
big), the gravity wins over the expansion and
so one needs a pressure to hold against collapse, a particular case being the
Schwarzschild case
ext
= 0.
0.2 0.4 0.6 0.8 1
0.2
0.4
0.6
0.8
1

+ + + + +
+ + + +
+ + +
+ +
+
+
+
+ +
+
+
+
+
+ +
+ +
+ +
+
+
+
+
+
+
+
+
2M/a
M
2

ext
9
Figure 3 - The regions where P is negative and positive in a plot of 9
ext
(GM/c
2
)
2
as a function of the inverse of the relative size of the wormhole, i.e., 2GM/c
2
a (in
the gure we have used geometrical units G = 1 = c) are given. Inside the solid line
is the region of solutions. To the left of the dashed line P is a negative pressure,
i.e., a tension, to the right P is a positive quantity, i.e., is a pressure, see text for
more details.
Now we detail the procedure to plot the graph. Dene = 2GM/(c
2
a)
and = 9
ext
(GMc
2
)
2
, so that f = 1
2GM
c
2
a

ext
3
a
2
can be written as
1
4
27
2
. The zero of f, dening the corresponding horizon for the set
of parameters, gives as a function of . Varying from 0 to 1 one gets
the solid curve in Figure 3. The inside of this curve denes the region where
P takes values (outside the curve there are no wormholes, on the top left
hand part one mouth of the possible wormhole is causally disconnected from
the other forever, and on the top right hand part the wormhole has been
24
converted into a black hole ). We can check for the sign. To do this dene
g =

2

4
27
2
, see Equation (61). g = 0 denes the curve =
27
8

3
, which is
plotted in the graph as a dashed line. To the left of this dashed curve, P is
negative, and to the right it is positive.
One can have a term of comparison for the surface tangential pressure P
at the thin shell. Assuming that the thin shell has a width of approximately
r, one can consider a volumetric tangential pressure, orthogonal to the
radial coordinate, given by

P =
P
r
. (62)
Taking into account Equation (35), we see that the tangential pressure at
the mouth, with

int
= 0, is given by p(a) = (c
4
/16 Ga
3
) (b(a) b

(a)a).
Estimates of

P may be given in terms of p(a), by dening the following ratio

P
p(a)
= 2
a
r
1
b
a
b

GM
c
2
a

ext
3
a
2
_
1
2 GM
c
2

ext
3
a
3
. (63)
It is also interesting to nd the ratio to p(r
o
), the maximum pressure, given
by

P
p(r
o
)
= 2
r
2
o
a r
1
1 b

(r
o
)
GM
c
2
a

ext
3
a
2
_
1
2GM
c
2
a

ext
3
a
2
. (64)
One may nd numerical estimates, considering various choices of the shape
function, b(r), which will be done while considering specic solutions of
traversable wormholes.
(iii) Matching of the equations II: The radial pressure
To construct specic solutions of wormholes with generic cosmological
constant, one needs to know how the radial tension behaves across the junc-
tion boundary, S. The analysis is simplied if we consider two general solu-
tions of Equation (1), an interior solution and an exterior solution matched
at a surface, S. The radial component of the Einstein equations, Equation
(20), provides
b
int
r
3
=
8G
c
4

int
(r) +
int
+ 2
_
1
b
int
r
_

int
r
, (65)
b
ext
r
3
=
8G
c
4

ext
(r) +
ext
+ 2
_
1
b
ext
r
_

ext
r
. (66)
25
Taking into account the continuity of the metric at the junction boundary
one has obtained
int
(a) =
ext
(a) and b
int
(a) = b
ext
(a). For simplicity, we
are considering

int
(a) = 0. Using again the following relation,

ext
(r) =
_
GM
c
2
r
2

ext
3
r
_
/
_
1
2GM
c
2
r

ext
3
r
2
_
, and taking into account Equation (61),
we verify that Equations (65)-(66), provide us with an equation which governs
the behavior of the radial tension at the boundary, namely,

int
(a) +
c
4
8G

int
=
ext
(a) +
c
4
8G

ext
+
2
a
Pe
(a)
, (67)
where we have put e
(a)
=
_
1 2GM/(c
2
a)
ext
a
2
/3. Equation (67), al-
though not new in its most generic form [2], is a beautiful equation that
relates the radial tension at the surface with the tangential pressure of the
thin shell. A particularly interesting case is when P = 0. In this situ-
ation M =
ext
c
2
a
3
/(3G), and

ext
(a) = 0. Since by our construction

int
(a) = 0,

is continuous across the surface and the solution is reduced to

a boundary surface. From Equation (57) one nds that the shape function
at the junction is given by b(a) =
ext
a
3
. Thus, Equation (67) simplies
to,
int
(a) +
c
4
8G

int
=
ext
(a) +
c
4
8G

ext
. If one considers a matching of an
interior solution of a wormhole with generic , given by Equation (1), to an
exterior Schwarzschild solution, with
ext
= 0 and
ext
= 0, we simply have
the condition that
int
(a)+
c
4
8G

int
= 0 at the boundary surface. Matching an
interior solution to an exterior Schwarzschild-de Sitter or Schwarzschild-anti
de Sitter solution, with
ext
= 0 and
ext
= 0, we have the following relation-
ship,
int
(a) +
c
4
8G

int
=
c
4
8G

ext
at the boundary surface. These solutions
will be analyzed in greater detail in the next sections.
2.4 Spacetime diagrams
We now draw the spacetime diagrams, i.e., the Carter-Penrose diagrams,
corresponding to wormholes in spacetimes with
ext
= 0,
ext
> 0, and
ext
<
0. They are easy to sketch once one knows the corresponding diagrams for the
solution with no wormhole, i.e, the Carter-Penrose diagrams for Minkowski
spacetime, de Sitter spacetime and anti-de Sitter spacetime [8], respectively.
Each point in the diagram represents a sphere.
26
future null infinity
future null infinity
past null infinity past null infinity
past timelike infinity
future timelike infinity
a a 0
r
Figure 4 - The spacetime diagram for the wormhole with
ext
= 0, i.e, a worm-
hole in an asymptotically Minkowski spacetime, represented by two copies of the
Minkowski diagram joined at the throat.
past null infinity
future null infinity
c
o
s
m
o
l
o
g
i
c
a
l

h
o
r
i
z
o
n
c
o
s
m
o
l
o
g
i
c
a
l

h
o
r
i
z
o
n
past null infinity
future null infinity
a
a 0
0
r
r
a a 0
r
Figure 5 - The spacetime diagram for the wormhole with
ext
> 0, i.e, a wormhole
in an asymptotically de Sitter spacetime, with an innite number of copies (only
two are represented).
spacelike infinity
future timelike infinity future timelike infinity
past timelike infinity past timelike infinity
spacelike infinity a a
0
r
Figure 6 - The spacetime diagram for the wormhole with
ext
< 0, i.e, a wormhole
in an asymptotically anti-de Sitter spacetime, represented by two copies of the
anti-de Sitter diagram joined at the throat.
27
Since the wormhole creates an extra asymptotic region one has to duplicate
the original diagram through the throat. In Figures 4, 5 and 6 the diagrams
for a wormhole in an asymptotically at spacetime
ext
= 0, in an asymptot-
ically de Sitter spacetime
ext
> 0, and in an asymptotically anti-de Sitter
spacetime
ext
< 0, respectively, are drawn. Note the duplication of the
asymptotic regions.
3 Specic construction of wormholes with
generic
We will give some examples of traversable wormholes similar to those con-
structed in [1]. The dierence from the wormholes in that work is that the
wormholes here in general have an innitesimal thin shell with a tangential
pressure P = 0, and the exterior spacetime has a cosmological constant. We
discuss rst the case
ext
= 0, then
ext
> 0, and nally we mention briey
the case
ext
< 0. In all cases we put

= 0 in the interior region.

3.1 Specic solutions of traversable wormholes with

ext
= 0 (asymptotically at wormholes)
3.1.1 Matching to an exterior Schwarzschild solution, with P = 0
Here we consider a matching of an interior solution with an exterior Schwarzschild
solution (
ext
= 0 and
ext
= 0), and with the junction having zero tangential
pressure, P = 0. From Equation (67) one has at the junction

int
(a) +
c
4
8G

int
= 0 . (68)
Then, from Equation (34) (with

= 0) one gets
0 =
c
4
8G
b(a)
a
3
. (69)
Since b = 0, Equation (69) is only satised if a . This is one of the
cases considered by Morris and Thorne [1], in which the wormholes material
extends from the throat all the way to innity.
28
3.1.2 Matching to an exterior Schwarzschild solution, with P = 0
Matching the interior solution to an exterior Schwarzschild solution (
ext
= 0
and
ext
= 0) but considering P = 0, provides some interesting results.
The behavior of the radial tension at the junction is given by Equation (67)
and taking into account Equation (57) one nds that the shape function
at the junction simply reduces to b(a) = 2GM/c
2
. We will next consider
various choices for the shape function, b(r), which will give dierent wormhole
solutions.
(i) b(r) = (r
o
r)
1/2
Consider the following functions
(r) =
0
, (70)
b(r) = (r
o
r)
1/2
, (71)
where r
o
is the throat radius as before. Using the Einstein equations, Equa-
tions (33)-(35) we have
(r) (r) +
c
2
8G

int
=
c
2
8G
r
1/2
o
2r
5/2
, (72)
(r) (r) +
c
2
8G

int
=
c
4
8G
r
1/2
o
r
5/2
, (73)
p(r) p(r) +
c
2
8G

int
=
c
4
8G
r
1/2
o
4r
5/2
. (74)
The distribution of the material threading the wormhole is plotted in Figure
7. To nd an estimate of the surface pressure at the thin shell, one has

P
p(ro)
=
4r
2
o
GM
rc
2
a
2
_
1
2GM
c
2
a
_
1/2
(see Equation (64)).
From b(a) = 2GM/c
2
and b(a) = (r
o
a)
1/2
, one nds that the matching
occurs at
a =
(2GM/c
2
)
2
r
o
. (75)
Now in order that the wormhole is not a black hole one has to impose
a > 2GM/c
2
. Then, from Equation (75) one nds r
o
< 2GM/c
2
. From
Equation (75), we also extract the mass of the wormhole, given by M =
c
2
(r
o
a)
1/2
/(2G).
29
1 3 5 7 9 1 3 3 5 5 7 9 7 9 1
P
~
0.2
0.4
0.6
0.8
1
r
/
r
o
0.2
0.2
0.4 0.4
0.6 0.6
0.8 0.8
1 1
r
/ r
o
r
/
r
o

o
/

o
p
/
p

/
p
o
Figure 7 - Distribution of the material threading the wormhole for the case b(r) =
(r
o
r)
1/2
. The mouth of the wormhole is at a = 9 r
o
. The averaged surface pressure

P on the thin shell dened in the text is depicted qualitatively as

P/p(r
o
) in the
p/p(r
o
) graph as a point.
The interior metric, r
o
r a, is determined recalling that e
2(a)
=
_
1
_
r
o
/a
_
. It is given by
ds
2
=
_
1
_
r
o
a
_
c
2
dt
2
+
dr
2
_
1
_
ro
r
_
+ r
2
(d
2
+ sin
2
d
2
) . (76)
The exterior metric, a r < , is given by
ds
2
=
_
1

r
o
a
r
_
c
2
dt
2
+
dr
2
_
1

roa
r
_
+ r
2
(d
2
+ sin
2
d
2
) . (77)
The nal metric of the whole spacetime is given by Equations (76)-(77),
which are joined smoothly, as we have carefully worked out.
It is also interesting to briey consider the traversability conditions that
the absurdly advanced civilization might require to cross the wormhole from
one mouth to the other and back (see the Appendix B for details). One
nds that for an observer traversing the wormhole with a velocity v = 0.01 c,
the wormhole has a throat radius given by r
o
500 km. One can choose
r
o
= 500 km, and assume that the traversal time done by the spaceship is
approximately one year. Then, one nds that the matter distribution extends
from r
o
to a = 4.74 10
13
m 5 10
3
light years, with a being the size
of the wormhole. It is supposed that the space stations are parked there.
One also nds that the wormhole mass is M = 3.3 10
36
kg, six orders of
magnitude superior to the Suns mass. One may also nd an estimate for
Equation (64), giving

P/p(r
o
) 10
4
.
30
One may choose other parameters, for instance, so that the wormhole
mass is of the same order of the Suns mass. Considering a traversal with
a velocity v = 5.4 10
3
m/s, we may choose that the wormhole throat is
given by r
o
= 9 10
2
m. If we consider an extremely fast trip, where the
traversal time is given by
traveler
= 3.7 s, the matter distribution extends
from r
o
to a = 10
4
m. In this case the mass of the wormhole is given by
M 2 10
30
kg, which is the Suns mass. From Equation (64) an estimate
to

P is

P/p(r
o
) 5.7 10
3
. We have an extremely large surface pressure.
As the wormhole mass is decreased, one sees that a larger tangential surface
pressure is needed to support the structure.
(ii) b(r) = r
2
o
/r
Consider now,
(r) =
0
, (78)
b(r) = r
2
o
/r . (79)
Using the Einstein equations, Equations (33)-(35), one has
(r) (r) +
c
2
8G

int
=
c
2
8G
r
2
o
r
4
, (80)
(r) (r) +
c
2
8G

int
=
c
4
8G
r
2
o
r
4
, (81)
p(r) p(r) +
c
2
8G

int
=
c
4
8G
r
2
o
r
4
. (82)
The distribution of the material threading the wormhole is plotted in Fig-
ure 8. A qualitative estimate for P/p(r
o
), in this case given by

P
p(r
0
)
=
4r
2
0
GM
rc
2
a
2
_
1
2GM
c
2
a
_
1/2
(see Equation (64)) is also plotted.
From b(a) = 2GM/c
2
and b(r) = r
2
o
/r, the matching occurs at
a =
c
2
r
2
o
2GM
. (83)
Imposing a > 2GM/c
2
, so that the wormhole does not correspond to a black
hole solution, we have r
o
> 2GM/c
2
and the mass is given by M =
c
2
r
2
o
2Ga
.
The interior metric, r
o
r a, with e
2(a)
= [1 (r
o
/a)
2
], is given by
ds
2
=
_
1
_
r
o
a
_
2
_
c
2
dt
2
+
dr
2
1
r
2
o
r
2
+ r
2
(d
2
+ sin
2
d
2
) . (84)
31
The exterior metric, a r < , is given by
ds
2
=
_
1
r
2
o
ar
_
c
2
dt
2
+
dr
2
_
1
r
2
o
ar
_
+ r
2
(d
2
+ sin
2
d
2
) . (85)
The respective nal metric solution of the spacetime is given by Equations
(84)-(85), joined smoothly at a.
| |
0.8
0.6
0.2
1 3 5 1 3 3 5 5 1
P
~
0.4
1
0
r
/
r
o
0.2
0.2
0.4 0.4
0.6 0.6
0.8 0.8
1 1
r
/ r
o
r
/
r
o

o
/

o
p
/
p
/

2 4 2 2 4 4
p
o
/
Figure 8 - Distribution of the material threading the wormhole for the case b(r) =
r
2
o
/r. The mouth of the wormhole is at a = 5 r
o
.

P/p(r
o
) is depicted qualitatively
in the p/p(r
o
) graph as a point.
A comment on the interior metric (84) is in order. In fact, this inte-
rior solution is the same as the one found by Ellis [18]. Indeed, consid-
ering the following coordinate transformations,

t =
_
1 (r
o
/a)
2
_
1/2
t and
l = (r
2
r
2
o
)
1/2
, one nds that the metric is reduced to ds
2
= c
2
d

t
2
+
dl
2
+ (r
2
o
+ l
2
) (d
2
+ sin
2
d
2
) , where l is the proper radial coordinate,
ranging from to +. The properties are commented in [1] and [22].
Harris showed further that it is a solution of the Einstein equations with a
stress-energy tensor of a peculiar massless scalar eld [23].
With respect to the traversability conditions given in the Appendix B,
considering that the traversal velocity is v = 0.01 c, the wormhole throat
is r
o
= 10
6
m. If the traversal time is one year,
traveler
= 3.16 10
7
s,
the junction is at a = 4.74 10
13
m 5 10
3
light years. The mass of
the wormhole is M 1.4 10
25
kg, approximately an order of magnitude
superior to the Earths mass. An estimate for Equation (64), is given by

P/p(r
o
) 4.6 10
16
.
32
3.2 Specic solutions of traversable wormholes with

ext
> 0 (asymptotically de Sitter wormholes)
3.2.1 Matching to an exterior Schwarzschild-de Sitter solution,
with P = 0
In this section we will be interested in a matching of an interior solution
with an exterior Schwarzschild-de Sitter solution,
ext
= 0 and
ext
> 0, at a
boundary surface, P = 0. We verify from Equation (67) that the following
condition holds

int
(a) +
c
4
8G

int
=
c
4
8G

ext
, (86)
at the surface boundary. Considering Equation (34), we have
b(a) =
ext
a
3
. (87)
Substituting this value in Equation (58), one obtains the mass of the worm-
hole, given by,
M =
c
2
3G

ext
a
3
. (88)
We shall next consider identical shape functions as in the above section.
(i) b(r) = (r
o
r)
1/2
Consider the following functions
(r) =
0
, (89)
b(r) = (r
o
r)
1/2
. (90)
From b(a) =
ext
a
3
and b(a) = (r
o
a)
1/2
, one sees that the matching occurs
at
a
r
1/5
o

2/5
ext
. (91)
The mass can then be expressed as M = c
2
(r
o
a)
1/2
/(3G). It can be shown
that the interior solution, r
o
r a, is identical to Equation (76), i.e.,
ds
2
=
_
1
_
r
o
a
_
c
2
dt
2
+
dr
2
_
1
_
ro
r
_
+ r
2
(d
2
+ sin
2
d
2
) . (92)
33
The exterior solution, a r < , is given by the following metric
ds
2
=
_
1
2(roa)
1/2
3r

r
1/2
o
r
2
3a
5/2
_
c
2
dt
2
+
dr
2
_
1
2(roa)
1/2
3r

r
1/2
o
r
2
3a
5/2
_
+r
2
(d
2
+ sin
2
d
2
) . (93)
The spacetime of the nal solution is given by the metrics, Equations (92)-
(93), which have been smoothly joined at a.
The additional parameter now is the cosmological constant,
ext
, given
by
ext
= (r
o
/a
5
)
1/2
. For instance, consider a traversal velocity v = 0.01 c,
so that r
o
= 5 10
5
m. If the observer traverses through the wormhole
comfortably during a year,
traveler
3.16 10
7
s, and a = 4.74 10
13
m.
The mass of the wormhole is M 2.210
36
kg and the cosmological constant
has the value
ext
= 4.610
32
m
2
. The cosmological event horizon is then
situated at r
c
= 8.1 10
15
m 200 a.
(ii) b(r) = r
2
o
/r
Consider now the functions
(r) =
0
, (94)
b(r) = r
2
o
/r . (95)
From b(a) =
ext
a
3
and b(a) = r
2
o
/a one sees that the matching occurs at
a = r
1/2
o
/
1/4
ext
. The mass can be written as M = c
2
r
2
o
/(3Ga). The exterior
cosmological constant is then
ext
= (r
o
/a
2
)
2
.
Considering that the traversal velocity is v = 0.01 c, the throat radius is
r
o
= 10
6
m. If the traversal time is done in one year the matching occurs
at a = 4.74 10
13
m. The mass is then given by M = 9.5 10
24
kg, the
same order of magnitude as the Earths mass, and the cosmological constant
is
ext
= 1.98 10
43
m
2
. The cosmological horizon is situated at r
c
=
3.9 10
21
m 10
8
a.
Another example is provided by a traveler with velocity is v = 0.0001 c.
The throat radius is then r
o
= 10
4
m. Assuming that the traversal time is
done in
traveler
= 6.310
3
years, one has a 310
15
m 0.32 light years.
The mass is given by M = 1.5 10
19
kg, the mass of an asteroid, and the
cosmological constant has a value of
ext
= 10
54
m
2
, approximately the
present value.
34
3.2.2 Matching to an exterior Schwarzschild-de Sitter solution,
with P = 0
One can also match the interior solution with an exterior Schwarzschild-de
Sitter solution (
ext
= 0 and
ext
> 0) in the presence of a thin shell, P = 0.
From Equation (67), we have the behavior of the radial tension at the thin
shell, given by

int
(a) +
c
4
8G

int
=
c
4
8G

ext
+
2
a
P e
(a)
. (96)
The shape function at the junction is given by Equation (57). From Equation
(58), one veries that the mass of the wormhole is zero when b(a) =
ext
a
3
/3,
is positive when b(a) >
ext
a
3
/3, and is negative when b(a) <
ext
a
3
/3. One
can perform a similar analysis as done for the previous examples.
3.3 Specic solutions of traversable wormholes with

ext
< 0 (asymptotically anti-de Sitter wormholes)
3.3.1 Matching to an exterior Schwarzschild-anti de Sitter solu-
tion, with P = 0
From Equation (67), matching an interior solution with an exterior given
by the Schwarzschild-anti de Sitter solution (
ext
= 0 and
ext
< 0), at a
boundary surface P = 0, yields,

int
(a) +
c
4
8G

int
=
c
4
8G
|
ext
| , (97)
at the surface boundary. Considering Equation (34), we have
b(a) = |
ext
| a
3
. (98)
From Equation (26), we concluded that the shape function has to be positive
to guarantee that the factor
_
r/b 1 be real. Therefore, for the anti-de
Sitter exterior, i.e.,
ext
< 0, with P = 0 there is no solution. This problem
may be overcome by considering a matching to an exterior anti-de Sitter
solution with a thin shell, i.e., P = 0.
35
3.3.2 Matching to an exterior Schwarzschild-anti de Sitter solu-
tion, with P = 0
From Equation (57), one nds that b(a) is positive if
2GM
c
2

|
ext
|
3
a
3
. (99)
Then one can construct easily wormholes in anti-de Sitter spacetime, and
again perform a similar analysis as done for the previous examples.
4 Conclusions
We have considered Morris -Thorne wormholes, i.e., static and spherically
symmetric traversable wormholes, in the presence of a non-vanishing cos-
mological constant. Matching the interior solution with a vacuum exterior
solution, we have deduced an equation for the tangential surface pressure,
and another one which governs the behavior of the radial tension at the
boundary.
Specic solutions with various choices of the shape function were pre-
sented. Through the traversability conditions, we have obtained estimates
for the matching boundary, a, the mass of the wormhole M, and the tangen-
tial surface pressure P, by imposing values for the traversal velocity and the
traversal time.
Acknowledgements JPSL thanks Joseph Katz and Donald Lynden-Bell
for teaching many years ago how to do junctions in an easy way, thanks
Madalena Pizarro for alerting that a civilization that constructs a wormhole
is an absurdly advanced civilization, rather than arbitrarily advanced, and
thanks Observatorio Nacional do Rio de Janeiro for hospitality. FSNL thanks
many conversations with Paulo Crawford do Nascimento. The present ad-
dress of SQO is Instituto de Fsica, Universidade Federal do Rio de Janeiro,
CEP 21945-970, Rio de Janeiro. This work was partially funded through
project PESO/PRO/2000/4014 by Funda cao para a Ciencia e Tecnologia
(FCT) Portugal.
36
A Solutions of (1
2 GM
c
2
r

ext
3
r
2
) = 0
The roots of the cubic equation
x
3
+ ax + b = 0 , (100)
can be found. Indeed, dening
A =
3

b
2
+

b
2
4
+
a
3
27
, (101)
B =
3

b
2

b
2
4
+
a
3
27
, (102)
the solutions are given by
x
1
= A + B , (103)
x
2
=
A + B
2
+
A B
2

3 , (104)
x
3
=
A + B
2

AB
2

3 . (105)
(i) The Schwarzschild spacetime,
ext
= 0:
This case is trivial and is analyzed directly in the text.
(ii) The Schwarzschild-de Sitter spacetime,
ext
> 0:
The equation f(r) = (1
2GM
c
2
r

ext
3
r
2
) = 0 may be recast into the form
x
3
x + = 0 , (106)
with x = r/(2GMc
2
) and = 3c
4
/(4
ext
G
2
M
2
). Thus comparing (106)
with (100) one nds
A =
3

2
+

2
4


3
27
, (107)
B =
3

2
4


3
27
. (108)
37
To nd which two of the three solutions one must pick up, we consider the
case
ext
(GM/c
2
)
2
1, i.e., 1. The square root
_

2
4


3
27
may be
expanded as i

3/2

27
_
1
27
8
_
. Thus Equations (107)-(108) yield the following
approximation
A =
3

i

1/2

3
_
1 + i

27
6
1

1/2
_
, (109)
B =
3

i

1/2

3
_
1 i

27
6
1

1/2
_
. (110)
Dening =

1/2

3
and =

27
6
1

1/2
, and considering
3

i = e
i

6
=

3
2
+
i
2
,
3

i = e
i

6
=

3
2

i
2
, Equations (109)-(110) take the form
A =
__
3
2


2
_
+ i
_
1
2
+

3
2

__
, (111)
B =
__
3
2


2
_
i
_
1
2
+

3
2

__
. (112)
We are only interested in the positive solutions. Thus, substituting Equa-
tions (111)-(112) into Equations (103)-(105) and taking into account x =
r/(2GMc
2
), we have that x
3
and x
1
give, respectively
r
b
=
2GM
c
2
_
1 +
4
3

ext
_
GM
c
2
_
2
_
, (113)
r
c
=

ext
_
_
1
GM
c
2

ext
3
_
_
. (114)
We see that r
b
and r
c
correspond to the vacuum black hole horizon and to
the cosmological event horizon, respectively. In Equation (113) to nd the
numerical factor of the rst order term
ext
_
GM
c
2
_
2
it is easier to linearize
the solution by writing x = 1+, for small , and then with the help of (106)
one nds .
(iii) The Schwarzschild-anti de Sitter spacetime,
ext
< 0:
In the Schwarzschild-anti de Sitter spacetime, one nds that the equation
f(r) = (1
2GM
c
2
r
+
|ext|
3
r
2
) = 0 may be recast into the form
x
3
+ x = 0 , (115)
38
with x = r/(2GMc
2
) and = 3c
4
/(4|
ext
| G
2
M
2
), |
ext
|
ext
. Thus
comparing (115) with (100) one nds
A =
3

2
+

2
4
+

3
27
, (116)
B =
3

2
4
+

3
27
. (117)
The only real solution is x
1
= A + B. Indeed, for |
ext
| (GM/c
2
)
2
1, i.e.,
1 one nds
r
b
=
2GM
c
2
_
1
4
3
|
ext
|
_
GM
c
2
_
2
_
, (118)
which gives the black hole horizon. In Equation (118) to nd the numerical
factor of the rst order term
ext
_
GM
c
2
_
2
it is easier to linearize the solution
by writing x = 1 +, for small , and then with the help of (115) one nds .
B Traversability conditions
We will be interested in specic solutions for traversable wormholes and
assume that a traveler of an absurdly advanced civilization, with human
traits, begins the trip in a space station in the lower universe, at proper
distance l = l
1
, and ends up in the upper universe, at l = l
2
. We shall,
for self-containment and self-consistency, briey describe the traversability
conditions given in [1].
The cosmological constant does not enter the analysis directly. Indeed, al-
though the interior cosmological constant,
int
, can be incorporated into the
eective quantities, for
ext
= 0 the external parameters such as the mass of
the wormhole change. Thus, the cosmological constant enters the traversabil-
ity conditions indirectly. Ford and Roman [12] with their quantum inequal-
ities have imposed severe restrictions on the T

obeyed by matter, and in

particular found that the parameters of the Morris -Thorne - Yurtsever [63]
wormhole obeyed these inequalities. However, we now know that the energy
conditions can be even classically violated [14], and thus the Ford-Roman
inequalities have somehow lost their strength in this context. Therefore, our
39
parameters are not chosen in order to satisfy the quantum inequalities, but
rather follow the spirit of the Morris and Thorne work [1] where attention is
paid to the traversability conditions of a human being.
(i) Tidal acceleration felt by a traveler
The tidal accelerations between two parts of the travelers body, separated
by, say, 2 meters, should not exceed the gravitational acceleration at Earths
surface g
Earth
, (with g
Earth
10m/s
2
). From [1] and using our simplifying
assumption

= 0, one obtains the following inequality

2
2r
2
_
v
c
_
2
_
b

b
r
_

g
Earth
2c
2
1m

1
10
16
m
2
, (119)
with = (1 v
2
/c
2
)
1/2
, and v being the travelers velocity. This inequality
refers to tangential tidal accelerations. Radial tidal acceleration are zero for

= 0. From Equation (119) one sees that stationary observers with v = 0

measure null tidal forces.
(ii) Acceleration felt by a traveler
Another condition that needs to be respected is that the local gravita-
tional acceleration, |a| at the space stations should not surpass the terrestrial
gravitational acceleration, g
Earth
. The condition imposed in [1] (for

= 0)
is
|a| =

_
1
b
r
_
1/2

c
2

g
Earth
. (120)
For v =constant travelers, one has |a| = 0, of course!
(iii) Total time in a traversal
The trip should take a relatively short time, for instance one year, as
measured by the traveler and for observers that stay at rest at the space
stations, l = l
1
and l = l
2
, i.e.,

traveler
=
_
+l
2
l
1
dl
v
1 year, (121)
t
space station
=
_
+l
2
l
1
dl
ve

1 year, (122)
40
respectively.
Having set these conditions for general wormholes, we will now study two
particular cases.
(1) The wormhole with shape function b(r) = (r
o
r)
1/2
(i) Tidal acceleration felt by a traveler
In the interior region, r
o
< r < a, the tidal acceleration as measured by
a traveler moving radially through the wormhole is given by Equation (119).
For non-relativistic velocities, v c, we have 1, and substituting the
expression of b(r), i.e., b(r) = (r
o
r)
1/2
, in Equation (119), one can impose
a velocity for the traveler traversing through the wormhole, such that the
tidal accelerations felt are inferior to the terrestrial gravitational acceleration,
g
Earth
. With these conditions, one obtains a restriction for the velocity at the
throat, where the acceleration is severest, given by
_
v
c
_

2r
o
10
8
m
. (123)
If the observer traverses the wormhole with a non-relativistic velocity v c,
the accelerations at the beginning and end of the trip are negligible. If
we consider that the velocity is v 0.01 c, one can nd estimates for the
dimensions of the wormhole. The wormhole throat obeys r
o
500 km. For
deniteness consider the choice r
o
= 500 km. Taking into account Equation
(75) for asymptotically at spacetimes, the region of matter distribution will
extend to
a =
(2GM/c
2
)
2
(500km)
. (124)
Then, by choosing a, we nd the value of the wormhole mass. We now take
the steps to choose a. Similar procedures follow for
ext
= 0, see main text.
(ii) Acceleration felt by a traveler
Another condition that needs to be respected is that the local gravita-
tional acceleration |a|, at the space stations should not surpass the terres-
trial gravitational acceleration, g
Earth
. Considering non-relativistic velocities,
v c, and v constant, one has |a| 0, so that 1. Condition (120) is
immediately satised, the traveler feels a zero gravitational acceleration.
41
(iii) Total time in a traversal
The expressions for the total times in a traversal of the wormhole, mea-
sured by the traveler and observers at rest at the station, are given by Equa-
tions (121)-(122). l = l
1
and l = l
2
are the positions of the space stations.
For low velocities, v c, we have 1, and with =
0
, Equations
(121)-(122) reduce to

traveler

2l
v
= e

0
t
space station
(125)
Suppose that the space stations are placed in the neighborhood of the mouth,
at a, in the exterior side. It is convenient to place the space stations at large
enough radii, i.e., a r
o
, so that the factor 1 b(r)/r 1. Thus, one also
has that e
2(a)
= 1 (r
o
/a)
1/2
1, so that
traveler
t
space station
2a/v.
Assume that the traversal time should be less than a year, i.e.
traveler

1 year (1 year = 3.16 10
7
s). Therefore, from 2a/v 1 year, one extracts a
value for a, namely a 4.74 10
13
m.
(2) The wormhole with shape function b(r) = r
2
o
/r
Applying a similar analysis as before, Equation (119), with v c and
1, and considering the form function given by b(r) = r
2
o
/r, we have
_
v
c
_

_
r
2
r
o
_
1
10
8
m
(126)
Imposing a traversal velocity v = 0.01 c at the throat, r = r
o
, one nds
r
o
10
6
m. For =
0
the local gravitational accelerations at the space
stations are zero. The traversal time, with 1 is given by

traveler

2l
v
= e

0
t
space station
. (127)
By a similar analysis as above, we place the space stations at a r
o
, implying
that e
2(a)
= 1 (r
o
/a)
2
1, so that
traveler
t
space station
. Considering
that the traversal time is approximately one year,
traveler
= 3.16 10
7
s,
and taking into account that the traversal velocity is v = 0.01 c, the junction
surface is at a = 4.74 10
13
m.
42
References
[1] M. S. Morris and K. S. Thorne, Wormholes in spacetime and their use
for interstellar travel: A tool for teaching General Relativity, Am. J.
Phys. 56, 395 (1988).
[2] M. Visser, Lorentzian Wormholes: From Einstein to Hawking (American
Institute of Physics, New York, 1995).
[3] J. A. Wheeler, Geometrodynamics, (Academic Press, New York, 1962).
[4] S. W. Hawking, Wormholes in spacetime, Phys. Rev. D37, 904 (1988).
[5] R. P. Geroch, Topology in General Relativity, J. Math. Phys. 8, 782
(1967).
[6] B. K. Harrison, K. S. Thorne, M. Wakano and J. A. Wheeler, Gravita-
tional Theory and Gravitational Collapse, (University of Chicago Press,
Chicago, 1965).
[7] Ya. B. Zeldovich and I. D. Novikov, Relativistic Astrophysics, Vol. I:
Stars and Relativity, (University of Chicago Press, Chicago, 1971).
[8] S. W. Hawking and G.F.R. Ellis, The Large Scale Structure of Spacetime,
(Cambridge University Press, Cambridge 1973).
[9] G. Klinkhammer, Averaged energy conditions for free scalar elds in
at spacetime, Phys. Rev. D 43, 2542 (1991).
[10] T. A. Roman, Quantum stress-energy tensors and the weak energy
condition, Phys. Rev. D 33, 3526 (1986).
[11] L. H. Ford and T. A. Roman, Averaged energy conditions and quantum
inequalities, Phys. Rev. D 51, 4277 (1995).
[12] L. H. Ford and T. A. Roman, Quantum eld theory constrains
traversable wormhole geometries, Phys. Rev. D 53, 5496 (1996).
[13] L. H. Ford and T. A. Roman, The quantum interest conjecture, Phys.
Rev. D 60, 104018 (1999).
43
[14] C. Barcel o and M. Visser, Scalar elds, energy conditions and
traversable wormholes, Class. Quantum Grav. 17, 3843 (2000).
[15] B. McInnes, The Covariant Entropy Bound, Brane Cosmology, and the
Null Energy Condition, JHEP 0212, 053 (2002)
[16] D. Hochberg, A. Popov and S. Sushkov, Self-consistent wormhole so-
lutions of semiclassical gravity, Phys. Rev. Lett. 78, 2050 (1997).
[17] C. Barcelo and M. Visser, Traversable wormholes from massless con-
formally coupled scalar elds, Phys. Lett. B466, 127 (1999).
[18] H. G. Ellis, Ether ow through a drainhole: A particle model in general
relativity, J. Math. Phys. 14, 104 (1973).
[19] K. A. Bronnikov, Scalar-tensor theory and scalar charge, Acta Phys.
Pol. B 4, 251 (1973).
[20] T. Kodama, General-relativistic nonlinear eld: A kink solution in a
generalized geometry, Phys. Rev. D 18, 3529 (1978).
[21] G. Clement, Einstein-Yang-Mills-Higgs solitons, Gen. Rel. Grav. 13,
763 (1981).
[22] G. Clement, The Ellis geometry, Am. J. Phys. 57, 967 (1989).
[23] E. G. Harris, Wormhole connecting two Reissner-Nordstr om uni-
verses, Am. J. Phys. 61, 1140 (1993).
[24] M. Visser, Traversable wormholes: Some simple examples, Phys. Rev.
D 39, 3182 (1989).
[25] M. Visser, Traversable wormholes: The Roman ring, Phys. Rev. D
55, 5212 (1997).
[26] M. Visser and D. Hochberg, Generic wormhole throats, gr-qc/971001.
[27] N. Dadhich, S. Kar, S. Mukherjee and M. Visser, R = 0 spacetimes
and self-dual Lorentzian wormholes, Phys. Rev. D 65, 064004 (2002).
[28] V. P. Frolov and I. D. Novikov, Wormhole as a device for studying a
black hole interior, Phys. Rev. D 48, 1607 (1993).
44
[29] P. F. Gonzalez-Das, Ringholes and closed timelike curves, Phys. Rev.
D 54, 6122 (1996).
[30] G. Clement, Flat wormholes from cosmic strings, J. Math. Phys. 38,
5807 (1997).
[31] R. O. Aros and N. Zamorano, Wormhole at the core of an innite
cosmic string, Phys. Rev. D 56, 6607 (1997).
[32] F. Shein, P. C. Aichelburg and W. Israel, String-supported wormhole
spacetimes containing closed timelike curves, Phys. Rev. D 54, 3800
(1996).
[33] E. Teo, Rotating traversable wormholes, Phys. Rev. D 58, 024014
(1998).
[34] S. Nojiri, O. Obregon, S. D. Odintsov and K. E. Osetrin, Can primor-
dial wormholes be induced by GUTs at the early Universe?, Phys. Lett.
B458, 19 (1999).
[35] A. Saa, Nonexistence theorems for traversable wormholes, Mod. Phys.
Lett. A 14, 843 (1999).
[36] S. Krasnikov, Traversable wormhole, Phys. Rev. D 62, 084028 (2002).
[37] S. A. Hayward Wormholes supported by pure ghost radiation, Phys.
Rev. D 65, 124016 (2002).
[38] L.

A. Gergely, Wormholes, naked singularities, and universes of ghost
radiation, Phys. Rev. D 65, 127502 (2002).
[39] K. Bronnikov and S. Grinyok, Charged wormholes with non-minimally
coupled scalar elds. Existence and stability, gr-qc/0205131.
[40] P. K. Kuhttig, A wormhole with a special shape function, Am. J.
Phys. 67, 125 (1999).
[41] D. Hochberg and M. Visser, Null energy condition in dynamic worm-
holes, Phys. Rev. Lett. 81, 746 (1998).
45
[42] D. Hochberg and M. Visser, Dynamic wormholes, antitrapped surfaces,
and energy conditions, Phys. Rev. D 58, 044021 (1998).
[43] M. Visser, S. Kar and N. Dadhich, Traversable wormholes with arbi-
trarily small energy condition violations, Phys. Rev. Lett. 90, 201102
(2003).
[44] S. Kim, Schwarzschild-de Sitter type wormhole, Phys. Lett. A 166,
13 (1992).
[45] T. A. Roman, Inating Lorentzian wormholes, Phys. Rev. D 47, 1370
(1993).
[46] M. S. R. Delgaty and R. B. Mann, Traversable wormholes in (2+1) and
(3+1) dimensions with a cosmological constant, Int. J. Mod. Phys. D
4, 231 (1995).
[47] A. DeBenedictis and A. Das, On a general class of wormholes, Class.
Quantum Grav. 18, 1187 (2001).
[48] A. Chodos and S. Detweiler, Spherical symmetric solutions in ve-
dimensional general relativity, Gen. Rel. Grav. 14, 879 (1982).
[49] G. Clement, A class of wormhole solutions to higher-dimensional gen-
eral relativity, Gen. Rel. Grav. 16, 131 (1984).
[50] A. DeBenedictis and A. Das, Higher dimensional wormhole geometries
with compact dimensions, gr-qc/0207077.
[51] K. K. Nandi, B. Bhattacharjee, S. M. K. Alam and J. Evans, Brans-
Dicke wormholes in the Jordan and Einstein frames, Phys. Rev. D 57,
823 (1998).
[52] Y.-G. Shen, H.-Y. Guo, Z.-Q. Tan and H.-G. Ding, Wormholes in
Kaluza-Klein theory, Phys. Rev. D 44, 1330 (1991).
[53] S. Kar, Evolving Lorentzian wormholes, Phys. Rev. D 53, 722 (1996).
[54] L. A. Anchordoqui and S. E. Perez Bergliaa, Wormhole surgery and
cosmology on the brane: The world is not enough, Phys. Rev. D 62,
076502 (2000).
46
[55] C. Barcelo and M. Visser, Brane surgery: energy conditions, traversable
wormholes, and voids, Nucl. Phys. B584, 415 (2000).
[56] H. Koyama, S. A. Hayward and S. Kim, Construction and enlargement
of dilatonic wormholes by impulsive radiation, gr-qc/0212106.
[57] M. Visser, Traversable wormholes from surgically modied
Schwarzschild spacetimes, Nucl. Phys. B328, 203 (1989).
[58] E. Poisson and M. Visser, Thin-shell wormholes: Linearization stabil-
ity, Phys. Rev. D 52, 7318 (1995).
[59] M. Ishak and K. Lake, Stability of transparent spherically symmetric
thin shells and wormholes, Phys. Rev. D 65, 044011 (2002).
[60] C. Armendariz-Picon, On a class of stable, traversable Lorentzian
wormholes in classical general relativity. Phys. Rev. D 65, 104010
(2002).
[61] H. Shinkai and S. A. Hayward, Fate of the rst traversable wormhole:
Black hole collapse or inationary expansion, Phys. Rev. D 66, 044005
(2002).
[62] K. Bronnikov and S. Grinyok, Instability of wormholes with a non-
minimally coupled scalar eld, gr-qc/0201083.
[63] M. S. Morris, K. S. Thorne and U. Yurtsever, Wormholes, Time Ma-
chines and the Weak Energy Condition, Phy. Rev. Lett. 61, 1446
(1988).
[64] V. P. Frolov and I. D. Novikov, Physical eects in wormholes and time
machines, Phys. Rev. D 42, 1057 (1990).
[65] S. Kim and K. S. Thorne, Do vacuum uctuations prevent the creation
of closed timelike curves? Phys. Rev. D 43, 3929 (1991).
[66] S. W. Hawking, Chronology protection conjecture, Phys. Rev. D 46,
603 (1992).
[67] M. Lyutikov, Vacuum polarization at the chronology horizon of the
Roman spacetime, Phys. Rev. D 49, 4041 (1994).
47
[68] W. A. Hiscock and D. A. Konlowski, Quantum vacuum energy in Taub-
NUT-type cosmologies, Phys. Rev. D 26, 1225 (1982).
[69] L.-X. Li and J. R. Gott, Self-consistent vacuum for Misner space and
the chronology protection conjecture, Phys. Rev. Lett. 80, 2980 (1998).
[70] W. A. Hiscock, Quantized elds and chronology protection, gr-
qc/0009061.
[71] M. Visser, The quantum physics of chronology protection, gr-
qc/0204022.
[72] K. S. Thorne, Closed timelike curves, in General Relativity and Grav-
itation, Proceedings of the 13th Conference on General Relativity and
Gravitation, edited by R. J. Gleiser et al (Institute of Physics Publish-
ing, Bristol, 1993), p. 295.
[73] S. A. Hayward, Black holes and traversible wormholes: a synthesis,
gr-qc/0203051.
[74] J. G. Cramer, R. L. Forward, M. S. Morris, M. Visser, G. Benford and
G. A. Landis, Natural wormholes as gravitational lenses, Phys. Rev.
D 51, 3117 (1995).
[75] D. F. Torres, G. E. Romero and L. A. Anchordoqui, Wormholes,
gamma ray bursts and the amount of negative mass in the Universe,
Mod. Phys. Lett. A 13, 1575 (1998).
[76] M. Safonova, D. F. Torres and G. E. Romero, Microlensing by natural
wormholes: theory and simulations, Phys. Rev. D 65, 023001 (2002).
[77] M. Safonova, D. F. Torres, G. E. Romero, Macrolensing signatures of
large-scale violations of the weak energy condition, Mod. Phys. Lett.
A 16, 153 (2001).
[78] J. P. S. Lemos, Three-dimensional black holes and cylindrical General
Relativity, Phys. Lett. B352, 46 (1995).
[79] J. P. S. Lemos and V. T. Zanchin, Charged Rotating Black Strings and
Three Dimensional Black Holes, Phys. Rev. D 54, 3840 (1996).
48
[80] J. P. S. Lemos, Rotating toroidal black holes in anti-de Sitter space-
times and their properties, in Astronomy and Astrophysics, Recent De-
velopments, Proceedings of the Xth Portuguese Meeting on Astronomy
and Astrophysics, edited by J. P. S. Lemos et al (World Scientic, 2001),
p. 88, gr-qc/001092.
[81] C. W. Misner, K. S. Thorne and J. A. Wheeler, Gravitation (W. H.
Freeman and Company, New York, 1973).
[82] W. Israel, Singular hypersurfaces and thin shells in general relativity,
Nuovo Cimento B 44, 1 (1966); Nuovo Cimento B 49, 463 (1967).
[83] A. Papapetrou and A. Hamoui, Couches simple de mati`ere en relativite
generale, Ann. Inst. Henri Poincare A 9, 179 (1968).
[84] A. Taub, Spacetimes with distribution valued curvature tensors, J.
Math. Phys. 21, 1423 (1980).
49