gr-qc/yymmnnn
Is dark matter an extra-dimensional effect?
M. E. Kahil∗
Mathematics Department, Modern Sciences and Arts University, Cairo, Egypt
T. Harko†
arXiv:0809.1915v2 [gr-qc] 26 Mar 2009
Department of Physics and Center for Theoretical and Computational Physics,
The University of Hong Kong, Pok Fu Lam Road, Hong Kong
(Dated: March 26, 2009)
We investigate the possibility that the observed behavior of test particles outside galaxies, which is
usually explained by assuming the presence of dark matter, is the result of the dynamical evolution
of particles in higher dimensional space-times. Hence, dark matter may be a direct consequence
of the presence of an extra force, generated by the presence of extra-dimensions, which modifies
the dynamic law of motion, but does not change the intrinsic properties of the particles, like, for
example, the mass (inertia). We discuss in some detail several possible particular forms for the
extra force, and the acceleration law of the particles is derived. Therefore, the constancy of the
galactic rotation curves may be considered as an empirical evidence for the existence of the extra
dimensions.
PACS numbers: 04.50.+h, 04.20.Jb, 04.20.Cv, 95.35.+d
I.
INTRODUCTION
There is a large amount of observational evidence showing that the standard gravitational theories cannot describe
correctly the large scale dynamics of massive astrophysical systems. The observed rotational curves of spiral galaxies
cannot be explained by applying Newtonian or general relativistic mechanics to the visible matter in galaxies and
clusters. Neutral hydrogen clouds are observed at large distances from the galactic center, much beyond the extent
of the luminous matter. A large number of independent observations have shown that the rotational velocities vtg (r)
of these clouds tend toward constant and slightly rising values of the order of vtg∞ ∼ 200 − 300 km/s as a function of
the distance r from the center of the galaxy [1, 2, 3]. This is in sharp contrast to the Newtonian inverse square force
law, which implies a decline in velocity. For clouds moving in circular orbits with velocity vtg (r), the balance between
2
the centrifugal acceleration vtg
/r and the gravitational attraction force GM (r)/r2 allows to express the mass-distance
2
relation M (r) in the form M (r) = rvtg
/G. In the constant rotation velocity region this leads to a mass profile
2
M (r) = rvtg∞ /G. Consequently, the mass within a distance r from the center of the galaxy increases linearly with r,
even at large distances where very little luminous matter can be detected.
This behavior of the galactic rotation curves is usually explained by postulating the existence of some dark (invisible)
matter, distributed in a spherical halo around the galaxies. The dark matter is assumed to be a cold, pressureless
medium. There are many possible candidates for dark matter, the most popular ones being the weakly interacting
massive particles (WIMP) (for a review of the particle physics aspects of dark matter see [4]). Their interaction cross
section with normal baryonic matter, while extremely small, are expected to be non-zero and we may expect to detect
them directly. It has also been suggested that the dark matter in the Universe might be composed of superheavy
particles, with mass ≥ 1010 GeV. But observational results show the dark matter can be composed of superheavy
particles only if these interact weakly with normal matter, or if their mass is above 1015 GeV [5].
Another interesting candidate for the dark matter is the dilatonic dark matter - the fundamental scalar field which
exists in all existing unified field theories. The cosmological implications of the dilatonic dark matter have been
explored in [6], where a higher-dimensional generalization of the standard big-bang cosmology has been proposed. It
was shown that the missing-mass problem as well as the horizon problem, and the flatness problem of the standard
model can be resolved within the context of this unified cosmology. The possibility that the dilaton plays the role
of the dark matter of the universe was investigated in [7]. The condition for the dilaton to be the dark matter
strongly restricts its mass to be around 0.5 keV or 270 MeV. For the other mass ranges, the dilaton contradicts the
cosmological observations. The 0.5 keV dilaton has a free-streaming distance of about 1.4 Mpc and is an excellent
∗ Electronic
† Electronic
address: kahil@aucegypt.edu
address: harko@hkucc.hku.hk
2
candidate for warm dark matter, while the 270 MeV one has a free-streaming distance of about 7.4 pc and is a
candidate for cold dark matter. An experiment to detect the relic dilaton using the electromagnetic resonant cavity,
based on the dilaton-photon conversion in strong electromagnetic background was proposed in [8]. The density of the
relic dilaton, as well as an estimate of the dilaton mass for which the dilaton becomes the dark matter of the universe
were calculated. The dilaton detection power in the resonant cavity were also obtained, and they were compared
with the axion detection power in similar resonant cavity experiment. Based on the fact that the scalar curvature
of the internal space determines the mass of the dilaton in higher-dimensional unified theories, the dilaton mass can
explain the origin of the mass, and resolve the hierarchy problem [9]. Moreover, cosmological observations put a
strong constraint on the dilaton mass, and requires that the scale of the internal space to be larger than 10−9 m.
Scalar fields or other long range coherent fields coupled to gravity have also intensively been used to model galactic
dark matter [10, 11, 12, 13, 14, 15, 16, 17, 18, 19].
However, despite more than 20 years of intense experimental and observational effort, up to now no non-gravitational
evidence for dark matter has ever been found: no direct evidence of it and no annihilation radiation from it. Moreover,
accelerator and reactor experiments do not support the physics (beyond the standard model) on which the dark matter
hypothesis is based.
Therefore, it seems that the possibility that Einstein’s (and the Newtonian) gravity breaks down at the scale
of galaxies cannot be excluded a priori. Several theoretical models, based on a modification of Newton’s law or of
general relativity, have been proposed to explain the behavior of the galactic rotation curves. A modified gravitational
potential of the form φ = −GM [1 + α exp (−r/r0 )] / (1 + α) r,with α = −0.9 and r0 ≈ 30 kpc can explain flat
rotational curves for most of the galaxies [20, 21].
In an other model, called MOND, and proposed by Milgrom [22, 23, 24, 25], the Poisson equation for the gravitational potential ∇2 φ = 4πGρ is replaced by an equation of the form ∇ [µ (x) (|∇φ| /a0 )] = 4πGρ, where a0 is a fixed
constant and µ (x) a function satisfying the conditions µ (x) = x for x << 1 and µ (x) = 1 for x >> 1. The force law,
giving the acceleration ~a of a test particle, becomes µ (a/a0 ) ~a = ~aN , where ~aN is the usual Newtonian acceleration.
√
a = aN for aN >> a0 and a = aN a0 for aN << a0 . The rotation curves of the galaxies are predicted to be flat,
and they can be calculated once the distribution of the baryonic matter is known. The value of the constant a0 is
1/2
given by a0 ≈ 1.2 × 10−8 cm/s−2 ≈ cH0 /6 ≈ c (Λ/3) /6, where H0 is Hubble’s constant and Λ is the cosmological
constant. MOND is a purely phenomenological theory, but still it can explain most of the galaxy rotation curves
without introducing dark matter. But despite its achievements, MOND has many problems of its own, like, for
example, the lack of conserved quantities, like energy, and a theoretical justifications to the MOND phenomenology.
A relativistic gravitation theory for MOND was proposed by Bekenstein [26]. In this model gravitation is mediated
by the tensor field gαβ , a scalar field φ and a vector field Uα , all three dynamical. For a simple choice of its free
function, the theory has a Newtonian limit for non-relativistic dynamics with significant acceleration, but a MOND
limit when accelerations are small. A tensor-vector-scalar theory that reconciles the galaxy scale success of modified
Newtonian dynamics with the cosmological scale evidence for cold dark matter (CDM) has been proposed by Sanders
[27]. The theory provides a cosmological basis for MOND by showing that the predicted phenomenology only arises
in a cosmological background.
Alternative theoretical models to explain the galactic rotation curves have been elaborated recently by Mannheim
[28, 29], Moffat and Sokolov [30], Brownstein and Moffat [31, 32] and Roberts [33]. The constancy of the tangential
velocity of test particles orbiting around galaxies can be also explained in the brane world models [34, 35], where the
effects of the projection of the Weyl tensor from the bulk plays the role of the dark matter [36, 37, 38, 39, 40, 41], in
the f (R) modified gravity models [42, 43, 44, 45], and by assuming that dark matter is in the form of a Bose-Einstein
condensate [46, 47, 48, 49, 50, 51, 52], or of an Einstein cluster [53, 54].
It is the purpose of the present paper to show that there is an alternative general physical interpretation of the ”dark
matter” paradigm. More exactly, the constancy of the galactic rotation curves can be obtained from the assumption
that the motion of the test particles in circular orbits around galaxies is non-geodesic. The galactic dynamics of test
particles is a direct consequence of the presence of an extra force f µ , which modifies the dynamic law of motion. Such
a scenario, in which the galactic rotation curves are explained by the presence of an extra force may be called EFDOD
(extra force dominated orbital dynamics).
One of the most interesting possibilities is that the extra force is due to the presence of the extra dimensions. In
such a model, which we may call multidimensional EFDOD, the motion of the particles takes place on geodesics in
higher dimensions. A comprehensive geometric treatment of Kaluza - Klein type unifications of non-Abelian gauge
theories with gravitation was first introduced in [55], where the appearance of a cosmological constant was also noted,
and further developed in [56, 57]. The possible modifications of Einstein’s theory of gravitation due to the fifth force
generated by the Kaluza - Klein dilaton were discussed in [58], including the effects on the gravitational redshift, the
deflection of light, the precession of perihelia, and the time-delay of radar echo around a spherically symmetric black
hole in multidimensional space times. The long-range effect of the higher-dimensional fifth force is characterized by
the dilatonic charge carried by the black hole even when it is neutral. In [59] it was emphasized that in the Brans-Dicke
3
theory it is the Pauli metric, not the Jordan metric, which describes the massless spin-two graviton. Similarly, in the
Jordan-Brans-Dicke theory, based on Kaluza-Klein unification, only the Pauli metric can correctly describe Einsteins
theory of gravitation. This necessitates a completely new reinterpretation of the Kaluza-Klein cosmology, as well
as of the Brans-Dicke theory. More significantly, this analysis shows that the Kaluza-Klein dilaton must generate
a fifth force, which could violate the equivalence principle. Recent torsion-balance experiments [60] have tested the
gravitational inverse-square law at separations between 9.53 mm and 55 µm, thus probing distances less than the
dark-energy length scale d = 85 µm. It has been found with 95% confidence that the inverse-square law holds down
to a length scale of around 56 µm, and that an extra dimension must have a size ≤ 44 µm.
However, it is known for some time that the effects of extra-dimensions on the trajectory of test particles as observed
in four dimensions can be modeled in terms of an extra force f µ , for both compactified and non-compactified spaces
[7, 61, 62]. The presence of such a force may explain the phenomenology and behavior of the galactic rotation curves.
We investigate in detail the dynamics of the test particles in extra-dimensional models, and we find the conditions
which must be satisfied by the five-dimensional metric tensor in order to explain the observed rotation curves. As a
physical test of our model we suggest that the lensing effects could be able to find evidences for the multi-dimensional
geometry.
We also investigate the acceleration law in EFDOD, and we find that it has a striking similarity with the acceleration
law in MOND. This leads us to the conclusion that the MOND theory, which can be considered, from a physical point
of view, as describing the non-geodesic motion of a test particle in a gravitational field under the action of an extra
force generated by the supplementary vector and scalar fields introduced in the model, is a particular case of the
EFDOD models.
The present paper is organized as follows. Physical models determining a non-geodesic motion of particles are
considered in Section II. The behavior of test particles in stable circular orbits in multidimensional models is considered
in Section III. In Section IV we consider the possibility that dark matter is an extra-dimensional effect, and we obtain
the general metric tensor in the flat rotation curve region and we check the consistency of the model. The acceleration
law and the relation of our model with MOND is discussed in Section V. We conclude and discuss our results in
Section VI. Throughout the paper we use the Landau-Lifshitz conventions [63] for signature and metric.
II.
SCALAR FIELD GENERATED EXTRA FORCE MODELS
There are several physical situations in which an extra force may be present, determining a non-geodesic motion of
the particles, like, for example, the case of a real scalar field minimally coupled to gravity and interacting with matter,
the case of the extra-force generated by the non-trivial coupling between matter and geometry in the f (R) modified
gravity models, and the cases of the extra forces generated by the presence of the compactified and non-compactified
higher dimensions of the space-time, respectively. In the present Section we will review briefly the first case.
In order to give a systematic treatment of the extra forces in the presence of a scalar field we will use the Bazanski approach for obtaining the geodesic equation [64]. According to this approach, the equation of motion in any
dimensions can be obtained by applying the action principle to the Lagrangian [65]
L = m(s)gAB uA
DΨB
+ fA ΨA , A, B = 0, 1, ..., D,
Ds
(1)
where ΨB is the deviation vector, uA is the tangent vector to the geodesic and fA and m(s) are functions which depend
C D
on the specific physical models. The covariant derivative DΨB /Ds is defined as DΨB /Ds = dΨB /ds + ΓB
CD Ψ u .
The equation describing a real scalar field ψ minimally coupled to gravity and interacting with matter can be given
as [66]
∂U
∂J
−
,
(2)
∂ψ
∂ψ
where U = U (ψ) is the self-interaction potential and J = J xβ is the source term of the scalar field. In the following
we neglect the effect of U (assumed to be of a breaking symmetry type). As for the source term J we assume that
it is of the general form J = 4πGg(ψ)Tµµ /c2 , where Tµµ is the trace of the energy momentum tensor of the matter
and g is a coupling function satisfying the conditions g (ψ∞ ) = 0 and ∂g (ψ∞ ) /∂ψ 6= 0, respectively, where ψ∞ is the
value of the scalar field at the minimum of the potential. In four dimensions the equation of motion of a test particle
in the presence of a scalar field can be derived from the Lagrangian
∇α ∇α ψ = −J −
L = m(s)gµν uµ
DΨν
+ m(s),µ Ψµ ,
Ds
(3)
4
and is given by
1
duµ
+ Γµαβ uα uβ =
ds
m(s)
g µσ m,σ −
dm µ
u .
ds
(4)
By assuming that the effective mass is of the form m ∼ exp (−g (ψ) ψ) we obtain [66]
duµ
d (g (ψ) ψ) µ
+ Γµαβ uα uβ =
u − ∂ µ (g (ψ) ψ) .
(5)
ds
ds
R
√
This equation of motion can also be derived from the variational principle δ mc gµν uµ uν ds = 0. The force
f~ has two components, one proportional to the velocity ~v of the particles, and which, being perpendicular to the
acceleration, does not give any contribution in Eq. (40) and therefore can be neglected, and a second component
given by f~k = −∇ [g (ψ) ψ]. By assuming that the scalar field has a spherical symmetry, ψ = ψ (r), evaluating the
force f~k around ψ = ψ∞ gives fk ≈ − [∂g (ψ∞ ) /∂ψ] ψ ′ ψ, where ′ denotes the derivative with respect to r. In order
to obtain concordance with MOND, which means a constant a0 , it is necessary that fk ∼ −1/r, which implies that
p
∂g (ψ∞ ) /∂ψ = g ′ (ψ∞ ) > 0 and ψ (r) = ψ0 ln (r/R0 ), with ψ0 , R0 = constant.
As for the scalar field we assume that is satisfies the equation [66]
∆ψ =
4πG ′
g (ψ∞ ) ψ∞ ρ,
c2
(6)
where ρ is the mass density of the matter fields other than ψ. With the obtained form of the scalar field it follows
that the density of the matter interacting with the scalar field has a density profile given by
ρ (r) =
ψ0
1 8 ln2 (r/R0 ) + 6 ln (r/R0 ) − 1
c2
.
16πG g ′ (ψ∞ ) ψ∞ r2
ln3/2 (r/R0 )
(7)
Therefore EFDOD may be due to a scalar field interacting with a matter distribution, whose density varies according to Eq. (7). The presence of such a field could explain the constancy of the galactic rotation curves and the
corresponding MOND phenomenology.
III.
MOTION OF TEST PARTICLES IN STABLE CIRCULAR ORBITS IN MULTIDIMENSIONAL
SPACE-TIMES
The above approach can naturally be implemented in Kaluza-Klein theory [6, 58]. In the following we will consider
the case of the extra forces generated by the presence of the compactified and non-compactified higher dimensions of
the space-time.
Let the coordinates of the five-dimensional manifold, with metric tensor γAB , be xA (A = 0, 1, 2, 3, 4). The 5D
interval is given by dS 2 = γAB dxA dxB . Usually it is assumed that the first four coordinates xµ are the coordinates
of the space-time xµ (µ = 0, 1, 2, 3), while x4 = ξ is the extra-dimension. Setting γµ4 = γ44 Aµ and γ44 = εΦ2 , where
Aµ and Φ are the vector and scalar potentials, respectively, and ε = ±1, we may write the line element without any
loss of generality as [61, 62]
2
dS 2 = gµν dxµ dxν + εΦ2 (dξ + Aµ dxµ ) ,
(8)
where gµν = γµν − εΦ2 Aµ Aν .
In the non-compact Kaluza-Klein theories, like, for example, the brane world models [34], all test particles travel
on five dimensional geodesics, but the observers, bounded to the usual four-dimensional space-time, have access
only to the 4D part of the trajectory. Mathematically, this means that the equations governing the motion in 4D
are projections of the 5D equations on the 4D-hypersurfaces orthogonal to some vector field ψ A . Generally, the
background metric in 5D can be written as [61, 62, 68]
dS 2 = γµν (xα , ξ) dxµ dxν + ǫΦ2 (xα , ξ) dξ 2 ,
(9)
where γµν is the induced metric in 4D. In brane world theory the physical space time four dimensional metric gµν is
generally identified with γµν . However, in some approaches, the physical metric in 4D is assumed to be conformally
related to the induced one,
dS 2 = Ω (ξ) gµν (xα , ξ) dxµ dxν + ǫΦ2 (xα , ξ) dξ 2 = Ω (ξ) ds2 + ǫΦ2 (xα , ξ) dξ 2 ,
(10)
5
where Ω (ξ) > 0 is called the warp factor [62].
In both compact and non-compact Kaluza-Klein theories, the motion of test particles takes place in higher dimensions (usually the number of dimensions of the space-time is assumed to be five), along the geodesics lines, and with
the equation of motion given by
duA
B C
+ Γ̂A
BC u u = 0.
dS
(11)
where uA = (dxµ /dS, dξ/dS) is the five-velocity and Γ̂A
BC are the Christoffel symbols formed with the 5D metric
[58, 61, 62, 65].
In order to obtain results which are relevant to the galactic dynamics, in the following we will restrict our study to
the static spherically-symmetric five-dimensional metric given by
dS 2 = eν(r,ξ) c2 dt2 − eλ(r,ξ) dr2 − r2 dθ2 + sin2 θdφ2 + ǫΦ2 (r, ξ) dξ 2 = ds2 + ǫΦ2 (r, ξ) dξ 2 ,
(12)
where the coordinates have been chosen so that xA = (ct, r, θ, φ, ξ). The components of the five-velocity U A are given
by U A = dxA /dS. In particular, U 4 = dξ/dS. In the following we also denote uA = dxA /ds, which represent the
components of the five-velocity with respect to the four-dimensional space-time with
q interval ds. The four-dimensional
interval ds is related to the five-dimensional interval dS by the relations dS = ds/ 1 − ǫΦ2 (r, ξ) (U 4 )2 or, equivalently,
q
q
dS = ds 1 + ǫΦ2 (r, ξ) (u4 )2 . The velocity uA is given as a function of U A by uA = U A / 1 − ǫΦ2 (r, ξ) (U 4 )2 .
The Lagrangian L of a massive test particle traveling in the five-dimensional space-time with metric given by Eq.
(12) is
"
2
2
2
2 #
2
dθ
dr
dξ
dφ
cdt
2
2
λ(r,ξ)
2
ν(r,ξ)
+ ǫΦ (r, ξ)
−e
−r
+ sin θ
.
(13)
2L = e
dS
dS
dS
dS
dS
Since the metric tensor coefficients do not explicitly depend on ct, θ and φ, the Lagrangian (13) gives the following
conserved quantities (generalized momenta) in five dimensions:
eν(r,ξ)
cdt
dθ
dφ
= E = const., r2
= Lθ = const., r2 sin2 θ
= Lφ = const.,
dS
dS
dS
(14)
where E is the total energy of the particle (in five-dimensions) and Lθ and Lφ are the components of the angular
moment, respectively. With the use of conserved quantities we obtain from Eq. (12) the geodesic equation for material
particles as
eν+λ
ds
dS
2
dr
ds
2
2
L2
= E2,
+ eν 1 + 2T − ǫΦ2 U 4
r
(15)
where we have denoted L2T = L2θ + L2φ / sin2 θ. Eq. (15) can be written as
eν+λ
ds
dS
2
dr
ds
2
+ Vef f (r, ξ) = E 2 ,
(16)
where
Vef f (r, ξ) = e
ν
L2T
4 2
2
,
1 + 2 − ǫΦ U
r
(17)
is the effective potential of the motion, which also contains the effects of the presence of the extra-dimension. If Φ ≡ 0,
we obtain the well-known four-dimensional expression.
For the case of the motion of particles in circular and stable orbits the generalized potential must satisfy the
following conditions: a) dr/ds = 0 (circular motion) b) ∂Vef f /∂r = 0 (extreme motion) and c) ∂ 2 Vef f /∂r 2 !extr > 0
(stable orbit), respectively. Conditions a) and b) immediately give the conserved quantities as
2
L2
E 2 = eν 1 + 2T − ǫΦ2 U 4
,
(18)
r
6
and
rν ′ e−ν
r ∂ h 2 4 2 i
L2T
,
ǫΦ U
= E2
−
2
r
2
2 ∂r
respectively. Eqs. (18) and (19) allow us to express the constants of the motion in the equivalent form
eν
1 ∂ h 2 2 4 2 i
2
E =
,
r ǫΦ U
1−
′
2r ∂r
1 − rν2
(19)
(20)
and
L2T =
1
r3 ν ′
′
2 1 − rν2
e−ν ∂ h ν 2 4 2 i
,
e ǫΦ U
1− ′
ν ∂r
(21)
respectively.
We define the tangential velocity vtg of a test particle in four dimensions, measured in terms of the proper time,
that is, by an observer located at the given point, as [63]
"
"
2
2 #
2
2 #
2
dθ
dθ
dS
dφ
dφ
2
2
2
−ν 2 2
−ν 2 2
vtg = e r c
=e r c
+ sin θ
+ sin θ
.
(22)
cdt
cdt
dS
dS
cdt
By using the constants of motion from Eqs. (14) we immediately obtain
2
vtg
L 2 eν
= T2 2 .
2
c
E r
(23)
By eliminating the constant quantity L2T /E 2 between Eqs. (20) and (21) gives
h
i
e−ν ∂
ν
2
4 2
2
e
ǫΦ
U
1
−
′
′
vtg
ν ∂r
νr
h
i .
=
c2
2 1 − 1 ∂ r2 ǫΦ2 (U 4 )2
(24)
2r ∂r
An alternative expression for the tangential velocity can be obtained directly from the line element Eq. (12), by
using the constants of motion. The result is
2
vtg
e−ν E 2 − 1 + ǫΦ2 U 4
=
c2
e−ν E 2
2
.
(25)
With the use of Eq. (18) we can express the tangential velocity as
2
vtg
L2
i.
h T
=
2
2
c
L2T + r2 1 − ǫΦ2 (U 4 )
(26)
In order to completely solve the problem of the stable circular motion of the test particles in extra-dimensional
models we need an equation determining U 4 . This can be taken as the fifth component of the geodesic equation, and
is given by
1 ∂gαβ α β
d
ǫΦ2 U 4 =
U U .
dS
2 ∂ξ
(27)
Taking into account that U 1 ≡ 0, and that the g22 and g33 metric tensor components do not depend on ξ, we have
1 ∂g00 0 0
1
∂
d
ǫΦ2 U 4 =
U U = − E 2 e−ν ,
dS
2 ∂ξ
2 ∂ξ
where we have again used the conservation law for the energy.
(28)
7
IV.
DARK MATTER AS AN EXTRA-DIMENSIONAL EFFECT
The galactic rotation curves provide the most direct method of analyzing the gravitational field inside a spiral
galaxy. The rotation curves have been determined for a great number of spiral galaxies. They are obtained by
measuring the frequency shifts z of the light emitted from stars and from the 21-cm radiation emission from the
neutral gas clouds. Usually the astronomers report the resulting z in terms of a velocity field vtg . The observations
show that at distances large enough from the galactic center
vtg ≈ 200 − 300 km/s = constant.
(29)
This behavior has been observed for a large number of galaxies [1].
For a test particle traveling on a stable circular orbit in a multi-dimensional space-time, the tangential velocity
is given by Eq. (26). In a purely four-dimensional space-time, Φ2 ≡ 0, and the tangential velocity is given by
2
vtg
/c2 = L2T / L2T + r2 . In the limit of large r, r → ∞, and taking into account that L2T is a finite (conserved)
quantity, we obtain limr→∞ vtg = 0. However, the situation is quite different in the multi-dimensional models. If the
condition
1 − ǫΦ2 U 4
2
=
C 2 (ξ) L2T
,
r2
(30)
holds true for large r, where C 2 ≥ 0 is an arbitrary function of the fifth coordinate, then the tangential velocity of a
test particle in circular stable motion around the galactic center is given by
2
vtg
1
=
.
2
c
1 + C 2 (ξ)
(31)
If C 2 (ξ) is a true (galaxy-dependent) constant, then the tangential velocity is an absolute constant, too. Therefore,
the constancy of the galactic rotation curves can be explained naturally in the multi-dimensional physical models,
without the necessity of introducing the ad hoc hypothesis of the dark matter. Of course, there are several other
choices in Eq. (26) which may lead to constant or slightly increasing rotation velocity curves.
Since the tangential velocity is a constant or fifth-dimension dependent quantity, one can solve Eq. (24) and
find the value of the metric tensor component exp (ν) in the constant rotational curves region. After some simple
transformations Eq. (24) can be written in the form
!
2
2
vtg
vtg
∂ h 2 4 2 i
1
1
′
ν =2 2 + 1− 2
ǫΦ U
,
(32)
2
c r
c
1 − ǫΦ2 (U 4 ) ∂r
giving
2
2
eν = D (ξ) r2vtg /c ǫΦ2 U 4
2
−1
2
−(1−vtg
/c2 )
,
(33)
where D (ξ) is an arbitrary integration function. In the four-dimensional limit Φ2 ≡ 0 we obtain the well-known result
2
2
exp (ν) = Dr2vtg /c , which has been extensively used to discuss the properties of dark matter [10, 11, 12, 13, 14, 15],
[36, 37, 38, 39]. Hence, the expression of the metric tensor component g00 in the constant rotation curves region can
be obtained in an exact form.
An important particular case corresponds to the situation in which eν is independent of ξ. Then, the geodesic
equation Eq. (28) can be immediately integrated to give
ǫΦ2 U 4 = B 2 = constant.
(34)
Together with Eq. (30), the above first integral of the equations of motion allows the determination of the functional
form of the metric tensor component Φ2 in the constant rotation curves region as
ǫΦ2 =
B2
.
1 − C 2 L2T /r2
(35)
In this case B, C and D are true constants, and they are independent of the fifth dimension ξ. This model
corresponds to a compactified Kaluza-Klein type theory, in which all the metric tensor components are independent
of the fifth coordinate.
8
The multi-dimensional effects are dominant in the vacuum at large distances from the galaxy. In the presence of
matter, that it, inside or at the vacuum boundary of the galaxy, these effects are very small, as compared to the
gravitational effect of the normal four-dimensional matter. Hence we may assume that at the boundary of a galaxy
with baryonic mass MB and radius RB the four-dimensional geometry is approximately the Schwarzschild geometry,
and therefore
eν |r=RB ≈ 1 −
2GMB
.
c2 RB
(36)
This matching condition (approximately) determines the constant D in Eq. (33).
V.
THE ACCELERATION LAW IN EFDOD
We start by assuming that the motion of a test particle in a 4D space-time with metric gµν is given by
duµ
D(4) uµ
≡
+ Γµαβ uα uβ = f µ ,
ds
ds
(37)
where uµ = dxµ /ds is the usual four-dimensional velocity of the particle and Γµαβ are the Christoffel symbols constructed by using the 4D metric. The presence of the extra force f µ makes the motion of the particle non-geodesic.
For f µ ≡ 0 we recover the geodesic equation of motion. All the usual gravitational effects, due to the presence of an
arbitrary mass distribution, are assumed to be contained in the term aµN = Γµαβ uα uβ . In three dimensions and in the
Newtonian limit, Eq. (37) can be formally represented as a three-vector equation of the form
~a = ~aN + ~af ,
(38)
where ~a is the total acceleration of the particle, ~aN is the gravitational acceleration and ~af is the acceleration (per
unit mass) due to the presence of the extra force. If ~af = 0, the equation of motion is the usual Newtonian one,
~a = ~aN , or, equivalently, ~a = −GM~r/r3 .
2
In the following we denote by v 2 = ~v · ~v = |~v | the magnitude of a vector ~v . Taking the square of Eq. (38) gives
~af · ~aN =
1 2
a − a2N − a2f ,
2
(39)
where · represents the three-dimensional scalar product. Eq. (39) can be interpreted as a general relation which gives
the unknown vector ~aN as a function of the total acceleration ~a, of the acceleration ~af due to the extra force, and of
the magnitudes a2 , a2N and a2f , respectively. From Eq. (39) one can express the vector ~aN , as one can easily check,
in the form
~aN =
~a
1 2
~ × ~af ,
a − a2N − a2f
+C
2
~af · ~a
(40)
~ is an arbitrary vector perpendicular to the vector ~af . In the following, for simplicity, we assume C
~ ≡ 0.
where C
The mathematical consistency of Eq. (40) requires ~af ·~a 6= 0, that is, the vectors ~af and ~a cannot be perpendiculars.
Generally, ~af · ~a = af a cos α, where α is the angle between ~af and ~a. Again, for simplicity, we take α = 0, that is, we
assume that the vectors ~af and ~a are parallels. For ~af = 0, Eq. (40) gives a2N = a2 , as required.
Therefore, we can represent the gravitational acceleration of a test particle in the presence of an extra force as
~aN =
~a
1 2
a − a2N − a2f
.
2
af a
In the limit of very small gravitational accelerations aN << a, we obtain the relation
!
a2f
1
1
~a.
~aN ≈ a 1 − 2
2
a
af
(41)
(42)
By denoting (1/2af ) 1 − a2f /a2 = 1/aE , Eq. (42) immediately gives
~aN ≈
a
~a,
aE
(43)
9
which is similar to the equation proposed phenomenologically
by Milgrom [22, 23, 24]. From this equation we obtain
√
√
2
/r, where vtg is the rotational velocity of the test
a ≈ aE aN , and since aN = GM/r2 , we have a ≈ aE GM /r = vtg
√
2
2
4
particle. Therefore, it follows that vtg → v∞ = aE GM , giving the Tully-Fisher relation v∞
= aE GM ∼ L, where L
is the luminosity, assumed to be proportional to the mass [22, 23, 24].
However, in the framework of EFDOD, aE is not generally a universal constant, but it may be a position, acceleration
or galaxy characteristics dependent quantity.
Generally, from the given definition
of aE , we can formally represent the extra acceleration as a function of a and aE
p
as af /aE = −a2 /a2E ± (a/aE ) 1 + a2 /a2E . Then, by means of some simple calculations, Eq. (41) can be represented
as
a
aN 2
a
+ 1 ~a,
(44)
F
~aN =
aE
aE
a
where
F
a
aE
1
=
2
a
aE
−1
a
∓
aE
s
a2
1+ 2
aE
!−1
.
(45)
4
With the use of the equation a2 = v∞
/r2 = aE GM/r2 we obtain the following expression for aE :
aE ≈
a2f r2
+ 2af .
GM
(46)
If af ∼ GM a0 /r, where a0 is a constant, then in the large r limit, when af → 0, aE ≈ a20 is a constant, whose
numerical value is determined by the physical properties of the extra force. If the extra-force is universal in its nature,
than aE is a universal constant.
Therefore the MOND paradigm (which also postulates the existence of a universal constant a0 ) is equivalent, and
can be derived, from the assumption of the non-geodesic motion of the test particles around the galactic centers under
the action of a specific force.
VI.
DISCUSSIONS AND FINAL REMARKS
In the present paper we have shown that the dynamics of the test particles in circular orbits around galaxies can
be attributed to the presence of an extra force, generated by the presence of the extra dimensions, which modifies the
standard (Newtonian or general relativistic) motion, by giving a supplementary contribution to the acceleration. The
total acceleration has a general form, which is formally identical to the one proposed on a phenomenological-empirical
basis in MOND. In an equivalent formulation, MOND is the result of the non-geodesic motion of particles under the
influence of a specific force. On the other hand, depending on the physical nature of the extra force, more general
physical models than MOND can be obtained. We have considered several possibilities for the extra force. The extra
force may be generated, for example, by a scalar field coupled with matter. However, in such a scenario, some extra
(dark?) matter is required, and, in order to obtain a constant a0 , a specific density profile for the matter is necessary.
On the other hand, for the scalar field to influence the galactic rotational velocity, it has to be massless. But generally
massless scalar fields could not exist in nature.
However, one could relate extra dimensions to dark matter through the dilatonic dark matter [6, 59]. As an essential
part of higher dimensional metric the dilaton plays a crucial role to determine the higher dimensional geodesic. But
this higher dimensional geodesic equation, expressed in the lower dimension, contains a non-geodesic force, created
by the dilaton, which requires a modification of general relativity. Generally, in 4D the motion can be described as
taking place under the effect of a tensor (metric) field and of a vector and scalar field, respectively. Interestingly
enough, the relativistic version of MOND [26] requires exactly such a modification of general relativity, but with the
extra-fields introduced by hand. However, for this model to be true, it is necessary that the extra- force from the extra
dimension has to be long ranged (in a galactic scale). On the other hand the fifth force obtained in the framework of
the compactified Kaluza-Klein theories cannot be long ranged [7]. Some possibilities of overcoming these difficulties
may be by assuming that the extra-dimensions are large, as is the case in the brane-world models [34, 61, 62].
In order to explore in more detail the connections between EFDOD, MOND and dark matter, some explicit physical
models are necessary to be built. This will be done in some forthcoming papers.
10
Acknowledgments
We would like to thank to the anonymous referee for comments and suggestions that helped us to significantly
improve the manuscript. The work of T. H. was supported by the GRF grant No. 7018/08P of the government of
the Hong Kong SAR.
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
J. Binney and S. Tremaine, Galactic dynamics, Princeton, Princeton University Press, (1987).
M. Persic, P. Salucci and F. Stel, Month. Not. R. Astr. Soc. 281, 27 (1996).
A. Borriello and P. Salucci, Month. Not. R. Astr. Soc. 323, 285 (2001).
J. M. Overduin and P. S. Wesson, Phys. Repts. 402, 267 (2004).
I. F. M. Albuquerque and L. Baudis, Phys. Rev. Lett. 90, 221301 (2003).
Y. M. Cho, Phys. Rev. D41, 2462 (1990).
Y. M. Cho and Y. Y. Keum, Class. Quantum Grav. 15, 907 (1998).
Y. M. Cho and J. H. Kim, Phys. Rev. D79, 023504 (2009).
Y. M. Cho, J. H. Kim, S. W. Kim and J.H. Yoon, arXiv:0708.2590 (2007).
U. Nucamendi, M. Salgado and D. Sudarsky, Phys. Rev. Lett. 84, 3037 (2000).
T. Matos and F. S. Guzman, Class. Quant. Grav. 18, 5055 (2001).
E. W. Mielke and F. E. Schunk, Phys. Rev. D66, 023503 (2002).
B. Fuchs and E. W. Mielke, Month. Not. R. Acad. Sci. 350, 707 (2004).
X. Hernandez, T. Matos, R. A. Sussman and Y. Verbin, Phys. Rev. D70, 043537 (2004).
D. Giannios, Phys. Rev. D71 103511, (2005).
A. Bernal and F. Siddhartha Guzman, Phys. Rev. D74, 063504 (2006).
F. Dahia and C. Romero, Phys. Lett. B651, 232 (2007).
C. H. Coimbra-Araujo and P. S. Letelier, Phys. Rev. D76, 043522 (2007).
C. H. Coimbra-Araujo and P. S. Letelier, Class. Quantum Grav. 25, 015001 (2008).
R. H. Sanders, Astron. Astrophys. 136, L21 (1984).
R. H. Sanders, Astron. Astrophys. 154, 135 (1986).
M. Milgrom, Astrophys. J. 270, 365 (1983).
J. Bekenstein and M. Milgrom, Astrophys. J. 286, 7 (1984).
M. Milgrom, Astrophys. J. 599, L25, (2003).
M. Milgrom, New Astron. Rev. 46, 741, (2002).
J. D. Bekenstein, Phys. Rev. D70, 083509 (2004).
R. H. Sanders, Month. Not. R. Astr. Soc. 363, 459 (2005).
P. D. Mannheim, Astrophys. J. 419, 150 (1993).
P. D. Mannheim, Astrophys. J. 479, 659 (1997).
J. W. Moffat and I. Y. Sokolov, Phys. Lett. B378, 59 (1996).
J. R. Brownstein and J. W. Moffat, Astrophys. J. 636, 721 (2006).
J. R. Brownstein and J. W. Moffat, Mon. Not. R. Astron. Soc. 367, 527 (2006).
M. D. Roberts, Gen. Rel. Grav. 36, 2423, (2004).
L. Randall and R. Sundrum, Phys. Rev. Lett 83, 3370 (1999).
L. Randall and R. Sundrum, Phys. Rev. Lett 83, 4690 (1999).
M. K. Mak and T. Harko, Phys. Rev. D70, 024010 (2004).
T. Harko and M. K. Mak, Annals Phys. 319, 471 (2005).
T. Harko and K. S. Cheng, Astrophys. J. 636, 8 (2005).
C. G. Boehmer and T. Harko, Class. Quantum Grav. 24, 3191 (2007).
T. Harko and K. S. Cheng, Phys. Rev. D76, 044013 (2007).
S. Pal, S. Bharadwaj and S. Kar, Phys. Lett. B609, 194 (2005).
O. Bertolami, C. G. Boehmer, T. Harko and F. S. N. Lobo, Phys. Rev. D75, 104016 (2007).
C. G. Boehmer, T. Harko and F. S. N. Lobo, Astropart. Phys. 29, 386 (2008).
C. G. Boehmer, T. Harko and F. S. N. Lobo, JCAP 03, 024 (2008).
C. S. J. Pun, Z. Kovacs and T. Harko, Phys. Rev. D78, 024043 (2008).
S. J. Sin, Phys. Rev. D50, 3650 (1994).
S. U. Ji and S. J. Sin, Phys. Rev. D50, 3655 (1994).
J.-W. Lee and I.-G. Koh, Phys. Rev. D53, 2236 (1996).
C. G. Boehmer and T. Harko, JCAP 06, 025 (2007).
J.-W. Lee, arXiv:0801.1442 (2008).
J.-W. Lee, arXiv:0805.2877 (2008).
J.-W. Lee, S. Lim and D. Choi, arXiv:0805.3827 (2008).
K. Lake, arXiv:gr-qc/0607057 (2006).
C. G. Boehmer and T. Harko, Mon. Not. Roy. Astron. Soc. 379, 393 (2007).
11
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
Y. M. Cho, J. Math. Phys. 16, 2029 (1975).
Y. M. Cho, Phys. Lett. B186, 38 (1987).
Y. M. Cho, Phys. Rev. D35, 2628 (1987).
Y. M. Cho and D. H. Park, Gen. Rel. Grav 23, 741 (1991).
Y. M. Cho, Phys. Rev. Lett. 68, 3133 (1992).
D. J. Kapner, T. S. Cook, E. G. Adelberger, J. H. Gundlach, B. R. Heckel, C. D. Hoyle, and H. E. Swanson, Phys. Rev.
Lett. 98, 021101 (2007).
J. Ponce de Leon, Phys. Lett. B523, 311 (2001).
J. Ponce de Leon, Gen. Rel. Grav. 36 1333 (2004); Gen. Rel. Grav. 36, 1335 (2004).
L. D. Landau and E. M. Lifshitz, The Classical Theory of Fields, Oxford, Pergamon Press (1975).
S. L. Bazanski, J. Math. Phys. 30, 1018 (1989).
M. E. Kahil, J. Math. Phys. 47, 052501 (2006).
J. P. Mbelek, Astron. Astrophys. 424, 761 (2004).
P. S. Wesson and J. Ponce de Leon, Astron. Astrophys. 294, 1 (1995).
D. Youm, Phys. Rev. D62, 084002 (2000).